Text

Nonproprietary Mods to Critical Flow Model in RELAP5YA
	ML20148G543
Person / Time
Site:	Yankee Rowe, Maine Yankee, 05000000
Issue date:	01/15/1988
From:	; YANKEE ATOMIC ELECTRIC CO.
To:	;
Shared Package
ML19341D824	List: ML20148G458; ML20148G543;
References
	NUDOCS 8801260538
	Download: ML20148G543 (158)
	v • d • e

{{#Wiki_filter:w c qe [- ( ATTACHMENT A { MODIFICATIONS TO THE CRITICAL FLOW MODEL IN ( RELAPSYA l r i "!R298ai 8*eM8 ~ c 2, P PDR ............J

s-s W I Modifications to the Critical Flow Model in RELAP5YA ABSTRACT The RELAP5YA program contains two options for the calculation of critical (choked) flows--the RELAPS/ MODI model and the Moody model. In code assessment work, the RELAPS/ MODI model is generally used because it is expected to provide o more realistic estimate of critical flow rates. However, it has been found that when the area at the choking location is less than the upstream flow area, the model overpredicts critical flows at high steam quality, by approximately 20% at near single phase vapor conditions. This report presents a modification to the critical flow model that corrects this discrepancy. 1 ( ( A-1 l

1. INTRODUCIION - In LOCA analyses, the calculation of mass and energy release rates at the break location is of primary importance. The flow at the break is usually choked (critical) for the entire duration of postulated accident sequences. The RELAP5YA h computer program (Reference 1) contains two options for the calculation of critical flows: (a) the RELAP5/ MODI critical flow model (References 2, 3) and (b) the ( Moody critical flow model (Reference 4). In plant licensing calculations, the Moody model is used as dictated by regulatory requirements. In these applications, the intent is to maximize the break discharge flow for a given break flow area. Mcwever, in assessing code capabilities of modeling LOCA scenarios against appropriate cxperimental data, the intent is to provide a "best-estimate" calculation of thermal-hydraulic phenomena. The RELAP5/M)D1 critical flow model has been assessed against a variety of separate effects experiments (References 1, 2, 6 3) and found to be reasonably accurate. Hence, in RELAPSYA assessment against experiments, the general practice has been to use the RELAP5/ MODI critical flow model to calculate the break discharge flow rates. The assessment of the RELAPS/ MODI critical flow model (References 1, 2, & 3) has attempted to address the full range of fluid conditions that can be expected et the break location in postulated LOCAs. However, most of the attention was focussed on the case of subcooled liquid or low quality two phase flows. It his been found that at higher qualities (typically in the range of 50% to 100%), f when the break flow area is smaller than the upstream flow area, the RELAP/ MODI nodel yields critical flow rates higher than the Moody model. A.t near single phase ( vcpor conditions, the RELAP5/ MODI critical flow has been found to be about 20% higher than the Moody critical flow calculation. At these conditions, the Meody codel yields realistic estimates of critical flow. Hence, it is believed that the RELAPS/ MOD 1 model overpredicts critical flow at high qualities. A review of the RELAPS/ MODI critical flow model has indicated an error in the estimation of sonic velocities at flow paths where the area of the flow [ path is smaller than the upstream flow area. The error is in the calculation { A-2 1

of local fluid thermodynamic properties (Pressure, Internal Energy) needed in the calculation of sonic velocity. Section 2 of this report provides details of the review and identifies the problem more clearly. In Section 3, a modification is proposed that corrects the error in the RELAPS/ MOD 1 model. Sample calculations and results are also discussed in Section 3. l ( [ i i I A-3 9

J n 2. BACKGROUND Critical flow (or choked flow) is defined as the condition where the flow rote becomes independent of the downstream conditions. It represents a condition of maximum flow rate for a given set of upstream conditions and flow geometry. The reason for choking is that acoustic signals (which transmit information en fluid pressure) no longer propagate upstream. This occurs when the fluid vslocity is equal to or greater than the sonic velocity. ( One method of calculating critical flow is to numerically solve the conservation equations (mass, momentum, energy) in the vicinity of the choking location. { This method requires that the conservation equations adequately describe the physics of critical flow. The method also requires the use of a fine calculational tesh for numerical solution to handle the relatively steep gradients of pressure end flow near the choking location. One requirement of this method is that in two phase flow, the conservation equations must include an adequate description of interphase transport of mass, momentum and energy, which are not easily modeled. In addition, the computational cost of this method can be high due to the fine nodalization required. { An alternate method, developed by Ransom and Trapp (Reference 3), eliminates the need for fine nodalization near the choking location and has been demonstrated to provide a reasonably accurate calculation of critical flow. Hence, this aethod was implemented in RELAP5/ MOD 1 (Reference 2). Ransom and Trapp (Reference

3) studied signal propagation in a two phase medium and used the, method of characteristics to solve the basic conservation equations. For the case of non-homogeneous, cquilibrium two phase flows, they were able to derive an approximate choking criterion (Reference 2) ass (2-1) g) + (a /p )) - a

((a /c )v + (a /o )v 1/I(a /p f f gg g g g f f f g where a is the homogenous equilibrium sound speed. HE A-4 {

J s In RELAPS/ MODI, the above equation is used to check if chokin' exists. \\ If so, the above equation is solved simultaneously with the momentu.r. difference g cquation for the phase velocities at the choking location. (The momentum difference L equation is obtained by taking the difference between the vapor and liquid phase comentum equations). [ The use of this method requires knowledge of the homogeneous equilibrium { sound speed, aHE, at the choking location. It could be determined from thermodynamic properties at the choked location. However, these properties are not available cs a part of the normal hydrodynamic solution of RELAPS/ MOD 1 and must be estimated. In RELAP5/ MODI, the pressure and internal energy at the choked location are dstermined by considering the conservation of mechanical energy and total energy j b2 tween the choked location and an upstream location. These properties are sufficient to determine other needed thermodynamic properties in the estimation of the sound speed. However, the method does not account for compressibility offects. Hence when the area at the choking location is smaller than the upstream flow area, as the fluid becomes more compressible (higher quality) the error is larger and the pressure at the choking plane is overestimated. It is believed that this is the primary reason for the overprediction of critical flow at higher qualities by RELAPS/ MOD 1. ( l A-5 (

J 3. PROPOSED MODIFICATIONS TO THE CRITICAL FLOW MODEL a j The thrust of the proposed modifications to the RELAP5/ MODI critical flow L model is to improve the estimation of the homogeneous equilibrium sound speed et the choked location. This is carried out by estimating the pressure and internal energy at the choked location by conserving mechanical and thermal cnergy similar to the RELAPS/ MOD 1 method, except compressibility effects are included. The result is that the pressure at the choked location is lower than { that pred!cted by the RELAPS/ MOD 1 method (at higher qualities) and the sound speed is correspondingly decreased. The method first determines pressure and internal energy at the choked location by a "homogeneous-frozen" type of model. In this model, the main assumptions cre that (a) interphase heat and mass transfer are neglected, (b) the vapor phase behaves as a perfect gas and (c) the process is adiabatic with the vapor phase obeying the relationship: (P/p )T = constant g For single phase vapor conditions, the pressure and internal energy frvm this model are used directly to compute the sound speed. However, for two phase upstream conditions, direct use of this model by itself will underpredict the pressure at the choked location. Hence, for upstream qualities between 0.2 f cod 1.0, the pressure and internal energy are determined by quality-weighting between this model and the RELAP5/ MOD 1 model. For qualities below 0.2, the { RELAP5/ MOD 1 is used because previous assessment (References 1, 2, & 3) has shown good comparison with data in this range of qualities. A sample calculation has been carried out to test the modified critical flow model. Figure 3-1 describes the problem set-up. In this problem, the upstream pressure was held constant at 7 MPa and the upstream quality varied in steps from 1.0 to 0.0. At each upstream quality, steady-state conditions were obtained and the critical mass flux at the exit junction recorded. The problem was executed with (a) the RELAPS/ MODI critical flow model (b) the Moody model and (c) the modified critical flow model. The results are shown as a A,6

plot of mass flux versus quality in Figure 3-2. Also shown is the mass flux cs would be calculated by the homogeneous-frozen model. As can be seen, the RELAP5/ MODI model yields higher critical flow rates than the Moody model for qualities higher than 0.5, with a maximum difference of about 20% at single Phase vapor conditions. The modified critical flow model provides lower flows than the RELAPS/ MOD 1 model for qualities higher than 0.2 and also consistently Icwor than the Moody model. At single phase vapor conditions, the new model opproaches the Moody model, which is more realistic than the RELAP5/ MOD 1 model. ( ( l A-7 I

s W 4 ,e= = . o. T? [ a os Eo W k N' ii.i o,' go g [a 2O s s#3tS& y a134o"2 K hN 2R u4 e <=*,3or aa 5 g a ( = y 2 s v 2 N k' 1I IIok E bE d o o o 30 +44 g6 ,,.,.,si.g e og e, 2 T E k w s N'M pgg I:

r 2

dJ 0 6 o k h ~o wt t nn O 8

d.,dac*=4 A

$8 + ~9 2 [ 5 284 mi x a Il el ll 5 T4 A \\ A-8

.___.__y._____. _...._...t.. flG. 3-2 : RLS U LTS HCM Tesr PRO B Li t=t ...__S._O g j ... _.. g _ _ _. .. \\. _ _. l. _1_Z __~~_- Z ~1 ~ - - _ _ j- ~ - _Z.f._^ ~ - ~ ~ __. j i i ._-_._l 3 --Al+s Te t Att --htiss utt :- 1 +1hn. - _.. _ _ _ _ _a n ..____.__4__.._ . 5 y ..M_cx.__R.n.r..r e_._ _._. _._ _ _ ~ y" O_+ _____._p_.__.___. 4 g . _ _ _. y _ __1_. s ____g___.. ._____..j. C h 1 1 g j__ ' _. _' 3;1-j t 3 _\\ d, q _.} .] g .ag J. I ._j f m .g _ ..s _.. 3 ,__ g l +, _1 .C. _j

_\\. 9f cc o v l

20 1- \\ j._ j s 3 ._- g ~. _.... _.., \\ ' ; ~ .N_., ' s y Rc!c5 0 ?fLTHcD. P _ _..... _ _ _ N, ' s R 3. 4 3 .s y _ ~~ \\ ~.~f T T._. .)-- RELA 95]MO i { y_ j..._. _.. _ _ _ _ _ _.. _.%...g. J .j ._a 4 g {

g.

w +. g N h W q .Js s.4 N .__..]. ____..q.. -i. - N,s -i l 44J g _ f. 0 -Hen. Rog gn-- - s *. '--n ~ '-.% E us i:.4 u. g _ _ _.. _____ a ____..____] ..____.___.2 ____..___.4 -________..___;_.________j 4 00 _ Z - ~ ~~~ T ~~T i ~' _. ' l. l l ._l .[ ~3 T ~ . - 0.4. j

l. o j

-00 _._...__.3 1 . ),H TMLA M _ d rA rte A stA L I T Y ---> l A-9 i

4. CONCLUSIONS i A method has been developed to improve the prediction of critical (choked) flow in the RELAP5YA computer program. The previous method was based on the RELAP5/ MODI critical flow model which tended to overpredict critical flow rates et high qualities wnen the fJow area at the choking location was lower than the upstream flow area. This was due to the omission of compressibility effects in estimating sound speed at the choking location. The new method attempts to correct this deficiency by including the effects of fluid compressibility. A quantitative assessment of the new method against appropriate critical flow dsta is yet to be performed. However, results of sample calculations indicate that the new method would provide a more realistic estimation of flows over a wide range of upstream fluid conditions. ( 0 A-10

,s J

REFERENCES:

1. Fernandez, R. T., et. al., "RELAP5YA-A Computer Program for LWR System Thermal Hydraulic Analysis," Vol. I and II, YAEC-1300P, December 1982. 2. Ransom, V. H., et. al., "RELAPS/ MOD 1 Code Manual," Vol. 1 and 2, NUREG/CR-1826, March 1982. 3. Ransom, V. H. and Trapp, J. A., "The RELAPS Choked Flow Hodel and Application / to a Large Scale Flow Test," Proceedings of the ANS/ASME/NRC International Topical Meeting on Nuclear Reactor Thermal-Hydraulics, Saratoga Springs, New York, October 5-8, 1980, pp. 799-819. 4. Moody, F. J., "Maximum Flow Rate of a Single Component Two-Phase Mixture," Journal of Heat Transfer, Trans. ASME, 87,1, February,1965. I ( l 1 i A-ll

) F ATTACHMENT B t RELAPSYA CRITICAL FLOW MODEL: ADDITIONAL ASSESSMENT l l I \\

RELAP5YA CRITICAL FLOW H0 DEL: ADDITIONAL ASSESSMENT In Attachment A, we presented a sample steady-state calculation test problem showing results as executed with the RELAPS/ MODI critical flow model, the Hoody W model, and the modified critical flow model. Those results showed the tendency cf the RELAPS/ MOD 1 model to overestimate the break flow in comparison to expected results as derived by the Moody model. The improvement of the revised model was evident for the selected steady-state conditions. To provide further assessment of the modified critical flow model, we have performed studies using the General Electric Level Swell Test. The test selected for the assessment of RELAPSYA and the results of that work are presented in Volume III of "RELAPSYA A Computer Program for Light Water Reactor System Thermal-Hydraulic Analysis", October 1982. Figure 1 shows the major elements of the test configuration. Figure 2 presents the RELAP5YA model for the GE Top Blowdown Test 1004-3 and reviews some key features of the modeling which cre pertinent here. The previous results showed that excellent agreement for the vessel pressure history was obtained when a contraction coefficient of 0.55 was used at the break orifice in the blowdown line. One reason for the use of this value was that the vessel wall was not modeled as a passive heat source in the previous RELAP5YA representation for this test. The current RELAPSYA representation of this test (Figure 2) includes the vessel wall heat structure. Based on our ( racent work, we conclude that the influence of the RELAP5/ MODI critical flow model also contributed to the underprediciton of system pressure. Figure 3 shows a comparison between the pressure response as seen in the test and the RELAP5YA predictions using the original and the revised critical flow model. In each prediction the break orifice discharge coefficient was set equal to 1.0. The improvement of the pressure response prediction when the revised model is employed is evident. B-; s

J j Small blowdown ve s s el bb L O O ,g ,l Vent oice 1:.0 f Blowdown ofilice 11.0 h 10 f t.- 9.0 _pm P) h s liquid 5 8 f t.- 7.0 kP) \\ 6 f t.- D .0 B wdown O x e m-( = 3.0 P) 2 f t.- 1.0 0.51 j O f t.- i f t.- Pressure vessel Suppees sion pool FIGURE 1 GE Level Swell Test Configuration B-2 1 L

d FIGURE 2 )} Y R E L A P 51A MDblL 0F GrE TDP blo6@ ceed '76$T L [ rv fio i > TRPVLV fc5' l L tit.D NhY sy if6 0 ST k.A $5V300 ~ 150 ? I y (CO. ~O l%.0 .y y 00 -)) 121 2 (-.t- ~

t 0. o e

Jg - ICQ.0 i --j-o 7 4 -9 I Comments: 96.o 9 2 (1) TV 910 is set at Initial System Pressure _g of 1011 psia for initialization. n., k 7- -i - (2) TV 900 is set at atmospheric pressure. (3) For 0<t<10, TRPVLV 275 is closed ~' and '5PVLV 905 is open. This is the k 12.o initialization period. D j _g (4) For t>10, TRPVLV 905 is closed and 5 TR?VLV 275 is open. This is the t 4-Sc.o ,d-transient calculation. /,.g (5) TRPVLV 275 represents the blowdown 48.o ) orifice. A discharge coefficient can be chosen for this orifice to match n.9 _9 the system pressure transient during 36.o 5 blowdown. The value will also depend 7 4-on whether the vessel wall is $ -3 modeled as a passive heat slab. d (6) The wall is 0.688" thick, Carbon 24 0 U~" Steel. / l

1

1' doo

/ -g EL 0 0 fy B-3 o

) l ~ T~~' l l i i i ~~

I I

. '~l'9 i i ~~~! FIGURE 3 i f2CR'

r i

i f l t l l l 8 ' $E _ hlOP _N0Wbb ik {5 7_. fDDk

b '

\\ \\ I 'a i iii i t i s t I i l ! $I l i 1 I l l-kh I f I ! j_,_jj _j __ j. l. ' 1 I 1 f I 8 t i ( I l. .J J J i I L i f00A.,i i, i l . g. _.... _ _._ _ __g. _. .. q L.!_!_ _i _ _ _4 7. g. .~ k _ .a_.. Y -

__._q m ;......

_ yi L 800 _5_J 4 _ ~. .-'... ~ ..s_ i I i i i i i -j ).. _ L._; . __ 4 ..!. q _. ___. -

7. q i

i _,..... _ _ _ _.,.. t..- (. y. ._4 i j l f ,l '[ J__+ 7;_. _i i _j ,,h -- l, j-

_ s.--

l. 2 NT-- - d E , i. i-i \\ i 7 7--kk-i--!- ; - ggg l l ' % % " /,-- t l l H 0 1 .Ny%y d i_ _ i 6-t i i o i i \\1 RetApsrA Worn hen 400- > \\ \\ .._j.C4mcAtil10w_No.hEt I _.l !. '.!...[ g I I % !_Qx14 _i__.____'__!__. 1 ._. 4 _. ._ Q l ..i_._l __.I I . ;J_.4 bJo..l'm, !_L L;. .. }....j. _ q 4 ;.. j... ..p i t i i i e N d '2 l .I I i I i 1- 'I iNNm% J 'i i 7l 200_ 1 6 i i L l I \\" l,I +l i i i l ~ 1 i i i i j ) 2 i I e i I l i j i i i i l d i I 2 i i } i l l ! I i l i } I k i I i i i i l '+ ! _.I I I r 1 1 O l l 'I j l I ) l l l O 40 90 (20 160 l i TlHE, sac --> B-4

s J L r L [ V ATTACHMENT D ASSESSMENT OF RELAPSYA AGAINST LOFT l MUCLEAR EXPERIMENT L5-1 I

.= y e. ~'; -:.y y &, 4, yllaQ::, [ ~ s ASSESSMENT OF RELAP5YA AGAINST LOFT NUCLEAR EXPERIMENT L5-1 F g- = E ' l.0 Introduction 1 \\= GY I L Two Loss of Fluid Test (LorT) Small Break tests were simulated hE r h using the RELAP5YA computer code as part of the integral code , gg.g h-f. f,+ ' assessment program for licensing analysis of PWR SBLOCAs. The r ~..;.. a.g tests were LOFT L3-6 and LOFT L3-1. The assessment results were

