You are on page 1of 328

i f Si

UC-10 —Chemical Separations Processes
for Plutonium and Uranium
TID-4500 (45th ed.)






operated by
for the

This report was prepared as an account of work sponsored by an

agency of the United States Government. Neither the United States
Government nor any agency Thereof, nor any of their employees,
makes any warranty, express or implied, or assumes any legal
liability or responsibility for the accuracy, completeness, or
usefulness of any information, apparatus, product, or process
disclosed, or represents that its use would not infringe privately
owned rights. Reference herein to any specific commercial product,
process, or service by trade name, trademark, manufacturer, or
otherwise does not necessarily constitute or imply its endorsement,
recommendation, or favoring by the United States Government or any
agency thereof. The views and opinions of authors expressed herein
do not necessarily state or reflect those of the United States
Government or any agency thereof.

Portions of this document may be illegible in

electronic image products. Images are produced
from the best available original document.
This report was prepared as an account of Government sponsored work. Neither the United
States, nor the CommlsBlon, nor any person actii^ on behalf of the Commission;
A. Makes any warranty or representation, expressed or implied, with respect to the accu-
racy, completeness, or usefulness of the information contained In this report, or that the use
of any information, apparatus, method, or process disclosed in this report may not infringe
privately owned rights; or
B. Assumes any liabilities with respect to the use of, or for damages resulting from the
use of any Information, apparatus, method, or process disclosed in this report.
As used in the above, "person acting on behalf of the Commission" Includes any em-
ployee or Contractor of the Commission, or employee of such contractor, to the extent that
such employee or contractor of the Commission, or employee of such contractor prepares,
disseminates, or provides access to, any information pursuant to his employment or contract
with the Commission, or his employment with such contractor.

Contract No. W-7405-eng-26



u r mjcLEAR scimcE ABSTRACTS



for Period Ending May 31, 1965

D. E. Ferguson - Division Director

R. G. Wymer - Chemical Development
Section A Chief
R. E. Blanco - Chemical Development
Section B Chief
K. B. Brown - Chemical Development
Section C Chief
M. E. Whatley - Unit Operations Section Chief
H. E. Goeller - Process Design Section Chief
R. E. Brooksbank - Pilot Plant Section Chief



Oolc Ridge, T e n n e s s e e
operated by
for the



1.1 Hydrolysis of Uranium and Thorium Carbides
1.2 Development of P r o c e s s e s for Graphite-Base Fuel
1.3 Studies on the P r o c e s s i n g of Uranium-Plutonium Oxide Fuel
1.4 Studies on the P r o c e s s i n g of Thorium-Uranium Fuel
1.5 Chloride Volatility Studies with Uranium and Plutonium
1.6 Development of Mechanical P r o c e s s e s
1.7 Development of Close-Coupled P r o c e s s e s
1.8 Conceptual Plant Studies


2.1 Molten-Salt P r o c e s s i n g of Uranium-Aluminum Alloy Fuel
2.2 Volatilization of P u F from Molten Salts Containing High Concentrations
of Uranium
2.3 Recovery of Plutonium and Uranium from Falling Molten-Salt Droplets and
from Beds of Solid P a r t i c l e s by Fluorination
2.4 Fluidized-Bed Volatility P r o c e s s Development for Stainless-Steel- and
Zircaloy-Clad UO^ F u e l s
2.5 Design of a Fluidized-Bed Pilot Plant for Zirconium- and Stainless-Steel-Clad
UO Power-Reactor F u e l s
2.6 Sorption of P u F ^ by Metal Fluorides
2.7 Sorption of UF^ by N a F
2.8 Continuous In-Line Monitoring of Gas Streams in Fluoride Volatility Processin
2.9 Vapor-Liquid Equilibria of UF -NbF System
2.10 P h a s e Equilibria Studies Pertinent to Fluoride-Volatility P r o c e s s i n g of F u e l s
or Mixtures Containing Chromium
2.11 Volatilization of UF,, from Sol-Gel-Derived ThO -UO^
o 2 1
2.12 General Corrosion Studies


3.1 Conversion of High-Level Radioactive Wastes to Solids
3.2 Disposal of Intermediate»Level Radioactive Waste


3.3 Treatment of Low-Level Radioactive Wastes 119

3.4 Engineering, Economic, and Safety Evaluation 134


4.1 Development of Chemical P r o c e s s e s 140
4.2 Development of P r o c e s s Equipment 148
4.3 Design and Fabrication of P r o c e s s Equipment 151
4.4 Construction of the Transuranium P r o c e s s i n g Plant 152


5.1 P r o c e s s Development 157
5.2 Curium Recovery Facility: Equipment and Flowsheet Testing 160
5.3 Solvent Extraction P r o c e s s i n g of ^''^Am and ^^''Cm 161
5.4 Ion Exchange Purification and Separation of ^''^Am-^''''Cm 166
5.5 P r o c e s s i n g of ^"'AmO2 Targets 168
5.6 Development of Alternative P r o c e s s e s : Separation of Lanthanides and
Actmides 172


6.1 " ^ U Storage and Distribution Facility 174
6.2 Development of the Sol-Gel P r o c e s s 174
6.3 Development of Methods for Producing Microspheres 177
6.4 Design of the Thorium-Uranium Recycle Facility 182
6.5 Development of Equipment for the Thonum-Uranium Recycle Facility 185


7.1 Extraction of Metal Chlorides by Amines 193
7.2 Extraction of Metal Nitrates by Amines 193
7.3 Reagents for Separating Polonium(IV) from Bismuth 195
7.4 New Extraction Reagents 197
7.5 Performance of Degraded Reagents and Diluents 197
7.6 Biochemical Separations 200
7.7 Extraction Properties of Lanthanide and Actinide Complexes 201
7.8 Recovery of Rubidium from Ores 203
7.9 Recovery of Beryllium from Ores 205
7.10 Recovery of Acid by Amine Extraction 207
7.11 Recovery of Thorium from Granitic Rocks 207
7.12 Separation of Biological Macromolecules 210
7.13 Extraction of Alkaline Earths by Di(2-ethyhexyl)phosphoric Acid 210
7.14 Intermolecular Bonding in Organic Mixtures 211
7.15 Synergism in the Extraction of Strontium by Di(2-ethylhexyl)phosphoric Acid 213
7.16 Synergism and Diluent Effects in the Extraction of Cesium by 4-sec-
Butyl-2-a-methylbenzylphenol (BAMBP) 214

7.17 Kinetics of Extraction of Iron by Di(2-ethylhexyl)phosphoric Acid 217

7.18 Activity Coefficients of the Solvent P h a s e s 218


8.1 Cesium 222
8.2 Isolation of Cerium and Promethium 224
8.3 Ruthenium 226
8.4 Treatment of Waste that Contains Dissolved Stainless Steel or Large

Amounts of Aluminum Ion 226

8.5 Engineering Studies 226


9.1 Extraction from Sulfuric Acid Solutions 228
9.2 Absorption Spectra in Sulfuric Acid Solutions 230
9.3 Recovery of Protactinium 231
10.1 Radiolysis of Water Adsorbed on Silica Gel 233
11.1 Fabrication Status of the High-Temperature, High-Pressure Spectrophotometer
System 236
11.2 Spectral Studies of Ionic Systems 237
11.3 Measurement of Liquid Densities at High Temperature and High P r e s s u r e s 241
11.4 Computer Programs for Spectrophotometer Studies 242


12.1 Activities of the Three Components in the System: Water—Nitric Acid-Uranyl
Nitrate Hexahydrate at 25°C 248
12.2 Activities of Tributyl P h o s p h a t e in Mixtures Containing Tributyl Phosphate,
Uranyl Nitrate, and Water 248
12.3 Computer Resolution of Spectrophotometric Data on Uranyl Nitrate—Nitric
Acid—Water Systems 250


13.1 The Stacked-Clone Contactor 252
13.2 Effect of Ionizing Radiation on Coalescence in Liquid-Liquid Systems 253
13.3 In-Line Detection of P a r t i c l e s in Gas Streams by Scattered Light 256
13.4 Solvent Extraction Engineering Studies: Correlation of Pulsed-Column
Flooding Data 259


14.1 Studies of the Cost of Shipping Reactor F u e l s 261
14.2 Studies of the Cost of P r o c e s s i n g Fuel 263
14.3 Studies of the Cost of Preparing Fuel by the Sol-Gel P r o c e s s 266
14.4 Computer Code for Studies of Overall Power Costs 268

14.5 The TRUFIZ Computer Code 269

14.6 Manual of Shipping-Cask Design 269
14.7 Criticahty Control in Fuel Cycle P l a n t s 269


15.1 Project Coach 271
15.2 Focusing Electrophoresis for Separation of Heavy Elements 272
15.3 Hypervelocity Jet-Sampler Development 272
15.4 Feasibility of Producing Chemicals with Nuclear Explosives 272
15.5 Stimulation of Natural Gas Wells by Nuclear Devices: Project Gasbuggy 273
15.6 Magnesium Ores 273
15.7 Copper Ores 274


16.1 Fundamental Studies of Thoria Sols and Gels 275
16.2 Surface and Sintering Properties of Sol-Gel Thoria 281


17.1 Eurochemic A s s i s t a n c e Program 285
17.2 Projects for Improving ORNL Waste Systems 285
17.3 A Demonstration of the Disposal of Solid, High-Level Radioactive Waste
in Salt 286
17.4 Disposal of Radioactive Waste by Hydrofracturing 287
17.5 Construction and Startup of the High Radiation Level Analytical Laboratory 288
17.6 High-Level Alpha Laboratory 289
17.7 Storage Facility for " S y , Building 3019 289
17.8 Filter T e s t s in the Nuclear Safety Pilot Plant 290
17.9 Studies of Radiation Shields for Fission Sources 291
17.10 Irradiation and Decontamination Evaluation T e s t s on Selected Protective
Coatings 296


18.1 Continuous Fluorination of Molten Salt 300
18.2 Distillation of Molten Salt 301
18.3 Reconstitution of Fuel for a Molten-Salt Breeder Reactor 302
18.4 Chromium Fluoride Trapping 302

18.5 Alternative P r o c e s s i n g Methods for a Molten-Salt Breeder Reactor 303





1. POWER REACTOR FUEL PROCESSING impurity (up to 8 wt %), the amount of hydrocarbons
produced in hydrolysis decreased, and the produc-
In the development of new nuclear power reactor tion of hydrogen and waxes increased. The im-
concepts, fuels that can be used for long times purities were present a s WUC and W C (in
and at very high temperatures are continuously thorium).
being sought. This sometimes leads to the use of
fuels for which the chemical reactions encountered
in fuel processing are not well understood. The 1.2 Development of Processes for
development of new recovery p r o c e s s e s for such Graphite-Base Fuel
fuels is directed toward the demonstration of chemi-
cal methods and engineering practices for recover- Work on graphite-base fuels last year was devoted
ing and decontaminating the uranium, thorium, and to the development of burn-leach p r o c e s s e s for
plutonium. In the studies described here, two ap- pyrolytic-carbon-coated ThC -UC and ThO -UO^
proaches are being followed: the development of fuel particles contained in graphite. Two alterna-
both mechanical and chemical head-end methods tives were studied: (1) burning at 750°C in fluid-
designed to produce suitable aqueous feeds for ized beds followed by leaching in boiling nitric
solvent extraction and the development of non- acid and (2) submerged combustion in oxygenated
aqueous p r o c e s s e s such a s chloride volatility, dilute nitric acid at 300°C and 1500 to 2000 psi.
which may either precede or circumvent solvent The latter process is called the Pressurized Aque-
extraction or fluoride volatility p r o c e s s e s . ous Combustion (PAC) p r o c e s s .
In the burn-leach process, uranium and thorium
are recovered from graphite-matrix fuel by burning
1.1 Hydrolysis of Uranium and Thorium Carbides the graphite in a fluidized bed of alumina and dis-
solving the resultant uranium and thorium oxides
Fundamental studies of the hydrolysis of uranium in a strong acid leachant to provide feed for solvent
and thorium carbides are continuing. High-purity extraction processing. Prototype HTGR fuel com-
samples of the uranium carbides UC, U (C ) , and pacts were burned and leached in engineering-scale
UC when hydrolyzed in 4 to 16 M HNO at equipment, with 99.7% recovery of the uranium and
90°C, produced uranyl nitrate, soluble organic thorium. The burning was done in bench s c a l e in a
acids (mellitic and oxalic), and an off-gas con- 2-in.-diam fluidized bed of alumina at a rate of
s i s t i n g of NO^, NO, and CO . The samples were 20 g-moles of carbon per minute per square foot of
passive to treatment in 0.001 to 0.5 M HNO^. Tho- c r o s s section to produce a bed containing 28% of
rium carbide samples in the range of ThC to combined metal oxides. The bed was leached at
The were prepared; below ThC the samples 110 to 118°C in a 1.5-in.-diam upflow column, with
contained Th metal and ThC . The dicarbide was provision for recycling the leachant (13 M H N O ^ -
found to have the composition Th rather than 0.1 M F - - 0 . 0 4 A/A13+).
ThC 2 . When hydrolyzed in water between 25 In parallel laboratory leaching t e s t s on mixtures
and 99°C, ThC produced hydrogen and methane, of alumina and combustion ash containing 6%
and T h e produced C to C hydrocarbons and UjOg, 25% ThO^, and 69% Al^Oj, 99.5% of the
waxes. When samples of uranium and thorium car- uranium and thorium were recovered as 0.6 M
bides were prepared with tungsten carbide a s an Th(N03)_, in 5 to 7 hr in boiling ^ 4 A/ HNO3-O.O2


to 0.05 M H F - 0 . 0 to 0.1 W M{NO^)^. About 2% of fluoride, 2% of the alumina, and 16 to 85% of the
the alumina dissolved. Since coated-particle ThO - iron in the bed product was a l s o leached; u s e of
UO^ microspheres s e t t l e to the bottom of a fluidized higher HNO concentrations increased the solu-
bed, leaching t e s t s on undiluted microspheres were bility of iron. To reduce the amount of fluoride
made; complete dissolution to 0.5 M Th(NO ) was in the decladding product, the bed was pyrohydro-
achieved in 3 to 6 hr in boiling 13 M HNO^-O.OS M lyzed with steam or with 0 ^ - 3 % H^O at 200 to
H F . Lower nitric acid concentrations were much 600°C for 3 hr. About 85% of the fluorine was re-
less effective as l e a c h a n t s . moved. The beds were then leached for 10 hr in
PAC process t e s t s were made in both an auto- boiling 10 M HNO3 to produce 0.2 M HNO^-O.OS
clave and in a semicontinuous chemical reactor. M HF solutions; sufficient alumina was also
In this process, the nitric acid is not consumed leached to complex the H F . Uranium recovery
when an e x c e s s of oxygen i s present. In 24-hr amounted to 99.7 to 99.9%, and the iron content
autoclave t e s t s with HTGR fuel compacts contain- was reduced to 6%.
ing mixed thorium-uranium carbide in 2 M HNO , As to the Sulfex decladding process, hot-cell
all the carbon was burned to CO , and all the fuel t e s t s with fast-reactor fuel samples (irradiated up
dissolved when the O -to-C and NO ~-to-(Th + U) to 99,000 Mwd per ton of uranium), containing 20%
ratios were = 1 and =50 respectively. The high PuO —80% UO clad in s t a i n l e s s s t e e l , showed
NO ~-to-(Th + U) ratio i s required b e c a u s e aque- that it cannot be applied satisfactorily to this type
ous uranium and thorium nitrates hydrolyze readily of fuel, since 1.5% of the uranium and plutonium in
at high temperatures. The pyrolytic carbon coat- the fuel was dissolved by the boiling 6 M H SO de-
ings on fuel particles are not attacked until all cladding reagent. Thus, the waste solution produced
matrix carbon is oxidized. The oxidation rate for by the Sulfex process would have to be specially
unfueled moderator-grade graphite is much lower, processed to recover this l o s s . In addition, over
0.15 to 0.24 g - g " i - h r - ' , than the rate for fueled half of the ' ^ ' C s appeared in the decladding waste
graphite. In t e s t s with carbon-coated 97% ThO —
solution; the presence of this long-lived gamma
3% UO^ ^ microspheres, complete dissolution was
emitter would materially increase the c o s t of per-
obtained in 24 hr in 2 A/ HNO3 at 300°C when the
manent disposal of decladding w a s t e .
N03""-to-(Th + C) mole ratio was 100. Under simi-
Parallel t e s t s of the shear-leach process on iden-
lar conditions, only 49% of pure ThO was d i s -
tical fuel samples were very successful. The ir-
solved in 24 hr.
radiated samples dissolved at practical rates in
During the first half of last year, development of boiling 3.5 to 10 M HNO without fluoride catalyst,
a burn-leach and a burn-volatility process for which was not the c a s e with t e s t s on unirradiated
graphite-base Rover fuel was completed. UO -PuO samples. Leaching of the oxide core
from both Sulfex-declad pellets and chopped fuel
pins was accomplished in 3 hr with boiling 5 M
1.3 Studies on the Processing of
HNO . The small residue of refractory material
Uranium-Plutonium Oxide Fuel
remaining with the s t a i n l e s s s t e e l hulls retained
In the laboratory, acid-leaching studies were made only 0.13% of the uranium and plutonium originally
on stainless-steel-clad oxide fuel declad by a new in the fuel sample.
process developed under the Fluoride Volatility Related dissolution t e s t s with unirradiated high-
Processing program (see Sect. 2). In this process, density PuO microspheres in boiling 14 M HNO —
the fuel assemblies are contacted with 40% HF— 0.5 M HF resulted in the very low dissolution rate
60% O^ at 600°C in a fluidized bed of alumina. of 7 X 1 0 " ^ mg m i n ~ ' cm~^; only 8.5% of the
The cladding i s converted to " s t a i n l e s s s t e e l " PuO^ dissolved in 37 hr.
oxides and fluorides and the UO -PuO fuel to
U . O -PuO powder. A typical fluidized-bed de-
1.4 Studies on the Processing of
cladding product contained 25% U^O , 13% stain-
Thorium-Uranium Fuel
l e s s s t e e l compounds, 5% fluorine, and 67% Al O .
When leached for 5 hr in boiling 1 to 15 A/ HNO Sulfex decladding and shear-leach studies were
that was S i A/ in A1(N03)3, 99.6% of the uranium made with both unirradiated and irradiated stain-
was dissolved to produce a leachate 0.2 M in ura- less-steel-clad ThO —4% UO^ prepared by the sol-
nium and 0.6 M in fluoride. About 80 to 90% of the gel process. Hot-cell t e s t s on the dissolving of

irradiated samples [3000 to 98,000 Mwd per metric UO^-PuOj pellets with 85% 0 1 ^ - 1 5 % CCl^ was
ton of (Th + U)] in boiling Thorex reagent [13 M unsuccessful. Therefore, before chlorination, the
HNO3-O.O4 M H F - 0 . 0 4 M A1(N03)3] indicated that pellets were oxidized at 500 to 800°C in oxygen
irradiation markedly i n c r e a s e s the dissolution rate to produce U 0 -PuO powder, but only pellets
over that for unirradiated oxide. For example, more containing l e s s than 20% PuO were affected by
than 99.8% of the thorium and uranium was d i s - this treatment. Pure UjO^ was e a s i l y converted
solved in 24 hr, as compared with 75 hr for nearly to volatile uranium chlorides at 400 to 500°C.
complete dissolution of unirradiated material. The P e l l e t s containing 99% U 3 0 g - 1 % PuO^ were nearly
macro amounts of fission products from t h e s e highly unreactive at 300 to 350°C; at 400 to 500°C all
irradiated specimens were soluble in the product the uranium and 50% of the plutonium were volati-
solutions, which were 1 A/ in Th(NO ) , 10 A/ in lized in 3 to 4 hr. Complete volatilization was
HNO3, and 0.05 M in total fission products. Cyclic achieved at 600°C, but sintering and melting
dissolving, which left a 6% h e e l of undissolved (probably of UO CI ) were observed. About 3.3 x
thoria-urania, produced a 1 A/ thorium product solu- 10^ and 100 moles of chlorine were required to
tion after 5 hr. The residues consisted of undis- transport 1 g of Pu at 427 and 727°C respectively.
solved thoria-urania, cladding, corrosion products,
and contaminants from the fuel rod, such a s insula-
1.6 Development of Mechanical Processes
tion and alumina. The thorium and uranium could
be recovered from the residue by leaching with
Since second-generation power-reactor fuel a s -
fresh Thorex reagent. In the Sulfex decladding
semblies cannot be sheared satisfactorily when in-
t e s t s , there was a 0.33% loss of thorium and ura-
tact, the rods must be removed before they are
nium to the decladding solution.
sheared. Pulling single rods is too slow, and
Protactinium was successfully recovered by ad- pulling all rods at one time crumples the assembly
sorption on unfired Vycor from thorium-uranium sheath. Therefore, a device was designed and
nitrate fuel solutions prepared from specimens ir- built for pulling single rows of rods. L o o s e arrays
radiated to 75,000 Mwd/metric ton and cooled 26 of 64 pulled rods were easily sheared in the 250-
days before processing. In three hot-cell experi- ton prototype shear. Shearing t e s t s on 3-in.-diam
ments, the specimens were first dissolved in re- Zircaloy-2-clad uranium-metal annular fuel rods
fluxing 13 M HNO3-O.O5 M H F , and the protactinium were equally satisfactory. The shearing force was
was then preferentially adsorbed on columns of 82 tons; 30% of the area was sheared through be-
60- to 80-mesh unfired Vycor g l a s s . More than fore the rod was broken. A new feeding device,
99.4% of the protactinium was adsorbed, and more external to the shear, successfully replaced the in-
than 95% was eluted with 0.5 M oxalic acid. The ternal stop on the shear, which was hard to repair
protactinium was decontaminated of zirconium- with remote manipulators.
niobium, total rare-earth beta emitters, and ruthe-
A mockup was used to demonstrate the remote
nium by factors of 9, 6000, and 2800 respectively.
handling and transfer of a filled sheared-fuel basket
Evaluation of other inorganic adsorbents was between the shear and dissolver. All operations
continued. The distribution coefficients for pro- went smoothly. Also, experiments and calcula-
tactinium were measured for s i l i c a gel and Bio-Rad tions showed that the fission product heating of
ZP-1 and for 0.5- to 12-M HNO3 solutions con- stored sheared fuel will not be e x c e s s i v e and will
taining 4, 50, and 100 g of thorium per liter. The not c a u s e pressure surges in the dissolver when
coefficients ranged from 1400 to 5800 for the s i l i c a quenched.
gel and from 100 to 29,000 for Bio-Rad Z P - 1 . As Abrasive-disk sawing of Zircaloy-2 assembly end
much as 95% of the protactinium was unadsorbable fittings in air was shown to present a definite fire
by Bio-Rad Z P - 1 in the more dilute nitric acid hazard; sawing under water sprays was somewhat
solutions. safer but produced fines in a s i z e range (=400 fi)
considered dangerous. However, a 7:1 automatic
dilution by abrasive material is believed to l e s s e n
1.5 Chloride Volatility Studies with Uranium
the risk.
and Plutonium Oxides
Leaching of s t a i n l e s s - s t e e l - c l a d UO sheared
In laboratory studies at temperatures up to 400 fuel in boiling 7 M HNO 3 in a 90-liter batch thermo-
to 600°C, the direct chlorination of unirradiated siphon leacher was complete in 2 hr. Leaching of

ThO^-UOj in 13 M HNO3-O.O4 A / F - - 0 . 1 M Al^* tion was contacted first with 3% T B P in n-dodecane

was much slower. The leaching rate decreased to extract uranium and then with 30% T B P in n-
with thorium loading and increased with tempera- dodecane to recover thorium. Uranium and thorium
ture. The length of the sheared section had little recoveries were 99.9 and 87%, respectively; ura-
effect on the rate. nium decontamination factors from Th, total rare
A leached-hull monitor, using delayed-neutron earths (TRE), Zr-Nb, and Ru were 16, 2.6 x 10",
activation a n a l y s i s , was developed. It has a s e n s i - 300, and 500 respectively. Thorium decontamina-
tivity of about 1 mg of " S u , 2 3 9pu^ ^^ 2 3 3u per tion factors from U, TRE, Zr, and Ru were 7 x lO'',
kilogram of leached h u l l s . This corresponds to a 16, 6, and 4 respectively. One feature of the new
loss of l e s s than 0.01% of the fuel in a typical low- unit i s that it can be shut down and restarted very
enriched uranium power-reactor assembly. quickly. Tentative flowsheets for the recovery of
U and Th either separately or together in a mixer-
settler containing three extraction, one scrub, and
1.7 Development of Close-Coupled Processes three strip s t a g e s were evolved; di-sec-butyl phenyl-
phosphonate (DSBPP) in diethyl benzene (DEB)
When uranium, thorium, and plutonium are recycled was used. About 99% U recovery was obtained,
in nuclear reactors, short-lived isotopes are built with a thorium-decontamination factor of 6000. New
up which decay rapidly through a s e r i e s of high- flowsheets, calling for either solvent extraction
energy gamma-emitting daughters. Since t h e s e i s o - (DSBPP in DEB) or ion exchange (Dowex 50), were
topes cannot be separated easily from the major developed to separate stored ^ s a y ffQjj, (-he daughter
isotopes in which they are contaminants, the re- isotopes of ^^^U. Immediately after processing,
covered fuels are very gamma active and must be the solvent extraction process resulted in a DF of
reconstituted into replacement fuel assemblies by 1000, and the ion exchange process resulted in a
remote manipulation. Thus, high decontamination DF of 2. However, after two weeks, the activity of
from fission products may not be necessary. A both products was the same. The reduction of
new program was begun this year to develop low- Pu(IV) to Pu(III) in nitrate solution was s a t i s f a c -
cost, low-decontamination p r o c e s s e s capable of torily accomplished with 4% H —Ar in a platinized-
achieving factors of 10 to 100 from fission products. alumina bed. This method avoids the use of ferrous
Emphasis has been placed primarily on seeking a sulfamate or uranous ion.
simple process which will effect reductions in
A new method was developed to remove nitrate
capital, operating, and analytical c o s t s . Attempts
from thorium-uranium nitrate solutions, a required
are also being made to close-couple the head-end step in the sol-gel p r o c e s s . The nitrate was ex-
and separations steps with subsequent operations tracted with either 0.2 M Primene JM-T in n-dodec-
such as the sol-gel process. ane or 0.2 M 1-nonyldecylamine in n-dodecane.
Laboratory experiments with thorium-uranium The nitrate-to-thorium mole ratio in the aqueous
nitrate fuel solutions in studies to substitute pre- phase approached 0.2 a s a limit. Sol-gel ThO
cipitation, ion exchange, or electrodialysis for prepared from the amine-extraction product i s now
solvent extraction were generally unsatisfactory, being evaluated. Amine extraction may be a pos-
and in all c a s e s decontamination factors ( D F ' s ) sible alternative to presently used steam denitra-
l e s s than 10 were obtained. Precipitants evaluated tion.
included dibasic ammonium phosphate, oxalic acid,
and 30% hydrogen peroxide. Ion exchangers tested
included Dowex 1, Dowex SOW, J P - 1 , and s i l i c a 1.8 Conceptual Plant Studies
gel. Although electrodialysis was also u n s u c c e s s -
ful in decontaminating fuel solutions, it may be Two conceptual plant studies were prepared dur-
useful in separating fission products in aqueous ing the past year. The first, for a small on-site
wastes into useful fractions; for example, a mixture plant associated with a 500- to 3000-Mw (electrical)
was separated into fractions containing Zr-Nb and boiling-water reactor station using slightly en-
Ru; Eu, Ce, and Sr; La; and Cs by appropriate pH riched fuel, was based on use of existing technol-
control. ogy. The second, concerning a plant for processing
A simple, efficient differential solvent extraction HTGR fuel, provided a design for a bum-leach
contactor was developed and satisfactorily tested. head-end process to be a s s o c i a t e d with a solvent
A 0.2 M acid-deficient 1.5 A/ T h - O . l M U fuel solu- extraction plant.

2. FLUORIDE VOLATILITY PROCESSING pilot-plant operations. Uranium l o s s e s ranged

from 0.1 to 0.9%, and averaged about 0.6%. De-
The investigation of fluoride-volatility processes contamination from fission products was quite
at ORNL is part of an intersite program to develop good. In one run, decontamination factors were
an alternative to aqueous methods for the recovery generally in the 10'^ to 10^ range; in all other
of values from spent nuclear reactor fuels. The runs, they ranged from 10* to 1 0 ^ ° . Chemical
program here is presently in transition between purity of the product UF was lower than desirable;
molten-salt and fluidized-bed methods. principal contaminants were fluorides of molyb-
Except for laboratory experiments on removal denum, aluminum, and sodium. Dissolver corrosion
of plutonium and perhaps protactinium from molten during the program was about 5 mils, c o n s i s t e n t
s a l t , work on molten-salt methods is finished. with previous experience. The r e l e a s e of radiation
The final study in the pilot plant was a demonstra- emitters to the atmosphere and the radiation dosage
tion of the recovery of uranium from fully irra- to personnel were controlled satisfactorily, even
diated, short-decayed aluminum-base fuel elements. though decay at the time of processing was l e s s
The AEC has announced a goal of having a than would normally be expected.
fluoride-volatility technology for processing low-
enrichment UO fuel completely developed through
" c o l d " engineering, " c o l d " semiworks, and " h o t " 2.2 Volatilization of P u F , from Molten Salts
pilot-plant programs by July 1, 1969. Our princi- Containing High Concentrations of Uranium
pal contribution to this effort will be the installa-
tion and operation of a fluidized-bed volatility
The volatilization of plutonium is an e s s e n t i a l
pilot plant in Building 3019. F i r s t priority in
part of a molten-salt volatility process for low-
pilot-plant experiments will be given to studies of
enrichment fuel. Although P u F ^ recoveries greater
the HCl process for removal of Zircaloy cladding
than 99% have been demonstrated in previous
as developed at Argonne National Laboratory.
laboratory t e s t s , 20 to 30 hr of fluorination was
Design of the pilot plant has begun. Supporting
required. This presents a serious corrosion prob-
laboratory- and bench-scale s t u d i e s are a l s o being
lem. A possibly useful method for shortening the
conducted. Thus far, laboratory s t u d i e s here have
time required for volatilization would be the main-
been mostly concerned with the u s e of H F - 0
tenance of a high concentration of uranium in the
for removing s t a i n l e s s s t e e l cladding from UO
melt by using a UF -F mixture during fluorination.
fuel elements. The retention of plutonium by
In one t e s t , with a melt containing initially 25%
alumina was studied in a miniature t e s t reactor
uranium, 32% of the P u F was evolved in 1 hr of
only \ in. in diameter.
fluorination, compared with about 15% in a previous
t e s t with no uranium present. T h e s e r e s u l t s , how-
ever, were not encouraging enough to warrant
2,1 Molten-Salt Processing of
further study.
Uranium-Aluminum Alloy Fuel

Development of a molten-salt fluoride-volatility

process for aluminum-base fuels has been com- 2.3 Recovery of Plutonium and Uranium from
pleted. Phase-diagram data for the system KF- Falling Molten-Salt Droplets and from Beds of
ZrF -AlF have been refined in the region of Solid Particles by Fluorination
interest. The effect of NaF addition to waste
salt was determined. Finally, a pilot-plant de- Additional laboratory development work was con-
velopment program was e s s e n t i a l l y completed. ducted to t e s t the feasibility of recovering fissile
This program culminated in the processing of fuel material from molten-fluoride mixtures by spraying
cooled only 25 days. the melt downward into fluorine gas at high tem-
The molten-salt fluoride-volatility process con- peratures. Such a method would reduce corrosion
s i s t s in dissolution of fuel elements in a molten and achieve the large ratio of fluorine gas to mol-
fluoride s a l t , fluorination to s e p a r a t e the uranium ten s a l t that is n e c e s s a r y for rapidly recovering
(as UF ) from the s a l t and most of the fission plutonium from power-reactor fuels. This type of
products, and further purification and recovery of molten-salt—gas contactor may a l s o be useful in a
the UF . The process proved satisfactory in continuous processing scheme.

In uranium fluorinations, volatilization of more ding agent for either Zircaloy or s t a i n l e s s s t e e l ,

than 99.9% of the contained uranium was achieved and of plutonium and uranium volatilization in the
with molten-salt droplets 150 fi or l e s s in diameter. subsequent fluorination treatment. A small fluid-
Fluorinations were made in a tower 5 ft long, and ized bed was developed to expedite some of the
the temperatures were between 550 and 650°C. t e s t s . This mini-test unit i s V, in. in diameter and
Fluorinations of five s a l t blends have been com- u s e s only about 5 to 8 g of bed material.
pleted. A correlation equation applicable to this Use of H F - 0 mixtures for decladding s t a i n l e s s
type of uranium fluorination is presented. s t e e l or Zircaloy-clad fuel results mainly in sin-
As to plutonium fluorination, five experiments tered oxides that convert only slowly to the cor-
were made in which plutonium was volatilized from responding fluorides. The oxides retain to a large
falling droplets of 50-50 mole % NaF-ZrF^ con- extent the shape of the original metal. Treatment
taining initially 0.258 or 0.0026 wt % plutonium as with fluorine rapidly converts the " s t a i n l e s s s t e e l "
P u F 3 . The extent of the best removal was 87.6% oxides to the fluoride form, with accompanying
for 75-jLi drops at 640°C. T h e s e drops had a resi- complete physical degradation to a fine powder;
dence time of about 7.3 s e c in the 52-in.-long for ZrO this is not s o . Extensive volatilization
fluorination section of the falling-drop column. of chromium fluoride occurs while fluorinating
The degree of plutonium removal was independent s t a i n l e s s steel residue.
of the initial plutonium concentration. An equation The presence of 10% or more of s t a i n l e s s s t e e l
based on diffusion control is presented for correlat- residue in a fluidized bed of alumina hinders
ing the plutonium data. fluidization during the fluorination period. This is
Another possible method for recovering uranium a l s o true of UO when it is first hydrofluorinated to
from fluoride salt without e x c e s s i v e corrosion con- UF^ or UO^F^ in the HF-O^ decladding treatment
s i s t s in fluorinating s t a t i c beds of particles. In lab- and then fluorinated in a second s t e p . Presumably
oratory t e s t s , about 99% of the uranium was removed the presence of a high fluoride content in the
from s p h e r e s , 60 to 150 /x in diameter, by fluorina- alumina bed promotes sintering and caking.
tion for 2 to 3 hr at temperatures between 450 and Good plutonium-volatilization results were ob-
480°C (solidus temperature, 512°C). However, tained with alumina beds spiked with PuO^, al-
during fluorination, the s a l t sintered into a hard, though the temperature of fluorination is quite
porous m a s s . Thus, fluorination of solid particles important. Recoveries of better than 99.5% were
does not compare favorably with the spray-fluorina- demonstrated in short-time t e s t s at and below
tion scheme and will not be investigated further. 500°C. At higher temperatures (up to 650°C),
The work with uranium is now complete; after diffusion of lower plutonium fluorides (PuF or
a few more experiments, the plutonium work will P u F ) into the alumina leads to greater plutonium
be terminated. A few similar experiments with retention in the bed, at l e a s t in short-cycle testing.
protactinium will be made in the near future. We Further development studies will be of factors
conclude that although plutonium is more difficult affecting plutonium and uranium retention in the
to volatilize from a falling drop than uranium, decladding-fluorination reactor. A related study
falling-drop fluorination is a chemically feasible is also planned of the factors that cause the bed
method for recovering plutonium and uranium from material to sinter or c a k e .
molten fluorides.

2.5 Design of a Fluidized-Bed Pilot Plant

2.4 Fluidized-Bed Volatility Process
for Zirconium- and Stainless-Steel-Clad
Development for Stainless Steels and
U O , Power-Reacfor Fuels
Zircaloy-Clad U O , Fuels

A fluidized-bed fluoride-volatility pilot plant is

Bench-scale laboratory studies have been carried being designed to study the processing of UO
out in support of development of fluidized-bed power-reactor fuels at high levels of radioactivity.
volatility processing from low-enrichment UO These studies will be made to obtain data needed
fuel clad in Zircaloy-2 or s t a i n l e s s s t e e l . T h e s e to design a full-scale commercial plant.
studies have been mainly of chemical and physical F i r s t priority will be given to processing Zir-
effects observed in the use of H F - 0 a s a declad- caloy-clad UO by reaction with HCl, followed by

a two-zone oxidation-fluorination s t e p . The use of concentration of fluorine is required in the UF -

H F - 0 is preferred for decladding and oxidizing P u F gas stream passing through an NaF bed at
s t a i n l e s s - s t e e l - t y p e fuels. If results of earlier 325 to 400°C to prevent formation of an NaF-UF^
developmental studies are favorable, the use of complex.
HF-0 will also be studied for processing zir-
conium-clad fuels. Although the method for s e p -
arating UF , P u F and volatile fission products 2.7 Sorption of U F ^ by N a F
is not definite at present, thermal decomposition
of P u F ^ to solid P u F ^ , followed by distillation of Sodium fluoride is used to separate UF^ from
the UF and volatile fission product fluorides, is other volatile fluorides by a process of s e l e c t i v e
presently favored. A possibility e x i s t s that BrF sorption-desorption. Sorption data for UF • NaF,
vapor will be used instead of fluorine to volatilize WF . N a F , and M o F , - N a F that will be useful in
6 6
only the uranium, thus effecting an early separa-
defining separations-process conditions are re-
tion of the uranium from the plutonium. The plu-
ported. The capacity and durability of NaF pel-
tonium would later be converted to P u F by fluo-
lets made at the Paducah Gaseous Diffusion Plant
(for about $ l / l b ) were compared with the values of
The facility is sized for a charge of about 40 kg these properties determined for the pellets pur-
of UO and 200 g or more of plutonium, depending chased from Harshaw Chemical Company at $ 5 / l b .
on nuclear-safety limitations. The primary re- Performance of the two types was the same.
actor will be 8 in. in inner diameter.
Several design studies are under way, and de- 2.8 Continuous In-Line Monitoring of Gas
sign and layout for major head-end equipment
Streams in Fluoride-Volatility Processing
items are progressing. Engineering of instrumenta-
tion has begun, as has development of a critical-
For proper control of the various phases of the
path schedule.
fluoride-volatility p r o c e s s e s , continuous in-line
Modifications to improve the containment features monitoring devices must be developed for use with
of the Building 3019 penthouse have been de- multicomponent corrosive gas streams. A spectro-
signed, and craft work is about to begin. scopic method for UF determination and a gas
chromatographic method for hydrofluorinator off-
gas monitoring are being developed.
The feasibility of using ultraviolet spectrometry
2.6 Sorption of P u F , by Metal Fluorides to continuously monitor UF concentrations in gas
streams was studied preliminarily in the laboratory.
A possible method for recovering P u F in fluo- The absorption coefficient of UF at 3686 A was
ride-volatility processing might c o n s i s t in the use found to be 5.5 liters m o l e " ' c m ~ ' for the pres-
of solid sorption similar to the manner in which sure range 10 to 100 mm Hg. An approximate value
UF is trapped on N a F . Accordingly, the sorbability of 3610 ± 1000 liters m o l e " ' c m " ' for the absorp-
of P u F on a large number of inorganic fluorides tion coefficient was determined at 2140 A for a
and alumina was examined. As yet, no sorbent previously unreported absorption peak. More work
has been found from which desorption is feasible, is needed before this method can be utilized.
at least up to 500°C. All the alkali and alkaline-
earth fluorides, with the exception of MgF , sorbed
significant amounts of P u F , indicating the exist- 2.9 Vapor-Liquid Equilibria of U F ^ - N b F j System
ence of stoichiometric complexes. Fluorides of
other elements showed that sorption was greater As a part of continuing s t u d i e s on fluoride
at higher temperatures. volatility methods of reprocessing nuclear fuels,
The most highly developed sorbent, NaF, has the separation of UF from volatile impurities by
been used at 325 to 400°C with a high degree of distillation is being considered. For this reason,
s u c c e s s in quantitatively sorbing P u F from a the vapor-liquid equilibria of UF and one such
stream of UF and fluorine, but efforts to desorb impurity, NbF are being determined. The 150°C
the plutonium at temperatures up to 600°C in isothermal equilibrium curve was determined from
dynamic t e s t s have been unsuccessful. A high total-pressure—liquid-compos ition measurements,

and an isobaric (6 atm) equilibrium curve is being was used for all t e s t s b e c a u s e of its exceptionally
determined by direct measurements of vapor and high degree of corrosivity. The results of experi-
liquid compositions. The critical constants of ments with coupons in the 4-in.-diam containers
NbF are being determined. showed that conditions could be obtained for def-
initely reducing corrosion. Argon and helium were
the most effective sparge g a s e s . In the individual
2.10 Phase Equilibria Studies Pertinent to 2-in.-diam containers, however, only slight protec-
Fluoride-Volatility Processing of Fuels tion was obtained.
or Mixtures Containing Chromium Because of the reorientation of the overall pro-
gram toward fluidized-bed p r o c e s s e s and the indif-
Studies of phase equilibria of systems contain- ferent results of the experiments, corrosion studies
ing CrF were conducted in attempts to establish as related to molten-salt p r o c e s s e s were discon-
a suitable mixture of molten fluorides for use as tinued.
a solvent in the processing of s t a i n l e s s - s t e e l - c l a d
or -base fuels. Attention was focused on the solu-
bility of one constituent, chromium, and informa-
tion was obtained specifically on the 3 L i F . C r F -
3KF • CrF and NaF-CrF3 s y s t e m s . The latter was
of s p e c i a l interest because of the use of NaF as 3.1 Conversion of High-Level Radioactive
a sorbent for the volatile chromium fluorides formed Wastes to Solids
in volatility processing, and further work on this
aspect of the process will be continued. No addi- Engineering Studies. — Development of methods
tional work on the development of molten-salt for converting high-level radioactive liquid w a s t e s
solvents for " s t a i n l e s s s t e e l " fluorides is planned to solids using simulated w a s t e s was continued.
because of the emphasis on the use of fluidized- The development of the pot calcination processes
bed reactors. was successfully completed with t e s t s in 16-in.-
diam pots which confirmed the mathematical model
for predicting filling rates a s a function of pot
2.11 Volatilization of U F ^ from Sol-Gel-Derived diameter. Two successful Rising-Level P o t g l a s s
ThOj-UOj t e s t s were made using simulated Purex waste at
processing rates equivalent to the rate of waste
A possible method for recovering uranium from production from a plant processing 2 tons of ura-
thoria-urania sol-gel microspheres would c o n s i s t nium per day. These t e s t s were made with an
in the direct fluorination of the uranium from the induction heating furnace. Based on present oper-
oxide at high temperature. A small-scale examina- ating experience at ORNL with r e s i s t a n c e and in-
tion of this possibility initiated during the past duction heating furnaces, r e s i s t a n c e heating is
year demonstrated that spheres 149 to 210 /x in recommended for use with pot s y s t e m s .
diameter containing 3.9 wt % uranium cannot be An initial test of the continuous process in a
successfully fluorinated at reasonable tempera- 20-in.-diam horizontal continuous melter using a
tures. Treatment for 3.5 hr at 650°C removed only r e s i s t a n c e furnace was successful from an oper-
15% of the uranium. ational viewpoint. However, the melter failed due
to e x c e s s i v e corrosion after 250 hr of operation.
Further development of this process has been
2.12 General Corrosion Studies terminated for the immediate future.
High-Level Radioactive Waste Disposal: Labo-
The relatively high rate of corrosion of both the ratory Studies. — A computer code was developed
hydrofluorinator (dissolver) and the fluorinator is a for retrieving and correlating the mass of data
serious deficiency of the molten-salt fluoride- generated in the program for developing glassy
volatility process. Work was continued on intro- melts for the fixation of high-level w a s t e s . Initial
ducing an inert gas just below the surface of the readout was obtained on ORNL melt compositions.
molten fluoride to help reduce corrosion of the Compositions received from Hanford will also be
hydrofluorinator at the molten-salt—gas interface. included in the code. Melts were prepared which
A melt of 52-37-11 mole % N a F - L i F - Z r F , at 650°C contained up to 32% total waste oxides, including

simulated fission products expected from fuel after about six months. The leach rate for '°®Ru
irradiated as high a s 35,000 Mwd/metric ton. was 7 x 1 0 " * g cm~^ d a y " ' after six months.
T h e s e materials melt at about 800°C and are suf- The products showed sufficient radiation stability
ficiently fluid at 850 or 900°C to permit their use to a dose of 10* rads. Leach rates were not af-
in a continuous melting p r o c e s s . Melts containing fected by this radiation dose.
Purex FTW-65 waste with simulated fission products Design of a pilot plant to incorporate 100 gal
from a fuel burnup of 10,000 Mwd/metric ton were of ORNL waste-evaporator bottoms in emulsified
successfully run in semiengineering-scale con- asphalt per 8-hr shift has been completed and in-
tinuous melters. In t h e s e t e s t s , phosphate vola- stallation started.
tility was generally <0.5%, sulfate volatility
<5.0%, molybdenum volatility 1 to 2%, and iron
and ruthenium volatilities 20 to 30%. Some anom- 3.3 Treatment of Low-Level Radioactive Wastes
alous results were observed.
Corronel 230, 50% Ni-50% Cr, Inconel, and Scavenging-Precipitation Foam-Separation Proc-
Nichrome V show some possibility a s substitutes e s s . — Pilot Plant Studies. — The Scavenging-
for platinum as a material of construction for a Precipitation Foam-Separation process for treating
continuous melter. Corrosion t e s t s are s t i l l in low-level radioactive waste water was studied on a
the early s t a g e s . A thermal conductivity probe pilot plant s c a l e — 300 gal/hr for scavenging-
was developed for the measurement of levels of precipitation and 120 gal/hr for foam separation.
solution, c a l c i n e , and glass in the rising-level The process c o n s i s t s in (1) precipitating calcium,
(RL-Potglass) process. magnesium, and radiation emitters in a sludge-
blanket clarification step that includes the use of
Pilot Plant Design. — Active liaison with Han-
Grundite clay for sorption of cesium and (2) a-
ford personnel on the design and startup of the pot
chieving further decontamination in a countercurrent
calcination phase of the Hanford Waste Solidifica-
foam column. The foam-separation equipment
tion Engineering Prototypes h a s continued. The
includes a 2-ft-square by 8-ft-high countercurrent
equipment is being tested to verify the design
foam column, three 2-ft-square foam stripping
before installation in the Hanford F u e l s Recycle
columns in s e r i e s for recovering surfactant from
Pilot Plant in August of 1965. Design studies
the decontaminated waste, an air-recirculation
conducted recently include pressure transients in
system to provide air for foam generation, a cen-
the pot calciner off-gas line, thermocouple s e l e c -
trifuge foam breaker, and three orifice foam breakers.
tion for the calciner and furnace, the composition
of Nuclear Fuel Services w a s t e s , a review of the It was more difficult to generate a s t a b l e foam in
furnace design, and participation in the scheduling the pilot plant foam column than it was in the
of runs for the hot pilot plant experiments. smaller columns (maximum diameter, 6 in.) used
for laboratory and engineering s t u d i e s . The foam
section was baffled to prevent foam circulation by
filling the countercurrent s e c t i o n with 1.5-in.-diam
3.2 Disposal of Intermediate-Level
Radioactive Waste
by 1.5-in.-long P a l l rings, and the concentration
of surfactant in the liquid pool at the bottom of
the column was kept much higher than it had been
Nuclear installations generate large volumes of
in the smaller columns. Under t h e s e conditions
intermediate-level radioactive w a s t e s , such a s
the foam column alone provided strontium decon-
evaporator concentrates, second- and third-cycle
tamination factors of up to 200, compared with
solvent extraction raffinates, and slurries or solids
factors of l e s s than 20 found previously.
(residues from low-level treatment p r o c e s s e s ) .
A process has been developed for incorporating The strontium decontamination factors were pro-
these w a s t e s in emulsified asphalt. Use of free- portional to the parameter V/LD (where V and L
flowing emulsified asphalt permits mixing of w a s t e s are gas and liquid flow rates, respectively, and D
and asphalt at ambient temperature or above. is an effective bubble diameter), a s in the lab-
Asphalt products containing 20 to 80 wt % solids oratory s t u d i e s . The use of the parameter V/LD
from waste have been prepared from simulated as a b a s i s for process control appears valid for
ORNL waste-evaporator bottoms. Leach rates for a large system even though the foam contains
'^'^Cs and sodium were 5 x 1 0 " ' ' g cm~^ d a y " ' bubbles that vary widely in s i z e .

A test of the entire process was made that lasted installation. Then, after p a s s a g e through an
93 hr at steady s t a t e . The overall decontamination atomic installation, the water would be decontami-
factors were 1050, 8, 4.2, 1.3, 20, 2.2, and 50 for nated, demineralized, and recycled for reuse. F e a -
Sr, C s , Ru, Co, Ce, Sb, and Zr-Nb respectively. sibility studies were made in the laboratory with
The concentrations of t h e s e elements in the treated neutral ORNL LLW. The system included: (1) clar-
product, in terms of percentage of the maximum ification of LLW using zeta-potential-controUed
permissible concentration in water, were 2, 0.6, additions of coagulant (alum) and coagulant aid
0.3, 0.1, 0.2, 0.02, and 0.0001 respectively. Thus, (activated silica), (2) demineralization and decon-
the process is suitable for use at many nuclear tamination of the supernatant by cation-anion ion
installations where low-level radioactive w a s t e s exchange (Dowex 50W-Dowex 1), and (3) treatment
contain comparable amounts of radioactivity. The with a column of activated carbon. Demineralization
scavenging-precipitation equipment was operated capacity (0.1% specific conductance breakthrough)
at a feed rate of 5 gpm and the foam-separation is expected to be as high a s 2500 volumes. Break-
system at 2 gpm. The ratio of the volume flow through of some radionuclides occurred either
rate of the liquid feed to the volume flow rate of simultaneously with bulk ionic breakthrough or
the condensed foam was 3 1 . The capacity of the soon after, as indicated by specific-conductance
foam separation system is apparently limited by measurements. Thus, continuous monitoring of
the low capacities of the air-supply equipment and specific conductance could be a convenient indi-
the centrifugal foam breaker now in use. cator of bed exhaustion. Overall decontamination
factors ( D F ' s ) ranged from 10^ to 10"* for individual
Laboratory Studies. — The two-step process for activities. For ' ° S r and ' " C s , the D F ' s were
decontaminating low-level radioactive waste water comparable to or better than those obtained in the
(LLW) was tested. This process c o n s i s t s in pre- Scavenging-Precipitation Ion-Exchange process;
cipitating and eliminating most of the hardness for ' " ^ R u and *°Co, the D F ' s were 10 to 1000
(calcium and magnesium) and radioactivity in a times higher. Thus, the proposed system appears
suspended-bed sludge column and producing further very promising.
decontamination in a foam column having an 80-
cm-high countercurrent section. Overall decon- Selective Sorption of Phosphates on Activated
tamination factors for Sr, Co, Ru, and Ce were Alumina. — To make the scavenging-precipitation
>3.7 X 1 0 ^ 2.3 to 4, 2 to 5, and 50 to 180 respec- ion exchange process economically attractive,
tively. The cesium decontamination factor was phosphates, which interfere with the precipitation
20 when 60 ppm of Grundite clay (baked at 600°C s t e p , were removed by sorbing them on activated
for 20 min) was added to the LLW during the pre- alumina. For example, results of experiments
cipitation s t e p . The height equivalent to a theo- with tap water spiked with normal and dehydrated
retical s t a g e for strontium removal in the foam phosphates showed that when the water is p a s s e d
column was about 3 cm. The ratio of surface to downflow through Alcoa F-1 activated alumina,
liquid flow rate is the crucial factor governing more than 99% of the normal phosphate can be
satisfactory strontium decontamination in the foam removed from 3000 bed volumes (BV) when the
column. The metal-ion decontamination factor in phosphate concentration is 5 ppm. Dehydrated
the foam-separation step is proportional to the ion forms of phosphate were removed to a greater ex-
distribution coefficient and inversely proportional tent. The volume reduction factor (BV of product
to the volume reduction and liquid throughput. water per BV of regenerant w a s t e ) was 2000, a
In the search for biodegradable surfactants, two very favorable ratio of throughput to regenerant
were found that exhibited very good removal of w a s t e . Of the aluminas t e s t e d , the F-1 grade was
strontium from solutions. F a s t and economical best.
analytical methods for the determination of sur-
factants and radionuclides in LLW were developed The simplicity and low cost of the process may
and tested. make it attractive for removing phosphates from
water streams that have been polluted with phos-
Water-Recycle Process. — A Water-Recycle phate builders used in household detergents. In-
P r o c e s s is being developed for low-level radio- creased levels of phosphates interfere with stand-
active waste (LLW) treatment. It is based on the ard water-treatment procedures and contribute to
demineralization of all water entering a nuclear greater t a s t e and odor problems in potable water.

