ML20087L189

Pre-Test Analysis for Panda Test M3

ML20087L189
Person / Time
Site:	05200004
Issue date:	08/31/1995
From:	Bandurski T, Fitch J, Healzer J GENERAL ELECTRIC CO.
To:
Shared Package
ML20087L192	List: ML20087L189
References
NUDOCS 9508250055
Download: ML20087L189 (40)

Text

-

r v

PRE-TEST ANALYSIS FOR PANDA TEST RU August,1995 T Bandurski(PSI)

JR Fitch(GENE)

JM Healzer(PSI)

J Morales (IIE)

M Stempniewicz(KEMA) j 3

6 9508250055 950821 PDR ADOCK 05200004 A

PDR J

m 4

.m a

D' 4:

TABLE OF CONTENTS Eage

1.0 INTRODUCTION

1 2.0 TEST FACILITY AND TEST MATRIX 2

3.0 APPLICABILITY OF DATA TO SBWR 4

4.0 TRACG MODEL AND NODALIZATION 5

5.0 TEST SIMULATION 25

. 6.0 RESULTS OF PRE-TEST CALCULATIONS 27 7.0 DISCUSSION 34

8.0 REFERENCES

i 35 I

W

O-b I

LIST OF TABLES I

East 4-1.

PANDA /FRACG VSSL01 Component Breakdown 10 4-2.

PANDAffRACG Components Connecting Two VSSL01 Cells 10 4-3.

PANDAfrRACG RPV Components 11 4-4.

' PANDAfrRACO PCCS Components 12 4-5.

PANDA /TRACG ICS Components 13 L

4-6.

Other PANDA /TRACG Components 13 4-7.

PANDA Heater Power vs. Time 14

+

l 4-8.

Comparison of PANDA and SBWR Component Nodalizations 15 f

5-1.

Initial Conditions for PANDA Test M3 26 i

a s

s il r

e a -

i LIST OF FIGURES EAEC 2-1.

_ PANDA Test Facility Schematic 3

i 4-1.

~ PANDA Vessel Component Nodalization Diagram 16 i

4-2.

RPV Component Nodalization Diagram 17 4-3.

RPV and Connected Piping Nodalization Diagram 18 4-4.

Components used for RPV Level Control during Startup 19 4-5.

PCCS Nodalization Diagram 20 4-6.

ICS Nodalization Diagram 21 4-7.

PCCS and ICS Pools Nodalization Diagram 22 4-8.

Main Vent Nodalization Diagram 23 4-9.

Vacuum Breaker Nodalization Diagram 24 6-1.

DW and WW Pressures for PANDA Test M3 28 6-2.

DW Temperatures for PANDA Test M3 29 6-3.

WW Temperatures for PANDA Test M3 30 6-4.

PCCS Inlet Flows for PANDA Test M3 31 6-5.

DW1 Air Mass Fraction in PCC1 Inlet Region 32 6-6.

PCCS Heat Removal for PANDA Test M3 33

)

i ik

l ACKNOWLEDGMENT i

i I

The authors of this report wish to acknowledge and thank Dr. Paul Coddington of the I

Paul Scherrer Institute for his significant contributions to the PANDA test and analysis program and, specifically, for developing the TRACG model of the PANDA test facility which forms the basis for this study. Dr. Coddington's early recognition of the importance of a strong analytical efront to accompany the test program and his perseverance in developing and verifying the TRACG model were crucial to the ability of the authors to complete this study prior to the initiation of PANDA transient testing.

i

+

1 i

i 5

I i

IV

1.0 INTRODUCTION

As part of the validation effort for the TRACG code for application to the SBWR, calculations will be performed for the various test facilities which are part of the SBWR design and technology certification program. These calculations include both pre and i

post-test calculations for tests in the PANDA test program. The analyses of the tests are being performed by an SBWR PANDA analysis team, with participation from PSI in Switzerland, where the tests are being performed, KEMA and ECN in the Netherlands, IIE in Mexico and GE in the USA. The PANDA TRACG input model described in this j

memorandum was developed as a joint effort of the team members, led by PSI. Pre and post-test calculations of the PANDA test results are a shared responsibility of the team members with final documentation and quality assurance records maintained by GE.

4 This report has two main purposes. The first is to provide a description of the PANDA TRACG input model. The second purpose is to present and discuss the results of a pre-test calculation, performed with the Level-2 version of the TRACG code, for PANDA matrix test M3. TRACG is being used for the calculation oflong-term SBWR containment response to a postulated LOCA. The PANDA test facility includes all of the features of the SBWR containment required for an integrated system simulation of long-term LOCA response. Most notably, the PANDA facility includes a detailed representation of the passive containment cooling system (PCCS) utilized for long-term decay heat removal. A satisfactory comparison between TRACG calculations of PANDA response to a simulated LOCA, and the test data, will provide strong evidence for the suitability of TRACG to predict containment response to a postulated LOCA in the 4

SBWR.

The overall PANDA TRACG qualification activity is summarized in Table A.3-13 of Reference 1. It includes " double-blind" pre-test calculations for matrix tests M2, M3, MS, f

l and M9. A double-blind calculation is defined as one in which neither the exact test conditions nor the test data are available. This report presents the results of the pre-test analysis for Test M3. As described in Paragraph A.3.1.3.4.1 of Reference 1, Test M3 is the " base case" of the PANDA test matrix. The initial conditions for Test M3 simulate I

the state of the SBWR at one hour from the initiation of a guillotine rupture of one of the main steam lines. Test M3 will be the first test in the PANDA sequence.

This report includes a description of the PANDA TRACG model. Subsequent pre-test reports (for Tests M2, M5, and M9) will describe modifications to the model relative to the M3 version. In addition to the modifications required to simulate specific tests (closure of one steam line for Test M2, introduction of DW spray for Test MS, etc.),

model changes are likely to incorporate the results of facility characterization tests (heat loss and pressure drop). It was not possible to incorporate the results of the facility characterization tests in the M3 analysis and still be certain of completing the analysis before Test M3 was run.

I i

The remainder of this report is organized as follows. Section 2.0 presents a brief description of the PANDA test facility and test matrix. Section 3.0 discusses applicability

^

of the data to the SBWR including key features of the scaling of the PANDA facility.

