129

&Modeling the Solvent Extraction of Oilseeds

G. A b r a h a m * , R.J. Hron Sr. and S.P. Koltun
Southern Regional Research Center, AR$/USDA, P.O. Box 19687, New Orleans, LA 70179

A c o m p u t e r m o d e l and a n e x p e r i m e n t a l p r o c e d u r e of stages is known. The number is put into the model

for g e n e r a t i n g data n e e d e d in t h e m o d e l h a v e b e e n and the residual lipids in the flakes extracted by t h a t
d e v e l o p e d for t h e o i l s e e d e x t r a c t i o n p r o c e s s . The number of stages are calculated. If, on the other
e x p e r i m e n t s are r e l a t i v e l y s i m p l e a n d are per- hand, an extractor is to be designed to achieve a par-
formed with a bench-top extractor. Experimental ticular residual lipid value, then this residual lipid
results and m o d e l i n g c a l c u l a t i o n s are p r e s e n t e d value is put into the model and the number of stages
for t h e e x t r a c t i o n o f c o t t o n s e e d u s i n g h e x a n e , required to achieve this value is calculated. While the
i s o p r o p a n o l and e t h a n o l . The c a l c u l a t i o n s s h o w analyses of both deep bed and shallow bed extractors
t h a t i n an a l c o h o l e x t r a c t i o n u s i n g a chill s e p a r a - are similar, this study focuses on the mathematicallysimpler
t i o n , i s o p r o p a n o l ' s g r e a t e r oil miscibility a l l o w s shallow bed extractor (9). In summary, as input to the
for a l o w e r s o l v e n t - t o - f e e d r a t i o t h a n d o e s e t h - model the following quantities are needed:
anol. U s i n g t h e latter s o l v e n t , h o w e v e r , a c h i e v e s • The mass flow rate of feed flakes.
l o w e r r e s i d u a l lipids in t h e e x t r a c t e d meal b e c a u s e • The lipid and moisture content of feed flakes.
r e c y c l e d e t h a n o l c o n t a i n s less oil t h a n recycled • Miscella/solids equilibrium data.
isopropanol. • The number of extractor stages or desired residual
lipids in extracted flakes.
For an alcoholic extraction the following input also is
Solvent extraction of oilseeds accounts for over 90% needed:
of the vegetable oil produced in the U.S. In recent • Oil/ethanol solubility data at the extractor and
years there has been considerable interest in both separator temperatures.
industry and government in a suitable replacement • The percent solids in the marc leaving the press.
for hexane, the solvent currently used (1-5). At pres- The output from the model contains the following:
ent, ethyl and isopropanol (6) appear to be the most • The mass flow rate of solids, solvent and oil in all
likely candidates. Hron et al. are developing a process streams.
t h a t uses ethanol in the solvent extraction of cotton- • The residual lipids in the extracted flakes or the
seed (7,8). Scaling up the process from the laboratory number of extractor stages.
to a pilot plant facility, formerly used for hexane
