OAK RIDGE NATIONAL LABORATORY

OPERATED BY

For Internal Use Only

 

UNION CARBIDE CORPORATION

NUCLISION 0 R N L
CENTRAL FILES NUMBER

OAK RIDGE, TENNESSEE 37830

 

68-1-42

 

 

 

DATE: January 214., ]_968 COPY NO.

SUBJECT: Protactinium Removal from Molten Salt
Breeder Reactor Fertile Salt

 

 

TO: Mo w. ROSenthal

FROM: J. S. Watson
M. E. Whatley

ABSTRACT

A study has been made of promising flowsheets for removing
protactinium from the blanket of a two region Molten Salt Breeder
Reactor by reductive extraction into liquid bismuth. Although
none of these flowsheets have been adequately tested, the
equilibria data presently available suggest that a relatively
simple and economic flowsheet can be used. 1If interest in two
region molten salt reactors persists, further development of
the reductive extraction process to obtain better equilibria
data, and to demonstrate its engineering feasibility is
recommended. |

NOTICE This document contains information of a preliminary nature
and was prepared primarily for internal use at the Oak Ridge
National Laboratory. It is subject to revision or correction and
therefore does not represent a final report. The information is
only for official use and no release to the public shall be made
without the approval of the Legal and Information Control Depart-
ment of Union Carbide Corporation, Nuclear Division.

 
CONTENTS

ABSTRACT . . « « « ¢« o & &
INTRODUCTION . . .
DESCRIPTION OF THE FLOWSHEETS.

EQUILIBRIUM DATA . . . . . . .
RESULTS . . . « ¢ . « « + .
CONCLUSIONS AND RECOMMENDATIONS.
REFERENCES . . . . . . . . . . .
DISTRIBUTION . . . . « « .+ « .

S
O W = e

18
27
30

31
INTRODUCTION

To obtain a high breeding ratio in molten salt reactors, it is
necessary to maintain a low protactinium concentration in regions of
high neutron flux to avoid capture by the protactinium before it
decays to =33y. As presently conceived, a molten salt reactor will
use a two-fluid concept with the blanket stream circulating in
separate channels through the high flux core region, as well as
through the "blanket region" outside the core. The protactinium
concentration in this blanket stream can be kept low in two ways.
First, the blanket salt can be processed rapidly to remove protac-
tinium soon after it is formed leaving little time for neutron
capture. Or secondly, the total blanket salt volume can be large
so that any sample of blanket salt spends only a small portion of
its circulating cycle within the high neutron flux of the core region.
The first approach adds additional processing equipment capital and
operation costs to the reactor system, while the second approach
requires larger salt inventory and storage costs. In selecting the
proper reactor and processing system, the relative costs of these
approaches must be compared with each other and with the economic
penalty of a lower breeding ratio associated with a higher protac-
tinium concentration. This penalty increases with increasing
protactinium concentration, but process (or blanket salt inventory)
costs decrease with increasing concentration. Thus, an optimization

of the protactinium concentration is needed.

The purpose of this memo is to present initial results calculated
for protactinium removal by one process, reductive extraction into
liquid bismuth. This is not the only processing scheme which has
been considered, but at this time, there are more data available to
suggest feasibility of reductive extraction than any other process.
This memo presents only calculations of concentrations and flow
rates for a given group or class of flowsheets. If interest in two
region MSBR's persists, economic optimization of these flowsheets

will be made as cost data become available.
DESCRIPTION OF THE FLOWSHEETS .

The basic flowsheet chosen for this study is shown in Fig. 1.
Two possible modifications shown in Figs. 2 and 3 are also considered.
The flowsheet in Fig. 2 is a modification using two extractors (or
salt metal contactors) but requiring a smaller reducer and possibly
a smaller decay tank. This is the preferred flowsheet. Figure 3
shows a modification of the process which could be used if develop-

ment of a reliable reducer proves more difficult than expected.

In the basic flowsheet, shown in Fig. 1., a salt stream from
the blanket (labeled stream 1) is contacted with a liquid bismuth
stream (labeled stream 3) saturated with thorium metal (approximately
0.003 mole fraction). The bismuth contains a particular lithium
concentration such that no thorium or lithium will transfer between
the metal and the blanket salt. The protactinium concentration in
both phases is much lower than the concentrations of lithium or .
thorium, so protactinium can thus be treated as a minor component
not affecting the lithium-thorium equilibria. 1If the proper lithium
composition is chosen in the metal, then no significant quantity of
lithium or thorium will transfer between the phases (i.e. be added
or removed from the blanket). This is a desirable condition because
with the high processing rate required for the blanket (approximately
two blanket volumes per day is considered a likely processing rate)
any significant readjustment of the blanket composition will be

expensive.

