ML20094E041

Forwards Nonproprietary Version of Tracg Containment Model Road Map as Agreed to at 950821 & 22 Meeting

ML20094E041
Person / Time
Site:	05200004
Issue date:	11/01/1995
From:	Quinn J GENERAL ELECTRIC CO.
To:	Quay T NRC (Affiliation Not Assigned), NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
MFN-222-95, NUDOCS 9511070036
Download: ML20094E041 (14)

Text

_ __ _._ _ _ _ _ _ ._. _.

f GENuclear Energy James E. Quinn,1%jects Manager LAIR and SBWR1% grams Phone (408) 923-1003 Fax (408) 925J991 Novem aer 1,1995 MFN 222-95 Docket STN 52-004 Document Control Desk U. S. Nuclear Regulatory Commission Washington DC 20555 A' ion: Theodore E. Quay, Director Standardization Project 1)irectomte

Subject:

SBWR - TRACG Containment Model " Road Map" (Non-Proprietary).

Reference:

GE/NRC TRACG Containment Meeting, August 21 and 22,1995, at GE SanJose, CA.

The enclosure to this letter is the non-proprietary version of the TRACG Containment Model

" road map" as agreed to in the referenced meeting. The " road map" shows how the models described in the TRACG Model Descdption Licensing Topical Report are applied for SIlWR containment analysis. The enclosure describes the components of the SilWR containment, shows the TRACG input model used to characterize the containment, and details the specific physical models and correlation's used in the major containment regions and components.

Should you have any questions concerning the Subject document please contact Ilharat

Shiralkar of our staff on 408-925-6889.

Sincerely.

I I

i 3 James E.Quinn e

Enclosure:

TRACG Containment Model (Non-Proprietary) )

i l

cc

(1 paper copy w/ encl. and E-Mail w/o encl. except as noted below)

P. A. Ilochnen (NRC/ACRS) (2 encl.)

I. Catton (ACRS)

A. Drozd (NRC)

J. A. Kuddck (NRC) i S. Q. Ninh (NRC) (2 encl.)

J. II. Wilson (NRC) i 9511070036 951101 1\

PDR ADDCR 05200004 A PDR

)

MFN 222-95 Enclosure 1

4 TRACG Containment Model (Non-Proprietary)

The purpose of this document is to provide a " road map" that shows how the models described in the TRACG Model Description Licensing Topical Report are applied 4 for SBWR containment analysis. The first section of this document describes the components of the SBWR containment. The second section shows the TRACG input i j model used to characterize the containment. Finally, the third section details the specific l physical models and correlations used in the major containment regions and l

components.

1. Containment Components i

l Figure I shows a schematic of the SBWR containment and reactor pressure vessel 4

(RPV). The main components of the containment are:

• Drywell The drywell is composed of an upper drywell, bounded by the drywell head, top

! slab, containment walls, and the diaphragm floor separating it from the wetwell.

The upper drywell (indicated by 1 in the figure) constitutes the largest portion of the drywell volume. A break in the Main Steam Line as well as the opening of the

, Depressurization Valves (DPVs) would discharge flow into this region. The annulus region of the drywell (indicated by 2) comprises the region between the RPV and the inner wall of the wetwell horizontal vent duct system. A break in the Gravity Driven Cooling System (GDCS) line would be expected to discharge flow into this region. The lower dgwell is a separate region that is connected to the dr>well annulus by 14,0.8 m OD, annulus to lower drywell connecting vents . Liquid discharged into the upper drywell or the annulus region 1 (e.g. from a broken GDCS line connected to a GDCS pool) will drain into the lower dnwell. A break in the bottom drain line could discharge flow to the lower dr>well.

• Wetwell The wetwell consists of the suppression pool (4) and the wetwell vapor space (5).

The wetwell is bounded by the diaphragm floor on top, containment outer wall and wetwellinner wall on the sides and the floor of the containment. During blowdown flow from the Safety Relief Valves (SRVs) is directed to the suppression pool and quenched via the SRV discharge lines. Flow from the LOCA break and DPVs is directed from the dowell to the suppression pool and quenched via the suppression pool horizontal vent system. Any flow through the Passive Cooling Condenser (PCC) vents is also discharged to the suppression pool.

