ML20199B225

Proposed Resolution Approach NRC Bulletin 96-03 Hope Creek Generating Station Technical Evaluation Report

ML20199B225
Person / Time
Site:	Hope Creek
Issue date:	09/10/1997
From:	Rao D SCIENCE & ENGINEERING ASSOCIATES, INC.
To:	NRC
Shared Package
ML20199B172	List: ML20199B167 ML20199B202 ML20199B225
References
IEB-96-003, IEB-96-3, SEA-97-3703-A:2, NUDOCS 9711180250
Download: ML20199B225 (14)

Text

. - -. _ . . . - - . . . . . . . _ . . -. . _ . _. . - _ .

~

. .. ATTACHMENT SEA 97-3703-A:2 Proposed Resolution Approach-NRC Bulletin 96-03 Hope Creek Generating Station

-Technical Evaluation Report September 10,1997 l

i Prepared by Dasari V. Rao

Science and Engineering Associates,Inc.

6100 Uptown Boulevard NE, Suite 700 Albuquerque, NM 87110 l

l l

l l Prepared for U.S. Nuclear Regulatory Commission Washington, DC 20055 anc.

l

' 354 #

D A 05 G 8'DR $

~ '

TECHNICAL EVALUATION OF

• ' FPROPOSED RESOwTION APPROACH - NRC 96-03 POTENTIAL FOR PLUGGING Or ECCS SUCTION STRAINERS BY DEBRIS, HOrt CREEK GENERATING STATION" 1.0 INTRODtJCTION ,

In response to Bulletin 96-03, Public Service Electric and Gas Company (PSE&G) has submitted

' Proposed Resolution Approach - NRC Bulletin 96-03; Potential Plugging of ECCS Suction Straine s by Debris Hope Creek Generating Station' for US Nuclear Regulatory Commission (NRC) review [Ref.1). Science and Engineering Associates, Inc. (SEA) is tasked by NRC to review the submittal.

De resolution option is based on installation of passive large capacity suction strainers, designed and manufactured by the Performance Contracting, Inc. (PCI). The submittal estimated debris loading on the strainer following a postulated ' worst case' break using methodology prosided by the Boiling Water Reactors Owners Group (BWROG) in the Utility Resolution Guidance (URG) document (Ref. 2]. Estimates for quantities of fibrous debris transported to the strainer were evaluated on a plan'.-specific basis, but usirig guidance provided in the URG On the other hand, generic estimates provided in the URG were used for sludge, paint-chips, dust and dirt, and rust from unpainted ,tructures. No allowance was made for unqualified or indeterminate coatings or for transportable foreign material such as clothing or plastic sheet.

He plant intends to design strainers subject to single failure analysis, which resulted in the availability of one (3) Residual Heat Removal (RHR) and two (2) Core Spray (CS) pumps for injection into the core. The large capacity strainers to be installed on the RHR and CS pumps will be sufficiently large to accommodate the debris and result in head lo:ses less than the available NPSH Margin of 1.511 water for the RHR pumps and 1.2 ft water for the CS pumps.

The corresponding flow rates are 10500 GPM and 4015 GPM for RHR and CS, respectively.

The flow rates are consistent with ECCS design criteria and accident analysis sections of the Utility Final Safety Analysis Reports (UFSAR) [Ref. 3]. Estimates of Available NPSH are based upon 'no credit for containment overpressure' and are consistent with Regulatory Guide (RG) 1.82, Rev. 2 which cites RG 1.1.

The utility did not provide actual sizes of the strainers to be installed on the RHR and CS pumps.

Instead, it provided a methodology proposed kc its contractor for sizing the strainers. It is the intent of the utility to finalize the strainer desig. upon receiving NRC approval of the ' Proposed Criteria for ECCS Strainer Design' and, if necessary, to forward the design configuration to the NRC staff when the design is completed.