Protactinium however, does transfer to the metal phase. Thus
the salt stream 2 which leaves the extractor and returns to the
blanket has a lower protactinium concentration but essentially
identical lithium and thorium concentrations as the blanket salt.
Likewise, the metal stream 4 leaving the extfactor differs from
stream 5 going to the extractor only in its higher protactinium

concentration.
 

ORNL DWG 67-9765

  
     
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fo (Salt) |
X2Li X5pPq
X Kot
- OFF-GAS 4-—-7 325 ThFg
28e ' STh MAKE-UP
@ ’:':'::-:-@
Fe (Salt
X66(PG | F~ (Salt)
X6Li Fo i?Pg S
o Fg(Salf) |
X8Pa |
X8Li
) XgTh
® [
F (Salt) Fa(Bismuth) 3
X1Pa X4paq F3(Bismuth) Fg(Salt)
XL XaLi X3 P _ X9Pa
X1Th X4Th X3Li X9Li
X1Be X3Th X9Th

 

 

 

Fig. 1 Basic Flowsheet.
ORNL DWG 68-787

F2 (SALT)

 

Fa (METAL

 

-

Fs (SALT)

 

   
   
     

   
 
 

 

 

 

     
 

 

 

 

 

 

 

 

 

 

H OXIDIZER
x
Q UF
O | 6
g
(0 ot
-
- X
BLANKET w
SALT
DECAY |p——eet| &
Fi (SALT) Fq(METAL) TANK =
FERTILE BLANKET REDUCER <
SALT x
F7 -2 (SALT) ¥ 2
L @ F2
H
x
O
f-
O
- §
1 o
= (SALT)
x F7
I Fq-2 (METAL)

 

 

Fig. 2 Modified Flowsheet.
ORNL DWG 68-788

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

F3 (METAL) Na F ADDITION
Fs5 (SALT)
: _-OXIDIZER
O
5 UF6
- §
(1 e
-
>
BLANKET w
SALT DECAY || &
, O
Fi (SALT) | , TANK :
FERTILE BLANKET <
SALT 'REDUCTANT a
. ADDITION K~_7F_,/ o
TO WASTE ceme] 3
F7-2 K Li° The L je—F2
(8 o
O
-
O
g
(8 o
»
5 e F7 (SALT) A
F4-2 (METAL)

 

 

Fig. 3 "Throw-away" (or no reducer) Flowsheet.

 
The metal stream from the extractor contains protactinium and
flows to a hydrofluorinator where it is oxidized either electrolytically
or with an HF-Hs> mixture to convert all of the thorium, protactinium,
and lithium to fluorides. Bismuth is not oxidized by this process
and is recycled to the extractor after the proper amounts of thorium
and lithium reductant are added. This reductant may be added
electrolytically by reducing all or part of a recycle salt stream

from the decay tank (stream T).

The salt mixture formed by oxidation of stream 4 can be stored
for decay of the protactinium and eventually fluorinated to remove
the resulting 233y. However, this mixture would have an undesifably
high melting point. The melting point is high because the mixture
has a substantially higher thorium composition than the near eutectic
mixture used in the blanket. To lower the melting point it is
necessary to add lithium to the mixture. This is accomplished by
recycling salt from the decay tank through the reducer and then into

the oxidizer. This recycle stream is labeled stream 6 in Fig. 1.

One disadvantage of the flowsheet in Fig. 1 is the large amount
of protactinium which must be recycled to the reducer from stream T
and thus enter the metal (stream 3) returning to the extractor. A
high protactinium concentration in stream 3 limits the fractional
protactinium removal per pass through the extractor since the salt
leaving the extractor and returning to the blanket will have a protac-
tinium concentration equal to or greater than that required for
equilibrium with stream 3. One method of reducing the protactinium
concentration in stream 3 is illustrated in Fig. 2. The metal stream
from the extractor (stream 4) is contacted in a second extractor with
the salt stream recycled from the decay tank. This transfers most
of the recycled protactinium in stream 7 to the metal in stream L
which is subsequently oxidized. The protactinium then returns to
the decay tank without entering stream 3. This appears to be a
superior flowsheet. The additional complication of a second extractor
appears to be worthwhile due to the reduced protactinium recycle
and/or decay tank size restrictions. Both flowsheets will be discussed

