WCAP-15862 APP-SSAR-GSC-588 AP 1000 · 2012. 11. 19. · WCAP-15862 APP-SSAR-GSC-588 API000 l0...

WCAP-15862APP-SSAR-GSC-588 AP 1000

Figure 9.C.2-22 Perspective view, vertical and upper horizontal cross-sections through BMCwith main steam feedline

(reprinted from L.Wolf, M.Gavrilas, K. Mun, "Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAS", University of Maryland atCollege Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765")

Revision I5956rd c4-.Od-4 1404

9.C-101

I

WCAP-15862APP-SSAR-GSC-588 AP I000APP-SSAR-GSC-588 APIOOO

-.-J4

K>

Figure 9.C.2-23 Perspective view, vertical and upper horizontal cross-sections through BNICwith auxiliary steam feedline

(reprinted from L.Wolf, M.Gavrilas, K. Mun, "Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAS", University of Maryland atCollege Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765")

9.C-102 Revision I9.C-102 Revision I

59560lc4.wpd 041404

WCAP-15862APP-SSAR-GSC-588 API000

l0

Dampfeinspeisung:D F1 271D TL271D PL 271DTLOOOL02

Helidampt f. Sumpfbrw. Hilfs-Dampfeinspciu'ns:DF1 270

0. ISO*

Figure 9.C.2-24 Instrumentation plan for BMC - sump heatup test RX4

(reprinted from L.Wolf, M.Gavrilas, K. Mun, "Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAs", University of Maryland atCollege Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765")

Revision I5956r -c5.vpd-041404

9.C-103

I

WCAP-1 5862APP-SSAR-GSC-588 API000

--- ---- - -- .

a1Wa..

I-

-d'4

K-)

0. 1. 2. 3. 4. S.Time [hours]

Figure 9.C.2-25 Battelle-Experiment RX4: Sump and atmospheric temperatures


K>9.C-104

Revision I9.C-104 Revision I

5956r] c5.wpd-041404

WCAP-1 5862APP-SSAR-GSC-588 AP1000

C~

V

CI

0. 1. 2. 3. 4. 5.Trune [hours]

Figure 9.C.2-26 Battelle-Experiment RX4: Velocities in side vents

(reprinted from L.Wolf, M.Gavrilas, K. Alun, "Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAs", University of Maryland atCollege Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765")

Revision I5956r0.cS.wpd-041404

9.C-105

-1

WCAP-15862APP-SSAR-GSC-588 API 000

/I . ._

A-

aI.

dL-

0. 1. 2. 3. 4.Trime [bours]

5.

Figure 9.C.2-27 Battelle-Experiment RX4: Temperatures in center compartment and dome

(reprinted from L.Wolf, I.Gavrilas, K. Mlun, "Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAs", University of Maryland atCollege Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765")

9.C-106 Revision I

9.C-106 Revision I5956r] -c5 .,pd-04 1404

WCAP-1 5862I -- ee A" D rl cO0 AP 1000A rjr-a 3tM-%j 0 0

9?LMMIO 9 PLM A W20 TL NON 0 9 YLVI A 129TLM9 go

MINK "_lato O" 'M ZMAV

ITL 9TICrMAV

,I'LM"II*'ILM. .

ITLCA 315 S 'O' PWTTLL M 1. 3. STL2.A 12.TL - A 36 ITLZMI MW

CTLOCAN ,TLZMA 316WOTLMR32 CTLMAMWITLOWN" TLrMR 36- elSTLIMS So n2psan-t,ITLMA Vrowmanm I! VILOWLE

Ts id is 9JK2MA3*,TIMA 3,0

'TL HK 200 A ItI it No A M 9TK3"A0 TL on U 40 2*A MW:vr!uu a TL"OA3*W

TL7*aSTLSIGS " TL" I

TL'M wa TLMA 3.:V

TL3nU

IrL 00 I. " VT= "a

3?L'3. M JLZ

?SVDM a

U

OE .ga :

v9TL32SUOS--0 9TL70R10A 9TL27OR22M 9TIA27flR36-0 9TL270K60 Ž

c;

0. 1. 2.Time Ihouru)

3. 4. S.

Figure 9.C.?2-28 Battelle-Experiment RX4: Tcmperatures in external annulus

(reprinted from L.Volf, M.Gavrilas, K. Mun, "Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAs", University of Maryland atCollege Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765")

K>9.U-IIJI

Revision I5956rl .cS.wpd-04 1404

s.c;-107U


0

I.

?-

00

0

0U

0. 1. 2. 3.Time [hoursi

5.

Figure 9.C.2-29 Battelle-Experiment RX4: Hydrogen concentration along circulation path


K>9.C-1 08 Revision 1

5956r1-cS.wpd-041404


Cr4

C!

C-;U

C .0I-

v 4KP60OM32 _ _o 8KP090M17 .i.A 6KP270M17II 9KP32SUO5 ___ ___ = -_____e 9KP270R10 - ~ -

IH2-E nl s=e

1. 2. 3.Time [hours]

4. S.

Figure 9.C.2-30 Battelle-Experiment RX4: Hydrogen concentrations in intermediatesubcompartments

(reprinted from L.Wolf, NI.Gavrilas, K. Mlun, "Overview of experimental results for long-term,large-scale natural circulations in LWR-containments after large LOCAs", University ofMaryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765")

Revision I5956rl -c5.wpd-041404

9.C-109


amnex MV 3 in29A11

, L On H U6 O CLSSA MW

ai.s mu _ 1 .22n 3 M

VlL2NA 3W

MCAlle a s 3n,

a7snLuu IBM

2 .23I U .6W39a

0

01

-s

I\..-

0. 1. 2. 3.Time [hours]

5.

Figure 9.C.2-31 Battelle-Experiment RF4: Relative humidity in the containment atmosphere

(reprinted from L.Wolf, M.Gavrilas, K. lMun, "Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAs", University of Alaryland atCollege Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765")

K2_9.C-1I0


5956r I -. N pd 041404

WCAP-1 5862APP-SSAR-GSC-588 AP I OQO

9.C.2.1.4 References

1. Fischer, K., Kanzleiter, T., Schall, M., Wolf, L., 1989"CEC Thermal-Hydraulic Benchmark Exercise on FIPLOC Verification Experiment F2 inBMC-Long-Term Heatup Phase - Specification for Phase I, Battelle-Institut e.V., Frankurt/Main,Germany, Sept. 1989

2. Fischer, K., T., Schall, M., Wolf, L., 1991"CEC Thermal-Hydraulic Benchmark Exercise on FIPLOC Verification Experiment F2 inBMC-Long-Term Heatup Phase - Results for Phase 1, Final Report, EUR 13588 EN, 1991

3. Fischer, K., Kanzleiter, T., Schall, M., Wolf, L., 1990"CEC Thermal-Hydraulic Benchmark Exercise on FIPLOC Verification Experiment F2 in BMC,Experimental Phases 2, 3, and 4 - Specification for Phase II, Battelle-Institut e.V., Frankurt/Main,Germany, July 1990

4. Fischer, K., T., Schall, M., Wolf, L., 1993"CEC Thermal-Hydraulic Benchmark Exercise on FIPLOC Verification Experiment F2 in BMC,Experiment F2 in BMC, Experimental Phases -Results forPhase 2,3 and 4, Results of Comparisons,Final Report, EUR 14454 EN, 1993

5. Fischer, K., Hafner, W., Holzbauer, H., Kanzleiter, T., (1994)"Experiments for Concerning Natural Convective Flows in Battelle Model Containment,Documentation of Measured Data"(In German), Battelle lngenieurtechnik GmbH, Eschborn, Germany, Technical Report V68270.2,Oct. 1994

6. T. Kanzleiter, 1988"FIPLOC-Verification Experiments",(In German), Battelle-Institut e. V., Frankfurt/Main, Germany, Final Report BleV-R-66.614-01,March 1988

7. Langer G., Jenior R., Wentlandt H.G., 1979"Experimental investigation of the hydrogen distribution in the containment of a LWR following aLOCA"(In German), Battelle-Institut e. V., Frankfurt/Main, FRG, Report BF-R-63.363-3

Revision 1 9.C-1I I5956rl -9c.upd-041404

WCAP-15862APP-SSAR-GSC-588 AP1000

8. Langer G., Baukal W., 1982"Experimental investigation of the hydrogen distribution in a model containment (Preliminary

Experiments 11)"(In German), BMFT-Research Contract 150.375, Battelle-Institut e. V., Frankfurt/Main, FRG,

Report BF-R-64.036-3

9. Petersen, K., Pamme, H., Seyffarth, L., Wolf, L., (1994)"Hydrogen Mixing by Natural Convection in PWR Containments"

(In German), atw 39 (1994), 758-769

10. L.Wolf, H. Holzbauer, M. Schall, 1994,

"Comparisons Between Multi-Dimensional and Lumped-Parameter GOTHIC Containment Analyseswith Data" International Conference on New Trends in Nuclear System

Thermohydraulics, Pisa, Italy, May 30th - June 2nd, vol. 2, pp. 321 - 330.

11. L. Wolf, M. Gavrilas, K. Mun, 1996,

"Overview of Experimental Results for Long-Term, Large-ScaleNatural Circulations in LWR-Containments after Large LOCAs"

Final Report for DOE - Project, Order Number. DE-AP07-961D 10765, Department of Materials andNuclear Engineering, University of Maryland at College Park, July 1996.

9.C.2.2 DESCRIPTION OF THE AVAILABLE NUPEC DATA BASE

9.C.2.2.1 MI-7-1 Test

NUPEC's Hydrogen Mixing and Distribution Test M-7-1 is used as OECD/CSNI sponsored International

Standard Problem Exercise ISP-35 (report NEAICSNI/R(94)29 ) to compare and validate the performanceof various computer codes. The WGOTHIC code is included among the codes that have been compared to

this test (see ref. report NEA/CSNI/R(94)29 ). A detailed presentation of the M-7-1 test is not provided

because the test is not directly applicable to the AP600 or AP1000. Internal sprays, which are not used in

the passive containment design, are active during the M-7-1 test. A more relevant NUPEC experiment, theM-4-3 test (reference T. Hirose, 1993), is performed without internal sprays. This test investigates the

mixing behavior of hydrogen and steam injected into the lower containment compartment.

9.C.2.2.2 M.-4-3 Test

The test facility represents one quarterofa linearly scaled PWR (four-loop) containment model with 25 inner

compartments. The approximate volume is 1600 rn3 with a height of 17.4 m and an inner diameter 10.8 m.

Of the total volume, only 1300 m3 is used. The dome compartment is approximately 70 percent of the total

9.C-112 Revision I5956ri -9c.wxpd-041404


volume. Twenty-four smaller compartments occupy from 0.1 percent to 4 percent of the total volume. The

containment walls are made of steel plates (concrete structures are not present), so that the response of the

heat sinks was very fast. The external containment wall is 12 mm thick steel and is insulated. Three floors

are located at the 3.2,5.4, and 7.3 m levels. A large sump is located below the lowest floor. A gas supply

system for steam and helium, a containment internal spray supply system, and measuring systems fortemperature, helium concentration, and pressure are installed.

The M4-3 test simulates a break inside the lower D loop of the steam generator compartment. This

compartment is at the first level of the containment. Figures 9.C.2-32, 9.C.2-33, and 9.C.2-34 illustrate the

containment compartments and their corresponding numbers on each level. Figure 9.C.2-32 also shows the

location of the break and typical circulation patterns.

The M-4-3 test starts with a pressure of 101.35 kPa and an initial containment temperature of 300C. Amixture of steam and helium is injected at a constant flow rate during the first 30 minutes. The mass flow

rates simulating a small break are 0.33 kg/s and 0.027 kg/s for the steam and helium, respectively. The

inflow temperatures are 1400C and 200C for the steam and helium, respectively.

Experimental data is recorded for 2 hours. The atmosphere and wall temperatures, the helium concentrations

in the various compartments, and the pressure inside the dome are recorded. The pressure inside the domeincreases almost linearly during the release and reaches 1.6 bar after 30 minutes (end of the release). The

total mass of the steam released is 594 kg. Following the release, pressure decreases slowly to 1.5 bar (after

2 hours - see Figure 9.C.2-35). The dome temperature increases from 303 to 341 OK during the release andthen decreases to 3340K (see Figure 9.C.2-36). The dome helium concentration history is presented in

Figure 9.C.2-37.

The temperature and helium concentrations increase first near the release point and the compartments aboveit, indicating an upward flow direction. This is followed by increases in the dome and steam generatorcompartments at the opposite side of the containment, indicating a downward flow loop direction. The

global circulation loop is closed in the lower level of the containment. The consecutive positions of thehigher temperatures and the helium concentration fronts indicate the flow path due to the global circulation

during the break release (see Figure 9.C.2-38, from 0 - 1800 seconds).

A comparison of the temperatures and helium concentration histories in various compartments,Figure 9.C.2-38, shows the direction ofthe formed global circulation loop during the break release. One flow

path starts in compartment 8, followed by compartments 15, 21, 25 (dome), 14, and 7. Following breakrelease, the temperature field stratifies but the temperature difference between the dome and first floor level

is not greater than 14IC at the end of the experiment. After the break release, the temperatures in all the

compartments slowly decrease, despite the fact that the outer surface of the containment is insulated.

Revision I 9.C-1 135956rl -9c.wpd-041404

I


Figure 9.C.2-38b shows that the concentration in the containment starts to homogenize after the end of therelease (1800 seconds). This indicates that natural circulation effects are active after the end of release.

Note that the external walls of the containment are insulated and the interior walls are made of steel. This

limits their ability to accumulate energy and act as a heat sources (which would promote natural circulation)later in the test. The natural circulation that exists after the end of release is most likely formed by the

presence of the hot sump at the bottom of the containment.

Throughout the test, global circulation is present inside the containment and contributes to the

homogenization of the atmosphere over time.

Application to AP600 and APIOOO

This test is similar to the AP600 and APIOOO plant configurations in both the position of the injection point(break at the low level) and the geometry of the containment (70 percent of the volume is dome). However,overriding the similarities are the absence of concrete structures (able to store heat and later act as heatsources) and cooling on the outer surface of the containment shell. These features make the test very

dissimilar to the passive containment design. The external cooling and the hot concrete structures, positioned

in the lower portion of the passive containment produce strong global circulation and decrease the effectsof thermal stratification.

<2,9.C-1 14 Revision I

5956rl-9c.wpd-041404

WCAP-1 5862APP-SSAR-GSC-588 APIOOO

BREAKRELEASE -

LOCATION

Figure 9.C.2-32 Model containment used in the test, break-release position, and circulationflow paths (see arrows) ..

(Figure prepared according to T. Hirose documentation for AX4-3 test data, NUPEC Nuclear PowerEngineering Corporation, ISP35-035, 30 April, 1993)

Revision I5956rl cS.%pd-041404

9.C-115

WCAP-I 5862APP-SSAR-GSC-588 APIOOO

can't' '*1ep

K-'

Figure 9.C.2-33 Detailed arrangement of each floor

(Figure prepared according to T. Hirose documentation for M14-3 test data, NUPEC Nuclear PowerEngineering Corporation, ISP35-035, 30 April, 1993)

9.C-l 16 Revision 1

9.C-1 16 Revision I5956r] -6.uzpd-041404

WCAP-15862APP-SSAR-GSC-588 APIOOO

3rd Floor

2nd Floor

1st Floor0 * Compartmnt number

Figure 9.C.2-34 Distribution of the compartments at each floor

(Figure prepared according to T. Hirose documentation for A1-4-3 test data, NUPEC Nuclear PowerEngineering Corporation, ISP35-035, 30 April, 1993)

Revision I5956rl -c6.wpd-04 1404

9.C-1 17

I


. :.

Co0

-~ O 0

CD

CL.

'..... .....................

CD3

a.O

O 1.8 3.6 - 5.4

Time (sec) 2

K>~

Figure 9.C.2-35 Pressure history

(Figure prepared according to T. Hirose documentation for A14-3 test data, NUPEC Nuclear PowerEngineering Corporation, ISP35-035, 30 April, 1993)

K1-)

9.C-l 18 Revision 1

9.C-1 18 Revision 15956rl -c6.wpd-041404


oo

CO

_ O

CD .=~~~ X .*-**-****.....

o ;

CDcv,

E CCDCD

0

oo00

T 1.8 3.6Time (sec)

* .. I ...... * * .-.

* . I 11 ._ .

Figure 9.C.2-36 Dome temperature history

(Figure prepared according to T. Hirose documentation for 1-4-3 test data, NUPEC Nuclear PowerEngineering Corporation, ISP35-035, 30 April, 1993)

Revision I 9.C-1 19

Revision I5956r] -c6.supd-041404

9.C-1 19

I

WCAP-1 5862APP-SSAR-GSC-588 APIOOOAPP-SSAR-GSC-588 APIOOO

CujC

O LO

Eax 0 . .

E :

(D CD.

E0

1.8 3.6 5.4

Time (sec)

0

Figure 9.C.2-37 Dome Helium Concentration history

(Figure prepared according to T. Ifirose documentation for M-4-3 test data, NUPEC Nuclear PowerEngineering Corporation, ISP35-035, 30 April, 1993)

9.C-120 Revision I

9.C-120 Revision I5956rl -c6.,%pd0411404

WCAP-15862APP-SSAR-GSC-588 APIOOnv

0

-..

I-(D

a)E

0a)

-0a-

C)

E

I

b)

Time (sec)

L(ocmj

Time (sec)

Figure 9.C.2-38 a) Temperature histories for compartments forming a circulation path (8, 15,dome and 7) b) Helium concentration histories for compartments forming acirculation path (8, 15, dome and 7)

(Figure prepared according to T. Hirose documentation for M4-3 test data, NUPEC Nuclear PowerEngineering Corporation, ISP35-035, 30 April, 1993)

Revision I 9.C-121

Revision I5956rl I6.vwpd 041404

9.C-121


K) '9.C.2.23 References

1. NEAICSNIIR(94)29: Final comparison report on ISP-35: NUPEC hydrogen mixing and distribution

test (Test M-7-1), Committee on the safety of nuclear installations OECD Nuclear Energy Agency,

December 1994

2. Tadashi Hirose (1993): Letter, floppydiskwhich containstheM-4-3 test data, an attached document,

and a video-cassette, NUPEC Nuclear Power Engineering Corporation, ISP35-035, 30 April, 1993

KI

9.C-122 Revision I

9.C-122 Revision I5956r]-9c.wpd-041404


9.C.23 DESCRIPTION OF THE AVAILABLE CAROLINAS VIRGINIA TUBE REACTOR

CONTAINMENT (CVTR) DATABASE

The results of the simulated DBA of CVTR containment system are presented in (Schmitt et al., 1970). The

CVTR containment is made of reinforced concrete (see Figure 9.C.2-39). It is cylindrical in shape with aninternal diameter of 57 ft 11 - 1/2 in and an inside height of 114 ft. The cylindrical wall is 2 feet thick. Atthe top of the cylinder is a hemispherical dome made of 1/2-inch thick steel covered by 20.5 inches of

concrete. The structure is divided in three regions (operating, intermediate, and basement). The free volumeis about 227,000 ft3. A water spray system is installed at the 360 foot elevation (6 inches above the

containment bend line).

Three simulated DBA tests are performed. The first test does not use the water sprays. The pressure andtemperature histories are obtained to establish the effects ofthe sprays on pressure reduction. The remaining

tests include active water sprays with flow rates 290 and 500 gpm, respectively. The water spray flow rateis adjusted by changing the total number of nozzles in the spray header. The spray flow is initiated

30 seconds after the end of steam injection and remains in operation for 12 minutes. The second and thirdtests are not applicable to passive containment DBA situations due to the spray effects.

Superheated steam, at approximately 10 0F above the saturation temperature, is injected into the containment.The steam is released above the operating floor through a diffuser made of a 10 foot pipe installed vertically

on the discharge nozzle of the steam line (see Figure 9.C.2-40). The end of the pipe is capped. The pipecontains approximately 126 holes (with 1-inch diameter) that direct the steam horizontally (see

Figure 9.C.2-40). The injection continues until the containment pressure reaches 17.5 psig above the

atmospheric pressure.

To obtain the containment response, measurements of the interior pressure are made using seven fastresponse pressure transducers and a Heise gauge. The containment atmospheric temperature is measured

by 34 thermocouples and 15 resistance thermometers. Two additional thermocouples and pressuretransducers are located in two partially isolated regions (the reactor header cavity and the refueling canal)to determine their thermal and pressure response.: Twenty three thermocouples are positioned at the

containment liner envelope and the interior surfaces to obtain the temperature response of the interior

concrete sections. The heat transfer through the liner wall of the operating region of the containment is

established using heat flow devices along with thermocouples that measure the temperature distribution inthe cross-section of the wall. The weight of condensate collected from the selected areas of the cylindrical

liner walls is also measured.

The mass flow rates of steam for the three DBA tests range from 360,000 to 380,000 lb/hr (see

Figures 9.C.2-41 a, b, c). The injection steam conditions are near to saturated conditions with an enthalpy

of 1196 Btu/lb. The steam is injected in the operating region at 335 ft elevation. The duration of the release

Revision I 9.C-1235956rl -9cuwpd-04 1404


is approximately 180 seconds. The measurements are taken for approximately two hours, but only the first

50 minutes are presented in the report.

Containment pressure histories for the three DBA tests are presented in Figure 9.C.2-24. The influence ofthe spray application and the intensity of the sprays on the decrease of the containment pressure is obvious.

The containment vertical temperature profiles for various tests are presented in Figures 9.C.2-43, 9.C.2-44

and 9.C.2-45, respectively. For the first test, large vertical temperature gradients exist for the duration of the

measurements (about two hours), indicating temperature stratification. The containment air recirculation

system is activated one and half hours after the beginning of the first test and it decreases the verticaltemperature gradients (but not so efficiently as water sprays).

The spray tests (tests 2 and 3- which are not applicable for passive containment designs) contribute towards

the uniformity of the containment atmosphere by decreasing the dome temperatures. After one hour, thedome temperatures were 177, 155, and 137OF for tests 1, 2, and 3, respectively.

One hundred minutes of test data from thermocouples mounted at the same horizontal level inside theoperating region are analyzed. Significant temperature gradients are not established in the horizontal

direction. This is consistent with the view that stratification gradients are mainly in a vertical direction, withnegligible horizontal gradients, except near plumes and wall boundary layers.

Liner vertical temperatures are almost the same as the temperatures in the upper regions of the containment

atmosphere, while they are higher than temperatures in the atmosphere of the lower regions (seeFigures 9.C.246 and 9.C.2-47 for lower regions). An explanation for the higher liner temperatures in the

lower regions is that the steam, in its downward movement, passed through the annular gap near thecontainment wall (surrounding the operating floor) releasing heat predominantly to the wall instead of the

atmosphere.

Data from six thermocouples embedded in concrete surfaces in the containment are presented inFigure 9.C.248 for the case without sprays. The operating region structures have a large temperature

increase, while the basement region is almost unaffected. The results are in agreement with the atmosphericand liner temperature data and show stratification above the break elevation.

The results from two thermocouples installed in two partially isolated regions (the reactor header cavity andthe fuel transfer canal) are presented in Figure 9.C.249. A small temperature increase is seen during the

steam injection. After the injection, temperatures decrease very fast, resulting in the very small final

9.C-124 Revision I5956ri-9c.wpd-041404


temperature increase. The data also indicates that dead-ended compartments get some steam during the

pressure increase.

The reasons for the established temperature stratification in the case of the CVTR containment testperformed without internal sprays are:

The position of the release point is above the 335 foot elevation of containment height seeFigure 9.C.2-39). This could be considered as a release in the upper region of the containment

(H/H, = 0.525).

The external walls and the dome of the containment were made of 23-3/4" thick concrete which

accumulated a portion of the heat released in the upper regions (operational region and intermediateregion - see Figure 9.C.248 for vertical temperature distribution inside the liner). The upper

portions of the walls remain hot as containment cools, producing boundary conditions for a stable

stratification.

External spray cooling of the dome and walls (upper half exposed to the air) is not applied.

The steam condensate is collected in the buckets so that weight of the condensate can be measured.This eliminated hot sump evaporation and heat source effects. The other experiments (BMC and

HDR) indicate that even small temperature differences between the sump (as a heat source at thelowest level) and the containment atmosphere generate global natural circulation inside the

containment.

Revision I5956r] -9c.wpd4041404

9.C-125

WCAP-15862A ODD-C A 0n.C._rRR AP 1000rr-VL AW-O , _ I

KJ

Figure 9.C.2-39 Carolinas Virginia Tube Reactor (CVTR) containment structure

(reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg, "Simulated design basis accident testsof the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. AtomicEnergy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230)

<-I~

9.C- 126 Revision I9.C-126 Revision 1

5 95 6r I c6.%Npd 041404

WCAP-I 5862APP-SSAR-GSC-588 AP1000

#'�=1��i �*

Figure 9.C.240

.. - . :

CVTR steam line diffuser


Revision I5956rl c6.wpd-041404

9.C-127

WCAP-15862APP-SSAR-GSC-588 . API000

- X

a

K>

Figure 9.C.241a Steam injection histories test 1


9.C-128 Revision I

9.C-128 Revision 15956rl yc6.,wpd-D414D4

WCAP-15862APP-SSAR-GSC-588 A 1n IAAr r I uuu

e

I

Figure 9.C.2-41b Steam injection histories test 2


Revision I5956ri c6.wpd-04 1404

9.C-1 29

I


'a8a

Figure 9.C.241c Steam injection histories test 3

(reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg, "Simulated design basis accident testsof the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. AtomicEnergy Commission, Idaho Operations Office, Under Contract No. AT(l0-1))1230)

9.C-130 Revision I

9.C-130 Revision I5956r)-c6.wpd-041404


In ~~

z

e0.

* O

a.

Spray On

tolural Pressure Decoy

' r - 290 qpm Pressure Sippression

, Sproy Off - - - _

500 epon Pressure Suppression Spra_

S

I I t. .. .. . .I-I '1 . *.. '~ '~ A '*tn*'n

-o 10 Time Irminl D*u

Figure 9.C.2-42 CVTR containment pressure response (Heise gauge)


Revision I 9.C-131

Revision I5956r] -0.wpd 0414W4

9.C-131

-I


250

200

I

A2 ISO

i3

'OD

50

KJ

0

Tb-_ 6. .I

Figure 9.C.243 Atmospheric vertical temperature profile, test 1


9.C-132 Revision I

9.C-132 Revision 15956rl Sc6.wpd-01404

NVCAP-I 5862APP-SSAR-GSC-588 AP1000

250

.III

4000

Figure 9.C.2-44 Atmospheric vertical temperature profile, test 2

(reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg, "Simulated design basis accident testsof the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. AtomicEnergy Commission, Idaho Operations Office,'Under Contract No. AT(10-1)-1230)

Revision I 9.C-1 33

Revision I5956r]-c6.wpd 041404

9.C-133


250

200

zI

I, 130

100

O

2000

Tk.. (seeI

3500 0 400

Figure 9.C.245 Atmospheric vertical temperature profile, test 3


9.C-134 Revision 1

9.C-134 Revision 15956rl wc6.wpd-041404

WCAP-1 5862APP-SSAR-GSC-588 AP1OOO

250

200

100

500 500 1000 1500 2000 2500 3000 3500 4000

V.w be, I

Figure 9.C.246 Surface and atmosphere temperatures - intermediate region, test 1


Revision I 9.C-135

Revision 45956rl c6.spd-041404

9.C-135


\-)

E

I

I

,og p I I I I

TC. Tl..c_

TC -s 0mm *jo4Fh J .. E - 284 ti

* TC-32 85l .f Sofg cE - 299 11

200

*50

5v____________

I Soo KM0 r500 2000

T~laectl

2500 3000 3500 4000

Figure 9.C.247 Surface and atmosphere temperatures - basement region, test 1


9.C-136 Revision I5956ri c6.wpd-041404


250

200

L

i

2000

To. (s.e

Figure 9.C.248 Concrete surface temperatures, test 1


KJRevision I5956rl-c7.wpd-041404

9.C-137

I

WCAP-1 5862APP-SSAR-GSC-588 AP1000-- --

250

200

115

.I

100

0 500 1000 1500 2000 2M00

Tim. ses

3000 3500 4000

Figure 9.C.249 Isolated containment region temperatures


V~~9.C-138

Revision I9.C-138 Revision 1

5956flc7.wpd-041404


9.C.2.4 DESCRIPTION OF THE AVAILABLE HDR DATABASE

The HDR containment (Figure 9.C.2-50) is 20 m in diameter and 60 m high with a total free volume of11,300 m3 . It is made of steel and concrete with a total surface area of 30,000 m3 available for heat transfer(condensation). The dome volume, including a crane structure, is about 42 percent of the total volume or

5,000 m3 . The main and spiral staircase form the dominant vertical flow paths connecting the dome with

sump. The containment can be subdivided into 62 to 72 subcompartments containing pipes, pumps, and

valves etc. The subcompartments are interconnected by numerous openings. The numbers and locations of

thermocouples and anemometers for the HDR-E test series are presented in Figure 9.C.2-5 1.

9.C.2.4.1 HDR ElI Test Series

Test group El I (the hydrogen distribution and mixing experiments) is included as a part of the HDR-Safety

Program Phase III that was conducted in the summer of 1989 (Cron and Schrammel, 1993). This test series

simulated severe accident scenarios. The tests provide insights on DBA circulation and stratification. A

review of the natural circulation flows formed in experiments El 1.0 -El 1.5 is presented in Wolfet al., 1996.Circulation is indicated by examination of the recorded temperatures and velocities.

Seven of the eight El I experiments are initiated with small steam releases. The one exception, the El 1.5

experiment, is initiated with a large-break, two-phase LOCA blowdown. The tests provide data for the

transient and spatial distribution of the gas concentration (hydrogen/helium mixture) under various initial

and boundary conditions.

The following features are considered:

The large scale of the experimental facility and the multi-compartment geometry with a relatively

large dome

High hydrogen release rates typical of severe accidents

. ' Superheated steam injection into the containment'- single and multiple steam and hydrogen injectionphases

Internal concrete and metal structures' and surfaces

* Different axial positions for hydrogen and steam releases

Revision I 9.C-1395956rl -9c.vpd-04 1404

I


The efficiency of hydrogen mitigating features, including venting (for example blockage of onemajor vertical flow path)

The impact of internal and external sprays

Sump heat up effects

* Air heater operation effects

9.C.2.4.1.1 Experiment E11.0

Experiment E l 1.0 is performed with an external steam supply and without a gas mixture release. However,the external steam supply is not able to produce typical severe accident conditions. Therefore, a blowdownof the primary system is used for all other experiments. The results of El 1.0 are not discussed here.

9.C.2.4.1.2 Experiment E1I.1

Experiment El 1.1 includes a high-positioned, small break LOCA, followed by additional steam and gasreleases. The effectiveness of the internal spray for decreasing the pressure and temperature inside thecontainment is also tested. Since this experiment included internal sprays that are not a part of the passive

containment design, the results are not discussed here.

9.C.2.4.1.3 Experiment El 1.2

Experiment El 1.2 also includes a high-positioned release point. However, instead of internal spray as inexperiment E I 1. 1, external spray is applied to the dome shell. A detailed presentation of this experiment is

given in Wolf, Mun, Floyd, 1995. This experiment, together with E 1 1.4, provides results used as benchmarkproblems to validate numerical models (Wolf et al., 1994. The verification of the GOTHIC code byHolzbauer and Wolf, 1994, is also based on El 1.2 results.

Figure 9.C.2-52 provides the sequence of the experimental procedures. This includes the duration of theheatup and the blowdown of steam (p=l 10 bar and T=318 0C) in the compartment RI805 (in the mainstainvay at elevation 21.05-25.3 m, see Figure 9.C.2-5 1, right side). The consecutive steam and gas (mixtureof 85 percent He and 15 percent H2) injections are also presented in Figure 9.C.2-52. Compartment 1405,in the spiral stairway at elevation 4.5 m (see Figure 9.C.2-52 right side), receives steam at 770 minutes.

9.C-140 Revision I5956rl -9c.Awpd-041404


The change in the containment pressure with time is shown in Figure 9.C.2-53. The temperatures, steam

concentrations, and He/H 2 gas mixture concentrations at various axial positions of the main staircase are

presented in Figures 9.C.2-54, 9.C.2-55, and 9.C.2-56, respectively. Temperature differences between the

same axial positions of the two vertical flow paths (main and spiral staircases) are pronounced during the

heatup phase. Later, temperature differences decrease and change signs. The crossing temperature histories

in Figures 9.C.2-57, 9.C.2-58, 9.C.2-59, 9.C.2-60 indicate a change in the flow directions.

A steep thermal gradient indicates the presence of the stratification that exists between the lower

compartments and the containment dome (see Figure 9.C.2-61). Even an additional steam release in the

lower portion of the containment (compartment RI405) at 770 minutes is not able to completely eliminate

the vertical thermal gradient, although it is greatly reduced.

The injection of steam into the lower level, together with the effects of cooling the dome with the externalspray, causes redistribution of the He/H 2 gas mixture (see Figure 9.C.2-56). The increase in the gas

concentration in the upper part of the dome is a consequence of the steam condensation on the inside dome

surface. Later, the decrease in temperatures in the upper dome promote global circulation (causing unstable

stratification) and homogenize the dome atmosphere. Even after the external spray is stopped, global

circulation affects the lower parts of the containment (see Figures 9.C.2-54, 9.C.2-55, 9.C.2-56). Overall,

the external spray causes a reduction of 0.44 bar in the containment pressure (see Figure 9.C.2-53).

The application of external sprays at the dome surface promotes global circulation, homogenizes hydrogen

concentrations, and decreases the containment pressure. The water film applied to the external surface of

the AP600 and APIOOO shell produces similar effects.

Revision I 9.C-1415956rl -9cwpd-04 1404


I !)

MAIN

279' 9s.

Figure 9.C.2-50 Cross-sections, elevations and numbers of compartments of the HDRcontainment, left (180°-0°0 cross-section), right (270°-900 cross-section)

(reprinted from: L. Wolf, K. Mun, J. Floyd, "IlDR hydrogen mixing evaluation for containment safetyevaluations" - Phase 1, University of Maryland at College Park, Final Report for DOE-Project OrderNumber: DE-AP07-95lD81401, July 1995)

K-y.

9.C-142 Revision I

9.C-142 Revision I5956fl-0.vupd-041404


LOCATIONS OF THERMOCOUPLES LOCATIONS OF ANEMOMETERS

Figure 9.C.2-51 Locations and Enumerations of Thermocouples (left) and Anemometers (right)for HDR-E-Test Series

(reprinted from: L. Wolf, Al. Gavrilas, K. Mun, "Overview of experimental results for long-term,Large-scale natural circulations in LWR-containments after large LOCAs", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

Revision 15956rl-c7.wpd-041404

9.C-143

I


K.'_

Minutes

Event

HUP

STI

GM

ST2

NCISP6SPSSPIOSP12NC2VENEND

Timeminutes

0-690

690-750

740-770

770-959

959-975975-10951095-11551155-11I51185-12001200-14451445-15111536

_s ___ , . IE Ia

MR -blowdown plus externalsteam supply with m - 2.06 kg/hReduced external seam withm - 1.19 kg/r in Rl.805HJile gas mixture release intoRlSO05 with 0.0O9 kg/sSteam releas into RIA05with 2.06 kg/sNatural cooldovwnExternal spray m - 6 kg/sExternal spray m - 8 kg/sExternal spray m - 10 kghiExternal spray m - 12 kg/Natural cooldownContainment ventingEnd of daft a isition

Event Description

aiN-AI

I I

---. --- ,… …----III

\' '.1a I

I

1; --- ---,-Th1..xM

Figure 9.C.2-52 HDR-E11.2 Sequence of Experimental Test Procedures

(reprinted from: L. Wolf, K. Alun, J. Floyd, "HDR hydrogen mixing evaluation for containment safetyevaluations" - Phase 1, University of Maryland at College Park, Final Report for DOE-Project OrderNumber: DE-AP07-95lD81401, July 1995)

9.C-144 Revision I

9.C-144 Revision I5956rl -0.%pd-041404

WCAP-1 5862APP-SSAR-GSC-588 A 1) IAA

irL lU vu

-z

a

'tPn

,, I I . ....

f

0 TEME 690 770 959 975 1155 1200 1 1536

I r-1 i I.- N1 n I * ItN C2 I I I _

= bM ST6 | SP1hp2\ V&> 'ENDNCI SP8 NC2

Figurc 9.C.2-53 HDR-E11.2 Containment pressure history

(reprinted from: L. Wolf, K. Mun,J. Floyd, "HDR hydrogen mixing evaluation for containment safetyevaluations" - Phase 1, University of Maryland at College Park, Final Report for DOE-Project OrderNumber: DE-AP07-951D81401, July 1995)

Revision I5956rl -c7.wbpd-04 1404

9.C-145


2.

