ORNL-TM-1730.txt

 

i.fié;-j

>

 

 

o abrteen 1 e
ot ‘

™~

Contract No. W-T405-eng-26

CHEMICAL TECHNOLOGY DIVISION

‘ ACTOR
TO PROCESSING OF MOLTEN-SALT BREEDER RE

L. E. McNeese

 

ORNL-TM-1T30
CESTI PEICES

HC. scflfl; MN__-é:f_

I

 

 

Y

BELEASED FOR ANNOUNCEMEKT
- "IN NUCLEAR SCIENCE ABSTRspRg

 

MARCH 1967

B racy, Completeness, o usefulness of the information ed, v Ith reapect 4, the accy-

Tt, or that the use

t_flscloa_ed in this reporg may not infringe

If of the Commisasop Includes iuh_v em-

e " { OT employee of Such eoniractor, ¢o the
S ;;wb amplqygg OF contractor of the Commission, o employee of guch co:m'lcto premat
' MiBBeminatag, o Provides accegs to, Ormation " Frepares,

OAK RIDGE NATTONAL LABORATORY
.~ 0ak Ridge, Tennessee
 _operated by |
UNION CARBIDE CORPORATION
- for the |

aay tion Pursuant to g o
i88lon, or hig *mployment with gyep contractor mplomt o contracy

' ‘U S. ATOMIC ENERGY COMMISSION

 
 

 

 

 

 

 

 

$

 

 
 

 

iif

CONTENTS

AbstraCtl e & -8 @ » -8 -8 - e« - .8 * * - e -0 * -9 - . . e " e .

lo IntrOdUCtion ¢« -5 ® s & 8 -% @» LR L I L I A I AL I

2, Distillation at Low Pressure . . v «.« o « o« o o o s o «
2.1 Equilibrium Distillation . . . . v ¢ ¢ .o v o o & .
2,2 Molecular Distillation .« .. o o « o o o = « « o o »

2.3 Mean Free Path . . . ; s s s e s s s e e s v e e
2,4t Langmuir Vaporization Rate . . « v v o «. v o + « 4
, 2.5. Probable-opérating Mbde for MSBR‘Processing.; o o

'3, Relative Volatility . . . ¢ o v ¢ ¢ .0 0 o ¢ e o o o s s

3.1 ‘Relative Volatility for Equilibrium Distillation .

‘3,2 Relative Volatility for Molecular Distillation . .

3.3 .Comparison of Experimental and Calculated Relative
Volatilities for Rare Earth Fluorides . . .. . .

4, -Separation Potential of‘Various.Distillation'MEthods . o

4.1 Continuous Distillation . . « & ¢ v ¢ ¢ 0o v v o .0 s
4.2 Semicontinuous Distillation with Continuous Feed
4.3 Semicontinuous Distillation with Rectification . .
4L 4 Batch DLSEL11ation o . v s o v v b oie e e e
4.5 Semicontinuous Distillation Followed by Batch
: Distillation ... . .,.,; b e e e ee b nieeen
'Vh'6 Comparison of Methods COnsidered . ¢ eie.e.ae

6. Conclusions and Recommendations  ;I.];.......;.....,,.,.

"""Referenc_es v s e tooi" 7!.-"-“'-"'"'T"‘,""'-" *

-Page

Vi EF W W N M

10
11
12
b

16

16
16
19
21
28

29
.32

5. Prevention of Buildup of Nonvolatiles at a'Vapprizing Surface

by Liquid Phase. Mixing .a.‘;.. ...'.,.,....,f. o le ea .36

L3

- 43
e b Wl

 

 

 
t)

0

S

wh

 

CONSIDERATIONS OF ‘LOW PRESSURE DISTILLATION AND ITS APPLICATION
TO. PROCESSING OF MOLTEN-SALT BREEDER REACTOR FUELS

_L.-E.-McNeese
| ABSTRACT‘

Distillation at low ‘pressure was examined as a: method
for removing rare- earth fluorides from the fuel stream of
-a molten-salt breeder reactor,' ‘It was concluded that 'dis- - .
tillation allows: adequate rare earth fluoride removal with
‘the simulténeous recovery of more ‘than 99,5% of the fuel
salt. Characteristics of equilibrium and molecular =
distillation were noted and expressions for the relative !
volatility of rare earth fluorides were derived for these
types of distillation. 7 . e -

- Expressions for the separation potential of several
modes of distillation were derived and reported rare
earth fluoride relative volatilities were shown‘to:allow_,

& great deal of latitude in still design and operational

- mode. It was concluded that a single contact stage such
as a well mixed liquid pool provides adequate rare earth.

-fluoride removal and that rectification is not required

. The buildup of rare earth fluorides at the vapori-
zation surface was shown to seriously reduce the R
effectiveness of a distillation system, .Liquid circu-

 lation was shown to be an effective means for preventing
"buildup of rare earth fluorides at vaporization surfaces.'

"lgjlleRObUCTION |

The molten*salt breeder reactor (MSBR) is a reactor concept

having ‘the possibilities of economic nuclear power production and

simultaneous breeding of fissile'material using the thorium-uranium -

'ri'fuel cycle. The reactor will be fueled with a mixture of molten

fluoride salts’ which will c1rcu1ate continuously through the reactor

core where fission occurs and through the primary heat exchanger"

';where most of the fission energy is removed The reactor will employ'r

a blanket of molten fluorides containing a fertile material. (thorium)

 
 

in order. to: increase the neutron- economy of the system by the con-.

version of thorium to fissile uran1um-233 A close-coupled processing

facility,for removal of fission products, corrosion products, and
fissile materials from these fused fluoride mixtures will be an

integral part of the reactor system..

1t has been proposed that the rare earth fluorides (REF) and
fluorides of Ba, Sr, ‘and Y be removed from the fuél stream by -
vacuum distillation. The purpose of this report is to examine
various factors pertinent to such an operation and to compare several

methods for effecting ‘the . distilIation_ .
2. }DISTILLATION’AT,LOW‘PRESSURE -

The vaporization of a liquid is normally carriedlout under . -
conditions such that the 1liquid and vapor phases are essentially '
in thermodynamic equilibrium. -This condition may cease to exist
1f the distillation pressure is reduced sufficiently; and phenomena

peculiar to low pressure distillation may be observed.

