ML13078A448
| ML13078A448 | |
| Person / Time | |
|---|---|
| Site: | Salem  |
| Issue date: | 04/05/2013 |
| From: | John Hughey Plant Licensing Branch 1 |
| To: | |
| Hughey J | |
| Shared Package | |
| ml13078a442 | List: |
| References | |
| GL-04-002 | |
| Download: ML13078A448 (22) | |
Text
{{#Wiki_filter:UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 AprilS, 2013 LICENSEE: PSEG Nuclear, LLC FACILITY: Salem Nuclear Generating Station, Units 1 and 2
SUBJECT:
SUMMARY
OF FEBRUARY 28, 2013, PUBLIC MEETING WITH PSEG RE: FINAL SUPPLEMENTAL RESPONSE TO GENERIC LETTER 2004-02 FOR SALEM NUCLEAR GENERATING STATION, UNITS 1 AND 2 On February 28,2013, a Category 1 public meeting was held between members of the U.S. Nuclear Regulatory Commission (NRC) staff and representatives of PSEG Nuclear, LLC. The purpose of the meeting was to discuss the PSEG response to Generic Letter (GL) 2004-02, "Potential Impact of Debris Blockage on Emergency Recirculation During Design Basis Accidents at Pressurized-Water Reactors," submitted on April 27, 2012, for Salem Nuclear Generating Station, Units 1 and 2.1 NRC staff and PSEG representatives discussed the draft responses provided by PSEG to the questions included in the public conference call summary issued on January 30,2013. 2 The draft responses are included as Enclosure 2 to this letter. 3 The NRC staff and PSEG representatives disclJssed specific questions and draft responses from Enclosure 2 as follows: Question #1 : The NRC staff is considering the validity of the assumption that no fibers greater than 500 microns will pass through the filter. What would the effect be if fibers greater than 500 microns were to pass through the filter? PSEG representatives noted that conservatism has been added to the analysis to account for this possibility, and they will provide additional information regarding this assumption. Question #3: The NRC staff asked if most of the fibers were deposited on the screen, or in the bottom of the flume. PSEG representatives responded that the material was wetted and was easy to pick up since it stuck together. No follow-up information to the NRC is required. Question #5: The NRC staff is considering issues regarding the effects of non-uniform flow. PSEG representatives acknowledged that there is an embedded assumption of uniform flow. However, PSEG also stated that non-uniform flow is a relatively short-lived event. No follow-up information to the NRC is required. 1 Agencywide Documents Access and Management System (ADAMS) Accession No. ML121290536. 2 ADAMS Accession No. ML1301 OA325. 3 ADAMS Accession No. ML13052A794.
-2 Question #7:
The NRC staff is considering the fact that only one type of tap water was tested to determine if there are effects of testing with tap water instead of buffered borated water. Different types of tap water have been shown to perform differently in head loss testing. PSEG will review available industry-sponsored testing. Question #11: The NRC staff asked about the accuracy of the scale used to weigh the screen. PSEG responded that the submitted supplemental response stated that the scale used was accurate to a tenth of a gram for the test being used to determine the Salem bypass design basis values. No follow-up information to the NRC is required. The licensee also provided a brief presentation at the beginning of the meeting which has been placed in the Agencywide Documents Access and Management System. 4 No regulatory decisions were made during the conference call. Participants included a representative of the New Jersey Department of Environmental Protection. No members of the public attended the meeting or participated on the teleconference bridge line. A list of the meeting attendees is provided in Enclosure 1. Please direct any inquiries to me at 301-415-3204 or John.Hughey@nrc.gov. P()~ John D. Hughey, Project Manager Plant Licensing Branch 1-2 Division of Operating Reactor Licensing Office of Nuclear Reactor Regulation Docket Nos. 50-272 and 50-311
Enclosures:
- 1. List of Attendees
- 2. PSEG Draft Responses Regarding Salem Bypass Testing For PSEG cc w/encl: Distribution via ListServ 4 ADAMS Accession No. ML13064A323.
LIST OF ATTENDEES FEBRUARY 28, 2013, CONFERENCE CALL WITH PSEG NUCLEAR. LLC SUPPLEMENTAL RESPONSE TO GENERIC LETTER 2004-02 SALEM NUCLEAR GENERATING STATION, UNITS 1 AND 2 NRC John Hughey Stewart Bailey Stephen Smith Paul Klein PSEG Paul Duke Emily Bauer Kiran Mathur Sameh Markos Gregory Sosson Robert Peterson (Sargent & Lundy) Helmut Kopke (Sargent & Lundy) Tony Ryan (Sargent & Lundy) New Jersey Department of Environmental Protection Elliot Rosenfeld Enclosure 1
ENCLOSURE 2 Draft Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28,2012
PSEG Nuclear February 21, 2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012
- 1) The screen used to catch the fiber was 0.31mm or 310 micron. How was it ensured that fiber did not bypass the screen considering that Nukon is 7 micron diameter and many of the bypassed fiber pieces are less than 250 microns in length with almost all less than 500 microns? Any fiber bypassing the screen may have been caught on the strainer on its next pass. Please provide the bypass amounts and debris sizes that would be expected to pass through the strainer and captured on a 100% efficient filter, and the method and assumptions used to calculate these values.
During both the 2006 and 2008 bypass testing, the bypassed the strainer were collected on a 0.31 mm stainless steel mesh maLX]!Ql+Ill diagonal opening of 0.44 mm. The bypass fiber size distribution of the,.strainer and m of the fiber bypass capture screen was obtained . .~ :tijf~lnples for bypass Test 3a in 2006 and for bypass Tests 1,2 and 3 in 2008. B von Bypass Tests 1 and 2 from 2008, which are expected to b\' most repres~ri:f4t;i:ve of the actual fiber size distribution that bypasses the straitt'.. ~'6% of th~';:fit~~that bypass the strainer are less than 0.5 mm in length (Table.3f.. \;.;Z"Z,of 20i~Supplemental Response).