SEA performed a preliminary review of the licensee submittal and forwarded a Request for Additional Information (RAI) to the utility. In a conference call with the utility, SEA prosided detailed explanations of the RAI. The utility response to the RAI is contained in their letter (dated: July 18,1997), ' Response to Request for Additional Information Regarding Proposed Resolution Approach' [Ref. 4]. The utility response provided clarifications regarding (a) debris generation estimates, (b) strainer sizes and (c) rationale for not including unqualified paint coatings in the debris estimates. However, the utility did not provide the actual methodology to be followed by its contractor to qualify the strainers for Hope Creek, although the letter contained a brief description of the proposed approach ne focus of SEA's review of the submittal (Ref.1) and the supporting documentation [Ref. 4],

is:

Scienn &

k 1 SEA-3704-010-A:2, Rev. I

., TECHNICAL EVALUATION OF j

PRorostD RLsottfrioN APPROACH -NRC 96-05 POTENTIAL FOR PLUGGING or ECCS SUcrION '

STRAINEks BY DEBRIS, HorE CRFEK GENER ATING STATION" L To assess the adequacy / accuracy of the utility estimated debris loading to be used in sizing the strainer. -In particular, SEA focused on determining if the breaks were selected in accordance with Regulatory Guide 1.82, and if the methodology used by the utility has been

, previously approved by NRC to provide reasonable estimates for debris generation and transport. Finally, has the plant made efforts to ensure that the head loss predicted for the

' worst-case' break is, in fact, higher than that corresponding to all ether possible breaks.

2. To conduct order of magnitude calculations regarding the adequacy of the proposed strainer designs to handle debris loading derived from the step above. These calculations were performed to give an approximate estimate for head loss that can later be refined. These calculations will be finalized once the utility finalizes its design and forwards the final submittal.

SEA did not devote any effort to examining the validity of the licensee's assumption related to hydrodynamic loads. The licensee has not included any such calculations for review, although a brief description of the methodology is provided in Rei. 4.

To achieve these goals SEA performed a series'of calculations. Calculations related to debris generation are summarized in Appendix A, whereas calculations related to the stacked disc strainer assessment are documented in Appendix-B. The following sections summarize SEA's findings.

2.0 CONTRACTOR FINDINGS 2.1 Selection of the Break:

The utility explicitly stated that the break location wss selected based on criterion of High Energy Line Breaks (HELB), which refers to locating the break close to the high stress points. A total of 120 breaks that formed the licensing basis of Hope Creek were evaluated. The rest of the breaks were assumed to be 'non-credible' and were screened out. This selection is not con::istent with NRC position stated in RG 1.82, Rev.1. The utility has stated that it would revisit this issue during their seventh refueling outage (RF07). At that time the utility will evaluate if any of the

'non-credible breaks' would generate debris higher than the 120 breaks analyzed so far.

The utility did not analyze any Medium size breaks as required by Regulatory Position 2.3.1.5 of RG 1.82, Rev. 2, which requires the licensee to identify 'the medium and large breaks with largest potential particulate to insulation ratio by weight.' 'Ihe stacked disc strainer is not expected to have the ' thin-bed effect' and hence Regulatory Position 2.3.1.5 of Regulatory Guide 1.82, Rev. 2 does not apply.

SEA concludes that selection of the break is inconsistent with the guidance provided in RG 1.82, Rev. 2. Installation of strainers sized based upon the present set of breaks would not meet 10 CFR 50.46 guidance, iflater analyses were to reveal that other breaks (described as non-credible in this submittal) would generate a higher amount of debris. Furthermore, such a design would j also pose a financial risk, since it may have to be redesigned.

2.2 Debris Generation:

, The utility used a modified version of Method 3 of the URG to estimate the zone ofinfluence i (ZOI) and the quantity of debris generated by the jets. This revised approach assumes that all l ,5amce &

^88*dak8 A 2 SEA-3704-010-A:2, Rev.1 l

, TECHNICAL EvAtt!ATioN or

' # PROPOSED RESOLUTION APPROACH- NRC 9643 POTENTIAL roR PLOGGING c7 ECCS St crioN STRAIN ERs my DEaRis, IloPE CREEK GENERATING STATION" insulation contamed m a dynamic pressure isobar of 5.2 psi would be damaged instead of the 10 psi value recommended in the URG. De rei.ulting Z01 used by the licensee for estimating fibrous debris generation is a sphere whose radius is approximately 12 x Diameter of the Recirculation Line (12D). His ZOI is about 4 times larger than the ZOI used in NUREG/CR.

6224 [Ref. 5) in volume. The licensee estimated that about 1402 3fl of insulation would be damaged of a total of 4500 ft3(see Appendix A). De licensee screened out RMI used on the  :

reactor vessel because their analysis concluded that no paths existed for its transport.

SEA concludes that the use of Method 3 of URO is acceptable and that licensee has appropriately estimated the quantity of debris generated. De estimated quantities of fibrous debris appear to provide a reasonable upper bound (provided, of course, that one of the unanalyzed breaks does not result in more debris).

