THE WATERSHED-SCALE OPTIMIZED AND REARRANGED LANDSCAPE
DESIGN MODEL (WORLD) FOR CELLULOSIC FEEDSTOCK PRODUCTION AND
ADVANCED LOCAL BIOMASS PROCESSING DEPOTS (LBPD) FOR SUSTAINABLE
BIOFUEL PRODUCTION: INTEGRATED LIFE CYCLE ASSESSMENTS
By
Pragnya Lavanya Eranki

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILSOPHY
Chemical Engineering
2012

ABSTRACT
THE WATERSHED-SCALE OPTIMIZED AND REARRANGED LANDSCAPE
DESIGN MODEL (WORLD) FOR CELLULOSIC FEEDSTOCK PRODUCTION AND
ADVANCED LOCAL BIOMASS PROCESSING DEPOTS (LBPD) FOR SUSTAINABLE
BIOFUEL PRODUCTION: INTEGRATED LIFE CYCLE ASSESSMENTS
By
Pragnya Lavanya Eranki
Interest in commercially viable cellulosic biofuel production has greatly increased due to
concerns regarding the sustainability of petroleum fuels. These biofuels can help fulfill
escalating demands for liquid fuels and mitigate the environmental impacts of petroleum-derived
fuels. Two key factors in their successful large- scale production are pretreatment (in biological
conversion processes) and a consistent supply of feedstock. While research into solving the
technical issues is ongoing, much less attention has been paid to solving supply chain challenges
such as low bulk density of cellulosic biomass, compositional variability and seasonality of the
feedstock.
Currently large biorefineries face many logistical problems because they are centralized
facilities in which all units of the conversion process are present in a single location. These
logistical problems can be addressed using a system of distributed processing networks called
Regional/Local Biomass Processing Depots (RBPDs, LBPDs or depots). These depots are
strategically distributed facilities that procure, pre-process /pre-treat and densify biomass into
stable intermediate products that are compatible with existing bulk commodity logistical
systems.
On the agricultural production side, an array of feedstocks such as corn stover,
switchgrass, miscanthus, native prairie grasses etc. are being evaluated as potential raw materials

for cellulosic biofuel production. Additionally, management practices such as the use of marginal
lands, no-till and double-cropping, riparian buffers, when incorporated in the feedstock module
of the biofuels supply chain, may enhance overall system sustainability. However, thorough
assessments are required in real landscape settings on regional levels before these feedstocks can
be cultivated and sustainable practices can be implemented. Likewise biofuel production should
be maximized and negative environmental impacts should be minimized in growing these new
feedstocks.
This research has two primary objectives: to propose designs of sustainable optimized
cellulosic feedstock landscapes for biofuel production and to conduct integrated systems-wide
life cycle analyses of these optimized landscapes combined with distributed processing and
associated auxiliary processes (such as transport operations). It also aims to address pertinent
current issues in the bioenergy production sector such as: avoiding indirect land use change
impacts (iLUC) and the “feed vs. fuel” controversy, maximizing ecosystem services and
improving the quality of water bodies. The watershed-scale optimized & rearranged landscape
design (WORLD) model was created to estimate land allocations for different cellulosic
feedstocks at biorefinery scale while paying attention to the aforementioned issues.
In summary, this research answers several key questions in the biofuel production
process regarding the advantages of distributed processing systems, the technical potential of
landscapes and maximizing the benefits of these landscapes plus processing systems for
environmental, economic and social incentives. The WORLD model and integrated LCAs can be
used as decision making tools by growers, industries or policy makers.

DEDICATION

I dedicate this dissertation to my loving family- my parents, my sister and my husband.

iv

ACKNOWLEDGEMENTS

First, I wish to thank my advisor, Dr. Bruce Dale for his guidance, encouragement, support, and
patience.
I would also like to thank Dr. Seungdo Kim for his assistance on the subject of life cycle
assessments and Dr. Bryan Bals for his suggestions about modeling and distributed processing.
I wish to thank my parents, Annapurna and Ravi Prakash Eranki, without whom I would not
have made it this far.
I would like to thank my sister, Ambika Eranki, for her unconditional support and
encouragement.
I would like to thank my husband, Krishna Sai Pappu, who has been my pillar of strength during
tough times, and the best research companion and sounding board throughout these five years.
This work was funded by DOE Great Lakes Bioenergy Research Center
(http://www.greatlakesbioenergy.org) supported by the US Department of Energy, Office of
Science, Office of Biological and Environmental Research, through Cooperative Agreement
DEFC02-07ER64494.

v

“The important thing in science is not so much to obtain new facts as to discover new
ways of thinking about them.”- Sir William Henry Bragg

vi

TABLE OF CONTENTS

List of Tables .................................................................................................................................. x
List of Figures ............................................................................................................................... xii
List of symbols and abbreviations .............................................................................................. xvii
Chapter 1: Introduction and literature review ................................................................................. 1
1.1. The rationale behind alternate energy production, specifically biofuels. ................ 3
1.2. From grain ethanol to cellulosic ethanol .................................................................. 4
1.3. Integration of cellulosic biofuel production with livestock feeding operations ...... 7
1.4. The logistics of cellulosic biofuel production .......................................................... 8
1.4.1. Regional Biomass Processing Centers .............................................................. 9
1.5. Land efficient biofuel and animal feed production ................................................ 12
1.6. Research approach ................................................................................................. 15
1.7. REFERENCES ...................................................................................................... 19
Chapter 2: Comparative life cycle assessment of centralized and distributed biomass processing
....................................................................................................................................................... 26
2.1 Introduction ............................................................................................................. 26
2.2. Structure and methods............................................................................................ 28
2.2.1 Calculations...................................................................................................... 28
2.2.2 Landscape analysis and LCA description ........................................................ 30
2.3. Results .................................................................................................................... 41
2.3.1. Base case scenario........................................................................................... 41
2.3.2 Sensitivity Analysis ......................................................................................... 44
2.4. Conclusion ............................................................................................................. 48
2.5. Appendix A ............................................................................................................ 52
A1. Crop inputs used in agricultural budget spreadsheets ........................................ 52
A2. Densification ...................................................................................................... 56
2.6. REFERENCES ...................................................................................................... 59

vii

Chapter 3: Advanced Regional Biomass Processing Depots - A key to the logistical challenges of
the cellulosic biofuel industry ....................................................................................................... 65
3.1. Introduction ............................................................................................................ 65
3.1.1. Current biomass supply system: Problems and some prospective solutions .. 65
3.2. Regional Biomass Processing Depots .................................................................... 68
3.2.1. Decoupling pretreatment from biofuel production ......................................... 68
3.2.2. Potential pretreatments for RBPDs ................................................................. 71
3.3. Discussion .............................................................................................................. 73
3.3.1. Overcoming the challenges of an RBPD network .......................................... 73
3.3.2. Enhanced Depots: Multiple technologies ....................................................... 76
3.4. Designing and implementing RBPDs .................................................................... 81
3.5. Conclusion ............................................................................................................. 83
3.6. REFERENCES ...................................................................................................... 85
Chapter 4: Life cycle assessment of watershed-scale cellulosic feedstock landscapes integrated
with distributed processing ........................................................................................................... 89
4.1. Introduction ............................................................................................................ 89
4.2. Structure and methods............................................................................................ 92
4.2.1. Technologies in the processing system ........................................................... 92
4.2.2. Feedstocks ....................................................................................................... 93
4.2.3. Landscape system ........................................................................................... 94
4.2.4. Life Cycle Assessment .................................................................................... 95
4.3. Results and discussion ......................................................................................... 100
4.3.1. Base case ....................................................................................................... 100
4.3.2. Base case results ........................................................................................... 101
4.3.3. Sensitivity analyses ....................................................................................... 109
4.3.4. Sensitivity analyses results ........................................................................... 110
4.4 Conclusion ............................................................................................................ 111
4.5. Appendix B .......................................................................................................... 114
B.1. The “watershed weightage factor” method for division of RIMA-MI into
watersheds ........................................................................................................................... 114
B.2. The “geometric factor” method for assignment of watershed land area portions
to depots .............................................................................................................................. 114
B.3. Land uses in different watersheds ................................................................... 116
B.4. Transport calculations ..................................................................................... 121
B.5. Nutritional values and co-product credit calculations ..................................... 121
viii

B.6. Results of energy and emission burdens as absolute values ........................... 123
4.6. REFERENCES .................................................................................................... 126
Chapter 5: The watershed-scale optimized and rearranged landscape design (WORLD) model
and local biomass processing depots for sustainable biofuel production: An integrated life cycle
assessment ................................................................................................................................... 131
5.1. Introduction .......................................................................................................... 131
5.2. Structure and methods.......................................................................................... 132
5.2.1. The watershed-scale optimized and rearranged landscape design model ..... 132
5.2.2. Local Biomass Processing Depots and Biorefinery ...................................... 140
5.2.3. Life Cycle Assessment .................................................................................. 141
5.3. Results and discussion ......................................................................................... 145
5.3.1. Reconfigured landscapes, animal feed and environmental impacts ............. 145
5.3.2. Net energy yields and net GHG emission reductions ................................... 153
5.3.3. Sensitivity analyses ....................................................................................... 162
5.4. Conclusion ........................................................................................................... 163
5.5. Appendix C .......................................................................................................... 168
C.1. A summary of technologies in the local biomass processing depots [LBPD] 168
C.2. Transport calculations ..................................................................................... 175
C.3. Co-product credit calculations ........................................................................ 178
C.4. Results of the WORLD model consistency checks (performing multiple runs in
each scenario)...................................................................................................................... 179
5.6. Appendix D .......................................................................................................... 195
D.1. Instructions to run the WORLD model and perform the integrated LCA ...... 195
D.2. Macros ............................................................................................................. 199
5.7. REFERENCES .................................................................................................... 211
Chapter 6: Conclusions and recommendations ........................................................................... 216
6.1. Conclusions .......................................................................................................... 216
6.2. Recommendations ................................................................................................ 218
6.3. REFERENCES .................................................................................................... 223

ix

LIST OF TABLES

Table 2.1: Nutritional value of feedstocks .................................................................................... 37
Table 2.2: Unit processes in the LCA and their energy inputs and emission outputs ................. 40
Table 2.3: Data sources and results of sensitivity analyses .......................................................... 46
Table A1. Corn grain and stover –inputs (all inputs for stover are same as that for corn) ........... 52
Table A2. Winter rye inputs.......................................................................................................... 53
Table A3. Switchgrass inputs ....................................................................................................... 54
Table A4: Miscanthus inputs ........................................................................................................ 55
Table A5: Values of parameters used in densification ................................................................. 57
Table 3.1: List of desirable properties for both cellulosic feedstocks and supply chains ............. 66
Table 3.2: Comparison of logistics for a typical 5000 ton/day corn stover biorefinery vs. a 100
ton/day regional biomass processing depot (RBPD) in Iowa ....................................................... 67
Table 4.1: Scenario description- Reconfiguration of RIMA-MI land area and technologies used
in each scenario ............................................................................................................................. 98
Table 4.2: Transportation distances of feedstocks, raw-material and co-products ....................... 99
Table B1: Yields of currently available feedstocks .................................................................... 115
Table B2: Yields of currently unavailable feedstocks estimated from literature values ............ 115

x

Table 5.1: Summary of optimization categories (tests) .............................................................. 139
Table 5.2: Details of perennial grasses in reconfigured landscape and changes in the original
landscape ..................................................................................................................................... 149
Table 5.3: Ranges of NEY and NGER in different scenarios, ranges of NEY and NGER at the
farm gate ..................................................................................................................................... 158
Table C1: Acreages of different land categories in the current RIMA landscape ...................... 169
Table C2: Feedstocks, management practices and yields simulated in the EPIC model ............ 170
Table C3: Nutritional values of feedstocks assumed in the model to fulfill animal feed
requirements ................................................................................................................................ 172
Table C4: Environmental impacts of feedstocks from the EPIC model .................................... 173
Table C5: Energy inputs of different feedstocks from literature and agricultural budgets ........ 174
Table C6: Cellulosic ethanol and electricity yields of different feedstocks assumed in the model
..................................................................................................................................................... 175
Table C7: Transportation distances for all raw-materials and co-products ................................ 177
Table C8: Transportation energy requirements and emissions ................................................... 178
Table D.1: List of variables and their interpretation in the WORLD model .............................. 198

xi

LIST OF FIGURES

Figure 1.1: Parallels between past recessions and oil price spikes. For interpretation of the
references to color in this and all other figures, the reader is referred to the electronic version of
this dissertation. .............................................................................................................................. 2

Figure 1.2: Total energy consumption and CO2 e. emissions (historical and projected values) in
the United States. ............................................................................................................................ 4
Figure 1.3: A flow-diagram representation of the original RBPC configuration proposed by
Carolan et al. in 2007. ................................................................................................................... 11
Figure 1.4: Ethanol production and GHG reductions from the current US agricultural system
considered in this analysis (114 million ha), that same acreage configured for maximizing
ethanol production, and that same acreage configured for maximizing GHG reductions. ........... 14
Figure 1.5: A flowchart depicting the research approach for this project ................................... 16
Figure 2.1: Block diagram of a simple Regional Biomass Processing Depot (RBPD) used in this
study. All processes confined within the dashed line represent operations present in the RBPD. 28
Figure 2.2: System boundaries for comparative LCA - (a) Distributed processing system, (b)
Centralized processing system. Both processing systems are combined with the mixed feedstock
landscapes and animal operations for an integrated system-wide analysis .................................. 32
Figure 2.3: Fractional yields of feedstocks in different configurations. The land area remains the
same in each configuration but feedstock acreages vary. Yields of cropping systems vary
proportionally with their acreages in each configuration. Corn system acreages as % of total land
area: Configuration 1 = 65%, Configuration 2 =15%, Configuration 3=0. Combined perennial
grass acreage Configuration 1 =35%, Configuration 2 = 85%, Configuration 3= 100%, divided
equally between switchgrass and miscanthus. HG represents high yields and LG represents low
yields of the two perennial grasses ............................................................................................... 35
Figure 2.4: Net Energy Yield (NEY) per hectare of the collective system (landscape, transport,
processing) in different configurations. The NEY (MJ) of the collective system comprises
differences in energy inputs and outputs of each processing system (distributed or centralized)
xii

combined with crops and transport. HG represents high yields and LG represents low yields of
the two perennial grasses .............................................................................................................. 42
Figure 2.5: Net Carbon Emission Reductions (NCER) per hectare of the collective system
(landscape, transport, processing) in different configurations. The NCER (kg CO2 eq.) of the
collective system comprises differences in GHG emissions generated by processing system
(distributed or centralized) combined with transport, and animal operations and carbon
sequestered by feedstocks. HG represents high yields and LG represents low yields of the two
perennial grasses ........................................................................................................................... 44
Figure 3.1: Comparison of potential RBPD systems (top) vs. no RBPDs (bottom). Advantages
of RBPDs include (A) smaller collection radius for individual RBPDs combined with a larger
total collection radius for the refinery, (B) individual RBPDs collecting enough biomass in
marginal cropland areas in locations where a single biorefinery may not be profitable, and (C)
allowing flexibility in feedstock allocation to multiple final locations, such as woody materials
(brown) to a co-fired power plant while grass (green) is sent to a refinery. ................................. 71
Figure 3.2: Process Flow Diagram of an RBPD with all technologies. Green squares represent
the landscape, blue squares represent operations within the RBPD, gray squares represent
products/intermediates, and red squares represent outside operations. Red arrows are energy
flows, green arrows are feedstock flows, purple arrows are material flows within the refinery,
and blue arrows are final products. ............................................................................................... 77

Figure 4.1: (a) Energy burdens of system modules, percent of total (b) CO2 eq. GHG emission
burdens of system modules, percent of total ............................................................................... 102
Figure 4.2: Land area distribution of different feedstocks and biomass multiplier. .................. 105
Figure 4.3: (a) Net energy gain [NEG]- LBPD profiles and averages (b) Net carbon emission
reductions [NCER] - LBPD profiles and averages. The base case for these values are set on the
vertical line at zero. ..................................................................................................................... 106

Figure 4.4: Soil N2O loss and NO3- leaching (based on functional unit [RIMA-MI land area]
and on unit ethanol production) - watershed profiles and averages............................................ 109
Figure 4.5: Results of the sensitivity analyses as a percentage increase or decrease from values
of original scenario 3 (on which the sensitivity analyses were performed) ................................ 111
Figure B.1: Land use in the St. Joseph watershed in RIMA-MI ............................................... 116
xiii

Figure B.2: Land use in the Kalamazoo watershed in RIMA-MI ............................................. 117
Figure B.3: Land use in the Thornapple watershed in RIMA-MI ............................................. 118
Figure B.4: Land use in the Upper Grand watershed in RIMA-MI........................................... 119
Figure B.5: Land use in the St. Joseph watershed in RIMA-MI ............................................... 120
Figure B.6: Absolute energy inputs of the different system modules ....................................... 123
Figure B.7: Absolute emission outputs of the different system modules .................................. 124
Figure 5.1: Modules, material flows and tools used in this analysis ......................................... 134
Figure 5.2: Animal feed nutrition provided in new configuration (vs. original) and cellulosic
ethanol production ...................................................................................................................... 146
Figure 5.3: (a) Percentage of total land area occupied by each feedstock in the “technical
potential of landscape” scenario (S1) (b) Percentage of total land area occupied by each
feedstock in the “opposition to conversion from conventional landscape” scenario (S2) .......... 147
Figure 5.4: (a) Percentage increases in beneficial environmental impact categories (b)
Percentage decreases in detrimental environmental impact categories under various scenarios in
reference to the baseline. Notation: T = test, S = scenario. The figure shows tests 1 (maximize
ethanol), 2a (minimize erosion), 2b (maximize Δ SOC), 3a (minimize N loss), 3b (minimize P
loss), 3c (maximize CWUE), 4 (minimize FGE) in combination with the two scenarios (S1technical potential of landscape, S2- opposition to conversion of conventional landscapes) .... 152
Figure 5.5: (a) Relative and absolute net energy yields (NEY) and biomass multiplier (BM).
NEY = Energy outputs of ethanol and electricity - energy inputs of feedstock, transport and
processing modules (b) Relative and absolute net greenhouse gas emission reductions (NGER)
and biomass multiplier (BM). NGER = CO2 e. emissions displaced due to ethanol and electricity
minus CO2 e. emissions generated due to feedstock, processing and transport modules. BM is
defined as the ratio of total cellulosic biomass generated in the reconfigured landscape to that
generated in the original landscape. ............................................................................................ 155

xiv

Figure 5.6: Correlations between the biomass multiplier and various system parameters. ∆SOC –
change in soil organic carbon. P loss – Phosphorus loss. N loss – Nitrogen loss. (NEY)abs –
absolute values of net energy yield. (NGER)abs – absolute values of net GHG emission
reductions .................................................................................................................................... 159
Figure 5.7: (a) Energy burdens of various modules- feedstock, processing and transport (b)
Emission burdens of various modules- feedstock, processing and transport.............................. 160
Figure 5.8: Results of sensitivity analyses performed in the WORLD model on Test 2b
(maximizing ∆SOC) in Scenario 1(technical potential of landscape). This test, scenario
combination was chosen to perform sensitivity analyses since this is the case where the greatest
NEY and NGER are obtained. .................................................................................................... 163
Figure C.1: Consistency check for T1S1 ................................................................................... 180
Figure C.2: Consistency check for T1S2 ................................................................................... 181
Figure C.3: Consistency check for T2aS1 ................................................................................. 182
Figure C.4: Consistency check for T2aS2 ................................................................................. 183
Figure C.5: Consistency check for T2bS1 ................................................................................. 184
Figure C.6: Consistency check for T2bS2 ................................................................................. 185
Figure C.7: Consistency check for T3aS1 ................................................................................. 186
Figure C.8: Consistency check for T3aS2 ................................................................................. 187
Figure C.9: Consistency check for T3bS1 ................................................................................. 188
Figure C.10: Consistency check for T3bS2 ............................................................................... 189
Figure C.11: Consistency check for T3cS1 ............................................................................... 190

xv

Figure C.12: Consistency check for T3cS2 ............................................................................... 191
Figure C.13: Consistency check for T4S1 ................................................................................. 192
Figure C.14: Consistency check for T4S2 ................................................................................. 193

xvi

LIST OF SYMBOLS AND ABBREVIATIONS

RBPD- Regional Biomass Processing Depot
AFEX™ - Ammonia Fiber Expansion (pretreatment process)

NEY- Net Energy Yield (MJ)

NCER- Net Carbon Emission Reduction (kg or Mg CO2 e.)

TPD- Tons per day
GHG- Greenhouse gas
SOM- Soil organic matter
SOC- Soil organic carbon
WSS- Web soil survey
LCA- Life cycle assessment
IBSAL- Integrated Biomass Supply Analysis and Logistics
GREET- Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation
LBPD- Local biomass processing depot
HUC- Hydrologic unit code
NEG- Net energy gain (GJ)

xvii

iLUC – Indirect land use change
BASINS- Better Assessment Science Integrating Point and Non-point Sources (model)
EPIC- Environmental policy integrated climate (model)
NLCD- National land cover database
LPC- Leaf protein-concentrate (extraction process)
RIMA- Regional intensive modeling area
WORLD- Watershed-scale optimized and rearranged landscape design (model)
AL, PL, ML- Arable, pasture and marginal lands
ESV- Early succession vegetation
TDN- Total digestible nutrients
TNPDN- Total non-protein digestible nutrients
ΔSOC- Change in soil organic carbon
CWUE - Crop water use efficiency
FGE- Farm GHG emissions
N, P loss- Nitrogen, Phosphorus losses
BM- Biomass multiplier (ratio)

(NEY)rel, (NEY)abs- Relative and absolute net energy yields

xviii

(NGER)rel, (NGER)abs- Relative and absolute net GHG emission reductions

xix

CHAPTER 1: INTRODUCTION AND LITERATURE REVIEW

The reliance on fossil fuels all over the world has triggered socio-economic crises in
recent time such as:
A.

Economic disorder due to high and variable costs: Average national

gasoline prices increased by $0.46 per gallon between the end of December 2011
and the end of February 2012 [1]- in just a two month period. Gasoline consumers
in the United States are particularly feeling the effects of oil shocks [2] because of
high consumption and endure the consequences of these increasing prices. Rising
gasoline prices have also been pointed out as one of the reasons for the current
global economic recession [3]. Figure 1.1 shows parallels between economic
recessions in the past and spikes in oil prices [4].
B.

Declining fossil fuel supplies: There are large differences in the assumed

size of oil resource in different existing projections. However, these dissimilarities
make relatively little difference to the timing of a global peak in oil production. A
significant risk has been predicted due to the peaking of conventional oil
production before 2020. Forecasts that predict the delay of this peak to beyond
2030 may be optimistic but inaccurate [5].

1

Figure 1.1: Parallels between past recessions and oil price spikes. For interpretation of the
references to color in this and all other figures, the reader is referred to the electronic version of
this dissertation.
C.

Unstable imports: Petroleum imports in the United States increased from

about 20% in 1960 to nearly 60% in 2005. However, with a decline in
consumption in 2008, and enlarged domestic production, reliance on imports
decreased to 45% in 2011 [6]. In 2011, the crises in the Middle East and North
Africa contributed to higher crude oil prices, consequently increasing gasoline
prices. In 2012 again, tensions with Iran have contributed to rising crude prices,
which in turn are increasing gasoline costs [1]. Such instabilities place a great deal
of stress on both prices and supplies of transportation fuels.
D.

Climate and health concerns: The combined climate-change and health

costs of producing and combusting gasoline amount to $469 million (quantified

2

based on billion ethanol-equivalent gallons of fuel produced) in the USA [7].
Particulate matter emission, a form of air pollution, from gasoline combustion has
been related to cardiovascular disease and premature mortality [8]. Greenhouse
gas (GHG) emissions from petroleum combustion have been linked with global
climate change [9]. Furthermore, the Exxon Valdez oil spill in 1989 and the BP
oil spill in 2010 were the biggest and most environmentally destructive of oil
discharges in the history of North America [10], that lead to questions regarding
both direct and indirect environmental impact concerns of fossil fuels.
1.1. The rationale behind alternate energy production, specifically biofuels.
In spite of these concerns, the world energy consumption of liquid fuels continues to
grow steadily with projected increases to 97.6 million barrels per day in 2020 and 112.2 million
barrels per day in 2035 from 83.6 million barrels per day in 2005 and 85.7 million barrels in
2008. The United States is the chief consumer of transportation energy among the organization
for economic co-operation and development (OECD) nations [11]. As seen in Figure 1.2, steady
increases are projected in energy consumption as well as CO2 equivalent emissions in the energy
sector [15]. In the liquid fuel consumption sector, light duty vehicles (dominated by gasoline
powered vehicles in the US) are projected to consume about 56% of energy among all modes of
transportation [12].
In order to mitigate these problems and diversify liquid transportation fuels it is necessary
to seek alternatives. Biofuels (generated from natural organic matter such as plants) are a favored
alternate source of liquid transportation fuels since they are easy to transport, compatible with
existing infrastructure and possess relatively high energy densities. The most commonly used

3

biofuel in the United States currently is ethanol mainly due to its property of blending with
gasoline in “Flex Fuel” vehicles (that are able to combust E-85: 85% ethanol and 15% gasoline)
[13]. In 2011, using 5 billion gross bushels of corn, 13.9 billion gallons of ethanol and greater
than 39 million metric tons of livestock feed (distiller’s grain, corn gluten feed and meal) were

120

6400
6200
6000
5800
5600
5400
5200
5000
4800

Energy use

100
80
60
40
20
0
2009

2010

CO2 emissions

generated in the US [14].

2035

Total Energy Use (Quadrillion btu)
Carbon Dioxide Emissions (million metric tons CO2 e.)
Figure 1.2: Total energy consumption and CO2 e. emissions (historical and projected values) in
the United States.
1.2. From grain ethanol to cellulosic ethanol
The production of an alternate fuel from a food crop has been a source of great
disagreement in recent times with the two chief, albeit questionable, arguments being the “food
vs. fuel” controversy [16] and that biofuel generation from food crops cause long-term GHG
emission increases due to indirect land use change impacts [17]. Due to these negative
connotations to grain ethanol as well as due to limits on amounts of crop lands in the United
States, it became necessary to examine other sources of biofuel production from non-food
components of agricultural crops (such as residues) as well as non-food crops (such as grasses).
4

The Energy Independence and Security Act [18] mandated a Renewable Fuel Standard (RFS-2)
to produce 36 billion gallons of biofuels in 2022, of which no more than 15 billion gallons can be
generated from grain ethanol and requires the production of 16 billion gallons of biofuel from
cellulosic feedstock [19]. In addition it imposes a 60% life cycle GHG reduction threshold
(compared to gasoline) on cellulosic biofuels as well as emphasizes the use of feedstocks
expected to have minimal land use change impacts such as agricultural and forest residues,
secondary annual crops such as winter cover crops and perennial grasses. Moreover, certain
pieces of legislation and regulations such as the Renewable Portfolio Standard (RPS), and the
regional Greenhouse Gas initiative (RGGI) have also been formulated that appear to have some
probability of being enacted in the near future [20, 21].
Based on the negative perception of grain biofuels, some characteristics of desirable
biofuel raw-materials are stability in price and availability, low cost, consistent composition,
generating favorable co-products, being either environmentally neutral or beneficial, being
storable and posing no threat to food production. Cellulosic feedstocks already possess most of
these characteristics and have been gaining impetus in the past decade. Corn stover is one of the
most abundant lignocellulosic resources available in the U.S and is widely considered to be a
primary feedstock for cellulosic biofuels [22, 23]. Sheehan et al. [24] showed in a study that corn
stover ethanol in E85 fuel reduced total fossil energy use and GHG emissions by 102% and
113% respectively on a life-cycle basis considered in that study. The perennial grasses
switchgrass and miscanthus have also been attracting interest as potentially important feedstocks
for biofuel production and are also known as dedicated energy crops [25-29]. These grasses
require lower inputs (such as fertilizers, etc.) compared to conventional annual crops and
monocultures and are also known to improve soil quality [30]. Moreover, Schmer et al. [31]
5

showed that using switchgrass in field trials of 3–9 ha on marginal cropland, with average yields
of 5.7 -12.21 tons / ha, 540% more renewable energy is generated than non-renewable energy
consumed (equivalent to a Fossil Energy Ratio of 5.4). It was also estimated that the GHG
emissions were 94% lower compared to gasoline. More recently, unmanaged (low-input high
diversity) native prairie grasses [32] are also under investigation as potential biofuel feedstocks.
Such grasslands can offer valuable ecosystem services such as improved carbon sequestration,
wildlife habitat conservation, pollination, recreational benefits, water quality maintenance and
nitrogen fixation. However, a chief concern regarding the use of these grasslands as a biofuel
feedstock is whether or not they can provide sufficient biomass and processing yields. Garlock et
al. [33] suggest that these grasses may have the potential to produce yields equal or greater than
certain grass monocultures, generate significant quantities of sugars at pretreatment conditions
similar to corn stover, be highly profitable to both the biorefinery and the farmer. Other crops
that are currently under consideration as potential feedstocks for cellulosic ethanol production
include sweet sorghum, energy sorghum [34, 35].
Additionally, the implementation of sustainable practices such as the use of marginal
lands (defined as any land not being used to grow commercial or conventional crops or
abandoned and degraded cropland that may be capable of growing low-maintenance biomass
feedstocks [36]) and no-till farming for erosion abatement can help obtain greater environmental
benefits associated with lignocellulosic biomass production [31, 37-39]. The use of cover crops
(such as rye, winter wheat and hairy vetch) either as double crops or as “green manure” to
maintain or increase soil carbon reserves and to reduce nitrate seepage and phosphorus runoff
after harvesting agricultural residues has also been attracting great interest [40-43].

6

1.3. Integration of cellulosic biofuel production with livestock feeding operations
Grain ethanol has been favored as a biofuel not only because of its compatibility with
gasoline but also because it provides an equally important co-product in the form of distiller’s
grain for livestock feeding operations. Similarly it is also possible for cellulosic feedstocks to
produce animal feed along with biofuels. Bals [44] stated that it is possible for early harvest
switchgrass to generate nearly as much protein (0.5-1.5 Mg protein/ ha) as soybean which is the
leading source of the protein component in animal nutrition. Producing protein concentrates from
high protein cellulosic biomass has also been reported in literature by Enochian et al. [45] since
more than two decades ago, however such processes have not been commercially successful as
stand-alone operations. Integrating animal feeding operations with biofuel production may act as
a secondary income source for biorefineries and help in reducing the economic risks involved in
this new industry [44, 46].
Sendich et al. [47] compared several crop and animal simulation models such as DayCent
[48], IFSM [49] and I-Farm to incorporate the most suitable one (IFSM) in an integrated analysis
of biorefinery and farming operations. Sendich and Dale [50] then reported a model called the
Biorefinery and Farm Integration Tool (BFIT) in which a basic economic and environmental
analysis was conducted for the integrated biofuel production system. The model reported that the
Midwest United States is particularly suitable for a cellulosic ethanol industry and that
integrating such a refinery with agricultural operations augments farm incomes and decreases
feedstock emissions. However, this model included only individual farm-level operations. In
more realistic scenarios, biomass production modeling must be developed on regional and local
scales i.e. an aggregation of local farms.

7

1.4. The logistics of cellulosic biofuel production
Just like petroleum refining, biorefining is the integrated, industrial scale production of
fuels and higher value products but from biomass feedstock [51]. The current model for large
biorefinery contains all operations at a single centralized location, handling enormous tonnages
of mixed biomass per day. This may be unfeasible from a logistical standpoint, since although
biofuel production is centralized, biomass production is decentralized and local. For instance,
Perlack et al. [52] estimated that for a facility handling 4000 dry tons/day of feedstock for 300
operating days and with 10% storage and handling losses, a collection area (for corn stover only)
2

of 14250 mi would be required. This translates to a one way haul radius of 62 mi, including a
30% road winding factor. Additionally, the impending and rapid growth in the cellulosic biofuel
industry presents many questions related to reaction of societies, environmental concerns, and
benefits to local and rural communities. Whether biofuels can simultaneously contend and assist
with current and potential production scenarios while meeting societal requirements has become
a global question [53].
There are some disadvantages associated with cellulosic feedstocks particularly from the
logistical standpoint. They do not possess the inherent property of flow-ability akin to grain
crops such as corn. Moreover they have low bulk densities resulting in inefficient transport.
Other problems with feedstocks include their compositional variability and tendency to
decompose over time. These concerns have an impact on feedstock costs since they are not just
dependent on processing but also on biomass quality specifically sugar content and recalcitrance
of biomass as well as on logistics issues such as storage, transport and handling efficiencies [54].
While there is adequate focus on improving conversion technology, supply chain challenges
remain largely unsolved.
8

Nonetheless, the importance of connections between farms, logistics and processing is a
familiar one and the problems associated with delivering feedstock to biorefineries have been
acknowledged. Whereas certain biomass such as forest material may be left in its place of origin
and acquired as needed, herbaceous cellulosic feedstocks including perennial grasses with
comparatively delayed harvest timings, need to be stably stored before they can be used by the
biorefinery. One solution to this challenge has been offered in the form of intermediate facilities
known as satellite storage locations, satellite storage facilities or satellite depots. These sites are
either confined storage locations for the transitory storage and loading of round bales typically
using hauling equipment from farms [55, 56] or sometimes include moderate processing, usually
densification of biomass, before delivering it to the biorefinery [57]. Such depots are been being
employed by the Tennessee Biomass Supply Co-operative (part of the University of Tennessee
Biofuels Initiative) [58]. The Idaho National Laboratory bioenergy program is also focusing on
developing an ‘advanced uniform format feedstock supply system’ to create a homogenized,
consistent and stable commodity from cellulosic biomass [59].
1.4.1. Regional Biomass Processing Centers
While the concept of distributed storage and intermediate facilities in the supply chain of
biomass may be a familiar one, none of these facilities employs processing steps more advanced
than basic size reduction and densification which usually take place at the farm-gate; nor do they
include any form of chemical pretreatment. Carolan et al. [46] proposed the concept of
distributed biomass preprocessing and pretreatment facilities called ‘regional biomass
preprocessing centers’ (RBPC) in 2007 to bridge the gap between feedstock suppliers and
biorefineries. This study evaluated the technical and financial feasibility of a simple RBPC using

9

the ammonia fiber expansion (biochemical) pretreatment (AFEX™) process and integrating
animal feed and biorefinery operations.
The study suggested that RBPCs can provide solutions to the logistical concerns of
centralized biorefineries such as contracting with thousands of individual farmers, potentially
interrupted feedstock supplies, large transport and storage costs of feedstock and other business
and market power issues. These centers are envisioned to interact with growers on a local scale
and contract with them individually thus reducing complications in transactions. The proposed
RBPCs are isolated preprocessing (namely cleaning, separating and sorting, chopping, grinding,
mixing/blending, moisture control) and pretreatment facilities. In their simplest configuration
these centers produce pretreated biomass that can be supplied to the biorefinery for further
conversion. In their mature form the RBPCs are hypothesized as flexible processing facility
pretreating and converting biomass into various intermediate and final products such as fuels,
chemicals, electricity, and animal feeds. Figure 1.3 shows this original RBPC concept where the
facility interacts with other modules in the biofuel production supply chain [46].
This study suggested biorefinery capacities in the range of 5000-10000 tons/day but
considered only corn stover (which has significantly lower biomass yields compared to perennial
grasses) as raw-material and predicted large land area requirements (in the order of 0.875-1.75
million acres of corn land).
The results showed capital and operating costs of the centers and specified that RBPCs
can be economically feasible with gross margins of $3.32/ton in the worst case and $31.71/ton in
the best case scenarios. The study also stated that co-products such as animal feeds may reduce
ethanol prices by 9-20 cents per gallon, thereby supporting the concept of integration of animal
10

feeding operations with biorefineries. However, this study did not include energetic or
environmental evaluations of RBPCs. It suggested the need for further analyses of more complex
centers that are location specific and supply to a multitude of industries.