extraction, required changes in the existing physical MATERIAL BALANCESAND EQUIUBRIUM EQUATIONS
facilities and in operating conditions. To assure There are typically two types of equations needed in
proper sizing of new equipment and as a guide in the the modeling of a process: material balance equations
selection of operating conditions, a computer model and equilibrium equations. To model its function,
of the extraction process was developed. This paper every unit or piece of equipment in the process will
describes t h a t model. The results of laboratory exper-
need one or more material balances and, depending
iments performed to generate the input d a t a needed on the nature of the unit, one or more equilibrium
for the model also are presented. In its present form, equations. Consider, for example, the surge t a n k
this model is a useful tool for providing the kind of shown on Figure 1. For an alcohol extraction there
information mentioned above. It can also be used to are three streams entering the t a n k and one stream
make economic comparisons a m o n g alternate pro- leaving. The function of this t a n k is to collect the
cessing conditions and strategies. Such a study is incoming material and provide a single output stream
currently underway. However, more experimental to the extractor. Two material balances are needed to
d a t a are needed if the model is to be used in the
fully describe this operation:
design of new commercial extraction facilities, as an
expert computer system for the day-to-day manage- Sum of mass flow rates of
ment of an existing oil mill, or for incorporation in an oil in incoming streams -
automatic computer control system. Mass flow rate of
oil in outgoing stream = 0
STRUCTURE OF MODEL Sum of mass flow rates of
solvent in incoming streams -
The model, written in FORTRAN, uses a G a u s s i a n Mass flow rate of
elimination to solve a set of material balance and solvent in outgoing stream = 0
equilibrium equations. The flow rates and concentra- Now consider the chill separation used in the alcohol
tions of solids, solvent and oil in all streams entering extraction process. Once the miscella leaves the ex-
and leaving each unit are calculated. Within the tractor it is cooled to about 5 C. Two phases are
extraction unit there is a stage-to-stage calculation formed, a h e a v y phase consisting of mostly oil and
where one of two values is computed, depending on gums with about 15% ethanol and a light phase con-
w h a t is to be accomplished with the model. If an sisting of ethanol with about 3% oil. These two phases
existing extractor is to be modeled, then the number are then separated by centrifugation. Four equations
are needed to describe this operation. In addition to
1Presented in part at the AOCS meeting in Honolulu, HI, in
1986. material balances, similar to the previous two equa-
*To whom correspondence should be addressed. tions, two equilibrium equations are also needed:

