ORNL-5143.txt

ORNL5143

ORNT.-51L3
Dist. Category UC-T64

Contract No. W-TL05-eng-26

CI-IEMICAL TECHNOLOGY DIVISION

MEASUREMENT OF MASS TRANSFER COEFFICIENTS IN A MECHANTCALLY
AGTITATED, NONDISPERSING CONTACTOR OPERATING WITH A

MOLTEN MTXTURE OF LiF-BeFB-‘]th_ ANND MOLTEN BISMUTH .

C. H. Brown, Jr,
J. R. Hightower, Jr.
J. A, Klein

Date Published: November 1976 ,.

OAK RIDGE NATTONAL LABORATCRY
Oak Ridge, Tennessee 37830
operated by
UNION CARBIDE CORPORATION
for the :
ENERGY ERESEARCH AND DEVELOPMENT ADMINISTRATTION

REPRODUCED BY: m

U.3. Department of Commerce
National Technical Information Service
Springfield, \firginia 22161
CONTENTS
«ABSTRACT - . . [ ] - * . - - & .. - . - - - - - » * *
1. INTRODUCTION. . v ¢ & v o o 2 o « « = . -
2. EXPERIMENTAT. EQUIPMENT. . . « + o« + o « « o
2.1 Flow Diagram of Saglt-Bismuth System. . . .
2.2 Contactor Vessel . . . . . . .
0.3 Feed and Catch Tanks for Salt and Bismuth.
2.4 Treatment Vessel for Salt and Metal. .
2.5 5zZmPlers o v v v e e s 4 s e e e e . .
2.6 Freeze Valves and Lines. . . . . . .
2.7 Instrumentation and Control., . . .
2.8 Gas Purification and Supply Systems. . . .
3. EXPERIMENTAL PROCEDURES . & v « v & o .

3.1 Reductant Addition . . « « . .
3.2 Tracer Irradiation and Addition.
3.3 Run Procedld¥@. o v « ¢ o o+ o @
3.L Trestment with Hydrogen-Hydrogen
3.5 Sample Preparation and Analysis.

L, EXPERIMENTAL RESULTS. . « + « « & « .
5. INTERPRETATION OF RESULTS . . . . . .
6. CONCLUSIONS AND RECOMMENDATIONS . . .
T. ACKNOWLEDGMENT. . - « . + « . « .

8., REFERENCES. . . + + o« v o o o « « « o
APPENDIX A. v v v v « o o o o o o = « 4

APPENDIX B, . & ¢« v v « o« 4 =« o o o = « =

APPENDIX C - . - - L] - - . " - - » - . - *
APPENDIX D. & & ¢ v v v v o o o v o o o =
APPENDTIX E. & v v 4 v v o« ¢« o o o o o o &

=N

o o M

10
12

1k

ik

33

35
bk
MBASUREMENT OF MASS TRANSFER COEFFICTENTS TN A MRECHANTCAITY
- AGITATED, NONDISPERSING CONTACTOR OPRRATING WITH A
MOLTEN MTXTURE OF IiF-BeF,-ThF, AND MOLTEN BISMUTH

C. H. Brown, Jr.
J. R. Hightower, dJr.
J. A, Kiein

ABSTRACT

A mechanically agitated, nondispersing contactor in
which molten fluoride salt and molten bismuth phases were
contacted has been built and operated. The mass traansfer
performance of the contactor was evalusted over a range of
agitator speeds under conditions in which the major resist-
ance to mass transfer was in the salt phase. The measured
mass transfer rates were compared with rates predicted by
literature correlations, The equipment necessary to contain
the salt and bismuth at ~ A00°C is described along with the
complete set of experimental data obtained during operation.

1. INTRODUCTION

A molten-salt breeder reactor (MSBR) will be fueled with 2 molten
fluoride mixture that will circulate continuously through the blanket
and core regions of the reactor and through the primary'heat exchanger.
Methods are being developed for use in a close-coupled processing facility
for removing fission products, corrosion products, and fissile malerials

from the molten fluoride mixture.

The proposed MSBR processing plant is based on fluorinaticn to
remove urénium, reductive extraction to remove protactinium, and the
metal transfer process to remove the rare~earth fission products. The
type of two-phase contactor being considered for the iatter two steps in
the processing plant is a nondispersing, mechanically agitated contactor
in which & molten-salt phase and molten-bismuth phase are contacted to

effect the desired separation.
A facility was instelled for measuring mass transfer rates across a
salt-metal interface in a mechanically agitated, nondispersing contactor
using a molten mixture of LiF-BeF,~ThF) (72-16-12 mole %) as the light
phase, and molten bismuth as the heavy phase., Mass transfer rates for
237U and 2! 7r tracers were measured at nine different agitator speeds.
The purpose of the experiments was 1o provide measurements of mass trans-
fer coefficients in a fluoride salt-bismuth system with which existing
correlations could be compared, and to provide data for developing new
correlations for mass transfer coefficients which would allow large-

scale contactors to be designed.

Included in this report is a complete description of the experimental
equipment, operating procedures, experimental data, and interpretation of
the results.

2. EXPERIMENTAL EQUIPMENT

Mass transfer rates between molten salt and bismuth in the mechani-
cally agitated contactor were measured in steady-state experiments in
which salt and bismuth streams flowed through the contactor. Concentra-
tions of components which transferred between phases were measured in
inlet and effluent streams. The equipment used to make these measure-
ments consisted of the contactor véssel; feed and catch tanks for salt
and bismuth; a vessel for purification of the salt and bismuth inventory:
provisions for withdrawing samples of each phase from various locations;
Treeze valves for salt and bismuth flow contrel; instrumentation for
temperature, pressure, and gas flow measurement and control; and gas
supplies and purification systems. A description of the equipment
follows.,

2.1 TFlow Diagram of Salt-Bismuth System

A flow diagram of the system for flowing salt and bismuth streams
through the contactor is shown in Fig. 1. Salt and bismuth were metered
from the salt feed tank and the bismuth feed tank (vessels T-1 and T-3,
respectively) by controlled pressurization of these tanks. The salt and
*POTTBISUT J090B3U0D PalRLIIB ATTBOTUBYOSW I
UG TM £3TTTORM UYSNOAUGMOTL YINWSTE-1TBY 9UT JO WeITeTp Mot T *STd

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TIASSIA LNIWLVEYL MNVL NOILLOZTTI02  MNVYL NOILLIOITI0D YOLOVLNOD
1AW ONV 1VS ANV (334 1vS  dNV a33d VLAW ..
| B ! f!
| 5L “ mn#“?» |} LR
| t
I, Iy ! “ ” ““
i ] ! n | ! i -“
by b ] i
! oy R
i i — —
i i L
v-nd g f1
e A |
I i [ s
i - 1 _
flll%dlnlill.w_ Y L- D | | = mnu,ul
Qens — L-oT-go-mEE |
| i
S
WAL
4 Hinwsig”  H390YS
1Vs

WP69-bL 9MA “INYO
bismuth flowed to the contactor vessel; each phase entered below the
surface of the contactor inventory of that phase and left the contactor
through an effluent line at the salt-bismuth interface elevation. The
interface thus was continuously renewed and mass-transfer inhibiting
films were removed. The combined effluent stream was separated, and
each stream flowed through a flowing stream sampler and then to the
salt and bismuth catch tanks (vessels T-2 and T-4, respectively). The
feed and catch tanks for each phase were concentric tanks to conserve
space in the hood (see Sect. 2.3)., The salt and bismuth inventory
could be sent to a graphite-lined treatment vessel (vessel T-5) for

periodic treatment with H_-HF mixtures for removal of impurities and

2
adjustment of distribution coefficients.

2.2 Contactor Vessel

A diagram of the contactor 1s shown in Fig. 2. The contactor was
& 6-in. (152-mm)-diam low-carbon steel vessel conteining four l-in.
(25-mm)-wide vertical baffles. The agitator consisted of two 2-7/8-in.
(73-mm)-diam turbines with four 3/L-in. (19-mm)-wide straight blades.
A 3/k-in. (19-rm)-diam overflow at the interface allowed the removal of
interfacial films with the salt and metal effluent streams. Salt and
bismuth were fed to the contactor below the surface of the respective

phase.

2.3 Feed and Catch Tanks for Salt and Bismuth

The duplex feéd and catch tanks for salt and bismuth were identical
in construction. The feed tank, an inner cylinder of 8-in. sched 30
pipe, was designed to operate at pressures up to 50 psig (345 kPa) at
600°C, Both the inner feed tank and the outer catch tank had a capacity
of about 20 liters of fluid; however, only about 15 liters of salt and

15 liters of bismuth were used.

The top of each feed tank contained seven ports: (1) an inlet
port (1/2-in. pipe with a fitting for 3/8-in. tubing), which did not
extend into the tank; (2) an outlet line (1/2-in. pipe with a fitting
for 3/8-in. tubing), which extended to within 1/2 in. (13 mm) of the
ORNL DWG 76-584

AGITATOR DRIVE AND
M SEAL ASSEMBLY

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

BISMUTH
|NLET>——9£::::§\ <« SALT INLET
| AGITATOR -
| | ]~ T SHAFT -
™1 [
SALT 1 /
'SURFACE\ / —~
TR - 1
U ; 2
h 14 —aSALT
\ U; ( QUTLET
N Lt | |Eh
INT = —
, ¢ — > BISMUTH
T ) OUTLET
FLAT-BLADED . 7 7
TURBINES ARy

BAFFLES (4)

 

 

 

Fig. 2. ©SOchematic diagram of mechanically agitated molten-salt—
bismuth contactor.
bottom of the tank; (3) a sparge and pressurization port (with a fitting
for 3/8-in. tubing), which extended to within 1/2 in. (13 mm) of the
bottom of the tank; (4) a 1/2-in. pipe (with a fitting for 3/8-im.
tubing) used as a thermocouple well, which extended to within 1/2 in.
(13 mm) of the bottom of the tank; (5) a 1/2-in. pipe with a fitting

for a 1/2-in. ball valve and sampler and a fitting for 1/U-in, tubing
below the valve; (6) a l-in. pipe with a 1-in. ball valve as an addition
port; and (7) a 1/2-in. capped pipe as a spare port. Each catch tank
had the same ports as the feed tanks except that no addition port was
provided. The outer surfaces of the feed and catch tanks were flame

sprayed with nickel aluminide to retard oxidation.

2.t Treatment Vessel for Salt and Metal

The treatment vessel consisted of a 30LI, stainless steel pressure
vessel that held a graphite crucible. The cylindrical portion of the
pressure vessel was 26.5 in. (0.67 m) long [1/h-in. (6.L4-mm) wall thick-
ness] by 18 in. (0.46 m) OD and with 18-in., (0.46-m)-OD by 1/lU-in.
(6.4-mm)-thick dished heads on each end. It was designed to withstand

HE-HF at 600°C at a pressure of 50 psig (345 kPa).

The inner crucible, machined of graphite,® had an outer diameter
of 16.75 in. (0.43 m) and was an overall 26.75 in. (0.68 m) high. The
wall thickness tapered from 1.75 in. (44 mm) at the bottom to 0.75 in.
(16 mm) at a point 16.75 in. (0.43 m) from the bottom, and was uniform
from there to the top. The bottom of the crucible was 1.75 in. (44t mm)
thick. The crucible had a 16.75-in. (0.43 m)-diam 1id, whose thickness
varied from 1 in. (25 mm) at the rim tc 0,5 in. (13 mm) at the center.
The graphite crucible rested on a support plate ingide the pressure
vessel, and the 1lid was held loosely in position by three studs pro-
jecting from inéide the top of the pressure vessel, The vessel had 13

nozzles, which are described in Table 1.