-In discussing digtillation at lowfpressure,,it'is convenient

to'make-a distinction between two modes of distillation-' equilibrium |

distillation and molecular distillation. During equilibrium
distillation, a kinetic equilibrium exists at the liquid-vapor
interface owing to the presence of a vapor atmosphere above the
liquid which has the net effect of immediately returning most of the
vaporizing molecules to theoliquid surface. In contrast, molecular
‘distillation is carried out in the absence of such an atmosphere

and the vaporizing molecules reach the condensing surface without
experiencing collisions ‘with other gas molecules or with the walls
of the system. In the following sections, consideration will be
given to characteristics of these modes of distillation, to. values ‘i
of the mean free path under conditions of Anterest for MSBR , N
’ processing, and to calculated values of maximum vaporization rates

to be expected

 
 

ry

cl

 

2.1 Equilibrium Distillation

Equilibrium distillation can be further divided into ebullient

distillation and evaporative distillation. Ebullient distillation N

occurs when bubbles of vapor form within the bulk of the liquid
which remains at a temperature such that -the vapor preSSure is
equal to the total external pressure acting on the liquid (in the

absence of other gases Boiling‘prOmotes mixing in the liquid

.and the surface from which vaporization occurs is not depleted in

the ‘more- volatile species,

“Evaporative distillation occurs when the distillation is carried
out at a temperature below.the bdiling‘point of the liquid. Under

‘these conditions there ie no formation of bubbles at points below

the ‘1iquid surface and no visiblemovenent‘of the ‘1iquid surface,
Transfer of the more volatile species to the. liquidisurface occurs

by a combination of molécular’diffusionfiand convective mixinglso:'

that depletion of this species in“the viéinity of the- surface is

possible, _However, the rate of distillation is relatively low owing
to the kinetic equilibrium which exists ‘at the liquid-vapor -interface
and the liquid surface may have ‘a composition near that of the bulk

‘1iquid.

‘o.2 'Molecularr‘Distillation |

Mblecular distillation is quite similar to evaporative distillation

-in that vaporization occurs only from a quiescent liquid surface

end in that ‘the vaporizing species 1s transferred .to the surface by

- molecular. diffusion and convective mixing. However, few of the

vaporizing molecules are returned to ‘the liquid surface by

"collisions in the vapor space above the liquid and vaporizdtion -
: 1proceeds at the greatest rate’ possible at the operating temperature.

| f In order to achieve this condition, the distance between the

vaporizing surface andstheieondensing-surface should_theoretically

l,beqleSS'than_the "meanflfree"path“of-thehdistilling moleculeS. This

condition - is seldom realized in practice, however ‘the distance should

 
 

he”no‘gréater than a few mean free paths,
greater buildup of rare earth fluorides at the 1iquid surface than |

do those- of evaporative distillation at the ‘same temperature where
vaporization is impeded by the vapor - atmosphere above the liquid |
‘which serves to Teturn most of the vaporized molecules to the liquid lfi,

surface,

“In a distillation‘system, the gases in the region between the -

vaporizing liquid and the condenser normally consist of a mixture

2.3 Mean Free Path

These conditions favor a

of ‘the distilling molecules and molecules of noncondensable gases.-“

. The calculation of the mean free path in this region is complicated

by the fact that the vaporizing molecules, which have a net velocity

 component directed away from the liquid surface, pass into 8as
vhose molecules are in random motion.
‘1 molecule in type 2 molecules may be obtained from a relation

given bijoeb2 as

' 01,0z = collision diameters of type 1 and.type:2*molecules',

M,2

- The :mean free path of a type

 

oo+ 022, €SP
v (2422 “41*6?5

mean free path of a type 1 molecule moving among type 2
molecules

)

‘n = number of type 2 molecules per unit volume -

By making appropriate substitutions into this relatiofi, one can obtain
the following relation for the mean free path-of a type 1 molecule

in type 2 molecules at a pressure P, both gases.being.at;thec,;'
temperature T. | o

C1,Cz = average velocity of -type 1 and type 2 molecules.

@

 
 

[

n}

&1

 

where ;
R = gas constant, (mm Hg) (em®)/(°K)(gmole)
03,02 = collision;diameters of type 1 and 2 molecules

CMy,Mp = molecular weights of type 1 and 2 m°le°files"~

Values for the mean free path of LiF in Ar and in LiF at 1000 C

:;are given in Fig. 1. It should be noted that the mean free path of
'LiF at a pressure of 1 nm Hg-1s approximately 0. O4 cm and that at a
pressure of 0.0l mm Hg, the mean free path of LiF is approximately
4 cm., These distances are probably quite small in comparison with '

~ the distance between the condenser and the surface from which

vaporization will occur- in an MSBR distillation system. ‘Hence, the
rate of distillation in an MSBR system will be set by the pressure

:-drop ‘between the liquid surface-and the condensing surface. . The
~values for the mean free path are sufficiently large that slip-flow

may be of importance in presgure drop considerations,

2.4k Langmuir Vaporization Rate

‘The maximum rate of evaporation of a pure substance was shown

by Lsngmiur5 to be =¥r__ |
=
W = 0.0583 EP - (3)
where |
W = evaporation rate, gms /cmZ 1 sec
M= molecular weight =~
‘T= absolute temperature, éKV -

'7P-e vapor pressure, mm Hg.

'_'A derivation of this relation will be given in order to show the
”region of its applicability., Consider a plane liquid gsurface at -

a temperature below ite’ boiling point. At equilibrium, the rate

*of vaporization from the surface will equal the rate of condensation
- on the surface. Langmiur postulated that the rate of vaporization

;_in a high vacuum is the same as - the rate of vaporization in the

presence of a saturated vapor . and that the rate of condensation in

 
 

PRESSURE (wn Hg)

 

  
  

 

 

 

B | o ORNL DWG 67-19

10 ¢ T |fii-1i|], T T T 11T — T T T T T 1T

-

0.1 b— —
e - «

N 2

r- -

= i -

oorl 0 vl v vl 4

0.01 0.1 o !
L | " MEAN FREE PATH {em)

Fig. 1. Mean Free Path of LiF in Ar and in LiF at 1000°C.

g o i

e . T

 
 

4]

o

1))

so that

 

" a high vacuum is determined by the pressure of the vapor. At

equilibrium the rates of vaporization and condensation are equal

and the rate of vaporization can be calculated from the rate of

-_condenSatlon.

‘The vapor contained in a—onit;cube in contact with the liquid

“surface is in equilibrium when the number of molecules ‘moving

toward the surface equals the number moving away from the surface.

‘For n molecules of mass 'm in the volume v, the quantity of vapor

approaching the liquid surface will be

i omn 1 o o . :

where p is the vapor density._ The average component of velocity
of molecules moving toward the surface is é U, where U.is the

arithmetic mean velocity of the molecules. The mass of vapor

- striking a unit area ‘of the liquid surface per unit time is then

T I

If the vapor is an ideal gas,

PM - -
and '
PV=RT=3-mnC.. | B _ | (7)
Solving for (Ca) /2 yields _“__h : - |
. 170 r = e T |
(02)/ ifi‘ | - - ®
- vwhere M = pv. The mean veloc1ty U is related to the root mean

7square velocity, (—E)l/e :?:i~;7l”-"

,-"hfl-;;‘u-'hf-?;Bl ..-”n e
.' U=J%fi@ S : ,. L e (10)

 
 

Thus
or |

W = 0.0583 P—T P
The assumptions implicit in the use of this relation for cal- : ..
-culating the rate of vaporization from a liquid surface include P
the following: '
(1) The liquid surface is plane.
(2) The liquid surface is of infinite extent, i.e. collisions

-of molecules with the vessel walls in the vapor space must -

exert a negligible influence on the rate of vaporization. a e

(3) The vapor behaves as an ideal gas.ki,

-F

(4) Every part of the liquid surface is within a fraction of fl- o

the mean free path from every other part or from a
condensing surface,-i.e., thereffect of collisions between
evaporating molecules on the ratehof vaporization is
negligible. | S - «

(5) The number of molecules-leaving the liquid surface is not
affected by the number striking the surface. |

(6) Vapor molecules striking_therliquid surface are absorbed
and revaporized in a direction given by a cosine relation
which is independent of the direction of aoproach atithe

moment of absorption.