. . ):;:; ~)'f:/ "<'<c;-;/~:!Jf{:;c5:'-;>;;>>_;.-;-
However, thIS percent<1gtl)s based on a:fihercount :rat1:ier than fiber mass. e" /../ ;:};" ____ '"' ;__. ,".;;,;-" I;n-j J>"
<c:v- --:-~':-~f':/>:- .<~-~
The percent o~h~sed fib\~t;s by mass tha~.a~~ in Size Class I (0.1 rom to 0.5 mm in length) in Table 3' .2.2 .. ij the 2012'Sii~plemental Response is calculated for 2008 B~;~W1ti~n below. nM'C>:p.rltlif(tf": of fiber in each size class by mass of fiber Pf= fiber density D = fiber diameter P percentage of fibers in each size class by count Subscripts 1, 2, 3, and 4 denote size classes from Table 3f.4.2.2.2-2 of the 2012 Supplemental Response Page 1 of 17
PSEG Nuclear February 21,2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012 The fiber density (pj) and fiber diameter (D) are constant across all size classes; therefore, the equation above is reduced to the following: Values of percentage by count (P) are taken from Table 3f.4.2.2.2-2 of the 2012 Supplemental Response. The average fiber length of each S\t:e class is used as the fiber length (L). For Size Class 4 (>1.5mm) an averageiIDfg~);l~ngth of 1.5 mm is used, which conservatively reduces the percentage of ";// by mass) in Size Class 4. Therefore, _ 0.26*0.3 M 1 0.26*0.3 + 0.39*0.75 + 0.28 *1.25 + The percent of bypassed fibers by mass that length) for 2008 Bypass Tests 1 and 2 was cal!CUliatC!Jto If it is assumed that fibers that ha ...pg.$less than (500 microns) are not collected by the fiber bypass capturJ;<scr;~H~?ilienQle bypass debris reported is potentially underes' ted. Therei6r~'ft(P;~~~6'ilil~Jor the potential for increased fiber bypass of strlillf ..*.. . , ough the n.31 mm me~h, the mass of fibers that bypass the strainer wiq2tf¢rwcreasid~y 9%.'";, , ;;"; :
';-:~~;0n
- 2) The 10cat1l&I1,A~f the s~~",!,,!>;~~,lgptwas not clear. Could turbulence in the
(,;~::-:~;;-~_~ --~:;<:>>!:;-:~---v;: :;:*/i..... ':;,:;:~-0!{<- -.. -, --:-<: ". >y4-:-- flumr;.;IJavepYt!¥.If1JJ,!d sO~~JJjber from collecting on the screen or washed some of thif.:ffi,er from thi~~{m?fC r-_'j~</;;_ ~f;>>_ <,. :-;:;.</ :;?~N~tt";~:i;j
;-t-;~;k:2';' , -,;.,:<-(:/ ;:/)
";'i',n_,y, {;F:':~\ /,.,
The fil'~J:)ypass captUt~ '/Y'. {J.'." x,<-; ~ scree~ is located in the test flume between the sparger and strainer. ./,-"-__ ,./) 'ii"',
-/:;}Y;}'-~-
~~s;::li::':>
It is possible th~t)~bulence in the test flume could cause some fiber redistribution on the fiber bypass' capture screen. This is not expected to result in reduced bypass quantities due to the procedures in place to collect fiber that may have fallen off of the fiber bypass collection screen prior to determining the mass of bypass fibers (See Question 3). Page 2 of 17
PSEG Nuclear February 21, 2013 Salem Units I & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28,2012
- 3) Discuss the procedures for handling the collection screens. How was debris ensured not to fall off the screen when it was removed, during drain down, handling, drying, etc?
Prior to test initiation, a clean fiber bypass capture screen is weighed and installed in the clean test loop. The clean test loop is then filled and the test commences. The fiber bypass capture screen is not removed during the test, thus minimizing the chance for captured fiber to fall off the capture screen. Furthermore, the fibers are continually compressed on the fiber bypass capture scrct;:.!1'byithe flow through the flume, thus helping to keep the fibers on the capture scr~J.i~h
<'j\3~2;j, Once the test is complete, the flow is stopped al'lg.the~test lOopi!~4rained. Then the fiber bypass capture screen is carefully remoy(ja:<</Plior to clea~iiij,the test loop, the
- - \;;;, ~\,~_-?f< -~,
walls and floor of the MFTL upstream oq~j < r byp~ss capture screeware' carefully checked for residual fiber. Any fiber th~t iSi~J:~~n~~.t~~i~~luded in th~'lotal bypass quantity which ensures that any fiber which';~~f:fal1 off the debris bed during draindown is accounted for. };!~~\tW'l¥';;';:\!;$;;;, Following removal of the fiber byp~ss;"f screen aM' checking the walls and floor for residual fib~r,.}he fiber bYJl~s~~P<n dried. The fiber bypass "is capture screen is I?lltce(I~~?:~zontallY in}~rt' oven, tQllI~ preventing fibers from falling off during the,~ing p;~s. Once th~ . f Iber bWass capture screen is dry (see response to QG~~{~~;\11 ),rtbe screen is:~ighed in the horizontal position, thus fibers f;oUjki:!!~;,9:m:ip.g.jthe weighing process. The difference
<"i',S> ' "<;*>;~';:/:'
bet:we:~n;tn(~;;tiI!~,;~ld inifftl~.;1?ypass capture screen mass is the mass of bypassed fiber. Page 3 of 17
PSEG Nuclear February 21, 2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012
- 4) How was it ensured that the samples were representative ofan average amount that would be present downstream of the strainer over the sample interval? For example, what was the timing of the samples compared to the debris additions?
