EPA-66 0/2-74-925
May 1974
Environmental Protection Technology Series
Protein Production
From
Acid Whey Via Fermentation
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodology to repair or prevent environmental
degradation from point and -non-point sources of
pollution. This work provides the new or improved
technology required for the control and treatment
of pollution sources to meet environmental quality
standards.
-------
EPA-660/2-74-025
May 1974
PROTEIN PRODUCTION FROM ACID WHEY
VIA FERMENTATION
By
Sheldon Bernstein/ Ph.D.
Thonas C. Everson, Ph.D.
Grant #8-800747
Program Element 1BB037
Project Officer
Kenneth A. Dostal
National Waste Treatment Research Program
National Environmental Research Center
U.S. Environmental Protection Agency
Corvallis, Oregon 97330
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON D. C. 20460
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 2M02 - Price $1.25
-------
EPA Review Notice
This report has been reviewed by the Office
of Research and Development, EPA, and
approved for publication. Approval does not
signify that the contents necessarily reflect
the views and policies of the Environmental
Protection Agency, nor does mention of trade
names or commercial products constitute
endorsement or recommendation for use.
ii
-------
ABSTRACT
Prom the operation of a demonstration pilot plant over
extended periods of time, it has been shown that yeast may
be grown on an acid whey or sweet whey medium in a contin-
uous, deep tank aerated fermentor. Variations in
fermentation conditions, strain selection, and medium com-
position produced cell concentrations of several billion
cells per milllliter. By a process of evaporation and spray
drying the whole fermented whey mass and the utilization of
the evaporator condensate to dilute Incoming condensed whey,
a high grade, non-toxic, protein feed material may be produced
without any effluent streams. Amlno acid analyses and protein
efficiency ratios are presented for this feed material.
Economic estimates show that while a large capital investment
and low cost raw material are required for the commercial
feasibility of this fermentation process, it will be compet-
itive with other methods for the manufacture of single cell
protein. This whey fermentation is one means of converting
large quantities of a potential environmental pollutant into
a useful and needed product.
ill
-------
CONTENTS
Section
Page
II
III
IV
V
VI
VII
VIII
IX
X
XI
XII
XIII
Summary and Conclusions - International
Minerals Corporation Preliminary
Studies
Conclusions & Recommendations
Amber Laboratories
Introduction
International Minerals Corporation
Materials and Methods
International Minerals Corporation
Experimental Results
International Minerals Corporation
Introduction
Amber Laboratories
Materials and Methods
Amber Laboratories
Results and Discussion
Amber Laboratories
Summary
Amber Laboratories
Acknowledgements
References
Glossary
Appendices
6
11
28
29
35
58
63
64
65
67
-------
FIGURES
PAGE
GROWTH OP SACCHAROMYCES FRAGILIS
AT DIFFERENT INOCULUM LEVELS UNDER
CARBOHYDRATE FEED CONDITIONS MAIN-
TAINING 5-6# CHO IN THE BROTH 18
GROWTH OF SACCHAROMYCES LACTIS FROM
VARIOUS INOCULUM LEVELS, GROWN UNDER
CONDITIONS OF MAINTENANCE OF 5-6# CHO
IN THE BROTH 20
3 SEMI -CONTINUOUS OPERATION, SOP, 0
NH^OH, SACCHAROMYCES FRAGILIS 22
4 CONTINUOUS OPERATION OF SACCHAROMYCBS
FRAGILIS WHEY FERMENTATION IN A 3 STAGE
SYSTEM
5 BLOCK DIAGRAM OF CLOSED-LOOP SYSTEM FOR
MINIMIZING FERMENTATION EFFLUENTS
6 SEMI -CONTINUOUS GROWTH OF SACCHARQMYCES
FRAGILIS USING S0P MEDIUM-2
7 CONTINUOUS GROWTH OF SACCHAROMYCES
FRAGILIS USING SOP MEDIUM-2 (59 HR) 46
8 CONTINUOUS GROWTH OF SACCHAROMYCES
FRAGILIS USING SOP MEDIUM-2 (220 HR) -4$
9 RAT GROWTH RATES ON CASEIN AND FERMENTED
WHEY PRODUCTS 53
vi
-------
TABLES
No. Page
1 Summary of Experimental Data 12
2 . Experiment #18. Continuous operation of
a Whey Fermentation with Saccharomyces
fragilis in a 3-Stage System 24
3 Experiment #4. Amino Acid Analyses,
Inoculum Level Effect 2!?
4 ~ Summary of Fermentation Experiments
EPA Project S-800747 36
5 Comparison of Gross Composition and Amino
Acid Content of Various Single-Cell Proteins 50
6 Amino Acid Content of Whole Wheat, Commercial
Brewers Yeast and Saccharomyces fragilis
Yeast compared to FAO Profile 5!
7 Protein Efficiency Ration (PER) Assays of
Several Typical Saccharomyces fragilis
Yeasts 52
8 Annual Production of Fermented Whey Mass
Solids vs Fermentor Size 55
9 Production Cost Estimates of Various Single-
Cell Proteins 57
vii
-------
SECTION I
SUMMARY AND CONCLUSIONS
INTERNATIONAL MINERALS CORPORATIONrPRELIMINARY STUDIES
1. A aeries of experiments were conducted in 44 liter, pilot
scale fermentors using several strains of lactose-fermenting
Saccharomyces. The process of conversion of lactose in whey
to yeast cellular material was reasonably efficient with con-
version rates of 45-55# recorded. Therefore, the literature
conversion rate of 55# was attained (ll).
2. The fermentation process was completed in 8-10 hours and
could be operated either in a serai-continuous manner with
minimum "downtime" or continuous manner without "downtime".
3. Cell yields, cell counts, and crude protein concentrations
were approximately equivalent for the various fermentation
processes.
4. From a fermentation standpoint the process is simple to
operate and maintain (non-sterile) and should become an
effective means of converting whey to a marketable commodity.
-------
SECTION II
CONCLUSIONS AND RECOMMENDATIONS
AMBER LABORATORIES
1. Saccharomyces fragilia can be produced at growth levels
of several billion cells per milliliter using a deep tank
aerated fermentor operated in a continuous manner on acid or
sweet whey.
2. The complete whole fermented whey mass may be dried
eliminating effluents and produce a satisfactory feed
supplement. The yield of fermented whey mass (PWM) was 0.675
to 0.75 pounds per pound of whey solids.
3. A superior quality yeast, high in protein and low in ash
content, may be produced by harvesting the yeast cells. The
product should find commercial applications in specialty feeds
and in foods as a food additive.
4. The fermentation could be carried out on permeates from
ultraflltratlon (UF) of acid or sweet whey. However, numerous
problems were encountered using the UF unit such as low flux
rates due to membrane fouling, inability to properly sanitize
the UF unit, and extremely low solids in the lactose permeate.
Therefore, the fermentation of lactose permeate from the UF
system studied does not look economically feasible.
5. A large capital Investment and low cost, raw material are
required for the commercial feasibility of the fermentation
process. These studies were limited to the laboratory and
pilot development of protein production from acid and sweet
-------
whey via yeast fermentation. It is recommended that the
sealed-up testing of protein production from whey fermentation
in commercial sized fermentora be examined which would permit
full-scale, long term evaluation of the chemical, physical,
biological and economic aspects of the process.
-------
SECTION III
INTRODUCTION
INTERNATIONAL MINERALS CORPORATION
The preliminary investigation and shake flask studies were
done by International Minerals Corporation, Libertyville,
Illinois. Various Standard Operating Procedures were devel-
oped and used by Amber Laboratories in the operation of the
500 gallon, Demonstration Pilot Plant fermentor. The results
of the Amber Laboratories studies are presented in Sections
VI-IX.
The carbon of whey lactose can be converted by a variety of
lactose-fermenting yeasts to yeast cellular components of
which 50$ are protein. An exogenous source of inorganic
nitrogen is usually required as a supplement since whey pro-
tein is not a readily available N-source for these saccharo-
lytic strains. It has been considered that if fermentation
processes were devised which could result in near theoretical
conversion of lactose to yeast protein in a suitably short
fermentation cycle then such a process might prove a feasible
means of producing a low cost protein feed supplement from
the high tonnage of whey which is currently considered waste
material.
As discussed in a preliminary report (international Minerals
Corporation, Section XIII Appendices) shake flask experiments
demonstrated that sufficient doublings of yeast populations
in short fermentation cycles could be obtained on simple whey
-------
media to warrant further investigation of the process in pilot
fermentors (15 gallon). Plant throughputs of 2-3X fermentor
capacity per 24 hour period appeared feasible. The shake
flask experiments also demonstrated that such a process might
be run under non-sterile conditions.
The following series of experiments was carried out in fer-
mentors as an extension of this work. A progression from
batch to semi-continuous to continuous fermentor operation was
carried out to demonstrate the optimum method for use of
fermentor capacity and feasibility of tshe whey to yeast con-
version process.
-------
SECTION IV
MATERIALS AND METHODS
INTERNATIONAL MINERALS CORPORATION
Two yeast strains, Saccharomyces fragilis (NURL, Y-1109) and
S.. lac t is (NURL, Y-1140), were used throughout the ferment or
experiments. The strains were maintained on a weekly transfer
schedule on sterile whey agar.
Whey concentrate as received contained 35-40# total solids and
a lactose content of 35-38#. The high lactose concentrate Has
prepared from locally collected raw acid whey by Amber Lab-
oratories at Juneau, Wisconsin, and transported hot to the IMC
fermentation pilot plant. Whey for the experiments described
later was diluted to appropriate levels of lactose with tap
water and then used to prepare Medium 5102-A or Medium 5102-B.
Prior to dilution, whey concentrate was heated to Just below
boiling to insure complete solubillzation of lactose.
Whey agar was prepared by diluting whey concentrate with tap
water to 5% - Q.5% reducing sugar content and adding the com-
ponents of 5102-A Medium described later. The pH was adjusted
to 7-0 with 10$ XaOH and agar added to a 2£ concentration.
The medium was heated to melt the agar, tubed in suitable
aliquots and sterilized for 10 minutes at 120°C.
Two basal media were used throughout these studies. Medium
5102-A consisted of whey concentrate that was adjusted to
lactose concentration. The diluted concentrate was made to
-------
contain (W/V) 0.5$ (NH^)2SOK; 0.5$ K2HPO^ and 0.1$ Amber BYP-
300. The pH was adjusted to lj.. 5 for S. fragills and 5-5 for £>.
laotis with 10$ H3POK. The medium was not sterilized for
fermentor use. Medium 5102- B was prepared from whey concen-
trate adjusted to "-6$ lactose but made to contain (W/V) 0.9$
NH^OH; 0.3$ Amber BYF-300; 0.001$ PD-82 an t if o am (Hodag).
A series of nine 15-gallon (l^ liter) fermentation vessels was
used throughout these experiments. The fermentors were of the
jacketed type equipped with automatic pH, antifoam, temperature
and nutrient feed controls. They were aerated by compressed
sterile air entering through a bottom sparger and agitated by
a standard shaft and impeller system powered hydraulic ally
through magnetic shaft couplings. Steam could be directly in-
jected into the vessel or circulated through the jacket. Cool-
ing water could also be circulated through the jacket. The
vessels were fitted with various inoculating and sampling ports
and could be fitted to handle oxygen or carbon dioxide sensors.
