ML17275B149
| ML17275B149 | |
| Person / Time | |
|---|---|
| Site: | Columbia  |
| Issue date: | 07/09/1981 |
| From: | Bouchey G, Bouchey R WASHINGTON PUBLIC POWER SUPPLY SYSTEM |
| To: | Schwencer A Office of Nuclear Reactor Regulation |
| References | |
| G2-81-183, GO2-81-183, NS-L-02-CDT-81, NS-L-2-CDT-81, NUDOCS 8108040092 | |
| Download: ML17275B149 (52) | |
Text
REGULATOR lNFORMA'TION DISTRIBUTION TEM (RIDS)
ACCESSION NBR ~ 8 1 08040092'OC ~ DATE:: 8 1 /07/09 NOTARIZED:<<NO, . DOCKEiT FACIL<r 50 397 WPPSS Nucl ear Pr oj ect F, Unit- 2E Nashington Pub l i c Powe 05000397 AUTH', NAME" AUTHOR AFFILIATlION BOUCHE Y> G ~ D ~ Washington Public Poweri Supply System RECIP ~ NAMEI RECIPIENT AFFILiIAT'ION SCHNENCER F A~ L1i censing, Branch 2
SUBJECT:
For war ds responses to Round Two questions from: Conte'ment1 Sys Branch. Responses will bel incorporated into FSAR amend wi thin>> 4 months ~ GE proprietary f igur es sent under sepal ate cover by 81 0709 l tr ~
DISTRIBUTION CODEi: B001S COPIES RECEIVED:LTR I. ENCL 'IZE::
PSAR/FSAR AMDTS and Re ated Corr espondence 1
'ITLE" NOTES! PM: 2', copi es'f a 1 l mater'l ~ 1 cy.'BNR-LRG PM(L'. RIB) 05000397 REC I PI ENT COPIES RECIPIENT .COP IES'TTR ID CODE/NAMEI LTTR ENCL ID CODE/NAME ENCL>>
ACTION:" A/Di LICENSNG 1 0 LIC BR ¹2 BC 1 0 L'I CI BR ¹2'A 1 0 AULUCKR R ~ 04 1 1 INTERNAL'. ACCID EVAL 1 1 AUX SYS BR 27 1 1 09 BR26'HEM ENG BR 1 1 1 ~
1 CONT SYS BR 1 1 CORE PERF BR 1 0 1 1 EFF TR SYS BR12 1 1 EMRG PRP DEV 351 1=- 1 EMRG PRP L IC 36 3 EQUIP QUAL BR13>> 34 3 FEMA REP DIV 39 1 1.
GEOSC I ENCES 28 2 2 HUM FACT ENG 40 1 1.
HYD/GEO BR 30 2. 2 I 8C SYS BR 1 I LEI 16'IC 3 3 GUID BR 33 1 1 QUAI 'R 06'I C 32'. 1 1 NATL~ ENG BR 1 7 1 1 iilECH ENG BR 1 8 1 1 MPA 1 0 NRC'DR 02'. 1 1 OELD 1 0 OP" LIC'R 34 1 1 PONER SYS BR 1 9 1 1 PROC/TST'EV 20 ~ 1 1 QA BR 21 1 1 S BR22'. 1 1 REAC SYS BR 23 1 1 WEG 0 1 1 1 S I T" ANAL OR 24 1 1 C T ENG BR251 1 1 EXTERNAL: ACRS 4], LPDR 03 1 1 NSIC 050 1 11 NTI S 1 1>>
'ggG p
'l j%~
65 @'o TOTAL NUMBER OF COPIES'EQUIRED: LTTR PZ ENCL'
K Fr" ~ ~
f ff ~
K"
~
I I ~ f It 'i<< ~
Kf
+ << tr r
KK<<>K K K' K>> ir 4 ci Nlf e - ~
<<>>,1' f r m>>>>I>>>>
IK ffff
(> IK*
t
~K t "<< ~ KK' I I F.I K
~ .t K ~
K
Washington Public Power Supply System P.O. Box 968 3000George Washington Way Richiand, Washington 99352 (509) 372-5000 Docket No. 50-397 July 9, 1981 G02-81-183 NS-L-02-CDT-81-007 Director, Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D.C. 20555 Attention: Mr. A. Schwencer, Chief Licensing Branch No. 2 Division of Licensing
Subject:
SUPPLY SYSTEM NUCLEAR PROJECT NO.