JAOCS, VoL 65, no. 1 (January 1988)

130

G. ABRAHAM, R.J. HRON SR. AND S.P. KOLTUN

miscetla drains rapidly. As time increases, the drain-

ing slows a n d eventually M / I reaches a constant
C~UDEOIL
RECYCLESOLVENT jRANE
ME, value. T h e r e m a i n i n g oil and solvent are either bound
!SEPARATION RECYCLE
MISCELLA within the cells of the flakes are bound to the surface
of the flakes. T h e time to reach a c o n s t a n t value can
be considered a residence time for an extraction in a
~ FEEDMISCELLA
given stage because it is the time it takes free miscella
PRODUCTMISCELLA C~
FAI at the top of the bed to drain t h r o u g h to the bottom of
_I
MARC
EXTRACTOL-
R -I the bed. For the shallow bed extractor, residence time
should therefore correspond to the time it takes the
FEED bed to travel from one s p r a y to the next.
\i FLAKES
It is possible t h a t different oil concentrations in the
CONDENSED CRUDE
MISCELLA I '-' SOLVENT OIL miscella cause changes in the binding forces such
t h a t the c o n s t a n t or final value of M / I discussed
~ASHED above changes. To calculate the number of stages
MARC
necessary to achieve a desired residual lipids content
PRESS
MISCELLA P ~
\ DESOL'ENT,,
FODTS
.I-' or to calculate the residual lipids for a given n u m b e r
of stages it is n e c e s s a r y to know the final value of
EXTRACTED REFINED
M / I as a function of the total concentration of oil
MEAL OIL bound to the flakes (xF). Further, it is necessary to
know x F as a function of the concentration of oil in
the miscella drained from the flakes (xS). There are at
FIG. 1. D e t a i l e d s c h e m a t i c o f e t h a n o l e x t r a c t i o n pro-
cess. least two reasons w h y the values for x F a n d xS are
not the same. I f the flakes exert a preferential adsorp-
tion for either the oil or solvent, then x F will differ
T h e mass flow rate of oil in the light p h a s e / from xS. Second, if the oil is only partially miscible in
the total mass flow rate of the light phase = F(T) the solvent then the oil a n d solvent bound to the
flakes consist of two phases, a homogeneous oil-
T h e mass flow rate of oil in the h e a v y p h a s e / solvent phase and an undissolved oil phase. Since x F
the total mass flow rate of the h e a v y phase = G(T) refers to the total c o n c e n t r a t i o n of oil in both of these
Where F(T) a n d G(T) represent functions of tempera- phases t a k e n together it differs from xS, which refers
ture which, for ethanol at 5 C, are equal to 0.03 a n d only to a homogeneous phase. This last point is
0.85, respectively. There are two more types of equili- e x p a n d e d upon when the experimental results are
brium equations associated with the streams contain- discussed, later in this study. T h e details of how these
ing flakes. To u n d e r s t a n d these equations it is helpful equilibrium relationships are used in the stage calcu-
to consider w h a t occurs when the solvent comes in lation will not be discussed here because this topic is
contact with the flakes. If oilseed flakes are com- discussed elsewhere (10). The two p h e n o m e n a just
pletely wetted with a solvent and allowed to drain, a discussed, t h a t is, the a m o u n t of miscella bound to
plot of the weight of solvent and oil on the flakes per the flakes after extraction versus the concentration of
weight of insoluble solids (M/I) versus drain time oil in t h a t miscella a n d the concentration ofoil in the
could be obtained as shown in Figure 2. Initially, the drained miscella versus the c o n c e n t r a t o n of oil in the
miscella bound to the flakes, give the two equilibrium
equations needed for the streams containing flakes.
To summarize, several kinds of equilibrium infor-
m a t i o n are needed to model the solvent extraction
process:
• The weight of miscella bound to the flakes per unit
weight of solid as a function of the concentration of
oil in the bound miscella.
• The concentration of oil in the drained miscella
"t
I
versus the c o n c e n t r a t i o n of oil in the bound mis-
cella.
Further, if the oil is only partially miscible in the
BOUND solvent, as is the case with ethanol and isopropanol,
MISCELLA then the following solubility d a t a must be known:
• The m a x i m u m solubility of oil in solvent at the
extraction temperature.
RESIDENCE • The m a x i m u m solubility of oil in solvent at the
[> TIME cooling temperature.
• The m a x i m u m solubility of solvent in oil at the
TIME cooling temperature.
There is, at present, no reliable theoretical method for
FIG. 2. R a t i o o f w e i g h t o f m i s c e l l a o n f l a k e s (M) to obtaining the above information. It must be deter-
w e i g h t o f i n e r t s o l i d s (I) v e r s u s d r a i n t i m e . mined experimentally (11).

JAOCS, Vol. 65, no. 1 (January 1988)

