ORNL-TM-1852.txt

  

OAK RIDGE NATIONAI. lABORATORY
- - operated by

UNION CARBlDE CORPORATION
-”'"""lj;-;'NUCLEAR DlVlSlON B
e for the s
U S ATOM'C ENERGY COMM'SS'ON
- f;ORNL TM"IBSZ

 

- K

   

—"7';':C°P" NO fc‘n:f? L

~ oaTe- ,_June_ 1967

  

ool e 'FUEL AND BLANKET PROCESSING DEVELOPMENT S ERSTI BR!C':’S |
= w1 TR MOLTEN SALT BREEDER REACTORS ..*-T' | o

" M. E. Whatley H‘G‘ ‘-i—i‘-f 1 --——§‘

   
  

 

. /g - Evaluations of Molten Salt Breeder Reactors have pointed
Lo © o the direction of desirable design, research, and development
. _for improving the economic and breeding performance of these
...+ gystems, In design, the conventional concept of processing.
oo fuel din facilities that are separate from the reactor plant
. needs to be abandoned in favor of integrating the processing -
- =" operations dlrsctly into the reactor building. For the MSBR,
"7 " only a small space is required adjacent to the reactor cells
oo for process1ng cells. - This arrangement allows signlficant
q_;;{*'fls_rsaV1ngs in cap1tal operatlng, shlpplng, and 1nventory costs.

 

    

e Research and develepment have shown that 1rrad1ated MSBR_“ o
. fuel can be- decontaminated in a four- step process con51st1ng B
% 5\\ o - “of fluorination, UFg sorptien, vacuum distillation, and - .. .
N fm;;:fi}a;;g;reductlon-reconstltutlon. These operations recover not only SR
7 .o - - the uranium but also: theeLiF-Ber carrier, Fluorlnatlon,and .

. sorption technologies are well developed for batchwise opera-
© . tion, having advanced through the pilot plant stage.. However,r:,'

- . for MSER appllcatlon, these two steps,as well as distillation
.. and reductlon-reeonstltutlon,must be developed for- contznuous |
~~ . 'pperation. These four unit Pprocesses are rather. 51mp1e and-

n'gedastralghtforward' however, vacuum distillation requires a hlgher
. temperature (~1000°GC): than has been encountered prev1ously in 'f}n~-

",‘*"f R f.-molten salt: processing;' L | s

gé - - . . - E

T o . NOTICE This document eontains informotion of a preliminary noture .. . . - 7 :
8 Leed L T e and was prepared pnmun!y for mtemul use at the Oak Ridge National -~ - oL

L T e  Laboratory.- 1t is. wb;ec‘t fo. revmon or correction nnd therefora does B O

T - not represent a fmd report.. : Sl s

  

 

  

oy
 

 

: LEGAL NOTICE

 

 

This report was prupcrcd as an uceoum ef Govornm-m sponsorod work. Noiihor the Um!ed Siflhs,

nor the Commission, nor any person acting on behalf of the Commission:

A. Makes any warranty or nprnunfuhon, .xpr.ss-d or implisd, with respect to the aceuracy,
completeness, or usefulness of the ll'lformuflon contained in this report, or that the use of
any information, apparatus, mathod or process disclosed in this report may not infi'ingl .
privately owned rights; or -

B. Assumes any licbilities with respect to the use of, or for damages resulting from the use of
any infermation, apparatus, method, or process disclosed In this report. s

As used in the above, "person acting on behalf of the Commission® includes any employee or -

contractor of the Commission, or employse of such contractor, to the extent that such smployee

or controctor of the Commission, or employse of such contractor prapores, disseminates, or
provides access to, any information pursuant 10 hil -mploymom or contract vmh the Commission, -
or his cmployment with such contracter, :

 

 

&

¥

 

 

 

 

 

:
z
i
'
) i':*fi »)

-3 -

» Fertile stream processing steps consist of protactinium
?emoyal, fluorination, and UFg sorption. Protactinium removal
is singularly important because of the improved breeding
performance withlow protactinium concentrations in the fertile
stream. Relatively high processing rates are required.

Xenon poisoning is kept low by injecti
. : g ker ng a purge gas
d1rectly£;nto the circulating fuel stream. An in-lgnegstripper
removes the gases, which are routed to a charcoal
retention and decay. roost system for

- This docum?nt describes the fuel and blanket processes
for ?hg MSBR, giving the current status of the technology and
outlining the needed development. It is concluded that the

‘principal needs are to develop the vacuum distillation and

Protactinium removal operations, which have been demonstrated

in the laboratory but not on an engineering scale., A program to
develop'continuous fluoride volatility, liquid-phase reduction-
r?constltution, improved xenon control, and special instrumenta-
tion should also be a major developmental effort. An estimate of
manpower and cost for developing MSBR fuel and fertile processes
indicates that it will require 288 man-years of effort over a
7-year period at a total cost of about $18,000,000. |

 

 

 

 

LEGAL NOTICE

This report was prepared as an account of Government sponsored work, Neither the United
‘States, nor the Commisaion, nor sny person acting on behslf of the Commission: . :
% A, Makes any warranty or representation, expressed or implied, with respect to the accu-
““racy, completeness, or usefulness of the information ocontained in this report, or that the use
_of say information, apparatus, method, or process disclosed in this report may not infringe -
. privately owned rights; or - - ' s ’ .
~i % P. Assumes any liabilities with respect to the use of, or for damages resulting from the
-use of any information, apparatus, method, or process disclosed in this report. -
" As used in the above, *person acting on behalf of the Commission® includes any em-
ployee or contractor of the Commission, or employee of such contractor, to the extent that
_such employee or contracter of the Commission, or employee of such contractor prepares,
 disseminates; or provides sccess to, any information pursuant to his employment or contract
. with the Commission, or his employment with such contractor. i _ -

o R T

RISIRIBLTION QF TS ROCUMENT IS unymires
 

 

 

 

 

_ L -

TABLE OF CONTENTS

Page No.
Abstract -------------------------------f----—f—-? ------------------ 1
INtroduction ====e==--=-ommcmmmm e e cce e esm—— - --—-- 6
Elements and Requirements of MSBR Processing --—--~+--------?--~?--- 7
Objectives ====-croomrm e ee 7
Design Features =-----=-mcecccccrcmcmccccccmcr e e e -- 8
Process Operations ----- e e cmcmccmcceeeeee i 8
Description of Process -----=c-c-cemcmmmme e —————— -9
Fuel Stream Process --;f ----- mmem e cc e e - cteaae -== 9
Fertile Stream Process ------- ————— cmmmemem—————— mmmmmm—mm e e 17
Off-Gas Treatment ~-------ecmomccaacaaan e 18
WasteVStarage ------------------------------------------------- 19
Status of Processing Technology and Needed Development =-----=--=--=- 20
Fluoride Volatility Process --=-=-----mcmmcccccccmc e e 20
Continuous Fluorination =------eeceammcmmcmcccccccccccccmcee e 25
UF6 Purification by Sorption ~-----—---—cmemmcmm o 26
Cold Traps -----=—==mmemmemc e mmmm e 28
Vacuum Distillation ------cccecmmccmmrerrcr e rme e e 31
Puel Reconstitution ---------ccccmm e - 33
Fuel Clarification ===---me-momm oo ceemeeeeeeeee= 35
Protactinium Removal ~-------c-e-c-mecmmmoccmcmmmccmmmcme—mann- 36
Waste Handling and Disposal -------m-ccmmmcmmmmcccmc oo 36
Process Control ----=cececmcmcreccmcccacrccmmcccce e e e 38
Development of Alternative Processes and Improvements -------------- 39
Iodine Stripping ---------c-cecmmmmcm e - 39
Use of Additives in Vacuum Distillation ----c-ccccomccmacc- -- 39
Reduction-Coprecipitation and/or Electrolysis -----------c----- L0
Ion Exchange with Nonfluoride Solids ~-—=---r-—mrmecccmcmcaeaaa 1]
Schedule of Manpower and Cost for Development Program -----------za- )j2
Continuous Fluorination ==e-----==eemcmmmmocmmcomomceocaeecen
Sorption =-er--e-ccemcncrcccnc e e ———— Ll

Carrier Salt Distillation and the MSRE Experimentr-é ----------- Ll

L«
L

)

)

=)

- c -

TABLE OF CONTENTS (cont.)

Page No,
Reconstitution of Fuel -=-==m==m-comcomoo oo ee oo oo L5
~ Blanket Processing (Pa) ~----cmmmmm oo oo e L5
Filtration and Salt Cleanup =------------ oo 46
Process Off-Gas Handling'#--;—--4---—-—----—---—--4-----¥ ------ L6
Alternative Processing Schemes and Process Improvements ------- L6
Special Instrumentatibn'and'Rrocéss Control -=-------- e L7
Waste Handling and Disposal ==-=eeeememcecmccecmcccccmmceecen== |7
General Design ——--=see-tcmmamocmicdom e a e —————— L7
Salt Production =-=-c-eecmm o e e - 1,8
Engineering Test Unit e mmmemmmmmm - ——————————— L8
Summary of Development Costs ===-==-mmcmmmommo e eeemeemee 148

References ---------- e e L9

 
 

-6 -
 INTRODUCTION

The breeding potential of a molten salt reactor operating on the
thorium-uranium cycle has been recognized for a number of years, and in
1959 a study was initiated to compare the MSBR on the same bases with
other thorium-uranium breeders, namely, aqueous hamogeneous,'liquid bismuth,
and gas-cooled reactors. The comparisons'were made on the bases of fuel
yield (percent of fissile inventory bred per year) and fuelrcyéle cost, and
the results® confirmed the belief that the MSER was attractive in both
regards. Evaluations of MSBR concepts havecontinued.toward optimizing
nuclear and economic performance and have resulted in the generation of
more reliable information on the system. These evaluations have also
sefvéd to point the direction of desirable development and résearch.

Whereas the initial studies were made on the bases of (a) on-site and (b)
central processing plants using existing processing techniques, the savings
that would result by using an integrated processing plant,. i.e., (a), soon became
apparent. Furthermore, a new processing concept for recovering the fuel
carrier salt by vacuum distillation had been shown by K.'ell:y'LL to be feasible,
considerably enhancing the economic appearance of the MSBR. These aspects
of the fuel cycle were studied by Scott and Carter- for a 1000-Mw (elec-
trical) system. Protactinium removal from the fertile stream was not
considered as a processing step in these studies because no satisfactory
process was apparent; however, more recent experiments have suggested a
mechanism of protactinium removal, whereby significant. improvement in the
breeding performance of an MSBR can be achieved. | |

These newer design and processing concepts have been incorporated into
more recent physics and economic calculations to define an optimum MSBR.

The discussion that follows treats processing of MSBR fuel and fertile
streams in this same respect. First, processing requirements for an
economic molten salt breeder are discussed; second, the process is described;
| third, the status of the technology and needed development for each process
step are described; fourth, attractive alternative processes to the main-
line operations are discussed; and, finally, a schedule of manpower and

costs for a developmental program to solve critical MSBR processing problems

W) (\‘ ‘

T
"C?

)

M

-~ 7T =

and to provide the processes and equipment for the processing plant for an
Engineering Test 'Unit is included.