Sampling has been noted to miss higher concentration clouds offiber that pass the strainer during debris introduction. This may not be important if the sample results are not usedfor evaluation offiber amounts over time. The downstream grab samples were taken immediately JplYowing the first debris addition and every 3 minutes thereafter. However, the,~k sample results were not used to determine the bypassed fiber amount ase~~~us~~~l;B.elow. Therefore, the timing of the grab samples compared to the debri~cadditions i~rl~t'relevant.
;~;<;rl-;:'>
-';:;~;iW;';'
The bypass results (in terms of quantity of;~~I>~ss per<unit strainerif~~~~!;1baSed on fiber bypass capture screen results (not s~~pij~~). SR~i!!cally, Figur;~;;3f.4.2.2.1-1, 3f.4.2.2.1-2, 3f.4.2.2.2-l and 3f.4.2.2.2-2iri:.1~c:tio~' 3f.4.2.2 ~f the 2012 Supplemental Response were gel'l~tated based on'ih~quantity of fibers collected on
/,/, ,':>'/
the fiber bypass capture screen. "it;";,.
;~; ;'i;<~'~!i;~;;
Downstream grab samples were takg4 /:5--,' 0:;.'- tod~~enlt.bet~ansient
--/o"'\}:::;)~;:_
downstream fiber concentration in39Qti~~~r Figure 3f.~j2.2.1-3 ilFSection 3f.4.2.2 of the 2012 Supplemental R~~~nse)/~:r~Jo determi~~Jhe size distribution of the bypassed fibers in both 2006fJ~d~'.f008 t~~e Tables i~4i2.2.l-1 and 3f.4.2.2.2-2 of the 2012 Supplemental RespJrt~~ '" ;~ ..~....p'
'/';j{2';'::>':;"": F
- 5) hest rep?~~te;il~~~i~ies~;re about 80 times lower than the maximum
. . . . velocity fo;<~~ Sal~j;,a}Jr~iner according to the vortex evaluation. How do veloilfi::gr:adients
<- ,;,;:---t;,
of;~~.magnitude affect bypass?
" f //7
-/'~(--:-Y: / :l~~:
Maximum .~{:ted J:fiiocity for Salem Strainer The maximum aij5foach velocity used in the vortex evaluation is 0.582 mls based on two pump operation (See Section 3.f.3.1.1 of 2012 Supplemental Response). However, at Salem the approach velocity is not equivalent to the penetration velocity due to the pocket design of the CCI strainer. Approach velocity can be converted to penetration velocity by dividing the approach velocity by the ratio (12.4) of the total pocket area (0.124587 m2) to the pocket approach area (0.010008 m2 = 120 mm x 83.4 mm). Page 4 of17
PSEG Nuclear February 21,2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconfereuce Held on November 28, 2012 In addition, the approach velocity used in the vortex evaluation is based on the flow to the closest one third of the closest module to the sump. While this approach is used to ensure the absence of vortexing, it is too conservative for bypass which is an integrated phenomenon rather than a localized phenomenon such as vortexing. Therefore, using the spreadsheets developed as part of the strainer head loss calculation (see section 3f.9 of 2012 Supplemental Response), the maximum penetration velocity across the entire first module for a clean strainer head loss condition is determined to be 0.027 mls. This is only higher than the maximum penetration velocity tested by eel in 2006. How Velocity Gradients at Salem Affect Bypass The maximum computed penetration velocitigsdir~ussed above on a clean strainer. As soon as some fiber reaches t .ou
>'7 velocity immediately decreases. Therefore, only a 1~1Sl~ant~~;)8f debris wil ".' fexposed to the maximum computed velocities. In fact, mo~t!tiebti;\¥iiI be exposed/to penetration velocities both significantly less.;~han the ma~ffu~ velocity and close to the penetration velocities tested. <?i '::l;;i'?dii,.,
o ;:/::(~;t/1{tj;\, The 2008 eel fiber bypass tests det~rmin~~tfi~~fll~.~tgnbasis fiber bypass quantity based on a uniform ' tion velociiY'fr0(0.0014jpJs over all strainer modules per eel Report 68 TD 90 157~} While / these tests underestimate the penetration vefbbi.~0::thro ,several of*~~~. modules closest to the sump, they overestimate the pettt~iY:~19~ity throllgh most of the modules further from the
',>..!fj7' , ~'*>***,';j:Cc~!::~},k~:{:,'{~~~~;f;:',~~{~?'
