environmental assessment of bio-based fuels and chemicals ... · environmental assessment of...
TRANSCRIPT
Environmental Assessment of Bio-based Fuels and Chemicals Using LCA Methodology
A Dissertation Presented
By
Mahdokht Montazeri
To
The Department of Civil and Environmental Engineering
In partial fulfillment of the requirementsFor the degree of
Doctor of Philosophy
In the field of
Environmental Engineering
Northeastern UniversityBoston, Massachusetts
May 2017
ii
ABSTRACT
Based on US EPA and DOE projections, biomass derived fuels and chemicals will supply
up to 17% and 10% of total demand for transport fuels and basic chemicals, respectively,
during the coming decade. Such large-scale production requires environmental assessment
at a systems level using tools such as life cycle assessment (LCA). This dissertation
combines chemical engineering process modeling with LCA to assess environmental
impacts of novel bio-based products and synthesis routes. Four projects of this dissertation
include, (1) a statistical meta-analysis of life cycle GHG emission and energy use results
for priority bio-based chemicals; (2) a process design and LCA analysis of a novel catalytic
depolymerization process for production of aromatics from the lignin fraction of woody
biomass; (3) an industry-sponsored assessment of the net environmental benefits of
substitution of renewable chemical building blocks in the formulation of wood flooring
coatings; and (4) evaluation of integrated fuel, energy, and chemicals production from a
microalgal biorefinery, considering time-dependent fractional growth kinetics of
freshwater and marine microalgae.
In general, assessment results of bio-based fuels and chemicals were found to be sensitive
to process and LCA model parameters, especially the choice of conversion process, co-
product allocation method, and inclusion/exclusion of emissions from land use change. Net
GHG emissions results for most sugar-derived chemicals met existing sustainability
thresholds, while thermochemical conversions routes typically did not. High-yield
conversion of lignin to catechol via catalytic depolymerization is environmentally
preferable, when coupled with upstream process modifications such as use of lignin-rich
iii
sources and recovery/substitution of chlorinated solvents and ozone-depleting substances.
Such a modified pathway showed 6%-80% reduction in impacts, compared to fossil-based
catechols. For use of renewable building blocks in industrial coating formulations,
substitution of corn-derived chemicals with identical chemicals derived from corn stover
reduced impacts by more than 50% across impact categories, primarily due to reductions
in on-field emissions. Finally, time-dependent microalgal biorefinery designs were
optimized through simultaneous consideration of on-site energy production and protein
recovery, in addition to conventional lipid-derived biofuel. Overall, this dissertation
develops novel LCA modeling methods and provides guidance for bio-based product
design and development and policy.
iv
ACKNOWLEDGEMENT
First and foremost, I would like to express my gratitude to my advisor, Dr. Matthew J.
Eckleman, for his guidance during the past four years. Undoubtedly, this research would
not have been completed without his assistance. I would also like to extend my appreciation
to my doctoral committee members, Dr. Matthias Ruth, Dr. Annalisa Onnis-Hayden, and
Dr. Richard West for their help and recommendations upon completion of this dissertation.
I would like to thank amazing members of my research group, all former and current PhD
students. Their feedbacks, cooperation and of course friendship helped me to be a better
scientist and a better human being.
My deepest gratitude to my mother, Manije Karajibani, and my sisters, Mahboubeh
Montazeri and Mahshid Montazeri who stood by my side every step of this journey. Their
love and support make everything possible for me. Finally, I would like to dedicate this
dissertation to the most supportive man in my life, my beloved father who is not amongst
us today, Ali Montazeri.
This dissertation was supported by multiple grants from USEPA (award FP-91717301-0)
USDA (award NIFA-2010-38202-21853) and NSF CAREER award (Grant No. CBET-
1454414).
v
TABLE OF CONTENT
Chapter 1: .......................................................................................................................... 1Introduction to Bio-based Products and Their Environmental Assessment ....................... 1
1.1. Introduction to bio-based products....................................................................... 11.1.1. Market, demand and application of bio-based products ............................... 11.1.2. Environmental Implications of bio-based products .................................... 10
1.2. Life Cycle Assessment (LCA) of bio-based products........................................ 121.2.1. Introduction to Life Cycle Assessment (LCA) ............................................... 121.2.2. Life cycle impact assessment (LCIA) methods .............................................. 161.2.3. Gaps and Challenges................................................................................... 22
1.3. Motivation and Summary of Chapters ............................................................... 24Chapter 2: ........................................................................................................................ 29Meta-Analysis of Life Cycle Energy and Greenhouse Gas Emissions for Priority Bio-based Chemicals................................................................................................................ 29
2.1. Introduction ........................................................................................................ 302.2. Methods.............................................................................................................. 352.3. Results and Discussion....................................................................................... 44
Chapter 3: ........................................................................................................................ 61Life Cycle Assessment of Catechols from Lignin Depolymerization .............................. 61
3.1. Introduction ........................................................................................................ 623.2. Methods.............................................................................................................. 69
3.2.1. Goal and Scope ............................................................................................... 693.2.2. Process Description......................................................................................... 703.2.3. Catalyst Preparation ........................................................................................ 733.2.4. ASPEN Plus Simulations................................................................................ 743.2.5. Life Cycle Inventory ....................................................................................... 773.2.6. Alternate Extraction Processes ....................................................................... 773.2.7. Life Cycle Assessment.................................................................................... 78
3.3. Results and Discussion....................................................................................... 803.3.1. Solvent Waste Treatment ................................................................................ 843.3.2. Alternate Lignin Extraction Method............................................................... 853.3.3. Alternate Lignin Source.................................................................................. 863.3.4. Uncertainty and Additional Considerations.................................................... 88
Chapter 4: ........................................................................................................................ 91
vi
Life Cycle Assessment of UV-Curable Biobased Wood Flooring Coatings .................... 914.1. Introduction ........................................................................................................ 924.2. Methods.............................................................................................................. 97
4.2.1. Goal and Scope ............................................................................................... 974.2.2. Life Cycle Inventory ..................................................................................... 1004.2.3. Life Cycle Impact Assessment...................................................................... 102
4.3. Results and Discussion..................................................................................... 103Chapter 5: ...................................................................................................................... 112Evaluating Microalgal Integrated Biorefinery Schemes: Empirical Controlled Growth Studies and Life Cycle Assessment ................................................................................ 112
5.1. Introduction ...................................................................................................... 1125.2. Materials and Methods ..................................................................................... 118
5.2.1. Chemicals and materials: .......................................................................... 1185.2.2. Algal Growth Experiments: ...................................................................... 1185.2.3. Algal Sampling and Harvesting: ............................................................... 1195.2.4. Extraction and Analyses: .......................................................................... 1205.2.5 Life Cycle Assessment:.................................................................................. 121
5.3. Results and Discussion..................................................................................... 1245.3.1. Algal Growth and Composition:............................................................... 1245.3.2. Fatty Acid Methyl Ester Content and Composition:................................. 1255.3.3. Biochemical compositions: Lipid, protein, starch: ................................... 1295.3.4. Life Cycle Assessment: Energy consumption, greenhouse gas emissions, and eutrophication potential:................................................................................... 1315.3.5. Implications for Microalgal Integrated Biorefinery Schemes .................. 136
5.4. Conclusions ...................................................................................................... 137REFERENCES.............................................................................................................. 138APPENDIX A ................................................................................................................ 157APPENDIX B ................................................................................................................ 178APPENDIX C ................................................................................................................ 191APPENDIX D ................................................................................................................ 195
vii
LIST OF TABLES
Table 1- List of major bio-based products, their producers and market size ..................... 8
Table 2- Literature sources for life cycle energy use and GHG emission results ............ 36
Table 3- ANCOVA and ANOVA summary results for bio-based chemicals meta-data. 54
Table 4- 1 Way Analysis of Variance (ANOVA) for factor, ‘Conversion Platform’ for
response variable absolute greenhouse gas emissions ...................................................... 56
Table 5- 1 Way Analysis of Variance (ANOVA) for factor, ‘LCA Coproduct Handling
Method’ for response variable relative non-renewable energy use .................................. 56
Table 6- Global lignin resources and current production/cultivation levels .................... 64
Table 7- Design parameters for alternate lignin extraction methods ............................... 78
Table 8- Summary of products and allocation methods................................................... 79
Table 9- Total environmental burden of lignin-based and petroleum-based TBC .......... 84
Table 10- Potential catechol production from different resources ................................... 87
Table 11- Relative LCA of BRC wood flooring coating compared to control UV-cured
coatings (per m2 of coating) ............................................................................................ 104
Table 12- Conditions of algal cultures at harvest on day 8/9 during exponential growth
phase for four species (two freshwater and two marine) in nitrogen deplete and replete
conditions. Uncertainty values represent standard error between triplicates.................. 125
Table 13- Lipid profiles of N. oleoabundans (Neo), C. sorokiniana (Chl), N. oculata (Nan),
and T. suecica (Tet) grown under nitrogen replete and deplete conditions. The lipid profiles
of other established biofuel feedstocks from (Moser, 2008) are included for comparison.
......................................................................................................................................... 128
viii
LIST OF FIGURES
Figure 1- Biomass resources, conversionn and products (adapted from (eXtension, 2013))
............................................................................................................................................. 2
Figure 2- Potential building blocks from processing biomass (partially adapted and
modified from (Werpy et al., 2004))................................................................................... 6
Figure 3- Life cycle of a product (adapted from (Rebitzer, 2002)) ................................. 14
Figure 4- Percent change in life cycle GHG emissions of (a) chemicals derived from
carbohydrate content of corn feedstock, (b) chemicals from lignin content of biomass
feedstocks, and (c) chemicals derived from carbohydrate content of non-corn feedstocks,
compared to their petrochemical counterparts. Dashed lines present GHG reduction
thresholds for each category compared to the fossil-based counterparts. Note: the range
shown in each figure represents relative GHG values with negative numbers indicating
GHG emissions reductions and positive numbers indicating GHG emissions increases. 46
Figure 5- Relative NREU values for (a) chemicals derived from sugar content of corn
feedstock, (b) chemicals derived from sugar content of non-corn feedstocks and (c)
chemicals derived from lignin content of non-corn feedstocks, compared to their petroleum
counterparts. Note: the range shown in each figure represents relative GHG values with
negative numbers indicating GHG emissions reductions and positive numbers indicating
GHG emissions increases. ................................................................................................ 51
Figure 6- Life cycle energy use (NREU, CED and fossil fuel input) vs. GHG emissions
for bio-based chemicals .................................................................................................... 52
Figure 7- Lignin polymer and three main monomers (adapted from http://www.ir
nase.csic.es)....................................................................................................................... 62
Figure 8- Process flow chart of bio-based production route............................................ 73
Figure 9- ASPEN Plus process flow diagrams for (a) catalyst synthesis and (b) lignin
depolymerization............................................................................................................... 75
Figure 10- Flow diagram of petroleum-based TBC......................................................... 77
Figure 11- Process contribution for 1 kg TBC production, considering nuts cultivation and
preparation, lignin extraction and catalytic depolymerization, and catalyst synthesis ..... 81
Figure 12- Process contribution of TBC production from petroleum based phenol........ 82
Figure 13- Process environmental burden considering different extraction method ....... 86
ix
Figure 14- System boundary for 1 m2 of control and BRC coatings ............................. 100
Figure 15- Contribution of layers and UV-curing process in environmental impacts of (a)
BRC and (b) control coatings ......................................................................................... 106
Figure 16- Life cycle comparison between layers of BRC and control coatings .......... 108
Figure 17- Mass flows through life cycle stages included in the scope of the study as
described by A) and detailed for each growth scenario (species/N-loading) in B) where for
N. oleoabundans (Neo), C. sorokiniana (Chl), N. oculata (Nan), and T. suecica (Tet) under
N-deprived (-) and N-replete (+) growth conditions....................................................... 122
Figure 18- FAME content and productivity of algal species, N. oleoabundans (Neo), C.
sorokiniana (Chl), N. oculata (Nan), and T. suecica (Tet), with nitrate replete (solid
symbols) and nitrate deprived (outlined symbols) growth conditions. Error bars represent
standard error between experimental replicates.............................................................. 126
Figure 19- Fatty acid methyl ester profile of lipid extracts for N. oleoabundans (Neo), C.
sorokiniana (Chl), N. oculata (Nan), and T. suecica (Tet) under N-deprived (N-) and N-
replete (N+) growth conditions....................................................................................... 128
Figure 20- Lipid, protein, and starch profiles (as percent dry mass) of N. oleoabundans
(Neo), C. sorokiniana (Chl), T. suecica (Tet), and N. oculata (Nan).............................. 131
Figure 21- Life cycle impacts for GHG emissions, eutrophication, and primary energy use
per kg of biodiesel for N-replete and N-deplete growth conditions ............................... 135
1
Chapter 1:
Introduction to Bio-based Products and Their Environmental Assessment
1.1. Introduction to bio-based products1.1.1. Market, demand and application of bio-based products
Bio-based products, as defined by the United States Secretary of Agriculture in the Farm
Security and Rural Investment Act of 2002, are commercial or industrial products, other
than food and feed, that are composed of biological content in significant parts.(Farm
Security and Rural Investment Act, 2002) They can be derived from municipal solid waste,
marine organisms, agricultural and forestry feed stocks, including wood, wood waste and
residues, grasses, crops and crop by-products.(Mohanty, Misra, & Drzal, 2002; Van Dam,
De Klerk-Engels, Struik, & Rabbinge, 2005) Bio-based products (including fuels and
chemicals) can be produced either as alternatives to fossil-based platforms, using
developed infrastructures and value chains, or as advanced platforms that require new
infrastructures and value chains.(Vennestrøm, Osmundsen, Christensen, & Taarning,
2011) The goal is to provide the same molecule or a different molecule with the same or
superior chemical properties, including function and reactivity.(Octave & Thomas, 2009)
As a primary renewable energy source in US, biomass is just behind hydropower(Chum &
Overend, 2001), but it sets itself aside from other renewable resources because it uses up
atmospheric CO2 through photosynthesis and store its energy in form of chemical bonds.
(S. V. Mohan, Modestra, Amulya, Butti, & Velvizhi, 2016; Vennestrøm et al., 2011) This
characteristic makes biomass suitable for multiple applications other than heat and power
2
generation, such as chemical conversion to alternatives for fuel and chemical
industries.(Vennestrøm et al., 2011) The land and agricultural resources of the United
States are sufficient enough to meet current domestic and export demands for food and
feed while producing surplus amount of biomass for the bio-based industry.(National
Research Council, 2000) The main constituents of biomass are carbohydrates (including
sugar, starch ,cellulose and hemicellulose), lignin, protein and fats which represent 95% of
its mass.(Octave & Thomas, 2009) Currently, 105 million tons of cellulose, the most
abundant biopolymer on earth, is produced annually, and only 150 million tons, are used.
Lignin, the second most abundant biopolymer on earth, has annual production of 50 million
tons, and most of it is not utilized at all.(Van Dam et al., 2005) Figure 1 shows the
resources, conversions and product categories from biomass processing.
Figure 1- Biomass resources, conversionn and products (adapted from eXtension, 2013)
3
Every fraction of biomass can be converted to useful products, depending on the feedstock
and composition. Biofuels can be produced from several biomass resources using multiple
conversion routes. When biomass is used for fuel production, biochemical routes use
microorganisms to convert cellulose and hemicellulose components to sugars, prior to their
fermentation to ethanol.(Sims, Mabee, Saddler, & Taylor, 2010) Thermochemical routes
include technologies such as gasification and pyrolysis which produce a synthesis gas
(CO+H2) that can be further processed to produce a wide range of long chain carbon fuels
based on Fischer-Tropsch conversion.(Schmidt & Dauenhauer, 2007; Sims et al., 2010)
Lignin content of biomass cannot go through biological conversions so it is typically
burned onsite to produce heat and power. For the case of thermochemical conversion, the
whole biomass, including lignin, is converted into synthesis gases.(Mu, Seager, Rao, &
Zhao, 2010)
Bio-based chemical conversion, on the other hand, is challenged by the lack of economic
conversion technologies, infrastructure for large-scale production, and abundance of
chemical targets.(Bozell & Petersen, 2010) Among potential chemicals derived from
biomass, biopolymer production is industrially more developed compared to fine
chemicals.(Mohanty et al., 2002) Synthetic bioplastics have been around for about 150
years.(L. Shen, Haufe, & Patel, 2009) In the 1930s and 1940s, various biopolymer
formulations were invented but the revolution of crude-oil extraction and refineries in
1950s provided a source of cheap synthetic polymers and as a result, impeded further
progress of bio-based products.(L. Shen et al., 2009) Currently, biopolymers, such as
polylactic acid (PLA), polyhydroxyalkanoate (PHA), polyhydroxybutyrate (PHB),
4
cellulose acetate propionate (CAP) and cellulose acetate butyrate (CAB) and their blends
are applied in paint, packaging, plastics, coatings and automotive industry.(Mohanty et al.,
2002) Production of non-polymer renewable chemicals, on the other hand, is not well-
developed and is dominated by existing markets with stable demand rates, such as ethylene,
propylene, acrylic acid, and epichlorohydrin.(Vennestrøm et al., 2011)
Production of biofuels and bio-based chemicals separately uses selective methods,
converting a specific fraction of biomass while generating a significant amount of residues/
wastes. The concept of a biorefinery, in an analogy to petrochemical refineries, has been
developed to avoid waste generation and aims to convert all components of biomass into
valuable products.(Demirbas, 2009; Fernando, Adhikari, Chandrapal, & Murali, 2006;
Kamm & Kamm, 2004) Biorefineries would provide energy (such as biofuels and heat),
molecules (such as commodity and fine chemicals and nutraceuticals), materials (such as
plastics and composites), and also food ingredients.(De Jong, Higson, Walsh, & Wellisch,
2012; Octave & Thomas, 2009)
Considering the concept of multi-product generation from biomass feedstock, several
chemicals can be produced from each fraction. Biomass processing in biorefineries can
produce syngas, biogas, carbohydrate derivatives, and refined lignin.(Cherubini &
Strømman, 2011; De Jong, Higson, et al., 2012; Kurian, Nair, Hussain, & Raghavan, 2013)
Syngas is produced through thermochemical conversion of biomass. It is used to produce
heat, power, hydrogen and olefins or go through fermentation to generate methanol and
ethanol. (De Jong, Higson, et al., 2012; Haro, Villanueva Perales, Arjona, & Ollero, 2014;
5
Meerman, Ramírez, Turkenburg, & Faaij, 2011; Munasinghe & Khanal, 2010) Biogas is
produced from anaerobic digestion of biomass with the purpose of energy applications.(De
Jong, Higson, et al., 2012) Wet milling of biomass produces extracts/ residues with
significant amount of carbohydrates, proteins, amino acids and enzymes, makes it a rich
stream for further processing.(Moncada, Posada, & Ramírez, 2015) Carbohydrates can be
broken down to C5 and C6 sugars. C5 sugars are mostly sourced from hydrolysis of
hemicellulose and are used in production of xylitols, furfurals and ethanol.(Werpy et al.,
2004) C6 sugars are obtained from sucrose, cellulose and starch and can be processed for
carboxylic acids, alcohols, acetones, sorbitols and hydroxymethyl furfurals.(Bozell &
Petersen, 2010; Cherubini & Strømman, 2011) Oil fraction of biomass can be converted to
various categories of products including food, biofuels, fatty alcohols, lubricants and care
products.(Moncada et al., 2015) The lignin fraction is the residual stream of biomass after
hydrolysis of cellulose and hemicellulose. Its aromatic structure provides a good source for
benzene, toluene, xylene, ethyl benzene, vanillin and phenol production.(De Jong, Higson,
et al., 2012; Zakzeski, Bruijnincx, Jongerius, & Weckhuysen, 2010) Figure 2 shows
different components of biomass and potential building blocks, in detail. Primary building
blocks can be produced directly from processing of syngas, sugars and aromatics while
secondary chemicals are derivatives of building blocks and require further processing.
6
Figure 2- Potential building blocks from processing biomass (partially adapted and modified from Werpy et al., 2004)
According to The Technology Road Map for Plant/ Crop, published by the US National
Renewable Energy Lab (NREL), at least 10% of the US chemical feedstock demand should
be met by plant-derived materials, by 2020. This portion will increase to 50%, upon
development of processing methods, production and market penetration, by 2050.(NREL,
1999) The ACS Green Chemistry Institute Formulators’ Roundtable is working with
industries to develop green formulations for various categories of products. Recently, ten
categories were identified for priorities in formulation development including
antimicrobials, solvents, small amines, chelates and sequencing agents, boron alternatives,
7
fragrance raw materials, corrosion inhibitors, alkalonamides surfactants and UV-
screens.(Jessop et al., 2015) Biofuels production, on the other hand, is mostly regulated
through Renewable Fuel Standard (RFS) program, developed by USEPA. Total biofuel
production in US is mandated to reach 36 billion gallons by 2022,(USEPA, 2015) 55%
more production compared to 2015.(USEIA, 2015a, 2015b, 2016) Table 1 lists bio-based
chemicals, their market size and producers.
Tab
le 1
-Lis
t of m
ajor
bio
-bas
ed p
rodu
cts,
thei
r pro
duce
rs a
nd m
arke
t siz
e
Prod
ucts
Mar
ket t
ype
Mar
ket s
ize
Maj
or p
rodu
cers
Feed
stoc
k
Bio
fuel
s(M
Mga
l.y-1
)
Bio
etha
nol
Exis
ting
1480
6
(USE
IA, 2
015b
)
AD
M, P
oet,
Val
ero
Ener
gy C
orpo
ratio
n
(Far
m In
dust
ry N
ews,
2016
)
Cor
n/w
heat
(Bio
fuel
.org
.uk,
201
6)
Bio
dies
elEx
istin
g12
68
(USE
IA, 2
016)
RB
F Po
rt N
eche
s LLC
, REG
Gre
y H
arbo
r
LLC
, Lou
is D
reyf
us A
gric
ultu
ral I
ndus
tries
(Far
m In
dust
ry N
ews,
2016
)
Soyb
ean
(Bio
fuel
.org
.uk,
201
6)
Bio
-bas
ed c
hem
ical
s(M
Mt.y
-1)(
Ven
nest
røm
et a
l., 2
011)
Ace
tic a
cid
Exis
ting
9-
Etha
nol
Acr
ylic
aci
dEx
istin
g4.
2A
rkem
a, C
argi
ll/N
ovoz
ymes
Gly
cero
l/gl
ucos
e
C4
di-a
cids
Emer
ging
(0.1
-0.5
)B
ASF
/ Pur
ac/ C
SM, M
yria
nt
Glu
cose
Epic
hlor
ohyd
rinEx
istin
g1
Solv
ay, D
owG
lyce
rol
Ethy
lene
Exis
ting
110
Bra
ksem
, Dow
/Cry
stal
sev,
Bor
ealis
tG
luco
se
Ethy
lene
gly
col
Exis
ting
20In
dia
Gly
cols
, Dan
chen
g In
dust
rial
Glu
cose
/ xyl
itol
Gly
cero
l Ex
istin
g 1.
5A
DM
, P&
G, C
argi
llV
eget
able
oil
5-H
ydro
xym
ethy
lfufu
ral
Emer
ging
--
Glu
cose
/fru
ctos
e
3-H
ydro
xypr
opio
nic
acid
Emer
ging
(>0.
5)N
ovoz
ymes
/Car
gill
Glu
cose
Isop
rene
Exis
ting/
em
ergi
ng0.
1 (0
.1-0
.5)
Dan
isco
/Goo
dyea
rG
luco
se
Lact
ic a
cid
Exis
ting/
em
ergi
ng0.
3 (0
.3-0
.5)
Car
gill,
Pur
ac/A
rkem
a, A
DM
/ Gal
actic
Glu
cose
Levu
linic
aci
dem
ergi
ng
(>0.
5)Se
getis
, Mai
ne B
iopr
oduc
ts, L
e C
alor
ieG
luco
se
Ole
oche
mic
als
Exis
ting
10-1
5Em
ery,
Cro
da, B
ASF
, Van
tage
Ole
oche
mic
als
Veg
etab
le o
il/ fa
t
1,3-
Prop
aned
iol
Emer
ging
(0.1
-0.5
)D
upon
t/ Ta
te &
Lyl
eG
luco
se
Prop
ylen
eEx
istin
g80
Bra
skem
/ Nov
ozym
esG
luco
se
Prop
ylen
e gl
ycol
Exis
ting/
em
ergi
ng1.
4 (>
2)A
DM
, Car
gill/
Ash
land
, Sen
ergy
, Dac
heng
Indu
stria
l
Gly
cero
l/so
rbito
l
Poly
hydr
oxy
alka
noat
eEm
ergi
ng(0
.1-0
.5)
Met
abol
ix/ A
DM
Glu
cose
*M
arke
t siz
e of
an
exis
ting
mar
ket i
s giv
en a
s its
cur
rent
size
incl
udin
g pr
oduc
tion
from
foss
il re
sour
ces,
for e
mer
ging
mar
kets
, the
exp
ecte
d m
arke
tsi
ze is
repo
rted
in p
aren
thes
is
10
1.1.2. Environmental Implications of bio-based products
Globally, transportation and industry are responsible for about 47% of total GHG emissions
and 50% of primary energy use. (IPCC, 2015; USEIA, 2011) while on industrial level,
chemical industry, ranks first in energy use and third in GHG emissions.(Broeren, Saygin,
& Patel, 2014; IEA, 2012) These statistics highlight the significant share of petrochemicals
and fuels in two main environmental concerns of past decade. Besides GHG emissions and
energy use, VOC emissions and aquatic toxicity are other known impacts from fossil-based
fuels and chemicals.(Furuholt, 1995; USEPA, 2016b) In an effort to create a consistent
platform for resolving environmental issues, green chemistry principles were introduced in
1990s. (Anastas & Eghbali, 2010) These principles are trying to address environmental
concerns of current formulations while motivating for design improvements and use of
renewable feedstock.(Anastas & Eghbali, 2010) Biomass-derived products are expected to
mitigate environmental impacts associated with conventional fuels and chemicals while
delivering the same or superior functions. However, environmental preference of bio-based
products is not fully promised, considering their impacts from increased agricultural
activities such as land use change and eutrophication.(Börjesson & Tufvesson, 2011;
Noble, Bolin, Ravindranath, Verardo, & Dokken, 2000) Moreover, large-scale production
from biomass requires maintaining the balance for food and feed production.(IEA, 2009;
Popp, Lakner, Harangi-Rákos, & Fári, 2014) In order to maximize environmental benefits,
minimize transition of environmental impacts from one category to another, and meet the
demand of biomass production for multiple applications, sustainable product development
is required.
11
Sustainable development is defined as “development that meets the needs of the present
generation without compromising the ability of future generations to meet their own
needs”. (Brundtland et al., 1987) One important implementation of this general concept is
that of Triple Bottom Line (TBL) assessment of products and processes, which aims to
balance among social equity, economic prosperity, and environmental
protection.(Elkington, 2001) These three categories of considerations are interconnected
and there is a growing number of national and international efforts to motivate bio-based
industries and develop a secure economy for their maturation. USDA bio-preferred
program is one of the initiatives in developing stable markets for bio-based chemicals
through mandatory federal purchasing and “USDA Certified Bio-based Product” labeling.
(Golden, Handfield, Daystar, & McConnell, 2015) This program links environmental
preference to economic growth, motivating industries to invest on sustainable
formulations. Since 2002, 97 categories of products (14000 products) have been covered
by this programs, contributing $369 billion in US economy and four million jobs
creation.(Golden et al., 2015) Europe has also developed a bio-economic program called
Bio-based Industries (BBI) which is a €3.7 billion public private partnership between EU
and Bio-based Industries Consortium (BIC). This program provides funding for innovative
projects focusing on biomass valorization, production optimization, development of new
value chains and biorefinery design in Europe and it creates job opportunities in rural
areas.(Biobased Industries Consortium, 2012) Mentioned programs are focused on socio-
economic aspects of sustainability development, while environmental protection is a factor
that should be considered from early stages of product development.
12
Assessing the environmental impacts of products goes all the way back to 1960s and 1970s,
especially on a comparative basis, such as which product uses less energy.(Guinee et al.,
2010) However, detailed analysis of products showed that for many products, a large
portion of overall energy use is associated with upstream processes, such as production or
distribution of a product rather than its use. As the assessment tools have developed, the
scope of environmental impacts has expanded beyond just energy use, into several other
categories related to emissions (such as global warming, acidification, and eutrophication),
non-renewable resource use, land use, biodiversity and noise.(Guinee et al., 2010) Life
cycle assessment (LCA) is a standardized tool framed by International Standardization
Organization (ISO), for measuring environmental impacts of a specific product. This tool
is widely used as a measure of sustainability in academic and industrial research and design
projects. In this dissertation, LCA is the main methodology for environmental analysis of
the products that have been studied. The following section introduces the framework and
basis of LCA methodology.
