ORNL-4574.txt

 

<y

¥,

Y

n

-

 

 

ORNL-L5TL

Contract No. W-TL405-eng-26

Chemical Technology Division

MOLTEN-SALT FLUORIDE VOLATILITY PILOT PLANT: RECOVERY OF

ENRICHED URANIUM FROM ALUMINUM-CLAD FUEL ELEMENTS

W. H. Carr’

L. J. King

F. G. Kitts

W. T. McDuffee
F. W. Mileg™

 

*Present address: Allied Chemical Nuclear Products, Inc., Florham
Park, N. J. 07932

**Present address: General Electric Company, Midwest Fuel Recovery
Plant, Morris, I11l. 60450

 

 

{ This report was prepared as an account of work

" | sponsored by the United States Government, Neither

the United States nor the United States Atomic Energy

Commission, nor any of their employees, nor any of

: o their contractors, subcontractors, or their employees,

APRI L -I 97] | makes any warranty, express or implied, or assumes any .

- - | ‘legal liability or responsibility for the accuracy, com-

‘ pleteness or usefulness of any Information, apparatus,

product or process disclosed, or represents that its use
“would not infringe privately owned rights.

 

 

 

OAK RIDGE NATIONAIL LABORATORY
Oak Ridge, Tennessee
operated by
UNION CARBIDE CORPORATION
' for the
U.S. ATOMIC ENERGY COMMISSION

OISTRIBUYION OF THIS DOCUMENT 1S UNLIMITES

(ol
 

M e - et

e
 

Y

n

«}

iii

CONTENTS

Ab StraCt » . . . . . . - - . . . * . . - [ - . . * - - - . . .

Introduction « &« ¢ o« o« o« o o o o o o s s s s s o & o o »
Description of the Molten-Salt Volatility Flowsheet . . .
Preparation of the K~Zr-Al Fluoride Dissolvent Salts .

Descriptive Data for the Irradiated U-Al Fuel Elements
Processed in Runs RA-1, -2, =3, and =4 . . . . . . .

4.1 Configuration, Composition, and Method of Handling .
4.2 Irradiation History . .« « ¢« o o ¢ ¢ o o o o « o o o &
4.3 Significant Fission Products . . « « « « ¢« v v « « &

Dissolution of Fuel Elements, and Volatilization of UFg
from the Melt by Fluorination . . . . ¢« ¢« « ¢« « .«

5.1 Dissolution of Spent Assemblies in Molten K-Zr-Al
Fluoride Salts . « ¢« o o o o ¢ o s o s s s o o o+

5.2 Conversion of the Dissolved UF, to UFg, and Transfer
of the Volatilized UFg from the Melt to NaF
Absorber Beds Using Elemental Fluorine . . . . . .

5.3 Distribution of Significant Fission Products During
Dissolution, Fluorination, and Desorption . . . . .

Quality of UFg Product, and Uranium Inventory . . .

6.1 Fission Product Decontamination Factors for the UFg
Product o « « ¢ & 2 &+ o s o s o o 2 s s s s s & v o

6.2 Primary Radioactive Contaminents in the Product from
Run RA-L & v v v v vt 6 o o e e s e n e e e e e

6.3 Removal of Technetium, Neptunium, and Plutonium from
the UFg Product « « « « ¢ v ¢ o o ¢ o o o o & &

6.4 Nonradioactive Cationic Conteminants . . . . . . . .

6.5 Cumulaetive Material Balances for Salt and Uranium . .

Release of Fission Products to the Environment . . . +
ACknOWJ.e d@ent’s * - 9 ® » » - » - . - - . ° . » » * 2 . *

Re ference S . - - - . - - . - * . . » - - - . - * . » . . L ]

11
11

12
13

1T

17

22
26
34

37
40
L
L3
L7
50

o2
 

 

_....,...______‘.._.........,_‘____..,_...._________

 

10.

Appendixes . . .

10.1

Appendix A:

Fission Product Content of Irradiated

Fuel Elements .
Dissolution of Fuel Elements .

Appendix B:
Appendix (:
ment . .
Appendix D:
Appendix E:
Appendix F:
Books .

Decontamination of Pilot Plant Equip-

*

iv

.

»

*

Corrosion of Vessels .
Radiation Safety .

Index of Volatility Pilot Plant Lo

»

*

.

»

»

»

s
 

oy

2}

«h

MOLTEN-SALT FLUORIDE VOLATILITY PILOT PLANT: RECOVERY OF
ENRICHED URANIUM FROM ALUMINUM-CLAD FUEL ELEMENTS

W. H. Carr, L. J. King, F. G, Kitts,
W. T. McDuffee, and F. W. Miles

ABSTRACT

We have developed and successfully demonstrated a molten-
salt fluoride-volatility process for recovering decontaminated
uranium from spent uranium-gluminum allcy ORR and LITR fuel
elements clad in aluminum. The facilities and the process
were essentially the same as those used for zirconium- and
Zircaloy-clad fuels except that an aluminum-potassium-zirconium
fluoride mixture was used as the carrier salt. The development
progran included the processing of both unirradiated and irrad-
iated fuel elements. Fission product decontamination factors
(fuel to UFg product) for the UFg products in the four hot rums
were generally 10% to 1010, The uranium concentration in the
salt after fluorinstion ranged from less than 0.1 to 9 ppm;
total nonrecoverable losses in scrubbers and waste salt averaged
less than 0.9%. Dissolution of the fuel elements required 8
to 17 hr for 90% completion, and 12 to 25 hr for 100% comple-
tion; average dissolution rates were 0.6 and 0.4 kg of aluminum/
hr, respectively. The release of fission products to the atmos-
phere during the first three hot runs was confined to 120 mCi
of technetium, 5 mCi of ruthenium (which occurred in one run),
lélzsldetectable amounts of !31I, and 47 to 60 Ci (calculated) of

Kr.

In the fourth run, an ORR element that had been cooled
less than four weeks was processed. Radiation exposures to
personnel were controlled within tolerable limits. The de-
contamination factors (DF's) in this run ranged from 2 x 105
to 108. Two major exceptions were the DF's for %Mo and
125gh, which were 36 and sbout 500 respectively. The gggduct

U.

‘had a high radicactivity level due to the presence of

The uranium concentration in the salt after fluorination in
this run was approximately 0.1l ppm, and the total nonrecov=~
erable loss was 0.4%. In the short-cooled run (RA-L), 24 Ci
of 85Kr and 2260 Ci of 133Xe (calculated) were released to
the atmosphere during hydrofluorination; 20 Ci of technetiums
along with barely detectable asmounts of 1311 (1072 ci)

and ruthenium (10~3 Ci), were released during fluorination.
The 8%Kr and 133Xe were released over an extended period so
that actual ground-level concentrations did not exceed a

small fraction of the maximum permissible concentrations

(MPC's) at any time.
 

 

1. INTRODUCTION

At ORNL, uranium-zirconium alloy fuels containing highly enriched

uranium and irradiated to burnups of 32% have been successfully pro-

cessed,using the molten-salt fluoride-volatility process, after cooling

periods as short as six m.onths.l However, the anticipated quantity of
spent uranium-zirconium slloy fuel is insufficient to justify a molten
salt-fluoride volatility processing plant. To be economically practi-
cal, such & plant would have to be capable of processing other types of

nuclear reactor fuel as well.

In order to provide a larger processing load (and thereby improve
plant economics) for a molten-salt fluoride-volatility plant, a devel-
opment program was undertaken to extend the use of the voletility pro-
cess mentioned above to the decontamination and recovery of unburned
uranium in uranium-aluminum alloy (U-Al1) fuel elements. This program
culminated in the processing of highly enriched fuels, with approxi-
mately 30% of the initial 235U fissioned, within 25 days of discharge

- from & reactor operating at a flux of greater than 2 x 10'* neutrons

=2

em 1

sec

Five cold runs and four hot runs were made in the Volatility Pilot
Plant* (VPP) at ORNL. In the first two cold runs, multiplate aluminum
dummy fuel elements were dissolved; dummy fuel and unirradiated UF,
wére processed in the remaining three cold runs. In the hot runs, fuel
from the Low Intensity Test Reactor (LITR) and the Oak Ridge Research
Reactor (ORR) was processed after cooling periods ranging from less
than 30 days to 18 months. The hot runs were followed by an additional
dummy dissolution, four barren salt flushes, agueous decontamination,

corrosion measurements, and, finally, equipment sectioning.

The purpose of this report is to present the information obtained
in the VPP runs; primary emphasis is on the four hot runs (LITR fuel
cooled 18 months snd ORR fuel cooled 6 months, 3 months, and 25 days
respectively). The molten-salt fluoride-volatility flowsheet and &

summary of the operational experience and results for the processing

 

¥Located in cells 1 and 2 and adjacent areas, Bldg. 3019.
 

 

 

 

 

"

13

%

(4) recovery of the UFg product by solid-phase condensation. A more

of aluminum-base fuels in the VPP are included. The irradiation his-
tories of the fuel elements that were dissolved, the compositions of
the molten salts employed, and the two principal chemical reaction steps
of hydrofluorination and fluorination are discussed in detail. Special
attention is given to the distribution and release of fission products.
Daeta regarding the purity of the UFg products, the nonrecoverable uran-
ium losses, the uranium and salt balances, and the radiation experience
(radiation intensity measurements as well as personnel exposures) are
also presented. Equipment design and performance, and operating proce-
dures that differ from those used in the earlier zirconium program, are
described elsewhere.2 Complete discussions of eguipment and procedures
in the zirconium progrem may &lso be found elsewhere.l’B’h
In this report, the molten-salt fluoride-volatility process as
applied to U-Al fuels is referred to as "the volatility process,”" even
though it is only one of many volatility processes. This particular
volatility process essentially consists of four steps: (1) dissolution
of fuel elements in a molten fluoride salt, by reaction with anhydrous
HF, to produce UFy and AlFj; (2) removal and partial decontamination of
the uranium by fluorination with fluorine, which converts the UFy to
UFg; (3) further purificatioh of the UFg by passage throuéh beds of NaF

and MgF,; pellets, utilizing sorption and desorption techniques; and
detailed description of the process follows in the next section.
2. DESCRIPTION OF THE MOLTEN-SALT VOLATILITY FLOWSHEET

A simplified schematic diegram, or flowsheet, of the equipment
used in the VPP is shown in Fig. 2.1. In accordance with this flowsheet,
each irradiated fuel'elément was brought into the pilot plant in the
shielded carrier-chargef (FV-9501). The carrier-charger was centered
over the charging chute, and the multiplate element was lowered (by a
zirconium wire) directly into the 5-in.-diam bottom section of the
dissolver (FV-1000), where it rested on the distributor plate. 1In

most instances, a second element was placed on top of the first. The

 
 

Barren Salt
Transfer Tank
FV-1500

 

 

 

 
 

 

 

 

Fuel
> Careler- Charger
FV-950!
06

 

 

 

 

 

 

 

 

 

 

 

  
 

Condenser

   
  
 

 

 

 

 

 

 
 
 
   

 
 

E?rr‘:ple NoF Fill Line
Rod-Out

 

drofluorinator
Dissolver)

-1000 -

 

    

  
  
 
 

  
    

   

HF
Reboiler
FV-1005

Nasny
L Fiuorinator
Fv

 

A A Waste Salt | '.
Can -
-FV=-112

 

 

   

KOH
Surge Tank
Fv-152

  
     

 

   

 

ORNL Dwg 64-9234 R-|

O O >
ot >

 
    

To Off-Gas
Caustic
Scrubber
FV-150 -

  

 

 
   
   

 

 

 

 

4
—Small

     
   
 
   
 

Product-
Receiver
Fv-223

 
 
    

 
 

  
 
  

 

 

 

 

  

 

 

  

 

>-Decontamination

 

 

 

ey

%
SRR

 
  
  

 

 
  
     

 

 

 

 

Chemical Trop
{NaF)Fv-12]

  

Cold Trap
FV-220_

 
  

Decontamination
Connection Esn‘é?.':’?&
Fv-1207

 
 

  
  
 

 

 

 

 

 

 

 

 

 

 

 

 

NOTES: Shaded vessels ore those

which were deconfaminaied.
NORMAL FLOW LEGEND

 

y

HF --+—a-
F2 -—
Mollen Solt At

Service —_—

Fig. 2.1, - Schematic Flowshéet of Equipment in Volatility Pilot Plant.

 
 

 

 

)

b

K-Zr-Al1 fluoride salt was mixed as a powder, added to the barren salt
transfer tank (FV-1500), melted (mp = ~600°C), sampled, and transferred
to the dissolver (also called the hydrofluorinator), which had been pre-
heated to 600-650°C.

Anydrous HF was distilled batchwise into the system through the
HF cooler (FV-200hk) to the HF accumulstor (FV-1006). A stream of li-
quid HF was pumped from the accumulator to the HF vapor generator
(FV-120T7), where it was vaporized; the vapor was superheated to 100°C,
metered, and fed to the dissolver beneath the distributor plate. The
HF dissolved in the salt and reacted with the elements to produce
AlF; and UF, (which became part of the melt), and hydrogen. The hydro-
gen, unreacted HF vapor, and inert gases from instrument purges left
the dissolver and entered the flash cooler (FV-100l1), where they con-
tacted a second stream of liquid HF pumped from the HF accumulator.
Solids that had been entrained from the dissolver were removed here,
in the condenser (FV-2001), and in the HF catch tank (FV-1003). Solids
that collected in the catch tank remained there until the end of the
dissolution step, when the contents of the tank were transferred to
the caustic neutralizer (FV-1009). In turn, the contents of the neu-
tralizer were pumped to the hot chemical waste. The HF was distilled
from the HF reboiler (FV-1005), and was collected in the HF accumulator
after passing through filter FV-TOOlC and the HF cooler. The hydrogen
and inert gases passed through the -50°C HF condenser (FV-2005), which
removed traces of HF, and were then bubbled through approximately 2 M
KOH in the caustic neutralizef. This off-gas stream then Joined the
cell off-gas stream, received‘another caustic scrub, passed through
sbsolute filters (AEC type), and was then released to the atmosphere
through the 3020 stack.

- After dissolution was complete (i.e., there was no further decrease
in off-gas volume or HF inventory),the HF recirculation was stopped,
and the melt was sparged.with nitrogen to remove the dissolved HF. The
molten salt (mp = ~550°C) was then transferred to the fluorinator
(FV-100), which had been preheated'to_afiproximately 600°C; enough salt

was left behind to fill the horizontal section of the connecting line
 

 

and thus form a plug,or "freeze valve, to separate the hydrofluorina-

tion equipment and the fluorination equipment.

The fluorinator was sparged fiith nitrogen to mix the new charge
with any "heel" that remained from the previous run, and then the salt
was sampled by lowering a copper ladle (on a chain) directly into the
molten salt and "dipping" & small volume from beneath the surface of
the melt. From such a "feed salt" sample, the uranium and fission
product concentrations after hydrofluorination @nd before fluorination)
could be determined. After the sampling procedure was complete, ele-
mental fluorine was passed through the melt to convert the UFy to UFg
and to thereby remove it from the melt. The only important higher
fluorides of fission products that were formed during fluorination and

were not retained by NaF were MoFg, TeFg, and TcFg.

Fluorine at 12 to 60 psig was supplied by a tank trailer parked
outside Bldg. 3019; it entered through a NaF trap (inlet end heated to
100°C), which removed HF. The purified fluorine flowed into the fluori-
nator through a draft tube, which induced circulation of the melt and
improved gas-liquid contacting. Volatile UFg, volatile fission product
fluorides, and unreacted fluorine passed out of the flfiorinator through
the movable bed absorber (FV-105); the higher fluorides of most of the
fission products are nonvolatile, and they remained in the salt. This
gas stream passed, first, through a section of the movable bed absorber
containing NaF pellets at 400°C. Here, the bulk of the fission product
fluorides that were volatilized or entrained were deposited; the Fjp,
essentially all of the U, Mo, Np, and Te, and significant quantities of
Zr, Nb, Ru, I, and Te proceeded to the next section containing NaF
pellets at 150°C. The UFg and most of the contaminants were sorbed
under these conditions, while the fluorine, MoFg, and some tellurium
fluorides passed on to the chemical trap (FV-121), which contained NaF
at ambient temperature. The MoFg and any traces of UFg were removed
by this trap. Fluorine was removed in a caustic scrubber (FV-150).

The off-gas was then vented to the cell off-gas system (which included
another caustic scrubber) and was filtered before being exheausted.

Generally, a small amount of tellurium was released in the off-gas.
x}

D)

)

Desorption of UFg (but not fission products) from the 150°C NaF
pellets in the movable bed absorber was achieved by heating to 400°C
in a fluorine sweep. This gas stream passed, first, through the
impurity trap (FV-120), containing MgF, at 100°C, for the removal of
any technetium and neptunium present and, then, through the product
filter (FV-723) into the small product cylinder (FV-223) maintained
at -T0°C by dry ice--trichloroethylene slush. About TO to 100% of the
UFg was removed; the remainder was deposited in the UFg cold traps
(FV-220 and FV-222) held at -50 to -60°C. The off-gas exited through
the chemical trap (FV-121) to remove any traces of uranium and then
passed to the caustic scrubber as previously. After HF had been
flashed from the UFg product under vacuum at 0°C, the small product
cylinder was removed from the system, weighed, sampled, and assayed

to confirm weight, composition, and enrichment of the product.

After fluorination, the melt in the fluorinator was sampled to
determine the degree of removal of UF, from the salt, Analytical re-
sults were received (usually <3 pg of uranium per gram of salt) before
a portion of the NaF pellets from the LO0°C section of the movable bed
absorber was dumped into the fluorinator. In the event that the uran-

ium concentration was higher than desired, the fluorination could be

continued until an acceptable value of residual uranium was obtained.