TO ROUND TWO (lUESTIONS 2'ESPONSES CONTAINMENT SYSTEMS BRANCH
Reference:
G02-81-184, letter, G. D. Bouchey to A. Schwencer, "Responses to Round Two guestions, Containment Systems Branch, GE Proprietary Figures", July 9, 1981
Dear Mr. Schwencer:
Enclosed are sixty (60) copies of the responses to Round Two questions from the Containment Systems Branch. These responses are to be formally incorporated into the FSAR in an amendment within four months.
Seven (7) copies of the GE proprietary figures were sent under separ ate cover by the referenced letter.
Very truly yours, G. D. BOUCHEY Director, Nuclear Safety GDB:CDT:ct Enclosure cc: WS Chin, BPA TA Mangelsdorf, Bechtel J Plunkett, NUS Corporation NS Reynolds, D&L JJ Verder ber, B&R (NY)
JA Satir, B&R (NY)
AD Toth, NRC R Auluck, NRC slosoeoova aso7oeW PDR ADOCK 050Q0397I A
I
,l l
WNP-2 Q. 022.053 The Caorso test results discussed in your r eport on the .
saf ety/reli ef valve (SRY) Loadsr "SRV Loads-Improved Definition and Application Nethodology for, Nark EI Contain-mentsr" exh ibit sone high frequency pressure spikes in the boundary pr essure measurenents during the initiaL phase of.
the air cle aring transient. Since these spikes were not observed in pr evious quencher test data (i.e.i the German, quenc her t e sts) r this ohenomenon suggests .ha he specif ic type of que ncher design may be important in determining the characteris ties of air clearing loads. Accor dinglyr provide a detailed description of the WNP"2 quencher
'.ncludino a of the hub design. Compare the geometry'escription geometry o the WNP-2 quencher design with the device tested in Caorso. Discuss any differences that may exist'between them. Indi cate how these dif erences might infLuence air clear ing lo ads ~
Response
As can be seen 'in Table 022.053-1r the geometric properties of the quencher de'vices used in WNP'-2 and Caorso are essentiaL'Ly
~denticaL. The arns are of the same diameter and thickness
'and differ in length t>y only 3l4". The holes in the quencher arms are identical in numbers sizei spacings or',entation and Location with respect to the center Line of the quencher. The huks are of the same size for both quenchers. WNP-2 has a slightly .hicker quencher hub than Caorso. The conical transi-tion piece f rom SRVDL to quencher hub has a slightLy more gr aduaL taper for WNP-2 as conpared to Caorso (10 vs- 13.5 ) .
Since the quencher configurations, are nearly identicaL as outlined above'her e is no reason to expect a difference in the characteristics of air clearing 'Loads.
The only geometric difference in the pLan orientation of the quencher arms is i Llustrated in Figure 022.0S3-1. 0 WNP-2 quencher arms forn two centr al angles of 80 and two centraL angles of 100 . At Caorso the quencher arns form three cen.ral angles o- 80 and one central angle of 120 0 . This results in two of four arms being shifted 20 with respect to Caorso quencher arms. This minor di ference in arm orientation should have no effect on air cLearing Loads.
The SRV report defines two types of time histories for the forcing function: one which'exhibits.some ini ial high ~
.f requency pressure spikes and another which Looks more Like
'a typi ca l single f r equency wave form (c la imed to be observed during German quencher tests). Both types were observed during Caorso tests. Consequentlyi since the MNP"2 design basis time history specifies use of bothy this should take care of the specified concern.