5G

,____ LO IO or Tun Mdo co n

KJ

I

*sfl5rWUIU 690 770 959 975 1155 1200 445 1536i 1,11 S L N C Z M I t l I I 1

HUP (~749' SO \ So I S;1041\O NNOI SP$ NC2-

Figure 9.C.2-54 HDR-E1l.2 Atmospheric temperatures along main stairway

(reprinted from: L. Wolf, K. Mun, J. Floyd, "HDR hydrogen mixing evaluation for containment safetyevaluations" - Phase 1, University of Maryland at College Park, Final Report for DOE-Project OrderNumber: DE-AP07-951D81401, July 1995)

K>_J9.C-146


5956rl Sc7.wpd-041404

WCAP-15862A l) CI! A D' oCtl _QQ APl000

* 1 Hz+31m .2H- +48m _3H-+40m 4

i 4 H=+ 6m * r[5Ht+ Cm

'6 H-- 4m

----------~--r - i ^'

K-'

TIW IN MAMTI-T---

mw

l . N, l 4 s.

ST(74Z&M z \ SI |i10 1oNP12\ VN E2NDMCI No :C

Figure 9.C.2-55 HDR-E11.2 Steam concentrations along main stairvay

(reprinted from: L.Wolf, K. Mun, J. Floyd, "HDR hydrogen mixing evaluation for containment safetyevaluations" - Phase 1, University of Maryland at College Park, Final Report for DOE-Project OrderNumber: DE-AP07-951D81401, July 1995)

K>n r-.,t;Revision I

5956rl-c7.wpd-041404Y.L;-114/


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

\K)

N8

5!-

883vL

- -O I

, IO I

02: H =+48 m0 ,_ 3 H + 40 m '___

4 H=+ GmS Hu+ Om6 H- - 4m

------------.I .... ... I

0 I 1IC 1 85 1 51TdE~frMxIUxE 690 Tn0 959 9175115 10 1445 1536

I I 9 Itl . U I I 9 .M l A I I -l- -

IUP STrU741>M SA2 \ S)6 I S1`0 PIA v& 'EN BDNCI SP$ N02

Figure 9.C.2-56 HDR-E11.2 H2/He Gas concentrations along main stainvay

(reprinted from: L.Wolf, K. Mun,J. Floyd, "HDR hydrogen mixing evaluation for containment safetyevaluations" - Phase 1, University of Maryland at College Park, Final Report for DOE-Project OrderNumber: DE-AP07-951D81401, July 1995)

9.C-148 Revision 1

9.C-148 Revision I5956rl c7.wpd-041404


U

II

1111.2 C 2 norm IS n .co

S I I1~1O ' ~ '1. 1I, I £11.2 I1510 " W II El

91. I a' a" IT a TS_

2b ___- f---------i -'

d § I ; I. -

0 TDEM IN NUN=ji I I IIII 750 1095 1185

690 770 959 975 1155 12001 1 , II ., U I * lqk

fluK-I

ST('740%GM Sn \s I~ P1 >Pio1P\

NCI Sn NC2

I II I

a"crewa cowCH=cowa"MMMMCrM

arrwrwaca,"

Figure 9.C.2-57 HDR-E11.2 Comparison between temperatures in main and spiral stairwaysat +31.0 m level

(reprinted from: L. Wolf, Al. Gavrilas, K. Mun, "Overview of experimental results for long-term,Large-scale natural circulations in LWR-containments after large LOCAS", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

Revision I5956rl -c7.wpd-04 1404

9.C-149


\Kj

0

gz

I

-- Il 4~L

I I II I IIK- t - 1 1 t -t -1-

I I II * I I I

I I I III I I

I I i I I II I .!III 1 , , I.

D 750 ~ 1095 1185C) TI INUIE 690 770 959 975 1155 1200I 1tl '.. N I ilk

JIUP STr70M 2 \ SN6C SPP8NC2NCI SP8 NC2

I


(reprinted from: L. Wolf, 51. Gavrilas, K. Alun, "Overview of experimental results for long-term,Large-scale natural circulations in LWR-containments after large LOCAS", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-APO7-961D10765", July 1996)

K2-9.C-150 Revision 1

5956rl-c8.spd-041404

------

WCAP-15862APP-SSAR-GSC-588 AP I 000

0

L)

I

0 TIEINMUrEI II I II I II 750 09 105 1185

690 770 959 975 1155 12001* t K N 1K I t }a

ST1 740 GM ST2n \ S;6 | PISP12NCI P

HUP


(reprinted from: L. Wolf, MI. Gavrilas, K. Alun, "Overview of experimental results for long-term,Large-scale natural circulations in LWN'R-containments after large LOCAS", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

Revision I 9.C-151

Revision 45956rl BcS.wpd-041404

9.C-151


'AS

I El 1.2 CTI 01 PICR rs TS 1.02 Eli.2 Co1 M am TS 1.03

I ' i i I* 7.OSM.C

2 -:Lt

U.2 E!2-UzV�-

9W i� '-" IH �

' I I ' 1 L. a______-… __*____ aim-.1-- 0"

I.

I I1 I I I I r�r�

-�…F�1 W�e�

I I II I I

I I I II I I

-- I- - - - - -

I II I I K>

=

Il II III

I " - 4--

I II II I I

0

I I

0 TIE IN MINfEP

IIII 750690 770

1I 11 A

II l 115I959 975 1155 1200

M t I t ItK., , , , v s

IHUP ST(74 M ST2 \S I 5POSP1\

NC1 SPS NC2

Figure 9.C.2-60 HIDR-E11.2 Comparison between temperatures in main and spiral stairvaysat +6 m level

(reprinted from: L. Wolf, M. Gavrilas, K. Mun, "Overview of experimental results for long-term,Large-scale natural circulations in LWR-containments after large LOCAS", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

9.C-152 Revision I5956r -cS.wpd-041404


IME IN MIVMUIS

D- 1200. . 690. v.. . -5. a .-- 959.

975.

30. 60. 90.TENUERATURE IN GRAD C

30. 60. 90.TEMPERATURE IN GRAD C

O

LOCNM o bo x

Figure 9.C.2-61 HDR-EII.2 Axial temperature profiles in both major flow paths at specifictimes, left in spiral stairwvay, right in main stairway

(reprinted from: L. Wolf, M. Gavrilas, K. Mun, "Overview of experimental results for long-term,Large-scale natural circulations in LWNR-containments after large LOCAS", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

Revision I5956r -cS.wpd-041404

9.C-153

-I



Experiment E I 1.3 includes a low-positioned small break and gas release, as well as a closed entrance to thespiral stairway at the upper-deck. Although detailed boundary conditions are not available in the literature,

this case is included to qualitatively indicate the influence of the injection position on the formation of globalcirculation. Even with the main vertical flow path from the top of the spiral staircase to the dome closed,global circulation is generated through other horizontal connections between the compartments (seeFigures 9.C.2-62 and 9.C.2-63).

Temperature differences exist between the vertical axial flow channels during the heatup phase, the gasrelease phase, and the long cooldown period. The axial temperature profiles indicate that the temperaturefield was close to homogeneous with a AT of only 7'C, except for the region below the release, as indicatedby CT8 101 in Figure 9.C.2-64 (see Figures 9.C.2-64 to 9.C.2-68).

V)

9.C-154 Revision I5956rl-9c.wpd-041404

WCAP-1 5862APP-SSAR-GSC-588 --n----

Art uuu )

El 1.3 Main Stairway VELOCITY IN MISSCALE: 0. +.- 3.

tr 431

%_ c" 1_

ZOCAnc* WAU4OT 0331

direction

seuasor failed at 480 min.

TlE IN NMN=

El 1.3 Spiral Stairway VIELOCrTy IN MISSCALE:- 0. +13.

exit to dome dosed!X 432

__,,-, Il

LOAIIWE WM¢ anus

U4IOI

.0`11VI -ou

_=11 . 4.

.. . . . .

* B -

.0 241e'I NMINUIES

Th4E IN MINUMES

. . . .

"as. *roe

j" -% _,� .. __ - - - - - - - -rz vurc 7...h-OL tI1)K-hiJ.3 Velocities in

during containment heatiboth major flow paths at different axial positions

up .

(reprinted from: L. Wolf, M. Gavrilas, K. Mun, "Overview of experimental results for long-term,Large-scale natural circulations in LWVR-containments after large LOCAS", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

D 6 ,:..: 4

5

Revision I5956rl-c8.wpd 041404

9.C-155

I

WCAP-15862APP-SSAR-GSC-588 AP I OOO

E I .3 main surwity VELOCITy LN MISSCALE: 0. +- 2.

Cr 431

(rim GEW1~m.m

MM210

TUV!M IN MNquIES

El 1.3 SpIrai Stalrw yaj.~r IN M/% 2

ezit to doic dosed!

K>~

t2M9 UKI5.TIME IN MINIUrES

138L '44&0

Figure 9.C.2-63 HDR-E1.3 Velocities in both major flow paths at different axial positionsduring H2/He gas mixture release

(reprinted from: L. Wolf, AM. Gavrilas, KI Nun, "Overview of experimental results for long-term,Large-scale natural circulations in LWR-containments after large LOCAS", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

9.C-156 Revision I9.C-156

5956r1 -c8.wApd-04 1404

WCAP-15862APP-SSAR-GSC-588 AP1OO0

U

3

0 run M i;ie 2I880 -

Figure 9.C.2-64 HDR-Ell.3 Comparison between temperatures in main and spiral stainvaysat +31.0 m level

(reprinted from: L. WNolf, Al. Gavrilas, K. Mun, "Overview of experimental results for long-term,Large-scale natural circulations in LWNR-containments after large LOCAS", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

Revision I 9.C-157

Revision I5956r I c8oxpd1041404

9.C-157

WVCAP-15862APP-SSAR-GSC-588 AP I000

IJ

0 rime in Minutes 2880

Figure 9.C.2-65 HDR-E1.3 Comparison between temperatures in main and spiral stairvaysat +26.5 m level

(reprinted from: L. Wolf, Al. Gavrilas, K. Mun, "Overview of experimental results for long-term,Large-scale natural circulations in LWR-containments after large LOCAS", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

VJ

9.C-158 Revision I5956r1-c8,wpd-041404

WCAP-15862I DD C'C! AnD fr'.r COO AP I000Arr-SbanK-u -o o

U

z

w,

'V/

0 me r in Minurcs 2880

Figure 9.C.2-66 HDR-E1l.3 Comparison behveen temperatures in main and spiral stainvaysat +16.5 m level

(reprinted from: L. Wolf, M. Gavrilas, K. Mun,"Overview of experimental results for long-term,Large-scale natural circulations in LWR-containments after large LOCAS", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

Revision I5956r]-c8.wpd-041404

9.U-1 DY

I


gc;

z

KU~

28800 Tmin Minutes

Figure 9.C.2-67 HDR-Ell.3 Comparison between temperatures in main and spiral stairwaysat +6 m level

(reprinted from: L. Wolf, A. Gavrilas, K. MNun, "Overview of experimental results for long-term,Large-scale natural circulations in LWR-containments after large LOCAS", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

9.C-I 60 Revision I

9.C-160 Revision 1595 6r1 -8.upd-41404

WCAP-l 5862APP-SSAR-GSC-588 API000

TIME IN MINUrE

I g3--1470. 0-.,, 1200.* ...... -5. E11.3

CT0438

CT0436

CT0431

CT9302

CT5sol

CT7701

CT6603

CT5301

CT4602

CT3B02

i . i.i: t: : :

9: : :::

Mai S:irw.

60 90.TEMPERATURE IN GRAD C

0. 30. 60. 90. 120.TEMPERATURE IN GRAD C

Figure 9.C.2-68 HDR-E11.3 Axial temperature profiles in both major flow paths at specifictimes, left in spiral stairway, right in main stainvay

(reprinted from: L. Wolf, Al. Gavrilas, K. Mun, "Overview of experimental results for long-term,Large-scale natural circulations in LNNR-containments after large LOCAs", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

K>Revision I5956r1.c9.upd-041404

9.C-161

I



Experiment E 11.4 includes a low-positioned break and an active external spray. The spiral stairway entranceto the dome is open. A detailed presentation of this experiment is given by Wolf, Mun, Floyd, 1995. Datafrom this test is used to validate various numerical containment codes (Wolf et al., 1994, Pisa and Holzbauerand Wolf, 1994, Pisa).

Figure 9.C.2-69 and Table 9.C-5 present the experimental procedure. This includes the duration of heatupand blowdown of steam (p=l 10 bar and T=3180C in the compartment R1405, (see Figure 9.C.2-51 and9.C.2-69) and the consecutive steam and gas injections.

The containment pressure history is'shown in Figure 9.C.2-70. This figure shows the effects of the externalspray application (after 2800 minutes) on the decrease of the containment pressure (by 0.25 bar).

Temperatures at the various axial positions in the main staircase are presented in Figure 9.C.2-7 1. Note thatthe shapes of the temperature curves for various levels are similar, but the values are different. Globalcirculation is indicated by the small temperature differences from the top to the bottom of the dome,AT"2-60 C, and from the dome to the sump, AT=10-26°C. Figures 9.C.2-72 and 9.C.2-73 providesteamandgas concentration histories for the various levels in the main staircase.

During all phases of the experiment, temperature differences exist between the two vertical channels (mainand spiral staircases) enabling the global circulation (see Figures 9.C.2-74 - 9.C.2-77). The temperaturedifferences decrease during the cooldown phase. As in experiment El 1.3, the axial temperature distributionindicates a nearly uniform temperature profile, except below the release point (see Figure 9.C.2-78).

During the El 1.4 test, the release into the lower-level compartment promotes global circulation andhomogenizes the containment atmosphere. In addition, the application of the external sprays causes adecrease in the containment pressure (see Figure 9.C.2-70) and a reduction in the axial temperature gradient(see Figure 9.C.2-71, curves 1-6). The external spray also promotes a more uniform gas concentration (seeFigure 9.C.2-73 from 2800 to 3000 minutes).

9.C-162 Revision I5956rl-9c.%%pd-041404


Table 9.C-5 Events for E1.4 Eperiment

Event Time (minutes) Event Description

HUP 0-2040 RPV-blowdown plus external steam supply with m = 2.06 kg/s

STI 2040-2100 Reduced external steam with m = 1.19 kg/s in R1.405

GMI 2090-2100 H2/He gas mixture release into R1.405 with 0.089 kg/s

ST2 2120-2180 Steam release into R1.405 with 2.06 kg/s

DH 2180-2550 Dry heat addition into Rl.308

GM2 2193-2195 H/He gas mixture release into RI.405 with 0.2 kg/s

GM3 2195-2205 H2/He gas mixture release into RI .405 with 0.089 kg/s

5T3 2205-2505 Start of steam release into Rl.45 with m = 1.19 kg/s

ST4 2550-2620 Steam release into RI.405 with m = 2.06 kg/s

SUB 2620-2800 Sump boiling

NCI 2800-2805 Natural cooldown

SP6 2805-2925 External spray m = 6 kg/s


SPIO 2985-3015 External spray m = 10 kg/s


NC2 3016-3229 Natural cooldown

VEN 3229-3300 Containment venting

END 3360 End of data acquisition

(reprinted from L. Wolf, M. Gavrilas, K. Mun, "Overview of experimental results for long-term, large-scale

natural circulations in LWR-containments after large LOCAS," University of Maryland at College Park,

Final Report for DOE - Project, Order Number DE-AP07-961D 10765")

Revision I 9.C-l 63

Revision I5956r] -9cwspd-041404

9.C-1 63

I


(PSkL20 I

n-.

Wy Matht. R=I.", ..

Htat - up PhasR1405

external atrm

T - 25 'Cp - 11 bar

Andadd * rpdtantow

btawdowa (3 26 mm)T I 3SI'Cp 11- bar

-I

-

-*to

Frr1.1-

1-Z

OUf89 41

a C

4r 4~ :~ a 4 is *4 W on So

t~ ftu

LO

a,

$,

I.

.;

i 'aKto

"I

n_ I _ _ _ _ I.. I I..-.U-

0

33 30; A 1316-I414 i 36EP 1

I 1 36;,P2202040o 2126 219'

2090 21802100 2193

S ST2 S1,T 1 I 11-

5 25052550

26T3 1 S}4

, I.

20

2800 2985 3229 Minutes2805 301I5 3300 Niue

2925 3016 3360Nf~ I l 8 SP12 jN

I "IUPJ1I -

' GMI IGMt2_0 GM3

| SDH

SUB SP6 SPIO NC2NEND

mawbww .u.

Figure 9.C.2-69 HDR-E 11.4 Sequence of Experimental Test Procedures

(reprinted from: L. Wolf, K. Mun, J. Floyd, "HDR hydrogen mixing evaluation for containmentsafety evaluations" - Phase 1, University of Maryland at College Park, Final Report for DOE-ProjectOrder Number: DE-AP07-951D81401, July 1995)

`K>

9.C-164 Revision I5956rl-c9.wpd-041404


Figure 9.C.2-70 IIDR-Ell.4 Containment pressure history

(reprinted from: L. Wolf, K. Mun, J. Floyd, "HDR hydrogen mixing evaluation for containmentsafety evaluations" - Phase 1, University of Maryland at College Park, Final Report for DOE-ProjectOrder Number: DE-AP07-95ID81401, July 1995)

K>Revision I

9.C-l 65Revision I5956rl -6.wpd-041404

9.C-165


.3

49

K2

Figure 9.C.2-71 HDR-E11.4 Atmospheric temperatures along main stairway

(reprinted from: L. Wolf, K. Mun, J. Floyd, "HDR hydrogen mixing evaluation for containmentsafety evaluations" - Phase 1, University of Maryland at College Park, Final Report for DOE-ProjectOrder Number: DE-AP07-95lD81401, July 1995)

9.C-l 66 Revision I

9.C-1 66 Revision I5956rloc9.wpd-041404


Figure 9.C.2-72 HDR-E1I.4 Steam concentrations along main stairway

(reprinted from: L. Wolf, K. Mun, J. Floyd, "IIHDR hydrogen mixing evaluation for containmentsafety evaluations" - Phase 1, University of Maryland at College Park, Final Report for DOE-ProjectOrder Number: DE-AP07-951D81401, July 1995)

Revision I5956ri -c9.wpd-04 1404

9.C-1 67

I

WCAP-15862APP-SSAR-GSC-588 APIOOOAPP-SSAR-GSC-588 API 000

VC 1.00 4 1lO.4e OD 50 IL.

W LOG * EtL

-ccwnonmuam Vc u.ecOuS M cm w 1. I

cxnu ma 1.0 _. _

II I I _

I I t tI I I t ----- ~- --I -c4- - -. -

14 . U...+a~m I 5

I2 H=+27mIH , 3H+17m I a

4 H=+ 6mI 15 H=+ am

a6 H=- 4m m- ---- --- -

I 1 - --- - -- - - -- - -- - - -- - - I--- - -

I I1 I Il

KU

0

Figure 9.C.2-73 HDR-E11.4 H./He Gas concentrations along main stainvay

(reprinted from: L. Wolf, K. Mun, J. Floyd, "HDR hydrogen mixing evaluation for containmentsafety evaluations" - Phase 1, University of Maryland at College Park, Final Report for DOE-ProjectOrder Number: DE-AP07-951D81401, July 1995)

9.C-1 68 Revision I5956rI-c9.Npd-041404


15 1.0 j~l

B

I E11.42 [11.43 E11.4

Cn 432 POR EZMlCnLIol no amElliot no [21a

15 1.00

15 1.00 IEi 1Y_ _ . . .

,5_ .

U :-

S

I I I

I I I I

* I I* I I* I I

I I I

I

II

l

0Th PIE hMN

IFP

I I I I, I 1 s I I2040 2180( 2505 2800 2925 3229

2090 2193 2550 2805 2985 33002100 2195 2620 S 3015 3360,1i 2 Q 2205 ,, ._I 3016,

I -

GMM2 H1 ST4 NCI P2Hpl

Figure 9.C.2-74a HDR-EI1.4 Comparison between temperatures in main and spiral stairwaysat +31.0 m level

(reprinted from: L. Wolf, M. Gavrilas, K. Mun, "Overview of experimental results for long-term,Large-scale natural circulations in LWR-containments after large LOCAs", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

Revision I5956rl-c9.wpd-041404

9.C-169


'K)R

Figure 9.C.2-74b HDR-EI1.4 Comparison between temperatures in main and spiral stairwaysat +31.0 m level


K\J9.C-1 70 Revision I

5956rI-c9.mpd-G41404


'--;--1-

'j n11-1 - ^11-

CM$ __2 11(' 30"66nL~VINT g B

0-


(reprinted from: L. Wolf, Al. Gavrilas, K. Mun, "Overview of experimcntal results for long-term,Large-scale natural circulations in LWR-containments after large LOCAs", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

Ku~Revision I5956r-c9.wpd-041404

9.C-171

I


K\)

z

F-

\K)

20 Vus or ** * * 28( I 29I 3122192043 2N1 8 2505 2800 2925 32290 llsN UTS 2090 2193 2550 2805 2985 3300° 2100 2195 2620 e 3016 3360

GM sa 1 ST4 Ni C

Figure 9.C.2-76 IIDR-E1l.4 Comparison between temperatures in main and spiral stairwaysat +16.5 m level


9.C-172 Revision I

9.C-172 Revision 05956r I -. wpd4D41404

WCAP- 15862APP-SSAR-GSC-588 APIOOO

U

a,

I

I"u

, ,,, .M . IN I I ,I" ,204G 21SO 2505 2800 2925 3229

2090 2193 2550 2805 2985 33002100 2195 2620 3015 3360

S21s0 1Z 20 ,, , SP 30161 ,,

GMr I IGM2 ST3 ST4 NCI 1 P' 2

BHP

. .

Figure 9.C.2-77 HDR-E11.4 Comparison between temperatures in main and spiral stairwaysat +6 m level

(reprinted from: L. Wolf, AM. Gavrilas, K. Alun, "Overview of experimental results for long-term,Large-scale natural circulations in LWR-containments after large LOCAs", University of Marylandat College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

Revision I 9.C-l 73

Revision I5956rl -Osvpd-041404

9.C-173

WCAP-1 5862- - C An C- Q AP1000

TIM IN MDNUMS

o- 3012. 2033. _ 2100.*.. ... -L 2610. E11.4

'1 )MMlMRn IN GRAD C _ EMPERAnlR IN CRAD) C

eto' ¢n _ _oCrmnCCo" W

CMM- |sr r

Figure 9.C.2-78 HDR-E1I.4 Axial temperature profiles in both major flow paths at specifictimes, left in spiral stairway, right in main stairway

(reprinted from: L. Wolf, Al. Gavrilas, K. Mun, "Overview of experimental results for long-term,

Large-scale natural circulations in LWR-containments after large LOCAs", University of Maryland

at College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765", July 1996)

KCVISlOfl I

9.C-174 KeV6S-On 45956rl -c I0.%%-pd-041404


V9.C.2.4.1.6 Experiment E1I.5

Experiment E 11.5 includes a large-break LOCA at a low position, with subsequent small-break releases. The

temperature differences between the vertical flow channels is large enough to obtain global naturalcirculation. The temperature difference changes sign at some axial positions during some phases.

A more detailed presentation of the El 1.5 results will be given in the next chapter, together with the

presentation of the other experiments related to large break LOCA.

Conclusions from the HDR Ell Test Series

Wolf, et al., 1996, concludes that global natural circulation patterns are established in all El I experiments,

despite the different release positions, the applications of internal and external sprays, and the changes to the

flow path geometries. Also, an almost homogeneous temperature field is generated for the low position of

the release points (experiments El 1.3, El 1.4 and El 1.5). Higher release positions generated stratified

temperature fields.

9.C.2.4.2 HDR Large LOCA Experimental Data Base

Experimental results for long-term, large-scale natural circulation in containments after large LOCAs

presented byWolf and Mun, 1996. An assessment of experiments T31.5, V21.1, V43, and El 1.5 isprovided.

Experiment T31.5 simulates a design basis large LOCA and hydrogen release in the upper section of the

containment. Experiment V21.1 simulates a design basis large LOCA in the middle section of the

containment, while experiment El 1.5 simulates the release of steam and hydrogen in the lowest containmentsection. Experiment V21.1 is a subcooled blowdown, while all others use steam.

The influence of the break and hydrogen release locations on the containment atmosphere is established by

comparing the results of these tests.

9.C.2.4.2.1 Experiment T31.5

The results of containment experiment T31.5 performed by Valencia, 1987 and Wenzel et al., 1987 arereported by Wolf and Valencia, 1988 and 1989. Experiment T31.5 includes a short-term, superheated steam

blowdown followed by a long-term, low steam release rate and a H2/He gas mixture injection. Theshort-term steam mass flow release into the room 1704 of the HDR-containment takes 25 seconds (with the

approximate average flow rate of 1200 kg/s - see Figure 9.C.2-79). Room 1704 is at elevation 17.55-25.3 m

in the upper section of the containment, near the spiral staircase (see Figure 9.C.2-50 and 9.C.2-79). The

Revision 1 9.C-1755956ri -9c.wpd-04 1404

I

WCAP-1 5862APP-SSAR-GSC-588 API OO

K Ilarge break position is at +22.3 m (about 8 m bellow the upper deck) and the blowdown pipe is pointingupward.

From 21 to 36 minutes, steam is released at an average mass flow rate 2.3 kg/s (see Figure 9.C.2-80). From36 to 48 minutes, an H2/He gas mixture is released into the containment with an approximate average massflow rate of 0.24 kg/s (see Fig. 9.C.2-80). All the releases can be considered as steam blowdowns into ahigh-positioned compartment in the containment.

The containment pressure, temperature, hydrogen concentration, and velocities histories over 20 hours arepresented in Figures 9.C.2-81 - 9.C.2-85. The two pressure peaks correspond to the periods of steaminjections (see Figure 9.C.2-81).

Temperature histories at various axial positions are presented in Figure 9.C.2-82. The axial temperaturedistribution is uniform above the break position, with the exception of local temperatures measuring thebreak plume. A linearly decreasing temperature (stratification) is present below the break position. Thevertical temperature gradient decreases with time due to the global natural circulation, which develops bothabove and below the break location. The major vertical flow paths are through-compartments in the mainand spiral staircases, where small temperature differences of about 5*C exist at equal heights. Horizontalconnections exist at discrete elevations to provide circulation between the vertical flow paths.

The measured hydrogen concentration distributions at various times are presented in Figure 9.C.2-83. TheH2/He gas mixture concentration increases to a maximum value (13 percent volume, see Figure 9.C.2-84)in the regions above the release position. This concentration remains uniform over several hours in the upperportion of the dome (above 35 m - see Figure 9.C.2-84), while in the lower positions (34, 31, and 25m) itdecreases after reaching maximum values. At the highest position of the dome (49 m), the decrease in themixture concentration starts about 300 minutes after the start of the experiment and continues until about360 minutes. After 360 minutes, the concentration above 12 m is a uniform mixture concentration of10 percent vol. The steady-state profile of the mixture concentration is reached after 900 minutes.

For the first 12 minutes, upward and downward flows are present in the spiral and main staircases,respectively. The measured velocities are presented in Figure 9.C.2-85. The velocities are recorded againafter the start of the steam injection (after 21 minute). Staircase flow starts again at approximately200 minutes, at the same time that the mixture concentration in the upper portion of the containment startsto decrease. It then decreases again around 350 minutes, when upper region reaches a uniform concentration.

9.C-176 Revision I5956r] -9c.wpd-041404


Interpretation of Data

Even for the case of the high release point with an initially stratified distribution, the final gas concentration

through most of the containment is nearly homogenized due to the global natural circulation effects. Only

a small, steady stratification gradient exists (about I OC overthe height of dome, about 180C over25 mbelowthe break, gradient of about 7.5 to 8.8 percent H2/He over the height of dome).

Applicability to the AP600 and AP1000

Since the AP600 and AP1000 have significant cooling over the containment dome (which somewhat

corresponds to the HDR dome region), additional driving forces for circulation above the operating deck

exist.

Revision I 9.C-l 77

Revision I5956rl -9c.svpd4041404

9.C-177


.00TnM AFTER BREAK MN SEC

Figure 9.C.2-79 IIDR-T31.5 Blowdown mass flow rate into the containment

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V43,T31.5 and El1.5", University of Maryland at College Park, for DOE- ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

9.C-178 Revision I5956r1-c10.,pd-0414D4


TIME AFTER BREAK IN MINUTES

Figure 9.C.2-80 HDR-T31.5 Steam and H2/He gas mass flow rates into the containment

(reprinted from: L. Wolf, K. Alun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. 11: Assessment of HDRExperiments V2 1.1, *'43, T31.5 and El 1.5", University of Maryland at College Park, for DOE - ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

K>Revision I5956r cI0.vwpd-041404

9.C-1 79


IzWWD06W.

1�

Ti AFTER BREAK IN MIN * 10'

Figure 9.C.2-81 IIDR-T31.5 Pressure history after large LOCA over 1200 minutes

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V43,T31.5 and El l.5", University of Maryland at College Park, forDOE- ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

V>

9.C-I 80 Revision 1

9.C-180 Revision I5956rl -c I0.%%-d-041404

WCAP-1 5862APP-SSAR-GSC-588 AP I 000

100-

C.)

50'

100.

C.)z

150

9! .

T31.5

200 On

1100 Min

C.)

zw

I

T31.5

D . . . . . . . . . . . ..O 2500

HEIGHT IN CM5000

Figure 9.C.2-82 HDR-T31.5 Axial temperature profiles along spiral stairway (left) and mainstairway (right) for different instants in time

(reprinted from: L. Wolf, K. Mlun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. 11: Assessment of HDRExperiments V'21.1, V43, T31.5 and El1.5", University of Alaryland at College Park, for DOE - ProjectHDR Hydrogen Alixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-96ID10765", April 1996)

K>Revision I5956rl -0c O.pd.041404

9.C-181


I.-

0CC Ia.tu

-I0)

T31.5n60 M

40 li

,, _

. .

u-- - - - -

0 2500HEIGHT IN CM

5000

2

i

us= 10am-U1

-j0

T31.D

200 Mmn

360 Min

1100 Mn

2-

0a.zU)

-J0)

200 Min

I)

l n ......%, o 2500HEIGHT IN CMA

5000 0 2500HEIGHT IN CM

5000

Figure 9.C.2-83 HDR-T31.5 Axial gas concentration profiles along spiral stairway (left) andmain stairivay (right) in upper containment region at different instants in time

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. 11: Assessment of HDRExperiments V21.1,V43,T31.5 and El1.5", UniversityoflMaryland at CollegePark, for DOE- ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D310765", April 1996)

9.C-I 82 Revision I

9.C-1 82 Revision I5956rl Tc I 0.%Npd-04 1404


I T31.s2 T1.55 T31.5

C1A431 PMR EZAWWCC432 M EZAW

CC1433 PHDR EZW

YU 10.00 4 31.5Vii 10.00 5 31.5VY 10.00 6 T31.5

CC1434 PHDR ZW YuCC1435 PNREZAW Q WC01436 PH)R MW VU

.

8

C._o

0

C4_0.U

CD

I---- 4-

____L_1.

.1---- 4-

I.

to.

I I I I I I I I I-I I I I I I I I I

I I I I I I I II

I I I I I I I I . I

I I - I - - - - I

or

(.4

1-1.20 .00 1.20 2.40 3.60 4.80 6.00 7.20 8.40

ETM AFTER BREAK IN MIN * 102

9.60 10.80 12.00

Figure 9.C.2-84 HDR-T31.5 Gas concentrations inside the containment dome

(reprinted from: L. Wolf, K. Mlun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V2 1.1, V43, T31 .5 and El1.5", University ofAlaryland at College Park, for DOE - ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

Revision I5956rl -c I 0.u-pd-04 1404

9.C-1 83


MlIaS CWa 1ZW cuding *a 4IX$ u.IM wSu 1m u I *IZILS *UaSflglA 0l03IM MOO 513.1 UPI(412AV WINW313,1 .ffruila123u 00u5W *a

IisI I I I I I I i s

8 __ I\4-q----F- I ------ >- I __

…. _ I I, , I a, .

a a a a * a a a aB f I a a a m a a as a m m a a * a aa a a m a a a a a

…a a a a a a a , a…I a a I a . I a ar S a a a i a . a8 a a a a a a a .3 3 I

. I I I I I 4-I-I-I-i-

'J)

-LO0 .00 &00 ILO0 I380 24.00 X00TIUE AFrER BREAK IN MIN

.00 42.0 4M 5400 Sao

Figure 9.C.2-85a HDR-T31.5 Velocities over 60 minutes along spiral stairways and associateddome regions

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. 11: Assessment of HDRExperiments V21.1,V43,T31.5 and El l.511,University of Alaryland at College Park, for DOE-ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

K)

9.C-1 84 Revision I5956rl-c10.vpd-041404


4

I llt. a46M Z $1 1rw xnol rJweoo 413. IJW t l7? Qrtt $W TF 1 oa31113 U410111Zlt GUa00V 0000O~t 1UIUW 030f8 a llis * 4aI u ANt tw arz tS ti S UI. U 1U2 r? Ert IA 0110f 14to

31315

Ut2P. W z ueINO titc

2 ~ i ' j j-- '- lla -' - L I I ., , r I

8 -i i---i…-t-i-

-- --- I- ----------- --- ---

* _ _ a _ 1 i a a* o a a, a e a. a. !. ! : a .

.w ..w IZ.Mu ID.W 0.00 ,

TIME AFTER BREAK It MIN300 42.00 48.00 54.00 6000

Figure 9.C.2-85b HDR-T31.5 Velocities over 60 minutes along main stainvays and associateddome regions

(reprinted from: L. %Nolf, K. Alun, "Overview of experimental results for long-term, Large-scalenatural circulations in LUWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V43, T31.5 and El 1.5", University of Maryland at College Park, for DOE - ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

K>Revision I5956r] -cIO.wpd-041404

9.C-1 85

II


9.C.2.4.2.2 Experiment V21.1

The blowdown for the V2 1.1 experiment (see Kanzleiter T., Valencia, L., 1984) is initiated inside the RI 603compartment, which is in the middle section of the HDR containment (see Figure 9.C.2-50 and 2-86). Thearrangement of reactorpressure vessel, blowdown pipe (break diameter453 mm), andjet impingement plate

for the subcooled water blowdown is shown in Figure 9.C.2-86. Other characteristics of the experiment are

given in Table 9.C-6. The total mass and enthalpy of the exiting fluid are 50157 kg and 72320 MJ,respectively. The test could be characterized as a large break cross-section and a short-term blowdown

(30 seconds). The maximum peak pressure was 3 bar because of the much higher mass and energy releases

during the steam/liquid mixture blowdown (Figure 9.C.2-87). The transient atmospheric temperature for the

dome, main and spiral stairway are presented in the Figures 9.C.2-88, 9.C.2-89, and 9.C.2-90.

At the beginning ofthe V2 1.1 experiment, the blowdown results in fairly uniform distribution of temperaturein the dome (Figure 9.C.2-88). After about 10 seconds, a stratification gradient exists. The maximumtemperature gradient over the above-deck height is 10°C at about 100 minutes. Until 420 minutes, thethermal stratification inside the dome volume exists (see Figure 9.C.2-88). At the end of the break release

(after 30 seconds) the temperatures inside the dome are decreasing (slower near the top). After 420 minutes,

temperature differences inside the dome volume are only around 2-3 'IC, indicating the presence of natural

circulation. The natural circulation in the above-deck region acts to reduce the gradient with time.