3.4 Engineering, Economic, and Safety to the operation of the HFIR. Up to a milligram
Evaluations will be recovered in the Curium Recovery Facility
beginning in January 1966, and the remainder will
A detailed estimate of the c o s t s of " p e r p e t u a l " by processed through the TRU a few months later.
tank storage of w a s t e s from processing power re- Design and development work on the target ele-
actor fuels is being undertaken to e s t a b l i s h a ments, to be remotely fabricated in the TRU, is
basis for comparison with c o s t s for alternate waste under the direction of the Metals and Ceramics
management schemes. In this study the concept Division. The major p h a s e s of the project, includ-
of double containment of radioactivity is applied ing design and construction of the building, the
throughout, and in all instances the design and development of the chemical separations p r o c e s s e s ,
operation of the facility emphasize safety over any and the design, fabrication, and installation of the
potential savings in c o s t s . chemical process equipment are the responsi-
Storage of Purex and Thorex solvent extraction bilities of the Chemical Technology Division and
raffinates a s both acid and alkaline solutions is are reported here.
considered, and storage of Zirflex and Sulfex de-
cladding w a s t e s in alkaline form is postulated.
Tanks are enclosed in steel-lined concrete vaults 4.1 Development of Chemical Processes
for secondary containment and are similar to those
at Savannah River. Decay heat is removed by During the past year the recovery and decon-
circulating water through banks of coils submerged tamination of americium and curium were demon-
in the waste. strated at full-scale radioactivity levels in the
The estimated capital c o s t s of tank farms for Curium Recovery Facility (CRF); in addition,
storing acid raffinate waste range from $21 mil- americium and curium were further purified by LiCl
lion to $47 million, and farms for storing alkaline anion exchange and separated by the carbonate
raffinate cost from $27 million to $50 million. In precipitation method. This work, a s well a s lab-
terms of the electricity originally produced from oratory support given to C R F processing, is re-
the fuel, the c o s t s are 0.007 to 0.015 mill/kwhr ported in Chapter 5. Investigations were continued
and 0.009 to 0.016 mill/kwhr for acid- and alkaline- on developing and testing techniques for intra-
waste farms respectively. T h e s e c o s t s are to be actinide separations by ion exchange; additional
used in a computer code that c a l c u l a t e s the total ^''^Pu was converted to dense oxide and incor-
present-worth storage c o s t s based on the optimal porated into HFIR targets; and sol-gel methods for
tank s i z e for a given rate of interest on capital preparing actinide oxides suitable for HFIR targets
and return on investment. were studied.

An adequate solvent extraction process to s e p -

arate transcalifornium elements is not available,
and chromatographic elution from cation resin
4. TRANSURANIUM ELEMENT PROCESSING with a-hydroxyisobutyrate solution has been the
only reliable method for separating californium,
The High Flux Isotope Reactor (HFIR) and the einsteinium, and fermium from each other. Since
Transuranium P r o c e s s i n g Plant (TRU) have been scale-up of this process to 10- to 100-mg quantities
built at ORNL to produce large quantities of the may be complicated because of disturbance of the
heavy actinide elements as part of the USAEC resin bed by radiolytic g a s e s , additional separa-
Heavy Element Production Program. These ma- tion methods would be desirable.
terials will be used in b a s i c research in labora- In an effort to develop a new separation system,
tories throughout the country. The HFIR is sched- the anion exchange behavior of the transcurium
uled to go critical by August and to power by elements with ethylenediaminetetraacetic acid
December 1965, while the installation of equipment (EDTA) and other actinide complexing agents was
in the TRU should be completed and startup pre- investigated. It was demonstrated that EDTA
parations under way by the same d a t e . Special anionic complexes of the heavy actinides, loaded
irradiations of the available ^''^Pu, being done in on anion exchange resin, can be used in a separa-
a Savannah River production reactor this year, tion scheme. Einsteinium-californium and fermium-
will produce a few milligrams of californium prior einsteinium separation factors of 1.41 and 1.90

were achieved; in a laboratory demonstration of used to prepare uniform microspheres that densify
the process, an einsteinium product fraction was when fired.
obtained which contained 99% of the einsteinium For similar reasons, a sol-gel method for prepar-
and no detectable californium or fermium. ing americium and curium oxide for incorporation
The dissociation constants of lanthanide com- into HFIR targets would be convenient, and lab-
plexes with 1,2-diaminocyclohexanetetraacetic acid oratory efforts to prepare lanthanide sols a s stand-
(DCTA) corresponded to cation exchange separa- ins for americium and curium have been initiated.
tion factors of 2.0 to 2.5 between adjacent heavy Stable lanthanide sols capable of producing dense
lanthanides and indicated that DCTA might be microspheres have been made, and the methods
useful in obtaining transcurium element separa- will be evaluated for the production of americium
tions by cation exchange. The dissociation con- and curium s o l s .
s t a n t s of the DCTA complexes with Am, Cm, Bk,
Cf, E s , and Fm were therefore determined. How-
ever, it was found that unlike the heavy lanthanide 4.2 Development of Process Equipment
elements, the Cf, E s , and Fm dissociation con-
s t a n t s are not appreciably different in this system; Engineering studies of the process equipment
cation exchange separation factors of only 1.23 are continuing in order to test equipment, make
for Es/Cf and 1.35 for F m / E s could be obtained. necessary modifications, and develop operating
High Flux Isotope Reactor targets will be made procedures. Remote mechanical handling t e s t s of
from pressed pellets of aluminum powder and equipment racks and components in a full-scale
actinide oxide, and it is e s s e n t i a l that the alumi- mockup of the processing cell are nearly complete.
num phase be continuous in order to ensure s a t i s - Flowsheet t e s t s of the Tramex solvent extraction
factory heat transfer during irradiation. When very process for separating a c t i n i d e s from lanthanides
fine actinide oxide particles are mixed with —325 were conducted in glass pulsed columns. Extrac-
mesh aluminum powder and pressed into cermets, tion of americium was consistently good; the l o s s
the oxide phase is continuous and the thermal was l e s s than 0.1%, and the stage height was 6 to
conductivity of the pellet is low. But, with oxide 10 in. Scrubbing of europium was efficient, con-
particles ranging from 20 to 200 /z in diameter, the firming previous t e s t s (decontamination factor
aluminum phase will be continuous, and conduc- greater than 2000). In the stripping column an un-
tivity will be satisfactory for irradiation in high expectedly high loss of americium caused by
neutron fluxes. entrainment in the w a s t e organic was noted. This
A process employing hydroxide precipitation for loss was reduced to acceptable values by increas-
preparing dense, coarse particles of PuO has been ing the concentration of hydrochloric acid in the
reported previously; during this past year the strip reagent from 1 to 4 A/.
process was used to prepare 180 g of ^''^Pu as The first phase of the hydraulic testing of the
PuO for incorporation into 18 HFIR targets. full-scale metal columns was completed. After
These targets are being irradiated at the Savannah temporary modification of lines and flow elements,
River Plant. design flow rates were attained, and good control
of feed rates and interfaces was possible.
As part of the preparation of PuO by hydroxide
precipitation, the dried oxide must be ground and
screened, about 25% of the resulting particles are
l e s s than 20 ji in greatest diameter and must be 4.3 Design and Fabrication of Process Equipment
recycled. Preparation of PuO by a sol-gel method
would eliminate the dissolution and valence ad- The TRU facility will be equipped with two sol-
justment necessary for recycle, and the prepara- vent extraction c y c l e s ; a head-end system, includ-
tion of uniform microspheres from a plutonium sol ing a dissolver; feed-adjustment tank and centri-
would eliminate grinding and screening. For t h e s e fuge for feed clarification; ion exchange columns
reasons, efforts to prepare plutonium sols have for cleanup and s p e c i a l separations; and miscel-
continued, and sphere-forming techniques have laneous equipment.
been investigated. This work is still in progress; Design of the equipment is nearing completion,
however, stable plutonium s o l s have been prepared and shop fabrication of the components and con-
in the laboratory, and plutonium s o l s have been necting jumper piping is nearly 80% complete. The

only equipment not yet designed is that for separat- development during this report period was confined
ing the heavy actinides (californium, einsteinium, to problems a s s o c i a t e d with full-scale processing
fermium) from each other and the equipment for in the Curium Recovery Facility at high levels of
preparing curium oxide for the recycle rods. This radioactivity.
design of the equipment, just started, has been As to process improvements, the entrainment of
held up pending completion of flowsheet develop- product during stripping operations in the Tramex
ment. and Clanex processes was markedly reduced by in-
c r e a s i n g the mixer speed and the ratio of the strip
and solvent volumes.
4.4 Construction of the Transuranium Other laboratory studies indicated that the solu-
Processing Plant
bility of the rare earths e x c e e d s 44 g/liter in a
feed solution 12 M in LiCl, 0.2 M in HCl, and 0.1
The Transuranium P r o c e s s i n g Plant was com-
M in SnCl . That concentration of rare earths is
pleted in late May, twenty-three months after start
not expected to be reached in any of the feed
of construction and six weeks beyond the sched-
solutions. Also, the radiolysis of feed solutions
uled completion d a t e . No major problems arose.
and the consequent oxidation of cerium to the
Installation of the process equipment is just under
tetravalent (extractabie) s t a t e were studied. It was
way and should be completed by late fall.
found that methanol does not consistently prevent
radiolysis nor the oxidation of trivalent cerium,
both of which must be avoided as much as p o s s i b l e .
5. CURIUM PROCESSING Other studies showed that ^'*''Cm-^''^Am solutions
received from the Savannah River Plant can be
A joint program between the Isotopes Division processed, even when the sodium nitrate content
and the Chemical Technology Division was e s t a b - is a s high a s 0.5 M.
lished to produce curium heat sources for use in
thermoelectric converters. The Chemical Tech-
nology Division is responsible for providing proc-
e s s technology, facilities, and operations to i s o - 5.2 Curium Recovery Facility:
late gram amounts of curium. Nearly 17 g of ^'''*Cm Equipment and Flowsheet Testing
was purified, separated from ^''^Am, and transfer-
red to the Isotopes Division for incorporation into Further changes were made in the process equip-
an experimental heat source and for distribution ment to incorporate piping and v e s s e l s made of
to other AEC installations. Six irradiated targets more suitable structural materials. Cold testing
of ^'*'AmO in an aluminum matrix were also was completed, and hot operation began in Novem-
processed to separate ^"^^Cm from the aluminum ber 1964. During five months of hot operation with
and fission products, but not from residual ameri- radioactive feed material, the equipment performed
cium. To d a t e , 3.56 g of purified ^''^Cm has been quite well, with the exception of several failures
delivered to the Isotopes Division, and another of the mixer-settler drive train and several failures
14 g will be processed early next fiscal year. of the feed transfer pump.
Changes in the flowsheets were made to over-
come problems arising from Zircaloy-2 corrosion
5.1 Process Development products, interfacial s o l i d s , e x c e s s i v e solvent
extraction l o s s e s , cerium oxidation and extraction,
The chemical p r o c e s s e s for isolating curium and the radiolytic effects of high power d e n s i t i e s .
include dissolving the target in dilute hydro- Final forms of the flowsheets were satisfactory
chloric acid; using the Clanex process to convert for processing ^''''Cm and were s u i t a b l e for proc-
certain feed solutions from the nitrate form to the e s s i n g ^"^^Cm up to a limit of 6 w / l i t e r (0.05 g/liter)
chloride and/or to remove aluminum; using the in the chloride (Tramex) system. In the nitrate
Tramex process to separate americium-curium from system (Clanex), power density was not limiting
fission products; and precipitating Am^ (as a up to 16 w/liter. It is hoped that higher power
double carbonate) to s e p a r a t e it from curium. Most d e n s i t i e s can be achieved in the chloride system
of the laboratory development and t e s t i n g of these with more development, thus permitting a greater
p r o c e s s e s was completed l a s t year, and process capacity for ^''^Cm. Contrary to e x p e c t a t i o n s .

adding methanol to the Tramex feed failed to in- the curium and residual 2'"Am. During processing,
hibit radiolysis of the free acid and did not prevent 1.31 g of ^''^Cm decayed. Of the remaining 5.44 g,
gassing. This is believed to be due to the presence 3.56 g of high purity was sent to the Isotopes
of impurities such a s iron, nickel, and copper in Division and 1.13 g, mostly recycle material, was
the actual process solutions but not present in the saved for process-development s t u d i e s . Radioly-
laboratory experiments. tic destruction of the SnCl reductant was much
too great at power levels above 6 w/liter. For ex-
ample, at higher radiation l e v e l s the solutions
5.3 Solvent Extraction Processing of could be processed without mechanical difficulties,
^^^Am and 244c„ but cerium could not be separated from the a c t i -
In the initial processing campaign of the Curium
Recovery Facility, 24 g of ^'*''Cm contaminated
with about 100 times that amount of rare earths 5.6 Development of Alternative Processes:
was processed; 19 g of product was recovered.
Separation of Lanthanides and Actinides
About 19 g of ^''^Am was a l s o recovered. Decon-
tamination factors of over 1000 for the rare earths
The Talspeak process s e p a r a t e s trivalent a c t i -
were achieved in one cycle when SnCl was used
nides from lanthanides and many other elements
as a reductant for Ce(IV) formed by radiolysis;
by two cycles of extraction with di-(2-ethylhexyl)-
methanol was ineffective in inhibiting acid de-
phosphoric acid (HDEHP) from l a c t i c acid solu-
struction and gassing in the feed solutions. The
tions containing sodium diethylenetriaminepenta-
americum-curium mixture was transferred to cell 3
acetate (Na DTPA). Neodymium is the l e a s t ex-
of the Curium Recovery F a c i l i t y , where it was
tractable lanthanide and californium apparently
further purified and the ^'*'*Cm was separated from
the most extractabie trivalent actinide, slightly
t h e 2 43Am.
more s o than einsteinium and fermium. Laboratory
t e s t s of a flowsheet designed for processing feed
5.4 Ion Exchange Purification and Separation solutions containing high concentrations of lan-
of 243Am.244c„ thanum were successful in separating americium
from nearly all contaminants. Preliminary t e s t s
During initial solvent extraction processing, indicate excellent removal of contaminants in feed
americium-curium product was obtained which re- preparation by extraction with tributyl phosphate
quired additional purification. Glass ion exchange followed by stripping with lactic acid.
equipment was installed in cell 3 of the Curium
Recovery Facility, and additional decontamination
was obtained by LiCl anion exchange. In addition,
^''''Cm was separated from ^'^^Am by precipitating
americium as K AmO (CO ) .
A total of 16.6 g of ^''''Cm was purified by ion FUEL CYCLE
exchange, separated from ^''^Am, and transferred
to the Source Fabrication Facility of the Isotopes The Thorium Fuel Cycle work is part of the
Division for incorporation into an experimental Thorium Utilization Program of ORNL. The over-
heat source and for distribution to other AEC lab- all objective is the development of the fuel-cycle
oratories. The ^''^Am was converted to the oxide, technology required for economical power produc-
most of which was also distributed to other AEC tion in thorium-fueled reactors. Much of the work
laboratories for use in research programs. An ad- carried out in this program by the Chemical Tech-
ditional 1.9 g each of ^''''Cm and ^''^Am were nology Division has been in c l o s e cooperation
purified and will be used for process development. with the Metals and Ceramics Division. Oxide
fuels and the preparation of microspheres of those
5.5 Processing of ^ ^ ' A m O j Targets fuels are being emphasized, but some work on prep-
aration of thorium-uranium carbides and other
Six targets of irradiated AmO containing a total systems is being carried out to exploit the ver-
of 6.75 g of ^''^Cm were processed for recovery of satility of the sol-gel process.

6.1 ^^•'U Storage and Distribution Facility ward increasing the capacity to that level. In this
laboratory, equipment and methods will be devel-
During the past year, design of a ^ 3 3 ^ gtojage oped for the production of coated microspheres of
and purification facility was completed and instal- oxides and carbides by completely remote opera-
lation was begun. Storage wells for solids such tions. T h e s e p r o c e s s e s will be conducted in the
a s " ^ U 3 0 g in amounts up to about 100 kg of " ^ U Thorium-Uranium Recycle Facility.
have been installed, and tanks and other equipment
for liquid storage are being made and installed.
With the purification facility (1 kg of ^^^U per day) 6.4 The Thorium-Uranium Recycle Facility
being fabricated and the e x i s t i n g solvent extraction
system used for Thorex process demonstration, a A major effort has been directed toward design-
versatile intermediate-scale facility for producing ing and getting construction started on the Thorium-
•^^^U highly purified from all but isotopic impuri- Uranium Recycle Facility (TURF). Construction
t i e s and attendant decay products is now available. is under way, and completion of the entire facility
is scheduled for March 3 1 , 1967. Completion of
the construction contract is scheduled for Decem-
6.2 Development of the Sol-Gel Process ber 1966. Equipment design, construction, and
testing is proceeding concurrently with the facility
In the laboratory, sol-gel process studies were construction. Equipment initially installed will
carried out on methods to control the porosity of provide a capability for producing up to about 50
thoria and thoria-urania materials and to prepare kg/day of metal-clad thorium-uranium oxide fuels.
nitrides of thorium. Experiments were also carried Enough s p a c e has been provided to permit instal-
out to prepare zirconia by sol-gel methods. lation of a versatile, close-coupled head-end
processing complex and to accommodate a com-
plete line of equipment for making coated micro-
s p h e r e s . Equipment fabrication and operation of
6.3 Development of Methods for
this facility are shared with the Metals and Ceramics
Producing Microspheres

Continuous preparation of microspheres of thoria

from a 3 A/ thoria sol was demonstrated on an en- 6.5 Development of Equipment for T U R F
gineering s c a l e in a tapered-column system employ-
ing a two-fluid nozzle at a thoria feed rate of 230 g In the Chemical Technology Division, equipment
of ThO per hour. Over 72% of the microspheres development and testing, beyond that being carried
had diameters of 230 ± 20 fi, and over 85% were in out in connection with the Coated P a r t i c l e Devel-
the range 200 ± 50. This represents excellent s i z e opment Laboratory, h a s centered about a vertical
control. In addition to the results obtained with tube calciner for oxide to be vibratorily packed
the two-fluid nozzle, encouraging results were into metal fuel tubes. The calciner meets capac-
obtained with a rotating disperser. This kind has ity, product quality, and other requirements im-
the advantage over the two-fluid nozzle in that it posed on it, but there has been a tendency for the
permits a significantly higher production rate. oxide to bridge near the top of the furnace. Present
As part of the development of a process for mak- efforts are directed toward solving this problem.
ing microspheres, equipment for a coated-particle Equipment to simulate the TURF c e l l s has been
development laboratory was designed and installed. installed, and t e s t s in this c e l l mockup have
This laboratory was built and will be operated as started.
a cooperative undertaking with the Metals and
Ceramics Division. It includes equipment for
forming, drying, calcining, and carbon-coating 7. SEPARATIONS C H E M I S T R Y RESEARCH
oxide and, eventually, carbide microspheres.
Equipment is sized to produce about 5 kg/day of New separations methods and reagents are be-
microspheres. At present the coater does not have ing developed, principally for u s e s in radiochemi-
that capacity, but development work under way in cal p r o c e s s i n g but a l s o for other purposes extend-
the Metals and Ceramics Division is directed to- ing from extractive metallurgy to biochemical

separations. The current emphasis is on solvent group B metal ions were 0.5 to 1 for mercury,
extraction technology, although other separation 0.002 to 0.005 for selenium, and l e s s than 0.001
methods are receiving a growing share of attention. for antimony, with little dependence on nitric
Reagents that were developed in the former raw- acid concentration. Continued study of nitrosyl-
materials program at ORNL have continued to ruthenium extraction by TLA suggested identifica-
show extended utility. Additional reagents have tion of the extractabie s p e c i e s at high aqueous
a l s o been discovered in more recent evaluation nitric acid concentration a s HRuN0(N03)^-H^O
studies. The present program in separations and RuNO(N03)-2H20. On dilution from 12 to
chemistry can be divided into three interdependent 3 M HNO , they hydrolyze to nonextractable s p e c i e s
types of research activity: (1) Descriptive Chemi- with first-order rate constants of 0.10 to 0.15
cal Studies (Sect. 7.1 to 7.7) of the reactions m i n " ' and 0.01 m i n " ' . They are extracted to form,
of s u b s t a n c e s to be separated and of separa- probably,
tions reagents, of the controlling variables in
particular separations, and of new compounds (R3NH)2RuNO(N03)
that may be potential reagents. (2) Develop-
ment (Sects. 7.8 to 7.12) of selected separations
into specific complete p r o c e s s e s , both where no R3NHRuNO(N03)^ -H^O .
workable process yet e x i s t s and where existing
p r o c e s s e s are l e s s than satisfactory, carried where The latter is then converted to the former with a
warranted to the point that large-scale performance rate constant of 0.075 m i n " ' .
can be predicted. (3) Fundamental Chemical
Studies (Sects. 7.13 to 7.18) of the equilibria and
reaction mechanisms involved in separation s y s - 7.3 Reagents for Separating Polonium(IV)
tems, both to increase knowledge and to help de-
from Bismuth
fine potential applications.
Several types of amines and neutral organo-
phosphorus compounds showed a potential for
7.1 Extraction of Metal Chlorides by Amines separating 2'°Po(IV) from bismuth by solvent
extraction from hydrochloric acid solutions. For
As part of a program on surveying the extraction example, from 8 M HCl, nearly all of the reagents
characteristics of many metals from various s y s - studied gave efficient polonium extractions; and
tems with representative amines data were ob- in some c a s e s , the polonium-bismuth separation
tained for the extraction of 31 different metal ions factor was greater than 1000. The data suggest a
from HCl and L i C l - 0 . 2 M HCl solutions over the number of potentially practicable process flow-
range 0.5 to 10 A/ total chloride. Combined with s h e e t s . The development of p r o c e s s e s would be
previous work, this makes a total of 60 metal ions greatly facilitated if a nonradioactive stand-in
that have been studied in t h e s e s y s t e m s . With could be found for polonium. Preliminary t e s t s
few exceptions, the extraction power of the amine indicate that tellurium will be of limited useful-
for the metal varied in the order: Aliquat 336 n e s s for this purpose.
(quaternary amine) >Alamine 336 (tertiary amine)
>Amberlite LA-1 (secondary amine) > Primene JM
(primary amine).
7.4 New Extraction Reagents

7.2 Extraction of Metal Nitrates by Amines Fourteen new substituted phenols, many of
which are structurally similar to the efficient
In survey extractions of metal nitrates by tri- cesium extractant 4-sec-butyl-2-(a-methylbenzyl)-
laurylamine nitrate (TLA), extraction coefficients phenol (BAMBP), were tested for their ability to
of the group IB metal ions by 0.2 M TLA in toluene extract cesium from dilute c a u s t i c solutions. Al-
decreased in this order: gold » silver » copper, though some of the phenols under certain condi-
being about 50, 0.01, and 0.0005 from 3 M HNO3 tions extract cesium more strongly than BAMBP,
and 10, 0.2, and 0.0001 from 12 M HNO3. Corre- none appears superior to BAMBP a s a process
sponding extraction coefficients for some other extractant.

7.5 Performance of Degraded Reagents experiment, the position of 16 specific t-RNA's

and Diluents in the eluate fractions was e s t a b l i s h e d by radio-
chemical amino acid biochemical a s s a y .
Studies were continued on the s t a b i l i t i e s of A sample of mixed t-RNA's was chromatogram-
alkylbenzenes, especially diethylbenzenes, and med, and the phenylalanine-accepting t-RNA frac-
of the reagent di-(sec-butyl) phenylphosphonate. tions were pooled and rechromatogrammed. The
The principal products of the reactions between bulk of the material eluted in a symmetrical ultra-
diethylbenzene and dilute nitric acid are ethyl- violet-absorbing peak which coincided with the
acetophenones and a-nitroethyl ethylbenzenes. phenylalanine radioactivity peak, indicating clean
Minor degradation products include ethylbenzoic separation. A purity of at l e a s t 74% was calcu-
acid. Very little nitration of the benzene ring was lated after making a reasonable assumption for
detected. The combined products are thermally the molar extinction coefficient of phenylalanyl-
stable to 160°C as determined by differential t-RNA.
thermal analysis (DTA). A new variation of the Experiments are under way to prepare larger
DTA method was used to determine an approximate quantities of the phenylalanine-accepting t-RNA
value for the heat of degradation of diethylbenzene for further biochemical and physiochemical in-
(50 to 70 kcal/m.ole). Degradation of a T B P - vestigations.
diethylbenzene solution did not lower its flash
point. T e s t s proved that degraded impurities, and
not the major degradation products, account for
much of the fission product extraction power of 7.7 Extraction Properties of Lanthanide
some degraded diethylbenzenes. Commercial and Actinide Complexes
alkylbenzenes made from a-olefins were not as
stable as n-nonylbenzene and diethylbenzene. Di(2-ethylhexyl)phosphoric acid (HDEHP) ex-
Undiluted di-(sec-butyl) phenylphosphonate was ir- tracts lanthanum from lactic acid solutions as a
radiated while being stirred with nitric acid, and lactate complex with varying ratios of lactate to
the fission product extraction power of the irra- lanthanum, such that the maximum loading capacity
diated material was l e s s than that of tributyl phos- of HDEHP is about 1.5 equivalents of lanthanum
phate in a comparable t e s t . Nitration of both re- per mole of HDEHP.
agents occurred, but the products have not yet
been identified. L e s s than 1 part of phenol per
250 parts of the phosphonate was detected, show-
ing little tendency for phenol formation at the 7.8 Recovery of Rubidium from Ores
potentially active s i t e ortho to the phosphorus
bond. The phenol extraction (Phenex) p r o c e s s , origi-
nally developed for recovering cesium from reactor
w a s t e s and from ores, was applied to recovery of
rubidium from alkaline-ore leach liquors. Rubidium
7.6 Biochemical Separations is extracted with 4-sec-butyl-2-(a-methylbenzyl)-
phenol (BAMBP) in hydrocarbon diluent and strip-
A new reverse-phase column (extraction) chroma- ped with dilute acid to give a pure rubidium s a l t .
tography system has been devised which affords
significant fractionation of the different transfer
ribonucleic acids (t-RNA's) from E. coli. The
column packing c o n s i s t s of a quaternary ammonium 7.9 Recovery of Beryllium from Ores
liquid extractant, dimethyldilaurylammonium chlo-
ride in isoamyl a c e t a t e , supported on hydrophobic Further studies of a primary-amine extraction
diatomaceous earth. The mixed t-RNA's were process for recovering beryllium from sulfuric
dissolved in the mobile aqueous phase and were acid digests of low-grade beryllium ores continued
sequentially eluted from the column with increas- to give favorable r e s u l t s . The development of a
ing concentration of NaCl in the e l u a t e . Magne- more economical stripping method, which u s e s
sium ions had a pronounced effect on the chroma- dilute ammonium fluoride solutions, has been of
tographic behavior of specific t-RNA's. In one particular interest.

7.10 Recovery of Acid by Amine Extraction

U^* + -(aHA-bNaA) ^ M A .4HA + (2-66)H* + 66Na *
y ^
A tentative process for recovering and purifying
phosphoric acid by tertiary amine extraction was
tested with a samn'e of commercial, wet-process for calcium, strontium, and barium at S A = 0.125 M
phosphoric acid. Phase-separation problems, and for magnesium at S A > 0.25 M; and
o b s t a c l e s to commercial applications, were en-
countered. M^"^ + -(aHA.6NaA) :^MA2-2HA + (2-46)H* + 46Na* ,

7.11 Recovery of Thorium from Granitic Rocks with probably some water also extracted, at lower
HA concentrations for magnesium and calcium,
In a bench-scale demonstration of an improved for strontium and barium from dilute aqueous solu-
flowsheet for treating granite, about 60% of the tion (0.5 M NaNO ), and for beryllium under all
thorium and 65% of the uranium were recovered conditions tested.
from a composite of three drill cores from the
Conway formation. The thorium and uranium were
dissolved by an acid-cure—percolation-wash treat- 7.14 Intermolecular Bonding in Organic Mixtures
ment of the crushed rock and were recovered from
solution by amine extraction. Processing costs Qualitative and quantitative information about
for treating granite of this particular composition intermolecular association in the organic solutions
are estimated at about $35 per pound of thorium- of several solvent extraction s y s t e m s was obtained
plus-uranium recovered. through infrared spectra, nuclear magnetic reso-
nance, diluent vapor-pressure depression, and
dielectric constant. Association was found be-
tween tributyl phosphate (TBP) and several di(2-
7,12 Separation of Biological Macromolecules
ethylhexyl)phosphates (SrA^, SrA2-4HA, NaA),
and between 4-sec-butyl-2-(a-methylbenzyl)phenol
A versatile cold room for preparing large quan-
and di(2-ethylhexyl)phosphoric acid, octanol-1,
tities of biologically important macromolecules
and acetophenone. Dielectric constants of T B P -
was designed and is being built. The facilities
water mixtures, measured at 2°C to obtain water
consist of a room controlled at a temperature of
contents significantly above the 1:1 mole ratio,
about 4°C, refrigerated centrifuges, and a variety
showed a small but definite break at the 1:1 ratio,
of auxiliary equipment. Initial activities will
indicating that the controversial association
include the preparation of purified transfer ribo-
Species TBP-H O does actually exist, although
nucleic acids (from E. coli), for which there is a
probably very weakly bound.
great need in fundamental biochemical research
7.15 Synergism in the Extraction of Strontium
by Di(2-ethylhexyl)phosphoric Acid
7.13 Extraction of Alkaline Earths by
Di(2-ethylhexyl)phosphoric Acid Since previous work showed that NaA.3HA and
SrA •4HA are principal organic-phase s p e c i e s in
Extraction equilibria of Be, Mg, Ca, and Ba from the extraction of strontium from sodium nitrate
sodium nitrate solutions by benzene solutions of solutions by solutions of di(2-ethylhexyl)phosphoric
di(2-ethylhexyl)phosphoric acid (HA) were de- acid (HA) alone, an obvious hypothesis to explain
termined a s functions of pH and HA concentration, Synergism from added tributyl phosphate (TBP) is
in comparison with previously reported equilibria that T B P replaces some or all of the molecular
of strontium extraction. The order of extractability HA to form a more favorable strontium complex.
i s : Be » Ca > Mg = Sr > Ba. The suggested ex- This hypothesis was supported by (1) shift in the
traction reactions (dotted underlines for organic- extraction maximum toward higher pH, and its
phase s p e c i e s ) are relative decrease at high T B P concentration;

(2) shift of the power dependence of extraction 7.17 Kinetics of Extraction of Iron by
coefficient on pH from the s e r i e s of values ex- Di(2-ethylhexyl)phosphoric Acid
pected for SrA • 4HA to values expected for
S r A 2 . 2 H A - x T B P and for S r A ^ . y T B P ; and (3) Iron(III) is one of several metal ions that are
change of the power dependence of extraction coef- very slow to reach equilibrium in extraction by
ficient on HA concentration from 3 (for SrA •4HA) di(2-ethylhexyl)phosphoric acid (HA). The ex-
to 2 (consistent with SrA2-2HA-xTBP). If the c o - traction rate is first order with respect to aqueous
ordination number of the strontium ion does not iron concentration in acid perchlorate solutions.
change, so that T B P replaces HA 1:1, then x The rate constant k is directly proportional to the
(above) should be 2 for the formation of SrA •2HA. interfacial area but nearly independent of stirring
2 T B P in the synergistic extraction at low pH. rate. This and the relatively high activation energy
Estimates of the power dependence of extraction (=10 kcal/mole) indicate that the rate is controlled
coefficient on T B P concentration do not contradict by a chemical reaction at the interface. The rate-
this hypothesis, but they show nonlinear effects, controlling step (or the pathway) must change at
indicating that some other variable not yet ac- an extractant concentration of about 0.4 M HA in
counted for a l s o varies with the T B P concentra- n-octane, s i n c e several parameters change rather
tion. abruptly at about that concentration: at [HA] = 0.2
Af, jt « [HA]°-^ and [H"*]"', activation energy = 15
kcal/mole; at [HA] ^ 0.5 A/, A: « [HA]'-^ and
[H "*]'•*, activation energy = 10 k c a l / m o l e . The
7.16 Synergism and Diluent Effects in the rate constant d e c r e a s e s with increasing aqueous
Extraction of Cesium by 4-sec-Butyl-
perchlorate concentration, c o n s i s t e n t with an a s -
sociation quotient

Continued study of synergistic cesium extraction

by 4-sec-butyl-2-(a-methylbenzyl)phenol (BAMBP) Q = [FeClO^ 2 +]/[pe3 + ] [ c i o * " ] = 1 .
mixed with organic a c i d s , by means of equilibrium
distribution, infrared absorption, and nuclear
magnetic resonance, indicates that the synergism 7.18 Activity Coefficients of the Solvent Phases
is due to the availability of the proton from the
much stronger acid for the cation exchange, while Further evaluations of departures from ideality,
BAMBP molecules continue to provide most or all expressed either a s activity coefficients (y) or
of the solvation that contributes to organic-phase as apparent average aggregation numbers (n) of
solubility of the cesium. With a dialkylphosphoric real or hypothetical polymeric s p e c i e s assumed to
acid and two carboxylic a c i d s (HX), the principal be ideal, were obtained by both direct measurement
extract s p e c i e s formed at low pH are probably of diluent vapor-pressure lowering and i s o p i e s t i c
CsHX^ • 2BAMBP, CsHX^ • 2(BAMBP)2, and balancing. As previously found for tri-n-octylamine
[Cs.2(BAMBP)2]'^HX2". Comparison of diluents (TOA) and its bisulfate (TOAHS), ff for its normal
of different hydrogen-bonding capabilities showed sulfate (TOAS) was nearly independent of organic-
that in chloroform (acceptor-type, not self-hydro- phase water content up to 90% of saturation, in-
gen-bonded), extraction was somewhat lower than c r e a s i n g only from 0.85 at a = 0 to 1.10 at a =
in carbon tetrachloride (no hydrogen bonding). 0.902. The activity coefficient of TOAS in dry
In octanol-1 (donor-type, self-hydrogen-bonded) benzene decreased nearly linearly from 1 at infi-
and in acetophenone (donor-type, not self-hydrogen- nite dilution to 0.769 at 0.355 m; TOAS thus de-
bonded), extraction at BAMBP concentrations parts l e s s far from ideality than does the free
above 1 M were considerably lower because of base TOA (y = 0.628 at 0.354 m).
stronger competition from BAMBP-diluent a s s o c i a - To resolve a discrepancy between published
tion; at lower BAMBP concentrations the depend- v a l u e s , the activity of tributyl phosphate (TBP)
ence on BAMBP concentration changed from third when saturated with water at 25°C was determined
to first power, showing that t h e s e donor-type by means of an isopiestic s e r i e s . At saturation
diluent molecules can replace BAMBP molecules (a^ = 1), mole fraction ''^^J-BP ^ 0.489 and activity
in solvating the cesium BAMBPate. a = 0.515, with pure T B P as the standard

s t a t e . At lower water contents (X = 0.36), a - low (0.2 to 0.4 M) nitric acid concentrations.
X ^ „ „ ± 0 . 0 0 3 , and a = (2.073 ±"O.01)A: . Partial Separation factors with 2-ethylhexylphenylphos-
TBP W ^ ^ w phonic acid were a l s o about 2 . 5 , but precipitation
molal free energies of solution and e x c e s s partial of the extractant—rare-earths complex occurred in
molal free energies were calculated for both com- some t e s t s . Separations with a primary amine in
ponents. a sulfate system were much l e s s effective.
In continuing improvement of the equipment and
techniques for precisely measuring small vapor-
pressure differences, the possibility of using an 8.3 Ruthenium
inverted cup floating in mercury was examined by
derivation of its equation of motion and by trial A tentative process was outlined for recovering
of a preliminary model. A model for detailed t e s t - ruthenium from acid Purex waste by extracting
ing has been designed that should have about 40 with a tertiary amine and stripping with dilute
times the sensitivity of a simple mercury manometer. c a u s t i c solution.


BY SOLVENT EXTRACTION 8.4 Treatment of Waste That Contains
Dissolved Stainless Steel or Large
8.1 Cesium Amounts of Aluminum Ion
The fission product recovery p r o c e s s e s originally
In batch extraction t e s t s , contaminant metals
developed for treating Hanford Purex w a s t e s are
such a s Ca, Ba, Mo, and Tc were not extracted
being studied for application to w a s t e s of different
from alkaline solutions by 4-sec-butyl-2-(a-methyl-
compositions. In batch t e s t s , strontium, rare
benzyl)phenol (BAMBP), further confirming the
earths, and cesium were recovered effectively from
high selectivity of this extractant for cesium.
simulated " s t a i n l e s s s t e e l " w a s t e s complexed
BAMBP was degraded extensively upon contact
with tartrate and diethylenetriaminepentaacetate.
with acid nitrite solutions, but the reaction was
Complexing with relatively large amounts of tar-
avoided by adding sulfamic acid to the system to
trate also allowed treatment of w a s t e s that contain
destroy the nitrite. BABMP showed no reaction
large amounts of aluminum nitrate (wastes from the
with alkaline nitrite solutions.
TBP-25 process).

8.2 Isolation of Cerium and Promethium

8.5 Engineering Studies

Cerium oxidation with silver-catalyzed persulfate The basic engineering data for designing mixer-
occurred at maximum rate with about 2 M HNO settlers for recovering cesium from Purex-type
in the aqueous phase. More than 98% of the cerium waste by solvent extraction with a substituted phe-
was oxidized and subsequently extracted with nol, 4-sec-butyl-2-(a-methylbenzyl)phenol (BAMBP),
di(2-ethylhexyl)phosphoric acid in 20 to 30 min were determined in a single-stage prototype. The
at 60°C, using 1.8 moles of persulfate per mole of flow capacity of the settler was nearly independent
cerium. of mixing and at the normal flow ratios was 290
The separation of promethium from other rare gal h r - i ft~^ for extraction and 730 gal hr~^ ft"^
earths was studied with a number of different for stripping. Stage efficiencies greater than 90%
types of extractants. Batch studies indicated that, were e a s i l y attained for both extraction and strip-
in the previously developed flowsheet for separat- ping at a power input of l e s s than 10 hp per 1000
ing rare earths by extraction with tributyl phos- gal and a residence time of l e s s than 2 min.
phate (TBP) from concentrated nitric acid, flows
of scrub acid and solvent can be greatly reduced
to make the process more attractive for large-scale 9. CHEMISTRY OF PROTACTINIUM
use. In addition, promising results were obtained
with di(2-ethylhexyl)phosphoric acid. Separation Solvent extraction studies of protactinium in
factors between adjacent rare earths of about 2.5 sulfuric acid solutions were continued. Evidence
were obtained with this extractant at relatively has been obtained that in 2.5 M H SO relatively

extractable polymeric s p e c i e s occur for protactin- equipped with a digitized punched-card output. The
ium concentrations between a few thousandths and effects of nitrate complexation and hydrolysis on
a few hundredths of an mg per ml, but at higher the spectra were also investigated. Computer
concentrations l e s s extractable s p e c i e s appear. codes for the IBM-7090 and CDC 1604-A computers
T h e s e l e s s extractable s p e c i e s are probably larger were written for various types of manipulations of
polymers in slow equilibrium with the extractable spectral data, for calculating reaction rates, for
species. resolving complex spectra into individual adsorp-
Spectrophotometric studies in both sulfuric and tion bands, for spectral band model and profile
hydrochloric acid solutions are being started, and s t u d i e s , for spectral band overlap s t u d i e s , and for
preliminary data for concentrated sulfuric acid the convolute smoothing of digitized spectral data.
solutions of protactinium from three different
sources have each given a different spectrum.
Recovery of the ^ ^ ' P a borrowed from England 11.1 High-Temperature, High-Pressure
for use in preparing ^^^U has been completed, and Spectrophotometer System
all the material has now been returned to England.
The equipment used for p r o c e s s i n g the ^•'^U some A high-temperature, high-pressure spectrophotom-
three years ago was exhumed from the burial eter and a s s o c i a t e d c e l l s and other s y s t e m s were
ground, and 1.65 g of protactinium was recovered specially designed for ORNL by the Applied Phys-
from the dissolver v e s s e l . i c s Corporation, but fabrication is not yet com-
plete. The main spectrophotometer is scheduled
for delivery by August. It is anticipated that in-
stallation of the locally constructed subsystems
can be completed before the arrival of the main
spectrophotometer. The design is such that, ulti-
mately, solutions containing high levels of alpha
The radiolysis of water adsorbed on s i l i c a gel
emitters of interest in heavy-element chemistry
was investigated in an attempt to determine whether
can be studied. The spectrophotometer system
or not energy transfer from the gel to the adsorbed
will permit the study of the solution chemistry be-
water took place. G(H ) values a s a function of
havior and equilibria, reaction kinetics, and s p e c -
water coverage were measured and found to be
tral properties of elements and complex s p e c i e s in
greater than the value for pure water, indicating
solution up to the vicinity of the water critical
either energy transfer or s e n s i t i z a t i o n of the ad-
point (372°C). Work has continued on the develop-
sorbed water.
ment of equipment and techniques for studying
aqueous solutions at high temperatures and pres-
s u r e s . The prototype spectrophotometric absorp-
11. HIGH-TEMPERATURE CHEMISTRY tion cell constructed as part of the subcontract
development work has performed satisfactorily in
The major objectives of this program are to de- t e s t s at ORNL. Conditions included temperatures
velop and to exploit various spectrophotometric up to 365°C and pressures to 8000 psi, applied
techniques for studying the properties of aqueous simultaneously.
solutions, primarily those of uranium and the trans-
uranium, rare-earth, and transition-group elements.
The principal effort i s directed at construction of a 11.2 Spectral Studies of Ionic Systems
high-temperature, high-pressure spectrophotometer
and associated equipment that can be used to study Studies of the effects of ionic s p e c i e s (complex-
spectra of t h e s e elements in solutions at tempera- ing ions) and temperature on the spectra of the
tures up to the vicinity of the critical point of uranyl ion in perchloric acid systems at fixed ionic
water. A s p e c i a l autoclave and a temperature- strengths were continued. Experimental conditions
control-and-measuring system were built for meas- of interest to us are: temperature, metal ion (uranyl)
uring liquid d e n s i t i e s at high temperatures and and ligand concentrations, acidity (hydrolytic com-
pressures by means of x-ray photography. Prelimi- plications), ionic strength, and isotope effects of
nary spectral s t u d i e s of U O ^ and P u ' ' ^ i o n s were the solvents (light and heavy water). Experiments
made up to 95°C with existing spectrophotometers have been carried out for the uranyl ion in per-

chloric acid and perchlorate systems in the acid- results on uranyl perchlorate solutions agreed with
e x c e s s region, at the stoichiometric point, and in pycnometric results to within 0.4%. Up to about
the region of hydrolysis over the temperature range 300°C dilatometer results on water agree with
of 25 to 95°C at several ionic strengths and at steam-table values to within 0.5% or better, using
several uranyl ion concentrations. As the acidity a conventional thermocouple system. At higher
is lowered through the stoichiometric point into temperatures the error increased because the tem-
the region where the uranyl ion becomes progres- perature was not known accurately enough. Equip-
sively more hydrolyzed, significant spectral changes ment installed recently permits the measurement of
are observed. Also, as the acidity i s lowered, the the temperature to within about 0.05°C.
effect of increased temperature in promoting hy-
drolysis becomes more pronounced.
11.4 Computer Programs for Spectrophotometric
Improved computer techniques developed for the
mathematical resolution of complex overlapping
spectra were applied to a study of the fundamental
A generalized nonlinear least-squares computer
parameters of the absorption spectra of the UO ^ ""^
program (MR0C0S) for the mathematical resolution
ion and related systems. Our results show that
of complex overlapping spectra has been com-
some prior inferences, from unresolved spectra,
pleted. The program will resolve complex over-
concerning the U O ^ band i n t e n s i t i e s (and half-
lapping spectral bands and fine structure, and is
band widths) are in error. We have, to our knowl-
now available for the lBM-7090 computer and auto-
edge, for the first time resolved the absorption
matic off-line plotting equipment. Using this code,
spectrum of the uranyl ion by nonlinear least-
the spectra of a number of uranyl ion systems have
squares computational means over the 3300- to
now been resolved.
5000-A region into 14 absorption bands. The ef-
A set of computer programs and subroutines were
fects on the spectra of U O j ^ of the progressive
written for the CDC 1604-A computer to smooth
complexation of UO^"*^ by NO ~ were studied as a
spectral data. A least-squares convolution-smooth-
function of several experimental parameters. Even
ing technique i s employed. Output from the very
under conditions where hydrolysis effects are non-
flexible set of programs can be in many forms, in-
existent, changes of varying magnitudes are ob-
cluding binary coded decimal and binary tapes,
served, for example, in all three parameters of each
tabulations, punched cards, or CALCOMP graphi-
band with all experimental conditions invariant
cal plotting. All necessary combinations of experi-
except temperature.
mental spectral data, cell-balance data, and coni-
cal-screen attenuator data can be handled and
11.3 Measurement of Liquid Densities at
High Temperatures and High Pressures

As an important adjunct to the program for the 12. MECHANISMS OF SEPARATIONS PROCESSES
spectrophotometric study of solutions at high tem-
peratures and high pressures, a method has been The thermodynamics of solvent extraction by
devised for measuring the densities of aqueous tributyl phosphate (TBP)—hydrocarbon-diluent solu-
solutions at accurately measured temperatures and tions are being studied by means of vapor pressure
pressures up to the solution critical points. A measurements and analytical partition data. Vapor
high-temperature, high-pressure autoclave (dilatom- pressures of water and nitric acid over the three-
eter) and related facilities were designed and built component system water—nitric acid—uranyl ni-
to permit the accurate measurement of liquid densi- trate hexahydrate and corresponding pressures
ties under these conditions. The volume of a over the two-component systems water—nitric acid
weighed solution of a known composition is deter- and water—uranyl nitrate hexahydrate were ana-
mined in the dilatometei by taking an x-ray photo- lyzed by integrating the Gibbs-Duhem equation.
graph to show the position of the vapor-liquid inter- This integration was performed by using an em-
face in a calibrated section of the dilatometer. pirical four-parameter function which is then used
This system has now been completely installed to calculate the activity coefficients of uranyl
and has been operated. At 95°C the dilatometer nitrate hexahydrate. These parameters describe

the variation of water and nitric acid activities 13.2 Effect of Ionizing Radiation on Coalescence
with acid and uranyl nitrate hexahydrate concentra- in Liquid-Liquid Systems
tions with a standard deviation of about ±10%.
The work on the effect of ionizing radiation on
The distribution of uranyl nitrate and water be- c o a l e s c e n c e of drops on a plane interface was
tween T B P and aqueous uranyl nitrate solutions completed and a mechanism which explains the
was used to calculate the a c t i v i t i e s of tributyl data was postulated. Data are presented on the
phosphate by means of the Gibbs-Duhem equation. effect of temperature, liquid properties, and drop
Rational activities of T B P are tabulated for aque- diameter, as well as radiation type and intensity.
ous uranyl nitrate concentrations ranging from 0 to Alpha and fission product irradiation alone are
2.0 M. These a c t i v i t i e s , together with literature effective in enhancing c o a l e s c e n c e , apparently
data on the activities of water and uranyl nitrate by chemically disturbing the surface tension over
hexahydrate, were used to calculate the thermo- a very small region during impact.
dynamic equilibrium constant for the partition
reaction. Partial molar volumes in the organic
phase were calculated and used to c a l c u l a t e a free
energy change of —7.47 kcal/mole for the partition
reaction. Organic-phase d e n s i t i e s are presented 13.3 In-Line Detection of Particles in
as an empirical equation, with the molar concen- Gas Streams by Scattered Light
trations of water and uranyl nitrate a s the inde-
pendent variables. The detection of particles in a gas stream by
scattered light depends on the intense illumination
A generalized computer program for the resolu- of a well-defined, very small volume. Recently,
tion of spectrophotometric data into a s e r i e s of a helium-neon laser was tested to replace the other
Gaussian distributions, with least-squares-adjusted conventional light s o u r c e s , and it increased the
parameters corresponding to band area, position, signal-to-noise ratio by a factor of 2 for particles
and half-width, is described. A second nonlinear in the micron range. Also, a technique was de-
program, useful for c a s e s difficult or even impos- veloped for electronically shaping the pulse from
sible to converge by the modified Gauss-Newton the photomultiplier detector so that the pulse can
method, is also mentioned. As a result of this be satisfactorily analyzed by a multichannel pulse-
work, preliminary information is presented on the height analyzer.
variation in area of spectral bands for 0.1 M uranyl
nitrate a s a function of nitric acid concentration
(0.01 to 7.9 M).

13.4 Solvent Extraction Engineering Studies:

Correlation of Pulsed-Column Flooding Data

13. C H E M I C A L E N G I N E E R I N G RESEARCH The literature and all available sources were

exploited to accumulate data on flooding in pulsed
13.1 The Stacked-Clone Contactor columns, and 2300 data points representing 23
different chemical systems were accumulated and
A s e v e n - s t a g e s t a i n l e s s s t e e l experimental used to test over 200 proposed correlating equa-
stacked-clone contactor was used to determine tions by a multiple-regression a n a l y s i s . Up to 13
operating performance with a variety of solvent variables were used, including column diameter,
systems including 100% TBP—1 M sodium nitrate, pulse-plate spacing, hole s i z e , free area, pulse
18% TBP—3 M sodium nitrate, and hexone-water. amplitude and frequency, solution d e n s i t i e s and
The unit was run at flow ratios ( A / 0 ) as high as v i s c o s i t i e s , and interfacial tension. Most correla-
50, with good c a p a c i t i e s and stage efficiencies tions grouped the variables in dimensionless groups.
consistently over 60%. The Purex flowsheet was When equations of the type originally proposed by
successfully demonstrated, a s was the ability of Thornton and Pike were modified to include new
the device to recover more than 99.9% of the coefficients generated by the regression a n a l y s i s ,
uranium. they fitted the data well.

14. REACTOR EVALUATION STUDIES chemical explosives at the Vincentown, New

Jersey, test area under a purchase order to Frank-
This program is supported jointly by the Reactor ford Arsenal. The t e s t s showed that copper or
Division and the Chemical Technology Division iron targets were focused into hypervelocity j e t s
and includes studies on various proposed advanced- and that the target material traversed the length
reactor and fuel-cycle s y s t e m s to determine their of the test chamber at a velocity comparable to
feasibility and economics. The work in this Divi- or faster than those at which the spherical shock
sion during the past year included calculation of front develops and moves out from a nuclear deto-
the c o s t s of fuel shipping and processing and of nation.
preparing sol-gel oxide in connection with the The production of industrial chemicals by utiliz-
advanced converter evaluation, ^^^U value study, ing underground nuclear explosions does not a p -
and other reactor evaluation programs. Also in- pear economically attractive. Unless yields from
cluded were the development of computer codes the nuclear device are about a megaton, the cost
for calculating shipping and processing c o s t s and of the heat energy or pressure generated by the
the overall cost of nuclear power, and for calculat- device is much too high.
ing individual and gross concentrations and radio- Laboratory tests were started to determine the
activities of thorium, uranium, transuranium e l e - rate of exchange of tritium and hydrogen in systems
ments, and fission products produced during reactor containing methane and water. This investigation
operation at constant flux or constant power. A is n e c e s s a r y to help evaluate the feasibility of
manual on the design of c a s k s for shipping spent stimulating gas production from wells by using
fuel was published. Criticality problems and nuclear explosives.
neutron-gamma shielding problems in fuel c y c l e Studies were begun on problems that would arise
plants v/ere also studied. from the presence of radiation emitters during the
processing of ore bodies that have been fractured
by a nuclear explosive. A feasibility study is
15. CHEMICAL APPLICATIONS OF being made on the recovery of magnesium oxide or
NUCLEAR EXPLOSIONS magnesium from large deposits of olivine (about
48% MgO) that may be fractured by nuclear ex-
Laboratory development of p r o c e s s e s for re-
covering transplutonium elements from the proposed Studies were begun of problems that might arise
Project Coach detonation was completed. The from the presence of radioactivity in the processing
tentative process flowsheet was tested on kilogram of copper ore bodies that have been fractured by
samples of radioactive debris from the Gnome nuclear explosives.
detonation, which is presumed to be similar to
that for Project Coach, and was satisfactory for
obtaining a primary concentrate of the transpluto- 16. PREPARATION AND PROPERTIES OF
nium elements. The process includes water leach- ACTINIDE-ELEMENT OXIDES
ing the debris to remove salt, which leaves a
water-insoluble residue that contains the trans- The chemistry underlying the current sol-gel
plutonium elements; then, acid leaching the water- process for making dense nuclear-fuel oxides is
insoluble residue to put the transplutonium ele- the surface chemistry of coUodial thoria and of
ments into solution; and, finally, concentrating other collodial oxides being prepared analogously.
the transplutonium elements by carrier precipitation The purpose of the Actinide Oxides Program is to
with calcium oxalate. Focusing electrophoresis determine the properties that control the behavior
did not appear promising as a method for separat- of t h e s e oxides during their preparation.
ing milligram quantities of transplutonium ele- The adsorption of nitric acid on crystalline
ments. In experiments in which lanthanides were thoria is limited to one nitrate ion per two surface
used as stand-ins, separation bands were always s i t e s , assuming equal exposure of the (100) and
overlapped. (111) faces. The free energy of adsorption is a
T e s t s to demonstrate a jet-sampler method for function of the fraction of reacted surface s i t e s ,
removing a target irradiated about a meter from a being about —5.7 kcal per mole of nitric acid at
detonating nuclear device were conducted with half of saturation. The enthalpy of adsorption,

determined from the temperature coefficient of the Laboratory Facility (also for the Analytical Chem-
surface equilibrium, is estimated to be about —7.6 istry Division); consultation to the Operations Di-
kcal/mole. vision on the construction and startup of the two
Hydrous amorphous thoria has been found to plant-waste-improvement projects; design and
undergo spontaneous crystallization at room tem- fabrication of equipment for the Health P h y s i c s
perature when allowed sufficient time, about 24 Division experiments on waste d i s p o s a l in salt
months in the present study. and w a s t e d i s p o s a l by hydrofracturing; design and
The viscosity of colloidal thoria in nitric acid installation of a ^^^U storage facility in Building
solutions at pH 2 implies that the fundamental 3019; design and installation of a test unit for
c r y s t a l l i t e s exist in the sol a s s o c i a t e d together evaluating absolute filters and charcoal b e d s in
in weak, reversible, highly unsymmetrical floes. the Nuclear Safety Pilot Plant for the Reactor
Viscometry of s o l s in pH 3.6 nitric acid a l s o im- Division; calculations, using advanced shielding
plies the e x i s t e n c e of highly anisotropic floes of c o d e s , of gamma and neutron s h i e l d s for special
crystallites which are only partially reversibly point isotropic fission s o u r c e s ; and irradiation-
dissociated. contamination-decontamination t e s t s on selected
Weight-loss data for dehydrating gels of nitrate- commercial protective coatings and p l a s t i c s .
stabilized thoria at temperatures below 600°C were
correlated by equations from absolute-reaction-
17,1 Eurochemic Assistance Program
rate theory, which showed the enthalpy and entropy
of activation of dehydration to be 10 kcal/faole
The Laboratory continued to coordinate the e x -
and —45 eu respectively. The enthalpy and entropy
change of technical information between Euro-
of dehydration were 9.1 kcal/mole and +22.2 eu
chemic and the AEC production s i t e s and National
Laboratories for the AEC Division of International
Sintering of — 400-mesh compacts of thoria-gel Affairs. E. M. Shank completed his third year at
particles at temperatures between 600 and 1100°C Mol, Belgium, a s U.S. Technical Advisor to Euro-
could not be correlated by any single-mechanism chemic during the construction and startup of the
sintering model, probably because of the existence Eurochemic fuel-processing plant. Construction
of an ultrafine (less than 1 fx) s i z e fraction, which of the plant and its auxiliary facilities is now
superimposed interparticle shrinkage on intra- about 90% complete. Cold startup will begin in
particle shrinkage. Compacts made from coarser mid-1965, and hot startup is scheduled for the end
gel particles, from which the ultrafine fraction of 1966,
had been removed, gave sintering results that
correlated with the previously established surface-
diffusion sintering model. At temperatures between 17.2 Projects for Improving ORNL Waste Systems
1000 and 1120°C, erratic growth, breakdown, and
regrowth of the grain structure of sintering thoria The Chemical Technology Division continued to
gel were observed in the early s t a g e s of sintering. supply consultation to the General Engineering
T h i s process was observed repeatedly, but its and Construction and the Operations Divisions dur-
c a u s e is still obscure. ing construction and startup of the two plant-waste-
system improvement projects. The first of t h e s e ,
the Melton Valley waste collection and transfer
system, was completed in February 1964 and is
17. ASSISTANCE PROGRAMS now in s e r v i c e . The second, the intermediate-
and high-level-waste evaporator and high-level-
During the past year, several engineering efforts waste storage t a n k s , will be completed and cold-
for others have been carried out by the Chemical tested by January 1966, This project was delayed
Technology Division. T h e s e include a continua- b e c a u s e of the late delivery of the two 50,000-gal
tion of the Eurochemic A s s i s t a n c e Program; liaison high-level-waste storage tanks. A preliminary
on the construction and startup of the High Radia- design and c o s t estimate was prepared for venting
tion Level Analytical Laboratory for the Analytical and filtering the off-gases from the six concrete
Chemistry Division; preparation of two preliminary intermediate-level-waste t a n k s , W-5 to W-10, to
design and c o s t e s t i m a t e s for a proposed Alpha the main plant s t a c k .