MNP "2 TABLE 022,053-1 CONPARiSON OF QUENCHER GEONETRY VN'P'-2 'C'a o'r s o Number of Arms Length of Arms (f rom h Hub) 4'-11 1/4" 4 I-10 1/2" Diameter 8 Thickness of Arms 12" sch. 80 12" sc h. 80 Number of Holes per Arm 1496 1496 Size of Holes (8) 0.39" 0.39" Spacing of Holes 1.96" 1.96" Quencher to F'irst Row of Holes 1 '"10 3/4" 1'-10 3/4"
'I Hub Size (8) 24" 24 II Hub Thickness 2.3" 2.0"
<<t WNP-2 Q. 022.054 The Caorso test result s discussed in the SRV report cited above'ndicate that t he size of the vacuum breaker on the SRV Line is important in determining the refLood tr ansient after valve closure an di consequent Lye the subseque nt vaLve actuation Loads ~ Indi cate whether the sizer number and characteristics of the vacuum breakers instaLLed on the SRV Lines of the WNP-2 fac ility are similar to those of the Caorso pLant. If ther e are di f f erencesr di scuss wh at effects you expect these diffe rences may have on your faci l ity>>
State how there differ ences and their effects wiLL be incor-porated in your load d efinition for the WNP-2 facil ity.
Response
Two vacuum breakers in parallel are instaLLed on each of the 18 main steam reL ief valve discharge Lines in WNP In designing the vacuum breakersi the effects of the sizings opening timei pressure losses and delta P were considered for each discharge Line. From these considerationsi a set of conservative characteristics necessary to provide protection for the discharge Lines for subsequent actuation was determined.
A comparison of the characteristics of the WNP-2 breakers with those in the Caorso plant is as foLLows'.
/pe ~4g ~
WNP "2 Caorso Manufacturer GPE Controls Atwood 5 Norrill Co.
Siz,e (in.) 10 10 Type S ingle wafer Single straight type with through wi th swinging disk swinging disk Number 18 16 FLow area (in. 2 ) 38.48 78.54 Design Conditions.
FLow 1 966 S C Fil* 14000 CFN Pressure (psig) -0.5 to 2 0.0 Delta P (psid) 0.115 7 Set point (psid) 0.1 Not available Opening t ime (sec. yt ps id) 0.21 9 0.4 Not avaiLable>
A/iK (ft. ) 0.278 (40 'in. ) 0.72 (104 in. )
- 'Acceptance point at'0.115 psid.
I Span WNP-2 Differences between. the values Listed are apparent; the flow arear design flows delta Pr and A/~K of the Caorso breakers are Larger. This arises because they are expressed in terms of quantities occurring at different operating points of their respective operating characteristic curves. The WNP-2 breakersr for examples are expressed in terms of values occurring at delta P = 0.115 psidr whereas those in Caorso are expressed in terms of values occurring at delta P = 7 psid. At these unequal differential pressuresr the flowsr pressure Losses (~K) and area res i stance coef f i c i ents (A/0 K) are expected to be different. Xf the 14r000 cfm flow through the Caorso vacuum breakers is scaled down by the square root of the ratio 0.115 psid/7 psid (since the f Low through the u'<<~
breakers varies approximately as the square root of delta P) i the Caorso breakers are caLculated to pass approximately the same equi v Lent f low's the WNP".2'reakers. As a results the two, break rs 'should behave in ess ntially- the same manner.
ua~~ been As has indicatedi vacuum breakers'etermine the:reflood transient within the SRV Line. The refLood transient estab-Lisnes the in,itiaL wa.ter Level condi'tion w,ithin the SRV Line prior to an SRV discharge events i.e.r an initiaL condit-ion for the event. Test data for a diversity of ini'tiaL SRV Line water Level conditions (Lowi normaL and high water Level)r
- were gathered during the Caorso tests. The WNP-2 SRV Load "definition envelopes the Loads observed at.Caorso.. As such>
differences which might exist in vacuum,br. akers are accounted for by this bounding approach to Load definition which envelopes test data obtained for Lowe normal and high water Level initial conditions.
- See Reports: NEDE-25100-P and NEDE-24757-Pr Caorso Phase i and II test reports.
<<t 'I 1
MiNP-2
- a. 022.055 The design vaLues for the transient SRV Loads in tne WNP-2 f ac i ity are based L on single va Lver subsequent actuation data from Caorso in-plant tests. These design values are then used in Load cases involving multiple valve actuations based on the assumption that, actuation of mu lti.pie vaLv s oc.curs only for the first actuation of the SRVs. State whether this assumption can be supported by a transient analysis of the worst transient event expected in the WNP-2 facility. Zf this is not the caser revise your Load definition to consider he multiple vaLve effect on 'the design basis pooL boundary Loads.
