ML20054F137
| ML20054F137 | |
| Person / Time | |
|---|---|
| Site: | Clinch River |
| Issue date: | 05/31/1982 |
| From: | Everson W, Lowrie R WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| To: | |
| Shared Package | |
| ML20054F129 | List: |
| References | |
| NUDOCS 8206150245 | |
| Download: ML20054F137 (47) | |
Text
{{#Wiki_filter:. o .. s.
~
VERIFICAilUi & NATLWL CIRCLU.T!G1111 CUN01 RIER BEEEP EACTOR FIA<T
- A't TTATE -
PAY 1982 PRE?eaED BY: R. R. LOGIE g PERFOR'#4CE k;ALYSIS & Q RELIABillW ('ESTINGHSISEELEC nuv4CED hCTORS MPROVEDBY: k . W.J.Ga/Scu,IkacEa - PERFORr# ICE hhtYSIS & WESTINGHOUSE ELEC. C 8206150245 820610 l PDR ADOCK 05000537 A PDR l
VERIFICATION OF NATURAL CIRCULATION IN CLINCH RIVER BREEDER REACTOR PLANT - AN UPDATE , EXECUTIVE
SUMMARY
The natural circulation verification program was initiated in April 1976 with the issuance of " verification of Natural-Circulation in Clinch River Breeder Reactor - A Plan". This plan was implemented to verify the findings of "A Preliminary Evaluation of the CRBRP Natural Circulation Decay Heat Removal Capability" (issued March 1976) that adequate decay heat removal capability by natural circulation existed in CRBRP and to verify the computer codes required to model the response to a natural circulation transient. The purpose of this document is to present a status report on the verification work completed to date and an update of the verification plan for the work required to complete this verification. The emphasis for this verification effort is on testing of individual components and of their interaction in the system during the transition to and operation in the natural circulation operating regime. Plant wide natural circulation testing at FFTF and EBR-II has not only confirmed component modelling acceptability but also has demonstrated that all pertinent system / component parameters have been included in the models to adequately define the plant response to a natural circulation event. Figure 1 highlights the tests supporting the verification of the DEMO, COBRA-WC, and FORE-2M codes which are used to characterize the response of CRBRP to a natural circulation transient. The conclusion drawn is that adequate testing is in place to characterize the pressure drop and heat transfer of components and the dynamic response of the thermal heads during a natural circulation event, and, therefore, verify the codes used to predict the CRBRP response to the event. t O
I . 9 i CRDRP NATURAL CIRCULATION l VERIFICATION PROGRAM . DVERALL ST$itti MtJtitS_ M??8t_ Snt!M_vialfna.flim f
- Cll#: Plant Systems e (gu-li fests e Ctb4A-nfC: Corv Systee *
- IIIf M.C. Aueptante Tests ggg e 10ft 2M: loot lied !
(Coteles 0[te/CW RA.uc
- Ventfor Atteptance fests
*id f usti-2M in Pietest ;
e ion.ii vest, ' obh I [ us d fue(md o . rnr resis f
. rien.ineet to,acis, M ai, sis i e f luw St ratlit tation/Haldis tribut t on I
M I ptpp - tests at Ntl e vendor Water Acteptance Tests f e h e Sodline Tests 9 5Fff TFIF Tests (ft.tt.odulogy Deffaltion) - y =e e 3 8 --- I A' m fi.
-c '
' at Tests at $CTI -
l , E , EU.IIl P3!N3 - j e
,,f y ,
0 ML $cale Meel $odlW t r Mising/ Stratification # , I j e BCt Scale Moel Vater [l , { e 5;talag/5tratification fests -w [ ( I ( ,
/ - " --'"- gistytAf_f tpg ptsy_
e ihnt1115 aP Tests e Ituat C,yatity Study , (Y s $( -
- Vender .halysts
' Museriest Myers Code Analysee _
~
/ ~
h \ - s L._ g
~ o C081-grinc o thecry liest Tests (pggg yn g
- ML $tratificatinn - aP Tests
{lf8'R-4.IPPA-II(N0f/8-IF,
,) e 16* (IfIF) Flew Test
- IEA*ll Secondary toop fests
- PU T 1esting
- LIM /C0itt Assembly /ftudwie aP Tests ) b* (f ff f) flw Test Iktalled Analy%Is (Based ' WaCHF Analysis 8 fHit-ll fests (Fluw bedistellivtim) e e tilF = Natural Ctreviatten fests a tiedet/ fuel evidle tieag l'ensfer/Mirfag Tests ...
u .
TABLE OF CONTENTS Page 1 "
1.0 INTRODUCTION
3 I 2.0 REVIEW OF ANALYSIS METHODS . 8 3.0 SENSITIVITY ANALYSES 8 3.1 UPPER PLEAUM MODELLING , 9 3.2 DECAY POWER 11 3.3 CORE INTER- AND INTRA-ASSEMBLY FLOW REDISTRIBUTION AND ASSOCIATED HEAT TRANSFER EFFECTS 15 3.4 LOOP FLOW RESISTANCE 17 3.5 FLOW COASTDOWN 18 3.6 IHX HEAT TRANSFER PERFORMANCE FLOWS AND 18 3.7 TRANSIENT INITIAL CONDITIONS: TEMPERATURES 20 4.0 VERIFICATION OF SUBSYSTEM CHARACTERISTICS 20 s.1 ....CTOR THERMAL HYDRAULIC ASPECTS 20 4.1.1 REACTOR STRUCTURES AND ASSEMBLIES OVERALL REACTOR PRESSURE DROP 20 4.1.2 CORE FLOW REDISTRIBUTION AND HEAT TRANSFER 26 4.1.3 REACTOR UPPER PLENUM MODEL 25 4.2 DECAY POWER 27 4.3 PLANT COMPONENT THERMAL HYDRAULIC CHARACTERISTICS 28
*4.3.1 PRIMARY AND INTERMEDIATE PUMPS 29 4.3.2 IHX 30 4.3.3 CHECK VALVES 30 4.3.4 PIPING STRATIFICATION AND MIXING 31 4.3.5 WATER / STEAM NATURAL CIRCULATION 35 5.0 VERIFICATION OF WHOLE PLANT ANALYSIS 35 5.1 FFTF TESTING 37 5.2 EBR-II TESTING 40 6.0 SCHEDULE 41
7.0 REFERENCES
4 4 4 ( 9
i . j
1.0 INTRODUCTION
I The CRBRP has been designed to remove decay heat from the reactor to the steam generator system by natural circulation in the primary and intermediate sodium loops as well as in the steam
- drum to evaporator recirculation loops. This has been achieved through the arrangement of components with respect to each other and through the specification of component characteristics
! important to natural circulation. A preliminary evaluation of the CRBRP natural circulation decay heat removal capability has l been performed, which concludes that adequate circulation will be provided by the plant systems as presently specified. An overall plan to verify this natural circulation capability for the CRBRP was initially presented in 1976 (Reference 1). This update of the 1976 plan presents the results of verification efforts that i have been completed to date and an updated plan to complete the l verification of the conclusions drawn in the preliminary evaluation that adequate natural circulation capability exists for the CRBRP. Three underlying assumptions to this verification are: 1) that prior to plant startup testing conclusions drawn with respect to i plant wide natural circulation performance of the CRBRP will be j based on analyses; 2) that a sound basis for this analysis is the
- understanding and verification of models for system and component behavior under natural circulation conditions; and 3) that the evaluation of data from FFTF and EBR II natural circulation testing using CRBRP methodology will contribute significantly to the overall verification of the analytical techniques used in
, na.tural circulation analyses. On this basis, the verification
- plan has been developed to provide the technical support required for reasonable confidence in the systems analysis tools. The
! program has four parts: A. A review and selection of acceptable analytical methods (codes) used to calculate flows and temperatures to identify and provide visibility of assumptions, capabilities and limitations which may be important to natural circulation analyses. B. A series of sensitivity studies to provide a quantitative understanding of the influence of key parameters on flows and temperatures during natural circulation transients. C. Component testing to determine the thermal / hydraulic characteristics which are empirical in nature and to supplement this testing with detailed component analysis where testing is not programmatically feasible. Data from these tests provide the technical basis verifying the component modelling conservatism (orin the system analysis code (s). D. Comparison of predictions (both pre-test and post-test) with data from FFTF and EBR-II natural circulation tests 1
- w'ith the predictions being made with the same methodology and modelling techniques employed on CRBRP in order,to demonstrate the systems modelling conservatism in the system analysis code (s).
S 4 O 2
2.0 REVIEW OF ANALYSIS METHODS First, The purpose of a review of analysis methods is two-fold. used it is to focus the verification efforts on the codes that, together, can completely define the responseit of Secondly, is the
'ta) CRBRP evaluatetothe a a natural circulation transient.
capabilities of various test facilities to provide the . experimental data required by the Natural Circulation Verification Program. (NCVP). , The codes to be verified by the NCVP will be DEMO (Reference 3), FORE-2M (Reference 4), and COBRA-WC (Reference 5). Three codes 3 were selected, because it was concluded that one code alone cannot meet, nor should a singular code be developed to meet, the program objectives. Such a code would be too large a code to be DEMO is a system wide of practical use in the LMFBR industry. code used to predict the coupled performance of the COBRA-WC reactor, heat is a transport systems, and the steam generator system. whole core code that predicts how the flow and heat is distributed between the numerous parallel flow paths in the core. FORE-2M is a detailed core code used to predict " hot channel" parameters in the core. Both COBRA-WC and FORE-2M use input from the DEMO overall plant analysis and FORE-2MTherefore, uses further input as a unit from the results of the COBRA-WC analysis. these three codes provide the detailed model required to predict A brief summary of natural circulation transients in the CRBRP. each of these codes follows: A. DEMO DEMO is a system-wide code which predicts the coupled performance of the reactor, heat transport systems and the steam generating system (including piping and plena heat capacity ef f ects) . To provide accurate results for the plant as a whole, yet maintain the code as an amenable tool with regard to computer storage requirements and running time, localized phenomena (which do not aff ect the system as a whole) are not given detailed modelling. An example of this is an item such as local flow / heat redistribution between and within core assemblies at low flow. To consider localized phenomena I in the core, with the required resolution, takes a separate computer code itself (i.e., COBRA-WC was r selected f or this purpose) . In general, DEMO provides a l prediction of the overall. system state variables such as I net flow through the reactor and bulP temperatures i entering and exiting the core. I B. COBRA-WC t The COBRA-WC code, which accounts for core inter- and intra-assembly flow and heat redistribution, predicts the ( boundary conditions for a peak rod or cluster of rods in fuel and blanket assemblies given the reactor boundary i 3 1 l -
conditione auch cs total roactor flow, proecuro drop cnd core inlet temperature from DEMO, individual core assembly powers, and individual core assembly thermal-hydraulic characteristics. Analyses of this type phenomena under natural convection cooling requires detailed core / reactor modelling because of the strong interaction between the fuel assemblies, blanket assemblies, control assemblies, plus other core regions and bypass flows which all act as highly coupled parallel flow paths with heat transfer between them. C. FORE 2M Given the localized peak channel flows and heat interchange as a function of time, FORE-2M predicts the hot channel coolant temperature for natural circulation which includes considering uncertainties in the prediction.s. This hot channel coolant temperature is used as a the basis for determining the acceptability of natural circulation for decay heat removal _ of LMFBR's.