• GDCS Pools Three GDCS pools (6) are located in the upper drywell. During the GDCS phase of the post-LOCA transient, the GDCS pools discharge into the RPV downcomer, following the opening of squib valves and the check valves in the 3 divisionally separated GDCS lines. During the intermediate and long term phases of the

'MFN 222-95 Enclosure l

l I

post-LOCA transient, the GDCS pools receive condensate from the P units. One l GDCS pool receives the condensate from two PCC units; one receives condensate from the third PCC unit; and the third is not connected to the Passive Containment Cooling System (PCCS). Each PCC unit condensate return line is designed with a loop seal to prevent reverse flow of steam or noncondensibles in the condensate return line.

• PCC Pools '

The three PCC pools (7) are located outside (above) the containment. Each contains a PCC unit. The three pools are interconnected.

• PCC Units The SIlWR has three PCC heat exchanger units (8). Each is comprised of two modules with inlet and outlet headers and 248 tubes in parallel. The PCC units are connected to the top of the upper drywell and discharge condensate into the GDCS pools. Noncondensibles and uncondensed steam are vented to the suppression pool. The vent submergence is 0.9 m less than that of the top horizontal LOCA vent. Thus drywell noncondensibles and uncondensed steam are purged preferentially through the PCC vent line.

Isolation Condenser (IC) Pools The three IC pools (9) are located outside (above) the containment. Each contains an IC unit. The three pools are interconnected with each other and with the PCC pools.

• IC Units The SIlWR has three IC heat exchanger units (10). Each consists of two modules with inlet and outlet headers and 248 tubes in parallel. The IC units are connected to stub tubes, which are attached to the RPV steam dome. Condensate is discharged into the downcomer of the RPV. Noncondensih!es can be vented from the upper and lower IC headers to the suppression pool. This venting processes requires manual action by the operator.

• Depressurization Valves There are 6 DPVs (11) in the SIlWR. Four DPVs are on the RPV stub tubes. (The steam supply lines for the three IC units are also connected to three of these stub

tubes). The other two DPVs are on the Main Steam Lines. The DPVs discharge into the upper drywell.

• Safety Relief Valves (SRVs) and quenchers Eight SRVs (12) relieve RPV pressure by discharging steam into the suppression pool. Steam is discharged though quenchers to minimize chugging and condensation loads. The quencher submergence is greater than that of the top row i of horizontal vents.

MFN 222-95 Enclosure

• Horizontal vent system The SBWR has eight sets of horizontal vents between the drywell and the suppression pool. Each set of three vents consists of three horizontal vents (13) attached to a vertical vent pipe. The top row of horizontal vents is approximately 0.9 m below the bottom of the PCC vents.

• GDCS Equalizing lines Three GDCS equalizing lines (14) connect the suppression pool to the RPV downcomer. During the long term portion of the post-LOCA transient, the squib valves in these lines will open if the levelin the downcomer drops to 1 m above the top of the active fuel and a time delay of 30 minutes has elapsed.

• Vacuum Breakers The SBWR has three vacuum breakers (15) connecting the upper drywell to the wetwell vapor space. The vacuum breakers will open to relieve a negative pressure difTerence between the drywell and the wetwell.

2. TRACG Nodalization The TRACG model used for SBWR analysis includes a representation of the RPV and the containment. All the major components referred to in the previous section are present in the model. Table 1 details how each component is represented in the TRACG model.

3. TRACG Physical Models used in Containment Regions The SBWR containment is represented by a combination of S-dimensional and '

ldimensional TRACG components. All these components utilize the same conservation equations and constitutive correlations. The major physical models used in each region are discussed below.

Break 5

The critical flow is calculated based on the upstream pressure, enthalpy and void fraction. The void fraction will depend on the position of the two-phase level in the downcomer.

In many thermal hydraulic codes (RELAP, other versions of TRAC), the kinetic energy terms in the energy equation are eliminated using the momentum equation. This leads to a form of the energy equation which is nonconserving when discretized; i.e. the energy leaving the RPV is not exactly equal to that deposited in the containment. In TRACG, the kinetic energy terms have been retained in the energy equation and the

. discretization is in a conserving form.

Later in the transient, the flow through the break will no longer be choked.