;j'iyt

{ presented in References 1 ar.d 2. While both tests exhibited many 9.d. L[ of the major phenomena in a PWR small break accident and hence, NW Bihaggann provided a wide data base for code assessment, no core 1.g uncovery/ recovery was encountered. The LorT Test L5-1 was chosen C,p ;.N, a s y to supplement the integral code assessment because it provides tf these phenomena.; .\\: -l_' In the assessment presented, two calculations have been carried out. The first is a base case best estimate calculation with the subcooled and two phase discharge coefficients equal to 1.0, while the second is a sensitivity on the discharge coefficient at the break location. 2.0 racility Description The Loss of Fluid Test (LOFT) facility at the Idaho National Engineering Laboratory is a 55 MW(t), volumetrically scaled pressurizer water reactor (PWR) used to simulate the thermal-hydraulic responses expected in a commercial reactor during postulated Loss of Coolant Accidents (LOCAs). The facility includes five major systems: the reactor vessel, the intact loop, the broken loop, the emergency core cooling system (ECCS), and the blowdown suppression tank. The LOFT major f-components are shown in rigure 2.1 D-1 ]

s u J The LorT reactor vessel has an annular downconer, a lower plenum, [ lower core support platec, a nuclear core and an upper plenum. The downconer is connected to the cold legs of the intact and broken loop. The upper plenum is connected to the hot legs of the intact and broken loop. The core contains 1300 unpressurized fuel rods. The fuel rods have an active length of 1.67a and an outside diameter of 10.72mm. The intact loop simulates three loops of a commercial, four loop PWR and contains a steam generator, two primary coolant pumps in parallel, a pressurizer and connecting piping. The broken loop consists of a hot leg and a cold leg which are connected to the reactor vessel and the blowdown suppression tank header. Each leg consists of a break plane orifice, a quick-opto.ng blowdown valve, a recirculation line, an isolation valve and connecting piping. The broken loop hot leg also contains a simulated steam generator and a simulated pump. For the LS-1 test, the steam generator and pump simulator were not attached to the broken loop, but were replaced by a blind flange. The LOTT ECCS simulates the ECCS of a commercial PWR. It consists of two accumulators, a high-pressure injection system and a low pressure injection system. The blowdown suppression tank (BST) contains the blowdown effluent as a temporary storage volume and provides a simulated containment back pressure. A more complete description of the LorT facility is provided in Reference 3. 3.0 RELAP5YA Model for Experiment L5-1 The RELAPSYA input used to model the LorT Experiment L5-1 is based D-2 \\

J on the EG&G RELAP5/Modl deck used to simulate the experiment s LP-SB-3 (Reference 4). The original nodalization is presented in rigure 3.1. Various changes to the input were necessary in order to model the Experiment L5-1. The final nodalization used to simulate the test is presented in rigure 3.2. The changes are described in this section. The blowdown valve was modified to reflect the geometry specific to the test L5-1. In order to represent the core power, the reactor kinetics option was used. The kinetics parameters were obtained from the data presented in the LorT System description (Reference 3) and The Experimental Data Report (Reference 5). Heat slabs representing the hottest pin were added to the core volumes to allow calculation of the peak ( cladding temperature. Based on data presented in Reference 5, the hot rod was chosen to be rod SM3. The downcomer was modeled as two parallel annulus components. One component connects to the broken loop and the other connects to the intact loop. The top of the two annulus components are connected by a single junction. The ECC system was modeled such that it reflects the test data. The HPSI and LPSI pumps are modeled as time dependent junctions. The accumulator and the injection line were remodeled using the information received from EG&G in Reference 6. The ECCS injection line was directed into the intact loop cold leg as per the test configuration. Actual temperature was used for the ECCS i

water, D-3 a

J Main Steam Valve (MSV): From the L5-1 Experimental Data Report (Reference 5), the MSV starts to close at 0.17 r seconds and is closed at 12.1 seconds. Since the valve L did not act as a relief valve in this experiment, but ( simply closed following scram, it was adequate to model the MSV as a motor valve which starts to close at 0.17 seconds at a rate that will result in it closing by 12.1 seconds. No leakage through the MSV was modeled. The volumes 502 and 500 in the steam generator secondary in the EG&G nodalization (figure 3.1) were combined into a new volume 500. Modeling the steam generator with the nodalization shown in rigure 3.2 was considerably more stable in null-transient conditions than the base nodalization shown in rigure 3.1. The reactor coolant pump flywheel system with its variable inertia was modeled by inputting the pump speed versus time from Reference 5 instead of through a code modification Initial conditions were entered to match the measured initial conditions of the L5-1 test. Initial conditions are given in Table 1 together with the corresponding RELAP5YA input values used for performing the Base Case ) calculation and the Sensitivity Study. 1 4.0 LOFT Experiment L5-1 i The Experiment L5-1 was conducted on September 24, 1981. The break location for this experiment was in the broken loop cold leg. The experiment simulated a shear break in one of four Emergency Core Cooling System (ECCS) injection pipes [28.5cm (11.2 in.) inside diameter] in a commercial four loop PWR. The 4 Experiment L5-1 was designed to provide data to evaluate the effectiveness of degraded ECC System performance (only three of D-4

four accumulators available) during an intermediate size break LOCA. The accumulator pressure was changed from a nominal f setpoint of 4.22 MPa (600 psig) to 1.66 MPa (228 psig). This pressure setting is used in some Combustion Engine 3 ring PWR ( designs, which is the limiting condition for worst case ECC System operation. The liquid volume of a single LOFT facility [ accumulator was scaled to represent three of four accumulators available for a large commercial PWR. 4.1 Experimental Procedure-Chronology of Events The reactor was operated at full power for approximately 36 hours prior to the start of experiment. The experiment was initiated by opening the cold leg QOBV. A flow limiting nozzle 2 (1.73 F.-3 m ) provided a break mass flow scaled to a ruptured 14-in. accumulator line in a commercial PWR. The reactor scrammed on low hot leg pressure (14.19 MPa) at 0.17 seconds. (break initiation is defined as "time zero") When the control rods were fully inserted, the primary coolant pumps (RCPs) were manually tripped (4.0 sec.). The pumps then coasted down under the f influence of the flywheel system. Coastdown was completed at 19.3 sec., when the flywheels were decoupled from the RCPs. HPIS "A" injection, initiated by a hot leg pressure setpoint of 10.6 Mpa, commenced at 2.88 sec. The main feed pumps tripped and the steam generator steam control valve commenced to ramp shut on receipt of the reactor scram signal; the main steam flow control valve was fully closed at 12.1 seconds. Fluid saturation in the upper plenum occurred at 0.2 seconds and voiding began in the broken loop cold leg (BLCL) at 10.5 seconds. The pressurizer emptied at 15.5 seconds. The primary system pressure dropped below the secondary pressure at 53.0 seconds. A thermal excursion began at the top of the active core region at 108.4 seconds. The vessel liquid level reached its lowest point in the core at 184 seconds. The core was 95% uncovered when accumulator injection was initiated at a cold leg pressure of 1.66 MPa at 186 seconds. A maximum fuel cladding temperature of 715.0 K was reached at 198.0 D-5

~ seconds. The core reflood was assisted by the scaled LPIS which [ began injecting ECC water to the intact loop cold leg when the primary coolant system pressure had decreased to 1.08 MPa at 201 L seconds. Core reflood was complete and the experiment was terminated at 214 seconds when all fuel cladding temperatures indicated at or below the saturation temperature. 5.0 L5-1 Transient Calculation The method for performing the experiment prediction was to and begin with a best estimate nominal calculation (base case) then perform sensitivity calculations on phenomena that are known to have larger uncertainty. The phenomena of larger uncertainty examined in the experiment prediction was the critical flow through the break. The sensitivity calculation used a subcooled discharge coefficient of 0.85 and a two-phase discharge coefficient of 0.9 instead of the nominal calculation with both coefficients of 1.0. Both calculations are presented. i 5.1 Base Case Simulation of Experiment L5-1 The RELAP5YA calculation for L5-1 predicted the important phenomena occurring during the transient in the proper sequence, as shown in Table 2. Figure 5.1 compares the calculated and measured primary system pressures. The uncertainty on the measured pressure is + 0.14 MPa). The analysis agrees well with the data for the first 100 seconds of the transient. Between 100 seconds and 200 seconds, the system depressurizes more slowly than calculated. Since the accumulator injection in LS-1 is initiated by a low pressure reading, accumulator injection is thus calculated to begin at 154.5 seconds which is about 30 seconds earlier than in the experiment. The cause of the discrepancy between calculated and measured depressurization was attributed to the overestimation of the break flow and underprediction of D-6

J primary inventory at later times. q [ rigure 5.2 shows the calculated and measured break flows during L5-1. The break flow during the first 50 seconds of the transient is a bit high compared to the data. Discharge C coefficients of 1.0 were used for both subcooled and saturated break flow for the base case. The break flow is not well-known experimentally; the quoted uncertainty is 3 16 Kg/sec. For this test, with the primary coolant pumps tripped, the ability of the system to continue to provide core cooling depends in large part on che primary coolant inventory remaining in the system. The experimental mass inventory in the primary system is not available in the data tapes. The plot labeled "DATA" was taken from Reference 7. Comparison of calculated and measured inventories throughout LS-1, given in rigure 5.3, shows fairly close agreement until about 50 seconds, when the calculated inventory begins to fall far below the experimental inventory. j This is most likely due to the overprediction of the break flow for the first 50 seconds of transient. By the time of accumulator actuation (154.5 seconds), the calculated inventory is only 623 Kg, while the measured inventory is about 950 Kg. The effect of the accumulator injection is seen as an increase in primary system f inventory; however, the increase is not as sharp as seen in the measured inventories. The reason can be seen in rigure 5.4 which presents the accumulator injection rate versus time. The experimental accumulator injection rate is not reported in the data report. The estimation of the experimental accumulator flow was made by Sandia in Reference 8, from the accumulator liquid level. As can be seen, the accumulator flow in the experiment increased sharply from 0 to 20 Kg/sec in 10 seconds, reaches a peak of about 24 Kg/see at 200 seconds and drops to 0 at 220 seconds. In the calculation, the accumulator injection flow never exceeds 13 Kg/sec, it starts at 154.5 seconds and remains on for the D-7

J remainder of the transient. Due to lower calculated system } J pressure, the accumulator injects about 330 Kg more than the data. L The calculated rates of core uncovery and refill are shown in rigure 5.5. The collapsed liquid level drops at an average rate ( of 0.041 m/sec during the period of core uncovery, much higher The core refill is than the measured uncovery rate of 0.02 m/sec. calculated to occur at a rate of 0.008 m/sec, much slower than the measur'ed rate of 0.055 m/sec. Also, the core starts uncovering earlier than the data. By about 90 seconds, the core is totally empty while in the test the liquid level from the upper plenum drops into the core by 108 seconds. The early core uncovery is attributed to the overestimation of break flow. Early core uncovery leads to early clad heatup. rigures 5.6 through 5.9 show the calculated and measured clad temperatures for four core elevations 0.20 m, 0.76 m, 1.14 m and 1.57 m. The clad thermal excursion was predicted early at all core elevations due to earlier, deeper and faster core uncovery. The clad quench is predicted to occur about 20 seconds early f for the 0.28 m elevation and 10 seconds early for the 0.76 m elevation. The core is totally quenched at 250 seconds, 35 seconds later than the data. The early quench at the lower core elevations is due to early accumulator injection. A peak clad temperature (PCTS) of 899.23 K at 170.5 seconds was calculated vs. 715 k measured in the experiment. The peak clad temperature was conservatively calculated at all core elevations. The experimental core exit temperature shows that the core produced significant amounts of superheated steam. Figure 5.10 presents the calculated and measured temperature at the core exit. The new code version was able to calculate significant steam superheat, in excess of the measured one, rigures 5.11 and 5.12 compare the calculated and measured intact and broken loop cold densities. The calculated and measured densities are D-8

s generally in reasonably good agreement until accumulator actuation when the density in the cold leg intact loop deviates from the measured data. In the intact loop cold leg, a measured density increase at about 190 seconds was caused by the accumulator ( injection, followed by a decrease when the liquid drains into the downconer. The calculation showed a larger density int 9ase, but { not the decrease. This difference may be attributed to a small calculated draining of liquid from the intact loop cold leg into the downconer and to the different behavior of accumulator flow presented in rigure 5.4 5.2 Summary of Results The base case calculation of RELAP5YA against LorT Experiment L5-1 indicates the following: a. The overprediction of break flow led to a significant difference in the primary mass inventories during the latter part of the transient also led to faster depressurization and early accumulator actuation. b. The rates of core uncovery and refill were conservatively calculated. The code calculated early thermal excursion at all core elevations. The PCT was overpredicted at all elevations, c. Significant steam superheat was calculated at the core exit. i The underprediction of the primary pressure and of the system mass inventories was attributed to the overprediction of break flow. A sensitivity study, using break discharge coefficients was performed. The results of this study are presented below. D-9 i J

6.0 Sensitivity Study: Experiment L5-1 r L In order to improve the prediction of system pressure and, consequently, accumulator actuation time, a calculation was performed by reducing the discharge coefficient at the break. Values of 0.85 and 0.9 were used for the subcooled and saturated discharge coefficient, respectively. The calculated results were an improvement over the base case calculation. The depressurization rate, presented in rigure 6.1 was lower and the accumulator came on later in time namely at 170.4 seconds. The calculated system pressure at accumulator actuation is 0.27 MPa lower than the data which compares well with the experimental uncertainty of 1 0.14 MPa. The comparisons of break flow and primary system mass inventory are presented in Figures 6.2 and 6.3. The effect of the break discharge coefficient can be clearly seen in the prediction of the primary system inventory. There is only 160 Kg difference between the inventory measured and calculated at the time of accumulator actuation. Given the high uncertainty of i 16 Kg/sec on the break flow, this difference in inventory, at 170 seconds in the transient, is considered to be well within the measurement uncertainties. The accumulator injects about 778 Kg of liquid, 150Kg more than the data. The accumulator flow, presented in Figure 18 does not match the experimental flow. The overprediction of total mass injected by the accumulator is due to slightly lower calculated system pressure at the time of accumulator actuation. The rates of core uncovery and recovery were similar to the previous calculation, however the time of total core uncovery, presented in rigure 6.5, was delayed, by about 20 seconds, (compared to previous calculation) to 95 seconds. The early core uncovery leads to higher calculated clad D-10

ls L 0 temperatures. A PCT of 846.9 K was calculated to occur at 178.6 seconds which is conservative compared to the measured value of 713 K. The clad temperatures in the hot rod, at four axial r l elevations are presented in rigures 6.6 through 6.9. All calculated temperatures at all elevations are well above measured ( values. The core exhibits early thermal excursion and, due to the intermittent a cumulator injection, is quenched later in the transient. The other calculated parameters such as core exit temperatures cold leg densities, presented in rigure 6.9 through 6.11, were similar to the previous calculation. ) In conclusion, reducing the discharge coefficients from 1.0 to 0.85 in the subcooled blowdown region and 0.9 in the two-phase region provided better system pressure and accumulator actuation time predictions than the previous base case calculation. t 7.0 conclusions The following conclusions can be drawn based on the assessment of RELAP5YA against LOFT Experiment L5-1: 1. The LS-1 calculation correctly demonstYates the l availability of adequate core cooling from HPIS, LPIS and accumulator injection, even with the pumps off and when a major portion of the primary fluid was lost in an intermediate break. 2. The Sensitivity Study improved the prediction of primary system pressure, system mass inventories, and the timing of core uncovery and accumulator actuation. i 1 3. The code predicts conservative rates of core uncovery D-11 ~

J and recovery. As a consequence, the peak clad s temperature is consistently higher than the data at all ( core elevations. t h D-12

s REFERENCES s r 1 1. Fernandez, R. T. et al., "RELAP5YA-A Computer Program for LWR System Thermal-Hydraulic Analysis," YAEC-1300P, October 1982. 2.

Letter, G. Papanic (YAEC) to J. A. Zwolinski (NRC) on Transmittal of Responses to 43 NRC Questions on RELAP5YA, November 1, 1985.

3.

Reeder, D.

L., LOTT System and Test Description (5.5 ft. Nuclear Core 1 LOCES), TREE-1208, July 1978. 4. Grush, W. H., et al., "Best Estimate Prediction for the OECD LOTT Project Small Cold Leg Break Experiment LP-SB-3, OECD 1 LOTT-T-3603, February 1984. 5.

Jarell, D.

B.,

Divine, J.

M., Experimental Data Report for LOTT Intermediate Break Experiment L5-1 and Severe Core Transient Experiment LB-2", NUREG/CR-2398, EGG-2136, November, 1981. 6. Letter, S. Modro (EG&G) to L. Schor (YAEC) on Accumulator "A" For L5-1, SMM-08-87, March 16, 1987. 7.

Adams, J.

P., "Quick-Look Report on LOFT Nuclear Experiment L5-1 and L8-2", EGG-LOFT-5625, October, 1981. 8.

Orman, J.

L.,

Kmetyk, L.N.,

"RELAPS Assessment LOTT Intermediate Break L5-1 and L8-2", NUREG/CR-3406, SAND 83-1575, August, 1983. D-13

TABLE 1 I s L EXPERIMENT LOTT LS-1 INITIAL CONDITIONS PARAMETER EXPERIMENT RELAP5YA [ LS-1 CORE POWER (MF) 45.91 2 45.9 1 HOT LEG PRESSURE (MPa) 14.931 08 14.92 0 MASS FLOW RATE (kg/s) 308.21 0 312.3 4 COLD LEG TEMPERATURE (K) 552.310.9 551.59 PRESSURIZER LIQUID LEVEL (m) 1.131 03 1.12 0 SG SEC. WATER TEMPERATURE (K) 537.810.8 537.8 SG SEC. PRESSURE (MPa) 5.051 06 5.06 0 3.22 0.02 3.19 1 SG SEC. LIQUID LEVEL (m) SG SEC. FEEDWATER FLOW (kg/s) 25.31 6 25.3 0 D-14

s l r s TAPLE II L L5-1 SEQUENCE OF EVENTS r EVENT TIME (S) L MEASURED CALCULATED BASE CASE SENSITIVITI STUDY ( EXPERIMENT L5-1 INITIATED 0.0 0.0 0.0 PRESSURE REACHED 14.2 MPa 0.17+0.01 <0.02 <0.02 r (REACTOR SCRAM, FEEDWATER ~ TRIP, STEAM VALVE TRIP) / UPPER PLENUM SATURATED 0.2+0.1 <0.5 <0.5 HPSI INJECTION STARTS 2.9 2.5 2.8 PUMP TRIP 4. 0,+0. 5 4.0 4.0 SUBCOOLED BREAK FLOW ENDED

10. 5,+0. 5 17.0 19.0 PRESSURIZER INDICATED EMPTY 15.5+0.5 21.0 17.0

~ (1.38E-3m) (2.6E-3m) PRIMARY PRESSURE BELOV 53.0+1.0 32.0 34.0 SECONDARY ACCUhULATOR TRIP 185.8,+0.5 154.5 170.4 CLAD QUENCH STARTED 188.1+0.5 168.0 178.0 MAXIMUM CLAD TEMPERATURE 198.g2.0 170.5 178.6 REACHED (715 K) (899.23 K) (846.9 K) LPSI ON (PRESSURE 1.08 MPa) s201.0 220.0 275.8 CLAD QUENCH COMPLETE 213. 0,+1. 0 250.0 275.0 l l 1 l l D-15 l

LOFT Systsm Configuration intact loop Broken loop ~ r 0"3*k 'P'"3"9 steam Sleam 9'"*I"' N simulator,%'e "~VgN n, s Break plane h ~ ~ Pressurizer 4 t' Isolation V lla y N *** 1 c Li Break plane D 'O--A w Pump simulator i y ) q IE 1 ECC injecilon ') fli',: umps Q location r l Ots -'l E Downcomer

s. Q p,,pp,,,,,,,

Reactor Core vessel ,7,,g Lowar ple usen Reactor vessel Figure 2.1: I,0FT System Configuration

5 J M b 4 1 [ T

p. i -m. r.

