ML19344A345

Monthly Operating Rept for Dec 1970

ML19344A345
Person / Time
Site:	Midland
Issue date:	01/15/1971
From:	IDAHO NUCLEAR CORP.
To:
Shared Package
ML19344A344	List: ML19344A343 ML19344A345
References
NUDOCS 8008180409
Download: ML19344A345 (20)

Text

_.

=, m,

N l

w

.^*

t

+

-I

'i

~IDA110 NUCLEAR CORPORATION

~

. Nuclear Safety Program Division l-t-

l b

i I

i

'4 MONTHLY _ REPORT

.-December 1970 y

l The information contained ;in, this report is preliminary and ' subject to further evaluation.-

lgo o

?..

- )

F v-t TABLE OF CONTENTS-i.

v Page p

' A.-

LOFT Integral Experiments and Radiological Studies 1

" B.~ -. Loss-of-Coolant-Accident. Analysis.'.-...................

3 i.

C.

-Technica1' Assistance in Re' actor Safety Analysis.

6

- D.-

. Separate Effects. Testing 8.

4 s

<:V i-5 E

P l

i r

ar " #

1 t

l-f 4

I i

oc 1

"/g t

' )t p G

.4 4

,} '

' _b

___n_

c

~

1 December 1970 HIGHLIGHTS r

LOFT Integral Experiments and Radiological Studies: The initial phase of the LOFT experimental program plan has been completed. This plan will become part of the LOFT program base-line documentation, and will be published as the " LOFT Experimental Requirements Document" early in 1971.

Seven sections of the LOFT PSAR and six LOFT SDD's were reviewed during December.

Computer programs for fission-product release analyses based on the Idaho Nuclear Corporation steady-state release model and convectional-flow model have been written in FORTRAN IV.

Loss-of-Coolant Accident Analysis: The report Comparison of Predictions from the Reactor Primary System Decompression Code (RELAP3) with Decompression Data from the Semiscale Elowdown and Emergency Core Cooling (ECC) Project, IN-1444, was published.

Technical Assistance in Reactor Safety Analysis: A revision of the two-phase pressure-drop subroutine and two other, minor changes in CONTEMPT-PS have increased the accuracy with which the code predicts peak dry-well pressures in the Humboldt Bay tests.

Separate Effects Testing: Semiscale Tests 846 and 847 involving emergency core cooling (ECC) were performed with the ECC injected directly into the

. vessel inlet plenum.

Preliminary analysis for Test 846 indicates little or no core cooling by the emergency core coolant. Test 84S was performed late in December.

A time-dependent correlation for heat transfer directly to the coolant was-developed from BWR-FLECHT stainless steel test data.

Use of this correlation produced predictions that quite accurately represented the Zircaloy bundle (Test Zr-2) experimental results.

The final PWR-FLECHT Zircaloy bundle test (Test 9473) was conducted, completing the planned PWR-FLECHT experimental program.

Preliminary data inspection shows that Zircaloy cladding temperature reises are higher than those of stainless steel cladding for the same test conditions.

The rapid cooling exhibited by the previous Zircaloy bundle Test 8874 did not occur in this test.

4 111 i

Decemb:r 1970 A.

LOFT Integral Experiments and Radiological Studies l.

Highlights The initial phase of the LOPr experimental program plan has been completed. This plan will become part of the LOFT program base-line documentation, and will be published as the " LOFT Experimental Requirements Document" carly in 1971.

Seven sections of the LOFT PSAR and six of the LOFT Project SDD's were reviewed during December.

Computer programs for fission product release analyses based on the Idaho Nuclear Corporation steady-state release model and convectional-flow model have been written in FORTRAN IV.

Studies nearing completion at the Chemical Processing Plant show that the hydrolysis of molecular iodine, I, in dilute solutions and the subsequent 2

decomposition of HIO can be followed by using a specific ion electrode for iodide.

2.

Technical Activities Program Requirements and Experimental Plan for LOFT Integral Tests:

A summary description of the reference of the experimental program plan, which has been under development for the LOFT tests has been issucd for

, interval review. The reference program plan as well as detailed informa-tion in support of the plan is to be contained in the LOFT Experimental Requirements Document to be published early in 1971.

The test Program has been formulated to provide a direct, timely and economical means of obtaining LOCA information.