Our concern is that the Caorso test resuLts indicate that pressure Loads from multiple-valve actuations are greater than those from single-valve actuations under similar first actuation conditions..
Response
The SRV report states thai a multiple valve actuation case is more Likely to be an initiaL actuation rather than a subsequent a'ctuation. This statement is only made to show added conser-vatism in the design value since initial actuations are
'expected to be Lower in peak pressure ampLitude than subsequent actuations.
There is insuf- icient data from the Caorso tests to provide a statistical bas e for muLtipLe vaLve actuations. For comparisor.
purposes though the maximum measured pressures at P"<9 may be examined:
P max for singLe valve (valve A) r initiaL actuations 5.96 psi (See Table 4.1 of SRV report.)
P max for nultipLe valvei initia'L actuations = 5.87 psi (Se Table 4.5 o- SRV report.)
If the initial. actuation tests measured at Caorso a re used as the data base for comparison purposes~ it is found that a single valve test has recorded the highest pressure amp L i tude.
If aLL of the Caorso tests are used as the data bas er it is again found that a single valve test has recorded t he highest pressure amplitude.
s WNP -2 Furthermorer in comparing design envelope frequency spectra with multiple valve actuation data'rom the Caorso Question 022.056m Figure 022.056-2)i conservative tests'see resuLts are observed.
lt was concluded thatr since single valve actuations of Caorso tests yields higher maximum pressure amplitudes and since the frequency spectra enveLope of aLL multipLe valve tests performed at Caorso is enveLoped by the design envelope frequency spectral the Load definition is indeed adequate.
N WNP-2 Q. 022. 056 In Figure 6.8 of the SRV .report cited abover you indicate that the frequency spectrum of the design pressure-t ime histories can bound the experimenta l pressure-time traces at a statis-ticaL confidence Level'of 90 percent/90 percent and also bounds the envelope of the single valve subsequent actua ion ~
pressure traces fron the Caorso tests. Provide similar comparisons for Leaky valve (LV) first actuation data and for mul t i p'Le valve ac tuat i on (HVA) dat a. i s that the distinct differences in 'the characteris.ic concern Our in LV data (e.g.i the freque'ncy and amplitude) and the greater number of initial pressure spikes in NVA data.
Response
Comparison of Desi n Envelope Versus Leaky Valve Actuation Data A compar i son between the design envelope respon se spectra and the L ea'Jly va Lve f ir st actuation response spectr a (measured at P19) is shown in Figu re 022.056-1. The design curve completely envelopes the Leaky v alve f irs. actuation enveL ope recorded from Caorso data ther eby justifying the desion curve as adequate.
The Leaky valve f irst actuation envelope is dev e Loped f rom testsr 4ir 42r 43 and 44 of Caorso Phase II tes ts.
Comparison of Design 'Envelope Versus Nultiple Valve Actuation Data In the case of sin'g Le va Lve actuat i onsr va Lve "A" was considered ands as shown in'he SRV reports r data recorded by sensor P19 was cons idered adequate to,represent the Loca l as well as the globaL boundary pressure frequency conten . In the case of
~
vaLve actuationi in order to obtain a vaLid comparison 'ultiple with the des ign (g loba L) .pressures 'one must determine a g L oba L (averaged) frequency content about'he quenchers which are actuated. Pressure sensors correspondino approxinately to the position of P19 in reference to valve "A" are selected for each of the actuated quenchers. Their readings could be used as a measure of loca l boundary pressure frequency content. Howevers their average is used for comparison with the design (global) pressure. Caorso tests were instrumented to record Local data for quenchers "A" (sensor P19)r "E" (sensor P50) and "U" (sensor P51). Thereforer for each test in which quenchers Ar Er and/or U are actuatedr the Locally recorded frequency spectra are averaged to give a measure of global response for that test. An averaged frequency spectrum envelope is then generated for the multiple valve actuation tests consideredi and is
t 4>>
h'NP-2 compared to the design enveLope in Figure 022.056-2 with conservative results. The multiple valve actuations f r equency spec t rum envelope i s developed irom ". est s 32@ 45-1 and 45-2.