The axial temperature distributions in the mean and spiral staircases are presented in Figures 9.C.2-89 and9.C.2-90. Immediately after the end of the blowdown, a stratification gradient develops. The transient

histories of the temperatures at the same axial levels (in the spiral and main staircases) are presented in

Figure 9.C.2-9 1. The large blowdown injected into the middle of the both staircases results in almost equaltemperatures in both flow channels, therefore, global natural convection can not be formed. A slightly higher

temperature is recorded in the main staircase. A temperature difference of 1-2°C starts at 200 minutes and

continues until 800 minutes.

Interpretation of Data

The large break with the blowdown penetrating into the middle of both staircases (mean flow paths)

minimizes temperature differences among the same axial positions and suppresses the development of thenatural circulation.

The small temperature differences of about 1 °C (Figure 9.C.2-90) and the slight global natural circulationappear to be established later due to the different heat release rates of the various wall structures. Thesestructures are heated to various temperatures during the early phase of the transient.

9.C-186 Revision I5956rl -9c.upd0441404


The large LOCA HDR tests described should be carefully applied to the passive containment design. The

majority of the passive containment compartments are at the lower levels of the containment. Heating of thestructures (walls, floors, and equipment) in the middle of the containment height or above is not possible

because they do not exist. The formation of the stable stratified layers in the upper portion of the passive

containment due to the accumulated heat (in fact the higher temperatures) in the concrete and steel structuresis not possible. The temperature of the containment shell, which is the only solid structure in the upper

portion of the containment, tends to be lower than the internal containment temperature. This promotesinternal global circulation.

Application to AP600 and APIOOO

Small temperature differences inside the dome, even without dome cooling or additional releases after

blowdown, are sufficient to drive natural convection. This leads to more homogenous conditions over time.

Revision I 9.C-l 87

Revision I5956rl -9c.%%-pd0441404

9.C-1 87

-I


Table 9.C-6 HDR-V21.1 Test Conditions

MEASURED INITIAL CONDITIONS INREACTOR PRESSURE VESSEL -

Pressure 110.4 bar

Temperature 315 -319 0 C

Liquid Level filled

Liquid volume prior to blowdown 74.5 m3

Liquid Mass prior to blowdown 50233 kg

Liquid level in reactor pressure vessel after blowdown empty

Steam volume after blowdown 74.5 me

Steam mass after blowdown 76 kg

Mass of exiting fluid 50157 kg

Enthalpy of exiting fluid 72320 MJ

Jet impingement plate distance (LD) 2

Jet impingement plate inclination angle 00

Break nozzle diameter 453 mm

Break nozzle length 6091 mm

VENT FLOW OPENINGS FROM BREAK

SUBCOMPARTMENT

U0140 open

U0143 closed

U0162 6.1 m2

U0177 open

U0178 closed

Extemal spray off (all the time)

K>i

(reprinted from: L Wolf, K. Mfun, "Overview of experimental results for long-term, Large-scale natural circulations In LWR-containments after large LOCAs, Vol. 11: Assessment of tiDR Experiments V21.1, V43, T315 and El l.5", University of Marylandat College Park, for DOE - Project IIDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural GlobalCirculation. Order Number- DF-AP07-96iDn 76O5"- April 19961

9.C-l 88 Revision I

9.C-188 Revision I'5956rl-9c.wpd-041404

I


Zulaufvom VKL

In

zum Versuchskreislauf(VKL)

Figure 9.C.2-86 HDR-V21.1 Arrangement of reactor pressure vessel, blowdown pipe and jetimpingement plate for water blowdown test

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments :after large LOCAs, Vol. 11: Assessment of HDRExperiments V2 1.1, V43, T3 1.5 and El 1.5", University ofMaryland at College Park, for DOE - ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

Revision I5956rlcI0.upd-041404

9.C-1 89

WCAP-1 5862APP-SSAR-GSC-588

I V 211 Ca62I Pi2 Y 43 am2IR P

83 rT4 CPM2

P1

z ~I2I

4C4EU

AP1000

KJ

TIME AFTER BREAK IN MINUTE * 102

Figure 9.C.2-87 IIDR-V21.1 Containment pressure history

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWVR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1, V43,T31.5 and El 1.5", University of Maryland at College Park, forDOE-ProjectIIDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D 10765", April 1996)

K>~

9.C-190 Revision I5956r1 -c i O.pd-0414 04


0

u

U

UzH

W

aEE-

IV 211 aT402 Ezut IS 2.O0 4V 211 CIJ05 [ZU 1S 2i2V 211 Cr403 EZM IS 2.00 SY 211 CT It! EZ1A iS 2*00 CONW3Y 211 CTl4m Ena iS 200

I I

------- -- r- -------.3 CT 403

. 1-

I CT 404

CT 405. ne

CDICDC 0

… 1---C 40

II

.W 4.UU *U.w 12.00

TIME AFTER BREAK IN MINUTES *102

Figure 9.C.2-88 HDR-V21.1 Temperatures in the containment dome

(reprinted from: L Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V43, T31.5 and El 1.5", University of Maryland at College Park,forDOE -ProjectHDR Hydrogen Mixing Evaluation for Containme'nt Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

Revision I 9.C-191

Revision I5956rl -c10 .wpd-041404

9.C-191

I

WCAP-I 5862APP-SSAR-GSC-588 AP1000-AOG

0ro

ri

2

'-4

I~z2

W .

TIME AFTER BREAK IN MINUTES .io2

Figure 9.C.2-89 HDR-V21.1 Temperature along the main staircase at different axial positions

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1, V43,T31.5 and El 1.5", University of Maryland at College Park, forDOE - ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

K\ >9.C-I 92

Revision I9.C-192 Revision 1

5956r] - I .mpd-041404

WCAP-1 5862APIOOOAPP-SSAR~jL-.SC-5S

0

p4

U

14

4.00 8.0TIME AFTER BREAK IN MINUTES .io2

Figurc 9.C.2-90 HDR-V21.1 Tcmperatures along spiral staircase at differcnt axial positions

(reprinted from: L. Wolf, K. Mun, "Overview of cxperimental results for long-term, Large-scalcnatural circulations in LWVR-containments after large LOCAs, Vol. 11: Assessment of HDRExperiments V21.1,V43, T31.5 and El 1.5", University of Mlaryland at College Park, forDOE -ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

Revision 1 9.C-1935956rl-c I .vpd-041404

KU

I


K-)

0

a'-4

adKY

TIME AFTER BREAK IN MINUTES *1.02

Figure 9.C.2-91 HDR-V21.1 Comparison betveen temperatures in spiral staircase (H=22.5 m)and main staircase (H=25 m)

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V43,T31.5 and El1.5", University of Maryland at College Park, for DOE-ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

V>9.C-194


5956rlc I 1.wpd-041404


9.C.2.4.23 Experiment E11.5

Experiment El 1.5 is performed with a large break LOCA at the lowest possible position (room R1405) with

subsequent small-break releases (see Fig. 9.C.2-92). An overview of experimental results is presented byL. Wolf, K. Mun, 1996.

The containment pressure during the large break heatup phase is shown in Figure 9.C.2-93. At the end of

the initial steam blowdown (50 seconds), the pressure reaches the peak of 2.56 bars (see Figure 9.C.2-93),

then continuously decreases due to the condensation at the cold surfaces. Note that the external dome

cooling is not applied during this experiment.

The 15-hour (900 minute) containment pressure history is presented in Figure 9.C.2-94. The gas release and

steam injection following the initial blowdown slows the depressurization (see Figure 9.C.2-94). Over the

time intervals 155-219 minutes, 245-545 minutes, and 695-820 minutes, the pressure slowly increases inside

the containment due to the additional gas and steam releases. During the sump heating phase, the pressurefirst decreases and then increases. The increase in pressure coincides with the moment when the sumpreaches saturated conditions and starts to evaporate. After reaching a second peak value of 1.6 bar at the end

of the steam release into the subcompartment R1405, containment pressure decreases to atmosphericpressure. The containment depressurizes through the vent lines at the 0 m and 10 m elevations.

Temperatures along the main stairway during the heatup phase are presented in Figure 9.C.2-95. The

combination of steam and gas releases during the heatup phase results in thermal stratification. This is

shown by a 250C temperature differential over the height of the containment after 80 minutes and a 120Cdifferential from top to bottom of the dome - see Figure 9.C.2-95.

Figure 9.C.2-96 presents the temperatures in both major flow paths (main and spiral staircases) at the samethree axial positions. The temperature differences are relatively constant through time at each elevation,

although smaller at the high elevations. Circulation is sufficient to keep the temperature difference between

the two chimneys small (e.g., AT is approximately 5°C at the +16.5 elevation), for the various steam, dry

heat, and sump boiling releases that occur in the lower portion of the containment.

The temperature distributions for the entire containment are presented in Figures 9.C.2-97 and 9.C.2-98 overtime intervals 2.0 to 75.0 minutes and 123 to 345 minutes, respectively. After the initial rise, the

temperatures decrease quickly during the first 25 minutes (Figure 9.C.2-97). The decrease then continuesat a slower rate (see Figure 9.C.2-98). The temperatures stratify in the main staircase, while in the spiral

staircase they are almost uniform. Temperature differences between the top of the containment and the

release level are approximately 10°C in the main staircase and 5 °C in the spiral staircase. The temperature

Revision I 9.C-1 955956rl-9c.wpd-041404

---------- I-


difference below the main stair case, between the level of the release and the bottom of the containment(betveen CT5301 and CT3802) is around 15 0C.

The temperature decreases slowly in the upper region of the dome. The effects of the global circulation onthe upper region of the dome are noticeable after the end of the steam release. This is evident from acomparison of the H2/He concentrations during the gas mixture release (Figure 9.C.2-98). After the release,the temperature in the containment is relatively uniform, except for below the releases (see temperaturedistribution on the main stairway side).

During the sump heatup phase (but before the start of the boiling, 570 - 655 minutes), a convectioncirculation loop affects the region up to 15 m. This region has a uniform temperature gradient, while theregion above is stagnant. After sump evaporation begins, the convection circulation loop affects the entirevolume and homogenizes the temperatures.

The steam and gas concentrations over time correspond to the temperature distributions. Note that there areno large local changes in gas concentrations in the regions (radially) away from the two staircases, see

Figures 9.C.2-99 and 9.C.2-100.

The flow patterns in the compartments may be inferred by comparing the velocities in the two staircases (seeFigures 9.C.2-101, 9.C.2-102, and 9.C.2-103). Upward flow is recorded inboth stairways at the early stageof the blowdown (first 30 seconds, see Figure 9.C.2-101). The upward flow in the spiral staircase is theresult of the unplanned vent connection with the release point. A pressure wave destroyed the separatingwall during the blowdown. This indicates that the break compartment pressurization forced flow out of allthe openings in the break compartment.

After 80 seconds (end of blowdown is at 50 seconds), upward and downward flows are recorded in the mainand spiral staircases, respectively (see velocities in Figure 9.C.2-1 01). During the steam releases ST 1, ST2,ST3, ST4, and ST5 (see Figures 9.C.2-103 and 9.C.2-104), upward flows are recorded in both staircases.During the gas mixture release phases GT I, GT2, and GT3, the flow is downwards in the spiral staircase (seeFigure 9.C.2-103).

During the sump heatup phase the flow is upwards in the main staircase and downwards in the spiral staircase(see Figure 9.C.2-104). The weak circulation loop reaches the 15 m level before the start of sump boiling.After the start of sump boiling, the circulation loop extends over the total containment height and the flowvelocities are higher (see Figure 9.C.2-104 for 655 - 695 minutes).

9.C-196 Revision 15956r] -9c.Tpd-041404


Various global circulation flow patterns are established during the E I .5 experiment as a result ofthe various

steam and gas release phases, as well as the sump heatup and boiling phases. The temperatures, steam, and

gas concentrations follow similar distributions. Global circulation effects contribute to reduced gradients

in the vertical direction of temperature and concentration fields.

Application to AP600 and AP1000

The release position in experiment El 1.5 is the closest to the possible AP600 and API000 LOCA cases.

However, in the HDR, chimney and heat storage effects (heat sink and source effects) along the two main

flow paths cause dominant global circulation flow patterns below the operating deck. After the blowdown,

the plume rises from the break compartment and entrains laterally, yielding global circulation.

In the passive containment, upward flow patterns are formed in the core of containment (depending on the

break location). Downward flow is present near the vertical walls due to the relatively high cooling rates ofthe steel shell. The overall effects are shallowervertical density gradients in the above-deck region compared

with El 1.5 test. The shallow gradients are the result of the additional wall layer entrainment that is not

present in HDR.

K>

Reiso . 9.-l

Revision I5956rl -9c.wpd-04 1 40

9.C-197

II

WCAP-15862APP-SSAR-GSC-588 A Tb AA

Ar I UUJ

I. . j

-. 9

-. i I- 1I 11. .[ i* ~ 125 219

0 1 75 155

HUP ST ST3

11I9 24527228 GM3

GM2

- I I -- I if II I I II545-. 820

570 695 826 890

JS&B I ST5 II EN

finutes

-ST4

,DH

900

,

Figure 9.C.2-92a IHDR-E1 1.5 Sequence of Experimental Test Procedures

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperimentsV21.1,V43,T31.5and El1.5", University of Maryland at CollegePark, forDOE- ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

l)

9.C-198 Revision I5956rl1-cI Ibpd.041404


Event

BDHlP STIST2

0MI

ST3

DHGM2

GM3

ST4

SUBST5

VENSD -

Timominutes

0-11.7575-135

125-155

155-219

2 19-570227-228

22S-245

245-545

5704695695-820

S26-890900

Event Description

Blowdown in R1.405Steam rupply with rn - 2.06 kg/sReduced extena- steam withrn - 1.18 kg/s in RI.405Heleo gas mixture release intoRl.405 with 0.089 kg/sSteam release into Rl.405with 2.06 kg/sDry beat addition into Rl.30SIHVe gs mixture release intoRl.405 with 0.2 kg/sHAHo gas mixture release intoRl.405 with 0.089 4/sStanl of steam rekaso intoRIA05 with rn - 1.8 k/JsSump boilingSteam rleae into RIA05with m - 2.06 kg/sContainment nenSngEnd of daft acquisition

'1Ril

Figure 9.C.2-92b HDR-E11.5 Sequence of Experimental Test Procedures

(reprinted from: L. Wolf, K. Alun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21 .1,V43, T31.5 and El 1.5",;University of Aaryland at College Park, forDOE - ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

K'Revision I5956rl1 - L.vvpd-041404

9.C-199


.t

40.00

TIME IN MINIiES

Figure 9.C.2-93 HDR-E1 1.5 Containment pressure

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. 11: Assessment of HDRExperimentsV21.1, V43,T31.5 and El 1.5", University of Maryland at CollegePark, for DOE-ProjectIlDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-APO7-961D10765", April 1996)

9.C-200 Revision I5956r-clI I.vpd-041404


. L.S VP I PtHD EZIA 10/21 PS t.72*..t11.5 -VP 2 t EVzA 12/io po 9.72 |_E I 1p

.1!

im

Z o

z

9! =

W -

U

rnrnZi on

0w

op0

CO

yd _,

z .Z.0

W

y; o

9 _,

.. -

1.-

p 9 _-90.00. * .00 90.00 180.00 270.00 360.00 450.00

125 219 2450 135 227 IUMEINMINMIE

1 75 -.22S

540.00

545570

630.00 720.00 80.00 900.00

820 890695 826 900

- - -. X. - I I X II - -II

BD Sri SlZ GMI N \M2HUP D

'S14 'II StlB sT5 *- VEN END

I

Figure 9.C.2-94 HDR-EII.5 Containment pressure history

(reprinted from: L. Wolf, K. Mlun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1, V43, T31.5 and El l.5", University of Mlaryland at College Park, forDOE - ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

Revision I5956r -c1 I.vpd-041404

9.C-201

I


N>

0

*Z

PKnATow6 of nlERMOaOM

.00

TIME IN MINUTES

Figure 9.C.2-95 HDR-Ell.5 Containment temperatures along main stairway duringcontainment heatup phase

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LW1R-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V43,T31.5and El 1.5", University of Maryland at CollegePark, for DOE- ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

9.C-202 Revision I

9.C-202 Revision I5956rI c1 I Dupd-041404


11.3 01.V.*.11 c- _

2 t9 1.1 013M rW [21

IS r.r0e

Is *.00 |E1 Ig

un 80 _

z o6

r<-r-

P8r- -

_r r -----

26. I

I- - - - - - - - - - - - - - - - - -

27r ' or

LOnATIO OF MMMxOUME

o:

125 219 245 55 65 " 0

I - U4 . I I I iD

BD ST) m7 )GNUHUP I D

Figure 9.C.2-96a HDR-EI1I.5 Temperatures in both major flow paths at the same axial positionsfor three heights in the containment

(reprinted from: L. W\olf, K. Alun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V721.1;V43,T31.5 and El l.5"1,University of Nlaiyland at College Parlc,for DOE -ProjectHDR Hy drogen Mlixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765"1, April 1996) I a

Revision I5956rl -01.%-d-41404

9.C-203

I

WCAP-15862APP-SSAR-GSC-588 AP1000APP-SSAR-GSC-588 API 000

U2

C z3

C-v

2M 6 W

LOCATIOe OU 1 7tMOMOM"

125 219 245 5452 7 35 2272 Tom i D4= IJ M

I' . ,*t5 122

BD STI ST2 GMI N 3 1\ 570HUP I DH

320 890*26 900

Figure 9.C.2-96b HDR-E1 1.5 Temperatures in both major flow paths at the same axial positionsfor three heights in the containment

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V413,T31.5 and El 1.5", University of Maryland at College Park, for DOE- ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

KJd9.C-204

Revision 19.C-204 Revision I

5956rl-c I ].%%Tpd-041404

WCAP-15862APP-SSAR-GSC-588 An I fk^

_ --r I uJV)

nblowo co~ffCIO43O-' /-CT0431

U1202 - -CS

CnMOI __nClI

n _ ., 101-_* 13 n01 l v a~

n~s-n nd=

SE f

C.)

z

-c

F r

LOCAT1OE OF rnMMOCOMn

<2 I I .

125 219 245 5450° 75 135 227 Th&flN D s7c

'DH

t29 . *90926 900

.. | I

stml

Figure 9.C.2-96c IHDR-EI 1.5 Temperatures in both major flow paths at the same axial positionsfor three heights in the containment

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. 11: Assessment of HDRExperiments V21.1, V43, T31.5 and Eli .5", University of Maryland at College Park, forDOE -Project

HDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996) -

Revision I5956r -0c I.%vpd-041404

9.C-205

I


_~~ ~ - - -

'

Time in Minutes

r __ 75.00 oo -2.00 O -_

2.00 _25.00

50.00

. rCL8430

CL8435

CL8432

CL8092

CLt8084

CL8078

CL81 66

CLSO0 I

CL8043

:. *. *

: I :

* I

*.:

i.6

:t

T. ) .

*t

.+ | t *t.d82>

CL8438

CL8436

CL8431

CL8093

CL8085

CLS8077

CL8066

CL8053

.* CL8038

60.0 80.0 100.0

I

I C -- t "

t_ t . a"

. f. .m a"

§ I_ a .I

I

I

I | . I . .

. , . I

.0 20.0 40.0

STEAM VOL. %

.0 20.0 40.0 60.0 80.0 100.0

STEAM VOL. %

Figure 9.C.2-97a IIDR-E 11.5 Axial steam concentration profiles in spiral (left) and main (right)stairways for different instants in time for large break LOCA release in lowercontainment region

(reprinted from: L. Wolf, K. .Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V43,T31.5 and El1.5", University of Mlaryland at College Park, forDOE- ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

9.C-206 Revision I

9.C-206 Revision I5956r]lc I1.wpd-041404


TME IN MINUTES

0 - 75.00 a _____...... . -2.00, -

2.00 . ____25.00

50.00

CT0430

CTr0435

CT0432

CT9202

CT8401

CT7801

CT6604

CT1101

* if

* p-

I *r- Ii

CT0438

CT0436

Cr0431

CTr9302

CT8501

C 7701

Cr6603

Cr5301

CT3802

..

o - I

II --- CrMI

I _ . 11M

lb '

6I*.

H.

lb -I

Id:

II1I47

,

.0 30.0 60.0 s0.0 120.0TEMPERATURE IN GRAD C

.o 30.0 60.0 90.0 120.0lEMPERATlRE IN GRAD C

Figure 9.C.2-97b HDR-E11.5 Axial temperature profiles in spiral (left) and main (right)stairways for different instants in time for large break LOCA release in lowercontainment region

(reprinted from: L. Wolf, K. Alun, "Overview' of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. 11: Assessment of HDRExperiments V21.1,V43,T31.5 and El1.5", University of Maryland at College Park, forDOE -ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

Revision I5956rl c1 .wpd-041404

9.C-207

WCAP-I 5862APP-SSAR-GSC-588 AP I 000

_ ,, ,,, 122.00 __Z45.00 .154.00

225.00

CC14)0

CC143S

CC1432

CC1Oon

CC 1084

CC1078

C1 1 6s

C00101

Cro104

-********1*********�1�****�** Y - I - .7

IT* ./1 *

; 4* +* .'t.

' I * I *

* 4* 4?* . II* 4: 4

us:'4

* .1±

.0 5.0 10.0 55.0 20.0 25.0

HrHe CONCENTRATION VOL % HrHc CONCENTRATION VOL %

`Ky/

Figure 9.C.2-98a HDR-E1I.5 Axial H2 -He gas mixture concentration profiles along spiral (left)and main (right) stairways during gas release phase (total mass 291 kg)

(reprinted from: L. Wolf, K Mun, "Overview of experimental results for long-term, Large-scale naturalcirculations in LWR-containments after large LOCAs, VoL II: Assessment ofHDR Experiments V21.1,V43, T31.5 and E11.5", University of Maryland at College Park, for DOE - Project HDR HydrogenMixing Evaluation for Containment Safety Evaluations Natural Global Circulation, Order Number:DE-AP07-96ID10765", April 1996)

K29.C-208 Revision I

5956r1 -cI2.upd-041404


TIME IN NUN245.00 . - - - 34500 '-

a -= 12Z.00 . --. 554.00

Figure 9.C.2-98b HDR-Ell.5 Axial steam concentration profiles along spiral (left) and main(right) stair ways during gas release phase (total mass 291 kg)

(reprinted from: L. Wolf, K. Mlun, "Over view of experimental results for long-term, Large-scalenatural circulations in LWVR-containrments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V43, T31.5 and E 11.5",University of Maryland at College Park, for DOE - ProjectHDR Hydrogen Mlixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

Revision I5956rl-cI2.w pd-041404

9.C-209


TIME I MNe- 245.00 - 345.00 _ -,. _ 123.00 .. ___, 154.00

22,.00

CT0430

CT0435

CT0432

CT9202

CT8401

CT7001

CTseo

CT1 101

. *. .

00.0 60 o900e 120.0

TEMPERATURE IN GRAD C

CT0436

CT04J3 *

CT0431 .'

C793<02 . *

CT8501 . .

CT770t . , .

CTS 3 ..

crsa *.

CT3602

.0 30.0 60.0 s0.0 120.0

TEMPERATURE IN GRAD C

LoucA1o or nHInM0natnu

Figure 9.C.2-98c H1DR-E11.5 Axial temperature profiles along spiral (left) and main (right)stainvays during gas release phase (total mass 291 kg)

(reprinted from: L. Wolf, K. Alun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1, V43,T31.5 and El1.5", University of Maryland at CollegePark, forDOE- ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

KJ9.C-2 10


5956rl-c12.,Apd 041404

WCAP-15862APP-SSAR-GSC-588 - - API000

0<3

z0

e80

V.

8D00L 360Q00 480.00

llME IN MlNUTE

&NMM 54570 6950

125

7S 135I751

219 245227

2281. I

TMMERIN 820 890S 826 900

iST 1 Vi "D', , , . , , . , , .

BD Sr1 ST2 GMI S3GM2HUP I

\ST4I I

SbB-r

NDH

Figure 9.C.2-99a HDR-EI1.5 Gas concentrations in the major flow paths and in regions andsubcompartments away from the stairways (H=22m up to 27 m)

(reprinted from: L. Wolf, K. Alun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V43,T31.5 and El1.5", University of Aaryiand at College Park, for DOE -ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

Revision I 9.C-21 1

Revision I5956r] -c I2.wpd-041404

9.C-211

WCAP-15862APP-SSAR-GSC-588 A n --.

_Al'IUW

-i0

z0

zU3z0LiUX

K)~

0

s-BD

125 219 245135 227

STI S GMIHUP

TMT IN M1NUrE 545

"ST4 r- -- I

695-t-

820 890826 900

'45 1 DsbB

'DH

rigure Y.L.L-99b HDR-EI1.5 Gas concentrations in the major flow paths and in regions andsubcompartments away from the stainvays (H=22m up to 27 m)

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for. long-term, Large-scalenatural circulations' in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V43,T31.5 and El1.5", University of MIaryland at College Park, forDOE- ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

K\Jn el f-'

7.-X-Zl I Revision I5956r1-c12.wpd-041404


>0

z0Uz0X

TIME IN MNUE

0

BD

125 219 245135 227

7S H -5 228

STl ST2 GM I b tGM2

AE Di WNUTE 545570

820 890695 826 900

'ST4

-"'DH

I I StJB SO " VEN 'ND

Figure 9.C.2-lOOa HDR-E11.5 Gas concentrations in the major flow paths and in regions andsubcompartments away from the stairways (H=4.5m up to +2.0 m)

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containmeits after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1, V43, T31.5 and El1.5", Uniiersity of lairyland at CollegePark, for DOE- ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

K>Revision I5956r -c12.wpd-041404

9.C-213

WCAP-I 5862APP-SSAR-GSC-588 API000

J

-J

z0

zE

z0

C41X

.00 120.00 240.00 360.00 480.00 600.00 7125 219 245

0 135 227 MM dINMNnA 545r1 75 228 70 695

BD STi ST2 GM1I T GM2 ST4 |HUP I -

820 890826 900

XS VIN D

UH

Figure 9.C.2-lOOb HDR-El1.5 Gas concentrations in the major flow paths and in regions andsubcompartments away from the stainvays (H-4.5m up to +2.0 m)

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations. in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V43,T31.5 and El1.5", University of Mlaryland at College Park, for DOE- ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-APO7-961D10765", April 1996)

V29.C-2 14


595&rlvc2.vbpd-041404

WCAP-1 5862APP-SSAR-GSC-588 AP1000-- --

Ca"1o" IN &.~OM

TIME AFTER BREAK IN SECONDS

Figure 9.C.2-lOla HDR-E11.5 Velocities in the main stairway during blowdown

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in-LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1,V%43, T31.5 and El 1.5", University of Maryland at College Park, for DOE - ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

Revision I 9.C-21 5

Revision I5956rl c I2.wpd4041404

9.C-215

I

WCAP-15862APP-SSAR-GSC-588 --- AAAAP IUUU

Spiral StairwayVELOCIY IN M1SSCALE: 0. +/- 20.

CF 432

._

eW ffi W;

i0M

_ , _s

OAMN OF AMOU

CF9202

CF8401

Cr7801

CF6604

XXI 101Sensor Failure

Sensor Failurexx4301

0 20.0 40.0 60.0TIME AFTER BREAK IN SECONDS

80o

11.;- -- 9 s n L.

rigurc e~wll HDR-Ell.5 Velocities in the spiral stainvay during blowdown

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. 11: Assessment of HDRExperiments V21.1, V43,T31.5and El 1.5", University of Mlaryland at College Park, forDOE -Project

HDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D 10765", April 1996)

<21-Y.;-Z 10 Revision I

59S6r0 -c I2.%pd-041404

WCAP-1 5862APP-SSAR-GSC-588 AP 1 000

. . . -" 3

v "M KMwo- S -- -amvry"l- _ a"

|4>T er ga

a

TIME AFTER BREAK IN MINUTES

Figure 9.C.2-102a HDR-EII.5 Velocities in the main stainvay during containment heatup phase

(reprinted from: L. Wolf, K. Alun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWVR-containmeints after'large LOCAs, Vol. 11: Assessment of'HDRExperiments V21.1,V43, T31.5 and El 1.5", University of Maryland at CollegePark, for DOE - ProjectHDR Hydrogen Mixing Evaluation for'Containment Safety Evaluations Natural Global Circulation,OrderNumber: DE-AP07-961D10765",Aprill996) '

Revision I 9.C-21 7

Revision I5956rl -c I3.vupd-41404

9.C-21 7

WCAP-1 5862APP-SSAR-GSC-588 A ngenA,ArIvuU

Spiral StairwayVELOCITY IN MISSCALE: 0. +/- 3.

CF 432

-- Owl3 CF8401

- CF7801

CF6604

Im� I --- - ____AM SMIM I

XWT~ o.,MO XXI101

XX4301

0 20.0 40.0 60.0TIME AFTER BREAK IN MINUTES

80.0

Be __ _ _ _ ^ < AS w

Itigure 9.C.2-lU2b HDR-E1 1.5 Velocities in the spiral stairway during containment heatup phase

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of IIDRExperiments V21.1, V43,T31.5 and El 1.5", University of Maryland at CollegePark, for DOE-ProjectLIDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE--AP07-96ID10765", April 1996)

Y.1-;iI Z Revision I5956rl-c13.upd-041404


TIME IN M[NUrES

Figure 9.C.2-103a 'HDR-EIl 1.5 Velocities in the main stairvay during H2-He gas mixture releasephase

(reprinted from: L. Wolf, K; Mun, "Overview of experimental results for long-term, Large-scalenatural 'circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments '21.1, V43, T31.5 and El 1.5", University of Maryland at College Park, forDOE -ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

Revision 1 9.C-2195956r1.c13.wpd-041404

K>

I


_,,. z _--at

OM _ A

CF 432

CF9202

CF8401

CF7801

CF6604

XX1 101

XX4301

Spiral Stairway VELOCrrY IN M/SSCALE: 0. + 2.

1 m nI

TmE IN MINUTES

Figure 9.C.2-103b HDR-El 1.5 Velocities in the spiral stainvay during H2-He gas mixture releasephase

(reprinted from: L. Wolf, K. Mun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. 11: Assessment of HDRExperiments V21.1,V43,T31.5 and El1.5", Universityof laryland at College Park, for DOE-ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

9.C-220 Revision 1

9.C-220 Revision I5956rl -c I3.ApdO041404


an:MMM

Xre LOU

TIME IN MINUTES

Figure 9.C.2-104a HDR-E11.5 Velocities in the main stairway during sump heatup and boilingphase

(reprinted from: L. Wolf, K. Mlun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments'after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1, V43, T31.5 and E11.5", University of Maryland at College Park, for DOE-ProjectHDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

K>Revision I5956rl -c13.wpd-041404

9.C-221

I


Spiral Stairway VELOCllY IN M/SSCALE: O.+/-2.

CF 432

F9202

ven"_ § i as

vnal L - awl

MMo ii Y CF7801

Ux-oft4of '^nnnk

XX1101

XX4301

. __

Sensor Failure

Sensor Failure

I I

600.o 60.0 720.0

TME IN MINUTES

K2

Figure 9.C.2-104b HDR-E11.5 Velocities in the spiral stainvay during sump heatup and boilingphase

(reprinted from: L. Wolf, K. Alun, "Overview of experimental results for long-term, Large-scalenatural circulations in LWR-containments after large LOCAs, Vol. II: Assessment of HDRExperiments V21.1, V43,T31.5and El 1.5", University of Alaryland at College Park, forDOE-ProjectIIDR Hydrogen Mixing Evaluation for Containment Safety Evaluations Natural Global Circulation,Order Number: DE-AP07-961D10765", April 1996)

9.C-222 Revision 1

9.C-222 Revision I5956rl -c I3.-pd-041404


9.C.2.4.3 References

1. T. Cron, D. Schrammel, (1993)"Investigations on Hydrogen Distribution in a Reactor Containment, Quick Look Report, Test GroupEl i, Experiments El 1.0-6"(In German), PHDR Technical Report PHDR 111-92, 1993

2. H. Holzbauer, L. Wolf, (1994)"GOTHIC verification on behalf of HDR-Hydrogen mixing experiments"International Conference on New Trends in Nuclear System Therrmohydraulics, Volume II,pp.331-340., Pisa, Italy, May 30th - June 2nd, 1994

3. T. Kanzleiter, L. Valencia, (1984)"Blowdowvn-Experiments in a Reactor Containment, Quick Look Report, Test Group CONW andCOND, Experiments V21.1, V21.3, V45"(In German), Technical PHDR Report No. 49/84, Nuclear Center Karlsruhe, Germany, May, 1984

4. L. Valencia, (1987)"Design Report, Blowdown and Hydrogen Distribution Experiments, HDR-Test Group CON,Experiments T31.4-5"(In German), PHDR-Working Report No. 3.516/87, Nov. 1987

5. H. H. Wenzel, R. Grimm, L. Lhr, (1987)"Test Report, Blowdown and Hydrogen Distribution Experiments, HDR-Test Group CON,Experiments T31.5"(In German), PHDR-Working Report No. 3.520/88, Dec. 1987

6. L. Wolf, L. Valencia, (1988)"Results of the Preliminary Hydrogen Distribution Experiment at HDR and Future Experiments forPhase 1II"16th Water Reactor Safety Information Meeting, Gaithersburg, MD, USA, Oct. 24-27, 1988.

7. L. Wolf, L. Valencia, (1989)"Experimental Results ofthe PreliminaryHDR-Hydrogen DistributionTestT31.5" 4th Intl. TopicalMtg. on Nuclear Reactor Thermal-Hydraulics, Karlsruhe, Oct. 10-13, 1989, vol. 2, pp. 967-973.

8. L. Wolf, H. Holzbauer, T. Cron, (1994)"Detailed Assessment of the HDR-Hydrogen mixing experiments El 1"

Revision I 9.C-2235956r] -9cwvpd441404

I


International Conference on New Trends in Nuclear System Thermohydraulics, Volume 11,pp. 91-103., Pisa, Italy, May 30th - June 2nd, 1994

9. L. Wolf, K. Mun, J. Floyd, (1995)

"HDR hydrogen mixing evaluation for containment safety evaluations, Final Report, DOE-ProjectOrder No.: DE-AP07-951D81401, Dept. of Materials and Nuclear Engineering, University ofMaryland, College Park, MD, Sept. 1995

10. L. Wolf, K. Mun, (1996)"Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAS - Vol. II: Assessment of HDR Experiments V2 1.1, V43, T31.5 andEl1.5" _

DOE - Project, HDR Hydrogen Mixing Evaluation for Containment Safety Evaluations NaturalGlobal Circulation, Order Number: DE -AP07 - 961D10765University of Maryland at College Park, April 1996

11. L. Wolf, M. Gavrilas, K. Mun, (1996)

"Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAS"Final Report for DOE - Project, Order Number: DE - AP07 - 961D 10765

University of Maryland at College Park, July 1996

9.C-224 Revision I5956rl -9c.wpd-041404


9.C.2.5 CONCLUSION

Eleven of the twenty experiments presented have a low release position HI/H, < 0.2. These experiments may

be compared to the AP600 and APIOOO.

They include the seven BMC experiments with a low release position (F2 set, Phase 2, 3 and 4, Test 2 -Phase 1, Test 20 -Phase 11 and RX4), the NUPEC experiment M4-3, and the three HDR experiments (El 1.3,

El 1.4 and El 1.5). Most of these tests (except two - BMC Test 20 and NUPAC M4-3) have globalcirculation through the dome which contributes towards homogenization of temperature and concentration

fields.

The two experiments with the low release position where stratification is recorded are BMC Test 20 and

NUPEC M4-3. As already noted, the stratification occurs due to the special circumstances (boundary

conditions). BMC Test 20 stratifies due to the initially stratified temperature field. Upper compartmentsare maintained at the higher temperature for several days. Global circulation starts first in the lowercompartments, resulting in higher concentrations of the released gas mixture being recorded. Later, the

circulation flow path penetrates the upper thermally-stratified layers and the vertical concentration gradients

are smaller.

In the M-4-3 NUPEC test, the temperature stratification occurs after the end of the release, although injectedgas mixture homogenized slowly. The thermally insulated shell of the NUPEC containment could be onecause for thermal stratification. The homogenization of the gas mixture concentration indicates that some

circulation inside of the containment existed, probably due to the presence of the sump.

The external cooling of the AP600 and AP 1000, as well as the hot concrete structures positioned in the lowerportion of the containment, will produce global circulation.