1.2. Life Cycle Assessment (LCA) of bio-based products
1.2.1. Introduction to Life Cycle Assessment (LCA)
Achieving sustainable production requires measuring tools to quantify environmental
impacts of products (goods or services). Every product has a ‘life’, starting with the
design/development of the product, followed by resource extraction, production
(production of materials, as well as manufacturing/provision of the product),
use/consumption, and finally end-of-life activities (collection/sorting, reuse, recycling,
waste disposal).(Rebitzer et al., 2004) Each stage in the life of a product, requires raw
13
sources from the environment as inputs, and releases emissions back to the environment,
as outputs. So, it is crucial to develop a life cycle perspective for estimating environmental
impacts of products. Figure 3 shows resource use and emissions in life cycle of a product.
LCA is a framework for evaluating environmental impacts attributed to the life cycle of a
product, back to the raw material acquisition and down to the waste handling
scenarios.(Rebitzer et al., 2004) LCA can assist in identifying opportunities to improve
the environmental performance of products at various points in their life cycle, informing
decision-makers about strategic planning, priority setting and product or process design or
redesign, developing relevant indicators of environmental performance, including
measurement techniques and marketing (e.g. implementing an eco-labelling scheme,
making an environmental claim, or producing an environmental product declaration).(ISO,
2006) In addition, LCA helps determining environmental trade-offs between several
products with the same functionality, makes it suitable as a comparative tool.
14
Figure 3- Life cycle of a product (adapted from (Rebitzer, 2002))
As described in ISO standards (14044:2006), every LCA study comprises of four phases:
a) Goal and Scope Definition: In this stage, objective, actors, system boundary,
functional unit, and scope of the study are specified. System boundary identifies the
processes that are going to be included in the analysis; functional unit is the basis for
the analysis that enables alternative goods, or services, to be compared and
analyzed.(Rebitzer et al., 2004) The actors involved, include the commissioner of the
work, the analyst, the intended audience for the results, and any other relevant
stakeholders.(ISO, 2006, p. 14) Studies are typically scoped to be from “cradle to
grave” (including use and disposal) or “cradle to gate” (only through final
production).
15
b) Life Cycle Inventory: Input/output data for all processes within the studied system are
collected and integrated as the life cycle inventory.(ISO, 2006, p. 14)
c) Life Cycle Impact Assessment: Resource inputs and emissions data are linked to the
potential environmental impacts using coupled fate-transport-exposure-effect
models.(Owens, 1997) These models track emissions from their sources to final sinks
in order to estimate the physical or biological changes caused in the receiving
environments.
d) Interpretation: Results of inventory analysis and life cycle impact assessment are
summarized and discussed as a basis for conclusions, recommendations and decision-
making in accordance with the goal and scope definition.(ISO, 2006, p. 1) This phase
can be integrated with each of the three phases described above.
LCA has been used in industrial and research purposes to evaluate environmental
performance (benefits and trade-offs) of various products, including renewable and fossil-
based products. Carbon footprint (CF) is one of the primary indicators for comparison of
sustainable products upon their conventional counterparts. A product carbon footprint is
the sum of all direct and indirect greenhouse gases (GHG) emitted over its life cycle. For
biomass- derived products, carbon dioxide sequestered through photosynthesis is
considered as a credit in the carbon footprint counting, if the life-time of product is long
enough to store sequestered carbon as a permanent storage.(Houghton, Meira Filho, Lim,
Treanton, & Mamaty, 1997) Here, the main assumption is that atmospheric CO2 acts as
carbon pool, supplying carbon content of feedstock during growth phase while this carbon
will be captured in final product long enough to compensate for the growth period. Several
16
standard protocols have been developed for the CF analysis, WRI/WBCSD, PAS 2050
(BSI, Carbon Trust, DEFRA), and ISO 14067, each specifying parameters that should be
considered in the assessment. Direct emissions are those from the main production
processes, while indirect emissions include GHG emissions associated with the production
of purchased energy and upstream and downstream processes.(WBCSD, 2011)
In addition to carbon footprint, other categories of environmental impacts play critical role
in definition of sustainable products such as eutrophication, acidification, ozone depletion,
land and water use and human health impacts. Life cycle impact assessment methods are
used to link the potential impacts to the actual quantitative values that can be compared
among alternatives. While the impacts are quantified, LCA models allow handling of co-
products and count for their share in overall environmental impacts, through three methods,
mass allocation- distributing the environmental burdens based on the mass of output
streams (main product and co-products)- economic allocation- assigning environmental
impacts based on the economic values of the output streams- and system expansion-
expanding the boundaries of the studied system to include the impacts of alternative
production of exported functions.(Ekvall & Finnveden, 2001)
1.2.2. Life cycle impact assessment (LCIA) methods
Life cycle impact assessment (LCIA) methods are developed to quantify a broad range of
environmental impacts in life cycle of a product.(Frischknecht et al., 2007) Environmental
impacts can be assessed based on two approaches of mid-point and end-point indicators. A
midpoint indicator is defined as a parameter in the cause and effect network for a particular
impact category that is between the inventory data and the category endpoint.(J. C. Bare,
17
Hofstetter, Pennington, & De Haes, 2000) For example, ozone depletion is a midpoint
indicator that can lead to skin cancer, immune system suppression, marine life damage,
material damage and crop damage as endpoint impacts.(J. C. Bare et al., 2000) Most of the
current impact assessment methods focus on midpoint indicators.
Based on ISO 14042, (Environmental Management - LCA– Life Cycle Impact
Assessment), there are three broad groups of impact categories that should be considered
in the LCA study, referred to as AoPs (Areas of Protection). AoPs include resource use,
human health consequences and ecological consequences.(Pennington et al., 2004)
Developed impact assessment methods address the AoPs using defined impact categories.
The following LCIA methods are peer-reviewed and in common use: (Frischknecht et al.,
2007)
CML 2001
Cumulative energy demand
Cumulative exergy demand
Eco-indicator 99
Ecological footprint
Ecological scarcity 1997
Ecosystem damage potential - EDP
EDIP’97 and 2003 - Environmental Design of Industrial Products
EPS 2000 - environmental priority strategies in product development
IMPACT 2002+
IPCC 2001 (climate change)
18
TRACI
Selected Life Cycle Inventory indicators
The selection of environmental impact categories is designed to consider local and
regional, and global effects, as well as short-term and long-term effects. These categories
are not meant to be equivalent in scale or severity, but rather represent a broad range of
environmental and public health issues. Each category is measured in equivalent units,
that is, in relation to a reference chemical whose fate and transport and subsequent effects
are well-understood and documented. A familiar example is that greenhouse gases (GHGs)
are emitted and cause global warming potential (GWP) based on their potential for
radiative forcing in the atmosphere. GWP in total is expressed relative to that of carbon
dioxide, that is, in units of CO2 equivalents, (CO2 eq.). Depending on whether impacts are
local, regional, or global in nature, the LCIA models used to calculate characterization
factors will use biogeochemical models designed for that scale. Equation 1(Pennington
et al., 2004) and Equation 2(Pennington et al., 2004) show the methods used for
calculation of category indicators and characterization factors, where s denotes the
chemicals, i is the location of emission, j is the location of exposure receptor, and t is the
time period during which the potential contribution to the impact is taken into account.
Equation
Equation 2
19
TRACI- Tool for the Reduction and Assessment of Chemicals and Other Environmental
Impacts- is the impact assessment method used in the LCA analysis of this dissertation.
TRACI allows the quantification of stressors that have potential effects in environment and
human health. This impact assessment method has been used in many applications such as
US Green Building Council’s LEED Certification and the National Institute of Standards
and Technology’s BEES tool. However, the selection of the impact categories is a
normative decision depending on the purpose of the individual use.(J. Bare, 2011) Impact
categories included in TRACI are as below:
Acidification: it refers to the increasing concentration of hydrogen ions (H+) within
a local environment. Acidifying substances are often air emissions which can
deposit on soil and water and cause damage to building materials, paints, and
human-built structures, lakes, rivers and various plants and animals. Nitric acid and
sulfuric acid are popular pollutants in this category. Acidification potential is
expressed in units of kg SO2 equivalent.(Pennington et al., 2004)
Eutrophication: this environmental impact is defined as the “enrichment of an
aquatic ecosystem with nutrients that accelerate biological productivity and an
undesirable accumulation of algal biomass”.(USEPA, 2008) phosphorous and
nitrogen play important role in this category. Eutrophication impact is expressed in
units of kg nitrogen equivalent.(Pennington et al., 2004)
Global warming potential: global climate change/global warming is an average
increase in the temperature of the atmosphere near the Earth’s surface as a result of
increased emissions of greenhouse gases from human activities.(USEPA, 2016a)
The main GHGs are carbon dioxide, methane, nitrous oxide, sulfur hexafluoride,
20
and certain fluorocarbons. The GWPs are estimated with a 100-year time horizon
and expressed in unit of kg CO2 equivalent.(Pennington et al., 2004)
Ozone depletion potential: Ozone within the stratosphere provides protection for
solar radiation, and low concentration of this compound can lead to increased
frequency of skin cancer. In addition, ozone has been documented to have effects
on crops, plants, marine life, and human-built materials. Chlorofluorocarbons
(CFCs) are the most known substances causing ozone depletion in stratosphere, so
they are used as base compounds in quantifying impacts of this
category.(Pennington et al., 2004)
Human health criteria: this category of impacts deals with particulate matter and
precursors to particulates, which has the ability to cause respiratory illness and
death.(Pennington et al., 2004) Primary particulate matter are directly emitted to
the atmosphere or produced through a series of chemical reactions. The most
common precursors to secondary particulates are sulfur dioxides (SO2) and nitrogen
oxides (NOx).(Pennington et al., 2004) Fossil fuel combustion, wood combustion,
and dust particles from roads and fields are sources of primary and secondary
particulate matter.(Breysse et al., 2013) The impact of this category is expressed
based on kg PM 2.5 equivalent.(Pennington et al., 2004)
Human cancer, non-cancer and ecotoxicity: Human health and ecotoxicity in
TRACI is represented by three impact categories of cancer, non-cancer and criteria
pollutants, according to the structure of the EPA regulations and the chemical and
physical behaviors of the pollutants of concern. The USEtox model is used to
develop human health cancer and non-cancer toxicity potentials and freshwater
21
ecotoxicity potentials for over 3,000 substances including organic and inorganic
substances.(Pennington et al., 2004) Impacts for cancer and non-cancer are
estimated based on the unit of CTUh while ecotoxicity is measured based on
CTUe.(Pennington et al., 2004) CTUh, comparative toxic unit, provides the
estimated increase in morbidity in the total human population per unit mass of a
chemical emitted.(Rosenbaum et al., 2008) CTUe, on the other hand, is the
comparative toxic unit relating to ecosystem and provides an estimate of the
potentially affected fraction of species integrated over time and volume per unit
mass of a chemical emitted.(Rosenbaum et al., 2008)
Photochemical smog formation: Ground level ozone is created as a result of
reactions of nitrogen oxides (NOx) and volatile organic compounds (VOCs) in the
presence of sunlight. It can cause a variety of human-health issues including
increasing symptoms of bronchitis, asthma, and emphysema.(Pennington et al.,
2004) Permanent lung damage can be a result of long-term exposure to ozone.
Quantitative measurement of ground level ozone is expressed in term of kg O3
equivalent.(Pennington et al., 2004)
Fossil fuel depletion: This category is different from total fossil fuel consumption
because total non-renewable energy consumption does not fully address potential
depletion issues associated with these flows. It is counting for the continued
extraction and production of fossil fuel that tends to use most economically
recoverable reserves first and further extraction will require more energy and cost.
So this category is represented by the MJ surplus of energy associated with the
production of target product.(J. Bare, 2011)
22
LCA approaches and assessment methods have been applied on various renewable
products in the past few years. Currently, there is an active research area in finding the
most efficient sources and conversion methods for fuels and chemicals production from
biomass. The present dissertation focuses on life cycle assessment of several products,
including renewable fuels and chemicals, while addressing scientific gaps in previous
researches. The next sections present more details about existing gaps and challenges, the
motivations behind each project, and the approach taken toward filling the gaps.
1.2.3. Gaps and Challenges
LCA is used as one of the main methodologies in evaluating environmental burdens and
benefits of renewable fuels and chemicals. Most of the LCA studies have found a
significant net reduction in GHG emissions and energy consumption when bioethanol and
biodiesel are compared to their diesel and gasoline counterparts.(Kim & Dale, 2005; Punter
et al., 2004; Von Blottnitz & Curran, 2007) However these results were shown to be
feedstock dependent and varying the source could have a significant effect on overall
environmental preference. In 2005, the U.S. EPA developed a consistent platform for
production capacity and characterization of biofuels in U.S, called RFS (Renewable Fuel
Standard). RFS targets the production volume and minimum GHG emissions reduction for
fuels produced from various categories of biomass.
Bio-based chemicals, on the other hand, are currently in the phase of research and
development and except some cases of biopolymer production, most of them are not
produced in large-scale. Bio-based building blocks, produced through fermentation
processes are mostly environmentally attractive, considering GHG emissions and non-
23
renewable energy use.(Patel et al., 2006) However, depending on the type of feedstock,
mentioned environmental benefits can change significantly. The challenge in this case, is
to develop efficient processes for the collection, handling and pretreatment of biomass and
for the selective conversion of biomass feedstock.(Vennestrøm et al., 2011) The
quantitative analysis of environmental impacts of bio-based chemicals are scarce due to
lack of processes and incomparable due to different assumptions and boundary conditions.
Correspondingly, there is no such sustainability metric for comparison of bio-based
chemicals with their fossil-based counterparts. Current regulatory programs are not
comprehensive and consistent, regarding the measurement methods and sustainability
criteria. There is a need to integrate the LCA results of bio-based chemicals and evaluate
the state of knowledge and gaps of current analysis. (Hermann, Blok, & Patel, 2007; Mila
i Canals et al., 2011)
While many LCA studies consider GHG emissions and energy use as main indicators of
sustainability, it should be noted that other categories of environmental impacts can play
an important role in sustainability of bio-based products, such as local air pollution,
acidification, eutrophication, ozone depletion and land use change. Some case studies have
shown that bio-based products are associated with higher impacts in categories other than
global warming potential and energy use, such as acidification and eutrophication, which
is an important factor in large-scale decision making.(Larson, 2006) While current
research targets minimizing these trade-offs through use of efficient bio-feedstock
(agricultural and forest residues) and development of efficient conversion methods (higher
24
conversion rate with less solvent use), this area suffers from lack of in depth data on
potential impacts.
The biorefinery concept is an approach to minimize cost and environmental impacts and
maximize production of value added chemicals. Biorefinery is a facility that integrates
biomass conversion through production of fuels, power and chemicals.(Smith &
Consultancy, 2007) Additional challenge is the design of biorefineries that process multiple
input feedstock in such a way that the byproducts and waste of one stage could be sold as
a high value commodity or be used as a feedstock or energy source for other stages.(Octave
& Thomas, 2009) In this way, biofuels and bio-based chemicals can be produced in one
facility, using every fraction of biomass feedstock, and resulting in less waste, more
products and less environmental impact.
1.3. Motivation and Summary of Chapters As mentioned above, there are some gaps and challenges in life cycle assessments of
renewable products. This dissertation is trying to address some of the ongoing challenges
in large-scale development of bio-based products.
Chapter 2 provides a review and meta-analysis of LCA studies conducted for bio-based
chemicals to date. The goal is to collect the available LCA data for bio-based chemicals,
analyze the difference in scope, extent, conversion and the method of LCA analysis, find
the gaps in current data and finally compare available results with the sustainability
thresholds developed for biofuels. This review highlights the areas where more research
25
is required and determines the state of knowledge for developing a Renewable Chemical
Standard (RCS). Moreover, statistical analysis of the collected data uncovers patterns as to
which biofeedstock and conversion platforms, and which target bio-based chemicals, are
most promising in terms of reducing fossil energy use and GHG emissions compared to
existing petrochemicals. The analysis provides additional statistical insights on methods
development, specifically what scope and allocation rules could be used in developing a
consistent basis in LCA analysis of bio-based chemicals. These specifications can be
integrated into PCRs (Product Category Rules) and be adopted by LCA practitioners in
academia and industry, to develop a Renewable Chemical Standard (RCS). This is the first
study of its kind for bio-based chemicals overall. Results of this study can guide future
LCA research to fill the gaps in life cycle assessment of bio-based chemicals.
Chapter 3 is a case study of aromatic chemicals production from agricultural residues.
From chapter 2, catechol is found as one of the bio-based chemicals that suffers from lack
of data. In this chapter, renewable and non-renewable catechol production pathways are
modeled using ASPEN plus, a process design software, and their environmental burdens
are compared. Lab-scale data are used for simulating the actual large-scale production
process. Comparing environmental impacts of renewable and non-renewable catechols
highlights the trade-offs in environmental impacts and required modifications for the
production process, including the choice of bio-feedstock and process design parameters.
The assessment results can guide further research in order to improve the synthesis process,
as it specifically indicates the need for low solvent use, substitution, recovery and reuse of
highly potent solvents. This project does not develop any new methods, but rather presents
26
a case study for a new bio-based synthesis pathway, and indicates that the process will
achieve the intended environmental goals of reducing resource use and environmental
impacts on a life cycle basis. The study is also innovative in its integration of laboratory
experiments, chemical process simulation, and LCA, which are used in concert to evaluate
progress toward technological and sustainability goals in early stages of product
development.
Chapter 4 focuses on LCA of a renewable formula for wood flooring coating applications.
This project is conducted in collaboration with PPG industries. The renewable formula
consists of 30% renewable chemicals and zero-to-low volatile organic compounds (VOC)
and is supposed to be produced as an alternative to the conventional wood flooring
coatings. Data are sourced from the PPG Coatings and Resins R&D Center and its
sustainability analysis is required prior to pilot-scale production of the coatings. The results
demonstrate the contribution of chemicals in overall environmental burden of each coating,
highlight the components that need to be considered for further research and provide
recommendations on how to maximize benefits of renewable formulation. The analysis
indicates the importance of expanding current LCI databases to include specialty
chemicals, as many commercial formulations are non-existent within public and
commercial databases. In this case, more than 40 new unit processes are created, which can
be used by the worldwide LCA community in assessing chemical and formulated products.
Chapter 5 focuses on biorefinery design, estimating the environmental benefits associated
with production of biofuels and value added chemicals from algal biorefinery. Composition
27
of algae, as main feedstock, was analyzed under 9-day cultivation and two feeding regimes.
This study is scoped to include conversion of lipid, protein and carbohydrate to biofuels,
animal feed and on-site energy. Maximum environmental benefits obtainable from various
scenarios are evaluated using GREET model. My master’s thesis is an extension on chapter
5, considering the factor of time. There, cultivation of algae is analyzed under 3, 6, 9, and
12 days, considering two feeding regimes and the co-product valorization scenarios. This
project uses the concept of dynamic growth and harvesting in design of biorefinery
schemes, providing valuable insights on how the environmental performance can be
maximized. This aspect of dynamic optimization in microalgal biorefineries is applied
here for the first time to non-lipid fractions. Final results emphasize the choice of co-
product valorization under each scenario, while identifying the best case in terms of target
species, feeding regime, and harvesting cycle time.
This dissertation develops novel data sets, develops new, dynamic LCA methods, and
assesses potentially breakthrough bio-based chemical syntheses in a multi-faceted
investigation of the sustainability of bio-based chemicals. The work builds a foundation
for further LCA analysis, process modification, and finally commercial development of
bio-based products. Each chapter of this dissertation is trying to address current challenges
using novel approaches with the twin goals of advancing environmental assessment
methods and providing process-level results specific to the bio-based chemicals,
formulations, and processing schemes under study. For the first time, existing data on bio-
based chemicals are collected in a meta-analysis, in order to provide insights on which
chemicals need further research and which existing processing routes will meet
28
environmental objectives (such as a Renewable Chemical Standard); environmental
performance of renewable building blocks is then studied in the context of novel
formulations for high demand products, providing recommendations on the choice of
feedstock and operational conditions; and finally, environmental benefits of bio-based
products are studied using the concept of biorefineries using a novel dynamic assessment
approach that integrates models from microbial kinetics.
29
Chapter 2:
Meta-Analysis of Life Cycle Energy and Greenhouse Gas Emissions for Priority Bio-based Chemicals
This study has been published
Montazeri, M., Zaimes, G. G., Khanna, V., & Eckelman, M. J. (2016). Meta-Analysis of Life Cycle Energy and Greenhouse Gas Emissions for Priority Bio-based Chemicals. ACS Sustainable
Chemistry & Engineering.
Research and development for bio-based chemicals production has become a strategic
priority in many countries, due to the widespread availability of renewable feedstocks and
the potential for reduced life cycle greenhouse gas (GHG) emissions and fossil energy use
compared to petrochemicals. These environmental benefits are not assured, however, as a
multiplicity of processing features (i.e., biofeedstock, conversion platform, energy/solvent
recovery) and life cycle modeling factors (i.e., coproducts, allocation scheme, study scope,
location) influence the overall GHG emissions and energy use of a bio-based chemical
production scheme. Consequently, there has been high variability in reported
environmental impacts of bio-based chemical production across prior life cycle assessment
(LCA) studies. This meta-analysis considers 34 priority bio-based chemicals across 86
discrete LCA case studies. Most bio-based chemicals exhibited reduced GHG emissions
and net energy use compared to petrochemical counterparts, with exceptions including. p-
xylene, acetic acid, and adipic acid. Seven priority bio-based chemicals had no reported
results, predominantly lignin-derived. GHG emissions reductions were compared against
proposed thresholds from the Roundtable on Sustainable Biomaterials (RSB), the
International Sustainability & Carbon Certification (ISCC), and those applied to U.S.
biofuels under the Renewable Fuels Standard (RFS2) program. ANCOVA and ANOVA
30
statistical tests were utilized to identify process and life cycle modeling factors that
contribute significantly to environmental metrics. Conversion platform was found to be
statistically significant (α=0.1) for GHG emissions, with thermochemical routes having the
highest results, while LCA coproduct allocation scheme was significant for non-renewable
energy use. Recommendations for harmonizing and prioritizing future work are discussed.
2.1. Introduction Biofuels and bio-based chemicals have received significant interest as a potential low-
carbon and environmentally sustainable alternative to conventional fossil-based fuels and
petrochemicals. As defined by the US Secretary of Agriculture in the Farm and Rural
Investment Act of 2002, bio-based products are commercial or industrial products that are
composed of biological products, renewable agricultural and forestry materials or
intermediate feedstocks, in whole or in significant parts.(Farm Security and Rural
Investment Act, 2002) The annual production of bio-based chemicals (excluding fuels) is
estimated to be 50 million tons,(De Jong, Higson, et al., 2012) dominated by bio-based
polymers (55%), oleochemicals (20%) and fermentation products (18%).(NNFC, 2014)
Commercialization of bio-based chemicals is still nascent, and their penetration rate in the
global market will be strongly dependent on development of bio-refineries.(Hatti-Kaul,
Törnvall, Gustafsson, & Börjesson, 2007) The US Department of Agriculture (USDA)
estimates that the global chemicals industry is projected to grow 3-6% annually through
2025, with the bio-based chemicals share of that market rising from 2% in 2006 to 22% or
more by 2025.(Williamson, 2010)
31
In order to prioritize research and development efforts, the US Department of Energy
(DOE) published a two-volume report listing target bio-based chemicals.(Holladay, Bozell,
White, & Johnson, 2007; Werpy et al., 2004) The first volume of this report investigated
bio-based chemical candidates derived from the carbohydrate content of biomass (sugar,
cellulose, and starch). 300 candidates were evaluated based on potential markets and the
technical complexity of the synthesis pathways. The synthesis routes were examined as
two-part pathways: transformation of sugars to building blocks; and conversion of building
blocks into secondary chemicals or families of derivatives.(Werpy et al., 2004) The second
volume of the report considered potential candidates derived from the lignin portion of
biomass. Three categories of products were studied, including: fuel and syngas;
macromolecules and aromatics; and miscellaneous monomers. Candidates were chosen
based on their technical difficulty of production, market risk, building block utility, and
whether a pure material or a mixture would be produced.(Holladay et al., 2007)
The chemical sector is the largest industrial energy user with ~10% of global primary
energy use,(Broeren et al., 2014) and ranks third among industrial sectors for direct CO2
emissions, after iron and cement.(IEA, 2012) The expectation is that bio-based chemicals
require less energy to produce, with fewer associated emissions and a more favorable
environmental profile than their petrochemical counterparts. Numerous Life Cycle
Assessment (LCA) studies have quantified environmental trade-offs from switching to bio-
based production of fuels and chemicals, considering impacts of land use change, fertilizer
and pesticide runoff besides fossil energy use and air emissions.(Hall & Scrase, 1998;
Joslin & Schoenholtz, 1997; Matson, Parton, Power, & Swift, 1997; Miller, 2010; Petrou
32
& Pappis, 2009) For example, Groot and Boern conducted an LCA of polylactic acid
(PLA) production from sugarcane in Thailand, and compared the results with that of fossil-
based polymer. The study is a cradle to gate analysis including sugarcane cultivation,
sugarcane milling, auxiliary chemicals production, transport, and production of lactide and
PLA. On a mass basis basis, bio-based PLA had lower associated GHG emissions and less
material and non-renewable energy use compared to the fossil-based polymers; however,
PLA had higher impacts in acidification, photochemical ozone creation, eutrophication and
land use categories due to agricultural activities, compared to the fossil-derived polymer.
(Groot & Borén, 2010)
In an effort to reduce US dependence on petroleum-based transportation fuel, heating oil,
and jet fuel, the national Renewable Fuel Standard (RFS) program was created under the
Energy Policy Act of 2005, which sets explicit sustainability criteria for renewable fuels.
In 2007, the Energy Independence and Security Act expanded upon this program to
establish RFS2 by mandating that 36 billion gallons of renewable fuels be added to the
transportation fuel mix by 2022. In addition, RFS2 established relative life cycle GHG
emission reduction thresholds for three categories of biofuels—conventional biofuel
(primarily corn-based), biomass-based diesel, and cellulosic biofuel―compared to the
emissions baseline of the gasoline or diesel they replace. As defined by RFS2, life cycle
GHG emission reductions of 20%, 50% and 60%, are required for conventional, biomass-
based and cellulosic biofuels, respectively.(USEPA, 2015)
33
The RFS2 criteria can be useful as benchmarks for bio-based chemicals. There are also
initiatives that propose sustainability criteria specifically for renewable chemicals, such as
the Roundtable on Sustainable Biomaterials (RSB), USDA BioPreferred, International
Sustainability & Carbon Certification (ISCC), and Bonsucro. RSB and ISCC are multi-
stakeholder coalitions that measure sustainability of different renewable fuels and
chemicals and specify GHG emissions reduction thresholds, as one of the primary criteria
in their sustainability measures. GHG reduction thresholds for both programs are assigned
based on cradle-to-gate system boundary while inclusion of transport and distribution of
target chemical is mandated in ISCC scope but not in RSB. Land use change (LUC) and
carbon sequestered in growth phase of biomass are also included in the scope of both
standards. GHG reduction thresholds are at least 10% and 35% for RSB and ISCC,
respectively.(ISCC PLUS, 2011; RSB, 2015; USDA, n.d.) The BioPreferred program
developed by USDA is another program that encourages the use of bio-based products,
consisting of mandatory purchasing requirements for federal agencies and their contractors
and a voluntary labeling initiative for bio-based products. Primary sustainability criteria in
this program is at least 25% bio-based content in the composition of the final product.