24-27'9'0/
WNP -2 Q. 022.057 Nany of the Caorso subsequent actuation tests were conducted with one of the two vacuum breakers blocked. You used the results from these particular tests to derive the design values of SRV pressure transients for the WNP-2 facility.
Howevers the maxi.mum pool boundary pressure measured in the Caorso'ests is from a subsequent actuation test with both vacuum breakers operating (Tes't 22A02)i which we believe to be prototypical for Nark Ii plants. The maxinum measured value of the peak positive 'pressure is 8.7 psi and the mean value is 5.9 psi for subsequent actuation tests w',th only one vacuum breaker functioning. They are 9.4 psi and 7.3 psi~ respectivelyi when two vacuum breakers are functioning.
This represents a potentiaL nonconservatism in the data bas used in the derivation of design .values of SRV pressure tran-sients for the WNP-2 facility. Accordinglyi discuss this phenomenon and its effect on the data evaLuationr incLuding your derivation of the design basis SRV Loads..
Response
As ind icated in the response to Question 022.054'acuum breake rs affect SRV .Loads by influencing the ref Lood transi ent which in turn establishes initiaL SRV Line water Lev'el'nitia i.e.r the vacuum breakers infLuence .one of the L condi ions for SRV discharge. The Caorso test matrix includes few subsequent actuations with both vacuum breake rs operating; too few actuations to make any statis-ticaL conclusions. Howevers the maximun pressure amplitude 1 s OT interest. Tne P 90/90 value for Caorso data equaLs 9.37 p Sl i The max imum pressure amp L i tude r ecorded a t Caor so for sub-sequent actuations for cases involving the op ration of either one or .two vacuum breakers is 9.4 psi which is equivalent to the statistically derived P 90/90 value.
It cani therefor ei be concluded that the approach used in the deve looment of a design pressure amplitude for WNP-2 is indeed a justified and conservative approach since the P 90/90 pressure amplitude used in the determination of a WNP"2 design pressure amplitude is not exceeded by any pressure amplitude measured during the Caorso tests. For further discussioni see the response to Question 022.054.
)
WNP-2 022.058 In order to account for the differences betw een the W NP-2 des ign conditions and the C'aorso test condit ions (e.c the pool geonetryi the number of SRVs and th e in i t'i a l pooL tern perature) r. you used a pressure amplitude mul t ipLie r based on a correlation.in the Desi'gn Forcing Funct ion Repor t (DiFR) to ob.ain the WNP"2 design values for SRV Lo ads. Thi s pro cedure involves the extrapolation of pres sure ampl 1 tud D s rea sured in the Caorso .ests wi.h respec. to som e par amet er va ues (e.g.i .he SRV stean flow rate) .o WN P-2 desig L
con ditions using the trepds established in t he D iiR.
Acc ordinglyr p'rovide justification -or your pos i tion tnat the trends used in this extrapolation can be sup porte d by ava i lable Caorso data.
Response
The DFFR,established trends for pressure amplitude variation with Gi ferent parameters such as steam f Low ratei initiaL pool .empera urer etc.r are supported by avaiLable Caorso
~
data.
As can be seen ,rom Ficure 022.058-1 the DFFR predicted pressure amplitudes'increase with steam flow rate. Caorso Phase II Tests n35 and ~38 were selected, since all tes.
conditionsr other than steam flow rater were maintained approximately equaL during these tests:
p p predicted = '1.45 psi measured = 1.5 psi (DFFR)
Similar lyi'ioure 022.058-2 illustrates tha the DFFR pre- ~
dieted press'ure amplitude increases with init ia l pool temper" ature. This is also verified by available Caorso data Caorso Phase I Tests ~7 and ~1301 were selected in thi s case since alt. conditionsr other than the initial pool temp eraturer were maintained approx inate ly equal during these tests p
predicted = 0.4 psi predicted = 0.5 psi (DFFR) (Caorso Phase I SYA)
It is concluded from the abov examples that DFFR esta blushed trends are supported by available Caorso data and coul d be used to account for the differences between ~ he WNP-2 design conditions and the Caorso test conditions.
