ML19339K496

Final Rept, Assessment of Ex-Vessel Fuel-Coolant- Interaction Energetics for GE Abwr.

ML19339K496
Person / Time
Site:	05200001
Issue date:	02/28/1993
From:	Michael Corradini, Hossein Esmaili, Khatibrahbar AFFILIATION NOT ASSIGNED, WISCONSIN, UNIV. OF, MADISON, WI
To:	NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
Shared Package
ML19339K495	List: ML19339K496
References
CON-FIN-L-25422, CON-NRC-92-04-045, CON-NRC-92-4-45 ERI-NRC-93-203, NUDOCS 9303300323
Download: ML19339K496 (41)

Text

. ..

I

~ ,

. )

i e  :

i l l

i l

RN R55ESSMENT OF EX-VESSEL FUEL-COOlRNT-INTERRC ENERGETICS FOR THE GENEARL ELECTRIC RDVANCED BOluNG LURTER REACTOR 5

Final Report 4

february 1993 Energy Research, Inc.

6290 Montrose Road

• Rockville, Maryland 20852 r

.< T Work Performed Under the Auspices of the United States Nuclear Regulatory Commission Office of Nuclear Regulatory Research Washington, D.C. 20555 Contract No. 92-04-045 FIN L-25422

. . umu e 3

9303%C323 930308 P  ;

CF 4iDOCK 05200001 Ee CF Q . - - . . - __. . . - - . . . - - .

F i

i t

AN ASSESSMENT OF EX-VESSEL FUEL COOLANT INTERAC ENERGETICS FOR THE GENERAL ELECTRIC ADVANCED l BOILING WATER REACTOR .

i Final Report  !

i i

February 1993 ,

t

' M. Khatib-Rahbar -

Pnncipal Investigator H. Esmaili. M. Corradini'. J. Ptacek, R. Vijaykumar Energy Research. Inc.

i 6290 Montrose Road Rockville Maryland 20S52 l

Work Performed Under :ne Auspices of the l

l United 5stes Nuclear Regulatory Commissuin - '

t 0:6cc of Nuclear Repuietory Researen -

Washington. D. C. 20555 Contract No. 92-04-645. FIN L-25422 -

l l

Department of Nuciear En tneering and Engineering Physics. Univeru:y r r Wisconsi a w

'ir s

TABLE OF CONTENTS I

1. INTRODUCTION .

. 2 1.1 Objectives .

, 2 1.2 Description of the TEXAS.II Fuel. Coolant-Interaction Model

2. ABWR DESIGN FEATURES .

. . ..... 5

3. INITIAL AND BOUNDARY CONDITIONS . ..

. .... 5 3.1- Review of the Mark-1 Liner Failure Study .. . .

. . . . . ....... 7 3.2 Best Estimate Case . .

. 9 3.3 Conservative Case. .

.. . . . .... . 10 f ABWR EX-VESSEL FCI CALCULATIONS . . . . .

4

. . 10 l Results of the Best Estimate Case 4.1 . 21 4.1.1 Sensitivity Calculations for the Best Estimate Case . . ....

l 21 4.1.1.1 Sensitivity to Melt Superheat . . . .. .  ;

24  ?

4.1.1.2 Sensitivity to Water Pool Level .. ..

. . .. 27 l 4.2 Results of tne Conservative Case . . ..

. . .. . . 5 CONCLUDING REMARKS . . +- ,,

5. . .

..... .... .. . . . .......37 l 6. REFERENCES . . .. .... .

?

\

l l

l l

l  :

l l

ERl/NRC 93-203 Energy Research. Inc.

1 i i

9 1

LIST OF TABLES l

\

Table 1 Comparison of ABWR Design Features to Existing BWRs I f l l Table : Summary of the Merk-I Liner Failure Conditions [1] l

. ........ .. 11 l Table 3 Summary of TEXAS-Il input parameters .

l ABWR Ex-Vc. .el Local Pressure Impulse (Best Estimate) .. .. . 20  :

i Table 4 Table 5 Best Estimate Local Pressure impulse Sensitivity to Melt Superheat

... .... .. )  !

(75 K) . ....

Best Estimate Local Pressure impulse Sensitivity to a Higher Pool Depth l Table 6 ' -

of 5 m  !

ABWR Ex-Vesse! Local Pressure impulse (Conservative Estimate) . . 23 Table 7 M t Table S Summary of ABWR Ex-Vessel Local Pressure impulse Calculations i i

e t

i l

I l

I l

ERl/NRC 03-203

- Energy Research. Inc.--

- . .. .. _ ,, ~ ^ ~ ~ ~~ - - - - - - = - - - - _ .

11

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

1 I

l t l .

l I

c LIST OF FIGURES Figure 1 Unfragmented jet length as a function of jet pour. time for the Best , J2 l Estimate mixing calculations (1 m pool depth) . .

j Figure : Local said fraction for the Best Estimate mixing calculations (! m pool

. 13 l depth) .

f Figure 3 Local explosion pressure for the Best Estimate explosion calculations j4 (l . .

m pool depth) ... .. .. .. . . . . . . . .

l; Figure 4 Temporal development of the local explosion pressure for the Best

........ .... . 15  ;

Estimate explosion calculations (1 m pool depth) t Figure 5 Unfragmented jet length as a function of jet pour time for the Best

.. . 16  :

Estimate mixing calculations (3 m pool depth) . . . . . . . . . .

Figure 6 Local void fraction for the Best Estimate mixing calculations (3 m pool ,

......... .... ... . . 17 depth) . . ..  :

t Figure 7 Local explosion pressure for the Best Estimate explosion calculations (3

.. . . .. . 18  ;

m pool depth) .. ..

Figure 8 Temporal development of the local explosion pressure for the Best

........... . . . 19 l Estimate explosion calculations (3 m pool depth)

Figure 9 Temporal development of the local explosion pressure for the Best Estimate (melt superheat) se isitivity calculations (1 m pool depth): 75 K

.. . . . . . . ... , . 23

! melt superheat. . ... .. .

Figure 10 Temporal development of the local explosion pressure for the Best  :

Estimate (melt superheat) sensitivity calculations (3 m pool depth); 75 K

.. 23  :

l melt superheat. . .... .

i 1

Figure 11 Local explosion pressure for the Best Estimate pool depth sensitivity 25 l calculations (5 m pool depth). . . . ....... ............

j Figure l'! Temporal development of the local explosion pressure for the Best

. . 26 l Estimate (pool depth) sensit' vity calculations (5 m pool depth). .

i t

I ERl/NRC 93-203. j Energy Research. Inc.  ;

_ j;; _ _ _ ,

t 1

I

1. INTRODUCTION The potential for Fuel-Coolant-Interactions (FCI) exists when the hot molten fuel teorium interacts with water during the in-vessel and/or the ex-vessel phase of severe accidents in a Lig In either case, these pheromena may lead to an energetic (vapor Water Reactor (LWR).

explosion) or a non-energetic (steam spike) interaction which could potentially challenge th containment integrity. The hot liquid (fuel) will rapidly transfer heat to the colder, more volatile liquid (coolant), generating steam, which may lead to a high local pressure. An energetic F (steam explosion) will occur if the pressure rise time scale is short compared to the time scal for inertial pressure release (Mach number > I). A non-energetic FCI (steam spike) does not have the shock wave characteristics of the steam explosion, and the pressure rise time scale is

)

considerably longer. The processes leading to an energetic FCI can be characterized by the following four phases:

(1) premixing of fuel and coolant, (2) triggering, (3) propagation, and (4) expansion.

The premixing phase of the interaction is complex; however, experiments and numerical simulations / analyses continue for cetter understanding. The premixing phase describes the interval during which a hot liquid material, initially as a coherent glob or as a stream pour.

penetrates a cooler liquid, breaks up into small panicles, and is dispersed into the volume o cooler liquid. This ' premixing' is believed to aid in increasing the energetics of an explosion.