When applied to the vaporization of LiF-BeF2 mixtures, the

poorest .of these assumptions is likely that of considering the vapor “i__m

to behave as an ideal gas since it is known that gaseous_LiF ‘tends

to associate. The vaporization rate given by Eq. 11 represents the
maximum rate at which vaporization will occur and hence sets an
upper limit on the vaporization rate. Values forvthe-Langmulr
vaporization rate of LiF are given in Fig. 2. The vaporisation‘rates-
observed in practice may be considerably lower than the Langmuir rate

since the fourth assumption is rarely met,

 
 

 

 

 

+

VAPORIZATION RATE (g/cm. sec)

. 0.0001

0.04

- 0.0

 

ORNL DWG 67-20R1

 

 

 

 

 

800

| TEMPERAIURE (-c) Sl S e

| Fig2 La.ngmuir Vaporization Rate for LiF.

 
10

2.5 Probable QOperating Mode for MSBR Processing

The mode of distillation currently envisioned for proceSSing
MSBR fuel salt is that of single stage equilibrium distillation at
950°~1050°C and at a pressure of -approximately 1 mm Hg. The -
composition of liquid in equilibrium with vapor having the composition
of MSBR fuel salt (64 mole 4 LiF - 36 mole ¢ BeFy) is approximately
88 mole % LiF - 12 mole % BeF. The vapor pressure of liquid of
this composition is ~ 1.5 mm Hg at~lQOO°C_.5 -Hence, evaporative
distillation, with surface vaporization onli, will occur if the
distillation is carried out at a pressure greater than 1.5 mm Hg.
However, if the distillation is carried out at a pressure lower than
‘1.5 mm Hg, boiling could occur below the liquid surface. At a pressure
of 0.5 mm Hg, boiling could occur to a depth of about 0.7 cm. The
actual depth to which boiling would occur is-dependent on the
vertical variation of liquid temperature and composition (and hence
vapor pressure) and on the extent of superheating of the liquid;
the value of 0.7 cm assumes a constant temperature and concentration
and no superheat throughout the bulk of the liquid. Boiling in the
vicinity of the liquid surface would promote convective mixing
which would result in a lower rare earth fluoride concentration
at the liquid surface than would be observed without such mixing.

The lower surface concentrations would in turn decrease the relative

rate of volatilization of REF with respect to LiF.

 

.Molecular distillation offers two advantages over either type
of equilibrium distillation in.that (1) the distillation proceeds
at the maximum rate, and (2) a greater separation of rare earth

fluorides from the MSBR fuel salt is possible as will be discussed

~ in the section on relative volatilities. 1Its chief disadvantages
are the low pressure required to achieve this type of ‘distillation
and the increased likelihood of an undesirable buildup of rare eafth

fluoride at the liquid surface.

The MSBR distillatlon system will probably be operated at the

vapor pressure of the liquid at the vaporization surface or possibly

 
oy

L)

 

11

.at a pressure:0,5-1.0 mm Hg lower than the vapor pressure. A decrease

in pressure could yield an increase :in distillation rate-and/or -a
deerease-in operating-temperature. Entraimment -at the lower pressures
should'be»consideree.'\It'is improbable thatjthe'advantages to be
gained'by'molecnlar'distillation justify the;efforthnecessary to attain

this mode of operation.

" %, " RELATIVE VOLATILITY

- The relative volatility;is;a convenientrform4for,presenting

‘data relating the compositionnof‘liQuid and vapor phases at equilibrium

and,is-defined as

 

Yy,
AR
Qe = . (12)
AB xAle |
vhere | o ‘ - | .
:aAs =:relativepvolatility,of componentiA,referreo to oomponent'B
Yprg =.mole fraction of component A, B in vapor
X, 2%y = mole fraction of component A, B in liquid.
If the conoentration of component A is low. compared to that of
‘the major component (B), the relative volatility can be expressed
in a useful approxrmate form |
% "T, @)
where o I
C, = moles A/unit volume of .condensed vapor
1Cg-= moles:A/unit volnmé‘of'lionid"

. and where the condensed vapor and liquid are at the temperature at which
| '~vaporization is . carried out, In a binary system, the error ‘introduced -

_in relative volatility by -this approximation depends on. the concentration

~of component A, the relative volatility, and the. relative molar volumes

-of components A and B. The error can be evaluated as follows.

 
 

12

Let O denote‘the aotual relative "oolatility as defined by - -
- Eq, .12, and- a* denote the relative vdlatility in the approximate
form defined by Eq. 15. From Eq. 135, '

'Y, X +(1 "}i')v

 

a* = ‘A A
X, Y, V, + (1 -¢¥ )v
The relati.on between Y and X 1is given by Eq..12 and its use with

A
the expression for a* yields the fractional error in o as

o .o X (1 a)(v /v )

o o S
rrfrac error = o 1 "X [1 a(v /V)] .

The fractional error in afilis given in Eig.‘j as a function of'_X.A
for various values of & for the case where the molar volumes of A

and B are equal. It should be noted that the error in @ introduced
by Eq. 13 is less than 184 for X < 0.15 mole’ fraction if @ < 2,46,
For rare earth fluorides in LiF, the error in.d' will be approximately
three times the values shown for & < 1072 since the molar volume

of rare earth fluorides 1is ~approximately ‘three times that of LiF.

The definition of relative volatility given by Eq..12 has . -
‘been used throughout ‘this report except in Section 5 where the
definition given by Eq. 13 was used - |

The appropriate forms of the relative voIetility will be
derived in the following sections for both equilibrium and molecular
distillation, and a comparison of experimental and calCuleted'values
of relative volatility for several rare earth fluorides in LiF will

be made.

3.1 Relative Volatility for-‘Equllibriinn,‘ Distillation -

- In equilibrium distillation, the relative volatility relates
-the composition of liquid and vapor which are -in thermodynamic
equilibrium. .For the ith component of a system which obeys Raoult'

Law, one can write = S o C T , -

 
 

 

 

A

&x¥

I

1Y

| CWt-0
a
5

 

 ORNL DWG 67-21 -

 

ST T T T T
0-6—

04 b

o
»N
I

 

 

B |

a <10

  

 

 

-0.4 |

 

 

 

 

- 001 . o
| | X A. finofe froc!lon of cmpomnl A In liguid)

0.5

Fig | 3. - Error Introduced in Relative Vola'bility by Use of

Approximate Form of Relative Volatility. S

 
 

1k

Ty =By
vhere |
o = total pressure
P, = vapor pressure of component i
Vg = mole fraction of i in vapor
*1'=;9°Ie fraction 'of i in liquid.

