5

T

S

2

e

o

 

S

s
i

w

 

I

|

|

 

I

|

S
....n..&uuv...rwru"..-....r.. .,u.flu..r

R
- .,“..;Nu"..u(.m.m..mu..mw.}.. S s

|

|

|

ADGE N@Igmmflglfl_flmEOHATwY LIBRARIES
3 4455 0550621 7

      

 

paKE

.""...
2 o
SR

 
 

Printed in the United States of America. Available from
National Technical Information Service
1J.S. Departiment of Commierce
52865 Port Royal Road, Springfield, Virginia 22161
Price: Printed Copy $5.45; Microfiche $2.25

 

 

 

This report was prepared as an account of work sponsored by the United States
Government. Neither the United States nor tiiec Energy Research and Developiment
Administration, nor any of their employzes, nor any of their contractors,
subcontractors, or their employees, makes any warranty, express or impliad, or
assumes any legal liability or responsibiiity for the accuracy, completeness or
usefulness of any information, apparatus, product or process disclosed, or represents
that its use would not infringe privately owned righis.

 

 
ORNL-TM~4863
UC~76 — Molten Salt Reactor Technology

Contract No,'W—74OS—eng—26

CHEMICAL TECHNOLOGY DIVISION

ENGINEERING DEVELOPMENT STUDIES FOR MOLTEN-SALT
BREEDER REACTOR PROCESSING NO. 19

Compiled by:

J. R. Hightower, Jr.

Other Contributors:

C. H. Brown, Jr.
W. L. Carter
R. M. Counce
J. A. Klein
H. C. Savage

JULY 1975

NOTICE This document contains information of a preliminary nature
and was praparad primarily for intarnal use at the Qak Ridge Mational
Laboratory. It is subject to revision or correction an'd therefore does
not represent a final report,

OAK RIDGE NATIONAL LABORATORY
Oak Ridge, Tennessee 37830
operated by
UNION CARBIDE CORPORATION
for the
ENERGY RESEARCH AND DEVELOPMENT ADMINISTRATION

OAK RIDGE NATICNAL LABDRATORY LIBRARIES

RURBATINARIEE

3 445k 055021 7

 

 

 

 

 

 
ii

Reports previously issued in this series are as follows:

ORNL-TM~3053 Period ending December 1968
ORNL~-TM-3137 Period ending March 1969
ORNL-TM~-3138 Period ending June 1969
ORNL~TM-3139 Period ending September 1969
ORNL-TM~3140 Period ending December 1969
ORNL~-TM—-3141 Period ending March 1970
ORNL-TM-3257 Period ending June 1970
ORNL~TM~3258 Period ending September 1970
ORNL~-TM-~3259 Period ending December 1970
ORNL~TM-3352 Period ending March 1971

ORNL-TM~4698 Period January through June 1974
iii

CONTENTS

| Page
SUMMARIES - - - - - . - - . . L » - - - . » . - - - - - . - - e - V
INTRODUCT ION - . . * L] . . . . . I‘ . - . - * . - . - » . - - - » l
DEVELOPMENT OF THE METAL TRANSFER PROCESS. .« « + « « &+ + o o« « » 1
2.1 Examination of MTE~-3 Equipment and Materials. . . . . . . . 2
2.2 Installation of Metal Transfer Experiment MITE-3B. . . . . . 6
2.3 Design of the Metal Transfer Process Facility . . . . . . . 9
SALT-METAL CONTACTOR DEVELOPMENT: EXPERIMENTS WITH A MECHANICALLY
AGITATED, NONDISPERSING CONTACTOR USING WATER AND MERCURY. . ., . 11
3 - 1 Theory . . . » » » . - - » . * - . - - . » - - » o . . * - . 11
3.2 Experimental Apparatus. . . . ¢« . 4+ + + « « 4 e+ . o« s . . 18
3.3 Results and Conclusions . . « « 4+ ¢ &+ ¢ & 4 & o + « « +» « o 18
SALT~-METAL CONTACTOR DEVELCPMENT: EXPERIMENTS WITH A MECHANICALLY
AGITATED, NONDISPERSING CONTACTOR IN THE SALT-BISMUTH FLOWTHROUGH
FACILITY &« « 4 v » v = o o o« o o o o o o o o o 2 o 2 o s o« o « 21
FUEL RECONSTITUTION DEVELOPMENT: DESIGN OF A FUEL RECONSTITUTION
ENGINEERING EXPERIMENT ¢« & +v &« ¢ + o o o s o o o o = 2 o« o« o« « o« 38
R—EFERENCES - - - . - - . - . - - - . > - a * * - - - " - - » - . 4 3
SUMMARIES

DEVELOFMENT OQF THE METAL TRANSFER PROCESS

We have continued fabrication and assembly of the new carbon steel
vessels for metal transfer experiment MTE-3B; during this report period,
the vessels were completed and installed in the metal transfer experiment
facility in Bldg. 3541. The vessels previously used in experiment MTE-3,
along with the salts and bismuth they contained, were removed and sent to
the burial ground for disposal. We are renovating the metal transfer
hood to improve operating accessibility and assure that all egquipment

is in good operating condition.

Examination of the vessels. and analyses of the salt and metal phases
from the previocusly operated experiment MTE-3 have been comp leted,
Limited visual examination indicated that the internal surfaces exposed
to salts and bismuth were in excellent condition. Failure of the
oxidation~resistant protective coating on the external surfaces allowed
significant oxidation of these surfaces at the 650°C operating tempera-
ture, but was not extensive enough to affect the vessel integrity. A
different protective coating with superior air-oxidation resistance was

applied to the MTE-3B vessels.

X-ray fluorescence analyses of the Li~Bi phase from the rare-earth
stripper at the LiCl--Li~Bi interface contained significant amounts of
iron and thorium. Scanning electron photomicrographs of this interface
at magnification of 20, 100, 500, and 2000X were taken. No identifiable

film or foreign material can be seen at the interface.

Review of the design criteria for the Metal Transfer Process Facility
has been initiated. This facility is to be located in Bldg. 7503 (MSRE

site).
vi

SALT-METAL CONTACTOR DEVELOPMENT: EXPERIMENTS WITH A MECHANICALLY
AGITATED, NONDISPERSING CONTACTOR USING WATER AND MERCURY

Five runs were made in the 5- by 7-in.-Plexiglas contactor with phase
volumes, agitator speed, and initial mercury-phase zinc concentrations
held constant at 1.8 liters, 150 rpm, and 0.1 M, respectively. The ini-
tial aqueous-phase lead concentration was varied from 0.02 M to 0.10 M
to determine which phase contains the limiting resistance to mass transfer
and the concentration of lead ions in the aqueous phase at which the con-

trol of mass transfer changes from one phase to the other.

A model was developed for this system under the assumption that the
reaction takes place entirely at the water—-mercury interface and is
instantaneous and irreversible. It was found that for the conditions
under which these and all previous runs were performed, the resistance

to mass transfer was not in the mercury phase as was previously believed.

SALT-METAL CONTACTOR DEVELOPMENT: EXPERIMENTS WITH A MECHANICALLY
AGITATED, NONDISPERSING CONTACTOR IN THE
SALT-BISMUTH FLOWTHRCUGH FACILITY

A 6-in. diam low-carbon steel stirred interface contactor has been
installed in the Salt-Bismuth Flowthrough Facility in Bldg. 3592, which
was previously used to study salt-bismuth flow in packed columns. Six
tracer runs have been completed to date using 97Zr and 237U tracers. The
stirrer rate was varied from 121 rpm to 205 rpm.

Results from the first six runs indicate that the salt-phase mass
transfer coefficient based cn 237U counting data is 37 + 3% of the value
predicted by the Lewis correlation for runs 1, 2, 3, and 5, and is 116
+ 10% of the Lewis value for runs 4 and 6. The relatively high values
for the latter two runs are probably due to dispersal of salt into the
bismuth. Experiments with water-mercury and water-methylene bromide
systems support this belief. The mass transfer coefficients based on
97Zr counting data are felt to be less reliable than those based on 237U
because of the inability to correct for self absorption of the 743.37

keVv 8— in the solid bismuth samples.
vii

FUEL RECONSTITUTION DEVELOPMENT: DESIGN OF A FUEL
RECONSTITUTION ENGINEERING EXPERIMENT

We are beginning engineering studies of fuel reconstitution. Egquip-
ment is described for carrying out the reaction of gaseous UF6 with UF4
dissolved in molten salt and the subsequent reduction with hydrogen of

the resultant UFS' The experiment will be a flowthrough operation, and

the main vessels will consist of a 36-liter feed tank, a UF. absorption

6
vessel, a hydrogen reduction cclumn, and a receiver vessel.
1. INTRODUCTION

A molten-salt breeder reactor (MSBR) will be fueled with a molten
fluoride mixture that will circulate through the blanket and core regions
of the.reactor and through the primary heat exchangers. We are developing
processing methods for use in a close-coupled facility for removing
fission products, corrosion products, and fissile materials from the

molten fluoride mixture.

Several operations associated with MSBR processing are under study.
The remaining parts of this report discuss:
(1) results of inspection of equipment used in experiment MTE-3

for demonstrating the metal transfer process for removal
of rare earths from MSBR fuel carrier salt,

(2) status of the installation of equipment for metal transfer
experiment MTE-3B,

(3) results of studies with mechanically agitated, nondis-~
persing contactors using water and mercury,

(4) results of studies with mechanically agitated, nondis-
persing contactors using molten salt and bismuth, and

(5) a description of equipment being fabricated for a fuel
reconstitution engineering experiment.
This work was performed in the Chemical Technology Division during

the period June through September 1974.

2. DEVELOPMENT OF THE METAL TRANSFER PROCESS

H. C. Savage and W. L. Carter

We plan to continue studying the steps in the metal transfer process
for removing rare earths from molten-salt breeder reactor fuel salt. 1In
this process, fuel salt, which is free of uranium and protactinium but
contains the rare eérths, is contacted with bismuth containing reductant
to extract the rare earths into bismuth. The bismuth phase, which con-
tains the rare earths and thorium, is then contacted with lithium chloride.

Because of favorable distribution coefficients, significant fractions of
the rare earths transfer to the lithium chloride along with a negligible
amount of thorium. The final steps of the process extract the rare earths
from the lithium chloride by contact with bismuth having lithium concen-
trations of 5 and 50 at. %.