~
To have a verified hot rod prediction, one must properly account for such effects as: a) fuel restructuring; b) fuel-cladding gap conductance; c) stored heat in the rod; d) statistical significance of physics and engineering uncertainties in power, flow, dimensions, properties, ect.; e) localized hot spots on the rod due to the wire wrap or pellet eccentricity; f) locaj?.ned decay heat variations; and g) localized rod power variations in the assembly. Since FORE-2M is used for core hot rod analysis with input from DEMO and COBRA-WC, extreme detail in nodalization can be used to assure accurate temperature predictions. This level of detail is not practical for the COBRA and DEMO predictions since the storage requirements and computer running times would make the codes impractical. (Note: In addition, the code provides a detailed nuclear physics prediction for the prompt and delayed neutron fission power as a function of time from shutdown with the capability of considering doppler, sodium density, and radial / axial core expansion feedback plus the effects of control system worth and PPS functions. This information is used to verify the DEMO prediction of core power versus time) . During the transition to and operation in the natural convection cooling mode, the effect of an increasing power-to-flow ratio approaching or exceeding that of steady state can be experienced. Consequently, core temperatures increase and natural convection phenomena such as inter- and intra-assembly flow redistribution become significant once low flow conditions are reachgd. In the CRBRP or FFTF, the core thermal head becomes increasingly significant relative to the form and friction losses across the core for flows below 5% of full flow. Coupled with the flow redistribution, significant heat redistribution on an inter- and intra-assembly basis occurs throughout the core due to radial 4
i tcmparature differcntiolo and cn increacsd flow trencport tim 3. Both of these effects (i.e., natural convection flow and heat redistribution) have been found to significantly reduce maximum core temperatures. Figure 2-1 shows the interaction between the three codes used in the CRBRP natural circulation analysis procedure. The-first code (DEMO) provides core delayed neutron powers, dynamic average region temperatures and inlet flow, and average core pressure drops. The last two codes (COBRA-WC and FORE-2M) provide the two-stage calculational approach for the core analyses under natural circulation conditions. Utilizing DEMO boundary conditions, detailed whole-core flow and heat redistribution analyses of all the parallel core assemblies and bypass regions is performed by the COBRA-WC code. These data are then used for a detailed analysis on a single, " hot rod" using the FORE-2M code. These latter analyses include effects of localized phenomena and uncertainties in nuclear / thermal- hydraulic / mechanical data. A linkage between the COBRA-WC and FORE-2M codes has been developed and verified to incorporate the inter- and intra-assembly phenomena into the localized hot rod natural circulation analyses. For each axial node of the hot rod modelled in FORE-2M, a heat balance is performed using the expression for the heat transferred to the coolant at that section, Qc(X, tau) as: Oc (X , tau) =Qr (X , tau) +Qex (X , tau) where: Or(X, tau) = heat transferred from the rod surface at axial location X and' time tau-Qex(X, tau) = coolant heat input or loss due to radial conduction and mixing heat transfer and flow redistribution to adjacent coolant channels; directly input from COBRA-WC. Coupled with this, the axial mass flow rate for each axial node is also input from COBRA-WC analyses. Boundary conditions for the COBRA-WC (e.g., plenum-to-plenum pressure drop and coolant inlet temperature) are furnished by the plant-wide code, DEMO for several cases: the "best estimate" or nominal case as well as cases with uncertainties. Likewise, a corresponding modelling of the core parallel flow network, with regard to pressure drop and decay heat uncertainties, can be used in the COBRA-WC* analyses for input to the FORE-2M hot rod temperature predictions. To evaluate the effects of core uncertainties, hot channel factors are used. Similar to those applied in steady state 5
.m
. .e g
CODE INPUI .
- AV! RAGE Cont lit GIGN DECAY HE AT
- AVERAGE CORE nl GIUM FOW(8 Ol5tRl8Ull0N PL ANT-W10E CODE.
- AvinAGE Cont HIGION 3r VtR5US FLOW OtMD
- AVERAGE CORE REGION fl0W Att0 Call 045
- Ot1 Alt s Of PL ANI SY5f tM/CONitGUR All0N l
' OYNAMIC CORE Ott AYf D NEUIRON POWERS
- PHYSICS TitD8ACK COEfilCI(Mis l DYN AMIC AVERAGE FUtt AND St ANRIT CHANNit if MPER ATUfit 5 WINLEi DI ggqLgy gg
- ROD PAnAMIit ns - G AP CONOUCI ANCE, RISTRUCTURING ,
itMPERATUnt5 AND p s b AND Cp,Ilg g g g. E T C. AVf RAGE CORE* SP hl , 17 C0llt AVERAGE SPhi AND (Ff f CFivt ) l i
- L OWE RIUPPE R INit RN A15 AND EYPA55 REGION HYOR AUtlC /
Cil AR ACil R151105 Wilott-COR[ g AVillAGE CORE THERMAL ORIVING N( AS g l
- IN0lV10UAt A55tMgtY HYOR AUtlC CH Afl AtitRISTICS
- INDIVIOUAL A55tMeLY POWills AND Ol5f Al5Uil0N -
SYSTEM CODE CD8tl A-WC
\]in0M OtMO ARE
{l5R10WRIO.
, NOT CORRICf,IltRAll
____A l SII ADY SI Af t & TRAN5ttNT (DECAY lit All j :
- INDIVluuAt A55tMetY & SYPA55 GPERAflNG TLOW RAf ts- ,
INiil At CON 0ll10N f OYNAMIC CORE Pt AR AND AVERAGE CHANett FLOWS
- Cont Gt0MtInic M00L11NG DAT A f
- MINING PARAMtilf;5 ;
' INTER-AND INTRA-ASSEMBLY FLOW AND ltCAT REDIST
, OYN AMIC PE AR COOL ANT ItMPIR ATURES
- UNCERT AINTY F ACTOH5
- L OC At otC AY Hi A15 1901 tilANNtt/ PIN kOEMO inAN51 TNT POWERl CALCUL All04]
- PIIYSICS 7 Et OS ACK COE F flCIE Ni$ ANKLYSIS COUL -( IS NOT COMRECT,lif fl Afl0N 100P IS
- LOC AL 1801 RnD P AR AMIIIRS (f utt fitsf 71UCTURING, f ullt -24 \ntilUIRf D~~~~~~~~~~J GAP CONDUCI ANCE, t IC I
*
- MODELING A55t55 MEN 15 FOR TfMt IN titt Ef f ECI5 OYNAMIC Hy IDEt, St ANRti AND CONTROL A55tM9tY, j
- A55tM81Y Ct0Mtintt M00tUNG D AT A CHANNttinUO itMPERATURE5; VERiflC All0N ENI AltS
, COMPARING 19415 ItMPER AIUnt 10 ENPERIMENI A,t S Afd.
o IICVP ANAL.YSIS PROCEDURE FIGURE 2-1:
calculations, tho dirGet cnd atotictieni typa factor'c are concorvatively ccmbin;d by the s mi-otatictical mathod. To support this analytical method, an initial review of the testing requirements was performed. The results of this study of testing needs to support code verification are presented in the " Review of Experimental Facilities and Testing Requirements" (Reference 2). Based on this review of experimental facilities for testing and verifying the above models for natural circulation predictions, a testing program has been outl-ined for code verification. The specifics of this testing will be outlined in the verification discussions that follow. O 7 82-0304 .. _ _ - - . . _ _ _ _- _ _ .__________ _
3.0 SENSITIVITY ANALYSES The objective of the sensitivity analyses is to identify those system and component parameters which have the greatest effect on the temperatures and flows during a natural circulation transient. This aids in indicating which components require the closest attention during the design process, aids in defining testing requirements, and aids in indicating where detailed modelling is required. . Sensitivity analyses in most cases are based on the outer bounds of parameter uncertainties. These analyses give early indication and visibility to the degree of conservatism in base case calculations'and in nominal expected predictions of plant natural circulation capabilities. The base case for these studies is the three loop design natural circulation case from rated plant conditions presented in Reference 11. The parameters for which the temperature and flow sensitivity is determined, are based on previous CRBRP calculations, FFTF experience, and component / system testing capabilities. Results of initial sensitivity analyses was reported in Reference
- 2. This report defined the areas which required verification by testing and areas which required further study to determine the magnitude of their impact and therefore establish the need for the continuing sensitivity studies and for the testing required
'for component and system verification. 3.1 UPPER PLENUM MODELLING Upper plenum characterization takes both upper internal structure modelling and outlet plenum mixing modeling into account. The base case analysis of natural circulation in Reference 11 was based on a stratified model of the upper plenum without the upper internal structure (UIS) included. Later sensitivity analysis showed the importance of including the UIS model for natural circulation analyses. Following a reactor trip and during the transition to natural circulation, the degree of sodium mixing in the upper plenum determines the sodium density profile along the flow path in the plenum and primary hot leg piping. The integrated density around the whole primary loop flow path determines the natural circulation driving head. It follows that plenum mixing assumptions influence the plenum and primary hot leg contribution to the thermal driving head and, therefore, the natural circulation flows. Results of an initial sensitivity study on the outletg plenum mixing (UIS excluded) impact on hot channel temperatures was reported in Reference 2. The range of impacts for full stratification to full mixing in the outlet plenum was shown to berelativegyminor. Blanket assembly temperatures varied from
+12 F to -0 F while fuel assembly temperatures varied f rom +0 F 8
0 to -15 F. The addition of the UIS to the model for natural circulation , analyses was found to have a significant impact on the natural circulation transient. The major impact of this addition was on the transport time for the transient to reach the plenum through the chimneys. This resulted in a negative impact on thermal heads causing the primary flow to drop further and rec 6ver slower than predicted in the base case analysis. The impact on maximum sogiumtemperatureinthecorewasanincreaseofapproximately 70 F in the calculated value. A sensitivity study has recently been performed which evaluated the impact that the upper plenum modeling assumptions have on plant transients. The basis for this comparison was the reactor outlet temperature calculated for a normal plant scram (duty cycle event U-1B) . These evaluations centered on two central characteristics, the upper plenum sodium mixing efficiency and the upper internals structure (UIS) impact. The models used in this comparison were: A. A one node fully mixed model. B. A stratified model. C. A detailed two region mixing model based on the ANL-PLENUM-2A code but including mass accumulation in the plenum. D. A detailed three region mixing model which is based on ) the ANL-PLENUM-3 code and includes the UIS. i j E. A modified version of the two region model above that l
' includes the UIS and a more accurate geometric
- representation of the flows and volumes in the outlet plenum.
l These models were compared against the results of a two dimensional plenum calculation with the VARR-II code (Reference
- 19) and the results of this comparison are shown in Figure 3-1.
