technical, economic, and environmental feasibility of
TRANSCRIPT
The Pennsylvania State University
The Graduate School
College of Engineering
TECHNICAL, ECONOMIC, AND ENVIRONMENTAL FEASIBILITY OF
WASTEWATER-DERIVED DUCKWEED BIOREFINERIES
A Dissertation in
Environmental Engineering
by
Ayse Ozgul Calicioglu
2019 Ayse Ozgul Calicioglu
Submitted in Partial Fulfillment
of the Requirements
for the Degree of
Doctor of Philosophy
May 2019
ii
The dissertation of Ayse Ozgul Calicioglu was reviewed and approved* by the following:
Rachel A. Brennan
Associate Professor of Environmental Engineering
Dissertation Co-Advisor
Co-Chair of Committee
Tom L. Richard
Professor of Agricultural and Biological Engineering
Director, Institutes of Energy and the Environment
Dissertation Co-Advisor
Co-Chair of Committee
Charles T. Anderson
Associate Professor of Biology
John M. Regan
Professor of Environmental Engineering
Deborah L. Sills
Associate Professor of Environmental Engineering
Special Member
Patrick J. Fox
Professor of Civil And Environmental Engineering
Department Head, Civil And Environmental Engineering
*Signatures are on file in the Graduate School
iii
ABSTRACT
Duckweeds (Lemnaceae) are efficient aquatic plants for wastewater treatment due to their
high nutrient uptake capabilities and resilience to severe environmental conditions. Combined
with their rapid growth rates, high starch, and low lignin contents, duckweed could be used as a
viable feedstock for bioprocessing into fuels and chemicals in a biorefinery system.
In this study, several of the knowledge gaps preventing the establishment of integrated
wastewater-derived duckweed biorefineries were addressed. Technical, economic, and
environmental evaluations were performed to enable the simultaneous utilization of duckweed as
a reliable nutrient recovery tool and as a feedstock for bioenergy generation.
A naerobic bioprocesses (bioethanol fermentation, acidogenic digestion for volatile fatty
acid (VFA) production, and methanogenic digestion for biomethane production) were
sequentially integrated to maximize the carbon-to-carbon conversion of wastewater-derived
duckweed biomass into bioproducts. Duckweed was fed to reactors raw (dried) after liquid hot
water pretreatment or enzymatic saccharification. At the end of each bioprocess, the target
bioproduct (i.e., bioethanol, VFAs, or methane) was separated from the reactor liquor (i.e., by
vacuum extraction of ethanol, or membrane separation of VFAs) and the remaining reactor
components were subjected to further anaerobic bioprocesses. The highest total bioproduct
carbon yield of 0.69±0.07 grams per gram of duckweed carbon was obtained by sequential
acidogenic and methanogenic digestion. Nearly as high yields were achieved when three
bioprocesses were integrated sequentially (0.66±0.08 grams of bioproduct carbon per duckweed
carbon). For this three-stage value cascade, yields of each process in conventional single-stage
units were: 1) 0.186±0.001 grams ethanol per gram duckweed; 2) 611±64 mg acetic acid
equivalent of volatile fatty acids per gram of volatile solids; and 3) 434±0.2 ml methane per gram
of volatile solids.
iv
Following experimental studies, the techno-economic analysis of a hypothetical large-
scale duckweed production/wastewater treatment and biorefinery system was performed. Annual
duckweed yield was simulated as 51 dry Mg per hectare after losses, over an area of 141 ha fed
with municipal wastewater primary effluent, when 80% of the mat is harvested weekly.
Discounted cash flow analysis results revealed that minimum biomass selling price of $25 per dry
Mg with a 10% internal rate of return could be achieved if the system boundaries consider
wastewater treatment as a credit. Modification and downscaling of the National Renewable
Energy Laboratory 2011 Report on lignocellulosic biorefineries revealed a minimum ethanol
selling price of $8.2 per U.S. gallon, with a 2.45% internal rate of return. For the calculation of a
more realistic minimum ethanol selling price, a rigorous mass and energy balance must be
performed.
Life cycle assessment of the base case scenario used in techno-economic analysis showed
that the recovery of nutrients from wastewater into duckweed biomass produced a net benefit on
reducing eutrophication potential. The environmental impacts of duckweed biorefinery products
to substituted products (i.e. gasoline, natural gas, and chemical fertilizers) were found to
generally depend on biorefinery size: the larger the biorefinery, the smaller the environmental
impacts. In terms of global warming potential (GWP), distillation for ethanol production appears
to cause the highest environmental burden; however, a credit for marketing of process residues as
a synthetic fertilizer substitute results in a net 13% reduction in GWP, more than compensating
for the distillation burden.
v
TABLE OF CONTENTS
LIST OF FIGURES ......................................................................................................... viii
LIST OF TABLES ........................................................................................................... x
ACKNOWLEDGEMENTS ............................................................................................. xii
Chapter 1 Introduction .................................................................................................... 1 General Background ........................................................................................................ 6
Lemnaceae (duckweed) ............................................................................................ 6 General characteristics...................................................................................... 6 Utilization of duckweed in wastewater treatment processes ............................ 8 Duckweed as a bioenergy feedstock ................................................................. 10 The potential for integrating duckweed production with wastewater
treatment in ecological systems ................................................................ 16 Anaerobic bioprocesses ............................................................................................ 18
Methanogenic anaerobic digestion ................................................................... 18 Acidogenic anaerobic digestion (AAD) ........................................................... 23 Ethanol fermentation ........................................................................................ 26
Life cycle assessment ............................................................................................... 31 State-of-the-art for Duckweed Bioconversion through Anaerobic Bioprocesses ............ 32
Chapter 2 Proof of Concept - Sequential Ethanol Fermentation and Anaerobic
Digestion Increases Bioenergy Yields from Duckweed .................................................. 36 Abstract ............................................................................................................................ 36 Introduction ...................................................................................................................... 37 Materials and Methods ..................................................................................................... 39
Analytical methods ................................................................................................... 39 Plant material and cultivation ................................................................................... 40 Inocula ...................................................................................................................... 41
Yeast strain ....................................................................................................... 41 Anaerobic Seed ................................................................................................. 41
Fermentation experiments ........................................................................................ 42 Biochemical methane potential (BMP) assays ......................................................... 42 Overall bioenergy yields .......................................................................................... 43
Results and Discussion ..................................................................................................... 44 Fermentation experiments ........................................................................................ 44 Biochemical methane potential (BMP) assays ......................................................... 45 Overall bioenergy yields .......................................................................................... 47
Conclusion ....................................................................................................................... 49
Chapter 3 Additional Product in the Grid: Effect of pH and Temperature on
Microbial Community Structure and Carboxylic Acid Yield during the Acidogenic
Digestion of Duckweed .................................................................................................... 50 Abstract ............................................................................................................................ 50 Introduction ...................................................................................................................... 51 Materials and Methods ..................................................................................................... 53
vi
Analytical methods ................................................................................................... 53 Plant material and growth conditions ....................................................................... 54 Inoculum .................................................................................................................. 55 Acidogenic digestion ................................................................................................ 56 Carbon balance ......................................................................................................... 57 DNA extraction, PCR amplification, and high-throughput sequencing ................... 58 Bioinformatics .......................................................................................................... 59 Statistical analysis .................................................................................................... 61
Results .............................................................................................................................. 61 Acidogenic digestion performance ........................................................................... 61 Carbon balance ......................................................................................................... 66 Microbial community analysis ................................................................................. 69
Discussion ........................................................................................................................ 75 Effect of pH and temperature on acidogenic digestion performance ....................... 75 Effect of operating conditions on microbial community diversity and
composition ...................................................................................................... 77 Alpha diversity ................................................................................................. 77 Beta diversity .................................................................................................... 78 Composition ..................................................................................................... 81
Relationships between operating conditions, microbial community structure,
and end products ............................................................................................... 82 Conclusions ...................................................................................................................... 87
Chapter 4 Anaerobic Bioprocessing of Wastewater-Derived Duckweed:
Maximizing Product Yields in a Biorefinery Value Cascade .......................................... 89 Abstract ............................................................................................................................ 89 Introduction ...................................................................................................................... 90 Materials and Methods ..................................................................................................... 93
Analytical methods ................................................................................................... 93 Plant material, cultivation, and pre-processing ........................................................ 94 Inocula ...................................................................................................................... 96
Yeast strain ....................................................................................................... 96 Acidogenic anaerobic seed ............................................................................... 97 Methanogenic anaerobic seed........................................................................... 97
Anaerobic bioprocessing scenarios in a biorefinery system ..................................... 97 Ethanol fermentation and distillation ............................................................... 98 Acidogenic anaerobic digestion and membrane separation ............................. 99 Biochemical methane potential (BMP) assays ................................................. 100
Overall duckweed-to-bioproduct conversion yields and carbon balances ............... 101 Duckweed-to-bioproduct conversion yields and carbon balances in
individual reactors .................................................................................... 101 Duckweed-to-bioproduct conversion yields and carbon balances of
sequential processes .................................................................................. 102 Fertilizer potential assessment ................................................................................. 103 Statistical analysis .................................................................................................... 104
Results and Discussion ..................................................................................................... 104 Ethanol fermentation and distillation ....................................................................... 104 Acidogenic anaerobic digestion ............................................................................... 105 Biochemical methane potentials ............................................................................... 107
vii
Overall duckweed-to-bioproduct conversion yields and material balances ............. 109 Duckweed-to-bioproduct conversion yields and carbon balances in
individual reactors .................................................................................... 109 Duckweed-to-bioproduct conversion yields of sequential processes ............... 111
Fertilizer potential .................................................................................................... 113 Conclusions ...................................................................................................................... 115
Chapter 5 Techno-economic Analysis and Life Cycle Assessment of Wastewater-
Derived Duckweed Biorefinery Supply Chain System .................................................... 116 Abstract ............................................................................................................................ 116 Introduction ...................................................................................................................... 117 Methodology .................................................................................................................... 119
Supply chain components ......................................................................................... 119 Feedstock production and harvesting ............................................................... 120 Feedstock drying and transportation ................................................................ 124 Biorefinery processes ....................................................................................... 125
Techno-economic analysis overview ....................................................................... 129 Duckweed production and harvesting .............................................................. 130 Feedstock handling and transportation ............................................................. 135 Biorefinery processes ....................................................................................... 135
Life cycle assessment overview ............................................................................... 139 Goal and scope definition ................................................................................. 139 Life cycle inventory (LCI)................................................................................ 140 Life cycle impact assessment (LCIA) .............................................................. 141
Results and Discussion ..................................................................................................... 142 Techno-economic analysis ....................................................................................... 142
Duckweed production and harvesting .............................................................. 142 Biorefinery processes ....................................................................................... 143
Life cycle assessment ............................................................................................... 145 Conclusion and Future Work ........................................................................................... 147
Chapter 6 Conclusions, Significance and Future Work .................................................. 148
REFERENCES................................................................................................................. 152
Appendix A Chapter 2 Additional File ........................................................................... 171
Appendix B Chapter 3 Additional File ........................................................................... 173
Appendix C Chapter 4 Additional File ........................................................................... 179
Appendix D Chapter 5 Additional File ........................................................................... 185
viii
LIST OF FIGURES
Figure 1-1: A schematic of the Penn State Eco-MachineTM. .................................................. 18
Figure 1-2: Sequential stages of the MAD process (Modified from: McCarty, 1964). .......... 20
Figure 1-3: Sequential stages of the acidogenic anaerobic digestion process (Modified
from: McCarty, 1964). ..................................................................................................... 24
Figure 1-4: Types of feedstock for ethanol production with example crops (Source: Khan
& Dwivedi, 2013). ........................................................................................................... 27
Figure 1-5: Schematic flow diagram of a simultaneous ethanol fermentation and MAD
process. ............................................................................................................................. 30
Figure 1-6: Phases of an LCA (Source: ISO, 2006). ............................................................... 32
Figure 2-1: Cumulative methane production (ml CH4/ g volatile solids added) in batch
reactors fed with raw Eco-MachineTM duckweed (EM), raw Living-Filter duckweed
(LF), fermented Eco-MachineTM duckweed (FEM), fermented Living-Filter
duckweed (FLF) at different substrate-to-inoculum (S/I) ratios and with and without
the addition of Vanderbilt Medium (VM): A) S/I = 0.5, without VM; B) S/I = 0.5,
with VM; C) S/I = 1.0, without VM; D) S/I = 1.0, with VM. .......................................... 46
Figure 3-1: Volatile Fatty Acid profiles of the acidogenic duckweed reactors over 21
days. Legend: Reactors were operated under: A) Acidic Mesophilic, B) Acidic
Thermophilic, C) Basic Mesophilic, D) Basic Thermophilic conditions. Narrow
stacked columns represent blank reactors (no inoculum) whereas thick stacked
columns represent active (with inoculum) reactors. Error bars are cumulative
standard deviations of the individual stacked bars. .......................................................... 62
Figure 3-2: Cumulative biogas, hydrogen, methane, and carbon dioxide yields of the
acidogenic duckweed reactors over 21 days. Reactors were operated under: A)
Acidic Mesophilic; B) Acidic Thermophilic; C) Basic Mesophilic; D) Basic
Thermophilic conditions. Blank (no inoculum) reactors are represented as empty
bullets whereas active (with inoculum) reactors are represented as solid bullets. ........... 64
Figure 3-3: Carbon balance of the acidogenic duckweed reactors. Total carbon percent
contributions from initial duckweed, inocula, and alkalinity, and final soluble (<0.2
µm), particulate (>0.2 µm; <340 µm), solid (>340 µm), and gaseous phases of the
reactors under: A) Acidic Mesophilic, B) Acidic Thermophilic, C) Basic Mesophilic,
D) Basic Thermophilic conditions. Error bars are cumulative standard deviations of
the individual measurements. ........................................................................................... 67
Figure 3-4: Class-level relative abundance taxonomic bar plot. .............................................. 71
Figure 3-5: A) Weighted and B) unweighted PCoA plots. ...................................................... 80
ix
Figure 4-1: Volatile Fatty Acid profiles of the acidogenic duckweed reactors over ten
days. Reactors were fed with: A) raw; B) pretreated; C) saccharified; D) saccharified
and fermented duckweed.................................................................................................. 105
Figure 4-2: Cumulative methane yields of the methanogenic duckweed reactors over 42
days. Reactors were fed with raw, pretreated, saccharified, and saccharified and
fermented duckweed: A) not subjected to acidogenic digestion; B) subjected to
acidogenic digestion and membrane separation. Control biomethane yields were
subtracted from each case. ............................................................................................... 107
Figure 4-3: Percent initial and final carbon contents of the bioreactors fed with raw,
pretreated, and saccharified duckweed and subjected to: A) fermentation; B)
acidogenic digestion; C) methanogenic digestion. The desired product in each
process was ethanol (A), VFAs (B), or methane (C). ...................................................... 110
Figure 4-4: Carbon-to-carbon conversion yields as a result of individual bioprocesses,
two sequential bioprocesses, and three sequential bioprocesses for: A) saccharified;
B) pretreated; C) raw duckweed. ..................................................................................... 112
Figure 4-5: Fertilizer potentials of reactor residuals in terms of total nitrogen (TN as N),
total phosphorus, and potassium concentrations on a dry basis. Stacked bars
represent total ammonia nitrogen (TAN) and other nitrogen species. ............................. 114
Figure 5-1: System boundaries of the conceptual supply chain. Downstream processes are
excluded. .......................................................................................................................... 120
Figure 5-2: Illustration of the dynamic Stella Architect model used for duckweed growth
and harvesting. ................................................................................................................. 123
Figure 5-3: Potential biorefinery process scenarios. The solid red line shows the scenario
presented in this chapter. .................................................................................................. 126
Figure 5-4: A breakdown summary of the capital (A) and operating (B) expenses of a
wastewater treatment – duckweed production system. .................................................... 142
Figure 5-5: Breakdown of costs and revenues for the discounted cash flow analysis for
minimum biomass selling price of 25.2 USD .................................................................. 143
Figure 5-6: Minimum biomass selling price at differenc considerations of wastewater
treatment credits ............................................................................................................... 143
Figure 5-7: Minimum ethanol selling price at different daily processing capacities. .............. 144
Figure 5-8: Contribution of life cycle phases of wastewater-derived duckweed
biorefinery supply chain to environmental impact categories when duckweed is
grown in land-based ponds in Florida, USA. ................................................................... 146
x
LIST OF TABLES
Table 1-1: Compositional analysis of different duckweed species. ......................................... 8
Table 1-2: Summary of nutrient removal studies using duckweed. ......................................... 9
Table 1-3: Duckweed type, enzymes, and microorganisms used in other fermentation
studies and their associated end products. ........................................................................ 12
Table 2-1: Bioethanol, biomethane, and bioenergy yields from Eco-Machine (EM) and
Living-Filter (LF) duckweed biomass through separate and coupled ethanol
fermentation and anaerobic digestion processes. ............................................................. 49
Table 3-1: Final volatile fatty acid yields of the blank and active reactors under acidic
mesophilic, acidic thermophilic, basic mesophilic, and basic thermophilic
conditions. ........................................................................................................................ 63
Table 3-2: Alpha diversity metrics for microbial populations in duckweed acidogenically
digested under different environmental conditions. ......................................................... 69
Table 3-3: Relative abundance (R.A.) and Cumulative Abundance (C.A.) of top five
genera in each reactor group operated under acidic conditions. ...................................... 72
Table 3-4: Relative abundance (R.A.) and Cumulative Abundance (C.A.) of top five
genera in each reactor group operated under basic conditions. ........................................ 73
Table 3-5: Relative abundance (R.A.) and Cumulative Abundance (C.A.) of top five
archaeal genera in each reactor group. ............................................................................. 74
Table 3-6: Summary of microbial populations and end product profiles under various
operating conditions. ........................................................................................................ 85
Table 5-1: Wastewater treatment – duckweed production pond specifications. ...................... 121
Table 5-2: Wastewater quality change in duckweed ponds ..................................................... 122
Table 5-3: Duckweed liquefaction unit specifications. ............................................................ 127
Table 5-4: Saccharification unit specifications ........................................................................ 128
Table 5-5: Fermentation unit specifications............................................................................. 128
Table 5-6: Anaerobic digestion unit specifications .................................................................. 129
Table 5-7: Total direct expenses of a duckweed production/wastewater treatment system. ... 131
Table 5-8: Total indirect expenses of a duckweed production/wastewater treatment
system. ............................................................................................................................. 131
xi
Table 5-9: Total capital expenses of a duckweed production/wastewater treatment
system. ............................................................................................................................. 132
Table 5-10: Operating expenses of wastewater treatment - duckweed production system. .... 133
Table 5-11: Wastewater treatment credit ................................................................................. 134
Table 5-12: Input data for discounted cash flow rate of return analysis of wastewater
treatment - duckweed production system. ........................................................................ 134
Table 5-13: Total direct expenses for the biorefinery. ............................................................. 136
Table 5-14: Total indirect expenses for the biorefinery. .......................................................... 137
Table 5-15:Total capital investment for the biorefinery. ......................................................... 137
Table 5-16: Fixed operating expenses ..................................................................................... 137
Table 5-17: Variable operating expenses ................................................................................. 138
Table 5-18: Input data for discounted cash flow rate of return analysis of wastewater
treatment - duckweed production system. ........................................................................ 139
xii
ACKNOWLEDGEMENTS
I would like to express the deepest appreciation to my dissertation advisers, Dr. Rachel
Brennan and Dr. Tom Richard, for their insights, guidance, and encouragement on my way to
becoming an independent researcher. I would also like to express my gratitude to other members
of my Ph.D. committee, Charles Anderson, John Regan, and Deborah Sills, for dedicating time to
serve on my committee and providing their guidance and expertise.
I am also grateful for the mentorship and help I received toward completion of this
dissertation work from our laboratory coordinator, David Jones, and fellow graduate students,
Michael Shreve, Anahita Bharadwaj, Boya Xiong, Sarah Cronk, Travis Tasker, Benjamin Roman,
and others. I want to express my gratitude to the Office of Student Disability Resources of Penn
State as well, for generously funding assistance for tasks requiring high visual acuity, and the
undergraduate assistants, Kara Slocum, Nicole Urban, Kayla Wirth, and Mpila Nkiawete, who
helped me complete these tasks.
I would also like to thank my parents Gulay Calicioglu and Bahri Can Calicioglu, not
only for their love and support but also for raising me as an environmentally-conscious and
socially-aware global citizen. Thank you also, my brother, Kaan Calicioglu, for complementing
me so well to bring out my potential. The creativity and sense of humor we developed together
have come in handy along the way.
Finally, I would like to express my dearest gratitude to Mert Yigit Sengul, my partner in
life and my “life coach”, who has a role in every single accomplishment of our “team” over the
last decade. Thank you for teaching me to put my true self into all I do in the world. Without this
attitude, none of my accomplishments would have been possible. I am looking forward to our
future endeavors together.
1
Chapter 1
Introduction
Pre-industrial societies mainly relied on renewable resources to supply necessary food
and non-food products. After the Industrial Revolution took place in the early 18th century, energy
and resource demands continued to increase, and massive amounts of nonrenewable resources
such as coal, crude oil, and gas were consumed as primary energy and chemical precursors.
Modern economies utilize renewable resources only to fulfill a minor fraction of their total energy
and chemical demands (Hatti-Kaul et al, 2007). The repetitive oil crises, the volatility of oil
prices, and rising energy demand worldwide led to an increased awareness of the unsustainable
nature of fossil fuel-based economies over the last half century (Aiello-Mazzarri et al. 2006).
Furthermore, increases in fossil-based resource consumption have historically caused
environmental disasters such as the London Great Smog in 1951, and the major oil spills of the
20th and 21st centuries. In addition to episodic events, the increase in atmospheric carbon dioxide
levels due to petroleum-based fuel combustion is known to cause global warming and climate
change (Aiello-Mazzarri et al., 2006). Additionally, recalcitrant petrochemical products pose
stress to aquatic and terrestrial environments. A fossil fuel-based economy is therefore not only
economically, but also environmentally, unsustainable.
The economic and environmental disadvantages of fossil fuels have led to increased
efforts in finding alternative resources to fulfill energy and chemical needs. Given that renewable
alternatives should be abundant, inexpensive, and complementary to the food system, non-edible
plant-based raw materials (biomass) have gained attention recently as a potential substitute for
petroleum products. Conversion of biomass into biofuels can potentially decrease carbon dioxide
emissions, increase energy security, and contribute to rural development (Cherubini, 2010).
2
So far, most biomass-to-energy work has focused on the production of a single valuable
end product, usually bioethanol or biodiesel. In contrast, current fossil fuel production
technologies involve the production of a main product along with co-products and byproducts
through a complex and integrated refinery system. In a conventional petroleum refinery,
relatively small volumes of high value coproducts such as industrial chemicals and lubricants
represent a large fraction of the product portfolio’s value. For economically feasible large-scale
production of biofuels, a conceptually similar biorefinery approach may have similar advantages,
by separating and increasing the value of different biomass components and by targeting a variety
of end products (Biddy et al., 2016). A viable biofuel production system should lead to the
production of other lower-value, yet higher volume, products such as animal feed or fertilizers, as
well as an additional energy products such as heat or electricity, in addition to higher-value
products (Cherubini, 2010).
Biomass sources, composition, availability and costs are of particular importance for
providing a sustainable and reliable feedstock for biofuel production. This is primarily because
future biorefineries are likely to be supply limited, in contrast to traditional refineries which are
demand-limited. The ideal feedstock, therefore, should enable biofuel production with minimum
social, economic, and environmental challenges, in order to ensure a continuous and robust
supply. First generation biofuels (mainly bioethanol and biodiesel derived from edible food
crops), while they are the economically most feasible alternative so far, have not been socially
accepted as they raise “food vs. energy” debates. Although second generation biofuels, derived
from non-edible plant biomass, do not pose such obvious ethical concerns, they do need to be
carefully integrated into the food system as energy crops could displace food crops on prime
agricultural land. These second generation “cellulosic” biofuels have not been commercialized
yet largely due to the cost and environmental impacts associated with pre-processing
requirements, which are needed for lignocellulosic biomass sourced from woody materials, crop
3
residues and herbaceous grasses whose high lignin content makes it recalcitrant to microbial and
enzymatic degradation. The production and harvesting costs associated with microalgae
production, potentially a third generation biofuel resource, also bring challenges for its
commercialization. It has been argued that microalgae production must be coupled with
wastewater treatment to increase economic feasibility (Dale et al., 2010).
Lemnaceae (duckweed), a family of simple, fast-growing, floating aquatic plants, is a
promising option for biofuel production and holds several advantages over other bioenergy
feedstocks: (1) it can accumulate up to 43% of its biomass as easily degradable starch; (2) it does
not require prime agricultural land for production; (3) its cell walls contain very little lignin, and
so do not require energetically- or chemically- intensive pretreatments prior to bioconversion into
fuels and chemicals; (4) its small size (1 mm – 1 cm) and uniform structure greatly reduce the
need for grinding or milling; (5) it can easily be harvested from the water surface (in contrast to
microalgae); and (6) it can be grown using nutrients derived from wastewater, and therefore can
convert a common waste stream directly into a valuable resource (Cui and Cheng, 2015).
The conversion of duckweed, grown as a byproduct of wastewater treatment, into
biofuels has been previously investigated, but for a limited set of pathways implementing the
thermochemical and sugar platforms that focus on a single primary process and biofuel product.
These prior studies have mostly focused on the technical viability of duckweed-based bioethanol
production using laboratory- and pilot-scale enzymatic saccharification and fermentation
experiments (Cui and Cheng, 2015). In contrast, the feasibility of a duckweed-to-biofuel system
has not yet been analyzed from the perspective of a complete biorefinery concept.
This dissertation has attempted to address the key knowledge gaps preventing the
establishment of integrated wastewater-derived duckweed biorefineries. Technical,
environmental, and economic evaluations were performed to enable the simultaneous utilization
of duckweed as a reliable wastewater treatment strategy and as a feedstock for bioenergy
4
generation. Technical evaluations included selection of bioconversion technique(s) through
anaerobic bioprocesses, including methanogenic digestion, ethanol fermentation, and acidogenic
digestion, in a sequential value cascade. Economic evaluation of the duckweed biorefineries
concept was performed with a major focus on the construction of a supply chain model for large-
scale application scenarios, in order to demonstrate the techno-economic feasibility of the
integrated system of interest. Environmental evaluation of the system focused on the life cycle
performance of duckweed production, as a co-product of wastewater treatment.
This dissertation consists of seven chapters. The first chapter (Chapter 1) provides
general background on related scientific concepts, fundamental methodologies, and literature
review, as well as the current state-of-the-art in duckweed research. The next five chapters are
grouped into two phases, each focusing on either experimental (technical), or modeling
(environmental and economic) studies for the evaluation of duckweed biorefineries, as follows:
Phase 1: Technical evaluation of wastewater-derived duckweed bioconversion through
anaerobic bioprocesses, in a biorefinery concept.
Chapter 2: Proof of Concept - Sequential Ethanol Fermentation and Methanogenic Anaerobic
Digestion of Duckweed.
This manuscript has been published in Bioresource Technology:
Calicioglu, O., Brennan, R.A., 2018. Sequential ethanol fermentation and
anaerobic digestion increases bioenergy yields from duckweed. Bioresour.
Technol. 257, 344–348. doi:10.1016/j.biortech.2018.02.053
Chapter 3: Additional bioproduct in duckweed biorefinery - Acidogenic Digestion of Duckweed
Using Mixed Anaerobic Cultures to Maximize Carboxylic Acid Yields.
5
Note that the molecular techniques applied in Chapter 3 were performed by Michael J. Shreve in
a collaborative effort to characterize the microbial communities in acidogenic digestions of the
aquatic plant duckweed under various pH and temperature conditions.
This manuscript has been published in Biotechnology for Biofuels:
Calicioglu, O., Shreve, M.J., Richard, T.L., Brennan, R.A., 2018. Effect of pH
and temperature on microbial community structure and carboxylic acid yield
during the acidogenic digestion of duckweed. Biotechnol. Biofuels 1–19.
doi:10.1186/s13068-018-1278-6.
Chapter 4: Biorefinery Value Cascade - Maximizing Product Yields from Anaerobic
Bioprocessing of Wastewater-Derived Duckweed in a Biorefinery System.
This manuscript is in review for publication in Bioresource Technology:
Calicioglu, O., Richard, T. L., and Brennan, R. A. Maximizing product yields
from anaerobic bioprocessing of wastewater-derived duckweed in a biorefinery
system. Bioresource Technology, 2019, in review.
Phase 2: Techno-economic and enviornmental evaluation of integrated wastewater-derived
duckweed biorefineries supply chain.
Chapter 5: Techno-economic Analysis and Life Cycle Assessment of Wastewater-Derived
Duckweed Biorefineries.
The contents of Chapter 5 will be submitted for publication in the Journal of Cleaner
Production.
Authors: Ozgul Calicioglu, Chris Mutel, Deborah L. Sills, Tom L. Richard, and
Rachel A. Brennan
The last chapter (Chapter 6) concludes the dissertation and lays out potential future work.
6
General Background
Lemnaceae (duckweed)
Lemnaceae (duckweeds) represent a family of simple, fast-growing, floating aquatic
plants, with five genera (Landoltia, Lemna, Spirodela, Wolffia, and Wolfiella). Within these
genera, 37 species have been identified as the most widely accepted classification (Xu et al.,
2014; Cui & Cheng, 2015). In this section, general characteristics of duckweed, its potential for
wastewater treatment, and general background on utilization of duckweed as a bioenergy
feedstock are presented.
General characteristics
Duckweeds can grow in a wide range of environmental conditions, including polluted
and/or eutrophic water bodies and saline waters. Duckweed is reported to tolerate a broad range
of pH conditions from 5 and 9 but grows best within the range of 6.5-7.5 (FAO 1999). Some
strains can live in all climatic regions, such as Lemna gibba, which can grow in temperatures
from 5° C up to 34° C, with an optimum range of 18 °C – 30 °C (Oron et al., 1986). In winter,
duckweed stays dormant by forming turions (i.e., a starch-rich frond with relatively higher
density), and sinking to the bottom of water bodies.
Duckweeds have adapted to aquatic habitats and therefore lack distinguishable stems and
leaves; instead, they consist of simpler physical structures called fronds and simple roots. The
frond sizes may vary between 1 mm and 1.5 cm (Chaiprapat et al., 2005). Each plant produces
approximately 20 daughter fronds throughout its life cycle, alternately through the meristem
regions of each frond (Oron et al., 1986). Duckweeds have high growth rates and short doubling
times, as low as 16 – 24 h under ideal conditions (J. Xu et al., 2012). Duckweed prefers ortho-
7
phosphate as its primary phosphorus source and achieves a higher growth rate and more nitrogen
accumulation in the presence of ammonium as a nitrogen source rather than nitrate or nitrite.
However, the existence of ammonia above pH 9 can be inhibitory to duckweed growth (Culley et
al., 1981).
Duckweed is reported to have highly variable, yet manipulatable, starch content, ranging
between 3 – 43% of the dry biomass among different species and strains (Cheng and Stomp,
2009). Starch accumulation in duckweed can be triggered by adjusting environmental conditions
to meet those of its dormant state, i.e., creating stress conditions by altering the pH, inducing
nutrient starvation, reducing temperature, or reducing the photoperiod / light intensity. The effect
of nutrient starvation on starch accumulation is well documented. J. J. Cheng & Stomp (2009)
achieved a 45.8% (dry basis) starch content in duckweed, by growing S. polyrrhiza on
anaerobically treated swine waste and then transferring the plant to tap water for 5 days. In
another study, Spyrodela polyrrhiza was transferred from nutrient-rich swine lagoon wastewater
to well water to achieve a final starch content of 29.8%, representing a 64.9% increase over initial
conditions (J. Xu et al., 2011). Similarly, stress conditions were induced in another duckweed
(Landoltia punctate) by transferring the biomass from nutrient rich solution to distilled water,
thereby increasing the starch content from 3% to 45% within 7 days (Huang et al., 2014). The
light intensity and photoperiod effects on starch accumulation in duckweed have also been
demonstrated. McCombs and Ralph (1972) reported that the starch content of duckweed (S.
polyrrhiza) left in the dark for 6 days increased three fold compared to photosynthetically active
biomass grown on the same medium. The effect of nutrient starvation and light limitation was
studied by transferring Lemna minor from swine wastewater to a glucose-rich but nutrient
deprived solution in the dark, resulting in up to a 36% increase in starch content (Ge et al., 2012).
Cui et.al. (2010) examined the effect of temperature on starch accumulation of duckweed and
reported higher starch accumulation at a temperature of 5 o C than at 15 o C and 25oC. They
8
concluded that at 5 oC, with a photoperiod of 12 h, the duckweed starch content increased by
59.3% in two days, through transfer of Spirodela polyrrhiza from a nutrient-rich solution to well
water. Table 1-1 shows examples of different duckweed compositions.
Utilization of duckweed in wastewater treatment processes
Due to their high nutrient uptake capabilities, growth rates, and adaptability to a broad
range of nutrient concentrations, duckweed species such as Lemna minor, Spirodela polyrrhiza,
and Lemna aequinoctialis have been widely studied for nutrient recovery from domestic and
agricultural wastewaters (Cheng and Stomp, 2009). Studies on nutrient removal using duckweed
have been summarized in Table 1-2.
Table 1-1: Compositional analysis of different duckweed species.
Constituents (% dry weight) Landoltia punctata Landoltia punctata Lemna minor
Extractives 13.04 ± 1.98
Lipid 8.7± 0.6
Crude protein 16.27 ± 0.12 21.5 32.2 ± 0.7
Starch 24.59 ± 0.67 47.8 10.3 ± 0.8
Cellulose 13.31 ± 0.41 14.26 9.4 ± 0.5
Xylene 1.61 ± 0.01
Xylose 2.7 + 0.6
Galactose 3.46 ± 0.32 1.4 ± 0.1
Arabinose 1.32 ± 0.02 2.1 ± 0.5
Apiose 3.1 ± 0.3
Lignin 1.16 ND
Acid insoluble lignin 5.55 ± 0.36
Ash 3.48 ± 1 17.7 ± 0.1
Reference (Chen et al., 2012) (Su et al., 2014) (Ge et al., 2012)
9
Utilization of duckweed for nutrient removal is also beneficial for greenhouse gas
abatement through photosynthesis. In order to sustain efficient removal of nutrients and
atmospheric carbon dioxide, duckweed decomposition and the release of these accumulated
products back into the environment should be avoided by regular harvesting of the biomass.
Sustainable management of the harvested biomass is the key point for accomplishing nutrient
removal with a net positive impact on the environment in terms of improvement of water quality
and greenhouse gas emissions. As a result of nutrient uptake, the produced duckweed biomass is
rich in protein and starch content; therefore, it can be used as an animal feed (J. J. Cheng &
Stomp, 2009; and J. Xu et al., 2012), or as a feedstock for conversion into synthetic biofuels
(Baliban et al., 2013), or bioethanol through thermochemical or biological processes (J. J. Cheng
& Stomp, 2009; Chen et al., 2012; Ge et al., 2012; J. Xu et al., 2012; C. Yu et al., 2014; Cui &
Table 1-2: Summary of nutrient removal studies using duckweed.
Species Wastewater Removal Rate Removal
Efficiency
Duckweed
Growth Rate
Reference
Lemna minor 50%, 33%,
25%, and
20% swine
lagoon liquid
2.11 g m–2 day–1 TN;
0.59 g m–2 day–1 TP
- 29 g m–2 day–1 (Cheng et
al., 2002)
Spirodela oligorrhiza
6% swine
lagoon water
- 83.7% TN;
89.4% TP
- (Xu and
Shen,
2011)
Lemna minor Swine lagoon
wastewater
- 100% NH4+–N;
75% NO3-–N;
74.8% PO43-– P
3.5 g m–2 day–1 (Ge et al.,
2012)
Spirodela polyrrhiza
Anaerobically
treated swine
wastewater
1.3 g m–2 day–1 NH3-N;
0.09 g m–2 day–1 PO4-P
- 9.25 g m–2 day–1 (Xu et al.,
2012)
Lemna minor Stormwater - 79 ± 3 % NH4+–N;
86 ± 2 % NO3-–N;
56 ± 7 %
orthophosphate
- (Sims et
al., 2013)
Lemna aequinoctialis
Domestic
wastewater
- 80% TN
95% TP
4.3 g m–2 day–1 (Yu et al.,
2014)
10
Cheng, 2015). Biofuel production processes from duckweed are discussed further in the
Duckweed as a bioenergy feedstock section.
Duckweed as a bioenergy feedstock
Several advantages of duckweed have led to its widespread cultivation, including: (1) one
of the highest growth rates of higher plants; (2) a longer growing period compared to other plants
(Chaiprapat et al., 2005); (3) world-wide distribution with adaptation to a wide variety of
environmental/aquatic conditions; and (4) contribution to water quality enhancement during the
growth process by intensive nutrient uptake (Culley et al., 1981). These advantages originally led
to the use of duckweed for waste management and agricultural purposes, and more recently have
resulted in the consideration of the plant as a potential bioenergy feedstock.
According to the results of a pilot-scale wastewater-derived duckweed production study
conducted by Xu et al. (2012), the dry matter production rate could reach 27.3 Mg ha-1 year-1,
with an average starch content of 18.6%. These calculations yielded duckweed starch of 5.08 Mg
ha-1 year-1 within a 9-month growing season under the climate conditions of North Carolina. The
starch yield of duckweed was comparable to that of corn, calculated in the same study as 5.7 Mg
ha-1 year-1.
Duckweed provides several benefits when compared to other energy crops. The
production of duckweed rich in starch and cellulose content does not require agricultural land or
fresh water for cultivation; therefore, its utilization as a biofuel resource does not raise ethical
concerns about food security, as compared to the conversion of starch crops such as corn and
sugarcane into bioethanol. Moreover, the low lignin content of duckweed species makes it a
feasible alternative for conversion into bioethanol, since it does not require intensive
pretreatments prior to saccharification as lignocellulosic agricultural residues and energy crops
11
do. Given these advantages, the popularity of duckweed as an environmentally and economically
sustainable feedstock for the production of biofuels has been increasing. The duckweed-to-biofuel
conversion alternatives demonstrated by others so far include biological processes yielding
bioethanol and other higher alcohols like biobutanol, and thermochemical processes which yield
oil, biochar, or synthetic biofuels.
The technical feasibility of duckweed-based bioethanol production has been
demonstrated in laboratory- and pilot-scale enzymatic saccharification and fermentation
experiments. It was reported by Xu et al. (2011), that up to 96.8% of the theoretical glucose could
be recovered by saccharification of S. polyrrhiza starch using the enzymes α -amylase,
pullulanase, and amyloglucosidase for hydrolysis. After saccharification, 97.8% of the theoretical
ethanol yield could be achieved by fermentation with a yeast loading of 6.2 g dry weight L-1. In
another study, the starch content and ethanol fermentation of duckweed (Lemna aequinoctialis)
grown in either Schenk & Hildebrandt (SH) growth medium or sewage wastewater were
compared by Yu et al. (2014). The final starch contents were 39 ± 1.95 and 34 ± 1.62 for
duckweed grown on SH medium and sewage wastewater, respectively. Both duckweeds were
then subjected to enzymatic saccharification and fermentation, resulting in nearly equivalent
sugar recoveries (94.1% and 94.6%) and ethanol yields (0.44 and 0.45 g g-1 (as glucose)) for
duckweed grown on SH medium and sewage wastewater, respectively. Table 1-3 summarizes the
duckweed type, enzymes, and microorganisms used in previous fermentation studies and their
associated end products.
12
Several researchers have focused on increasing the starch content of duckweed, in order
to increase the bioethanol production yield. For example, Xu et al. (2011) transferred Spirodela
polyrrhiza grown in a pilot-scale pond of diluted swine wastewater to well water for 10 days,
resulting in an enhanced starch accumulation with 64.9% increase. This biomass was then
fermented by yeast, and 94.7% of the theoretical starch was converted into bioethanol. Assuming
a duckweed harvesting frequency of three times a week, this conversion was estimated to
correspond to an annual bioethanol yield of 6.42 Mg per hectare per year, which is 50% higher
than that of the annual maize-based bioethanol yield.
Other than starch, other cell wall materials in duckweed, such as cellulose, can be
converted into simple sugars for fermentation. Zhao et al. (2012) used mixtures of commercially
available cellulase enzymes to recover glucose from the cell walls of duckweed. Results of the
Table 1-3: Duckweed type, enzymes, and microorganisms used in other fermentation studies and
their associated end products.
Duckweed Enzymes Microorganisms End product Ref.
Lemna minor α -amylase (Sigma A4582)
α -amyloglucosidase
(Sigma A7095)
cellulase (Sigma C2730)
Novozyme 188 (Sigma
C6105)
Saccharomyces cerevisiae
strain, ATCC 24859
Ethanol (Ge et al.,
2012)
Lemna aequinoctialis
strain 6000
α -amylase (Sigma A4582),
α -amyloglucosidase
(Sigma A7095)
pullulanase (Sigma P1067)
Angel Yeast (Angel Yeast
Co., Ltd, China)
Ethanol (Yu et al.,
2014)
Landoltia punctata α- amylase
glucoamylase
pectinase
xylanase
Clostridium acetobutylicum,
Mutant Saccharomyces
cerevisiae AH109.
Bioengineered strains of
Escherichia coli
Butanol and
other higher
alcohols
(Su et al.,
2014)
S. polyrrhiza α -amylase
pullulanase
amyloglucosidase
Saccharomyces cerevisiae
(ATCC 24859) cells
Ethanol (Xu et al.,
2011)
Landoltia punctata α -amylase
Glucoamylase
pectinase
Saccharomyces cerevisiae
strain CCTCC M206111
Ethanol (Chen et al.,
2012)
13
study revealed that approximately 0.6 % of the fresh weight of duckweed can be converted into
glucose. Chen et al. (2012) applied pectinase pretreatment and achieved a 142% increase in
saccharification efficiency, using 26.54 units of pectin transeliminase dosing per gram of
duckweed mash at 45 °C for 300 min. In the same study, fermentation experiments were also
performed, resulting in a concentration of 30.8 ± 0.8 g/L of ethanol at a production rate of 2.20
g/L/h and a 90.04% fermentation efficiency. Ge et al. (2012) cultivated Lemna minor on two
different media: swine lagoon wastewater and Schenk & Hildebrandt (SH) medium for starch
accumulation and fermentation experiments. The starch content was increased up to 10–36%
(w/w) in duckweed biomass by nutrient starvation or growing in dark with addition of glucose.
Enzymatic hydrolysis performed with the addition of both α–amylase and cellulase resulted in
96.2% (w/w) of glucose. The hydrolysates were then fermented by self-flocculating yeast
(SPSC01) and conventional yeast (ATCC 24859) and the highest ethanol yield of 0.485 g ethanol
g-1 (glucose) was achieved.
In addition to bioethanol, duckweed has been studied for its potential conversion into
higher alcohols, such as butanol, after acid or enzymatic pretreatment by organisms such as
Clostridium acetobutylicum, mutant yeast strains, and bioengineered strains of Escherichia coli.
Following acid pretreatment, fermentation by C. acetobutylicum CICC 8012 provided butanol
and total solvent concentrations of 12 and 20 g/L, respectively. Fermentation of enzymatic
hydrolysate by the same strain provided similar results of 12 and 20 g/L. The mutant yeast strain
produced 24 g/L ethanol and 680 mg/L of isopentanol from duckweed, which is a 15 times higher
ethanol yield compared to conventional yeast. Bioengineered strains of E. coli produced 16 mg/L
butanol, 25 mg/L isopentanol, and 196 mg/L pentanol from the acid hydrolysate of duckweed (Su
et al., 2014).
Another method that has been tested for the bioconversion of duckweed into biofuels is
anaerobic fermentation to produce biohydrogen. Fermentation of acid-pretreated duckweed,
14
harvested from a swine wastewater treatment system, was found to produce up to 75 mL
biohydrogen per gram of dry duckweed in 7 days, at a concentration of 42% of the headspace gas
produced (Xu and Deshusses, 2015).
Although anaerobic digestion of organic materials into biomethane is an efficient
technology that has been widely studied for a variety of organic wastes and feedstocks, recent
studies on the anaerobic digestion of duckweed are rather limited. In one study, duckweed
(Lemna sp.) was anaerobically fermented into methane in mesophilic (37oC) and thermophilic
(60oC) completely stirred tank reactors with a 26 days retention time and 5% solids loading. It
was reported that 25-34% and 32-46% of the energy value in the duckweed was recovered under
mesophilic and thermophilic conditions, respectively. Low bioconversion efficiency was
attributed to the absence of steady state conditions as well as the presence of non-biodegradable
portions of the biomass, which may require pretreatment for increased biomethane yield (Wise et
al., 1979).
The high nutrient and metal uptake capabilities of duckweed has encouraged the use of
duckweed for the recovery of metals and their subsequent supplementation into anaerobic
digestion processes of feedstocks. Similarly, Jain et al., (1992) investigated the capacity of
duckweed (Lemna minor) to adsorb iron, copper, cadmium, nickel, lead, zinc, manganese, and
cobalt. It was shown that iron and manganese did not cause toxicity; however, copper, cobalt,
lead, and zinc did show toxicity during the anaerobic digestion of heavy metal duckweed. Yet, the
methane content was higher than that of non-contaminated duckweed biomass. The highest
biogas yield was reported as 176 L/kg, with a methane content of 60%, from the anaerobic
digestion of manganese-contaminated duckweed. In another study, the effect of iron-enriched
duckweed supplementation into laboratory scale batch and semi-continuous reactors was
investigated by Clark & Hillman (1996). The results revealed that iron-rich duckweed addition
facilitated biomethane production in batch reactors, reducing the time until peak methane
15
production was achieved from 40 days to 15 days compared to no duckweed addition. In the same
study, about a 44% increase in gas production was reported in semi-continuous mode operation,
compared to no duckweed supplementation. In another study, duckweed (Landoltia punctata)
was used as a supplement to enhance biomethane yields obtained from the anaerobic digestion of
dairy manure (Triscari et al., 2009). Various concentrations of duckweed addition were tested in
batch reactors operated under mesophilic conditions (35 oC) for 20-40 days. The outcomes of the
study revealed that a blend of 2% duckweed on a dry mass basis increased the methane and total
gas production of dairy manure slurries.
In addition to the biological conversion processes described above, several researchers
have investigated the thermochemical conversion of duckweed into various products. Xiu et al.
(2010) treated duckweed (Lemna minor) for oil production through liquefaction. The highest oil
yield, with a heating value of 34 MJ/kg, was achieved at a reaction temperature of 340oC with a
retention time of 60 min. In another study, the production, characterization and catalytic
application of bio-char obtained from the pyrolysis of duckweed (Lemna minor) was reported by
Muradov et al. (2012). In their study, duckweed biomass was successfully converted into bio-
char, and the treatment of bio-char with CO2 at 800oC increased its surface area. Baliban et al.
(2013) performed gasification of duckweed for the production of gasoline, diesel, and kerosene.