 

*No. 8735, Speer Carbon Company, a Division of Air Reduction Company,
St. Marys, Pennsylvania,
 

 

Teble 1. Description of nozzles on treatment vessel
Nozzlie
Mo. Purposs Description
i Bismrth charging 2-in. sched Lo pipe, flanged at the top o
gccommodate o chute for loading bismuth. The
greghite 1id below this nozzle hes & 1.625-in, -
dism hole with & removable plug.

2 Bismuth sampling; salt 0.5=in. sched L0 pipe with bail valve and
sampling; ges-phase szmpler. The 1id is fitted with & 1-in.-ID
pressure connection graphite pipe into which the 0, 5-in, pipe slips,

The graphite pipe extends through the gravhite
1id and into the crucible for a distence of
1 in,

3 Retwrning salt from the 0.5-in. sched 10 pipe nozzle conteining e

galt rcceiver sleeved 0.375-in.-0D tube, PRelow the cerbon
steel=-to-molybdenium transition, the 0.375-ixn.-
0D molybdenum tubing extends It in. belcw the
graphite 1id.

ly Returning bismuth from the Identical to nozzle No. 3.

Dismuth receilver -

5 Transferring bismith to the 0.5=-in. sched L0 pipe nozzile containing a

bismuth feed tank sleeved Q,370-in, 0D tube thet extends to
within 0.5 in. of the bottom of the erucible.
The tubing that extends into the crucible is
made aof molybdenmum.

6 Transferring salt to the Similer to nozzle No. 93 set so that 15 liters
galt feed tank of =21t can be transferred to the g2l feed

tank, leaving 2 0.5-in. heel of salt on top of
the bisruth.

T Monitoring liquid levela Similsr to nozzle No. S.

8 Sparging with H,-HF gimilar to nozzle No. 5.

9 Adding salt Similay to nozzle No. 3.

10 Spare Similer to nozzle No. 3.
11 Thermocouple well 0.5-in. sched 40 pipe with fittings for 0.3735-
in, 0D tubing. _
iz Meking miscellanecus addi- I1-in, sched LO pipe, with ball valve.
tions, or vessel veniing .
13 Draining vessel C.5=in. sched 40 pipe extending from the

hottom of the pressure vessel; this line is
capped.

 

| &'Acfis as a bubbler type of liquid-level monitor,
2,5 Samplers

The treatment vessel and the feed and catch tanks were each pro-
vided with a lfz—in. sched 40 pipe nozzle fitted with a ball valve and
sample port. These tank sample ports held four sample capsules attached
to capillary tubing that extended through a Teflon plug in the top.
These capsules were lowered (while the system was under avgon pressure)
through the ball valve into the tank below, and samples were drawn into

the capsules by vacuum,

In addition to the five sample ports on the vessels, there were twe
Tlowing-stream sample ports that operated in a manner similar to that of
the tank sample ports. These flowing-stream sample ports allowed seven
samples to be taken from each of two flowing streams during operation.
One sample port was located on the salt return line (between the con-
tactor and the salt catch tank), and one was located on the metal return

line (between the contactor and the metal cateh tank).

The filtered sample capsules, which were used to take bismuth and
salt samples, were made from 1/L-in. (6.L-mm)-diam stainless steel rod
that was 3/h in. (19 mm) long. The sample capsules were fitted with a
porous 347 stainless steel filter on one end and 1/16-in. (1.6-mm)-diam
caplllary tubing on the other. Figure 3 shows a schematic diagram of a

sample capsule and a typical tank sample port.

2.6 Freeze Valves and Lines

Salt and metal flows through the facility were directed by four
freeze valves in the transfer lines, located as indicated in Fig. 1.
These valves were simply dips (in the carbon steel tubing) that were
fitted with air cooling lines. Those freeze valves that had to be closed
before any salt or metal could be transferred from the treatment vessel
were equipped with small reservoirs (about 50 cmS) upstream and down-
stream from the valve. The facility, which was of welded constructiofi,
contained approximately 200 £t (A1 m) of salt and metal transfer lines

(3/86~ and 1/2-in. pressure tubing).
- PLASTIC TUBING

fl iy
b

SAMPLERS

 

7 b TEFLON PLLEG

 

 

 

 

 

 

 

 

VENT PURGE

SAMPLE HOLDER

BALL VALVE

TOP GF VESSEL

   

 

"'f/lfl]/ff’[’} FH
I IT TP IITLT,
-t et

    
  

  
   

 

 

 

 

 

A
Pl d 711
;i et T Lt rnt Tt

LT e

 

 

¢
it
i
L

I
: !
i

 

 

 

 

 

 

 

L,
TP T LTI Y e rrd

T
i

 

 

Fig. 3.

—=T0 ARGON AND VACUUHM SUPPLIES

N

=

ORNL OWG 72-10408

‘r,a-I/lfi-in. STAINLESS STEEL

CAPILLARY TUBIMG, 40 in. LONG

TYFICAL SAMPLER

 

 

 

3/4 in.

 

1

 

N ‘Q\\C\\Q:.

 

 

 

 

3/18-in. DRILL

1/4-in.-DIAN
STAINLESS STEEL ROD

FOROUS METAL FILTER,
¢0-p PORE SIZE,
347 STAINLESS STEEL

Typical tank sample port and sample capsule.
10

2.7 Instrunentation and Control

The principal objective of the instrumentation and control system
was to provide closely reguleted flows of bismuth and molten salt to the
contactor. The range of flow rates for both bismuth and molten salt was
nominally L0 tc 500 ce/min, corresponding to experiment durations of
about 5 to 0.5 hr. Pressures and liquid levels in the five vessels
(treatment vessel and feed and catch tanks) of the facility were sensed
by Foxboro differential-pressure transmitters, which sent signals to
miniature pneumatic recorders or controllers. Liquid level was inferred
from the pressure of the argon that was supplied to a dip-leg bubbler in
each tank., Flow rates of bismuth and salt to the contactor were con-
trolled by regulating the rate of change of liquid level in the two feed
tanks., The feed and catch tanks, the treatment vessel, and the contactor
were maintained at the desired temperatures by automatic controllers;
transfer-line temperatures and temperatures of small components were
controlled by manually regulating the appropriate voltage transformers

that supplied power to Calrod tubular heaters.

Figure 4 is a schematic diagram of the control system that regulated
the flow of bismuth or salt to the coantactor. It was designed to circum-
vent the flow-control problems that sometimes occur when gas pressure is
used to maintain a constant flow of liguid from a heated feed tank. An
adjustable ramp generator and an electric-to-pneumatic converter were
used to linearly decrcase the set point of a controller that sensed
liquid level in the feed tank, The level was controlled by controlling
the flow rate of argon to the gas space of the feed tank. The result
wes a uniformly decreasing liquid level and, hence, a uniform discharge
rate of bismuth or salt from the tank. This control system was unaffected
by small increases in back pressure, partial plugging of transfer lines,
decreasing feed tank level, etc., or leakage of argon (a small argon
bleed was provided to improve pressure control). Small gas pressure
oscillations caused by temperature cycling was minimized by using time-
proportioning controllers. Rates of transfer of salt and metal between
the collection tanks and the treatment vessel were not required to be

closely regulated; therefore, manual control of pressurization was used.
ORNL DWG 7O-4357iRt

 

 

 

 

 

FOXBORO
ADJUSTABLE RAMP | _ | o verTer E/p |hrids RECORDER
GENERATOR (E vst) CONTROLLER

 

 

 

 

 

 

 

 

B

2%

Al
I -

o o be

A
TR

 

 

d P sl bl 4 5
/ a e [Z4 £ Izl i ¥ " "
CELL

 

 

 

 

 

Y
a

ARGON

 

BISMUTH QUT «@——
50-500 ml/min.

 

 

 

    
  
 
 
 

LEN
TEMPERATURE
RECORDER
CONTROLLER
(TIME
PROPORTIONING)

 

 

 

 

 

 

 

 

 

 

Fig. L. Schematic diagram of control system for metering bismuth
from the pressurized feed tank, T-l.
12

Heating circuits were controlled manually for 11 transfer lines and
the two flowlng stream samplers. On the transfer lines, the Calrods
rated at 230 V were operated at 140 V or less, and provided up to 185 W
per foot (600 W/m) of line, Typically, temperatures at three points on
each line were recorded. The temperature of approximately 100 points

was recorded for the system.

2.8 Gas Purification and Supply Systems

Three gases were required for the experimental facllity: anhydrous
hydrogen fluoride (HF), hydrogen, and argon. Because of the highly
deleterious effect of small amcunts of oxygen or webter vapor, the nomi-
nally pure bottled hydrogen and argon were further purified to remove
traces of oxygen or water vapor. The anhydrous hydrogen flubride that
was used only in the treatment vessel for hydrofluorination of the metal
and the salt was given no additional purification. A schematic diagram

for each of the three supply systems is shown in Fig. 5.

Highly purified argon was used for all applications requiring an
inert gas (e.g., pressurization of tanks for transferring bismuth and
molten salt, dip-leg bubblers for liguid-level measurements, and purging
of epparatus for sampling bismuth and galt). Cylinder argon with a
minimm purity of 99.995% was first fed to & bed of molecular sieves
(Fig. 5a), which reduced the water vapor content to about 2 ppm [-100°F
(=73°C) dew point]. The argon then flowed through a bed of uranium
metal turnings where the remaining oxygen and water vapor were removed.
A porous stainless steel filter removed any uranium oxide dust that
might have been carried from the uranium bed by the gas stream. The
maximum argon flow rate, based on the capacity of the molecular sieve
bed, was about 6 sefm (2.8 x 1073 std ma/sec).

The hydrogen purification system was a commercially available deviceX

that purified hydrogen by the selective diffusion of hydrogen across a

 

*Serfass hydrogen purifier, product of Milton Roy Company, St. Petersburg,
Florida.
13

ORNL OWG 70-280% /!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

{a)
MOISTURE
> MONITOR
&
X - 2
Cf’ ] MOLECULAR SIEVE BED | X URANIUM CHIF BED ~360
¥ bl (WATER REMOVAL) A ! (OXYGEN REMOVAL) FILTER > ccth
~25°( ~E50°C [{max.)
ARGON
CYLINDERS {4} ARGON PURIFICATION SYSTEM
' IMPURITIES BLEED §
{b) . T .
i | SERFASS : ~45
e HYDROGEN : sefh
PURIFIER (mex.)

 

 

HYDROGEM
CYLINDERS (2)

HYDROGEN PURIFICATION SYSTEM
{c)

.~ Eib/hr

| £ut t:‘:;'__:i . {max.)
E—#HOT WATER SUPPLY L e PK-HUT AR BATH

FOR CAFILLARIES

 

  
 

 

 

 

HF
TANK

 

 

 

 

HOT WATER BATH PRESSURE

RECORDER
CONTROLLER

 

 

 

HF SUPFLY SYSTEM

Fig. 5. BSimplified diagram of the gas supply systems.
14

palladium alloy barrier. Impurities, along with a small flow of hydro-
gen, were bled continuously from the upstream side of the barrier. The
capacity of the unit wes 15 scfhi (1.2 x 107% std m3/sec). Controls for

the purifier were self-contained.

The anhydérous HF supply system utilized small capillaries for
metering; a pneumatic controller maintained a specified pressure drop
across a capillary by controlling the HF gas supply pressure. This was
achieved by regulating the temperature of the water bath in which the HF
supply tank was suspended (Fig. 5¢). Accidental overheating of the HF
supply tank was prevented by a switch that released cold water into the
bath if the temperature exceeded 60°C. The minimum flow range for the
HF supply was nominally O to 0.25 1b of HF per hour (0 to 0.15 g/sec);
the maximum range was O to 6 lb/hr (0 to 3.7 g/sec).