Based on this study, the modified two-region model has been used to predict outlet plenum performance for FFTF natural circulation test pretest predictions and is planned for use in predicting CRBRP outlet plenum performance. The results of these studies provide the nasis for the verification testing discussed in Section s.l.3. 3.2 DECAY POWER g The thermal load from decay. power is one of the more important factors determining natural circulation performance. As discussed in Section 4.2, values of reactor system and l 9 OS_A1AA l_._____. ._
o . . .
.I
~
i
' iif- ;1, g -
! 1
.I .
i.is ii , l l l ,' i ll 1Y t g
~
- ,lI l
I . a q r / r B
> l 5 w
/ l- 3 s :)
5 El I * = 4
= /
r 8 o
==
ew x
~
W l. n = -===_ i u
/
5 E
- =- ~.
J > I .E e~c--ua-x u ba / 7 " [ OU o a SI TNENh5 w n c...w . > h, Q { $ E O 5 - f , ' E m/l / . 9I?3't 9 l
'/
E U c: f
/ v '
u /l:,."/ i
/
/ .
]l,' ,) .
\ , - p <r , s*
& /l ' / /
w l'l 9 - 5 8
-/'i ll d a
/
g
- w f.- /
i u
,o .
/. .. *
)!; f~ .
r- /- 8
..> ~
O**A' / .3-c' At' ~ W/ - ,
, r/ , . .
. . . . .o
- 33S/E7 - MD l'cK31 YH 11X3 ' A* H ,
10 1 .-- _
individuni cscenbly daccy h ct includ3 tico d3 pendant uncertainties which cro coxinum at tho tino cf cerca cnd d crocce The in magnitude with increasing time after scram hours initiation. after scram decay heat uncertainties f rom scram to two (2) are summarized as follows: Uncertainty - (a) Scram (b) 2 minutes (c) 2 hours 31.0% 12.3% 8.8% Reactor System Fuel Assembly 32.4% 13.7% 10.3% Inner Blanket Assembly 28.6% 9.1% 5.6% Radial Blanket Assembly 27.6% 10.1% 7.2% These values are for end-of-equilibrium cycle conditions at rated power operation. Based on an uncertainty in decay heat of +0% to -20% used in the Reference 2 sensitivity study, the effect on blanket temperatuS*5 during g natural circulation transient was shown to be frgm t0 F to -g37 F and the ef f ect on fuel temperatures was from +0 F to
-130 F. ,
3.3 CORE INTER- AND INTRA-ASSEMBLY FLOW REDISTRIBUTION AND ASSOCIATED HEAT TRANSFER EFFECTS For the core natural convection cooling. mode, the effect of dynamically approaching low flow with worst case decay heat loads results'in a power-to-flow ratio greater than one. Consequently, core temperatures increase and natural convection phenomena such as inter- and intra-assembly flow redistribution due to different thermal heads and hydraulic characteristics of the core i assemblies become important. In general, the core thermal head becomes significant relative to the form and friction loss across l the core below approximately 5% of full flow. Coupled with the flow redistribution, significant heat redistribution on an inter-and intra-assembly basis occurs throughout the core due to large temperature differentials and an increased heat transport time (low power assemblies can have a transport time of over 20 seconds). These effects (i.e., natural convection flow and heat redistribution) are found to signi'ficantly reduce maximum core temperatures as demonstrated in the EBR-II natural circulation experiments (Ref. 14). Independent studies outside the CRBRP Project have be!n published which also show that reactor flow redistribution causes a significant decrease in predicted maximum core temperatures during natural circulation conditions. For example,,Brookhaven National Laboratory (Agrawal, et al., in Ref. 15), using the 11 f R?-ninA
ssc-L code, predicted localizcd flow increases as large es 20% in the hot fuel assembly and 40% in the hot blanket assembly for the CRBRP during natural convection. Temperature reduction on the - order of 16% and 22% and 210 F) were shown for the hot: ^ fuel and blanket assemblies, respectively, relative to the maximum temperatures predicted without flow redistribution. Similar results were found in Reference 16 using the CURL-L co'de. These studies do not include inter-assembly heat transfer effects, or intra-assembly flow redistribution and heat , conduction effects. However, if included, these effects would. further reduce the maximum predicted core temperatures. Preliminary studies have been performed for the CRBRP to demonstrate the effect of inter-assembly flow redistribution for the heterogeneous core design. The effects of inter-assembly heat transfer and intra-assembly flow and heat redistribution, were neglected in this case. Figure 3.2 shows the results of c these analyses for the peak fuel, peak inner blanket and p'eak radial blanket assemblies. Figure 3.3 shows results for a. myeivat utificing zone for the fuel, inner blanket an'd radial blanket assemblies. Consistent with other natural circulation studies, the flow increase to the hotter core regions is apparent. This effect, along with the other natural convection l phenomena, will significantly decrease the maximum hot rod 1 temperatures in the core. As discussed in detail in Section 2.0, the effect of all natural convective cooling phenomena (i.e., inter- and intra-assembly ! flow redistribution and heat transfer) on the maximum transient coolant temperatures in the CRER core will be assessed using three computer codes (DEMO, COBRA-WC, and FORE-2M). Four cases have been analyzed to illustrate typical results l predicted by this technique for a high temperature fuel rod: l l o Case 1--Fuel assembly peak coolant channel transient flow and temperature calculations including bothfinter- and intra-assembly flow and heat redistributioni nominal-conditions (no uncertainty f actors applied); o Case 2--Fuel assembly peak coolant channel transient Llow and temperature calculations including both inter- and I intra-assembly flow and heat redistribution; l uncertainty factor conditions; l , o Case 3--Fuel assembly peak coolant channel transient flow and temperature calculations only including inter-assembly flow redistribution; uncertainty factor conditions; and g o Case 4--Fuel assembly transient flow and temperature calculations without inter-assembly and intra-assembly flow and heat redistribution; uncertainty factor l conditions. l l 12 n, n,n. l
.' I I! .
, 2 t .
0 1 5 1 S N 0 l
} f
- f ,
/ ' 0 N
l e O C 0 k G I 4 n a
//
W 1 i ' Y l l l l t g A l a M N i t ( l i s
/ $
S A F O R a E L i
? t A 0 t t
t i I I 2 n K T T I 1 a N MN l Y L
/ A L
S R f i T WA W I W k l e n a
'/
O l E OO I M t M L L r t 8 F f e ) S 1 T i 0 n A R E I 0 n t 1 l A L t I t E N N l P I I r U L L u I K A (
~- A A 1
0 O 1 I T
)
S D N l' k a O P
%_ F O O N N F
I 0 8 C E S c l' r t ni, i O I 0 l I i T f E n ics nn N\ C C A A M i a N R R F F I T g nf r
= a l
ie i n l
) t i
o f 0 0 i s t I ( 1 6 a x l F F ' F nln F. H f t i
) v f
" 0 N
wir n t 1 i F ia n 7 I 0 lvi a t i 4 nN i T c g E uins N N A
- A. u r l L eI S
L A I 7 0 f lo ic l t s i Oy I al A Ru t s / 7 i r n ie i Kt As s / IiW v E P A 2
/
s 3
- - /A -
8 7, O e r u 0 9 g I I. o 0 a i I l I F _ e u.1 5 T g 4x9es
)
'^
i l! l l ii i l,l - i ;! 1: i
l E o .K
.= is e
== C r.. 1 p m
E
~ * ( -
5 n J
\ \s w E
O
\ E
~~
% .5
=
ed w &
- 5 E N E "5 : = - I E E 5 z s 2 - e
\ e E
- g g o o
- N s,'
g
=-
\ E a a
- - 5
- - /
w w 5 4
\ & =' en E Eg
\ =g
=
N d .a - O
> 4 4 *- C Y'
.a - e- g -
e O
-go ,y 2 a -
- - -u :
e i o e o _ M O E l E
=z ' c e- - -,
u u e u a
< e.,
m-a _ gw
~ e -: -
a . . w g a / - - :E
~ o r.
/.3 t
/ -
o - S
=
3
= :
n .
. / -
w .4 r .
/ i 4 .
5
=
[ - 3 i I 5- - e>
\
w .: z :
~
a y N
\ - e
~ .
\ ,
=__
\ _
\ :5 I I
\ I 1
~
a w"-
* = n
= e e w
- o v
(0133/(2)d d o S O. b g g O 14
R2culto for thsco four cacao aro pres 2nted in Figure 3.4 so normnlizcd tcmparatura difforcnc22 relativo to tho otoedy ctote temperature difference in Case 1, i.e., l [Tc (tau) -TM(tau) ] i S= i= case 1,2,3-or 4 [Tc (0) - Tm(0)] i where: Tc(tau) = maximum hot channel coolant temperature; Tin (tau) = inlet temperature. As can be noted by comparing Cases 2, 3 and 4, accounting for inter- and intra-assembly flow and heat redistribution effects significantly decreases the predicted transient coolant temperatures in the hot channel. It can also be seen that the uncertainty factors cause a significant increase in the predicted hot channel coolant temperatures (i.e., Case 1 versus Case 2). In summary, natural convection cooling of the core exhibits 6ne of the few CRBRP core design transients where low power /high temperature conditions exist. Due to the long coolant transport time and low pressure drop for the. core while descending into and - operating in this mode, core inter-assembly and intra-assembly flow / heat redistribution; a) becomes significant with regard to accurately predicting temperatures; and. b) significantly decreases the maximum hot rod temperatures in all core regions, when compared to analyses which neglect these effects. 3.4 LOOP FLOW RESISTANCES The primary and intermediate loop flow resistances have a direct effect on the quasi-steady state flows and hot to cold leg during natural circulation. During the transition from forced to natural circulation following a pump trip, the coastdown rate is determined by the system's initial kinetic energy and its rate of energy. dissipation in the pump and system. The hydraulic dissipation rate is directly proportional to the system flow resistance. The plant operating conditions prior to the transient are also directly influenced by the primary system resistance. Resistances lower than that used for the base case analysis (Reference 11) would result in higher flows and, therEfore, lower initial primary and intermediate system temperatures. The results of the base case analysis show that the primary , 15 82-0304
. ~ . .