In their study, production and conversion scenarios were set and optimization of a hypothetical
biorefinery was performed, in order to determine the cost of dry duckweed below which the
duckweed refineries can not compete with the price of crude oil. The compatible price of
duckweed was estimated as $30/metric ton of dry biomass, to be comparable to crude oil prices
above $95/barrel.
16
The potential for integrating duckweed production with wastewater treatment in ecological
systems
As a low-cost option for wastewater treatment, it has been argued in the World Bank
Technical Report on Duckweed Aquaculture (Skillicorn et al., 1993) that duckweed wastewater
treatment systems are ideal alternatives to conventional treatment plants in developing countries.
The integration of duckweed into low-cost ecological wastewater treatment systems will not only
contribute to improvements in water quality by increasing the potential for nutrient removal, but
will also provide a sustainable and reliable production system of duckweed biomass for
conversion into biofuels and other valuable bioproducts.
Ecological engineering is the integration of ecological principles into engineering design
for the benefit of both human societies and the natural environment. The primary goals of
ecological engineering are to: (1) restore ecosystems that have been significantly disturbed by
human activities; and (2) deploy new sustainable ecosystems, from which both society and the
natural environment benefit. In this respect, ecological wastewater treatment can be defined as the
remediation of wastewater by integration of the process into an engineered ecosystem. As
opposed to conventional wastewater treatment technologies developed by environmental
engineering principles, ecological wastewater treatment systems involve less energy and human
manipulation to control the treatment process, as the main driving force is provided by the natural
activities in the ecosystem (Mitsch and Jørgensen, 2003). Constructed wetlands are typical
examples of ecologically engineered treatment systems, as they utilize the natural biological,
physical, and chemical processes of land and aquatic plants to purify wastewater (Shao et al.,
2013).
Due to management simplicity and low operating costs compared to conventional
wastewater treatment systems, ecological wastewater treatment systems are economically
affordable and environmentally sustainable alternatives in rural areas and as decentralized
17
wastewater treatment systems (Yoon et al., 2008; Z. Xu et al., 2010). One of the most promising
designs for ecological wastewater treatment is the Eco-MachineTM, which was first developed by
John Todd in the 1980’s. The Eco-Machine™ technology first originated from a basic
aquaculture system aligned in series, which later evolved into a wastewater treatment technology
that utilized the biological processes occurring in such aquaculture systems to effectively treat
wastewater. This mechanically simple and biologically complex process utilizes sunlight and
biodiversity for realization of biological, physical, and chemical processes such as sedimentation,
plant uptake, and microbial degradation for the treatment of wastewater (Todd and Josephson,
1996).
The Penn State Eco-MachineTM
The Pennsylvania State University has one of the few pilot-scale ecological wastewater
treatment systems (Eco-Machine™) in the United States, operated for research purposes. Due to
climatic conditions in Pennsylvania, the treatment process takes place in an enclosed greenhouse.
The Eco-Machine™ system at Penn State is composed of a series of open and closed tanks, and a
constructed wetland, hosting a variety of life forms including terrestrial and aquatic plants,
microorganisms, and macroinvertebrates in the treatment process (Figure 1-1). The sequence of
treatment stages are as follows: (1) anaerobic holding tank; (2) two closed anoxic tanks; (3) three
open aeration tanks; (4) a clarifier; and (5) a constructed wetland. The self-sustaining system
takes advantage of the natural biological reactions performed by plants, animals, and
microorganisms without requiring the addition of chemicals.
18
Anaerobic bioprocesses
Methanogenic anaerobic digestion
Methanogenic anaerobic digestion (MAD) is a process in which microorganisms
decompose organic matter (substrate) in the absence of molecular oxygen (O2), resulting in the
production of methane (CH4), carbon dioxide (CO2), and inorganic nutrients (McCarty, 1964).
Anaerobic treatment of wastes typically results in the conversion of organic matter into biogas,
which consists of approximately 20-30% CO2, 60-79% CH4, 1-2% hydrogen sulfide (H2S), and
other gases (Parkin and Owen, 1986; Sperling et al., 2007; Themelis, 2002; Yilmazel and
Demirer, 2011). In these systems, up to 90 % of the organic portion of waste can be converted
into methane (McCarty, 1964).
Bioconversion of wastes and biomass through MAD technology has been practiced as a
sustainable and renewable energy production method for several decades (Chynoweth et al.,
1993; Angelidaki et al., 2009). MAD of organic feedstocks is advantageous, particularly
because: (1) bioconversion can be achieved under non-sterile conditions with high quality end
products; (2) there is no strict requirement of intensive feed pre-processing (i.e. drying or
Figure 1-1: A schematic of the Penn State Eco-MachineTM.
19
pretreatment) (Chynoweth et al., 1993a); (3) it is suitable for treatment of biomass with high
moisture content; (4) it is a simple, naturally occurring process; (5) the technology is robust and
applicable on smaller scales; and (6) it is suitable for a variety of wastes and organic materials
(Appels et al., 2011).
MAD consists of four consecutive steps, namely: (1) hydrolysis (liquefaction); (2)
acidogenesis; (3) acetogenesis; and (4) methanogenesis. Each stage involves microbial flora
adapted to anaerobic environments. In the first stage, bacteria excrete hydrolytic enzymes that
break down complex organics into simpler forms, such as sugars, long chain fatty acids (LCFA),
and amino acids. This step is rate limiting for substrates with high solid contents. After the
organic matter is solubilized in the first stage, fermentative acidogenic bacteria in the second
stage provide conversion of hydrolyzed waste into acetic, propionic, butyric, and other short
chain volatile fatty acids (VFAs), as well as alcohols. In the third stage, fermentative acetogenic
bacteria then convert the VFAs synthesized in the previous phase into H2, acetate, and CO2.
Elevated hydrogen concentrations cause inhibition of methane formation and increase organic
acid concentrations (Parkin and Owen, 1986). In the fourth and final stage of MAD, methanogens
simultaneously produce biogas from the end product of the previous stage. Methanogens are strict
anaerobes and are sensitive to environmental conditions (McCarty, 1964; Speece, 2008);
therefore, in order to achieve a robust MAD process, operating conditions should be controlled to
favor their growth and maintenance. Figure 1-2 illustrates the sequential stages of MAD
(McCarty, 1964).
20
Biochemical methane potential (BMP) assays
As a result of the growing popularity of application of MAD technology, determining the
ultimate biogas potential has emerged as a requirement for a variety of solid substrates
(Angelidaki et al., 2009). Biochemical Methane Potential (BMP) assays have been developed for
the determination of the ultimate convertibility of an organic substrate into CH4 (Chynoweth et
al., 1993a), which is a key parameter for assessing the technical and economic feasibility of full-
scale applications of MAD (Owen et al., 1979; Angelidaki et al., 2009).
BMP tests are conducted in batch mode under anaerobic conditions, after inoculation of
the substrate with fresh seed (i.e., anaerobic cultures obtained from operating anaerobic
digesters). The CH4 yield associated with the substrate added is determined as the difference
between CH4 yields of test reactors and seed control reactors without substrate. In order to
determine the ultimate sample biodegradability, any limitation of microbial activity should be
avoided by selection of optimum environmental conditions and operating parameters.
Figure 1-2: Sequential stages of the MAD process (Modified from: McCarty, 1964).
COMPLEX ORGANICS
- CARBOHYDRATES
- PROTEINS
- LIPIDS
SIMPLER SOLUBLE ORGANICS
SHORT – MEDIUM CHAIN
VOLATILE FATTY ACIDS
- PROPOINATE
- BUTYRATE ETC.
H2 + CO2 ACETATE
CH4 , CO2
21
Important environmental conditions and operating parameters in BMP assays
Anaerobic degradation efficiency is directly related to balanced microbial activity during
the MAD process. Operating parameters which must be controlled to achieve optimum growth of
microbial flora in BMP assays, including pH, temperature, growth medium, and substrate-to-
inoculum ratio, as discussed below.
pH
The MAD process involves a variety of microorganisms, the majority of which are
inhibited by acidic conditions, and the growth of methanogens in particular is strictly dependent
upon pH. The optimum pH range for the overall MAD process is 6.6 – 7.6 (McCarty, 1964).
Unless the system is well buffered, high amounts of organic acid produced by hydrolyzers and
acidogenic microbes create a tendency towards pH levels lower than 6, which is inhibitory for
methanogenic activity. In the MAD process, bicarbonate is the predominant alkalinity species
that creates buffering capacity, which suppresses pH drop. On the other hand, excessive presence
of alkalinity may also damage CH4 production by favoring ammonia-N toxicity (Parkin and
Owen, 1986). Alkalinity concentrations between 2000 and 4000 mg/L are typically sufficient to
sustain neutral pH (Tchobanoglous et al., 2003).
Temperature
Three temperature ranges are used for anaerobic digesters, namely, psychrophilic (0-20 ͦ
C), mesophilic (30-38 ͦ C), and thermophilic (50-60 ͦ C) (McCarty, 1964). Most conventional
digesters are operated in the mesophilic range (Parkin and Owen, 1986).
Medium
Inorganic nutrients are essential for growth and maintenance of both aerobic and
anaerobic microorganisms. Major nutrients that must be supplied in sufficient amounts are
nitrogen (N) and phosphorous (P). N is required for protein and amino acid synthesis, whereas P
is necessary for the synthesis of nucleic acids such as DNA and RNA, as well as energy structures
22
such as ATP. Hence, nutrients, especially N and P, must be provided sufficiently for a balanced
MAD (Speece, 2008). The optimum C/N ratio for MAD is 20/1 to 30/1 (Yen and Brune, 2007)
and the optimum N/P ratio is 5/1 to 7/1 (Parkin and Owen, 1986). Other than N and P, nutrients
such as iron, nickel, cobalt, sulfur, calcium, and some trace organics are required in lower
amounts. Compared to aerobic systems, anaerobic systems usually require a lower nutrient
supply; however, in some cases, an external source may be necessary (Angelidaki et al., 2009).
Toxicity
The toxicity of a substance depends on its nature, concentration, and the acclimation of
the system. Generally, many substances are tolerable at low concentrations but become inhibitory
as their concentrations increase. Alkali and alkaline earth-metals, heavy metals, ammonia-
nitrogen, sulfide, and some other inorganic and organic compounds such as sodium, potassium,
calcium, magnesium, copper, chromium, nickel, formaldehyde, chloroform, ethyl benzene,
ethylene dichloride, kerosene, and detergents are toxic to anaerobic digestion. Microorganisms
can improve their resistance to toxic compounds through acclimation (Parkin and Owen, 1986).
Substrate - to – Inoculum Ratio
In a BMP assay, any potential substrate toxicity must be avoided, in order to determine
the maximum extent of biomass conversion. Moreover, the inoculum concentration should not be
limiting (Owen et al., 1979). Therefore, the substrate-to-inoculum ratio (S/I) is one of the key
parameters that influences determination of anaerobic digestibility and biomethane production
potential (Chynoweth et al., 1993a). S/I does not only affect total CH4 yield, but also its
production rate (Eskicioglu and Ghorbani, 2011). Determination of the optimum S/I could
provide information on start-up protocols for continuous bioreactors (Alzate et al., 2012). The S/I
on a standard BMP procedure is 1.0 on a volatile solids basis. In general, lower S/I values provide
slightly more rapid bioconversion of substrates rich in cellulose and low in soluble sugars into
23
CH4, and an S/I of 0.5 typically yields the maximum conversion for such substrates (Chynoweth
et al., 1993a).
Pretreatment methods for enhanced anaerobic digestion
Hydrolysis of particulate organic matter is usually the rate-limiting step in anaerobic
digestion. The aim of pretreatment is to facilitate the hydrolysis of wastes with high solids
contents. Pretreatment methods include, but are not limited to, the single or combined application
of: (1) physical pretreatment, such as milling and grinding; (2) physicochemical pretreatment,
such as steam, thermal, hydrothermolysis, and wet oxidation; (3) chemical pretreatment, such as
alkali, acid, or oxidizing agents; and, (4) biological and electrical pretreatment.
Each pretreatment method has various benefits and shortcomings in terms of efficiency,
energy intensity, and impact on the environment. Thermal pretreatment has been proven to be
effective for many feedstocks such as corn stover, municipal organic wastes, and other complex
materials (Liu, 2010). Moreover, thermal pretreatment at low temperatures (< 100 oC) has been
stated to be the most effective method in terms of efficiency, economic cost, and environmental
impact. During low temperature thermal pretreatment, complex molecules are not broken down
into simpler forms; however, disintegration of particulate matter is achieved. As a result, up to
78% higher biogas production has been reported for various organic wastes when thermally
pretreated at low temperatures (Ariunbaatare et al., 2014).
Acidogenic anaerobic digestion (AAD)
Acidogenic Anaerobic Digestion (AAD), is the decomposition and fermentation of
organic material into organic acids, such as carboxylic acids or volatile fatty acids (VFAs), as
well as CO2 and H2, by complex acidogenic microbial consortia in the absence of oxygen. AAD
is an intermediate stage in MAD, as methanogenic microorganisms require VFAs as substrates.
When these two stages are uncoupled by inhibition of methanogenic activity, a mixture of short
24
and medium chain fatty acids can be obtained. The short-chain VFAs (acetate, propionate, lactate,
and n-butyrate) are the main end products of undefined mixed cultures through AAD. These
fermentation products can then be separated and utilized for the production of various
biomaterials and biofuels, including olefins, alkanes, alcohols, and esters (Datta, 1981; Agler et
al., 2011). Figure 1-3 illustrates the AAD process, uncoupled from complete MAD.
Acidogenic digestion is advantageous over alcohol fermentation due to: (1) increased
direct utilization potential of cellulose, without pretreatment or enzyme addition due to effective
cellulolytic hydrolyzers in the acidogenic microbiome; (2) production of a single class of end-
products from multiple precursor molecules including cellulose, hemicellulose, starch, sugars,
protein and lipids; (3) the absence of sterilization requirements; and (4) convertibility of end
products into higher-value chemicals and fuels. However, these systems also have some
drawbacks, such as: (1) requirement of process control to avoid a shift into methanogenic
Figure 1-3: Sequential stages of the acidogenic anaerobic digestion process (Modified from:
McCarty, 1964).
COMPLEX ORGANICS
- CARBOHYDRATES
- PROTEINS
- LIPIDS
SIMPLER SOLUBLE ORGANICS
SHORT – MEDIUM CHAIN
VOLATILE FATTY ACIDS
- PROPOINATE
- BUTYRATE ETC.
H2 + CO2 ACETATE
CH4 , CO2
25
activity; (2) slower conversion rate compared to alcohol fermentation; and (3) requirement of
more intensive downstream processing (Datta, 1981). AAD has been considered as the
“carboxylate platform” for conversion of lignocellulosic feedstocks such as energy crops,
agricultural residues, and organic wastes into short and medium chain VFAs by utilization of
non-sterile mixed cultures under anaerobic conditions (Agler et al., 2011).
Important Environmental Conditions and Operating Parameters for AAD
The bioconvertibility of organic materials using AAD depends on the nature of the
inoculum and the substrate, as well as several operating parameters, including the solid loading
rate. AAD can be negatively influenced by product inhibition. The key condition for the
fermentation of VFAs from organic biomass with high yields is prevention of the process from
shifting towards methanogenesis, which can be achieved by adjustment of environmental
conditions such as temperature and pH. The environmental conditions and operating parameters
for a balanced AAD process are discussed below.
pH
The pH of the AAD process can be set below 5.5 to prevent methanogenesis. However,
recent studies show that VFA production at pH 10 is also feasible, with prevention of methane
generation (G.-H. Yu et al., 2008).
Temperature
It was reported that methanogenic activity can be inhibited at temperatures lower than
mesophilic region. However, it was also reported that the highest VFA production yields can be
achieved under thermophilic conditions (55oC) (Shin et al., 2004).
Toxicity
Acidogenic microorganisms are relatively more tolerant of low pH values compared to
methanogens. However, high concentrations of VFAs are reported to be self-inhibitory for acid
forming bacteria. Therefore, continuous separation and removal of produced VFAs is required to
26
improve bioconversion performance (Ghosh et al., 1975; Datta, 1981; Herrero, 1983; Xiong,
2014).
Solid Loading
Optimal solid loading in the AAD process is essential for achieving maximum VFA
yield. Low solid loading rates may lead to larger reactor volumes and higher quantities of
wastewater generation. However, above a certain solid loading concentration (i.e., 70 g/L), VFA
yields may drop due to product self-inhibition or relatively poor hydrolysis (Xiong, 2014).
Inoculum Characteristics
Since the AAD process involves complex microbial consortia, the selection of inocula
has a direct effect on the production and yield of VFAs. Most commonly, rumen fluid, silage,
wastewater sludge, soil, marine sediments, swamp material, and other sources of anaerobic
biomass have been used as inocula (Thanakoses et al., 2003; Aiello-mazzarri et al., 2006; Yue et
al., 2007, 2008; Chen et al., 2008; Xiong, 2014).
Ethanol fermentation
Ethanol has various uses as a raw material, solvent, and fuel, and is utilized in large
quantities in the chemical, pharmaceutical, and food industries. Equation 1-1 shows the
conversion reaction of hexoses to ethanol and carbon dioxide through glycolysis by yeast.
C6H12O6 2C2H5OH + 2 CO2 (Equation 1-1)
The theoretical yield of this process is 0.51 g ethanol / g glucose. However, the actual
maximum ethanol yield over glucose is approximately 90–95% of the theoretical yield, due to the
accumulation of byproducts that inhibit the fermentation process, including glycerol, succinic
acid, and acetic acid. The optimum temperature for ethanol fermentation is 30o- 35 oC for
mesophilic and 50° - 60°C for thermophilic organisms. The optimum pH range is 4-6. Ethanol
fermentation takes place under anaerobic conditions, although trace amounts of oxygen (0.05–0.1
mm Hg) are necessary for lipid biosynthesis and maintenance of cellular processes (Concepts,
27
n.d.). For industrial-scale production of ethanol, yeast species, in particular, S. cerevisiae, has
been widely used for the bioconversion of hexoses into ethanol. However, different substrates
may require different species for successful bioconversion (Badger, 2002).
Biomass as a Feedstock for Ethanol Fermentation
A variety of agricultural, forest, or municipal waste products with considerable quantities
of sugar or materials that can be converted to sugars, such as starch and cellulose, can be utilized
as feedstock for ethanol production. The feedstocks which can be converted into ethanol through
fermentation can be classified in three main groups: (1) sugary materials; (2) starchy materials;
and (3) lignocellulosic materials (Shapouri & Salassi, 2006; Naik, Goud, Rout, & Dalai, 2010).
Examples of each class are provided in Figure 1-4.
Figure 1-4: Types of feedstock for ethanol production with example crops (Source: Khan &
Dwivedi, 2013).
28
Bioconversion of sugary materials into ethanol is a simple process. Yeast can directly
ferment hexoses to ethanol and carbon dioxide, when necessary physical conditions are provided
(Khan et al., 2013). Starchy biomass cannot be directly fermented into ethanol. However, as
starch is a polymer of glucose, its breakdown into glucose is a relatively simple process which
involves two consecutive enzymatic processes. The first step, called liquefaction, is the
hydrolysis of starch into short chains, i.e., dextrin and oligosaccharides, by the aid of amylase
enzyme. In the second step, called saccharification, the produced short chain compounds are
further hydrolyzed into glucose, maltose, and isomatose. After saccharification, the slurry can be
subjected to fermentation by yeast addition and adjustment of required conditions. The ethanol
produced from starchy biomass is currently being produced as a first generation biofuel, mostly
derived from corn. However, first generation biofuels raise ethical concerns related to trade-offs
between food and energy supplies, potentially causing an increase in food prices (Naik et al.,
2010).
Lignocellulosic biomass is composed of three main constituents, namely hemicellulose,
lignin, and cellulose. Cellulose consists of polymers of fermentable D-glucose molecules;
however, their availability for enzymatic hydrolysis by cellulase enzymes is very limited until the
lignin matrix has been degraded, by application of one or more pretreatment methods (Kumar et
al., 2009). Therefore, production of ethanol from lignocellulosic biomass involves an additional
initial step of pretreatment, followed by enzymatic hydrolysis of cellulose and fermentation of
hexoses (Bondesson, 2008). Ethanol derived from lignocellulosic biomass is usually regarded as a
second generation biofuel. Since the feedstock is usually an organic waste rich in cellulose, or
non-edible fraction of agricultural products, second generation ethanol does not cause ethical
concerns as compared to first generation biofuels. However, costs associated with the
pretreatment of the lignocellulosic biomass for the separation of lignin has been considered as the
major drawback of this process (Naik et al., 2010; Khan & Dwivedi, 2013).
29
Downstream wastewater management alternatives originating from second generation
bioethanol production processes have not been well documented in the literature. However, the
environmental impacts, associated costs, and energy requirements of their mitigation must be well
mapped out prior to full commercialization of large scale bioethanol plants, in order to ensure
their sustainability (Bondesson, 2008).
Sequential Ethanol Fermentation and Methanogenic Anaerobic Digestion
As discussed in the previous section, utilization of biomass as a resource for bioethanol
production has not been proven to be sustainable in terms of social, economic, and environmental
aspects. Therefore, researchers have been focusing on increasing the overall energy efficiency
and lowering the environmental impacts of second generation biofuels, which do not pose ethical
debates around the food-energy nexus and which can be obtained from crops grown on marginal
land. For this reason, the anaerobic digestion of the waste streams associated with second
generation ethanol production has gained popularity in the last decade.
It was argued by Bondesson (2008), that considerable quantities of biomethane can be
obtained from MAD of ethanol fermentation residues of wheat straw, depending on the process
configuration. The configuration yielding the highest net bioenergy yields in terms of bioethanol
and biomethane produced is given in Figure 1-5. By using this configuration, 83% of the
theoretical ethanol fermentation yield was achieved from silage, as well as 754 ml CH4/g VSadded.
The net energy gain in the form of ethanol and methane was reported as 60%, when the energy
input of the process was taken into account.
30
Another sequential ethanol fermentation and MAD study was conducted by Dererie et al.,
(2011) to evaluate the effect of the coupled process on the net energy gain obtained from oat
straw. Thermochemically pretreated oat straw recovered 85-87% higher heating value from the
biomass in the coupled process, which is 28–34% higher than direct biogas digestion. Rabelo et
al. (2011) reported similar results by integration of MAD into pretreatment and bioethanol
production from bagasse residues. They recovered 63–65% of the energy by combining ethanol
production with the combustion of lignin and hydrolysis residues, along with the MAD of
pretreatment liquors. This combined process yielded 72.1 L methane/kg bagasse, which was
twice of that which could be obtained by sole ethanol production. In another study, a three stage
bioconversion process of food waste (FW), which involved saccharification, fermentation of
saccharified liquid, and MAD of saccharification residue, was developed (Wu et al., 2015). As a
result, 61.7% sugar recovery, 0.9 g / L.h ethanol productivity, and 252.6 mL/g VSadded of methane
yield was achieved. It was concluded that the three stage process increased the FW
decomposition rate by 27.5 %, decreased the energy requirement of the process by 51.8%, and
increased the total energy yield by 17.6%.
Figure 1-5: Schematic flow diagram of a simultaneous ethanol fermentation and MAD process.
31
Life cycle assessment
The negative impact of industrial activities on environmental quality have increased the
awareness of the unsustainable nature of manufacturing and consumption. Therefore, the demand
for “green” products, and the interest in the development of methods to assess the environmental
performance of products and processes have increased. One of these methods is life cycle
assessment (LCA).
LCA is a tool that is used to evaluate the potential environmental impacts of a product,
process, or service throughout its life cycle, from “cradle to grave”. LCA provides both a holistic
picture of a product's environmental impacts and comparisons between stages of the product’s life
(Dong and Adams, 2012). The most up-to-date International Standards set forth for the
implementation of LCA are ISO 14040:2006 and 14044:2006, each providing guidelines and a
framework for a high-quality assessment. The latter, ISO14044, also provides several
requirements and recommendations to increase the comparability of different LCAs, although the
comparison is only possible for studies with equivalent assumptions and contexts (ISO 14044).
LCA studies consist of four iterative phases: 1) goal and scope definition; 2) life cycle
inventory (LCI) analysis; 3) life cycle impact analysis; and 4) interpretation. The goal and scope
definition phase involves description of the product, process, or activity of interest, identification
of the functional unit and system boundaries including sub-units, inputs, and outputs. In the
second (LCI) phase, information related to inputs and outputs such as energy, water, materials
flow, and environmental emissions are studied by a collection of necessary data. During the third
phase, life cycle impact analysis, the potential impacts of the system on the environment are
assessed, based on the data gathered during LCI process. In the final, interpretation phase, the
results of the previous phases are used for deriving conclusions and decision making in the
32
context of the goal and the scope defined previously (Rios et al., 2007). Figure 1-6 summarizes
the phases of an iterative LCA procedure.
State-of-the-art for Duckweed Bioconversion through Anaerobic Bioprocesses
Current duckweed-to-bioenergy literature is quite narrow, focusing mostly on ethanol
production. Research on methanogenic anaerobic digestion of duckweed is especially scarce. So
far, the studies on this topic are limited to a few trials with no focus on process improvement.
Furthermore, combined anaerobic digestion and ethanol fermentation has gained popularity
recently, in order to improve both environmental and economic performance of bioethanol
production processes, especially from lignocellulosic biomass (Bondesson et al., 2013;
Bondesson, 2008; Dererie et al., 2011; Wu et al., 2015). However, conventional ethanol
Figure 1-6: Phases of an LCA (Source: ISO, 2006).
33
production processes are reported to have a risk of nutrient (especially nitrogen but possibly
phosphorus) deficiency, limiting successful anaerobic bioconversion of these feedstocks.
Considering the high nutrient content of duckweed, investigation of a combined ethanol and
biomethane generation system is another important research area to be addressed.
In contrast to ethanol production from lignocellulosic feedstock through the sugar
platform, the carboxylic acid platform represents an alternative bioconversion pathway. The
production of carboxylic acids from organic wastes and lignocellulosic biomass has received
attention recently as an intermediate for the production of biofuels and other chemicals. The
carboxylic acid platform can be an alternative anaerobic bioconversion process for duckweed as
well. To the best of our knowledge, no previous investigation has tested the potential for
carboxylic acid production from acidogenic anaerobic digestion of duckweed. Therefore,
determination of the carboxylic acid production potential from duckweed and optimization of
operating parameters is necessary.
To frame out a complete biorefinery approach to deliver a competitive product to the end
user markets, a robust, reliable, and sustainable biofuel supply chain is essential. For this reason,
a variety of work has been conducted on biofuel supply chain networks, consisting of the raw
material (biomass) production processes, biorefineries, storage facilities, blending stations, and
end users (Awudu and Zhang, 2012). As opposed to supply chains of industrial goods which
consider consumer demand, biorefineries are restricted by the supply, and therefore require
different modeling strategies. The applicability of duckweed-based bioenergy technologies must
therefore be analyzed by optimizing the network as a whole. A holistic approach would enable
evaluation of the economic feasibility of the biomass supply when its production is coupled with
wastewater treatment, to ensure efficient and effective delivery of the end products to blending
facilities.
34
Duckweed-to-bioenergy research requires further study to address not only the technical
limitations of converting duckweed into various end products through individual or coupled
processes, but also the sustainability of optimizing cultivation and bioconversion processes.
Coupling wastewater treatment and feedstock production addresses ethical issues related to
agricultural resource allocation for fuel production. Moreover, the integrated systems not only
reduce the risk of food insecurity, but also may be the only option for sustainable biofuel
production from aquatic biomass, such as microalgae. Similarly, it has been shown by several
studies that life cycle impacts of microalgal biofuels are dominated by the cultivation phase, if
wastewater is not used (Clarens et al., 2010). In addition, Murphy & Allen (2011) have discussed
that an uncoupled microalgal biodiesel system requires seven times higher energy for wastewater
management than is produced from the biodiesel. Therefore, wastewater treatment systems must
be considered as downstream units of anaerobic bioprocesses. Parallel conclusions are essential
for duckweed-based biofuels in order to evaluate the feasibility of the process and its
commercialization potential. A logical step would be to perform a life cycle assessment of an
integrated wastewater treatment, duckweed production, and the biorefinery supply chain, in order
to ensure sustainability of the system by comparison with conventional wastewater treatment
processes and petroleum refineries.
In this study, the knowledge gaps preventing the establishment of integrated wastewater-
derived duckweed biorefineries were addressed. Technical, environmental, and economic
evaluations were performed to enable the simultaneous utilization of duckweed as a reliable
contaminant bioremediation tool and as a feedstock for bioenergy generation. The following five
phases of work were completed to meet this goal:
1. Coupled ethanol fermentation and methane production from duckweed was evaluated in terms
of the net bioenergy yield difference, compared to individual processes.
35
2. Duckweed was evaluated as a feedstock in the carboxylic acid platform by optimizing the
operating conditions of the anaerobic acidogenic digestion process.
3. The potential of producing an array of products in a value cascade was evaluated in a
biorefinery system targeting ethanol, carboxylates, methane, and fertilizer from wastewater-
derived duckweed biomass through anaerobic bioprocesses.
4. A general supply chain framework was designed for duckweed biorefineries, under a large-
scale production scenario. The supply chain was established to determine cost, environmental
impact, and efficiency both upstream and during operations of the hypothetical biorefinery, in
order to provide: the best spatial and temporal options for cultivation, harvesting, and transport
of duckweed; and bioconversion of duckweed into the most feasible end product. Data from
previous sections of this work were used to develop this supply chain framework
5. Life cycle assessment (LCA) of an integrated wastewater treatment and bioenergy production
process using duckweed was performed using outcomes from the abovementioned work.
The overall goal of this study was to develop a sustainable and reliable energy supply
network, without posing a risk to food or water security, and while decreasing dependence on
petroleum-based fuels and chemicals. The results should provide insight into the feasibility of
such a system with the minimum negative impact at the food–energy–water nexus.
36
Chapter 2
Proof of Concept - Sequential Ethanol Fermentation and Anaerobic
Digestion Increases Bioenergy Yields from Duckweed
This chapter has been published as follows:
Calicioglu, O., Brennan, R.A., 2018. Sequential ethanol fermentation and anaerobic
digestion increases bioenergy yields from duckweed. Bioresour. Technol. 257, 344–348.
doi:10.1016/j.biortech.2018.02.053.
Abstract
The potential for improving bioenergy yields from duckweed, a fast-growing, simple,
floating aquatic plant, was evaluated by subjecting the dried biomass directly to anaerobic
digestion, or sequentially to ethanol fermentation and then anaerobic digestion, after evaporating
ethanol from the fermentation broth. Bioethanol yields of 0.41 ± 0.03 g/g and 0.50 ± 0.01 g/g
(glucose) were achieved for duckweed harvested from the Penn State Living-Filter (Lemna
obscura) and Eco-MachineTM (Lemna minor/japonica and Wolffia columbiana), respectively. The
highest biomethane yield, 390 ± 0.1 ml CH4/g volatile solids added was achieved in a reactor
containing fermented duckweed from the Living-Filter at a substrate-to-inoculum (S/I) ratio (i.e.,
duckweed to microorganism ratio) of 1.0. This value was 51.2 % higher than the biomethane
yield of a replicate reactor with raw (non-fermented) duckweed. The combined bioethanol-
biomethane process yielded 70.4 % more bioenergy from duckweed, than if anaerobic digestion
had been run alone.
37
Introduction
The economic and environmental disadvantages of fossil fuel consumption have
increased the search for alternative resources to fulfill world’s growing energy and chemical
needs (Jung et al., 2016). At the same time, conventional bioenergy crops have also been posing
social, economic, and environmental challenges. Duckweed (Lemnaceae), a family of fast-
growing, simple, floating aquatic plants, consisting of 38 species in five genera (Les et al., 2002),
has been demonstrated to be a technically feasible alternative feedstock for bioethanol production
due to several advantages: it can accumulate high amounts of starch (up to 46% of dry mass)
under nutrient starvation (Zhao et al., 2015); has relatively little lignin content (1%-3%); its
small size (0.1 cm to 1 cm) eliminates the need for milling; and, because it floats, the harvesting
process is relatively simple (Cui and Cheng, 2015). Duckweeds are resilient to a broad range of
nutrient concentrations; therefore, they can be grown on wastewater steams (Cheng and Stomp,
2009).
Due to its high and manipulatable starch content, duckweed is regarded as a promising
bioethanol feedstock in the current literature. The studies conducted to date have focused on the
utilization of the starch component only (Xu et al., 2011; Yu et al., 2014), or the fermentation of
cell wall carbohydrates as well (Ge et al., 2012; X. Zhao et al., 2014). The high level of
variability in wastewater compositions, however, may cause uncertainties in starch and
bioethanol potentials from wastewater-derived duckweed biomass. By comparison, a more
resilient pathway for duckweed valorization could be anaerobic digestion, since this process
converts not only sugars, but also proteins and lipids into biomethane. In addition, anaerobic
digestion can be used to stabilize residual organics in the ethanol fermentation broth, and thereby
help to compensate for the costs of ethanol production and distillation (Wu et al., 2015). Indeed,
the sequential process of ethanol fermentation and anaerobic digestion has been shown to
38
increase the overall bioenergy yield of several other substrates such as food waste (Wu et al.,
2015), oat straw (Dererie et al., 2011), and corn stalks (Vintilǎ et al., 2013). This combined
approach may improve the sustainability of large-scale biorefineries.
Although some work has focused on ethanol production from duckweed, reports on its
anaerobic digestibility are limited to a very few studies. An early study on anaerobic digestion of
manganese-contaminated duckweed produced a maximum biogas yield of 176 ml/g with a
methane content of 60% (Jain et al., 1992). Other work conducted on duckweed has focused on
its co-digestion with other substrates, such as dairy manure (Triscari et al., 2009), to help balance
the C/N ratio.
To ensure that neither limitations nor inhibition will occur during anaerobic digestion due
to substrate loading, the substrate-to-inoculum ratio (S/I) should be optimized (Chynoweth et al.,
1993a). The S/I not only affects total methane yield, but also its production rate (Alzate et al.,
2012). In the current study, the potential of increasing bioenergy yields obtained from duckweed
grown in an ecological wastewater treatment system for nutrient removal was investigated using a
sequential process: fermentation of duckweed and distillation of the resulting bioethanol,
followed by anaerobic digestion of the residual fermented duckweed. In addition, the effects of
S/I ratio on anaerobic digestion performance were evaluated through biochemical methane
potential (BMP) assays.
39
Materials and Methods
Analytical methods
Total solids (TS), volatile solids (VS), total suspended solids (TSS), volatile suspended
solids (VSS), and volatile dissolved solids (VDS) were determined according to Standard
Methods No. 2540 (APHA/AWWA/WEF, 2012). The suspended portion of samples was
separated on glass fiber filters (AP40; Millipore, Billerica, MA, USA) using a vacuum filtration
apparatus. Chemical Oxygen Demand (COD) was measured according to the closed reflux
colorimetric method as described in Standard Methods, No. 5220 (APHA/AWWA/WEF, 2012).
Glucose and ethanol quantification were performed using a Waters high performance
liquid chromatograph (HPLC) equipped with a refractive index detector (Waters, Milford, MA)
and a Bio-Rad Aminex HPX-87H column (300 mm × 7.8 mm; Bio-Rad, Richmond, CA) with 0.8
ml/min of 0.012 N sulfuric acid as the mobile phase. The detector and column temperatures were
constant at 35 °C and 65 °C, respectively. Prior to HPLC analysis, samples were centrifuged at
4°C for 20 min at 5,200 x g and the supernatant filtered through 0.2 μm nylon syringe filters.
Theoretical maximum glucose concentration was calculated according to Gulati et al. (1996).
Headspace gas volumes of anaerobic reactors were measured at 25 °C using a water
displacement device filled with 0.01 M hydrochloric acid to prevent microbial growth. Volume
readings were reported at standard temperature and pressure. Volumetric methane concentrations
were determined by withdrawing headspace from the reactors using a 250 μL airtight syringe
(Hamilton, Reno, NV, USA) and injecting into a gas chromatograph (model SRI310C, SRI
Instruments, Torrance, CA, USA) equipped with a 6 foot molecular sieve column (Altech,
5605PC, MD) held at 80 ◦C.
40
Plant material and cultivation
Duckweed used in this study was obtained on May 27, 2015, from two sources: 1) an
open tank dedicated for growing duckweed in the Penn State Eco-Machine™ (EM), which is a
pilot-scale ecological wastewater treatment system receiving on average (n = 4) 3.6 ± 1.1 mg/L
phosphate, 0.1 ± 0.0 mg/L ammonia, and 11.1± 3.0 mg/L nitrate; and 2) an open pond within the
effluent spray fields of the Penn State Wastewater Treatment Plant, a.k.a. the “Living-Filter”
(LF), receiving on average (n = 3) 2.2 ± 0.4 mg/L phosphate, 2.3 ± 0.9 mg/L ammonia, and 7.8 ±
0.8 mg/L nitrate. In both sources, duckweed was naturally present and had not been subjected to a
frequent harvesting regime.
To identify the duckweed species present in each source, total DNA was extracted from
duckweed tissue using a PowerPlant® Pro DNA isolation kit (QIAGEN, Hilden, Germany), and
then amplified using a two-barcode PCR protocol (Borisjuk et al., 2014). After amplification, the
DNA fragments were purified using a GeneJET PCR purification kit (ThermoFisher, Waltham,
MA), and sent to the Genomics Core Facility (The Pennsylvania State University) for
processing. Following a BLAST-based protocol for duckweed species identification (Borisjuk et
al., 2015), the EM duckweed was identified as a co-culture of Lemna japonica/minor (100%
sequence identity to accession numbers KJ9211760.1 and DQ400350.1, respectively, in the NCBI
database) and Wolffia columbiana (99.6 % sequence identity to accession number GU454371.1);
whereas the LF duckweed was identified as a monoculture of Lemna obscura (100% sequence
identity to accession number GU454331.1).
For use in these experiments, harvested duckweed was rinsed with tap water and dried at
50 ± 2 oC to a constant weight over two days. The composition of the dried duckweed was
determined by first grinding and sieving through mesh No. 20 (850 mm opening size), and then
sending to Dairy One Wet Chemistry Laboratory (Ithaca, NY). The composition of EM
41
duckweed was reported as 16.9 % cellulose, 23.9% hemicellulose, 4.3 % starch, 2.0 % lignin,
26.0 % crude protein, and 0.73 g VS per g TS. The composition of LF duckweed was reported as
17.0 % cellulose, 18.1 % hemicellulose, 15.9 % starch, 1.1 % lignin, 17.0 % crude protein, and
0.81 g VS per g TS.
Inocula
Yeast strain
For fermentation of duckweed, Saccharomyces cerevisiae (ATCC 24859) was enriched
in culture medium with the following constituents (concentrations in parentheses are g/L):
glucose (20); yeast extract (Difco, Sparks, MD) (6); CaCl2·2H2O (0.3); (NH4)2SO2 (4);
MgSO4·7H2O (1); and KH2PO4 (1.5). The culture was grown at 30 °C for 24 h before being
transferred to fermentation flasks as the inoculum.
Anaerobic Seed
Anaerobic seed was obtained from the Penn State Wastewater Treatment Plant secondary
anaerobic digester. The inoculum was starved for two days prior to use in the BMP assays. The
TS of the starved seed was 23.9 ± 0.5 g/L, and the VS was 15.7 ± 0.7 g/L, which is 65.8 ± 5.1 %
of the TS.
42
Fermentation experiments
Enzymatic saccharification of the duckweed was performed in 500 ml flasks with 200 ml
distilled water and 10 g duckweed (dry weight). The pH was adjusted to 7.0 ± 0.1 with 2 M
hydrochloric acid prior to liquefaction by autoclaving at 95 °C under 103 kPa for 1 h. Flasks with
EM and LF duckweed received 0.6 ml and 1.98 ml of α–amylase (Sigma Aldrich, A3403, USA)
respectively, based on the starch content of each duckweed type, to achieve an amylase loading of
5000 units/g starch. Following liquefaction, the pH was adjusted to 4.8 ± 0.1 with glacial acetic
acid. After pH adjustment, 60 mg and 198 mg glucoamylase (Sigma Aldrich 10115, USA) were
added to each flask containing EM and LF duckweed, respectively. In addition, all flasks received
2 ml cellulase (60 filter paper unit/g cellulose). Saccharification was then performed at 50 °C,
while mixing at 120 rpm for 24 h in flasks sealed with cotton stoppers and parafilm. All
experiments were conducted in triplicate under sterile conditions.
Following saccharification, the pH of each flask was increased to 7.0 ± 0.1 by dosing
with 2 M sodium hydroxide, and then 2 ml yeast culture was added. Flasks were incubated at 30
°C while mixing at 120 rpm for 48 h. Glucose and ethanol concentrations before and after
fermentation were quantified. Fermented ethanol was then evaporated by vacuum extraction after
the pH was increased to 7.8 ± 0.1 by 2 M sodium hydroxide addition, in order to avoid escape of
volatile fatty acids (VFAs) from the slurry. The triplicates for each duckweed type were then
combined and subjected to BMP assays.
Biochemical methane potential (BMP) assays
The BMP assays with duckweed were carried out based on the protocol proposed for
bioenergy crops and organic wastes (Angelidaki et al., 2009) with slight modifications. Batch
43
reactors (160 ml total volume, 120 ml working volume) were filled with 24 ml inoculum, and
substrate (either raw EM or LF duckweed, or residual fermentation slurries, FEM or FLF), to
provide an S/I of 0.5 or 1.0. To account for the effect of endogenous gas production by the
anaerobic inoculum, control bottles were prepared with the same amount of anaerobic seed, but
without substrate. Blank bottles were prepared with duckweed, but without inoculum addition.
To determine if the duckweed reactors were lacking in alkalinity or other nutrients for microbial
growth, the effect of basal medium addition (Vanderbilt Medium, VM) (Uludag-Demirer et al.,
2008) was also tested. After the initial pH was adjusted to 7.2 ± 0.3 by adding 2 M solutions of
hydrochloric acid and sodium hydroxide, the bottles were purged with a 80/20 (by volume)
mixture of N2/CO2 gas for 3 min prior to sealing with butyl rubber septa and aluminum crimp
tops. Reactors were incubated at 35 ± 0.5 °C for 45 days. Gas volumes and contents were
quantified periodically, until the weekly gas production was less than 5 % of the cumulative
value. Test and control reactors were run in triplicate, whereas blank reactors were run in
duplicate. Biogas volumes in control bottles were subtracted from those of tests before reporting.
Overall bioenergy yields
The overall bioenergy yields of ethanol fermentation, anaerobic digestion, and the two
processes coupled together were calculated for both duckweed sources, using lower heating
values of ethanol and methane of 29.7 MJ/kg and 35.8 MJ/kg, respectively (Wu et al., 2015). For
these calculations, the yields of ethanol (Table 2-1) and biomethane (Figure 2-1) were considered
on a TS basis. The energy input and output associated with enzyme, yeast, and pH adjustment
were assumed to be negligible.
44
Results and Discussion
Fermentation experiments
Ethanol fermentation potentials of duckweed obtained from the EM and LF were
quantified in terms of glucose recovery, glucose recovery efficiency, ethanol concentration in the
fermentation broth, fermentation efficiency, and ethanol yield (Table 2-1). The results revealed
that only 55.5 % of the glucose could be recovered from EM duckweed after enzymatic
saccharification. Since α-amylase was added in proportion with the starch content, EM duckweed
received lower quantities of the enzyme. Therefore, the poor glucose yield for EM duckweed can
be attributed to a slower rate of liquefaction due to lower α-amylase availability.
Despite relatively low glucose recoveries, the ethanol concentration observed in the EM
duckweed fermentation broth was 3.2 g/L, which corresponds to an ethanol yield of 0.50 g/g
glucose recovered. This relatively high conversion efficiency might be a result of ongoing
enzymatic activity, which may have increased glucose availability during the fermentation
process and consequently boosted its simultaneous conversion into ethanol. By comparison, the
glucose recovery for the LF duckweed was 17.9 g/L, corresponding to 97.6 % of the theoretical
value. This value is similar to the sugar recovery reported by Xu et al. (2011), as 96.8 % of the
theoretical glucose saccharification of S. polyrrhiza starch using the enzymes α-amylase,
pullulanase, and amyloglucosidase for hydrolysis. The ethanol concentration in the LF duckweed
fermentation broth after 48 h was 7.3 g/L, which corresponds to a yield value of 0.41 g ethanol/ g
glucose recovered. This result is slightly lower than the average value reported by Yu et al.
(2014) as 0.44 g/g (as glucose) for duckweed grown on Schenk & Hildebrandt medium and
sewage wastewater, following sugar recoveries of 94 %.
45
Biochemical methane potential (BMP) assays
Approximately 90 % of the total biogas production was observed in the first 20 days in
all reactors. The biogas production was proportional to the VS concentration of substrate
provided. The biomethane yields of the reactors varied between 141 to 390 ml CH4/g VSadded
(Figure 2-1A-D), which is comparable to that reported by Jain et al. (1992) as 176 ml CH4/g
VSadded. No methane production was observed in blank reactors (data not shown).
The raw EM duckweed yielded slightly lower biomethane (234 ml CH4/g VSadded),
compared to that of LF duckweed (260 ml CH4/g VSadded) at an S/I value of 0.5. However, for an
S/I of 1.0, EM and LF duckweed yielded similar biomethane (258 and 259 ml CH4/g VSadded
respectively). Compared to BMP assays conducted on other raw bioenergy crops, these values are
consistent with the literature. For instance, lignocellulosic feedstock such as straw, yielded a
methane potential between 180 and 320 ml CH4/g VS, whereas starch crops showed higher, yet
comparable, methane yields of 250 to 406 ml CH4/g VSadded for corn, and 310 to 430 ml CH4/g
VSadded for potatoes. In general, both raw and fermented EM duckweed reactors yielded less
biomethane than their LF duckweed counterparts. This could be explained by the lower readily
biodegradable (i.e., starch) content and higher recalcitrance (i.e., lignin) of EM duckweed.
46
Figure 2-1: Cumulative methane production (ml CH4/ g volatile solids added) in batch reactors
fed with raw Eco-MachineTM duckweed (EM), raw Living-Filter duckweed (LF), fermented Eco-
MachineTM duckweed (FEM), fermented Living-Filter duckweed (FLF) at different substrate-to-
inoculum (S/I) ratios and with and without the addition of Vanderbilt Medium (VM): A) S/I =
0.5, without VM; B) S/I = 0.5, with VM; C) S/I = 1.0, without VM; D) S/I = 1.0, with VM.
0
100
200
300
400
0 10 20 30 40 50
ml
CH
4/g
VS
ad
ded
Time (Days)
S/I = 0.5; Without VM
EM 0.5 LF 0.5FEM 0.5 FLF 0.5
A)
0
100
200
300
400
0 10 20 30 40 50
ml
CH
4/g
VS
ad
ded
Time (Days)
S/I = 0.5; With VM
EM 0.5 VM LF 0.5 VM
FEM 0.5 VM FLF 0.5 VM
B)
0
100
200
300
400
0 10 20 30 40 50
ml
CH
4/g
VS
ad
ded
Time (Days)
S/I = 1.0; Without VM
EM 1.0 LF 1.0FEM 1.0 FLF 1.0
C)
0
100
200
300
400
0 10 20 30 40 50
ml
CH
4/g
VS
ad
ded
Time (Days)
S/I = 1.0; With VM
EM 1.0 VM LF 1.0 VM
FEM 1.0 VM FLF 1.0 VM
D)
47
Interestingly, basal medium (VM) addition had a negative effect on biomethane yields.