3. EXPERIMENTAL PROCEDURES

In order to measure mass transfer rates in the eguipment previously
described, 1t was necessary to perform several operations. The proper
distridbution coefficient of the transferring materials was adjusted by
adding reductant thorium and lithium to the bismuth. Tracers (237U‘and
972r) were prepared by irradiating 236U and 96Zr in the Oak Ridge
Research Reactor (ORR) and these were added to the salt feed before
each run. The salt and bismuth were fed {through the contactor. Samples
were taken of salt and bismuth and were prepared for analysis. The salt
and metal phases were treated periodically with mixtures of hydrogen and

HF. Details of these procedures are described in this section.

3.1 Reductant Addition

Pericdic ddjustment of the reductant inventory in the bismuth phase
was necessary in crder to replenish reductant loss due to oxidation,.
since even the high-purity argon which was used as a cover gas still
contained a small amount of oxygen. The reductant inventory in the

bismuth also required adjustment after H,-HF treatment of the salt and

2
bismuth. The method used for adjusting the reductant inventory was
15

electrolytic dissolution of beryllium ions in the salt phase while the

.salt and bismuth were in contact in the treatment vessel.

© A schematic diagram of the experimental apparatus used for electro-
lytic beryllium addition is shown in Fig. 6. A 3/8-in. (8.2-mm) diam by
6~in, (152-mm) long berylliium rod was suspended in the treatment vessel
and immersed in the salt phase. The beryllium rod was connected electri-
cally to the positive terminal of a 12-V lead-acid storage battery vis
wire and a stainless steel rod, which is insulated electrically from the
treztment vessel. To couplete fihe circuit, the negative pole of the
battery was connected to an ammeter, a variable resiétance, and finally
to the treatment vessel. The bismuth phase served as the cathode in

+this electrolysis.

The overall reaction that tekes plaece in this electrolytic cell
when current is pessed between the beryllium anode and bismuth cathode

ig:

3

3
= o
5 Be” + U

F_(salt) +—Be2+F2(salt) + U°(bismuth) . (1)

3(
Thus, the electrolysis resulted in the oxidaticn of beryllium at the znode
(the beryllium rod) and the reduction of uranium at the cathode (the
bismuth surface). The reduced uranium dissolved in the bismuth and the
concentrations of thorium and lithium dissolved in the bismuth adjusted

to Sétisfy the equilibrium conditions that were reported previously.l

3.2 Tracer ITrradiation and Addition

Mass transfer rates between the sali and bismuth phases were

23Ty or

determined from the extent of transfer of and of Zr tracers that

were dissolved in the salt phase prior to an experiment. The salt and
bismuth were at chemical equilibrium with respect to the nonradiozctive
uranium and zirconium,

o7

The Zr tracer was prepared by irradiating a T.5-mg quantity of

Q
’6Zr02 enclosed in a quartz ampul in the ORR for ~ 2L nhr. The 9TZrO2

was then transferred to a 0.75-in. (19-mm)-diam steel capsule after an
*[9SSaA qUSWIBAIY YINWSIQq-1TBS 3Y3

ut eseyd 3TeS 9U3 0% WMTTTAISQ JO UOTAIPPE JIOJ OTHBWAYUDS °9 '8BT4

13SS3A
INIWIVIYL
HLNWNSIG-11VS

aod WAWNITTAY3H

Ad311vHd

 

 

LNNSIE
171VS

JOVHO1S
daiov-avai

 

 

 

 

 

 

313NNV
¢

 

 

«§§¥T|l@|lw_ *

dO1SIS3Y 318VIHVA

 

 

I1I86-92 'OMA TINYO
17

3gs activity from the smpul) to

18-hr decay period (for decay of
facilitate addition of the tracer to the salt phase. The capsule was
then immersed in the salt phase in the salt feed tank while argon was
sparged through the capsule to circulate galt through the capsuie and
enhance mixing. ' |

236

Uranium-237 tracer was prepsred by irredisting v 1 mg U (as

236U308) encased in a quartz empul in the ORR for v T2 hr. As with
the zirconium tracer, the uranium was then transferred to a steel =ddi-

tion capsule and loaded into the salt phase in the salt feed tank.

3.3 Run Procedure

A1l runs were performed using the same procedure. While the
fluoride =salt and bismuth were in contact in the trestment vessel
(T5), sufficient beryllium was added to the salt electrolytically to

produce the desired distribution coefficient (D).

Prior to a run, the salt and bismuth phases were separated by
pressurizing the salt-bismuth treatment vessel and transferring salt
and bismuth to their respective feed tanks. Approximately T mCi of
?Tzr-"Timo and 50 to 100 nCi of *>Tv,0g were alloved to diszolve in the

salt phase sbout 2 hr prior to an experiment.

Salt and bismuth streams were passed through the contactor vessel
at the desired flow rates by controlled pressurization of the sgalt and
bismuth feed tanks. The contactor was maintained at 520 to 600°C for
gll runs. Both phases exited through é common effliuent line, separated,

and returned to the salt and bismuth catch vessels.

3.4 Treatment with Hydrogen-Hydrogen Fluoride Mixture

Periodic treatment of the salt and bismuth phases with HF-H2 mix-
tures was necessary to remove oxides from the salt, and dissolved
reductants and impurities (thorium, lithium, and iron) from the bismuth.
The treatment procedure alsc served to adjust the equilibrium distri-

bution of uranium and zirconium between the salt and bismuth phases.
18

The procedure followed was to sparge an HF—F2 mixture into the salt
phase while both the salt and bismuth were in the treatment vessel (T5).
In order to prevent excessive attack on containers and piping, the
hydrogen fluoride concentration was kept below 30 mole % by dilution
with hydrogen, although an attempt was made to keep the HF concentration
as near 30% as possible to afford the maximum oxide removal rate. The
total nominal flow rate was 30 scfh. The hydrogen fluoride flow rate
was set by the pressure drop across a capillary, and the H, flow rate

2

was set by a rotameter which was calibrated with H From the treatment

vegssel, the HF-H

, stream passed through a sodium fiuoride bed (to remove
HF) and then was exhausted to the atmosphere outside of Building 3592.
The feed and off-gas from the treatment vessel were analyzed by
diverting a small portion of the stream through a small agueous scrubber
and a 0.05 ftB/revolution (1.4 x 10_3 m3/revolution) wet-test meter,
which were connected in series. The concentration of HF in the gas
stream was determined by passing the gas through 250 ml of a 0.k N NaOH
solution in the scrubber. When 0.05 £65 (1.1 x 1073 m3) of H, had
passed through the wet-test meter, the gas flow was stopped and the
solution was removed for analysis. The HF concentration in the gas was
determined by titrating small samples of the scrubbing solution with
0.1 N HC1. Utilization of the HF was calculated from the feed and

discharge concentrations.

3.5 Sample Preparation and Analysis

In an effort to avoid contamination of the samples obtained in each
run with extraneous material, the sample capsules were cleaned of foreign
material before analysis by the following procedure. Gross amounts of
salt or bismuth were first removed with a file, the sample capsule was
then polished with emery cloth, and, finally, the capsule was washed

with acetone.

The samples were analyzed by first counting the sample capsules for
the activity of 51U (207.95 keV B~) and the activity of °  |Zr-> Nb
(743.37 keV and 658.18 keV B, resvectively) after secular equilibrium
was reached between QTZT and its daughter 97Nb. The material in the
19

237

sample capsules was then dissolved, and the activity of U was counted

97Zr~97Nb activity had decayed toc a very low level. This

again after the
was done to correct for self-absorption in the solid samples. Mass
transfer rates were than calculated from the ratios of tracer concentrstions

as discussed in Appendix B.
Iy, EXPERIMENTAL RESULTS

Nine runs were made in the experimental equipment to measure rates of

mass transfer of 23TU and o1

Zr between selt and bismuth. In these runs,
+the agitator speed was varied over the range of 68 to 24k rpm, and the
operating temperature was held in the range of 590 to 600°C. Concentra-

r Mgy and 237U were messured (as deseribed in Sect. 3.5) in the

tions o
salt input and both the salt and bismuth effluent streams. The counting
dats obtained in a2ll runs are given in Appendix A. Using these concen-
trations, three different equations were used to calculate the mass
transfer coefficient between_the salt and biémnth in the contactor.

The derived equations are given in Appendix B. The calculeted mass
transfer coefficients for all the runs are summarized in Table 2. The
.values given in Table 2 are the average of the values calculéted from the
three equetions [Egs. (B-18)-(-20)] with the standard devietion. Values
are given both for results based upon the uranium counting and for results

based upon the zirconium counting.

Run TSMC-1 was mainly a preliminary run designed to test the procedure,
Salt and bismuth flows were approximately 200 ce/min, and the stirrer rate
was 123 rpm. Unfortunately., the distribution coefficient (defined in
Appendix B) was too low to effect any significant mass transfer, and mess
transfer rates could not be determined accurately; thus, no results are

shown for this run in Table 2.

Operation of the equipment during run TSMC-2 was very smooth. The
salt and bismuth flow rates were 228 and 197 ce/min, respectively. The
‘distribution coefficients were higher than for the previous run, but were
lower than desired, resulting ian much uncertainty. Several determinations
of the distribution coefficient of uranium D_ were made that ranged from

)
C.9Lh to 3L, One determination was made of the distribution coefficient
20

.Afio\mo.l HV = PIJIISDISUBIY JI30BIY GOfipomhmd

 

 

 

-—- gOT°0 + T2T'O 76°0 - I < he 19T 69T 6-0DHSL
- 0T00*0 + 2200°0 G20 -- of < 0~ 79T 2sT §—OWST
——= 2T00°0 + LS00°0 o0 -- L6 < 89 OLT 2sT L~OWST
TO°0 + 020°0 G00°0 + 6£0°0 719°0 2 gLt < 09T ¢eT 902 9=OHSL
6STO0 + €9TO0 £T00°0 + $600°'0 GE*0 e €n < LTA GLT 612 G—OWST,
20°0 + G&£0°0 200 + H®S0'O gL'0 e gL < S02 T 0LT H1—OWSL
-—= €00°0 + 2T0°0 0$°0 - € < 29T €LT 991 € -OWST
¢G00'0 + £800°0 g2600°0 = 6500°0 LT°0 960 HE-46°0 12T L6T gee 2—OWSL

IMTUODITZ wniue.ln pPSII8I SUBLYL MNQ Dm Aamhv {(utm/29) Anfla\uov uny

uo poseq U0 paseg I90BIY 998X MOTJI Ujnmstg MOTJI 371BS
Aomm\.fiov mum TUOTFOoOBILL I3JIATIG

 

J040BIUOCD YLnUSIQ-4TeS 23 UT SqUallsINSeBSW J9JSUBLY S8BW JO sSyTnsad TeauswTJIodxy
21

of zirconium DZr-that indicated Dy.. was 0.96. Consequently, only a range
of possible values for the overall mass transfer coefficient could be
stated for the results based on uranium. A value for overgll mass transfer
coefficient based on zirconium is given, but since the value for DZr is
uncertain, there is more uncertainty in the mass transfer coefficient

than is indicated by the reported standard deviation given in Table 2.

A bismuth line leask occurred immediately preceeding run TSMC-3.
During the resulting delay for repairs. the 9TZr decayed and only the
23TU tracer could be used. The remainder of the run went smoothly. Salt
and bismuth flow rates were 166 and 173 cc/min, and the stirrer rate was

162 rpm. A high value for Dy (greater than 34) was maintained for this

- run.

Tn run TSMC-L, flow rates of 170 and 1L4 cc/min were set for the
salt and bismuth flows, and a stirrer rate of 205 rpm was maintained.
The distribution coefficients, which had been determined from samples
taken before, after, and during the run, were greater than 172 for DU
and greater than 24 for DZr' These values are sufficiently large so that
Eqs. (B-18)-(-20) in Appendix B are valid. lLarge distribution coefficients
cannot be determined precisely due to the inability to determine very
smgll amounts of uranium in the sallt phase. No problems arose during

this run.