Substitution of this relation into the definition for'reletive
volatility of component A referred to component B yields

% =T ey T F,

_ *A B
Raoult ¢ Law implies the absence of chemical interaction between the
components under consideration. Interaction may be taken into
account if information is available on the activity of the components

‘since one can write for the ith component
Ty = Bxg o (@9)
where

_71 = activity coefficient of conponent-i,in liquid of the

composition under consideration.

‘The relative volatility may then be written as

P

 

(o
AB 7B 7B

3.2 Relative Volatility for Molecular Distillation

With molecular distillation, the liquid and vapor phases are’
not "in thermodynamic equilibrium,,instead, the composition of the
vaporiting~material i{s related to that of the liquid by a dynamic
equilibrium which is dependent on the relative rates for naporization
of the various components of the liquid. An expression for the

rate of vaporization of a pure liquid was derived in Section 2.4,

PA"A/PB"B P -
| ,(111) |

/S o (e)

:\ r () £

 
 

15

Division of the- rate equation (Eq. 11) by M, the molecular .

weight ‘yields the molar rate of vaporization as

P

 

g
L

==

F

. - . (n
Thus, the molar rate of vaporization for any substance, atlaigiven'
‘temperature, is governed by the ratio of P/{M. "In a binary system,

if Raoult's Law is assumed,

Py =% By - (18)
where _ :
X, = mole fraction of component i in the liquid
Pi,='vapor pressure of pure component i
pi = partial pressure-of-componentii at liquid:surface. .

- The mole ratio of components vaporizing from the liquid surface
is then ' ‘ B

o ATA r MB A (19)
- T ;] R T x13 | |
Since the quantity /mB is related to the ratio of the mole

fraction of components A and_B in the vapor as

(20)

it
‘éfilfifi

oF ifi»g |

where L o
~yi~; mole fraction of component i in vapor

_i one obtains af the relative volatility as'

o = A/x:fié‘l_; (&)

- This should be compared with the relation for relative volatility
for equilibrium distillation which was P /P | |

 

 
 

 

16

3.3 Comparison of Experimental and Calculated Relative .
Volatilities for Rare Earth Fluorides
‘Relative volatilities for several rare earth fluorides in LiF
have been measured at 1000°C bj‘Hightoweré and the relative
volatility of LaFs in 87.5 -11.9 - 0.6 mole % LiF-BeFp-LaFg has
been measured at '1000° and 1028°C by Cantor.7 These - experimental .
data and calculated relative volatilities for rare earth fluorides -

8 »9,10

for which vapor pressure data '’ are available are given in

Tebie‘l. Calculated values were obtained using Eq. 14 and Eq. 21.
4. SEPARATION POTENTIAL OF VARIOUS DISTILLATIONIMETHODSr :

Several modes of operetion~are=availab1erfor.the.distillation

of MSBR fuel selt; the_choice_betweenttheee-will_ipvolve coneideration .

of their separation potential as well as factors 'such as degree of
complexity, economics, etc, In the following sections, a compariseh “
will be made of the separation potential of distillation systems
employing continuous, batch, and semicontinuous methods. A semi-

continuous system employing rectification will also be considered.

4.1 Continuous Distillation

Consider a continuous distillation syetem,of the type shown in
Fig. 4. 8Salt containing Cf moles REF/mole LiF is fed to the |
system at a rate of F moles LiF/unit time. A vapor stream
containing OC moles REF/mole LiF is withdrawn at the rate of v moles
‘LiF/unit time and salt containing C moles REF/mole LiF is discarded
at the rate of F-v moles LiF/unit time, A material balance on REF
yields the relation | '

FC wa + F-V)C.

£ -
The fraction of the REF removed by the distillation system is then

given as

QO

 
 

Table 1. Experimental and Calculated Relative Volatilities of- Several Rare
Earth Fluorides Referred to LiF at 1000°C

 

Vapor Pressure o

mm Hg

 

.~ Relative Volatility

 

 

',Rare Earth'E1fiotide__ .

*Smfia: e
,cngf
. ,LaFa

o

' g:1.6-x 10f§fl-'”“'

1.3 x 104
. .2.8x105

8 ._6 X ‘_210-"4
1.09 x 103

| Calculated
” - ~ Equilibrium = Molecular .
*gExperimental31-Distillation - Distillation
6x10%  31x10% 1.1x10*
5x10% - - -
3x103 = 24k x104 8.7x 105
3x10%  52x10° 1.9x10°

b

*Measured in 87.5-11.9-0.6 mole % LiF-BeF,-LaFs at 1028°C.
‘bMéasuréd;ifi'87,54;1!9-0.6'm01e ¢ LiF-BeFy-LaF4 at 1000°C.

LT

 
 

 ORNL DWG 67-22

Vapofi.ze:d Salt
v, a C :

 

 

L Discard 'Sc_a_,ltzr, |
. F-v, C

 

 

 

Feed Salt I |

F, Cf

Fig. 4. Continuous Still Having a Uniform Concehti'ation of Rare
Barth Fluoride in Liquid. : .

 
.

:fOr‘V CQnstant

 

19

fraction of_REF removed = (E_- V)C.= 1

_ (22)
FCg 1+

 

If the fraction of the LiF fed to the system which is vaporized

.is denoted as f, where |

 

v
t= F
‘then
fraetion'of REF removed = ———;—EE-. (23)
LT

- The fraction of REF. removed was .calculated using Eq. 23 for

various values of £ andtx -Ls shown in Fig. 5.

4,2 Semicontinuous Distillation with Continuous Feed

Consider a single-stage distillation system which contains V

-moles LiF at any time and a quantity of BeF, such that vapor in o

equilibrium with the liquid has the composition of the MSBR fuel

~ salt, Assume that MSBR fuel salt containing X moles REF /mole

LiF is fed to the system at a rate ‘of F moles LiF/unit time where

-1t mixes with the ‘liquid in the- system, Let the initial REF

concentration in the'liquidube'x moleS'REF/mole LiF and let the

concentration at any time t be X moles REF/mole LiF. From a mass

balance on REF, _

: _' s - ‘E(VX) = FXO "FOfX IR = (24)
¢Wh1thhasxthe solution_
Xe=g -@-a)exp(-Fae/V)] - (25)

The total quantity of REF fed to the system at time ‘t is

-(Ft + V)X and ‘the quantity of REF remaining in the liquid at ‘that

 
 

L0 g

o
oo

FRACTION OF RARE EARTH FLUORIDE NOT VAPORIZED
e
o

04—

20

| ORNL DWG &7-2

 

 

, T ——
@ =0.0001
\ 30 _00_057

 

 

 

02 I | |
0.945 0.97 0.98 0.99 1.0
' FRACTION OF LiF VAPORIZED

Fig. 5. Rare Earth Fluoride Removal in a Continuous Still as.a

Function of LiF Recovery and Rare Earth Fluoride Relative Volatility.
 