I;s I H ( [mii-i-i,.) 1 1 l t at I I r, 1 1I E l t a l - -t m l - 3 e d a g l g l g a s a hh h l-e u 7 %,Ei = 5 p is s s gl =! T q =, -... - - -=_ _ g s a g,- 3.s. c -a i-i-m g; =,. = _, - .a -j a 8 g,. = m el ~ 3 e mm. S gi so gi ffj M, I Q "g I t a i' 4-,; s 5 g fri F.I i l 8,_ I g7 3 3., m 3 I I w 3 i 2 E 5 -1,, t,

=

c J s a 5 g j 1, L_ e !! m B rf 5 t ~ g

1 e

s e

t 3~3*

? 2 u I

r.(-- gl

--s 3 g I 5 5 2' u3 - > <c=(m 3 .t-g 'I' ,,,~-& $llE I , p m==c i =. I i

5. l' y

s s D-17 i

s, A s 1 3 I R R I J t R 3( a>. .g I {.1 I i, l .L.1 u " * " * ' ~ ,j, 4 j 3 e m g 7 en saa n r 4 E m p l g .,,,~.< a g< c c E c, g Jg is f. y [-[he sh ~.. a gL l hie 7 g ~i-% g, 4 a i ..,. u j k. j h jf d o m s Q us we an o i u e R l j F 3 t-- .2 @ A 9 S 'l I t v I I _.. gl i w -j M, { T E l l' l .c se p .E E S, zi E i 1. E j: ~" E ~ u

=

j s s [ l OE 3 I "~ m f j ! a, r y G sm s 3 9 ?, 2 5 m = s.i.i-. g .s T@ 'i e 16 tr g -Qs c g, - Y, u,- J' 5 5 3 m g lj h. 1-j I< _t .f [ g 1 d. g! mQs 9" 5{ . = = = ,, ~ m-l i I t D-18

. 0 ~ '. '. '. ". _. S '~ . 5 PRESSURE INTRCT LOOP COLD LEG o

N 3 _::: R5 I:: _.0 7

C: JE:G4: 7 { 7::\\ 5) I I i 1 1 os m o T b -E] l o_ - 2-PE-PC-001 v o e--e P 185010000 o-y w-o- i m o m y g-o- Q_ 9 E C

)

u o i I I I I 0.0 50.0 100.0 150.0 200.0 250.0 300.0

v E

= 7:: E R Ri'::L R E [ SEC0\\JS ) Figure 5.1: Comparison of Measured and Calculated Cold Leg Pressure (Base Case)

g m r, r,,gr, g ._, U ..._ J. aJ_.. BRERK FLOW o ::\\ 3J::: R5:~: 0 J0 J E : G4: 2 ( 7:: 45) o 1 I I I I O ^ 03 l CLi_] o. U) o _ W N-I cs o rR M -BRK e-* IRONJ N m o Z o-o O 1 o a_ cn * (D U

.A4%A M vrM A

.m_. m d .L 1 ima uv %,r = m i -, a q V9 vi u a o i si i r 1 JIP IJ gIII v I I L 0.0 50.0 100.0 150.0 200.0 250.0 300.0

E P'::ER _RL 3':.URE C SECC\\DS

) Figure 5.2: Comparison of Measured and Calculated Break Flow (Base Case)

M N N N m 7 I - 1 1 1 J ' c 0,~.' '.' ~.' J.. aJ a ToTRL MRSS

N PL:'. : R5:::'.

0 7 C: E:G4: 7(P:.NL5) 6000. t i I I I I 5000. { \\ 4000. DATA (E7) e-e CNTRLVfR 000000072 mg un 7 "' 3000. ~b 0 2000. d r, r,_ 1000. 0.0 2do.o 2so.o 300.0 s6.0 1do.o 1so.o o.o a es.- y r, m 7 7, 3r,- 7 { p a - R.\\ \\w ,\\ L aU .m w Figure 5.3: Comparison of Measured and Calculated Primary System Ir.ventory (Base Case)

_.0 71' " E S" L5-:. RCCUMULRTOR FLOW o :: \\ 'J": R5::".07 C0JE:G67(7"\\5) fl I I I I n OEd O W o" _ l xm C9 L51RCC V s-e f1FLONJ 615010000 O ? "d_ ~ x [i.l f O ] 3 1 O Li_ U I

E C

C C C C W O C IO ~ l l I I I I I 0.0 50.0 100.0 150.0 200.0 250.0 300.0 "::

E R=TER RU'"L RE

( SECONES ) Figure 5.4: Comparison of Measured and Calculated Accumulator Flow (Base Case)

m r i r i LOFT TEST LS-1 R5ITLOF RSf1GG4F REf1CTOR VESSEL LEVEL O. g e 8. I o. I -3 I I i l l I V r N w O J f. 3. ... ~..j ..... }: 0 (J y taJ o J o TOP OF CORE B i i i O. ~. B Pt OF CORE i i ~ l o. o v v v s 0.0 60.0 120.0 180.0 240.0 300.0 TIME (SEC) Figure 5.5: Reactor Vessel Collapsed Liquid Level (Base Case)

l 0 0 ~ ) 0 ) 3 5 ) 0 N 1 10 0 T 0 .S ) 0 e 0D F 1 s 13 I a 5N 12 C ( 0 C 2 e I F 6P O sa 0 B M4 5D C ( ET 0E T dM 8G TH e 2: a t8 .Sl a2 0 u0 1 T 0E e c I 0 lL DI 2( CaE 5 O t d a LE nae RC u E r d u U 0R et ra T .J ur se 0 ap TR 1 em R 5T Me S E Pf 1 T I od EP J nl a TM R id oC s E 0 ro T aR p T F 0 R mt oo j0 Cl FD l E O_ O 0 1 1 R T6 5 T F T e O I 0R rug H5 0 i F R E 5 1 M ~ T I i U T r 0 P b 0 N ] I < ~ [_ OOO~ O dO0 od o.O0 o OOW 0 - l _m1rn>r[ V1 o l i I r w 5 llIl1

07:: TES:: .5- :. m HOT ROD TEMPERATURE 0.76M

N 'U": R5:::: 07 CC D E : G4: 2(7T\\5) m O

1 I I I I I O O ~ m O Cb h l O 00 l 1 v WO Z V Ol3 3 H (ff T r, / a-J o' O k.., 3 y w t OOOh TE-ses-030 EW e-a HTTDP 231000310 f_2 3 O I w OO \\ CQ sb.o 150.0 i do.o 2$0.0 250.0 300.0 o.o

"E 22:: ER RL 3::_ RE

( SECC\\:S ) Figure 5.7: Comparison of Measured and Calculated Hot Rod Clad Temperature at EL 0.76 M (Base Case) g.,. g

~ I a 0 " '. ~ ~ %.o* 5 .J a HOT ROD TEMPERATURE 1.14M

N 'U::: R5I::107 C0 J E : G4: 7("T\\5)

O g i i g i O O .-e m O v f O_ O CD v h oO A x y ,Ol 4 "H8 m CE N T A O .J y O_ O - TE-5E-045 Eh e-a HTTEtF 231000410 - HTTEff 231000510 MO s O 1 O (N iso o 2do.o 2so.0 300.0 o.o sb.o do.o i KN E = =::ER L 3:: RE ( SEC0\\JS ) l Figure 5.8: Comparison of Measured and Calculated flot Rod Clad Temperature at EL 1.14 M (Base Case)

-_~m m .JPT

ES'.:

.5-:. HOT ROD TEMPERATURE 1.57M D ::\\P ':: R5::TLor CC J E :G4: 2 ( 2::\\ 5 ) 1 I i I I OO A G No TE-506-062 y a1 o--e HTTDP 231000610 j O ~ as Et3 o K a e >8 (a _ n _ s to i g 0-N.. / 2 a O_ O.h rS L'l o H o'x O\\ C\\1 o.o sb.o ido.o 15o.0 2do.o 25o.0 300.0 X ? E 3 7:: ER RU3:: URE ( SEC0\\JS ) Figure 5.9: Comparison of Measured and Calculated llot Pod Cicd Temperature at EL 1.57 M (Base Case)

x C' 7 L a d.. adc ~ ' a FLUID TEMPERRTURE CORE EXIT o ::\\"L::=R5::': a: C: J E-G4: 2(F:: N5) 7 o I I I I I o CN m l o V O[ l G co v \\l V M i_L3 o T l ,o sm a-U- -kb_ I ') x o W TE-StF-001 O

-" TEMPG 230060000 O_ S 1~

Mo H ~ oo t9 _ l i I I I I I 0.0 50.0 100.0 150.0 200.0 250.0 300.0

E

= 7': ER RF"L RE ( SEC04JS ) f Figure 5.10: Comparison of Measured and Calculated Fluid Temperature at the Core Exit (Base Case)

w u u a LI 1 . mSr1 _. 5 _.s. r1 p rys 1 FLUID DENSITY INTRCT LOOP COLD LEG - o .. \\ 3L c,... R5. m.1 m0 C0J__,.n,: ( r, \\ 5 ) o. O I I I I I DE-PC-001FI M" s....e DE-PC-0018 y e--e DE-PC-001C 4 y ID

----A RHO 185010000 e'

E tN

4 4

[ N G O!

i i

O E L i i

i j

} !! o =: ID 'l.',i i !i h $ = !! ii i !i :: 5: i !! it 5 i. h i!i k ii !! !! ! j!! i ':L $is. i E Fi 1 m k ii I O $ lii i :i i: l j !!.~: i! ! i i V. - } i !! !! ! ! ! Vi' i e, 1 'ii '@kk 4 >-i ! ii *! [ ii b 853 i f.; i ik k !*! ii Ii i b !, 's : 1 3 i .c

r. 3 5 1 :: i:

(D O il i,. i - i 4 s $ Y S't h O LDN H io 1 1 I I I I I L 0.0 50.0 100.0 150.0 200.0 250.0 300.0 a qw 3,33a ( 3rC0\\JS ) v n s t a Figure 5.11: Comparison of Measured and Calculated F]uid Density in Intact Loop Cold Leg (Base Case)

_ ~ O ~' ' ' '~' r S 5' FLUID 05'NSITYBkOKENL60PCOLDLEG

\\ 3 J": R5:~".0 7 C0JE:G47(7"N5) o O

g i g i i M ~_ X l to 3 I N( e: T y NO ) DE-BL-001R e ---e DE-BL-0018 E@ _ lllg l

l M,

e--o DE-BL-001C

-aRHO 345010000 O

!! 's +. ? o !! f' ;:" t 8 m > t: B l4ii; I ilij : i - t .I J, E-' \\l\\ k.j '- t, .n T 1 tn Om l i ~@ o io 1 I I I I I I ti_ 0.0 50.0 100.0 150.0 200.0 250.0 300.0 l ":~

E

= FT ER RUPi' R E ( SEC0\\JS ) Figure 5.12: Comparisson of Measured and Calculated Fluid Density in Broken Loop Cold Leg (Base Case)

m m u r, m r, s 0 mr, J J m.J PRESSURE INTFICT LOOP COLD LEG o

N 3 J": R5::"C0F.

C0 J E : G4: 7(7"\\5) d_ i I i 1 I cx1 o C Lf5 - ] o_ E Z-PE-PC-001 -o c aP 185010000 Z o_ Li_1 - T ) U3 07 9 o m-s T O_ 9 GL O O

]

o 5b.0 100.0 150.0 200.0 250.0 300.0 0.0 ":: # E = 71' E R R _ 3" R E _: SECONJS ) Figure 6.1: Comparison of Measured and Calculated Cold Leg Pressure (Sensitivity Study)

~~ ~~ 0 ~:: ::ES:: .5- :. m BREFlK FLOW o :: N 3..:: R5::'.007 C0 J E : G4: 7 ( 7': \\ 5 ) o 1 I I I I om o Li_] O cn o - x$ L cs M / h a FR-BL-BRK r -eNFLOWJ 365000000 V, m m g r~ o1o u o_ cn

cn

. ik M1buifDNAF'n W3 l i i ni ryy n n v i i r, y i s 0.0 50.0 100.0 150.0 200.0 250.9 300.C ":: Y E = 7:: ER RL 3::_ RE C SECO.\\ JS ) Figure 6.2: Comparison of Measured and Calculated Break F. low (Sensitivity Study)

u w v, 7pr1 {_a r1 _., O p r -e aJ.. J I HOT ROD TEMPERATURE 0.28M R5:::?00= G4:= { 2:: N5) c' C3 g l l l l { O O ,-e ^ 1 O C$ - O TE 'iG6-011 00 m-m HITEMP 231000110 v

-*HTTEtF 231000210 MO K

9 al] O b_ m 0-N ^ Z 3 ~ O 6 y O_ O - o l T Ei3o l H o O 63 I I I I I I 0.0 50.0 100.0 150.0 200.0 250.0 300.0

Y E 2'7:: E R L ~3': E{

SEJONJS ) Figure 6.6: Comparison of Measured and Calculated Ilot Rod Clad Temperature at EL 0.28 M (Sensitivity Study)

~ w ru w a LOFT TEST L5-1 RSITC0F R5RGG4F RER__C_ TOR VESSEL LEVEL O. e 1 1 O. _............................................................................................................................................................. i l O e w m o I Ed l Ed 1 J i TOPiOF CORE i i w o.................................. BO'4r0M OF CORE ~ o. o a s a u 0.0 60.0 120.0 180.0 240.0 300.0 TIME (SEC) Figure 6.5: Reactor Vessel Collapsed Liquid Level (Sensitivity Study)

m m 0 mr, _, J. J_.. r, m g r, { s a ACCUMULRTOR FLOW

\\ "J::: R5::::CO2 C0 J E : Gf: 2 ( F \\ 51 o

d I I I I I n C] [.i_1 O W d-l xm C3 L51FEC V m-eNFLOWJ 615010000 l o b-k u . -g 3 N ] l a w 'Eu i fO _ O O W W [g 5o .-1 I I l' I l l 0.0 50.0 100.0 150.0 200.0 250.0 300.0

v E ? =::E R R _3 E

.: SECJ\\JS ) Figure 6.4: Comparison of Measured and Calculated Accumulator Flow (Sensitivity Study)

_m - m LOF": ::ES:: .5- :. TOTRL MRSS

N3 '::R5::::007 C0]E : M 7 ( 2:: N5) 6000.

I i l I I 5000. -f e 4000. _ GRAPH t RET 61 H CNTRLVfR 000000072 mg 75 M} 3000. - I 4 O 2000. - }- 1000. I I I I I 0.0 50.0 100.0 150.0 200.0 250.0 300. 0 gr v ,T qi 3-, qr 32C0\\JS ) v r 1_., a .igure 6.3: Comparison of Measured and Calculated Primary System Inventory (Sensitivity Study) {

02:: ::ES' _5- :. m HOT ROD TEMPERATURE 0.76M

LP L ':: R 5 "{ f 0l2 C0 J E : G4: 7 ( 7::\\ 5 )

o g i o o m o o_ I 83 / m Wo ) oa M .oE ~ i g E-i w _r, T o fv L C [] o o_ o - TE-ses-03c E P m-m HTTEMP 23100C310 Wo E-i oo C\\1 1 I I I I 0.0 50.0 100.0 150.0 200.0 250.0 300.0

E

='7:: ER R P'.:LRE ( SECJNJS ) } Figure 6.7: Cemparison of Measured and Calculated Hot Rod Clad Temperature at EL 0.76 M (Sensitivity Study)

~ m m l _ 0 ? '.' '? E S " _. 5 - :. HOT ROD TEMPERATURE 1.14M

\\ ' ": R5: "C0 7 C0 J E : G2: 7: 7"NS) o o

O l m o V o_ d o y oo 't y e, s a oJJ m W8 C ~ cc J Z o &n i ,J o Q._ O - TE-sm-045 E@ e--e HTTEf1P 231000410 oo - HTTEff 231000s10 9 oo~ l I I I I i_ 0.0 50.0 100.0 150.0 200.0 250.0 300.0 :~v- - -.1R _L .'_ R E C S ECO \\ JS ) Figure 6.8: Comparison of Measured and Calculated flot Rod Clad Temperature at EL 1.14 M(Sensitivity Study)

__-__->m m - r, m Q r s [ s _., O m r _, J. __. J HOT ROD TEMPERATURE 1.57M . :: N " : R5:::700' C0]E : G4: 2 ( : N5) o I I I I I oo m o V TE-5G6-062 o_ m-eHTTEt1P 231000610 m o Lilo 7 ] W7 ~ . = k c5_[W /O r $o l I / t., r a o o eo r? Lil o s oow I I I I l 0.0 50.0 100.0 150.0 200.0 250.0 300.0

E

= =T E R U3 RE ( SECONJS ) Figure 6.9: Comparison of Measured and Calculated Hot Rod Clad Temperature at EL 1.57 M (Sensitivity Study)

m LO 7: "ES" .5-: FLUID TEMPERATURE CORE EXIT 9 :: N 3 =R5::"C0F C0 J E=G4: 7 ( 7:: N5) O i i g l O O C3 m O_ O V 03 O TE-5tP-001 v o - s-eTEMPG 230060000 o8 l QC c, i o ~ g ) O EH N i a-9 [I l l MO_ b Li3 @ [Q o_ O ~ 1 E i y g_ t~ s ch n O P 250.0 25o.0 300.0 55.0 150.0 15o.0 l o.o

v E

= ="ER L 3"L RE ( SEC0\\JS ) Figure 6.10: Comparison of Measured and Calculated Fluid Temperature at Core Exit (Sensitivity Study)

W l f U L 0,r, r,., e, FLUIDDENSITYiNTRCTLdOPCOLDLEG r, T r, 3 ,r, 2 O .. = .s. a 1. = O i i i g N DE-PC-001R o - e OE-PC-001B in I 8--e OE-PC-001C X .g EN P

-- =a RHO 185010000 i

g l (9 O()q ii !i I O i l .!l I !! ji i i; !!:ii LD s' 'i is ij !jiii i-l! f : il !!. !!!i

i !! j!i!