in more detail in the following sections.
The second modification of the basic flowsheet is shown in Fig.
5. 1In this flowsheet, fresh lithium-7 and thorium metal are added
to bismuth to form stream 3; no reducer is used. The metal stream
from the first extractor is again contacted with a recycle from the
decay tank, as in Fig. 2, before going to the oxidizer. The recycle
salt stream in this case is subsequently discarded to waste. To
reduce the liquidus temperature of the decay salt, lithium fluoride
is added to the oxidizer. Since this lithium will not return to
the reactor blanket, natural lithium can be used. Also, if it will
not interfere with protactinium transfer in the second extractor,
another alkali metal fluoride, e.g. sodium fluoride, can be used.
Some important considerations in this flowsheet are reductant
composition control and discard losses of thorium,lithium~T7, and

protactinium.

EQUILIBRIUM DATA

The distribution of Th, Li, and Pa between molten fluoride
blanket salt and liquid bismuth has been studied by Shaffer and
M.oulton.l Their data are reported as apparent reduction potentials,

t

EO. The exchange of two metals, M; and M-, between the metal and

salt phases may be described as follows:

camem— Tm——— i — 1 f— M ]-t ]‘
V1 M5 Fvl(Salt) + v M (Bl) <—V M (Bl) + v 2 EVg(Sa ) ( )

where v, and Vo are valences of metals M; and M. The difference

between the apparent reduction potentials of M; and Mo may be defined

 

as
1 1 1
Vi Vo - -
e g JBLg Lu(Bi) “p(salt) _ BT -D-l—\i (2)
0 o F 1 1 ~ F 1
1 2 Vi Vo 5 Ve
Xl(salt) Xg(Bi) L 2 .

 

 

 
10

X, ...
and Di =  1(Bi (3)

xi(salt)

where X = mole fraction

R = gas constant
F = Faraday's constant
D = distribution coefficient

Moulton and Shaffer established a standard value for the apparent
redfiction potential of one metal (Li), and then from measurements
of the distribution of all materials of interest, they assigned
values for these metals. The results are summarized in Table 1 for
measurements made at 650°C. The differences between these numbers,
not their absolute values, are important to this study, so the
choice of a standard state is not important for our purposes. The
apparent reduction potentials differ from the reduction potentials
usually defined because mole fractions are used in the place of

activities.

Table 1
!
REDUCTION POTENTIALS ( Eo) FROM
MOLTEN BLANKET SALT TO LIQUID BISMUTH

"y

 

 

!

Metal E (volts)
Li '1 080
Th "'1 o)'l'T
Pa -1 032
U -1 028

 

 

CALCULATIONS OF FLOW RATES AND COMPOSITIONS

To evaluate the relative merits of these three flowsheets, the
flow rates and compositions of all streams were calculated under a

wide range of conditions. This section describes the material

 
‘4

11

balances and equilibrium relations used. For these calculations,

the lithium and thorium compositionsimlthé blanket were fixed at

0.72 and 0.28 mole fraction respectively, and the protactinium
generation rate was fixed at 10.6 g moles per day. (This corresponds
to a 2220 MW(t) reactor.) Protactinium was always considered a
minor component of the system and assumed to not affect the thorium

and lithium.

With the above assumptions, there are five remaining independent
variables in the basic flowsheet which are under our control. The
five variables considered in the study of the first flowsheet were
processing rate (labeled Fy), concentration of protactinium in the
blanket (XlP), number of stages in the extractor (N), composition
of protactinium in the bismuth stream to the contactor (X P)’ and
the lithium concentration in the decay tank (X L). This last
variable determines the liquidus temperature o? the decay salt.
Several other combinations of five variables could have been selected
to describe this system. These particular variables were selected
primarily because they appeared to provide a straight forward approach
to the calculations. The thorium-lithium composition of the blanket
was not treated as a variable because the liquidus curve prevents
substantial deviations from the eutectic composition without

operating at higher temperatures.

The lithium~-thorium concentrations in the metal stream to the

extractor (X
sL

lithium-thorium composition of the salt (X I and X T). Thus, with
1 1

and X T) were adjusted to be in equilibrium with the
3

protactinium being a minor component, no significant change occurred
in the lithium-thorium concentrations in the blanket or in the metal.
The metal stream was assumed to be saturated in thorium (approximately
0.003 mole fraction). The lithium content of the metal was then

calculated from Moulton and Shaffer's equilibrium data at 650°C.