sump, ;iliSy1t'j KC:;,;.,;;" '
" '{~:,,;;(.,.'-
- "'i<~
J6~r~i!J~strate this')~~Mtt, of the expected penetration velocities with appi6~~~telY 2.3% d!~O" bed thickness) and 3.5% (1132" theoretical bed thickn~~~1~{)f the Unit j;:(fiber reaching the strainer is shown below (see question 18 for discussio~~{);p. the a~jhty of thin fiber beds to filter debris). Note, while the strainer and sump pitiil1~~~%:water before recirculation begins, approximately 2.3% of the fiber will reach ilitstrainer (based on the volume of water in the strainer compared to the volume of water in the sump pool at recirculation initiation assuming uniform debris concentration). The Unit 1 debris load is selected since it has a less uniform penetration velocity distribution due to a lower overall debris head loss than Unit 2 (see Section 3f.1O.2 of 2012 Supp lemental Response); however, the debris load is conservatively applied to the Unit 2 strainer since it has higher velocities. The plot of expected penetration velocity in each module was generated using the spreadsheets developed as part of the strainer head loss calculation (see Section 3f.9 of 2012 Page 5 of 17
PSEG Nuclear February 21, 2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012 Supplemental Response). Based on a uniform debris concentration in the smnp pool, two pump flow, and the strainer head loss being directly proportional to the quantity of debris on the strainer, it is expected that the penetration velocity profiles shown in the figure below would develop quickly after recirculation initiation. 0.012 --e- Expected Penetration Velocity with 2.3% of Unit 1 Fiber Reaching Strainer (HL Debris= 0.28 mbar 0.009 It) 0.011 0.010 --O--Expected Penetration Velocity with 3.5% of Unit 1 Fiber Reaching Strainer (HL Debris= 0.43 mbar=0.014 It) 0.009 0.008 - 2 0 0 8 Bypass Testing Base Velocity
-0(J) 0.007 c: 0.006 00 0
(J) 0.005 c: (J) 0.004 n. 0.003 0.002 0.001 0.000 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Strainer rvIodule (higher #'s closer to sump) Distribution As shQ;w;O!" at recirculation initiation (when 2.3% of Co II prd ~>.~!e~y 75% of the strainer modules will experience p~~tration velociti 000 we;'~tnan those tested in the 2008 CCl Bypass Tests. Furth&~re, by the t~';~ 3.5%/of the total Unit 1 fiber debris load has reached the strainer"(hp~roximatel~~ 25 seconds after recirculation begins), the penetration
.. <<;:;:;/'";"'_. ':"',:<1 velocities af'tlfe;modu~~s <:!:>-> v:,:, ,,->':*i nearest the sump are already significantly lower.
Non-uniform ve166ities could result in modules nearer the sump having a greater quantity of debris than modules farther from the sump due to the higher module approach velocities near the sump. The higher velocities could result in potentially more bypass in the modules near the sump. However, the results of bypass testing which give the mass of bypassed fibers per unit area based on a uniform debris bed/velocity show that additional fiber bypass does not occur beyond a saturation value which was reached during testing (See Question 6). Therefore, application of Page 6 of 17
PSEG Nuclear February 21, 2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012 the test results (bypass/area) could result in conservative results for modules farther from the sump since those modules may not achieve a saturation fiber quantity. For these reasons, the net effect of assuming a uniform velocity gradient when determining the fiber bypass quantity is expected to result in a best-estimate assessment of the total fiber bypass per unit area. Therefore, the velocity gradient across the strainer train will not adversely impact the design basis fiber bypass results.
- 6) Related to the velocity question above, is using bypa~}'~~;'iitrainer area valid?
Does a larger plant strainer area compared to t"'t!)l~st,strainer area result in linearly greater bypass or is it some other funci;~n?),(~~/d the larger plant strainer result in a less uniform debris deposi!i(JiI,,.e~ulting i~)i4t~~nge in bypass? How well do the bypass tests conducted valitldi; this ~elationship?'t'i~';.
'::;#i~:B't-;---;;,}/
The use of the existing strainer bypas~;6~1~, fnl);jarger strainersJ~fth velocity gradients within the strainer train is address~d~iifthe response t~ Question 5. Therefore, this response focuses o'll,;~he other portio'n§19f,this question.
~:{<_ '--:~);(-<n:~-~~:>i;:J:":- _ .-:--~~;);:>~-(~:~~);
The Salem fiber bypass test results deInQAsi,rate thattlie strainers reach a fiber saturation point as tes~s~ith Signific;iltlY~ff~f&rt~~ifO~lY distributed debris loads yielded similar re~~ltS~~;;fu~'~llS' the Sai~tests valid~te the use of bypass per area instead of OtheI;;~easur~ift~ts (e.g. bypa~s equal to a fraction of the total debris). Since the Salem;'fi~pypas~~easurement~~are obtained in tests with uniform debris beds (due/0~.* ~~me*ri~~t!p~Yftli~;~~S~l~lipresent a reasonable upper bound for the fiber 4ypais¥~Mty per:ar~a; i.e:' it;;ts~t\nlikely that any more fiber could bypass the 1i!0 0
;/ <:;~;:;;::;:-~~-> _ ;:,: -Pi:~:
str(iiner than was~observed in4:he tests.
,;f)iJf;~t£~~:; --~\(l~r::;~~"~. -"'{~;~f('~~~~;/:);/
'.'?:~f.1~~:-,~: -,:;~t)~;{:, ,_'<<0 In a l~~\strainer wi~:a potentially non-uniform debris bed, some modules would have mof~d.~lJris than in;;the test, while others would have less debris. Given that the
":;::;;I::~t,:,??>\ :~>~,;
maximum amo,tW:-to~bYPass is established by the saturation fiber quantity (obtained in the tests prio't't?yieaching the full fiber debris load), the modules with more debris than in the testcwill bypass a similar amount of fiber to the modules in the test. However, the modules with less debris (if insufficient to be saturated with fiber) will bypass less fiber than the modules in the test. Thus, extrapolating the fiber bypass results from the tests to a large strainer is conservative even if the larger strainers have a non-uniform debris bed. Page 7 of 17
PSEG Nuclear February 21, 2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012 Further analysis of the 2006 and 2008 fiber bypass testing data also validates both the use of bypass per strainer area as a measurement and the application of test results to a larger strainer. Although both the 2006 and 2008 tests utilized a fixed test strainer area, they also tested markedly different debris loads. Tests 1-4 in 2006 tested debris loads ranging from approximately 100 fe to 2600 fe of equivalent NUKON and resulted in similar bypass quantities (i.e. Test 2 did not experience 26 times the bypass of Test 4). Similarly, Tests 1-3 in 2008 tested debris loads ranging from approximately 10 fe to 500 ft3 of equivalent NUKON and~t!~ltlted in similar bypass quantities (i.e. Test 1 did not experience 50 times the bypJ~s dfTest 3). The tests with smaller debris loads result in debris beds equivaI~fit{[to those which would be obtained by testing a larger debris load on a much laf~er str;ii~eJ:,ceteris paribus. For instance, a test with 100 ft3 of equivalent NU~PNt~sillts in aiiiiml~~ debris bed to a 2600 ft3 equivalent NUKON debris load o?(~~trainerapproximat~ly~~.?times larger than the test strainer assembly. Thus, t4et~tcI~sult~ j~~!idate the appt~~ation of the results to strainers larger than the test module.'" ...'\':;: 'J
;-{~?~t~,:':\ .