The series of fermentors were inter-connected by header lines
allowing complete flexibility of material transfer from vessel
to vessel. The series of 15-gallon fermentors was served by
three 15-liter New Brunswick fermentors which served as seed
vessels. The New Brunswick fermentors were of standard in-
ternal baffle design and were connected to the 15-gallon
fermentor header line.
Two procedures were used for inoculum and are designated
appropriately in the experimental summary (Table 1). The
-------
procedures are outlined below:
Inoculum Buildup Procedures:
I
Stock Slant
II
Stock Slant
Wash growth off with
10 ml medium, inoculate
into
liters 5102-A in 15 liter ferraentor
liters air/m; 200 rpra; 0.5 PSIG
back pressure; 30°C
1
8 hours J.
Add 6 liters 5102-A
16 hours I 1 gal to:
9 gal 5102-B in 15 gal seed fermentor
200 rpm; 2.32 V/V air/m; 10 PSIG
back pressure; 30°C
7-8 hours
Use as seed as indicated
I
500 ml 5102-A In 2
liters
Incubator on
rotary shaker at
30°C for 2t| hours
8 gal 5102-A in 15 gal
seed fermentor 200 rpm;
2.32 V/V air/m; 30°C;
1U-16 hr
Use as seed at 1-2x10
cells/ml as indicated
Fifteen gallon fermentors were inoculated from either of the
trains indicated above as noted for individual experiments.
The 10$ level was standard. Fermentor operating conditions
unless otherwise noted were 200 rpm; 2.32 V/V air/min; 10 PSIG
back pressure; 30°C. The fermentors were not sterilized nor
was pH controlled, Fermentors were loaded with 9 gal of medium
and inoculum transferred from a seed vessel to bring the total
fermentor volume to 10 gal (10$ inoculum).
A three stage continuous fermentation procedure in the 15
gallon fermentors utilized an inoculum development as in I
above. Three fermentors were simultaneously loaded and
8
-------
inoculated. After six hours of inoculum development the fer-
mentors were loaded to 9 gallon volume with 5102-B medium and
a continuous feed of 5102-B medium to the first stage initiated.
The rate of feed was calculated to accomplish total utilization
of lactose by completion of the third stage. The rate of con-
tinuous feed was ^3.18 gal/hr (lLj.,0 liters/hr).
Solids determinations were made as described below: A sample
representing approximately 1 gram of dried solids was weighed
into a previously dried, cooled, and weighed.5 cm circular
aluminum flat-bottomed dish containing reagent grade seasand
(20/lj.O mesh). The dish and contents were heated at 60°C over-
night at full vacuum (-^30"). The dish was cooled in a dessi-
cator and then reweighed. Protein was determined in accord
with the A.O. A.C. method (llth Ed. 1970 #16.035 ref 2.651).
A factor of 6,38 x percent-N was used as recommended for dairy
product protein. Fat was determined in accord with the A.O.A.C.
method (llth Ed. 1970 #16.052) for dairy products. Ash was
determined in accord with the A.O.A.C. method (llth Ed. 1970
#16.031^) for dairy products. Free and total reducing sugar
was determined using a Technioon Autoanalyzer. The Technicon
Ferricyanide colorimetric method was used to determine reducing
sugars before and after hydrolysis. Total amino acids were
determined by over-night hydrolysis in 6N HC1 followed by
column chromatography in a Beckman 120C Amino Acid Analyzer.
Cell count analyses on samples of appropriate dilutions of
whey fermentor broth were made in 0.1 M phosphate buffer, pH
-------
7.0. Both chambers of a hemooytometer (Spencer Bright Line
Improved Neubauer 1/10 mm deep) were loaded with the same
diluted sample and only the large subdivided center square was
used In counting.
Q
If cell numbers were <10 /ml the entire square was counted,
the microscopic counts from both chambers were averaged and
cells/ml calculated. If the two counts deviated from each
other, the hemooytometer was reloaded, two more counts taken
and the four counts averaged. In counting samples with high
cell numbers, three rows of five squares each were counted for
both chambers, the six counts averaged and cells/ml calculated
in accord with the following equation:
cells/ml = count x dilution x k x 10 o
' ^ -
# small squares counted
Cell volumes were determined by centrifuglng a 10 ml broth
sample in a conical, graduated centrifuge tube at 25>0 rpm for
30 minutes in an International PR-2 centrifuge equipped with
a fixed rotor. Volume of the pellet was read, in ml, directly
from the tube.
10
-------
SECTION V
EXPERIMENTAL RESULTS
INTERNATIONAL MINERALS CORPORATION
The series of experiments described below were performed to
determine the effects of inoculum size (Bxpt's 4,5), medium
Ingredients (Expt's 6-14), carbohydrate level (Expt's 12-1?)
and physical conditions (Expt's 16-17) on the two strains of
yeast under study. In certain of the experiments operation
of the fermentation under semi-continuous (Expt's 11,12,15)
or continuous (Expt 18) conditions was examined. Experimental
conditions and parameters are summarized In Table 1. The data
obtained for maximum cell yield, total solids, dry weight of
protein (as is), dry weight of washed protein, % conversion
and % alcohol (ethanol) are provided where such determinations
were made on a particular run.
The effect of amount of Inoculum transferred to the fermentor
on cell growth and yield were determined in Expts 4 and 5 for
S. frag11is and S. lactls. respectively. Growth of S. frag11is
_ T* ————' —
under different inoculum size conditions is summarized graphi-
cally in Figure 1. The experiments were conducted under con-
ditions in which carbohydrate level was continuously maintained
at 5-6£. The data obtained were interpreted as reflecting the
inhibitory effect of excess whey on growth of the yeasts which
had been observed previously in shake flasks. The effect was
not grossly apparent during the first four hours of fermenta-
tion. Doubling times for the yeasts calculated for this time
11
-------
Table 1. SUMMARY OP EXPERIMENTAL DATA
ro
Max, Pop. %
cells/ral
T.S.
Condition Medium xlCP "as is"
Experiment #4. Effect of Inoculum
5-6$ CHO
W-10
*
44$ Inoculum
25$
15$
5$
10$ >
t
10$ Whey only
4
10
10
10
8
1.155
1.0?
1.07
.885
.795
.805
Experiment #5- As in 4 S. lactis
W-10
44$ Inoculum
25$
15$
5$
10$ 4
p
2$ Whey only
10
10
10
10
10
10
8.75
6-5
9.5
3.2
8.4
3.3
size on
13.05
11.30
10.40
8.9
8.8
8.2
12.7
10.5
9.5
9.1
8.1
8.9
Dry Weight
Protein Protein $ $
''as is" washed Conversion Alcohol
S. fragllis fermentation
18.03 41.99
19.08
19.84
20.42
22.22 43.60
15.51
"Numbers in parentheses refer to hoars of fermentation at which maximum cell population
was reached.
-------
Table l. (eont.)
Experiment #6. Effect of level of BYP and ammonium salts on S.. fragilis fermentation
W-10
In
BYP
0.1*
0.3
0.5
0.1
0.3
0.5
1.0*
1.0
1.0
0.5
0.5
0.5
10
10
Q
Q
ft
Q
1.31
1.78
1.40
1.56
1.30
1.03
6.15
6.45
6.55
5.51
6.00
6.25
36.42
27.95
28.13
43.58
42.49
23.74
29.15
28.34
30.88
27.33
36.23
Experiment #7. As in 6 S. lactis
BYP
0.1*
0.3
0.5
0.1
0.33
0.5
(MH4)2S04
1.0*
1.0
1.0
0.5
0.5
0.5
10
10
10
10
10
10
2.25
1.97
1.97
2.37
2.32
2.10
Experiment #8. Effect of N-source
BYP
0.3*
4r
-NltyOH-KHgPO^-NS
0.5* 0.5*
0.7 0.5
1.0 0.5
0.5 1.0
0.45* 0.5
0.90 0.5
10
10
10
10
10
8
2.91
1.97
2.35
1.90
1.64
(1.57)
6.15
6.55
5.30
5.20
5.40
5.40
and P- level
6.15
6.00
6.00
5.35
5.05
6.00
33.44
32.88
35.00
26.44
26.62
28.00
on S.
26.25
27.50
26.25
38.31
27.81
34.50
45.36
45.74
44.78
46.93
48.23
46.45
fragilis whey
45.74
37.51
W.09
42.75
45.38
46.25
46.84
37.77
39.26
fermentation
36.83
61.03
56.15
-------
Table 1. (cont.)
Experiment #9. Aa in 8 S. lactia
N-OH
-
-
-
•
0^90
W T)f^» UQ
njLinf\Jh HO
. 5* 0 . 5*
.7 0.5
1.0 0.5
0.5 1.0
0.5
0.5 - 1
Experiment #11. If fed
10
10
10
10
8
1 8
3.15
3.11
3.08
3.26
(2.17)
(2.39)
t of NH^OH/ft
,J
5.4o
5.95
6.20
5.55
5*80
PO. medium
29.67
25.01
27.88
36.43
28.97
34.32
on S.
41.09
36.11
37.13
fragilis
53
55
54
.72
.66
.36
fermentation
NH^OH adj. HgPO^
1 0.
1A
IB
2
2A
2B N
3 o.
3A
3B
4
4A
4B >
5 0.
9t
v
45
f
9
5A 1
5B +
8) 1.10
10 1.08
9
10
10
9
8
10
9
8
10
Q
10
10
Q
1.22
1.12
1.31
1.08
1.23
1.08
1.29
1.05
1.01
1.22
1.03
1.10
1.20
_
6.80
6.65
6.40
7.40
_
31.58
25.47
24.88
28.96
_
33.30
33.56
33.56
36.17
50
55
52
52
.09
.5
.42
.11
-------
Table I. (oont.)
Experiment f 12
. As In 11 S. lactis
MH^OH
1 0.
1A
IB
2
2A
2B >
3 0.
3A
3B
4
4A
4B >
9*
f
45
f
5 0.9
5A
5B ^
\
,
Experiment #13
«v°*
0.5*
:
-
-
MH^OH
0.9*
0.45
0.90
0.90
0.45
0.90
94
9
9
Q
Q
9
9
Q
Q
9
9
9
0
Q
Q
3.09
3.65
3.30
3.01
3.40
3.00
2.57
3.10
3.00
2.98
2.75
2.80
3.00
3.20
3.00
6.95
6.50
6.30
7.05
7.00
_. Effect of K* "source and
Lactose
5-6 1
5-6
5-6
10-11
10-11
10-11
[10
10
14
14
14
1.49
1.32
1.29
1.23
1.23
1.25
7.90
6.25
7.15
11.75
10.40
10.60
31.
31.
29.
30.
33.
lactose
29.
32.
32.
26.
25.
27.
33
96
92
11
30
level
54
98
98
16
01
88
35.91
34.46
39.08
39.08
37.50
on S.
34.83
40.38
38.73
31.64
36.24
34.13
49
48
40
40
45
fragilis
37
28
33
26
21
24
.05
.72
.10
.24
.07
fermentation
.07
.10
.60
.80
.34
.30
2.05 PWC
2.33
2.63
1.75
0.78
1.05
-------
Table 1. (eont.)
a\
Experiment #14. As in 13 S. lactis
KHgPO^ HH^OH Lactose
0.5#
v
f
1.0* 5
0.5
1.0 x
I
1.0 10
0.5
1.0
\
I
10
10
10
12
14
12
2.85
2.88
4!oo
4.70
3.30
Experiment #15. Effect of lactose
6.76
5.66
6.86
10.43
10.15
10.43
level on
35.80
36.04
34.40
31.16
26.40
31.16
35.91
29.50
25.93
36.53
37.47
36.56
39
27
29
35
24
27
.45
.83
.74
.56
.12
.91
semi-continuous operation of S.