This phase could be described qualitatively by the condition that the two liquids disperse into one another. The phenomena that characterize this mixing process consist of (1) heat transfer from the hot liquid to the coolant due to film boiling, (2) evolution of steam and hydrogen, and (3) fuel breakup by the relative-velocity-induced fragmentation.

l Triggering is a local small-scale phenomenon which initiates the fragmentation of the fuel. Most j

fuel-coolant-interactions appear to be initiated by the collapse of the vapor tilm layer or bubble l

in a localized region which may arise spontaneously, or it could be triggered by an external l

pressure pulse. The fuel-coolant mixture can produce high pressure vapor when undergoing a vapor explosion, and do work against its surroundings. Once the explosion trigger is initiated.

the pressure pulse propagates through the mixture. The explosion propagates spatially with a l velocity which is greater than the speed of sound in the region ahead of the shock front in a somewhat similar fashion as a chemical detonation wave. During the explosion expansion. work is done on the surroundings that could cause damage to the structures.

ERl/NRC 93-203 Energy Research, Inc.

1 . ._

l t

a

i t

it has been established that the necessary minal condition for a steam explosion in LWR is the Thus, the tragmentation model will formation of a coarse mixture of fuel and coolant.

eventually decide the predictive course of FCI processes durmg severe accidents.

i 1.1 Objectives l j

The objective of this report is to investigate the potential loads resulting from ex-vessel fuel-coolant-interactions on the General Electric (GE) Advanced Boiling Water Reactor (ABWR) l containment. This investigation is limited to a pouring contact mode" only. As part of the j

initial phase of this study, containment loads resulting from ex-vessel FC!s will be assessed, j

assuming a best estimate and a conservative variation from the best estimate assumptions, usmg j

the probability distribution functions of relevant parameters established by Theofanous, et al. [IJ.

j These probability distribution functions will be adjusted for ABWR-speci6c conditions as l  ;

' needed. The fuel-coolant-interactions analyses are based on the TEXAS-Il computer code [2]

i with auxiliary calculations, thermodynamic analysis, and simple parametric calculations. r 1.2 Description of the TEXAS-II Fuel-Coolant-Interaction Model The TEXAS-II computer code is based on a one-dimensional transient model for hydrodynamic i

calculations developed at Sandia National Laboratories (SNL) and modided for fuel-coolant-

! interactions by Chu [2]. TEX AS-Il solves the one-dimensional. Inree-6 eld equations desenbing the fuel-coolant-interactions and hydrodynamics. Two 6 elds represent the coolant as liquid and l i

vapor, and one field represents the discrete fuel particles. The liquid and vapor Gelds are solved using the Eulerian technique and the particle phase is treated using the Lagrangian formulation.

j In this model, the governing consetration equations for each phase (i.e., liquid, vapor, and j

particle) are written separately, which allows thermal and mechanical nonequilibrium between the phases to exist. The effects of condensation, evaporation, and interfacial momentum - l l

transport are included as source terms in the partial differential equations. A hydrodynamic I fragmentation model based on the Rayleigh-Taylor instability mechanism is implemented in TEXAS-II. It is posttdated that this mode of instability is dominant in FCis [2]. The dynamic fragmentation model implemented in TEXAS-Il predicts the Lagrangian particle size at a new l l  :

time (n+ 1) using the 6 eld variables at the old time level (n) without any reference to the history '

of the particle. The particle diameter at the new time level is then given by the expression (1) >

D "* ' = D " ' l - C} T* We V) where AT' is the dimensionless time step, We is the Weber number evaluated by the relative velocity and density ratio of the continuous-dispersed phase at the old time level. and C, and C.

are empirically determined constants. This linear fuel breakup model is developed from a complete theoretical nodel [:). The TEXAS-Il code reuuires me user to de6ne the sptem geometry. the initial and the bouncary concitions. The fuel entry moce mio the pool can ne ERUNRC 93-203 Enerev Research.Jnc. _

2 i

I

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

i

[

modelled as a' coherent jet or in the form of discrete prefragmented particles. To presenbe the ,

initial conditions in the TEXAS-Il simulation of a FCI. the radius of the particle / coherent iet l In along with the fuel mitial velocity and thermophysical properties must be established. ,

Three  ;

addition, the water pool conditions and _ initial vapor vord fraction are also required.

l options exist to determine the proper boundary conditions at the top or bonom surfaces, i.e.,

l reflective boundary condition free gradient boundary condition. or the constant pressure-  !

boundary condition.  !

e i

i i

i l

h i

l l

l I

l l

l l

i ERl/NRC 93-203 Energy Research, Inc. _ _ _ , _

- - - ~-

{-

i l

6 4

2. ABWR DESIGN FEATI'RES i l

A summary of plant and containment system design features important to core me;t progressmn. i and containment response with particular imerest to ex-vessel steam explosions is provided in '

Table 1. Also listed in the table are design data.for three current generation BWRs. namely. -

Peach Bottom (Mark-1 containment). LaSalle (Mark-ll containment), and Grand Gulf (Mark-l!!

l-l containment). It can be seen that the ABWR is most similar to BWRs with Mark-I containment l insofar as design features are concerned, but it should be noted that the geometry of the ABWR containment design bears a marked resemblance to the Mark-Ill containment. ,

t l Table 1 Comparison of ABWR Design Features to Existing BWRs ABWR Peach LaSalle Grand Feature i Bottom Gulf Power level. MW(t) l 3.926 l 3.293 l 3.323 l3.S33  ;

Volume of suppression pool water, m' l 3.580 l 3.898 3.590 l3.851

! 39.640 Volume of suppression chamber, m' l 6.0004** 3.591 S.050 Volume of drywell, m' 7.350 l 4.563 l 5.930 l 7.650 l Mass of UO,, kg 171.600*' 159.412 158.490 l 165.878 f Mass of zircalloy kg 78.550** l 65.491 70.534"' l ' 33.107 l Suppression pool water volume / power, ,

1.0  :

3 m /MW(t) 0.9 1.2 1.1 i

Containment free volume / power, '

3 3.4 2.5 a.1 12.3 m /MW(t)

Zr mass / containment free volume. kg/m3 l 5.9 8.0 l 5.0 l 0.7 Fuel mass!comainment free volume, ,

12.9 19.6 11.3 3.5 ,

kg/m 3 Low water level

• Nnt dew;n data trom GE repon CEB92 12 (issue 21 provaied by NRC "8 Ctadding 09.696 kg). Canister 00.S38 kg) i Energy Research. Inc. ER1/NRC 93-203

i i I

)

3. i INITIAL AND BOUNDARY CONDITIONS '

i i

3.1 Review of the Mark-l Liner Failure Study l

t Two different possible acciden: progression secuences were considered in the evaluation of  ;

Mark-1 shell melt-through failure issue [1]. Scenario I was based on the massive slumping of i the core.

This scenario assumed that a large amount of UO, was present in the initial melt-composition. A large initial release was predicted, due to sudden massive core slump, followed  !

by a more gradual release later in time, dictated by the decay heat. Scenario 11 assumed that

[

the initial melt was mostly metallic, since these materials were found to be the first to melt due I to their lower melting temperature in comparison to that of UO 2 . This melt was then found to  !

be released into the cavity. '

i The scenario I was based on predictions of the M AAP computer code developed by Fauske ani Associates, Inc. In the MAAP simulation of the core melt accident, UOrZr mixture melts and l relocates into the lower core plate. After a user-specified time (1 minute default), the core pli fails and the UOrZr-Stainless Steel mixture relocates mto the lower plenum, where the melt  !

attacks the penetrations in the lower vessel head. This entire volume will be released after the  !

failure of the lower head. The melt pour was found to include a large amount of UO,.

The scenario 11 was based on predictions of the BWRSAR computer code developed at Oaki Ridge National Laboratories (ORNL). BWRSAR calculations showed tnat the initial melt i l

composition was mostly metallic. which then proceed to the lower plenum.

allowed to quench, and then remelt. The melt was 5 instrument tube), leading to the gradual release of melt.At this point there was a loci The calculation then permitted the  :

remaining material to be released as soon as it became molten, allowing for more gradua l

! rates than seen in scenario 1. The BWRSAR analysis was found to lead to low melt supei (of approximately 50 K [3]), and the melt was found to be composed of mostly metallic constituents.