ELEMENTS AND REQUIREMENTS OF MSBR PROCESSING

One of the most attractive features of a two-region, molten salt
reactor is the ease with which the fuel and fertile streams can be processed
for removal of fission products and recovery of bred material. The fluid
streams are easily removed from and returned to the reactor without disturb-
ing operations, and the processing methods are relatively simple and straight-
forward., On-site processing is a primary requirement of a breeder reactor,
which would be at an extreme disadvantage if a sizable inventory of fissile

material were held up in decay cooling and transit.
Objectives

The fuel stream of the MSER is a mixture of LiF-BeF,-UF) in the
approximate molar proportions 63.6-36.2-0.22 mole %; the fertile stream
is a mixture of LiF-BeF,-ThF, of approximate composition 71-2-27 mole %.
The primary objective of the fuel process is to recover both uranium and -
carrier salts sufficiently decontaminated from fission products to ensure
attractive breeding performance of the reactor. To discard the carrier salt
with the fission products is not economical since both lithium and
beryllium”are expensive campohents (the former being enfiched to about
99,995 at, 7Ll). For the blanket processing, the objectlve is to remove
the bred fissile materlal on a sufflciently rapid cycle to. minimize the

inventory of fissile materlal the flSSlon rate, and concentration of

| fission products in the blanket Asalt. This can ‘be accompllshed by
.iremoving the 233U on a- relatlvely fast cycle but even more effectively

by remov1ng the protactinlum precursor rapldly enough so that its con-

233

centration in the blanket is. 1ow. A low Pa concentratlon is desirable

because each capture of a neutron by an atom of 33Pa results in the net

2334y,

loss of two neutrons (effectlvely two atoms of bred

 
 

 

 

-8 -
Design Features

An essential element in the design of an MSBR processing plant is

| that of keeping the ofit-of-reactqr inventory low to decrease inventory
charges and improve the fuel yield, which is the fraction of fissile'
inventory bred per year. The logfcai way to achieve this low inventory

is to élose-couplerthe processing plant with the reactor, and‘it is pro-
posed to integrate the two operations. Processing.equipment"Will be
located in cells adjacent to the reactor cell, and a small portion of

each circulating stream will be metered semicontinuously to the processing
plant. Most processing operations are continuous: the fuel stream being
purged of fission products, fortified with makeup fuel and carrier, and
returned to the reactor core; the fertile stream being stripped of its |
protactinium and uranium, fortified with makeup thorium and returned to
the blanket., It is not necéssary tb allow long decay periods before
processing since the reactants are either gaseous (F2) or solid (granular
NaF) and are not affected by strong radiation fields. Cooling periods
before processing no longer than one day and perhaps as low as a few hours
should suffice. The length of this period depends upon the design of the
continuous fluorinator and the ability to control the fluorination tempera-
ture. Removal of fission product decay heat is a principal consideration
throughout the plant. -

In the integrated plant all services available to the reactor are
available to the chemical plant. These include mechanical equipment, |
compregsed gases, heating and ventilating equipment, electricity, shop
services, supervision, etc. The cost savings for an integrated facility -
arerimmediately'apparent'when one considers the large amount of duplica-
tion required for separate plants. VOnly-a relatively small space is needed
for‘the processing equipment as compared to that needed for the reactor and

power conversion equipment so that the additional building cost is small..
Process QOperations -

Four major operations are needed to sufficiently decontaminate the

fuel stream of an MSBR. These are fluorination, sorption of UF6’ vacuum

-
f\(:* 9

)

n

-9 -

distillation, and salt reconstitution.  The fertile stream requires
fluorination, sorption of_UFé, and for maximum effectiveness includes
protactinium removal, These operations represent the most straightforward
processing for achieving a high-performance molten salt breeder. The
technology for fluorination and sorption is well developed {through the
operation of the fluoride volatility pilot plant at ORNL) the other opera-
tions have been demonstrated in small engineering experiments and/or in
the labératbry. The process for eachIStream_is-capable of economically
recovering more4than,99.9% of the uranifim, 94% or more of the LiF-BeF2

in the fuel carrier, and more than 99% of the LiF in the fertile stream.
DESCRIPTION OF PROCESS

- The'prOCessing facilitj must be capable of removing the major portion
of the fission products from the molten fuel salt and returning the purified
' 233

salt to the fuel system after reconstitution with U and carrier salts,
In blanket processing, the facility must recover bred uranium, minimize
parasitic neutron loss to protactinium, and minimize the loss of carrier
and fertile salts with the waste. | |

In Fig. 1 a flow diagram is presented to show the steps in processing
a molten salt breeder. The flow rates were obtained from physics calcula-
tions for a 2225-Mw (therhal) reference reactor., The core cycle time is
52 days, and the blanket cycle time is about 22 days for the uranium
reGOVery'step;-protactinium removal is on a l-day cycle. The core power

~is 2160 Mw (thermal ); the blanket power is 65 Mw (thermal). A flowsheet,

similar to Fig. 1 and exCludingithefprotactinium removal step, was used in
a design-and cost studyl of a processing plant for a 1000-Mw (electrical)

Fuel Stream Process

| Dedqz'Holdup,

Irradiated fuel is femovéd:difectly fram the Cifculating'fuel stream
for processing, and, as such, is only a few seconds removed from the |

fission zone. The gross heat generation rate is shown in Fig, 2. It is

 

 
 

10

 

" ORNL DWG #7-3834 R

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   
   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

| Makeur  B3ur
F2 TO _ ‘ 7 F» TO _
RECYCLE ’ ' - RECYCLE

4 o 5 _ -
| UFR; + Fp | SORPTION | COLD TRAP “Urg+F,+ FP | SORPTION | COLD TRAP .
- NoF |MgF - | NoF {MgF, -
v =yproDUCT| !
| WASTE ~45 g/DAY WASTE
NoF+MgFo| PERREACTOR | | [NoFsMgR+FP e [Pl |
DISCARD| |MAKEUP
| | | LiF+BeF, !
~630 FTYDAY/REACTOR  4.3FTYDAVREACTOR DISTLLATE | *
FLUORINATION |  Pa BLANKET. DECAY|[FLUORINATION]  [VACUUM] [URe—~UR -
‘ REMOVAL| _ | CooL | STt REDUCTION | <
> ~600°C '
1-2_1500-550°C ~1000°C | Hp— -
. DAYS ~§ mm Hg
g &J LiF-ThF, OR T | o |
| | LiF-Bef-Thi 2 UFe
v - LiF-BeFa-Uf;- FP o
WASTE NO.REACTORS/STATION = 4. | | FISSION PRODUCT
30-YEAR POWER/REACTOR 1 L AcCuMuLATOR
DISCARD FUEL VOL./REACTOR & 225FTS - ~400FT3 | -
OF SALT BLANKET VOL./REACTOR & 627FT® ~6 Mw
U INVENTORY/REACTOR & 190kg LiF-BeF, -UF, RECYCLE FUEL

Fig. 1. Processing Diagram for a Molten Salt Breeder Reactor. B

’\\
/
o W
 

n;“:T o

F'.

38

11

ORNL DWG 63-938iRI

 

 

105

 

104

 

103

 

HEAT GENERATION RATE (btu/hr-ft3)

 

 

 

 

T S0 TN VU T T TIny T T T 1 |Hil

 

10! " :
MINUTES ERNEETIY |
Vo 10 30 - _ : 1

' HOURS :

 

 

 

 

10 20

YEARS

Lot |1|||n| L ;|:t1n| 9 Jlnjln L 1;1|;.n.| 4 Linn

 

 

% 0 ot o e e
' TiME AFTER DISCHARGE FROM REACTOR (dcys)

103

Flg 2. FlSSlon Product Decay Hea’c in MSBR Fuel Strea.m for a

1000 Mw (electrical) Reactor. The curve is calculated for the gross amouht

of fission products in the discharged fuel

T
 

-12 -

necessary to allow fission products to decay a few hours before fluorination
so that the flubrinating temperature can be controlled at the desired value.
This holdup will be carried out in a metering tank in the reactor cell where

there is convenient access to the reactor cooling system.

Contlnuous Fluorlnatlon

 

After the cooling perlod the fuel salt flows into the top of 2 |
columnar vessel where it is contacted by a countercurrent,strsam of fluorine
gas. The temperature of fluorination is controlled at about 550°C. Uranium
tetrafluoride in the irradiated salt reacts{quahtitativeiyswith fluorine to
give volatile UF6, whichris carried ovefhead by excess fluoriné."The'
chemical reaction is 7 -

UFL +F, UFé

Certain fission product fluorides are also volatile and will leave
the fluorinator with the UFg, The principal metallic elements that form
volatile fluorides are ruthenium, niobium, molybdenum, technetium, and
tellurium. Since prefluorination time is so short, appreciable amouhts of
fission product iodine and bromine will also be present., These elements
are oxidized by fluorine to volatile interhalogen Cdmpounds and exit in
the UF6 stream. Zirconium fluoride also has a relatively high vapor
pressure and has been observed in the overhead.product. Sorption techniques

are used to purify the UF6 product.

 

gEé Purification by Sbrption

Volatile fission product fluorldes and UFB are separated in a series
of sorptlon and desorption steps. These are batch steps, but the process
is made contlnuous by using parallel beds alternately. :

The first separation is made when the gas passes through a bed of
pelletized NaF. The system consists of two distinct zones, one held at
LO00°C and one at about 100°C. In the higher-temperature zone, most of the
fission products, corrosion products, and entrained salt are irreversibly
removed while the UF6 and some fission products pass through to the lower-
~ temperature zone. In this zone,'UF6 and MbFB are sorbed, - The barren
~ fluorine carrier passes through the next bed of pelletized.Mng, which is

O,
*

)

i

o

- 13 -

believed to be effective for sorbing technetium; however, available data
are not altogether conclusivé. The volatile halogens are expected to pass

through the sorbers and remain in the récycle fluorine stream. These are

 controlled by decay and by a small gas purge to the off-gas disposal system.

About 104 of the circulating fluorine will have to be discarded to purge
these fission products. |

When the low-temperature zone of the NaF bed is loaded with UF, the
bed is taken off stream, and the temperature of the cold zone is slowly
_raised.while the bed is'swepf with fluorine gas. At temperatures around
150°C, molybdenum fluoride desorbs and is carried away, thereby separating
it from UF6' The temperature is raised higher to about 40O°C for complete
desorption of uranium, which is collected for recycle to the reactor, The
NaF and MgF beds are reused until loaded with retained fission products;
at this p01nt they are dlscharged to waste and filled with fresh material,
The operation of the sorption system is diagrammed in Fig. 3.

The rate of heat generation by fission products deposited on the
sorbers might be as large as 30%'bffthe gross rate shown in Fig. 2. There-

fore, the beds will require cooling to prevent local overheating.

226 Collection - - B | |
Desorbed UF, is carried by fluorine into a primary cold trap held at
about -40°C where it is collected for the fuel reconstltutlon step."The

-1,0°C trap is backed up by a. colder (-60°C) trap to catch any UF, that
mlght pass through the first trap. For additional safety a NaF chemical

7 trap is 1ncluded to trap any UFS that might get through the cold traps.
"~ When a cold trap is loaded'w1th UF6, the trap is warmed to triple point
. conditions (90°C and L6 p51a) to melt UF and allow it to drain to a

~receiver for feeding the redupt;on unit.