Seven experiments have high release positions (H/H, > 0.5), two BMC (Test 4 and Test 12), all three CVTRexperiments, and two HDR experiments (El 1.2 and T3 1.5). Thermal stratification is present in six of the

seven tests, with the exception of BMC Test 12. BMC Test 12 includes only hydrogen injection.

Due to the uniform initial temperature field (boundary conditions) and circulation patterns formed by

hydrogen injection, the concentration field was uniform in BMC Test 12. In the second and third CVTRexperiments, the application of the internal sprays decreases the pressure and vertical temperature gradients.

In HDR experiments El 1.2 and T31.5, the temperature and concentration fields stratify at the beginning of

the experiments. Later, the global circulation decreases the vertical gradients.

Revision I 9.C-2255956r1 -9c.wpd-04 1404


In experiment El 1.2, additional steam release in the lower compartment and the application of externalsprays generates global circulation and decreases vertical gradients.

A review of the tests indicates that global circulation and atmosphere homogenization will occur if:

1) The position of the steam or hydrogen release (H,/H < 0.2) is low

2) External sprays are applied3) Internal sprays are active

4) Openings between compartments are large5) The temperature field is not initially stratified

6) Heat sources, such as hot concrete walls or a sump, are at the low positions

7) Compartments are connected (not dead-ended)

In the case of AP600 and APIOOO passive containment designs, all conditions except 3) and partially 7) aresatisfied.

9.C-226 Revision I

9.C-226 Revision I5956rl 9c.Apd-041404


U9.C.3 APPLICATION OF LUMPED-PARAMETER CODES FOR MODELING LARGE

CONTAINMENT FACILITIES

Results of codes using the lumped parameter approach are summarized in this chapter. The ability of the

lumped parameter codes' to model large facilities is assessed and guidelines to improve predictions are

presented.

Tests from several facilities included in an international database are used to compare and validate the results

of the lumped parameter containment analysis computer codes. Among the test facilities in the database areBMC, NUPEC, and HDR. Results from these experiments have been described in sections 9.C.2.1, 9.C.2.2,

and 9.C.2.4, respectively.

9.C.3.1 VALIDATION OF THE LUMPED PARAMETER CONTAINMENT ANALYSIS

COMPUTER CODES BASED ON BMC EXPERIMENTALRESULTS

The BMC F2 tests are used to validate various containment codes (Fisher et al., 1991 and Fisher et al., 1993).

BMC tests 2, 4, 6, 12, and 20 are used to compare various lumped parameter and distributed-parameter

GOTHIC models (L. Wolf, H. Holzbauer, M. Schall, 1994).

9.C.3.1.1 F2 Experiments - Natural Convection Phenomena Inside the Multi-Compartment

Containment

A comparison of the experimental results of the F2 experiments with the results of various codes is presented

by L. Wolf, M. Gavrilas, K. Mun, 1996. The comparison is based on thermal-hydraulic benchmark exercises

by Fisher et al., 1991 for Phase I and by Fisher et al., 1993 for Phases 2, 3, and 4. Many different codesincluding FUMO, JERICHO, FIPLOC, WAVCO, CONTAIN, MELCOR, and COBRAIFATHOMS arecompared. The comparisons demonstrate the state of containment code development with respect to multi-compartment thermal-hydraulics.

Figures 1.17 through 1.37 (in Wolf et al., 1996) illustrate both the experimental data and the results of

various containment analysis codes for F2-Experiment Heatup Phase (Phase I) over a long period of time(48 hours). Single-node models are specified for this exercise. This results in three mean sources ofdeviations between'measurement and code results:'

I) Single-node models are not able to model stratification (due to the penetration ofthe steam front)

inside the dead-end compartments.

Revision I 9.C-2275956rl -9c.upd-04 1404

WCAP-1 5862APP-SSAR-GSC-588 AP 1000

2) One-node lumped parameter models cannot model the stratification front passing throughhorizontal vents. One-node models provide artificially perfect mixing between the connectedcompartments, while in reality only a portion of the vents are available for circulation.

3) Sump liquid level and sump temperature are not predicted well due to the instantaneoustransport of high temperature condensate into sump and stratification phenomena inside the sump.(Note that the temperature of the condensate is too high.)

The influence of all three discrepancies between the experimental and numerical results can be decreasedby applying a series of nodes in a vertical direction or subdivision, as in the GOTHIC distributed parameter

model.

Figures 1.50 through 1.55 (in Wolf et al., 1996) compare the experimental data for the F2-ExperimentPhase 11 with the results of various codes. All compartments except the external annulus have almosthomogeneous temperatures due to the presence of the natural circulation. A comparison of the measured andpredicted velocities through the vents is in agreement for the majority of the codes, which supports theconclusion (L. Wolf, M. Gavrilas, K. Mun, 1996) that containment codes based on lumped parameter modelscan predict fully-developed natural circulation flows.

However, some codes produced results that are not correct. It was established that the incorrect codesmissed or oversimplified the buoyancy terms (see K. Fischer et al., 1993).

9.C.3.1.2 Influence of Initial Temperature Distribution, Location of Hydrogen Injection,

Duration of Injection, and Size of Vent Openings on Hydrogen Distribution

(BMIC Tests 2, 4, 6, 12, and 20)

A comparison of the multi-dimensional and lumped parameter GOTHIC containment analyses and the BMCexperimental data is presented in Wolf et al., 1994b. Tests 2, 4, and 6, described in Section 9.C.2.1.2, areperformed with only the central compartments R I, R2, and R3. Only twvo-dimensional GOTHIC results arepresented.

Three different models are used to compare the results with multi-compartment experiments 12 and 20.Test 12, which uses a uniform initial temperature distribution, results in homogenized hydrogenconcentrations in the entire containment. The lumped parameter model simulates hydrogen concentrationhistories in the various compartments (see Figure 9 in Wolf et al., 1994b, see also Figure 9.C.2-20 in thisappendix). To account for recirculation flows, the applied lumped parameter model uses double-junctions in the horizontal direction.

9.C-228 Revision I5956r] -9c.wpd.041404


Test 20 has a stratified initial temperature distribution, which results in a higher hydrogen distribution in thelower rooms. The conventional lumped parameter model, with one junction connection between two

subvolumes, results in a hydrogen concentration profile (in RI and R2 rooms) that is opposite to that

measured (see Figure 9 in L. Wolf, H. Holzbauer, M. Schall, 1 994b). Also, the experimental results indicatea stratified hydrogen concentration during the first three hours, while the computed generated hydrogen

distribution is more uniform. The calculation performed'with double-junction lumped parameter

modeling (second model) does not improve results. This model also results in a high degree ofhydrogen homogenization inside the containment.

The third type of model uses a two-dimensional model (distributed parameter model) for all containmentrooms except room R2 and uses double-junctions for horizontal connections. The results are presented in

Figure I I in L. Wolf, H. Holzbauer, M. Schall, 1994b and Figure 9.C.2-21 of this appendix. Improvement

between experimental and computed results is obvious (except in the source room R6). A finer nodalization

of the source compartment R6 could further improve the results, because the processes in this compartmentaffect the hydrogen distribution phenomena in all other portions of the containment. Therefore, a more

complete momentum formulation, such as that in GOTHIC distributed parameters, is needed to accurately

model H2 distributions.

K>Revision 1I5956ri-9c.wpd-041404

9.C-229


9.C.3.2 VALIDATION OF THE LUMPED PARAMETER CONTAINMENT ANALYSISCOMPUTER CODES BASED ON NUPEC EXPERIMENTAL RESULTS

9.C.3.2.1 IM-7-1 Test

A comparison of the M-7-1 test results with the various computer codes results is presented in report

NEA/CSNIIR(94)29. The WGOTHIC (Westinghouse modified GOTHIC) code results are also included (seepages 125, 130, 135, blind test calculation results - pages 246-253, open test calculations - pages 322-327,

preheating calculation - 374-378). However, the M-7-1 test is not directly applicable to the passivecontainment design, because internal sprays are active during the test.

9.C.3.2.2 M14-1 Test

The experimental results of NUPEC M-4-3 test (T. Hirose, 1993) are used to validate various lumpedparameter codes and to check the input geometry and boundary conditions, before applying them to blind

and open comparisons with the NUPEC M-7-1 database. WGOTHIC is also compared with NUPEC M-4-3experimental data. A comparison of the WGOTHIC results with the M-4-3 test, which is performed without

internal sprays, is presented by R. P. Ofstun, J. Woodcock, D. L. Paulsen, 1994. The location of the break

is at a low position in the first level of the containment. A mixture of steam and helium is injected at aconstant flow rate during the first 30 minutes and experimental data is recorded for 2 hours.

A comparison of the experimentally and numerically obtained pressure of the dome is illustrated in

Figure 9.C.3-1. Figures 9.C.3-2 through 9.C.3-5 present comparisons of the temperatures and helium

concentrations inside the compartments 8, 15, dome, and 7. These compartments are affected by the global

circulation loop. Lumped parameter modeling methods developed for NUPEC M-4-3 are summarized inSection 9.2.4. Good agreement exists between the numerical and experimental results for the multiple,

connected, lumped parameter compartments in the circulating regions.

However, a paper by R.P. Ofstun, J. Woodcock, D. L. Paulsen, 1994 shows a discrepancy between the

measured and calculated (lumped parameter model) helium concentrations in dead-end compartments.

Calculated values for in-core chase node and two pressurizer nodes (which are all dead-end compartments)are presented and compared with M-4-3 test data (see Figures 3-8, in R.P. Ofstun, J. Woodcock,

D. L. Paulsen, 1994). The discrepancy shows that the lumped parameter model has difficulty predictingcirculation effects within dead-end compartments.

It is postulated that asymmetric temperatures on the vertical walls of the dead-ended compartments induce

a natural circulation within the compartment that is not modeled in the lumped pammeter model. To improve

the results for dead-end compartments, a WGOTHIC distributed parameter model is applied to the

<2/9.C-230 Revision I

5956r0-9c.wpd-041404

WCAP-1 5862APP-SSAR-GSC-588 API 000

dead-ended compartments and interfaced to the lumped parameter model (see Figures 9 - 14 in R. P. Ofstun,

J. Woodcock, D. L. Paulsen, 1994).

The lumped parameter model predicts pressure, temperature and helium concentrations inside the

compartments affected by the global circulation loop. However, a distributed parameter model is necessary

to improve predictions inside the dead-end compartments.

Revision I5956rI-9c.wpd-G41404

9.C-231

WCAP-15862APP-SSAR-GSC-588 AP1000APP-SSAR-GSC-588 APIOOO

Test M-4-3 Comparison-N-UMERICAL RESULTSso@*.E.XPERMENTAL RESULTS

-

Ia-

lua

0toT_

1V*

es

8

0_

rime (see)

Figure 9.C.3-1 Dome pressure history

(Numerical results are produced in the framework of: R.P. Ofstun, J. Woodcock, D.L. Paulsen,"Westinghouse-GOTHIC modeling of NUPEC's hydrogen mixing and distribution test M-4-3", TheThird International Conference on Containment Design and Operation, Volume 1, Toronto, Canada,October 19 - 21, 1994)

9.C-232 Revision I5956r1-c14.wpd-041404


Test M-4-3 Comparisona TV8 DC15X * o --

£a)

C-

co)bw

MCL

CO'

0,

z

a)'

Time (sec)

Figure 9.C.3-2a Temperature history inside node 8

(Numerical results are produced in the framework of: R.P. Ofstun, J. Woodcock, D.L. Paulsen,"Westinghouse-GOTHIC modeling of NUPEC's hydrogen mixing and distribution test A14-3", TheThird International Conference on Containment Design and Operation, Volume 1, Toronto, Canada,October 19 - 21, 1994)K>Revision I5956rl -c14.wpd-041404

9.C-233

I

WCAP-15862APP-SSAR-GSC-588 A DI nn

__ _r I uuu

Test M-4-3 Comparison3R8 DC44

- - -

cC3C:00E

IcoCD

*00z

b)

Time (sec)

Figure 9.C.3-2b Helium concentration history inside node 8

(Numerical results are produced in the framework of: R.P. Ofstun, J. Woodcock, D.L. Paulsen,"Westinghouse-GOTHIC modeling of NUPEC's hydrogen mixing and distribution test M14-3", TheThird International Conference on Containment Design and Operation, Volume 1, Toronto, Canada,October 19 - 21, 1994)

9.C-234 Revision 1

9.C-234 Revision I5956rl tc I 4.,Apd-041404


-Y

a,

13

co,a)2

a)0z

a)

Test M-4-3 Comparison0 TV15 DC22

coC4- -..

C,,I

0 *

co0cm

0

o,C); |:1

Time (sec)



Revision I 9.C -235

Revision I5956rl -0 4.v96pd 041404

9.C-235

WVCAP-15862APP-SSAR-GSC-588 APIOOO

- - - -

0

C00E

a)

U)

0z


cm I4 -

- . - a. ......

00

up -N

LO - a

T. /o

_ .,

b)

LO

0

Time (sec)



V>9.C-236

Revision 19.C-236 Revision 1

59560 -6 4.,ipd-041404

I


Test M-4-3 Comparison0

%e.

EDE

L.6.

E

a,E0a

a)

Time (sec)

Figure 9.C.3-4a Temperature history inside dome

(Numerical results are produced in the framework of: R.P. Ofstun, J. Woodcock, D.L. Paulsen,"Westinghouse-GOTHIC modeling of NUPEC's hydrogen mixing and distribution test M-4-3", TheThird International Conference on Containment Design and Operation, Volume 1, Toronto, Canada,October 19 - 21,1994)

Revision I5956r -c1 4.wpd-04 1404

9.C-237

I

WCAP-I 5862AP1000

Test M-4-3 Comparisona

0-

a,

C0)0

a,20

0

b)J ;.j2

Time (sec)

..

Figure 9.C.34b Helium concentration history inside dome

(Numerical results are produced in the framework of: R.P. Ofstun, J. Woodcock, D.L. Paulsen,"Westinghouse-GOTHIC modeling of NUPEC's hydrogen mixing and distribution test M4-3", TheThird International Conference on Containment Design and Operation, Volume 1, Toronto, Canada,October 19 - 21, 1994)

AI9.C-238 AUvblslon I5956rl sc i 4.upd-04I404

WCAP-15862APP-SSAR-GSC-588 3-8 -AP1000

Test M-4-3TV7 DC14

i:

Comparison

U'Y-

a)

-

a)CL

a)

r-o0,z .i-

a)

Time (sec)


(Numerical results are produced in the framework of: R.P. Ofstun, J. Woodcock, D.L. Paulsen,"Westinghouse-GOTHIC mnodeling of NUPEC's hydrogen mixing and distribution test M-4-3", TheThird International Conference on Containment Design and Operation, Volume 1, Toronto, Canada,October 19 - 21, 1994)

Revision I 9.C-239

Revision I5956d -c I4.%N-d-041404

9.C-239

I

WCAP-15862APP-SSAR-GSC-588 A------

Ar I UUU


LO 1 1- - -._C*J r-

0

-

cJ

0

0

z

b)

Time (sec)


(Numerical results are produced in the framework of R.P. Ofstun, J. Woodcock, D.L. Paulsen,"Westinghouse-GOTHIC modeling of NUPEC's hydrogen mixing and distribution test MA4-3", TheThird International Conference on Containment Design and Operation, Volume 1, Toronto, Canada,October 19 - 21, 1994)

YI.a-4(U Revision I5956rl1-c4.wpd-041404


9.C.3.3 VALIDATION OF THE LUMPED-PARAMETER CONTAINMENT ANALYSIS

COMPUTER CODES BASED ON HDR EXPERIMENTAL RESULTS

Comparisons of the results of best-estimate open post-test predictions of various containment analysis

computer codes (as RALOC, WAVCO, CONTAIN, MELCOR, and GOTHIC) for El 1.2 and El .4 HDRexperiments presented by L. Wolf, H. Holzbauerand T. Cron, 1994a. Detailed comparisons and verifications

of the blind and open GOTHIC results for the El1.2 and El1.4 HDR experiments are presented by

H. Holzbauer and L. Wolf, 1994. More details about El1.2 and El1.4, and comparisons with the

experimental results of large blowdown tests T3i.5, V21.1, and E1 1.5, are presented in the report by

K. Fischer, M. Schall, L. Wolf, 1995.

Following is a summary of observations from the comparison of the experiments with codes:

1) Scenarios with homogeneous containment atmosphere (like El 1.4 and El 1.5) can be simulatedsuccessfully with lumped parameter models.

The results of the El 1.4 computation agree with the corrected experimental input data (open test -see Figures 17-21 in H. Holzbauer and L. Wolf, 1994).

The same conclusion is valid for accidents initiated by a large-break LOCA in the lower positions

ofcontainments, accompanied by subsequentsteam and gas releases (as in El1.5). Good agreementhas been achieved by the GOTHIC lumped-parameter model using a modest number of nodes (see

Figures 10.9 - 10.22 in K. Fischer, M. Schall, L. Wolf, 1995). A comparison of the experimentally

and numerically obtained pressure histories is presented in Figure 9.C.3-6. The calculated pressurehistory is slightly higher than the experimental results. Figure 9.C.3-7 presents a comparison forsump temperature. Note that the sump temperature sensor is exposed to the containment atmosphere

for the first 8 hours. It is then submerged by higher sump water level. The history of sump boilingis well simulated. Comparisons of temperatures in main and spiral stairways at 6 m level are

presented in Figures 9.C.3-8 and 9.C.3-9, respectively. Gas mixture concentration comparisons forthe same staircase elevation are presented in Figures 9.C.3-10 and 9.C.3-11. The measured velocity

in spiral stairway at 15 m elevation is presented in Figure 9.C.3-12. Since the order of magnitudeof the computed velocities matches the data, it can be concluded that trends in the directionof the flow are predicted well; however, predicted velocities differ by as much as factor of two.

2) The lumped parameter method is not capable of predicting the hydrogen distribution in astratified containment atmosphere (as in E11.2 test with high-positioned release).

The pressure history is well predicted after applying the correct steam inflow experimental data and

K.. accounting for energy sink of the sensors cooling system. Comparisons between the experimental

Revision I 9.C-2415956rl-9c.upd-041404


data and the "optimized" post-test prediction for containment pressure are presented inFigure 9.C.3-13. GOTHIC lumped parameter results overpredict pressure by 0.25 bar (11%).

Temperatures, steam, and gas concentrations are underestimated and overestimated above and below

the break location,, respectively, (see Figures 7 and 8 in H. Holzbauer and L. Wolf, 1994).Therefore, for a high-positioned release, the steam and gas transport to the lower parts of thecontainment are overpredicted.

Artificial limitation of convective flows (by reduction of flow path areas) in all modeled flow

paths improves the prediction of temperatures and gas concentrations in the lowercontainment regions (Figures. 13, 15, and 16 in H. Holzbauer and L. Wolf, 1994), butoverestimates the containment pressure and temperatures in the upper containment (see results

of parametric calculations NAlS andNA16, Figures 11 and 12 in H. Holzbauerand L. Wolf, 1994).

3) Application of distributed parameter models may improve prediction of a stratifiedcontainment atmosphere (as in El 1.2 test).

A study of the influence of the number of nodes applied for a simple volume proportional to theHDR El 1.2 facility volume using a distributed parameter model is presented by J.S. Narula and

J. Woodcock, 1994. The authors conclude that in situations requiring detailed flow distribution in

non-homogenous environments, a fine mesh may be required to obtain realistic results. A coarsemesh (small number of nodes) may reduce the accuracy of the results, despite a highly accuratecontainment model in all other aspects, including momentum representation.

A detailed study of the influence of the nodalization, and a comparison of the lumped parameter

results with distributed parameter model results for El 1.2 case are presented by K.K. Mun, 1996.

This report also includes a comparison of the pressure, temperature, steam and hydrogen

concentrations histories (see Figures 67 - 91 in K.K. Mun, 1996). The author concludes that, ingeneral, the modeling approach of merging the distributed and lumped parameter models improves

the prediction. However, to improve the prediction further, a proper distributed parameter

nodalization for the lower break compartment and dome, together with the correct connections for

the associated flow paths, are required.

All specified observations 1,2, and 3 for GOTHIC applications are in agreement with comparisons

of the results of other well-known codes (see L. Wolf, H. Holzbauer and T. Cron, 1994a).

K-I9.C-242 Revision I

5956ri-9c.wpd-041404


0

'-i

.0.

&-4

Ci20)1.-,

CP04011

GOTHIC

. . . . . . . . . . ... .. . . . . . . . . . . .. .. . . . . . . . . .

.. . .. . .:.. . . . . .. . . . . . . . .. . . . . . . . .

I0

0. 20000. * 40000.

Time [s]

Figure 9.C.3-6 HDR-EI1.5: Comparison between measured long-term containment pressure andGOTHIC open post-test prediction

"Reproduced with permission from a report funded by the EPRI GOTHIC Advisory Group: K.Fischer, M. Schall, L. Wolf, " Simulations of GOTHIC Large Scale Containment Experiments",Battelle Ingenicurtechnik GmbH, Eschborn, Fachbericht BF -V - 68317 - 01, October 1995)

K>Revision I5956r -c14.wpd.041404

9.C-243

II

WCAP-15862APP-SSAR-GSC-588 Apin00JAP1AA�1

c

I-

1-4

U

0. 20000. 40000.

Time [s]

Figure 9.C.3-7 HDR-E11.5: Comparison between measured sump temperature and GOTHIC openpost-test prediction

"Reproduced with permission from a report funded by the EPRI GOTHIC Advisory Group: K.Fischer, M. Schall, L. Wolf, " Simulations of GOTHIC Large Scale Containment Experiments",

. Battelle Ingenieurtechnik GmbH, Eschborn, Fachbericht BF - V - 68317 - 01, October 1995)

9.C-244 Revision I

9.C-244 Revision I5956r I c I4.mpd-041404

WCAP-1 5862APP-SSAR-GSC-588 APIOQO

K)~

I~

,-' 00CZ)

tM .

S

6

Time [s]

Figure 9.C.3-8 HDR-EII .5: Comparison between measured temperature in main stairivay (6m) andGOTHIC open post-test prediction'

"Reproduced with permission from a report funded by the EPRI GOTHIC Advisory Group: K.Fischer,M'N. Schall, L. Wolf, " Simulations of GOTHIC Large Scale Containment Experiments",Battelle Ingenicurtechnik GmbII, Eschborn, Fachbericht BF - V - 68317 - 01,' October 1995)

Revision I 9.C-2455956r0 c14.wpd-041404


U00

c)

Q.

E'&-4

6

Time [s]

Figure 9.C.3-9 HDR-E11.5: Comparison betwveen measured temperature in spiral stairway(6m) and GOTHIC open post-test prediction

"Reproduced with permission from a report funded by the EPRI GOTHIC Advisory Group: K.Fischer, M. Schall, L. Wolf, " Simulations of GOTHIC Large Scale Containment Experiments",Battelle Ingenicurtechnik GmbH, Eschborn, Fachbericht BF - V - 68317 - 01, October 1995)

V29.C-246


5956rl -c I4.%%pd 041404

WCAP-15862APP-SSAR-GSC-588 Ai~innnJAPiAAn

0

0.-

C#)0C7K)

6 0. 20000. 40000.

Time [s]

Figure 9.C.3-10 IIDR-E11.5: Comparison between measured gas mixture concentration inmain stairway (6m) and GOTHIC open post-test prediction

"Reproduced with permission from a report'funded by the EPRI GOTHIC Advisory Group: K.Fischer, Al. Schall, L. Wolf,-" Simulations of GOTHIC Large Scale Containment Experiments",Battelle Ingenieurtechnik GmbH, Eschborn, Fachbericht BF - V - 68317 - 01, October 1995)

K>Revision I5956r -c 5.wpd-041404

9.C-247


6

4-i

Z-4

0

C:0

.~q

00U -

cu

V

6'-4

6 L.0. 20000.

Time [sI

Figure 9.C.3-1 1 HDR-E1I.5: Comparison between measured gas mixture concentration inspiral stairway (6m) and GOTHIC open post-test prediction

"Reproduced with permission from a report funded by the EPRI GOTHIC Advisory Group: K.Fischer, M. Schall, L. Wolf, " Simulations of GOTHIC Large Scale Containment Experiments",Battelle Ingenieurtechnik GmbH, Eschborn, Fachbericht BF - V - 68317 - 01, October 1995)

9.C-248 Revision I5956r I-c 15.wpd-041404

WVCAP-I 5862APP-SSAR-GSC-588 AP] 000

IW.

-4001

, '-1' n11 I" ! III*

. .

, I1-.CF7801

c.- .... GOTHIC 10. 20000. 40000.

Time [s]

Figure 9.C.3-12 HDR-E11.5: Comparison between measured velocity in spiral stairway (15m)and GOTHIC open post-test prediction

"Reproduced with permission from a report funded by the EPRI GOTHIC Advisory Group: K.Fischer, A1. Schall, L. Wolf, ".Simulations of GOTHIC Large Scale Containment Experiments",Battelle Ingenicurtechnik GmbH, Eschborn, Fachbericht BF -V 68317 - 01, October 1995)

K.>

Revision I5956rl c15.wpd-041404

9.C-249

WCAP-15862APP-qRAR-G5SC-SR9 API000

I'

L.

-o

L.

a,

a)

C,,-

.0C/.13

4.80 i

Time in min * 102

Figure 9.C.3-13 HDR-E1 1.2: Comparison between measured long-term containment pressureand various "optimized" open post-test predictions

(Reproduced with permission from a report funded by the EPRI GOTHIC Advisory Group: K.Fischer, I1. Schall, L. Wolf, " Simulations of GOTHIC Large Scale Containment Experiments",Battelle Ingenieurtechnik GmbH, Eschborn, Fachbericht BF - V - 68317 - 01, October 1995)

9.C-250 Revision 1

9.C-250 Revision 15956rl - I5.wpd{041404

WCAP-15862APP-SSAR-GSC-588 AP IO00Q

U9.C.3.4 CONCLUSIONS

Lumped parameter models can predict pressure, and in some cases temperature, steam, and hydrogenconcentrations inside the containment. Good predictions exist for the situations where the global circulation

loop, which contributes towards the homogenization of atmosphere, is present inside the containment (see

comparison with BMC F2 experiments and BMC test 12 experiments, NUPEC M-4-3 tests and HDR El 1.4

and El 1.5 tests). Additional improvements are possible if double-junctions between compartments in thehorizontal direction are applied (allowing simulation of recirculation flows).

Single-node, lumped parameter models are not able to model stratification inside dead-end compartmentsdue to the steam front penetration. They also do not include cases where the stratification front passes

through horizontal vents, nor do they simulate sump liquid level, temperature, and stratification inside the

sump. (See the BMC F2 experiments and the NUPEC M-4-3 results for dead-end compartments (R. Ofstun

et al., 1994). The remedy can be an increase in the number of the nodes in the vertical direction or theapplication of distributed parameter model for dead-ended compartments (R. Ofstun et al., 1994).

For a stratified initial temperature distribution (BMC test 20) and a stratified containment atmosphere (HDR

El 1.2 test), the lumped parameter models do not predict the correct hydrogen distribution (they overmix the

atmosphere). Theartificiallimitation ofthe size ofthe flow path areas in theHDREl1.2 test improved theprediction of temperatures and gas concentrations in the lower containment regions, but overestimated the

containment pressure and temperatures in the upper levels.

Application of distributedparametermodels may improve prediction capability. See BMC test 20 (L. Wolf,

H. Holzbauer, M. Schall, 1994b; HDR El 1.2 test - J. S. Narula and J. Woodcock, 1994; and K. K. Mun,

1996). All three publications conclude that proper distributed parameter nodalization for break and domecompartments is required to improve results. However, the increase of the number of nodes in all important

regions of containment increases required memory and computing time beyond current practical limits.

The next step towards better space resolution and modeling would be the application of commercial CFD

(Computational Fluid Dynamics) codes withbuilt-in capabilities formulti-phase flows simulation, and multi-

block, unstructured, or embedded grid options. Considering that three-dimensional transient muiti-phase

flow simulation would be required, only codes prepared forparallel multi-processormachines wouldproduce

results in a reasonable time. In addition, there are limitations related to noding convergence (see Gavrilas

et al., 1997). Finally, an appropriate set of constitutive models (see Royl et al., 1997), such as condensation

and sump evaporation, are not presently available in many CFD codes.

Revision I 9.C-2515956rd -9c.upd-04 1404


For the AP600 and API000 with a low break release position and an established global circulation patternthat improves homogenization (see conclusion of the review of international experimental data base), acareful application of lumped parameter models could produce results which would be in agreement with

test data.

In summary, biases that should be addressed when applying lumped parameter codes are:

1. Single-node models are not able to model stratification, or the passing of a stratification front

through horizontal vents.

2. Sump liquid level and sump temperature are not predicted well.

3. Some codes produced results that are not correct due to missing or oversimplified buoyancy terms.

4. To account for recirculation flows, the applied lumped parameter model uses doublejunctions in thehorizontal direction. (This does not help in the case of an elevated release and resulting stratified

containment.)

5. For releases low in containment, typical for the AP600 and API000 DECLG LOCA, the lumpedparameter model predicts pressure, temperature, and helium concentrations inside the compartments K.)that are affected by the global circulation loop. Improvements are needed to account for postulated

circulation effects inside dead-ended compartments.

6. Scenarios with homogeneous containment atmosphere (HDR El1.4 and El1.5) are simulated

successfully with lumped parameter models. (Such conditions typically result from breaks located

below 20 percent of the containment height.)

7. Circulation effects due to sump boiling (releases generated at the bottom of containment) are well

simulated.

8. The order of magnitude of computed velocities matches the experimental data. Trends in the

direction of the flow are predicted well; however, predicted velocities differ by as much as a factor

of two.

9. The lumped parameter method is not capable of predicting the hydrogen distribution in a stratified

containment atmosphere, as in HDR El 1.2 with high-positioned release. In a break scenario witha buoyant plume (released at about 50 percent of containment height), the steam and gas transport

to the lower parts of the containment are over-predicted. (When convective flows are artificially

K.)

9.C-252 Revision I5956r -9c.wpd-041404


limited by decreasing flow areas, predicted concentrations in the lower regions improve, but thecontainment pressure and temperatures in upper compartments are overestimated.)

9.C.3.5 REFERENCES

1. Fischer, K., T., Schall, M., Wolf, L., 1991, "CEC Thermal-Hydraulic Benchmark Exercise on

FIPLOC Verification Experiment F2 in BMC-Long-Term Heatup Phase - Results for Phase I," Final

Report, EUR 13588 EN, 1991

2. Fischer, K., T., Schall, M., Wolf, L., 1993, "CEC Thermal-Hydraulic Benchmark Exercise on

FIPLOC Verification Experiment F2 in BMC, Experiment F2 in BMC, Experimental Phases -Results for Phase 2, 3, and 4, Results of Comparisons," Final Report, EUR 14454 EN, 1993

3. K. Fischer, M. Schall, L. Wolf, 1995, "Simulations of GOTHIC Large-Scale Containment

Experiments," Battelle Ingenieurtechnik GmbH, Eschbom Fachbericht BF -V - 68317 - 01, October

1995

4. M. Gavrilas, B. T. Mattingly, M.J. Driscoll, 1997, "WGOTHIC Noding and Parametric Studies ofa Passively-Cooled Small Rating PWR Containment," Proceedings of the International Topical

Meeting on Advanced Reactors Safety, Volume 1, Orlando, Florida, June 1-5, 1997, pp. 549 - 560.

5. H. Holzbauer, L. Wolf, 1994, "GOTHIC Verification on Behalf of HDR-Hydrogen MixingExperiments," International Conference on New Trends in Nuclear System Thermohydraulics,

Volume ]1, pp.331-340., Pisa, Italy, May 30th - June 2nd, 1994

6. T. Hirose, 1993, Letter, floppy disk which contains the M-4-3 test data, an attached document, and

a video-cassette, NUPEC Nuclear Power Engineering Corporation, ISP35-035, 30 April, 1993

7. K. K. Mun, 1996, "Modeling of HDR Experiment El 1.2 with the Distributed Parameter Feature of

the GOTHIC - Code," M.S. Scholar's Paper, University of Maryland at College Park, September

1996

8. J. S. Narula, J. Woodcock, 1994, "Westinghouse - GOTHIC Distributed Parameter Modeling ofHDR Test El 1.2," The Third International Conference on Containment Design and Operation,

Volume 1, Toronto, Canada, October 19 - 21, 1994

Revision 1 9.C-2535956rl -9c.kpd-04 1404


9. NEA/CSNIIR(94)29: Final Comparison Report on ISP-35: NUPEC Hydrogen Mixing and

Distribution Test (Test M-7-1), Committee on the Safety of Nuclear Installations OECD NuclearEnergy Agency, December 1994

10. R. P. Ofstun, J. Woodcock, D. L. Paulsen, 1994, "Westinghouse - GOTHIC Modeling of HDR

NUPEC'S Hydrogen Mixing and Distribution Test M-4-3," The Third International Conference on

Containment Design and Operation, Volume I, Toronto, Canada, October 19 - 21, 1994

11. P. Royl, Breitung, J.R. Travis, H. Wilkening, L. Seyffarth, 1977, "Simulation of Hydrogen Transport

with Mitigation Using the 3D Field Code Gasflow," Proceedings of the International Topical

Meeting on Advanced Reactors Safety, Volume 1, Orlando, Florida, June 1-5, 1997, pp. 57 8 - 588.

12. L. Wolf, H. Holzbauer, T. Cron, 1994a, "Detailed Assessment of the HDR-Hydrogen MixingExperiments El 1," International Conference on New Trends in Nuclear System Thermohydraulics,Volume 11, pp. 91-103., Pisa, Italy, May 30th - June 2nd, 1994

13. L. Wolf, H. Holzbauer, M. Schall, 1994b, "Comparisons Between Multi-dimensional and Lumped

Parameter GOTHIC Containment Analyses with Data", International Conference on New Trends in

Nuclear System Thermohydraulics, Volume 11, pp. 321-330., Pisa, Italy, May 30th - June 2nd, 1994

14. L. Wolf, M. Gavrilas, K. Mun, 1996, "Overview of Experimental Results for Long-Term Large-

Scale Natural Circulations in LWR-containments after Large LOCAS," DOE - Project, OrderNumber:

DE - AP07 - 961D 10765, University of Maryland at College Park, July 1996

K\)

9.C-254 Revision I

9.C-254 Revision I5956rl-9c.Nvpd 041404

Appendix 9.D

Basis for Assuming Homogeneous Bulk Conditions forAP600 and AP1000 Containment Pressure

Design Basis Analysis

<2

5956r] -9d.,wpd-041404


TABLE OF CONTENTS

LIST OF TABLES ................ ivLIST OF FIGURES ................ iv

9.D.1 INTRODUCTION ................... 9.D-I9.D.2 APPROACH ................... 9.D-I9.D.3 REGION I HORIZONTAL GRADIENTS ADDRESSED THROUGH

TURBULENT BUOYANT PLUME ANALYTICAL MODEL ..... ........... 9.D-29.D.4 REGION 2 HORIZONTAL GRADIENTS ADDRESSED THROUGH LST

TEMPERATURE DATA AND SEPARATE EFFECTS ENCLOSURE TESTDATA . ............................................................ 9.D-2

9.D.5 REGION 3 HORIZONTAL GRADIENTS ADDRESSED THROUGHTURBULENT BOUNDARY LAYER ANALYTICAL MODEL ..... ......... 9.D-4

9.D.6 COMPOSITE REGION DISCUSSION .............. I....................... 9.D-59.D.7 TURBULENT BUOYANT PLUME ANALYTICAL MODEL - SPECIES

CONCENTRATION ................................ 9.D-69.D.8 ENTRAINMENT INTO THE BREAK PLUME AND WALL LAYER, AND

THE EFFECT ON MASS TRANSFER RATE ............................. 9.D-99.D.9 TURBULENT BOUNDARY LAYER ANALYTICAL MODEL .9.D-179.D.10 INFLUENCE OF HORIZONTAL GRADIENTS ON MASS TRANSFER

COEFFICIENTS ................ 9.D-289.D.1 I CONCLUSION ................ 9.D-299.D.12 REFERENCES ................ 9.D-29

Revision I iii5956rl -9d.Npd-041404


LIST OF TABLES

Table 9.D-1 Analytical Solution to the Turbulent Buoyant Plume Model(Reference 9.D-5, Table 9-7) .................. 9.D-3 1

LIST OF FIGURES

Figure 9.D-I

Figure 9.D-2aFigure 9.D-2b

Figure 9.D-3

Figure 9.D-4Figure 9.D-5Figure 9.D-6Figure 9.D-7

Interactions Between Containment Regions During Post-BlowdownLOCA (Low Momentum) . ...................................... 9.D-32Detail of Steam Concentration Distribution in Boundary Layer ... ..... 9.D-33Qualitative Radial Concentration Profile from Plume Centerline toShell at Mid Elevation . ........................................ 9.D-33Period of Application of Analytical Models for Turbulent WallBoundary Layer and Turbulent Buoyant Plume .... ................. 9.D-34Turbulent Buoyant Plume Analytical Model ..... .................. 9.D-35LST Measurement Locations ......... .......................... 9.D-36LST Temperature Profile - Test 220.1 @ Quasi-Steady Conditions ..... 9.D-37LST Temperature Profile -Test 217.1 @ Quasi-Steady Conditions ..... 9.D-38

'\y'iv Revision I

5956r0-9d.wpd-04144


9.D.1 INTRODUCTION

During the blowdown phase of a DECLG LOCA, forced jet momentum from the break source providesvigorous mixing of steam and air inside containment. The jet momentum helps establish homogeneousconditions within a few seconds (Reference 9.D-12, Section 10.2.1.5). Consequently, treating thecontainment as homogeneous or well-mixed during the blowdown phase is appropriate.