See the responses to Questions 6 for a presenta~pn of the bypass test results
;~~!"S';;;~;.. ';;J' and debris loads, respectively.
- 7) Was sensitivity to water chemistry ev~lluat~~a?
>-;;~~!~@¥r~f:~~~~.-~.;::.-;>_-.
~::
The CCI B)'lJ;.... ests,W:~ich used for all testing, did not evaluate sensitivity to water::el;t.~mis~~.A at the October 18-19, 2012, NEI PWR s~g~,t~rfo;*+*'~~~((jS; summarizes the Texas A&M uni~.~I'Sr&*&~~L.*~xtIts conducted by STP to study the sensitivity of water 1YIJ~9n debris byPa~~~>Thindicated that there is no significant difference in oYJj~quantity whe~f~ing tap ~ter instead ofbufferedlborated DI-Water.
- 8) What w;';Jii:l~he batch 'iizes and what was the time interval between each batch?
(Batch size ifitl'jt .4ressed as a theoretical debris bed thickness). Large batch sizes could res eee e.
- .:;/
;"a debris bed forming more quickly than would actually occur in the plant, resulting in and less bypass.
Table 8-1 gives the batch sizes and time intervals between debris additions for the 2008 Fiber Bypass Tests. Page 8 of 17
PSEG Nuclear February 21, 2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012 Table 8-1: Bypass Testing Batch Size Summary Number of Elapsed Incremental Turnovers Theoretical Time Theoretical Bypass After l't Nukon, Kaowool, Fiberglas, Debris Bed Portion! Bucket From I" Debris Bed Test Debris kg kg kg Thickness, Addition, Thickness Addition in min Added, in (Note b) 0 0 0.496 0.09 0.09 2 4(') 1 0.16 0.07 3 8(') 3 0.18 0.02 4 Ii') 4 0.29 0.11 5 16 6 0.31 0.02 30 II 0.42 O.ll 2 38(') 14 0.11 2008 2 46(') Bypass 3 17 0.07 Test I 4 54 20 0.03 54 20 0.11 3 2 86(') 0.11 3 118 0.02 I 125 0.10 4 2 135(') 0.10 3 145 0.02 0.08 0.552 0.14 0.06 2008 0.552 0.21 0.06 Bypass 0.669 0.24 0.04 Test 2 0.669 0.28 0.04 0.182 0.089 0.32 0.04 0.03 0.01 Average Incremental Theoretical Debris Bed Thickness: 0.02 lapsed.~e is linearly interpolated between the portion start and end times .
. .e is~pproximately 2.7 minutes.
The batch sizes used for Bypass Tests 1,2 and 3 in 2008 resulted in theoretical debris bed thickness additions less than 118 inch (0.01 to 0.11 inches) and therefore are not expected to result in unrealistically low fiber bypass values. Furthermore, the debris prepared in each bucket was added to the test loop with a pitcher (see photo 10-2 in response to Question 10). Thus the actual addition sizes are even smaller. Page 9 of17
PSEG Nuclear February 21,2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012
- 9) What were the test results in lb of bypass? How will the results be used? Will a single maximum value be used or will a time dependent debris load be calculated?
Bypass testing results are used in the Downstream Wear Calculation (S-C-RHR MDC-2089) and Fuel Deposition (In-Vessel Downstream Effects) Calculation (S-C RHR-MDC-2295). Some of the results presented graphically in of the 2012 Supplemental Response are repeated below. Table 9-1: 2006 Results Summary ass Test 1 ass Test 2 ass Test 3a ass Test 4 (a) As-measured density of Nukon (l: fe /1000 t r . ; / (b) The conversion from Ibm / 1000 ft2 Cto . ...(t2 is performed using a density of 3 Ibdfe. This densityisthe average o~,they<,s.:)) /density of Nukon (1.941bm/ fe) and the as-meas~~~tl~:sity of KaowoolJ4. I Ibm / ft3j.'* The average fib ass ~Il1 the four 2a2~pon-Iatent debris load tests (1.015 it? / 1000 ft2) presente ..... :,.T(l, plus n,lltfgin, is used in the downstream wear 1 valif~l1i&e IS . ***~1g~ft~liooo ft2). As discussed in Section 3m.3 of
- ~!ttalR~~?nse, Appendix B of the Salem downstream wear
.91 ........ ~ion calcui<lf~~.~ fiber~~~ypass fraction of 1.4 % (using the data above),
resuJ~f,in an initial;~~r rem6~al efficiency of 98%. For conservatism, a value of 97% ~a;) ed in thea,(;mg term wear effects evaluation. The Downstream Wear not n;ti~ed to utilize the more prototypical 2008 bypass results since It~are conservative relative to the 2008 results.