1.25
2.55
0.68
0.33
0.53
0.20
fragilis
fermentation
Lactose
1
1A
IB
2
2A
W-l
W-3
W-5
W-2
W-4
2B
2C
2D
2E
2P
W-l
W-2
w-3
W-4]
5
-6
Sub-1 1
Sub -2 J,
10-11
Sub-1
Sub-2
Sub-3
| Sub-4
> Sub-5
» Sub-6 N
v
8?
c
12
10
11
10
9^F
•• i
12
0.98
1.11
0.92
1.26
1.15
1.37
1.45
1.40
1.44
1.40
5.48
9.87
9.12
46.53
37.48
39.91
40.00
34.14
33.16
56
54
50
.04
.00
.63
1.42
1.53
1.76
2.86
3.38
4.40
3.93
4.30
4.85^
(0.65
-------
Table 1. (cont )
Experiments
(16)
(17)
CHO
5-6*
10-11
#16
AIR
50*
60
80
30
40
0
50
60
80
30
40
0
and #17. Effect
RPM
200
300
350
100
150
50
200
300
350
100
150
50
10) 4.1
10%) 1.19
10*A5 1.31
10) 0.73
10
10
13
11 '
10
1?
1 9
12
0.88
0.16
1.24
1.70
1.23
0.6?
0.89
0.15
of aeration
6.93
7.51
7.70
7.15
7.44
6.61
9.08
9v34
9.04
8.36
8.92
-
conditions
37.80
38.21
38.96
37.48
38.57
39.63
33.04
32.12
33.18
35.88
33.63
at 2
35.91
37.27
41.48
34.16
37.07
31.47
34.83
32.9
39.51
33.72
32.70
carbohydrate
36.25
42.06
41.56
29.72
26.63
13.6
30.2
36.02
35.47
21.56
25.34
levels
2
1
1
3
3
4
3
2
4
4
2
.28
.76
.36
.26
.26
.65
.02
.26
.68
.55
.17
.48
-------
Figure 1,
Growth of S. fragills at different
inoculum levels under carbohydrate
feed conditions maintaining $-< "
OHO in the broth.
1 x 10
1 x 10
8
0
Time (Hours)
18
-------
period Indicated that doubling time decreased as inoculum
size decreased. Thus, Inoculum at 44#, 25#, 15#, 10#, 5# had
doubling times of 100, 86, 55, 50 and 35 minutes, respectively.
Extrapolating from these growth rates indicated that normal
plant inoculum levels of 10-20 percent would give the cell
yields desired (>10^/ml) In a 8 hour period or 3 batches per
24 hour period. Subsequent experiments without maintenance
of 5-6# carbohydrate gave no evidence of inhibition due to
excess whey when the 10$ inoculum level was adopted for routine
use. Whey inhibition was not observed with S_. lactls (Figure
2)» S_» laetis also responded differently in initial growth
rate. A lag phase in growth of S_. lactia was obvious.
The effects of yeast supplementation (Amber BYP) and different
sources and levels of N and P were Investigated (Expt's 6-14,
Table l). Several possible trends were extrapolated from the
data. First "As Is" proteins appeared higher when higher
levels of N were included in the medium. When NHjjOH was used
as source of N, the percent conversion of lactose to yeast
cellular material appeared higher than when (Nlty^SO^ was used
at equivalent N levels. With NH^OH as 17 source higher conver-
sions were apparently obtained at the higher N-levels (Expt1 s
8,9,12.13,14). The higher level,, 0.9# NlfyOH was then used as
SOP for the semicontlnuous and continuous fermentation opera-
tion.
Effftota of variation of aeration and agitation conditions in
19
-------
Figure 2. Growth of S. 1actIs from various
inoculum levels, grown under
conditions of maintenance of 5-
GHO in the broth.
1 x 10
1 x 10
1 x 10
8
0
Time (Hours)
20
-------
the fermentor were Investigated at standard (6$) and high
(11$) carbohydrate levels in the medium with S. fragilis
(Expt's 16 and 17, Table 1). These data indicate that the
higher aeration conditions £0$ Scale Air &2.3>2 V/V/mln), 200
rpm to 80$ Scale Air, 3f>0 rpm Increase conversion efficiency
and perhaps % protein in the washed product. They have little
beneficial effect on total solids or cell numbers. Conditions
below experimental SOP (f>0$ Scale, 2£0 rpm) result in lower
growth and cell yield. The 60$ Scale Air, 300 rpm condition
would appear optimum and the proper departure point for scaleup
to larger equipment.
In order to avoid costly fermentor "downtime" in a short cycle
process, the fermentation was tested under semi-continuous
(Expt's 11, 15) and continuous (Expt 18) operating conditions.
In semi-continuous operation the fermentors were allowed to
run until carbohydrate level reached or dropped below 0.5$.
At this point 90 percent of the beer was pumped into a holding
vessel and an equal amount of fresh medium pumped Into the
fermentor. In short, 10$ of the fermentor beer was retained
as Inoculum for the next batch. Continuous operation of the
fermentation was carried out as described under Materials and
Methods. Figure 3 provides data for serai-continuous operation
of the process under a high (0.9$ NH^OH) and low (0.1^5$ NH^OH)
nitrogen condition. Performance for all conditions is for all
practical purposes identical and maximum cell growth is obtained
In all vessels by 10 hours. In regard to this point, It should
21
-------
ro
ro
Figure 3. Semioontlnuous operation, SOP, 0.9$ NH^OH, S. fragllis
10
0
0 2
(Hours)
-------
be notea that the trend for the Initial vessel set in each
series (2,1,1|,3) was to reach maximum cell growth at 8 hours
while the subcultures were a little slower. This was not due
to initial lag. Slightly lower cell levels were reached at
8 hours and the maximum reached between 8 and 10 hours at a
lower growth rate. The reason for this is not obvious, but
there are sufficient differences in individual subculture
growth patterns to prohibit drawing conclusions until the re-
producibility of this type of operation is explored further.
The fermentation was observed to operate well in a three-
stage continuous system. Data taken on operation of a 60-
hour run are given in Table 2 and Figure 14.. Medium was fed
at a rate of^3 gal/hr into the first fermentor. This rate
is equivalent to 2? gal/9 hr, thus 9 gal/fermentor/9 hr. This
approximates batch fermentation time per unit fermentor, but
eliminates the fermentor downtime which would have to be taken
in batch fermentation operation. The system equilibrated 12
9
hours after a 1 x 10 cell population was reached and the
initial dilution effect was overcome (22 hours). At equilib-
rium time the equivalent of Ij. batch fermentors had already
been harvested. Cell yield out of the third stage fermentor
Q
was equivalent to batch level, oa. 1 x 10 cells/ml, with only
minor deviation from this level. About 1-1.9$ of ethyl alcohol
were accumulated in the third stage of this process. This is
lower than the 2-1$ levels obtained in batch or semi-continuous
operation (Table 1). The effect of process changes on alcohol
-------
TABLE 2.
to
Experiment #18. Continuous operation of a whey fermentation with
S. fragilis in a 3-stage system
Vessel
Hours of
Operation
AI 0
4.5
6.0
10.0
14.0
iQ.o
21.0
22.0
23.0
31.0
33-0
35.0
37.0
39.0
41.0
43.0
45.o
47.o
49.0
53.0
55.o
59.0
TERM59.7
#1
Stage
#2
Cells/mlxlO0 % Cells/mlxlC
Alcohol
1.26
3.6
7.0
9.7
9.1
4.7
5.o
3.5
fc?
3.8
3.8
3.4
3.2
3.0
2.6
2.9
2.8
3.1
3.5
3.4
4.2
4.8
0.33
0.43
0.54
0.34
0.26
0.14
0.11
0.15
0.20
0.73
-
1.09
1.23
1.55
0.94
3.7
6.5
9.2
9.2
8.6
6,8
6.2
5.7
6.8
6.2
6.3
4.7
5.1
4.2
4-2
3.9
4.6
5.7
5.1
6.5
6.4
7.3
)° %
Alcohol
1.0
1.09
1.18
.91
.77
• 30
• 41
• 45
1.05
-
148
1.83
1.84
#3
Cells/mlxlO0 %
Alcohol
0.81
3.5
5.5
10.5
10.9
12.8
12.8
11.5
13.6
11.4
n.5
11.8
12.0
lk.O
8.3
9.2
8.6
10.6
9.6
11.2
11.5
11.3
1.27
1.45
1.36
1.31
1.30
1.10
1.07
1.05
1.90
•
1.79
1.97
1.97
-------
Figure U.. Continuous operation of S. fra
IV)
VJl
gills whey fermentation In a 3-stage system.
-------
accumulation were not investigated. Total solids and protein
analyses performed on broths from batch and semi-continuous
operations were not determined on samples from Experiment 18.
Amber Laboratories will pursue single stage continuous opera-
tion and provide chemical and bacteriological data as required
on product quality.
Amino acid analyses were performed on samples from various
experiments. Data is shown for Experiment 4, in Table 3*
Since less than 30£ of the protein values were recovered as
amino acids some N-containing entity other than amino acid is
present in the broth. The N-containlng entity could be excess
medium nitrogen that was not incorporated into the cell or
amide by-products from the yeast fermentation.
26
-------
TABLE 3. Experiment #k. Amlno Acid Analyses, Inoculum Level Effect
ro
Experimental _,_.., _ _
Conditions *°% Inoculum
10$ Inoculum
25/o I«wuulum As ls Washed Enzyme hydrolyzed
Lysine
Histidlne
Arginine
Aspartic Acid
Threonine
Serine
Glut ami c Acid
Proline
1/2 Cystine
Glyoine
Alanine
Valine
Methionine
Isoleucine
Leucine
Tyrosine
Phenylalanine
1.00
0.25
0.36
1.37
0.77
0.79
3.71
1.01
0.15
O.k9
0.66
0.76
0.67
2.22
0.37
0.51
1.09
0.25
O.k2
1.58
0.87
0.85
2.69
1.13
O.Ik
O.kk
0.80
o.7k
0.28
0.61}.
1.27
0.14.1
o.55
Amino Acid %
1.27
0.31
0.51
1.70
0.96
0.93
3.05
1.2k.
0.16
0.50
0.93
0.80
0.32
0.71
I.k3
O.k8
0.62
3.22
0.77
I.k8
0.96
O.k8
0.51
1.95
2,16
0.11
0.55
0.69
0.60
0.27
0.58
1.70
0.71*
1.00
3.17
0.77
1.76
3.12
1.61
0.09
1.25
2.0k
l.kl
0.36
1.08
2.31
0.92
1.26
16.
. 15
15.92
17.77
29.12
-------
SECTION VI
INTRODUCTION
AMBER LABORATORIES
The primary goal of the fermentation studies was to develop a
method for the production of low cost, single cell protein
(SCP). The SCP was produced by Saccharomyces fragilis grown
on lactose from whey and to be used for animal feed supplemen-
tation. The specific objective of the yeast fermentation
studies conducted in the Demonstration Pilot Plant was to
scale-up a process Standard-Operating-Procedure (SOP) devel-
oped by the subcontractor, International Minerals Corporation
(IMC).