Bonsucro standard, on the other hand, is mostly used for chemicals derived from sugarcane
and set field-to-gate GHG reduction threshold for sugarcane production and processing
(<0.4 t CO2/t sugar- for agriculture + milling + processing). (BonSucro, 2014) While there
is not consistency in scope and extent of sustainability thresholds for life cycle GHG
reduction for bio-based chemicals, several authors have suggested testing bio-based
chemicals against RFS2-like criteria, initially focusing on bioethanol/bioethylene, as the
34
most common intermediates in production of renewable building blocks.(Carus, Dammer,
Hermann, & Essel, 2014; Posen, Griffin, Matthews, & Azevedo, 2014)
In this study, we reviewed published results for life cycle GHG emissions and energy use
for 34 priority bio-based chemicals, including those identified by DOE, compared against
their fossil-based counterparts. Prior meta-analyses of bioenergy systems showed that there
are several factors controlling environmental benefits from GHG emission and energy use,
from biomass carbon cycle and soil carbon change to selection of appropriate fossil
reference systems, homogeneity of input parameters, and co-product handling
schemes.(Cherubini et al., 2009; Cherubini & Strømman, 2011) The present meta-analysis
is conducted for bio-based chemicals, focusing on collection and interpretation of existing
LCA results with statistical analysis. It does not attempt a harmonization of various cases
but rather aims to identify trends across the many feedstocks and processing routes that
have been considered, while examining the statistical effects of modeling factors such as
co-product allocation. The main goals of this work are to evaluate a potential ‘Renewable
Chemical Standard’, to identify gaps in the assessment literature, and to synthesize the state
of knowledge for net energy and life cycle GHG emissions assessment of bio-based
chemicals. This work can support high-level policy-making that requires effective, broad-
based synthesis of existing knowledge.(Philp, 2015)
35
2.2. Methods
A detailed literature review was conducted on life cycle assessments (LCA) of eleven
sugar-based and eight lignin-based building blocks identified by the US DOE
reports.(Holladay et al., 2007; Werpy et al., 2004) In addition, sixteen other bio-based
chemicals were identified from the literature as research priorities. These additional
chemicals can be produced from either sugar or non-sugar components of biomass and are
categorized as ‘secondary chemicals’ by US DOE. Table 2 shows the bio-based chemicals
included in this study. Figure 2 in Chapter 1 shows a chemical synthesis tree for each
group of chemicals considered.
A survey of LCA studies was conducted including journal papers, academic dissertations,
conference papers, industrial reports, and patents published over the time period 2003-
2016. Several criteria were considered in screening LCA studies and reports, as follows.
Following the RSB and ISCC standards, all of the LCA studies were cradle-to-gate, taking
account of life cycle processes from raw material acquisition up to and including the
manufacture of the target bio-based chemical. Several studies reported cradle-to-grave
results; where the breakdown of results was included in the original studies, energy use and
GHG emissions from use and end-of-life stages were excluded in order to maintain
consistency with other cradle-to-gate studies. Selected studies focused on existing rather
than future scenarios, so the final results reported here, show life cycle GHG emissions and
energy burdens associated with currently developed agricultural and conversion methods.
Tab
le 2
-Lite
ratu
re so
urce
s for
life
cyc
le e
nerg
y us
e an
d G
HG
em
issi
on re
sults
Prio
rity
Bio
-bas
ed
Che
mic
al C
ateg
ory
Che
mic
al N
ame
Cas
e St
udy
Ref
eren
ces
Maj
or P
rodu
cers
DO
E ca
rboh
ydra
te-
base
d ch
emic
als
Ara
bini
tola
n/a
Asp
artic
aci
da
n/a
But
yrol
acto
ne b
[der
ivat
ive
of B
DO
]Fu
mar
ic a
cid
b[d
eriv
ativ
e of
succ
inic
aci
d]Fu
ran
dica
rbox
ylic
aci
d/Po
lyet
hyle
ne fu
rand
icar
boxy
late
(P
EF)
(De
Jong
, Dam
, Sip
os, &
Gru
ter,
2012
)(E
erha
rt, F
aaij,
& P
atel
, 201
2)
Glu
caric
aci
dc
[pre
curs
or o
f adi
pic
acid
]R
iver
top
rene
wab
les
(De
Jong
, Hig
son,
et a
l., 2
012)
Glu
tam
ic a
cid
c[p
recu
rsor
of N
-met
hylp
yrol
lidon
e]G
loba
l Bio
tech
, Mei
hua,
Fu
feng
, Juh
ua
(De
Jong
, Hig
son,
et a
l., 2
012)
Itaco
nic
acid
(Nus
s & G
ardn
er, 2
013)
Qin
gdao
Keh
ai B
iche
mis
try
Co.
, Ita
coni
x(D
e Jo
ng, H
igso
n, e
t al.,
201
2)M
alei
c ac
id b
[der
ivat
ive
of su
ccin
ic a
cid]
Prop
ioni
c ac
id(J
. Dun
n, 2
014)
(Ekm
an &
Bör
jess
on, 2
011)
(Tuf
vess
on, E
kman
, Sar
dari,
Eng
dahl
, &
Tuf
vess
on, 2
013)
Car
gill
(De
Jong
, Hig
son,
et a
l., 2
012)
Sorb
itola
n/a
Roq
uetta
, AD
M(D
e Jo
ng, H
igso
n, e
t al.,
201
2)Su
ccin
ic a
cid
(J. D
unn,
201
4)(B
ioA
mbe
r, 20
13)
(Cok
, Tsi
ropo
ulos
, Roe
s, &
Pat
el,
2014
)
Bio
Am
ber,
Myr
iant
, B
ASF
/Pur
ac, R
ever
dia
(DSM
/Roq
uetta
), PT
T C
hem
/ Mits
ubis
hi C
C
3
(Pat
el e
t al.,
200
6)(H
erm
ann
et a
l., 2
007)
(De
Jong
, Hig
son,
et a
l., 2
012)
Xyl
itol
(T. S
hen,
201
2)(X
IVIA
, 201
0)D
anis
co/ L
enzi
ng, X
ylito
l C
anad
a(D
e Jo
ng, H
igso
n, e
t al.,
201
2)
DO
E lig
nin-
base
d ch
emic
als
Bip
heny
lan/
a
Cre
sol/
Res
orci
nola
n/a
Cyc
lohe
xane
an/
a
Met
hano
l/Dim
ethy
l eth
er(G
oepp
ert,
Cza
un, J
ones
, Pra
kash
, &
Ola
h, 2
014)
Bio
MC
N, C
hem
rec
(De
Jong
, Hig
son,
et a
l., 2
012)
Phen
ol
(Gal
lard
o H
ipol
ito, 2
011)
Styr
ene
(Y. Z
hang
, Hu,
& B
row
n, 2
014)
Van
illic
aci
da
n/a
Van
illin
(M
atos
& P
etro
v, n
.d.)
(Mod
ahl,
Bre
kke,
& R
aada
l, 20
09)
Oth
er si
gnifi
cant
m
arke
t che
mic
als
Ace
tic a
cid
(Pat
el e
t al.,
200
6)(H
erm
ann
et a
l., 2
007)
Wac
ker
(De
Jong
, Hig
son,
et a
l., 2
012)
Acr
ylic
aci
d(A
dom
, Dun
n, H
an, &
Sat
her,
2014
)C
argi
ll, P
erst
op, O
PXB
io,
Dow
, Ark
ema
(De
Jong
, Hig
son,
et a
l., 2
012)
Adi
pic
acid
(Pat
el e
t al.,
200
6)(H
erm
ann
et a
l., 2
007)
(Van
Duu
ren
et a
l., 2
011)
Ver
dezy
ne, R
enno
via,
B
ioA
mbe
r, G
enom
atic
a(D
e Jo
ng, H
igso
n, e
t al.,
201
2)B
utan
edio
l (B
DO
)(A
dom
et a
l., 2
014)
Gen
omat
ica/
M&
G,
Gen
omat
ica/
Mits
ubbi
shi,
Gen
omat
ica/
Tate
& L
yle
(De
Jong
, Hig
son,
et a
l., 2
012)
3
But
adie
ne
(Ces
pi, P
assa
rini,
Vas
sura
, & C
avan
i, 20
16)
Ethy
l lac
tate
(Mue
ller,
2010
)V
erte
c B
ioSo
lven
t(D
e Jo
ng, H
igso
n, e
t al.,
201
2)i-B
utan
ol(A
dom
et a
l., 2
014)
But
amax
, Gev
on-
But
anol
(Pat
el e
t al.,
200
6)(H
erm
ann
et a
l., 2
007)
Cat
hay
Indu
stria
l Bio
tech
, B
utam
ax, B
utal
co,
Cob
alt/R
hodi
a(D
e Jo
ng, H
igso
n, e
t al.,
201
2)H
igh
dens
ity p
olye
thyl
ene
(HD
PE)
(Tsi
ropo
ulos
et a
l., 2
015)
Low
den
sity
pol
yeth
ylen
e (L
DPE
)(L
ipto
w &
Till
man
, 201
2)(P
osen
et a
l., 2
014)
Poly
ethy
lene
(PE)
(Ado
m e
t al.,
201
4)B
rask
em, D
ow/M
itsui
, so
ngyu
an J’
ian
Bio
chem
ical
(De
Jong
, Hig
son,
et a
l., 2
012)
Solv
ay(L
. She
n et
al.,
200
9)Po
lyhy
drox
yal
kano
ate
(PH
A)
(Pat
el e
t al.,
200
6)(H
erm
ann
et a
l., 2
007)
(Tab
one,
Cre
gg, B
eckm
an, &
Lan
dis,
2010
)(Y
u &
Che
n, 2
008)
(Ado
m e
t al.,
201
4)(L
ipto
w &
Till
man
, 201
2)
Met
abol
ic E
xplo
rer
(Met
ex),
Mer
idia
n pl
astic
s (1
03),
Tian
jin G
reen
B
iosi
ence
Co.
(De
Jong
, Hig
son,
et a
l., 2
012)
Bio
mer
, Mits
ubis
hi G
as,
PHD
indu
stria
l, P&
G(P
atel
, Mar
sche
ider
-W
eide
man
n, S
chle
ich,
Hüs
ing,
&
Ang
erer
, 200
5)Ti
nan,
Tel
les,
Kan
eka,
PH
B in
dust
rial
(L. S
hen
et a
l., 2
009)
3
Poly
hydr
oxyb
utyr
ic a
cid
(PH
B)
(Gal
lard
o H
ipol
ito, 2
011)
(Har
ding
, Den
nis,
Von
Blo
ttnitz
, &
Har
rison
, 200
7)(K
im &
Dal
e, 2
008)
(Pat
el e
t al.,
200
6)Po
lyla
ctic
aci
d (P
LA)
(Pat
el e
t al.,
200
6)(G
alla
rdo
Hip
olito
, 201
1)(G
root
& B
orén
, 201
0)(H
erm
ann
et a
l., 2
007)
(Vin
k, R
abag
o, G
lass
ner,
& G
rube
r, 20
03)
Pura
c, N
atur
eWor
ks,
Gal
actic
, Hen
an Ji
dan,
B
BC
A(D
e Jo
ng, H
igso
n, e
t al.,
201
2)C
argi
ll D
ow L
LC, H
ycai
l,To
yota
(Pat
el e
t al.,
200
5)Pr
opan
edio
l (PD
O)
(Ado
m e
t al.,
201
4)(P
atel
et a
l., 2
006)
(Her
man
n et
al.,
200
7)(U
rban
& B
aksh
i, 20
09)
DuP
ont/T
ate
& L
yle
(De
Jong
, Hig
son,
et a
l., 2
012)
p-X
ylen
e(L
in, N
ikol
akis
, & Ie
rape
trito
u, 2
015)
Gev
o, U
OP,
Vire
nt(D
e Jo
ng, H
igso
n, e
t al.,
201
2)aPr
iorit
y ch
emic
al w
ith n
o LC
A re
sults
bA
ssoc
iate
dLC
A re
sults
foun
d fo
r bui
ldin
g bl
ock
cA
ssoc
iate
dLC
A re
sults
foun
d fo
r der
ivat
ive
3
40
Several studies considered multiple production scenarios for an individual chemical; each
scenario is reviewed here as an individual case. Three studies, by Adom et al.,(Adom et
al., 2014) Hermann et al.(Hermann et al., 2007) and Patel et al.(Patel et al., 2006) had the
largest number of discrete cases. These studies assessed life cycle GHG emissions of
several bio-based chemicals from sugar and non-sugar content of biomass resources, by
considering extraction of non-renewable energy sources, agricultural production and
biomass pretreatment, and finally conversion (mostly bio-processing). Patel et al.(Patel et
al., 2006) (reporting results from the BREW project) was among the most comprehensive
LCA studies, supported by many industrial partners, which investigated production of
sixteen different alcohols, carboxylic acids, N-compounds, H2 and polymers from corn
starch, sugarcane and lignocellulosic sources. Among the studied chemicals, bio-based
PHA, PLA, and acetic acid showed higher GHG emissions compared to their
petrochemical counterparts, particularly when maize starch was used as the sugar
source.(Patel et al., 2006)
All energy uses and GHG emission results were scaled to the common functional unit of 1
kg of target chemical. GHG emissions were typically reported using Global Warming
Potential (GWP) 100-year characterization factors, though the values recommended by the
Intergovernmental Panel on Climate Change (IPCC) have been revised over time,
particularly for methane. Results from the literature on life cycle energy use were typically
expressed in one of three metrics: CED, NREU, and fossil fuel input. CED values include
both renewable (biomass, wind, solar, geothermal and water) and non-renewable (fossil,
nuclear) sources while NREU estimations focuses on non-renewable sources, categorized
41
above, and fossil fuel input estimates energy consumption based on fossil fraction of non-
renewable sources. Absolute results for GHG emissions and net energy use are presented
in Appendix A (Table A1). All results were considered as the relative difference between
bio-based chemical and petrochemical equivalents rather than as absolute results, thus
allowing for differences in GWP values and energy metrics used to be compared across
studies.
Results for petrochemical equivalents were sourced from the same studies, where provided.
For cases where comparative results for petrochemicals were not reported, appropriate
fossil-based counterparts were chosen and analyzed, as follows. For five
cases―polyethylene furandicarboxylate (PEF) from starch crops, polyhydroxyalkanoate
(PHA) from corn grain, p-xylene from corn grain, p-xylene from red oak, and styrene from
forest residues―cradle to gate energy use and GHG emission results for corresponding
petrochemical counterparts―polyethylene terephthalate (PET), high-density polyethylene
(HDPE), p-xylene, and styrene, respectively―were estimated using the CED 1.08 method
and IPCC 2013 GWP factors.(Frischknecht et al., 2007) Equivalent energy indicators were
used for comparative analysis. For these cases, petrochemical counterparts are chosen
based on most commonly reported substitutions in literature. Generally, each of the
building blocks may have several counterparts depending on their functionality and end
use purposes. For example, based on collected studies, PHA can substitute high and low
density polyethylene, polystyrene, polypropylene and polylactic acid.(Hermann et al.,
2007; Patel et al., 2006; Yu & Chen, 2008) Table A2 and Table A3 in Appendix A, list
42
each bio-based chemical under consideration with its petrochemical counterparts. Thirteen
chemicals from Table 2 can be produced from either corn or non-corn feedstocks.
Several GHG reduction thresholds were considered for GHG emissions comparison
between fossil-based and bio-based chemicals. Based on RFS2 thresholds, corn-based
chemicals were compared with a hypothetical 20% reduction threshold (mirroring that
mandated for corn-based biofuels) while non-corn derived chemicals were compared with
a hypothetical 50% reduction threshold with fossil-based counterparts as the baseline. RSB
(10% reduction) and ISCC (35% reduction) thresholds were also included regardless of the
type of feedstock. The same comparative analysis was conducted for life cycle energy
estimates based on available data points. Twelve out of 86 cases did not report energy use
in their LCA results.
Some but not all of the compiled cases accounted for carbon sequestered during biomass
cultivation. Among those studies that considered biogenic carbon, various estimation
methods and accounting methods were used, including the DayCent model, PAS2050, and
simple equivalence with the carbon content of target building block chemical. In order to
maintain a consistent framework for this study, cases that did not originally account for
biogenic CO2 were adjusted by reducing their GHG emissions values by the molar
equivalent of the carbon content of target chemical. Carbon sequestered in bio-based
chemicals can be re-emitted at end-of-life, but it is difficult to apply end-of-life scenarios
consistently and realistically across all bio-based chemicals under study due to multiple
potential end uses across chemical types as well as for individual chemicals; however, a
43
sensitivity analysis was performed that assumed a simple end-of-life scenario across all
chemicals for conversion of all contained carbon to CO2.
Prior LCA studies on biofuels and bio-based chemicals have shown that certain modeling
assumptions can have a decisive effect on overall life cycle results. (Daystar et al., 2015;
Patel et al., 2006; Posen, Jaramillo, & Griffin, 2016; Zaimes & Khanna, 2014; Zaimes,
Soratana, Harden, Landis, & Khanna, 2015) Accordingly, for each of the bio-based
chemicals, specific modeling variables were noted for subsequent statistical analysis:
biomass resource (e.g., corn, sugarcane, switchgrass, algae, woody waste, and pulp and
paper waste streams); conversion method (e.g., catalytic, biochemical, thermochemical,
chemical, and hybrid); location; inclusion of direct and indirect land use change
(dLUC/ILUC); and handling of co-products (e.g., economic allocation, mass allocation, or
system expansion). Reliance on laboratory-scale versus commercial-scale data was also
considered. Several statistical tests were performed to investigate the influence of these
variables on the GHG emissions and NREU results, including Analysis of Covariance
(ANCOVA) and 1-Way Analysis of Variance (ANOVA). NREU was chosen as the
primary measure of life cycle energy use because more than half of the cases used this
metric for their analysis.
In addition, covariates of molecular complexity or molecular weight were also investigated
for statistically significant effects on the mean absolute or relative GHG emissions and
NREU. In this context, absolute GHG emissions were defined as life cycle GHG emissions
in units of carbon dioxide equivalent normalized per kg of bio-based chemical (kg CO2
44
eq/kg chemical), while relative GHG emissions were defined as the percent change in life
cycle GHG emissions of the bio-based chemical relative to a standard reference
petrochemical. Similarly, absolute NREU is defined as non-renewable energy use per kg
of bio-based chemical (MJ-NREU/kg chemical), while relative NREU is defined as the
percent change in non-renewable energy use of the bio-based chemical relative to a
standard reference petrochemical. In addition, several measures have been proposed to
quantify the complexity of a molecule based on its structure, bond connectivity, diversity
of non-hydrogen atoms, and symmetry; including the Bertz Index,(Bertz, 1981) the
Bonchev-Trinajstic Index,(Bonchev & Trinajstić, 1977) and Randic Index.(Randić &
Plavs̆ić, 2003) These information-theoretic indices characterize the complexity of
chemical compounds and are generally based on the concept of Shannon entropy. In this
study, values for the molecular complexity of specific compounds were obtained online via
PubChem, and are provided in Appendix A, Table A4. ANCOVA and 1-Way ANOVA
tests were performed using the statistical software package Minitab v.17; for all statistical
tests the significance threshold was set at =0.10. For statistically significant factors, post
hoc multiple comparisons using Tukey’s test were performed to determine if pairwise
differences between factor level means are statistically significant, and the family error rate
for post hoc tests was set at =0.10.
2.3. Results and Discussion Cradle-to-gate energy use (NREU, CED, and fossil fuel input) and GHG emission results
were identified for eighty-six (86) discrete cases. Figure A1 in Appendix A shows the
increasing number of bio-based chemical LCA studies from 2005, the year that the RFS
45
program was established. Among the priority bio-based chemicals, succinic acid, adipic
acid, polyethylene (including PE, LDPE and HDPE), propanediol and
polyhydroxyalkanoate (PHA) were the most studied chemicals with more than five cases
each. No LCA results could be found for the carbohydrate-based chemicals aspartic acid,
sorbitol and arabinitol, nor for the lignin-based chemicals biphenyl, cyclohexane,
cresol/resorcinol, and vanillic acid, revealing significant gaps in the literature. These gaps
are particularly notable considering the identification of these compounds by the US DOE
as priority bio-based chemicals. In general, technological options have been more
thoroughly compared for carbohydrate-based chemicals than for lignin-based chemicals,
likely due to the former’s greater variety of potential feedstocks and conversion methods
and actual production capacity.(Smolarski, 2012)
Reported values for cradle-to-gate life cycle GHG emissions of different bio-based
chemicals were compared with their petrochemical counterparts and plotted against
hypothetical thresholds for GHG reduction in Figure 4. Chemicals listed in Table 2 have
been reorganized into carbohydrate-based (corn and non-corn, with 20% and 50%
emissions reduction thresholds, respectively) and lignin-based (with a 50% emissions
reduction threshold). Two thresholds of 10% and 35% GHG reduction, were also
considered representing existing standards for bio-based chemicals. Error bars represent
the full range of relative GHG emission values reported for each of the chemicals- with
negative values being reduction potential. Solid dots, on the other hand, represent average
values for the reported data.
46
Figure 4- Percent change in life cycle GHG emissions of (a) chemicals derived from carbohydrate content of corn feedstock, (b) chemicals from lignin content of biomass feedstocks, and (c) chemicals derived from carbohydrate content of non-corn feedstocks, compared to their petrochemical counterparts. Dashed lines present GHG reduction thresholds for each category compared to the fossil-based counterparts. Note: the range shown in each figure represents relative GHG values with negative numbers indicating GHG emissions reductions and positive numbers indicating GHG emissions increases.
As illustrated in Figure 4(a) for carbohydrate-based chemicals from corn, relative GHG
emissions results varied from a >300% increase for p-xylene production from corn (with a
mean value of 371% increase in GHG emissions) to a >100% decrease for PHB production
from corn (with a mean value of 177% decrease in GHG emissions), when compared to
their fossil-based counterparts. For the corn-based chemicals, Figure 4(a), PHA and p-
xylene data showed wide ranges of reported values compared to the average, while most
of the other chemicals in this category had their results distributed within the 50% of the
47
average values. Based on reported results, succinic acid, ethyl lactate and PHB had the
largest potential for GHG emissions reduction (84%, 87% and 177% reductions,
respectively) when using corn as feedstock, while p-Xylene showed significant increase in
GHG emissions compared to its petrochemical counterpart. Based on collected data, more
than half of the chemicals in this category meet all three GHG reduction thresholds (RFS,
RSB, and ISCC). Carbohydrate-derived glucaric and glutamic acids were studied not as
target chemicals but as intermediates for the production of adipic acid and N-
methylpyrollidone.(Diamond, Murphy, & Boussie, 2014; Lammens, Potting, Sanders, &
De Boer, 2011) Results for these chemicals showed decreases in GHG emissions compared
to corresponding petrochemicals, but similar results were not available for production of
glucaric and glutamic acids.
Figure 4(b) presents the results for GHG change of lignin-derived chemicals. The RFS
threshold for this group was 50% reduction, since all of the collected cases were sourced
from agricultural and forest residues known as non-corn feedstock. Three out of five
chemicals with reported results in this category were studied in a single study while phenol
and vanillin both had two sets of results. (GHG results for phenol were within 10% of the
average value, so the range of reported results was not wide enough for error bars to be
visible.) Bio-based adipic acid and phenol had the highest and the lowest potential in GHG
emission reduction, 143% and 35%, respectively. No reported values were found for
lignin-derived biphenyl, cyclohexane, cresol or vanillic acid. Other chemicals in this
category had two data points at most, which make the average results less reliable and
emphasize the need for more LCA studies in this category. Vanillin, methanol, styrene, and
48
adipic acid were reported to have more than 50% reduction (Goeppert et al., 2014; Van
Duuren et al., 2011; Y. Zhang et al., 2014) while phenol was shown to have less potential
for GHG emission reduction. However, all of the chemicals in this category meet RSB and
ISCC GHG reduction thresholds.
Figure 4(c) presents the life cycle GHG results for carbohydrate-based chemicals produced
from non-corn feedstocks. PHA, in this category, demonstrated highly varied GHG results,
which can be interpreted by the features of production pathways. Based on Patel et al.,
fermentation is the primary conversion method for this chemical, followed by various
downstream processing such as solvent extraction, oxidation, homogenization, enzymatic
solubilization or solvent extraction and enzymatic solubilization, combined.(Patel et al.,
2006) Evaluation of production pathways showed that synthesis of mid-chain length PHA
from fermented dextrose using oxidizing agents minimizes GHG emissions.(Patel et al.,
2006) This production pathway represents the lower end for reported GHG estimates. The
high boundary corresponds to solvent extraction of fermented rapeseed oil. Low PHA level
(up to 8% PHA/dry weight) in rapeseed oil along with coproduction of significant amount
of residues in solvent extraction process, led to high levels of GHG emissions.(Patel et al.,
2006) Reported GHG emissions of PHB, propionic acid, and succinic acid, on the other
hand, were distributed within 30% of their average values. Among the chemicals included
in this category, sorbitol, arabinitol, and aspartic acid had no LCA results at all, while PEF,
PHB, propionic acid, PHA, butadiene, acetic acid, p-xylene and adipic acid showed less
than 50% reduction in GHG emissions, on average; However, PEF and PHB met both RSB
49
and ISCC thresholds. Average values for the remaining chemicals showed more than 50%
reduction in life cycle GHG emissions.
As mentioned earlier, several parameters such as choice of feedstock, conversion method,
and co-product handling can have significant effects on life cycle emissions and energy
use for bio-based chemicals. Results for LDPE provide a useful case study to this effect.
According to reported results, non-corn LDPE can meet all three GHG reduction thresholds
but the estimates vary significantly across studies, and hence highlight the sensitivity of
results to the above parameters. Posen et al. in a series of studies (Posen et al., 2014, 2016)
examined variation in results for GHG emissions due to uncertainties in modeling
parameters. The authors showed that ethylene and polyethylene production from cellulosic
and advanced feedstocks (sugarcane and switchgrass in particular) can result in lower
emissions than their fossil-based counterparts, but these results have high uncertainty
mainly due to limited data for commercial-scale production. Corn-based PE on the other
hand, shows higher relative GHG emissions and more confident final results because of
the data availability in large-scale. For each of the mentioned feedstocks, fertilizer N2O
emissions, land use change and co-production of on-site energy from residues, cause
significant variations in estimated GHG savings.(Posen et al., 2014, 2016)
Considering life cycle energy use, comparative results between energy use values (CED /
NREU / fossil energy input) demonstrated wide ranges of estimates for both sugar-based
and lignin-based chemicals (Figure 5). As mentioned earlier, energy use of both bio-based
and fossil-based chemicals were compared based on equivalent indicators. For
50
carbohydrate-based chemicals, PHB from corn and xylitol from non-corn feedstock, had
the highest reduction in consumption of non-renewable energy sources (>85%), while
styrene with about 100% reduction was the most favorable compound among lignin-based
chemicals. Average values of energy use reported for both sugar-based and lignin-based
chemicals varied from 97% reduction for PHB to more than 100% increase for propionic
acid, PEF and p-xylene. PDO, acetic acid, p-xylene, PHB and adipic acid had a wide range
of results due to different sources and conversion methods. The expectation is that
chemicals with less non-renewable energy use demonstrate lower GHG emissions, as well.
However, this correlation depends on other factors such as conversion pathway or co-
product handling method.
51
Figure 5- Relative NREU values for (a) chemicals derived from sugar content of corn feedstock, (b) chemicals derived from sugar content of non-corn feedstocks and (c) chemicals derived from lignin content of non-corn feedstocks, compared to their petroleum counterparts. Note: the range shown in each figure represents relative GHG values with negative numbers indicating GHG emissions reductions and positive numbers indicating GHG emissions increases.
52
Figure 6- Life cycle energy use (NREU, CED and fossil fuel input) vs. GHG emissions for bio-based chemicals
Figure 6 presents the relationship between absolute values of GHG emission and indicators
of life cycle energy use. Blue and orange dots represent sugar-based chemicals while green
dots show lignin-based compounds. As expected, NREU and fossil fuel input have strong
positive correlations with life cycle GHG emissions (with a slightly higher correlation
coefficient for fossil fuel input). Statistical results for CED have fewer data points and
show a weak linear correlation, perhaps as this indicator includes renewable sources as
well as non-renewable sources in estimating life cycle energy use. Corresponding linear
regression equations are shown in Figure 6, with 95%-confidence intervals for the slope
of regression line are demonstrated using the curved bands.
53
Table 3 provides a summary of ANCOVA and 1-way ANOVA results for model
parameters. The results from Table 3 indicate that for response variable GHG emissions
only factor ‘Conversion Platform’ is shown to be statistically significant at the 90%
confidence level, while for response variable non-renewable energy use factors
‘Conversion Platform’, ‘LCA Coproduct Handling Method’, and ‘Land Use Change’ are
significant, i.e., the p-values for these factors are less than the significance level (α=0.10).
In total, these results indicate that the choice of ‘Conversion Platform’ has a statistically
significant effect on mean life cycle GHG emissions, while the choice of ‘LCA Coproduct
Handling Method’ has a statistically significant effect on mean non-renewable energy use.
This is important as the choice of LCA scheme for handing coproducts is subjective, and
contingent on the judgment of the LCA practitioner, yet can highly influence the results.
Additionally, statistically significant differences in the environmental performance
between conversion platforms can help guide and prioritize research into specific
conversion and upgrading technologies. Accordingly, Tukey tests were performed to
determine if pairwise differences between factor level means are statistically significant.