17,3 A Demonstration of the Disposal of Solid, contractor (H. K, Ferguson Company) during con-
High-Level Radioactive Waste in Salt struction of the High Radiation Level Analytical
Laboratory for the Analytical Chemistry Division.
A s s i s t a n c e was provided by the Division on the Installation of equipment was completed late in
experiment concerning the disposal of solid waste 1964, and the facility is in u s e . A s s i s t a n c e in-
in s a l t (Project Salt Vault). The experiment is cluded liaison on the completion of construction
being conducted by the Health P h y s i c s Division by H. K. Ferguson Company and ORNL field forces
in a salt mine at Lyons, K a n s a s . The first t e s t s and the preparation of operating and maintenance
will include the use of fully irradiated, 90-day- procedures and manuals.
cooled Engineering T e s t Reactor fuel assemblies
to simulate the heat and radiation from solid-
waste containers. All equipment for transporting 17,6 High-Level Alpha Laboratory
the fuel assemblies from Idaho to Kansas and
handling them at the mine has been designed and A preliminary design and c o s t study was made
built, and the first experiment is scheduled for for a new Alpha Laboratory Facility for joint use
September 1965. by the Analytical Chemistry, Chemical Technology,
and Metals and Ceramics Divisions, The labora-
tory would contain twenty-eight 16- by 32-ft alpha
17,4 Disposal of Radioactive Waste by laboratories, four supporting " c o l d " laboratories,
Hydrofracturing a 32- by 64-ft deep-bay area for large alpha experi-
ments, and office and service facilities. The
The engineering on the Health P h y s i c s Division facility, which would be administered by the
experiments on the d i s p o s a l of intermediate-level Analytical Chemistry Division, is estimated to
radioactive aqueous w a s t e s by hydrofracturing is cost about $5,000,000. At the request of manage-
being coordinated by the P r o c e s s Design Section ment, an alternative design was prepared in which
of this Division. In this program, the aqueous half the proposed facility was added to the High
w a s t e s are mixed with cement and clay, and then Radiation Level Analytical Laboratory and the
they are pumped at high pressure into an approxi- other half to the Thorium-Uranium Fuel Cycle
mately 1000-ft-deep hole in the ORNL test area Development Facility. This proposal was aban-
in lower Melton Valley. When sufficient pressure doned when it became apparent that s u c h an al-
is exerted, the rock strata are fractured at the ternative was considerably more expensive than
bottom of the hole, and the waste is distributed in the original proposal.
a thin s l a b around the hole, where it s e t s up a s a
concrete s h e e t .
17,7 Storage Facility for " 3 u , Building 3019
Cores of the grout s h e e t s of previous injections
have been obtained and analyzed. The mixes Oak Ridge National Laboratory now serves a s
seem satisfactory; radionuclide retention is quite a national storage and d i s p e n s i n g facility for
good, The injection facility was modified, and reactor-produced ^^^U fuel formed by neutron ir-
another injection was made to t e s t an improved radiation of thorium. Therefore, expanded storage
mix and changes in equipment and procedure. capacity for ^^^U nitrate solution was installed
Further coring will be done, and the facility will in the Building 3019 pipe tunnel, an additional
be modified for regular disposal of laboratory bank of storage wells for dry s o l i d s (^^^U 0 )
waste. was drilled into a cell dividing wall in the 3019
penthouse, and a small-scale solvent extraction
17.5 Construction and Startup of the High facility (1 kg of uranium per day) was installed
Radiation Level Analytical Laboratory in the newly constructed alpha-contained laboratory
located in the former solvent makeup room in the
The Chemical Technology Division continued 3019 operating gallery. During the past year, about
to supply the necessary technical liaison between 20 kg of stored 2 3 3jj ^ g g separated from the radio-
the Laboratory, the AEC, the designers (Vitro active daughters of ^^^U for designated AEC u s e s .
Corporation), the construction contractor (Foster The ^^^U inventory has been increased by addition
and Creighton Company), and cost-plus-fixed-fee of 40 kg containing 45 ppm of ^^^U and 80 kg

containing 200 ppm of ^^^U. Prior to the new ac- 18. M O L T E N - S A L T R E A C T O R PROCESSING
quisitions , nearly 50 kg of ^ ^ ^U with a ^ ^ ^U content
of 38 ppm was on hand. 18.1 Continuous Fluorination of Molten Salt

The most attractive process for removing the

uranium from Molten-Salt Breeder Reactor (MSBR)
17.8 Filter Tests in the Nuclear Safety Pilot Plant fuel involves continuous fluorination to produce
volatile U F . Although technology e x i s t s for batch
At the request of the Reactor Division, the fluorination, corrosion rates are high. A frozen-
Chemical Technology Division designed and in- wall technique for protecting a tower in which fluo-
stalled an absolute-filter—charcoal-bed combina- rine may be bubbled countercurrent to the fuel s a l t
tion unit for testing in the Nuclear Safety Pilot is proposed. For a reference 1000-Mw (electrical)
Plant (NSPP), The system includes a Demister, reactor, the fuel stream, cooled for 1.5 days, will
absolute filter, and a charcoal bed all incorporated have a specific heat generation rate of 3 x lO'*
in a remotely removable canister, a sealed com- Btu h r ~ ' ft~^, which will give an adequate heat
pressor for gas circulation, and three May-pack flux to support a frozen layer on a cooled wall.
samplers. T h e system is designed for a flow of Experimental work has been started on a l-in.-diam
20 cfm of steam and gas at 300°F and 50 psig nickel continuous countercurrent fluorinator with a
pressure. The radioactive g a s e s and particles for salt depth of 48 in., from which mass transfer data
the N S P P t e s t s will be generated by melting ir- can be obtained.
radiated fuel elements with a plasma torch in a
steam-air atmosphere.

18.2 Distillation of Molten Salt

The proposed Molten-Salt Breeder Reactor will

17,9 Studies of Radiation Shields for use a fuel carrier s a l t of ' L i F and B e F . After the
Fission Sources uranium has been removed by fluorination, this mix-
ture can be partially separated from the fission
Advanced shielding codes for an IBM 7090 com- product poisons, particularly the rare earths, by
puter were used to calculate the radiation d o s e
semicontinuous low-pressure distillation. At
rates from fission sources through s h i e l d s that
1000°C the still will operate at a pressure of about
are commonly used in cell w a l l s , cell windows,
1 mm Hg, and s i n c e B e F has a vapor pressure
and shipping c a s k s . The dose rates were calcu-
100 times that of L i F , the still pot will contain
lated for various concretes, water, CH , and mix-
little B e F ^ . A 1000-Mw (electrical) MSBR would
tures of H O or CH with lead, iron, and boron for
require a processing rate of about 15 ft^ of s a l t
shield thicknesses up to 150 cm.
per day. Preliminary experimental work h a s been
directed toward the determination of the relative
volatilities of the s a l t and contaminants. A cold-
finger technique for sampling the vapor phase of
17.10 Irradiation and Decontamination Evaluation an equilibrium still was employed. The trivalent
Tests on Selected Protective Coatings rare earths had relative volatilities, with respect
to L i F , ranging from 0.01 to 0.05. Cerium tetra-
Various commercial protective coatings and fluoride showed a relative volatility of about 0.15;
p l a s t i c s were evaluated in irradiation-contamina- and C s F , RbF, SnF , and Z r F are more volatile
tion-decontamination t e s t s . Gamma radiation was than L i F and cannot be separated this way. The
used, and the irradiations were carried to the l e s s volatile fission products are accumulated in
coating-failure level, Epoxy coatings were the the still bottoms and will be discharged on a time
most resistant, and vinyl coatings were the most cycle that will probably be determined by the heat
easily decontaminated. The samples were ir- generation rate. Small prototype s t i l l s have been
radiated both in air and in deionized water in a constructed to study the problems of heat transfer,
^"Co source. condensation, and boiling.

18.3 Reconstitution of Fuel for a Molten-Salt Experimental extractions were made with various
Breeder Reactor molten-metal phases, including mixtures of metallic
lithium, magnesium, aluminum, and tin. Little
Since the carrier s a l t and U F are purified sepa- s u c c e s s was observed in assimilating rare earths
rately, they must be recombined for return to the into these metal p h a s e s . However, intermetallic
reactor. This is best accomplished continuously compounds of the form Be M were found that were
in a column in which the barren salt and U F , are highly insoluble in the s a l t phase or any metal
phase tested. T h e s e compounds form granular
introduced at the bottom, along with salt contain-
solids which are e a s i l y removed by sedimentation
ing U F , which is recycled from the top. The UF
or filtration. Studies to exploit this possibility for
first reacts with the UF^ to form a nonvolatile
fuel-salt processing are continuing.
intermediate fluoride, which is then reduced again
to U F in the upper part of the column with hydro-
gen. Three t e s t s were made in a 4-in.-diam reac-
tion v e s s e l where the off-gas could be monitored
to detect UF^ that had not been absorbed. Com-
plete absorption of the U F had occurred, to the Oak Ridge National Laboratory is carrying out
limit of detection. Better data on the rate of ab- a program of basic research on the properties of
sorption and subsequent hydrogen reduction are water and its solutions under the auspices of the
necessary for optimal design. Office of Saline Water, Department of the Interior.
This program has as its long-range goal the devel-
18.4 Chromium Fluoride Trapping opment of methods for the economical purification
of water in such amounts and of such purity that
Among the problems associated with the terminal it may be used for irrigation and for drinking. Work
recovery of the uranium from the fuel salt of the in the program is under the direction of K, A, Kraus
Molten-Salt Reactor Experiment (MSRE) is the and is interdivisional. It is reported in part in a
separation of chromium fluoride from UF . The s e r i e s of reports issued annually by the Depart-
chromium is introduced by corrosion. Packed beds ment of the Interior. The most recent report in
of pelleted N a F at 400°C have been found to effec- the s e r i e s is Saline Water Conversion Report,
tively trap the volatile CrF . Eight t e s t s were 1964. Because the work is reported formally in
made with a sodium fluoride bed 1'/ in. in diameter the above s e r i e s , only this abstract of the work
with an effective length of 6 ft. The bed accommo- carried out by Chemical Technology Division
dated gas rates (CrF^ in fluorine) of 0.4 to 0.9 std personnel is presented in this report.
liter/min. Over periods from 2 to 5 hr, 11 to 20 g
Studies carried out in the Chemical Technology
of chromium fluoride was deposited. Different
Division have followed several l i n e s . One has
kinds of sodium fluoride pellets were tested, one
been measurements of thermodynamic properties
of which is found to be quite satisfactory for use
of water-organic liquid mixtures, in the hope that
in processing the MSRE fuel.
such mixtures will exhibit the properties of, and
so serve as models for, salt-excluding organic
membranes. This hope has been reinforced by
18.5 Alternative Processing Methods for a the finding that the distributive selectivity and
Molten-Salt Breeder Reactor the activity coefficients of inorganic s a l t s in the
mixtures are similar functions of their water con-
A scheme that may be competitive to the d i s - tents over a wide range of organic structures, from
tillation method for purifying the fuel salt is e s t e r s to amides and organic s a l t s . Correlations
reduction of the fission product poisons to form of activity coefficients and miscibility gaps with
metals that might be either assimilated into a molten dielectric c o n s t a n t s , s a l t types, and water thermo-
metal phase or filtered from the s a l t . Since the dynamic activity have contributed to the informa-
free energies of formation of the rare-earth fluo- tion necessary to e s t a b l i s h the utility of organic
rides lie between those of L i F and B e F , for re- liquids a s membrane models.
duction of rare-earths to take place it i s necessary In another line of study, methods are being
that stable intermetallic compounds be formed. sought to alter the hydrodynamic conditions near

a solid surface to promote r a t e s of mass transfer A third line of study is directed to metals and
of a solute between the surface and the circulating metal oxides and the relation of their surface
fluid. (This study is under the direct supervision properties to the adhesion of inorganic s a l t s pre-
of D, G. Thomas, Reactor Division, ORNL.) This cipitated from aqueous solution onto them. T h i s
research is specifically intended to reduce the work has shown that the adhesion of s a l t s pre-
concentration polarization problems encountered cipitated from solutions containing orthophosphate
in reverse osmosis and electrodialysis of brackish is profoundly influenced by pH and orthophosphate
w a t e r s . The study is focused upon measurements concentration. Further, a fixed stoichiometry
of mass-transfer promotion from wires or cylinders e x i s t s between the phosphate molecules which are
supported above the mass-transfer surface near sorbed on the metal oxide surfaces and the number
the edge of the hydrodynamic boundary layer. Us- of metal oxide s i t e s on the surface. The stoichi-
ing an air-naphthalene system, Thomas has found ometry was found to be determined by the shape
that such detached promoters may give average of the phosphate molecule, the concentrations of
mass-transfer rates twice a s high a s those ob- the various dissolved phosphate s p e c i e s (e,g.,
tained without the promoters, and at certain points P 0 ^ ^ ~ , H P O ^ ^ ~ , H ^ P O ^ " ) , and the geometric
the rates may be three or four times as high. array of the reactive s i t e s on the surface of the
T h e s e studies are now being extended with an solid. The kinetics of phosphate sorption on
aqueous system with properties closer to those thorium, cerium(IV), and zirconium oxides was
found in desalination p r o c e s s e s . Initial results found to be described by a rate equation which
with this system indicate that the percentage in- contains both a linear depletion term and a term
c r e a s e in mass-transfer rates in the aqueous system exponential in the extent of coverage. This is
will be at l e a s t a s high a s that observed in the interpreted to mean that the activation energy for
gaseous system, although there are not enough sorption of phosphate i n c r e a s e s linearly with
data yet to predict what the optimum conditions the quantity of phosphate sorbed.
will be or what degree of promotion ultimately can
be obtained.
1. Power Reactor Fuel Processing

Laboratory and engineering-scale development T h 0 2-U02 fuel samples irradiated to 75,000 Mwd
of p r o c e s s e s for recovering fissionable and fertile per metric ton of thorium. Alternative adsorption
material from spent power reactor fuels i s contin- s t u d i e s with silica gel and zirconium phosphate
uing. Considerable emphasis was placed on b a s i c (Bio-Rad ZP-1) were also made. Unirradiated
chemical s t u d i e s ; however, major efforts were also UO,—17% PuO„ fuel pellets were successfully
continued on chemical applications, engineering oxidized to U j O g - P u O j powder in O2 at 800°C;
development, and small-scale hot-cell t e s t i n g of p e l l e t s containing more than 20% PUO2 were
the more promising fuel-recovery processes. unreactive. The resulting U O -PuO powders
B a s i c chemical studies on carbide fuels included were then chlorinated at 500°C in 85% C l 2 - 1 5 %
nitric acid hydrolysis experiments with U C , , the CCl ; volatilization of uranium and plutonium chlo-
preparation of pure ThC, ThC^, and ThC-ThCj rides was only partially successful. Reduction of
mixtures and studies of their hydrolytic charac- Pu(IV) to Pu(III) in nitrate solution with Ar-4% H2
t e r i s t i c s in water, and the effect of tungsten was satisfactorily demonstrated. The c a t a l y s t was
carbide impurity on the hydrolysis of both uranium platinized alumina.
and thorium carbides. P r o c e s s development for Development of the shear-leach p r o c e s s with
thorium-uranium-graphite fuels included alumina unirradiated fuel elements on a full engineering
fluidized-bed burning studies on pyrolytic-carbon- s c a l e was continued. Work l a s t year included:
coated ThC^-UCj and Th02-U02 fuel particles (1) removing fuel rods from second-generation
contained in graphite. Nitric acid leaching s t u d i e s fuel assemblies by remote row pulling; (2) shearing
were conducted with the fuel a s h . Wet-combustion— t e s t s on loose bundles of fuel rods and on Zircaloy-
leaching studies in oxygen and dilute nitric acid 2-clad uranium metal (annular fuel); (3) further
at 250° C and 1500 to 2000 psi were also begun. safety evaluations on the mechanical head-end
Development of burn-volatility and burn-leach handling of zirconium fines; (4) evaluations of
p r o c e s s e s for graphite-based Rover fuel was handling problems of, and fission product heat
completed. generation in, b a s k e t s of sheared fuels; and (5)
further leaching t e s t s on sheared UO2 and T h 0 2 -
Hot-cell comparison t e s t s of the Sulfex and
UO2 fuels. Also, a monitor for determining the
shear-leach p r o c e s s e s were made with s t a i n l e s s -
residual " * U , " ' P u , or " S y content of leached
steel-clad UO2-20% P u O j and T h 0 2 - 4 % UO^
metal hulls by delayed-neutron activation a n a l y s i s
fuel samples irradiated to 99,000 Mwd per metric
was developed. The monitor i s s e n s i t i v e to a
ton of fuel. Stainless-steel-clad oxide fuels were
residue equal to l e s s than 0.01% of the fissile
successfully declad with HF-O^ at 600°C. Nitric
material before leaching.
acid leaching studies were made on the declad
A new program for developing low-decontami-
material both with and without a prior pyrohydroly-
nation close-coupled p r o c e s s e s for Th-U recycle
s i s step at 600°C (with steam or 0 2 - 3 % Hfi)
fuels was begun. Results of scouting t e s t s to
designed to remove residual fluorides formed
evaluate the capability of e l e c t r o d i a l y s i s , ion
during decladding and to reduce the solubility of
exchange, and precipitation processes were
iron during leaching.
relatively unsuccessful; therefore, the main effort
The adsorption of ^^^Pa on unfired Vycor and was devoted to the development of low-cost
i t s subsequent elution with oxalic acid was differential solvent extraction equipment and
successfully demonstrated with 26-day-cooled methods. Amine solvent extraction was evaluated


a s an alternative to steam denitration of thorium Hydrolysis of Uranium Carbides in Nitric Acid^'^

nitrate for the preparation of thorium s o l s .
Conceptual plant studies were completed for Dissolution of uranium carbides in 4 and 16 M
(1) a small-capacity (1/10 to 1/6 ton/day) fuel HNO3 at 90°C was investigated to determine the
processing plant using conventional head-end and composition of the g a s e s evolved. Other s t u d i e s
solvent extraction technology and (2) an HTGR were conducted to determine the behavior of the
graphite-base fuel processing plant employing a carbides in 0.001 to 0.5 M HNO3 solutions. Rel-
bum-leach—solvent-extraction flowsheet. atively high-purity uranium monocarbide (UC),
sesquicarbide [U^(C2)3], ^""^ dicarbide (UCj gg)
samples were used in all c a s e s . Each carbide,
1.1 HYDROLYSIS OF URANIUM AND when contacted at 90° C with 4 and 16 M HNO 3,
THORIUM CARBIDES yielded uranyl nitrate, soluble organic a c i d s ,
NO 2, NO, CO2, and t r a c e s of N2O. No hydrogen,
Uranium, thorium, and plutonium carbides are CO, or gaseous hydrocarbons (the hydrolysis
starting to b e used a s reactor fuels, both as the products with water alone) were produced. With
pure compounds and a s ingredients in fuel-graphite 4 M HNO 3, NO was the predominant nitrogen oxide,
mixtures. For example, uranium monocarbide in whereas NO 2 was the chief nitrogen oxide evolved
NaK-bonded s t a i n l e s s steel tubes i s being used in reactions with 16 M HNO 3 (Table 1.1). Between
as the second-core fuel for the Consumer's Public 50 and 80% of the carbide carbon was converted
Power (Hallam, Nebraska) reactor, and pyrolytic- to CO ; the remainder was converted to oxalic acid
carbon-coated UC2-ThC2 p a r t i c l e s in a graphite (0 to 11%), mellitic acid (2 to 9%), and unidentified
matrix are being used as fuel in the High-Temper- highly substituted aromatic compounds.
ature Gas-Cooled reactor (HTGR, Peachbottom,
The carbides were almost completely p a s s i v e
Pennsylvania). The TARGET-type reactor will
in boiling 0.001 to 0.5 M HNO 3 and in solutions
u s e undiluted carbon-coated ThC2-UC2 p a r t i c l e s
of sodium, calcium, uranyl, aluminum, and thorium
packed in longitudinal holes in graphite fuel rods; nitrates where the nitrate concentration was
plans are also under way to u s e carbon-coated 0.05 M. This behavior h a s not yet been explained.
T h 0 2 - U 0 2 microspheres a s fuel. Development
efforts on the processing of fuel carbides are
d i s c u s s e d in this section; the work on p r o c e s s Preparation and Hydrolysis of Thorium Carbides '
developments for graphite-base fuels i s reviewed
in Sect. 1.2. Thorium carbide specimens with total-C/Th
B a s i c chemical studies on the preparation and atom ratios varying from 0.8 to 2.1 (4 to 10 wt %
hydrolysis of uranium, thorium, and other carbides carbon) were prepared by arc-melting high-purity
were continued. This work included s t u d i e s on thorium metal and carbon. T h e s e specimens were
(1) the hydrolysis of U C j in nitric a c i d ; * ' ^ ' ^ (2) examined by chemical, x-ray diffraction, and
the preparation and hydrolysis of ThC, T h C j , metallographic a n a l y s e s , and their hydrolysis
and ThC-ThC2 mixtures in water; "'^ and (3) the properties studied.'*-* The thorium monocarbide
effect of tungsten carbide impurity on the hydrol- p h a s e was found over a range of compositions
y s i s of uranium and thorium carbides. from about ThC^ ^^ (4.6% carbon) to at l e a s t
ThCg gj (4.0% carbon) and probably lower. The
maximum combined-C/Th atom ratio obtained by
L. M. F e r r i s and M. J . Bradley, Off-Gases from the
Reactions of Uranium Carbides with Nitric Acid at 9(Pc, arc-melting with graphite electrodes was 1.95
ORNL-3719 (December 1964). (9.2% carbon) rather than the expected 2.00 (9,4%
L. M. F e r r i s and M. J . Bradley, " R e a c t i o n s of the carbon). Specimens with compositions varying
Uranium Carbides with Nitric A c i d , " to be published in from T h C p g g to T h C ^ g g (4.9 to 8.9% carbon)
the Journal of the American Chemical Society.
3 were two-phase mixtures of the mono- and di-
J . R. Flanary et al., Hot-Cell Studies of Aqueous
Dissolution Processes for Irradiated Carbide Reactor carbides. There was no evidence for any signif-
Fuels, ORNL-3660 (September 1964).
4 icant range of composition for the dicarbide.
M. J . Bradley and L. M. F e r r i s , " H y d r o l y s i s of
Thorium Carbides Between 25 and 9 9 ° C , " to be pub-
lished in the Journal of Inorganic and Nuclear Chemistry.
^M. J, Bradley and L. M. F e r r i s , " T h e Effect of M, J, Bradley and T, M, Kegley, J r , , Correlation of
Tungsten on the Hydrolysis of Uranium D i c a r b i d e , " Composition with the Microstructures of Arc-Cast
to be published in Inorg. Chem. 4 (May 1965), Thorium Carbides, in preparation.

T a b l e 1 , 1 , Approximate Compositions of the Gases Evolved in Reactions of

the Uranium Carbides with Nitric Acid at 90 C

HNO 3 Gas C omposition (mole %)

Carbide C Dncentration

UC 4 74 12 2 12

UC" 16 7 88 0 5

"4(^2)3 4 74 8 4 14

U,(C2)3 16 26 62 1 11

4 68 11 3 18
UCi.85 16 21 65 1 13

Reaction time in this experiment was only 6 hr.

The thorium carbides reacted with water to ORNL-DWG 65-9700

produce thorium oxide and hydrocarbons,'* Thorium

monocarbide produced mostly methane, and also
two moles of hydrogen for each mole of thorium
in e x c e s s of the 1:1 thorium-to-carbon atom ratio.
E 80
Thorium dicarbide (ThC^ g^) produced C j to Cg
hydrocarbons, wax, and hydrogen. The prepon-
derance of hydrocarbons with an even number of
carbon atoms in the gas from the hydrolysis of the o
a 40
dicarbide i s to be expected b e c a u s e of the pres-
ence of discrete C^ units in the dicarbide crystal
l a t t i c e . ' Carbon spacing in the thorium monocar-
bide lattice h a s not been studied, but the carbon
is probably present a s single C units a s in uranium
0 6 0 8 \.0 1.2 (4 (6 t.8 2.0 2.2
monocarbide,* which also yields methane when TOTAL-C/xh ATOM RATIO
hydrolyzed.^ Varying the reaction temperature
between 25 and 99°C caused no change in the F i g . 1.1. Effect of C / T h Atom Ratio on the Volume of
hydrolysis products from either the thorium mono- Gas Evolved in the Hydrolysis at 80 C of As-Cast
or dicarbides, although the reactions were con- Thorium Carbides.
siderably faster at the higher temperatures. The
total volume of gas evolved decreased from 110
ml ( S T P ) / g to 49 ml/g as the t o t a l - C / T h atom ane and an increase in the amount of C2 to Cg
ratio increased from 0.81 to 1.95 (Fig. 1.1); there hydrocarbons a s the total-C/Th atom ratio in-
was no further change in volume a s the t o t a l - C / T h creased ( F i g . 1.2). Within experimental error,
atom ratio increased above 1.95. Specimens with the specimen with the total-C/Th atom ratio of
compositions between ThC and T h C j ^^ gave 2.14 (combined-C/Th ratio of 1.95) gave the same
the products expected for ThC-ThC^ ^^ b i n a r i e s , products as the specimen with the t o t a l - C / T h
showing a regular d e c r e a s e in the amount of meth- ratio of 1.95.

Effect of Tungsten Impurity on Hydrolysis

E . B. Hunt and R. E . Rundle, / . Am. Chem. Soc. of Uranium and Thorium Carbide
73, 4777 (1951).
A. E. Austin, Acta Cryst. 12, 159 (1959).
M. J . Bradley and L. M. F e r r i s , Inorg. Chem. 3, 189
Tungsten, as an impurity in uranium and thorium
(1964). dicarbides, had a pronounced effect on the hydrol-

ORNL-DWG 65-9701 y s i s reaction, causing a marked d e c r e a s e in the

amount of gaseous C2 to Cg hydrocarbons and
.r^^ an i n c r e a s e in the amount of wax and free hydrogen
20 \ ^
(Table 1.2).'*'' This emphasizes the importance
of preparing and using high-purity carbides for
n 1 1
80 hydrolysis s t u d i e s and indicates that some of
the d i s c r e p a n c i e s in the literature on carbide
60 Cj - C s H
\/ / chemistry may be caused by relatively small
amounts of impurities in the specimens used by
40 / the investigators. The tungsten in the uranium
dicarbide specimen was present primarily a s
^ 20 _^ WUC2, which was recovered as an insoluble
Q > ^
^ 1^ residue after hydrolyzing the carbide and d i s -
O O solving the resulting U 0 2 ' x H 2 0 in hydrochloric
<^ 100
C/5 acid. About half the tungsten in the thorium
=) ^
o / dicarbide specimen was recovered a s W2C after
uj 80
< ^ hydrolysis; it i s not known which chemical s p e c i e s
/ ^
of tungsten (W, WjC, or WC) reacted.
60 i ^ ^

40 \

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 Development of p r o c e s s e s for recovering fuel
TOTAL-C/.f^ ATOM RATIO materials from spent thorium-uranium-graphite
fuels h a s included studies on the burning of
F i g , 1.2. Effect of C / T h Atom Ratio on the Gaseous pyrolytic-carbon-coated ThC2-UC2 and T h 0 2 - U 0 2
Products from the Hydrolysis at 80 C of As-Cast Thorium fuel particles contained in graphite (TARGET
Carbide. reactors, or the Peachbottom reactor). Burning

Table 1.2, Effect of Tungsten Impurity on the Reaction of Uranium and

Thorium Dicarbides with 80 C Water

W/(U or T h ) Atom R a t i o for:

Thorium Dicarbide Uranium Dicarbide

0.00005 0.09 0.00003 0.10

Volume of g a s evolved,

ml (STP) per gram of carbide 48 67 39 51

Gas composition, vol. %

Hydrogen 30 74 33 75

Methane 3 3 17 7

C™ to Cj, hydrocarbons 67 23 50 18

Carbon d i s t r i b u t i o n , 7o of carbon r e a c t i n g

Methane 1 1 4 3

C j to Cg g a s e o u s h y d r o c a r b o n s 51 30 33 20

Wax 12 52 20 52

U n a c c o u n t e d for 35 17 43 25

was done in a fluidized bed of alumina; the result- b e c a u s e it provides uniformly controlled combus-
ing mixtures of alumina and fuel oxides were tion and b e c a u s e it i s easily adapted to continuous
satisfactorily leached in HNO^, catalyzed with operation. The ThC2-UC2 fuels bum to a T h 0 2 -
HF, to recover the fuel. A pressurized aqueous UgOg ash, which i s finely divided and becomes
combustion (PAC) process is also being evaluated. dispersed uniformly through the AljO^ fluidizing
Here, the fuel is contacted with dilute nitric acid medium. The T h 0 2 - U 0 2 sol-gel microspheres do
and oxygen to produce CO 2 and a fuel nitrate not break up during burning and collect at the
solution. This i s done at a pressure of 1500 to bottom of the fluidized bed b e c a u s e of their high
2000 p s i , at temperatures up to 300°C. Develop- density. T h u s , for the latter fuel, very little of
ment of both aqueous and volatility p r o c e s s e s for the AI2O3 follows the Th02-U02 into the leaching
Rover fuels was completed. Both p r o c e s s e s step.
employ a common fuel-combustion head-end The conditions used in a typical burning run
operation. in a 2-in.-diam fluidized-bed burner are given in
Table 1.3. The product bed was green-gray and
consisted of slightly discolored A ^ O , and finely
Burn-Leach Process divided ThOj-UgOg powder; the alumina did not
powder. No sintering or clinker formation occurred,
The work on the burn-leach process for Rover and the bed material was free flowing. The s i e v e
fuels ^° h a s been modified and extended to apply a n a l y s i s of the product i s given in Table 1.4. The
the process to pyrolytic-carbon-coated ThC2-UC2 dry bulk density of the bed, 28% of which consisted
and ThOj-UOj fuels. This effort has included of fuel oxides, was 2.25 g/cm^.
engineering-scale t e s t s of fluidized-bed burning Except for a short time during startup, the off-
(alumina beds) and leaching of crushed HTGR g a s , a s shown by the LIRA infrared absorption
(Peachbottom) fuel c o m p a c t s * ' and laboratory- analyzers, consisted of more than 95% C O , and
s c a l e leaching t e s t s on both HTGR-fuel burner l e s s than 3% CO. Utilization of O2 was almost
ash and Th02-U02 microspheres. The mixed complete. Temperature control was very e a s y , and
oxide (or carbide) fuel particles are 150 to 500 no cooling was needed during most of the run.
H in diameter, and the pyrolytic carbon coatings After all the fuel had been charged, burnout of
are about 100 fi thick. the carbon took about 3 hr. During the first hour,
Burning can be done in fluidized beds of alu- the CO 2 concentration was higher than 90%; during
minum oxide or similar inert material as the heat the second hour, it fell slowly to about 60%, and
transfer agent. The fluidized bed i s preferred the CO content rose slightly. In the l a s t hour,
there was a rather abrupt d e c r e a s e in both CO2
and CO content; zero reading on the LIRA's was
Rover Fuel Reprocessing Development, June 1, reached in about 45 min.
1963, to July 1, 1964, ORNL-CF-64-12-5 (confidential).
H T After the bed was cooled, it was transferred
Unit Operations Quarterly Report for April-June,
1965 (ORNL-TM report, to be published). to a 1.5-in.-diam g l a s s fluidized-bed leacher and

T a b l e 1.3. Fluidized-Bed Burning Run Conditions for H T G R Fuel

Fuel Crushed ( Z-in.-diam) HTGR " A " fuel compacts, each containing
52 g of Th, 10 g of U, and 285 g of C; total charge 2.7 kg; 200 g
charged initially; 100 g charged every 15 min thereafter

Alumina Equal weights of 60- and 90-mesh Norton RR grade; 0.6 liter,
1.21 kg, 12 in. settled height

Temperature 750 to 775°C

O flow rate 12 l i t e r s / m i n (STP); 1.15 f t / s e c at 750°C

Product A l j O j - T h O j - U j O g ; 0.7 liter, 1.664 kg


T a b l e 1.4. Sieve Analysis of Product from with boiling reagent, more than 99.5% of both
F l u i d i z e d - B e d Burner the uranium and thorium were recovered when the
nitric acid concentration in the leachant was
Sieve Mesh Size Weight Percent 4 M or higher and the hydrofluoric acid concen-
tration was 0.02 to 0.05 M (Table 1.5). The
-30+50 21.2 presence of up to 0.1 M A1(N03)3 in the leachant
-50+60 12.7 had no adverse effect on the recoveries. Each
- 6 0 + 100 30.2 product solution was about 0.6 M in thorium, and
-100 + 140 16.4 l e s s than 2% of the alumina had been dissolved
-140 + 170 3.6 in each c a s e . Uranium and thorium recoveries
-170+200 2.5 were not acceptable when the bed material was
-200 13.5 leached with either 13 M HNO3 alone or with
2 M HNO3-O.O5 M HF (Table 1,5), indicating the
need for the fluoride c a t a l y s t and relatively high
nitric acid concentration in the leachant. L e a c h i n g
leached with boiling, circulating Thorex dis- at temperatures below the boiling point will be
solvent (13 M HNO3-O.O4 M H F - 0 . 1 M Al^"^). investigated in future s t u d i e s .
After 1 hr, the Th02-U308 was 90% dissolved; Another type of graphite-base fuel being con-
after 7 hr, dissolution was 99.7% complete. The sidered c o n s i s t s of carbon-coated Th02-U02
leached bed was finally washed with water. The microspheres dispersed in a graphite matrix.
combined leach-wash solution was 0.26 M in Th. Although no extensive experimentation with this
The maximum p o s s i b l e concentration of thorium type of fuel has yet been done, the oxide particles
in the blended stream would be about 0.35 M for are not expected to be affected during combustion
bed material of this loading (28 wt %), the limita- of the graphite matrix and should s e t t l e to the
tion being the solids-disengagement volume re- bottom of the burner b e c a u s e of their high den-
quired at the overflow. The useful range of sity.*^ All the T h 0 2 - U 0 2 microspheres could
fluidization velocity was 0.58 to 0.92 c m / s e c . then be withdrawn from the burner with very little
At l e s s than 60% bed expansion, corresponding dilution by alumina. Thus, the leaching step
to a solution velocity of 0.58 c m / s e c , the bed was would e s s e n t i a l l y involve dissolution of undiluted
not fully fluidized by the Thorex solution. T h 0 2 - U 0 2 microspheres.
The leaching equipment included a 1 Vj-in.-diam In recent t e s t s , nearly theoretically d e n s e s o l -
g l a s s column in which the bed material was gel T h 0 2 microspheres, 300 to 600 ^ in diameter,
supported on a perforated Teflon disk covered were dissolved in 3 to 6 hr in boiling 13 M HNO3—
with ^, ,-mesh A L O , . The overflow solution from 0.05 M H F (Table 1.6); the product solutions
10 2 3
were about 0.5 M in thorium. Similar r e s u l t s were
the top of the column was continuously airlifted
obtained even when the microspheres were dissolved
to a standpipe that provided head for returning
in the presence of a large e x c e s s of Norton RR
the solution through a s t a i n l e s s s t e e l heater and
alumina. In contrast to the finely divided oxide
rotameter. A hot oil jacket on the 1/'2-in- g l a s s
product from burning of fuel that contains Th-U
section permitted c l o s e temperature control of
dicarbide p a r t i c l e s , Th-U dioxide microspheres
the bed. Disengagement of solids in the 3-in.-diam
probably cannot be dissolved in a short time in
enlarged upper section of the column was s a t i s -
solutions of low nitric acid concentration (Table
factory a s long a s the solution was not permitted
1.6). In 6-hr experiments, the amounts dissolved
to boil actively. Total entrainment of s o l i d s in
increased with increasing nitric acid concentration,
the column overflow amounted to l e s s than 1% of
other conditions being the same.
the weight of leached s o l i d s , and c a u s e d no
operational difficulties.
Laboratory-scale studies have also shown that
E. L. Nicholson, L. M. Ferris, and J. T. Roberts,
excellent uranium and thorium recoveries can be Bum-Leach Processes for Graphite-Base Reactor Fuels
achieved by leaching the product bed with various Containing Carbon-Coated Carbide or Oxide Particles,
paper presented at the EURATOM symposium on "Fuel
boiling, fluoride-catalyzed nitric acid solutions. Cycles of High-Temperature Gas-Cooled Reactors,"
The burner product bed consisted of 6% UjOg, Brussels, Belgium, June 10-11, 1965, ORNL-TM-1096
(Apr. 2, 1965).
25% T h O j , and 69% Mfiy In 5- to 7-hr l e a c h e s

T a b l e 1.5. Results of Laboratory-Scale Leaching of F l u i d i z e d - B e d Material with Boiling

H N 0 3 . H F - A I ( N 0 3 ) 3 Solutions

Bed material: 6% UgOg, 25% T h 0 2 , 69% Norton RR alumina

Leaching time: 5 to 7 hr; final solutions were about 0.6 Af in Th

L e a c h a n t Composition (Af) Amounts L e a c h e d (%)

HNO, HF Al(NO,), u Th Al

2 0.05 0 83.6 83.0 0.2

4 0.05 0 99.9 99.9 1.9

4 0.05 0.1 99.7 99.5

13 0.0 0 27.7 14.8

13 0.01 0 61.1 57.6

13 0.02 0 99.9 99.6

13 0.05 0 99.9 99.9 1.9

13 0.05 0.1 99.9 99.9

T a b l e 1.6. Dissolution of Sol-Gel T h 0 2 Microspheres in Boiling Fluoride-Catalyzed Nitric Acid

Reaction time: 6 hr

Microsphere Diameter Reagent Composition (Af) Amount Dissolved

HNO, HF (%)

250-300 2 0.05 20
420-600 2 0.05 33
250-300 4 0.05 57
420-600 4 0.05 73
250-300 13 0.05 100
420-600 13 0.05 100

Pressurized Aqueous Combustion or for the oxidation of c o a l . ^ ' T e s t s have been

of Graphite Fuels made at temperatures up to 300°C both in a batch
autoclave and in a system in which g a s e s were
Combustion of the carbon matrix and subsequent continuously admitted and removed. Most of the
dissolution of the uranium and thorium from
graphite-base fuels in a pressurized v e s s e l con-
taining nitric acid and oxygen i s being studied 13 F , J, Zimmerman, Chem. Eng. 65, 117 (1958).
a s a low-temperature alternative to the burn-leach 14rG. H. T e l e t z k e , "Wet Air O x i d a t i o n , " presented at
process. Such a process appeared potentially 56th Annual Meeting, AIChE, Houston, T e x , , Dec, 1-5,
attractive based on results reported for the Zim-
' ^ T . R. Savich and H. C, Howard, Ind. Eng. Chem.
merman process for sewage or w a s t e d i s p o s a l ' ^ • ^ ' ' 44, 1409 (1952).

T a b l e 1.7. Effects of O j / C and N 0 3 / ( T h + U) Mole Ratios on the

Combustion-Dissolution of H T G R Fuel in Pressurized Nitric Acid Solutions

Temperature: 300°C
F u e l composition; 12% Th, 6% U, 82% C

Mole Ratios Amounts Burned or Dissolved (%)

Reaction Time
Experiment No,
(hr) Oj/C NO^3/(Th + U) C Th U

1 24 0.58 63 89.3 6.0 8.0

2 24 0.71 63 91.4 8.0 19.0
3 24 1.1 57 99.6 99.9 99.7
4 24 1.6 61 100.0 100.0 100.0
5 24 1.7 11 99.9 38.0 33.0
6 24 2.0 12 100.0 50.0 53,0
7 24 1.8 32 99.6 60.0 96,0
8 6.3 2.1 57 86.8 17.0 12.0
9 6.0 1.9 113 95.0 36.0 45.0
10 24 0.0 53 44.0 0.0 2.0

effort was devoted to the nitric acid oxidation- high N 0 3 / ( T h + U) ratio was not surprising,
dissolution of unfueled graphite and simulated considering that thorium and uranium contained
HTGR fuel (carbon-coated Th-U dicarbide particles in nitrate solutions hydrolyze at high temperatures.
dispersed in a graphite matrix); however, some Previous s t u d i e s of the hydrolysis of thorium
work also was done with other aqueous solutions nitrate s o l u t i o n s ' * showed that a minimum NO / T h
and with T h 0 2-U02 microspheres. mole ratio of about 25 was required to keep thorium
In the batch autoclave t e s t s , the effects of the in solution at 300° C in 2 M HNO 3.
O j / C and N 0 3 / ( T h + U) mole ratios on the oxi- In those experiments where only about 90% of
dation and dissolution of HTGR fuel samples the total carbon was oxidized (Table 1.7, experi-
(12% Th, 6% U, 82% C) were investigated. The ments 1, 2, and 8), the residue appeared, under
coated fuel particles in the sample were about the microscope, to consist solely of unattacked
500 ^ in diameter (400-fi-diam dicarbide kernels carbon-coated p a r t i c l e s . The same behavior was
with 50-|:i-thick pyrolytic carbon coatings); about noted in 6-hr reactions with 2 and 4 M HNO . Most
90% of the total carbon was present a s the graphite of the matrix was oxidized, but the p a r t i c l e s re-
matrix. In 24-hr reactions in nitric acid solutions mained almost intact (Table 1.7, experiments 8 and
at 300°C, complete combustion of the carbon and 9).
solubilization of the uranium and thorium were Additional studies of the oxidation of unfueled
achieved only when the O 2 / C and N 0 3 / ( T h + U) graphite were made in a reactor designed for the
mole ratios were greater than about 1 and 54, continuous addition and removal of g a s e s or
respectively (Table 1.7). The pressure of the solution (Fig. 1.3). However, s i n c e the volume
system at 300°C was 1500 to 2000 p s i . Com- of gas in the reactor was minimized to avoid long
bustion of the carbon apparently occurred accord- time l a g s , most of the t e s t s were made with batch
ing to the overall reaction: C + O j — C O j . Further charges of solution and continuous flows of g a s e s
confirmation of the stoichiometry was obtained only. The graphite, contained in a b a s k e t made
from gas a n a l y s e s , which showed mainly CO 2 and with screen, was loaded at room temperature
e x c e s s oxygen, and from a n a l y s e s of the solutions, through the flanged joint, nitric acid was added
which verified that little or no nitric acid was
converted to nitrogen oxides when sufficient
^^HRP Quart. Progr. Rept. July 31, 1960, ORNL-3004
oxygen was present. The need for maintaining a (October 1960).

ORNL-DWG 6 4 - 7 2 3 6 4

- t X h — S A M P L E S OF







Fig. 1.3. Schematic Arrangement of the Pressurized Aqueous Combustion Reactor System for Processing Graph-

ite Fuels at 3 0 0 ° C and 2 0 0 0 p s i .

with t h e solid or by a high-pressure metering ature, and surface area of the graphite. T h e rates
pump, and oxygen or gas mixtures were metered for 2 M HNO^ were about twice those for 1 M
into t h e reactor from gas cylinders. The gas, HNO3, and the rates at 300°C were about 1.5 times
which bubbled up through the solution, provided those at 275°C. As the surface area of the graph-
the necessary agitation. T h e removal of gas at ite w a s increased, the reaction rate increased,
room temperature was controlled manually by a but not proportionally. In two t e s t s , the reaction
needle valve. The CO 2 content of the effluent rates were proportional to the external surface area
gas was monitored continuously by a thermal- to the 0.4 and the 0.7 power. The degree of agita-
conductivity analyzer and was also determined tion also seemed to have a pronounced effect. It
by a n a l y s e s of gas samples. Twenty-one runs was not possible to vary the agitation in this system
were made with i^- and /j-in. cubes or V i n . - without varying one or more of the other variables;
square s t i c k s of moderator-grade graphite. T h e however, several different effects could be ex-
results were analyzed in terms of reaction r a t e s : plained if agitation i s important. For the V- by
grams of graphite oxidized to CO2 per hour per '/-in. graphite c u b e s , the reaction rate constants
gram of graphite present. ( h r ~ ' ) , assuming t h e reaction to be first order,
In t h e s e t e s t s , the oxidation rates increased were practically constant, regardless of the number
with increasing nitric acid concentration, temper- of cubes present or of whether the cubes were d i s -

tributed over a larger volume by mixing them with tory s c a l e ' ^ and on an engineering s c a l e ' * h a s
s t a i n l e s s s t e e l Raschig rings. The rate constants been completed. A conceptual plant s t u d y " based
calculated for moderator-grade graphite are: on the development work has also been completed.
B e c a u s e this work i s classified, it cannot be
Rate Constant (hr~ ) reviewed here. Unlike l a s t year, when a supple-
Temperature ( C) ment to the annual report covering the Rover work
IMHNO3 2AfHN03 4AfHN03
was i s s u e d , ^ " no supplement will be i s s u e d this
275 0.06 0.10-0.15 0.19-0.25 year s i n c e the more recent work is fully covered
300 0.08-0.12 0.15-0.24 in the topical reports noted above.

Very high rates of graphite oxidation were ob-

tained in a run with an HTGR fuel specimen. The 1.3 STUDIES ON THE PROCESSING OF
high gas flow rate, the increased agitation, the URANIUM-PLUTONIUM OXIDE FUEL
friability of the specimen, and the presence of
the uranium and thorium dicarbides in the compact
may have all contributed to the higher rate. The B a s i c chemical and hot-cell s t u d i e s on the
consumption of nitric acid was greatly decreased processing of pressurized-water-reactor UO2 and
by starting the oxygen flow when the reactor fast-reactor UO2-PUO2 fuels were continued.
temperature reached 180 to 200° C and continuing A new decladding process for s t a i n l e s s s t e e l
the flow until the reactor had cooled to below fuels, in which the cladding i s destroyed with
240°. In previous runs, where the flow of oxygen H F - O j at 600°C, was developed, and leaching
was controlled to achieve operation at constant of the treated fuel in nitric acid was demonstrated
temperature, the effluent nitric acid concentrations with unirradiated fuel samples. Since the resulting
were often l e s s than 25% of the charge concen- fuel solutions contained e x c e s s i v e amounts of
trations. This was probably due to a deficiency fluorides, a preleaching pyrohydrolysis step was
of oxygen during heatup and cooldown without developed. It eliminated 85% of the fluoride a s
oxygen flow. volatile H F . Also, dissolution experiments with
Reagents other than nitric acid were briefly nearly theoretically dense PuO 2 microspheres
tested in the autoclave with HTGR fuel compacts demonstrated that they dissolve very slowly in
at 300°C. Water, NaOH solutions, 2 M N a N 0 3 , HNO3-HF; hot-cell t e s t s will be run to determine
and H2S0^-KMn04 were tried, but none was a s whether irradiated material d i s s o l v e s faster.
effective a s dilute nitric acid. Sulfex process hot-cell t e s t s on s t a i n l e s s - s t e e l -
Preliminary autoclave experiments were a l s o clad UO2-20% PuO2 fuel samples irradiated to
conducted on the dissolving of sol-gel ThOj-UOj 99,000 Mwd per metric ton of U + Pu at cladding
microspheres in pressurized nitric acid solutions. temperatures up to 700°C resulted in high (1.5%)
Complete dissolution of high-density, 300-ff-diam uranium and plutonium l o s s e s and high (60%)
97% ThOj—3% UOj ^ microspheres was achieved r e l e a s e of ' ^ ^ C s to the decladding solution.
in 24 hr in 2 A/ HNO 3 at 300°C when the N O 3 / P a r a l l e l t e s t s with the shear-leach p r o c e s s showed
(Th + U) mole ratio was about 100. Pure T h 0 2 it to be the superior processing method for this
microspheres (260 to 300 11 in diameter) were type of fuel.
much l e s s reactive. In 24-hr reactions with 2 M
HNO3 at 300°C, the amounts dissolved were 17,
20, 35, and 49% when the N O j / T h mole ratios
were 10, 30, 50, and 100 respectively. L. M, F e r r i s , A Burn-L,each Process for Recovery
of Uranium from Rover Fuel: Terminal Report of Lab-
oratory Development, ORNL-3763 (March 1965).
R. W. Horton, summary of engineering-scale d e v e l -
opment on p r o c e s s e s for r e p r o c e s s i n g of Rover fuels
Development of Processes for Rover Fuel will be i s s u e d a s a topical report (confidential).
E . L. Nicholson, private communication.
Development of a burn-leach and a burn-volatility Rover Fuel Reprocessing Development, June 1,
p r o c e s s for graphite-base Rover fuel on a labora- 1963, to July 1, 1964, ORNL-CF-64-12-5 (confidential).

Leaching of U O j from Stainless-Steel-Clad fluoride), and 69% AI2O3. Uranium recoveries

Fuel Disintegrated by H F - O j better than 99.6% were obtained by leaching for
5 hr with boiling 1 to 15 W HNO3 (Table 1.8).
Studies of the leaching of mixtures of alumina, To enhance the leaching of fluorides from the
U30g, and " s t a i n l e s s s t e e l " oxides, resulting bed and to partially complex the fluoride, A1(N03)3
from the decladding of s t a i n l e s s - s t e e l - c l a d reactor was added to the leachant; but in concentrations
fuel with H F - O j , were begun. The leachant was up to 1 M, it had no effect on the recoveries.
a solution of HNO3 and A1(N03)3 in various
proportions. In this new p r o c e s s , the cladding The fluoride must be complexed to inhibit
on s t a i n l e s s - s t e e l - c l a d UO2 fuels can be d i s - corrosion of s t a i n l e s s steel process v e s s e l s .
If the bed product used in t h e s e studies can be
integrated and oxidized by reaction with HF-O2
considered typical, the fluoride concentration in
mixtures (see also Sect. 2.4, fluoride volatility
the product solution will be about three times the
p r o c e s s e s ) . Reaction at 600 to 650°C, preferably
uranium concentration; in t h e s e experiments, the
in a fluidized bed of inert alumina, results (1)
product solutions were about 0.2 Af in uranium and
in the disintegration and conversion of the stain-
0.6 M in fluoride.
l e s s s t e e l cladding to oxides and (2) in the
oxidation of the UO2 to UjOg powder. The Generally, 80 to 90% of the fluoride was leached
product always contains some fluorine, present from the bed irrespective of the leachant used
a s U O j F j and " s t a i n l e s s s t e e l " fluorides. (Table 1.8). The amount of iron dissolved in-
The bed material used in most of the experi- creased from about 16 to 85% a s the nitric acid
ments was produced by contacting simulated concentration increased from 1 to 15 A/ (Table
Yankee fuel (UO2 pellets clad in type 348 stain- 1.8). If the alumina fluidizing medium i s not to
l e s s s t e e l tubes) with 40% H F - 6 0 % O2 at 600°C be reused, leaching with dilute acid would prob-
in a fluidized bed of Norton RR alumina. (We are ably be desirable to avoid large amounts of iron
indebted to J . J . Reilly and C. B. Eartlett of in the raffinates from subsequent solvent extrac-
Brookhaven National Laboratory for supplying tion. The leaching of nickel and chromium h a s
this material.) It was composed of 25% UjOg, not yet been studied. In all c a s e s , l e s s than
13% s t a i n l e s s s t e e l oxides, 5% fluorine (as 2% of the alumina was dissolved.