 Vacuum Distillation

. After fluorinatidn, fhebarren carrier, containing‘the bulk of the
fission products; flows to a still which is operated at about 1 mm Hg
pressure and 1000°C. The LiF and BeF, volatilize, leaving fission products

in the still bottoms. This residue consists largely of rare earth fluorides,

 
 

F, AND Ry, Te, |, B, Mo) |
FLUORIDES TO GAS ,
' SCRUBBER AND/OR - ORNL DWG 65-3015 #4
RECYCLE |

RECYCLE F,

 
 
  
     

      

   

    
 

 

 
 

 

 

 

 

F,, UF,, TcF
. r ¢ & | COLD TRAP,
100°C DWRING . -60°C
SORPTION, NaF PRIMARY COLD
400°C DURING ABSORBER TRAP, -40°C
DESORFTION
400°C
FLUORINATOR > v
OFF-GAS OR F _
WASTE REDUCTION UNIT
(Tc)
NafF WASTE

(Nb,Zr)

Fig. 3. UF6 Purification System with Disposition of Volatile
Fission Products, . |
.t ( 3y

»

-

)

- 15 -

" yhich are the principal neutron poisons. Available data indicate that the

relative volatilities of thefrare earth fluorides, compared to LiF-and'Ber,,

“are low and that a good separation can be achieved in a single-step distilla-
 tion without rectifieatibn} The folloW1ng data (Table 1) have been obtained

~ for several fluorlde cempounds.

‘Table 1.- Relative'Vblatilities_of Rare Earth Fluorides,
ZrF) , and fieF2,With.Respeet to LiF

Tefiperature = 1000°C
_ Pressure .= 1.5 mm Hg

 

 

 

Component ~ ~  Mole Fraetion‘in,Liquid . Relative Volatility
7 .
Zrf) S 0.0096 ok
PrF, . 0.085 | 2.5 x 107
NdF, : o 0,06 - | - <3x 107
S o 0.05‘ | - <3 x 107

Tt is proposed to operate the still semicontinuously, allowing the

._fiesion products to collect'in”the'still bottoms, the bottoms, in turn,

are sent to the f1531on product accumulator tank (see Fig. 1) for dilution

}?and storage. ‘The accumulator contalns about 400 £t° of 88-12 mole %
"LaF-BeF mlxture,'mhlch 1s the equlllbrlum comp031t10ncfi'the11quid in the

2

© still. The accumulator salt is recycled through the stlll by adding it

into the stlll feed comlng from the fluorlnator. Since the recycle/feed

:=1jrat10 can- have any de81red value, it is adjusted to exerclse control over
"f}the heat generation rate in the st111 as the fission products contlnuously
- concentrate at that p01nt Calculations 1ndlcate that a volume ratlo as

. low as 1:1 is adequate for control of the heat generatlon._..-l o

The hOO ft3 of LiF-BeF2 mlxture 1n the accumulator 1s'be11eved to be:

_sufflelent for.collectlng.flssion products for the entire 30-year lifetime

of the reactor. At the end of this time the mixture can be processed for

 
 

 

 

-16 -

_ i} —
recovery of LiF and BeF, and transfer of fissionhproducts to,some inexpensive \sJ

- medium for permanent dlgposal.

Although CsF and RbF have higher vapor pressures than LiF and'would.be
expected to contaminate the recycle carrier, no difficulty is expected with
-these fission products because each has & gaseous precursor which is removed
on a very fast cycle in the gas sparging operation. Another fission product
that has a relatively high vapor pressure is ZrFL, and it might be necessary
to discard a small fraction of the carrier distillate to purge this poison.
There is some evidence that the activity of ZrFL is low in the still bottoms;
thus, most of the ZrFL'wlll remaln in these bottoms. Even if all the erh
distilled, no more than 5% of the distillate would have to be ‘discarded to
control its poisoning effect. , S ,

Although fluorination is expected to recover greater than 99 9% of
the uranium in the fuel stream, any small fraction that reaches the vacuum
still is not entirely lost. The vapor pressure of UFh at 1000°C is slightly
greater than that of LiF allowing part of this uranium to accompany the

carrier salt.

Fuel Reconstitution

 

The principal operation in reconstituting the fuel is to recombine
the distillate, the recycle and makeup uranium, and the makeup carrier
salt in the proper proportions for feed to the reactor. Uranium hexa-
fluoride from the cold traps is mixed with a portion of the,bredrUFé from
the blanket process, and the mixture is dissolved at 550 to 600°C in
recycle fuel salt. The mixture is kept at the proper UEh concentration
by introducing LiF-BeF2
into the mixture to reduce UF6 to UFh" The entire operation can_be carried
out continuously in & single unit.

- A filter after the reduction unit will be used to clarlfy the fuel
salt before it goes to the reactor. Reducing conditions in the columnfiW111

distillate and makeup. Hydrogen is then introduced

precipitate any colloidal nickel or iron, which will contaminate the:salt

(as a result of equipment corrosion).
.'“:j mj_

¥

")

#}

-17 -

- Xenon Removal -

Because of the low solublllty of the rare gases in molten fluoride
salts, xenon can be strlpped from the fuel salt by sparglng with an 1nert
gas such as hellum or nltrogen.' The sparglng is done in an in-line gas

13 A very short

strlpplng unit located between the pump and the reactor,
cycle time of the order of 30 sec is required to keep xenon p01son1ng at
the des1red 1ow level whlch is 1ess than 0.5% fractional poisoning. The
1nert carrler gas tranSports the fission gases to charcoal adsorbers where
they are retalned for decay,_the carrler gas is recycled ' Because of the
very short cycle tlme, this proce551ng must be accomp11Shed with equipment
in the reactor prlmary system and is con31dered to be a part of the reactor
system, rather than a part of the fuel proce551ng sy stem. ’

An alternative method-of-ellmlnatlng xenon is to processsthe reactor

.fluid for removal of iodine, the precurSOr of xenon. A method for removing

iodine has not been developed- however, laboratory‘experlments have g1ven

promising results..l The method is to sparge the fuel salt with an HF- H

-mixture, allowing the iodine ion (the form present under reactor condltlons)

to react with HF forming HI, which is removed as a gas, Considering the
fact that 1351 has a half—life;Of 6.7 hr, it is seen that the processing

rate must be fast to make the process effective for removing the Xxenon

f 35xe that can be purged

daughter. There is a lower limit to-the amount o
135Xe iS

by removing the iodine precursor,_the independent fission yield of

- quite 31gn1f1cant (removzng 35I accounts for only about 80% of the xenon

vp01son1ng)

“”TFértile“Stream-Process

The fertlle (or blanket) stream process has three prlmary operations:

"1{protact1n1um removal, 'contlnuous fluorination, and UFB sorptlon,_ The
a,assumptlon 1s made that the uranlum is not removed.W1th the protactlnlum,
a:so the latter two operations are analogous to the correspondlng fuel stream

Mb-qperatlons dlscussed above but are carrled out at a hlgher volumetrlc rate.

"The fertile stream is: cycled through fluorlnation once every 20 to 25 days

to keep a low uranium concentratlon in the blanket, thereby keeping the

 
 

 

 

. 18 -

fission rate low. The low fission rate ensures a low fission product
accumulation rate so that it is unnecessary to remove them on“fhe'same
cycle as uranium, In fact, a '3i0-year ,:disoa'rd cycle of the barren fertile
stream is a2 sufficient purge rateefor fission products. At the end of
this time the waste tank would be processed to recover ThFh, the flSSlon
products would be permanently stored in an inexpensive medium.

If the ‘process for removing protactinlum.from the fertlle stream also
removes all the uranlum, fluorlnatlon of the fertile stream will not be
necessaryu A fluorination step 1s, however, requlred in the sflbsequent
process1ng of the protactinium to separate the uranium. |

The bred.uranlum is recovered in cold traps as UF6. A portlon of |
this is used to refuel the core, and the remainder is sold.

Protactinium Removal

 

~ The breeding performance of the reactor can be significantly improved
if the protactinium concentration is kept low. The proposed process is to
contact the fertile salt at about 650°C on & rather fast cycle, perhaps as
fast as once every 10 to 60 hr, with a stream of liquid'metalroarrier; €.,
bismuth, containing 3000 to 4000 ppm thorium. The thorium metal - ‘reduces
protactinium and the two nuclides exchange between the two streams. A
hydrofluorination or sorption treatment is then used to extract protactinium
from the bismuth stream into a second salt mixture, in which it is allowed to
~ decay to uranium. Later the salt is fluorinated to recover the uranium.
Preliminary data indicate that 96% of the protactinium can be removed
from the blanket by this method. The behavior of uranium has not yet been
- determined. Thorium, of course, is the ideal reductant beoause the removed

protactinium is replaced by eqnlvalent fertile material.
off-Gas Treatment

Mbst of the off-gas from the processing plant comes from ‘the continuous
fluorlnators, smaller amounts are formed in various other process vessels.
The gases are treated to prevent the release of any contained fission
products to the atmosphere. Excess fluorine used in the fluorinators is-
»)

»

w)

#)

- 19'_

recycled through a surge chamber, and a small side-stream of the recycling
fluorine is sent through a caustic scrubber to prevent gross buildup of
fission products., Each of the processing vessels and holdup tanks has off-
gas lines that 1eadrto'the scrubber for removing HF, F2, and volatile
contaminants. o _ . |

The scrubber operates as a contlnuous countercurrent, packed bed unit

“with re01rculat1ng aqueous KOH. A small side-stream of KOH solution is

sent to waste, and the scrubber off-gas is contacted with steam to hydrolize

fission product tellurium. A filter removes the hydrolized product. Non-

condensable gases are sent to the facility that treats gases generated by

the reactor, ultlmately the gases are dlscharged to the atmosphere after

con31derable dllutlon.
“Waste Storage

Waste streams'from thetproceSSing facility are:"(l) aqueous waste from

| the KOH scrubber for purifying recycle fluorine, (2) molten salt from fertile

stream discard, (3) NaF and MgF sorbent from UF6 purlflcatlon, (4) daughter
products of Kr and Xe that accumulate as solids in the gas sparging system of
the reactor, and (5), if needed, molten salt discard of L:LF--BeF2 carrier to
purge Zth; Thermaterial’in.the fission product accumilator (see Fig. 1) is
not considered a waste because it is not removed from the process even

though it contains fission products.=,Eventua11y a portion of this volume

- .becomes waste. at the end of the reactor lifetime and after proce851ng to

sreclalm as much as p0831b1e of the LlF-BeF content. Each of the waste

2

 streams would be stored in: underground facllltles at the reactor site; after
-_”approprlate decay perlods and waste treatment procedures, these could be .