For factor ‘Conversion Platform’ and response variable absolute greenhouse gas emissions,
Tukey tests reveal that the means for factor levels ‘Biochemical’ as well as ‘Hybrid’ are
statistically different from ‘Thermochemical’. Moreover, grouping information using the
Tukey method indicate that factor level means for ‘Thermochemical’ platforms are
comparatively higher than that of ‘Biochemical’ or ‘Hybrid’, (6.68 kg CO2e/kg as
compared to 2.02 and 0.90 kg, respectively), detailed results are provided in Appendix A,
see Table A5 and Table A6. For factor ‘LCA Coproduct Handling Method’ and response
variable relative non-renewable energy use, Tukey tests reveal that factor level means for
54
‘Mass’ are statistically different from ‘Hybrid’, detailed results are provided in Appendix
A, see Table A26 and Table A27. These results reinforce the need for a standardized
approach for dealing with coproducts in a life-cycle framework, so as to accurately
benchmark the sustainability of bio-based chemicals, and to provide a fair basis of
comparison between LCA studies. Detailed 1-way ANOVA results for ‘Conversion
Platform’ and ‘LCA Coproduct Handling Method’ is provided in Table 4 and Table 5,
respectively.
Table 3- ANCOVA and ANOVA summary results for bio-based chemicals meta-data
Parameter Covariate or Factor
Factor Levels Response Variable P-
value
Statistically Significant (α=10%)
Complexity Covariate - GHG Absolute 0.525 No Complexity Covariate - GHG Relative 0.788 No Molecular Weight Covariate - GHG Absolute 0.106 No Molecular Weight Covariate - GHG Relative 0.91 No Feedstock Factor 13 GHG Absolute 0.933 No Feedstock Factor 13 GHG Relative 0.184 No Composition Factor 2 GHG Absolute 0.499 No Composition Factor 2 GHG Relative 0.415 No Conversion Platform Factor 5 GHG Absolute 0.087 Yes Conversion Platform Factor 5 GHG Relative 0.77 No Geography Factor 5 GHG Absolute 0.242 No Geography Factor 5 GHG Relative 0.954 No LCA Coproduct Handling Method Factor 4 GHG Absolute 0.439 No LCA Coproduct Handling Method Factor 4 GHG Relative 0.742 No Land Use Change Factor 3 GHG Absolute 0.511 No Land Use Change Factor 3 GHG Relative 0.274 No Complexity Covariate - NREU Absolute 0.12 No Complexity Covariate - NREU Relative 0.874 No Molecular Weight Covariate - NREU Absolute 0.363 No Molecular Weight Covariate - NREU Relative 0.26 No Feedstock Factor 13 NREU Absolute 0.214 No Feedstock Factor 13 NREU Relative 0.367 No Composition Factor 2 NREU Absolute 0.83 No Composition Factor 2 NREU Relative 0.68 No Conversion Platform Factor 4 NREU Absolute 0.954 No Conversion Platform Factor 4 NREU Relative 0 Yes
55
Geography Factor 5 NREU Absolute 0.689 No Geography Factor 5 NREU Relative 0.809 No LCA Coproduct Handling Method Factor 4 NREU Absolute 0.757 No LCA Coproduct Handling Method Factor 4 NREU Relative 0.075 Yes Land Use Change Factor 3 NREU Absolute 0.585 No Land Use Change Factor 3 NREU Relative 0.027 Yes
A growing body of scientific work has suggested that GHG emissions resulting from
changes in the above and below-ground carbon pools as well as soil organic carbon cycles
as a result of direct or indirect transformation of land coverage may negate the carbon
neutrality of bio-based products.(Fargione, Hill, Tilman, Polasky, & Hawthorne, 2008;
Searchinger et al., 2008) As such, ANOVA tests were performed to determine if the
inclusion of land-use change impacts had a statistical effect on mean GHG emissions for
bio-based chemicals. Twelve studies out of the 86 discrete cases evaluated in this study
included LUC impacts, and highlight the large variability in scope and system boundary
between cases; however, results from Table 3 indicate that incorporation of LUC impacts
did not have a statistically significant effect on mean GHG emissions estimates. It is
important to note that the results of this analysis are constrained by a relatively small
sample size. As such, additional statistical findings may be gained as more data becomes
available in the literature. Detailed ANOVA and ANCOVA results for all parameters are
provided in Appendix A, see Table A7- Table A23 and Table A30-Table A52.
56
Table 4- 1 Way Analysis of Variance (ANOVA) for factor, ‘Conversion Platform’ for response variable absolute greenhouse gas emissions
Source DF Adj. SS Adj. MS F-Value P-Value Conversion Platform 4 162.1 40.54 2.11 0.087 Error 79 1516 19.19 Total 83 1678.1
DF: Degrees of Freedom; Adj. SS: Adjusted Sum of Squares; Adj. MS: Adjusted Mean Squares Response Variable: Greenhouse Gas Emissions (Absolute) Factor: Conversion Platform; Factor Levels: Biochemical, Catalytic, Chemical, Hybrid (i.e., a combination of conversion strategies), and Thermochemical Table 5- 1 Way Analysis of Variance (ANOVA) for factor, ‘LCA Coproduct Handling Method’ for response variable relative non-renewable energy use
Source DF Adj. SS Adj. MS F-Value P-Value LCA Coproduct Handling Method 3 3.247 1.0825 2.49 0.075 Error 39 16.959 0.4348 Total 42 20.206
DF: Degrees of Freedom; Adj. SS: Adjusted Sum of Squares; Adj. MS: Adjusted Mean Squares Response Variable: Non-renewable Energy Use (Relative) Factor: LCA Coproduct Handling Method; Factor Levels: Economic, Mass, System Boundary Expansion, Hybrid (i.e., a combination of two or more) Two other factors were considered in this meta-analysis. The first is the use of laboratory-
scale versus commercial-scale data in the original LCA studies. Scale is an important
consideration in LCA modeling, as commercial facilities tend to be better integrated and
optimized, for example using solvent recovery processes and on-site energy production in
large-scale plants, which tends to result in lower energy use and GHG emissions compared
to laboratory results. In this review, only 13 out of 86 collected cases were found to have
relied on bench-scale production for their LCI data. A corresponding statistical analysis
indicated that “Plant Capacity” is statistically significant for both absolute (p-value =
0.022) or relative GHG emissions (p-value = 0.095) estimates. For absolute GHG
emissions, Tukey tests find that factor levels "Pilot Scale" and "Commercial Scale" are
statistically different while for relative GHG emissions, Tukey tests do not find any
57
significant differences in factor level means. Table A47 - Table A52 in Appendix A show
the results of the analysis.
Finally, a sensitivity analysis was performed for the expansion of scope from cradle-to-
gate to cradle-to-grave to see if consideration of end-of-life (EOL) shifts the environmental
preference or causes bio-based chemicals to miss threshold values for GHG emissions
reductions. A single end-of-life scenario was applied so that, for both bio-based and fossil-
based chemicals, the carbon content of the chemicals is assumed to be released as CO2. For
those cases where the bio-based chemicals are identical to their fossil-based counterparts
(51 cases), these emissions from EOL are the same. For those cases for the bio-based
chemicals which were compared with functionally but not chemically equivalent
counterparts (30 cases), CO2 emissions from degradation of bio-based chemicals were
found to be lower than those of the counterparts in all cases (details in Table A53 of the
Appendix A). This will increase the advantage of bio-based chemicals in absolute terms;
however, EOL emissions generally make up a larger proportion of cradle-to-grave GHG
emissions for bio-based chemicals than for fossil-based counterparts, which can reduce the
advantage of bio-based chemicals in relative terms. These relative results for cradle-to-
grave GHG emissions values are reported in Table A54. Ideally, in a more application-
specific context, the length of the use phase and the actual end-of-life disposition of bio-
based chemicals would be known so that the benefits of long-term carbon storage could be
assessed.
58
In summary, this review revealed that the majority of LCA studies on bio-based chemicals
have focused primarily on sugar-based chemicals, while comparatively little attention has
been placed on lignin-derived chemical compounds. Analysis revealed that most, but far
from all, bio-based chemicals were able to achieve RSB, ISCC and RFS2-like reductions
in GHG emissions relative to baseline petrochemicals. Further, statistical analysis revealed
that the choice of conversion platform and LCA coproduct handling method had
statistically significant effects on mean GHG emissions and NREU estimates, respectively.
Furthermore, the system boundary, scope of the analysis carbon-accounting scheme, and
the choice of petrochemical counterpart play an important role in our findings. In order to
create a consistent platform for integration of LCA cases, the system boundary of this
study, was set to be cradle-to-gate excluding GHG emissions and energy use during use
phase and end of life of building blocks. Biogenic carbon was considered for the bio-based
chemicals while scope and boundaries of the fossil-based counterparts were adapted from
the reference literature. However, for specific studies of LCAs of bio-based chemicals with
known application, life time and end of life scenario, current results can be further
improved by accounting for GHG emissions from landfill or incineration processes, and
using more accurate methods for estimation of biogenic carbon such as DayCent and
PAS2050. (BSI, 2011; Necpálová et al., 2015)
In light of these findings, several recommendations are provided for future work. First,
given the lack of available data, future assessment work should emphasize bio-based
chemicals from lignin-based sources. Further, chemicals derived from sugar and lignin
content of non-corn feedstock may provide lower GHG emissions related to baseline
59
petrochemicals and merits further investigation. Second, this work shows that the choice
of LCA coproduct handling method has a non-trivial impact on non-renewable energy use
estimates. As such, a standard allocation method should be agreed upon and applied for
bio-based chemicals in order to report and corroborate results between studies. In the
context of RFS for biofuels, the recommended LCA method for coproduct handling is
avoiding allocation using system expansion.(USEPA, 2007) However, research has shown
that system expansion can produce distorted LCA results for biofuel systems in which
coproducts constitute a significant fraction of total economic value, energy flow, or mass
flow.(Wang, Huo, & Arora, 10; Zaimes & Khanna, 2014; Zaimes et al., 2015) To avoid
such pitfalls, it is recommended that LCA practitioners, sustainability scientists, and the
chemicals industry collaborate to form a consensus on a standardized LCA approach to
account for coproduct flows for bio-based chemicals, perhaps through the creation of
industry-wide product category rules. Third, estimations of potential GHG reductions are
dependent on the choice of conversion platform, thus categorical differences between
conversion platforms may be taken into account for a potential Renewable Chemical
Standard. Fourth, single metric-based policies fail to capture broader environmental
externalities, such as ecological or health-related trade-offs, and may result in unintended
environmental consequences. Accordingly, multiple LCA metrics should be concurrently
analyzed to ensure that biochemical production does not shift environment impacts across
domains or outside of the analysis boundary. For example, a single score LCA study for
biofuels production by Daystar et al.(Daystar et al., 2015) found that impact categories
other than GHG emissions such as ecotoxicity, carcinogenics and non-carcinogenics,
largely determined the score values and as a result the environmental preference of target
60
fuels. Finally, while bio-based chemicals have the potential for GHG reductions relative
to their petrochemical equivalent, further collaboration between industry leaders,
sustainability scientists, and policy makers are needed to assess the technical and
commercial feasibility as well as broader environmental consequences of a potential
Renewable Chemicals Standard.
61
Chapter 3:
Life Cycle Assessment of Catechols from Lignin Depolymerization This study has been published
Montazeri, M., & Eckelman, M. J. (2016). Life Cycle Assessment of Catechols from Lignin Depolymerization. ACS Sustainable Chemistry & Engineering, 4(3), 708-718.
Lignin is the second most abundant natural polymer on Earth. The aromatic structure of
lignin makes it a promising platform for bio-based chemicals. Catalytic depolymerization
of lignin has been demonstrated with high yields and selectivity, resulting in efficient
conversion to target products. In this study, we performed a comparative process
simulation and life cycle assessment (LCA) of catechol-derived products from lignin
contained in candlenut shell with those conventionally derived from petrochemical phenol.
The modeled bio-based production pathway includes candlenut cultivation, nutshell
separation and preparation, lignin extraction and purification, catalytic depolymerization
of lignin, and catalyst synthesis. Commercial-scale process modeling was done in ASPEN
Plus based on experimental data, while life cycle environmental burdens were modeled
using the USEPA’s TRACI 2.1 impact assessment method, covering ten categories of
resource use and impact. Comparison of bio-based and fossil-based results showed an
overall reduction in environmental impacts for the lignin route of 2%, 7%, and 59% in
global warming potential, ecotoxic effects, and fossil fuel depletion, respectively. In other
environmental impact categories, particularly ozone depletion, the fossil-based route was
shown to be preferable. Dichloromethane, used as solvent in purification of extracted
lignin, and electricity use during depolymerization of lignin are the dominant contributors
in total environmental burdens of bio-based route. A complementary analysis was
62
conducted to consider the relative impacts of an alternate extraction method. The overall
results emphasize the need for further work in developing conversion processes and also
considering several parallel scenarios to find the most beneficial use of lignin on a life
cycle basis.
3.1. Introduction
Lignocellulosic biomass is the most abundant renewable biological resource on Earth with
a yearly growth of 200 billion tons.(Y.-H. P. Zhang, 2008) Lignin accounts for
approximately 25–35 % of the organic matrix of wood.(Kleinert & Barth, 2008) Lignin
binds cellulose-hemicellulose matrices while adding flexibility. The molecular structure of
lignin polymers has significant diversity but the primary structure is aromatic (benzene
rings with methoxyl, hydroxyl, and propyl groups) interconnected by
polysaccharides.(Paster, Pellegrino, & Carole, 2003) Long chain lignin as shown in Figure
7 has three monolignol monomers: p-coumaryl alcohol, coniferyl alcohol and sinapyl
alcohol, that may serve as bio-based platform chemicals.(Norman, 1969)
Figure 7- Lignin polymer and three main monomers (adapted from http://www.ir
nase.csic.es)
Coumaryl Coniferyl Sinapyl
63
There are several current and potential sources of lignin. Energy crops and their residues
such as stalks and stover, woody residues from agriculture and forestry, and paper wastes
are major resources. Each of these resources has different lignin content, with a potentially
different chemical structure, and thus can be targeted for production of specific bio-based
chemicals. Table 6 lists different resources, their lignin contents and isolation methods,
major supplier countries, and present production capacity based on total mass produced or
total area harvested for each resource.
Tab
le 6
-Glo
bal l
igni
nre
sour
ces a
nd c
urre
nt p
rodu
ctio
n/cu
ltiva
tion
leve
ls
Res
ourc
e C
ateg
ory
Lign
in S
ourc
eLi
gnin
Con
tent
(wt%
)Li
gnin
Isol
atio
n M
etho
dM
ajor
Pro
duce
r7
(“FA
OST
AT,
” 20
13)
Prod
uctio
n/
Cul
tivat
ion
Am
ount
(Mt)
Har
dwoo
d
Euca
lypt
us
24.4
( Kaw
aoka
, Nan
to, I
shii,
&
Ebin
uma,
200
6)26
.91
(USD
OE,
201
5)
Kla
son
ligni
nTo
tal l
igni
n (A
STM
E -
1721
&T-
250)
--
Popl
ar
25.6
(San
nigr
ahi,
Rag
ausk
as, &
Tu
skan
, 201
0)24
.8(U
SDO
E, 2
015)
C N
MR
Tota
l lig
nin
(AST
M
E-17
21 &
T-25
0)C
anad
a28
,300
(Mha
cul
tivat
ed)
Will
ow25
.6(D
unfo
rd, 2
012)
-R
ussi
a2,
850
(Mha
cul
tivat
ed)
Softw
ood
Bam
boo
26.8
(Sek
yere
, 199
4)A
cid
inso
lubl
e lig
nin
Chi
na1,
230
Lobl
olly
pin
e
28(Z
hu &
Pan
, 201
0)25
.9(U
SDO
E, 2
015)
- Tota
l lig
nin
(AST
M
E-17
21 &
T-25
0)-
-
Spru
ce28
.3(Z
hu &
Pan
, 201
0)K
laso
n lig
nin
--
Agr
icul
tura
l pro
duct
s/
resi
dues
Coc
onut
shel
ls36
-44
(Men
du e
t al.,
201
1)A
cid
inso
lubl
e lig
nin
Indo
nesi
aTo
tal c
ocon
ut: 2
1.6
Dry
cot
ton
stem
s>4
0(B
ell,
1986
)C
hina
Tota
l cot
ton:
6.8
6
Ric
e hu
sks
34(N
dazi
, Nya
hum
wa,
&
Tesh
a, 2
008)
U.S
.A.
Tota
l ric
e: 2
04H
usk
(wt%
): 20
%(S
antia
guel
, 201
3)
Suga
rcan
e ba
gass
e
22(B
oopa
thy
& D
awso
n,
2008
)24
.09
(USD
OE,
201
5)
Tota
l lig
nin
(AST
M
E -17
21 &
T-25
0)B
razi
l
Tota
l sug
arca
ne: 7
34B
agas
se (w
t%):
17%
(Cha
ndel
, da
Silv
a,
Car
valh
o, &
Sin
gh,
2012
)
Cor
n st
over
7-21
(Red
dy &
Yan
g, 2
005)
20.2
4(U
SDO
E, 2
015)
Lign
osul
fona
tes t
o K
raft
ligni
nTo
tal l
igni
n (A
STM
E-
1721
&T-
250)
U.S
.A.
Tota
l cor
n: 2
37C
orn
to re
sidu
e ra
tio: 1
:1(“
BIO
SAT,
” 20
11)
Whe
at st
raw
16(D
el R
ío e
t al.,
201
2)16
.85
(USD
OE,
201
5)
Kla
son
ligni
nTo
tal l
igni
n (A
STM
E -17
21 &
T-25
0)U
.S.A
.
Tota
l whe
at: 1
26Le
af: 2
5 -50
%
Stem
: 45-
70%
(Y.-H
. P. Z
hang
, 200
8)
Bar
ley
stra
w11
(Mac
greg
or, 2
000)
Aci
d in
solu
ble
ligni
n
Aus
tralia
Tota
l bar
ley:
17
Bar
ley
to re
sidu
e ra
tio: 1
:1.2
(“B
IOSA
T,”
2011
)
Cor
n st
alks
9.3
(Lec
hten
berg
, C
olen
bran
der,
Bau
man
, &
Rhy
kerd
, 197
4)
U.S
.A.
47
Alfa
lfa si
lage
8.4
(Del
Río
et a
l., 2
012)
U.S
.A.
-
Alfa
lfa h
ay7.
6(D
el R
ío e
t al.,
201
2)U
.S.A
.-
6
Cot
ton
fiber
s<1
(Fan
, Hu,
Yan
g, &
Li,
2012
)
Kla
son
ligni
nU
.S.A
.3.
5
Soyb
eans
-C
hina
82So
ybea
n m
eal
-C
hina
57
Pere
nnia
ls
Switc
hgra
ss
16.8
(San
nigr
ahi e
t al.,
201
0)17
.56
(USD
OE,
201
5)
C N
MR
Tota
l lig
nin
(AST
M
E-17
21 &
T-25
0)-
-
Mis
cant
hus
10-3
0(B
ross
e, D
ufou
r, M
eng,
Su
n, &
Rag
ausk
as, 2
012)
C N
MR
--
Swee
t sor
ghum
12-2
0(C
arls
on, C
arr,
&
Cun
ning
ham
, 198
3)11
.34
(USD
OE,
201
5)
Aci
d-so
lubl
e lig
nin
Tota
l lig
nin
(AST
M
E-17
21 &
T-25
0)U
.S.A
.6
Gra
in so
rghu
m
10.3
(Frit
z, C
antre
ll,
Lech
tenb
erg,
Axt
ell,
&
Her
tel,
1981
)
U.S
.A.
6
6
67
The first step in biomass processing for bio-based chemicals is separation of cellulose,
hemicellulose, and lignin fractions. There are several methods currently available for this
process including steam explosion, liquid hot water, dilute acid, ammonia fiber explosion,
alkali, and organosolv pretreatment; however, only a few of these separate lignin rather
than destroying its structure and can thus be used to extract lignin fraction of lignocellulosic
biomass.(Sherman & Gorensek, 2011) Conde-Mejía et al.(Conde-Mejía, Jiménez-
Gutiérrez, & El-Halwagi, 2012) looked at five extraction methods mentioned earlier, for
bioethanol production from lignocellulosic biomass. The results showed steam explosion
treatment to be the most efficient method for carbohydrate conversion (85% conversion
yield) with energy cost of about 19 $/ton of dry biomass (not considering recycling of water
and material). Fractional conversion of biomass was also included in the same
study(Conde-Mejía et al., 2012), displaying that organosolv and alkali/LIME pretreatment
methods are capable of separating lignin partially. Conversion yields for lignin are reported
to be 74% and 15%, for organosolv and LIME extraction, respectively. Since these
methods can extract all three components of woody biomass moderately, output lignin
stream needs purification in order to eliminate unwanted components. An example of this
process is described in detail in Methods section. There are other pretreatment methods
designed for lignin separation such as filtration(Arkell, Olsson, & Wallberg, 2014;
Toledano, García, Mondragon, & Labidi, 2010) and solvent extraction.(Sherman &
Gorensek, 2011) These methods aim for lignin fraction of woody biomass in particular and
result in a relatively pure lignin stream.(Jørgensen, Vibe‐Pedersen, Larsen, & Felby, 2007;
D. Mohan, Pittman, & Steele, 2006; Sun & Cheng, 2002)
Transformation/depolymerization of lignin is the next step. Here, the goal is breaking down
68
the lignin polymer while preserving the phenolic structure for production of complex
aromatic chemicals. Most routes currently have low conversion factors and are relatively
energy intensive.(Barta, Warner, Beach, & Anastas, 2014; Kleinert & Barth, 2008) In
general, there are several options for thermochemical transformation of biomass for
production of bio-based chemicals, such as pyrolysis for bio-oil and bio-chemical
production, liquefaction and acidic/basic hydrolysis for bioethanol production,(Jørgensen
et al., 2007; D. Mohan et al., 2006; Sun & Cheng, 2002) and more selective
depolymerization through catalytic processes(Barta et al., 2014; Huber, Iborra, & Corma,
2006) for bio-chemical production.(Zakzeski et al., 2010)
Potentially several chemicals can be produced from lignin such as syngas products
(methanol and dimethyl ether), hydrocarbons (BTX and higher alkylates, cyclohexane,
styrene and biphenyl), phenols (phenol and catechol), oxidized products (vanillin, aromatic
and aliphatic acids, cyclohexanol) and macromolecules.(Holladay et al., 2007) Current
barriers for industrial scale production of these chemicals are mainly due to the variable
structure of lignin, which can result different final products depending on biomass source
and processing route, thus requiring pre-conditioning of lignin streams.(Holladay et al.,
2007)
Here we considered catalytic depolymerization of lignin from candlenut shells. These
nutshells have 12% (w/w) organosolv lignin in their structure.(Barta et al., 2014) Candlenut
(Aleurites molucanna) is a common agroforestry tree species found throughout Indonesia
and the Asia Pacific region. Products of these trees are seeds with 50% oil content and 30%
69
of product weight.(Norman, 1969) The remainder is mainly thick woody shells which are
typically burned or piled as waste. Target products resulting from extraction and
depolymerization of candlenut shell lignin are catechol derivatives including 4-
propylcatechol, 4-(3-hydroxypropyl) catechol, 2,3-dihydro-1H-indene-5,6-diol and 4-(3-
methoxypropyl) catechol. For modeling purposes tert-butyl catechol (TBC) was chosen as
the representative target chemical. Catechol has the molecular formula C6H4(OH)2 and is
primarily used in the production of pesticides, the remainder being used as a precursor to
fine chemicals.(Fiege et al., 2000) Worldwide consumption of catechol is estimated to be
about 20,000 metric tons, and is mainly produced in France, Japan, Italy and the UK.
(Krumenacker, Costantini, Pontal, & Sentenac, 1995) In the present study, process
simulation coupled with life cycle assessment (LCA) was conducted in order to compare
energy and environmental impacts associated with this bio-based production scheme at
industrial scale, compared with TBC produced through petrochemical route.
3.2. Methods
3.2.1. Goal and Scope
Every LCA study includes four standard steps,(Klöpffer, 1997) namely goal and scope
definition, inventory analysis, impact assessment, and interpretation. Goal and scope
specifies the aim and depth of the study. Inventory analysis counts for all activities related
to the production of one functional unit and impact assessment is used in order to transform
the quantitative data collected in the inventory table into (potential) impacts.(Horne, Grant,
& Verghese, 2009) This LCA is a field-to-gate analysis, focusing on the life cycle impacts
of catechol derivatives. Functional unit for the study is set to be 1 kg of target chemical,
70
tert-butyl catechol (TBC). Besides the main production steps, all major upstream processes
were considered, including candlenut cultivation, harvesting and transport, nutshell
separation and milling for the bio-based route and oil and gas extraction, transport, and
petrochemical refining for the fossil-based route. Waste treatment of solvents was
considered in separate scenarios for landfilling or incineration of residual solvents. The
environmental burdens of each process step were allocated based on mass of target stream
compared to the residual/side streams, while overall burden of each of the bio-based and
fossil-based routes was estimated accounting for salable co-products using system
expansion. The objective of the work is to evaluate the relative preference of bio-based or
fossil-based TBC across a range of impact categories and to identify specific processes and
material sources that contribute significantly in overall impacts, highlighting opportunities
for research and development.
3.2.2. Process Description
The overall process consists of nut cultivation and harvesting, nutshell preparation
(separation, air drying and milling), lignin extraction with methanol followed by lignin
depolymerization in the presence of Cu-doped porous metal oxide (Cu-PMO) catalyst. An
integrated flow chart of the modeling scope is shown in Figure 8. The data provided in this
study are based on previous experimental work by Barta et al.,(Barta et al., 2014) who
reported on catalytic depolymerization conducted under different temperatures, residence
times, and catalyst doses while final products were specified using gel chromatography.
The highest conversion rate (92%) was reported at 180°C, a reaction time of 14 hours, with
a catalyst-to-lignin weight ratio of 0.5:1.
71
The first modeling step is cultivation of candlenuts. We accounted for fertilizer
consumption as the main contributor in environmental impacts of cultivation. Based on
Food and Agriculture Organization (FAO) data, average per hectare usage of fertilizer on
groundnuts in India (our sample source) is 24.4, 39.9 and 12.9 kg for N, P2O5 and K2O
fertilizers, respectively.(FAO, 2005) Reported production yield is 0.98 ton/ha(Nautiyal,
2002) and a shell weight percentage of 70% is assumed for candlenuts.(Sustainable Tarde
and Consulting, 2009) Carbon sequestration was evaluated from the carbon content of
lignin. The assumed lignin formula of C20H26O6 requires 3.1 kg CO2 sequestered per kg
TBC produced. Field N2O emissions were also considered, using the IPCC value of 1.3%
of N2O-N emissions per unit of N-fertilizer applied.(Bouwman, 1996)
Material and energy use in harvesting and preparation of nutshells were derived from the
developed unit processes of husked nuts and wood chopping in the ecoinvent 3.1 life cycle
inventory database. These two unit processes estimate energy consumption for harvesting
of nuts and transport of them followed by grinding the nutshells in mobile wood choppers.
We excluded drying of nuts since they go through air drying which happens naturally.
Based on experimental data, lignin is extracted through organosolv treatment, a method for
fractionation of lignocellulosic biomass in presence of an organic solvent, usually methanol
or ethanol.(Zhao, Cheng, & Liu, 2009) (An alternate extraction method is considered in
Section 2.6.) Extracted lignin is purified further in the presence of ethyl acetate and
dichloromethane. The purpose of structural purification is to eliminate organic and
inorganic impurities(Vishtal & Kraslawski, 2011) as well as cellulose and hemicelluloses,
72
which constitute an obstacle in depolymerization.(Prado, Erdocia, Serrano, & Labidi,
2012) Lignin extraction and purification were modeled by scaling up bench-scale
processes(Barta et al., 2014) and estimating energy consumption. Reported cooling and
heating energy values for organosolv pretreatment of softwood from Conde-Mejía et
al.(Conde-Mejía et al., 2012) were modified based on the solvent to wood ratio and used
in this analysis.
Ethyl acetate soluble lignin is fed to the depolymerization process, where it is broken down
during a single-step hydrogenation reaction in the presence of methanol and Cu-doped
metal catalyst. As described in the Supporting Information of Barta et al.,(Barta et al.,
2014) ethyl acetate cannot be removed completely even after several steps of lignin
filtration and prolonged drying. So, we considered less than 0.1% of consumed ethyl
acetate reacting in this step with hydrogen. The output stream has TBC as the main product
and ethanol as a co-product. TBC is a representative for four different catechol derivatives
resulted from the experimental analysis. Energy consumption for catalytic
depolymerization, as well as catalyst synthesis were estimated using industrial-scale
simulations.
73
Figure 8- Process flow chart of bio-based production route 3.2.3. Catalyst Preparation
During catalyst synthesis, metal nitrates (aluminum, magnesium and copper) are mixed
with Na2CO3·H2O and converted to metal oxides after calcination at 460 ˚C. A
hydrotalcite-like Cu-doped porous metal oxide (PMO) is produced with no char formation.