After the blowdown phase, however, the physical processes are different. The forced jet momentum quicklydecreases and actually ceases for a brief period during the refill phase before the break source reestablishesitself as a buoyant plume. Although the forced jet momentum ceases during the refill phase of a DECLGLOCA, wall boundary layers are quickly established, on the order of seconds (see Reference 9.D-12,Section 4.8) and according to work by Hartley, quasi-isotropic turbulence resulting from the jet momentumcontinues mixing enclosed volumes for some time after the source jet ceases. The turbulent wall boundarylayers provide: entrainment of the bulk region inside containment and hence contribute to mixing of thecontainment atmosphere as well.

This appendix addresses the buoyant plume, wall boundary layers, and enclosure test data as they relate tomixing inside containment during the post-blowdown time period during a DECLG LOCA.

9.D.2 APPROACH

Thermal and concentration gradients in the horizontal direction, including the boundary layer next to theshell, are addressed using results of 1) theoretical, or first principles, calculations 2) LSTtemperature/concentration data, and 3) separate effects enclosure test data. The wall boundary layer steamconcentration profile is examined using turbulent boundary layer models. Enirainment and boundary layerprofile calculations are used to estimate the magnitude of the steam concentration in the wall layer and risingplume as compared to the bulk average steam concentration at a given elevation. Test data, where available,are used to support the conclusions.

Horizontal gradients in AP600 and AP 1000 during the post-blowdown LOCA periods are addressed for thebase case scenario which assumes that the source mass rises from the east steam generator compartment asa buoyant plume, having lost its momentum as it passes around the equipment and structure in thecompartment.

The potential magnitude of horizontal gradients are examined in three radial regions: (1) the turbulentbuoyant plume (2) the bulk region between the plume and boundary layer and (3) the turbulent boundarylayer. The three radial regions are depicted in Figure 9.D-1. This three-region approach is similar to thatsuggested by Peterson (Reference 9.D-1). In the region-wise discussions, area fractions are developed andprocesses which affect horizontal temperature and concentration gradients within each region are described.

Revision 1 9.D-l5956rl -9d.mpd.041404


The test bases for the discussions include the LST, as well as smaller scale separate effects naturalconvection enclosure tests. The general interaction between the regions via entrainment is then used tosummarize how the integral system performs.

9.D.3 REGION 1 HORIZONTAL GRADIENTS ADDRESSED THROUGH TURBULENTBUOYANT PLUME ANALYTICAL MODEL

Referring to Figure 9.D-1, Region I is comprised of the rising plume and its entrained flow. In Region 1,the concentration profile within a turbulent buoyant plume is characterized by an exponential (i.e., e'57(rh) 2

profile at any given elevation, with the steam concentration at the boundary of the plume equal to that of thebulk (Region 2). Such a profile is summarized by Blevins (Reference 9.D-2). The turbulent buoyant plumeanalytical model, based upon Reference 9.D-2, shows that the plume centerline steam concentration at theexit of the plume is within 10 percent (0.04/(1-0.6) = 0.10) (see Section 9.D-7) of the bulk (Region 2),assuming uniform temperature and concentration in Region 2. With the assumption that the buoyant plumedevelops freely to the dome region, the plume occupies about 10 percent of the cross-sectional area at thetop of containment. An axisymmetric round plume with a divergent angle of 7.5 degrees from the exit of thesteam generator compartment is assumed. The calculated plume equivalent diameter at the top ofcontainment is 38 feet based on the net exit area of the trapezoidal steam generator compartment opening,the containment radius of 65 feet, and plume height of 121 feet. It is recognized that negatively buoyantplumes may descend from the dome region and limit the free development height of the positively buoyantplume. However, since this phenomenon results in an additional mixing mechanism (i.e., counter-currentplume mixing of positively and negatively buoyant plumes), it is conservatively neglected in the simpleanalytical plume model.

9.D.4 REGION 2 HORIZONTAL GRADIENTS ADDRESSED THROUGH LSTTEMPERATURE DATA AND SEPARATE EFFECTS ENCLOSURE TEST DATA

The horizontal gradients in Region 2 can be assessed based on indications from LST temperature data andseparate effects enclosure test data and by examining the postulated condition of a non-uniform horizontalgradient.

The LST has been examined for evidence of horizontal gradients outside the wall boundary layer.Time-averaged thermocouple rake data for LST tests in the LOCA configuration at tvo different steam flows(220.1 at 0.5 lbm/sec; 217.1 at 1.0 Ibm/sec) show a near-zero horizontal temperature gradient except abovethe steam plume and over the distance within one inch of the wall (see Figures 9.D-6 and 9.D-7). Since theconvective processes operating in the bulk gas region mix temperature differences and gas species at aboutthe same rate, the observation that there is no horizontal temperature gradient also shows that there is nohorizontal gas concentration gradient, and therefore, no density gradient. The data show that all of themeasurable temperature drop in the vessel wall boundary layer occurs between the wall and the thermocouple

9.D-2 Revision I5956r1-9d.wpd-041404


located one inch away from the wall. This is consistent with boundary layer calculations which indicate thatthe majority (60 percent) of the gradient occurs within the first 0.25 inch of the 7-inch boundary layer at theLST operating deck level.

LST thermocouple rake data at successive, incremental data collection times, shows evidence of buoyancyforces eliminating horizontal gradients. Data from the thermocouple rake was taken every 90 seconds duringthe LST matrix tests. The data at some point in time may show that the temperature at a measurementlocation may deviate less than ten degrees Fahrenheit from the horizontal average. Data at the next timeinterval indicate that horizontal gradients are not maintained for more than the 90-second data acquisitioninterval. This is evidence that a perturbation which creates a nonhomogeneous density at a given elevationalso creates a local relative density driving force which tends to level out the horizontal density gradient.This is to be expected since gravity will neutralize any gradients that attempt to form, tending to result in atime-averaged flat profile.

Separate effects enclosure test data show that as the Rayleigh number, Ra, increases from 3.5x 1O0 to Ix I o6,

the circulation mechanism fully transitions to turbulent, and the horizontal gradient becomes primarilyconcentrated in the thin wall boundary layer (Section 9.C.1.1.2.1, see for example Figure 9.C.1-5). Thesetests show that at Ra > 106, the horizontal gradient in the bulk region is nearly zero in two-dimensionalenclosures. Data from three-dimensional enclosures suggest that the transition to turbulence may occur atRa < I o6, due to vortices which affect mixing inside the enclosure by communicating between the front andback walls through the middle of the enclosure. The Rayleigh number between the plume centerline and thecooled wall for both AP600 and API000 is 4.2 x 1012 based on a relatively low value of 9'F temperaturedifference. Therefore, both the AP600 and API 000 are fully turbulent and would be expected to show littleor no horizontal gradient in Region 2.

Negatively buoyant falling plumes have been shown to occur with upper horizontal surfaces cooler than theenclosure and to result in an additional mixing mechanism (Figures 9.C.1-15 and 9.C.1-16) which wouldfurther homogenize the containment gases. The presence of plumes of cooler gases descending from theunderside of the dome, suggested by the large vertical Grashof number on the order of 2x10W3, based on arelatively low temperature difference of 90F (Section 9.C.1.3.1), are conservatively ignored.

Separate effects enclosure test data span up to Ra equal to 109. Numerical studies by Markatos andPericleous predict that the heat transferbehavior is constant (Nusselt number linearly increases with length)at increasing Ra over many decades up through Ra of IO6 (Figure 9.C. 1-6), suggesting that increased heattransfer is primarily a result of thinning of the boundary layer. The implication is that the Nusselt numbervaries monotonically over that range, and thus no unexpected performance is introduced at higher Ra values.

In a bulk region such as Region 2 in the AP600 or AP 1000, while one may not rule out some minor transienthorizontal temperature or concentration gradients, a horizontal gradient outside the relatively small volumesoccupied bytheplume and falling wall layeris difficult tosustainathigh Ra number. That is, anypostulated

Revision I 9.D-35956rl-9d.wpd.041404


deviation from horizontal uniformity would tend to be readily flattened by buoyant forces. Therefore,horizontal gradients may be assumed to occur solely within the plume and wall layer.

9.D.5 REGION 3 HORIZONTAL GRADIENTS ADDRESSED THROUGH TURBULENTBOUNDARY LAYER ANALYTICAL MIODEL

Region 3 consists of the negatively buoyant turbulent wall boundary layer. Equations for the velocity,temperature, and air concentration boundary layer profiles are presented in Section 9.D.9. The profiles arebased on turbulent boundary layer 1/7 power laws, as used by Eckert and Jackson. More recent work in thearea of turbulent boundary layer profiles has led to the refinement of the boundary layer into three or morelayers, such as the viscous sublayer, logarithmic-law layer, and outer layer, to attempt to increase accuracy.Such multi-layer models may indeed increase the accuracy; but results show the 1/7 power profile to bereasonable to use for simple calculations.

The results of the boundary layer calculation for the LOCA post-blowdown atmosphere are compared tocalculations for the LST in the LOCA configuration, that is, having a diffuser below the steam generatorcompartment. The LST includes the effects of both temperature and concentration as driving forces fornatural convection. Conclusions regarding the thin part of the boundary layer near the wall, over which mostof the gradient occurs are consistent with observations of the internal thermocouple rake data from LST.

The normalized steam concentration profile versus normalized distance through the boundary layer is plotted Ad.Jin Figure 9.D-2a. The parameter, C, represents local boundary layer steam concentration, and thesubscripted values represent: s for surface, and b for bulk. The normalized value, il, is x/8, where x isdistance from the wall and 8 is the boundary layer thickness. From Figure 9.D-2a, it can be seen that theaverage steam concentration in the boundary layer is only 12.5 percent lower than the steam concentrationin the bulk.

The maximum boundary layer thickness has been calculated from the Eckert and Jackson style model to beless than 31 inches at the operating deck level in AP600 (assuming that a turbulent boundary layer exists forthe full AP600 containment height). From the calculated profile, it is found that 60 percent of the changefrom bulk to wall conditions occurs over the first 2Y/2 percent of the boundary layer, or over less than the firstinch in AP600. Furthermore, these boundary layer calculations conservatively overpredict wall layerthickness due to neglecting the effects of suction at the boundary (velocity normal toward the wall due tocondensation). The boundary layer occupies less than 9 percent of the AP600 cross-sectional area at theoperating deck elevation, based on the boundary layer thickness calculations and the AP600 containmentradius of 65 feet. The region of significant concentration difference occupies less than 0.5 percent of theAP600 and APIOOO cross-sectional area.

9.D4 Revision I5956rI-9d.wpd-041404


9.D.6 COMPOSITE REGION DISCUSSION

Referring to Figure 9.D-1, Region I is comprised of the rising plume and its entrained flow; Region 2 is thebulk region of low velocity and nearly zero horizontal temperature/concentration gradient; and Region 3 is

the falling turbulent wall boundary layer and its entrained flow.

Region 1 occupies about 10 percent of the cross-section at the top, and Region 3 occupies about 9 percentof the cross-section at the bottom, the maximum cross-section for each. Thus, it is reasonable to neglectmomentum-related interaction between the rising plume and the falling wall layer in simple first principlescalculations. Since the plume occupies about 10 percent of the cross-sectional area at the top, and even lessat lower elevations, as a first approximation the effects of the enclosure walls and dome on plumeentrainment over the plume height is neglected. Similarly, with the even smaller area occupied by the fallingwall layer, the effects of the enclosure are neglected when calculating wall boundary layer entrainment rates.

At a given instant in time, the rising plume of Region I is supplied to the above-deck volume from the topof the steam generator cavity. Plume entrainment calculations (Section 9.D.8) show that 5 to 14 times thesource flow is entrained from Region 2 over the height of a buoyant plume above the operating deck inAP600. Therefore, the plume average steam concentration would be expected to be near that of the bulk bythe time it discharges to the top of Region 2. As discussed previously, entrainment is shown to result in acenterline plume steam concentration at the top that is within 10 percent of the bulk average steamconcentration.

The flow leaving Region I spreads horizontally and feeds Region 2 at the top. Since higher order mixingmechanisms have been neglected, by continuity, the vertical flow velocity at the top of Region 2 is a lowvelocity net downward flow with a steam concentration reduced by condensation and heat transfer on thedome to a value lower than the steam concentration discharged from Region 1.

In Region 3, condensation develops a liquid film over the full height of the containment shell and anegatively buoyant gas boundary layer that grows with distance down the shell. Noting that at quasi-steadyconditions, the volumetric condensation rate on the wall is just equal to the source flow rate, Q., wallboundary layer entrainment calculations (Section 9.D.8) show that the volume of steam condensed on theshell is only 1/6 to 1/13 of the volumetric flow rate of gas entrained into the falling wall layer. Such largeentrainment rates relative to the condensation rate suggest that the average steam concentration exitingRegion 3 would be near the bulk steam concentration of Region 2. These results are consistent with wallboundary layer profile calculations, discussed previously, which show that the average steam concentrationthrough the boundary layer is only 12.5 percent below the steam concentration in the bulk.

Because global, orlarge-scale, circulation through the operating deck reduces stratification in the above-deck

region (Section 9.C. 1.4.2.4), global circulation through the operating deck is conservatively neglected for

this discussion. By continuity, the flow exiting the bottom of Region 3 spreads out over the operating deck

area, rising in Region 2 with a low net upward velocity and with the same steam concentration as the exit

Revision I 9.D-5595 6r1-9d.wpd-0-41404


flow of Region 3. Because the top of Region 2 is fed by the plume and the bottom of Region 2 is fed by thewall layer, Region 2 will have a higher steam concentration at the top than at the bottom (neglecting

negatively buoyant plumes in the dome region).

In this simplified model, as one moves from the top down in Region 2, entrainment into both the plume andthe wall layer on either side of Region 2 steadily reduces the downward flow; similarly the upward flowfrom the bottom is reduced as one moves upward from the operating deck. There is, therefore, a neutralplane in Region 2 where the vertical velocity approaches zero.

A simplified representation summarizing horizontal gradients through the three regions, consistent with theabove'discussion, is shown in Figure 9.D-2b.

9.D.7 TURBULENT BUOYANT PLUME ANALYTICAL MODEL - SPECIES

CONCENTRATION

This section addresses the assumptions, methodology, equations, and results for the turbulent buoyant plumeanalytical model - species concentration.

9.D.7.1 TIME PERIOD OF APPLICATION

The turbulent buoyant plume analytical model is applicable in the post-refill phase (when break flow intocontainment is re-established) and beyond, as shown in Figure 9.D-3.

9.D.7.2 NOMENCLATURE (SEE FIGURE 9.D-4)

u velocity in plume

C concentration of air in plume

T = temperature in plume

U. = bulk velocity

C_ = bulk air concentrationT_ = bulk temperature

Cd = plume centerline air concentration

Subscripts:

c = bulk

o = source

CL = centerline

9.D-6 Revision I5956r]-9d.wpd-041404

WCAP- 5862APP-SSAR-GSC-588 AP 1000

9.D.7.3 MAJOR ASSUMPTIONS

I. Plume is turbulent and buoyant (i.e., density inside plume is less than the bulk region, Region 2).

2. Plume is not confined by wall surfaces or descending negatively buoyant plumes. It freely develops

until it reaches dome region. It is recognized that descending negatively buoyant plumes may begenerated in the dome region, and may limit the development height ofthe positively buoyant plumegenerated at the outlet of the steam generator compartment. However, the interaction of the

positively and negatively buoyant plumes further enhances mixing, and therefore, this potential

phenomenon is neglected in this simple analytical model.

3. Bulk conditions are uniform (i.e., homogeneous reservoir)

4. Plume is axisymmetric/round

9.D.7.4 SOLUTION

The analytical solution to the turbulent, buoyant plume model with the above assumptions is already known,

and the results are summarized in Table 9.D-1 (Reference 9.D-5, Table 9-7 - Turbulent Plumes in ConstantK./ Density Reservoirs).

9.D.7.5 APPLICATION OF TURBULENT BUOYANT PLUME RESULTS TO

LST/AP600

Applying the plume centerline species concentration analytical solution, an expression for plume centerlineair concentration, CCL is obtained. (Note that C replaces T, and z replaces x from Table 9-7 of

Reference 9.D-5)

CCL = C_ - II(Q0ACQ) B- 3 z 513

B = gQ 0 | = BuoyancyFlux

where: P

AC = Difference in concentration of air in plume and ambient

Q = Volumetric flow rate of source

Applying the above expression to LST Test 213. 1 B and AP600 at the same total pressure (i.e., 29.8 psia) and

air concentration (i.e., 0.60), the air concentration values near the top of containment can be obtained:

Revision I 9.D-75956r1 -9d.wpd-041404


For LST Test 213.1 B:

p.= 29.8 psia, p = 17.9 psia, p,,. = 11.9 psia, T. = 218°F, z = 13.2 ft

C_ = psi//p. = 0.60, p, = 0.03 Ibm/ft3, pi,= 0.073 Ibm/fl3, pt,,,,0 = 0.067 Ibm/ft3

111o 0.54 Ibm/sec= O = 8 ft3/sec

P,,,,o 0.067 Ibm/ft3

B =32.2x 8x0.35=90.2

P.. Pstf.o= 0.35

P-

where:

P_ = Pair + Psm

AC.=C -0 =0.60

AC = C x 11 x 8 x (90.2)13 x (I3 .2 )5n =0.16

For AP600: (assuming the same bulk air concentration as LST 213.1)

p. 29.8 psia, p.i, = 17.9 psia, pblk = 11.9 psia, T_ = 2020F, z = 110 ft

C._= pi/p. = 0.60, p., 0 = 0.0723 Ibm/ft3

lb. 15.4 Ibm/secQ =- 213 ft/secPmo,0 0.0723 Ibm/ft3

B = 32.2 x 211 x 0.305 = 2072

P. - PsIu.o = 0.305

P.-

AC =0.60 x 11 x 213 x (2072)" x (1 10)5 = 0.04

9.D-8 Revision I5956rl-9d.wpd-041404


Air Concentration, Cc,. at the Top Centerline of Buoyant Plume

Pressure P - Pst Q. | z B(atm) P.. (ttPlsec) - C| (ft) (ft4IsSc 3 ) AC CCL

LST 2 0.35 8 0.60 13.2 90.2 0.16 0.44

AP600 2 0.305 213 0.60 110 2072 0.04 0.56

9.D.7.6 TURBULENT BUOYANT PLUME CONCLUSIONS

1. AP600 turbulent buoyant plume should provide better dilution of steam compared to LST due to theincreased height above the source in AP600, and the zest dependence for mixing/entrainment.

2. Based on the assumption ofa well-mixed Region 2, the centerline concentration in the plume is close

to ambient conditions for both AP600 and LST near the top of containment. Due to turbulent

mixing, the average concentration in the plume should be even closer to the Region 2 bulk

conditions.

9.D.8 ENTRAINMENT INTO THE BREAK PLUME AND WVALL LAYER, AND THE

EFFECT ON MASS TRANSFER RATE

The purpose ofthis calculation is to determine the volumetric entrainment rate into the break plume and into

the negatively buoyant wall layer in AP600 containment during a large LOCA. The plume and buoyant wall

layer entrainment models recommended by Peterson for mixing in large stratified volumes are used. Theentrainment rates are used to calculate the dilution of the plume and buoyant layer to get the steam/airconcentrations at the top (dome) and bottom (deck elevation) of containment. With the steam/air

concentrations, the effect on the mass transfer rate in the'dome region and at the operating deck level isdetermined.


The entrainment models for the buoyant plume are applicable in the post-refill period and beyond and modelsfor the wall boundary layers are applicable to the post blowdown period and beyond, as shown in

Figure 9.D-3.

Revision I 9.D-9

Revision 159560 -9d.wepd 041404

9.D-9


9.D.8.2 KEY ASSUMPTIONS

9 Bulk fluid (Region 2) is at a constant density

9.D.8.3 ENTRAINMENT INTO A BUOYANT PLUME

The volumetric flow rate entrained is:

Q = 0.l5B3 Z5 1 3

where: Z = 108 ft. the height of the plume (conservatively low value of height from steamgenerator cavity outlet to dome)

B = g Q.(P~b - P.)/Pb, the buoyancyQO = the break steam volumetric flow rate

p0 = the break steam density (assumed to be the saturation density at the total pressure)P=,b = density of the ambient gas, the same density as the entrained gas, pn,

It is convenient to define the plume entrainment ratio:

rp Q /Q0 = 0. 15 g"i (Apfp)"3 Zn Q;-2'

For entrainment into the plume, conservation of mass iii, = iti", + M. with 1In = Qp, gives QouPoul = Q0tPat +Qp0. Conservation of mass for each species produces the relationships

QOULPSiOUI = QcntPs n + QOp,,O and Qoutpout = Qtpen,, where the s and a subscripts indicate the partial densitiesof the steam and air. With partial density and pressure pi = P/RIT, so mass conservation for each speciesbecomes QouPo/T. = Q.AP,5 ./rT + QOPO/TO and QutP.JOUIO.. = Q.P,Ie,/Tt,. Since the absolute

temperature does not differ significantly throughout the system, the species equations can be approximatedas Q..P,.o = QP,. + QOP. O and QutPot = QtP. nt. Note that the entrained flow is several times larger thanthe source flow, so the effect of the higher source temperature is not very significant. These latter equationspermit the calculation of the steam and air concentration at the outlet of the plume.

Conservation of mass on a molar basis n0u, = nf, + no. With fi = PuA, A = P/RT, and Q = uA, conservation ofmoles can be written PouZQuOJTOJ - PQ,.JRTQt+ POQJRTO. Since the total pressure, P is the same for each,

and the absolute temperature T only differs by a small amount, conservation of mass on a molar basis can beapproximated as Qut = Qt + Q0.

With conservation of mass in terms of steam partial pressure, P. out = (QtP, , + Q0P.0 )/Qou,, and Q0,t =Qct

+ Q0, the outlet steam partial pressure is P, ou, = (Q.PS. + Q0PO)/(Qt, + Qj). Written in terms of the plumeentrainment ratio rp, P,. = (rpP,,ct + P,O)/(rp + 1).



where:

rh = mass flow rate

n = molar flow rate

p = molar density

Q = volumetric flow rate

p = mass density

R = gas constant

T = absolute temperature

Subscripts:

a = air

s = steam

ent = entrained

out = outlet

o = source

stm steam

bl = boundary layer

Values for densities and break flow rates at different times using values representative of post-refill and peak

pressure are as follows:

90 sec 1200 sec units

Pmb 0.13584 0.16325 Ibm/ft3

pO = 0.10487 0.13373 Ibm/ft3

P, = 46 60 psia

P.,, = 26.94 40.30 psia

(Pmb- Po)/Pamb - 0.2280 0.1808g~bdc~o = 200 45 Ibm/sec

Q0 = -lbk,.Po 1907 336.5 ft3 /secB = g Qo(P b - Po)/P.mb 14,000 1959 ft4/sec3

Q.= 0.15 B'm Z' = 8855 4597 ft3/secrp 4.64 13.66

P,.,= (rpP,, + P,O)/(rp + 1) = 30.32 41.64 psia

Revision I 9.D-I I5956ri-9d.wpd-041404


Solving the equations for AP600 at 90 seconds and at the time of peak pressure (1200 sec.) results in thefollowing:

Steam Partial Pressure at Top of AP600 Plume from Entrainment Relations

Q.Qnt pstmtopTime sec Ap/p ft/sec ft/sec r-psia

90 0.2280 1,907 8,855 4.6 30.3

1200 0.1808 336.5 4,597 14 41.6

9.D.8.4 ENTRAINNIENT INTO A NEGATIVELY BUOYANT WALL LAYER

The volumetric flow rate entrained is:

0.0979vGrz 5 Pp.

%nt (1 +0.494Pr23)2 5 Pr 5

where:

Z = 121 ft. the height of the wall (deck to dome)Gr, = g(Pub-Pw)Z3/(Pnbv2), the Grashof numberPr = the Prandtl number of the boundary layer

= boundary layer thicknessv = [L/p, the boundary layer kinematic viscosity based on the average of bulk and surface

values of dynamic viscosity and densityeu = the wall perimeter (7D) = 408.4 ft

P = the total gas density adjacent to the liquid film surfacePDb = the density of the ambient gas, the same density as the entrained gas,pr,V = volume

It is convenient to define the boundary layer entrainment ratio in terms of the entrained and condensedvolumetric flow rates:

rbl _ Q./Qnd

For entrainment into the boundary layer, the equations for conservation of mass ti.O, = -cn , II1d, and on amolar basis n., = rcm, - ft~nd can be developed as was done for the plume. The results are the importantrelationships for the steam partial pressure at the outlet of the wall layer:

for volumetric flow Qo. = Qnl - QCond, the outlet steam partial pressure is P,. = (Q.P 1'.. -

Qcondps.d)/(Q., - Qond)- In terms of the plume entrainment ratio rp, P,,,ut (rpPen, - P,.O)/(rp - 1).

9.D-12 Revision I5956r] -9d.wpd-041404


Values for densities and break flow rates at different times using representative values for post-refill (90 sec)and peak pressure (1200 sec) are as follows:

Pamb

Pw

Pcond

P.

Pr

pV = PI/p

(Pw" Pab)/Pamb

Gr= g(pw-Ppb)ZI/(P.mbv%)

I + 0.494 Pr5

Qcnt

tcond =Igfbrkso d 1,es

Qcond = Icond/PcoNd

rbI = Q../Qnd

P PUt = (rbIP.,.I-P,,CO,,)/(rbI-l)

=

=

-

=

. .

= 85=

90 sec0.135840.190200.104874626.940.811.23xl0-5

0.163027.55xl0-5

0.40014.o0xlO"1.42935098

40.5810.5.6.2923.34

1200 see0.163250.202290.133736040.300.831.24xl0-5

0.182776.78xl0-5

0.23912.97xl0"1.43634004

unitsIbm/ft3

Ibm/ft3Ibm/ft3

psiapsia

Ibm/sec-ftIbmIft3

fi2/sec

fl3/sec

302.813.2238.69

Ibm/secft3/sec

psia

K j Summary Tables:

-. Time, Plotal PSmb 1) Qcond

sec psia psia ft2/sec Ap/p Gr Pr. ft3 se

90 46 26.9 7.55xl0-5 0.4004 4.0OxlO" 0.81 810.5'

1200 60 40.3 6.78xl0- 5 0.2391 2.98xl0'5 0.83 302.8

P . .,bo. 8b1 Vbl

Time, sec ftP/sec . psia ft ft

90 5098 6.3 23.34 2.2 63,200

1200 4004 13.221 38.69 2.3 65,500

9.D.8.5 EFFECT OF ENTRAINMENT ON MASS TRANSFER RATE

The condensation heat flux for given total pressure, steam partial pressure, and bulk-to-surface temperaturedifference is determined and used to calculate the influence of the concentration differences.

K>Revision I5956rl -9d.wpd04 1404

9.D-13


IRepresentative bulk-to-surface temperature differences are:

Tbuk

AT = (TbUfk - Td)

90 sec

= 244.2

= 165.1

= 79.1

1200 sec

267.7 OF

232.2 0F

45.5 0F

The refill time phase, from 30 to 90 sec., has no source, but does have a wall layer. Refill is represented byits end state that is assumed to have the same conditions as calculated at 90 sec, except that without a plumethe steam partial pressure and heat flux at the top is the same as the ambient bulk conditions.

The steam partial pressure values at the top, middle, and bottom correspond to the steam partial pressure outof the plume, the bulk steam partial pressure, and the steam partial pressure out of the wall layer.

The average heat flux is a simple, unbiased average of the top, middle, and bottom values, that is

q = q,.W4 + q3",2 + qb,/4.

Summary of Mass Transfer Effects Due to Vertical Concentration Gradients

Time P,-,, AT -. Imtop 'lwamb Psm.bo t o qsmb q.1 Averagesec psia *F psia psia psia B/sec-ft2 B/sec-ft2 1B/sec-ft2 B/sec-ft2

refill 46 79 26.9 26.9 23.3 1.85 1.85 1.4 1.74

90 46 79 30.3 26.9 23.3 2.25 1.85 1.4 1.84

1200 60 43 41.6 40.3 38.7 1.5 1.4 1.3 1.4

The results in this table show that the, upper estimates of vertical concentration differences in AP600 resultin a 6 percent reduction on the net heat transfer to the shell during refill and less than I percent effect on netheat transfer during the 90 to 1200 sec time period.

9.D.8.6 BOUNDARY LAYER THICKNESS AND VOLUMNIE

The thickness of the negatively buoyant wall layer can be calculated from the integral equations presented byPeterson (Reference 9.D-1).

0.565(1 +0.494Pr2 13)"'0 z5bi ~ Gr '/10 Pr 8/15

K>9.D-14

Revision I9.13-14 Revision I

5956rl -9d.-.%pd-041404


The equation can also be integrated over its height, with the simplifying assumption that all properties areconstant over the height, Z. This assumption is reasonable since the Prandtl number only changes a fewpercent over the range of conditions inside containment, and the only other parameters are contained in theterm Ap/(pv2) which is estimated to change less than a factor of 2 over height, and when raised to the1/10 power has only a 7 percent effect. Consequenily, the product ofthe integral and the circumference givesa reasonable estimate of the boundary layer volume that can be used with the entrainment rate to estimateboundary layer transit times, or fill time.

The product of the circumference and the integral of the equation above is;

V 0.565(1 +0.494Pr 2n)11'0(v2 /g)1 0r D IZ07dzbl- (Ap/p)"'l0 Pr8 /15

0.565in (I +0.494Pr 23)1/O (v2/g)1'ODH 1.7

1.7 (AP/P) "'0Pr81 15

Evaluating this equation with the values from Section 9.D.8.3;

Prv

(pw- Pnb)/PnbI + 0.494 Pr2'3

V (Equation 4)

90 sec= 0.81= 7.55xl0-5

= 0.4001= 1.4293= 63436

1200 sec0.836.78xl 0-0.23911.436364553

units

fl2/sec

ft3

Summary Table:

AP600 Bounda Layer Volume and Fill Times

Boundary Layer FillTime, sec

Time, sec V, ft3 Qfnt 9 ft3/SeC r/Q,,t

90 63436 5098 12.4

1200 64553 4004 16.1

Reiso . ,,...1 5.._.......Revision 45956rl -9d~v6Td-041404

9.D-15


kJ9.D.8.7 SUBLAYER PENETRATION TIMIE

A measure of the response time of the boundary layer temperature to a change in the environment is thesublayer penetration time. This represents the time it takes for steam to diffuse through the laminar sublayer,where most of the mass transfer resistance is located. The transient diffusion equation:

ao = D, ap

at aY2becomes =

DvT at' ay.2

with the substitutions:

A = p.0p Y sY* t = Tt*

where:

t

y

TD,

00

molar densitytimedistance along the normal to the surfaceboundary layer thicknessair-steam gas diffusion coefficienttime constantvalue at a large distance from surfacedimensionless variable

If the time constant is defined T = Sm2/Dv, the coefficient on the left side of the dimensionless equation = 1,as required. The mass transfer sublayer thickness, 8#m. is related to the heat transfer sublayer thickness, ai,by Nu/Sh = (Pr/Sc)'. The heat transfer sublayer thickness is 8

h = h/k. Other assumed values are:

hk

Dv,PrScSh =.h/k

a,, = 8h(Pr/Sc)"

T = S.2/D,

90sec.2.610.01640.5370.810.510.00630.00730.357

1200 sec.2.510.01730.4640.830.51

0.00690.00810.509

units

B/hr-ft2 -FB/hr-ft-Fftl/hr

ftftsec

Consequently, the sublayer penetration time is on the order of I sec or less. This is very rapid in comparisonto the structure time constants that are on the order of 100 sec, and the system pressurization time constant

9.D-16 Revision I

9.D-16 Revision I5956r] -9d.N pd-041404


&that is on the order of 1000 sec. Even the boundary layer fill time of 16 sec is short compared to both thestructure and pressure time constants.

9.D.9 TURBULENT BOUNDARY LAYER ANALYTICAL MODEL

This section addresses the assumptions, methodology, equations, and results for the turbulent boundary layeranalytical model. The analytical model is not intended to be an independent verification of LST, rather, it isused to estimate boundary layer thickness and temperature/concentration profiles for AP600 containment andLST. Measured thermal/concentration boundary conditions from LST tests 217.1 and 220.1 are applied inthe analytical model since these represent realistic conditions to apply to the boundary layer model.


The turbulent boundary layer analytical model (Figure 9.D4) applies to the time period when quasi-steadyconditions are established inside containment. The time period of applicability is shown in Figure 9.D-3.

9.D.9.2 KEY ASSUMPTIONS

* Surface is vertical flat platea- Containment shell radius of curvature is large so it can be treated locally as a flat plate.

Velocity, temperature, concentration boundary layer profile is turbulent- This is appropriate since boundary layer is turbulent within a few feet of the top of the

containment shell.

* Bulk (Region 2) fluid velocity L. 0- This is true after blowdown period since the break source transitions from jet to buoyant

plume and bulk containment area is large relative to plume/boundary layer.

* Condensate film is impermeable to noncondensible gases.- This is conservative since absorption by the film removes air from containment volume

which enhances heat/mass transfer of steam.

Condensate film is stationary relative to gas boundary layer.- This is appropriate since condensate film velocity is smaller than gas boundary layer velocity,

and conservative because a moving film enhances heat transfer.

Suction effect at wall is neglected.- This is conservative because suction thins the boundary layer.

Revision I 9.D-175956ri-9d.wpd-04 1404


Thermal boundary layer thickness is the same as concentration boundary layer thickness.

- This is appropriate since Le"3 = (Pr) - 1.0.

Saturation conditions exist at film surface.

- This is appropriate since steam is condensing at the film surface.

Bulk fluid (Region 2) temperature is uniform above the break source elevation.

- Refer to LST test 217.1 and 220.1 temperature profiles in (Figures 9.D-6 and 9.D-7).

Bulk fluid (Region 2) concentration is uniform above the break source elevation.

Surface Temperature is approximately constant for purposes of applying the simple analytical model.

Wall heat flux is modeled as forced convection using equations from Reference 9.D-6.