The preliminary study conducted by IMC using shake flasks and
pilot fermentors produced several conclusions and SOP's for
the Amber Laboratories studies. The most encouraging rec-
ommendation indicated that the whey fermentation process could
be run under non-sterile conditions. In addition preliminary
results indicated that operation of the process in a continuous
manner might be feasible.
However, the detailed SOP for the continuous process, Includ-
ing product quality, yield and cost estimations required
study at Amber Laboratories. The following experiments were
conducted to demonstrate a method for simple and economical
conversion of lactose from whey into a high quality feed
supplement.
28
-------
SECTION VII
MATERIALS AND METHODS
AMBER LABORATORIES
Saccharomyces fragilis (NURL, Y-1109) was used in the majority
of experiments except for two preliminary studies involving
Saccharomyces lactis (NURL, Y-1140). Morphologically S_. lactis
was smaller in size than S.. fragilis and it was found that the
fermentation broths of ,S. lactis did not separate efficiently
in commercial sized, yeast separators. Therefore, S_. lactis
was not included in further studies.
In general, concentrated, acid whey was obtained from the
manufacture of cream cheese and condensed to 45-50$ total
solids (T.S.). Acid whey from cream and cottage cheese has pH
values in the range of 4.2-4.9 in comparison to sweet whey
from Cheddar, Italian, and Swiss cheese with pH values of 5«7-
6.2. It was found that the organic nitrogen content of cream
cheese whey was lower than other wheys tested. Wasserman (12)
found that only 25# of the organic nitrogen could be utilized
by S_« fragilis, therefore, we assumed that the fermentation
characteristics of acid and sweet whey would be similar as
long as a slight excess of inorganic nitrogen was added to the
media.
The whey was diluted to appropriate levels of lactose with tap
water and then used to prepare the various media described
below:
29
-------
Whey agar was prepared for sterile transfers of S.. lactls and
S.' fragllls. Acid whey was diluted to 10# T.S. and the follow-
ing ingredients added so the media contained (w/v) 0.5# (Nltyte
304, 0.5$ f^HPOip and 0.3# Amber BYF 100 (see Glossary). Agar
was added to the whey media at a 4# level and the mixture of
whey and agar dispensed in suitable aliquots to 20 x 150 mm
test tubes. The whey agar tubes were pasteurized at 85-90C
for 20 minutes. Saccharomyces lactis and S.. fFagilia were
aseptically transferred on whey agar slants at weekly inter-
vals.
Two basal media were used throughout these studies. SOP
Medium-1 was used for inoculum build-up in shake flasks. SOP
Medlum-2 was used for Inoculum build-up In the New Brunswick
fenaentors and for studies utilizing a 500 gallon, pilot fer-
ment or.
One part whey concentrate was diluted with three parts tap
water and the following Ingredients added so the medium con-
tained (w/v) 0.5# (NH4)2 SOjp 0,5# KgHPOj^ and 0.3# Amber BYP
100. The pH was adjusted to 5.5 for Si. lactis and 4.5 for S_.
fragilis with 8556 H^PO^. The SOP Medlum-1 was not sterilized
for shake flask use.
Whey concentrate was adjusted to various lactose concentra-
tions by dilution with tap water. The diluted whey was made
to contain (w/v) 0.9# NH^OH, 0.3# Amber BYP 100,0.Q5#
-------
and adjusted to pH 5.5 for S_. lactls and 4.5 for J. fragilla
with 30# HC1. The medium was heated to 800, held for 45 min-
utes and cooled. Deviations in the SOP Medlum-2 are described
in the Experiment Condition for each Experiment Number, (Table
4).
A 500 lb/hour, ultrafiltration (UP) pilot plant manufactured
by Abcor. Inc. was utilized to fractionate acid or sweet whey
into a protein concentrate and lactose permeate. The lactose
permeates were prepared for the fermentation studies by
following the SOP Medium-2 format. The permeates were sub-
stituted for diluted whey in preparing the media. Analyses
of ultrafiltration permeates from both acid and sweet whey
were as follows:
Ultrafiltration
Gross Composition Permeate
Average Range
% Total Solids 4 3-5
% Lactose (dry basis) 88 86-90
% Protein (dry basis) 4 3-5
% Ash (dry basis) 8 7-9
The whey permeate was produced from acid or sweet whey using
the Abcor ultrafiltration unit (model #UF-8o-S) operated under
the following conditions:
Gross Composition
of Raw Whey Acid Whey Sweet Whey
pH (as is) 4.6-4.9 5.7-6.0
% Lactic Acid (as is) 0.50-0.65 0.12-0.19
% Total Solids 6.0-7.0 5.5-6.5
% Protein (dry basis) 5.0-9.0 12.5-14.0
% Ash (dry basis) 8.0-12.5 6.0-8.0
31
-------
Operating Conditions
Inlet Temp, of Whey 120-124°F
Inlet Pressure 55 PSIG
Outlet Pressure 25 PSIG
Flux Rate 2.0-3,0 gal/hr
Membrane Surface Area 70 ft2
The closed-loop fermentation studies were made utilizing media
prepared according to the SOP Medium-2 format. However coa-
densate water was substituted for tap water to dilute the whey
concentrate. The condensate water, derived from the evapora-
tion process of fermentation broths, contained up to k% alcohol
(w/v) and less than one part per million nitrogen.
A 500 gallon, fully baffled, deep tank fermentor was used
for the semi-continuous and continuous studies. The media
were aerated by compressed air entering through a bottom
sparger and agitated by a standard shaft and impeller system.
(See Appendices for photographs of equipment). Constant
temperature control of the media was obtained by circulation
of water and steam through the jacket of the fermentor.
Three 15-liter New Brunswick fermentors served as the seed
vessels for the 500 gallon fermentor. The New Brunswick
fermentors were manufactured to standard design and contained
air agitation and temperature controls.
Inoculum for the 500 gallon fermentor was prepared according
to the following procedure:
32
-------
\f
Stock Slant
Wash growth off with 5 ml
medium, Inoculate into:
200 ml SOP Medium-1
Shake Flasks-Ambient con-
ditions for 16 hours. In-
oculate 1,000 ml into:
10 liters SOP Medlum-2
New Brunswick fermentor-
10 liters air/rain, 30C.
Inoculate 30 liters into:
60 gallons SOP Medium-2
500 gallon ferraentor - One
volume air /volume medium,
30 C. Inoculate 60 gallons
into:
240 gallons SOP Medlum-2
The semi-continuous fermentations were made by drawing off 9O£
of the fermentation broth and using the remaining 10$ to seed
the next fermentation batch. The number of consecutive draw-
down batches are indicated in each experiment (Table 4). The
continuous fermentations were begun when the cell count of the
fermentation broth reached 1 x lOVml and the lactose concen-
tration, 0.50-0.755^ (w/v). The rate of continuous feed and
removal was maintained at 37 gallons/hour which corresponded
33
-------
to a batch fermentation cycle of about 300 gallons/8 hours
(dilution rate 0.125 hrT ). The number of hoars the fer-
mentation was maintained continuous Is shown In each experi-
ment (Table 4).
The conditions of operation for the 500 gallon fermentor were
0.5-1.0 volume of air/volume of SOP Medi«ni-2j agitator speed
200 rpra; 30°C; pH 5.5 for S.. lactis and pH 4.5 for S. fragilis.
The fermentor was not sterilized however, the SOP Medlum-2
was heated to 80C for 45 min as described earlier.
Solids determinations were made according to the Mojonnier
Method (?). Protein was determined in accordance with the
A.O.A.C. Method (l). A factor of 6.25 x percent - M was used
as recommended for yeast protein. Ash was determined In
accordance with the A.O.A.C. Method (l). Lactose concentra-
tion was determined using the Anthrone procedure described
by Umbreit, et al. (9). Aralno Acid analyses performed by
Wisconsin Alumni Research Foundation using the method des-
cribed by Moore, et al. (8). Official rat Protein Efficiency
Ration assay performed by Wisconsin Alumni Research Foundation
using the A.O.A.C. procedure (l). Cell count analyses made
according to International Mineral Corporation Method (3).
Results reported as yeast cells/ml of fermentation broth.
-------
SECTION VIII
RESULTS AND DISCUSSION
AMBER LABORATORIES
The effects of variation in the type of acid (phosphoric or
hydrochloric) used to adjust the SOP Medium-2 to pH 4.5 on
yeast cell population, gross composition of yeast and conver-
sion of lactose to yeast material are shown in Experiment
Numbers 1-7, (Table 4). The use of phosphoric acid to adjust
the pH of the medium prior to and during fermentation pro-
duced yeasts with ash levels of 35-43$. The inclusion of
hydrochloric acid (HCl) in the medium resulted in a lower ash
content that approximated the 9-13$ ash level typical of
commercial yeasts. The use of low cost HCl did not adversely
affect the fermentation and produced a more economical product
than that obtained with phosphoric acid.
The 0 = 9$ (w/v) ammonium hydroxide level recommended by IMC
investigators and used in Experiments 1-7, 15 and 18, (Table
4) as a supplementary nitrogen source, resulted in S_. fragilis
yeasts with crude protein levels that were abnormally high.
The effects of 0.3-0.7$ (w/v) ammonium hydroxide concentra-
tions on growth of ;S. fragilis are shown in Experiments 8-13,
(Table 4). It was found that the crude protein levels approx-
imated commercial yeasts with ammonium hydroxide levels of
0.3-0.7$ and the conversion of lactose to yeast material re-
mained constant with varied ammonium hydroxide levels. A
comparison of net protein concentrations (($ crude nitrogen-
35
-------
Table U. Summary of Fermentation Experiments
EPA Project S-8007U7
Number
1 (4?
2 (5)
\ ^ f
3 (2)
Experiment
Condition
S. fragilis
Semi- continuous
SOP Medium-2
Phosphoric Acid
Acid Whey
6-300 gal.
Tank Perm.
S. fragilis
"Semi- continuous
SOP Medium-2
Phosphoric Acid
Acid Whey
8-300 gal.
Tank Perm.
S. fragilis
Batch
SOP Medium-2
Acid Whey
1-300 gal.
Tank Perm.
Lactose
Cone.
7-Qfo
7-Qfo
6-7$
Max . Pop .
Cells/ml
X 10V
o.55
o.75
o.3o
T.S.
PWM
7.1
7.9
5.5
fo Dry Basis
Crude
Protein Ash
kl*k
1*8.6
56.8
35.0
lj.2.6
20.9
% Conversion
gm T.S./
gm lactose
0.96
1.01
0.85
UJ
a
Numbers in parenthesis refer to raw data experiments
-------
Table Ij. (cont.)
Number
M6)
5 (13)
6 (15)
7 (16)
Experiment
Condition
S. fragilia
"Semi-continuous
SOP Medium-2
Acid Whey
7-300 gal.
Tank Perm.
S. fragilis
Continuous
SOP Medium-2
Acid Whey.
Perm. Operated
59 k*1. cont.
S. fragilis
Continuous
SOP Medium-2
Acid Whey
Perm. Operated
90 hr. cont.
_S. fragilis
Continuous
SOP Medium-2
Acid Whey
Perm. Operated
220 hr. oont.
Lactose
Cone.
8-9$
5-6$£
5-6*
5-6^
Max. Pop.
Cells/ml
X 10V
1.52
2.50
2.50
2.2l|
T.S.