Overall retention time for preparation of 1 kg catalyst is about 76 hours.(Hansen, Barta,
Anastas, Ford, & Riisager, 2012) The final product is used in the depolymerization process,
based on 0.5:1 catalyst-to-lignin ratio.
Catalyst Synthesis
Nuts Cultivation & Harvesting
Lignin Extraction
Lignin Purification
Lignin Depolymerization
MethanolExtraction residues
(cellulose, hemi-cellulose and others)
Dichloromethane Ethyl acetate
Na2CO3NaOH
Al(NO3)3Cu(NO3)2
CatecholsEthanol
Inputs Modeled Processes Outputs
Fertilizer
Preparation (Nutshell Milling)
Nuts
74
3.2.4. ASPEN Plus Simulations
Process modeling for catalyst synthesis and lignin depolymerization steps was performed
at commercial scales using ASPEN Plus v8.6 in order to estimate energy use at each step,
based on experimental data previously reported.(Barta et al., 2014; Hansen et al., 2012)
Solvent and residual reactants separation and recovery was modeled fully, however catalyst
separation simulation was excluded due to lack of data in industrial scale. Process flow
diagrams are shown in Figure 9(a) and (b). Resource use (material and energy) for
cultivation, harvesting, extraction, and purification steps were based on literature values
for subsequent LCA modeling.
Data from the NREL biomass ASPEN data bank(Wooley & Putsche, 1996) were used to
estimate the enthalpy and heat capacity of lignin, based on a model compound for lignin
content of lignocellulosic biomass, C7.3H13.9O1.3. The reported value for enthalpy of
formation is ΔHs= -1.6 x 109 (kJ/mol) and heat capacity value is calculated based on
below.(Wooley & Putsche, 1996)
(for C7 <T<C8) Equation 3
where C1=31,400, C2=394, C3=0, C4=0, C5=0, C6=0, C7= 298.15 K, and C8=1000 K.
The proposed molecular formula for candlenut lignin is C20H26O6, based on our
experimental data. Simulations were based on input lignin flow rate of 75 tons/day
(matching current US industrial scales for lignin processing). The input nutshell flow rate
was scaled according to below, resulting in 625 tons/day of biomass.
75
Equation 4
Catalyst lifetime of 5000 hrs.(O. G. Griffiths et al., 2013) and a methanol recycling rate of
99% were assumed during solvent separation. High recycling rate of methanol is due to the
fact that the solvent is not reacting during depolymerization, it only solubilizes input lignin
for the hydrogenation step. Daily production of TBC is 54 tons while ethanol is a salable
co-product of the depolymerization process with daily production of 30 tons. A co-product
credit of 0.44 kg ethanol per 1 kg TBC is therefore assigned.
Figure 9- ASPEN Plus process flow diagrams for (a) catalyst synthesis and (b) lignin depolymerization
(a)
(b)
76
The fossil-based route was also simulated in ASPEN Plus as a two-step process for
production of TBC from phenol (Figure 10). The first step is based on an industrial patent
for production of catechol and hydroquinone from phenol.(Drauz, Kleeman, Prescher, &
Ritter, 1991) The process consists of phenol hydroxylation with hydrogen peroxide in the
presence of SeO2 as a catalyst. Catechol and hydroquinone are components of the output
stream from first reaction with the ratio of 1.8:1, based on below.(Sienel, Rieth, &
Rowbottom, 2000) Complete conversion of H2O2 is assumed.
Equation 5
The second step is making TBC from reaction of catechol and isobutanol in the presence
of xylene and trifluoromethanesulfonic acid as a catalyst(Rajadhyaksha & Chaudhari,
1987), as shown in below. The reported reaction rate is very low, approximately 35%
conversion for catechol.
Equation 6
For both steps, catalyst synthesis was considered as well as reaction and separation of
phenol and xylene from final products. 97% recovery was assumed for unreacted phenol
and xylene streams and catalyst separation based on experimental methods and modeled in
ASPEN Plus.
77
Figure 10- Flow diagram of petroleum-based TBC 3.2.5. Life Cycle Inventory
Life cycle inventories were compiled based on material and energy consumption data.
Chemical inputs were scaled up based on lab-scale data and literature for bio-based and
fossil-based routes, respectively. Recovery of solvents was considered with recycling rates
of 97%, except for methanol used in depolymerization, which has higher recycling rate of
99%. Energy consumption, on the other hand, was estimated based on ASPEN Plus
simulations and literature. The inventories were set for both routes using SimaPro 8.01
LCA software (PRé Consulting, Amersfoort, Netherlands) and ecoinvent 3.1. life cycle
inventory unit processes adjusted for the US energy system (US-EI database, Earthshift,
Huntington, VT). Full LCI tables are provided in Table B1 to Table B4 of Appendix B.
3.2.6. Alternate Extraction Processes
In order to test the sensitivity of the results to the choice of lignin extraction process, a
complementary analysis was performed assuming an alternate extraction method, designed
for lignin separation, solvent extraction. The lignin stream leaving this process is relatively
78
pure so there is no need for further purification. This method can extract 10-87% of lignin
depending on feedstock category and solvent volume-to-feed weight ratio.(Sherman &
Gorensek, 2011) For this study we chose 16:1 ratio with a 60% conversion reported for the
softwood feedstock, loblolly pine. Data for the alternate pretreatment method is sourced
from a US patent(Sherman & Gorensek, 2011) summarized in Table 7. Energy and
chemical use in this method was estimated based on a pilot-plant with the capacity of 1 ton
of dry biomass/day, normalized based on the target product of 1 kg of TBC.
Table 7- Design parameters for alternate lignin extraction methods
Extraction method Energy input Material input
Solvent extraction Electricity: 3.98 kWh Sulfuric acid (kg): 1.13
Ammonium hydroxide (kg): 7.70
3.2.7. Life Cycle Assessment
Life cycle assessment (LCA) is a standardized systems modeling tool that inventories the
emissions and the consumption of resources along a product’s life cycle and links these to
potential environmental and health impacts.(Rebitzer et al., 2004) As described in goal and
scope section, both bio-based and fossil-based routes were modeled based on cradle to gate
LCA, for 1 kg of target product. For the bio-based route, three sets of allocation were
considered which include both mass allocation and system expansion. While for the fossil-
based route, the default economic allocation employed in the ecoinvent LCI database was
used. Table 8 summarizes products and allocation methods used for each of the processes
in bio-based and fossil-based routes.
79
Table 8- Summary of products and allocation methods
Processes Products Allocation Method
Bio-based Route
Nuts cultivation & harvesting Nutshellsa
Nutsb
Mass allocation based on nutshells to nuts weight ratio and lignin content of nutshells
Wood chopping Ground nutshell a Mass allocation based on lignin content of nutshell
Lignin extraction Lignina
Cellulose, hemicellulose, residualsb
Mass allocation based on lignin content of nutshells
Catalyst synthesis Catalyst with lifetime of 5000 hrs.
Allocation over lifetimec
Lignin Depolymerization Tert-butyl catechola
Ethanolb System expansion for ethanol
Fossil-based Route
Catechol Production Catechola
Hydroquinoneb Economic allocation based on market values
Catalysts synthesis Catalyst with 97% recovery Mass allocation based on recovery ratio
a represents the main product from each process b represents the side stream/ residues from each process c ASPEN Plus simulations are based on continuous flow (defined per hour) so the energy consumption is 1/5000th of estimated value
The U.S. EPA’s Tool for the Reduction of Chemical and Other Environmental Impacts
(TRACI) 2.1 life cycle impact assessment model was used for estimation of overall
environmental and human health impacts based on fate-transport-exposure-effect models
developed for the US.(J. Bare, 2011) Investigated environmental impact categories (with
equivalent units in parentheses) are ozone depletion (kg CFC-11 eq.), global warming (kg
CO2 eq.), smog formation (kg O3 eq.), acidification (kg SO2 eq.), eutrophication (kg N eq.),
carcinogenic and non-carcinogenic health effects (CTUh), respiratory effects (kg PM2.5
eq.), ecotoxicity (CTUe) and fossil fuel depletion (MJ).
80
3.3. Results and Discussion ASPEN Plus simulations provided estimates for energy use for industrial scale processes
of depolymerization and catalyst synthesis for bio-based TBC and two-step production
process for petroleum-based TBC. Energy use for catalyst synthesis is estimated as 21.8
kWh/kg Cu-PMO catalyst, while depolymerization requires 10 kWh/kg TBC. The high
energy consumption for catalyst synthesis is mainly due to high temperature and pressure
operating conditions. Energy use for petroleum-based route is estimated as 2.93 kWh/kg
TBC. These values were combined with literature values in the subsequent LCA modeling,
and results for the bio-based and petroleum-based routes are shown in detail in this section.
First we present results for each route separately, broken down by process step. Figure 11
demonstrates the relative process contribution, normalized in each category, for the field-
to- gate production of TBC from lignin content of candlenut shells. Results show that
dichloromethane used for lignin structural purification drives life cycle impacts, followed
by electricity use during depolymerization. Dichloromethane, produced from methyl
chloride at high temperatures (400-500 °C), is highly volatile and an ozone-depleting
substance (characterization factor of 6.7E-5 kg CFC-11 eq./kg emitted). Thus, its
contribution to life cycle ozone depletion is >99% of the total, even assuming a 97%
recovery rate. Elimination of dichloromethane can decrease overall environmental impacts
significantly, from 5% for eutrophication up to 99% for ozone depletion potential.
Electricity use for maintaining operational conditions during depolymerization is the other
major contributor to impacts, both in terms of fossil energy use and for environmental and
health effects stemming from power plant emissions. Impacts of fertilizer consumption are
81
more significant in eutrophication category due to runoff during cultivation process.
Nitrogen fertilizer release as N2O is about 2% of overall GHG emissions. Heat
consumption and ethyl acetate release during extraction and purification processes
contribute significantly in fossil fuel depletion. Contribution of ethyl acetate mainly comes
from its building blocks, ethanol and acetic acid, both producing from fossil consumptive
sources.
Figure 11- Process contribution for 1 kg TBC production, considering nuts cultivation and preparation, lignin extraction and catalytic depolymerization, and catalyst synthesis
Catalyst contribution is not significant compared to other processes, ranging from 0.02%
in ozone depletion category up to 3.5% in non-carcinogenic health effects. Production of 1
kg TBC requires 0.56 kg of catalyst that can be recovered and reused for 5000 hours of
operation. Just considering the catalyst synthesis, electricity use for maintaining reactor
82
conditions contributed more than 50% in all investigated impact categories. A detailed
process contribution chart is provided in Figure B1 of Appendix B.
Life cycle impacts for petrochemical-based TBC were also analyzed, with process
contribution results shown in Figure 12.
Figure 12- Process contribution of TBC production from petroleum based phenol
Phenol production is the dominant contributor in all environmental impact categories as its
production requires significant electricity, heat, and organic chemical feedstocks, primarily
cumene. Hydrogen peroxide shows significant contribution in carcinogenic health impacts
due to upstream tetrachloroethylene and dichloromethane use in its production process.
83
Ozone depletion results highlight the contribution of isobutanol due to the consumption of
high pressure natural gas in its manufacturing process.
Finally, a comparison of fossil-based and lignin-based TBC illustrates energy and
environmental trade-offs in the results. Table 9 shows total impacts for each production
route. Bio-based TBC has higher environmental impacts in ozone depletion, smog
formation, acidification, eutrophication, carcinogenics, non-carcinogenics and respiratory
effects. The complexity and resistance of the lignin structure require intense operational
conditions and strong solvents for conversion to simpler aromatic compounds, which drive
negative impacts across impact categories. As mentioned above, utilization of
dichloromethane as a solvent for purification of crude lignin is a major contributor and its
substitution should be a target for further research. Prado et al. (Vishtal & Kraslawski,
2011) studied different green solvents that can purify lignin, showing IL (ionic liquids),
water and [BMI][MeSO4] to be preferable for organosolv lignin. However, the yield of
purification by IL reported as 60% which is about 30% lower than for the dichloromethane
purification process considered here. In the three impact categories of global warming
potential, ecotoxicity and fossil fuel depletion, lignin-based TBC was shown to be
preferable to the conventional petrochemical route.
84
Table 9- Total environmental burden of lignin-based and petroleum-based TBC
Impact Category Unit Total (lignin-based catechol)
Total (fossil-based catechol)
Ozone depletion kg CFC-11 eq. 1.01E-04 5.67E-07 Global warming kg CO2 eq. 1.34E+01 1.35E+01 Smog kg O3 eq. 9.85E-01 5.59E-01 Acidification kg SO2 eq. 9.88E-02 5.33E-02 Eutrophication kg N eq. 4.36E-02 2.84E-02 Carcinogenics CTUh1 7.35E-07 5.89E-07 Non carcinogenics CTUh 1.06E-06 3.65E-07 Respiratory effects kg PM2.5 eq. 2 8.04E-03 3.72E-03 Ecotoxicity CTUe3 1.80E+01 1.93E+01 Fossil fuel depletion MJ surplus 1.88E+01 4.59E+01
1Comparative toxic units (CTUh), providing the estimated increase in morbidity in the total human population per unit mass of a chemical emitted(Rosenbaum et al., 2008) 2PM2.5 is particulate matter with diameter of 2.5 micrometers or less 3Comparative toxic units (CTUe) that provides an estimate of the potentially affected fraction of species (PAF) integrated over time and volume per unit mass of a chemical emitted (PAF m3 day kg−1) (Rosenbaum et al., 2008)
3.3.1. Solvent Waste Treatment
The treatment and disposal of residual solvents is case dependent and is considered here
through two scenarios. Life cycle inventories for solvent treatment, specifically landfill
and incineration are presented in Table B7 and Table 8, while its contributions to overall
estimated results are presented Table B9 in Appendix B. Including solvent waste
treatment is particularly important for the global warming impact category, as both
landfilling and incineration result in additional GHG emissions. Assuming landfilling of
waste solvents increases environmental impact results for both routes from 0-8% (highest
for global warming), while the relative comparisons between routes are virtually
unchanged. Incineration of waste solvents, however, changes the baseline results more
significantly. Global warming results increase by 30% for the bio-based route and 40% for
85
the fossil-based route, thus enhancing the preference for bio-based catechols when only
considering life cycle GHG emissions.
3.3.2. Alternate Lignin Extraction Method
As described in the Methods section, we looked into an alternate extraction method for
separation of lignin and compared its results with our base case scenario. Figure 13 shows
environmental burdens associated with the production of 1 kg TBC using solvent
extraction. Legend titles are based on the extraction methods but represent burden of the
overall bio-based route including preparation, extraction and depolymerization. Reported
results show significant decrease in all investigated impact categories ranging from 99%
for ozone depletion to 18% for eutrophication potential compared to the early bio-based
route. Observed decrease is mainly due to elimination of purification process and
substitution of energy intensive organosolv extraction method with an efficient method, in
spite of lower conversion factor (60%) reported for the solvent extraction. This new
pathway shows lower impacts compared to the fossil-based route except for eutrophication,
acidification and respiratory effects categories. Sulfuric acid and ammonium hydroxide, as
main intermediates for extraction, drive negative impacts in these three categories.
86
Figure 13- Process environmental burden considering different extraction method 3.3.3. Alternate Lignin Source
In this study, we focused on candlenut shells as our primary lignin resource. But different
sources of lignin can be used for bio-chemical production. Resources with higher lignin
content tend to consume less energy and material for extraction and structural purification
when normalized to the functional unit of 1 kg TBC. Although it is known that different
0.E+00
4.E-05
8.E-05
1.E-04kg
CFC
-11
eq.
0
5
10
15
kg C
O2
eq.
0
0.4
0.8
1.2
kg O
3 eq
.
0
0.04
0.08
0.12
kg S
O2
eq.
0
0.02
0.04
0.06
kg N
eq.
0.E+00
2.E-07
4.E-07
6.E-07
8.E-07
CTUh
0
0.002
0.004
0.006
0.008
0.01
kg P
M-2
.5 e
q.
0
4
8
12
16
20
24
CTUe
0
10
20
30
40
50
MJ s
urpl
us
Bio-based TBC-OS
Bio-based TBC- Solvent Extraction
Fossil-based TBC
0.0E+00
4.0E-07
8.0E-07
1.2E-06
CTUh
Ozone Depletion Potential
Global Warming Potential Smog
Acidification Eutrophication Carcinogenics
Non-carcinogenics Respiratory Effects Ecotoxicity
Fossil Fuel Depletion
87
lignin resources have different types of lignin polymers and may not result in the same
target products and conversion efficiencies under the same process, here we estimate
potential TBC production for reported resources shown in Table 6. The exact products
from each source of lignin must be verified experimentally. Certain agricultural resources
such as coconut shells and cotton stems have approximately 40% lignin in their structure
so in these cases, lignin can be extracted under milder conditions and it is expected to
consume less energy when normalized based on 1 kg of target product. Table 10 reports
potential TBC that can be produced from 1 ton of each source. These values are estimated
theoretically and scaled up based on the lignin content of original feed source.
Table 10- Potential catechol production from different resources
Reference source Potential TBC production (ton/ton biomass)
Coconut shell 0.22-0.28 Cotton stem 0.26 Rice husk 0.22 Softwood 0.16-0.23 Switch grass 0.16 Sugarcane bagasse 0.15-0.18 Hardwood 0.11-0.16 Corn stover 0.11-0.14 Wheat straw 0.1 Barley straw 0.07 Corn stalks 0.06 Candlenut shell 0.04
A final modeling scenario considered the same simulation for the most efficient resource
(coconut shells-36% lignin) and the alternate extraction method (solvent extraction) to
estimate total potential reduction in ten impact categories of interest. A detailed life cycle
88
inventory of this pathway is included as Table B6 in the Appendix B. Potential TBC
production is about 0.22 ton/ton of biomass, more than five times that for candlenut shells,
so less biomass will be required for production of 1 kg TBC. Accordingly, less energy and
material use is required along the entire life cycle. TBC production from coconut shell
presented fewer impacts compared to candlenut shell except for eutrophication impacts, as
the contribution of fertilizer input is higher for plantation coconuts than for candlenuts. In
this case, the field N2O emissions contribution is 4% of overall GHG emissions. For all
other impact categories, more than 40% reduction was observed from using coconut rather
than candlenut shells. Comparative impact assessment details are included as Figure B2
of Appendix B.
3.3.4. Uncertainty and Additional Considerations
As mentioned previously, there is uncertainty in the lignin structure and the exact products
of conversion processes. Chemical reactions that occur during lignin extraction and
purification are not known exactly. The proposed molecular formula in this study
(C20H26O6) is based on experimental analysis, while for simulation purposes we used
enthalpy and heat capacity reported by the NREL database(Wooley & Putsche, 1996) for
an empirical molecular formula of C7.3H13O1.3. Several studies have proposed model
compounds for lignin conversion(Binder, Gray, White, Zhang, & Holladay, 2009;
Hofrichter, 2002; Zakzeski et al., 2010) but with variation in results according to source,
implying that process simulation models should use feedstock-specific results wherever
possible. Even for the same process of solvent extraction, separation yield will likely be
different from one source of softwood (candlenut shells) to the other source (coconut
shells).
89
In addition to uncertainties associated with the feedstock itself, there are many other
parameters that can induce higher levels of uncertainty such as cultivation yields,
operational conditions, catalyst lifetimes, and conversion or separation yields for different
processes. There is not enough laboratory data for the depolymerization process to
determine probability distributions required for a robust statistical uncertainty analysis, but
here we give a qualitative discussion. Scale-up of laboratory methods to industrial
operations was modeled linearly for feedstock and chemical inputs, while energy
requirements were simulated in ASPEN Plus rather than extrapolated. Actual scale-up will
involve economic optimization, leading to more efficient use of solvents and extension of
catalyst lifetimes where technically feasible. Though methanol is specified in large
quantities for the laboratory experiments, its more efficient use at commercial scales will
not noticeably change the results, as methanol’s contribution to overall environmental
impacts is small for all impact categories. As mentioned previously, more efficient use of
dichloromethane (or its complete substitution) could significantly improve the bio-based
route, while more efficient use of xylene in the petrochemical route could reduce the
impacts of this option by up to 5-10% depending on impact category.
Scale is an important consideration at the cultivation stage as well. We have assumed here
that the primary feedstocks for lignin, candlenut shells and coconut shells, are unused
byproducts from established agricultural operations, so that direct and indirect land use
change were not considered. The development of high value-added chemical uses for the
lignin fraction of agricultural residues may incentivize production of high lignin-content
crops, leading to both direct and indirect land use change that should be considered.
90
Presently, however, the lignin resource described in Table 6 far outstrips global demand
for catechols.
As a potential production platform for bulk chemicals, lignin has many competing uses. As
discussed in Scown et al.(Scown, Gokhale, Willems, Horvath, & McKone, 2014),
production of bio-based chemicals from lignin should be studied in parallel to more
common uses, primarily direct combustion of lignin for on-site heat or combined heat and
power. Here we have compared bio-based and petrochemical TBC while considering co-
products only during the allocation and system expansion procedures. Further work could
instead consider a reference flow of lignin resource and compare multiple conversion
routes and end-products to find the most beneficial uses of lignin on a life cycle basis.
91
Chapter 4:
Life Cycle Assessment of UV-Curable Biobased Wood Flooring Coatings
For most people in industrialized countries, the great majority of time is spent indoors.
Indoor air quality is thus a vital concern for human health and productivity. Paints and
coatings have been formulated with the goal of eliminating or significantly reducing VOC
emissions, while the chemicals industry has been developing bio-based alternatives to
fossil-based building blocks in many applications. In this study, a bio-renewable content
formulation for wood flooring coating is analyzed using a life cycle assessment (LCA)
framework and compared to a petrochemical formulation of equivalent performance. This
formulation has 30% bio-based ingredients and zero-to-low VOC emissions, and was
developed by PPG Coatings and Resins R&D Center. This is a cradle to gate analysis and
is scoped to consider biomass cultivation and crude oil extraction and refining for
renewable and non-renewable chemical inputs, formulation, transport, and application of
1 m2 of each coating, followed by UV-curing. Comparative results showed more than 30%
reduction in six out of ten impact categories, using the USEPA TRACI 2.1 impact
assessment method, with smog formation, acidification, eutrophication and respiratory
effects showing increase in environmental impacts for the bio-renewable content
formulation. Epoxy resin (type-A) and corn-derived monomers are the most impactful
chemicals in the composition of conventional and bio-renewable wood flooring coatings,
respectively. The contribution of various building blocks to the environmental impacts of
both coatings are presented in detail, potentially guiding further formulation research and
development. It is shown that modifying BRC formulation using corn stover instead of
corn grain for synthesis of sugar-derived building blocks, will minimize trade-offs and
92
improve environmental profile of BRC formulation. The results highlight that meeting
targets for bio-based content can have multiple secondary benefits to the environment and
human health, which depend on the particular biofeedstock and conversion processes as
well as the petrochemical components that are being replaced.
4.1. Introduction Biomass from dedicated production or residues from forestry, agriculture and aquaculture,
could serve as environmentally sustainable feedstocks for fuels and chemicals, provided
that, production routes offer reductions in energy and material use and emissions on a life
cycle basis. Global production of bio-based chemicals (excluding biofuels) is estimated to
be 50 million metric tons (De Jong, Higson, et al., 2012), the largest category of which is
synthetic bio-based polymers (~55%).(NNFC, 2014) Renewable chemical building blocks
have been targeted to substitute for petrochemicals in various applications,(Holladay et al.,
2007; Montazeri, Zaimes, Khanna, & Eckelman, 2016; Werpy et al., 2004) including paints
and coatings, one of the major markets for chemicals and polymers. Active research and
development in this sector has facilitated application of bio-based chemicals in products,
such as the use of proteins as biopolymer binders,(Derksen, Cuperus, & Kolster, 1996)
vegetable oils as binder constituents in coatings formulations,(Derksen et al., 1996) non-
drying oils including soybean, sunflower and linseed oils as automotive finishes,(Athawale
& Nimbalkar, 2011) and production of powder coatings and alkyd resins using bio-
renewable ingredients.(Van Haveren et al., 2007) Various biomass fractions have been
utilized as feedstocks for renewable polymers,(Gross & Kalra, 2002; Shakina, Lekshmi, &
Raj, 2012) including polyesters, polyurethane, polyamides, epoxy resins and vinyl
93
copolymers.(Meier, Metzger, & Schubert, 2007) In this study, we investigate application
of renewable building blocks in composition of wood flooring coatings. Wood coatings,
with global market size of 100 million gallons (378 million liters) in 2005,(Kimberly Davis
& Swanson, 2005) are applied on the surface of the wood in order to enhance its natural
beauty, protect wood from abrasion and degradation, and provide a cleanable
surface.(Williams, 1999) Wood flooring is an important building product and many of the
green building rating systems, including LEED, GBTool, Green Globes, and CASBEE, are
supportive of coatings that minimize VOCs and other indoor air pollutants,(K.M. Fowler,
2006) while LEED assign a credit, specifically, for use of rapidly renewable materials in
coating formulations.(USGBC, 2006)
Before the development of modern petrochemicals, agricultural sources were used widely
for ingredients in wood coating applications.(Derksen et al., 1996) Plant proteins, linseed
oil and soybean oil were all used historically as building blocks in coating
formulations.(Derksen et al., 1996) With the widespread availability of synthetic polymers
(Bardi, 2009), polystyrene, polyurethane and polyvinyl chloride were introduced in
coatings with customizable physical properties (Deaner, Puppin, & Heikkila, 1996; Meier-
Westhues, 2007), while later on, acrylates combined with isocyanates and melamines
added high UV durability and hardness to the coatings.(Maldas & Kokta, 1991) In the late
1970s, the US Occupational Safety and Health Administration (OSHA) issued regulations
to help control indoor emissions and maintain safe indoor air quality (IAQ) levels,
primarily targeting volatile organic compounds (VOCs).(Safety & Administration, 2015)
Exposure to high concentrations of VOCs in indoor environments can trigger membrane
94
irritation, liver and kidney disease and cancer, depending on the contaminant and the level
of exposure.(Niu & Burnett, 2001) As a result of new standards, VOCs were targeted for
substitution in the development of low-solvent and solvent-less adhesives and
coatings.(Linak, 2009)
Such formulations may reduce VOC exposure for workers and building inhabitants,
reducing potential health effects, while the inclusion of bio-renewable ingredients reduces
the need for non-renewable petrochemical inputs. While these direct benefits are obvious,
there are many other types of hazards and potential environmental impacts to consider,
such as total energy use for production and application, or greenhouse gas (GHG)
emissions. The goal is developing sustainable coating formulations that provide equivalent
functionality as conventional formulations, while mitigating associated environmental
impacts overall. In order to ensure that new formulations do not have unintended
environmental or health impacts, either from emissions during production of novel
ingredients, or during product use and eventual disposal, it is necessary to apply a holistic
assessment tool that compares formulations on a life cycle basis. In addition to current
efforts in decreasing fossil fuel inputs and addressing human health issues, there are various
environmental programs that encourage enhancing ecosystem health through consumption
of renewable building blocks.
Life Cycle Assessment (LCA) is a tool to assess the potential environmental impacts and
resources used throughout a product's life cycle, considering all potentially hazardous
emissions and multiple categories of health and environmental impacts that result from
95
those emissions.(ISO, 2006) By identifying the processes or materials in a product life
cycle that contribute the most or the most hazardous emissions overall, LCA can be used
to investigate the most important contributors to environmental impacts. Thus, it can
deliver information for designers to guide material selection, assist in supply chain
management efforts, compare alternate designs or formulations, and provide product-level
assessments that can be used for technology development and marketing.
LCA has been used extensively in the chemicals and formulated products sectors, including
coatings.(Bidoki, Wittlinger, Alamdar, & Burger, 2006; Häkkinen, Ahola, Vanhatalo, &
Merra, 1999; Hofland, 2012; Papasavva, Kia, Claya, & Gunther, 2001) Hakkinen et
al.(Häkkinen et al., 1999) investigated environmental impacts of thirteen water-borne and
solvent-borne commercial coatings for outdoor applications in Finland, using LCA
framework. The cradle-to-grave analysis was framed in a 100-year period including
maintenance and renewal, in addition to final disposal of the coatings. The results showed
that water-born acrylic coatings had the lowest VOC emissions, as expected. Results for
other environmental impact categories were mixed, as several formulations of water-born
coatings were shown to have higher energy use and CO2, NOx and SOx emissions when
compared to the solvent-born counterparts. These results also highlighted that the
manufacturing of coating components inputs are a critical consideration in determining
environmental impacts of a coating over its lifetime, and not just emissions that occur
during product application.