- i
,. .. l 1.8 -
T CASE 4 CASE 3
"' / %
g e 1.6 - f \ e 8 /7 6"' o I g i 1.4 - g CASE 2
- i
+ 1.2 -
Y l CASE 1 . r w m -% w N 1.0 m
- w
- c [ /
5 0.8 L
/
/
I I w
* /
/
I 0.6 CASE 1 - INTER AND INTR A-ASSEMBLY ! 3 N REDIST RituTl0N :h0 MIN AL Y [ CASE 2 - INTER AND INTRA-ASSEMBLY RtoisT RisuTiON:3 a nCr g o, 4 /
/ 7 CASE 3 - INTER ASSEMBLY REDISTRI5UTION: -
j 30HCF CASE 4 - NO INTER OR INTRA. ASSEMBLY REDIST RituTl0N.30 HC F y v f l i 1 1 1 J I I I 140 160 0.0 80 100 120 40 60 0 20 TIME (SECONDS) l Figure 3.4 Typical Ell'ect of t'ncertaintin ami Inter Intra Awembly Figw Heat Redistribution on F. A Peak Cnolani Temperaturn 1)uring Natural Consertion Cooling . m O 16 esum=
- _ ,- , - - - - - - - , , -. - - - _ _ _ , - -- --m--,
gystcm flow has th9 highest effect cn both initial conditions cnd natural circulation conditions. A3 lcng as tha IHTS resistencas are such that the IHTS flow rate at natural circulation . conditions is higher than the PHTS flow rate, the actual IHTS pressure drop uncertainties have less of an impact on the transient (primary flows) than do the PRTS pressure drop uncertainties. Therefore, sensitivity studies have focused on the primary system. The overall primary loop system pressure drop is composed of that in the reactor (45%), the pump with an assumed stationary impeller (25%), the IHX (5%), the check valve-(20%),
-and the piping (5%). The percentages are based on worst case specifications at natural circulation flows for the respective components and are referenced to the base 3-loop natural circulation case. For natural circulation on less than 3 loops, the percentage drop from the loop components would increase.
These pressure drop fractions, which show the relative impact of the components, plus the results of the sensitivity study, which defines the relative importance of the overall pressure drop, have been used to define component testing which will verify the pressure drop correlations and have been used as a basis for the sensitivity evaluation of the primary system resistance reported in Reference 2. Theimpactofa+17%to-32%primaryloopglow resgstanceuncertaintyrangeat3%flowwasfoundtobe+g0Fto
-7 0 F in blanket assembly hot channel temperature and +66 F to
-100 F in fuel assembly hot channel temperature. The sensitivity
,of temperatures and flows during a natural circulation transient to the flow coastdown rate and to the initial conditions are specifically addressed in the sensitivity studies discussed in Sections 3.5 and 3.7.
3.5 FLOW COASTDOWN l The sensitivity of the maximum cladding temperature for the fuel j and blanket assembly hot rods to various pump and pump drive system stored energy values were performed during the first half of 1975. This study was used to provide a basis Increasing for the I specification of the pump and its drive system. stored energy in the pump results in lower fuel assembly peak temperatures and times after scram at which these peak temperatures occur. As reported in Reference 1, the coastdown characteristics of the pump are a significant factor in determining the flow integral and, therefore, the peak fuel and blanket cladding temperatures. Within the narrow design window allocated for pump inertia, the impact of using the maximum pump inertia in natural circulation analyses, which is 11% higher than that used in the base case (minimum inertia) analysis, was reported by Reference 2 tg reduce the blanket assembly hot channel temperatu5e byTherefore, 15 F and the thepump fuel assembly coastdownhot channel temperature by 20 F. g characteristic which depends on the effective inertia, pumping torques and friction torques, will be tested, as discussed in Section 4.3.1, to determine the actual values for use in natural circulation analyses. . 17
i 3.6 IHX HEAT TRANSFER PERFORMANCE j The IHX is designed to promote mixing on the primary shell side. j However, during a natural circulation transient with low primary flows, one may speculate that buoyancy induced flow maldistributions may exist. Such maldistribution could alter the l unit's heat transfer effectiveness, change the frictional j pressure drop, and also change the unit's contribution-to the thermal driving head. Assuming a range in IHX performance of
+33% to -73% (heat transfer performance) the Reference 1 4
sensitivity study predicts g blankeg assembly hot channel U temgerature variation of +0 F to -7 F and a variation of +7 F to
-13 F in fuel assembly hot channel temperature. This study was j made by assuming that only a specific fraction of the tubes are
! active and the remainder are thermally and hydraulically , isolated. To implement this analysis, the effective heat l transfer area was varied in direct proportion to the assumed ! effective flow fraction. Thus, extremes in IHX flow ! maldistribution show insignificant impact on natural circulation transients. . 3.7 TRANSIENT INITIAL CONDITIONS: FLOWS AND TEMPERATURES i The preliminary evaluation of the CRBRP natural-circulation . capability was based on a set of plant initial conditions for the l primary, intermediate, and steam generator systems that are
- j believed to result in a conservative set of temperatures and
- associated flows. The primary flow used corresponds to the
- minimum flow which would be expected at 100% pump speed since it
. is related to the intersection of the minimum pump head / flow characteristic with the maximum system resistance curve. The ! intermediate loop flows and temperatures were based on the i thermal hydraulic design flow with an arbitrary upward adjustment
- of 20 F on the hot and cold leg temperatures.
The actual plant conditions, however, can vary significantly from f the conditions discussed above. The primary flow depends on the "as built" loop impedance characteristics and pump head / flow characteristics. The IHTS conditions are controlled by the heat transfer conditions between it and the SGS. It is important to note that conditions of fouling and plugging for the steam generator modules will vary with plant lifetime. From a performance prediction standpoint, uncertainties in heat transfer coefficients and the transition from nucleate to film boiling in the evaporator can require the IHTS conditions to. vary significantly. At rated power, for example, the required intermediate loop flow may be at any value between 11.5 and 13.5 million lb/hr. The primary loop flows can vary between 13.8 and 15.9 million 1b/hr. The corresponding variation in hot and cold leg temperatures along with the possible variations in system resistances have an affect on the flow decay curves in the PHTS and IHTS and, therefore, the behavior of the IHX and ultimately have a contribution to variations in thermal heads. , 18 4 82-0304
Sats of tsmparature end flow conditiene will ba selectcd to bracket the possible varictione in those canditions and the sensitivity of the natural circulation controlling parameters will be determined for each case. This sensitivity study of plant initial conditions will be performed using the DEMO computer code and the results will be presented as a part of the natural circulation assessment supporting the final sys, tem designs. 9 D I i l T l 19 l
4.0 VERIFICATION OF SUBSYSTEM CHARACTERISTICS 4.1 TEACTOR THERMAL HYDRAULIC ASPECTS 4.1.1 REACTOR STRUCTURES AND ASSEMBLIES OVERALL REACTOR DELTA P The reactor vessel dynamic pressure drop correlations used in DEMO are as outlined below: Delta P Dynamic = Delta P Total = Delta P Static (4-1) Demo Cobra Demo j O where: Delta P Total = total reactor pressure drop calculated Cobra by COBRA for given reactor flow; Delta P Static = static reactor pressure drop as Demo calculated by DEMO for the same flow conditions; and Delta P Dynamic = dynamic pressure drop to be used for Demo determining the pressure drop , correlations. Using experimental data on pressure drop under low flow conditions through each of the reactor components, the total pressure drop for a series of steady state calculations at various flows with the power-to-flow ratio near one are made. The dynamic pressure drop is then calculated from Equation 4-1. Test programs are either underway or completed to experimentally determine the hydraulic characteristics of all reactor components over the range of expected flow rates. Table 4.1 lists the major reactor pressure drop components and the applicable test programs. These test programs will confirm that the flow rate in
# " 9f*" *#
eachCRBRPassemblywillbeachievedwithi287[# than the value assumed for design analyses This test data
! will determine the reactor pressure drop characteristics down to approximately 2% of rated flow to cover natural circulation operation. Figures 4.1 and 4.2 show typical data for fuel and blanket rod bundles, respectively. As can be noted, a smooth transition between turbulent and laminar flow occurs.
4.1.2 CORE FLOW REDISTRIBUTION AND HEAT TRANSFER Flowandheatredistributionverificationdataarereduiredat high and low steady state operating conditions and during transients. The tests listed in Table 4.1 will provide the necessary data to verify the reactor assembly hydraulic characteristics. Verification of temperature distributions at j 20
. ~_ ..
e.e. . -e
, . . .- . . . . . . > . s ..
-. _ -}
t.. .! L __u.i., l ~... -{. ._a._. l t.-
.e.. 6._ -l 4.. . .. .. a.
i
.i .. . . . _ . _1.
I_.: 1: J..
. . :_1 1- .! ! . .! , . , >
. . ' i ;. . .l..,. .
. q. i.; : .,
+
. . . _ 4. ,..i .,. ., , .. .
- 4. ..
. _.7 s..
....w........,.... * *
,"{ g e.
' I t
;- l :_.__..-
i 1.l ' 1. l i : . . , .
.l
.: , _. _I ..
.l.. ; l.. j l . ,
f":
.il! p; _.1
' - 4_..t..
' s t-
) .
>: s ..e a 'j.
e .. .
. l I ; , I, J 4 I ,; ; . i : i *
- 4. . p _ ;. ,
f' . .. _
._.....i._._..__. ,
. t e l .
. . I .
i, .
- I i .
se
. 1 i l g .
- . .! _ .J e . . . .
..I'..... 4.
. I.g. * ' ' i ,
. -{ 1.. g:
;!r. j
' l l
~E
_- _s . =:. g ! i i l i i; : #
'l'I : . .
5_ -5 *I 5s i . ...
.j. . . ., i I
i e . 9 . . . _ e
. ... . .. . . . . I
,e .;
I. . . ,. .7.....,.. a. .
. . . . . . . , .7. .,_. i.. ., ,. ..g3.
. .. ..4 p.. r _7. ..!.
. . . .. . .: . .j .y.. j .p.
i _ . .; .
._.I
, .j . . H' .1 l.L.;. .
- k, .
L; *m I.' tI1 l i f, 3 t
. t . . , _ .
e .__...,_. t
> m. 1 i
I r. . 1... eja . . ,, E y. .m . I.
._g. *
. - r .
I
.I_l m._' m wT l
. ....S 4
.ii - . l * \ ___ __ _ i I E l ll1 I
.. . . m l * - l- t ! { ,. f. Iy, .
E' O 1 -
** .a .
{ --
. ' . .ase . j ;
-
- j.i. :l '
- l. .
- e. ,
. i
- l. .. .. i.l g . . . ..._
.. i . in.
t 5
! t ;: i ;
- ri , 4. .
gE i 1 iI j /** ii 4 :
.r e., ga l- i .
l
;4 p t . .. _ _
.=
s u .-e . l _._...t ... _ . . _ - . .u . .
- 2. __. ..
. g..
n , . _ _ _ _ .. . . . _ . _ _ _ . _ . .
. y .=... L_ .a, _ __. .,. .
. . u. . , _
.r.. -g.-l .
.. _'*4j*
t 3*,- 3 p. . g . l
.t . . g. . .s . . p y 7. .. .t.
. /_ . : . I.. . _. .
.j
. _- i, i
.i2..
. .i_;
i 4 . ;-
- .. i
._ t .. .
/. . . _ .
- j. p.. l
- i. ! .); '!
i . ,/' .. . , , 5 i l s t. i l. .
.R a
. ~. . l .
. i i , .
- f
, i. t .__ . e
.. i. _ . , .
i I
., .w.m . . w.m muu ..
m
.;_ _ .. _ r. ..s 2,
a_
..:...__.__- u _
e l .-
- . . s . .