]

v a-i

j ::

? O t j j:: : :i i 1:ll A:i p: ii !I !!:li si n ,::it : ~ M i f l

i:i 5

N H i :: (D O .j ( g 3 g,; g i jj jg i: e /.:))

5)

ii

.h[- f '3)- 1 O LD \\ C\\1 3 O 1 I I I I I I E 0.0 50.0 100.0 150.0 200.0 250.0 300.0 -, r q q 0-, qr ( 3 00\\JS ) - v a w -w -.s a a Figure 6.11: Comparison of Measured and Calculated Fluid Density in Intact Loop Cold Leg (Sensitivity Study)

- -m m 0 ,~ 7 g,' 5' FLUID DENSITY BITOKEN LdOP UOLD LEG

N 3 J': R5::'.:00 =

CEJE:G67: 7:'\\ 5 ) O O l l l l l M ~_ X 1 II) T. ~ i DE-BL-001A U '.Ya e--e DE-BL-0018 \\ GO ll l J-d. e--e DE-BL-001C 2 I0 !! j, l a--* RHO 345010000 G

l j

!!b,,*; to 03 ll 3 ll t l Y i" i k I,P i T j z I I [ t.] $ I O i O H 1 O 3 I I I I I ti_ 0.0 50.0 100.0 150.0 200.0 250.0 300.0 :.

E F E E R R P

R E ( SEC0\\JS ) Figure 6.12: Comparison of Measured and Calculated Fluid Density in Broken Loop Cold Leg (Sensitivity Study)

J L F L ATTACHMENT E ASSESSMENT OF RELAPSYA AGAINST SEMISCALE SMALL BREAK TEST S-LH-1 }

s s ASSESSMENT OF RELAPSYA AGAINST g .SEMISCALE SMALL BREAK TEST S-LR-1 s [

1.0 INTRODUCTION

The assessment of RELAP5YA for licensing analysis of PWR SBLOCAs was presented in References 1 through 7. Although the assessment covered the spectrum of phenomena expected in PWR SBLOCAs, there was no direct assessment against an integral test designed primarily for PWR SBLOCAs in which core uncovery/ recovery was encountered. The Semiscale Small Break Test S-LH-1 provides these phenomena. Hence, RELAP5YA has been assessed against this test. In the assessment presented, two calculations have been carried out. The first is a base case best-estimate calculation with the subcooled and two-phase discharge coefficients equal to 1.0, while the second is a sensitivity on the two-phase discharge coefficient with a value equal to 0.80.

2.0 BACKGROUND

Semiscale Test S-LH-1 (Reference 8) siculated a five percent centerline break in a PWR cold leg. The phenomena observed in these tests and a post-test analysis of the test using the RELAPS/ MOD 2 computer code (Reference 9) have been presented in Reference 8. The test was conducted in the Semiscale MOD-2C facility (Reference 10). The input deck used for the RELAP5YA simulation of this test was derived from the input deck used in the RELAP5/ MOD 2 analysis of the f acility (Reference 11) and an earlier RELAPS/ MOD 1 model (Reference 12). Various changes to the input were necessary because of the different input requirements of RELAP5/ MOD 2 and RELAP5YA. Some additional changes have also been made to reduce the computational costs. The changes are described in the next section. E-1 1

3.0 RBLAP5YA BE MODEL FOR TEST S-LH-1 Figure 3-1 shows the nodalization used in the RELAPS/ MOD 2 analysis of Test S-LH-1 (Reference 8). The changes required to convert RELAP5/ MOD 2 input to appropriate RELAPSYA input were as follows: Conversion of cross-flow junctions to normal junctions because a. RELAP5YA does not have this feature, b. Corrections and modifications to various control systems. These control systems accessed features not available in RELAP5YA. For example, in RELAP5YA, input for heat structures cannot be modified during restarts. Several additional changes were also made from the perspective of modeling philosophy and computational cost. These are explained below, Heat structures simulating heat losses and those simulating a. external heaters in the vessel and loop piping have been retained. However, the number of radial mesh points has been considerably reduced. This change was primarily made to decrease computational time. b. The number of radial mesh points in the core heater rods was reduced from eighteen to six. Also, the number of radial mesh points in the steam generator tubes was reduced from five to two. Again, we believe this would result in substantial savings in computational cost without overly compromising the accuracy of the simulation. The number of hydrodynamic nodes in the core was increased from six c. to twelve to provide better resolution of the core thermal-hydraulic response. The change results in core nodes of 1.0-foot lengths compared to the 2.0-foot lengths used in the RELAPS/ MOD 2 analysis. E-2

a d. The top region of the downcomer is split into two parallel annulus components. One component connects to the broken loop and the s L other connects to the intact loop. The top of the two annulus components are connected by a single junction. We believe this is [ a more realistic modeling of the test downcomer top region. ( e. A motor valve is used to simulate the test break assembly. By using a motor valve, we were able to better match break flow during the early part of the test. f. The model for the secondary side of the steam generators was derived from an earlier RELAP5/ MOD 1 model of the semiscale facility. This final model is similar to the models expected to be used in PWR analysis using RELAP5YA. The above changes are generally consistent with the modeling concepts that will be used in RELAP5YA analysis of PWR SBLOCAs. The sequence of events for Test S-LH-1 and the test results h:ive been outlined in Reference 8. The results of the test are p.esented in the next section. 4.0 RELAPSYA BE SIMULATION OF TEST S-LH-1 { The semiscale 5% SBLOCA Test S-LH-1 is characterized by two major core uncovery periods. The first uncovery was of short duration and was caused by the manometric head balance associated with liquid holdup in the steam generators and pump suction clearing. The second uncovery was of longer duration and was caused by core bolloff. The core liquid level recovery in the first depression occurred after the pump suction seals cleared in both the intact and broken loop. This occurred at about 300 seconds. The second core liquid level recovery began after accumulator injection. l E-3

s a h Therefore, in this section, the prediction of the phenomena which occur in the first 300 seconds of the test will be presented first followed by the system hydraulic and thermal response along with the break flow characteristics. 4.1 Base Case: BE Simulation of Test S-LH-1 When the transient was initiated due to the 5% SBLOCA, the pressurizer ( began to drain until it emptied at about 40 to 50 seconds. The pressurizer drain comparison is shown in Figure 4.1. Good agreement can be seen between experiment and calculation. The drainage from the upper head is shown in Figure 4.2. The liquid fraction in the bypass line that connects the downcomer to the upper head is shown in Figure 4.3. The bypass is calculated to clear at about 180 seconds. This occurred at about 240 seconds in the test. When the bypass line cleared, steam flow from the upper head was able to vent through to the break. Until this happened, however, the steam was blocked due to the presence of liquid in the steam generator tubes and the pump suction pipes. The test data comparisons of liquid head in the upside and downside sections of the broken loop steam generator and the corresponding calculated r.ollapsed liquid levels are shown in Figures 4.4, 4.5, and 4.6 respectively. The test data for the short and long broken loop steam generator tubes are shown in Figures 4.4 and 4.5. The downside of the broken loop steam generator tube is calculated to clear completely at about 100 seconds, while this occurred at about 140 seconds in the test (Figure 4.5). The intact loop upside and downside steam generator tube behavior is shown in Figures 4.7, 4.8, and 4.9. The test data for the short and long intact loop steam generator tubes are shown in Figures 4.7 and 4.8. The downside of the intact loop steam generator tube is calculated to clear at about 150 seconds, while this occurred at about 140 seconds in the test (Figure 4.8). E-4 j

After the steam generator tubes drained in the test, liquid that y collected in the pump suction continued to block steam flow. A manometric depression of liquid level in the vessel resulted. In the test, the intact loop cleared at about 180 seconds, and the broken loop seal cleared later at [ about 280 seconds. Before the intact loop seal cleared, the vessel level had dropped to about 40 percent of the core height and caused a core heatup. When the intact loop seal cleared, the core level recovered to the top of the Subsequently, when the broken loop seal cleared at 280 seconds, the core. vessel level rose some more. The test data comparison of liquid head in the broken loop pump seal upside section and the calculated collapsed liquid levels in the upside and downside sections are shown in Figure 4.10 and 4.11 ( respectively. The broken loop seal is calculated to clear much earlier than the test indicates, at about 150 seconds, as can be seen in Figure 4.10. This is a direct consequence of the calculated early broken loop steam generator tube drain (see Figure 4.5). On the other hand, the intact loop pump seal is calculated to clear much later than in the test, at about 250 seconds, and can be seen in Figures 4.12 and 4.13. Therefore, the order of loop seal clearing in the calculation is the reverse of what occurred in the test. It shculd be noted, however, that in both the calculation and the test, the loop seals were cleared by 300 seconds. As a result of loop seal plugging, the core liquid level was depressed to about the same point in both the calculation and the test, as can be seen in Figure 4.14. By 300 seconds, when both loop seals have cleared, the calculated liquid level rose to the level obtained in the test. After 300' seconds, the liquid in the vessel boiled off. The steam readily flowed to the break location. At about 500 seconds in the test, the system pressure reached the accumulator setpoint. The accumulator was actuated, and the vessel level recovered by the end of the test at about 1,000 seconds. In the calculation, however, the accumulator was actuated earlier at about 400 seconds. This is shown in Figures 4.16 and 4.17. Because of the early accumulator actuation, the core did not uncover as much as in the test, as shown in Figure 4.14. The early accumulator actuation.is caused by a faster depressurization rate. Figure 4.20 shows the system pressure comparison. Before about 200 seconds, excellent agreement was obtained between the calculation and the test. Starting at 200 seconds, the calculated E-5

~ depressurization rate is much faster than the test, with the consequence that the accumulatar setpoint is reached early on, at about 400 seconds. 4 Figures 4.21 and 4.22 show that the fluid conditions and flow rate at L the break are reasonably calculated. In addition, Figure 4.23 shows that the integrated break mass is also reasonably calculated. Figures 4.24 and 4.25 show the broken loop and intact loop secondary { pressure comparison. The secondary pressure prediction is seen to be better for the intact loop. There were two core heat-up periods in the test that were a consequence of the vessel level response, as shown in Figure 4.14. The first was due to the manometric core depression before loop seal clearing and was short-lived. The second was due to boiloff and was terminated shortly after accumulator ( injection began. Figures 4.26 through 4.30 show the clad temperature comparison et four axial locations in the core. During the manometric period, RELAP5YA consistently calculated rod heatup at the 181 cm, 228 cm, and 253 cm elevations, while the test heatup occurred only at the 253 cm elevation, as can be seen in Figures 4.26 through 4.29. The second heatup, due to core boiloff, however, was not calculated as shown in Figures 4.26 through 4.30. This is due to the higher-than-observed calculated depressurization rate when the break was uncovered (see Figure 4.20); leading to an earlier actuation of the accumulators (see Figures 4.16 and 4.17). 4.2 Summary of Results The assessment of RELAP5YA against Semiscale Test S-LH-1 indicates the following: a. The pressurizer drain is well simulated. b. The liquid holdup in the broken loop steam generator is not well simulated. It is calculated to clear much earlier than in the test. However, the liquid holdup in the intact loop steam generator is reasonably calculated. E-6 l

n c. Although the phenomenon of loop seal clearing was simulated, the ~ order was not. In the test, the intact loop seal clears first followed by the broken loop seal. In the calculation, the reverse is true. This is a direct consequence of Item b. L d. The phenomenon of core liquid level depression during the manometric period is well simulated. The thermal response during that period, including heater rod temperature excursion, is also well simulated. e. The phenomenon of break uncovery, as well as the time at which it occurs, is well simulated. In addition, the integrated mass leaving the break is also well dimulated. f. The depressurization rate before break uncovery is well simulated. However, after break uncovery, the calculated depressurization rate is much faster than in the test. g. The calculated accumulator actuation time is 100 seconds earlier than the test. Consequently,, the calculated core liquid level depression, due to core boiloff, was not as severe as the test, [ thus, resulting in the prediction of no heatup. The calculated rapid depressurization rate, when the break uncovers, is similar to the result obtained in the post-test analysis using the RELAP5/ MOD 2 computer code (Reference 8). Basically, the primary pressure is an indicator of the energy in the system and is affected by primary-to-secondary heat transfer, fluid-to-primary piping heat transfer, and energy carried out of the system by break flow. Therefore, there is obviously a connection between how well the primary pressure is predicted and the calculation of core heatup. The calculation of secondary pressures suggests that, after break uncovery, the broken and intact loop secondaries act, if anything, as sources of heat to the primary. Therefore, this component of energy transport should contribute towards decreasing the primary depressurization rate. The E-7

fluid-to-primary piping heat transfer has been modeled in this calculation by taking into account heat losses and the presence of external piping heaters. Therefore, we believe that this component of energy transport is being accurately simulated. Since the density and fluid mass flow at the break are approximately the same as in the experiment after the break uncovers, the calculated primary system depressurization should therefore be the same as in the experiment. The observed deviation points to the problem that the energy carried out of the system by break flow must be too large. This suggests that the simulation of this particular test facility is very sensitive to slight perturbations in break flow. 5.0 SENSITIVITY CASE: BE SIMULATION OF TEST S-1.H-1 In order to improve the prediction of system pressure and, consequently, accumulator actuation time, a calculation was perforced by reducing tiie two-phase discharge coef ficient f rom 1.0 to 0.80. The objective was to see if RELAPSYA, with this change, would yield lower calculated depressurization rate after break uncovery, )ater accumulator actuacion time, and, hence, improved results with respect to core heatup. The calculated results were an improvement over the previous BE calculation. The depressurization rate after break uncovery was lower, and the accumulator came on later in time, namly, at about 470 seconds. This is shown in Figures 5.1 through 5.3. However, more liquid mass was calculated to remain in the core than was obtained in the previous calculation. This is shown in Figure 5.4 where the vessel level at the start of core boiloff (300 seconds) was higher than in the experiment. The previous calculation had a vessel level identical to the experiment at the start of core boiloff. The cause of the hightr liquid mass appears to be the reduction of the two-phase discharge coefficient from 1.0 to 0.8. The effect of that can be seen in Figure 5.6 where the integrated liquid mass leaving the break is less than the previous calculation and the axperiment. On the other hand, the fluid density and flow at the break match the experiment better than the previous calculation. This is shown in Figures 5.7 and 5.8. The other calculated phenomena, such as pressurizer, upper head, and steam generator tube drain, in addition to loop seal clearing, were similar to the previous calculation. f E-8

s s The calculated core liquid level depression during the manometric period and the corresponding heatup were similar to the previous calculation. L However, heatup was not calculated during the later core boiloff period, as shown in Figures 5.9 through 5.13. This appears to be due to the extra liquid y ( mass mentioned previously that remained in the core. In conclusion, reducing the two-phase discharge coefficient from 0.8 to 1.0 provided better system pressure and accumulator actuation time predictions than the previous BE calculation. However, more liquid mass was calculated to remain in the core than was observed in the experiment, thereby preventing core heatup during the boiloff period.

6.0 CONCLUSION

S The following conclusions can be drawn based on the assessment of RELAP5YA against Semiscale Test S-LH-1: 1. The two major core uncovery periods observed in the test were predicted. The events causing the first uncovery, such as licuid holdup in the broken and intact loop steam generator tubes and pump loop seal plugging, were alt,o predicted. The phenomenon of core boiloff during the second core uncovery was partially c21culated. 2. In the base case BE calculation, the integrated mass leaving the break was well predicted. However, the primary side depressurization after break uncovery was overpredicted. It is suspected that this is mainly due to an overestimation of break energy outflow. The sensitivity study utilizing a two-phase discharge coefficient of 0.8 at the break substantially improved the primary pressure prediction, but left more mass in the core than was observed in the test. Therefore, either the mass or the j energy flow leaving the break is predicted, but not both. An { improved calculation has to be able to lower the energy flow without lowering the mass flow leaving the break. E-9

y 3._ Although the phenomenon of loop seal clearing was predicted, the timing of some of the early events was different than observed in the test. The liquid held up in the broken loop st+:am generator was calculated to drain early. Consequently, the broken loop seal ( ~ plugging was of shorter dura'. ion than in the test. In addition, the order of loop seal clearing in the calculation was the reverse of the experimental sequence. 4. The simulation of Test S-LH-1 is very sensitive to slight perturbations in the break flow. This is because the semiscale facility la tall and skinny, and hence, small changes in fluid mass tre.nslate to relatively large change 6 in liquid levels. In larger scale facilities, small perturbations in cass are not expected to { have such a significant impact on system characteristics. E-10