PWM
8.2
5.1
5.19
% Dry Basis
Crude
Protein Ash
73.5
76.9
69.7
*
5.89
68.5
15.6
17.2
15.2
17.6
% Conversion
gm T.S./
gm lactose
0.96
0.92
0.99
0.93
U)
-------
Table l| (oont.)
Exp.
Number
8 (9)
9 (10)
10 (7)
» * /
Experiment
Condition
S. fragilis
"Semi-continuous
SOP Medium-2
0.7$ NHhOH
Acid Whey
8-300 gal.
Tank Perm.
S. fragilis
Semi-continuous
SOP Medlum-2
Acid Whly
6-300 gal.
Tank Perm.
S. fragilis
"Semi - oo lit" i nuo us
SOP Medium-2
Acid Whiy
8-300 gal.
Tank Perm.
•
Lactose
Cone .
5-6$
8-9$
l|-5$
Max. Pop.
Cells/ml
X 10
1.22
1.57
0.88
T.S.
PWM
5.38
8.3l
U.98
% Dry Basis
Crud e
Protein Ash
66.1
51.7
1|7.9
13.2
13.9
22.8
$ Conversion
gm T.S./
gm Lactose
1.08
0.96
1.15
oo
GO
-------
Table ij. (cont.)
Number
11 (8
12 (21
13 (26)
Experiment
Condition
S. fragilis
"Semi-continuous
SOP Medlum-2
0.5$ NB.OH
Acid Whey
10-300 gal.
Tank Perm.
J3. fragilis
Continuous
SOP Medium-2
0.3$ NB.OH
Acid Whly
Perm. Operated
61$ hr. cont.
S. fragilis
Continuous
SOP Medium-2
0.3$ NHLOH
Acid Whty
Perm. Operated
56 hr. cont.
Lactose
Cone.
lj.-5$
lj.-5$
7-8$
Max . Pop .
Cells /ml
X 10 v
1.21
2.17
2.16
T.S.
FWM
k*&
5.13
U.9U
$ Dry Basis
Crude
Protein Ash
5U.2
Ij4.6
32.8
11^.0
i
11.2
25.9
$ Conversion
gm T.S./
gm lactose
0.91
1.11
0.68
U)
VO
-------
Tgble ^ (cont.)
Humber
U (25)
15 (l^)
16 (22)
Experiment
Condition
S. fragilis
Continuous
SOP Medium-2
Acid Whey -
Permeate
0,3$ NHi.OH
Perm. Operated
50 hr. cont.
S. fragilis
"Semi-continuous
SOP Medium-2
Sweet Whey
7-300 gal.
Tank Perm.
j3. fragilis
Continuous
SOP Medium-2
Sweet Whey
0.3$ NHL OH
Perm. Operated
76 hr. cont.
Lactose
Cone .
3-li$
ll- 5$
6-7$
Max. Pop.
Cells/ml
X 10 v
2.20
1.5l|
2.1*3
T.S.
PWM
3. Ok
5.38
5.38
$ Dry Basis
Crude
Protein Ash
51.6
81.0
52.2
16.6
11.8
11^.5
$ Conversion
gm T.S./
gm lactose
0.83
1.26
0.82
-------
Table LJ. (oont.)
Number
17 (24)
18 (12)
19 (27)
Experiment
Condition
S. fragllls
"Continuous
SOP Medium- 2
Sweet Whey-
Permeate
0.3$ NHL OH
Perm. Operated
32 hr. cont.
S. fragllls
"Semi-continuous
SOP Medlum-2
Acid Whey
Closed Loop
Condensate -
lj..l$ w/v
Ethyl Alcohol
5-300 gal.
Tank Perm.
S. fragllls
"Continuous
SOP Medium-2
Acid Whey
0.3$ NHL OH
Closed Loop
^^ ~st
Condensate-0 . o>
w/v Ethyl Alcohol
Perm. Operated
38 hr. cont.
Lactose
Cone.
3-1$
IH*
6-7 fo
Max. Pop.
Cells/ml
X 1
-------
% ammoniacal nitrogen) x 6.25)) was made between the whole
fermented whey mass from a media with 0.3-0.7$ ammonium hy-
droxide and a media with 0.9# ammonium hydroxide. The net
protein values were 30-35# for both levels of supplemented
nitrogen. It was apparent that the 0.9# ammonium hydroxide
level did not increase the nitrogen up-take into cellular
protein. Therefore, ammonium hydroxide levels of 0*3-0.7#
(w/v) were used for production of S>. fragilis yeast on whey.
Although lactose derived from acid whey was used throughout
the studies as the main carbon source for yeast fermentation
experiments were performed using lactose from sweet whey
(Experiments 15-16, & 17, Table 4). No significant differ-
ences in fermentation characteristics were noted between acid
and sweet whey (compare Experiment 4 (Acid Whey) with Experi-
ment 15 (Sweet Whey) and Experiment 12 (Acid Whey) with Ex-
periment 16 (Sweet Whey)).
A corollary to the fermentation study of acid and sweet whey,
was to study the fermentation characteristics of deproteinated
acid or sweet whey from an UP unit. The results shown in Ex-
periments 14 and 17, Table 4, indicated that yeast would fer-
ment lactose from an ultraflltration permeate and produce
similar cell counts and crude protein concentrations as yeast
propagated on whole acid or sweet whey. Under the conditions
of these experiments a number of difficulties were experienced
in using ultrafiltration as a pre-treatment for acid or sweet
-------
whey to be used for fermentation. The raw material used for
the UP unit was whey (acid or sweet) having a solids content
of 5.5-7.0^. The resultant permeate was very low in solids
(total solids 3-5#) which would be uneconomical to ferment.
If the solids of the raw whey were raised by evaporation or
partial dilution of condensed whey, considerable membrane
fouling and clean-up problems were experienced. Operation of
the UF unit under these conditions reduced flux rates by 50-
75#. In such a situation, mechanical cleaning (with a sponge
ball) of the membranes was necessary which probably decreased
membrane life. Moreover, the inability to adequately sanitize
the UP unit resulted in extremely high plate counts after
completion of each experiment. Therefore, It was concluded
that the use of UP pre-treatment of whey prior to fermentation
would be unattractive.
A study of the feasibility of using a method for minimizing
waste streams from the fermentation operation was conducted
in Experiments 18 and 19, (Table 4). The block diagram des-
cribing the closed-loop system is shown in Figure 5. A closed
loop design was used which consisted of dilution of acid whey
concentrate with condensate water derived from the evaporation
process of fermentation broths followed by fermentation of
the whey. Theoretically, the closed-loop could be repeated
as often as new whey concentrate was added to the cycle.
Since the entire condensed fermentation broths were spray
dried, no waste streams would be obtained from the process.
-------
Condensed Acid
Whey @ \£>% T.S.
Diluted Whey
Medium @ 12$ T.S,
Fermentation
Process
Condensate
Water
Evaporation
Process
Spray Dry
Process
Product
BLOCK DIAGRAM OF CLOSED-LOOP SYSTEM FOR MINIMIZING
FERMENTATION EFFLUENTS
-------
The fermentation results were not significantly different
from acid whey fermentations using tap water for dilution
although ethyl alcohol concentrations as high as 4# (w/v)
were present in the condensate water.
Although the emphasis of the studies was to develop a fermen-
tation method for production of high quality feed supplements,
several yeasts were washed to produce food yeasts. A bland
food yeast was obtained with crude protein concentrations of
52-55# and ash concentrations of 6-10$. If the higher grades
of yeast products are produced, the additional processing
through centrifugation produced a supernatant effluent that
would have to be processed further by some form of waste
treatment. These supernatant streams from the centrifuges
contained 5 day BOD values that averaged 10,100 mg/1 (average
for 16 experiments, range 5100-12,700 mg/l). The principal
organic materials in these streams were ethyl alcohol (average
less than 2# w/v) and lactose (average 0.5#). These
supernates could not be used to dilute Incoming condensed
whey (as in the closed-loop process) because high concentra-
tions of inorganic salts were present in the supernatant
which interfered with the fermentation. If a centrifuged
product was desired, it was found that yeast slurries could
be centrifuged successfully after partial evaporation (to 15-
\1% T.S.) which reduced supernate volume substantially with -
out raising the BOD value (still 10,000 mg/l). The BOD value
did not increase because some of the ethanol was removed
-------
during partial evaporation.
Throughout the entire study, non-sterile, semi-continuous,
and continuous operation of the fermentation process was dem-
onstrated. The results shown in Figures 6, 7, and 8 are
typical of the high cell populations (l x 10^ cells/ml) ob-
tained with S. fragilis using the SOP Medium-2.
A comparison of the gross composition and amino adld content
of several single-cell proteins is presented in Table 5«
Another comparison of the amino acid composition of several
yeasts, whole wheat, and the FAO (PAO Committee, World Health
Organization, United Nations) profile Is shown in Table 6.
The high lysine and threonlne contents of S.. fragilis and
commercial Brewers' yeast and low values for wheat make the
yeasts valuable supplements for cereal diets. The low con-
tent of sulfur amino acids, especially methionine, in S_.
fragilis and commercial Brewers' yeast is an obvious deficit
for a feed supplement. It has been reported by Klhlberg (4)
that yeast with added methionine produced a high quality feed
supplement.
Feeding studies with growing rats were conducted on centri-
fuged (commercial, nozzle type, yeast centrifuges) fermented
whey mass (FWM) and whole FWM which was fed as the only
nitrogen source in an otherwise adequate diet. The results
are reported in Table 7 and plotted in Figure 9. The Protein
Efficiency Ratio (PER) of centrlfuged, FWM was greater than
-------
3ELL COUNT
ACTOSE,
10
8
0 2
Figure
6 8 10
6. Semi-continuous
0 2 l|
TIME
Growth of
8 10
6 8 10 0 2 l± 6
(Hours)
Saooharomyoes fragllis using SOP Medlum-2.
-------
(DO
0
u
10 20 30
TIME (Hours)
Figure ?, Continuous Growth of Saooharomyoes fragllis Using SOP
Medium-2 (£9 Hr.).
-------
•IS-
vo
0 It
20
60
80
160 180 200 220
100 120
TIME (Hours)
Figure 8. Continuous Growth of .Sacoharomyoes fragllis Using SOP
Medium-2 (220 Hr.).
-------
Table 5. Comparison of Gross Composition and
Amlno Acid Content of Various Single-Cell Proteins
VJI
o
Typical Analysis
Protein
Pat
Ash
Moisture
Amlno Acid
Lysine
Threonine
Methionine
Valine
Leucine
Isoleucine
Tyros ine
Phenylalanine
Tryptophan
Esso
Yeast (10)
54.0
10.0
10.0
a
7.0
3.9
1.2
4.0
5.9
3.6
a
3.7
0.5
Wheast (5)
57.2
1.1
8.5
5.0
?