The NaF pellets that were transferred to the fluorinator were
from the lower section (400°C) of the absorber; since these pellets

were the first to be contacted by the fluorination off-gas,'they had

the highest concentration of sorbed fission products. The salt was

sparged with nitrogen to aid in the pellet dissolutionj then another
waste salt sample was taken to determine the amount of uranium held
by the pellets. After this uranium snalysis (usually <8 ppm) was
received, the waste salt was transferred to a waste salt can (FV-112)
located inside a shielded carrier. Enough salt was left in ‘the trans-
fer line to form a freeze valve, as was done for the molten salt line
between the dissolver and fluorinator. After cooling, the waste salt
carrier was transported to the burial ground, where the waste salt can

was dropped into an underground vault for long-term storage.
 

 

‘3. PREPARATION OF THE K-Zr-Al FLUORIDE DISSOLVENT SALTS

The ternary salt KF-ZrFy-AlF3 waé considered for the dissolvent in
>

the aluminum cempaign in the VPP since Thoma, Sturm, and Guinn” had shown
it to be the most suitable solvent system for the processing of alumi-
num-uranium fuels. The use of this salt would permit us to operate at
relatively low temperatures, thus minimizing corrosion and avoiding

the difficulties that would otherwise result from the low melting point

(660°C) of aluminum.

A portion of the revised triangular plof of liquidus temperature
as a function of composition for the system KF-ZrF,-AlF3; is shown in
Fig. 3.1. This portion includes the only region with melting points
less than 600°C. It is easily seen that any dissolution path (e.g.,
héamy,*daéhed lines) that is chosen to maximize capaéity will start at
the maximum allowable melting point, cross a region of lower melting
point, and terminate at the maximum melting point that is allowable
during fluorination. Obviously, the higher the temperature that can
be tolerated, the greater the capacity of the salt for aluminum. A
maximum melting point of 600°C was chosen for the beginning salt. This
pernmitted operation at a temperature slightly above the melting point,
and still allowed for a reasonable temperature rise (due to reaction
heat) without attaining the melting point of aluminum. The melting
point at the end of dissolution was held to 550°C to limit the corro-

sive effect of elemental fluorine on the nickel fluorinator.

For all four hot runs, barren salt containing 64.3 to 63.0 mole
% KF and 35.T to 37.0 mole % ZrFy (mp, “600°C) was transferred to the
hydrofluorinator. These salts, when mixed with the small "heels"
carried over in the hydrofluorinator, gave the desired initiasl composi-
tions. The binary salts were prepared by dry-mixing commercial grades
of K,ZrF, (containing 27% potassium and 32.1% zirconium, by weight) and
ZzrF, (54 to 54.5% zirconium). The granular salts and mixtures were
handled in air; no special precautions were taken, except that a rea-
sonable effort was made to minimize the time during which the salt was

exposed to moisture.
 

= ORNL Dwg. 64 -7982
50 9

)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

- 80,
8O
90/ LN N NS X
10 20 30 | 40 50
% sz4
COMPOSITION, mole % COMPOSITION, mole %
POINT | KF | ZrF4 | AIF3 | POINT | KF ZrF4 | AIf
€4.7 | 2t.1 | 14.2 $5.5 | 28.5 |} 16.0

J

57.0 | 100} 33.0 K €64.3 | 35.7 0.0
75.3 (247 | 00 M 55.0 | 30.6 | 14.4
66.0 | 34.0 | 00 N 56.9 | 18.6 | 24.5
€39 |32.9 3.2

 

 

 

 

 

 

 

 

 

 

CjE® M MP>

- - Fig. 3.1, Portion of the KF-ZrF,-AlF3 System That is Applicable to

 

the Dissolution of U-Al Fuels.

=)
 

10

The mixture was melted in a closed vessel; a nitrogen purge was
maintained through the vapor space, and a nitrogen sparge was used to
promote mixing when melting began. There was no evidence (thermal) of
Hy0 evolution at any temperature, although the odor of HF was quite
evident. The melt was clear and had a low viscosity; transfer of the

barren salt into the system was accomplished without difficulty.

Although the binary salt just described provided a highly satis-
factory starting material for processing U-Al fuels, we wanted to de-
monstrate dissolution with a salt initially containing aluminum. The
aluminum could easily be supplied by leaving an aluminum-rich heel in
the hydrofluorinator. The remainder of the charge would then consist
of K2ZrFg and KF; the substitution of low-cost KF for ZrF, would greatly
reduce the cost of the salt components. Although the addition of solid
KoZrFg-KF, as outlined, could not be done in VPP equipment because of
design limitations, partial transfer (i.e., terminating a molten salt
transfer at a specified point) had been demonstrated during the zir-

conium campaign.

We encountered difficulty in all early attempts to prepare a ter-
nary Kp-ZrFy-A1F3 barren salt. Although salt materials (i.e., commer-
cial KpZrFg and KF, and specially-dried AlF3) were carefully pre-mixed,
fusion was always incomplete. A sediment, having the consistency of
coarse sand, was evident in the bottom of the melt vessel, while a
layer of undissolved material floated on the surface of the melts.

When the ternary phase diagram (liquidus temperature vs composition)

was examined, revised data showed that these melts had liquidus tempera-
tures sbout 85°C higher than those indicated by the former triangular
plot. However, mixtures whose compositions had been adjusted to the
dats of the new diagram exhibited thé same characteristics. Although
the dissolvent salt could not be fused completely, experiments showed
that the product salt, after aluminum dissolution and HF sparging, was

a single—phasé melt and could be transferred readily in the liquid
state. Hence, for initisal operations in the\VPP, the ternary salt

wes prepared by mixing, melting as much as was possible, allowing the

salt to freeze, and breaking the frozen salt into chunks of
«)

LB

0

11

non-homogeneous solid. These chunks were then dropped into the dissol-
ver. This experimental procedure was used only for cold runs (liquid
binary salt was used in the hot rums); it was reasonably satisfactory
for cold runs, but would be hazardous if used to charge a highly con-

taminated hydrofluorinator.

After the last run in which irradiated fuel was processed (RA-k4),
a cleanout run was made using a dummy aluminum element. In the latter
run, & molten KF-ZrFy-AlFj3 ternary salt was transferred successfully
into the system. The target composition was 63-21-16 mole % KF-ZrFy-AlF3,
which was thought to be the composition of a ternary eutectic with the
lowest-melting salt in the immediate area of interest, Heating to
650°C appeared to clarify the melt; however, a few inches of sediment
was found in the bottom of the melt vessel. When ZrF, was added, to
increase the ZrF, content to 22.33 mole %, the sediment disappeared.
It did not reappear upon cooling to 600°C. Thus the final composition
of the barren salt that was transferred into the system was 61.94-22,33-
15.73 mole % KF-ZrF,-AlFj.

4. DESCRIPTIVE DATA FOR THE IRRADIATED U-Al FUEL ELEMENTS
PROCESSED IN RUNS RA-l, -2, -3, AND -k

4.1 Configuration, Composition, and Method of Handling

The irradiated U-Al fuel elements proéessed in this campaign were
of the multiplate box type, with curved fuel plates, that is used in
the Oak Ridge Research Reactor (ORR) and the Low Intensity Test Reactor
(LITR). When received, the elements had overall lengths of 26 to 27
in. (the end boxes had been removed), and the ovérall cross section of
each fias essentially 3.0 by 3.1 in. There were'fiineteen 2.8-in.-wide
fuel-bearing plates brazed or swaged between two slotted (inert) side
plates. The cladding and structural material was aluminum, and the
"active" core slloy was approximately 18% uranium--82% sluminum. The
17 "inside" fuel plates were 24.625 in. long (active length, 23.625 in.)
and 0.050 in. thick overall (core alloy, 0.020 in.; cladding, 0.015 in.
 

12

each side). The two fuel plates comprising two sides of the box had
the same alloy core as the "inside" plates, but were 2.520 in. longer
(overall), and the aluminum cladding was 1-1/2 times as thick (for en

overall thickness of 0.065 in.).

Prior to processing, the elements were stored in a canal under
approximately 10-1/2 ft of water. To transfer the elements to the pro-
cessing plant, & shielded (9 in. of lead) charger-carrier was lowered
to the bottom of the canal. A 0.060-in.-diem zirconium wire, which
had been threaded between the plates of the elements and through an
opening in the closed end of the carrier, was used to slide the ele-
ments, horizontally, from a loading table into the cavity of the charger-
carrier. The carrier was closed, hoisted from the canal, and, after
appropriate procedures to prevent the spread of contamination, brought
into the processing plant. It was placed on the enclosure that shielded
the top of the charging chute (centered above the dissolver), with its
exis vertical and the closed end up. The carrier drawer and the charg-
ing chute valves were opened, and the elements were lowered directly
into the 5~in.-diam section of the dissolver by means of the zirconium
wire that was used to load the elements into the carrier. The wire was
cut while still under load (the elements were allowed to drop a few
inches); the severed wire withdrew into the charging chute sufficiently

to clear the valves closing the top of the chute.
4.2 Irradistion History

The elements processed in runs RA-2, -3, and -U4 had been used in
ORR cores, while the RA-1 charge was part of an LITR fueling. In the
ORR, elements are usually irradiated for periods of 10 to 1k days;
then they are removed from the reactor and allowed to cool for 12 to.
180 days before being reinserted for the next cycle. Burnups of 2 to
10% per cycle are accumulated over four or five cycles until a total
burnup of sbout 28 to 32% is achieved; (Here, burnup is defined as the
number of atoms of 235U that have fissioned, divided by the number of
atoms of 235y initially present; thus burnup does not include the 235y
depletion suffered in the formation of 23%U by neutron capture.)
 

 

*)

9]

13

The irradiation cycles for the fuel elements that were processed
in the VPP during the aluminum campaign are shown in Table 4.1. For
fuel elements T4C, L7C, 60D, 58D, and 93D, each line represents an
irradiation cycle for a specified time (shown in hours) in the reactor
at the indicated average neutron flux for the element's particular
location. Both the bfirnup achieved during each cycle and the cooling
(or decay) period observed prior to the subsequent cycle are given.
The decay time listed for the last cycle for each element is the time
from reactor shutdown to the beginning of the dissolution step in the
VPP. This value is used throughout this report to designate the cool-

ing time for the element(s) in each run.

In the LITR, elements are placed in the core, where they remain
until the desired burnup is achieved. New elements are initially placed
on the periphery of the core and later, during the appropriate shutdown,
are transferred to the center section. The irradiation times listed in
Table 4.1 for the fuel elements processed in run RA-1 are the total
times the elements were in the reactor. These times are much longer

than those for ORR fuels because of the much lower flux in the LITR.
4.3 Significant Fission Products

The quantities of the principal fission products that were present
in the irradiated fuel elements at the time of their dissolution are
shown in Table 4.2. These fission products are divided into two
slightly overlepping categories: nuclides contributing & significant
part of the total radioactivity, and nuclides that, due to volatility
at various.times during processing, must receive special consideration
regardless of the amount present. The quantities given in the table
were calculated, using the CRUNCH code, and take into account6the alter-

(The

complete tabulation of the quantities of fission products present is

nate periods of irradiation and decay described in Sect. k.2,

given in Table A-l in Appendix A.)

The most important determinant of the radioactivity contributed

by a particular nuclide at a specific ctooling time is its half-life
 

 

1k

Table 4.1. Irradiation History of Fuel Elements Processed in the VPP
During the Aluminum Campaign

 

 

Fuel Flux Irradistion Decayb Date of Date of
Run Element (10i" neufrons Time Burnupa Time Reactor VPP -
No. No. m 2 gec }) (ar) (%) - (days) Discharge Processing
RA-1 1103 0.18 91T © 219 552 3-5-63 9-9-6L
RA-1 N3k9 0.18 1T 25.3 582 2-5-63 9-9-64
RA-2 ThC 1.48 2h2 6.0 69.1
1.08 185 2.7 16.5
2.17 261 8.6 131.0
2.17 22l 6.7 47,0
2.31 291 8.3 17%.0 y-23-64
1203 32.3 10-1h-64
RA-2 470 1.10 296 5.4 180.2
1.42 k15 9.2 15.9
1.7h 276 6.9 23.h4
2,22 18 0.5 118.8
2.27 201 8.4 174.0 4_23-64
1296 30.54 10-14-6L
RA-3 60D 1.48 80 2.0 13.0
1.48 229 5.5 32,k
1.08 325 5.4 17.3
1.72 292 7.3 111.7 "6l
2.09 276 7.6 7.0 8-24-
1202 27, 11-11-64%
RA-3 58D 1.65 59 1.6 43.4
1.30 361 T.7 1h.3
2.16 310 10,0 116.9 \gl
2.35 276 8.6 T7.0 B2k
1006 27.9 11-11-6%
RA-h 93D 1.58 290 7.9 26.4
1.38 297 7.6 12.8
0.88 301 6.2 1344
2.35 2ho 6.4 24,8 11-15-64
1128 28.1 12-9-64

 

 

aBurnu.p =

atoms of 23%y rigsioned

atoms of 4°°U initislly present’

bDecay time is cooling time between irradiation cycles or between final discharge of the fuel
element and the start of processing in the VPP,
 

 

Table 4.2. Principal Fission Products Present in Irradiated Fuel Elements Processed

Fission product activities are corrected to time of dissolution

 

Nueclides Contributing Significantly to the Total Activity of -
Irradiated U-Al Fuels

Calculated Activity in Fuel

Charge et Time of Processing

 

 

Volatile and/or Otherwise Troublesome Nuclides
Calculated Activity in Fuel
Charge at Time of Processing

 

 

 

. Half-Iife (curies) e Half-Life (curies)

Nuclide Years Days RA-1 RA-2 RA-3 RA-4P Nueclide Years Days RA-1 RA-2 RA-3 RA=-4°
85kr 10.27 LT ¢ 85kp 10.27 47.3 59.5 54,1 24.2
89gr 54 2252 6036 6306 %0 2.8 a 148
0gy 28 377 450 403 103py 41 0.5 T8 2766 3986
y 58 3131 7973 7928 106gpy 1.0 167 Lys ko1 2h2
35zre 63 - 31 387k 9189 8333 125gp 2.7 9.3 1h.7 k.1 6.5
ISyt 35 70 7342  1k505 7523 127gy 3.9 30.9
103py k1 T48 2766 3986 127qer 90 1.7 51.4 100 48,2
106gy 1.0 167 hbs 491 1297¢ 33 33.8 1M1 341
1311 8.05 2293 131y 8.05 8.5 2293
133%e 5.27 2261 132q¢ 3.2 232
137¢gc 33 . 323 383 342 133xe 5.27 2261
140pg 12.80 331 82L3

l4loe 32 T61 4059 8559
143pp ' 13.7 Liz2 8378

l44ce 290 2338  T703 8964 4550

147x4 ©11.3 , 3126
147ppf 2.6 1081 1730 1676 734
Cooling Cooling

Time _ Time

(Gays) 56T 174 7 25 (days) 567 1Tk 7 25

 

 

 

aThe first nuclide in each decay chain that contributes s significant fraction of the total metivity at the time of process=-
ing is tabulated. A "true" total activity would also include the activity of the short-lived daughters in secular equilibrium

with the nuclides listed.

b'I‘he fuel charge in this run was only one element. Two elements were used in each of the other three runs.

®Less than 1/2% of the total of the activities tabulated.

dLess than 0.0l curie.

eIn addition to contributing significantly to the total dose, the nuclide is also volatile under certain conditions.
fA second significant nuclide is shown because the parent-daughter relationships do not fulfill the requirements for secular

equilibrium,

¢T

 
 

 

© 16

(tl/2)' The radiation intensity [determined by its decay constant

(A) = %i%%é-sec-l] of a long-lived nuclide is relatively low. On the
other hand, & short-lived nuclide may almost completely disappeer (i.e.,
it decreases by a factor of ~103 after ten half-lives) in a relatively
short time; however, as long as any significant fraction remains, the
radiation intensity may be quite high due to the higher A. For these
reasons, the quantities of long-lived nuclides shown in Table 4.2 de-
crease only slightly as decay times increase. Examples are 9°Sr,

106Ru, and 137Cs, which comprise a large fraction of the activity in
the long-cooled run, (RA-l), but are reletively insignificant in the
short-cooled runs (RA-3 and RA-4). Nuclides with intermediate half-
lives (tl/2 = 30 to 90 days), such as 32-day “1Ce, do not appear until
shorter cooling times. Insignificant in run RA-1, l4lce increases from
minor significence in RA-2 to become the highest listed intemsity in

run RA-4. Four important short-lived (tl/2 = <6 days) nuclides —

%o, 127g5p, 132Te, and !33Xe — are not encountered until run RA-k;

four others — 1311, 140Bg, 143pr ang 147N3 — with slightly longer half-
lives (tl/2 <14 days), are hardly notable until run RA-4, Obviously,
the decay time is the variable that determines which nuclides will have

decayed to insignificance and which remain.

The cyclic irradiation has en effect on the fission product spec-
trum at the time the fuel is discharged, but it is not nearly as impor-
tant as the cooling time prior to processing. Since the cooling time
between irradiation cycles is generally less than 135 days, the concen-
trations of long-lived (tl/2 >290 days) nuclides increase steadily.