+ W w REFERENCES F 1. Fernandes, R. T., et al., "RELAP5YA - A Computer Program for LWR System L Thermal-Hydraulic Analysis," YAEC-1300P, October 1982. 2. Letter, J. A. Kay (YAEC) to J. A. Zwolinski_(NRC) on Transmittal of Responses to 33 NRC Questions on RELAP5YA, April 30, 1985. 3. Letter, J. A. Kay (YAEC) to J. A. Zwolinski (NRC) on Transmittal of { Responses to 26 NRC Questions on RELAP5YA, April 30, 1985. 4. Letter, G. Papanic (YAEC) to J. A. Zwolinski (NRC) on Transmittal of hosponses to 80 NRC Questions on RELAPSYA, July 1, 1985. 5. Letter, G. Papanic (YAEC) to J. A. Zwolinski (NRC) on Transmittal of Responses to 15 NRC Questions on RELAP5YA,' August 15, 1985. 6. Letter, G. Papanic (YAEC) to J. A. Zwolinski (NRC) on Transmittal of Responses to 43 NRC Questions on RELAP5YA, November 1, 1985. 7. Letter, R. W. Capstick (YAEC) to D. R. Muller (NRC) or Transmittal of Responses to 39 Additional BWR Questions on RELAP5YA, December 31, 1985. f 8. Yoomis, G. G., and J. E. Streit "Results of Semiscale MOD-20 Small Break (5%) Loss-of-coolant Accident Experiments S-LH-1 and S-LH-2." NUREG/CR-4438 or EGG-2424. November 1985. 9. Ransom, V., et al., "RELAPS/ MOD 2 Code !!anual," EGG-SAAM-6377, April 1984. 10. Beucher, T. J., J. R. Wolf, "Experimental Operating Specifications for Semiscale MOD-2C Feedwater and Steam Line Break Experiment Series," EGG-SEMI-6625 May 1984. 11. Personal Conraunication, G. G. Loomir (EG&G) to M. A. Langerman (ITI) and R. K. Sundaram (YAEC), June 1986. 12. Leonard, M. T., "RELAP5 Standard Model Description for the Semiscale MOD-2A System," EGG-SEMI-5692, December 1981. 9 e i

jll1 i t i ,l.a u. 3. d g5 t, 2 s 7-- aQr-i c c s. 3 r i-Q 'T 8 s i-2. 3 e g1).'_- jO ga n o P o o / m o m _ O r im " c. _ L Q(e, g e } 6 i T C o A j n T I J = _. P I 3._ _ i m f r I } f ? _ n L p e U a I J2.' c

i

, c" s ,a o s n m I s o c &,s o e i 2 z h, c ~ c i a t H c e 7 0 a m c,, i-1 l 4 s 7Q, ' - J u o 4 i' . c < c . a e .J 7 l la E C my %a a F S o P g 3 ss cb[c A 3 c a e. L c' E c o R 3 t 7, [ s c 1J l o Eg sI r _tde, e - f g l l o. 4 s i s, n 6 _d s c c c 3 . f e C' c c m o r' c' ~ C 1 r a 2 ~ Ip [ c c l' ga i q a., ,^ 2 7 i _33 ll D i c ~, I 2 i pF n s o o c l' l i i c t 1. s a c z 1 c i l a o d 1 i-l w o 1 t N 2 s u v>- r c o c t c a A c. vu i c c 1 o ' f. w 3 1 mS 3 8 o' s fa c t e 0 c r >L c "' I-i

l s

3 eest 3' s ~ u e 5 sg 3 et; s g 2 at c oe i l* J o P F 3 L m m r NP

o. o f C

i I e c c p D I c m O 6 l L vs O 7 c H N . M f' i J gp# 4 I . e @iI I c E e*: c .e a k t O n w R o,f s af u L a -, _=EG . x &p' >_*_.n. bm 3 w LL i. r @c c a Mb~ \\

e a r-SEMISCALE TEST S-LH-1 INPUT:R5IJS26 CODE:G4F(FTN5) - e.. ' "?. PRESSURISER DELTAP O,; e n n 1: m s. -a u o { g_ sc Q._ a9 s m- -'~ LPRZ U a-o cNTRt. VAR 0000005N c3 i., or 0 ~ w v, o - ora u, u, o. ww w-o_ c .-u o. I i i I I i -50.0 0.0 50.0 100.0 150.0 200.0 250.0 300.0 TIME RFTER BRERK (S) Figure 4.1: Comparison of Measured and Calculated (RSACC4F,BE) Pressurizer DELTAP

~ SEMISCALE TEST S-LH-1 INPUT:RSIJS26 CODE:G4F(FTN5) i UPPER HEAD LEVEL 9 I I I I I I Osn ~~~~~ --s,,' ~\\ 13 O O s. ~ - - 8-n v 2- .s,- o-9 o_ _._) W W" s. y Wo N _)0-am ~ y xW LUH 'N ~ T O, a-o CNTRLVfM 000000005 s. 's o- 'N g in ~~ gm C ',N D g g g 3 O-m 9o* ~ I I I I I I -50.0 0.0 50.0 100.0 150.0 200.0 250.0 300.0 TIME RFTER BRERK (S) Figure 4.2: Comparisce of Measured and Calculated (RSAGG4F,BE) Upper Head Levels

m SEMISCALE TEST S-LH-1 INPUT:R5IJS26 CODE:G4F(FTN5) ~ BYPRSS LIOUID FRRCTION m. i i i i i i i R z-o H oo ^%. g g_ m u o e m 55-O o VOIDF 181010000 ] VOIDF 181020000 u, s - 1 u, c Q_5" O-o. O -100.0-sb.o d.o s5.0 ido.o is'o.o 2do.o 2s'o.o 300.0 TIME RFTER BRERK (S) Figure 4.3: (RSAGG4F,BE) Bypass Liquid Fractions

SEMISCALE S-LH-1 TEST ~ INPUT:RSIJS26 C00E:G4F(FTNS) 4 BLSG UPSIDE DELTRP 2: 9 o i i i i i o N ... r o ..,..,,.--w.....,3.v-, tn _ N -~ O, -v./ ~ms.i > / nm.s,rv.~. i ^Go 1:: to i O_ ~ li Mc - LBPLH N-i:! CNTRLVf1R 000000511 w i O_ - i __. LBPSH gg i E-. __1 8 - (u,, y-l i. C3 R m J. o in _ .'.,,s e C3 N t, Q_ (f) o* r., D tn ' ^ b., 5 r......... ~--.. 9 tn - O o Sb.0 100.0 150.0 200.0 250.0 300.0 ~ -100.0-5b.0 0.0 TIME RFTER BRERK (S) Figure 4.4: Comparisc1 of Measured and Calculated (RSAGG4F,BE) BLSC Upside DELTAP i

4 't. ..3 MigE,b s (,* .u- , i .a O L I 1 1 I I I 1 ,i n n L / m_ i C. ( C [ m N O Z an e H ogs -8 as dSLb L m -e do L_ ~ \\ me HT w O. CD E3 (D G m L.j.. C 'i .C H (.L.3o d -' El x 8 cc = f*' ;I -o =$. E L.) O (- l O t U O. g-) w y a o,1 En ~~... ~O l Oy y *, (p CD ,,,,a;. - ~w aw H 3y y

'e.....f"2...

L_

o. <

a 3., 38. - O. C 50 J g 3. C u. o o (o f."r -@ U (D (N g- .3 z y 2.< - e,n a 9, ..:t.............................................._'_.. r H (n LD \\'c. u t C3 53 f .,;. -Q.. ~~3 O E ae S. h

e$

HOa

)i

./ Z 'J:. O. 1 i [ k i e 7 O. x O ( k O ( i .e ( [ l l l l l I l O*09 0'0S 0*0k 0*0E 0'0E 0'01 0'0 0' 01 F 0E-( (UdM) d81'130 30lSNM00 ) E-17

4 ) ' ',t.$:n '. ! o e ) o. L g i i r (, _~ = o. O ~ tn EE E M m 5b $8_- a

^

.-4, Z --@ D 3 L 3 l g N" j Ie@~ a 1 CD M M ~ l C O m O @g-s.~, a s a CC = Ho g O. D g D D o a y rd H p u H 8 L-f, y __) m C 5 f' Cod N oU C.3 l-(nq>a i./. u s..) ~8 r s [ r tn g _ / ~H 2 U$m L o (n g o N .5) o m r og ( p z g O ,/*** -S ( ,,,eees.*****" C.o f i i I 0*ObOI 0'009 0'009 0'00) 0'00E 0'O 0'00E-i (WO) S13A31 0InDIl 3901 SS19 E-18 1

m 'l W 3 Gt".: ,t. t. u. [ + o. 'I I i 1 l { s' ( 4 4 l I / )

o..

.i , _a 81 l\\ i ln E z o gr. be$Y r H 5 L 2 \\f 0y m k 't ^lo ' H va I . ( O C.D (n

4

- y.1 I,I ~?2 T HW r ~o w a. H-a mM ~oJ ,lf 5Sh $U Q w T W ,,.,2 O 00

Q 1

C-o m g oe =w ~ D ,.

l

~U Nf Q. 1 r H 14 J /' 1 = a-0g ( D O (.O ( \\

o. C G

m m (n ~ l \\ -o w r9 W ww ( ry1 ~ w~~ 9 y ww r E O !/ g { o H O t G-L Z ~ o. -o W r i l 4 o. .I o I o ( 0'00E 0*NSE 0'NOE 0*NSI 0*N01 0*bS O'O (UdM) dull 30 30ISdn g-19

n

</ +.. 4. t.~. s. s ;-

,.m u o _d t I t-1 I I g k, I' O. I I_ o-g r ( ) 1 1 = .m h C 3 [- Z 9 M a - o. H L. E 5' a< d !3 a a w-da L Hvg o, O ve am 03C m

~ '

W g i. l o = in M H M y-e 1 -[ g O U y "Oa s'....... 4-j} iO y ,, ', 111'. ~, C CD I y ~" fC ~8 x

2

~ (M h- [.-..'.S am U o% H ( Z O. c - < ( U E L_ gg s, gn -J o '...,.) U5 C a -O W O CD tn m (n g.D - (,'.~~" p g O (N C Jf. ~ E7 L1 y ( gg-l o e a .I .s ee er9* 3 a { CL i .': $l o. Z f-M 's -o LA } f. ) ( i. ) o. l-i l' 8 s' 1 I i i i I i [ 0'04 0'09 0'OS O'Ok O'0E 0*02 0*01 0'0 ( (UdM) d81130 30ISNM00'

we + eenm E-20

c-w - v , w SEMISCRLE TEST S-LH-1 INPUT:RSIJS26 CODE:G4F(FTN5) ILSG TUBE LEVELS; UP(SOLID 1,00HN(OflSH) ~ Q I f t I I 9 e oe T \\ r O o-o \\ g f^. m s, N. (n \\ '/ 1 a. s. y > 8-i,i CNTRLVAR 000000007 j* CNTRLVAR 000000008 O9 ~ \\ ~o_ 3? a m ~ b d 9 o g-oa D H ~ Q l C.D m (n a- __J l - o. 8 i i i ? 0.0 60.0 100.0 150.0 200.0 250.0 300.0 i i TIME RFTER BRERK (S) Figure 4.9: (RSAGG4F,BE) ILSG Tube Levels

SEMISCRLE S-LH-1 TEST INPUT:RSIJS26 CODE:G4F(FTN5) BL PUMP UPSIDE DELTAP .x 0 c; i i i i i i i n lb M l-o ..w.-,.-..,, m-OJ [ t O_ o M c" - "N 4*6*. ! i O_[R ~ ,v:t I H J m-A# D m 4 61 o BLSUP C d. _. s-e CNTRLVfR 000000514 N ~.- (D O_ l D o.- ~ C3 o-1 I I I I I I I -100.0 -50.0 0.0 50.0 100.0 150.0 200.0 250.0 300.0 TIME RFTER BRERK (S) j Figure 4.10: Comparison of Measured and Calculated (RSAGG4F,BE) BLPS Upside DELTAP l

b me b 9 4. b 0 r= L Q 1 l I t i i .ml l C. EE 2 r m 5 z2 m m i -8 m ? (u->g!f S-3 (s/ u D a s C uO j .N y a Co s o g HU-C O. z CD m. C w _.1 O g 8 o-H g x g e w a_ L_ v _) o.. 3 C w QU Z. C) b _>_1 m U / 7 / ~O- ,W .\\ g ( EW d~ yZ a ,mq g x eu ,.8* sue ,...je tn m O. ( _O I i i i i 0'00E 0'00E 0'001 0*0 0*001- 0'00E-0'000- [ (WO) S13A31 0InDIl Sd19 E-23 / )

~ ,m SEMISCALE S-LH-1 TEST INPUT:RSIJS26 CODE:G4F(FTNS) IL PUMP UPSIDE.DELTAP T og i i i i i i i m a o a a o -1 g_ W m a-G- M o m-- -_ m - v;- x a- >.? \\. e- _.2 o i oo i y 1 L_1 o - I -+ .._. ILSUP i O e-eCNTRLVAR 000000510 i ^ j (n O-

1..

j 9o ,_&--...-.o u;- o. o -100.0-s5.0 d.o sb.o 10o.0 15o.0 2do.o 2s'o.o, 300.0 TIME RFTER BRERK (S) Figure 4.12: Comparison of Measured and Calculated (RSACC4F,BE) ILPS Upside DELTAP

ww f SEMISCALE TEST S-LH-1 J. INPUT:R51JS26 CODE:G4F(FTN51 ILPS LEVELS;UP(SOLID,DOWN(DflSH) 4 ) o n o i i i i i o -f.t o 0 rn o CNTRLVf1R 000000010 e-e CNTRLVfR 000000009 u) o _ _J o L_] L3 i ]o l ~ o l o0 l ~o- _a o v, ' O_ o 1 ~o-gmb 8 ) i i i i 0.0 50.0 100.0 150.0 200.0 250.0 300.0 TIME RFTER BRERK (S) Figure 4.13: (RSAGG4F,BE) ILPS Levels ..in.i-.. .i--

0 '4#f e. _ a4.c :* -, 5

s 4

D

e e

O. O 1 1 I I N 8 1 ~ \\(.t.t O _g c) r., 's. O ~ co m O a -m g n =- I.?> 9 a da _O ~C R ^ 12 ~ i o D eo .'r,N. Zv e5 J G.. u J O wa i g c . Y. @Q \\ OJ O HO .*~s, f.L3 E$ o m N. 9O Y (1.) 3 g O em H 0-W (& O$ O o ~ m A.) J / g H a8 _J J ./ 96 2% ceu _Oc 3 "M o tN w - 3 O 0mm ~ 7. U /g*[ t U ~ O. - 3 Em y ct _OhA e r ( H a v M { ( z i 'O } ~ I O _O \\ s u YYA-m q _O O ~ O. O I I I I ( O'001-0'002-0'00E-0'00k-0*00S- 0'009-(WD) 13A31 OIn0Il 1]SS3A (.- E-26

m -- s SEMISCRLE TEST S-LH-1 ~ INPUT:R5IJS26 CODE:G4F(FTNS) DOWNCOMER LIOUID LEVEL 8 a. I I I I I I I I f O ++,. r -I i p a ) d-g

~ r EY

,f s O ..!!]Vil a l w g -- !I" a ?, p,L&4' D?' y '? xo l .p:+ N. / w d_ Eo f ,m Oy J O l z W - L29578 Z e--e CNTRLVfR 000000003 ]gg_ o T O b8 i i r i i i i i 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.0 1000.0 fIME RFTER BRERK [S) Figure 4.15: Comparison of Measured and Calculated (RSAGG4F,BE) Downconer Liquid Level

i 't .s se g

)

I s L a g 4./' g s a b ( g w d 3 o -i -N 88 [#- s d ^ M z.0 1 i oO w l H 18 IL ee J[@- my v5 = C e i O O,C.D L _ i oe e< W g8 em je ( w e 0% sU a as Uer ND-T L t$ ( [ J (.O O C eo C70 O a< U bU tn [. W e.. J - @E uD O. b e Oz -8 ( 8 m m g a M O 8 d n a. t O~b 9 ga a E o. O m i i Ol'0 90 0 90 0 FO*O E0'0 00'0 ( (S/SM) M0lJ 80191RWR008 ll DM'" ( E-28 /

e

s i

e - t, l 4 .s f e e W (_ s*,, ~ p.e****#,. ...F a=.__ _ -g -e m <?*.y a M..* 9 G ^ O 0 i>? Ln z ~c =o8 H t. O Mb .f m IL jf a m,8 Ie c. m aw Af &O5 I y3 ^4,' %, m to y jj WOL ~~ O . o. c y 8 wau HOM og e< g o ja O-H in CD m U C w ( ( H J oy G E on o -8 H n O- / _] O eL wg COO C E.< c.3 m x o e a: to 7J OM .-.co - Og EO H OH 7 Os e -0 D O G 1 8 m u z a g -a O o. ta. E O da -a og C O. OO ~ l l l l Ol'0 90'0 90'0 FO*O E0'O 00'O (S/S>l) M0lJ 80181nWn008 19 p E-29 l

\\lllj)ll f'- 0 0 ~ 0 0 ~ ~ 1 0 0 00 0 4,W i 0 0 t 0 9 00 0 1 4 0 ) 5 i 0 m"~ - I IJ 8 N PW d HO e T IL 0 t F WW 0 la ( i F e 0 u I c 7 l 14 ) a n Ce - G

0. S t

i d a H:E i) 0 nR i LE 0 a T I w 6K - D d o A R el SO R p

0. E rF u

C sI 0R aP T i WI 0 eH S 5 B M O L E L fI T

0. R o)

F E n B, E 0 o E 0T i sF I I 4 F i4 L P . !i : :' rG A

0. R aG H

pA f C6 mS 0E oR S2 Li C( i 0 M IS 3 I MJ ~ I

0. T 8

EI 1 SS 0 4 R i 0 i 2 e ~ r T ug 0 U i F 0 P 0 I N I 1 I .\\iN 'r- 0 2 i: 0 I - " 0 00 1 ,F ^ r3d m o o g o o o8a oog +o9o moo 8oRo o C I 3C Qxex yyg 'oJ_L nV(2 ( ,i !l md,o i\\l;l) ltf1lll 1

1i l

.;f: -

J . P. - t J 1. 0 I_ 0 ~ ~ 0 0 1 1 C 0 0 00 0 0 0 i m I 0 9 00 0 1 if 5 ) 0 i f 0 S I IJ 8 N k' PW HO d T + BL 0 e F FF t y ( f NN 0 a i l F pp o 0 u I 7 c 14 y ) l g c ae

0. S

- G i, Ct a H:E 0( dR LE 0 n 1 TI aw - D 6 K o A dl SO

0. R eF R

r C E uI T 0R sP WI 0 aH S 5B e O ML E /

0. E

.B L f R T F o)E E / 0T n B, 0 i o I I sF .L 4F P i4 A

0. R rG H

aG C6 pA mS S2 0E oR / LI i 0 MJ 3 M C( IS B I j

0. T EI 9

1 SS l 0 R / 4 i 0 I 2 e r T 0 u U g i P 0 F i 0 N I 1 I bla-l' 0 i C 0 I C 0 0 0 1 ~ N-o.s ag s o.M o* o*~ ^ bw, Ltt% 3Od nn[E ).. ,nN (( ( ^- ,d~ ~ .1l l(

' ~ ' W e e L' o. og i E o o .o 1 O

]

o. o o n Q M. -. CD g .e W () O.51 9 8 o 8 i a R 30 I G t ~ lo qm 7, ~z r[_u 1 a a-

== o eE JL M-oM $E I fh

o. c W

ag memu ~@ T Ea H M mQ m o A-m o^ a

c 0%

a$. H D

y o

Z -8 s n weS_ +t as age

o. C

.a 6" J c2z / -g u l 0mmE-n& R n e. E&aD OH =* Ua d Cn a-a g z m = M 4 ) ~ o, E -8 0 -d ( C ( a C3 o [ ci i i i i i S*21 0'SI S*EI 0'01 S*Z O'S S*C 0*O wam sanssasa j ~ e-n {

o O. O U f f I I g x [ O O OO _OO O. CD f C - l O NW r" d -O D cn c l. C b h 9 If -8 is iE o ~ b n-y ~- 3 90 0) bb5 <{1 Je5_ _g-is p ex at iw e C TE Wo o a HO3 - dd m ._1 - .( O a= U k D (l E O cr y2 H r M es w E_ 3 _O H J A S h