2.3 Debris Trarfporti ne licensee has determined that 100% of the generated debris for the ' worst case' break was contained above the lowest drywell grating; which appears reasonable considering that the

' worst case' break is located in the congested region of the drywell about 25 ft above the lowest grating. De licensee estimated the total quantity of debris reaching the strainer using the URG combined generation / transport factor of 0.28. nis value was recommended by URG and includes 100% transport of small debris, which amounted to 22% of the total, and 6.25%

transport of the remaining large debris due to erosion. A transport factor of 1.0 was used for all other debris ger.erated in the drywell.

A transport factor of 1.0 was used for suppression pool transport.

SEA believes that these transport fractions are reasonable and that estimated quantities appear to provide a reasonable upper bour.d for Ge quantity of debris loading on the strainer, provided:

(a) ne lowest floor grating does not contain large discontinuities that allow advection1 oflarger pieces ofinsulation torn out from the canvass covers by the jets, or (a) The ECCS does not run unthrottled in excess of 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> folloving a LOCA, et which point erosion would be larger than that assumed in deriving the 28% number recommended in the l URG.

l l

' Air bome transport of debris during blowdom sdence 4 -

3 SEA 3704-010- A:2, Rev. )

l l

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

, TECHNICAL EVAltfATioN or

• Paorosto Ru.ottfrioN APrnoACH- NC 96-03 POTENTIAL son Pttica:No or ECCS SterloN STRAlvens av DEskis, Hort Cnttx GENERATING STATION" i

2.4 Dehris Leading on the Strainer: ,

he following table summarizes the debris loading assumed and its basis for each type of debris. l Type of Deb'rls Quantity Remarks Fibrous Debris (Nukon) 393 ft3 Break selection not consistent with RG 1.82, Rev.2. Otherwise reasonable.

RMI- Stainless Steel 0 There may be some small quantitles of RMI.

Sludge 300 lb. Generation of150 lb. for two years 2, Dust and Dirt 150 lb. URG Number. Also used in NUREG/CR 6224 Rust from Unpainted Surface 50 lb. URG Number.

Paint inorganic Ztne 0 lb.

10Z with Epoxy topcoat 85 lb. Paint located outside the conical let exp.msion Epoxy 0 lb. area was excluded, whether it is qualified or Unqualified Paint Coating 0 lb. unqualified.

Other (Trans. For. Mat.) 0 fl2 These numbers and their rationale for usage appear reasonable, except:

1. He estimate for paint chips is based on qualified paints only No allowance is made for generation from unqualified or indeterminate paint chips. Could be a major issue if future NRC efforts were to show that unqualified paints fail by processes other than blistering.

2 ne licensee did not include other transportable foreign materials such as clothing and plastic bags, its impact is expected to be minimal.

25 STRAINER DESIGN CONSIDERATIONS ECCS Operating Parameters The utility provided the following parameters for ECCS pumps:

System Flow Rate NPSif..tsitable NPSli equired R T po.: Comments 8 Pumps (GPM) (ft water) (ft water) op RilR (4) 10500 6.0 4.5 212 Long term CS (2) 4015 11.2 10 i 212 Long term De utility took credit for decreased ECCS flow after core reflood (=10 minutes post. LOCA).

During long term operation the suppression pool temperature is assumed to be 212 0F.

He NPSilavailable was estimated in accordance with RG 1.1 taking no-credit for containment pressurization.

• There is no cornmitment on the part of the utility to clean the pool every two years. Instead, the utility stated in response to RAl#4 that the Torus Water Cleanup system operational at ilope Creek would likely keep the sludge volume to values lower than assumed. If suppression pool surveys were to indicate quantities higher than the assumeJ 300 lb., then the utility will clean the suppression pool This appears to I be a reasonable approach, muur

A58a d*8e* A - 4 SEA-3704-010-A

2, Rev. I

TECHP ICAL EVALLIAT10% OF l cPaorosEn RL50LtmoN APPROACH- NRC 403 POTENTI AL ion PLticr,asa or ECCS SticTioN l STRAINLRS BY DEBuls, Hort Currx GENERArtso STATsoN i De values used for flow rate and suppression pool temperaterc akicar reasonable and are consistent with UFSAR numbers.

Single Failure Analysis The single failure analysis by the plant assumes the availability of three RHR pumps and one CS pump following a LOCA and simultaneous failure of Channel A DC Source.

He assumed single design failure appears reasonable.