The nuclides with intermediate lives (30 days < tl/2 <90 days) alter-
nately grow in and decay to give, when plotted, a "sawtooth" curve
having gradually increasing maxima., The short-lived (tl/2 <14 days)

nuclides decrease to near 50% or below on each cooling cycle.
17

5. DISSOLUTION OF FUEL ELEMENTS, AND VOLATILIZATION OF
UFg FROM THE MELT BY FLUORINATION

5.1 Dissolution of Spent Assemblies in Molten K-Zr-Al
Fluoride Salts

During the dissolution (hydrofluorination) of U-Al fuel elements,
HF is vaporized in a steémejacketed vessel, superheated to 100°C in an
electrically heated coil, and fed to the dissolver (hydrofluorinator)
through a line maintained at an average temperature of approximately
500°C (by autoresistance heating). The HF, which enters the dissolver
beneath a distributor plate, is assumed to be at the temperature of the
salt by\the time it contacts the element. Although the HF feed rates
may appear 1ow7(20 to 130 g/min), they represent fairly high volumetric
rates at 400 to 600°C. For example, each gram of HF, fully dissociated,
represents about 0.1 ft3 at 410°C and 0.127 £t3 at 600°C (average salt

temperature).

The conditions and results (including data for the fuel charge,
salt compositions, dissolver temperatures, dissolution rates and times
required to attain 90% and 100% completion, and HF feed rates, consump-
tion, and utilization) for the ten aluminum dissolutions in the VPP
are shown in Table 5.1. The runs were made in the chronological order
shown. The runs in the RA (Radioaétive, Aluminum) series are listed in
order of decreasing cooling time for the fuei elements processed: RA-1,
approximately 540 days; RA-2, approximatély 180 days; RA-3, approximately
80 days; and RA-4, 25 days.' A cleanout run, DA-3, is discussed only be-
cause it involved a'dissolution-of aluminum comparable'to others in the

series.

The elements dissolved in the DA (Dummy, Aluminum) and UA (Uranium-
Aluminum)’runs'were 17-plate alumihum dummies cut to the appropriate
length to give the desired weight. In the UA runs, the uranium was
added as finely_powdered UFy, packaged in alumirum foil that was less than
1 mil thick. A double charge of UFy was added in run UA-l so that a mea-
surable amount of UFg product (>100 g) could be withdrawn from the system

after a uranium inventory near the normal steady-state value was established.
P

 

. Table 5.1.

 

 

18

 

 

 

 

 

 

 

Hydrofluorination® Conditions end Results for VEP Rune -

Weight Hydrofluorinator L Dissolution Times , - ' HP Otilization

of_Fuel Molten Salt CompositioLLMIe s) Terperature(°c)° HF Flow For 90% . For 100% Al Dissolution Rates (kg/hr) HF_Consumed - At 90% At 100%
Run Al U Initial® : L Final Salt Vapor Rate Completion Completion To 90% To 100% Quantity % of Dissolution Dissolution
¥o. (xg) (&) K Zr n K- Ir © A Section Bection  (g/min) « {nhr) {(hr) Dissolution Dissolution (xg) Theoretical (%) (%)
DA-1 7.81 0 6h.0 - 360 O 53.8 3.2 16.0 610 k8o 125-80 23.1 28.% 0.30h 0.275 19.2 110.1 9.6 9.3
DA-2 79 0 ‘6!;;!; 22.0  13.6. - 55.1 18.8 B 26.1 616 . 500 70-120 8.9 12.8 0.799 0.616 17.6 114 29.8 22.8
U1 6.79 02 63.9  20.8 | 15.3 55.1 18.0  26.9 60k k60 100-60 12.3 16.8 o.uoT 0.403 16.9' 106.4 21.8 19.2

- UA-2 6.62 36h | 63.2 33.5 3.3 55.0 2.1 15.9 620 505 130-50 22 26 0.271 0.255 - 15.h 104.8 8.7 8.1

UA-3 k.2 0h 643 35T 0 57.9 32.2 9.9 615 k75 125-h0 16.8 22.6 0.225 '0.186 10.9 115.7 8.3 7.8%
RA-1  8.80 323 64,3 35.7 o 55.0 30.6 | 1h.h 608 L6T 125-75 16.6 '25.0 0.h9h 1 0.36h 20.2 100.d 18.0 14.3
RA-2 8.24 306 ~ 62.h 346 3.0 55.4 . 30.8  13.8 . 603 475 100 15.5 2k.5 0.503. 0.353 19.7 o 102.5 18.6 13.1
RA-3 8.7 34  62.8 36.8 0.  S5.1 3.3 12.6 515 K70 ko, 100 8.0 15.5 0.969 0.555 20.4 106.6 50.7 24,2
RA-4 2 159. ' 62 1 36.h 1.5 55.5 . 32.6 1n.9 575 kho 20-100 10.0 12.3 6.378 0.343 9.8 . 105.4 18.7 16.0
DA-3 _3.96 N 1 61.5 23.0 15.5  56.1 21.0 22.9 585 k90 . 60, 100 | 9.7 lo.7 0.37 0.37 9.9 112;5 : 1h.9 1k.9

 

'Here, *hydrofluorination" and "dissolution are used interchangeably.
COrrected for heel from previous rum. ' -

®Average. Represents cyclic behavior (within 10 to 15°C of the stated value)

 
 

)

19

In the RA runs, the charges were LITR (RA-1l) and ORR elements (RA-2
through RA-lI) that had been irradiated to burnups of 22 to 32%.

The barren salts for all the runs except three (DA-2, UA-1, and
DA-3) were binary mixtures consisting of approximately 64 mole % KF
and 36 mole % ZrFy. They were prepared from commercial-grade K;ZrFg
and ZrFy. The "initial" salts (Table 5.1) in four of the runs con-
tained 0.4 to 3.3 mole % AlF3 supplied by heels left in the hydro-

fluorinator from the previous runs. In runs DA-2 and UA-1l, the high

 

initial Al1F3 was obtained by adding a ternary salt to the hydrofluori-
nator as a solidj; in run DA-3, the target composition was a ternary
eutectic with a melting point sufficiently low to permit the ternary
salt to be transferred as a liquid. All the "final" melts contained 5k
to 58 mole % KF and either 29 to 32.5 mole % ZrFy--10 to 16 mole %
AlF3 or 18 to 21 mole % ZrFy--23 to 27 mole % AlFj3, depending on the

initial AlF3 content; each was readily transferable.

Temperature was not investigated as a process variable; the tem-

' peratures used were simply those which we believed would approach the

minima needed to give acceptable operational performance.

The range of the HF dissolvent flow rate was also chosen from the
standpoint of operational experience, Systematic changes in this
varieble were not made, and no attempt at optimization with respect
to any particular parameter was made. It was not necessary to keep the
HF flow rate low, since the HF that was not utilized in a particular
pass was recycled; the only additional costs related to high flow rates
were those connected with pumping and hesting (or cooling). At very
high gas flow rates, selt entrainment could be troublesome. In the
equipment at Bldg. 3019, the practical upper limit of the HF flow rate
(v130 g/min) is determined by the refrigeretion capacity at low HF uti-
liZatidn. At high utilization, pressurization of the off-gas system

cen impose an upper limit on HF feed rate. Heat evolution during dis-

solution was never a problém during the aluminum campaign. A lower

1imit would probebly be imposed by the control system, but this was
not explored; the equipment functioned very satisfactorily at the lowest

HF rate used, 20 g/min.
 

 

 

20

The 9- to 23-hr periods required for 90% dissolution, and the 13-
to 28-hr periods required'for 100% dissolution, are based on the actual
volumes of HF consumed, which were recorded as a function of time dur-
ing each run. These volumes ranged from 100 to 116% of the theoretical

HF consumptions.

The tabulated HF utilization values renge from 8 to 50% and corre-
late rather well with the HF flow rates; that is, the higher the flow
rate, the lower the utilization. This means that the rate of HF sparging
was too rapid to be effective. However, if this added throughput in-
creased the reaction rate even slightly, by such means as increased tur-
bulence, better film coefficients, gas phase reaction, or improved

temperature distribution, it might be justified.

Corfelation of aluminum dissolution rates with run parameters is
difficult since three aluminum-uranium systems, two salt compositions
(as well as intermediate ones), and two schemes of feeding HF were used.
The aluminum dissolution rate and the HF feed rate are plotted vs run
time for each of the ten dissolutions in Appendix A. Figures A-1 through
A-10 show that the average rates and times for dissolution are not en-
tirely representative of what is taking place. Some runs show high rates
initially, while others start low and gradually increase for several |
hours. ©Some decrease rapidly, whereas others tend to become constant
at different levels. All are characterized by a "tailing out" as total
dissolution is approached, although the length of this tailing varies
widely.

As mentioned earlier, aluminum dissolution rates and the completion
of dissolution are inferred from periodic readings of the inventory of
liquid HF in the recirculating system. These readings are subject to
the usual instrument and reading errors; this may eXplain some of the
roughness of the plots, which were smoothed to some extent by averaging.
The initial dissolution rate is probably the least reliable of all
since changes in inventories in pipes, vapor spaces, the salt itself,
etc. are also involved. The point at which dissolution is complete is
very difficult to detect. The 90% level of dissolution is considered
to be the point at which 90% of the HF required to achieve 100%
 

n

0

)

21

dissolution (the point after which no further decrease in HF inventory
occurs) has been consumed. This time is then used to calculate the

90% dissolution rate.

Although the datsa for 100% dissolution are admittedly not precise,
and the dissolution rate curves are somewhat erratic, the dissolution

times and average rates based on 90% dissolution are'fairly accurate

and meaningful and can be used to draw some valid conclusions regarding

the principal variables. By comparing runs DA-1 and DA-2, which were
similar except for the aluminum contents of the initial salt charges,
we must conclude that the salt with the higher aluminumICOntent exhibits
dissolution rates two to three times that of salt containing no alumi~
num initially. The short dissolution time in run DA-3 (i.e., 9.7 hr to
90% completion) confirms this. (Note that the dissolution rate in DA-3
is only one-half that of DA-2; however, the times are consistefit since
the weight of fuel used in run DA-2 was twice that in DA-3 and hence
twice the area was available for dissolution.) The salt dissolved in
run UA-1 was similar to that dissolved in DA-2, except that it contained
a relatively large amount of UFy. We believe that the longer dissolu-
tion time that was required to achieve 90% completion in run UA-1 (see

Table 5.1) is a statistical variation since the RA- runs give no indi-

cation that the presence of urenium inhibits dissolution. Run UA-2 was

similar to DA-1 excépt that small amounts of aluminum and uranium were
included in the initial salt; the same slow dissolution was observed.
From a comparison of runs UA~3 and RA-1, we conclude that irradiation
has no discernible effect on dissolution rate; dissolution times.for
equal-sized salt charges were equal, although the weight of metal in
fihe charges differed by a factor of 2. Further justification for
this conelusion was provided by-the:fesults pbtained'in run ‘RA-2, which
was a duplicate of run RA-1l except that the salt for RA-2 initially
éontained 3 mole %‘A1F3. - | B |
Dissolution rates were considerably higher (from the standpoint
of total time required or in terms of rate per unit area of aluminum)

in runs RA-3 and -4 than in runs RA-1 and -2. Three variables, which
 

 

 

22

could cause diffefent dissolution rates, were: +the ZrFy content, the
temperature, and the HF flow rate. The effects of these parameters

on dissolution rate have not been studied. Based on results in other
runs and on.previous operating experience, we‘would not expect the

first two to cause significant changes in dissolution rate. In all the
runs prior to RA-3, the HF flow rate was increased up to the maximum
(125 to 130 g/min) as soon as possible. As cooling times became shorter,
however, we decided to react the bulk of the elements at lower flow
rates and not risk the production of "bursts" of off-gas, which, in

turn, would cause pressurization of the off-gas system. It is possible

that local cooling effects of the excess gas or hydraulic effects

(especially while the multichannel configuration is still intact) could
affect dissolution rate,

5.2 Conversion of the Dissolved UF, to UFg, and Transfer of the
Volatilized UFg from the Melt to
NaF Absorber Beds Using Elemental Fluorine

During the fluorination step, elemental fluorine is admitted at
the bottom of the fluorinator through a draft tube; The purpose of
this tube is to increase circulation within the melt. In the fluorinsa-
tor, the fluorine contacts the molten salt and reacts with the dissolved
UFy. Both the UFg that is formed and the excess fluorine pass up through
the vapor space into the moveble-bed (NeF pellets) absorber. The lower,

or nearly horizontal,* section of this ebsorber, which is maintained

 

¥The movable-bed sbsorber was originally constructed with a horizon-
tal section and a vertical section. During initial testing, it became
apparent that tilting the unit down (with respect to the entry point)
would result in a longer effective length of horizontal section due to
the angle of repose of the pellets (v45°). Absorption in the horizontal
section, rather than in the vertical section, is desirsble since the
horizontal piston can easily move any sintered pellets in this section.
Hence, the unit was tilted 15° in the VPP, resulting in the displace-

"ment of the horizontal and the vertical sections 15° from the horizontal

and the verticel axes, respectively.
 

o

»

"

23

at L00°C, removes most of the fission product fluorides that are vola-
tilized from the melt; the'UFs passes to the upper, or vertical, section
(meintained at 150°C), where it is sorbed guantitatively. At 150°C the
MoFg passes through to a 20 to 30°C NaF trap, which is used to remove any
UFg that could have passed through the 150°C trap. The off-gas is
scrubbed with an aqueous caustic solution (thus removing the fluorine)
and then combined with the cell off-gas; the resulting off-gas stream

is scrubbed with caustic and, finally, is filtered and vented to the

stack.

The conditions and results for seven fluorination-sorption and de-
sorption runs and two simulations are presented in Table 5.2. In each
run, the melt-gas interface was within the lower 15-1/4-in.-ID cylin-

drical section of the fluorinator. The temperature of the salt was

measured by a thermocouple in a well situated beneath the surface of

the melt. This temperature was held only as high as was considered

necessary to maintain the salt as a liquid; the purpose of such tempera-

" ture control, of course, was to minimize corrosion. The density of the

salt was measured by a differential-pressure cell that was placed across

tvo bubbler probes stationed 5 in. apart.

The uranium concentration in the fluorination salt was obtained

by analyzing a sample of the salt in the fluorinator. BSalt samples

~were taken from beneath the surface of the melt by using a copper ladle

suspended on & Monel chain. Each sample that was withdrewn was actually
a comp051te of the dissolution product from the current run (1nclud1ng

the heel left in the hydrofluorinator from the previous run) and the

~ heel left in the fluorinator from the previous run. The latter heel

contained very little uranium; thus it acted only as & diluent.

Fluorination rates for the hot runs were 6 and 11 std liters/min.

- The lower rate was used in the first parts of the runs in an effort to

produce more unlform loadlng of the NaF pellet bed during the period

- when UFg evolutlon was mos+t rapid. The higher rate, which served to

improve the circulation of the salt through the draft tube, was used

later in the runs in an attempt to minimize the amount of uranium lost
 

Table 5.2.

Conditions and Results for Fluorination-Sorption and Desorption Runs in the VPP

First two runs were practice runs with no uranium; solid UF, was added to the salt in the next three runs;
in the last four runs, irradiated fuel elements were processed.

 

 

 

 

Fluorination Salt Fa F1 Fa U in
Level Above Temp- U 2 Fiow Utildi- Waste
Run . Volume Draft Tube erature Density Concentration Ratg Time zation Absorber'TeEpgraturec (°C) Salt’
No. Cycle (11ters) (in,) (°c) (g/ce) {g/kg salt) (swM”) (min) (%) Zone 1  Zome 2 Zone 3 Zone & {ppm)
DA-1 Fl-Sorp T7.0 10.9 605 2.40 6, 13 17, 15 395 170 150 110
Desorp 1 100 392 300 300 290
DA-2 Fl-Sorp  T3.2 9.7 615 2,20 1, 18 8, 30 180  Amb.%  Amv.d am.d
UA-1  Fl-Sorp 63.3 6.4 560 2,24 2,314 6, 15, 11 101, 20, 19 2,8 392 125 160 1k
Desorp 1 230 392 375 380 375 2.6
UA-2  Fl-Sorp Th.0 10.0 586 2.46 1.808 T, 12 81, 2k 3.7 398 155 137 143 <1
Desorp 1 218 koo 375 418 Loo 10
UA-3  Fl-Sorp 63.3 6.4 €00 2.46 0.976 6, 11 76, 24 2,0 Loz 130 171 150 1.b
Desorp 2T Lob 375 375 375 1.6
RA-1 Fl-Sorp T70.6 8.8 €10 2.45 1.048 6, 11 T1, 24 2.5 ko1 149 146 146 1.6
Desorp ‘ 1 191 4oo hoo k30 380 1.6
RA-2 Fl-Sorp  69.7 8.5 550 2,46 0.616 * 6, 11 73, 22 1.5 396 175 154 138 1.5
No. 1
Fl-Sorp sh.3 3.3 550 2,50 1.433 6, 11 68, 22 2.9 393 162 15k 136 5.2
No, 2
Desorp 1 255 395 398 428 400 16.7
RA-3 Fl-Sorp 64.8 6.9 550 2,52 0.994 6, 11 65, 20 2.5 Lo5 146 160 150 8.65
No. 1
Fl-Sorp Lo.0 ~1.6 550 2,48 0.769 6, 11 60, 18 1.3 406 148 170 1Ly 6.95
No. 2
Desorp 1 226 403 385 433 398 13.80
RA-%  Fl-Sorp 54,3 3.3 550 2,51 1.351 6, 11 55, 20- 3.2 398 154 147 153 <0.1
Desorp * 215 Loo 395 430 398 5.35

 

%1 = fluorination; Sorp'- sorptiony

bSLM = gtandard liters per minute.

Desorp = desorption.

Where more than one rate is given, the rates correspond to the times listed in the adjacent column,.