 131
MODELING THE SOLVENT EXTRACTION OF OILSEEDS

EXPERIMENTAL possible at the given temperature. Further extrac-

tions did not increase the oil concentration in the
Although the major t h r u s t of this study was to model miscella. After the third extraction, the drained mis-
an ethanol extraction process, equilibrium d a t a was cella was weighed, desolventized in a r o t a r y vacuum
o b t a i n e d for h e x a n e a n d i s o p r o p a n o l as well as evaporator a n d the weight of the r e m a i n i n g oil found.
ethanol. T h e solvents used in the experiments were In order to determine the a m o u n t of petroleum ether
commercial grade. Mississippi cottonseed was used. solubles contained in the oil it was necessary to redis-
It was dehulled, flaked to about 0.03 cm (0.01 inch), solve the desolventized oil in petroleum ether and fil-
a n d dried to about 2% moisture in an oilseed extrac- ter out and weigh a n y insoluble material. The percent
tion pilot plant facility. Drying the flakes is neces- petroleum ether solubles can then be calculated, and
s a r y in an alcohol extraction so t h a t the solvent does it is this q u a n t i t y t h a t is used for oil solubility. To
not pick up moisture and reduce its miscibility with determine solubilities at the chilling temperatures,
oil (12).
the oil was extracted at 75 C and cooled to the desired
Oil-ethanol solubility. A given vegetable oil can temperature, thus forming two phases. The solubili-
h a v e various f a t t y acid a r r a n g e m e n t s in the trigly-
ties in both phases are then determined as previously
ceride. It can also h a v e different amounts of free f a t t y
described.
acids a n d other nontriglyceride components, depend- Oil-flake equilibrium. The other required data for
ing on the method of extraction. This variability the model, as was discussed earlier, are information
affects its solubility in various solvents. To determine relating both the a m o u n t of miscella bound to the
the o i l / e t h a n o l solubility for cottonseed oil in an flakes after draining a n d concentrations of oil in
ethanol extraction process the following experiments drained miscella versus concentrations of oil in bound
were performed at different temperatures: A bench- miscella. These d a t a were obtained using the bench
top extractor (Fig. 3) was charged with 350 g of full extractor discussed above. After heating the extrac-
fat cottonseed flakes, dried to about 3% moisture, and
tor to the desired operating temperature for a given
500 g of 95% aqueous ethanol. The extractor, 30 cm solvent, it was charged with 350 g of flakes. Heated
long and 10 cm in diameter, was made of stainless solvent (600 g) was then added and circulated through
steel and was jacketed. Water from a circulating con-
the flakes for 30 min. A sparger was used at the inlet
s t a n t temperature b a t h was used as the heating to provide an even distribution of solvent over the
medium. The top of the extractor was covered with a flake bed. At the end of the extraction period the mis-
metal plate and gasket to keep the system vapor tight. cella was drained from the extractor. After draining,
The solvent was circulated through the extractor for the flakes were removed from the extractor and imme-
30 min while the system was kept at the desired diately weighed. This was done quickly a n d with lit-
temperature. A peristaltic pump was used for solvent tle exposure of the wet flakes to the a t m o s p h e r e so
circulation. At the end of the extraction time, the mis- t h a t solvent vaporization was minimized. The flakes
cella was d r a i n e d a n d the e x t r a c t e d flakes were were air desolventized for about 12 hr and then in a
removed. The extractor was t h e n recharged with
forced draft oven for 2 h r at 101 C. After desolventiza-
another 350 g of full fat flakes, and the previously tion the flakes were reweighed to determine their sol-
drained miscella was added and circulated for another vent content immediately after draining. A modified
30 rain. This procedure was then repeated a third time AOCS residual lipid analysis was performed on these
to assure t h a t the miscella contained the most oil flakes and, using this value along with the weight of
solvent, two desired quantities were then calculated:
EXTRACTORINLET the weight of miscella bound to the flakes per unit
weight of solids a n d the concentration of oil in the
b o u n d miscella. T h e c o n c e n t r a t i o n of oil in t h e
HEATING o--,- drained miscella was determined as described pre-
WATEROUT <] / viously for the solubility experiments. To obtain the
.JACKETED quantities just calculated for a different oil concen-
/
-/ /
yEXTRACTOR tration of the bound miscella the experiment described
/ above was changed slightly. After the miscella was
/
drained from the flakes it was discarded and a second
b a t c h of fresh solvent was added to the flakes. This
new solvent was also circulated t h r o u g h the flake bed
RECIRCULATETO for 30 min. The rest of the experiment and the analy-
EXTRACTOR --o<~ HEATING sis were the same as described for the single charge of
INLET
t ~ATER IN solvent. The oil concentrations for both drained and
bound miscella were lower when two charges of fresh
solvent were used t h a n when a single charge was
used. Three or more charges were used to obtain
equilibrium d a t a at still lower oil concentrations.
EXTRACTOR Hexane, 91% isopropanol and 95% ethanol were used
DRAIN as solvents. The extraction temperatures were 60 C
for hexane, 78 C for isopropyl and 75 C for ethanol.
FIG. 3. B e n c h t o p e x t r a c t o r u s e d to d e t e r m i n e e q u i l i b r i - Drain times. A final set of experiments was per-
um relationships. formed to determine the length of time needed to

JAOCS, Vol. 65, no. 1 (January 1988)