;~Y Table 9-2: 2008 Results Summary Measured Fiber Bypass From Testing I [Ibm / 1000 tr]
i 2008 Bypass Test 1 0.85(a)
. 2008 Bypass Test 2 0.51 I 2008 Bypass Test 3 0.68 12008 Bypass Test (Average) 0.68 (a) MaXImum value from 2008 bypass testmg IS used m Fuel DepOSItIOn CalculatIOn Page 10 of 17
PSEG Nuclear February 21, 2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012 The fiber bypass quantity used for the existing Salem Fuel Deposition Calculation is 0.85 lbm / 1000 fe. This is based on CCl Bypass Test I in 2008 (see Figure 3f.4.2.2.2 2 in the 2012 Supplemental Response). The results of Bypass Test 1 in 2008 were used since they result in a conservatively greater bypass quantity than the results of Bypass Tests 2 and 3 in 2008.
- 10) Discuss the procedures for debris control during the test~~;;it!l0w were the fibers ensured to all make it into the test tank after weighing a1J:~1i;;;;aration?
The dry fiber material was weighed in separate buck,ej~~~ ".. ',were labeled with tape (see Photo 10-1 below). The dry material was th~l.l?@itper di~l~into batches which were soaked with water prior to being blast~dWitha' water jei::iS'I1le fiber batches were placed in larger bucket(s) in order t<:)a¢~ommo~te water frorii:i~~~.jeJ blasting. The quantity of debris in these larger buckef . . . als~;r¢sorded and pl~td on a tape label on the bucket (red circle in Photo 1O-2):'+~§'f~rge'buckets we~e not drained during the water jet blasting to that no fibb~g:;:fil1es were lost. The prepared
,_ -/h.///
fiber debris bucket was then nearth~Jv1I;'TL and the fiber slurry
- hlu'luc-t method'~iiefe a pitcher was used to
.l.U::"':J.!i~'~" {0-2).
Photo 10-1 Photo 10-2 Furthermore, given that the results of the fiber bypass tests indicate that the strainer reaches a fiber saturation value beyond which additional fiber does not result in significant bypass, the loss of a few fibers during preparation and addition would be inconsequential. Page 11 of 17
PSEG Nuclear February 21, 2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012
- 11) Discuss the controls in place for verifying an accurate weight of fiber in the collection screen. How was the drying process for the screen controlled before being weighed and inserted into the loop?
Prior to each fiber bypass test, the clean fiber bypass capture screen was dried in an oven at 1000e (212°F) for 24 hours to ensure the dryness of the screen prior to measuring the initial (pre-test) weight of the screen. Once the fiber bypass capture screen was removed fr~~!~~e test loop (post-test), it was dried in an oven at 1000e (212°F) for 24 ho~~aiIfi' being removed and weighed. This is long enough to ensure a comp!7t~!yr~4rY fi '~Jj:~.~ss capture screen based on eel's experience which has shown;;'th~t'70oe fori~;*bpurs is actually T~***;'Z*j. sufficient for complete drying. The final~~ight of\pe fiber bypassC 1:tp"e screen (including dried fibers) was then comparedild;~~initi~~tJ?.re-test) weigjb"determine the mass of bypassed fiber collected on the cap~e;.::>c't~e~: In the unlikely event that the fiber bypass capture screen not completeliiiY after 24 hours, it would result in conservative fiber bypass the residuar\Water mass would be included
";-~:"J~~~}/
in the bypass mass. 'j{::.; the fiber bypass capture
- 12) At whq~$f. reach a small constant value? Is this dep~iif1~~t on enetration velocity? Was it determined that bypass
;::~~?
h!i'J;ff?$topped or r l small constant value prior to the tests being
;ec:tP(~~. If not, wha -,:;3;, re thtftermination criteria?
-~j/,-//'_->'- -
"--t:~~m>~_,__ ';~~((
Based on th(* suIts of.~e 2008 Fiber Bypass Tests, the amount of fiber bypass is not proportional~. t of fiber used in the testing beyond a "saturation" value of fiber on the stra! ockets. Therefore, the values used in the 2008 tests are sufficient to show total bypass. This is supported by Figure 3f.4.2.2.2-2 of the 2012 Supplemental Response which shows the relationship between bypass quantity per strainer area and the total mass of fiber added to the test loop. No specific termination criteria were used in the fiber bypass tests. However, it is very likely that fiber bypass stopped prior to the test termination for all tests based on the following: Page 12 of 17
PSEG Nuclear February 21, 2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012
- Each of the three fiber bypass tests performed in 2008 had a test loop turnover time of approximately 2.7 minutes. During the 2008 bypass testing, the time between the final debris addition and test termination was 20.3 hours, 19.3 hours, and 5.9 hours for Bypass Tests 1,2, and 3, respectively. Therefore, the test loop was able to turnover approximately 451, 429, and 131 times between the final debris addition and test completion for Bypass Tests 1, 2 and 3 in 2008, respectively. At Salem, a maximum of 54 pool turnovers would occur before hot leg switchover (based on the maximum tiI'Q~;!o hot leg switchover, maximum ECCS flow, and minimum pool volume),pj/ ..
,> '<;~~:.: .
/" xj/ A;:>-"t;,
- The total fiber bypass quantity in Test 3 is similar t~'th~ total bypass quantity .
in Test 1 and Test 2, despite Test 3 beingp1nrOr 131 ti1I4i~"ers instead of 451 (Test 1) or 429 (Test 2) turnovers (see p,;jgtir63f.4.2.2.2-2i~~;!~>. Based on the number of pool turnovers that occ tween the final debris addition and test completion and clear wa t test terminaii~);i" e time chosen to terminate the tests was appropriate. Longer bs:t~§ting times ~lltd;not be expected to result
":\" -'/-:;-<~{1;?:';-- :'/:.", /. "'/;'[;0 in greater fiber bypass test r e s u l t s . ; *...