9.D.9.3 BOUNDARY LAYER PROFILES

The boundary layer profiles, used by Eckert and Jackson, are based on experimental data for turbulent

boundary layers:

Boundary Layer Velocity Profile (Reference 9.D-7, Equation 24)

1rU = U(1-ri)4l 7 where 11=r

where:

u = local velocity in boundary layer

S = boundary layer thickness

r = coordinate direction normal to condensing surface

Boundary Layer Temperature Profile (Reference 9.D-7, Equation. 25)

T - T = (Tsurf - T) (1 -

9.D-18 Revision I5956r] -9d.upd-041404


where:

T = temperature

Boundary Layer Air Concentration Profile

C - C. = (Csurf C 1)

where:

C air concentration

Boundary Layer Thickness (Reference 9.D-8, Equation. 10.121b)

Bzt

where:

B, n are constants to be determined

Maximum Velocity in Boundary Layer (Reference 9.D-8, Equation. 10.121 a)

U = Azn'

A, m are constants to be determined

9.D.9.4 GOVERNING EQUATIONS

Momentum Equation

df'u (T - Tdr

= gfp (T - T) dr---dz Jp

Revision I 9.D-195956rl-9d.wpd-041404


where:

W = shear stress at condensing surfaceP = volumetric thermal expansion coefficientp = bulk density

g = local gravitation acceleration

Energy Equation

f su (T - T)dr

dz pC p

where:

CP = specific heat at constant pressure

q " = wall heat flux

Conservation of Mass Equation

Not used in this calculation since we are not calculating entrainment volumes/rates, but rather, KJtemperature/concentration profiles.

Boundary Conditions

Thermal and concentration boundary conditions are from LST tests are applied at the shell surfaceand in bulk (Region 2) locations because the boundary conditions are expected to be representativeof AP600.

9.D.9.5 BOUNDARY LAYER MIONIENTUNI EQUATION DEVELOPMENT

Wall shear stress in the momentum equation is modeled using a correlation for forced turbulent convection.Eckert and Jackson argue the Tr is similar to forced convection in the near surface region and can berepresented as follows:

_ = 0.0225 [uspU2 Lv

9.D-20 Revision I5956r]-9d.wpd-041404


where:

v = kinematic viscosity

r,,, = 0.0225pU ( Uo)

wvhere:

U = is a characteristic velocity interpreted by Eckert and Jackson as the maximumvelocity in boundary layer.

The integral momentum equation then becomes:

d Ju'drrl

o = gfp(Y"-7)dr- 0.0225U 2 [o d t bp

Applying turbulent velocity and thermal boundary layer profiles, it can be shown that:

s 1r'u 2 dr = 8 j [U(Io 0

j)4 r7 dTl = 0.0523158U 2

and

I I

(T_ - T)dr = Sf (T_ - Tu,)(1-Tl 7 )dr) = 8(T - T..)8

0

The boundary layer momentum equation becomes:

d (0.052315U52) = gp~(T - Tsf) 8 - 0.0225U- 2 4dx8 [ UV8J

Revision I5956rl -9d.upd-04 1404

9.D-21


Applying Reynolds Analogy for heat/momentum transfer

where:

h = fEFpCPU 2

f, = local friction factor = 0.045 IVF1UTs51

(Reference 9.D-7, Equation 5)


El, = ratio of turbulent eddy diffusivity of heat to that of momentum

h = heat transfer coefficient

from Colbum Analogy:-2

Ell= Pr I (Reference 9.D-7, Equation 6)

where:

Pr = Prandtl numberK>I-

Therefore,

h 0.045pCPU 2

r' -2V!J' Pr3

1U5]

li -2= 0.0225 I v I; Pr3

L Ts J

Now, since

q = A = h(T - Tw)c AlWf

heat flux can be modeled as follows:

qw

pCp= 0.0225 (

r I 2- T surf)U I .va--IPr '

K>9.D-22 Revision I

5956r -9d.mpd-041404


where:

Asf = heat transfer surface area

Now,

f u(T-T_)dr = sjlU(I _-r )4 r 7 lT - T I - 0i7)dr 1 = O.036633(T - Te)8U0 0

The boundary layer energy equation becomes:

2

d [0.03663(T - T0.f)0U] = 0-0225Pr 3 (T - Tsu )U-YV

Substituting expressions for U and S into boundary layer momentum and energy equations, differentiating theresulting equations, solve for exponents m and n by matching exponents. This procedure results in m=1/2 andn=7/10 which can also be found in Reference 9.D-8, page 335.

Therefore:

U=Az2 and 5=Bz '0

Substituting the expressions for U and S into boundary layer equations and performing the differentiation, weobtain: -

(. 7 _ 5

17[0.0252315A 2] = 0.125gp(T~ - T;.)-0.0225A 4 B 4 vI/4

and

32 -5I

6 [0.03663A]Pr3 = 0.0225A 4B 4

K.>Revision 15956rl-9d.wpd-04 1404

9.D-23


\--1Solving for constants A and B, an expression for boundary layer thickness can be obtained:

7 1 2 I

3 = Bz '° = 0.565z(Grz) lO(Pr)IS5l + 0.494Pr3 lo (Reference 9.D-7, Equation 36)

The above equation agrees with that obtained in Reference 9.D-1, Equation 18.

From the expression for 3, the velocity, thermal and concentration boundary layer profiles can be calculated,where:

G,=g'(T,-TsUr)z3GrV =v2


9.D.9.6 BOUNDARY LAYER THICKNESS RESULTS

Rather than calculate Gr. and 3 for a mixture, the calculation is done using air or steam properties to examinethe range of possible values.

AP600 Prediction (for thermodynamic properties based on LST test 217.1)

The following parametric values are associated with LST test 217.1 and are used to calculate boundary layerthicknesses below:

LOCATION I T. I T,,rf Pr (air)

0.70

v (air)

0.84 x I0O ftl2 sec *1E I 230 0F I 188 0 F

Based on AIR properties (Reference 9.D-10) and noting , = I/T:

32.2 ft (2300F-188F)( I ( (1 loft)3

Gr, = sec \ 690JR)

(0.84 x 10-4 sec 2

= 3.7 x 1014

o = 0.565 (110 ft) [3.7 x 10141-1/11[0.7]4-/I5[1+0.494(0.7) 23]-13 0 = 30.5 in.

K>9.D-24 Revision I

5956rI-9d.upd.041404


Based on STEAM properties:

32.2(230 - i88 )(9 I.) (110)3Gr2 = = 3.7 x 1014

(0.84 x 104 )

8 = 0.565(110 fl)[3.4x1014Y-"/0[1.1- 8115[1+0.494(1.1)2 13]-"' = 23.9 in.

LST Prediction (for thermodynamic properites based upon LST test 217.1)

Based on AIR properties:

32.2(230-188)( 690)(13.2)3Gr = = 6.4 x 10"

(0.84 x 10-4 )

at Point E in LST:

8 = 0.565 [13.2 ft] (6.4 x ioll)-10 10 (o.7o)-8"5 [1+0.494 (0.70)2/311/1o = 6.9 in.

Based on STEAM properties (Pr = 1.10, v = 0.87 x 10' ft2/sec):

32.2(230 - 188)(-!--) (13.2)3(690)

(0.87 x 10-4Gr = = 5.9 x 10"

at Point E in LST: . - I

8 = 0.565 [13.2 ft] (5.9 x 1011)111o [1+0.494 (1.1)2j'1"o = 5.4 in.

Revision I 9.D-25

Revision 15956rl -9d.wvpd-04 1404

9.D-25


The following parametric values are associated with LST Test 220.1 and are used to calculate boundary layerthickness below:

| LOCATION T. I Tu.f I Pr (air) I v (air)

E 200OF I 1550F 1 0.70 | 1.0 x 10 4 fl2 /sec

Based on AIR properties:

32.2(200-155)( 660 (I3.2)Gr = = 5.0 x 10"

(1.0 X 10-4 Y

at Point E in LST:

S = 0.565 [13.2 ft] (5.0 x l0Ii)yl/ (0.70)-8"5 [1+0.494 (0.70)'3J'I'o = 7.1 in.

9.D.9.7 CONCENTRATION PROFILE RESULTS

The concentration profile and integrated average in boundary layer are calculated from the turbulent analyticalmodel for LST. The concentration at the surface is determined from inside surface temperature data andapplying saturated conditions near the surface. See Figure 9.D-5 for relative measurement elevations. Radialpositions are shown in Figures 9.D-6 and 9.D-7.

LST Test 220.1 P,Ow = 32 psia, dii = 0.5 Ibm/sec

Average LST Boundary Layer Air Concentration Based Upon Measured Bulk and WVall Conditions

T,,f (measured) (measured) ClC C..Location (OF) (OF) (calc) (measured) (calc)

Dome 180 _ 0.77 0.34 0.39

A 180 230 0.77 0.44 0.48

B - 230 - -

C 177 230 0.78 0.48

D - 230 - -

E 155 200 0.87 0.72 0.74

K>9.D-26

Revision 19.D-26 Revision I

5956rl -9d.wpd-01404

WCAP-15862APP-SSAR-GSC-588 AP1000APP-SSAR-GSC-588 AP 1000

LST Test 217.1 PL w = 43 psia, `t = 1.0 Ibm/sec

Average LST Boundary Layer Air Concentration Based Upon Measured Bulk and Wall Conditions

TIed (measured) T. (measured) Cr C.Location (OF) (OF) (calc) (measured) (calc)

Dome 200 . 0.73 0.33 0.38

A 210 260 0.67 0.35 0.39

B 207 260 0.67 - 0.39

C 205 260 0.70 . 0.395

D 198 260 0.74 - 0.40

E 188 240 0.79 0.67 0.69

Notes:

1. CC,,, is calculated from applying Dalton's law of partial pressures, assuming saturation conditions

exist at the condensing surface, and using measured surface temperatures and total pressure from LST

tests.

2. CBL, which represents the average air concentration in the boundary layer, is calculated from

performing an integrated average which results in CBL= (Crff+ 7C)/8.

3. Refer to inside wall temperature data (Table 4.8-1 forLST test217.1 and Figure 4.1 1-6 forLST test

220.1) in Reference 9.D-1 I for surface temperature data.

4. Bulk fluid temperatures shown in the above tables represent spatially-averaged values based upon

attached LST time-averaged temperature profiles for LST tests 220.1 and 217.1 at quasi-steady

conditions.

9.D.9.8 BOUNDARY LAYER CONCENTRATION PROFILE RESULTS - LST

1. Comparison of C. and CBL shows that the difference in air concentration is small (within 15 percent)

in the radial (horizontal) direction for LST. This along with temperature profile data indicate that

LST is well-mixed in the radial direction throughout the above-deck region, outside of the relatively

thin laminar sublayer.

2. The measured bulk fluid temperature in both horizontal and vertical directions shows little difference

(i.e. a few degrees) between the dome region and region D as seen in Figures 9.D-6 and 9.D-7.K)Revision I

9.D-27Revision I5956rl -9d.N%-d-041404

9.13-27


There is however about 200 F difference in the bulk temperature between regions D and E, for LST

test217.1 andabout30oFdifferenceforLST test220.1. Thedata indicate some level of stratificationbetween regions D and E, which is more pronounced in LST test 220.1. Since the LST does not have

a flow connection from the simulated SG compartment, there is no global circulation. As a result,the stratification below the source elevation is not surprising.

3. Region E temperature/concentration is notably different than the regions above. This indicates an

interface or gradient region exists between the higher temperature, steam-rich upper region (abovethe plane of the break) fed by the buoyant plume, and the lower temperature, air-rich region fed by

the wall boundary layer.

4. Turbulent boundary layer mixing is significant such that boundary layer average properties are nearly

at bulk conditions. It is expected that this will be the case for AP600.

9.D.9.9 BOUNDARY LAYER ANALYTICAL MODEL RESULTS- BOUNDARY

LAYER THICKNESS

1. The maximum boundary layer thickness calculated from the analytical model for LST turbulent

boundary layer.- 7 inches. (LST 217.1 and 220.1)

2. The maximum total boundary layer thickness calculated from the analytical model for AP600 at

similar thermodynamic conditions - 31 inches.

3. The calculated boundary layer thicknesses are conservative because suction effects due to steam

condensation at the wall were not included. It is well known that suction reduces boundary layerthickness (refer to Schlicting, Boundary Layer Theorvy. Therefore, actual boundary layers should be

thinner than calculated.

4. The boundary layer horizontal profile is rather "flat" due to mixing effects of turbulence.Consequently, most of the horizontal boundary layer gradient is contained in the much smaller region

near the condensing surface.

9.D.10 INFLUENCE OF HORIZONTAL GRADIENTS ON MASS TRANSFER

COEFFICIENTS

Since only free convection is assumed throughout the design basis containment transients, the velocitiescalculated in the lumped parameter model are not used in calculating mass transfer rates. Therefore,justification of the approach taken for mass transfer is based on steam concentration gradients.

9.D-28 Revision I5956r1-9d.wpd.04 1404


The horizontal concentration gradients in the post-LOCA containment atmosphere are consistent with thelumpedparameternoding and associated mass transfercoefficients used in the WGOTHIC Evaluation Model.The scale of significant concentration gradients near the wall are much less than the 2-foot thick calculationcell (node) used in the WGOTHIC Evaluation Model. .Thus, the wall cell is large enough in the radialdirection that cell properties can be used to represent the bulk condition for use with boundary layer heat andmass transfer correlations which are described in Reference 9.D4.

9.D.11 CONCLUSION

The following conclusions apply to the containment atmosphere during the post blowdown period for aDECLG LOCA:

1. The turbulent buoyant plume (Region 1) entrains a significant volume from the bulk (Region 2)such that it is nearly at bulk thermal/concentration conditions in the dome region.

2. Based upon LST data and enclosure test data, the large bulk region (Region 2) is well-mixedhorizontally and above the break source elevation, vertically as well.

3. The turbulent wall boundary layers entrain volume from the bulk region such that the averagethermallconcentration in the boundary layer is nearly at bulk conditions.

9.D.12 REFERENCES

9.D-1 Peterson, P.F., "Scaling and Analysis ofMixing in Large Stratified Volumes," International Journalof Heat and Mass Transfer, Vol. 37, Suppl. 1, pp 97-106, 1994.

9.D-2 Blevins, "Jets, Plumes, Wakes, and Shear Layer," Applied Fluid Dynamics Handbook 1984.

9.D-3 WCAP-1 4407, Rev. 2, "WGOTHIC Application to AP600,"April 1998.

9.D14 WCAP-14326, Rev. 2, "Experimental Basis for the AP600 Containment Vessel Heat and MassTransfer Correlations," April 1998.

9.D-5 Blevins, "Jets, Plumes, Wakes, and Shear Layers", Applied Fluid Dynamics Handbook, 1984,pp. 247-251.

9.D-6 Eckert and Jackson, "Analysis Turbulent Free Convection Boundary Layer on a Flat Plate," NACAReport 1015, 1951.

9.D-7 Corradini, "Turbulent Condensation on a Cold Wall in the Presence of a Non-Condensible Gas,"Proceedings on Nuclear Reactor Thermal Hydraulics, Vol. 1, 1983.

Revision I 9.D-295956r -9d.wpd-041404

I


9.D-8 Kakac and Yener, Convective Heat Transfer, 2nd edition, CRC Press, 1995.

9.D-9 Kays and Crawford, Convective Heat and Mass Transfer, 3rd Edition, McGraw-Hill.

9.D-10 Kreith, "Principles of Heat Transfer", 3rd Edition, Appendix.

9.D-1 I WCAP-14135, "Final Data Report for PCS Large-Scale Tests, Phase 2 and Phase 3," Revision 1,April 1997.

9.D-12 WCAP-14845, "Scaling Analysis for AP600 Containment Pressure During Design BasisAccidents," Revision 3, March 1998.

9.D-30 Revision 15956r1-9d.wpd-041404


Table 9.D-1 Analytical Solution to the Turbulent Buoyant Plume Model (Reference 9.D-5,Table 9-7)

Notation: b = half-width of the plume: i.e., transverse distance for the axial velocity to fall to one-halfthe centerline value: B = specific buoyancy flux, see Eqs. (9-50) and (9-54): 0 = volume flow rate.per unit depth for plane plume; 00 = initial volume flow rate: r = radial distance from plume axis:AT = species concentration relative to reservoir level; ATO = Initial relative species concentration:

u = axial velocity; u. = centerline axial velocity: v. = transverse velocity of reservoir fluid intoplume: x = axial distance from origin of the plume: y = transferse distance from center plane:

v = kinematic viscosity. (Refs. 9-1.9-73.) The uncertainty in the coefficients is approximately±10%.

(a) AxisymmetricPlume Characteristic Plane Plume (i.e., round) Plume

1. Centerline velocity, u 1.7 B/ 3.15 B 1 x 3

2. Width, b 0.097 x 0.11 x

3. Axial velocityprofile(b),u/u m

e774(y/x)e-57(r/x)2

4. Volume flow rate, Q 0.34 B1 /3 x 0.15 B x5/3

5. Centerline species 2.4 (Q AT )B1/ 3 x-1 11 (Q AT)B 1/3 x 5/3concentration, AT * "o o 0 x

6. Species concentration 0.13 x . 0.10 xwidth, bet

7. Species concentration 2 2profile(b), AT/AT e-41(y/x) e-69(r/x)

m

6. Entrainment velocity, 0.10.u 0.041 uv m me

9. Reynolds number, B1/3x B1/3 2/3u b/v 0.17 v .35 v

(a)In fully developed region.

(b)These are consistent with the corresponding width.

K>Revision I*5956rl -9d.wpd-041404

9.D-31


, Containment SteelShell

-Region 3

Negatively BuoyantGas Boundary Layer K.)}

Q00q,P0

e,BL

OP

(BL

Source plume volumetric flow at exit of steam generator compartment

Volumetric plume entrainment from Region 2

Volumetric wall boundary layer entrainment from Region 2

Volumetric plume flow feeding top of Region 2

Volumetric boundary layer flow feeding bottom of Region 2

Figure 9.D-1 Interactions Between Containment Regions During Post-Blowdown LOCA (LowMomentum)

9.D-32 Revision I5956ri-9d.wpd.041404


1

0.9

0.8

NormalizedSteam

Concentration,C-Cs

Cb- Cs

0.7

0.6

O0

0.4

0.3

02

0.1

O.

Normalized Boundary Layer Thickness,1 = x/

Figure 9.D-2a Detail of Steam Concentration Distribution in Boundary Layer

0

C0

Eto

co

I

Radius - -* Shell

plumeCL

I 'Figure 9.D-2b Qualitative Radial Concentration Profile from Plume Centerline to Shell at MidElevation

Revision I 9.D-335956r1 -9d.v6pd-041404


-- --

CL

L-

._

a.41)

a,

4-i-

0

C-

LL

Q

L-

a)CL

0-

-4-I

a)EC

~4-J

00

K>_

Time after Break (seconds)

Figure 9.D-3 Period of Application of Analytical Models for Turbulent Wall Boundary Layer andTurbulent Bouyant Plume

9.D-34 Revision I5956r1-9d.wpd-041404


Containment shell inside surface

r

AZIControl Volume

-4-

4-

Bulk FluidRegion 2

1I -4-

4-

4-

C",

C = concentration of air

T = temperature,,

qw = heat fluxU = velocity

4- Turbulent Air/Steam Boundary LayerRegion 3

Laminar Condensate film

K�) Figure 9.D4 Turbulent Buoyant Plume Analytical Model

Revision I5956ri-9d.wpd-041404

9.D-35

WCAP- 15862APP-SSAR-GSC-588 API1000

Springline -

Elevation

SourceElevation

'j

Operating Deck

Figure 9.D-5 LST Measurement Locations


WVCAP-1 5862APP-SSAR-GSC-588 API000

(a,c)

KJ Figure 9.D-6 LST Temperature Profile -Test 220.1 @ Quasi-Steady Conditions

Revision I5956rl-9d.wpd-041404

9.D-37


(ac)IJ

Figure 9.D-7 LST Temperature Profile - Test 217.1 @ Quasi-Steady Conditions V~9.D-38 Revision I

5956rl -9d.wpd-04 1404

Section 10

Nominal Inputs andCorrelations Sensitivities

5956r] -10.wpd-04 1404

---


TABLE OF CONTENTS

LIST OF TABLES ........................................................ iii

LIST OF FIGURES ........................................................ iii

10.1 INTRODUCTION ...................................................... 10-1

10.2 SENSITIVITY STUDY RESULTS ........................... 10-110.2.1 Sensitivity Study Cases for LOCA ....................................... 10-510.2.2 Sensitivity Study Case for MSLB ....................................... 10-10

10.3 CONCLUSIONS .. 10-10

10.4 REFERENCES .. 10-11

LIST OF TABLES

Table 10-1 Nominal Inputs and Correlations Sensitivity Results ......................... 10-2

LIST OF FIGURES

Figure 10-1 LOCA Sensitivities to Nominal Inputs and Correlations ...... ................ 10-3

Figure 10-2 MSLB Sensitivity to Nominal Inputs and Correlations ........................ 104

Figure 10-3 Comparison of Evaluation Model and Nominal Mass Release .................. 10-8

Figure 10-4 Comparison of Evaluation Model and Nominal Energy Release ................ 10-9

Revision I iii5956ri-10.pd-041404

WVCAP-15862APP-SSAR-GSC-588 API000

10.1 INTRODUCTION

The input values for the WGOTHIC Evaluation Model have been biased to ensure a conservative prediction'of containment pressure. A subset of these parameters has been selected to determine the impact of eachparameter on the calculated pressure for the LOCA and MSLB transients, and to provide a quantification

of the total conservatism in the Evaluation Model associated with these parameters. For these parameters,nominal values have been assumed.

Seven sensitivities were run for the LOCA transient to determine the additive sensitivity to each parameter.One sensitivity was run for the MSLB transient, which was a composite of six ofthe parameters investigatedfor the LOCA.

The results of these studies show that there is over 11.5 psi margin inherent in the AP600 ContainmentEvaluation Model for the LOCA peak pressure calculation due to the parameters studied. There is over 13psi margin associated with the post-blowdown peak pressure, and for the final nominal case, the maximum

pressure shifts to the blowdown phase. There is at least 4.9 psi margin in the MSLB calculation due to theparameters investigated. If nominal mass and energy releases were assumed for the MSLB case, even moremargin would be shown.

10.2 SENSITIVITY STUDY RESULTS

An estimate of the amountofconservatism, quantified as the change in containment pressure and representedby some of the significant assumptions made for the Evaluation Model follows. These are not single-effectsensitivities. These results are cumulative, in that each additional modification is stacked upon those thatimmediately preceded it. These sensitivities provide insight into the effect of each parameter individuallyby comparison to the preceding case, as well as the total conservatism represented by these parameters..

The AP600 Containment Evaluation Model calculations for the LOCA and MSLB described in Section 4were used as the basis for these studies, with the parameter changes described below, made in each'

succeeding case. Table 10-1 summarizes the basis for each case and the calculated pressure results for eachcase. Seven sensitivity cases were analyzed for the LOCA transient, as described below. Figure 10-1 showsa composite pressure curve forthe LOCA cases. One MSLB sensitivity was run including all the parameters

investigated for the LOCA, except for nominal mass and energy releases. Figure 10-2 shows the pressure

results for the MSLB case.

Revision I 10-15956r1-1O.wpd-041404

---------- L_


Table 10-1 Nominal Inputs and Correlations Sensitivity Results

LOCA SENSITIVITIES

Post-BlowdownBlowdown Peak Pressure atPressure Pressure 24 Hours

Case Case Description (psig) (psig) (psig)

AP600 Evaluation Model 34.4 43.9 18.9

I leat & Mass Transfer Multipliers on the 34.4 42.5 16.8Containment Shell,

2 Nominal Initial and Ambient Conditions 33.3 39.8 11.0(plus case 1)

3 Nominal Clime Material Properties (plus 33.3 39.0 10.4Cases 1, 2)

4 Nominal Steel-to-Concrete Gap 33.3 38.8 10.4Thickness (plus Cases 1,2, 3)

5 Nominal External Annulus Loss 33.3 38.7 10.2Coefficients (plus Cases 1, 2, 3,4)

6 Condensation on Dead-Ended 33.3 36.7 10.1Compartment Heat Sinks Considered(plus Cases 1, 2,3,4, 5)

7 Nominal Mass and Energy Releases 32.4 30.6 9.2(plus Cases 1,2,3,4,5, 6)

Total Conservatism Represented by 2.0 13.3 9.7Above Assumptions for LOCA.Transient

_MISLB SENSITIVITY

PeakPressure

Case Case Description (psig)

Evaluation Model . 44.8

8 Nominal Inputs and Correlations with 39.9Conservative Mass and Energy Releasesand Stratification Bias

. Conservatism Shown 4.9

KI

10-2 Revision I5956rl -I O.Apd-01404

-o 50

On 40 10 ,0 0,0 0,0

40

~ 20

10

0.

-I*1001,6000 10,000 100,000UTme (sec)

Evaluation Nominal Heat and Mass Transfer Nominal Initial Nominal Clime MaterialModel Multiplier Condition Properties

Nominal Steel-Concrete Gap Nominal External Annulus Turn On Heat Transfer to Nominal Mass and EnergyThickness Loss Coelfldents Dead-ended Compartments Release

................

I


/

M S L BM S L B

_ -_-- MSLB

50 -_

- 40-en ,/_- 7e A r

S e n sEva I ua l iNomi no I

i t i v i tyon ModelCond i t i on

Ana I ys i s

Cn

a)W

nL

J'J

20

1 0

'J

02000 3000

Time (seconds)5000

Figure 10-2 51SLB Sensitivity to Nominal Inputs and Correlations

104 Revision I5956rl-1lO.wpd-041404

WCAP-1 5862APP-SSAR-GSC-588 APlOOO

10.2.1 Sensitivity Study Cases for LOCA

Case I considered nominal heat and mass transfer on the containment shell. Conservative multipliers havebeen applied to the heat and mass transfer coefficients calculated on the inside and outside surface of thecontainment shell in the Evaluation Model to ensure a conservative calculation. 'A multiplier of[ isassumed on the inner surface, where free convection only is considered, and a multiplier of [ ]"I isassumed on the outer surface, where mixed convection is considered. For this sensitivity, both multiplierswere assumed to be 1.0, which represents a more nominal fit to the heat and mass transfer data with onlya slight conservative bias remaining. (Ref. 10.1). The heat transfer regime assumptions were not changed.

The nominal heat and mass transfer modeling resulted in a 1.4 psi reduction in the post-blowdown peakpressure and a'2.1 psi reduction in the pressure at 24 hours. There was no significant impact on the peakblowdown pressure, since the heat removal from the shell has a small impact during this time period.

Case 2 considered the impact of nominal initial and ambient conditions. The parameters discussed inSection 5 were varied for this case, except they were set to a nominal set of conditions. The sensitivity caseused initial conditions inside containment of 900F, 14.7 psia, and 30 percent relative humidity as comparedto 120 0F, 15.7 psia, and 0 percent relative humidity in the Evaluation Model. The sensitivity case usedambient conditions of 70'F and 50 percent relative humidity compared to I 15OF and 22 percent relativehumidity in the Evaluation Model. The sensitivity case PCS water temperature was set to 700F comparedto 1201F in the Evaluation Model. The initial temperature assumptions for the conductors in thecontainment 'and annulus were consistent with their environment. The individual contribution of each ofthese parameters is discussed in Section 5.

As discussed in Section 5, the internal containment conditions and the PCS water temperature have the mostimpact on containment pressure. This sensitivity case showed a 1.1 psi reduction in containment pressureduring blowdown, a 2.7 psi reduction in the post-blowdown phase, and a'5.8 psi reduction at 24 hours.

Case 3 added the assumption of nominal material properties (thermal conductivity and emissivity) in theclime modeling. In the Evaluation Model the emissivities were reduced by 10 percent to bound the rangefound in the literature'for the materials used, and the conductivity of the carbo-zinc paint was reduced bya factor of four to conservatively account for the effects of oxidization. In this sensitivity case the 0.9multiplier on the emissivities was removed and the conductivity of the carbo-zinc paint was set to its nominalvalue.

The nominal clime property modeling resulted in a 0.8 psi reduction in the peak post-blowdown pressure anda 0.6 psi reduction in the pressure at 24 hours. There-was no significant impact on the peak blowdownpressure, since the heat removal from the shell has a small impact during this time period.

K>Revision 1 10-55956rH -10.wpd-041404


Case 4 modified the assumption of the gap between the steel liner and the concrete on the applicable internalheat sinks.

The steel surface plates are one-half inch thick and are connected to each other by vertical trusses at 30-inchcenters. The steel surface plates are connected to the concrete by embedding these trusses and also bysix-inch long welded studs. The surface plates are stainless steel for the IRWST and refueling canalboundaries and are carbon steel elsewhere. The shear studs on the stainless steel plates are spaced at10 inches horizontally and 8 inches vertically. The shear studs on the carbon steel plates are spaced at10 inches horizontally and 9.6 inches vertically. The welded studs plus trusses result in a direct steelconduction path across the interface between the surface plate and the concrete. The contribution of this steelwith much higher thermal conductivity has been neglected in the WGOTHIC Evaluation Model

Concrete is placed into the structural modules and will initially bond to the surface plates. The wet concreteweight will load the steel plates and result in permanent outward deformation of the plates spanning betweenthe trusses. Mechanical and thermal loading and shrinkage may break this bond due to relative motion.However, the welded studs will keep the surface plates in contact with the concrete at the stud locations andgaps between the studs would be very small. The Evaluation Model conservatively assumes a five mil gapfor the steel-jacketed concrete heat sinks. A one mil gap was assumed for this sensitivity.

The results of this case show a 0.2 psi reduction in the post-blowdown peak pressure. The blowdown peakpressure was not affected significantly, since the heat capacity of the concrete has little impact on energyremoval during this time. The pressure at 24 hours was not affected, since the heat sinks are effectivelysaturated at this time and the pressure is dictated by the balance between releases to the containment and heatremoval via the PCS.

Case 5 included a modification to the pressure loss coefficient in the external annulus. The loss coefficientassumed in the Evaluation Model includes a 30 percent increase over the value derived from the test program.For this study, a loss coefficient 10 percent greater than measured was modeled.

The annulus loss coefficient has a small impact on the pressure, reducing the post-blowdown peak pressureby 0.1 psi and the pressure at 24 hours by 0.2 psi. There was no significant impact on the peak blowdownpressure since the heat removal via the PCS has a small impact during this time period.

Case 6 included the effects of heat sink utilization in the dead-ended compartments. As discussed inSection 9, condensation and convection on the heat sinks in the dead-ended compartments are not creditedaflerblowdown to ensure aconservative treatment oftheirutilization. With onlya small thermal asymmetry,convection would cause these heat sinks to continue to be exposed to steam after this time and wouldcontinue to provide a condensation surface. In this case, the structures were allowed to absorb energy basedon conditions predicted by the WGOTHIC code. The WGOTHIC calculation limits the amount of steamentering these compartments, since circulation does not occur in lumped volumes, so the utilization of these

10-6 Revision I5956rd -l 0.%,pd-041404


heat sinks is not expected to be grossly over-predicted. It is therefore acceptable to include the utilization

of these heat sinks in these sensitivities.

The additional heat absorption capability of the heat sinks in the dead-ended compartment, with an initialtemperature of 900F, resulted in a 2.0 psi reduction in the post-blowdown peak pressure. There was nochange in the blowdown peak pressure since the Evaluation Model already considers condensation on thesesurfaces during blowdown. The reduction in the pressure at 24 hours was only 0.1 psi, since the heat sinksare effectively saturated at this time in both cases.

Case 7 includes the use ofnominal mass and energy release rates. The nominal mass and energy release ratesare compared to those used in the AP600 Evaluation Model (described in Section 4) in Figure 10-3 andFigure 10-4, respectively. The following modifications were made in the mass and energy releasecalculation:

* Nominal full-power temperatures and pressure without uncertainty

* Nominal RCS volume with no uncertainties (thermal expansion was considered)

* Core licensed power without adding 2 percent for calorimetric error

. Nominal core stored energy without adding 15 percent

* 1979 ANS decay heat standard without uncertainty for an 800 day average bumup

* Steam generator energy was assumed to be released over a period of one hour for the broken loop andtwo hours for the intact loop rather than thirty minutes and one hour, respectively

* The refill period was modeled

* No heat generation due to zirc-water reaction

* Nominal initial accumulator, CMT, and IRWST fluid temperatures were used in the post-blowdownreleases

Revision 1 10-75956r1 - O.wpd-041404


,

Mass ReleaseNomi no I

---- Evaluation Model

Rate Comparison

510

41 0

0

0

0

a0

0

a,a,

ra

:310

21 0

I10

01 0

K>

I 10 10

T ime ( sec )

Figure 10-3 Comparison of Evaluation Model and Nominal Mass Release

10-8 Revision I5956r]-I O.wpd-041404


E n e

_ 7 - -

rgy Re I easeNomi no IEvoluot ion Model

Roate Comparison

8m 10

I-

cno

- 60 1 0

4)

en

° 1 0

a: . 4

-10

c 3L 10

* ~ . 1

-- X .'. .I

4 , - - --* W4J I' ~ 'WU W Wi

0

1 0 102

1 0 10

T-i me10

( s e c )5

1 06

1 07

1 0

Figure 10-4 Comparison of Evaluation Mf odel and Nominal Energy Release

Revision I5956rl-10.wpd-041404

10-9


This case demonstrates that the mass and energy input to the containment has the most impact on theresulting pressure and represents significant conservatism in the Evaluation Model. The blowdown peakpressure was reduced by 0.9 psi, the pressure during the post-blowdown period was reduced by 6.1 psi, andthe pressure at 24 hours was reduced by 0.9 psi. The decrease in pressure is sufficient to shift the time ofpeak pressure for the entire transient from the post-blowdown phase to the blowdown phase. The largesensitivity during the post-blowdown phase is expected, since the mass and energy release models aredeliberately biased to maximize the releases during this time in order to ensure that a conservative peakcontainment pressure is calculated.

10.2.2 Sensitivity Study Case for MISLB

A case was also run to provide quantification of the conservatism in some of the parameters for the MSLBtransient. Case 8 included the nominal inputs and correlations embodied in cases I through 6 of the LOCA.Nominal mass and energy releases were not included in this calculation. The results of this case indicate thatthe conservative assumptions studied in the sensitivity case constitute 4.9 psi conservatism. Furtherreduction in containment pressure would result from the use of nominal mass and energy releases and a lessconservative stratification bias. These results also do not credit the significant improvement to shell masstransfer, which would result from forced convection.

10.3 CONCLUSIONS

The results of the sensitivities in this section illustrate that the Containment Evaluation Model provides aconservative prediction of pressure. Some of the conservative nominal inputs and correlations in theEvaluation Model were removed to determine the impact on the calculated pressure for AP600 for thelimiting LOCA and MSLB cases. These sensitivities are not intended to portray a best estimate calculation;other conservative features and modeling techniques have not been included that would further reduce thepredicted pressure.

However, the conservatism represented by the parameters that were studied is significant. The LOCA casesinvestigated indicate a reduction of 2.0 psi during blowdown, 13.3 psi during the post-blowdown phase, and9.7 psi at 24 hours. The summation of these cases switches the time of peak pressure from the post-blowdown period to the blowdown period. Comparison of the maximum pressures in the two cases shows11.5 psi reduction in pressure. The MSLB case shows 4.9 psi margin due to nominal inputs included in thisstudy. Incorporating other nominal conditions, such as mass and energy release rates and forced convectionenhancements, would result in further pressure reduction.

10-10 Revision I5956rl-IO.wpd.041404

WCAP-15862APP-SSAR-GSC-588 APIOO0

-- --

10.4 REFERENCES

10.1 WCAP-14326, Revision 2, "Experimental Basis for the AP600 Containment Vessel Heat and MassTransfer Correlations," April 1998.

Revision I 10-11

Revision I5956rl -10O.wpd-04I404

]0-11

Section 11

Timestep Sensitivity

5956r -I lu.pd-040204


TABLE OF CONTENTS

LIST OF TABLES ..................... iii

LIST OF FIGURES ..................... iii

11.1 INTRODUCTION ................. 11-1

11.2 METHODOLOGY ................... 11-1

11.3 RESULTS ................... 11-2

11.4 SUMMARY ................... 11-3

11.5 REFERENCES ................... 11-6

LIST OF TABLES

Table I 1-1 Timestep Sensitivity Results .11-2

Table 11-2 Timestep Limit Results .11-3

LIST OF FIGURES

Figure 11-1 AP600 Containment LOCA Transient Pressure Prediction

with Full, One-Half, and One-Quarter Timesteps .11-4

Figure 11-2 AP600 Timestep Comparison, Percentage Difference

in Predicted Pressures Between Successive Code Versions .11-5

K>Revision I5956rd-I I.4pd-040204

iii


11.1 INTRODUCTION

To establish WGOTHIC as an acceptable tool for use in the licensing process, it is necessary to perform asensitivity to the timestep size selected by the code. The timestep size determines how far the transient ispermitted to progress from calculational step to calculational step. The algorithm that is used to determinethe maximum allowable timestep size should produce a timestep value that results in a stable, suitablyaccurate solution without prohibitive computer execution times.

