Process Design and Economic Studies of

Alternative Fermentation Methods for the

Production of Ethanol

GERALD R. CYSEWSKI* and CHARLES R. WILKE,

Department of Chemical Engineering and Lawrence Berkeley
Laboratory, University of California, Berkeley, California 94720

Summary
Cell recycle and vacuum fermentation processes are described for the continuous
production of ethanol. Preliminary process design studies are employed to make an
economic comparison of these alternative fermentation schemes with continuous
and batch fermentation technologies. Designs are based on a production capacity of
78,000 gal 95% ethanol (EtOH)/day employing molasses as the fermentation sub-
strate. The studies indicate that a 57% reduction in fixed capital investment is
realized by continuous rather than batch operation. Further decreases in required
capital investment of 68 and 71% over batch fermentation were obtained for cell
recycle and vacuum operation, respectively. However, ethanol production costs
were dominated by the cost of molasses, representing over 75% of the total manu-
facturing cost. But, when a reasonable yeast by-product credit was assumed, the net
production cost for 95% ethanol was estimated at 82.3 and 80.6 cent/gal, for the cell
recycle and vacuum processes, respectively.

INTRODUCTION

Presently, 9.8 x lo5 tons/year of industrial ethanol are produced

in this country for use as a solvent and a chemical feed stock. Over
98% of this industrial ethanol is produced from the catalytic con-
version of ethylene. However, with the impending petroleum short-
age there has been renewed interest in the production of ethanol
via fermentation. The overwhelming advantage of fermentation is
that the raw materials are renewable. Any fermentable sugar can
be used. The recent developments in acid and enzymatic hydrolysis
of cellulose to fermentable sugars may possibly allow the economic
production of fermentation ethanol from the vast and renewable

* Present address: Dept. of Chemical and Nuclear Engineering, University of

California, Santa Barbara, Calif. 93 106.

Biotechnology and Bioengineering, Vol. XX, Pp. 1421-1444 (1978)

@ 1978 John Wiley & Sons, Inc. OOO6-3592/78/0020-1421$01.OO
1422 CYSEWSKI AND WILKE

quantities of cellulose on the earth.'S2 Ethanol could then serve not

only as a chemical feed stock but as a liquid fuel.3
The renewed interest in ethanol fermentation has led to the op-
timization of old fermentation processes, and to the proposal of
several new innovative fermentation scheme^.^,^ The aim of this
paper is to provide an economic comparison of these new fermen-
tation technologies with conventional fermentation processing.
Traditionally, ethanol fermentations have been operated batch-
wise, even though continuous fermentation would produce substan-
tial savings by eliminating fermentor down-time between the batch
fermentations. The two main reasons continuous fermentations
have not been more extensively employed are possible yeast mu-
tations and the problem of maintaining a high fermentation rate
during continuous fermentation. The problem of deleterious muta-
tions is particularly serious for the beverage industry. The slightest
change in fermentation products can impart noxious flavors in fer-
mented beverages. This is not a problem for the production of
industrial ethanol. In fact, continuous ethanol fermentations have
been maintained in the laboratory in excess of 60 days without any
signs of deleterious mutations.6 During these extended experiments
ethanol yields of over 90% of the theoretical yield were obtained-
although the product did have a noxious taste.
The low fermentation rates sometimes obtained in continuous
ethanol fermentations have been shown by numerous workers to be
caused by a lack of ~ x y g e n . ~ *Even
~ , * though ethanol fermentation
is an anaerobic process, trace amounts of oxygen are required for
biosynthesis. The oxygen requirement, however, can be eliminated
by adding an unsaturated lipid, ergosterol, to the fermentation
broth,I6 but the cost of ergosterol is prohibitive for industrial-scale
processes. Recent work using Saccharomyces cerevisiae (ATCC
No. 4126) has shown that an oxygen tension of 0.07 mm Hg was
optimal for ethanol p r o d ~ c t i o n As
. ~ shown in Figure 1, below an
oxygen tension of 0.07 mm Hg the yeast became oxygen starved
and the fermentor ethanol productivity (g EtOH producedfliter fer-
mentor volume hr) decreased; whereas, a t high oxygen tensions the
yeast metabolism began to shift from anaerobic to aerobic and less
ethanol was produced with a corresponding increase in cell mass
production. As a result, if the oxygen tension was maintained at
0.07 mm Hg by constantly sparging a small amount of air through
the fermentor, high fermentation rates were obtained with contin-
uous culture. If, however, air was not continuously sparged through
the fermentor and only an air-saturated feed was used, the oxygen
 DESIGN AND ECONOMICS OF ETHANOL PRODUCTION 1423

L I 1 I I I I I I
0 0.5 I .o 1.5 2.0
Oxygen tension ( m m Hg)

Fig. 1. Productivities as a function of oxygen tension at a dilution rate of 0.22

hr-l. 8.9% glucose feed, T = 35°C; pH = 4.0. (0)Fermentor EtOH productivity; ( x )
fermentor cell mass productivity; (m) specific EtOH productivity.