SJ f

CO$ 19 O oss 3 _d U 89 W 3 7 [ S*U$ Y 05 [ 2 9H m m -O 2- ..o ? a z .i O C h _g 7.g -m C N O O i l i I 0'0001 0'009 0'009 0'00t 0*002 0*O (EW/SN) IllSN30 0In1J MU389 E-33

w 1w SEMISCALE TEST S-LH-1 INPUT:RSIJS26 CODE:G4F(FTN5) BRERK MRSS FLOW RATE m D I I I I I I I I f f s ^ O. _ (n x-C9 Mn i -m LJ ~ H[o i WBRK [U-ki ~ e-o TLOWJ 375000000 ~ z O i 3 to t k h_ I i o W (D i 1. (D @ I !,b; ; O_y,?l%l yF Cn._ i = ? k \\,gl!. O i, r. i.- !l p M h I C

ji in

? LJ m i g 5 (D in g i O I E,%'~~Y O .:Yglta.*w-+-J. r f 1[V ? c o E ~O i i i I I -50.0 0.0 50.0 100.0 150.0 200.0 250.0 300.0 350.0 400.0 450.0 500.0 i TIME RFTER BRERK (S) Figure 4.22: Comparison of Measured and Calculated (RSAGG4F,BE) Break Mass Flow Rate

2 SEMISCRLE TEST S-LH-1 f INPUT:RSIJS26 CODE:G4FIFTNS) c 4. ' INTEGRATED BRERK MRSS 9 in i i i t i I i I I D o. g~ O tn y~ " O. ~ D tn _ W ". C E o. E @a-M M M CD o. in - ri O C3 " M IFBRKI E-* O e--e CNTRLVAR 000000600 C M sn _ O M E-o Z g* _ H N t l I 9 I 1 3 0, i i i i i i i -50.0 0.0 50.0 100.0 150.0 200.0 250.0 300.0 350.0 400.0 450.0 500.0 TIME RFTER BRERK [S) Figure 4.23 Comparison of Measured and Calculated (RSAGC4F,BE) Integrated Break Mass

I'd b - W t II 8-e + W s o. r u ( i e i gm o o* Eo ~ n h '2 : ~ o o 8 $5 f ~O ,c. u a o -c m "I -8 s to lW f ca Ih: m. o Je8-x

"o i

m oC*U E$ D86 / -8 e a f om m 5 H o c. / m co oc g i ca g=* (n e wh w i / -g u t*0 fs O 8 U / " 'N 06 I / k oC J o / [ CRM O/'/ -8 o O (n ME N (.0 7 co -c;m- ? / RH E m.. co f e4 y$ _o D l O O E / O .N A tl o .Z () 3 o -= I O mms ( en O u m k Ln ) c I' A CD Q -g l m g) lac O, c. i la, 8 [] l l l l l l l l l 0*8 0*/ 0*9 0'S 0*k O'E 0*E 0*I 0'0 I waW '3anssasa E-36

~ n vw SEMISCALE TEST S-LH-1 INPUT:RSIJS26 C00E:G4F(FTNS) ILSG SECONDARY PRESSURE o.

.._G*

I 1 I I i i i i l' I ^* o t .__= n- ~... [ ~ ~ _O c j _g gg_

y-c.

a-Q- r in - wo e 4- ? 5 PILSGS t: mo o--oP 611010000 y g- __.---- R O V S E T o_ o o J-o. o -100.00.0 100.0 200.0 300.0 400.0 SNO.0 6d0.0 7N0.0 800.0 900.01000.0. TIME RFTER BRERK, SEC Figure 4.25: Comparison of Measured and Calculated (RSAGG4F,BE) ILSG Secondary Pressure i,,,

1' ~. i-0 0 00 1 00 . 0 30 0 i 0 0 i 0 9 0 u 00 0 d, R 0 ee i 0 t r f 1V 8 au i 8L l t 1R ua 3T cr CN 0 l e )6 Tc ap 5 = 0 Cm E 0) e i 1N dT i D 7 = S n T O c ad F ( a H( N dl LF 0 eC r 0K u6 - 4 M i U 0R s SG Ci 6 ae E ed F E Mo R N 1 TC f! 8

0. B f

S0 / o) 1 EC 0 E n B,M i 0R T o .i 5 sFC L E i4 E E6 T rG1 aG8 L2 L

0. F pA1 mS S1 0R oRL A

f J i CI Ti 0 C( E N 4 SS E E IR M M 6

0. I 2 M:

I T EU R ~ i 0T 4 E ~ 0 e SP i P 3 r N X u I g E i 0 F 0 i 0 i 2 = 0 0 i 0 c 1 ~ 0 i \\ \\I 0 _c oo= g oo OoR odow 0dg od? oHcMwGEww Q 3_o c Tw-lll ll

lill l w ~ 0 0 I 0 0 1 0 1 03 0 0 0 I 0 9 I 0 g 000 0 d,ee 0 t r R I f 0 au 8 V. o 8 l t I 2t ua 2R cr 3T l e cN 0 ap C m 7 TC 3 0 e ) S E

0) dT I

n 1N 7 ad I D = S - T O O a F ( dl HI N eC LF L 0 r u7 0K i\\i s - 4 I} I M 0R ae D.1 SG:CI 6 ed EMo N E 0.R f T0 8 / B o) 2 / E S0 2 / 0 n B, M E C 0RosFC I i /'[ Ei4rG8 T 5 1 LE ,/;I

0. T aG2 E6

/ pA2 F mS N L 2 oRL L 0R C( E S A J R I 0 CI T1 4 E N SS 7 E M2 IR M

0. I M:

4 I T 0T EU R e I E 0 r SP P 3 u i N g X i I E F 0 0 I 02 I 0 0 I 0 l 1 'lii 0 _ 0 I _l ~ o,Co 9oom O,gk o do* o go R8v ) t g" $aHEoLoE1H o[)__O ca _ L ?w* 11\\I< tIIjl:

1J .w ~ ~ 6 ~ q 0 . ~ 0 00 1 2 0 03 n 0 I0 0 0 I 00 O 9 00 0 R 0 A 0 I 8V C 8 d 2L ee 2R t r 3T au CN 8 TC 0 l t ua e 0 cr E i l e

0) ap 1)

I D - 5 7 Cm S e O WC N H N [ dT T n 0 LF ad a M 0Kdl ( SF CI 0Rr eC 4 6 G E u8 T 8 / s 2 / R ae 2 /

0. B M o ed SE

/ N 0 /l/ 0R o) f T .I 62 /l/ 5 L E E E n B, M ES / T o /l sF C LJ L

0. F i 4 I

rG8 A R 0R aG 2 5 CR TI i 0 pA 2 m5 N 4 S: E E oR L C( E IT M M UP

0. I M

I EN R SI I i 0T8 E 0 2 P 3 4 XE e 0 rug i 0 i 0 F I 2 0 0

0 I

= 1 0 i- .i\\i 0 ~ _ c 9 COG 98e 9gs O 8e R gn 98# 3 t y hgHTMWOrws Q[3O m1o ll

l\\!ll ~ y *- ~ o 0 0 0 0 1 m 2 0 03 i 0 0 0 0 I 9 0 d-00 0 0 d, R 0 ee i 0 t r f 8I 3V w 8 A au ) 5L l t 5 2R ua E 5T cr ~ N BN ~ l e T 0 D TC ~ ap 1 O o Cm F i 0 e NI 0) dT c 7 n ( .c S ad F a H i [ dl 4 L M s.' g' 0 eC G 0K u8 r S:C i 0R ae b s I E 6 E ed 3 Mo D R N T S O S 2 /.

0. B o )E f

C E / 0 n B,M i 0R o T .I 5 sFC L / E i4 rG3

0. T E

E aG5 6 / F pA2 L /' ! 0R oRL mS 2 A L 0 C( E S C RI N

0. M J

4 E S T I I N 9 5 2 M E I R 4 0T E M ~ ~ i S: ~ 0 e II T .~ 3 r R u U g E i P 0 F P N X I e 0 EI 0 l' 2 ii i J 0 i 0 0 I a 1 0 N 0 t ?og ooN odoW O oom ?g o CO 3 ( c 3TJ_CJ + Ma H[oWo_ EWE c C m1~ jlll

~ ~ 1 1 SEMISCRLE TEST S-LH-1 INPUT:RSIJS26 CODE:G4F(FTN5)

'J.

EXPERIMENTRL EL. 351 CM, NODE 11 -T o C3 I I I I I I I I i o m .o- -o y* TB3351 e-o HTTEtiP 150001106 Mo QC 3 g-ss /. e N 0-s. oc / W / s o_ o / N Ed[\\ \\ rii y o-C Hw / a 6A.. ~ Q- _3 o. c3 o _ c b "g -_ m o d I I I I I I I I I l 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.0 1000.0 l TIME RFTER BRERK (S) Figure 4.30: Comparison of Measured and Calculated (RSAGG4F,BE) Node 11 Clad Temperature, EL 351 CM 1

l I l(!l l z, *.- 0 ~ ,0 0 0 1 r 0 0 0 m 0 i 0 0 9 I 0 1 0 0 3 0 i 36 0 11 8 I P d U 0 e PP t 0 ae i o 0 l r ) 7 uu I cs 5 ) l s 1N c

0. S ae

- T Cr F 0( P d H( E \\ i 0 nm LF R 6K au i U n 4 S

0. R d e SG S el E

rP

E 0R u

E 0 sr T0 R s B ae i Pt 5 ep S0 \\ Mp EC M U

0. R f

T U E o) 0 E N 0T n B, i E o E5 Li 4F sF ~ E .,x

0. R L 2 i4 P

S rG R J aG CI R pA 0E mS i 0 M oR N S5 P ~ 3 C( IR P 1 I M: U T

0. T EU 1

0 SP 0 5 i ~ N 2 e 1 ,m I ru 0 g i 0 F i 0 1 i ~ 0 i 0 0 I 0 0 ,1 o- "W mu ag m-o ? m* m o.m ao5 MKommMma_ rA~ n iIli[ l

~- ,w SEMISCRLE TEST S-LH-1 INPUT:RSIJS25 C00E:G4F(FTN5) o IL RCCUMULRTOR FLOW

7 I

i t I I I I I I I o m (f) co N U. _ CD o M ik j 2 O8 3 k / d c;- [, l \\ f1CFLIL 1 MFLOWJ 430010000 1 1 Hw U T U D l l E I \\, D i U ON [ U. _ O } g l ~ 8 f i I i i i I i i i -100.0 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.0 1000.0 TIME RFTER BRERK (S) Figure 5.2: Comparison of Measured and Calculated (RSAGG4F,BE) IL Accumulator Flow

\\ l1fI J !, * :y t. 1 f T I 0 0 T 0 \\' 0 f \\ s\\ 1 \\'s 0 i M ii 0 3 ij' 0 i 1 ? 9 J,E' 0 m Ilf i 0 e E 0 p d l 8 ta E ) ^ l 0 uw 5 co N 0 ll T l i 0 aF C F 7 r 1 ) d o ( - F nt H4

0. S aa l

0( d u LGW / i 0 em i

O

- EL j' 1 6K ru uc SDF !ll R sc aA

0. E O

l e TCR i ML OI 0R B S 0 f T 5 B o) E R E n B, T LU

0. R osF E

i4 M 0 E 0T rG UI i aG L C 4F pA R5C m5 C2R

0. R oR C(

S S JL 0E I IB I i 0 M 3 M5 3 I 5 ER S:

0. T e

T r U 0 0 u P 0 i 0 g I i 0 2 F N 0 I 1 0 0 03 0 5 i 0 I 1 LJ 8W 0 LO FL CF i 0 I RN 0 00 1 o",O 0O. @.o <o. mO o 85 D DNY-1O]__L MOH[gaEoET 1CD ( m4* ,,llll1'l i

y-w* 9 w* 1 [ O. O I i 1 - O O [ \\ 9 .(, O [ C - O m g "?., O ,e 3 E O _ O => 2 ,-4

s tn u.

m" t, ro s h 5- ? 8-k". -b U N .m I (,s 9 g c' IP r,, N \\', Q u U U.. w \\ OM a't. i u Oc too>- \\s. j w O 0J W HQ w o x.'f Og o^ l' Ch N D O& 8t: ~ 0~ i M Uy J / E O[ .d y J J ou [mu OC Uv O Nw-( O i-wm y <'r# 0 7W r ED ,/ Om 5 ~~> e WW y OH 8 z O a (fy H (s M to w D ...I g o Z sr,, a - O c..... 4 O _O O ( O. O i 4 i I 0*001-0'00E-0'00C-0'00E-0'00S- 0'009-(WD) 13A31 0100I1 13SS3A E-46

b 5 a};e -- { ~ h Y a 5 e L O. O I i 1 I o [ O 4 O. v. [ a~> _O O G T*h .o G ed ed

Ni?

O aW M > =:E., e I O 4; e> O S 3 "d 63 m $~ = z w, 25 O.

i. 5a

-S n~%. -8 ,*w w a o N (n 8 i o_

t.

EO I*d li. JT> o. Eg i iw u 4 Ox zg I (D a J-Oc e O Wy og a [-* O - '>.s y c so O CD E$ 3.> CD D -e w a W s to y .n M H O6 O L_ ~3 d CC y G LD a "O ON CD$o-

y Z

~1 z e Ow n E LA Q ds 3 M5 w O (n w F) g',. -) Z O. M O aN O. -g D. ~****=ses_ C i i i I l O*O 0*001-0'002-0'00E-0'00F- 0'00S-( (WD) 13A31 83WODNM00 E-47 k

~ ~w + SEMISCALE TEST S-LH-1 INPUT:R5IJS25 CODEiG4F(FTN5) o BRERK MRSS FLOW RATE .}. t I I I I t i I t 1 .~ 9 a g m-y~ g ul,15_ u2 cc E o. M @- T~ Li.3 M G3 a. in - c., 5 C3 " LJ MFBRKI Eo e--o CNTRLVAR 000000600 [ @- ~ Me Li.] E-o U C i i i i i i i i i I -50.0 0.0 50.0 100.0 150.0 200.0 250.0 300.0 35G.0 400.0 450.0 500.0 TIME RFTER BRERK [S) l l Figure 5.6: Comparison of Measured and Calculated (RSAGG4F,BE) Integrated Break Mass i

~ r, ...v SEMISCRLE TEST S-LH-1 INPUT:RSIJS25 CODE:G4F(FIN 5) BRERK UPSTRERM FLUID DENSITY g i i i i i i i i i i o tn a t N 8-CD w W. 9-4 % ' jy H x ~ ffk wo H c5_ l ~ ~ l I RB79M c e RHO 362010000 o O g ~ h_ ~ 7 O G w 3 0 __3 L_ M O CE o - ) om QC 00 i \\ '*4h%%uus 9 o I I I I I I I I I I I -100.0 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.0 1000.0 TIME RFTER BRERK (S) l Figure 5.7: Comparison of Measured and Calculated l (RSAGG4F,BE) Break Upstream Fluid Density I

-1 I-i _ _ f ' t. r SEMISCALE TEST S-LH-1 INPUT:RSIJS25 C00E:G4F(FTN51 BRERK MRSS FLOW RATE in "i i i i i i i i i ^O (D x-C9 Mn i - n2 LJ ~ H[o M o. - MFBE z" M. o-oMFLOWJ 375000000 O i _) tn i. w %- \\ ? o i 8 D (D[O I .! j.. l,, E ! 5 MI.f ?[{IYd, A1/ !?g ? y [ tn I:' S l6 Z CD 9 i ~ O O f v 'I f 1 % N Is N 8 v. 4 '***'# h "_ v+>i~ ~.r,p oo j o -:- ry -50.0 0.0 50.0 100.0 150.0 200.0 250.0 300.0 350.0 400.0 450.0 500.0 TIME RFTER BRERK (S) Figure 5.8: Comparison of Measured and Calculated (RSAGG4F,BE) Break Mass Flow Rate i I l

lI1l

i?.

0 m 0 00 1 0 0 030 0 0 i 1 0 00 C 9 00 0 R d, A 0 ee I i 1V 0 t r 8L C 8 au 1R l t 3T ua 6 CN cr 0 ) TC l e ap 5 3 E 0 Cm 1N D I 0) dT e - T O c 7 n S H (F N ad ( a LF dl 0 eC - 4 M 0K u6 r SG Ci I 0R s 6 ae E E ed TD \\' Mo 1 8 R N SO /

0. B f

1 / EC / o) i /; 0 E T .e n B,M L I 0R o E 5 sFC E E5 i4 T rG1 L2 L aG8 S1

0. F pA1 A Jf 0R oRL mS CI TI I 0 C( E N

SS 4 E IR E M ~ M 9 M: I EU R ~

0. I T

5 0T E 1 0 r SP e t P N X 3 u I E g i 0 F 0 i 1 02 0 0 t 1 0 O 1 0 . i iit 0 I O'c8 o ~ O= 9oO O* o= 9gm O 8 *r, ,, x- )

CgWQEW[4 OE]O E

md" l l I

S 9 $8 g 9 l - a C 1 t i I _o O -8 [ l O o O 'Eei %E 8 35 0 $2 ~ -v nw N es n O Om $g W W-i i U-O U ts W e e-() SO O I *D Z r w i [ 1.e l O yn C fy g, iW ^ m85-C,,,.. ge n u 2 O CD x.' s. M Ym H N w Og -p g j 5 W N j-mW ' ~... o e8= m 's. g <~ u,d OL EUa JG dx E x.. w - ,1 O C) z T l y WHW I E ~Dr We '/ C-CL Ez-fp e (D o_ - O y W n w W y C. -gn O. OO [] O. O 1 I" i I 0'006 0'009 0'004 0'009 0'00S 0'00k (M) 380lW83dW31 0910 E-52

~ c, ^ SEMISCALE TEST S-LH-1 2 INPUT:RSIJS25 CODE:G4F(FTN5)

..: ii o.