'. ;("'.
ged mor;'~Ij;an once<:f~ring the test? lfso, when and
<;?j'
- 13) Were the filters (~~,.e how were the .g!fJ~Il~~.es ed? ,~
()t;?J~~ged during the tests. It was installed
-*;ji4i)2~*_
The fiber bypass cap~~;~ '/<:;".'
'0_~?';}~?----'_'.--" <:i)}P\:~~;';_ ',-"
prior tQ';test'i~li~1i()n ari'dC~oveda "test completion.
;;t5 F
"-"/'1:3:£ ;;:Si{A~'~/~_
->~~~;:~P~~;!:;_"~_'~_'fP
- 14) tnouni~~!ir the graphs included in the Salem supplemental 2.2 bfi~ed upon (filter results or sampUng)?
- {;t The bypass .*.. ts
"'::n-;t?V'"'x, -</:~f (JJ~erms of quantity of bypass per unit area) are based solely on fiber bypass caIi~ screen results. Specifically, Figures 3f.4.2.2.1-1, 3f.4.2.2.1-2, 3f.4.2.2.2-1 and.~f.4.2.2.2-2 in Section 3f.4.2.2 of the 2012 Supplemental Response were generated based on the mass of fibers collected on the fiber bypass capture screen.
Note, Figure 3f.4.2.2.1-3 in Section 3f.4.2.2 of the 2012 Supplemental Response documents the transient downstream fiber concentration and is based on grab sampling. However, this data was not used to compute a fiber bypass quantity. Page 13 of 17
PSEG Nuclear February 21,2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012 Fiber size results presented in Tables 3f.4.2.2.1-1 and 3f.4.2.2.2-2 are based on grab sampling. However, these tables are not used to calculate the bypass amount.
- 15) How were the debris amounts in table 3f.4.1.3.4-1 calculated? For example, in test 1 I calculate that 7.3 kg of debris should have been added. 1212.5 frI180.8x2.4Ib/frI2.2Iblkg = 7.3 kg, not 5.9 kg.
The debris amounts in Table 3f.4.1.3.4-1 were \"<1I',",Ul"lt""+j.~I~HJ:b density of the Nukon (31.13 kg/m 3 or 1.94 1bn/ft3 , supplied to eel for testing as shown in the equations 1"\"'1,,,...~t*, As-measured: - 1212.5 .1.94~.1.0kg = 180.8 ft3 2.2tb Nominal: 1212.5 ft3. 2.4~.1.0kg = 7.3k 180.8 ft3 2.2tb g The as-measured density is the me fiber supplied to eel for the 2006 bypass testing. ""v~",,,,t~rI (data sheet) as-fabricated fiber Page 14 of 17
PSEG Nuclear February 21,2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012
- 16) Did two sided strainer tests result in a different amount of bypass when other conditions were similar to the single sided test?
The single-sided (2006) and two-sided (2008) strainer module tests are described in Sections 3f.4.1.3, 3f.4.1.6, and 3f.4.2.2 of the 2012 Supplemental Response. Parameters which differed for the tests are provided below. Table 16-1: Com arison of2006 & 2008 Fiber B Parameter 2006 B Test Module Baked Fiber Debris? No Yes*' << . ; ON & Kaowool) No (Anlii'S:weat Fiber lass) Tested Fiber Types NUKON, 1 test wunJ~~aowoOJ NUKON,~Q'Ypol, Anti
~$weat Fiber 'f~~s Plant Fiber Debris Test 1: Test 1: 268.8 ft NUKON Load Test 2: 37.0 ft3 Kaowool (summation, L, is 54.6 ft3 Fiberglas equivalent NUKON 12.5 ft3 Latent load where applicable LequivNUKON 495 ft3 .
based on §3f.4.1.5.8 Test 2: 20.9 ft3 NUKON i of2012 .. ::;Jt: *..., 29.2 ft3 Kaowool Supplemental 47.0 ft3 Fiberglas Response) 12.5 fe Latent LequivNUKON 208 ft3 Test 3: 12.5 ft3 NUKON Debris added in small m
~2.7 minutes Thus, there w\g'tests performed where the only difference between the tests was
/'l:'
the strainer module in the test loop. Therefore, the impact of the strainer module by itself cannot be ascertained with certainty. Page 15 of17
PSEG Nuclear February 21, 2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012
- 17) It is difficult to evaluate the results as presented for the 2008 tests because they are given in units of volume. There were 3 different types of fiber used, all with different densities. How does each of these contribute to bypass? How were the volumetric values determined? Was some sort ofaverage or composite density used that assumed an equal bypass ofeach type offiber?
The 2008 Bypass Testing results are presented in Figures 3f.4.2.2.2-1 and 3f.4.2.2.2-2 of the 2012 Supplemental Response. Figure 3f.4.2.2.2-l pre~nts the bypassed as fabricated fiber volume per 1000 ft2 strainer area whileri~;~~"3f.4.2.2.2-2 presents the mass of bypassed fiber per 1000 ft 2 of straine:t::;i;11i~:?mass based results are obtained directly from the bypass tests since the testfinea~:fit~the mass of bypassed fibers. The volume based results are obtained by,.cO:rt:VCrting thei~s based results to as-fabricated volume using the as-fabricat~¢>densiiy of Nuko~;;'(2A Ib/ft 3). This conversion method is conservative since.N~kon has the lowest denJ~:pfithe tested fiber debris (density of Kaowool = 8.0 Ib/ft3 ,:de~§~tyj6fAnti-Sweat Fib~rglass 4.0 Ib/ft 3) and therefore results in the greatest bypas~ed\ftber volume associated with the bypass mass measured in the test~i!i,;:?,i'
'"'1:"~::';~~f~"~Jgii;~;.."