tension dropped to zero. Under these conditions complete fermen-

tation of even an 8.9% glucose feed was not possible with contin-
uous operation.
Fermentor ethanol productivities in both batch and continuous
culture are limited by two factors: ethanol inhibition and a low cell-
mass concentration. This is illustrated in Figure 2 for continuous
cu1tu1-e.~As the sugar feed concentration was increased, the specific
ethanol productivity (g EtOH produced/g cells hr) decreased be-
cause more ethanol was produced at high sugar concentrations and
ethanol inhibition increased. At low sugar concentrations, ethanol
inhibition was lessened, but the cell-mass concentration decreased.
These two counterbalancing effects produced an optimum fermentor
productivity at 10% sugar feed. At low sugar feeds, the fermentor
productivity was limited by a low cell density; while at high sugar
feeds, the productivity was limited by ethanol inhibition.
To overcome the low cell-density limitation a cell recycle system
can be employed. A portion of the cells are separated from the
fermented beer and returned to the fermentor. This increases the
cell density in the fermentor and produces higher ethanol produc-
tivities. The results of a cell recycle fermentation are shown in
Figure 3.5 The cell density was increased four times over simple
1424 CYSEWSKI AND WILKE

1 8 , I , I I I I , ,9 ,0.9

I6- 18 10.8
L
L
14- -7 ' -07:
?
m
12- - 6cs -06 $
-
2- 10- >
- 5 9 -05:
t

D
e e
-
8
8-

6-
-4:
-3:
W
c
-04
-03
-
-
E a
v)
4- - 2 2 -02

'0 2 4 6 8 10 12 14 16 18
Per cent sugar in feed
Conditions at "Complele" Substrate Utilization

Fig. 2. Effect of glucose concentration on continuous fermentation.

continuous operation resulting in a fourfold increase in ethanol

productivity.
In order to eliminate ethanol inhibition, ethanol must be removed
from the fermentation broth as it is formed. This is easily done by
taking advantage of ethanol's high volatility and boiling off the
ethanol from the fermentation broth. But the fermentation must be
run under a sufficient vacuum to obtain boiling at a temperature
compatible with the yeast. It is then possible to combine vacuum
and cell recycle operations and eliminate both limitations of con-
ventional continuous fermentation. The results of a vacuum-cell
recycle fermentation are shown in Figure 4.5 Here a 33.4% glucose
feed was fermented to less than 0.3% residual glucose. This was
not possible with atmospheric operation owing to ethanol inhibition.
A final cell density of 124 g dry wtniter was obtained in the vacuum-
cell recycle system producing an ethanol productivity of 82 g h t e r
hr or almost 12 times that obtained with simple continuous opera-
tion.
Yeast viability was not affected by vacuum operation as long as
sufficient oxygen was present. But to supply the required trace
amounts of oxygen, pure oxygen had to be sparged through the
fermentor. The only discernible difference noted with vacuum fer-
mentations was a 50% lower cell yield factor as compared to at-
 DESIGN AND ECONOMICS OF ETHANOL PRODUCTION 1425

mospheric pressure fermentations. The lower yield factor may be

a result of higher maintenance energy requirements of the yeast in
vacuum.
A summary of optimum values of parameters and maximum
ethanol productivities for each of the above operation modes is
shown in Table I . Complete details of the laboratory fermentation
systems and a full discussion of experimental results are given
elsewhere . 4 - 6

GENERAL PROCESS DESIGN BASES

Preliminary process designs of industrial-scale ethanol fermen-

tation plants were made employing the aforementioned modes of
operation: batch, continuous, continuous with cell recycle, and
vacuum with cell recycle. Each plant was based on a production

Fig. 3. Effect of increasing cell density by use of cell recycle in continuous

fermentation, 10% glucose feed.
1426 CYSEWSKI AND WILKE

80 -

- - I40
2
5 60-
-
a

-?
u
-
0
3
YI

-g
40- -80
-
U
0
- - 60

20 - - 40

- - 20
1 1 I I I I I I I I I I 0
0 20 40 60 80 100 120 140 160 180 200 220 240 260

capacity of 78,000 gal 95% EtOHlday using cane molasses as the

fermentation substrate. Molasses was assumed delivered to the
fermentation plants as a 50% sugar solution with 97% of the sugar
fer~nentable.~
The laboratory fermentation kinetics described above were used
for the process design studies. In each design optimum fermentation
conditions of temperature, pH, and oxygen tension have been as-
sumed.
Surely, optimal laboratory conditions cannot be expected in in-
dustrial operations. However, designs based on laboratory data
should yield a valid comparison between the various processing
schemes, although the absolute magnitude of the cost figures may
be optimistic. This is especially true with regard to the fermentation
substrate. Analytical grade glucose was used in the laboratory ex-
periments, while molasses is proposed in the industrial-scale pro-
cesses. It is not suggested that fermentation kinetics obtained with
glucose will be identical with results obtained with molasses.
Rather, that the extent the kinetics, or ethanol productivities, are
changed by employing molasses will be equally reflected in each
fermentation system, and a relative comparison between each sys-
tem remains valid. Further, each design is the result of a computer
process model and represents an optimum design. Thus, economic
 DESIGN AND ECONOMICS OF ETHANOL PRODUCTION 1427

TABLE I
Some Results of the Laboratory Fermentation Systems"
Cell-mass
Optimum Optimum concentration Maximum
oxygen sugar at optimum EtOH
tension concentration conditions productivities
Fermentation system (mm Hg) (%) (g dry wtkter) @/liter hr)

Batch b - 5.6' 2.26

Continuous 0.07 10.0 12.0 7.0
Continuous with cell 0.07 10.0 50.0 29.0
recycle
Vacuum with cell recycle - 124.0 82.0

a Optimum pH = 4.0 and optimum temperature = 35°C in all cases.