This program provides a demonstration of ECC effectiveness under the most severe accident conditions and at the same time a basia for assessing the adequacy of analytical models under the same, most demanding conditions.

It will therefore provide key information on the major issues of water reactor safety at the earliest possible time. The proposed test sequence provides for continuous development of the analytical models such that they will have a high probability of successfully predicting the most demanding test when it is conducted.

The test program consists of both nonnuclear and nuclear test series with simulated ruptures in the primary coolant system piping, j

The general. ordering of the tests proceeds from the nonnuclear tests to tests with increasing potential for high cladding temperatures and thus j

cladding annealing or distortion that could preclude fuel reusability.

.This approach is followed to: (1) result in maximum fuel reusability, thereby reducing program costs', (2) provide reasonable assurance that the experimental system will respond to LOCE conditions as planned, and (3) determine plant response and. evaluate key areas of the experiments prior to the more demanding Test Series III and IV which, respectively, will investigate principal LOCA-ECC safety issues for ECCS desigh ccnditions and ECCS margin conditions.

/

Tha rzf r:nca test 1, gram, which r:prcsonts cn cccc:

ant cf tccto to. b3 perf rmed in LOFT, la bas d en curr nt knowicdga cf wctcr rc ctor safety issues.

Information needs of AEC regulatory agencies and the nuclear industry relative to the safety of large PWR's have been factored into the program.

In addition, maximum practical versatility has been provided in the_ test assembly experimental capabilities to assure accom-modation of tests to provide safety information which is not now apparent, but could be required as a result of unexpected results from the early LOFT _ tests or by_ the changing needs of the nuclear industry for safety information.

During Decemb tr, Paven sections of the draf t of the LOFT Preliminary

-Safety Analysis Repert were reviewed and comments were transmitted to LOFT Project.

In addition, as?en of the LOFT Project SDD's were reviewed to insure the systems, as described, were consistent with the requirements of the Program Requirements Document, PRD-1B.

Included in this effort was an indepth review of the blowdown system essential for providing the necessary flexibility in test initiation conditions.

Analytical Models Describing Fission Product Behavior: Computer codes have been written in FORTRANIV to describe the behavior of fission products in the containment building under natural response conditions and the release of fission products to the fuel-cladding gap during steady-state reactor operation. The models are based on the Idaho Nuclear Corporation convectional-flow model(l> and the steady-state, release model(2), respectively.

The computer program for the behavior of fission products under natural response conditions in the containment building, FLOMOD, calculates the airborne half-life for molecular iodine and acrosols, the maximum velocity of the natural convection currents in the containment building, and the tempera-ture drop between the bulk air in the containment building and the wall of the containment building.

The code requires information en the geometry of the building and the steam condensation rate as a function time. The latter information is obtained from a computer code such as CONTEMPT.

The computer code, FPFM, which describes the release of noble gases from operating fuel pins can be used to calculate the fraction of the noble gases present in the fuel-cladding gap af ter normal operation.

The nobic gas release fraction obtained can then be used as the maximum release fraction 'for iodine-131. ~ais code requires information on the average linear heat rating for. 1 the pins in the core and the radius of the fuel pins. The equations used in both FLOMOD and FPFM produce results that are conservative for nuclear safety calculations.

-Fission Product Behavior Experiments: Studies to evaluate the use of an iodide specific-ion electrode to measure the kinetics of iodine hydrolysis as a function of pH, initial iodine concentrations, and tempera-ture are continuing. The results to date show that at a pH of 8 and at 250C the hydrolysis of iodine in a 10-5 molar 1 -solution and the subsequent 2

decomposition of HIO can be followed with the specific ion electrode.

Tests at other conditions are in progress. These tests are part of a program for determining the mechanisms leading to the formation of hypoiodous acid

.from iodine. released in a loss-of-coolant accident. The program is to be concluded during' January.

(1)

W. A.-Yuill and V. F. Baston, "A Model for' Fission-Product Deposition Under Natural Response'Cenditions in Containment Buildings," Nuclear Safety,10(6) (November-December 1969) pp 492-498.

L(2)

W. A. Yuill, V. F. Baston, and J. H. McFadden, " Release of Noble _ Cases From Operating Fuel Rods," Quarterly Technical _ Report Nuclear Safety Program Division, October l'- December 31, 1969, IN-1320. (August 1970).

2 4

~

s

, Decemb3r 1970 B.