Strainer Design The utility solution is based on replacing existing ' stacked disk' strainers with large capacity passive stacked disk strainer modules. Module details are:

Length =8A Diame'~ = 45 inches Plate Area = 220 fl2 lloleSize = 1/8 inch Each LPCI suction will have three (3) modules and each LPCS suction will have one (1) module.

Corresponding loadings on each module are:

Quantity IParameter RHR Strainer CS Strainer Module Module Nukon in ft'(ib. @ 2.4 lb/ft') 34.8 (83.52) 39.9 (95.76)

RMI (ft') 0.0 0.0 Studge (ib.) 2S.6 30.5 Dust and Dirt (ib.) 13.3 15.2 Rust from unpainted (ib.) 4.33 5.1 Paint (ib.)

Inorganic IOZ with Epoxy topcoat 7.6 8.6 Epoxy Unqualified Paint Coats (Ib,)

Other (ft )

Particulate fiber ma:.s ratio 0.62 0.62 Flow (GPM) 3500 4015 Surface Area (ft') 220 220 Diameter (inch) 45 45 Active Length (ft) 8 8 The CS strainer module draws more flow per square foot and also has more debris loading. Thus the CS suction line poses the limiting situation.

Head Loss Correlation and the Strainer Sising Criteria:

Ihe plant intends to use vendor provided head loss correlation and sizing methodology described briefly in Ref. 4. This method relies on applying the NUREG/CR-6224 correlation in a modified form suitable for stacked disk geometry. To maintain conservatism, the utility proposes to use a filtration efficiency of 1.0.

4 he. 5 SEA-3704-010 A:2, Rev.1

TECHNICAL EvAlltATION OF "PaorosEn REsott' Tion APPRO ACH - NRC 96-03 POTLNTI AL FOR plt 1GGING OF ECCS St!Crl0N SinAlstRs av DraRis. Hort CartK GEstnATING STATios"  ;

SEA has not reviewed the vendor methulole gy to render detailed comments. Selected calculations were performed to obtain approximate estimates of head loss, which suggest that the proposed strainers may have sufficient area to handle the debris loads anticipated. These calculations are summarized in Appendix B.

3.0 DEriCIENCIES AND RECOMMENDATIONS Two possible deficiencies with potential to have serious impacts on the overall resolution were identified:

• The proposed modules are 8 A long and raise serious concems regarding hydrodynamic loads. This issue should be addressed in more detail than has thus far been provided to the NRC for review.

• The estimate of paint chips is based on the rationale that failure of unqualified paints located '

outside the conicaljet expansion region is by blistering processes that occur ove the long-term aner blowdown and thus would not likely to be transported to the strainer surt:.

It is recommended that the utility examine its plant layout, plant configuration and EOPs to ensure that:

1. 14 west grating is continuous to prevent 100% oflarge debris from entering the vents and reaching the suppression pool. If the grating is not continuous, then sare should be taken to increase the estimate of debris transported to suppression pool.

2. Operators are allowed to, and are trained to throttle ECCS within 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> aner a recirculation line break. If not, a higher fraction of the large debris captured on the gratings would be eroded, resulting in transport of greater than 28% of the insulation contained in the 201.

3. Breaks r.ot analyzed as pan of the present analysis do not generate more debris than the 1400 n3 used in the present design.

4.0 CONCLUsloNs The utility selection of breaks for analysis is not consistent with RG 1.82, Rev. 2 guidance and may not be in compliance with 10 CFR 50.46. Otherwise, the utility employed reasonable methods to estimate the quantity ofinsulation debris generated in the drywell and transponed to the ECCS =ucCon strainer. The utility rationale for non insulation debris also appears reasonable, liowever, the utility did not provide the actual methodology to be followed to estimate head loss coss the strainer. They were encouraged to submit the finalized design of the strainers and the methodology used to estimate head loss across them for review and approval.

5.0 REFERENCES

1. Public Senice Electric and Gas Company,' Proposed Resolution Approach-NRC Bulletin 96-03; Poter.tial Plugging of ECCS Suction Strainers by Debris llope Creek Generating Station,' Docket No. 50 354, May 20,1997,

2. BWROG, ' Utility Resolution Guidance for ECCS Suction Strainer Blockage,' NEDO-32686, November 20,1996.

3. Updated Final Safety Analysis Report,llope Creek Generating Station.

4. Public Senice Electric and Gas Company, ' Response to Request for AdditionalInformation Regarding Proposed Resolution Approach- NRC Bulletin 96-03; Potential Plugging of ECCS Suction Strainers by Debris llope Creek Generating Station,' Docket No. 50 354, July 18,1997