®Zone 1; 400°C zone; zone 2: temperature transition zone; zone 3: UFg sorber; zone l4: cleanup section (in flow-path only during desorption).

d‘Amb = gmbient.

fie

 
 

a0

e

L}]

25

in the waste salt. Fluorine utilization was calculated by assuming
that one mole of fluorine reacted per mole of UF, present in the salt.
The absorber temperatures shown in Table 5.2 are averages over the

fluorination period.

The most important values in Table 5.2 are the uranium concentra-
tions in the waste salt (shown in the last column). The first value
listed for runs UA-2 through RA-1 and for RA-4 is the uranium content
of‘the melt after fluorination; the second value is the concentration
after the UFg had been desorbed from the NaF-filled movable-bed absorber
and a batch of NaF pellets had been discharged from the horizontal sec-
tion, which is maintained at 400°C, into the fluorinator. In runs RA-2
and -3, two fluorinations were made without an intervening desorption
or pellet discharge. All salt samples (eXcept in run UA-1l) were taken

from the fluorinator in the manner described earlier.

. : . . 1 . o
As in the zirconium campaign,” the uranium concentration in the

melt after fluorination seemed to vary in a random fashion when it was

~only a few parts per million. Any correlation of this concentration

with fluorination time, temperature, initial concentration, irradia-
tion, batch size, or melt depth was not obvious in the range of the

variables investigated.

‘The amount of urahium returned to the fluorinator in the NaF pellets
from the 400°C sbsorber bed should be a function of sorption and desorp-
tion conditions, but not of fluorination conditions. The higher losses
in runs RA-2 and -3 were probably the result of loadlng the bed via two
fluorlnatlon-sorptlon cycles prlor to desorptlon P0351b1y, longer
times at the meximum desorption temperature (400°C) would have reduced
the amount of uranium that was returned. In instances where the amount
of uranium returned might be greater than that which could be economlcally_
dlscharged, the melt could be fluorinated repeatedly_untll the uranium

concentration reached the desired terminal level.
 

26

5.3 Distribution of Significant Fission Products During

Dissolution, Fluorination, and Desorption

During the dissolution step, unreacted HF was condensed into a
catch tank and then revaporized and returned to the recycle system.
Any particulate matter was collected in this catch tank. Hydrogen (a
reaction product) and noncondensables passed through to a tank contain-
ing caustic (2 M KOH), called the caustic neutralizer (KN), where the
gaseous mixture was bubbled through approximately 3 ft of liquid. At
the end of each run, .the HF heel in the catch tank was discharged to
the KN; thus, all of the fission product materiasl that wés collected
from the hydrofluorination off-gas was represented by samples from the
KN. In a somevhat similar manner, the unreacted fluorine from the
fluorinator, after passing through an NaF absorber and a urénium
clean-up trap, entered a scrubbing tower where it was reacted with cir-
culating 2 M KOH. A sample of the scrub solution was representative
of the radiocactive material that was evolved during fluorination or

desorption and removed in the scrubber.

The amount of radiocactive material that was released to the stack
was calculated from the analyses of a charcoal trap through which a sam-
ple of gas from the bottom of the stack had been drawn. The radio-
active noble gases were assumed to be released upon dissolution (measure-
ments of the radiocactivity level of the off-gas being discharged through
the stack were made, but the relatively small amount of radioactive
material that was released during dissolution precluded quantitative
results using this method ). The molten salt was not sampled until it
had been transferred to the fluorinator after hydrofluorination. Here
it waé sampled both before and after fluorination, end after the NaF
pellets had been discharged from the absorber bed; Agreement between
samples teken before and after fluorination was not goods In general;
the velues obtained for most fission products were higher (as predicted
by laboratory investigation) after fluorination; this was unexpected,
based on representative samples of a homogeneous melt. In one run,

radiochemical analyses were made of samples withdrawn before and after
 

*)

n

27

the NaF pellet discharge, but the sensitivity of the results was not
high enough to permit an estimation of the activity returned to the
melt by the pellet discharge. In all cases, the "salt" activity

values presented later will be the higher of the pre- and postfluorina-

tion samples.

The amounts of significant fission products found in the salt,
HF, fluorine, off-gas scrubber solutions, the UFg product, and the
stack effluent are tabulated later in this section for each hot run.
As mentioned in Sect. 4.3, a particular fission product may be impor-
tant because it is troublesome (volatile, etc.) or because it contri-
butes significantly to the general radiation background. In the earlier
runs, the long-lived fission products were the most important; however,
as cooling times decreased, the nuclides with intermediate and short
half-lives received more attention. In run RA-U, emphasis was on nu-
clides with half-lives on the order of eight days or lessj; these
nuclides had not been present previously in appreciable quantities in
molten salt processing. In comparing quantities of the various nuclides
expected at the time of processing (machine calculation by the CRUNCH
code) with the totals actually found by radiochemical analyses, we con-

‘cluded that agreement of these values within a factor of 2 is satis-

factory and that agreement within an order of magnitude is not to be
considered grossly in error.
In run RA-1 (two LITR elements cooled 19 months, Table 5.3), the

long-lived nuclides 9°Sr, 1°5Ru, 137Cs, and 14%Ce provided the major
percentage of the radioactivity; in addition, some 957r-95ND (tl/2 =

65 days) and 127Te'(tl/2 = 90 days) were still present after a cooling

period of 570 days. Only a low yield of !?5gb (0.023%) was obtained,
but this nuclide has a long half-life (2.7 years) and is expected to
be volatile (as in the zirconium campaigh)l during hydrofluorination.
Résuits confirmed that the antimony and the teliurifim were volatile,

while the other nuclides remained in the salt.

In run RA-2 (two ORR elements cooled 175 days, Teble 5.4), three

additional nuclides - 89Sr, 1°3Ru, and 129Te - are present; however,
Table 5.3. Distribution of Significant Fission Products in Run RA-1

 

 

Distribution of Radioactivity®»P in Run

 

 

Calculated
Caustic Caustic _ Radioactivity
Used in Used in 0ff~Gas UFg Released ~of Fuel
Fission Salt HF System F, System Scrubber Product to Stack Total Element Anal./Calc.
Product (curies) {curies) (mCi) (mCi) (pci) (mCi) {curies) (curies) (%)
90gy 46 <1 46 377 12
957y 14 | «1° 1k 31 L5
I5Nb 37.7(98.5)  0.59(1.5) <1° <1® 38.3 70 | 55
106Ry 4,5(97T) 0.03(0.7) 108(2.3) 88 4.64 167 2.8
125gy <2,0%2¢ 0.95(99.9) 1(0.1) < 0.95 9.3 10.2
1277 0.02(65)  0.01(29) 1 1 <1® 0.03 1.7 1.9
137¢g 237(99.9) 0.23(0.1) ¢ 237 323 73
litpe 1162 ’ 1162 2338 ‘ 50

 

aQua.ntities were determined by analysis.

bNum'be::"s in parentheses are percentages of the total, as determined by analysis.
®Below analytical limits.

dQuantity ignored in total,

®Quantity of !3!I also less than 1 mCi.

g8c

 
( h .‘ | .I | | .U (

Teble S.4. Distribution of Significant Fission Products in Run RA-2

 

 

 

| Distribution of Radioa.ctivitya"b in Run Calculated
Caustic Caustic : Radioactivity
Used in Used in UFg Released of Fuel
Fission Salt HF System F; System Serubber Product to Stack Total Element Anal./Cale.
Product (curies) (curies) {mCi) (mCi) (uci) (mCi) (curies) (curies) (%)
89gr 1850 1850 2252 82
0gr 403 | 1 430 450 90
9zr 2716 | | «° 2716 3874 70
95%L  6036(99.2) 51.4(0.84) «a° 5 6087 7342 83
103y 19.8 - 19.8% TL8 2.6
106gy  31.8(98.7)  0.15(0.47) 259(0.81) Th 32,2% 145 7.2
125gy, 209¢ | 1.31(99.2) 10{0.8) <1© 1.32 14.7 8.8
127,123%mpg L.24(93.1) 0.15(3.3) 122(2.7) 10(0.22) 30(0.66) 4.55 85 5.k
137¢s  527(99.9) 0.44(0.08) - <1 527 383 138
14hce 3626 | <1° 3626 7703 L7

 

aQu_antities were determined by analysis.
bNumbers in parentheses are percentages of the total, as determined by analysis.
®Below analytical limits.

41n addition to the Quantities listed, the following quantities of fission products were found on the nickel wool trap (FV-15h4)
between the F, system caustic scrubber and the off-gas scrubber: Uk uCi of 129Te, 164 uCi of 198Ru, and 1T uCi of 1%3Ru.

®Value was disregarded since it was later found that niobium coextracts with antimony in the sanalysis.

62
 

 

 

30

they do not represent any new chemical species. In instances where we
are interested in two isotopes of the seme element (e.g., strontium
end ruthenium), we did not feel that it was necessary to obtain a com-
Plete set of analyses for both nuclides since it was assumed that their
chemical behavior is similar. The mater;al balances for the nonvola-
tile elements in this run are considerably better than in run RA-l.
Again, antimony and tellurium were volatilized during hydrofluorina-
tion, and tellurium and ruthenium were partially removed from the salt
during fluorination. A small quantity of tellurium was carried through

the off-gas scrubber and released to the stack.

In run RA-3 (two ORR elements cooled 80 days, Table 5.5), three
new species (21Y, 1311, and 140Ba) and one additional isotope (1*1lce)
were present; of these, only the 1317 proved to be of any concern.
The iodine balance in this run was excellent; greater than 90% remained
in the salt, while about 9% was found in the caustic used in the HF

system. Traces were also found in the other two scrubbers and on the

stack sampler. As previously, the tellurium was volatilized in both
major processing steps, and about 5% of the total found was released
to the stack; a small amount of ruthenium was also released. Most of
the antimony and some ruthenium and niobium were found in the caustic
used in the HF system. The material balances for all nuclides except
103Ry, 108Ry, and 127°1297¢ ranged from 67 to 152%, which is considered

to be excellent.

In run RA-4 (one ORR element, cooled 25 days), the presence of
greater than 2-1/k4 kilocuries of !33Xe made the total activity from
the noble gases 100 times that of the 85Kr; however, high dilution
fectors and high MPC values made rapid release safe. In RA-lL, the
total release required more than 10 hrj if it had been accomplished
over a 3-hr period, the groundrlevel concentration would still have
been less than 1% of the MPC.

The significant fission products for run RA-4 are shown in Table
5.6. Three short-lived nuclides, sbsent in runs RA-1l, -2, and -3, were

present in this run: 9%Mo, l1lAg, and !3%Te. The !32Te represented
Table 5.5, Distribution of Significant Fission Products in Run RA-3

 

 

 

Distribution of Radioactivity®:P in Run Calculated
Caustic Caustiec _ Radiocactivity
Used in Used 1n Off-Gas UFg Released of Fuel
Fission Salt HF System F, System Scrubber Product to Stack Total Element Anal. /Calc.
Product (curies) (curies) (mCi) (mCi) (uci) (mCi) (curies) (curies) (%)
83sy  6,836(100) 0.28 <0.1 <1 % 6,836 6,036 113
905,  Ns® 0.04(n0.01) Ns® ng® <«
9ly  8,636(100) 0.28 <0.2 19 <2 8,636 7,973 108
957¢  13,950(100) 0.28 <0.1 <1 <«2? 13,950 9,189 152
M - 20,860(99.3)  152(0.7) <1 15 <? 21,012 14,505 145
103gy  77(98.5) 1.1(1.4) Ns® 59(0.08) <1 n5C 78.16 2,766 2.8
106y  27(98.1) 0.32(1.2)  136(0.5) 68(0.3) 23 27.53 491 5.6
1255,  1,8477 10.3(99.9) 3(0.03) 13(0.10) 8 10.32 14,1 T3
1275129, 1,0(5T.6) 0.56(32.3)  21(1.21) 65(3.Th) 3 90(5.18) 1.7 271 0.6h
1311 7.8(91.1) 0.76(8.9) < <1 <y¢ <1 8.56 8.5 101
1375 411(100) 0.02 <0.1 a <2 411 3h2 120
1405,  453(100) 0.02 <0.1 <1 a? 453 331 137
1¥loe  2,727(100) 0.15 <0.1 <1 ¢ 4,059 2,727 67
lihce  7,773(100) 0.33 .1 < «d 7,773 8,964 87

 

8Quantities were determined by analysis.

bNumbers in parentheses are percentages of the total, as determined by analysis.
As the total of two isotopes.
dBelow analytical limits.

°NS = not sought.

fValue was disregarded since it was later found that niobium coextracts with antimony in the analysis.

EFifty-five microcuries of tellurium was found on the charcoal trap (FV-155); 52 uCi was found on the L00°C Ni wool trap (FV-15k).

There was & total of <1 uCi of ruthenium on both traps.

T¢

 
 

Table 5.6. Distribution of Significant Fission Products in Run RA-b

 

 

Distribution of Radiomctivity®:P in Run

 

 

Caustic Caustic Rag:i:zi:t;:&
Used in Used in Off-Gas UFg Released of Fuel

Fission Salt HF System Fy System Scrubber Product o Stack Total Element Apal,/Calc.
Product (curies) (curies) (curies) (mCi) (mCi) curies (curies) (curies) (%)

83g, 7,960(100) 0.311 0.004 3 0.47 ¢ 7,960 6306 126

9y 9,965(100) 0.226 <<0,001¢ 14 0.32 e 9,965 7928 126

957, 13,400(100) 0.311 <0.0014 1 0.07 e 13,400 8333 161

95yp 12,400(99.4)  75.5(0.6) 0.005 11 0.23 ¢ 12,476 7523 166

%0 65(90:5) 2.80(3.90) 2.26(3.15) 113(0.16)  1660(2.31) e T1.8 148 k9
103gy 11(69.2) 0.032(0.20)  4.,75(29.9) 108(0.68) <0.1% <0.003% 15.9 3986 0.40
106Ry 2.8(64.3) 0.019(0.44)  1.5(34.5)  35(0.80) <0,02% e 4.35 2k2 1.80
N1 22(100) <0.001% 0.006 <102 0.5% ¢ 22 14 157
125g <0,54 2.08(82.5)  O.MB(1T.4) <2 5.2(0.2) ¢ 2.52 6.5 39
127mq e e e e e e e L8 e
1290, e e e e e 13.5 e 3k e
1311 <0.43% 109(99.2) 0.85(0.8) <30% 0.25 <0.01% 110 2293 4.8
1327 37(64.1) 0.586(1.02) 12.8(22.3) 770(1.34) 0.15 6.32(110) 57.5° 232 25
137¢g 232(99.9) 0.126(0.1)  <0.0062 <1 0.32 c 232 152 153
140Bs 8,B856(100) 0.351 <0.001¢ <1 <0.0k e 8,243 8856 107
14lge 7,050 e e e e ¢ 7,050 8559 82
Ihhce 5,617(100) 0.076 <<0,001% <1 0.13 ¢ 5,617 4550 123

 

SQuantities were determined by analysis.

bNumbers in parentheses are percentages of total, as determined by analysis.

®Not detected by gamma scan of charcoal in stack sampler,
dpelow snalytical limits,
eSpecific analyses are not avallable; quantities are- assumed to be proportional to the nuclide for which data are presented.

IThis does not inelude about 1 curie (total) of 1275129,132mg that is held up on the 400°C wool trap in the exit gas line for the
caustic serubber used in the fluorine system,

et

 
 

33

e 60% increase of the 12721297 total, and the 127Sb activity was a
factor of 5 times the customary activity of 125Sb. The presence of
3%Mo presented a serious problem since special procedures had been
necessary to prevent stable molybdenum from following the UFg product.
In general, the results obtained in RA-U4 confirmed the trends observed
in the earlier runs; however, the data of RA-L should be considered to
be more significant because of the appreciably higher activity levels.
(Note that in Table 5.6 some of the units are larger than in preceding
tables.) As previously, tellurium and ruthenium were volatilized
principally during fluorination; antimony and iodine, along with frac-
tional percentages of niobium and cesium, were evolved during the
hydrofluorination step. Some antimony also appeared in the UFg product
and in the caustic used in-the fluorination system. Although greater
than 90% of the 99Mo was found in the salt, the remainder was well dis-
tributed. Analyses showed that greater than 3% of the 9o was present
in each of the two system scrubbers, 0.16% was present in the off-gas
scrubber, and 2.3% was found in the UFg product. None was detected

in the stack sampler. About 20 curies of tellurium and traces of 1311
and ruthenium were released through the stack, but the ground-level
concentration wes calculated te be much less than 5% of MPC, assuming
a 40-min release time and a staek-to—ground dilution factor of 10°

(see Sect. T).

With only one exception, no unusual or unexpected events occurred
as.a result of processing at the higher activity level of run RA-L
_(dbout 1-1/2 times that in RA-3, e#en though twice as many fuel elements
were processed ineRA—S); Upon sparging the salt prior to withdrawing
the first samples from the fluorinator, there was a rapid increase in
the background radiation in the vicinity of the chemical trap (FV-121).
The increase continued until sparging was stopped; the rediation inten-
sity then decreased in a typicel manner (i.e., an exponehtial decey
curve). The rate of decrease indicated that 132I (t,,, = 2.4 hr), the
deughter of 132Te, was the source of the activity. The salt had been
ellowed to remain static (i.e., only low-rate nitrogen purges were

allowed to flow) in the hydrofluorinator for several hours while
 

 

 

34

temperatures were adjusted preparatory to transferring the salt to the
fluorinator. Evidently, during this period, a sizable amount of 1327
grew in from the decay of 132Te, and largely remained in the salt until
fluorination. Sparging in the fluorinator caused the 1321 to be carried
through the 150°C NaF absorber and heated piping to the ambient-tempera-
ture trap (FV-121), where it deposited on the relatively cold surface.