Optimum procedure was to initially air saturate broth and use an aerobically
grown 2.0% inoculum with no air sparging during fermentation.
At end of batch fermentation.
Assumes 6 hr fermentor down-time between 16 hr batch fermentation.
Oxygen tension could not be determined. Optimum procedure was to sparge
pure oxygen through the fermentor at a rate of 0.10 v/v/m at STP.

comparisons are based on the optimum design for each operation

mode.
The ethanol produced in each fermentation process was concen-
trated to 95 wt% by a single distillation step. Since only one distil-
lation is performed, the distillate product will contain approximately
0.5% fusel oil.l0 The fusel oil has no effect on the product value if
the ethanol is used as a liquid fuel. Also, the obnoxious taste
imparted by the fusel oil may eliminate the need for addition of
denaturants.

Economics
The f.0.b. process equipment costs were estimated from two main
sources, Peters and Temmerhause" and Guthrie,12 and Marshell-
Stevens indices were used to up-date the cost figures. The total
fixed capital cost was estimated as a multiple of the f.0.b. purchased
cost of the principal items of equipment. In the present case a
multiplier of 3.1 was used, except in the case of the fermentors for
which the multiple was increased to 4.24 to reflect the additional
instrumentation, piping, and installation costs associated with fer-
mentors.
Plant operating costs were divided into fixed charges and direct
production costs. A summary of the fixed charges is shown in Table
11. Here a 10 year straight-line depreciation was assumed and local
1428 CYSEWSKI AND WILKE

TABLE I1
Fixed Operating Costs
Percent of fixed
capital
investment per
year
(%I
Depreciation 10.0
Taxes 4.0
Insurance 1.0
Maintenance and repair 3.0
Operating supplies 1 .o

Total 19.0

taxes have been included. The total fixed charges amount to 19.0%
of the fixed capital investment per year.
Direct production costs were estimated according to Peters and
Temmerhause." A base labor rate of 5.60 dollarhan hour and a
8,500 hr year was assumed throughout the cost calculations. The
base utility rates are shown in Table 111. Electric power require-
ments were calculated assuming an 80% efficiency of electric-to-
mechanical power conversion. Also, an 80% efficiency was taken
for adiabatic gas compression and pumps.

PROCESS DESCRIPTIONS

Continuous Fermentation

The design basis of the continuous fermentation process is shown

in Table IV. Figure 5 shows a schematic process flow diagram of

TABLE 111
Base Utility Rates
Utility Rate
Cooling water 0.128 dollar/103 gal
Electric power 3 cent/kWhra
Steam
Low pressure, 45 psia 2.81 dollar/103Ibb
High pressure, 600 psia 3.45 dollar/103lbb
~

a Bought from public utility.

Self-generated from low sulfur fuel oil.
 DESIGN AND ECONOMICS OF ETHANOL PRODUCTION 1429

TABLE IV
Continuous Fermentation Design Basis
Specification Value
Sugar concentration 10%
Dilution rate 0.17 hr-'
Temperature 35°C
Cell yield factor, Y,,, 0.10
EtOH yield factor, Y,,, 0.45
Cell concentration in fermentor 10.0 dry wtfliter

the continuous fermentation process to produce 78,000 gal 95%

EtOH/day. The principal items of equipment corresponding to the
flow sheet are listed in Table V.
The molasses is first diluted to a 10% sugar solution and mineral
supplements are added. Sterilized by steam injection, the molasses
solution is distributed to eight 1.89 X lo5liter continuous fermentors
(working volume of each fermentor is taken at 80% of total fermen-
tor volume), each operating at a dilution rate of 0.17 hr-l. A low
flow of air (2.5 X v/v/m) is sparged through each fermentor to
maintain the oxygen tension at the optimum level of 0.07 mm Hg.
The fermented beer then passes to four continuous centrifuges and
the yeast is removed. (Only two centrifuges are shown in Fig. 4.)
The yeast is subsequently dried and stored for sale as a protein feed

Fig. 5. Flow diagram and mass balance for continuous fermentation. Capacity is
78,000 gal 95% EtOH/day (flows in lo3 Ibhr) cw = cooling water.
1430 CYSEWSKI AND WILKE

TABLE V
Major Items of Equipment for Continuous Ethanol Fermentation Planta
No. of C o s t h i t
Item Unit specification units (d01lar)~
-
Ethanol fermentation
Fermentor 1.89 x lo5 liter vol., SS cons. 8 90,500
Agitator 14 HP, SS cons. 8 6,500
Air compressor 91 HP, centrifuger type, 30 1 38,000
PSk
Air filter 0.4 X 0.3 m, glass fiber 8 2 10
Media sterilizer 8.7 x 1.2 m, insulated SS 1 11,700
Pipe
Preheat exchanger coupled 10,000 ft2, SS cons. 1 112,500
with sterilizer
Cooler exchanger coupled with 4,100 ft2, SS cons. 1 64,000
sterilizer
Heat removal exchanger 410 ft2, SS cons. 8 14,500
coupled with fermentor
Solid feeders screw conveyor, 4 todday 4 1,600
Nutrient mixing tank and 1.03 x lo5 liter vol, SS cons. 1 45,300
agitator
Sugar solution storage tank 2.48 x lo6 liter vol, SS cons. 1 270,000
In-plant beer storate 4.14 x lo5 liter vol, SS cons. 1 26,700
Centrifuge nozzle-type bowl, 40 H P 4 62,000
Yeast spray dryer 18 ft diam SS cons. 2 31,100
Product alcohol storage tank 5.9 x lo5 liter vol., CS cons. 1 21,600
Yeast storage tank 1.0 x lo4 liter vol, SS cons. 1 7,900
Pumps and drivers 10 4,800