Section 4.0 describes the PANDA TRACG model and compares the model nodalization to that used in the SBWR TRACG model for long term, post-LOCA containment pressure and temperature analyses. Section 5.0 describes the TRACG simulation of the 1

steps employed to run Test M3 (e.g., the manner in which the initial heater power is established following the preconditioning of the PANDA vessels). Section 6.0 presents t

the results of the pre-test calculation and Section 7.0 discusses the results.

i 2.0 TEST FACILITY AND TEST MATRIX I

The PANDA Test Facility test matrix have been described in References 1 to 3. A summary description is included here in the interest of keeping the present document 1

reasonably self-contained. A schematic of the test facility is shown in Figure 2-1. The facility was designed to model the long term cooling phase of the loss-of-coolant accident 4

(LOCA) for the SBWR. It is a 1/25 volume-scaled, full-height simulation of the SBWR l

primary system and containment. Included in the facility are the major components necessary to model the SBWR system response during the long-term phase of the LOCA.

These components include the containment drywell (DW), the wetwell (WW) or suppression chamber, the reactor pressure vessel (RPV) including the core and those safety systems that would operate during the long-term phase of the LOCA. The RPV is represented by its own vessel in PANDA, while the DW and WW are represented by pairs of vessels, connected by large pipes. This double-vessel arrangement permits improved simulation of spatial distribution effects relative to a single vessel arrangement.

Important passive safety systems modeled in PANDA include the passive containment cooling system (PCCS), isolation condenser system (ICS) and the gravity driven cooling j

i system (GDCS). The PANDA PCCS is a direct representation of the SBWR PCCS with three separate loops, each containing a 1/25-scaled condenser unit. The PANDA ICS has l

one loop and condenser, scaled to represent two of the three ICS loops in the SBWR. The GDCS pool is represented by a separate vessel in PANDA, which communicates with the DWs through pressure equalization lines. Other SBWR components represented in i

PANDA include the vacuum breakers (VBs) between the DW and the WW and the i

equalization line (EQL) between the suppression pool and the RPV.

The PANDA test matrix includes steady-state tests of the performance of one of the PCC units and a series ofintegrated system tests intended to simulate the long-term cooling phase of the post-LOCA transient. As described in Reference 3, Test M3 is the base case for the transient series. The initial conditions for Test M3 are based on the calculated SBWR conditions at one hour from the occurrence of a guillotine rupture ofone of the main steam lines. Reference 3 provides a description of the full set of transient tests and how the variation of key parameters will be used to address specific thermal-hydraulic

{

phenomena considered to be of potential importance for prediction oflong-term post-LOCA behavior in the SBWR, i

d I

2 l

l l STEAM VENT IC PCC PCC PCC

~_"

.1.

2 1

PCC POOL 7

IC POOL T"=

=,:

==' t IC-DRAIN i

i PCC 3 VENT IC-SUPPLY l

1 I

i o

S Y

PRESSURE l[ f 2

EU 3

IC PCC UNE X:: HC:(W XX j:j:-

f

~

~

~

q:{

{

XX l

VENT VENT 1 SAFETY GDCS POOL VALVES GDCS DRAIN DRYWELL 1 DRYWELL 2 MA N MSL STEAM 2

RPV UNE1 VP VP BP VACUUM 8P yg J BREAKER J

J

,gDOWN-COMER SUPPRESSION SUPPRESSION CHAMBER 1 y

CHAMBER 2 y MAIN MAIN

/

VENT RISER /

GDCS

=_-

s'

~

SUPPRESSION l SUPPRESSION POOL 1 l

POOL 2 ELECT.

HEATER EQUAllZATION LINE h

f i

}

i t

i Figure 2-1. PANDA Test Facility Schematic 3

i

l 3.0 APPLICABILITY OF DATA TO SBWR j

l l

As stated in Reference 3, the objectives of the PANDA Test are: (1) provide additional data to qualify TRACG to predict quasi-steady PCCS heat rejection rate and to identify j

the effects of scale on PCC performance; (2) provide a data base to confirm the capability of TRACG to predict SBWR containment system performance (including potential i

systems interaction effects); and (3) demonstrate startup and long-term operation of the PCCS.

j i

The integral system response tests are focused on the objective ofinvestigating the highly ranked phenomena identified as " Qualification Needs" in Table 6.1-1 of Reference 3. The i

testing philosophy for the PANDA prograra is that this objective can best be accomplished by identifying a " Base Case Test" around which perturbations are made to i

assess the effects of specific systems, systems interactions, and phenomena ofinterest.

j The base case selected is a simulation of the long-term cooling phase following a LOCA caused by a guillotine rupture of one of the main steam lines. This LOCA scenario has been shown to lead to the highest long-term containment pressure in the SBWR.

t The long-term phase of the LOCA is defined as starting at one hour from the occurrence of the break. At this time, the effect of subcooling of RPV water by GDCS injection is just on the verge of being overcome by the decay power. When this occurs, the PCCS will be called upon to remove the energy added to the DW. To fulfill this role, the system

)

must first purge residual noncondensable gases from the DW to the WW. The performance of the PCCS under these conditions represents the single most important element of the PANDA test program.

Tests performed in the large-scale PANDA facility are expected to provide a set of data which will be directly applicable to qualification of TRACG for analysis of post-LOCA behavior of the SBWR containment. The facility has been carefully scaled to retain the i

important characteristics of the SBWR as described in References 4 and 5. A major feature of the facility scaling is the inclusion of an explicit representation of each of the l

three PCCS loops. Initial conditions for the tests in the PANDA matrix have been l

selected to directly correspond with post-LOCA conditions in the SBWR. The j

thermodynamic state of the SBWR RPV and containment at the time chosen for the i

initiation of the PANDA tests (one hour from the instant of LOCA andjust prior to the l

startup of the PCCS) is quasi-steady. This means that there is very little compromise made by setting up a steady state at these conditions in the PANDA facility prior to the initiation of the test. Finally, the instrumentation of the PANDA facility has been selected and located with the specific objective of providing the necessary data for the qualification of TRACG.

e 4

_=

b-e 1

5 4.0 TRACG MODEL AND NODALIZATION This section of the report provides a detailed description of the PANDA TRACG input i

model. The description includes a comparison with the nodalization used for the SBWR TRACG containment model. The DW, WW and GDCS vessels of the PANDA TRACG input model are represented by a three-dimensional "VSSL" component (see Reference 6 for a description of TRACG components). The RPV, the PCCS, the ICS and the piping l

which interconnects the PANDA vessels are represented by a combination of 85 one-l i

dimensional " PIPE", " TEE", "VLVE", "CHAN", "BREK" and " FILL" components.