The current experiments utilize mechanically agitated contactors as

1,2 The

an alternative to packed columns for the metal transfer process.
Lewis-type contactor appears to have the potential for achieving accept-
able rare-earth mass transfer rates with minimum dispersal of the salt

r

and bismuth phases. This is an important factor, since entrainment

of bismuth into processed fuel salt that is returned to the reactor can-
not be tolerated. The contactor has two agitators——-one in the salt phase
and one in the bismuth phase--~that are located well away from the salt-
bismuth interface. These agitators are operated in such a manner that the

phases are mixed as vigorously as possible without dispersing one phase

into the other.

An engineering experiment (MTE-3) in which the salt flow rate was
about 1% of that required for processing the fuel salt from a 1000 MW (e)
MSBR was operated during 1972 to measure rare—earth mass transfer rates
across the salt-bismuth interfaces.5 These experiments will be continued
in a new experiment, designated MTE-3B, that will use new process vessels,

salts, and bismuth.

2.1 Examination of MTE-3 Equipment and Materials

We completed the examination of the vessels and removal of samples of
the salt and bismuth phases from various locations in the contactor and
stripper vessels. The vessels, along with the salt and bismuth contained
in them, were removed from the hood in Bldg. 3541 and were sent to the

burial ground for disposal.

As previously reportedl significant amounts of iron were found in
samples of the Bi~Th, Li-Bi, and LiCl phases taken at or near the salt~
metal interfaces. Additional samples of the Bi~Th, Li-~-Bi, and LiCl were
taken at various distances from the interfaces and analyzed for iron con-

tent. The samples were obtained by drilling through the tank walls with
a l-in.-diam hole saw, and were generally taken from material approxi-
mately 1/2 in. from the inside-wall surface. Results of iron analyses
are shown in Table 1. As seen in this table, iron concentrations in
samples taken some distance (2 to 8 in.) from the interfaces are signifi-

cantly lower than those at or near the interface.

A section of the Li-Bi~-LiCl across the interfacial area in the
stripper was examined by scanning electron microscopy. Scanning electron
photomicrographs at magnifications of 20, 100, 500, and 2000X were taken
of this interface, and are shown in Fig. 1. These photographs were taken
of the rough, unpclished surface, and no identifiable film or foreign
material is visible. (From a visual examination of the as-removed section
of this interface, there appeared to be a layver of material ~ 1/32 in.

thick of a different structure than the LiCl phase or the Li-Bi phase.)

Analyses by X-ray fluorescence were made at five locations across
this interface, which are identified as 1, 2, 3, 4, and 5 on the 20X
photomicrograph of Fig. 1. Results are given below:

{1) In the LiCl some distance from the Bi-LiCl boundary, some

thorium and a small amount of bismuth and lanthanum were
detected.

(2) 1In the LiCl near the Bi-LicCl boundary, a very small
amount of thorium, iron, and bismuth was detected.

(3) ©On the Bi~LiCl boundary, thorium and iron, as well as
bismuth and chlorine, were detected.

(4) In the bismuth area near the Bi-LiCl boundary, large
amounts of thorium, iron, and chlorine, as well as
the bismuth, were detected. '

(5) In the bismuth area some distance from the Bi-LiC1

boundary, only bismuth was detected.

The presence of iron (or iron oxides) at or near the interface is
consistent with the relative densities of bismuth (9.6 g/cm3) and iron
(7.6 g/cmB) or possibly iron oxides (v 5.5 g/cm3) that would be expected
to precipitate on cooling of the bismuth phase. Iron might be expected
to concentrate at the interfaces due to the densities of the iron and

ircen oxides (v 5 to 5.5 q/cm3), and the salt and bismuth phases (v 3.5
Table 1. Iron content in samples of metal and salt phases
from metal transfer experiment MTE-3 at various
distances from the metal~-salt interfaces

 

Material Source Fe (ppm)

 

Contactor, LiCl side

 

Bi-Th N~ 1/8 in. from intexrface 2500
v 2 in. from interface 18
v 4-1/2 in. from interface 17

Contactor, fluoride salt side

 

Bi~Th ~v o 1/4 in. from interface 1100
v 2 in, from interface 37
~v 4 in. from interface 9
Stripper
Li-Bi < 1/8 in. from interface 1400
v 2 in. from interface 108
v 5 in. from interface 7
StriEBer
LiCl v 1/2 in, from interface 650
v 1l in. from interface 200
v 6-3/4 in. from interface 115
v 8-3/4 in. from interface 35
Contactor
LiF-BeF ,~ThF, v 1/8 in. from interface 320,88%

(72-16-12 mole %)

 

aResults of two different samples. BAll other results are for one sample.

g/cm3 for fluoride salt, 1.5 g/cm3 for LiCl, and ~v 9.6 g/cm3 for bis-

muth) .

When sampling the LiCl in the stripper about 1 in. above the LiCl--
Li~Bi interface, a small amount of black material was seen. We were able
to separate some of this black material (< 1/2 g) from the sait for
chemical analyses. The following results were reported (wt %): Bi -

52.7; Th - 18.7; C1 - 4.1; 1Li - 2.8; F ~ 2.4; Fe ~ 0.26; Be - 1 ppm.
ORNL DWG. 74-10ll6

MAG: 2000X MAG. 500X

      

 

Fig. 1. Scanning electron photomicrographs of the interface between
LiCl and Li-Bi in the stripper from experiment MTE-3. Unpolished.
These constituents total 81%, leaving 19% unaccounted for. No other
elements, except possibly those introduced by oxygen or water contamina-
tion, are known to be present in the system. If it is assumed that the
remainder is predominantly oxygen, this would be sufficient to form oxides
and/or hydroxides of all of the cationic species (bismuth, thorium,
lithium, and iron). The gram equivalent of these species is ~ 0.015,
while the anion-gram equivalent of fluorine and chlorine is only 0.0025.
For 19 wt % oxygen in the material, the oxygen gas equivalent is 0.024,

or 0.011 of hydroxide.

It is not now possible to draw firm conclusions about the relation-
ship of our current observations to the low mass transfer rates seen in
experiment MTE-3. The transfer of fluoride salt into the chloride salt
just prior to shutdown, and the length of time between shutdown and
inspection (from February 1973 to February 1974), have caused much uncer-

tainty in the interpretation of the analyses.

Additional samples of fluoride salt, LiCl--Bi-Th, and LiBi from inter-
facial surfaces in the contactor and stripper have been submitted for X-
ray diffraction analyses in an attempt to identify compounds which might
be present. (Nothing other than beryllium was identified on one sample

of Li~-Bi from the stripper previously examined by X-ray diffraction.)

2.2 1Installation of Metal Transfer Experiment MTE-3B

Fabrication and assembly of new carbon steel vessels were completed
during this report period. An oxidation-resistant protective coating
(METCO No. P443-10, v 0.015 in. thick) was applied to the outside surfaces
of the carbon steel vessels and interconnecting lines to prevent air oxi-
dation at the operating temperature of 650°C. A new pump for transferring
fuel carrier salt between the salt reservoir and contactor was fabricated
and installed. New molybdenum agitator shafts and blades have alsoc been
obtained, and the agitator assemblies were installed in the contactor and

stripper.

The vessels and their contents (fuel carrier salt, lithium chloride,

and bismuth) previously used in experiment MTE-3 were removed from Bldg.
3541 and sent to the burial ground for disposal. The new vessels have
been installed, and the experimental area in which the metal transfer
experiment is located is being rénovated. This includes rerouting some
of the service lines to improve access to sampling stations, relocation
of some pressure gages, flowmeters, and valves for better visibility

and acc¢ess, calibration and replacement (where required) of pressure
gaqes,;flowmeters, and solenoid fialves, and rgcalibration of temperature

controllers and recorders.

We have completed tests on fiwo different oxidation-resistant protec~-
tive coatings using a different method of application for each type of
coating. The coatifigs were applied to test séctions made from longitudinal
half-sections of 6-in. sched 40 mild steel pipe. The test sections were
coatedion all exposed surfaces. One test piece was coated with the nickel
alumide material (METCO No. M405fi10 wire) by flame spraving with a wire
gun to obtain a ccating thickness of v 0.010 in., thereby duplicating the
coating used on the MTE-3 vessels. The second test piece was coated with
a nickel chromium alioy containing 6% aluminum (METCO No. P443-10 powder)
applied with a plasma spray gun to a thickness of ~ 0.0l0 in., as recom-
mended by the manufacturer. The two pieces were placed on fire bricks
inside a furnace and heated to the test tempefature. During the test,
the pieces were thefmally cycled‘several times by alternateiy cocling the

furnace to " 100°C and returning it to the test temperature.

The initial teét temperature was 700°C. After v 400 hr at 700°C with
eight thermal cycleé, the pieces were examined, weighed, and photographed
(Fig. 2)" The M405-10 coating was covered extensively with a rust-colored
oxidation product, but there wasfno sign of spalling of the;coating,
whereaé the piece cfiated with P44B—10 had a muéh better appearance with
only one edge showing a rust color. Weight gain of the P443~10 coated
test piece was about 7.3 mg/cmz, while the M405-10 coating gained about
10.3 mg/cm2

The test temperature was then increased to 815°C and was continued
for 520 hr with eight thermal cycles to 100°C. The test pieces were again

examined, weighed, and photographed (Fig. 2). The M405-10 coating had
ORNL DWG. 74-10117

 

 

MNICKEL AL UMINIDE FOUCFEL CHROGME 4 6% ALUMIEIGM
METCC AMI0S-10 METCC PA43-10

 

PLFCKEL ALLISAINITY
PAETC O M40 5 -10

 

(b)

Fig. 2. Photographs of 6-in. sched 40 mild steel pipe with oxida-
tion resistant protective coatings: (a) after 400 hr at 700°C in air,
(b) after 500 additional hr at 815°C in air.
deteriorated to the point where spalling had occurred, while the P443-10
coating had no spalling except along one edge where there was some rust-
colored oxidation product and some spalling. Total weight gain of the
M405-10~coated test piece was 69.2 mg/cmz, and 20.5 mg/cm2 for the
P443~10 test piece.

Metallographic examinations were made of the best and worst appearing
areas of the two specimens. In the best appearing areas, both coatings
were adherent, but a thin layer of oxide formed at the metal-coating
interface of the specimen coated with METCO M405-10; this indicates that
oxygen is diffusing through the coating. The metal-coating interface of
the specimen coated with METCO P443-10 was free of visible oxide, indi-
cating that it is a good barrier to oxygen diffusion. For the worst
areas of both coatings, oxide at the metal-coating interfaces caused
some spalling where the coating may not have been thick enough to prevent
oxygen diffusion through the coatings. The oxide formed underneath the
P443-10 coating {in a spalled area occurring only along one edge) appeared
to be more dense and protective than that formed on the M405-10-coated
specimen. This indicates that some of the elements in the P443-10 coating
may have diffused into the base metal and imparted a measure of protection,

even though the coating had separated in this area.