Loss-of-Coolant Accident Analysis l.

Highlights The report, Comparison of Predictions From the Reactor Primary

~

System Decompression Code (RELAP3) with Decompression Data From the Semiscale Blowdown and Emergency Core Coolinn (ECC) Proicct, IN-1444, was published.

The main exhaust stack and the process building at the Chemical Processing Plant were modeled and analyzed for their structural response to the influence of the El Centro carthquake.

-2.

Technical Activities llydraulic Model Development: Preliminary RELAP3 predictions of Semiscale-ECC Test 846 were made. The. initial conditions and a description of this test are presented in Section D Separate Ef fects Testing, of this report.

The comparison of RELAP3 results with the semiscale results indicates that-RELAP3 is not providing adequate predictions of blowdown phenomena during ECC inj ection.

Figure 1, an example of experimental and analytical data, shows a comparison of analytical and experimental pressure data for Test 846 and, as shown, the RELAP3 predicted' pressure behavior deviates from the experimental data fter ECC injection begins.

RELAP3 predicts a shorter blowdown than actually occurred.

Figure 1 is an example of experimental and analytical data which shows a comparison of the analytical and experimental pressure data from semiscale Test 846. The figure shows the agreement between the RELAP3 prediction of pressure and the experimental data is relatively poor in the period which follows injection of ECC fluid (beyond about 10 seconds for Figure 1).

RELAP3 predicts a shorter blow-down time than was observed in the experiment.' The difference between the RELAP3 predictions and the~ experimental data during ECC injection is attributed to inadequate calculation of the mass flux of low quality fluid at the break and nonequilibrium fluid conditions that cannot be accounted for by RELAP3.

Investigations are underway to determine means of correcting RELAP3 to -provide better predictions of semiscale blowdown tests with ECC injection.

Core Thermal >bdel Development: The report describing the computer program THETA 1-B was submitted to Technical Editing for publication.

Development continued on the MOXY code to predict BWR-FLECHT heat transfer data' and reactor rod bundle thermal behavior. Provision was made to include a variable cladding oxide thickness and th'e fraction of the metal-water reaction taking place as initial conditions. Modification is also being continued to divide the canister into_ several separate nodes to more closely represent the BWR-FLECHT experiments.

3

i 6

0 9

~

22 3

0 7

8 1

i B

C N

I 07 6

48 ts s

i i

0 i

6 e

la

)

c s

i m

c e

e 0

S s

i 5

(

g n

e i

r r

u u

t D

a p

t u

r a

0 R o

i D

4 v

i r

a 6

e h

4 t

e f

B 8

a e

t r

se 0 e u

n s

i T

3 r

s i

e

~

T r

~

P

~

l n

e o

s it s

a d

0 e

i e

2 V

t i

t n

c s

1 i

I d

s e/

s n

r g

o i

P f

i

\\

F t

c.

,,i l iI!1 e

\\i lI 0

j

' l ;

iI l 1

n l

I a

s\\

3 u

\\'

P C

tc C

A

\\

A E

L E

I l

iI't R

t O

0 0

0 0

0 0

0 0

O-1 0

0 0

0 0

0 0

8 4

0 6

2 8

4 2

2 2

1 1

3g E ci t&n s. ' osU2 i

i

1 December 1970 Work continues on the development of a correlation for the prediction of heat transfer coef ficients during flooding. The correlation mentioned in the last monthly report (3) is beingmodified to predict coefficients for a full length PWR core. A nonlinear least squares computer program is being used to evaluate the form of the correlation best suited to predicting tae PWR-FLECHT sata. The correlation will be based on the

~

FWR-FLECHT test parametern.

s.

Experimental data related to critical heat flux has been accumulated in an Idaho Nuclear data library to assist in developing and determining the adcquacy of corre1ations for predicting critical heat flux. About 250 CHF data points are being prepared for addition to the CHF data library.

These data will bring the total number of points in the library to about 3700.

t

. Structural Model Development: The main exhaust stack and the process build. g at the Chemical Processing Plant (CPP) were modeled and analyzed for their structural response to the El Centro earthqucke. The main exhaust stack is a 250-foot reinforced concrete structure having a diameter varying from 25 feet at the bottom to 12 feet at the top.

The process building is a massive reinforced concrete building 240 ft by 100 ft and 43 ft high.