Tables 4-1 through 4-6 list the components used in the PANDA input model, their connecting junction numbers, and a brief description of what they represent in the test j

facility.

i 4.1 Wetwell, Drywell and GDCS Pool 1

The PANDA WW, DW, and GDCS vessels are modeled with TRACG component l

VSSL01. The VSSL01 component has eleven levels, two rings and five azimuthal sectors l

(Figure 4-1). In each level, the individual " cells" are numbered as shown. The first seven l

levels of the VSSL01 component are used to model the two PANDA WWs. Levels 1 through 3 represent the WW pools and the remaining four levels represent the vapor l

space above the pools. Cells 1,2,6 and 7 of these seven levels represent WWI and cells 3,4,8 and 9 represent WW2. In the first seven levels, cells 5 and 10 are blocked off from the neighboring cells and not used in the model. Level 8 represents the space between the j

WW and DW vessels and is modeled as an adiabatic interface. Levels 9 through 11 3

represent the DW and GDCS vessels. The cells which represent WWI and WW2 in levels I to 7 now represent DW1 and DW2 and cells 5 and 10 represent the GDCS pool.

1 Pipe components (PIPEI1,12 and 13) are used to model the connections between WW1 and WW2 and between DW1 and DW2. The pressure equalization lines between the DWs and the GDCS vessel are modeled by "VLVE" components (VLVE43 and 44). To represent containment spray, FILL components (FILL 47,48,56 and 57) are connected to j

the uppermost cells in DW2. With this arrangement, the two WW pools are represented i

by twelve cells each, the WW vapor spaces by sixteen cells each, the DWs by twelve cells each and the GDCS pool by six cells. If the data from the PANDA tests indicate i

the vessels are not well mixed and there is significant stratification or other spatial distribution effects, the nodalization of any of the PANDA vessels can be increased by l

addition oflevels and rings to the VSSL01 component.

4.2 RPV and Associated Piping The RPV in PANDA is modeled with a combination of TEE components, and the electric j

heaters are modeled with a "CHAN" component. Figure 4-2 shows the nodalization j

scheme used for the RPV. TEES are used for the lower plenum (TEE 60), the downcomer(TEE 59 and 61), the chimney (TEE 62) and the upper plenum (TEE 63 and 66) 5 I

1

l regions of the vessel. A total of eleven cells model the chimney above the core and another ten cells model the region containing the two-phase level and vapor space above l

the top of the core shroud. A CHAN component (CHAN08) is used to model the electric j

heaters in the vessel with two heated cells. The two-cell heater simulation is expected to be adequate because, under anticipated test conditions, there is very little boiling in the i

core and much of the steam is produced by flashing in the chimney and in the two-phase i

mixture region above the shroud.

Figure 4-3 shows the RPV together with its connecting components. These include the l

TEE and VLVE components which model the steam lines to the two drywells (TEE 67, VLVE68 and 64), the VLVE components representing the drain line from the GDCS pool (VLVE45 and 46), the TEE and VLVE components (VLVE65 and TEE 69) which represent the EQL between the RPV and the two WW pools, ud the VLVE components representing the ICS inlet and drain lines (VLVE24 and SI). (For completeness, the full ICS is shown, including the condenser unit (PIPE 94 and 98 and TEE 28) and the vent line (VLVE54). The lower header (TEE 28) in the figure has been displaced from the condenser unit for clarity.)

4 i

The RPV model also includes a number of components specifically for the purpose of i

simulating its approach to steady-state conditions prior to the initiation of transient testing. These components include BREK50 and VLVE70 at the top of the vesse!

(pressure control), and TEE 03, VLVE02, VLVE05, FILL 04, and BREK06 which interconnect with the EQL to the RPV via TEE 74 (level control). Figure 4-4 shows a detailed blowup of the level control components. The functioning of the components used

}

for the approach to steady-state is described in Section 5.

i 1

{

4.3 PCC and IC Condensers and their Pools i

2 The TRACG component arrangement used to model one of the PCCS loops is shown in Figure 4-5 and a stand-alone representation of the model of the ICS loop is shown in Figure 4-6. In the PANDA input model, each of the three PCCS loops has a component

)

arrangement similar to that shown in Figure 4-5. In the heat exchanger section, a single PIPE component is used to model all of the tubes in each heat exchanger. Ten axial cells are used to model spatial distribution effects along the length of the condenser tubes. The inlet and outlet headers are modeled with a PIPE and TEE component, respectively, and VLVE components are used to model the inlet, vent and drain lines. The pools are modeled with TEE, PIPE and BREK components as shown in Figure 4-7. Heat is transferred from the PIPE component representing the heat exchanger tubes to the PIPE component representing the part of the pool adjacent to the tubes. The pool water boils, inducing natural circulation in the pool, with up-flow in the PIPE receiving heat and down-flow in the TEE representing the balance of the pool. The steam formed from i

boiling in the pool is vented to the BREK component at the top of the pool which is set to maintain constant pressure at the pool surface.

6

l 4.4 Main Vents and Vacuum Breakers l

The main LOCA vents in the PANDA input model (Figure 4-8) are represented with VLVE components (VLVE15 and 16), one connecting DW1 with the pool in WWI and the second connecting DW2 with the pool in WW2. Flow through the LOCA vents will l

]

occur when the DW pressure exceeds the WW pressure by more than the hydrostatic head from the surface of the WW pool to the terminus of the vent line within the pool.

The two PANDA VB lines are modeled with a combination of TEE and VLVE

]

components as shown in Figure 4-9. There is one VB line connecting the WWI vapor space to DW1 and a second VB line connecting the WW2 vapor space to DW2. The VBs open and reclose at WW to DW pressure differences of 3040 Pa and 2060 Pa, respectively. As shown in Figure 4-9, the PANDA VB lines include valved bypass lines (VLVE19 for VB2) to simulate leakage between the SBWR DW and WW vapor space.

4.5 System Line Flow Resistance 4

The TRACO code allows user input oflocal loss coeflicients, but calculates its own losses for piping friction based on user input roughness-to-diameter ratio. The flow path l

resistances used in the PANDA input model are based on handbook values for the loss i

coeflicients of various fittings and piping components. A roughness to diameter ratio of 0.005 was found to cover most piping sizes in PANDA and this value has been used for all piping. This results in fully developed turbulence for the flows expected in most flow j

l paths, giving a friction factor of about 0.03. A detailed summary of the flow path losses j

and comparison to similar flow paths in the SBWR can be found in Reference 5.

The pressure drop modeling for most flow paths in PANDA will be confirmed by tests to l

characterize the system behavior. These characterization tests will provide data to confirm or adjust the flow path loss coeflicient in the PANDA TRACG input model. The revised flow path modeling will be available for post-test predictions of the PANDA data and for some of the later pre-test calculations. Comparison of the measured and predicted flow path pressure drops and any recomrrendations to modify the flow path loss coeflicient will be reported separately.