Total 1,856,700

Ethanol recovery
Distillation column 11.2 ft diam 45 sieve trays, 1 76,200
CS cons.
Condenser 4,700 ftz CS cons. 1 69,900
Reboiler 2,600 ftz CS cons. 1 47,800
Preheat exchanger 4,200 ft2 CS cons. 1 64,500
Reflex tank 1.13 x lo3 liter vol, CS cons. 1 3,800
Ethanol absorber 7.2 ft diam, 26 ft high, 1 in. 1 25,700
rasching rings
Pumps and drivers 5 2,300

Total 299,400
a Capacity was 78,000 gal 95% EtOH/day. SS = stainless steel; cons. = construc-

tion; CS = carbon steel; vol = volume.

Costs are estimated for the second quarter 1975, Marshell-Stevens Index =
445.6.
 DESIGN AND ECONOMICS OF ETHANOL PRODUCTION 1431

supplement. The clarified beer from the centrifuges is next distilled

to concentrate the ethanol to 95 wt%. The distillation column con-
sists of a tower 11.2 ft in diam having 45 sieve trays and a reflux
ratio (mol reflux/mol product) of 7.33 is used to produce the 95 wt%
ethanol product. As a result, the steam requirement for distillation
is large amounting to 26.1 Ib steadgal 95% EtOH produced. An
absorber, using the distillate bottoms as the absorbing liquid, is
employed to recover ethanol lost in the exit gases (air and CO,)
from the fermentor. The ethanol-rich stream from the absorber is
also fed to the main distillation unit for final ethanol recovery.

Continuous Fermentation with Cell Recycle

The cell recycle process is identical to the continuous process
described above except that a portion of cell concentrate from the
centrifuges is returned to the fermentors. This increases the cell-
mass concentration in the fermentors permitting a higher volumetric
ethanol productivity. As a result, the total fermentation volume is
reduced in the cell recycle process requiring only three fermentors
of 1.22 X lo5 liter in volume to produce 78,000 gal 95% EtOH/day.
The reduction in fermentation volume is offset somewhat by an
increased load on the centrifuges. Seven centrifuges are needed in
the cell recycle fermentation to maintain a yeast cell concentration
of 50 g dry wtiliter in the fermentors. All other process equipment
is the same as that listed in Table V for conventional continuous
operation. A summary of the design basis of the cell recycle process
is listed in Table VI.

Batch Fermentation
The batch fermentation process parallels the continuous process
shown in Figure 5 , however, the fermentors are operated batchwise

TABLE VI
Continuous Fermentation with Cell Recycle Design Basis
Specification Value
Sugar concentration 10%
Dilution rate 0.7 hr-I
Temperature 35°C
Cell yield factor, Yx,s 0.10
EtOH yield factor, Y,,, 0.45
Cell concentration in fermentor 50.0 g dry wtniter
1432 CYSEWSKI AND WILKE

TABLE VII
Batch Fermentation Design Basis
Specification Value
Sugar concentration 10%
Fermentation time 16 hr
Fermentor down-time per cycle 6 hr
Temperature 35°C
Cell yield factor, Yxls 0.056
EtOH yield factor, Y p l s 0.477

instead of continuously. A 16 hr fermentation time was assumed

with an additional 6 hr required to fill, drain, and sterilize each
fermentor. The net result is that 26 fermentors of 1.85 X lo5 liter in
volume are required for the batch production of 78,000 gal 95%
EtOH/day. In addition, two 4.9 X lo4 liter seed fermentors are
required to produce a 2% inoculum for the main fermentors. All
other processing equipment is identical to that listed in Table V for
continuous operation. The design basis of the batch fermentation is
shown in Table VII. As indicated in Table VII, a 10% sugar feed
(for which fermentation kinetics were available from a past study6)
has been specified for batch fermentation. No attempt was made to
optimize batch fermentation with respect to sugar concentration
since, as shown below, the economics of continuous processes
proved far superior to batch processing.

Vacuum Fermentation with Cell Recycle

Figure 6 shows a schematic flow diagram of a vacuum fermen-
tation process to produce 78,000 gal 95% EtOH/day. The design
basis of the vacuum process is listed in Table VIII. The principal
items of equipment corresponding to Figure 6 are shown in Table
IX.

TABLE VIII
Vacuum-Recycle Fermentation Design Basis
~

Specification Value
Sugar concentration 50%
Dilute rate 0.18 hr-I
Temperature 35°C
Pressure 55 mm Hg
Cell yield factor, Y,,, 0.058
EtOH yield factor, Ypls 0.475
 DESIGN AND ECONOMICS OF ETHANOL PRODUCTION 1433

Fig. 6. Flow diagram and mass balance for continuous vacuum fermentation with
cell recycle. Capacity is 78,000 gal 95% EtOH/day (flows in lo3 Ibibr) cw = cooling
water.

Full strength molasses is first mixed with mineral supplements.