Theofanous, et al. [1] assigned uncertainty distributions in the form of Discrete Distributions (DPDs) to the amount of melt released in terms of the tot .'

on MAAP calculations, the melt volume was found to be as high as 12.5 m' (more than 50%

l of the total core volume), and the authors believed that this value represented al '

i bound. This value was then used in the determination of the probability distribution fun for the amoum of melt released. The most probable range for the volume of melt rel; 3

i assumed to be 11-13 m . The authors also determined a probable range for the amount of Z ,

in the UO/ZRO ceramic mixture by assuming that the debris released is at core co 2

ti.e.. Zr composition corresponds to 30% by weight and if 30% of this weremoxidized . est a

with 20% metallic Zr in the mixture would be obtained (1]). The most probable rangei '

percentage of Zr metal in the melt was found to be 10-20% by weight.

Energy Research, Inc.  !

1 ERl/NRC 93C03  ! 1 i

l i

l t

?

The melt superheat was assumed [1] to be small for the following reasons:

I (a) Convective heat transfer to boundaries could be sustamed at decay heat levels.

i (b) Continuous meltmg and incorporation of boundary material tends to keep the meh temperature near it's melting point, and (c) Heat losses to water and control guides during relocation to the lower plenum; i s

Therefore, the most probable melt superheat was assumed to be in the range of 25 K to 50 K.

A workshop on BWR Lower Head Failure in 1989 (Reference [1], page 13) determined that the '

only potential pathways for failure were the instrument tubes (diameter of 2.5 cm) and the drain plug (diameter of 5 cm). It is stated in Reference [1] that "the timir.g of such a release is .

usually based on a 2.5 cm failure with area enlargement during the release process (ablation).' l However, the very nature of the accident scenario implies that there will be multiple penetrations i failing simpitaneously. The pour rate for an initial 5 cm hole was found ta be = 0.5 m'/ min. l

[1] (based on a nn a Mb of about I m in the lower plenum). Under these conditions, melt l pour rates up to 4.0 m'/mit (1) were expected which corresponded to eight penetration failures based on an initial 5 cm opedne diameter.

The melt release rates for scenario II were based on results from BWRSAR. The early penod l was approximately the first two hundred minutes of the sequence and the averaged short term '

release rate was found to be = 0.15 m'/ min [1]. This initial perioc consists of periods of melt accumulation (no release) followed by periods of high pour rates (peak pour rate of 0.7 m'/ min i is cited in Reference [1]). The peak pour rate, however, occurred approximately 100 minutes -

after the initial pour of molten debris from the reactor vessel began. For scenario 11. the amount  !

of Zr in the melt was observed to be much larger, and was given a range of 20-40% of the total l melt release. The melt superheat was estimated to be between 75 and .100 K. A summary of. ,

the initial conditions for the Mark-1 liner failure study [1] is presented in Table 2. The initial j hole diameter of 5 cm for scenario I is based on the MAAP calculation and the initial hole size of 2.5 cm is based on the results of BWRSAR [3] where penetration failure occurs via an instrument tube.

The most important initial condition for an ex-vessel fuel-coolant-interaction is the initial melt

• release immediately after lower head failure. The most energetic interactions occur when the initial melt ejected from the lower head interacts with water. Therefore. the longer term behavior of the melt pour (and the peak release rate of 0.7 m'/ min after 100 minutes) considered in Reference [1] is not relevant to the assessment of ex-vessel steam explosions. and it is not included in the present analysis.

Energy Research, Inc. ERl/NRC 03-203 6 - .- _ - .. --

p ..a r,. s er

}

Table 2 Summary of the Mark-1 Liner Failure Conditions 11]

l Scenario 11 l  :

f Propenies l Scenario 1 p l l 0.5--LO t m'!mm s  ! 0.15 (m'imm) l Release rate 67% Stainless Steef f i Melt composition 20% Zr 70% UO, and 10% ZrO, 33% Zr l  ;

(weight %)

l 25-50 K l75-100K l Superheat 5.0 cm 2.5 cm  ;

Hole size  :

l1 4

Number of penetrations ll-8 f. -

r The initial vessel breach diameter would not be relevant to the FCI calculations. As the meit ejects from the vessel it will be accelerated by gravity and the diameter of the jet at the pool surface will be smaller (thus keeping the mass dow rate constant). l It should also be noted here that the conditions for the FCI calculations are not meant to exactly mimic the conditions applicable to the shell melt-through issue, but rather obtain insights to i Two cases develop appropnate initial and boundary conditions for the present FCI analyses.

representing a "Best" estimate and a -Conservative" variation of the best estimate are considered in the present study and they are discussed in the following sections.

r 3.2 Best Estimate Case i l

The 'best-estimate" case proposed for use in the present calculations is based on scenario 11 f i

(i.e., the initial debris composition is metallic). The melt is assumed to be composed mostly '

of zircalloy and stainless steel, at a relatively low temperature. The initial melt temperature is assumed to be 1833 K (the melting temperature of the pour material is about 1783 K [3]). which corresponds to an initial melt superheat of 50 K. This superheat value is slightly lower than the values given in Reference {l); however, it is more compatible with the melt superheat value given in Reference {3). The Zr and stainless steel content of the melt is assumed to be basec on the initial melt pour composition reported by Hoege [3] (i.e. 67 w/o stainless steel and 33 .

w/0 Zr). P The quantity of water in the lower drywell region of ABWR under severe accident conditions '

remains uncertain. It is stated in the General Electric SSAR P] that under most acciden conditions the lower drywell region is expected to be dry, and that the most probable accidem '

scenario under wnica water can exist in the lower drywell is a small break Loss of Coolant Accident (LOCA) in the lower vessel head region. In these accident scenarios, the steam . leaving -

the lower head escapes directly into the lower drywell volume. Most of the steam will Cow and ERI/NRC 93-203 Energy Research, Inc.  :

7 6

, _. , p. 4-,

. , p

- t I

t l

condense in the pressure suppression pool. and the remainder will condense m the lower dryweil region. In the present analysis, the 6nal, post blowdown pressure of the lower drywell region l

is assumed to be equal to the sum of the original pressure of the drywell and tne hydrostatic j

head betweta the vent and the suppression pool surface (see Figure 15. Reference [4]). The water pool formed in the lower drywell and the steam-mtrogen mixture above the pool are

• assumed to be at thermal and mechanical equilibrmm (i.e.. equal temperatures and pressuress.

The temperature of the lower drywell atmosphere is the saturation temperature of steam l

(corresponding to its partial pressure) in the atmosphere. It is estimated that the total pressure i

of the lower drywell atmosphere is approximately 1.4 bars. The saturation temperature of steam at its panial pressure (0.4 bars) is 349 K. Thus, at vessel breach. it is assumed that the pedestal l

water pool has a temperature of 349 K. '

There are other possible sources of water in the drywell and these include water flow from the wetwell-drywell connecting vents and the passive Dooder in other accident scenarios (Section l

X.2.7.6.2.2.1 of Reference [4]). It is stated in Reference [4] that neither of these caths are i credible sources of water to the lower drywell~for a possible FCl. However, it is our intention l

to vary the height of water in the lower drywell to study the effect of the water pool depth upon the FCI energetics. t 11 can be assumed that in low pressure accident scenarios the melt ejection velocity (from the RPV) can be estimated using the gravity head between the vessel lower head and the water pool surface in the lower drywell. The maximum velocity of the melt jet at the pool surface, V,, ,

i can be calculated as  !

C)  !

V,=/2;.:.1H where .aH is the distance from the lower vessel head to the surface of ine water pool on the pedestal floor.

Section X.2.7.5.2 of Reference [4] states tnat the water level in the lower dryweil will not be l j

! greater than the suppression water level during a severe accidenL The actual height of the water l

i pool is related to the volume of water condensed in the lower drywell during the course of the accident. The distance from the bottom of the reactor vessel to the bottom of the cavity is i1.5 l l

m (from Figure 1.2-3c of ABWR schematics) and the diameter of the drywell is equal to 10.6

m. The molten corium pour rate is estimated to be about 16.7 kg/s (which corresponds to a 3

conum volumetric pour rate of 0.15 m / min), and the _ water pool depth will be parametrically j

varied from 1 to 3 m. Based on this corium pour rate and the coherent jet velocity at the pool 11 should be j surface. the jet diameter is calculated to yield the mass now rate of 16. ig: 3.

i noted that the volumetric corium pour rate of 0.15 m /mm is chosen for tne best estimate case. l l

since any energetic fuel-coolant-interaction is expected to occur during the very early period of i

corium pour from the vessel. The peak flow rate of 0.7 m / min (Table 3.3 of Reference [1]) j

~

ERl/NRC 93-203  !