The decay product was stable xenon, which could not be traced.

Table 5.7 presents a comparison of the important data (from the |
standpoint of radiation background) from Tables 5.3-5.6, expressed on

a percentage basis.
6. QUALITY OF UFg PRODUCT, AND URANIUM INVENTORY

The uranium hexafluoride product was collected by desublimation
as it was desorbed from the NaF pellets in the movable-bed absorber
(FV-105). As the bed was heated from 150 to 400°C, the UFg-2NaF
complex decomposed. The UFg was carried, via.fluorine, through & porous
metal filter into the product cylinder, which was cooled with a dry
ice--trichloroethylene mixture. After the product had been collected,
HF was removed from the cylinder by flashing (7 min at 0°C under vacuum);
the cylinder was then weighed and sampled. Table 6.1 gives the weights,
the uranium contents, and the isotopic analyses of the products col-
lected in the four hot runs. Results for an additional run (UA-3) are
included to show an "unburned" isotopic composition. Obviously, all
the uranium added to the system in & particular run does not appear in
the product for that run since, in this seven-run campaign, about one-
third of the uranium was held up in cold traps, NaF traps, salt heels,
ete. This holdup led to & blending, or an overlap, of products from
run to run. The analysis of the uranium product, as reported, is the
result of a coulometric determination in which all the uranium was
assumed to be 233U; the values in the table have been corrected for
the appropriste isotopic analysis. All of the values are in reasonable
agreement with'the theoretical uranium content of UFg (67.36%) for the

assays involved.
.

Table 5.7. Comparison of Fission Product Distributiens in Runs RA-1, -2, -3, and -4®

 

(

Total FP Found in Fuel
Calculated Quantity of FP in Fuel

(Total FP Found in HF System Caustic) (Total FP Found in Salt)

Total FP Found in Fuel

Total FP Found in Fuel

 

Total FP Found in F2 System Caustic)

Total FP Found in Fuel

 

 

?iiiiifi RA-L RA-2 FA-3 RA-F RA-1 RA-2 RA-3 RA-}  RA-1 RA-2 RA-3  FRA-J RA-1 RA-2 RA-3 RA-L
89y 82 113 126 b b ~100° 100 100 b b
Vsr 12 90 b ~100°  100°
9y 108 126 b b 100 100 b b
Bz 45 70 152 161 b b A100°  ~100° 100 100 b
By 55 83 145 166 1.5 0.8 0.7 0.6 98.5 99.2  99.3 99.% b b b b
o kg 3.9 90.5 3.1

103p, 2.6 2.8 0. 1.k 0.2 ~100° 98,5  69.2 29.9
106py 2.8 T.2 5.6 1.8 0.7 0.5 1.2 .4 97 98.7 98,1 643 2.3 0.8 0.5 3k.5
Hlgg 157 b 100 0.0k
1255,  10.2 8.8 73 39 99.9 99.2 99.9 g2.h  <68d a a4 <20  0.10 0.8 0.03 17.4

12721297¢ 1.9 5.4 0.6 29 3.3 32.3 65 93.1  5T.6 3 2.7 1.2
131y 101 4.8 8.9 99.2 91.1  <0.4 b 0.8
132me 25’ 1.02 64.4 22.3
137¢cs 713 138 120 153 0.10 0.08 b 0.05 99.9 99.9 100 100 b b
140pq 137 107 b b 100 100 b b
loe 67 82 b 100  ~200° b
l4hoe 50 47 87 123 b b a100¢  ~100° 100 100 b b

 

%Dats were obtained from Tsbles 5.3-5.6 and are expressed as percentages.
PLess then 0,01%.

cT’hese values are approximate because only analyses of the salt were obtained; however, subsequent runs showed these species to be essentially non-
volatile.

dThese velues were meaningless since an error in the analytical procedures was discovered during the RA-4 analysis.

Ge

 
 

 

Teble 6.1. Uresnium Input and Product Data

36

 

 

 

Run No.
RA-1 RA-2 RA-3 RA-k4 UA-3
Feed:
Uranium in fuel after
burnup, g 323 306 324 159 304°%
Product:
Weight gein of UFg
product cylinder, g 300 470 410 100 300
% U in UFg product 65.64 66.76 67.05 63.94 67.33
Isotopic analyses, %
23hy 0.98 1.16 1.17 1.15 0.82
233y 86.87 83.00  82.52  82.54  91.k0
236y 3.22 5.93 6.09 5.89 0.32
238y 8.93 9.91  10.22  10.k42 T.46

 

®as UFy; unirraediated.

C .
37

6.1 Fission Product Decontamination Factors
for the UFg Product

Most of the fission products remained ifi the melt and were dis-
carded with the waste salt. Portions of thé fission products that vere
volatilized—dhring fluorination were sorbed on the NaF pellets in the
movable-bed absorber (FV-105). The small fractions of these radio-
nuclides that were desorbed along with the UFg usually‘constituted the
contaminants in the product cylinder. (However, it is possible that
some of these contamihantsrwere deposited in earlier runs and were
acquired from the connecting piping.) The decontamination factors
(DF's) for various fission products in the four hot runs are shown'in

Table 6.2.

The first of the four columns for each run lists the activity for
each nuclide in the fuel, as calculated by the CRUNCH code. (Techne-
tium-99 is omitted from the table and is discussed in Sect. 6.3.)

This value, which is given on a per-milligram-of-uranium basis, was
obtained by dividing by the number of milligrams of uranium remaining
in fhe fuel as determined by the reactor operations burnup calculations.

(This is also the uranium content of the fuel for accountability pur-

poses. )

The second and third columns show the results 6f the analysis of
the UFg product for radioactive conteminants. Multiple samples of the UFg
product (as‘a liquid) were enalyzed; the second éolumn gives the acti-
vity level (on the date of analysis) that was thought to be most repre-
sentative for éach nuclide. Sampiing‘fiaS'ddne by heating and tipping
the product cylinder; hence the‘tabulated values do not reflect the
degree of.purification that would have been'afforded.by vapor sampling
of the UFg. In the third column, the activity is placed on a per-
milligram-of-uranium basis. | ”

The DF's in the first three runs were quite satisfactory, generally
ranging from 107 to 1010, The 125gp DF, which decreased to 1.4 x 106
in run RA-3, is an obvious exception. The 196Ry DF was low (108 to 107)

in the first two runs, but its improvement in RA-3 (to 1.8 x 108) and
 

 

 

 

 

 

8

Table 6.2. Decontamination Data for UFg Products in Runs RA-1, -2, -3, and -b

Activities of feeds were calculated by CRUNCH code; activities of products were obtained by analysis.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

: ' : Rnn RA-12 ' : | Run RA-2b'. | Run RA-3° ' . : | Run Ra-44

Fission Fuel Product o Fuel __Product 7 Fuel Product __Fuel o Product

Product dpm/mg U dpn/g UFg . dpm/mg U DF ‘dpm/mg U dpm/g UFs  dpm/mg U DF dpg/mg U dpm/g UFg dpm/mg U DF dph/mg U dpm/g UFg dpm/mg U DF .
89g,. , 4.1k x 1010 7 : ‘ | 8.8L x 1019 9.7 x 106 1.63 x 10* 5;h x 108

- ‘; : : <T400 <11.0 >h,0 x 109

%0sr 2.59 x 10° 215 0.42 6.2 x10° 3.27 x 10° kséo . 7.3 h5x108 2.76 x 10°
511 ') | o o , _ | ; 5.46 x 1010 <9500 <1%,1  >3.9 x20° 1.1l x 10!l 6.6 x 106 1.09 x 10% _'1.0 x 107
957y 2.1k x 108 <600 <0.91  >2.h x 10 2.81 x 1010 <5400 81 >3.5 x 10° 6.30 x 1010 <9300 <13.8  >4.6 x 109 1.16 x 1011 1.k x 105 2.34 x10® 5.0 x 107
9w, k.80 x 108 <350  <0.53  >0.1'x10® 5.33 x 1010 2.35x 10% - 35 1.5 x 10°  9.9% x 1010 @000 - . <13.3 >7.5x10% 1.05 x 101! L.k x 108 7.8 x 103 1.3 x 107
Mo . N | | o | | 2.07 x 10° 1.2 x 1010 5,70 x 107 36.
03p - o | | - S np® 5.57 x 1010 <1,9 x 108 <3.26 x 103  >1.7 x 107
106gy 1,15 x 102 | 5.90 x 205 900 . 1.3 x105 3,23 x10° 3.k x 105 510 6.3 x 106 3:36 x 1010 1.2k x 105 184 1.8 x 108 3.38 x 209 <3.3 x 105 <520 >6.5 x 106
1115, | : | o | o ' . ’ - 1.89 x 108  <5.0 x 10% <1.70 x 10% >1.1 x 0%
125g, 6.b1 x 107 | 5200 7.9 8.1 x10% 1,07 x 108 <6500 <9.7  >1.1 x 107 9.69 x 107  14.50 x 104 67.5 1.h x 206 9.12 x 107 1.1k x 10% 1.78 x 105 . 510

1271129q, - | a ) | | ‘ 0 1.85 x 102 1.75 x 10% 25.9 7.1 x 107 5.4 x 10° ’
131y | - - | S | | 5.83 x 107  <1.8 x 10*  <26.7 >2.2 x 105 3.2p x 10}0 3.k x 106 8.52 x 103 3.8 x 106
132pe | _ " | 3.2L~x 102 9.9 x 105 5.10 x 103 6.4 x 105
137¢s 2.22 x 109 - <350 . <0«53  >k.2 x 10° 2.78 x 10° 1200 1.8 1.5 x 10° 2.35 x 102 <9200 <13.6  >1.7 x 108 2.1% x20% 7.0 x10% 1.09 x 10* 2.0 x 105
140pq 1 | : - _ o - ~ o 2.27 x 109 <5300 <7.9 >2.9 x 108 1.15 x 1011 <h.1 x 10° <1.éo x 103  >9.6 x 107
14hce o . - S 5.59 x 1010 <850 © <1.3  >k.3x 1010 6,14 x 1010 <6100 <9.0 >6.8 x 10° 6.3% x 1010 2.85 x 105 4.59 x 20° 1.k x.107

 

SFuel processed in this run was cooled 19 months.
Pruel processed in this run was cooled 175 days.
®Fuel processed in this run was cooled 80 days.
dFuel procéssed in this run was cooled 25 days.

€103y (t = 41 days) and 14lpe (t1/2 = 32 days) were not detected by gamma scan. _ : ‘ ‘ i

1/2
|

 

 
 

39

the fact that 1°3Rquas not detected lead to the conclusion that 108Ru
may have been acquired from incompletely decontaminated (from the pre-
vious campaign) process piping. Tellurium and iodine decontamination,
although not high, was satisfactory; the principal concern with these
radionuclides is release to the atmosphere, which is discussed in Sect.
T. The DF for cerium, representing the rare earths, was consistently

high.

In run RA-L4, the decontamination factors were somewhat lower than
in the first three runs, generally ranging from 2 x 10° to 108. The
low antimony DF mey have been caused by the transfer of a large amount
of antimony into the fluorinator as a result of the shorter hydrofluo-
rination period employed in RA-4. The low 29Mo DF was not surprising
since the chemical properties of MoFg are quite similar to those of UFg.
The "molybdenum stripping" feature of the process was intended to keep
the bulk of the stable molybdenum from contaminating the UFg product,
but was not expected to effect the complete removal required for radio-
chemical decontamination. The 29Mo (tl/2 = 2.8 days) would not be pre-
sent in fuel processed after a cooling period of 60 days; it would decay
out of the product (to <0.1%) 30 days after processing. Such a cooling
period might be desirable to allow for decay of the unavoidable 237y,
The lllAg is believed to be of no concern since none of the scrubber
solutions showed any appreciable concentrations of it; a low yield and
a high limit of detection combined to give an indeterminate DF for this

nuclide.

Generally speaking, the DF's obtained in the first three runs were
gbout one order of magnitude lower than thosé achieved in the earlier
zirconium progra.m.l We believe that this is due, at least in part, to
fluorinating at 550°C rather than at 500°C., The lower DF's in run RA-4
may be due to one or more of thé following factors. First, if the
sintered-metal filter located just upstream of the product cylinder had
failed, NaF fines could have carried fission producfis'into the UFg
cylinder. Second, in run RA-4, the additional HF sparge carried out

after dissolution was complete was shortened; therefofe, the molten

 

 
 

 

 

Lo

salt may have carried a higher percentage of the fission products into
the fluorinator. Finally (and this factor seems most significant), the
quantity of NaF pellets discharged from the movable-bed absorber may
have been insufficient. In the aluminum program, 3.8 xg of NaF (about
11% of the bed) was discharged during each run, as compaied with about
6.5 kg in the zirconium program, Insufficient bed discharge would
result in an upward migration and eventual breakthrough of some fission

products.

6.2 Primary Radioactive Contaminants in the
Product from Run RA-L

The gamma scan for the product samples in run RA-L (the short-
cooled run) indicated that the radioactivity of the product was almost
entirely due to the presence of 2°Mo and 237U. The 237U content of the
product was equivalent to a radiocactivity level of 2.53 x 10! gisin-
tegrations per minute per gram of UFg at a cooling time of approximately

39 days.

The gross beta activity of the product in run RA-l4 was 2.7 x 101!
disintegrations per minute per gram of UFg, assuming a 10% geometry
during analysis. This value is 102 to 103 times the gross beta activity
of the product in run RA-3 (3.66 x 10® disintegrations per minute per

gram of UFg).

Radiation readings for the product cylinders in runs RA-3 and RA-k,
taken with a cutie pie at various times after the run was completed,

were as follows:

Approximate Cutie Pie

Cooling Time Reading
Run No. (days) (r/hr)
RA-3 85 0.5
RA-4 28 (completion of run) 50
RA-L 36 6.2

RA-L L9 2
 

 

41

The decrease in the readings of the RA-Y4 product reflects the decay
of both ?%Mo ana 237y.

6.3 Removal of Technetium, Neptunium, and Plutonium
from the UFg Product

The removal of technetium, neptunium, and plutonium from the UFg
product is discussed separately in this section because of the incom-

pleteness of the data. This, in turn, is due to the difficulty of

-obtaining valid analyses of highly radiocactive material. Most of the

data for these three elements were discarded as meaningless; for example,
one set of analyses indicated that the 237Np content of the fuel was

20 times as great as the total uranium content.

The recovery of plutonium in the processing of highly enriched
uranium fuels is not economically feasible. Hence the only concern,
with respect to plutonium, in the runs discussed in this report is
assurance that the plutonium content of the UFg product is within rea-
sonable limits. Past experience in molten-salt volatility processing
has shown excellent plutonium decontamination; the data from the alumi-

num program are given later in this section.

In general, technetium and neptunium fluorides tend to follow UFg
through the sorption-desorption cycle in the NaF bed, although the data
below suggest that the cycle might be arranged to achieve some decon-
tamination. Thus & separate purification step, in which TecFg and NpFs

were sorbed on MgF,; pellets, was employed.

Some technetium apparently remained in the off-gas stream during

sorption; for example, analyéis of the NaF in the chemical trap down-

- stream from the sorber showed 3.6 ppm of technetium at the entrance and

0.5 ppm at about the middle of the trep at the conclusion of runs RA-1,
-2, and -3, However, most of the technetipm was found on the MgFs.
Inlet and outlet samples of the sorbent from the MgF, bed showed the
following 997¢ concentrations: RA-1, -2, and -3 (combined), 1540 ppm
and not greater than 0.15 ppm respectively; RA-4, 8.8 ppm and 1.7 ppm;
 

 

 

L2

and T-18,%¥ 970 ppm and 2 ppm. Results of the analyses of samples of
the composite bed for the entire series indicated that a total of 390
mg of 22Tc was present in the 973-g bed. The products in the first
 three runs contained very little 2Tc: 4.6, 2.4, and 1.1 parts of
technetium per million parts of uranium respectively. However, in
run RA-U4, the product analyzed 440 ppm. This discrepancy has not, as

yet, been explained.

The results for neptunium were similar to those for technetium.
Some neptunium reached the chemical trap during runs RA-1, -2, and 43;
the 237Np concentrations in the NeF at the inlet and at the middle of
the trap were 23 ppm and 1 ppm respectively. Analyses of samples of
MgF, (inlet and outlet respectively) showed: RA-1, -2, and -3, 2310
ppm and 40 ppm; RA-4, 490 ppm and 150 ppm; and T-18,%¥ 1140 ppm and 420
ppm. Analyses of samples of the composite bed indicate that 785 mg of
237§p was present in the bed. Analyses of the UFg products for runs
RA-1 through -4 showed no consistent trend: 140, less than 12, 1670,
and 1860 parts of 237Np per million parts of uranium.

Plutonium data are availsble from runs RA-3 and RA-4, Data from
run RA-3 confirm the difficulty of fluorinating plutonium from a molten
salt. For the first fluorination in run RA-3, the feed salt and the
waste salt showed essentially the same plutonium concentration (1.19
x 10° and 1.25 x 10 counts per minute per gram of salt, respectively);
the feed salt concentration was equivalent to 1.2 x 10° counts perrminute
per gram of uranium. For the second fluorination, the plutonium radio-
activity in the feed salt was 1.26 x 106 counts per minute per gram of
salt (equivalent to 1.6 x 102 counts per minute per gram of urenium),
but thé analysis of the waste salt showed only 5.1 x 10° counts per
minutes per gram of salt. Radiochemical analyses of four product (UFg)
samples from run RA-4 for plutonium yielded the following results:

3.1 x 10%, 5.1 x 105, 1.7 x 10%, and 4.5 x 10" counts per minute per
gram of UFg. If we discard the second value, the average is 3.1 x 10"

 

®
T-18 was a cleanup run (using barren salt) made following run RA-L.
 