5. NUREG/CR 6224.

sdom

• • • d*W. 6 SEA 3704-010-A:2, Rev.1

, TECHNICAL EVALLtATION Or

• PaorosED Rr_sotteloN ArrmoArH - NRC 403 POTEN11 AL FOR PLl1GGING Or ECCS SterlON STRAINtns av DEDRis, IlorE CREEK GENERATING STATION" Appendix A Confirmatory Calculations Conducted by SEA to Assess Accuracy of Utility Estimates of Debris Leading on the Strainer A.1 Selection of tbe Break The utility explicitly stated that potential break locationswere selected based upon the licensing basis, which in turn is based on the Branch Technical Position MEB 31, " Postulated Rupture Locations in Fluid System Piping Inside and Outside Containment." Apparently a total of 120 potential break locations inside the containment formed the sub set for which debris generation calculations were performed. Bey included breaks in main steam, feed water and recirculation lines. De remaining potential break locations were excluded from the analysis, as they were considered non-credible, [Ref.: See Response to Question #1 of Request for Additional Information).

During our discussions, the utility stated that this sub set of breaks would most likely contain btcaks with largest amounts ofinsulation contained in the zone of influence. Dat is, this set will most likely contain the ' worst case break' which bounds debris generation estimates. The utility also stated that remaining 'non credible' breaks will be evaluated if they would produce debris volumes larger than the break selected in the submittal for analysis [Ref.: See Response to Question #1 of Request for Additional Information).

Although it was not explicitly stated, it is likely that the break selected for analysis is located in the mid region of the containment with the largest number of potential piping targets. Likely targets: a) Recirculation Line in which break occurred, b) Recirculation Manifold, c)

Recirculation Risers, d) Feed Water Lines, e) Main Steam Lines and f) ECCS Injection Line.

[Ref. Piping Drawings, UFSAR).

e It is SEAS opinion that the utility did not include all thepotential break locations to reasonably ensure that selected F cak would bound the volume ofdebris generatedfrom all other break.r. The utility she. ' include other break locations as required by Regulatory Guide 1,82, hev. 2, before)1nt :ing the strainer design.

• Proceeding as proposed may c #esent afinancial risk, ofone ofthe unanal):ed breaks result in debris loading higher than the break anal):cd (i.e., theplant may beforced to redesign strainersfor higher debris loads or redo the analysis).

A.2 Debr's Generation Estimates:

Utility used Method 3 ofthe URG to estimate zone ofinfluence for each postulated break with the following assumptions:

1. All breaks are unrestrained and fully offset both axially and radially.

2. All breaks result in double jets resulting in spherical zone ofinfluence

-, anc- 7 SEA-3704-010-A:2, Rev.1

~

\

, 'I rcusicAL EVALSATsoN or l "Paorosta REsot.trrioN ArraoAcH - NRC 96-03 POTENTIAL FOR PLL'GGING or ECCS SticrioN STRAINths sY Dthnis, Hort CRttx GENERATING STATION" The following calculations were performed as a ' sanity check' of utihty estimates:

Insu.'ation Quantitles in the Containment:

Type Volume Largest Pipe 14eation (ft3) OD (in)

Nukon 4,500 28 + 7 Recirculation line.

'H! ""' 'r:: *tr' Inside the biological shield. Screened out Stainless Steel RMI ss located behind the shield wall. Contribution of this insulation was neglected because paths available for transport are limited. Appears reasonable, but should have shown that transport and accumulation of small amounts of RMI would riot cause any noticeable increase in head loss.

Zcme ofinfluence (201) 2.,timates:

The utility performed ZOI estimates for NUKON are different from the guidance provided in the URG.1he difference pertains to scaling BWROG sir jet tests to DWR plants. The URG Table 1 Note 2 recommends that no-pipe diameter correction is required for NUKON insulation and that failure pressure of 10 psig is applicable to all pipes sizes. NRC is reviewing this issue. The utility performed calculations assuming that failure pressure is a function of pipe diameter, which resulted in lowering the destruction pressure to 5.2 psig (instead of 10 psig). The differences are shown quantitatively in the following tables, in which entries corresponding to each criterion are explicitly marked.