E.nergy Research. Inc. . -__ _ _

1  ?

I l "mw vm, .

does not occur until approximately 100 mmutes after vessel breach, which is not relevant to the problem at hand.

3.3 Conservative Case 1 i

l Based on the past severe accident analyses reviewed in Section 3.1 beenano 1). the conserva estimate for the molten corium pouring out of the vessel is assumed to be given by a l composition of UO:/ZrO (assumed S0/20 w/o) at a temperature of 50 K above its liquidus I 2

This temperature is consistent with the assumed oxidic temperature (assumed to be 2800 K).

melt pour, and is based on the upper bound value of the MAAP input parameter tile. The .

corium pour rate is estimated to be about .540 kg/s, which is based on simultaneous failure l S penetrations (with initial opening diameters of 5 cm). This value corresponds conservative' in this studyto the maxi:

pour rate of 4 m3/ min in scenario 1. This pour estimate is termed ~l since it involves a large coherent pour over a short interval of time. The drywell . initial'and l

boundary conditions are assumed to be the same as the be'st estimate case, and the water  !

depth is parametrically varied from 1 to 3 m.  :

l i

l l

i I

i i

i

}

t ERl/NRC 93-203 Energy Research. Inc.

9

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

l i

l . ,

l i

4. ABWR EX VESSEL FCI CALCULATIONS l 1he TEXAS-Il FCI model [2] is used in tne simulation of fuel-coolant mixing to determme the extent of the corium breakup and mixing before an explosion is triggered. A numoer of l

calculations are performed to observe the overall behavior of the TEXAS-Il model. Because the pool has such a large surface area (8S.2 mh. it is unrealistic to consider the entire cross-l sectional area for mixing, because the melt would pour in the central region 2 where the fuel-coolant mixture is generated. For the nominal calculations, an area of 2 m is assumed. l l i i

The initial location of the coherent jet iniection for the TEXAS-Il calculations is 3.5 m above

~

the lower drywell basemat (corresponding to the free fall height of 8 m). The jet entry velocity

'or both the best estimate and the conservative cases is determined to be 10 m/s (see Equation l l

3) by the acceleration due to gravity over a fall height of 5 m. The tacit assumption in this calculation is that the below vessel structures reduce the jet momentum (and thus its velocity) l l

by a small amount (approximately 20%) corresponding to an effective free fall height of 5 m. '

The jet diameter at the pool surface is calculated based on the jet entrance velocity to tne pool t and the corium mass flow rate. The depth of the water pool is varied from I to 3 m in these calculations. As mentioned previously. the most important uncertainties are associated with the Table 3 lists the l corium pour rates and the quantity of water available for the interactions.

numerical values of the major variables used in the TEXAS-Il simulations of the fuel-coolant-interaction for the General Electric ABWR.

t i

4.1 Results of the Best Estimate Case The results of the ABWR ex-vessel calculations for the best estimate case are shown in Figures .

I through 8. The analyses indicate that the leading edge of the of the fuel jet penetrates the

~

water pool and mixes rapidly; however, the unfragmented jet slowiy makes its way down to the basemat. As the jet enters the water pool. the leading edge of the jet is fragmented and it is broken up into smaller size particles: the unfragmented jet length signi6es the length of the coherent jet measured from the surface of the water pool. It is estimated that it takes approximately I second for the jet to reach the bottom of the shallow I m pool, and it takes l longer for the deeper 3 m pool. Thus, given the initial conditions for the ABWR situation. it is determined that the most critical time for an explosion to occur is about one second after the pour is initiated. For a shorter time period, the mixed mass of the fuel would be less. and longer time periods would cause the jet to reach the pool bottom and accumulate on the pedestal floor and in the process the melt jet would be partially solidified. The explosion trigger is, therefore, set at I second for both the shallow and deep water pools.

i ERl/NRC C3-:03 Energy Research. Inc.

~ 10

~ -- - -

~

i

i. ,

i i

i i

I l

Table 3 Summary of TEXAS-Il input parameters ,

l'roperties Best Estimate Conservative l f 0.15 (m'/ min) 4 (m'/ min) . l Melt Volumetric Pour Rate l 16.7 kg/s l 540 kg/s-Total Melt Pour Rate UO; 80 wio  !

Melt Composition Steel 67 w/o Zr 33 w/o Zro, 20 w/o i i

Melt Superheat 50 K 50 K i i

I Melt Temperature 1833 K. 2850 K Melt Density l 6545 kg/m' 8000 kg/m' Coherent Jet Velocity 10 m/s l 10 m/s ,

Coherent Jet Diameter . l .S cm 8

3.3 cm' .

Above the Pool Surface Number of Coherent Jets l1 l8 j i The coherentjet diameter at the pool surface is calculated based on the given pour rate  ;

of the melt, the melt density, and the jet velocity before it enters the pool. The jet velocity is calculated using Eouation (2) and the jet accelerates as it leaves the vesse!.

and thus it will have a smaller diameter.

.j i

ERl/NRC 93-203 -

Energy Research. Inc. __ _

11

- ^' y-tn e-

i

. b

,i k

b

. i s 1 .

! h I

0.8 - -

o ,.

v

= s 5 -  :

E u

0.6 -

J  ;

U  !

l -=  ;

2 i

= ,

u

= -

-o 0.4 --

l a i c

=

U  !

L 0.2 -

l 0'!' '

0.4 0.6 0.8 1 O 0.2 Jet Pour Time (s)

'ntragmented jet .engtn as a Tuncuan et jet pour um: tot tne cest esumate

-:.gura 1 7

mixine calculations (1 m pool dep:h)

.. 1 i

ERl/NRC 93-203 E.nergy Research._Inc. _

l f

- -. , r n .

I L

1 F

1 1

l

.______ _ _ _ ,_____i___ _

3 ,7____i__

i t=0.25 s :

t=0.50 s ---

t=0.75 s -

t=1.00 s :+-

0.8 -

0.6 -

=

=

=

u

_.=

-=

=

s 0.4 -

0.2 -

I , ,

m

_ _ a_i i

O -_

0 1 2 3 4 5 Dis:2nC: from the Bottom of Pool (m)

-DOSI _51; male n:K:ng C2lcula*.lonS t i ~ 300i COO:n!

~0021 VO;C :~2CDon ICT*T.

-I~lEUTC -

Enercy Research.- Inc. -- . _

rRT.cN.

- R C.C... i 'O,' _

4e

\

i i '

l le47 , 4 i 4 a t=1 ms -+--

t=2 ms *--

t=3 ms -

t=4 ms -x-t=5 ms + - _

Se46 -

=

c _

~

o 6:46 -

m W

O

=

=

m 5 ae46 -O x

' *J A ,

li 2

y  %

8  %

Il  %

2e46 3' '+s

\

lf"1 r i .-

l' c 5

\s a.a '=

x...%, ... .r s. ks l\ ,

l m . c ,%, (= ...%_+_.._.n.

. M_ \, _

0 . - -- ,

1.5 2 2.3 a 2.3 -

0 0.5 1 Distance from the Bottom of Pool (m) l Cigure Local explosion pressure for Se Best Estimate explosion calce!ations (1 m pool depd)

- Enerry Research, Inc.

ERl/NRC 93-203 A*

l

, t I l t

i j

l i

.I i

l' t

-l r

t e

i le+07 i i . . . 4 6 8 x=0.52 m i l x=0.92 m ----  ;

x=1.32 m -- --  !

x = 1.72 m - - -- i x=2.10 m - - -  ;

Se46 -  :

i t

i 7

~- -

i 6e+06 - l l

o_  ;

=

= r I Z t l

Q s 1

.=x r

ix 4e+06 -

I

1. \ ,

I \ -

l 2e46 -

's i

1 ,i  %

r s

I s I s

/ s

/ $

~

/ s, y ' ~~- ._

______.==~ ___ -

m_ ,

0 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01 Time (s) 1 Temporal cevelopment of :he 10:21 explosion cressure for tne nestesuman

-. j

'rigure a explosion calculanons U m pool cep:n. ) l i

C Rl/NRC O' '03

-- 1

-:n=rgy Research, Inc. ------ - -

15 l l

l 4

1

- , , _, ,_,i

1 s

.. f i

s i

i i

t

., , i . .