In the WGOTHIC code, the algorithm that is used to-determine the maximum allowable timestep size isbased on two stability criteria: the amount of time it would take to completely replace the mass within agiven volume (Courant limit), and one half of the natural period of oscillation for a gravity-driven system(gravitational limit). In addition, the algorithm also includes limits on timestep growth, checks onnonphysical results, and limits on the rates of change of primary variables. These limits are all discussed inReference 11.1, Enclosure 2, Section 12.7.

11.2 METHODOLOGY

Subroutine timstp.f in the WGOTHIC solver program contains the timestep selection algorithm. Thissubroutine calculates the maximum allowable timestep size that can be used in the WGOTHIC transientcalculations.

Subroutine timstp.f was modified to perform the timestep sensitivity study. The maximum allowabletimestep size was reduced to study the effect of using a smaller timestep on the results of the WGOTHICtransient calculations. Use of timesteps that are smaller than the maximum allowable value maintainsnumerical stability.

Twvo new versions of subroutine timstp.f wvere created; in WGOTHICS version 4.1.1, the maximumallowable timestep size was halved, and in version 4.1.2, it was quartered. The timestep selection processwas not impacted by these changes and there were no other differences between the three versions ofWGOTHICS.

The AP600 containment evaluation model input deck described in Section 4 was used for the timestepsensitivity analyses. The loss-of-coolant-accident (LOCA) transient was chosen since it contains both rapidand extended pressurization and depressurization transients. The LOCA is a long transient that relies on thePCS for a substantial amount of energy transfer to the environment.

Revision I 11-15956r -Il .Lpd-040204


11.3 RESULTS

The LOCA transient containment response analysis was performed with each of the three code versions.Table 11-I indicates the number of calculational steps taken, the peak predicted pressure, and the predictedpressure at 24 hours for each code version.

Table 11-1 Timestep Sensitivity Results |

Number of Peak Pressure Pressure atWGOTHIIC S Code Version Calculational Steps (Psig) 24 Hours (psig)

Version 4.1 - l.0*At 187380 43.85 18.86

Version 4.1.1 - 0.5*At 374414 43.76 18.88

Version 4.1.2 - 0.25*At 748680 43.72 18.85

The one-half timestep version of the code takes approximately twice the number of steps to complete thetransient and the one-quarter-timestep version of the code takes about four times the number of steps tocomplete the transient. Since the timnestep selection logic relies on the volume conditions at the end of eachstep, it is extremely difficult to get exactly twice or four times the number of steps. With the smallertimesteps, the changes in fluid conditions and other parameters used in the timestep selection process aresmaller. The smaller changes permit larger timesteps to be allowed. As a result, slightly less than two timesand four times the number of calculational steps are used in the modified versions of the code.

The reduced timestep versions predicted slightly lower peak pressures than the base version of the code. Thepredicted peak pressure for the quarter-timestep version was 0.04 psi lower than the one-half timestepversion, and 0.13 psi lower than the full-timestep version. This indicates that the calculated peak pressuresolution converges from above as the timestep is reduced.

In addition to the number of timesteps taken to complete the problem, the code output contains the numberof times that a timestep was limited by some phenomenon. In all three cases, the timestep size was limitedby either the phase change limit, the Courant limit, or the gravitational limit. Table 11 -2 presents the numberof times that the timestep was limited in each case.

11-2 Revision 1

11-2 Revisiond I5956rl-l1 .%-pd{040204


Table 11-2 Timestcp Limit Results

WGOTIIIC-S Code Version Courant Limit Phase Change Limit Gravitational Limit

Version 4.1 - 1.0 * At 4000 21 91612

Version 4.1.1 - 0.5 * At 8070 I 183305

Version 4.1.2 - 0.25 * At 16108 9 366639

During the pressurization phase of the LOCA transient, the timestep size was primarily limited by theCourant stability limit. This was not unexpected, since the volumetric break flow is so large during thisphase. Since the circulation is primarily buoyancy-driven during the much longer depressurization phase,the time step size was primarily limited by the gravitational stability limit after the peak pressure wasreached.

Note, the sum of the number of limiting timesteps in Table 11-2 does not add up to the total number ofcalculational timesteps given in Table 11-1 because, in some instances, the timestep size was not limited byany of the limits tracked in the output file. This could occur, for example, if the upper limit imposed by userinput was less than the value calculated by the timestep selection algorithm.

The LOCA transient containment pressure predicted by each of the three code versions is shown inFigure 11-1. There is very little difference in the results. The transient results are further examined bycomparing the differences in the predicted transient pressure from each of the three code versions. Thepercentage difference in the predicted transient pressure between the full-timestep version and the one-half-timestep version, and the percentage difference in the predicted transient pressure between the one-halftimestep version and the one-quarter-timestep version is presented in Figure 11-2. The differences in thepredicted transient pressure between successive versions of the code are very small (less than 1% over theentire 24-hour period).

11.4 SUMMARY

Two versions of WGOTHICS were created for this sensitivity study. The calculational timestep size wasreduced to a value that was either one-half (Version 4.1.1) or one-quarter (Version 4.1.2) of the maximumallowable value calculated by the timestep selection algorithm.

A comparison of the predicted AP600 LOCA transient containment pressure response between the baseversion and the two versions created for the sensitivity study shows that there is very little difference in theresults. Using smaller timesteps does not significantly affect the transient response. Therefore, the timestepselection algorithm produces a timestep value that results in a stable, suitably accurate solution.

Based on the timestep study results presented herein, it is concluded that the timestep selection logic forWGOTHICS is acceptable.

Revision 1 11-35956ri-1 I.vpd-040204

1 .

0 >

_. -V

C ra 0

CA

CA

nI--q,,

Cj

I>

3> Z7

C) "

900'-'

T i me sFull

- - - - Half- -- - Ouor

t epT i me s t eTimes te

ter Time

S e n spps

(Ve r(Ve r

tep ( V

44e

i

r

1I

t

4

i V I t y S t u d y

I1.2)

C.)

a'Cn

CL

50

40

30

20

10

0

10LA

0%

4-.

kO

I~J

i :,.6 _

2 ,3 410 1i0 10

Time (seconds)

51 0

( C

( ( C

u.

t'J

T :.= 0

C. T i me s- Fu I- - Ha I f

t e pto Ha Ito Qua

S e n s t i

lpEn

(bwoo~oo0000

v i t y S t udyrr

Times t epter Timestep

e8 0o qc. 3

-f:

e g

U .,#0

a-

°U

.6

a

u._

C

cm

0)

0)

(nCn

C3-

.4

.2

0

-.-2

-.-4I

1 02 3 4

10 10 105

10

T ime ( seconds )u.

>040


11.5 REFERENCES

11.1 Westinghouse Letter NTD-NRC-95-4563, B. A. McIntyre to Quay (NRC), "GOTHIC Version4.0 Documentation," September 21, 1995.

11-6 Revision 1

11-6 Revision I5956r]l- I.vpd 040204

Section 12

Sensitivity to Clime Noding

5956r1-12.%wpd-040204


TABLE OF CONTENTS

LIST OF TABLES ............... ; iii

LIST OF FIGURES ...... iv

12.1 INTRODUCTION .. 12-112.1.1 Background . 12-112.1.2 Problem Statement . 12-312.1.3 Approach . 12-312.1.4 Selected Parameters . 12-412.1.5 Success Criteria . 12-4

12.2 MODEL DESCRIPTIONS . 12-512.2.1 Simple Annulus Clime Model . 12-512.2.2 AP600 Containment Model . 12-7

12.3 RESULTS ............................. 12-812.3.1 Simple Annulus Clime Model .......................... 12-8

<2 12.3.2 AP600 Containment Model .......................... 12-10

12.4 SUMMARY .......................... 12-1212.4.1 Simple Annulus Clime Model Noding Study .......................... 12-1212.4.2 AP600 Containment Model Clime Noding Study .......................... 12-12

12.5 REFERENCES ........................................ 12-12

LIST OF TABLES

Table 12-1 Input Parameters for Annulus Clime Model Sensitivity Study . 12-6Table 12-2 Predicted Heat Removal Rates for Various Clime Noding Schemes .............. 12-9

Revision . iii5956rl -1 2.wpd-040204

WCAP-15862APP-SSAR-GSC-588 APIO0O

LIST OF FIGURES

Figure 12-1Figure 12-2Figure 12-3Figure 124Figure 12-5Figure 12-6Figure 12-7Figure 12-8Figure 12-9Figure 12-10Figure 12-11Figure 12-12Figure 12-13Figure 12-14Figure 12-15Figure 12-16Figure 12-17Figure 12-18Figure 12-19Figure 12-20Figure 12-21Figure 12-22Figure 12-23Figure 12-24Figure 12-25Figure 12-26Figure 12-27Figure 12-28Figure 12-29Figure 12-30Figure 12-31Figure 12-32Figure 12-33

Figure 12-34

Figure 12-35

Westinghouse-GOTHIC Clime Wall Source Term Models .... ............... 12-13Simplified Line Diagram of Annulus Clime Model ..... .................... 12-14Noding Diagram, 8 Clime Node Model ....... ............................ 12-15Comparison of Transient Heat Transfer Rates; Case I ..... .................. 12-16Comparison of Transient Heat Transfer Rates; Case 2 ..... .................. 12-17Comparison of Transient Heat Transfer Rates; Case 3 ..... .................. 12-18Comparison of Transient Heat Transfer Rates; Case 4 ..... .................. 12-19Comparison of Transient Heat Transfer Rate Differences; Case I .... .......... 12-20Comparison of Transient Heat Transfer Rate Differences; Case 2 .... .......... 12-21Comparison of Transient Heat Transfer Rate Differences; Case 3 .... .......... 12-22Comparison of Transient Heat Transfer Rate Differences; Case 4 .... .......... 12-23Comparison of Heat Flux Profiles; Time t = 2000 Seconds; Case I .... ......... 12-24Comparison of Heat Flux Profiles; Time t = 2000 Seconds; Case 2 .... ......... 12-25Comparison of Heat Flux Profiles; Time t = 2000 Seconds; Case 3 .... ......... 12-26Comparison of Heat Flux Profiles; Time t = 2000 Seconds; Case 4 .... ......... 12-27Comparison of Film Temperature Profiles; Time t = 2000 Seconds; Case I ...... 12-28Comparison of Film Temperature Profiles; Time t = 2000 seconds; Case 2 ...... 12-29Comparison of Film Temperature Profiles; Time t = 2000 seconds; Case 3 ...... 12-30Comparison of Film Temperature Profiles; Time t = 2000 seconds; Case 4 ...... 12-31Comparison of Air Temperature Profiles; Time t = 2000 seconds; Case I ... ..... 12-32Comparison of Air Temperature Profiles; Time t = 2000 seconds; Case 2 ... ..... 12-33Comparison of Air Temperature Profiles; Time t = 2000 seconds; Case 3 ... ..... 12-34Comparison of Air Temperature Profiles; Time t = 2000 seconds; Case 4 ... ..... 12-35AP600 Containment Model Clime Noding Pattern ..... ..................... 12-36AP600 Containment Model Double Vertical Clime Noding Pattern ... ......... 12-37AP600 Containment Model Double Stack Clime Noding Pattern ............. 12-38AP600 Containment Model Double Mesh Point Clime Noding Pattern ... ....... 12-39Pressure History, AP600 Containment Model; Double Clime .... ............. 12-40Wet Heat Flux vs. Clime; AP600 Containment Model, Base Case .... .......... 12-41Wet Heat Flux vs. Clime; AP600 Containment Model, Double Clime ... ........ 1242Dry Heat Flux vs. Clime; AP600 Containment Model, Base Case .... .......... 12-43Dry Heat Flux vs. Clime; AP600 Containment Model, Double Clime ... ........ 12-44Film Temperature vs. Clime; AP600 Containment Model,Base Case.......................................................... 1245Film Temperature vs. Clime; AP600 Containment Model,Double Clime....................................................... 1246Dry Surface Temperature vs. Clime; AP600 Containment Model,Base Case.......................................................... 1247

iv Revision I5956rl-12.wpd-04024


LIST OF FIGURES (Cont.)

Figure 12-36

Figure 12-37

Figure 12-38

Figure 12-39Figure 12-40Figure 12-41Figure 1242Figure 12-43Figure 12-44Figure 12-45

Figure 1246

Figure 1247

Figure 12-48

Figure 12-49

Figure 12-50

Figure 12-51

Figure 12-52

Dry Surface Temperature vs. Clime; AP600 Containment Model,Double Clime ................................................... 1248Annulus Pressure vs. Clime; AP600 Containment Model,Base Case .......................................................... 1249Annulus Pressure vs. Clime; AP600 Containment Model,Double Clime ................................................... 12-50Air Temperature vs. Clime; AP600 Containment Model, Base Case ..... ....... 12-51Air Temperature vs. Clime; AP600 Containment Model, Double Clime .... ..... 12-52Air Density vs. Clime; AP600 Containment Model, Base Case ..... ........... 12-53Air Density vs. Clime; AP600 Containment Model, Double Clime ..... ........ 12-54Air Velocity vs. Clime; AP600 Containment Model, Base Case ..... .......... 12-55Air Velocity vs. Clime; AP600 Containment Model, Double Clime ..... ....... 12-56Heat Rejection History Comparison, AP600 Containment Model;Double Clime Sensitivity Case .................. ....................... 12-57Integrated Heat Rejection Comparison, AP600 Containment Model;Double Clime Sensitivity Case .................. ....................... 12-58Pressure History Comparison, AP600 Containment Model;Double Stack Sensitivity Case . ......................................... 12-59Heat Rejection History Comparison, AP600 Containment Model;Double Stack Sensitivity Case . ......................................... 12-60Integrated Heat Rejection Comparison, AP600 Containment Model;Double Stack Sensitivity Case . ......................................... 12-61Pressure History Comparison, AP600 Containment Model;Double Mesh Sensitivity Case . ......................................... 12-62Heat Rejection History Comparison, AP600 Containment Model;Double Mesh Sensitivity Case . ......................................... 12-63Integrated Heat Rejection Comparison, AP600 Containment Model;Double Mesh Sensitivity Case . ......................................... 12-64

K>Revision I5956rl -12.wpd-040204

v


12.1 INTRODUCTION

This section provides background associated with the WGOTHIC clime model, a statement of the issueassociated with the application of the clime methodology, a description of the approach developed and takento address the issue, a summary of the results of the analyses and the conclusions drawn from a review ofthose analyses.

12.1.1 Background

The GOTHIC code (Refs. 12.1 and 12.2) is a state-of-the-art program for modeling multi-phase flow forcontainment analysis. The code solves the integral form of the conservation equations for mass, momentum,and energy for multi-component, two-phase flow. The conservation equations are solved for three fields:continuous liquid, liquid drops, and a steam/gas phase. The three fields may be in thermal non-equilibriumwithin the same computational cell. Relative velocities are calculated for each field, as well as the effectsof two-phase slip on pressure drop. Heat transfer between the phases, surfaces, and the fluid are alsoallowed.

As described in Section 3 of this report, the GOTHIC containment analysis code was modified byWestinghouse to include mechanistic convective heat and mass transfer correlations, a liquid film trackingmodel, a one-dimensional wall conduction model, and wall-to-wall radiant heat transfer to model heatremoval by the PCS. The code with these modifications is called Westinghouse-GOTHIC, and is abbreviatedas WGOTHIC.

A solution technique that includes wall-to-wall radiation at the conditions expected for the AP600 andAP 1000 plant designs necessitates a close coupling ofthe participating walls. This coupling is accomplishedby assigning boundaries that define the portions of the various walls that radiate to one another. Consistentvith the basic formulation implemented for the GOTHIC code that considers conductors or heat sinks to be

energy and mass sink or source terms, code modifications that include wall-to-wall radiant heat transfer canbe thought of as the addition of a special type of conductor group. This special conductor type or groupconsists of a set of walls that radiate to each other and interface with GOTHIC fluid cells through mass andenergy source terms. The term clime, meaning region, is used to differentiate and distinguish this specialconductor type from those already existing in GOTHIC terminology.

A three-conductor clime is shown schematically in Figure 12-1. For the containment model, a clime is ahorizontal slice of the containment structure consisting of the steel shell, the baffle, and the shield buildingand liquid films which may form on those solid surfaces, and representing the following transport processes:

The heat and mass transfer source terms from the containment volume to the shell

Revision 1 12-15956ri-12.wpd-040204


Liquid film mass and energy conservation and thermal resistance on shell, baffle, or shield buildingsurfaces

Conduction through the shell

Heat and mass transfer source terms from the exterior shell to the riser air flow channel

Radiation from the exterior shell to the interior baffle

Heat and mass transfer source terms to the interior baffle from the riser air flow channel

Conduction through the baffle

Heat and mass transfer source terms from the exterior baffle to the downcomer air flow channel

Radiant heat transfer from the exterior baffle to the interior surface of the shield building

Heat transfer source terms to the interior surface of the shield building from the downcomer air flowchannel

Conduction through the wall of the shield building

Both radiant and convective heat transfer from the exterior surface of the shield building to theenvironment

As shown schematically in Figure 12-1, the internal containment vessel volume, riser air flow channelvolume, downcomer air flow channel volume, and environment volume are separate computational cells orfluid volumes in the GOTHIC model. The shell, baffle, and shield building walls are one-dimensionalconductors representing solid wall structures between the computational cells. These conductors are furthersubdivided into regions of different materials with different numerical mesh sizes.

The climes are stacked vertically through the PCS to model the effects of changing properties both insideand outside the containment shell. Usually there are at least two stacks of climes a wet stack and a dry stack.The only difference between a wet and dry stack is that a time-dependent, water flow rate boundary conditionis specified for each conductor surface of the top clime in a wet stack. Because condensation can occur oneither wet or dry conductor surfaces, an initially dryi stack of climes could contain some wet conductorsurfaces and/or a partially wet conductor surface due to condensation. Likewise, an initially wet stack ofclimes could contain some dry conductor surfaces and/or a partially dry conductor surface due toevaporation.

12-2 Revision I5956rl-42.upd-040204

WCAP-15862APP-SSAR-GSC-588 APlOOO

The user must specify values for the area and circumferential perimeter for each conductor of each clime inboth the wet and dry stacks. The input values for the area and circumferential perimeter for the climeconductors in the wet stacks are based on measurements of the water coverage from the full-scale waterdistribution tests. The water coverage input calculation method for the containment Evaluation Model isdescribed in Section 7.

The WGOTHIC clime model calculates the temperature, flow rate, and thermal resistance ofthe water filmson the various conductor surfaces of a clime. Liquid mass is conserved whenever the film reaches the bottomclime in a stack or a conductor surface dries out. The clime model takes the film flow rate from eachconductor surface of the previous clime in the stack as input, then adds the local condensation rate, orsubtracts the local evaporation rate to determine the output water flow rate on each of its correspondingconductor surfaces. Any liquid film remaining on the conductor surfaces of the last clime in a stack is addedto the liquid field of the WGOTHIC cell in contact with the conductor surface, or an alternate drain cellspecified by user input.

Dryout occurs when either the film flow rate is low enough or the heat flux is high enough to result incomplete evaporation of the film before it can exit the conductor. The clime model calculates theevaporation heat and mass transfer and the location of the dryout elevation; the remainder of the conductorsurface is treated as a dry surface.

Kt The details of the clime equations and integration into the GOTHIC code are described in Sections 3.4 and3.5.

12.1.2 Problem Statement

Simply stated, the objective of this effort is to demonstrate that the clime calculations are insensitive toincreased numerical resolution relative to the noding pattern employed in the Evaluation Model.

12.1.3 Approach

The approach taken to address the problem statement given above is to:

Perform sensitivity calculations using a simple two-channel annulus model with a constanttemperature boundary condition. The simplified model is used to better isolate effects of climenoding. For this task, the number of climes in the model will vary from 4 to 16. Although notintended to simulate or scale to the AP600 or API 000 plants, this two-channel annulus model willprovide for thermo-fluid conditions similar to those expected in the annulus ofthe passive plant, i.e.,evaporation into a buoyancy-driven air flow.

Perform sensitivity calculations with an AP600 containment model for which the number of climes,stacks, and conductor layers are varied to confirm the results of the simpler models.

Revision 1 12-35956r -1 2.wpd.040204


, .

12.1.4 Selected Parameters

The following thermal-hydraulic parameters are used in evaluating the comparison of various modelpredictions:

Containment pressures (for the AP600 model only)

Temperatures - cooling air and liquid film

* Heat fluxes and/or heat rates from the shell

These parameters are primary indicators of the heat transfer process and were selected as a basis forevaluating the sensitivity of the calculated results to noding patterns.

12.1.5 Success Criteria

The noding pattern used for a model may affect the results of a calculation. For the purpose of this study,the following success criteria are used to evaluate the significance of change in calculated results withincreasing detail in clime noding:

Success Criteria: The change between results calculated with two noding patterns is defined to benegligible; if:

1) The variation in results between two successive noding patterns is less than[ ]'ll percent.

2) The variation in results from successively finer noding patterns isdecreasing.

]B~

Thus, the criteria listed above establish that the variation between results obtained from different nodingpatterns must be small, and the variation must be converging as the noding pattern is increased.

I 2-4 Revision 1

124 Revision I595 6ri -I2.%-pd-040204


12.2 MODEL DESCRIPTIONS

The sensitivity of the thermo-fluid calculations ofthe WGOTHIC clime methodology to the numberofnodesassociated with the flow channel was investigated using a two-channel annulus model. This section presentsdescriptions of the model and the boundary conditions used in this investigation. This simplifiedtwo-channel model is not intended to scale to the plant. It is a simplified model with fixed boundaryconditions that allows study of changes caused by clime noding detail. The fixed boundary conditions allow

direct comparison of clime heat removal predictions. A detailed comparison of the results calculated for thismodel to those calculated in the AP600 or AP1000 is not meaningful. However, the trends for the simplemodel, relative to the success criteria, are applicable to the passive plant design, as confirmed with thesensitivity calculations using an AP600 containment model as a basis (Section 12.3.2).

12.2.1 Simple Annulus Clime Model

Model Description

A simple two-channel model was developed to study the effects of increasing the number of climes over awide range of film flow rates and film temperatures and is shown schematically in Figure 12-2. The modeledheated height was [ ]3' feet. This height was chosen to promote the calculation of velocities in the modelrepresentative of the lower bound of the range expected for AP600. The number of climes (andcorresponding annulus cells) in the heated section of the model varied between 4 and 16. Figure 12-3presents the noding structure for the 8 clime model. The noding pattern for the 4 and 16 clime models is

similar to that of the 8 clime model, having one-half and double the number of axial cells, respectively.

The annulus clime model is connected to two stacks of lumped parameter cells, similar to the EvaluationModel. One stack represents the don comer volume and the other represents the riser volume. The volumesof the riser and downcomer are [ a]c cubic feet and [ ]J cubic feet respectively. The downcomervolume was arbitrarily selected to be about twice the value of the riser volume. A set of equally spacedelevation planes crosses both the riser and downcomer to form the two stacks of lumped parameter cells.The volume, height, and vertical flow area of each cell in the riser stack and each cell in the downcomer stack

is the same.

A natural draft flow of air from the downcomer through the riser develops as the riser channel is heated.Friction acts to retard the increase in air velocity. Except for the turning location and exit, the frictionlengths for each flow path are equal to the cell height. The friction lengths at the riser entrance and exit areset to one half the cell height to conserve the total friction length and fL/D values between models. Losscoefficients of[ ]' at the downcomer entrance and [ ]' at the riser outlet are used to model the formlosses representative of a contraction and an expansion, respectively.

Revision 1 12-55956r] -12.%wpd-O040204

WCAP-15862APP-SSAR-GSC-588 API OOO

For the 8 clime model shown in Figure 12-3, a thermal conductor located within the heat source (Volume 9),provides a [ ]"C constant temperature boundary condition for the model. A single stack of climes is usedto thermally connect the heat source with the riser and downcomer volumes. There is one clime per cell inthe riser. An additional clime at the bottom of the stack is used to model the runoff film flow. The last climein the stack is connected to three dummy volumes. This modeling is used to allow the runoff from the lastclime to collect in the drain volume (Volume 18) without affecting the heat removal in the active section ofthe annulus.

Each clime has two conductors; the first one represents a [ ]'¢ thick steel plate and the otherrepresents an acrylic cover. The perimeter and heat transfer area is the same for each conductor on allclimes.

To prevent the drain volume from overfilling, it is connected to a flow rate boundary condition. Theboundary flow rate for the drain volume is controlled with trips based on the liquid level in the drain volume.

These features were also included in both the 4 and 16 clime models.

Boundary Conditions

Four cases were considered for this study. The independent variables for each case are listed in Table 12-1.

-1 a>}-

I I

4 U

4 4 4 4 I

4. 4. 4. 4. I

4 4 4 4 I

I 2-6 Revision 1

12-6 Revision I5956rl -12.%%pd-040204


The film boundary conditions are selected to cover a range oftemperatures and flow rates considered typicalfor the passive plant. For Case 4, the film mass flow rate is reduced to force the prediction of dryout aboutmidway down the plate. Both the downcomer and riser are connected to a fixed pressure boundary set at

[ ]3

' psia.

12.2.2 AP600 Containment Model

A schematic of the AP600 containment model clime noding pattern is shown in Figure 12-24. The dashedlines represent divisions between the [ ]3' climes (and cells) in the annulus.

The AP600 clime noding sensitivity cases were performed using a preliminary Evaluation Model input deck.The differences between the AP600 containment model that was used for these sensitivity studies and theEvaluation Model, as described in Section 4, are irrelevant to the clime noding sensitivity study because eachofthe sensitivity cases used this same AP600 containment model as its basis. The major differences betweenthe sensitivity model and the Evaluation Model, are summarized below:

- The mass and energy releases in the sensitivity model are lower during the peak pressure phase andhigher during the long-term phase. The different phases of the LOCA containment pressurization

transient are defined in Section 3.4.2.2 of Reference 12.3. [

The PCS flow rate input is higher in the sensitivity model for the first three hours, but lower for theremainder of time.

aic

* The internal noding structure is different. [

]3,c

The annulus initial conditions in the sensitivity model (15.7 psia, 1200F) are different from thosein the Evaluation Model (14.7 psia, 1150F).

* [ :

] 3@c.

Revision 1 12-75956rd -1 2.vpd-040204


* The annulus loss coefficients are lower in the sensitivity model. [

]ac

The base case AP600 containment model was modified for the clime noding sensitivity study to examine thefollowing effects:

Doubling the number of climes (or vertical segments) from [ ]^ to [ ]. (see Figure 12-25)

Doubling the number of stacks (or radial segments) from [ ]' to [ ]^' (see Figure 12-26)

Doubling the number of numerical mesh points through the thickness of the clime conductors(see Figure 12-27)

As with the simple annulus model noding study described in Section 12.2.1, the heat transfer parametersidentified in Section 12.1.4 were used to evaluate the sensitivity of the calculations to changes in the nodingpattern used for the calculations. In addition, plots comparing the annulus air pressure, density and velocityprofiles are included for the vertical clime noding sensitivity comparison.

12.3 RESULTS

12.3.1 Simple Annulus Clime Model.

The four test cases described in Table 12-1 of Section 12.2.1 were run using WGOTHIC Version 4.1.Transient calculations were performed until the time when little or no change in one or more governingparameters (pressure, film temperature, etc.) was predicted. Since constant boundary conditions were used,a steady-state solution was eventually reached for each case. All four cases were run out to 2000 secondsof transient time.

Plots of the predicted heat removal rate from the plate surface, shown in Figures 12-4 through 12-7, indicatethat the four cases were close to steady-state conditions at the end of the 2000-second transient period. Thedifference between the heat rejection rates of the 4 and 8, and 8 and 16 clime models is shown inFigures 12-8 through 12-11. From these plots, it is noted that the predicted transient heat removal rate isquite insensitive to the level of clime noding detail; a change of less than 1 percent is observed for increasingthe axial nodal pattern from 4 to 8 climes. The difference observed when the axial nodal pattern is increased

from 8 climes to 16 is even less.

12-8 Revision I5956rl1 - 2.%Npd-040204

WCAP- 5862APP-SSAR-GSC-588 APIOOO

A comparison of the predicted heat removal rate from the plate surface at the end of the transient is shownin Table 12-2.

n a,c

I I '''' I I I

For the first three tests (all dry or all wet), the steady-state heat removal rate increases slightly as the numberof climes is doubled from 4 to 8. Doubling the number of climes from 8 to 16 increases the heat removal rateagain, but by a much smaller amount. Therefore, for these cases, the predicted heat removal rate isconverging from below.

For test case 4 (half wet), the steady-state heat removal rate decreases slightly as the number of climes isdoubled from 4 to 8. Doubling the number of climes from 8 to 16 also decreases the heat removal rate, butby a smaller amount. Further comparisons show that the predicted heat removal rate decreases on the top[ ]"' feet, but increases on the bottom [ ]' feet of the model surface as the number of climes is increased.The decrease on the top [ ]' feet is larger than the increase on the bottom [ ]' feet. Therefore, the totalpredicted heat removal decreases as the number of climes increases in case 4. At the low flow rate for thistest, the increased resolution of the subcooled heat flux with more climes yields a slightly lower estimate ofthe predicted heat removal rate.

Figures 12-12 through 12-15 compare the calculated axial heat flux profiles, and Figures 12-16 through 12-19compare the axial film temperature profiles at time t = 2000 seconds, where "film temperature" for the drycase represents the surface temperature. Note, the data from the 8 and 4 clime models is represented as 2and 4 points respectively'on the plots to match the 16 points from the 16 clime model.

For the wet tests, smoother axial heat flux and film temperature profiles are calculated as the number ofclimes increases. In the cold film tests (cases 2 and 4), the heat flux decreases and film temperature increasesas the film flows from the first clime down. In the hot film test (case 3), the heat flux increases and the filmtemperature decreases as the film flows from the first clime down. The heat flux remains constant after the

Revision I 12-9

Revision I5956rl -I2.%pd 04 0204

12-9

WCAP-15862APP-SSAR-GSC-588 API OOO

film reaches a temperature at which evaporation dominates. For the partially wet test (case 4), the heat fluxdecreases rapidly and the film temperature increases to the dry surface temperature in the clime where dryoutoccurs.

The heat flux and surface temperature profiles are adequately represented with either 4, 8, or 16 climes forthe dry test (case 1). The heat flux decreases linearly from the entrance (last clime) to the exit (first clime)of the channel while the surface temperature remains essentially constant.

Figures 12-20 through 12-23 compare the calculated axial air temperature profiles at time t = 2000 seconds.In all tests, the air temperature increases from the entrance (last clime) to the exit (first clime) of the heatedchannel.

12.3.2 AP600 Containment Iodel

The long-term mass and energy release input and PCS flow rate input were not finalized at the time thesensitivity cases were made, therefore, sensitivity results are presented up to the time the IRWST inventorywas depleted (43500 seconds). This was a sufficient transient duration to determine if the AP600containment model would be sensitive to doubling the number of climes, the number of stacks, or theconductor mesh points.

Clime Sensitivity Results

The number of climes and volumes in the annulus ofthe base AP600 containment model were doubled (whilemaintaining the same total volume and heat transfer area) and the case was run using WGOTHICVersion 4.1. Figure 12-28 shows a comparison of the transient pressure for the double clime case with thebase case. There is essentially no difference between the two cases. Figures 12-29 and 12-30 present theaxialwetheat flux profile forthe( ])'and( ]')' clime cases at transient time t=0, 30, 1500, and43500 seconds. The profiles are similar, but the wet heat flux profile for the [ ]ac clime case has betterresolution of the subcooled region at the top of the dome at 1500 seconds and of the dryout elevation at43500 seconds.

Figures 12-31 and 13-32 present the axial dry heat flux profile for the [ ]"' and [ ]' 'clime cases at transienttime t=0, 30, 1500, and 43500 seconds. The profiles are similar.

Figures 12-33 and 12-34 present the axial external film temperature profile for the [ ]' and [ ]' climecases at transient time t=0,30,1500, and 43500 seconds. The profiles are similar, but the film temperatureprofile for the [ ]"' clime case has better resolution of the subcooled region at the top of the dome at1500 seconds and of the dryout elevation at 43500 seconds.

Figures 12-35 and 12-36 present the axial external dry surface temperature profile for the [ ] ' and [ ]'.'

clime cases at transient time t=0, 30, 1500, and 43500 seconds. The profiles are similar.

12-10 Revision I5956rI -I2.wpd-040204

WCAP-15862APP-SSAR-GSC-588 . API000

Figures 12-37 and 12-38 present the axial annulus air pressureprofile for the [ ]' and [ ]ac clime cases attransient time t=0, 30, 1500, and 43500 seconds. The pressure quickly decreases to the boundary conditionvalue (14.7 psia) in both cases. The pressure at the bottom of the annulus is slightly higher than at the topdue the density head.

Figures 12-39 and 1240 present the axial annulus air temperature profile for the [ ]' and [ ]a, 'clime casesat transient time t=0, 30, 1500, and 43500 seconds. The downcomer and riser air temperatures eventuallyestablish similar profiles. The air temperature increases slightly as it flows downward, then increases rapidly(by about 30'F at 1500 and about 18OF at 43500 seconds) in the riser as it flows upward.

Figures 12-41 and 1242 present the axial annulus air density profile for the [ ]3' and [ ]t' clime cases attransient time t=0, 30, 1500, and 43500 seconds. The profiles are similar.

Figures 12-43 and 12-44 present the axial annulus'air velocity profile for the ]4'c and [ ]3C clime cases attransient time t=0, 30, 1500, and 43500 seconds. The profiles are similar. The air velocity decreases at thetope of the riser as it enters the large open area above the dome.

Comparisons of the transient heat rate and heat releases integrated over height of the model are shown inFigures 1245 and 1246. Again, there is essentially no difference between the two cases.

By this comparison, it is concluded that the transient pressure and shell heat removal rates calculated by theAP600 containment model are not sensitive to the number of climes.

Stack Sensitivity Results

The number of stacks of climes in the base model were doubled (while maintaining the same total heattransfer area) and the case was run using 3YGOTHIC, Versi6n 4.1. Figure 12-47 shows a comparison of thetransient pressure for the double stack case with the base case. There is essentially no difference betweenthe two cases. Comparisons of the transient heat rate and heat releases integrated over the height of themodel are shown in Figures 1248 and 12-49. Again, there is essentially no difference between the twocases.

These comparisons show the transient pressure and shell heat removal rate calculated by the AP600containment model are not sensitive to the number of stacks.

Clime Conductor Numerical Mesh Point Sensitivity Results

The number of numerical solution mesh points through the thickness of each of the three conductors (shell,baffle, and concrete) that make up each clime in the base model was doubled and the case was run usingWGOTHIC Version 4.1. Figure 12-50 shows a comparison ofthe transient pressure for the double mesh casewith the base case. There is essentially no difference between the two cases. Comparisons of the transient

Revision 1 12-115956r0 -12.Apd-040204


heat rate and heat releases integrated over the height of the model is shown in Figures 12-51 and 12-52.Again, there is essentially no difference between the two cases.

From these comparisons, it is concluded that the transient pressure and shell heat removal rates calculatedby the AP600 containment model are not sensitive to the number of numerical mesh points within theconductors that comprise each clime.

12.4 SUMINIARY

12.4.1 Simple Annulus Clime Model Noding Study

A two-channel annulus model was exercised over a range of film flow rates, and film temperatures with thecode calculating air velocities associated with natural draft heating. For the cases considered, increasing thenumber of climes was observed to have no significant effect on the predicted heat removal rate. That is, thepredicted heat removal rate was insensitive to the number of clime nodes used in the model. The predictedheat removal rate converges in all cases considered. Increasing the number of climes resulted in smootheraxial heat flux and film temperature profiles for the wet tests. Adequate axial heat flux and temperatureprofiles were predicted without increasing the number of climes for the dry cases.

12.4.2 AP600 Containment Model Clime Noding Study

Using an AP600 containment model similar to, but not exactly the same as the Evaluation Model, the effectofdoubling the number of clime nodes, doublingthe number ofstacks, and doubling the numberofconductormesh points was studied. In each case, code calculations were found to be unaffected by the variations inthe model features.