This result may be related to the higher buffering capacity and higher pH values in reactors
supplemented with VM, compared to reactors with no VM supplementation. Indeed, final pH
measurements revealed pH values from 7.2 to 7.6 for VM-supplemented reactors, compared to
pH values from 6.5 to 7.0 for reactors with no VM addition (data not shown). High pH
conditions may have resulted in an “inhibited steady state”, during which the ammonia
concentrations may have risen to levels high enough to cause process instability and temporary
VFA accumulation (Montingelli et al., 2015).
In general, higher biomethane yields were observed in reactors with a larger S/I of 1.0
(Figure 2-1A–D). The highest biomethane yield among all reactors was 390 ± 0.1 ml CH4/g
VSadded, in the reactor with fermented LF duckweed (FLF) without VM addition, at an S/I of 1.0.
This value was 51.2 % higher than the corresponding raw duckweed reactor with no VM addition
at an S/I of 1.0 (LF 1.0). The superior biomethane production in reactors fed with fermented
duckweed indicates that upstream ethanol fermentation had a positive impact on methanogenic
activity. This has previously been attributed to direct interspecies electron transfer pathways
triggered by the presence of ethanol in methanogenic digesters (Zhao et al., 2017), which enhance
the synthrophic metabolism of VFAs such as propionate and butyrate (Zhao et al., 2016).
Biomethane produced with both fermented duckweed types was higher than that reported for the
anaerobic digestion of food waste fermentation residues of 248 ml CH4/g VSadded (Wu et al.,
2015).
Overall bioenergy yields
Overall bioenergy yields of EM and LF duckweeds by separate and sequential processes
of ethanol fermentation and anaerobic digestion are summarized in Table 2-1. Comparison of the
48
separate processes revealed that biomethane production from duckweed provides higher energy
gain. Therefore, 100% of the duckweed biomass allocated to biomethane production was used as
the basis of comparison for the energy yield performance of the coupled process. However, it is
important to note that relative market values of bioethanol and biomethane may lead to a
difference in the allocation of duckweed end products. The highest bioenergy yield in this study
was obtained from LF duckweed subjected to the coupled sequential bioethanol and biomethane
process, which provided 70.4 % higher overall energy yield compared to sole biomethane
production. This value is comparable to the literature. For example, thermochemically pretreated
oat straw recovered 85 to 87 % higher heating value from the biomass in the coupled process,
which is 28 to 34 % higher than direct anaerobic digestion (Rabelo et al., 2011). Based on these
results, the coupled process seems more attractive for enhancing bioenergy gain. Techno-
economics of the coupled process must still be taken into account to arrive at a definitive
conclusion.
49
Conclusion
In this study, it was demonstrated that significant methane production from duckweed is
possible. Contrary to the current literature, from an energy yield standpoint, anaerobic digestion
of duckweed seems to be a more reasonable approach than its fermentation into ethanol.
Nevertheless, upstream ethanol fermentation results in even higher (51.2 %) biomethane yields
when compared to anaerobic digestion of raw duckweed, increasing the overall energy gain by
70.4 %. To further demonstrate the technical feasibility of a coupled system, mass and energy
balances, as well as a techno-economic analysis of the coupled system, must be performed.
Table 2-1: Bioethanol, biomethane, and bioenergy yields from Eco-Machine (EM) and Living-
Filter (LF) duckweed biomass through separate and coupled ethanol fermentation and anaerobic
digestion processes.
Eco-Machine™
(EM)
Living-Filter
(LF)
1 Bioethanol production
a Theoretical maximum glucose (g) 11.8 ± 0.7 18.3 ± 0.9
b Glucose recovery (g/L) 6.5 ± 0.8 17.9 ± 0.6
c Glucose recovery (%) 55.5 ± 6.7 97.6 ± 3.4
d Ethanol produced (g/L) 3.2 ± 0.3 7.3 ± 0.3
e Ethanol yield (g ethanol / g glucose) 0.50 ± 0.01 0.41 ± 0.03
f Ethanol yield (g ethanol / g TS) 0.07 ± 0.01 0.15 ± 0.01
2 Biomethane production
a Raw duckweed methane yield (ml CH4/g VS) 258 ± 0.0 259 ± 0.3
b Raw duckweed methane yield (ml CH4/g TS) 183 ± 0.0 192 ± 0.2
c Fermented duckweed methane yield (ml CH4/g VS) 328 ± 0.1 390 ± 0.1
d Fermented duckweed methane yield (ml CH4/g TS) 261 ± 0.0 289 ± 0.0
3 Bioenergy production
a Ethanol from raw duckweed (kJ/g TS) 1.9 ± 0.2 4.3 ± 0.2
b Methane from raw duckweed (kJ/g TS) 6.8 ± 0.0 7.5 ± 0.0
c Net ethanol recovered after distillation (kJ/g TS) 1.4 ± 0.2 3.7 ± 0.2
d Methane from fermentation residue (kJ/g TS) 8.9 ± 0.0 9.1 ± 0.0
e Total energy yield of coupled process (kJ/g TS) 10.3 ± 0.2 12.8 ± 0.2
f *Energy gain of coupled over separate processes (kJ/g TS) 3.5 ± 0.2 5.3 ± 0.2
* 3f = 3e – 3b (Energy gain of coupled over separate processes has been compared to the maximum energy
gain potential of the separated process)
50
Chapter 3
Additional Product in the Grid: Effect of pH and Temperature on Microbial
Community Structure and Carboxylic Acid Yield during the Acidogenic
Digestion of Duckweed
This chapter has been published as follows:
Calicioglu, O., Shreve, M.J., Richard, T.L., Brennan, R.A., 2018. Effect of pH and
temperature on microbial community structure and carboxylic acid yield during the acidogenic
digestion of duckweed. Biotechnol. Biofuels 1–19. doi:10.1186/s13068-018-1278-6.
Note that molecular techniques applied in this chapter were performed by Michael J.
Shreve in a collaborative effort to characterize the microbial communities in acidogenic
digestions of the aquatic plant duckweed under various pH and temperature conditions.
Abstract
In this study, a series of laboratory batch experiments were performed to determine the
favorable operating conditions (i.e., temperature and pH) to maximize carboxylic acid production
from wastewater-derived duckweed during acidogenic digestion. Batch reactors with 25 grams
per liter solid loading were operated anaerobically for 21 days under mesophilic (35oC) or
thermophilic (55oC) conditions at an acidic (5.3) or basic (9.2) pH. At the conclusion of the
experiment, the dominant microbial communities under various operating conditions were
assessed using high-throughput sequencing.
The highest duckweed-to-carboxylic acid conversion of 388 ± 28 mg acetic acid
equivalent per gram volatile solids was observed under mesophilic and basic conditions, with an
average production rate of 0.59 grams per liter per day. This result is comparable to those
reported for acidogenic digestion of other organics such as food waste. The superior performance
51
observed under these conditions was attributed to both chemical treatment and microbial
bioconversion. Hydrogen recovery was only observed under acidic thermophilic conditions, as
23.5 ± 0.5 ml per gram of duckweed volatile solids added. More than temperature, pH controlled
the overall structure of the microbial communities. For instance, differentially abundant
enrichments of Veillonellaceae acidaminococcus were observed in acidic samples, whereas
enrichments of Clostridiaceae alkaliphilus were found in the basic samples. Acidic mesophilic
conditions were found to enrich acetoclastic methanogenic populations over processing times
longer than ten days.
Introduction
Throughout the industrial era, population growth and increased consumption have
resulted in a steady increase in the demand for energy. This demand has been met mainly by
nonrenewable fossil-based resources (i.e. coal, crude oil, crude gas) (Hatti-Kaul et al., 2007),
which generate excessive CO2 emissions and other environmental concerns (Aiello-Mazzarri et
al., 2006). As a renewable and sustainable alternative, advanced biomass energy approaches have
been attracting increasing attention (Jung et al., 2016). However, feedstock sustainability,
availability, and affordability issues remain a serious concern. In this context, an environmentally
friendly, socially acceptable, and economically feasible biomass crop could overcome the
challenges faced by the majority of biofuels on the energy market.
Lemnaceae (duckweeds) represent a family of simple, fast-growing, floating aquatic
plants, with five genera (Landoltia, Lemna, Spirodela, Wolffia, and Wolfiella) and 38 species
classified to date (Cui and Cheng, 2015; Xu et al., 2014). Production of duckweed rich in starch
and cellulose can be integrated into wastewater treatment systems, which can improve the
economics of the feedstock production process (Cheng and Stomp, 2009). Moreover, the low
52
lignin content of duckweeds relative to lignocellulosic agricultural residues and traditional energy
crops make them an attractive alternative for conversion into bioethanol, since they do not require
intensive pretreatment prior to saccharification. Previous studies with duckweed have
investigated its use as a feedstock to produce either sugar or syngas intermediates; these two
platforms have dominated most of the public funding as well as private investment in advanced
biorefineries. Thermochemical conversion of duckweed into syngas demonstrated pathways to
gasoline, diesel, and jet fuel (Baliban et al., 2013). Biochemical conversion of duckweed starch
and cellulose into simple sugars and fermentation into alcohols has also been demonstrated, and
been applied at both laboratory and pilot scales (Su et al., 2014; Xu et al., 2011).
A third biomass–to–biofuel conversion strategy has been termed the carboxylate platform
(Holtzapple et al., 1999). This platform utilizes mixed cultures for anaerobic degradation of
organic matter into carboxylic acid intermediates, a process that has been termed acidogenic
digestion. During acidogenic digestion, 2 to 5 carbon volatile fatty acids (VFAs) are initially
produced, and can be converted into longer chain fatty acids consisting of six or more carbon
atoms through chain elongation via mixed cultures (Steinbusch et al., 2011). These longer chain
fatty acids have a higher energy density than short term VFAs, and are precursors of higher-value
chemicals and biofuels such as esters, alcohols, and alkanes (Agler et al., 2011).
Acidogenic digestion is advantageous over alcohol fermentation due to: (1) the potential
to directly utilize feedstocks such as duckweed without requiring pretreatment; (2) production of
a single class of end-products; (3) the absence of sterilization requirements; and (4) convertibility
of longer chain products (3-carbon and higher) into higher-value chemicals and fuels (Holtzapple
and Granda, 2009). However, there do not appear to be any prior published studies on processing
duckweed through the carboxylate platform.
Carboxylate platform systems also have some drawbacks, such as requiring process
control to avoid a shift into methanogenic activity (Datta, 1981). Methanogenic activity is
53
normally inhibited by either chemical addition or avoiding the conditions which favor
methanogens (e.g., maintaining pH outside the range of 5.5 – 8.5, which methanogens prefer).
Indeed, the literature suggests that higher VFA concentrations can be achieved under alkaline
conditions of pH 9 to pH 10 (Yu et al., 2008), which should simultaneously suppress
methanogenic activity. Under high ammonia concentrations present in reactors at elevated pH,
anaerobic bacteria are expected to outcompete methanogenic archaea (Appels et al., 2008).
However, the behavior of acidogenic microbial consortia at high pH is not well understood.
The objectives of this work were: (1) to evaluate the effect of operating conditions such
as temperature and pH on the acidogenic digestion of duckweed, (2) to quantify conversion rates
and the associated carboxylic acid yields, and (3) to characterize the dominant microbial taxa
present under various operating conditions. This study is the first to determine the performance of
duckweed during acidogenic digestion under various operating conditions, with an emphasis on
investigating the resulting acidogenic microbial consortia.
Materials and Methods
Analytical methods
The moisture, total solids (TS), and volatile solids (VS) contents of duckweed and the
inocula were determined according to the National Renewable Energy Laboratory (NREL)
Laboratory Analytical Procedure (LAP) for biomass and total dissolved solids of liquid process
samples (Sluiter et al., 2008). Ash content was measured according to NREL LAP for
determination of ash in biomass (Sluiter et al., 2004). Carboxylic acids (i.e., VFAs) were
quantified using Gas Chromatography (GC) (SHIMADZU, GC-2010 Plus, Japan) with a flame
ionization detector. The final total VFA yields were calculated in terms of acetic acid equivalents
54
per gram duckweed volatile solids added (HAceq g VSadded-1) (E.M. Siedlecka, J. Kumirska, T.
Ossowski, P. Glamowski and J. Gajdus, Z. Kaczyński, 2008). Carbon quantification of samples
were performed using a total carbon (TC) analyzer (SHIMADZU, TOC-V CSN, Kyoto, Japan)
equipped with solid sample module (SHIMADZU, 5000A, Kyoto, Japan). Total ammonia
nitrogen (TAN) concentrations were measured by selective electrode method as described in
Standard Methods No. 4500 (APHA/AWWA/WEF, 2012), using an ammonia probe (Orion,
9512, USA). Headspace pressure in the reactors was measured using a pressure gauge (Grainger,
DPGA-05, USA). If the pressure was found to be negative or zero, no volume readings were
performed in order to avoid disturbance of the headspace gas composition. The gas volumes of
reactors were measured using a water displacement device filled with 0.02 M hydrochloric acid.
Since the measurement process was quick, the headspace temperature was assumed to be constant
and equal to 35oC (El-Mashad, 2013; Theodorou et al., 1994). Volume readings were reported at
standard temperature and pressure. Volumetric methane (CH4) and hydrogen (H2) concentrations
were determined by extracting headspace from the reactors using a 250 μl airtight syringe
(Hamilton, Reno, NV, USA) and injecting onto a GC (SRI Instruments, SRI310C, Torrance, CA,
USA) equipped with 6-foot molecular sieve column (SRI 8600-PK2B, USA) in continuous mode
at 80oC with argon as the carrier gas. Volumetric carbon dioxide (CO2) concentrations were
quantified using an identical GC equipped with 3-foot silica gel packed column (SRI, 8600-
PK1A,USA) in continuous mode at 60oC with helium as the carrier gas.
Plant material and growth conditions
Duckweed was collected on May 29, 2016, from an open pond within the effluent spray
fields of the Pennsylvania State University Wastewater Treatment Plant, a.k.a. the “Living-
Filter”, receiving on average (n = 9): 2.3 ± 0.5 mg L-1 carbonaceous biological oxygen demand;
55
1.5 ± 0.1 mg L -1 phosphorus; 0.6 ± 0.9 mg L-1 TAN; 5.8 ± 1.5 mg L-1 nitrate; 0.3 ± 0.2 mg L-1
nitrite; and 1.3 ± 0.4 mg L-1 total Kjeldahl nitrogen. The duckweed species in the pond was
identified as a monoculture of Lemna obscura (100% sequence identity to accession number
GU454331.1, in the NCBI database) through DNA extraction and sequencing as described
previously (Calicioglu and Brennan, 2018). Prior to using in these experiments, the duckweed
was rinsed with tap water and dried at 45 ± 3oC to a constant weight over two days. Duckweed
was then analyzed for its moisture (5.0 ± 0.4%), and VS (85.6 ± 0.4%) contents. The composition
of duckweed was determined as (% per dry weight): cellulose (11.8 ± 0.9); hemicellulose (20.5 ±
1.0); starch (9.8 ± 0.9); lignin (1.6 ± 1.2); water soluble carbohydrates (19.9 ± 0.2); and crude
protein (18.2 ± 0.2) (Dairy One Wet Chemistry Laboratory, Ithaca, NY). A separate batch of
duckweed was used to enrich the inoculum, which was previously collected from the same pond,
and dried at 45 ± 3oC to a constant weight. Subsamples of dried duckweed were collected and
stored at -80oC for future DNA analysis.
Inoculum
A combination of mesophilic and thermophilic seeds were collected to prepare the
inoculum: silage, rumen fluid, and anaerobic wastewater sludge were used as mesophilic seeds;
and compost was used as a thermophilic seed (Fong et al., 2006; Hamelers, 2001; Tuomela et al.,
2000). Silage and rumen fluid were obtained from the Pennsylvania State University Dairy Farm
(University Park, PA). Anaerobic wastewater sludge was obtained from the Pennsylvania State
University Wastewater Treatment Plant’s secondary digester. Compost was obtained from the
Pennsylvania State University composting facility. Silage (360 g) and compost (180 g) were each
blended separately in 1 L of 25 mM phosphate buffered saline (PBS) at pH 6.8. Rumen fluid was
centrifuged at 2880 rgf for 30 minutes and the pellet was re-suspended in 1 L of 25 mM PBS at
56
pH 6.8. All three sources were incubated separately overnight at 35oC. Solids from 1.5 L
anaerobic sludge were collected by centrifuging at 2880 rgf for 30 minutes (Eppendorf, 5804 R,
Germany) and were re-suspended in 25 mM PBS at pH 5.0, incubated at 35oC overnight, and
boiled for 1 h to inhibit methanogenic activity (Arslan et al., 2013; Fernandes et al., 2009).
All four sources were screened through a sieve with 150 µm opening. The permeates
were blended in equal parts (on a VS basis), previously harvested duckweed was added at a
substrate-to-inoculum ratio of 0.1, and the cultures were acclimated to acidic (pH = 5.3) or basic
(pH=9.2) conditions for five and seven days, respectively, at 35 oC until substantial biogas
production was observed. The final slurries were both centrifuged for 30 minutes at 2880 rgf and
the inoculum solids collected. An aliquot of each inoculum was collected and stored at -80oC for
later DNA extraction. The final compositions of the two inocula were: 84.0 ± 0.1% moisture, and
74.4 ± 1.2% VS of TS for the acidic inoculum; and 84.5 ± 0.2% moisture, and 60.3 ± 0.2% VS of
TS for the basic inoculum.
Acidogenic digestion
Batch reactors (300 ml working volume) were fed with duckweed to achieve a total solids
content of 25 g L-1, and inoculum was added at an inoculum-to-substrate ratio of 0.1 on a VS
basis. Initial pH values were adjusted to either pH 5.3 or pH 9.2. Reactors to be operated under
basic conditions were supplemented with 4.0 g L-1 sodium carbonate as buffer, which is
equivalent to about 5% of the duckweed carbon added and was quantified in the carbon balance
accordingly. All reactors were purged with nitrogen gas for three minutes and sealed to provide
anaerobic conditions. Reactors were operated under mesophilic (35oC) or thermophilic (55oC)
conditions for 21 days. Once every two days, headspace gas volume and composition were
measured, liquid samples were taken, and the pH was adjusted to either 5.3 or 9.2. Test reactors
57
were run in triplicate, and controls (with no substrate) were run in duplicate. The observed biogas
values in control reactors were subtracted from those observed in active reactors. The VFA
production values, however, were found to be negligible compared to those achieved in active
reactors; therefore, they were not subtracted. Duplicate blank reactors (with no inoculum) were
also operated to evaluate the acidogenic digestion potential of microorganisms naturally
associated with the duckweed, which as previously described was air dried at 45oC, and not
sterilized.
At the end of reactor operation period, samples for microbial community analysis were
obtained under axenic conditions. Prior to sacrificing the reactors, 6 ml of liquid were withdrawn
and centrifuged sequentially (2 ml at a time) in 2 ml Eppendorf tubes, discarding the supernatant
after each cycle to concentrate suspended solids for DNA extraction. Samples for DNA extraction
were stored at -80oC until processed. The rest of the reactor constituents were wet sieved by
pressing through a 340 µm opening. The screenings were analyzed as reactor liquids, and the
retentates were analyzed as reactor solids. TAN of the liquids were measured.
Carbon balance
Initial and final TC concentrations of the headspace, liquids, and solids were reported.
Headspace TC was calculated as the sum of CO2 and CH4 recovered over the 21-day operation
period, and the amounts remaining in the headspace at the end of operation. The VFA losses
during solids drying were estimated as 95% for acidic reactors and 55% for basic reactors
(Vahlberg et al., 2013). The sampling losses were calculated as 24 sampling events of 2 ml each.
The mass closure has been calculated as the ratio of the final to initial total carbon values
(Appendix B).
58
DNA extraction, PCR amplification, and high-throughput sequencing
DNA was extracted from approximately 100 mg each of acclimated inoculum and
suspended biomass from final (day 21) reactor contents using a Mo Bio PowerSoil DNA
extraction kit (MO BIO Laboratories, Inc., Carlsbad, CA, USA) according to the manufacturer’s
protocol. Microbial DNA was isolated from dried duckweed samples using the same kit, by
adding approximately 25 mg of plant tissue and following the manufacturer’s protocol. The V4
region of the 16S rRNA gene (bacteria and archaea) was PCR-amplified using the primers 515F-
Y (5’- GTGYCAGCMGCCGCGGTAA-3’) and 806RB (5’- GGACTACNVGGGTWTCTAAT-
3’) (Apprill et al., 2015; Parada et al., 2015). Forward and reverse overhang adapters were
appended to the 5’ end of the locus specific primers to accommodate the addition of sample
indices via a second PCR step (Forward overhang: 5’-
TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG-3’; Reverse overhang: 5’-
GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG-3’). Each 20 uL PCR reaction
contained 1X Invitrogen Platinum SuperFi Master Mix (Thermo Fisher Scientific, Waltham, MA,
USA), 0.2 µM of each primer, and 0.25 ng uL-1 of template. PCR thermal cycling conditions were
as follows: initial denaturation at 98 °C for 2 min;, followed by 25 cycles of 98°C for 10 s,
56.5°C for 20s, and 72 °C for 15s; and a final extension at 72°C for 5 min. No-template,
mismatched template (fungal DNA), and positive controls were included for all PCR reactions.
PCR was carried out in triplicate for each sample and the reaction products pooled. PCR products
were submitted to the Huck Institutes of the Life Sciences (Huck), Genomics Core Facility (The
Pennsylvania State University, University Park, PA) where sample indices were added via a
second PCR step (10 cycles) using the Illumina Nextera XT Index Kit (Illumina, Inc., San Diego,
CA) following the manufacturer’s protocol. Sample libraries were then normalized using a 96-
well SequalPrep Normalization Plate Kit (ThermoFisher Scientific, Waltham, MA) following the
59
manufacturer’s protocol. Samples with a normalized concentration of approximately 1.25 ng µl-1
were pooled and checked for quality using an Agilent 2100 Bioanalyzer (Agilent, Santa Clara,
CA) in conjunction with a High Sensitivity DNA Kit (Agilent, Santa Clara, CA). The final pooled
library was quantified using a Kapa Library Quantification Kit (KK4835; Kapa Biosystems,
Wilmington, MA) according to the manufacturer’s protocol. The pool was loaded at a final
concentration of 7pM. The pool of libraries was sequenced on an Illumina MiSeq using 250 x 250
paired-end sequencing but utilizing MiSeq Reagent Kit v3 (600 cycle). The raw sequencing reads
were deposited in the Sequence Read Archive (SRA) of the National Center for Biotechnology
Information (NCBI) database under accession number SRP150539.
Bioinformatics
Paired-end sequencing data was received in an already de-multiplexed format. Primer
sequences were trimmed from the forward and reverse reads using cutadapt (Martin, 2011) before
joining the paired-end reads using fastq-join (Aronesty, 2013) with a minimum overlap of 30 nt
and a maximum difference of 30% in the overlap region. Joined reads were then filtered by length
to include only those of the expected size (251-256 nt retained). The Quantitative Insights Into
Microbial Ecology (QIIME; version 1.8.3) (Caporaso et al., 2010) workflow
multiple_split_libraries_fastq.py was then used to quality filter the remaining reads, retaining
reads which were 95% of their original length after truncation at the first base call with a Phred
quality score below 20. Quality-filtered sequences were checked for chimeras against the
ChimeraSlayer reference dataset (version microbiomeutil-r20110519) using VSEARCH (Rognes
et al., 2016).
Downstream analysis of chimera-free quality-filtered sequence sets was carried out using
QIIME. Open reference operational taxonomic unit (OTU) clustering using the
60
pick_open_reference_otus.py workflow was used to cluster sequences using a combination of de-
novo and reference based methods against the GreenGenes reference database (version 13_8) at
97% sequence similarity. The uclust (Edgar, 2010) clustering method was used and only OTUs
containing two or more sequences were retained. When using the GreenGenes database to assign
taxonomy to 16S rRNA amplicon sequences derived from plant associated samples, mispriming
(and amplification) of plant DNA can be revealed through sequences classified as chloroplast at
the class level (Hanshew et al., 2013). All OTUs classified as chloroplast at the class level were
filtered from the OTU table using the QIIME script filter_taxa_from_otu_table.py prior to
diversity analysis and taxonomic summary steps.
Alpha diversity, beta diversity, and taxonomic analysis was performed using the
core_diversity_analysis.py workflow at a rarefaction depth of 29,500 sequences per sample (other
settings default). Additional alpha diversity metrics were calculated using the alpha_diversity.py
script in QIIME. Principal Coordinate Analysis (PCoA) was carried out on the weighted and
unweighted UniFrac distance matrices generated by core_diversity_metrics.py, using the
cmdscale function in base R (version 3.4.4) in order to produce more suitable plots. To identify
differentially abundant taxa between the main treatment groups (acid vs. basic and mesophilic vs.
thermophilic), the OTU table was collapsed to the genus level using the QIIME script
summarize_taxa.py. The collapsed table was then filtered to exclude genera present in less than
25% of samples and those whose total abundance within the table was less than 150 counts.
Filtering was performed using the QIIME script filter_otus_from_otu_table.py. Differentially
abundant taxa were identified using the QIIME script group_significance.py, and comparisons
were made using a nonparametric t-test. The QIIME script compare_categories.py was used to
analyze the strength and statistical significance of sample groupings (acidic vs. basic and
mesophilic vs. thermophilic) in terms of beta diversity. Both weighted and unweighted UniFrac
61
distance matrices were used for comparing groupings under both conditions and the test method
was permanova with 999 permutations.
Statistical analysis
Data are presented as the mean ± standard deviation of triplicate samples. Significant
differences between means were tested using one-way analysis of variance (ANOVA) and least
significant difference (LSD) tests at a significance level of p<0.05 (Appendix B), using Minitab
statistical package (Version 3.1, Minitab Inc., USA).
Results
Acidogenic digestion performance
All reactors produced VFAs, ranging in final concentrations from 1.1 ± 0.1 to 9.0 ± 0.7
mg L-1 (Figure 3-1). The highest VFA production was observed under basic mesophilic
conditions (Figure 3-1-C), where the average composition consisted of 83.0% acetic, 6.3%
propionic, 3.6% isobutyric, 2.7% n-butyric, and 4.4% isovaleric acids. These results correspond
to a total of 388 ± 28 mg VFA as HAceq g VSadded-1 (334 ± 24 mg VFA as HAceq g TSadded
-1, Table
3-1). Approximately 80% of the final VFA values were achieved by day 13, with an average
production rate of 0.59 g HAceq L-1 d-1 under these conditions.
62
Figure 3-1: Volatile Fatty Acid profiles of the acidogenic duckweed reactors over 21 days.
Legend: Reactors were operated under: A) Acidic Mesophilic, B) Acidic Thermophilic, C) Basic
Mesophilic, D) Basic Thermophilic conditions. Narrow stacked columns represent blank reactors
(no inoculum) whereas thick stacked columns represent active (with inoculum) reactors. Error
bars are cumulative standard deviations of the individual stacked bars.
63
The lowest final VFA concentrations were observed in the active reactors operated under
acidic mesophilic conditions, in which the acetic acid concentration increased until Day 9 and
then gradually disappeared (Figure 3-1-A), presumably converted into CH4 and CO2 (Figure 3-2-
A). In order to avoid bias on evaluation of acidogenic digestion performance, it was assumed that
the acetate produced had been converted into equal moles of CO2 and CH4. According to this
stoichiometry, the loss in the VFA yield could be back-calculated as 200 ± 20 mg VFA as HAceq
g VSadded-1 (171 ± 17 mg VFA as HAceq g TSadded
-1), in which case the “actual” yield under acidic
mesophilic conditions would have been 256 ± 23 mg VFA as HAceq g VSadded-1 (219 ± 20 mg
VFA as HAceq g TSadded-1).
Table 3-1: Final volatile fatty acid yields of the blank and active reactors under acidic mesophilic,
acidic thermophilic, basic mesophilic, and basic thermophilic conditions.
Acidic
Mesophilic
Acidic
Thermophilic
Basic
Mesophilic
Basic
Thermophilic
VFA Yields
(mg VFA as HAceq g VSadded-1 )
Blank: 218 ± 7.7 a 116 ± 28 ab 256 ± 37 a 86 ± 22 b
Active: 55 ± 3.7 a 117 ± 7.9 b 388 ± 28 c 341 ± 2.8 d
Note: Mean VFA yields were compared separately for blank and active groupings using TUKEY test at a significance level of p<0.05.
Superscript letters indicate the resulting statistical groupings within reactor class.
64
Among the blank reactors, the final VFA concentrations varied between 2.1 ± 0.5 and 5.9
± 0.8 mg L-1; however, when comparing the yields for all blank reactors, a statistically significant
difference was found only between the conditions with the highest (basic mesophilic) and lowest
(basic thermophilic) yields (Table 3-1; Appendix B). In contrast, the final VFA compositions
Figure 3-2: Cumulative biogas, hydrogen, methane, and carbon dioxide yields of the acidogenic
duckweed reactors over 21 days. Reactors were operated under: A) Acidic Mesophilic; B) Acidic
Thermophilic; C) Basic Mesophilic; D) Basic Thermophilic conditions. Blank (no inoculum)
reactors are represented as empty bullets whereas active (with inoculum) reactors are represented
as solid bullets.
65
varied between operating conditions (Figure 3-1-A,B,C). Potential reasons for these observations
are considered in the Discussion.
Temperature had an adverse effect for blank reactors with no inoculum under basic
thermophilic (55oC) conditions, as their VFA yield of 86 ± 22 mg VFA as HAceq g VSadded-1 (74 ±
19 mg VFA as HAceq g TSadded-1) was about one-third of the value observed under mesophilic
conditions, observed as 256 ± 37 mg VFA as HAceq g VSadded-1 (219 ± 32 mg VFA as HAceq g
TSadded-1). The effect of temperature was less pronounced for active reactors operated under basic
conditions. Similarly, increased temperature had a negative impact on the average final VFA
yield in blank reactors under acidic conditions. Active acidic reactors were more prone to VFA
loss due to methanogenic activity; however, the back-calculation of the acetate yields taking the
CO2 and CH4 productions into account show that the mesophilic (35oC) conditions would have
yielded higher VFA concentrations compared to those of thermophilic conditions for the active
reactors as well.
Control reactors without duckweed produced negligible amounts of VFAs, in part
because the inocula were pretreated, enriched, and starved prior to the experiments. Also, the
substrate-to-inoculum ratio of 10 used in this study was significantly lower than the common
value used for anaerobic digestion trials, which typically varies between 0.5-2 for substrates rich
in cellulose (Chynoweth et al., 1993b). Therefore, results pertaining to the effects of endogenous
respiration have been omitted.
In both blank and active reactors operated under basic conditions, the acetic acid fraction
of VFAs was higher than under acidic conditions, where larger fractions of longer chain VFAs
(i.e. propionic, butyric, valeric, caproic) were observed. For instance, under thermophilic
conditions, acidic reactors had a final composition of 69.6% acetic and 30.4% butyric acids,
whereas basic reactors had a final composition of 78.0% acetic, 5.1% propionic, 4.3% isobutyric,
and 6.4% n-butyric acids.
66
Biogas production was observed in all reactors to some extent (Figure 3-2); however, the
quantities and the compositions varied greatly among treatments. The highest biogas production
was recorded in the active acidic reactors operated under mesophilic conditions (124 ± 8.6 ml g
duckweed VSadded-1). In these reactors, the predominant gas species recovered was CO2 (59.2% of
the total gas recovered), followed by CH4 (21.3% of the total gas recovered) (Appendix B). The
CH4 recovery started by Day 9 and reached a cumulative yield of 26.6 ± 3.8 ml g duckweed
VSadded-1. High CO2 release (61.4% of the total gas recovered) was also observed in the acidic
mesophilic blank reactors, but CH4 was not produced in the absence of inoculum.
Biogas recovery was minimal in basic reactors under both mesophilic and thermophilic
conditions, and was only observed in the first 9 days, mainly as CO2 (Figure 3-2-C, D). Over
time, the headspace gas compositions changed and the final contents in active reactors were
found to be 1.6 ± 0.04% CO2 and 52.5 ± 6.1% CH4 in the basic mesophilic reactors, and 2.3 ±
0.3% CO2 and 56.8 ± 2.2% CH4 in the basic thermophilic reactors. However, significant
cumulative recovery of biogas was not observed under either of these conditions.
In contrast to the other three treatments, no CH4 was observed under acidic thermophilic
conditions. Instead, this was the only condition under which H2 was produced (Figure 3-2-B),
with an observed yield of 21.8 ± 4.6 and 23.5 ± 0.5 ml g duckweed VSadded-1
in blank and active
reactors, respectively. These values correspond to 33.1% and 43.8% of the total gas recovered
from blank and active reactors.
Carbon balance
The fractions of initial and final solid, particulate, soluble, and gaseous TC were
compared for both blank and active reactors, as percentage of the initial TC content in each
67
reactor (Figure 3-3). The average mass closure values on TC basis varied between 82.9 ± 6.7%
and 102.2 ± 1.9% among different operating conditions with and without inoculum addition.
Figure 3-3: Carbon balance of the acidogenic duckweed reactors. Total carbon percent
contributions from initial duckweed, inocula, and alkalinity, and final soluble (<0.2 µm),
particulate (>0.2 µm; <340 µm), solid (>340 µm), and gaseous phases of the reactors under: A)
Acidic Mesophilic, B) Acidic Thermophilic, C) Basic Mesophilic, D) Basic Thermophilic
conditions. Error bars are cumulative standard deviations of the individual measurements.
68
The carbon balance results revealed that the highest solubilization efficiency (i.e. highest
increase in the soluble TC content) was achieved under basic mesophilic conditions (52.7%). The
lowest final solids content was also observed under these conditions (27.4%). An average of
61.0% of the soluble TC was VFA-carbon, accounting for 34.5% of the duckweed TC added in
these reactors. The lowest average percentage of soluble TC was found in the acidic mesophilic
active reactors; however, the solids were instead converted to particulate and gaseous TC at a
higher extent in these reactors compared to others.
The major TC loss to the gaseous phase was observed in the active acidic mesophilic
reactors, due to the highest biogas recovery, which consisted of both CO2 and CH4 (Figure 3-2-A,
3-A). For the rest of the acidic (active and blank) reactors, CO2 was the predominant gas.
Although not recovered in significant quantities, residual CH4 in the reactor headspace
constituted most of the TC lost to the gaseous phase in the basic active reactors (Appendix B).
In the acidic blank reactors, the particulate TC concentration was below detection and
insignificant compared to the soluble and solid TC values. However, more particulate matter was
observed in the active counterparts, supplemented with inoculum. Overall, particulate TC
concentration was higher in the basic reactors, with the value observed in basic thermophilic
blank reactors (average 18.6%).
Overall, the active reactors exhibited better solids reduction compared to their blank
counterparts, except for acidic thermophilic conditions, where the opposite held true (Figure 3-3-
D). In parallel, the final biogas yield was higher in acidic thermophilic blanks, compared to the
actives.
69
Microbial community analysis
Good’s coverage ranged from 0.957 to 0.999, indicating that a majority of the microbial
diversity was captured at the rarefied depth of 29,500 sequences per sample (Table 3-2). OTU
richness varied widely across all samples both in terms of observed OTUs (128-4815) and the
chao1 richness estimator (191-6856). Samples with the lowest OTU richness included blank
reactors and the active acidic thermophilic reactors. The highest OTU richness was observed for
control reactors and inoculum. Samples were ranked similarly with regard to the Simpson
diversity index (0.179-0.993) and Shannon diversity index (0.605-9.16), which also account for
evenness.
Table 3-2: Alpha diversity metrics for microbial populations in duckweed acidogenically digested
under different environmental conditions.
Sample Type Good's
Coverage
Observed
OTUs Chao1
Shannon
Diversity
Index
Simpson
Diversity
Index
Acidic
Mesophilic
Blank 0.997 ± 0.001 789 ± 86 943 ± 81 4.53 ± 0.21 0.891 ± 0.017
Active 0.990 ± 0.002 1819 ± 82 2495 ± 54 6.04 ± 0.07 0.955 ± 0.003
Control 0.967 ± 0.003 2688 ± 13 3772 ± 240 7.77 ± 0.23 0.974 ± 0.009
Acidic
Thermophilic
Blank 0.999 ± 0.000 135 ± 9 221 ± 42 0.85 ± 0.35 0.261 ± 0.117
Active 0.993 ± 0.001 981 ± 89 1511 ± 129 3.64 ± 0.07 0.809 ± 0.018
Control 0.989 ± 0.000 2492 ± 235 2960 ± 258 6.66 ± 0.40 0.945 ± 0.028
Basic Mesophilic
Blank 0.993 ± 0.003 1155 ± 298 1481 ± 381 5.65 ± 0.68 0.947 ± 0.017
Active 0.983 ± 0.002 2145 ± 235 3155 ± 236 6.23 ± 0.41 0.947 ± 0.018
Control 0.960 ± 0.002 4226 ± 98 6141 ± 0 8.61 ± 0.08 0.988 ± 0.002
Basic
Thermophilic
Blank 0.993 ± 0.001 1135 ± 151 1463 ± 146 4.62 ± 0.26 0.826 ± 0.025
Active 0.986 ± 0.001 2251 ± 222 3156 ± 361 5.77 ± 0.20 0.916 ± 0.012
Control 0.960 ± 0.004 4626 ± 267 6568 ± 407 9.15 ± 0.01 0.993 ± 0.000
Acidic Inoculum 0.976 ± 0.004 2637 ± 242 3706 ± 49 7.12 ± 0.06 0.974 ± 0.001
Basic Inoculum 0.961 ± 0.000 3421 ± 92 5129 ± 1 8.48 ± 0.03 0.988 ± 0.000
Duckweed 0.984 ± 0.002 1539 ± 159 2053 ± 129 7.07 ± 0.17 0.971 ± 0.006
70
All reactors were dominated by the class Clostridia, within the phylum Firmicutes, which
averaged 70.5% relative abundance (min 35.8%; max 99.6% in blank acidic thermophilic
reactors) (Figure 3-4). Members of the class Clostridia were rare on duckweed (< 2% relative
abundance), but dominant in the inoculum (average 43.3%) suggesting that inoculum mainly
contributed to the presence of Clostridia in active and control reactors. However, acidic
thermophilic blank reactors were dominated by Clostridia, which is likely due to duckweed
associated Clostridia outcompeting other taxa under these extreme conditions. Other dominant
classes of bacteria were: (1) Bacteroidia (phylum Bacteroidetes), present mainly in mesophilic
reactors and the acidic inoculum; (2) Gammaproteobacteria (phylum Proteobacteria), present
mainly in the acidic mesophilic group (8.5-21.9%%), but also prominent on duckweed (average
24.4%%); (3) Bacilli (phylum Firmicutes), present in higher abundance in all control reactors,
active basic mesophilic reactors, and both acidic and basic inocula, with the highest relative
abundance in basic inoculum (42%). The taxonomic profile of duckweed microbes is clearly
distinct from both the inocula and the reactors. In addition to Gammaproteobacteria (mentioned
above), the dominant bacterial classes associated with duckweed include Alphaproteobacteria
(22.2%), which seemed to persist in blank basic reactors (mesophilic and thermophilic), and
Betaproteobacteria (13.5%). In addition, Nostocophysideae (phylum Cyanobacteria),
Flavobacteriia (phylum Bacteroidetes), and Epsilonproteobacteria (phylum Proteobacteria)
exhibited moderate relative abundance on duckweed (5-10%), but were low in abundance or
absent in reactors. The bacterial class Actinobacteria (phylum Actinobacteria) was present at a
moderate relative abundance across inoculum samples (average 8%), but was largely absent from
reactors, aside from controls.
71
The top five genera for each reactor group operated under acidic and basic conditions are
given in Table 3-3 and Table 3-4 respectively, and significant archaeal taxa (relative abundance >
0.01%) are summarized in Table 3-5. However, taxa outside of the top five may contribute
important biochemical pathways (see Discussion). In general, the top five genera in reactors
accounted for 32.9-99.5% of the observed OTUs (average 63.8%) and the total richness captured
by the top five genera showed a strong inverse correlation with alpha diversity metrics, as
expected. The top five genera in the inocula were dominated by members of the phylum
Firmicutes, while those associated with duckweed were dominated mainly by members of the
Figure 3-4: Class-level relative abundance taxonomic bar plot.
0%
25%
50%
75%
100%
Clostridia
Bacilli
Bacteroidia
Gammaproteobacteria
Alphaproteobacteria
Actinobacteria
Betaproteobacteria
Nostocophycideae
Coriobacteriia
Planctomycetia
Flavobacteria
Epsilonproteobacteria
Erysipelotrichi
Methanobacteria
Thermomicrobia
All < 0.5% R.A.
Other
Active
Bla
nk
Contr
ol
Active
Bla
nk
Contr
ol
Active
Bla
nk
Contr
ol
Active
Bla
nk
Contr
ol
pH
5.3
pH
9.2
Enrichm
ent
Feed
Duckweed Inoculum
Acidic
Mesophilic
Acidic
Thermophilic
Basic
Mesophilic
Basic
Thermophilic
Re
lative
Ab
un
da
nce
72
phylum Proteobacteria. If the top ten genera are considered, an additional 10-20% of the OTU
richness is described.
Archaea were absent from the top 5 genera in all reactors except the active basic
thermophilic reactors, which contained 3.6% Methanobacteriaceae
methanothermobacter. Overall, the archaeal content of the reactors was low, ranging
from none detected up to approximately 4% relative abundance in the active basic thermophilic
reactors. Other dominant archaea (> 1% relative abundance) included Methanosarcinaceae
Table 3-3: Relative abundance (R.A.) and Cumulative Abundance (C.A.) of top five genera in
each reactor group operated under acidic conditions.
Co
nd
itio
n
Ty
pe
Taxa
R.A
.
(%)
C.A
.
(%)
Aci
dic
Mes
oph
ilic
Bla
nk
c__Bacteroidia;o__Bacteroidales;f__Prevotellaceae;g__Prevotella 33.1
82.1
c__Gammaproteobacteria;o__Enterobacteriales;f__Enterobacteriaceae;g__ 20.7
c__Clostridia;o__Clostridiales;f__Lachnospiraceae;Other 15.2
c__Clostridia;o__Clostridiales;f__Ruminococcaceae;g__Ruminococcus 9.1
c__Clostridia;o__Clostridiales;f__Veillonellaceae;g__Megasphaera 3.9
Act
ive
c__Bacteroidia;o__Bacteroidales;Other;Other 13.4
53.8
Unassigned;Other;Other;Other;Other;Other 11.4
c__Gammaproteobacteria;o__Enterobacteriales;f__Enterobacteriaceae;g__ 11.3
c__Bacteroidia;o__Bacteroidales;f__Prevotellaceae;g__Prevotella 8.9
c__Clostridia;o__Clostridiales;f__Ruminococcaceae;g__Ethanoligenens 8.8
Co
ntr
ol
c__Gammaproteobacteria;o__Pseudomonadales;f__Pseudomonadaceae;g__Pseudomona
s 11.6
32.9 c__Clostridia;o__Clostridiales;f__Veillonellaceae;g__Acidaminococcus 6.9
c__Clostridia;o__Clostridiales;f__Veillonellaceae;g__Succiniclasticum 5.0
c__Bacteroidia;o__Bacteroidales;f__Porphyromonadaceae;g__Parabacteroides 4.9
c__Bacteroidia;o__Bacteroidales;f__Prevotellaceae;g__Prevotella 4.5
Aci
dic
Th
erm
oph
ilic
Bla
nk
c__Clostridia;o__Clostridiales;f__Clostridiaceae;g__Thermoanaerobacterium 85.5
99.5
c__Clostridia;o__Clostridiales;f__Clostridiaceae;g__Clostridium 10.0
c__Clostridia;o__Clostridiales;f__Ruminococcaceae;g__Ruminococcus 2.5
c__Clostridia;o__Clostridiales;f__Ruminococcaceae;g__Ethanoligenens 1.3
c__Bacilli;o__Bacillales;f__Planococcaceae;g__ 0.1
Act
ive
c__Clostridia;o__Clostridiales;f__Ruminococcaceae;g__Ethanoligenens 43.0
90.6
c__Clostridia;o__Clostridiales;f__Clostridiaceae;g__Thermoanaerobacterium 20.6
c__Clostridia;o__Clostridiales;f__Ruminococcaceae;g__Ruminococcus 20.6
c__Clostridia;o__Clostridiales;f__Clostridiaceae;g__Clostridium 4.0
c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__Coprococcus 2.4
Co
ntr
ol
c__Clostridia;o__OPB54;f__;g__ 20.0
59.9
c__Clostridia;o__SHA-98;f__D2;g__ 11.2
c__Clostridia;o__;f__;g__ 11.0
c__Clostridia;o__Clostridiales;f__Clostridiaceae;g__Caloramator 9.1
c__Clostridia;o__SHA-98;f__;g__ 8.6
73
methanosarcina (2% in active acidic mesophilic reactors) and Methanobacteriaceae
methanobrevibacter (1.5% in active basic mesophilic reactors). In general, all acidic thermophilic
reactors, blanks from all conditions, and duckweed samples exhibited negligible fractions of
archaea.
Table 3-4: Relative abundance (R.A.) and Cumulative Abundance (C.A.) of top five genera in
each reactor group operated under basic conditions.
Co
nd
itio
n
Ty
pe
Taxa
R.A
.
(%)
C.A
.
