Available online at www.sciencedirect.

com

ScienceDirect
Energy Procedia 61 (2014) 2729 – 2732

The 6th International Conference on Applied Energy – ICAE2014

Ethanol fermentation under dissolved carbon dioxide control

Sijing Feng and Yen-Han Lin*
Department of Chemical and Biological Engineering, University of Saskachewan, Saskatoon, SK, Canada

Abstract

Dissolved carbon dioxide (dCO2) evolved from Saccharomyces cerevisiae was measured, controlled and
used to drive repeated batch operation for ethanol fermentation. The experiments were conducted at four
glucose feeds (150, 200, 250 and 300 g/l) under three dCO2 control levels (no control, 750 ppm and 1000
ppm). Results showed that the evolution patterns of dCO2 depend on the extent of glucose being utilized
as well as the dCO2 control level. Repeated batch was successfully operated for four glucose feeds under
dCO2 control. Controlling dCO2 level at 750 ppm, the self-cycling period for each repeated batch is
6.6±0.3, 7.3±0.2, 12.12±1.12, and 21.70±5.98 h, as glucose feed changes from 150, 200, 250 and 300 g/l,
respectively. When dCO2 level increases, the self-cycling period increases as well.
© 2014 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
© 2014 The Authors. Published by Elsevier Ltd.
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and/or peer-review under responsibility of ICAE
Peer-review under responsibility of the Organizing Committee of ICAE2014

Keywords: dissolved carbon dioxide, repetaed batch, fermentation,

Nomenclature
Bp annual biomass productivity, tons/year
Cg initial glucose concentration, g/l
Cb final biomass concentration, g/l
Ce final ethanol concentration, g/l
Ep annual ethanol productivity, tons/year
Gc annual glucose consumption, tons/year
Td down time, h
Tf fermentation time, h
Tw annual operating time, h
Vw working volume, m3

* Corresponding author. Tel.: +1-306-966-4764; fax: +1-306-966-4777.

E-mail address: yenhan.lin@uasask.ca.

1876-6102 © 2014 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Peer-review under responsibility of the Organizing Committee of ICAE2014
doi:10.1016/j.egypro.2014.12.290
2730 Sijing Feng and Yen-Han Lin / Energy Procedia 61 (2014) 2729 – 2732

1. Introduction
Repeated batch is a cyclic process operating on a feedback control strategy. In theory, cycles in a
repeated batch operation are based on the metabolic rates of the fermenting cell population thus taking
into account the biochemical aspect of any bioprocess. The details of the repeated process fermentation
have been reported by several authors [1, 2, 5, 6, 10,]. If there is no external parameter controlling the
cyclic process, this process also gives the name self-cycling fermentation (SCF) [1, 2, 6,]. For a
fermentation process to be self-cycling, a parameter that is representative of the state of the cells in the
fermenter need to be measured. Strategies for fresh media feeding and broth removal are then developed
accordingly. Prior research relevant to SCF processes has been discussed for various bacteria and yeast
strains other than saccharomyces cerevisiae [1, 2, 6,]. Feng et al (2012) designed a repeated batch ethanol
fermentation where the fermentation redox potential was used to drive self-cycling process [3]. As
fermentation proceeded, they reported that the self-cycling period reduced, resulting in high ethanol
productivity. However, as sugar concentration in the feed stream increases, the self-cycling time has to be
manually determined instead of settling by yeast itself. To overcome this problem a new fermentation
controlling parameter was sought in order to drive self-cycling process.
As a major by-product of ethanol fermentation, dCO2 measurements during the fermentation
provided integrated information with respect to glucose utilization, ethanol production and yeast activity.
With earlier literatures showing the application of CO2 evolution rate in the off-gas stream to determine
microbial activity, the ability of dCO2 to do the same has never been shown except in few cases which
were more focused on the novel measuring device rather than the characteristics of the observed dCO2
profiles [4, 7, 8,]. The empirical relations established so far between CO2 evolution, glucose consumption
and ethanol production no longer seem to be applicable for VHG environments. In order to make up for
the gap of phenomenon of dissolved carbon dioxide in ethanol fermentation, a series of dissolved carbon
dioxide monitoring and control experiments were designed and performed by our group. We reported
dCO2 concentration profiles during VHG fermentation [9] and has successfully incorporated a dCO2
based control methodology to improve otherwise sluggish batch fermentation for glucose concentrations
higher than 200 g/L. We hypothesized that the repeated batch operation of a fermenting system can run
through SCF under both of with and without dissolved CO2 control condition and result in higher ethanol
productivity. In order to test our hypothesis, fermentation was carried out utilizing four feed glucose
concentrations with three dCO2 control levels.