The structure is highly compartmented with walls up to 6 feet in thickness.

A consistent mass matrix generator has been added to STRAF-MS.

Calculations were made by using STRAP-MS with the consistent mass matrix generator and a model of a four-loop Pun system having 180 degrees of freedom.

These calculations ware compared to STRAP-D culculations for the same problem.

The fifteen lowest frequencies calculated by the two techniques agree within 6%.

1 l

4 (3) Nuclear Safety _ Program Monthly Report - November (Hai-436-70)

December 14, 1970.

5

.P, 0.'.

DIcemb r 1970 1

Technical Assistan'e in Reactor Safety Analysis c

~

C..

1.~. : Highlights _.

A resision of :the -two-phase pressure-drop-subroutine and the correction

~

' of two minor programming' errors.in CONTDiPT-PS have increased the accuracy

. with which 'the code pred%ts peak' dry-well pressures ir

. ::umboldt Bay tests.

-2.

. Technical Activities Containment Response Analys'is : A review of the capability'of the C0!TIDIPT-PS. computer code to handle long-term (hours to days) containment calculations resulted in the.following conclurions:

- (l) The present model, in which water.is boile, off from the

. primary system f rom decay heat and metal-water reaction, does not account for continuous ECC injection and possibic overflow.

-(2) Containment spray systems of several types can be handled, but not concurrently.

(3) The' calculation time required for running long-term problems is.not excessive, but could be shortened.

A schedule for modifying CONTDIPT-PS to handle these long-term problems

~

has been set up.

Studies have been carried out to compare calculations from the most recent' version of the CONTEMPT-PS code with calculations from an earlier ver'sion..-The-differences between the two versions are as follows: A revised two-phase pressure drop subroutine that utilizes a modification of the'Baroczy-two-phase frictional pressure drop correlation was incor-porated into the ' latest version of CONTDiPT-PS, as previously described (

Also,' two coding crrors existed in the older version of the code.

Equation (25) in tl$e CUTDIPT-PS document (S), namely xp)i*("x~"q)

• Uf (O

O

+C t

+

t x

p

.L p '

'-AP I- *-

Mf x_

.xp.

TPF g -

2

.x 2g

)

was' coded.in the program as ft (O9f

[C

+C-

~

~

r

-:1.43 6P L~

x XP AZ

-l p... p.

(u

-u-)

x xp TPF g

X j

g X

XP c;

e 1

~(4). ; Nuclear Safety Division Monthly Report - September (llai-384 -70).

October 14;~1970.

(5)- C. F. :Carmichael and S. A. Marko,- CONTD!PT-PS - A -Digital Computer Code.for Predicting the Pressure-Temperature Historv Ufthin n

Pressure-Suppression Containment Vessel in Resconso en n T n"-n f-

~

? Coolant Accident,-IDefl7252,(April 1969).

e

.6 2

~

~

v m

~

'pecember 1970 In the: olderiversion.of the code, the two-phase pressure drop (dPTPF) iwas. multiplied by 1.43 to account for scatter in the_two-phase exper mental i

1 data and to-calculate a_ design pressure'rather than a most likely pressure.

The'. multiplier _(1.43) has'bcon changed to 1.0 in the most recent version

~

of CONTEMPT-PS. The sign

change in:the-density-averaging term of Equation (25) wasLmade:to correct a, coding error. The change in results due to l=

this ' change is not :likely to bc. large because of -the small height changes encountered _ in most pressure-suppression systeem.-

The.cffect-on peak dry-well pressure caused by these corrections' and

~

the modification ~ of:the Baroczy two-phase pressure drop correlation was studied by; comparing resultn of computer runs made with the new version -

of the code with results_from the previous. version.

Humboldt' Bay Tests l'

15,16,17, and 33 were used for this comparison because they cover a wide l

- range of. vent - flow' conditions. '

'The results of the study are presented in Table II.

Each of. the.

. computer runs was based on the assumptions of no heat transfer,' maximum water carryover (PC0=1) and early air removal from the drywell.

These-assumptions are all conservative.(that is, they contribute to an over-l-

prediction of. dry well pressure). The results of this study do not alter the previous" conclusion'that CONTEMPT-PS is conservative (overpredicts peak dry well pressures) with respect to availabic test data (Humboldt Bay and Bodega Bay) when the-computer runs are made using the heat transfer,-

l tair carryover, and-water-carryover assumptions centioned.