7.4
5.2
1.5
6.3
7.6
5.2
3.1
3.7
1.4
Brewers
Yeast
% (As Is)
44.9
0.7
6.9
3.0
i of Total 1
6.8
5.9
1.5
4.7
5.8
3.6
2.7
3.4
1.1
Torula
Yeast
47.0
1.2
6.9
5.0
!»T»ot:e1n
8.5
5.1
1.5
5.6
8.0
6.4
4.3
5.1
1.1
S. fragilis
FWM
44.6
1.1
11.2
3.5
6.9
5.8
1.9
5.4
6.1
4.0
2.4
2.8
1.0
b(8.8)
(5.5)
(1.5)
(6.6)
(9.9)
(5.5)
a
(3.9)
(1.5)
a Not Reported
b Dried Pragllis Yeast, Standard Brands Inc., 1967 (6)
-------
Table 6. Amino Acid Content of Whole Wheat,
Commercial Brewers
fragilis Yeast
PAO
Amino Acid (7) Profile
Lysine 14.. 2
Threonine 2.8
Methionine 2.2
Valine l±.2
Leucine [4.. 8
Isoleucine i|.2
Tyro sine 2.8
Phenylalanine 2.8
Tryptophan 1.4
Histidine
Arginine
Aspartic Acid
Serine
Glut ami c Acid
Proline
Glycine
Alanine
Cysteine 2.0
Yeast and Saocharomyoes
compared to PAO Profile
Whole Brewers S.
Wheat (4) Yeast
_ df, n-P Tn-l-nT Pv»n4-nii-i ___ .
2.8 6.8
2.9 5.9
1.5 1-5
If. 4 l|..7
6.7 5-8
3.3 3.6
2.7
if.5 34
1.1 1*1
2.1
74
8.3
\ i i
T 4«-'
12.1
U.2
3.9
5-7
a
t \
7
fragilis
Expt. £5
6.9
5-7
1.9
54
6.1
l±.Q
24
2.8
1.0
2.1
3.7
8.6
• ,
M-.
134
if.i
3.6
5.7
a
a
Not Reported
51
-------
Table 7. Protein Efficiency Ratio (PER) Assays
of Several Typical
Sample
Fermented Whey Mass
Experiment #16
Centrlfuged,
Fermented Whey Mass
Experiment #12
ANRC Casein
Sample
Fermented Whey Mass
Experiment #16
Centrlfuged,
Fermented Whey Mass
Experiment #12
ANRC Casein
Sample
Fermented Whey Mass
Experiment #16
Centrifuged,
Fermented Whey Mass
Experiment #12
ANRC Casein
Saccharomyces fragilis Yeasts.
Average
1 Week 2 week
3 15
18 25
20 24
Average
1 week 2 week
3-6 7-1
5-2 7-7
5-4 7.7
Average
As is
1.97
2.59
2.86
Weight Gain - gm
3 week 4 week
18 17
22 27
25 31
Protein Consumed
3 week 4 week
8.1 8.1
10-5 12 1
9-7 12.2
Total
53
92
100
- gm
Total
26.9
35 5
35-0
PER Value - 4 week
1
Corrected
1.72
2.26
2.50
Corrected to 2.50 as PER value of Casein
(as is value) x (2-50/2.86) = Corrected
-------
100
90
AKRC Casein
Centrlfuged, Fermented
Whey Mass
Fermented Whey Mass
Figure 9.
1 2 Weeks
Rat Growth Rates on Casein and Fermented
Whey Products.
53
-------
of the PER value for the standard casein (ANRC-Animal
Nutrition Research Council). Thus, it appeared that £3. fragllis
yeast was utilized efficiently by the animal without methlonine
supplementation. However, further studies are necessary with
each animal in question to estimate the true effectiveness of
§.- fragilis single cell protein as a feed supplement.
Prom the data derived in the pilot plant operations, it is
possible to calculate preliminary costs of producing yeast for
food or feed purposes by the fermentation of whey. Certain
assumptions must be made as to substrate cost, size of equip-
ment, hours of operation, and capital investment. However,
from these assumptions, it is possible to examine the commer-
cial feasibility of such a process.
Prom the experimental results in the 500 gal fermentor, run
continuously for extended periods of time, the conversion rate
is 0.9 to 1.0 pounds of solids in the whole fermented whey
mass (PWM) per pound of lactose in the original whey. Since
the lactose content of the whey solids is approximately 75#
one would therefore realize 0.68 to 0.75 pounds of PWM (dry
basis) from each pound of whey solids in the original medium.
Assuming fermentation vessels of various size operated con-
tinuously, as in the pilot plant, and using a 300 day produc-
tion year, the production capacities of different size commer-
cial operations may be seen in Table 8.
-------
Table 8, Annual Production of Fermented Whey Mass Solids vs
Fermentor Size
Size of Fermentor Fermented Whey Mass Solids
(gal Ions) (Tons/yr)
5,000 900
10,000 1,800
20,000 3,600
30,000 5,*00
40,000 7,200
The cost of the medium for the fermentation Is based on a 10-
12J* total solids level and will be primarily determined by
the price of whey. While it is not difficult to foresee a
time when this material will have a negative value due to the
necessary cost of waste treatment and pollution abatement, at
the present time it may be possible to obtain the needed whey
for no cost or for fche cost of transportation only. Using a
cost of whey of zero as a low value and 2|/pound solids as a
high value, medium costs will be in the range of 2.5-5.7^/
pound of product.
The cost of production of FWM material is directly affected
by the size of the operation. The cost of labor and opera-
tion decreases on a per pound, finished product basis as the
55
-------
size of the equipment and Its degree of sophistication in-
creases. However, as the capacity increases so does the
capital investment with its connected charges for deprecia-
tion, taxes, insurance and physical facilities.
A plant capable of an annual production of 4,000 to 10,000
tons per year would cost 5 to 15 million dollars depending on
the design. A plant this size is considered a small fermenta-
tion plant, in light of those being designed for the produc-
tion of single cell protein from hydrocarbons (100,000 ton,
annual capacity).
The operating costs of a yeast-whey fermentation, including
those for utilities, power, aeration, recovery and processing
are lower than other processes for producing similar materials
from other substrates. The aeration requirements for the fer-
mentation of hydrocarbons is 2.5 to 5.0 times that of the
fermentation of a carbohydrate such as lactose (10). Because
of the solubility of the whey substrate, agitation and power
needs are lower, as Is the amount of cooling. A cost compar-
ison between acid whey, hydrocarbon, and cellulose fermenta-
tions Is shown in Table 9. The basis of the production
capacities are 100,000 tons per year for the products from
hydrocarbons or cellulose raw material and 4,000-8,000 tons
per year for PWM from whey.
56
-------
Table 9. Production Cost Estimates of Various Single-Cell Proteins
Acid Whey Hydrocarbon (10) Cellulose (2)
Annual Capacity (Tons)
4,000-8,000 100,000
100,000
Type of Cost
Medium
Operating & Utilities
Aeration & Agitation
Cooling
Recovery
Dryi ng
Other
Capital Investment
Labor
Total
Cents per pound
2.
2.
9.
5-5.7
o.5o
0.50
0.30
0.90
0.60
0-4.0
1.90
2- 14.4
2.0-4.0
1.25-2.00
1.00
0.25-0.50
0.40-0.60
0.20-0.40
1.20-2.10
0.60
6.9-11.2
3-0
0.50
0.10
0.20
0.50
0.30
1.60
0.60
6.8
-------
SECTION IX
SUMMARY
AMBER LABORATORIES
The experimental data and experience obtained operating the
demonstration pilot plant over an extended period of time,
leads to the following observations:
Saccharomycea fragllls may be grown on an acid (or sweet) whey
medium In a continuous, deep tank, aerated, fermentor. While
similar fermentations have been described and demonstrated In
the literature for some time, this study has shown that varia-
tions In fermentation conditions, strain selection and medium
composition produced cell counts of several billion cells per
millillter, that may be maintained for extended periods of
time.
The fermentation Itself has many advantages easily recognized
by the experienced Investigator. By operating at a low pH
(4.5) and with a large seed size and a high cell count, con-
tamination Is no problem and therefore, sterile or special
aseptic equipment or techniques are not necessary. The aera-
tion requirements are not excessive with adequate agitation
and efficient baffling, nor Is there any problem In foam con-
trol. Wang (10) reported that hydrocarbon fermentations re-
quired 2.5-5.0 times the amount of oxygen as a carbohydrate
fermentation. Temperature control, despite the rapid growth
*f
rate, was surprisingly easy and a low level of cooling water
was needed. The medium is simple In composition and, at the
58
-------
concentrations tested, the carbohydrate (lactose) Is completely
soluble. The absence of potential toxic substances in the
medium eliminates the necessity of harvesting the cells by
centrifugation. Thus the production of a dried whole fer-
mented mass precludes additional processing of waste streams
from yeast separators and increases the yield of the fermenta-
tion. As a result of the evaporation process of the whole fer-
mented mass prior to drying, condensate water is obtained that
can be used to dilute Incoming condensed whey and thereby
operate a completely closed system with no effluents.
However, if a high protein, food grade yeast is desired, the
cells may be harvested from the medium by centrif ugation.
The yeast cells are sufficiently large for efficient centrif-
ugatlon on standard yeast separators.
These products may be obtained from the fermentation of whey
and are described as follows: A dried whole fermented whey
mass (PWM), a dried cream obtained from the centrifugation
process of the FWM, and a dried cream from centrifuged,
washed, PWM. A dried whole PWM would make a good feed Ingre-
dient. It has a crude protein of 40-5<#, a light color, a
free flowing characteristic and a pleasant "dairy" odor. The
protein shows an excellent amino acid profile, high in lysine,
although somewhat low in the sulfur containing amino acids.
The quality, as indicated by PER determined in rat feeding
59
-------
tests, is good (70% of casein PER) although it is much better
if the fermented whey mass is centrifuged (90% of casein PER)*
Actual feeding tests with animals would have to be run to
determine its true value as a feed ingredient. Saccharomyces
fragilis is an accepted, non-toxic material and may be used
in feed formulations.
When the cells are harvested by centrifugation and the creams
dried or when the creams are further washed and centrifuged,
two superior products are obtained. The dried creams have a
lower ash content than the dried FWM and a PER almost equiv-
alent to casein (PER for centrifuged FWto-2.26; PER for casein
2.50). The uses for such a product may be in specialty feeds
that can pay for the added cost of processing. The washed and
dried cells, give an excellent food grade yeast that compares
favorably with the protein and ash levels (If7% protein and 1%
ash) of Torula and Debittered Brewers Yeasts now being sold
commercially.
The economics of the fermentation is dependent on many factors
and should compare favorably with other procedures for the pro-
duction of single cell protein* By beginning with a soluble,
inexpensive carbohydrate source (lactose in whey) many distinct
advantages should be recognized over those processes that use
hydrocarbons as the carbon substrate. A few of the advantages
are lactose is more soluble than hydrocarbons; lactose
60
-------
fermentations require 1/3 to 1/2 the amount of oxygen necessary
for hydrocarbon fermentation; lactose fermentation broths do
not require separation or solvent removal; lactose fermenta-
tions require less agitation and cooling than hydrocarbon
fermentation* In addition the ease of processing and its
present acceptability as a feed and food ingredient are
valuable considerations.
The cost of production of such an ingredient is primarily de-
pendent on the cost of whey and the capital investment.
Obviously tremendous amounts of excess whey are available in
this country. The figures are well known and do not need to
be repeated. Excess whey must be removed as a contaminant or
a pollution source from our environment and in one way or
another the cheese producer will be required to pay the expense
of pollution abatement. For these reasons the cost of whey
to a whey processor will be low or non-existent. At the pre-
sent time, there are a number of cheese producers who pay
others (either whey processors or municipalities) to dispose
of their excess whey. Therefore, the cost of whey should be
realistically set at zero and possibly a negative value in
the future.