71
~.
e . C l L 4 4 io.c
\ M"' f l R ActAL BLANKET ME AT TR A885F ER TEST - NA (
, vsATE R TEST ROOM Tf WER ATumE 4.c - Q CHlU & TOOmE A5
- e g S TO 1 SCALE AIR FLOW TEST !
c m E MM E P/D = 1.125 _ wop uAp.N K P K ths3 P/D = 1.32 2.o *-- i I,s, 1.o e-- \e T R AN5tTION . e C .8 *=- TuarukE NT h y b :
=
e % C
- LAMth AR si .
l . E ! 2 -- ,1; - . a ... .
=
l
.s o- a
.oss os - !
.oe'-
l . t i I
.ca -
5 e' i . I tl I ; i tt i kiI? ? I a a a tooooo
.cs -
1 a s a tooo 2 4 s: 10o0c 2 ! a s a too 2 to 2 400 SUNOLE REYNOLDS NO. t Figure I. 2 Average Friction Factor Data for a Wire Wrapped Rod Bundle: P/D = 1.08. 4" Wire-Wrap Pitch Sicilar to CRSR Elanket Assemblies 03691
! 77
BBLE 4.1 RFX70R COMPCNENT PRESSURE DROP VERIFICATI0t1 TESTS Pressure Drop Component Test Status - J.nlet Plenum & Modules Inlet Plenum Feature Model Hydraulic testing cmpleted 1976 (1/4 scale) - Ar ' Plenum and Upper Integral Reactor Flow Delta P data otuupleted
..als Model (IREM) 1977 (1/4 scale)
IREM Core and Permanent Complete testing Reactor Structure 'Ibermal in 1982. Report to Striping Test be issued 1983. Radial Blanket Orificing Radial Blanket Flow Orifice Calibration Orificing completed 1979, complete testing LIM in 1982. Report to be issued 1982. Fuel Assenblies: Inlet Nozzle, Shield F/A Inlet Nozzle Flow Completed full and Orifice scale water test 1979 Cavitation and Orifice Testing ccrnpleted in Calibration Test 1982. Report to be issued in 1982. Rod bundle Inlet and F/A Flcw and Vibration Full scale water Bundle Outlet Nozzle Test test ccrnpleted 1980 Outlet Nozzle F/A Outlet Nozzle Flow Full scale water test Test test completed 1979 l Blanket Assemblies: Rod Bundle, Rod Bundle Radial Blanket Assembly Full scale water test Inlet and Assembly Flow and Vibration Test ccxnpleted in 1982. Outlet Report to be issued in 1982. Control Assemblies: Control Assembly Hydraulic
~
Full scale water test Tests completed 1982. Report to be issued in 1982. Renovable Radial Shield Assenblies: 1 RRSA Orifice Pressure RRSA Orifice Pressure Full scale water Drop Test Drop Tests test ccznpleted 1977 9 23 82-0304 - - -- - . _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ -
high reactor flow rates in fu21 castmblic will drcw upon tha inrgo body of cub-channel mixing cnd haat transfer test data collected in support of both FFTF and CRBRP. Calibration and verification of temperature distributions in blanket assemblies will be provided by the full scale blanket heat transfer test program with sodium coolant. This blanket testing by W-RM provides detailed data on intra- and inter-assembly heat transfer for a wide range of representative blanket power distributions at low, intermediate and high flow rates. Transient forced and natural circulation data in the natural circulation flow range are also available. Fuel assembly low flow intra-assembly flow and heat redistribution data are available from the Fuel Assembly Natural Circulation Sodium Heat Transfer Test at HEDL. Detailed steady state and limited transient heat transfer data in sodium on fuel rod bundles of 61 pins are also available from the heat transfer test program conducted at ORNL for a wide range of flow and heat input conditions. An extensive wire wrap bundle mixing test data base is available which has been utilized to improve the understanding of the thermal / hydraulic phenomena within wire
.._yp.3 ::d bundles for calculation of subchannel analysis computer codes. References 12 and 13 describe many of these tests.
Aside from the single assembly data described above, valuable core systems data for confirming the verification of the core inter- and intra-assembly flow and heat redistribution analyses is available from the EBR-II and FFTF natural circulation tests. In EBR-II detailed flows and temperatures have been measured in the instrumented XXO7 and XX08 assemblies which were prototype EBR-II fuel assemblies located in the core during the testing. Similar FFTF data is available from Open Test Assemblies (OTA's).
- A high flow /high power zone and a low flow / low power zone l
prototypic fuel assemblies were located in two OTA's during the FFTF testing. Both the EBR-II XXO7 and XX08 assemblies and the FFTF OTA's had flow meters and numerous thermocouples at various axial and radial locations throughout the rod bundles. These l data sources are extremely valuable for validating the local, as i well as the core wide, temperature calculations made by the l COBRA-WC and FORE-2M computer codes. Many irradiation experiments have been performed in EBR-II to verify the fuel restructuring characteristics under prototypic fast reactor conditions. This includes fuel and blanket rod type irradiation data. Similar data will be available from FFTF to support the CRBRP FSAR. This information is being used to verify the steady state fuel behavior code LIFE (Reference 17) l which in-turn is used to validate the restructuring model in the transient core hot rod analysis code FORE-2M. The restructuring validation is necessary to assure that phenomena which offset the rod stored heat release / inertia are satisfactorily modelled. Tabl'e 4.2 summarizes the verification approach for the COBRA-WC and FORE-2M analyses described above. . 24
TABLE 4.2 COBRA-WC/ FORE-2M CODE VERIFICATION APPROACH , o Apply steady state single assembly data'taken at conditions indicative of natural circulation (High Power-to Flow Ratios, Low Flow and High Power Skews) HEDL 217-Rod F/A low flow data in sodium (includes data with one side of bundle heated and other side unheated)
-- ARD 61-Rod B/A data in sodium (Steady State and Transient Data includes bypass channels to typify inter-assembly heat transf er)
-- ORNL 61-Rod F/A data in sodium (Steady State and Transient Data)
-- Fuel, Blanket and Control Assembly hydraulic cha.racterization from 100% to low flows o FFTF and EBR-II Quasi-Steady State and Dynamic Data for Coupled Core System Low power and flow quasi-steady state tests Dynamic total loss of power tests from icitial high power / flow operation a) Single XXO7 and XX08 instrumented assembly for accurate flow and rod bundle temperature data in EBR-II tests b) Multiple Open Test Assembly (OTA) data for accurate flow and rod bundle temperature data in FFTF tests o Irradiation data from EBR-II and FFTF to confirm fuel restructuring, fuel / cladding gap conductance, stored heat in rod, etc., for fuel and blanket type hot rods (by direct comparisons and by LIFE computer code calibration) t 0
25 R?-010A
l 4.1.3 REACTOR UPPER PLENUM MODEL Sensitivity studies have shown the necessity to verify the outlet plenum thermal transient behavior during a natural circulation event. This temperature response at the outlet nozzles can directly influence the overall reactor plant performance during the event. , Although sensitivity studies showed that mixing / stratification in the outlet plenum had a minor impact on overall natural' circulation in CRBRP, a significant number of tests to assess the outlet plenum characteristics have been performed. These tests include Argonne National Laboratory (ANL) 1/10-scale model water tests (Ref erence 21) , ANL 1/15-scale model water and sodium tests (Reference 22), and Battelle-Columbus Laboratory (BCL) 0.55 scale model water tests (Reference 23). It should be noted that water testing in small scale models in adequate to simulate full scale CRBRP sodium conditions. The ANL experiments indicate that no significant difference exists between water and sodium testing (1/15 scale models). Additionally, comparison of the BCL 0.55 scale, ANL 1/10 scale, and ANL 1/15 scale model test results indicate little difference in transient behavior. l The data obtained in the water natural circulation i were compared to predictions obtained from VARR-II(ggperiments ,a . numerical hydro code. This code has been verified and used to predict outlet plenum transient responses at the' outlet nozzle for all natural circulation events that will be considered for
, evaluation in CRBRP. The outlet nozzle responses predicted by the numerical hydro code will be used in conjunction with the overall plant performance code such that accurate feedback, due to plenum transient behavior, will be included in the system model. .
The tests discussed above also showed that the UIS chimneys
~
mitigate' flow stratification and improves fluid mixing. However,
- as discussed in Section 3.1, the major impact that the UIS has on the natural circulation event is due to its impact on natural I heads / loop pressure drop and therefore flows. The thermal head's retarding of flows in the DIS makes the value of the hydraulic resistance across the gap between the bottom of the UIS skirt and the removable radial shield (RRS) nozzles important. Therefore, 1/4 scale model IRFM water tests have recently been conducted to determine the pressure drop characteristic of this gap. The results of these tests will be used to set the range of gap resistances in system analyses.
l l Confirmation testing for CRBRP models is not possible from FFTF l natural circulation tests. This lack of verificationgis due to the differences in the upper internal structures. The FFTF instrument tree and CRBRP UIS chimneys have significantly different impacts on both the outlet plenum mixing and the UIS impedence/ thermal head affects, however, sufficient tests are in place without FFTF test input to demonstrate that th~e outlet t l 26
- plenum models that are importent to natural circulation analy202 j in CRBRP.
~
4.2 DECAY POWER The post-shutdown reactor power generation is comprised of the i fission power and decay power. The fission power is calculated l in DEMO and FORE-2M with a point kinetics model. COBRA-WC uses
- the DEMO calculated fission power. The decay power, on the other i hand, uses data from separate decay power calculations and is input in tabular form as a function of time, with separate tables
- for the fuel and non-fuel assemblies. Thus, the codes (analysis
- methodology) and input data used to generate decay powers for DEMO, COBRA-WC and FORE-2M analyses are included in the
- verification of these system codes. ,
i ! Decay power predictions of the reactor core include contributions from fission product, transuranic, actinide, and neutron activation isotope decay energy release. Isotopic fission and > capture rates in the CRBRP fuel, inner blanket, and radial blanket assemblies are used in the prediction of the reactor system decay power. Verification of the data for fission product j isgggpegggayenerg58
,U , and U release due to fast neutron fission in has been provided by detailed comparison to
- Pu i experimental data. The transuranium isotope decay energy release l is provided by decay data from Nuclear Data Sheets or Table of
! Isotopes and exists as reference data. Verification of isotopic l neutron fission rates and neutron capture rates is provided ! through an extensive critical experiment program. 1 i Time dependent uncertainties associated with fuel and blanket I assembly decay power result in approximately a 28-33% increase in j predicted decay heat at shutdown and approximately a 6-10% i increase at 2 hours after scram. Actual uncertainties are l dependent on the type of assembly, i.e., fuel, inner blanket, or l radial blanket. The fission rate Sata uncertainty is included directlyinhotchanne}3gactorsusedfortheanalysisofthehot rod in an assembly. U capture rate uncertainties are verified i values and uncertainties are included in transuranium isotope
- decay energy release data.