Regional biomass processing: Supply chains

Farms/
Forests

Woody materials
Grasses
Crop residues
Specialty crops
Nitrogen
Fertilizer
Minerals
Energy
feeds
Proteins

Animal
Feeders
Treated
biomass
Animal
products
Materials
producers

Ammonia

High value uses

Proteins
Enzymes
Neutraceuticals

Regional Biomass
Processing
(AFEX)

Densified
biomass

Power
plant

Minerals
Water
Treated
biomass
Sugar(s)
Enzymes

Electricity

Biorefinery

Ethanol
Chemicals
Electricity

Automotive parts
Composites
Furniture

Figure 1.3: A flow-diagram representation of the original RBPC configuration proposed by
Carolan et al. in 2007.

11

The following section contains portions of a reformatted version (to fulfill the dissertation
document requirements) of the paper: Biofuels Done Right: Land Efficient Animal Feeds Enable
Large Environmental and Energy Benefits. Dale, B.E., Bals, B.D., Kim, S., Eranki, P.L.,
Environmental Science and Technology, 2010, 44, 8385–8389.
1.5. Land efficient biofuel and animal feed production
As mentioned previously, there has been some strong opposition to biofuel production in
the form of the “food vs. fuel” controversy. The argument is regarding what the potential scale of
biofuel production could be without creating adverse effects on food supply. However, the
proponents of this argument assume that biofuel production is imposed on an agricultural system
that is static. The premise of the “biofuels done right” model proposed by Dale et al. [41]
conversely is that agriculture can change, utilizing new technology and approaches to meet and
reconcile demands for food, biofuels and environmental services. The model was designed to
investigate the technical potential for changes in US agriculture to meet the demand for large
scale biofuel production. It was developed keeping current food production and exports constant
and using a combination of existing and emerging technologies. Soil fertility was maintained or
increased and large GHG reductions were obtained simultaneously. Producing the same amount
of food on current agricultural land eliminates the so-called indirect land use change (iLUC)
effect. Over 80% of total agricultural acreage in the U.S is currently used to feed animals
(especially ruminants) [60]. Therefore, two land-efficient animal feed technologies are
considered in the study: ammonia fiber expansion (AFEX™) pretreatment to produce highly
digestible cellulosic biomass and leaf protein concentrate (LPC, an animal feed protein)
production. Only 114 million ha of current U.S cropland (amounting to less than one third of the
current U.S cropland, grassland and range) was used in order to produce animal feed, corn
12

ethanol, and exports. The feedstocks used were corn grain and stover, soy, canola, winter wheat
as a double crop, alfalfa and cellulosic biomass crops. Cellulosic biomass crops (CBCs) were
also considered including perennials, annual or a mixture of both (e.g., mixed prairie species)
that can be processed for animal feed or biofuel production. The DayCent model [48] was used
to simulate environmental impacts of all the feedstocks. Animal feeding operations were adapted
to the new feeds, thereby freeing land for biofuel production under three basic feed constrains to
balance animal diets – digestible energy (calories), protein, and rough fiber. A nonlinear
optimization model was then used to determine the two main of the greatest importance:
maximum biofuel production and maximum GHG reduction.
The current use of 114 million ha used to produce food, feed, and some biofuels when
coupled with a more land efficient approach, uses that same acreage to generate an equal amount
of food and animal feed while also providing much larger quantities of biofuels of approximately
400 GL/y. Additionally, the two main objectives stated above (ethanol and GHG reductions)
harmonize well with each other and also with production of land efficient animal feeds as seen in
Figure 1.4.
However, the model also predicted that nitrate releases may increase by 60% in the future
scenario compared to base-case. These results, based on DayCent model simulations, should be
validated with other biogeochemical models. These increases can be combated with better
agricultural practices and landscape design. The study also suggested that some candidate
cellulosic biomass crops such as Miscanthus × giganteus have reported increased nitrogen use
efficiencies and might exhibit low nitrate emissions.

13

The biofuels done right study concluded that by using enhanced forage feeds for
ruminants including leaf protein concentrates, pretreated forages, CBCs and double crops, it is
possible to maximize the productivity of the United States farmland. Using less than 30% of total
US cropland, pasture, and range, 400 billion liters of ethanol can be produced annually without
decreasing domestic food production or agricultural exports. This approach also reduces US
greenhouse gas emissions by 670 Tg CO2-equivalent per year, or over 10% of total US annual

EtOH Production (GL/yr)

500

800

400

600

300
400
200
200

100
0

0

Current

Max EtOH
Production

Max GHG
Reduction

Scenario
EtOH Production

CO2 Reduction

Figure 1.4: Ethanol production and GHG reductions from the current US agricultural system
considered in this analysis (114 million ha), that same acreage configured for maximizing
ethanol production, and that same acreage configured for maximizing GHG reductions.

14

GHG Reduction (Tg CO2 eq/yr)

emissions, while increasing soil fertility and promoting biodiversity.

1.6. Research approach
This project is divided into three categories based on the research needs identified in
literature. A representation of the research needs and how these gaps are filled is shown in Figure
1.5. The Carolan et al. study [46] formed the basis for a further investigation into the concept of
distributed processing of biomass. First, it was necessary to compare the basic energetic and
environmental performance (in terms of energy yields and GHG emissions, respectively) for the
distributed processing network vs. the current paradigm of centralized processing. In Chapter 2, a
comparative life cycle assessment (LCA) of distributed and centralized processing systems
combined with farm-scale multi-feedstock landscapes is presented. After concluding from this
study that a simple depot setup consisting only of pretreatment and densification processes is
fundamentally feasible, an exploration of the concept of potential advanced depot configurations
containing more technologies and generating valuable co-products via synergies among these
technologies was essential. Chapter 3 describes this concept of advanced regional biomass
processing depots as a potential solution to the logistical challenges of the cellulosic biofuel
industry.
It was established in the Dale et al. study [41] that generating feed and biofuel from the
same land area on national-scales using new biomass feedstocks is beneficial not only from the
energy yield perspective but also from the environmental standpoint. But, conditions such as
climate, soil, etc.) vary to a large extent on national scales. Moreover, the Sendich et al. study
[50] showed that the integration of biorefinery and animal operations is advantageous in
maximizing farm profits as well as in minimizing GHG emissions. However this study was
conducted on a farm-scale whereas in reality biomass feedstock comes from an aggregation of
many local farms.
15

Additionally, since there are multiple crops, grasses, legumes, forages and woody
material under consideration as suitable biofuel feedstock, an analysis into the land areas that
should be allotted to a combination of these feedstocks within biofuel generating landscapes is
required. Therefore, it was essential to create a regional/local scale model based on biofuel,
animal feed and environmental constraints to maximize land use efficiency and avoid iLUC
impacts that can allot optimized land areas to various biofuel feedstocks.

Investigate the
concept of
advanced depots
for distributed
processing

Distributed
processing

Identify
research
needs from
literature

Land
efficient fuel
and feed
production

Fill
research
gaps

A regional-scale
model to
maximize landefficiency while
providing animal
feeds
Integrated Life
Cycle
Assessments of
systems-wide
modulefeedstock,
processing and
transport

Integrated
farm and
biorefinery
assessments

Figure 1.5: A flowchart depicting the research approach for this project
Furthermore, thorough LCAs are also required to estimate the renewable fuel energy returns on
investment of fossil fuels and various environmental impacts in addition to GHG reductions for
the overall feedstock-processing-transport arrangement in biofuel production. In Chapter 4, an
LCA was conducted for watershed-scale cellulosic feedstock landscapes integrated with local

16

biomass processing depots. In this study, land areas were allocated manually under different
scenarios to the different feedstocks. This study formed a basis for the watershed-scale optimized
and rearranged landscape design (WORLD) model and the integrated LCA including advanced
depot configurations, described in Chapter 5. Chapter 6 presents conclusions from this project
and provides recommendations for future analyses.

17

REFERENCES

18

1.7.REFERENCES

1.
Nerurkar, N. and R. Pirog, Rising Gasoline Prices. Congressional research service
Report, 2010. R42382.
2.
Hamilton, J.D., Historical Oil Shocks. Handbook of Major Events in Economic History,
2011. The National Beureau of Economic Research (16790).
3.
Lahart, J., Rising Oil Prices Raise the Specter Of a Double Dip. The Wall Street Journal,
2011. (Accessed, August 10, 2012)
4.
Rubin, J., Just how big is Cleveland? CIBC World Markets, 2008.
http://research.cibcwm.com/res/Eco/EcoResearch.html.(Accessed, August 10, 2012)
5.
Sorrell, S., et al.,Oil futures: A comparison of global supply forecasts. Energy Policy,
2010. 38(9): p. 4990-5003.
6.
Behrens, C.E. and C. Glover, U.S. Energy: Overview and Key Statistics. Congressional
research service Report, 2012. R40187.
7.
Hill, J., et al., Climate change and health costs of air emissions from biofuels and
gasoline. Proceedings of the National Academy of Sciences of the United States of America,
2009. 106(6): p. 2077-2082.
8.
Chow, J.C., et al., Health effects of fine particulate air pollution: Lines that connect.
Journal of the Air & Waste Management Association, 2006. 56(10): p. 1368-1380.
9.
IPCC, Climate Change 2007: Synthesis Report. Intergovermental Panel on Climate
Change, Geneva, 2007. (Accessed, August 10, 2012)
10.
Gill, D.A., J.S. Picou, and L.A. Ritchie, The Exxon Valdez and BP Oil Spills: A
Comparison of Initial Social and Psychological Impacts. American Behavioral Scientist, 2011.
56(1): p. 3-23.

19

11.
DOE/EIA, International Energy Outlook. Report- DOE/EIA-0484, 2011. (Accessed,
August 10, 2012)
12.
DOE/EIA, Annual Energy Outlook 2012. Report- DOE/EIA-0383 ER, 2012. Table A7:
Transportation Sector Key Indicators and Delivered Energy Consumption. (Accessed,
August 10, 2012)
13.
NREL, Renewable Energy: An Overview. Energy Efficiency and Renewable Energy102001-1102, 2001.
14.
Dinneen, B., 2012 Ethanol Industry outlook. Renewable Fuels association, 2012.
Washington D.C. (Accessed, August 10, 2012)
15.
DOE/EIA, Annual Energy Outlook 2012. Report- DOE/EIA-0383 ER, 2012. Table A1:
Total Energy Supply and Disposition Demand. (Accessed, August 10, 2012)
16.
Runge, C.F. and B. Senauer, How Biofuels Could Starve the Poor. Foreign Affairs, 2007.
http://www.foreignaffairs.com/articles/62609/c-ford-runge-and-benjamin-senauer/howbiofuels-could-starve-the-poor. (Accessed, August 10, 2012)
17.
Searchinger, T., et al., Use of US croplands for biofuels increases greenhouse gases
through emissions from land-use change. Science, 2008. 319(5867): p. 1238-1240.
18.
Energy Independence and Security Act of 2007. Public law # 110–140, 2007. (Accessed,
August 10, 2012)
19.
Schnepf, R. and B.D. Yacobucci, Renewable Fuel Standard (RFS): Overview and Issues.
Congressional research service Report, 2012. R40155. (Accessed, August 10, 2012)
20.
DOE/EIA, Annual EnergyOutlook 2011. Report- DOE/EIA-0383(2011), 2012.
(Accessed, August 10, 2012)
21.
EPA, EPA Finalizes Regulations for the National Renewable Fuel Standard Program for
2010 and Beyond, 2010. http://www.epa.gov/oms/renewablefuels/420f10007.htm. (Accessed,
August 10, 2012)

20

22.
Aden, A., et al., Lignocellulosic biomass to ethanol process design and economics
utilizing co-current dilute acid prehydrolysis and enzymatic hydrolysis for corn stover, in
Technical Report, NREL, Editor. 2002, Storming Media. (Accessed, August 10, 2012)
23.
Hettenhaus, J., Biorefinery Feedstock Supply: Implementing Sustainable Collection of
Straw and Corn Stover. Biotechnology Industry Organization, 2005: p. 1-26.
24.
Sheehan, J., et al., Energy and Environmental Aspects of Using Corn Stover for Fuel
Ethanol. Journal of Industrial Ecology, 2003. 7(3-4): p. 117-146.
25.
Lewandowski, I., et al., The Development and Current Status of Perennial Rhizomatous
Grasses as Energy Crops in the US and Europe Biomass & Bioenergy, 2003. 25(4): p. 335-361.
26.
McLaren, J.S., Crop biotechnology provides an opportunity to develop a sustainable
future. Trends in Biotechnology, 2005. 23(7): p. 339-342.
27.
McLaughlin, S.B. and L. Adams Kszos, Development of switchgrass (Panicum virgatum)
as a bioenergy feedstock in the United States. Biomass and Bioenergy, 2005. 28(6): p. 515-535.
28.
Murnen, H.K., et al., Optimization of ammonia fiber expansion (AFEX) pretreatment and
enzymatic hydrolysis of Miscanthus x giganteus to fermentable sugars. Biotechnology Progress,
2007. 23(4): p. 846-850.
29.
Sokhansanj, S., et al., Large-scale production, harvest and logistics of switchgrass
(Panicum virgatum L.) - current technology and envisioning a mature technology. Biofuels,
Bioproducts and Biorefining, 2009. 3(2): p. 124-141.
30.
Powlson, D., A. Riche, and I. Shield, Biofuels and other approaches for decreasing fossil
fuel emissions from agriculture. Annals of Applied Biology, 2005. 146(2): p. 193-201.
31.
Schmer, M., et al., Net energy of cellulosic ethanol from switchgrass. Proceedings of the
National Academy of Sciences, 2008. 105(2): p. 464.
32.
Ceotto, E., et al., Grasslands for Bioenergy Production: A Review. Sustainable
Agriculture. 2009, Springer Netherlands. p. 141-151.

21

33.
Garlock, R.J., et al., Potential for Biofuel and Animal Feed Production From Native
Prairie Grasslands. AIChE Symposium Series, 2009. Sustainable Engineering Forum- 119(d).
(Accessed, August 10, 2012)
34.
Türe, S., D. Uzun, and I.E. Türe, The potential use of sweet sorghum as a non-polluting
source of energy. Energy, 1997. 22(1): p. 17-19.
35.
Monk, R.L., F.R. Miller, and G.G. McBee, Sorghum improvement for energy production.
Biomass, 1984. 6(1-2): p. 145-153.
36.
Dale, V.H., et al., Biofuels: Implications for Land Use and Biodiversity. Ecological
Society of America, 2010. Biofuels and Sustainability Reports. (Accessed, August 10, 2012)
37.
Fargione, J., et al., Land Clearing and the Biofuel Carbon Debt. Science, 2008.
319(5867): p. 1235-1238.
38.
Hill, J., et al., Environmental, economic, and energetic costs and benefits of biodiesel and
ethanol biofuels. Proceedings of the National Academy of Sciences, 2006. 103(30): p. 1120611210.
39.
Lal, R., Soil Carbon Sequestration Impacts on Global Climate Change and Food
Security. Science, 2004. 304(5677): p. 1623-1627.
40.
Fronning, B.E., K. Thelen, and D.H. Min, Use of Manure, Compost, and Cover Crops to
Supplant Crop Residue Carbon in Corn Stover Removed Cropping Systems. Agronomy Journal,
2008. 100(6): p. 1703-1710.
41.
Dale, B.E., et al., Biofuels Done Right: Land Efficient Animal Feeds Enable Large
Environmental and Energy Benefits. Environmental Science & Technology, 2010. 44(22): p.
8385-8389.
42.

Robertson, G.P., et al., Sustainable Biofuels Redux. Science, 2008. 322(5898): p. 49-50.

43.
Sullivan, P., Overview of Cover Crops and Green Manures: Fundamentals of Sustainable
Agriculture, 2003. http://attra.ncat.org/attra-pub/covercrop.html. (Accessed, August 10,
2012)

22

44.
Bals, B., Assessment of options for the integration of food and fuel production in
cellulosic ethanol refining, in Chemical Engineering. 2009, Michigan State University: Lansing.
p. 249.
45.
Enochian, R.V., et al., Producing Pro-Xan (Leaf Protein Concentrate) from Alfalfa:
Economies of an emerging technology. Report no. 445, 1980. USDA Agricultural Economics,
Washington DC, USA
46.
Carolan, J.E., et al., Technical and Financial Feasibility Analysis of Distributed
Bioprocessing Using Regional Biomass Pre-Processing Centers. Journal of Agricultural & Food
Industrial Organization, 2007. 5(10): p. 1-29.
47.
Sendich, E.D., B.E. Dale, and S. Kim, Comparison of crop and animal simulation options
for integration with the biorefinery. Biomass & Bioenergy, 2008. 32(12): p. 1162-1174.
48.
Del Grosso, S.J., et al., DAYCENT model analysis of past and contemporary soil N(2)O
and net greenhouse gas flux for major crops in the USA. Soil & Tillage Research, 2005. 83(1): p.
9-24.
49.
Rotz, C.A., M. Corson, and C. Coiner, Integrated Farm System Model: Reference
Manual. Version 2.1, 2007. http://ars.usda.gov/naa/pswmru. (Accessed, August 10, 2012)
50.
Sendich, E. and B.E. Dale, Environmental and economic analysis of the fully integrated
biorefinery. Global Change Biology Bioenergy, 2009. 1(1): p. 331-345.
51.
Kimes, L., Biomass Conversion: Emerging technologies, feedstocks and products.
Sustainability Program, Office of Research and Development, U.S EPA, 2007.
52.
Perlack, R.D. and A.F. Turhollow, Feedstock cost analysis of corn stover residues for
further processing. Energy, 2003. 28(14): p. 1395-1403.
53.
Cundiff, J.S., et al., Logistic Constraints in Developing Dedicated Large-Scale Bioenergy
Systems in the Southeastern United States. Journal of Environmental Engineering, 2009.
135(11): p. 1086-1096.
54.
Wright, C., K. Kenney, and J. Hess, Feedstock Supply System Design and Economics for
Herbaceous Biomass Below 15% moisture. 29th Symposium on Biotechnology for Fuels and
Chemicals, 2007. Feedstock preprocessing and supply logistics- 5A-01. (Accessed, August 10,
2012)
23

55.
Cundiff, J., R.D. Grisso, and J. Judd, Operations at satellite Storage Locations (SSL) to
deliver round bales to a biorefinery plant. Paper no: 095896. ASABE Annual International
Meeting, 2009. (Accessed, August 10, 2012)
56.
Judd, J., et al., An Optimal Storage and Transportation System for a Cellulosic Ethanol
Bio-energy Plant. Paper no: 1009413. ASABE Annual International Meeting, 2010. (Accessed,
August 10, 2012)
57.
Sokhansanj, S. and J. Fenton, Cost benefits of biomass supply and pre-processing. A
BIOCAP Research Integration Program synthesis paper, BIOCAP Canada, 2006. (Accessed,
August 10, 2012)
58.
Tiller, K., Demonstration of a Large Scale Integrated Switchgrass Supply Chain
Presented at Biomass Conference, March 30, 2010. Washington DC, USA. (Accessed, August
10, 2012)
59.
Hess, J., et al., Uniform-Format Solid Feedstock Supply System: Commodity-Scale
Production of an Infrastructure-Compatible Bulk Solid from Herbaceous Lignocellulosic
Biomass. Report INL/EXT-09-15423, 2009. Idaho National Laboratory, Idaho Falls, ID.
(Accessed, August 10, 2012)
60.
Dale, B.E., et al., Protein feeds coproduction in biomass conversion to fuels and
chemicals. Biofuels, Bioproducts and Biorefining, 2009. 3(2): p. 219-230.

24

The following chapter is a reformatted version (to fulfill the dissertation document
requirements) of the paper: Comparative life cycle assessment of centralized and distributed
biomass processing systems combined with mixed feedstock landscapes. Eranki, P.L., Dale,
B.E., Global Change Biology Bioenergy, 2011, 3(6), 427-438.

25

CHAPTER 2: COMPARATIVE LIFE CYCLE ASSESSMENT OF CENTRALIZED AND
DISTRIBUTED BIOMASS PROCESSING

2.1 Introduction

Lignocellulosic biofuels are an environmentally superior alternative to petroleum-derived
fuels. Abundant and renewable cellulosic feedstocks provide important solutions to fulfill
escalating demands for alternate liquid fuels. However their highly complex physical structure
impedes conversion into useful end-products when using biological conversion routes. As a
result, pretreatment forms the core of biomass conversion processes [1]. Biorefining is the
integrated, industrial scale production of fuels and higher value products from biomass
feedstock, similar to the petroleum refining approach [2]. A centralized, fully integrated
biorefinery includes all biomass conversion processes (i.e. size reduction, pretreatment,
hydrolysis, fermentation, distillation) in a single location. Production of large quantities of
biofuels at optimal scales for efficient capital investment requires biorefineries handling
enormous tonnages, probably of mixed feedstocks. This implies contracting with thousands of
individual farmers, potentially interrupted feedstock supplies (due to drought, etc.), large
transport and storage costs of feedstock and other business and market power issues [3].

This gap between feedstock suppliers and biorefineries can be bridged by a network of
smaller scale pre-processing facilities called “Regional Biomass Processing Depots” (RBPDs),
or just “depots” in this paper. These strategically distributed depots interact with farms producing
feedstock and with animal production operations as well as with the biorefinery and power plants
[3]. RBPDs can potentially provide benefits in environmental, economic and social

26

sustainability. In one simple configuration a depot procures, pre-processes/pre-treats, densifies
and delivers feedstock to the biorefinery and also returns animal feed to farms.

This study focuses on this simple depot configuration consisting of feedstock
pretreatment for bioethanol production and return of a single by-product (animal feed) to farms.
Biomass is procured from farms and undergoes conditioning and size reduction. Employing the
Ammonia Fiber Expansion Process (AFEX™) [4, 5] in the processing depots offers multiple
advantages as described in the previous chapter [3].Transporting the pretreated solids over long
distances requires densification to reduce transport costs and its associated environmental
impacts as well as to facilitate handling of pretreated biomass. Therefore a densification step
such as pelletization is imperative in the depots. AFEX™ pretreatment prior to densification can
improve the binding properties of lignocelluloses and enhance pellet characteristics, thereby
providing stability during storage and transport[6, 7]. Part of the pretreated solids is used as
animal feed [8]. A block diagram of this proposed RBPD is shown in Figure 2.1. Depots can also
be configured to accommodate multiple technologies such as leaf protein concentrate production,
thermochemical conversion and stem-leaf separation to deliver additional valuable by-products
to end users. For example, if wet biomass is acquired from farms it can be pulped and pressed to
extract protein concentrate [9], before being sent to the pretreatment reactor.

The primary objective of this study is to perform a comparative life cycle assessment of
distributed and centralized processing systems. Additionally, the effect of apportioning land area
to different feedstocks within a landscape on the net energy yields and greenhouse gas emissions
of the combined landscape-processing systems is assessed.

27

Auxiliary processes
Conditioning
and size
reduction
AFEX Pretreatment

Dry
biomass
Landscape

Solids Densification

(Crops and
animals)
Biorefinery

Pre-treated and densified biomass

Figure 2.1: Block diagram of a simple Regional Biomass Processing Depot (RBPD) used in this
study. All processes confined within the dashed line represent operations present in the RBPD.
2.2. Structure and methods
2.2.1 Calculations

1. Net Energy Yield (NEY) is calculated as the difference between total energy outputs
from and inputs to the cropping, transport and processing systems. All inputs are considered only
in terms of non-renewable fossil fuel energy used. Energy embodied in the feedstock is included
in the outputs only in the form of ethanol, electricity and co-product feeds generated.

28

2. Net Carbon Emissions Reduction (NCER) is calculated as the difference between CO2
equivalent greenhouse gas emission (GHG) outputs from the transport and processing systems
and the carbon sequestration effects of the agricultural systems. In this study the carbon
sequestration effect of the agricultural system is essentially the net carbon (kg CO2 equivalent)
resulting from gasoline displacement by ethanol plus the soil organic matter (SOM) sequestration
due to crop residue and no-till practices and carbon emissions during cultivation and harvest plus
annual soil organic matter losses (based on an yearly SOM maintenance parameter for different
farm location). All SOM values are converted to soil organic carbon (SOC).

3. Calculation of relative differences of NEY and NCER between distributed and
centralized processing systems:

i)

% difference in NEY (%∆ NEY) = [((NEY cent – NEY dist )/NEY cent)*100]

ii)

% difference in NCER (%∆ NCER) = [((NCER cent – NCER dist)/NCER cent)*100]

Where cent represents the centralized processing system and dist represents the
distributed processing system.

All values are in terms of dry tons wherever applicable. Co-product credit calculations,
explained further in section 2.2.2, show the conversion of non-energy co-products into energy
and emission -compatible values which are subsequently included in the NEY and NCER
calculations respectively.

29

2.2.2 Landscape analysis and LCA description

All energy and emission parameters are calculated on the basis of one kg dry feedstock.
The functional unit is a 5000 TPD (tons/day) biorefinery and the systems for comparative LCA
are: System 1: 9 RBPDs, 500 TPD each (with pelletization) + a 5000 TPD Biorefinery that
th

contains a single (10 ) 500 TPD RBPD (with no pelletization).

System 2: Centralized 5000 TPD biorefinery (no pelletization). Figure. 2.2 a, b show the
system boundaries of both systems under consideration for LCA.
The feedstocks included in this study are a continuous “corn system” consisting of corn
grain, stover and a winter double crop (rye in this case) as well as two perennial grasses –
switchgrass and miscanthus. Corn stover removal rates of 70% are based on literature values
[10]. Both perennial grasses are assumed to have average stand lives of 10 years [11, 12]. Stover
is one of the most abundant lignocellulosic resource available and is widely considered to be a
primary feedstock for cellulosic biofuels [13, 14]. Switchgrass and miscanthus have also
attracted interest as potentially important feedstocks for biofuel production [15-18]. Part of the
pre-treated perennial grasses and stover is also used as animal feed.

A total landscape area of 280 hectares (~700 acres) (in Barry County, SW Michigan) was
selected using the Web Soil Survey (WSS) (a Geographic Information System tool provided
online by the United Stated Department of Agriculture Natural Resources Conservation Service
(USDA NRCS)). Although the biorefinery size is fixed at 5000 TPD, for purposes of comparison
only a fraction of the biomass generated from the land area under consideration is investigated
here. This area is considered as a representative fragment of a larger landscape that would be
30

required to satisfy biorefinery feedstock requirements. It is intended to compare the energy yield
and greenhouse gas (GHG) emissions for this fixed land area but with different acreage
distributions among the primary crops of interest. To achieve this objective, different
configurations with varying acreages allotted to different feedstocks are combined individually
with each processing system and these landscape-processing systems are then compared with
each other.

The amount of biofuel (bioethanol in this case) and electricity generated from each of
these configurations is the same irrespective of the type of feedstock used because of the
unchanging scale of biorefinery; therefore they can be compared without ambiguity. Similarly,
the basis of 1 kg dry feedstock was chosen since one of the two most important aspects of this
study is feedstock allocation within a given landscape. This allows for fair comparison because
the land area used remains constant for all configurations.

For this fixed land area three configurations were formulated to evaluate the effect of
decreased acreage in the corn system and increased perennial grasses. It is assumed that this
fixed land area is a “clean-slate” where any crops grown on this land area prior to this analysis
are ignored. This assumption is valid based on the impetus that the use of marginal lands to grow
cellulosic biofuel feedstocks is gaining [19, 20]. Marginal land is defined as any land not being
used to grow commercial or conventional crops or abandoned and degraded cropland [21] that
may be capable of growing low-maintenance biomass feedstocks such as perennial grasses [22].

31

(a)

Fertilizers, fuel, all
other inputs

Landscape

Process inputs

9 RBPDs
Feedstock
Transport

Densified
material
Transport

Process inputs

Large Scale
th
Biorefinery + 10
RBPD without
pelletization

Animal feed

Ethanol, electricity

(b)
Fertilizer, fuel, all
other inputs
Landscape
Feedstock
Transport

Process Inputs

Integrated/ Centralized
Biorefinery

Animal feed

Ethanol, electricity

Figure 2.2: System boundaries for comparative LCA - (a) Distributed processing system, (b)
Centralized processing system. Both processing systems are combined with the mixed feedstock
landscapes and animal operations for an integrated system-wide analysis

32

In Configuration 1, 65% of the acreage was dedicated to the corn system and the
remaining acreage was divided equally between the two perennial grasses. In Configuration 2
only 15% of the acreage was allotted to the corn system and the remaining 85% was divided
equally between the grasses. In Configuration 3 all the acreage was divided equally between
these two perennial grasses. There is a subdivision within each configuration acting as an
embedded sensitivity analysis to assess the impact of grass yields on the results.

A high grass yield (HG) of 10 tons/ac (24.7 tons/ha) yield was chosen for switchgrass and
12 tons/ac (29.6 tons/ha) for miscanthus while a lower grass yield (LG) of 7 tons/ac (17.3
tons/ha) yield was chosen for switchgrass and 8 tons/ac (19.8 tons/ha) for miscanthus. The high
and low yields were selected based on literature values from various publications for switchgrass
[23, 24] and for miscanthus [25, 26]. The perennial grasses were assumed to have an
“establishment period” of 3 years in which they have negligible yields but post- establishment
they have significantly higher yields and much lower maintenance requirements than the annual
crops. Yields are assumed to be on a dry mass basis and with moisture contents of 15% at
harvest [12, 27]. Yields of both perennial grasses are averaged over their entire stand-life.

Yields for the corn system were obtained from the Web Soil Survey (WSS) for available
crops or estimated for unavailable crops based on the yields of similar crops present in the area
of interest. For example, winter rye yields were calculated on a weight basis compared to winter
wheat and as mentioned previously perennial grass yields were obtained from literature. After
obtaining yields crop cultivation and harvest energy and emission values were estimated using
crop budgeting spreadsheets (K Thelen, 2010, Personal communication). An extensive literature
review was conducted for each crop to obtain inputs including fertilizers, insecticides, fuel used

33

in cultivation and harvest, seeds required or roots transplanted (data available in Tables A1-A4,
Appendix A). The budget spreadsheets provide details such as increases in soil organic matter
(SOM) when using no-till practices and losses in SOM based on farm locations. Figure 2.3
shows fractions of biomass yields of each crop in each configuration for the fixed land area. The
corn system contributes the greatest biomass in Configuration 1 as expected since it occupies the
largest portion of land. Moreover, its biomass fraction increases relatively in low grass yield
scenarios compared to that in high grass yield scenarios in all the applicable configurations.
Similarly, the fraction of biomass from perennial grasses increases with an increase in their land
area allowance. On-farm animals form an integral part of these landscape analyses. It is assumed
that ruminant animals are present within the landscape at a stocking rate of 2 animals/ac
[28].These ruminant animals consume part of the pretreated feedstock and their methane
emissions are calculated, converted to CO2 equivalent and included in the NCE impacts.

Carolan et al. [3] state that animal feed is an important by-product of the AFEX™
pretreatment method. Therefore, co-product credit calculations for animal feed are an important
aspect of this LCA. Here it is necessary to calculate displacement ratios for all the lignocellulosic
feedstocks included in this study. The displacement ratio is defined as the amount of
conventional animal feeds (corn and soybean meal) that pretreated lignocellulosic feedstock can
replace based on animal nutritional requirements.

34

Corn system

Switchgrass

Miscanthus

1
0.9

0.8

Fraction of total yield

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1 HG

1 LG

2 HG

2 LG

3 HG

3 LG

Figure 2.3: Fractional yields of feedstocks in different configurations. The land area remains the
same in each configuration but feedstock acreages vary. Yields of cropping systems vary
proportionally with their acreages in each configuration. Corn system acreages as % of total land
area: Configuration 1 = 65%, Configuration 2 =15%, Configuration 3=0. Combined perennial
grass acreage Configuration 1 =35%, Configuration 2 = 85%, Configuration 3= 100%, divided
equally between switchgrass and miscanthus. HG represents high yields and LG represents low
yields of the two perennial grasses
These calculations were performed for stover, switchgrass and miscanthus based on the
amount of energy, protein and fiber (in case of grasses replacing hay) replaced by lignocellulosic
feedstock compared to the conventional feedstock. The following illustration shows
displacement ratio calculation for stover.