Runs TSMC-5 and -6 were performed without incident. The distribution
coefficients were maintained at high lefiels for both runs. TSMC-5 had a
stirrer rate of 12U rpm and salt and bismuth flows of 219 and 175 ce/min.,
TSMC-6 selt and bismuth flows were 206 and 185 cc/min, respectively, with

a stirrer rate of 180 rpm.

Prior to run TOMC-T, two leaks developed in the blsmuth transfer line
from the bismuth feed tank to the contactor vessel. This transfer line
was completely replaced slong with the associated Calrod and thermal
insulaticn. The volumetric flow rates of salt and bismuth during the
run were 152 cm3/min and 170 cmS/min, respectively. The stirrer rate was
set at 68 rpm for the run. The uranium distribution coefficient was
greater than 97. Seven sets of salt and bismuth flowing stream samples

were taken from the contactor effluent during the run,.
22

Run TSMC-8 was performed with salt and bismuth flow rates of 152 cc/
min and 164 cc/min, respectively. The uranium distribution coefficient
was maintained at a level (> 40) for this run which was greéter than the
minimum desired value of 20. It was presumed that the agitator operated
at 241 rpm, which is high enough to produce mild dispersion of the phases
in the contactor and, therefore, a high measured mass transfer rate.
However, results from this run indicated that very little (v 25%) of the
237U tracer was actually transferred from the salt to the bismuth.
Inspection of the magnetically coupled, agitator drive assembly indicated
that an accumulation of a highly viscous carbonaceous material between
the upper carbon bearing and the agitator drive shaft had prevented
proper rotation of the shaft. The drive assembly was cleaned of all foreign

material, was reassembled, and was found to operate satisfactorily.

The ninth tracer run, TSMC-Q, was performed as a repeat of the eighth
run. Salt and bismuth flow rates were set at 169 cc/min and 164 ce/min,
respectively. The agitator was operated at 2L4 rpm during this run. A
high stirrer rate was maintained to determine the effects on the mass
transfer rate of dispersal of one phase in the other, and to determine
if large amounts of bismuth and salt are entrained in the other phase
after settling in the contactor effluent line, Entrainment results from
this run and another similar run are given in Appendix D. The uranium
distribution coefficient was greater than 47 during this run. No

systematic problems were encountered and the run was performed smoothly.
5. INTERPRETATION OF RESULTS

In this section, the mass transfer coefficients measured in the salt-
bismuth system are compared with typical mass transfer coefficlents
measured in aqueous-organic systems at comparable agitator diameters and
speeds and with mass transfer coefficients measured in a water-mercury
system. The mass transfer coefficients measured in this study are also
compared to predictions made by three mass transfer correlations taken
from the literature that were developed from data measured in aqueous-

organic systems.
23

EBTU given in Teble 2 &are prcobably

T

The mass transfer coefficients for

more reliable than the results given for

237

U was greater than 80%, whereas that for 9TZr was

Zr because, In all cases, the
maeterial balance for
consistently about 60%. Because of this, and also because more useful

data points were obtained for 237U than for o

Zr, the interpretation of
results presented in this section is based mainly on the 237U measurements.

The magss trangsfer coefficients for 237

U are compared in Fig. 7 with
some typical mass transfer coefficients measured in aqueous-organic systems
and with water-side mass transfer coefficients measured in a water-mercury

39'1*!'

systemn. The figure shows that the mass transfer coefficients measured
in the salt-bismuth system (curve A) are quite high compared to the agueous-
organic and water-mercury results measured in cells of comparable size

and at comparable agitator speeds (curves B through E). Except for curve
E all the mass transfer coefficient data can be divided into two regimes:
(1) at low agitator speeds the mass transfer coefficient is proportional
to the agitator speed raised to a power'less than 1.0 (0.9 for the salt-
bismuth results and 0.7 for results represented by curves.BAthrough E):
-and (2) at higher agitztor speeds the mass transfer coefficient is pro-
portional to the agitator speed raised to a power significantly grezter
than 1.0 (1.5 for curve D, 1.95 for curves B and ¢, and 3.0 for curve A).
Olander asnd Benedict2 suggest that the sudden change in expofient for their
data (in the absence of phase dispersal) is caused by a laminar-turbulent
transition at the interface. Observation has shown that phase dispersal
did not occur at the break points of curves B, C, and D, but it was not
possible to determine unequivocally when dispersal occurred in the salt-
bismuth system since measurements of entrainment were Incconclusive cn

5,6

this point (see Appendix D). However, previous work with water-mercury
and with agueous-organic systems Indicates that, for the agitator diameter
used, dispersal of molten salt into bismuth should begin to cccur at an
agitator speed of about 170 rpm. This entrainment would cause the apparent
mass transfer coefficient to be greater than the mass Transfer coefficient
that would have résulte& if phase dispersal had not occurred. The increase
would be due to the increased arez for mass transfer. since the apparent

mass transfer coefficient is based upon the area of The undisturbved inter-

face. Since dispersal was expected, and because the dependence on agitator
2l

speed is so different from the aqueous-organic data and the water-mercury
data, it is concluded that dispersal of salt into bismuth occurred at an
agitator speed between 160 and 180 rpm, even though entrainment measure-
ments do net support this.

91

The mass transfer coefficients for Zr are shown in Fig. 8 compared
to curve A from Fig. 7. In all but two cases (runs TSMC-2 and —5), the
zirconium mass transfer coefficient was lower than the uranium mass trans-
fer coefficient. This difference is probably related to the inability to
correct for the self-absorption of the Ti3.37 keV B~ in the analysis of
97Zr in the solid bismuth samples. In run TSMC-2, all the resistance to
mass transfer of uranium was in the salt phase, whereas there was sig-
nificant resistance to mass transfer of zirconium in both phases. Never-
theless, the overall mass transfer coefficients for uranium and zirconium

were of comparable magnitude in this run.

The mass transfer coefficients were compared with three literature
correlations for mass transfer coefficient in stirred cells that were
developed for aqueous-organic systems. The properties of the fluoride
salt that were used to evaluate these correlations are given in Appendix
D. The properties of bismuth at 600°C that were used are:

s = Q.66 g,/cm3 and

Np; = 1.0 x 1072 g/cm-sec.
Lewis7 investigated mass transfer rates in mechanically agitated,
nondispersing contactors, all of the same size, using several agueous-

orgahic systems. He fitted his results with the following empirical

 

equation:

60 k, 6 N, 1.65

= 6,76 + 10 Re. + Re,, — + 1, (1)
V , 1 2 1N
1 1
where
I\TL2
Re = Reynoclds number —E—E- )

= stirrer speed, rps ,
= gtirrer length, cm ,

= mass transfer coefficient, cm/sec ,
25

ORNL DWG.Fe-582

 

ledl

Ix 109

% 1G9

MASS TRANSFER COEFFICIENT (cm/sec)

 

 

1 { 1 1T 1T iT

 

237
U TRANSFER IN FLUORIDE SALT
(B HEAVY PHASE) 73 mm AG!TATOR
QUINONE TRANSFER IN WATER {(REF. 3,4)
(Hg HEAVY PHASE) 89 mm AGITATOR
QUINONE TRANSFER IN WATER (REF. 3,4)
(Hg HEAVY PHASE)E4 mm AGITATOR

WATER IN 20% TBP-70% n-HEXANE (REF 2)
(WATER HEAVY FHASE) ?6 mm AGITATOR
WATER IN ISOCBUTANCL (REF. 2)
(WATER HEAVY FHASE) 76 mm AGITATOH

llllll

el

i

 

 

(x 107
.

Fig. T.

water-mercury coefficients.

Comparigon of

! |
20 40

€6 80 100 200 400 600
AGITATOR SPEED {rpm) .
237

to run number in the salit-bismuth experimental facility.

IOOO

U mass transfer coefficients in fluoride
salt with representative agueous-organic mass-transfer coefficients and
Numbers in parentheses on top curve refer
26

ORNL DWG. 76-578

 

1x10"!

 

 

 

= N
w o
o
£ L
Q0
e 01(4)
w N
o
w
w - 0 (8)
3 0(s)
o
L -
» Ix10
2 i o(2)
o L
-
wn L
N
I
= L
2,"0'3 1 I 1 [ I i 1 1 1
0 20 40 60 80 100 200 400 600 800 1000

AGITATOR SPEED (rpm)

Fig. 8. Effect of agitator speed on salt-side mass transfer coef-
ficient of 9T%r in the salt-bismuth contactor. Numbers in parentheses
refer to run number.
27

p = density, g/cm3 .
n = viscosity, g/em sec ,
. = kinematic viscosity, n/p. cm?/sec, and

1.2 = phase being considered .

For the case in Which,Nl = N2 and Ll = L2, the above equation can be

reduced to the form:

60 k &

L = 6.76 x 10” Re. | 1 + -2 +1 . (2)
Ul 1 pl

 

For Lewis' work, where the densities of the various phases varied from
0.8 to 1.2 g/cm3 but the stirrer length was kept constant, this correla-
tion effectively uses only the Reynolds number of the phzse being con-

sidered.

The wranium mass transfer coefficients are compared to the Lewls
correlation in Fig. 9. At agitator speeds bélOW’lTO rpm, the Lewis corre-
lation overprediclts the mass transfer coefficient for uranium; it zalso
shows a stronger dependence of mass transfer coefficient on the agitator
speed than the data indicate., No dependence of molecular diffusivity is
shown by fihis correlation. The omission of a term contasining molecular
diffusivity from the correlation has been criticized in the recent liter-

8,9 9

sture. It has been shown theoretically and in recent experiments
that the mass transfer coefficlent should be proportional to molecular

diffusivity raised to a power near 1/2.

Mayerslo developed a slightly more involved correlation of the

following form:

171 i/2 (N2 Mo ~1/6
S = 0.1896 (Bel Reg) 0.6 + = (scl) . (3)

1 Ny 1

When both phases are stirred with identical paddles at the same speed,

this equation reduces to:
28

ORNL DWG 76-583

 

 

 

 

 

1000 ! v
-
00 —
T’ -
x_ — -y
gl - -
~ o(3) .
o(5)
=T o7 -
l L 1 1 1 Lt 1' 1 i 1 i L it ll
10 100 000
n .68
676 x 107 [ Re, + Re, (=2 )]
i

Fig. 9. Comparison of uranium mass transfer coefficients with the
Lewis correlation. Numbers in parentheses refer to run number.
29

L - | -
60 kL n, 1.4 n, 2.h 6o 0.5
——— = 0.1896 (Re;) | == 0.6 + —= (5c,) (L)
1 My Ny \P1
wnere
S¢ = Schmidt number, n/p& .

This correlation, which is based on data. covering a limited renge of den-
sities (0.8 to 1.2 g/cm3), indicates that the viscosity and dénsity of

each phase affect the mass transfer ccoefficient.

A comparison of the uranium mass transfer coefficients with the Meyers
correlation is shown in Fig. 10. At agitator speeds below iTO rpm, the
Mayers correlation also overpredicts the mass transfer.coefficient, but
the predicted dependence of mass transfer coefficient on agitator speed
is more nearly in accord with the data than it is in the Lewis correlation,
Note that the dependence on the Schmidt number (molecular diffusivity) is

fairly weak.

McManamey ™ correlated his data and the results obtained by Lewis
by using the following expression, whilch i1s similar to that used by Lewis

but includes the Schmidt number:

 

60 k N. Re
_I —
—= = 0.0861 (Rel)0'9 1+ -2 fi (se, ) 0.37 (5)
1 N &
where
S¢ = Schmidt number, n/p® , and
® = diffusion coefficient, cm?/sec.

This equation can be reduced ta:

60 k | _ P o A
L - 0.0861 (Re )O'9 1+ = (Sc )‘O'37 . ' (6)
1 1 Py .