1/h
_ _ 0.003
Xp=X1= 0.016 X L (=% T) : (L)

 
12

then
X =X _ =0.00 5)
F2 = Fl = FB 9 (6)
and F3 = F4 = EM J (7)

The distribution of protactinium between the salt phase and a metal

phase with this composition is:

D = “Pa-metal = 21.1 (8)

XPa-salt

This performance of the extraction column was calculated in a manner
suggested by Foust, et ;,11...;2 for conditions where the equilibrium

distribution coefficient is a constant.

 

 

XlP - X.2P E N + 1 - E (9)
X .~ _N+1
X _aP E - 1
1P D
where EM
E = (=)D (10)
FB

and N = number of stages in the extractor.

This equation was solved implicitly for E, and the metal circulation

rate was calculated as:
F = FB B (11)

The flow rate from the hydrofluorinator and subsequent down-

stream rates were calculated in the following manner. A lithium

 
15

balance around the hydrofluorinator gives:
F X_. =F _X_ +F X : (12)

Note that stream 6 is in equilibrium with the metal stream leaving
the reducer. Since this metal stream is also in equilibrium (in
lithium and thorium) with the blanket, stream 6 has the same lithium-

thorium composition as the blanket, e.g.

X =X, - (13)
Solving equation (12) for F
6

F X - F
F = 5 5L MX&L (l)-l-)

An overall salt balance around the hydrofluorinator gives

 

 

 

F +F (X_ +X_)=F (15)
6 M»( 4L 4T) 5
Substituting for F , this becomes
6
FX_. -F
— - 5 5L SL 1
Fs = FM(X4L x4T) + % (16)
1L
Solving equation 16 for F ,
5
X L
F(X_  +X, -2
L T X F |[X (X -1) +X X
F o= T L 2y m Y T XKy (17)
5 X X, -X L
] - 5 1L 5
X
1L
1L

Since stream 6 is in equilibrium with stream 3,

XGP = XSP/D :

An overall balance around the reducer shows that

=F -F (X . +X
Fe 5 M( 4L 3T)

A protactinium balance around the oxidizer gives

F X _ +FX
=P F ’
5

and a protactinium balance around the reducer gives

+
FM XaP Fest
X}P = T .
5

 

The fraction of protactinium entering the decay tank which

decays before going to the reducer is defined as F

X p - X p
F. = —————1
R X
sP

The gram moles of salt in the decay tank is then

F5 FR

MD - KPa(l - FRj

where A q = protactinium decay constant = 0.025 days-l.

P

thorium composition in the decay tank will have to be near the blanket

composition to maintain reasonable melting temperatures, so the

R’.

(18)

(19)

(20)

(21)

(22)

The lithium-

density of the decay tank salt can be assumed to be close to that of

the blanket salt. Then the volume of the decay tank can be estimated

L1
15

as

M (23)

when MD is in gram moles and VD in cubic feet.

In calculating the flow rates and compositions for the modified
flowsheet (Fig. 2), it is not convenient to specify XSP since this
value is quite low. Instead of specifying XsP and XSP, we chose to
specify F and V. Again any five independent variables could have
been chosgn including the combination used in the basic flowsheet

calculations. But at this point, it seemed desirable to change.

A simplifying assumption was made in the second extractor. No
lithium or thorium was allowed to transfer between phases. This is
a satisfactory assumption if the decay tank composition is essentially
that of the blanket salt. The requirement of a low melting salt in
the decay tank will force us to operate where this assumption is
justified. Calculations were made for only one stage in the second
extractor. Any number of stages may be used, but these results will

show that one stage is adequate for most purposes.

Calculations of the metal flow rate and compositions in the first
column were made just as they were made in the basic flowsheet, except
a lower value was used for XSP. As a first guess XsP was assumed to
be zero and EM and X4P then calculated as before. The protactinium
concentration in the stream to the decay tank can be calculated by
noting that at steady state, the rate of decay of protactinium in
the decay tank must be equal to the rate of extraction in the first

column. Or
FM(X4P - st) = FS(XSP - X7P) : (24)

Dividing both sides of equation 24 by F and solving for X p slves
5 5

e
16

From the definition of F_ given in equation 22, one can thus evaluate

R

X, =X _(1L-F,) .
P SP( R)

(26)

A protactinium balance around the second extractor gives

(X p, = Xp)

For a single stage (second) extractor

X = X D .
7P2 4P2/

Combining these last two equations and solving for X

 

F
M
X , +X 5 (=
x o zP " T4P (7
7P> DF,,
1+ ——
F .
5

(27)