The medium is sterilized by steam injection and fed to a single 1.89
x lo5 liter vacuum fermentor operating at a total pressure of 55 mm
Hg and 35°C.Pure oxygen is sparged through the fermentor at a
rate of 0.1 v/v/m at STP to satisfy the trace oxygen requirement of
the yeast. As the fermentation proceeds, ethanol and water are
boiled away from the fermentation broth. The vapor from the fer-
mentor is compressed to 340 mm Hg and condensed in the fermentor
reboiler to supply the energy for the vaporization of ethanol and
water in the fermentor. After the vapor recompression cycle, the
uncondensable gases (carbon dioxide and oxygen) are compressed
to 760 mm Hg and cooled to 35°C to condense additional ethanol
and water. The fermentation gases are finally fed to an absorber
where the last traces of ethanol are removed. The fermented beer
is pumped to atmospheric pressure and fed to two continuous cen-
trifuges where the yeast concentrate is removed. A portion of the
yeast concentrate is returned to the fermentor. The remaining yeast
is spray dried and packaged for sale. The clarified beer from the
centrifuges and the condensation products are fed to a distillation
column where the ethanol is concentrated to 95%. A portion of the
bottoms product from the distillation column is cooled and fed to
the absorber. The compressors employed in the vacuum fermenta-
tion are driven by high-pressure steam turbines (600 psia). The
exhaust steam from the turbines is used in the distillation column
1434 CYSEWSKI AND WILKE

TABLE IX
Major Items of Equipment for Vacuum Ethanol Fermentation Plant”
No. of Cost/unitb
Item Unit specification units (dollar)
Ethanol fermentation
Fermentor 1.89 x lo5 liter vol, SS cons. 95,500
Agitator 110 HP, SS cons. 16,400
Compressor for vapor 3,000 HP, centrifuger type 523,800
recompression
Compressor for CO, 500 HP, centtifuger type 1 122,000
compression
Fermentor reboiler 4,400 ft2, SS cons. 66,500
Oxygen filter 0.5 x 0.3 m, glass fiber 360
Media sterilizer 8.7 x 0.64 m, insulated SS 5,500
Pipe
Preheat exchanger coupled 2,800 ft2, SS cons. 1 50,500
with sterilizer
Cooler exchanger coupled with 1,200 ftz, SS cons. 1 28.900
sterilizer
Solid feeders screw conveyors, 4 tontday 4 1,600
Nutrient mixing tank with 2.9 X 105 liter vol, SS cons. 1 27,200
agitator
Sugar solution storage tank 7.09 x lo5 liter vol, SS cons. 121,000
In-plant beer storage tank 1.18 x lo5 liter vol, CS cons. 11,900
Gas-liquid separators 1.5 x 103 liter vol, CS cons. 3,600
Secondary vapor condenser 100 ft2 SS cons. 25,800
Centrifuge nozzle-type bowl, 40 HP 62,000
Yeast spray dryer 18 ft diam, SS cons. 31,100
Yeast storage tank 1.0 x lo4 liter vol, SS cons. 7,900
Product alcohol storage tank 5.9 x lo5 liter vol, CS cons. 21,600

Total 1,382,300

Ethanol Recovery
Distillation column 10.2 ft diam 51 sieve trays, 1 74,300
CS cons.
Condenser 3,700 ft2, CS cons. 1 59,700
Reboiler 1,900 ft2, CS cons. 1 39,400
Preheat exchangers 200 ftz, CS cons. 2 9,100
Reflex tank 1.13 X lo3 liter vol, CS cons. 1 3,800
Ethanol absorber 9.5 ft diam, 70 ft high, 1 in. 1 58,900
rasching rings
Pumps and drivers 5 2,300

Total 265,800
a Capacity was 78,000 gal 95% EtOH/day. SS = stainless steel; cons. = construc-
tion; CS = carbon steel; vol = volume.
Costs are estimated for the second quarter 1975, Marshell-Stevens index =
445.6.
 DESIGN AND ECONOMICS OF ETHANOL PRODUCTION 1435

reboiler and supplies 63% of the distillation steam requirement

shown in Figure 6.

PROCESS ECONOMICS

Table X compares the fixed capital investment required for the

various fermentation processes to produce 78,000 gal 95% EtOH/
day from molasses. As shown in Table X, a substantial decrease in
capital investment is experienced in going from batch to continuous
processing. A batch fermentation process requires 225 dollars of
capital/gal EtOH produced each day. This figure is cut in half for
continuous fermentation requiring only 96 dollar/gal/day . The vac-
uum process requires the lowest total capital expenditure of only
5.2 million dollars or 66 dollar/gal/day . Whereas, the continuous-
cell recycle process requires the lowest capital expenditure for the
fermentation equipment; this is offset by an increased expenditure
for centrifuges used for yeast recovery to maintain a high cell
density in the fermentors.
Although the vacuum system requires only one fermentor, a
higher capital expenditure is needed for the fermentation equipment
in the vacuum system than in the cell recycle fermentation. This is
due to the large compressors and fermentor reboiler needed in the
vacuum process. The increased fermentation equipment cost is,
however, balanced by a reduction in the size of auxiliary equipment
(media sterilizer, storage and mixing tanks, centrifuges, and distil-
lation column) because a more concentrated sugar solution (50%)
is used in the vacuum process.