i ! ENDF/B-IV nuclear data (fission product isotope decay constants, i energy yields, and branching ratios) for 824 fission product isotopes and fast fission yields for 6 fissionable isotopes for
; application to CRBRP were provided upon completion of a i contributing base development program. This data bank was used I
to provide decay energy release data for each fissionable isotope for use in a design methodology that considers the detailed j spatial and time variation of isotopic fission rates (or full
- power, partial power, or load. follow conditions. Verification of the ENDF/B-IV data base was achieved by prediction of decay heat I
experiments where individual fissionable isotopes were irradiated and decay energy release was measured by spectrometric or l calorimetric methods. Use of the design methodology method (S-4M 27 on_n, -
or RIBD-II computer prcgram) to predict exparimtntG CDUncertainty a function oftimeofirradiggjonandgggaytimewasperformsd. and Pu have been completed and documented i assessments for U This information, as well as analysis of l in References 24"35. current experiments, provides verification of decay energy release data. , I verification of the codes and input data used to generate decay l heat values for DEMO, COBRA-WC, and FORE-2M analyses will be j complete by July, 1983. . !. 4.3 PLANT COMPONENT THERMAL HYDRAULIC CHARACTERISTICS 1 4.3.1 PRIMARY AND INTERMEDIATE PUMPS The pump characteristics i.nportant to the analysis of natural circulation events are those which influence the flow coastdown and the resistance to flow after the pump has stopped turning. j The pump characteristics are second only to theThe reactor pump decay coastdown power in their effects on peak temperatures. affects the peak core temperatures in three ways. First, the integral of the flow coastdown determines the amount of sensible Secondly, the time heat that has been taken out of the reactor. 4 at which the pump stops determines the reactor decayThirdly, power Igvelthe at which transition to natural circulation occurs. stopped pump impedance represents approximately 25% of the loop pressure drop at natural circulation flows. The initial primary flow coastdown (down through 10%) is largely determined by the pump and pump drive system stored energy and l
- this follows the system pressure drop curve down to low flow.
This coastdown (initial flow coastdown and subsequent pump tailoff) affects the power to flow history and hence the i l temperature transients in the core. The preliminary natural circulation calculations have been based on minimum specified flow coastdowns with conservative assumptions on the pump tailoff . (high values for the loss torques at low speed). l The pump pressure drop, when the shaft stops rotating, has been J calculated based on an equation which treats the pressure drop as fully turbulent (flow exponent of 2.0) and the maximum The specified coefficient on the normalized flow has been used. relationship implies a pressure drop of 54.6% of design head, should design flow be forced through the stopped pump in the normal flow direction. Based on natural circulation experiments l at EBR-II, analysts have revised estimated pump pressure drops to significantly higher values. A flow exponent of 1.8 was j recommended together with transition to laminar at high Reynold's ' number appro::imately (50,000) which resulted in a turbulent to laminar transition at about 6% flow, which is well above the 3% flow' range expected for CRBRP primary flow in the natural circulation mode. Experiments to measure the locked rotor pressure drop of the FFTF pump at LMEC indicate, however, that a i 28 on n,ne l ._ . -. - - ._ .. _.
for the F.FTF pump at least, no cuch transition to laminar flow was evident for flows down to approximately 2%. j Preliminary performance data has been developed, based on vendor water test data. This data has been used to make a determination of the coastdown characteristics and stopped rotor impedance for use as interim data until sodium test data is available. l During prototype pump testing at the Sodium Pump Test Facility (SPTF), a series of'coastdown tests will be conducted to verify that the equations modelling the decay in flow are, in fact, conservative. By running a series of tests, data will be generated to determine a mean and variance for the pump coastdown and locked rotor resistance. These statistical values can then be used to verify the conservatism used in the system simulation. Prototype pump tests at SPTF began in 1982. Therefore, both the calculated values and preliminary water test data used in the model will be verified during sodium testing at SPTF. Water tests of production pumps will be conducted at the vendor's l facility. These tests are expected to confirm the repeatability of the prototype pump characteristics particurlarly with respect to the pump "tailoff" and the locked rotor resistance. The method used to develop the pump characteristics from test data , has been verified (Reference 9). Excellent comparison of DEMO l calculations and measured speed and flow coastdowns at FFTF were l obtained. 4.3.2 IHX Verification of the IHX model in the DEMO code is important for two reasons. The thermal performance of the IHX is important to , assure that the overall plant response is correct and secondly, to assure that the structural design transient applied to the IHX l in the equipment specification is adequate for structural i analyses.
- Both of these considerations are dependent on the assumption that a one dimensional model adequately describes the hydraulic
' characteristics of the IHX in the natural circulation flow range (1% to 4% flow) . Verification of the one dimensional flow model can be accomplished most efficiently as an integral part of the l whole plant performance analysis verification which is discussed j in Section 5. It should be noted that the existance of localized j phenomena such as buoyancy effects or flow maldistribution are of no concern unless they impact the, natural he~ ads and loop flow rates. Therefore, system testing is adequate to verify the IHX model used in DEMO.
~
, The adequacy of the one dimensional flow model for thg secondary l (tube) side of the EBR-II IHX.has been demonstrated by comparisons to test data (Reference 6). The tests at EBR-II i centered on determining the acceptability of the heat transfer j model and the one dimensional flow model. The results of this } 29 l 4 no_nana
testing was close agrecmsnt batwocn nodal predictions cnd measured EBR-II data, verifying both the product of the overall heat transfer coefficient times the effective surface area (UA) and tube side one dimensional flow assumption. Post-test
> analysis of FFTF natural circulation tests is also expected to contribute to the primary side one dimensional flow model-verification. Pretest predictions for natural circulation testing at the FFTF facility have been issued in Referdhee 7.
l 4.3.3 COLD LEG CHECK VALVES - 4 ! The hydraulic design of the CRBRP 24" Cold Leg Check Valve (CLCV) i is based on hydraulic similitude with the 16" FFTF Check Valve. The flow testing performed on that program, 6" model size and full flow tests on a 16" prototype has provided the data for the 24" valve design. Due to this similarity in design between the ' FFTF tested valve and the CRBRP valve, no testing of the CRBRP i valve was performed. A detailed report (Reference 8) has been prepared which justifies the applicability of existing test data
--' ;"r-tifies the uncertainties in the extrapolation of test data to the CRBRP valves.
The cold leg check valve characteristic is an important part of 1 the verification of natural circulation because it represents approximately 20% of the primary loop resistance at natural . circulation flows. Based on the results of the detailed i evaluation of the CLCV, the model used in the preliminary i evaluation of the natural circulation capability of CRBRP, was
- found to be overly conservative. Therefore, flows predicted were
! lower than expected. 4.3.4 PIPING STRATIFICATION AND MIXING, AND HEAT TRANSFER In DEMO' calculations, thermal driving head is calculated as a 1 summation of the sodium densities around the respective primary i and intermediate flow loops. Fluid stratification and fluid to j structure heat transfer can potentially have a significant impact
- on natural circulation transients. This impact would be not only j on the transient rate, but on the thermal driving head. The total primary system thermal driving head, with the design hot to
> cold leg temperature difference, is only approximately 1 ft. Fluid stratification in piping during the transition to and I operation at natural circulation conditions has been investigated by test at ANL. The tests demonstrated that for temperature and i flow transients during natural circulation events, one j dimensional models are adequate for use in codes (such as DEMO) l to make predictions of natural circulation performance of piping systems in LMPBRs. The significance of the test results and I fluid stratification in LMFBRs during natural circulation events is discussed in a report on the effects of natural circulation induced thermal stratification on plant system simulation i (reference 36). .
~
t 30 I n, n,n.
i An discussed in Section 4.3.2, tests Et EBR-II and FFTF hava algo 1 j shown that a one dimensional flow model is adequate to define the
; natural circulation flow phenomena. On the local scale there may I be some stratification, however, these tests have shown that a one dimensional model adequately defines the response of the 1
system as a whole. j A detailed evaluation of the importance of sodium to structure
- heat transfer was performed. This study (Reference 10) .
> determined that it is important to account for piping /IHX plena / pump metal to fluid heat transfer in the analysis of low flow l transients where thermal heads are important. Therefore, the
" bucket brigade" delay model previously used in DEMO analyses has been replaced. Verification of the sodium to structure models
- has been accomplished through comparisons with EBR-II data (Reference 6) and will be further verified by data from the FFTF
] natural circulation tests. 4.3.5 WATER / STEAM SYSTEM NATURAL CIRCULATION j The natural circulation performance of the steam generator system j (SGS) does not have a significant impact in the peak core temperatures which occur immediately (approximately 10 minutes) i after shutdown. This is due to the long (approximately 10 minutes) loop transport times from the SGS back to the reactor resulting from the low sodium flow rates inherent in the natural i circulation mode. However, the SGS does affect the rate of heat transfer from the IHTS later in the transient and, consequently, influences the thermal head and resulting flow in the IHTS. The IHTS flow determines the elevation of the effective thermal center in the IHX and, therefore, is one of the important factors which determine the PHTS thermal head and resulting flow. i Therefore, the analytical models for the SGS must be verified to ] insure that the system will provide adequate natural circulation i flow to remove decay heat.
- Literature surveys have been conducted to determine the j appropriate heat transfer and pressure drop correlations to use i in the analytical model of the SGS. As a result of these surveys
. it has been determined that well established correlations are l available for the SGS loop components (piping, valves, and pump) so that additional verification is not necessary. However, data in the literature for the thermal hydraulic performance of heat transf erring components at low flowrates is not as well developed. Therefore, it has been concluded that additional data i are required to adequately verify the analytical techniques for the steam generator modules to be used in the DEMO computer code. Continuing verification efforts are concentrating on (1) follow of ongoing industry tbermal hydraulic studies (2) comparisions of calculations with test data from EBR-II t i . 4 4 31 1 _ . - - N Y." i _ .