(%)
Bas
ic M
eso
ph
ilic
Bla
nk
c__Clostridia;o__Clostridiales;f__Clostridiaceae;g__Alkaliphilus 14.1
48.8
c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__Coprococcus 13.0
c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__ 7.5
c__Bacteroidia;o__Bacteroidales;f__Porphyromonadaceae;g__ 7.3
c__Clostridia;o__MBA08;f__;g__ 6.9
Act
ive
c__Clostridia;o__MBA08;f__;g__ 23.9
66.7
c__Clostridia;o__Clostridiales;f__Ruminococcaceae;g__ 19.2
c__Bacilli;o__Bacillales;f__Bacillaceae;g__Natronobacillus 9.6
c__Bacteroidia;o__Bacteroidales;f__Porphyromonadaceae;g__ 8.6
c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__ 5.4
Co
ntr
ol
c__Clostridia;o__MBA08;f__;g__ 25.9
43.9
c__Bacilli;o__Bacillales;f__;g__ 6.2
c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__ 4.8
c__Clostridia;o__Clostridiales;f__;g__ 3.6
c__Clostridia;o__Clostridiales;f__Ruminococcaceae;g__ 3.5
Bas
ic T
her
mo
ph
ilic
Bla
nk
c__Clostridia;o__Clostridiales;f__Caldicoprobacteraceae;g__Caldicoprobacter 45.1
80.2
c__Clostridia;o__Clostridiales;f__[Tissierellaceae];g__Tepidimicrobium 17.6
c__Alphaproteobacteria;o__Rhodobacterales;f__Rhodobacteraceae;g__Rhodobacter 7.5
c__Alphaproteobacteria;o__Rhizobiales;f__Rhizobiaceae;g__Agrobacterium 6.8
c__Acidimicrobiia;o__Acidimicrobiales;f__C111;g__ 3.3
Act
ive
c__Clostridia;o__Clostridiales;f__Caldicoprobacteraceae;g__Caldicoprobacter 24.5
73.2
c__Clostridia;o__Halanaerobiales;f__Halanaerobiaceae;g__ 19.1
c__Clostridia;o__OPB54;f__;g__ 13.6
c__Clostridia;o__MBA08;f__;g__ 12.4
c_Methanobacteria;o_Methanobacteriales;f_Methanobacteriaceae;g__Methanothermo
bacter 3.6
Co
ntr
ol
c__Clostridia;o__MBA08;f__;g__ 10.8
33.7
c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__ 8.0
c__Bacilli;o__Bacillales;f__;g__ 6.6
c__Clostridia;o__Clostridiales;f__Ruminococcaceae;g__ 4.1
c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__Butyrivibrio 4.1
74
Table 3-5: Relative abundance (R.A.) and Cumulative Abundance (C.A.) of top five archaeal genera
in each reactor group. C
on
dit
io
n
Ty
pe
Taxa R.A.
(%) C.A. (%)
Aci
dic
M
eso
ph
ilic
Bla
nk
None None
Act
ive
c__Methanomicrobia;o__Methanosarcinales;f__Methanosarcinaceae;g__Methanos
arcina 2.03
3.03
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methano
brevibacter 0.83
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methano
bacterium 0.13
c__Thermoplasmata;o__E2;f__[Methanomassiliicoccaceae];g__vadinCA11 0.04
Co
ntr
ol
c__Thermoplasmata;o__E2;f__[Methanomassiliicoccaceae];g__vadinCA11 0.77
2.35
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methano
bacterium 0.69
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methano
brevibacter 0.45
c__Methanomicrobia;o__Methanosarcinales;f__Methanosarcinaceae;g__Methanos
arcina 0.19
c__Thaumarchaeota;o__Nitrososphaerales;f__Nitrososphaeraceae;g__Candidatus
Nitrososphaera 0.16
Aci
dic
T
her
mop
hil
ic
Bla
n
k
None < 0.1
Act
iv
e
c__Methanomicrobia;o__Methanosarcinales;f__Methanosarcinaceae;g__Methanos
arcina 0.03 < 0.1
Co
ntr
o
l
c__Methanomicrobia;o__Methanosarcinales;f__Methanosarcinaceae;g__Methanos
arcina 0.02 < 0.1
Bas
ic
Mes
op
hil
ic
Bla
nk
c__MCG;o__pGrfC26;f__;g__ 0.02 < 0.1
Act
ive
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methano
brevibacter 1.45
1.49 c__Thermoplasmata;o__E2;f__[Methanomassiliicoccaceae];g__Methanomassiliico
ccus 0.02
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methano
bacterium 0.01
Co
ntr
ol
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methano
brevibacter 0.35
0.73
c__Thermoplasmata;o__E2;f__[Methanomassiliicoccaceae];g__Methanomassiliico
ccus 0.14
c__Thaumarchaeota;o__Nitrososphaerales;f__Nitrososphaeraceae;g__Candidatus
Nitrososphaera 0.13
c__Methanomicrobia;o__Methanosarcinales;f__Methanosarcinaceae;g__Methanos
arcina 0.05
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methano
sphaera 0.0
75
Discussion
Effect of pH and temperature on acidogenic digestion performance
The experiments revealed high variations in VFA production potentials at different pH
and temperature values. The highest VFA yield observed 388 ± 28 mg VFA as HAceq g TSadded-1
(332 ± 24 mg VFA as HAceq g TSadded-1) under basic mesophilic conditions is similar to the
findings of a study conducted by Yuan et al. (Yuan et al., 2006) on acidogenic digestion of
activated wastewater sludge at pH 10 and ambient temperature. The authors reported 233 mg
VFA as HAceq g VS-1, attributing the high performance to the availability of soluble proteins
under these conditions. The superior performance achieved in our study might be due to the high
carbohydrate content of duckweed biomass, in addition to proteins. In parallel, basic mesophilic
conditions resulted in high acetic acid content (up to 83% of total VFAs). Apart from its effect on
Bas
ic
Th
erm
op
hil
ic
Bla
nk
None 0.00 < 0.1
Act
ive
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methanot
hermobacter 3.58
3.93
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methano
bacterium 0.28
c__Thermoplasmata;o__E2;f__[Methanomassiliicoccaceae];g__Methanomassiliico
ccus 0.02
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methano
brevibacter 0.02
c__Thaumarchaeota;o__Nitrososphaerales;f__Nitrososphaeraceae;g__Candidatus
Nitrososphaera 0.01
Co
ntr
ol
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methano
brevibacter 0.27
0.97
c__Thermoplasmata;o__E2;f__[Methanomassiliicoccaceae];g__Methanomassiliico
ccus 0.24
c__Methanobacteria;o__Methanobacteriales;f__Methanobacteriaceae;g__Methanot
hermobacter 0.15
c__Thaumarchaeota;o__Nitrososphaerales;f__Nitrososphaeraceae;g__Candidatus
Nitrososphaera 0.12
c__Methanomicrobia;o__Methanosarcinales;f__Methanosarcinaceae;g__Methanos
arcina 0.10
76
protein solubilization, high pH also has a chemical pretreatment effect on cellulosic and
hemicellulosic biomass, causing the release of acetyl groups, which could explain the high acetic
acid concentrations observed under these conditions. The same effect has also been reported by
other researchers for the acidogenic digestion of food waste at elevated pH (Dahiya et al., 2015).
The relative effects of biotic and abiotic conversion mechanisms on high acetic acid yields are
further discussed below in this section.
In addition, H2 recovery observed under acidic thermophilic conditions (up to 23.5 ± 0.5
ml g duckweed VSadded-1) was comparable to a study on swine wastewater-derived duckweed
(Lemna minor) mesophilic fermentation to biohydrogen, which resulted in 13 ml H2 g-1 dry
duckweed for non-pretreated biomass (Xu and Deshusses, 2015). The higher values observed in
our study could be due to thermophilic conditions. In the same study, the researchers reported up
to 42% H2 content, which was also in agreement with our findings of 33.1-43.8%. These results
are also within the range of specific H2 production potentials of materials characteristic of the
organic fraction of municipal solid waste, such as cabbage, carrot, and rice, reported as 19.3-96.0
ml H2 g VS-1 with 27.7-55.1% H2 content (Okamoto et al., 2000).
Overall, although the acetate produced under acidic mesophilic conditions was lost in the
form of CH4, the mesophilic reactors produced more VFAs than the thermophilic reactors in both
acidic and basic reactors, with and without inoculum supplementation. As observed for activated
sludge by Yu et al. (Yu et al., 2008), the present study with duckweed also found that pH has a
more significant impact than temperature on VFA production. Yu et al. attributed this observation
to enhanced substrate availability due to chemical hydrolysis under alkaline conditions at both
mesophilic and thermophilic temperatures (Yu et al., 2008). However, our observation might also
be due to the presence of alkaliphilic thermophiles originating from compost and the absence of
acidophilic thermophiles in the enriched inoculum mixture.
77
Effect of operating conditions on microbial community diversity and composition
Alpha diversity
Within each tested condition, blank reactors without inoculum were found to be less
diverse than active reactors, which were in turn less diverse than control reactors without
duckweed (Table 3-2). The lack of diversity in blank reactors is likely due to the fact that the sole
source of microbes in these reactors was from duckweed, which was harvested from an aerobic
environment. These aerobic microbes, introduced into an anaerobic environment, are not
expected to flourish. In general, the diversity of blank basic reactors (both mesophilic and
thermophilic) was similar to, but slightly lower than, the diversity of the duckweed microbes,
while acidic conditions (especially thermophilic) led to a decrease in the diversity in those blank
reactors. Controls had the highest alpha diversity within each treatment group and were generally
similar to the inoculum for acidic control reactors, but diversity slightly increased from the basic
inoculum to the basic controls. Since inoculum was the sole source of microbes in the control
reactors, it is reasonable that the diversity would be similar, but the reasons for the slight increase
in diversity observed in the basic controls is unclear. In active reactors, the decrease in diversity
from the inoculum (presumably the major source of microbes in active reactors) is reasonable
given the potential selective pressures of an active microbial community in the presence of
substrate (duckweed biomass). In general, diversity increased among the active reactors as
follows: acidic thermophilic << acidic mesophilic < basic mesophilic ≈ basic thermophilic. The
very low diversity in acidic thermophilic reactors is reasonable given the extreme conditions
present there. Low diversity has previously been noted for thermophilic cultures (Gaby et al.,
2017). Similar trends were observed for blank and control reactors across treatment groups with
78
respect to all diversity measures, except for acidic thermophilic controls, which suffered less
diversity loss in relation to acidic mesophilic conditions than their blank and active counterparts.
Beta diversity
Principle coordinate analysis (PCoA) using both abundance-weighted and unweighted
UniFrac distances showed reasonable clustering effects (Figure 3-5-A, B). All replicates clustered
closely together except blank basic mesophilic replicates, which were still reasonably associated.
PCoA of weighted UniFrac distance explained more of the variation (PC1-24.75% and PC2-
21.37%) compared to unweighted distances (PC1-18.37% and PC2-11.17%); however,
unweighted UniFrac PCoA clustered very clearly according to sample group. In the unweighted
PCoA plot, the most prominent clustering effect is by pH regime (PC2), with duckweed samples
clustering with all basic samples. PC1 appears to separate the samples based on sample type
(blank, active, control). Blank reactors without inoculum are clearly more similar to duckweed
samples, and control samples without duckweed cluster very tightly with the inoculum, which
was the only source of microbes in these reactors. Acidic thermophilic controls diverge
somewhat from the acidic inoculum. Comparing active reactors, it appears that temperature had a
greater effect on differentiating acidic reactors than basic reactors (degree of separation, PC2).
The same appears to be true for blanks.
The weighted PCoA plot still shows significant clustering by pH regime;
however, all acidic thermophilic reactors appear to cluster more closely with basic inoculum,
active basic mesophilic reactors, and basic controls. The reasons for this are unclear. The
weighted PCoA plot also shows a greater degree of separation between acidic mesophilic and
acidic thermophilic reactors than does the unweighted plot, and duckweed appears to be more
distinct from blank reactors on a weighted basis. In the literature it has been noted that qualitative
79
measures such as unweighted UniFrac distances better reveal the effect of different founding
populations and the ability of microbes to survive under different conditions, while quantitative
measures (weighted UniFrac) better show the effect of transient factors (e.g. nutrient availability)
(Lozupone et al., 2007). Here, the weighted PCoA analysis does not seem to reflect the various
VFA profiles as well as the unweighted PCoA. Statistical analysis of sample groupings (acidic vs.
basic, and mesophilic vs. thermophilic) confirmed the significance of these groupings. Analysis
based on weighted UniFrac distances revealed statistical significance for both pH and temperature
groupings (p-value 0.001); however, the test statistic for the pH grouping was slightly higher
(7.01 vs. 6.11), indicating a stronger effect. Grouping by temperature was significant on the basis
of unweighted UniFrac distances as well, but to a lesser degree than with weighted distances (p-
value 0.009; test statistic 1.94), while pH grouping was deemed to be very significant under both
measures (unweighted p-value 0.001; test statistic 3.81). These results back-up the clustering
observed in the PCoA plots and indicate that pH had a stronger effect in determining the
microbial community composition (both qualitatively and quantitatively) than did temperature,
for the conditions tested.
80
Figure 3-5: A) Weighted and B) unweighted PCoA plots.
A)
B)
81
Composition
Analysis of differential abundance at the genus level (acidic vs. basic and mesophilic vs.
thermophilic) was performed using reactor samples only (i.e. no inoculum or duckweed samples)
and revealed a greater number of differentially abundant taxa across pH regimes than across
temperature regimes (71 vs. 22 based on FDR-corrected p-values < 0.05).
Of the differentially abundant taxa across temperature regimes, only five were enriched
in thermophilic reactors and all were members of the phylum Firmicutes. These included
Clostridiaceae thermoanaerobacterium, Tissierellaceae tepidimicrobium, Planococcaceae
lysinibacillus, and Thermoanaerobacterales thermovenabulum, along with unidentified members
of the order OPB54. The percent difference in the mean counts of these genera between the two
conditions exceeded 80% in each case (average 95%), indicating that temperature was strongly
selective for these microbes. On the other hand, 17 of the genera with temperature dependent
differential abundance were enriched in mesophilic reactors. Among those with the largest
increase in observed counts under mesophilic conditions were Prevotellaceae prevotella,
unidentified genera in the families Enterobacteriaceae and Porphyromonadaceae, and
unidentified members of the order Bacteroidales (percent difference in mean counts > 99%). Only
one member of the kingdom archaea was differentially abundant across temperature regimes
(Methanobacteriaceae methanobrevibacter), preferring mesophilic conditions.
Differentially abundant taxa across pH regimes are too numerous to detail, but some key
taxa that support the validity of the differential abundance analysis include the enrichment of
Veillonellaceae acidaminococcus in acidic samples, and Clostridiaceae alkaliphilus in the basic
samples. In fact, nearly half of the 14 genera enriched in the acidic samples belong to the family
Veillonellaceae. Others belong mostly to the class Clostridia, with two examples from the class
Bacteroidetes. The remaining 57 genera fared better under basic conditions. Some were only
82
moderately enriched under basic conditions (e.g. unidentified genus in the family
Lachnospiraceae; 71% difference in mean counts), while others were completely absent from
acidic reactors - Bacillaceae natronobacillus, Clostridiaceae natronincola_anaerovirgula, and
Bacteroidaceae bacteroides for example. Overall, phyla enriched in basic reactors were more
diverse, including Firmicutes, Bacteroidetes, Tenericutes, Actinobacteria, Cyanobacteria and
Proteobacteria. Only two genera of archaea were found be differentially abundant across pH
regimes, both preferring basic conditions - Methanomassiliicoccaceae methanomassiliicoccus
and Methanobacteriaceae methanothermobacter.
Relationships between operating conditions, microbial community structure, and end
products
The differences in operational parameters of the reactors provided unique environments
which led to distinct microbial communities and the production of different end products under
each condition (Figures 1-4). The effects of pH and temperature on microbial populations and end
product profiles during acidogenic digestion of duckweed have been summarized in Table 3-6.
For example, the genus Acidaminococcus, a mesophilic anaerobic gram-negative cocci which can
ferment amino acids (Shetty et al., 2013), was observed only in acidic mesophilic reactors.
Thermoanaerobacterium, a genus with members which can degrade starch, cellulose, and sucrose
for H2 production, favors slightly acidic conditions (Prasertsan and O-thong, 2009), and was
observed here as one of the most dominant genera under acidic thermophilic conditions. In
contrast, basic conditions were dominated by cultures originating from alkaline environments.
For instance, Natronobacillus, a genus of alkaliphile anaerobic species with the capability to fix
nitrogen (Sorokin et al., 2008), was identified in the basic mesophilic reactors. The negative gas
pressure reported in these reactors (Appendix B) may have been caused by the fixation of the
83
nitrogen gas by these organisms. Another family of bacteria which was abundant in basic
mesophilic reactors, Porphyromonadaceae, have been previously isolated from mesophilic
anaerobic reactors (Müller et al., 2016). In addition, some uncultivated bacterial lineages such as
MBA08 [Clostridia] and OPB54 [Clostridia] which were previously detected in anaerobic
digesters, were present in the basic reactors tested here. Tepidimicrobium, a xylanolytic genus
with thermophilic and alkali-tolerant members (Niu et al., 2009) was detected under basic
thermophilic conditions, along with Halanaerobiacea, a thermophilic genus found in agricultural
biogas plants (Maus et al., 2017). Some genera, such as Coprococcus, Ethanoligenens, and
Clostridium were observed under both acidic and basic conditions.
The high acetic acid yields observed under basic conditions were very likely augmented
by homoacetogenesis. The presence of hydrolytic and fermentative taxa such as the families
Porphyromonadaceae (Müller et al., 2016) and Ruminococcaceae (Sträuber et al., 2012), and the
genera Prevotella (Hung et al., 2011), and Caldiocoprobacter (Müller et al., 2016), might have
theoretically resulted in the production H2 and CO2. In contrast, the biogas recovery observed was
negligible (24 ml/g duckweed VSadded), which suggests that the produced H2 and CO2 might be
converted to acetate by homoacetogenic bacteria. While it is not possible to positively identify
homoacetogenic species given the resolution of the current data set, taxonomic groups which are
known to contain homoacetogens were abundant in the basic thermophilic reactors. These include
the genus Clostridium (2.8% relative abundance), within which thermophilic homoacetogenic
species have been identified in the literature (e.g. C. thermoaceticum and C.
thermoautotrophicum), and the order Thermoanaerobacterales (4.5% relative abundance), which
is known to encompass thermophilic homoacetogens of the genus Thermoanaerobacter (e.g. T.
kivui) (Ljungdahl, 1986; Onyenwoke and Wiegel, 2015). The negative headspace pressures
recorded in both mesophilic and thermophilic reactors at pH 9 also support this conclusion
(Appendix B), and indicate that homoacetogens are not inhibited at pH 9. However, an evolution
84
of CH4 was also observed in the headspace, especially after Day 5, where no biogas was
recovered, but rather the headspace H2 and CO2 contents decreased. In both basic mesophilic and
basic thermophilic reactors, hydrogenotrophic methanogenesis was observed, potentially due to
the activity of genera such as Methanobrevibacter, which was also reported by Gaby et el. (Gaby
et al., 2017) in anaerobic digesters fed with food waste. However, the absence of acetotrophic
genera such as Methanosarcina, along with high acetate concentrations, show that at pH 9,
neither 35oC or 55oC favored acetoclastic methanogens.
Although it was previously reported that methanogenic activity could be inhibited under
pH 6 (Luo et al., 2011), the reduction of acetate and generation of CH4 under mesophilic
conditions here revealed acetoclastic methanogenic activity, which is likely related to the
presence of Methanosarcina sp. Methanosarcina are capable of both acetoclastic and
hydrogenotrophic methanogenesis (Table 3-4), so this finding could also explain the absence of
H2 in the headspace. This outcome might be a result of effective solids reduction and hydrolysis
(Figure 3-3-A), leading to high protein degradation and subsequent ammonium release, which
provided local pH increases and buffering capacity, thereby creating a suitable environment for
methanogens. In fact, others have similarly reported that co-fermentation of food waste and
excess sludge provided favorable conditions for high solubilization, leading to higher ammonia
concentrations and slight VFA loss to CH4 during acidogenic digestion (Wu et al., 2016). The
present results also indicate that methanogenic activity could not be permanently inhibited by
heat pretreatment, similar to the findings of Luo et al (Luo et al., 2011). However, the biomethane
recovery observed in this study (Figure 3-2-A, 26.6 ± 3.8 ml g duckweed VSadded-1) was not
comparable to biochemical methane potential studies of raw duckweed reported in the literature,
which were 158 ml g VSadded-1 from Lemna minor (S.K. Jain et al., 1992) and 259 ml g VSadded
-1
(Calicioglu and Brennan, 2018) from Lemna obscura. This could be primarily because of the ten-
fold higher substrate-to-inoculum ratio provided in the present study, which may have caused
85
simultaneous substrate inhibition due to ammonia and VFA accumulation (Ozgul Calicioglu and
Demirer, 2017). These conditions might have led to an inhibited state at which the process was
stable, but yielded lower CH4 (Chen et al., 2008). In fact, the free ammonia concentrations
reported in this study (Appendix B) have been previously reported to have potential inhibitory
effects (Yenigün and Demirel, 2013). In addition, the higher CO2 recovery observed in this study
could be due to the activity of syntrophic bacterial populations (producing CO2 and H2 from
acetate), such as some members of Coprococcus and Clostridium (Esquivel-Elizondo et al.,
2016), which also might have acted as a sink for acetate.
Table 3-6: Summary of microbial populations and end product profiles under various operating
conditions.
Conditions Key findings
Acidic
Mesophilic
Susceptible to VFA loss due to acetoclastic methanogenic activity (Methanosarcina, 2.03%).
High biogas-CO2 content, suggesting fast hydrolysis, resulting in TAN release.
Low CH4 yield (26.6 ± 3.8 ml g duckweed VSadded-1) compared to literature, likely because the
very high ammonium concentrations required as buffer were inhibitory.
Acidic
Thermophilic
H2 recovery up to 23.5 ± 0.5 ml g-1 duckweed solids added.
Least diverse microbial communities (α diversity).
Acetate and butyrate were predominant VFA species.
Basic
Mesophilic
Highest VFA yields (388 ± 28 mg VFA as HAceq g VSadded-1).
Competition between homoacetogenesis and hydrogenotrophic methanogenesis over H2.
Low biogas recovery (23.7 ± 6.2 ml g duckweed VSadded-1) compared to literature, suggesting
presence of internal sinks for headspace H2 and CO2.
Basic
Thermophilic
Highest final particulate matter formation (18.6 % of initial total carbon) in the absence of
inoculum, suggesting chemical (alkaline) pretreatment augmented VFA production.
Low biogas recovery (29.7 ± 6.3 ml g duckweed VSadded-1), suggesting presence of internal
sinks for headspace H2 and CO2.
Overall
conclusions:
Within 9 days, more than 80% of the final day VFA concentrations were achieved.
Species richness (α diversity) was higher in basic reactors.
pH has a more significant impact than temperature on both the composition of microbial
communities (β diversity) and VFA production.
86
In contrast to acidic mesophilic conditions, acidic thermophilic conditions may have
inhibited methanogenic activity, potentially due to lower solids solubilization efficiency (Figure
3-3-B). This may be why lower ammonia concentrations were observed under acidic
thermophilic conditions (Appendix B), and local increases in pH were not favored. This is
consistent with the literature: methanogenic activity is known to be more easily suppressed under
thermophilic conditions in the presence of lower ammonia concentrations (Chen et al., 2008).
Furthermore, the acidic thermophilic reactors were heavily dominated with genera containing H2-
forming members such as Ethanoligenens (Tang et al., 2012) and Clostridium (Collet et al.,
2004); as well as Ruminococcus (Tian et al., 2014) and Thermoanaerobacterium (Prasertsan and
O-thong, 2009), which include sugar fermenting thermophiles that can produce acetate and
butyrate (Figure 3-1-B).
The activity originating from the microorganisms associated with duckweed may have
significantly affected solubilization of the biomass and its conversion into VFAs (Figure 3-3), as
VFA production was also observed in blank reactors containing duckweed to which no inoculum
was provided. This suggests that the biofilm present on duckweed can serve as suitable
environment for anaerobic microorganisms. The VFA production was higher in mesophilic blank
reactors, compared to those operated under thermophilic conditions. This might be because the
mesophilic operating conditions are closer to the natural habitat of duckweed. Under acidic
thermophilic conditions, the blanks were dominated by spore-forming bacteria, which might have
survived in the natural habitat of duckweed biofilm until favorable conditions prevailed. In most
cases, addition of inoculum resulted in higher reactor performance in terms of solubilization and
VFA production. Only in acidic thermophilic reactors was better solubilization efficiency
observed for blanks (with no inoculum); however, the VFA yields were still slightly lower than in
the actives. The lowest VFA production performance was observed in blank reactors under basic
thermophilic conditions. This suggests that under basic conditions, VFA production was mainly
87
carried out through biotic processes, rather than as an effect of chemical pretreatment releasing
acetyl groups from hemicellulose (Kumar et al., 2009), as has been previously observed during
alkaline pretreatment of cellulosic biomass (Hendriks and Zeeman, 2009). However, another
interesting point to note in the thermophilic blank reactors is the increase in the particulate matter
fraction. The particulates were only evident in basic blank reactors, and were markedly higher in
concentration under thermophilic conditions. This may imply that the basic conditions augmented
acetate production by a chemical pretreatment effect, which increased the efficiency of hydrolysis
and in turn increased the bioavailability of the biomass for microbial conversion. Overall, the
results indicate that the enhanced VFA production observed under basic conditions was an
outcome of a synergy between chemical pretreatment and biological activity.
Conclusions
This study demonstrated that 33.2 ± 2.4% of duckweed biomass can be converted into
VFAs with a mixed culture microbiome under basic mesophilic conditions. The superior
performance observed under these conditions was attributed to both chemical treatment and
microbial bioconversion. Final yield and composition of the VFAs primarily depended on the pH
and much less on the temperature of the reactors. The composition of the microbial community
under these different conditions was also affected more by pH than temperature, with temperature
effects enhanced under acidic conditions as compared to basic conditions. Depending on the end
product of interest, pH can be adjusted either to produce longer chain VFAs and H2 (under acidic
conditions), or to maximize total VFA yields (under basic conditions). VFAs can be further
processed into medium chain fatty acids, which are building blocks for high-value advanced
biofuels. Avoidance of the pH window which favors methanogenic activity during acidogenic
88
digestion would enable downstream processing of carboxylic acid production residuals through
methanogenic anaerobic digestion to maximize energy recovery.
These results indicate that duckweed is a technically feasible alternative feedstock for the
production of advanced biofuel precursors. In addition, the residual biomass from the VFA
production process could be valorized through conversion into biogas and biosolids. To more
completely access the feasibility of this process, studies on the conversion of duckweed into
multiple end products in a complete biorefinery system are necessary.
Acknowledgement: This research was generously funded by a seed grant from the Penn
State Institutes of the Energy and Environment. This project was also supported by Agriculture
and Food Research Initiative Competitive Grant No. 2012-68005-19703 from the USDA National
Institute of Food and Agriculture. The findings do not necessarily reflect the view of the funding
agencies.
89
Chapter 4
Anaerobic Bioprocessing of Wastewater-Derived Duckweed:
Maximizing Product Yields in a Biorefinery Value Cascade
This manuscript is in review as follows:
Calicioglu, O., Richard, T. L., and Brennan, R. A. Anaerobic “Bioprocessing of
Wastewater-Derived Duckweed: Maximizing Product Yields in a Biorefinery Value Cascade”,
Bioresource Technology, 2019.
Abstract
In this work, the potential for integrating sugar and carboxylate biochemical conversion
platforms was investigated to enhance feedstock bioconversion in biorefinery systems. Two or
three anaerobic bioprocesses (bioethanol fermentation, acidogenic digestion for volatile fatty acid
(VFA) production, and methanogenic digestion for biomethane production) were sequentially
integrated to maximize the carbon-to-carbon conversion of wastewater-derived duckweed
biomass into bioproducts. Duckweed was fed to reactors raw (dried) after liquid hot water
pretreatment or enzymatic saccharification. At the end of each bioprocess, the target bioproduct
(i.e., bioethanol, VFAs, or methane) was separated from the reactor liquor (i.e., by vacuum
extraction of ethanol, or membrane separation of VFAs) and the remaining reactor components
were subjected to further anaerobic bioprocesses. The highest total bioproduct carbon yield of
0.69±0.07 grams per gram of duckweed carbon was obtained by sequential acidogenic and
methanogenic digestion. Nearly as high yields were achieved when three bioprocesses were
integrated sequentially (0.66±0.08 grams of bioproduct carbon per gram duckweed carbon). For
90
this three-stage value cascade, yields of each process in conventional single-stage units were: 1)
0.186±0.001 grams ethanol per gram duckweed dry matter; 2) 611±64 mg acetic acid equivalent
of volatile fatty acids per gram of volatile solids; and 3) 434±0.2 ml methane per gram of volatile
solids.
Introduction
Modern economies utilize renewable resources to fulfill only a minor fraction of their
total energy and chemical demands, and rely instead on nonrenewable resources such as coal,
crude oil, and gas (Hatti-Kaul et al, 2007). However, the economic and environmental
disadvantages of fossil fuels have led to increased efforts to find alternative resources to fulfill
energy and chemical needs (Jung et al., 2016). Among the alternatives, biomass is the only
renewable resource for chemicals. In order to utilize biomass as an alternative to fossil-based raw
materials, it must be processed in integrated, complex biorefineries, analogous to petroleum
refineries, by targeting an array of end products with different market values, chemical properties,
and quantities (Biddy et al., 2016).
Although current biorefineries generally target ethanol or other liquid biofuels as the
primary end product, methanogenic (anaerobic) digestion (MAD)of fermentation residues is a
common practice in order to improve both environmental and economic performance of ethanol
production processes (Bondesson et al., 2013; Dererie et al., 2011). However, these residues
could also be processed into higher-value compounds. One alternative pathway suitable for
establishing such a product value cascade is the carboxylate platform, which utilizes mixed
cultures for acidogenic anaerobic digestion (AAD) of organic matter with carboxylic acids (i.e.
volatile fatty acids, VFAs), as products and/or precursors of higher-value chemicals and biofuels
such as esters, alcohols, and alkanes (Agler et al., 2011). Although chemical inhibition of
91
methanogenic activity is often performed to ensure the stability of the carboxylate platform,
another inhibition method is to operate acidogenic digestion process at high pH (9-10) values,
which in turn gives higher VFA yields (Calicioglu et al., 2018) at short residence times of up to
ten days. This inhibition technique also allows remaining residues to be bioprocessed into
methane further, if the high pH control is stopped to drop the pH to neutral. Under this scenario, a
sequential biorefinery process train with a value cascade of end products, integrating the sugar
and carboxylate platforms, would sequentially produce ethanol, VFAs, and methane.
The overall yield of biomass-to- co-products has been reported in the literature as the
cumulative energy content of the co-products (Bondesson, 2008; Wu et al., 2015). However, for a
biochemical biorefinery that includes co-products sold into other market segments, a mass
approach for calculating the actual process yields as a function of theoretical potential might be
more suitable. In this study we consider the carbon-to-carbon conversion of a feedstock into
bioproducts, which not only provides a common set of units for system input and outputs, but
also reveals how the atmospheric carbon sequestered in biomass “fractionates” among the
bioproducts in the output portfolio.
Biomass composition and availability is of particular importance for providing a reliable
feedstock for biorefining, along with its social acceptance and environmental performance for
long term sustainability. Given that renewable alternatives should ideally be abundant,
inexpensive, and complement rather than compete with food production, there is a preference for
non-edible plant-based raw materials (biomass) as a feedstock for biroefineries (Cherubini, 2010).
One alternative feedstock which fulfills these criteria is duckweed (Lemnaceae), a family of fast-
growing, simple, floating aquatic plants, consisting of 38 species in five genera (Les et al., 2002).
Duckweeds can accumulate high amounts of starch (up to 46% of dry mass) under nutrient
starvation (Zhao et al., 2015). In addition, due to their relatively low lignin content (1%-3%),
duckweeds do not require harsh chemical pretreatments prior to processing. Because they float,
92
duckweeds are easy to harvest, and their small dimensions (0.1 cm to 1 cm in the largest
dimension) eliminates the need for size reduction (Cui and Cheng, 2015). Furthermore,
duckweeds are resilient to a broad range of nutrient concentrations; therefore, they can be grown
on wastewater steams (Cheng and Stomp, 2009) and require minimal agricultural inputs. The
advantages duckweed possesses as a feedstock has encouraged several prior research studies,
focusing on three platforms of biorefineries: (1) thermochemical conversion into syngas, as well
as gasoline, diesel, and jet fuel (Baliban et al., 2013); (2) sugar platform conversion into alcohols
(Ge et al., 2012; Zhao et al., 2012; Su et al., 2014); and (3) carboxylate platform conversion into
VFAs (Calicioglu et al., 2018).
It is known that valorizing process residues from fermentation effluents is technically
feasible for duckweed (Calicioglu and Brennan, 2018). However, an integrated biorefinery value
cascade has not previously been investigated for this renewable feedstock. The integration of
various anaerobic bioprocesses involving sugar and carboxylate platforms might be particularly
advantageous in for nutrient rich feedstocks like duckweed. A feedstock high in nutrients reduces
the need to import and supplement nutrients for various fermentation processes, and any excess
nutrients remaining after anaerobic bioprocesses are complete could also be valorized as one of
the end products.
This study utilizes wastewater-derived duckweed as a model biomass feedstock to
investigate the potential for the sequencing of anaerobic bioprocesses (i.e., ethanol fermentation,
acidogenic digestion, and methanogenic digestion) in an integrated biorefinery system. The aim
of the study is to determine the optimum combination of the number and type of bioprocesses for
the maximum carbon-to-carbon conversion efficiency, while producing fertilizers as a side
product.
93
Materials and Methods
Analytical methods
The moisture, total solids (TS), and volatile solids (VS) contents were determined
according to the National Renewable Energy Laboratory (NREL) Laboratory Analytical
Procedure (LAP) for biomass and total dissolved solids of liquid process samples (Sluiter et al.,
2008). Ash content was measured according to NREL LAP for determination of ash in biomass
(Sluiter et al., 2004).
Glucose and ethanol quantification were performed using a Waters high performance
liquid chromatograph (HPLC) equipped with a refractive index detector (Waters, Milford, MA)
and a Bio-Rad Aminex HPX-87H column (300 mm × 7.8 mm; Bio-Rad, Richmond, CA) with 0.8
ml/min of 0.012 N sulfuric acid as the mobile phase. The detector and column temperatures were
constant at 35 °C and 65 °C, respectively. Prior to HPLC analysis, samples were centrifuged at
4°C for 20 min at 5,200 x g and the supernatant filtered through 0.2 μm nylon syringe filters.
VFAs were quantified using Gas Chromatography (GC) (SHIMADZU, GC-2010 Plus,
Japan) with a flame ionization detector. The final total VFA yields were calculated in terms of
acetic acid equivalents per gram of duckweed volatile solids added (HAceq g VSduckweed-1)
(Siedlecka et al., 2008), and as grams of carbon in VFAs per gram of total carbon in duckweed
added (g VFA-C g TCduckweed-1).
Carbon quantification of liquid and solid samples were performed using a total carbon
(TC) analyzer (SHIMADZU, TOC-V CSN, Kyoto, Japan) equipped with solid sample module
(SHIMADZU, 5000A, Kyoto, Japan).
Headspace gas pressures in acidogenic and methanogenic reactors were measured using a
pressure gauge (Grainger, DPGA-05, USA). If pressures were positive, volumes of gas
94
production from the acidogenic and methanogenic reactors were measured at ambient
temperature using a water displacement device. Carbon dioxide gas production from the ethanol
fermentation reactors was also measured to complete the carbon balance. The device was filled
with 0.01 M hydrochloric acid solution to prevent microbial growth. The headspace temperature
was assumed to be constant and equal to 35oC during the measurement, due to the rapid sampling
process (El-Mashad, 2013; Theodorou et al., 1994). Volume readings were reported at standard
temperature and pressure. Volumetric methane (CH4) and hydrogen (H2) concentrations were
determined by collecting headspace gaseous samples using a 250 μl airtight syringe (Hamilton,
Reno, NV, USA) and injecting onto a GC (SRI Instruments, SRI310C, Torrance, CA, USA)
equipped with 6-foot molecular sieve column (SRI 8600-PK2B, USA), operated in continuous
mode at 80oC with argon as the carrier gas. Volumetric carbon dioxide (CO2) concentrations were
quantified using an identical GC (SRI Instruments SRI310C) equipped with 3-foot silica gel
packed column (SRI, 8600-PK1A, USA) in continuous mode at 60oC with helium as the carrier
gas. Carbon loss in reactors in the form of CO2 was expressed as grams of carbon in CO2 per
gram of TC in duckweed added (g CO2-C g TCduckweed-1).
The raw and processed duckweed were analyzed at the Penn State Agricultural Analytical
Services Laboratory for fertilizer potential. Total ammonia nitrogen (TAN) was determined by
specific ion electrode method. Total nitrogen was quantified by combustion. Total phosphorus
and total potassium were quantified by microwave-assisted acid digestion method (Peters, 2003).
All fertilizer tests were performed in duplicate.
Plant material, cultivation, and pre-processing
Duckweed (Lemna obscura, 100% sequence identity to accession number GU454331.1,
in the NCBI database (Calicioglu and Brennan, 2018) was collected on September 20, 2016, from
95
an open pond within the effluent spray fields of the Penn State Wastewater Treatment Plant, also
known as the “Living-Filter”. In August and September, the pond received on average (n = 9): 1.7
± 0.5 mg L-1 carbonaceous biological oxygen demand; 2.2 ± 0.2 mg L -1 phosphorus; 0.2 ± 0.0
mg L-1 TAN; 15.1 ± 1.5 mg L-1 nitrate; 1.2 ± 0.1 mg L-1 nitrite; and 1.3 ± 0.4 mg L-1 total
Kjeldahl nitrogen. The average water quality of three grab samples obtained from the surface of
the pond at the harvest day was reported as: 35.1 ± 1.3 mg L-1 total chemical oxygen demand,
17.5 ± 0.7 mg L-1 soluble chemical oxygen demand, 2.0 ± 0.3 mg L-1 TAN, 2.5 ± 1.3 mg L-1
nitrate and 1.5 ± 0.6 mg L-1 phosphate. After harvest, the duckweed was wet sieved with tap
water to remove smaller and coarser impurities, dried at 45 ± 3oC to a constant weight over two
days, and analyzed for its moisture ( 6.9 ± 1.3% wet basis), and VS (85.8 ± 1.2% of TS) contents.
The composition of duckweed was determined by wet chemistry analyses as (all values on a %
dry weight basis): cellulose (12.6 ± 0.2); hemicellulose (21.0 ± 0.5); starch (10.8 ± 0.1); lignin
(0.8 ± 0.2); water soluble carbohydrates (20.1 ± 0.1); and crude protein (18.3 ± 0.3) (Dairy One
Wet Chemistry Laboratory, Ithaca, NY).
Enzymatic liquefaction and saccharification of the duckweed was performed in four 2-L
flasks with 1 L total working volume. Prior to liquefaction, 50 g duckweed (dry weight basis),
equivalent to 20.3 ± 0.15 g TC, was sterilized by autoclaving with 945 ml water for 30 minutes at
121oC. Then the pH was adjusted to 7.0 ± 0.1 with 2 M hydrochloric acid. Once the slurry was
cooled to 90oC, α–amylase (Sigma Aldrich, A3403, USA) was added at a loading of 5000 units g
starch-1. The flasks were incubated for one hour at 90oC for liquefaction. Following liquefaction,
the pH was adjusted to 5.2 ± 0.1 with sodium citrate buffer, yielding 25 mM in the total working
volume. After pH adjustment, 334 units of glucoamylase g starch-1 (Sigma Aldrich, 10115, USA)
and cellulase (Novozymes, Cellic® CTec2, Denmark) with 60 filter paper unit g cellulose-1
loadings were added to each flask, and then sealed with rubber stoppers. Saccharification was
then performed at 50°C, while mixing at 120 rpm for 24 h. All experiments and sampling were
96
conducted under sterile conditions. Glucose and ethanol concentrations were quantified before
and after liquefaction, and after saccharification. The theoretical maximum glucose
concentrations of glucose and starch components of duckweed was calculated according to Gulati
et al. (1996), and the water soluble sugar content of the duckweed was considered as glucose (i.e.
fermentable by Saccharomyces cerevisiae) for a conservative estimate of the maximum
theoretical glucose yield. Saccharified duckweed was utilized in individual fermentation,
acidogenic digestion, and methanogenic digestion processes, or in the first stage of sequential
processes of the value cascade.
Liquid hot water pretreatment was carried out in a 500 ml stainless steel Parr reactor
(Parr Instrument Company, model 4575, Moline, IL), with a pressure limit of 345 bar. The vessel
was filled with 30 g duckweed (dry weight) and 270 g distilled water. The temperature was
ramped up to 150oC within 15 minutes, followed by pressurization with nitrogen gas for 5
minutes which was monitored using a digital pressure transducer (Tasker et al., 2016).
Inocula
Yeast strain
The yeast, Saccharomyces cerevisiae (ATCC 24859), was enriched in basal medium
containing (g L-1): glucose (20); yeast extract (Difco, Sparks, MD) (6); CaCl2·2H2O (0.3);
(NH4)2SO2 (4); MgSO4·7H2O (1); and KH2PO4 (1.5). The culture was grown at 30 °C for 24 h,
centrifuged at 2880 relative centrifugal force (rcf) for 20 minutes (Eppendorf, 5804 R, Germany),
and the pellet refrigerated for less than two hours before being used to inoculate the fermentation
reactors.
97
Acidogenic anaerobic seed
A mixture of silage, rumen fluid, anaerobic wastewater sludge, and compost was used as
acidogenic seed. Silage and rumen fluid were obtained from The Pennsylvania State University
Dairy Farm. Anaerobic wastewater sludge was obtained from the Pennsylvania State University
Wastewater Treatment Plant’s secondary anaerobic digester. Compost was obtained from the
Pennsylvania State University composting facility. All sources were mixed and acclimated to
basic conditions (pH 9.2) as described in detail previously (Calicioglu et al., 2018). The VS
content of the final acidogenic inoculum was 52.2 ± 1.1% of the TS, and the moisture content was
84.2 ± 0.5%.
Methanogenic anaerobic seed
Methanogenic seed was obtained from the Penn State Wastewater Treatment Plant
secondary anaerobic digester. The inoculum was starved for two days prior to use in the
biochemical methane potential (BMP) assays. The final composition of the starved methanogenic
seed was: 98.0 ± 0.0% moisture, and 75.1 ± 3.2% VS of TS.
Anaerobic bioprocessing scenarios in a biorefinery system
Raw, pretreated, and saccharified duckweed were anaerobically processed into a value
cascade of end products (i.e. ethanol, VFAs, and methane, respectively) through two or three
sequential anaerobic bioprocesses. The single end product yields of individual processes were
also quantified. The potential of producing fertilizer as a side product from the final residuals
was evaluated. After each step, the desired end product was recovered from the process liquids,
and the residues were further processed.
98
Total carbon content, rather than VS, was used as a basis of reactor dosing for various
substrates in anaerobic bioprocesses due to the following advantages: (1) duckweed sequesters
atmospheric carbon, and therefore determining the fate of the carbon through bioprocesses is
important; (2) VS determination for process residues high in ethanol and VFAs (i.e. stillage and
acidogenic digestate) can be inaccurate since these volatile compounds are underestimated during
the determination of solids content (Vahlberg et al., 2013); (3) calculating VS equivalence of
methane as an end product is not practical while constructing material balances, and therefore TC
provides a uniformly applicable platform for comparison; (4) inorganic carbon can also be
consumed and converted to other forms during acidogenic and methanogenic digestion. The
carbon to VS ratio for the raw duckweed and other substrates was calculated and used to dose the
same amount of carbon in the feedstock for each unit operation; namely, pre-processed duckweed
for single processes or the initial stage of a cascade, or the residues of the upstream anaerobic
bioprocesses for subsequent stages of a cascade. The solid and liquid residues of each process
were carried to the next process, keeping the same ratio of solids to liquids for subsequent stages.
Details on the substrates used (i.e. the type of pre-processed duckweed), operation of the
bioreactors, end product separation for each anaerobic bioprocess, as well as the overall product
yield calculations for the value cascades, are provided in the following sections.
Ethanol fermentation and distillation
Only saccharified duckweed was subjected to fermentation; the raw and pretreated
duckweed were excluded from the assay, since they lack the monosaccharides that are
fermentable by standard yeast. Following saccharification, a 0.8 g yeast pellet (dry weight) was
added to each fermentation flask, which was then incubated at 32 °C while mixing at 120 rpm for
24 h. The produced gas was vented out from an outlet through the rubber stopper, and its carbon
99
dioxide was captured in 10 M sodium hydroxide solution. Glucose and ethanol concentrations at
0 h, 12 h, and 24 h were quantified. Ethanol yields were expressed as g ethanol g glucose
recovered-1, and g ethanol g TSduckweed-1. In order to compare the ethanol yields to those of other
products, grams of carbon in ethanol per gram of TC in duckweed added (g ethanol-C g
TCduckweed-1) was also calculated.
The constituents of the fermentation flasks were then combined and transferred to a
vacuum evaporation setup. In order to keep the volatile fatty acids in the stillage, the pH was
increased to 7.8 ± 0.1 by 5 M sodium hydroxide addition. The ethanol distillation was performed
by keeping the slurry temperature at 80oC. After distillation, a portion of the stillage was tested
for fertilizer potential. The remaining stillage was subjected to acidogenic digestion or BMP
assays.
Acidogenic anaerobic digestion and membrane separation
Batch reactors (300 ml working volume) were fed with raw, pretreated, or saccharified
duckweed, or with fermentation residues to achieve a total substrate carbon loading of 10.1 ± 0.1
g L-1, which is equivalent to the carbon loading of 25 g L-1 raw duckweed. The VS variation
between reactors was less than 18%. The inoculum was added at a substrate-to-inoculum ratio of
10:1 on a VS basis calculated for raw duckweed. Initial pH values were adjusted to 9.2 after the
reactors were supplemented with 4.0 g L-1 sodium carbonate as a buffer, which is equivalent to
about 5% of the duckweed carbon input and was quantified in the carbon balance accordingly. All
reactors were purged with nitrogen gas and sealed with rubber stoppers and aluminum crimps.
Reactors were operated under mesophilic (35oC) conditions for 10 days. Once every two days,
headspace gas volume was quantified, liquid samples were collected, and the pH was adjusted to
9.2. Test reactors were run in triplicate, and controls (with no substrate) were run in duplicate.
100
The VFA production in the control reactors was found to be negligible compared to that achieved
in the active reactors, and therefore, were not subtracted.
Following acidogenic digestion, the digestates were centrifuged at 2880 rcf for 30
minutes (Eppendorf, 5804 R, Germany). The supernatants were filtered through a 0.2 µm nylon
filter and their pH values were adjusted to 4.0 using 5 M and 1 M hydrochloric acid prior to
membrane separation of the VFAs. Pellets were saved to be combined with the reactor liquids
following membrane separation.
Nanofiltration of the digestates were performed as described by Xiong et al. (2015) in a
200-ml dead-end nanofiltration vessel (Amicons, Stirreed Cell 8200, USA) at ambient
temperature, using a thin film membrane (GE Osmotics, DL, USA) with an effective filtration
area of 28.7 cm2. The vessel was pressurized to 0.5 MPa using nitrogen gas. At the beginning of
each filtration process, membranes were flushed with deionized water for 30 min. Approximately
70 ml of each digestate was added to the continuously-stirred vessel. Once 70% of the original
digestate volume was collected as permeate, the same amount of deionized water was added to
the vessel and re-collected, again equaling 70% of the original volume. The recovery efficiency
was calculated through a VFA balance over retentate, first permeate, and second permeate, on a
VFA carbon basis. The retentate volume was made up to its original value of approximately 70
ml by deionized water, and was mixed back with the pellets, to be used for BMP and fertilizer
assays.