96
The benefits of including renewable building blocks in coating formulations was examined
in a comparative LCA study by Gutaffson and Borjensson.(Gustafsson & Börjesson, 2007)
Four different formulations, two wax-based and two lacquers using ultraviolet light for
hardening (UV lacquers), were investigated. Wax-based coatings included one 100%
fossil-based coating sourced from crude oil and one renewable wax ester produced from
rapeseed oil, while UV lacquers consisted of one 100% UV coating-100% solid content-
and one water-based coating. The results of cradle-to-grave LCA showed that 100% UV
coating is the most environmentally benign alternative followed by water-based UV. For
global warming potential, the fossil wax had the highest contribution while acidification
and eutrophication potential were mostly dominated by renewable wax.(Gustafsson &
Börjesson, 2007) Consumption of pesticides and fertilizers during biomass cultivation
played a key role in ecotoxicity, acidification, and eutrophication impacts of renewable
wax, highlighting the importance of considering multiple impact categories, not just global
warming, when evaluating bio-based products. As recommended by the authors, the 100%-
UV coatings could be further improved by substituting epoxides and diacrylates with
renewable building blocks.
Supporting the results of Gutaffson and Borjensson, several other comparative LCA studies
between renewable building blocks and their fossil-based counterparts have shown that use
of renewable alternatives can cause trade-offs in overall environmental impacts, lowering
impacts in GHG emissions and non-renewable energy use, while shifting burdens to other
impact categories (due to increased agricultural activities and inefficient or energy-
intensive conversion methods).(Huijbregts et al., 2006; Montazeri et al., 2016; Tabone et
97
al., 2010) Therefore, minimizing these trade-offs through the choice of feedstock and
conversion method, is a key element in ongoing research. Previous research has shown that
use of agricultural and forest residues as feedstock can decrease the impacts associated
with agricultural activities,(Cherubini et al., 2009; Vink et al., 2003) while process
modifications such as less solvent use, recycling /substitution of hazardous input materials
and catalyzed reactions can result in more efficient conversions.(Fernando et al., 2006)
The present study is a cradle-to-gate LCA study of a new 100% UV-cured wood flooring
coating with 30% bio-renewable content (BRC) and zero-to-low VOC content.
Environmental profile of this formulation is compared with the conventional low-VOC
UV-cured wood flooring coating. The proposed formulation was developed by Pittsburgh
Paint and Glass (PPG) Coatings and Resins R&D center. Cradle to gate LCA results are
compared across multiple impact categories in order to highlight potential environmental
benefits or impacts of the new formulation and provide recommendations for further
improvements.
4.2. Methods As described in relevant ISO standards (14044:2006), goal and scope definition, life cycle
inventory, life cycle impact assessment and interpretation are the four main stages in
each LCA study.(ISO, 2006)
4.2.1. Goal and Scope
The goal of this study is to evaluate life cycle environmental impacts of a new UV-cured
coating formulation for wood flooring applications with equivalent performance to the
98
existing UV-cured control formulation. The proposed formula has 30% bio-renewable
content (BRC) made up of three main renewable monomers (components a, b and c) from
corn and soy. Bio-based monomers can be processed further to produce renewable
polymers. These compounds and their derivatives will replace acrylate groups, one of the
VOC sources of conventional UV-cured coatings. The abrasion-resistant sealer, sanding
sealer and topcoat are main layers of the coating where the sealer layers prevent abrasion
and seal interior surface of the wood (Mireles et al., 2011), and the topcoat is the finishing
solvent applied in order to inhibit surface degradation of the wood.(George, Suttie, Merlin,
& Deglise, 2005) Typically, all of the three layers have acrylates as main components, as
acrylate groups increase adhesion and show resistance to breakage and attack by chemical
solvents.(Moore, 1990)
In order to ensure equivalent functional unit, the proposed formula has been tested upon
standard protocols for hardwood flooring finishes. Two sets of tests were conducted,
including 1) flooring performance tests and 2) required tests for acceptance by wood
flooring industry. The first set included Cross Hatch Adhesion (ASTM D3359), Belmar
Loop (ASTM D2197), Gloss Retention (ASTM 2486), Taber Adhesion Resistance (ASTM
D4060) and Stain Resistance (ASTM D1308). The second set consisted of Hoffman
Scratch (ASTM D5178), Coefficient of Friction (ensures proper floor safety), Impact
Resistance (in-house method accepted by flooring customers), Steel Wool Scratch
Resistance (in-house method accepted by flooring customers), and Cold Check Resistance
(in-house method accepted by flooring customers that assesses coating flexibility and
ensures no coating failure under variations in temperature and humidity). The new
99
formulation was determined from these tests to have equivalent characteristics and
functionality during the coating use and maintenance phases. The functional unit of the
study was thus set to 1 m2 of coatings.
The LCA is scoped to account for impacts associated with raw material acquisition
(including crude oil extraction and refining for fossil-based building blocks, and biomass
cultivation, fractional extraction and conversion for renewable building blocks),
intermediate chemicals synthesis, layer assembly and UV-curing processes, a cradle-to-
gate assessment. The system boundary of this LCA is shown in Figure 14 for both BRC
and conventional coatings. This study focuses on formulation comparison and cradle-to-
gate life cycle assessment of the conventional and BRC wood flooring coatings, so
environmental impacts of use and end-of-life phases of the two coatings were excluded.
Two studies, from the VTT research center in Finland(Häkkinen et al., 1999) and
Gustaffson et al.,(Gustafsson & Börjesson, 2007) have shown that manufacturing energy
use and emissions and the durability of coatings are key factors in the life cycle
environmental impacts, while impacts stemming from end-of-life treatment and disposal
are relatively insignificant. In this study, durability is considered to be comparable between
the two coatings, based on standard performance testing, and thus the coatings are
compared for a single application.
100
Figure 14- System boundary for 1 m2 of control and BRC coatings 4.2.2. Life Cycle Inventory
Life cycle inventories are compiled based on material and energy consumption data. Composition and thickness of layers were given by PPG Coatings and Resin R&D Center.
Integration of the input parameters and life cycle inventories of the coatings are modeled
in the commercial LCA software package SimaPro v8.05 (Amersfoort, the Netherlands).
101
The inventories are developed using ecoinvent life cycle inventory database adjusted for
the US energy system (US-EI database, Earthshift, Huntington, VT). Inventory of both
coatings are partially based on existing unit processes.
For most cases, the exact chemical/compound is not available in the database, so either
approximate unit processes are used or new unit processes are created and added to the
database. For the purpose of this project, nearly 40 new unit processes are created in
ecoinvent including both intermediate and final compounds. MSDS (Material and Safety
Data Sheet) is the primary source for developing new unit processes. MSDSs typically
specify CAS number, chemical structure, production paths, and characteristics of the target
compound. Additional literature sources(Hess, Kurtz, & Stanton, 1995; Sienel et al., 2000)
are used besides MSDS. Target chemicals and their upstream processes are modeled up to
the point where the precursors are available in ecoinvent.
The proposed bio-renewable oligomers substitute for petroleum-based oligomers of
acrylate resins. Three corn-derived chemicals (two target chemicals (a and b) and one
intermediate) are modeled based on industrial data from literature, while soy-based
component (compound c) is modeled using the existing unit process from ecoinvent. As
the coating is partially bio-based, non-renewable compounds are handled using
approximate unit processes, substituting target compounds, or creating new ones. In
addition to the main formulations, control and BRC coatings, an alternative scenario was
modeled for the BRC coating, substituting corn-derived chemicals with the identical
counterparts obtained from corn-stover. While the inventories are created, the density and
102
film width values are used to convert mass-based inventories to fraction of each content in
unit area covered by the coatings. Final LCI data are all scaled based on 1 m2 of each
coating.
4.2.3. Life Cycle Impact Assessment
Ten environmental impact categories are considered in the life cycle comparison, including
(with equivalent units in parentheses) global warming (kg CO2 eq.), ozone depletion (kg
CFC-11 eq.), smog formation (kg O3 eq.), acidification (kg SO2 eq.), eutrophication (kg N
eq), carcinogenics (CTUh), non-carcinogenics (CTUh), respiratory effect (kg PM2.5 eq.),
ecotoxicity (CTUe) and fossil fuel depletion (MJ surplus), following the US EPA’s Tool
for the Reduction of Chemical and Other Environmental Impact (TRACI 2.1) life cycle
impact assessment method.(J. Bare, 2011) Impact assessment methods use coupled fate-
exposure-effect models to connect each life cycle emission to environmental or health
midpoints (physical changes) or endpoints (damages), considering a range of ecosystem
and public health issues.(Jolliet et al., 2003) Following GHG accounting conventions for
durable products,(WBCSD, 2011) we assume that the entire carbon content of purely bio-
based chemicals is supplied by atmospheric CO2. The amount of sequestered carbon is
calculated from the chemical formula of the bio-based compounds a and b, while carbon
content of refined soy-oil is used as an approximation for soy-based compound c.(Omni
Tech International, 2011)
In order to count for the share of impacts attributed to the co-products, allocation is
considered when necessary. Creation of new unit processes takes the same approach as
existing ecoinvent processes, assigning environmental impacts to several products using
103
mass allocation. For the simulation of corn-derived building blocks, impacts of upstream
processes of corn cultivation and wet milling are allocated between corn grain and corn
stover using economic allocation.(Luo, Van der Voet, Huppes, & De Haes, 2009)
Economic allocation gives higher share of impacts to the corn grain, compared to the
mass/energy allocation, and it creates an upper bound for environmental impacts of corn-
derived compounds. Besides, processing of agricultural residues is nascent technology, so
this allocation method leads to more realistic results. For the primary source of this study,
corn grain, share of impacts in agricultural and milling processes is about 88% while for
the alternate source, corn stover, this fraction is about 12%.
4.3. Results and Discussion The comparative cradle to gate life cycle results showed lower impacts in six impact
categories, when renewable feedstock (corn and soybean) is used in coating formulation.
Trade-offs of BRC formulation were significant in four impact categories of smog
formation, acidification, eutrophication and respiratory effects, with acidification showing
more than 27 times more impacts compared to the control coating. Table 11 presents
relative impacts of the BRC coating compared to the reference control coating. Human
toxicity (non-carcinogenic) and fossil fuel depletion show significant reductions (>50%)
over the life cycle of BRC coating. Impact reductions for ozone depletion, global warming,
human toxicity (carcinogenics) and ecotoxicity are less pronounced. These four categories
are estimated to take less credit from introduction of renewable building blocks, due to the
contribution of energy and chemical inputs in upstream processing of corn-derived
chemicals and residual acrylate groups. Agricultural activities, including diesel burned in
trucks and also surface runoff of N and P compounds to local water bodies due to fertilizer
104
use, are important trade-offs to recognize in assessment of BRC formulation. (Hill, Nelson,
Tilman, Polasky, & Tiffany, 2006; Hottle, Bilec, & Landis, 2013) Table 11- Relative LCA of BRC wood flooring coating compared to control UV-cured coatings (per m2 of coating)
Impact category Unit % Change (BRC relative to control)
Ozone depletion kg CFC-11 eq.
-31%
Global warming kg CO2 eq. -42%
Smog kg O3 eq. 617%
Acidification kg SO2 eq. 2771%
Eutrophication kg N eq. 35%
Carcinogenics CTUh -29%
Non-carcinogenics CTUh -74%
Respiratory effects kg PM2.5 eq. 1241%
Ecotoxicity CTUe -38%
Fossil fuel depletion MJ surplus -51% Figure 15 shows the comparative results broken down by relative contribution to overall
impacts of the three coating layers and the curing process. (Absolute results and the
chemical composition of various layers are discussed in the next section.) Green and gray
bars represent BRC and control coatings, respectively. Figure 15(a) shows the breakdown
of results for the BRC coating and demonstrates that the abrasion-resistant sealer is the
primary driver of negative impacts, contributing 58-83% of the total across impact
categories, followed by the sanding sealer. This behavior can be partially explained by
considering the mass fraction and the composition of each layer. The abrasion-resistant
sealer has the highest mass fraction in the coating, while the sanding sealer and topcoat are
105
ranked second and third. Corn-derived substitute for acrylate groups (compound a), with
the second highest mass fraction in the composition of both BRC abrasion resistant sealer
and BRC sanding sealer, contributes the most in overall environmental impacts. The
topcoat contributes less than 20% in all investigated categories. Electricity for UV-curing
adds the least impact to the overall burden, less than 1%.
Figure 15(b) shows the contribution of layers for the control coating. As for the BRC
coating, the abrasion-resistant sealer again shows the highest contribution, 50-81% of
overall impacts, mainly caused by the extensive use of epoxy acrylates resins in this layer.
Mass fraction of acrylates in control abrasion resistant sealer is about 50%. The only
exception is ozone depletion potential, where the sanding sealer is controlling the impacts.
Use of liquid chlorine in synthesis of diol precursors plays the key role in contribution of
this building block in ozone depletion of control sanding sealer. The impacts of control
sanding sealer is mostly controlled by production of epoxy resin, major component of
coating formulation with thermoplastic behaviors.(Aouf et al., 2013) Again mirroring the
BRC coating results, estimated impacts are mainly caused by the synthesis of intermediate
chemicals and the production of each coating, while electricity use in the UV-curing
process is shown to have <1% contribution to overall impacts.
106
Figure 15- Contribution of layers and UV-curing process in environmental impacts of (a) BRC and (b) control coatings
Various layers of BRC and control coatings are compared in absolute terms in Figure 16.
The BRC layers (green bars) have lower impacts compared to fossil-based counterparts
(gray bars), except for the impact categories of smog formation, acidification,
eutrophication and respiratory effects. As presented in Figure 16, environmental impacts
of BRC and control layers show the same behavior relative to their counterparts, however,
there are some exceptions in ozone depletion potential and eutrophication categories.
107
Ozone depletion potential is one of the impact categories estimated to show environmental
benefits when renewable sources are used. Closer look at the layers’ comparison in Figure
16, shows that higher impacts of BRC abrasion resistant sealer is compensated by
environmental benefits of BRC sanding sealer and BRC topcoat. Primary contributor of
ozone depletion in abrasion resistant sealers is an acrylate derivative, used for radiation
cure purposes, which is common for both BRC and control layers. For the BRC
formulation, impacts of this compound is closely followed by corn-derived substitute
(compound b), adding more impacts to the BRC layer. Eutrophication, on the other hand,
is an impact category showing environmental trade-offs due to introduction of renewable
feedstock. It is expected that use of renewable feedstock induce more eutrophication
impacts in BRC layers, but control sanding sealer is not following this trend. Coal burned
power plants that supply energy for upstream processing of epoxy resin and acrylates
derivatives, main components of control sanding sealer, drive eutrophication impacts.
BRC abrasion resistant sealer, sanding sealer and top coat shows the maximum
environmental benefits in impact categories of non-carcinogenics and fossil fuel depletion,
with impact reduction of more than 50% for all three layers. Substitution of epoxy resin in
abrasion resistant sealer and sanding sealer and acrylate derivatives in topcoat with
environmentally benign counterparts led to the significant impact reductions.
108
Figure 16- Life cycle comparison between layers of BRC and control coatings
The results of Figure 15 and Figure 16 highlight key points in understanding the cradle-
to-gate environmental impacts of both coatings. Acrylate derivatives, corn-derived
compound b and soy-based compound c are shown to be major contributors to the
environmental impacts of BRC formulation. Some of the acrylates are mutual compounds
between BRC and control coating formulation. Chlorinated solvents used in upstream
processing of the acrylates, trigger environmental impacts in all categories. Environmental
impacts of corn-derived chemicals, as mentioned earlier, is caused by the diesel burned in
109
agricultural equipment and fertilizer run off during cultivation of corn, in the first place.
Soy-based compound c shows the same behavior in upstream processing. Fertilizer and
pesticide consumption in soybean agriculture, and aromatic, aliphatic and chlorinated
compounds added during soybean crushing and degumming and soy oil refining induces
high levels of environmental impacts.(Omni Tech International, 2011) Environmental
burden of upstream activities for corn and soybean derived component, emphasizes that
the environmental performance of renewable building blocks are highly dependent on the
choice of bio-feedstock and extraction/conversion processes.
Specific types of epoxy resins (here is called type A) and acrylate groups drive negative
impacts of the conventional coating formulation. Both of these compounds have been
studied frequently for their levels of toxicity. Precursors of type-A epoxy resins are known
to be an endocrine-disrupting chemical, causing developmental, metabolic, and
reproductive systems malfunctioning.(Flint, Markle, Thompson, & Wallace, 2012)
Acrylate groups, on the other hand, are classified as mutagenic and/or carcinogenic
compounds(Lithner, Larsson, & Dave, 2011) and even trace amount of these chemicals
show significant contribution in overall impacts. Type-A epoxy resin, with 20% mass
fraction in control sanding sealer, contributes up to 47% contribution in ecotoxicity. Its
content and contribution in other layers are less than 2% and 8%, respectively. Acrylate
groups are more common in both control and BRC layers. Between 40-75% of control
layers are composed of acrylate derivatives which shows significant impacts in different
categories, from 25% contribution in ozone depletion potential of control sanding sealer
up to 96% contribution in smog formation of abrasion resistant sealer.
110
Further improvement in environmental performance of BRC formulation can be achieved
by using agricultural and forest residues as biomass sources, since these sources are mostly
piled as waste or burned on site to produce energy. There is an active research area in
finding the most efficient resources using comparative LCA between various feedstock
choices. In order to evaluate this proposed scenario, an alternate BRC formulation is
modeled, substituting corn-derived chemicals with their identical counterparts from corn-
stover. The results show that if corn-derived chemicals were produced from corn-stover,
environmental impacts of BRC formulation would decrease significantly. The relative
reduction between the proposed BRC formulation and control coating, would be between
20-60% in nine out of ten categories, Table C2 in Appendix C. Eutrophication is the only
category that shows increase in impacts and even in that case, the relative value is 1%
increase. Carcinogenics and non-carcingenics are the only impact categories that show
more reduction when corn is used as main feedstock. Carcinogenics impact is triggered by
additional energy for processing of corn stover, supplied by hard coal burning plants. The
main reason for higher non-carcinogenics is the upstream mercury use in production of a
pretreatment solvent, used for separation of soluble and insoluble solids of corn stover.
This complementary analysis highlights that further modification in feedstock choice can
minimize expected environmental trade-offs and should be considered as next steps for
development of the formulation. Our study is focused on production of BRC and control
coatings, considering their use phase would have the same environmental profile.
However, direct exposure via inhalation of particulate and gaseous emissions from sanding
has been a major motivator of reformulation efforts. Future work would benefit from
111
empirical data through monitoring and characterization of emissions from sanding and re-
application of coatings.
In summary, bio-renewable formulations have been prioritized in research and
development phase by PPG and many other chemical companies. Comparative LCA results
for a 30%-BRC wood flooring coating show trade-offs in four categories of smog
formation, eutrophication, acidification and respiratory effects on a life cycle basis,
compared to a conventional control coating with equivalent performance and durability.
Parallel analysis on substituting hazardous components with renewable counterparts and
maximizing environmental benefits by modifying bio-feedstock choice and processing
conditions, should be considered as key steps in modifying new formulations.
Consideration of renewable building blocks in the design of buildings is an innovative
approach for sustainability that merits further research and development by industrial
sectors and policy makers.
112
Chapter 5:
Evaluating Microalgal Integrated Biorefinery Schemes: Empirical Controlled Growth Studies and Life Cycle Assessment
This study has been published
Soh, L., Montazeri, M., Haznedaroglu, B. Z., Kelly, C., Peccia, J., Eckelman, M. J., & Zimmerman, J. B. (2014). Evaluating microalgal integrated biorefinery schemes: empirical controlled growth studies and life cycle assessment. Bioresource technology, 151, 19-27.
Two freshwater and two marine microalgae species were grown under nitrogen replete and
deplete conditions evaluating the impact on total biomass yield and biomolecular fractions
(i.e, starch, protein, and lipid). A life cycle assessment was performed to evaluate varying
species/growth conditions considering each biomass fraction and final product substitution
based on energy consumption, greenhouse gas emissions, and eutrophication potential.
Lipid for biodiesel was assumed as the primary product. Protein and carbohydrate fractions
were processed as co-products. Composition of the non-lipid fraction presented significant
trade-offs among biogas production, animal feed substitution, nutrient recycling, and
carbon sequestration. Maximizing total lipid productivity rather than lipid content yielded
the least GHG emissions. A marine, N-deplete case with relatively low lipid productivity
but effective nutrient recycling had the lowest eutrophication impacts. Tailoring algal
species/growth conditions to optimize the mix of biomolecular fractions matched to desired
products and co-products can enable a sustainable integrated microalgal biorefinery.
5.1. Introduction As renewable energy sources increase in their prevalence and use, the research and
adoption of efficient processes and technologies is vital for the field to sustainably expand.
113
Microalgae have been indicated as a robust potential alternative to traditional fuel resources
due to their ability to be used as a feedstock for a variety of biofuels and other value-added
chemicals.(Pienkos & Darzins, 2009) Microalgal biomass production offers a number of
advantages over conventional biomass production, including higher productivity, use of
otherwise nonproductive land, reuse and recovery of waste nutrients, use of saline or
brackish waters, and reuse of CO2 from power plant flue gas or similar sources.(Pienkos &
Darzins, 2009) While algal biofuels are promising, particularly for the production of
biodiesel, current practices and technologies are not sufficient to make large-scale
production energetically or economically favorable with liquid fuel as the sole salable
product. Thus, improvements and innovations to the biofuel production process must be
achieved including the development of biorefinery approaches to recover energy and
nutrients as well as accommodate the non-lipid fractions (i.e., carbohydrate, protein) of
algal biomass.
Currently, there is a significant focus on growing microalgae specifically for biofuel
applications. Possible fuel products include biocrude, biogas, biohydrogen, bioethanol,
and biodiesel,(Brennan & Owende, 2010) each of which have advantages and
disadvantages due to feedstock processing and limitations. The feedstock requirements for
these processes can vary significantly, and optimization of the microalgal biomass will
differ based on the process and target fuel selected. For example, bioethanol production is
optimal with a high starch (carbohydrate) feedstock, while biodiesel is produced from
triglycerides found in the lipids.(Mata, Martins, & Caetano, 2010) Biocrude oil and biogas
can be produced through thermochemical conversion processes where the maximization of
114
total biomass is ideal (Brown, Duan, & Savage, 2010), yet these processes do not allow for
the harvesting of other valuable co-products such as protein for animal feed or
nutraceuticals and reduce the potential for nutrient recycling.(Spolaore, Joannis-Cassan,
Duran, & Isambert, 2006) In fact, business models have shown that algae cultivated for
biofuel alone will yield comparatively low profits or returns since 1) biofuel is relatively
low value commodity and 2) only a fraction of algal biomass can be utilized for biofuel
leaving a significant percentage of “waste” if not managed for further value recovery.
(Subhadra & Edwards, 2011)
Thus, it is economically and environmentally critical to expand the downstream processing
of biomass to other finished products besides fuels in a biorefinery setting.(Stephens et al.,
2010) This multiproduct paradigm aligns with the model used by crude oil refineries where
multiple value-added fuels and chemicals are produced. This type of approach has been
explored for a biorefinery in a life cycle assessment (LCA) of switchgrass by (Cherubini
& Jungmeier, 2010)where it was found that significant GHG and fossil energy savings
could be achieved when compared to a fossil reference system, although there are
potentially larger eutrophication and acidification impacts. The study did not compare the
impacts of a fuel only versus a biorefinery model, which will be important in demonstrating
the benefit of a biorefinery configuration versus a singular focus on an individual end
product. In another study, a coproduct market analysis and water footprint, not considering
energy or GHGs, was conducted for an algal biorefinery (Subhadra & Edwards, 2011)
demonstrating clear advantages for a multiproduct paradigm to attain high operational
profits.
115
In order to successfully implement the biorefinery model, technological innovation as well
as gains in efficiency must be made. These efficiencies may be realized through cultivation
of the appropriate strain and optimization of the growth conditions for the intended end
products. There are wide ranges observed for lipid, protein, and carbohydrate composition
depending on algal species as well as the growth conditions.(M. J. Griffiths & Harrison,
2009) For example, growing microalgae in N-deplete conditions promotes cellular lipid
accumulation in many species. (M. J. Griffiths & Harrison, 2009) Attempts to exploit this
high lipid content for the production of biodiesel while simultaneously reducing nutrient
costs, however, is challenged by a low total biomass growth in microalgal cultures.(Rodolfi
et al., 2009) This trade-off presents a challenge towards optimization of strain and nutrient
loadings for the appropriate mix of desired outputs (i.e.total biomass, high lipid, protein,
or starch content) while minimizing resource inputs and environmental impacts.
Previous LCA studies have evaluated the embedded energy, water use, and environmental
impacts associated with many aspects of microalgal biofuel production process including
co-product production.(Brennan & Owende, 2010; Brentner, Eckelman, & Zimmerman,
2011; Campbell, Beer, & Batten, 2011; Clarens, Resurreccion, White, & Colosi, 2010;
Jorquera, Kiperstok, Sales, Embiruçu, & Ghirardi, 2010; Lardon, Hélias, Sialve, Steyer, &
Bernard, 2009; Shirvani, Yan, Inderwildi, Edwards, & King, 2011; Sills et al., 2012;
Subhadra & Edwards, 2011) These studies generally concluded that, although microalgae
are a promising fuel feedstock, system improvements are necessary for them to become
economically viable and sustainable. One crucial finding has been that effective utilization
of non-fuel co-products is essential for the overall system to achieve a positive energy
116
return on investment (EROI).(Sills et al., 2012) Differences in EROI ratios from
previously published studies are a result of inconsistencies in functional units, system
scope, boundaries, key parameters, and other assumptions.(Jorquera et al., 2010; Liu,
Clarens, & Colosi, 2012; Sills et al., 2012) Further, many of these studies assume non-
specific or freshwater algal species. Of the few studies that considered marine
species,(Campbell et al., 2011) described a coastal algae production system based on
pumped seawater and assumes similar lipid production and profiles to freshwater species,
which is likely unrealistic based on (M. J. Griffiths & Harrison, 2009) and the findings
reported below. (Jorquera et al., 2010) assessed different reactors for marine algal growth
and obtained positive net energy ratios (NERs) for oil production in both closed reactors
and open ponds; however, downstream processing of the lipid and oilcake was beyond the
scope of their study. Finally, (Yang et al., 2011) considered life cycle water and nutrient
reductions associated with seawater rather than freshwater for cultivation but used reported
growth results only for C. vulgaris (a freshwater strain) at N-replete conditions. The work
presented here encompasses controlled growth studies for multiple freshwater and marine
species where the biomass composition is well characterized for not only lipid production
but also starch and protein contents across different growth conditions as not previously
seen in the literature.
Previous LCAs have also considered the trade-off between high and low nitrogen growth
conditions.(Campbell et al., 2011; Lardon et al., 2009) In an LCA of C. vulgaris, significant
differences were reported in cumulative energy demand for N-replete and N-deplete
conditions. (Lardon et al., 2009) It was found that N-deplete conditions yield more lipids
117
for less cumulative energy demand, in part as a result of reduced fertilizer input. However,
this reduction in life cycle energy requirements was offset by a decrease of 55-65% in
embodied energy remaining in the oilcake after lipid extraction. In a setup where the
oilcake is combusted or anaerobically digested for biogas, this decreased energy content
reduces on-site heat and power meaning that external fuel sources must instead be used.
This result is important because fertilizer inputs have been shown to significantly impact
the overall energy and GHG balance of algal fuels (Clarens et al., 2010) while much of the
research on nutrient-limited conditions has focused on the biomolecular composition and
productivity of the lipid fraction only. (Campbell et al., 2011; Lardon et al., 2009)
In general, previous LCA studies have focused on a single production scheme (typically
lipid for biodiesel or starch for bioethanol), rather than considering trade-offs among each
fraction of algal biomass for production of multiple salable co-products. Many researchers
have used common assumptions about various algae strains, including lipid content,
volumetric productivity, and nutrient inputs, based on stoichiometric requirements and
ideal conditions rather than empirical data.(Clarens et al., 2010; Liu et al., 2012; Shirvani
et al., 2011; Subhadra & Edwards, 2011)
In this work, we use experimental results from controlled growth studies (nitrogen replete
and deplete) to provide LCA data for four different microalgae strains – two freshwater
and two marine species – while considering material, energy, and media inputs. Variation
in nutrient inputs specifically occur between replete and deplete conditions and between
source water types. The resulting biomolecular composition (lipid, starch, protein) and
118
biomass productivity values are used to quantify the life cycle environmental impacts from
the algal biorefinery for three key critical environmental midpoints: cumulative energy
demand, GHG emissions, and eutrophication potential. This LCA considers lipid-based
biodiesel as the primary product and carbohydrate-based bioelectricity and protein-
substituted animal feed as co-products within the biorefinery. In this way, targeted
cultivation and species selection can be evaluated to inform potential microalgal integrated
biorefinery schemes.