4.6 Component Heat Loss and Heat Capacity The heat losses from the PANDA vessels and piping to the environment or to other components have been included in the PANDA input model. Component-to-component heat transfer is accounted for between the components used to represent the PCC and IC and other containment components. As noted earlier, heat transfer from the PCC and IC tubes, upper and lower headers and inlet steam lines to the condenser coolant pools is modeled. Heat transfer from the PCC drain lines to the fluid in the GDCS vessel is modeled. Heat transfer from the PCC and IC vents to the fluid in the WWs is modeled.

Heat losses to the environment from all PANDA facility vessels (DWs, WWs, GDCS tank and RPV) and from lines connecting the vessels are also modeled.

7

[

+

In general, heat transfer between cells in different TRACO components is through one dimensional heat slabs. The heat transfer is calculated based on heat transfer correlations built into the code and selected based on conditions in the cell receiving or losing the heat. 'Ihe user can control the heat slab thickness and heat transfer area. The user can l

also elect to input the liquid and vapor heat transfer coefficients and sink temperatures when the heat is transferred from the fluid in all cells of a component to a constant temperature sink, such as the test facility outside environment. For most vessel heat j

losses, the double sided heat slab model in TRACG is used to model the metal wall of the l

vessel. The insulation and outside film resistance are combined into an overall outside heat transfer coefficient and the outside ambient temperature is specified. The same is l

done for other components losing heat to the environment such as the main vent lines, the

~

steam lines, V/B lines, the EQLs and the GDCS drain line.

System characterization heat loss tests will be conducted to better define heat losses from the PANDA vessels (the major source of facility heat losses). The results of these system characterization tests will be used to either confirm or modify the simulation of heat loss in the PANDA TRACG input model. The revised heat loss modeling will be available for post-test calculations of the PANDA data and for some of the later pre-test calculations.

Comparison of the measured and predicted system heat losses and any recommendations to modify the heat loss modeling will be reported separately.

4.7 Decay Heat i

To represent decay heat in the PANDA TRACG input model, the SBWR decay heat design specification (Reference 7) was v. sed. The magnitude of the decay heat was scaled by the PANDA volume factor of 0.04. The decay heat is modified by a time-dependent multiplier to account for the sensible energy released by structural components in the SBWR RPV which are not included in the PANDA facility. This multiplier was based on TRACG calculations using the SBWR containment model, which includes a representation of the RPV structural components. The enrgy released by these components was separately edited and converted to a time-dependent multiplier on the decay power. Over the course of a PANDA test, the multiplier vatics from 1.07 to 1.007.

The modified decay heat input is entered in tabular form in the PANDA input model (see Table 4-7).

4.8 Comparison with SBWR Containment Model The component arrangement in the PANDA input model is similar to the TRACG input model for the SBWR. The major difTerence between the SBWR and the PANDA TRACG input models is that the SBWR model uses the VSSL component to represent the RPV and the PCCS and ICS pools. These are modeled with a combination of one-dimensional PIPES, TEES and BREKs in the PANDA input model. Since PANDA does not simulate the blowdown phase of the LOCA, a simpler representation of the RPV is justified. The use of one-dimensional components to represent the PANDA PCC pools, as j

l 8

l compared with the use of the VSSL component in the SBWR TRACG model, is not regarded as a significant distinction. Both models amount to the use of a single column of l

cells to represent the heated region (the " riser") adjacent to the condenser tubes and a i

parallel column of cells (the "downcomer") to represent the balance of the pools. The use of a VSSL component has the advantage of allowing intermediate crossflow between the two major flowpaths but this should be relatively unimportant in PANDA where the i

pools are maintained in a saturated state throughout the tests. The crossflow branch at the i

top of the PANDA model has been located such that it will remain below the pool c el j

over the duration of the test simulation.

i l

A comparison of the SBWR and PANDA component nodalizations is given in Table 4-8.

I Differences in nodalization result partly from the histccical origins of the two models and j

pantly from differences between the PANDA as-built layout and the SBWR conceptual j

layout. As an example of the latter, the VB lines in PANDA must physically connect individual DW and WW vessels whereas, in SBWR, the VBs are located in the diaphragm floor separating the DW and WW regions of the containment. On the whole, the nodalizations of the one-dimensional components are comparable and would not be expected to lead to better predictions with the PANDA model than are possible with the SBWR model.

Significant differences in the two nodalizations appear in the first four rows of Table 4-8, j

covering the RPV, fuel and heaters, upper DW, and WW. As described above, the i

PANDA RPV is constituted of one-dimensional components in contrast to the use of the VSSL component for the SBWR. 'Ihe use of one-dimensional components requires more cells to accommodate the multiple RPV piping connections. The fuel in the SBWR model has a more refined nodalization than the PANDA heaters because of the need to calculate the axial void fraction profile during full-power operation prior to the LOCA simulation.

l The axial and radial nodalization of each of the PANDA DW and WW vessels is i

comparable to SBWR. The total number of cells for the PANDA vessels is significantly larger than for SBWR because PANDA has two DWs and two WWs, each of which has two azimuthal sectors. The more extensive DW and WW nodalization in PANDA is a j

direct consequence of the fact that one of the major objectives of the PANDA facility and test program is to evaluate the significance of asymmetric behavior relative to the prediction oflong-term containment pressure and temperature. An underlying assumption of the SBWR model is that asymmetric effects are not important.

i 4

6 9

J Component Levels Cells Description VSSL01 1 to 7 1,2,6,7 WWI VSSL01 1 to 7 3,4,8,9 WW2 l

VSSL01 9 to 11 1,2,6,7 DW1 VSSL01 9 to 11 3,4,8,9 DW2 VSSL01 9 to 11 5,10 GDCS pool 4

Table 4-1. PANDAfrRACG VSSL01 Component Breakdown i

1 i

Component J1 J2 Description FILL 47 901 DW1 spray (L11, Cl)/spili(L9, C7) i FILL 48 501 DW1 spray (Li1, C7)

FILL 56 902 DW1 spray (L11, C6)

FILL 57 903 DW1 spray (L11, C2) 101 201 lower WWl/WW2 connection PIPEI1

, 102 202 upper WWl/WW2 connection PIPE 12 PIPE 13 103 203 DWl/DW2 connection VLVE 15 111 110 DW1 to WW1 vent line VLVE16 211 210 DW2 to WW2 vent line VLVE43 141 142 DWl/GDCS pressure equalization VLVE44 241 242 DW2/GDCS pressure equalization Table 4-L "ANDAfrRACG Components Connecting Two VSSL01 Cells i