These studies demonstrate that results obtained with the AP600 containment model are not sensitive tochanges in the clime and annulus noding. Therefore, it is concluded that [ I' climes are adequate to predictthe PCS performance for the Evaluation Model.

12.5 REFERENCES

12.1 Westinghouse Letter NTD-NRC-954563, B. A. McIntyre to T. R. Quay (NRC), "GOTHICVersion 4.0 Documentation," September 21, 1995.

12.2 Westinghouse Letter NTD-NRC-95-4462, N. J. Liparulo to T. R. Quay (NRC), EPRI ReportRA-93-10, "GOTHIC Design Review, Final Report," May 15, 1995.

12.3 WCAP-14812, Revision 2, "Accident Specification and Phenomena Evaluation for AP600 PassiveContainment Cooling System," April 1998.

12-12 Revision I5956r-1 2.vwpd-040204

- --

WVCAP-1 5862APP-SSAR-GSC-588 APinnnAPP-SSAR-GSC-588 APlAAA

InsideContailnent

IIIIIIIIIIIIIIIIIII

AirIU fIlow

C-hannel

IIIIIIIIIIIIIIIIIt17

AirDownflowrChannel

Environment

ShieldBuilding

,-- Wall

qeonvect ~convect Qcon%,eclli ~

______

Qconvee Q9convect

Q9oden1se Q Q

Qradfnet)

* Clime- Boundarv

i III I I !

GOTHIC Climes

CUME Heat and Mass Transfer Routines Interface I G COTHIC Volume Properties are Used In

with Westinghouse-GOTHIC Volume Source Trrm Ithe CUME Heat and Mass Transfer Correlations. I-

K> Figure 12-1 Westinghouse-GOTIIIC Climie Wall Source Tcrm Models

Revision I5956rl -1 2.vpd-040204

12-13

I


I

OUT 1. 0 -NKr%

KINm= 0.5

.-

i

TW=

200aF

RIS'

*

ER * I'l Q

I`-DOWNCOMER

I $

Turning Location

Figure 12-2 Simplified Line Diagram of Annulus Clime Model

12-14 Revision II 5956r1-12.wpd-040204

K>


- ---

Riser Downcomer

Clime ClimeConductors Conductors

* t 1 I' Ii- ' 7I3 -u==~--J ~Jruu-

. | ________ J -- 1 -

.1 1uu---T--- - rpI. I _' I_

Heat

Source

,1 4-:L…nA.1 1 LfIJJ----UH_

- - - - - - - - - - - I 16

'L- - - -'.___…---ai-184n

_ r~ _ _ ~Jiw . .

3F. Drain

Figure 12-3 Noding Diagram, 8 Clime Node Model

Revision I5956rl -1 2.vpd-040204

12-15


5.25

5

rA

-

E- 4.75

m

C)

C:

12 4.5

4.25

40 500 1000 1500 2000

Time (sec)16 Climes 8 Climes 4 Climes

O3 -... a..-- 3

2500

Figure 12-4 Comparison of Transient Heat Transfer Rates; Case 1

12-16 Revision I5956rl - 2.vpd-040204


117

116.51-S1

u 116

._

M 115.5

: .

A.,.'----A

.......... 9......... 9.........0

1150 500 1000 1500 2000 2500


-B -_.- .....

K> Figure 12-5 Comparison of Transient Heat Transfer Rates; Case 2

Revision I5956r -12.vpd.040204

12-17


84

83.75

-

c 83.5

U

z 83.25

:4

'.E ..... ..... .... .. ..... --- -'0 .. .

I I II

.I

830 500 1000 1500


B ._ _- ..e....

2000 2500

Figure 12-6 Comparison of Transient Heat Transfer Rates; Case 3

12-18 Revision I5956r-1I 2.%Npd-040204

WCAP-15862APP-SSAR-GSC-588 API. 00

-- --

57

56.75

i; 56.5

a) 56.25

'a 56

55.75

55.5 Lo 500 - 1000 1500 2000


B -- --- -- e.

2500

K> Figure 12-7 Comparison of Transient Heat Transfcr Rates; Case 4

Revision I5956rl -12.wpd-040204

12-19


0.9

0.8

,_ 0.7

- 0.6U9 0.5

0 0.4

C 0.3

pa 0.20MX 0.21

-0.1

-0.20 500 1000 1500 2000

Time (sec)16 vs 8 Climes 8 vs. 4 Climes

G---A--.

2500

Figure 12-8 Comparison of Transient Heat Transfer Rate Differences; Case I V>12-20 Revision I

5956rl1-2.wpd-040204


0.6

% 0.5

0U)r.)

. 0.4

.)

04D

U' 0.3

0.2 1 l

0 500 1000 . 1500 2000


-2-

2500

K> Figure 12-9 Comparison of Transient Heat Transfer Rate Differences; Case 2

Revision I5956r1-12.%pd-040204

12-21

VCAP-15862A DDCQ A D-r.0_RRI2 API1000IAr-or- - -IL-

0.3

Z 0.2U

I-

~U- 0.1

C:

M O

-0.10 500 1000 1500 2000 2500


CB---

Figure 12-10 Comparison of Transient Heat Transfer Rate Differences; Case 3

12-22 Revision I5956ri-12.wpd-040204


0

- -0.1

C.)c)

E) -0.2

2) -0.3C:

Z -0.4

I,

,''

0 ,

-0.50 500 1000 1500


- -- A---

2000 2500

K> Figure 12-11 Comparison of Transient Heat Transfer Rate Differences; Case 4

Revision I5956r0-12.vwpd-04020W

12-23

WCAP-15862APP-qRAR-GSC-5R API000

170

160

II 150

-aE- 140

o 130x

[I 120

110

100

0...- .-- o.

1 , 1 -ILL I I I I I I I I I I

'\\y

I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Clime Number16 Climes 8 Climes 4 Climes

C] --- A .-- . ..... 0.....

Figure 12-12 Comparison of Heat Flux Profiles; Time t = 2000 Seconds; Case I 'K)y

12-24 Revision I5956r1-12.vwpd-040204


-- --

12000

11000

-IsAt l0000

, 9000

E 8000m

*- 7000

0° 6000

: 5000

G 4000

3000

2000I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16


-B --- A . -e.....

K> Figure 12-13 Comparison of heat Flux Profiles; Time t - 2000 Seconds; Case 2

Revision I5956r& -12.wpd-040204

12-25


--- ----

K2

3500

-c 3000Cq

1.I

- 2500

E-M- 2000:3

0

-1 1500

Z 1000

500

I I

* a0_

.1 I I I (. - -'K)

I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16Clime Number

16 Climes 8 Climes 4 Climes- .- A . .. ..........

Figure 12-14 Comparison of Heat Flux Profiles; Time t = 2000 Seconds; Case 3

12-26 Revision I5956r] -I 2.wpd-040204


4000

3500

r, 3000

2500

0, 2000

X 1500

,: 1000

500

JO ' I I I -I II r r T T

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16


- -B --- Ae

K> Figure 12-15 Comparison of Heat Flux Profiles; Time t = 2000 Seconds; Case 4

Revision I5956r] -1 2.wpd-040204

12-27


K-,-

198.9

198.8

_Iz

B 198.7

0

E 198.6

198.5

198.41 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Clime Number16 Climes 8 Climes 4 Climes-a --- .- . . ...... .....

Figure 12-16 Comparison of Film Temperature Profiles; Time t = 2000 Seconds; Case I

12-28 Revision I5956r1-12.%vpd-040204


180

a 170

Co 160

tZ 150

140 Ll I I I I I t I tl I I I I

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16


- . - ----.. ....

Figure 12-17 Comparison of Film Temperature Profiles; Time t = 2000 seconds; Case 2

Revision I5956r1-12.wpd-040204

12-29


188

186

184

as 182

a.E 180

Ea

3 178

176

174

K>

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Clime Number16 Climes 8 CHies 4 Climes

B -......-....

Figure 12-18 Comparison of Film Temperature Profiles; Time t = 2000 seconds; Case 3

12-30 Revision I5956rl-12.wpd-040204

--------

WCAP-15862APP-SSAR-GSC-588 APIMON

-- --

195

C,190

p-

g185

V4 180

1751 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Clime Number16 Climes '8 Climes 4 Climes

B .-.-..-- -

K> Figure 12-19 Comparison of Film Teniperature Profiles; Time t = 2000 seconds; Case 4

Revision i5956r1-12.vKpd-040204

12-31


. IN

165

160

_ 155LT.,1-)1- 150

c) 145

EO 1401-4

¢ 135

130

125

K)o

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Clime Number16 Climes 8 Clmes 4 Climes

Figure 12-20 Comparison of Air Temperature Profiles; Time t = 2000 seconds; Case I

12-32 Revision I5956r 1I2.%%pd.040204

WCAP-15862APP-SSAR-GSC-588. APIOOO-- - -

170

165

160

C.) 155

as 150a-

' 145c)

Hi 140

135

130

125 I I I . I . I l I I I I I I I I I I

1 2 3 4 5 6 7. 8 9 10 11 12 13 14 15 16


-B --- A---- -. ..-

\> Figure 12-21 Comparison of Air Temperature Profiles; Time t = 2000 seconds; Case 2


12-33


K>

155

150

e 145

2P.140

9 135

130

125

\-i,

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Clime Number16 Climes 8 Cries 4 Climes

B -- -

Figure 12-22 Comparison of Air Temperature Profiles; Time t = 2000 seconds; Case 3J

12-34 Revision I5956rl -I 2.%%pd-040204

C(7 (C

-o ClmeNube

CD~

160 C 8 e l

aA0

155-

05

*0 ~L145

CD

14

~13

13

<135

130 9 10 11 12 1

1 2 3 4 56 7lme 8 Clme 10 C2l3i1m1e1ClimeNumbe


(a,c)

I>

K-)~

12-36 Revision I5956r]-12.wpd-040204


U) (ac)

Revision 1 12-375956r1-12.wpd-040204

WCAP-1 5862APP-SSAR-GSC-588 API OOO

(ac)

12-38 Revision 15956r1-I 2.wpd-040204


;

Inside t I Air A i Air t I!niromentContainment upnor Dawnoaw t

I t Chzannl t t Channel tI t t I t I

Vessel II Shield

I, t !--- tII

ClimeI L I__ I I I Boundary

meoundary

iI i" t Ise Yol{18 Meshes I I meshes 16 Meshes

AP600 Containment Model

N

I I I I II

Inaide I A Jdr A I Air I I EnironmentContainment upno t Do-nflo t

I I C a el I Channel I I

Vessel t B e Ptdingwall wan

ClimeI - - _ I _ Boundary

Clime.

---- - - - -Bondary

32 Mese22 MZeshes 3o Yeher

AP600 Mesh Sensitivity Model

Figure 12-27 AP600 Containment Model Double Mesh Point Clime Noding Pattern

Revision 1 12-395956rl-12.%pd-040204


, N

IK "(a,c)

K�)1240 Revision I

5956rl -1 2.wpd-040204


Rvso I(a,c)

Revision I 1 2415956rl -12.vwpd-040204

WCAP-15862APP-SSAR-GSC-588 APIO00

(a,c) &

1.,,

12-42 Revision I5956rl -1 2.wpd-040204


(ac)

K>Revision I5956rl -1 2.pd-040204

1243


(a,c)

K�)12-44 Revision I

5956r -12.% pd-040204

WCAP-I 5862APP-SSAR-GSC-588 APInnn-- ----vv

(ac)

K>Revision I5956r] -1 2.wpd-040204

1245


(a,c)

K212-46 Revision I

5956r] -12.wpd-040204


(a,c)

I


1247


(ac)

V>1248 Revision I

5956r] -I 2.wpd-040204

WCAP-I 5862APP-SSAR-GSC-588 APInnn

(a,c)

Revision I5956rl -12.upd.040204

1249


(a,c) `_J

12-50 Revision 15956r I-12.wpd-040204


(a,c)

K)Revision I5956rl -12.wpd-040204

12-51

WCAP-15862APP-S5AR-SCr-5RR AP 1000

(a,c)

K>12-52 Revision I

5956rl-I2.wpd-040204

WCAP-15862APP-SSAR-GSC-588 - APIOOO

(a,c)

KU

Revision I5956rl -1 2.Nwpd-040204

12-53


(a,c)K)~

k~lj

12-54 Revision I5956r -1 2.wpd-040204


(a,c)

Revision 1 12-555956ri-12.wpd.040204


(a,c)

K>

12-56 Revision I5956rl-12.wpd-040204

WVCAP-I15862APP-SSAR-GSC-588 APIOQO

(a,c)

Revision 1 12-575956rl-12.wpd-040204


(a,c)

12-58 Revision I5956r1-12.%wpd-040204


(ac)

Revision I5956r] -1 2.vwpd-040204

12-59


(ac)

I

12-60 Revision I5956r] -12.wpd-040204

WCAP-1 5862APP-SSAR-GSC-588 APInAAn

(a,c)

K>Revision I5956rl-12.wpd-040204

12-61

Il

WCAP-15862APP-SSAR-GSC-588 AP1 nonAPP-SSAR-.GSC-588 AP1AAfl

45

40

35

Q 30

C5 25

3 20CT,

t 1s

10

5

0

K>

I I ... . .1 . . I11-1 . . I. ... .1 . .. ..... I. 1 . . . I I3.....

I 10 100 1000 10000 100000

Time (sec)Normal Mesh DoublMesh

E3 -_ _

Figure 12-50 Pressure History Comparison, AP600 Containment Model; Double MeshSensitivity Case

12-62 Revision I5956rl-1 2.upd-040204

(i C Ctot

o110 10c00o00 100

75

o ~00

- ~60 0

300

~1 o

a Time (sec)

~~| ] Normal Mesh Doublq ,Mesh |

~~30 ~ 4..J>

I>- (iQ u

(2 O 02 3 05

X00

00U 0.6

o - oilMshDul eh|

(. 0.(

0.

0.4I

- 0.2

00

0 10 20 30 40 50

0 Time (see)Thousands

Normal Mesh Double Mesh

C;CC

Section 13

Description of AP1000 Plant Geometry inWGOTHIC Evaluation Model

5956rl -1 3.wpd-041404


--- ----

This entire section (13) is proprietary to Westinghouse Electric Company.(a,c)

K>Revision I5956r1-13.wpd.041404

Section 14

LOCA Mass and Energy ReleaseCalculation Methodology

5956r] -14.v6,pd-041404


TABLE OF CONTENTS

LIST OF TABLES ................... iii

LIST OF FIGURES ................... iii

14.1 INTRODUCTION .14-1

14.2 BLOWDONVN MASS AND ENERGY RELEASE CALCULATION .14-2

14.3 POST-BLOWDOWN MASS AND ENERGY RELEASE CALCULATION . 14-2

14.4 COMPARISON OF SATAN/SPREADSHEET RELEASES WITHWCOBRA-TRAC CALCULATED RELEASES .14-9

14.5 SUMMARY AND CONCLUSIONS .14-14

14.6 REFERENCES .14-14

, .

LIST OF TABLES

Table 14-1 DECLG LOCA Energy Release Modeling .14-4

Figure 14.3-1Figure 14.4-1Figure 14.4-2Figure 14.4-3Figure 14.4-4Figure 14.4-5Figure 14.4-6Figure 14.4-7

LIST OF FIGURES

Spreadsheet Mass and Energy Release Schematic ........................... 14-8Mass Release Comparison .................................. 14-10Energy Release Comparison .................................. 14-10Mass Release Comparison .................................. 14-11Energy Release Comparison .................................. 14-11Integrated Mass Comparison ............... ................... 14-12Integrated Energy Comparison .................... .............. 14-12LOCA Transient Comparison ................. ................. 14-13

Revision I5956rl-1 4.wpd- 1404

iii

WCAP-1 5862APP-SSAR-GSC-588 . AP I000

14.1 INTRODUCTION

This section describes the method used to calculate the loss-of-coolant accident (LOCA) mass and energyreleases that are used to calculate the containment pressure response for the Design Basis Analysis (DBA).The method maximizes both the magnitude and the rate of the mass and energy release to containment. Thisintroduces substantial conservatism in the prediction of the peak containment pressure and temperatureresponse following a LOCA event.

Section 10, Figure 10-1 presents a comparison of the WGOTHIC AP600 Containment Evaluation Modelsensitivity to various input values and correlations. These are not single-effect sensitivities; the results arecumulative' Each additional sensitivity case modification is stacked upon those that immediately precededit. These sensitivities provide insight into the effect of each parameter, individually, by comparison to thepreceding case, as well as the total conservatism represented by these parameters. 'Case 7 (the heavy dashedline in Figure 10-1) represents the effect of removing some of the conservatism in the input mass and energyreleases.

For Case 7, the mass and energy release calculation input was modified as follows:

The nominal (without adding 5 F uncertainty) full power RCS temperatures were usedThe nominal (without adding 30 psi) full power RCS pressure was usedThe nominal (without adding uncertainty) RCS volume was usedThe nominal (without adding 2% uncertainty) full core power was usedThe nominal (without adding 15%) full power core stored energy was usedThe nominal CMT, accumulator, and IRWST fluid temperatures were used.

In addition to the input changes described above, the following calculational changes were made during thepost-blowdown phase:

The 1979 ANS decay heat standard for an 800-day average bumup (without adding the 2 sigmauncertainty) was used

The post-blowdown refill period was modeled

* The broken loop steam generator energy was assumed to be released over a 1-hour period

* The intact loop steam generator energy was assumed to be released over a 2-hour period

* No zirc-water reaction energy was added.

A comparison of the mass and energy releases for Case 7 and the AP600 Containment Evaluation Model isshown in Figures 10-3 and 104.

Revision I 14-15956rl-14.wpd-041404


As shown in Figure 10-1, by eliminating only a portion of the conservatism in the calculation of the massand energy releases, there is a substantial decrease in the post-blowdown peak pressure (about 6 psi). Thecalculated peak containment pressure is reduced from the previous case by about 4.3 psi and is shifted to theend ofblowdown. The largest effect of any other single change to one of the containment model input valuesis a reduction of about 3 psi. Therefore, the conservatism in the mass and energy calculation outweighs anyother single modeling parameter.

14.2 BLOWDOWN MASS AND ENERGY RELEASE CALCULATION

The primary differences between the conventional plant design and the passive plant design are theengineered safety features. The safety features of conventional operating plants include both passive andactive systems; the AP600 and API 000 safety features are all passive. This difference only affects the long-term inventory makeup systems, not the system behavior during the blowdown phase.

The methodology for calculating the mass and energy release to containment during the blowdown phase isnot affected by the passive safety systems. The only safety system that injects water during the blowdownphase is the accumulator system; the gravity-driven core makeup tanks (CMTs) cannot inject into thecommon direct vessel injection line against the pressure of the gas-charged accumulators. Accumulators areincluded in both the passive and active plant designs and are modeled with the NRC-approved DBA LOCAmass and energy release methodology. The AP600 uses spherical accumulators, whereas currently operatingWestinghouse designed plants use cylindrical accumulators. The accumulator inventory is depleted wellbefore the time of peak pressure so any difference in discharge rate associated with the different accumulatorgeometry would have an insignificant effect on the calculation for peak containment pressure.

The blowdown mass and energy releases are calculated using the SATAN-VI computer code. TheSATAN-VI code, associated modeling assumptions, and noding structure for the mass and energy releasecalculations are documented in WCAP- 10325-P-A (Reference 2). It has been reviewed and approved by theNRC for use in calculating the blowdown mass and energy releases for Westinghouse PWR containmentintegrity DBA. The blowdown phenomena modeled by the SATAN VI code are the same for allWestinghouse PWRs, including the AP600 design.

The variable noding structure of the SATAN model allows the user to simulate current and advanced reactorcoolant system geometry with generalized control volumes. The standard Westinghouse PWR reactorcoolant system noding structure is modified to specifically model the passive plant reactor coolant systemgeometry, which includes two cold legs in the coolant loops and a direct vessel injection (DVI) line to thedowncomer.

14.3 POST-BLOWDOWVN MASS AND ENERGY RELEASE CALCULATION

The post-blowdown phase consists of three periods: refill, reflood, and long-term cooling. The refill periodoccurs just after blowdown. During the refill period, the accumulators refill the downcomer and lower

14-2 Revision 15956ri-14.upd-041404


plenum of the vessel with water. There are little or no releases during the refill period. The reflood periodoccurs just after refill. During the reflood period, the core is refilled with a 2-phase mixture of steam andwater. Following reflood, the RCS fluid inventory and long-term cooling are maintained by the pumpedsafety injection system and recirculation system (conventional plants) or the core makeup tanks and gravity,draining of the IRWST (passive plants).

The approved codes for conventional plant post-blowdown mass and energy release analyses would have hadto have been significantly modified for them to have been used for passive plant analyses. Instead, acalculational method was developed to conservatively predict the AP600 and AP I 000 post-blowdown massand energy releases in a manner that is consistent with the current approved methods for DBA. A spreadsheetwas used for the calculations to simplify accounting of the mass and energy sources and releases.

Mass Sources

Figure 14.3-1 illustrates the basic model for the AP600 and API 000 post-blowdown mass and energy releasecalculation. All of the injection sources (accumulators, CMTs, and IRWST) are assumed to be delivered tothe DVI line at a mass flow rate, mij,,, Flow from both accumulators refills the vessel following blowdown.The CMTs begin to inject just before the accumulators empty. The IRWST begins to gravity drain to the vessel(and spill from break) just before the CMTs empty.

The injected liquid mass flow rate delivered to the DVI line is calculated with a simplified resistancenetwork. Two cases are considered: one with the line resistances maximized (to produce the minimumaccumulator, CMT, and IRWST flow rates) and one with the line resistances minimized (to produce themaximum accumulator, CMT, and IRWST flow rates). The resistance network also calculates a flow ratearound the loops and a flow rate out the broken cold leg stub, however, these are not used in the mass andenergy release spreadsheet calculation. Instead, as shown in Figure 14.3-1, all of the injection flow that isdelivered to the DVI line is assumed to pass through the core, hot legs, steam generators and out of the break.This maximizes the SG heat release rate, and is a source of conservatism in the analysis.

Energy Sources

The various sources of energy that are considered in the calculation, and their assumed release rates aresummarized in Table 14-1. The release rates are biased to release energy (in the form of steam) to thecontainment early in the transient. The containment pressure response is dependent on the steam mass flowrate, so maximizing the steaming rate is conservative. The RCS metal energy release provides sensibleheating of the injected liquid mass flow. The core decay heat and SG energy releases provide both sensibleand latent heat to the injected liquid mass flow rate.

The SG equilibration time input value in the spreadsheet calculation is adjusted to maximize the steamingrate during the CMT injection phase. The SG equilibration times that are used for API000 are longer thanAP600 because both the core decay heat and SG stored energy are higher in the API 000..

Revision 1 14-35956rH-14.vpd-041404


Table 14-1 DECLG LOCA Energy Release Modeling _

Time Frame Source of Energy Energy Release Rate

Accident initiation to end of blowdown (30 Core, RCS and SGs SATAN VI Calculationseconds)

End of blowdown + 180 seconds The Zr-112O Reaction Energy for Constant1% of the Fuel Cladding

End of blowdown to end of reflood Core stored energy Constant(400 seconds)

End of blowdown + SGB equilibration SG energy - broken loop Exponential

End of blowdown + SGI equilibration SG energy - intact loop Exponential

End of blowdown + 60 minutes RCS metal energy Exponential

End of blowdown to 72 hours Fission Product Decay ANS 1979 + 2 sigma

Description of Calculational Method

Three basic steps in the calculational method determine the DBA mass and energy releases as a function oftime after the blowdown phase. First, the individual contributions for each of the various sources of energyare calculated. Second, the vessel internal mass and break mass flow rate are calculated. Third, the breakmass flow rate is partitioned into a steam and a 2-phase component. A check for conservation of mass andenergy is performed at the end of each time step. This method for calculating the mass and energy releasesafter the blowdown phase is described in more detail below.

Calculation of Sources of Energy

1. The 1979 decay heat standard with 2 sigma uncertainty, all U-235 fission, and 3 years full poweroperation is used to calculate the core decay heat energy release rate at each time step.

2. [

K-i

]6 The core stored energy is released at aconstant rate over the reflood period.

3. The Zr-H20 reaction energy, corresponding to 1% clad oxidation, is released at a constant rate overthe first 180 seconds of the reflood period.

4. The RCS metal energy is released exponentially. Using an exponential rate is conservative since thisassumes the heat transfer rate is much larger than the rate of conduction through the metal. Theequation for determining the stored metal energy as a function of time is:

14-4 Revision I

14-4 Revision I5956rl -I4.NTAd041404

WCAP-15862APP-SSAR-GSC-588 AP IOOO

Q(t) = Qri, + (QmiiDi - Qfl)*e e

where: Qfinal = stored metal heat at equilibration (BTU)

Qnitial = stored metal heat at the end of blowdown (BTU)a = inverse of the metal time constant (sec-')t = time (sec)

b = end of blowdown time (sec)

therefore, the release rate is:

dQ a * (Qi1i, - Qfina) * e -a~t-b)dt

The stored metal heat and time at the end of blowdown are obtained from the SATAN output tables.The stored metal heat at equilibration, Qain,, is'determined by the metal heat capacity and theequilibration temperature, M*c *Tq. The metal heat capacity, M*c,, is obtained from the SATANinput tables. The RCS temperature is assumed to reach an equilibrium value at the end of the refloodperiod. Using an equilibration temperature, Teq, of [ ]'c was found to produce good agreementbetween the spreadsheet calculational method and the code calculated mass and energy releases fora conventional plant. This equilibration temperature value corresponds to [

']a. the AP600 containment design pressure. A higher equilibrationtemperature value of[ ],C is used for API 000 since the containment design pressure is higherthan AP600. The equilibration temperature is held constant for the post-blowdown calculation.

The metal time constant is the metal heat capacity divided by the heat transfer rate. Because the heattransfer rate is variable, the time constant is arbitrarily increased to release 99% of the RCS metalenergy (in excess of the equilibration temperature) over 60 minutes.

5. The steam generator stored energy at the end of blowdown is released using the same equationshown above for the RCS metal energy release rate. The stored energy and time at the end ofblowdown are obtained from the SATAN output tables. The stored energy at equilibration isdetermined by the metal and fluid heat capacity and the equilibration temperature, {(M*cp)nuid +

(M*cp)m.,. ) * T.q. The steam generator fluid and metal heat capacity input values are obtained fromthe SATAN input tables.

The stored energy in the broken loop and intact loop steam generators is released at different rates.The time constants for AP600 are set to values similar to standard plant analyses. Because theAP1OOO CMT injection flow rate limits the steaming rate, the APIOOO time constants are adjustedto provide a more mechanistic reprsentation of the steam generator energy release relative to theavailable core makeup flow, and to maximize the steaming rate during the CMT injection period.

-'Revision 1 14-55956r] -14..pd-041404


Calculation of the Vessel Liquid Mass

6. During the reflood period, the mass of liquid in the vessel is assumed to increase exponentiallyaccording to the following equation:

M(t) = Mrn,, + (M~idai - Mrmai) * e c(t-b)

where: Mr.. = RCS mass at the end of reflood (Ibm)Minitial = RCS mass at end of blowdown (Ibm)c = inverse reflood time constant (sec-')t = time (sec)b = end of blowdown time (sec)

At the end of the reflood period, the RCS is assumed to be at a saturation temperature of[ Y^with a vessel void fraction of[ ]' I. These assumptions were found to produce good agreementbetween the spreadsheet calculational method and the code calculated mass and energy releases fora conventional plant.

The RCS liquid mass at the end of the reflood period, Mfina,, is calculated using the followingequation:

Mrin.1 = V*(l-alpha)*rho

where: V = Vessel volume (ft3)

alpha = Vessel void fraction at end of refloodrho = Saturated liquid density at [ adC (Ibm/ft3 )

The temperature and amount of mass in the vessel at the end of reflood are held constant throughoutthe long-term cooling phase. These assumptions maximize the steam release rate because there isno additional filling of the vessel and, because the temperature does not decrease, the latent heat ofvaporization does not increase.

An initial value for the RCS fluid mass is given in the SATAN summary table. The RCS initial massinput for the spreadsheet calculation, Minid.1, is artificially instantaneously increased to eliminate thetime delay associated with refilling the downcomer and lower plenum of the vessel. Eliminating therefill period shifts the mass and energy releases earlier in time and results in an increase in thepredicted peak containment pressure.

The reflood time constant is related to the vessel volume, the injected mass flow rate and thesteaming rate. The time constant is set to allow the vessel mass to reach 99% of the Min,1 value in

14-6 Revision II 5956rl-14.wpd-041404

WCAP- 15862APP-SSAR-GSC-588 APIOQO

approximately 400 seconds. This time frame is consistent with reflooding times for conventionalWestinghouse PWR designs with similar vessel volume and injection rates.

Calculation of the Break Vapor and 2-Phase Mass Flow Rates

Steps 7 through 10 describe the method that is used to calculate the break mass flow rate and to partition itinto the vapor and 2-phase flow components for input to the WGOTHIC Containment Evaluation Model.

7. The break mass flow rate is defined as the difference between the injected mass flow rate and thetime derivative of the vessel liquid mass, as shown in the equation below:

mbrcak = Inj.., - C*(Mfinal -Minifia-) * e -c(-b)

Following the end ofthe reflood period (400 seconds), the total flow released from the break is equalto the injected flow; no additional inventory is stored in the RCS.

8. The core inlet enthalpy is calculated by dividing the RCS metal energy release rate by the injectedliquid mass flow rate, and summing this with the enthalpy of the injected flow, lc.,¢ = hbnij< +Qmc./minjct. After injection stops, the core inlet enthalpy is set to a constant value. This value isbased on the upper bound fluid temperature in the containment sump.

9. The total break flow rate is partitioned into a steam and a 2-phase component. Because thecontainment pressure is dependent on the steam release, the steam mass flow rate is maximized byassuming 95% of the core energy and all of the steam generator energy is used to produce steam.The steam and 2-phase fluid break mass flow rates are calculated using the following equations:

. rri,.a = min { mbk, (0.95 * Qco. + QsG )( hg - h.) }m2-phaw mbcAk - Mstmcam

where: Qcorc = Sum of the core decay heat, stored energy and Zr-H20 reaction release rates(BTU/s)

QSG = Sum of the broken and intact SG energy release rates (BTU/s)hg = Steam enthalpy at the equilibration temperature (BTU/Ibm)hcore = Core inlet enthalpy (BTU/s)

mbncak = Total break mass flow rate (Ibm/s)

Note, because the steam enthalpy is assumed to remain constant at the equilibration temperature, thesteam generator energy is used to produce additional steam release, rather than superheat the steam.This maximizes the steam release rate.

Revision 1 14-75956rl -14.wpd-041404


7-

10. The enthalpy of the 2-phase break fluid is calculated using the following equation:

h2 .phas = min { hg, h.. + 0.05 * Qr, / M2-phae }

where all terms have been defined previously. The limit is applied to prevent the 2-phase enthalpyfrom exceeding hg when the 2-phase mass flow rate is very small.

Check for Energy Conservation

11. Conservation of energy is checked at the end of each time step. [

I2C

VESSEL

idol

Figure 14.3-1 Spreadsheet Mass and Energy Release Schematic

14-8 Revision I5956rl-14.wpd-041404


14.4 COMPARISON OF SATAN/SPREADSHEET RELEASES W'ITH WCOBRA-TRACCALCULATED RELEASES

The API 000 DECL LOCA mass and energy releases calculated using the SATAN/spreadsheet method werecompared to the WCOBRA-TRAC (WCT) code calculated results to demonstrate the amount ofconservatismin the calculation. The comparison was run out 2000 seconds; this is beyond the calculated time of peakcontainment pressure. Although this version ofthe WCT code was not designed to produce LOCA mass andenergy releases that would be conservative for use in calculating containment peak pressure, it shouldnevertheless mechanistically calculate a reasonable estimate of the releases.

The blowdown mass and energy comparison is shown in Figures 14.4-1 and 14.4-2. SATAN predicts a higherblowdown mass and energy release rate than WCT. The post-blowdown mass and energy comparison isshown in Figures 14.4-3 and 14.4-4. The spreadsheet method calculates a higher release rate than WCTbecause the spreadsheet releases the SG energy at a much higher rate than WCT.

The integrated mass and energy releases are compared in Figures 14.4-5 and 14.4-6. The SATAN/spreadsheetmethod predicts approximately a 50-percent higher integrated mass and energy release than WCT, by thetime peak pressure occurs. This difference results in a 20 psi higher calculated peak pressure as shown inFigure 14.4-7.

Revision I5956rl-14.wpd 041404

14-9


\-Z--/

Mass Release Comparison

80000-

7 70000

60000-

E 50000 -

coWC

40000 -W T.... SATAN

MO 30(000 - -_ _ _ _ _ _ _ _ _ _ _ _

u~20000

100-

0 10 20 30 40 50Time (sec)

Figure 14.4-1

Energy Release Comparison

4500000040000000

: 35000000-

m 30000000-

W 25000000 - - WcT

= 20000000 - ....-- SATAN

> 15000000

ID 10000000-

5000000-

0 10 20 30 40 50

Time (sec)

Figure 14.4-2

14-10 Revision 15956rl1-4.wpd-041404


Figure 14.4-3

-Energy Release Comparison

500000

450000 =

- 400000 - lm 350000 - _ _ _ _

, 300000 -_______ I W CT I250000 - I

_ __ ....... SATAN200000 - _ _ _ _ _ _

=> 12°50°0°O s . ............... ...........150000w 100000 -

50000 -

0 -

0 500 1000 1500 2000

Time (sec)

Figure 14.44

Revision I5956rl -14.%Npd-041404

14-11

WCAP-15862APP-SSAR-GSC-588 API000 __

Figure 14.4-5

Integrated Energy Comparison

800000000

_700000000

m 600000000

6 500000000co

aO 400000000

,= 300000000c-

200000000wU:

00100000000

0

WCT..SATAN |

0 500 1000 1500

Time (sec)

2000

Figure 14.4-6

14-12 Revision I5956rl -14.wpd-0414G4


LOCA Transient ComparisonWCAP-15612 Scoping M&E Release

- - -- SATAN + Spreadsheet M&E Release------- WCT M&E Release

60 -

50-

F540

_ 30 -----------------

as20-

10-

0~ _ …

D3o-

Figure 14.4-7

Revision I5956r1-14.%pd-041404

14-13

I


14.5 SUMMNIARY .AND CONCLUSIONS

The methodology, used to predict mass and energy releases to the containment during the blowdown phase

of the Large LOCA accident for the AP600 and AP1000 containment DBA, is consistent with the

methodology approved by the NRC for current operating plants. The approved methodology yields

conservatively high total energy and energy release rates to the containment atmosphere, consistent with

regulatory guidance for Design Basis Analyses.

A spreadsheet calculation method is applied to calculate the post-blowdown mass and energy releases for

the AP600 and AP1000. Because the containment pressure is dependent on the steam release, the break

steam mass flow rate is maximized in the calculation. The transient flow rates for the accumulators, CMTs

and IRWST, are calculated using a resistance network. These flow rates, along with the conditions at the end

of the SATAN blowdown mass and energy release analysis, are input to the spreadsheet. The spreadsheet

accounts for the same sources of energy and uses similar energy release rates as the conventional plant

analysis. A comparison of the releases predicted by the spreadsheet and the approved mass and energyrelease analysis codes/models show that the spreadsheet provides a conservatively high prediction of the

release to containment. Therefore, the spreadsheet method for calculating the post-blowdown mass and

energy releases used in the AP600 and AP1000 containment analysis is consistent with the methodologyapproved by the NRC for current operating plants.

14.6 REFERENCES

1. WCAP-10325-PA, "Westinghouse LOCA Mass and Energy Release Model for Containment Design

March 1979 Version," May 1983

14-14 Revision 15956rI-14.wpd-041404

WCAP-15862 APP-SSAR-GSC-588 AP 1000 · 2012. 11. 19. · WCAP-15862 APP-SSAR-GSC-588 API000 l0...

Documents

Transcript of WCAP-15862 APP-SSAR-GSC-588 AP 1000 · 2012. 11. 19. · WCAP-15862 APP-SSAR-GSC-588 API000 l0...