Biochemical methane potential (BMP) assays
The BMP assays with duckweed were carried out based on the protocol proposed for
bioenergy crops and organic wastes (Angelidaki et al., 2009) with slight modifications. Batch
reactors (160 ml total volume, 64 ml working volume) were filled with 18 ml inoculum,
101
equivalent to a substrate-to-inoculum ratio of 2.0 for raw duckweed on a VS basis. All test
reactors were provided with substrate yielding 4.1 ± 0.03 g L-1 TC, which is equivalent to the
value for 10 g L-1 raw duckweed. Sodium bicarbonate (4 g L-1) was provided to reactors as a
buffer. After the initial pH was adjusted to 7.2 ± 0.3 by adding 2 M solutions of hydrochloric acid
and sodium hydroxide, the bottles were purged with a 80/20 (by volume) mixture of N2/CO2 gas
for 3 min prior to sealing with butyl rubber septa and aluminum crimp tops. Reactors were
incubated at 35 ± 0.5 °C for 42 days, until the weekly incremental gas production was less than
5% of the cumulative value. Test reactors were run in triplicate, and the controls (without
substrate) were run in duplicate. Biogas volumes in control bottles were subtracted from those of
tests before reporting the biomethane yields. However, the absolute biogas values were used for
carbon balances as these balances explicitly included the inorganic carbon inputs (e.g. from
buffer solutions) in the controls. Biomethane yields were expressed as ml per gram of VS
duckweed added (ml CH g VSduckweed-1, and as grams of carbon in CH4 per gram of TC in
duckweed added (g CH4-C g TCduckweed-1).
Overall duckweed-to-bioproduct conversion yields and carbon balances
Duckweed-to-bioproduct conversion yields and carbon balances in individual reactors
In all bioprocesses, liquid, solid, and gaseous TC were quantified. The losses associated
with sampling events were estimated by taking into account the sampling volumes . The VFA
losses during solids drying were estimated as 55% for basic reactors (Vahlberg et al., 2013). The
mass closure has been calculated as the ratio of the final to initial total carbon values.
Initial and final fractionation of TC among individual triplicate reactors were reported.
Initial TC consisted of substrate (raw, pretreated, or saccharified duckweed, or the residues of the
102
previous bioprocess), inoculum (yeast, acidogenic seed, or methanogenic seed) and buffer
(sodium citrate, sodium carbonate, or sodium bicarbonate) for each bioprocess. Final TC
consisted of the target bioproduct (ethanol, VFAs, or methane), slurry excluding the target
chemical, and the losses in the gaseous form (i.e. carbon dioxide for fermentation and
methanogenic digestion, methane and carbon dioxide for acidogenic digestion).
Duckweed-to-bioproduct conversion yields and carbon balances of sequential processes
Overall product yields were calculated by taking the recovery efficiencies of the products
after separation into account. The fraction of the TC recovered in the form of a target product was
calculated by multiplying the TC fraction of the target chemical with the recovery efficiency. The
remaining (i.e. non-recovered) TC of the target product in the reactor was added to the TC value
of the slurry, and accounted for in the fraction of the residue for a given bioprocess. This
adjustment was done since unrecovered product could be the substrate in the next process.
Carbon-to-carbon conversion yields of the sequential processes were calculated using Equations
4-1 to 4-3.
𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠
𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑= ∑ (
(𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡 𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑
)𝑖
1+𝛽𝑖(𝑇𝐶𝑏𝑢𝑓𝑓𝑒𝑟
𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑 )𝑖
)𝑛𝑖=1 (1)
(𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡 𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑
)𝑖
= 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑_𝑝𝑟𝑜𝑑𝑢𝑐𝑡,𝑖 (1 +𝑓𝑎𝑑𝑑𝑖𝑡𝑖𝑣𝑒𝑠𝑖 𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒𝑖
)∏ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒𝑗 (1 +𝑓𝑎𝑑𝑑𝑖𝑡𝑖𝑣𝑒𝑠𝑗
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒𝑗 )𝑖−1
𝑗=0 (2)
(𝑇𝐶𝑏𝑢𝑓𝑓𝑒𝑟
𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑 )𝑖
=𝑓𝑏𝑢𝑓𝑓𝑒𝑟
0
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒0
+ ∑ [(𝑓𝑏𝑢𝑓𝑓𝑒𝑟
𝑖
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒𝑖
)∏ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒𝑗 (1 +
𝑓𝑎𝑑𝑑𝑖𝑡𝑖𝑣𝑒𝑠𝑗
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒𝑗
)𝑖−1𝑗=0 ]𝑛
𝑖=1 (3)
Where:
𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠 = TC recovered in the products (ethanol VFA and/or methane),
𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑 = TC in initial substrate (raw, pretreated, or saccharified duckweed) added,
103
𝑇𝐶𝑏𝑢𝑓𝑓𝑒𝑟 = buffer carbon introduced in a given bioprocess,
𝑓𝑖𝑛𝑜𝑐𝑢𝑙𝑢𝑚 = fraction of TC in the inoculum, 𝑓𝑏𝑢𝑓𝑓𝑒𝑟 = fraction of TC in the buffer,
𝑓𝑎𝑑𝑑𝑖𝑡𝑖𝑣𝑒𝑠 = fraction of TC in the additives (i.e. sum of the inoculum and buffer fractions),
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒 = fraction of TC in the substrate (i.e. raw, pretreated, saccharified duckweed, or the
residues of the previous bioprocess) initially fed to a given bioprocess, 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑_𝑝𝑟𝑜𝑑𝑢𝑐𝑡 =
fraction of the reactor TC recovered in the form of a particular product (ethanol, VFAs or
methane) after the bioprocess, and
𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒 = fraction TC remaining in a given bioprocess, to be subjected to sequential processing.
Since inocula have negligible product yields, their influence on the calculations was
neglected. However, buffers used in anaerobic processes were often a significant part of the
carbon mass and could be converted into bioproduct. This effect has been taken into account as a
correction (Equation 1), by introducing the term βi, the buffer assimilation potential of a given
conversion process, utilizing the accumulated buffer in the reactor, which takes the value of zero
for fermentation and one for acidogenic digestion and methanogenic digestion.
Fertilizer potential assessment
Raw, pretreated, and anaerobically processed duckweed samples were subjected to
fertilizer tests. The scenarios involving saccharified duckweed without ethanol being one of the
end products were excluded from the assessment, as this route would not be economically viable
for VFA or methane production due to enzyme costs.
104
Statistical analysis
Data are presented as the mean ± standard deviation of triplicate samples unless specified
otherwise. Significant differences between means were tested using one-way analysis of variance
(ANOVA) and least significant difference (LSD) tests at a significance level of p<0.05
(Appendix C), using Minitab statistical package (Version 3.1, Minitab Inc., USA).
Results and Discussion
Ethanol fermentation and distillation
Total maximum theoretical glucose yield in the reactors were calculated as 0.46 ± 0.0 g
glucose g TSduckweed-1. After saccharification, actual glucose yield reached 0.38 ± 0.1 g glucose g
TSduckweed-1 (18 ± 0.1 g L-1 in final reactor volume), which corresponds to 83.4 ± 0.2 % glucose
recovery efficiency. This value is lower than the sugar recovery reported by Xu et al. (2011),
which was 96.8 % of the theoretical glucose saccharification of S. polyrrhiza starch. The slightly
low efficiency observed in our study could be due to our assumption that water soluble
carbohydrates in the duckweed biomass were glucose.
The ethanol concentration observed in the fermentation reactor after 24 h was 8.7 ± 0.1 g
L-1, which corresponds to an ethanol yield of 186 ± 1.0 g ethanol kg TSduckweed-1 . This result is
comparable to the average value reported by Soba et al. (2015), who achieved an ethanol yield of
170 g kg-1 of dry Wolffia globosa biomass after simultaneous saccharification and fermentation
(SSF) using the α-amylase, amyloglucosidase, and dry yeast. Our results on a glucose basis were
found to be higher than those reported by Yu et al. (2014) as 0.44 g g-1 (as glucose) for duckweed
grown on Schenk & Hildebrandt medium and sewage wastewater, after 94% sugar recovery.
105
Acidogenic anaerobic digestion
All reactors produced VFAs (Figure 4-1), and approximately 80% of the final VFA
values were achieved by day 5. Acetic acid was found to be the predominant VFA in all reactors
(>73%), as was also observed by another acidogenic digestion study of duckweed at high pH
values (Calicioglu et al., 2018). High production of acetic acid can be attributed in part to the
release of acetyl groups from hemicellulose under these conditions (Dahiya et al., 2015).
Figure 4-1: Volatile Fatty Acid profiles of the acidogenic duckweed reactors over ten days.
Reactors were fed with: A) raw; B) pretreated; C) saccharified; D) saccharified and fermented
duckweed.
106
The final VFA concentrations ranged from 5.9 ± 0.7 to 12.3 ± 1.6 mg L-1 (Figure 4-1).
The highest VFA concentration was observed in reactors fed with saccharified duckweed (Figure
4-1C), which produced a maximum rate of 4.8 g HAceq L-1 d-1 and an average rate of 1.23 g
HAceq L-1 d-1. The average final VFA composition in saccharified duckweed reactors consisted of
78.3% acetic, 16.3% propionic, 1.5% isobutyric, 2.0% n-butyric, and 1.9% isovaleric acids. These
results correspond to a total of 620 ± 82 mg VFA as HAceq g VSadded-1, under these conditions,
which is 2.2 times higher than that of raw duckweed. Since hydrolysis is the rate limiting step
under anaerobic conditions (Ariunbaatar et al., 2014), the higher conversion efficiencies observed
with saccharified (i.e. enzymatically hydrolyzed) duckweed is reasonable. The highest yields
achieved were comparable to another acidogenic digestion study performed on a 1:1 mixture of
primary and secondary wastewater treatment sludge, which achieved the highest VFA
concentrations at pH 10 as 0.62 g VFA g VSadded -1 (Jankowska et al., 2015). The yield observed
in our raw duckweed reactor, 288 ± 38 as HAceq g VSadded-1, is comparable to the findings of a
study conducted by Yuan et al. (2006) on acidogenic digestion of activated wastewater sludge at
pH 10 and ambient temperature (233 mg VFA as HAceq g VSadded-1). The fermented duckweed
also produced similar amounts of VFAs (12.0 ±1.3 g L-1). Although fermented substrate is also
previously saccharified, it is possible that the yeast cells present might not be as readily
biodegradable. The yield on a VS basis (611 ± 64 mg VFA as HAceq g VSadded-1), however, was
not statistically different than that of saccharified duckweed, since the volatile solids content for
the same amount of substrate carbon provided was lower in fermented duckweed residues.
Pretreatment also had a positive effect on VFA production, increasing the concentration
by 44% to 8.5 ± 1.0 g L-1 and the yield by 46% to 419 ± 51 mg VFA as HAceq g VSadded-1,
compared to raw duckweed.
Biogas recovery was minimal (< 30 ml g VSadded-1) as expected under alkaline conditions
in acidogenic digesters (Garcia-Aguirre et al., 2017), and mainly consisted of CO2. Over time, the
107
final headspace gas compositions in the reactors changed, and the final contents were found to be:
8.6 ± 4.0% CO2 and 0.0 ± 0.0% CH4 for raw duckweed; 3.0 ± 0.0% CO2 and 0.7 ± 0.1% CH4 for
pretreated duckweed; 5.7 ± 0.3% CO2 and 0.0 ± 0.0% CH4 for saccharified duckweed; and 3.4 ±
0.3% CO2 and 1.6 ± 0.4% CH4 for fermented duckweed. Hydrogen was not observed in the final
headspace gas mixture of any reactor.
Biochemical methane potentials
In all reactors, approximately 90% of the total biogas production was observed in the first
21 days (Figure 4-2). The biomethane yields ranged between 227 and 434 ml CH4 g VSadded-1 at
the end of 42 days (Figure 4-2A-B), which is higher than the 114 ml CH4 g VSadded-1 and 176 ml
CH4 g VSadded-1 reported for the anaerobic digestion of raw duckweed (Ran et al., 2018; Jain et al.,
1992).
Figure 4-2: Cumulative methane yields of the methanogenic duckweed reactors over 42 days.
Reactors were fed with raw, pretreated, saccharified, and saccharified and fermented duckweed:
A) not subjected to acidogenic digestion; B) subjected to acidogenic digestion and membrane
separation. Control biomethane yields were subtracted from each case.
108
Overall, substrates subjected to acidogenic digestion and membrane separation
(Figure 4-2B) yielded higher biomethane per VS than their counterparts subjected to less
(or no) pre-processing, (i.e. raw, pretreated, saccharified, or fermented, Figure4- 2A),
after their acidogenic digestion and recovery of VFAs. This general trend is due to lower
VS contents per TC of the anaerobically processed substrates, although same initial TC
concentration was provided to each reactor. The highest biomethane yield among all
reactors was 434 ± 0.2 ml CH4 g VSadded-1, in the reactor with saccharified, fermented,
and acidogenically-digested duckweed. This value was 62% higher than the
corresponding acidogenically-digested raw duckweed reactor (268 ± 0.1 ml CH4 g
VSadded-1), and 91% higher than the lowest observed value in the reactor fed with raw
duckweed (227 ± 0.1 ml CH4 g VSadded-1). Considering that the VS content would also be
low after two sequential bioprocesses and biopropduct recoveries, this result is
reasonable. However, comparison per TC carbon added offers a more generalizable
baseline for evaluating reactor performance and is reported in next section.
The highest biomethane yield observed for substrates without prior anaerobic
bioprocessing was 348 ± 0.3 ml CH4 g VSadded-1 for saccharified duckweed. The
biomethane value observed by the saccharified and fermented duckweed was 327 ± 0.3
ml CH4 g VSadded-1 which is higher than that reported for the anaerobic digestion of food
waste fermentation residues of 248 ml CH4 g VSadded-1 (Wu et al., 2015). However, the
observed value is slightly lower than previous findings on a sequential ethanol
fermentation and anaerobic digestion study on various duckweed sources, which reported
390 ml CH4 g VSadded-1 (Calicioglu and Brennan, 2018). Yet, in that study, the ethanol
yield was lower, which potentially left more readily biodegradable materials for the
109
downstream methane production. In addition, reactors fed pretreated duckweed provided
the next highest yields (301 ± 0.3 ml CH4 g VSadded -1) to those of reactors fed fermented
duckweed, with a 33% increase compared to raw duckweed biomethane yields (227 ± 0.1
ml CH4 g VSadded -1). However, all results obtained are lower than the yield of 468 ml
CH4 g VSadded -1 previously reported for co-digestion of Lemna gibba biomass with excess
sludge at a 50:20 mass ratio (Gaur et al., 2017). This might be due to the varying carbon
to nitrogen ratio in these studies, as this parameter can have significant effects on the
biomethane yields obtained from nitrogen-rich substrates, and can improve anaerobic
digestibility if balanced with a co-substrate (O. Calicioglu and Demirer, 2017).
Overall duckweed-to-bioproduct conversion yields and material balances
Duckweed-to-bioproduct conversion yields and carbon balances in individual reactors
The comparison of theoretical bioproduct yields on a carbon basis for all
bioprocesses individually are illustrated in Figure 4-3 in terms of initial and final %TC
distribution in the reactors. The mass closure difference between initial and final TC
values in the reactors were calculated as 4.3 ± 0.2% for fermentation, and ranged between
4.3-18.0% for acidogenic digestion, and between 3.5-8.0% for methanogenic digestion.
110
Saccharified duckweed produced the highest carbon-to-carbon conversion, for both
acidogenic digestion (53.5± 0.04%) and methanogenic digestion (22.6 ± 0.6%) reactors.
Fermentation resulted in a similar yield as methanogenic digestion on a carbon-to-carbon
basis.
Figure 4-3: Percent initial and final carbon contents of the bioreactors fed with raw, pretreated,
and saccharified duckweed and subjected to: A) fermentation; B) acidogenic digestion; C)
methanogenic digestion. The desired product in each process was ethanol (A), VFAs (B), or
methane (C).
(A)
(B)
(C)
Raw Pretreated Saccharified Initial
Initial Raw Pretreated Raw
+
AAD
Pretreated
+
AAD
Sacch. Sacch.
+
AAD
Sacch +
Ferm. Sacch +
Ferm.
+ AAD
111
The yields reported under section 3.4.1 are theoretical; i.e., they do not take into
account the availability of substrate from one process to the next in a sequential
application, or the recovery efficiencies of the target bioproducts. The actual yields,
taking into account these essential aspects for a biorefinery, are provided in the following
section.
Duckweed-to-bioproduct conversion yields of sequential processes
The duckweed-to-bioproduct conversion yields of sequential processes were
calculated by using Equation 1 for single, two, and three processes (Figure 4-4). The
recovery efficiencies were reported as 83.0 ± 0.7% for fermentation, and ranged between
94.7% and 98.3% for acidogenic digestion. The recovery efficiencies were assumed as
100 ± 5% for methanogenic digestion, since only the gaseous (i.e. already separated)
methane was used for the yield calculations and dissolved methane has not been taken
into account. All values used in the calculations (Equation 1-3) for individual reactors are
provided in Appendix C.
112
Figure 4-4: Carbon-to-carbon conversion yields as a result of individual bioprocesses, two
sequential bioprocesses, and three sequential bioprocesses for: A) saccharified; B) pretreated; C)
raw duckweed. Means that do not share a lowercase letter are significantly different.
Methane from fermentation and AAD residues
Methane from AAD residues
Methane
VFAs
Methane from fermentation residues
VFAs from fermentation residues
Ethanol
113
When individual processes are compared, the highest conversion efficiency (on a TC
basis) was observed by acidogenic digestion of saccharified duckweed, as 0.57 g
TCproducts g TCduckweed-1. This value was followed by acidogenic digestion of pretreated
duckweed (0.39 g TCproducts g TCduckweed-1), which corresponds to a 56% increase
compared to its untreated (raw) counterpart. The lowest carbon conversion value was
achieved by fermentation (0.19 g TCproducts g TCduckweed-1), which could be partially
increased by improving the separation efficiency.
In all sequential scenarios, the residuals of upstream bioprocesses were successfully
valorized. The highest overall conversion yield among all scenarios was 0.68 g TCproducts
g TCadded-1, which was achieved by subjecting duckweed sequentially to acidogenic
digestion and then methanogenic digestion. This scenario was very closely followed by
another sequential scenario involving three anaerobic bioprocesses, cascading in the
order of ethanol fermentation, acidogenic digestion, and methanogenic digestion (0.66 g
TCproducts g TCduckweed-1). The slightly lower yield observed in these three sequential
processes might be due to prior carbon losses occurring in the fermentation process.
Since one mole of CO2 is released per mole of ethanol produced, less carbon for the
downstream processes might remain, whereas the carbon losses in acidogenic digestion
were reported to be minimal (Figure 4-3B).
Fertilizer potential
The values for total nitrogen, TAN, total phosphorus, and total potassium of
reactor effluents from individual and sequential processes are provided in Figure 4-5.
114
Overall, the nutrient contents on a mg kg-1 basis increased proportionally with the number
of sequential processes. This was an expected result since the carbon in the biomass has
been recovered in the form of bioproducts. Two exceptions to this observation were
pretreated and fermented duckweed, as their TN concentrations were higher after
acidogenic digestion, compared to sequential acidogenic and methanogenic digestion.
This might be due to high VFA recoveries observed under these two conditions, followed
by a dilution of the nutrient concentrations with seed sludge before methanogenic
digestion.
As expected, the TAN concentrations also increased in parallel with the degree of
bioprocessing (Möller and Müller, 2012). However, the TAN content of the acidogenic digesters
Figure 4-5: Fertilizer potentials of reactor residuals in terms of total nitrogen (TN as N), total
phosphorus, and potassium concentrations on a dry basis. Stacked bars represent total ammonia
nitrogen (TAN) and other nitrogen species.
Total Nitrogen
(excluding TAN)Total
PhosphorusTotal
Potassium
Total Ammonia
Nitrogen (TAN)
115
was relatively low, which might be due to volatilization of ammonia at operating pH values of
9.2, and thus would result in a loss of fertilizer capacity.
In general, the nitrogen, phosphorus, and potassium concentrations observed in the study
were in alignment with the literature. For instance, Mulbry et al. (2005) reported N, P, K
concentrations of 45, 7.3, and 9.1 g kg-1 respectively, for algal turf scrubber biomass grown on
anaerobically digested dairy manure, and Wilkie and Mulbry (2002) reported 79.2, 15.4, 11.3 g
kg-1respectively for dried benthic freshwater algal biomass grown on digested dairy manure.
Conclusions
In this study, up to approximately 70% of the biomass carbon could be valorized by
sequential anaerobic bioprocessing of wastewater-derived duckweed biomass, targeting VFAs
and biomethane as end products. This value was closely followed by three sequential processes to
produce ethanol, VFAs, and biomethane. Saccharified duckweed showed the highest performance
both for individual and sequential processes in terms of carbon-to-carbon conversion. While these
technical conversion rates appear promising, it will be important to compare the economic
feasibility of two and three sequential processes. To this end, an economic analysis considering
market values of the end products and the operational costs of two and tree sequential
bioprocesses is needed. Similarly, life cycle analysis (LCA) could provide useful information on
the environmental performance of the system, and the fertilizer potential of various byproducts
could be confirmed through plant tests in greenhouses or in the field.
Acknowledgement: This project was supported by Agriculture and Food Research
Initiative Competitive Grant No. 2012-68005-19703 from the USDA National Institute of Food
and Agriculture. The findings do not necessarily reflect the view of the funding agency.
116
Chapter 5
Techno-economic Analysis and Life Cycle Assessment of Wastewater-
Derived Duckweed Biorefinery Supply Chain System
Abstract
Duckweeds are efficient aquatic plants for wastewater treatment, due to their high
nutrient uptake capabilities, growth rates, and resilience to severe environmental conditions. The
high starch and cellulose and low lignin contents of duckweed species make them an attractive
alternative for conversion into biofuels and biochemicals. In contrast to lignocellulosic
agricultural residues and energy crops, duckweed’s composition reduces or eliminates the need
for intensive pretreatment prior to saccharification. Experimental studies have shown that
sequential anaerobic bioprocessing of duckweed into ethanol, carboxylates, methane, and soil
fertilizer/amendment in a biorefinery system is feasible. However, studies on the economic and
environmental implications of such an integrated wastewater-treatment and biorefinery system
are lacking. This study aims to fill this knowledge gap toward the application of large-scale
wastewater-derived duckweed biorefineries.
A cradle-to-gate Life Cycle Assessment (LCA) was performed on a hypothetical supply
chain consisting of duckweed production ponds, harvesting, transportation, and biorefinery
operations. Duckweed supply was determined by incorporating a harvesting module into an
already existing duckweed growth model available in the literature. The most suitable end
products from wastewater-derived duckweed biomass were determined in a series of laboratory
batch experiments performed previously, and those results were used to estimate the bioproduct
yields during the hypothetical operation of a large-scale biorefinery. These experimental data
were supplemented with values from the Ecoinvent database where necessary. The impact
117
categories evaluated were: eutrophication potential; global warming potential; water depletion
potential; human health impact; and land use impact.
Introduction
Lemnaceae (duckweed), a family of simple, fast-growing, floating, aquatic plants, is a
promising option for biofuel production and holds several advantages over other bioenergy
feedstocks: (1) it can accumulate up to 43% of its biomass as easily degradable starch; (2) it does
not require direct agricultural land to produce; (3) its cell walls contain very little lignin, and so
do not require energetically- or chemically-intensive pretreatments prior to bioconversion into
fuels and chemicals; (4) its small size (1 mm – 1 cm) and uniform structure greatly reduce the
need for grinding or milling; (5) it can easily be harvested from the water surface (in contrast to
microalgae); and (6) it can be grown using nutrients derived from wastewater, and therefore can
convert a common waste stream directly into a valuable resource.
The conversion of duckweed grown as a byproduct of wastewater treatment into biofuels
has been previously studied through the thermochemical (Baliban et al., 2013) and sugar
platforms of the lignocellulosic biorefineries concept. These prior studies have mostly focused on
the technical viability of duckweed-based bioethanol production using laboratory- and pilot-scale
enzymatic saccharification and fermentation experiments (Cheng and Stomp, 2009; Ge et al.,
2012; Yu et al., 2014). While single-product studies are critical for process feasibility assessment
and optimization, this study evaluates an integrated value cascade biorefinery with multiple
synergistic product streams.
In order to frame out a complete biorefinery approach to deliver a competitive product to
the end user markets, a robust, reliable, and sustainable biofuel supply chain is essential. For this
118
reason, a variety of work has been conducted on biofuel supply chain networks, including the
raw material (biomass) production processes, storage facilities, biorefineries, blending stations,
and end users (Awudu and Zhang, 2012). In contrast to supply chains of industrial goods which
must adapt to consumer demand, biorefineries represent a small and desirable fraction of very
large markets but their size and capacity are often restricted by the regional biomass supply, and
therefore require different modeling strategies. The economic feasibility and commercial
applicability of duckweed-based bioenergy technologies must therefore be analyzed by
considering the network as a whole. A holistic approach would enable evaluation of the economic
feasibility of the biomass supply when its production is coupled with wastewater treatment, to
ensure efficient and effective delivery of the end products to blending facilities.
Duckweed-to-bioenergy research requires further study to address not only the technical
limitations of converting duckweed into various end products through individual or coupled
processes, but also the sustainability of an integrated cultivation and bioconversion system for
which wastewater provides water and nutrients inputs to the process. Coupling wastewater
treatment and feedstock production addresses ethical issues related to agricultural resource
allocation for fuel production. Moreover, integrated systems not only reduce the risk of food
insecurity, but also may be the only option for sustainable biofuel production from aquatic
biomass, such as microalgae. Similarly, it has been shown by several studies that life cycle
impacts of microalgal biofuels are dominated by the cultivation phase if wastewater is not used
(Clarens et al., 2010). In addition, Murphy & Allen (2011) have discussed that an uncoupled
microalgal biodiesel system requires seven times higher energy for wastewater management than
is produced from the biodiesel product. Therefore, wastewater treatment systems must be
considered as upstream units of anaerobic bioprocesses. This conclusion also likely applies to
duckweed-based biofuels, in that a stand-alone system may not be financially or environmentally
viable. Thus an integrated system will maximize the potential feasibility of the process and its
119
commercialization potential. A logical approach would be to perform both techno-economic and
life cycle assessment of an integrated wastewater treatment, duckweed production, and
biorefinery supply chain, in order to evaluate the sustainability of the system by comparison with
conventional wastewater treatment processes and petroleum refineries.
In this chapter, a general supply chain framework was designed for duckweed
biorefineries, under a large-scale production scenario. The supply chain was established to
determine cost in the upstream (duckweed production, handling and transportation) and the
operations of a hypothetical biorefinery. The goal of this analysis is to understand and compare
spatial and temporal options for cultivation, harvesting, and transport of duckweed; and
bioconversion of duckweed into the most feasible end products. Data from previous sections of
this dissertation research were used to develop this supply chain framework. The cost calculations
were demonstrated for a single value cascade scenario, with centralized wastewater treatment and
biorefinery processes, converting fresh duckweed into ethanol, methane, and soil amendment.
This study also utilized life cycle assessment (LCA) as a tool to analyze the
environmental impacts and energy consumption of an integrated ecological wastewater treatment
and biorefinery system, in order to assess the life cycle impact of a biorefinery supply chain
which utilizes wastewater-derived duckweed biomass as a feedstock to produce a value cascade
of end products (i.e. ethanol, methane, and soil fertilizer/amendment).
Methodology
Supply chain components
In this section, the design assumptions and details are presented for the following three
stages of the supply chain (Figure 5-1): (1) feedstock production and harvesting; (2) feedstock
120
handling and transportation; and (3) biorefinery processes. The design period of the integrated
wastewater-derived duckweed production and biorefinery system was set at 30 years, and applies
to all components of the supply chain. The feedstock production and harvesting component has
been used for the determination of the minimum biomass selling price. The calculated minimum
biomass selling price was used to determine minimum ethanol selling price of the biorefinery
component. The cost calculations were demonstrated for a single scenario, with centralized
wastewater treatment and biorefinery processes, converting fresh duckweed into ethanol, methane
and soil amendment.
Figure 5-1: System boundaries of the conceptual supply chain. Downstream processes are
excluded.
Feedstock production and harvesting
Pond Design and Wastewater Treatment:
The duckweed production and wastewater treatment design utilized here consists of three
100-acre (47 ha) ponds. This pond design enables “decentralizing” the wastewater treatment
Biomethane
midstreamupstream downstream
Biochemicalshttps://greenheatug.files.wordpress.com
Animal Feed
Duckweedproduction
Harvesting and handling
Blendingfacility
Bioethanol
Electricity
Heat
Blendingfacility
processingfacility
Carboxylateshttps://greenheatug.files.wordpress.com
Fertilizer
Bioethanol
Transportation
121
system into three plants and separating the central biorefinery operations as an alternative
scenario. In this section the scenario for central wastewater treatment and integrated biorefinery is
discussed in detail, while addition options are discussed in the following section that focuses on
life cycle analysis. For the dual functions of wastewater treatment and duckweed production, each
pond was divided into twelve plug flow modules, each with 40,000 m2 surface area and a length
to width ratio of 20, as recommended for free water surface wetlands (Jørgensen, 2009). The
depth of water was selected as 0.3 m as previously reported for duckweed ponds (Y. Zhao et al.,
2014), and the hydraulic retention time was 18 days, which required a total flowrate of 23,500 m3
d-1 (6 million gallons per day, MGD) over the 36 modules. This flowrate is equal to a wastewater
treatment plant demand for a population of 62,117, assuming wastewater generation is a typical
100 gallons per day (GPD) per capita.
Table 5-1: Wastewater treatment – duckweed production pond specifications.
Specification Value (unit)
Total area: 141 ha (3 x 100 acre)
Water depth: 0.3 m
Residence time: 18 days
Total flowrate: 23,500 m3 d-1 (6.21 MGD)
Treatment efficiency of the ponds were estimated using Equation 1 (Jørgensen, 2009) for
free water surface wetlands kinetics (Equation 5-1). Influent wastewater quality was assumed to
be equal to typical values (Metcalf and Eddy, 2003) for primary effluent, and for the base
scenario, to be constant throughout the year. The required hydraulic retention time was fixed as
18 days to match the necessary hydraulic conditions, and used to calculate the associated effluent
concentrations and removal efficiencies of wastewater components.
𝑙𝑛 [(𝐶𝑒−𝐶
∗
𝐶𝑖−𝐶∗)] =
𝑘
𝑞 Equation 5-1
122
In equation 5-1 Ce is the effluent target concentration (mg l-1); Ci is the influent
concentration (mg l-1); C* is the background concentration (mg l-1), k is the first order areal rate
constant (m d-1); and q is the hydraulic loading rate (m d-1, q = Q/A, where Q = daily flow in m3 d-
1 and A = area of the wetland in m2).
The background pond water quality was set to typical effluent characteristics of free
water surface wetlands and temperature was assumed as 200C for the decay coefficient, k (m d-1).
The annual treatment efficiency of the system for BOD was used to determine the substitution
credits for an equivalent conventional wastewater treatment plant. Duckweed was assumed to
uptake 60% of influent ammonium nitrogen and phosphorus. The rest was assumed to be
removed by microbial activity.
Table 5-2: Wastewater quality change in duckweed ponds
Influent
concentration
(mg l-1)
k at 20oC
(m d-1)
Background
concentration
(mg l-1)
Effluent
concentration
(mg l-1)
Removal
efficiency
(%)
BOD 140 0.093 10 10 92.9
TSS 70 0.027 5 5 92.9
NH4+-N 25 0.049 0.1 0.1 99.6
TP 6 0.033 0.1 0.1 98.3
TN 35 0.06 3 3 91.4
Reference
(Metcalf and Eddy,
2003) (Jørgensen, 2009) (Jørgensen, 2009) (Jørgensen, 2009)
Duckweed yield model:
Duckweed growth dynamics were simulated using Stella Architect (Version 1.1.2),
according to the intrinsic growth model developed by Lasfar et al., (2007), considering the mat
density as a variable for the intrinsic growth rate (Figure 5-2). This assumption was particularly
important in our case, as harvesting would change the mat density frequently. Other parameters
used for the description of duckweed growth function were nitrogen and phosphorus
123
concentrations, temperature, initial mat density, limiting mat density, and initial mat density, and
photoperiod, as well as the proximity of actual values to their optima. For the estimation of
photoperiod as a function of day length and calendar day, a model estimating the day length as a
function of geographic coordinates was used (Forsythe et al., 1995). Both duckweed growth and
day length models used are presented in Appendix D. Florida (FL) was selected as a hypothetical
location due to the nearly optimal conditions for duckweed growth throughout the year. The
coordinates of Fort Myers, FL, were used in the model, due to the potential availability of
sufficient wastewater. The temperature data was retrieved from National Centers for
Environmental Information database for Fort Myers, FL. The nutrient values used were the
average of the influent and effluent of the ponds, as N and P.
Figure 5-2: Illustration of the dynamic Stella Architect model used for duckweed growth and
harvesting.
Harvesting:
The optimum harvesting fraction and frequency of the duckweed mat was determined by
incorporating a harvesting module (Figure 5-2) into the duckweed growth model (Lasfar et al.,
124
2007). This was implemented using Stella Architect (Version 1.1.2). The optimum harvesting
fraction and frequency combination was found to be once in every seven days and 80% of the
pond, in order to achieve the highest biomass yield annually. Harvesting was assumed to be
performed using conventional machinery for aquatic weed harvesting. The harvester used in this
simulation had the capacity to skim 7037 m2 h-1, which was calculated to require 4.5 hours to
complete harvesting of a single pond unit, or 1.7 units per eight hour day. At a total number of 36
modules, the current design would require three machines. A quote for this equipment has been
provided in Appendix D.
Feedstock drying and transportation
For drying microalgal slurry with an initial mass of m (kg), Ali and Watson,
(2015)provided the following equation for calculating the heat requirement (kJ):
Equation 5-2
where Cp is the specific heat and equal to 4.179 and 4.762 J/g oC for water and green
algae respectively. Same study reports that reducing the moisture content from 90.26% to 20%
takes 12 hours. Assuming same specific heat values for duckweed with 92% initial moisture
content at ambient temperature (25 oC), and the initial 2% moisture difference would cause an
extra hour for drying, the energy requirement would be roughly equal to 10.9 MWh to decrease
the moisture content to 20%. This energy requirement corresponds to a power requirement of
0.91 MW.
Depending on system economics, drying the feedstock for transportation purposes could
be beneficial. In the base scenario the transportation stage has been omitted by centralizing the
biorefinery and the wastewater treatment ponds. However, for potential decentralized scenarios,
125
round trips within a 50 km radius would require 37 liters of diesel per round trip and 111 liters
per day, considering a 9.1 MT load per truck and 0.37 L km-1 diesel consumption, for 20 dry ton
of duckweed dried down to 20% moisture.
Biorefinery processes
Design basis
The process outlined in Chapter 4 sequentially subjects duckweed into anaerobic
bioprocesses in a value cascade that starts with ethanol fermentation, then proceeds to acidogenic
digestion, then to methanogenic digestion, and finally evaluates the valorization potential of the
residuals as soil amendment. The base case scenario in this chapter focuses on production of
ethanol, methane, and soil amendment from duckweed. The potential process flows, along with
the base case presented in this section are schematically shown in Figure 5-3. The methane
produced is sent to a boiler / generator to supply process heat and electricity, and if in excess, is
sold as electricity to the grid. Wastewater treatment processes are excluded from the boundary, as
the majority of the nutrients are valorized and no harsh chemicals are involved in the biorefining
process. For some design specifications such as energy requirements and residence times, values
from the US Department of Energy’s National Renewable Energy Laboratory Lignocellulosic
Biomss Biorefinery 2011 report (NREL, 2011) were used.
126
Figure 5-3: Potential biorefinery process scenarios. The solid red line shows the scenario
presented in this chapter.
Plant size
The overall quantity of duckweed (20.6 dry ton d-1) was determined by assuming the
biorefinery and algae production was coupled with a medium-size municipal wastewater
treatment plant (WWTP) that treats approximately 6 MGD. The biorefinery was assumed to be
functioning for 350 days a year (97% uptime). In the current scenario, the biorefinery was placed
next to the WWTP so there was no need for biomass drying or trucking.
Feedstock composition
The feedstock composition was assumed to be equivalent to the duckweed used for the
experimental studies in Chapter 4. The cellulose and starch components were converted into
ethanol by six-carbon sugar utilizing Saccharomyces cerevisiae.
Theoretical yields and conversions
The theoretical yields of products observed in Chapter 4 were used consistently in this
chapter. No degradation losses were assumed in the processes.
Feedstock Handling
Liquefaction
SaccharificationC6
FermentationDistillation
Anaerobic Digestion 1
Anaerobic Digestion 2
Drying
Liquid Hot Water
PretreatmentCentrifuge
Membrane Separation
99.5 % EtOH
Carboxylates
Biomethane
Fertilizer
Carboxylates and ethane scenario
Ethanol, carboxylates and ethane scenario
Base scenario (ethanol and methane)
127
Process overview
The process analyzed in this chapter consists of feedstock handling, liquefaction,
saccharification, fermentation, distillation, anaerobic digestion, and storage units. Details are
presented further in this section.
Feedstock handling
The duckweed feedstock would be delivered directly after harvesting to a feedstock
staging area in the biorefinery, so that minimal storage and handling would normally be required.
From this staging area the feedstock is further conveyed to the liquefaction unit. The moisture
content was assumed to be 92%, as no drying or transportation was included in this scenario.
Liquefaction
In this unit, the biomass is autoclaved for sterilization (121oC, 30 min) then cooled down
to 90oC and held for two hours after the addition of alpha amylase and a negligible amount of
water. Moisture content was assumed to be 92%. The original NREL report has an ammonia
conditioning tank at 1300C for 30 minutes, and this unit is used for downscaling the equipment
requirements. Table 5-3 shows the specifications of liquefaction unit.
Table 5-3: Duckweed liquefaction unit specifications.
Enzyme loading 0.3 % of total solids
Residence time 2 hours
Temperature 90oC
Pressure 1 atm
Total solids loading 8 wt%
Sacchrification
In this unit, duckweed is subjected to saccharification by the addition of glucoamylase
and cellulose simultaneously. The retention time in this unit was assumed to be 24 hours.
Temperature is held at 50oC and the pH is held at 5.2. The saccharification unit of NREL design
is 48 hours, and this difference has been considered during scale-down.
128
Table 5-4: Saccharification unit specifications
Enzyme loading 0.3 % of total solids
Residence time 24 hours
Temperature 50oC
Pressure 1 atm
Total solids loading 8 wt%
Fermentation
Fermentation takes place in batch reactors with separate batch cultivation and addition of
Saccharomyces cerevisiae at a loading of 1.6% of feedstock weight on dry basis. The
fermentation residence time was assumed to be 48 hours, in alignment with the NREL design.
The resulting duckweed beer is then sent through the ethanol recovery train. For the production of
yeast, corn steep liquor and sorbitol were used in a seed reactor. Fermentation losses due to
contamination were neglected.
Table 5-5: Fermentation unit specifications
Yeast loading 2 % of total solids
Residence time 48 hours
Temperature 32oC
Pressure 1 atm
Total solids loading 7 wt%
Distillation and rectification
The beer is separated into ethanol, water, and residual solids by distillation and solid-
liquid separation. Ethanol is distilled to a nearly azeotropic mixture with water and then purified
to 99.5% using vapor-phase molecular sieve adsorption. Solids and other liquids recovered from
the distillation bottoms are sent to anaerobic digestion.
Anaerobic digestion
Retention time in the anaerobic digestion unit is 20 days. Temperature is kept at 35oC and
pH is kept at neutral. The methane-rich biogas from anaerobic digestion is sent to the combustor.
129
Table 5-6: Anaerobic digestion unit specifications
Residence time 20 days
Temperature 35oC
Pressure 1 atm
Total solids loading 5 wt%
Soil amendment recovery
Since duckweed moisture content in the base scenario was 90%, it was assumed that the
digestate was directly applied to land in the vicinity of the biorefinery, and the costs from the
solids recovery are excluded.
Storage
This area provides bulk storage for chemicals used and produced in the process, including
corn steep liquor (CSL), enzymes, sorbitol, caustic, hydrochloric acid, water, and ethanol.
Combustor, boiler and turbogenerator
The biogas from anaerobic digestion is combusted to produce high-pressure steam for
electricity production and process heat. In the original NREL design, 36% of the
combustor/boiler and generator system was fed with biomethane, as that system also receives
residual process solids and wastewater sludge, which are excluded in our case. This difference
has been taken into account while down-sizing the unit.
Techno-economic analysis overview
A spreadsheet-based model was developed to perform the techno-economic analysis of
the duckweed biomass supply chain for a biorefinery targeting ethanol, methane, and soil
amendment as end products. The techno-economic analysis reported here uses what is known as
“nth-plant” economics. The key assumption implied by nth-plant economics is that our analysis
does not describe a pioneer or “first of a kind” plant; instead, it assumes that several facilities
130
using the same technology have already been built and are operating. Based on that experience,
the expectations is that capital and operating costs will have gone down and reliability has
increased so that the system performs as designed. In contrast, a pioneer plant is likely to have
major cost overruns and operational difficulties, which need to be factored into the deployment of
new biorefinery technologies.
Duckweed production and harvesting
Capital expenses:
In this study the NREL report on process design and economics for algal biomass
production (Davis et al., 2016) was used as the primary guide and design basis for pond design
and techno-economic evaluation. As the design of the pond is similar to the 50-acre design case
of the report, the scaling of the cost components were relatively straightforward. One major
difference between the designs are that in NREL model, pond construction included the
installation of paddlewheels for mixing the algal ponds. This portion of the design was modified,
as duckweed ponds would not require mixing. Instead, gravitational flow of wastewater
throughout the system was assumed. The breakdown of total direct expenses are given on Table
5-7. Total indirect expenses were calculated as percentage of total direct costs, with the factors
provided in the NREL report (Table 5-8). Working capital was assumed as 5% of the fixed
operating cost, and the land value was assumed as $3000 acre-1 for the calculation of total
investment cost (Table 5-9).
131
Table 5-7: Total direct expenses of a duckweed production/wastewater treatment system.
Component Value Unit Reference
Total Direct Expenses:
Total Installed Costs:
Pond production:
Civil: 910,000 $ / 100 acre (Davis et al., 2016) 9,100 $/acre
22,500 $/ha
3,200,000 $
liner LDPE (unit) price: 13 $/m2 (Beal et al., 2015) 1,410,000 m2
1,900 $/ha
267,000 $
Piping: 70,000 $/100 acre (Davis et al., 2016) 700 $/acre
1,730 $/ha
244,000 $
Subtotal: 3,700,000 $
Harvesting:
179,980 $ ea.
Machinery requirement: 3 units
Subtotal: 540,000 $
Total Installed Costs: 4,222,000 $
Additional direct costs:
Warehouse: 1 % of pond
construction
(Beal et al., 2015)
Additional direct costs: 48,000 $
Total Direct Expenses: 4,270,000 $
Table 5-8: Total indirect expenses of a duckweed production/wastewater treatment system.
Component Value Unit Reference
Field expenses: 5 % Total Direct Cost (Beal et al., 2015)
Home office and construction: 8 % Total Direct Cost (Beal et al., 2015)
Project contingency: 10 % Total Direct Cost (Beal et al., 2015)
Other costs: 1 % Total Direct Cost (Beal et al., 2015)
Total Direct Cost: 4,800,000 $ Total Direct Cost Total indirect expenses: 1,170,000 $
132
Table 5-9: Total capital expenses of a duckweed production/wastewater treatment system.
Component Value Unit Reference
Fixed capital investment: 5,3400,000 $ Working Capital: 5 % FCI
265,000 $ Land: 3,000 $/acre (Davis et al., 2016)
7,400 $/ha
1,045,000 $ total (Davis et al., 2016) Total Capital Investment: 6,619,000 $
Operating expenses
Since the duckweed production system was assumed to be passive (i.e. gravity driven),
electricity demand was neglected. For the harvesting operations, the fuel requirements of the
aquatic weed harvesters were taken into account. The associated variable operating costs are
given on Table 5-11.
Labor salaries were also taken from the NREL report and the labor burden was applied as
90% as suggested. This labor covers items such as safety, general engineering, general plant
maintenance, payroll overhead (including benefits). The labor demand was down-scaled to meet
the requirements of the current design. Property insurance and tax was assumed to be 0.7 % of the
fixed capital investment. The maintenance of the pond was assumed to require 0.5 % of its capital
cost annually, and the maintenance requirement of the harvesting machinery was assumed as 4 %
of their capital cost annually. Total fixed operating costs are presented in Table 5-10.
133
Table 5-10: Operating expenses of wastewater treatment - duckweed production system.
Variable operating cost: Pond:
Electricity: Neglected Harvesting:
Fuel requirement: 8.0 h/day
2 L/h
147 Gal/yr
3.1 $ diesel/ hgal
4,600 $/ year Fixed operating costs:
Pond maintenance: 0.5 % of capital cost (Davis et al., 2016)
18,400 $/yr Harvesting:
Maintenance: 3 % of capital cost
16,200 $/yr Labor: (Davis et al., 2016)
Manager: 155,400 $/yr Technician: 82,000 $/yr Technician: 60,000 $/yr
Module operator: 81,000 $/yr
378,000 $ Labor burden: 90 %
Labor subtotal: 719,000 $ Property insurance & tax: 0.7 % FCI
37,000 $ Total fixed operating costs: 795,000 $
For the determination of by-product credit, the amount of BOD treatment achieved in the
system was taken as the basis for the calculation of associated activated sludge unit construction
and operating costs. The construction and operating costs of WWTPs were estimated using
approximations provided by Fraas and Munley (1984), and the deflator indices were used to
estimate the values in 2018 as $5.92 kg BOD-1 year-1 and $0.85 kg BOD-1 year-1, respectively
(Fraas and Munley, 1984). The activated sludge contribution to construction was assumed as
20%, and for the operation as 50%. The associated wastewater treatment credits are given on
Table 5-11. Both the construction and operation credits are considered as annual credits in our
design.
134
Table 5-11: Wastewater treatment credit
Component Value Unit Reference
BOD removal per year: 1,042,000 kg/year
Wastewater treatment construction cost: 5.92 $/ kg BOD/ year (Fraas and Munley, 1984)
Total construction substitution: 6,173,000 $/ year
Activated sludge portion: 0.20 $/$ of WWTP
Construction credit: 1,235,000 $/ year
Wastewater treatment operation cost: 0.85 $/ kg BOD/ year (Fraas and Munley, 1984)
Total operation substitution: 882,000 $/ year Activated sludge portion: 0.50 $/$ of WWTP
Operation credit: 441,000 $/ year
Discounted cash flow analysis
For this analysis, the discount rate, which is also the internal rate of return (IRR) in this
analysis, was set to 10% and the plant lifetime was set to 30 years. For this analysis, it was
assumed that the plant would be 40% equity financed. The terms of the loan were taken to be 8%
interest for 10 years. The principal was taken out in stages over the two-year construction period.
This is all consistent with the assumptions used in the NREL design. The discounted cash flow
analysis is given in Appendix D.
Table 5-12: Input data for discounted cash flow rate of return analysis of wastewater treatment -
duckweed production system.