10

l 2

I J

P Component J1 J2 J3 Description BREK%

373 sink for FILLO4 following startup BREK50 999 pressure control during startup CHAN08 161 162 169 electric heaters FILLO4 371 level control during startup FILLO9 169 CHAN08 bypass leakage (= 0)

TEE 03 371 370 372 allows switching FILLO4 from RPV i

to BREK06 following startup TEE 59 159 160 175 lower downcomer TEE 60 160 161 166 lower plenum TEE 61 164 159 168 upper downcomer l

TEE 62 162 163 164 chimney above core TEE 63 165 998 150 steam dome TEE 66 163 165 155 two-phase mixture region above core shroud TEE 74 166 369 368 RPV connection to equalization line and "startup" components VLVE02 368 370 turns offlevel control liquid flow after startup period VLVE05 372 373 opens flow path to BREK06 i

following startup VLVE70 998 999 tums off steam flow from RPV to atmosphere following startup period 4

Table 4-3. PANDAfrRACG RPV Components d

i 4

11

a Component J1 J2 J3 Description BREK33 51 air space pressure control for PCCI pool BREK37 61 air space pressure control for PCC2 pool BREK83 71 air space pressure control for PCC3 pool PIPE 31 53 54 PCC1 pool riser PIPE 35 63 64 PCC2 pool riser J

PIPE 81 73 74 PCC3 pool riser PIPE 91 321 322 PCC3 upper header PIPE 92 121 122 PCCI upper header PIPE 93 221 222 PCC2 upper header PIPE 95 322 323 PCC3 condenser tubes PIPE 96 122 123 PCC1 condenser tubes PIPE 97 222 223 PCC2 condenser tubes TEE 25 323 324 325 PCC31ower header TEE 26 123 124 125 PCCl lower header TEE 27 223 224 225 PCC2 lower header TEE 30 51 52 53 PCCl pool airspace and upper section of downcomer TEE 32 52 55 54 PCCl pool lower section of downcomer TEE 34 61 62 63 i PCC2 pool air space and upper section of down comer TEE 36 62 65 64 PCC2 pool lower section of downcomer TEE 76 90 91 55 PCCI pool interconnection TEE 77 91 92 65 PCC2 pool interconnection TEE 78 92 85 75 PCC3 poolinterconnection l

TEE 80 71 72 73 PCC3 pool airspace and upper section of downcomer TEE 82 72 75 74 PCC3 pool lower section of downcomer j

VLVE21 320 321 PCC3 inlet line VLVE22 120 121 PCCI inlet line VLVE23 220 221 PCC2 inlet line VLVE52 125 126 PCCI vent line VLVE53 225 226 PCC2 vent line i

VLVE55 325 326 PCC3 vent line i

VLVE71 124 127 PCCl drain line VLVE72 224 227 PCC2 drain line VLVE73 324 327 PCC3 drain line Table 4-4. PANDA /TRACG PCCS Components 12

d l

Component J1 J2 J3 Description 1

BREK87 81 air space pressure control for IC pool 1

PIPE 85 83 84 IC pool riser PIPE 94 421 422 IC upper header PIPE 98 422 423 IC condenser tubes TEE 28 423 424 425 IC lower header TEE 84 81 82 83 IC pool airspace and upper section of downcomer TEE 86 82 85 84 IC poollower section of downcomer VLVE24 421 155 IC inlet line VLVE51 424 168 IC drain line VLVE54 425 426 IC vent line Table 4-5. PANDAfrRACG ICS Components i

Component JI J2 J3 Description TEE 17 231 233 281 DW2 vacuum breaker connection TEE 20 234 232 282 WW2 vacuum breaker connection TEE 38 131 133 181 DW1 vacuum breaker connection TEE 41 134 132 182 WW1 vacuum breaker connection TEE 69 267 367 167 WWI and WW2 equalization line connections TEE 67 151 152 150 connects steam lines to RPV VLVE64 151 157 steam line to DW1 VLVE68 152 158 steam line to DW2 VLVE18 233 234 VB2 valve VLVE19 281 282 VB2 bypass valve VLVE39 133 134 VB1 valve VLVE40 181 182 VB1 bypass valve VLVE45 172 173 upper portion of GDCS drain line VLVE46 173 175 lower portion of GDCS drain line VLVE65 367 369 RPV equalization line connection Table 4-6. Other PANDA /TRACG Components i

13

l*

i 1

Time (sec)

Heater Power (MW) 0 0.0 1

1 1.130 500 1.130 550 1.120 900 1.082 1900 1.007 2900 0.9392 3900 0.8911 4100 0.8817 4400 0.8709 4900 0.8527 5900 0.8238

~

6900 0.7971 8900 0.7721 s

11300 0.7550 11900 0.7480 14900 0.7134 16900 0.6941 21900 0.6602 25700 0.6350 26900 0.6292 1

32900 0.6026 36900 0.5873 46900 0.5551 56900 0.5301 I

66900 0.5084 76900 0.4906 83300 0.4802 86900 0.4745 96900 0.4608 116900 0.4367 Table 4-7. PANDA Hester Power vs. Time 14 l

Component No. of Cells PANDA Model SBWR Model RPV 36(1) 16 Fuel / Heaters 2

8 i

Upper DW 24 10(3)

WW 56 8

GDCS tank 6

4 PCC steam line 9

5 PCC upper header 2

3 PCC tubes 10 8

PCC lower header 2

3 PCC drain line 10 8

PCC vent line 13 6

IC steam line 12 12 IC upper header 2

3 IC tubes 10 8

i IC lower header 2

3 IC drain line 19 15 IC vent line 14 10 Main steam lines 9

2,11(4)

Main LOCA vents 18 18 (5)

GDCS drain line 17 10 Equalization lines 14,16 (2) 4 V/B lines 17 4

Table 4-8. Comparison of PANDA and SBWR Component Nodalizations 8

Notes:

(1)

The PANDA RPV model (apart from components included specifically for the j

simulation of the startup transient) encompasses TEE 59 (2 cells), TEE 60 (7 cells),

TEE 61 (6 cells), TEE 62 (11 cells), TEE 63 (5 cells), and TEE 66 (5 cells).

(2)

The two equalization lines in the PANDA model have slightly different nodalizations.

(3) including DW head.

)

(4)

Broken MSL (2 cells); intact MSL to MSIV (11 cells).

l i

(5) including three rows of horizontal vents.