TRACG cffectively calculates the minimum of the Bernoulli flow from the momentum

'MFN 222-95 Enclosu -

. equation and critical flow. The flow calculated from the momentum equation cannot  !

i exceed the critical flow. At low pressures, the flow will not be limited by critical flow.

Drywell The drywell is modeled as a 3-dimensional region, with 4 radial rings in the upper drywell and 2 radial rings in the annular and lower drywell regions. This allows natural

circulation patterns to develop, if calculated, with upflow in one ring and downflow in

another. The three-dimensional conservation equations for mass, momentum and energy are applied in this region.

Specific models are discussed below:

i i Tudmlent Shear between Cells 1

The TRACG model for turbulent shear between cells at cell boundaries is not being

used. Thus, there is no shear between adjacent cells. All flows in the drywell are driven j by buoyancy and wall shear.

l N<mcondensibleDistribution l TRACG has a mass continuity equation for one species of noncondensible in

addition to steam. The noncondensible is treated as a perfect gas and its properties are l specified in terms of the gas constant, R and the specific heat cpg. The noncondensible j gas (or mixture of gases) has the same temperature and velocity as the steam in a given cell. The partial pressure of the noncondensible gas is calculated based on the i temperature and mass of gas in a cell (Perfect Gas Law). Dalton's law relates the partial j pressures of steam and noncondensible to the total pressure. Note that there are no j requirements for the steam to be at saturation conditions corresponding to its partial

pressure.

! The TRACG model for molecular diffusion of noncondensibles driven by concentration gradients is not used. Noncondensibles are transported solely by bulk convection. Diffusion effects will be small for nitrogen and air. Transport by diffusion could be significant for helium, and to a lesser degree, for hydrogen. Buoyancy effects are not treated at a local level; i.e., steam and noncondensibles have the same velocity in  !

a cell. However, buoyancy effects will be accounted for on a global level. For example, if ,

a light noncondensible is injected into a cell, a natural circulation pattern will develop  ;

between adjacent rings, and lighter fluid will rise to the upper regions. (GE plans to -

implement an additional mass balance equation for a second noncondensible gas species l in TRACG.)

L l WallFriction Correlations i

j The flow regime in the drywell is mostly single phase vapor. In some cells, a j dispersed droplet high void fraction regime may exist. This corresponds to cells wherc l

1

- i

~

MFN 222-95 Enclosure liquid from the break or from the GDCS pool with a broken line is falling to the lower -

regions of the drywell. In some cells, a liquid film can form on the wall because of condensation. The single phase friction factor is utilized. The Reynolds number is calculated based on the axial velocity in the cell adjacent to the wall and the hydraulic diameter of the cell in the direction of the wall. In case a two-phase flow regime is present a two-phase multiplier will be applied.

InterfacialShear Correlations For the droplet flow regime, the models will be employed to calculate the interfacial shear between vapor and droplets. For cells with wall liquid films, the annular flow correlations are used.

WallHeat Transfer The important modes of wall heat transfer in the drywell inc!ude forced and free convecdon to vapor and condensation heat transfer.

For forced convecdon, TRACG uses the Dittus-Boelter correlation, based on the cell velocities and properties. The hydraulic diameter of the cellin the direction of the wallis used in the correlation. The vapor properties are calculated at the cell fluid ,

temperature.

For free convection, the McAdams correladon is used. Again, the cell temperature is used for the calculation of vapor properties and the cell hydraulic diameter for the calculation of the Gr.tshof number.

TRACG will evaluate both the free and forced convection correlations and use the higher of the two calculated values. The same correlations are used for horizontal surfaces.  !

The condensation correlations are included. A Nusselt condensation correlation is used with multiplicative factors for shear enhancement and degradation by noncondensibles. In these equations, the liquid film Reynolds number Rei si defined as:

Rei = 4F/ 3, where F is the condensate flow rate per unit perimeter of surface and pi is the liquid viscosity. The Reynolds number Remis based on the cell hydraulic diameter, vapor velocity and vapor properties. The correlation accounts for the degradation by noncondensibles through the use of the Vierow-Schrock degradation factor. The ratio of the mass fraction of the noncondensible species to the mass fraction of steam in the cell is used as the correlating parameter.