TABLE X
Required Fixed Capital Investment for Different Modes of Operationa
Fixed capital investment (lo3dollars)
Continuous- Vacuum-
Batch Continuous cell recycle cell recycle
Fermentation 14,900 4,808 2,484 3,366
EtOH recovery 928 928 928 824
Yeast recovery 962 %2 1,362 194
Storage 81 1 811 81 1 23 3
Total 17,601 7,509 5,585 5,217
a Plant capacity was 78,000 gal 95% EtOH/day.
1436 CYSEWSKI AND WILKE

ETHANOL PRODUCTION COSTS

Ethanol production costs excluding the cost of sugar and medium

supplements are shown in Tables XI and XII. As was expected,
reduction of capital investment is reflected in the production costs.
The largest decrease in production cost is achieved by using con-
tinuous operation rather than batch. Batch ethanol production costs
are 27.5 cent/gal while only 17.3 cent/gal is required for continuous
fermentation. A further decrease of production cost to 16.1 and 12.7
cent/gal is obtained in the continuous-cell recycle and vacuum fer-
mentations, respectively. As shown in Table XII, the low fermen-
tation cost of the continuous-cell recycle process is partially offset
by an increased cost for yeast recovery.
Almost 3.5 cent/gal can be saved with the vacuum system com-
pared to conventional continuous fermentation. The advantage of
the vacuum process is twofold. The 50% molasses solution must
not be diluted and the ethanol distillation cost (as reflected by the
steam cost) is reduced.
The reduced distillation cost in the vacuum process is due to the
preliminary concentration of ethanol achieved in the vapor re-
compression cycle used to maintain the fermentor vacuum. The
ethanol concentration is increased from a mole fraction of 0.03 in
the fermentation broth to 0.21 in the condensed vapor. The resulting
increased ethanol concentration in the feed to the distillation column

TABLE XI
Ethanol Production Costs for Different Operation Modes”
Production cost (cent/gal)
Continuous- Vacuum-
Batch Continuous cell recycle cell recycle
Investment related costs 10.3 4.9 4.0 3.5
Operating labor 3.2 0.9 0.5 0.4
Supervision and clerical 0.2 0. I 0.1 0.1
Utilities
Water 0.6 0.6 0.6 0.4
Power 1.2 0.6 0.9 0.6
Steam 10.1 9.5 9.5 6.8
Oxygen - - - 0.5
Laboratory changes 0.1 0.1 0.1 0.1
Plant overhead 1.8 0.6 0.4 0.3
Total 27.5 17.3 16.1 12.7
a Plant capacity was 78,000 gal 95% EtOH/day from 50% “cane” molasses sugar
solution.
 DESIGN AND ECONOMICS OF ETHANOL PRODUCTION 1437

TABLE XI1
Ethanol Production Costs for Different Modes of Operationa
Production cost (cent/gal)
Continuous- Vacuum-
Batch Continuous cell recycle cell recycle
Fermentation 16.8 6.6 5.0 4.8
EtOH recovery 8.1 8.1 8.1 6.7
Yeast recovery 2.0 2.0 2.4 1 .O
Storage 0.6 0.6 0.6 0.2
Total 27.5 17.3 16.1 12.7
a Plant capacity was 78,000 gal 95% EtOH/day

allows the use of a lower reflux ratio for the final concentration of
ethanol to 95%. By decreasing the required reflux ratio (i.e., mol
reflux/mol product) from 7.33 for atmospheric fermentations to 5.66
for vacuum fermentation, an overall steam savings of 17% is ob-
tained. This includes the steam required for compressor operation
in the vacuum system.
As discussed above, another advantage of the vacuum fermen-
tation scheme is elimination of end-product inhibition by boiling-off
ethanol as it is produced. However, by changing the fermentation
pressure, the equilibrium ethanol concentration is altered and the
ethanol concentration in the fermenting broth may be adjusted to
any desired level. Figure 7 shows the effect of ethanol concentration
on production costs for the vacuum system. When the ethanol
concentration of the broth is low, the equilibrium vapor concentra-
tion is also low. A high boil-up rate is thus necessary to remove the
ethanol produced during fermentation. This increases the vapor
compression costs in the recompression cycle and product costs
increase. At high ethanol concentrations the compression costs are
reduced, but fermentation costs increase because the yeast becomes
inhibited by the ethanol. As shown in Figure 7, these two competing
effects produce a rather flat production-cost curve between ethanol
concentrations of 5.0 and 8.0%.
The production cost of vacuum fermentation without recycle is
also shown in Figure 7. Production costs without cell recycle rise
more rapidly with increased ethanol concentration than when cell
recycle is employed. This stems from the overall mass balances.
When the ethanol concentration is high, the boil-up rate necessary
to remove the required amount of ethanol is low and, from the mass
balance, a large bleed rate from the fermentor is necessary. But, a
1438 CYSEWSKI AND WILKE

16

-
\
&
I
: Without cell recycle

With cell recycle

10 I 1 1 I I I I I

substantial amount of cell mass is removed with the bleed stream

and the yeast concentration in the fermentor decreases. Since the
cell-mass concentration decreases , the fermentation rate per unit
volume decreases and the fermentation costs increase. This, of
course, is not the case when cell recycle is employed because the
biomass concentration in the fermentor is maintained at a high level
by returning a portion of the yeast to the fermentor.