2.5 -

i I

= , _.

i

. _o-l

=

^t

_8 i U

l = 1.3 -

-l i 2 .>

' =  ;

! w  :

j' =

• j t

=

=

c= -

1 D  !

i I

0.5 -

i 0

0 0.2 0.4 - 06 0.8 1 Jet Pour Time (s)

';igure 5 Unfragmented jet length as a function of jet pour time for me Best Esti?. ate .

i l

mixin_a :alculations G m pool depth) .

l ERl/NRC 93-203 Encrev - - Researc.h.

. - - Inc.

16 f

I I

r ,- 1 3, e~v ,-t ---r-- , yw14r w h

i i

1 , , y____i-__

l' t=0.25 s o t=0.50 s ---

t=0.75 s -D --

t= 1.00 s -*--

0.8 -

! 0.6 -

5

=

w 0.4 -

0.2 l

0

_r._. . . . _i_ _ _ _ ..

0 1 2 3 4 5 Distance from the Bottom of Pool (m) pool deotn >

,_oca,l . . . . .

t -

-Irur: o vo:c t acuan tor tne nest sumate mamg calcmanons te l

l Ener;y Research; Inc. - - -- -

. ERUNRC 93-203 _

j I a

l

, i i le47 6 * ' a t=1 ms O t=2 ms ----

t=3 ms -D --

t=4 ms - x----

t=5 ms +-- _

Se46 -

o -

_c.

6:46 E 4 m

G. 9 -

m . ,

o .

m. .

'm

= .

C m .

m o -* -

+-

= .

de46 . i ,..

n x

,l

'+. ' + , . . . '.

m y ~-y.

.a

.8 x...x.'xx*.v >V, ~'~ s .s. s rn y

x.

w -

e46 ,+

, ..50

u. .~

t

.a . 'y 1

• s s

,2 -

.- . E ,<@ . e.

Q l -

..? 5 4...D. 9

../ ~%,..k..

L-

+..

a -

. , -==v, .4 ~

s s..,,.u._w.s2 <

2 2.5 3 3.5 4 0 0.5 1 1.5 Dis:ance kom the Bottom of Pool (m) rigure <- 7_ocal explosion pressure for the Best Estimate explosion ca! ulat:ons t.' m pool l

depth)

Energy Research, Inc. ERl/NRC 93-203 i - - -

1 i

i

. i i

i

?

r i

r 1

i . .

le47 i i . 4 i a x=0.52 m x=0.92 m ----  !

x=1.32 m --- -  !

x= 1.72 m -----

' x=2.10 m --- - - ,

i Se+06 m 1 es L

o 6e+06 -

l' ss W

0 <,

L o

- r, s ,

i F

. .C. \ >

> w

~

.: \ -

', i

i. 4e46 '- l ,.. . . .

x , *

. s.

• g .

,*, '. rt I

3,. .

. s

'.e. ,

• ' . k. * . .

.t

[-e , .. ,

. \* * .- s t\ l

, }-

I.; , f -

s 9 -

,. p /

.A*%. ,

I :,.

. , g .r ~.

. , ~~.

.e I r<-. .

2:,*w r- , s!

I.: ' 's,

, g, '.

i l.

• g t, w.

\s ll ,* II *-- *-.' 2 . ~-* ..n. n! .

s.s. . . . *- .= .--- . . . . ::: .a.nt<m ~.-

0' O 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01 -

l ume (s) l l

l

-l

_ temporal cevelopment of the local explosien pressure for the Best Estima:e l

gure d .

explosion calculan.ons o. m pool aeo. th) .

I i

- ERl/NRC 93-203  !

! E.nergy Res arch. Inc. ~-

19 l

i

in the case of the shallow pool, the jet reaches the bottom of the pool, while m the deep pool the edge of the jet is halfway through the ater. Figure; 2 and 6 show the void fraction :n the yol for the 1 and 3 m water pool depths, respecuvely. The void fraction is defined as the ratio of the vapor volume to the temi volume. At the star

• of the calet.!ation. the twl depth is either

. m nhallow pool) or 3 m ideep pool). The correspondmg mid nLetion m the pool region is then given a value of zero. Above the pool surface, only vapor is present and the initial void fraction is one. As the interaction progresses, the melt thermal energy w'ient is transported to water leading to generation of steam. The local void fraction in the water pool is increased in time as more vapor is generated winin each computational cell. The peak prenures from the FCI calculations for the shallow and deep pools are 40 cars (I rc .liseconds after the exp!osion trigger, see E" ore 3) and 60 bars (3 milliseconds after the explosion trigger, see f'igure 7),

respectively. It should be remembered that these ace tne local peak pressures, and the pressure at the arywell wall is expected to be redeced by approximately 1/r (where r is the radial disance The to the drywell wall), as the pressure wavr expands radially within the subcooled water.

rationale for this pressure variation comes from both theory and experiment. Sim"le theoreticaf; analysis based on irrotational flow theory for the motion of an initial spnencal cavity produced by an underwater expl'sion is provided in Batchelor [5] and McCormack and Cranc [6]. The anal, ical pressure distabution shows that the pressure at a distance away from the cavity center decays as 1/r. The numerical experiments using PM-ALPHA computer code by Theofanous et al. [7] have also shown the same behavior. In addition, the experimental measurements [8]

based on scaling laws for conventional explosives have provided more .unvincing evider-the lir-dependence for pressure expansion sway from the location of explosion. In ~ c gn explosive experiments. both the pressure and the pressure impulse vary as (1/ry', where or is an emp:rically determined coefficient that range from 0.8 to 1.2. Figures 4 Isd S show the time development of the pressure wave at different axial loca' ions within the pool. The impulse loads j

at these locations are ca!culated and tabulated in Table 4 Table a ABWR Ex-Vessel Lacal Pressure impulse (Best Est; mate) i Pressure lopulse (kPa-sec) Pressure I.mpulse (kPe-sec)

Axial Distance from the Bottom of the Pool, m for the 1 m Pool Depth p for the 3 m Pool Depth 3-10.S l 13.6 0.52 l ._

6.7 h 13 4 y 0.92 0 d 13.2 l 1.32 l l 1.72 h l

2.10 h h . _ .1 l f

ERl/NRC 93-203 Energy Research, Inc.

i i

4.1.1 Sensitivity Calculations for the Best Estin. ate Case i

A series of sensitivity calculations for the best e,timate case are performed whtch melude:

(1) Sensiuvity to a higher melt superheat of 75 K. and ,

C) Sensitsvity 'o a higher water pool level of 5 m. l t

The higher melt superheat value of 75 K corresponds to the conditions assessed by

[ et al. [1]. The results of these sensitivity calculations are described below.  ;

l t

4.1.1.1 Sensitivity to Melt Superheat  :

l The results of the FCI explosion calculations for the higher melt sup"iieat of 75 K are sho in Figures 9 and 10 for both the 1 m deep water pool and 3 m deep water pool in drywell. The calculated local void fraction and local pressures were in As observed seen to be best estimate case (with the melt superheat of 50 K) and are not shown here.  !

Figure 9, the explosion results are similcr to the 50 K melt superheat calculation pool. The results of the calculations for the 3.o pool (Figure 10) shows that tl l

e:plosion pressures are slightly higher than the corresponding pressures for the l superheat of 50 K. Table 5 lists the local pressure impulse as a function of distan i l bottom of the pool. As ae melt superheat is increased from 50 K to 75 K for the best el case, the local pressure impulse is mereased by less than 2% for tne 1 m pool and byl I

approximately 10% for the 3 m pool.

l

! Table 5 dest Esti-ete 1.ocal Pressure impulse Sensitivity to Melt St.perneat c5 K) , j i

l Presse e Impulse skP2 see)

Axial Dis'.ance from the [ Pressure impulse (kPa-ser; , far the .; m Pool Depth 4

Bo.2m of the Pool, m l for the 1 m Pool Depth 11.0 h 14.9 f 0.52 6.S l li.S 0.92 l .

l4.6 l 1.32 j --

[

-- l 14.4

! 1.72 _

h l i3.3 2.10 -

l i-ERl/NRC 9t:03

~

- iinergy Research.-lac. -- ._ _ -- -- - - - - - - ~ . _

l I .. .. .-

i le47 i , , 4 i i 4 , ,

x=0.52 m x=0.92 m ----

x=1.32 m ----- l x=1.72 m -- ,

l x=2.10 m - - - _

8e+06 -

l M

O 6ed6 -

o

=

N u

1 1

=

.2, o -

=. 4e+06 -

x

s .