This limited test clearly demonstrated the superiority of the plasma
spray P443-10 coating over the M405-10-wire gun sprayed material prev-
iously used for the MTE-3 vessel; therefore, the plasma spray coating has
been applied to the external surfaces of the MTE-3B vessels that will

operate at elevated temperatures.

2.3 Design of the Metal Transfer Process Facility

Design of the metal transfer process facility (MTPF) in which the
fourth metal transfer experiment (MTE-4) will be carried out was underway
when the MSR program was terminated in 1973.6 Briefly reviewed, MTE-4 is
an engineering experiment that will use salt flow rates that are 5 to 10%
of those required for processing a 1000-MW(e) MSBR. Conceptual designs

of the three-stage salt-metal contactor, made of graphite, and its
10

containment vessel were completed.6 The primary purposes of MTE-4 are:

(1) demonstration of the removal of rare-earth fission
products from MSBR fuel carrier salt, and accumulation
of these materials in a lithium-bismuth solution in
equipment of a significant size,

(2) determination of mass transfer coefficients between
mechanically agitated salt and bismuth phases,

(3) determination of the rate of removal of rare earths
from the fluoride salt in multistage equipment,

(4) evaluation of potential materials of construction,
particularly graphite,

(5) testing of mechanical devices, such as pumps and
agitators, that will be required in a processing
plant, and

(6) development of instrumentation for measurement and
control of process variables, such as salt-metal
interface location, salt flow rate, and salt or
bismuth liquid level.

We are currently reviewing the design of metal transfer experiment
MTE-4. A mathematical model of the system has been devised, and a com-
puter program (METTRAN) has been written to simulate transient operation
of the experiment. Computations can be made to determine the concentra-
tion of each nuclide being transferred at each stage and at the feed and
receiving vessels as a function of operating time. The program allows the
experimenter to make parametric studies for such design features as inter-
facial area, number of stages, flow rates, agitator speed, and inventories
of materials. METTRAN is being used to analyze the MTE-4 experiment to
ascertain the significance of various design features on metal transfer

rates in order to fix optimal design conditions.

Due to space limitations in Bldg. 4505, we are planning to locate
MTE-4 and several of the engineering experiments on molten-salt proceésing
in the MSRE Building (7503). General cleanup and checkout of existing
building services (electrical circuits, ventilation, and air-filtration
systems) is underway. A 480-V, 3-phase, 60-Hz, 300-kW diesel generator

set and necessary controls will be installed in the existing generator
11

building at the MSRE site. This will provide emergency power for main-
taining portions of engineering experiments which contain salt or bismuth
at temperatures above the liguidus temperatures. A purchase corder for the

generator set has been issued, and the system design has been completed.

3. SALT-METAL CONTACTOR DEVELOPMENT: EXPERIMENTS
WITH A MECHANICALLY AGITATED, NONDISPERSING
CONTACTOR USING WATER AND MERCURY

C. H. Brown, Jr.

A critical part of the proposed MSBR processing plant is the extrac-
tion of rare earths from the fluoride fuel carrier salt to an intermediate
bismuth stream. One device being considered for performing this extrac-
tion is a mechanically agitated, nondispersing contactor in which bismuth
and fluoride salt phases are agitated to enhance the mass transfer rate
of rare earths across the salt-bismuth interface. Previous reportsz’7’8
have shown that the following reaction in the water-mercury system is

sultable for simulating and studying mass transfer rates in systems with

high density differences:
+ 2+
Pb2 [H20] + Zn[Hg] - Zn [H20] + Pb{[Hgl . (1}

A large amount of data have been reported7 for the water-mercury system in
which it was assumed that the limiting resistance to mass transfer existed
entirely in the mercury phase, as suggested by literature correlations.
During this report period, a series of experiments was performed in the
water-mercury contactor to determine which phase actually controls the
rate of mass transfer and, also, the concentration of Pb2+ at which the

control of mass transfer changes from one phase to the other.

3.1 Theory

The reaction under ceonsideration, Eg. (1), is a liquid-phase jonic
reaction that occurs entirely at the mercury-water interface; this is

pecause zinc metal and lead metal are inseoluble in water and there can
12

be no ionic lead or zinc in the mercury. Since this is a typical ionic
reaction, it is assumed to be essentially instantaneous and irreversible.

The equilibrium constant for the reaction is given by the following equa-

tion:
LPb CZn2+
Kt e ore, (2)
ph2t “zn
where
K = equilibrium constant,
CPb = concentration of Pb metal in mercury, g-mole/liter,
CPb2+ = concentration of Pb iocons in water, g-mole/liter,
CZn = concentration of 2n metal in mercury, g-mole/liter,
CZn2+ = concentration of Zn ions in water, g-mole/liter.

The equilibrium constant is very large, inplying that at equilibrium the

ionic lead and metallic zinc cannot coexist at appreciable concentrations
at the interface. Since it is assumed to be an extremely fast reaction,

the equilibrium relation near the interface must be satisfied at all

times.

For the instantaneous irreversible reaction discussed previously,
two situations could occur near the liquid-liqguid interface, depending on
the relative magnitudes of the individual-phase mass transfer coeffi-
cients and the bulk-phase concentrations of the transferring species in

each phase.9 Figure 3 illustrates these two conditions.

In Fig. 3(a), the limiting resistance to mass transfer is assumed to
occur in the mercury phase. It can be shown that the product of the bulk
phase concentration of reactant and the individual-phase mass transfer
coefficient in the phase where the limitation occurs must be less than
the product of the bulk-phase concentration of the other reactant and
the individual-phase mass transfer ccefficient in the other phase.9 The

concentration of zinc in mercury near the interface decreases from the
13

ORNL DWG. 74-{0/IOR|

LIQUID-LIQUID
INTERFACE

     

 

 

 

1
: Cpb ,B
| |
CZn,B % E (a)
! 2+
MERCURY i Cpb i ! WATER
:Cani E
M A e !
INTERFACIAL FILMS
POSTULATED BY LEW!S AND WHITMAN
| .
CZn,B | : 2+
i i Cep ,B (b)
I
MERCURY ! ! WATER
: Czn'i — I
; A

 

 

Fig. 3. 1Interfacial behavior for an instantaneous irreversible
reaction occurring between two liguid phases. (a) Mercury -phase con-
trolling mass transfer. (b) Water-phase controlling mass transfer.
14

bulk concentration to near zero at the mercury-water interface. The con-
centration at the interface is very small because of the instantaneous
irreversible reaction that occurs at the interface. At the interface in
the water phase, the concentration of Pb2+ has a finite value and

increases through the interfacial film to the bulk phase value.

Fig. 3(b) illustrates the condition in which the limiting resistance
to mass transfer is assumed to occur in the water phase. The explanation
of the concentration gradients in the interfacial films is entirely analo-
gous to the case explained above. The concentration profiles of lead in
the mercury, Zn2+ in the water, and NOB_ (the anion) in the water are not

shown in Fig. 3.

Several correlations have been developed and presented in the litera-
ture for predicting individual-phase mass transfer coefficients in nondis-
persing stirred interface contactors. For the mercury-water system, all
of these correlations predict that the mercury-phase mass transfer coeffi-
cient would be smaller than the water-phase coefficient.

In all previous work performed with the mercury-water system,2'7'8

the concentrations of the reactants were equal. This condition, coupled
with the fact that the mercury-phase mass transfer coefficient was pre-
dicted to be significantly smaller than the water-phase ccefficient,
indicated that the limiting resistance to mass transfer should occur in

the mercury phase.

In order to test the assumption that mass transfer is controlled by
the mercury phase, we can write the following relations for transfer of
the reactants from the bulk phase to the interface where they react, based

on the two-film representation shown in Fig. 3:

Non = ¥ug®Con.p ™ Can,i’ ¢ (3)
Npp2t = kH20A(CPb2+,B = Cppet i) (4)
where

k = individual-phase mass transfer coefficient, cm/sec,
15

N = rate of mass transfer to the interface, g/sec,
. . 2
A = interfacial area, cm ,
C = concentration, (B denotes bulk-~phase concentration, i denotes

interfacial concentration), g/cm3, and

subscripts Hg and H. 0 refer to the phase being considered. As stated

2
above, we assumed that the rate at which reaction (1) proceeds is con=-

trolled by the rate of transfer of zinc through the mercury phase to the
interface. The necessary conditions for this assumption to be valid are:

(1) The equilibrium constant for the reaction must be large
(i.e., the reaction should be irreversible).

(2) The product of the mass transfer coefficient times the
bulk concentration of reactant in the phase where the
rate of mass transfer is limiting must be less than the
product in the other phase.

. . . . 2+ .
Since 1 mole of zinc is equivalent to 1 mole of Pbh according to Eg. (1),
NPb2+ = NZn' Substituting Egs. (3) and (4) into this expression, and
assuming that the controlling resistance is in the mercury phase, the

following expression is obtained for the apparent mercury-phase mass

transfer coefficient:

 

+ — +
kHZO(Csz , B chz ,i)
K - (5)
Hg,2 cZn,B
where
ng A = the apparent individual-phase mass transfer coefficient for
’

the mercury phase, cm/sec,

H 0 = the true individual-phase mass transfer coefficient for the

water phase.

In the above equation, the actual mass transfer coefficient differs from
the apparent mass transfer coefficient only if the mercury phase doass not
control the rate of mass transfer. The transient method used to determine

the mass transfer coefficient has been described previcusly.
16

By an argument similar to that which led to the development of Eq.
(5), an equation- -for the apparent agueous-phase mass transfer coefficient

can be written for the case where the limiting resistance is in the water.