Analysis Type Pdest3 PIPe Stre4 Pdest AS Reelrc. 6 Final A (psi) Correction (psig) Correction llope Creek Calculation 10 0.52 5.2 6750 1.0 6750 URG Recommended 10 1.00 10. 4708 1.0 4708 9M! c+'m ct '

Equholent Sphere Parameters:

The equivalent sphere diameters for the zone ofinfluence are given as follows:

Analysis Volume of the ZOI Equivalent Radius Type D' (ft')  % of Containment (D) (ft) llope Creek Cales. 6750 (85x10') =40% =12D =28 URO Recommended 4708 (60x10') =30% = 10.5D =24.5 Comment:

1. The zone ofinfluence used by the utility is much larger than that recommended by URG; sphere with 12D versus 10.5D.

3 Table 2 of Utility Resolution Guidance (Page #46)

• Note #3 to Table 2 in Utihty Resolution Guidance (Page #47)

Table #1, Radial Off-set > 3D/2 of the Utihty Resolution Guidance (Page #45)

• Note #5 to Table I in Utility Resolution Guidance (Page #46) mur Ases, k. 8 SEA-3704-010-A:2, Rev. I

)

, TECHNICAL EVAlt!ATioN OF j

~'

"Puorosta REsott! Tion ArrROACH - NRC 96-03 POTENTI AL FOR PLLIGGING Or ECCS St)CTION STRAINtns av DisnI% Hort CatEK GENERATING STATION" l

2. De zone ofinfluence appears to be able to bound debns generation. Note that for NUKON, NUREO/CR 6224 assigned a zone ofinfluence of about 7D compared to 12 D used by the utility, insulation Debris Generated (or Targeted):

Utility reponed that thejets originating from the analyzed break target a total of 1402 ft 3 of the insulation. SEA did not perform independent evaluation of the pipe segments contained in the spherical zone ofinfluence (12D in radius) because the plant drawings are not readily available.

Also llope Creek is substantially different from NUREG/CR-6224 reference plant and, therefore, the reference plant estimates can not be scaled to llope Creek.

It is SEA's opinion that the utility estimate of NUKON debris generated is based on usage of an acceptable zone ofinfluence. The ZOI method as applied is more conservative than the method recommended by the Utility Resolution Guidance nese calculations appear reasonable A.3 leestion of the Debris Generated Utility stated that no debris is generated below the lowest grating. his appears reasonable for recirculation breaks located in the mid region of the containment and is consistent with past submittals. Also note that the utility reported to have analyzed those breaks that resulted in debris generation below the lowest grating. Ilowever, such breaks apparently resulted in lower debris being transported to the suppression pool than the break analyzed. We need to look at full calculations to confirm this finding.

A.4 Debris Transport The utility assumes that 28% of the targeted fibrous insulation reaches the suppression pool as a result of blowdown and washdown. He basis for 28% is Section 3.2.3.2.5 of the URG, Table 5.

The NRC confirmatory research has shown that larger gantity of debris may transport if:

(b) Uc lowest floor grating contains large discontinuities that allow advection oflarger pieces of insulation tom out from the canvass covers by the jets, or (c) ne ECCS runs unthrottled in excess of 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> following a LOCA, at which point erosion was found to be larger than that assumed in deriving the 28% number recommended in the URG.

Licensee should ensure that ilope Creek plant layout and EOPs are consistent with their transport assumptions.

No settling was allowed in the Juppression pool. All debris was assumed to be transported to the strainer.

A.5 Debris Loading on the Strainer Fibrous debris reaching the strainers: 1402 3ft x 0.28 (transport factor) = 393 f1 3 (or 942 lb.).

ne utility assumptions related to paint-chips generation is as follows:

• Chips from qualified coatings were estimated as recommended in the URG by assuming that all the paint located in the conical region ofjet expnasion becomes air bome.

• Although a total of 275 lb. of unqualified paints are located in the containment degradation outside thejet region was assumed to be long term resulting in negligible transport.

mnr nepneedw h-W- 9 SEA-3704-010-At2, Rev.1

TECHNICAL EVALL'ATioN OF "Pmorosta REsottrrioN ArenoACH .'CC M-03 POTENTIAL FOR plt 1GGING or ECCS StictioN

'. STRAl%F.Rs BY Dennis Hort CntEK GENtnATING STATION"

'Ihe net loading on the strainers is as follows:

Type of Debris Quantity Rationale Fibrous Debris (Nukon) 393 ft3 Methodology arpears reasonable.

RMI- Stainless Steel 0 There may be small quantities of RMI.

Sludge 300 lb. Generation of 150 lb. for two years 7 Dust and Dirt 150 lb. URG Number. Also used in NUREG/CR 6224 Rust from Unpainted 50 lb. URG Number.