.-

 

el

| tifiehierXf Thus, the fraction of the ‘REF not vaporized at. time t: is

 

 

o 'l'e,' VX e 1 - (1 - a) exp (-EJCIV)
fReF = (Ft + Nk, T (%1:_ . 1) ,, (26)

where

fREF = fraction of REF not vaporized at time t, -

,The fraction of the LiF vaporized at time t. is given by the relation

__Ft
iR (2m)

 

where

f = fraction of LiF vaporized at time t.

rSubstitution of Eq. 27 into Eq. 26 yields the desired relation:

'kafj= L= [l -r(l - @) exp (-af/l - f)] o (28)

Values for the fraction of REF not volatilized as a function of the.
fraction of the LiF volatilized are - shown in,Fig 6 for various
values - of REF relative volatility '

‘H;5 ‘Semicontinuous Distillationrwith Rectification

Consider a distillation system as shown in Fig. T, which

consists of a reboiler and one ‘theoretical plate to which reflux is»

returned, The feed stream to the reboiler consists of F moles

LiF/unit time plus X moles REF/mole LiF and Z moles BeFa/mole

';LiF. The following assumptions will be made-'

3;(1) At any time the reboiler contains V moles LiF, where V. is
L s constant. e ' S
l'*;f(z) The initial REF concentration in the reboiler 1is X
. .moles REF/mole LiF.-_' ) '
(3) The concentrations of BeFfi%‘moles BeFz/mole LiF) throughout
_the system are- ‘the steady state values, i, e., the values
o which WOUId“be ofitained at steady state. with a feed ‘Stream :
containing Z moles BeFalmole LiF and no REF

 
 

22

ORNL DWG 67-24

 

1.0

 

 

08—

0.4 [—

FRACTION OF RARE EARTH FLUORIDE NOT VAPORIZED

0.2}

 

 

 

0.965 057 0.98 0.9 1.0
' FRACTION OF i.iF VAPORIZED

- Fig. 6. Rare Earth Fluoride Removal in & Semicontinuous Still as
e Function of LiF Recovery and Rare Earth Fluoride Relative Volatility.

 

 
 

 

23

~ ORNL DWG" 67-25

 

 

~ Condenser

 

(R + fl F
'aREF'x‘-. i

 

 

—P»F %er*y: 7,
| rRF Crieriodin
| ®reF*y

Zo

 

v

 

 

 

X

One Theoretical Plate

 

Z S of Rectificaflon

 

Y .

ererXs
CBeF, Zs

2 -

 

A

X,xv Z1 .

 

 

Pl

- Xo

 

 

| rRebOIIer S

 

 

”'_zo_

' Fig. 7. Semicontinuous Distillation System with Rectification.

 
 

2k

(4) The vaporization rates for LiF and BeFy are unaffeéted_by
fhe presence of REF and hencéiare the-steady state
vaporization rates. This assumption is;made-ifi_view-of
the-lowféoncentration of REF in thé vapor from the

' reboiler and hence in the liquid on the plate and in the
distillate. - o

(5) The holdup in all parts of ‘the system excluding the

| ~ 1liquid in the reboiler is negligible. '

(6) The relative volatilities of BeFp and REF referred to

| 'LiF are constant.
For Ealéulatidh of tfie»sféadf ététe‘Bng concentration and the
eteady state vaporization rates fo;;BéFa and LiF, it is assdfié&
- that the heat input rate to the-reb§iler-is equal to the heat
withdrawal rate in the condensef-(fiegligiblé;heat of mixing) which
18

= (R + 1)F (lLiF'T'ZoxBeFé)

heat input rate of reboiler

qH O
n

= reflux ratio

"

feed rate of LiF to reboiler, moles/unit time

if = molar heat of vaporiiation for' LiF

F
xBeF = molar heat of vaporization for Bng
2
Z
o

 

= concentration of BeFs in feed and distillate, moles BeFp

 

 

mole/LiF. ,
The rates of vaporization of LiF and BeF, from. the reboiler are =
then
moles LiF vaporized/unit time = $———7 Q 7 - (29)
- - Muar * %Ber, Zs MBers
g, 2 '
moles BeFp vaporized/unit time = +_a? 7 (30)
ALiF Bng s 7\BEFE
vhere

aBéF = relative volatility of BeF, referred to LiF

 
 

e st e AR B0 SRR & gL gt 1t

 

where

 

25

‘Zs.= concentration of - Bng_in liquid in reboiler, moles. .
BeFs/mole LiF,

A material_balanceeon_LiF around plate7leyie1ds

9

‘ _ + RF = (R_-l-‘l)F. +X
MaF * %Ber, Zs MBer,

 from which

‘ . Q _

X = - - F

Mar * %er, z_s‘ NpeF,

X = moles Li.F returned to reboiler/unit time from plate 1.

A material balance on BeF2 around the reboiler gives '

' QZ1

7\Li. BeF 2 Zg ?\Bng

. z
- FZy +FZ) = e
. 7\L:i.F BeF2 s ,7\BeF2

'Substitution of the definition of Q from Eq. 29 with the reletti.on~

 

 

- Zo‘
'_ZJ_ =__a[A
o | BeFo
into Eq. 33 yields
R + aBeF
2 - azo_ — R+l 7\L:LF * % 7‘13:-;@'2

“Bera [RBeF - ]
e ek BeF2 ?\LiF o xBng - R+1

A material balance around plate 1 for REF yields

 

 

 

" Fa “REF’ l)(7‘Luf.1? + 0 Fg Zg Npep,) * Q

(31)

(32)

(33)

- (34)

”'-‘(35)_

(36)

(R + 1)Fa Fxl ¥ QXl CEG e
7\LiF Bng s ?\Bng ' '
qa e
| REF _.

= .+ RFQ 1

R 7\L;I.F BeFa z 7‘BeF2 _REFX
 from vhich R

K= REF L

 
 

26

. A material balance around the reboiler on REF yields

 

dX_ Q.. X
"REF "8 - . ,
V=—>=FK_ - +Xx (37

de °  Mur * %ers Zs Mger,
which can be written as

& o
8 _F t; _ SR
it -y (Ko - FXQ) S B®

where

(Apyp + zol‘L ) O‘REF

2 * Frer Oeg~t -~ 1
R + 1 7\L1F+7\BeF2[z +R+l aBeFaz]

The solution to Eq._58, with the,condition X = X when t = 0 is

B =

 

8

X =-- [1 - (1 - B) exp (-Fat/v)] - (39)

from which the fraction of the total REF not vaporized is given by o

b = S5E 1L - (1 - 8) exp (-BEAL - £)] (ko)

 

where _
fREF = fraction of total REF fed to éystem whlch is contained
in liquid in reboiler
= fraction of total LiF ‘added to system.which is vaporized

Values for the fraction of REF not vaporized as a function of
‘the fraction of the LiF vaporized are shown in Fig. 8 for various
values of the rare earth fluoride relative volatility. Data used

in the calculation were as follows
ZO = 0.45 moles BeFalmole LiF
R=1

| rlaBEFg =5
kLiF = 53.8 kcal/mole

Ngep, = 50-1 keal/mole

 
 

 

 

 

27

 

 

 

 

 

 

 

 

 

 

 

oS
L ORNL DWG 7-26
R - S a001 |
0.96 p—
0.88}—
S
&
x
g .
O
=
&
£
& (.80 {—
L
Q
Z
6
g
g
0.72 b
S 0.965 ,--0.97.—,;‘:;_:;;_5 Sl 098 08 1.0

FRACTION OF LiF VAPORIZED

o Fig. 8. Rare Earbh Fluoride Removal in & Semicontinuous S5till
' Having One Theoretical Stage wit.h Rectification a.nd a Reflux Ratio
of Unity. . .