5.2. Materials and Methods 5.2.1. Chemicals and materials:
All chemicals for media growth were supplied by either Sigma-Aldrich or J.T. Baker and
were of reagent grade quality. Seawater was harvested from Long Island Sound, filtered,
and pasteurized as described in (UTEX, 2011). Solvents chloroform, acetone, ethanol, and
methanol were supplied by J.T. Baker. CHROMASOLV® heptane and LC-MS
CHROMASOLV® 2-propanol were supplied by Sigma-Aldrich and Fluka, respectively,
for chromatographic analysis.
5.2.2. Algal Growth Experiments:
Algae were cultivated in triplicate 1 L Erlenmeyer flasks filled with 500 ml of culture
media and supplied with 0.75 L/min air enriched with 2% carbon dioxide bubbled into the
reactors. Algae strains were purchased from the Culture Collection of Algae at the
University of Texas at Austin (UTEX, 2011) and grown in the specified media. Freshwater
strains - Neochloris oleoabundans (Chantanachat and Bold 1962, UTEX #1185) and
Chlorella sorokiniana (Shihira and Krauss 1965, UTEX #260) - were grown in Bold 3N
119
Medium (Harold C. Bold, 1970) and marine strains - Nannochloropsis oculata ((Droop)
Hibberd 1981, UTEX#LB 2164) and ‘Tetraselmis suecica ((Kylin) Butcher, UTEX #LB
2286) - in Enriched Seawater Medium (Harold Charles Bold & Wyynne, 1978) without the
specified nitrogen content which was modified according to the experimental protocol for
N-replete and N-deplete conditions. The nitrogen concentration of the background media
was taken into account, and potassium nitrate was used to bring the total nitrogen
concentration up to 10mg/L (N-deprived) and 100 mg/L (N-replete) - as N. The microalgae
were supplied with 14 h light and were constantly mixed with a magnetic stir bar. In order
to account for the volume of reactor contents harvested for analyses, a reactor for each N-
and N+ condition was simultaneously cultivated and used to replenish the harvested
volume to maintain a constant reactor volume.
5.2.3. Algal Sampling and Harvesting:
Optical density measurements at 610 nm of the cell cultures were taken daily and
correlations with cell dry mass and cell number concentration estimated by via calibration
curves. For these calibration curves, cell mass per unit volume was measured for
lyophilized cells. Cell number per volume was measured by counting under microscope
with a hemocytometer. For all analyses, cells were harvested in late exponential growth
phase as determined by previously established growth curves, which correlates to 8 or 9
days of growth depending on the species. During harvesting a fixed volume of culture was
transferred into falcon tubes, centrifuged for 5 mins at 12,000 rpm and 4˚C, decanted, and
transferred to microcentrifuge tubes in which samples were frozen at -20˚C until extraction
and further analyses.
120
5.2.4. Extraction and Analyses:
All analyses were run in duplicate on each of the three replicates. Lipid: Lipid extraction,
transesterification, and fatty acid methyl ester analysis was performed as detailed by the
conventional solvent extraction method in (Soh & Zimmerman, 2011). Glyceryl
nonadecanoate was used as an internal standard by addition to a subset of each strains’
samples and was used to calculate the extraction efficiency for each strain per cell mass.
Protein: Thorough method development was performed to insure maximal protein
extraction and replicable analyses varying extraction solutions, homogenization timings,
number of extractions and background analyses. A solution of 0.1 M NaOH and 0.25 mL/L
Tween 20 as well as 0.07 g of 0.5 mm and 0.2 g of 0.1 mm ceramic beads were added to
the pelleted cells and homogenized on a bead beater for 1 min similar to (Meijer & Wijffels,
1998). The samples were then centrifuged for 1 min at 7000 rpm and the supernatant
collected. For complete extraction, this process was repeated two more times, and the
combined supernatants were then analyzed for protein using the Pierce® BCA Protein
Assay Kit (Thermo Scientific), following the manufacturer’s instructions. Analysis was
performed in 96-well microplates using the provided bovine serum albumin standard to
make a calibration curve on each plate. In order to mitigate any interference with pigments,
each sample was also run with a sample blank and the background absorbance subtracted
before calculation of protein concentration. Starch: Similar method development
procedures were followed as for protein extraction. In the end pelleted cells were pre-
extracted with acetone and ethanol in order to remove interfering substances as in
(Fernandes et al., 2012). Acetone was added to the algae, and the samples were
homogenized for 1 min on a bead beater. The samples were then centrifuged at 14,000
121
rpm for 1 min and the acetone was discarded. This step was repeated with ethanol until no
further visible pigments were extracted from the cells. After pre-extraction, the cells were
then completely transferred to glass test tubes for starch extraction. The tubes were
centrifuged at 4000 rpm for 1 min and the supernatant discarded. Analysis was performed
using a starch assay kit (Sigma SA 20) following manufacturer’s instructions for starch
extraction with DMSO and HCl with all analyses scaled down to the appropriate volume.
Nitrate: Nitrate concentrations were measured from the filtered supernatant of centrifuged
cells using a nitrate test kit (Nitrate Elimination Company, Inc.). Analysis was performed
as per the manufacturer’s instructions for both freshwater and seawater species in 96-well
microplates.
5.2.5 Life Cycle Assessment:
A life cycle assessment was performed comparing the various production schemes:
freshwater and seawater species under both nitrogen replete and deprived conditions, with
downstream processing of each biomass fraction into a target product: lipid to biodiesel,
starch to bioelectricity, and protein to animal feed (Figure 17). Material and energy
requirements for each scheme were estimated using the Algae Process Description (APD)
module of the GREET 2012 rev2 model (Frank et al., 2011) with minor modifications as
follows. GREET assumes internal recycling of water and nutrients from dewatering and
anaerobic digestion (AD) back to cultivation. Here, GREET-specified water quantities and
pumping requirements were preserved, while nutrient inputs were altered according to the
media recipes used in the experimental set-up. Nutrient recipes for each media type (in
g/L) were modeled using primarily unit processes from the ecoinvent 2.2 LCI database and
adjusted for background levels of N in freshwater and seawater. Where data did not exist
122
for specific nutrients, new unit processes were created using current primary industrial
production routes and stoichiometric equivalents. Unit process proxies from the ecoinvent
database were used in the remaining few cases. (Further details on modeling nutrient inputs
can be found in Appendix D)
Figure 17- Mass flows through life cycle stages included in the scope of the study as described by A) and detailed for each growth scenario (species/N-loading) in B) where for N. oleoabundans (Neo), C. sorokiniana (Chl), N. oculata (Nan), and T. suecica (Tet) under N-deprived (-) and N-replete (+) growth conditions. The APD module assumes as a reference flow 1 kg of bio-oil, with protein and
carbohydrate fractions as co-products. Subsequent transesterification to biodiesel was
modeled, including chemical and energy inputs, was modeled using the main GREET
Cultivation
Harvesting
Lipid Extraction
Residue Management
water
Digestion
Protein Extraction
Lipid Fraction
Protein Fraction
Methane
Fertilizer (50% of C and P; 24% of N in x5 flow)
a1
a2
a3
a4
a5
x1
x2
x3
x4
x5
Flow from Algae Process
Mass BalanceEquation
Cultivation x1
Harvesting x2=0.9*x1
Lipid Extraction x3=x2-a1
Protein Extraction x4=x3-a2
Digestion x5=x4-a3
Residue Mgmt x5=a4+a5
a1=1.04 kg bio-oil
solvent FlowNeo+ Neo- Chl+ Chl- Tet+ Tet- Nan+ Nan-
x1 13.3 3.5 6.5 4.6 72.4 9.1 6.9 5.6
x2 12.6 3.3 6.2 4.4 68.8 8.6 6.5 5.3
x3 15.0 1.8 3.7 2.4 73.5 9.2 4.3 4.1
x4 12.3 1.4 3.2 2.1 38.8 5.5 2.8 3.7
x5 6.2 1.3 2.8 1.9 35.7 4.1 3.0 2.3
a1 1.04 1.04 1.04 1.04 1.04 1.04 1.04 1.04
a2 2.7 0.4 0.6 0.3 34.8 3.6 1.4 0.3
a3 2.4 0.2 0.5 0.3 7.7 0.9 0.6 0.8
a4 0.3 0.1 0.1 0.1 1.3 0.2 0.1 0.1
a5 5.9 1.2 2.7 1.8 34.4 3.9 2.8 2.2
A)
B)
123
model. Following the algal biorefinery model suggested in (Brune, Lundquist, &
Benemann, 2009), lipid extraction is followed by protein extraction for animal feed, with
digestion of the remaining starch and residues for electricity, heat, nutrient, and sludge co-
products. Internal nutrient flows for N and P followed GREET assumptions, including 5%
N volatilization for open ponds, 95% re-utilization of N and P in AD supernatant,
displacement of N and P fertilizers, and C sequestration in soils by AD solids. Nutrient
flows were adjusted for each species and growth regime as were model parameters for the
protein-starch-lipid fractions determined experimentally. These fractions also affect the
production of methane in the AD by changing the relative inputs of C and N to the unit,
which was modeled in GREET following the biogas model of (Sialve, Bernet, & Bernard,
2009) with recommended adjustments for the digestible fraction. (Frank, Han, Palou-
Rivera, Elgowainy, & Wang, 2011)
The baseline harmonized Algae Process Description model assumes an open pond reactor
system due to the high degree of variability in reported energy requirements for mass
transfer in photobioreactors; however, an air-lift tubular reactor is also specified in the
model with zero mixing energy. In order to preserve comparability with reported results,
open ponds were modeled here, though the empirical growth studies relied on bench-scale
closed reactors with openings for air exchange. GREET model outputs for energy and
chemical use were matched with ecoinvent LCI data, as detailed in appendix A. Life cycle
impact assessment was carried out for three specific environmental impact categories:
cumulative energy demand (CED 1.08), greenhouse gas emissions (IPCC 2007 GWP100),
and eutrophication (using the TRACI 2 LCIA method). These endpoints were chosen to
124
consider trade-offs associated with nutrient use, lipid productivity, and co-product
generation using freshwater and marine species.
5.3. Results and Discussion 5.3.1. Algal Growth and Composition:
The four algal strains were chosen to represent both freshwater (N. oleoabundans and C.
sorokiniana) and marine (N. oculata and T. suecica) species, which have all been
previously studied for their use as biofuel feedstock.(Mata et al., 2010) As indicated in
Table 12, the culture density and mass yields of biomass for the N-replete conditions were
much higher than the N-deprived set as expected. (M. J. Griffiths, van Hille, & Harrison,
2012) For all N-deprived conditions nitrate concentrations were near or below the method
detection limit confirming that the availability of nitrate is a limiting factor for cell growth
in this system. For the N-replete set (starting at 100 mg/L as N), the freshwater species’
nitrate supply is significantly depleted though complete nitrogen starvation is not yet
reached. The marine species still show a significant portion of nitrate left in the media
despite nearing the end of their exponential growth implicating the limitation of another
key nutrient for growth.
125
Table 12- Conditions of algal cultures at harvest on day 8/9 during exponential growth phase for four species (two freshwater and two marine) in nitrogen deplete and replete conditions. Uncertainty values represent standard error between triplicates.
Algae Strain Nitrogen Cell density (cells/mL x 106)
Mass density (g/L)
Nitrate concentration (mg/L as N)
Fre
shw
ater
Neochloris oleoabundans (Neo)
deplete (-)
39.9 ± 7.5 0.69 ± 0.13 b.d.l.a
replete (+)
104.7 ± 11.9 1.83 ± 0.21 11.3 ± 6.3
Chlorella sorokiniana (Chl)
- 9.26 ± 0.38 0.28 ± 0.01 0.16 ± 0.01
+ 69.4 ± 13.7 2.18 ± 0.43 14.9 ± 8.1
Mar
ine
Tetraselmis suecica (Tet)
- 1.00 ± 0.21 0.30 ± 0.08 b.d.l.*
+ 5.59 ± 0.81 1.54 ± 0.25 61.3 ± 2.3
Nannochloropis oculata (Nan)
- 60.5 ± 5.7 0.56 ± 0.05 b.d.l.*
+ 188.6 ± 9.8 1.79 ± 0.09 67.69.9
5.3.2. Fatty Acid Methyl Ester Content and Composition:
The fatty acid methyl esters (FAME) that could be produced from extracted lipids form
each algal species was quantified. The derived FAME content per cell mass and
productivity was determined for each algal species and fell within typical ranges (M. J.
Griffiths et al., 2012). In all cases nitrogen limitation led to higher FAME content per dry
algae mass (between 8 – 75% higher) but the FAME productivity per volume was often
much lower due to the significantly inhibited biomass growth (Figure 18). This feature is
most evident with C. sorokiniana in N-deplete growth conditions where the high lipid
content per cell (35%, mg FAME produced/ mg cell mass) does not compensate for the
low total biomass growth in terms of total lipid production per volume (90 mg FAME/ L
cell culture). It is interesting to note that C. sorokiniana grown in N-replete conditions has
126
the highest FAME productivity per unit volume (580 m/L) due to the high biomass yield
even though the FAME content is moderate. N-deprived N. oleoabundans had the highest
FAME content of 42% and in this instance yielded higher lipid productivity per volume
(350 mg/L) than the N-replete condition (220 mg/L). A similar trend is observed for one
of the marine species, T. suecica (80 mg/L for N- vs. 50 mg/L in N+). Though not yielding
the highest lipid productivity on a per volume basis, the observed enhanced FAME
productivity per volume given the lower nutrient requirements will present an interesting
tradeoff in terms of resource use and environmental impact to be quantified by the LCA.
Figure 18- FAME content and productivity of algal species, N. oleoabundans (Neo), C. sorokiniana (Chl), N. oculata (Nan), and T. suecica (Tet), with nitrate replete (solid symbols) and nitrate deprived (outlined symbols) growth conditions. Error bars represent standard error between experimental replicates.
The composition of FAME produced via transesterification of the lipid extract varies
significantly between each species and in some instances with nitrate loading (Figure 19).
The variation in FAME profile between species can potentially be used as a means to
0
100
200
300
400
500
600
700
0 10 20 30 40 50
FAM
E p
rodu
ced/
cul
ture
vol
ume
(mg/
L)
FAME produced/cell mass (mg/mg, %)
Neo Chl
Tet Nan
127
control the biodiesel and co-product characteristics; that is, specific strains and certain
growth conditions may be chosen for certain desired end products. The FAME profile is
further quantified in terms of FAME properties including average chain length, percentage
polyunsaturated fatty acids (> 1 double bond), the average degree of unsaturation, and the
percentage of unsaturated fatty acids in Table 13 for each species under N-replete and N-
deplete conditions. These metrics for other commonly used biodiesel feedstocks are also
listed for comparison.
Figure 19- Fatty acid methyl ester profile of lipid extracts for N. oleoabundans (Neo), C. sorokiniana (Chl), N. oculata (Nan), and T. suecica (Tet) under N-deprived (N-) and N-replete (N+) growth conditions.
128
Table 13- Lipid profiles of N. oleoabundans (Neo), C. sorokiniana (Chl), N. oculata (Nan), and T. suecica (Tet) grown under nitrogen replete and deplete conditions. The lipid profiles of other established biofuel feedstocks from (Moser, 2008) are included for comparison.
Biomass Feedstock NitrogenAverage chain length
% Polyunsaturated fatty acid
% Unsaturated FAME
Fre
shw
ater
alg
ae
Neochloris oleoabundans (Neo)
deplete (-)
17.24 ± 0.09 50.73 ± 2.52 62.14 ± 5.50
replete (+)
16.38 ± 0.15 67.92 ± 3.91 71.63 ± 7.55
Chlorella sorokiniana (Chl)
- 17.16 ± 0.02 47.50 ± 3.16 56.14 ± 5.57
+ 17.12 ± 0.04 68.37 ± 1.56 76.91 ± 2.62
Mar
ine
alga
e Tetraselmis suecica (Tet)
- 17.06 ± 0.05 11.49 ± 2.82 48.21 ± 8.06
+ 17.16 ± 0.15 9.15 ± 4.65 21.96 ± 13.28
Nannochloropis oculata (Nan)
- 16.40 ± 0.03 11.87 ± 0.47 59.85 ± 2.74
+ 16.21 ± 0.03 12.72 ± 0.42 54.68 ± 4.31
Cro
p-ba
sed
Canola N/A 17.93 27.8 92.3
Palm N/A 17.10 10.4 51.8
Soy N/A 17.79 61.3 85.4
Sunflower N/A 17.96 8.2 90.2
FAME profiles and characteristics can inform the type of products that would be preferable;
for instance, biodiesel properties such as oxidative stability and cold flow are extremely
important in defining biodiesel use.(M. J. Griffiths et al., 2012; Moser & Vaughn, 2012)
In general the FAME from algae are shorter than that of canola, palm, soy, and sunflower
oils with an average chain length across all species and growth conditions of 16.8 compared
to 17.9, 17.1, 17.8, and 18.0 for the oils respectively. These shorter chain lengths may be
beneficial for cold flow properties and viscosity(Knothe, 2005; Moser & Vaughn, 2012),
but are still long enough to not significantly affect the heat of combustion and cetane
129
number (Knothe, 2005; Moser & Vaughn, 2012). In fact, the four algal strains have almost
non-existent amounts (< 1.6% of total FAME production) of very long chain fatty acids (>
20 carbons), which have significant effect on fuel viability; if found in high concentrations,
these long chain FAME will cause the fuel product to suffer in terms of kinematic viscosity,
derived cetane number, and cold flow properties. (Moser & Vaughn, 2012) The percentage
of polyunsaturated fatty acids (% PUFA) ranges from 9.2 – 50.7%, which is within the
range of the conventional biomass feedstocks (8.2 – 61.3%) except for the two freshwater
strains in N-replete conditions (~68% PUFA). Minimizing % PUFA is necessary to ensure
oxidative stability of the biodiesel product. (Moser & Vaughn, 2012) Further, the
unsaturated lipid percentages (48.2 – 76.9%) fall in the range of the other feedstocks (51.8
– 92.3%) except for T. suecica with 22.0% due to large amounts of methyl palmitate
(C16:0) and methyl stearate (C18:0). The percent of unsaturated FAME needs to be high
enough to favor cold flow while polyunsaturated FAME need to be moderated as they have
poor oxidative stability. (Moser & Vaughn, 2012) These results suggest that the effects of
strain and growth conditions can play an important role in the properties of the final fuel
product and must be chosen carefully to allow for an efficient and effective biodiesel
production process.
5.3.3. Biochemical compositions: Lipid, protein, starch:
As seen in Figure 20, the lipid, protein, and starch compositions vary significantly between
species and less so between nitrogen conditions. For instance, T. suecica, while lower in
lipid (11.6% FAME for N-, 1.6% for N+) than the other species, is high in protein (41.9%
protein for N-, 49.3% for N+ compared to an average of 14.4% for the other species) and
thus may be considered for purposes other than fuel such as animal feed.(Spolaore et al.,
130
2006) As expected due to the necessity of nitrogen for amino acid synthesis, the protein
content of the N-deficient algae is lower than the N-replete. When considering the
appropriateness of microalgae for a given application, this composition must be weighed
to find the most economical and environmentally preferable strain and product pair. For
instance, the high biomass density that is attained by the freshwater species may be ideal
for thermochemical conversion into biocrude (Brennan & Owende, 2010), though the high
nutrient inputs and challenges in loss of ability to isolate co-products may unfavorably tip
the economic, energy, and resource balance. Alternatively, the species with high starch
content may be more suitable for anaerobic digestion or biofermentation and subsequent
production of bioethanol. (Mata et al., 2010)
Figure 20- Lipid, protein, and starch profiles (as percent dry mass) of N. oleoabundans (Neo), C. sorokiniana (Chl), T. suecica (Tet), and N. oculata (Nan)
131
From these controlled growth studies and subsequent analyses, it is clear that there are
different opportunities associated with different algal compositions, productivities, growth
conditions, and product endpoints. The following life cycle assessment was done in order
to compare the potential environmental benefits and impacts associated with an algal
biorefinery considering different compositions of the biomass feedstock. It is important to
consider that the data reflected here represent results from bench-scale growth studies
where the productivities are not necessarily representative of what may be obtained at large
scale. Depending on several growth parameters including reactor size, configuration, and
orientation, productivities may vary significantly, though it has been found that
optimization of these parameters may in fact preserve high productivities when growing in
large volumes.(Ugwu, Aoyagi, & Uchiyama, 2008) Scale-up of these processes remains a
major challenge in terms of bio-process engineering for algal growth systems. However,
the fast pace of reactor development has been accompanied by an increase of biomass
productivities in large-scale reactors, which are starting to approach those observed on a
smaller scale. The following LCA study based on empirical data is helpful to provide a
fundamental basis for the analysis and subsequent results.
5.3.4. Life Cycle Assessment: Energy consumption, greenhouse gas emissions, and eutrophication potential:
The controlled growth studies show clear differences between the algae species considered;
these differences are also reflected in the LCA results for energy consumption, GHG
emissions and eutrophication impacts. Figure 21 shows LCA results for each growth
scenario, with contributions from each life cycle stage to the right of the y-axis and avoided
burdens (or environmental benefits) due to co-products to the left of the axis.
132
Several patterns are evident. First, it appears that while all of the growth scenarios produce
biodiesel with positive net GHG emissions, the N-replete freshwater scenarios Chl+ and
Neo+ have among the lowest at 2.4 and 0.5 kg CO2e per kg of biodiesel, respectively.
These cases had lipid contents significantly lower than their N-deplete counterparts but
total volumetric lipid productivities were among the highest of all scenarios with Chl+
leading at 550 mg/L. The Neo+ result is equivalent to 13 g CO2e per MJ of fuel, well under
the 50% reduction threshold set by the RFS Baseline Renewable Fuel Standard (RFS) as
compared to the RFS Baseline for petroleum diesel (U.S. Environmental Protection
Agency, 2010). The GHG results underlines the fact that using nutrient deprivation to
enrich a particular biomass fraction while sacrificing total lipid productivity may not be
desirable when considering the entire system.
In all cases but one (Nan-), production of nutrients was the largest contributor to GHG
emissions, ranging from 27% for Tet+ to 64% for Chl-, even though the GREET model
includes internal recycling of N and P after biogas digestion. The exception is for the
marine species Nan-, where N-deplete conditions and a fairly high nutrient recycling rate
of >60% drive down impacts of nutrient production, leaving electricity use for mixing, CO2
and water delivery as the largest contributor to GHG emissions. Co-product credits for
GHG emissions and primary energy use in all cases is due to surplus electricity generation
from biogas production through anaerobic digestion, whereas substitution of algal protein
for soybean meal is the largest credit to eutrophication impacts.
133
Considering primary energy use, energy consumption exceeds energy delivered in
biodiesel (assuming biodiesel HHV of 37 MJ/kg) leading to EROI<1 for all cases with one
exception, where the Neo+ scenario has an EROI of 1.2. Neo+ has the highest volumetric
productivity of 2.9 g/L after 9 days but a lipid content of only 7%. Therefore, the co-
products derived from the lipid-extracted algae deliver most of the energy in the form of
credits from avoided use of electricity, heat, fertilizer, and animal meal. Interestingly, the
Tet+ scenario, with its extremely low lipid productivity, has the lowest EROI of 0.08. In a
microalgal integrated biorefinery setup where all lipid, starch, and protein fractions are
effectively utilized, there are important trade-offs that occur among processing choices for
these fractions that are not independent. In the GREET harmonization study (Davis et al.,
2012), the protein fraction is assumed to be digested, and a portion of the nutrients from
the resulting residues recycled back to cultivation, while another portion is used to offset
fertilizer use and sequester carbon on fields. However, when the protein fraction is instead
extracted to serve as animal feed, increased synthetic nutrients are required for algae while
simultaneously reducing agricultural nutrient inputs for soybeans from substituted meal.
Anaerobic digestion and subsequent combined heat and power production typically
supplies >100% of the heat and a substantial portion of the total electricity requirements
for cultivation up to lipid transesterification. If a different biorefinery setup were used that
processed the starch fraction into bioethanol rather than employing anaerobic digestion,
on-site biogas production would likely be reduced, potentially requiring external energy
sources to make up the lost heat and electricity. (Further integration is also possible, for
example by utilizing surplus heat from other industrial processes located with or near the
biorefinery.)
134
It was hypothesized that the existence of nutrients and micronutrients in seawater would
lower the impacts associated with growth media for marine algal species, as fewer synthetic
chemicals would be required. Background concentrations of nitrogen in the feed water and
buffers contributed toward total nitrogen specified in the growth media. These levels were
insignificant in the N-replete cases, but comprised 17% (marine) and 42% (freshwater) of
total N (without recycling) in the N-deplete cases. The concomitant reductions in required
NaNO3 inputs therefore had relatively small benefits relative to total impacts in the N-
replete cases, while for the N-deplete cases, life cycle impacts were driven by the
production of chemicals other than NaNO3, and these differed between marine and
freshwater media. In particular, the eutrophication impacts seen in the freshwater cases, C.
sorokiniana and N. oleoabundans, (Figure 21) are largely driven by the production of
mono- and dipotassium phosphate, which has much higher life cycle eutrophication
impacts than the sodium glycerophosphate required for the marine media.
135
Figure 21- Life cycle impacts for GHG emissions, eutrophication, and primary energy use per kg of biodiesel for N-replete and N-deplete growth conditions
Neo (N+)
Neo (N-)
Chl (N+)
Chl (N-)
Tet (N-)
Nanno (N+)
Nanno (N-)
-200 -150 -100 -50 0 50 100 150 200 250Primary Energy Use (MJ)
-690
Neo (N+)
Neo (N-)
Chl (N+)
Chl (N-)
Tet (N-)
Nanno (N+)
Nanno (N-)
-15 -10 -5 0 5 10 15 20GHG Emissions (kg CO2e)
Neo (N+)
Neo (N-)
Chl (N+)
Chl (N-)
Tet (N+)
Tet (N-)
Nanno (N+)
Nanno (N-)
-0.10 0.00 0.10 0.20 0.30Eutrophication (kg N)
conversion
ImpactsAvoided Impacts
cultivation extraction recoveryprotein meal
dewateringelectricity
natural gas
fertilizer + C storage
Tet (N+)
Tet (N+) 1180
-55 113
% Lipids 22%
18%
13%
2%
26%
19%
35%
9%
Growth Scheme
136
5.3.5. Implications for Microalgal Integrated Biorefinery Schemes
The orientation of the GREET model and most LCAs of algal biomass is toward biodiesel
or green diesel; however, algal biorefineries need not be optimized for lipid productivity.
For example, the growth scenario with the lowest lipid productivity (only 2% for Tet-)
clearly had the highest impacts for both GHG emissions and eutrophication, due to the
large quantities of this microalgae required to produce a unit of biodiesel. Instead of
biodiesel, this species may be well suited to fermentation into ethanol or direct use as an
animal feed supplement due to its large starch and protein content (Figure 20).
Consequently, avoiding the significant energy and chemical resource consumption
associated with lipid extraction and conversion.
While this study has emphasized microalgae cultivation, it is important to consider other
life cycle aspects of this comparison between freshwater and marine species and N-replete
and N-deplete cases. Several other LCA studies (Campbell et al., 2011; Jorquera et al.,
2010) that have modeled coastal production with marine species have assumed various
combinations of fertilizers for the growth media, with most omitting micronutrients under
the assumption that these are already present in non-limiting quantities in seawater. If
micronutrient sources are indeed flexible, then recipes for growth media may be optimized
using a strategy of minimizing high-impact synthetic chemicals. In this study the growth
of the marine species did not appear to be nitrate limited in the N-replete conditions, and
other nutrients were likely limiting growth such as iron for marine microalgal species.
Optimization of the media will likely increase the biomass density and decrease overall life
cycle impacts. Other implications involved with cultivation in untreated seawater must
137
also be considered, such as the presence of contaminating biota which could necessitate a
resilient algal strain or energy intensive seawater pretreatment such as filtering or
pasteurization as recommended by (UTEX, 2011).
5.4. Conclusions Maximizing productivity of a single algae fraction does not a priori lead to optimal
environmental outcomes. Microalgae with higher lipid productivity do not necessarily lead
to lower environmental impacts. However, engineered increases in lipid productivity
should be carefully balanced against intended uses of the non-lipid fractions, particularly
given the significant benefits realized through anaerobic digestion of the starch fraction.
Targeted extraction of high-value compounds for pharmaceutical or chemical industries
may greatly improve the economic performance of algal production systems, while
beneficial use of remaining fractions for lower-value end-uses can improve overall
biorefinery performance in environmental terms.
138
REFERENCES
Adom, F., Dunn, J. B., Han, J., & Sather, N. (2014). Life-Cycle Fossil Energy Consumption and Greenhouse Gas Emissions of Bioderived Chemicals and Their Conventional Counterparts. Environmental Science & Technology, 48(24), 14624–14631.