Values for digestible fiber, energy and protein for both replacing and replaced animal
feeds are shown in Table 1 (BD Bals, 2010, Personal communication).

35

Oil content of soybean =0.196, meal = 1-0.196 = 0.804.

Based on equalizing nutritional values:

For corn stover displacing corn and soybean meal:

0.887*C+ (0.654/0.804)*S = 0.756

0.094*C+ (0.386/0.804)*S = 0.172

Where C and S represent corn and soybean meal respectively

Solving for C, S:

C= 0.638476 and S= 0.233298

Similar calculations were performed for perennial grasses. Due to absence of nutritional
value data for miscanthus its displacement ratios are assumed to be the same as switchgrass.
Displacement ratios were then incorporated in equations (adopted from Edward and Anex [29]
and modified for this study) for co-product credit calculations. The following illustration shows
co-product credit calculation for stover:

Corn stover energy credit for displacing corn [MJ] = [animal feed production (kg feed/kg
stover) * feed displaced (kg corn displaced/kg animal feed)] * corn production energy (MJ/kg
corn)

36

Corn stover energy credit for displacing soybean [MJ] = [animal feed production (kg
feed/kg stover) * feed displaced (kg soybean displaced/kg animal feed)] * soybean production
energy (MJ/kg soybean)

Table 2.1: Nutritional value of feedstocks
Nutritional value

Corn

a

Soybean Switchgrassb

Stover

parameter

Grass
hay

Energy

0.887

0.756

0.654

0.630

Protein

0.094

0.172

0.386

0.014

0.819

(including non-protein
nitrogen)
Fiber

-

-

-

Oil

-

-

0.577

0.196

a. Stover is assumed to displace corn and soybean meal.
b. Perennial grasses (switchgrass and miscanthus) are assumed to displace corn, soybean
meal and fiber from grass hay.

Assumed animal feed production =1 kg/1 kg of stover, energy inputs for corn production
is obtained from literature review and displacement ratios (for feed displaced) are calculated as
stated previously. Similarly energy and emission credits were obtained for each lignocellulosic
feedstock and included in NEY and NCER calculations.

Displacement ratios are based on direct substitution of un-pretreated feedstock for lack of
actual data from animal feed trials (i.e. feeding animals with pre-treated lignocellulosic feedstock
used in this study). However, AFEX™ treated rice straw fed to dairy cows has shown higher
37

neutral detergent fiber intake and milk yield [30]. Also, preliminary analysis has suggested that
100% of beef cattle nutritional requirements and up to 70% of dairy cattle nutritional
requirements (along with a protein supplement and grain silage) can be met with by using
AFEX™ treated animal feed depending on the age of the cattle [8, 31]. Until large-scale animal
feed trials are conducted, these displacement ratios are assumed to be applicable. Animal feed is
included in system boundaries of both processing systems (as were transportation and landscape
values) since this is an integrated system-wide analysis. Although this does not affect
comparative calculations in this case, it is possible in future analyses that only certain RBPDs
send back animal feed to farms based on their location, feedstock type processed and
technologies included, in which case results from comparative studies may vary.

Processing energy and emissions were obtained from the NREL/Dartmouth Aspen plus
biorefinery model [8]. This is the principal simulation model for US cellulosic ethanol
production in a centralized biorefinery. The model contains all the conversion processes for
ethanol production namely feedstock handling, pretreatment, biological conversion (hydrolysis
and fermentation), product recovery, utilities production, and waste treatment [32]. The RBPD
energy and emissions were calculated by isolating processes applicable to the depots and scaling
them to its lower capacity (compared to the fully integrated centralized biorefinery). Processes
absent in the depots such as biological conversion and ethanol purification and recovery were
excluded from the model. The process energy and emissions for densification were obtained
from literature [33] for all 9 depots in the network except for the tenth one which is co-located
with the biorefinery. While incorporating densification values from literature, it was ascertained
that that the properties of pre-treated feedstock are compatible with the conditions required for
densification [34] (densification details available in Table A5, Appendix A). Similarly, energy
38

requirements and emissions generated during transportation of densified or non- densified
feedstock were added to the distributed and centralized processing system, respectively, using
literature data [33].

The energy and emission inputs and outputs from all the sources discussed above were
aggregated and the NEY and NCER values were calculated for each cropping system, combined
with distributed or centralized processing and their respective transportation information. Table
2.2 shows all the modules and inputs included in this LCA.

There are two chief assumptions in this study. Firstly, it is assumed that the distributed
processing network taken as a whole is as energy self-sufficient as the centralized system.

Accordingly all the energy inputs for the distributed system are estimated as a direct
scale-down of the integrated biorefinery model. This is true if lignin-rich process residues are
burnt for electricity [32] or if other energy sources such as heat/electricity from thermochemical
conversion or methane-rich biogas from anaerobic digestion of manure [35] are present in the
distributed networks. Second, the transport differences between the two processing systems are
accounted for in terms of transporting bales (where no densification is involved-in the
centralized biorefinery nor in the tenth depot) versus pellets (where densification is present- in
the 9 depots) for an average transportation radius of anywhere between 20 and 100 km for both
processing systems as found in literature [33].

39

Table 2.2: Unit processes in the LCA and their energy inputs and emission outputs
System module

Input

Energy input and Emissions output per kg dry
a

feedstock . E= Energy (MJ), C = Emissions (kg
CO2 eq.)
Feedstock

Energy for

production and

cultivation and

b

d

E

C

4.274

-0.854

Switchgrass

0.235

-0.916

Miscanthus

0.176

-0.551

Processing

Process

E

C

energy

Single RBPD (excluding

0.74

0.0001

0.05

Negligible

E

C

Densified

0.2954

0.0233

Non- densified

1.0434

0.06

Cropping system
Corn System

harvest, animals

RBPD network

c

harvest.

pelletization)
Pelletization
Densified and non

Transport

densified biomass

energy

transport
Biorefinery

Processing

E =13.24, C =0.0018

energy
a. All values are based on kg dry feedstock and represent biomass derived only from
the landscape area under consideration.
b. The values for cropping systems are entered as an average of all three configurations
and for the perennial grasses as an average of high and low grass yields over all three
configurations.
c. All values for combined cropping system of corn grain, stover and rye.
d. All emissions from crops are in terms of sequestered kg CO2 equivalent and hence
have a negative sign.

40

For the distributed processing network it is assumed that the transport distance between
the farms and the depots is negligible compared to the transport distance of pellets from depots to
the biorefinery. In the case of the centralized biorefinery it is assumed that all biomass is directly
transported from the farms to the refinery in the form of conventional bales.

2.3. Results
2.3.1. Base case scenario

According to this analysis, Configuration 1, where 65% of the acreage was allotted to the
corn system, with high grass yields (HG) has the greatest NEY because it exhibits the greatest
total biomass production from all sources (grain, stover, rye and perennial grasses). Where the
corn system dominates the acreage (for both low and high grass yields) the centralized
processing system has a greater NEY than the distributed system (in Configuration 1). Rye and
grain, which make up approximately 40% of total biomass in this configuration, are not densified
prior to transport, thereby eliminating a primary advantage of the depots. As the acreage
dedicated to the corn system gradually decreases and the total amount of densified biomass
increases in Configurations 2 and 3 (with 85% and 100% of the entire acreage dedicated to
perennial grasses respectively), the NEY of the distributed processing system surpasses that of
the centralized system. On average, miscanthus has 15% and 56% greater NEY than the corn
system and switchgrass respectively because of its greater biomass yield. The corn system on the
other hand has 48% greater NEY than switchgrass since the combined biomass from the corn
system is greater than that from switchgrass alone. Figure 2.4 shows the NEY of the two
processing systems in different landscape configurations per unit land area.

41

Perennial grasses can contribute large quantities of soil organic matter over time because
of their thick root masses and since they do not need to be replanted each year [36]. Moreover,
no-till agricultural systems for corn also increase the amount of carbon stored in soil [37].
According to this analysis, switchgrass has the greatest potential for carbon sequestration among
all the cropping systems. The corn system has a comparable carbon sequestration potential to
switchgrass.

Distributed

1.2E+05

Centralized

1.0E+05

NEY (MJ/ha)

8.0E+04

6.0E+04
4.0E+04
2.0E+04
0.0E+00
1 HG

1 LG

2 HG

2 LG

3 HG

3 LG

Figure 2.4: Net Energy Yield (NEY) per hectare of the collective system (landscape, transport,
processing) in different configurations. The NEY (MJ) of the collective system comprises
differences in energy inputs and outputs of each processing system (distributed or centralized)
combined with crops and transport. HG represents high yields and LG represents low yields of
the two perennial grasses
Switchgrass and the corn system both have about 30% greater sequestration potential
than miscanthus on average for all configurations. Although both switchgrass and miscanthus are
perennial grasses, their different sequestration potential is due to the differences in inputs and
42

planting practices. Switchgrass and the corn-system are assumed to be cultivated using no-till
whereas miscanthus is planted conventionally since it is a rhizome and must be propagated
asexually [38]. Moreover, the carbon sequestration potential is related to yields because the total
above-ground residue declines with decreasing yields. Consequently, the total residue
contributing to soil organic matter increase decreases, thereby reducing carbon sequestration
potential. This is the reason for differences in sequestration potential between low and high
yielding grasses. It is to be noted that to calculate NCER due to gasoline displaced by ethanol
for individual processing systems, a reasonable evaluation would be based on equivalent service
provided by the two fuels; however this is a case of relative comparison between sequestration
potentials of different cropping systems. Therefore as stated previously this “closed-system”
assumption is valid because the land area remains constant in all configurations and a
comparison is made within this land area between different cropping systems.

The NCER results for each configuration correspond to the sequestration potentials of
each cropping system. The highest NCER occurs when the entire acreage is allotted to high
yielding grasses (Configuration 3 HG) mainly because high yielding switchgrass is dominant in
this configuration. In contrast, the lowest emission reduction occurs with all acreage allotted to
low yielding grasses (in Configuration 3 LG). This is because the sequestration potential of the
combined corn system is comparable to that of the grasses and at low yields the sequestration
due to grasses does not exceed that of the corn system. Figure 2.5 shows the NCER of combined
landscape and processing systems in different configurations per unit land area.

43

1 HG

1 LG

2 HG

2 LG

3 HG

3 LG

NCE (kg CO2 eq/ha)

0.0E+00

-5.0E+03

-1.0E+04

-1.5E+04

-2.0E+04
Distributed

Centralized

-2.5E+04

Figure 2.5: Net Carbon Emission Reductions (NCER) per hectare of the collective system
(landscape, transport, processing) in different configurations. The NCER (kg CO2 eq.) of the
collective system comprises differences in GHG emissions generated by processing system
(distributed or centralized) combined with transport, and animal operations and carbon
sequestered by feedstocks. HG represents high yields and LG represents low yields of the two
perennial grasses
2.3.2 Sensitivity Analysis

Sensitivity analyses were performed to identify significant system variables. Table 2.3
summarizes the results of variations in NEY and NCER between the two processing systems.

2.3.2.1 Densification

The base case scenario established the fact that pretreated perennial grass densification is
a key contributor to the distributed processing network. Densification reduces both the
44

environmental impacts and economic costs of transportation. Hence choosing the right
densification method is essential. Three separate densification processes were considered in this
sensitivity: briquettes, (“PAKs”) as produced by Federal Machine (Fargo, ND), pelletizing as
performed as part of the Pro-Xan process on dehydrated alfalfa pellets and a ring-die
densification process (see Table 2.3 for references). The energy requirements and emissions for
these processes were incorporated into the distributed processing system calculations. The
energy requirements for densification in these methods differ by 25, -67 and 78% respectively
compared to base-case energy requirements for pelletization. The emissions generated are not
significantly different compared to base-case.

The densification method can be a considerable source of variation in NEY as seen in
Table 2.3 and causes small deviations in the NCER differences between the two processing
systems. Selection of a densification method will depend largely on process economics.

2.3.2.2. Transport

The base-case scenario incorporated pelletized biomass transport energy and emission
information obtained from the IBSAL model (Integrated Biomass Supply Analysis and
Logistics) [41]. In this sensitivity analysis, we used transportation emissions and energy inputs
for non-densified biomass from the GREET (Greenhouse Gases, Regulated Emissions, and
Energy Use in Transportation) model for comparison. As seen in Sensitivity 1, using different
methods of densification alters the NEY values of depots. Similarly, this sensitivity indirectly
illustrates the variations in NEY and NCER when densification is not used in distributed
processing networks.

45

Table 2.3: Data sources and results of sensitivity analyses
Parameter

References

%∆

%∆
a

b

NEY

NCER

S0- Base case

[33]

0.09

-3.7

S1,1- PAKs process

(DJ Marshall, 2010,

2

-4.5

[9]

-5

-4.5

S1,3-Ring/Die process

[39]

34

-4.5

S2- Transport

[40]

10

0

S3- Credits for conversion to

-

0.05

-2.4

-

0.14

-3.6

Personal Communication )
S1,2- Pelletization from ProXan process

perennial grasses
S4- Absence of double-crop

a. Percentage difference in NEY (%∆ NEY) = [((NEY cent – NEY dist )/NEY cent)*100].
b. Percentage difference in NCER (%∆ NCER) = [((NCER cent – NCER dist)/NCER
cent)*100]

2.3.2.3. Credit for conversion to perennial grasses

Growing perennial grasses instead of annual crops on the same land area can result in
environmental improvements. For example, eliminating annual cultivation and monocultures can
benefit farmland biodiversity. Perennials can increase soil organic matter content thereby
improving carbon sequestration and soil quality [23]. Moreover, energy inputs and maintenance
costs of annual crops are higher than for perennial crops .The base-case scenario was based on
46

the assumption of a “clean-slate land area” i.e. there were no crops present on the landscape prior
to this analysis. It is also necessary to analyze possible changes resulting from land area
conversions from existing agricultural system to perennial grasses [42]. Savings in energy inputs
and carbon sequestration due to growing each perennial grass instead of the corn system were
calculated. The corn system has greater energy inputs than both perennial grasses (Table 2.1,
section 2.2); therefore an energy gain on an average of 2.7 MJ/ kg dry biomass and 2.3 MJ/ kg
dry biomass is predicted for switchgrass and miscanthus respectively. In Configuration 3 where
no corn system was initially present the same energy inputs as corn system in Configuration 1
were assumed. Carbon sequestration increases for all configurations except for miscanthus in
Configuration 1. As mentioned in section 2.2.2 miscanthus has lower carbon sequestration
benefits than the corn system because of different tillage practices. Therefore in Configuration 1
growing miscanthus instead of the corn system on the allotted acreage is unfavorable for
sequestration. On average, carbon sequestration gains of 0.75 kg CO2 eq / kg dry biomass and
0.36 kg CO2 eq/ kg dry biomass are observed for switchgrass and miscanthus, respectively. The
relative differences in NEY and NCER between the two processing systems are nearly the same
(Table 2.3).

2.3.2.4. Absence of double crop

Double crops are attracting interest as a method to maintain or increase soil carbon
content after harvesting agricultural residues, mainly corn stover [43, 44]. This is the primary
reason for including the winter rye double crop in the base-case scenario. In this sensitivity, the
double crop was removed from the corn system and this cropping system was reduced to only
corn grain and stover production. Removing the double crop from the system, which undergoes
47

densified transportation, has its primary negative effect on the NEY of the distributed processing
system. The NCER values are practically unaltered.

Both sensitivity analyses 3 and 4 involve varying major components of the cropping
system. These components are integral to emission reductions and contribute significantly to
system energy consumption. Therefore it is important to assess individual variations in NEY and
NCER compared to the base case (further discussed in Section 2.4) because the effects of these
changes are not apparent in the relative differences between the two processing systems.

2.4. Conclusion

Based on this analysis, to achieve NEY and NCER objectives, the entire acreage should
be dedicated to perennial grasses only when their yields are high. But when perennial grass
yields are low, it is more advantageous to adopt a landscape configuration containing mostly
perennial grasses but including some corn system acreage. The distributed processing system has
consistently greater NEY and NCER than the centralized system when combined with perennial
grasses. Additionally, different perennial grass yields change NEY values by 15 to 50% and
NCER values by 20 to 65% in each configuration for both processing systems. On average, the
distributed processing system has practically the same NEY as and a 3.7 % greater NCER than
the centralized processing system.

This study also highlights the fact that distributed processing networks when combined
with densified high yielding perennial grasses have consistently greater energy yields as well as
larger emission reductions than centralized processing systems. Therefore dedicated energy
feedstock landscapes (using perennial grasses) work best where grass yields are high and some

48

form of densification is involved. However, if most feedstock is trucked as bales and if grass
yields are low it is unlikely that the NEY of distributed processing systems will exceed that of
centralized processing systems. Evaluating the impacts of landscape conversion from highmaintenance annuals to low-maintenance perennials is also important because of the reduced
energy inputs and carbon sequestration benefits of the latter systems. For sensitivity analysis
cases 3 and 4 it is more effective to look at the individual differences compared to the base case
for NEY and NCER rather than examining relative differences in the two processing systems.
The third sensitivity analyses (credit for conversion to perennial grasses) better highlights energy
inputs and emission reductions due to growing perennial grasses instead of annuals if each
processing system were evaluated individually. It emphasizes the importance of a detailed
analysis to assess the energy and carbon sequestration characteristics of each cropping system.
Evaluations compared to the base case showed increased NEY values ranging between 13 to
33% and increased NCER values ranging between 8 and 53% for the different configurations
averaged over the two processing systems and low and high grass yield cases. Similarly,
although in the fourth sensitivity analysis (absence of double crop) the relative differences in
emissions and energy yields of the overall systems are nearly the same, individual evaluations
averaged over the two processing systems and low and high grass yields shows decreased NEY
values ranging from 8 to 22% and decreased NCER figures ranging between 5 and 21%.
Analyses such as these can help determine the most sustainable land configurations within mixed
feedstock landscapes in the RBPD context.

The economic performance of these depots is an important factor but is outside the scope
of this analysis, nor do the data and tools used in this analysis permit the evaluation of other
environmental impacts such as water quality or biodiversity. The conclusions from this study
49

probably apply to systems containing similar combinations of crops and land areas. Landscapes
with different soil conditions, cropping systems and yields will almost certainly require similar
analyses. Hence further research using more advanced tools such as ArcGIS for landscape
studies is underway. There is a necessity for the development of flexible models for sustainable
landscape configurations combined with distributed processing based on varying yields, soil
conditions, landscape sizes and processing technologies. Distributed processing networks using
densified biomass may be able to catalyze the formation of commodity cellulosic biomass
markets, thereby providing grower incentives and advancing biofuel production. Modeling the
logistics and conversion technologies and performing integrated systems investigations is a
stepping stone in the successful establishment of large-scale sustainable lignocellulosic biofuel
industries.

50

APPENDIX A

51

2.5. Appendix A
A1. Crop inputs used in agricultural budget spreadsheets
Table A1. Corn grain and stover –inputs (all inputs for stover are same as that for
corn)
Corn grain and stover

Quantity

Unit

% SOM

1.70

% SOM

SOM per Acre

3.E+04

lbs/ac

SOM Lost per year

1020

lbs/ac/yr

Corn Grain

106.64

bu

Biofuel Energy Component

298.592

gal/ac

Weight of Crop

5971.84

lbs/ac

Harvest Efficiency

0.50

Shoot-to-Root Ratio

8.5

Total Aboveground Biomass

11944

lbs/ac

Aboveground Residue

5972

lbs/ac

Belowground Residue

1405

lbs/ac

Total Residue

7377

lbs/ac

Residue Contribution to SOM

1475

lbs/ac

No-till or perennial crop SOM credit

15.00%

% SOM increase

Seed

0.24

units

12

lbs/ac

31

lbs

N - NH3
Insecticides

52

Table A1 (Cont’d)
Imidacloprid

0.32

oz

Atrazine

0.5

qt

Glyphosate

12

oz

Drying

106.64

bu

Fuel, oil, lube

3.15

gal

Herbicides

Table A2. Winter rye inputs
Winter rye

Quantity

Unit

% SOM

1.70%

% SOM

SOM per Acre

3.E+04

lbs/ac

SOM Lost per year

1020

lbs/ac/yr

Biomass

2.5

ton

Biofuel Energy Component

180

gal/ac

Weight of Crop

5000

lbs/ac

Harvest Efficiency

0.95

Shoot-to-Root Ratio

11.6

Total Aboveground Biomass

5263

Aboveground Residue

263

Belowground Residue

454

lbs/ac

Total Residue

717

lbs/ac

Residue Contribution to SOM

143

lbs/ac

53

lbs/ac

Table A2 (Cont’d)
No-till or perennial crop SOM credit 15.00% SOM increase
Seed

2

bu

90

lbs/ac

N - urea

60

lbs

Drying

89.3

bu

Fuel, oil, lube

1.63

gal

Table A3. Switchgrass inputs
Switchgrass

Quantity

Unit

% SOM

1.70%

% SOM

SOM per Acre

3.E+04

lbs/ac

SOM Lost per year

680

lbs/ac/yr

Biomass

10

ton

Biofuel Energy Component

720

gal/ac

Weight of Crop

20000

lbs/ac

Harvest Efficiency

0.80

Shoot-to-Root Ratio

0.5

Total Aboveground Biomass

25000

Aboveground Residue

5000

Belowground Residue

50000

lbs/ac

Total Residue

55000

lbs/ac

Residue Contribution to SOM

11000

lbs/ac

No-till or perennial crop SOM credit

15.00%

SOM increase

54

lbs/ac

Table A3 (Cont’d)
Seed (1st year only)

7

lbs

N - NH3

88

lbs

Atrazine

1.5

qt

Drying

333.33

bu

Fuel, oil, lube

9.1

gal

Table A4: Miscanthus inputs
Miscanthus

Quantity

Unit

% SOM

1.70%

% SOM

SOM per Acre

3.E+04

lbs/ac

SOM Lost per year

1020

lbs/ac/yr

Biomass

12

ton

Biofuel Energy Component

864

gal/ac

Weight of Crop

24000

lbs/ac

Harvest Efficiency

0.75

Shoot-to-Root Ratio

1.2

Total Aboveground Biomass

32000

Aboveground Residue

8000

Belowground Residue

27350

lbs/ac

Total Residue

35350

lbs/ac

Residue Contribution to SOM

7070

lbs/ac

No-till or perennial crop SOM credit

0.00%

SOM increase

55

lbs/ac

Table A4 (Cont’d)
Root

4000

root

40

lbs/ac

N - NH3

63

lbs

P2O5

41.2

lbs

K2O

17.352

lbs

Drying

399.96

bu

Fuel, oil, lube

2.25

gal

A2. Densification
A typical pelleting operation of un-pretreated biomass consists of drying, size-reduction
and compaction. In the case of pre-treated biomass, the feedstock will already have undergone
preliminary size reduction and drying before pretreatment. The ground material is treated with
super-heated steam at temperatures above 100°C before compaction. The superheated steam
increases moisture and temperature of the mash causing the release and activation of the natural
binders present in the biomass .In this study this is achieved by the use of AFEX™
pretreatment. The following table summarizes densification conditions and energy requirements
[33, 34]:

56

Table A5: Values of parameters used in densification
Parameter

Values

Moisture content of incoming biomass

~10%

Energy requirement of processes (GJ/t)
Drying

0.35

Hammer mill (size reduction)

0.1

Pellet mill

0.268

Pellet cooler

0.013

Temperature of outgoing pellets before cooling

70-90 oC

Temperature of final pellets after cooling

Within 5 oC of ambient

57

REFERENCES

58

2.6. REFERENCES

1.
McMillan, J.D., Pretreatment of Lignocellulosic Biomass. Enzymatic Conversion of
Biomass for Fuels Production, 1994: p. 292-324.
2.
Kimes, L., Biomass Conversion: Emerging technologies, feedstocks and products.
Sustainability Program, Office of Research and Development, U.S EPA, 2007.
3.
Carolan, J.E., et al., Technical and Financial Feasibility Analysis of Distributed
Bioprocessing Using Regional Biomass Pre-Processing Centers. Journal of Agricultural & Food
Industrial Organization, 2007. 5(10): p. 1-29.
4.
Balan, V., et al., Fundamental Understanding of Biomass Pretreatment Technologies:
The Case of Ammonia Fiber Expansion (AFEX). 2006.
5.
Sendich, E., et al., Recent process improvements for the ammonia fiber expansion
(AFEX) process and resulting reductions in minimum ethanol selling price. Bioresource
Technology, 2008. 99(17): p. 8429-8435.
6.
Sokhansanj, S., et al., Binderless Pelletization of Biomass. Paper no: 056061.ASAE
Annual International Meeting, Tampa, FL.. St. Joseph, MI. (Accessed, August 10, 2012)
7.
Dale, B.E., Densifying & handling AFEX biomass : A cooperative research project.
Northern Plains Biomass Economy: What makes sense?, 2009. (Accessed, August 10, 2012)
8.
Sendich, E. and B.E. Dale, Environmental and economic analysis of the fully integrated
biorefinery. Global Change Biology Bioenergy, 2009. 1(1): p. 331-345.
9.
Enochian, R.V., et al., Producing Pro-Xan (Leaf Protein Concentrate) from Alfalfa:
Economies of an emerging technology. Report no. 445, 1980. USDA Agricultural Economics,
Washington DC, USA
10.
Graham, R., et al., Current and potential US corn stover supplies. Agronomy Journal,
2007. 99(1): p. 1.

59

11.
Duffy, M. and V. Nanhou, Costs of Producing Switchgrass for Biomass in Southern
Iowa. Iowa State University, Department of Economics, 2002. (Accessed, August 10, 2012)
12.
Scurlock, J.M.O., Miscanthus: A Review of European Experience with a Novel Energy
Crop. Publication No: 4845. ORNL Environmental Sciences Division, 1999. (Accessed, August
10, 2012)
13.
Aden, A., et al., Lignocellulosic biomass to ethanol process design and economics
utilizing co-current dilute acid prehydrolysis and enzymatic hydrolysis for corn stover, in
Technical Report, NREL, Editor. 2002, Storming Media. (Accessed, August 10, 2012)
14.
Hettenhaus, J., Biorefinery Feedstock Supply: Implementing Sustainable Collection of
Straw and Corn Stover. Biotechnology Industry Organization, 2005: p. 1-26.
15.
Lewandowski, I., et al., The Development and Current Status of Perennial Rhizomatous
Grasses as Energy Crops in the US and Europe Biomass & Bioenergy, 2003. 25(4): p. 335-361.
16.
McLaren, J.S., Crop biotechnology provides an opportunity to develop a sustainable
future. Trends in Biotechnology, 2005. 23(7): p. 339-342.
17.
McLaughlin, S.B. and L. Adams Kszos, Development of switchgrass (Panicum virgatum)
as a bioenergy feedstock in the United States. Biomass and Bioenergy, 2005. 28(6): p. 515-535.
18.
Murnen, H.K., et al., Optimization of ammonia fiber expansion (AFEX) pretreatment and
enzymatic hydrolysis of Miscanthus x giganteus to fermentable sugars. Biotechnology Progress,
2007. 23(4): p. 846-850.
19.
Hill, J., et al., Environmental, economic, and energetic costs and benefits of biodiesel and
ethanol biofuels. Proceedings of the National Academy of Sciences, 2006. 103(30): p. 1120611210.
20.
Fargione, J., et al., Land Clearing and the Biofuel Carbon Debt. Science, 2008. 319
(5867): p. 1235-1238.
21.
Dale, V.H., et al., Biofuels: Implications for Land Use and Biodiversity. Ecological
Society of America, 2010. Biofuels and Sustainability Reports. (Accessed, August 10, 2012)

60

22.
Schmer, M., et al., Net energy of cellulosic ethanol from switchgrass. Proceedings of the
National Academy of Sciences, 2008. 105(2): p. 464.
23.
Powlson, D., A. Riche, and I. Shield, Biofuels and other approaches for decreasing fossil
fuel emissions from agriculture. Annals of Applied Biology, 2005. 146(2): p. 193-201.
24.
Goddard, K.J., Switchgrass for Biofuels. UT Biofuels Initiative, 2007. (Accessed, August
10, 2012)
25.
Gibson, L., Miscanthus Hybrids for biomass production. Iowa State University,
Department of Agronomy, 2007. (Accessed, August 10, 2012)
26.
Thelen, K., et al., Biomass production: Switchgrass and Miscanthus x giganteus.
Michigan Agri-energy conference, 2009. (Accessed, August 10, 2012)
27.
Sokhansanj, S., et al., Large-scale production, harvest and logistics of switchgrass
(Panicum virgatum L.) - current technology and envisioning a mature technology. Biofuels,
Bioproducts and Biorefining, 2009. 3(2): p. 124-141.
28.
Sendich, E., Modeling and analysis of the biorefinery integrated with the agricultural
landscape, in Department of Chemical Engineering and Material Science. 2008, Michigan State
University: Lansing. p. 184.
29.
Edwards, K. and R.P. Anex, Co-product Allocation in Life Cycle Assessment: A Case
Study. Extended Abstract, Trinity consultants, 2009. (Accessed, August 10, 2012)
30.
Weimer, P.J., et al., FIBEX-treated rice straw as a feed ingredient for lactating dairy
cows. Animal Feed Science and Technology, 2003. 103(1-4): p. 41-50.
31.
Dale, B.E., Biofuels, food, economic development and the environment: Designing the
biofuels future. USAIN conference, 2008. (Accessed, August 10, 2012)
32.
Laser, M., et al., Coproduction of ethanol and power from switchgrass. Biofuels,
Bioproducts and Biorefining, 2009. 3(2): p. 195-218.
33.
Sokhansanj, S. and J. Fenton, Cost benefits of biomass supply and pre-processing. A
BIOCAP Research Integration Program synthesis paper, BIOCAP Canada, 2006. (Accessed,
August 10, 2012)
61

34.
Mani, S., et al., Economics of producing fuel pellets from biomass. Journal Name:
Applied Engineering in Agriculture; Journal Volume: 22; Journal Issue: 3, 2006; page(s) 421426.
35.
Laser, M., et al., Projected mature technology scenarios for conversion of cellulosic
biomass to ethanol with coproduction thermochemical fuels, power, and/or animal feed protein.
Biofuels, Bioproducts and Biorefining, 2009. 3(2): p. 231-246.
36.
Post, W.M. and K.C. Kwon, Soil carbon sequestration and land-use change: processes
and potential. Global Change Biology, 2000. 6(3): p. 317-327.
37.
Bernacchi, C.J., S.E. Hollinger, and T. Meyers, The conversion of the corn/soybean
ecosystem to no-till agriculture may result in a carbon sink. Global Change Biology, 2005.
11(11): p. 1867-1872.
38.
Pyter, R., et al., Growing giant miscanthus in Illinois. University of Illinois 2007.
http://miscanthus.illinois.edu/wp-content/uploads/growersguide.pdf. (Accessed, August 10,
2012)
39.
Kaliyan, N., et al., Roll Press Briquetting and Pelleting of Corn Stover and Switchgrass.
Transactions of the ASABE. American Society of Agricultural and Biological Engineers, 2009.
52(2).
40.
Wang, M.Q., Development and use of GREET 1.6 fuel-cycle model for transportation
fuels and vehicle technologies. 2001. (Accessed, August 10, 2012)
41.
Sokhansanj, S., A. Kumar, and A.F. Turhollow, Development and implementation of
integrated biomass supply analysis and logistics model (IBSAL). Biomass & Bioenergy, 2006.
30(10): p. 838-847.
42.
Jensen, K., et al., Farmer Willingness to Grow Switchgrass for Energy Production.
American Agricultural Economics Association (New Name 2008: Agricultural and Applied
Economics Association), 2006. (Accessed, August 10, 2012)
43.
Fronning, B.E., K. Thelen, and D.H. Min, Use of Manure, Compost, and Cover Crops to
Supplant Crop Residue Carbon in Corn Stover Removed Cropping Systems. Agronomy Journal,
2008. 100(6): p. 1703-1710.

62

44.
Dale, B.E., et al., Biofuels Done Right: Land Efficient Animal Feeds Enable Large
Environmental and Energy Benefits. Environmental Science & Technology, 2010. 44(22): p.
8385-8389.

63

The following chapter is a reformatted version (to fulfill the dissertation document
requirements) of the paper: Advanced Regional Biomass Processing Depots: a key to the
logistical challenges of the cellulosic biofuel industry. Eranki, P.L., Bals, B. D., Dale, B.E.,
Biofuels, Bioproducts and Biorefining, 2011, 5(6):621-630.