 
 

 

 

 

28

L. Batch Distillation

Consider 2 liquid of uniform composition consisting of components

1, 2, and 3 which have ‘molecular weights Ml, Ma, and Mg, respectively.

Let the weights of the components in the liquid at any time be Wl,

Wg, -and Ws, respectively.

For the vaporization of a differential quantity of
the number of moles of each component in the vapor thus
is d (—1§ (—2), and d (W )e' Since the mole fraction
1 in the vapor fs given by -

a(w, )

) o) o)

 

one can write

 

Yi ¥a
In the liquid
WafMa _Xa i _Za'd
= g = ! .
WM X M Xy W
Thus
My WM Viza Wy
Wz ya Mg y3 X W3

Then

d(log W;) ¥,/ o
d(log W3) x/xg ~ i3

the liquid
produced-

of component

)

(v2)

(43)

(bb)

_where ¢, is the relative volatility'of component i‘fiith'respect to-

i3
component 3. The quantitytxi3_is-defined as

afif (moles i/mole 3), vapor
i3 =

_ _ (mole frac i/mole frac 3), vapork,hs)
(moles i/mole 55:'liguid (mole frac i/mole frac 3) liquid'

 
 

 

 

 

E3l

| range of LiF recoveries since for o

 

29

Ifa13 is constant during the distillation, Eq. 44 can be integrated,

Thus if Qi and Q3 are the initial weights of»components.i and 3 in
the liquid, |

W, Wy
[ d( log Wi) = aij_ ! d( log W5)
W fw\ %3
"\

where the quantity W /Qi represents the fraction of component i
which remains in the liquid -1f the components 1, 2, and 3 are
now designated to be a rare earth fluoride (REF), BeF2 and LiF ;
respectively, |

R - ‘a
MREE i (WLiF REF
Qir/
- o«
VBer, _ (WLiF‘ ‘BeFz
Re™

The fraction of the rare earth fluoride not vaporiaed as a

function of the fraction of the LiF vaporiZed for various values

of aREf was calculated from Eq. 48 and is shown in Fig. 9. The;

- vaporization of BeF2 can be regarded as- complete for ‘the: probable

Be F ~LiF = 5.0, VaPDrization'

of 90% of the LiF results in vaporization of 99. 9% of the Bng._

h 5 Semicontinuous Distillation Followed by Batch Distillation'

(46)

(k)

(48)

(49)

Consider a still containing V‘ moles LlF If a feed stream o

consisting of F moles LiF/unit time which contains X moles

REF/mole LiF and which may also contain BeFp is fed to the still
'with ‘the condition. that the initial REF concentration in the
~still is X o’ the concentration of REF in the liquid at time t is

given by

 
 

1.0

0.96

0.88

FRACTION OF RARE EARTH FLUORIDE NOT VAPORIZED

0.72

- 30

'J -  ORNL DWG 67-27

 

 

a=0.001

-

@=0005

 

“a=QOI

 

      

a=0.02 -

 

 

0.965

0.97 - 0.9 _ 0.99 | S 1.0
FRACTION OF Lif VAPORIZED | . o

, Fig. 9. Rere Earth Fluoride Removal in a Batch Still as a
Function of LiF Recovery and Rare Earth Fluoride Volatility. =

 
 

 

¥

 

31

e

X=52[1-(1-a)exp (-E/Y)] (50
where

REF-COncentratiqn in still liquid, moles REF/mole LiF

n

REF concentration in feed, moles REF/mole ‘LiF
relative volatility of REF referred to LiF
feed rate, moles LiF/unit time

o QMM
"

= time

V' = moles LiF contained in liquid in still,

1f at time t the feed is stopped and batch distillation is carried
out on the liquid in the still to give a.final liquid.nolume con-
taining V moles LiF, the fraction of the REF which was present in
the still at the begrnning of batch distillation which remains in
the still is then (V/V') . The moles of REF at the beginning of
batch operation is VX 0 that the moles of REF in the still at

the end of batch operation is V'X(V/V ) The total moles of REF
fed to the system isu(Ft,+,Vq)xo-_‘Henceethe fraction of the totalrn
REF which remains in the still is

which can be written in the form_

m [l .'.L - Ot) exp (-rfl)]g - (52)
| fREF = fraction of REF which remains in still .
g - vV . .

= final LiF content ofiEtiifi,—noleem
'V = initial LiF. content of still ~moles -

n = Fe/V', number: of st111 volumes fed prior to batch operation
' Ei; feed rate to st111 during semicontinuous operation,

.' ' moles/unit time ;_¢;};~- s |
'tt = time , ia“ ) T ,

@ = relative volatility of REF referred to LiF.

 
 

32

The fraction of the LiF which is vaporized by both methods of

operation is given by

.ij_F— 1 n +1 , L - (53)
where
fLiF fraction of LiF vaporized by both methods of distillation.

The fraction of REF not volatilized as e function of the fraction
of the LiF volatilized for ¢ = 0.1 (final volume of 10% of still
volume) is shown in Fig. 10 for various values of the REF relative
volatility, " |

4.6 Comparison of Methods Considered

A comparison of the various distillation methods must take

account of numerous factors such as separation potential

operability, economics, etc. Consideration of all of these factors

is beyond the scope of this report, however, the two topics of

separation potential and operational simplicity will be discussed

For currently envisioned processing rates, the required removal.
efficiency for the rare earth fluorides is approximately 90% for
the more important neutron poisons (Pm, Nd, Sm) and less,for other
members of this group. It should be recalled that the real
requirement is the rate at which neutron poisons are removed from

the reactor system, hence the "required removal efficiency"” can be

-lowered at the expense of an increased throughput in the processing

plant. With this in mind,!a rare earth‘fluoride;removal'efficiency
of 90% will be used as a basis for discussion of separation potential
for the various distillation methods.

As shown in Table 1, recent measurements indicate the relative
volatilities (referred to LiF) of some of the fore important rare |
earth fluorides to be 6 x 10 % for NdFs, 5 x 10 % for SmFs, and
3 x 10 # for LaFs at 1000°C in LiF, The relative volatility of CeFg
was found to be 3 x 10-3,.however; the required removal effiCienoy.' o
for CeFs is only 84.