, 2+ . C ..
The concentration of Pb in the water below which the limiting resis-
tance to mass transfer is in the water (at a fixed concentration of zinc
. . 2+ .
in the mercury) can be determined as follows: If the Pb concentration

is sufficiently high, the limiting resistance to mass transfer will be in

2+

the mercury, and the interfacial concentration of Pb will have

+
CPb2 P11
a finite value. If CPb2+ 5 is lowered by an amount A (not large enough to

!
cause the limiting resistance to change phases), Eg. (5) then indicates

that CPb2+ i must also decrease by the same amount A, keeping the differ-
’

ence CPb2+,B "“ch2+,i constant. The water-phase mass transfer coeffi-

cient is assumed to remain constant (although of unknown value). If

+ 1 > 1Ci : ;i +
ch2 B is reduced sufficiently, CPb2+,1 will drop to zero; if CPb2 B

is further reduced, CPb2+ i will remain zero, and the limiting resistance
!

will change from the mercury phase to the water phase. At high values

+
of CPbZ B

remain constant as C

the apparent mercury-phase mass transfer coefficient will
Pb2+ B is lowered to the point where the limiting
r

resistance moves into the water phase. As CPb2+ 3 is further reduced,
!

CPb2+,i will be zero or very near zero. Equation (5) shows that the
apparent mercury-phase mass transfer coefficient will vary directly with
CPb2+,B (since CZn,B is fixed, and the true water-phase mass transfer
coefficient is assumed to be constant). This dependence of the apparent
mercury-phase mass transfer coetfficient on the concentration of Pb2+ in
the water is shown in Fig. 4. The dependence of the.analogously defined
apparent water-phase mass transfer coefficient is also shown. Thus, by
calculating the apparent mercury-phase mass transfer coefficient as a
function of the initial agueous-phase lead concentration for a single
agitator speed, the transition from mercury-phase controlling to agueous-
phase controlling should be identifiable by a line-slope change determined

by plotting the apparent mercury-phase mass transfer coefficient vs the

initial aqueous—phase lead concentration.
17

ORNL DWG. 74-10lll

 

 

» WATER PHASE MERCURY PHASE
2 CONTROLLING 4——II}-—-—---P CONTROLLING
= | £
- | / HQ,A
w |
v W
< O
T T :
i '
& |
S 0 !
O
= o
2w |
8 w |
= W |
z |
- !
Z o |
W = |
a |
<t |
a |
a |
T |

 

 

 

INITIAL AQUEOUS—PHASE LEAD CONCENTRATION

Fig. 4. Theoretical relationship between the apparent mass transfer
coefficients and the initial agueous~phase lead concentration.
18

3.2 EXperimental Apparatus

A series of mass transfer experiments was performed in a mercury-
water contactor to determine the point at which contreol of mass transfer
changes from the mercury phase to the aqueous phase. The experimental
apparatus used in this study is shown schematically in Fig. 5. The equip-
ment consists of a 5- by 7-in. Plexiglas vessel, 10 in. high, containing
two phases, each of which is 3 in. deep. An agitator shaft to which two
four-vaned, flat paddles are attached is suspended from a variable speed
motor. The paddles are positioned at the midpoint of each phase. In
this study, the two phases were 1.8 liters of clean mercury containing
0.1 M Zn, and 1.8 liters of distilled water containing from 0.02 to (.10
M Pb(NO,) .

3.3 Results and Conclusions

Five experiments were run to measure the apparent mercury-phase mass
transfer ccoefficient as a function of the initial aqueous-phase lead
concentration. The initial mercury-phase zinc concentration was held
constant at 0.1 M. Phase volumes and agitator speed were also held

constant at 1.8 liters and n 150 rpm, respectively.

The experimental results are presented graphically in Fig. 6. The
apparent mercury-phase mass transfer coefficient decreased for all initial
lead concentrations less than in previous experiments. This indicates
that under the specified conditions, the limiting resistance to mass
transfer is apparently not in the mercury phase. However, based on the
one point at the lowest Pb2+ concentration, the apparent aqueous-phase
mass transfer coefficient does not seem to remain a constant as would be
predicted by the model described above. This point must be clarified by

additiocnal data.

From these results, the following conclusion can be drawn: For the
conditions under which these and all other runs were performed, the
resistance to mass transfer was not in the mercury phase as was previously
believed. Further studies are needed to test other assumptions employed

in analyzing results from the transient mass transfer equipment.
19

ORNL DWG. 74-10112

 

 

 

 

— 7y
b <X
T 3
L =
> @ X
S & S0
O &5
Qo = - <
o B €
Wi — = A ~ Z
Qo % Xz o > w
ws T .
c
O =
5 z o
a
< ” ]
x 5
m 'va
m.‘.
)
H.&,..,..,
! A .u:‘l
=\ - L
9 o
= o
— ————— AhTi X
AT
S
b w%w%
LR
PR
X ,m.,n.w.hz., )
. - .Havfiy»fiq,M.i_\f’...
4
vt
=<
w i
o=
o )
58
Ul
2y
—z<
O
<TG
g

Schematic diagram of the Plexiglas contactor used for the

5.
mercury~-water system.

Fig.
20

ORNL DWG. 74-10109

 

 

 

 

 

 

1 1 1 |
V- APPARENT MERCURY~
T 0.0I10 PHASE MASS -
Eg TRANSFER COEFFI~
CIENT
~— 0O O- APPARENT AQUEOUS—
— PHASE MASS TRANS-
Z FER COCFFICIENT
Y 0.008k O i
s O
w
.
w
W O
. O
a 0.006F -
w 1
1
wn
&
a Vv
k- 0004+ ~
o
wn v
<T
=
O
w 0.0021 _
5
v Vv
<t
W
P
00 J I ] 1
0.0 0.02 0.04 006 008 0.10

INITIAL

Fig. 6. Experimental results showing the apparent mercury-phase
mass transfer coefficients as a function of the initial agueous-lead

e el

AQUEOQUS — PHASE LEAD CONCENTRATION (

concentration at constant agitator speed.
21

4. SALT-METAL CONTACTOR DEVELOPMENT: EXPERIMENTS WITH
A MECHANICALLY AGITATED, NONDISPERSING CONTACTOR
IN THE SALT-BISMUTH FLOWTHROUGH FACILITY

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

Mechanically agitated, nondispersing contactors are being considerad
for extracting protactinium and the rare-earth fission products from the
fuel carrier salt of a molten-salt breeder reactor into molten bismuth

containing dissolved reductant.

Mass transfer rates were measured in a mild steel contactor which
was fabricated and installed in the Salt~Bismuth Flowthrough Facility.
This experimental system allows (a) periodic purification of the feed salt
and metal, (b) removal of surface contamination from the salt-metal inter-
face, and (c) varying of the distribution ratio of the material of inter-

est between the salt and bismuth.

The new stirred interface contactor makes use of the existing piping
equipnent and instrumentation in the Salt-Bismuth Flowthrough Facility
that were used for studying salt-bismuth flow in packed columns.lo”l4 A
simplified flow diagram of the facility is shown in Fig. 7. The facility
contains five vessels, the associated piping, and the contactor assembly.
The vessels are a salt~feed tank, a salt-collection tank, a bismuth-feed
tank, a bismuth-collection tank, and a treatment vessel (T5) for sparging
both salt and metal with mixtures of HF and H2' To conserve space, feed
and collection tanks for both salt and metal are concentric. Feed and
collection tanks and piping are made of low carbon steel. The treatment

vessel is made of stainless steel with a graphite liner.

A diagram of the contactor is shown in Fig. 8. The contactor is made
of a 6-in.-diam low-carbon steel vessel containing four l-in.-wide verti-
cal baffles. The agitator consists of two 2-7/8-in.-diam stirrers with
four 3/4-in.-wide noncanted blades. A 3/4~in.-diam overflow at the
interface allows the removal of interfacial films with the salt and metal
effluent streams. Salt and bismuth are fed in the contactor below the
surface of the respective phase, and both phases were equilibrated prior

to an experiment. The system was operated in essentially the same manner
ORNL DWG 74-694R¢

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

sé‘f;EAM
SAMPLER Blsfi%té;n
SAMPLER
-t
(t-—fm=——> ) FV=1 J
1 {
i i
1 H-————-!-———-] FV—ZU
T 1 I FV-3 J 1
| i
| - - !
P I
| : | : FV—4U ' ’
l i i
¥ , ! b
i ; i g ' l
1 ‘1 1y L by
1k 1 ; ' I i i '
L bl b | i l
i vl i .
t ! 1 ¢4 i
o by i : T-‘HT-E; | -5 |
§ ‘ § ‘ i i q ;
CONTACTOR METAL FEED AND SALT FEED AND SALT AND METAL
COLLECTION TANK COLLECTION TANK TREATMENT VESSEL

Fig. 7. Flow diagram of the Salt-Bismuth Flowthrough Facility with
the mechanically agitated contactor installed.

ce
23

ORNL DWG 74-10073

T

-fi§$§£::tfinflnuhums
e A=y 8
AP ~ 3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A-Bismuth inlet
8-Sait inlet
C~Four l-in. baffles
D-2-7/8-in. x 3/4-in. blades
E-Gas-sait interface

F-Bismuth-salt inferface
G-Interface removal
H-Salt effluent

I-Bismuth effluent

Fig. 8. Diagram of the mechanically agitated, nondispersing con-
tactor installed in the Salt-Bismuth Flowthrough Facility.
24

10~
as was employed with the packed column. O-14 After transfer of the salt

97 237
and metal phases to the feed tanks, Zr and U tracers were added to
the salt. With this technique, the rates at which zirconium and uranium
tracers transfer from the salt to the bismuth could be measured in a

system that was otherwise at chemical egquilibrium.

The experimentally determined data from the system are sufficient
to allow three independent expressions for the overall mass transfer
coefficient to be derived for the contactor in the Salt-Bismuth Flow-
through Facility from steady-state material balance relationships.

These expressions for the overall mass transfer coefficient are given

 

 

below in terms of the measured guantities Cl' CS, Cm' Fl' F2, D, and A:
Cs
(- &)
1 1 Cl
K = ’ (6)
B! A(fi)+fl(ia)(i),5(i)
D
Cl D C1 F2 F2
Cm
— ] F
),
K = , and (7)
s , C F C
r=A{2)NF) - E)E)
F
Cl 1 Cl D
C
2\ g
(),
K, T T ®)
T (2)5)
C D
s
where
Fl = flow rate of salt, cm3/5ec,
F2 = flow rate of metal, cm3/sec,

Q2
i

. i ) ) 3
1 tracer concentration in salt inflow, units/cm ,
25

. : . 3
C = tracer concentration in salt outflow, units/cm ,
. : . 3
Cm = tracer concentration in metal outflow, units/cm™,
. . 2
= interfacial area, c¢m , and

D = distribution coefficient = ratio of concentraction in metal
phase to concentration in salt phase at equilibrium,
moles/cm3
moles/cm3
The subscripts 1, 2, and 3 on KS indicate the eqguation used to evaluate
the overall mass transfer coefficient, KS. The overall mass transfer
coefficient is related to the mass transfer coefficients in the individual

phases through the following relation:
]./KS = )L/kS + 1/ka, (9)

whare

Ks = overall mass transfer coefficient based on salt-phase con-

centrations, c<m/sec,

kS = individual mass transfer coefficient in salt phase, cm/sec,
and
km = jndividual mass transfer coefficient phase in metal, cm/sec.