Component Value Unit Reference
Biomass production rate: 7200 Mg/yr
BMSP: 38.8 $/mg
Equity: 40 % (Davis et al., 2016)
Interest rate: 8 % (Davis et al., 2016)
Loan term: 10 years (Davis et al., 2016)
Inflation rate: 0 % (Davis et al., 2016)
Plant life: 30 years (Davis et al., 2016)
Discount rate (irr): 10 % (Davis et al., 2016)
General plant depreciation: 200 % (Davis et al., 2016)
General plant recovery period: 7 years (Davis et al., 2016)
Federal tax rate: 35 % (Davis et al., 2016)
Construction period: 2 years modified
First year expenditure: 60 % modified
Second year expenditure: 40 % modified
Working capital: 5 % FCI (Davis et al., 2016)
Start-up time: 0.5 years (Davis et al., 2016)
Variable costs during start-up: 75 % (Davis et al., 2016)
Fixed costs incurred during start-up: 100 % (Davis et al., 2016)
Start-up yield: 50 % modified
135
Feedstock handling and transportation
Although not included in the current scenario and discounted cash flow analysis, the
duckweed dewatering unit to bring the moisture content down to 80% was calculated by
considering the purchase (5,300 $/m3/h) and installation (70 $/m3/day) costs provided by Beal et
al., (2015). Three centrifuge units at $942 (installed) were considered as the pond system can
either be centralized or decentralized. For maintenance, 3% of this cost would be assumed as a
fixed operating expense. For further drying scenarios, the lignin cake drying scenario on NREL’s
2002 lignocellulosic biorefinery design (Aden et al., 2002) could be considered.
Biorefinery processes
For the biorefinery techno-economic analysis, the NREL 2011 lignocellulosic ethanol
report was taken as a basis (NREL, 2011), without application of cost year indices for the
estimation of the current value. The process described in the report uses co-current dilute-acid
pretreatment of lignocellulosic biomass (corn stover), followed by enzymatic hydrolysis
(saccharification) of the remaining cellulose, and fermentation of the resulting glucose and xylose
to ethanol. The process design also includes feedstock handling and storage, product purification,
wastewater treatment, lignin combustion, product storage, and required utilities. Since the
duckweed design is similar, yet also has significant differences, modifications were made when
necessary. For example, the duckweed liquefaction stage replaces NREL pretreatment, as our
design produces ethanol from both starch and cellulose with the addition of alpha amylase, but
without pretreatment. The biomass processing capacity for the NREL design is about 100 times
larger (2000 metric tonnes dry mass per day) than our duckweed biorefinery case. The NREL
136
design obtained a detailed quote for the biorefinery totaling approximately $20 MM for the whole
system,
Capital expenses
A factored approach in which multipliers are applied to the purchased equipment cost
was considered for the calculation of scaled, purchased, and installed costs considering the quotes
for the NREL biorefinery as a starting point. However, very likely this is an overestimation due to
large differences in scale (up to 200 times in some units). Scaling factors were applied using
Equation 5-2. Table 5-14 summarizes the calculated installed costs as total direct expenses along
with the additional costs, considering the additional cost factors as a percentage of total installed
costs. Further breakdown of the equipment costs is provided in Appendix D.
𝑁𝑒𝑤 𝐶𝑜𝑠𝑡 = (𝐵𝑎𝑠𝑒 𝐶𝑜𝑠𝑡) (𝑁𝑒𝑤 𝑆𝑖𝑧𝑒
𝐵𝑎𝑠𝑒 𝑆𝑖𝑧𝑒)𝑛
Equation 5-3
Where n is a characteristic scaling exponent (typically in the range of 0.6 to 0.7).
Table 5-13: Total direct expenses for the biorefinery.
Component Value Unit Reference
Total Direct Costs: Storage and Handling: 1,964,000 $
Liquefaction totals: 78,000 $ Saccharification and Fermentation: 810,000 $
Distillation and Rectification: 1,555,000 $ Anaerobic Digestion: 1,060,000 $
Storage: 96,000
Boiler and Turbogenerator: 5,927,000 $ Total Installed Costs: 11,490,000 $
Additional direct costs: (NREL, 2011)
Warehouse: 4 % of installed cost site development: 9 % additional piping: 5 %
Additional direct costs subtotal: 2,011,000 $ Total Direct Cost Subtotal: 13,501,000 $
137
Table 5-14: Total indirect expenses for the biorefinery.
Component Value Unit Reference
Total Indirect Costs: (NREL, 2011)
Profitable expenses: 10 % Total Direct Cost Field expenses: 10 %
Home office and construction: 20 % Project contingency: 10 %
Other costs: 10 % Total indirect costs subtotal: 8,101,000 $
Table 5-15:Total capital investment for the biorefinery.
Fixed Capital Investment: 21,602,000 $ Working Capital: 5 % FCI (NREL, 2011)
1,080,000 $ Land: 1,800,000 $ (NREL, 2011)
25 decreasing factor
72,000 $ total Total Capital Investment: 22,754,000 $
Operating expenses
The recommended number of employees in the NREL report (60) were scaled down to
meet the requirements of the duckweed biorefinery (11). A labor burden of 90% was applied to
the total salary. The labor cost breakdown is presented in Appendix D. The maintenance was
assumed to take 3% of total installed costs, and the property insurance tax would cost 0.7 % of
the fixed capital investment. A breakdown of total fixed operating costs is provided in Table 5-
16.
Table 5-16: Fixed operating expenses
Component Value Unit References
Fixed operating costs: Labor: 1,107,0008 $
Maintenance: 3 % total installed cost (NREL, 2011)
345,000 $
Property insurance & tax: 0.7 % (NREL, 2011)
151,000 $ Total fixed operating costs: 1,603,000 $
138
The variable operating costs include feedstock, chemical, and energy requirements of the
biorefinery. In the base scenario, the methane produced by anaerobic digestion was used to
supply the energy requirements of the biorefinery processes. This energy requirement was
assumed to be 2% of the NREL design case, as a conservative estimate. However, the actual
energy requirement of this system will likely be less than 2% of the NREL design, if a mass and
energy balance were to be performed. The by-product credits for additional electricity to the grid
assumes 0.07 $/kWh credit, which is consistent with NREL report.
Table 5-17: Variable operating expenses
Component Value Unit References
Variable operating costs: Feedstock cost:
25.2 $/Mg Excluding enzyme production: 23179 kW requirement (NREL, 2011)
Scale: 50
464 kW required
1.5 MW provided
-982 kW extra -8,3 kW/yr extra
-469979 $/year credit Chemicals: 4.900 $/yr.
Total Variable operating costs: 4,900 $
Byproduct credits: -470.000 $/year credit
Discounted cash flow analysis
For this analysis, the discount rate, which is also the internal rate of return (IRR) in this
analysis, was set to 2.45% and the plant lifetime was set to 30 years. For this analysis, it was
assumed that the plant would be 40% equity financed. The terms of the loan were taken to be 8%
interest for 10 years. These data are all consistent with NREL biorefinery design. The
construction period was modified to be two years (Table 5-18). This is all consistent with the
assumptions used in the NREL biorefinery design. The discounted cash flow rate of return details
are given in Appendix D.
139
Table 5-18: Input data for discounted cash flow rate of return analysis of wastewater treatment -
duckweed production system.
Component Value Unit References
Feedstock cost: 176,000 $/yr.
Ethanol production rate 439.000 gal/year
Equity: 40 % interest
Interest rate: 8 %
Loan term: 10 years
Inflation rate: 0 %
Plant life: 30 years
Discount rate (Internal rate of Return): 2 %
General plant depreciation: 150% %
General plant recovery period: 7 years
Federal tax rate: 35 %
Construction period: 2 years
First year expenditure: 60 %
Second year expenditure: 40 %
Working capital: 5 % FCI
Start-up time: 0.5 years
Variable costs during start0up: 75% %
Fixed costs incurred during start-up: 100% %
Start-up yield: 50% %
Byproduct credit: 577,000 $
Life cycle assessment overview
The LCA of the integrated duckweed production, wastewater treatment and biorefinery
system was conducted according the standards set forth by the International Organization for
Standardization (ISO) ISO 14040:2006 and ISO 14044:2006. Brightway2, an open source LCA
framework was used for Ecoinvent 3.3 database communication and LCA processing (Mutel,
2017).
Goal and scope definition
The goal of this LCA was to assess the environmental impacts associated with the life
cycle of municipal wastewater-derived duckweed biorefineries, producing bioethanol,
biomethane, and soil fertilizer/amendment over a 30 year design period. The system boundary
140
was defined as cradle-to-gate, including the construction and operation of wetlands for duckweed
production, and excluding the biorefinery end product distribution. The components of the system
were identical to those described in detail in Chapter 5, Techno-economic Analysis of
Wastewater-Derived Duckweed Biorefinery Supply Chain System of this dissertation.
Functional unit
The functional unit was selected as one square meter (m2) for duckweed
production/wastewater treatment, in order to facilitate a comparison of the effects with those of
other feedstocks. All calculations were made both for wastewater treatment and biorefinery
sections taking the functional unit into account.
Life cycle inventory (LCI)
The life cycle inventory was performed for the following phases of the biorefinery supply
chain: pond construction and operation; duckweed cultivation; transportation; drying; biorefinery
construction; fermentation; distillation; anaerobic digestion; and solids recovery for soil
fertilizer/amendment.
The wastewater availability and wetland sizing was calculated according to typical
wetland design parameters (IDNR, 2007). The wetland construction material inventory was based
on the aerated lagoon dataset in Ecoinvent 3.3. The total duckweed yield was calculated by
incorporating a harvesting module into a duckweed growth model (Lasfar et al., 2007), using
Stella Architect (Version 1.1.2). For duckweed transportation, a radius of 50 km was assumed,
but the transportation and drying components were excluded from the boundary in the results
section, to be consistent with the techno-economic analysis conducted in this dissertation work.
The biorefinery processes were designed based on a production capacity of 20.1 ton dry
biomass per day, considering wastewater availability for duckweed production as a limiting
141
factor. Product yields were based on a previous laboratory study for sequential ethanol
fermentation and anaerobic digestion of duckweed (Calicioglu and Brennan, 2018). The end
products (i.e. bioethanol, biomethane, and soil fertilizer/amendment), were assumed to substitute
for gasoline, natural gas, and synthetic nitrogen fertilizer (liquid ammonia), and the associated
impacts with the production of commercial fuels and chemicals were credited to our system. The
biorefinery processes consisted of liquefaction, saccharification, fermentation, anaerobic
digestion, and solid recovery as soil fertilizer/amendment. The energy requirements for the
biorefinery were calculated based on the NREL biorefinery model (NREL, 2011), using
appropriate scaling factors for each process. Materials inventory for the biorefinery was estimated
from Ecoinvent 3.3 dataset, and the details are provided in Appendix D.
Life cycle impact assessment (LCIA)
Impact categories
The impact categories used in this study were: global warming potential (IPCC 2013,
climate change, GWP 100a); eutrophication potential (ReCiPe Endpoint, freshwater
eutrophication); water depletion potential (ReCiPe Midpoint, water depletion); human health
impact (ReCiPe Endpoint, human health, total); and land use impact (ReCiPe Endpoint, natural
land transformation).
142
Results and Discussion
Techno-economic analysis
Duckweed production and harvesting
Figure 5.4 shows the breakdown of the capital and operating expenses of wastewater
treatment – duckweed production system. It was found that the largest contributor of the
duckweed cultivation capital expenses is the pond construction (55.6%), followed by the land
cost (15.8%). Within the total lifetime of 30 years, the operational expenses are more significant
compared to capital expenses (Figure 5.5).
Figure 5-4: A breakdown summary of the capital (A) and operating (B) expenses of a wastewater
treatment – duckweed production system.
Discounted cash flow rate of return results revealed that minimum duckweed biomass
selling price of $25 per dry Mg with an 10% internal rate of return could be achieved if the
system boundaries consider wastewater treatment as credit. This price is slightly lower than those
of agricultural residues such as corn stover ($40) (Brown and Brown, 2014). The minimum
55.6%
8.2%
0.7%
15.7%
4.0%
15.8%Pond construction
Harvesting
equipmentAdditional direct
costTotal indirect
costsWorking Capital
Land
A) Capital Expenses
0.6%2.3%
2.0%
90.4%
4.7%Harvesting fuel
Pond
maintenance
Harvesting
maintenance
Labor
Property
insurance & tax
B) Operating Expenses
143
biomass selling price was calculated considering the wastewater treatment credits as a by-
product, and this assumption caused a major drop in the prices (Figure 5-6).
Figure 5-5: Breakdown of costs and revenues for the discounted cash flow analysis for minimum
biomass selling price of 25.2 USD
Figure 5-6: Minimum biomass selling price at differenc considerations of wastewater treatment
credits
Biorefinery processes
Using the yields provided in the experimental studies of Chapter 4, the techno-economic
analysis of a hypothetical large-scale duckweed production/wastewater treatment and biorefinery
system was performed. Modification and downscaling of National Renewable Energy Laboratory
-$25,000,000 $0 $25,000,000
$ (USD)
Fixed capital investment
Land + working capital
Initial interest
Biomass sales
By-product credit
Variable operating costs
Fixed operating costs
Income tax
Loan payment
0
100
200
300
Activated sludge
construction and
operation credit
Activated sludge
construction
credit
Activated slude
operation credit
No creditMin
imum
bio
mas
s se
llin
g p
rice
($/
dy M
g)
Wastewater treatment plant construction and operation credits
144
2011 Report on lignocellulosic biorefinery to a daily processing capacity of 20.1 Mg dry weight
of duckweed biomass revealed a minimum ethanol selling price of $8.4 per U.S. gallon with a
2.45% internal rate of return, which is a four times higher price than their findings for the ethanol
biorefinery (Figure 5-8), and more the four times 2018 ethanol market prices. This high price may
be due to an overestimation of costs associated with capital expenses and energy requirements
during scale down, but may also indicate that much larger facilities are needed to achieve
economies of scale with this configuration of technologies, or even that marginal operating costs
alone are too great to justify this approach. Figure 5.8 illustrates the minimum ethanol selling
price versus plant capacity curve, which would be be more steep if capital costs were the primary
driver. But for a duckweed production system and integrated biorefinery, labor is the primary cost
so fewer economies of scale are expected. For the calculation of a more realistic minimum
ethanol selling price, a rigorous mass and energy balance and detailed labor and management
analysis must be performed.
Figure 5-7: Minimum ethanol selling price at different daily processing capacities.
To improve the overall economic feasibility of the system, higher value products such as
proteins could be targeted upstream of ethanol production. For example, one more end product,
0
1
2
3
4
5
6
7
8
9
15 25 35 45 55 65Min
imu
m E
than
ol
Sel
lin
g P
rice
($
/gal)
Daily capacity (Mg/day)
NREL's
145
mixed carboxylic acids, could be added to the value cascade grid of the biorefinery, and should be
included in the assessment as another scenario.
As an alternative to ethanol as the first stage of this value cascade, two-stage anaerobic
digestion (i.e. where acidogenic digestion effluents are subjected to methanogenic digestion)
could provide high biomethane conversion yields. The produced biogas could be converted to
renewable natural gas. This two stage anaerobic digestion strategy had one of the best carbon-to
carbon conversion results in Chapter 4,
Life cycle assessment
Figure 5-8 shows the contribution of life cycle phases of wastewater-derived duckweed
biorefinery supply chain to environmental impact categories when duckweed is grown in land-
based ponds in Fort Myers, Florida. The contribution of life cycle phases of a wastewater-derived
duckweed biorefinery supply chain revealed a strong net benefit on eutrophication potential, due
to the recovery of nutrients from wastewater into duckweed biomass. This produced a net benefit
on reducing eutrophication potential.
Overall, however, the environmental impacts of a duckweed biorefinery appear to be
higher than that of the substituted products. The environmental impacts of duckweed biorefinery
products relative to substituted products (i.e. gasoline, natural gas, and chemical fertilizers) could
therefore depend on biorefinery size: the larger the biorefinery, the smaller the environmental
impacts.
146
Figure 5-8: Contribution of life cycle phases of wastewater-derived duckweed biorefinery supply
chain to environmental impact categories when duckweed is grown in land-based ponds in
Florida, USA.
At the scale analyzed, the highest contribution to environmental impacts in the land use
category was associated with the construction of the duckweed growth ponds. Since the pond
construction impacts were estimated using a dataset for aerated lagoons, the wastewater treatment
phase requires further analysis and inventorying for a more realistic result. In addition, vertical
farming of duckweed could be an option to minimize the land use impact. The largest negative
human health impact is originated from the distillation unit, due to the volatile organic compound
losses during the process. The duckweed fermentation unit revealed the highest impacts on water
depletion potential, due to the water demand associated with the production of yeast.
-100% -50% 0% 50% 100%
Percent contribution
Pond-construction Water-quality-changeDuckweed-cultivation Biorefinery-constructionLiquefaction SaccharificationDuckweed-fermentation DistillationDuckweed-anaerobic-digestion Gasoline-substitutionNatural-gas-substitution Nitrogen-fertilizer-substitutionWWTP substitution
Global warming potential
Euttophication potential
Water depletion potential
Human health
Land use
147
Conclusion and Future Work
Earlier chapters in this dissertation demonstrated that duckweed is a technically feasible
alternative feedstock for the production of fuels and chemicals, but did not address environmental
impacts. Integrating duckweed production with wastewater treatment has positive impacts on
eutrophication mitigation. However, pond construction brings significant burden in terms of land
use, and this issue could be addressed by investigating vertical farming options as another
production scenario. Offsets for gasoline, natural gas and fertilizer substation by biorefinery
products as well as WWTP offsets reduce the global warming potential of the system, but not to
zero. Downscaling an already-existing biorefinery model for the estimation of the life cycle
burden associated with the system at hand may not have been sufficient to properly assess these
impacts at a much smaller scale. Therefore, a LCA inventory and analysis at a higher resolution is
needed to come up with more realistic impact assessment of the biorefinery processes.
Primary data for the appropriate scale of biorefinery processes must be gathered from
vendors and process simulation tools such as Aspen Capital Cost Estimator. A sensitivity analysis
on duckweed yield, duckweed carbon content, and end product yields must also be performed. In
addition, seasonality of the wastewater quality and treatment efficiency must also be taken
account for better precision in duckweed yield estimations. Some strategies that might improve
the economics of the system include: vertical farming of duckweed; larger decentralized ponds
and a central biorefinery; duckweed drying and transportation alternatives; integration of higher-
value end products to the biorefinery; hemicellulose fermentation, carbon capture and storage
after biorefinery processes; heat recovery from biorefinery processes.
Acknowledgement: This project was supported by Agriculture and Food Research
Initiative Competitive Grant No. 2012-68005-19703 from the USDA National Institute of Food
and Agriculture. The findings do not necessarily reflect the view of the funding agency.
148
Chapter 6
Conclusions, Significance and Future Work
This study addressed the gaps to evaluate the technical, economic, and environmental
potential of establishing integrated wastewater-derived duckweed biorefineries targeting
biochemical (carboxylic acids) and two energy carrier end products (bioethanol and biomethane),
along with the potential of valorizing the residuals as a soil amendment. The performances of the
production of individual products were compared to other feedstock in literature, and the overall
yields were also evaluated for a particular duckweed biorefinery system.
This work studied the performance of duckweed during acidogenic digestion under
various operating conditions, with an emphasis on understanding the behavior of acidogenic
microbial consortia. This study is particularly significant in terms of understanding substrate
assimilation potential at high-pH conditions, as the literature on acidogenic digestion under basic
conditions is scarce. It is known that free ammonia is toxic to microorganisms. The interaction
between microbial species and the elevated ammonia concentrations under basic conditions was
not in the context of this study, and requires future work such as anaerobic toxicity assays
specific to the acidogenic communities. Simultaneous acidogenic digestion and ammonia
recovery could also be an interesting future study.
The biorefinery experimental results demonstrated that two (ethanol and VFAs) or three
(ethanol, VFAs, and methane) bioproducts could be targeted to maximize product yield on a
carbon basis. This study also outlined a framework for the evaluation of carbon-to-carbon
conversion for the evaluation of laboratory-scale bioproduct value cascade experiments. A buffer
assimilation capacity term was introduced to account for the potential interference of end product
yields obtained from duckweed, due to the conversion of buffer added to a system in the upstream
or midstream of a particular anaerobic bioprocess (e.g. conversion of citrate buffer added in
149
fermentation process into volatile fatty acids during abiogenic digestion). The value of this
coefficient was taken as equal to one in this study, which implies that all buffer added was
assumed to be assimilated. This is a conservative estimate, and more work can be conducted to
come up with the empirical values.
The potential for duckweed yields when the system is coupled with wastewater treatment,
by dynamic modelling were computed. This study assumed constant nitrogen update as percent
available in the ponds, rather than considering the actual kinetics of nitrogen uptake, and
considering the implications of harvesting (i.e. absence of coverage on the surface of the ponds)
on the nitrogen availability and treatment efficiency. Therefore, a better nitrogen balance can be
performed in parallel with further experimental work to validate the assumptions of the model
developed in this study. Similarly, a more comprehensive mass balance over BOD removal
would be useful in the determination of the treatment efficiencies, and in turn, wastewater
treatment credits associated with the integrated wastewater treatment-duckweed production
design. In addition, a detailed analysis of the BOD removal mechanisms would also give insight
on the potential methane emissions caused by the lowered rates of oxygen penetration to the
ponds (due to lowered diffusion potential and the absence of light for algal growth) in the
presence of duckweed. Determination of methane loss potential is particularly important for the
evaluation of the environmental impacts of the system.
Designing shallow ponds would enable better diffusion efficiency and may avoid
anaerobic conditions resulting in methane release. In addition, shallow ponds in a vertical farming
setting could increase area availability for duckweed production, and could reduce the need for
the utilization of primary wastewater effluent, as the slower yields obtained on secondary
treatment effluents could be compensated. Such a design would decrease the uncertainty about
treatment efficiency for BOD, and avoid the risks of methane emissions. However, eliminating
primary treatment would decrease the wastewater treatment credits, as only a smaller portion of
150
BOD removal will be achieved in the secondary effluent. Yet, if duckweed efficiently removes
nutrients below strict nutrient limits (e.g. <3 mg/L TN and <0.1 mg/L TP in the Chesapeake Bay
area), the credits can be still significant since such low nutrient limits require costly treatment
technologies. However, such a design receiving secondary effluent in shallow trays needs
experimental evidence for validation.
In the techno-economic analysis, ethanol production capacity was found to be too small
(about 1:100) compared to commercial lignocellulosic biorferineries, when the system is coupled
with wastewater treatment. This fact had a drawback in terms of the economics of scale. Mass
and energy balances at the original scale of the design could reveal more realistic economic
performance of the system. While designing the system, the process configurations could be
selected differently for the specific units, compared to conventional methods (e.g., VFA
production results in Chapter 3 suggest that a batch fermentation system would be more logical
than continuous, unlike the conventional acidogenic digestion processes). In addition, carbon
dioxide credits for capture and utilization or geological storage of that CO2 could improve the
economics of the particular scenario presented in Chapter 5.
This study was the first to evaluate the environmental performance of duckweed
production on wastewater and its conversion into valuable fuels and chemicals in a biorefinery
concept, and revealed positive impacts on eutrophication mitigation. However, its high negative
impact on land use when grown in ponds suggests that vertical farming options should be
investigated. In Chapter 5, it was assumed that the ponds would be covered with liners, which
brought additional negative environmental impacts. In the current context these plastic liners
might not be necessary and could be excluded from the design, which would partially improve the
environmental impacts of pond construction. In the biorefinery end, the distillation process was
found to have high environmental “costs”. High-solids fermentation might improve the outcomes,
but might require dewatering and drying of duckweed to obtain higher solids content, which
151
might bring additional environmental burdens in terms of energy consumption. A sensitivity
analysis on duckweed moisture reduction and high solids fermentation would be necessary to
understand the relationship between solids content and associated environmental impacts.
Apart from the potential ways to improve the current design’s economics and
environmental performance as mentioned above, other scenarios could be evaluated for the
integrated wastewater treatment-duckweed production pathways. For example, similar to the
approach of the experimental work as detailed in Chapter 4, individual and sequential processes
targeting one or more end products must be simulated for a more comprehensive TEA and LCA.
Such an approach could reveal interesting outcomes if the fuel and chemical production trains are
excluded and the duckweed is utilized as a high protein feedstock only. Therefore, a more
valuable product could be targeted in a biorefinery system, such as proteins. In this example,
however, the system might not be suitable for coupling with municipal wastewater treatment due
to social acceptance. In such a scenario, the wastewater might need substitution (with fertilizers
or at least a more homogeneous waste stream as opposed to municipal wastewater) for the growth
of duckweed, and its impacts on the system economics might be significant. Sensitivity analysis
over production (pond vs. vertical growth, wastewater vs. fertilizer), conversion processes
(producing one or more end products) and market prices for the product portfolio under multiple
scenarios is required.
152
REFERENCES
Aden, A., Ruth, M., Ibsen, K., Jechura, J., Neeves, K., Sheehan, J., Wallace, B., Montague, L.,
Slayton, A., Lukas, J., 2002. Lignocellulosic Biomass to Ethanol Process Design and
Economics Utilizing Co-Current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis for
Corn Stover.
Agler, M.T., Wrenn, B.A., Zinder, S.H., Angenent, L.T., 2011. Waste to bioproduct conversion
with undefined mixed cultures: The carboxylate platform. Trends Biotechnol. 29, 70–78.
doi:10.1016/j.tibtech.2010.11.006
Aiello-Mazzarri, C., Agbogbo, F.K., Holtzapple, M.T., 2006. Conversion of municipal solid
waste to carboxylic acids using a mixed culture of mesophilic microorganisms. Bioresour.
Technol. 97, 47–56. doi:10.1016/j.biortech.2005.02.020
Ali, M., Watson, I.A., 2015. Microwave treatment of wetalgal paste for enhanced solvent
extraction of lipids for biodiesel production. Renew. Energy 76, 470–477.
doi:10.1016/j.renene.2014.11.024
Alzate, M.E., Munoz, R., Rogalla, F., Fdz-Polanco, F., Perez-Elvira, S.I., 2012. Biochemical
methane potential of microalgae: Influence of substrate to inoculum ratio, biomass
concentration and pretreatment. Bioresour. Technol. 123, 488–494.
doi:10.1016/j.biortech.2012.06.113
Angelidaki, I., Alves, M., Bolzonella, D., Borzacconi, L., Campos, J.L., Guwy, A.J., Kalyuzhnyi,
S., Jenicek, P., Van Lier, J.B., 2009. Defining the biomethane potential (BMP) of solid
organic wastes and energy crops: A proposed protocol for batch assays. Water Sci. Technol.
59, 927–934. doi:10.2166/wst.2009.040
APHA/AWWA/WEF, 2012. Standard Methods for the Examination of Water and Wastewater.
American Public Health Association, Washington D.C.
153
Appels, L., Baeyens, J., Degrève, J., Dewil, R., 2008. Principles and potential of the anaerobic
digestion of waste-activated sludge. Prog. Energy Combust. Sci. 34, 755–781.
doi:10.1016/j.pecs.2008.06.002
Appels, L., Lauwers, J., Degrve, J., Helsen, L., Lievens, B., Willems, K., Van Impe, J., Dewil, R.,
2011. Anaerobic digestion in global bio-energy production: Potential and research
challenges. Renew. Sustain. Energy Rev. 15, 4295–4301. doi:10.1016/j.rser.2011.07.121
Apprill, a, McNally, S., Parsons, R., Weber, L., 2015. Minor revision to V4 region SSU rRNA
806R gene primer greatly increases detection of SAR11 bacterioplankton. Aquat. Microb.
Ecol. 75, 129–137. doi:10.3354/ame01753
Ariunbaatar, J., Panico, A., Esposito, G., Pirozzi, F., Lens, P.N.L., 2014. Pretreatment methods to
enhance anaerobic digestion of organic solid waste. Appl. Energy 123, 143–156.
doi:10.1016/j.apenergy.2014.02.035
Aronesty, E., 2013. Comparison of Sequencing Utility Programs. Open Bioinforma. J. 7, 1–8.
doi:10.2174/1875036201307010001
Arslan, D., Steinbusch, K.J.J., Diels, L., De Wever, H., Hamelers, H.V.M., Buisman, C.J.N.,
2013. Selective carboxylate production by controlling hydrogen, carbon dioxide and
substrate concentrations in mixed culture fermentation. Bioresour. Technol. 136, 452–460.
doi:10.1016/j.biortech.2013.03.063
Awudu, I., Zhang, J., 2012. Uncertainties and sustainability concepts in biofuel supply chain
management: A review. Renew. Sustain. Energy Rev. 16, 1359–1368.
doi:10.1016/j.rser.2011.10.016
Badger, P., 2002. Ethanol from cellulose: A general review. Trends new Crop. new uses 17–21.
Baliban, R.C., Elia, J.A., Floudas, C.A., Xiao, X., Zhang, Z., Li, J., Cao, H., Ma, J., Qiao, Y., Hu,
X., 2013. Thermochemical Conversion of Duckweed Biomass to Gasoline, Diesel, and Jet
Fuel: Process Synthesis and Global Optimization BT - Industrial & Engineering Chemistry
154
Research. Ind. Chem. Res.
Beal, C.M., Gerber, L.N., Sills, D.L., Huntley, M.E., Machesky, S.C., Walsh, M.J., Tester, J.W.,
Archibald, I., Granados, J., Greene, C.H., 2015. Algal biofuel production for fuels and feed
in a 100-ha facility: A comprehensive techno-economic analysis and life cycle assessment.
Algal Res. 10, 266–279. doi:10.1016/j.algal.2015.04.017
Biddy, M.J., Davis, R., Humbird, D., Tao, L., Dowe, N., Guarnieri, M.T., Linger, J.G., Karp,
E.M., Salvachúa, D., Vardon, D.R., Beckham, G.T., 2016. The Techno-Economic Basis for
Coproduct Manufacturing to Enable Hydrocarbon Fuel Production from Lignocellulosic
Biomass. ACS Sustain. Chem. Eng. 4, 3196–3211. doi:10.1021/acssuschemeng.6b00243
Bondesson, P.-M., Galbe, M., Zacchi, G., 2013. Ethanol and biogas production after steam
pretreatment of corn stover with or without the addition of sulphuric acid. Biotechnol.
Biofuels 6, 11. doi:10.1186/1754-6834-6-11
Bondesson, P.M., 2008. Combined production of bioethanol and biogas from wheat straw. J.
Enviromental Manag. 86, 481–497.
Brown, R.C., Brown, T.R., 2014. Biorenewable Resources: Engineering New Products from
Agriculture: Second Edition, Biorenewable Resources: Engineering New Products from
Agriculture: Second Edition. doi:10.1002/9781118524985
Calicioglu, O., Brennan, R.A., 2018. Sequential ethanol fermentation and anaerobic digestion
increases bioenergy yields from duckweed. Bioresour. Technol. 257, 344–348.
doi:10.1016/j.biortech.2018.02.053
Calicioglu, O., Demirer, G.N., 2017. Carbon-to-nitrogen and substrate-to-inoculum ratio
adjustments can improve co-digestion performance of microalgal biomass obtained from
domestic wastewater treatment. Environ. Technol. (United Kingdom) 0, 1–11.
doi:10.1080/09593330.2017.1398784
Calicioglu, O., Demirer, G.N., 2017. Carbon-to-nitrogen and substrate-to-inoculum ratio
155
adjustments can improve co-digestion performance of microalgal biomass obtained from
domestic wastewater treatment. Environ. Technol. (United Kingdom).
doi:10.1080/09593330.2017.1398784
Calicioglu, O., Shreve, M.J., Richard, T.L., Brennan, R.A., 2018. Effect of pH and temperature
on microbial community structure and carboxylic acid yield during the acidogenic digestion
of duckweed. Biotechnol. Biofuels 1–19. doi:10.1186/s13068-018-1278-6
Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K.,
Fierer, N., Peña, A.G., Goodrich, J.K., Gordon, J.I., Huttley, G.A., Kelley, S.T., Knights, D.,
Koenig, J.E., Ley, R.E., Lozupone, C.A., McDonald, D., Muegge, B.D., Pirrung, M.,
Reeder, J., Sevinsky, J.R., Turnbaugh, P.J., Walters, W.A., Widmann, J., Yatsunenko, T.,
Zaneveld, J., Knight, R., 2010. QIIME allows analysis of high-throughput community
sequencing data. Nat. Methods 7, 335–6. doi:10.1038/nmeth.f.303
Chaiprapat, S., Cheng, J.J., Classen, J.J., Liehr, S.K., 2005. Role of internal nutrient storage in
duckweed growth for swine wastewater treatment. Trans. Asae 48, 2247–2258.
doi:10.1016/j.ecoleng.2013.12.055
Chen, Q., Jin, Y., Zhang, G., Fang, Y., Xiao, Y., Zhao, H., 2012. Improving production of
bioethanol from duckweed (Landoltia punctata) by pectinase pretreatment. Energies 5,
3019–3032. doi:10.3390/en5083019
Chen, Y., Cheng, J.J., Creamer, K.S., 2008. Inhibition of anaerobic digestion process: A review.
Bioresour. Technol. 99, 4044–4064. doi:10.1016/j.biortech.2007.01.057
Cheng, J., Landesman, L., Bergmann, B. a, Classen, J.J., Howard, J.W., Yamamoto, Y.T., 2002.
Nutrient Removal from swine lagoon liquid by Lemna minor 8627. Trans. ASAE 45, 1003–
1010.
Cheng, J.J., Stomp, A.M., 2009. Growing Duckweed to recover nutrients from wastewaters and
for production of fuel ethanol and animal feed. Clean - Soil, Air, Water 37, 17–26.
156
doi:10.1002/clen.200800210
Cherubini, F., 2010. The biorefinery concept: Using biomass instead of oil for producing energy
and chemicals. Energy Convers. Manag. 51, 1412–1421.
doi:10.1016/j.enconman.2010.01.015
Chynoweth, D.P., Turick, C.E., Owens, J.M., Jerger, D.E., Peck, M.W., 1993a. Biochemical
methane potential of biomass and waste feedstocks. Biomass and Bioenergy 5, 95–111.
doi:10.1016/0961-9534(93)90010-2
Chynoweth, D.P., Turick, C.E., Owens, J.M., Jerger, D.E., Peck, M.W., 1993b. Biochemical
methane potential of biomass and waste feedstocks. Biomass and Bioenergy 5, 95–111.
doi:10.1016/0961-9534(93)90010-2
Clarens, A.F., Resurreccion, E.P., White, M.A., Colosi, L.M., 2010. Environmental life cycle
comparison of algae to other bioenergy feedstocks. Environ. Sci. Technol. 44, 1813–1819.
doi:10.1021/es902838n
Clark, P.B., Hillman, P.F., Fellow, M., 1996. Enhancement of Anaerobic Digestion Using
Duckweed (Lemna minor) Enriched with Iron. Water Environ. 10, 92–95.
doi:10.1111/j.1747-6593.1996.tb00015.x
Collet, C., Adler, N., Schwitzguébel, J.P., Péringer, P., 2004. Hydrogen production by
Clostridium thermolacticum during continuous fermentation of lactose. Int. J. Hydrogen
Energy 29, 1479–1485. doi:10.1016/j.ijhydene.2004.02.009
Concepts, B., n.d. Bioprocess Engineering Basic Concepts.
Cui, W., Cheng, J.J., 2015. Growing duckweed for biofuel production: A review. Plant Biol. 17,
16–23. doi:10.1111/plb.12216
Culley, D.D., Rejmankova, E., Kvet, J., Frye, J.B., 1981. Production , Chemical Quality and Use
of Duckweeds ( Lemnaceae ) in Aquaculture , Waste Management , and Animal Feeds. J.
World Maricul. Soc. 12, 27–49. doi:10.1111/j.1749-7345.1981.tb00273.x
157
Dahiya, S., Sarkar, O., Swamy, Y. V., Venkata Mohan, S., 2015. Acidogenic fermentation of
food waste for volatile fatty acid production with co-generation of biohydrogen. Bioresour.
Technol. 182, 103–113. doi:10.1016/j.biortech.2015.01.007
Dale, B.E., Bals, B.D., Kim, S., Eranki, P., 2010. Biofuels done right: Land efficient animal feeds
enable large environmental and energy benefits. Environ. Sci. Technol. 44, 8385–8389.
doi:10.1021/es101864b
Datta, R., 1981. Acidogenic fermentation of corn stover. Biotechnol. Bioeng. 23, 61–77.
doi:10.1002/bit.260230106
Davis, R., Markham, J., Kinchin, C., Grundl, N., Tan, E.C.D., Humbird, D., 2016. Process Design
and Economics for the Production of Algal Biomass: Algal Biomass Production in Open
Pond Systems and Processing Through Dewatering for Downstream Conversion.
doi:10.2172/1239893
Dererie, D.Y., Trobro, S., Momeni, M.H., Hansson, H., Blomqvist, J., Passoth, V., Schnürer, A.,
Sandgren, M., Ståhlberg, J., 2011. Improved bio-energy yields via sequential ethanol
fermentation and biogas digestion of steam exploded oat straw. Bioresour. Technol. 102,
4449–4455. doi:10.1016/j.biortech.2010.12.096
Dong, B., Adams, E.E., 2012. Life-Cycle Assessment of Wastewater Treatment Plants by Life-
Cycle Assessment of Wastewater Treatment Plants by.
E.M. Siedlecka, J. Kumirska, T. Ossowski, P. Glamowski, M.G., J. Gajdus, Z. Kaczyński, P.S.,
2008. Determination of volatile fatty acids in environmental aqueous samples. Polish J.
Environ. Stud. 17, 351–356.
Edgar, R.C., 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics
26, 2460–2461. doi:10.1093/bioinformatics/btq461
El-Mashad, H.M., 2013. Kinetics of methane production from the codigestion of switchgrass and
Spirulina platensis algae. Bioresour. Technol. 132, 305–312.
158
doi:10.1016/j.biortech.2012.12.183
Eskicioglu, C., Ghorbani, M., 2011. Effect of inoculum/substrate ratio on mesophilic anaerobic
digestion of bioethanol plant whole stillage in batch mode. Process Biochem. 46, 1682–
1687. doi:10.1016/j.procbio.2011.04.013
Esquivel-Elizondo, S., Parameswaran, P., Delgado, A.G., Maldonado, J., Rittmann, B.E.,
Krajmalnik-Brown, R., 2016. Archaea and Bacteria Acclimate to High Total Ammonia in a
Methanogenic Reactor Treating Swine Waste. Archaea 2016. doi:10.1155/2016/4089684
Fernandes, T. V., Klaasse Bos, G.J., Zeeman, G., Sanders, J.P.M., van Lier, J.B., 2009. Effects of
thermo-chemical pre-treatment on anaerobic biodegradability and hydrolysis of
lignocellulosic biomass. Bioresour. Technol. 100, 2575–2579.
doi:10.1016/j.biortech.2008.12.012
Fong, J.C.N., Svenson, Æ.C.J., Bowman, J.P., Chen, Æ.B., Glenn, Æ.D.R., Neilan, B.A., Rogers,
Æ.P.L., 2006. Isolation and characterization of two novel ethanol-tolerant facultative-
anaerobic thermophilic bacteria strains from waste compost 363–372. doi:10.1007/s00792-
006-0507-2
Forsythe, W.C., Rykiel, E.J., Stahl, R.S., Wu, H., Schoolfield, R.M., 1995. A model comparison
for daylength as a function of latitude and day of year 80, 87–95.
Fraas, A.G., Munley, V.D., 1984. Municipal Wastewater Treatment Cost. J. Enviornmental Econ.
Manag. 11, 28–38.
Gaby, J.C., Zamanzadeh, M., Horn, S.J., 2017. The effect of temperature and retention time on
methane production and microbial community composition in staged anaerobic digesters fed
with food waste. Biotechnol. Biofuels 10, 302. doi:10.1186/s13068-017-0989-4
Garcia-Aguirre, J., Aymerich, E., González-Mtnez. de Goñi, J., Esteban-Gutiérrez, M., 2017.
Selective VFA production potential from organic waste streams: Assessing temperature and
pH influence. Bioresour. Technol. 244, 1081–1088. doi:10.1016/j.biortech.2017.07.187
159
Gaur, R.Z., Khan, A.A., Suthar, S., 2017. Effect of thermal pre-treatment on co-digestion of
duckweed (Lemna gibba) and waste activated sludge on biogas production. Chemosphere
174, 754–763. doi:10.1016/j.chemosphere.2017.01.133
Ge, X., Zhang, N., Phillips, G.C., Xu, J., 2012. Growing Lemna minor in agricultural wastewater
and converting the duckweed biomass to ethanol. Bioresour. Technol. 124, 485–488.
doi:10.1016/j.biortech.2012.08.050
González-Fernández, C., Sialve, B., Bernet, N., Steyer, J.P., 2012. Thermal pretreatment to
improve methane production of Scenedesmus biomass. Biomass and Bioenergy 40, 105–
111. doi:10.1016/j.biombioe.2012.02.008
Gulati, M., Kohlmann, K., Ladisch, M.R., Hespell, R., Bothast, R.J., 1996. Assessment of ethanol
production options for corn products. Bioresour. Technol. 58, 253–264. doi:10.1016/S0960-
8524(96)00108-3
Hamelers, H.V.M., 2001. A mathematical model for composting kinetics.
Hanshew, A.S., Mason, C.J., Raffa, K.F., Currie, C.R., 2013. Minimization of chloroplast
contamination in 16S rRNA gene pyrosequencing of insect herbivore bacterial communities.
J. Microbiol. Methods 95, 149–155. doi:10.1016/j.mimet.2013.08.007
Hatti-Kaul, R., Törnvall, U., Gustafsson, L., Börjesson, P., 2007. Industrial biotechnology for the
production of bio-based chemicals - a cradle-to-grave perspective. Trends Biotechnol. 25,
119–124. doi:10.1016/j.tibtech.2007.01.001
Hendriks, A.T.W.M., Zeeman, G., 2009. Pretreatments to enhance the digestibility of
lignocellulosic biomass. Bioresour. Technol. 100, 10–18.
doi:10.1016/j.biortech.2008.05.027
Holtzapple, M.T., Davison, R.R., Ross, M.K., Albrett-Lee, S., Nagwani, M., Lee, C.M., Lee, C.,
Adelson, S., Kaar, W., Gaskin, D., Shirage, H., Chang, N.S., Chang, V.S., Loescher, M.E.,
1999. Biomass conversion to mixed alcohol fuels using the MixAlco process. Appl.
160
Biochem. Biotechnol. 77–79, 609–631. doi:10.1385/ABAB:79:1-3:609
Holtzapple, M.T., Granda, C.B., 2009. Carboxylate platform: The MixAlco process part 1:
Comparison of three biomass conversion platforms. Appl. Biochem. Biotechnol. 156, 95–
106. doi:10.1007/s12010-008-8466-y
Huang, M., Fang, Y., Xiao, Y., Sun, J., Jin, Y., Tao, X., Ma, X., He, K., Zhao, H., 2014.
Proteomic analysis to investigate the high starch accumulation of duckweed (Landoltia
punctata) under nutrient starvation. Ind. Crops Prod. 59, 299–308.
doi:10.1016/j.indcrop.2014.05.029
Hung, C.H., Chang, Y.T., Chang, Y.J., 2011. Roles of microorganisms other than Clostridium
and Enterobacter in anaerobic fermentative biohydrogen production systems - A review.
Bioresour. Technol. 102, 8437–8444. doi:10.1016/j.biortech.2011.02.084
IDNR, 2007. Constructed Wetlands Technology Assessment and Design Guidance Iowa
Department of Natural Resources Constructed Wetland Technology Assessment and Design
Guidance.
Jain, S.K., Gujral, G.S., Jha, N.K., Vasudevan, P., 1992. Production of biogas from Azolla
pinnata R.Br and Lemna minor L.: Effect of heavy metal contamination. Bioresour.
Technol. 41, 273–277. doi:10.1016/0960-8524(92)90013-N
Jain, S.K., Gujral, G.S., Jha, N.K., Vasudevan, P., 1992. Production of biogas from Azolla
pinnata R.Br and Lemna minor L.: Effect of heavy metal contamination. Bioresour.
Technol. 41, 273–277. doi:10.1016/0960-8524(92)90013-N
Jankowska, E., Chwiałkowska, J., Stodolny, M., Oleskowicz-Popiel, P., 2015. Effect of pH and
retention time on volatile fatty acids production during mixed culture fermentation.
Bioresour. Technol. 190, 274–280. doi:10.1016/j.biortech.2015.04.096
Jørgensen, S.E., 2009. Applications in Ecological Engineering 380.
Jung, H., Baek, G., Kim, J., Shin, S.G., Lee, C., 2016. Mild-temperature thermochemical
161
pretreatment of green macroalgal biomass: Effects on solubilization, methanation, and
microbial community structure. Bioresour. Technol. 199, 326–335.
doi:10.1016/j.biortech.2015.08.014
Khan, Z., Dwivedi, A.K., Engineering, C., College, U.E., 2013. Fermentation of Biomass for
Production of Ethanol : A Review Abstract : 2 . Potential of Biomass. Univers. J. Environ.
Res. Technol. 3, 1–13.
Kumar, P., Barrett, D.M., Delwiche, M.J., Stroeve, P., 2009. Methods for pretreatment of
lignocellulosic biomass for efficient hydrolysis and biofuel production. Ind. Eng. Chem.
Res. 48, 3713–3729. doi:10.1021/ie801542g
Lasfar, S., Monette, F., Millette, L., Azzouz, A., 2007. Intrinsic growth rate: A new approach to
evaluate the effects of temperature, photoperiod and phosphorus-nitrogen concentrations on
duckweed growth under controlled eutrophication. Water Res. 41, 2333–2340.
doi:10.1016/j.watres.2007.01.059
Les, D.H., Crawford, D.J., Landolt, E., Gabel, J.D., Kimball, R.T., Rettig, J.H., 2002. Phylogeny
and Systematics of Lemnaceae, the Duckweed Family. Syst. Bot. 27, 221–240.
Liu, Y., 2010. Green algae as a substrate for biogas production - cultivation and biogas potentials.
Linköping University, Linköping, Sweden.
Ljungdahl, L.G., 1986. The autotrophic pathway of acetate synthesis in acetogenic bacteria.
Annu. Rev. Microbiol. 40, 415–450. doi:10.1146/annurev.mi.40.100186.002215
Lozupone, C.A., Hamady, M., Kelley, S.T., Knight, R., 2007. Quantitative and qualitative beta
diversity measures lead to different insights into factors that structure microbial
communities. Appl. Environ. Microbiol. doi:10.1128/AEM.01996-06
Luo, G., Karakashev, D., Xie, L., Zhou, Q., Angelidaki, I., 2011. Long-term effect of inoculum
pretreatment on fermentative hydrogen production by repeated batch cultivations:
Homoacetogenesis and methanogenesis as competitors to hydrogen production. Biotechnol.
162
Bioeng. 108, 1816–1827. doi:10.1002/bit.23122
Martin, M., 2011. Cutadapt removes adapter sequences from high-throughput sequencing reads.