Paint inorganic Zinc 0 lb.

lOZ with Epoxy topcoat 85 lb. Paint located outside the conical jet region was Epoxy 0 lb. excluded, whether it is qualified or unqualified.

Unqualified Paint Coating 0 lb.

Other (Trans. For. Mat.) 0 fi2 The debris loading on the strainer appears reasonable, except for paint chips. In the case of paint chips, the utility assumed that failure of unqualified paints outside the conicaljet region will occur through the ' blistering' process over long term after a LOCA. Their transport would have to be by containment spray or break flow water. This allows for debris to settle down either in the drywell pool or in suppression pool. Design of strainers based on this rationale appears logical, provided of course present knowledge base holds good, llowever, there is an implicit risk involved in the design process, if future research were to find that unqualified paint chips could fail well outside the ZOI during blowdown itself, then the strainers have to be redesigned.

For example there appears to be ample evidence that some of the unqualified (or improperly applied qualified) paints are peeling-off during normal operation. A substantial fraction of the loosely attached pieces may become air bome and transported to suppression pool during blowdown. It is advisable to design strainers to handle higher debris leading, to avoid possible modifications in the future. This same concept holds tme for foreign materials.

' Here is no cornmitment on part of the utility to clean pool every two years. Instead, the utility stated in response to RAl#4 that Torus Water Cleanup system operational at Hope Creek will likely keep the sludge volume to values lower than assumed. If suppression pool surveys were to indicate quantities higher than the assumed 300 lb., then the utility will clean the suppression pool. His appears like a reasonable approach.

s,dence W k- 10 SEA 3',04-010-A:2, Rev.1

, TECHNICAL EVALL'ATioN or ePRoN>sEJ REsottrrioN APPROACH- NRC 96-03 POTENTIAL ron PLticctNG or ECCS SUcrioN STRAINERS BY DrBRIS, Hort CREEK Gr.NERATING STATION" Appendix H Confirmatory Calculations Conducted by SEA to Assess Licensee Assumptions Related to Head Loss B.1 Break Detalls The licensee has provided information for one break. The following table provides expected debris loading on each suction line. These values were estimated based upon the following assumptions:

• The debris loading on each suction line will be proportional to the ratio of flow rate through it to the total ECCS Dow rate.

• Single failure results in loss of a RIIR pump, with 3 RilR pumps and 2 CS pumps operational. Each pump has one suction line into the suppression pool.

Type of Debris Quantity Reaching each suction Total RHR#1 RHR#2 RHR#3 css 1 CS#2 Suction Suction Suction Suction Suction Nukon (ft') 393 104.4 104.4 104.4 39.9 39.9 RMI (ft') 0 0.0 0.0 0.0 0.0 0.0 Sludge (Ib.) 300 79.7 79.7 79.7 30.5 30.5 Dust and Dir1 (Ib.) 150 39.8 39.8 39.8 15.2 15.2 Rust from unpainted (Ib.) 50 13.3 13.3 13.3 5.1 5.1 Paint (ib.)

Inorganic 0 IOZ with Epoxy topcoat 85 22.6 22.6 22.6 8.6 8.6 Epoxy Unqual. Paint Coats (ib.) 0 2

Other (ft ) O Particulate fiber mass ratio 0.625 0.625 0.625 0.625 0.625 0.625 Flow (GPM) 39530 10500 10500 10500 4015 4015 Tahir headings:

• Totat: Total quantity of debris introduced into the pool. Reproduced from Table in Section A.S.

• RIIR Suetion: Quantity approaching the RIIR suction calculated as: Total *(RHR Flow of 10$00 GPMS'et ECCS Flow of39330 GPM) e CS Suetion: Quantity approaching the CS suction calculated as: Total *(CS Flow of 4013GPMS'et ECCS Flow of39330 GPM) k- 11 SEA 3704-010-A:2, Rev. I

e , ,

TECHNICAL EVAlt!ATioN or "Paorosto REsottrtioN ArraoACO- NRC 96-03 POTENTirt ton PLticcING or ECCS SttersoN STRAINERS BY DEBkis, Hort CRttK GENERATING STATION" 8.2 Strainer Details

'the utility stated that three strainer modules would be used for each RHR suction and one module for the core spray. He design details (i icluding debris loading for each strainer) are as follows:

QuantityIParameter RHR Strainer CS Strainer Module . Module 3

Nukon in ft (ib. @ 2.4 lb/ft') 34.8 (83.52) 39.9(95.76)

RMI (ft') 0.0 0.0 Sludge (ib.) 26.6 30.5 Dust and Dirt (ib.) 13.3 15.2 Rust from unpainted (Ib.) 4.33 5.1 Paint (Ib.)