132

G. ABRAHAM, R.J. HRON SR. AND S.P. KOLTUN

drain a flake bed saturated with miscetla which, as
discussed earlier, corresponds to the residence time in
a n extraction stage. T h e e x t r a c t o r was charged with
flakes to a height of 10 cm, and ethanol was circulated
t h r o u g h it for 30 min. It was then allowed to drain for
1 min. The flakes were removed and the weight of
miscella r e m a i n i n g on the flakes per unit weight of
solids was determined. T h e r e a c t o r was t h e n re-
c h a r g e d a n d the above e x p e r i m e n t was r e p e a t e d
except t h a t the drain time was increased to 2 min.
This experiment was performed repeatedly for drain
times up to 7 min. In order to determine the drain
time as a function of bed height, two more sets of
experiments were t h e n performed in a m a n n e r sim-
ilar to the above except t h a t the extractor was charged
with 18- and 25-cm-high flake beds.
RESULTSAND DISCUSSION
Solubilities. At its boiling point, 78 C, 95% ethanol
has a 12% oil solubility. At 75 C, a realistic operating
temperature, the solubility is 10%. The 12% solubility
of cottonseed oil in 95% aqueous ethanol, measured in
this work, is greater t h a n previously reported values
(13,14). T h e s e o t h e r studies, however, h a v e used
expeller-produced crude oil or commercially refined
oil in their d e t e r m i n a t i o n s whereas in this study
ethanol-extracted crude oil was used. Apparently,
some of the nontriglyceride components extracted by
the ethanol are also soluble in petroleum ether. At 5 C
the solubility of oil in the lighter phase in 3%. T h e
heavier phase contains 85% oil. Decreasing the tem-
perature below 5 C does not appreciably affect the
solubility.
E q u i l i b r i u m curves. The results of the equilibrium
experiments are shown in Figures 4 and 5. Data points
clustered together are results of the repetition of one-
charge, two-charge and three-charge experiments. In
the ethanol extractions four- a n d five-charge experi-
ments also were performed. For each solvent at least
five repetitive experiments were performed for each
n u m b e r of charges. Some of the d a t a points overlap
on the plots because several of the repetitions pro-
duced very close results. In Figure 4, as the oil con-
c e n t r a t i o n s increase, the slope of the line on the
drained miscella versus bound miscella curve for
ethanol is greater t h a n t h a t for the isopropanol curve
which in turn is greater t h a n t h a t for the h e x a n e
curve which has a c o n s t a n t slope of 1. This behavior
can be explained by considering t h a t on a molecular
level there is a driving force, or potential, which
causes the oil to dissolve in the solvent. If the oil is
only partially miscible in the solvent, t h e n this force
decreases as the concentation of oil in the solvent
increases a n d goes to zero at the s a t u r a t i o n concen-
tration. Since isopropanol h a s a greater saturation

70

25.0
65
~ 0 - ~ o ISOPROPANOL
-

ETHANOL
"~. 20.0
50

U
rr

f...l 55
Z
15.0 nr
0 I
Z

/
d 0
UJ < 50
U
CD
10.0 -3
llJ / "
rj / "
O3
Z
~E .~>,/././"/ HEXANE
45
tlJ
Q
i'1
d
5.00

/" 40

.000 I I I I
.000 5.00 10.0 15.0 20.0 35 ................. I I I I I
• 00 5.0 10. 15. 20. 25.

LIPIDS IN DRAINED MISCELLA (% W/W]

LIPIDS IN MISCELLA ON MARC (% ~/W)
FIG. 4. E x p e r i m e n t a l l y d e t e r m i n e d v a l u e s for t h e lipid
concentration of the miscella remaining on the marc F I G . 5. E x p e r i m e n t a l l y d e t e r m i n e d v a l u e s f o r t h e
after draining versus the lipid concentration of the weight percent of miscella on the marc after draining
drained miscella. v e r s u s the lipid c o n c e n t r a t i o n o f t h a t m i s c e l l a .