MNP,-2 Q. 022.059 The proposed ve rticaL pressure dist ribut ion in the SRV report cited constant 'between th e bottom of the suppressi on and the qu encher el eva.ion and 'then decreases above'ool Linearly to zero at the poo L surf ace. You stat ed that you derived this particular s pat ial variation by rev iewing the maximum p.essures measured at var ious e leva ions in t he Caorso tests.
~ Howevers as shown in F1g ure 38a of the c i te d reports this proposed pressure dis tr i but i on cannot bound he"naxinum measured values
~
of pressure Tol alL Caorso tests. Furthermorer the use of the ni ax 1 mi ui.l mi e a 5 ure d pressure val's in the ef fect o c b ubble vertical motio n on the measured pressure the compar-',son canno reveal
~
distribution Our concerns are tha t .he bubble vertical motion wi L L result 1 il a more severe pressu re distribution,'n the Later par of the SRV .ransient and your model may not yield the
~
correct pres sur e dis,ribution. Spe c ifi ca l lyr the c r oss-corre-La ion coef f 1C ent o
~ 1 . pressure .rac es measured at di f f erent ei.evazions i s L ess 'than 1 (about 0. 9) which indi cat es that there may be so me effect from bubbL e mot ion on the pressure di st r ibut ion According lyr modi, y your proposed vertical pressure dis t r'i butioni as requiredr .o assure conservatism in the desig n L oad specif icati.ons or SPV transients in the WNP-2 facili ty.
Response
- 1. Proposed Vertical Pressure Distribution Figure 3.8a of the SRV report may more appropria tely be shown, as in Figure 022.059-1 such that the measu red pressures are normalized to the maximum value of pressure sensor P19 instead of P13r as was used in he or iginaL ~
SRV Load de inition reports since P19 correspond s to the sensor selected in he development of the Load d ef in i t i on.
~
The proposed vertical distribution is compared a ga inst the average and maximun values o- pressure measu red at Tive loca.ions during six representative tests Tes.s 4402r 2325r 2324r 2305'202'nd 1104 were chos en as representative since these tests measured, some o f the highest pressures at P19.
This alterna.e. figure shows the proposed vertica L distri-bution as adequate.. The pressure dis ribution w hi ch would ~
envelope the maxinum values shown in the f igure 1S adequa.ely represented by the proposed pressure d1strlbut1one
WNP-2
- 2. Bubble Vertical Notion The 'suggested "bubble verti cal motion" wiLL not result in a more severe pressure dis ~ ribution in .he .Latter par ~
of the transient..Xt can be shown'sing Test 2202 as an examples that the vertical pressure distribution is preserved. Figure 022.059-2,compares the normalized first peak positive and negative naximum pressure anpli.udes of pressure sensors Located along the ver ticaL face of the containnent with the proposed vertical pressure distr ibu" tionr with satisfactory results. The second positive and necative peak amplitudes as well as the third positive and nega ive peak amplitudes for he applicable pressure sensors
~ ~
plotted in Figures 022.059-3 and 022.059-4r respectivelyr again show the distribution as adequate.
~
~
WNP -2 Q. 022.061 You use the methodology in the OFFR to establish the c ircum-ferential pressure distribution for the WNP-2 facility.
State whether you use the Line-of-sight and square-root"of-the-sum-of-the"squares (SRSS) assumptions of the OFFR in your calculations. Ef sos provide justification for using these assumptions in the WNP-2 facility. Your justification shouLd be based on the Caorso test results. indicate to what extent these assumptions affect the WNP"2 Load'ases.
Provide representative f igures showing the pressure distri-butions on the basematr the pedestal waLL and the containment waLL for the various SRV discharge cases considered in WNP-2 plant design assessment.
Response
circumferential pressure e distribution actuaLly used for The the WNP-2 facility is derived from a conservative and more realistic application of the pressure distribution on the suppression pool boundary recommended in DFFR (Reference I).