Table 1.8. Leaching of a Product Bed Resulting from the Disintegration of Stainless-Steel-Clad UO2 by H F - O 2

Composition of bed: 25% U 3 0 g , 13% " s t a i n l e s s s t e e l " oxides, 69% AI2O3, 5% F

Samples were leached 5 hr with boiling leachant; final solutions were 0.2 Af in U

Leachant CompcDsition (M) Amounts Diss olved (%)

Experiment No.
HNO 3 A1(N03)3 U Fe Al F

1 1 0 99.9 16 0.9 84
2 2 0 99.9 24 1.5 84
3 2 0.5 99.7 24 84
4 2 1.0 99.8 19 89
5 4 0 99.7 27 1.2 84
6 4 0.5 99.7 31 90
7 4 1.0 99.6 16 88
8 10 0 99.9 60 1.5 99
9 10 0 99.9 68 1.8 86
10 10 0.25 99.9 50 80
11 10 0.5 99.7 55 79
12 15 0 99.8 85 1.0

Pyrohydrolysis of the declad fuel prior to leach- the bed prior to leaching with dilute acid would
ing with nitric acid considerably reduced the not only minimize corrosion in the leacher and
fluoride content of the bed. In 3-hr reactions other process v e s s e l s but also would allow nearly
with steam, the fluorine content of the bed was quantitative uranium recovery without dissolving
reduced from about 5 to 0.07% a s the steam temper- much of the s t a i n l e s s s t e e l oxides.
ature was increased from 200 to 600°C (Table
1.9). A sample of bed contacted with steam for
Sulfex-Process Tests with Irradiated
3 hr at 600°C was leached for 5 hr with boiling
10 M HNO 3. The uranium recovery was greater 20% P u O 2 - 8 0 % UO2 Fuel
than 99.9%. The product solution was 0.2 M in
uranium and only 0.06 M in fluoride. About 86% Prototype fast-reactor, s t a i n l e s s - s t e e l - c l a d fuel
of the fluorine had been removed by pyrohydrolysis. s a m p l e s , containing either swaged or pelleted
Also, only about 40% of the iron was leached, 20% PuO 2-80% UO2, were made and then irradi-
showing that pyrohydrolysis d e c r e a s e s the reac- ated at cladding temperatures up to 700°C to
tivity of iron somewhat (compare this value with bumups ranging from 5000 to 99,000 Mwd/metric
the data in column 5 of Table 1.8). ton. This phase of the work was done by the
Comparable results were obtained by heating General Electric Company at Vallecitos Atomic
a bed sample for 5 hr at 600°C in a stream of Laboratory. T h e s e specimens were subsequently
97% 0 2 - 3 % HjO (produced by bubbling oxygen used in hot-cell s t u d i e s at O R N L ^ ' to determine
through water at 25°C) and leaching the residual the feasibility of processing this fuel by the
s o l i d s for 5 hr with boiling 10 M HNO3. The Sulfex p r o c e s s , in which the cladding is dissolved
uranium recovery was 99.7%, and the product in H2S0^ followed by dissolution of the P u O j -
solution was 0.2 M in uranium and 0.05 M in U O j core in HNO 3 or HNO3-HF. Similar t e s t s to
fluoride. About 82% of the fluorine had been evaluate the shear-leach process for t h e s e fuels
removed by pyrohydrolysis. During leaching, are reviewed in Sect. 1.3.3.
sufficient alumina (2%) was dissolved to make In hot-cell t e s t s of the adaptability of the Sulfex
the product solution 0.05 M in aluminum. (Ade- process to this type of fuel, the s t a i n l e s s s t e e l
quate complexing of the fluoride is achieved with cladding was removed from the fuel specimens
an A l / F mole ratio of 1.) In a similar experiment, by refluxing in boiling 6 M HjSO^ for 4 hr at
the residue was leached with 2 M HNO 3 after 105°C; the s t a i n l e s s s t e e l was not p a s s i v a t e d .
pyrohydrolysis. The uranium recovery was 99.8%; L o s s e s of uranium and plutonium to the decladding
about 85% of the fluoride was removed, and, solution were e x c e s s i v e l y high in all runs and
significantly, only 6% of the iron had dissolved. averaged 1.55%. Nearly 60% of the fission product
The results of t h e s e preliminary experiments cesium from each sample also appeared in the
suggest that removal of most of the fluorine from decladding solution, something not previously
encountered in Sulfex process t e s t s .
The presence of s o much cesium in the cladding
solution is attributed to the high temperature of
Table 1.9. Fluorine Removal by Pyrohydrolysis from a the fuel samples during irradiation. The cladding
Product Bed from the H F - 0 - Disintegration of temperatures were 700°C, and the fuel center-line
Stainless-Steel-Clad UO, Fuel temperatures were considerably higher, about
2000°C. At this temperature, much of the '^^Xe
Reagent, steam; reaction time, 3 hr that forms in the PUO2-UO2 can e a s i l y migrate
to the voids in the fuel, where it decays to ' ^ ' C s
Temperature Fluorine Content of (ref, 22) and is ultimately dissolved during chem-
(°C) Residual Solids (%)

Original sample 5.3

J . H. Goode, Hot-Cell Dissolution of Highly-Irradi-
5.6 ated 2OP/0 Pu02-80fYo UO^ Fast Reactor Fuel Speci-
200 mens, ORNL-3754 (in preparation).
400 3.4 ^^B. F . Rider and C. P . Ruiz, "Determination of
Atom Percent F i s s i o n in Uranium F u e l , " Progr. Nucl.
600 0.07 Energy, Ser. IX 3(1—3), pp. 25—60, Pergamon, New
York, 1962.

ical decladding. This phenomenon h a s not been the dissolvent. The lower rate for sheared fuel
observed with other fuel samples since they have (Fig. 1.4) i s due to the much smaller area of oxide
been irradiated in water-cooled reactors, where exposed to the dissolvent.
temperatures are much lower.
Leaching of the sample irradiated to 99,000
Sulfex processing of this type of fuel does not Mwd/metric ton with 5 M HNO3 for 3 hr produced
appear economically attractive b e c a u s e the ex- a solution 0.52 M in U02(N03)2 and 0.12 M in
c e s s i v e l o s s of plutonium and uranium to the Pu(N03)^; it contained 3.60 mg of fission products
decladding solution will force the u s e of additional per milliliter. Although the bulk of the fission
processing s t e p s to recover the fissionable mate- products was dissolved in the leaching acid,
rial. In addition, the presence of over half of an insoluble residue of uranium, plutonium, and
the long-lived '•'^Cs in the decladding solution fission-product molybdenum and zirconium oxides,
would introduce additional problems in the manage- equivalent to 2.2% of the fuel oxide, remained
ment of intermediate-radioactivity-level waste after leaching. Only 0.13% of the uranium and
b e c a u s e the cooling capacity of waste storage plutonium originally present in the fuel sample
tanks for this stream was based mainly on the was found in the residue. T h e s e results suggest,
presence of ^''Co. The additional decay heat however, that some form of feed clarification i s
due to cesium would require either greater dilution required prior to solvent extraction of this type
of the decladding waste and larger waste tanks of fuel.
or u s e of expensive water-cooled tanks.

In previous s t u d i e s ^ ' ' ~ ^ ' on the dissolving of

unirradiated, high-density UO2, P u O j , and U O j -
P u O j , it was found that mixed oxides, containing
a s much a s 35% P u O j , could be dissolved in
Shear-Leach Process Tests with Irradiated nitric acid without the aid of fluoride c a t a l y s t
20% PuO2-80% UO2 Fuel only when the PuO 2 was in solid solution in the
UOj. Recent studies with unirradiated, high-
Companion hot-cell dissolution s t u d i e s with density PuO 2 microspheres of nearly 100% theo-
20% PuO 2-80% UO2 fast-reactor fuel samples retical density indicate this material to be ex-
irradiated from 9000 to 99,000 Mwd/metric ton tremely difficult to dissolve. In t e s t s with
at cladding temperatures up to 700°C indicate refluxing 14 M HNO3-O.O5 M H F , the dissolution
that the shear-leach process may be successfully rate was only 0.007 mg m i n ~ ' cm~^; only 8.5%
applied to this fuel. ^^ Dissolution rates for of the P u O j dissolved in 37 hr. Hot-cell t e s t s
both swaged and pelleted oxides were determined with irradiated samples will be made to determine
by leaching sheared s t a i n l e s s - s t e e l - c l a d s p e c i - whether irradiation results in increased r a t e s .
mens and by dissolving Sulfex declad specimens
( s e e Sect. 1.3.1) in strong nitric acid (3.5 to
10 A/). Fluoride catalyst was not required in
most i n s t a n c e s . R. E. Blanco and C. D. Watson, " H e a d - E n d P r o c -
e s s e s for Solid F u e l s , " chap. 3, Reactor Handbook,
vol. II, Fuel Reprocessing, I n t e r s c i e n c e , New York,
Both swaged and pelleted oxides were dissolved 25
completely in l e s s than 5 hr at all nitric acid L. G. R u s s e l l , " P l u t o n i u m Ceramics a s Nuclear
F u e l s , " chap. 3-2, Progr. Nucl. Energy, Ser. IV 4,
concentrations greater than 3 M (Fig. 1.4). T h e Pergamon, New York, 1961.
dissolution rates were independent of the density ^^F. L. Culler and R. E. Blanco, " F e e d Preparation
for Aqueous P r o c e s s e s , " P / 1 9 3 0 , Proc. U.N. Intern.
of the oxide before irradiation (maximum, 96%), Conf. Peaceful Uses At. Energy, 2nd, Geneva, 1958
irradiation level, and original concentration of 17, 259 ff. (1958).
W. M. Coshin, Fast Oxide Breeder — Fuel Irradi-
ation Experiments, KAPL-1784 (August 1957).
A. L. Uriarte and R. H. Rainey, Dissolution of
^^J. H. Goode, Hot-Cell Dissolution of Highly-Irradi- High-Density UO ^, PuO ^, and UO^-PuO^ Pellets in
ated 2(f7o PuO^—80% l/Oj Fast Reactor Fuel Specimens, Inorganic Acids, ORNL-3695 (April 1965).
ORNL-3754 (in preparation). 29
J . M. Gerhart, GE-VAL, personal communication.

ORNL-DWG 6 5 - 7 8 2 5 A ton of Th + U. Both the adsorption and subsequent

elution with oxalic acid were s u c c e s s f u l , and
^ more than 95% of the dissolved ^^^Pa was recov-
/ / 1
"^ i
ered. Silica gel and zirconium phosphate were
/ a l s o t e s t e d a s adsorbents.
Dissolution Studies on Stainless-Steel-

N I1

Clad ThOj-UOj

Hot-cell head-end studies were made to further

compare and evaluate the adaptabilities of the
/ Sulfex and shear-leach p r o c e s s e s to the processing
/ of highly irradiated s t a i n l e s s - s t e e l - c l a d ThO —4%
/ UO2 f u e l s . ^ ° ' ^ ' T h e s e t e s t s were made with
i 1 sol-gel-prepared, pelleted, and arc-fused T h O j -
1 1
UO2 fuel samples irradiated in water-cooled
reactors up to 98,000 Mwd per metric ton of
Th + U. Data were also obtained on the nature
of undissolved refractory residues and on the
solubilities and rates of dissolution of the macro


1 2 3 4
TIME ( hr )

Effect of Head-End Treatment on Dissolu-

5 6 7
amounts of fission products present in the highly
irradiated samples.^^ Much of this work was done
in cooperation with General Electric s c i e n t i s t s at
Vallecitos Atomic Laboratory. ^^~^*
In the Sulfex t e s t s , sectioned fuel samples were
declad in boiling 6 M H2S0^. The uranium and
thorium l o s s to the decladding solution averaged
tion of Irradiated 80% U O 2 - 2 0 % P u O j in Boiling 3.5 to
0.33%, which is about 10 times a s great as l o s s e s
obtained in previous t e s t s with prototype Consol-
idated Edison pelleted-fuel samples irradiated
to 20,000 Mwd per metric ton of Th + U. The
sol-gel irradiation samples used in the current
L. M. F e r r i s and J . W. Ullmann, Dissolution of
The Sulfex and shear-leach p r o c e s s e s were Sol-Gel-Derived and Arc-Fused ThO^-UO^ Fuel Par-
compared in t e s t s with T h 0 2 - 4 % UOj fuel samples ticles in HNO -HF Solutions: Laboratory Studies,
that had been irradiated in a water-cooled reactor ORNL-TM-867 (May 19, 1964).
to bumups of 3000 to 98,000 Mwd per metric ton ^^M. E. Whatley et al., Unit Operations Section
of Th + U. Three types of samples were u s e d ; Monthly Progress Report for March 1964, ORNL-TM-887
(September 1964).
sol-gel, pelleted, and arc-fused. The shear-leach 32
process, even though fuel dissolution rates were J . H. Goode and J . R. Flanary, Dissolution of
Irradiated, Stainless-Steel-Clad ThO -UO^ in Fluoride-
somewhat lower, was superior s i n c e the Sulfex
Catalyzed Nitric Acid Solutions: Hot-Cell Studies on
process resulted in a thorium-uranium l o s s of Pelletized, Arc-Fused, and Sol-Gel-Derived Oxides,
0.3% to the decladding solution. In all t e s t s , ORNL-3725 (January 1965).
R. W. Darmitzel, Post-Irradiation Examination of
irradiated sol-gel T h 0 2 - U 0 2 dissolved much faster ORNL Fuel Cycle Capsules, GEAP-4397 (September
than unirradiated sol-gel oxides. 1963).
^'*D. T. Ikeuye et al., ORNL-RML Services Program
The adsorption of ^^^Pa from Th-U-Pa nitrate Progress Report for January 1964, GEAP-4472 (1964).
fuel solutions on unfired Vycor was demonstrated ^^D. T. Ikeuye et al., ORNL-RML Services Program
in hot-cell t e s t s with 26-day-cooled T h 0 2 - U 0 2 Progress Report for February 1964, GEAP-4500 (1964).
^*D. T. Ikeuye et al., ORNL-RML Services Program
fuel samples irradiated to 75,000 Mwd per metric Progress Report for July 1964, GEAP-4549 (1964),

t e s t s were originally made from T h 0 2 - U 0 2 that sol-gel fuel irradiated to 17,000 Mwd per ton
was 60% - 1 0 +16 mesh, 15% - 1 6 +200 mesh, and of Th + U, steady s t a t e was reached after four
25% —200 mesh. After irradiation and d e c l a d d i n g , 5-hr c y c l e s . The total heel was 6%, and the
the T h 0 2 - U 0 2 s i z e distribution averaged 96.7% thorium concentration in the solution was 1 M.
+10 mesh, 2.1% - 1 0 +20 mesh, and 1.2% - 2 0 The final heel was completely dissolved in two
mesh, which indicates that considerable sintering 5-hr heel-cleanout cycles with Thorex dissolvent.
and agglomeration of the fuel oxide occurred during Most of the fission products in the irradiated
irradiation. fuel samples were soluble in the dissolver solu-
The dissolution rates of the irradiated T h O j - tions; however, small amounts of Sb, Mo, Ru,
UO fuels, whether sol-gel derived, pelleted, or and Zr-Nb were found in the residues. These
arc-fused, were all nearly the same in boiling r e s i d u e s , which comprised about 0.75% by weight
13 M HNO3-O.O4 M H F , with or without 0.04 M of the initial T h 0 2 - U 0 2 , contained 0.06% of the
Al(NO ) . The Sulfex-declad specimens dissolved original T h 0 2 - U 0 2 , s t a i n l e s s s t e e l cladding
faster than the sheared samples, but only b e c a u s e fragments, corrosion products, and alumina and
larger oxide areas were available for reaction. silica (Fiberfrax insulation). To improve solvent
For example, Sulfex-declad sol-gel samples extraction, the residues should be removed from
irradiated to 17,000 Mwd/ton were 97% dissolved the feed solution.
in 7 hr and 99.8% dissolved in 24 hr. Previous
t e s t s with identical unirradiated material resulted
Studies on the Adsorption of Protactinium
in 73% dissolution in 7 hr and complete dissolu-
tion in 75 hr^° (Fig. 1.5). In all c a s e s , the d i s - The presence of 27.4-day ^^^Pa accounts for
solver solution (1 M Th(NO3)^-10 M HNO3) was a large fraction of the beta and gamma radioactivity
a satisfactory solvent extraction feed. Solutions a s s o c i a t e d with short-decayed thorium-uranium
obtained from irradiated fuel were 0.05 M in
reactor fuels. To reduce the radioactivity to
fission products.
acceptable levels for solvent extraction and to
B e c a u s e of the very long total-dissolution times ensure a more complete recovery of the ^^^U, such
found in earlier t e s t s with unirradiated fuel, a fuels are usually allowed to decay 150 to 180
cyclic dissolution procedure was d e v e l o p e d . ^ ° ' ^ ' days before processing. To avoid this long decay
With this method, steady-state conditions for period, an adsorption process for the separation
unirradiated fuel were achieved in two c y c l e s of of protactinium from thorium and uranium on
20 hr each, after which a 13% heel (13% of the unfired Vycor has been developed. ^^ Significantly,
total Th + U added in two charges) remained. The the recovered protactinium would also be a source
dissolver solution was 0.8 M in thorium. With of isotopically pure ^^^U.
During the past year, hot-cell experiments were
made to determine the effects of macro amounts
ORNL-DWG 65-10006A
of fission products and high levels of radioactivity
100 on the p r o c e s s . Additional studies were made
80 to determine the protactinium distribution coef-
- 60 ficients between thorium—nitric acid solutions
and s i l i c a gel or zirconium phosphate (Bio-Rad
3 ZP-1).
o o ARC-FUSED 3 0 0 0 Mwd/METRIC TON
» SOL-GEL 4 0 0 0 Mwd/METRIC TON
Adsorption of Protactinium on Unfired Vycor. —
• SOL-GEL 16,000 Mwd/METRIC TON Hot-cell experiments showed that protactinium
20 1 SOL-GEL 17,000 Mwd/METRIC TON can be recovered from short-decayed thorium-
uranium fuel solutions by adsorption on unfired
Vycor g l a s s . ^® Thorium-uranium fuel specimens
0 1 2 3 4 5 6 7 8
^^C/iem. Technol. Div. Ann. Progr. Rept. May 31,
1964, ORNL-3627, pp. 1 3 - 1 6 .
Fig. 1.5. E f f e c t of I r r a d i a t i o n on D i s s o l u t i o n of Sol-
J . G. Moore et al., Adsorption of Protactinium on
Gel-Prepored ThOj-UOj in 13 M HNO3-O.O4 M KF-0.1
Unfired Vycor: Initial Hot-Cell Experiments, ORNL-
M AI(N03)3. 3773 (April 1965).

irradiated to 75,000 Mwd/metric ton in an average had been adsorbed, corresponding to about 0.5
neutron flux of 2 x 10^* and allowed to decay mg of ^^^Pa per gram of unfired Vycor. About
26 days were dissolved in refluxing 13 M HNO^— 0.1% of the ^^^Pa was lost when the column was
0.05 M H F . More than 99% of the fuel dissolved, washed with 6 column volumes of 10 M HNO3—0.1
leaving a residue containing l e s s than 0.1% of M A I C N O J ) . Elution with 8 column volumes of
the uranium and up to 20% of the protactinium. In 0.5 M oxalic acid removed 97.7% of the adsorbed
p r o c e s s application, the protactinium l o s s e s could protactinium. About 90% of the original ^^^Pa
be minimized by cyclic dissolution. In each of was recovered at a concentration 5.5 times that
three experiments, more than 99.4% of the d i s - in the feed. The protactinium had been separated
solved protactinium was adsorbed by p a s s i n g the from zirconium-niobium, total rare-earth beta,
dissolver solution through a bed of 60- to 80-mesh and ruthenium by decontamination factors of 9,
unfired Vycor, 0.9 cm in diameter and 17 cm long. 6030, and 2870 respectively. The protactinium
The flow rate in the first experiment was 0.3 concentration profile of the effluent and eluate
ml/min but was increased to 1 ml/min for the (Fig. 1.6) was similar to that previously reported
second and third runs. The feed to the third for 2 3 i p a s o l u t i o n s . ^ '
experiment was made 0.1 M in A1(N03)3 to complex Although there were several i n c o n s i s t e n c i e s
the fluoride in the feed. noted in the protactinium a n a l y s e s , the results of
The high degree of radioactivity of the feed had the hot-cell experiments showed the feasibility
little effect on the adsorption p r o c e s s . In the of recovering protactinium from short-decayed,
third run, the feed contained Th, 69 mg/ml; U, highly irradiated thorium-uranium fuel by adsorbing
1.80 mg/ml; and ^^^Pa, 0.032 mg/ml. It was it on unfired Vycor. Additional hot-cell experi-
10.1 M in HNO3, 0.05 M in H F , and 0.1 M in ments are planned for specimens irradiated to
A K N O , ) , , and each ml contained gamma emitters
equivalent to 1.7 x 10^* counts/min. Based on ^^Chem. Technol. Div. Ann. Progr. Rept. May 31,
a n a l y s e s of the effluent, 99.6% of the protactinium 1963, ORNL-3452, pp. 1 3 - 1 4 .

ORNL-DWG 6 5 - < 2 ( 4

' 2
__^. — — •

« 5
T3 FE ED: (0.( M HNO3, 0.05 Af HF, 0.( M Al (NOjjj ,
69 mg/ml Th, (.80 mg/ml U
fe 2 SH-. (0 M HNO3, 0.( M t 1 ( ^ 0 3 ) 3
UJ •
OWRATE: 0.97 ml/min
z 5

tto'° _ , 2 33 D „ «v \
u ><*>Lc r*
^ 5 yT
_,„^ . • « \
/ \ \
\DS0RP1 3H-»|-.

20 40 60 80 (00 (20 (40 (60 (80 200 220


Fig. 1.6. Concentration of Pa in Column Effluents During Adsorption on Unfired Vycor, Washing, and Desorption.

5000 Mwd per metric ton of Th + U to determine Th per liter; 500 to 4200 for a concentration of 50
the upper capacity limit of unfired Vycor for g/liter; and 400 to 3400 for a concentration of 100
protactinium adsorption, the maximum concen- g/liter. Although the distribution coefficients are
tration factor obtainable with oxalate elution, higher for Bio-Rad Z P - 1 than for unfired Vycor,
and the effect of aluminum on the solubility of preliminary experiments showed that Z P - 1 i s
protactinium during the dissolution of thorium- attacked by 0.5 M oxalic acid; therefore, other
uranium fuels in 13 M HNO3-O.O5 M HF. agents must be found for eluting the protactinium
from this material.
Adsorption of Protactinium on Other Inorganic
Exchangers. - As in earlier i n v e s t i g a t i o n s , ^ "
the results of adsorption experiments with s i l i c a
gel and Bio-Rad Z P - 1 (a zirconium phosphate 1.5 CHLORIDE VOLATILITY STUDIES
exchanger) continue to show a good potential for WITH URANIUM AND PLUTONIUM
separating protactinium from irradiated thorium.
The adsorption properties of s i l i c a gel and Chlorination is being studied a s a volatility
Bio-Rad Z P - 1 were compared by contacting them method for recovering uranium and plutonium from
with 0.5 to 12 M HNO3 solutions containing 4.50 UO^-PuO, fuels. Quantitative recovery of the
and 100 g of Th per liter and ^^^Pa equivalent to uranium and plutonium and decontamination from
1 to 5 X 10* counts min~ ^ ml""'. Except for fission products by a factor of at l e a s t 50 to 100
slight minima at 3 to 5 M HNO3, the distribution is required so that the product chlorides can be
coefficients obtained with silica gel increased a s reconverted to oxide fuel, possibly by a sol-gel
the nitric acid concentration increased or a s the method. Direct chlorination with 85% 0 ^ - 1 5 %
thorium concentration decreased. The coefficients CCl_j of U O j - P u O j fuel pellets containing up to
ranged from 2100 to 5800 for solutions containing 35% PUO2 was studied briefly. In all c a s e s , the
4 g of Th per liter; from about 1900 to 4000 for a rate of reaction was too low to be of practical
Th concentration of 50 g/liter; and from 1450 to use in the temperature range of interest, 400
2380 for a Th concentration of 100 g/liter. The to 600° C.
amount of unadsorbable protactinium ranged from Prior work with UOj f u e l ' " showed that rapid
0.6 to 1.9% and was independent of the nitric chlorination could be achieved after converting
acid or thorium concentration. With Bio-Rad Z P - 1 , UOj pellets to U30g powder with oxygen at
the coefficients ranged from 100 to 29,000 for 500 to 800°C. Oxidation not only results in the
thorium solutions (4 g/liter) in which the molarity conversion of the uranium to a higher valence
of the HNO3 ranged from 0.5 to 12 M. Although s t a t e but a l s o yields a product of high specific
45% of the protactinium was unadsorbable from surface area. Similar oxidation experiments
0.5 M HNO3, l e s s than 0.4% was unadsorbable showed that UO2-PUO2 fuel pellets generally
from solutions that were ^ 3 W in HNO3. With could be converted to U3O.-PUO2 powder when
solutions containing 50 and 100 g of Th per liter the P u O j content of the pellet was 17% or l e s s
and 0.5 to 12 M in HNO3, the coefficients ranged (Table 1.10). However, pellets containing 20%
from 400 to 19,000 and 200 to 8000 respectively. or more P u O j were generally unaffected by oxygen,
At both thorium concentrations, more than 95% of even after long reaction periods at temperatures
the protactinium was unadsorbable from solutions up to 800°C. A more critical evaluation of the
1 M in HNO3. With 3 M HNO3, 3% was unadsorb- oxidation behavior of UO2-PUO2 pellets cannot
able when the Th concentration was 50 g/liter, be made s i n c e , in general, the pellets used had
and 15% when the concentration was 100 g/liter. not been carefully characterized. However, t h e s e
Nearly all the protactinium was adsorbable at preliminary data do indicate that an oxidation-
higher nitric acid concentrations. chlorination process for fuel containing more than
To complete the comparison, approximate d i s - 20% PuO^ may not be feasible.
tribution coefficients obtained for unfired Vycor Chlorination, with several reagents, of the
are: 600 to 5000 for a solution containing 5 g of U j O . - P u O , powders resulting from the reaction

^ Chem. Technol. Div. Ann. Progr. Rept. May 31, *^Chem. Technol. Div. Ann. Progr. Rept. May 31,
1964, ORNL-3627, p. 16. 1964, ORNL-3627, pp. 1 7 - 1 9 .

of U O j - P u O j pellets with oxygen was studied chlorides was achieved at 600°C with an e x c e s s
in an effort to find the best chlorinating agent. of chlorine. Calculations^^ based on equilibrium
Of the reagents t e s t e d , which included chlorine, data for the reaction'*^
CCl - N , mixtures, phosgene, and CO-CCl^ mix-
tures, a mixture of 85% 0 1 ^ - 1 5 % CCl^ (produced PuCl- 3( s_) ,+ ,V, C. L. (, ^^) _ PuCl/v)
by bubbling chlorine through liquid CCl^ at 25°C)
gave the best r e s u l t s . Pure UjOg was readily showed that about 3.3 x 1 0 ' and 100 moles of
converted to volatile uranium chlorides at 400 chlorine are required to transport 1 g-atom of
to 500° C (Table 1.11). With a 1% P u 0 2 - 9 9 % PuCl at 427 and 727°C respectively. Thus,
U30g powder, the rates of chlorination were too
low to be practical at 300 to 350°C. At 400 to
500° C, nearly complete volatilization of the T. A. Gens, Thermodynamic Calculations Relating
uranium was achieved in 3 - to 4-hr reactions, but to Chloride Volatility Processing of Nuclear Fuels. II.
The Capacity of Chlorine for Transporting Plutonium
only up to 50% of the plutonium volatilized. Tetrachloride Vapor During Reaction of U O -PuO
Complete volatilization from the hot zone of the with Carbon Tetrachloride, ORNL-3693 (October 1964).
reaction v e s s e l of both uranium and plutonium ''^R. Benz, / . Inorg. Nucl. Chem. 24, 1191 (1962).

T a b l e 1.10. Reactivity of U 0 2 - P u 0 2 P e l l e t s in Oxygen at High Temperatures

PuO 2 Content Oxidation Reaction Time Oxidation of

Experiment No. of P e l l e t (%) Temperature (*" C) (hr) UO2 t ° U3O8

1 0 400 0.5 Yes

2 0 600 3.0 Yes
3 1 600 1.0 Yes
4 1 600 3.0 Yes
5 5 600 1.0 Yes
6 17 600 3.0 Yes
7 17.8 600 1.0 Yes
8 20 600 19.0 No
9 20 800 20.0 No
10 20* 600 3.0 Yes
11 35 800 18.0 No

° T h i s pellet was fired to only 1300°C during manufacture; the others were fired at 1600 C.

T a b l e 1.11. Reaction of U 3 0 g - P u 0 2 Powders with 85% C l 2 - 1 5 % C C I 4 at Various Temperatures

Amount Volatilized (%). ^°^^^ °^ ^h "^^'^

Experiment No ^"'-'2 Content of Temperature Reaction Time per Gram-Atom of
Mixture (%) (°C) (hr) Pu Pu Volatilized

0 400 5.5 100

0 500 3.5 100
300 5.0 14 5 3X 10"
350 2.75 35 2 6 X 10*
400 4.0 99 23 8x lO''
500 3.0 93 49 1 x 10''
600 3.0 59 29 3x 10'
600 6.0 100 100 7 x 10''

failure to achieve complete volatilization of carried out over the past several years on the
plutonium chloride at the lower temperatures may process.
have been due to a deficiency of chlorine.
The optimum temperature for the chlorination of
U j O . - P u O , powders has not yet been determined Fuel Disassembly Tests
but will probably be l e s s than 500°C, where large
amounts of chlorine are required to transport Since some second-generation power-reactor
plutonium chloride. During chlorination at 500° C, fuels cannot be sheared satisfactorily without
definite evidence for sintering and some indication first being disassembled, mechanical equipment
of melting were obtained. At 600°C, formation (Fig. 1.9) consisting of a hydraulic cylinder,
of a liquid phase (possibly UO^Clj or a similar gripper head, bumper, cradle, ejector, and e l e -
compound) was always noted during the chlo- vating j a c k s was built to demonstrate multitube
rination of powders containing up to 17% PuO^. withdrawal from the " e g g c r a t e " of the Consol-
idated Edison core B fuel assembly. After the
inert end adapters are removed from a fuel assem-
bly by sawing with an abrasive disk, the remaining
" e g g c r a t e , " containing the fuel rods, i s placed
in the cradle, which in turn is placed on the rollers
on the rod puller platform and clamped into posi-
Development of the shear-leach p r o c e s s with
tion. The fuel assembly i s then forced against the
unirradiated fuel assemblies on a full engineering
ejector plate by the bumper and hydraulic cylinder
s c a l e was continued. In this process for fuels
which pushes one row of 13 or 14 fuel tubes about
clad with s t a i n l e s s s t e e l or Zircaloy-2, the
3 in. through a slot in the bumper into the jaws of
multirod assemblies are sheared in a 250-ton
the gripper head. The tubes are finally clamped
shear, either with or without disassembly, and
with the gripper head and pulled from the parent
the sheared fuel is leached in nitric acid to
assembly by a hydraulic cylinder. The procedure
produce solvent extraction feed. Significant
is repeated until the assembly i s emptied of t u b e s .
developments d i s c u s s e d in this section are
After each row of tubes is withdrawn, the assembly
summarized in the next paragraph.
is raised with elevating screws to properly align
B e c a u s e second-generation power-reactor fuels
the next row with the ejector plate and the slot
may not always be satisfactorily sheared intact
in the bumper. The loose tubes are then loaded
in the 250-ton shear, a useful prototype fuel-rod
into an open-top envelope for shearing.
row puller was built. Shearing t e s t s on bundles
of loose rods and on an annular Zircaloy-2-clad In t e s t s with this equipment, a force of 332 to
uranium-metal fuel rod were satisfactory. Studies 420 lb was required to eject a row of 14 UO^-filled
on the remote handling and fission product heat tubes from the assembly into the gripper head,
generation in b a s k e t s of sheared fuel showed that and a force of 238 to 315 lb was then required to
no serious problems exist there. Head-end s t u d i e s withdraw them. It i s expected that the force
on the safety a s p e c t s of zirconium fines indicate required to displace fuel tubes may be slightly
that a fire or explosion hazard may exist in the higher after the assembly h a s been irradiated;
removal of assembly end boxes by sawing with roughened surfaces will develop b e c a u s e of the
abrasive d i s k s . Leaching t e s t s on metal-clad formation of oxides.
ThO - U O , were concluded. A delayed-neutron
activation analysis monitor for determining the
residual fuel content of leached hulls showed Shearing Tests
that l o s s e s in the range of 0.01% can be con-
sistently detected. Shearing t e s t s with the 250-ton prototype fuel
A pictorial flowsheet of the shear-leach p r o c e s s shear were made this year on both Zircaloy-2-clad
a s applied to second-generation fuels where fuel U O j multitube fuel assemblies and on coextruded
disassembly is required is shown in F i g . 1.7. An Zircaloy-2-clad uranium metal.
optimized layout of equipment for the same fuel A stacked 8 x 8 array of unsecured Zircaloy-2-
type i s shown in Fig. 1.8. This conceptual layout clad U O , fuel tubes (0.4395 in. in outer diameter,
is based largely on the results of the developments with a 0.32-mil wall) was sheared into 1-in.






Fig, 1.7. Flowsheet for Shear-Leach Process for Second-Generation Power-Reactor Fuels.
i I \ , i I ^ ^.~i 1 k,-M(


ORNL-OWG 6 5 - 9 7 0 2



Fig. 1,8. Optimized Layout of Equipment for the Sheor-Leach Process.


ORNL-DWG 64-7235R












Fig. 1,9. Equipment for Withdrawing Rows of Fuel Rods from Second-Generation Power-Reactor Fuel A s s e m b l i e s .

lengths. It was observed that there w a s l e s s of the exposed uranium. Some sparking occurred
end closure of the brittle zirconium than with the during shearing, but there appeared to be no real
more ductile s t a i n l e s s s t e e l , which results in hazard involved in shearing this type of fuel.
the dislodgement of considerably more fuel. Some
sparking, about one to three sparks per cut, w a s
observed. The shearing force required w a s from Shear Evaluation Tests
50 to 60 tons. From t h e s e t e s t s , it appears that
zirconium-clad fuels will b e e a s i e r to shear than A rack-and-pinion feeding mechanism was
s t a i n l e s s - s t e e l - c l a d fuels. installed and tested on the 250-ton prototype
Fuel assemblies consisting of 3-in.-diam shear, replacing the hydraulic system. Replace-
coextruded Zircaloy-2-clad uranium metal were ment w a s made so that the length to b e sheared
successfully sheared into \ -, 1-, 1 V-, and 2-in. from a fuel assembly could be controlled by the
lengths. One fuel assembly w a s sheared at a external feed mechanism itself rather than by an
time; the fuel w a s " g a g g e d " or clamped during internal stop inside the shear, which experience
shearing, using semicircular i n s e r t s in the fixed had shown to be difficult to repair. T h e new
blade holder and gags. There was no discernible mechanical device operated satisfactorily in
difference in the power required or appearance of shearing t e s t s with various types of fuel-assembly
the sheared s e c t i o n s when using a straight or prototypes.
tapered semicircular moving blade. The shearing Inspection of the shear after 10,100 c u t s revealed
force varied from 45 to 82 tons and was independ- an appreciable increase in the galling of the gibs
ent of length sheared. About 30% of the c r o s s (type 420 s t a i n l e s s steel) a s compared with t h e
section of the fuel w a s sheared, and the remainder galling after 7600 c u t s . It i s estimated that the
was broken. There was only a small amount of shear h a s been stroked an additional 10,000 times
wiping of the Zircaloy-2 cladding over the face for various reasons, resulting in a total of nearly

20,000 strokes. Some of the galling is attributed The thermal conductivity k (Btu h r ' ^ f f ^ " F - ^
to misalignment in the shear; this has been cor- for sheared s t a i n l e s s - s t e e l - c l a d U O , i s being
rected. For the most part, the shear was operated determined experimentally to aid in the evaluation.
dry, but a lubricant, such as a molybdenum di- The measured heat transfer coefficient for d i s -
sulfide suspension in water, is desirable. The lodged UO fines in an air atmosphere, consisting
gibs were cleaned by grinding. Measurements of a mean particle s i z e of 120 fi and 38% voids,
indicated a hardness of 53 Rockwell C for the increased linearly from 0.256 Btu h r " * ff-^ o p " '
Stoody-I liners and 47 for the gibs. The liners at 200° F to 0.448 at 1600° F.
were only moderately galled. In initial calculations, a /: of 0.2 and 0.4 was
s e l e c t e d , and a typical second-generation UOj
fuel, having a specific power of 23.5 w / g and a
Remote-Handling Tests with burnup of 20,000 Mwd/ton, was chosen. The
Fuel Leaching Baskets cooling time was assumed to be 180 d a y s . It
was further assumed that a nominal 8-in.-diam
In the batch shear-leach process, a solid con- leaching basket would be filled to a depth of 6 ft
sumable liner of 10-mil carbon s t e e l i s inserted with 1-in. lengths of sheared fuel. The filled
in the perforated basket to confine the sheared b a s k e t s would be stored in an open array approxi-
fuel during shearing, storage, and transport. To mately as shown in F i g . 1.8 (item 11). Heat
avoid losing fuel, there must be a dust-tight s e a l transfer from the sealed b a s k e t s to the cell atmos-
between the liner and perforated basket. A mockup phere, at an ambient temperature of about 120° F ,
cell was built and successfully operated to demon- would occur by both radiation and natural con-
strate the feasibility of using a Magneforming vection.
machine capable of producing up to 4500 ft-lb of It was a l s o assumed that during the initial
energy to expand the liner into the groove. To 15-min contact in the acid, the carbon-steel liner
attain maximum efficiency with the Magneformer, and all dislodged UOj fines are dissolved, and
a highly conductive material was taped to the that all s e n s i b l e heat is transferred from the
inside of the liner in the area to be expanded. sheared fuel. In a leacher sized to dissolve 1
Aluminum tape with an adhesive backing was the metric ton of uranium, a maximum of about 55,000
most satisfactory. The cell-mockup operation kcal of heat would be released from a total of
a l s o demonstrated the use of a hinged Tube-Turn six b a s k e t s during this time. Assuming good
closure a s a means of sealing the top of the mixing of the leachant, a s little as 300 gal could
b a s k e t during transport; a Grayloc connector absorb this amount of energy in being heated
provided a satisfactory s e a l between the b a s k e t from 9 0 ° F to about 190°F. The mass of s t a i n l e s s
and leacher. In concluding t h e s e t e s t s , a short s t e e l in leachers of this s i z e represents an ad-
color movie was made to illustrate the simplicity ditional heat sink of about 30,000 kcal between
and feasibility of these remote manipulator oper- initial and final temperatures of 120 and 190° F
ations. respectively. Therefore, the amount of acid it
would be p o s s i b l e to flash during t h i s initial
period would probably not c a u s e any significant
Heat Transfer Tests with Fuel Leaching Baskets pressure rise. The results of this preliminary
calculation are summarized in Table 1.12.
Leaching b a s k e t s filled with sheared, highly A more exact calculation will be made pending
irradiated, spent reactor fuel will, b e c a u s e of completion of k measurements and evaluation of
heat generated by the fission products, reach the induced activity.
some s t e a d y - s t a t e temperature in storage while
awaiting transfer into the leaching equipment.
Estimates were made of the amount of s e n s i b l e Safety Studies on Finely Divided Zirconium
heat in a typical basket of s t a i n l e s s - s t e e l - c l a d
UOj fuel and of the possible consequences of a To evaluate the safety of sawing and shearing
sudden r e l e a s e of this heat from several such zirconium, fuel tubes filled with UO^ were sawed
b a s k e t s of fuel in contact with nitric acid in a and sheared; the fines are being metallurgically
large leaching v e s s e l . examined.

T a b l e 1.12. Estimated Heat Release from a Basket of Fully Irradiated Sheared, Stainless-Steel-Clad
UO2 During Initial 15-min Immersion in 6 M HNO3 at 7 2 ° F

Heat R e l e a s e (kcal)
Heat Source Weight Heat R e l e a s e
(kg) (kcal/kg) 0.2 k= 0.4

U O — s t a i n l e s s s t e e l fuel quenching 261" -4510° -3550^^

Decay heat 261 -1.46 -380 -380
Mild s t e e l dissolution'^ 2.22 -997 -2210'' -2210"

U O j fines dissolution 70.2 -31.1 -2180^ -2180^

Total heat r e l e a s e -9280 -8320

Weight of 1-in. sheared lengths of UO,-SS contained in a b a s k e t liner 7.75 in. in inner diameter and 6 ft high.
Uranium weight is about 175 kg.
Basket temperature profile i s for a volume heat source; Cp for 75% UO —25% s t a i n l e s s s t e e l = 0.072 Btu lb""
deg "" Center-line temperature of fuel in basket, 7 9 0 ° F ; basket liner temperature, 435°C; b a s k e t surface temper-
ature, 355'^F; ambient temperature, 120"F.
The iron and UO would not normally dissolve at 120°F, placing the overall calculation on the conservative
''Mild s t e e l liner i s 0.010 in. thick; dissolution reaction: Fe + 6HNOj — FeCNOj)^ + 3H2O + 3NO2.
' R e a c t i o n : UO^ + 4HNO3 "* UOjCNOj)^ + 2H2O + 2NO2.

Zircaloy-2 rods (^^ in. in outer diameter) were Table 1.13. Particle-Size Distribution of Z i r c a l o y - 2
cut with an abrasive disk saw with and without Fines Produced by Wet Sawing of /^-In.-diam Tubes
coolant water, with a medium-hard abrasive blade
(Manhattan No. 225) rotating at about 10,450 sur- P a r t i c l e Size P e ] rcentage of F i n e s
face feet per minute. The rods were fed through (M) (wt %)
the saw at 3 in./min. Dry cutting of the zirconium
presents a definite fire hazard, a s evidenced by <2000 99.04
the brilliant stream of sparks and dense smoke <840 97.7
produced. Therefore, wet cutting i s recommended, <420 95.3
especially where the abrasive sawing will be done <250 90.0
in the presence of uranium metal or other materials, <149 77.5
whether pyrophoric or combustible. Of the parti- <88 62.4
c l e s collected from the dry abrasive sawing, 79 <74 55.7
and 31 wt % of them were l e s s than 400 and 44 <44 38.7
fi in s i z e respectively. The volume ratio of abra-
s i v e blade consumed to zirconium particles pro-
duced was 7.3 to 1.
Sawing with a wet abrasive disk (three water
T e s t s by the Bureau of Mines'*'' have e s t a b l i s h e d
spray nozzles directed on the s a w blade) decreased
that dry zirconium powder (90% = 6 /z) can ignite
the spark stream to an occasional spark. The
explosively and spontaneously when dispersed
amount of zirconium fines from wet sawing i s
in air at room temperature or when heated in air
twice that produced in dry cutting b e c a u s e of the
to 190°C. Since 39% of the Zircaloy-2 fines from
burning during the dry cutting. With wet cutting,
wet sawing are smaller than 44 /i, it is assumed
the weight of Zircaloy-2 particles smaller than
that a considerable amount i s in the 6-/x range and
400 fi (the s i z e range where s p e c i a l handling i s
recommended in hot c e l l s ) was 94.5% of the total
produced; 39% of them were smaller than 44 fi 44
Zirconium Fire and Explosion Hazard Evaluation,
(Table 1.13). TID-536S (Aug. 7, 1956).

should be kept under water until discarded. Once Engineering-Scale Leaching Studies
the Zircaloy-2 powders are wet, they should be
kept wet b e c a u s e powders containing 16% or l e s s Batch leaching t e s t s with sheared, unirradiated
moisture are considered particularly hazardous. s t a i n l e s s - s t e e l - c l a d UO^ and U 0 2 - T h 0 2 fuels were
Special procedures must be followed in handling continued in an engineering-scale 90-liter Pyrex
the fine powder even when stored under water and s t a i n l e s s s t e e l leacher (Fig. 1.10) in which
b e c a u s e the powder s e t t l e s into a compact mass the leaching section i s 9 in. in inner diameter and
that can exclude much of the water. F i r e s have 10 ft tall. Heat is supplied by a steam j a c k e t with
started'*' when such packed powders were d i s - an inside surface area of 6.6 it^. In the leacher,
turbed. Since the Zircaloy-2 particles are highly boiling induces a counterclockwise circulation
diluted with the abrasive particles (volume ratio, through a side arm arranged to impinge dissolvent
1 to 6.6; weight ratio, 1 to 2.9), it i s probable on the bottom of the leacher; this produces the
that this dilution provides a builtin safety factor same general effect a s would be attained with an
in preventing spontaneous or static-spark reactions. air-sparge draft tube. Fuel charges in a slotted
fuel basket varied from 18 to 36 kg.
The literature on the relationship between fine- The consumable carbon-steel liner dissolved in
n e s s of the powder (and other factors) and fires about 2 min. As the liner dissolved, 14 to 38% of
or explosions shows the following: The dry pow- the core was discharged from the basket and s e t -
der i s ignitable at room temperature by s t a t i c tled to the bottom of the leacher.
electricity.'*' Powders containing 16% or l e s s T h e overall heat transfer coefficient, based on
moisture may explode spontaneously in a vac- the inside area of the jacket, is fairly constant
uum.'*'* The surface area of powders in square for both water and a solution containing 250 g of
meters per gram is within 10% of the reciprocal Th per liter at steam pressure of 20 to 60 psig.
of the average particle s i z e in microns. When For water, the coefficient varied from 164 to 180
the area per gram approaches 0.1 m^, careful Btu hr""^ ft~2 ° p - i ^ and for the thorium solution,
handling is necessary.'** Violent explosions, from 200 to 210.
which have resulted in fatal injuries, have all In one t e s t , in which UO^ was leached from
involved particles smaller than 62 /x (230 mesh).'*' 1-in.-long sheared sections of s t a i n l e s s s t e e l
Scrap finer than 20 mesh (800 fi) and coarser than cladding with boiling 7 M HNO3, total dissolution
120 mesh (125 /z) should be collected under water was attained in 2 hr. B e c a u s e UOj d i s s o l v e s
and mixed half and half with sand or grinding-wheel rapidly, variables affecting the rates were studied
grit.**' Powders that have a water content of more satisfactorily with slower-dissolving T h O , -
3 to 16% and that have been stored for a fairly UO^, using boiling (120°C) 12.7 M HNO3-O.O4
long time are the most hazardous.'** Crystal-bar M F ~ - 0 . 1 M Al^"^ a s the solvent. Typical UO^-
Zircaloy-2 irradiated to 540 Mwd showed no ap- ThO leaching data are presented in F i g . 1.11,
preciable amount of hydrogen pickup. The micro- which shows the effect of dissolution time on
hardness of Zircaloy-2 irradiated to 15.2 x 1 0 ' ^ terminal thorium loading, boilup rate a s repre-
nvt increased from 135 to 173 DPH a s measured sented by steam pressure, and a comparison of
at a 50-kg load.'*' v.- and 1-in. sheared lengths. The duration of all
t e s t s was sufficient to ensure that 99.9% of the
uranium and thorium would d i s s o l v e . With the
'*'National Safety Council Data Sheet No. 382, rev., exception of one run, all fuel was sheared to
1957. 1-in. lengths.
W. R. DeHoUander, An Evaluation of the Zirconium Leaching times for k - and 1-in. sheared lengths
Hazard, HW-44989 (Aug. 15, 1956).
47 were nearly equal (18 hr and 19 hr). Leaching
W. W. Allison, Zirconium, Zircaloy, and Hafnium times for equivalent thorium loadings in the solu-
Safe Practice Guide for Shipping, Storing, Handling,
Processing, and Scrap Disposal, WAPD-TM-17 (Rev.) tions were roughly doubled by reducing the steam
(December 1960). pressure from 60 to 20 psi. At the same steam
'**F. B. Holt and S. G. Wilson, A Code of Practice
for the Handling of Zirconium and Its Alloys, No. pressure, the time required to achieve a given
IGR-TD/S-578, United Kingdom AEC (July 1957). thorium loading was approximately linear between
M. L. Bleiberg and L, S. Castleman, Effects of 0.5 and 0.7 M Th but increased rapidly at higher
Neutron Bombardment upon the Properties of Zirconium loadings. Dissolution data obtained with a per-
and of a Zirconium-Tin Alloy, V/APD-78 (Mar. 17, 1953).

ORNL-DWG 6 4 - 4 i e 4 R 2

SHEARED FUEL 1 8 - 3 6 kg OF UOa


SLOTTED BASKET 0 015 x 0 3 4 3 in

HEIGHT I 6 V 4 in
WALL 0 0 2 5 in

' V i 6 - n - D I A M HOLES 6 3 % FREE AREA
HEIGHT l e ' / ? in
WALL 5/(6 in

3-in - I D



Fig. 1,10. Engineering-Scale 9 0 - l i t e r Batch Leacher for D i s s o l v i n g Oxide Fuel from Sheared Cladding

forated b a s k e t containing '/i-in. holes on k-in. t e s t i n g the leached hulls. Several methods for
centers (4.2% free area) were nearly the same as such monitoring were evaluated, and a prototype
the data for the slotted basket shown in F i g . 1.10. was developed and t e s t e d . It can detect about
This type of basket discharges the same amount 1 mg of " 5 u ^ " 9 p y ^ Of 233u pgf kilogram of
of fines a s the slotted basket does during the leached hulls. This corresponds to a fuel l o s s
dissolution of the liner. of l e s s than 0.01% for typical low-enriched ura-
In producing a dissolver product approximately nium power-reactor fuels. The proposed monitor
1 A/ in Th or U when using a dissolvable 10-mil- i s based on delayed-neutron activation a n a l y s i s
thick carbon-steel basket liner, the resultant and is* applicable to either s t a i n l e s s s t e e l or
solution is nearly 0.1 M in F e (about 11 g of iron Zircaloy-2 hulls.
per kilogram of core). The prototype monitor, capable of testing 20
(150 cm^ or 200 g) leached hulls at a time, was
successfully demonstrated. A larger unit, capable
Monitoring of Leached Hulls 50
of monitoring a 7-in.-diam fuel basket in s u c c e s -
sive 4-in. vertical increments (2.5 liters or 2.7 kg
To be assured that the leaching of uranium,
of hulls) i s now being made. An illustration of
plutonium, or thorium from s t a i n l e s s s t e e l or
the smaller test unit is shown in Fig. 1.12, and
Zircaloy-2 cladding in the shear-leach process
of the larger unit in Fig. 1.13.
is complete, it would be very advantageous for
In the t e s t mockup, separate chambers made of
a fuel processor to use a reliable monitor for
paraffin blocks were used for irradiating and
counting. The chambers were separated to prevent
the boron in the B F ^ neutron detection tubes from
^"W. J . R o s s , J. W, Landry, and J . E. Strain, "Moni-
toring of Leached F u e l Elements with a Neutron being depleted by the neutron flux in the irradiation
G e n e r a t o r , " Nucleonics (August 1965). chamber. Operation consisted in placing 20 hulls

ORNL-DWG 64 - 7509

10 (2 H 16

Fig. 1.11. Effect of Thorium Loading, Dissolution Temperature,and Length of Sheared Fuel in the Batch L e a c h -
ing of Unirradiated U O - - T h O - from Sheared Stainless-Steel-Clad Power-Reactor Fuel Elements.

ORNL-DWG 6 5 - 9 7 0 3


10 X 12 X 28 in


8 X 12 X 2 8 in


Fig. 1.12. Mockup Monitor for T e s t i n g L e a c h e d Fuel Cladding for Residual Fuel

m the irradiation chamber (in the position marked B F j counting system was only a few tenths of
" 1 " in F i g . 1.12) and irradiating them with thermal a percent b e c a u s e the BF3 tubes were unmatched.
neutrons for about 20 s e c . Within the following B a t c h e s of 20 hulls containing weighed amounts
2 s e c , the floor of the chamber (the slide) w a s of ^^^U gave t h e results indicated in Table 1.14.
opened, and t h e activated hulls were dropped into It w a s found that a 7-in.-diam basket of leached
the counting chamber (position " 2 " in F i g . 1.12), hulls must b e monitored in vertical increments
where the delayed neutrons from the activated not exceeding 4 in. This specification w a s ar-
hulls were counted for the next 30 s e c . T h e aver- rived at in the following way: The thermal-neutron
age thermal-neutron flux during irradiation w a s flux in t h e irradiation chamber (6 cm in diameter
1.89 X 10*, and t h e average fast flux (14-Mev and 8 cm long) w a s mapped by manganese d e t e c -
neutrons) w a s 1.67 x 10*. The efficiency of the tors. T h e counts obtained from ^ ^ ' U in a hull

0RNL-DW6 66-9704






Fig. 1.13. Prototype P l a n t Monitor for T e s t i n g Leached Fuel Cladding for Residual F u e l .

can be determined within a variation of ±15% when Other methods which were evaluated but which
the chamber i s filled with hulls and when the were l e s s s e n s i t i v e or completely inadequate
position of the hull that contains the ^ s s y ^g included: (1) monitoring the residual 2.18-Mev
varied throughout the volume of the chamber. gamma rays from ^^''Ce-Pr either by gamma s p e c -
Axial variation was ±15%, and radial variation trometry or by detecting the neutrons generated
±5%. The sensitivity of the instrument decreased by t h e 'Be(y,n)*Be reaction; (2) monitoring the
a s the hull was moved along the axis away from low-energy gamma and x-ray emission from ^^*U
the center of the chamber. and ^ ^ ' P u ; (3) comparing the gamma-ray emissions
It IS expected that the larger monitor (Fig. 1.13) from I3 4,i3 7(-;g g ^ j eoQoj ^^^ (4) comparing the
will retain the 1- to 5-mg ^ss^j sensitivity ob- absorptions of two gamma-ray energies, that from
tained in t h e mockup t e s t and that it will operate a ' ^ ' C s source with that from a ^""Am source.
on about a 10-sec cycle. Table 1.15 shows the The instrument may also have general u s e a s
b a s e s for t h e s e conclusions. a w a s t e monitor. For example, b e s i d e s monitoring

T a b l e 1.14. Neutron-Counting Results Obtained in Prototype Leached-Hull Monitor

Number and Type of Hulls Counted Counts per Minute

20 empty h u l l s ; t h i s i s the background 6

19 h u l l s empty and 1 with 1 mg of ^^^U 16
19 h u l l s empty and 1 with 5 mg of ^^^U 47
18 hulls empty, 1 hull with 1 mg of ^^^U,
and 1 hull with 5 mg of ^^^U 49
19 hulls empty and 1 with 100 mg of ^^^U 650

Background h a s not been subtracted.