2. Reproducibility of dCO2-driven repeated batch fermentation

The experiments were initiated in batch mode followed by SCF operation. Once the operation was
activated, half of the working volume of spent media in the fermenter was withdrawn and refilled with an
equal volume of fresh media. The SCF operation was then repeated until the media in the feed constrainer
was exhausted. In the dCO2 controlled cases, once dCO2 became lower than the set point value, an
appropriate amount of filter-sterilized air, determined by the PID control algorithm, was sparged into the
fermenter to maintain dCO2 at the desired level.
As similar as former redox-potential driven repeated batch system [3], the dCO2 self-regulated
repeated batch system settled after 3-4 cycles and the resulted were presented in Table 1. Results
suggested that final biomass concentration under dCO2 control was much higher than that of no control
case. The high final biomass concentration resulted in a low production efficiency, and short fermentation
time. It is also important to note that there were no residual glucose left at each cycle for all sets of
experiments. These conclusions were agreed with previous publications related to yeast metabolism [3,
9].
 Sijing Feng and Yen-Han Lin / Energy Procedia 61 (2014) 2729 – 2732 2731

Table 1 Summary of dCO2-driven repeated batch fermentation under four feeding concentrations with
three dCO2 control levels.

~150 g glucose/l ~200 g glucose/l

dCO2 control level no Control 1000 ppm 750 ppm no Control 1000 ppm 750 ppm
Final ethanol, g/l 74.6±2.2 59.6±2.7 54.2±4.0 102.5±2.5 70.7±3.4 66.7±1.6
Final biomass, g/l 3.2±0.2 10.7±0.2 10.1±0.3 4.2±0.3 10.9±0.2 10.1±0.3
Stable cycle time, h 11.3±1.1 6.6±0.3 6.2±0.2 14. 7±2.7 8.7±0.5 7.3±0.2
Efficiency 0.91±0.05 0.67 ±0.07 0.62±0.08 0.91±0.07 0.68±0.06 0.69±0.03
Viability 0.94±0.03 0.98±0.01 0.99±0.01 0.90±0.08 0.99±0.0 0.97±0.01
~250 g glucose/l ~300 g glucose/l
Final ethanol, g/l NA 85.12±4.34 83.24±2.95 NA 113.54±7.90 94.29±10.32
Final biomass, g/l NA 9.94±0.73 9.92±0.20 NA 7.42±0.99 8.76±1.48
Stable cycle time, h NA 14.94±1.93 12.12±1.12 NA 31.47±6.97 21.70±5.98
Efficiency NA 0.65±0.01 0.69±0.3 NA 0.68±0.02 0.62±0.08
Viability NA 0.97±0.02 0.99±0.01 NA 0.73±0.08 0.92±0.06

3. Applicability of dissolved carbon dioxide-driven and controlled repeated batch fermentation

In order to determine the effects of dCO2 control levels on dCO2 driven repeated batch ethanol
fermentation, a comparison of glucose utilization, biomass and ethanol production among no control,
1000 ppm control level and 750 ppm control level was performed. The formulae used to calculate glucose
consumption, biomass and ethanol productivities are shown in Eqn (1), and the calculated results were
compiled in Table 2.