The new version of CONTEMPT-PS is recommended for use in the analysis of pressure-suppression type containments.

I.

TABLE II L

COMPARISON OF THE NEWEST. VERSION OF CONTEMPT-PS (CONPS)

WITH THE PREVIOUS VERSION Humboldt Bay Peak Dry Well Pressure (psia)

Test Test Old CONPS New CONPS i

15:

47.7 55.0 51.2 1

-16.

78.7 87.6 81.1:

i 17-104.7

125.4 116.8-33 142.7' 158.1; 148.3 1

A' literature, search.was initiated to-locate-experimental-data'

.l describing two-phascisteam-water.and. air-water' flow.

These data, if j

f i

i ound, will2b'c 'used in further evaluation-of the-CONTEMPT-PS homogeneous.

~

Event flow ~model-and for; future.cvaluation:of the slip flow vent model.

- The Dukler. two-ph'ase flow data bank.has bcen ordered.

j 1

A 7:

j w

^

~

i m

(c-

.s

y

'g'

e. -

Decemb.er 1970 D.

Separate Effects Testing.

l.

Highlights-l 1Semiscale Tests -846-and 847 involving emergency core cooling (ECC)'

were performed with the' ECC injceted directly into the vessel inlet plenum.

Preliminary analysis for Test 846 indicates little or no core cooling by

.the emergencyLcoolant.- Test 847 was performed late in December with test conditions similar to those of Test 846. However, for Test 847 orifices were installed in the -loop ;to more closely typify pressure drops in a

~

large PUR.

'A. time-dependent correlation'for heat transfer.directly to the coolant was developed from BWR-FLECHT' stainless steel test data.

Use of.this.correla-tion produced predictions that quite accurately represented the Zircaloy bundle (Test Zr-2) experimental _ results.

The final PRR-FLECHT'Zircaloy bundle test (Test 9573) was conducted,

. completing the planned PWR-FLECHT experimental program.

Preliminary data

inspection'shows that-_Zircaloy cladding temperature rises are higher than

'those of stainless steel cladding for the same test conditions. The rapid

cooling exhibited by the. previous Zircaloy bundle Test 8874 did not occur in.this test.

-2.

Technical ~ Activities Sing 1c-Loop Semiscale: The second semiscale test involving emergency core cooling, Test 846, was performed with a high inlet break.

The break area-to-system volume _ ratio was 0.0007 ft-1 The test was conducted from

. initial conditions of 2290 pgig system pressure, 220 gpm primary fluid

. flow rate, and 5930F-and 555 F fluid temperatures 'for the hot and. cold legs, respectively. The ECC accumulator fluid was at 660 psig and 155oF; ECC_ fluid was injected directly into the inlet plenum.

Core _ power was shut off 7 seconds after rupture.

Decompression was essentially complete 55 seconds after system ~ rupture.

Figure 2 presents the' vessel. inlet plenum-pressure and the pressure

. differential-between the1 ECC accumulator and ' the inlet ' plenum for Test

-846.

ECC injection started 9.5 ' seconds af ter ' rupture-when the inlet plenum

. pressure:had: decreased to-650 psig. The pressure' differential between

' the = accumulator and ' inlet plenumf reached a. maximum of 210_ psi 22.5 seconds

.after: rupture and dropped to about'170_ psi at 39 seconds when ECC fluid

~

injection was complete.

- The:ECC. flow' rate data for Test 846 are_ presented-in Figure 3.

The 3

2.6a f t fof ECC fluid- (sufficient.to cover the core) was injected.

The-1 3

average ECC: flow. rate-for~the period 22.5 to 39: seconds was 45 gpm (0.1 ft /sec).

__ -The path lfollowed by.the ECC. fluid subsequent to injection into the

inlet / plenum is' reflected in; fluid density data shown in Figure 4.

By 9 seconds-~afterDrupture; densitiesiin the inlet nozzle', inlet plenum, and

~

' core had ' decreased to' about ?l.'5 'lb/ft3 Immediat'ely'after.the start of 8'

v

~^

Aw

4

'O'

'l 2400 g

g g

g g

g.

g g

,3 1-l 350-

. ~{

ECC Starts

- 300 I A!! ECC Injected I

2000 e.