Within experimental error, the distribution coefficient, D, can be
set as desired. In order to minimize effects of uncertainties in the
value of D on the calculated value of the overall mass transfer coeffi-
cient, it is desirable to make D fairly large. For the values of con-
centrations and flow rates used in these experiments, the terms con-
taining D in Egs. (6)-(8) are less than 5% of the values of the other
terms for values of D greater than 20, and can be disregarded with
little error. Assuming that the terms containing D can be omitted,

Egs. (6)-(8) reduce to
20

8
e
s, TR | T | o
LR
Cl

N

1
KS = "K" c ’ and (ll)
b -ENE)
Fl Cl
F C
k == B, (12)
s A C
3 s

uncertainties in the distribution coefficient do not affect the accuracy
of the overall mass transfer coefficient. However, as shown by Eg. (9),
when D is very large, the overall mass transfer coefficient is essentially
the individual salt-phase coeffiéient, since resistance to mass transfer

in the metal phase is negligible in comparison.

Six runs have been completed to date. Four runs (TSMC-1 through -4}
have been discussed previously,7 and two runs (TSMC-5 and ~6) were com-
pleted during this report period. We have recalculated results for all

runs and will discuss them together in this report.

All runs were performed using the same procedure. While the fluoride
salt and bismuth are in contact in T5 (the treatment vessel), sufficient
beryllium is added to the salt electrolytically to produce the desired
distribution coefficient, D. Since it is impossible to completely exclude
oxidants, addition of beryllium must be carried out periodically in order
to maintain a relatively high distribution coefficient. Periodic transfer
of salt and bismuth throughout the system are also performed to keep the

system in equilibrium.

Prior to a run, the salt and bismuth phases are separated by pressur-

izing the salt-bismuth purification vessel and transferring salt and
27

, . o7 o7
bismuth to their respective feed tanks. Approximately 7 mCi of Zr-~ Nb

and 50 to 100 mCi of 237U308 are allowed to dissolve in the salt phase

about 2 hr prior to an experiment.

Salt and bismuth streams are passed through the contactor vessel at
the desired flow rates by controlled pressurization of the salt and bis-
muth feed tanks. The contactor is maintained at approximately 590°C for
all runs. Both phases exit through a common overflow line, separate,
and return to the salt and bismuth catch vessels. Periodic sampling of
both exit streams is accomplished by means of flowing-stream samplers
installed in the exit lines from the contactor.

Sample analysis was accomplished by first counting the sample cap-

2 - .
sules for the activity of 37U (207.95 kev B ) and the activity of 97Zr~

97Nb (743.37 keV and 658.18 keV B, respectively) after secular equilib-
rium was reached between 97Zr and its daughter 97Nb. The material in

the sample capsules was then dissclved, and the activity of 237U was
counted again after the 97Zr-97Nb activity had decayed to a very low
level; this was done to correct for self~absorption in the solid samples.
Mass transfer rates were then calculated from the ratios of tracer con-~
centrations as discussed previously.

The counting data obtained during runs TSMC-2 through -& are shown

in Tables 2~6. Counting data are given for 237U (207.45 kev 8 ) and

27 237

Zr (743.37 keV 8") in the solid salt and bismuth samples, and for U
after the samples were dissolved. All the results are given in terms of

counts per gram of phase.

Results of measurements of overall mass transfer coefficients are
shown in Table 7. Three different equations for calculating mass trans-
fer coefficients were used for calculating the results for any one run,
and the results are presented as an averade of three calculated values
with the standard deviation. Values are given both for results based
on the uranium counting data and for results based on the zirconium

counting data.
Table 2. Counting data obtained from run TSMC-2

 

Solid analysis Soluticn analysis Solid analysis Solid analysis Solution analysis Solid analysis
Sample for 237y for 237y for ¥7zr Sample for 237y for 237y for ¥7zr
coded (counts/g) {counts/g) {counts/g) coded (counts/g) (counts/g) (counts/g)

 

Samples taken prior to run

 

 

 

88-B~5 < 3.3 x 102 < 2.8 x 108 < 6.37 x 10! 86-5-5 < 8.5 x 103 < 1.8 x 10 < 4.5 x 102
93-3-1 < 3.2 x 107 <1.9 x 103 < 1.03 x 102 87-5-5 < 6.8 x 103 < 1.8 x 10 < 7.7 x 102
94-B~1 < 2,2 x 102 < 2.3 x 103 < 6.6 x 10! 89-5-3 < 2.9 x 103 < 6.4 x 103 < 3.9 x 102
- - 90-S-3 < 5.6 x 10 < 6.8 x 10% < 6.6 x 102
Samples taken prior to rum, but after addition of tracers
91-5-3 3.01 x 10° 2.79 x 105 1.85 x 105
92-5-3 1.96 x 10° 2.09 x 10° 1.79 x 105
Samples taken during run
93-B-FS 1.80 x 104 3.49 x 10% 2.01 x 10% 100-S~FS 2.32 x 10° 2.36 x 105 1.48 x 105
94-B-FS 1.46 % 10% 3.29 x 0% 1.69 x 10" 101-S-FS 2.41 x 10° 2.70 x 105 1.52 x 105
95-B-FS 1.91 x 1o0% 2.80 x 10% 2.00 x 1p“ 102-5-FS 2.44 x 109 2.57 x 10° 1.71 x 108
96~8~FS 2.24 x 104 4.28 x 10" 2.10 x 10% 103-8-FS - - -
97-8-FS 2.15 x 10% 3.87 x 10 1.94 x 104 104-5-FS 2.65 x 10° 2.87 x 10° 1.90 x 103
98-B-FS 1.86 x 10 3.91 x 104 1.97 x 104 105-S=FS 2.74 x 105 2.77 x 19° 1.64 x 103
99-5-FS 3.21 x 1o“ 5.01 x 10% 2.74 x 104 106=-5-FS 2.91 % 1n° 2.57 x 105 1.65 x 103
. Samples taken after run
107-B-1 5.05 x 103 1.08 x 1o% 4.40 x 103 111-5-3 5.00 x 10 5.09 x 10° 1.94 x 10°
108-B-1 4,18 x 103 1.04 x 10% 4.05 x 103 112-3-3 5.14 x 10° 5.51 x 107 1.93 x 103
109-B-2 1.48 x 104 3.29 x 1p% 1.61 x 1Q% 113-5-4 2.41 x 10° 2.82 x 105 1.29 x 105
110-B-2 1.96 x 10% 3.76 x 10 1.83 x 10" 114-8-4 2.55 x 10° 2.91 x 10° 1.36 x 10%
115-B-5 2.76 x 10% 5.52 x 10" 2.40 x 10% 117-5-5 1.53 x 103 < 1.9 x 103 6.63 x 10“
116-8-5 5.15 x 10% 6.34 x 10 2.54 x 10" 118-5-5 9.59 x 1o“ < 2.6 x 10° 7.95 x 10

 

3Each sample is designated by a code corresponding to A-B-C, where A = sample number; B = material in sample (B = bismuth, S = salt); and
€ = sample origin; 1 = T1l; 2 =T2; 3 =1T3; 4 =7T4; 5= T5; F§ = flowing stream sample.

8¢
29

Table 3. Counting data obtained from run TSMC-3

 

 

 

 

 

 

Solid analysis Solution analysis Solid analysls Solution analysis
Sample for 237y for 237y Sample for 237y for 2370
code?® (counts/g) (counts/g) code® {counts/g) (counts/g)
Samples taken prior to run
141~-B~5 < 1.13 x 102 < 5.1 x 103 143-8-5 < 3.3 x 103 < 5.3 x 10%
142-B-5 <1.64 x 10° <s5.5 x 103 144~5=5 < 3.2 x 103 <7.6 x 10%
147-B~5 <6.26 x 10! < 3.4 x 103 145-5-3 < 3.4 x 103 < 7.6 x 10%
148-B-5 < 1.00 x 102 <4.0 x 103 146-5-3 < 3.2 x 0% < 5.6 x 10
Samples taken prior to run but after addition of tracer
149-5-3 6.13 x 105 9.37 x 10°
150-5-3 6.33 x 10° 9.29 x 10°
Samples taken daring run
151-B~FS 4,52 x 10* 1.13 x 10° 158~5-F5 3.48 x 10° 4.89 x 105
152-B-F§ 4,14 x 10" 1.29 x 103 159-5-F5 - .
153-B~F5 - 4.12 x 10% 1.33 x 10° 160-S-FS 2.96 x 10° 4.36 x 10°
154-B-FS 4.51 x 10° 1.44 x 10° 161-5-F§ 3.08 x 10° 4.56 x 10°
155-B~F3 6.32 x 10° 1.26 % 10° 162-S-F§ 3.02 x 10° 3.99 x 10°
156-B~FS 3.88 x 10 1.44 x 10% 163~S-F3 3.19 x 105 4,43 x 105
157-B-FS 4.00 x 10% 1.49 x 10° 164-5-FS - ——
Samples taken after rum
165-B-1 4.38 x 102 4.8 x 103 167-5-3 6.68 x 105 1.19 x 105
166-8-1 6.53 x 102 4,7 x 103 168-5-3 - -
169-B-2 3.19 x 10* 1.61 x 10° 171-5-4 2.64 x 10° 4.71 x 10°
170-B-2 2.86 x 10" 9.56 x 10% 172-8-4 2.70 x 10% 4.12 x 105
173-B-5 5.47 x 10% 1.43 x 10° 175-8-5 < 6.4 x 10° < l.4 x10°
174-B-5 5.24 x 0% 1.44 x 10° 176-5-5 < 6.4 x10° < 1.4 x 10%

 

8Each sample 1s designated by a code corresponding to A-B-C, where A = sample number; B = material in
sample (B = bismuth, S = salt); and C = sample origin; 1 = Tl; 2 = T2; 3 = T3; 4 = T4; 5 = T5; FS =
flowing stream sample.
Table 4. Counting data obtained from run TSMC-4

 

 

 

 

 

 