)

7P2’

(29)

The salt rate to the oxidizer can be obtained from an overall

balance around the reducer

F =F -F (X, +X . 30
= F_ - E(X g +X ) (30)
A protactinium balance around the reducer gives
X P

F X = F +FX _=F + F ===, 31
5 7P2 aP e ef aP e D (51)

Combining equations 30 and 31 to eliminate F and solving for X p

5 3
F X

X - g (52)

aP

z”.’l
+
Ub’;‘.’l o

&)
tN

L7

In calculations of the metal rate and compositions in the first
column, X p was assumed to be zero. TIf the new value.just calculated
for X3P is large enough (relative to X4P) to affect the calculations
for the first column, those calculations were repeated using the new

value of X . The entire process was then repeated to obtain a new

3P

value of X p to be compared with the previous result. This process
3

was repeated until X _ was known accurate enough to determine the

sP
performance of the first extractor.

Once an accurate value for X p was obtained, the remaining
3
compositions could be evaluated directly as follows:

X p = XBP/D , (33)
X4P2 = X4P + FS(X7P - X7P2)/FM , (34)
and X5L = (Fs XlL + FMX4L)/F5 f (35)

Flow rates and compositions in the "no reducer" flowsheet shown
in Fig. 3 may be estimated from a special case of the second flowsheet.

That is when X p is zero (or essentially zero) and when F has the
3 5
proper value. If the decay tank salt is to have the same liquidus

temperature as the blanket,

XST
5
1T

The protactinium loss is that contained in stream F , e.g. F X .
7 -2 5 7P2
The fraction of protactinium loss is the loss rate divided by the

production rate or

F X p
Pa loss = _ETE%EZ' . (37)
18

The lithium-T and thorium losses will be respectively (in g moles/day): :
Li-T loss = F X (38)
g sl
and Th loss = F_ X . - (39)
3 3
RESULTS

These relations were coded for automatic computation on the
CDC-1604 computer, and the results are shown in Figs. L4 through 9.
All of these calculations were based upon a single processing rate
of 5.81 x 10® g moles/day of blanket salt. This isvh750 ft3/day
(25 gal/min) or approximately two blanket volumes per day.

Figures 4, 5, and 6 show the metal flow rate in the first
extractor given in g moles/day (multiply by 3.9 x 10 © to get gal/min)
as a function of the number of stages in the extractor. Three
different protactinium concentrations in the blanket were considered, :
5x 108 10 x 108, and 20 x 10 8. These results are shown
respectively in Figs. 4, 5, and 6. These blanket concentrations
may be compared with the value of approximately 80 x 10 © mole
fraction which would result in the blanket of a "no-Pa-removal"
reactor system which has been considered.5 In the basic flowsheet
(shown in Fig. 1), a substantial fraction of protactinium is recycled
back into this metal stream. -‘The recycled protactinium concentration
X _ is a parameter in Figs. 4 through 6. 1In the modified flowsheets

3P

X p is reduced to or near zero and the corresponding curves for
3

these flowsheets are also shown in these figures.

Raising the protactinium concentration in the blanket raises
the equilibrium (or maximum) metal loading and, for a given protac-
tinium generation rate, lowers the required bismuth rate. Also

lowering the protactinium concentration in the metal recycled to

the extractor, X p improves the extractor performance and lowers
3

the required metal flow rate. In the basic flowsheet, shown in
19

ORNL DWG 68-789

 

F, =58l x _1_96 g MOLES/DAY =PROCESSING RATE
X, =5.0x10 = MOLE FRACTION Pa IN BLANKET

 

536 x 105

4.02x 1075

2.68x 105

5 1.34 x 10~5

— 2 COLUMNS
_

 

 

 

l'-'M , BISMUTH RATE ,gMOLES/DAY

 

 

Lol | L Ll L1
| 2 3 4 5 678910 2 3 4 5 0678910

N, NUMBER OF STAGES

 

Fig. 4 Metal Rate for Processing a Blanket with 5 x 10 6
Mole Fraction Protactinium.
20

ORNL DWG 68 —790

 

 

 

 

 

 

 

 

 

10
8 -
-
6 hun
- F, =581 x10° _ QMOLES/ DAY = PROCESSING RATE
% 4F X p=1.0x 105 MOLE FRACTION Pa IN BLANKET
2
n
i
o
on
w \ .38 x10"¢
’—
g s
210 \ 1.035x10™¢
5 8 S~ 6.90 x 1078
=
» 6 3.45x10°5
@ :\ 2 COLUMNS
wt g4 L
2 -
IO‘» I oot | Lo ] L
| 2 3 4 5678910 2 3 4 56789I10°

N, NUMBER OF STAGES

Fig. 5 Metal Rate for Processing a Blanket with 1 x 10 °
Mole Fraction Protactinium.
Fu BISMUTH RATE, gMOLES /DAY

21

ORNL DWG 68-79I

 

| R I [ LR

T 11
!