-~.

! t g '

2: 4 5 - l 's s s I g

$ \

! j g l r

1 I

i g l ~

f %m' -

L, ----_____,___ .

0'  !

0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01 1

- ume i

. ip;: 9 Temporai development of the local explosion pi :sure for the Best Estimate tmelt superheat) sensitivity calculations (1 m pool d.,,th): 5 K melt superheat.

t ERl/NRC 93-203 l _ _

_ E.nergy Research, Inc. - _. .. .

l

! l i

l l

I

4 l

J i

1 1

1 I

t 1

i i , , ,

le47 i i , i i ,

x=0.52 m .

x=0.92 m ----

x= 1.3 2 m - - - - - ,

x:1./ , m x=2.10 m - - - -

8e46 - F m i

=- - i

~

o 6:46 -

O es a

A

1. 'g

1. 4 U I \

.- g i t m-

-J  %

I \

i ,. t

.2 8 *

• t s!) * *

\ -

O l ' ' -

=

4e46 -

,i .- .

s i, -

K - ',

=: *^

. 1 m .s l ~*)17 f, l , k. .

9

• 1 f .- A \. , '. g1 1

• ,' \

sg

• 1\

.g t

• $ .b

• l ' 4 2e46 -!!- - l

'\*s.

n ,. Iy ~. % ...

- l I' s g *%.~~'"*'*y*

hl I %lL,%.

s '*,

. , I 1:

,e y i 4s e

h**

• r-r "%.

l' . f *******.!*=g

[

g'! *

. p .

. ****-=.. -.W.1,g.r:.

p .. .. . ... . ,

s,, . - ------ - - - . . .

, , . . i 0 ,

0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01 -

ume (s)  !

?

4 emporal cevelopment or ine :ocal excioston pressure for n Ees: Es: mz;e e melt

:-gure .0 3 superneau se .sitivity calculadons O m pool dep:n); ~4 K ne? supernea:  !

. ERl/NRC 93-203 Enert.v Research..Inc. - -

4 b

i

[

i

+ 1.1.2 Sensitivity to % ater Pool Lesel ,

' he results of the FCI calcu.auons for the 5 m deep water pooi in the ABW R lower dry well are nown in Figures . and ::. For :nis case. the nr. sing cakulations are run :or 1.4 -econos to the seiore the explosion is triggerec. The local soid fraction temporal r.:> tory w.a simii .: ,

oest estimate case and will noto' e meluded here. Comparison of Figures i1 and 7 shows that the maximum explosion pressures for both the 3 m pool and the 5 m pool are approximately 60 '

bars. In this case, however, the maximum explosion pressure occurs approximately 5 ms after i

the explosion is triggered as compared to 3 ms for the 3 m deep water pool. The local pressure impulse for this case is shown in Tzble 6. Although. the dynamics of this interaction is siightly -

different from the best estimate case twith a 3 m deep pool). nevenne!ess, tne local pressure impulses are r.ot significantly altered.  :

Table 6 Best Estimate 1.ocal Pressure impulse Sensitivity to a Higher Pool Depin of 5 m 7  !

i Axial Distance from the Bottom of the Water Pool, m l Pressure impulse (kPa-sec) . l  :

03: l  ::.s N ,

0.72 l :2.2 [

l

' 72 l 20.5 l l .

{  :.32 l 19.0 I

1 ..

e. .i l

_ .m_e ]

l 17 ' _

l,,

l . . l _' ,

I!  !

'l ._.. t I'

I i

1 ,

I l

I l

l ERl!NRC :03 _,

Energy Research. In:. ,. - . . . _ _ . _ ._ _.

1

~

l

(, I I

h

Ie47 *

• i i

!=1 ms s'-

=2 mS ---

t=3 ms _. . - .

=4 ms 9:----

t=3 ms '"- -

Se -

= -

~

o 6e+06 fx X z i .#.

u  :

.Xr

~

-s 4 E  ! x 'z.

! u_

=.

0 8 X

2. 4 -06 + -

'a

-,b*-N ,+" %E j -X

• 7W .+ '

/0 s$.m

-_ -f:L":"

x , .c.L_.

=: - s .- .: - 0 .

2 X s -; O' is

- v - m -

I ' 'XM 3 W- ,'ra o K

l 2e--06 +L , 7- .

l t f <

0 1

, ,s Ts m  ;

L e 0 , X -

0 s

I '

/ a .X

,2 ~.. .m-1 l x ,o ~..,

l m s

O s= . . -

.lj. f . . . A

-~- ~ ~

%ERREEEEEEEEEEN' i O . , - . , i e S i

0 1 2 -

2. _

vis:2n00 Irom theD000m OI .-001 m

,- - < . . . . , ,. 3 3, - . ; ._..v.

y . . .,

. s . 3.

. . ... . . r. . . , .; . . s . y. s

..:;. .e. ,_ ... ._ o _.c. . ._ . . p : c s t o,. s . ., ~.._ . s 3.. . ... . ., . .._ . . ..: ._.

_ s.. . ..

.,0_, .: c .. ~.

-- ERENRC C.-:03 Energy Resear:n.- in:.

c

L f

i i . . . 4 i le47 , , ,

x=0.32 m x=0.72 m ----

! x=1.72 m -----

x=2.32 m x=2.72 m -----

I 8e+05 - x=3.12 m --- -

' x=3.72 m -- -- -

i l

l l

e.  ;

c.

o 6:45 -

,a ',~ ,

=-  !

A m b u.

u

,a

=

.=_

,s ..... .

m , '  !

4e45 -

,/

x

a . ..<.<.,-

. . - ,. .i.,.;.yss,s.. . . . .

a .! .:,.

, - ~. ~.

i - ). .o , s.. s. . . . .

-,. - s.

'\

i .i ,

i

,r. .. ~. '. .- .. ~

2:45 - :. ,,! l .: l s

. ~. .

.. .. .. ... . ... . ... ..\ .s ..

,. , . . ~.

e

. .i

s l ...- -........\s. ~ . . . .

, s s ,~..--

w-. .

x,. ..5~.....-..-

g, , , , ,

0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.00S 0.009 0.01 -

inne (s) t r

I i

Temp ral development ci she io:al :xpicsion pr:ssure fe the Best Esd=n: (poc!  !

. ign 12 depthi s:nsid ity c& iadons (5 m pool depth!.

- - . Energy Reserch. In:. _ _

ERI/NRC 93-203

,D

l i

4.2 Results of the Conservatise Case l In tnts The results of the conservatise case calculauons are shown in Figures 13 through 18.

case. the melt pour was assumed to occur through S penetranons and the total pour rate of the tne corium was 540 kg/s. For Ine 1 m pool depin, and a mixmg cross-secuonal area of .' m local void fraction was too high for an explosion to propagate. Thus, the mixing cross-secuonal f 2 t area was increased artincially to 6 m (increasing the area to 4 m also did not result in an explosion) so that an explosion propagation calculation could be performed, and the presente i results pertain to this cross-sectional area (increasing the cross-sectional areaThe artificially decreases the local void fraction within the one-dimensional computational framework).2 For mixing cross-sectional area for the 3 m pool dep:n was kept at the nominal value of 2 m . i the sake of comparison, calculations wue also performed for the 3 m 2pool depth and cross-I sectional areas of 4 and 6 m2 . The explosion calculations for the 6 m area were terminated 2  !

For the 4 m after 7 msec due to a sudden decrease in the time-step (reduced to Ix10* second).

cross-sectional area, the maximum local impulse was calculated to be 28 kPa-sec.