 

Vv

for the case where Ll = L2 and Nl = Ng. Note that the numérical constant

in Eqgs. (5) and (6) has the dimension of reciprocal centimeters.
30

ORNL DWG 76 - 580

 

 

 

 

1000 | 1
- _
r-— e~
® (9)
100 4
- ®{4) ]
n : ®(6) -
—.l -
4 -
i
co - g
®(3)
®

10— (5) .
. ®(7) J
L_ —

| o ' |

! 0 100

oums( e, e, )* (221" (064 387" (sc, 7™

Fig. 10. Comparison of uranium mass transfer coefficient with the
Mayers correlation. Numbers in parentheses refer to run number.
31

The uranium mass transfer coefficients are compared with the.MbManamey
correlation in Fig. 11. - This correlation shows good,agreementlwith the
data at agitator speeds below 170 rpm. It must be pointed.out, however,
that the Schmidt number for diffusion of uranium in molten fluoride selt
was estimated by using correlations based upon materials with solution
behavior that is gquite different from molten salt solutions. Hence, the

very good agreement shown here should be considered somewhat coincidental.
6. CONCLUSIONS AND RECOMMENDATIONS

The following conclusions and recommendations are based upon the

experimental results and analysis presented in this report.

(1) At comparable agitator speeds, salt-side mass transfer coeffi-
cients for uranium are higher than water-side mass transfer k
coefficients for quinone measured in water-mercury systems,
and higher than mass transfer coefficients for other Coumponents
in aqueous—-organic systems. Furthermore, the dependence of
salt-side mass transfer coefficients on agitator spesd seems
to be somewhat stronger than Tor the water-mercury system at

low agitator speeds and with similsr diameter'agitators.

(2) The change in slope of the mass transfer coefficient vs agitator
speea curve at 170 rpm is probably caused by the onset of phase
dispersal. The occurrence of dispersal at this speed is in
reasonable agreement with data measured in water-mercury and

aqueous—-organic systems.

(3) There is & large increase in mass transfer rates with only
Slighthhase dispersal. Tt may be possible Tc achieve mass
transfer rates required in the MSER processing plant by
operating under these cbnditidns without suffering bismuth
entrainment in the salt. The data on bismuth entrainment
'presented in the Appendix suggest this; but further testing

is required to confirm it.
32

ORNL DWG 76-579

 

1000 T T T T T T 7] T T T T T

L1 11
i _J. 1

- ®(9) . 1

 

 

 

 

T
£ _
x|
wh -
10 ==
-1 -
- -
-1 -
I L Llllllll 1 ) b Lk 1
| 10 100

0.0861( Sc, )37 (Re, )08 ( | +22 Bez )
) Re,

Fig. 11. Comparison of the uranium mass transfer coefficients with
the McManamey correlation. Number in parentheses refers to run number.
33

(4) The mass transfer coefficients measured at agitetor speeds
below 170 rpm.provide“data Which,should be compared with new
correlations for stirred nondispersing contactors. The
McManamey correlation correlated these data much beter than
twd other litefature‘éorrelations; however, it should be used
with extreme caution for scaleup and design if extrapolations

to untested conditions are reguired.
T. ACKNOWLEDGMENT

The authors wish to acknowledge the corntribution of Mr. J. Beams,
technician assigned to this project, for his diligenée and skill in
operating the rather cantankerous experimental equipment used for this
work. Appreciation is also due to My, Max Montgomery, pipefitter, for

maintaining the equipment in working condition.
8. REFERENCES

1. L. M. Ferris et al., J. Inorg. Nuel. Chem. 32, 2019-35 (1970).
D. R. Olander and M. Benedict, Nucl. Sci. Engr. 1k, 287-9L (1962).
C. H. Brown, Jr.., in Engineering Development Studies for Molten-Selt
Breeder Reactor Processing No. 2L, ORNL/TM-5339 (in preparstion).

4, J. Hernanz et al., Determination and Correlation of Mass Transfer
Coefficients in a Stirred Cell, ORNL/MIT-22 (in preparation).

5. dJd. A, Klein and C. H. Brown, Jr.., unpublished dats.
6. H. 0. Weeren and L. E. McNeese in Engineering Development Studies

for Molten-Salt Breeder Resctor Processing No. 10, ORNL/TM-3352
(December 1972), p- 52.

 

T. J. B. Lewis, Chem. Eng. Sci. 3, 248-59 (195L4).

8. D. R. Olander, Chem. Eng. Seci. 18, 123-32 (1963).

9. W. J. McManemey et al., Chem. Eng. Sci. 28, 1061-69 (1973).
10. G. R. A. Mayers, Chem. Eng. Seci. 16, 69-T75 (1961).

11. W. J. McManamey, Chem. Eng. Sei. 15, 251-5k (1961).
3k

l2. A. S. Foust et al., Principles of Unit Operations, p. 210, Wiley,
New York, 1960.

13. C. R, Wilke, Chem. Eng. Progr. 45, 218-24 (1949).

14. R. B. Bird et al., Transport Phenomena, pp. 514-15, Wiley, New York,
1960,

15, &. Cantor in Molten-Salt Reactor Program Semisnn. Progr. Rep. for

Aug. 31, 1969, ORNL-4LLk9, p, 1b5,

16. 5. Cantor in Pnysical Properties of Molten-Salt Reactor Fuel,
Coolant, and Flush Salts, ORNL/TM-2316 (August 1968), p. 28.

17. S. Cantor in Molten-Salt Reactor Program Semiann. Progr. Rep. for

Aug. 31, 1969, ORNL-4hhg, p. 1LT.

18. R. B. Lindauer, personal communication, Feb. 28, 1973.
35

APPENDIX A.
Sample Analyses

The counting data obtained during runs TSMC-2 through -9 are pre-
sented in Tables AI-A8. Counting data are given for 237U (207.L5 xev R7)
and ° 7z (743.37 keV B ) in the solid salt and bismuth samples, and for
2BTU after the samples were dissolved. All results are givén in terms

of counts per minute per gram.
36

*aTdwes weaxis FUTMOTT = S 4L = G ‘Hl

= €L =€ ‘8L =2 ‘TL = T ‘utrdtio aofdmes = )
pue ¢(378S = § ‘Uynwslq = g) 97dwes UT TBTISIBW = ¢ tTaqumu oTduBS = ¥ 8J9UM ‘H-g-Y¥ 07 FuTpuodsaiaoo spod B £q poqwuBTSep ST oTdmss sowmm

 

 

 

 

 

 

40T X G6°L gOT X w.m“M £0T X 656 G=5-gTL #0T X f6°e #OT ¥ ®E°9 LO0T ¥ ST S G-g-91T
70T X £€9°9 0T X 6°T > 0T X €6°T G-8-LTT 40T X Of"¢ /0T X 26" ¢ ,O0T X 9L°2 G-g-ST1
¢OT X 98°T gOT X 16°c gOT X 662 H=g-%1T 70T ¥ £8°T 70T X 9L ¢ 70T X 96°T c—4-0TT
gCT X 621 ¢OT X cg8°¢c ¢O0T X T2 h-S—€T1 x0T X T9'T 40T X 62°'€ yOT X g1°T 2—-g-60T
0T X €6°T gOT ¥ 1I6°6 gOT X HT°G £-g-21t 0T X G0'f x0T X RO°T ¢0T ¥ gT°K T-9-g0T
0T X 76T 0T ¥ 60°G gOT X 00°§ £-5-T1T ¢0T X Of"} 40T X QO0°T 0T X 60* 4 T-g-L0T
I JI299J8 U2YB3 Sordweg

0T ¥ 49°T 0T X LG°¢ 0T X T6'e Sd~8-90T 70T X wl'2a 40T X T0°'S §OT X et Sd-g-66
gOT X ®9°'T 0T X Li*g gOT X fL°2 Sd-5-60T x0T X L6°T 40T X T6°¢ /0T X 98°T SJd-g-g6
¢OT X 06°1 0T X lg-e gOT X 69°C SA-5—-40T ;0T X 76°T 40T X lg-¢ a0T X L A Sd-g-L6

- -= - S4-5-£0T x0T X 0T"¢ 40T X g2 % x0T X fc'c Sd-g-96
cOT X TL'T 0T X [G°2 ¢OT X "¢ Sd-g-c0t £0T X 00°¢ #0T X 082 70T X 16T Sd-g-66
OT X 25T 0T ¥ 0L°e c0T X Ix*c Sd-5S-T0T 70T X 69T #0T X 62°¢ {OT X o' T Sad-g-16
gOT X g°T 0T X 9¢°2 0T X gt°¢ Sd4-8-00T 40T X T0'¢ §OT X 61" € x0T X 08°T Sd-g-€£6

TUhd PULTAND UsyNB] So[duBg
gOT X 6L°T ¢0T ¥ 60°2 gOT ¥ 96°1 g-g-cé
GOT X 68T 0T X 6L°2 ¢OT X T0°¢ £=-5-16
SI90BIG JO UOTIFIPPRE J27JB 4nq ‘uUng 03 J0TJad usys] Sof g
0T X 9°9 > 0T X g9 > #OT X 9°6> £-5-06 _ _ _
0T X 6°¢ > e0T X %°9 > g0T X m.m.w £-5-68 (0T X 9°9 > 0T X £°¢ > 0T X ¢°¢ > T-d-4%6
Z0T X P.N.W x0T X 8°T > gOT X 879 > S-s-lg 20T X mo.H.W 0T x m.HAW z0T X m.m.w T-g-£6
0T X G°'f7 > 40T X @°T > gOT X G°Q > 5—-5-9¢8 10T X LE"9 > 0T X g'c > 0T X €°€E > c-9-58
unI o3 JoTJd Ud¥e] SaTdweg
(uTw-38/s3unod) (utw-3/sjunca) (uTw-%/s3unco) g9P02 (uTw-3/squnoco) (utw-g/squnoo) (uTwm-% /S3Unoo) g9POD
I1Z,¢ 103 Nz ¥0F N;gz 207 oTdmeg IZ,6 I0F O,¢7 107 N;gz %03 oTdursg
sTsfTBUB DPITOS STSATEBUR UOTINTOQ STSATBUB PLITOY STSATBUB PITOQ STSATBUB UCTINTOYG sTSATBUB PITOg

 

Z=0WSI UnI woJgJ psulBlqo ®BIBP JFUTIUNOD

*I-V STqEL
37

Teble A-2. Counting data ¢bteined from run TEMC=3

 

Solid =znalysis Solution analysis Solid anslysis Sclution enalysis

 

 

 

 

 

Sample for 237y for 237y Sarmple for 237y for 2377
code? (counts/g~min) (counts/e-min) code® {ecunts/g-min} {counts/g-min)
Semples taken prior Lo run
141-B-5 < 1.13 x 102 < 5.1 x 1083 1355 < 3.3 =x 108 <5.3 x 10"
142-B-5 < 1.6k x 102 < 5.5 x 108 ibkl-8-5 < 3.2 x 109 < 7.6 x 1o%
147-B-5 < §6.26 x 10} < 3.L x 108 145-5-3 < 3.k x 103 < 7.6 x 10“*
148-B-5 < 1.00 x 102 <Lko0 x10° 146-5-3 < 3.2 = 10% <5.6 x 10%
Samples tzken prior to run but sefter addition of tracer
1hg-g-3 6.13 x 105 9.37 x 10°
150~8-3 6.33 x 10° 9.20 x 10°
Semples taken durinsg run
151-B-FS k.52 x 1o% 1.13 x 10° 158-8-F8 3.48 x 10° %.89 x 10°
152-B-FS b1k x 10% 1.29 x 103 159-8-FS - -
153-B-F8 4,12 x 10" 1.33 x 10 160-S-FS 2.96 x 10° h.36 x 107
154-B-FS L.51 x 10% 1.bhk x 103 161-8-F8 3.08 x 10° k.56 x 10°
155-B-FS £.32 x 10° 1.26 x 109 162-5-FS 3.02 x 10° 3.99 x 10°
156-B-FS 3.88 x 10% 1.5 x 10 163-8-FS 3.15 x 10° h.k3 x 10°
157-B-FS k.00 x 10% 1.49 x 10° 16k4-5=FS - —
Semples tzken after run
165-B-1 L.38 x 102 4.8 x 103 167-8-3 6.68 x 10% 1.19 x 105
166-B-1 6.53 x 102 4.7 x 109 168-5-3 _— -
169-B-2 3.19 x 10% 1.61 x 10° 171-S-L 2.6% x 10° L.71 x 10°
170-B-2 2.86 x 10" 9.56 x 10% 172-8-k 2.70 x 10° L.i2 x 10°
173-B~5 5.47 x 10% 1.h3 x 107 175-9-5 < 6.b x 103 <1.,b x 109
17h-H_5 5.24 x 10% 1.hh x 10° 176-8-5 < 6.h x 103 <1.L x 105