 

 

 

 

 

 

 

 

6 -
— F,=58x10° gMOLES /DAY = PROCESSING RATE -

4 X 2.0x |0°= MOLE FRACTION Pa IN BLANKET -

3 " -

2 |- - | —

| _\ _
- 3.068 X 10~4 —

8 | _

6 \ _
B 2.30 X10"¢ _

q T .53 X 10" -
\ _
— —  7.67X107% —
T 2 COLUMNS

2 - | ~

| | ] ] 1 1 1 11l ] | | ] | S .
| 2 3 4 5678910 2 3 4 5 6789102

N, NUMBER OF STAGES

Fig. 6 Metal Rate for Processing a Blanket with 2 x 10 °
Mole Fraction Protactinium.
22

Fig. 1, this is accomplished at the expense of decay tank volume.
Increasing the number of stages in the contactor also improves the
performance of the extractor and lowers the required metal rate,

but Figs. 4 through 6 show that the advantages gained here will be
relatively small after two or three stages. In general, one can
expect metal flow rates to be in the neighborhood of 0.3 x 10° to

1 x 10° g moles/day or 0.1 to 0.4 gal/min depending upon which
flowsheet is adopted and upon the protactinium concentration accepted
in the blanket. This rate is important because it affects the
extractor design, but even more importantly, this rate sets the
reducer capacity (or lithium-7 and thorium consumption in the "no-
reducer" flowsheet). A relatively low metal rate is desirable since
reducer anode surface may be expensive; this is one of the reasons

flowsheets 2 and 3 are attractive.

The limited bismuth rates for a large number of stages are
summarized in Fig. 7. Here the metal rate is plotted as a function
of the protactinium concentration in the metal recycled back to the

extractor, X for each of the three blanket concentrations

2
3P
considered. The metal rate, or course, approaches infinity as X p
3
approaches the concentration which is in equilibrium with the

blanket salt, X The asymptotes are marked on Fig. 7. Similarily
1

there is a horizgntal asymptote for each of these curves which
corresponds to an X p of zero. These asymptotes are also marked.
The two extractor f?owsheets, Figs. 2 and 3, lower the recycled
protactinium concentration in the metal and thus reduces the bismuth

flow rate to, or essentially to, the horizontal asymptote.

The salt recycle rate to the oxidizer is determined by the
desired liquidus temperature. (The "no-reducer" flowsheet shown
in Fig. 3 is an exception since there is no recycle.) As noted
earlier, the salt which would be formed by direct oxidization of
stream 4 ( or L-2) will be richer in thorium than the reactor blanket

salt. 1In order to avoid a high liquidus temperature, this salt is
23

ORNL DWG 68-792

 

 

 

 

- ! ! | T 1 1 l i1 1 | i i 1T 1V 3 l_
- Fi=5.8/x10 gMOLES/DAY= PROCESSING RATE -
6 - N =10 STAGES / -
~ / | -
/
4 |- // ll -
x 3| / / -
g 2 o // / I, -
= -6 :
S Xp =5 X10 / _ I
oOn
LL.I. ,I
.5 -
EIO _-LTMIT AT X3p=0 / -
-8 r Xp =10°3 -
E - IP -
2 °r --—-“"‘-" -5 I
& [LIMITAT X3p=0 Xip = 2 X10 -
® 4 - —
Th - - -
CLIMIT AT Xap = O
2 - -
| | ] ] i 1 + 1111 1 ] ] 1 1 111

 

 

 

|0~ 6 2 3 4 56789

X3 p, PROTACTINIUM CONCENTRATION IN METAL RECYCLE

3 4 56789100% 2

Fig. T Metal Rate for a Large Number of Stages.