T a b l e 1.15, Comparison of Operating Characteristics of 150-cm Prototype Monitor and Proposed Plant Model

150-cm^ Proposed
Prototype Plant Model

14-Mev neutron flux 1,,67 X 10 IQll

Thermal-neutron flux 1,89 X 10 1.3 X l o '

Neutron flux utilization efficiency, % 1.13 13

B F , counting delay, s e c 2 0

B F , counting efficiency, % 0.42 4-5

leached h u l l s , it may be useful for finding fuel started this year to develop low-cost recovery
shards in fluoride volatility processing waste methods, adaptable to remote operation, which
and for determining fuel l o s s e s in liquid w a s t e s will provide decontamination factors of 10 to 100
from solvent extraction operations. from fission products.
The processing cycle for recycle reactor fuel
will probably c o n s i s t in removing the cladding
1.7 DEVELOPMENT OF CLOSE- a n d / o r moderating material by a physical head-end
COUPLED PROCESSES p r o c e s s (e.g., shear-leach, shear-burn, e t c . ) ;
separating the neutron poisons from the fuels by
The presence of 74-year ^^^U in recycled ^^^U, a chemical method; reconstituting the fuel, prob-
1.9-year " ^ T h in recycled ^^^Th, and 13-year ably by the sol-gel route (Sect. 6); and remotely
^ ' " P u in recycled ^ ^ ' P u makes the high-decon- refabricating the fuel a s s e m b l i e s . The product
tamination processing of recycle reactor fuels solution from the decontamination step must a l s o
unnecessary b e c a u s e of the high gamma radiation meet the restrictions imposed by the subsequent
background of the daughters of t h e s e short-lived sol-gel p r o c e s s regarding the type and amount of
i s o t o p e s in t h e recovered fuel. That i s to say, impurities present.
s i n c e the recovered material must be handled by Work done this year to find a s u b s t i t u t e for
remote manipulation in any c a s e , the decontam- solvent extraction included research on methods
ination from fission products need not be great — such a s electrodialysis, precipitation, and ion
just enough to prevent their accumulation during exchange, which might provide cheaper ways of
multiple fuel c y c l e s . Therefore, a program was obtaining low-decontamination recovery of the

fuel. So far, the outlook for t h e s e methods has original solution volume at 0.4 mole of NOj per
not been promising. Therefore, a reexamination mole of Th to 60% at 4 moles of N O j " per mole
of solvent extraction chemistry and equipment was of Th.
begun in an attempt to achieve c o s t s a v i n g s Ion Exchange. - Preliminary ion exchange
through the use of simplified equipment and s t u d i e s were also made to determine if the fission
operating methods capable of obtaining lower products could be adsorbed from acid-deficient
decontamination factors. One result of t h e s e thorium-uranium solutions. When a solution that
s t u d i e s , which could reduce the subsequent w a s t e - was 0.49 M in Th, 0.025 M in U, and 0.15 M in
management c o s t s , has been the satisfactory N O , " and which contained 2.1 x 10® counts m i n " '
reduction of Pu(IV) to Pu(III) in nitrate solutions m l " of gross gamma was p a s s e d through columns
with Ar—4% H^ in a column of platinized alumina. containing various adsorbents, l e s s than 20% of
Another result h a s been the development of s a t - the activity was adsorbed on Dowex 50W and Z P - 1 .
isfactory solvent extraction and ion exchange None of the fission products were adsorbed on
flowsheets for removing ^^*Th and its daughters s i l i c a gel or Dowex 1.
from " ^ U . Electrodialysis. -^ Thorium-uranium nitrate
In the sol-gel p r o c e s s , T h O j s o l s , about 0.2 M solutions and mixed-fission-product solutions
in nitrate, are currently prepared by steam deni- were electrodialyzed in an attempt to remove and
tration of Th(NO )^ solution. An alternative separate fission products. Electrolytic p r o c e s s e s
p r o c e s s , in which the sol i s prepared by amine have an advantage over others b e c a u s e they do
extraction of nitrate from Th-U nitrate solutions, not require the addition of reagents that may give
was developed. Sol-gel materials produced by this difficulty in subsequent operations. In electro-
new method are now being evaluated. d i a l y s i s , ion exchange membranes are used to
separate the compartments of the electrolytic
cell and thereby increase the efficiency of ion
Studies of Decontamination by Precipitation, separation.
Ion Exchange, and Electrodialysis During a 6-hr electrodialysis of solutions that
were about 0.5 M in Th, 0.02 M in U, and 0.3
Scouting experiments on the use of precipitation, M in N O g " and which contained 2 x 10* counts
ion exchange, and electrodialysis of acid deficient m i n " ' m l " ' of gross gamma, 50 to 55% of the
solutions as alternatives to solvent extraction thorium and uranium and 10% of the gamma activity
to effect low decontamination of reactor fuels precipitated on the cathodic membrane. When the
from fission products were generally unsuccessful. cell was maintained at 2.5 v, the initial current
Precipitation. - Thorium-uranium precipitation was 198 ma; but during the 6-hr experiment, the
s t u d i e s were made by using dibasic ammonium current dropped to 22 ma. About 35% of the gamma
phosphate, oxalic acid, and 30% hydrogen peroxide activity, mostly ' ^ ' ' " ' ^ ' ' C s , transferred to the
a s precipitants. Dibasic ammonium phosphate cationic compartment, and l e s s than 0.1% t r a n s -
produced the most quantitative precipitation; ferred to the anionic.
however, most of the fission products were carried Attempts to separate fission product elements
with the thorium-uranium precipitate. For example, from a thoria sol by electrodialysis were a l s o
more than 99.6% of the thorium and uranium and unsuccessful, due both to limited migration of the
98.8% of the fission product gamma activity were fission products and to the instability of the sol
precipitated by adding 0.32 to 2 moles of d i b a s i c in the electrolytic field. Only 5 to 10% of the
ammonium phosphate per mole of thorium to a fission products migrated to the cathodic compart-
solution that was about 0.5 M in Th, 0.026 M in ment and none to the anodic after 5 hr at a poten-
U, and 0.37 M in N 0 , ~ and which contained 1.6 x tial of about 25 v. The positively charged ThO
10* counts min"'' m l " ' of gross gamma activity. sol migrated to the cationic membrane, where it
With solutions in which the nitrate-to-thorium precipitated. Very limited s u c c e s s in preventing
ratio was varied from 4:1 to 0.4:1, the amount of the precipitation was achieved by dilution and
thorium and uranium precipitated was independent stirring. Addition of sodium ethylenediamine-
of the nitrate concentration. However, the volume tetraacetate (EDTA) or sodium citrate increased
of the precipitate increased a s the nitrate concen- the removal of fission products to better than 30%
tration increased, ranging from about 15% of the but decreased the stability of the sol.

The EDTA complexes of the long-lived fission investigated for this application. One device
product elements have been separated — in the being evaluated is a differential extractor, which
a b s e n c e of thorium — in an electrodialysis cell is a simple one-stage mixer-settler unit in which
by varying the pH of the s o l u t i o n . ^ ' The pH's at a s t a t i c aqueous phase is maintained in the equip-
which 90% of the Zr-Nb, Ru, Eu, Ce, Sr, and La ment while the organic phase is slowly p a s s e d
transferred a s anions were < 0 . 5 , 1.4, 1.8, 2.2, through the unit. The two p h a s e s are stirred
2.4, and 3.6 respectively. Cesium remained together vigorously in the mixer section and
cationic at pH values a s high a s 11. Separation continuously circulated to the s e t t l e r , where they
into groups containing Zr-Nb and Ru; Eu, Ce, and disengage. The organic phase overflows con-
Sr; L a ; and Cs were obtained by controlling the tinuously from the equipment, and the aqueous
pH during e l e c t r o d i a l y s i s . This migration of phase is circulated back to the mixer. A back-
fission products may have application in isotope extraction unit may be operated similarly or may
production or in waste disposal but does not be modified to operate with the organic a s the
seem practical in fuel processing. s t a t i c p h a s e . Due to the nature of the contactor,
concentration of the product ion in the organic
Solvent Extraction Studies phase is at near saturation during a large portion
of the recovery cycle, thereby increasing decon-
Since largely negative results were obtained tamination from impurities. The simplicity of
in scouting t e s t s to evaluate precipitation, ion this type of contactor results in satisfactory oper-
exchange, and electrodialysis a s p o s s i b l e alterna- ation with a minimum of care and makes it adapt-
t i v e s to solvent extraction a s low-decontamination able to intermittent operation, such as one shift
p r o c e s s e s , emphasis has been increased this year per day. The simplicity of the operation also
on the development of simple and v e r s a t i l e solvent minimizes analytical requirements.
extraction equipment and flowsheets designed to
In a typical t e s t with a prototype unit, a 0.2
d e c r e a s e operating and analytical c o s t s . Cost
M acid-deficient solution 1.5 M in Th and 0.1 M
a n a l y s e s showed that no substantial capital saving
in U was contacted first with 6 volumes of 3%
can be made by decreasing the c o s t of the solvent T B P in n-dodecane and then with 7 volumes of
extraction contactor but that savings can be made 30% T B P in n-dodecane. More than 99.9% of the
by decreasing the complexity of the operation, uranium and 87% of the thorium were extracted.
thereby decreasing the number of analytical sam- The uranium product was decontaminated from
p l e s required and by having one operations crew thorium, rare-earth elements, zirconium, and
perform various t a s k s in sequence rather than ruthenium by factors of 16, = 2 . 6 x 10'', 300,
concurrently ( s e e Sect. 1.8.1). and = 500 respectively. The thorium-product de-
Work in this area h a s included the development contamination factors from uranium, rare-earth
of an efficient and reliable single-stage differential elements, zirconium, and ruthenium were 7 x 10"*,
contactor and the development of flowsheets for 16, 6, and 4 respectively. Better separation from
u s e in d e c r e a s e d - s t a g e mixer-settlers. New rare earths is expected with more radioactive
solvent extraction and ion exchange flowsheets feeds.
were also developed for the cleanup of stored
The equipment was operated easily, and oper-
^^^U, and a new method for reducing Pu(IV) to
ation was nearly automatic. Also, the unit could
Pu(III) with Ar—4% H j was demonstrated.
be stopped and restarted e a s i l y . For example,
Development of Differential Contactor. - P r e s e n t
several experiments were shut down overnight by
solvent extraction p r o c e s s e s for recovering and
turning off the solvent flow, allowing the solvent
decontaminating fissionable and fertile materials
in the system to collect in the disengaging s e c t i o n ,
from fission products u s e complex contactors that
which required about 15 min, removing the accu-
require closely supervised operation and many
mulated solvent from the tube, and turning off the
chemical a n a l y s e s . Therefore, cheaper and
mixer. The next morning the system was restarted
simpler, but l e s s efficient, contactors are being
by simply turning on the mixer and the solvent
pump. B e c a u s e of the high saturation of thorium
A. F a c c h i n i and R. H. Rainey, The Separation of in the solvent during the first portion of the thorium
Long-Lived Fission Product Elements by Using Ethyl- c y c l e , two organic p h a s e s were present, but the
enediaminetetraacetate in an Electrodialysis Cell,
ORNL-TM-1035 (1965). equipment continued to operate satisfactorily.


In both the uranium and the thorium c y c l e s , the in diethylbenzene (DEB) ( F i g . 1.15). The ex-
solvent was nearly saturated during extraction of traction product w a s stripped with water; most of
the first 75% of the material. Then, after a short the third-stage strip solution w a s recycled to t h e
transition period, the concentration in both the scrub and extraction section. T h e product solu-
solvent and the aqueous p h a s e s decreased ex- tion contained about 32 g of U and 0.16 g of Th
ponentially ( F i g . 1.14). Operating variables may per liter; the decontamination factor for uranium
therefore be controlled with a minimum of anal- from thorium was about 6000. Decontamination
yses. factors from fission products were not determined
Countercurrent Recovery P r o c e s s e s . — In ad- in this t e s t ; however, subsequent t e s t s , in which
dition to the differential extraction s t u d i e s d i s - the stage samples were spiked with fission prod-
c u s s e d above, tentative flowsheets were outlined ucts and reequilibrated, indicated that decon-
for recovering uranium alone, and for the simul- tamination factors from rare e a r t h s , zirconium-
taneous recovery of uranium and thorium in a niobium, and ruthenium of > 1000, > 100, and
countercurrent (mixer-settler) system with three > 1 0 0 , respectively, could be expected.
extraction, one scrub, and three stripping s t a g e s . The number of s t a g e s and their arrangement
The process for recovering uranium alone w a s were e s s e n t i a l l y the same for corecovery of
successfully demonstrated in a batch countercur- uranium and thorium, but the solvent w a s 1 M
rent t e s t with an 8 M H^ dissolver solution that T B P in diethylbenzene. The thorium product
contained 6.9 g of U and 216 g of Th per liter. solution w a s withdrawn from the first stripping
About 99% of the uranium was recovered with s t a g e and the uranium product from the third.
0.25 M di-sec-butyl phenylphosphonate (DSBPP) Development of Process for Purifying U. —
Since ORNL i s a national depository for ^^^U,
large quantities (several hundred kilograms) will
3 ORNL-DWG 6 5 - 9 7 0 5
soon be stored in new storage facilities being
ORIGINAL SOLUTION: 3 2 0 g Th per liter, ^
2 7 g U per liter, 2 M H N O j ^
installed in Building 3019 ( s e e Sect. 17.7). After
U EXTRACTED WITH 3 7 o T B P - N D D its initial separation from the thorium in which
it was produced, or after i t s decontamination
following u s e a s a reactor fuel, the stored ^^^U
rapidly becomes radioactive again a s a result of
the buildup of the daughters of the ^^^U contami-
nant. T h e 2.6-Mev gamma from ^os-pj jg (.j^g
controlling activity for shielding requirements,
but other daughters also emit considerable gamma
activity. Consequently, before t h e 2 3 3y jg shipped
to u s e r s , it i s advisable to remove the ^^^U
daughters immediately before shipping.
Although adequate cleanup of ^^^U can be ac-
complished in the existing Thorex solvent extraction
pilot plant in Building 3019, the procedure is awk-
ward and expensive since a large amount of thori-
um must be added as a neutron poison; the equipment
i s not geometrically safe. T h e added thorium also
s e r v e s a s a salting agent in the extraction of
the ^^^U with 3% di-sec-butyl phenylphosphonate
in diethylbenzene (DSBPP in DEB). For decon-
tamination of large withdrawals, u s e of the pilot
plant i s probably warranted on a temporary b a s i s ;
but for small shipments, new flowsheets, using
SOLVENT THROUGHPUT f VOLUMES S O L V E N T \ either solvent extraction or ion exchange, were
developed, and a new solvent extraction system
Fig. 1.14. Concentration of Uranium and Thorium using small, geometrically safe equipment w a s
During Extraction in a Differential Contactor. provided for this service.

ORNL-DWG 65-9706



(30) U 77 U 71 U 0028
SOLVENT- Th 0 78 Th 0 0 4
H+ 0 18

U 0.04 U 20 U 6 5 U 3)7 (5) U 24

Th 109 Th 5 7 Th 0 16 (30)
H+ I ( H+ 0 24



Fig. 1.15. Flowsheet for Selectively Recovering Uranium from U-Th Fuel Solution with D S B P P in D E B in a
Reduced-Stage M i x e r - S e t t l e r . Aqueous f e e d : prepared by d i s s o l v i n g u n i r r a d i a t e d U O j - T h O - fuel in 13 M HNO_ —0.04
M H F ; 8 M In H , and c o n t a i n e d . In grams per l l t e r , 6 . 9 U, 216 T h , 6.7 F e , 1.0 A l , and 0.37 Cr. Numbers In the
b l o c k s show the s t e a d y - s t a t e molar c o n c e n t r a t i o n of H and the u r a n i u m and t h o r i u m c o n c e n t r a t i o n s , in grams per
l i t e r , of stage samples from batch c o u n t e r c u r r e n t t e s t . Numbers in p a r e n t h e s e s show the r e l a t i v e v o l u m e f l o w s .

Laboratory experiments with the new p r o c e s s e s When the solution contained 28 g of U per liter,
showed that the ^^^U may be decontaminated from the maximum decontamination factor was about
the daughters of ^^^U either by extracting with 12, and the capacity was greater than 4.6 g of U
30% tributyl phosphate in n-dodecane (TBP in per milliliter of resin. When the solution contained
NDD) or by adsorbing the ^^*Th (the immediate 123 g of U per liter, the maximum factor was about
daughter of ^^^U) on Dowex 50W ion exchange
resin. However, solvent extraction is preferred
b e c a u s e of better decontamination. The solvent
extraction flowsheet feed stream compositions and )0^
ORNL-DWG 6 5 - 9 7 0 7

rates are as follows:


Feed: 0.43 M U02(N03)2. 0.3 M A1(N03)3; \ [ 1 11

\ \ \
1 vol 2
Scrub: 0.2 M A1(N03)3, 0.1 M Al(OH)(N03)2; O 2 \ ,o • 1ON EXCHAN GE
t-, )0'
0.1 vol r^ JX
Solvent: 30 vol % T B P , 70% NDD; 1.2 vol ° 5 H
L\ ^
About 4.5 hr after extraction, the reduction of h V\
^"^Tl activity was nearly a factor of 600 but ^ ^r
frl *0
decreased to about 100 after 44 hr and to 10 after
\\ y^ T
19 days (Fig. 1.16). ^ j ^

T h e ion exchange experiments c o n s i s t e d in

> r\ V.
/ r°Sr>
p a s s i n g solutions of various uranium concen- 2
^ / VS\ s

trations through a column containing Dowex 50W

resin, 50 to 100 mesh. T h e first day after treat-
)0'= 2 5
ment, the decontamination factor from ^os-pj ^ g g TIME (hr)
only about 2, but it increased to about 10 after
12 days (Fig. 1.16). Decontamination from ^os-pj Fig, 1.16.
Comparison of Decontamination of U
and the resin throughput capacity both decreased from the Daughters of U by Solvent Extraction or Ion
a s the concentration of uranium was increased. Exchange.

10 (Fig. 1.17), and the capacity was about 3.6 the column was about 10 min. More than 99% of
g of U per milliliter of resin. the plutonium was reduced and transferred to the
T h e s e data (Fig. 1.16) indicate that even though aqueous p h a s e . When 100% hydrogen was used,
solvent extraction provides a greatly superior some of the uranium was reduced, but there was
initial decontamination factor (nearly 1000, com- no noticeable reduction of uranium with the diluted
pared with 2 for ion exchange), the radioactivity hydrogen. The Ar—4% H j mixture i s a l s o safe
of the 2 3 3jj pfoduct for all times beyond two weeks s i n c e it i s below the explosive limit of hydrogen.
after cleanup is the same for both methods. In plant operation, the organic-aqueous mixture
Reduction of Plutonium by Hydrogen. — To from this plutonium reduction step would be fed
achieve satisfactory separation of uranium and into a partitioning column, where the separation
plutonium in the partitioning cycle of the Purex of the uranium and plutonium would be completed.
solvent extraction p r o c e s s , the plutonium i s
reduced from Pu(IV) to Pu(III). In present oper-
a t i o n s , ferrous sulfamate or uranous nitrate i s Denitration of Thorium Nitrate by Amine Extraction
normally used a s the reductant. Use of the former
introduces iron and sulfate into the intermediate- The extraction of nitrate from thorium nitrate
level waste system, which should be avoided, if and from thorium-uranium nitrate solutions with
p o s s i b l e , for good waste management. If uranous amines is being investigated in the laboratory a s
nitrate i s used, its ^^*U content must match that an alternative to the currently used steam-deni-
of the uranium in the fuel; this can lead to in- tration process for the preparation of Th (and
creased cost. In experiments done this year, an Th-U) oxide sols for u s e in the sol-gel p r o c e s s .
Ar—4% H j mixture was found to be a satisfactory In single-batch extraction experiments, an
plutonium reductant without the disadvantages of aqueous solution, about 0.4 M in Th, 0.02 M in
either of the ones presently used. U, 1.8 M in N 0 3 ~ , and 0.1 A^ in H"^ and containing
In laboratory experiments, 30% T B P (0.1 A^ in gamma emitters equivalent to 2 x 10® counts
HNO3) containing 95 g of U and 1 g of Pu(IV) min""' m l ~ ' (gross gamma) was contacted with
per liter, an aqueous solution 0.5 N in HNO3, and 0,2 M Primene JM-T in n-dodecane for 2 hr at
Ar—4% H j were p a s s e d cocurrently through a 90° C. When the initial mole ratio of amine to
column packed with alumina pellets containing aqueous nitrate was varied from 0.25 to 1.9, the
0.5% platinum. Holdup time of the solutions in nitrate became l e s s extractable a s the solution

ORNL-OWG 6 5 - 9 7 0 8

TIME ( h r )

233 208
F i g . 1.17. Effect of Feed Concentration on Separation of U from T l by Ion Exchange.

became more acid deficient. In t h e s e t e s t s , the ated with an existing solvent extraction plant,
aqueous nitrate-to-thorium mole ratio approached for HTGR fuel (ThC^-UC^-graphite).
a value of 0.2 asymptotically. L e s s than 0.5%
of the thorium, uranium, and gamma activity was
extracted with the nitrate. Increasing the contact Studies on the Design of Small Plants
time from 2 to 6 hr, or using multiple c o n t a c t s
rather than a single contact, did not reduce the A design and cost study of a small on-site
final nitrate-to-thorium ratio below 0.2. Extractions plant for processing reactor fuel is nearing com-
at room temperature led to emulsions. A 1- pletion. The plant is designed to be operated in
nonyldecylamine solution also satisfactorily ex- conjunction with a 500- to 3000-Mw (electrical)
tracted nitrate from thorium-uranium nitrate solu- boiling-water reactor station operating on slightly
tions. enriched fuel. A number of design and c o s t s t u d i e s
The thorium s o l s prepared by amine extraction already exist for chemical processing plants for
were dried and fired at 1150° C. Although both the variety of fuels from power-reactor complexes
amines yielded a product solution with nearly expected during the next decade. However, in
the same thorium-to-nitrate ratio, the solutions almost all t h e s e s t u d i e s , central plants of large
prepared by extracting the nitrate with 1-nonyldec- capacity were assumed.
ylamine contained only about )'^ as much carbon There has been recurring interest in the capital
a s those prepared with Primene JM-T. Contacting and operating cost of a small plant at a single
10 vol of 0.2 M 1-nonyldecylamine in n-dodecane power station of modest s i z e , from 500 to 3000
with 1 vol of 0.4 M Th(N03)^ for 2 hr at 90°C Mw (electrical). In the immediate future, plants
produced a sol 0.4 M in Th, 0.16 M in N 0 3 ~ , and of only relatively small processing capacity will
0.01 W in C at a pH of 5.13. When fired at 1150° be needed. Furthermore, the diversity of fuel-
for 4 hr in air, it produced a dull-gray g l a s s y oxide element d e s i g n s and the widely scattered d i s -
with a density (measured by mercury penetration) tribution of power reactors, in operation or pro-
of 9 . 8 1 , a porosity of 2.00%, and a surface area posed, may favor the small, single-purpose
of 0.004 m ^ / g . The same sol when fired in Ar—4% processing plant, operated in c l o s e conjunction
H^ produced a white oxide with corresponding with the power station, a s opposed to large,
values of 9.72, 2.66%, and 0.017 m V g . central, multipurpose plants. If the chemical
The acid-deficient thorium-uranium nitrate solu- plant i s small enough, many of i t s auxiliaries can
tions prepared in t h e s e experiments were also be t h o s e of the reactor station whose fuel it will
used in some of the fission product decontam- p r o c e s s . Stack facilities, waste treatment, main-
ination t e s t s , where precipitation, ion exchange, tenance s h o p s , and administrative facilities such
and electrodialysis were tried (Sect. 1.7.1). a s offices and lunchrooms can be shared by the
reactor and chemical plant with s a v i n g s to both.
This small-plant design study reflects operating
and maintenance procedures and standards con-
1.8 CONCEPTUAL PLANT STUDIES s i s t e n t with power-reactor i n s t a l l a t i o n s . It pro-
vides reasonably accurate cost data for fuel
The preparation of conceptual plant s t u d i e s processing in a plant operating on the s i t e of a
and cost estimates for present and new p r o c e s s e s power station, which, for the purpose of this
was continued this year. Three s t u d i e s were study, was assumed to be the Oyster Creek nuclear
made. The first, a conceptual plant for a burn- power station proposed by the J e r s e y Central
leach and a bum-volatility p r o c e s s for Rover Power and Light Company. The zirconium-
fuel, is reported e l s e w h e r e ' ^ for classification sheathed, slightly enriched fuel element of this
r e a s o n s . The second included the designing and reactor is similar to those used in previous c o s t
costing of a conventional small on-site processing s t u d i e s , making comparison with those r e s u l t s
plant (VjQ to '/^ ton of U per day) for a boiling- reasonably valid.
water reactor station of the Oyster Creek type A fairly conservative Purex-type flowsheet was
(proposed by J e r s e y Central) with an output of assumed for the 23,000-Mwd/ton, 120-day-cooled
about 3000 Mw (electrical). The third study was fuel. Equipment design was b a s e d on the rack
for a bum-leach head-end facility, to be a s s o c i - concept throughout, to facilitate both maintenance
and installation. It was assumed that most heavy servative estimates were used throughout this
maintenance would be purchased by contract, a s study, and a reasonably accurate detailed engi-
would most of the analytical s e r v i c e s , to avoid neering design estimate might well result in a
the high cost of elaborate shops, analytical somewhat lower figure. But the probable d i s -
facilities, and the corresponding staff personnel. parity does not justify the expense of a detailed
Capital-cost estimates were based on detailed design and engineering cost estimate.
conceptual equipment design and flowsheets, but
the capital c o s t s of buildings were estimated on
Design Studies on Processing of HTGR Fuel
the b a s i s of factors derived from Commission
experience. The estimates of operating c o s t s
A preliminary plant design and c a p i t a l - and
do not include in-process inventory charges or
operating-cost estimate was made for a head-end
credit allowance for the plutonium product. The
facility for processing spent HTGR fuel elements
capital and operating c o s t s are summarized in
(ThCj-UCj-graphite; s e e also Sect. 1.2). The
Table 1.16.
plant would provide for the following s e r v i c e s :
A power station consisting of two 500-Mw (elec-
trical) reactors of the Oyster Creek type will 1. receiving and storing the spent fuel elements,
discharge fuel at a rate of about 51,000 kg of 2. crushing them,
uranium per year. B a s e d on the above estimated
3. burning the graphite in a fluidized bed,
operating c o s t s , the fuel-processing unit cost is
$51 per kg of U. Similarly, assuming a burnup of 4. leaching the burner ash,
23,000 Mwd/ton and a thermal efficiency of 30%, 5. adjusting the leachant to acid-deficient con-
the fuel-processing-cost contribution to the c o s t ditions suitable for solvent extraction,
of power would be 0.29 mill/kwhr.
6. storing the adjusted leachant, and
This cost study will be completed by estimating
the capital and operating c o s t s of a supplemental 7. storing the raffinate from the solvent extraction
addition to the small plant for converting plutonium plant.
and uranium nitrates to oxides, and fabrication
of complete fuel assemblies for recycle to the This head-end facility was assumed to be located
reactor. at a conventional fuel processing plant, such a s
While it is recognized that considerable un- the one built by Nuclear Fuel Services, Inc. (NFS).
certainty e x i s t s in cost estimates that are not At this plant, the uranium and thorium would be
based on detailed design and development, con- recovered separately by solvent extraction, the

T a b l e 1.16. Capital and Operating Costs for Small On-1 ite Processing Plant Using Established Technology

Capital Cost
Building structure $2,257,000
Equipment and installation 2,964,000

P l a n t total $5,221,000

Contingency (25%) 1,305,000

Startup and training (1 y e a r ' s direct operating c o s t s ) 930,000

Total capital investment $7,456,000

Annual Operating Cost

Operating cost — direct $ 930,000
Capital investment — annual charge 1,680,000

Total annual operating cost $2,610,000

E x c l u s i v e of fuel in-process inventory charges and exclusive of plutonium credit.


thorium returned to the head-end facility for decay bed burning, leaching, and feed adjustment. At the
storage, and the fission product waste concen- nominal throughput rate, such a facility could
trated for d i s p o s a l . handle the fuel from reactors having a combined
The study was made for a fuel that contains installed capacity of 10,000 Mw (electrical). The
sol-gel oxide microspheres coated with pyrolytic c e l l s for unloading and storing fuel were designed
carbon, which was assumed to be the favored for remote operation and limited remote mainte-
HTGR fuel of the future. The fuel element was nance; the chemical process c e l l s were designed
assumed to be a 4.5-in.-diam, 20-ft-long graphite for direct maintenance. Analytical, administrative,
" l o g " from which the fueled particles are not chemical supply, waste d i s p o s a l , and plant utility
e a s i l y separable; thus, provisions are made for s e r v i c e s were assumed to be provided by the
c m s h i n g and burning the entire element. Each a s s o c i a t e d solvent extraction plant (not c o s t e d ) ,
element contains 107 kg of carbon and about with appropriate enlargements where n e c e s s a r y .
10.9 kg of thorium plus uranium, before irradiation. The estimated capital cost of the head-end
After an assumed burnup of 50,000 to 80,000 facility is $9,040,000 (Table 1.17). T h i s includes
Mwd/metric ton in about six years of irradiation $1,260,000 for the second processing line, an
and six months' decay for fission product heat expenditure that could be postponed until required
reduction and protactinium decay, the element if the plant were started on a fraction of i t s design
contains about 10 kg of thorium plus uranium, the load. The standby operating c o s t , that i s , the
b a l a n c e having been converted to fission products. minimum cost of labor, u t i l i t i e s , and overhead for
The nominal processing capacity i s 40 elements maintaining the plant when fuel is not being
a day for 225 days a year. Two parallel lines of processed, is estimated to be $115,000 a year.
equipment were provided for c m s h i n g , fluidized- Additional labor and overhead c o s t s when fuel

T a b l e 1.17. C a p i t a l - C o s t Estimate for H T G R Head-End F a c i l i t y

Item Cost

Building and s e r v i c e s $ 550,000

Building equipment 130,000
Cell structure 1,295,000
Cell s e r v i c e s 565,000
Cell equipment 247,000
P r o c e s s equipment 1,098,000
P r o c e s s piping 906,000
P r o c e s s and radiation instrumentation 350,000
Outside equipment 346,000
Site improvements 50,000
Utilities 85,000

Subtotal $5,622,000

Engineering and inspection (20%) $1,124,000

Subtotal $6,746,000

Contingency (25%) $1,687,000

Subtotal $8,433,000

Interest during construction, startup c o s t s , and working c a p i t a l $ 607,000

Total $9,040,000

Installed c o s t , including contractor's overhead and profit.


is being processed are estimated at $350 a day. waste disposal, e t c . , were not estimated separately
The cost of oxygen and alumina is estimated at s i n c e they were considered to be a part of the
$846 a day for one burner line, or $1020 for two normal solvent extraction c o s t s . The c o s t s of
l i n e s . The lower unit cost for two lines is the t h e s e items should be nearly the same a s those
result of lower oxygen c o s t s at the high usage for standard metal-clad oxide fuel on an equivalent
r a t e s . The c o s t s of nitric acid, other c h e m i c a l s . throughput b a s i s .
2. Fluoride Volatility Processing

The investigation of fluoride-volatility p r o c e s s e s Building 3019. Laboratory- and b e n c h - s c a l e studies

at ORNL i s part of an intersite program to develop in support of the program will also be conducted.
an alternative to aqueous methods for the recovery F i r s t priority will b e given to the p r o c e s s using
of values from spent nuclear reactor fuels. Fluo- anhydrous gaseous HCl for removing Zircaloy
ride-volatility methods effectively separate uranium cladding. T h e use of H F - 0 for removing s t a i n l e s s
(and plutonium, when present) from fission products s t e e l and Zircaloy claddings will be studied in
by formation of the highly volatile UFg and P u F g . that order if smaller-scale t e s t s now in progress
Studies here have ranged from laboratory-scale to show that the methods based on the use of H F - 0
pilot-plant demonstrations with irradiated fuel are feasible. The development work for the fluo-
elements. ride volatility fluidized-bed method has been per-
The development of fluoride-volatility methods at formed mostly at Argonne National Laboratory
ORNL is presently in transition between the u s e (particularly the work with HCl), with l e s s e r
of molten-salt and fluidized-bed methods. Previ- amounts at Brookhaven, ORNL, and the Oak Ridge
ously, emphasis was on (1) processing molten-salt Gaseous Diffusion Plant.
reactor fuels or metallic fuels that had been dis-
solved in molten salt, and (2) on the u s e of solid T h e p o s s i b l e advantages of fluoride-volatility
sorbents for removing volatile fission product p r o c e s s e s over aqueous methods are a s follows:
fluorides from the UF produced by the reaction (1) Chopping of the fuel rods from power reactors
of fluorine with the U F ^ in the melt. Specifically, will probably not be necessary. (2) Radiation
p r o c e s s e s were developed for recovering highly damage to reagents will not be a problem when
enriched uranium from the molten salt used in the processing short-decayed fuel. (3) Problems of
Aircraft Reactor Experiment, from zirconium-clad nuclear safety are l e s s severe. (4) Fewer process-
alloys of zirconium and uranium, and, most recently, ing s t e p s are required. (5) Highly radioactive
from aluminum-clad alloys of aluminum and ura- waste from fluoride-volatility processing i s dry and
nium. Work on molten-salt methods was terminated highly concentrated.
with the successful demonstration in the Volatility
Fluidized-bed methods for processing low-enrich-
Pilot Plant of the recovery of uranium from alu-
ment UO power-reactor fuels appear more promis-
minum-base fuel elements. One exception i s a few
ing than molten-salt methods b e c a u s e of the ex-
more s t u d i e s of the removal of plutonium (and
c e s s i v e corrosion of the reaction v e s s e l expected
probably protactinium) from droplets of molten s a l t
while plutonium i s being fluorinated from a bath
falling through a fluorine atmosphere.
of molten salt. Use of a spray tower or frozen-wall
Work on the newly emphasized fluidized-bed reactor would l e s s e n this corrosion, but the s t a t e
methods begun during this reporting period was of development of both t h e s e methods i s consider-
prompted by the announced AEC goal of having a ably l e s s advanced than the fluidized-bed approach.
fluoride-volatility technology for low-enrichment
fuel reprocessing completely developed through Sections 2 . 1 , 2.2, 2.3, and 2.12 are concerned
" c o l d " engineering, " c o l d " semiworks, and " h o t " with molten-salt methods; s e c t i o n s 2.4, 2.5, and
pilot-plant programs by July 1, 1969. Our principal 2.11 are specific to fluidized-bed work, and the
contribution will be the installation and operation other s e c t i o n s deal with problems related to
of a pilot plant in several of the shielded cells of fluoride-volatility methods in general.


2.1 MOLTEN-SALT PROCESSING OF absorber are disposed of by discharge into the

URANIUM-ALUMINUM ALLOY FUEL molten fluorides from which the uranium has been
volatilized. The effect on liquidus temperatures
Aluminum-base fuels are a major portion of the of NaF addition to molten KF-ZrF - A l F , was de-
4 3
anticipated processing load from reactors fueled
termined to ascertain how much NaF could be dis-
with enriched uranium. Since the anticipated
posed of in this way.
volume of zirconium-base fuels i s too small to
The System KF-ZrF .-AlF,: Refinement of Phase-
justify a plant for processing t h e s e fuels alone,
Equilibrium Diagram. — Liquidus temperatures in a
the molten-salt fluoride-volatility process previ-
portion of the KF-ZrF .-AlF system were redeter-
ously developed for zirconium-base f u e l s ' was
mined. Very pure materials are required for such
adapted to the processing of uranium-aluminum
measurements, partly b e c a u s e the visual determina-
alloy fuel. Laboratory development and engineering
tion of the liquidus point of the melt depends on
studies of the molten-salt fluoride-volatility proc-
the appearance of a precipitate, and partly b e c a u s e
e s s for aluminum-base fuels have been reported.^'''
some impurities markedly change the liquidus
T h i s section describes additional laboratory de-
temperature. For example, near K F - Z r F , oxides
velopment and pilot-plant processing of fuel that
in concentrations up to 0.25 mole % i n c r e a s e the
had cooled for only a short time.
liquidus temperature about 5°C for every 0.01%
Development work directed toward adaptation of
oxygen (as oxide). The pure materials required
the molten-salt fluoride-volatility process to alu-
for measurements of liquidus temperatures were
minum-base fuels has been completed. Liquidus
obtained by subliming ZrF and AlF and by dis-
temperatures for the system KF-ZrF -AlF have
tilling KF from a melt with 2.3 mole % U F ^ . ^
been more precisely defined in the composition
A phase diagram of the system KF-ZrF -AlF
ranges in which melting difficulties had been
had been published previously.* B e c a u s e of dif-
encountered. Composition and temperature limita-
ficulties in keeping salt molten in the pilot plant,
tions were determined for discharging NaF into
additional data were obtained to more precisely
waste salt.
define the 550 and 600°C isotherms (Fig. 2.1).
Pilot-plant t e s t s included the processing of
"dummy" aluminum elements and of irradiated fuel *R. E. Thoma, B . J . Sturm, and E , H. Guinn, Molten-
Salt Solvents for Fluoride-Volatility Processing of
that had been cooled for as short a time as four Aluminum-Matrix Nuclear Fuel Elements, ORNL-3594
weeks. T h e pilot plant i s being decontaminated (August 1964).
in preparation for corrosion evaluation and dis- ^Chem, Technol. Div, Ann, Progr. Rept. May 31,
1964, ORNL-3627, p. 4 0 .

ORNL-DWG 65-2601
Investigation of Phase Equilibria

In pilot-plant experiments, solidification and pre-

cipitation were encountered with salt mixes that
presumably should have been molten. To l e s s e n
this difficulty, more precise liquidus values were
obtained for KF-ZrF - A l F , melts with composi-
tional ratios of pilot-plant interest.
At one stage of the molten-salt fluoride-volatility
p r o c e s s , waste NaF pellets from the movable-bed

^Chem. Technol. Div. Ann. Progr. Rept. May 31, 1964,

ORNL-3627, p. 29.
^Ibid., p. 40.
^M. R. Bennett et al., " F u s e d - S a l t Fluoride-Volatility
P r o c e s s for Recovering Uranium from Spent Aluminum-
B a s e d F u e l E l e m e n t s , " Ind, Eng. Chem,, Process
Design Develop, (in p r e s s ) .
Fig. 2.1. Portion of K F - Z r F . - A l F , Phase Diagram.
''work performed by B . J . Sturm and R. E . Thoma,
Reactor Chemistry Division. C o m p o s i t i o n s in mole %.

Even though the shift from earlier data was slight ORNL-DWG 6 5 - 9 7 0 9
from the standpoint of composition, changes in
liquidus temperatures were significant. The pre- NaF ADDED TO.
c i s e location of point A was important b e c a u s e X 5 2 . 0 K F - 3 1 . 0 Z r F 4 - 1 7 . 0 AIF3
with increasing KF concentration the liquidus O 51 2 K F - 2 3 . 3 Z r F 4 - 2 5 . 5 AIF3
temperature r i s e s very steeply to the K ZrF _-
K A l F , join. B e c a u s e of the resemblance of struc-
tures of cubic K A l F , [isostructural with
(NH^)3AlFg] and K j Z r F ^ [isostructural with
(NH ) Z r F _ ] , ^ considerable solid solution e x i s t s
along this join.
Based on earlier data, we had thought that the
desirable procedure was to operate along the line 3
AB. In light of the new data, which show a sharper
temperature rise near A and l e s s room for aluminum
addition between A and B, starting with about
64-36 mole % KF-ZrF i s obviously preferable. 600
Effects of NaF on KF-ZrF^-AJFj Melts. - Sodium
fluoride was added to KF-ZrF - A l F , melts to de-
4 3
termine i t s effect on liquidus temperatures. While
this study served as part of the search for a lower-
melting solvent, capable of dissolving more AlF ,
i t s principal purpose was to determine the effect
of adding a part of the NaF in the movable-bed
absorber to the fluorinator. T h e compositions 20
chosen for the study, 52.0-31.0-17.0 and 51.2-23.3- NaF (mole 7o
25.5 mole % KF-ZrF -AlF , were approximately
those obtained at the end of the volatility process Fig. 2.2. Liquidus Temperatures at Various NaF
by, respectively, a single-step dissolution of Concentrations in K F - Z r F , - A l F , , Showing the Effect of
aluminum in 62.5-37.5 mole % KF-ZrF^ solvent N a F Discharge into V o l a t i l i t y Pilot Plant Waste Salt.
and a two-step dissolution with additional KF
provided for the second step. T h e melts were
found to dissolve substantial amounts of NaF at nonirradiated simulated fuel and with highly burned
600°C, the process temperature (Fig. 2.2). The fuel elements. The program culminated in the
liquidus temperatures were lowered to about 500°C successful processing of highly enriched fuel,
at NaF concentrations of 10 and 13 mole % for 26% burned, after only 25 days cooling. (The
the respective solvents. They were raised to amount burned i s the percentage of ^ ^ ' u fissioned.)
600°C at 17 and 20 mole % respectively. Description of Process for Uranium-Aluminum
Alloy Fuel as Studied in the Pilot Plant. - T h e
Pilot-Plant Processing of Short-Cooled process used in the pilot plant for recovering
Uranium-Aluminum Alloy Fuel uranium from aluminum-base fuels was e s s e n t i a l l y
the same a s that used earlier for zirconium-base
P r o c e s s development in the pilot plant was the fuels. Since that p r o c e s s was described last
final phase of the ORNL program to adapt the year, ^ this report will only briefly outline most of
molten-salt fluoride-volatility p r o c e s s to aluminum- the process; changes will be described in detail.
b a s e fuels. Both the process and equipment were The flowsheet showing equipment and flow pat-
e s s e n t i a l l y those used in the earlier zirconium terns that were used for processing uranium-alumi-
program. Pilot-plant s t u d i e s were carried out with num alloy fuel elements i s shown in F i g . 2.3.

R. W. G. Wyckoff, Crystal Structures, vol. II, c h a p . '^Chem. Technol, Div, Ann, Progr, Rept. May 31, 1964,
IX, p. 3 1 , I n t e r s c i e n c e , New York, 1951. ORNL-3627, p. 30.
ORNL LR DWG 5 5 2 3 6 R 6 A R


FOR S C R U B B E R — — ^ ' " y ^ '

. T ^ — - • IT

• • — F2




Fig. 2.3. Process Flow Diagram of V o l a t i l i t y Pilot P l a n t .


This flowsheet encompasses the dissolution of but allowed some of the remaining contaminants to
fuel elements in a molten fluoride salt, fluorination flow out into the off-gas system.
to separate the uranium (as UF ) from the salt One of the contaminants discharged to the off-
and from most of the fission products, and further gas system was T e F , so a tellurium trap was
purification and recovery of the UF . T h e s e s t e p s installed in the off-gas line downstream from the
are described below in the same order. fluorine scrubber. The trap consisted of nickel
Fuel atid Salt Charging. — F u e l elements, such mesh heated to 400°C, a gas cooler, and an
a s ORR a s s e m b l i e s , were lowered into the dis- activated-charcoal bed near ambient temperatures
solver and then covered with a molten fluoride ( l e s s than 50°C). Although appreciable amounts
salt at about 600°C. [The composition of the salt, of tellurium were trapped, a significant fraction
either KF-ZrF or KF-ZrF -AlF j , may be any point was carried through to the off-gas s t a c k . Some of
within the 600°C isotherm (Fig. 2.1) provided that the tellurium was released after evolution from the
it permits the desired aluminum dissolution without fluorinator had c e a s e d ; thus, the trap extended the
causing the liquidus temperature to exceed 550°C.] r e l e a s e time of tellurium to the atmosphere.
One convenient salt mixture was 64-36 mole % After fluorination and sorption were completed,
KF-ZrF , made by melting commercial K . Z r F the entire movable-bed absorber was heated to
and Z r F , . When this mixture was added as a 400°C. The U F ^ was desorbed, p a s s e d through
liquid to the heel remaining from the preceding a bed of MgF pellets for removal of technetium
run, the resulting salt had a low enough liquidus and neptunium, and condensed a s a solid in cold
temperature to operate a " s a f e d i s t a n c e " above traps.
the liquidus temperature and below the melting
Results of Pilot-Plant Runs. — T h e aluminum
point of aluminum (about 655°C).
processing program in the pilot plant included ten
Fuel Dissolution. — The metallic fuel p l a t e s dissolution runs, ten uranium fluorinations, and
were dissolved by sparging anhydrous HF through eight desorptions. The dissolution runs, in chron-
the melt at about 600°C (575°C in two runs, to ological order, were: two dummy dissolutions
620°C in run) converting the metal to AlF^ and (aluminum only), three dummy dissolutions spiked
UF . Sparging rates have ranged from 20 to 130 with UF , four runs with irradiated fuel, and a
g/min, but were generally about 100 g/min. Times dummy run for cleanout. Fluorinations and de-
required for complete dissolution ranged from 11 sorptions started with the first spiked run; but in
to 29 hr. After dissolution was complete, the salt the second and third hot runs, the salt was fluo-
was transferred to the fluorinator for uranium re- rinated in two b a t c h e s . Since the runs with non-
moval. irradiated material served primarily to establish
Uranium Fluorination. — Salt was held in the system operability, data are presented below only
fluorinator at 30 to 50°C above i t s highest meas- for the four hot runs.
ured transition temperature; the temperature of the Fuels Processed. — Two fuel elements from the
last five fluorinations was 550°C. Fluorine contact Low-Intensity T e s t Reactor were processed in the
for about l \ hr converted the U F ^ to U F ^ . The first hot run, RA-1; the burnup amounted to 24%,
volatile UF., entered the movable-bed absorber, and the elements had cooled for about 1 \ years.
leaving most of the fission products in the salt. There were two Oak Ridge Reactor (ORR) elements
Product Purification and Recovery. — The UFg in each of the next two runs; RA-2 fuel was 31%
leaving the fluorinator was further purified in a burned and had cooled for six months, and RA-3
sorption-desorption cycle. T h i s resulted in the fuel was 28% burned and had cooled for 80 days.
discharge of some contaminants to the off-gas Only one ORR element was processed in RA-4;
system, retention of some on MgF^ p e l l e t s , and this element was 28% burned and was processed
return of the remaining ones to the waste s a l t . after only 25 days of cooling.
Most of this purification was done in the movable- Uranium Losses. — After the six fluorinations,
bed absorber, which was filled with NaF p e l l e t s . the uranium remaining in the salt ranged from l e s s
The pellets in the first section, at 400°C, retained than 0.1 to 8.7 ppm. T h e s e l o s s e s , added to other
some of the fission and corrosion products but l o s s e s , resulted in nonrecoverable uranium l o s s e s
allowed the UF to p a s s through. In the next for the four runs of 0.1, 0.8, 0.9, and 0.5% of the
section, held at 150°C, the NaF sorbed the U F ^ uranium charged in the individual runs.

Table 2 . 1 . O v e r - A l l Decontamination Factors in V o l a t i l i t y P i l o t P l a n t Aluminum Program"

Nonvolatile Isotopes Volatile Isotopes
Run. Time
No. Sr" 95 9 5, 99, 131
(days) 'Cs "Ba 'Ce Zr Nb Mo Ru Sb Te°

RA-1 ~570 6X10' >4xl0' >2xlo' >9 x l o ' 1x10* Sxio'*

RA-2 175 SXIO* 2 Xl o ' >4 x 1 0 ' ° >3 x l o ' >2 x l o ' 6x10*' >1 x l o '

RA-3 80 >4X10' 4X10' >1 X l o ' >2 x 10* >6 X l O ' >4 X l o ' >7 X l o ' 2 x lo' 1x10* 7xl0' >2xia*

RA-4 25 5X10* ixio' 2x10^ >9 x l o ' 1 x lo' 5 x lo' 1x lo' 36 >6xl0* 1x lo' 6x10* 4X10*

Decontamination factor i s defined a s the ratio of quantity of nuclide in fuel per gram of uranium to quantity m product per gram of uranium.

' ^ ' S r and " S r .



Decontamination Factors. — F i s s i o n product de- and the analytical results appeared quite reason-
contamination factors ( D F ' s ) are shown in Table able. Most of the cationic contaminants, including
2.1. T h e D F ' s are satisfactory with the exception Cr, Cu, F e , Ni, and Zr, were present to the extent
of those for antimony and 67-hr " M O in RA-4. of l e s s than 50 parts of cation per million parts
Generally speaking, the D F ' s listed for the first of uranium. Boron content w a s 75 ppm, and potas-
three runs are about one order of magnitude lower sium was about 500. Three others were present
than those achieved in the zirconium program. We in macro quantities; approximate values were: Mo,
believe that this i s due, at least in part, to fluo- 0.1%; Al, 0.3%; and Na, 1%. Molybdenum content
rinating at 550°C rather than at 500. T h e reduced of the product w a s dependent on conditions of
D F ' s in RA-4 may b e due to one or more of t h e operation of the movable-bed absorber. Either the
following factors: First, if the sintered-metal temperature during absorption should have been
filter just upstream of the product cylinder had increased slightly or the bed should have been held
failed, N a F fines could have carried fission prod- at 150°C for a longer time before desorption w a s
ucts into the UF cylinder. Second, in RA-4, the started. T h e high sodium content led u s to s u s p e c t
additional HF sparge at t h e end of dissolution was that the sintered-metal filter had failed, allowing
shortened, and a higher percentage of t h e fission NaF fines to b e carried through. The high alumi-
products may have reached t h e fluorinator. Finally, num content h a s not yet been explained.
and this factor seems most significant, discharge Release of Radiation Emitters to the Atmos-
of NaF from t h e movable-bed absorber may have phere. — Some radioactivity was released through
been insufficient. In the aluminum program, 3.8 kg the stack to the atmosphere during t h e hot runs.
of NaF (about 11% of the bed) w a s discharged each Rare g a s e s were released during dissolution but
run, compared with about 6.5 kg in t h e zirconium constituted only a fraction of the maximum permis-
program. Insufficient bed discharge would result sible concentration (MPC) at ground level. T h e
in an upward migration and eventual breakthrough only other detectable r e l e a s e s occurred at the
of some fission products. time of fluorination. In RA-4, the total '"^Ru
r e l e a s e was l e s s than 3 mc, and t h e total ^^H was
Cationic Contaminants. — Cationic contamination l e s s than 10 mc. T h e total tellurium r e l e a s e w a s
of the UF product was higher than desirable. about 20 curies; of t h i s , 6 curies was 78-hr ' ^ ^ T e ,
Analyses from individual runs (and even from mul- which decays to 2.3-hr '^^I and to stable xenon.
tiple samples from the product of a single run) Again, the ground-level radioactivity during the
were too inconsistent for valid interpretation. release w a s only a fraction of t h e MPC.
After the products from all hot runs were combined Personnel Exposure to Radiation. — Radiation
for shipment, the composite product was sampled. exposure was watched closely to determine t h e

effectiveness of personnel-exposure control during in the u s e of a high uranium concentration in the

the processing of short-cooled fuel. Even during melt to act a s a fluorine carrier, an approach
the " h o t t e s t " run, exposures were reasonable and, studied during this reporting period.
with one exception, were within the maximum limits Previous work showed that more than 99% of the
for routine radiation exposure. T h i s exception P u F , could b e volatilized from molten fluoride
arose from the need for immediate handling of the o

product cylinder, principally for sampling. Radio- s a l t s in 20 to 30 hr. The use of s a l t s having a
activity from the RA-4 product cylinder was high high initial concentration of uranium ( a s UF ) was
(about 50 r/hr at contact), due primarily to 67-hr studied a s a way of increasing the rate of evolution
' % o and 6.75-day ^^^U. As a result, the operator of P u F , . At an initial uranium concentration in
handling the product cylinder at the time of the the melt of about 25%, the percentage of P u F
run received an accumulated d o s e of 150 millirads volatilized in 1 hr increased to about 32%, com-
during the week of the run. No other operator re-
pared with about 15% volatilized in an hour in an
ceived 100 for the week.
earlier t e s t with no uranium present. These
Corrosion. — Corrosion data from laboratory- and results were not good enough to warrant further
engineering-scale studies have been reported.^ work.
Measurements of corrosion in the pilot plant have
been started. F i r s t preliminary results [l36 thick- Volatilization Rate and Equilibrium Constant
n e s s readings (ultrasonic meter)] indicate that Generally Increased as Initial UF Content
dissolver corrosion during the aluminum program was Raised
averaged 5 mils; if corrosion during decontamina-
tion i s ignored, this i s equivalent to about ]/ mil Apparatus and Operation. — The t e s t s were made
per dissolution. A few readings a s high as 30 mils in nickel reactors, each 8 in. high by 1 in. in outer
are believed to indicate total depth of p i t s . diameter, containing 50 g of L i F - N a F - Z r F (31-
24-45 mole %) and about 1 g of plutonium added
a s P u F . In a s e r i e s of five 1-hr fluorination
2.2 VOLATILIZATION OF PuF^ FROM MOLTEN t e s t s , the uranium concentration in each t e s t was
SALTS CONTAINING HIGH CONCENTRATIONS adjusted by adding U F ^ to the extent of about 5,
OF URANIUM 10, 15, 20, and 25 wt %. All t e s t s were made at
600°C, with a fluorine flow rate of 100 ml/min
One of the key operations in the fluoride-volatil- (STP). R e s u l t s of the t e s t s (Table 2.2) show a
ity process i s the recovery of plutonium as the significant i n c r e a s e in the P u F volatilization rate
hexafluoride. However, P u F ^ is not easily vola- and corresponding equilibrium constant K (K is the
tilized from the molten fluoride by direct fluorina- ratio of moles of plutonium in the gas p h a s e di-
tion. One method for achieving this may consist vided by the moles of plutonium in the salt phase).