The local void fractions at various times are shown in Figures 13 and 16 which indicate a After 1 second of interaction. there is 540 kg f signi5 cant change from the best estimate case.

of corium in contact with water which is considerably larger than the best estimate case. The local peak pressures for the shallow and deep pools were approximately 62 bars (l milliseconds '

after the explosion trigger, see Figure 14) and S5 bars (3 milliseconds after tne explosion '

trigger, see Figure 17), respectively. The time history of the local pressures at various axial loc.adons are shown in Figures 15 and 18. Table 7 summarizes tne local pressure impulse at l

these locations. Again, it should be noted that these are the local pressures, and the wall I pressure will be subsequently smaller due to pressure wave expansion. i These calculations have used an assumed triggering tinke of about ims Iwas tuceea second.

to be appropriate based on the integretation that the explosion would most likely be triggered l at or near the time of contact with the basemat floor or where the fuel debris oegan to cuench in their passage through the water pool. The former limit has been ooserved empirically in l

expriments cut to the postulated effect of the bottom surface allowing for water entrapment.

The second criterion of fuel cebris cuench is also empirically observed and is postulated to occur l For the conditions because a spontaneous film boiling collapse triggers the explosion locally.

of the present configuration, geometry and scenario.- these conoitions are satisfied at zoout 1 l second or shortly thereafter. If the trigger time is decreased, the void fraction in the pool anc the co-ium mass mixed with water will also d. crease. The reduction in void fraction tends toI increase thelocal pressures and thus the impulsive loads, because a smaller void fraction reouces the mixture compressibihty leading to the pressure rise. On the omer hanc. tne reducuan m participating corium mass tends to cecrease the impulse ioat, because there is less fuel available to thermally drive the explosion. Several validation calculations indicate inat tne impulse For initially rises due :o ine void fraction effect and then falls as the corium mass cecreases.

I ERl/NRC 93-203 Energy Research, Inc. _.

i

t

. }

I i,

example the maximum local pressure impulses for the 1 m pool cand a o m' cross-sectional area) are 43 kPa-sec texplosion trigger at 0.75 second), 50 kPa-sec texplosion ingger at 0.5 The maumum local pressure second), and 35 kPa.sec texplosion ingger at 0.25 seconal.

2 l impulses for the 3 m pool (and a 2 m cross-sectional area) are 31 LPa-see (explosion tngger at  ;

0.75 second). 53 kPa-see texplosion ingger m 0.5 second). and n0 kPa-sec texplosion ingger l

at 0.25 second). Therefore, considenng the stochastic nature of the Inggenng process ano the .

uncertainties associated with the timing of the explosion trigger (with notable influence on the

[

explosion impulse), the numerical values for the impulse loads at the pedestal wall should be l used with caution. The results that are presented in this report correspond to our best judgement for the timing of explosion trigger.

1 Simple independent parametric calculations [9] have also veri 6ed the results of the TEXAS-Il l

calculations, i.e., the peak explosion pressure was observed to' be a strong runction of the void i

fraction and showed good agreement with the TEXAS-il calculations, i r

Table 7 ABWR Ex-Vessel Local Pressure impulse (Conservative Estimate)-

t -

i Pressure impulse (kPa- Axial Distance Pressure impulse ,

Axial Distance I from the Bottom (kPa-sec) for the 3 m from the Bottom sec) for the 1 m Pool of the Pool, m Pool Depthd "  !

of the Pool, m , Depth" 0.52 l 18.4 0.12 l 13.6 11.5 0.92 I 18.5 l f 0.32 l 1.32 l.- 18.6 0.52 l 13.6 1.72 .l 18.5 f 0.72 h 19.5 1 2.10 l 13.7 l l l 0.92 l 18.9 - ll t

l. .-

l- l 1.12 l 14.4 .

7 l # A mixm; cross-r.ectional area of 6 m is used.

[ ,

M A mixm; cross-sectional area ni 2 m: is used For a 4 m: cross- ectinnal-een. me maumum incel pressure  !

f

' impulse is 2S kPa-see tot an n.ual dW.4nce ut 1.72 from the bottom ut the pond.

l l

i Enerzy Research. Inc. ERl/NRC 73-203

- _..._ __ _ _ *- b9' rh ._4 i

L

,, .- -w e a.,g y.-s . -- m -a-w-,-en- r ~a.g-.

i l

l I

I j

,- [ , _ _,_ _ _ _ _i_t=0.25 _ ys -+-

' t=0.50 s ---

': N l,:

t=0.75 s -e --

' - t=1.00 s -* --

,I *-

,f '. _

0.8 - ,

8 *

+: i .

.t .

i t -  !

0.6 -

.x. l:

= '

x a ,e :  :

a .

.:. i

l: i

=. l 0.4 + x l:

,u'

. - J I

4 G >

ED l b y ll y -

0.2 l ,

f *.l W .

l' l'

h '

\li v.

/ '.. $

. ~ . ..

's D. , y , i , i

, j -; e , ,

0 '

3 3.5 4 1.5 2 2.5 0 O.5 1 Distance from the Bonom of Pool (m)

Firure 13 Loca] void fraction for th: Conservative Estimate mixir.g :alculations () m pool cepth)

Energy Research,. lac. ._ _ _ __ . ERI/NRC 93-203

. O.

, i led)7 i i 6 ' '

t=1 ms 4~

t=2 ms ----

t=3 ms -e --

t=4 ms -*- -

t=5 ms +-- _

8e+06 -

-n

" 7 -

6e+06 -

E R

5. * ,

- ,o ,

u ,

_ i

_. , i

• r

.== u s e-i

,e o i. -

Ex 4e+06 - -

+ i o i

• = l  :

1

\n, :.:.

D '.

i lsi -

t '9'M

\ :

2e46 - x i . ,

.p ...x..- 1X-i.

1

!j,,$lhs_y.. o-

- \ ' '. i i

.i . .:,t 1

- n .

1 1 .,,o r.;

fi g v',

Q' _i..

4W+ . . . . __. - _ _ _ ._ ._

1.3

,, 3 a ,3 2._ - .

0 0.3 1 - _.

Distanee from the Bo::om of Pool (m) 4 i

n.

--:gure 1,. 3 oeal explos. ion pressure Ior tne Conservative estimate explosion ca:culations (l m pool deptn)

ERil.NRC 93-203 Energy Research. Inc.

30 i

i

9 9

I p

1e+07 i , , i i 4 4 4 x=0.12 m x=0.32 m ----

x=0.52 m -----

x=0.72 m x=0.92 m ----- ,

Se+06 - x=1.12 m --- -

.i

.o i 0

4 1 o  !,. i e

a--

u $. <

a u 6e46 - i .a g.

u 1

i w 4 0

m -u. ,. s-\

• i g s .

t p

G I 1  ;

~

-i *s i

-  !, , s t

=

o ,

i

._o ,.u, .

.s. -

t m , . ., s. s.  :

o -

--=. 4e46 - -

i

,'s g

< .I . t. s . -  !

'M .I f

\, g s

f  % ,

I \, . #

\. g..  %,

t.lal g

. .>L .~.,

s ,

r f

s

/

.I-;s' T' s gs,*

- l e

6w .l f

On. *s . ..,y ' ,

.a .

a .

~..,..........,. ,

,! ,! u.. . ., ,, .

-. . n . . ~%.. . .~= .s.-."n = ::,g:... s. :=:=:.= ~. _

\'o. i *

,/ **'- .7=:: u._..-.= i

.i i

0' 0 0.001 0.002 0.003.0.004w 0.005 0.006 0.007 0.008 0.009 0.01 r l

Alme(s) 1 1

1 Figure 15 Temporal development of the io:al explosion pressure for the Consenative .

estimate explos. ion calculations (1 m pool depth)

Energy Research, Inc. _ ERl/NRC 93-203 31

i

[

l 4 4 ,e 6'  : ---

1 ,

t=0.2ds 0 l t=0.5 s +--

..X-x- F, x,~x 0 75 s -

g i ts,1.p0 t= s -x-i

• g '

X -

0.8 - a; , l ,

D  :

. t i .

a l g X

.R. ., , .

D l  :

0.6 -  :. . ,

=  :

.2 e

x l .

.M ,. , s l l  : . .

-: l -l  : X-X

'5 l ;i  :

0.4 -

Et &  :

- .$. ,,o '.

a, m.

n .

Y >*:, i 1

s  !.

t i s '. l p ', i ,. .