'_sent to permanent storage or disposal s1tes,_-

-]Agueous'Waste :

The aqneouS'waste stream is small and Would.be comblned'w1th similar

twastes from the reactor 1n a s1ngle underground faclllty.-

 
 

- 20 -

Fertile Stream Discard

 

This waste is stored in an underground tenk which can be cooled by
air circulation in natural convection. The specific heat genera‘cion"rate
~1s low because of the small concentration of fission products presént. Tt
may be desirable to store the fertile stream discard inside the processing
cell. The salt volume is only about 2500 £t , and, if inside the cell,
cooling could be conveniently prov1ded by the "same medlum that is used in
the processn.ng operatlons. B

 

NaF and MgF Pellets

These sorbents, which contain the volatile flss.mn product fluorldes ’
are stored in cylindrical cans in an underground concrete vault.. The spent
pellets are discharged from the sorbers into cans inside the process:.ng
cell. When several cans have been-i‘llled,‘,t.he contents are moved from the
¢ell to the vault. Decay heat is removed by forced air circulation through
the vault. |

Kr and Xe Daughter Products

 

Krypton and xenon, sparged from the reactor,' are precursors of solid
fission products (Rb, Sr, Y, Zr, Cs, Ba, La, Ce) that settle as fines and
dust throughout the gas processing system. These particulates must be ' .
removed to afi'oid clogging and overheating gas lines. A possible method for
doing this is to periodically flush lines and gas holders with a molten
salt, which dissolves and accumulates the decay products. The accumulator
could be 2 vessel (in the processing cell) in which these fission products
are retained until permanent disposal.

STATUS OF PROCESSING TECHNOLOGY AND NEEDED DEVELCPMENT
Fluoride Volatility Process
Experience at (RNL with the processing of molten salt fuels by the

fluoride volatility process dates from 1954 and included all phases of
laboratory and development work through the successful operation of a
 

13

)

"

- )

- 21 -

- pilot plant. Durlng this”pe'riod the processfwa's‘ adapted for use With the

fluid fuel, NaF- Zth_UFh » of the Aircraft Reaotor Exper:.ment , and two alloy

fuels, Zr-U and Al-TU. Although the alloy fuels were SOlld they were

‘dissolved in the molte_n fluonde salt and their processing was carried out

by methods analogous. to: processing a molten' fluoride fuel. After the initial

dissolution step for the solid element processing of llquld and solid fuels

" becomes identical within the llmlts 1mposed by different compositions.

The fluoride volatn.llty process takes ‘its name from the principal
operation, the volat lllZl.Ilg. of uranium as the hexafluorlde. The advantages

of fluoride lvolatility are: the small volume of fission product waste, the

convem.ent form of the UF6 product which is easily converted to UFLL for

fuel recycle , the. v:Lrtual ellmnatlon of prooess:.ng orltloallty hazards as

opposed to aqueous process:.ng of enrlched fuels, and the high decontamina-

tion factors atta:.nable for the UF6 produot Unattractive features of the

process are: the corrosiveness of fluorlne_-molten{ salt mixtures, and, to a

lesser degree, the high processing temperatures._ A diagram of the ORNL
Fluoride Volatility Pilot Plant is shown in Fig. L.

In addition to the molten' fluoride prooessing*work at ORNL, similar
research has been carrled out at Argonne ‘National Laboratory. Efforts at
the two installations were complementary, with the ANL work being directed
prlmarlly toward hydrofluorination and i_‘_:l_uorlnatlon techniques and reagents

~as well as equipment“ development., In addition,' molten salt process develop-
~ment was supplemented by an extens:l.ve oorros:.on-testlng progran under sub-
'contract W:Lth Battelle Memorlal Instltute (l959-l963) Evaluatlons Were

| .made of potential constructlon mater:.als for the hydrofluorlnatlon and

fluorlnation equlpment'-the metals of most J.nterest were found to be

_':;-.I-Iastelloy N, Alloy 79-)4 (known also as HyMu 80 and" Moly Pennaloy) , and
"'Nlokel 201 (L-N:Lckel) A1) these metals are satlsfactory construotlon "
o materlals for molten salt process:Lng equlpment with the poss1ble exceptlon
":of the vaouum stlll. _ The 1000°C d:n.stlllatlon temperature is beyond the

o 5""range at Whlch materlals have been evaluated.

The proposed process for the MSER inoludes several operatlons in

| add:.tion to those w:Lth whloh there has been reprocessmg exper:r.enoe. The
-'most important of these are°‘ contlnuous fluorlnatlon, vacuum distillation,

reduction of UF6 to UF s and removal of protactinium from the fertile

 
 

  

BARREN-SALT
TRANSFER TANK

 

 

  

 

: /—mn. ELEMENT

. CARRIER -CHARGER

Nof Ve
CHARGE LINE
g 2y

 
  
      
       
  

CONML=LR-DWG 45235¢R0AN

B i ) s

 

  

 

FLUDRINE
SCRI:PUEEN '

OFF GAS . l

 

 

 

 

 

. uF
NEUTRALIZER

  
 
   

MF SUPPLY '—-N“‘

HF CODLER  {

   

 

 

 

 

   

- — ——

 

 

 

 

 

-

X wr vapoR GENERATOR

 

TO0
ORAIN

 

Fig. L. Schematic Diagram of ORNL Fluoride Volatility Pilot Plant

_pewote HESD PUMPS

MOVASLE-8ED
ABSORBER

 

 
    
  
 

 

 

 

- RECEWER

SMALL PRODUCT

UF, COLD TRaP

 

b 4
UF. coLn fRAP - j

 
  
   

 

 

 

 

 

 

 

 

  
    

 

 

RECEIVER

 

————————
e
PRODUCT : . ‘
_

NORMAL FLOW LEGEND

U

HF

F

MOLTEN SALT
SERVICE, ETC.

¢é

 
 

 

 

b5

"y

I

L]

- 23 =~

stream. The folIOW1ng paragraphs review the experlence with ARE, Zr-U,

"and n-u fuel processing and delineate the status of exlstlng technology

for the operatlons involved in MSER processing.

ARE Fuel Processing B

 

Initial development of a process for molten fluoride fuel was motivated
by the need to recover unburned hlghly enrlched uranlum.from the irradiated
fuel of the Alrcraft Reactor Experlment. This fuel was a 48-49.5-2.5 mole %
mixture of NaF-ZrEh-UFh | Beglnnlng in 195} the fluoride volatility process

 was developed for this recovery and culminated in successfully processing
the ARE fuel during the period December 1957 to March 1958. Since the

reactor operated in November l95h, the 1rrad1ated fuel was about three years
old. The recovery operations con51sted of batchw1se UF6 volatilization,
followed by its decontamlnatlon on granular beds of sodium fluoride, and 1ts
recovery in cold traps.g Barren Waste,"contalnlng the bulk of the fission

products, was drained into metal cans and buried. Better than 99% recovery

of uranium was attalned 1n thls 1n1t1a1 hot operatlon of the fluoride

L

volatlllty,pllot plant Decontamlnatlon factors of at least 10" were

attained, as indicated by the" gross gamms. activity of - the_product.9

Zr-U and Al-U Fuels

 

In the initial stages of fluoride volatility development it was
recognlzed that the process had appllcatlon to fuels other than molten
fluorlde salts. Concurrently"W1th the development of a process for ARE

'p,'fuel 1aboratory research was pursued toward adaptlng the process for
e treatlng 1rrad1ated ercaloy-clad Zr-U alloy fuel from submarlne reactors.
N,f“Thls research resulted 1n the development of a hydrofluorlnatlon operat1on

in whlch the solid fuel element 1s dlssolved by reactlon with anhydrous HF

,)1n molten LlF-NaF Zth, the remalning steps of the process were the same as
_ffor ARE fuel LT o | o

Proces31ng experlence led to improvements in the pllot plant and

equlpment before operatlon'w1th Zr-U fuel In addltlon to ‘the hydro-
" fluorinator, an improved fluorlnator (Fig. 5) and NaF sorber were incor-

porated in the modified plant, The plant was operated with irradiated

 

 
 

 

 

 

 

 

2

€
./,»—\ .
- .

. ORNL-LR-DWG 39150-R2

_PRODUCT OUTLET TO
S MOVABLE BED ABSORBER

     
  

{I—— — PRODUGT OUTLET
SAMPLE LINE- {5~ (ALTERNATE) .

* (+in NPS SCH. 40)

" TUBULAR ELECTRIC
HEATERS

 
    
     
     
 
   

APPED NOZZLE

 

 

 

 

 

 

 

 

 

 

 

) ' THERMOCOUPLE ’
{ WELL =
FURNACE LINER—|___ [
)
TREPANNED SECTION iy AT ALl
{4-in. OD) sl O ) R - o
gpal f s .| ~FLUORINATION CHAMBER
CRE . B Mk  (16-in.0D)
b ‘ !I ,‘1 E - _ W . .
DRAFT TUBE \\gh_*; Wy Es |
pgt- r‘ | % ,‘g'!‘;
SIS 740 =R -,
WASTE-SALT OUTLET—{IE. %4 AN 4 bi2)|-[— FLUORINE INLET
o - 0 ARy " g.\\—-——TREPANNEDSECTION |
[ MATERIAL | s . (4-in.0D)
_%-in. NICKEL 201 ‘ 0 5 . 10 B 20
R j ~ INGHES
' FURNACE
. Fig. 5. Volatility Pilot Plant Fluorinator Vessel |
 

 

- 28

13}

“’;_ submarine_reactor fnels during the periodIMarch 1962 to September 1963.
Fuel'elements that had cooled*as,long'as 6.5 years and as short as

=)

6 months were succeSSfuIIY'proceSSed.' Uranium recovery and decontamina-
“tion in all runs were excellent,
The flnal phase of operatlon of the ORNL Vblatlllty Pilot Plant was
assoc1ated.W1th proce531ng alumlnum-clad ‘Al-U. alloy fuel. 10. It was found
“that the solid elements eould~be-satlsfactorily dissolved by reaction'with
arhydrous HF in molten 66-22-12 mole %'KF-ZthquFB.
the process were the same as for Zr-U alloy processing, and the same pilot

Other operations in

plant equipment was used. A number of pilot plant runs were made with
- irradiated Al-U fuel during the period July 196L to January 1965 at which
© time the Volatility Pilot Plant was shut down. During that period, Al-U
' fuel cooled for only four weeks was satisfactorily processed. The same

exoellent recovery and decontamination factors were obtained as with ARE
and Zr-U fuels; uranium recoveries of 99, 9% or better, and decontamination

factors of 106 were typlcal

»

. - . .. Continuous Fluorination

7 Contlnuous fluorlnatlon is des1rable for pr006551ng both the core and
' blanket of an MSHR because of the need to mlnlmlze the out-of -reactor
1nventory and the costs of proce381ng. Engineering development is compli-
cated by the corrosiveness of the molten-salt—fluorine mixture and by the
dlffloulty -of reproduolng in the laboratory the high 1nternal heat
generatlon rate of a recently 1rrad1ated fuel.

| The technOlogy for batchw1se fluorlnatlon has been rather thoroughly

e developed and applled in the Fluorlde Vblatlllty Pllot Plant but there is
llttle experlence with contlnuous fluorlnatlon. One technlque of continuous
fluorlnatlon 1nvolves spraylng or Jettlng small droplets of molten salt
through a fluorlne atmosphere in suoh a way that ‘they do not contact the

7 metal walls of the fluorlnator.z Some very prellmlnary tests w1th 51ngle
\fi) : salt droplets have been suocessful in achieving high uranium recovery.