T a b l e 2.2. Effect of Uranium Concentration on Rate of V o l a t i l i z a t i o n of P u F ^ from Molten

L i F - N a F - Z r F ^ (31-24-45 Mole %) at 6 0 0 " C

Uranium in Salt Plutonium in Salt

PuF, Half-Time Equilibrium
Test (wt %) (ppir 0 D

Volatilized (hr) Constant, K

Initial Final Initial Final (%)

1 4.74 0.0004 1140 972 14.8 4.4 0.32

2 9.19 0.0015 1060 854 19.1 3.3 0.43

3 14.8 0.130 805 607 23.3 2.6 0.69

4 19.4 0.125 627 576 8.2 8.1 0.20

5 24.5 0.25 805 553 31.6 1.8 0.94


T h e cause for the erratic results in test 4 i s un- blends. R e s u l t s clearly indicate that this approach
known, but it does not appear to signify any to recovering the uranium can lead to much faster
change in the general trend of the data. The ef- removal than that provided by bubbling fluorine
fective half-time of 4.4 hr at about 5 wt % initial through a pool of the molten fluorides.
uranium i s comparable with values obtained in
With respect to the volatilization of plutonium by
t e s t s where the uranium concentration varied be-
fluorinating falling droplets of fluoride melts, five
tween zero and 0.5 wt %. At about 25 wt % uranium
laboratory experiments were made. The results
(Table 2.2), this value was reduced to l e s s than
led to the conclusion that a spray-fluorination
process for plutonium recovery i s chemically feas-
ible. Most of the plutonium work i s now complete
2.3 RECOVERY OF PLUTONIUM AND URANIUM and will be discontinued after a few more experi-
FROM FALLING MOLTEN-SALT DROPLETS ments with the low-plutonium-content salt. A few
AND FROM BEDS OF SOLID PARTICLES similar experiments with protactinium will be per-
BY FLUORINATION formed in the near future.

To study the possibility of volatilizing fissile

A laboratory investigation of the fluorination of
material by fluorinating b e d s of powdered fluorides,
UF^ and P u F ^ from falling droplets of molten fluo-
five laboratory-scale fluorinations were made.
ride was made to determine the feasibility of a
Although this approach results in much l e s s cor-
falling-drop tower.
rosion, the salt sinters at the operating tempera-
With this method, the rate of corrosion of the
ture. For this and other reasons the method had
contactor should be l e s s than that for a fluorinator
little to recommend it, and the work was terminated.
operated batchwise with the salt phase continuous.
The possible recovery of uranium from fluoride
salt by fluorination of solid salt particles was also Volatilization of Uranium from Falling Droplets
investigated in the laboratory. Compared with
bubbling fluorine through molten fluoride, this The t e s t s to study the volatilization of uranium
method also would be much l e s s corrosive to the from falling droplets of molten salt used the equip-
container. ment described last year. ' In this equipment, the
Data on uranium volatilization from falling drop- fluorination zone was 3 in. in diameter and a max-
l e t s of molten salt have been extended from the imum of 56 in. high. T h e s e additional t e s t s ex-
work reported p r e v i o u s l y ' to include five salt tended the previous information to cover the s a l t
compositions given in Table 2.3.
Chem. Technol, Div. Ann. Progr, Rept, May 31, 1964,
A correlation equation that fits the data for all
ORNL-3627, pp. 6 1 - 6 4 . five s a l t s , with appropriate changes in constants.

T a b l e 2.3. Correlation Equation Constants for Five Salt Blends

Salt Composition
(mole %) Weight Fraction Initial Liquidus Temperature
Log k
Uranium (°C)
NaF LiF ZrF, UF,

X 10

31.7 31.7 31.7 5 0.132 455 15.78 1.45

48.75 48.75 2.5 0.0542 510 21.58 2.84

48.0 48.0 4 0.0843 515 21.58 2.84

45.5 45.5 9 0.1733 550 28.79 4.35

37.4 56.1 6.5 0.119 630 40.25 7.16


h a s been developed. The equation follows, and t h e droplets are 200 fj. or l e s s in diameter. The
the values of the constants are also listed in temperature must b e about 600°C and the fluorina-
Table 2.3: tion column 5 to 6 ft long.

1 1 kt Volatilization of Plutonium from Falling Droplets

D^ The plutonium falling-drop fluorination t e s t s were
identical to the uranium t e s t s except that glove
where boxes enclosed the top and bottom of the fluorina-
D = droplet diameter, fi, tion column and the fluorination section was only
52 in. long and 3 in. in diameter. Table 2.4 sum-
T = fluorination temperature, °K,
marizes the results of five falling-drop fluorina-
t = fluorination time, s e c , tions of s a l t s 50-50 mole % in NaF and Z r F ^ ,
C o,u = original uranium concentration,' wt fraction,' containing either 2.58 or 0.026 mg of plutonium,
a s P u F , per gram of salt. Note that plutonium
C = final uranium concentration, wt fraction,
removals are much l e s s dependent on fluorination
a.,k = empirical c o n s t a n t s . temperature and drop diameter than are uranium
All data indicate that a spray-fluorination proc- removals. Also, the fraction of the plutonium
e s s for recovering uranium from molten fluoride removed i s independent of the initial plutonium
salt can remove more than 99% of the uranium when concentration.

T a b l e 2.4. Results of Fluorination of Plutonium from F a l l i n g Droplets of Molten Salt

Initial salt: 50-50 mole % N a F - Z r F , -I- P u F

4 3

Average Run Plutoniu m Concentration

Plutonium Fluorination
Temperature Sire Range (mg/g) Removal Time
(°C) Initial Final (%) (sec)

506 63-88 2.58 1.232 52.2 7.3

88-105 1,45 43.8 4.5
105-12 5 1.58 38.7 3.2
125-149 1.09 57.8 2.2
149-177 1.34 48.2 1.6

543 53-63 2.58 0.79 69.5 12.4

63-88 1.19 53.8 7.3
88-105 1.56 39.3 4.5
105-125 2.04 21.0 3.2

609 53-63 2.58 0.45 82.7 12.4

63-88 0.50 80.4 7.3
105-125 0.93 64.0 3.2
125-149 1.06 58.8 2.2

640 53-63 2.58 0.33 87.4 12.4

63-88 0.32 87.6 7.3
88-105 0.39 85.0 4.5
105-125 0.48 81.3 3.2

624 63-88 0.026 0.0156 40.0 7.3

88-105 0.0055 79.0 4.5
105-125 0.0088 66.0 3.2

The plutonium-volatilization data were s u c c e s s - ORNL DWG 65 9710

fully correlated by u s e of an equation developed
by Skelland and Wellek'" for spray extraction:
dC A
— = -k.-C ,
dt dV 60

where k, is calculated from the Sherwood number,

C = average plutonium concentration in the drop- TESTS WITH EQUIMOLAR NaF ZrF^ AT INITIAL
let, mg/g, or any convenient units, 3 20 Pu CONTENTS ( m q / q of s o l t )
O 2 58
A = droplet surface area, cm ^, A 0 026
\ [ i I L^__
V = droplet volume, cm^, and 500 550 600 650
t - fluorination time, s e c
The diffusion coefficients used in the Sherwood Fig. 2.4. V o l a t i l i z a t i o n of Plutonium from 97-/:i Drops
number were estimated by the Stokes-Einstein of Molten NaF-ZrF. Falling Through a 52-in.-High
e q u a t i o n . ' ' All other required data are from Cohen Fluorination Column.
A typical plot of the experimental data vs the Volatilization of Uranium from Static Beds of
equation-predicted line is shown m F i g . 2.4. From Solid Particles
this plot we can calculate the length of column
required to volatilize more than 99% of the plutonium The particles for use in the solid-salt fluori-
from a drop 97 ^ in diameter. At 650°C, a column nations were first formed by feeding pulverized,
20 ft long IS required. Although this is a much sieved salt through a helium-filled tower heated
longer column than would be required for similar above the liquidus temperature of the s a l t . The
uranium volatilization, it still is a practical length molten droplets were then frozen, collected, and
for a production-scale plant. resieved for use in t e s t s of the idea that uranium
could be volatilized from beds of solid p a r t i c l e s .
The fluorinations were conducted in V-in.-OD
A. H. P . Skelland and R. M. Wellek, A.I.Ch.E, J,
10, 491 (1964). nickel U-tubes.
The results of the five fluorinations are given in
^^R. B . Bird, W. E. Stewart, and E . N. Lightfoot,
Transport Phenomena, p. 514, Wiley, New York, 1960. Table 2.5 and indicate that very good uranium
' ^ S . I. Cohen, W. D. P o w e r s , and N. D. Greene, A removals are possible m 2 to 5 hr. Unfortunately,
Physical Property Summary for ANP Fluoride Mixtures, each time that good uranium removal was attained,
ORNL-2150 (Aug. 2 3 , 1956).
the powder sintered into a hard, porous m a s s .

Table 2.5. Results of Static-Bed Fluorinations

Initial s a l t 8.4 wt 7o uranium, as U F , in 50-50 mole % N a F - Z r F

Fluorination Size Range Uranium

Time Condition of Salt
Temperature of P a r t i c l e Removed
(hr) After Run
(°C) (M) (%)

400 105-125 2.0 Nil Free flowing

452 53-63 2.0 98.5 Sintered

473 63-88 3.0 98.2 Sintered

474 125-149 2.0 99.1 Sintered

480 63-88 5.0 99.3 Sintered


2.4 FLUIDIZED-BED VOLATILITY PROCESS Stainless-steel-clad fuel will be treated with

DEVELOPMENT FOR STAINLESS-STEEL- HF-O^ at about 600°C to convert the cladding to
AND ZIRCALOY-CLAD UOj FUELS the oxides of s t a i n l e s s s t e e l .
Pulverization of the UO2 fuel previously declad
This study is intended to complement chemical- with HCl can be accomplished by reaction with
process development work being done at Argonne oxygen to form U 0 and PuO , or H F - 0 can be
and Brookhaven National Laboratories on a used to form U F ^ , P u F ^ , and UO^F^. When HF-O^
fluidized-bed method of recovering uranium and is used for removing either zirconium or s t a i n l e s s
plutonium from zircaloy- or s t a i n l e s s - s t e e l - c l a d , s t e e l cladding, pulverization is done by continued
low-enrichment UO power-reactor fuels. Work at treatment with HF-O to again form U F , P u F ,
Argonne is primarily on the HCl process for and U O ^ F ^ .
zirconium-base fuels, while Brookhaven personnel The oxides or fluorides of uranium and plutonium
are concentrating on studying operational problems are then converted to the volatile hexafluorides by
related to the H F - 0 process at an engineering reaction with fluorine in the same fluidized bed of
s c a l e . While the intersite goal is the development alumina. The UF and P u F are subsequently
of the overall technology for the volatility collected in a cold trap. Methods for separating
processing of low-enrichment fuels, ORNL's and purifying the two products are still tentative.
principal contribution will be a pilot-plant-scale The ANL reference flowsheet, dated January 1,
demonstration of the process with highly irradiated 1965, c a l l s for decomposing P u F to P u F at
fuels. T h e s e laboratory studies are specifically 350°C in the absence of fluorine, followed by
in support of the latter effort. distillation of the U F . Beds of NaF and MgF^
Bench-scale laboratory s t u d i e s of the proposed 6 2

fluidized-bed volatility process have been directed would be used for further purification.
mainly toward observing the important chemical The use of BrF is also being considered as a
effects and making small-scale process runs with way of separating the plutonium from the uranium
both uranium and plutonium present, in addition to early in the p r o c e s s . Gaseous BrF would react
cladding materials. Most of the preliminary work with uranium compounds to form U F without
has been carried out in a V,-in,-diam fluidized-bed reacting with the plutonium. The latter would be
unit whose s i z e is especially convenient for the removed by treatment with fluorine. Details of
plutonium t e s t s . Work on the H F - 0 decladding this process are being explored at ANL.
method for s t a i n l e s s s t e e l has been emphasized to
d a t e . Future work will extend to u s e of H F - 0 in Mini-Test Fluidized-Bed Unit
Zircaloy runs and to u s e of HCl a s a Zircaloy
decladding method. Also to be studied are factors A s m a l l - s c a l e , disposable fluidized-bed (mini-
affecting plutonium and uranium retention and bed t e s t ) unit c o n s i s t s simply of a specially machined
sintering or caking in the decladding-fluorination 30° cone bottom welded to V,-in.-OD nickel tubing
reactor. (Fig. 2.5), No disengaging section has been found
n e c e s s a r y with this unit if linear velocities of
0.75 f t / s e c or l e s s are employed. The normal
Description of Fluidized-Bed Volatility Processes
alumina bed loading is about 5 g. A principal
In fluidized-bed volatility p r o c e s s e s the de- advantage in its use is that the entire bed can be
cladding, oxidation, and pulverization are conducted taken a s a sample, thus avoiding the sampling
in a fluidized bed of alumina. The bed a c t s mainly uncertainties present in handling larger q u a n t i t i e s .
a s a medium for removing the heat derived from the
various chemical r e a c t i o n s .
Zircaloy-clad fuel will be treated with gaseous Oxidation of Zircaloy-2 by HF-O, Mixtures
HCl at about 400°C to convert the cladding to
volatile ZrCl , or the cladding will be oxidized to Boat t e s t s with Zircaloy-2 specimens (wafers
ZrO^ by reaction with g a s e o u s HF-O^ at 600°C. 0,56 in. in diameter and 0,125 in, thick) gave
In the HCl c a s e , the ZrCl will be converted to oxidation rates of 32 to 41 mils/hr at 600°C with
ZrO by reaction with steam in a separate fluidized 40% H F - O ^ , At 650°C the rate decreased to 18
bed of alumina or sand. m i l s / h r . At 600°C with 67% H F - 0 the rate was

PHOTO 6 9 5 6 0
t e s t s indicated a much greater degree of degradation
of the s t a i n l e s s s t e e l oxides than of the ZrO^.

Uranium Volatilization Tests in Mini-Test Unit

Preliminary results with the mini-test fluidized-

bed unit indicate that some loss of fluidization
occurs at UO concentrations in the alumina bed
(Norton RR, 90 to 120 grade) of only 10%. This
effect is promoted by having present an equal
weight of s t a i n l e s s s t e e l oxides formed by the
HF-0 decladding p r o c e s s . The effect is most
prevalent when 100% fluorine is introduced at
temperatures above 500°C; the effect can be
minimized by using a mixture of fluorine and
nitrogen (about 1% fluorine) for 15 to 30 min at the
beginning of a run in order to condition or convert
part of the oxide to the fluoride form.

Plutonium Volatilization in Mini-Test Unit

(tilM'>r'T*f't'l»l'Mt' fM'CI'l
1 2 3
T e s t s similar to the uranium runs were carried
out with PuO additions to alumina in the mini-test
fluidized-bed unit (Fig. 2.5). A s e r i e s of four 1-hr
Fig. 2,5. Miniature Fluidized-Bed Reactor, fluorination t e s t s were made using 5 g of alumina
(Norton RR, 120 mesh) and a fluidization gas
velocity of 0.5 fps. Starting with about 125 mg of
PuO in the bed (about 2% plutonium), the
retentions at temperatures of 450, 500, 550, and
33 mils/hr. X-ray-diffraction t e s t s on the product
600°C were, respectively, 0.4, 0.36, 4,5, and 48%
showed the presence of both ZrO and ZrOF , with
of the initial amount. Corresponding material
wet-chemical a n a l y s e s indicating a predominance
balances were 103, 101, 117, and 134%, A
of the latter. The product oxide or oxyfluoride
subsequent rerun of the 600°C t e s t , in which
seemed to have considerable strength, indicating
cross-contamination was indicated, gave a retention
that little breakup would occur in fluidized-bed
value of 18,8%. The noticeable increase in
plutonium retention on the bed with increasing
temperature indicates appreciable diffusion of
Form of "Stainless S t e e l " Oxides Resulting plutonium into the alumina particles at tempera-
from HF-O2 Decladding Step tures of about 550°C and above. Two additional
t e s t s at 550°C, in which the effect of time on
Analyses have indicated that the fluoride content plutonium retention was studied, indicated that
of type 304L s t a i n l e s s s t e e l oxide material, re- retention at that temperature reaches a relatively
sulting from HF-O^ treatment at 550 to 650°C, constant value and that removal of the plutonium
varies from 1 to 5%, dependent on the process time. is not limited by the amount of fluorine gas
The conversion from oxide to fluoride is relatively employed. The l o s s e s incurred in t e s t s of 30-min
slow due to unfavorable thermodynamics rather and 2-hr duration were, respectively, 4.88 and
than to k i n e t i c s . The opposite is true when t h e s e 4.15%. Modifications of certain conditions, such
oxides are contacted with fluorine at 600°C. Here, as increasing the particle s i z e of the alumina from
rapid conversion and physical degradation occur 120 to 90 mesh and preconditioning the 120-mesh
along with volatilization of a large fraction of the alumina with fluorine at 500°C for 30 min prior to
chromium (60 to 80%). Comparative fluorination P u F volatilization, caused no significant changes

in plutonium retention compared with results of converted to ZrO by reaction with steam in a
previous t e s t s . In the first c a s e , with 90-mesh separate fluidized bed of alumina or s a n d . Fol-
alumina at 500°C for 1 hr, 0.40% of the plutonium lowing pulverization of the UO pellets by reaction
was retained. In the second, where preconditioning with oxygen (HF-O in the c a s e of s t a i n l e s s s t e e l ) ,
was employed, the retention was 0.56 wt %. the uranium and plutonium compounds will be
Further t e s t s of this type are planned to examine converted to their hexafluorides by reaction with
the many variables of interest in the proposed fluorine. The method for separating U F , P u F ,
fluidized-bed volatility pilot plant. T h e s e will and volatile fission products is not definite at
include the effect of uranium, cladding materials, present, but thermal decomposition of P u F to
and fission products on the completeness of solid P u F , followed by distillation of the U F
uranium and plutonium volatilization. Process and volatile fission product fluorides, is presently
variables, such a s grade of alumina, bed recycle, favored. A possibility e x i s t s that BrF vapor will
temperature, and flow conditions, will also be be used instead of fluorine to volatilize the
examined. uranium first. The plutonium would then be con-
verted to P u F by fluorine, (See Sect, 2,4 for more
details of the process.)

2.5 DESIGN OF A FLUIDIZED-BED PILOT The facility is sized to process a charge of

PLANT FOR ZIRCONIUM- AND STAINLESS- about 40 kg of UO and 200 g or more of plutonium,
STEEL-CLAD UO2 POWER-REACTOR depending on nuclear-safety limitations. The
FUELS primary reactor will be 8 in. in inner diameter and
about 12 ft high and will accommodate a fuel
A fluidized-bed fluoride-volatility pilot plant element 60 in. high with a 5.4-in. square cross
( F B V P P ) is being designed for installation in section.
Building 3019. The processing of both Zircaloy-
and s t a i n l e s s - s t e e l - c l a d UO power-reactor fuels Process Engineering
will be studied in the facility at irradiation levels
up to about 15,000 Mwd/ton. Design studies are under way on fuel procure-
The primary goal of s u c h s t u d i e s will be to ment and handling, on safety measures for handling
obtain data needed in the design of a full-scale plutonium compounds, on the solution of accounta-
commercial plant. Some of the specific benefits bility problems expected in the facility, and on
to be derived include the following: a determi- methods for calculating the fission product and
nation of the behavior of actual fission products in transuranium-element content of the fuel to be
the process, particularly during the purification of p r o c e s s e d . The stoichiometry of the process h a s
plutonium; acquisition of experience in operating been coded in FORTRAN for calculation by a
and maintaining a fluidized-bed processing plant digital computer. A study is under way at BMI to
under conditions of intense radioactivity; and assemble and evaluate data available for predicting
demonstration of the safety of the fluidized-bed corrosion of major process equipment planned for
fluoride-volatility method. the pilot plant.
Design studies are under way, and a preliminary In addition to the preparation of the preliminary
flowsheet (combination process and engineering) flowsheet previously mentioned, equipment design
has been prepared (Fig. 2.6). Design of the major and layout studies are under way for the primary
equipment items required in the decladding and reactor (Fig. 2.7), its off-gas filter (Fig. 2.8), and
pulverization-fluorination phases h a s started. This the pyrohydrolyzer. This work is being done at
phase of the plant should be installed during the Oak Ridge Gaseous Diffusion Plant by their
calendar year 1966. Engineering Development Department and Plant
Engineering Division. Fuel-charging equipment i s
being designed and incorporated into the equipment
Process Description
layout, and an instrumentation flowsheet and
The Zircaloy cladding will be removed by reaction instrument requirements are being developed. The
with either HCl or H F - 0 ; the latter reagent is project will be scheduled by the critical-path
preferred for removing s t a i n l e s s s t e e l cladding. method; development of the network diagram is
The ZrCl formed in the Zr-HCl reaction will be under way.
ORNL- DWG 65-8201




Fig, 2,6. Process Flow Diagram of Fluidized-Bed V o l a t i l i t y Pilot Plant.


ORNL-DWG 65-8199



^8-in. SCHED-40 PIPE

COILS FOR - 2 0 X 8 - i n . ECCENTRIC


-8-in. SCHED-40 PIPE,

7 f t 10 in. LONG, NICKEL


F2 INLET ( T Y P I C A L ;




O2, HF-O2, ETC.

F i g . 2.7. Primary Reactor Proposed for FBVPP,


ORNL-DWG 6 5 - 8 2 0 0


2^/4 in. OD X 4 f t ^AlgOg
WifW :


Fig. 2.8. Main Filter for Off-Gas as Proposed from Primary Reactor and Pyrohydrolyzer (FBVPP),

Building Preparation range 100 to 360°C, and the desorption character-

i s t i c s at temperatures up to 500°C. The second
Modifications to improve the containment features part of the program, now under way, c o n s i s t s in
of the Building 3019 penthouse have been designed, further quantitative testing of the more promising
and craft work is about to begin. Air locks at sorbents over a broader temperature range under
entrances are being added, and cracks will be dynamic flow conditions.
s e a l e d to help increase the difference between Static Sorption and Desorption T e s t s , — The
atmospheric pressure and that inside the penthouse, equipment for the semiquantitative s t a t i c t e s t s
c o n s i s t e d of a 3-in.-diam, 15-in.-high reactor
connected to a vacuum pump and a source of
fluorine and nitrogen. The reactor was heated at
the bottom and top. In each t e s t , up to 21 sorbents
(2 mg each) were placed in 1-cm^ dishes arranged
The possibility e x i s t s that a sorption-desorption
and supported 10 in. from the bottom of the reactor.
method can be developed for P u F similar to the
The reactor was then s e a l e d , flushed with nitrogen,
method whereby UF is sorbed by and desorbed
from N a F . A program aimed at evaluating the and filled with fluorine. The bottom of the reactor
sorption and desorption of P u F with various was heated to 600°C for 30 min to convert the
materials in both s t a t i c and dynamic t e s t s is being P u F to the volatile P u F ; during this time the
carried out. sorbents reached a temperature of 360°C, The
Examination in s t a t i c t e s t s of 31 different in- temperature at the t e s t array was then reduced by
organic fluorides and alumina showed that ap- cooling (circulating water through coils about the
preciable P u F sorption occurs only with the reactor) to 125°C in about 30 min, at which temper-
alkali and alkaline-earth elements at temperatures ature the samples remained for 15 min. The e x c e s s
up to 360°C. This agrees generally with results P u F was removed by vacuum and flushing with
of dynamic t e s t s reported p r e v i o u s l y . ' ^ Attempts nitrogen. For t h e s e sorption s t u d i e s , the samples
at desorption at temperatures up to 500°C in the were removed at this point, and the plutonium
s t a t i c t e s t s were unsuccessful. Sodium fluoride at content of each was determined by gamma counting.
350°C was shown in dynamic t e s t s to be highly Calibration with alpha counting was obtained on a
effective in separating P u F from U F when sufficient number of samples to e s t a b l i s h the
sufficient e x c e s s fluorine was present to prevent reliability of the simpler gamma-counting method.
the formation of UF . However, desorption of the Metal a n a l y s i s was used in some c a s e s (KF, R b F ,
plutonium from the N a F does not appear feasible. C s F , and B e F ) to accurately ascertain the amount
The sorption results are not favorable to the of salt involved.
development of a process specifically aimed at For the desorption s t u d i e s , the same equipment
purifying P u F . While sorption of P u F by N a F and test materials were used, except that at the
(and possibly other group I and II elements) is end of the sorption s t e p the temperature of the
feasible, recovery would require aqueous leaching test array was raised to at l e a s t 500°C over a
or dissolution. T h u s , a sorption process could period of about 1 hr, during which, at 100°C
perhaps best be employed as a means of de- intervals, the gas was evacuated down to a pres-
contaminating UF from traces of plutonium. sure of about 1 mm Hg and then replaced with a
fresh atmosphere of fluorine. Cooling and sample
a n a l y s i s were the same a s before.
Exploratory Sorption Program
The sorbents were s e l e c t e d on the b a s i s of
The first part of the exploratory sorption program, availability, c o s t , and previous sorption d a t a . ' ^
which is completed, consisted in semiquantitatively Each sorbent was tested at l e a s t twice for sorption
examining in s t a t i c t e s t s the P u F sorption and desorption; when the data appeared incon-
characteristics of 31 s u b s t a n c e s in the temperature s i s t e n t , additional samples were t e s t e d . For none
of the t e s t materials did desorption appear to have
occurred; therefore, only sorption data appear in
13 Table 2.6. T h e s e data show that significant
Chem. Technol, Div, Ann, Progr. Rept. May 31, complex formation occurs with all the alkali-metal
1964, ORNL-3627, pp. 5 9 - 6 0 .

Table 2.6. Sorption of P u F , on 31 Metal Fluorides and Alumina

Ratio of w e i g h t s , P u / m e t a l fluoride or alumina

Metal Fl uorides^

Li 0.8 Be 1 Al 0.02 Sn 0.005 Cr 0.01 Mn 0.01 Fe 0.001

Na 1.4 Mg 0.03 In 0.01 Pb 0.005 Co 0.001

K 2.2 Ca 0.3 Tl 0.002 Ni 0.005

Rb 0.6 Sr 0.5

Cs 0.6 Ba 0.4 Sc 0.01 Zr 0.01

Y 0.05 Hf 0.015

Cu 0.005 Zn 0.005 Th 0.001

Ag 0.002 Cd 0.002 La 0.02

Ce 0.02



Categorized roughly according to periodic groups.

fluorides and all the alkaline-earth fluorides bed. By comparison, 99.91% of the P u F was
except MgF . For a few of the poorer sorbents, sorbed by NaF at 400°C, with nearly all of it in
cadmium, copper, and nickel, the plutonium content the first section.
appeared to increase after the "sorption s t e p , " A desorption t e s t was run with 148 mg of P u F
and reruns verified the data. T h e s e results are sorbed on a 2-g bed of N a F . P a s s a g e of fluorine
interpreted to mean that at the higher temperatures (150 ml/min) through this bed at 600°C for 2 hr
those sorbents more readily react with P u F . (The gave 'only a 0.32% desorption, a s measured by
PuF adsorbed during the desorption t e s t was trapping on a second N a F bed at 350°C. This
derived from the fluorination of previously deposited irreversibility was confirmed in a special s e t of
P u F or P u F in the upper part of the reactor and dynamic sorption runs where 10 mg of P u F was
from desorption from other samples.) passed through a 2-g NaF bed. At 100, 200, 300,
Dynamic Sorption and Desorption T e s t s . — In 400, 500, and 600°C the respective amounts of
dynamic t e s t s , the sorption capacity of N a F (porous plutonium not trapped were 1.4, 0 , 2 1 , 0,20, 0.07,
type, made from N a F ' H F ) relative to C a F was 0.05, and 0,04%, The trapping effectiveness of
demonstrated. P a s s i n g 135 mg of P u F in a large N a F under the conditions of this t e s t appeared
e x c e s s of fluorine through a 4-g C a F bed (12 to approximately constant in the 400 to 600°C range.
20 mesh) at 100°C resulted in sorption of only The separation of P u F ^ from UF^ was studied in
39%. The concentration of the trapped plutonium in
several t e s t s in which they were passed together
the four separate 1-g sections of this bed was
with e x c e s s fluorine through 2-g beds of N a F
quite uniform, confirming that the trapping efficiency
(12 to 20 mesh) held at 350°C. In three t e s t s ,
was low. In a similar t e s t with C a F but at a
2 starting with 0,5 g of 20% P u F ^ - U F ^ , the respective
higher temperature (400°C), 67% was sorbed and sorptions for P u F on the first bed were 99,7,
there was almost twice a s much plutonium trapped 99.9, and 99.72%. The corresponding l o s s e s of
in the first section as in the l a s t section of the uranium by cosorption were 0.24, 0.34, and 0.21%.

In all the high-temperature t e s t s with N a F , a Evaluation of Pellets Mode at Paducah Gaseous

blue-green coloration was observed after P u F had Diffusion Plant
been sorbed, possibly indicating autodecomposition
of the PuFg to P u F ^ , The difficulty of desorbing P e l l e t s prepared at Paducah during the early
plutonium is possibly related to stabilization of the s t a g e s of the study had satisfactory sorption
tetravalent s t a t e in a P u F -NaF complex. The c h a r a c t e r i s t i c s , but lacked the structural stability
mechanism of P u F sorption probably involves n e c e s s a r y for repeated u s e in fixed-bed sorbers.
formation of a P u F -NaF stoichiometric complex, Preliminary t e s t s of sorption c h a r a c t e r i s t i c s and
followed by reduction to P u F , and the liberation structural stability were made with the gasometric
of fluorine. system described previously,^* With this system,
the quantity of UF^ sorbed by pellets exposed to
pure U F at a pressure of about 100 mm Hg and a
prescribed temperature was e a s i l y measured. The
pellets were exposed to three cycles of sorption-
2.7 SORPTION OF UF^ BY NaF desorption, and the intercyclic changes in sorption
characteristics were measured. Degradation of
In the fluoride-volatility p r o c e s s , fixed beds of pellet structure was observed at the end of the
NaF pellets are used for sorbing U F from a gas test.
stream that may also contain fluorides of corrosion
products, fission products, or both. Data on the Later, pellets were prepared by agglomeration of
rate of removal of UF from gas streams by fixed an NaF-powder—water mixture followed by drying
beds of N a F and data on changes in pellet at 250 to 350°C and sintering at 650 to 700°C.
characteristics after repeated sorption-desorption Preliminary t e s t s of t h e s e newer Paducah pellets
are necessary for design of sorber s y s t e m s . Also, indicated good sorption characteristics and struc-
knowledge of the sorption and desorption character- tural stability. In view of these r e s u l t s , more
i s t i c s of possible contaminants in the gas stream extensive t e s t s were made, a s described below.
is necessary for selecting optimum conditions for Two s e r i e s of t e s t s were conducted to determine
operating the fixed beds of N a F to obtain the the effectiveness of the bed for UF^ removal and
desired purity consistent with reasonable uranium its capacity for UF . In t h e s e t e s t s , a quantity of
l o s s e s . At present, the pelleted N a F used in the the N a F was placed in a 2-in.-diam sorption v e s s e l
process is purchased from the Harshaw Chemical between 5-in.-long entrance and exit s e c t i o n s that
Company for about $ 5 / l b . Since the starting ma- had been filled with V-in.-diam Monel shot. A
terial, powdered N a F , i s available for about flow of 0.0043 g-mole/sec of gas having a UF^
15();/lb, a study of pellet production from N a F concentration of 1.69 mole % was p a s s e d through
powder was carried out on a subcontract b a s i s the v e s s e l . The v e s s e l was kept at a temperature
with the Paducah Gaseous Diffusion Plant to of 100°C, and the concentration of UF^ in the off-
determine whether or not the cheaper material was gas w a s measured. The absorbed U F was then
satisfactory for process u s e . desorbed at 400°C with an F^ flow of 0.5 standard
liter/min. Three sorption-desorption runs were
Experimental determination of U F capacity and
made in each s e r i e s . The time variation of the
stability against degradation of the Paducah pellets
UF concentration in the effluent during the
has been completed. We concluded that the 6
Paducah material will perform a s well a s the second s e r i e s of runs, where an N a F bed depth of
Harshaw material for the present application. The 14.5 cm was used, is shown in F i g , 2,9. The
most important factor in making a choice between effectiveness of the bed for UF^ removal w a s
the two types will be relative cost; initial e s t i -
slightly greater during the second and third runs
mates indicate a cost of l e s s than $ l / l b for the
than during the first. Little difference in sorption
Paducah material, contrasted to $ 5 / l b for the other.
behavior was observed between the last two runs;
this indicates that little degradation of the N a F
The gasometer used for characterizing the N a F
pellets was also used to prepare and measure the 14
S. Katz, Apparatus for the Gasometric Study of
dissociation pressure of the compounds U F - N a F , Solid-Gas Reactions: Sodium Fluoride with Hydrogen
WF^ - N a F , and MoF - N a F , Fluoride and Uranium Hexafluoride, ORNL-3497 (Oct.
6 0
15, 1963).

ORNL-DWG 65-2996A MoFg . 2 N a F , the compounds U F ^ - N a F , WF^ - N a F ,

• CYCLE I, 8 R - 3 2 ~

o CYCLE 3, B R - 3 6 ]
and MoF - N a F were prepared in the gasometer.
T h e s e new complexes were made by first preparing
U F . 2 N a F , W F g - 2 N a F , and M o F g . 2 N a F by
reaction of the metal hexafluoride with thin layers
of dispersed N a F at the respective temperatures of
150, 165, and 185°C. (The sodium fluoride had
previously been dispersed by decomposition of
N a F • 4HF or U F • 2NaF.) The MF^ • 2NaF complex
was then treated with the appropriate metal hexa-
fluoride at 75°C to form the M F ^ - N a F complex.
The equations for the dissociation pressures of the
new complexes are:

log P = 11.06
^ mm

- (3.48 X 10^)/T for U F ^ over U F ^ - N a F ,

log P = 7.44
° mm
80 (20 200
TIME (mm) - (2.13 X 10^)/T for WF^ over WF^ . N a F ,

F i g , 2,9, Effect of Cyclic Sorption-Desorption on Bed log P = 7.29

Capacity for Paducah N a F P e l l e t s and a Comparison ° mm
- (1.83 X 10^)/T for M o F , over MoF - N a F ,
with First-Cycle Capacity of Harshaw NoF Pellets, ^ ' D O

Second series of tests. where P is pressure in mm of mercury, and T is

temperature in °K.
was occurring. The effluent curve for a bed of GAS STREAMS IN FLUORIDE VOLATILITY
pelleted NaF of the Harshaw type having the same PROCESSING
weight a s the bed of Paducah material i s a l s o
shown for comparison. The initial UF^ concentra- Continuous a n a l y s i s of volatile fluorides in gas
tion in the effluent from this bed is lower than that streams is of considerable importance in the
from the Paducah material, however, the final U F ^ operation of the fluoride-volatility p r o c e s s e s .
capacity i s slightly greater. The differences Proper control of the various p h a s e s of t h e s e
between bed capacities and initial rates of UF^ p r o c e s s e s and the evaluation of the equipment
removal are considered negligible for the present require that the concentration of one or more of
application. these fluorides be the primary controllable
After three cycles (BR-32, -34, -36, F i g . 2.9) variable.
the two beds of N a F were examined visually for Severe demands are made on the analytical
pellet degradation. Almost no powdered N a F was instruments used in fluoride-volatility processing.
observed, and the pellets did not appear to have The instrument must measure concentrations in
lost structural stability. Overall, the Paducah multicomponent mixtures of chemically similar
pellets were satisfactory. s p e c i e s , it must withstand the corrosive action of
fluorides, and it must accept remote calibration

Preparation of Complexes of U F , , WF,, or MoF^

with NaF and Measurement of Dissociation ^^Chem. Technol. Div. Ann. Progr, Rept. May 31,
1964, ORNL-3627, p. 52.
^®S. Katz, " U s e of High-Surface-Area Sodium Fluoride
to Prepare M F ^ ' 2 N a F Complexes with Uranium,
Supplementing the previously reported 1 5 , 1 6 0
Tungsten, and Molybdenum Hexafluorides," Inorg.
preparation of UFg.2NaF, WFg-2NaF, and Chem. 3, 1958 (1964).

and repair when used for measuring radioactive A schematic of the transfer system i s shown in
materials. The signal from the instrument should F i g , 2,10. All lines are made of V.- or ^/'-in,
be a simple function of concentration, and it copper tubing. Connections are made with Monel
should not reflect effects of any corrosive action Swagelok fittings. The N a F trap is made of nickel
by the g a s e s . and holds about 1 kg of N a F , The Sanborn pneumatic
The instruments previously used have not all differential-pressure indicator and the Robertshaw-
met the requirements listed above. We are presently Fulton Bourdon-tube gage are made of Monel,
developing an ultraviolet spectroscopic method for Valve 3, attached to the cell by a short length of
monitoring UF , and a column-chromatographic nickel tubing, is a Hoke M-482 Monel valve. All
method for monitoring H F . Both methods fulfill others are Hoke 413 Monel v a l v e s .
most of the requirements for fluoride-volatility
instrumentation. Although other volatile fluorides Before using the cell we completely removed any
have been successfully analyzed in-line by gas traces of moisture by flushing with CIF , and then
chromatography, HF presents a s p e c i a l problem. conditioned the cell with U F at 15 psig and
In laboratory a n a l y s e s it reportedly gives poor 75°C for 16 hr. The transfer system was evacuated
peaks and reacts with the column material. How- with valves 2, 3, 4, and 7 open (Fig, 2.10), flushed
ever, we feel that these effects will be minimized with U F , and then reevacuated, A sample of gas
by repeated automatic sampling, since each sample was then admitted to the cell, its pressure being
will effectively pretreat the column for the next indicated on the g a g e s . We assumed the gas in the
sample. cell to be at room temperature.

Preliminary s t u d i e s have been made to determine The Beer-Lambert law of absorption defines the
the requirements for an ultraviolet spectroscopic "molar absorptivity" of a s u b s t a n c e by the fol-
method for continuously monitoring U F concen- lowing equation:
trations in gas streams. In this study, s t a t i c
samples of U F were analyzed to confirm and
expand b a s i c data about the U F spectrum and to log ^ M = aC^z , (1)
examine the n e c e s s a r y techniques for uv a n a l y s i s .
Two peaks that might be useful in monitoring
applications were investigated; more, work is where
needed before either can be utilized.
/ = light intensity,
A fluidized-bed reactor for the HF-catalyzed
destructive oxidation of s t a i n l e s s s t e e l cladding /g = initial intensity,
material is being built, and it requires instruments
to monitor oxygen, hydrogen fluoride, water vapor, C. = concentration of s p e c i e s i, moles/liter,
and possibly CrO F in the off-gas stream. We
z = path length, cm, and
installed a dual-column chromatograph (Barber-
Colman, s e r i e s 5000) for this purpose. The instru- a *: molar absorptivity (a function of wavelength),
ment is particularly useful for this program be- liters mole~^ c m ~ ' .
c a u s e it can indicate the presence of unexpected
s p e c i e s in the off-gas and also give quantitative Three s e t s of samples of pure U F were analyzed
analyses. for a quantitative measure of its absorptivity at
3686 A. The values of molar absorptivity were
calculated by Eq. (1) from readings of log ^ Jl)
Measurement of Absorptivity of U F , In given by the spectrophotometer, the known concen-
Ultraviolet Region tration of U F , and the known path length of the
c e l l . The value of the molar absorptivity at 3686 A
The U F spectrum was measured with a Cary was calculated to be 5.5 liters mole"* c m " ' . This
spectrophotometer (model 14M). The spectroscopic value was independent of pressure in the range
cell used in this study was made entirely of 10 to 100 mm Hg. We found a previously unreported
nickel; the windows were sapphire, ''/ in, in peak for U F at 2140 A and determined an approxi-
diameter and / in, thick; and the 0-ring s e a l was mate value for the molar absorptivity — about
made of Teflon, 3610 +1000 liters m o l e " 1 c m " ' .

ORNL-DWG 6 4 - 7 6 0 3 A

( 0 - 2 0 0 0 mm OF Hg )
( 0 - 1 0 0 mm OF Hg)


F i g . 2.10, System for Transferring Gas Sample into Spectroscopic C e l l ,

2.9 VAPOR-LIQUID EQUILIBRIA OF UF^-NbFj determination of critical constants has been made.
SYSTEM The experimental work still to be done includes
the remainder of the isobaric data and confirmation
During the recovery of uranium from spent nuclear of the critical c o n s t a n t s . Design, fabrication, and
fuels by a fluoride-volatility p r o c e s s , the UF is operation of equipment suitable for precise work
contaminated by other volatile metal fluorides. with t h e s e materials in the temperature and pres-
Distillation is being considered as a method for sure region shown have been very difficult and
separating the U F from those impurities; but, to time consuming.
design the distillation system, more must be known
about the vapor-liquid equilibria of the impurities
Experimental Work
and the U F , One such fluoride impurity is NbF ,
In high-bumup fuels, it would be produced in Two equilibrium stills were constructed and
significant quantities as a fission product, and it operated. The first was of an isothermal design
would also present a problem when the fuel ele- that was used to determine the total-pressure-vs-
ment contained niobium initially. Hence, an in- liquid-composition relationship at 150°C (Fig.
vestigation is under way to determine the vapor- 2.11). The data were fitted with the Scatchard-
liquid equilibria of U F -NbF in the pressure Hamer e q u a t i o n s : ' '^
region 0 to 12 atm at 80 to 250°C.
The isothermal determination of total vapor logy^ =/l +2z^[{V^B/V^)-A]z
pressure for the entire composition range is com-
plete, and experimental data points on vapor logy^ = 5 +2z^[iV^A/V^)-B]z
composition vs liquid composition at constant
pressure have been determined at a few scattered 17
points in the composition range. Preliminary G. Scatchard and W. J . Hamer, / . Am. Chem, Soc.
57, 1 8 0 5 - 9 (1935).

ORNL-DWG 65-9711
(4 ORNL-DWG 65-97(2
A^-- • - ^


B (0

§ 8
— " -
7 ^ - HAMER EQUATION A = 0 6 , B =0 7 5 - ^ 06 -
^• c 05
? 4 o
^ : o
02 03 04 05 06 07 08 09 (0 1 03 SYSTEM FOLLOWS D A L T O N ' S LAW
^UP (mole fraction of UFg in liquid) B THEORETICAL RELATIONSHIP,
1 1 1 1
Fig. 2.11, Comparison of Isothermal Data for U F , -
NbFj System with Scatchard-Hamer Equation Using
0 01 02 03 04 05 06 07 08 09 (0
Dalton's Law Assumption.
.r^jp (mole fraction of UFg in liquid )

Fig, 2.12, Vapor-Liquid Equilibria of UF,-NbFc

where System at 1 5 0 ° C ,

Yj - activity coefficient of / ,
^ = ^ / [ x , +(7/F^)x^] , critical temperature to be between 480 and 520°C
and the critical pressure to be between 40 and
V • - molar volume of / , 50 atm.
In support of the above experiments, an apparatus
X. = mole fraction of i, for preparing and purifying about 100 g of NbF^
A,B- c o n s t a n t s , which were varied to produce per day was built. The NbF^ is prepared by
fit of data, and assuming that contacting niobium metal shavings with fluorine at
400°C. The resulting product is purified by
P=yxP° + y x P ° , distillation in an 18-in,-high packed column and
'\ 1 1 '2 2 2 '
collected overhead.
where P is total pressure and P ? is vapor pressure
of i at 150°C, With uranium hexafluoride a s
component 1, the values of A and B that gave the 2.10 PHASE EQUILIBRIA STUDIES PERTINENT
best fit were: A = 0,6; B = 0,75, This relationship TO FLUORIDE-VOLATILITY PROCESSING OF
was used to calculate (assuming that Dalton's FUELS OR MIXTURES CONTAINING
law held) the 150°C isotherm shown in F i g . 2.12. CHROMIUM^^
As soon a s the critical constants of NbF can be
determined, the x-y diagram will be recalculated by As part of an attempt to develop a fluoride-
u s e of the Lewis and Randall fugacity rule, which volatility process for s t a i n l e s s - s t e e l - c l a d or -base
provides better accuracy for vapor mixtures at reactor fuels, phase equilibria s t u d i e s were con-
pressure above 1 atm. The second equilibrium tinued in search of a molten fluoride mixture
still i s of an isobaric design and i s now being capable of dissolving the fluorides of the constitu-
used to determine both the x-y diagram and the ents of s t a i n l e s s s t e e l . Also, a brief study of the
boiling-point curve at 6 atm. phase and chemical equilibria in the system N a F -
The critical constants of NbF are being ob-
tained in an apparatus that measures the constant-
volume pressure-temperature relationship up to 18 Work performed by B . J . Sturm and R. E . Thoma,
650°C and 2000 psi. Preliminary data indicate the Reactor Chemistry Division.

CrF was made. Data from the latter study might 1000
also be useful in relation to a possible solvent
system for " s t a i n l e s s s t e e l " fluoride, but, probably
more importantly, they may be pertinent to defining HOO
the capacity of an NaF sorption trap for volatile ir
chromium fluorides produced in the fluorination of f'OO
chromium-bearing melts. hi
n 6CXJ
Work on the development of a fluidized-bed III

process for fuels that contain or are clad in 500

s t a i n l e s s s t e e l is described in Sect. 2,4, and other
studies of the chromium fluoride problem are LiF 20 40 60 80 FeFj
described in Sect. 18, "Chemical P r o c e s s i n g for mole {%)
the Molten-Salt Reactor Experiment."
Fig. 2.13. The System L i F - F e F , .
As reported last year, molten K F - Z r F was an
acceptable solvent for s t a i n l e s s steel fluoride, but
the reaction between s t a i n l e s s s t e e l and HF was
quite slow. During the current report period, the phase diagram. ^^ The use of K F - L i F mixtures is
phase diagram of the L i F - F e F system was deter- suggested, therefore, as a solvent convertible into
mined, and the effect of CrF was examined across multicomponent, low-melting-point KF-LiF-NiF -
the composition section 3 L i F • CrF -3KF • CrF . F e F ^ - C r F j - C r F ^ solutions. The solubility of
Although the results were encouraging, work on CrF in the K F - L i F eutectic is expected to be low
solvents for s t a i n l e s s s t e e l fluoride will be at processing temperatures of about 600°C because
deemphasized in favor of studies more directly of the high liquidus temperatures that prevail
applicable to a fluidized-bed fluoride-volatility across the composition section 3 L i F • CrF -
process. 3 K F . C r F g (Fig, 2.14), Chromium is a minor
Some study of phase equilibria in the system constituent of the alloy, however, and i t s solubility
NaF-CrF was made as a possible guide to deter- is expected to be adequate.
mining the limit of the melting or sintering temper-
ature imposed when using N a F beds for sorbing the
Separation of Chromium Fluoride from U F .
volatile chromium fluorides. Further work on
Adsorbed by NaF
chromium fluorides will be restricted to studies
connected with the development of a fluidized-bed
Preliminary examinations were made of phase
and chemical equilibria in the system NaF-CrF
because of i t s relation to possible solvent systems
Solvents for Use with Fuels That Have a (now deemphasized) a s well as to the separation
Stainless Steel Matrix of UF from N a F sorber b e d s , A p h a s e diagram of
the system NaF-CrF , constructed from visual and
Development of a molten-salt solvent for thermal a n a l y s i s experiments, is shown in F i g . 2.15.
processing s t a i n l e s s s t e e l fuels would most
probably be based on use of one or more of the
alkali-metal fluorides to act as melting-temperature
d e p r e s s a n t s . The phase diagram of the L i F - F e F 2.11 VOLATILIZATION OF UFg FROM
system (Fig. 2.13) was determined by direct visual SOL-GEL-DERIVED ThOj-UOj
observation and thermal a n a l y s i s of molten
mixtures. The diagram is characterized by a The possibility of recovering uranium from sol-
eutectic mixture containing 38.5 mole % F e F at gel-derived thoria-urania by direct fluorination at
620°C. In the LiF-NiF^ system, F e F ^ and N i F ^ high temperature was investigated in the laboratory
were found to be nearly equal in their effects on during the past year. This work was aimed at
the liquidus.
To depress the melting point further, KF would
'^R. E. Thoma (ed.). Phase Diagrams of Nuclear
be particularly effective, a s noted in the K F - L i F Reactor Materials, p, 15, ORNL-2548 (Nov. 6, 1959).

ORNL-DWG 65-97J4

• ^

1 / 1 1

1 /

1 1 1 1
F3 \ s .
3 L i F - C rFj + LiF • 2KF • C'F3 u


Li,CrF, K,CrF, (mole%)

Fig. 2.14. Section of the System K F - L i F - C r F j : The Quasi-Binary System 3Li F'CrFj-SKF-CrFj.

developing a nonaqueous process for recovering Their diameters ranged from 149 to 210 /z, and the
fissile material from spent fuel elements. fluorination temperatures were between 480 and
Three fluorinations were conducted with thoria- 660°C.
urania (3.9 wt % uranium) sol-gel microspheres. The t e s t s showed that fluorination of uranium
from thoria-urania containing small amounts of
urania will be very difficult unless the particles
ORNL-DWG 65-97(5 are l e s s than 10 \i. in diameter. Also, better
s u c c e s s would be expected if the uranium content
1400 .---' were higher. Sintering may not be a serious
! ' ^' problem s i n c e it was nearly absent in one
1 •
.> / successful fluorination.
(200 — 1 x'
Future work will include fluorination of smaller
~ ((00 / 1; microspheres and urania-thoria of different r a t i o s .
/rS - //
/ 1
/ 1

E (000 ,' \ 1

^/ \ 1

S 900 - •

V ' 1
Uranium Not Readily Removed by Fluorinating
800 \h
T h O j - U O , Microspheres

700 Fluorinations were performed on 3- to 4-g samples

in V-in,-OD nickel U-tubes, After fluorination,
0 4
ro the samples were analyzed for thorium, uranium,
500 D
0 nickel, and fluoride.
400 The results of the three runs performed to date
20 40 60 80 CrF,
mole (%) are summarized in Table 2.7. Fluorination of the
microspheres proceeds quite slowly due to their
Fig. 2.15. The System NaF-CrFj. very low porosity and the large amount of thorium.

T a b l e 2.7. Results of Sol-Gel Microsphere Fluorinations

Initial thoria-urania composition: 3.9 wt % uranium

Fluorination Microsphere Uranium Fluorine Condition

Temperature Size Range lime Removed Penetration After
(°C) (/:i) ^'""^ (%) (fJL) Run

480 149-177 3.5 Little ~0 Same a s initial

647 149-177 3.5 15 4 Sintered

657 177-210 1.0 7 2.5 Slightly sintered

both of which impede the diffusion of fluorine into duction of an inert gas just below the surface of
the microspheres. For example, c a l c u l a t i o n s , the melt. Results of the electrochemical studies
based on the amount of uranium removed, showed and the first five of the inert-gas studies were
that the depth of penetration was only about 4 fi in summarized last year, ^' and the electrochemical
the 3.5-hr mn, and about 2.5 ji in the 1-hr run. studies were reported.^^'^^ This period, the
inert-gas studies were finished and a letter report
Experiments with coupons and inert g a s e s in
4-in.-diam Hastelloy B containers showed that
The relatively high rate of corrosion of both the
conditions could be obtained to definitely reduce
hydrofluorination and fluorination v e s s e l s i s a
corrosion. In individual 2-in.-diam containers
serious fault of the molten-salt fluoride-volatility
made of INOR-8, however, only slight protection
p r o c e s s . Studies related to the molten-salt process
was obtained. Argon and helium were more ef-
for aluminum-base fuels have been made at
fective than nitrogen, hydrogen, or a mixture of
laboratory and engineering s c a l e s here, and at
hydrogen and argon. A melt of 52-37-11 mole %
a laboratory s c a l e at Battelle Memorial Institute
N a F - L i F - Z r F , was used at 650°C for all t e s t s
(BMI). Results are mentioned in Sect. 2.1 of this
b e c a u s e of its exceptionally high corrosivity.
report. This section (2.12) is concerned with the
concluding experiments of a s e r i e s made at BMI Results of the study showed that introduction of
to determine the effectiveness of introducing an an inert gas below the surface of the salt melt
inert gas just below the surface of the salt melt in may have useful possibilities for reducing rates of
minimizing corrosion during hydrofluorination. attack by H F and molten fluorides on reaction
As a part of process development, early studies vessels. The optimum procedures for its u s e ,
(mostly on a laboratory s c a l e at BMI) were aimed however, have not been established thus far.
at determining the corrosiveness of different s a l t Apparently an engineering type of study would be
melts to various metals. Much effort has also been necessary to determine the effect of v e s s e l design,
devoted to the determination of corrosion during
engineering- or pilot-plant-scale^° t e s t s of the
^^Chem. Technol. Div. Ann. Progr, Rept. May 31,
1964, ORNL-362 7, p. 70,
During the last two years, two possible ways of ^ ^ P . D. Miller, L. K. Matson, and E . F . Stephan,
minimizing corrosion during hydrofluorination were Cathodic Protection of the Hydrofluorinator, BMI-X-312
(Sept. 25, 1964).
studied: electrochemical protection and intro-
L. K. Matson et al., Cathodic Protection in Fused
Fluoride Salts at 1200°F. presented at National
Association of Corrosion Engineers Conference, Mar.
E. L. Youngblood et al,. Corrosion of the Volatility 1 5 - 1 9 , 1965.
Pilot Plant INOR-8 Hydrofluorinator and Nickel 201 ^ ' ' P . D. Miller, L. K. Matson, and E . F . Stephan,
Fluorinator During Forty Fuel-Processing Runs with Corrosion Protection of the Hydrofluorinator by an Inert
Zirconium-Uranium Alloy, ORNL-3623 (March 1965). Gas Sparge, BMI-X-329 (Feb. 5, 1965).

gas flow pattern, and gas flow rate on rates of Details of equipment, test procedures, and results
attack. Because of the reorientation of the overall have been reported in ref. 24, available upon
program toward fluidized-bed p r o c e s s e s and the request to the Director, Chemical Technology
indifferent results of the experiments, corrosion Division, ORNL.
s t u d i e s as related to molten-salt p r o c e s s e s were
3. Waste Treatment and Disposal

The waste treatment and disposal development 3.1 CONVERSION OF HIGH-LEVEL

program was designed to develop a comprehensive RADIOACTIVE WASTES TO SOLIDS
waste management system for nuclear w a s t e s ,
including their final disposal, and to estimate Three solidification processes have been devel-
the cost of this operation. The effective e c o - oped in the ORNL program using nonradioactive
nomic management of radioactive effluents is a simulated w a s t e : the pot calcination process
prerequisite to the natural growth of a nuclear (Potcal), the rising-level process (RL-Potglass),
power industry. The proposed system is shown in and the continuous potglass process (Con-Pot-
F i g . 3 . 1 . High-level radioactive w a s t e s (HLW), g l a s s ) . ' " ^ The P o t c a l and R L - P o t g l a s s p r o c e s s e s
which contain nearly all the fission products, are will be tested with radioactive waste in a pilot
converted to s o l i d s . Intermediate-level radio- plant at Hanford. ( " G l a s s " includes true g l a s s e s
active w a s t e s (ILW), characterized by their gen- and rocklike crystalline products.) The primary
erally high salt content, are incorporated in concept of the pot processes is that an evaporator
asphalt; any contained water is simultaneously and a solidification unit should be used together
volatilized. Low-level radioactive w a s t e s (LLW)
are treated to remove the radioactivity, and the
decontaminated water is discharged to the envi- F . L. Culler, Jr., et al., Chem. Technol. Div. Ann.
Progt. Rept. May 31, 1964, ORNL-3627.
ronment. The recovered solids are combined
with the ILW. Both the HLW and ILW solid ^M. E. Whatley et al.. Unit Operations Sect. Monthly
Progr. Rept. May 1964, ORNL-TM-937.
products can be shipped to an ultimate disposal Waste Treatment and Disposal Semiann. Progr. Rept.
site. July-December 1964, ORNL-TM-108] (in p r e s s ) .