Another factor that will determine the cost of the final pro-
duct is the capital investment required. To be truly efficient
and competitive, a sophisticated plant is required with a
large fermentation capacity. A plant that produces 3,600 tons
61
-------
of material annually would require an investment of over J>
million dollars* The production of the product in the nec-
essary amounts to be competitive, would require a plant with
an annual capacity of 7*000 to 10,000 tons and an investment
of 10-15 million dollars.
The economics of the market place will ultimately determine
the industrial feasibility of this process. Several factors
favor the use of such an ingredient in animal feeds and may
be listed as follows: The shortage of high quality protein,
the high cost and continued shortage of feed grade non-fat
dried milk, and the constant reformulation and special
vitamin requirements of the new high energy feeds. The need
for such a valuable feed ingredient will continue in the
foreseeable future and should increase* The use of the des-
cribed process should produce material to fill this need as
well as provide a useful outlet for a potential pollutant.
62
-------
SECTION X
ACKNOWLEDGMENTS
The support of the sub-contractor, International Minerals
Corporation, and In particular Dr. Ralph Anderson, Dr. Martin
Rogoff, and Mr. Doug Slsson is acknowledged with sincere
thanks.
The following personnel are acknowledged for their valuable
assistance In performing the laboratory and pilot plant
studies at Amber Laboratories, Mrs. Leslie Oberts, Mr. Oliver
Justman, and Mr. Percy Love, who supported the work directed
by Dr. Tom Everson and Dr. Sheldon Bernstein.
The support of the project by the Environmental Protection
Agency, and in particular Mr. Kenneth Dostal, the Grant
Project Officer Is acknowledged with,sincere thanks.
63
-------
SECTION XI
1. Association of Official Analytical Chemists. 1970.
Official Methods of Analysis, llth Edition. Published
by the Association, Washington, D.C.
2. Calllhan, C.D. and Dunlop C.E. 1969. The economics of
Microblal proteins produced from cellulostic wastes.
Louisiana State University.
3. International Mineral Corporation. 1972. Cell count
analyses using hemocytometer . IMC Plaza, Llbertyvllle,
Illinois.
4. Kihlberg. 1972. The microbe as a source of food, p 437
in Annual Review of Microbiology, C.E. Clifton, Raff el,
S., and Starr, M.P., Ed., Vol. 26, Annual Reviews Inc.,
Palo Alto, California.
5. Knudsen Milk Products Co. 1970. Analysis of yeast
products. Knudsen Milk Products, 715 No. Divisadero,
Visa 1 la, California.
6. Mateles R.I. and Tannenbaum S.R. 1968. Single-Cell
Protein. The M.I.T. Press. Massachusetts Institute of
Technology, Cambridge, Massachusetts, and London, England.
7. Mojonnler Bros. C. 1925. Instruction Manual for setting
up and operating the Mojonnler milk tester. Bull. 101.
Mojonnier Bros. Co., Chicago, Illinois.
8. Moore, Stein and Spackman. 1958. Analytical Chemistry
30:1190-1206.
9. Umbereit, Burris and Stauffer. 1957. Manoraetric Tech-
niques. Burgess Publishing Company, 426 S. Sixth St.,
Minneapolis 15, Minnesota.
10. Wang, I.C. 1968. Proteins from petroleum. Che*.
Engineering. 75:99-108.
11. Wasserman, A.E. 1960. Whey Utilization. II Oxygen re-
quirements of Saecharomyces frag 11 is growing in whey
med ium . Appl . Microblol 5:291-293 •
12. Wasserman, A.E. I960. Whey utilization. IV. Availability
of Whey Nitrogen for The Growth of Saecharomyces fragilis.
J. Dairy Sci. 43:123-34.
64
-------
SECTION XII
GLOSSARY
y " The Product remaining after removal of caaein and
fat from cream in the process of making cream cheese, or from
skim milk in the process of making cottage cheese.
Amber BYF 100 - Autolyzed yeast fraction: Spray dried yeast
supernatant.
Amber BYF ^00 - Autolyzed yeast fraction: Spray dried yeast
cream.
A.N.R.C. Casein - A standardized casein product adopted for
use by Animal Nutrition Research Council for Official Associa-
tion of Analytical Chemists rat P.E.R. assays.
B.O.D. - Five Dav - Biological Oxygen Demand: Procedure per-
formed in accordance with methods for the Examination of Water
and Wastewater, 13th Ed., 1971, pp. 489-495.
PD-82 Hodag Antifoam - Fermentation defoamer from Hodag
Chemical Corporation, Chicago, Illinois.
Medium 5102A - A whey medium used for production of Sac-
charomyces fragilis or S.. lactis yeast. Medium consisted of
whey with a lactose concentration of 6%; 0.5% (NHj,)2 SOh (w/v)
(w/v) and 0.1$ Amber BYF 300.
Medium 5102B - A whey medium used for production of Sac-
charomyces fragilis or S_. lactis yeast . Medium consisted of
whey with a lactose concentration of 6%; 0.9% NItyOH (w/v); 0
Amber BYF 300; 0.001# FD-82 Hodag antifoam.
N.U.R.L., Y-1140 - Northern Utilization Research Laboratory,
Strain Y-1140, Saccharomyces lactis.
N.U.R.L., Y-1109 - Northern Utilization Research Laboratory,
Strain Y-1109, Saccharomyces fragilis .
Protein Efficiency Ratio - Assay Method used to determine
efficiency of utilization of nitrogen by rats. Single-cell
Protein fed to growing rats as only nitrogen source in an
otherwise adequate diet.
Scale Air - Arbitrary method of measuring the amount of com-
pressed air injected into the fermentation vessel.
S.O.P. Medium-1 - A whey medium used for production of
Saccharomyces fragilis of S_. lactis yeast, medium used for
65
-------
yeast studies in shake flasks and consisted of whey with a
solids content; 0.5# (NHii)2 SOh (w/v); 0.5# KoHPCii (w/v); and
0.3# Amber BYP 100 (w/v).
S.O.P. Medium - 2 - A whey medium used for production of
Saccharomyces fragilis or S_. lactis yeast. Medium used for
yeast studies in New Brunswick and 500 gallon fermentors and
consisted of whey with a 10# solids content: 0.9# NHhOH (w/v);
0.3# Amber BYF 100 (w/v), 0.05# HoPOj, (w/v); and sufficient
30^ HC1 to adjust pH to 5.5 for S_. lactis and to 4,5 for S..
fragilis fermentation.
Sweet Whey - The product remaining after removal of casein
and fat from milk in the process of making Italian, Swiss and
Cheddar Cheese.
66
-------
SECTION XIII
APPENDICES
Title: Whey Fermentation Investigators: M. Rogoff
N. Janosko
D. Sisson
Project: Contract Research Location: Libertwille
-5000
Cost
Center: Microbiology ) 0741 Period: Oct-Dec 1971
Objective:
Develop an economic process to convert whey lactose to yeast
protein without deproteinlzing the whey. A high quality pro-
tein feed supplement is the desired product.
Background
Processes based on utilization of whey lactose by yeasts on
deproteinized or whole whey substrates have been previously
described (l). Some result in rather efficient utilization
of whey lactose, e.g. the process described by Wasserman
while others are apparently less so. Among such processes
in the latter category are the "Wheast" process (2), that of
Metwally et al (3) and that of Amundson (4). If Wasserman's
24 g/litre yield of yeast is taken as approaching the
theoretically feasible yield of yeast (55% weight conversion
from lactose) then the other processes might be ranked Wheast,
Metwally and Amundson in decreasing order of efficiency.
The more efficient processes in terms of lactose conversion
are characterized by a short cycle time, 6-8 hours, and a
single doubling of the yeast cells in the fermentor. This
type of process requires a 50# inoculum to the fermentor.
Use of such a high Inoculum level has several drawbacks in
operation of an economical fermentation process as follows;
l) The seed vessel must be at least half the size of the
fermentor.
2) Inoculum to the seed vessel must be sufficient to complete
seed development in a time s fermentor cycle time. Unless
this criterion is met, more than one seed vessel per fer-
mentor or a semicontinuous seed operation is required.
3) The entire seed train must be high volume or long in dura-
tion to obtain the required number of cells.
67
-------
4) If operated as In the "Wheast" process requires a centrl-
fuged seed which constitutes an additional operating cost
and is equivalent to recovering half the volume of the
ferment or twice.
5) The economics of operating with the fermentation /turnaround
time ratio very low may be unfavorable.
The feasibility of operating a process based on a 5< inoculum
level is. minimally, questionable, and not representative of
normal fermentation operations. The latter are usually
operated at a nominal 15# maximum inoculum level.
The present study was undertaken in order to demonstrate
operation of a whey lactose to yeast conversion process which
would be commercially feasible and reproducible for ready
scaleup.
Status at Beginning of Quarter
Project initiated this quarter.
Progress During Quarter
Shake flask studies were completed for determination of the
following:
1. Strain. Six yeast strains were screened for growth on
diluted whey concentrate. Two cultures Saecaromyces
lactia NRRL Y-1140 (S^ lactis H. ) and Si. fragilis NRRL Y-
1109 (£[. fragilis W) were retained for scaleup to pilot
plant fermentors.
2. Medium. Effect of N-level and Yeast Extract additions
were tested. SOP for these to FPP was 0.5 and 0.1#, re-
spectively.
3> Effect of carbohydrate level. Serial feeding of lactose
was tested to determine whether sugar was limiting in the
system. An indication of increased cell numbers was
obtained by pure lactose additions either in the batch or
by serial feeding.
4. Growth rate. Generation times were examined under several
conditions. It was found that at 10# inoculum levels one
log increase could be obtained in 8 hours to yield maximum
population obtainable on or 5# lactose in flasks. These
data indicated feasibility of an 8 hour fermentation time
in larger equipment thus 3 cycles/24 hour period.
68
-------
Results and Discussion
Methods. Shake flask experiments were carried out as follows:
l) Stock cultures were maintained on whey agar slants (5%
lactose) pH 5.0 and transferred at weekly intervals.
2) Inocula were prepared by washing the growth from a 24 hour
whey agar slant into 5 ml of whey medium and inoculating
1.5 ml into 50 ml of whey medium in a 250 ml flask. Flasks
were incubated on a rotary shaker at 270 rpm at 30 C.
Media were not sterilized in any of the procedures used
with the exception of whey agar for stock slants.
3) Shake flasks were charged with 90 ml of whey medium pre-
pared as follows: whey concentrate diluted to 5# lactose
(or as indicated in individual experiments) and 0.5#
(NHu)2SO&, 0.5# K2HPOU and o.l# yeast extract added; pH
as infiictted in individual experiments. Flasks were in-
oculated with 10 ml of 16 hour growth from suitable in-
oculum flasks. All Incubations were at 270 rpm, 30 C.
Optimal pH for T. lactoaa was 3.5- S. lactis 5.5 and S_.
fragilis 4.5. These pHfs were used for the respective
cultures.
4) Analytical, a) Cell count, microscopically, by standard
dilution technique in a blood counting hemocytometer,
b) Cell pack - 10 ml of broth were spun down in volumetric
tapered centrifuge tubes for 30 minutes at 3000 rpm.
Volumes were read directly.
5) Reducing sugar (lactose) values were determined by a
standard Technicon Autoanalyzer colorimetric method.
6) Dry weights were determined by shell freeze drying of appro-
priate aliquots of broth.