EXPERIMENTRL EL. 228 CM, NODE 8 o i t l 1 l t i f l

{ e,, ' *!?

o

L-m o

o- -o ~ w y v o TC3228 g e-e CNTRLVAR 000000302 -) o _ j HNt ct K i ta / 1 O_ o / ma E ,i w 8-w Hw l C / a ~~ ~. j cI-3 o, O o_ r g __ onn 1 1 o i o 1 1 I I I I I I I l 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.0 1000.0 i TIME RFTER BRERK (S) Figure 5.11: Comparison of Measured and Calculated (RSAGG4F,BE) Node 8 Clad Temperature, EL 228 CM

l, 1 4 t.- c ~ 0 000 1 20 0 i0 3 0 i 0 0 0 0 g 9 0 d, 0 ee 0 t r au R ) l t 8i IF 0 ua 5 c 0 cr i l e 8 ap E ~ N Cm D e T dT 0 O n F 1 Ni ad 0 a ( i 0) dl F 7 eC H c S r 4 L M u8 i ( G s S:C 0 ae ed E 0K Mo i N i 3 0R T 0 6 f S 0 E o)E /,'/ S 2 R n B,M E C o

0. B sFC T

i4 .i 0 rG3 L ./ 0R aG5 i E E / 5 pA2 E mS /,/,/>- 5 oRL L T C( E 2

0. F R

L S C Ri 0R J 2 S T i 0 1 I 4 I N E S 5 M E E R M e M

0. I r

S: u Ii 0T g T i 0 i R U 3 F E P P N I 0 XEi 0 i L 02 0 i 0 i C 0 1 0 !Iii 0 _ t 9gm q 8= ~ 98o O gm ob@ x _ t$HEtL'rws oE)_o a rJ m ll ll

.. ( M t (.., O. s-' ... i.:i ', m O. e l l g O F O e m O d _O ~ O 0) e /,- _OO [] co oe y

3. u m

[ Ln ~ 9 58 zs / _O u mbu_ E o N (n e { a 1 m .e a e.n. z O. c =o em iw* ( _OM .o G E_ C 0- [ Mc s 0 O [ o 'N-W m-Eu HU

a. 1 m

co g; (n m W n s N O x [ 1 Y g U g c in W y \\ H g w-5 [ __] \\

a. L cmp i

01 $85 t \\ O s<m om g_ sn W \\ vW OMJ (n 7 Z k) Ws UE O. ~ Zm W Z *g _OH m ~ (n " w - O c_ M E r C c x 3 [ w v O y ~ d = ~ O \\. _O m [ 8 a N o [ su Ob o s [ O u 0'006 0'000 0'002 0'009 0'009 0'00h I i wi [ (N) 3801983dW31 0W10 ( E-55

E M ATTACHMENT G CLASSES OF SMALL BREAK LOCAs AND THE EFFECT OF NONCONDENSIBLES ON THE PRIMARY SYSTEM BEHAVIOR { ( { l ( { I }

4 Classes of Small Break LOCAs and ~ the Effect of Noncondensibles .on the Primary System Behavior f The small-break LOCA, as generally defined includes any break in the PWR 3 2 pressure boundary that has an area of about 0.5 ft or less. This range of break areas encompasses all small lines Gat penetrate the reactor coolant system (RCS), including relief and safety valves, chcrging and letdown lines, drain lines and various instrumentation lines. Classification of Small Break LOCA by Pressure Response Three major classes of small-break LOCA are readily identified: (1) breaks { that are large enough to depressurize the RCS to the setpoint pressure of the cccumulators, (2) smaller breaks that lead to a quasi-steady pressure plateau for a relatively long time, and (3) breaks that may lead to RCS repressurization. [ Small Break LOCAs Resulting in PCS Depressurization For plants like Maine Yankee, the relatively large small-break LOCAs range [ in area from about 0 1 to 0.5 ft (Reference 1). While the exact size range for this class of breaks has not been determined for the Yankee Nuclear Power Station (YNPS), the lower end is on the order of 4 inch ID (Reference 2). The exact break size is determined by: Physical Design of NSSS o o Core Power f o HPS1 Pump Capacity o Accumulator injection setpoint Figure 1 presents an RCS pressure response for a typical break in this class. The break causes a rapid depressurization in the RCS, tripping the reactor ( cad generating an SIAS signal. The rate of depressurization slows somewhat cs the saturation pressure is reached (Point A). Steam is being produced by flashing of the liquid and by boiling in the core. Because of the large temperature dif ference across the steam generators and the reduced core power, sufficient cnergy is removed from the RCS to continue the depressurization. As the RCS pressure and temperature approach the secondary side values, the depressurization slows and then reaches a plateau. G-1

w P ge 2 The cessation of depressurization (Point B) occurs because the volumetric a flow out the break is insufficient to accommodate the steam production in the care region. The steam generator heat sink condenses the volume of steam not cecommodated by the leak. Once the liquid / vapor interface falls below the break olevation, the volumetric break flow increases and the RCS continues to depressurize (Point C). The depressurization continues until the safety injection tanks (SITS) initiate flow (Point D). Figure 2 presents the pressure in the primary system and the pressure in the steam generator (SG) secondary in the intact and broken loop for a break in this category at Maine Yankee. Figures 3 and 4 present the temperatures en the primary and secondary side of the steam generator tubes for the intact End broken loop, respectively. The system saturation temperature is also plotted. Primary / Secondary Heat Transfer In Reference 3 we assessed the effect of noncondensible gases on condensation hrat transfer. In order to have condensation heat transfer, the vapor on the primary side has to be at saturation conditions and the temperature in the SG sscondary has to be lower than the temperature on the primary side. Figures 3 and 4 show that condensation heat transfer can occur only for the first 160 ssconds of the transient. /,t this point in time, the accumulator has not actuated end there is no fuel failure. The clad temperature is well below 1600 F, the { point at which metal-water reaction becomes significant. Whatever hydrogen is generated by metal-water reaction is well compensated by the assumptions derived for this assessment. That is, we considered all air dis' solved in the RWST tank and assumed that no dissolved hydrogen in the primary coolant leaves at the break. As the primary system heats up, superheated steam will be formed in the core and the steam generator will alternate between being a sink or a source of heat for the primary system. During this time, the metal-water reaction rate would increase and fuel failure with release of fission gas may occur. The heat transfer modes on the steam generator primary side could be: G-2

P ge 3 e

1) Laminar Natural Convection

2) Turbulent Natural Convection

3) Forced Convection (Dittus Boelter)

Appendix A presents the heat transfer ratios in the SG assuming pure vapor or pure nitrogen or air on the primary side of the steam generators at various pressures and temperatures. The ratios of these heat transfer coefficients are within data uncertainties for these correlations. Hence, even if there are 100% noncondensibles on the primary side of the steam generators, the degradation of the single phase heat transfer is within data uncertainties ( (cbout i 10% to 130%) for these correlations. Depressurization Effects Appendix B presents the increase in primary system pressure produced by the presence of these noncondensibles. ( Accumulator actuation pressure for the break presented in Figures 2 through 4 is about 220 psia. The maximum increase in system pressure due to the presence ( of noncondensibles would be about 6.5 psia. This conservative estimate includes nl1 the fission and fill gas, all hydrogen in the primary coolant (including pressurizer), a calculated value for injected RWST air, and a calculated value for hydrogen produced by metal-water reaction. This increase in presstro, for a depressurization rate of about 1.25 psia /sec, would delay accumulator injection by five seconds. This delay would not affect peak cladding temperatures which occur beyond 100-200 seconds after accumulator actuation. In reality, the pressure increase will be smaller since most of the hydrogen dissolved in the primary coolant will be removed at the break during subcooled and two phase blowdown. Appendix B also presents the increase in pressure at the end of the transient. If all the noncondensibles are assumed to remain in the system and the system is assumed to be at 150 psia and at saturation condition, then the noncondensibles will produce an increase in pressure of about 21 psia. This increase in pressure assumes that all the noncondensibles remain in the system. By this time, the core is quenched and all the decay heat is removed at the break. G-3 ________A

L r L Page 4 Ve conclude that for relatively large SB LOCAs, noncondensibles gases have a minor impact on both system pressure and steam generator heat transfer. Small Break LOCAs Resulting in a Pressure Plateau A second class of small break LOCAs is defined as those in which the core recovers before the RCS pressure falls to the value that would actuate 2 the SIT's. For Maine Yankee these breaks range in area from about 0.02 ft 2 to about 0.1 ft. For YNPS, the range is about 0.5 inch to 4 inch ID break. In this class of breaks, the high pressure safety injection pumps and the steam gtnerator heat sinks are very important. Figure 5 presents RCS pressure versus time for a break size in this class. The initial portion of the transient is basically the same as for the larger breaks; a rapid depressurization accompanied by reactor trip and SIAS signal gsneration. The RCS pressure then establishes a plateau dictated by secondary-water tcmperature while liquid is still passing through the break. However, when the break is uncovered and steam is discharged (point 1), only a short period ( of rapid depressurization occurs. This is followed by very slow depressurization, which continues as long as heat is removed by the steam generators. The RCS pressure remains above that of the secondary side of the steam generators until the core decay heat is reduced sufficiently for the flow out the break to match the rate of steam generation in the reactor core. For medium-size small break LOCAs the steam generators must' be maintained as a heat sink until the core decay heat falls low enough for all heat to be r: moved through the break. G-4

Prge 5 For this class of breaks the amount of noncondensibles present in the primary system was calculated in Reference 3. The degradation of condensation heat I transfer due to the presence of noncondensibles was estimated and was shown to have no significant ef fect on the primary system behavior. For this class of breaks there will be no accumulator injection and no fuel failure is expected. The clad temperatures are not expected to exceed 1600 F, so the amount of metal-water reaction will be small. The assumptions in Reference 3 which included all hydrogen dissolved in the primary system and all the air in RWST tank (for an 8000 seconds transient, only 2% of the tank will be injected) more than mmpensates for the small amounts of hydrogen produced by metal-water reaction at clad temperatures below 1600 F. As in the previous discussion, Reference 3 assumes that noncondensible gases do not leave the system via the break. Small Break LOCAs Resulting in RCS Repressurization Following a LOCA, the RCS can repressurize by one of three ways: o Loss of steam generators as a heat sink o Isolation of the break f High-head injection flow exceeding break flow o ( Repressurization of the RCS following a very small break is expected to happen and its effects are considered in developing procedures for long-term cooling. Repressurization occurs because more mass is being injected than is being expelled via the break. The rise in pressure may be very rapid and is terminated in one of two ways: (1) as the pressure increases, the break flow increases and the injection flow decreases to where a balance is reached; or (2) the pressure may continue to rise and open a primary relief valve thereby [ increasing the leak flow. Whatever the cause, the system behavior and the remedial cetions are the same whether the RCS pressure rises to the relief valve setpoint ( or not. Figure 6 presents the pressure versus time behavior for the RCS for a break that repressurizes. For all breaks resulting in the RCS repressurization, a point in time is reached where injection exceeds the break flow (Point 1). Once this occurs, the liquid / vapor interface elevation starts to rise. G-5 l

Page 6 L If the liquid / vapor interface was initially below the break, it will rise to f tha break elevation, causing two phase fluid to exit the break. Assuming the brcak is small enough, its capacity for passing liquid will be less than the For larger breaks, the liquid / vapor interface will remain at injection rate. However, for the tha break elevation and no repressurization will occur. smaller breaks, the liquid / vapor interface will continue to rise above the break clevation condensing or compressing the steam region. Eventually, the RCS will be refilled (i.e., packed) and the pressure will spike upwards (Point 2). The [ pressure rise will be stopped when the break flow matches the injection rate. This balance is achieved because as the RCS pressure rises, both the leak flow increases and the injection rate decreases. [ The RCS is now subcooled and the core decay heat is being removed by natural The conclusions drawn in Reference 3 also are valid for this circulation. class of breaks. l l ( { l G-6

REFERENCES -1. W. E. Burchill, "Physical Phenomena of a Small Break Loss-of-Coolant Accident in a PWR", Nuclear Safety, Vol. 23, No 5, September-October 1982. 2. Yankee Nuclear Power Station Core XIII Performance Analysis, YAEC, December, 1977. 3. Letter, G. Papanic (YAEC) to John A. Zwolinski, "Response to 80 Questions on RELAP5YA", July 1, 1985. 4. R. T. Fernandez, et al., "RELAP5YA-A Computer Program for Light-Water Reactor System Thermal Hydraulic Analysis, Volume I: Code Description", YAEC-1300P, October 1982. l l l 1 t G-7 l

W 5 F L / r L 2400 2000 -A 1600 5 3 E' 1200 t / l W ~C s 5! 23 ( E_ 800 f ~~' 400 D ( 0 0 500 1000 1500 2000 2500 TiliE,' SECONDS EXAMPLE OF SYSTEM PRESSURE RESPONSE FOR A SMALL FIGURE 1 BREAK RESULTING IN COMPLETE DEPRESSURIZATION G-8

c, r r w 9 k II 9 k-a-aP 300010000 o-o P 906010000 9 A -A P 956010000 5 @- m -* ~ ! Broken Loop u 6 (SGSecondary $o w _ E'~ s h a Intact Loop q 5G"Suir6ndsfy g-i / Prfmary System

Pressure 9

~ Accumulator Accudation o 2d0.0 N.0 m.0 m.0 0.0 TIME (SEC) EXAMPLE OF PRESSURE RESPONSE FOR A SMALL BREAK FIGURE 2 RESULTING IN ACCUMULATOR ACCTUATION AT MAINE YANKEE l 1 1

J .= m f--- 1 O i i 8 i i SG Primary i / o H- ' a--a TEMP 360010000 i / ~~ e--e TEMP 954010000 l n/ i

--* TEMP 21010000 t

o J iA,. 'j. i g._ j-- j c l j j ll .i j - i 3 ra o i. 4 g g_' j SG Secondady j 's i la ih 11 sij l s _ _/_ _ _ _ _ - - - E < b l l )l rei,.. o.

....q' I.

eg_ i Y hin -o i i j u i ,I l 5.. o J i,. -4 m- . ll. -g-l, -,m i i sat j o h 2$0.0 4d0.0 8d0.0 800.0 0.0 TIME (SEC) FIGURE 3 INTACT LOOP STEAM GENERATOR TEMPERATURE RESPONSE I MAINE '!ANKEE \\

~ m m o 1 -- 1 O. k i I .il i. o. 5 B,.- [ SG Primary I i e 2 i. e--e TEMP 310010000 n e-eTEMP 904010000 i h

-* TEMP 21010000 i

pj 1 i w ta o SGSofcondary oc U' s_ _ ~,f b< [ r g_ 'ia yq i $9 9 lg 3 u i Ii g.. d .l 9, i 2do.0 4do.0 edo.0 800.0 0.0 TIME (SEC) FIGURE 4 BROKEN LOOP STEAM GENERATOR TEMPERATURE RESPONSE - MAINE YANKEE

r L [ 2400 (' 2000 IC00 E 1203 I f' [I -m s c. 800 N 400 0 0 1200 2400 3500 4800 6000 T!!E SECONDS l FIGURE 5 EXAMPLE OF SYSTEM PRESSUP.E RESPONSE FOR A SMALL BREAK RESULTING IN A PRESSURE PLATEAU G-12

d L [ [ 2400 2000 l \\ 1600 5 i m t. 1200 w L 300 400 i i 1 0 0 2000 4000 6000 80C0 Tit 1E, SECORDS FIGURE 6 EXnMPLE OF SYSTEM PRESSURE RESPOSSE FOR A SMALL BREAK RESULTING IN A REPRESSURIZATIO_N G-13

Appendix A Effect of Noncondensibles on t Single Phase Heat Transfer r L As the core uncovers, superheated steam will be produced. This steam will enter the steam generator tubes. {- Three modes of heat transfer will be present in the steam generator tubes during periods of steam superheat. Laminar Natural Convection Turbulent Natural Convection Forced Convection to Steam A calculation was performed tc examine the ratio of the heat transfer coefficients for vapor divided by that for noncondensibles. The limiting caras of pure vapor cod pure noncondensibles at the same wall and fluid temperatures are assumed. Laminar Natural Convection The heat transfer coefficient for laminar natural convection in RELAP5YA (Reference 4) is given by: k O.25 ( (# h NAT De CONV LAM where T +T p f T =

fgy, 2

fgy, (T -T ) (p/9)fgy, Gr = g(D,/2) B g f G-14 l

The ratios of heat transfer coefficients for vapor and noncondensibles is given L by: i (k )4(S ) (Pr )("N} f h y y N LNC - vapor = kNC-noncond ( )( ) (Pr }( } (2) N Turbulent Natural Convection The heat transfer coefficient for turbulent natural convection in RELAP5YA is given by: = 0.26 k TURB film (Gr Prfgy,) (3) NAT De CONV. where Ty+Ty T = film 2 The ratio of heat transfer coefficients for vapor and noncondensibles is given by: (O ) (Pr ) (v ) 2 W kNC-vapor (b) V y N (4) = f h ( ) ( ) (Pr ) ( ) N TNC-noncond._ Forced Convection For forced convection RELAP5YA uses the Dittus Boelter correlation f,Pr Re ($) h = 0.023 ( where the physical properties are evaluated at the fluid temperature. ( G-15 \\

/ Th3 ratio of heat transfer coefficients for vapor and noncondensibles is given l by (h) ( #V)0.4 (Rey )0.8 (6) kB= vapor = kB-noncond. (b)(Pr} 'N N The The ratios given by equation 2, 4, and 6 are evaluated in Table A-1. properties of air are used in the evaluation since air and nitrogen will comprise the bulk of the noncondensibles. t f l 1 { ( G-16 (

4 5 4 0 2 0 1 0 0 2 1 6 d a n 5 0 2 0 0 4 r o 1 0 0

0..