ORNL DWG 6 4 -8256A



99.997c, FP's < 0.017,, FP's TRACE FP's

r f



-10"^ g / c m 2 . d a y "^ 10~^ g /cm^-day
c;ni in=; Awn FPV

' '

F i g . 3 . 1 . O R N L Waste Management Program.


in a loop. Advantages of this system have been Based on t e s t s performed at ORNL, either
described previously and a flowsheet given, ^ induction or r e s i s t a n c e heating can be used for
illustrating the interchangeability of the three the Potcal and P o t g l a s s p r o c e s s e s . Resistance
pot p r o c e s s e s in a fixed loop system. heating appears to offer fewer operating problems
The development of the Potcal process has been and is recommended for the pot s y s t e m s .
successfully completed (total of 80 experiments). Pot Calcination Process. — In the P o t c a l proc-
In the final test with FTW-65-X3 (27 gal of waste e s s , the concentrated aqueous waste is fed to
per ton of uranium) Purex w a s t e , the process the calcination pot continuously, and the aqueous
was successfully demonstrated using an 8-in.-diam level in the pot is controlled by a proportional
pot and a 15-in. susceptor as a heater in an thermocouple probe. Solids build up on the walls
induction furnace. Previous t e s t s with 16-in.- radially until the pot is filled. The aqueous feed
diam pots confirmed the mathematical model for is then stopped, and the solids are heated to
predicting filling rate as a function of diameter 900°C to calcine them. The Potcal process has
and showed the impracticability of using the pot been successfully demonstrated previously with
wall a s the susceptor. In this c a s e the pot walls simulated Purex, TBP-25, and Darex w a s t e s in
failed in two t e s t s due to inability to maintain 88-in.-high pots having 6- and 8-in. diameters,
uniform heating. Based on t e s t s at ORNL with using r e s i s t a n c e h e a t i n g . ^ " ^
both heating s y s t e m s , r e s i s t a n c e heating is the Two Potcal t e s t s were made with a 16-in.-diam
preferred method. pot as the susceptor in the 150-kw furnace^
Development of the R L - P o t g l a s s process is (Tables 3.1 and 3.2). T h e s e two t e s t s resulted in
continuing in both laboratory and engineering- pot failures from local overheating of the calciner
scale studies. Principal objectives are to (1) because of the difficulties in controlling the in-
develop " n e u t r a l " glass-ceramic phosphate or duction flux and heat transfer in the nonuniformly
s i l i c a t e formulations for Purex w a s t e s (with or insulated inductive susceptor. After the failure
without sulfate) which have low melting points of the two 16-in.-diam pots, when used a s direct
and low corrosivities and (2) to more precisely s u s c e p t o r s , a 15-in. susceptor was installed in
define operating conditions in full-scale equip- the induction furnace to act a s a heat radiator.
ment, including control of dusting and monitoring Eight-inch pots were used with this susceptor.
of phase levels in the pot. Further development One successful calcination test and two rising-
of the Con-Potglass process h a s been terminated level t e s t s (see below) were made with the 15-in.
for the immediate future. susceptor and 8-in.-diam pots. The average feed
rate in the P o t c a l test was 8.8 l i t e r s / h r . The
Engineering Studies bulk density of the solids obtained from induction
heating of the calciners was 1.10 to 1.42 g/cm^
for the calcined s o l i d s . T h e s e are comparable
The engineering development program was prin-
to the bulk d e n s i t i e s of solids obtained with the
cipally concerned with t e s t i n g induction heating
resistance-heated calciner.
for the calcining and glass-forming systems con-
currently with continued development of the RL- One important result from the 16-in.-diam t e s t s
P o t g l a s s and Con-Potglass p r o c e s s e s . The was the confirmation of the mathematical model
furnace consisted of a 15-in.-diam susceptor for solids deposition. Figure 3.2 shows good agree-
contained in 16-in.-diam c o i l s . Three Potcal ment between the predicted solids filling rate and
t e s t s were made which completed the development the filling rate obtained for t e s t R-84, a 16-in. c a l -
of this p r o c e s s . Two successful RL t e s t s were cination t e s t . The average filling rates were 24 and
made using FTW-65-X3 simulated Purex waste 14 l i t e r s / h r in the two t e s t s . The filling rate in
(s=27 gal per ton of uranium processed) and in- the second test was low because the lower coils
duction heating. Further t e s t s are required to of the induction furnace failed midway through
the t e s t .
define operating conditions more closely, using
both sulfate-bearing and nonsulfate types of Purex Rising-Level Potglass Process. — In the RL-
waste a s well a s aluminum w a s t e s . A single Con- P o t g l a s s process the concentrated aqueous feed,
P o t g l a s s run was satisfactory on an operating containing glass-forming additives, is fed directly
b a s i s , but the corrosion rate was too high in into the pot, the liquid is vaporized, and the
s t a i n l e s s s t e e l equipment. solid residue is melted to a glassy solid. The

additives can be added directly to the evaporator simulated waste as well a s the final composition.
or continuously to the feed line entering the pot. The glass-forming additives were added directly
With proper glass formulation the operating tem- to the w a s t e , whereas the calcium nitrate was
perature should be 900°C or lower. As the level added through a separate addition line at a con-
r i s e s , three phases are present: liquid glass at trolled volume ratio. The amount of off-gas dust-
the bottom, a small solid interface, and a small ing and the pressure surges were small during
aqueous supernatant pool. The overhead vapors both t e s t s . The feed rate for these t e s t s varied
are condensed and returned to the evaporator, thus from 8 to 10 liters/hr, equivalent to the w a s t e
preserving the advantages of the loop concept. from processing 2 tons of uranium per day. The
Two rising-level, e s s e n t i a l l y duplicate, t e s t s dense glass formed in t h e s e t e s t s appeared to be
were made in 8-in.-diam by 9-in.-high type 304L satisfactory (Fig. 3.3).
s t a i n l e s s s t e e l pots (Tables 3.1 and 3.2). The Continuous Glass P r o c e s s . — The initial t e s t
purpose of these t e s t s was to determine the of the 20-in. horizontal continuous melter (type
amount of solids entrained into the off-gas, to 304L s t a i n l e s s s t e e l ) was successful from an
determine the magnitude of pressure surging in operational viewpoint, even though the melter
the off-gas line, to determine the approximate failed due to corrosion. The melter was operated
maximum production rate, and to examine the qual- with feed for a total of 80 hr and was held at
ity of glass made by using lithium and phosphate temperature (900 or 700°C) for 250 hr. During
as the major glass additive components. The this time the average feed rate with FTW-65-X3
feed composition in Table 3.2 shows the calcium (27 gal per ton of uranium) was 26 l i t e r s / h r , with
nitrate and glass-forming additives added to the maximum feed rates up to 50 liters/hr (26 l i t e r s / h r

Table 3.1. Pot Tests Using 18-in. Induction Furnace and FTW-65-X3 Simulated Purex Waste

Product Specifications Calcination

Test Average Nitrate
Feed Rate Bulk gal Vol Liquid Waste , , ,, -, in Solids
No. Process 2 Melt Time
(liters/hr) ^.'^"sity per ton of (ppm)
(g/cm ) uranium Vol Solid Waste (hr)

R-83 Potcal-16 24 1.3 2.56 10.6

R-84 Potcal-16 14.6 2.60 10.4 30,000

R-85 P o t c a l 8-in. 1.42 2.64 10.2 10,000

pot susceptor

R-86, R L - P o t g l a s s S-in. 2.13'= 3.1 11.3 10

-87 susceptor

R-88 R L - P o t g l a s s 8-in. 9.9 2.3' 3.33 10.5 12" 200

, f

Equivalent to tons of uranium originally processed.

Pot wall failed; prematurely terminated,
"^Low because of e x c e s s foaming.
Time to reach 900°C after liquid feed terminated.
^Viscosity was 27.5 poises at 850°C m laboratory t e s t s .
Viscosity was 15.1 poises at 900°C in laboratory t e s t s .

is equivalent to the waste from processing 6.1 tons Steel appeared to be promising materials of con-
of uranium per day). The FTW-65-X3 glass (Table struction for the horizontal melter. A new melter
3.2) melts at 750°C and has an acceptable fluidity was fabricated of type 310 s t a i n l e s s s t e e l b e c a u s e
at 850 to 875°C for use in the continuous melter of its commercial availability. The purpose of
(viscosity is 27.5 poises at 850°C and 15.1 this first continuous t e s t was to gather design
poises at 900°C). The run was terminated when operating data, to t e s t the new FTW-65-X3 glass
the v e s s e l failed at the interface level. Of the mix for compatibility, and to obtain corrosion data
corrosion specimens tested in this experiment, on a number of metal specimens.
Corronel 230 and types 446 and 310 s t a i n l e s s The 20-in. continuous melter system c o n s i s t s of
the melter located in a 90-kw Globar furnace, the
feeding equipment, the off-gas handling equipment
(which c o n s i s t s of a condenser and a gas-scrubbing
Table 3.2. Composition of Synthetic Purex Waste column), and a condensate dilution tank ( F i g . 3.4).
and the Glass Formed from It This system is very similar to the 8-in. melter re-
Simulated FTW-65-X3 w a s t e , 27 gal ported previously. The melter (Fig. 3.5) is 20 in. in
per ton of uranium diameter, 40 in. long, and h a s four water-cooled
equally spaced feed l i n e s . The operating melt level
is 3 in. deep. A baffle in the rear of the melter,
Component Concentration which has a l ' / - i n . clearance, keeps the floating
calcine from flowing into the discharge pipe. Fig-
0 riglnal Waste
ure 3.5 shows the melter, without insulation, posi-
HNO3 1.5 M tioned in the Globar furnace. The 90-kw Globar
Fe(N03)3 0.3 M

NKNOj)^ 0.03 M
0RP1L-0WG 6 4 - 7 8 8 0 A
A1(N03)3 0.15 M

Cr(N03)3 0.06 M

Na^SO^ 0.45 M
Addition to Prevent V o l a t i l i z a t i o n of Sulfate
(in Potcal Process)

Ca(N03)^ 0.20 g-mole per liter of feed

Additions to Form Glass 1500

(in R L - P o t g l a s s Process)

1.5 g-moles per liter of feed

LiOH 1.5 g-moles per liter of feed

A1(N03)3 0.5 g-mole per liter of feed

Composition of Glass

N O 3, 0.02 wt %
3+ 500
Fe 5.28 wt %

Al 3 + 8.37 wt %

Na'' 7.65 wt %
SO 21.2 wt %
(00 <50
PO 45.3 wt % RUN TIME (hr)
Li + 3.6 wt' % Fig. 3.2. Actual and Calculated Volumes of Feed vs
2+ 7.2 wt'% T i m e for Potcal T e s t R-84. Calculated volumes were
obtained by use of the radial-deposition model.

PHOTO 61(09

Fig. 3.3. Simulated Purex Glass.

ORNL-DWG 6 4 - 4 6 5 9 A


9 0 - k w G LOBAR




Fig. 3.4. Process Flowsheet of 2-in.-diam Continuous Melter.


F i g . 3.5. 2 0 - l n . Continuous Melter.

furnace has three z o n e s . The hot vapors pass that can be concentrated rapidly in a small
into the updraft condenser. The condensables area of a susceptor, which can c a u s e an ex-
p a s s down the off-gas pipe, flushing off the s o l i d s , tremely high heat flux and temperature. The
into the condensate receiver (Fig. 3.6). The feed more probable c a u s e s of uncontrolled high
system c o n s i s t s of a hold tank, which has a recir- temperatures are the failures of s e n s i n g devices
culation pump to keep solids in the suspended such a s thermocouples or power r e l a y s . If a
feed. A separate pump is used to feed the four malfunction does occur, positive feedback to
water-cooled feed l i n e s . The feed is monitored the induction coil from the overheated s u s c e p -
with rotameters and controlled manually, using tor does not occur if the insulation is adequate
ball v a l v e s . Needle and globe valves were tested to prevent e x c e s s i v e heat loss under normal
for throttling but were unsuccessful because solids operation. In contrast, the r e s i s t a n c e fur-
accumulated in the valves. nace can be safer because the heating coil
Evaluation of Induction and Resistance Heating. is hotter than the heated v e s s e l . Most com-
— Either induction or r e s i s t a n c e heating can be monly used r e s i s t a n c e coils have a failure
used for the Potcal and P o t g l a s s p r o c e s s e s . temperature of about 1250°C or lower, which
R e s i s t a n c e heating appears to offer fewer operating is lower than the melting temperature of the
problems and is recommended for the pot s y s t e m s . metals usually used for h e a t e d - v e s s e l construc-
Further evaluation is required in heating units con- tion. The failure of a r e s i s t a n c e coil in pref-
taining all of the refinements required for radio- erence to the heated v e s s e l , in c a s e of mal-
active operations to make a final c h o i c e . function of the control system, is a desirable
The following observations are made:
1. An induction furnace can be dangerous be- 2. Operations with the present induction equipment
c a u s e of the large amount of electrical power have been unpredictable. Cross coupling

ORNL-DWG 6 5 - 2 5 9 0 A


//////// /fH.





F i g . 3.6. D e t a i l of Vapor C o n d e n s a t e Equipment for C o n t i n u o u s Waste M e l t e r .

between adjacent coil zones has prevented the High-Level Radioactive Waste Disposal:
maximum u s e of heating capacity in a given Laboratory Studies
zone when required; up to 10 to 15% of the
energy could be transferred to, and c a u s e Laboratory studies included the development of
e x c e s s i v e heating in, an adjacent zone. suitable g l a s s e s ( " g l a s s " in this report includes
both true g l a s s e s and rocklike crystalline s o l i d s )
3 . The heat loss for the induction system was 49 for the incorporation of a variety of waste types
to 86% vs 10 to 29% for the r e s i s t a n c e furnace. derived from the Purex solvent extraction process
4. The high-frequency current from the induction (Table 3.3); the development of a computer code
for retrieval and correlation of data generated in
unit interfered with other nearby electronic
the melt developmental program at ORNL and at
other s i t e s ; the accurate measurement of melt
5. The heating units used in the ORNL t e s t s did viscosity and its correlation with the operational
not include all of the refinements such a s characteristics required of the melt in the RL-
forced air or water cooling tubes required for Potglass and Con-Potglass p r o c e s s e s ; the oper-
use with highly radioactive w a s t e s . Hence, ation of a small-scale continuous melter to study
final evaluation of such items as cost, in-cell volatilities of various elements and the corrosion
space requirements, weight, and cooling char- of materials of construction for a continuous
a c t e r i s t i c s and temperature control must await melter; and the development of a thermal con-
comparative t e s t s with radioactive w a s t e s in ductivity probe for measuring the levels of aqueous
specially designed resistance and induction solution, solid calcine, and molten glass in a
units. melter.

Table 3.3. Composition (g-moles per liter of waste) of Simulated Purex Waste Solutions
Without Fission Products

FTW-65 Waste
Constituent Representing
FTW.65-X3^ Low Iron High Iron
82 gal per ton
of Uranium

H^ 0.5 1.5 3.0 6.76

Na^ 0.30 0.9 0.6 0.20

Al^^ 0.05 0.15 0.2

Fe^^ 0.10 0.30 0.05 1.35

Cr^^ 0.02 0.06 0.006 0.005

Ni^ + 0.01 0.03 0.006 0.002

Hg^^ 0.0035 0.0105

so,^~ 0.15 0.45

po/~ 0.005 0.015 0.02
SiOj^~ 0.01 0.03

F~ 0.0005 0.0015

NO^-" ~1.0 '^^3.1 -^4.3 ~11.03

T h e FTW-65-X3 represents the h i g h e s t concentration of this waste type that can be conveniently simulated by
direct makeup from laboratory r e a g e n t s without heating the solution.
' ' j . M. Holmes, Acid 1st Cycle Waste, ORNL-TM-1013 (Dec. 7, 1964). T h e composition as listed is too c o n c e n -
trated to be a stable solution.
Enough NO ~ to balance in each w a s t e .

Glass Formulation. — Objectives of the g l a s s - 5. develop universal matrix g l a s s e s that can be

formulation development program are to: used with a wide variety of waste compositions;
6. operate safely and economically.
1. produce a glass that is not corrosive toward
conventional materials of construction; P a s t work h a s been concentrated primarily on
2. hold fission products and sulfate in the solid phosphate g l a s s e s because (1) phosphates are
product and eliminate side streams; more readily handled in acid solution than are
s i l i c a t e s , (2) low-melting g l a s s e s are more read-
3 . achieve a large volume reduction and a high ily prepared from phosphates than from s i l i c a t e s ,
concentration of waste oxides in the g l a s s ; (3) phosphate appears to be more compatible with
4. produce a solid product that is stable under sulfate retention in the glass than is s i l i c a t e ,
storage conditions, mechanically strong, insol- and (4) phosphate is compatible with the use of
uble, and a good heat conductor; phosphite a s a reductant for ruthenium.

OF?NL-DWG 6 5 - 9 7 ( 6
A computer code was developed for retrieving
and correlating the data from the development
of melts for fixation and ultimate storage of highly
radioactive w a s t e s . Initial readout was obtained
on 397 ORNL melt compositions, and a search for k\ • ^ ^
meaningful correlations has begun. The code can \
accommodate all data currently generated with
respect to compositions, properties, and conditions
of preparation of such s o l i d s . Provision is made
for future entry of additional data in the code.
X\ y,. V'>„ •-.; \
The code is a l s o designed to provide a conven- \~ ^k
_. X?/. >s!
ient method for exchanging data between groups Sv-^ \
working on the problem at different s i t e s . A
\% ^k ^
^^NT-^ N
s e r i e s of compositions of the high-phosphate N
(Brookhaven) type were received from Hanford for
coding. Specific details of the computer program \
are given in a topical report now in preparation.
Melt-development studies are continuing, with
emphasis on decreasing melting temperatures,
700 800 900 (000
melt v i s c o s i t i e s , and corrosivities. For this study, TEMPERATURE (°C)
the upper limit on operating temperature of a melter
h a s been chosen a s 900°C, s i n c e the rate of Fig. 3.7. Viscosity of Phosphate Melts Incorporating
reaction of steam with s t a i n l e s s s t e e l increases Purex FTW-65 Waste.
rapidly with increasing temperature, beginning at
about 910°C.'' Most of the melt development has
been done with high-sulfate Purex w a s t e , primarily
the FTW-65-X3 type, but some work was with non-
Calculation of the amounts of fission products
sulfate Purex wastes containing very low to high
resulting from fuel burnups of 10,000 and 35,000
concentrations of iron (Table 3.3).
Mwd/metric ton, using the data compilation of
A melt previously found to be suitable for the Blomeke and Todd, ^ resulted in appreciable
rising-level method of fixation (melt 1, Table 3.4) changes in both distribution and total amounts of
was found to be too viscous at 900°C for use in fission products from the actual composition used
a continuous melter. The addition of simulated (Table 3.5). The revised compositions will be
fission products equivalent to fuel burnups of used in future ORNL work.
10,000 and 35,000 Mwd/metric ton increased both
Satisfactory melts were readily developed for
the melting temperature and the viscosity of the
fixation of nonsulfate low-iron Purex waste with
melt. Dilution of the melt with fluxing agents,
10,000 Mwd/metric ton equivalent fission products
primarily sodium, lithium, phosphate, and alu-
(melts 5 and 6, Table 3.4). Scoping studies on
minum, together with the omission of calcium,
the development of melts for the fixation of non-
reduced both the melting temperatures and the
sulfate high Purex waste are designed to lower
v i s c o s i t i e s of the melts to acceptable ranges for
the melting temperatures and the v i s c o s i t i e s , and
continuous operation, even in the presence of
to increase the volume reductions obtained using
the simulated fission products (melts 2 and 3 ,
a s reference the melts developed for the s t a i n l e s s
Table 3.4 and Fig. 3.7). Preliminary viscosity
measurements have been mad^ with a modified
falling-sphere a p p a r a t u s ' (Fig. 3.8). More data
F . L. Culler, Jr., et al., Semiann. Progr. Rept, Chem-
are necessary for definition of the viscosity ranges ical Technology Division, June 1 to Nov. 30, 1964,
required for u s e in the R L - P o t g l a s s and Con- ORNL-TM-1081; s e e also L. Shartsis and S. Spinner,
" V i s c o s i t y and Density of Molten Optical G l a s s e s , "
Potglass processes. J. Res. Natl. Bur. Std. 46, 176 (1951).
®J. O. Blomeke and M. F . Todd, Uraniam-235 Fission-
Product Production as a Function of Thermal Neutron
Flux, Irradiation Time and Decay Time, ORNL-2127,
H. H. Uhlig, Corrosion Handbook, p. 514, Wiley, Part II, vols. 1, 3, Summations of Individual Chains,
New York, 1955. Elements and the Rare Gas and Rare Earth Groups (1957).

T a b l e 3.4. Compositions of Mixes for Producing Purex Glasses

With and Without Simulated Fission Products

Additives FTW-65-X3 Nonsulfate

(g-moles per liter of waste)

"3PO4 1.5 3.6 11.2 11.2 2.0 2.0

LiOH 1.5 2.4 7.5 7.5

NaOH 2.7 8.45 8.45

Ca(N03)2 0.2

A1(N03)3 0.5 0.75 2.35

Pb(N03)2 0.2 0.3

Wt % w a s t e oxides (theoretical)
No F P ' s 39.6 20

10,000 Mwd/metric ton F P ' s ^ 52 30 24 22

35,000 Mwd/metric ton F P ' s ' ' 30.1 32.2

Approx. mp ( C)
No F P ' s 800 700
10,000 Mwd/metric ton F P ' s 1000 800 800 800

35,000 Mwd/metric ton F P ' s 800 800

Sulfate L o s s (wt %)
No F P ' s 19.6 10.2

10,000 Mwd/metric ton 10.6

35,000 Mwd/metric ton

^Using Hanford-BNL simulated fission product composition (Table 3.5).

U s i n g ORNL simulated fission product composition (Table 3.5).
'^From thermogravimetric t e s t s , assuming total weight loss to be sulfate. Semiengineering-scale fixation t e s t s nor-
mally give a much lower value.

Steel (Darex) w a s t e . ' R e s u l t s are s t i l l in a pre- incorporated a storage chamber in which the prod-
liminary s t a g e . uct could be retained in molten form until the
Volatility and Corrosion T e s t s in a Continuous end of the d a y ' s operation (Fig. 3.9). T h e melting
Melter. — A s e r i e s of continuous fixation experi- chamber itself was 4 in. in diameter and 24 in.
ments (Con-Potglass process) were run on Purex long. Fixation rates obtained in this melter were
FTW-65-X3 waste using a s the continuous melter a s high as 7 liters of feed (waste plus additives)
a fixation pot modified by the addition of a bottom per hour; about 4 l i t e r s A r appeared to be a more
outlet with an overflow weir. Later models also realistic figure for steady-state operation. A rate
of about 1 liter/hr was used in most of t h e s e
experiments, s i n c e it was difficult to start up,
W. E . Clark and H. W. Godbee, Laboratory Develop-
ment of Processes for Fixation of High-Level Radioac- reach steady s t a t e , and operate for an appreciable
tive Wastes m Glassy Solids: Wastes Containing (1) length of time during a normal 8-hr working day.
Aluminum Nitrate and (2) the Nitrates of the Constituents
of Stainless Steel, ORNL-3612 (July 1964). Consequently, the experiments were operated on





- • ^



Fig. 3.8. Restrained-Sphere Apparatus for Determining the Melt V i s c o s i t y .

an intermittent rather than a long-term continuous experiments done under rising-level conditions.^
basis. Sulfate volatility was erratic, ranging from 0.67
The objectives of these experiments were to to 14.8%. This variation was probably caused by
investigate the volatility of various constituents upsets in operating conditions, though these were
and the corrosion of the melter during the con- not obvious at the time, possibly because of the
tinuous melting process as well as to define brevity (<8 hr) of each operating period. Con-
practical operating problems such as the required sistently, 1 to 2% of the molybdenum volatilized.
ranges of melt viscosity, the possibility of opera- Minimum values of entrainment determined from
tional difficulty from solids building up in the the total amounts of lithium, aluminum, and/or
melter, and the deposition of solids in the off-gas phosphate reporting to the condenser, off-gas l i n e s ,
line. and scrubbers were rather consistently 0,2 to
A continuous melting process requires a more 0.5% of the total solids present. It is expected
fluid (less viscous) melt than the rising-level that these solids would be retained in a loop s y s -
(RL-Potglass) process. Continuous operations tem. Operation of the melter in conjunction with
for about four days and intermittent operation for a close-coupled evaporator should hold the amount
total periods of about 38 hr without opening the of volatile and entrained solid from the loop to
melter resulted in no buildup of solids in the
melter using melt 2, Table 3.4 (Purex FTW-65-X3
•^W. E . Clark, H. W. Godbee, and C. L. Fitzgerald,
waste). The results of the studies of volatility Laboratory Development of Processes for Fixation of
and entrainment (Table 3.6) showed that iron and High Level Radioactive Waste in Glassy Solids. 3.
Wastes from the Purex Solvent Extraction Process,
ruthenium had the highest volatilities, since 20 OKNL.3640 (in p r e s s ) .
to 30% of the total amounts of these elements ^C. W. Hancher, J . C. Suddath, and M. E . Whatley,
either reported to the condensate or collected a s Engineering Studies on Pot Calcination for Ultimate
Disposal of Nuclear Waste: Formaldehyde-Treated Purex
solids in the off-gas line. The volatility results Waste for 1965 (FTW-65), ORNL.TM-715 (Jan. 6, 1965);
for iron are doubtlessly too high because severe s e e also C. W. Hancher and J . C. Suddath, Pot Calcina-
tion of Simulated Radioactive Wastes with Continuous
corrosion occurred in the s t a i n l e s s s t e e l off-gas Evaporation, ORNL-TM-117 (Dec. 26, 1961); R. E .
Blanco and F . L. Parker, Waste Treatment and Disposal
line and condenser c o i l s . High iron volatility i s , Progr. Rept. for February to April 1963, ORNL'TM-603
however, consistent with the results of earlier (Dec. 16, 1963).

l e s s than 0.1% of the solids present, especially The unavoidable presence of ruthenium in a
if a small amount of reductant is added to the s t a i n l e s s s t e e l off-gas line will doubtless a c c e l -
evaporator to lower the ruthenium volatility. The erate corrosion if the line is kept at a temperature
off-gas line may require periodic flushing in low enough to allow condensation of nitric acid.
any c a s e . Corrosion of " d r y " s t a i n l e s s s t e e l (i.e., at
~ 1 5 0 ° C ) is negligible in this off-gas, and the
corrosion of titanium in the " w e t " off-gas and
T a b l e 3.5. Simulated Fission Product Concentrations condensate is also quite l o w . ' °
g-moles per liter of waste for 333 liters of waste per ton Platinum, Corronel 230, 50% N i - 5 0 % Cr, H a s t e l -
of uranium (80 gal per ton of uranium) loys C and F , Inconel, Ni-o-nel, Nichrome V, BMI-
HAPO 20, and s t a i n l e s s s t e e l types 304L, 310,
10,000 Mwd/Metric 35,000
446, and Carpenter No. 20SCb were exposed in
Ton Equivalent Mwd/Metric Ton
the melter in the form of long s t r i p s , rods, or wires
which extended from the vapor s p a c e in the top
HanfoTdl ' Equivalent,
„-„a ORNL" ORNL"
of the melter, through the solution and calcine
l a y e r s , and into the melt. Penetration rates were
determined by micrometer measurements at numer-
Zr 0.021 0.045 0.1675 ous positions along the length of the specimens
Mo+ T c 0.025'^ 0.044"' 0.1643'^ before and after exposures.
(Mo) (0.035) (0.1321) All materials tested except platinum showed
(Tc) (0.009) (0.0322) appreciable rates of attack, the maximum gener-
ally occurring at or near the position of the melt
Y + lanthanides 0.045 0.081^ 0.2966^
interface. The high-nickel alloys, Corronel 230,
(Ce) (0.022) (0.0741)
Nichrome V, Inconel, and 50% N i - 5 0 % Cr were the
(Others) (0.059) (0.2225)
most resistant of the conventional alloys tested.
R b + Cs 0.014 0.025 0.0982 The maximum penetrations of t h e s e alloys, based on
(Rb) (0.007) (0.006) (0.02 09) exposure times at operating temperature, varied
(Cs) (0.007) (0.019) (0.0773) from 0.176 mil/hr for Corronel 230 at 900°C to
Sr + Ba 0.011 0.022 0.0832 0.945 mil/hr for Inconel at 950°C (Table 3.7).
(Sr) (0.0055) (0.014) (0.0503) Level Detection During Formation of Melt. —
(Ba) (0.0055) (0.008) (0.0329) A modified "thermal conductivity probe"^^ was
0.016' 0.084l'
used to determine the type of material ( g l a s s ,
Ru, Rh, Pd 0.028'
calcined solid, liquid, or gas and vapor) filling
0.0017 0.0068 various zones or regions in a pot during melt for-
mation. T h i s device can make use of the thermo-
T o t a l g-moles 0.144 0.2347 0.8871
couples already in the pot, so that little or no
additional paraphernalia will be required inside
P e r s o n a l communication from G. Rey, HAPO, to
BNL Waste D i s p o s a l Group, Mar. 10, 1964. F i s s i o n the pot. It h a s the additional advantage that the
product distributions are for the Plutonium R e c y c l e T e s t thermocouples do not lose their utility a s temper-
Reactor ( P R T R ) .
ature s e n s o r s . The probe c o n s i s t s basically of a
Calculated according to J . O. Blomeke and M. F .
Todd, Uranium-235 Fission-Product Production as a heat sink, which could be the wires of a thermo-
Function of Thermal Neutron Flux, Irradiation Time and couple itself, and a temperature-sensing device
Decay Time, ORNL-2127, part II, v o l s . 1 and 3, "Sum-
mations of Individual Chains, E l e m e n t s , and the Rare
Gas and Rare-Earth G r o u p s , " assuming 3 x 10 nvt
(thermal), decay time of 10 s e c , irradiation time of
3 X 10^ s e c , N ° j = 5.1 x 10^^ (2% ^^^U). ' " w . E . Clark, P . D. Neumann, L. R i c e , and D. N.
H e s s , Laboratory Development of Processes for Fixation
'^Calculated from ORNL-2127, part II, v o l s . 1 and 3 of High-Level Radioactive Wastes in Glass. 4. Corro-
(see b above), assuming 10 1 3 nvt (thermal), decay time sion Studies on Candidate Materials of Construction,
ORNL-3816 (in preparation); s e e also R. E . Blanco and
of 10^ s e c , irradiation time of 10* s e c , N ° j = 1.72 X F . L. Parker, Waste Treatment and Disposal Quart.
10^® (67o ^^^U). Progr. Rept. August-October 1962, ORNL.TM-482,
pp. 1 8 - 1 9 .
' ' A U added a s MoO^.
^Added a s "didymium n i t r a t e " plus cerous nitrate. H. S. Carslaw and J . C. Jaeger, Conduction of Heat
in Solids, 2nd ed., p. 345, Oxford University P r e s s ,
' A U added a s RuCl . London, I9S9.

ORNL-DWS 65-1850A










Fig. 3.9. Continuous Melters Used in Semiengineering-Scale F i x a t i o n Experiments.

such a s a thermocouple. The probe i s heated (or positions. The materials tested at various tem-
cooled), and the response of the temperature peratures in the pot were air, a magnesia powder
sensor is recorded. This response is a measure (to simulate calcined material), and a phosphate
of the effective thermal conductivity of the medium g l a s s . The heater dissipated about 20 w/lin in.
surrounding the probe. Representative results with one of the thermo-
Experiments were carried out with a probe 1 in. couples at 850°C when surrounded by each of the
in diameter and 22 in. long, concentrically located above different materials are shown in F i g . 3.10.
in a pot 4 in. in diameter and 24 in. long. T h e T h e s e results show that such a probe h a s promise
probe was heated by a Nichrome-wound heater a s a device to permit discrimination between
that extended the full length of the probe. Chro- regions filled with g l a s s , calcine, or g a s e s . A
mel-Alumel thermocouples were attached to the heater capable of dissipating more power per unit
1-in. tube at different radial and longitudinal of length than the one used in these studies will
T a b l e 3.6. V o l a t i l i t y and Entrainment of Various Elements During Fixation of Purex FTW-65 Waste
in a Continuous Melting Process

Experiment No.

Melter Discharge Open Melter Discharge Closed

Simulated fission products present Yes Yes No No

(10,000 Mwd/metric ton equiv) 700

Approximate softening temp., ° C 700 700 600 600 700 700

Maximum operating temp., ° C 875^ 850 850 950 900

Melt viscosity at operating temp., p o i s e s 10 6.5 6.5 '^4.1 7.4

Volatility and entrainment tility (condensate Volatility Entrainment Volatility Entrainment

(wt % of total present) plus scrub solution) (cond. + scrub) (off-gas lines) (cond. + s c r u b ) (off-gas lines) °^

SO, 1.4 14.8 1.1 0.73 1.16 1.0 2.16 7.19 1.69 8.88

P O 4. 0.5 0.35 0.23 0.24 2.1 2.10 0.24 0.24

Fe 3.7 2.5 1.9 8.41 12.91 31.32 15.8 10.6 26.4
f I
21.73' 16.6
Ru 3.92 20.52

Mo 0.76 0.45 1.21 1.29 0.52 1.81

Al^ 0.24 0.24

Li« 0.36 0.05 0.41

^Melt additives (g-moles per liter of waste): 3.6 H^PO^, 2.7 NaOH, 2.4 LiOH -H^O, 0.75 AKNOj)^ ' 9 H^O.
Figure 3.9a.
"Figure 3.96.
BNL-Hanford composition (see Table 3.5).
There was a short temperature excursion to 925°C.
The bulk of the ruthenium in both condenser and off-gas lines was recovered a s s o l i d s . T h e s e were collected for a n a l y s i s .
"Analyzed for as indicators of entrainment.

shorten the time required to distinguish between Pilot Plant Design

substances. T h e time required in t h e s e s t u d i e s
to make t h e s e distinctions was about 1.5 min Active liaison with Hanford personnel on the
( F i g . 3.10). A remaining problem in the use of design and startup of the pot calcination and
such a probe i s the effect of films or layers that rising-level P o t g l a s s equipment for the Hanford
will build up on the probe during the filling of a Waste Solidification Engineering Prototypes P i l o t
pot with melt. Plant c o n t i n u e s . T h i s equipment is presently

Table 3.7. Corrosion During Fixation of Purex FTW-65 Waste

in a Continuous Melting Process^

Maximum Corrosion R a t e
Scoping T e s t s : (mils/hr)
Candidate Material T = 875-900°C; Remarks
T e s t 1: T e s t 2:
Time — Variable
950°C, 900° C,
35V hr 36V hr
4 4

Platinum No attack; no weight change 0 No attack in any t e s t

Corronel 230 0.176 No marked local attack;

maximum attack at melt

Inconel Slight attack 0.943 0.508 Nearly all attack near melt

50% N i - 5 0 % Cr 0.356 0.566 Nearly all attack near melt


Nichrome V 0.288 Visible local corrosion in

and just above melt
BMI-HAPO 20 Severely attacked at melt 1.78 E x c e s s i v e corrosion at melt
interface interface
Type 446 SS 0.867 Considerable attack over
entire specimen; maximum
attack at melt interface

Ni-o-nel Slight attack Excessive Completely destroyed in melt

Hastelloy F Attack varied from slight to Eliminated in scoping t e s t s


Haste Hoy B Very severe pitting attack C o n s i s t e n t l y poor in a l l


Hastelloy C E x c e s s i v e local attack Eliminated in scoping t e s t s

Carpenter No. 20SCb Heavy pitting attack Eliminated in scoping t e s t s
Type 310 SS Heavy attack m vapor Consistently poor m vapor
Type 304L SS Attack in melt and at melt Insufficiently r e s i s t a n t for
interface construction of a continu-
ous melter

Melt 357 in computer readout. Additives (g-moles per liter of w a s t e ) : 3.6 H P O , 2.7 NaOH, 2.4 L i O H ' H O,
0.75 Al(NOj), -9 H^O. ^ ^

undergoing design-verification t e s t s . It will be
installed in the Hanford F u e l s Recycle Pilot Plant 2-in VAPOR LINE
in August of 1965.
An analog computer study was made of the
transient response to surges in vapor rate gener-
ated while feeding liquid waste to a pot calciner
operating at temperatures up to 900°C.
system currently installed for the design verifi-

cation t e s t s at Hanford is shown in Fig. 3 . 1 1 .

Vapors from the pot calciner flow through a 2-in.
vapor line to the condenser, where the steam and
nitric acid are condensed. Noncondensable vapors
flow to the evaporator through a 2-in. line which
may contain a jet. This jet would be used to PROPOSED ORIFICE-
increase vapor circulation rates if required. The
SEAL POT 8 in DIAM, 36 in HIGH-
condensate flows by gravity from the condenser
to the evaporator. A water s e a l pot is installed POT CALCINER 8 in DIAM, 90 in HIGH'
in the vent line from the pot calciner to serve as
an emergency relief valve in c a s e the normal vapor
line plugs. The maximum allowable height of
liquid in the water s e a l before it vents is 24 in. F i g . 3 . 1 1 . Pot-Calciner Vapor and Vent System.
The s e a l vent is also connected to the evaporator,
which serves as a common plenum for both the
condenser and s e a l pot v e n t s .

ORNL-DWG 65-97(7 The system was simulated on a PACE TR-10

analog computer, using the following basic assump-
/ tions :
The pressure in the evaporator would be con-
09 /
stant and at a slight vacuum.
/ 2. The normal vapor rate would be about 148
ypo\i TOER" cfm, which approximates the maximum boilup
« 0 7
/ rate from a 12-in.-diam pot calciner during the
U06 / y
y initial feeding period. The vapor-line pressure
differential at this rate would be about 6 in.
/ ' H O without installation of the jet downstream

/ /
of the condenser, and 10.2 in. H O with the
u / ^
o jet installed.
0 4
" G L A SS
/ The prime c a u s e of pressure fluctuations in
03 / / the system is the surges of vapor at i n s t a n t a n e -
/ > ous rates considerably above the normal rate.
0 2 /^.

V - -
T h e s e surges are probably due to fluctuations
in the calciner feed rate, liquid level, and wall
temperature. Vapor surges were simulated a s
<y s t e p functions in the normal vapor rate for the
20 40 60 80 (00 (20 (40 (60 (80 purposes of this study.
TIME (sec)
The level in the s e a l pot could be simulated
F i g . 3.10. T h e Response of a Chromel-Alumel Thermo- by treating the system a s a manometer with a
couple at 850°C as a Function of the Surrounding fluctuating pressure on the pot calciner s i d e
Medium when Heated at a Uniform Rate. and a constant pressure on the evaporator s i d e .

R e s u l t s of a typical run are shown in F i g . 3.12, vapor-surge frequency on the s e a l differential

where the vapor rate, pot and evaporator pressure, height. The data indicate that the system could
and manometer differential height are plotted vs take surges 70% above the normal rate without
time. This run shows that for a 30% increase in venting the s e a l . The area of the s e a l restriction
vapor rate at a 2-sec frequency, the pot pressure (4.54 in.^) is about 35% greater than the area of
increased to —20 in. H O from —27 in. H O , and the inlet pipe to the s e a l , s o no severe restriction
the manometer differential height reached 23 in. would be added to the flow of vapor through the
H O ; so the s e a l was within 1 in. of venting. seal-pot vent system.
Results of all the computer runs are summarized As a result of this study, a recommendation was
in F i g . 3.13, which shows the maximum s e a l dif- made that the jet in the condenser vent line be
ferential height as a function of the vapor rate removed and that the area under the s e a l baffle be
increase and the step-function frequency. The reduced to 4.54 in.^ in order to dampen out the
latter is expressed in terms of 0 , defined as the effects of the vapor fluctuations on the seal-pot
absolute value of the difference between an odd level.
integer, n = 1, 3 , 5, . . . , and the ratio of the s t e p - A study of thermocouple materials for the pot
function period to the natural period of the s e a l calciner and its furnace was made to determine
pot. As the period of the step function approached the most economical and reliable materials for
an odd multiple of the seal-pot natural period the Hanford pilot plant. Two c l a s s e s of thermo-
( 0 -"O), the adverse effect of vapor rate increase couple service were specified. For the furnace
was intensified. Curves B and C in F i g . 3.13 liner and susceptor, the thermocouples must be
indicate that the s e a l was l e s s likely to vent when stable for at l e a s t 100 hr at 1100°C and for 1000
the restriction caused by the I'^^-in. jet was re- hr at 1000°C. T h e s e couples will be used for
moved. Considerable improvement was obtained many t e s t s and therefore could be constructed of
by inserting an orifice with an area of 4.54 in.^ in more costly materials such a s platinum. For the
the s e a l pot at the bottom of the baffle. This pot calciner the thermocouples must be stable for
tended to dampen the manometer fluctuations, as at least 100 hr at 900°C and for very long times
shown in curve C, and also removed the effect of at temperatures below 700°C.

ORNL-DWG 6 5 3043

F i g . 3.12. Typical Computer Plot of Pot-Calciner Pressure-Transient Study.


ORNL-DWG 6 5 - 3 0 4 4

42 (6 20 24 28

F i g . 3.13. Pot-Calciner Pressure-Transient Study: Effect of Step Function on Woter-Seal Differential Height.

Thermocouples recommended were as follows: which is essentially a nickel-silicon binary alloy.

1. Furnace liner and susceptor — Pt vs Pt—10% Improved h i ^ - t e m p e r a t u r e oxidation r e s i s t a n c e is
Rh thermocouple wires sheathed in Inconel or claimed for t h e s e newer materials.
Hastelloy X, with MgO insulation. A survey^ ^ ' ' ^ of the composition and quantity
of w a s t e s to be generated at the Nuclear F u e l s
2. Pot calciner — Chromel-Alumel thermocouple Services Reprocessing Plant, near Buffalo, New
wires sheathed in type 310 s t a i n l e s s s t e e l . York, was made to enable t e s t i n g of this w a s t e
Since t h e s e thermocouples will probably be type in the hot pilot plant. Since fuels reprocessed
used only once during a calcination, they must in this plant will have very high burnups (up to
be a s inexpensive a s possible, consistent with 20,000 Mwd per metric ton of uranium), the w a s t e s
the service specifications.
Further recommendations included the use of 20
gage a s the minimum wire diameter, ungrounded J . M. Holmes, Survey of Nuclear Fuel Services, Inc.,
wires to obtain longer life, and investigation of Waste Composition'^, ORNL-TM.1013.
the new Chromel-P wire, which contains a small R. E. Blanco and F . L. Parker, Waste Treatment and
quantity of niobium and a s p e c i a l type of Alumel Disposal Semiann. Progr. Rept. July-Dec. 1964, ORNL-

will contain high concentrations of stable and The w a s t e s — incinerator a s h e s , concentrated

active fission products and will generate high aqueous solutions, or filtered sludges with variable
internal heat fluxes in the calcine or glassy s o l i d s . water content — are gradually introduced into
Compositions were obtained for two types of Purex, hot asphalt (about 240°C) with violent stirring.
Thorex, and TBP-25 w a s t e s . Most of the work at Mol has been with s l u d g e s .
Formulation of a detailed schedule for pot c a l - At Marcoule, F r a n c e , ^ ' a continuous process
cination and rising-level glass t e s t s in the hot has been developed for incorporating sludges in
pilot plant was completed recently with Hanford asphalt. The sludges to which emulsifying agents
personnel. During the initial operating phase, have already been added are mixed with hot asphalt
three pot calcination t e s t s will be run on Purex (about 125°C). The major portion (80 to 90%) of
w a s t e s having approximately 1000 to 2000, 5000, the water in the sludges is exuded and thus s e p -
and up to 6000 Btu hr~^ ft~^ of internal heat- arated from the coated w a s t e s . The residual
generation rates. Two rising-level t e s t s will be water in the asphalt mixture is boiled off and
made: one on Purex waste containing tracer levels the mixture heated to about 130°C. Since most
of radioactivity, and the other with up to 6000 soluble activity would be in the exuded water,
Btu h r ~ ' ft~^ of internal heat-generation rate. the process is applicable only to w a s t e s in which
Both sulfate and nonsulfate Purex w a s t e s will be the radiation emitters are tied up in the sludge
tested in this initial phase. T e s t s on other w a s t e s s o l i d s . At ORNL, a process (Fig. 3.14) h a s been
such at TBP-25, Thorex, and Darex will be con-
sidered during the second operating phase, which
is scheduled tentatively for calendar year 1967. P . Dejohghe, N. Van de Voorde, J . Pyck, and A.
Stynen, Insolubilization of Radioactive Concentrates
by Asphalt Coating, final report No. 2, 1st part, con-
cerning proposal 167, April 1, 1961, to March 3 1 , 1963,
RADIOACTIVE WASTE J . Rodier, G. LefiUatre, and J . Scheidhauer, Bitumen
Coating of the Radioactive Sludges from the Effluent
Treatment Plant at the Marcoule Center, Review of the
Nuclear installations generate large volumes of progress reports 1, 2, 3, and 4, CEA report 2331, 1963
intermediate-level radioactive w a s t e s (ILW), such
a s evaporator concentrates, second- and third-
cycle solvent extraction raffinates, and slurries
ORNL-DWG 6 5 - 6 5 9 A
or solids (residues from low-level treatment proc-
e s s e s ) . T h e s e wastes are characterized by their
modest levels of radioactivity, s o that heating
and radiation dose levels do not present serious
problems, and by their high salt or solids contents,
which prevent their efficient treatment by con-
ventional ion exchange, precipitation, or foam-
separation methods. The ORNL intermediate-level
waste treatment program includes development of
processes to remove the water from t h e s e w a s t e s
and to incorporate the dried solids in a cheap,
inert, insoluble material before burial or contained
storage. Incorporation of t h e s e wastes in asphalt
or other plastic and e l a s t i c substances appears
to be simple, inexpensive, and relatively insen-
sitive to the type of material being incorporated,
and, within limits, independent of local geologic
and hydrologic conditions. Asphalt has good
adhesive and coating properties and is insoluble
in water, but its sensitivity toward radiation limits
the activity level of included material.
At Mol, Belgium,^'* a batch process has been Fig. 3.14. Flow Diagram of Process to Incorporate
developed for incorporating the w a s t e s in asphalt. ILW in Emulsified Asphalt.

developed for incorporating w a s t e s in emulsified 60 wt % solids from waste flowed freely from the
asphalt. Attractive features of the process include: evaporator at 130°C. The product containing 80
wt % s o l i d s from waste was removed with diffi-
1. u s e of emulsified asphalt, which flows readily
culty at 195°C.
at room temperature;
Since the product containing 60 wt % s o l i d s
2 . evaporation at low temperatures s o that entrain- from w a s t e (Fig. 3.15) represented a good corn-
ment of activity from the evaporator and thermal
degradation of the asphalt are minimized;
3 . agitation at modest rates (100 to 500 rpm, which T a b l e 3.8. Composition of Simulated O R N L
provides adequate mixing and keeps the heated Intermediate-Level Radioactive Waste Solution
surfaces of the evaporator clean; Incorporated in Emulsified Asphalt

4. operation in batch or continuous manner; Density of solution at 25°C = 1.335 g/ml

5. incorporation of soluble and insoluble activity

with equal effectiveness. Component Molarity
Asphalt products containing 20 to 80 wt % solids
from waste have been prepared. The waste used Na 6.61
in t h e s e studies was simulated ORNL w a s t e - NH * 0.19
evaporator bottoms (Table 3.8), and the asphalt 4

was an emulsified asphalt used in the surface 0.22

treatment of roads. (This asphalt emulsion is 0.35
defined in Federal Specifications SS-A-674b a s 0.056
type RS-2: a rapid-setting high-viscosity emulsi-
fied asphalt. The material used contained about OH" 2.06
63 wt % base asphalt, 35 wt % water, and 2 wt % NO ~ 4.64
emulsifying agent.) Products containing 20 to

PHOTO 67974R

1 inch

F i g . 3.15. Solids-Asphalt Product Incorporating 60 wt % Solids from Simulated O R N L ILW Solution. Nominal
composition of product (wt % ) : 47.8 N o N G j , 9.1 N a O H , 4.6 A l j C S O ^ j j , 0.4 N a C i , 0.3 N o j S O ^ , and 37.8 a s p h o l t .
D e n s i t y - 1.5 g / c m . Volume r e d u c t i o n (volumes of w a s t e per volume of product) = 2 .

promise of properties — volume reduction, v i s c o s -
ity, ductility, e t c . — it was studied in most detail. \
The process consisted in (1) adding the waste \
directly to emulsified asphalt with modest stir- \
ring — about 100 rpm, (2) boiling off the water, \
(3) raising the temperature of the product to the \
desired final temperature — usually 165°C, and
(4) draining the product into an appropriate con-
V. y
, _ c ; o n i i i M Awn ' ^ ' r < ; F R O M P R O D U C T
tainer. Although the process described above i s T A i K M R i ^ i~r\ I. OT t-rs\ \r\c-

a batch operation, it h a s been carried out with

slight variation in procedure as a continuous
Probably the best measure of how well the waste
particles have been coated with asphalt is how
r e s i s t a n t the product i s to leaching by water.
T h u s , asphalt products containing 20 to 60 wt %
solids from simulated ORNL waste-evaporator
bottoms have been leached with water for periods \
of up to nine months. As expected, these leach \ /
t e s t s indicate surface contamination of the sam- \ / 106- .

p l e s , which d e c r e a s e s rapidly during the first CONTAINING 6 0 w t % SOLIDS

\ /
several weeks ( F i g s . 3.16 and 3.17). The term f
\ ^
80 (20 (60 200 240 280
TIME (days)
ORNL-DWG 65-(4 a

Fig. 3.17. Leach Rates of Cs, Sodium, and Ru

from Asphalt Product Containing 60 wt % Solids from
O R N L ILW Solution.

\(\ l , B n wt \ SO i n s
" l e a c h r a t e " rather than " d i s s o l u t i o n r a t e " i s
used, since the asphalt matrix does not d i s s o l v e .
Leach rates were calculated by using the ex-
Vi i ; ^ ( F L OWING hj . n i
\ " • ^ J pression
\ ^« i r
\ \
> :
;Owt \ SOLIDS --
activity l e a c h e d / c m ^ / d a y
a c t i v i t y / g of product
= g cm ^ day '
\ Dwt% SC

T e s t s were made with stagnant (static t e s t ) and
flowing (dynamic t e s t ) distilled water at 2 5 ° C .
\\ The water was sampled and completely replaced
\X /y
t - / with fresh water at the end of each week.
The asphalt product containing 60 wt % solids
from waste (Fig. 3.15 and Table 3.8) reaches an
apparent steady-state leach rate for ' ^ ' C s and
sodium of 3 X 10""* g cm~^ d a y " ' in s t a t i c t e s t s ,
and about 5 x 1 0 " ' ' g cm~^ d a y " ^ in dynamic
10" t e s t s after about six months (Fig. 3.16). The
0 40 80 120 (60 200 240 280
difference between leach rates determined by
TIME (days)
gamma counting of ^^'^Cs and flame-photometric
1 37 determination of sodium in the same leachate was
Fig. 3.16. Leach Rates of Cs and Sodium from
Products Incorporating O R N L ILW Solution. negligible. The product containing 20 wt % solids

(Fig. 3.16) reaches an apparent steady-state leach A dose of 10^ rads caused it to swell very slightly
rate for ^ ^ ' C s and sodium about half that of the (Fig. 3.19), while a dose of 10^ rads c a u s e d the
product with 60% solids — 1.5 x lO""* g cm~^ material to increase about 36% in volume (Fig.
day"'. Unlike the results with the product con- 3.20). Leach t e s t s on a sample irradiated to a
taining 60% s o l i d s , the results with the product con- dose of 10* rads indicated that the irradiation
taining 20% solids are not influenced by whether h a s no effect on the leach rate.
the water is flowing or stagnant. T h e s e results Design of a pilot plant to incorporate 100 gal
suggest that there is a r e s i s t a n c e to diffusion of ORNL waste-evaporator bottoms in emulsified
due to concentration in the leachate in the first asphalt per 8-hr shift by the above batch process
c a s e (60%) that is not apparent in the second has been completed and installation has started.
(20%). The asphalt product containing 60 wt % Studies have been started to determine the long-
solids from waste reaches an apparent steady-state term effect of incorporated radioactive elements
leach rate for ' ""^Ru of 7 x 1 0 " ^ g cm"^ d a y " ' on the physical (in particular, leach rate) and
(Fig. 3,17). The fact that the above leach rates chemical (in particular, radiolytic g a s e s formed)
for '•'^Cs and sodium are 50 to 100 times higher properties by incorporating low and high concen-
than leach rates reported in Belgium''* and trations of radionuclides in asphalt samples.
F r a n c e ' ^ implies that leach rate is a function of F a c t o r s being investigated include the effect of
whether the incorporated activity is soluble or stirring rate, temperature, and time at temperature
insoluble. on the final product; of radiation dose level on
To evaluate the r e s i s t a n c e of the asphalt product leach rate; of different base a s p h a l t s and emulsi-
to irradiation, samples of the product containing fying agents on the quality of the final product;
60 wt % solids from waste were irradiated by and of solubility on the leach rate of a given com-
means of a *°Co source to absorbed d o s e s ranging pound, namely, Sr(NO^^ vs SrCO . A hazards
from 10^ to 10^ rads. A dose of 10® rads had survey and evaluation program has also been
negligible effect on the asphalt sample (Fig. 3.18). started.