64

CHAPTER 3: ADVANCED REGIONAL BIOMASS PROCESSING DEPOTS - A KEY
TO THE LOGISTICAL CHALLENGES OF THE CELLULOSIC BIOFUEL INDUSTRY

3.1. Introduction
3.1.1. Current biomass supply system: Problems and some prospective solutions
Biofuels are a near-term opportunity for the United States and the rest of the world to
reduce or eliminate our reliance on petroleum for transportation. Currently, grain ethanol and
oilseed biodiesel represent only a small fraction of United States transportation use, but the
addition of cellulosic biofuels could greatly increase this fraction. The Energy Independence and
Security Act of 2007, for example, mandates the production of 16 billion gallons of cellulosic
ethanol by 2022. A shift from petroleum-based fuels to biofuels would be a tremendous
undertaking, requiring hundreds of billions in investment and large shifts in land use. However,
such a shift would produce substantial benefits in terms of economic stability, environmental
sustainability, and national security.
Unfortunately, a paradigm shift to cellulosic biofuels is unlikely to occur unless several
challenges can be met, including reconciling biofuels with food production, avoiding harmful
land use changes, and assembling the supply chains. Some desirable characteristics of both
feedstocks and feedstock supply chains are listed in Table 3.1.
In particular, the local economic and environmental sustainability of the supply system is
vital in order to advance the farmers’ interests. Appropriately priced, low supply risk feedstocks
are needed to insure the growth of the nascent biofuel industry, as is the ability to inexpensively
transport and store the biomass. Local economic and environmental concerns must be resolved in
the feedstock supply chain, and diversifying the market for cellulosic feedstocks via valuable coproducts would be advantageous as well. Unfortunately, not all of these properties are currently
65

demonstrated by cellulosic feedstocks. A major purpose of this chapter is to demonstrate how
low cost processing at the regional level can help cellulosic feedstocks more fully achieve these
characteristics.
Table 3.1: List of desirable properties for both cellulosic feedstocks and supply
chains
Desirable Cellulosic Feedstock Properties

Desirable Supply Chain Properties

Low cost

Low transportation cost

Price stability

Multiple markets available

Consistent composition

Uniform, consistent (commodity) feedstock

Easily stored

Provides local economic opportunities

Dense or easily densified

Satisfies local and global environmental

Not competitive with food crops

criteria

Potential for co-product generation

The current paradigm for cellulosic biofuel production envisions large biorefineries
(approximately 2000-5000 dry tons of feedstock per day) which are fully integrated, centralized
facilities containing all biofuel conversion unit operations in a single location. Such facilities
may be impractical from a logistical standpoint, as seen in Table 3.2. For example, a large scale
biorefinery would require that one truck filled with biomass arrive every three minutes (for a full
24 hours per day), contract with thousands of farmers, and raise hundreds of millions of dollars
to begin operations. The difficulties with this system are exacerbated by a short harvest season,
requiring herbaceous biomass to be stored for months to insure a year-round biofuel production.
In contrast, a much smaller facility might require relatively little investment and would be a
simpler, more easily manageable operation. While there is substantial interest and investment to
66

improve conversion technologies, supply chain challenges, which are further exacerbated by the
low bulk density of feedstock, its compositional variability, and tendency to decompose, are
largely unaddressed and unresolved.
Table 3.2: Comparison of logistics for a typical 5000 ton/day corn stover biorefinery
vs. a 100 ton/day regional biomass processing depot (RBPD) in Iowa
5000 t/d refinery

100 t/d RBPD

Collection radius

48.2 miles

6.8 miles

Frequency of trucks

1 truck/3 minutes

10 trucks per day

Farmers to contract with

3900

78

380 ha

7.6 ha

$347 million

$3 million

Storage footprint
Capital cost

b

a

The RBPD calculation was performed using US Census of Agriculture to assume average farm
size of 276 acres, 38% of total land is in corn, and 4.6 tons/acre of corn stover are produced.
Also assumes 35% corn stover removal and 60% of farmers participating.
2
a. Assumes 180 tons require 195 m storage space for bales, and area open for access and
machinery storage is equal to total storage space. A one year supply of corn stover must
be stored in each location.
b. Cost of refinery determined from Laser et al. [1]. Cost of RBPD is an internal estimation.

One approach to help resolve these problems is to create satellite storage locations,
satellite storage facilities, or satellite depots. These facilities are currently envisioned as
transitory storage locations for square or round bales typically using standard farm equipment [2,
3] and sometimes including moderate processing, usually densification of biomass, before
delivering biomass to the biorefinery [4]. As mentioned in Chapter 1, such depots are currently
being used by the Tennessee Biomass Supply Co-operative (part of the University of Tennessee
Biofuels Initiative) where these facilities perform functions such as aggregation, storage, preprocessing (size-reduction) and intermediate processing (densification) of switchgrass and corn
67

cobs [5]. The preprocessed biomass is then sent to a cellulosic ethanol biorefinery operated by
DuPont-Danisco Cellulosic Ethanol (DDCE) which can produce 250,000 gallons per year of
ethanol. The Idaho National Laboratory (INL) bioenergy program is also focusing on
developing an “advanced uniform format feedstock supply system” that attempts to imitate the
current grain commodity supply system. The INL design proposes to locate specialized depots in
regions to collect biomass with similar characteristics which is subsequently sent to and blended
at a common shipping terminal to create a homogenized, consistent and stable commodity [6].
3.2. Regional Biomass Processing Depots
3.2.1. Decoupling pretreatment from biofuel production
While distributed storage and intermediate facilities in the biomass supply chain have
been somewhat explored, this concept is explored further in this chapter by using distributed
biomass pre-processing and pretreatment to bridge the gap between feedstock suppliers and
biorefineries. These Regional Biomass Processing Depots (RBPDs) are in essence isolated
preprocessing and pretreatment centers which, in their simplest configuration, produce pretreated
and densified biomass. The biomass is then shipped directly to a local biorefinery or,
alternatively, transported to a shipping terminal and sold to the global market.
A major objective of the RBPD network is to process and pretreat low density and often
unstable biomass into stable, dense intermediate products compatible with current established
commodity logistics systems, allowing the densified biomass to be transported economically
over much longer distances. Various densification methods are available, such as pelletizing,
briquetting, or cubing, which all have different capital and energy requirements. In general, a
high bulk density (pelletization) requires higher energy costs, more unit operations, and more
finely ground material than low bulk density material (cubing) [7]. Because biomass is naturally
68

heterogeneous, different harvests of a particular species may have different physical properties
that can impact biofuel production. Grinding, densifying, mixing, and storing the biomass
produces a consistent product that can be supplied to biorefineries or other markets while
standardizing the supply system schedule and logistics. RBPDs should provide relatively low
capital cost processing to densify and stabilize heterogeneous feedstocks into more uniform,
useful commodities. RBPDs would interact with local farmers and contract with them
individually within certain regions, perhaps using the well-established mechanism of farmer
cooperatives (“co-ops”). This co-op ownership model, in which the owners are the biomass
suppliers themselves, including perhaps the local community, would help bolster rural job
creation and income.
The RBPD concept can also be expanded to advanced configurations in which they
provide intermediates and products beyond those required for biochemical and thermochemical
biofuel production, such as higher value animal feeds, nutraceuticals, biocomposite materials,
etc., thereby leveraging the capital and expertise of these well-established industries. Adopting
RBPD systems that generate several products can potentially lead to greater per acre productivity
and diminish concerns about direct and indirect land use change. Alfalfa, for example, can
produce more protein per acre than soybeans, and extracted protein concentrates are similar in
feed quality to soybean meal. Thus, alfalfa can conceptually replace soybean land, generating
similar levels of protein while simultaneously producing fiber for biofuel production [8]. These
advanced configurations could allocate biomass resources to their optimal final products. For
example, in the “Biofuels Done Right” study [9] enhanced forage feeds for ruminant animals and
leaf protein concentrate (LPC, an animal feed protein) production were combined with
aggressive double cropping (planting a winter cover crop after corn or soybean harvest that can

69

be harvested prior to the next year’s planting) to maximize productivity on United States
farmland. In this study, U.S. farmland was sufficient to displace 50% of current gasoline use
while maintaining current food production, reducing U.S. annual carbon emissions by 10%,
increasing soil fertility, and promoting biodiversity [9]. An RBPD could effectively produce the
animal feed co-products while providing incentives to integrate unconventional cropping systems
such as double cropping and dedicated energy crops into traditional corn/soybean production
systems.
Figure 3.1 illustrates the advantages of the RBPD supply chain compared to more
conventional supply chains. This figure shows three potential landscapes (A, B, and C) and the
difference between a centralized refinery (bottom) and an RBPD network (top). By densifying
the biomass close to the farm, the practical collection radius for bulky biomass decreases,
decreasing transportation costs. This allows a refinery to collect biomass from a larger radius
(landscape A), increasing the size of the refinery and thereby improving its ability to exploit
economies of scale. Likewise, the ability to densify biomass near the feedstock source and to
transport this densified biomass cheaply over long distances may allow biorefineries to be
constructed in landscapes with limited biomass productivity (landscape B), where biomass
transportation costs to a centralized biorefinery would otherwise be prohibitive. Finally,
landscapes will likely contain a mixture of different feedstocks that would optimally be used for
different products (landscape C). With a centralized system, it may not be economically feasible
to allocate resources to different outputs, but instead allocate all biomass to a single product.
However, flexible RBPDs might be able to upgrade the biomass into different intermediate
commodity-like products, thereby allowing the optimal use of different feedstocks.

70

Figure 3.1: Comparison of potential RBPD systems (top) vs. no RBPDs (bottom). Advantages
of RBPDs include (A) smaller collection radius for individual RBPDs combined with a larger
total collection radius for the refinery, (B) individual RBPDs collecting enough biomass in
marginal cropland areas in locations where a single biorefinery may not be profitable, and (C)
allowing flexibility in feedstock allocation to multiple final locations, such as woody materials
(brown) to a co-fired power plant while grass (green) is sent to a refinery.
3.2.2. Potential pretreatments for RBPDs
While traditional approaches to intermediate supply facilities may include densification, a
key element of RBPDs as envisioned in this chapter is to include a chemical pretreatment as
well. Pretreatment significantly alters the morphological structure of biomass while increasing
the susceptibility of carbohydrate polymers to enzymatic attack. These changes can also be
synergistic with the primary purposes of the RBPD, namely densifying and storing the biomass
while producing valuable co-products. Ideally, a pretreatment operation at an RBPD would have
71

a low capital cost, simple catalyst recovery, create few if any harmful degradation products, and
produce a stable intermediate that could be upgraded to multiple products. Some potential
pretreatments for herbaceous and woody materials include dilute acid, hot water [10], ammonia
recycle percolation (ARP) [11], steam explosion, lime [12] and SPORL pretreatment [13].
Of particular interest is ammonia fiber expansion (AFEX™) pretreatment [14, 15]. In
the AFEX™ process, hot concentrated anhydrous ammonia (temperatures of 70-200°C) is
mixed with biomass under pressure (20-30 bar) for residence times of 15-30 minutes. Both
ammonia and water loading range from 0.4-2.0 g/g dry feedstock. The pressure is then rapidly
released causing the system to cool and the ammonia is recovered. During this reaction,
recalcitrant lignin is depolymerized and some is relocated to the surface of the biomass,
hemicellulose is hydrolyzed and cellulose is partially decrystallized, thereby making the sugars
in the biomass more accessible to enzymatic breakdown. This reaction is performed under high
solids loading (30-75% solids on a total weight basis) and produces no separate liquid phase
while causing very little biomass degradation due to moderate conditions of temperature and pH.
AFEX™ pretreatment has several features that may make it almost uniquely suitable for
RBPDs [16]. Since it is a relatively “dry” pretreatment, the resulting substrate is inert and stable.
Some lignin, a natural adhesive, is removed to the surface of the biomass during AFEX™ and
improves the binding properties of the biomass. This increases the ease of densification,
eliminating the need to purchase binders or operate at high temperatures/pressures, thereby
significantly reducing densification costs while simultaneously improving the transportation
characteristics of pretreated biomass. AFEX™ can also add value to pretreated biomass as an

72

animal feed by increasing the digestibility of fiber while simultaneously adding non-protein
nitrogen in the form of acetamide. AFEX™ treated corn stover and late-harvest switchgrass
showed improved rumen digestibility by 52% and 128%, respectively over untreated material
[17]. Also, preliminary analysis suggests that a very large fraction of both beef cattle and dairy
cattle nutritional requirements can be met with AFEX™ treated animal feed depending on the
age of the cattle [18]. Thus, AFEX™-treated feedstocks might be viable alternatives to
traditional forages and could help convert diverse biomass feedstocks into more uniform, salable
commodities. The AFEX™-treated biomass could also be used for thermochemical conversion
as a third potential market, although the additional cost might be problematic.
3.3. Discussion
3.3.1. Overcoming the challenges of an RBPD network
Despite the potential advantages of an RBPD network, several challenges to acceptance
and eventual commercialization remain. These include compounded dry matter losses due to
multiple storage locations and transport, potentially high energy costs of operations such as size
reduction and densification, possible under-usage of equipment in each depot leading to capital
cost intensification, and low total capacity leading to disadvantageous economies of scale.
Shrinkage, defined as dry matter losses between harvest and end use of the raw material,
is a detriment to biomass refining and should be minimized. Shrinkage can occur through
damaged bales, microbial contamination during wet storage conditions, or losses in preprocessing and transport. A stable, dry storage facility for bales can reduce losses due to
moisture effects, but grinding biomass for densification can also cause shrinkage. Jannasch et al.

73

[19] reported losses up to 57% of dry matter during a test of grinding equipment, but noted that a
system with the ability to recycle fines would expect only 5-10% losses. When combined with
shrinkage from storage, a total of 10-15% of dry matter could be lost during the process,
representing a significant loss of revenue.
In addition, grinding and densification can be costly steps in the process, both in terms of
economics and energy requirements [20]. The cost of densification can be up to $25/Mg
biomass, a cost which likely eliminates any savings in transportation costs. Such costs would
probably not occur in an RBPD using AFEX™ due to the synergism between AFEX™
pretreatment and densification. Evidence suggests that particle size reduction is not beneficial to
the effectiveness of AFEX™ treatment, and thus small particles are not required for
downstream processing [21]. Furthermore, large particle sizes (>5 cm) are important for animal
feed purposes, as large fibers are more effective at providing a rumen mat. Cattle feeding guides
suggest that 75% of the fiber in the diet should be long pieces rather than ground particles [22].
Because of the improved binding characteristics of AFEX™ treated biomass, these larger
particles can be easily compacted. Federal Machine (Fargo, ND) has successfully densified
switchgrass and corn stover of up to 8 mm particle size into briquettes (called PAKs) with a bulk
density of 330 kg/m [4]. Preliminary estimates suggest this process can be performed for
approximately $5-10/Mg biomass (D Marshall, 2008, Personal Communication).
The added cost of densification is acceptable if it can abate both storage and
transportation costs. Densifying biomass allows for more weight per truck, thereby decreasing
the number of trucks required. Alternatively, RBPDs can be placed along rail lines, replacing
road transportation from the RBPD to the refinery with cheaper rail transportation. Furthermore,

74

storage of pellets or briquettes is simpler than storing bales. Vertical structures similar to grain
storage tanks or silos can be used instead of horizontal bale storage options. A single storage
2

tank (with a surface area of ~600 m ) may hold over 5000 Mg of biomass briquettes, reducing
the total storage footprint for a full size refinery by a factor of 20 (see Table 3.2), assuming a
year’s supply of biomass in reserve.
In order for an RBPD to be feasible, the overall capital cost must be low and the
technologies used must be scalable to smaller capacities (e.g., 100-1000 Mg/day). AFEX™
pretreatment is generally considered to be cost competitive compared to other pretreatments [23]
and the core processes are readily scalable. Much of the cost in AFEX™ pretreatment is due to
the ammonia recycle system, and efforts are currently underway to design and validate ammonia
recycling operations suitable for an RBPD setting.
Finally, the seasonality of feedstocks is another challenge involved in RBPD systems.
Biomass harvesting is performed during a short window, forcing processing to be performed for
only a few months out of the year while leaving the equipment idle for the rest of the year. For
example, in an economic study of leaf protein production, the length of the processing season
was a primary driver of profitability [24]. By creating a flexible RBPD network that can
incorporate numerous feedstocks and processing technologies, the seasonality issue can perhaps
be abated. Capital and overhead costs can be spread among multiple processes, and processing
different feedstocks expands the available growing season. For example, double crops could be
harvested and processed in the spring and summer, corn stover in autumn, switchgrasss in late
autum and miscanthus in late winter.

75

By minimizing the challenges and potential costs associated with RBPDs, the added
processes, energy costs, and complexity required by pretreatments might provide benefits
beyond improved logistics. If densification can be performed cheaply and with low energy
requirements, the transportation costs can be greatly reduced. As seen in the Chapter 2, a
comparative life cycle assessment of an RBPD network (employing pretreatment and
densification in all depots located away from biorefinery) vs. a centralized biorefinery showed
that the RBPD network yields practically the same net energy while generating significantly
lower greenhouse gas emissions compared to the centralized biorefinery in which all processing
operations occur in one location [25]
3.3.2. Enhanced Depots: Multiple technologies
To further enhance the functionality and efficiency of the depots, the concept of “tailormade” or enhanced depots is proposed: facilities that employ specific technologies that depend
primarily on regional feedstock availability and biomass characteristics as well on synergies
among these technologies. Figure 3.2 illustrates feedstocks and technologies that might be
included in a depot and the various products and co-products that might be generated. It is
unlikely that all depots will contain all technologies. The choice of appropriate technologies
should be based on detailed system wide analyses starting at the landscape producing the
feedstocks and going on through the exit of the biorefinery producing liquid biofuels and
electricity.
The enhanced depots would likely process between 100 to 1000 dry tons per day of
feedstock and should require a fairly low capital investment so that local ownership is facilitated.
Feedstocks include agricultural residues (such as corn stover), grain/fiber mixes (such as corn
silage), dedicated energy crops (mixed prairie grasses, switchgrass, and miscanthus), forages
76

(alfalfa), and woody material (forestry waste, poplar). Utilizing many feedstocks can reduce risks
for the depot network. Just as product diversification is important for the economic viability of
the biorefinery, so is feedstock diversity in the upstream logistics systems.

Forests and Woody
Crops

Pyrolysis

AFEX™

Cropland

Co-fired
Power Plant

Char
BioOil
Wood
Pellets
Densification

Biorefinery
Anaerobic
Digestion

Leaves
Protein
Extraction

AFEX
BM

Stems

Leaf/Stem
Separation
Fiber

Protein

Oil Refinery

Manure
Animal
Operations

Figure 3.2: Process Flow Diagram of an RBPD with all technologies. Green squares represent
the landscape, blue squares represent operations within the RBPD, gray squares represent
products/intermediates, and red squares represent outside operations. Red arrows are energy
flows, green arrows are feedstock flows, purple arrows are material flows within the refinery,
and blue arrows are final products.
Two important technologies - AFEX™ pretreatment and densification - were discussed
in the previous sections in this chapter. Other relatively low capital cost technologies that might
be included in the depots are as follows:

77

3.3.2.1. Leaf Protein Concentrate (LPC)

Forages such as alfalfa have the potential to produce more protein per hectare than
soybeans, the primary source of protein meal in the United States. Unfortunately, the high fiber
content of these forages make them unsuitable to swine and poultry diets unless the protein is
separated to produce LPC. The industrial processes to manufacture LPC have been studied for
decades but have not been widely adopted due to a short operating season and low value of the
remaining fiber. By integrating this LPC process with other operations at an RBPD, both of these
limitations might be significantly reduced. Alfalfa has been the usual feedstock for LPC
technology, but the process can be applied to any high protein cellulosic biomass such as double
crops and early season perennial grasses or grass/legume mixtures. The process involves pulping
and pressing wet biomass to squeeze out a protein rich juice which is subsequently coagulated
and dried to obtain a high protein powder [24]. The LPC extract is estimated to produce a profit
of $27/Mg feedstock [26], potentially providing a valuable co-product to fiber processing.

3.3.2.2. Leaf/stem separation

This technology includes using simple gravity separation methods such as air-separation
of heavier fractions (stem portions) from the lighter fractions (leaves). Techniques similar to
“green leaf threshing” (GLT) in the tobacco industry will most likely be adopted [27]. The
processed raw-material conditions in GLT are similar to those of feedstocks arriving from the
farms to the depots, thereby making this technique potentially useful for leaf stem separation.
Leaf stem separation can precede LPC extraction while the stems can be sent directly to
pretreatment or other processing. Alternatively, the leaves can be fed directly to nonruminants,
given that the leaves of forages tend to have 2-3 times greater protein content and 50% less fiber
78

than the forage as a whole [28]. While the protein content remains low (28%), this could be a
low capital cost alternative to LPC production. Feedstocks that might undergo this process
include high protein forages such as alfalfa mixes, native prairie grasses and double crops.

3.3.3.3. Pyrolysis

This thermochemical conversion technique wherein biomass is broken down under
anaerobic conditions can be applied to woody biomass and other highly recalcitrant feedstocks.
Fast pyrolysis involves moderate temperatures (in the range of 500°C) and short residence times
(in the order of seconds), producing 60-75% bio-oil, 12-20% biochar and 13-20% producer gas.
Low process temperatures and long residence times (slow-pyrolysis) direct the reaction toward
the production of biochar, whereas on the other extreme high temperatures coupled with long
residence (gasification) times push the product distribution toward producer gas generation [29].
Producer gas can be used for process heat in the RBPD. Biochar (with its associated mineral
content) can be sent back to farms as a soil-amendment, improving soil quality and decreasing
overall carbon emissions. Field research and historical observations show that applying biochar
to soil augments plant growth and reduces water runoff, soil erosion, and gaseous emissions from
soil. Biochar can enhance the delivery of nutrients to crops and also improve physical and
organic properties of soil. Even with high-input agriculture, application of biochar has shown to
enhance fertilizer efficiency [30]. These features might make biochar an important co-product in
the RBPD network. Biochar is a very stable compound with a high intrinsic energy value, and
thus could also be combusted in a power plant or at the biorefinery to provide heat and power.
The liquid component of fast-pyrolysis, bio-oil, is a relatively unstable, viscous, and
corrosive component that does not exist in thermodynamic equilibrium at storage temperatures.

79

It has high water and oxygen content leading to a low energy density. Bio-oil must be upgraded
or “reformed” before it can be used as a blended additive in petroleum products or processed into
valuable chemicals. Various hydro-deoxygenation, catalytic, emulsification and steam reforming
processes can be applied to improve the properties of bio-oil [31]. The petroleum industry
already possesses the expertise and equipment required to perform the upgrading process.
Therefore bio-oil will in all likelihood be shipped to a petroleum refinery. A central question that
might arise in the implementation of the pyrolysis technology in the RBPDs is whether it would
be more favorable to optimize pyrolysis for bio-char production (slow-pyrolysis) vs. optimizing
it for the bio-oil (fast pyrolysis). Furthermore, bio-oil would probably need to be stabilized at the
facility, likely through the addition of hydrogen. Regardless, including pyrolysis in the depot
network will probably benefit the system by helping to satisfy RBPD energy requirements, by
providing valuable co-products, and by increasing feedstock flexibility.

3.3.3.4. Anaerobic digestion

Since the RBPD network is already coupled with agricultural and animal feeding
operations, it is logical to maximize the raw materials available as feedstocks to the RBPDs and
also to produce profitable co-products whenever possible. While RBPDs can send animal feed
and biochar to the farms, they can also obtain manure from animal operations along with
cellulosic biomass feedstocks. In an anaerobic digester, the manure is liquefied by bacteria and
subsequently undergoes acidogenesis and methanogenesis to produce biogas which can be sold
or used on-site at the RBPDs for heat [32]. The end product can be relatively easily separated
after anaerobic digestion into solid and liquid fractions. The solid fraction may be used as
recycled bedding for animals or applied to farm-lands to enhance soil properties. Since the

80

animal feed is generated internally in the RBPD network, using manure in anaerobic digestion
from the same animals tends to create a “closed-loop”, more sustainable system of industrial
ecology in which wastes from one sub-system are used as raw-material for others.
3.4. Designing and implementing RBPDs
The technologies mentioned above are not meant to represent an exhaustive list. New
technologies that are easily scaled to 100-1000 tons/day and are relatively low capital cost can
also be included as these are developed. This includes further fractionating the biomass for
higher value products (biochemicals, nutraceuticals, etc) as they become commercially viable.
Likewise, the potential synergies and tradeoffs among all technologies must be determined in
order to successfully establish these RBPDs. For example, lignin is relocated to the surface of the
biomass during AFEX™ pretreatment, and a low cost lignin extraction might be performed to
generate higher value products. If the lignin is removed, however, then it cannot be used as a
binder during densification, thus requiring a redesign of the densification process. Alternatively,
the anaerobic digester can also take in waste streams (for example, excess water from the leaf
pulping process) as well as excess biomass that may not be suitable for other technologies to
increase biogas production. Heat integration between technologies should reduce the heat
demand for the overall process, and excess energy from pyrolysis and anaerobic digestion might
also support the remaining technologies.
It is necessary to design these RBPDs at the correct scale and suite of technologies to
adapt to the landscape surrounding it. Not all depots may operate at the same scale, nor will they
all contain the same technologies. For example, if a certain region has a concentration of animal
operations along with substantial land in forage, then a depot situated in this area might consist
of LPC extraction, AFEX™ pretreatment, and densification, but not use fast pyrolysis. On the
81

other hand, in regions containing only forest residues and marginal lands (where it is possible to
grow low-maintenance perennial grasses), a depot consisting only of pretreatment, fast pyrolysis,
and densification might be preferred. Proper locations for these RBPDs also must be considered.
The sites would preferentially be located in or near rural communities, along a rail line if
possible and potentially close to other industries to share heat needs. If a large animal feed
operation is already present in the community, it may be advantageous to co-locate the RBPD
there to share resources from manure or the sale of animal feed.
Implementing an RBPD network can also help solve the "chicken and egg" problem of
establishing biomass production systems for a dedicated biorefinery. Early RBPDs can be
focused primarily on providing animal feed or biomass for power production in a co-fired coal
plant. Output would be scaled up as biorefineries are built. Because the biomass is densified, it
can be transported long distances at relatively low cost. Thus, biorefineries will not necessarily
need to secure all of their feedstock from the local area, and RBPDs do not necessarily need to
be built around planned biorefineries. If so, this would allow the RBPD network to grow
organically as the demand for lignocellulosic biomass increases. Alternatively, the RBPDs could
be developed by the biorefinery owners as part of their supply chain. Such an approach would
likely reduce the input from farmers and possibly reduce the economic value to the local
community, it would ensure that RBPDs are built in the surrounding area and improve the supply
chain for specific refineries. While this approach may be preferable in some instances, it would
probably be more difficult to grow a commodity supply chain of pretreated lignocellulosic
biomass from RBPDs controlled by the biofuel producers.

82

3.5. Conclusion
As cellulosic biofuel technologies improve, there will be increasing emphasis on
commercial deployment of these technologies. Logistical and socio-economic challenges will
then become increasingly important. Regional Biomass Processing Depots can help bridge the
gap between rural concerns (rural economics, food security, environmental quality) and
biorefinery supply concerns (steady supply, uniform feedstock properties, stable feedstock costs,
and low transportation costs). By producing valuable co-products such as animal feed close to
the farm level, the RBPDs can help satisfy local interests while simultaneously increasing the
value and utility of cellulosic biomass feedstocks. Furthermore, RBPDs can also homogenize
feedstocks and simplify the supply chain while reducing the overall risk to the biorefinery and
spreading the capital costs of biofuel production over a greater number of interested participants.
Developing flexible RBPDs into viable industries will help society achieve greater economic and
environmental benefits from the nascent cellulosic biofuel industry.

83

REFERENCES

84

3.6. REFERENCES

1.
Laser, M., et al., Coproduction of ethanol and power from switchgrass. Biofuels,
Bioproducts and Biorefining, 2009. 3(2): p. 195-218.
2.
Cundiff, J., R.D. Grisso, and J. Judd, Operations at satellite Storage Locations (SSL) to
deliver round bales to a biorefinery plant. 2009 ASABE Annual International Meeting, 2009.
3.
Judd, J., et al., An Optimal Storage and Transportation System for a Cellulosic Ethanol
Bio-energy Plant. Paper no: 1009413. ASABE Annual International Meeting, 2010. (Accessed,
August 10, 2012)
4.
Sokhansanj, S. and J. Fenton, Cost benefits of biomass supply and pre-processing. A
BIOCAP Research Integration Program synthesis paper, BIOCAP Canada, 2006. (Accessed,
August 10, 2012)
5.
Tiller, K., Demonstration of a Large Scale Integrated Switchgrass Supply Chain
Presented at Biomass Conference, March 30, 2010. Washington DC, USA. (Accessed, August
10, 2012)
6.
Hess, J., et al., Uniform-Format Solid Feedstock Supply System: Commodity-Scale
Production of an Infrastructure-Compatible Bulk Solid from Herbaceous Lignocellulosic
Biomass. Report INL/EXT-09-15423, 2009. Idaho National Laboratory, Idaho Falls, ID.
(Accessed, August 10, 2012)
7.
Sokhansanj, S. and A.F. Turhollow, Biomass densification - Cubing operations and costs
for corn stover. Applied Engineering in Agriculture, 2004. 20(4): p. 495-499.
8.
Dale, B.E., et al., Protein feeds coproduction in biomass conversion to fuels and
chemicals. Biofuels, Bioproducts and Biorefining, 2009. 3(2): p. 219-230.
9.
Dale, B.E., et al., Biofuels Done Right: Land Efficient Animal Feeds Enable Large
Environmental and Energy Benefits. Environmental Science & Technology, 2010. 44(22): p.
8385-8389.

85

10.
Kim, Y., et al., Liquid Hot Water Pretreatment of Cellulosic Biomass Biofuels, J.M.
Walker, Editor. 2010, Humana Press. p. 93-102.
11.
Kim, T.H., et al., Pretreatment of corn stover by low-liquid ammonia recycle percolation
process. Applied Biochemistry and Biotechnology, 2006. 133(1): p. 41-57.
12.
Xu, J.L., et al., Lime pretreatment of switchgrass at mild temperatures for ethanol
production. Bioresource Technology. 101(8): p. 2900-2903.
13.
Wang, G.S., et al., Sulfite pretreatment to overcome recalcitrance of lignocellulose
(SPORL) for robust enzymatic saccharification of hardwoods. Biotechnology Progress, 2009.
25(4): p. 1086-1093.
14.
Balan, V., et al., Lignocellulosic Biomass Pretreatment Using AFEX Biofuels, J.M.
Walker, Editor. 2010, Humana Press. p. 61-77.
15.
Chundawat, S.P.S., et al., Multi-scale visualization and characterization of
lignocellulosic plant cell wall deconstruction during thermochemical pretreatment. Energy &
Environmental Science. 4(3): p. 973-984.
16.
Carolan, J.E., et al., Technical and Financial Feasibility Analysis of Distributed
Bioprocessing Using Regional Biomass Pre-Processing Centers. Journal of Agricultural & Food
Industrial Organization, 2007. 5(10): p. 1-29.
17.
Bals, B., et al., Ammonia fiber expansion (AFEX) treatment of eleven different forages:
Improvements to fiber digestibility in vitro. Animal Feed Science and Technology, 2010. 155: p.
147-155.
18.
Sendich, E. and B.E. Dale, Environmental and economic analysis of the fully integrated
biorefinery. Global Change Biology Bioenergy, 2009. 1(1): p. 331-345.
19.
Jannasch, R., Y. Quan, and R. Samson, A process and energy analysis of pelletizing
switchgrass, 2011.
http://reapcanada.com/online_library/feedstock_biomass/11%20A%20Process.pdf.
(Accessed, August 10, 2012)

86

20.
Sokhansanj, S., et al., Large-scale production, harvest and logistics of switchgrass
(Panicum virgatum L.) - current technology and envisioning a mature technology. Biofuels,
Bioproducts and Biorefining, 2009. 3(2): p. 124-141.
21.
Chundawat, S.P.S., B. Venkatesh, and B.E. Dale, Effect of particle size based separation
of milled corn stover on AFEX pretreatment and enzymatic digestibility. Biotechnology and
Bioengineering, 2007. 96(2): p. 219-231.
22.
Ensminger, M.E. and C.G. Olentine, Jr, Feed Allowance, in Feeds & Nutrition-complete.
1978, The Ensminger Publishing Company: Clovis. p. 956-961.
23.
Eggeman, T. and R.T. Elander, Process and economic analysis of pretreatment
technologies. Bioresource Technology, 2005. 96(18): p. 2019-2025.
24.
Enochian, R.V., et al., Producing Pro-Xan (Leaf Protein Concentrate) from Alfalfa:
Economies of an emerging technology. Report no. 445, 1980. USDA Agricultural Economics,
Washington DC, USA
25.
Eranki, P.L. and B.E. Dale, Comparative life cycle assessment of centralized and
distributed biomass processing systems combined with mixed feedstock landscapes. GCB
Bioenergy. 3(6): p. 427-438.
26.
Bals, B. and B.E. Dale, Economic comparison of multiple techniques for recovering leaf
protein in biomass processing. Biotechnology and Bioengineering, 2010. 108(3): p. 530-537.
27.
Manickavasagan, A., J.J. Gunasekaran, and P. Doraisamy, Trends in Indian Flue Cured
Virginia Tobacco (Nicotiana tobaccum) Processing: II. Threshing, Packing and Warehousing.
Research Journal of Agriculture and Biological Sciences, 2007. 3(6).
28.
Dale, B.E., Biomass refining- protein and ethanol from alfalfa. Industrial & Engineering
Chemistry Product Research and Development, 1983. 22(3): p. 466-472.
29.
Bridgwater, T., Biomass for energy. Journal of the Science of Food and Agriculture,
2006. 86(12): p. 1755-1768.
30.
Gaunt, J.L. and J. Lehmann, Energy Balance and Emissions Associated with Biochar
Sequestration and Pyrolysis Bioenergy Production. Environmental Science & Technology, 2008.
42(11): p. 4152-4158.

87

31.
Zhang, Q., et al., Review of biomass pyrolysis oil properties and upgrading research
Energy conservation and management, 2007. 48(1): p. 87-92.
32.
Lazarus, W.F. and M. Rudstrom, The economics of anaerobic digester operation on a
Minnesota dairy farm. Review of Agricultural economics, 2007. 29(2): p. 349-364.