Solid analysis Solution analysis Solid analysis Solid analysis Solution analysis Solid analysis
Sample for <37y for 237y for %7zr Sample for 237y for 237y for ¥72r
coded (counts/g) {counts/g) {counts/g) code? {counts/g) (counts/g) (counts/g)
Samples taken prior to run
191-B=5 <1.03 x 103 < 5.1 x 103 < 4.4 x 107 189-S-5 < 5.9 x 103 < 3.6 x 10% < 1.5 x 103
192-B-5 < B.83 x 102 < 6.7 x 103 < 3.3 x 10° 190-5-5 < 4.0 x 103 < 3.8 x 1l0% < 6.4 x 102
195-8-1 <1.00 x 103 < 4.9 x 103 < 3.3 x 10° 193-5-3 < 4.7 x 103 < 3.6 x 10% < 1.4 x 103
196-8-1 <1.19 x 103 < 4.7 x 10° < 5.7 x 10?2 194-5~3 < 4.3 x 103 < 1.0 x 104 < 9.5 x 102
Samples taken prior to run but after addition of tracers
197-S-3 1.22 % 1086 1.44 x 108 1.92 x 10°
198-5-3 1.21 x 10% 1.49 x 108 1.99 x 10°
Samples taken during run
199-B-FS 9.67 x 10" 2.91 x 10° 3.63 x 10* 206-5-FS 1.46 x 103 2.97 x 10° 4.58 x 10
200-B-FS 1.45 x 105 3.57 x 10° 4.50 x 104 207-S-FS 2.64 x 10° 3,23 x 105 6.33 x 10
201-B-FS 1.70 x 10° 4.82 x 10°% 5.66 x 10% 208-8-FS 2.98 x 1n° 2.88 x 105 4.61 x 10%
202-B-FS 1.85 x 105 - 4.72 x 10° 6.34 x 10% 209-S~FS 2.97 x 105 5.01 x 105 6.07 x 10%
203-B-FS 2,19 x 105 5.01 x 10° 6.38 x 104 210-5-FS 2,76 x 105 3.34 x 105 6.96 x 10%
204-B-FS 1.83 x 10° 5.36 x 10° 6.43 x 104 211-5-FS 2.63 x 105 3.10 x 105 8.48 x 10
205-B-FS 1.57 x 10° 5.32 x 10° 7.12 x 10% 212-S-FS 2.52 x 105 3.14 x 105 6.84 x 10"
Samples taken after run
213-B-1 2.06 x 103 8.7 x 103 1.9 x 10 217-5-3 1.31 x 108 1.48 x 106 2.39 x 10°
214-B-1 1.73 x 103 7.7 x 103 1.1 x 103 218-58-3 1.25 x 10® 1.54 x 108 2.36 x 10°
215-B-2 1.08 x 10° 2.84 x 10° 3.69 x 10“ 219-S-4 2.29 x 10° 2.61 x 105 4.88 x 10%
216-B-2 1.07 x 10° 3.17 x 105 3.68 x 10% 220-5-4 2.38 x 105 2.71 x 105 4.67 x 10"
221-B-5 7.86 x 10% 2.23 x 10° 2.53 x 104 223-5-5 8.3 x 10% < 4.7 x 10% < 2.2 x 10°
222-B-5 B.28 x 10% 2.38 x 10° 2.62 x 104 224-8-5 8.0 x 10" < 4.9 x 10% < 3.9 x 103

 

a . . .
Each sample is designated by a code corresponding to A-B-C, where A = sample number; B = material in sample (B = bismuth, S = salt); and
C = sample origin; 1 = Tl; 2 = T2; 3 =T3; 4 = T4; 5 = T5; FS = flowing stream sample.

o€
Table 5. Counting data obtained from run TSMC-5

 

Solid analysis Solution analysis Solid analysis ' Solid analysis Solution analysis Solid analysis
Sample ‘ for 237y ‘ for 237y for 972r Sample for 237y for *37y for ¥zr
code? {counts/g) (counts/g) {counts/g) code® (counts/g) {counts/g) {counts/g)

 

Samples taken prior to run

 

 

 

227-B-5 < 5.64 x 102 L - < 1.4 x 102 225-5-5 < 7.4 x 103 - <1.6 x 103
228~B-5 < 8.49 x 10° - < 2.2 x 102 226-5-5 < 6.9 x 103 - < 8.9 x 102
231-B-1 < 4.83 x 102 — 1.1 x 102 229-5-3 < 4.9 x 108 - <1.2 x 103
232-B-1 < 5.91 x 102 - < 1.0 x 102 230-5-3 < 6.0 x 103 -~ < 1,4 x 103
Samples taken prior to run but after addition of tracers
233-5-3 2.54 x 108 3.44 x 108 2,10 x 10°
234-5-3 2.57 x 108 3.48 x 108 2.20 x 10°
Samples taken during run
235-B-FS 1.04 x 103 2.87 x 10° 1.85 x 10" 242~5-FS 1.83 x 108 2.13 x 10° 2.47 x 10°
236-B-FS 1.30 x 10° 4.13 x 10° 2.62 x 104 243~S-FS 1.64 x 106 2.00 x 105 1.79 x 10°
237-B-FS 1.53 x 10° 4.60 x 103 2.94 x 10% 244~5-FS 1.82 x 10° 2.00 x 10° 2.64 x 10°
238-B-F§ 1.35 x 10° 4.61 x 10° 2.89 x 10% 265-5-FS 1.78 x 10° 2.10 x 10% 2.43 x 10°
239-B-FS 1.35 x+10° 4,83 x 10° 2.78 x 1pY 246~5-FS 1.96 x 108 2.01 x 10° 2.25 x 10°
240-B-FS 1.67 x 10° 4,67 x 10° 3.04 x 10" 247-8-FS - 1.79 x 10° 2.08 x 10° 1.92 x 105
241-B-FS 1.41 x 10° 5.31 x 10° 2.92 x 10% 248~5-FS 5.40 x 105 - 5.36 x 10%
Samples taken after run

249-8-1 < 1,38 x 103 -— < 4.0 x 102 253~5-3 3.11 x 108 -~ 2,96 x 103
250-B-1 < 1.05 x 10° - < 3.5 x 104 254~5-3 3.28 x 10° -~ 3.19 x 1g°
251-8-2 7.70 x 10 - 1.95 x 10% 255-5-4 1.52 x 106 - 1.19 x 19°
252-8-2 1,00 x 10° - 1.80 x 10% 256~5-4 1.10 x 108 - 1.37 x 10°
257-B-5 1.71 x 10° - 2.47 x 10% 259-5-5 < 9.2 x 103 - < 2.6 x 103
258-5-5 1.34 x 10° -- 2.00 x 10 260~5-35 < 1.1 x ig% - < 3.5 x 103

TE

 

8Fach sample is designated by a code corresponding to A~B-C, where A = sample number; B = material in sample (B = bismuth, § = salt); and
C = sample origin; L = Tl; 2 =7T2; 3 = T3; &4 = T4; 5 = T55 FS = flowing stream sample,
Table 6. Counting data obtained from run TSMC-6

 

Solid analysis Solution analysis S0lid analysis Solid analysis Solution analysis Solid analysis
Sample for 237y for 237y for 37zr Sample for 237y for 237y for 97zr
code? {counts/g) {counts/g) (counts/g) code? {counts/g) (counts/g) {counts/g)

 

Samples taken prior to run

 

 

 

261~B-5 1.75 x 10% 5.32 x 10° -- 259-5-5 < 6.2 x 103 <1.3 x 10% <2.2 x 103
262-B-5 2,37 x 10 5.55 x 10" 2.51 % 103 260-5-5 <6.1 x 103 <1.3 x 1o* <1.3 x 103
265-8-1 1,95 x 10% 5.78 x 10% 8.96 x 102 263-5-3 <1.3 x 10°¢ <1l.4 x 10% <i.6 x 103
266-B~1 1.99 x 10% 5.47 x 10" 9.48 x 101 264-5-3 <5.6 x 103 < 1.4 x 104 <2.1 x 103
Samples taken prior to run but after addition of tracers
267-5-3 1,12 x 108 1.29 x 106 3.29 x 105
268-5-3 1.07 x 10° 1.34 x 106 3.16 x 105
Samples taken during run
269-B-FS 7.69 x 10 2.39 x 10° 6.78 x 10% 276-5-FS 4.11 x 105 3.53 x 105 1.75 x 102
270-B-FS 1.13 x 10° 3.63 x 10° 6.70 x 10% 277-S-FS 4.63 x 10° 4.55 x 105 2.30 x 10
271-B-FS 1.25 x 10° 4.16 x110° 7.54 x 10" 278-5-FS 4.59 x 10° 3.83 x 105 2.11 x 10°
272-B-FS 1.43 x 10° 4,17 x 10° 8.01 x 104 279-$-F§ 4.03 x 10° 4,38 x 10° 2.16 x 10°
273-B~FS 1.55 x 10° 4.36 x 10° 7.79 x 104 280-S~FS 5.00 x 10° 4.26 x 105 1.85 x 10°
274-B~FS 1.54 x 10° 4.27 x 10° 9.01 x 10% 281-5-¥5S 4.83 x 10° 4.66 x 105 2.03 x 10°
275-B-FS 1.53 x 10° 4.53 x 10° 8.85 x 10" 282-5-FS 4.63 x 105 3.40 x 105 2.07 x 10°
Samples taken after run

283-B-1 2,28 x 10" 6.79 x 104 3.73 x 103 287-5-3 1.32 x 105 1.60 x 106 3.47 x 103
284-B-1 2,51 x 10% 6.59 x 10" 3.44 x 10° 288-5-3 1.20 x 108 1.34 x 105 3.57 x 10°
285-B-2 1.07 x 103 3.34 x 10° 5.33 x 10" 289-5-4 4.04 x 10° 4.15 x 105 8.40 x 10%
286-B-2 1.13 x 103 3.11 x 105 5.59 x 104 290-S-4 3.90 x 105 3.80 x 105 8.89 x 10"
291-B-5 1.05 x 10° 3.41 x 10° 4.36 x 10% 293-5-5 < 9.2 x 103 < 1.8 x 10% <5.7 x 10
292-B-5 1,33 x 109 3.31 x 10° 4.45 x 10% 294-5~5 < 8.1 x 103 < 1.7 x 103 <5.0 x 103

 

a ) , X
Each sample is designated by a code corresponding to A~8-C, where A = sample number; B = material in sample (B = bismuth, S = salt): and
C = sample origin 1=T1 2=T2 3=17T3 4=7T4 5=7T5 FS = flowing stream sample.