€l

 
 

 

- 26 -

A second method, which appears to be the most Promising, is to carry
out the fluorination in a liquid-phase continuous system using.a 1/2- to
3/li-in.-thick layer of frozen salt to protect the vessel walls. The initial
' 3

work on such a system was done at Argonne National Laboratory;”.currently,
the development is being pursued in the Chemical -Technology Division at
ORNL. Preliminary results, employing units that did not havé'frdfienawall
protection, have demonstrated that high uranium recoveries can be effected

in continuous, countercurrent operation.

 

Needed Development

'Engineering development of & continuous fluorinator should be given
high priority. Areas requiring study include: mass transfer information;
process flow control; methbd.of és£ab1ishing and maintaining a frozen-salt
wall; mist deentrainment in gas stream; and, efficiency of’gas¥liquid o
contact, 'A_problem in particular need of attention is the design of nozzles
that admit and_remoie the molten salt and fluorine gas., It-might be
difficult to protect these areas from corrosive attack by establishing a
frozen-salt wall. In the development it should be recognized that there
is a large difference in capacity'between the fuel and fertile stream
fluorinators and that the internal heat generation in the fertile stream
is appfeciébly less than in the fuel. A conceptual design of a continuous

fluorinator is shown in Fig., 6.
UF6 Purification by Sorption

The NaF and M’gF2 sorption units provide adequate decontamination for
UFE. The batchwise units can be operated satisfactorily for both fuel and |
- fertile streams of the MSER. Development on this part of the process is
not critical to successful operation of a molten salt breeder experiment.

Current Status

The UF6 purification system:has been bperated as an integral partiof

the Fluoride Volatility Pilot Plant, producing a product that can be handled
~ directly. Also, it has been demonstrated that uranium recovery from the
sorbers is essentially quantitétife;

O,
 

u)(" "

4

27

ORNL~DWG 65~3037RA

_a—F2,UFfg, F.P. TO SORPTION SYSTEM

DE-ENTRAINMENT SECTION

COOLANT

 

PO

QO QA AN

 

'FROZEN *
WALL OF SALT

| GRAVITY
Y2 TO 3/a in. THICK

LEG——~

hd

G AR S AN o . 0 P

 

 

 

 

 

 

 

 

NaK COOLANT

Flg. 6. Continuous Fludrihatlon. NaK coolant, fIGW1ng through the

'Jacket freezes a layer of salt on the inner surface of the column, thus

protecting the Alloy'79-h from corrosive attack by the molten-salt-fluorine

mlxture.
 

 

 

- 28 -

 

Needed Development

Although no extensive development is needed to ensure successful
operation of the sorption system, there are several areas in which performance
data should be obtained. The sorption characteristics and capacity of MgF
for technetium are not Well known; the methods for removal of tellurium and
ruthenium are also not well known; and the chemistry'of molybdenum in the
sorption system needs better definition. From an éngineering viewpoint, a
sorption system that could be operéted continudusly'wduld have advantages
over a batch system, and such a system should be developed for large scale
'MSER 1nstallat10ns. |

Heat generation in the sorber beds will bé.appreciable because of the
short decay time before fluorination. A thorough evaluation must be made
of this problem {ncluding a satisfactdry cob1ing mechaniSm)"Since the
_sorbents become wastes that are removed from the processing cell for
storage, it is essential that a means of doing this be develqped (one that
precludes frequent entry into the cell ),

A diagram of a succeésfully operated unit is shown in Fig. 7.

Cold Traps

Current Status

- Cold traps, like the one shown in Fig. 8, have been used at ORNL and
ORGDP for the collection of UF6. The unit has to be operated batchwise,
but this is no deterrent to its use in a breeder experiment or a full-scale

installation.

Needed Development

 

No development is needed on this phase of the process on'a:figar-térm
basis; however, ultimately it will be desirable to have a continuous cold

trap system.
")

T

29

ORNL-LR-DWG 5045iR-3

YNOF CHARGING CHUTE

l‘/z-m NPS, SCHED~40 { AIR COOL~

   
  
 
   
 
 
 
  

A
- : \ LT ING AND THERMOCOUPLES)
TO UF;- TRAPS VBN
DESORPTION CYCLE N
5-in. NPS, SCHED-40 INCONEL
100°0R 400°C.

TRANSITION
ZONE

INCONEL-X _PISTON 7

S-m. NPS, SGHED 80 INCONEL j’

  

HYDRAUL#G cvu NDER

 

i . Movable-Bed Temp:e:z'-'é.ture -Zoned Abéorber.
" of the bed becomes loaded with fission products, the hydraulic cylinder
operates the piston to discharge that portion of the bed into the waste

Fn.g. Ts

 carrier. Fresh NaF is added at the top.
tested in the CRNL pilot plax_:t._ -

Whén 'thé lower zone

This apparatus has already been
 

 

30

QORNL-LR-DWG !9091 R-l

N OUTLET END
- HEATER

    
    
 

- FILTER
. . GARTRIDGES

REFRIGERANT TUBES (4)

» 0. 12 24

SCALE (in))

CALROD HEATERS (6)
INLET END HEATER

Fig. 8. Cold Trap for UFg Collection. This d,esig.n- has aiready been

successfully used in the ORNL pilot plant.

L Q.
"C’

h

nsemlcontlnuous operatlon of the still as dlagrammed in Flg. 1 is the most
_ 'satlsfaotory method.foraocumulatlng and storlng the flss10n products. The
- general problems and.technology of operatlng a piece of equipment at 11000°C

- 31 -
Vacuum Distillation

The vacuum distillation concept for separating LiFnBeF2 fuel carrier
from fission products is feasible from an engineering viewpoint. However,
considerable development is needed to perfect the operation for a breeder
experiment, and the most significant advancement in molten salt processing

can be made by developing vacuum distillation.

 

Current Status

Vacuum distillation has been seriously considered for purifying the
carrier since laboratory tc-‘ss‘l:s"L showed that decontamination factors for_
rare earths in.the range 100 to 1000 could be achieved. These bench-scale
tests have been continued to-develqp basic data for the relative volatility

(see Table 1) of rare earth fluorides as well as operating technology at

 

high temperature and low pressure. Also underway is the fabrication of a

large-scale distillation experiment that will be an extensive demonstration

- of the operation employing nonradioactive mixtures; the last part of the

experiment will be a demonstration run employing radiocactive spent fuel

from the MSRE. The equipment for this experiment is shown in Fig. 9.

Needed Development

 

Research has only recently-begun to explore vacuum distillation of

molten salts, and a hlgh prlorlty should be given to this critical step in

) the - proce331ng scheme. Although a con51derable amount of phys1cal and
chemical data for this system has been measured, more data, especially
:relatlve volatlllty, ect1v1tygdand distillation rate measurements, are

.. needed before a full- scale plant unlt can be designed. It appears that

and 1 mm Hg need to be determlned S0 thet thelr influence on design can be
recognlzed. Extensive, nonradloactlve testlng of the MSRE experimental

still should answer some of these questlons.v

 
ORNL DWE 66-10982M

 

&

VENT
LEVEL STILL VENT
\__.{E'__C_%!T_RPE ______ Vol.= 12 liters '
“““““““ 3 Pulmm H
] T wi000°®

 

 

ARGON INLET

 

 

 

  

 

   
 

COLD TRAP

 

 

 

':..!'.€&ll_‘.'l3' e Ty L e ORI
P SRR NN, 1
I

2€ -

T DECONTAMINATED
- LIF - BeFp - ZrFy

    

QQETAM"‘JATEQ FEEQ‘_’G"‘fl
LIF = 65.6 mois %
BeFp 429.4 © o
sz4 @ 30 ¢ »
FiISSION PRODUCTS

 

 

   
  
  

b
canfet

 

 

 

 

SAMPLER

 

 

 

 

.

 

P

 

 

 

 

 

 

 

 

e ¥ pr
RSy 3 —
FEED TANK E":%-_’
e el - Vol.= 48 titers
F % ' - Pm0.5mmHg
a500°C

'Fig. 9. Vacuum Still for MSRE Distillation Experiment. Contaminated
feed is fed continuously to the annular space in the still by pressurizing
the feed tank with argon. The feed rate is made equal to the vaporization
rate so that the still level remains constant. The course of the distilla-
tion is followed by periodic sampling of the condensate. The bulk of the -
fission products remains in the still residue. S ‘

.C.

 
..._33-

Even though the*accumnlafed fission products generate a considefable
~amount of heat; the use of the accumulator vessel lowers the specific heat
'generation_rate_to a value thatiintroduces no difficulty in designing the
heat removal system. The calculated magnitude of this heat generation rate
for a conceptual still design for a 1000-Mw (electrical) MSER is shown in
Fig. 10. | - |

~ The 1000°C operating_temperature'of the vacuum still introduces
metallurgical problems that have not been ehcbuntered.previously in molten
salt processing. Equipment for the MSRE distillation experiment is made
of Hastelldy.N; howéver, it ié_belieVed that this material is not the best
for_theserconditions and that development of a suitable material should be
sfiarted. Molybdenum or a high_molybdenum alloy are likely candidates and
should be evaluated. Not only must the material be compatible with molten
fluoride salt but also with the coolant, which will probably be Nak or
another molten salt, and the surrounding atmosphere. Duplex or clad material

may be applicable.
Fuel Reconstitution

The reduétion of UF6 to fiEh and recombination of the UFh with LiF--BeF2
carrier requires engineering development. This work should have a high
priority, ranking in importancé_with continuous fluorination and vacuum
distillation. B
© The reduction of UF, by hydrogen in an H,-F, flame is a well known

-reaction; the reduction is rapid and quantitative. However, the powdery
| UF), product is difficult to handle remotely. Liquid-phase reduction in
which UFé'is reacted.with H2 in;moltgn salt is more suitable for continuous,
ftrduble-free operation. A limited amount 'ofpwork5 has been done on liquid-
_phase_feducticn; and.thejdata“indicatethat the method:is highiy stitab1e
:tdthis_application. R | '
o

_ ORNL DWG 67-5268
10

 

HEAT GENERATION RATE (Mw)
o

 

 

 

 

10" | , 1
BT | o o wt - 10°
OPERATING TIME OF VACUUM STILL {days) |

‘Fig. 10. Heat Generation Rate in the Vacuum Still Residue Due to
Fission Product Buildup. The integrated heat generation rate of decaying
fission products in the still residue makes it necessary to accumulate
these nuclides in a2 relatively large volume in order to lower the specific
heat generation rate. The above curve is for gross amounts of fission
products from processing 15 ft /day of fuel from a 1000 Mw (electrical)
reactor system. Elapsed time between dlscharge from the reactor and entry
to the still is 2.58 days. '
O

N

-

- 35 -

Needed Development

 

Little is known about the ‘kinetics of the absorption and reduction of
UF6 in the molten fluorlde mixture other than that it is fast. The mechanlsn
may be reaction of UF6 with UFL to form an intermediate fluoride, followed by
reduction with H An,experxmental progrem.ls needed to study reaction
rates, temperature, UFh concentratlon, nozzle ‘design, contactor de31gn, and
gas-llquld separatlon. There mlght be a corrosion problem in the reduction
unit, and considerable attentlon'W1ll be needed in the de51gncfi‘theun1t to

minimize this effect.

Fuel Clarification
Reducing conditions'that;exist in the UF, reduction unit are conducive
to the formation of metallic species, particularly nickel and iron, which
fiay be in a finely divided or colloidal state. Before returning to the
reactor, the fuel will need olarification. Development of a solid-liquid
separation system for the breeder experiment is not a critical item; |

however, it is a problem that,fiill have to be sclved eventually.