43

counts per minute per gram of UFg, or 4.6 x 10* counts per minute per
gram of uranium. Based on 2 51% counting geometry, 4.6 x 10" is equiva-

lent to a 23%Pu concentration of 0.63 ppm.
6.4 Nonradioactive Cationic Contaminants

Contamination of the UFg products (Table 6.3) with nonradioactive
cations was higher than desirable. Analyses of the products from indi-
vidual runs (even when multiple samples from the product of a single
run were used) were too inconsistent for valid interpretation. However,
results of the analysis of the composite product, consisting of a com-
bination of these products in a form suitable for shipment, appear
quite reasonable. The concentrations of most of the cationic contami-
nants, including Cr, Cu, Fe, Ni, and Zr, were lesé than 50 parts per
million parts of uranium. The boron and potassium contents were about T5
prm and 500 ppm respectively. Three bther metals were present in the
following (approximate) quantities: molybdenum, 0.1%; aluminum, 0.3%;
and sodium, 1%. The molybdenum content of the product was dependent on
the conditions under which the movable-bed absorber was operated. The
high sodium content led us to suspect that the sintered-metal filter
located Just upstream of the product cylinder had failed, allowing NaF
fines to be released. The high aluminum content has not been completely

explained, but is believed to be nonrepresentative.

6.5 Cumuletive Material Balances for Salt and Uranium

A cumulative salt balence was maintained throughout each run in
the aluminum-urenium campaign (Table 6.4). This allowed us to obtain
a uranium balance (uranium concentrations were reported as grams of
uranium per grem of salt); and it also provided a check on the instru-
mentation that was used to measure the level and density of molten salt.
It was imperativé that the volume of salt in the fluorinator be known
accurately in order to avoid overfilling the waste salt can. The
volume of waste salt that was transferred was measured by difference

in the fluorinator, since the waste can contained no instrumentation.

 
Table 6.3. Nonradioactive Cationic Impuritiesa"b in the

 

 

 

UFg Products
, RA-2 ‘
Cation UA-1 UA-2 UA-3 RA-1 A B C D E F G H Avg., RA~3 RA-k4 Composite
Al 1k 1,200 1,200 800 900 1,500 - 570 2,000 200 1,900 1,600 1,700 1;300 1,500 1,300 2,850 3,100
B 3.6 13,350 7,800 6,300 L4,900 L4,k00 1,000 3,500 560 k4,600 8,100 5,800 4,100 5,125 3,700 9,100c 5
Cr <0.57 ND ND 30 ND ND ND ND ND 146 | 90 210 55 ND 50 <3 3.5
Cu 0.45 9 11.5 L5 70 300 5 1,540 90 k20 20 30 300 koo 357 <3 22
Fe <0.57 ND ND 5 <50 <90 30 70 - XD 250 160 260 110 75 100 <30 L3
K 0.50 105 130 <50 200 300 T0 230 30 1,306 200 200 300 502 378 600 L5k
Mo 18 565 660 300 900 1,000 270 770 120 600 1,200 ThO 700 700 7,000 gso 1,560
Na 21 3,600 3,450 1,800 5,700 5,000 1,100 2,900 600 2,300 2,700 1,900 2,775  L,475 3,500 8,100 12,000
Ni 0.30 ND ND 30 ND ND ND ND ND 60 30 50 17 700 200 <30 36
Zr- 3.3 51 30 80 <40 <70 ND 10 N ND ND 20 13 23 105 96 <20
Total 62 18,880 13,280 9,500 9,670 13,500 16,690

 

%Values are in ppm.

cSample appeared to be contaminated.
ND = None detected.

bAll data except the values in the last column were taken from spectrographic analyses.” The data in the last column were obtained from chemical
analyses. .

h

 
.

Table 6.4, Material Balfinces for the Salt Charges Used in VPP Runs

 

 

 

Initisl  Barren Initial NeF Final FV-100 Final Cumulative
Heel in Salt AlF3 from Heel in from Total Heel in After Heel in Waste Recovery Material
FV-1000 Charge Dissolution FV-1000 FV-105 Input FV=-1000 DPT® FM-100 Salt T"?EE'BE%iT—" Belance
Run No. (kg) {kg) © (kg) (kg) (kg) {kg) (kg) (xg) {kg) (kg) kg (%)
Fvs® 0 3.1 0 0 0 3.1 0 14,7 2.5 12.2 14,7 45.8 45.8
DA-1 0 1k9.9 24.3 2.5 3.7 180.4 13.8 184.8 8.3 180.2 202.3 112.1 102.1
DA-2 13.8 132.7 24,6 8.3 2.6 182.0 20.5 161.0 3.3 160.3 184.1 101.2 101.8
UA-1 20.5 11k.0 21.1 3.3 2.5 161.k 19.3 141.8 0.7 143.6 163.6 101.% 101.7
UA-2 19.3 139.2 20.6 G.7 2.5 182.3 0 182.0 1.0 183.5 184.5 101.2 101.6
UA-3 ] 135.9 13.1 1.0 2.5 152.5 Q 155.7 0.3 157.9 158.2 103.7 102.0
RA-1 0 190.3 7.4 0.3 3.0 221.0 48,9 172,6 28.6 147.0 22k.5 101.6 101.9
ra-2°© 48,9 190.3 27,0 28.6 0 294.8 130.3 171.1 16.1 155.0 301.4 102.2 102.1
130.3 0 0 16,1 1.9 1k8.3 6.5 136.0 4.5 133.4 1Lk .4 97.4 101.8
RA-3 6.5 209.6 25.4 4.5 0 2L46.0 73.5 163.3 38.6 124,17 236.8 96.3 100.9
T3.5 o 0 38.6 k.0 116.1 15.% 99.2 1.0 102.2 118.6  102.2 101.1
RA-Y4 15.4 114%.3 13.1 1.0 4.2 148.0 9.6 136.3 1.0 139.5 150.1 101.4 101.1
Total 1408.3 196.6 26.9 1631.8 9.6 1.0 1639.5 1650.1. 101.1
“ppT = dissolver product transfer.
bFVS = freeze valve salt charging.
®Run was divided into two parts,

on

 
 

 

 

L6

Teble 6.4 shows the salt inputs for each run, as well as the total
input. Initial heels in the dissolver (FV-1000) and the fluorinator
(FV-100) are the quantities that were indicated by instruments on the
vessels; the quantities of barren sait and AlF3 were based on known
weights; and the NaF that was added to FV-100 from the absorber (FV-105)
- was measured by FV-100 instrumentation. The recovery for each run is
the total of the final heels (as measured by vessel instrumentation)
and the weste salt (as measured by "before and after transfer readings"

on FV-100 instruments).

The column labeled "FV-100 after DPT" is included to indicate the
degree of reliability of vessel instrumentation. Theoretically, the
value for each run should be equal to the weight of the initial salt
charge minus the sum of the weights of the NaF from FV-105 and the
final heel in FV-1000. Experience has shown that measurements of changes
in vessel contents are reasonably reliable when they are made using a
single set of instruments. The data in Table 6.4 indicate discrepancies

between instruments.

A very poor material balence was obtained for the freeze valve salt
charging (FVS). Since only a small quantity of material was charged and
since instrument heels and line volumes would account for some salt, a
low material balance was expected. However, the balance for the subse-
quent run compensated for the low balance; thus the weights of the final

heel after FVS and the initial heel for run DA-1 are probably incorrect.

Each of the two runs RA-2 and RA-3 consisted of a complete dissolu-.
tion, a partial transfer of salt from FV-1000 to FV-100, fluorination,
transfer of waste salt, transfer of remaining salt from FV-1000 to FV-100,
fluorination, discard of NaF pellets from FV-105, and transfer of waste

salt., Hence, & salt balance is shown for each fluorination.

An oversll material balance of 101.1% is within the anticipated
accuracy of the measurements involved. Therefore, the.reliability of
the instrumentation was adequate for process control, and experience
in the first few runs generated a high degree of confidence in the

measurements.
 

©

k7

The uranium content of the salt was calculated from the salt bal-
ance data presented in Table 6.4 and from analyses of salt samples.
This information was used, along with weights and‘analyses of effluents
(including the product) and accountability data for the material

charged, to obtain the uranium material balances shown in Table 6.5.
T. RELEASE OF FISSION PRODUCTS TO THE ENVIRONMENT

Various quantities of fission products were released through the
3020 stack to thé atmosphere during the fluorination step in the hot
runs (see Table T7.l). Noble gases were released during dissolution, but
they were not detected at ground level. In the least desirable case
(Run RA-b), less than 1 mCi of 108Ru and less than 10 mCi of 131I were
released. A total of dbout 20 Ci of tellurium was released; of this,
6 Ci was T8-hr 132Te, which decays to 2.3-hr !32I and, in turn, to
stable xenon. Calculations to determine the percentage of the maximum
permissible concentration (MPC) represented by these releases were made
by assuming a constant rate of releaseiduring the fluorination period.
The most unfavorable case was 1311 (<0.1% of MPC); howevef, even if the
entire quantity of 1317 had been released during a period as short as
1 min, the combined fraction of MPC for all three fission products
(1297¢, 1311, gna 198Ru) would still have been below 5%. We conclude
that fuel that has been cooled for as long as 25 days can be processed
by the molten-salt fluoride-volatility process with no hazard relative

to the release of fission products.

The quantity of tellurium (20 Ci) that was released during run
RA-4 represented only ebout 5% of the calculated quantity in the fuel
element. The major fraction of the tellurium that was volatilized
(i.e., less than one-half of the tellurium in the fuel) was removed by
the KOH scrubber solution. An attempt was made to remove tellurium in
& trap located downstream from the fluorine scrubber. This trap con-
sisted of nickel mesh heated to 400°C, a gas cooler, and an activated-

charcoal bed maintained near ambient temperature (<50°C). Although
 

Table 6,5, Uranium Balances for VPP Runs

 

 

 

 

Uranium Initisl Total Uranium Distribution of Final
Found in Holdup in in FV-100 and Nonrecoverable Uranium Losses Holdup in
Hydrofluorinator System FV-100 Product Product Product "Cold Trapped" KOH Total Produet
Initial Uranium Total Final by Salt Collection Collection in Collection Cylinder Waste Neutral- KOH in Collection
Holdup Charged Uranium Holdup Sampling System Systen Wt, U U 5u Salt izer Scrubber TFV-100 Systen
Run No. (g) {g) (g) (g) (g) (g) (g) {g) (%) (&) (%) (mg) (mg) (mg) (%) (g)
UA-1 0 592 592 264 328 0 328 325 6T7.06 218 91.98 373 0 0 0.11 110
UA=-2 264 304 568 239 329 110 L3g Lho 67.26 206 91.31 1835 0 240 0.63 lkl
UA-3 239 304 543 391 . 152 14 293 300 67.33 202 91.ko 253 0 171 0.28 91
RA-1 391 323 T1h 533 181 21 272 300 65.64 197 86,88 235 0 119 0.20 75
RA-2s 533 306 839 T34 105 T5 180 0 225 0 30 0.2h 180
RA-2b T34 0 T3k 539 195 180 375 k10 66.76 314 83.00 223k ‘55 1.17 59
RA-3a 539 324 863 TO1 162 59 221 0 - 1079 0 155 0.76 220
"RA-3b 701 0 T01 625 6 220 296 410 €7.05 275 B82.49 1hio 206 2.13 . 19
RA-k4 625 159 T84 600 184 19 203 100 63.97 64 B2.52 ThE 0 0 0.1 138
Cleanout 600 0 600 0? 138 607° 124°¢ o

 

8Total of T g unaccounted for (99.7% material balance).
PCollected as UFg, UOoF, solution, ete.
CGrams of uranium in salt flushes. This uranium could have been recovered by fluorination, but it was not considered worthwhile to do so.

8h

 
Table T.1.

Fission Products Released to the Atmosphere from the VPP During the Aluminum Cempaign

(Quantities expressed in curies)

 

Fission Products
Released
from Dissolver

Fission Products Released from Fluorinator

 

 

 

 

Noble Gases 127m¢ 129q, » 1311 103py, 106gy
Run No. Present Detgcted .Present Detected Present Detected Present Detected Present Detected Present Detected
RA-1 Wr.3 (85kr) 0 2 0 <0.01 0,007 <0.001 0 0.5 0 167 0
RA-2 60 (85kr) 0 518 0 34 ‘ 0.03 0.01 0 Tha 0 L5 0
RA-3 54 (85ky 0 100 0 1M ~0.09% 8.5 <0.001 2766 ~0.003 k91 0. 002
0.3 (133%e) 0 ,
RA-Y 2l (85Kn) 0 48 0 341 20P 2293 <0.01 3986 0 2k2 <0.001°
2261 (133xe) 0
wect, ] 4 x 1076 9 x 1079 8 x 1078 6 x 1072
uCi /em® (see d) (see &) (see d) (see d)
h
Apct> 2.8 x 1073 <6 1072 <1.0 x 1072
during RA-h ' 3 % v x

 

a'I'otal tellurium.

b127Te,+ 1297 + 132‘I‘e; all isotopes assumed to be 1297¢ in the calculation to determine the percent MPC released,

c103Ru + IOGR‘I.‘I..

dWorst case for insocluble material.

eWbrst case for z2oluble material.

fSource: Maximum Permissible Body Burdens and Maximum Permissible Concentrations of Radicnuclides in Air end in Water for Occupational Exposure,

Naticnal Bureau of Standerds Handbook ©9, U.S. Department of Commerce, 1959.

€rs determined at ground level.

The MPC's given are based on a 40-hr week.

hBases for calculation: (1) rate of eif flow through the stack was 43,000 cfm; (2) atmospheric dilution of eir discharged from the 3020 stack was 105;

(3) fission products were released at a constant rate during the entire fluorination period (40 min).

analytically in samples from charcoal traps used to sorb fission products from a sidestream of stack gases.

Quantities of fission products were determined

64

 
 

 

 

50

appreciable amounts of tellurium were held by the trap, a significant
fraction was allowed to pass through to the off-gas stack. Some of the
trapped material was subsequently released after evolution from the
fluorinator had ceased; thus, the trap helped extend the release period.
It appears that, although the quantity of tellurium that was dispersed
did not constitute a hazard, further developmental work in this area

would be advisable.

All of the effluent streams in run RA-U4 were analyzed to determine
the fate of iodine. Of the 2293 Ci of 1311 present at the time of dis-
solution, only 110 Ci was accounted for by analyses. ©Since laboratory
experience has indicated that there is a low probability of finding the
iodine in a fused salt, we believe that the remainder of the iodine was
either "plated out" in the vessels or was discharged in the waste salt.
Of the 110 Ci accounted for, 109 Ci was found in the caustic solution
used for scrubbing the dissolver off-gas and 0.8 Ci was found in the
caustic used for scrubbing the fluorinator off-gas. The quantities of
iodine found in the other streams were insignificant; essentially none
was released to the atmosphere. We conclude that molten—sélt fluoride-
volatility processing of nuclear fuels presents no iodine release
hazard, regardless of the length of the cooling period between irradia-

tion and processing.
8. ACKNOWLEDGMENTS

R. E. Brooksbank, Chief of the Pilot Plant Section of the Chemical
Technology Division, had overall responsibility for this pilot filant.
R. P. Milford was reponsible for coordinating the Volatility Project.
Other supervisors of the Volatility Pilot Plant who made major contri-
butions were: R. S. Lowrie, S. Mann, R. J. Shanhon, and E. L.

Youngblood.

‘Many people in other parts of the Laboratory provided aid and
assistance that made the operation and maintenance of this pilot plant

a success; chief among these were: G. E. Pierce and R. P. Beard of the
 

 

o

LA

o1

" Plant and Equipment Division, W. J. Greter of the Instrumentation and

Controls Division, E. I. Wyatt and C. E. Lamb of the Analytical Chemistry
Divigion, O. J. Smith of the Inspection Engineering Department, and
C. H, Miller and W. A. McLoud of the Health Physics and Safety Division.
The assistance of these, and of many other people at ORNL and at other
AEC installations, is gratefully acknowledged.

We also take this opportunity to express appreciation to Martha G.

Stewart for her invaluable editorial assistance.
 

 

 

 

10.

52

9. REFERENCES

Chem. Technol. Div. Ann. Progr. Rept. June 30, 1962, ORNI-331k,
Pp. 39-45; Chem. Technol. Div. Ann. Progr. Rept May 31, 1964 ORNL-
3627) PP. 29-40.

R. G. Nicol and W. T. McDuffee, Molten-Salt Fluoride Volatility
Pilot Plant: Eguipment Performance During Processing of Aluminum-Clad
Fuel Elements, ORNL-Lk251 (July 1968).

S. Mann and E. L. Youngblood, Decontamination of the ORNL Molten-

- Salt Fluoride Volatility Pilot Plant After Processing Irradiated

Zirconium-Uranium Alloy Fuel, ORNL-3891 (February 1966).

- E. L. Youngblood, R. P. Milford, R. G. Nicol, and J. B. Ruch,

Corrosion of the Volatility Pilot Plant INOR-8 Hydrofluorinator
and Nickel 201 Fluorinator During Forty Fuel-Processing Runs with
Zirconium-Uranium Alloy, ORNL-3623 (March 1965).

R. E. Thoma, B. J. Sturm, and E. H. Guinn, Molten-Salt Solvents
for Fluoride Volatility Processing of Aluminum-Matrix Nuclear

Fuel Elements, ORNL-3594 (August 1964).

M. P. Lietzke and H. C. Claiborne, CRUNCH - An IBM-T0L4 Code for
Calculating N Successive First-Order Reactions, ORNL-2958 (Oct.
24, 1960).