88

CHAPTER 4: LIFE CYCLE ASSESSMENT OF WATERSHED-SCALE CELLULOSIC
FEEDSTOCK LANDSCAPES INTEGRATED WITH DISTRIBUTED PROCESSING
4.1. Introduction
The use of petroleum-derived fuels has created much wealth and societal progress over
the past hundred plus years since these fuels became widely available. However, in recent
decades petroleum use is increasingly problematic is terms of price and price volatility, national
security and environmental issues. In order to mitigate these problems and diversify
transportation fuels, alternatives to petroleum are being sought worldwide. Liquid fuel
production from lignocellulosic feedstocks, an abundant and renewable resource, is one of the
most attractive alternatives to continuing our near exclusive reliance on petroleum fuels for
transportation.
However, the dominant existing conceptual model for cellulosic biofuel envisions large
biorefineries processing thousands of tons of biomass a day. These fully integrated, centralized
facilities contain all biofuel conversion unit operations in a single location. This model has
severe logistical constraints as discussed in the previous chapters [1, 2]. As a possible solution to
these logistical challenges, an alternative model is proposed in the form of advanced Local
Biomass Processing Depots [LBPDs] (previously referred to as “Regional Biomass Processing
Depots [2, 3]). These depots are decoupled pretreatment (for biochemical conversion of biomass
to ethanol), storage and formatting facilities containing other potentially advantageous
technologies to produce valuable end-products and co-products. A network of depots supplying
large biorefineries will have much smaller biomass collection radii, simplified logistical
arrangements, smaller biomass storage footprints and more tractable business arrangements (e.g.
contracting with farmers) compared to fully integrated, centralized biorefineries. For example, a
single small scale depot (100 dry tons per day (TPD)) will reduce the collection radius for
89

biomass by a factor of 7 and will also reduce by a factor of 50 the number of farmers with whom
contracts must be developed [2]. Furthermore, such a network of local depots also lends itself to
local ownership and therefore increased potential for local economic benefits.
As discussed in Chapter 3, the “base-technologies” included in all depots are
pretreatment and densification, including all pre-processing steps such as biomass handling, sizereduction, grinding, etc. The use of other technologies in the depots is dictated by the
characteristics of the landscape generating the biomass. To summarize the potential technologies
that may be included in depots: The Ammonia Fiber Expansion (AFEX™) process is a
pretreatment method where hot concentrated ammonia is mixed with moistened herbaceous
biomass under pressure. After the desired treatment time is complete, the pressure is then rapidly
released causing the system to cool and the ammonia is recovered[4]. AFEX™ is used as the
pretreatment method in all processing facilities because of its apparently unique suitability within
the depot arrangement. AFEX™ adds value to biomass as a highly-digestible fiber-based animal
feed, which is economically the most important co-product from the depots, and also improves
the efficiency of the densification process following pretreatment [1, 2]. Mechanical compaction
or densification of biomass into pellets or briquettes post-pretreatment can increase the bulk
3

3

density of cellulosic feedstocks from as low as 60 kg/m to as high as 800 kg/m , significantly
reducing transport, handling and storage burdens associated with the depots [3, 5]. Coupling
densification with AFEX™ pretreatment may also reduce the cost of densification because
lignin brought to the surface during AFEX™ eliminates the need for added binders (e.g., starch
or protein) to promote binding [1, 5, 6]. The Leaf Protein Concentrate (LPC) extraction process

90

involves pulping and pressing wet biomass to squeeze out a protein rich juice which is
subsequently coagulated and dried to obtain a high protein powder [7, 8]. LPC can be applied to
any high protein cellulosic biomass such as double crops and forage[2] to obtain a high protein
animal feed supplement. Pyrolysis is the thermochemical conversion of woody residues into
biochar, producer gas and bio-oil. Producer gas can be combusted for process heat in the LBPD
and bio-char is an important co-product that aids in soil amendment in the landscape[9].
Alternately, biochar can be used as a boiler fuel to substitute for coal[10]. Bio-oil however is an
unstable compound that must be further processed for use as transport fuel[11]. In anaerobic
digestion (AD) biogas is produced from manure and aqueous waste streams resulting from LPC
production. Solids produced by AD are sent back to farms to use as animal bedding or as a soil
amendment [12].
Not all depots necessarily contain all the other technologies in addition to the basetechnologies of AFEX™ pretreatment and densification. The depots can be imagined as
“custom-made” facilities varying in capacity and operational characteristics as well as in the
feedstocks processed and co-products produced, as dictated by the characteristics of the
landscapes (e.g., largely agricultural landscapes vs. heavily forested ones).
In chapter 2, an LCA study was conducted to compare the current paradigm of a
centralized biorefinery with the new concept of distributed biomass processing via depots [3].
This comparative LCA study was based on a small farm-scale landscape analysis but in more
relevant scenarios, biomass for processing is procured from an aggregation of local farms.
Therefore, the present study is based on seven digit watershed-scale landscapes. A group of 9
counties (Allegan, Barry, Branch, Calhoun, Cass, Eaton, Kalamazoo, St.Joseph and Van Buren)
in South-west Michigan termed the “Regional Intensive Modeling Area” (RIMA-MI) was
91

identified as a region of interest for modeling by in the Great Lakes Bioenergy Research Center
[GLBRC [13]]. The total area of this RIMA is 1.38 million ha (2010 estimate) and it has a
population of 0.89 million people (2010 estimate) [14]. The total harvested farmland area of the
RIMA amounts to 0.54 million ha (2007 estimate) [15].The RIMA was divided into five 7-digit
hydrologic unit code (HUC) watersheds (Upper Grand, Black-Macatawa, Kalamzoo, St.Joseph
and Thornapple) encompassing all counties contained within the RIMA. Cellulosic feedstocks
i.e. corn stover, forage (hay), switchgrass, miscanthus, native prairie and woody residues (woody
residues are only available in limited areas) are assumed to be cultivated in this RIMA in four
scenarios.
4.2. Structure and methods
The previous comparative LCA established that simple depots when combined with high
yielding perennial grasses and densification return comparable or sometimes greater energy and
environmental benefits compared to centralized systems. To better understand and assist the
development of this novel logistical system, a comprehensive LCA is required. Hence, in this
study the LCA of advanced depot configurations is combined with watershed scale landscapes
(7-digit hydrologic unit code [HUC]).
4.2.1. Technologies in the processing system
For the distributed processing system module in this analysis, an in-house technical
model [16] was developed to determine the processing energy required based on technologies
included in each scenario. This model uses a combination of literature values and experimental
results to form a basic material and energy balance around each unit operation. These balances
determine the fossil fuel and raw material input required. Pyrolysis was included in only one
advanced depot scenario and only in areas with relatively large forest cover. All processing
92

energy inputs and emission outputs for reforming bio-oil into gasoline and diesel [17] and fossil
fuel emissions displaced by bio-oil were included in the system boundary. As mentioned
previously, bio-oil is an unstable compound that has to be reformed before it can be further
converted into fuels. Just as distributed processing systems are being evaluated for bioethanol
production so it is likely that supply chains be put in place for distributed bio-oil reforming [18]
and that these facilities may be located in the vicinity of the biorefinery.
In the AD process, manure is procured from the animal operations within the landscape.
Using manure as a raw material within the depots enhances system sustainability. Animal
operations are already coupled with the depot network since animal feed is a principal coproduct. One definition of an industrial ecology system states that it is: “a change from linear
(open) processes to cyclical (closed) processes, so the waste from one industry is used as an
input for another” [19]. Therefore, providing feed to and using wastes from animal operations
creates a “closed-loop” or industrial ecology system in these distributed biofuel networks,
potentially enhancing the overall system sustainability[2]. For the biorefinery module the
processing energy for operation and energy gains due to electricity and ethanol produced were
determined using the NREL/Dartmouth biorefinery model [20, 21].
4.2.2. Feedstocks
The RIMA-MI is principally agricultural with forested areas present in some portions of
the more northern watersheds. The cellulosic feedstocks included in the study are agricultural
residues (corn stover), perennial grasses (switchgrass and miscanthus), forage (alfalfa and grass
hay), native prairie biomass and forest residue The agricultural residues and perennial grasses
were chosen because of their relevance as primary feedstocks of interest in lignocellulosic
biofuel production. Although low input high diversity (LIHD) native prairies do not yield large
93

quantities of biomass, they are have attracted interest as potential lignocellulosic feedstocks
because of their positive environmental effects and ability to generate biomass on
abandoned/degraded land [22-24]. A recent study connected simplified landscapes with
heightened pest infestations and insecticide use and concluded that perennial bioenergy crops
such as switchgrass and mixed prairie can counterbalance these effects [25]. In a related study, it
was found that diverse landscapes can promote better bio-control services, an important and
beneficial ecological provision [26]. Hence the use of complex landscapes containing a mix of
cellulosic feedstocks may have positive environmental benefits in areas such as biodiversity,
habitat preservation and reduced chemical use. Non-cellulosic feedstocks (e.g. corn grain) are
excluded from this analysis which focuses on lignocellulosic ethanol production. All associated
burdens (crop energy input and environmental output) were split between corn and stover on a
mass basis. Only the burdens of stover harvested were included in the system boundary. For
example, for 40% stover harvest, only the energy inputs/emission outputs to generate this mass
of stover were included in the system boundary.
4.2.3. Landscape system
Given the emerging importance of water use for bioenergy production [27, 28] and the
centrality of watersheds in a plant based ecosystem; this analysis is based on watershed scales.
The analysis is also structured in this way for other reasons a) Data availability- accessibility of
data for land use, feedstock yields, transport network, etc. for real regions of interest from
watershed-scale models, b) Ease of modeling- the data from these watershed models can be
connected with other map manipulation and data analysis software to create a flexible analytical
framework, c) Model validation [29]- an analysis on the similar scale provides an opportunity for
comparison and data exchange between various ongoing biogeochemical studies involving the
94

same region of interest, d) Impact assessment- environmental impacts such as nitrogen leaching
and soil nitrogen losses and impact assessment. The U.S. EPA Better Assessment Science
Integrating Point and Non-point Sources (BASINS) model [30], a multi-function open-source
GIS based environmental assessment model was chosen as the tool to obtain data such as land
availability and for identifying watersheds within the RIMA. This model contains 7-digit HUC
watershed data required by our analysis, including as land-use categories and their respective
areas (National Land Cover database), road network lengths, etc. BASINS is also easily
adaptable to other geospatial reference software such as ArcGIS. Data for all five watersheds
were obtained from BASINS and mapping projections and operations were performed in
ArcGIS. Maps of different land uses in the watersheds thus obtained are shown in Appendix B.
A watershed weightage factor method was developed to identify only watershed regions
encompassing the nine counties in RIMA-MI and a geometric factor method (discussed in
Appendix B) was established to apportion land areas to individual depots. Yields for available
feedstocks were determined using National Agricultural Statistics Service (NASS) data (Table
B1 in Appendix B). Yields of currently unavailable feedstocks (Table B2 in Appendix B) and
energy inputs for all feedstock operations were determined from extensive literature review.
4.2.4. Life Cycle Assessment
The functional unit in this analysis is one hectare of the RIMA-MI land area. LBPDbiorefinery distances were determined by combining the road network map in the US EPA
Basins model with the Arc GIS software to calculate the shortest routes from each LBPD to the
central biorefinery. Using this method the processing depots and central refinery were located
within the RIMA-MI. Distances of transport (collection areas and radii) of raw materials
(herbaceous feedstocks, animal manure and forest slash where applicable) to LBPDs and co95

products (pretreated animal feed, LPC solids, AD sludge and biochar where applicable) from
LBPDs back to farms were determined using formulae from the literature (Appendix B).
Transport distances for finished products (pretreated biomass to biorefinery and bio-oil where
applicable) were determined using the US EPA BASINS and Arc GIS softwares. Using these
distances (as seen in Table 4.2) and the Greenhouse Gases, Regulated Emissions, and Energy
Use in Transportation (GREET) model [31] for both densified and non-densified transport where
applicable, transport emissions were calculated. It was assumed that animal farming operations
were located conjointly with feedstock cultivation operations and therefore the same distances
were used for transport of all relevant co-products back to these animal farms.
The amount of animal feed returned and number of animals in the system were
determined based on forage displaced from the original land area in each scenario. The amount
of pretreated animal feed sent back is determined based on its nutritional value (total digestible
nutrients [TDN] and protein content) compared to the nutritional value of alfalfa [3]. Only
animals that consume this pretreated animal feed are considered to be within system boundary of
the LCA and only the manure generated by these animals is considered to be sent to the depots
for biogas production.

Methane emissions (converted to CO2 equivalent emissions) from only the animals
within the system boundary are included in the net greenhouse gas emission calculations. Coproduct credits for all animal related co-products based on nutritional values were calculated
using the displacement method and the general formula [3]:
Co-product credit (for y displacing x) = [animal feed production (kg feed/kg y)*(Kg x
displaced/kg animal feed)]* [x production energy] (MJ/kg x)

96

Where y is the new animal feed product namely pretreated stover, perennial grasses or
LPC extract from double crop and x is the original animal feed product being displaced (corn
grain and soybean protein and forage). For biochar in scenario 3, the environmental co-product
credits were calculated based on the avoided emissions from fertilizers [9, 32]. The heat and
electricity burdens of the biorefinery were allocated to each LBPD based on the amount of
pretreated biomass sent to the biorefinery from that particular LBPD. Similarly, the energy
output from the biorefinery was allocated to each LBPD based on the pretreated biomass it
provided to the biorefinery. The data from various system modules were consolidated and a life
cycle assessment was conducted for the combined landscape-transport-processing system to
determine NEG and NCER for the overall system.

97

Table 4.1: Scenario description- Reconfiguration of RIMA-MI land area and
technologies used in each scenario
Scenario

1

2a

2b

3

Base+AD+LPC

Base+AD+LPC+

Number (S)
Technology

a

b

Base +AD

Base+AD+LPC

c

suite

pyrolysis (in
forested region
depots)

Corn stover

40%
harvested

40% harvested

40% harvested

d

50% land converted
to perennial grasses,
60% of remaining
stover harvest

Forage

25%

50% land converted to

50% land

100% land

harvested

perennial grasses

converted to

converted to

perennial grasses,

perennial grasses

25% of remaining
forage harvested
Perennial

-

Grown on 50% of

Grown on 50% of

Grown on 100% of

grasses

converted forage land

converted forage

converted forage

(switchgrass

and on all marginal

land and on all

land and on 50% of

&

land

marginal land

marginal land

miscanthus)

98

Table 4.1 (Cont’d)
Winter rye

-

Grown on 30% of

Grown on

Grown on 75% of corn

corn land

30% of corn

land not converted to

land

perennial grasses

-

Grown on 50% of

(double
crop)
Native

-

-

prairie

marginal land

Woody

-

-

-

residue

Obtained from major
forested watersheds

Base stands for AFEX™ and densification technologies
AD stands for Anaerobic digestion
LPC stands for leaf protein concentrate extraction
In the corn stover row, % indicates percent of total corn land considered in RIMA-MI. ^
In the forage row, % indicates percent of total forage land considered in RIMA-MI
e. Marginal lands are defined as land not being used to grow commercial or conventional
crops that may be capable of growing low-maintenance, low-input biomass feedstocks
such as perennial grasses [3] as determined from BASINS as “transitional lands”
a.
b.
c.
d.

Table 4.2: Transportation distances of feedstocks, raw-material and co-products
LBPD

Distance from feedstock and animal farms to LBPD

Distance from LBPD to

no.

and vice-versa (mi)

biorefinery (mi)

S1

S2a

S2b

S3

S1, S2a, S2b, S3

1

21.53

13.78

13.24

20.09

51

2

25.58

15.88

15.17

23.19

36

3

25.58

15.88

15.17

23.19

37

4

25.58

15.88

15.17

23.19

68

5

12.83

8.88

8.58

12.66

48

99

Table 4.2 (Cont’d)
6

12.83

a

21.53

7

8.88

8.58

12.66

34

13.78

13.24

20.09

-

a. LBPD 7 does not generate animal feed nor does it use manure or generate AD solids as it
is co-located with the central biorefinery.
4.3. Results and discussion
4.3.1. Base case
The current situation (or base case) was assumed to produce no cellulosic biofuels. Then
the land use was repurposed or reconfigured to produce feedstocks for cellulosic biofuel
production. Scenarios were formulated starting from a base reconfiguration in scenario 1 (or S1)
which utilizes some corn stover and forage; to intermediate configurations in scenarios 2a and 2b
(or S2a, S2b) which make use of perennial grasses and marginal lands; to an advanced
configuration in scenario 3 (or S3) which employs the most land efficient techniques and new
biomass feedstocks. Table 4.1 summarizes these scenarios, land area reconfigurations and also
shows that the technologies included are directly dependent on the RIMA configuration, starting
from the base technologies and subsequently incorporating other technologies previously
discussed. Varying amounts of alfalfa hay, an important dairy cattle feed [33], are taken out of
the landscape and sent to processing instead of being used as animal feed in each scenario.
Twenty five percent of current-situation (or base-line) forage in S1, 50% in S2a, 75% in S2b and
100% in S3 is sent to processing in the LBPDs and is replaced with an equivalent amount of
pretreated biomass based on its nutritional value[3]. Henceforth any reference to the landscape
implies that animal operations are included within the landscape. Forage land was converted to
higher yielding perennial grasses in S2 and S3 in increasing amounts. Among perennial grasses
100

there was a 50%-50% distribution of land area between switchgrass and miscanthus. Restored
prairies were grown only on marginal lands in scenario 3 since they have low yields but also low
inputs.

More land is brought into production and used more efficiently in S3 by using more
forage land (after converting it to perennial grasses), more double cropping, use of marginal
lands and woodlands. Energy and emission inputs and outputs from all modules namely
landscape, processing and transport were consolidated in a life cycle assessment to determine
energy and (some) environmental impacts of the overall system.

4.3.2. Base case results
Figure 4.1 a, b shows the energy input and emission output burdens of the different
system modules. Appendix B contains absolute energy and emission burdens. Feedstock
production constitutes a large portion of overall energy burdens in scenario 1 due to the inclusion
of more energy intensive and less energy dense feedstocks (corn stover and forage) with lower
average yields. In scenario 3, including multiple processing technologies leads to greater
processing energy burdens generated in the LBPDs. In the LBPD processing module, the
AFEX™ process was found to have the largest energy requirement among all technologies in
the LBPDs per ton of biomass processed since it has a substantial natural gas requirement for
ammonia recovery.

101

100%
90%
80%

Processing
(LBPD)

70%
60%
50%

Transport

40%
30%
20%

Feedstock

10%
0%
S1

S2a

S2b

S3

(a)
Figure 4.1: (a) Energy burdens of system modules, percent of total (b) CO2 eq. GHG emission
burdens of system modules, percent of total

102

Figure 4.1 (Cont’d)

100%
90%
80%
Animal
operations

70%
60%

Processing
(LBPD)

50%
40%

Transport

30%
20%

Feedstock

10%
0%
S1

S2a

S2b

S3

(b)
Biorefinery processing inputs and emissions amount to zero since all its requirements are
covered by burning lignin to produce steam and electricity. The contribution of transport
emissions to total emissions increases in S3 because of extra transport burdens due to woody
residues and multiple co-products.
The total production of biomass per acre of the RIMA-MI increases substantially in the
more complex scenarios ( S2, S3) by utilizing previously unused or under-used land areas and by
implementing land management techniques such as the use of double crops (e.g.; winter rye). In
this analysis the “biomass multiplier” is defined as the ratio of total cellulosic biomass used in

103

each scenario to the cellulosic biomass currently generated in the RIMA (i.e. total forage land
yield). The denominator in the biomass multiplier ratio includes only forage (which is currently
used as animal feed), since no stover is currently harvested in RIMA-MI for the purposes of
either biofuel production or as a co-product (animal feed) and advanced cellulosic feedstocks do
not yet exist in the RIMA. With each intensification in the RIMA-MI reconfiguration, biomass
multipliers change from as low as 1.7 in scenario 1 to 11.7 in scenario 3. Potential usable land
area in the RIMA-MI includes the sum of all corn and forage land, all forested areas and
transitional lands. This amounts to only 21% in S1 and about 73% in S3. There is a 27% land
area that is unused which can also potentially be used for biofuel production but was not
considered in this study. Figure 4.2 a) shows the distribution of land area to different feedstocks
and the biomass multiplier. Figure 4.3 a) shows LBPD profiles for net energy gain (NEG) and
Figure 4.3 b shows carbon emission reductions (NCER) on a functional unit basis (one hectare of
RIMA-MI land area). NEG is defined as the difference between all renewable energy outputs
from ethanol and electricity produced by the system and the fossil energy inputs into agriculture,
transport and processing. NCER is defined as the difference between carbon sequestration
potential of the feedstock system and all emissions from harvest, transport and processing
systems. Carbon “sequestration” potential includes gasoline displaced by ethanol, soil nitrous
oxide emissions as well as gain or loss of soil organic matter. Because of the improved efficiency
with which the RIMA land area is used, NEG is increased by 88% and NCER is increased by
68% in S3 relative to S1.

104

12

Woody residues

300
10

Native prairie

8

Winter rye (on
corn land)

200
6
150

4

100

Biomass multiplier (ratio)

Land area (1000X ha)

250

Switchgrass

Miscanthus

Forage

50

2

Corn Stover

0

0

Biomass
multiplier

S1

S2a

S2b

S3

Figure 4.2: Land area distribution of different feedstocks and biomass multiplier.

Figure 4.4 shows the watershed-profiles nitrate (NO3-) leaching within the feedstock
landscapes on both the functional unit basis and also on the amount of ethanol generated in each
scenario. A 45% increase in nitrate leaching is observed on average from S1 to S3 on a RIMA
land area basis due to large acreages of land being dedicated to grow (fertilized) perennial
grasses (perennial grasses have nitrate leaching values comparable to a corn-cover crop system
with no-tillage as simulated by the Daycent model [34]).

105

10

45

9

40

8

35
LBPD1

30

6
25

5
20
4
15

3

10

2
1

LBPD3

LBPD4
LBPD5
LBPD6

LBPD7
Total

5

0

LBPD2

NEG (GJ)

NEG (GJ)

7

0
S1

S2a

S2b

S3

(a)
Figure 4.3: (a) Net energy gain [NEG]- LBPD profiles and averages (b) Net carbon emission
reductions [NCER] - LBPD profiles and averages. The base-case for these values is set on the
vertical line at zero.

106

Figure 4.3 (Cont’d)

5

0.8

4

0.6

3

0.4

2

NCER (Mg CO2 eq.)

6

1.0

NCER (Mg CO2 eq.)

1.2

LBPD1
LBPD2
LBPD3
LBPD4
LBPD5

LBPD6
LBPD7
Total

0.2

1

0.0

0
S1

S2a

S2b

S3

(b)
However, based on per unit ethanol production, S3 has the least nitrate leaching impact
due to greater amounts of ethanol produced in this scenario. Overall, NEG and NCER are both
greatest in the scenario in which the land area is used most effectively and a large amount of
perennial grasses and native prairies are cultivated (scenario 3). Various sensitivity analyses
were performed on scenario 3 to determine the most significant parameters affecting NEG and
NCER. The biofuel production rate from each scenario is 2037, 2676, 2555 and 3869 liters/ ha in
S1, S2a, S2b and S3 respectively. These values are calculated based on the amount of ethanol
107

produced from the biomass sent to the biorefinery in each scenario and on the total RIMA land
area used in each scenario.
The Energy Independence and Security Act [EISA] of 2007 [35] mandates the production
of 16 billion gallons of cellulosic biofuels by 2022. With the ethanol production rate of S3, it is
possible to generate this amount of ethanol using about 16 million hectares of land. This amounts
to approximately 14% of US cropland currently being used to produce animal feed, corn ethanol
and exports [36], while still providing the same amount of animal feed. All displaced animal
calorie and protein requirements are replaced and by producing animal feed in addition to
renewable fuel from the same land area, indirect land use change (ILUC) effects are avoided[37,
38]. With 6 Mg/ha NCER predicted in scenario 3, 16 million hectares of land area can act as a
sink for 96 Tg of CO2 eq. emissions annually. This amounts to approximately 2% of the total
annual GHG emissions of the U.S. [39] (this is the extrapolated total CO2 eq. emissions sink
after reconfiguring the land area of 16 million ha as in scenario 3. The base-line in this study is
“no cellulosic biofuel production”; therefore the assumption is that currently this land does not
act as a sink for CO2 eq. emissions from a cellulosic biofuels production perspective). In
addition the EISA regulates the use of feedstocks expected to have minimal land use change
impacts and those that do not adversely affect food production such as residues, cover crops or
perennial grasses. As mentioned previously, this constraint is thoroughly satisfied within the
scenarios analyzed here.

108

30

0.4

0.35
St.Joseph
0.3
20
0.25

15

0.2
0.15

10
0.1

Kg NO3-/gal ethanol produced

Nitrate Leaching (Kg NO3-)

25

Upper Grand
Kalamazoo
Black-Macatawa
Thornapple
Average

St.Joseph
Upper Grand
Kalamazoo
Black-Macatawa
Thornapple

5

0.05
0

Average

0

S1

S2a

S2b

S3

Figure 4.4: Soil N2O loss and NO3- leaching (based on functional unit [RIMA-MI land area]
and on unit ethanol production) - watershed profiles and averages.
4.3.3. Sensitivity analyses
Scenario 3 offers the best results for energy gains as well as for emission reductions.
Moreover, this scenario contains all representative technologies in a majority of the depots.
Therefore S3 was chosen as the base case from which to perform several sensitivity analyses.
The most important sensitivity analyses included: A) Varying feedstock cultivation and harvest
energy inputs and emission outputs by +/- 25%. B) Varying feedstock yields by +/-25%. C)
Varying AFEX™ process energy by +/-25%. D) Using corn stover cultivation inputs from the
GREET model to evaluate the effect of allocation (of inputs between corn grain and stover) on
system results. E) Optimizing the pyrolysis process for bio-char production rather than bio-oil

109

production. One question that arises during the development of the LBPD model employing the
pyrolysis process was whether it would be more favorable to optimize pyrolysis for bio-char
production vs. for bio-oil production. In the original scenario 3 the pyrolysis process in the
LBPD technological model is optimized to produce more bio-oil in order to meet the renewable
liquid biofuel goal.
Differences in results of sensitivity analyses compared to the base case were calculated
as: ∆NEG = ((NEG sensitivity S3 - NEG base-case S3)/ NEG sensitivity S3)*100,
∆NCER= ((NCER sensitivity S3 - NCER base-case S3)/ NCER sensitivity S3)*100

4.3.4. Sensitivity analyses results
Variations in feedstock yields have the largest effect on NEG (∆NEG=-33 to +20 %) and
NCER (∆NCER= +/-9%) whereas the other parameters do not seem to significantly impact the
system as seen in Figure 4.5. Feedstock yields affect all system modules including the amount of
feedstock cultivated and harvested (thus changing feedstock production inputs), amount of
feedstock transported (thus changing transport inputs) and amounts of feedstock processed to
finished products (thus changing LBPD and biorefinery inputs). Increased feedstock yields also
provide more energy production from raw material per unit area of the RIMA. This fact explains
the increased NEG in case of higher feedstock yields and a decrease in NEG with lower
feedstock yields. However, increased feedstock yields increase LBPD processing emissions
thereby decreasing overall NCER and vice-versa.

110

25% increase in feedstock yield

25% decrease in feedstock yield
25% decrease in AFEX™ process
energy
25% increase in AFEX™ process
energy
GREET model corn stover inputs

25% increase in feedstock operation
burdens
25% decrease in feedstock operation
burdens
Pyrolysis optimized for biochar
production
∆ NCER

∆ NEG

-30

-20

-10

0

10

20

30

Figure 4.5: Results of the sensitivity analyses as a percentage increase or decrease from values
of original scenario 3 (on which the sensitivity analyses were performed)
4.4 Conclusion
Large scale commercial cellulosic biofuel refineries are currently being built and many
more are expected in the future as crop production and processing technologies improve. This
study provides a foundation for a novel approach to cellulosic biomass logistics that may help
resolve biomass supply chain constraints as biofuel demand increases. Using this comprehensive
modeling approach, the most promising scenarios for cellulosic ethanol supply chains can be
identified. The depot network can help bridge the gap to biofuel commercialization by providing
valuable co-products for biofuels and thereby create more market demand to establish such
systems.

111

Greater energy and environmental benefits accrue to the overall system beyond simply
improved logistics when the challenges associated with biomass supply and processing are
minimized. By implementing land efficient techniques (such as use of marginal lands to grow
low input high diversity feedstocks, double-cropping, recycling wastes within the system) and
by producing valuable co-products (such as animal feed) close to the feedstock operations depot
networks can respond to local and global sustainability concerns while simultaneously
increasing feedstock value. The methodology adopted in this analysis can be applied to a wide
range of landscapes, feedstocks and flexible depots to evaluate the net energy returns and the
environmental benefits of this new approach to developing supply chains for cellulosic biofuels.
This analysis serves as a foundation to create the watershed-scale optimized and rearranged
landscape design (WORLD) model that will be discussed in Chapter 5.

112

APPENDIX B

113

4.5. Appendix B
B.1. The “watershed weightage factor” method for division of RIMA-MI into watersheds
Seven digit HUC watersheds covering the counties belonging to RIMA-MI were
identified using the BASINS model. The areas of these large watersheds were found to be greater
than the combined 9 county area thereby exceeding the boundaries of the area of interest in this
analysis, namely the RIMA-MI. Therefore, to allocate biomass yields and transport distances
only to watershed areas belonging within the 9 counties of RIMA-MI, a weightage method was
developed. The following calculations and map manipulations were done using the BASINS
model and the ArcGIS software: The total area of each complete watershed was calculated. The
land areas of each county were calculated. The percentages of county areas contributing to each
watershed were determined. A “watershed weightage factor” (wwf) was then developed as the
sum of all % county land areas belonging to each watershed. Actual land areas for different land
uses (cropland, pasture, forest lands, marginal lands, etc) were determined using the geographic
information retrieval and analysis system (GIRAS) database in BASINS. Land areas to be
considered for analysis (i.e equivalent to RIMA-MI land area) were then calculated using wwfs
on actual land areas.
B.2. The “geometric factor” method for assignment of watershed land area portions to
depots
The RIMA-MI was divided equally into land-area portions covering each RBPD (based
on a manual square grid division). The percentage of land area of each watershed dedicated to
each LBPD was determined. For example 48.3% of the Thornapple watershed land area is
apportioned to LBPD 5 and 51.7% of the Thornapple watershed land area is apportioned to
LBPD 6. Similarly all five watershed areas were divided between the seven LBPDs. Using these
114

percentages, different raw-materials and co-products (such as feedstocks and manure) from the
watersheds were allotted to respective LPBDs.
Table B1: Yields of currently available feedstocks
Watershed

Corn stover (tons/ac)

Forage (tons/ac)

40% harvest 60% harvest
St.Joseph

1.32

1.98

2.50

Upper Grand

1.26

1.88

1.79

Kalamazoo

1.27

1.90

2.34

Black-Macatawa

1.26

1.89

2.66

Thornapple

1.29

1.93

2.30

Yields are calculated using NASS data [40] and wwf method (B1). Stover yields were
estimated based on a 1:1 mass ratio from corn yields (converted from bushels to tons)
Table B2: Yields of currently unavailable feedstocks estimated from literature
values
Feedstock

Yield (tons/ac)

Sources

Winter rye

2.4

[41, 42]

Switchgrass

7

[43, 44]

Miscanthus

10

[45-47]

Native prairie

0.97

[22, 23]

Woody biomass

2.7

[48-50]

115

B.3. Land uses in different watersheds

Figure B.1: Land use in the St. Joseph watershed in RIMA-MI

116

Figure B.2: Land use in the Kalamazoo watershed in RIMA-MI

117

Figure B.3: Land use in the Thornapple watershed in RIMA-MI

118

Figure B.4: Land use in the Upper Grand watershed in RIMA-MI

119

Figure B.5: Land use in the St. Joseph watershed in RIMA-MI

120

B.4. Transport calculations
A 20% road winding factor was applied to these calculated values to account for nonlinear road networks. Densified and non densified transport were differentiated by modifying the
cargo payload (or truck capacity) suitably (higher payload for densified material vs. lower for
non-densified) in GREET 1.8b [31]. Formula for collection distances from farms[51]:
2

Collection area (mi ) = ß/ [δ* ρ*(1- І)* λ*640]
Where ß= biomass collected (dry tons/ yr), δ= density of feedstock acreage, ρ= % of
farmers selling feedstock, I = % of fields inaccessible, λ= biomass yield. Whereas δ, ρ and I
remain constant (~0.2 on average, 0.5, and 0.1 respectively), ß and λ vary in different scenarios
thus changing the collection area.
B.5. Nutritional values and co-product credit calculations
The following calculations are performed based on similar calculations in the
comparative LCA study [3].

For stover

It is assumed that pretreated corn stover displaces corn grain animal feed. Displacement
ratio = 0.653 (1 kg of stover can displace 0.653 kg of corn fed as animal feed based on TDN
content), Animal feed production ratio = 1 (1 kg of stover produces 1 kg animal feed). Using the
displacement method equation:
Co-product credit_corn = [animal feed production (kg feed/kg stover)*(Kg corn
displaced/kg animal feed)]* [corn production energy] (MJ/kg grain)

121

For perennial grasses

It is assumed that pretreated perennial grasses displace soybean protein animal feed.
Displacement ratio = 0.202 (1 kg of perennial grass (SG or miscanthus) displace 0.202 kg of
soybean protein), Animal feed production ratio = 1 (1 kg of grass produces 1 kg animal feed)
Co-product credit_soy = [animal feed production (kg feed/kg grass)*(Kg soy
displaced/kg animal feed)]*[soy production energy (MJ/kg)]

For rye forage

It is assumed that leaf protein concentrate produced from forage displaces soybean
protein animal feed. Displacement ratio = 0.804 (1 kg of perennial grass (SG or miscanthus)
displace 0.804 kg of soybean protein), LPC production ratio = 0.18 (1 kg of forage yields 0.18 kg
LPC)
Co-product credit_soy = [LPC production (kg feed/kg forage)*(Kg soy displaced/kg
LPC)]* [[(soy production energy-rye production energy) (MJ/kg)]

Carbon emission reduction co-product credits

Similar calculations for co-product credits of greenhouse gas emissions were performed.
For perennial grasses extra emission reduction for any corn or forage land displaced was also
included (in Scenarios 2a, 2b and 3)
Additional GHG emission sequestered (due to below ground root mass) = perennial grass
acreage grown on originally corn or forage land *(sequestration by perennial grass- sequestration
by corn or forage).

122

Soil nitrous oxide emissions were included in the feedstock emission output calculations
by converting to CO2 eq. using a factor of 298 for the global warming potential of N2O.