This conservative application is supported by Caorso test data as described b'elow.
a ~ The Line"of"sight assumption is not used for WNP-2. To calculate the boundary pressure at a. given Location resulting -from actuation of a-single--quencher the "straight Line" distance is used instead of the "Line-of-sight" di stance recommended in FFR: D
- 2. 2 P
B o for .ro < 1.0 r r p(r) 2 for o 1.0 where:
p(r) attenuated bubble pressure;
'B bubble pressure; r0 quencher radius; r straight Line distance from quencher center point to the location of interest
MNP-2 This results in finite pressure values being calculated over the entire suppression pooL boundaryi not only over the "viewed" portion of the boundaryr a fact verified by Caorso test data. Indeedy during Caorso Phase II Test 501 X (see Reference 2) quencher "V" was actuated and the available instrumentationi although Lo" ated in a shaded or "non-viewed" boundary area with respect to the actuated quencher as seen from Figures 4 1r 4 2 and 4 3 of Reference 2i recorded finite boundary pressure values as follows:
Sensor 13 14 15 17 18 Recorded pressure valuer (psi) 0.4 '.4 0.5 0.5 0.7 The bubble pressure value was not recorded during, this specif ic test but can be estimated to Lie in the range of vaLues recorded during Phase I Tests 19 and 20> and Phase "U" at II Test 22 A01 performed with quencher similar conditions: single val've first actuationsr cold piper normaL water leg and two 10-inch vacuum breakers (see References 2 and 3) . Then'he pre'ssure value recorded at Location of sensor P1 r Located approximately 180 from the actuated quencher "V 'r is estimated to be in the= range of 6.4% to 16.1% of the bubble pressure.
This is comparable to the 13% prediction calculated using the "straight line" distance assumption.
- b. The SRSS assumption is replaced with the more conservative Linear superposition (LS) assumption for the MNP-2 facility.
In -the case of two quenchers this LS rule becomes:
P = P1 + P2 <.PB where:
P = total pressure at the Location of interest; P1iP2 =,contributions from .quencher 01 and P2r respectiv'elyr calculated using eq. (1) r above.
Justification for the use of the LS assumption is provided by two Caorso Phase II two-valve tests: Test 24 and Test 25 (see Reference 2) . In- Table 022.061-1 attachedi pressures calculated at'different pool boundary Locations using'the LS rule and the .SRSS rule are identifi'ed. For pressures recorded at the same Locations during comparison'urposes the two tests are also listed. From examination of the data it is concluded thatthanthetheLS SRSS rule is adequate ands as expectedr rule.
more conservative
W NP.-2 Fi gur es 022.061 "1 and 022.061-2 i l lust r at e the pressure distributions along the suppression pooL f Loor and on the pedestal and containment waLLs for two SRV discharge the all-valves case and single outer valve These two cases'ase.
cases were determined to be the governing cases for the WNP-2 plant designr although the plant was also assessed for a single inner valve discharge caser a two vaLve discharge case and for the ADS case.
Figure 022.061-3 iLLustrates the circumferentiaL distribution of pressure Loading for the single outer valve case.
should be noted that there is no pressure variation in the It circumferentiaL direction for the all valves discharge case.
0
.~e "e rences
~ ~
I Ha rk ZZ Con" ain~ent Dyna'ic .o ~cun ng c tions Info~ation Report
¹v.
D~R) ~ ~ bFDO-24 063, 3 dated June 1978.
l I
- 2. ~ Yark ZX" Contain.=ent Super Prog aa. CAOPDO Sa Relief
'Valve .Discharge iests. Pha 'es t( 7-P, da te ety'e ZZ Repo H D Hay 5980, General Electric Company Proprietary.
- 3. "~TORSO SRV Discharge Tests. Phase T. Test -.=port," 8=-DB -25100-Pg dated N*y )979< Gene al =-1ectri c Co-panv ?=oprietary.
~ ~
~ t
~ ~
I I 0 ~
~ ~
~ ~
0 4 t
I 0 I 1 f ~ ~
0 4 0
0 0 E ~
0
h 0 0 e
I r
~ ~
~ o I ~
~ ~ ~ e P
~ 0 f
Q 0 379 ps i 0
gO 0.250 psi r Ck S4 e4 0 '54 psst.
~
+Y CD A O
/ O
\ '
0.604 ps-"
3.80 0 c~
~ ~
V 0 ~ 0 pS>
0 S~2 ps' 1
22aO PZGiSP= Q 022.06't-3 S:.IGL:- OU-'HR VALV>>- CASE ~ r
~ ~
WNP-2 Q. 022.062 Indicate what two valves are selected for the two-valve discharge case. State whether the two quenchers selected are in the inner or the outer circle. Provide justifi-cation for the two valves seLected and for their Location.