Effect of Feed Sugar Concentration

Ethanol production costs are plotted against the feed concentra-
tion of fermentable sugar in Figure 8 for conventional continuous
fermentation. There is a definite optimum production cost at 10%
sugar feed. Above a 10% sugar concentration ethanol inhibition
slows the fermentation rate. As a result, a larger fermentation vol-
ume must be used and as shown in Figure 8, the fermentation costs
increase. Below 10% sugar feed, the cell-mass concentration de-
creases, lowering the fermentor ethanol productivity which again
increases the fermentation costs. Also, at low feed sugar concen-
trations, dilute solutions of ethanol are produced. This increases
the distillation cost because more energy is required to concentrate
these dilute solutions to 95% ethanol.
An optimum sugar concentration of 10% also exists for continu-
ous-cell recycle fermentations; however, the fermentation cost
 DESIGN AND ECONOMICS OF ETHANOL PRODUCTION 1439

curve differs from that shown in Figure 8. When cell recycle is

employed, the cell-mass concentration in the fermentor is not a
function of sugar concentration. Therefore, at sugar concentrations
below 10% the productivity does not decrease but rather increases
because of less ethanol inhibition. This increased productivity is
somewhat counterbalanced by an increased centrifugation require-
ment since higher flow rates through the centrifuges are required
for low sugar concentrations. The net effect is that the fermentation
cost curve gradually decreases with decreasing sugar concentrations
below 10%. But, distillation costs sharply increase in this region as
shown in Figure 8. Hence, total ethanol production cost increases
as the sugar concentration decreases below 10%. Above 10% sugar,
ethanol inhibition becomes the controlling factor and drives up the
fermentation cost, and the total production cost in the recycle sys-
tem increases.
The effect of sugar concentration on ethanol production cost for
the vacuum fermentation process is illustrated in Figure 9. The
production costs shown are exclusive of any sugar concentration

0 4.0 8.0 12.0 16.0 20.0

Percent sugor in feed

Fig. 8. Effect of sugar concentration on EtOH production costs for continuous

fermentation. (0)Total; (A) fermentation; (0)
distillation.
1440 CYSEWSKI AND WILKE

I I I I I I
10 20 30 40 50 6c
Per cent sugor in feed

Fig. 9. Effect of sugar concentration on EtOH production costs for vacuum

fermentation. (0) Continuous fermentation; ( 0 )vacuum fermentation.

costs. As the sugar concentration increases, the fermentation cost

steadily declines since ethanol inhibition is not a problem in the
vacuum fermentation and high sugar concentrations allow a reduc-
tion in the size of process equipment. However, there is a practical
limit to the sugar concentration which may be employed. Extremely
concentrated solutions are difficult to pump because of high vis-
cosity. Also, as the sugar is concentrated, so are nonvolatile con-
stituents, such as minerals and salts. At high concentrations these
components may become toxic to the yeast. Without knowledge of
the exact optimum, a total sugar concentration of 50%, equal to
that of full strength cane molasses, was set in the vacuum system.
It should be realized, however, that molasses may contain sub-
stances (i.e., sulfur dioxide, hydroxymethylfurfural, potassium ini-
dodisulfonate) which inhibit fermentation, depending on the source
and manufacturing process of the m 0 1 a s s e s . ~Hence,
~ ~ ~ more work
is required to characterize fermentation kinetics on full strength
molasses before a design can be finalized. In this respect, the vac-
uum fermentation process design must be viewed as tentative, il-
lustrating the maximum potential of vacuum operation.
 DESIGN AND ECONOMICS OF ETHANOL PRODUCTION 1441

Total Ethanol Production Costs

The total ethanol production cost includes the cost of molasses
and any medium supplements that are required. Based on simple
stoichiometry, every centflb fermentable sugar costs adds 14.3 cents
to the manufacturing cost of 1 gal of 95% EtOH. Thus, it is easily
seen, if sugar costs are above 2 to 3 centflb, the sugar cost will
dominate the economics of ethanol production. Unfortunately, this
is the present situation.
A comparison of total ethanol production costs between contin-
uous-cell recycle and vacuum-cell recycle fermentations is given in
Table XIII. The cost of molasses was taken at 5.0 centflb sugar,14
and 97% of the sugar was assumed to be fer~nentable.~ The medium
supplement cost was based on a representative composition of cane
molasses and the mineral composition of the yeast.g The medium
costs are summarized in Table XIV. The only additions required
are mineral nitrogen, magnesium, and phosphorus. Sufficient quan-
tities of other minerals and biotin are present in the raw molasses.
A by-product credit of 10 centflb yeast was subtracted from the
total production cost. The yeast credit was reduced from the current
selling price of 40 centflb for S . cerevisiae,15 to reflect added costs
for marketing and distribution. But more important, if industrial
ethanol was produced entirely by fermentation, the tremendously
increased supply of yeast would drive down the selling price.
Hence, in the absence of a complete marketing study, a conserva-
tive price of 10 centflb yeast was assumed.

TABLE XI11
Total Ethanol Production Costsa
Continuous- Vacuum-
cell recycle cell recycle
Fermentation 5.0 4.8
EtOH recovery 8.1 6.7
Yeast recovery 2.4 1.0
Storage 0.6 0.2
Sugar 73.7 71.4
Medium supplements 5.8 2.7

Total production costs 95.6 86.8

Yeast credit 13.3 6.2

Net production costs 82.3 80.6

a In cent/gal 95% EtOH.
1442 CYSEWSKI A N D WILKE

TABLE XIV
Medium Supplement Costs for Continuous-Cell Recycle
Operationa
Raw
molasses
Component (gfliter) Dollar/ton Todday
Ammonium 21.1 90 22.1
Potassium phosphate 6.6 120 6.9
Magnesium sulfate 1.5 110 1.6
a Y,,$. = 0.10.