P. D. Miller et al., Corrosion Resistance of Nickel-Base Alloys
Under Hydrofluorinator Conditions with Aluminum Dissolving,
BMI-X-215 (Jan. T, 1963).

P. D, Miller et al., Corrosion Resistance of Nickel-Base Alloys
Under Hydrofluorinator Conditions with Aluminum Dissolving -
Part 1I, BMI-X-260 (Oct. 18, 1963).

P. D. Miller et al., Corrosion Investigation of HyMu 80, INOR-8,
and "L" Nickel Under Fluorinator Conditions for Processing Alumi-~
num Using the Molten-Salt Fluoride Volatility Method, BMI-X-316
(Oct. 30, 196k).

Chem. Technol. Div. Ann. Progr. Rept. May 31, 1966, ORNL-3945,

p. 63.
10, APPENDIXES
 

Sh

10.1 Appendix A: Fission Product Content
of Irradiated Fuel Elements

The radiocactivity levels of the fission products expected to be
present in the spent fuel elements that were processed in the four
hot runs of the uranium-aluminum campaign were calculated using the

CRUNCH code.6 The results are listed in Table A-l.
 

*

-

.

55

Table A-1l. Calculated Radioactivity Levels of Fission Products in U-Al Fuel Elements in
Runs RA-1, -2, -3, and -4 at Start of Processing®

 

Calculated Rediosctivity’ Level in
Run at Start of Processing

 

 

 

{curies) _ 4. © % of Total
RA-1 RA-2 RA-3 RA-L 1/2 RA-1 RA-2 RA-3 RA-4
85Ky 47.3 59.5 54.1 2,2 10.27 y 1.06 0.21 0.09- 0.03
895r 8.5 2252 . 6036 6306 51 0.19 T.77 10.48 8.56
30gr 377 450 %03 179 28 y 8.43 1.55 0.70 0.24
C 9y 17.0 3131 7973 7928 58 0.38 10.80 13.84 10.76
ISzy 31.1 3874 9189 8333 63 0.69 13.36 15.95 11.31
35nb 69.8 7342 14505 7523 35 1.56 25.32 25.18 10.21
M0 148 2.8 0.20
2970 0.03 0.06 0.06 0.02 2.12 x 105 ¥
103gy 0.5 T48 2766 3986 11 0.01 2.58 L.80 5.41
106gy 167 Lys Lo1 242 1.0 y 3.73 1.53 0.85 0.33
Nlpy 0.25 14 7.6 0.02
125gy, 9.3 1k.7 4.1 6.5 2.7y 0.21 0.05 0.02 0.01
127g, 30.9 3.9 0.0k
127pe 1.7 51.k% 100 48,2 90 0.04 0.18 0.17 0.07
1297 33.8 M 341 33 ' 0.12 0.30 0.46
1311 0.01 8.5 2293 8.05 0.01 3.11
132me 232 3.2 , 0.32
133%e 0.30 2261 5.27 3.07
13505 0.3k4 7.9 13 0.01
137¢cs 323 383 342 152 3y 7.22 1.32 0.59 0.21
140pg b2 331 8243 12.8 0.01 0.57 11.19
141ce 761 4059 8559 32 2.62 7.05 11.62
143pr 7.6 hh2 . 8378 13.7 0.03 0.77 11.38
lekce 2338 7703 896L 4550 290 52.28 26.57 15.56 6.18
1478q 0.51 T4 . 3126 " 11.3 0.13 L.24
147py 1081 1730 1676 73k 2.6 y 24,17 5.97 2,91 1.00
143ppy 7.7 2.25 0.01
151gy 1.3 1.26 1.28 0.73 T3 ¥ 0.03

 

&The first nuclide in each decay chain contributing a significant fraction of the total activity at the start
of processing is tabulated. A true total sctivity would also have to include the activity of the short-lived
daughters that are in secular equilibrium with the nuclides listed.

®) plank indicates <0.01 curie or <0.01%.

®The half-life is expressed in days unless followed by y, which designates years.
 

 

 

56

10.2 Appendix B: Dissclution of Fuel Elements

The sluminum dissolution rates are plotted against run times in
Figs. B-1 through B-10. The HF flow rate, the average temperature,

and the salt compositions for each run are shown.
 

L} ]

o7

ORNL Dwg 69-770I

— ITI 140

< SALT (mole %

S oW RATE] INITIAL FINAL 120

= | KF 64.0 53.8 _
t 2tF, 360  30.2 c

w i AlFs 0 16.0 ‘E

- Avg. Temp. 615° C {00 <

<« o

@ w

2 -

o 80 <m:

=

5 :

o ALUMINUM DISSOLUTION RATE €0 J

“ | W

0 T

40

=

2 90% COMPLETE

= 20

= END OF

< DISSOLUTION

 

' 0
0 2 4 6 8 10 42 {14 16 18 20 22 24 26 28 30
DISSOLUTION TIME (hr)

Fig. B-1. Aluminum Dissolution Rate as & Function of Fuel Element

Dissolution Time for Run DA-1l.

ORNL Dwg 70-12432

 

1.8 ' T 180
£
N KF ™ 63, 4 |160
o 2rFy 22.0 18.8
- AlFy 13.6  26.1

1.4 : > Temp, 6169 C 140 .
W =
212 120 E
e >
z . L™
° 1.0 RATE 100
E :
>
o 08
9 ) 90% COMPLETE 80 g
9 06 - 60 in
o ALUMINUM DISSQLUTION RATE - W

I
§ 0.4 ' 40
Z END OF
S 02 DISSOLUTION 20
S
O _
<, _ ' 0
0 2 4 6 B8 10 12 14 16 I8 20 22 24 26 28 30

DISSOLUTION TIME (hr)

Fig. B-2. Aluminum Dissolution Rate as a Function of Fuel Element

Dissolution Time for Run DA-2.
 

58

ORNL Dwg 6€95-7702

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

= 0.7 CONDITIONS 140
S SALT t{mole %)
S [ INITIAL  EINAL
o 0.6 KF €39 55.1 120
= Zrfy 20.8 18.0
| e T15.3 6042"669 =

w v emp. ‘-
Yos . = 100 E
§ o
Z 04 - 80 -
- S — ——_—l
E H FL(;E’ RAT | @
0 w
vy
= ALUMINUM DISSOLUTION RATE L

0.2 40
= [*T—90% COMPLETE I
5
Z |
= 0.1 Lt 20
> —
2 END OF DISSOLUTION
3 ] .

O 2 4 6 8 10 12 i4 16 18

DISSOLUTION TIME (hr)

Fig. B-3. Aluminum Dissolution Rate as a Function of Fuel Element

Dissolution Time for Run UA-1l.

ORNL Dwg 70-12433

 

2.6
£
% 0.6 3 120
= '
1 ~—
w INITI £
5 % i FLognaTE e e | S
m . .
AIFs 33 159 =
z Avg. Temp. 620° C
o 04 ALUMINUM DISSOLUTION RATE L 80
| <
5 END OF DISSOLUTION @
g 0.3 60 o
a 90% COMPLETE b
- w
© 0.2 40
= T
Z
= 0.1 20
>
2
° o 4 6 8 10 12 14 16 18 20 22 24°

DISSOLUTION TIME (hr)

Fig. B-4. Aluminum Dissolution Rate as a Function of Fuel Element

Dissolution Time for Run UA-2.
 

¥

)

29

ORNL Dwg 69-7703

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

*TH CONDITIONS 140
E SALT (mole %)
g 06 ' KF 64.3 57.9 120
= = Zrfy 357 32.2
| AlFy O 99 =
0.5 [ i Avg. Temp. 615° C £
a HF FLOW RATE ! 100 <
m Ao
s i END OF DISSOLUTION— "
04 8o U
o T 11
- 1 é
2 .
o 03 , 60 8
2 | _Le—TALUMINUM DISSQLUTION RATE ' W
o — 1 ! w
0.2 R Y l_ 40 u-
= | T
2 90% COMPLETE-——w
= 01 20
3
< -
© O

 

0 2 4 6 8 10 12 14 16 18 20 22 24
DISSOLUTION TIME (hr}

Fig. B-5. Aluminum Dissolution Rate as a Function of Fuel Element

Dissolution Time for Run UA-3.

ORNL Dwg 70-1243I

0.7 140
© ALT (mol
£ .
o 0.6 KF . 120
x ZrFy 35.7 30.6

AtFy © 14.4

w Avg. Temp. 610° C =
~ 0.5 100 ¢
2 s
2

04 w
= <t
> , , ‘ @
5’ 0.3 1 .——90%I co?nP_LlsTEl 60 S
m .
0 ALUMINUM END OF DISSOLUT! !
O

0.2 40
% T
<
S 0.1 20
3
<

 

0o

o

0 2 4 6 8 10 12 14 16 18 20 22 24 26
DISSOLUTION TIME (hr)

Fig. B-6. Aluminum Dissolution Rate as a Function of Fuel Element

Dissolution Time for Run RA-1.
 

 

 

60

ORNL Dwg €9-7704

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.4 CONDITIONS t40
w SALT {mole %
£ M‘F‘@! I FINAL
o 1.2 KF 624 554 20
x ZrE, 346 308

AlFs 30 138 <
Ll Avg. Temp. 605° C E
E 1.0 HF Fl‘.grv RATE - 100 ..\‘3'
S 0.8 go &
— <
S ] I @
. ALUMINUM DISSOLUTION RATE

§0.6 . ALUMINUM DISSOLUTIO . 60O
LTI :
) ] |t—so°/., COMPLETE

0.4 L 11| 40 b
= , B ’ END OF DISSOLUTION—I» T
2
Z 1 ,
= 02 ' X 20
2 L
< ° || ‘ o

0 2 4 6 8 10 12 14 16 18 20 22 24

DISSOLUTION TIME (br)

Fig. B-T. Aluminum Dissolution Rate as & Function of Fuel Element

Dissolution Time for Run RA-2.

s

ORNL Dwg 70-12434

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_CONDITIONS
= SALT (mole %)
< I iNITIAL  FINAL
o 1,2 KF "62.8 55.1
ax ZrFy 36.8 32.3
= AlFy 0.4 12.6
w Avg. Temp. ST7°C
= 1.0 T HF FLOW RATE
o "o

..'= '
S 0.8 :: i
B ¢ |
J o6 i +—909, COMPLETE
e iy}
7 : '
O 04 L ALUMINUM DISSOLUTION RATE
= |
2 END OF Dlssownon-—-l |
S o2
: ARl
=1
< [ -
0
0 2 4 6 8 10 12 i4 16

DISSOLUTION TIME (hr)

140

120

100

80

60

40

HF FEED RATE (g/min)

20

o

Fig. B-8., Aluminum Dissolution Rate as a Function of Fuel Element

Dissolution Time for Run RA-3.

U
L 3

61

ORNL Dwg €9-7705

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

07 | conomions 1140
P ALT (mole 9,
£ L T A wh
~ 06 KF 62.1 55.5 120
oo ZrF, 36.4 326
=. 90% COMPLETE — AlFy 1.5 1.9 c
| | Avg. Temp. 575°C 100 E
Wl .
: 0.5 HF FLOW RATE :9
a
W
< 80 -
o 0.4 :é
‘_
3 o
35 0.3 60 o)
& h
W
S 02 ___ ALUMINUM DISSOLUTION RATE 40L
>
—-— 1
E 0.1 — 1 E 20
>
= l l
' : : 0
05 2 4 5 8 10 12 14 16 18

 

 

 

 

DISSOLUTION TIME (hr)

Fig. B-9. Aluminum Dissolution Rate as a Function of Fuel Element

Dissolution Time for Run RA-4.

ORNL Dwg 70-12440

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.7 I ' CONDITIONS 140

= ALYUMINUM DiIgS | it mo_s_.L‘e ;
U OLUTION RATE INITIAL FINA

06 3 N KF 613 361 1120
s Zrfy 230 21.0
= AIFy 155 229
w : Avg. Temp. 585°C c
F 05 HF FLOW_RATE 100 ¢
o . o
S 04 80 w
E 5
o o
3
v ul
o) 1]
= w

0.2 40
= END OF DISSOLUTION — T
o
z
3 ol 20
>
3
<

0 0

0 2 4 6 8 ' 10 12

DISSOLUTION TIME (hr)

Fig. B-10. Aluminum Dissolution Rate as a Function of Fuel Element

Dissolution Time for Run DA-3.
 

 

62

10.3 Appendix C: Decontamination of
Pilot Plant Equipment

Decontamination methods similar to those described for decontami-
nating the plant after the zirconium program was com.pleted3 were used
to reduce radiation levels in excess of 5000 r/hr to levels sufficiently
low to permit the plant to be dismantled. Radiation dosage to indivi-
“duals who did the mechanical work did not exceed the quarterly 1l.3-r
allowance (see data in Table E-1). Final backgrounds (prior to equip-
ment removal) were 1 to 5 r/hr in the majority of cell 1 locations; a
maximum reading of 60 r/hr was obtained adjacent to the HF inlet line
to the dissolver. Backgrounds in bther areas of the plant were génerally
less than 1 r/hr.

The decontamination sequence-included flushing with a molten salt,
followed by treatment with three types of aqueous solutions for removing
‘salt film, metal scale, and deposits of radioactive nuclides. The
compositions of the three solutions were, respectively:

Ammonium oxalate, 0.3 to 0.35 M

Aluminum nitrate, 0.1 M; or A1(NO3)3*HNO3, 0.1 to 0.01 M

Sodium hydroxide--hydrogen peroxide--sodium tartrate, sbout
5-1-1 wt %.

This procedure is suitable for use with radiation levels at least
as high as those encountered in the operating period cited, and it does
not result in excessive personnel exposure during the decontamination

and subsequent equipment removal.
 

»

63

10.4 Appendix D: Corrosion of Vessels

Corrosion studies in the Volatility programs were directed pri-
marily at the dissolver (hydrofluorinator) and the fluorinator for
three reasons. First, these two vessels were located in a congested
"no-access" area that was accessible only after extensive decontamina-~
tion, whereas many of the other vessels were located in limited-access
areas. Second, these vessels, which were a part of the head-end por-
tion of the plant, were essential to the operation of the plant, and
their replacement would require a major shutdown (in addition to the
necessary decontamination). Third, corrosion observed on the other
vessels and components was slight, probably because the conditions
under which they were operated (especially temperature and contact with

corrosive chemicals) were less drastic.

Earlier Studies at BMI. Studies made at BMI (Battelle Memorial
Institute) of the dissolver under run conditions indicated a metal
loss of 0.2 mil/month for INOR~8 during sluminum processing, as com-
pared with 2.5 to 5 mils/month during zirconium processing.7 No
:Lntergra.nular attack was observed. Later studies indicated an INOR-8
corrosion rate of approx1mately 10 mils/month during aluminum processing
because of intergranular attack.8 The total time of HF exposure during

the-aluminum runs was 250 hr.

Studies at BMI indiceaeted a corrosion rate of approximately 100
mlls/month for the "L" nickel fluorinator during aluminum alloy pro-
cessing as compared w1th approx1mately 30 mlls/month durlng zirconium
proce531ng, The hlgher rate was at least partially attributed to the
higher operating temperature_[600°c rather than 500°C (ref. 9)] that
ves used in the alfifiinum proceesihg.

Corrosion of the Dissolver. The total corrosion of the dissolver
during the 10 aluminum runs was 5 mils, as previously'reported.lo - This
value was obtained by using pulse-echo and Vldlgage technlques. Previous

corrosion deta for the dissolver are reported elsewhere.
 

 

 

6L

Fluorinator. (a) Vessel. — Corrosion data for the fluorinator
during zirconium processing are reported elsewhere.h (However, the
data for the last 11 zirconium runs were not reported in that reference
because the fluorinator was not examined between the end of the zir-
conium program and the start of the aluminum program.) Corrosion
results obtained on examination of the fluorinator after the plant
was dismantled following the aluminum program are reported in Table
D-1. The values listed in Table D-1 are for 50 runs (40 zirconium

runs and 10 aluminum runs) and all aqueous decontamination sequences.

The region most vulnerable to attack was the salt region, as noted
previously.h The maximum corrosion was less than 3/4 mil per run; the
average was less fhan 1/2 mil per run. Visual examination of the in-
side of the fluorinator revealed no evidence of excessive attack. Com-
parison of the losses in total wall thickness (Table D-1) with data
obtained in the earlier corrosion studyh revealed that metal losses of
gpproximately 10 mils apparently occurred in the bottom of the vessel

(but not in the top) during the aluminum runs.

(b) Corrosion Rods. — The corrosion rods that were placed in
the fluorinator in late 1962 were removed when the plant was dismantled
following the aluminum program. Results of visual insPéction of the
rods are summarized in Table D-2., The rods were not examined metal-
lographically because of the termination of the Volatility program in
mid-1967.

Other Vessels and Components. (a) HFV-2207-1 (HF inlet line to
the dissolver). — Visual inspection of the 3-ft length of INOR-8 line,
including the elbow, showed only minor scratches. No evidence of
leakage was noted when this section of pipe was pressurized to 35 psig.
This portion of the line was examined because it had failed during
early volatility processing.h At that time, however, the line was

made of Inconel instead of INOR-8.

(v) Fluorine Supply Tanks. — Inspection of the inside of

tank NB-1432 on June 7, 1962, revealed no corrosion; no leakage occurred
 

P

65

Table D-1. Bulk Metal Losses from the Nickel 201 Fluorinator
During Fifty Runsa.’b and Associated

Aqueous Decontaminations in the VPP

Exposure times: 90.7 hr of fluorinec’d; 2962 hr of molten sa1t®

 

Corrosion Rate

Total Wall Mils per hour Mils per month

 

Section of the Thickness of F, Exposure of Molten Salt
Fluorinator Loss (mils) 2 Exposure
Measured® Max. Avg. Max. Avg. Max. Avg.