B.6. Results of energy and emission burdens as absolute values

1.E+07
9.E+06

Energy Inputs (GJ)

8.E+06
7.E+06

Feedstock

6.E+06

Transport

5.E+06

Processing (LBPD)

4.E+06

3.E+06
2.E+06
1.E+06
0.E+00
S1

S2a

S2b

Figure B.6: Absolute energy inputs of the different system modules

123

S3

8.E+05
7.E+05

Emission outputs (Mg CO2)

Feedstock

6.E+05

Transport
Processing (LBPD)

5.E+05

Animal operations

4.E+05
3.E+05
2.E+05

1.E+05
0.E+00
S1

S2a

S2b

Figure B.7: Absolute emission outputs of the different system modules

124

S3

REFERENCES

125

4.6. REFERENCES

1.
Carolan, J.E., et al., Technical and Financial Feasibility Analysis of Distributed
Bioprocessing Using Regional Biomass Pre-Processing Centers. Journal of Agricultural & Food
Industrial Organization, 2007. 5(10): p. 1-29.
2.
Eranki, P.L., et al., Advanced Regional Biomass Processing Depots: a key to the
logistical challenges of the cellulosic biofuel industry. Biofuels, Bioproducts and Biorefining,
2011. 10.1002/bbb.318: p. 1932-1031.
3.
Eranki, P.L. and B.E. Dale, Comparative life cycle assessment of centralized and
distributed biomass processing systems combined with mixed feedstock landscapes. GCB
Bioenergy, 2011. 10.1111/j.1757-1707.2011.01096.x: p. 1757-1707.
4.
Kazi, K.F., et al., Techno-Economic Analysis of Biochemical Scenarios for Production
of Cellulosic Ethanol. NREL Report No.TP-6A2-46588, 2010. (Accessed, August 10, 2012)
5.
Sokhansanj, S., et al., Large-scale production, harvest and logistics of switchgrass
(Panicum virgatum L.) - current technology and envisioning a mature technology. Biofuels,
Bioproducts and Biorefining, 2009. 3(2): p. 124-141.
6.
Sokhansanj, S., et al., Binderless Pelletization of Biomass. Paper no: 056061.ASAE
Annual International Meeting, Tampa, FL.. St. Joseph, MI. (Accessed, August 10, 2012)
7.
Bals, B. and B.E. Dale, Economic comparison of multiple techniques for recovering leaf
protein in biomass processing. Biotechnology and Bioengineering, 2010. 108(3): p. 530-537.
8.
Enochian, R.V., et al., Producing Pro-Xan (Leaf Protein Concentrate) from Alfalfa:
Economies of an emerging technology. Report no. 445, 1980. USDA Agricultural Economics,
Washington DC, USA
9.
Gaunt, J.L. and J. Lehmann, Energy Balance and Emissions Associated with Biochar
Sequestration and Pyrolysis Bioenergy Production. Environmental Science & Technology, 2008.
42(11): p. 4152-4158.

126

10.
Hu, Z.W., et al., Substitution of Biomass for Coal and Coke in Ironmaking Process.
Advanced Materials Research 2011. 236-238,77.
11.
Zhang, Q., et al., Review of biomass pyrolysis oil properties and upgrading research
Energy conservation and management, 2007. 48(1): p. 87-92.
12.
Balsam, J. and D. Ryan, Anaerobic Digestion of Animal Wastes: Factors to Consider.
ATTRA- National Sustainable Agriculture Information service, 2006. IP-219: p.
http://attra.ncat.org/attra-pub/anaerobic.html. (Accessed, August 10, 2012)
13.
Slater, S., K. Keegstra, and T. Donohue, The US Department of Energy Great Lakes
Bioenergy Research Center: Midwestern Biomass as a Resource for Renewable Fuels.
BioEnergy Research, 2010. 3(1): p. 3-5.
14.
USCB, U.S Census Bureau State and County Quickfacts. 2012.
http://quickfacts.census.gov/qfd/states/26/26159.html. (Accessed, August 10, 2012)
15.
USDA, Census of agriculture- MI county data 2007.
http://www.agcensus.usda.gov/index.php. (Accessed, August 10, 2012)
16.
Bals, B.D. and B.E. Dale, Developing a model for assessing biomass processing
technologies within a local biomass processing depot. Bioresource Technology, 2012. 106(0): p.
161-169.
17.
Jones, S., et al., Production of Gasoline and Diesel from Biomass via Fast Pyrolysis,
Hydrotreating and Hydrocracking: A Design Case PNNL-18284, 2009. Richland, WA (US):
Pacific Northwest National Laboratory (PNNL). (Accessed, August 10, 2012)
18.
Wright, M.W. and R.C. Brown, Distributed processing of biomass to bio-oil for
subsequent production of Fischer-Tropsch liquids. Biofuels, Bioproducts & Biorefining, 2008. 2:
p. 229-238.
19.
Tanzil, D., et al., Designing for Sustainability, in Transforming Sustainability Strategy
into Action. 2005, John Wiley & Sons, Inc. p. 163-197.
20.
Sendich, E. and B.E. Dale, Environmental and economic analysis of the fully integrated
biorefinery. Global Change Biology Bioenergy, 2009. 1(1): p. 331-345.

127

21.
Humbird, D., et al., Process Design and Economics for Biochemical Conversion of
Lignocellulosic Biomass to Ethanol: Dilute-Acid Pretreatment and Enzymatic Hydrolysis of
Corn Stover. National Renewable Energy Laboratory Report, 2011. NREL/TP-5100-47764.
22.
Tilman, D., J. Hill, and C. Lehman, Carbon-Negative Biofuels from Low-Input HighDiversity Grassland Biomass. Science, 2006. 314(5805): p. 1598-1600.
23.
Florine, S.E., et al., Yield and composition of herbaceous biomass harvested from
naturalized grassland in southern Iowa. Biomass and Bioenergy, 2006. 30(6): p. 522-528.
24.
Ceotto, E., Grasslands for bioenergy production. A review. Agron. Sustain. Dev., 2008.
28(1): p. 47-55.
25.
Meehan, T.D., et al., Agricultural landscape simplification and insecticide use in the
Midwestern United States. Proceedings of the National Academy of Sciences. 108(28): p. 1150011505.
26.
Gardiner, M.M., et al., Landscape diversity enhances biological control of an introduced
crop pest in the north-central USA. Ecological Applications, 2009. 19(1): p. 143-154.
27.
Schnoor, J.L., et al., Water Implications of Biofuels Production in the United States.
National Academy of Sciences, Washington DC, USA, 2007.
http://www.nap.edu/catalog/12039.html. (Accessed, August 10, 2012)
28.
Berndes, G., Bioenergy and water--the implications of large-scale bioenergy production
for water use and supply. Global Environmental Change, 2002. 12(4): p. 253-271.
29.
Zhang, X., et al., An integrative modeling framework to evaluate the productivity and
sustainability of biofuel crop production systems. GCB Bioenergy, 2010. 2(5): p. 258-277.
30.
USEPA, BASINS 4.0 (Better Assessment Science Integrating Point and Non-point
Sources). http://water.epa.gov/scitech/datait/models/basins/BASINS4_index.cfm, 2010.
(Accessed, August 10, 2012)
31.
Wang, M.Q., Development and use of GREET 1.6 full-cycle model for transportation
fuels and vehicle technologies. ANL/ES/TM-163, 2001. Washington , DC : U.S. Department of
Energy. (Accessed, August 10, 2012)

128

32.
Wood, S. and A. Cowie, A Review of Greenhouse Gas Emission Factors for Fertiliser
Production. IEA Bioenergy Task 38, 2004. www.ieabioenergy-task38.org/publications/GHG
Emission_Fertilizer %20Production_July2004.pdf. (Accessed, August 10, 2012)
33.
Gordon, C.H., et al., Preservation and Feeding Value of Alfalfa Stored as Hay, Haylage,
and Direct-Cut Silage. Journal of Dairy Science, 1961. 44(7): p. 1299-1311.
34.
Parton, W.J., et al., Generalized model for NOx and N2O emissions from soils. Journal of
geophysical research, 2011. 106(D15): p. 17403-17420.
35.
Sissine, F., Energy Independence and Security Act of 2007: a summary of major
provisions.A CRS Report for Congress. http://energy.senate.gov/public/_files/RL342941.pdf,
2007. (Accessed, August 10, 2012)
36.
Dale, B.E., et al., Biofuels Done Right: Land Efficient Animal Feeds Enable Large
Environmental and Energy Benefits. Environmental Science & Technology, 2010. 44(22): p.
8385-8389.
37.
Fargione, J., et al., Land Clearing and the Biofuel Carbon Debt. Science, 2008.
319(5867): p. 1235-1238.
38.
Kim, H., S. Kim, and B.E. Dale, Biofuels, Land Use Change, and Greenhouse Gas
Emissions: Some Unexplored Variables. Environmental Science & Technology, 2009. 43(3): p.
961-967.
39.
USEPA, Inventory of U.S greenhouse gas emissions and sinks: 1990 – 2009 EPA 430-R11-005, 2006. (Accessed, August 10, 2012)
40.
Iwig, W.C. and W.L. Schaible, The National Agricultural Statistics Service County
Estimates Program Indirect Estimators in U.S. Federal Programs. 1996, Springer New York. p.
129-144.
41.
Balkcom, K.S., Arriaga, F.J., Raper, R.L., Bergtold, J.S., Nitrogen Fertilizer
Requirements to Maximize Rye Cover Crop Biomass Production [Abstract]. AU Symposium
"Energy Solutions for Alabama Natural Resources". October 23-24, 2007. (Accessed, August 10,
2012)

129

42.
Lyndon L. Kelley and D.R. Mutch, Harvesting Irrigated Cover Crops as Forage.
Michigan State University Extension, Purdue Extension, 2008.
http://www.msue.msu.edu/objects/content_revision/download.cfm/item_id.534480/workspa
ce_id.-30/CCForage.pdf/. (Accessed, August 10, 2012)
43.
MSUExtension, Switchgrass variety testing for biofuel in 2008. UP Ag connection, 2008.
14, 7. (Accessed, August 10, 2012)
44.
Popp, M., Assessment of Alternative Fuel Production from Switchgrass: An Example
from Arkansas. Invited Paper, Southern Agricultural Economics Association meetings in Mobile,
AL, Feb. 3-8, 2007, 2007. (Accessed, August 10, 2012)
45.
Scurlock, J.M.O., Miscanthus: A Review of European Experience with a Novel Energy
Crop. Publication No: 4845. ORNL Environmental Sciences Division, 1999. (Accessed, August
10, 2012)
46.
Pyter, R., et al., Growing giant miscanthus in Illinois. University of Illinois 2007.
http://miscanthus.illinois.edu/wp-content/uploads/growersguide.pdf. (Accessed, August 10,
2012)
47.
Heaton, E. and T. Teske, Giant miscanthus for biomass production. Iowa State
University Extension, Department of Agronomy, 2010.
http://www.extension.iastate.edu/Publications/AG201.pdf. (Accessed, August 10, 2012)
48.
USDA, Woody biomass utilization desk guide. USDA forest service, National
Technology and Development Program. Washington, D.C, 2007 (Accessed, August 10, 2012)
49.
Eker, M., Assessment of procurement systems for unutilized logging residues for Brutian
pine forest of Turkey. African Journal of Biotechnology, 2011. 10(13): p. 2455-2468.
50.
Yoshioka, T., et al., Cost, energy and carbon dioxide effectiveness of a harvesting and
transporting system for residual forest biomass. Journal of Forest Research, 2002. 7(3): p. 157163.
51.
Perlack, R. and A. Turhollow, Assessment of options for the collection, handling, and
transport of corn stover. ORNL/TM-2002/44, 2002. Oak Ridge National Laboratory, Oak Ridge,
TN p. 29. (Accessed, August 10, 2012)

130

CHAPTER 5: THE WATERSHED-SCALE OPTIMIZED AND REARRANGED
LANDSCAPE DESIGN (WORLD) MODEL AND LOCAL BIOMASS PROCESSING
DEPOTS FOR SUSTAINABLE BIOFUEL PRODUCTION: AN INTEGRATED LIFE
CYCLE ASSESSMENT
5.1. Introduction
Several plant biomass types such as those derived from corn (Zea mays L.) stover,
switchgrass (Panicum virgatum L.), miscanthus (Miscanthus x giganteus), and native prairie
grasses, are under evaluation as “best feedstocks” for cellulosic biofuel production. Further,
management practices such as double-cropping or other practices tailored for marginal lands and
riparian buffers may contribute to the design of sustainable biofuel production systems.
However, since these feedstocks vary greatly in their net energy potential, co-product generation,
and environmental impact characteristics, it may prove most beneficial to grow a combination of
these feedstocks within specific landscapes. For example, a recent study connected low-diversity,
annual crop landscapes with heightened pest infestations and insecticide use and concluded that
perennial bioenergy crops such as switchgrass and mixed prairies may counterbalance these
effects [1]. In a related study, diverse landscapes were found to promote better bio-control
services, an important and beneficial ecological provision [2]. Moreover, since one of the key
concerns about biofuels is the so-called “food vs. fuel” conflict, sustainable landscape designs
that maintain the current food/feed production potential and also provide large amounts of
biomass for fuel must be explored and understood [3].
Distributed processing in the form of depots that preprocess and pretreat local biomass
form a solution to the logistical challenges of large- scale, centralized biorefineries. All depots
contain the ammonia fiber expansion (AFEX™) pretreatment and densification technologies,
and may also contain additional processing technologies (e.g., leaf protein concentrate (LPC)
131

extraction, anaerobic digestion, or pyrolysis (to produce a “bio-oil” plus a combustible gas and
biochar) based on the landscape characteristics. The depots vary in capacity, operational
characteristics, and in the feedstocks processed and co-products produced, as dictated by the
characteristics of the landscapes in which they are embedded (e.g., largely agricultural
landscapes vs. heavily forested ones). Appendix C summarizes the details of these technologies.
Watersheds constitute an ecologically-relevant scale at which to understand the potential
for integrating food and biofuel production with enhanced environmental performance. From the
results of all the previous analyses (Chapters 2, 4) where manual land allocations were made,
there is a requirement for an accessible model to predict how land might be allocated within this
area to provide cellulosic biofuel feedstocks while maintaining existing food/feed production
levels and enhanced environmental performance. Therefore, the watershed-scale optimized and
rearranged landscape design (WORLD) model was created.
The objective of this study is to describe and test the WORLD model for its ability to
allocate land areas to different feedstocks and combine these landscapes with the processing
technologies of LBPDs in a given region and thereby determine how this combination may
impact energy yields and the environmental performance of such landscapes. Figure 5.1 gives a
visual summary of the various modules, their inputs and outputs and the tools used in this
analysis.
5.2. Structure and methods
5.2.1. The watershed-scale optimized and rearranged landscape design model
WORLD is a user-friendly and flexible linear optimization model created to determine
optimal landscape configurations for a variety of feedstocks within real landscapes, for various
sets of optimization criteria. Instructions on setting up and running the WORLD model and
132

codes of macros used in the model are provided in Appendix D. The model is constrained by the
primary requirement that all of the food/feed provisioning services currently provided by the
landscape are provided by the new, rearranged landscape. Constraining the model in this way
avoids the so-called “food vs. fuel” issue and the associated indirect land use change (iLUC)
effect [4]. The analysis is structured for watershed-scale landscapes due to data availability, ease
of modeling and model validation [5]. The present study focuses on nine counties (Allegan,
Barry, Branch, Calhoun, Cass, Eaton, Kalamazoo, St.Joseph and Van Buren) in Southwest
Michigan (the “Regional Intensive Modeling Area” (RIMA)) [5, 6] discussed in Chapter 4. This
RIMA was divided into five 7-digit hydrologic unit code (HUC) watersheds (Upper Grand,
Black-Macatawa, Kalamazoo, St.Joseph and Thornapple) for the purpose of this analysis.

133

Cradle to Gate LCA
Inputs

Modules

Outputs

Raw
biomass and
manure

Fertilizers,
Pesticides,
Fuel for onfarm
operations

Landscape

Transport fuel

Transport

Heat,
Electricity,
Ammonia

Co-products
(pretreated
animal feed
LPC extract,
AD sludge,
biochar)

LBPD

Bio-oil

Transport
fuel
Heat,
Enzymes,
Nutrients

Pretreated
biomass

Transport

Ethanol
Electricity

Biorefinery

Figure 5.1: Modules, material flows and tools used in this analysis
134

Tools
The WORLD
model, USEPA
BASINS, ArcGIS,
EPIC model
Agricultural
budgets
Collection radii
ArcGIS, GREET,
GaBi transport
LBPD technical
process model
USEPA BASINS,
ArcGIS, GREET ,
GaBi transport
NREL/Dartmouth
biorefinery
model, NREL
report

The cellulosic feedstocks included in the study are agricultural residues (corn stover),
perennial grasses (switchgrass and miscanthus), forage crops (alfalfa [Medicago sativa L.] and
grass hay), native prairie grasses, winter wheat [Triticum aestivum L.] grown as a double crop
with corn/soy production, forest residues and perennial grasses used as buffer strips or riparian
buffers (a mix of switchgrass and miscanthus) along river banks and also canola (Brassica spp.).
The agricultural residues and perennial grasses were chosen because of their relevance as
primary feedstocks of interest in lignocellulosic biofuel production [7-10]. Native prairies have
attracted interest as potential sources of lignocellulosic feedstocks because of their positive
environmental effects and ability to generate biomass on abandoned/degraded land [11-13].
Canola is used to replace any displaced soybean oil in the reconfigured landscape.
The WORLD model obtains most of its inputs from GIS-based models such as the US
EPA BASINS model and the Environmental Policy Integrated Climate (EPIC) biogeochemical
model [14], as discussed in Chapter 4. EPIC is a field scale biofuel and crop simulation model
combining soil and crop components such as erosion, crop growth, nutrient balance and nitrogen
and carbon cycles [15]. These models provide data such as land availability and watersheds
within the RIMA, land use categories(from the National Land Cover database [16]) and road
network lengths, etc.
The primary inputs to the model are:
1. The type of land area in each watershed, namely marginal land (ML), pasture land (PL)
and arable land (AL). In this analysis, arable land is the land area used to grow
conventional crops such as corn and soy. Pasture land is defined as the land used to grow
forage crops such as alfalfa and grass hay. Marginal land areas were determined from the
national land cover database in GIS as “transitional lands” or those with land capability
135

class 7 within the National Resource Conservation Service definition [17]. Current
cropland and pastureland areas were determined using GIS software and from NASS
statistical surveys. Table C1, Appendix C contains details of current acreages of different
land category used in the model. The total land area used in the RIMA is the sum of these
three land categories. Forested lands and urban lands are not used in this analysis; only
cropland, pasture and marginal lands.
2. Biomass yields on each type of land category (arable, pasture or marginal) for each
watershed and the environmental impacts of each feedstock, as determined from the
EPIC model. Environmental impacts include changes in soil organic carbon (ΔSOC),
erosion, nitrogen (N) and phosphorus (P) losses, crop water use efficiency (CWUE) and
farm greenhouse gas emissions (FGE). ΔSOC is the total change in soil carbon after
growing a feedstock (positive ΔSOC indicates an environmental benefit as this indicates
carbon sequestration in the soil). Erosion is the average soil eroded over the land area.
Nitrogen and phosphorous losses are a result of surface runoff, sediment, lateral
subsurface flow and percolation below the root zone. Crop water use efficiency as
defined in this analysis is the ratio of biomass yield to evapotranspiration resulting from a
particular feedstock. Farm greenhouse gas emissions are the carbon dioxide (CO2) and
nitrous oxide (N2O) emissions (in CO2 equivalents) from crop production due to inputs
such as tillage, seed preparation, fertilizer production and use (including N2O generated),
and pesticide production. Improvements (decreases) in these environmental impacts
represent progress toward a more sustainable system [18, 19]. Table C2, Appendix C
shows the feedstocks used, their average yields over all watersheds and the management

136

practices assumed to produce these feedstocks in EPIC. Table C3, Appendix C shows the
assumed nutritional values of different feedstocks.
3. The RIMA currently produces conventional animal feeds such as corn grain, soy protein
and grasses. It is assumed that 41% of corn grain, 61% of soy and 100% of both alfalfa
and grass hay produced in the landscape are used as animal feed [20, 21]. Required
animal nutritional needs include total non-protein digestible nutrients (TNPDN), protein,
and fiber [3, 22, 23]. Oil produced from soybean in the landscape is also provided as an
input to the model so that any displaced oil due to growing new feedstocks instead of
soybeans may be replaced with canola oil.
4. Ethanol and electricity yields of each of the different feedstocks are from literature[24].
Table C4, Appendix C shows environmental impact values from the EPIC model. Table
C5, Appendix C shows the energy inputs for different feedstocks obtained from literature
and agricultural budgets. Table C6, Appendix C shows the assumed ethanol and
electricity yields of different feedstocks.
The following is a summary of the universal assumptions and constraints in the model:


Corn grain ethanol is excluded from the system boundary.



All marginal land (ML), pasture land (PL) and arable land (AL) used should be less than
or equal to the original configurations respectively and the sum of all ML,AL and PL
should equal the original



All fractions of feedstock going to animal feed should be between 0 and 1



Ratios of TNPDN, protein, fiber and oil coming from the new configuration to that of the
from the original configuration should be between 1 and 2

137



Stover and double crop land are excluded from the sum of arable land to avoid double
counting as they are grown on the same land area as corn



Early succession vegetation (ESV) or (cool season native prairie grasses simulated in
EPIC) is always allowed to grow only on marginal lands



Variable percentages of original feedstocks (corn, soy, grass hay, alfalfa) are assumed to
be sent to animal feed from the original configuration based on literature values



All biomass harvested from buffer strips is assumed to go to ethanol production (no
animal feed generation)



Original animal feed crops such as corn grain and soybean whenever displaced from the
landscape are replaced with nutritionally equivalent pretreated cellulosic feedstocks
Land areas of existing crops are varied or new feedstocks are added to the landscape on

different land categories- arable (AL), marginal (ML) or pasture (PL). The fraction of biomass
going either to ethanol production or to animal feed is varied in the model based on the
parameter being optimized in a particular scenario. The model also varies the amount of double
cropping and the amount of stover harvested from corn land. The model is configured to vary
these factors within a set of constraints. All marginal, pasture and arable land used in the new
configuration is not allowed to exceed the original areas respectively. In all cases, ESV is
allowed to grow only on ML but not on PL or AL. Conventional crops are never allowed to grow
on ML, similarly forage crops are never allowed to grow on AL; this is done in order to keep the
different types of land areas distinct. ESV is allowed to grow on ML only. In order to illustrate
the capabilities of this model, two scenarios were created:
Scenario 1 (S1): Technical biomass production potential of the landscape- In this
scenario there are no limitations set on the type of feedstocks grown on arable land. Double

138

cropping is allowed to take place on a significant but conservative amount of corn land of
between 20 and 30% [the upper limit for double cropping on corn land is around 50% of total
agricultural land (Kurt Thelen, personal communication, Feb, 2012)]. Riparian buffers can be
present in moderate amounts (up to 30% of available river bank areas) within the landscape.
Scenario 2 (S2): Biomass production potential of landscape in case of opposition to
conversion from conventional crops to novel bioenergy feedstocks- In this scenario, perennial
grasses are allowed to grow only on marginal lands and pasture lands while making no changes
to arable land currently used to grow conventional crops. There is considerably lower double
cropping (between 5 and 20%) and riparian buffers (10% of available river bank areas) used
within the landscape. These two scenarios were run on the model in combination with the
optimization categories (called tests) shown in Table 5.1,
The outputs from the model are the newly reconfigured land areas of different feedstocks
on different land categories- AL, ML, and PL.
Table 5.1: Summary of optimization categories (tests)
Tests

Description

T1:Fuel

Maximize ethanol production

T2:Soil

Minimize erosion
Maximize Δ Soil Organic Carbon (SOC)

T3:Water

Minimize Nitrogen losses
Minimize Phosphorus losses
Maximize Crop water Use Efficiency (CWUE)

T4:Emissions Minimize CO e. GHG emissions from feedstock production sector (FGE)
2

139

The original landscape contains corn, soybean, alfalfa and grass hay whereas the
reconfigured landscape contains a combination of switchgrass, miscanthus, winter wheat, and
ESV in addition to the original feedstocks.
The model also generates values of ethanol, electricity and animal feed produced and
environmental impacts of growing the feedstocks within the landscape. Environmental impacts
can be obtained either as absolute values for the new landscape or as percent changes compared
to baseline (where the original landscape forms the baseline).
In this analysis, instead of a low-input high- diversity native prairie, a fertilized cool
season native prairie also denoted here as ESV is simulated in EPIC and its resulting biomass is
also assumed to be used for ethanol production along with other feedstocks. Unfertilized native
prairie grasses may perhaps not generate substantial biomass yields but might nonetheless be
included in landscapes for their environmental benefits. Double cropping can enhance the
productivity of agricultural landscapes based on annual crops [3] and achieve some of the
benefits of perennial systems. Winter wheat is assumed to be grown as a double crop in the
analysis. Perennial grass buffer strips can minimize herbicide and sediment run-off and improve
water quality [25]. However, the analysis does not estimate improvements in water quality
versus systems without such buffers. Such buffers may minimize losses of sediment, nutrients
and pesticides by 50-70% [26]. Finally, the use of complex landscapes containing a mix of
cellulosic feedstocks will likely exhibit positive environmental benefits such as increased
biodiversity, increased wildlife habitat and reduced chemical use.
5.2.2. Local Biomass Processing Depots and Biorefinery
In all scenarios in this analysis, the processing module is assumed to contain a network of
depots providing pretreated biomass to the central biorefinery. For the distributed processing
140

system an in-house technical model [27] was developed to determine the processing energy
required based on technologies included in each scenario. This model uses a combination of
literature values and experimental results to form a basic material and energy balance around
each unit operation. These balances determine the fossil fuel and raw material input required.
For the biorefinery module the processing energy for operation and energy gains due to
electricity and ethanol produced were determined using the NREL/ Dartmouth biorefinery model
[28, 29].
5.2.3. Life Cycle Assessment
The functional unit in this analysis is the land area generating cellulosic biomass in the
RIMA and the reference flow is defined as one hectare of the RIMA land area. This is the
functional unit of choice since one of the primary goals of the analysis is to illustrate the
variations in land productivity with differences in types and amounts of feedstocks and the
management practices used in the RIMA. The system boundary in this study includes the
feedstock landscape (the only land area within the RIMA used to generate biomass), LBPDs, the
biorefinery and all associated transport operations between the feedstock and processing
modules. The production of only cellulosic ethanol (but not corn grain ethanol or soybean
biodiesel) is included in the system boundary. The energy and emission factors associated with
various feedstocks were obtained from agricultural budgets, extensive literature review[23] and
EPIC model results. The EPIC model results set stover harvest to an average of 60% (in all
cropping systems with stover removal). Conventional feedstocks such as corn and soy are only
included to determine arable land area, current animal feed requirements and oil production (in
the case of canola replacing soybeans).

141

Details of transportation distance estimation within the different watersheds are provided
in Table C7, Appendix C. Using these distances and the Greenhouse Gases, Regulated
Emissions, and Energy Use in Transportation (GREET) model [30] for both densified and nondensified transport where applicable, transport energy requirements and emissions were
calculated. Details of energy requirements and emissions of transportation of densified and nondensified material as well as those of bio-oil and biochar are provided in Table C8, Appendix C.
Manure from animal operations that goes to anaerobic digestion is also included in the
transportation calculations to depots. A 15% dry matter loss (DML) in transportation of raw
biomass from farms to LBPDs and a 10% loss of all other material (such as densified biomass
and solid co-products) is assumed in all scenarios.
As mentioned previously, one of the primary requirements in the WORLD model is to
provide equal or greater animal nutritional services in the newly reconfigured landscape as in the
original landscape. It is assumed that pretreated stover displaces TNPDN from corn grain,
perennial grasses switchgrass and miscanthus displace fiber from grass hay, and the double crop
and canola displace protein and oil from soy, respectively. Co-product credits for all animal feed
related co-products were based on nutritional values and were calculated using the displacement
method [23]. Details of co-product credit calculations (similar to those seen in Chapters 2 and 4)
are provided in Appendix C.
The data from various system modules i.e. the WORLD model for feedstocks, the LBPD
and biorefinery technological process models and transport modules are consolidated and an
integrated life cycle assessment is conducted to determine net energy yields (NEY) and net
greenhouse gas emission reductions (NGER) for the overall system. The absolute value of net
energy yield (NEY) is defined as the difference between total energy outputs (in the form of

142

ethanol, electricity and coproduct credits) and energy inputs of feedstock, transport and
processing systems of new cellulosic biofuel producing acreages in the newly reconfigured
RIMA. The absolute values of net greenhouse gas emission reduction (NGER) is defined as the
difference between total CO2-equivalent (CO2 e) GHG emissions displaced and total CO2 e
GHG emissions from feedstock, transport and processing systems of cellulosic biofuel producing
acreages in the newly reconfigured RIMA. GHG emissions displaced include emissions due to
ethanol displacing gasoline, emissions displaced due to surplus electricity generated in
biorefinery, and due to using new feedstocks (such as perennial grasses which increase soil
organic carbon) as animal feed as well as raw material for biofuel production in place of
conventional feedstocks. Fargione et al. [46] state that a “biofuel carbon-debt” may occur due to
the release of CO2 emissions as a result of converting native habitats into alternate forms such as
croplands. They define the debt as CO2 emissions released during the first 50 years of this land
conversion. A question regarding incurring “carbon-debt” while bringing previously unused
marginal lands into production may arise in the calculation of net carbon emissions. However,
incurring a carbon-debt is unlikely in this analysis for the following reasons:
1. Firstly, in the study stated above [46] it is seen that the C-debt accumulated is
greater if the marginal/abandoned land is converted to a conventional crop such as
corn to generate corn-ethanol. However this debt is low or negligible while
generating prairie ethanol on abandoned/marginal lands (about 6 Mg CO2/ha
when prairie ethanol is generated on abandoned cropland and 0 when prairie
ethanol is generated on marginal land). The study however does not give values
for marginal lands converted to perennial grasses but states that degraded and
143

abandoned agricultural lands when planted with perennials “incur little or no
carbon debt and can offer immediate and sustained GHG advantages”. Since in
the present analysis, it is assumed that marginal lands (defined as lands that are
already very low in carbon reserves) are only used to grow with perennials or
native prairie/ ESV, the risk of incurring a carbon-debt is negligible.
2. Moreover, in this analysis less than 1% marginal land (based on total RIMA land
area) is assumed to be used. This makes the C-debt negligible compared to overall
C-sequestration of perennial grasses.
All values of both NEY and NGER are also calculated as a change versus the baseline i.e.
(NEY) rel = (NEY) abs – Feedstock energy inputs of changing acreages of original
landscape.
(NGER) rel = (NGER) abs – Feedstock GHG emissions of changing acreages of original
landscape,
where (NEY) abs and (NGER) abs represent the absolute NEY and NGER of the
reconfigured landscape.
To obtain absolute values, only the energy inputs of cellulosic biomass generating land
areas in the new configuration are considered. To obtain relative values, only the energy burdens
of acreages that change from crop areas to cellulosic biomass generating land areas are
subtracted from absolute values. This is done to ensure that the functional unit does not change
and that the same types of areas (changing acreages) are included in both S1 and S2. However,
the model constraints dictate the amount of changing acreages in S1 and S2. These values are
normalized by considering the total RIMA land area as the reference flow in both scenarios.

144

5.3. Results and discussion
5.3.1. Reconfigured landscapes, animal feed and environmental impacts
Multiple runs were performed on each test and scenario combination in the WORLD
model to verify that the model converges each time, to determine how consistent the results are
and to find global maxima or minima. The model ran consistently each time with no significant
errors. The results from various runs in each scenario are provided in Appendix C. Pretreated
biomass can be used either to generate ethanol or animal feed but not both. Therefore, the ratio
of animal nutritional services provided in the new reconfigured landscape vs. the old
configuration is set to be between 1 and 2. This is done since no condition is set on ethanol
production in these scenarios; the model would otherwise be driven to produce large amounts of
animal feed which is contrary to our objective of generating large amounts of ethanol. Figure 5.2
shows the feed requirements met (in the new configuration vs. original) in the various categories
of TNPDN, protein, fiber and oil and ethanol produced, as determined by the WORLD model.
The new land configurations meet or exceed all animal feed requirements compared to that
generated by the original landscape.
Figure 5.3 shows the land ratios allotted to different feedstocks in the two scenarios.
Table 5.2 contains details of various modifications to the original landscape. The model
drives the landscape to grow more miscanthus than switchgrass especially on ML in all cases
except T4 (minimizing farm GHG emissions). This is because there are differences in the
characteristics of these grasses.
According to EPIC simulations, miscanthus has 36% greater biomass yields, 33% greater
CWUE, and 14, 31 and 38% lower N and P losses and erosion, respectively. Miscanthus also

145

generates 92% greater accumulations of soil organic carbon compared to switchgrass in the EPIC

2.0

300
250

1.5
200
1.0

150

100
0.5
50
0.0

TNDPN

Cellulosic ethanol (mil. gallons)

Ratio of animal feed in new configuration vs.
original

simulations.

0

Fiber

Protein

Oil

Cellulosic Ethanol

Figure 5.2: Animal feed nutrition provided in new configuration (vs. original) and cellulosic
ethanol production

146

T4S1
T3cS1
Switch grass

T3bS1

Miscanthus
Native prairie

T3aS1

Corn
Soy

T2bS1

Alfalfa
Grass hay

T2aS1

Canola

T1S1
0%

20%

40%

60%

80%

100%

% (of total) land area occupied by each feedstock
(a)
Figure 5.3: (a) Percentage of total land area occupied by each feedstock in the “technical
potential of landscape” scenario (S1) (b) Percentage of total land area occupied by each
feedstock in the “opposition to conversion from conventional landscape” scenario (S2)

147

Figure 5.3 (Cont’d)

T4S2
Switch grass

T3cS2

Miscanthus
T3bS2

Native
prairie
Corn

T3aS2

Soy
T2bS2

Alfalfa
Grass hay

T2aS2
Canola
T1S2

0%

20%
40%
60%
80%
% (of total) land area occupied by each feedstock
(b)

148

100%

Table 5.2: Details of perennial grasses in reconfigured landscape and changes in the
original landscape
Scenario 1
Test

Scenario 2

1

2a

2b

3a

3b 3c

4

2a

2b

3a

3b 3c

4

S_ML

a

0

0

0

0

0

46 47 15

0

0

8

1

0

b

4

21

0

16

18 20 40 25 25

0

0

33 33

92

c

38

0

38

41

25 25 24

-

-

M_ML

d

100

0

100 100 24 34

e

22

28

30

f

20

19

NP_ML

g

0

h

S_PL

1

1

-

-

-

8

1

0

23

28 27 22 26 31 100 100 26 26

0

29

19

19 19 18

-

100

0

0

75

63

86

i

75

64

j

26

S_AL

M_PL

0

-

-

10 28 100 100

-

-

76 65 54 40 58

0

0

84

63 64 73

-

-

-

-

-

-

-

83

66

63 61 48

-

-

-

-

-

-

-

49

30

40

46 47 62 52 61 100 100 60 63

0

68

61

77

70

60 60 61

7

9

14

14

8

9

0

l

40

9

67

60

9

12 34

-

-

-

-

-

-

-

m

59

41

73

44

39 37 15

-

-

-

-

-

-

-

M_AL

Cor_ch

Soy_ch
For_ch
L_ch

k

Cor_AF
Soy_AF
a.
b.
c.
d.