PHOTO 69225

I I \ I L I \ I I


rom O R N L ILW Solution After Irradiation to 10 Rods.

F i g . 3.18. Asphalt Product Containing 6 0 wt % Solids f


PHOTO 6 9 2 2 4


F i g . 3.19. A s p h a l t P r o d u c t C o n t a i n i n g 60 wt % S o l i d s from O R N L ILW S o l u t i o n A f t e r I r r a d i a t i o n to 10 Rods.

PHOTO 6 8 2 5 6

1 inch

F i g . 3.20. A s p h a l t P r o d u c t C o n t a i n i n g 60 wt % S o l i d s from O R N L ILW Solution After I r r a d i a t i o n to 10 Rods.


3.3 TREATMENT OF LOW-LEVEL pursued." Subsequently, a foam-separation

RADIOACTIVE WASTES pilot plant for treating ORNL LLW was designed,
i n s t a l l e d , " and operated.^" No further de-
Low-level radioactive w a s t e s (LLW) c o n s i s t of velopment of the foam process for treating LLW
waters with low solids and radioactivity contents. IS planned. Decontamination of LLW in the scav-
They are customarily treated to remove radionu- enging-precipitation foam-separation pilot plant ^°
c l i d e s and then discharged to streams. The ob- was accomplished in two s t e p s : (1) precipitation
jective of the ORNL program is to develop, evalu- of calcium, magnesium, and radionuclides in a
a t e , and determine the cost of alternative p r o c e s s e s scavenging-precipitation step that included the use
for decontaminating LLW to l e s s than the maximum of Grundite clay for sorption of cesium and (2)
permissible levels permitted for w a t e r , ' ^ M P C ^ , removal of strontium and other radionuclides by
for occupational exposure (168 hr/week). The complexing with a foaming agent, dodecylbenzene
retained s o l i d s and radionuclides can be combined sulfonate, and then removing the complex by foam-
with high-level radioactive wastes for ultimate ing in a countercurrent foam-separation column.
disposal, or they can be incorporated in asphalt or The scavenging-precipitation system could be
concrete and buried in restricted a r e a s . used alone, where decontamination factors for
Two p r o c e s s e s are being studied for treating strontium and cesium of about 10 were required,
LLW: the scavenging-precipitation foam-separation or in combination with a foam system to achieve a
process and the water-recycle ion exchange p r o c e s s . strontium decontamination factor of 1000. Cesium
T h e latter differs from the previously developed IS not removed in the foam system.
scavenging-precipitation ion exchange p r o c e s s '
m that it is designed to operate at neutral condi-
tions. Since very pure demineralized water is Scavenging-Precipitation Foam-Separation Process
produced in this process, recycle and reuse of this
water appear to be more economical than to d i s - Pilot Plant Studies. — The process flowsheets
charge it and continually demineralize fresh river for the scavenging-precipitation and foam-separa-
water. The s t e p s in this process are flocculation- tion s y s t e m s are shown in Fig. 3.21a and b. The
precipitation to remove colloids, demineralization scavenging-precipitation system was run contin-
by ion exchange, and sorption of ruthenium and uously under fixed operating conditions throughout
cobalt on activated charcoal. Initial t e s t s on a the t e s t program.
laboratory s c a l e have shown satisfactory r e s u l t s , Low-level waste from the ORNL equalization
demonstrating decontamination factors greater than b a s i n ^ ' was fed at 5 gpm to a flash mixer to which
1000. Semiengineering-scale t e s t s are scheduled Grundite clay was added continuously. Caustic,
for late in 1965. sodium carbonate, and copperas were added in a
second flash mixer to make the waste 0.005 M in
Foam separation has the desirable characteristic
NaOH, 0.005 M in N a ^ C O j , and 10 ppm in iron.
that separation efficiency is almost independent
The solution (pH 11.2) flowed to a flocculator
of concentration at very low concentrations. When
a metal ion is complexed with a surfactant in an
aqueous solution and the system is foamed by bub-
bling air through it, the metal ion is removed from "W. Davis, Jr., A. H. Kibbey, and E Schonfeld, Lab-
oratory Demonstration of the Two-Step Process for
the solution on the surface of the foam bubbles. Decontaminating Low-Radioactivity-Level Process Waste
The feasibility of applying foam separation to the Water by Scavengmg-Precipitation Foam Separation,
ORNL-3811 (in preparation).
specific problem of treatment of LLW was inves-
P . A. H a a s , Engineering Development of a Foam
tigated in the laboratory,'* and the development of Column for Countercurrent Surface-Liquid Extraction of
equipment for a large-scale separation plant was Surface-Active Solutes, ORNL-3527 (to be published).
L. J . King, A. Shimozato, and J . M. Holmes, Pilot
Plant Demonstration of the Removal of Activity from
Low-Level Process Wastes by a Scavenging-Precipita-
tion Foam Separation Process, ORNL 3803 (in prepa-
Maximum Permissible Body Burdens and Maximum ration).
Permissible Concentrations of Radionuclides m Air and
in Water for Occupational Exposure, U.S. Department of ^ ' K . E . Cowser, R. J. Morton, and E . J .,
Commerce, NBS Handbook 69 (June 5, 1959). " T h e Treatment of Large-Volume Low Level Waste by
the Lime-Soda Softening P r o c e s s , " Proc. U.N. Intern.
" F . L . Culler, Jr., et aU, Chem. Tecbnol. Div. Ann. Conf. Peaceful Uses At. Energy, 2nd, Geneva, 1958,
Progr. Rept. May 31, 1964, ORNL-3627. P / 2 3 5 4 , vol. 18, p. 161.

ORNL-DWG 65-9718






Fig. 3.21. Flowsheets for the Scavenging-Precipitation Foam-Separation Process for Decontaminating Low-
Level Radioactive Process Waste Water. N o t i c e that the two parts of the process ore c o n n e c t e d by a s i n g l e process
line, (a) S c a v e n g i n g - p r e c i p i t a t i o n s y s t e m , (b) F o a m - s e p a r a t i o n s y s t e m .

Table 3.9. Summary of Pilot Plant Tests for the Foam System

Countercurrent foam column: 2 ft square by 8 ft high

Recovery system: three 2-ft-square by 44-in.-high columns

Low-Level Waste T e s t s
Pall-Ring-:Filled Column
Tap Water Plus Reccjvery System
Tests Open Pall-Ring-FiUed
„ , b
Column Column Preliminary
Tests 93-hr Test

Liquid feed rate, gpm 2 1-2 1-2 1-4 2

Air rate to foam column, scfm 2-6 4-6 3-6 1-8 8

Air rate to each recovery s t a g e , scfm 4-6 1-3 1

DBS m feed, ppm'^ 25-220 55-400 100-470 47-490 120

DBS in bottom of column, ppm 10-130 10-290 35-350 61-413 128

DBS in condensed foam, ppm 1200-3850 1100-3600 1900-2900 1100-5600 1940

DBS in effluent, ppm 1-7 10-115 62

Volume reduction, feed/condensed foam 35-142 17-46 14-70 10-400 31

V/LD, cm"' 120-270 135-290 60-220 138

Strontium decontamination factor 2-200^ 10-200° 1-150° 105°

' D o e s not include decontamination factor for precipitation system (see T a b l e 3.10).
Foam column without recovery system.
" T h e s e items are set conditions; others are measured r e s u l t s
Includes both foam column and recovery columns.

where the hardness constituents (calcium and diam by 6-ft-high agitated sludge blanket clarifier,
magnesium), the iron, and some of the radioactive and 2-ft-diam by 33-in.-high Permutit vertical
ions precipitated a s the carbonate and hydroxides filters.''-2^'"
to form a floe which scavenged radiocolloids and The foam system (Fig. 3.216) was designed""
other colloidal impurities from the solution. The to have a capacity of at least 2 gpm of LLW, with
softened water was separated from the solids in a a possibility that the capacity might be as high a s
slowly agitated suspended-bed clarifier. The 5 gpm. Unlike the precipitation system, the oper-
solids were discharged as a slurry to the drain ating conditions for the foam system were varied
but could be collected by filtration in an actual during the t e s t program (Table 3.9). A solution of
waste-treatment plant for d i s p o s a l . When the foam-
separation equipment was being operated, the
desired amount of clarifier effluent was pumped 22
through a deep-bed anthracite filter for final polish- R. E. Brooksbank, et at., Low-Radioactivily-Level
Waste Treatment. Part II, Pilot Plant Demonstration
ing and then to the foam column. Otherwise, the of the Removal of Activity from I^oW'Level Process
clarified effluent was discarded. Except for the Waste by a Scavengmg-Precipitation Ion-Exchange Proc-
ess, ORNL-3349 (May 20, 1963).
additional flash mixer for adding Grundite clay, ^^R. E. Brooksbank, L. J . King, and J . T. Roberts,
the equipment for the precipitation system was the " P i l o t Plant Demonstration of the Removal of Radio-
activity from Low-Level P r o c e s s Waste Water by a
same as that used for the scavenging-precipitation Scavenging-Precipitation Ion-Exchange P r o c e s s , " pre-
ion exchange process,^^ a 5-gal Alsop flash mixer, sented at the Symposium on Radionuclide Exchange on
Soils, Minerals, and R e s i n s , American Chemical Society,
a 4-ft-diam by 3-ft-high agitated flocculator, a 4-ft- Philadelphia, P a . , April 10, 1964.

sodium dodecylbenzene sulfonate (5000 ppm DBS) exchange process, where the waste is made 0.01
was added in the feed line to achieve the desired M in NaOH and no Na CO is used. The decon-
concentrations in the feed. The feed was intro- tamination factor for strontium averaged 12, v/hile
duced at the top of the 2-ft-square foam column that for cesium was 1 without Grundite clay and
through a distributor consisting of six ^ - i n . - I P S about 9 with it. The averages of the turbidities
pipes with six 0.078-in.-diam orifices drilled along of the LLW and the filtered clarifier effluent were
each pipe. A small deflector plate below each 4.3 and 2.3 ppm, respectively, as equivalent SiO .
orifice prevented impingement of a liquid jet on Foam Separation. — Initial t e s t s were conducted
the foam. The feed solution drained down the at design conditions with tap water. The foam ap-
column countercurrent to about 8 ft of foam that peared to be s t a b l e , and the recovery system s a t i s -
was generated by bubbling air through nine 3-in.- factorily reduced the DBS concentration from 40
diam spinnerettes (each having 3949 holes 60 fi ppm in the feed to l e s s than 5 ppm in the effluent
in diameter) immersed in a 24-in.-deep pool of (Table 3.9). However, later t e s t s (see below)
water at the bottom of the column. The foam moved showed that actually up to 200 ppm of DBS was
up the column into a vertical foam-drying section required in the feed to achieve stable foam condi-
and then into a 2-ft-square by 5-ft-long horizontal tions and satisfactory decontamination values, and
foam-drying section. Liquid drained from the foam the recovery system was not designed to handle
and was collected and returned to the foam column this increased load. It was concluded that if
through a liquid distributor located under the feed properly matched equipment i s used, the DBS con-
distributor. The drained foam p a s s e d on to a centration can be reduced to satisfactorily low
centrifugal foam breaker, where it was condensed. values in effluent to be discharged.
When the surfactant recovery system was being The first t e s t s of the foam system with LLW
used, liquid from the bottom of the foam column were largely unsuccessful because of turbulence
was pumped continuously to the surfactant recovery and i n s t a b i l i t i e s . Varying the operating condition,
system where DBS was stripped both to conserve such a s reducing the feed rate from 2 to 1 gpm,
DBS and to reduce the DBS concentration in the increasing the DBS concentration, or disconnecting
treated w a s t e . This system c o n s i s t s of three 2- the recovery system did not establish steady-state
ft-square by 44-in.-high foam-stripping columns. conditions. Strontium decontamination factors
The water from the top column flowed by gravity varied from 2 to 200 (Table 3.9).
through the s t a g e s in s e r i e s and was then d i s - The column was filled with Pall-ring packing
carded. DBS was stripped as foam by bubbling air (1.5 X 1.5 in.) to act as baffles and to prevent
to each column through 9-in.-diam Micro Metallic channeling and circular turbulence. Further t e s t s ,
s t a i n l e s s s t e e l spargers immersed in a 24-in.-deep made in the foam column alone under a variety of
liquid pool. The average mean pore-size opening operating conditions, showed that steady-state
in the spargers is 10 (j. (grade G). The foam from conditions were achieved at a feed rate of 1 gpm,
each column was condensed while p a s s i n g through 200 ppm of DBS in the feed, and 4 cfm of air; the
an orifice, and the condensate collected in a single decontamination factor for strontium was 120 to
6-in.-diam glass column header. The recovery 200. A similar t e s t using the recovery column, with
columns were maintained at atmospheric pressure 1 cfm of air in each recovery s t a g e , also showed
and the glass column at '/, atm. s t e a d y - s t a t e operation with an average strontium
Scavenging Precipitation. — The scavenging- decontamination factor of 105 (Table 3.9).
precipitation portion of the plant was operated A final demonstration t e s t was made where
continuously for 91- and 45-day periods at a feed steady-state operation was maintained for 93 hr
rate of 5 gpm of LLW. Grundite clay was added (Tables 3.9 and 3.10). Both the liquid feed rate
only during the final 93-hr test at a rate of 0.77 and the air rate were doubled to 2 gpm and 8 cfm,
lb of clay per 1000 gal. respectively, to increase the processing rate while
The total hardness of the LLW averaged 109 ppm maintaining a constant extraction factor. The
as equivalent CaCO . The total hardness of the DBS concentration was decreased to 120 ppm.
clarifier effluent averaged 2 ppm and varied from The strontium decontamination factors across the
1 to 4. This is significantly lower than the hard- foam and precipitation system were 100 and 10,
n e s s in the product from the precipitation step respectively, giving an overall value of about
(6 to 10 ppm) in the scavenging-precipitation ion 1000.

Table 3.10. Decontamination Factors During the 93-hr Pilot Plant Test
of the Scavenging-Precipitation Foam-Separation Process

Foam Recovery Foam Scavenging-

Column System System Precipitation Overall

Average 35 3 105 10 1050

Sr 13
(range) (19-59) (2-4) (56-193) (8.5-10.8) (500-1500)
Average 0.9 0.9 1.0 8.8 7.6
Cs y
(range) (0.5-2) (0.8-1.3) (0.5-1.8) (5.13) (4-11)
Co 1.3

Sb 2.2

Ru 4.2

Zr-Nb >50

Ce >20

Cesium, antimony, and ruthenium were removed

- 555 A
only in the precipitation s t e p , and zirconium-
niobium only in the foam step, while strontium, 1
cobalt, and cerium were removed in both. The data OONLY FOAM COLUMN OPERATEC
for Sb, Ru, Co, Zr-Nb, and Ce were widely scattered, 200
L 30

s i n c e they were present in very low concentra-
tions. The concentrations of Sr, Cs, Ru, Co, Ce, 100
Sb, and Zr-Nb in the treated effluent were 2, 0.6, //
0.3, 0 . 1 , 0.2, 0.02, and 0.0001% of the maximum /
permissible concentration for water respectively. 50 f
d .
T h u s , this process is suitable for use at many n/
nuclear installations where low-level radioactive
w a s t e s contain comparable amounts of radioactivity. 20 /
Higher decontamination factors could probably be
attained in equipment designed to operate at higher
V/LD ratios ( s e e below). However, this could r // •
not be demonstrated in the present equipment be- 0
5 /-
c a u s e of limitation in the air supply and the c a p a c -
ity of the centrifuge.
Parameter V/LD. — The foam-column parameter / •
2 1
V/LD (where V and L are the gas and liquid rates
in the same units, and D is the effective bubble » •
diameter) is related to the extraction factor and
50 • 100 200 <000
h a s been shown to be useful a s a b a s i s for process V / L D (cm-')
control in laboratory t e s t s . ' * Figure 3.22 shows
the relationship between the strontium decontami- Fig. 3.22. Relationship Between the Decontamination
nation factor across the countercurrent foam column Factor for Strontium Across the 2-ft-square X 8-ft-high
and the parameter V/LD for all of the pilot plant Countercurrent Foam-Separation Column and the Operating
t e s t s . T h e s e data are in good agreement with the Parameter V/LD for All Pilot Plant T e s t s .

laboratory r e s u l t s . However, the utility of this foam s t e p . ) In this way a cesium DF of 20 was
parameter as a b a s i s for process control is limited achieved (Table 3.11). The height equivalent to a
by a lack of knowledge concerning the relation- theoretical s t a g e for strontium removal in the foam
ship between effective bubble diameter and other column was about 3 cm.
process variables. Flow rates, solution tempera- The crucial factor governing satisfactory stron-
ture, gas distributor parameters, and concentrations tium removal appears to be the regulation of gas
of the surfactant and other constituents all affect and liquid flow rates (i.e., V and L respectively)
the effective bubble diameter. Figure 3.23 shows and the average bubble diameter D, such that the
the relationship that existed in the pilot plant be- ratio V/LD will exceed 150. The V/LD for the
tween air flow rates and effective bubble diameter. above»mentioned demonstration runs was in the
The line indicates only the general tendency for range 200 to 270, with an average bubble diameter
bubbles to get larger at higher gas flow rates. of 0.06 to 0.08 cm. The liquid throughput was 40
Laboratory Studies. — The two-step scavenging- to 42 gal f t - 2 h r - ^ The effect of the V/LD
precipitation foam-separation process *'^^ for parameter on the strontium DF is shown in Fig.
decontaminating ORNL low-level radioactive 3.24. These data represent flowsheet runs wherein
process waste water was successfully demon- total phosphorus concentration in the water varied
strated on a laboratory s c a l e with the foam-recycle from < 0 . 1 to ' ^ 1 5 ppm, as orthophosphate, and
option omitted. First, most of the hardness, include actual waste water a s well as synthetic
initially 110 ppm as CaCO , and most of the radia- tap water solutions. No distinction could be made
tion emitters are precipitated in a suspended-bed between the strontium decontamination factor data
clarifier (76 mm in inner diameter and 25 cm high) obtained with columns that were 53 and 80 cm in
by making the water 0.005 M each in NaOH and length. The large amount of data involved in
Na CO in the presence of 6 to 20 ppm of ferric these calculations was programmed for processing
ion coagulant. Residual soluble hardness at this by the CDC 1604 computer.
point is l e s s than 5 ppm, which is low enough to The high-speed digital computer was also used
permit efficient strontium removal in the sub- for calculating the volume reduction factor (VR),
sequent foam-separation s t e p . Next, the almost another p r o c e s s variable that h a s direct bearing
clear supernatant is pumped to a countercurrent on process cost. The strontium DF, which is a
foam column (37 mm in inner diameter and 80 cm function of foam density and linear velocity of the
high, exclusive of 10-cm-ID x 10-cm-high foam-
drainage section; the foam phase in the counter-
current section was about 50 cm in height) where ORNL-DWG 6 5 - 5 5 4 A
dodecylbenzene sulfonate surfactant is added
simultaneously to the top and bottom of the column
in a ratio of about 5:1 to a concentration of roughly
90 ppm. Most of the remaining radioactivity is t ^ ^
removed in the foam generated by bubbling air < c ^^
through the solution. Overall decontamination \. ppm DBS
-I (.5
factors (DF's) obtained for strontium, cobalt, m 1 • ^ IN POOL
ruthenium, and cerium were >3.7 x 10^, 2.3 to 4, 3
m ^^% \. o 50
-^^^ a • <00
2 to 5, and 50 to 180 respectively. The process is T a (50
> 1.0
a l s o excellent for removing cesium, provided that * 200
I D 250
about 60 ppm of baked (600°C for 20 min) Grundite LU • 300
clay i s added to the water during the hardness- V 400
precipitation step. (Cesium is not removed in the 3 4 5 6 7 8

Fig. 3.23. Relationship Between the Effective

'^^Chem. Technol. Div. Ann. Progr. Rept. May 31,
Diameter of Bubbles in the 2-ft-square Countercurrent
1963, ORNL-3452, p. 86.
Foam-Separation Column and the Flow Rate of Air
^^W. Davis, J r . , A. H. Kibbey, and E. Schonfeld, Lab-
oratory Demonstration of the TwoStep Process for Through Nine 3-in.-diam Spinnerettes for all Pilot Plant
Decontaminating Low-Radioactivity-Level Process Tests. C o n c e n t r a t i o n of the s u r f a c t a n t DBS is I n c l u d e d
Waste Water by Scavenging-Precipitation and Foam
Separation, ORNL-3811 (in preparation). OS a parameter.
T a b l e 3 . 1 1 . Decontamination of Spiked O R N L LLW by the Foam Process

Feed: 0.005 M each NaOH and Na CO 1-^60 ppm Grundite clay ('^0.5 lb/10 gal) added m one run only;
3— -6 3
PO , < 0 . 7 ppm in absence of Grundite and 2.4 ppm with Grundite present
F l o w s : Surfactant feed ^ 9 0 to 100 ppm in foam column, added m a split stream with the ratio top/bottom = 5 to 5.5/1
2 1
Throughputs: Sludge column = 12 to 13 gal ft hr
2 —1
Foam column = 40 to 42 gal ft hr
Bubble diameter = 0.06 to 0.08 cm
V/LD: With no Grundite, 273 c m ~ ' ; with Grundite, 202 c m ~ '
Decontamma ition Factors
Sludge C.Dlumn Foam Column Overall
With Grundite No Grundite With Grundite No Grundite With Grundite No Grundite
Filtered^ Unfiltered^ Filtered Unfiltered Filtered'' Unfiltered Filtered'' Unfiltered Filtered Unfiltered Filtered Unfiltered

hardness 32 6 40 <8 -^^1.1 ^9 -^1 >9 41 53 30 43

Ca 29 5 34 <8 -^1.3 •^g ~1 >10 39 49 27 >40

«^Sr 20 16 70.5 42 338 212 12 13 7935 3725 630 705

'3'Cs 25 21 1.6 1.3 -x,! 1 1 1 21 20 1.3 1.3

eOCo 2.5 2.9 4.5 4.1 1.3 1.1 1.1 1.0 2.4 2.3 3.5 4

lO^Ru 1.7 1.8 1.3 1.5 3.5 3.8 1.4 1.3 4 6 1.8 2.0

'"^Ce 44 24 >775 75 6.4 4.4 -^2.2 1.3 280 118 >49 118.6

Zr-Nb (These a c t i v i t i e s were too low to measure)

F i l t e r e d effluents are expected to give higher D F ' s than unfiltered effluents, owing to removal of precipitated s o l i d s ; discrepancies are attributable
to the standard deviations of the analytical and sampling techniques.
T h e s e D F ' s are for soluble Ca and t o t a l h a r d n e s s ions only (no solids removal); this value i s the ratio of filtered sludge effluent to filtered
foam effluent a n a l y s e s .

ORNL-DWG 65-97(9 of the dye-DBS complex. The few-parts-per-mil-

D COLUMM HEIGHT, 8 0 cm, TAP WATER, FOAM CONDENSATE RECYCLED lion concentration is read directly from the stand-
ard color wheel of a Hellige comparator. Maximum
• • SAME AS ABOVE EXCEPT LLW USED INSTEAD OF TAP WATER accuracy (within 10%) for the method was attained
FEED in the neighborhood of 1 ppm. The greatest error
G GRUNDITE CLAY USED ( 0 5 l b / ( 0 0 0 gal) (about 50%) occurred at the still lower concentra-
P HIGH P o / " CONCENTRATE ( 5 TO (5 ppm ), IN ALL OTHER tion range of 0.1 to 0.2 ppm. The method was
CASES, 0 TO 5 ppm
adequate for DBS concentrations up to 2 ppm and
g was used for process control in the ORNL foam-
1 separation pilot plant.
/ A continued search for better surfactants indi-
i cated that the purportedly more biodegradable sur-
• 1 , c3 factants Nacconol 40 FX (National Aniline Divi-

r NF
sion, Allied Chemical Corporation) and Alipal LO
436 (General Aniline and Film Corporation) ex-
hibit desirable physical and chemical properties
// under foam-separation process conditions, and are
/ as good as the currently used DBS in effecting
/ strontium decontamination, for example, values of
( F / x ) 2 + °^ '•^^ order of 1.5 x 10~^ cm at a sur-
° 1 factant concentration of 0.5 times the critical
micelle concentration and at a pH of 4.
• NR
• 1 Determination of Nuclides in Solutions Con-
1 - taining Low Levels of Radioactivity. — One of
J ^r
the most difficult and expensive analyses a s s o c i -
p • B
ated with radioactive waste disposal is the
./ • determination of the amount of each of four to

f' . .
— ten radioactive nuclides when present at the very
low radioactivity levels of 10^ to 10^ /M/:ic/liter.
p •
1 The determination of nuclide concentrations by
least-squares resolution of the gamma-ray spectra
lO' 2 5 10^ 2 5 10^
V/LD with a high-speed computer was tested and was
found to provide a rapid, simple, and reasonably
Fig. 3.24. Strontium Decontamination Factor as a accurate method for these a n a l y s e s at a cost of
Function of V/LD, 5 to 25 times l e s s than that for conventional
methods.^* The savings increase with the number
of samples. The t e s t s were made with solutions
foam, varies inversely as both the VR and the foam of ^°Co, *'Sr, ^"^Ce, and ^"''Ce added a s tracers
column throughput rate, but is directly proportional to water containing ' ° S r , ^^Zr-Nb, ' " ^ R u , ' ^ ' ' C s ,
to the strontium distribution coefficient (F/x). and ' ^ ^ C s . The lowest concentration of radio-
Calcomp computer plots of VR vs strontium DF activity was about one-tenth of the lowest value
for various liquid throughput rates are given in previously determined by this technique, that i s ,
F i g . 3.25 for an assumed strontium distribution down to 10 counts/min. Evaporation by factors
coefficient of 3.4 x 10~^ c m ~ ^ up to 20 was also evaluated for improving accu-
racies of a n a l y s e s .
A limited support program for the foam-separation
pilot plant was also carried out. The need for a
cheap, rapid method for determining the concentra- ^*E. Schonfeld, A. H. Kibbey, and W. Davis, Jr., Deter-
tion of dodecylbenzene sulfonate (DBS) led to mination of Nuclide Concentrations in Solutions Con-
taining Low Levels of Radioactivity by Least-Squares
testing a methylene-blue colorimetric analytical Resolutions of the Gamma-Ray Spectra, ORNL-3744
method which u s e s a single chloroform extraction (January 1965).

ORNL-DWG 6 5 - 9 7 2 0

I— \^— ^1^ II
4= 4\
V- I ™v\ ^
\ \ 1 I

\ \ 1
\ \ \

(56 gal f - 2 h |(04 \5? \ (3

- ^
\ ^
T \

4\ I \
r\ \

1 ' 1
5 Q
1- \
< 2 \ \
z. in 4
2 F -\^ \1
if ^ j-
: \ T: \
\ \ tI \
\ \ \^ \
\ \^ 1I \
\ \ i I \
\ \ \ \ \t
-A\ N\ —> -\ H
5 \ \ ^3
\ _ „"_ \1
\ \
\ \ V \ \
\ \ ^ \, \, ^
V ^s^ ^^^
"^^ - = = ^ ^ ^ ^
10' (0^ 2 5 (0^ (0^

Fig. 3.25. Strontium Decontamination Factor as a Function of Volume Reduction.

For comparison, the solutions were also analyzed of the total activity and if ^'Sr is a minor com-
by standard radiochemical methods. Agreement be- ponent.
tween the two methods was excellent for ®°Co and
^^Sr for concentrations down to about 500 /x/^c/liter Water-Recycle Process
(scatter of ± 5 to ±20%, F i g . 3.26). Agreement
between the two methods was good for ^ ' C s at The objective in developing a water-recycle
activities down to 3000 to 4000 /x//c/liter. More process is to provide a water-treatment system to
dilute solutions were successfully analyzed only treat low-level radioactive waste with a low s a l t
if they were first concentrated by evaporation. content and to return deionized, well-decontam-
The method is not recommended for determination inated water to the system, thus providing a closed
of ' ° ^ R u at activities below 1000 to 2000 niic/li- circuit. Fresh water would be demineralized before
ter, unless this nuclide constitutes a major fraction introduction into the circuit (Fig. 3.27).

Status and Progress. — So far, the prospects and total rare earths). The D F ' s for ' °*Ru and
look very promising. The process, still in the ^''Co were 10 to 1000 times a s high a s previous
laboratory stage of development, h a s provided values obtained in the ORNL development program.
decontamination factors ( D F ' s ) ranging from 100 Next, the process will be tested in a micro pilot
to 10,000 using low-level radioactive w a s t e water plant with water that simulates recycled process
(LLW) (in the hundreds for ^ " " C e , ' ^ 4 , ' °«Ru, water; that i s , it will represent water that h a s
^ " S b , and '=Zr-Nb; in the thousands for ' ^ ' C s been demineralized and then contaminated through
and ' ° S r ; and in the ten-thousand range for ® Co plant u s e . T h e results will be used to predict
LLW processing c o s t s in full-scale plants. In
the past, micro pilot plant results have been
ORNL-OWG 6 5 - 8164 largely confirmed in large-scale pilot plant t e s t s .
Experimental Work and Results. — In i t s present
s t a t e , the process c o n s i s t s in zeta-potential-
controUed ( Z P ) batch clarification, cation-anion
/ exchange, and sorption by granular activated
carbon ( F i g . 3.28). Detailed studies are being
/ made of coagulants and coagulant aids that will
r ^ produce optimum clarification for a continuous
S 6
ho' The best clarification (jar t e s t s ) was obtained
when the water was made 20 to 30 ppm in Al2(S0^)j
• ' and 0.5 ppm in SiO^, in that order ( F i g . 3.29).
(0" / Here, the average Z P of the suspended solids
in the waste changed from i t s original —14 to
/ —17 mv to + 2 . 5 mv. There were only 20 particles
(0"' (0° (O' (0^ (0^ lO"* per cm^ of membrane filter area per ml of clarified
ACTIVITY OF ^^Sr BY MANUAL-STRIPPING COMPARISON waste, contrasted to 300 for the raw waste.
WITH STANDARDS (dis m m - ' m l " ' )
Demineralization capacity (0.1% of ionic break-
F i g . 3 . 2 6 . Comparison of Two Methods for Estimating through a s measured by specific conductance)
85 with mixed-bed ion exchange (Dowex SOW—
th< Sr Content of a Solution.

O R N L - DWG 6 5 - 9 7 2 (





Fig. 3.27. Water-Recycle F l o w s h e e t .


ORNL-DWG 65-1734A
ALUM 2 0 - 3 0 ppm Al2lS04)3

6 VOL, 6 M HNO,
1 0 VOL
\ ( 0 0 - 1 3 0 ppm AS CaC03


16 VOL, 2 M NaOH
16 VOL
8 VOL, 2 M NoOH

0 8 - 2 4 VOL SLUDGE ( 3 VOL



0 0 3 - 0 0 6 VOL WET CAKE



Fig. 3.28. Simulated Water-Recycle F l o w s h e e t - Laboratory Development Using O R N L L o w - L e v e l Waste.

Dowex 1) w a s about 470 bed volumes of clarified * Co w a s removed when a column of activated
LLW (total h a r d n e s s : 100 to 130 ppm a s CaCOg) carbon formed a part of the system. A s expected,
per bed volume (BV) of cation resin; with ORNL ion exchange alone removed most of the ^^^Cs
tap water, 560 B V ' s . T h u s , in the c a s e of ORNL and ' Sr. Specific conductance can be used a s
w a s t e , the salt content of the water w a s increased a simple indicator of bed exhaustion b e c a u s e ionic
by about 20% during plant u s e , producing a cor- breakthrough precedes breakthrough of the radio-
responding 20% decrease in demineralization elements (Fig. 3.30).
capacity. Hence, an increase in demineralization With separate beds, the cation column was regen-
capacity to 2500 BV's would be predicted when erated with 6 BV's of 6 W HNO^, leaving only
treating low-level waste that contained only the 0.1% of the radioelements on the resin. Regen-
s a l t s added through plant u s e ( F i g . 3.27). As eration of the anion column w a s not s o efficient:
expected, the u s e of separate cation and anion e x - 10 BV's of 2 M NaOH left 1% behind. Other
change columns yielded a capacity equal to about regenerants may be needed.
90% of that for a mixed-bed column and resulted in a To adapt the process to continuous operation,
lower-quality (but still satisfactory) product water a study is being made of coagulant aids for im-
(0.67 micromho/cm, compared with 0.33 for the proving the settling properties of the light alum
mixed-bed exchanger). Economically, separate floe. T h e change in the mutually repelling forces
beds may be desirable because regeneration related to the Z P of the suspensoids appears to
is easier. be an important criterion, since the neutralization
The overall D F ' s achieved with the present of t h e s e forces is e s s e n t i a l to achieving optimum
system ranged from 10^ to 10"* (Table 3.12). A coagulation and clarification.^'' T h e effects of
large part of the D F ' s were obtained by clarifi-
cation and ion exchange only. Significantly more ' T . M. Riddick, Tappi, 47, 171 (1964).

ORNL-DWG 65 - 9 7 2 2


< 2


H <

(00 2 0 ppm ALUM_

+ 3 ppm S1O2 + 2 0 ppm A L U M

a. + 2 0 ppm A L U M + 20ppmALUM +20ppmALUM
+ 2 ppm S1O2 + 1 ppm SiOg +05ppmSi02

-15 -13 4- -2

Fig. 3.29. Solids Filtered Out of Treated Waste Supernatant onto a Membrane Filter at a Zeta Potential of

-2.5 mv. Jor-test conditions: 3 0 m m of f l o c c u l a t i o n f o l l o w e d by 3 0 m m of s e t t l i n g .

some coagulant aids on the Z P are summarized 3 . Activated s i l i c a , which can be added to toughen
below: the fragile alum floe, increases the amount of
1. Alum neutralizes the normally negative Z P of alum needed to attain zero Z P . The Z P of
LLW (usually ranging from —12 to —18 mv). activated s i l i c a in tap water is about —15 mv,
Sufficient alum (and other coagulant aids) which is about the same as that for the s u s -
should be added to bring the Z P to zero, with pensoids in LLW.
a tolerance of ± 5 mv.
2 . Grundite clay gives bulk to the light alum floe
and slightly decreases the amount of alum sub- 4. An organic polyelectrolyte such a s Primafloc
sequently needed to attain zero Z P . The Z P C-3 (Rohm and Haas Company, Philadelphia,
of Grundite clay in tap water was about —10.8 P a . ) , which can be added during the floccula-
mv, which is l e s s negative than the average tion s t e p to a s s i s t in agglomerating the floe,
Z P of the suspensoids in LLW. Also, Grundite produces greater changes in Z P upon small
clay sorbs cesium selectively.^^ additions to the waste than those produced by
inorganic coagulants such as alum. Only 0.5
ppm of Primafloc C-3 shifted the Z P from about
Chem. Technol. Div. Ann. Progr. Rept. May 31,
1964, ORNL-3627, p. 95. - 6 to + 3 mv.

T a b l e 3 . 1 2 . Decontamination Factors Obtained in the Water-Recycle Process Using O R N L Process Waste

Overall D F ' s of 10^ to lo'*

Total Bulk
Individual Isotopes Gross Gross
Rare Ions
^^'•Ce «Vo ^"Cs l"l '°«Ru l^^Sb '°Sr "Zr-^Nb Earths ^'""'"^ ^^^^ (Specific

Raw waste a n a l y s i s , dis m i n ~ m l ~ or counts min *"

P r o c e s s 1° 11.3 136 12.2 25.4 0.8 29.1 1.2 15.1 77 52 330'

P r o c e s s 2^ 10.1 70.9 23.6 126.2 44.2 3.5 52.0 0.4 32.0 52 85 290'

Overall decontamination factors

P r o c e s s l'^ 150 6500 370 200 2 0* 1500 45* 1500 390 870 1000

Process 2 340 12,000 2600 430 200 300 6300 160® 11,000 130 1200 500

Alum-silica clarification, mixed-bed ion exchange, coconut-shell activated carbon. D F V a l u e s for 0.1% specific
conductance mce breakthrough.
Not determined.
The dimensions of specific conductance are microrahos/cm.
d Alum-silica clarification, separate-bed ion exchange, bituminous coal activated carbon. DF Values for 0.2% s p e -
c i f i c -conductance breakthrough.
T h e s e are minimum v a l u e s , since the radioactivity in the w a s t e water w a s reduced to the analytical limits of
detection (background).

Selective Sorption of Phosphates T h i s study can lead to a broader application:

on Activated Alumina the removal of phosphate from water at city water-
treatment p l a n t s . For example, household c l e a n i n g
The economical operation of the Scavenging- formulations soon will no longer contain practi-
Precipitation Ion Exchange p r o c e s s depends on cally indestructible d e t e r g e n t s , but many will s t i l l
keeping the concentration of phosphates in the contain the phosphate builders. T h e s e builders
feed water below about 1 p p m . ^ ' ' ^ ° Sorption of promote i n c r e a s e d biological activity that in turn
t h e s e phosphates on a column of activated alumina r e s u l t s in u n p l e a s a n t t a s t e and odor.^^ Also,
a s a first s t e p in the p r o c e s s e n s u r e s complete phosphates complicate water clarification and
precipitation of the h a r d n e s s ions in the s c a v e n g - softening.^ ^ ~ ^ ^
ing s t e p . ^^ Recent work, reported below, shows Experimental Work and Results. — The preferred
that Alcoa F - 1 activated alumina is the c h e a p e s t process c a l l s for downflow of the feed water
and most useful sorbent.^^ through a column of the activated alumina. How-

Ibid., p. 9 3 - 9 5 .
^ F . W. Gilcreas, " S y n t h e t i c Detergents and Water
30 Q u a l i t y , " Water Works and Waste Engineering 2, 66
R. R. Holcomb, Low-Radioactive-Level Waste Treat- (1965).
ment. Parti. Laboratory Development of a Scavenging-
Precipitation Ion-Exchange Process for Decontamination ^^R. S. Smith, J . M. Cohen, and G. Walton, " E f f e c t s
of Process Water Waste, ORNL-3322 (June 25, 1963); of Synthetic Detergents on Water C o a g u l a t i o n . " / . AWWA
R. E . Brooksbank et al., Low-Radioactive-Level Waste 48, 55 (1956).
Treatment. Part 2. Pilot Plant Demonstration of the
Removal of Activity from Low-Level Process Wastes ^ ' ' j . J . Morgen and R. S. Engelbrecht, " E f f e c t s of
by a Scavengmg-Precipitation Ion-Exchange Process, P h o s p h a t e s on Coagulation and Sedimentation of Turbid
ORNL-3349 (May 13, 1963). W a t e r s , " / . AWWA 52, 1303 (1960).
31 ^^J. F . Malina, J r . , and S. Tiyaporn, " E f f e c t s of
W. C. Y e e , The Selective Removal of Mixed Phos- Synthetic Detergents on Lime-Soda Ash T r e a t m e n t , "
phates from Water Streams by Activated Alumina, ORNL- J. AWWA 56, 727 (1964).
TM-1135 (in p r e s s ) .

ORNL-DWG 65-907A
80% of the phosphates was removed initially
<oo '
- (Fig. 3.32). Apparently, saturation capacity had
not been reached after almost 40,000 BV's of
50 water had passed through the column. Although
1 complete breakthrough was evident at 33,000
1 ^-bPtUII-IL BV's, phosphate sorption resumed. It is possible
O 20 1 that the composition of the process water had
1 changed to a higher proportion of those forms of
1 phosphate for which the bed has a h i ^ e r capacity
10 1
than for the ortho form (Table 3.13).
I S b , , _ a b , Jb 1
In a similar upflow test but with Alcoa H-51
—-f 1 / alumina (surface area, 400 m^/g), the maximum
I 1
Na* + / breakthrough was 75% at 33,000 B V ' s , and 100%
< 1 / breakthrough was not reached (Fig. 3.32).
LlJ 1 The column is regenerated by first passing 1 M

NaOH through it to elute the phosphates, rinsing

:a2 + with water, p a s s i n g 1 M HNO^ through to restore
60co "^ the alumina to the acid form, and finally rinsing
106 o ^ ^
^ Ru-.^ ^ ^ .^r• 1 /;
.3- 1 // with water. ^ ' The regenerating solutions and
> 1i , washes result in a volume equal to 12 BV's
o 0 5 - y <i
1 /' (Fig. 3.33). This leads to a favorable volume
y 1
reduction factor (BV of product water per BV of
regenerant waste) of about 2000. This value is
1 // based on a 1-ppm concentration of phosphates in
0 2
the feed. A volume reduction factor of 20 is con-

0 1
200 400 600 800
WASTE PROCESSED (cation resin volumes)
ORNL-OWG 6 5 - 4 7 6 0 A

Fig. 3.30. Breakthrough of Bulk Ions and Radio-

nuclides. Specific-conductance measurements signal
imminent breakthrough of radionuclides because bulk
ions break through first.

ever, if column plugging from suspended solids

becomes a problem, then the less-efficient method
of upflow through a fluidized bed can be used.
The column can be regenerated with NaOH and
HNO solutions, in that order.
In downflow operation with tap-water solutions
containing 5 ppm of phosphates (expressed a s
PO 3 - ) more than 99% of NaH^PO^ was removed
4 ^* 2 4
from 3000 bed volumes (BV) by Alcoa F-1 alumina
(surface area, 250 m^/g), while the same per-
centage of Na^P^O^, N a ^ P j O j g , and (NaP03)g ^ 0 2 4 6 8 10 (XIO'')
BED V O L U M E S (1 BV = 2 m l )
was removed from 4000 to 5000 BV's (Fig. 3.31).
Upflow operation in a fluidized bed was l e s s
Fig. 3.31. Phosphate Sorption on Alcoa F-1 Activated
efficient. In upflow operation with actual process
Alumina. The alumina has a lower capacity for ortho-
water that contained an average of 1 ppm of
phosphate than for its dehydrated forms.
mixed phosphates and suspended s o l i d s , only

ORNL-DWG 65-^853A

o 5
to 2
n or
UJ z 1 : . ^ ^ .-^^^ '^ «-•


s (A
3 1.0
() H

o 0 4

Q 0.2

o 0
f!? 80
UJ 60
u. 40
n •7'
tn 20
20,000 30,000 40,000
BED VOLUMES ( ( B V = 2 0 0 m l )

Fig. 3.32. Alumina Sorption of Mixed Phosphotes Contained in Raw L o w - L e v e l Radioactive Waste Water by
Upflow Through a F l u i d i z e d Bed.

Table 3.13. Comparison of the Capacities of the Individual Phosphates on Alcoa F-1 Activated Alumina
with the Amount of Mixed Phosphates Sorbed from L o w - L e v e l Waste ( L L W ) by Upflow
Through a Fluidized Bed of F-1 Material

Maximum Volume of F e e d Solution Used

Equivalent Volume
F e e d Solution Phosphate
Form of Phosphate Concentration Mode of Operation Actual If Only 1 ppm Capacity
(ppm as PO ~ ) Volume of P h o s p h a t e mg a s PO
(BV's) (BV's)
g F-1 Alumina

Ortho, N a H j P O ^ Packed bed, downflow 9,500 47,500

2 min residence time

Pyro, Na^jP^O^ 5 Same 11,000 55,000 >30

Tripoly, Na^PgO^g 5 Same 12,000 60,000 >30

Hexameta, ( N a P O , ) „ ^, , 5 Same 11,000 55,000 >25

3 8.4(av)
Mixed phosphates in LLW l(av) F l u i d i z e d bed, upflow 40,000 40,000 20
2 r e s i d e n c e time

ORNL-DWG 65 - 4 7 5 9 A T a b l e 3 14 Alumina Evaluation for Sorption of Mixed

Phosphates from Raw L o w - L e v e l Radioactive Waste

Conditions Upflow through a fluidized bed
Grade of Alumina''

Alcoa Alcoa Alcoa

F-1 F-20 H-51

Cost, if/lb 12 100 40

Surface area, m / g 250 250 400

Sorption capacity for mixed

phosphates, mg as PO ~ per
g of alumina 19 17 >27

Mixed phosphates eluted,

during regeneration, mg ds
3 -
PO per g of alumina 17 10 28

Regeneration-regenerant waste,
bed volumes 12 16 12

Amount of alumina dissolved

during regeneration, % of
original alumina 35

Products of the Aluminum Company of America, P i t t s -

burgh, Pa
Previously studied (see ref 31)

I 2 3 4


from the processing of power reactor fuels will
F i g . 3.33. Regeneration of Three Grades of Activated c o n s i s t in converting them to solids and disposing
Alumina. Two cycles of 1 M NaOH and one of 1 M of them permanently in s a l t mines, a possible
HNO^ were sufficient to regenerate the F 1 and the
alternative is long-term storage as liquids in
H 51 a l u m i n a .
tanks. As estimates of total fuel-cycle c o s t s
for advanced reactor concepts decline from previ-
ous highs of several mills/kwhr (electrical) to
l e s s than 1 mill/kwhr (electrical), the cost of
Sidered the minimum, according to water-treatment about 0 03 miU/kwhr (electrical) allocated to
workers. ^'^ waste management becomes increasingly signif-
A summary of the c o s t s , surface a r e a s , e t c . , icant. For these r e a s o n s , as well as for the
of the aluminas tested is given in Table 3.14. purpose of establishing a b a s i s for comparison
of the c o s t s of alternate waste management
schemes, a detailed estimate of the c o s t s of
3.4 ENGINEERING, ECONOMIC, AND SAFETY " p e r p e t u a l " tank storage of w a s t e s is being
EVALUATION undertaken.
The b a s i s of the study, the conceptual design
Although it is believed that the most responsible
of the tank farm, and the capital c o s t s that have
method of managing very radioactive liquid w a s t e s
been developed are summarized below. Detailed
cost data are being assembled for use in a com-
P e r s o n a l communication from D G Stephan, Ad- puter code that c a l c u l a t e s total storage c o s t s a s
vanced Water Treatment Program, Robert A. Taft Sanitary
Engineering Center, Cincinnati, Ohio. a function of tank s i z e for a given filling rate,

life expectancy of the tank, interest on capital, materials that they are neutralized for storage.
and return on investment. It is assumed that 0.1% of the gross fission
products and 50% of the ' ^ ^Cs appear in t h e s e
w a s t e s , together with induced ^^Zr-Nb and *°Co
Basis of Study
activity. They are stored separately in million-
As a b a s i s for this study, a fuel processing gallon tanks, the largest possible without requiring
plant having a nominal capacity of six tons of mechanical means for removing the decay h e a t .
fuel per day is assumed, handling the fuel from Purex and Thorex high-radioactivity solvent
an installed capacity of 22,400 electrical mega- extraction raffinate wastes are combined and
watts (70,000 thermal megawatts). The plant stored in tanks equipped with submerged cooling
p r o c e s s e s 1500 t o n s / y e a r of uranium converter coils. Storage of the raffinates under both acid
fuels irradiated to 10,000 Mwd/ton and 270 and alkaline conditions is considered.
t o n s / y e a r of thorium converter fuel irradiated It is assumed that the w a s t e s are produced con-
to 20,000 Mwd/ton. Of the uranium converter tinuously and are accumulated in a tank farm
fuels, 918 t o n s / y e a r are clad with 0.025-in.-thick during 20 years of plant operation. At the end of
Zircaloy, and 582 t o n s / y e a r are clad with 0.015- this accumulation period, the tank farm containing
in.-thick s t a i n l e s s s t e e l . 4,080,000 gal of highly radioactive acid w a s t e ,
In processing t h e s e fuels 120 days after d i s - or 24,480,000 gal of highly radioactive alkaline
charge from the reactors, w a s t e s of two general w a s t e , and 65,800,000 gal of alkaline cladding
categories are produced, each requiring e s s e n t i a l l y w a s t e s , is maintained for " p e r p e t u a l " containment.
permanent containment (Table 3.15). Cladding During the 20 years of waste accumulation, the
w a s t e s , consisting of Zirflex (from Zircaloy d i s - total fission product heat generation rate in the
solution) and Sulfex (from s t a i n l e s s s t e e l dissolu- tank farm rises to 1.03 x 10^ Btu/hr. Assuming
tion) w a s t e s , are so corrosive to ordinary structural no further accumulation after that time, the rate

Table 3.15. C h a r a c t e r i s t i c s of Wastes

Cladding Wastes Raffinates

Zirflex Sulfex Purex Thorex

Volume, gal per kg of fuel

Acid 1.7 1.34 100 200

Neutralized 1.8 2.0 600 1200

Production rate, g a l / y e a r
Acid 2 x 10^ 7.8 X 10^ 150,000 54,000
Neutralized 2.15 X 10® 1.14 X 10® 900,000 324,000

Radioactivity after 120 days of decay,

Acid 38 28 37,000 37,000
Neutralized 36 19 6,200 6,200

Heat generation rate after 120 days of decay,

Btu hr~ g a l ~

Acid 0.42 0.3 433 433

Neutralized 0.4 0.2 72 72


drops by a factor of 4 over the next ten y e a r s , waste area. The cladding waste are stored in
and d e c r e a s e s to insignificant levels over the alkaline form in tanks of about 1.1-million-gallon
next two to three centuries. capacity. Cooling is not provided for t h e s e
w a s t e s ; each tank is ventilated individually
through a filter at the surface. In addition to
Description of Facility filled tanks, an empty tank in each area is always
maintained on standby to receive the contents of
After the desired adjustments of volume and any tank which may have failed.
acidity or alkalinity have been made in the The tanks are similar in design to those in u s e
fuel processing plant, the wastes are transferred at the Savannah River Plant. They are cylinders
through concrete-encased s t a i n l e s s s t e e l pipes having a diameter-to-height ratio of 3 , are made
to a tank farm for long-term storage. In the design of 1/2- to 1-in.-thick s t e e l plate, and are housed
and operation of this tank farm, considerations of in steel-lined concrete vaults with walls 3 to 3 /
safety took precedence over any potential savings ft thick, buried under about 10 ft of earth. Stain-
in cost. Figure 3.34 is a conceptual layout of a l e s s s t e e l is used to contain acid w a s t e s ; carbon
" c o m p l e t e d " tank farm, containing a 20-year s t e e l is used for alkaline w a s t e s . The tanks are
accumulation of w a s t e . The farm is divided into constructed in accordance with the ASME code
two a r e a s , one containing tanks of high-level radio- for unfired pressure v e s s e l s , and all welds are
active waste and their associated cooling and completely radiographed. As to the mild s t e e l
ventilation facilities, and the other containing tanks, the welds are stress-relieved by heating at
tanks of cladding w a s t e s . In this farm the high- 1000°F for 1 hr.
level radioactive w a s t e s are stored a s acid solu- Heated air is circulated through the annular
tions in million-gallon tanks grouped around three space between the tank and vault for dehumidifi-
s i d e s of an operations building containing many cation. The occurrence and severity of leaks are
of the major equipment items of the cooling and determined by monitoring for both liquid and air-
ventilation s y s t e m s . A cooling tower and pumps, borne radioactivity in the annulus. The tanks
an emergency water storage tank, a water surge have steel-lined interval columns for support,
tank, and a stack and fans are a l s o located in this and, for storing highly radioactive waste, they





>\ I

Fig. 3.34. Layout of Tank Farm for Storing Waste.


are equipped with cooling coils which serve as to a secondary water cooling circuit, and then
the primary means for removal of radioactive recycled through a 250,000-gal surge tank (500,000
decay heat. A water-cooled condenser located gal for an alkaline tank farm) to the tank c o i l s .
in the operations building s e r v e s a s the secondary Thirteen heat exchangers and one spare are pro-
means for heat removal, and steam produced from vided, each with a rated heat duty of 9 x 10®
self-boiling w a s t e s during emergencies is vented Btu/hr. They are of the shell and tube type, with
to this condenser through a 4-ft-diam off-gas header external dimensions about 5 ft in diameter and
in the top of the tank. Air, which during normal 18 ft long, and with 2000 ft^ of heat-exchange
operation is swept through the vapor space in area. Thirteen centrifugal pumps are provided
the tanks at a rate of about 600 cfm to remove for water circulation in this circuit, each with
radiolytic hydrogen, is likewise vented t h r o u ^ a rating of 600 gpm at 150 ft of water head.
the off-gas header to fibrous glass filters and a
The concentration of ionic impurities in the
50-ft-high s t a c k . Instruments are provided to meas-