 
 

 

 

33

 ORNL DWG 67-28

 

 

 

 

 

 

 

| . a=0001
o
w
N
g .
O 0.96 | —
Z |
5
&
g a*0.005
T
- .
Ed
. A
w -
« 0.92 p— -
.
O
z
0
0
< , |
E L S . .
- - a=0.01
088 1 . ,.J
- 0965 057 098 :'*r_0.99»:-_:~ 1.0
FRACTION OF LtF VAPOR!ZED |
Fig. lO. Var:!.ation of Rare Earth Fluoride Removal in & Semi-

- continuous Still Using a Final Batch Volume Reduction to 10% of the
Still Volume. C ,

 
The fraction of REF not vaporizéd for the various distillatibn

~ methods considered is given in Table 2. The values were calculated

for a REF relative volatility'of.fi.x'10f4-and.a'LiF'recovery of
99.5%. As would be expected, the éontinuous system yields_the
-‘pporest-REF'removal.(91%) although this is an adequate removal
efficiency. The removal efficiency fof a semicontinuous system is.
somewhat higher (95.2%). The combination of semicontinuous dis-
" tillation followed by batch operation to yield a 10% heel results-
in a REF removal of 99.5% which is comparable to that obtained with
'batch'bpération (99.744). The effectiveness of rectification is
~ pointed. out by the REF removal (99. 99674) for a semicontinuous f ;
system containing one theoretical plate in addition to the reboiler.i

Table 2. Fraction of Rare Earth Fluoride not Vaporized for Several;
Distillation Methods for a REF Relative Volatility of 5 x 10 ¢
. and a LiF Recovery of 99.5%.

 

Distillation Method Fraction REF not Vaporized

 

Continuous 0.91

-Semicontinuous 0.952
Semicontinuous + batch (10%) heel 0.995
Batch 0.9974

Semicontinuous with Rectification 0.999967

 

 

The continuous distillation system is considered to be the -
simplest operationally. This system will operate with a near
constant heat load (due to fission product'decay) and liquid level
vhich will allow prediction and control of‘concentfafion-and‘ f '
temperature gradients in the liquid phése and will simplify‘instfu-
menting the system. Waste salt could be removed from~the?systémrat

.frequent intervals rather than continuously,

The semicontinuous system is considered to be slightly more

complex operationally than the continuous system; this complexity

 
 

 

35

arises primarily from the,transient“nature.ofhthisimode of operation
and the associated variation of.the‘fiSSion'product°heat load with
'time. Reduction of the liquid volume in the system near the end of
a cycle by batch operation will not complicate instrumenting the ”:
system if the heel volume is not less ‘than approximately 10% of the
initial still volume. - | | fi

Operation of a batch system W1ll be significantly more compli- _
cated than either of the above systems for a number of reasons.
The cycle time for this system should be relatively short in order
to maintain an acceptably low salt 1nventory, hence the system must
be charged and discharged frequently. Control of temperature and 7
concentration gradients in the liquid will be complicated by the
continual variation of both liquid level and heat generation per
unit volume, Near the end of a cycle, the liquid volume in the ”i'
system will be approximately 0., 5% of the initial volume° accurate
-measurement of liquid level at’ this point may be difficult '

Conceivably, a continuous distillation system employing
~rectification would be as simple operationally as a- ‘continuous
‘system without rectification._ However, a system using rectification_
probably requires a greater extension of present technology than any
of the systems considered Problems 'such as vapor- liquid contact
are aggravated by the necessity for low pressure, high temperature
operation. The very high rare ‘earth removal efficiency achievable

with rectification is not required for MSBR processing.ii:}

Besed on the factors considered the distillation methods can

be listed in order of decreasing desirability as follows-'-

51,,1continuous . . R i -
7"'2.”_semicontinuous With batch reduction to yield a 10% heel
- 3., semicontinuous S e '
4. batch ) L -
.5.' continuous with rectification’h

 
 

. |
i
i
1
|
A

36

5. PREVENTION OF BUILDUP OF 'NONVOLATILES AT A VAPORIZING SUREACE S 7 .
: BY LIQUID PHASE MIXING , o

During.vaporization of a multicomponent mixture, materials

 

less volatile than the bulk of the mixture tend to remain in the‘.v S |
liquid phase and are removed from the liquid surface by the processesjfiqq

-of convection and molecular diffusion. As noted in Section 11, Low

pressure distillation will result in little if any convective mixing

in ‘the liquid and ‘an appreciable variation in the concentration of

materials of low volatility is possible if these materials are " '_

~ removed by diffusion only. The extent of surface buildup and the -

effectiveness of liquid phase mixing will be examined for a continuous |

still although the phenomena 1is common to all types of stills.-

, Consider a continuous still of the type shown in Fig., . Fuel
carrier salt. (LiF-Bng) containing fission product fluorides is
fed to the bottom of the system continuously. Most of the LiF-BeFp
fed to the system is vaporized and a salt stream containing most _77
of the nonvolatile materials is withdrawn continuously. The S
positive x direction will be taken as vertically upward and the
liquid withdrawal point and the liquid surface will be located at _
x = 0 and x = £, respectively. Assume that above the liquid
‘withdrawal point, molten LiF containing rare earth fluorides (REF)
flows upward at a constant velocity V. At the surface, -a fraction |
v/V of the LiF vaporizes and the remaining LiF is returned to the
bottom of the still, o

Above the withdrawal point, the concentration of REF satisfies |
.the relation ' ' |

dec dc
DaxZ "~ Viax

oo
and the boundary conditions are: |

at x = £

-D = + VC_ = V¢ ; (Vv - v)C (55)
x = 8 s 8 :
vk

 

Recirculating.

Salt
V-V, c

. \.Figo- llo

 

 

_ ORNL DWG 67-29

| ____» Produ

 

X

T x=t
I
|
l

ct Salt

 

 

 

 

 

 

Wnthdrowal
Point

> Dlscord Salt

F-v, C

Continuous Still Having External Circulation and &

Nonuniform. Liguid Phase _Rare Earth Fluoride Concentration Gradient.

 
 

38

and at x = 0

-p £ Ix‘=‘ UG = (V- WG FC - (Fov)e,  (56)
where o , o
D= diffusivity of REF in molten salt of still pot eoncentretion,'
( cmzlsec A o - -

C.= concentration of REF in molten salt at position x, moles
REF/cm® salt | |

X = position in molten salt‘measured from liquid withdrawal
point, cm ' '

V = velocity of molten salt with respect to liquid surface,
cm/sec ' ' '

-cs;= concentration of REF at X = 3 moles REF/cm salt

‘em® LiF (1liquid) |
cm® vaporizing surface-.sec
O = relative volatility of REF referred to LiF

'co = concentration of REF at x = O, moles REF/cm® salt

Cs

Equation Sk has the solution

FC {1 -3 (1 - @) [1 - exp(-v(¢ - X)/Di

v.= LiF vaporization rate, .

= concentration of REF in feed salt, moles REF/cma salt .

 

 

v
1-3(1- @) [L - exp (-v2/D)]

 

C(x) = (57T
W + (F - v) {1 -¥(1-a)1 - exp(-Vfl/D)]}
The fraction of the REF removed by the still is
(F - v)C
fraction REF removed —_—
: 'FCf
= = — o (8)

F -v +

‘ The fractional removal of REF for a continuous system hav1ng a
perfectly'mixed liquid phase was derived earlier and is given by
Eq. 23. The ratio of the fractional removal of REF in a system

 
 

 

 

29

‘having a nonuniform_concentration to that in a still having a
uniform concentration will be denoted as ¢ and can be obtained by
- dividing Eq. 58 by Eq. 23.  Thus

:;li+ ¥

 

<

a
i
i

~ (59)

8!