-l 250 E

-1 I

~

I I

E l

- 200 :

}

l' 2

=

01600-l I

- B.

l 1

- 150 3 o-O

-l I

2 g

y ct l

l 100 0 5

e I

E1200 i

.l 1

E 50 e

I l

0 2

{ 800 O>

P A00 0

I I

I I

I I

~

-5 0

5

-10 15 20 25 30 35

~40 45 50 55

~ 60 Time of tes Rupture (sec) ise-a.47329 Fig. 2 Vessel Inlet Plenum Pressure and Pressure Differential Between.ECC Accumulator and Vessel Inlet Plenum During Semisca'.c Test 846 10 0 1

I i

'l-1 I

I i

I i

l l

j "e.

l 1

~*

@ 50 - '

$c.

t t

1 i

i I

I I

I I

I I

I 4!

I 1

I I

O

--5

. O f_

5'

'f0 IS

.20 25 30 35 40~

45-50 55 60 i

Time 'of ter Rupture. ( see) i INC-A-l?328

Fig. 3 ECC Flow Rate for Semiscale Test 446 b

9-t 2,sp

+

1 v

60-3 5

i i

i i

I.

I g'\\

I.l I

50 l l!

.I v

i

~

l I

j

'gI I

l 40 l

I l

E l

8-I i

I il l

.s B

l I l i

j l l I

~

I f30 1

l i

j o

(0-c-h I

l f

i Inlet Nozzle i

l 20

(

(D-14)

I l

I l

I-

\\

l I

I l

/

i io f

I i

' Inlet Plenum!

[.(D-I-1)

/

'l 3

/

t I

-#~

I I(

I '.

~

0

-10 0

10 20 30 40 50 60 70 80 Time offer Rupture (sec)

INC B-17330 Fig.4FluidDensityDuhing'SemisedleTest846 l

-l

.l

~

' December 1970

- ECC inj ection, the density at the inlet nozzle began gradually increasing while-the density in'the inlet plenum decreased and remained at about 1 lb/ft3 Note should be taken of the fact that the core fluid density exhibited only a small increase at the start of ECC injection and decreae.u to essentially zero for the remainder of the decompression, evidence that no significant amount of ECC fluid reached the core.

Shallar behavior was noted in the first ECC test reported previously(3).

of ECC fluid The density of the fluid in the inlet plenum began a sharp increase at about 24 seconds after rupture. By this time, the accomulator had injected about 50% of the emergency core coolant to the inlet plenum.

Apparently a gradual buildup of emergency core coolant in the inlet plenum occurred with part of the fluid being swept from the vessel, through the inlet nozzle, and out the break. Fluctuations in the temperature indicated by thermocouples located in the lower plenum indicate the presence of boiling fluid (see Figure 5).

Figure 5 also shows pin cladding temperature histories at representa-tive locations and elevations in the core.

DNB occurred first (3.5 seconds af ter rupture) at the top of the heated length on a pin (Pin 13) located next to the core flow skirt.

DNB occurred at all other locations at 5.0 0.5 seconds after rupture.

Pin cladding temperatures increased at about 150 F/sec until core power was shut off.

Shortly after ECC injection was completed, a momentary cooling was recorded at all core locations and wae reficcted by decreases in the inlet plenum fluid temperature with subsequent recovery to temperatures near those existing prior to the event. The cooling phecomenon is believed caused by the passage through the core of cool (about 500F) gas from the accumulator. The gas possibly also transported some entrained ECC fluid.

The test conditions of Test 846 were duplicated in Test 847 performed in late December. The significant difference between the two tests is that.for Test 847 the system was orificed at appropriate locations to more closely typify the pressure drops in a large PWR.

The results of Tests 846 and 847 are to be compared to determine whether the system behavior, particularly ECC behavior, is influenced by increased loop pressure drops.

Fabrication of vessel components and testing of prototype heater rcd: for the modified single-loop (66-inch core) system haye been initiated.

Two-Loop Semiscale: A small high-pressure system for testing the heater rods to be (med in the two-loop semiscale core has been completed.

Prototype heater rods (12 f t heated ler.gth and 23 f t overall length) have been received from one potential supplier; qualification tests will commence early in January 1971.