S_ML= % of ML acres used to grow switchgrass
S_PL= % of PL acres used to grow switchgrass
S_AL= % of AL acres used to grow switchgrass
M_ML= % of ML acres used to grow miscanthus
149

-

-

-

-

86 98 100

Table 5.2 (Cont’d)
e. M_PL= % of PL acres used to grow miscanthus
f. M_AL= % of AL acres used to grow miscanthus
g. NP_ML= % of ML acres used to grow switchgrass
Where; ML = marginal land
PL = pasture land
AL = arable land
h.
i.
j.
k.

Cor_ch = % of corn land that changes into new feedstock (perennial grasses)
Soy_ch = % of soy land that changes into new feedstock (perennial grasses or canola)
For_ch= % of forage land that changes into new feedstock (perennial grasses)
L_ch = the total % of original land area that changes crops (i.e. from conventional to new
feedstocks)
l. Cor_AF = % reduction in corn acreages used for animal feed (new vs. old landscape)
m. Soy_AF = % reduction in soy acreages used for animal feed (new vs. old landscape)
Miscanthus production also generates 10% greater N2O emissions compared to
switchgrass. Furthermore, the greatest changes in landscape (from conventional to new
feedstocks) that are especially prominent in S1 translate to greater net energy yields and
environmental benefits. In all tests in scenario 1, lower acreages of corn and soybean lands are
used to grow animal feeds since a large part of animal nutrition requirements are supplied by the
pretreated cellulosic feedstocks in this scenario. Figure 5.4 shows an increase and decrease,
respectively in desirable (ΔSOC and crop water use efficiency (CWUE)) and undesirable
(erosion, N and P losses, farm GHG emissions (FGE)) environmental factors compared to the
original landscape. The Y axis in this figure represents percentage change (either increase or
decrease) compared to baseline- the original landscape.
The FGE category is seen as more rounded than peaked in the figure since it is the only
category with comparable changes from baseline in both scenarios 1 and 2. Because the EPIC
results do not exhibit great differences in FGEs (combined CO2 and N2O e. emissions) between

150

different feedstock systems, the FGE category does not vary much between scenarios, so it is
seen as more rounded than sharp in the figure. Compared to scenario 2, scenario 1 shows
between 9 -20 fold increases in ΔSOC and CWUE (positive environmental effects) and between
2-4 fold decreases in detrimental impact categories as percent change from baseline). The
greatest gains in SOC are found not only in the test in which it is intended to be maximized
(T2bS1) but also in another scenario T3aS1 (minimizing N losses).
A similar result is seen with CWUE where greater percentage increases in this factor are
seen in T3aS1 and in T2bS1 (minimizing N losses and maximizing SOC respectively). Similarly,
not only are the greatest reductions in impacts such as P loss, and erosion seen in tests in which
this is the intended result, but great reductions are also seen when minimizing N losses and
maximizing ΔSOC are the intended results.
Also, significant improvements in environmental impact criteria are seen in the case
where ethanol production is maximized (T1S1). A correlation is observed between greater
changes in the original landscape (especially from corn and soybean land to perennial grasses)
and increases and decreases in beneficial and detrimental environmental impacts, respectively, in
the analysis.

151

T1S1

T4S2

T1S2

T4S1

T2aS1

T3cS2

T2aS2

T3cS1

T2bS1

T3bS2

T2bS2
T3bS1

ΔSOC

T3aS1
T3aS2
Crop water use efficiency (CWUE)
(a)

Figure 5.4: (a) Percentage increases in beneficial environmental impact categories (Scale ranges
from 0 to 120 % in the figure) (b) Percentage decreases in detrimental environmental impact
categories under various scenarios in reference to the baseline (Scale ranges from 0 to 105% in
the figure). Notation: T = test, S = scenario. The figure shows tests 1 (maximize ethanol), 2a
(minimize erosion), 2b (maximize Δ SOC), 3a (minimize N loss), 3b (minimize P loss), 3c
(maximize CWUE), 4 (minimize FGE) in combination with the two scenarios (S1- technical
potential of landscape, S2- opposition to conversion of conventional landscapes)

152

Figure 5.4 (Cont’d)

T1S1
T4S2

T1S2

T4S1

T2aS1

T3cS2

T2aS2

T3cS1

T2bS1

T3bS2

T2bS2
T3bS1

Erosion

P loss

T3aS1
T3aS2
N loss
Farm GHG emissions (FGE)

(b)
5.3.2. Net energy yields and net GHG emission reductions
As defined in Chapter 4, the “biomass multiplier” (BM) is the ratio of total cellulosic
biomass generated in each new reconfigured landscape to the cellulosic biomass currently
generated in the RIMA. In the reconfigured landscape cellulosic biomass comes from
switchgrass, miscanthus, native prairie grasses, stover, alfalfa, grass hay and buffer strips. The
current cellulosic biomass (the denominator in the biomass multiplier ratio) includes only forage
feedstocks grass hay and alfalfa since no stover is assumed to be currently harvested in RIMA

153

for either biofuel production or as a co-product (animal feed).The biomass multiplier varies from
as high as 9.9 in the greatest NEY case to 3.0 in the lowest NEY case. This implies that the new
reconfigured landscape can provide nearly 10 times as much cellulosic biomass as the original
landscape. Figure 5.5 shows the NEYs and NGERs in all cases both in absolute and relative
values (compared to original landscape as defined in the life cycle assessment section) for the
reference flow of one hectare of land area generating biomass in the RIMA. Figure 5.5 also
relates the biomass multiplier and NEYs, NGERs. The value of BM remains approximately a
constant in all tests in S2 because the landscape does not change significantly to generate
substantially greater biomass. Overall, the greater the value of BM, the greater the NEYs and
NGERs. A general trend consistent with all tests is that S2 (i.e.; opposition to conversion of
landscapes from conventional feedstocks) has lower NEYs and NGERs compared to S1. In S2
relatively more stover and forages and less perennial grasses are used to generate ethanol while
also meeting animal feed requirements.

154

60

10
9

50

7
6

30

5
4

20

3
2

10

Biomass multiplier (ratio)

8

40

NEY (GJ/ha)

(NEY)rel
(GJ/ha)

(NEY)abs
(GJ/ha)

Biomass
multiplier

1
0

0

(a)
Figure 5.5: (a) Relative and absolute net energy yields (NEY) and biomass multiplier (BM).
NEY = Energy outputs of ethanol and electricity - energy inputs of feedstock, transport and
processing modules (b) Relative and absolute net greenhouse gas emission reductions (NGER)
and biomass multiplier (BM). NGER = CO2 e. emissions displaced due to ethanol and electricity
minus CO2 e. emissions generated due to feedstock, processing and transport modules. BM is
defined as the ratio of total cellulosic biomass generated in the reconfigured landscape to that
generated in the original landscape.

155

Figure 5.5 (Cont’d)

9

7

8
7

6

6
5
5

4
4
3

3

2

2

1

(NGER)rel
(Mg CO2
e/ha)

(NGER)abs
(Mg CO2
e/ha)

1

0

Biomass multiplier (ratio)

10

8

NGER (Mg CO2 e/ha)

9

0

Biomass
multiplier

(b)
The energy generated from such minor changes to the landscape is not sufficient to
greatly exceed the energy inputs of producing, processing and transporting the feedstocks. This
is due to 78% lower biomass yields of corn stover and 52% lower yields of forages on average
compared to perennial grasses. Similarly, some NGERs in scenario 2 are low since emissions
reductions are not substantial in the reconfigured landscapes compared to the original
configuration. For example in T4S2, where the model does not generate any changes in

156

landscape (Table 1), the lowest NEY as well as the lowest NGER is observed. Table 2 shows the
ranges of relative and absolute NEYs and NGERs in the two scenarios.
Interestingly, in the ethanol production case (T1), maximizing the BM instead of directly
maximizing ethanol production in the model generates a greater NEY and NGER. Specifically,
18% greater NEY and 8% greater NGER are seen in the case where BM is maximized compared
to where ethanol production [i.e. biomass tonnage (ton) * ethanol yield (gal/ton)] is directly
maximized in T1. This is probably because ethanol yields (per kg feedstock) are assumed to be
similar in the model inputs for all feedstocks whereas there are significantly greater differences
between biomass yields of perennial grasses, especially miscanthus, compared to other cellulosic
ethanol feedstocks. Therefore, in the case where ethanol production is maximized, the model
does not drive the landscape towards any one particular feedstock. However, where BM is
maximized, the model drives the landscape towards generating larger amounts of perennial
grasses. Similarly, in the scenarios that focus on environmental impacts (such as minimizing N
losses or maximizing ΔSOC gains), the model drives the landscape towards growing more
perennial grasses due to large differences in environmental impacts of perennial grasses
compared to their conventional counterparts (such as corn).
Table 5.3 summarizes the range of NEY and NGER values observed in different
scenarios. Since the processing and transport assumptions may vary in different analyses but the
farm level assumptions remain constant for the given set of landscape configurations in this
analysis, Table 5.3 also contains NEY and NGER ranges at the farm-gate level for the different
scenarios.

157

Table 5.3: Ranges of NEY and NGER in different scenarios, ranges of NEY and
NGER at the farm gate
NEY
Relative

NGER

Absolute

--------- GJ/ha --------

Relative

Absolute

-------- Mg CO2 e./ha --------

Integrated LCA
S1
S2

31.1- 46.8 35.9-52.7

4.4-5.4

6.7-8.3

0.9-7.5

1.7-7.8

2.0-3.2

2.6-3.7

S1

34.9-53.7

39.7-59.6

5.5-7.0

7.9-9.9

S2

3.4-10.3

4.3-11.2

2.7-3.9

3.2-4.2

Farm-gate

The model also shows several significant correlations between various parameters, seen
in Figure 5.6. First, the model shows a direct positive correlation between beneficial
environmental impacts and the biomass multiplier, BM. This relationship is primarily driven by
the percentage of perennial grass land present in the landscape. Increasing the amount of
perennial grasses in the landscape leads to greater BM which in turn leads to increases in
beneficial environmental impacts and decreases in detrimental environmental impacts. NEY
increases parallel the increased biomass multiplier. This is because greater NEYs are generated
in scenarios with more perennial grasses which in turn are linked to greater BMs. Similarly, a
direct positive correlation is observed between increasing BM and increasing NGERs.

158

% change from baseline in environmental
impact

100

ΔSOC
Erosion

50

P loss
N loss

0

0

2

4

6

8

10

12

Crop water use
efficiency (CWUE)
Farm GHG emissions
(FGE)

-50

(NEY)abs (GJ/ha)
(NGER)abs (Mg CO2
e/ha)

-100

Biomass Multiplier (ratio)

Figure 5.6: Correlations between the biomass multiplier and various system parameters. ∆SOC –
change in soil organic carbon. P loss – Phosphorus loss. N loss – Nitrogen loss. (NEY)abs –
absolute values of net energy yield. (NGER)abs – absolute values of net GHG emission
reductions
The contribution of the feedstock module to the total energy inputs includes only energy
inputs of cultivating new cellulosic feedstocks grown on the landscape but not those of noncellulosic conventional croplands (corn, soy). The transport module includes transport of all nondensified raw material and pretreated densified products as well as co-products to and from
farms, LBPDs and the biorefinery. The processing module includes the energy inputs to the
LBPDs. As seen in Figure 5.7 a, in scenario 1, feedstock, processing and transport contribute
nearly equally (on average) to the total energy inputs of the overall system. In scenario 2, where
lower amounts of cellulosic feedstocks are generated in the landscape, the contributions of
159

feedstock and processing modules to the total energy inputs are lower compared to the transport
module. However, the feedstock module has consistently higher emission output contributions in
all cases as seen in Figure 5.7 b.

100%

% of total energy inputs

90%
80%

70%
60%
50%
40%
30%
20%
10%

0%

Processing (LBPDs)

Transport

Feedstock

(a)
Figure 5.7: (a) Energy burdens of various modules- feedstock, processing and transport (b)
Emission burdens of various modules- feedstock, processing and transport.

160

Figure 5.7 (Cont’d)

100%

% of total emission outputs

90%

80%
70%
60%
50%
40%
30%
20%
10%
0%

Processing (LBPDs)

Transport

Feedstock

(b)
Increased double cropping lowers the total energy co-product credit in this analysis since
the crop production energy inputs of the assumed double crop (winter wheat) management
scenario is greater (per kg of dry feedstock) than that of soy production. Since double cropping
can enhance both total biomass production and positive environmental impacts, it will be
important to understand how its energy efficiency can be improved. The biochar co-product
credit remains same in all cases since the amount of forest residue collected does not vary
between scenarios.
The number of LBPDs varies in each scenario ranging from as low as 8 LBPDs in the
low biomass production scenarios to as high as 18 in the high biomass production scenarios and
161

the depots range in average processing capacities from 600 to 750 (short) tons/day (tpd).
Biorefinery processing inputs and carbon dioxide emissions associated with processing in the
biorefinery are assumed to be zero since these requirements are covered by burning lignin to
produce steam and electricity.
5.3.3. Sensitivity analyses
Sensitivity analyses were performed on the scenario with the greatest NEY and NGER
namely, T2bS1 (maximizing ∆SOC). This was done by increasing and decreasing by 25%
parameters such as percentage of conventional animal feed (corn, soy, grass hay, alfalfa)
currently used as animal feed, ethanol and biomass yields of perennial grasses and doubling and
halving the constraint of ratio of new vs. old animal feed requirements in the categories of
TNPDN, protein, fiber and oil. Percentage differences between NEYs and NGERs obtained in
the sensitivity analyses vs. the original scenario, T2bS1 were calculated. Figure 5.8 shows all
eight cases of sensitivity analyses. A 25% decrease in perennial grass ethanol yield causes the
NEY to decrease by 58% and NGER to decrease by 27%. Since in T2aS1 perennial grasses
dominate the new landscape, a decrease in their ethanol yield leads to decreases in NEY and
NGER. For similar reasons, a 25% decrease in perennial grass biomass yields causes a 50%
decrease in NEY and a 26% decrease in NGER. A 25% increase in the amount of conventional
crops used as animal feed causes NEY to decrease by 54.5% and NGER to decrease by 32%.
This is because more pretreated cellulosic biomass has to be diverted from ethanol production to
compensate for increases in animal nutritional requirements, thereby decreasing NEY and
NGER. For similar reasons, doubling the upper limit of new/old animal nutritional requirements
(a constraint that diverts more pretreated biomass to animal feed instead of to ethanol
production) causes a 54% decrease in NEY and a 30% decrease in NGER. However, a 25%
162

increase in perennial grass biomass and ethanol yields or decreasing the amount of animal feed
requirements causes relatively less significant increases in NEY and NGER.

Half the upper limit of new to original animal
feed ratio
25% increase in perennial grass biomass yields
25% increase in perennial grass ethanol yields
25% decrease in conventional crops used as
animal feed
25% decrease in perennial grass biomass
yields
Double the upper limit of new to original
animal feed ratio
25% increase in conventional crops used as
animal feed

25% decrease in perennial grass ethanol yields
-60

-50

-40

-30

-20

-10

0

10

% ∆NGER

% ∆NEY

Figure 5.8: Results of sensitivity analyses performed in the WORLD model on Test 2b
(maximizing ∆SOC) in Scenario 1(technical potential of landscape). This test, scenario
combination was chosen to perform sensitivity analyses since this is the case where the greatest
NEY and NGER are obtained.
5.4. Conclusion
Depending on the cellulosic biomass production in different scenarios, either 970 million
liters of ethanol in the technical potential scenario or just 82 million liters of ethanol in the
opposition to conversion scenario can be generated from the same area of interest (RIMA). NEY
values in the range of 20-57 MJ/L of ethanol and NGER values in the range of 9-37 kg CO2 e. /L
of ethanol are obtained in new reconfigured landscapes within this RIMA area. Moreover,

163

maximizing cellulosic ethanol production leads to significant improvements in positive
environmental impact categories and substantial decreases in detrimental environmental impacts.
Simulations of realistic landscapes are needed to determine the potential of any region to
produce biofuels without displacing food/feed provisioning services or creating negative
environmental impacts. If the crops planted in a landscape are altered (in S1 where there is 6077% change in the land cover from conventional to new feedstocks) and sustainable management
practices (such as the use of marginal lands, reduced fertilizer use and no-till farming)
implemented, significant net energy yields may be obtained and GHG emissions may be reduced
while maintaining or potentially increasing animal feed production. Moreover, introducing
perennial grasses in the landscape will increase soil organic carbon by 120% (compared to
baseline) and also improve (by 20 -100%) other impact categories such as minimizing erosion,
improving crop water use efficiency, and reducing N and P losses and farm GHG emissions.
However, neither substantial net energy yields nor emission reductions and decreases in other
negative environmental impacts are obtained when the landscape is not changed (e.g. in scenario
S2 in which the landscape changes between 7- 14% from conventional to new feedstocks).
Additionally, minimizing environmental impacts (such as increasing ΔSOC and minimizing N
losses) leads directly to greater net energy yields and greater GHG emission reductions. The
opposite also holds true, where maximizing cellulosic ethanol production leads to lower
detrimental environmental impacts. Similarly, environmental impacts seem to be strongly linked.
Aiming to minimize one adverse impact also helps minimize others. These improvements in
environmental impacts are in turn linked to greater perennial grass production in the landscapes.
Commercial cellulosic biofuel refineries are currently being built and many more are
expected in the future as crop production and processing technologies improve. This study

164

provides a foundation for a novel approach to cellulosic biomass production that may help
resolve biomass supply and supply chain constraints as biofuel demand increases. Using this
comprehensive modeling approach, the most promising scenarios for cellulosic biofuel supply
chains can be identified. The depot network helps bridge the gap to cellulosic biofuel
commercialization by providing valuable co-products for biofuels and thereby creates more
market demand in order to establish such biofuel systems. Greater energy and environmental
benefits accrue to the overall system beyond simply improved logistics when the challenges
associated with biomass supply and processing are minimized. By implementing land efficient
techniques and by producing valuable co-products close to the feedstock operations, depot
networks can respond to local and global sustainability concerns while simultaneously increasing
feedstock value.
In summary, the scenarios presented here were chosen mainly to illustrate model
capabilities. In reality, the landscape may be somewhere between the various extremes shown
here. The WORLD model may be included in a larger-scale multi-optimization model to
determine the combined effects of energy yields and environmental impacts. Improvements can
also be made in the model with regard to calculating detailed water-quality impacts after planting
perennial grass buffer strips and simulating a true double crop (such as winter rye) in place of an
assumed double crop (winter wheat) used in the analysis. The model may also be expanded to
include economic aspects of the various modules in the overall system. Alternatively, the model
may be linked to agricultural economic models to assess profits earned by farmers under various
landscape configurations. The methodology adopted in this analysis can be applied to a wide
range of landscapes, feedstocks and flexible depots to evaluate the net energy returns and the
environmental benefits of this new approach to developing supply chains for cellulosic biofuels.

165

The WORLD model and its inputs in an integrated LCA can be applied to many regions of
interests and many different scenarios and might be further developed to serve as a decision
making tool for growers, industries or policy-makers.

166

APPENDIX C

167

5.5. Appendix C
C.1. A summary of technologies in the local biomass processing depots [LBPD]
The “base-technologies” included in all depots are pretreatment and densification,
including all pre-processing steps such as biomass handling, size-reduction, etc. The use of other
technologies in the depots is dictated by the characteristics of the landscape generating the
biomass.
The Ammonia Fiber Expansion (AFEX™) process is a pretreatment method where hot
concentrated ammonia is mixed with moistened herbaceous biomass under pressure. After the
desired treatment time is complete, the pressure is then rapidly released causing the system to
cool and the ammonia is recovered[31]. AFEX™ is used as the pretreatment method in all
processing facilities because of its apparently unique suitability within the depot arrangement.
AFEX™ adds value to biomass as a highly-digestible fiber-based animal feed, which is
economically the most important co-product from the depots [32, 33].
Mechanical compaction or densification of biomass into pellets or briquettes post3

pretreatment can increase the bulk density of cellulosic feedstocks from as low as 60 kg/m to as
3

high as 800 kg/m , significantly reducing transport, handling and storage burdens associated
with the depots [23, 34]. Coupling densification with AFEX™ pretreatment may also reduce the
cost of densification because lignin brought to the surface during AFEX™ eliminates the need
for added binders (e.g., starch or protein) or curing agents (e.g., steam) to promote binding [32,
34, 35].

168

The leaf protein concentrate (LPC) extraction process involves pulping and pressing wet
biomass to squeeze out a protein rich juice which is subsequently coagulated and dried to obtain
a high protein powder [36, 37]. LPC can be applied to any high protein cellulosic biomass such
as double crops and forage [33] to obtain a high protein animal feed supplement.
Pyrolysis is the thermochemical conversion of woody residues into biochar, producer gas
and bio-oil. Producer gas can be combusted for process heat in the LBPD and bio-char is an
important co-product that aids in soil amendment in the landscape[38]. Alternately, biochar can
be used as a boiler fuel to substitute for coal[39]. Bio-oil however is an unstable compound that
must be further processed for use as transport fuel [40].
In anaerobic digestion (AD) biogas is produced from manure and aqueous waste streams
resulting from LPC production. The remaining solid residues after AD are sent back to farms to
use as animal bedding or as a soil amendment [41].
Table C1: Acreages of different land categories in the current RIMA landscape
a

Watershed

Corn

Forage

Soy

Marginal

Buffer strips

St. Joseph

194506

42420

228603

2504

10152

Upper Grand

22701

12715

23967

115

781

Kalamazoo

153701

50344

125729

1087

7847

Black-

31707

11183

19011

231

2295

54782

31998

52274

0

2990

Macatawa
Thornapple

All acreages are shown in acres.
a. Forage includes grass hay and alfalfa. Based on Michigan agricultural statistical data,
alfalfa forms close to 75% of total forage land. Therefore, the remaining percentage of
total forage land is assumed to be grass hay.

169

b

Table C1 (Cont’d)
b. Potentially available acreages of buffer strip are determined based on river bank lengths
of major rivers present in the RIMA calculated using the BASINS model and ArcGIS and
the assumed width of buffer strips (0.015 mi). However, depending on the scenario,
conservative values of 30% (in S1: the technical potential scenario) or 10% (in S2: the
opposition to conversion of conventional landscapes scenario) are assumed to be used for
growing buffer strips.
Table C2: Feedstocks, management practices and yields simulated in the EPIC
model
Feedstock

Number of

Fertilizer

system

years

application

simulated
Switchgrass

Cropping

[N,P,K(kg/ha)]

24

Medium (60,0,0)

Switchgrass

Tillage

Avg.
a

yields(ton /ac)

No-till

4.7 - ML
5.4 - AL

Miscanthus

Miscanthus

24

Medium (60,0,0)

No-till

7.3 - ML
8.8 - AL

Native Prairie

Cool season

24

Medium (60,0,0)

No-till

4.25

24

High (175,24,34)

Chisel

2.5

12

Medium

No-till

1.6

Chisel

0.95

native prairie
Corn

b

Continuous
corn

Stover

Corn-soybean
rotation with

(110,24,34)

stover removal
b

Soy

Corn (c)-

12

High (c-135,24,34;

soybean(s)

s-0,10,0)

rotation

170

Table C2 (Cont’d)
Double crop

Corn(c)-

8

Medium (c-

soybean(s)-

Alfalfa/grassb

3.4

Chisel

1.05

No-till

6.0

70,24,34)

Alfalfa(A)-

15

corn(c) rotation

hay

No-till

0,10,0; ww-

wheat(ww)

1.03

150,24,34; s-

winter

No-till

Medium (A120,55,0; c49,21,30)

Canola

Corn(c)-

8

High (c-175,24,34;

soybean(s)-

s-0,10,0; wc-

winter

95,24,34)

canola(wc)
Riparian

Avg. of
c

buffer strips

24

Medium (60,0,0)

switchgrass
and miscanthus

ML = marginal land, AL= arable land.
a. Tons indicate short tons
b. Indicates all feedstocks present in original landscape (baseline).
c. The width of buffer strips was assumed to be about 25 m, as determined by values in
literature [42, 43].

171

Table C3: Nutritional values of feedstocks assumed in the model to fulfill animal
feed requirements
a

Nutritional values (g/g)

TNPDN

Protein

Fiber

Oil

Perennial grasses

0.616

0.014

0.819

-

Stover

0.689

0.067

-

-

Double crop, Alfalfa

0.314

0.5

-

-

Canola

0.19

0.202

-

0.42

Corn

0.793

0.094

-

-

Soy

0.268

0.4801

-

0.196

Grass hay

0.464

0.133

0.577

Alfalfa forage

0.387

0.202

0.396

(LPC)

b

a

TNPDN = Total non-protein digestible nutrients

b

LPC indicates the protein powder extracted from alfalfa

172

Table C4: Environmental impacts of feedstocks from the EPIC model
Feedstocks

∆SOC

a

(Kg

P loss

N

(Kg

loss(

b

Erosio
n (Mg

Kg

N2O

Kg

emissio

CO2/a

P/ac)

CEF (

ns(kg/a

soil/ac)

CO2/ac
N/ac)
)

c)

c)

c

GHG (
N2O
and

Water use
efficiency

(ton/ac/ML)

CO2)

Switch grass

79

1.01

11

1.63

129

2.02

79

1.01

Miscanthus

929

0.77

10

1.18

128

2.25

929

0.77

Native

-152

0.92

20

0.65

126

2.03

-152

0.92

-663

2.01

55

4.09

379

2.85

-663

2.01

-465

1.37

14

0.73

168

1.48

-465

1.37

Corn-Soy

-476

1.60

19

2.74

236

1.45

-476

1.60

Corn-soy-DC

-504

1.47

19

0.70

193

1.40

-504

1.47

Alfalfa-corn

-499

1.76

29

2.26

156

2.14

-499

1.76

Grass hay

-169

1.63

25

1.77

140

2.27

-169

1.63

Buffer strips

504

0.89

10

1.40

129

2.14

504

0.89

Canola

-190

1.60

26

1.97

271

1.48

-190

1.60

prairie (cool
season)
Continuous
Corn
Continuous
corn with
stover
removal

173

d

Table C4 (Cont’d)
a. ∆SOC = change in soil carbon (final-initial). Negative values indicate loss of carbon
which is detrimental to the soil. Perennial grasses are the only feedstocks that have a
positive ∆SOC. The native prairie in this case does not have a positive ∆SOC since it is
not an unmanaged prairie but a fertilized one. Therefore grass hay and the native prairie
have comparable ∆SOC values.
b. CEF =farm CO2 e. carbon emissions.
c. GHG emissions are calculated as: CEF+310*N2O emissions; Where 310 is the global
warming potential factor for nitrous oxide.
d. Water use efficiency is defined in this study as yield of plant product per unit of crop
water use (megalitres of water lost due to evapotranspiration)
Table C5: Energy inputs of different feedstocks from literature and agricultural
budgets
Feedstock

MJ/kg dry feedstock

Corn

1.776114

Stover

1.316816

Alfalfa/grass hay

0.772898

Switchgrass

0.311055

Miscanthus

0.145963

Native prairie

0.211205

Double crop

3.784182

Soy

2.57445

Canola

2.467036

Buffer strips

0.228509

174

Table C6: Cellulosic ethanol and electricity yields of different feedstocks assumed in
the model
Cellulosic feedstock

Ethanol yield (gal/ ton) Electricity yield (KWH/ton)

Stover

66

335

Switchgrass

66

334

Miscanthus

64

413

Native prairie

50

389

Grass hay/alfalfa(mixes)

66

294

Winter rye (double crop)

50

327

C.2. Transport calculations
LBPD to biorefinery distances were determined by combining the road network map in
the BASINS model with the Arc GIS software to calculate the shortest routes from each LBPD
to the central biorefinery. Using this method the processing depots and central refinery were
located within the RIMA. Transport distances (collection areas and radii) of raw materials
(herbaceous feedstocks, animal manure and forest slash where applicable) to LBPDs and coproducts (pretreated animal feed, LPC solids, AD sludge and biochar where applicable) from
LBPDs back to farms were determined using formulae from literature[44]. Transport distances
for finished products (pretreated biomass to biorefinery and bio-oil where applicable) were
determined using the US EPA BASINS and Arc GIS softwares. A 20% road winding factor was
applied to these calculated values to account for non-linear road networks. Densified and non
densified transport were differentiated by modifying the cargo payload (or truck capacity)

175

suitably (higher payload for densified material vs. lower for non-densified) in GREET 1.8b[30].
Formula for collection distances from farms [44]:
2

Collection area (mi ) = ß/ [δ* ρ*(1- І)* λ*640]
Where ß= biomass collected (dry tons/ yr), δ= density of feedstock acreage, ρ= % of
farmers selling feedstock, I = % of fields inaccessible, λ= biomass yield. δ is calculated for each
watershed based on acreages generating biomass (ratio of all the land area used to generate
biomass in each watershed to the total land area of the watershed) . ‘ß’ varies in all cases based
on land areas allocated to different feedstocks. ‘λ’ varies among feedstocks. In Scenario 1
(technical potential), ‘ρ’ is assumed to be 90% whereas in Scenario 2 (opposition to conversion
of conventional landscapes) ‘ρ’ is assumed to decrease significantly to 30%. ‘I’ is assumed to
remain constant in both scenarios at 10%.

176

Table C7: Transportation distances for all raw-materials and co-products
Watershed

Distance from feedstock and animal farms to LBPD

Distance from

number

and vice-versa (mi)

LBPD to
biorefinery (mi)

Scenario 1
T1

T2a

T2b

T3a

T3b

T3c

T4

All cases

1

10.75

10.99

11.60

10.79

10.89

10.87

9.65

50

2

2.34

3.40

3.53

3.77

3.30

3.52

2.64

20

3

8.65

9.08

9.79

10.21

8.97

8.94

7.49

50

4

3.13

3.87

4.57

4.45

4.06

4.03

3.34

20

5

5.34

5.84

6.47

6.29

5.58

5.80

5.30

40

Scenario 2
T1

T2a

T2b

T3a

T3b

T3c

T4

All cases

1

15.67

15.59

15.70

15.76

15.52

15.60

15.68

50

2

4.87

4.94

4.80

5.07

4.80

4.83

4.66

20

3

12.99

13.19

13.14

13.41

13.14

13.20

13.34

50

4

6.31

5.87

6.43

6.51

6.00

6.00

6.07

20

5

8.13

8.19

8.16

8.49

8.47

8.17

8.35

40

177

Table C8: Transportation energy requirements and emissions
Material transported

Units (energy-

Transport Transport

emissions)
Non-densified biomass

energy

emissions

(MJ/ton/mi - Kg

2302

0.204

734

0.003608

1511

0.03277

2.59

0.143317

CO2 e/ton/mi)
All other densified herbaceous material and co-

(MJ/ton/mi - Kg

products

CO2 e/ton/mi)

Forest slash

(MJ/ton/mi - Kg
CO2 e/ton/mi)

Bio-oil

3

(MJ/m /km - Kg
3

CO2 e/m /km)

C.3. Co-product credit calculations
Co-product credit calculations are performed based on the amount of animal feed
generated from each feedstock in the WORLD model. Similarly, co-product credit calculations
are performed for emissions displaced due to growing a less emission-intensive feedstock (such
as perennial grasses) and using this new feedstock as animal feed instead of a conventional
animal feed. Co-product credits are calculated as follows:

178

Where CPC = co-product credit for Y displacing X in MJ or kg CO2 e for energy and
emissions respectively, α = animal feed production (kg feed/kg Y), δ = displacement ratio based
on nutritional value of feedstock, τ = tonnage of animal feed (tons), ENx and ENy = Feedstock
production energy of X and Y (MJ/ton) respectively, EMx and EMy= Farming and sequestration
emission of X and Y (kg CO2 e. / ha) respectively, γ = yield of feedstock (tons/ha)
Here Y is the new animal feed product n and X is the original animal feed product being
displaced. The animal feed production value is set at 1. Displacement ratio is defined as kg X
displaced/kg Y animal feed produced based on nutritional values of X and Y.
For biochar, emission displacement credits for avoided emissions due to reduced fertilizer
use as well as due to avoided soil nitrous oxide emissions are assigned on a per acre basis [38,
45].
C.4. Results of the WORLD model consistency checks (performing multiple runs in each
scenario)
The standard error is calculated in each case as standard deviation/ [sq. root of sample size].
Although in this case all calculations were performed on the value of ethanol, these error
calculations can be performed with other desired parameters as well.