15

DRYWELLS AND GDCS 5

LEVELS 9 - 11 J 203 DW1 - DW2 CONNECTION MAIN VENT 1 MAIN VENT 2 J 232 VB2 J 103

'b.)

J 111 J 220

, ' ~ -

7 hi20 8

INLET LINE TO PCC2 t

s DW2

/

DW1 s

I

\\,-

,[/

,,9,,

J 132 N

i J 320 s

)

VB1 INLET LINE TO PCC3 l

1:

\\

4 1

o' 9 \\

/6 J 1$7

\\

5

/

MAIN STEAM LINE 1

'p MAIN STEAM LINE 2 o

j 343 J1 J4 DW1/GDCS 10 DW MDC GDCS GDCS DRAIN PCCS DRAINS J 172 J 127 J 227 J 327 J = JUNCTION NUMBER WETWELLS LEVELS 1 7 UQUID J 201

~

J 101 J 202 WW1 WW2 VAPOUR VB2

~"~

J 102 J 131 i

-t

, J.TO, MAIN VENT 1 J 226 7

8 PCC2 VENT

/

'N J 101_

/

3 2

BC VENT ww

' -/-

WWi

", \\

p J 126 l

J 236 4

y PCC1 VENT J 131 PCC3 VENT - ""'"~~

i 4

3 9 \\

/6 VB1

\\

/

g MAIN VENT 2 J 210 Figure 4-1.

PANDA Vessel Component Nodalization Diagram 16

O S

i 999 VLVE7gpBREK50 9 94**~

4 s

3 TSEtp al rop;;;

-r 1

S

.IEFils 2

.,ahr 4

j y

-El TEE 61 4

--1e fse TEE 62_

159 i

8 4EE59 2 175 s 11lC 1

s-

---.s IG]5-CHANdT FILL 9 g

1

=3 i

TEE R 1

i s

4 i

a Figure 4-2.

RPV Component Nodalization Diagram 17

O G2

[A IPE94 422f

[ PIPE 98 5

42.)

\\"**

VLVE24

, 424 f a

/

2 VLVE70J _,REK50 TEE 28 4

9 4

3 TEE 63 152,==_,

s 5

a

-51MEE67 wtan=*.

_. 155 22.i u io T

yLVE68_15e 3

TEE 6B 3=

_. m -.___a=_m157 a

VLVE64 y,

2 72 VLVE54

'and to: CQ io -

---4: TEE 61

/LVE51 e

si VLVE45 4

e

__:'~

e

'8 r

173 is 12 i

saa....

TEE 62__, _,3, ir l

g 1

m in t4

~EJ ES9 /LVE46 VLVE5

-_--I 3

n i

REK6 426

_I i*

VLV FILL 4 2-Ld TE

~[

_ TEE 74 388 VLVE65

~

3 77 g67

_ TEEM i

1-

^ - - - ""

1 1e} h ILL9 TEE 69 i in -

e 2

4 Jill Figure 4-3.

RPV and Connected Piping Nodalization Diagram 18 i

4 4

i 4

k i

4 1

4 i

4 4

{

1EE M -

EW5,373 r~0 a

a F '- w i i

.7 i

I M- )

4 i

a 4;

8 6

e WW_2 g 4 3P1 0

l

-r-i i

FN Itt3 Rf4 i

1 l

1 1

k 1

1 1

I A

i

]

Figure 4-4.

Components used for RPV Level Control during Startup 4

i I

r 3

4 6

19

1 4

1 n.

-7

f 3

i 4

_}ts PIPE 92

_f,122 3

.VLVE22 PIPE 96 4

..'B sd.'MS, I

1

120 i

li i

~

~

i g7 l

i t

i

VLVES2

\\

r l

VLVE71 l;l j

\\

I d

i I

l l

\\

i l

1 I

i l

\\

i l

I I

I

t2 h8 s2e Figure 4-5.

PCCS Nodalization Diagram 20

=

_.. ~. _..

Po I$

b

,,m ys

+

l i

a PIPE 94.

4 2

m de i

PIPE 98 f

2]

t J

l f) t TEE 28

' ^ 7s VLVE24

,a,

)

42[lt 3

3 a

I i

1' l

r

l

\\

'l 4

i i

4 0

l 4

1 i,

1 h

7 I

1g r

w_..__.

. __ g15s t

1*

l e

,s.-

i r VLVE54 1e e

10 l

l\\o VLVE51 1

1 is

{

5,,

1 l

1 1

llt l

l l

ll%

12 i

tesg.

l 1,1 l

l

s sn ng 13 14 426 1

Figure 4-6.

ICS Nodalization Diagram.

21

r PCCS-1 PCCS-2 PCCS-3 C

6 BREK33 BREK37 BREK83 BREK87 r

/

/

/

/

l l

e i

i.

l TEE 30 8 i

g

?

"l

1... s 1...

s 1-e e

y y

y fkE81 l*

ls PPE31 i_PtPE36 m

-.PPE85

.,g'

?

-r

-7 F

r-Trraa l'

~"

Trras :'

IEEsa j' u

i TEE 32fa

[,,

s I

e 3

_...___..___:...______-.:-.. __ m -,m_

l TEE 76 TEE 77 TEE 70 1

Figure 4-7.

PCCS and ICS Pools Nodalization Diagram 22

>+..a.+.

.v 211 e.2.F ;$

t t

4 1

t i

I

'1 1

i t

t t

I B

4 t

2... +VLV. E.18. -.... +.

+

i n

I

+

i a

lI I

t I

l=

I t

i 1B t

t

.i

?

I T

lI a

3B

1. 210 Figure 4-8, Main Vent Nodalization Diagram 23

VLVE19

,p - - - - -. :.,[I I

i s

1 I

b I,il

,1 su m1 1 z= -.... -y - - -.. y. s - + ' y ' + n a v,8 VLVE18

+

i I

I I

l4 f

l TEE 17

+

t p a +..... _ _ o' Mf a

i 8

TEE 20 I

I i

1

a I

I i

1

+

t 4

i, l

I

+

i 2n 4.... + s p..so. i Figure 4-9.