JAOCS, Vol. 65, no. 1 (January 1988)


MODELING THE SOLVENT EXTRACTION OF OILSEEDS
c o n c e n t r a t i o n t h a n ethanol, its driving force de-
creases more slowly t h a n ethanol over the same con-
centration range. Because hexane is completely mis-
cible with oil, its driving force is not a function of
concentration. In an alcohol extraction, due to a
reduction in the driving force, the miscella bound to
the flakes after draining consists of two parts, the
alcohol with dissolved oil and the unextracted oil. It
is not necessary to consider the unextracted oil and
actual bound miscella as two distinct species, but
rather to treat them as a bound miscella h a v i n g an oil
concentration greater t h a n saturation (7). For exam-
ple, at a concentration of 12% in the drained ethanol
miscella, the saturation point at the temperature of
the experiment, the bound miscella contains 24% oil.
Further, when the concentration of oil in the drained
ethanol miscella is 8%, the bound miscella concentra-
tion is 14%. This shows t h a t even though the drained
miscella is not saturated there is still unextracted oil
in the flakes because the concentration of 14% is
greater t h a n the saturation concentration of 12%. It is
not until the drained miscella contains less t h a n
about 5% oil t h a t the bound and drained concentra-
tions are approximately equal. Figure 5 shows the
amount of miscella left on the flakes as a function of
the a m o u n t of oil in t h a t miscella. The percentages
are calculated on a weight basis. It is interesting to
note t h a t as the percentage of oil in the miscella on
the marc increases, the amount of miscella bound to
the marc increases when hexane is the solvent but
decreases when alcohol is the solvent. This pattern
remains when the percentages are calculated on a
volume basis. A possible explanation for this is t h a t
30.0
25.0
20.0
I LL
0_
uJ
o z
o >
w
m .J
15.0
20.0
5.00
.000
l
2.00
DRAIN
I
A.O0
TIME
I
6.00
{MINUTES)
I
8.00
I
i0.0
FIG. 6. E x p e r i m e n t a l l y d e t e r m i n e d t i m e n e e d e d to d r a i n
t h e f r e e m i s c e l l a f r o m a bed o f f l a k e s for t h r e e d i f f e r e n t
bed h e i g h t s .
133
the more oil there is in a hexane miscella the more
viscous the liquid becomes and the less it drains. The
alcohols, on the other hand, are more polar t h a n hex-
ane a n d have a greater affinity for the flakes. The
addition of oil to the alcohol reduces the polarity of
the liquid and facilitates draining.
Drain curve. The results of the ethanol draining
experiments are shown in Figure 6. The points shown
represent the time it took flakes to drain free miscella
at each bed height. Draining for longer periods t h a n
shown did not decrease the ratio of bound miscella to
solids. As discussed earlier, these times represent the
optimum residence time in each stage. For example, if
the bed depth is 25 cm then it takes about 6.5 min to
drain the bed. Consequently, if the solvent spraying
heads are, say, 130 cm apart, then the bed speed
through the extractor should be 20 cm/min. It is
important to note t h a t the bed should not be allowed
to become dry and should be operated at near flood-
ing conditions.
E x a m p l e s of modeling calculations. U s i n g the
above laboratory data as input to the computer model,
various parameters of a pilot or commercial scale
extraction operation can be computed. Figure 7 shows
the calculated minimum solvent-to-feed ratio for the
three solvents considered in this study. It ranges
from a value of 1 in hexane to almost 4 in ethanol, the
least miscible of the three. These results are for a chill
separation of the alcohol solvents and an evaporative
separation of hexane. For the alcohol solvents, this
factor is a function of both the concentration of oil in
the fresh miscella and the solubility ofoil in the alco-
hol at the extraction temperature. T h a t is, at this
4.0
ETHANOL
3,0
Q
LU
UJ ISOPROPANOL
1--
IM
Q
@.0
Ul
HEXANE
i.O
! EXTRACTION SOLVENT
FIG. 7. Model c a l c u l a t i o n o f t h e s o l v e n t to f e e d r a t i o s
for t h e t h r e e s o l v e n t s c o n s i d e r e d in t h i s study.
JAOCS, Vol 65, no. 1 (January 1988)
 G. ABRAHAM, R.J. HRON SR. AND S.P. KOLTUN