Current Statue

Final clarification of -the fuel charge for the MSRE was accomplished

by filtration, which was carried out after an H2-HF treatment to reduce

oxides. Sintered nickel filters were used in a batch operation, amd after

| use the- fllters were dlscarded., This procedure‘was very effectlve in

clarlfylng the molten salt

1 'Needed Development

8o far, flltratlons have been carrled out on nonlrradlated salts, and

l_one dld not have to cope With heat generatlon from flSSlon product decay on
:Vthe fllters. Thls could be a real problem in a system that recycles fuel.

V'Present fllters probably do not have an adequate operatlng llfetlme in this
-type of serv1ce. “No technlques for handllng or backwashlng the filter cake

have been'worked out. A contlnuous clarification method would be highly

desirable to avoid frequentrentrles into the cell to change filters.
 

 

36 -
Protactinium Removal

Protactlnlum removal 1s an J_mportant operatlon in fertlle stream -
_processrpg and should be v1gorously'pursued The benefits are prlmarlly
in nrlprovement in the breeding ratio by decreasing pa.ras:.t:.c capturee) and
; reductlon in the thorium 1nventory. In order to be effectlve, the process-
" ing rate for protactlnlum‘removal must be 51gn1flcantly'faster than 1ts
natural decay rate (half-llfe = 27, L da.ys), this requlrement means a hlgh

volumetric processing rate for the fertile salt

Current Status

The work on protactinium removal has been coni'lned to laboratory

- development which has given some promising prellmlnary results, ' The
status of this work is discussed in the reportlz on the chemical develop-
ment program, CORNL-TM-1853,

Needed Development

 

Laboratory inve stilgations into protactinium removal are imderway , but
no engineering development has been started. As soon as the laboratory
work has reached the point at which engineering studies can be useful, the
studies should be started. They will involve the complete engineering
development of process and equipment. A primary need for the engineering
is an alpha-handling facility in which to do the work. Building such a
facility should begin soon in order to be ready when needed. - En'gin_eerihg
development must be oriented toward a2 simple and easily operated process
because of the high volumetric throughput required to malntaln a low

protactinium concentration.
Waste Handling and Disposal

Waste handllng for an MSBR poses some problems that have not been
encountered before in proce ssing operatlons.  These are brought about
prmarlly by having considerable amounts of fission products appear at

several po:.nts in the Process, . for example , in still bottoms » in NaF and

. Q).
¥ C 3

v

L

-t

- 37 -

MgF sorbents, and.throughOut"the'gas sparging system., In-process =

,collectlon and handling of these wastes are areas that need study and

development. On the other hand, waste disposal is not a critical problem
in the development of a molten salt breeder experiment. Bulk storage
procedures like those used for aqueous wastes can be used for the near '
term while research and development effort is applied in more-critical

areas. -

Current Status

The general toplc of Waste dlsposal has been the subJect of con51der-
able study at C(RNL for several years. Although molten fluoride wastes
were not a specific part of the study, a large amount of the developed

data and knowledge are appl1cable in this case. For example, the cost

data for permanent dlsposal and heat dlss1pat10n rate in underground

- storage facilities should‘be very good estimates for the fluoride salt

case, In addltlon there is considerable experience in handling .and
treating hot aqueous wastes in pilot plant operatlons at Hanford and the
Idaho Chemical Process1ng Plant to make glasses and calcined products.

. Needed'Development

 

A study isrbeingimaderof waste disposal for molten fluoride salts
to determine the most satisfactory progrem to follow. The fission product
accumulator, which is integrated directly into the vacuum‘distillation

_'step, is in effect a Waste tank'that stores ‘the bulk of the MSBR waste
~for 1ts 30-year llfetlme. At the end of life and follow1ng an 1nter1m

coollng period, this waste could be dlsposed of by'procedures current at

- that time. - Another hlghly radloactlve waste is the NaF ‘and MgF, sorbent
__ifrom purlfylng the UF6.' Since thls mater1a1 is already in solid, granular
| t_tform, it 1s probably sultable for permanent dlsposal 1f sealed in the

'_proper contalner. An 1nterlm coollng perlod would of course be necessary.

- In the reactor off-gas treatlng system, however, flSSlon products

are not conveniently collected in a meterlal that subsequently acts as a
'retentlon;medlum, 1nstead decaylng krypton and xenon dep051t solid daughter

products, perhaps as flnes and dust, throughout the off-gas system. The
 

'38]‘

problem of keeping the off—gasisystem clean of such products has not been
encountered in fuel processing, and considerable development is needed to

perfect a satisfactory method. -
Process Control

The chemical processing plant will need measufements of radiation,
temperature, pressure, uranium concentration, and flow rate. 'Development
of a flow metering device is of outstandlng importance because of the |
proposed contlnuous operatlons, and the work should be undertaken in the

near future.

Current Status

All of these measurements have been encountered in the development

program for the MSRE and much satisfactory instrumentation has been developed.

In the case of temperature and pressure control, satisfactory means of
control have been developed. However, the problem of controlling the flow
rate of a molten salt stream has not been solved. There:is no curfeht'
method of continuously measuring uranium concentration in a flowingmstream
of molten salt or UF6 concentration in a flowing gas stream. Concentrations
have been determined by sampling and laboratory analysis.

Needed Development

 

The primary need in instrumentation for processing the:MSBR is a flow
control device that can meter the flow of a molten salt stream. The problem
has been recognized as a difficult one and, in the past has been c1rcum-'r
vented by batch operation and liquid level measurements. HOwever, for
smooth operation of a continuous process, such a device Wlll be needed.
The usual method of using flow control valves is not suitable for a molten
salt and has been avoided. N o , |

Considerable emphasis must also be placed on the development of 1n~1ine
instrumentation for concentration measurements, particularly for uranium,.
lithium, and beryllium. Smooth operation of the plant mlght require a fast
analysis of a flowing stream.

.

. ( ) »
 

rt(!

A

- 39 -
DEVELOPMENT OF ALTERNATIVE PROCESSES AND IMPROVEMENTS

“The discussion in the preceding sections has presented the main-line
effort for the,development'of a processing-method for the molten salt

breeder reactor. In this sectlon several alternatlve schemes are described.

' These schemes have promise for 1mprov1ng or substltutlng for the primary

processing method and a concurrent investigation Should'berconducted.
Todine Stripping

Nb‘single fission_product has such a deleterious effect on the breeding
performance 337135xe.',Removal'of its iodine precursor from a side stream is
an indirect means of removihg-xenon; howe#er, since 35Xe has a direct
fission yield, the reductionrindpoison fraction by iodine removal is limited.
Iodine_stripping does not eppear_to have an advantage over direct stripping

of xenon, and ‘development should be subordinate to more important problems.

Current Status
Some 1aboratory studles have been made of the chemlstry of 1od1ne in

molten salts. They 1ncluded experlments in which 1od1ne was removed from

salt by sparglng w1th a stream of HF in hydrogen.

Needed DeVelopment

'_ Laboratory development should be continued to obtain the basic ‘data

for an iodine removal process., Englneerlng support is also needed to

-evaluate proposed processes. The englneerlng effort will 1nclude some

experlmentatlon as well as. process analy51s and calculatlons. Fission

*egases contain a surprlslngly large amount of the decay energy of the system,

and the 1mportance of this fact,needs to be assessed 1n the de31gn of a

\ "f1s31on gas dlsposal system.;f;t*fl L

Use offAdditd#es;ifi_Vecunm Distillation-"prf"'

The success of vacuum dlstlllatlon depends upon the fission product
fluorldes having a low relatlve volatlllty compared to LiF, and extant
 

- 40 -

~data indicate that this is the case with the rare earths (Sge Table 1).
However, there is a possibility that even lower relative VQlatilities can
be achieved through the use of additives that react with the rare earths,
zirconium, and other fission products. The additives could be introduced

 Just prior to the vacuum distillation step and'would remain in the still

residue, being sent to waste when the still is drained.

"This technique could be significantly helpful in improving zirconium
decontamination if the proper complexing additive were found, Zirconium
fluoride is more volatile than LiF so the vacuum still may not be very
effective in separating it from the carrier unless its activity in the
solution is lowered. -Zirconium.poisoning-can-bé economically controlled
by discarding carrier salt, but lowering Zthrvolatility'would decrease

the required amount of carrier discard.

Current Status

No work has beén done.

Needed Development

 

Reliable relative volatility data for the rare earths and zirconium
fluorides are being obtained, and this is the first step in accurately
assessing the effect of additives. Bench-scale development iS'the primary
need. If useful additives are found their effects can be dembnsfirated in

the general development of the distillation process.
Reduction—Coprecipitation and/or Electrolysis

Reductive copreclpltatlon6 of the lanthanide elements with berylllum
has been demonstrated in the laboratory. The reduction product is a
refractory, insoluble beryllide deposited at the salt-metal interface.
~ Metallic lithium can also be used as the reducing agent. __'
Electrochemical reduction instead of liquid metal contacting might
'also be used to control the concentration of such fission products as Mo,

Ru, Sn, and Pd and in removing corrosion product Fe, Ni, and Cr.

N

..
 

m(‘l

nt

Current Status
The liquid metal extractions have been limited to scouting experiments
in the laboratory, however, 81gn1flcant removals were observed for La, Sm,

Gd, &, and Eu u51ng a Li-Bi allqy extractant. Similar results were obtained
with Be instead of Li; however, a large excess of Be was needed to remove Zr.

Needed.Development

 

The data on these extractlon methods are pre11m1nary and insufficient
for Judglng the process., Back extraction of rare earths from llquld metal
carrier has not been tested., More basic laboratory experlments are needed
to more clearly define mechanlsms and products. As for electrolytlc
reduction, the electrochemistry of molten salt systems needs to be explored
as a prerequisite to a soundsevaluation of the method., Basic oxidation-
reduction relationships'betWeen'the metals:andrthe'rare’earth'compOnents'of
the salt phase need to be measured. Extensive ehgineering development will

be required if laboratory experiments indicate that the process is attractive.
Ton Exchange with Nonfluoride Solids

The prooesses inoioded'here“ere ooncerhed.with:the“selecfiive:removal
of rare earth fission products while not requlring any change of state in
the bulk of MSBR core stream. Conceptually, this is a most desirable

feature. Refractory sollds, based on carbides, phosphldes, nltrldes,

' 511101des, and.sulfldes, might be foundnwhich‘W1ll ‘exchange’ cations with
‘the rare ‘earths formlng 1nsolub1e compounds. Oxldes might also ‘be 1ncluded
‘in this class of materlals because of thelr extreme low solubillty in
Tmolten salt. Indeed ‘the oxide method.mlght be. partlcularly'useful for
-_proce381ng a fuel stream- contalnlng thorlum,as proposed 1n the alternative
.reactor concept. In 1ts 51mplest form, the operatlon can be ‘visualized as
~ a percolation of a small 51de stream of core fluld through a column packed

_ W1th the ion exchange materlal o

Current.status

7 with carbides (that

indicate such an exchange does occur)and a few incidental observations

There have been some exploratory experiments
 

 

- 42 -

~ on sulfides (made during the course of other studies). On the other hand;
there is a large body of 1nfomat3.on concern:l.ng the chemistry of oxides in
contact with molten salts. - Many oxides,. 1nc1uding those of fiss:.on products

~ as well as heavy metals, ‘have been studied to establ:.sh thelr solubillty

products in molten salt. . Also the removal of rare earths as ox:.des i‘rom
molten fluoride melts has been researched to assess the potential for a
processing scheme, However, it is not yet posslble to resolve all ‘the
, variebles to permlt a quantlta.tlve statement about the possib:u.lltles for

such 2 proce ss.