 

i

-
1 - %’-.-(1 = a)[1 - exp(-V4/D)]

1+

Values of ¢ calculated for"a still in which 99.5% of the LiF
fed to the still is vaporizedj(v/F - v = 199) and in which the |
relative'volatility.of'REFis.S.xth-* are given in Fig. 12. The
following two effects should be noted: -’

1. The value of ¢ is essentially unity for V£/D < 0.1 for
any value of v/Vf(fraction of LiF vaporized per circu-
‘lation through'still) Within this region, a near
uniform REF concentration is maintained by diffusion
of REF within the liquid and mixing by liquid circu—
lation is not required.

2. The value of ¢'is?strong1y dependent on v/V for
- VE/D > l.-7Within thisrregion, a near uniform REF
concentration can beJmaintainedronlygif liquid circu-
‘lation is-provided" For V£/D = 100, @ has a value -of
0.0055 with no 1iquid circulation and a value of 0. 99
1f 90% of the LiF is returned to the bottom of the still

h,fiAn actual still would probably operate in the region Vfl/D > 1lso

'.that the importance of liquid circulation can not be over emphasized
Liquid phase mixing by circulation is believed to be an essential

B feature of an effective distillation system. In the case . considered -

*.circulation was .provided by an external loop for mathematical con-t

'*'Vvenience. In an actual still internal circulation could be prov1ded

 

 

 

by a toroidal liquid flow path which for a liquid having ‘a strong
volume heat source (provided in the present case by fission product

decay), would result if more cooling were provided to the still in an

outer annular region than in the center.

 
 

 

. 'ORNL DWG 67-30

 

    

 

 

 

 

 

 

‘.0"- 7111[ \\_l I I3 .
o v | -
~ NG =01 :

Nyt cos i

h v 0.9

DLI S o
0.0' — ———
- \V 'nr'- ' -

- v .
0.001 L1t 1 ratal 1 P 1 1’1'5' L 4 ;'1'i‘t
KX v w0 D 100

D

Fig. 12. Ratio of Fraction of Rare Earth Fluoride Removed in
Still Eeving Nonuniform Concentration to That Still Having Uniform
Concentration. | | '

 
 

 

41

6. CONCLUSIONS ‘AND .:RECOMENDATIONS

" The following conclusions. have been drawn from the information

considered in- this report-

1.

Distillation"at'1ow.pressure.allows.the‘adequate"renoval

- of rare earth fluorides from MSBR.fuel salt and the
~simultaneous recovery of~more than 99.5¢ of.theafuel salt.

Recently reported relative volatilities of several rare

- earth fluorides allow a great deal of latitude in still

design and‘operetional mode. A single contact stage such
as a liquid pool is adequate and rectification is not

necessary.

The effectiveness\offa distillation system can be'seriously

decreased by the buildup of rare earth fluorides at-the

surface of the vaporizing liquid.

Liquid circulation can provide adequate liquid phase

- mixing and is an essential feature of an effective dis-

tillation gystem.

The following recommendations ‘are made-

jlo

Further consideration should be given to the use -of single-

stage,.continuous distillation for removal of rare.earth

- fluorides from=the fuel~stream of an MSBR. -

dThe ‘study of liquid phase temperature and concentration

'profiles ‘should be extended to stills having configurations

of interest for‘MSBR processing. ‘Methods :should be

gdevised for the calculation of velocity, temperature and

concentration in the liquidiphase of a three dimensional i,

-still'havingVa_distributed-heat source, The effect of

.'_rvariations;in heatigeneration rate ‘should be considered.

 
3. Removal of fission product decay heat from a distillation

42

system should be étudied. Heat removal_systems which

‘maintain the temperature within acceptable limits in the

event of failure of the primary cooling-systém-should;be?v,<i

 devised,

Factors which limit distillatibfi’raté1such~as opefétiné

- pressure, the presence of noncondensables, and entrainment

- should be examined.

 
S S P

WA

 

b3

REFERENCES

P. R. Kasten-et al., De sign Studies of 1000-Mw(e) Molten Salt
Breeder Reactors, USAEC Report ORNL 3996 Oak Ridge National
Laboratory, August 1966.

L. B. Loeb, The Kinetic Theory of Gases, McGraw-Hill Book Co.,
New York, 1951; p. 98 |

G. Burrows,‘Molecular Distillation, Clarendon Press, Oxford,
1960, p.‘20.~ '

Oak Ridge National Laboratory, Molten-Salt Reactor Program
Semiannual Progress Report for Period Ending February 1966,
USAEC Report ORNL-3936, p. 130.

Oak Ridge National Laboratory, Reactor Chemistry Division Annual
Progress Report for Period Ending December 31, 1965, USAEC Report

- ORNL-3913.

10.

Oak Ridge ‘National Laboratory, Unit Operations Section Quarterly

 

~ Progress Report for Period Ending September . 196 USAEC Report
o| RNL-LOT5. T

Oak Ridge National Laboratory, Molten-Salt Reactor Program
Semiannual Progress Report for Period Ending August 1966, USAEC

Report 0RNL-403T

 

Lim, M. J., Vapor Pressure and Heat of Sublimation of Cerium III

;Fluoride, UCRL-16150, University of California, 1965

Mar, R. W., Vaporization Studies of Lanthanum Fluoride, UCRL-l66h9,
University of California, 1966

Zmbov, K. F., and J. L. Margrave, Mass-Spectrometric Studies at -

High Temperatures. XI., The Sublimation Pressure of NdF. and
“the Stabilities of Gaseous Nng_and NdF, Journal of Chemical

,,]Physics, ks, 3167 (1966)

 
 

 

 

 
 

h

O O3 O\ W

-

29.

31 »
32,
33.
3,

35.

3,7-7
38,

29.

1,

L2,

b3,

b,

hg,
"h6

 

45

DISTRIBUTION

VMSRP Directors Office, 9d0h-l -

G. M. Adamso_n

C.

S.
E.
R.
- F.
E.

R.

‘R.

S.
W,

WO"#F’JMM

F.

E.

Baes
Beall = .
Bettis
Blanco
Blankenship
Bohlman

. Briggs
Brooksbank

Cantor

L.
I.

H..
L,
E.
M.

A,

R.
E.
N.
R.

W.

R.
J.

. Je

R.
T.

S,

B.

G.
F..

E.

1.
L. N

M.

Je

f ‘R.
,E..

S.

M.
. E.r.

Carter

Cathers
Cook

Culler, Jr,

Ferguson
Ferris

Friedman
E.
Grimes

Goeller

Guthrie
Habenreich
Hightower, Jr,

‘Horton

Kasten

Kedl
Kelly

Kennedy

Kerr
Kirslis
.Lindauer
‘McPherson

McDuffie
McNeese -

.Moore

icholson
Perry C

. Scott
’ .Ho
E.

'Schaffer
rSchilling
.Skinner _
‘Tallackson

Thoma
Watson -

Weinberg
Whatley

 
 

h7-48,

49-50

51-53.
5h.
55.

46

DISTRIBUTION (contd)

Central Research Library

. Document Reference Section
Laboratory Records

Laboratory Records~RC

Research and Development Div (ORO)

56-T0. DTIE

. C