BUR-FLECHT: Data from stainless steel Bundle 2.were used to develop a time dependent correlation for 1 eat transfer directly to the coolant at the bundle midpl.te during top spray cooling of a BWR fuel bundle. The

' coefficient defining this component of the heat transfer was obtained by the following steps in the data analysis process:

11 W

\\

~

l100 g

g g _

l g

s l

.1000 (g

Top of Heated Length of i

Outside Pin (TM-13-T)

Center of Heated Length of 900

{

Central Pin (TM-50-0)

I Bottom of Heated Length of I.

Center Pin (TM-61-8) l 800

[

i 1 -+--

A-ff l

t 700 f,,

=

l T

E I

l

{ 600 t

J j

e

'%4 E

l

's i

E500

\\

- 600 5-A N_

s y

400 J

- 500 b

\\

\\g' Y

1 o

y 300 b

E

,j 400 E

t..

\\

7""

g

~

\\g Lower Vessel (TF t-7 A \\'%

\\.

Y Plenum (TF-O-2) y Vessel Outlet

.?

Inlet Pienum T F-1-1

'It.

- 300 __

200

.w, 8

p h

i

' -10 0 200

-10 0

10 20 30 40 50 60 Time af ter - Rupture (sec)

INC - D - 17 331 Fig. 5-Typical Pin Cladding. Temperatures and Vessel Fluid Temperatures --

' Test 846 12 w

F

-3 i

o 1

v...

, Dec:mbar -1970 (1)! The' net heatfflux ifrom the; surface of a rod 'was calcu' lated by the DATAR: code by ; solving the inverse conduction equation.. This.

heat flux was converted to a total coefficient based on the icoolantisaturation temperature;'that is, 9neti" total ( rod' surface saturation}

. (1)

~

where; q = heat-flux, h = heat transfer. coefficient, T = temperature..

-(2) A radiative heat flux was calculated by the RADHT code by using temperatures on all' rods as input and assuming a transparent medium. ~This. radiative heat flux was converted to a radiation coefficient base'd on the coolant saturation-temperaeure; that is,-

(}

9 rad " " rad ( rod surface saturation *

~

(3) The difference 'between the' two coefficients thus obtained yields a third coefficient that represents the portion of the heat flux directly tol the coolant by means of convection, radiation to~ liquid droplets and vapor, and any conduction which might be occurring; that is',-

h

=h

(

h total rad coolant

• The hcoolant-term was found to correlate from run to.run when the results were normalized to a common reference time for canister quenching.

Thus, quench timeLof the canister is an important parameter.

This correlation,:INC~#1, has been used in'several versions of M0XY to-calculn* 'the respons_e of stainless-steel-and Zircaloy-clad heater

. bundles.

'ement v.ch the experimental data from the stainless steel-clad bundle tests with initial ~ temperatures from 800 to 1800*F is good.

~

The BWR form of.the MOXY code with the correlation, INC #1, was' modified slightly to, represent a Zircaloy clad. heater bundle with the. option thatL certain rod powers were set; equal to'zero at the times of actual rod electrical =

failures during:theitest. Figures 6 through. 9 show the calculated tr.mperature c response'and the ~ test data for the~ Zr-2 bundle. test.. As can be seen, the

- form and; magnitude of = the calculated results give confidence'in the capabilities.

of M0XY usingEthe INC #1~ correlation'for predicting the ECC phase of.s LOCA.

13 b

y

+1 ao

,4 e

yy p

e 4y -

e qr a

4-

I

~

o-i l

r. ?

. E' 2200 l

l l

6 I

i t

l

[

MOXY" t

I Q

DWR-FLECHT Dote I

2000 00 O O 5.

O o~n-n O

0 g

i

} 1800

>E I

0 OO 0

O 1600 i

I I

I I

I f

14 0 o l

_l 0

2 4

6 8

10 12 14 l

Time ( min )

inc.a.irsae l

i i

e t

Fig. 6 Comparison of Calculation with Data from Bundle Zr-2 Test, Rod 9, 6-ft Elevation i

2000 OO o

O O#

g O

1800 O

O 6

j O

C i

[isoo o

l E3 0

MOXY O

O BWR-FLECHT Octa i

1400 i

~

I I

I I

I 1200 O

2-

.4 6

6 10 12 14 Time (min) inc.4.orses

}

i Fig. 7 Coreparison of Calculation with Data from Bundle Zr-2 Test; Rod 32, 5.5-ft Elevation j

i' i

14 i

i 1

~

. e d

'i

~ - -

3 i

1 e.