ratio the miscella in the flakes leaving the extractor

has the same oil concentration as the fresh miscella 2,0
entering the extractor. Consequently, an increase in
this ratio has no effect on the residual lipids which
are greater than 1% for both alcohol solvents under
these conditions. Since h e x a n e is completely miscible
with oil, its solvent-to-feed ratio is governed by the
desired residual lipids, which is 1% for this case. 1.5 ISOPROPANOL ETHANOL
Figure 8 s h o w s the percent residual lipids remaining Lu
in the extracted flakes after desolventization as a LU
lL
function of the number of stages in the extractor.
These results are with pressing the marc in the alco- 0
W
hol extractions. Hexane requires seven stages to reach I.-,
ETHANOL
"~
cr 1.0 ISOPROPANOL
the minimum residual lipids. Ethyl and isopropanol 0
O. (PRESSED (PRESSED
require eight stages to reach minimums. Notice that
because isopropyl has a greater oil solubility, 6% at 5 W MARC) MARC)
C, t h a n ethanol, 3% at 5 C, the recyle stream in the Z
W
isopropyl process will have a higher concentration of .J
HEXANE
oil and hence leave more oil in the extracted flakes. 0
.50
Figure 9 s h o w s the pounds of solvent in the flakes
stream entering the desolventizer. The load on the
desolventizer is s i g n i f i c a n t l y increased with the
alcohol solvents. Even w h e n the marc is pressed,
there is about 25% more alcohol going to the desol-
ventizer t h a n there would be in a h e x a n e extraction.
Further, because the alcohols are less volatile t h a n
hexane, this increased load represents a large energy EXTRACTION SOLVENT
increase. Karnofsky (15) has pointed out the m a n y
difficulties in the desolventization of water miscible
solvents. This operation has received little attention F I G . 9. M o d e l c a l c u l a t i o n of the solvent load on the
in studies on alcohol extractions. However, it is desolventizer.

7.0

6.0
4.o T
5.0
ISOPROPANOL

3~ a.o T /
u) 4.0 ISOPROPANOL
D
H /
D_ /
J
3.0
2,0
121 ETHANOL
I--I -J
U~
W
II £D //~ET~ANOL
2.0
£r
l.O
1.0 HEXANE

.......... I I -~, I
3 4 5 6 7 8 9 t.0 2.0 3.0 4.0 5.0

NUMBER OF STAGES LZPIDS IN RECYCLE STREAM (% ~/W]

FIG. 8, Model calculation of the percent residual lipids

in the extracted flakes versus the number of extraction FIG. 10. Model calculation showing the effect of supple-
stages. mental oil removal on the percent residual lipids.

JAOCS, Vol. 65, no. 1 (January 1988)