4
 

 

 

Table 7. Experimental results from the salt-metal contactor
Salt Bismuth Stirrer Fraction K (cm/sec)
flow flow rate o 5 tracer S
Run {cc/min) {cc/min) (rpm) u Zr transferred® Based on U Based on Zr

TSMC-2 228 197 121 0.94-34 0.96 0.17 .0059 — 0.009%92 0.0083 + 0.0055
TSMC-3 166 173 162 >34 -- 0.5G .012 + 0.003 --=
TSMC-4 170 144 205 >172 >24 0.78 .054 + 0.02 0.035 + 0.62
TSMC-5 219 i75 124 >43 >24 C.35 0.0095 + 0.0013 0.0163 + 0.159
TSMC-5 206 i85 180 >172 >24 .04 .049 + C.02 0.020 + G.01

 

a ,
Fraction tracer transferred

{1 - Cs/cl)'

ge
34

A preliminary run, TSMC-1, was primarily designed to test the proce-
dure. Salt and bismuth flows were approximately 200 cc/min, and the
stirrer rate was 123 rpm. Unfortunately, the distribution ceoefficient
was too low to effect any significant mass transfer; thus, mass transfer
rates could not be accurately determined, and no results are shown for

this run.

Operation of the equipment during run TSMC-2 was very smooth. The
salt and bismuth flow rates were 228 and 197 c¢c/min, respectively. The
distribution coefficients were higher than for the previous run but were
lower than desired, and with a concomitant large degree of uncertainty.
Several determinations of DU were made with a range of 0.94 to 34; conse-
quently, only a range of possible values for the overall mass transfer
coefficient could be stated for the results based on uranium. One deter-
mination of DZr was made, indicating that DZr was 0.96. A value for
overall mass transfer coefficient based on zirconium is given based on
the value of this distribution coefficient, but since the value for DZr
must be considered uncertain, there is a larger error in the mass transfer

coefficient than is indicated by the reported standard deviation given in

Table 7.

If it is assumed that the ratio of salt~side mass transfer ceoeffi-
cient to bismuth-side mass transfer coefficient can be determined from
the Lewis correlation, then the salt-side mass transfer coefficients for
run TSMC-2 are higher than the overall mass transfer coefficient by a

factor of 1.55 for the results based on zirconium.

A bismuth line leak occurred immediately preceding run TSMC-3.
During the resulting delay for repairs, the 97Zr decayed and only the
237U 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 D__ (greater than 34) was maintained for this

U
run.
In run TSMC-4, flow rates of 170 and 144 c¢c¢/min were set for the salt
and bismuth flows, and a stirrer rate of 205 rpm was maintained. The

distribution coefficient determined from samples taken before, after, and
35

during the run, were dgreater than 172 for DU and greater than 24 for DZr.
This value is sufficiently large so that Egs. (10})~(12) are wvalid, Large
distribution coefficients cannot be determined precisely due to the inabil-
ity to determine very small amounts of uranium in the salt phase. No

problems arcse during this run.

Runs TSMC-5 and ~6 were performed without incident. The distribution
coefficients were maintained at high levels for both runs. TSMC~5 had a
stirrer rate of 124 rpm, and salt and bismuth flows of 219 and 175 cc/min.
TSMC-6 salt and bismuth flows were 206 and 185 c¢¢/min, respectively, with

a stirrer rate of 180 rpm.

As mentioned previously, when the distribution coefficient is large,
the overall mass transfer coefficient is essentially the individual salt-
phase mass transfer coefficient. Thus, results for runs TSMC~3 through
-6 can be compared directly to the Lewis correlation. The Lewis correla-

3,4

tion for mass transfer in the nondispersing contactor is

 

60 kl 6 n2
= (6.76 x 10 7) {Re. + Re_ -— + 1, (13)
v 1 2 n
1 1
where
k = individual mass transfer coefficient, cm/sec,
v = kinematic viscosity (n/p). cm2/sec,
n = vwviscosity, P,
. 3
p = density, g/fcm”,
2 . .
Re = nd /v, dimensionless,
d = agitator diameter, cm, and
n = agitator speed, rps.

The subscripts 1 and 2 refer to salt and bismuth, respectively. Figure
9 shows a comparison of the measured values of mass transfer coefficient

with the Lewis correlation. The effects of mass transfer resistance in
10000

1000

 

10

1.0

36

ORNL OWG 74 -833I1R3

 

Ty 1Td

1

 

1

 

T Tl lll] { I

T T TTTT] T
LEWIS

CORRELATION

N\

EXPERIMENTAL

SYSTEM
EXPERIMENTAL VALUES

SYSTEMS

® BASED ON U
A BASED ON 2Zr

1 lllllll 1 1 lllllll —

   

FROM MERCURY-WATER

FROM ORGANIC -WATER

U TT LA
’

b1 b il

VAt
v
g

veund

 

VALUES

o

]

Cl il

]

 

L1111l

 

10

Fig.

9.

100 1000
[6.76 x 107®] [Re; + Re; :;2

10000

]I.GS

Experimental results from the salt-bismuth contactor.
37

the bismuth phase cannot be accurately accounted for in the results of

run TSMC-2; hence, these results are not included in Fig. 9.

Figure 9 shows that twe runs produced values of mass transfer coeffi-
cient that were 30 to 40% of the values predicted by the Lewis correla-
tion, while there were two runs that produced values which were slightly
greater than those predicted by the Lewis correlation. The mass transfer
coefficients based on zirconium are consistently lower (slightly) than
the values based on uranium. It is felt that the mass transfer coeffi-
cients based on zirconium are less reliable than those based on uranium;
in all runs, only about 50% of the zirconium tracer could be accounted
for, whereas more than 80% of the uranium could be accounted for. This
discrepancy 1is probably related to self absorption of the 743.37 keV B

from 97Zr in the bismuth samples.

We believe that the initiation of salt entrainment into the bismuth
begins to occur at a stirrer speed between 160 and 180 rpm. The apparent
increase in mass transfer coefficient is a manifestation of an increase
in the surface area for mass transfer caused by surface motion. Experi-
ments with water-mercury and water-methylene bromide systems support

this belief.

If, as Fig. 9 seems to indicate, the mass transfer coefficients in
the salt-bismuth system are lower by a factor of about 0.35 than pre-
dicted by the Lewis correlation, this corrected correlation could be used
to estimate the area and bismuth flow rate required in the rare-earth
removal contactor in the present flowsheet. The Lewis correlation pre-
dicts that mass transfer coefficients are approximately proportional to
the agitator diameter, d, to the 3.30 power. Since it has been shown
that the allowable agitator speed below which there is no phase dispersal
is inversely proportional to the 1.43 power of the agitator diameter, at
speeds slightly below the limiting agitator speed the mass transfer
coefficient will apparently be proportional to the agitator diameter to
the 0.94 power. These results were used to estimate the area reqguired for
a single-stage contactor and the bismuth flow rate to remove cerium, which

has the shortest removal time (16.6 days) of the rare earths in the
38

reference flowsheet. For a salt flow rate of 0.88 gpm (corresponding

to the 10-day cycle time used in the reference process flowsheet) and a
distribution coefficient of 0.062 {the distribution coefficient that was
used in the reference processing flowsheet), an area for mass transfer of
10 ft2 will allow removal of cerium on a 16.6-day cycle if the bismuth
flow rate through the contactor is 30 gpm and a 3-ft-diam agitator is
used. This bismuth flow rate is only about 2-1/2 times the flow rate
specified in the processing plant flowsheet on the basis of eguilibrium
calculations. The use of a higher distribution coefficient, multiple
stages, etc., can result in a reduction of either the bismuth flow rate
or the mass transfer area. These changes, however, will result in
changes in other sections of the flowsheet, and more extensive calcula-

tions are reguired to evaluate these effects.

5. FUEL RECONSTITUTION DEVELOPMENT: DESIGN OF A
FUEL RECONSTITUTION ENGINEERING EXPERIMENT

R. M. Counce

The reference flowsheet for processing the fuel salt from a molten-
salt breeder reactor (MSBR) is based upon removal of uranium by fluorina-
tion to UF6 as the first processing step.l6 The uranium removed in this
step must subsequently be returned to the fuel salt stream before it
returns to the reactor. The method for recombining the uranium with the
fuel carrier salt (reconstituting the fuel salt) is to absorb gaseous
UF_ into a recycled fuel salt stream containing dissolved UF

6
the reaction:

4 by utilizing

+ = 2 (14)

Pew) T Pa@ T2 T

The resultant UF5 would be reduced to uF, with hydrogen in a separate

vessel according to the reaction:

+ 1/2 H + HF

2{(g) - UF4(d) (g) ° (15)

UFS(d)
39

We are beginning engineering studies of the fuel resconstitution step
in order to provide the technology necessary for the design of larger
equipment for recombining UF6 dgenerated in fluorinators in the processing
plant with the processed fuel salt returning to the reactor. Eguipment
for studying the fuel reconstitution process has been designed during

this report period, and is described in this report.

A flow diagram of the equipment to be used for the fuel reconstitu-
tion engineering experiment (FREE) is shown in Fig. 10. The equipment

for this experiment consists of a 36~liter feed tank, a UF,. absorption

6
vessel, a H2 reduction column, an effluent stream sampler, a 36-liter

receiver, NaF traps for collecting excess UF,_ and disposing of HF, gas

6

supplies for argon, hydrogen, and UF and means for analyzing the gas

6"
streams from the reaction vessels.

The experiment is operated by pressurizing the feed tank with argon
in order to displace salt from the feed tank to the UF6 absorption tank
at rates from 50 to BOchc/min. From the UF6 absorption tank, the salt
is siphoned into the H2 reduction column, and the salt then flows by
gravity through the effluent stream sampler to the receiver. The feed
salt will be LiF—BeF2~ThF4 (72-16-12 mole %) MSBR fuel carrier salt con-~
taining up to 0.3-mole % UF4. Absorption of UF6 by reaction with dig-
solved UF4 will occur in the UF6 absorption vessel, and the resultant
UFS will be reduced with hydrogen in the H2 reduction column. Uranium
hexafluoride flow rates from 60 to 360 scc/min,* and H2 flow rates from
60 to 720 scc/min will be used. The effluent salt will be collected in
the receiver for return to the feed tank at the end of the run. The off~
gas from the absorption vessel and the reduction column will be analyzed
for UF6 and HF, respectively. The salt in the feed tank and salt samples
from the column effluent salt will be analyzed for uranium. The perfor-
mance of the column will be determined from these analyses. The effluent

HF and any UF_ passing through the column will be collected on the NaF

o
traps before the gas is exhausted to the off-gas system.