EMBnet.journal 17, 10. doi:10.14806/ej.17.1.200
Maus, I., Kim, Y.S., Wibberg, D., Stolze, Y., Off, S., Antonczyk, S., Puehler, A., Scherer, P.,
Schlueter, A., 2017. Biphasic Study to Characterize Agricultural Biogas Plants by High-
Throughput 16S rRNA Gene Amplicon Sequencing and Microscopic Analysis. J. Microbiol.
Biotechnol. 27, 321–334. doi:10.4014/jmb.1605.05083
McCarty, P.L., 1964. Anaerobic Waste Treatment Fundamentals. Chem. Microbiol. 95, 107–112.
McCombs, P.J.A., Ralph, R.K., 1972. Protein, Nucleic Acid and Starch Metabolism in the
Duckweed, Spirodela oligorrhiza, Treated with Cytokinins. Biochem. J. 129, 403–417.
Metcalf, E., Eddy, H., 2003. Wastewater engineering: treatment and reuse. Tata McGraw-Hill
Publ. Co. Limited, 4th Ed. New Delhi, India. doi:10.1016/0309-1708(80)90067-6
Metcalf, Eddy, 1996. Wastewater Engineering. Treatment, Disposal and Reuse., Fourth Edi. ed.
McGraw - Hill Inc., New York.
Mitsch, W.J., Jørgensen, S.E., 2003. Ecological engineering: A field whose time has come. Ecol.
Eng. 20, 363–377. doi:10.1016/j.ecoleng.2003.05.001
Möller, K., Müller, T., 2012. Effects of anaerobic digestion on digestate nutrient availability and
crop growth: A review. Eng. Life Sci. 12, 242–257. doi:10.1002/elsc.201100085
Montingelli, M.E., Tedesco, S., Olabi, A.G., 2015. Biogas production from algal biomass: A
review. Renew. Sustain. Energy Rev. 43, 961–972. doi:10.1016/j.rser.2014.11.052
Mulbry, W., Westhead, E.K., Pizarro, C., Sikora, L., 2005. Recycling of manure nutrients: Use of
algal biomass from dairy manure treatment as a slow release fertilizer. Bioresour. Technol.
96, 451–458. doi:10.1016/j.biortech.2004.05.026
Müller, B., Sun, L., Westerholm, M., Schnürer, A., 2016. Bacterial community composition and
fhs profiles of low- and high-ammonia biogas digesters reveal novel syntrophic acetate-
163
oxidising bacteria. Biotechnol. Biofuels 9. doi:10.1186/s13068-016-0454-9
Muradov, N., Fidalgo, B., Gujar, A.C., Garceau, N., T-Raissi, A., 2012. Production and
characterization of Lemna minor bio-char and its catalytic application for biogas reforming.
Biomass and Bioenergy 42, 123–131. doi:10.1016/j.biombioe.2012.03.003
Murphy, C.F., Allen, D.T., 2011. Energy-water nexus for mass cultivation of algae. Environ. Sci.
Technol. 45, 5861–5868. doi:10.1021/es200109z
Mutel, C., 2017. Brightway: An open source framework for Life Cycle Assessment. J. Open
Source Softw. 2, 11–12. doi:10.21105/joss.00236
Naik, S.N., Goud, V. V., Rout, P.K., Dalai, A.K., 2010. Production of first and second generation
biofuels: A comprehensive review. Renew. Sustain. Energy Rev. 14, 578–597.
doi:10.1016/j.rser.2009.10.003
Niu, L., Song, L., Liu, X., Dong, X., 2009. Tepidimicrobium xylanilyticum sp. nov., an anaerobic
xylanolytic bacterium, and emended description of the genus Tepidimicrobium. Int. J. Syst.
Evol. Microbiol. 59, 2698–2701. doi:10.1099/ijs.0.005124-0
NREL, 2011. Process Design and Economics for Biochemical Conversion of Lignocellulosic
Biomass to Ethanol.
Okamoto, M., Miyahara, T., Mizuno, O., Noike, T., 2000. Biological hydrogen potential of
materials characteristic of the organic fraction of municipal solid wastes. Water Sci.
Technol. 41, 25–32.
Onyenwoke, R.U., Wiegel, J., 2015. Thermoanaerobacter, in: Bergey’s Manual of Systematics of
Archaea and Bacteria. American Cancer Society, pp. 1–29.
doi:10.1002/9781118960608.gbm00751
Oron, G., Porath, D., Wildschut, L.R., 1986. Wastewater Treatment and Renovation by Different
Duckweed Species. J. Environ. Eng. 112, 247–263. doi:10.1061/(ASCE)0733-
9372(1986)112:2(247)
164
Owen, W.F., Stuckey, D.C., Healy, J.B., Young, L.Y., McCarty, P.L., 1979. Bioassay for
monitoring biochemical methane potential and anaerobic toxicity. Water Res. 13, 485–492.
doi:10.1016/0043-1354(79)90043-5
Parada, A.E., Needham, D.M., Fuhrman, J.A., 2015. Every base matters: Assessing small subunit
rRNA primers for marine microbiomes with mock communities, time series and global field
samples. Environ. Microbiol. 18, 1403–1414. doi:10.1111/1462-2920.13023
Parkin, G.F., Owen, W.F., 1986. Fundamentals of Anaerobic Digestion of Wastewater Sludges,
Journal of Environmental Engineering. American Society of Civil Engineers.
doi:10.1061/(ASCE)0733-9372(1986)112:5(867)
Peters, J.B., 2003. Recommended Methods of Manure Analysis. Soils.Wisc.Edu.
doi:Recommended Methods of Manure Anaysis (A3769
Prasertsan, P., O-thong, S., 2009. Optimization and microbial community analysis for production
of biohydrogen from palm oil mill effluent by thermophilic fermentative process. Int. J.
Hydrogen Energy 34, 7448–7459. doi:10.1016/j.ijhydene.2009.04.075
Rabelo, S.C., Carrere, H., Maciel Filho, R., Costa, A.C., 2011. Production of bioethanol, methane
and heat from sugarcane bagasse in a biorefinery concept. Bioresour. Technol. 102, 7887–
7895. doi:10.1016/j.biortech.2011.05.081
Rios, L.M., Moore, C., Jones, P.R., 2007. Persistent organic pollutants carried by synthetic
polymers in the ocean environment. Mar. Pollut. Bull. 54, 1230–1237.
doi:10.1016/j.marpolbul.2007.03.022
Rognes, T., Flouri, T., Nichols, B., Quince, C., Mahé, F., 2016. VSEARCH: a versatile open
source tool for metagenomics. PeerJ 4, e2584. doi:10.7717/peerj.2584
Shao, L., Wu, Z., Zeng, L., Chen, Z.M., Zhou, Y., Chen, G.Q., 2013. Embodied energy
assessment for ecological wastewater treatment by a constructed wetland. Ecol. Modell.
252, 63–71. doi:10.1016/j.ecolmodel.2012.09.004
165
Shapouri, H., Salassi, M., 2006. The economic feasibility of ethanol production from sugar in the
United States. USDA Rep. 78. doi:10.1016/j.biortech.2007.11.013
Shetty, S.A., Marathe, N.P., Lanjekar, V., Ranade, D., Shouche, Y.S., 2013. Comparative genome
analysis of Megasphaera sp. reveals niche specialization and its potential role in the human
gut. PLoS One 8. doi:10.1371/journal.pone.0079353
Shin, H.S., Youn, J.H., Kim, S.H., 2004. Hydrogen production from food waste in anaerobic
mesophilic and thermophilic acidogenesis. Int. J. Hydrogen Energy 29, 1355–1363.
doi:10.1016/j.ijhydene.2003.09.011
Sims, A., Gajaraj, S., Hu, Z., 2013. Nutrient removal and greenhouse gas emissions in duckweed
treatment ponds. Water Res. 47, 1390–1398. doi:10.1016/j.watres.2012.12.009
Skillicorn, P., Spira, W., Journey, W., Riener, D.N., 1993. Duckweed aquaculture : a new aquatic
farming system for developing countries, Soil Science. Washington D.C.
Sluiter, A., Hames, B., Hyman, D., Payne, C., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D.,
Nrel, J.W., 2008. Determination of total solids in biomass and total dissolved solids in liquid
process samples. Natl. Renew. Energy Lab. 9. doi:NREL/TP-510-42621
Sluiter, A., Hames, B., Ruiz, R.O., Scarlata, C., Sluiter, J., Templeton, D., Energy, D. of, Dötsch,
A., Severin, J., Alt, W., Galinski, E. a, Kreft, J.-U., 2004. Determination of Ash in Biomass.
Microbiology 154, 2956–69. doi:TP-510-42622
Soda, S., Ohchi, T., Piradee, J., Takai, Y., Ike, M., 2015. Duckweed biomass as a renewable
biorefinery feedstock: Ethanol and succinate production from Wolffia globosa. Biomass and
Bioenergy 81, 364–368. doi:10.1016/j.biombioe.2015.07.020
Sorokin, I.D., Zadorina, E. V., Kravchenko, I.K., Boulygina, E.S., Tourova, T.P., Sorokin, D.Y.,
2008. Natronobacillus azotifigens gen. nov., sp. nov., an anaerobic diazotrophic
haloalkaliphile from soda-rich habitats. Extremophiles 12, 819–827. doi:10.1007/s00792-
008-0188-0
166
Speece, R., 2008. Anaerobic Biotechnology and Odor/corrosion Control for Municipalities and
Industries. Fields Publishing, Incorporated, Nashville, USA.
Sperling, M., Andreoli, C.V., Von, M., 2007. Sludge Treatment and Disposal, in: Biological
Wastewater Treatment. IWA Publishing, London, p. 10.
Steinbusch, K.J.J., Hamelers, H.V.M., Plugge, C.M., Buisman, C.J.N., 2011. Biological
formation of caproate and caprylate from acetate: fuel and chemical production from low
grade biomass. Energy Environ. Sci. 4, 216–224. doi:10.1039/C0EE00282H
Sträuber, H., Schröder, M., Kleinsteuber, S., 2012. Metabolic and microbial community dynamics
during the hydrolytic and acidogenic fermentation in a leach-bed process. Energy. Sustain.
Soc. 2, 13. doi:10.1186/2192-0567-2-13
Su, H., Zhao, Y., Jiang, J., Lu, Q., Li, Q., Luo, Y., Zhao, H., Wang, M., 2014. Use of duckweed
(Landoltia punctata) as a fermentation substrate for the production of higher alcohols as
biofuels. Energy and Fuels 28, 3206–3216. doi:10.1021/ef500335h
Tang, J., Yuan, Y., Guo, W.Q., Ren, N.Q., 2012. Inhibitory effects of acetate and ethanol on
biohydrogen production of Ethanoligenens harbinese B49. Int. J. Hydrogen Energy 37, 741–
747. doi:10.1016/j.ijhydene.2011.04.067
Tasker, T.L., Piotrowski, P.K., Dorman, F.L., Burgos, W.D., 2016. Metal Associations in
Marcellus Shale and Fate of Synthetic Hydraulic Fracturing Fluids Reacted at High Pressure
and Temperature. Environ. Eng. Sci. 33, 753–765. doi:10.1089/ees.2015.0605
Thanakoses, P., Black, A.S., Holtzapple, M.T., 2003. Fermentation of corn stover to carboxylic
acids. Biotechnol. Bioeng. 83, 191–200. doi:10.1002/bit.10663
Themelis, N.J., 2002. Anaerobic Digestion of Biodegradable Organics in Municipal Solid
Wastes. Found. Sch. Eng. Appl. Sci. Columbia Univ. Columbia University, New York, NY,
USA. doi:10.1016/j.biotechadv.2010.10.005
Theodorou, M.K., Williams, B.A., Dhanoa, M.S., McAllan, A.B., France, J., 1994. A simple gas
167
production method using a pressure transducer to determine the fermentation kinetics of
ruminant feeds. Anim. Feed Sci. Technol. 48, 185–197. doi:10.1016/0377-8401(94)90171-6
Tian, Z., Cabrol, L., Ruiz-Filippi, G., Pullammanappallil, P., 2014. Microbial ecology in
anaerobic digestion at agitated and non-agitated conditions. PLoS One 9.
doi:10.1371/journal.pone.0109769
Todd, J., Josephson, B., 1996. The design of living technologies for waste treatment. Ecol. Eng.
6, 109–136. doi:10.1016/0925-8574(95)00054-2
Triscari, P., Henderson, S., Reinhold, D., 2009. Anaerobic Digestion of Dairy Manure Combined
with Duckweed ( Lemnaceae ) Grand Sierra Resort and Casino. pp. 2–9.
Tuomela, M., Vikman, M., Hatakka, A., It??vaara, M., 2000. Biodegradation of lignin in a
compost environment: A review. Bioresour. Technol. 72, 169–183. doi:10.1016/S0960-
8524(99)00104-2
Uludag-Demirer, S., Demirer, G.N., Frear, C., Chen, S., 2008. Anaerobic digestion of dairy
manure with enhanced ammonia removal. J. Environ. Manage. 86, 193–200.
doi:10.1016/j.jenvman.2006.12.002
Vahlberg, C., Nordell, E., Wiberg, L., 2013. Method for correction of VFA loss in determination
of dry matter in biomass. Malmö.
Vintilǎ, T., Gherman, V., Bura, M., Dragomirescu, M., Ilie, D., Julean, C., Neo, S.I., 2013.
Biogas generation from corn stalks and corn stalks bagasse resulted from ethanol
production. Rom. Biotechnol. Lett. 18, 7212–7222.
Weightman, R., Sylvester-Bradley, R., Kindred, D., Brosnan, J., 2010. Growing wheat for
bioethanol production. HGCA Publ. 0300.
Wilkie, A.C., Mulbry, W.W., 2002. Recovery of Dairy Manure Nutrients by Benthic Freshwater
Algae Recovery of dairy manure nutrients by benthic freshwater algae. Bioresour. Technol.
8524, 81–91. doi:10.1016/S0960-8524(02)00003-2
168
Wise, D.L.L., Augenstein, D.C.C., Ryther, J.H.H., 1979. Methane fermentation of aquatic
biomass. Resour. Recover. Conserv. 4, 217–237. doi:10.1016/0304-3967(79)90002-7
Wu, C., Wang, Q., Xiang, J., Yu, M., Chang, Q., Gao, M., Sonomoto, K., 2015. Enhanced
Productions and Recoveries of Ethanol and Methane from Food Waste by a Three-Stage
Process. Energy and Fuels 29, 6494–6500. doi:10.1021/acs.energyfuels.5b01507
Wu, Q.L., Guo, W.Q., Zheng, H.S., Luo, H.C., Feng, X.C., Yin, R.L., Ren, N.Q., 2016.
Enhancement of volatile fatty acid production by co-fermentation of food waste and excess
sludge without pH control: The mechanism and microbial community analyses. Bioresour.
Technol. 216, 653–660. doi:10.1016/j.biortech.2016.06.006
Xiong, B., 2014. The Development of Carboxylic Acid Separation by Nanofiltration Membrane
for Carboxylate Platform Using Lignocellulosic Biomass.
Xiong, B., Richard, T.L., Kumar, M., 2015. Integrated acidogenic digestion and carboxylic acid
separation by nanofiltration membranes for the lignocellulosic carboxylate platform. J.
Memb. Sci. 489, 275–283. doi:10.1016/j.memsci.2015.04.022
Xiu, S.N., Shahbazi, A., Croonenberghs, J., Wang, L.J., 2010. Oil Production from Duckweed by
Thermochemical Liquefaction. Energy Sources, Part A Recover. Util. Environ. Eff. 32,
1293–1300. doi:10.1080/15567030903060408
Xu, J., Cheng, J.J., Stomp, A.M., 2012. Growing Spirodela polyrrhiza in Swine Wastewater for
the Production of Animal Feed and Fuel Ethanol: A Pilot Study. Clean - Soil, Air, Water 40,
760–765. doi:10.1002/clen.201100108
Xu, J., Cui, W., Cheng, J.J., Stomp, A.M., 2011. Production of high-starch duckweed and its
conversion to bioethanol. Biosyst. Eng. 110, 67–72.
doi:10.1016/j.biosystemseng.2011.06.007
Xu, J., Deshusses, M., 2015. Fermentation of swine wastewater-derived duckweed for
biohydrogen production. Int. J. Hydrogen Energy 40, 7028–7036.
169
doi:10.1016/j.ijhydene.2015.03.166
Xu, J., Shen, G., 2011. Growing duckweed in swine wastewater for nutrient recovery and
biomass production. Bioresour. Technol. 102, 848–853. doi:10.1016/j.biortech.2010.09.003
Xu, Y., Ma, S., Huang, M., Peng, M., Bog, M., Sree, K.S., Appenroth, K.J., Zhang, J., 2014.
Species distribution, genetic diversity and barcoding in the duckweed family (Lemnaceae).
Hydrobiologia 743, 75–87. doi:10.1007/s10750-014-2014-2
Xu, Z.-X., Wei, Z., Yin, H.-L., Huang, L.-H., 2010. Optimized design of natural ecological
wastewater treatment system based on water environment model of dynamic mesh
technique. J. Hydrodyn. Ser.B 22, 1–8. doi:10.1016/s1001-6058(09)60021-4
Yen, H.W., Brune, D.E., 2007. Anaerobic co-digestion of algal sludge and waste paper to
produce methane. Bioresour. Technol. 98, 130–134. doi:10.1016/j.biortech.2005.11.010
Yenigün, O., Demirel, B., 2013. Ammonia inhibition in anaerobic digestion: A review. Process
Biochem. 48, 901–911. doi:10.1016/j.procbio.2013.04.012
Yilmazel, Y.D., Demirer, G.N., 2011. Removal and recovery of nutrients as struvite from
anaerobic digestion residues of poultry manure. Environ. Technol. Middle East Technical
University, Ankara, Turkey. doi:10.1080/09593330.2010.512925
Yoon, C.G., Kim, S.B., Kwun, T.Y., Jung, K.W., 2008. Development of natural and ecological
wastewater treatment system for decentralized community in Korea. Paddy Water Environ.
6, 221–227. doi:10.1007/s10333-008-0109-y
Yu, C., Sun, C., Yu, L., Zhu, M., Xu, H., Zhao, J., Ma, Y., Zhou, G., 2014. Comparative analysis
of duckweed cultivation with sewage water and SH media for production of fuel ethanol.
PLoS One 9, 1–15. doi:10.1371/journal.pone.0115023
Yu, G.-H.H., He, P.J.P.-P.P.-J.J., Shao, L.-M.M., He, P.J.P.-P.P.-J.J., 2008. Toward
understanding the mechanism of improving the production of volatile fatty acids from
activated sludge at pH 10.0. Water Res. 42, 4637–4644. doi:10.1016/j.watres.2008.08.018
170
Yuan, H., Chen, Y., Zhang, H., Jiang, S., Zhou, Q., Gu, G., 2006. Improved bioproduction of
short-chain fatty acids (SCFAs) from excess sludge under alkaline conditions. Environ. Sci.
Technol. 40, 2025–2029. doi:10.1021/es052252b
Yue, Z.B., Yu, H.Q., Hu, Z.H., Harada, H., Li, Y.Y., 2008. Surfactant-enhanced anaerobic
acidogenesis of Canna indica L. by rumen cultures. Bioresour. Technol. 99, 3418–3423.
doi:10.1016/j.biortech.2007.08.010
Zhao, X., Elliston, A., Collins, S.R.A., Moates, G.K., Coleman, M.J., Waldron, K.W., 2014.
Enzymatic saccharification of duckweed (Lemna minor) biomass without thermophysical
pretreatment. Biomass and Bioenergy 47, 354–361. doi:10.1016/j.biombioe.2012.09.025
Zhao, Y., Fang, Y., Jin, Y., Huang, J., Bao, S., Fu, T., He, Z., Wang, F., Wang, M., Zhao, H.,
2015. Pilot-scale comparison of four duckweed strains from different genera for potential
application in nutrient recovery from wastewater and valuable biomass production. Plant
Biol. 17, 82–90. doi:10.1111/plb.12204
Zhao, Y., Fang, Y., Jin, Y., Huang, J., Bao, S., He, Z., Wang, F., Zhao, H., 2014. Effects of
operation parameters on nutrient removal from wastewater and high-protein biomass
production in a duckweed-based (Lemma aequinoctialis) pilot-scale system. Water Sci.
Technol. 70, 1195–1204. doi:10.2166/wst.2014.334
Zhao, Z., Li, Y., Quan, X., Zhang, Y., 2017. New Application of Ethanol-Type Fermentation:
Stimulating Methanogenic Communities with Ethanol to Perform Direct Interspecies
Electron Transfer. ACS Sustain. Chem. Eng. 5, 9441–9453.
doi:10.1021/acssuschemeng.7b02581
Zhao, Z., Zhang, Y., Yu, Q., Dang, Y., Li, Y., Quan, X., 2016. Communities stimulated with
ethanol to perform direct interspecies electron transfer for syntrophic metabolism of
propionate and butyrate. Water Res. 102, 475–484. doi:10.1016/j.watres.2016.07.005
171
Appendix A
Chapter 2 Additional File
Effect of Low Temperature Thermal Pretreatment on Anaerobic Digestibility of
Living Filter Duckweed
Thermal pretreatment has been proven to be effective for many feedstocks such as corn
stover, municipal organic wastes, and other complex materials (Liu, 2010). Moreover, thermal
pretreatment at low temperatures (lower than 100 oC) has been stated to be the most effective
method in terms of efficiency, economic cost, and environmental impact.
To evaluate the effect of heat on the anaerobic digestibility of duckweed and to compare
the biomethane production yields with those of fermentation effluents, heat pretreatment was
applied to LF duckweed. For this purpose, duckweed was subjected to the heat regimes of the
liquefaction and saccharification processes, without pH adjustment or enzyme addition. That is,
10 g dry duckweed was added to 200 mL distilled water, autoclaved for 1 h at 15 psi and 95 °C,
cooled, and then incubated at 50 °C while mixing at 120 rpm for 24 h. The resulting slurry was
also used as a substrate for BMP assays.
Thermal pretreatment did not show a significant impact and caused only a slight (4.2 %)
increase in biomethane yields from 258 to 269 mL CH4/g VSadded in reactors with an S/I of 1.0.
This value is quite low compared to others reported in the literature for low-temperature thermal
pretreatment of other biomass types. For instance, in a study conducted by Vintilǎ et al. (2013),
thermal hydrolysis of microalgae was found to be effective in biomethane production
enhancement, increasing the yield by 46 %. In another thermal pretreatment study, Scenedesmus
biomass at 70 °C and 90 °C produced similar organic material and ammonia release. However,
172
Figure A-1: Cumulative methane production in test reactors fed with raw Living-Filter duckweed (LF), fermented Living-Filter duckweed (FLF), and heat-pretreated Living-Filter
duckweed (HLF): A) S/I = 0.5, without Vanderbilt Medium (VM); B) S/I = 0.5, with Vanderbilt Medium (VM); C) S/I = 1.0, without Vanderbilt Medium (VM); D) S/I = 1.0, with Vanderbilt
Medium (VM).
the higher temperature yielded higher biomethane concentrations, as the damage to the cell wall
was higher (González-Fernández et al., 2012). In addition, considering high biomethane
productivities achieved in the fermented duckweed reactors even after some portion of the
biomass was converted and removed in the form of ethanol, it can be stated that low temperature
thermal pretreatment was not sufficient to damage the cell wall structure to an extent that it could
be hydrolyzed further during the anaerobic digestion process.
0
100
200
300
400
0 10 20 30 40 50
ml
CH
4/g
VS
ad
ded
Time (Days)
S/I = 0.5; Without VM
LF 0.5 FLF 0.5 HLF 0.5
A)
0
100
200
300
400
0 10 20 30 40 50
ml
CH
4/g
VS
ad
ded
Time (Days)
S/I = 0.5; With VM
LF 0.5 VM FLF 0.5 VM HLF 0.5 VM
B)
0
100
200
300
400
0 10 20 30 40 50
ml
CH
4/g
VS
ad
ded
Time (Days)
S/I = 1.0; Without VM
LF 1.0 FLF 1.0 HLF 1.0
C)
0
100
200
300
400
0 10 20 30 40 50
ml
CH
4/g
VS
ad
ded
Time (Days)
S/I = 1.0; With VM
LF 1.0 VM FLF 1.0 VM HLF 1.0 VM
D)
173
Appendix B
Chapter 3 Additional File
Effect of pH and Temperature on Microbial Community Structure and Carboxylic
Acid Yield during the Acidogenic Digestion of Duckweed
Table B-1: Final headspace and overall recovered biogas volumetric compositions in reactors at
final time point
Reactors: FINAL DAY HEADSPACE GAS
COMPOSITION (%)
OVERALL COMPOSITION OF
RECOVERED BIOGAS (%) Hydrogen Methane Carbon
Dioxide
Hydrogen Methane Carbon Dioxide
BAM1 6.70 0.00 84.10 2.77 0.03 56.32
BAM2 10.27 0.45 81.17 4.56 0.02 66.39
AM1 0.00 32.73 71.93 0.58 19.74 56.95
AM2 2.57 31.11 73.68 0.63 23.36 58.85
AM3 0.00 30.99 73.47 0.80 20.67 61.67
CAM1 0.00 1.51 3.53 0.00 0.02 1.46
CAM2 0.00 1.20 3.06 0.00 0.05 1.77
BAT1 54.95 0.00 51.80 39.16 0.00 47.62
BAT2 53.70 0.00 46.93 34.24 0.00 39.55
AT1 45.58 0.00 59.16 44.07 0.00 57.97
AT2 9.02 0.00 68.02 45.53 0.00 63.04
AT3 25.20 0.00 70.79 42.75 0.00 55.38
CAT1 1.37 0.00 4.30 0.00 0.00 3.12
CAT2 0.00 0.00 3.80 0.00 0.00 2.76
BBM1 0.00 0.00 1.35 0.00 0.00 1.37
BBM2 0.00 0.00 1.30 0.00 0.00 1.31
BM1 0.00 47.10 1.57 0.00 20.59 7.27
BM2 0.10 49.31 1.58 0.30 18.25 8.59
BM3 0.00 61.02 1.65 0.00 22.31 18.09
CBM1 0.00 0.95 1.35 0.00 0.05 4.38
CBM2 1.57 0.98 1.53 0.06 0.10 7.46
BBT1 0.00 0.00 1.58 0.42 0.00 1.84
BBT2 0.00 0.00 1.58 0.35 0.00 1.78
BT1 9.29 59.82 2.50 2.46 29.30 10.48
BT2 6.05 54.66 2.50 1.39 23.25 12.15
BT3 2.18 55.94 1.95 0.09 27.58 32.66
CBT1 0.00 0.74 1.91 0.00 0.00 2.31
CBT2 0.00 0.83 2.12 0.00 0.00 2.45
174
Table B-2: Total ammonia nitrogen and associated ammonium and ammonia concentrations in reactors at final time point
Reactors: TAN
(mg/L)
pH f_NH4+ f_NH3 [NH4
+]
(mg/L)
[NH3]
(mg/L)
BAM1 1528.20 5.30 0.99989 0.00011 1528 0.18
BAM2 1413.71 5.30 0.99989 0.00011 1414 0.16
AM1 1714.04 5.30 0.99989 0.00011 1714 0.20
AM2 1700.04 5.30 0.99989 0.00011 1700 0.20
AM3 1749.52 5.30 0.99989 0.00011 1749 0.20
CAM1 157.14 5.30 0.99989 0.00011 157 0.02
CAM2 164.38 5.30 0.99989 0.00011 164 0.02
BAT1 559.86 5.30 0.99989 0.00011 560 0.06
BAT2 588.08 5.30 0.99989 0.00011 588 0.07
AT1 1074.24 5.30 0.99989 0.00011 1074 0.12
AT2 1142.35 5.30 0.99989 0.00011 1142 0.13
AT3 1043.85 5.30 0.99989 0.00011 1044 0.12
CAT1 385.57 5.30 0.99989 0.00011 386 0.04
CAT2 359.62 5.30 0.99989 0.00011 360 0.04
BBM1 2191.90 9.20 0.52301 0.47699 1146 1045.51
BBM2 2019.39 9.20 0.52301 0.47699 1056 963.23
BM1 2182.93 9.20 0.52301 0.47699 1141 1041.24
BM2 2264.96 9.20 0.52301 0.47699 1185 1080.36
BM3 2311.85 9.20 0.52301 0.47699 1209 1102.73
CBM1 311.56 9.20 0.52301 0.47699 163 148.61
CBM2 297.82 9.20 0.52301 0.47699 155 142.06
BBT1 1466.83 9.20 0.52301 0.47699 767 699.66
BBT2 1898.98 9.20 0.52301 0.47699 993 905.79
BT1 2668.46 9.20 0.52301 0.47699 1396 1272.83
BT2 2614.33 9.20 0.52301 0.47699 1367 1247.01
BT3 2746.12 9.20 0.52301 0.47699 1436 1309.87
CBT1 398.42 9.20 0.52301 0.47699 208 190.04
CBT2 387.15 9.20 0.52301 0.47699 203 184.67
175
Table B-3: Carbon balance details of reactors at initial and final time points
INITIAL (TC AS % INITIAL TC) FINAL (TC AS % INITIAL TC) CLOSURE
REACTORS: Inoculum Alkalinity Duckweed Total Soluble Particulate Solid Gaseous Total (%)
BAM1 0 0 100 100 33.8 0 46.3 6.5 86.6 86.6
BAM2 0 0 100 100 34.6 0 46.2 6.5 87.3 87.3
AM1 11 0 89 100 14.3 15.4 48.9 13.7 92.4 92.4
AM2 11 0 89 100 14.9 18.1 28.2 16 77.2 77.2
AM3 11 0 89 100 15.2 8.1 41.6 14.3 79.2 79.2
BAT1 0 0 100 100 30.3 0 49.5 6.4 86.3 86.3
BAT2 0 0 100 100 27.1 0 59.4 4.5 91 91
AT1 11 0 89 100 19.9 3 65.9 5.4 94.3 94.3
AT2 11 0 89 100 22.9 5.5 56.4 5.8 90.6 90.6
AT3 11 0 89 100 20.6 5.3 58.9 5.9 90.7 90.7
BBM1 0 5 95 100 43.3 0 51.3 0.1 94.7 94.7
BBM2 0 5 95 100 52.8 1.4 43.3 0.1 97.5 97.5
BM1 10.2 4.5 85.2 100 51.4 0 26.9 2.3 80.6 80.6
BM2 10.2 4.5 85.2 100 52.9 0 27.9 2.4 83.2 83.2
BM3 10.2 4.5 85.2 100 53.9 6.9 27.5 3.6 91.9 91.9
BBT1 0 5 95 100 42.4 19 42.6 0.1 104 104
BBT2 0 5 95 100 41.2 18.2 40.9 0.1 100.3 100.3
BT1 10.2 4.5 85.2 100 53.1 5.5 36.1 3.2 98 98
BT2 10.2 4.5 85.2 100 50 5.2 41.3 2.8 99.3 99.3
BT3 10.2 4.5 85.2 100 48.7 6.9 35.4 4.3 95.3 95.3
176
Table B-4: Headspace pressure in reactors over time
DAILY PRESSURES (psi)
0.5 1 2 3 4 5 6 7 9 11 13 15 17 19 21
BAM
1
4.1
7
0 0.77 1.62 0 2.13 2.58 2.47 3.67 2.58 1.16 0.84 0.07 -
0.28
0.77
BAM
2
4.1
3
0 1.36 1.9 0 2.08 2.83 2.68 3.31 2.25 1.69 1.35 0.17 0 0.65
AM1 5.22
1.27
1.58 2.22 1.79 0.9 1.62 1.24 1.41 2.33 5.38 5.65 12.75
7.13 6.25
AM2 5.4
5
1.3
4
2.57 2.65 1.74 1.98 1.88 1.48 2.4 5.12 9.67 12.2
5
7.98 4.22 4
AM3 4.9 1.2
7
1.72 3.36 1.74 1.82 1.54 1.07 1.55 2.93 6.16 12.2
2
9.75 0.3 5.95
CAM
1
-0.5 0.9
1
-
0.09
0.09 -
0.21
0.22 0.03 0.19 -
0.34
0 0.07 -
0.25
-
0.59
-
0.63
-
0.23
CAM
2
0 0.72
-0.08
0.12 -0.28
0.15 0.14 0.14 -0.18
-0.06
0.07 0.09 -0.29
-0.32
0
BAT1 2.8
6
2.4 2.75 4.22 4.64 2.85 1.94 2.25 9.15 6.16 3.72 5.23 3.02 3.9
3
1.66
BAT2 4.5
4
2.4 2.39 1.98 1.37 0.73 1.44 1.5 4.28 2.58 2.26 3.8 3.48 4.13 2.05
AT1 14.9
1.84
8.67 7.04 2.51 2.4 1.18 1 0.16 0.14 0 1.01 0.92 3.4 0.11
AT2 14.
3
2.9
2
9.43 8.26 2.37 1.69 1.27 0.77 -
0.22
-
0.26
-
0.43
-
0.13
-
0.05
2.19 0
AT3 14.
5
3.3
3
7.22 7.19 2.69 1.87 0.98 1.12 1.63 1.15 0.87 2.76 0.71 0.84 -
0.31
CAT1 1.89
0.2 0.89 0.79 -0.59
0.62 -0.07
-0.59
-0.18
-0.32
-0.78
-0.31
-0.24
-0.19
-0.38
CAT2 1.0
1
0.2 0.8 0.2 -
0.63
0.51 -
0.29
-
0.21
-
0.08
-
0.29
-
0.98
-
0.31
-
0.31
0 -0.2
BBM
1
N/
A
1.0
5
0.82 -
0.24
-
0.06
-
0.63
0.14 0.06 -
0.06
0.53 -
0.05
-
0.14
-
0.11
0.45 0.11
BBM
2
N/A
1.01
1.19 0.34 -0.06
-1.1 -0.56
-0.71
-0.49
0.16 -0.05
0.07 0.22 0.14 0.14
BM1 N/
A
1.2
1
2 2.16 -
0.06
0.63 1.23 0.84 1.06 1.32 0.27 0.84 0.21 0.49 0.4
BM2 N/
A
1.2
3
2.19 1.87 -
0.64
-
0.27
0 1.11 0.98 0.62 0.45 0.71 0.56 0.55 0.63
BM3 N/A
1.18
1.98 2.03 4.41 4.82 -0.6 -0.32
-0.32
1.07 0.24 0.48 0.31 0.48 0.61
CBM
1
N/
A
0.7
7
0.44 -
0.29
0 -
0.45
-
0.06
-
0.14
-
0.14
0.2 -
0.27
-
0.27
-
0.11
-
0.05
0.06
CBM
2
N/
A
0.7
3
0.05 0.06 0.27 -0.3 0 0 -
0.05
0.34 -
0.29
-
0.26
-
0.14
-
0.09
0.09
BBT1 N/A
1.85
0.05 0.34 0.3 0.4 -0.27
0.07 0.16 0.09 0.16 -0.77
0.28 0.86 0.89
BBT2 N/
A
1.7
4
0.5 0.36 0.18 0.34 0.16 -
0.13
-
0.35
0 0.24 -0.2 0.59 0.58 1.27
BT1 N/
A
2.4
1
1.9 2.4 0.25 0.63 0.92 0.77 0.8 0.83 0.97 0.5 0.62 0.03 0.03
BT2 N/A
2.08
1.76 2.54 -1.08
-0.71
-1.01
-0.35
0.59 0.30 0.19 0.76 0.09 0.24 0.43
BT3 N/
A
2.3
9
2.41 7.49 4.74 6.13 -
2.81
-
2.37
-
2.37
0.56 0.82 -
0.47
0.39 0.39 0.02
CBT1 N/
A
1.4
8
0.62 0.66 0.41 0 0 0.09 -
0.05
0.24 0.08 -
0.76
-0.2 -
0.26
0
CBT2 N/A
1.21
0.71 0.19 0.25 0 -0.12
0 -0.56
-0.37
-0.57
-1.35
-0.58
-0.84
-0.42
177
Box B-1: One-way ANOVA and TUKEY comparison results of VFA Yields achieved in active (with inoculum) reactors
One-way ANOVA: AM, AT, BM, BT Method
Null hypothesis All means are equal
Alternative hypothesis Not all means are equal
Significance level α = 0.05 Equal variances were assumed for the analysis.
Factor Information Factor Levels Values
Factor 4 AM, AT, BM, BT
Analysis of Variance Source DF Seq SS Contribution Adj SS Adj MS F-Value P-Value
Factor 3 0.177185 98.93% 0.177185 0.059062 247.40 0.000
Error 8 0.001910 1.07% 0.001910 0.000239
Total 11 0.179095 100.00%
Model Summary S R-sq R-sq(adj) PRESS R-sq(pred)
0.0154510 98.93% 98.53% 0.0042972 97.60%
Means Factor N Mean StDev 95% CI
AM 3 0.04713 0.00388 (0.02656, 0.06770)
AT 3 0.09989 0.00832 (0.07932, 0.12047)
BM 3 0.3322 0.0294 (0.3116, 0.3528)
BT 3 0.29169 0.00290 (0.27112, 0.31226) Pooled StDev = 0.0154510
Tukey Pairwise Comparisons Grouping Information Using the Tukey Method and 95% Confidence
Factor N Mean Grouping
BM 3 0.3322 A
BT 3 0.29169 B
AT 3 0.09989 C
AM 3 0.04713 D Means that do not share a letter are significantly different.
178
Box B-2: One-way ANOVA and TUKEY comparison results of VFA yields achieved in blank (without inoculum) reactors
One-way ANOVA: BAM, BAT, BBM, BBT Method
Null hypothesis All means are equal
Alternative hypothesis Not all means are equal
Significance level α = 0.05
Rows unused 4 Equal variances were assumed for the analysis.
Factor Information Factor Levels Values
Factor 4 BAM, BAT, BBM, BBT
Analysis of Variance Source DF Seq SS Contribution Adj SS Adj MS F-Value P-Value
Factor 3 0.028830 87.80% 0.028830 0.009610 9.59 0.027
Error 4 0.004007 12.20% 0.004007 0.001002
Total 7 0.032837 100.00%
Model Summary S R-sq R-sq(adj) PRESS R-sq(pred)
0.0316508 87.80% 78.64% 0.0160283 51.19%
Means Factor N Mean StDev 95% CI
BAM 2 0.18705 0.00930 (0.12492, 0.24919)
BAT 2 0.0986 0.0338 (0.0365, 0.1607)
BBM 2 0.2187 0.0450 (0.1566, 0.2809)
BBT 2 0.0739 0.0275 (0.0117, 0.1360) Pooled StDev = 0.0316508
Tukey Pairwise Comparisons Grouping Information Using the Tukey Method and 95% Confidence
Factor N Mean Grouping
BBM 2 0.2187 A
BAM 2 0.18705 A B
BAT 2 0.0986 A B
BBT 2 0.0739 B Means that do not share a letter are significantly different.
179
Appendix C
Chapter 4 Additional File
Anaerobic Bioprocessing of Wastewater-Derived Duckweed: Maximizing Product
Yields in a Biorefinery Value Cascade
Table C-1: Descriptive table for values
Duckweed: Saccharified Pretreated Raw
Processes:
Variables::
Fer
men
tati
on
Aci
do
gen
ic D
iges
tion
Met
han
og
enic
Dig
esti
on
*
Fer
men
tati
on
Aci
do
gen
ic D
iges
tion
Met
han
og
enic
Dig
esti
on
Fer
men
tati
on
Aci
do
gen
ic D
iges
tion
Met
han
og
enic
Dig
esti
on
f residue,0 1.00 1.00 1.00 n.a. 1.00 1.00 n.a. 1.00 1.00
f redisue, 1 0.73 0.49 0.63 n.a. 0.65 0.70 n.a. 0.77 0.75
f residue,2 na 0.57 0.76; 0.83 n.a. na 0.81 n.a. na 0.79
f residue, 3 n.a. n.a. 0.84 n.a. n.a. na n.a. n.a. na
f recovered_product, 1 0.17 0.51 0.23 n.a. 0.34 0.19 n.a. 0.22 0.17
f recovered_product, 2 na 0.43 0.19, .15 n.a. na 0.15 n.a. na 0.15
f recovered_product, 3 na na 0.14 n.a. n.a. n.a. n.a. n.a. n.a.
f additives, 0 0.00 0.08 0.08 n.a. 0.00 0.00 n.a. 0.00 0.00
f additives, 1 0.10 0.15 0.37 n.a. 0.15 0.37 n.a. 0.15 0.37
f additives, 2 na 0.15 0.37; 0.37 n.a. 0.15 0.37 n.a. 0.15 0.37
f additives, 3 na na 0.37 n.a. na 0.37 n.a. na 0.37
f substrate 0 1.00 0.92 0.92 n.a. 1.00 1.00 n.a. 1.00 1.00
f substrate, 1 0.90 0.85 0.63 n.a. 0.85 0.63 n.a. 0.85 0.63
f substrate, 2 na 0.85 0.63; 0.63 n.a. na 0.63 n.a. na 0.63
f substrate, 3 na na 0.63 n.a. na na n.a. na na
f buffer, 0 0.00 0.08 0.08 n.a. 0.00 0.00 n.a. 0.00 0.00
f buffer, 1 0.08 0.04 0.03 n.a. 0.04 0.03 n.a. 0.04 0.03
f buffer, , 2 na 0.04 0.03 n.a. 0.04 0.03 n.a. 0.04 0.03
f buffer, 3 na na 0.03 n.a. na 0.03 n.a. na 0.03
beta1 0.00 1.00 1.00 0.00 1.00 1.00 0.00 1.00 1.00
beta2 na 1.00 1.00 na 1.00 1.00 na 1.00 1.00
beta3 na na 1.00 na na 1.00 na na 1.00
* If two values presented in one column, first is for sequential ethanol fermentation and methanogenic digestion, and
the second value s for sequential acidogenic digestion and methanogenic digestion processes.
180
Box C-2: Statistical Analysis
One-way ANOVA: SEVM, SEV, SVM, SEM, SE, SV, SM, ... ,
RVM, RV, RM * NOTE * Cannot draw the interval plot for the Tukey procedure. Interval plots for
comparisons are illegible with more than 45 intervals.
Method
Null hypothesis All means are equal
Alternative hypothesis Not all means are equal
Significance level α = 0.05 Equal variances were assumed for the analysis.
Factor Information
Factor Levels Values
Factor 13 SEVM, SEV, SVM, SEM, SE, SV, SM, PVM, PV, PM, RVM, RV, RM
Analysis of Variance
Source DF Adj SS Adj MS F-Value P-Value
Factor 12 0.95677 0.079730 23.49 0.000
Error 26 0.08824 0.003394
Total 38 1.04500
Model Summary
S R-sq R-sq(adj) R-sq(pred)
0.0582553 91.56% 87.66% 81.00%
Means
Factor N Mean StDev 95% CI
SEVM 3 0.6622 0.1007 (0.5931, 0.7313)
SEV 3 0.5567 0.0967 (0.4876, 0.6258)
SVM 3 0.6900 0.0823 (0.6208, 0.7591)
SEM 3 0.4070 0.0677 (0.3379, 0.4762)
SE 3 0.1909 0.0519 (0.1217, 0.2600)
SV 3 0.5693 0.0800 (0.5001, 0.6384)
SM 3 0.34150 0.01025 (0.27237, 0.41064)
PVM 3 0.5532 0.0373 (0.4840, 0.6223)
PV 3 0.3862 0.0332 (0.3171, 0.4553)
PM 3 0.29350 0.01625 (0.22437, 0.36264)
RVM 3 0.4528 0.0281 (0.3836, 0.5219)
RV 3 0.2516 0.0219 (0.1824, 0.3207)
RM 3 0.26152 0.00680 (0.19238, 0.33065)
181
Pooled StDev = 0.0582553
Tukey Pairwise Comparisons
Grouping Information Using the Tukey Method and 95% Confidence
Factor N Mean Grouping
SVM 3 0.6900 A
SEVM 3 0.6622 A
SV 3 0.5693 A B
SEV 3 0.5567 A B C
PVM 3 0.5532 A B C
RVM 3 0.4528 B C D
SEM 3 0.4070 B C D E
PV 3 0.3862 C D E
SM 3 0.34150 D E F
PM 3 0.29350 D E F
RM 3 0.26152 E F
RV 3 0.2516 E F
SE 3 0.1909 F Means that do not share a letter are significantly different.