Vacuum Breaker Nodalization Diagram i

i 24 i

b 5.0 TEST SIMULATION In addition to the PANDA facility geometry, the initial thermodynamic conditions in the system prior to initiation of the transient tests must be part of the PANDA /TRACG input model. For the long-term cooling phase of the LOCA transient, the PANDA tests begin approximately one hour after accident initiation. Conditions at this time in the transient were derived from SBWR TRACG calculations for the case of a guillotine rupture of one of the main steam lines. The results of the SBWR calculation at one hour from the time of the LOCA are given in Table A.3-11 of Reference 3. The initial conditions for PANDA Test M3, derived from the SBWR results, are given in Table A.3-10 of Reference 3, reproduced here as Table 5-1. The SBWR calculation shows that thermodynamic conditions throughout the system are relatively stable at this time in the LOCA transient.

The RPV blowdown is complete, the GDCS pools have dumped their inventory to the RPV, the decay heat has overcome the subcooling introduced by the injection of GDCS water, and steaming due to boil-off has resumed. The pressure difference between the RPV and DW is just sufficient to push the boil-off steam through the flow path provided by the break and the open depressurization valves. The pressure difference between the DW and WW is just sufficient to clear the PCC vents. From the Test M3 initial conditions shown in Table 5-1, pressures, temperatures, and liquid, vapor and air inventories have been defined for each of the components in the PANDA /TRACG input model.

The detailed procedures for initializing and initiating the PANDA transient tests had not been fully defined at the time of preparation of this document. The procedures employed in the TRACG simulation represent the authors' understanding, based on discussions with PSI testing personnel. A key feature of the initialization is the procedure used to establish steady steam flow conditions in the RPV. The calculation is initiated with the RPV saturated at its prescribed initial pressure and isolated from the remaining system components. The heaters (CHAN08) are ramped up to the prescribed initial power while the BREK50 component at the top of the steam dome (acting via VLVE70) is used to maintain pressure (see Figure 4-2). The FILL 04 component (acting via VLVE02), in conjunction with TRACG control system logic, maintains the RPV collapsed level. This set of components is connected to the RPV via TEE 74 as shown in Figure 4-3 and, in a detailed view, in Figure 4-4. The calculation is run for 500 seconds with the model in this configuration to ensure steady-state conditions in the RPV. The transient calculation is initiated by opening the valves in the main steam lines (VLVE64 and 68) while simultaneously closing VLVE70 and VLVE02. Additional components (VLVE05 and BREK06) were included to provide a sink for the flow from FILL 04 after VLVE02 is closed (Figure 4-4).

The PCC units are initialized with the tubes and headers filled with a saturated air-steam mixture at the pool temperature (373 K). This is comparable to the situation in the SBWR where, during the period leading up to one hour from the occurrence of the LOCA, the RPV steam production is almost totally suppressed by the subcooled inventory ofinjected 25

i GDCS water. During this period, the condenser temperatures drop towards the pool temperature and the tubes and headers fill up with a higher fraction of noncondensable gas than the DW as a whole. In PANDA, the main LOCA vent, the PCC vents and the GDCS drain line are all equipped with valves. In the TRACG simulation of Test M3, these valves are opened concurrently with the opening of the valves in the main steam lines. This was done on the assumption that the test procedures will require isolation between the DW and WW vessels and between the GDCS vessel and the RPV during the establishment of steady-state conditions in the RPV.

For subsequent PANDA /fRACG pre-test calculations, adjustments will be made in the simulation of transient test initiation as the test procedures become better defined. Any remaining discrepancies between the pre-test simulations and the manner in which the tests are actually run will be taken into account during the post-test evaluation.

RPV Drywell Wetwell GDCS PCC Pools Total Pressure (kPa) 295 294 285 294 101 Air / Nit. Pressure (kPa) 0 13 240 274 n/a Vapor Temperature (K) 406 405 352 333 n/a Liquid Temperature (K) 406 405 352 333 373 Collapsed WL (m)(1) 11.2 (2)

(3) 3.8 10.7 (2) 23.2 i

Table 5-1.

Initial Conditions for PANDA Test M3 Notes:

(1)

Water levels (WL) are specified relative to the top of the PANDA heater bundle.

(2)

PANDA RPV and GDCS water levels modified from SBWR one-hour values to minimize GDCS inflow and spill from the RPV at the start of the test.

(3)

The nominal DW condition is no water. However, a small amount of spill from the RPV to the DW at the start of the test is acceptable.

26

6.0 RESULTS OF PRE-TEST CALCULATIONS i

4 The results of the pre-test calculation for PANDA matrix test M3 are displayed in Figures 6-1 through 6-6. Figure 6-1 shows the DW and WW pressures. Both pressures initially rise as the initial inventory of DW air is purged to the WW via the PCCS. Both pressures subsequently decrease with the DW pressure decreasing at a slightly higher rate than the j

WW pressure. The DW pressure drops below the WW pressure at about 25,000 seconds j

(approximately 7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br />), and continues to decrease until the VBs open at 35,000 seconds (approximately 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />). The DW pressure then rises and remains above the WW pressure by the amount required to overcome the PCCS vent line submergence for the duration of the transient.

Figure 6-2 shows the temperatures in the two DW vessels. The temperatures in the two DWs vary in a nearly identical manner, rising during periods of PCCS purging and slowly decreasing when the PCCS is matched to decay power. Figure 6-3 shows the vapor temperatures in the two WW vessels. Initially, the WW vapor temperatures rise in response to the inflow of a hot steam / air mixture from the DW during PCCS purging. The WW temperatures subsequently decrease as energy is lost to the walls of the surrounding vessel. The WW2 temperature rises to a slightly higher value during PCCS purging. The l

WW1 temperature shows a small rate of decrease following the VB opening whereas the WW2 temperature remains essentially constant.

Figure 6-4 shows the inlet flows to the three PCC units. The flows oscillate during the purging period, then settle down to a gradual variation following the slowly decreasing i

decay power. The flows to the two PCC units attached to DW2 are nearly identical. At about 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> from the start of the transient, the flow to PCCI deviates from the flows to the other two units and starts decreasing at a noticeably slower rate. The PCCI inlet flow remains higher than the other two inlet flows until the VB opens. Immediately following the VB opening, the flows to all three units equalize and remain balanced for the duration of the transient.

Figure 6-5 shows the air mass fraction in DW1 in the region which feeds PCC1. The initial air inventory is purged to the WW in slightly less than two hours from the start of the test simulation. The air buildup following the opening of the VB is quite small in comparison to the initial inventory. Finally, Figure 6-6 shows total PCCS heat removal rate in comparison to the decay power. It can be seen that the condensers match decay power at the same time as the completion of the air purge from the DW. The degradation of condenser heat removal rate following the opening of the VB is clearly evident.

27 i

l 4

400 DW 350 r

V' l

300 WW 250 200<

4

.150-100-0 5

10 15 20 25 30 35 40 45 50 55 60 Time (IE3 sec)

Figure 6-1.