‘Needed Development

 

More laboratory development is needed on these ion exchahge 'proeesses
before there can be a full appreciation of their potentlal. Later in the
development, the work will have to be supported by engmeerlng exper:unents ’
- which to a large extent would be co‘n_cerwned_,_mth contactor development

SCHEDULE OF MANPGER AND COST FGR DEVELGPMENT PROGRAM

The cost and manpower commitment necessarj to develop on-site s
integrated processing methods for the fuel and fertile stree.ms of a molten
salt breeder reactor were estimated on the presumption that all cr:Lt:Lcal
questlons required answers by June of 1970. Costs were based on the pre-
dicted costs of work in the Chemical Technology Division (including a
steady 2% increase per year)and the estimated amount of work necessary to
solve each problem as required for comprehensive de sign and opti:mum,
operation. '

- Table 2 itemizes the manpower and costs for each subprogram w:.th totals
draw_n at the end of FY 7h. It is antieipa_ted that some sppport for the
reactor processing development program will be required after FY 7h, but |
that it will be small compared to the effort which brought the progra.m to
that point. . o

Comments on each subprogram, with the a.ntic1pated status at the end of

FY Th,are given below.

).

A
 

 

" Table 2,

WK

of a Molten Salt Breeder Reactor by the Chemical Technology Division

Manpwer and Cost Breakdowvn for the Development of the On-Site Processing for the Fuel and Fertile Salts

 

 

 

 

. ' Total
FY 68 . ) FL 70 T N T FY T3 FY T4 Through FY T4
MY #0007 My $a03 Mr 803 My $a073 Mr #0103 w8077 Wy #1053 M &ao”
Continuous Fluorimetion  Men 2.5 113.0 k0 1840 4.0 = 188.0 3.5 168,0 ‘3.0 147.0 .2.0 - 100.0 2.0 102.0 21.0  1002,0
' Unusual Expenses 25.0 150.0. . 100.0 50.0 : 5%0.0 50.0 50.0 L75.0
Total 128.0 334.0 288.0 218,0 “'197.0 150.0 152.0 . _ 1477.0
Sorption ‘Men - 1.0 3&0-0 . 1.5 7 6"'00 2.0 1@.5 005 2".0 1.5 73'_05 1-0 50-0 1.0 51-0 8-5 l}Ofi,O
thusual Expenses : ' ' - 50.0 50.0
_ Total i 0.0 64.0 12,5 2h.0 123.5 50.0 51.0 455.0
Carrier Salt " Men 2.5 0.0 2.5 112.5 4.0 185.0 6.0 3.0 6.5 328.0 L.0 1950 L0 199.0 29.5  1k38.5
Distiliation husual Ezpenses e 28,0 110.0 150.0 100.0 50.0 . 50.0 50.0 535.0
o - . 'I'o‘t.al - 135.0 202.5 333.0 .o 378.0 2k5.0 249.0 1973.5
MSHE Experiment = Men , L0 2200 3.0 14,0 7.0 361.0
‘ L Unusual Expenses 1000 - 30.0 130.0
. ‘Total - 3=0.0 © 170 : _ ‘ ‘ L . 451.0
- . Recombiner S Mem 0.5 20,00 2.5 1205 1.5 65,5 3.0 139.0 1.0 9.0 1.0 50.0 1.0 51.0 10.5 495.,0
' o i Unusual‘ihcpenses ‘ _ - - . 50.0 1 o - ' 50,0
‘ . S Total . 200 . 120.5 _65.5 ‘ 189.0 _ 9.0 7 50.0 51.0 545.0
Blanket Pmcessing Mem 1.0 0.0 b5 233.00 9.0 W0.0 9.0 478.0 9.0 489.0 12.0 650.0 9.0 511.0 53,5  28)1.0
(Pa) ‘ Unusual Expmses > P T70.0 100.0 ' 200.0 © . 2%0.0 300, 0 300 o . 1220,0
Filtra.tion and Salt Menn : ‘ ‘ 1.5 66,5 2.5 10.0 2.0 91.0 6.0 267.5
Clean.Up Unusual Expenses : 30.0 ¢ 20.0 ‘ €0.0
| Total _ 96.5 1k0.0 9.0 327.5
Frocess Off Gas Men ©1.0°  57.0 2.0 106.0 3.0 1640 3.0 136.5 1.0 k4.0 1.0 5.0 1.0 5.0 12.0 £98.5
Handling Unusual Expenses S : 20.0 %0.0 50.0 %0.0 50.0 50.0 260.0
‘ Total S 57.0 126.0 204.0 186.5 9.0 . 95.0 96.0 858.5
Alternative Processing .. - Men o 2.0 80.0 5.0 210.0 L.0 168.0 2.0 86.0 2.0 88.0 2.0 90.0 2.0 92.0 19.0 814,
Schemes and Process Unusual Expenses S S ‘
Impmement o ‘ Total ) 80-0‘ 210-0 m-o ] . 8600 BB!O . 90.0 . %-O Bllho
-Special Instrumentation = Mem . 1.0 5.0 2.0 105.0 2.0 0T.5 2.0 110.0 2.0 112.% 1.0 50,0 1.0 51.0 1.0 55%.
Total . 45.0 135.0 157.% 160.0 162.5 100.0 101.0 861..0
Waste Handling a.nd Men s 0.5 35.0 2.5 143.0 3.5 2001.5 0.5 18.0 7.0 n7.5
Disposa.l Unusual Expenses : 10.0 30.0 - 40.0
- Total : 35.0 153.0 231.5 38.0 ‘ k57.5
General nesign " Mem | . 1.0 5.0 1.0 72,0 1.5 1M.0 2.0 1%2.0 2.0 160.0 7.5 5140.0
; Unusual Expenses ' ' ‘ :
L . Total 45.0 T72.0 1.0 152.0 160.0 gh0.0
Salt Production Men * 1.0 .0 5.5 272.0 8.0 336.0 8.0 3440 8.0 32,0 8.0 360.0 8.0 388.0 h6.5 2077.0
. 3 Unusual Expenses ‘ : ' 0.0 200,0 200.0 ‘ 550.0
_ Total : 45.0 322.0 636.0 5840 . 352.0 360.0 268.0 26217.0
Ehgineering ‘I"egt © Men . 2.0 1%.0 2.0 114800 12.0 582.0 20,0 880-0 B.0 m"o 6-0 300:0 50-0 2!"5“’-0
Unit Operations Unusual Expemea i .
fibtal 1“00 11‘800 fito 880-0 muo 300.0 21“5,"0
Total Men 18.0 850.0 39.5 1975 5 47.0 2325.0 53.5 2659.5 56.0 2723.0 40.0 1990.0 35.0 17T1.0 289.0 14292.0
o Unusual Expenses - 150.0 800.0 700.0 500.0 500,0 . . : 3650,0
Total 1000.0 2!;75 5 2125.0 3350.5 3223.0 2490,0 227.0 17942.0

 

..E'T(..

 
 

 

 

Contimious Fluorination

The continuous fluorination development is primarily an engineering
problem'which will center around the development of small towers operated
with the liquid phase continuous. By the end of 1970 a good estimate of
the size of equipment necessary to give thé desired recofiery consistently
will be established., The feasibility of protecting surfaces from corrosion
by a layer of frozen salt will be established, although plant conditions
cannot be realized in experiment. Scouting work will have been done on
alternative methods for continuous fluorination, including spray fluorina-
tion technique. Studies will have been initiated to determine the rate-
controlling mechanisms. In years 1971 through 197L additional work will
be done to establish the mechanism of mass transfer to allow more nearly
optimum operation. Corrosion protection by a frozen layer of salt will
" be further Studiéd.with emphasis onrstartup and shutdowh and the specific
design of inlet and outlet ports.

Sorption

The absorption system to collect and purify UF6 from the fluorinator
does not require extensive work; however, the use of a supplemental
magnesium fluoride bed to retain technetium will be studied. All pertinent
problems will be answered by the end of 1970, Attention will be given to
the specific equipment necessary for the absorption problems in the reactor

processing.
Carrier Salt Distillation and'the MSRE Experiment

_ Work on carrier salt distillation is logically divided into:two parts.
There will be an experiment in the MSRE in which salt from this reactor will
be distilled and the separation from fission products demonstrated in a
short campaign of batchwise operation. This should be campleted»by the end
of 1969 and will require, including FY 67 expenses, about 8-1/2 man-yearsv
and $500,000. The other part of the distillation study is of a more general
nature and will ‘be aimed at providing an effective unit for use in the

.

Q)
 

oo C‘ 2

ap

- L5 -

breeder reactor. By the end of 1970 relative volatility of all significant
components in the fuel salt and blanket salt should be known with accuracy
necessary for design. The;behavior of evaporating salt systems under the
temperature and pressure of the still will be understood and calculations
will be made to deflnethe requ1rements for mass transport from the bulk of
the solutlon to_the evaporat;ng 1nterface. The feasibility of using

~distillation will be well established by this time and the approximate

size of the unit will be known; however, details on still design will not

have been fixed,
Reconstitution of Fuel

Work on the reconstitution of the fuel will comprise a minimum study
for understandlng the effect of oxidation state of the salt on its chemical
and phy31ca1 propertles, and for prOV1d1ng an effective design. There is-
no question as to the fea51billty of the llquld phase reconstltutlon of the
salt, o -

Blanket Processing (Pa)

Studies on the processing of the blanket salt to remove protactinium
cannot be SpecifiCally outlined at this time because of the uncertainties

~in the chemistry of the system.. It is anticipated that engineering work
_'would not be undertaken in earnest for at least two years, during Wthh
'[tlme chemlcal studies and feasibillty studies would proceedr(whlle the
-;necessary experimental facilitles are being constructed) By ‘the end of
-gl970 it is anticlpated that there will be a flchheet for protactlnlum

_.removal, it will 11kely‘1nvclve elther reduction of protactlnium'W1th its
'ffaccumnlatlon on metal surfaces, reductlon ‘and a531m11atlon 1nto ‘a molten
-_metal or the ion exchange of protactlnium'with some active insoluble h

| ;foxlde.- Cold engineering studleS'would have cammenced by 1970 and englneer-
r"1ng experlments in alpha facllltleS'w1ll be in an advanced.plannlng stage.
 

 

- 16 -
Fi]_.tratiofi_-énd Salt Cleanup

It is antlcn.pated that there will be partlculate matter circulatlng

- with the salt streams. It is poss:.ble that intentional preclpltatlon of

specific fission products_may be used, but in any case solid matter will
be present. A small effort will be devoted 'tor the development of methods
to fllter and clean the salt durlng reactor operation. By the end of 1970
an acceptable filter medium should be avallable and tests should have been
star_ted These will term:.nate after 1971,

Process Off-Gas .Handl;i.ng

Studies will be pursued to define methods of handling and treating
the gases produced by the chemical proceesi'ng plant. Among _-the.problems '

‘which hopefully will be solved by 1970 is the problem of accommodating .

the fission product decay heat in gas streams. By the end of 1970 all
problems in the handling of the off-gas will be anticipated and defined.