Experimental
The shake flask experiments described bel-ow were set up
following initial pure culture isolation of strains from in-
coming cultures and preliminary demonstrations of the ability
of raw diluted whey concentrate to support growth. The first
five experiments were carried out using all six strains of
yeast originally obtained. These included: Torula lactosa
NRRL Y-196; T. lactosa NRRL Y-1203; Saccharomyces lactis NRRL
Y-1140J S. fragilis ]*RRL Y-1156; S. fragilis NRRL Y-1109, S..
fragilis""l208 (from T. Everson). All subsequent experiments
in shake flasks included only T. lactosa Y-196, S. lactis
Y-1140 (S.. laetis H) and S. fragilis Y-1109 (S.fragilis W).
69
-------
Shake flask experiments were designed to Investigate factors
limiting the growth of the yeast. Following preliminary ex-
perlemnts to determine optimal pH for each test culture a
shake flask standard condition was set. Carbohydrate level
was investigated first. Experimental design was to supplement
whey concentrate diluted to a carbohydrate level of 5# with
pure lactose in five percent Increments. Carbohydrate supple-
mentation was made by either a batch procedure (SF 2,3) or
by a serial feeding of 5 percent increments at 8 hour In-
tervals (SF 4). An almost direct response to carbohydrate
level in terms of cell numbers obtained was observed for the
T. lactosa. S., lactla H (Figure l). Response of IS. fragllis
to Increased carbohydrate was not observed consistently
although a response trend might be extrapolated from the data.
Z- lactoaa did not utilize all carbohydrate provided; the
other two cultures evidenced usually less than 1% residual
carbohydrate in the broth.
The effect of nitrogen level and yeast extract supplementation
was investigated in an experiment in which SOP condition was
followed and an additional 5# lactose added to the flasks
after 8 hours to allow full growth potential (SF-5). Although
full utilization of carbohydrate was obtained no differences
in cell pack volume attributable to Increased N-levels were
observed. Since all carbohydrate was utilized it was assumed
other factors which might be limiting on the system e.g.
dissolved oxygen, might be repressing a nitrogen-response.
Pursuance of this line of experimentation was accordingly post-
poned until operation in pilot plant fermentors was initiated
and where 02 availability might be less limiting. Investigation
of P effect was also postponed.
A series of shake flask experiments were next carried out to
determine whether lactose in excess of 5# provided as additional
whey would evoke the same growth response as found on addition
of pure lactose (SF 6-7). Experimental format used was dilution
of concentrated whey to desired carbohydrate level (5-10#)
with no further sugar addition. N and Yeast Extract supple-
ment levels were varied. The results of these experiments in-
dicated that suboptlmum growth was obtained and that Incre-
mental growth response to carbohydrate was not observed. Pro-
viding additional carbohydrate as whey is apparently Increasing
the amount of some non-carbohydrate material in the medium
to an inhibitory level. Under this apparent inhibition no re-
sponse to adjuncts was observed. S_, laetla was perhaps more
sensitive to the inhibitory effect than was S_. fragilis.
Generation times for the shake flask fermentation were cal-
culated from cell counts taken on flask populations under
various conditions. Since it had been observed that carbohy-
drate depletion had taken place usually within 8 hours
70
-------
generation times calculated were all based on an 8 hour period.
Some of these data are summarized in Table 1. Slightly over
3 doublings are required to bring a 1O# Inoculum (of maximum
cell growth) to original cell levels; In short one log in-
crease in cell numbers. Examination of count data indicated
2.8-3.2 doublings of populations of S. lactis and S. fragllis
were obtained In 8 hours. This information indicates that
3 fermentor cycles per day may be feasible.
The data in Table 2 are a general summary of shake flask data
obtained. Details of Individual experiments are recorded in
IMC Laboratory Notebook #5587 assigned to M. Janosko.
Future Plans
1. Initiate Pilot Plant fermentor studies on shake flask SOP.
Characterize population response.
2. Investigate effects of N, Yeast or Yeast Extract, P and
carbohydrate levels in fermentors.
3. Investigate effects of physical parameters, e.g. , 02,
agitation, on the fermentation.
4. Investigate effects of Inouculum levels on yeast yield.
Investigate effect of retention of an aliquot of fermentor
final whole culture as Inoculum on semi-continuous
operation of the process.
5. Initiate characterization of yeast or product yields in
terms of carbohydrate, protein, fat, ash and moisture
under various fermentation conditions. Carry out amino
acid analysis on selected representative product samples.
References
1. Wasserman, A.E., Appl. Microbial. I960, 8. (5), 291.
2. Robe, K., Pood Processing, Chicago, 1964, 2£ (2), 95
3. Metwally, M.E., et al., J. Dairy Sci., 1964, 4£ (6), 680.
4. Amundson, C.H., Amer. Dairy Rev., 1967, 2£ (7), 22.
71
-------
Effect of Lactose Level on Yeast Cell Count (SP 3)
10-
o
tc
•=}•
CVi
a
w
u
10
8
5 10
TOTAL CARBOHYDRATE
15
72
-------
TABLE 1. Yeast Population Increases During 8 Hour
Fermentations Under Various Condtitions
Condition
SOP (see text)
SOP
SOP, Whey
adjusted to
[lactose]
= 12.6*
= 10.4
= 7.5
= 6.0
= 10.4
= 7-5
= 6.0
_ i o f-^oL
— J- d. . fJyo
* T.5
* 6.0
Initial
cells/ml
Yeast xlO8
T.
S.
S.
T.
I.
S.
T.
T.
T'.
|
S.
S.
s.
s.
lactosa
lactis
fragllis
lactosa
lactis
fragills
lac toga
lactosa
lactosa
lactosa
lactis
lactis
lactis
lactis
fragilis
fragilis
fragills
0.27
0.96
0.33
0.37
1,1
0.27
0.54
0.54
O ^SM
O ^M
1.2
1.2
1.2
1 2
0.32
0.32
0.32
AVERAGE
T.
S.
S.
Final
cells/ml
xlO65
No.
Doublings
2.04 3.0
10.75 3.4
4.75 3.8
2.3 2.6
17.5 4.0
6.7 4.5
0.85 0.6
1.10 1,0
2.0 1.9
2.15 2.0
7.05 2.5
6.45 2.3
8.35 • 2.8
7.55 2.6
3.1 3-2
4 . 05 3-6
3-25 3.3
4.85 3-9
GENERATION TIME:
lactosa =3-2 hrs
lactis s 2.4 hrs
fragllis=.2.0 hrs
73
-------
TABLE 2. Summary of Shake Flask Experiments
# Experimental Conditions
1
2
3
4
5
6
SOP
SOP
SOP
SOP
SOP
SOP
SOP
SOP
SOP
SOP
SOP
SOP
SOP
SOP
SOP
SOP
+ 5% lactose )Batch
* ?Llacto.se Isatch
•»• 1055 lactose)
)
* 5% lactose) 5$
serial feed
+ 10?6 lactose)
- As Is H4 Whey )
c" 2x (NHjj^SO^ }
As Is
c 2x Yeast Extract)
9 Con c . whey
adjusted to 1096
CHO
+ 0.4# N .05^6 Y
* 0.5# N 0.196 Y
+ 0.6# N 0.1556 Y
T. lactosa Y-196
Cells
xlO8
2.9
2.0
1.7
2.04
8.5
11.5
2.3
10.95
18.6
Resid
CHO %
3.76
.54
4.20
3.7
4.8
4.3
0.60
0.49
0.54
p*ck#*
ml T**
0.35
.30
.62
.77
.75
.775
.80
S.
Cells
xlO8
10.5
10.2
7.6
10.8
24.0
22.5
17.5
24.7
20.1
5.6
lactis
Resid
CHO %
.37
.40
.44
1.02
.47
.34
0.56
0.40
0.42
8.1
8.9
9.6
#
" Pack*»
ml T
0.55
.36
.64
.64
1.025
1.15
1.1
0.7
.675
0.75
S.
cell
xlO8
4.9
5.2
3.6
4.8
5.9
5.3
4.7
8.5
6.8
2.7
fragllls
s Resid
CHO
.29
.44
3.23
.66
.36
1.58
0.39
0.36
0.51
8.3
9.6
9.25
W*
Pack*«
ml T
0.5
.50
.70
.60
.775
.85
.70
0.6
0.5
0.5
-------
7 Dilution
of
various CHO
CHO 10
10
7
7
6
6
5
5
.0
.5
.5
.0
.0
.0
.0
whey to
levels
M 0.5
0.5
0.5
0.5
M
Y 0
0
0
0
.1
.1
_
.1
_
.1
-
7
6
8
7
.05
.45
.75
.55
6.1
5.8
3.1
2.1
.46
.k8
_
.05
.65
.65
.65
.60
.625
.60
.60
.60
3.1
3.4
4.1
2.8
3.3
3.6
4.9
5.8
6.1
5.9
2.4
2.2
-
•
-
•
.6
.68
.6
.6
.5
.5
.55
.55
* Minimally duplicate standard triplicate flasks were set for most experimental conditions,
obtained with three other yeast cultures in the first five experiments are not
included above (see text).
vn
** ml wet cells/10 ml broth
-------
Live culture of yeast asceptically transferred from
yeast agar slant to shake flask.
76
-------
Growth of yeast on whey media in shake flasks.
Growth of yeast on whey media in New Brunswick fermentors,
77
-------
Top view of 600 gallon fertnentor showing mechanical
foam breaker and yeast/whey fermentation broth.
Broth from continuous yeast fermentation of whey.
78
-------
Growth of yeast on whey media in 500 gallon deep tank
fermentor.
Progress of yeast fermentation followed by actual cell
counts.
79
-------
Pilot spray dryer for production of experimental lots
of dried yeast.
Spray dried yeast from continuous growth of yeast on whey.
80
-------
SELECTED WATER
RESOURCES ABSTRACTS
' TRANSACTION FORM
2.
3. Accession No.
w
at
PROTEIN PRODUCTION PROM ACID WHEY VIA FERMENTATION
7. Author(s) J
Sheldon Beenstein, Ph.D.; and Thomas p. Everson, Ph.D
Milbrew Incorporated
Amber Laboratories Division
Juneau, WI 53039
10. Project No.
S-800747
11. Contract/Grant No.
.tAl, Pro tec tiojo _ Agency;
15. Supplementary3 Notes *
Environmental Protection Agency report number, EPA-660/2-7U-025, May
16 Abstract '
Prom the operation of a demonstration pilot plant over extended periods
of time, it has been shown that yeast may be grown on an acid whey or
sweet whey medium in a continuous, deep tank aerated fermentor. Varia-
tions in fermentation conditions, strain selection, and medium cpmposi-
tion produced cell concentrations of several billion cells per milli-
liter. Up a process of evaporation and spray drying the whole fermented
whey mass and the Utilization of the evaporator condensate to dilute
Incoming condensed whey, a high grade, non-toxic, protein feed material
be produced without any effluent streams. Amlno acid analyses and
efficiency wetioff -are pre&j^ed -for thfc* feedr materialv
Economic estimates show that while": a large capital investment and low
cost raw material are required for the commercial,feasibility of this
fermentation process, it will be competitive with other methods for the
manufacture of singl« cell proteini This whey fermentation Is one
means of convertingllarge quantities of a potential environmental
pollutant Into a useful and needed! product.
l?a. Descriptors ^Illdua(:rl£|1 WaataSj *Fennentati
------- |