*
Standard cubic centimeters per minute.
UFg ~—im

40

ORNL DWG 74-il666

 

 

 

 

 

 

 

 

 

 

 

 

"

1
|
!
1
|
|
{
|
I

FEED TANK

 

 

 

 

B

 

 

e

 

 

 

 

UFg ABSORPTION
VESSEL

 

 

 

Hp, REDUCTION

COLUMN

 

 

 

 

 

 

 

 

 

 

 

o
- & —
UFG HF UFg
anaALYSIS | [anALYsis HF TRAP
TRAP g
-
i > BUILDING

OFF-GAS

 

 

A= SYSTEM

EFFLUENT
STREAM
SAMPLER

 

 

 

 

RECEIVER

Fig. 10. Flow diagram for equipment used in fuel reconstitution
engineering experiment.
41

The fuel reconstitution engineering experiment will be installed in
the high bay area, Bldg. 7503 (MSRE site). Scaffolding erected for a
previous experiment will serve the FREE equipment adegquately. The system,
except for the upper section of the column, will be enclosed by a splash

shield.

The UF6 absorption vessel will have an inside diameter of 4 in. and
an inside height of 11 in. The vessel will be coustructed from 4-in.
sched 40 nickel pipe mounted vertically with a 150~1b standard Monel
flange at the top and a 0.25-in. nickel plate bottom. Two 1/2-in.
nozzles provide for salt flow in and out of the vessel. Two 3/8-in.

nozzles provide for UF6(g) flow in and off-gas flow out of the tank.

TheH2 reduction column will have an overall height of 115 in. It
is constructed of 1-1/2-in. sched 40 nickel pipe mounted vertically with
a 150-1b Monel flange on the upper end, and a sched 40 nickel welded cap
at the lower end. Gas enters through a 3/8-in. sched 40 nickel side arm
at the bottom, while 1/2-in. and 3/8~in. nozzles provide for salt entrance

and an off-gas exit.

Because of the highly corrosive nature of dissolved UF5, all equip-
ment exposed to significant gquantities of UF5 will be gold or gold plated
after a start up and demonstration period using the equipment described

in this report.

The 36-liter feed tank will have an inside diameter of 10 in. and an
inside height of 33 in. It will be a nickel vessel with 0.25-in.-thick
walls, eguipped with two 1/2-in. nozzles for salt exit and entrance lines,
twe 3/8-in. nozzles for argon sparging and off-gas exit, and a 3/4~in.
sampling port. The receiver vessel will be essentially identical to the

feed tank.

All piping exposed to the liguid phase (MSBR carrier salt) or tempera-
ture in excess of 100°C will be nickel pipe or tubing. 0Off-gas lines

carrying UF_ or HF will be nickel tubing. All other gas lines will be

6
copper tubing. O0ff-gas lines and liquid phase piping will be 1/2-in.

tubing.
42

Standard brass valves can be used throughout the gas system, except
on gas lines carrying UF6 or HF where nickel bellows-seal valves are
required. Stainless steel, full-bore ball valves will be used to intro-

duce the sample ladles into the flowing stream salt sampler.

Two sodium fluoride traps are required for UF6 absorption and for HF
absorption. These traps are identical except for their lengths. The
UF6 removal trap has a length of 55-5/8 in., while the HF trap has a
length of 31-5/8 in. The traps are constructed of 4-in. sched 40 Monel
pipe mounted vertically with 150-1b standard flanges at both ends for
charing and removing NaF pellets. The gas enters from 1/2-in. nickel
tubing through the flanged end and exits through the lower flanged end.

A 6-in. section in the lower end of the trap contains 4-in.-diam Monel

York mesh to keep pellets from plugging the exit.

The experiment will be monitored by analyzing the off-gas from each
leg of the column. GOW-MAC gas density detectors are being considered
for this., The detector elements of this analysis system are not exposed

to the measured gas stream, which in this case is corrosive.

The liquid phase will be sampled periodically by withdrawing effluent-
stream salt samples from the hydrogen reduction column. The sampler is a
vertically mounted 3/4-in. sched 40 pipe, 20-3/4 in. in length; it has a
stainless steel full-bore valve located 12-1/4 in. from the lower end to
allow access to the flowing salt stream. The sampler is vented to the
off-gas system above and below the ball valve. The salt flows into the
lower end of the sampler from 1l/2-in. tubing and exits 1-1/4 in. from the
end, again using 1/2-in. tubing. This assures a minimum of 1/4 in. of
salt for insertion of a sampler. The sampler is a stainless steel 0.250~-
in. diam by l-in.-long tube with a 0.050-in. fritted metal filter on the
lower end. It is connected to a 1/16-in. stainless steel tube which in

turn is attached to a vacuum pump.

Engineering sketches for all of the equipment except the UF6 supply
have been completed. The scaffolding in Bldg. 7503 is being cleared to

make room for this experiment.
10.

11.

12.

13.

14.

15.

16.

43

6. REFERENCES

H. C. Savage, Engineering Development Studies for Molten-Salt Breeder
Reactor Processing No. 18, ORNL-TM-4628 (March 1975), pp. 23-36.

 

 

J. A. Klein et al., MSR Program Semiannu. Progr. Rept. Aug. 31, 19272,
ORNL-4832, p. 171.

 

J. B. Lewis, Chem. Eng. Sci. 3, 248-59 (1954).
J. B. Lewis, Chem. Eng. Sci. 3, 260-78 (1954).
H. 0. Weeren et al., Engineering Development Studies for Molten-3alt

Breedex Reactor Processing No. 9, ORNL-TM-3259 (December 1972}, pp.
205-15.

 

 

W. L. Carter et al., MSR Program Semiannu. Progr. Rept. Feb. 29, 1972,
ORNL-4782, pp. 224-25.

 

J. A. Klein, Engineering Development Studies for Molten-Salt Breeder
Reactor Processing No. 18, ORNL-TM-46%8 (March 1975), pp. 1-22.

 

 

J. A. Klein, in Engineering Development Studies for Molten—-Salt Breeder
Reactor Processing No. 15, ORNL-TM-401% (in preparation).

 

 

O. Levenspiel, Chemical Reaction Engineering, p. 387, Wiley, New York,
1962,

 

B. A. Hannaford, C. W. Kee, and L. E. McNeese, MSR Program Semiannu.
Progr. Rept. Feb. 28, 1971, ORNL~-467&6, p. 256.

 

 

B. A. Hannaford et al., MSR Program Semiannu. Progr. Rept. Aug. 31,
1971, ORNL-4728, p. 212.

 

B. A. Hannaford, C. W. Kee, and L. E. McNeese, Engineering Development
Studies for Molten-Salt Breeder Reactor Processing No. 8, ORNL-TM-
3259 (May 1972) p. 64.

 

 

B. A. Hannaford, C. W. Kee, and L. E. McNeese, Engineering Development
Studies for Molten—-Salt Breeder Reactor Processing No. 9, ORNL-TM-
3259 (December 1972), p. 158.

 

B. A. Hannaford, Engineering Development Studies for Molten-Salt Breeder

 

Reactor Pyocessing No. 10, ORNL-~TM-3352 (December 1972), p. 12.

 

H. O. Weeren and L. E. McNeese, Engineering Development Studies for
Molten-Salt Breeder Reactor Processing No. 10, ORNL-TM-3352 (December
1972}, p. 53.

 

 

D. E. Ferguson, Cnem. Technol. Div. Annu. Progr. Rept. March 31, 1972,
ORNL-4794, p. 1.
20.
21.
22,
23.
24.
25,
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41-66.
67.

45
ORNL~TM-4863

UC-76 — Molten Salt Reactor Technology

INTERNAL DISTRIBUTION

MSRP Director's Office 68. R.

W. Horton
C. F. Baes, Jr. 69. W. R. Huntley
C. E. Bamberger 70. C. W. Kee
J. Beams 71. A. D. Kelmers
M. Bender 72. J. A. Klein
M. R. Bennett 73. W. R. Laing
E. S. Bettis 74. R. B. Lindauer
R. E. Blanco 75. R. E. MacPherson
J. O. Blomeke 76. W. C. McClain
E. G. Bohlmann 77. H. E. McCoy
J. Braunstein 78. A. P. Malinauskas
M. A. Bredig 79. C. L. Matthews, ERDA-OSR
R. B. Briggs 80. A. S. Meyer
H. R. Bronstein 8l. R. L. Moore
R. E. Brooksbank 82. J. P. Nichols
C. H. Brown, Jr. 83. K. J. Notz
K. B. Brown 84. R. O. Payne
J. Brynestad 85. H. Postma
S. Cantor 86. M. W. Rosenthal
D. W. Cardwell 87. A. D. Ryon
W. L. Carter 88. H. C. Savage
W. H. Cook 90. C. D. Scott
R. M. Counce 9l1. M. J. Skinner
J. L. Crowley 92. F. J. Smith
F. L. Culler 93. G. P. Smith
J. M. Dale 94. I. Spiewak
F. L. Daley 95. M. G. Stewart
J. H. DeVan 96. 0. K. Tallent
J. R. DiStefanc 97. L. M. Toth
W. P. Eatherly 98. D. B. Trauger
R. L. Egli, ERDA-OSR 99. W. E. Unger
J. R. Engel 100. J. R. Weir
G. G. Fee 101. M. K. Wilkinson
D. E. Ferguson 102. R. G. Wymer
L. M. Ferris 103-105. Central Research Library
L. O. Gilpatrick 106, Document Reference Section
J. C. Griess 107-116. Laboratory Records
W. R. Grimes 117. Laboratory Records (ORNL~RC)
R. H. Guymon
J. R. Hightower, Jr.
B. F. Hitch
123.

124.

125-126.

127-230.

46

CONSULTANTS AND SUBCONTRACTORS

l18. J.
119. cC.
120. J.
121. Ww.
122. R.
EXTERNAL

Research and Technical Support Division, ERDA, Qak Ridge Opera-
E, Oak Ridge, Tenn. 37830
Director, Reactor Division, ERDA, Oak Ridge Operations Office,

tions Office, P. 0. Box

P. O. Box E, Oak Ridge,
Director, ERDA Division
Washington, D. C. 20545

For distribution as shown in TID-4500 under UC-76, Molten Salt

Reactor Technology

. Frye

Ice

Katz
Davis

. Richards

wR4gmxmn

DISTRIBUTION

Tenn. 37830

of Reactor Research and Development,