Alternative Processing Schemes and Process I_r@rovement

There will be a study of reasonable intensity on the "ei{algation of
alternatives to the fluorination-dietillation process. Although'it is not
anticipated that a more desirable alternative will be produced study =
should be undertaken to ascertain that we have not missed a Dbetter process
than the one we pla.n to use. . Alternative methods_ for removing xenon from.
the fuel salt , specificaliy the removal of precursor iodine, will be
evaluated. This su‘oprogra.m will also include study of methods of processing
an alternative reactor which contains-thorium'in the fuel salt. This oha_nge_

in chemistry produces chemical and engineering problems zm_ueh more formidable

than those associated with the system without thorlum. It is anticipated

~that the study would form a nucleus of technology fram whlch a.more intensive
- program could spring if such an alternative reactor were necess_ery. By the

end of 1970 both the reactor system and the processing method will be firmly
fixed and this subprogram will be reoriented to handle process chemibtry

ey
 

C

ab

- L7 -

problems as they arise. The work on process alternatives should yield

- " valuable 1n81ght into the phase equilibria of circulating impurities that

may be found in the reactor system.
Special -Instrumehtation and Process Control

There are many rplaces.,where;spe‘cialiinstrdmentation and control are
crucial. The feed to the continuous fluorinator and the feed to the dis-

, tlllatlon unit must be metered _therefore;; s !cnowledge of propert:l.es ) partlcu-

larly uranium concentrata.on and gross beta-gamna, is necessary for process

control. A minimal program will be manned to. do this, although it is

anticipated that most proce ss control problems will be work_ed out as part

of the development of particular units.
Waste Handling and Disposal

- One importent'. problem in. tfie successful operation of the chemical
processing plant is the safe , temporary acconnnodatlon of the separated
fission products. These 1nelude those stored in the flssa.on product
retention tank frhich will be close-coupled to the carrier salt still),
alkali metals @thh are accumulated along with several other fission
products from the xenon removal sysbem), and the sorber beds from the
fluorlnatlon step.  No. fundamenta.l dlfflcultles are foreseen, but several

‘materials-handling and heat-_-transfer problems must be worked out.

Our waste disposal prcgféfifis'fiat intended to make major contribution

) to the philosophy of long-term waste dlsposal. Our effort here will be

restrlcted o establishlng an acceptable way - of dlsposn.ng of the particular

['waste from thls expermental reactor. : The method will be selected and
| _eveluated by the end of 1969 and deta:Lls of the facility should be avallable
B :by ‘the end of 1970, | o s -

: ,’Ge;zere;l Design

Provision is made for funding a generel design group who will put
together the components in a rational form and do layout work, etc., and
 

| - 18 -

whorwill maintain liaison with designers. By the end of 1970, spa¢e
requirements and the 'general' ‘appearance of the facility should be reasonably
established, |

Salt Production

~ As a service to all of j-‘l_:he Iaboratory gréups working with reactbr-grade
MSR-type salt systems, a group will be set up to produce salt in adequate
quantities. It is ‘anticipated that the existing facilities at Y-12 will be
used for at least the first two years, after which time some modification
will be necessary to allow increased capacity. These modifications will be
done in 1969 and 1970. ' | |

Engineering Test Unit

An Engineering Test Unit for the chemical processing plant will be
built as part of the engineering test unit for the reactor. By the end
of 1970 a layout for this mockup 'will be fixed., The vfunds included in
this write-up cover the Title I design and liaison with the architect-
engineer who will make the Title IT design. They also cover the operation
of the facility and an engineering development program which will use the
facility as its major tool. The denoted funds do not cover either the
detailed design or the construction of this engineering test unit.

Summary of Development Costs

The total program through the end of 1970 (fiscal year) will cost

about $5,150,000 for support of people and §1,450,000 for- special equipment -

and experimental facilities.. The program through FY 7l is estimated at
$11;,300,000 for the support of people and $3, 650 000 for spe01a1 equlpment
and experimental facilities,

Q.
u-C* y

o .

10,

1.
" for Period Ending August. 31, 1965 , ORNL- 3872 > 127 ff (December 1965).

12.

- 13,

- L9 -
'REFERENCES

C. D. Scott and W. L. Carter, Prellmlnary Des:.gn Study of a Continuous
Fluorination—Vacuum Distillation System for Regenerating Fuel and Fertile
Streams in a Molten Salt Breeder Reactor, ORNL-3791 (January 19606 ).

F. L. Culler, Chemical Technology Division Annual Progress Report for

Period Ending May 31, 196l, ORNL-3627 (October 190l ).

R. W, Kessie et al., Process Vessel Design for Frozen Wall Containment
of Fused Salt, ANL-6377 (1961).

W. R. Gr:l.mes, Reactor Chemistry Division Annual Progress Report for
Period Ending December 31, 1965, ORNL-3913, pp. 35-0 (March 1966).

L. E. McNeese and C. D. Scott, Reconstitution of MSR Fuel by Reducing
_Ugé Gas to UF, in a Molten Salt, ORNL-TM-1051 (1965).

D. E. Ferguson, Chemical Technology Division Annual Progress Report for
Period Ending May 31, 1966, ORNL-3945 (September 1966 ).

 

 

~W. R. Grimes, Reactor Chemlstry Division Annual Progress Report for

Period Ending Jamlary 31, 1960 ORNL-2931, pp. 83-L (March 1961).

L. G. Alexander et al., Thorium Breeder Reactor Evaluation. Part I.

| Fuel Yield and Fuel c:yc1e Costs for Five Thermal Breeders, ORNL CF-

W. H. Carr, "Vola.tlllty Processmg of the ARE Fuel,¥ Chem. Engr, Sym.

Series 56(28), pp. 57-61 (1960).

M. R. Bemnett et al., "A Fused-Salt Fluoride-Volatility Process for
Recovering Uranium from Spent Aluminum-Based Fuel Elements,” published
in Ind. and Eng. Ghem. s Process Des:;gn and Development 1965

R. B Briggs, Molten Sal'b Reactor Program Semla.nnual Progre S8 Report.

 

'W. R. Grimes ; Chemical Research and Development for Molten Salt Breeder
Reactors, (RNL- TM-1853 (June 1967).

D.- Scot.t Jr., Comyonents and Systems Development for Molten Salt

Breeder Reactors, ORNL-TM-lBBS (June 1967).
 

&,

(

)’

.y

 

.hv "
) r\ 3 .

b

a¥

15.
16.

18,
19.

20.
2,
220
23.
2h.
25,
26.
27.
28,
29,

30.
31,

32.

33 -3L.

35,

37.,
38.
ko.

L3

..fi_..hs;.
k6.

L.

18,

L9,
50.

, K. Adams -
M. Adamson o

@, Affel -

- G,'Alexander»ffli:
"F. Apple

F. Baes
M. Baker

. Jo Ba-ll

F. Bauman

E. Beall
Bender

S. Bettis ,
F. Blankenship
E. Blanco

0. Blomeke
Blumberg

G. Bohlmann

~ J. Borkowski

E. Boyd
Braunstein
A. Bredig

B. Briggs

R. Bronstein
D. Brunton
A. Canonico
Cantor

Lo Car‘ber

I. Cathers
M. Chandler
L. Compere
H. Cook o
F. Cope -
T. Corbin

L. Crowley

L;_Culler,'Jr;_ ?5§f'”'

Mo .Dale o
G. Davis
J. Ditto

8., Dworkin

R. Engel
P. Epler
E. Ferguson

‘M, Ferris

P, Fraas

A.:Friedmah*i[r'”

H., Frye, Jr.,
H. Gabbard

B. Gallaher

-5l -

~ " DISTRIBUTION

51.
52,
53.

Sl

55.
56.
57.
58.
59.
60.
61,

62.

63.
6L,
65.
66.
67-
68.

69.
20

T1.

T2,
- 3.

L.
7.
76.
7.
78.
79.
80.
81.
82,

83,
8.

&

87,
88,
89,

900 '
9l.

92 *

. 93 *

9k
95.
96-110
111,

112,

113,

86.

El
R.
G.
H.
A.
H.

'-G.
S.

N.

A,
G,
, R.
W,

w.
L.

H.

R.
- d.
~T.
‘R,
W,
T,

S.

I'o .
W.

E.
A.

J.
B.
P.
1.
NQ‘
VG.

E.

- D.
-
L.
W.
McCoy

“McDuffie

E.
F_:'o

K.
Je
W,
, BE.
S.

L.

‘Giambusso, AEC-Wash.,

Goeller
Grimes

Grindell

Guymon
Hannaford

‘Harley
‘Harman
Harrill

Haubenreich
Heddleson

‘Herndon

Hightower
Hoffman
Horton

‘Hudson
Inouye =

Jordan -
Kasten
Kedl
Kelley :
Kelly
Kennedy
Kerlin -
Kbrr o
Kirslis
Krakoviak
Krewson
Lamb
Lane
Larkin, AEC-ORO
Lindauer .
Litman

,Lundln

Lyon -

-MacPherson o

thPherson‘
Martin =~
Mathews
Matthews
McClung

McGlothlan
McHargue

McIntosh, AEC-Wash,

McNeese
Meyer
Mbore
 

11k,
115.
116.
117.
118.
119,
120,
121,
122,
123,

12h.
125-17h.

175.
176.

177.
- 178,

179.
180.
181.
182-183,
18L.
185.
186.
187.
188.
189.
190.
191.
192,
193.
194,

195.

196.
197,
1980
199.
200.
201,
202,
203.
204 ..
205,
206,
207.
208,
1209,
210.
211.

H.
WO
J.
M.

M.

G.
w.
A.
_F.
G.
. 0-

P.

W.
I.
R.
H.
R.
Je.
E.
R.

) Jo.

C.
B.

A,

J.
W.
K.
M.
Je

- L.
- G.
H.

- Sé'_

DISTRIBUTION (cont.)

P. Nichols

L. Nicholson 211215,
C. Oakes 216-225,
Patriarca 226,
M. Perry 227-2l1.
B. Piper

E. Prince

L. Redford

Richardson |

C. Robertson

C. Roller

W. Rosenthal

M. Roth, AEC-CRO

C. Savage
E. Schilling

Dunlap Scott

E. Seagren

F. Schaffer

H. Shaffer -
Shaw, AEC-Wash.

J. Skinmner -

M. Slaughter :
L. Smalley, AEC-(RO
N. Smith

J. Smith
P. Smith

L. Smith

G. Snith

F. Spencer
Spiewak

C. Steffy

H. Stone _
F. Sweek, AEC-Wash.
R. Tallackson
H. Taylor

E. Thoma

S. Watson

F. Weaver

H. Webster

M. Weinberg
R. Weir

J. Werner
W. West

E. Whatley

C. White

V. Wilson
Young

C. Young

212-213.

Central Research Library

Document Reference Section

Laboratory Records Department
Laboratory Records, ORNL R.C.

Division of Technical Informa-
tion Extension .

» [ Mr>'?f‘