(3) cvaluation of tests to be run en the CRBRP prototyp2 steam generator in the Sodium Components Test Installation (SCTI) at ETEC o Steam Generator Modules A literature survey has been performed of available ' T information relevant to waterside heat transfer coefficients and heat transfer coefficients for natural circulation conditions. The literature reviewed included published information on natural circulation studies for light water reactors, including foreign facilities, as well as forced f circulation studies. x The correlations used in DEMO were also compared to correlations that have been used in other computer programs, including the TRAC computer program (a State of the Art two phase steam and water thermal hydraulicThe code developed review bytwo covered Los Alamos National Laboratory for NRC) . phase flow pressure drop correlations, heat transfer correlations for subcooled, nucleate boiling, film boiling and super heat regimes, and departure from nuclear boiling correlations. These correlations govern the heat transfer and flow through the evaporator and are among the most important features of the steam generator system The that " i influence sodium side natural curculation flow. conclusion of this detailed review was that the correlations used in DEMO are appropriate for the thermal hydraulic conditions during the transition to and during natural circulation. Additional confidence that these correlations are fully applicable to CRBRP will be gained from the SCTI tests of the prototype steam generator, o Steam Drum The preliminary analysis (Reference 11) assumed no mixing of auxiliary feedwater (AFW) with drum water during the natural circulation transient. This resulted in a significant cold The design sodium transient at the exit of the evaporator. of the steam drum has since been changed to include an auxiliary feedwater sparger which sprays the AFW into the steam space. This design significantly reduces sodium temperature transients. Additionally uncertainties in modeling the mixing of cold AFW water with the liquid in the steam drum are eliminated since the AFW spray droplets are heated to saturation temperature as they fall to the liquid Therefore, with the revised design, the surface in the drum. need for verifying AFW/ drum water mixing models is not required. o Recirculation Pump The SGS recirculation pump design, manufactured and testing has been completed. Vendor data for pump opera' ting 32 R?-nind
parformance, constdown charactorietica and stelled rotor hydtculic impendenca is being incorporated into the DEMO analytical model of the recirculation pump during the . transient to natural circulation. j o Piping and Valves Modeling of the SGS piping is being updated to the final arrangement. Vendor data for the various valves in the system will be incorporated into DEMO as the data becomes available from the vendor's design efforts. o EBR-II Tests I The DEMO steam generator model has been used to calculate EBR-II transients for a scram with loss of flow. As discussed in section 5.2, this comparision demonstrates that the sodium side and steam generator models are properly coupled and that they reliably predict overall system dynamic response. o SCTI Tests Tests on a full-scale prototype steam generator module are planned at the ETEC SCTI facility. Both steady state and transient tests will provide data which can be used in verifying the steam generator performance and the' validity of the DEMO analytical model of the SG module. Steady-state thermal hydraulic performance will be obtained at low levels representative of steam generator operation under plant emergency decay heat removal conditions to evaluate steam side two phase flow stability and sodium side This testing temperature stratification characteristics. will employ low flow sodium (shell) side forced circulation with water (tube) side natural circulation. A set of transient tests will be performed which will envelope the fastest and slowest sodium temperature decrease rates anticipated for initiation of natural circulation in the Clinch River plant. These transients have been developed to be within the design capabilities of the prototype steam generator. They will employ programmed changes in sodium flows and temperatures with water (tube) side natural circulation. Analyses of the steady state and transient tests will be conducted to verify the models used to predict CRBRP g natural circulation. O 33 O O h
1 R vicw of pertinent literaturo to dato end tho engoing DEMO verification efforts indicates that natural circulationThein the CRBRP SGS.is a viable method of removing decay heat. remaining efforts are focused on increasing the confidence in the accuracy of the calculations specific to the CRBRP SGS and its A components, particularly the steam generator modules. The approach being taken is to perform steady state and tran61ent thermal hydraulic performance tests on the CRBRP prototype steam generator water side operating in the natural circulation mode. The resulting data will be used to further verify the correlations and models of the steam generator to be used in the DEMO code. The correlations used in the model and a reference or j justification for their use will be provided in an updated Natural Circulation Report. 9 1
~
( 34
5.0 VERIFICATION OF WHOLE PLANT ANALYSIS Tests conducted at two separate reactor plant facilities have become an important part of the natural circulation verification program. Secondary loop and core testing conducted at EBR-II and plant testing conducted at FFTF provide a significant contribution to the verification of analysis methods used to predict CRBRP performance during a natural circulation transient. The comparison of predictions using the computer codes employed f or CRBRP transient analysis against FFTF and EBR-II testing data not only provides detailed verification of specific models, but also verifies that the computer codes prcperly integrate the models required to adequately characterize a loop type LMFBR over the range of plant conditions sustained by the plant during a natura) circulation transient. 5.1 FFTF TESTING A series of natural circulation tests were conducted at the FFTF as a part of its Acceptance Test Program. These tests not only demonstrated the FFTF's capability to remove decay heat by natural circulation but when coupled with pretest predictions, verified the analysis methods employed for making predictions of natural circulation behavior for conditions other than those tested. 'The FFTF tests which are being used to verify the codes used for CRBRP natural circulation analyses are as follows: A. A transition to natural circulation in both primary and secondary loops from 35% power and 75% flow. One of the secondary loop pony motors remained engaged during the-test. . B. A transition to natural circulation in both primary and secondary loops f rom 75% power and 75% flow.
~
C. A transition to natural circulation in both primary and secondary loops from 100% reactor power and flow. Pretest predictions and post-test analysis including comparisons with actual data from these tests are a key element in the whole natural circulation verification program. These tests have provided a unique opportunity for demonstrating that the DCMO, COBRA-WC and FORE-2M codes used f or CRBRP analysis provide adequate (conservative) predictions of the dynamic response of a similar, loop-type, sodium cooled reactor. From the standpoint of whole-plant analysis (using thg DEMO code), good agreement between pretest predictions and FFTF results will demonstrate: o that all models necessary for simulation are, included, 35
o that all phenomenn important to natural circulation predictions have bien includad, cnd o that component interactions (synergisms) do not produce unexpected, significant results. This activity will also verify pump dynamic models, pressure drop models, thermal models (for piping, pumps, heat exchangers) , as well as the methods to calculate thermal heads on a dynamic basis. . Since the FFTF employs instrumented assemblies, the above mentioned tests also provide the data necessary to check the intra and inter assembly heat and flow redistribution which are important modeling considerations in the COBRA-WC and FORE-2M codes. Finally, the verification tasks associated with these tests provide the means to demonstrate the practicability of the three-code concept. "ba-a ara recognized differences between the FFTF and CRBRP that are important to predictions of plant response to a natural circulation transient. The FFTF employs a dump heat exchanger (DHX) rather than a steam generator as a heat sink. Physical differences in the loops (e.g., piping lengths and elevation differences) alone result in a different effect on the secondary system cold leg temperatures and dynamic thermal heads. The' basic phenomena important to making predictions of the responses of these two loops, however, remain unchanged by the above mentioned physical differences. In the reactor area, the FFTF does not contain blanket assemblies and the flow patterns in the upper plenum are different from those predicted for CRBRP. As with the secondary loop, although physical differences in the FFTF and CRBRP reactors exist the phenomina that must be modeled to characterize the response to a natural circulation transient are unchanged. The basic plant response to a reactor scram with" simultaneous sodium pump trips (with de-energized pony motors) is the same for the two plants and the problems of calculating, on a dymanic basis, the thermal heads, flows, core flow redistribution and thus core (including individual pin) temperatures is the same. Only the physical data (mainly geometric) is different. Best estimate predictions were made, in addition to the design case predictions. The design case predictions were generated by applying uncertainties to the various parameters which can affect the temperatures and flows (such as individual pressure drop correlations, decay heats, etc) . Since the predictions were made using design data (i.e.; design information as opposed to plant test data) that existed prior to the conduct of any FMF plant tests, any predictions made for CRBRP prior to its operation will be based on data having the same level of validity (i.e.; data generated as part of the plant design effort) . If when using the same level of uncertainties for these data, predictions are made 36 09_ninA
for FFTF which cro chown to b2 conservativo, whtn ccmparcd to nctural circuletion toct dato, tha.orgum:nt 10 =Ed2 that th2 design case predictions for CRBRP will likewise be conservative. As a part of the post-test analysis effort for the FFTF tests, comparisons will be made with the best estimate calculations, and any differences between these calculations and actual plant data will be resolved. The intent will be to show that the current models are adequate or that one or more need to be modified to bring the best estimate calculations in line with the test data. The generation of the pretest predictions had to, of necessity, make assumptions with regard to the power history prior to initiation of the test, as well as the heat sink (DHX) boundary conditions. It turns out that these boundary conditions (DHX outlet temperature as a function of time as well as the power-history-influenced decay powers) were not the same for the actual tests as those used in the predictions. This, then, necessitates an upgrading of the predictions as part of the post test analysis effort. As discussed in the component verifications (Section 4.0) assumptions in reactor outlet plenum mixing, IHX flow, and loop flow modelling will be qualitatively verified. Although components are not instrumented in enough detail to develop exact models of the component response to transients, sufficient data
'will be developed to determine if the modelling presently contained in DEMO is satisfactory to characterize the overall plant response to the event.
5.2 EBR-II TESTING The results of EBR-II tests provide a significant basis for verification of parts of the DEMO, COBRA-WC, and FORE-2M models. For this test evaluation, the DEMO models of the EBR-II IHX and secondary loop hot leg piping were coupled (i.e.; the DEMO calculated IHX secondary outlet temperature was used as an input to the piping calculation); the measured IHX primary and secondary inlet flows and temperatures provided the boundary conditions, and the calculated temperature response at the superheater inlet was then compared with measured values. As discussed in detail in Reference 6, the close agreement between analsyes and measured data for this testing demonstrates that the numerical techniques embodied in the DEMO programming are adequate to define the IHX and secondary loop piping response to a. natural circulation transient event. The importance of l sodium to structure heat exchange in piping and plenu models, which was verified by closed f orm solution (Reference (10) , was demonstrated at EBR-II as well as the adequacy of a one dimensional flow model for piping analyses. EBR-II provided significant insight into the IHX secondary side (tube side) modelling requirements. The one dimensional hydraulic model of 37 OS ASAA
- - . - _ ~ _ - - - . - - - .
~
the IHX in DEMO cd;quatoly prcdictcd tha hydrculic rc7pon30 to trcnsiento. Thorcfero, if ctrcaming or tub 2-tc-tubo ficw oscillations exist, thcy are not cignificnnt cnough to impact the overall response of the unit to natural circulation conditions. The EBR-II staf f at ANL has merged a DEMO model of the EBR-II secondary loop and steam generator with their reactor and IHX
- to simulate the overall plant. This merged code model (NATCON) The ANL analysis of a LOF/ SCRAM at EBR-II is called NATDEMO. and comparisons with measured data with NATDEMO (Reference 20) serves to verify the whole-plant modeling capability of DEMO.