As shown in Table XIII, the total ethanol production cost for the
continuous-cell recycle system is 95.6 cent/gal and for the vacuum
system is 86.8 cent/gal. However, after the yeast credit has been
subtracted, the continuous-cell recycle fermentation appears more
attractive requiring only 82.3 cent/gal as compared to 80.6 cent/gal
for the vacuum fermentation. Thus, the net production costs are
almost identical for the recycle and vacuum systems, even though
total production costs are less in the vacuum system. This is due to
the lower cell yield factor and hence, reduced yeast by-product
credit obtained in the vacuum fermentation (see Tables VI and
VIII).
The cost of sugar does indeed dominate the ethanol production
cost, representing over 75% of the total manufacturing cost. How-
ever, the net production costs of 80.6 and 82.3 cent/gal, after the
yeast credit has been subtracted, compare favorably with the cur-
rent selling price for 95% EtOH of 1.10 dollar/gal.15
Since the processing costs represent about only 16% of the total
ethanol production costs, the effect of property taxes and labor

TABLE XV
Incremental Effect of Variables on Ethanol
Production Cost
Increase of
EtOH cost
Variables (cent/gal)
5.60 dollarhr + 10.0dollarhr increase 0.5
in labor costs
4.0% + 12% increase in taxes 3.6
5.0 centflb + 8.0 centflb increase in 44.2
molasses sugar costs
 DESIGN AND ECONOMICS OF ETHANOL PRODUCTION 1443

rates on the final ethanol cost is minor. This is shown in Table XV

together with the effect of sugar costs.

CONCLUSIONS

The cost of molasses represents over 75% of the total ethanol

production cost and thus will dictate the selling price of fermentation
ethanol. Viewed from this point, fermentation process improve-
ments have a minor effect on total production costs. However, the
process improvements described in this work have a substantial
effect on required fixed capital investment and will thus appreciably
change the profitability of ethanol production via fermentation. The
simple return on investment (ROI) before taxes for the various
fermentation processes is shown in Table XVI. The current selling
price of 1.10 dollar/gal 95% EtOH was assumed, and working cap-
ital, consisting of one month operating expenses and a one-month
surplus of product ethanol, was included in the total fixed capital
investment used for the ROI calculations. The ROI for each process
is shown both for the case when a 10 centnb yeast by-product credit
is taken and for the case when the yeast by-product credit is ne-
glected.
With a yeast credit of 10 cent/lb, batch fermentation yields only
18.5% return on investment (ROI). Since batch fermentation is the
currently accepted technology for ethanol fermentation, this low
ROI in part explains why ethanol is not presently manufactured by
fermentation. But if either of the continuous processes are consid-

TABLE XVI
Comparison of Return on Investment before Taxes between
the Various Fermentation Processes"
Percent return on
investmenth

Yeast credit
taken at 10 No yeast
Fermentation processes centilb credit
Batch 18.5 3.3
Continuous 55.6 27.9
Continuous with cell recycle 69.7 36.5
Vacuum with cell recycle 81.5 64.3

'' Selling price of 95% EtOH taken at 1.10 dollar/gal.

" Percent return on investment = (yearly profithotal capital
investment) x 100.
1444 CYSEWSKI AND WILKE

ered, the ROI ranges from 55.6 to 81.5%. In this light, fermentation
process improvements have a pronounced effect and could lead to
the profitable production of fermentation ethanol.
In conclusion, it should be remembered that this study has not
attempted to optimize actual fermentation processes which would
be found in an industrial environment. Rather, its purpose has been
to show the relative cost of various fermentation schemes. How-
ever, based on the preliminary (and tentative) process designs pre-
sented, ethanol production via continuous fermentation processes
appears presently profitable and will become more profitable as the
price of petroleum increases.

References
1. C. R. Wilke, R. Der Yang, and U. Von Stockar, Biotecknol. Bioeng. S y m p . .
6, 155 (1976).
2 . H. E. Grethlein, “The acid hydrolysis of refuse,” in Cellulose a s a Chemical
and Energy Resource. C. R. Wilke, Ed. (Wiley, New York, 1975).
3. W. A. Scheller, “Agricultural alcohol in automotive fuel,” presented at Eighth
National Conference on Wheat Utilization Research, 1973.
4. G. R. Cysewski and C. R. Wilke, Biotechnol. Bioeng., 18, 1297 (1974).
5. G. R. Cysewski and C. R. Wilke, Biotechnol. Bioeng.. 19, 1125 (1977).
6. G. R. Cysewski and C. R. Wilke, “Fermentation kinetics and process eco-
nomics for the production of ethanol,” Lawrence Berkeley Lab Report, LBL-4480,
March 1976.
7. T. W. Cowland and D. J. Maule, .I. Insr. Brew., 7 2 , 480 (1966).
8. J. White and D. G . Munns, Wallerstein Commun., 14, 149 (1951).
9. G. Reed and H. J. Peppler, Yeast Technology (Avi, Westport, Conn., 1973).
10. S. C. Prescott and C. G . Dunn, Industrial Microbiology (McGraw-Hill, New
York, 1959).
11. M. S. Peters and K. P. Temmerhause, Plant Design and Economics f o r
Chemical Engineers (McGraw-Hill, New York, 1968).
12. K. M. Guthne, C h e m . Eng., March 24 (1%9).
13. S. Burrows, “Baker’s yeast,” in The Yeasts, A. H. Rose and J . S . Harrison,
Eds. (Academic, New York, 1970), Vol. 3.
14. Wall Street Journal, May (1975).
15. Chemical Marketing Reporter, October (1976).
16. A. A. Andreasen and T. J. B. Stier, Cell Comput. Phys., 41, 23 (1953).

Accepted for Publication January 24, 1978