Fogging ofDrywell Vapor Heat transfer from the vapor in a cell will result in cooling of the vapor. If the temperature drops below the saturation temperature of the steam corresponding to its

- .-. - - . - .\

i' i .

MFN 222-95 Enclosure j .

partial pressure, condensation will occur. Generally, in this situation a cold wall will be ,

i present in the cell. A liquid film will form on the surface because of condensation. This i

will be typically the dominant form of condensation in the cell. If the temperature drops j

j. below saturation in a cell that has no heat transfer surfaces, liquid droplets will form j (fogging) by condensation of steam. In this situation, a droplet flow regime will exist.

j . Interfacial heat transfer between droplets and vapor will be calculated. Interfacial shear  !

j between the droplets and steam is calculated.

In general, heat transfer from the vapor is more likely to lead to condensation on the walls. Fogging is more likely to occur as a result of adiabatic expansion of steam from pressures higher than 30 bar.

WetwellVapor Space The wetwell vapor space is also represented by 3-dimensional cells. Typically, two rings and two axial levels are employed in the TRACG model. This would allow for natural circulation in this region. The flow regimes in this region will be the same as in

- the drywell: single phase vapor, dispersed droplets resulting from entrainment from the suppression pool, and a condensate film on the walls. The models discussed in the preceding section for the drywell for turbulent shear between cells; noncondensible distribution; wall friction; interfacial friction; wall heat transfer; fogging and interfacial heat transfer apply also in the wetwell vapor space. One other model is important for this region, namely the heat transfer at the suppression pool interface.

Interfacial heat transfer at pool interface The interfacial heat transfer coefficients on the vapor and liquid sides of the interface are included. The Sparrow-Uchida correlation is used to calculate degradation i of heat transfer at the pool surface due to noncondensible gases.

Suppression Pool The suppression pool is represented by 3-dimensional cells. At least two rings are used to represent the pool. The major phenomena ofinterest for the suppression pool include condensation of vapor bubbles, temperature distribution / thermal stratification and pool two-phase level.

Condensation ofvaporbubbles In the presence of noncondensibles, the bu21es will include steam and ,

noncondensibles. The partial pressure of steam and noncondensibles will be calculated as stated earlier. The interfacial heat transfer from the liquid to the vapor is calculated.

There is no degradation in heat transfer due to the pre ence of noncondensibles. This is based on large scale data showing complete condensation of steam in the bubbles.

- - - . - . - _ - -. - _ . . . _. - .. _ -. . . - - - ~ - - - _ - -

MFN 222-95 Enclosure Pwltemperature distnbution An empirical model is used to force thermal stratification below the lowest thermal source to the pool. This is done by effectively limiting the amount of water that participates in the absorption of energy to that above the lowest discharge location (i.e.,

lowest active horizontal vent, SRV quencher or PCC vent). Above this elevation, TRACG  !

will calculate circulation velocities which produce a well mixed region.

Pwllevel The two-phase level model is used to calculate the pool level. The liquid and vapor side interfacial heat transfer coefficients are calculated. When the liquid surface is subcooled, the condensation at the surface is reduced by a degradation factor based on the Sparrow-Uchida correlation.

GDCS Pools The GDCS pools 'are also modeled as part of the SD containment model. In practice, two pools are represented, with one accounting for the volume of two of the three pools. The representation is essentially 1-D, with each pool being characterized by one ring. The main phenomenon ofinterest for the GDCS pool is the pool level and the associated inventory of water in the pool. The two-phase level model referred to earlier is also applicable here. Heat transfer at the pool surface is modeled analogously to that for the suppression pool.

IC/PCC Pools The pools are represented as part of the SD TRACG region, partitionnd into the IC and PCC pools. The pools are allowed to communicate with each other at the bottom and the top. Two PCC pools have been combined into one, and the three IC pools into

one. The pools are modeled with two rings each and with several axiallevels. Heat transfer occurs from the PCC and IC headers and tubes to the water in the pools. Pool side heat transfer is calculated by the Chen correlation for boiling heat transfer.

(GE plans to use the pool boiling portion of the Chen correlation (Forster-Zuber) for the pool side heat transfer correlation in TRACG.)