179

2.5E+08
Ethanol (gal)

2.0E+08

1.5E+08

1.0E+08

5.0E+07

0.0E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.1: Consistency check for T1S1

180

7.E+07
Ethanol(gal)
6.E+07

5.E+07
4.E+07
3.E+07
2.E+07
1.E+07

0.E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.2: Consistency check for T1S2

181

Ethanol (gal)
1.8E+08

1.6E+08
1.4E+08
1.2E+08
1.0E+08
8.0E+07
6.0E+07
4.0E+07
2.0E+07
0.0E+00

Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.3: Consistency check for T2aS1

182

4.5E+07

4.0E+07
3.5E+07

Ethanol (gal)

3.0E+07
2.5E+07
2.0E+07
1.5E+07
1.0E+07
5.0E+06

0.0E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.4: Consistency check for T2aS2

183

2.5E+08

Ethanol (gal)

2.0E+08

1.5E+08
1.0E+08
5.0E+07

0.0E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.5: Consistency check for T2bS1

184

4.5E+07
4.0E+07

Ethanol (gal)

3.5E+07
3.0E+07
2.5E+07

2.0E+07
1.5E+07
1.0E+07
5.0E+06

0.0E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.6: Consistency check for T2bS2

185

2.5E+08
Ethanol (gal)
2.0E+08

1.5E+08

1.0E+08

5.0E+07

0.0E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.7: Consistency check for T3aS1

186

7.E+07
6.E+07
5.E+07
Ethanol (gal)
4.E+07
3.E+07

2.E+07
1.E+07

0.E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run
10
Figure C.8: Consistency check for T3aS2

187

1.8E+08

Ethanol (gal)

1.6E+08
1.4E+08

1.2E+08
1.0E+08
8.0E+07
6.0E+07
4.0E+07

2.0E+07
0.0E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.9: Consistency check for T3bS1

188

4.5E+07
4.0E+07

Ethanol (gal)

3.5E+07
3.0E+07
2.5E+07

2.0E+07
1.5E+07

1.0E+07
5.0E+06
0.0E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.10: Consistency check for T3bS2

189

1.80E+08

Ethanol (gal)

1.60E+08
1.40E+08
1.20E+08
1.00E+08

8.00E+07
6.00E+07
4.00E+07
2.00E+07
0.00E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10
Figure C.11: Consistency check for T3cS1

190

5.0E+07
Ethanol (gal)

4.5E+07
4.0E+07

3.5E+07
3.0E+07
2.5E+07
2.0E+07
1.5E+07

1.0E+07
5.0E+06
0.0E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.12: Consistency check for T3cS2

191

1.8E+08
Ethanol (gal)

1.6E+08
1.4E+08
1.2E+08
1.0E+08
8.0E+07
6.0E+07

4.0E+07
2.0E+07
0.0E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.13: Consistency check for T4S1

192

3.5E+07
Ethanol (gal)

3.0E+07

2.5E+07
2.0E+07
1.5E+07
1.0E+07
5.0E+06

0.0E+00
Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 Run 9 Run 10

Figure C.14: Consistency check for T4S2

193

APPENDIX D

194

5.6. Appendix D
D.1. Instructions to run the WORLD model and perform the integrated LCA

D.1.1. Running the WORLD model (MS Excel):

1. Click on the "initialize" macro button on model sheet
2. Select "data" tab> "solver", click on the "solve" button (If constraints need to be altered
this can be done in the solver window before running the model)
3. If the model does not converge and/or solution is not found go to step 1
4. If model converges and/or runs well go to step 5
5. Select "Home">"Insert">"Insert sheet" or select an empty existing sheet. Rename the
sheet as "Sheet1" (It is very important to rename the sheet in order for macros to run)
6. Click on the "consistency runs" macros button on model sheet. (Note: consistency macros
will not run if there is no sheet named "Sheet1")
7. Sheet 1 will now populate. Now rename "Sheet1" as "Run n"(where n is the number of
your consistency check run e.g. Run 1 or Run 2, etc.)
8. Make sure a new sheet called “Sheet1” is inserted in the excel workbook
9. Run each scenario multiple number of times (10- 20) following the instructions sheet in
the WORLD model.
10. Make sure each run within the optimization scenario is consistent (in ethanol production
range, number of LBPDs etc.)
11. Once all runs are complete, choose a single run from the results of which the integrated
LCA is to be performed. Then copy and paste these results into a new sheet and rename it
“Run”. This sheet now contains values of land allocations in different watersheds, ethanol
195

and electricity yields, animal feed fractions, number and capacity of LBPDs and absolute
values of environmental impacts as well as their relative percentage changes compared to
the original landscape. This is helpful if modifications are to be made to the model and to
summarize and compare results from various scenarios.

D.1.2. Conducting the integrated LCA:
1. All the following operations are performed in the sheet named “I-O sheet” in the
WORLD model spreadsheet
2. Obtain biomass values of different types of feedstocks from the sheet named “Run”
(Note: In the WORLD these numbers are in short tons. In the LBPD model they have to
be entered in metric tons. Therefore a conversion is required). Enter these values as input
in the input sheet of the LBPD technical model. The summary sheet in the LBPD
technical model provides diesel, electricity and natural gas requirements for each LBPD
in each watershed (already determined in WORLD). For the purpose of calculations, a
single LBPD can be assumed as a model LBPD in a particular watershed and the energy
inputs of this LBPD can be multiplied with the total number of LBPDs in that watershed
for total energy requirements and emissions from distributed processing in that
watershed. In this way, determine energy and environmental inputs of distributed
processing in all watersheds for the scenario.
3. The amounts of different co-products produced in LBPDs are also obtained from the
technical model. Copy these values into the “Transport” sheet of the WORLD model and
assume 10% dry matter losses.

196

4. Calculate total energy and environmental inputs of transport of all raw materials and coproducts using the methodology described in Appendix C and the values in Tables C7
and C8.
5. Calculate energy and environmental inputs for agriculture for all cellulosic feedstock
acreages and for unchanging conventional feedstock (corn, soy, grass hay and alfalfa
acreages) determined by the WORLD model for the new landscape configuration based
on values in Tables C4 and C5.
6. Obtain ethanol generated and electricity values from results of the WORLD model and
calculate their energy outputs (in MJ or GJ)
7. Perform co-product credit calculations as described in the methodology in Appendix C
8. Consolidate all inputs and outputs for energy and emissions to determine absolute and
relative values of NEY and NGER as described in Chapter 5. Environmental impacts
(absolute and percentage changes from original configuration (baseline) are already
present in the sheet named “Run”)
9. Summarize all results in a separate workbook for all tests and scenarios for the purposes
of comparison.

197

Table D.1: List of variables and their interpretation in the WORLD model
Variable name

Interpretation

SG (ML)

Switchgrass on marginal land

SG (PL)

Switchgrass on pasture land

SG (AL)

Switchgrass on arable land

M (ML)

Miscanthus on marginal land

M(PL)

Miscanthus on pasture land

M(AL)

Miscanthus on arable land

NP(M)

Native prairie on marginal land

A (PL)

Alfalfa on pasture land

Grass hay (PL)

Grass hay on pasture land

Corn (AL)

Corn on arable land

Soy (AL)

Soy on arable land

DC(AL)

Double crop on arable land

Canola(AL)

Canola on arable land

FeedFrac(SG)

Fraction of switchgrass to animal feed

FeedFrac(M)

Fraction of miscanthus to animal feed

FeedFrac(s)

Fraction of stover to animal feed

FeedFrac(dc)

Fraction of double crop to animal feed

FeedFrac(A)

Fraction of alfalfa to animal feed

FeedFrac(corn)

Fraction of corn to animal feed

FeedFrac(Soy)

Fraction of soy to animal feed

FeedFrac(grass hay)

Fraction of grass hay to animal feed
198

Table D.1 (Cont’d)

FeedFrac(canola)

a

Fraction of canola to animal feed

FeedFrac(np)

Fraction of native prairie to animal feed

a. Although, in the list of variables, the variable for canola is set up as an animal
feed fraction, it is only used to replace the oil that is displaced from soy
D.2. Macros
Multiple runs have to be performed on the model for each optimization test and scenario
in order to check for consistency, model robustness and to ensure global maxima and minima.
However, it is a laborious process to enter individual values of model variables as well as to copy
and paste the results from each run every time to compare results and to select a run for the
integrated LCA. Hence these macros were created to simplify the process. The following macros
are examples of only a particular optimization and need to be modified or adjusted as required.
D.2.1. The “Initialize” macro

In order to ensure that the model converges from any starting point, it is necessary to initialize
model variables for each single run using random numbers. Instead of entering a random number
for each individual model variable, the initialize macro was created to copy random numbers
[created in a separate sheet using the formula RAND()] and paste them in the cells reserved for
model variables. The subroutine for this macro in Visual Basic is as follows:
Sub Initialize()
'
' Initialize Macro
' Initializing sequence with random numbers

199

'

'
Sheets("Random numbers").Select
ActiveWindow.SmallScroll Down:=-9
Range("A1:A13").Select
Application.CutCopyMode = False
Selection.Copy
Sheets("model_TS 1").Select
Range("T2").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
ActiveWindow.SmallScroll Down:=6
Range("T20").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
ActiveWindow.SmallScroll Down:=12
Range("T38").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
ActiveWindow.SmallScroll Down:=30
Range("T56").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _

200

:=False, Transpose:=False
ActiveWindow.SmallScroll Down:=12
Range("T74").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
ActiveWindow.SmallScroll Down:=12
Sheets("Random numbers").Select
Range("B1:B5").Select
Application.CutCopyMode = False
Selection.Copy
Sheets("model_TS 1").Select
ActiveWindow.SmallScroll Down:=-78
Range("T15").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
Range("W15").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
ActiveWindow.SmallScroll Down:=18
Range("T33").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
Range("W33").Select

201

Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
ActiveWindow.SmallScroll Down:=21
Range("T51").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
Range("V51").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
ActiveWindow.SmallScroll Down:=21
Range("T69").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
Range("W69").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
ActiveWindow.SmallScroll Down:=12
Range("T87").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
Range("W87").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False

202

End Sub
D.2.2. The “Consistency runs” macro

After running the model (using solver), the consistency macro was created to automatically copy
and paste all important values from the model sheet such as land area allocations of each
watershed, the combined environmental impacts of these land areas and ethanol, electricity and
animal feed yields into a separate sheet. The subroutine for this macro in Visual Basic is as
follows:
Sub consistency()
'
' consistency Macro
'

'
ActiveWindow.ScrollColumn = 7
ActiveWindow.ScrollColumn = 6
ActiveWindow.ScrollColumn = 5
ActiveWindow.ScrollColumn = 4
ActiveWindow.ScrollColumn = 3
ActiveWindow.ScrollColumn = 2
ActiveWindow.ScrollColumn = 1
Range("A10:E22,I10:I22,Q10:Q22").Select
Selection.Copy

203

Sheets("Sheet1").Select
Range("A1").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
Sheets("model_TS 1").Select
ActiveWindow.SmallScroll Down:=18
Range("A30:E42,I30:I42,Q30:Q42").Select
Application.CutCopyMode = False
Selection.Copy
Sheets("Sheet1").Select
Range("A15").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
ActiveWindow.SmallScroll Down:=12
Sheets("model_TS 1").Select
ActiveWindow.SmallScroll Down:=18
Range("A50:E62,I50:I62,Q50:Q62").Select
Application.CutCopyMode = False
Selection.Copy
Sheets("Sheet1").Select
Range("A29").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False

204

ActiveWindow.SmallScroll Down:=15
Sheets("model_TS 1").Select
ActiveWindow.SmallScroll Down:=18
Range("A70:E82,I70:I82,Q70:Q82").Select
Application.CutCopyMode = False
Selection.Copy
Sheets("Sheet1").Select
Range("A43").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
ActiveWindow.SmallScroll Down:=18
Sheets("model_TS 1").Select
ActiveWindow.SmallScroll Down:=24
Range("A90:E102,I90:I102,Q90:Q102").Select
Application.CutCopyMode = False
Selection.Copy
Sheets("Sheet1").Select
Range("A57").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
Sheets("model_TS 1").Select
ActiveWindow.SmallScroll Down:=18
Range("A111:D116").Select

205

Application.CutCopyMode = False
Selection.Copy
Sheets("Sheet1").Select
Range("H22").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
Sheets("model_TS 1").Select
ActiveWindow.ScrollColumn = 2
ActiveWindow.ScrollColumn = 3
ActiveWindow.ScrollColumn = 4
ActiveWindow.ScrollColumn = 5
ActiveWindow.ScrollColumn = 6
ActiveWindow.ScrollColumn = 7
ActiveWindow.SmallScroll Down:=-87
ActiveWindow.SmallScroll ToRight:=2
Sheets("model_TS 1").Select
ActiveWindow.SmallScroll Down:=-6
Range("L2:N3").Select
Application.CutCopyMode = False
Selection.Copy
Sheets("Sheet1").Select
Range("I1").Select
Selection.PasteSpecial Paste:=xlPasteValuesAndNumberFormats, Operation:= _

206

xlNone, SkipBlanks:=False, Transpose:=False
Sheets("model_TS 1").Select
Range("O1:Q1").Select
Application.CutCopyMode = False
Selection.Copy
Sheets("Sheet1").Select
Range("I3").Select
Selection.PasteSpecial Paste:=xlPasteValuesAndNumberFormats, Operation:= _
xlNone, SkipBlanks:=False, Transpose:=False
ActiveWindow.ScrollWorkbookTabs Position:=xlFirst
Sheets("model_TS 1").Select
Range("Q3:Q7").Select
Application.CutCopyMode = False
Selection.Copy
ActiveWindow.ScrollWorkbookTabs Position:=xlLast
Sheets("Sheet1").Select
Range("I5").Select
Selection.PasteSpecial Paste:=xlPasteValues, Operation:=xlNone, SkipBlanks _
:=False, Transpose:=False
Sheets("model_TS 1").Select
ActiveWindow.ScrollColumn = 11
ActiveWindow.ScrollColumn = 12
ActiveWindow.ScrollColumn = 13

207

Range("AA1:AI21").Select
Application.CutCopyMode = False
Selection.Copy
ActiveWindow.ScrollWorkbookTabs Position:=xlLast
Sheets("Sheet1").Select
Range("N1").Select
Selection.PasteSpecial Paste:=xlPasteValuesAndNumberFormats, Operation:= _
xlNone, SkipBlanks:=False, Transpose:=False
End Sub
D.2.3. The “error” macro

In order to check for the consistency of the model and to ensure low error values, the error macro
was created in order to obtain values of the parameter from different runs in the same test and
scenario for which error is to be calculated. The subroutine for this macro in Visual Basic is as
follows:
Sub Error()
'
' Error Macro
'

'
Range("B13").Select
ActiveCell.FormulaR1C1 = "='Run 1'!R[-12]C[9]"

208

Range("C13").Select
ActiveCell.FormulaR1C1 = "='Run 2'!R[-12]C[8]"
Range("D13").Select
ActiveCell.FormulaR1C1 = "='Run 3'!R[-12]C[7]"
Range("E13").Select
ActiveCell.FormulaR1C1 = "='Run 4'!R[-12]C[6]"
Range("F13").Select
ActiveCell.FormulaR1C1 = "='Run 5'!R[-12]C[5]"
Range("G13").Select
ActiveCell.FormulaR1C1 = "='Run 6'!R[-12]C[4]"
Range("H13").Select
ActiveCell.FormulaR1C1 = "='Run 7'!R[-12]C[3]"
Range("I13").Select
ActiveCell.FormulaR1C1 = "='Run 8'!R[-12]C[2]"
Range("J13").Select
ActiveCell.FormulaR1C1 = "='Run 9'!R[-12]C[1]"
Range("K13").Select
ActiveCell.FormulaR1C1 = "='Run 10'!R[-12]C"
Range("K14").Select
End Sub

209

REFERENCES

210

5.7. REFERENCES

1.
Meehan, T.D., et al., Agricultural landscape simplification and insecticide use in the
Midwestern United States. Proceedings of the National Academy of Sciences. 108(28): p. 1150011505.
2.
Gardiner, M.M., et al., Landscape diversity enhances biological control of an introduced
crop pest in the north-central USA. Ecological Applications, 2009. 19(1): p. 143-154.
3.
Dale, B.E., et al., Biofuels Done Right: Land Efficient Animal Feeds Enable Large
Environmental and Energy Benefits. Environmental Science & Technology, 2010. 44(22): p.
8385-8389.
4.
Mathews, J.A. and H. Tan, Biofuels and indirect land use change effects: the debate
continues. Biofuels, Bioproducts and Biorefining, 2009. 3(3): p. 305-317.
5.
Zhang, X., et al., An integrative modeling framework to evaluate the productivity and
sustainability of biofuel crop production systems. GCB Bioenergy, 2010. 2(5): p. 258-277.
6.
Slater, S., K. Keegstra, and T. Donohue, The US Department of Energy Great Lakes
Bioenergy Research Center: Midwestern Biomass as a Resource for Renewable Fuels.
BioEnergy Research, 2010. 3(1): p. 3-5.
7.
Laser, M., et al., Coproduction of ethanol and power from switchgrass. Biofuels,
Bioproducts and Biorefining, 2009. 3(2): p. 195-218.
8.
Schmer, M., et al., Net energy of cellulosic ethanol from switchgrass. Proceedings of the
National Academy of Sciences, 2008. 105(2): p. 464.
9.
Hettenhaus, J., Sustainable Production of Agricultural Cellulosic Biomass: Collection of
Cereal Straw and Corn Stover as Biorefinery Feedstock. 2006, The Biotechnology Industry
Organization Industrial and Environmental Section. p. 2-32.

211

10.
Scurlock, J.M.O., Miscanthus: A Review of European Experience with a Novel Energy
Crop. Publication No: 4845. ORNL Environmental Sciences Division, 1999. (Accessed, August
10, 2012)
11.
Tilman, D., J. Hill, and C. Lehman, Carbon-Negative Biofuels from Low-Input HighDiversity Grassland Biomass. Science, 2006. 314(5805): p. 1598-1600.
12.
Florine, S.E., et al., Yield and composition of herbaceous biomass harvested from
naturalized grassland in southern Iowa. Biomass and Bioenergy, 2006. 30(6): p. 522-528.
13.
Ceotto, E., Grasslands for bioenergy production. A review. Agron. Sustain. Dev., 2008.
28(1): p. 47-55.
14.
USEPA, BASINS 4.0 (Better Assessment Science Integrating Point and Non-point
Sources). http://water.epa.gov/scitech/datait/models/basins/BASINS4_index.cfm, 2010.
(Accessed, August 10, 2012)
15.
Sharpley, A.N. and J.R. Williams, EPIC-erosion/productivity impact calculator: 1.
Model documentation. Technical Bulletin - United States Department of Agriculture, 1990.
1768: p. 1-235.
16.
Homer, C., et al., Completion of the 2001 National Land Cover Database for the
Conterminous United States. Photogrammetric Engineering and Remote Sensing, 2007. 73(4): p.
337-341.
17.
NRCS, Ecological and Interpretative Groups. National Soil Survey Handbook- Part 622,
2012. (Accessed, August 10, 2012)
18.
McBride, A.C., et al., Indicators to support environmental sustainability of bioenergy
systems. Ecological Indicators. 11(5): p. 1277-1289.
19.
Parish, E.S., et al., Multimetric spatial optimization of switchgrass plantings across a
watershed. Biofuels Bioproducts & Biorefining-Biofpr. 6(1): p. 58-72.
20.
USDA, World Agricultural Supply and Demand Estimates. World Agricultural outlook
Board, 2012. 1554-9089. (Accessed, August 10, 2012)

212

21.
ASA, Soy Stats 2011. http://www.soystats.com/2011/Default-frames.htm, 2011.
(Accessed, August 10, 2012)
22.
Dale, B.E., et al., Protein feeds coproduction in biomass conversion to fuels and
chemicals. Biofuels, Bioproducts and Biorefining, 2009. 3(2): p. 219-230.
23.
Eranki, P.L. and B.E. Dale, Comparative life cycle assessment of centralized and
distributed biomass processing systems combined with mixed feedstock landscapes. GCB
Bioenergy, 2011. 10.1111/j.1757-1707.2011.01096.x: p. 1757-1707.
24.
Egbendewe-Mondzozo, A., et al., Can Dispersed Biomass Processing Protect the
Environment and Cover the Bottom Line for Biofuel? - Staff Papers. Michigan State University,
Department of Agricultural, Food, and Resource Economics, 2011.
http://ideas.repec.org/p/ags/midasp/119348.html(119348). (Accessed, August 10, 2012)
25.
Blanco-Canqui, H., et al., Grass Barrier and Vegetative Filter Strip Effectiveness in
Reducing Runoff, Sediment, Nitrogen, and Phosphorus Loss. Soil Sci. Soc. Am. J., 2004. 68(5):
p. 1670-1678.
26.
Alfred Rankins Jr., David R. Shaw, and M. Boyette, Perennial grass filter strips for
reducing herbicide losses in runoff. Weed Science, 2001. 49(5): p. 647-651.
27.
Bals, B.D. and B.E. Dale, Developing a model for assessing biomass processing
technologies within a local biomass processing depot. Bioresource Technology, 2012. 106(0): p.
161-169.
28.
Sendich, E. and B.E. Dale, Environmental and economic analysis of the fully integrated
biorefinery. Global Change Biology Bioenergy, 2009. 1(1): p. 331-345.
29.
Humbird, D., et al., Process Design and Economics for Biochemical Conversion of
Lignocellulosic Biomass to Ethanol: Dilute-Acid Pretreatment and Enzymatic Hydrolysis of
Corn Stover. National Renewable Energy Laboratory Report, 2011. NREL/TP-5100-47764.
30.
Wang, M.Q., Development and use of GREET 1.6 full-cycle model for transportation
fuels and vehicle technologies. ANL/ES/TM-163, 2001. Washington , DC : U.S. Department of
Energy. (Accessed, August 10, 2012)
31.
Kazi, K.F., et al., Techno-Economic Analysis of Biochemical Scenarios for Production
of Cellulosic Ethanol. NREL Report No.TP-6A2-46588, 2010. (Accessed, August 10, 2012)
213

32.
Carolan, J.E., et al., Technical and Financial Feasibility Analysis of Distributed
Bioprocessing Using Regional Biomass Pre-Processing Centers. Journal of Agricultural & Food
Industrial Organization, 2007. 5(10): p. 1-29.
33.
Eranki, P.L., et al., Advanced Regional Biomass Processing Depots: a key to the
logistical challenges of the cellulosic biofuel industry. Biofuels, Bioproducts and Biorefining,
2011. 10.1002/bbb.318: p. 1932-1031.
34.
Sokhansanj, S., et al., Large-scale production, harvest and logistics of switchgrass
(Panicum virgatum L.) - current technology and envisioning a mature technology. Biofuels,
Bioproducts and Biorefining, 2009. 3(2): p. 124-141.
35.
Sokhansanj, S., et al., Binderless Pelletization of Biomass. Paper no: 056061.ASAE
Annual International Meeting, Tampa, FL.. St. Joseph, MI. (Accessed, August 10, 2012)
36.
Bals, B. and B.E. Dale, Economic comparison of multiple techniques for recovering leaf
protein in biomass processing. Biotechnology and Bioengineering, 2010. 108(3): p. 530-537.
37.
Enochian, R.V., et al., Producing Pro-Xan (Leaf Protein Concentrate) from Alfalfa:
Economies of an emerging technology. Report no. 445, 1980. USDA Agricultural Economics,
Washington DC, USA
38.
Gaunt, J.L. and J. Lehmann, Energy Balance and Emissions Associated with Biochar
Sequestration and Pyrolysis Bioenergy Production. Environmental Science & Technology, 2008.
42(11): p. 4152-4158.
39.
Hu, Z.W., et al., Substitution of Biomass for Coal and Coke in Ironmaking Process.
Advanced Materials Research 2011. 236-238,77.
40.
Zhang, Q., et al., Review of biomass pyrolysis oil properties and upgrading research
Energy conservation and management, 2007. 48(1): p. 87-92.
41.
Balsam, J. and D. Ryan, Anaerobic Digestion of Animal Wastes: Factors to Consider.
ATTRA- National Sustainable Agriculture Information service, 2006. IP-219: p.
http://attra.ncat.org/attra-pub/anaerobic.html. (Accessed, August 10, 2012)
42.
Liu, X., X. Zhang, and M. Zhang, Major Factors Influencing the Efficacy of Vegetated
Buffers on Sediment Trapping: A Review and Analysis. J. Environ. Qual., 2008. 37(5): p. 16671674.
214

43.
Gopalakrishnan, G., et al., Biofuels, Land, and Water: A Systems Approach to
Sustainability. Environmental Science & Technology, 2009. 43(15): p. 6094-6100.
44.
Perlack, R. and A. Turhollow, Assessment of options for the collection, handling, and
transport of corn stover. ORNL/TM-2002/44, 2002. Oak Ridge National Laboratory, Oak Ridge,
TN p. 29. (Accessed, August 10, 2012)
45.
Wood, S. and A. Cowie, A Review of Greenhouse Gas Emission Factors for Fertiliser
Production. IEA Bioenergy Task 38 2004. www.ieabioenergy-task38.org/publications/GHG
Emission_Fertilizer %20Production_July2004.pdf. (Accessed, August 10, 2012)
46.
Fargione, J., et al., Land Clearing and the Biofuel Carbon Debt. Science, 2008.
319(5867): p. 1235-1238.

215

CHAPTER 6: CONCLUSIONS AND RECOMMENDATIONS

6.1. Conclusions
Based on the comparative LCA study between distributed and centralized processing
systems on farm-scale landscapes, combining depots with high yielding perennial grasses can
achieve net energy yields comparable to those of a centralized biorefinery and with lower GHG
emissions. When modeling landscape conversions from current conventional croplands
containing annual crops such as corn to landscapes containing perennial grassland systems, it is
important to evaluate increases in (or accumulation of) soil carbon. These increases occur due to
soil carbon sequestration properties of perennial grass root masses and due to the absence of
plowing and tillage of soil each year. Low yielding perennial grasses lead to lower net energy
yields. In such cases, it is more beneficial to use other feedstocks such as corn stover along with
perennial grasses in combination with depots. Densification following pretreatment is crucial in
the successful establishment of distributed biomass processing systems since it is a major factor
in shrinking the transport and storage footprints of the overall system. Moreover, pretreated and
densified biomass has the potential to become a commodity that can be used as an end-product
(as animal feed) as well as raw material for multiple processes and industries such as biofuel
production. Products generated from biofuel landscapes such as pretreated biomass and solid
protein extracted from certain feedstocks such as double crops may satisfy animal feed
requirements for digestible nutrients, protein and fiber.
Implementing land-efficient techniques and producing multiple co-products can help
address sustainability concerns of biofuel production systems and increase the value of
feedstocks. Some of these techniques include, no-till farming, using marginal/idle lands to grow

216

low-input perennial grasses or native prairies, double cropping on the same agricultural land
areas used to generate conventional crops and to produce animal protein supplements, recycling
wastes within the different modules of the system and forming positive feedback loops between
different operations in the system (e.g. using animal manure in anaerobic digestion to generate
energy for different processes in the depot). Without bringing more land into production and by
utilizing idle land, landscapes can be modified to generate as much as ten-fold greater biomass
output. Overall, perennial grasses are inextricably tied with sustainable biofuel production. In
order to achieve substantial net energy yields and environmental benefits the landscape must
contain perennial grasses.
Depots can act as a link between the biorefining industry and farm operations. They can
standardize feedstocks with variable properties and streamline the logistics of biofuel production.
Enhanced depots (depots containing technologies in addition to pretreatment) can prove to be
more advantageous as they can better adapt to the characteristics of landscapes and can be
customized to include technologies on an as-needed basis and also since there are mutually
beneficial synergies among these technologies.
A comprehensive, integrated modeling and assessment approach is required to make
realistic and thorough evaluations of the biofuel production system. A sequential procedure is
necessary starting from land area availability and feedstock production to technology
requirements, co-product production and transport burdens for all raw-materials and products
included in the system. The WORLD model deals with the feedstock production and
sustainability aspects of the feedstock system whereas the integrated LCA ties the WORLD
model to the processing and transport modules. As a whole, this integrated analysis can be a
useful tool to evaluate biofuel production systems in real landscape settings.
217

6.2. Recommendations
Based on the results of this project, suggestions for further analysis in different categories
of interest are as follows:


Economics: Integrated LCAs give insight into the energy yields and environmental
aspects of distributed processing systems and shed light on the social aspects of the
overall biofuel production system. However, it is essential to assess the third facet of
sustainability in the form of economic analyses. A comprehensive economic model for
the combined feedstock-transport-processing system is needed. An economic component
should be built into the WORLD model based on agricultural budgets for existing crops
and literature values for costs of farming operations and projected prices for new
feedstocks such as perennial grasses. In the future, when carbon sequestration systems
receive economic incentives or when prices are set for co-products generated in the
depots, these components should be integrated into the economic model in order to
determine the overall economic and environmental benefits of biofuel production
systems.



Transport: This analysis primarily assumes the transport of raw materials and products
via trucks. However, other modes of transportation such as rail and barge should also be
taken into consideration and compared. Just as dry matter losses, road winding factors
and distances of transport from real road networks in GIS based systems were considered
in this analysis; realistic data should be used for future analysis of other transportation
systems wherever applicable. Moreover, energy and environmental burdens of
transportation of densified material were obtained from the GREET model. It may be
advantageous to compare these values with data from other models that are currently
218

being developed such as the Integrated Biomass Supply Analysis and Logistics (IBSAL)
model [1].


Biogeochemical modeling data: Due to a lack of comprehensive data for relatively new
feedstocks such as some perennial grasses, certain assumptions were made in this
analysis and literature values were used to fill these data gaps. However, as
biogeochemical modeling for these new feedstocks improves, the data used in the
analyses should be updated. For example, simulations of water quality improvements in
water-bodies on the banks of which riparian buffer strips are grown should be performed
and these results should be incorporated in environmental analyses. Similarly, data from
simulations of double crops such as winter wheat, rye and hairy vetch (Vicia villosa) are
required. Moreover, simulations of feedstocks systems other than switchgrass and
miscanthus such as energy sorghum (Sorghum bicolor) are needed so that the data from
these simulations can be used in integrated analyses. Simulation of forest systems is also
needed in order to obtain data for woody residues that are used in pyrolysis. The
possibility of incurring “carbon-debts” [8] must be analyzed and incorporated in
calculations while converting from one land-use to another (especially if there is a
conversion of marginal/abandoned/forage land to conventional crops such as corn). A
statistical analysis of data generated from biogeochemical models (usually containing
large data-sets) must be performed to assess errors. It may also be beneficial to compare
simulation data from different biogeochemical models such as EPIC, DayCent and
Biome-BGC [2] in order to standardize and validate results of these relatively new
feedstocks.

219



The WORLD model and integrated life cycle assessments: This analysis includes the
only a nascent stage of the WORLD model and primarily illustrates its capabilities by
simulating landscapes for a single region of interest- the Michigan RIMA. The model
should be applied to different regions of interest, for example, starting with the second
RIMA in central Wisconsin, under investigation in the modeling group of the GLBRC.
The model should then be extended to a completely different watershed in the United
States where cellulosic feedstocks are grown that are not dealt with in this analysis (such
as sugarcane bagasse or poplar) and with different soil and climatic conditions; so that the
model might be further tested and validated. Although the WORLD model assesses fuel,
soil, water and GHG emissions, the model should also be able to assess biodiversity and
ecosystem services. Using individual categories of sustainability measurement, for
instance biological control of insects (such as the soybean aphid) [3] or avian populations
or using an index that combines various categories of biodiversity, the overall ecosystem
services provided by biofuel feedstock landscapes should be determined and incorporated
in the WORLD model. For example, Meehan et al. [4] suggested that increasing the
production of annual bioenergy crops on marginal lands decreases avian richness
between 7% and 65% whereas substituting annual crops with varied perennial bioenergy
crops may increases avian richness between 12% and 207% while also recovering species
of conservational concern. Such effects should be incorporated in sustainability analyses
within the WORLD model. The WORLD model should also be extended to perform
multi-objective optimization. For example, trying to minimize nitrogen losses while
simultaneously obtaining large energy yields. Maximizing energy gains while incurring

220

the least amount of soil erosion and also improving soil quality by amassing large
amounts of soil organic carbon.
The LCAs performed in this project are cradle-to-gate; however this may be
extended to cradle-to-grave or well-to-wheel analyses by including transportation,
blending and end-use of the ethanol and other co-products generated. Also, taking into
account the nature of biofuels, Eco-LCAs (a framework to account for the role of
ecosystem goods and services in the life cycle of economic activities) [5, 6] may also be
appropriate for biofuel systems. Although comprehensive LCAs can be performed
independently, it is recommended that a LCA software platform such as GaBi [7] be used
for future assessments in order to standardize biofuel LCAs within different modeling
groups in the GLBRC. Using GaBi will also facilitate data exchange and comparison of
results using the same databases and system boundaries. All modules in the biofuel
production system should be built in GaBi for thorough assessments and scenario
analyses.

221

REFERENCES

222

6.3. REFERENCES

1.
Kumar, A.; Sokhansanj, S., Switchgrass (Panicum vigratum, L.) delivery to a biorefinery
using integrated biomass supply analysis and logistics (IBSAL) model. Bioresource Technology
2007, 98, (5), 1033-1044.
2.
Thornton, P. E.; Running, S. W., User’s Guide for Biome-BGC, Version 4.1.2 National
Center for Atmospheric Research, 2002, Numerical Terradynamic Simulation Group , University
of Montana. (Accessed, August 10, 2012)
3.
Landis, D. A.; Gardiner, M. M.; van der Werf, W.; Swinton, S. M., Increasing corn for
biofuel production reduces biocontrol services in agricultural landscapes. Proceedings of the
National Academy of Sciences 2008, 105, (51), 20552-20557.
4.
Meehan, T. D.; Hurlbert, A. H.; Gratton, C., Bird communities in future bioenergy
landscapes of the Upper Midwest. Proceedings of the National Academy of Sciences. (Accessed,
August 10, 2012)
5.
Bakshi, B.; Small, M. J., Incorporating Ecosystem Services Into Life Cycle Assessment.
Journal of Industrial Ecology 15, (4), 477-478.
6.
Ecologically-Based Life Cycle Assessment (Eco-LCA). Ohio State University, Center for
Resilience. http://resilience.eng.ohio-state.edu/eco-lca/index.htm. (Accessed, August 10,
2012)
7.
GaBi 4 manual, 2003. http://ocw.upm.es/ingenieria-quimica/produccion-limpiaecologia-industrial-y-desarrollo sostenible /contenido / Introduction _ to_ GaBi_ 4.
pdf,.(Accessed, August 10, 2012)
8.
Fargione, J., et al., Land Clearing and the Biofuel Carbon Debt. Science, 2008.
319(5867): p. 1235-1238.

223