 

S%ach semple
meterizl in
b =Th; 5=

is designated by a code corresponding to A-B-C. where A = sample number; B =
semple {B = bismuth, & = salt); and ¢ = sample origin; 1 = Tl; 2 = T2; 3 = T3;
T5; F8 = flowing streszm sample.
38

*oTdures wBaJI3S BUTMOTI = Sd 6L = 6 fl = f $€L =€ 2L =2 €11 = T ‘utdrIo ordumes = )

 

 

 

 

 

 

pue ¢(3T8BS = § ‘ynmwslq = g) S7dwes UT TBIIS}BW = g {Jaqunu 3Tdwes = Yy aI8ym ‘H~-g-y 03 Jurpuodsalaos apod B £q pojeudissp st oTdwes UoBH,
0T X 6°€ > 40T X 6°% > 40T X 0°8 ¢-5-t2e 40T X 29°¢ 0T X gE°2 40T X g2°¢ G-g-222
0T X 2'¢ > 0T X L' > LO0T X t°g g-g-tze 40T X ge e 0T X te'e 0T X 98" L S-g-T22g
40T X L9°} 0T X TL°¢ 0T X gt°¢ ~S=0c¢ 70T X g9°¢t g0T X LT°¢ 0T X L0°T c—d-9Te
70T X 88K 0T X 19°¢c 0T X 62°¢ f1-8-61¢ 40T X 69°€ ¢OT X %g°¢C ¢OT ¥ 80°T g-g-61e
gOT ¥ 9t*¢ 90T X }5°T g0T X G2°T £-s-Q12 0T X T°T 0T ¥ L°L 0T X EL°T T-9-1Te
0T X 6£°¢ 0T X gH°T 0T X TE°T £-g-L1¢ ¢0T X 6T 0T x L°'g ¢0T X 00°¢ T-g~£T1¢
ungd J533J8B USYEBY SSTAWBS
#OT X #4879 gOT X KT°¢ cOT X 26°¢ Sd-S-c1c 0T X 2T°L 0T X 2€°¢ 0T X LG°T Sd-g-502
50T X gh°8 0T X 0Tt 0T X £9°¢C SA-S-T12 ,0T X €7°9 0T X 9t ¢ ¢OT X £8°T Sd-g=w02
40T X 96°9 cOT X fE"E 0T X 9L°¢g Sd-5-0T<c 40T X g&°9 0T X T0°S 0T X 61" Sd-g-£0¢
40T X L0°9 0T X 10§ ¢OT X l6¢ Sd-5-60¢ x0T X f1£°9 0T X gl 'y 0T X 6§°T Sd-d-c08
40T X TO'q gOT X gg°'c 0T X g6°c 5d-5-g0c §0T X 99°¢ 0T X 28'x gOT X oL'T Sd-9-T0Z
#OT X £E°9 0T X £2°¢ gOT X ©9°¢S 8d4-5-L0¢ 0T X 05"t OT X L5°¢ 0T X ST Sd-g-002
40T X g4° % 0T X L6°¢ 0T X 9K°T Sd—-8-90¢ ;0T X £9°¢E 0T X T6°¢ 40T X L9°6 Sd-g=-66T
UNd SUldnp UoNe3 SoLdmeg
0T X 66T 0T X 67°'T g0T X Te'1 £-5-Q6T
gOT X 26°T g0T X " T g0T X cc'1 £-5-16T
SJI90%8J43 JO UOT3TpPP® J931J8 Jnq uni ©3 JoLJdd ueye] SoTdweg
20T X 66 > 40T X 0°T> 0T X €°f >  E-8-%6T 0T x L6 0T X L'yg> ¢0T X 6T°T > T-€-96T
0T X #°T > x0T X 9°t > 0T X b.:.w £-5-€61 20T X m.m.w g0T ¥ m.:.w ¢0T X 00°T > T-g-66T
0T X %79 > x0T X g°t > g0T X 0O°f > G-8S-06T 20T x m.m.w ¢0T X L°9 > 0T X €8°Q > G-g-261
0T X &°T > 40T X 9°¢ > ¢0T X 6°6 > G-S-68T 0T X #'f > 0T X T°6 > ¢0T X €0°T > G=g-T6T
und 09 JOTJd Usye]} SoTdweg
Anfls1m\mpzfioov {uTw-F/s3Uunod) fisflnTw\mpqzoou g9P00 AQHfllm\mpflfiouu hanTm\mvnfioov (uTw-3/squnov) goPO2
T, 0F N;¢z 103 fl,e7 I0J oTdureg 17,6 *0F 0,¢z 103 n £z I0J oTdureg

STSATBUB PITOS

ETafATBUR UOTINTOQ

stsAiBUB PIIOY

STSATEUB PITOS

stsiTBUR UOTINTOYS

stsATeus PITOS

 

{=~OWSIL UNJI WOIJ PauT®3qO ®I8P JUTIIUNOD

"£-Y 9TaBL
39

*oTdwss W8salS JUTMOTY = 8i SGL = § L = ¥ ‘€L = € %28l = g $TL = T ‘UTdiao ordums = J pus

 

 

 

 

 

 

(3188 = g ‘uyjnusTq = ) STAdwes UT TBTLSIBUW = g ‘Jaqunu sTduss = y aaeym ‘p-g-y 07 Julpuodssaaoo spoo ® Aq pejvudrssp sT oTdwms Yowd,
g0 ¥ S'E > - 0T % T'T >  $=8-093 40T * 002 -- gOT ¥ HE"T S—€-gse
gOT * 9'c > - 0T X 2°6 > G=5-642 yOT x Lf*g _ - gOT ¥ TL*T G~d-L52
gOT X LE'T - gOT ¥ OT'T {~-g-94¢2 40T X 08T - gOT X 00°T g-d-gsc
gOT X 61°1 - gOT X 84°T f~5-562 40T X mm.fi.l - 40T X ow.b.l g-t-16e
gOT * 6T'E - g0T X ge°t E=-g-152 20T X §°F > - 0T X mo.fi.w T-g-042
0T X 96'2 - g0T X TT°E £-g-€4e 20T X 0'ff > -— 0T X gE°T >  T-g-6%2

“UTWE A99.J8 UaiB] SoTdusg
40T ¥ 9£°6 - 90T X 0%°G Sd=-5-ghe 40T X 26°¢ gOT X TE'S gOT X TH'T Sd-g-The
gOT X 36°1 g0T X go°g 90T X 6L°T ' gd-g-Lhe 40T X #0°€E gOT X L9 0T ¥ L9°T SI-g-043
0T X 62'28 0T X 10'€ g0T X 96°T Sd=-5-9N2 4OT ¥ gL-2 0T X EQ'f gOT X SE'T Sd-g-6Eg
gOT X £f'e g0T ¥ 0T'¢ 0T X gL'T Sd-g~5he 40T X 6g°2 0T X T19°% 0T X GE°T Su-4-RL 2
gOT X 19°'¢ gOT X 00°2 0T X 28°1 gd-5-fihe 40T X 46°2 gOT X 09°% gOT ¥ ES°T Sad-g-LEe
gOT X 6L°T g0T ¥ 00'¢c g0T X #9°T SU-8-Ehe 40T X 29°¢ gOT X ET'H gOT X OE°T Sd~g~-9E2
0T X Li'g g0T X ET°¢ g0T X EQ°T Bd=G-2chg 40T X 4g°T 0T X 1g°¢ gOT x 70°T Si-g-GEC

NI JUTINP Us¥e] SsTduwRg

¢OT X @gc’¢ g0T X gh°t g0T x LG*2 E=g-fEc
gOT X 0T*2 90T X fif°t 90T X %5°'2 E-g-tte

S490HI] JO UOTJIPPE 48748 I00 und 07 JOTJdd uaie] SoTdWegy

0T X #°T > - 0T X 0°9>  £-5-0Ee 0T x 0TS -- 20T X T6°S > T-H-ZER
0T X 2°T > - g0T X 6% > £-5-622 0T x T°T > -- 20T X mm.:.w T-g~1te
20T X m.w.w - g0T X m.w.w ¢-g-92g 20T X m.mum - Z0T X 6f°g > S-g-gee
0T X 9°'T > - g0T X #*L > G-g~Scc 20T X w'T > - 20T X 19°'G > S-g-Le2

und 07 JI0TJIU Ul38q sajdwsg
(utw-3/83UNco) (uTw=3/s3tmoo) (uTw-3/83UN00) 59P0OD (uTw-3/sq.unoo) (utw=-3 /89UM00) (utw-8/s3unoo} 5oDOD
A7, 65 L0F gz I0F N7 70J ardureg a7, 4OF N,pg *OF ¢z 4OF a1duegy
STSATBUB PITOQ sTsATBUE UOIINTOY sTsdTeUB PTTOQ gsTedTRUB PITOY gTsiTBUB UOTANTOQ STSATBUB PITOQ

 

G-0OWSL Urnd WOIJ PAUTBIGO BBY BUTIUNOY =Y STUBL
%0

puw mfivflmm = g ‘yinwstq = q) aTdums UT TBTJ93BW

*o1dues weadds BUTMOTI = Sd 6L = ¢ {l = ‘€L =€ 2l =2 'TL = T ‘urdixo ordums = )
= g ‘aequmu sTdwes = Y 2J9UYsA ‘)-g-v 03 Jurpuodssaacc 9poos B £q parsBudisap sT oTdmss nodmd

 

 

 

 

 

 