IC/PCC Units The IC and PCC units are represented by 1-D components simulating the inlet piping, headers, condenser tubes, condensate discharge lines and vent lines. One dimensional forms of the mass, momentum and energy equations are applicable. Heat is transferred through the walls of the tubes and headers to the respective pools.

MFN 22245 Enclosure WallFriction Correlations The flow regime in the PCC and IC is single phase vapor at the inlet. Due to condensation, a liquid film forms on the walls. The exit conditions consist of a draining liquid film, and a gas mixture that is rich in noncondensibles. The single phase friction factor is obtained. The Reynolds number is calculated based on the axial velocity in the cell and the hydraulic diameter of the cell. In the condensing region, a two-phase multiplier will be applied.

InterfacialShear Correlations For cells with wall liquid films, the annular flow correlations are used.

Wall Heat Transfer The important mode of wall heat transfer in the PCC and IC is condensation heat transfer. Under conditions where condensadon heat transfer is severely degraded by a large amount of noncondensibles, forced convection from the vapor to the wall will become the mode of heat transfer.

The condensation correlations are discussed. A Nusselt condensation correlation is used with multiplicative factors for shear enhancement and degradation by noncondensibles. The Nusselt correlation is expressed. The liquid film Reynolds number Rei si defined as: Rei = 4F/ i, where f is the condensate flow rate per unit perimeter of surface and pi si the liquid viscosity. A shear enhancement factor is used.

The Reynolds number Remis based on the cell hydraulic diameter, vapor velocity and vapor properties. The correlation accounts for the degradation by noncondensibles through the use of the Vierow-Schrock degradation factor. The ratio of the mass fraction of the noncondensible species to the mass fraction of steam in the cell is used as the

correlating parameter.

I

! (GE plans to use the Kuhn-Schrock-Peterson form of the condensation correlation in TRACG.)

{ For forced convection, TRACG uses the Dittus-Boelter correlation, based on the cell

velocities and properties. The hydraulic diameter of the cell is used in the correlation.

The vapor properties are calculated at the cell fluid temperature.

, Depressurization Valves The DPVs are modeled using the VALVE component, which is a 1-D component.

i The TRACG control system will trigger the DPVs to open based on the sensed level in the

RPV downcomer. The primary TRACG model associated with the DPV is that of critical j flow, which was referred to earlier in connection with the break.

i l

MFN 222-95 Enclosure 4

Critical flow is calculated using a model which has been extensively qualified. The critical flow is calculated based on the upstream pressure, enthalpy and void fraction.

Correladons for interfacial shear used in the calculation ofinterfacial shear in the RPV are incorporated. The void fraction will depend on the posidon of the two-phase level in the downcomer. Validation of the void fraction and two-phase level models is performed.

SRVs and Quenchers The SRVs and associated piping are represented by 1-D components. TRACG will  ;

trigger the opening of the SRVs based on pressure or downcomer level. The quenchers are not modeled in detail. Condensation and chugging loads will m be calculated with ,

l TRACG. Critical flow models used for the SRVs have been discussed for the break and DPVs. Models for the condensation of SRV discharge were referred to in the section on the suppression pool.

HorizontalVents The horizontal vents are represented by 1-D TEE components. The 1-D level model r is used in the vertical pipe that connects to the three horizontal vents. As the level drops in the pipe to " uncover" the horizontal vent, the vent will pass two-phase flow to the suppression poc . TRACG is a used for calculation of the pool swell, vent chugging or .

condensation oscillation loads.

Row Regime The flow regime in the vents is single phase liquid, until the vent begins to uncover.

The flow to the vent is " donor celled" at the upstream conditions in the vertical pipe.

TRACG calculates a transition from stratified to dispersed flow based on a critical Froude ,

number.

PressureDmp Cmelations

{. The single phase friction factor is obtained. The Reynolds number is calculated l

based on the axial velocity in the cell and the hydraulic diameter of the cell. The

pressure drop in the vent is actually dominated by the inlet and exit form loss

coefficients. A two-phase multiplier will be applied for wall friction.