182
Box C-3: Derivation of formulas
Total bioprpopduct yield:
Given that the substrate oprocess i it the residue of process (i-1), such that:
𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0(𝑔) = 𝑇𝐶𝑠𝑢𝑏𝑠𝑡𝑎𝑟𝑡𝑒,1(𝑔) 𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒 + 𝑓𝑎𝑑𝑑 = 1
𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 = 𝑇𝐶𝑟𝑒𝑎𝑐𝑡𝑜𝑟,0 ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 𝑤ℎ𝑒𝑟𝑒 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 = 1
𝑇𝐶𝑎𝑑𝑑,0(𝑔) = 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔) ∗𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1
𝑇𝐶𝑟𝑒𝑎𝑐𝑡𝑜𝑟,1(𝑔) = 𝑇𝐶𝑎𝑑𝑑,1 (𝑔) + 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔)
= 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔) ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1)
𝑇𝐶𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,1(𝑔) = 𝑇𝐶𝑟𝑒𝑎𝑐𝑡𝑜𝑟,1 (𝑔) ∗ 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,1(𝑔)
= 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔) ∗ 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,1(𝑔) ∗ (𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1)
𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1(𝑔) = 𝑇𝐶𝑟𝑒𝑎𝑐𝑡𝑜𝑟,1 (𝑔) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1(𝑔)
= 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1(𝑔) ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1)
𝑇𝐶𝑎𝑑𝑑,2 (𝑔) = 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 (𝑔) ∗ (1 +𝑓𝑎𝑑𝑑,2
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2)
𝑇𝐶𝑟𝑒𝑎𝑐𝑡𝑜𝑟,2 (𝑔) = 𝑇𝐶𝑎𝑑𝑑,2 (𝑔) + 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 (𝑔)
= 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 (𝑔) ∗ (1 +𝑓𝑎𝑑𝑑,2
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2)
𝑇𝐶𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,2 (𝑔) = 𝑇𝐶𝑟𝑒𝑎𝑐𝑡𝑜𝑟,2 (𝑔) ∗ 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,2
= 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 (𝑔) ∗ 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,2 ∗ (1 +𝑓𝑎𝑑𝑑,2
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2)
𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔) = 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 (𝑔) ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1)𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,2 ∗ (1 +
𝑓𝑎𝑑𝑑,2𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2
)
𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2(𝑔) = 𝑇𝐶𝑟𝑒𝑎𝑐𝑡𝑜𝑟,2 (𝑔) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2
= 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 (𝑔) ∗ (1 +𝑓𝑎𝑑𝑑,2
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2
𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2(𝑔) = 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2
∗ (1 +𝑓𝑎𝑑𝑑,2
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2)
𝑇𝐶𝑎𝑑𝑑,3 (𝑔) = 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2 (𝑔) ∗ (𝑓𝑎𝑑𝑑,3
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,3)
𝑇𝐶𝑟𝑒𝑎𝑐𝑡𝑜𝑟,3 (𝑔) = 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2 (𝑔) ∗ (1 +𝑓𝑎𝑑𝑑,3
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,3)
𝑇𝐶𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,3 (𝑔) = 𝑇𝐶𝑟𝑒𝑎𝑐𝑡𝑜𝑟,3 (𝑔) ∗ 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,3
183
= 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2 (𝑔) ∗ (1 +𝑓𝑎𝑑𝑑,3
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,3) ∗ 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,3
= 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2 ∗ (1 +
𝑓𝑎𝑑𝑑,2𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2
)
∗ (1 +𝑓𝑎𝑑𝑑,3
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,3) ∗ 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,3
𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,3 (𝑔) = 𝑇𝐶𝑟𝑒𝑎𝑐𝑡𝑜𝑟,3 (𝑔) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,3
= 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2 ∗ (1 +
𝑓𝑎𝑑𝑑,2𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2
)
∗ (1 +𝑓𝑎𝑑𝑑,3
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,3) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,3
For process 1:
𝑇𝐶𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,1 (𝑔)
𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔)= 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,1 ∗ (1 +
𝑓𝑎𝑑𝑑,1𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1
)
For process 2:
𝑇𝐶𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,2 (𝑔)
𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔)= 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 ∗ (1 +
𝑓𝑎𝑑𝑑,1𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1
) ∗ 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,2 ∗ (1 +𝑓𝑎𝑑𝑑,2
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2)
For process 3: 𝑇𝐶𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,3 (𝑔)
𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔)
= 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2 ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1) ∗ (1 +
𝑓𝑎𝑑𝑑,2𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2
)
∗ (1 +𝑓𝑎𝑑𝑑,3
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,3) ∗ 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,3
=∑𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑,𝑖
𝑛
𝑖=1
∗ (1 +𝑓𝑎𝑑𝑑,𝑖
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,𝑖) ∗∏𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,𝑗
𝑖=1
𝑗=0
∗ (1 +𝑓𝑎𝑑𝑑,𝑗
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,𝑗)
where 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 = 1 and 𝑓𝑎𝑑𝑑,0 = 0 Additives:
𝑇𝐶𝑎𝑑𝑑,1 (𝑔) = 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔) ∗ (𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 ∗ (
𝑓𝑎𝑑𝑑,0𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,0
)
𝑇𝐶𝑎𝑑𝑑,2 (𝑔) = [𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1)] ∗ (
𝑓𝑎𝑑𝑑,2𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2
)
𝑇𝐶𝑎𝑑𝑑,3 (𝑔) = 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0 (𝑔) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2
∗ (1 +𝑓𝑎𝑑𝑑,2
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2) (
𝑓𝑎𝑑𝑑,3𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,3
)
𝑇𝐶𝑎𝑑𝑑,1𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0
= ∑(𝑓𝑎𝑑𝑑,𝑖
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,𝑖)
𝑛
𝑖=1
∗∏𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,𝑗
𝑖=1
𝑗=0
∗ (1 +𝑓𝑎𝑑𝑑,𝑗
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,𝑗)
184
Losses:
𝑓𝑙𝑜𝑠𝑠 + 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒 + 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑 = 1 𝑓𝑙𝑜𝑠𝑠 = 1 − (𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒 + 𝑓𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑒𝑑) 𝑇𝐶𝑙𝑜𝑠𝑠 = 𝑇𝐶𝑟𝑒𝑎𝑐𝑡𝑜𝑟 ∗ 𝑓𝑙𝑜𝑠𝑠
𝑇𝐶𝑙𝑜𝑠𝑠,1 = (𝑇𝐶𝑎𝑑𝑑,1(𝑔) + 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0(𝑔)) ∗ 𝑓𝑙𝑜𝑠𝑠
= 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0(𝑔) ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1) ∗ 𝑓𝑙𝑜𝑠𝑠,1
𝑇𝐶𝑙𝑜𝑠𝑠,2 = 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0(𝑔) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1) ∗ 𝑓𝑙𝑜𝑠𝑠,2 ∗ (1 +
𝑓𝑎𝑑𝑑,2𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2
)
𝑇𝐶𝑙𝑜𝑠𝑠,3 = 𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0(𝑔) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,1 ∗ (1 +𝑓𝑎𝑑𝑑,1
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,1) ∗ 𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,2 ∗ (1 +
𝑓𝑎𝑑𝑑,2𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,2
)
∗ 𝑓𝑙𝑜𝑠𝑠,3 ∗ (1 +𝑓𝑎𝑑𝑑,3
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,3)
∑𝑇𝐶𝑙𝑜𝑠𝑠𝑇𝐶𝑟𝑒𝑠𝑖𝑑𝑢𝑒,0
= ∑𝑓𝑙𝑜𝑠𝑠,𝑖
𝑛
𝑖=1
∗ (1 +𝑓𝑎𝑑𝑑,𝑖
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,𝑖) ∗∏𝑓𝑟𝑒𝑠𝑖𝑑𝑢𝑒,𝑗
𝑖=1
𝑗=0
∗ (1 +𝑓𝑎𝑑,𝑗
𝑓𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒,𝑗)
Derivation 2: For a given process:
(𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡
+𝑇𝐶𝑎𝑑𝑑𝑡𝑖𝑣𝑒𝑠𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡
)
−1
= 𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠
𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑 + 𝑇𝐶𝑎𝑑𝑑𝑖𝑡𝑖𝑣𝑒𝑠
where
𝑇𝐶𝑎𝑑𝑑𝑖𝑡𝑖𝑣𝑒𝑠𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡
=𝑇𝐶𝑎𝑑𝑑𝑖𝑡𝑖𝑣𝑒𝑠𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑
∗ (𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡
𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑)−1
1
𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡
∗ (1 +𝑇𝐶𝑎𝑑𝑑𝑖𝑡𝑖𝑣𝑒𝑠𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑
)=
𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑
1 +𝑇𝐶𝑎𝑑𝑑𝑖𝑡𝑖𝑣𝑒𝑠𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑
Overall:
∑𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠
𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑 + 𝑇𝐶𝑎𝑑𝑑𝑖𝑡𝑖𝑣𝑒𝑠 =
𝑛
𝑖=0
∑ (𝑇𝐶𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑
)𝑛𝑖=0
𝑖
1 + 𝛽𝑖 ∑ (𝑇𝐶𝑎𝑑𝑑𝑖𝑡𝑖𝑣𝑒𝑠𝑇𝐶𝑑𝑢𝑐𝑘𝑤𝑒𝑒𝑑
)𝑛𝑖=0
𝑖
where βi = buffer assimilation capacity of the ith process
185
Appendix D
Chapter 5 Additional File
Techno-economic Analysis of Wastewater-Derived Duckweed Biorefinery and
Supply Chain System
Box D-1. Duckweed growth model
{
𝑟𝑖 = 𝑅 ∙ 𝜃1
((𝑇−𝑇𝑜𝑝) 𝑇𝑜𝑝⁄ )2
∙ 𝜃2((𝑇−𝑇𝑜𝑝) 𝑇𝑜𝑝⁄ )
∙ 𝜃3((𝐸−𝐸𝑜𝑝) 𝐸𝑜𝑝⁄ )
2
∙ 𝜃4((𝐸−𝐸𝑜𝑝) 𝐸𝑜𝑝⁄ )
∙𝐶𝑃
𝐶𝑃 + 𝐾𝑃∙
𝐾𝐼𝑃𝐾𝐼𝑃 + 𝐶𝑃
∙𝐶𝑁
𝐶𝑁 + 𝐾𝑁∙
𝐾𝐼𝑁𝐾𝐼𝑁 + 𝐶𝑁
𝐷 =𝐷𝐿 ∙ 𝐷𝑂
(𝐷𝐿 − 𝐷𝑂) ∙ 𝑒−𝑟𝑖∙𝑡 +𝐷𝑂
⟺ 𝑟𝑠 =1
𝑡∙ ln (
𝐷
𝐷𝑂) =
1
𝑡∙ ln (
𝐷𝐿(𝐷𝐿 − 𝐷𝑂) ∙ 𝑒
−𝑟𝑖∙𝑡 + 𝐷𝑂)
where KP, KIP, KN, and KIN are the saturation and the inhibition constants of P and N,
respectively. CP and CN are the P–N concentration (mg/L), respectively. R is a constant
(maximum intrinsic growth rate in day-1), T is the temperature in °C with Top being the
optimum temperature; E is the photoperiod (h), ri and rs are the intrinsic and specific growth
rates (day-1), respectively. Do is the initial mat density (g dry/m2) of the duckweed and DL is
the upper limit of the mat density beyond which the duckweed growth is strongly inhibited; t is
the duckweed retention time (day), and θ1-4 are nondimensional constants smaller than 1.
Box D-2. Day length model
𝜃 = 0.2163108 + 2 ∙ tan−1[0.9671396 ∙ tan[0.00860 ∙ ( 𝐽 − 186)]]
𝜙 = 𝑠𝑖𝑛−1[0.39795 ∙ cos 𝜃]
𝐷 = 24 −24
𝜋∙ cos−1 [
sin𝑝𝜋180 + sin
𝐿𝜋180 ∙ sin𝜙
cos𝐿𝜋180 ∙ cos𝜙
]
where p is daylength definition in degrees, θ and ϕ are the revolution and sun’s declination
angles in radians (northern latitudes are positive, while southern latitudes are negative), D is
the daylength in hours (including twilight), and L is the latitude in degrees.
186
Figure D-1: Aquatic weed harvester specifications
187
Table D-1: Discounted cash flow rate of return details for wastewater treatment-duckweed
production system.
-1 0 1 2 3 4 5
Fixed Capital Investment 3,184,963 2,123,309
Land 1,045,255
Working Capital 265,414
Loan Payment 474,653 474,653 474,653 474,653 474,653
Loan Interest payment 152,878 101,919 254,797 237,209 218,213 197,698 175,541
Loan principal 1,910,978 1,273,985 2,965,107 2,727,662 2,471,221 2,194,266 1,895,153
Biomass Sales 136,099 181,465 181,465 181,465 181,465
By-product credit 2,066,214 2,066,214 2,066,214 2,066,214 2,066,214
Total aAnnual sales 2,202,313 2,247,679 2,247,679 2,247,679 2,247,679
Annual manufacturing cost
Other variable cost 4,587 4,587 4,587 4,587 4,587
Fixed operating costs 790,360 790,360 790,360 790,360 790,360
Total Product cost 794,946 794,946 794,946 794,946 794,946
Annual depreciation
General plant writedown 4% 7% 7% 6% 6%
Depreciation charge 199,060 383,257 353,531 328,051 303,102
Remaining value 5,109,211 4,925,014 4,954,741 4,980,220 5,005,169
Net revenue 953,509 832,267 880,989 926,984 974,089
Losses forward 0 0 0 0
Taxable income 953,509 832,267 880,989 926,984 974,089
Income tax 333,728 291,293 308,346 324,444 340,931
Annual cash income 598,985 686,786 669,733 653,635 637,148
Discount factor 1 1 1 1 1 1 1
Annual Present value 6,971,521 544,532 567,592 503,181 446,442 395,619
Total capital investment+
interest 4,821,406 2,225,227
Net Present Value 0
188
6 7 8 9 10
Fixed Capital Investment
Land
Working Capital
Loan Payment 474,653 474,653 474,653 474,653 474,653
Loan Interest payment 151,612 125,769 97,858 67,715 35,160
Loan principal 1,572,112 1,223,228 846,433 439,494 0
Biomass Sales 181,465 181,465 181,465 181,465 181,465
By-product credit 2,066,214 2,066,214 2,066,214 2,066,214 2,066,214
Total aAnnual sales 2,247,679 2,247,679 2,247,679 2,247,679 2,247,679
Annual manufacturing cost
Other variable cost 4,587 4,587 4,587 4,587 4,587
Fixed operating costs 790,360 790,360 790,360 790,360 790,360
Total Product cost 794,946 794,946 794,946 794,946 794,946
Annual depreciation
General plant writedown 9% 9% 5%
Depreciation charge 473,498 474,029 289,832
Remaining value 4,834,774 4,834,243 5,018,440
Net revenue 827,623 852,935 1,065,043 1,385,018 1,417,573
Losses forward 0 0 0 0 0
Taxable income 827,623 852,935 1,065,043 1,385,018 1,417,573
Income tax 289,668 298,527 372,765 484,756 496,151
Annual cash income 688,411 679,552 605,314 493,323 481,929
Discount factor 1 1 0 0 0
Annual Present value 388,590 348,718 282,384 209,217 185,804
Total capital investment+ interest
Net Present Value
189
11 12 13 14 15
Fixed Capital Investment
Land
Working Capital
Loan Payment
Loan Interest payment
Loan principal
Biomass Sales 181,465 181,465 181,465 181,465 181,465
By-product credit 2,066,214 2,066,214 2,066,214 2,066,214 2,066,214
Total aAnnual sales 2,247,679 2,247,679 2,247,679 2,247,679 2,247,679
Annual manufacturing cost
Other variable cost 4,587 4,587 4,587 4,587 4,587
Fixed operating costs 790,360 790,360 790,360 790,360 790,360
Total Product cost 794,946 794,946 794,946 794,946 794,946
Annual depreciation
General plant writedown
Depreciation charge
Remaining value
Net revenue 1,452,733 1,452,733 1,452,733 1,452,733 1,452,733
Losses forward 0 0 0 0 0
Taxable income 1,452,733 1,452,733 1,452,733 1,452,733 1,452,733
Income tax 508,457 508,457 508,457 508,457 508,457
Annual cash income 944,276 944,276 944,276 944,276 944,276
Discount factor 0 0 0 0 0
Annual Present value 330,963 300,876 273,523 248,657 226,052
Total capital investment+ interest
Net Present Value
190
16 17 18 19 20
Fixed Capital Investment
Land
Working Capital
Loan Payment
Loan Interest payment
Loan principal
Biomass Sales 181,465 181,465 181,465 181,465 181,465
By-product credit 2,066,214 2,066,214 2,066,214 2,066,214 2,066,214
Total aAnnual sales 2,247,679 2,247,679 2,247,679 2,247,679 2,247,679
Annual manufacturing cost
Other variable cost 4,587 4,587 4,587 4,587 4,587
Fixed operating costs 790,360 790,360 790,360 790,360 790,360
Total Product cost 794,946 794,946 794,946 794,946 794,946
Annual depreciation
General plant writedown
Depreciation charge
Remaining value
Net revenue 1,452,733 1,452,733 1,452,733 1,452,733 1,452,733
Losses forward 0 0 0 0 0
Taxable income 1,452,733 1,452,733 1,452,733 1,452,733 1,452,733
Income tax 508,457 508,457 508,457 508,457 508,457
Annual cash income 944,276 944,276 944,276 944,276 944,276
Discount factor 0 0 0 0 0
Annual Present value 205,502 186,820 169,836 154,397 140,361
Total capital investment+
interest
Net Present Value
191
21 22 23 24 25
Fixed Capital Investment
Land
Working Capital
Loan Payment
Loan Interest payment
Loan principal
Biomass Sales 181,465 181,465 181,465 181,465 181,465
By-product credit 2,066,214 2,066,214 2,066,214 2,066,214 2,066,214
Total aAnnual sales 2,247,679 2,247,679 2,247,679 2,247,679 2,247,679
Annual manufacturing cost
Other variable cost 4,587 4,587 4,587 4,587 4,587
Fixed operating costs 790,360 790,360 790,360 790,360 790,360
Total Product cost 794,946 794,946 794,946 794,946 794,946
Annual depreciation
General plant writedown
Depreciation charge
Remaining value
Net revenue 1,452,733 1,452,733 1,452,733 1,452,733 1,452,733
Losses forward 0 0 0 0 0
Taxable income 1,452,733 1,452,733 1,452,733 1,452,733 1,452,733
Income tax 508,457 508,457 508,457 508,457 508,457
Annual cash income 944,276 944,276 944,276 944,276 944,276
Discount factor 0 0 0 0 0
Annual Present value 127,601 116,001 105,455 95,868 87,153
Total capital investment+ interest
Net Present Value
192
26 27 28 29 30
Fixed Capital Investment
Land -1,045,255
Working Capital -265,414
Loan Payment
Loan Interest payment
Loan principal
Biomass Sales 181,465 181,465 181,465 181,465 181,465
By-product credit 2,066,214 2,066,214 2,066,214 2,066,214 2,066,214
Total aAnnual sales 2,247,679 2,247,679 2,247,679 2,247,679 2,247,679
Annual manufacturing cost
Other variable cost 4,587 4,587 4,587 4,587 4,587
Fixed operating costs 790,360 790,360 790,360 790,360 790,360
Total Product cost 794,946 794,946 794,946 794,946 794,946
Annual depreciation
General plant writedown
Depreciation charge
Remaining value
Net revenue 1,452,733 1,452,733 1,452,733 1,452,733 1,452,733
Losses forward 0 0 0 0 0
Taxable income 1,452,733 1,452,733 1,452,733 1,452,733 1,452,733
Income tax 508,457 508,457 508,457 508,457 508,457
Annual cash income 944,276 944,276 944,276 944,276 944,276
Discount factor 0 0 0 0 0
Annual Present value 79,230 72,027 65,479 59,527 54,115
Total capital investment+ interest -75,113
Net Present Value
193
Table D-2: Discounted cash flow rate of return details for the biorefinery system.
-1 0 1 2 3 4 5
Fixed Capital Investment 12,961,078 8,640,719
Land 72,000
Working Capital 1,080,090
Loan Payment -1,931,583 -1,931,583 -1,931,583 -1,931,583 -1,931,583
Loan Interest payment 622,132 414,754 1,036,886 965,310 888,009 804,523 714,358
Loan principal 7,776,647 5,184,431 12,066,381 11,100,109 10,056,535 8,929,475 7,712,250
Bioethanol sales 2,700,426 2,700,426 2,700,426 2,700,426 2,700,426
By-product credit 577,167 577,167 577,167 577,167 577,167
Total annual sales 3,277,593 3,277,593 3,277,593 3,277,593 3,277,593
Annual manufacturing cost
Feedstock cost 176023.113 176023.113 176023.113 176023.113 176023.113
Other variable cost 4,932 4,932 4,932 4,932 4,932
Fixed operating costs 1,602,520 1,602,520 1,602,520 1,602,520 1,602,520
Total Product cost 1,783,476 1,783,476 1,783,476 1,783,476 1,783,476
Annual depreciation
General plant writedown 14.29% 24.49% 17.49% 12.49% 8.93%
Depreciation charge 3,086,897 5,290,280 3,778,154 2,698,064 1,929,040
Remaining value 18,514,900 16,311,516 17,823,642 18,903,732 19,672,756
Net revenue -2,629,665 -4,761,473 -3,172,045 -2,008,469 -1,149,281
Losses forward -2,629,665 -7,391,138 -10,563,183 -12,571,652
Taxable income -2,629,665 -7,391,138 -10,563,183 -12,571,652 -13,720,933
Income tax 0 0 0 0 0
Annual cash income 457,232 528,807 606,109 689,595 779,760
Discount factor 1.02 1.00 0.98 0.95 0.93 0.91 0.89
Annual Present value 22,487,886 446,297 503,818 563,657 625,960 690,878
Total capital investment+
interest 13,989,762 9,055,473
Net Present Value 0
194
6 7 8 9 10
Fixed Capital Investment
Land
Working Capital
Loan Payment -1,931,583 -1,931,583 -1,931,583 -1,931,583 -1,931,583
Loan Interest payment 616,980 511,812 398,230 275,562 143,080
Loan principal 6,397,647 4,977,876 3,444,523 1,788,503 0
Bioethanol sales 2,700,426 2,700,426 2,700,426 2,700,426 2,700,426
By-product credit 577,167 577,167 577,167 577,167 577,167
Total annual sales 3,277,593 3,277,593 3,277,593 3,277,593 3,277,593
Annual manufacturing cost
Feedstock cost 176023.113 176023.113 176023.113 176023.113 176023.113
Other variable cost 4,932 4,932 4,932 4,932 4,932
Fixed operating costs 1,602,520 1,602,520 1,602,520 1,602,520 1,602,520
Total Product cost 1,783,476 1,783,476 1,783,476 1,783,476 1,783,476
Annual depreciation
General plant writedown 8.92% 8.93% 4.46%
Depreciation charge 1,926,880 1,929,040 963,440
Remaining value 19,674,916 19,672,756 20,638,356
Net revenue -1,049,742 -946,734 132,448 1,218,556 1,351,038
Losses forward -13,720,933 -14,770,675 -15,717,409 -15,584,962 -14,366,406
Taxable income -14,770,675 -15,717,409 -15,584,962 -14,366,406 -13,015,368
Income tax 0 0 0 0 0
Annual cash income 877,138 982,306 1,095,888 1,218,556 1,351,038
Discount factor 0.86 0.84 0.82 0.80 0.79
Annual Present value 758,571 829,207 902,964 980,026 1,060,591
Total capital investment+ interest
Net Present Value
195
11 12 13 14 15
Fixed Capital Investment
Land
Working Capital
Loan Payment
Loan Interest payment
Loan principal
Bioethanol sales 2,700,426 2,700,426 2,700,426 2,700,426 2,700,426
By-product credit 577,167 577,167 577,167 577,167 577,167
Total annual sales 3,277,593 3,277,593 3,277,593 3,277,593 3,277,593
Annual manufacturing cost
Feedstock cost 176023.113 176023.113 176023.113 176023.113 176023.113
Other variable cost 4,932 4,932 4,932 4,932 4,932
Fixed operating costs 1,602,520 1,602,520 1,602,520 1,602,520 1,602,520
Total Product cost 1,783,476 1,783,476 1,783,476 1,783,476 1,783,476
Annual depreciation
General plant writedown
Depreciation charge
Remaining value
Net revenue 1,494,118 1,494,118 1,494,118 1,494,118 1,494,118
Losses forward -13,015,368 -11,521,250 -10,027,133 -8,533,015 -7,038,897
Taxable income -11,521,250 -10,027,133 -8,533,015 -7,038,897 -5,544,779
Income tax 0 0 0 0 0
Annual cash income 1,494,118 1,494,118 1,494,118 1,494,118 1,494,118
Discount factor 0.77 0.75 0.73 0.71 0.70
Annual Present value 1,144,862 1,117,484 1,090,760 1,064,676 1,039,215
Total capital investment+ interest
Net Present Value
196
16 17 18 19 20
Fixed Capital Investment
Land
Working Capital
Loan Payment
Loan Interest payment
Loan principal
Bioethanol sales 2,700,426 2,700,426 2,700,426 2,700,426 2,700,426
By-product credit 577,167 577,167 577,167 577,167 577,167
Total annual sales 3,277,593 3,277,593 3,277,593 3,277,593 3,277,593
Annual manufacturing cost
Feedstock cost 176023.113 176023.113 176023.113 176023.113 176023.113
Other variable cost 4,932 4,932 4,932 4,932 4,932
Fixed operating costs 1,602,520 1,602,520 1,602,520 1,602,520 1,602,520
Total Product cost 1,783,476 1,783,476 1,783,476 1,783,476 1,783,476
Annual depreciation
General plant writedown
Depreciation charge
Remaining value
Net revenue 1,494,118 1,494,118 1,494,118 1,494,118 1,494,118
Losses forward -5,544,779 -4,050,661 -2,556,543 -1,062,426 0
Taxable income -4,050,661 -2,556,543 -1,062,426 431,692 1,494,118
Income tax 0 0 0 151,092 522,941
Annual cash income 1,494,118 1,494,118 1,494,118 1,343,026 971,177
Discount factor 0.68 0.66 0.65 0.63 0.62
Annual Present value 1,014,363 990,106 966,428 847,924 598,493 Total capital investment+
interest
Net Present Value
197
21 22 23 24 25
Fixed Capital Investment
Land
Working Capital
Loan Payment
Loan Interest payment
Loan principal
Bioethanol sales 2,700,426 2,700,426 2,700,426 2,700,426 2,700,426
By-product credit 577,167 577,167 577,167 577,167 577,167
Total annual sales 3,277,593 3,277,593 3,277,593 3,277,593 3,277,593
Annual manufacturing cost
Feedstock cost 176023.113 176023.113 176023.113 176023.113 176023.113
Other variable cost 4,932 4,932 4,932 4,932 4,932
Fixed operating costs 1,602,520 1,602,520 1,602,520 1,602,520 1,602,520
Total Product cost 1,783,476 1,783,476 1,783,476 1,783,476 1,783,476
Annual depreciation
General plant writedown
Depreciation charge
Remaining value
Net revenue 1,494,118 1,494,118 1,494,118 1,494,118 1,494,118
Losses forward 0 0 0 0 0
Taxable income 1,494,118 1,494,118 1,494,118 1,494,118 1,494,118
Income tax 522,941 522,941 522,941 522,941 522,941
Annual cash income 971,177 971,177 971,177 971,177 971,177
Discount factor 0.60 0.59 0.57 0.56 0.55
Annual Present value 584,180 570,210 556,574 543,264 530,273
Total capital investment+ interest
Net Present Value
198
26 27 28 29 30
Fixed Capital Investment
Land -72,000
Working Capital -1,080,090
Loan Payment
Loan Interest payment
Loan principal
Bioethanol sales 2,700,426 2,700,426 2,700,426 2,700,426 2,700,426
By-product credit 577,167 577,167 577,167 577,167 577,167
Total annual sales 3,277,593 3,277,593 3,277,593 3,277,593 3,277,593
Annual manufacturing cost
Feedstock cost 176023.113 176023.113 176023.113 176023.113 176023.113
Other variable cost 4,932 4,932 4,932 4,932 4,932
Fixed operating costs 1,602,520 1,602,520 1,602,520 1,602,520 1,602,520
Total Product cost 1,783,476 1,783,476 1,783,476 1,783,476 1,783,476
Annual depreciation
General plant writedown
Depreciation charge
Remaining value
Net revenue 1,494,118 1,494,118 1,494,118 1,494,118 1,494,118
Losses forward 0 0 0 0 0
Taxable income 1,494,118 1,494,118 1,494,118 1,494,118 1,494,118
Income tax 522,941 522,941 522,941 522,941 522,941
Annual cash income 971,177 971,177 971,177 971,177 971,177
Discount factor 0.53 0.52 0.51 0.50 0.48
Annual Present value 517,592 505,214 493,132 481,339 469,828
Total capital investment+ interest -557349.2929
Net Present Value
199
Table D-3: Biorefinery Equipment cost breakdown
EQ
UIP
ME
NT
TIT
LE
#
$
Year o
f Quo
te
Scalin
g V
al
Un
its
Scalin
g E
xp
Inst F
actor
New
Val
Size R
atio
Scaled
Pu
rch
Co
st
Scaled
Inst C
ost
Truck Scale 2 110,000 2009 94697 kg/hr 0.60 1.7 8561 0.09 26,007 44,213
Truck Dumper 2 484,000 2009 94697 kg/hr 0.60 1.7 8561 0.09 114,433 194,535
Truck Dumper Hopper
2 502,000 2009 94697 kg/hr 0.60 1.7 8561 0.09 118,688 201,770
Concrete
Feedstock
Storage Dome
2 3,500,000 2009 94697 kg/hr 0.60 1.7 8561 0.09 827,508 1,406,764
Belt Scale 2 10,790 2009 94697 kg/hr 0.60 1.7 8561 0.09 2,551 4,337
Dust Collection System
1 279,900 2009 94697 kg/hr 0.60 1.7 8561 0.09 66,177 112,501
Feedstock handling 1,155,365 1,964,120
Oligomer Hold
Tank Agitator 1 30,000 2009 264116 kg/hr 0.50 1.5 8561 0.03 5,401 8,102
Pretreatment Water Heater
1 92,000 2010 -8 Gcal/hr 0.70 2.2 0.03 7,903 17,386
Waste Vapor
Condenser 1 34,000 2009 2 Gcal/hr 0.70 2.2 0.03 2,921 6,425
Flash Tank Discharge Pump
1 30,000 2009 204390 kg/hr 0.80 2.3 6625 0.03 1,931 4,440
Oligomer Hold
Tank Discharge 1 17,408 2010 292407 kg/hr 0.80 2.3 9478 0.03 1,120 2,577
Hydrolyzate Pump
1 22,500 2009 402194 kg/hr 0.80 2.3 8561 0.02 1,034 2,379
Oligomer
Conversion Tank 1 203,000 2009 264116 kg/hr 0.70 2.0 8561 0.03 18,408 36,815
Liquefaction totals 38,717 78,124
Ethanol
Fermentor Agitator
1 52,500 2009 1 ea 1.00 1.5 0 0 1,050 1,575
Seed Hold Tank
Agitator 1 31,800 2009 40414 kg/hr 0.50 1.5 713 0.02 4,225 6,337
Beer Surge Tank Agitator
2 68,300 2009 425878 kg/hr 0.50 1.5 8561 0.02 9,684 14,525
Enzyme-
Hydrolysate Mixer
1 109,000 2009 379938 kg/hr 0.50 1.7 7637 0.02 15,454 26,272
Ethanol
Fermentor 12 2009 12 ea 1.00 1.5 0 0.02 202,560 303,840
1st Seed Fermentor
2 75,400 2009 2 ea 0.70 1.8 2 1.00 75,400 135,720
Fermentation
Cooler 12 86,928 2009 12 ea 1.00 2.2 0 0.02 1,739 3,825
Hydrolyzate Cooler
1 85,000 2010 8 Gcal/hr 0.70 2.2 0 0.00 0 0
Fermentor Batch
Cooler 1 23,900 2009 5 Gcal/hr 0.70 1.8
0.00
0 0.00 0 0
Fermentation Recirc/Transfer
Pump
5 47,200 2009 12 ea 0.80 2.3 0 0.02 2,064 4,748
Seed Hold
Transfer Pump 1 8,200 2009 43149 kg/hr 0.80 2.3 762 0.02 325 746
Beer Transfer
Pump 1 26,800 2009 488719 kg/hr 0.80 2.3 9824 0.02 1,177 2,707
200
Saccharification
Transfer Pump 5 47,200 2009 421776 kg/hr 0.80 2.3 8478 0.02 2,073 4,767
Seed Hold Tank 1 439,000 2009 40414 kg/hr 0.70 1.8 713 0.02 26,015 46,827
Beer Storage Tank
1 636,000 2009 425878 kg/hr 0.70 1.8 7518 0.02 37,689 67,840
Saccharification
Tank 8 3,840,000 2009 421776 kg/hr 0.70 2.0 7518 0.01 95,320 190,640
Saccharification and Femrentation 474,773 810,369
Beer Column 1 3,407,000 2009 30379 kg/hr 0.60 2.4 0.02 325,829 781,990
Rectification
Column Condenser
1 487,000 2010 23 Gcal/hr 0.60 2.8 0.02 46,574 130,408
Molecular Sieve
Package (9
pieces)
1 2,601,000 2009 22687 kg/hr 0.60 1.8 0.02 248,747 447,745
Pressure Filter
Pressing Compr 1 75,200 2009 808 kg/hr 0.60 1.6 0.02 7,192 11,507
Pressure Filter
Drying Compr 2 405,000 2009 12233 kg/hr 0.60 1.6 0.02 38,732 61,972
Scrubber
Bottoms Pump 1 6,300 2009 24527 kg/hr 0.80 2.3 0.02 276 634
Filtrate Tank
Discharge Pump 1 13,040 2010 31815 kg/hr 0.80 2.3 0.01 188 433
Feed Pump 1 18,173 2010 31815 kg/hr 0.80 2.3 0.01 262 603
Manifold Flush Pump
1 17,057 2010 31815 kg/hr 0.80 2.3 0.01 246 566
Cloth Wash
Pump 1 29,154 2010 31815 kg/hr 0.80 2.3 0.01 421 967
Filtrate Discharge Pump
1 13,040 2010 31815 kg/hr 0.80 2.3 0.01 188 433
Pressure Filter 2 3,294,700 2010 31815 kg/hr 0.80 1.7 0.01 47,533 80,805
Vent Scrubber 1 215,000 2009 22608 kg/hr 0.60 2.4 113 0.01 8,950 21,480
Filtrate Tank 1 103,000 2010 31815 kg/hr 0.70 2.0 219 0.01 2,524 5,048
Feed Tank 1 174,800 2010 31815 kg/hr 0.70 2.0 219 0.01 4,284 8,567
Recycled Water Tank
1 1,520 2010 31815 kg/hr 0.70 3.0 219 0.01 37 112
Pressing Air
Compressor Receiver
1 8,000 2010 31815 kg/hr 0.70 3.1 219 0.01 196 608
Drying Air
Compressor
Receiver
2 17,000 2010 31815 kg/hr 0.70 3.1 219 0.01 417 1,291
Distillation and Rectification 732,596 1,555,170
Anaerobic Digestor Feed
Cooler
1 83,863 2010 393100 kg/hr 0.60 1.0 8367 0.02 8,326 8,326
Biogas Emergency Flare
4 32,955 2010 393100 kg/hr 0.60 1.0 8367 0.02 3,272 3,272
Polymer
Addition System 1 9,300 2010 393100 kg/hr 0.60 1.0 8367 0.02 923 923
Caustic Feed System
3 22,800 2010 393100 kg/hr 0.60 1.0 8367 0.02 2,263 2,263
Evaporator
System 3,801,095 2010 393100 kg/hr 0.60 1.0 8367 0.02 377,358 377,358
Anaerobic Reactor Feed
Pump
4 231,488 2010 393100 kg/hr 0.60 1.0 8367 0.02 22,981 22,981
Centrifuge 3 6,493,500 2010 393100 kg/hr 0.60 1.0 8367 0.02 644,649 644,649
Anaerobic Digestion 1,059,772 1,059,772
201
Denaturant In-
line Mixer 1 3,850 2009 23154 kg/hr 0.50 1.0 159 0.01 319 319
Ethanol Product Pump
2 9,200 2009 22681 kg/hr 0.80 3.1 156 0.01 171 531
Firewater Pump 1 15,000 2009 8343 kg/hr 0.80 3.1 0.02 656 2,034
Gasoline Pump 1 3,000 2009 473 kg/hr 0.80 3.1 0.02 131 407
CSL Pump 1 3,000 2009 1393 kg/hr 0.80 3.1 0 0.02 131 407
Ethanol Product
Storage Tank 2 1,340,000 2009 22681 kg/hr 0.70 1.7 156 0.01 41,047 69,780
Gasoline Storage
Tank 1 200,000 2009 473 kg/hr 0.70 1.7 0.02 12,935 21,989
Storage 55,391 95,466
Burner Combustion Air
Preheater
1 Incl.
BFW Preheater 1 Incl
Boiler 1 2010 238686 kg/hr 0.60 1.8 0.02 2,730,386 4,914,695
Combustion Gas
Baghouse Incl.
Turbine/Generator
1 9,500,000 2010 -42200 kW 0.60 1.8 0.01 492,179 885,922
Hot Process
Water Softener System
1 78,000 2010 235803 kg/hr 0.60 1.8 0.01 4,041 7,274
Amine Addition
Pkg. 1 40,000 2010 235803 kg/hr 0.00 1.8 0.01 40,000 72,000
Ammonia Addition Pkg
1 Incl.
Phosphate
Addition Pkg. 1 Incl.
Condensate
Pump 2 Incl.
Turbine
Condensate
Pump
2 Incl.
Deaerator Feed
Pump 2 Incl.
BFW Pump 5 Incl.
Blowdown Pump 2 Incl.
Amine Transfer
Pump 1 Incl.
Condensate Collection Tank
1 Incl.
Condensate
Surge Drum 1 Incl.
Deaerator 1 305,000 2010 235803 kg/hr 0.60 3.0 0.01 15,802 47,405
Blowdown Flash
Drum 1 Incl
Amine Drum 1 Incl.
Boiler and Turbogenerator 3,282,408 5,927,296
Grand Totals 6,799,022 11,490,317
202
Table D-3: Labor breakdown in biorefinery processes
2010 Salary # Required 2010 Cost
Plant Manager 155,949 1 155,949
Plant Engineer 74,261 1 74,261
Maintenance Tech 42,435 1 42,435
Lab Technician 42,435 1 42,435
Shift Operators 42,435 4 169,740
Yard Employees 29,704 2 59,409
Clerks & Secretaries 38,191 1 38,191
Total Salaries 11 582,420
Labor Burden (90%) 524,178
TOTAL Labor: 1,106,598
Table D-4: Chemical demand breakdown in biorefinery processes
Raw Material NREL kg/hr scale factor kg/yr 2010 Cost ($ / ton)
Corn Steep Liquor 1899.71 0.004 63830.29 57.91281 3696.592
Sorbitol 0.290833 0.004 9.772005 1148.096 11.2192
Purchased Enzyme* 8.560714 0 65.81726
Caustic (as pure) 73.80419 0.01 6199.552 152.4021 944.8249
Makeup Water 4822.09 0.02 810111.1 0.263955 213.8331
Subtotal 4932.286
*Enzyme requirement was assumed as 1% of feedstockdry weight.
203
Table D-4: Life cycle Inventory and impact assessment inputs per functional unit
Activity Pond Construction
Exchanges
name amount unit database location
Pond Construction 1 unit Duckweed US
market for diesel, burned in building machine
0.19503
55 megajoule
ecoinvent
3.3 GLO
market for polyvinylchloride, bulk polymerised 0.1 kilogram
ecoinvent
3.3 GLO
market for concrete, normal 0 cubic meter
ecoinvent
3.3 RoW
market for waste polyvinylchloride -0.1 kilogram
ecoinvent
3.3 RoW
polyvinylchloride production, bulk polymerisation 0.1 kilogram
ecoinvent
3.3 RoW
gravel production, crushed
482.367
38 kilogram
ecoinvent
3.3 RoW
market for waste reinforced concrete 0 kilogram
ecoinvent
3.3 RoW
market for excavation, hydraulic digger
0.34543
52 cubic meter
ecoinvent
3.3 GLO
market for waste plastic, mixture 0 kilogram
ecoinvent
3.3 RoW
Transformation, to industrial area 1 square meter biosphere3
(Unknow
n)
Occupation, industrial area 30
square meter-
year biosphere3
(Unknow
n)
Transformation, from unspecified 1 square meter biosphere3
(Unknow
n)
market for polyethylene, low density, granulate 0.3 kilogram
ecoinvent
3.3 GLO
market for extrusion, plastic film 0.3 kilogram
ecoinvent
3.3 GLO
market for waste plastic, mixture -0.3 kilogram
ecoinvent
3.4 CH
Activity Water Quality Change
Exchanges
name amount unit database location
Water Quality Change -1 unit Duckweed US
BOD5, Biological Oxygen Demand
2.43E+0
1 kilogram biosphere3
(Unknow
n)
BOD5, Biological Oxygen Demand 0 kilogram biosphere3
(Unknow
n)
Phosphate
1.31852
82 kilogram biosphere3
(Unknow
n)
Ammonia 0 kilogram biosphere3
(Unknow
n)
Ammonium, ion
9.50E-
01 kilogram biosphere3
(Unknow
n)
Nitrate
3.79E-
01 kilogram biosphere3
(Unknow
n)
Nitrate 0 kilogram biosphere3
(Unknow
n)
204
Nitrogen
1.71E+0
0 kilogram biosphere3
(Unknow
n)
Phosphate 0 kilogram biosphere3
(Unknow
n)
Phosphate 0 kilogram biosphere3
(Unknow
n)
Phosphorus 0 kilogram biosphere3
(Unknow
n)
Water
-
4.66783
5 cubic meter biosphere3
(Unknow
n)
Water
175.557
45 cubic meter biosphere3
(Unknow
n)
Activity Duckweed Cultivation
Exchanges
name amount unit database location
Ammonium, ion
4.59574
5 kilogram biosphere3
(Unknow
n)
Phosphate
2.01131
4 kilogram biosphere3
(Unknow
n)
Carbon dioxide, in air -8.1 kilogram biosphere3
(Unknow
n)
Water
1.09937
4 cubic meter biosphere3
(Unknow
n)
Duckweed Cultivation -1 unit Duckweed US
Activity Duckweed Transportation
Exchanges
name amount unit database location
Duckweed Transportation 1 unit Duckweed US
market for transport, freight, lorry 7.5-16 metric ton,
EURO3 0 ton kilometer
ecoinvent
3.3 GLO
market for transport, freight, lorry 3.5-7.5 metric ton,
EURO5 0 ton kilometer
ecoinvent
3.3 GLO
Activity Duckweed Drying
Exchanges
name amount unit database location
Duckweed Drying 1 unit Duckweed US
market for electricity, high voltage 0 kilowatt hour
ecoinvent
3.3 RoW
market for heat, district or industrial, natural gas 0 megajoule
ecoinvent
3.3 CA-QC
water 0 cubic meter biosphere3
(Unknow
n)
Activity Biorefinery Construction
Exchanges
name amount unit database location
205
market for ethanol fermentation plant
3.55E-
08 unit
ecoinvent
3.3 GLO
Biorefinery Construction 1 unit Duckweed US
Activity Duckweed Liquefaction
Exchanges
name amount unit database location
Duckweed Liquefaction 1 unit Duckweed US
market for heat, district or industrial, natural gas
88.2291
9 megajoule
ecoinvent
3.3 CA-QC
Activity Saccharification
Exchanges
name amount unit database location
Saccharification 1 unit Duckweed US
market for electricity, high voltage
8.16780
8 kilowatt hour
ecoinvent
3.3 RoW
Activity Duckweed Fermentation
Exchanges
name amount unit database location
market for fodder yeast 1.53 kilogram
ecoinvent
3.3 GLO
Carbon dioxide, non-fossil
27.1659
6 kilogram biosphere3
(Unknow
n)
Duckweed Fermentation 1 unit Duckweed US
Activity Distillation
Exchanges
name amount unit database location
Distillation 1 unit Duckweed US
market for heat, district or industrial, natural gas
803.959
6 megajoule
ecoinvent
3.3 CA-QC
market for electricity, high voltage
5.72511
3 kilowatt hour
ecoinvent
3.3 RoW
NMVOC, non-methane volatile organic compounds,
unspecified origin
0.00121
6 kilogram biosphere3
(Unknow
n)
water -2.7E-05 cubic meter biosphere3
(Unknow
n)
Activity Duckweed anaerobic digestion
Exchanges
name amount unit database location
206
Duckweed anaerobic digestion 1 unit Duckweed US
Hydrogen sulfide
0.00913
9 kilogram biosphere3
(Unknow
n)
Carbon dioxide, non-fossil 14.9 kilogram biosphere3
(Unknow
n)
market for heat, district or industrial, natural gas
443.388
2 megajoule
ecoinvent
3.3 CA-QC
Activity Solid Recovery as Soil Amendment
Exchanges
name amount unit database location
Solid Recovery as Soil Amendment 1 unit Duckweed US
Water 0 cubic meter biosphere3
(Unknow
n)
Activity Gasoline Substitution
Exchanges
name amount unit database location
Gasoline Substitution 1 unit
Substitutio
n US
market for petrol, low-sulfur
28.5688
2 kilogram
ecoinvent
3.3 RoW
Activity Natural Gas Substitution
Exchanges
name amount unit database location
Natural Gas Substitution 1 unit
Substitutio
n US
market for natural gas, high pressure
32.5276
6 cubic meter
ecoinvent
3.3 US
Activity Nitrogen Fertilizer Substitution
Exchanges
name amount unit database location
Nitrogen Fertilizer Substitution 1 unit
Substitutio
n US
market for nitrogen fertiliser, as N
3.78473
09 kilogram
ecoinvent
3.3 GLO
Activity WWTP C_O Substitution
Exchanges
name amount unit database location
WWTP C_O Substitution 1 unit
Substitutio
n US
207
wastewater treatment facility construction, capacity
1.1E10l/year
1.135E-
07 unit
ecoinvent
3.3 RoW
market for electricity, low voltage
0.29106
4 kilowatt hour
ecoinvent
3.3 CH
market for heat, district or industrial, natural gas
1.34087
61 megajoule
ecoinvent
3.3 CH
208
VITA
Ayse Ozgul Calicioglu
EDUCATION
2014-2019 Ph.D. in Environmental Engineering, The Pennsylvania State University.
2011 - 2013 M.Sc. in Environmental Engineering, Middle East Technical University.
2007 – 2013 B.S. in Business Management, People’s Friendship University of Russia.
2005 - 2011 B.S. in Environmental Engineering, Middle East Technical University.
SCHOLARSHIPS AND AWARDS
Green Talents 2018 Competition Winner as a “promising young researcher in sustainable development
field”. Federal Ministry of Education and Research, Germany, October 2018.
National Federation of the Blind, Oracle Excellence in STEM Field Award. Orlando, July 2018.
Marley Fellowship, for academic success. Penn State, January 2017.
Virginia G. Rimer Memorial Scholarship, for academic success. Penn State, August 2016.
Delta Gamma Foundation Golden Anchor Scholarship for perseverance, and strength of character. Penn
State, May 2016.
Logan Ray Monkovski Award for 1st place in Paper Presentation, College of Engineering Research
Symposium. Penn State, April 2015.
Fulbright International Student Grant for Ph.D. studies. Fulbright Turkey, Fall 2013.
Course Performance Award for the highest departmental GPA in Master’s Program. METU, July 2013.
2nd graduate of Environmental Engineering Department Undergraduate Program. METU, July 2011.
High Honor Roll. METU, Spring 2011, Spring 2010, Fall 2009.
Honor Roll. METU, Fall 2010, Fall 2008.
SELECTED PUBLICATIONS AND CONFERENCE PROCEEDINGS
Calicioglu, O., Demirer, G.N., 2019. Carbon-to-Nitrogen and Substrate-to-Inoculum ratio
adjustments can improve Co-digestion performance of microalgal biomass obtained from
domestic wastewater treatment. Environmental Technology 40(5).
Calicioglu, O., Bracco S., Flammini A., Belu L., 2019. Future Challenges of Food and
Agriculture: An integrated Analysis. Sustainability, 11(1).
Calicioglu, O., Shreve, M. J., Richard, T. L., Brennan, R. A., 2018. Effect of pH and
Temperature on Microbial Community Structure and Carboxylic Acid Yield During the
Acidogenic Digestion of Duckweed. Biotechnology for Biofuels. 11(1).
Bracco, S., Calicioglu, O., Gomez San Juan, M., Flammini, A. 2018. Assessing the Contribution
of Bioeconomy to the Total Economy: A Review of National Frameworks. Sustainability, 10(6).
Calicioglu, O., Brennan, R. A. 2018. Sequential ethanol fermentation and anaerobic digestion
increases bioenergy yields from duckweed. Bioresource Technology, 257, 344–348.
Calicioglu, O., Demirer, G.N., 2016. Biogas Production from Waste Microalgal Biomass
Obtained from Nutrient Removal of Domestic Wastewater. Waste and Biomass Valorization.
Calicioglu, O., Demirer, G. N., 2015. Integrated Nutrient Removal and Biogas Production by
Chlorella vulgaris Cultures. Journal of Renewable and Sustainable Energy, 7 (3).
Calicioglu O., Hepgunes E., Firat F., Alp E. “Public perception and willingness to pay analysis
for the improved water quality in Ankara, Turkey.” CEST 2011, September 8-10, 2011, Greece.