Response
The two valves selected for the two"valve discharge case ar id ntified as 1-B and 3-A; one quencher lies in the inner circLe while the other Lies in the outer circ le.
There are a Large number of cases to be consideredi howevers certain DFFR criteria Limitthis selection. .One such criterion specifies that the required combination inc lude one Lowset valve and any adjacent va l.ve. This statement Limits the number of cases to be considered to seven. Of these seven tests analyzedr the arrangement chosen yields the most criticaL asymmetric Loading condition thereby establishing the basis for the seLection. Refer to Figure 022.062-1 for the Location of the two seLected-quenchers.
C ~ ~
COhTA'h; -'A lc.
co- ',"q (m ~p )
0 0
0 0 Ri j
0 4-bi PLAN "ZGUB= Q 022.062-1
<MM. NC OH PUBLIC PO4=+ QUPPr ~ guy .~ t-u NCHER LCcATION FOR TWO v'LvE
'.PJCi AR P PZPZ:-C ~. 2 DISCHARGE. gPSE
0- 1
~
- t
WNP-2 Q. 022.063 in your analyses of SRV transients in the WNP-2 facilityr you assume that the pool water is incompressible. Your only justi-fication for this assumption is that the cross-correlation co-efficients between pressure time histories measured at different Locations are high. Our concern is that this is insufficient justification since the relationship between the cross"correlation coefficient and the time phase shift has not been established; this relationship will inf Luence the ef ect of comoressibility on pressure measurements. You use the in-compressibLe flow assumption in your analyses addressing the f Luid-structure interaction (FSI) ef feet and in the WNP-2 structuraL anaLyses. Even though the incompressible f low assumption can be just.ified .for the Caorso plantr it is stiLL questionabLe whether it hoLds true for the WNP-2 faciLity.
Specificallyr our concern is that the fluid-structure coupling ef feet may be more significant in the WNP-2 plant which has a steeL containment than in the Caorso faciLity which has a concrete containment. -
Further'he velocity of sound in water is greatly reduced by the presence of air and steam bubbLes in the water; the conditions in the WNP-2 facility may differ to the extent that the amount of air and steam bubbLes in the pooL water wiLL be significantLy different for the two facilities. Accordinglyi since your assumption regarding the incompressibility of water in your analyses of SRV transients in the WNP-2 faciLity is important but not a'dequately supportedi provide additiona l justification on this matter.
Response
For any boundary pressurei Pi it can be shown that P = P. + P a
(See p. 32'RV report) where p.1 rigid p
waLL pressure a interaction pressure due to waL L flexibiLity Na in which N = hydrodynamic added mass a = walL acceleration lt was found during the Caorso tests that con tainment waLL acceleration measurements were very smaLL (se e Appendix 5-1r SRV report). One may conclude that the interacti on pressurer P a r is to be considered negligible for the .Caorso fa c ility so that
A WNP-2 The SRV Load definition was then based on the rigid walL pressures recorded at the Caorso facility.
In the design ass ssment phas of the WNP-2 f ac i L i tyr the rigid waLL pressurer P .r was applied at the pool boundary and since the acceleration of the containnen t waLL may no Longer be con-
'sider ed negligible. (WNP2r ste L c ontainment versus Caorsoi concrete containment)i the interac tion pressures P r is completeLy accounted for my a set of hydrodyn amic added masses a using the approach specified in Refer nc 1.
The high cross-correlation coeffic ient between pressure-tine histories measured at different Lo cations in the Caorso faciLity is merely a reinforcement of the f luid incompress ibi L i ty assumption for the Caorso plant.
0 WNP -2
REFERENCES:
- 1. Bedrosiani B.r "Analysis of a Nark EI Con ta inmen. S t rue tur e for Hydrodynamic Loads in Suporession Poo L" r P roc ceding sr Conference on StructuraL Analysis~ Design and Construe ion in NucLear Power PLantsi Vol. 2r Porto AL egrar Brazil>
Apr i l 1978.
Ev+' 4 4 4
4