Top 16-in.-diam 11 3.5 0.12 0.039 2.7 0.85
section

Top cone 13 8.4 0.1k 0.093 3.2 2.0

Neck 24 18.6 0.26 0.205 5.8 4,51

Bottom 16-in.- 27 19.2 0.30 0.212 6.6 4,66

diam section

Bottom cone 29 21.0 0.32 0.232 7.0 5.10

 

aForty zirconium runs and ten aluminum runs.
Psece ORNL-3623 (ref. 4), Table 12.

®Fluorine exposure time does not include vessel exposure during de-
sorption.

dyalues given include those presented in Table 12, ORNL-3623 (ref. k).
A breakdown of exposure times is as follows:

 

 

 

F» Exposure Time Molten Salt Exposure Time

No. of Runs ~ (nr) | (hr)
29 (value from 57.6 1922

ORNL-3623)
11 (finel runs in 18.1 604

Zr program) -
10 (runs in A1 - 15.0 | 436

program) :
50 (total runs) 90.7 2962

®Measurements were made at 1l-in. intervals in the south quadrant of
the vessel.
 

 

 

66

Table D-2. Condition of Corrosion Rods on Removal from

Fluorinator After the Aluminum Runs

 

 

Material Results of Visual Operation
INOR-8% Brown film; some corrosion in salt region.
"L" Nickel? Brown film; éome corrosion ét middle #nd bottomn.
HyMuFBOb Thin brown film over full length; some loose material

on top 4 in.; uniform diameter.

Specimen lb’c | Brown film; slight loss in width at bottom; warped.
Specimen 2b’c Same as for specimen 1.
Ni-Mgb ' Brdwn film; corrosion similar from top to bottom.

 

®Installed on November 13, 1962,
Prnstalled on December 11, 1962.

“Weld test units fabricated from "L" nickel and INOR-8, using the
following weld materials: Inco-82, INOR-8, Inco-61, and "L" nickel.
 

 

67

during a T5-psig pneumatic test. The 1/2-in. gage outlet on the rear
head appeared to be cavitating in the heated-affected zone of the weld.
The filled weld around this outlet on £he tank exterior was thought to
be adequate to take care of the condition on a temporary basis. All
stop valves in the manifold section were checked and reworked if
necessary. The external surfaces were found to be corroding; painting

of these surfaces was recommended.

Inspection of the inside of tank NB-1L433 on July 28, 1966, revealed
no corrosion; no leakage was observed when the tank was pressurized to
75 psig. Inlet and outlet stubs that had been welded to the rear head
showed no corrosion. All stop valves in the manifold section were
checked and overhauled. External surfaces of the unit were satisfactory;

the unit was repainted.

(c) Other Vessels. — No visible corrosion was detected visu-
ally in (1) the movable-bed absorber (FV-105), (2) the flash cooler
(FV-1001), (3) the HF condenser (FV-2001), or (4) the fuel element
charging chute on the dissolver (FV-1002).
 

 

 

68

10.5 Appendix E: Radiation Safety

Penetrating radiation from mafierials being processed required that
Veach piece of equipment containing the material be heavily shielded to
prevent exposure of operating personnel. Access to the procesé equip-
ment was carefully confrolled-at ali times during normal operation and
while maintenance or decontamination operations were in progress. The
physical form of the material involved (dust particles, liquids, or

_ gases) governed the type of protective clOthing and respiratory equip-

ment that was used in these shielded areas.

The average rate and the maximum rate of personnel exposure to
radiatiofi in the alufiinum program were about the same as those encountered
in the zirconium progrem. In each casé, exposure rates were highest
during plant decontamination procedures because temporary, unshielded
piping was connected to "dead end" process lines in normal work areas
for recycle of solutions. Even under these conditions, however, the
maximum exposure to any individual ‘did not exceed 75% of the recommended

meximum permissible dose to body organs for a calendar quarter.3

Unusual occurrences¥* were less frequent in the aluminum program
than in the zirconium program, largely because experience had been
gained in making design and operational changes. Because we anticipated
higher radiation levels from the shortef-cooled elements that were to be
processed'in the aluminum campaign, lead shielding was installed in many
areas that formerly did not require it; corrective actions were taken as

problems developed.

Personnel Exposure. — Radiation exposure to personnel was due pri-

marily to the presence of 125Sb, ?5Zr, and 95Nb in the dissolver off-gas

 

* : '
An unusual occurrence is defined as one which may result in (a)

personnel exposure in excess of the maximum permissible, (b) cleanup
costs or property loss in excess of $5000, (c) an incident of public-
relations significance, or (d) exposure of the off-site population to
radiation in excess of the maximum permissible.
 

¥

L

69

system and to the presence of 237y, 99Tc, 103py, 106Ry, °9Mo, and 237mp

in the fluorinator off-gas and product collection systems.

The aversge exposure rate remained fairly constant during the pro-
cessing of dummy and irradiated fuel elements, but increased by about

a factor of 2 during decontamination (Teble E-1).

During the processing of irradiated fuel elements, the maximum dose
received by an individual during a single day was 80 mrads. This expo-
sure occurred at the conclusion of run RA-4 when an operator removed the
product receiver from the receiving station in cell 2 and hauled it to
the sampling‘room in a lead-lined drum. The product was highly radio-
active because of the unusually high 237U content. The "cutie pie"
reading of the unshielded product cylinder was 25 r/hr at a distance of
1.5 in.

During the decontamination period, the maximum dose received by an
individual in & single day was 150 mrads. This exposure occurred during
replacement of a damaged rubber gasket on the waste salt nozzle sealing
jack. Removal of the gasket was slow because the nut that held the end
flange of the drain line tightly against the gasket was located on the

underside of the gasket holder and was not easily accessible.

In summary, the maximum radiation exposure received in the alumi-
num campaign did not exceed about 75% of the ORNL limit of 1.3 rem/quarter,
or 5 rem/year. The average VPP personnel radiation exposure was about

50% of the maximum.

Unusual Occurrences. — No unusual occurrences were experienced
during the processing of dummy or irradiated fuel elements in the alumi-
num program. However, two occurred at other times: one during prepara-

tion for startup, and one during vessel decontamination.

Just before the first dummy fuel element was processed, éome pre-
viously used air-operated valves were dismantled and deconteminated in
a "hot" sink preparatory to iepair. During this procedure, the hair
end the nostrils of the operator in charge were contaminated by a loose

powder found inside the valves. However, subsequent investigation
 

 

TO

Table E-1. Radiation Exposure of Personnel During the

Processing of Aluminum-Clad Fuel Elements

 

 

 

Type of Run  Exposure (mrads)
or Operation? Max. /Day Max./Week Max,/Quarter Avg./Quarter
DA 60 TS 450 200
UA | 50 120 500 130
RA 80 150 - 505 2ko
Decontamina~

tion 150 190 940 LL4o

 

 

®pA - dummy elements containing aluminum only.

UA - dummy sluminum elements "spiked" with unirradiated UFy.
RA - LITR and ORR fuel elements cooled 18 months to 25 days.
|

 

Tl

showed that the operator had received no internal exposure and only
negligible external ekposure. Analysis of the incident indicated that
a ventilated hood was needed over the sink and that a mask should be

worn when contaminated egquipment was being opened.

After the aluminum series had been completed, a radiochemical spill
(of material that had not completely drained from a pipe) occurred
during an attempt to remove a valve bellows assembly from an inactive
HF charging system pipe line south of Building 3019. The bellows as-
sembly was needed to replace one that had failed in the combination
caustic sampling and temporary decontamination solution recycle system
in cell 2. When the valve was opened slightly (a step in the valve dis-
mantling procedure), some radioactive liquid in the pipe flowed past
the operator, as signaled by a personal radiation monitor. This liquid
continued along the pipe to a previously dismantled valve and overflowed
to the ground. The blackiop area under the dismantled valve was decon-
taminated by flushing with water, chipping, and vacuum cleaning. Inves-
tigation showed that the three people in the vicinity at the time of the
incident received no internal exposure and only negligible external

exposure.

Control of Exposure. — Radiation exposure to personnel was controlled

 

by teking appropriate action following thorough, frequent checking of

the work areas during the runs. High radiation backgrounds were reduced

by shielding vessels and pipes with lead plate, by discerding contaminated

solutions, and by backflushing filters. The number of cell entries by
operating personnel was safely reduced by using revised operating proce-

dures. The amount of radioactive material that escaped to the atmosphere

- was decreased by decreasing the purge rates and the duration of the purges.

A 35-point radiation check of the area was made at least once per
8-hr shift during each run to determine any changes in the background.

Date are summarized for runs RA-1 through -4 in Teble E-2.

Some of the measures tsken to reduce the radiation exposure are
discussed below. For exemple, the installation of 1/2-in.-thick lead

shielding reduced the background radiation at the caustic sampler in
 

72

 

 

Table E-2. Radiation Backgrounds® in the Various VPP Work Areas During Runs RA-1, -2, -3, and -4

 

 

 

 

 

 

 

 

Run Fo.
"RA-1 RA-2 RA-3 RA-L
Before After Before After Before After Before After
Location Run Run Run Run Run Max. Run Run Max, Run
Cell 2, HF System
FV-TCO-1C, HF filter 3k 30 30 120 28 200 36 30 300 160
HCV-1003-1, HF catch tank drain valve T © 8 T 20 9 48 17T 16 60 42
Caustic sampling staticn 14 20 13 65 12 35 3 20 85 50
Suction line of FV-U4201, caustic pump 3R 30 35 65 26 %10 25 10 Lo 25
FV-h202, HF pump, remote head 18 20 22 48 24 100 28 25 60 48
FV-120T, HF vaporizer 10 8 9 15 8 16 9 12 k2 28
Cell 2, UFg System
Heated duct at enti'y from cell 1 43 ks i1 58 N T0 ks 46 200 100
FV-120-A, MgFz bed _ 8 7 8 15 9 ko 32 12 300 300
FV-723, product stream filter 5 6 5 8 6 30 13 12 120 95
FV-220, cold trap 5 S 5 6 12 10 9 35 27
FV-121-A, chemical trap 11 13 1k 49 40 120 105 13 65.000" 330
Penthouse, HF System
FV-T500, off-gas liquid trap 1 6 18 15 35 24 17 96 90
Qff-gas line from FV-1009, caustic neu- 2 2 3 3 10 L 5 15 10
tralizer '
FV-9500, flame arrestor 6 10 5 12 5 1k 6 8 60 35
Dissolver vent line at filter FV.T501 1 2 3 25 11 35 3 24 66 22
Transmitter rack (back) 1 1 1 8 2 6 L 3 200 82
Penthouse Fy System
FV-153, liquid trap in off-gas line : 5 8 6 32 1k 22 14 8 600 180
FV-150, caustic scrubber, top 19 30 20 205 110 140 35 1n 1,000 160
FV~150, caustic scrubber, bottom 4 5 5 25 - 15 120 10 T 800 170
FV-152, caustic surge tank 4 8 T 6 8 95 85 15 5,500 400
FV-450, caustic pump 45 Lo ks 155 Lo 100 100 Sk 1,800 730
Tellurium trsp, FV-154, nickel wool - - L 16 6 11 7 5 600 Lko
Tellurium trap, FV-155, charcoal - - - - 1 8 2 1 Lho Lho
0ff~Gas Scrubber System
FV-164, first stage of scrubber 14 22 17 17 .15 25 19 26 43 43
FV-16L, middie of scrubber 4 b y b L 6 6 6 23 23
FV-16L4, scrubber entrainment separator 1 2 2 2 2 y 4 y 20 20
FY=765, filter 6 5 6 6 6 8 8 9 22 22
FV-165, surge tank 2 2 2 3 2 T T 5 6k 6l

 

%values are given in mr/hr.
bArter nitrogen-sparge of salt in fluorinator (prior to fluorination).
 

13

cell 2 to 20% of the previous value. The same thickness of lead re-
duced the background at the NaF trap (FV-121) to about 25% of that

before installation. This same thickness also reduced the background

for the fluorinator off-gas scrubber (FV-150) to 10-25% of the unshielded
background. Approximately 1 in. of lead that was placed over the inlet
line for the caustic circulating pump (FV-L20l) in cell 2 reduced the

background to 6% of that observed earlier.

The dumping of fluorinator off-gas scrubber recycle caustic solu-
tion reduced the background for the storage tank (FV-152) to about 10%
of that prior to dumpiné. At the same time, the background for the
caustic circulating pump (FV-L50) was reduced by about 40 to 50%. Back-
flushing of the dissolver off-gas filter (FV-TOO1C), using waste liquid
HF, reduced its background to about 25% of that noted previously.
 

 

Th

10.6 Appendix F: Index of Volatility Pilot Plant Log Books

The log books listed below, along with the run sheets and the
recorder charts, comprise the primary record of the VPP operations
described in this report. The run sheets and the recorder charts will
be destroyed two months after this report is issued. The log books

will be retained permanéntly at Oak Ridge National Laboratory.

 

 

Laboratory
Records
- VPP Log Notebook
No. No. Inclusive Dates Subject Matter
1 to 15 “ 7/11/56-10/7/59 ARE Program
16 to 43 | 10/19/59-9/11/63 U-Zr Alloy Program

LY A-2964 9/12/63-1/22/64 Preparation for U-Al Pro-
gram

45 A-2226 1/22/6L4-3/26/64 Preparation for U-Al Pro-
gram

46 A-3389 3/30/64-5/1/6L Run DA~1 (aluminum, no
uranium)

L7 A-3390 5/1/64-6/2/64 DA-1

48 A-3h2p 6/2/64-6/26/6L DA-2 and UA-1 (uranium
and aluminum, non-
irradiated)

Lo A-3423 6/26/64-8/4/64 UA-1 and UA-2

50 A-34Th 8/4/6L4-9/10/64 UA-3 and RA-1 (irradiated)

51 ' A-3475 9/11/64-10/19/6L RA-1 and RA-2

52 A-3k476 10/19/64-11/16/64 RA-2, RA-3, and RA-L

53 A-34T7 11/16/6h-12/14 /6L RA-3 and RA-4 (short-

‘ : cooled)
5k A-34T8 12/14/6k-2/2/65 Cleanout and Shutdown

(cont.)
w

>

 

 

Laboratory
Records
VPP Log Notebook

No. No. Inclusive Dates SubjJect Matter
55 A-3T11 2/2/65-3/23/65 Cleanout and Shutdown
56 A-3T1h 3/2L4/65-5/27/65 Complete VPP Log Index

A-6105 Sample Log

A-6106 Sample Log

 
 

 

  

 

 
 
£0%

 

1T

ORNL-L5TL

UC-10 — Chemical Separations Processes
for Plutonium and Uranium

INTERNAL DISTRIBUTION

1. Biology Library
2-4., Central Research Library
5. ORNL — Y-12 Technical Library
Document Reference Section
6-25. Laboratory Records Department
26. Laboratory Records, ORNL R.C.
27. E. D. Arnold

28, J. E. Bigelow
29. R. E. Blanco
30. J. 0. Blomeke
31. R. E. Brooksbank
32. K. B. Brown

33. F. N. Browder
34. W. D. Burch

35. G. I. Cathers
36. W. L. Carter
37 J. M. Chandler
38. C. F. Coleman
39. D. J. Crouse
LOo. F. L. Culler
41. F. L. Daley

42. D. E. Ferguson
43, L. M, Ferris
Yy, B. J. Frederick
45. H. E. Goeller
46. A. T. Gresky
L7. M. B, Herskovitz
48. R. W. Horton
49. L. J. King

50. F. G. Kitts

51l. B. B. Klima

52. C. E. Lamb

53. R. E. Leuze

54, B. Lieberman

55. R. B. Lindauer
56. R. S. Lowrie

57T H. G. MacPherson

EXTERNAL DISTRIBUTION

58.
59.
60.
61.
62-63.
6h.
65.

»

BEHrEsYdSrraoqggEHg QOO W

’.:Utl:lta’-i"tll:l:l':d?:l.i:!_ztb<::£.‘tfl:‘fl

Mann

. McDuffee
McLoud

. McNeese
. Miliford
Miller

. Mossman
. Nichols
. Nicholson
. Nicol

. Parrott
. Pashley
. Pierce

. Ruch

. Ryon

. Schaffer
. Shannon
Shannon
. Shappert
. Skinner

oD QEYEDdEPEa

ey

. Martha Stewart

Stockdale

Sundberg

Unger

A. Vaughen

. Weinberg

. Whatley

Whitson

. Wyatt

Wymer

. Youngblood

. Emmett (consultant)
Katz (consultant)
Margrave (consultant)
. Mason (consultant)
Richards (consultant)

*

*

uimrt*tin:r*anEplytz(aga:bfa

-

96-98. W. H. Carr, Nuclear Project Engineering, Allied Chemical
Nuclear Products, Inc., P. O. Box 35, Florham Park, New

Jersey 07932

99, W. H. Lewis, Nuclear Fuel Services, Inc., Wheaton Plaza
Building, Suite 906, Wheaton, Maryland 20902

 
 

 

100.
101.

102,
103.
10k,

105-240,

78

F. W. Miles, General Electric Company, Midwest Fuel Recovery

Plent, Morris, Illinois 60450

S. Smolen, General Electric Company, Midwest Fuel Recovery
Plant, Morris, Illinois 60450 |

J. A. Swartout, Union Carbide Corporation, New York 10017
Laboratory and University Division, AEC, ORO

Patent Office, AEC, ORO

Given distribution as shown in TID-4500 under Chemical
Separations Processes for Plutonium and Uranium category
(25 copies — NTIS)