The mass based results (in Ib/1 000 ~i:of~b-~merlirea) ar~~dlrect1y used in the Salem c;: "; 'yr" r/-'"~r;<_--;>____ _ .' Fuel Deposition Calculation. TherefoJ,:e, /~e methppJor converting to volume of
- Y;.,i;;~,~:<" ~ :'*'" ,/'£";' " ,,'~'--';'
bypassed fibers pe~i~quar~;:t'90t is not geIplane.;' 18)Discuss why at~i;;'ifith a~tger 'c'~;~~:r:~:-, 4<:j!;J: e ..>--:;,.. fiber lo;d,;;(test 2) resulted in a lower bypass value
'~1:/!'
than a test with a IOWl!itfioer"lOii#i{lJ!~J3J;;;Did 2008 test 3 form a filtering bed over
"{~~U)i/<~~:;;: ;. '~</~?O-~f" .. '"'::'-":M,~{,:;*;;>,,:,,~t, the entjie;sffil~ti.'er:~. Coulilc~h,is be relaled to artificially fast arrival time for the fiber
~"t";'J '""?;;>;:~';!:~;~ '~-',. ;.'. ;;-~;:*;Y::,~.-*:'
as4(~cussed abo~e.{w~!h regtlr~tofiber batch size?
/';{~f~~fit... ...... ~i~~}ccJ:i')
Althou~:Jpe fiber arriy~l time for Bypass Test 2 is faster than Bypass Test 3 (Sec Questio;;(81~,tkc most l~~ely reason for Test 3 having more bypass than Test 2 is that Test 2 inclt:<'t~~r'Qthetjdebris types (e.g. Kaowool and Fiberglas). The inclusion of Kaowool is not~~fe:H~ that its presence likely inhibits bypass due to the nature of the insulation. Kao~ool is a needled insulation in which the individual fibers are not chemically bound, but are physically interlocked. The physical interlock is a result of fibers with "hooked" ends. The hooked ends could attach to any other fibers during debris preparation, effectively resulting in longer fibers which would be less likely to bypass. Page 16 of 17
PSEG Nuclear February 21,2013 Salem Units 1 & 2 Responses to NRC Questions Regarding Salem Bypass Testing During Teleconference Held on November 28, 2012 Bypass Test 3 in 2008 investigated bypass with only latent fiber (tested as NUKON) and resulted in a theoretical homogenous debris bed thickness of approximately 0.036 inches [==(30 Ibn/2.4 Ibm/ft3)/( 4656-500) ft2* 12 in/ft) or 1128 inch. Although this is much less than the theoretical debris bed thickness of 0.32 inches obtained in Test 2, it is sufficient to form a filtering fiber bed over the entire strainer. A filtering fiber bed is one which blocks fiber bypass, but does not necessarily result in significant head loss. It should be noted that some "fast" (i.e. low head loss) laboratory grade filters effective at blocking particles 15 !lm and larger on the order of300-500 !lm or 0.01 to 0.02 in. Page 17 of 17
-2 Question #7:
The NRC staff is considering the fact that only one type of tap water was te$ted to determine if there are effects of testing with tap water instead of buffered borated water. Different types of tap water have been shown to perform differently in head loss testing. PSEG will review available industry-sponsored testing. Question #11 : The NRC staff asked about the accuracy of the scale used to weigh the screen. PSEG responded that the submitted supplemental response stated that the scale used was accurate to a tenth of a gram for the test being used to determine the Salem bypass design basis values. No follow-up information to the NRC is required. The licensee also provided a brief presentation at the beginning of the meeting which has been placed in the Agencywide Documents Access and Management System. 4 No regulatory decisions were made during the conference call. Participants included a representative of the New Jersey Department of Environmental Protection. No members of the public attended the meeting or participated on the teleconference bridge line. A list of the meeting attendees is provided in Enclosure 1. Please direct any inquiries to me at 301-415-3204 or John.Hughey@nrc.gov. Ira! John D. Hughey, Project Manager Plant Licensing Branch 1-2 Division of Operating Reactor Licensing Office of Nuclear Reactor Regulation Docket Nos. 50-272 and 50-311
Enclosures:
- 1. List of Attendees
- 2. PSEG Draft Responses Regarding Salem Bypass Testing For PSEG cc w/encl: Distribution via ListServ 4 ADAMS Accession No. ML13064A323.
DISTRIBUTION: PUBLIC RidsNrrPMSalem Resource TWertz, NRR Branch Reading RidsNrrLAABaxter Resource TSantos, EDO Region 1 RidsAcrsAcnw_MailCTR Resou rce NSheehan, OPA Region 1 RidsNrrDssSsib Resource RidsNrrDorl Resource RidsRgnllViailCenter Resource RidsNrrDorlLpl1-2 Resource RidsNrrLAABaxter Resource ADAMS Accession Nos.: Package/ML13078A442; Meeting Notice: ML13036A435; Complete Set of NRC Quesf Ions: ML13052A794; M eef 109 Summary: ML13078A448 *via . emal'I !OFFICE DORLILPLI-2/PM DORLILPLI-2/LA
- DORLILPLI-2/BC DORLILPLI-2/PM NAME JHughey ABaxter iMKhanna JHughey DATE 03/28/2013 03/28/2013 ,04/01/2013 0410512013 OFFICIAL RECORD COpy}}