7 5 u c 5 P n 0 0 0 4 o 3 2 8 N 4 5 5 9 0 0 9 0 1 1 0 1 4 ~0 5 5 6 9 0

0.,

1 2 0 5 r a 5 0 2 0 4 7 r o 1 0 0

0.,

9 7 p u 5 a P 0 0 0 4 V 5 3 0 4 4 1 6 9 0 0 5 0 6 1 d a n 0 0 2 0 0 3 r o 0 0 0 0 7 8 u c 2 6 P n 0 0 0 4 o 4 4 N 3 1 0 2 3 1 3 1 1 1 1 5 - 0 6 1 7 0 0 9 0 6 1 3 r a 0 0 2 0 8 3 mP r o 0 0 0 0 9 3 p u 2 6 a 0 0 0 4 V ~0 5 4 1 2 0 2 1 6 3 0 0 d no 2 0 0 7 n r o 0 0 0 0

7.,

8 u c 0 0 P n 2 5 0 0 3 o N 7 2 2 0 3 7 3 4 1 3 1 1 1 1 ~0 6 1 6 2 r 0

0.,

3 1 2 0 5 2 ar o 0 0 0 0 0 4 p 0 0 a P 2 5 0 0 1 3 V m ) F t ) f c r e s h ) s e / / i u 2 d t ) t R t i r F B f u e ) ( ( ( p a ( V N V N l V N F o i g C C C C g r s 3 y y N V N B B b N N P p i g g f T e D D ( f f g r f L L h R h P T K 8 P v i0 ll

Appendix B System Pressure Increase Due to Noncondensibles This appendix presents estimates of the increase in primary system pressure due to the presence of noncondensibles at two points in the transient: J l a) at accumulator actuation and [ b) at end of transient The Dalton model is used to calculate the partial pressures of the noncondensibles. ( P=P + VAP noncond. (1) h RT P=PVAP

I M

V g where Primary System Pressure (Psia) P = Vapor Pressure (Psia) P = VAP 1545 f t-lb /lb mole Universal Gas Constant = g R = Absolute Gas Temperature ( R) T = 3 Total Gas Volume For Gas (ft ) V = Mass of Components (lb) m = g Molecular Weight of Components (lb/lb mole) M = g 2 As an example, we assume a break between 0.5 ft and 0.1 ft at the Maine Yankee Plant. Accumulator Actuation The noncondensible present at this time are: 2.7 lb 1. Dissolved hydrogen in the primary coolant. 2. Dissolved air in RWST tank (only 1.25% of the 0.7 lb ( tank liquid is injected) 0.9 lb 3. Hydrogen released from zirconium water reaction ( 0.04% cladding oxidation) C-18 )

s Appendix B - Page 2 He 10,.1 lb 4. Fission and fill gas in reactor fuel Ar 0.3 lb l N 0.3 lb Kr 5.5 lb Xe 61.4 lb 3.3 lb 5. Pressurizer steam space gas Assume the system is at 219 psia (accumulator actuation setpoint) and the ( gra and vapor temperature is 500 F (upper head temperature). The primary system volume occupied by the noncondensibles V = Primary Occupied by System Liquid 3 10,240 ft Y = V ~ Primary Lower System Plenum Using equation 1, the increase in pressure at accumulator actuation is calculated to be 6.5 psia. f The depressurization rate at accumulator actuation is about 1.25 psia /sec. The maximum delay in accumulator actuation is about five seconds (<2%). This Peak Clad Temperature is expected to have a minor effect on the system response. is calculated to occur about 100 to 200 seconds after accumulator actuation. End of Transient The amounts of noncondensibles assumed to be present in the system af ter i the accumulator empties are: i

1) Dissolved hydrogen in the primary coolant 2.7 lb

2) Dissolved nitrogen in the accumulator water 56.0 lb

"{ 65.1 lb

3) Dissolved air in RWST tank G-19 6

-Appendix B - Page 3

4) Hydrogen released from zirconium water reaction 21.7 lb

) ]

5) Fission and fill gas in reactor fuel He 10.1 lb Ar 0.3 lb

( N 0 3 lb Kr 5.5 lb { Xt 61.4 lb 3.3 lb

6) Pressurizer steam space gas The assumption is made that the system pressure is 150 psia and the system is et saturation. From Equation (1) the maximus. increase in pressure will be about 18 psia.

The calculation performed is very conservative. The maximum volume of noncondensibles which can be trapped in the primary system is limited to the volume of the pressurizer and the vessel volume above the upper head. For Maine Ycnkee, the pressurizer volume is 1572 f t and the vessel volume above hot leg 3 c:nter line is 1367 ft. At a minimum, noncondensible quantities which exceed 3 c volume of 2939 ft will mix with the steam vapor in the primary system and ( will flow to the break. ( G-20 E

L [ RESPONSES TO ADDITIONAL QUESTIONS IDENTIFIED IN REFERENCE (e) I ( I l f

r + Responses to Additional Questions On The Yankee Atomic RELAP5YA Submittal ( 1. In response to question Q1.19 of Reference 1. YAEC stated that under horizontal stratified choked flow conditions the Moody critical flow model is modified to account for vapor pull-through and liquid entrainment. The model computes modified junction vapor and liquid fractions using the same techniques developed in RELAP5/ MODI, ( Cycle 18. The assessment of the horizontal stratified choked flow model was also discussed in Q1.19, and was limited to a BE calculation of LOFT Test L3-1. Because this was a BE calculation, ( the Moody model was not used. Discuss how the use of the horizontal stratified choked flow model with the Moody model was assessed. Provide the results of assessment calculations, or justification for why assessment is not needed, for review. A.1.1 To assess the use of horizontal stratified choked flow model with the Moody critical flow model, LOFT L3-1 test was simulated. For this calculation the Moody critical flow model was used at the break junction and horizontal stratified flow conditions were seen to occur upstream of the break. The results of the calculation are presented in Figure 1.1 through Figure 1.8. The first three figures show the same trends as were seen in the answer to question 1.19 of Reference 1.1. Hence, as in Reference 1.1 (A1.19), we conclude that the fluid was horizontally stratified in the cold leg during the test and the calculation. During the two phase break flow period, the horizontal stratified flow model was used in RELAP5YA calculation. Evidence for this is given in Figure 1.4 which shows that the void fraction at the break junction is dif f erent f rom that for the upstream pipe volume. In conclusion, Moody critical flow model is augmented by the horizontal stratification model as intended. F-1

s REFERENCES 1.1 Letter, R. W. Capstick, YAEC, to V. L. Rooney, NRC, "Response to Additional NRC Questions of RELAP5YA Computer Code", November 4, 1986. (, l-( l s o ( F-2

il _O t !i ~ t p I n i 0 0 y 0 t 1 9 i 1 s L n E e D D OM 0 g W 0 en 0 Lo 9 2 i 1 dt F l a ol R 80 L C u 1 11 c fC 0A00 pl 0Y00 0 oa I T S-oC CPUC ,l ) L I. R PAPP C A C L-EtY CECC S cS Y DRDD ( aPtA D e E nL O O 0 MI E 0 I R M a 0 Tf 0 oo ( t n 1 o1 s - 3 i3 L 0 rL a 0 pT T. 00 mF FO oO %/ CL L R 0 1 Y g" 0 1 5 e 0 R I 4 er u ( g i F 0 i, 0 L 1 I 2 0 a 0 i og m*

8'

.R' n 5 a 0 d Bn s b3 k I l l

^- R5Yfi / LOFT L3-1 ( 1100DY CRITICfL FLOW N00EL 1 o g g' **=o-.3 ~ e-I $g a ~ SI' a 'jl LOFT L31 [ R' ik gh-k J N $ :I ~o ~ ki fl :l b ~ i T ce a .i %d' n.n JL aba 4. L, t u,qu tL n

1.

s 9 (ts ~ ' n 7 '"' U r3 " 4 9 ua $~ k!b r e.. 1 m' r g. r : r-ri \\@- g j,,j y g-

Iy;

o 4gdf;

~ ' d pjpg4p 4 g:;- If q s .0 so'0.0 10$0.0 2250.0 1450.0 1850.0 1 0 TIME (SEC) Figure 1.2: Comparison of Broken Loop Cold Leg Densi'..y in LOFT L3-1 to RELAP5YA Calculation .J

l l l l 1 u r m c 0 [ MT N Nb O a _a tn 8 O_ lj m sc O O LJ o ta _8 JE Z o 0 0 o .o [ d a _a a 7 ^ Y O a

O m2 Q

Og b J O o (f) o O J J l eL J O. - g C")a5 b h oE e M O gg cx O B O

i--4 O

o b o y E o f m o a 2": O. 2 ,_. _ s [a _zu e O w y- _g __.. g.-- - 8 Gll: o a l l i l i 0'09 0*0S 0*0) 0*0E 0'0E 0'01 0*O ( D3S/W81 ) 3188 M0lJ MB388 F-5 f

s o ) _ i a a a a a E y = [. P o I l [ y 8 o 4 h U h 5 ~ g I 11 d d N 3 ? o E (I .l - E mb y gn g oW i &n "Y _~ a -8 ( g-

3 ou o

h. -l [0 C ~ m o -n II l s, O *1 Sk 8 *O ik Sk sk ik tk Ek 1k 0 *0' S010A N0110inu 010A *)JUA F-6

0 ~ 00 = 8 1 1R 3Y L5 0 0 F 5 TU 8 WE y 1 LR M n 0 i 0 5 e 9 r 1 1 u L ssn ED eo O ri M 0 Pt a 0 ml N 5 2 eu O 1 t c L sl F ya SC LA yA C 0 rY I 0 aS T

5) mP 0

C i A. I. 1 R E rl C SPE ( R f YD E oo t O

0. M I n O

doTo1 M e s - i3 ( rL a 1 pT mF 3 oO L O. CL o T dO F O 5 L 1 / e A 0 r Y D u I 0 g 5 4 i R F ~ 0 0 d2 ~ 0 {h 0 >iI o o ".l-ag o o 8 o e ogog 2"ub 9g l-n 7" ~ ~ l jllll

j'l I 0 0 0 8 1 e 1R 3Y 0 LS 0 P 5 m iA 8 1 fL uE LR e n i 0 0 e 5 r 9 ) 1 u s s 2 e E0 rn 0 P o N 0 i 0 mt 5 ea W 2 tl O 1 su L yc F Sl a L yC 0 r FC ) aA 0 I 5 CdY T 9 E nS I. 1 S oP R ( cA C eL ESE Y M R D If O 0 Too o t O d n M e o1 s-( i3 rL 1 a pT 3 0 mF L oO D d CL ^ i e ru L 6 / 1 O. ~ e R o Y d r 5 e u R g i F N 0 o d2 I 0 i l 0 i J

a j 9g a

.g o N._ o of 5b 1 lllll\\ll

i 1 v 0 L 00 8 1 w 0 0 Y 58 1 a 1 -3 L L T I-0 F 0 O 5 L 4 1 1 n a i LE e 0 r 0 u 0 1 s 1 0 s W 52 e O 1 r e = L P F r L o A t n C 0 ao I 0 li 5 ) ut T 9 I 1 C ma I a R 4 E ul S cu u, C ( cc Y Al E a 00 O. MfC I o 0 o 1 \\ d T A i e nY 1 a oS ( = sP iA 1 rL = aE 3 pR L O. m o d oo i e Ct a fu L 1 - R 7 / 3Y 1 L 5 0 ~ R P D e Y Tr d r 1 5 FL e u a R O E g L R i e F 0 e D d 2 a 7 0 0 o b 8 o og .D "g .d aN ~ s E ggh:g 5m 1 ~ - e s llll 1

9- ,s m W M [ (e" R O ~ O h O de -? ys >C l 5

o u-

-o 36 7 E_ 3< o am n3 o et a a_ 8u se om t J -8 u j6 E_ [ s or et ez

o. ~H 6S

> U [ tn y _O l % o 8 9 O ~ [ -c a MB

o

-?

J tn O n_ m HC ( LJ a. O LJ az _g O N o U 1 i i I 0'9 0'S O*t 0*E 0'E 0'l I IJ 1 13A31 80191RWR309 F-10

2. In two of the three PWR integral assessment calculations (LOFT Test L3-6 and Semiscale Test S-LH-1), the system pressure was underpredicted. In the Test S-LH-1 assessment this resulted in the calculated accumulator injection beginning approximately 75 seconds before it began in the experiment. Because of the early accumulator injection the calculateJ PCT was slightly lower than the PCT measured in the experiment. In LOFT Test L3-6, there was no accumulator injection. If accumulator injection had been used in the experiment, calculated accumulator injection would have begun approximately 100 seconds earlier than in the experiment because of the lower calculated system pressure. Because early accumulator injection can limit the calculated PCT, or even preclude ( the calculation of a core heat-up as some RELAP/ MOD 2 assessment has shown (Reference 6), discuss why the system pressure was underpredicted in these two assessment calculations and what will be done to assure that the beginning of accumulator injection will be conservatively calculated in SBLOCA licensing analyses. A.2 To improve the RELAP5YA simulation of the integral small break tests, several improvements were made to the code representation of the LOFT and Semiscale test facilities. The code was assessed against the LOFT test LS-1 and Semiscale test S-LH-1. The prediction of pressure history and accumulator injection time will be discussed with respect to these recent simulations. The base case simulation of LOFT test L5-1 is discusse8 in detail in section 5.1 of Reference 2.1. In this case the system pressure, shown in Figure 2.1, was underpredicted between 100 and 200 seconds. The accumulator was calculated to inject approximately 30 seconds earlier than the experimental data (Figure 2.2). The overprediction ( of break mass flow rate, hence underprediction of system mass inventory, shown in Figures 2.3 and 2.4 respectively, was considered to be the cause of these discrepancies. To test this hypothesis, F-11

s w a calculation was performed with a subcooled discharge coefficient s' of 0.85 and two phase discharge coefficient of 0.9. The results of this calculation are shown in Figures 2.5 through 2.8. The pressure history shown in Figure 2.5 is closer to data than the e b base case calculation, hence the calculated accumulator injection time is within 15 seconds of the experimental data (Figura 2.6). ( The lower break discharge coefficients brought the calculated break flow and the system mass inventory closer to data as seen ( in Figures 2.7 and 2.8. The Semiscale calculations discussed in Reference 2.2 also indicate that a discharge coefficient of less than 1.0 significantly improves the prediction of pressure history and accumulator injection time. The effect of discharge coefficient on system pressure is similar to that of the break area. Since a break spectrum analysis will be performed for licensing small break calculations, the net effect of the tendency to underpredict the pressure may be to l shif t the l'imiting break towards a smaller break size. Since the LOFT LS-1 results indicate that the code conservatively predicts the system mass inventory and PCT, this shif t is not expected to affect the derivation of conservative results in the plant analysis. ( REFERENCES 2.1 "Assessment of RELAP5YA Against LOFT Nuclear Experiment LS-1", l Attachment D of letter from YAEC to USNRC, FYR 88-09 or MN 88-07, dated January 15, 1988. 2.2 "Assessment of RELAPSYA Against Semiscale Small Break Test S-LH-1", Attachment E of letter from YAEC to USNRC, FYR 88-09 or MN 88-07, dated January 15, 1988. F-12

LC 7:: "ES" 5-1 PRESSURE INTRCT LOOP COLD LEG o I I I I I oos o [ lb -{J o_ - E PE-PC-001 -o~ e--e P 185010000 o_ , L11 - " QC i CD O (D La_1 in - T O_ U ~ u O

)

o 5b.0 100.0 150.0 200.0 250.0 300.0 0.0

E 2 7"ER RL '::URE

{ SECCNJS ) Comparison of Measured and Calculated Figure 2.l* Cold Leg Pressure (Base Case) 1

<+,w r, e, L 3 me, RC5UMULRTUR FLOW o d1 I I I I I n O WO O d-xm C9 L51flCC M

-* fRORJ 615010000 O

I - g_ 2 a /bl o i 1 L 1 La_ O dC C O O O O <n cn ao. to 1 I I I I l 1 0.0 50.0 100.0 150.0 200.0 250.0 300.0

. I

E

= ~"E R RL 3:'l R E

SECJhJS

) Figure 2.2 : Comparison of Measured and Calculated l Accumulator Flow (Base Case) l

ljI 0 0 ,aI 03 . o' ) N a A.,I 0.S 0D I = A 5N r,8 2 I O J m' C m NO a L mf d 0E .=. e I i t c .S a l 1 n 0 u { A,. 2( C l c = 0 l ] a 1 5 Q uV dn 1 I a W A E I d) .]0 ee O J. .R rs I uA L J sC Mm 0L a TF ee I 5T Ms S K b, a 1 ]1 P fBo( ER A J nw oo TE M,, 0 R iF I sl R r ak B pa T A 0 me R or A 1 CB 0 F Y v 3 E 1 v 1 O w T L 3 F 2 0R eru g i 0 l F 5E I M I 0 T 0 ,M I o.Oog a.oW-o* a~ 0.ow o=0 ' 1n c nnI J 1 CfcN)cv ,. <f n f cc[ = l 1 fl

ff 0 0 1 5 0 3 ) 0.S 0D 5N I a. I L 2 O d C e) t e ~ 7-0E as l a uC .S c l e 1 0 as Ca I 0 B I 2( d( 5S = E n ay r do et rn ue S )= 0 R sv an R eI 7= 0L M TM m Wo 5T I f e ot S L I 1 s P ny ( ER oS s A. i iy - TT rr T aa A, 0R O pm T D c. mi or T 0 R CP I F 0 E 1 I 1 0 T 2 er F ug l i 0f F 0 l 5 E I M I 0 T 0 I t { \\- 0 0 0 o 0 0 0 0 0 o o 0 0 0 0 0 o 0 0 6 5 4 3 2 1 0 mg y 4E lll l

W O e 0 l 1 1 n m oo b i ?8 _OO aw tn z [ C.D y* "O E l w w O _3

a. a.

Q W {f D ~O _O (.D te J u

s

( lg o Ed O at NM =3 LD 9 l aa Q _ c5 7O x l23 o 5 J H WH $H ti L s W@ g aw >n 1 5.,", Z O g-1 Lt_ ta O J oz - o 1 D H9 in w ~* (D OC$' LJ = - @ Lt.] C' E I H O _O o.11 061 os 00 o.oe ca) Ersssu F-17

.0 7:: ::ES" .5- :. RCCUMULRTOR FLOW G I I I I I to m C.3 Et] O* CD l xm CS L51FICC V m-eifLOWJ 615010000 o f ? I hh] R V O L1_ E" I d[]_ O O cn ) C[o ( = to - 1 I I I I I 0.0 50.0 100.0 150.0 200.0 250.0 300.0 Y E =::E R L

)"

R E .: SEC0\\JS ) = Figure 2.6: Comparison of Measured and Calculated Accumulator Flow (Sensitivity Study)

) ] ] ( i Wl O$ jW : l L t )ZRD m or a ~ 0 0 0' ) 2 C 0 E I S0 I 5 / 1 G K 1 g 0 { uM e 7 I 3 ( 0 1 I 6 0 1 W 0 [ m0 I F 0I 5 S S f gc5 r i -h' fM e dvE R0 f f_P 1ba rH Etjy M g s 0,' W'- 1O O NOPC wwO'o aOO*n ~ OO ya.O "oP O U 1 m m' LOO /_JW J J D_ .MI U3-L J J (J ]

c L

l - l R g%5[ 3yG F2 gl.E A s ya y "" E G*hr l

r ll 0 7 00 d_ 3 ) c g 0 S .0D I I 5N )y 2 O du t S C dy 0 E ti et av . S li 1 0 ci ut i l s 0 an I 21 Ce 5_ S d(n 2 ay 7 Ed r S 0 o 0 et S )0 0Rrn 0 ue R 60 sv 0 0 J an TM F0 eI i 5 T .f M E S L R m R I 1 e P ot (f ER s V HL ny TT FR i oS FT s O RN Riy GC 0 rr aa T pm T mi 0RCPor F 0 i E I 1 O L T8 F2 0Re ru g i i0 F E 5 I M I T 0 .0 t 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 5 4 3 2 1 m3 l h l /)llll -}}

ML20148G543

Contents

Text

REFERENCES:

1.0 INTRODUCTION

2.0 BACKGROUND

6.0 CONCLUSION

Navigation menu

ML20148G543

Text

REFERENCES:

1.0 INTRODUCTION

2.0 BACKGROUND

6.0 CONCLUSION

Navigation menu

Search