Inorganic IOZ with Epoxy topcoat 7.6 8.6 Epoxy Unqualified Paint Coats (ib.)

Other (ft')

Particulate. fiber mass ratio 0.62 0.62 Flow (GPM) 3500 4015 Surface Area (ft') 220 220 Diameter (Inch) 45 45 Active Length (rt) 8 8 The CS strainer module draws more flow per square foot and also has more debris loading. Thus the CS suction line poses the limiting situation.

B.3 SEA Assessment of Utility IIcad Loss Correlation The utility stated that NUREG/CR 6224 correlation would be used to estimate head loss across the strainer, with two modifications to account for (a) the stacked-disk geometry and (b) cylindrical surface area. He modifications were supposedly validated using test data that was very close to llope Creek Da'.. In addition, the utility stated that they would use a filtration efficiency of 1.0 to maximize the head loss across the strainer. SEA has not, however, received the actual calculations. Herefore it can not assess accuracy of the methodology. Utility should consider providing these details such that SEA can assess head loss analyses used to finalize the strainer design.

B.4 Considerations for llend Loss Estimates Clean Strainer Head Loss and hallable NPSH Margin Assuming f 0% plugging, the utility estimates that approximately 1.2 ft 11 2 0 of NPSH margin is

. available for LPCS. His value probably will increase after the strainer is replaced because:

l

• The strainer surface area is increased by 4000%, resulting in an appr oach velocity (at the holes) of approximately 0.04 fVs instead of the presently assumed value of 1.59 ft/s (50%

plugging on old strainer).

e Ilead loss corresponding to such low approach velocities will be lower than corresponding to 1.59 fVs NW- 12 SEA-3704-010-A:2, Rev. I l

L

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

I .

TECHNICAL EvAtt!ArnoN or

,k' "PRorosEn REsot.ttiloN A*rROACH - NRC 96-03 PoTENTI AL roR plt 1GGING or ECCS SectioN STRAINERS aY DEaRIS, if OPE CREEK GENERATING STATION

Resulting savingE were estimated by SEA to be in the region of 0.25-0.75 ft water (more refinements are only possible if we SEA were to have actual utility head loss dua). Thir increases the availaole NPSil margin to 1.7 fi water instead of 1.2 fl. water. These savings are based on the assumptions that no changes are made apart from changing out the strainer.

BLUCKA GE Calculationsfor Fouled Strainer Head Loss The BLOCKAGE code was not developed for stacked disc strainers in mind, liowever, it can be applied to stacked disc geometry by a two etep process:

(a) Until Interstitial Gaps are Filled by Debris: Until that point the stacked disc strainer acts as plate strainer. For this case:

Astrainer = 220 fl2 Flow = 4015 GPM Veloci+" = 0.04 ft/s Thickness = 1.0 inch (0.0833 ft)8 Debris Volume = 18.3 ft3 The remaining parameters as per Table in Section B.2 above. For this case, BLOCKAGE predictions are about 0.25 fi water.

(b After the Interstitial Gaps are Filled and Occupiep The remaining volume (40 - 18.3 = 21.6 11 ) accumulates on the strainer circumscribed area which is estimated to be approximately 105 fl 2including the ends. For this case again BLOCKAGE run provided a head loss of 2 fi 2il 0.

Total IIcad Loss = licad Loss until interstitials are filled + llend loss during accumulation on circumscribed cylinder.

Total llead Loss = 0.542 + 0.243 = 0.3 fi water.

Calculational Uncertaintics The estimated head losses are approximately 0.8 fi water compared to avail sie margin of about 2 ft water 9. Clearly, the design allows for some margin of safety (= 1 ft water). Calculations suggest that:

• Lowering the fiber mass, while keeping the particulate mass constant does not cause a precipitous increase in the head loss, in fact, the increase in the head loss is minimal when the amount of fiber is halved from the present value of 40 ft3 .

• These strainers can not handle paint chips generated from unqualified paints (275 lb.) if all these chips were to be transponed to the strainer surface.

l -

l ' He thickness is half the gap width. Assumed 2. inch gap width.

' De same BLOCKAGE application methodology predicted Stacked disc head loss for run PS of URG aairly accurately. Derefore, the method as such can be seen to provide reasonable agreerrent.

5dence k- 13 SFA-3704-010-At2, Rev.1

- - . - .