 
2L

mixed with recycle salt from the reducer which has the same composition,

in lithium-thorium, as the blanket. Figure 8 shows the decay tank
composition, X () as a function of the ratio of the salt recycle
rate, F5 or F f to the metal rate, FMf Obviously with an infinite
recycle rate,7the lithium composition in the recycle salt will only
approach the blanket composition. Thus it is necessary to operate
with a lower lithium composition in the decay tank than in the
blanket. (Again the "no-reducer" flowsheet is an exception to this
rule). This will probably mean a higher liquidus temperature for
the decay salt. Just how high the liquidus temperature is allowed
to go is a question of economics. ILowering the liquidus temperature
requires higher salt recycle rates and, at least for the basic
flowsheet shown in Fig. 1, a larger decay tank. Raising the liquidus
temperature increases corrosion and eventually creep rates which may
raise the cost of the decay tank because of thicker walls, more
expensive materials of construction, and/or shorter tank life. The
approximate liquidus temperatures corresponding to the various

lithium compositions are indicated in Fig. 8.

One disadvantage of the basic flowsheet, Fig. 1, is the coupling
of the decay tank volume with the recycle metal composition, XSP-
This connects the bismuth rate and thus the reducer size with the
decay tank volume. The required tank volume is proportional to
XQP-l as shown in Fig. 9. The use of a second extractor as
suggested in the modified flowsheet of Fig. 2 in effect uncouples
the metal rate and decay tank volume because the first extractor
operates near the horizontal asymptotes of Fig. T, and the metal

rate is little affected by X The size of the decay tank is

3P’
governed by heat removal considerations rather than by fractional
decay considerations. There will be 6.7 MW of heat generated in
the blanket from protactinium decay. This suggests a minimum decay

tank volume of a few hundred cubic feet.

The important considerations unique to the "no-reducer" flow-

sheet are lithium-7 and thorium consumption and protactinium losses.

o

[.44
 

a O N oW

25
ORNL DWG 68-—793

 

 

 

 

| 1 | 1 | 1 1 ] | 1 | L 1

 

O ~N oW

59 60 61 62 63 64 65 66 67 68 69 70 TI 72

X gL

Fig. 8 Composition of Recycle Salt.
VOLUME OF DECAY TANK (F13)

»

$H

10

26

ORNL DWG 68-794

 

 

   
   

 

 

B DECAY SALT 70%(MOLE) LIF

- _

_

_ 675% LI F

- 6 | 65% LIF
PROCESSING RATE =581x10° MOLE /DAY
L1 L 11l 1 L1l L1 1 Lt L L1l

1073 Lo 10-3

X3P

Fig. 9 Volume of the Decay Tank.

»)

"
27

The lithium-T7 loss is 0.0037 EM or 370 and 90 g moles/day (5.7 and
l.h.lbs/day) respectively for blanket protactinium concentrations of
5% 10 © and 20 x 10 © mole fraction. The corresponding thorium
losses would be 0.005 Fy, or 150 and 38 1bs/day. These losses would
be tolerable. With 40O cubic feet decay tank, the protactinium

loss would be less than 0.02% of that produced in the reactor.

CONCLUSIONS AND RECOMMENDATIONS

We have looked at three conceptional blanket processes and
evaluated the pertinent flow rates and compositions to study the
feasibility of each process. Although all three flowsheets appear
feasible and even attractive, no step in the processes has yet been
satisfactorily developed or demonstrated. The reduction step appears
most likely to present difficulty, so a "no reduction" flowsheet
was included in the study for an alternate approach. No materials
of construction for the salt-metal (and possibly HF ) containing
vessels has been demonstrated, and all of the equilibrium relations
used in this study need to be checked to obtain more reliability.
One parameter which has been tacitly assumed to be favorable is the
protactinium solubility in bismuth. This is expected to be near
or above the solubility of thorium (0.003 mole fraction), but no
data are available to confirm this. The highest protactinium
concentration reached in this study was 0.0004 mole fraction.

Hopefully, the solubility will be at least this high.

Despite these unknown or untested aspects of the flowsheets,
there are no indications that any of the steps are impossible. All
three flowsheets appear to provide promising approaches to blanket

processing for two region molten salt reactor systems.

The "modified" flowsheet shown in Fig. 2 is the preferred
flowsheet, and further studies and development work for two region
processing should be aimed at this system. This "modified" flowsheet

provides a minimum size reducer and bismuth rate while placing little
28

restriction on the decay tank volume; the decay tank volume is govermned

by heat removal considerations. The protactinium concentration in
the blanket is governed by the metal rate only and not by the decay
tank size. This flowsheet also results in relatively low protac-
tinium concentrations in the reducer. Less radioactivity in the
reducer may mean lower maintenance costs since access for repairs

may become less restrictive.