The verification was based on the conclusion of this study, that ~ NATDEMO simulates the system well enough to accurately predictThe pa n natural circulation transients. DEMO provides further verification of the methodology of calculating pressure drops, thermal heads, and loop heat - capacities that are planned for use in evaluating CRBRP natural circulation transients. Verification of the COBRA-WC and FORE-2M models centered on the results of data collected Severalfrom two instrumented different categories ofcore assemblies transients are (XXO7 and XX08) . experienced with regard to primary and secondaryTwo flowsetsvariations of tests and initial power and primary flow conditions. were selected based on constant secondary system flow and high are These initial (pre-tripped) flow and decay power conditions. ! Test F of assembly XX07 and Test 7A of assembly XX08.- Test F from the XXO7 series involved a loss of primary forced l flow which was being provided by the auxiliary pump while the l l reactor was shut down and the fission-product The secondarydecay flow was power heldwas 1.6% of rated power (60 MWt) . Among the flow and l I constant at 2% of its rated value. temperature measurements through the instrumented mid-core and fueled driver near-core assembly, coolant temperatures at inlet, exit were used to compare with FORE-2M code predicted values. Test 7A from XXOB series initiated under steady state operating conditions of power at 17.1 MWt (28.5% rated power) and a flow rate of 2626 gpm (32.1% of rated flow) . The transient was initiated by interrupting the electrical power supply to the motor-generator set driving the primary pumps, the primary The auxiliary pump having been previously de-energized. secondary forced flow was Comparisons maintained during the initial three with code predicted minutes of the transient. coolant temperatures were made thr,ough "near center" elements (average of four fuel elements near the assembly center) . Comparisons were also made with COBRA-WC code predictions. Selected EBR-II experimental data on the assemblies Assembly averaged w@re compared values are with FORE-2M code calculations.As such, the heat, pressure drop and used for XXO7 comparison. flow redistribution within the assembly and the inter-assemblyIn the case of X heat transfer were not considered. 38 82-0304
data (ncar centor cubchennalc cro treatcd) tho intrc-oceccbly flow redistributien was simuletta by varying the ficw maldistribution factor. In general, the FORE-2M model over-predicts the maximum core coolant temperature reached during the natural circulation transients. By considering.the' flow . redistribution alone, the FORE-2M predictions agree very well ,
- r with XXO8 assembly test 7A data in "near center" subchannels. ,
The largest differences in coolant temperature comparison cases occur when near corner element measurements were compared with FORE-2M results. Edge channel overcooling by inter-assembly heat transfer and intra-assembly flow redistribution are the causes of the deviation. When these factors are taken into account, as was A done in the COBRA-WC model, the agreement was improved. refined FORE-2M model was established for FFTF natural circulation transient test calculations to incorporate these factors. Consideringthepowerandflowmeasurementuncertagntiesofabout '
+10% in these tests, it results in a Therefore, band of 10-20 F variation the FORE-2M resultsin s s
the measured coolant temperature. ,
;?
are within the region of test data uncertainty. More ' 1 importantly, in all cases, FORE-2M predictions of maximum \ temperatures are conservative. 9 I l f I - s I l [ . 39
6.0 Natural Circuhtien vermcauun mmuu,u CY 1981 CY 1982 CY 198 , Completed to Date 4 Component System Verification 3 L 4 1- 2 3 4 1 2 ~ 3 kd Ab Steam Generator Pump A is Check Valve /_k IHX d d Piping A d Decay Power d ; A [ Upper Plenum _b d bb d Reactor Flow Redistribution Reactor Pressure Drop k_dkkA AAA d Ahh
,g
- 14. ORNL 61-Rad F/A Testing.
- 1. Component Detailed Analysis.
- 15. 1/4 Scale Inlet Plenum Feature Testing.
- 2. Natural Circula tion Report
- 16. 1/4 Scale IRFM Reactor Testing.
- 3. SCTI Module Testing.
- 17. IRFM AP' Testing 0 Low Flows.
- 4. Prototype Pump Water testing by vendor.
, 18. LIM Testing and Radial Blanket Orificing.
- 5. Prototype Pump Sodium Testing at SPTF. Full Scale Inlet Npzzle Water Test.'
19.
- 6. Check Value Hydraulic Analysis Report,
- 20. Cavitation and Orificing Calibration Test.
- 7. EBR-II Test Verification.
FFTF Post Test Analysis Complete.
. 21. .F/A Flow and Vibration Water Test. '
! 8. '
- 22. F/A Dutlet Nozzle flow Test.'
- 9. Pipf ng StratWication Study Complete, R/B Assembly Flow and,Vibfation Water Test.
23. ^
- 10. Assessment of Fast Fission Decay Energy
- 24. Full Scale Control Assembly Hydraulic Water 1 Release Uncertainties.
- 25. Full Scale RRSA Orifice Pressure Drop Tests.
- 11. Isotopic Fission Capture Rate Verification Fm ZPPR Exp.
l 26. 1/4ScaleIRFMTesting(UIS-RRSGapAPatIb
- 12. Impicment Methodology in Design. Flows)
- 13. , Scale Model Testing.
e W~ J . - . . _ _ _ _ . _ _ -
2
7.0 REFERENCES
- 1. S:L:1195, A. R. Buhl to R. S. Boyd, " Natural Circulation Decay Heat Removal Verification Plan," June 21, 1976.
- 2. WARD-NC-3045-1, "LMFBR Natural Circulation Verification Program (NCVP) Review of Experimental Facilities and Testing Recommendations," by R. D. Coffield and B. P. Planchon, July 1977.
- 3. CRBRP-ARD-0005, " Clinch River Breeder Reactor Plant LMFBR DEMO Plant Simulation Model (DEMO) ," by W. H. Alliston, et al, February 197 8.
- 4. CRBRP-ARD-0142, " FORE-2M: A Modified Version of the FORE-II Computer Program for the Analysis of LMFBR Transients," by J. V. Miller and R. D. Cof field, May 1976.
- 5. FRL-3259, " COBRA-WC: A Version of COBRA for Single-Phase Multi-Assembly Thermal-Hydraulic Transient Analysis," by T.
L. George, et al, July 1980.
- 6. WARD-NC-94000-2, " Comparison of DEMO Calculations to EBR-II Secondary Loop Natural Circulation Measurements," by R.
Schurko, et al, April 1980.
~
- 7. WARD-D-0281, "CRBRP Natural Circulation Verification Program Pretest Predictions for FFTF Natural Circulation Tests Initiated at 100% and 75% Power / Flow Conditions," by A.
Cheung, et al, February 1981.
- 8. SAI-101-7 8-PGH, "A Review of the Hydraulic Characteristics of the Cold Leg Check Valve for the Clinch River Breeder Reactor," July 197 8.
- 9. WARD-NC-94000-1, " Comparisons of DEMO Calculations to FFTF Pump and Flow Coastdown Measurements," by H. P. Planchon, et al, August 197 9
- 10. WARD-NC-3045-2,
- Loop Heat Capacity Models and Their Effects j for DEMO Natural Circulation Transient Analysis," by H. P.
Planchon and W. R. Laster, September 1978
- 11. CRBRP- ARD-013 2, "A Preliminary Evaluation of the CRBRP Natural Circulation Decay Heat Removal Capability," by R. R.
Lowrie and W. J. Severson, November 1977
- 12. .Y. S. Tang, R. D. Coffield, Jr. and R. A. Markley, Thermal Analysis of Liquid-Metal Fast Breeder Reactor, Aqf Monograph, pp. 148-149, 1.978.
- 13. E. U. Khan, "LMFBR In-Core Thermal-Hydraulics: The State of i
the Art and U.S. Research and Development Needs," PNL-3337, - pp. 13 8- 13 9, 1980. 41 n, o,. n - , - . . . - - -
- 14. R. M. Sing 2r cnd J. L. Gillotto, "M3ccurcmsnto of Subneccmbly cnd Coro Tcmpercturo Dictributicna in cn LMPBR,*
AIchE Symoosium Series,11, p.97-104,1977.
- 15. A. K. Agrawal, et al., " Dynamic Simulation of LMFBR Plant Under Natural Circulation," ASME Paper 79-HT-6,197 9.
- 16. M. Khatib-Rahbar and K. B. Cady, " Establishment of Buoyancy-Induced Natural Circulation in Loop-Type LMFBRs," Trans.
Amer. Nucl. Soc. 28, pp. 432-433, June 197 8. -
- 17. M. C. Billene, et. al., " LIFE-III Fuel Element Performance Code User's Manual," ERDA-77-56, July 1977.
- 18. WARD-D-0050, "CRBRP Assemblies Hot Channel Factors Preliminary Analysis," October 1974.
- 19. WARD-D-0016, "VARR-II, A Computer Program for Calculating Time Dependent Turbulent Fluid Flows with Slight Density
"'d = tion," by J. L. Cooke and P. I. Nakayama, July 1975.
- 20. D. Mohr and E. E. Feldman, "A Dynamic Simulation of the EBR-II Plant During Natural Convection with the NATDEMO Code," pg 207, Decay Heat Removal and Natural Circulation in Fast Breeder Reactors edited by A. K. Agrawal and J. G.-
Guppy (1981).
- 21. ANL-CT-77-7, "CRBR Outlet Plenum Mixing Studies: Supressor Plate and bhear Web Tests," by P. A. Howard and J. J.
Lorenz, November 1976.
- 22. ANL-CT-76-18, "An Investigation of'LMFBR Outlet Plenum l Thermal-Hydraulic Behavior During Reactor Scram Transients,"
by J. J. Lorenz, et al. , September 1975.
- 23. FFTP Outlet' Plenum Buoyancy Effects Study. J. A. Carr and L. J. Flanigan, Battelle Columbus Laboratory, March 1975.
- 24. Fission Product Decav Library of the Evaluated Nuclear Data j File, Version IV (ENDF/B-IV). Available from, and l maintained by, the National Nuclear Data Center (NNDC) at l the Brookhaven National Laboratory.
l l 25. T. R. England and R. E. Schenter, ENDF/B-IV Fission Product Files: Summary of Major Nuclide Data, LA-6116-MS [ENDF-223] , October 1975. - ! 26. M. E. Meek and B. F. Rider, Compilation of Fission Product l Yields, NEDO-12154-1, January 1974. g
- 27. C. W. Reich, R. G. Helmer, and M. H. Putnam, Radioactive-Nuclide Decay Data for ENDF/B, ANCR-1157 (ENDF-120) , August 1974. ,
42
- 28. D. R. Marr, "A Ucar'c ManuOl for Cccputor C:d3 RIBD-II, A Fic2 ion Product Inv;ntory Cada," HEDL-TME 75-26, Ennford Engineering Development Laboratory report (January 1975) .
- 29. F. Schmittroth, Nucl. Sci. Eno., 11, 117 (1913),
I
- 30. F. Schmittroth and R. E. Schenter, Nucl. Sci. Eng., El, 276 (1977).
- 31. T. R. England, M. G. Stamatelatos, R. E. Schenter, and F. C.
Schmittroth, Fission-Product Source Terms for Reactor Applications, Conference 770708, Proceedings of Topical ' Meeting on Thermal Reactor Safety, Sun Valley Idaho (July 31
- Aug . 4 , 197 7 ) . Also LA-NUREG-6 917-MS, August 1977.
- 32. R. E. Schenter, F. Schmittroth, and T. R. England, Inteoral Determination of Fission Product Inventory and Decay Power, IAEA 213 Vol. 2 Proceedings of the IAEA Second Advisory Group Meeting on Fission Product Nuclear Data, Petten, Netherlands, September 1977.
- 33. T. R. England, R. E. Schenter and E. Schmittroth, Inteoral Decav-Heat Measurements and Comparisons to ENDF/B-IV and V, LA-7422-MS, August 137 8.
- 34. J. K. Dickens, A. F. Emery, T. A. Love, J. W. McConnell, K.
J. Northcutt, R. W. Peelle, and H. Weaver, Fission-Product Fnergy Release for Times Followino Thermal-Neutron Fission
# 'Pu Between 2 and 14000 Seconds, ORNL/NUREG-34, April af 1978. -
- 35. J. L. Yarnell and P. J. Bendg3gCalorimetric
### U. and Fission U,
Product Decay Heat Measurements for Pu, NUREG/ CR-0349, LA-7452-MS, September 197 8.
- 36. WARD-NC-94000-3, Fluid Stratification in Piping: An l
Evaluation of its Effect on LMFBR Plant System Natural l Circulation Analyses, by A. C. Cheung, January, 1982. t e e 43
,}}