(1)
Table 2 illustrated that maximal annual ethanol and biomass productivities were constantly observed
at 750 ppm dCO2 control level as the feed concentration was greater than 200 g glucose/l. The increased
productivity was partly related to shorter self-cycling time among three control levels. While high
osmotic pressure prolongs cell’s lag phase initially, ethanol inhibition results in low cell viabilities as well
as high residue glucose in the end of fermentation. Hence, high osmotic pressure and ethanol toxicity are
key factors that retard the application of VHG technology to the bio-ethanol industry. The reported dCO2
driven repeated batch process provides a way to alleviate above-noted. According to Table 2, the
developed new technology is expected to have at least 20% more ethanol annual productivity than
traditional fermentation.

Table 2 Comparison of annual glucose consumption, biomass and ethanol productivity for dissolved
CO2-driven repeated batch fermentation under ~150 and ~200 g glucose/l feed concentration conditions
with no control, 1000 ppm and 750 ppm dissolved CO2 control level. Vw = 100 m3, Tw = 7920 h.
2732 Sijing Feng and Yen-Han Lin / Energy Procedia 61 (2014) 2729 – 2732

Glucose (tons) Ethanol (tons) Biomass (tons) Glucose (tons) Ethanol (tons) Biomass (tons)
~150 g glucose/l ~200 g glucose/l
No control 5763 2614 112 5327 2761 113
1000 ppm 10361 3576 642 9627 3218 496
750 ppm 10451 3462 645 10455 3618 548
~250 g glucose/l ~300 g glucose/l
1000 ppm 6903 2256 324 4387 1429 93
750 ppm 7859 2720 263 5831 1721 160

4. Conclusion

Our experimental results proved that dCO2 driven repeated batch (self-cycling) system could
successfully applied into ethanol fermentation with three dCO2 control levels. There is no residual
glucose left at the end of each cycle for all tested conditions. Moreover, the maximal ethanol productivity
was observed at 750 ppm control level while a low production efficiency was detected in the same level.

5. Reference

[1] Brown W and Cooper DG. Self-Cycling fermentation applied to acinetobacter calcoaceticus RAG -1.
Appl Environ Microbiol 1991; 57 (10): 2901-2906
[2] Brown W and Cooper DG. Adapting the self-cycling fermentor to anoxic conditions. Environ Sci
Technol 1999; 33:1458–63
[3] Feng S., Srinivasan S. and Lin Y.-H. Redox potential-driven repeated batch ethanol fermentation
under very-high-gravity conditions, Process Biochem. 2012; 47(3): 523-527
[4] Kocmur, S., Corton, E., Haim, L., Locascio, G., and Galagosky, L. CO2-Potentiometric determination
and electrode construction, a hands-on approach. J. Chem. Ed 1999; 76 (9), 1253- 1255.
[5] Sauvageau D, Stromes Z and Cooper D. G. Synchronized populations of Escherichia coli using
simplified self-cycling fermentation. J Biotechnol 2010; 149 (1-2): 67-73
[6] Sheppard J, Cooper DG. Development of computerized feedback control for the continuous phasing
of Bacillus subtilis. Biotechnol Bioeng 1990; 36: 539–45
[7] Shoda, M., and Ishikawa, Y. Carbon dioxide sensor for fermentation systems. Biotechnol. And
Bioengng 1981; 23 (2), 461-466
[8] Sipior, J., Randers-Eichhorn, L., Lakowicz, J. R., Carter, G. M., and Rao, G. Phase flurometric optical
carbon dioxide gas sensor for fermentation off-gas monitoring. Biotechnol. Prog 1996; 12, 266-271.
[9] Srinivasan S., Feng S. and. Lin Y.-H. Dissolved carbon dioxide concentration profiles during very-
high-gravity ethanol fermentation. Biochem. Eng. J 2012; 69: 41-47.
[10] Yamakawa S, Yamada R et al., Repeated batch fermentation from raw starch using a maltose
transporter and amylase expressing diploid yeast strain. Appl Microbiol Bitechnol 2010; 87(1): 109-115.

Biography
Sijing Feng: MSc graduate student, Department of Chemical and Biological Engineering, University of
Saskatchewan, Canada
Yen-Han Lin: Professor, Department of Chemical and Biological Engineering, University of
Saskatchewan, Canada.