MODELING THE SOLVENT EXTRACTION OF OILSEEDS
p o t e n t i a l l y the m o s t t r o u b l e s o m e in a c h i e v i n g com-
m e r c i a l i z a t i o n w i t h one of t h e s e a l t e r n a t e s o l v e n t s . I t
is i n t e r e s t i n g to c o n s i d e r t h e effects on r e s i d u a l lipids
if t h e a m o u n t of oil in t h e recycle alcohol s t r e a m s is
reduced to a lower v a l u e t h a n is possible w i t h chilling.
T h i s c a n be done w i t h distillation or w i t h a m e m b r a n e
process (Fig. 1). T h e r e s u l t s of t h e c a l c u l a t i o n s are
s h o w n in F i g u r e 10. I t is a p p a r e n t t h a t for b o t h
ethanol and ispropanol the results show that, even
w i t h p r e s s i n g , t h e r e m u s t be less t h a n 1% oil in the
recycle s t r e a m if t h e r e is to be less t h a n 1% r e s i d u a l
lipids in the e x t r a c t e d flakes.
M a t h e m a t i c a l m o d e l uses. A m a t h e m a t i c a l model
of a p h y s i c a l process h a s a v a r i e t y of uses. R e s e a r c h -
ers a t t h e S R R C h a v e used t h e e x t r a c t i o n m o d e l dis-
c u s s e d h e r e in m o d i f y i n g a h e x a n e e x t r a c t i o n pilot
p l a n t for use a s a n e t h a n o l e x t r a c t i o n p l a n t . I t is also
a n aid in u n d e r s t a n d i n g t h e p r o c e s s a n d in p o i n t i n g
out a r e a s o f difficulty t h a t n e e d m o r e r e s e a r c h , s u c h
a s in d e s o l v e n t i z a t i o n . C o s t studies a n d e c o n o m i c
c o m p a r i s o n s of a l t e r n a t e p r o c e s s s c h e m e s are g r e a t l y
f a c i l i t a t e d b y t h i s k i n d of model. I t c a n also be incor-
p o r a t e d in a n e x p e r t s y s t e m for use in the d a y to d a y
o p e r a t i o n s of a n oil mill. S u c h a n e x p e r t s y s t e m
w o u l d a s s u r e u n i f o r m o p e r a t i o n of the p l a n t a n d
w o u l d be a n i m p o r t a n t aid to i n e x p e r i e n c e d opera-
tors. I n a m o r e a d v a n c e d f o r m a process m o d e l c a n be
used in a u t o m a t i c computer control. T h e potential down-
135
s t r e a m effects of a n u p s t r e a m d i s t u r b a n c e c a n be
predicted a n d corrected, a definite i m p r o v e m e n t o v e r
s t a n d a r d f e e d b a c k control.
REFERENCES
1. Jones, L,A., Cotton Gin and Oil Mill Press, May 10, 1986.
2. Johnson, L.A., and E.W. Lusas, J. Am. Oil Chem. Soc.
60:229 (1983).
3. Hron, R.J., S.P. Koltun and A.V. Graci, Ibid. 59:674A (1982).
4. USDA, Fats and Oils Situation (FOS/299), (May 1980).
5. Friedrich, J.P., and E.H. Pryde, J. Am. Oil Chem. Soc.
61:223 (1984).
6. Baker, E.C., and D.A. Sullivan, Ibid. 60:1271 (1983).
7. Hron, R.J., and S.P. Koltun, Ibid. 61:1457 (1984).
8. Hron, R.J., G. Abraham and S.P. Koltun, Oil Mill Gaz. 89:10
(1984).
9. Norris, F.A., in Bailey's Industrial Oil and Fat Products,
edited by D. Swern, Vol. 2, 4th edn., John Wiley and Sons,
New York, NY, 1982, pp. 215-251.
10. Treybat, R.E., Mass Transfer Operations, 3rd edn., McGraw
Hill Co., New York, NY, 1980, pp. 717-765.
11. Smith, C.T., J. Am. Oil Chem. Soc. 28:274 (1951).
12. Hassaanen, N.Z., L.A. Johnson, J.T. Farnsworth and E.W.
Lusas, Ibid. 62:619 (1985).
13. Magne, F.C., and E. Skau, Ibid. 30:288 (1953).
14. Rao, R.K., M.G. Krishna, S.H. Zaheer and L.K. Arnold, Ibid.
32:420 (1955).
15. Karnofsky, G., Ibid. 62:693 (1985).
[Received F e b r u a r y 20, 1987]
JAOCS, VoL 65, no. 1 (January 1988)