0T X 0°6 > 0T X L'T> 0T X T°g>  6-S-%62 40T X Gi°% gOT X TE"E 0T X €€°T ¢-g-262
g0OT ¥ L'G > 0T X @°T > 0T X 2°6 > G-g=£62 40T X 9E°f 0T X TR'E 0T X 60°T ¢-g-T162
40T X 68°8 cOT X 0g8°¢ ¢OT X 06°€ t~-S-062 0T X 65°4 gOT X TT°¢ OT X ET°T c—d-98¢
«OT X 0%°8 0T X STy gOT X ®0'f h=S-69¢ #OT X £€°6 0T X ®E°E ¢OT X LO°T c-d~5Q2
0T X LG°¢E 0T X RE'T g0T X 0¢°T £-8-gg<c ¢0T X fif°¢ 40T X 659 yOT X TS'e T-g-%g<c
0T X Ly°g 0T X 09°T 0T X CE°'T £=5-182 ¢OT X £L°¢E 70T X 6L°9 x0T X ge°¢ T-4-£Q2
unI I91JB UaNB} So[AWBS
¢OT X Lo°2 0T X Of'€ cOT X £9°% Sd-s-2ge 40T X 68°Q cOT X €5°y cOT ¥ €6°T Sd-g-612
0T X £0°¢ ¢0T X 99 0T X €g°% SI-9-TgC 0T X T0°6 0T X L2'q ¢OT X %SG T Sd-g-1Le
0T X 681 0T X 921 c0T X 00°6 S4-8-08e {07 X 6L°L cOT ¥ 9E°% ¢OT ¥ 66°'T sd-g-gle
gOT X 9T°¢ gOT * gty gOT X £0°f Ssd-g-6L2 x0T X T0°Q ¢OT X LT q gOT X £1°T Sd-g-cle
¢0T X TT'e ¢OT X £g°¢ cOT X 667K Sd-s-8le 50T X #5*L 0T X 9T°% OT X 62'T Sd-g-T.L2
g0T X 0t gOT X GG q gOT X £9°% Sd-s-L12 40T X 0L°9 0T X £9°¢ gOT X £T°T sd-4-0.2
0T X GL°T gOT X E4°E gOT X TT'¢ Sd-5-9.2 x0T ¥ 8L°9 0T X 6E°2 50T X 69°1 Sd-g~69¢2
unJgd SuTJnp uUa¥ed SoTdWeg
¢OT X 9T°€ 0T X HE"T 90T X 10°T £-5-g9¢
0T X 62°€ 90T X 62°1 g0T X cT'T £-5-L9¢
SJ908I9 JO UOTJTPP® J123Je 4nq UnNI 09 J0TId u=aye] SoLdueg
0T X T'2 > 0T X H°T > 0T X 9°6>  €-5-t92 {0T X g6 40T ¥ Li°¢ 40T X 66°T T-4-992
g0T X O°T > K0T X ®°T > 0T X &°T > £—-8-£92 0T X 96°¢ 0T X B8L*S 40T X G6°T T-g~592
0T X €°T > x0T X £°'T > 0T X T°9 > ¢~5—-09¢ ¢0T X TG°C x0T X 6§°6 40T X LE°2 G-g-2c92
0T X g'¢c > 70T X £ T>> 0T X 2°9 > g=s-66¢ - 0T X 2£°6 40T X SL°T G-g-192
und 09 J0TJId Us¥Bq SsSTAWEBG
(uTw-3/s3unoo) (uTw-3/s3unoo) (UTw-%/s3unod) »oPOO (uTw-3/squmoo) (uTw-8/squnod) {(UuTw-3/53UNo0) g°POD
17, IOF Nygz 20F Ng7 <OF aTdusg IZ,¢ 10 N,cqz ¥0F N,¢7 40F aTdmeg

STsATBUR PITOS

STSATBUB UOTINTOSG

sTsATBUB DPITO]

s1sATBUE PTITOQ

STSATBUB UOTINTOg

sSTsATBUB PITOQ

 

9=DWSL NI WOXJ paulelqe B1Bp JuTjuno)

*G-¥V

aTqEL
= 5d
ut Tetaajlew

mm.”m_ =

S

P = ¥ {EL

*{ § A9 §6°LOT) N,¢p FO uOTIdAOSAER-FTas 103 PIIDVAIOD goTdwes yanustd

£ ‘2L
g f{asqunu sTdwes = Y saoum ‘D-g-y¥ 03 bBurpuodsaaaoo spoo € Aq pejeubisep sT ordues yoed,,

m

g {TL =

1 futbtao sydwes

D pue !(31es

q

.mfimfimw weaxls butmoTs

g YISty = )

ta1dues

 

~

 

 

 

 

 

WOT X L°T> (0T X €°G > §-5-9vE ¢OT X BS°S c0T X G0°¥ S-B-VPE
0T X L°8 > g0T X 8'6 > S-S-GPE gOT X 5075 ¢OT % BE'V S-g-EbE
g0T ¥ 6T°2 0T X ¥9°T b-5-2ZveE gOT X 98°¢ gOT X 66 Z-g-BEE
—— - b-S-TPE gOT X BL'E gOT x £L°C C-d-LEE
- -- £E~S5-0VE gOT X B8T"C 0T X L¥°1 T-8-9E¢
- - £-5-6L¢€ gOT x g2 ¢ g0T X 06°T T-g-5¢¢
unI J8jje usyel SaTdwes
¢OT % LT°S gOT x 2e°E Sd-H-LZE
g0T X 06°€ g0T * €6°2 Sd-5-VEE gOT X 50°Ss gOT X ¥6°E SI-g=-9Z¢
g0T * 96°¢C g0T * ZB*C Sd-S-ELE gOT X 90°§ 0T X 9G°F Sd-g-6Z¢
g0T X L6°E 90T X* 68°C 5d-S-CEE 0T ¥ £0°S 0T X 86°E SI-H-vCE
g0T * 92°¢ gOT X SL°2 S3-S-TEE gOT * 1I8°V 0T X €T°E 5J-g-E£C¢
gOT * ZE°E g0T X 68°2 S4-5-0t¢E gOT X EE°V 0T X 66°¢ Sd-g-E2¢
g0T X 68°¢C g0T X TT°C 8d-5-62¢ gOT * T6°¢ gOT X v9°C SI-g-TCE
Und butanp Uede3 Sefdues _
g0T X T9°¢ g0T X G6°¢ £-5-02¢
¢0T X 97°6 g0T X S6°¢ €-S-6TE
SI190va] JO UOT4IPpPE A93JFF 3ng unit o3 I0Tad usyel saTdues
g0T X S°6 M 0T X 9°L M £-S-9T¢t {O0T * LV'9 x0T X §9°6 T-2-8T¢E
20T X 8°T > g0T ¥ 979 > £-S~5TE 40T ¥ 6L°9 0T X S6°V T-9-LTE
Q0T X &°T > 0T X T°5 > §-S~-CT¢E 0T X LL™9 40T X v9°V c-g-vTE
ZO0T ¥ 8°T7 > 0T X T°% > G=-5-TTE 40T X 6£°9 40T X 06°S S-g-tTt
una o3 I0Tad uslel Selawes

(utw-5/s3UNnoo) (uTw-56/83UN0D) 29POD (uTw~-H/s3UNOD) (uTw-56/53UNO2) pdP02
Nyge *93 N,ez 103 a1duwes N pg 393 N, gz 393 arduwes

stsAieur UOIINTOS

sTsATeu® pPTITOS

sTsATeUR UOTINTOS

q

sTsATeue pTITOS

 

L—-OHSL una

woxy peuteldo ejep buTjunod

“9-¥ STqRl
ho

* oT7dures
I = ‘€L = € ‘2L = ‘1L = T ‘ui@tao os7dus = 5 pur (3TES = § ‘UINWSIq = g)

uresa3s FUTMOTI = Sd ST = § c
tasqumu oTdures = ¥ SJ9UM ‘)-g-V 0% Jurpuodsaxioo apoo ® £q pelreUdTISSp sT aTdwes sommm

cordures UT TBTJI27BW = g

 

 

 

 

 

 

0T x 2> SOT X 2°T > G-S-T6€ ¢OT ¥ 99°€ cOT X 6£°T G-g-68€
K0T X #°c > 40T X 0°T > 6-5-06¢ 0T X 6T°€ 0T X 22°T C-€—QgE
0T X GE°2 o0T X 08" T #=-5-19¢ #0T X L0 70T X 94T c—gd-£g¢
0T X G1°2 0T X LL°T 1=5-9g¢ x0T X 2276 x0T X T6°T c—d-2g¢
g0T X gf*¢c 0T X 0T°¢ £-5-4Qt qf0T X gn°T ¢0OT X I8° 1 T-9-18¢
0T X [9°2 g0T X g2°¢ £—S-fg¢ 20T X LG°T ¢OT X gL°% T-9-0Q¢
una Js1Je usayeq Sotdured
90T X ft'¢ g0T X £f°¢ mmlmlmwm cOT X 90°T 70T X 90°¢t SA-g-cLE
90T X QT*2 90T X 06°T Sd-S-g.LE +0T X 65°6 40T X 66°2 Sd-g-TLE
0T X 06°T gO0T X #6°T Sd~-S-LLE 0T X 69°Q x0T X 98°¢ Sd-g-0L¢E
0T X 6c*c 90T X g0°¢C Sd-S=-9.LE x0T X £0°8Q 40T ¥ £°2 SA-4-69¢
0T X "¢ 0T X 9L'T SA-S~GLE 70T X 88°¢Q x0T X 99°¢ Sd-4-39¢t
g0T X ¢0°'T 0T X 676 SA-S—HLE +0T X Tc°¢ 0T X 8T 6 Sd-g-L9¢
cOT X ¢L°¢E ¢OT X #0°¢ Sd~S—ELE x0T X 00°T ¢0T X xL°2 Sd—~g-99¢
unJg SUTJnp us¥Bl] SaTdueg
90T X 16°¢2 g0T X If°¢ £-5-69¢
90T X 68°C 0T X ge* e €-8-19¢
JI90®JI] JO UOTZFTDPPE JI89IB 1ng UNJI 07 JOTJAA U3¥YE] SsSTAWERG
x0T X w.fi”M ¢0T X w.:”M £-8-TOE ¢0T X 90° % 0T X G2°'T T-€-¢9o¢t
x0T X 9°'T > ¢OT X w.j.w £-S-09¢ ¢OT X TS84 ¢0T X NN.H.I T-€-29¢
0T X 6°Q > ¢0OT X g°f > G-g=LGE ¢OT X €€ 0T X #°T > G-g-65¢€
c0OT X g°t > G—8=94¢E ¢OT X QTN 0T X L°T > S-g—-gse
unJg 0% JA0Tad Us¥eq SssTdweg
(uTw-3/53UNnod) (uTw~-3/s3Unod) poPOD Aflflalw\mpqfloov (UTw-3 /83uMoD) gSPOD
N,¢7 IO Ngg I0J aT1dureg 0;¢z 10 N,gz I0 aTdwreg
sTsATBRUB UCTANTOS sTsATeU®B PITOQ sTsATBUR UOTANTOG STSATBUB DPITOQ

 

Q-OWSL UNJd woJgj PauTejqo BIBP SuTjuno) *L-¥ STqEL
43

. *o1dwss wBaaANS
FUTMOTE = Qi SGL = § SHi = ft SEL =€ ‘8L =2 ‘T = T ‘furdrao ordwes = J pus (4788 = g ‘YINUWETQ = f)
“oTdurss UT TBTJA9BW = g (Jdaqunu ordums = Yy adeym )-g-y 01 Furpuodssdaon opoo v Lq pejvudrssp st ordwss sodad

 

 

 

 

 

 

40T X :.H“M g0T X :.m_M G—-g~-Lay gOT X G°g cOT X 6T"°T G-g~62h
ZOT X G°T > 0T X T°L > G-8-9¢t cOT X 62°¢ gOT X g0°T S-g-w2t
0T X 69'1 gOT X #5°T h—8-£3h gOT X 9g°¢ gOT * EX°T - c-g-6Th
gOT X 18°T gOT X #G*T h-g-gch gOT X G0°E gOT X LT T-g-8TH
cOT X 69°8 cOT X 2T'. g-g-Teh ¢0T X T°E > Z0T X 626 T-a-LTH
0T X 2L°g cOT X 68°9 £-g-024 g0T X L€ > z0T x ¢6°g T-g-9THk
unJg J99J8 uayw] ssTdweg
gOT X 69°1 0T X 69°T Sd~-8=-5Th gOT X #6°4 ‘ 0T X 262 SI-g-80%
gOT X g8°T 0T X %L°T Sd-g-HTH cOT X TL°G gOT X gg°'<¢ Sua~-g-1,0%
0T X 6L°T cO0T X g2'c S=-S—ETH gOT X 9E* 4 gOT X TE°E Sd-g-901
0T X 66°T cOT X 96T . Sd-8-cTh cOT X 62°% gOT X gl-g Sd=-g=-50t
cOT X 00°¢2 cOT X #0°¢ Sd-S-TTH gOT X ¢€°9 0T X 96°2 S-g-t01
gOT X 9T'¢ gOT X 90'¢c SJd—-8-0TH gOT X KLy gOT X GT°E Sd-d-E01
0T ¥ 76°T gOT x LT'2 Si—g-601 gOT X 02°% 0T X 60°€ Su-g-g0t
unJgd JUuTJUp UsyB] SBoTAWRY
0T X G2°€ - gOT X QT'¢ E-g-T0H
g0T X 68°¢ 90T X LT'2 €-8~001%
Jd90BI9. JO UOTETRPR J437JB Ing und 07 JO0TJId USYBT EITAWET
wOT X &°'T M 0T X 679 M g-g-L6t 0T X 0°¢ M Z0T X £4°g T-4-66€
40T X E°T > g0T X p.w.w £=-9-96¢ 0T X m.m.w 20T X :m.».l T-g-g6E
x0T X €°T > g0T X 8'9 > G-8-E6E ¢0T X 9°¢ > 20T x 2S°'g > G—g~G6t
xOT X £°T > 0T X g'9 > G-g-26¢t 0T X 2°¢ > 20T X g L > G-g~H6E
UL 09 JOTJ4d Ua¥B] SaTuweg
(uTw-2/83uUnoo) AanTm\mpnfioov 5oPOD (utw=-5/89unoo) Anflslm\mpmsoug goPOD
Ngg 40F Ngz 40 aTdureg Ngg 40J [ypg 40F aTdwsyg
sTsATBUB UOTINTOY STSATBUB PTTOQ grsATBUB UOTINTOL sTsATBUB PITOQ

 

6-0WSL UNJ WOIJ PauTelqo ©gep JuTqunoy 'g-y STUBI
Ly

APPENDIX B.