I I

2000 g

g g

g g

g WOXY S

18 0 o O

BWR-FLECHT Defe O

~

O w

t.'

O I

O p O-O.,,,,,0 o-O 31600 0

a

/0 0

j O

0 O

-Oo 5

0 1400 l Rod Electrical Facure I

e I

l 1

I I

'l O

2 4

G 8

I0 12 14 Time ( rnin )

$NC.A l?323 Fig. 8 Comparison of Calculation with Data from Bundle tr-2 Test; Rod 30, 6-ft Elevation g

g g

g

[

t

'2200 m

~

2000

,c WO.o O

~

O OO O

-0 O E 1800 O

e E

E O

MOXY 1600 BWR-FLECHT Deto i

I i

1 I

I O

2 4

6 8

10 12 14 Time (min).

INC=A=17326

- Fig. 9 Comparison of Calculation with Data fro:2 Bundle Zr-2 Test; Rod 10, 6-ft Elevation 4

e

/

15

'o

l_.

d:

F Decs.mber 1970-

~ _.

PWit-l'LECllT: The final PWR-FLECllT Zircaloy bundle test, Test 9573, was conducted with nominallinitial cladding temperatures of 2000 F and ja flooding' rate of 1 in./sec.

Cladding' temperatures were predicted'to reach 24000F after about-30 seconds at which time rod failures were fexpected to-occur.. Rod failures occurred starting-at'18.2 seconds; fif teen rods had failed' by,30. seconds; all but five of the 42 heated rods had failed when power. was shut. off. at 58 seconds.

Cladding temperatures continued to' rise until the time of_ power termination.

The peak recorded cladding temperature was.25750F 'at power : termination although the rod

-recording this temperv.ure failed:at 20 seconds.

Despite gross rod failure, total bundle power was maintained due to electrical arcing.

l The external Pt-13% Rh - thermocouples. failed at about 2200 F.

This

~

-fact, plus the limitations of Chromel-Alumel thermocouples, put the validity of any' data above about 24500F in question.

Furthermore, because

.of the rod failures,- data beyond 18 ' seconds should be used with' caution.

' The cladding temperature behavior after flooding for a central rod

.(Rod.3D)~in Test 9573 is presented in Figure 10 along with results from a stainless steel bundle test of essentially identical conditions (Test

~ 6553). arid' a pretest prediction of the"Zircaloy bundle test.

The tempera-ture rises for the Zircaloy-cladding are higher than for stainless steel.

The rapid cladding temperature drops exhibited in the previous Zircaloy bundle test (3,4,6-8) did not occur in this test.

LThe prediction.for'Zircaloy Test 9573 was made by using heat transfer coefficients obtained from the. stainless-steel test and assuming metal-water reaction energy to be 0.75 times the Baker-Just(9) prediction. As

.is apparent in Figure 10, a factor;of 0.75 appears to be nonconservative for predicting the experinental results.-

Test 9573 was the' final test in the planned PWR-FLECHT experimental program. ' Analysis of-the test data will continue.

[..

t

.(6) -Nuclear-Safety Program Division Monthly Report - July (Hai-325-70)

August.14, 1970..

(7) Nucle'ar Safe'ty Program Division Monthly Report - August (Hai-347-70),

~

-September. 15,.1970.

.(8)' Nuclear, Safety Program Division Monthly Report - October '(Hai-418-70)

November 16, 1970..

f(9) Louis Bakerf'Jr.. and L5uis C. Just,. Studies of Metal-Water Reactions at Hinh Temneratures,.ANL-6548 (May 1962).

26

[ e+

~

. (,_'~'

.6 e

t.

2500

.g o

Zircoloy Test 9573

/

/

/

/

I

. //

g-2400 7,

/

/

Test 9573 1

/

Rod Failed 7l

,/

f I

N

/

Zl.coloy Prediction 1

/

/

2300 f

/

/

/

/

p' '

Y

/

/

I

/~

k 2200

/

.p

/

Steinfess Steel 'est 6553 5'

/

0 l

/

/

/

/

/

2l00 f

/

/

/-

/

,If

/

/

2000 f

'8900 I

I I

O 10

-20 30 40 Time offer Flooding (sec)

NC B.t7327

. Fig. 10 PUR-FLECllT Cladding Thermal Response for Control Rod (Rod 3D) 17 h

y

,e-

+,,.

.,---r