0.2 -

l *.

x'

,1

, _s .D lQl

\ o

, s v

i

. ,. s i

s'I e  %

r !

X.y A ,+

. _ i - ., i . O, ., "si i e

0 4 5 0 1 2 3 Distance from the Bottom of Pool (m) i Figure 16 Local void fra:: ion for ine Conservative Estimate mixing :alculations (3 m pool deptn)

Energy Research. Inc. ERl/NRC 93-203

.b, . .

~ - +

i i le+07 i 6 t=1 ms -#e--

t=2 ms -*--

t=3 mS -O *

• t=4 ms -*--

GG '

t=5 ms + - -

8e+06 -: '-

. G.

.e

-C. .

W' 's D' n

s s '+ -

6e46 -

's, ' .

?._

=

a e s

\+ '. .

$ 3 '0

_- s.

- * +',

C -

1  %

\

j >

,_$^ e & -

4- 4e46 "5 hi'N s

< i .

w )

s l 0 '. N i

• s.

I . t u .

! O C.~ i i

1 -

2e+06

.x -X..."X-li%;i * '% [~X..X -,

,\

f p't 4.%

J d  ? \&-

/ 4 ss3 d

\'<~c_____________

0*~"~ 4 5 0 1 2 3 Distance from the Bonom of Pool (m)

I r . : cure :- Local explosion cressure for tne Consenative Esumate exciosion ::nculanons O m pool cepin) l t

Energy Research_,_._It. . ._ _ _ __ _ , _ERl/NRC 93-:03 23 l

i 1

4 l,

1 f

. l l

\ .

6 8

4. 4 4 i .

le47 . . . 6 x=0.52 m x=0.92 m ----

x = 1.32 m - - - - -

x=1.72 m l x=2.10 m - -- -

8e+06 -

r%

1. g i

l # \

i

> 0  %

i t s=% #

I

? o \, -

l ,

~

t o 6er06 -

1 t g s/J ' ,

4/5 i

\

U '

• ==* i g= *

. t

== , ,

• , . 1

== **

• 1 i 3 a

s-j

• 1 -

p ,

.=

4e+06 -

! 2' 'l i

x t

D ,

,a .i a1 e g e .

• t f eg '

l l# .4 e

• t i

g's

'.s

~ ' ~

e

_5

[ l,I[ ,,_

2:46 - ,

y

• .* tg*I

, h. . I

[

. of 4.

• %,~. .,'

lf .

. . ,.. . . i A._

-. % ' . .=.:. ~ ~; . ._,

...,:.,<.-~.  ;

.l ...'

. \ s~

• - a

.*f,/ l j

~, ,f

/ -'

M. . . . . . s. . ,

0 3 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01  ;

i ume (s) -' y t

rigure 2.

.g -

2 emporal c.evelopment of the 1o:21 explosion pressure for the Conservative Estimate explosion calculations (3 m pool cepin)-

ERl/NRC 93-:03 i Energy Research. Inc.

}^ .. _

_ _ _ . . ~ -~ - - -

h

• * - . - ~, - , ,.

I i

1

~

5. CONCLUDING RDIARKS The potential impacts of ex-vessel energetic FCis were imestigated based on a best estim and a consenative' sanation of the mitial and bouncary conditions in a General Ele,;tne conceptual ABWR con 6guration, using tre TEXAS-Il computer code. A number or parametnc sensitivity calculations were also performed to provide additional insights into the expectec r of impulse loacs resulting from steam explosions in the lower drvwell recion. $

in the best In the calculations reponed here, two corium pour scenarios were considered:

estimate" case the corium composition was assumed to be metallic, based on BWRSAR-type calculations, and in the consenative' case the corium composition was assumed to be oxidic, based on MAAP-type calculations. The corium pour rates for the 'best estimate and the

'conservadve cases were estimated to be 16.7 kg/s (1 penetration failure), and 540 kges (8 penetration failures), respectively.

Table 8 gives a summary of the results of the calculations performed for the best estimate case I and the consenative case, along with the sammary of the sensitivity results for the best estimate i

case.

The maximum local peak pressures for the best estimate and consen-atise cases were determined to be approximately 60 and 85 bars. respectively. However, the pressure at the drywell wall was estimated to be reduced approximately oy 1/r (where r is the distance to the pedestal wall and it equals 5.3 m) as the pressure wave expands. Thus. the maximum wall pressure loads were determined to be approximately 11 (best estimate), and 16 bars (consen estimate), respe:tively. These estimates correspond to maximum local pressure impulse loads of 13.6 (50 K melt superheat and I to 3 m deep water pool is referred.to as the best estimate) and 19.5 kPa-sec (50 K melt superheat and I to 3 m deep' water pool is referred to as tne consenative es-imate), respectively.

1 The melt supeieat was increased from 50 K [3] to 75 K [1] and the local pressure impuises ,

were calculatec bat showed a maximum increase of 2Tc-10ro for the 1 m and 3 m depth pools. '

respectively. Tne maximum local pressure impulse for 75 K melt m supern- (in the case of best '

estimate calculation) was calculated to be 14.9 kPa-see which vields a 2.S k?a-see impuise ioad at the wall. Tr.us, a 25 K increase in the melt superheat did not signi6cantly alter the results of the calculations.

A sensitivity caiculation was also performed for the best estimate case with a water pool depth i of 5 m. The maximum local pressure iapproximately 60 bars) predicted for this case was observed to be comparao'.e to the results of the calculations for the 3 m pool. However. occause i of te differences associated with the dynamics of the :nteracuan. the maximum impuise ;oad at me wall (4.: ipa-sea was approximately 60Fc higher man tnat of the 3 m oeep poo .  ;

- l ERl/NRC 03-203 Energy Researen. inc.

35 - _

l i

' ' ^ ~ ~ . . . , _,

i Table S Summary of ABWR Ev\ essel Local P essure impulse Calcula:.ons Maimum Pressure impulse at l Maumum Local Pressure impulse CJa-sec)

Ine Wall (kPa-se:) i l i Melt Water pool deptn m Me!t j Water pool dep:n. m  ;

Superheat superheat i 3 5 l 1 3 5 (g) _

(K)

Best Estimate 50 l 10.8 13.6 22.5 50 l2.0 l 2.0 - l 4.2 f 75 l 11.0 14.9 - 75 l2.1 l 2.s l -

i Conservative 50 19.5"' 18.6*- - 50 3.?" 3.5 *' - ,

Estimate ,

l

"' Based on a mixmg cross-secaonal area of 6 m .

• ' Based on a mixing cross-see::onal area of 2 m'.

An incependent assessment of the ABWR pedes al wall strue:.:ral strength is beyond 'he scope of the present study. The GE pedestal failure :imit estimate cased on a simpie elastoplastic  !

calculation for the ABWR has snown that the maximum pressure rise the peces:al can w:thstand l

~

is approximate.y 8.5 bars [4). It is further asserted [4] tha: the ABWR pedestal st ::ture is

' simdar, and at least as strong as that of the Grand Gulf pedestal; therefore. :r. stead of i performing ABWR-specific structural analyses. GE opted to use the results of the Grand Gulf structural analyses reported in se NUREG-1150 study [10]. It is stated in References [4] and

[7] that the smallest impulse load that will fail the pedestal wall is 25 kPa-see which could be used for comparison purposes only. This impulse load was cased on the ability of the Grand Gulf pedestal to withstand hydrogen detonation. and the containment response to dynamic loaos resulting from steam explosions was also assumed to be similar (see also Reference [10]). i The maximum impulse loads at se wall (2.6 kPa-sec for the rest estimate case and 3. ipa-see for the consenative estimate) calculated using the TEX AS-Il code were ocsened to 2e mucn ,

smalle- than 25 kPa-se; failure limit esubbsnec for Grand Gulf. i it is important to note that the underivina phenomena associated with the steam explosion issue are still not well understood and the current hypotheses remain technically controversial:

therefore, the present study was not intended to provide the SRC with a definitive quantitative l

estimate of the ex-vessel steam e :plosion energed:s for the ABWR co~"~ instea:. i: cas directed at providing a :alculational method and or framework for exploring tr.: possib!: range of impulsive loads based on our cut em undersanding of these comp!:x pnenemena.

ERl/NRC 93-203 Energy Research. Inc. _

l g

- . - - .