Vent Back Pressure As the vent discharges vapor into -he suppression pool, it will tend to move the j liquid in the pool above the vent upwards as it expands. The inertia of this liquid tends j to create a back pressure effect, reducing the discharge flow. This efTect is accounted for

! in the TRACG momentum equation. The liquid mass in the inner ring immediately l above the discharge lormion will have to be accelerated upwards as the vapor expands l into the pool.

I i

MFN 222-95 Enclosure Equalizing Line The equalizing lines are represented by a 1-D VALVE component. The correlations used for wall friction and singular losses are the same as described in the previous paragraph for the horizontal vents.

Vacuum Breakers The vacuum breakers (VII) are represented by 1-D VALVE components. Two VIls are lumped together as one component. The VIls are triggered open at a set negative pressure differential between the drywell and wetwell. They will close at a lower value of the pressure differential. The VIls tran: port flow from the wetwell vapor space to the drywell at conditions corresponding to the cell in the wetwell vapor space to which they are connected. The correlations used for the singular losses are the same as described in previously for the horizontal vents.

1 l

MFN 222-95 Enclosure Table 1. SBWR System TRACG Model REGION SBWR PLANT TRACG MODEL El'Y lower plenum multi-d multi-d (VSSL01: L1/R1 and R2) core multiple 1-d one 1-d (CHAN08) bypass Muldple 1-d multi <1 (VSSL01: L2/Rl) chimney multi 41 multi-d (VSSL01: L3 to L6/Rl) downcomer 1-d multi-d (VSSL01: L2 to L6/R2) steam dome multi-d multi-d (VSSL01: 17 and L 8/R1 and R2) dryers / separators multiple 141 models pressure loss SRVs eight l<l two 1-d (VLVE24 and VLVE28)

DPVs sixl-d three 1-d (VLVE12, VLVE13, VLVE19) steam lines two 141 one 141 + break (TEE 20,23,27 + PIPEl1)

EQLs three 1-d one 1-d (VLVE07)

IlW upper DW multi-d multi-d (VSSL01: L1 to L9/R1 and R2) lower DW multi <l 1-d (TEE 35) vacuum breaker three 141 two l<1 (VLVE10 and VINE 42) main vents eight 1-d with three one 1-d with three "T" connections "T" connections (TEE 02, TEE 03, TEE 04, TEE 09)

.GDCS pools three multi <1 two multi-d i i (VSSL01: L7 and L8/R5 and R6) l l

l injection lines three 1-d two 1-d (VLVE48 and VIXE49)

, PCCS inlet lines three 1-d two 1-d (PIPE 22 and PIPE 83)

) steam headers three multi-d two 1-d (PIPE 92 and PIPE 86) tubes 1488 l<1 two 1-d (PIPE 96 and PIPE 87)

.' condensate headers three multi-d two 1-d (TEE 26 and TEE 88) l drain lines three 1-d two 1-d (PIPE 46 and PIPE 84)

. vent lines three 1-d two l<1 (PIPE 52 and PIPE 85) 1

, pools three multi-d two multi-d l'

, (VSSL01: L11 to Ll7/R3 to R6) i f

J

MFN 222-95 Enclosure Table 1. SilWR System TRACG Model (Continued)

REGION SBWR PLANT TRACG MODEL LCS inlet lines three 1-d one 1<l (TEE 21) steam headers three multi-d one 1-d (PIPE 91) tube; 14401-d one 1-d (PIPE 95) condensate headers three multi 41 one l<l (TEE 25) drain lines three 1-d one l<l (VLVE47) vent lines three 1-d one 1-d (VLVE51) pools three multi <1 one multi <l (VSSL01: L11 to Ll7/R1 and R2)

}

l l

l

MFN 222-95 Enclosure 1

1 i

l l

- ? .. . . _ __ ___

f t_.y-

,, k o

k gma j mts ,p ,

m __

$ A, A='.= ----

};

,. ~_ z na -. .-

" 15 -

a ) .g

, h 13  :

., i 12 3s f-g,;r E

h _ .

(

y '.

,, M . ;> ,

S

lL,_ ;y, y, . to L ~

,+

E'. '.,,5

, ,g. ,

4.

s 4 D . , - sqmsgua y -

?(l_'I: _ %? r ,

f

'^

.~ . ,. m s, ,

cy ,, , y: u , .;, ,

d Figure 1. SBWR Conoioment

,