Calculation of Mass Transfer Coeffilicients

For a flow-through, continuously stirred contactor at steady-state

conditions, a mass balance on the salt phase yields:

= + -
F.C, =FC +J, (B-1)
where

F, = flow rate of salt, cm3/sec,

Cl = tracer concentration in salt inflow, units/cm3,

02 = tracer concentration in salt outflow, units/cmB,

= rate of transfer of tracer across the interface, units/sec.

Expressing the rate of transfer across the interface as the product
of an overall mass transfer coefficient and a driving force times the

area available for mass transfer yields:12

= - B-2
Jg=x/[c_-c /DA, (B-2)
where
= + -
1/Ks 1/ks 1/ka . {B-3)

and

KS = overall mass transfer coefficient based on salt phase, cm/sec,

ks = individual mass transfer coefficient in salt phase, cnm/sec,

km = individual mass transfer coefficient phase in metal, cm/sec,

D = distribution cocefficient = ratio of concentration in metal

phase to concentration in salt phase at eguilibrium,

moles/cm3
3 2

moles/cen
k5

C
m

A

tracer concentration in metal outflow. units/cms, and

. - 2
interfacial area, cm .

Taking an overall mass balance results in:

= ' ' =l
C.F + FF, CF, +CF,, | (B-L)
where
02 = {racer concentration in metal inflcw,-units/cm3, and
F, = flow rate of metal, cmS/sec.
Ir C, = 0, Eq. (B-L) can be rearranged to give the four following

relations:

= - ' (B-5)
C,F, = CF +CF, , -
Fy
¢, = c.+c |} (B-6)
i
Fy
Cs = C_ - Cm 71 and . ' (B-7)
]_ .
F F
c = ¢ =-cl]. | (B-8)
m o o

Combining Egs. (B-1), (B-2), and (B-8) yields:

K CAfF ) EcCA([F Y\ o
Pl = PO P ECA-TEO\R )T T (5, ) (5-9)

no

which can be rearranged to give:
KSA
+-..-......-._.
¢ o F, + -5 (F/F,) - _ (5 10)
s 1 K A_( ) -

 

8 Fl
Fl * KSA'+ D 'fig

Combining Eqs. (B-1), (B-2), and (B-T7) yields:
L6

 

 

 

F2 KSCmA
= - - —_—— B—
FiCy FiC; - CF, + K CiA - (K C_A) . 5 , (B-11)
which is rearranged to give:
KSA
C/Cy = F, (KA (B-12)
Fo v (KA 5=+
1
Combining Egs. (B-1), (B-2), and (B-6) yields
K CmA
F,C, +CF, = F.C +KCA-\—F ; (B-13)
which is arranged to give:
KSA
c_/C = e (B-1h)

m s KSA
Fo # _5-)

Rearranging Egqs. (B-10), (B-12), and (B-1k4) gives three expressions for

the overall mass transfer coefficient in terms of the measured quantities

 

Cl’ Cs’ Fl, F2, D, and A:
CS
LAt
v 1
3
8 s\ af%\/f1\ aff1
e )*ole JI5 ) - olF, (3-15)
1 1/\ "2 o
cll'l
g )
~ (B-16)

 
7

K= . | | (B-17)

The above eguations can then be used to calculete mess transfer
coefficients from experimental results (i.e., the ratio of tracer

concentration in any two of the salt of bismuth flows).

Within experimental error, the distribution coefficiefit D can be get
as desired. To minimize effects of uncerteinties in the velue of D on
the calculated value of the overall mass transfer coefficient, D should
be made fairly large. For the values of concentrations and flow rates
used in these experiments, the terms which contain D in Egs. (B-15),
(B-16), and (B-17) are less than 5% of the values of the other terms
for values of D greater than 20 and can be neglected with little error.
By assuming that the terms that contain D can be neglected, Eqs. (B-15),
(B-16), and (B-1T) reduce to

 

K. = (B~18)
kg = (B-19)
F. C
_ 2 m
K., = 1= G (B-20)
L8

Uncertainties in the distribution coefficient do not affect the
accuracy of the overall mass transfer coefficient. However, as shown
by Eq. (L), when D is very large, the overall mass transfer coefficient
is essentially the individual salt-phase coefficient, since resistance

to mass transfer in the metal phase is comparatively negligible.
49
APPENDIX C.

Calculation of Diffusivity of UF. and Zth in Molten Salt

3
| The diffusivities of UF; and ZrF) in molten salt (72-16-12 mole %)
LiFmBngnThFh were estimated from an empirical equation developed by

Wilke,13 which was based on the Stokes-Einstein equation:l

o () e
E%B = T7.bL x 10 ——7:?;?:3—— s (c-1)
A
where
213 = diffusion coefficient of specie A in solvent B, cm?/sec,
?B = assoclation parameter for solvent B,which is equal to 1.0
for an unassocisted liquid,
T = temperature, °K,
Mz = molecular weight of solvent B, g/g-mole,
%A = molar volume of solute A. cmB/g—mole,
n =  solubion viscosity, cP.

This equation is good only for dilute solutions of nondissociating solutes;

1k

for such solutions the error is within + 10% of the true value.

Semple caleulation

The diffusivity of UF3
mole %) at 600°C was calculsted as follows:

and ZrFLL in molten LiF_BeFQ-ThFh (72-16-12

T = STSOK,
n = 11.81 cp,’
M., = 63.16 g/g-mole, |
%fiFh = U6.4 cnd/g-mole at 600°C.1° This value was assumed to
apply to both UF3 and Zth,
wsalt assumed = 1.0,
_ 317
psalt = 3.35 g/cm”.
50

Substituting the above values in Eq. (C-1) results in:

 

 

1/2
& = T.4 x 10 = 4,45 x 10" em”/sec.
UF3 or Zth ~ galt (11.81) (h6.h)0'6 see
-2
8o = 1]2 _ (ll.8§ x 10 ° g/cm Zec)2 - 7920.
P (3.35 g/em™) (L.L5 x 107 em™/sec)
51
APPENDTX D.

Entrainment Studies

One hydrodynamic test was performed to determine the smount of fluo-
ride salt that might be dispersed and entfained.in the bismuth and the
amount of blsmuth that might be digspersed and entrained in the fluoride
salt effluent streams of the contactor at various agitator speeds. This
hydrodynamic run was performed with salt and bismuth flcw rates of 150
cc/min and 1L0 ce/min, respectively. The agitator was operated at three
different speeds during the run, 250 rpm, 310 rpm, afid 386 rpm. At 250
rpm gnd 310 rpm three sets of unfiltered salt and bismuth sampies from
.the-contactor effluent streams were taken ot L-min intervals. Three
sets of unfiltered effluent samples were also taken with the agitator
operating at 386 rpm, but the samples were taken &t P-min intervals. The
sample capsules were cleaned and the contents of each sample were removed
as described in Sect. 3.5 of this report. The contents of each sample
were inspected for evidence of gross entralnment of one fihase into the
cther., No such evidence of entrainmeni'was found. Results of chefiical
analysis of the bismuth and salt samples for beryllium.and,bismufih con-

tent. respectively, are given in Teble D-1.

The flowing~stream bismuth samples from runs TSMC-5, -6, and -9 were
also analyzed for beryllium content, and the results of these analyses
are given in Table D-2. Runs TSMC-5, -6, and -9 were performed with agi-

tator speeds of 124 rpm, 180 rpm, and 24l rpm, respectively.

The bismuth concentration meagured in the salt samples tzken during
the hydrodynamic run (Table D-1) shows a general decrease with increasing
stirrer speed, with very low values occurring at the highest stirrer speed.
It also seems evident that the bismuth concentration in the salit phase may.
have been a function of the run time. After the fourth sample, the bis-
muth concentration in the salt samples reamined at a relatively_constant
value of 50 + 11 ppm; this is quite different from the values reported

for the first four samples,which ranged from 1800 ppm to 155 ppm.
52

 

 

e Svv 0S 9ty 98¢
e 7A"44 SLT seEd 98¢
i74°] evd OTT PEY 98¢
e vy - £eEv 0Tt
€S 1vd 0T6 cev 0T¢
oLC o¥¥ S8 Ted 0T¢
esT 6ev S1c oev 0S¢
q0¢ 12184 QT 6Cv 0s¢
008T LEY qTC 8cv 0S¢
(udd) aTes *ON (udd) yanuwsiq *ON (udx)
ut ynwstyg oTdues 3Tes uT umtiTAiedg oTdures yaInwstyg poads I03e3THY

 

unx oTweulpoxpdAy 9yl HutaInp usNel soTdues
YInNWS I pue 3Tes 9yl JO STsATeur woxF sS3jtnssy *I-d ST9el
23

Table D-2. Results of analysis of flowing stream bismuth
samples for presence of beryliium

 

 

Run Agitator speed Bismuth flowing stream  Beryllium in
NO . (rpm) sample No. - bismuth (ppm)

TSMC=5 124 235 8l
124 236 131

124 237 86

124 238 ig1

124 ' 23¢ o 162

124 240 : 111

124 241 : i32

TSMC-6 180 269 264
180 270 315

180 271 | 118

180 272 125

180 : 273 86

180 274 278

180 275 113

TSMC-9 242 402" | 85
' 244 403 118

244 AQ4 41

244 405 | < 10

244 406 26

244 407 56

244 408 104

 
54

These results are significantly higher than those of Lindauer18 who
saw less than 10 ppm of bismuth in fluoride salt that was in contact with
bismuth in several different contacting devices. It is likely that sample
contamination is a contributing factor to the high bismuth concentrations
that were measured. Three possible sources of sample contamination have

been reported:l9

(1) sample contamination during sampling by withdrawing the samples

through a sample port that has been in contact with bismuth;

(2) sample contamination during sample handling and in the analytical
laboratory by the use of equipment that is used routinely for

bismuth analyses;

(3) sample contamination from a low-density bismuth material that

may be floating on the salt surface.

The beryllium concentration in the bismuth samples taken in the hydro-
dynamic run (Table D-1) and in runs TSMC-5, -6, and -9 (Table D-2) show
both high and low values with no discernible dependence on agitator speed.
Based on previous experiments with water-mercury and organic-mercury systems,
one would expect entrainment of the light phase into the heavy phase at
an agitator speed of about 170 rpm.
55

APPENDIX E.
Tocation of Originel Data

A1l data and operating records for the salt-metal contactor studies
are recorded in log book No. A-6886-G. Records for the Ffaeility up to
the time at which the stirred interfece contactor was installed are con-
tained in log books numbered: A-56U9-G, A~5965-G, A-6210-G, A-6LB2-G,
and A-6T722-G, and are available from the authqr.