An additional attractive feature of the "modified" flowsheet is
that the processing plant can be evolved from a "no-reducer” system.
That is, if development of a reliable reducer is more difficult than
expected, the system can first be operated in the manner shown in
Fig. 3 with no reducer. Instead fresh lithium-7 and thorium metal may
be purchased and consumed in the process. Then when a suitable
reducer (or combination electrochemical reducer -oxidizer ) becomes
available, it can be installed, and the system can be operated in
the manner shown in Fig. 2. The "throw-away" or "nmo-reducer"
concept is viewed as an alternative or interim mode of operation for

the system.

Although the system can be operated with "throw-away";
vigorous development of a reducer is still recommended. In addition
to having no inherent material consumption, the flowsheet shown in
Fig. 2 (with reducer) allows a large degree of self-regulation which
is not provided by the "no-reducer'" flowsheet of Fig. 3. The closed
loop in the process insures that no lithium (or thorium) will be
continuously removed from the blanket. That is, if due to an error
in the reducer, the metal stream 3 has a low lithium concentration,
this error cannot persist indefinitely. At first lithium will be
extracted from the blanket. But as this lithium accumulates in the
process streams, the recycle will raise the lithium concentration
in F and correct the error. If the total process plant volume is
smal? compared to the blanket salt volume, lithium-thorium concen-

tration control will be largely self-regulating.

9

[

S
29

If interest in two region reactors persists, further pursuit of
the chemical and engineering uncertainties in this process is
recommended. More chemical data are needed to confirm all of the
equilibrium relations involved (including solubilities) and to test
or develop materials of construction. Engineering studies should
include: (1) an economic evaluation and optimization of the flow-
sheet (to the extent possible at that time); (2) testing of the
reductive extraction, hydrofluorination, and reduction steps as
single units (with uranium as a "stand-in" for protactinium) to
demonstrate their feasibility and develop equipment designs; and

(3) demonstration of entire flowsheet in an integrated unit.

Much of these efforts will be also applicable to processing of
molten salt reactors with a single fluid core if a separate blanket
stream is still used or if reductive extraction is selected as a

processing step for the fissile-fertile core salt.

 
50

REFERENCES

"Molten-Salt Reactor Program Semiannual Progress Report" for
Period Ending February 28, 1967, ORNL-4119, July 1967, pp. 150-2.

Foust, A. §., Wengel, L. A., Clump, C. W., Maus, F., and Anderson,
L. B., Principles of Unit Operations, John Wiley and Sons, Inc.,

New York 1964, p. T7T.

Briggs, R. B., "Summary of the Objectives, the Design, and a
Program of Development of Molten-Salt Breeder Reactors," ORNL-

TM-1851, June 12, 1967.
\O OO W1 W D~

VRO ERNNGUUEEPSUHOZINEMOTNDQORLGIERRNO O

. G. Alexander
. F. Baes
. J. Barton

E. Beall
Bender

. Bettis

. Blanco

. Blomeke
. Bohlmann
. Borkowskil
. Boyd

. Briggs
antor
Carter

. Cathers

. Compere
Culler, Jr.
Ditto
Eatherly
Ferguson
Ferris
Frye, Jr.
‘Grimes
Grindell
Goeller
Hannaford
Haubenreich

QuwEHGOOMEW

Hoffman
Hor ton

Jordan

Kasten

. Kedl

T. Kelley

S. Kirslis

Lu?d?::&.z?u’zn?ffl?wmzw'fiht"t"Ht"

Hightower, Jr.

31

DISTRIBUTION

36.
37
38.
39.
LO.
h1.
Lo,
L3,
L,
45,
Lé.
L.
L8.
49.
50.
51.
52.
55
5h.
55.
56.
5T
58.
59.
60.
61.
62.
63 .
6L . .
65.
66.
67-68.
69.
T70-T1.

e R

CRPCCEIERGYORG PN NGO A

n

PR EHEOQOZR W 0N

2O

~ =

ap

O@mZOwunmd>mog o

Lane
Lin

. Lindauer
. Lyon
. MacPherson

MacPherson

. McCoy

McDuffie

. McNeese

. Nichols

. Nicholson
. Oakes

. Perry

. Roberts

. Rosenthal

Shilling
Scott

. Shaffer
. Skinner
. Stone

. Strehlow
. Thoma

. Watson

. Welr

Weinbery

. Whatley

. White

Gale Young

E. L. Youngblood

Central Research Library
Document Reference Section
Laboratory Records

T2. Laboratory Records-~RC