DW and WW Pressures for PANDA Test M3 (Cell 1 of Levels 7 and 11 in the TRACG model - see Figure 4-1)

F 4

28'

420 418 416

'414

- 412 e

410-400-

>=

404 402 -

400 0

5 10 15 20 25 30 35 40 45 50 55 60 Time (1E3 esc)

Figure 6-2.

DW Temperatures for PANDA Test M3 (Cells 1 and 3 of Level 11 in the TRACG model - see Figure 4-1) 1 29

)

\\

h

)

r l

380 1

WW2 358.

7 356 -

1 354 4

I 352 i

350 34e.

WWT P-348.

1 344.

j 342 J

340 0

5 10 15 20 25 30 35 40 45 50 55 60 Time (1E3 sec)

I

\\

1 i

)

i i

Figure 6-3.

WW Temperatures for PANDA Test M3 (Cells 1 and 3 of Level 7 in the TRACG model-see Figure 4-1) 4

)

4 30 i

~..

r i

i l

2.50601 r

j I:

2.00G01 I

1.50G01 l

a

-! i 1.00E01 I

1'7-97 5 00E02 -

4 0.00E4 ' I O

b 10 15 20 25 30 35 40 45 50 55 to Time (1E3 sec)

Figure 6-4.

PCCS Inlet Flows for PANDA Test M3 (e.g., Entrance of Cell 9 of VLVE22 in the TRACG model - see Figure 4-5) i 1

31

i 4

l i

0.06 0.Os.

i

• 0.04.

a 0.02 k-Om l

O 5

10 15 20 25 30 35 40 45 50 55 80 Time (1E3 sec) i 4

Figure 6-5.

DW1 Air Mass Fraction in PCCI Inlet Region (Cell 2 of Level 11 in the TRACG model - see Figure 4-1) i n

l i

I 32 i

.-m

- m.

1 o

I l

6 1400 1200 -

l heater power i

/

~

1000 4

g l f "" '

[

i 600 f

400-I 200 -

0-0 5

10 15 20 25 30 35 40 45

$0 55 60 Time (1E3 esc) l i

i 1

Figure 6-6.

PCCS Heat Removal for PANDA Test M3 1

33

l l

t h

i I

j 7.0 DISCUSSION i

The transient sequence depicted in Figures 6-1 through 6-6 can be described as follows.

At the start of the transient, steam flows from the RPV to the two DWs and mixes with the air in the DWs. The mixture flows into the three PCCS units. Some of the steam is 1

j condensed in the heat exchanger units while the remainder, along with the air from the DWs, flows to the WWs. The transfer of steam and air at DW temperature to the WW l

increases the WW pressure and temperature. The DW pressure rises correspondingly to keep the PCCS vent lines clear and sustain flow through the condenser units. The DW g

j and WW pressures reach maximum values of 327 kPa and 315 kPa, respectively, somewhere between 5000 and 6000 seconds from the initiation of the transient. The pressure peaks are not sharply defined but they occur at just about the time the air has i

j been purged from the DWs (Figure 6-4). The WW temperature then decreases by several degrees as energy is transferred from the fluid to the walls of the WW vessels. With the DWs essentially purged of air and the heater power decreasing with time, the combination of PCCS energy removal and heat transfer to the walls of the DW vessels is l

able to remove more energy from the DWs than is being added by RPV steam. This

)

results in a slow decrease of DW pressure which continues until it drops below the WW pressure by a sufficient amount to open the VBs at about ten hours into the transient.

The retum of air to the DWs via the VBs causes the PCCS energy removal rate to drop below the RPV steaming rate. This energy imbalance causes the DW pressure to increase q

to the point where the PCC vents are cleared and the DW air can again be purged to the WW. The DW pressure then levels out and both the DW and WW pressures remain nearly constant for the duration of the calculation. The DW pressure remains above the WW pressure by just the amount required to keep the water level in the PCCS vents near the vent-exit elevation. This enables air to be slowly bubbled into the WW pools at a rate l

equal to the rate at which it is ingested from the DWs minus the rate at which it is stored within the condensers. The condensers can slowly accumulate air without raising the DW pressure because the RPV heater power continues to drop in accordance with the j

programmed decay heat curve.

l The transient behavior exhibited by this calculation reveals two distinct types of PCCS purging behavior. The first type, leading to the VB opening at ten hours, is characterized by a relatively long period of purging with significant flow through the PCCS vents. This j

effectively cleans out the condensers and leaves them in a condition where they can remove energy at a higher rate than the scaled decay power. As a result, the DW pressure gradually decreases until a VB opens. This is called a " strong" purge. The second type of behavior is illustrated by the purge which follows the VB opening. In this case, the purge time is relatively short. The DW pressure rises until it is just sufficient to clear the PCCS vents. The purge is then accomplished by bubbling out the air at very small vent flow rates. This " weak" purge leaves the condensers with residual inventories of air which i

i 34

i l

l enable the system to regulate itself tojust match the decay power. In this mode of

^

behavior, the difference between the DW and WW pressures remains constant.

The strength of the purge can be associated with the maximum difference between the DW and WW pressures during the purge event or the amount by which this difference exceeds what is required tojust clear the PCCS vents. At the peak of the initial purge, the DW/WW pressure difference reaches 12 kPa or 3 kPa above the vent-clearing pressure difference. This results in sufficient cleansing to preclude self regulation and allows the subsequent energy removal rate to exceed the heater power. In the second purge, the DW/WW pressure difference peaks at 9 kPa which is just the amount required to open the vents and not sufficient to cause effective cleansing of the condensers.

8.0 REFERENCES

i 1.

GE Document 25A5587," PANDA Test Specification", (Rev.1), January,1995.

2.

PSI Document ALPHA-410," PANDA Steady-State PCC Performance Tests Test Plan and Procedures", February,1995.

3.

NEDC-32391P, "SBWR Test and Analysis Program Description", (Rev. B),

j April,1995.

4.

Yadigaroglu, G., " Scaling of the SBWR Related Tests", NEDC-32288 (Rev.1),

(to be issued).

5.

Huggenberger, M., " PANDA Experimental Facility Scaling of the System Lines",

i PSI Report ALPHA-412, January.,1995.

6.

NEDC-32192,"TRACG02V User's Manual", December,1993.

7.

Wilhelmi, F.E., " Decay Heat T.equirements, "SBWR Design Specification, Rev.C," MPL Item Al1-5283, General Electric Co., April,1991.

i 1

t i

35 4

- - -