Anastas, P., & Eghbali, N. (2010). Green chemistry: principles and practice. Chemical Society Reviews, 39(1), 301–312.
Aouf, C., Nouailhas, H., Fache, M., Caillol, S., Boutevin, B., & Fulcrand, H. (2013). Multi-functionalization of gallic acid. Synthesis of a novel bio-based epoxy resin. European Polymer Journal, 49(6), 1185–1195.
Arkell, A., Olsson, J., & Wallberg, O. (2014). Process performance in lignin separation from softwood black liquor by membrane filtration. Chemical Engineering Research and Design, 92(9), 1792–1800.
Athawale, V. D., & Nimbalkar, R. V. (2011). Waterborne coatings based on renewable oil resources: an overview. Journal of the American Oil Chemists’ Society, 88(2), 159–185.
B. Dunn, J., Adom, F., & Sather, N. (n.d.). Life-cycle Analysis of Bioproducts and Their Conventional Counterparts in GREET.
Bardi, U. (2009). Peak oil: The four stages of a new idea. Energy, 34(3), 323–326. Bare, J. (2011). TRACI 2.0: the tool for the reduction and assessment of chemical and
other environmental impacts 2.0. Clean Technologies and Environmental Policy, 13(5), 687–696.
Bare, J. C., Hofstetter, P., Pennington, D. W., & De Haes, H. A. U. (2000). Midpoints versus endpoints: the sacrifices and benefits. The International Journal of Life Cycle Assessment, 5(6), 319–326.
Barta, K., Warner, G. R., Beach, E. S., & Anastas, P. T. (2014). Depolymerization of organosolv lignin to aromatic compounds over Cu-doped porous metal oxides. Green Chemistry, 16(1), 191–196.
Bell, A. A. (1986). Physiology of secondary products. Cotton Physiology. The Cotton Foundation, Memphis, TN, 597–621.
Bertz, S. H. (1981). The first general index of molecular complexity. Journal of the American Chemical Society, 103(12), 3599–3601.
Bidoki, S. M., Wittlinger, R., Alamdar, A. A., & Burger, J. (2006). Eco-efficiency analysis of textile coating materials. Journal of the Iranian Chemical Society, 3(4), 351–359.
Binder, J. B., Gray, M. J., White, J. F., Zhang, Z. C., & Holladay, J. E. (2009). Reactions of lignin model compounds in ionic liquids. Biomass and Bioenergy, 33(9), 1122–1130.
139
BioAmber. (2013). BioAmber Biobased Platform Chemical Life Cycle Anlaysis. Retrieved from
http://www.bioamber.com/ignitionweb/data/media_centre_files/804/Ontario_Canada_LCA_04.16.2013.pdf
Biobased Industries Consortium. (2012). The bio-based industries vision accelerating innovation and market uptake of bio-based products.
Biofuel.org.uk. (2016). Major Biofuel Producers by Region. Retrieved from http://biofuel.org.uk/major-producers-by-region.html
BIOSAT. (2011). Retrieved August 16, 2014, from http://www.biosat.net/Agricultural.html
Bold, H. C. (1970). Some aspects of the taxonomy of soil algae. Annals of the New York Academy of Sciences, 175(1), 601–616.
Bold, H. C., & Wyynne, M. J. (1978). Introduction to the algae: structure and reproduction. Prentice Hall. Retrieved from http://afrilib.odinafrica.org/handle/0/22921
Bonchev, D., & Trinajstić, N. (1977). Information theory, distance matrix, and molecular branching. The Journal of Chemical Physics, 67(10), 4517–4533.
BonSucro. (2014). Retrieved from http://bonsucro.com/site/production-standard/ Boopathy, R., & Dawson, L. (2008). Cellulosic ethanol production from sugarcane
baggase without enzymatic saccharification. BioResources, 3(2), 452–460. Börjesson, P., & Tufvesson, L. M. (2011). Agricultural crop-based biofuels–resource
efficiency and environmental performance including direct land use changes. Journal of Cleaner Production, 19(2), 108–120.
Bouwman, A. (1996). Direct emission of nitrous oxide from agricultural soils. Nutrient Cycling in Agroecosystems, 46(1), 53–70.
Bozell, J. J., & Petersen, G. R. (2010). Technology development for the production of biobased products from biorefinery carbohydrates—the US Department of Energy’s “top 10” revisited. Green Chemistry, 12(4), 539–554.
Brennan, L., & Owende, P. (2010). Biofuels from microalgae—a review of technologies for production, processing, and extractions of biofuels and co-products. Renewable and Sustainable Energy Reviews, 14(2), 557–577.
Brentner, L. B., Eckelman, M. J., & Zimmerman, J. B. (2011). Combinatorial Life Cycle Assessment to Inform Process Design of Industrial Production of Algal Biodiesel. Environmental Science & Technology, 45(16), 7060–7067.
Breysse, P. N., Delfino, R. J., Dominici, F., Elder, A. C., Frampton, M. W., Froines, J. R., … Hopke, P. K. (2013). US EPA particulate matter research centers: summary of research results for 2005–2011. Air Quality, Atmosphere & Health, 6(2), 333–355.
140
Broeren, M., Saygin, D., & Patel, M. K. (2014). Forecasting global developments in the basic chemical industry for environmental policy analysis. Energy Policy, 64, 273–287.
Brosse, N., Dufour, A., Meng, X., Sun, Q., & Ragauskas, A. (2012). Miscanthus: a fast�growing crop for biofuels and chemicals production. Biofuels, Bioproducts and Biorefining, 6(5), 580–598.
Brown, T. M., Duan, P., & Savage, P. E. (2010). Hydrothermal liquefaction and gasification of Nannochloropsis sp. Energy & Fuels, 24(6), 3639–3646.
Brundtland, G., Khalid, M., Agnelli, S., Al-Athel, S., Chidzero, B., Fadika, L., Singh, M. (1987). Our common future.
Brune, D. E., Lundquist, T. J., & Benemann, J. R. (2009). Microalgal biomass for greenhouse gas reductions: potential for replacement of fossil fuels and animal feeds. Journal of Environmental Engineering, 135(11), 1136–1144.
BSI. (2011). PAS 2050, Specification for the assessment of the life cycle greenhouse gas emissions of goods and services (No. ICS code: 13.020.40). British Standard Institute.
Campbell, P. K., Beer, T., & Batten, D. (2011). Life cycle assessment of biodiesel production from microalgae in ponds. Bioresource Technology, 102(1), 50–56.
Carlson, K., Carr, M., & Cunningham, R. (1983). Lignin analyses on sweet sorghum samples. Transactions of the Illinois State Academy of Science, 76(3–4), 265–269.
Carus, M., Dammer, L., Hermann, A., & Essel, R. (2014). Proposals for a reform of the Renewable Energy Directive to a Renewable Energy and Materials Directive (REMD), Going to the next level: Integration of bio-based chemicals and materials in the incentive scheme. Nova Paper, (4), 2014–5.
Cespi, D., Passarini, F., Vassura, I., & Cavani, F. (2016). Butadiene from biomass, a life cycle perspective to address sustainability in the chemical industry. Green Chemistry, 18(6), 1625–1638.
Chandel, A. K., da Silva, S. S., Carvalho, W., & Singh, O. V. (2012). Sugarcane bagasse and leaves: foreseeable biomass of biofuel and bio�products. Journal of Chemical Technology and Biotechnology, 87(1), 11–20.
Cherubini, F., Bird, N. D., Cowie, A., Jungmeier, G., Schlamadinger, B., & Woess-Gallasch, S. (2009). Energy-and greenhouse gas-based LCA of biofuel and bioenergy systems: Key issues, ranges and recommendations. Resources, Conservation and Recycling, 53(8), 434–447.
Cherubini, F., & Jungmeier, G. (2010). LCA of a biorefinery concept producing bioethanol, bioenergy, and chemicals from switchgrass. The International Journal of Life Cycle Assessment, 15(1), 53–66.
Cherubini, F., & Strømman, A. H. (2011). Chemicals from lignocellulosic biomass: opportunities, perspectives, and potential of biorefinery systems. Biofuels, Bioproducts and Biorefining, 5(5), 548–561.
141
Chum, H. L., & Overend, R. P. (2001). Biomass and renewable fuels. Fuel Processing Technology, 71(1), 187–195.
Clarens, A. F., Resurreccion, E. P., White, M. A., & Colosi, L. M. (2010). Environmental Life Cycle Comparison of Algae to Other Bioenergy Feedstocks. Environmental Science & Technology, 44(5), 1813–1819.
Cok, B., Tsiropoulos, I., Roes, A. L., & Patel, M. K. (2014). Succinic acid production derived from carbohydrates: An energy and greenhouse gas assessment of a platform chemical toward a bio�based economy. Biofuels, Bioproducts and Biorefining, 8(1), 16–29.
Conde-Mejía, C., Jiménez-Gutiérrez, A., & El-Halwagi, M. (2012). A comparison of pretreatment methods for bioethanol production from lignocellulosic materials. Process Safety and Environmental Protection, 90(3), 189–202.
Davis, R., Fishman, D., Frank, E. D., Wigmosta, M. S., Aden, A., Coleman, A. M., Wang, M. Q. (2012). Renewable diesel from algal lipids: an integrated baseline for cost, emissions, and resource potential from a harmonized model. Argonne National Laboratory. Retrieved from https://www.researchgate.net/profile/Andre_Coleman/publication/254999713_Renewable_diesel_from_algal_lipids_An_integrated_baseline_for_cost_emissions_and_resource_potential_from_a_harmonized_model_ANLESD12-4_NRELTP-5100-5543_1_PNNL-21437/links/5502fc490cf2d60c0e64c415.pdf
Daystar, J., Treasure, T., Reeb, C., Venditti, R., Gonzalez, R., & Kelley, S. (2015). Environmental impacts of bioethanol using the NREL biochemical conversion route: multivariate analysis and single score results. Biofuels, Bioproducts and Biorefining, 9(5), 484–500.
De Jong, E., Dam, M., Sipos, L., & Gruter, G. (2012). Furandicarboxylic acid (FDCA), a versatile building block for a very interesting class of polyesters. Biobased Monomers, Polymers, and Materials, 1105, 1–13.
De Jong, E., Higson, A., Walsh, P., & Wellisch, M. (2012). Bio-based chemicals value added products from biorefineries. IEA Bioenergy, Task42 Biorefinery.
Deaner, M. J., Puppin, G., & Heikkila, K. E. (1996). Advanced polymer/wood composite structural member.
Del Río, J. C., Rencoret, J., Prinsen, P., Martínez, Á. T., Ralph, J., & Gutiérrez, A. (2012). Structural characterization of wheat straw lignin as revealed by analytical pyrolysis, 2D-NMR, and reductive cleavage methods. Journal of Agricultural and Food Chemistry, 60(23), 5922–5935.
Demirbas, M. F. (2009). Biorefineries for biofuel upgrading: a critical review. Applied Energy, 86, S151–S161.
Derksen, J. T., Cuperus, F. P., & Kolster, P. (1996). Renewable resources in coatings technology: a review. Progress in Organic Coatings, 27(1), 45–53.
142
Diamond, G. M., Murphy, V., & Boussie, T. R. (2014). Application of High Throughput Experimentation to the Production of Commodity Chemicals from Renewable Feedstocks. Modern Application of High Throughput R&D in Heterogeneous Catalysis; Hagemeyer, A.; Volpe, AF, Eds.: Bentham Science Publishers: Sharjah (UAE), 288–309.
Drauz, K.-H., Kleeman, A., Prescher, G., & Ritter, G. (1991). Method of producing catechol and hydroquinone.
Dunford, N. T. (2012). Food and industrial bioproducts and bioprocessing. John Wiley & Sons.
Eerhart, A., Faaij, A., & Patel, M. K. (2012). Replacing fossil based PET with biobased PEF; process analysis, energy and GHG balance. Energy & Environmental Science, 5(4), 6407–6422.
Ekman, A., & Börjesson, P. (2011). Environmental assessment of propionic acid produced in an agricultural biomass-based biorefinery system. Journal of Cleaner Production, 19(11), 1257–1265.
Ekvall, T., & Finnveden, G. (2001). Allocation in ISO 14041—a critical review. Journal of Cleaner Production, 9(3), 197–208.
Elkington, J. (2001). The triple bottom line for 21st century business. The Earthscan Reader in Business and Sustainable Development, 20–43.
eXtension. (2013). Woody Biomass Properties. Retrieved November 7, 2016, from http://articles.extension.org/pages/26517/woody-biomass-properties
Fan, L., Hu, W.-R., Yang, Y., & Li, B. (2012). Plant Special Cell–Cotton Fiber. INTECH
FAO (Trans.). (2005). Fertilizer use by crop in India. Land and Water Development Divison. Retrieved from ftp://ftp.fao.org/agl/agll/docs/fertuseindia.pdf
FAOSTAT. (2013). Retrieved May 19, 2014, from http://faostat3.fao.org/faostat-gateway/go/to/browse/Q/*/E
Fargione, J., Hill, J., Tilman, D., Polasky, S., & Hawthorne, P. (2008). Land clearing and the biofuel carbon debt. Science, 319(5867), 1235–1238.
Farm Industry News. (2016). The 5 largest ethanol producers. Retrieved from http://farmindustrynews.com/ethanol/5-largest-ethanol-producers
Farm Security and Rural Investment Act, Pub. L. No. 107–171 (2002). Fernandes, B., Dragone, G., Abreu, A. P., Geada, P., Teixeira, J., & Vicente, A. (2012).
Starch determination in Chlorella vulgaris—a comparison between acid and enzymatic methods. Journal of Applied Phycology, 24(5), 1203–1208.
Fernando, S., Adhikari, S., Chandrapal, C., & Murali, N. (2006). Biorefineries: current status, challenges, and future direction. Energy & Fuels, 20(4), 1727–1737.
Fiege, H., Voges, H.-W., Hamamoto, T., Umemura, S., Iwata, T., Miki, H., Paulus, W. (2000). Phenol Derivatives. In Ullmann’s Encyclopedia of Industrial Chemistry. Wiley-VCH Verlag GmbH & Co. KGaA.
143
Flint, S., Markle, T., Thompson, S., & Wallace, E. (2012). Bisphenol A exposure, effects, and policy: a wildlife perspective. Journal of Environmental Management, 104, 19–34.
Frank, E., Han, J., Palou-Rivera, I., Elgowainy, A., & Wang, M. (2011). User Manual for algae life-cycle analysis with GREET: Version 0.0. Retrieved from http://greet.es.anl.gov/publications
Frischknecht, R., Jungbluth, N., Althaus, H., Bauer, C., Doka, G., Dones, R., Köllner, T. (2007). Implementation of life cycle impact assessment methods. Ecoinvent Report.
Fritz, J., Cantrell, R., Lechtenberg, V., Axtell, J., & Hertel, J. (1981). Brown midrib mutants in sudangrass and grain sorghum. Crop Science, 21(5), 706–709.
Furuholt, E. (1995). Life cycle assessment of gasoline and diesel. Resources, Conservation and Recycling, 14(3), 251–263.
Gallardo Hipolito, M. (2011). Life Cycle Assessment of platform chemicals from fossil and lignocellulosic biomass scenarios: LCA of phenolic compounds, solvent, soft and hard plastic precursors.
George, B., Suttie, E., Merlin, A., & Deglise, X. (2005). Photodegradation and photostabilisation of wood–the state of the art. Polymer Degradation and Stability, 88(2), 268–274.
Goeppert, A., Czaun, M., Jones, J.-P., Prakash, G. S., & Olah, G. A. (2014). Recycling of carbon dioxide to methanol and derived products–closing the loop. Chemical Society Reviews, 43(23), 7995–8048.
Golden, J. S., Handfield, R. B., Daystar, J., & McConnell, T. E. (2015). An economic impact analysis of the US biobased products industry: A report to the congress of the United States of America. Industrial Biotechnology, 11(4), 201–209.
Griffiths, M. J., & Harrison, S. T. (2009). Lipid productivity as a key characteristic for choosing algal species for biodiesel production. Journal of Applied Phycology, 21(5), 493–507.
Griffiths, M. J., van Hille, R. P., & Harrison, S. T. (2012). Lipid productivity, settling potential and fatty acid profile of 11 microalgal species grown under nitrogen replete and limited conditions. Journal of Applied Phycology, 24(5), 989–1001.
Griffiths, O. G., Owen, R. E., O’Byrne, J. P., Mattia, D., Jones, M. D., & McManus, M. C. (2013). Using life cycle assessment to measure the environmental performance of catalysts and directing research in the conversion of CO 2 into commodity chemicals: a look at the potential for fuels from “thin-air.” RSC Advances, 3(30), 12244–12254.
Groot, W. J., & Borén, T. (2010). Life cycle assessment of the manufacture of lactide and PLA biopolymers from sugarcane in Thailand. The International Journal of Life Cycle Assessment, 15(9), 970–984.
144
Gross, R. A., & Kalra, B. (2002). Biodegradable polymers for the environment. Science, 297(5582), 803–807.
Guinee, J. B., Heijungs, R., Huppes, G., Zamagni, A., Masoni, P., Buonamici, R., Rydberg, T. (2010). Life cycle assessment: past, present, and future†. Environmental Science & Technology, 45(1), 90–96.
Gustafsson, L., & Börjesson, P. (2007). Life cycle assessment in green chemistry. The International Journal of Life Cycle Assessment, 12(3), 151–159.
Häkkinen, T., Ahola, P., Vanhatalo, L., & Merra, A. (1999). Environmental impact of coated exterior wooden cladding. VTT Technical Research Centre of Finland, Finland.
Hall, D., & Scrase, J. (1998). Will biomass be the environmentally friendly fuel of the future? Biomass and Bioenergy, 15(4), 357–367.
Hansen, T. S., Barta, K., Anastas, P. T., Ford, P. C., & Riisager, A. (2012). One-pot reduction of 5-hydroxymethylfurfural via hydrogen transfer from supercritical methanol. Green Chemistry, 14(9), 2457–2461.
Harding, K., Dennis, J., Von Blottnitz, H., & Harrison, S. (2007). Environmental analysis of plastic production processes: Comparing petroleum-based polypropylene and polyethylene with biologically-based poly-β-hydroxybutyric acid using life cycle analysis. Journal of Biotechnology, 130(1), 57–66.
Haro, P., Villanueva Perales, A. L., Arjona, R., & Ollero, P. (2014). Thermochemical biorefineries with multiproduction using a platform chemical. Biofuels, Bioproducts and Biorefining, 8(2), 155–170.
Hatti-Kaul, R., Törnvall, U., Gustafsson, L., & Börjesson, P. (2007). Industrial biotechnology for the production of bio-based chemicals–a cradle-to-grave perspective. Trends in Biotechnology, 25(3), 119–124.
Hermann, B., Blok, K., & Patel, M. K. (2007). Producing bio-based bulk chemicals using industrial biotechnology saves energy and combats climate change. Environmental Science & Technology, 41(22), 7915–7921.
Hess, W. T., Kurtz, A., & Stanton, D. (1995). Kirk-Othmer encyclopedia of chemical technology. John Wiley & Sons Ltd., New York.
Hill, J., Nelson, E., Tilman, D., Polasky, S., & Tiffany, D. (2006). Environmental, economic, and energetic costs and benefits of biodiesel and ethanol biofuels. Proceedings of the National Academy of Sciences, 103(30), 11206–11210.
Hofland, A. (2012). Alkyd resins: From down and out to alive and kicking. Progress in Organic Coatings, 73(4), 274–282.
Hofrichter, M. (2002). Review: lignin conversion by manganese peroxidase (MnP). Enzyme and Microbial Technology, 30(4), 454–466.
145
Hogle, B. P., Shekhawat, D., Nagarajan, K., Jackson, J. E., & Miller, D. J. (2002). Formation and recovery of itaconic acid from aqueous solutions of citraconic acid and succinic acid. Industrial & Engineering Chemistry Research, 41(9), 2069–2073.
Holladay, J., Bozell, J., White, J., & Johnson, D. (2007). Top value-added chemicals from biomass-Volume II–Results of Screening for Potential Candidates from Biorefinery Lignin.
Hong, E., Kim, D., Kim, J., Kim, J., Yoon, S., Rhie, S., Ryu, Y. (2015). Optimization of alkaline pretreatment on corn stover for enhanced production of 1.3-propanediol and 2,3-butanediol by Klebsiella pneumoniae AJ4. Biomass and Bioenergy, 77, 177–185.
Horne, R., Grant, T., & Verghese, K. (2009). Life cycle assessment: Principles, practice, and prospects. Csiro Publishing.
Hottle, T. A., Bilec, M. M., & Landis, A. E. (2013). Sustainability assessments of bio-based polymers. Polymer Degradation and Stability, 98(9), 1898–1907.
Houghton, J., Meira Filho, L., Lim, B., Treanton, K., & Mamaty, I. (1997). Revised 1996 IPCC guidelines for national greenhouse gas inventories. v. 1: Greenhouse gas inventory reporting instructions.-v. 2: Greenhouse gas inventory workbook.-v. 3: Greenhouse gas inventory reference manual.
Huber, G. W., Iborra, S., & Corma, A. (2006). Synthesis of transportation fuels from biomass: chemistry, catalysts, and engineering. Chemical Reviews, 106(9), 4044–4098.
Huijbregts, M. A., Rombouts, L. J., Hellweg, S., Frischknecht, R., Hendriks, A. J., van de Meent, D., Struijs, J. (2006). Is cumulative fossil energy demand a useful indicator for the environmental performance of products? Environmental Science & Technology, 40(3), 641–648.
IEA. (2009). Bioenergy–a sustainable and reliable energy source. IEA. (2012). Energy technology perspectives 2012-pathways to a clean energy system. IPCC. (2015). Climate change 2014: mitigation of climate change. ISCC PLUS. (2011). GHG Emission Requirements (ISCC PLUS 205-01). International
Sustainability & Carbon Certification. ISO. (2006). 14040: Environmental management–life cycle assessment–principles and
framework. London: British Standards Institution. J. Dunn, F. A., N.Sather, J.Han, S.Snyder. (2014). Life-cycle Analysis of Bioproducts and
Their Conventional Counterparts in GREET. Argonne Natinal Labratory. Jessop, P., Ahmadpour, F., Buczynski, M., Burns, T., Green Ii, N., Korwin, R.,
Omidbakhsh, N. (2015). Opportunities for greener alternatives in chemical formulations. Green Chemistry, 17(5), 2664–2678.
146
Jolliet, O., Margni, M., Charles, R., Humbert, S., Payet, J., Rebitzer, G., & Rosenbaum, R. (2003). IMPACT 2002+: a new life cycle impact assessment methodology. The International Journal of Life Cycle Assessment, 8(6), 324–330.
Jørgensen, H., Vibe�Pedersen, J., Larsen, J., & Felby, C. (2007). Liquefaction of lignocellulose at high�solids concentrations. Biotechnology and Bioengineering, 96(5), 862–870.
Jorquera, O., Kiperstok, A., Sales, E. A., Embiruçu, M., & Ghirardi, M. L. (2010). Comparative energy life-cycle analyses of microalgal biomass production in open ponds and photobioreactors. Bioresource Technology, 101(4), 1406–1413.
Joslin, J., & Schoenholtz, S. (1997). Measuring the environmental effects of converting cropland to short-rotation woody crops: a research approach. Biomass and Bioenergy, 13(4), 301–311.
Kamm, B., & Kamm, M. (2004). Principles of biorefineries. Applied Microbiology and Biotechnology, 64(2), 137–145.
Kawaoka, A., Nanto, K., Ishii, K., & Ebinuma, H. (2006). Reduction of lignin content by suppression of expression of the LIM domain transcription factor in Eucalyptus camaldulensis. Silvae Genetica, 55(6), 269–277.
Kim, S., & Dale, B. E. (2005). Life cycle assessment of various cropping systems utilized for producing biofuels: Bioethanol and biodiesel. Biomass and Bioenergy, 29(6), 426–439.
Kim, S., & Dale, B. E. (2008). Energy and greenhouse gas profiles of polyhydroxybutyrates derived from corn grain: a life cycle perspective. Environmental Science & Technology, 42(20), 7690–7695.
Kimberly Davis, K. V., Mary, & Swanson. (2005). Wood Finishes and Stains. Green Seal Inc. Retrieved from https://www.wbdg.org/ccb/GREEN/REPORTS/cgrwoodfinish.pdf
Kleinert, M., & Barth, T. (2008). Phenols from lignin. Chemical Engineering & Technology, 31(5), 736–745.
Klöpffer, W. (1997). Life cycle assessment. Environmental Science and Pollution Research, 4(4), 223–228.
K.M. Fowler, E. M. R. (2006). Sustainable Building Rating Systems Summary (No. PNNL-15858). Pacific Northwest National Lab (PNNL): Pacific Northwest National Lab (PNNL).
Knothe, G. (2005). Dependence of biodiesel fuel properties on the structure of fatty acid alkyl esters. Fuel Processing Technology, 86(10), 1059–1070.
Krumenacker, L., Costantini, M., Pontal, P., & Sentenac, J. (1995). Hydroquinone, resorcinol, and catechol.
Kurian, J. K., Nair, G. R., Hussain, A., & Raghavan, G. V. (2013). Feedstocks, logistics and pre-treatment processes for sustainable lignocellulosic biorefineries: A comprehensive review. Renewable and Sustainable Energy Reviews, 25, 205–219.
147
Lammens, T. M., Potting, J., Sanders, J. P., & De Boer, I. J. (2011). Environmental comparison of biobased chemicals from glutamic acid with their petrochemical equivalents. Environmental Science & Technology, 45(19), 8521–8528.
Lardon, L., Hélias, A., Sialve, B., Steyer, J.-P., & Bernard, O. (2009). Life-Cycle Assessment of Biodiesel Production from Microalgae. Environmental Science & Technology, 43(17), 6475–6481.
Larson, E. D. (2006). A review of life-cycle analysis studies on liquid biofuel systems for the transport sector. Energy for Sustainable Development, 10(2), 109–126.
Lechtenberg, V., Colenbrander, V., Bauman, L., & Rhykerd, C. (1974). Effect of lignin on rate of in vitro cell wall and cellulose disappearance in corn. Journal of Animal Science, 39(6), 1165–1169.
Lin, Z., Nikolakis, V., & Ierapetritou, M. (2015). Life cycle assessment of biobased p-xylene production. Industrial & Engineering Chemistry Research, 54(8), 2366–2378.
Linak, E. (2009). Paint and Coatings Industry Overview. Chemical Economics Handbook Report. SRI Consulting.
Liptow, C., & Tillman, A. (2012). A comparative life cycle assessment study of polyethylene based on sugarcane and crude oil. Journal of Industrial Ecology, 16(3), 420–435.
Lithner, D., Larsson, Å. & Dave, G. (2011). Environmental and health hazard ranking and assessment of plastic polymers based on chemical composition. Science of the Total Environment, 409(18), 3309–3324.
Liu, X., Clarens, A. F., & Colosi, L. M. (2012). Algae biodiesel has potential despite inconclusive results to date. Bioresource Technology, 104, 803–806.
Luo, L., Van der Voet, E., Huppes, G., & De Haes, H. A. U. (2009). Allocation issues in LCA methodology: a case study of corn stover-based fuel ethanol. The International Journal of Life Cycle Assessment, 14(6), 529–539.
Macgregor, C. A. (2000). Directory of feeds and feed ingredients. Hoard’s Dairyman Books.
Maldas, D., & Kokta, B. (1991). Surface modification of wood fibers using maleic anhydride and isocyanate as coating components and their performance in polystyrene composites. Journal of Adhesion Science and Technology, 5(9), 727–740.
Mata, T. M., Martins, A. A., & Caetano, N. S. (2010). Microalgae for biodiesel production and other applications: a review. Renewable and Sustainable Energy Reviews, 14(1), 217–232.
Matos, S., & Petrov, O. (n.d.). An integrative LCA approach to evaluate sustainable new technology development: The case of lignin based resin.
Matson, P. A., Parton, W. J., Power, A., & Swift, M. (1997). Agricultural intensification and ecosystem properties. Science, 277(5325), 504–509.
148
Meerman, J., Ramírez, A., Turkenburg, W., & Faaij, A. (2011). Performance of simulated flexible integrated gasification polygeneration facilities. Part A: A technical-energetic assessment. Renewable and Sustainable Energy Reviews, 15(6), 2563–2587.
Meier, M. A., Metzger, J. O., & Schubert, U. S. (2007). Plant oil renewable resources as green alternatives in polymer science. Chemical Society Reviews, 36(11), 1788–1802.
Meier-Westhues, U. (2007). Polyurethanes: coatings, adhesives and sealants. Vincentz Network GmbH & Co KG.
Meijer, E. A., & Wijffels, R. H. (1998). Development of a fast, reproducible and effective method for the extraction and quantification of proteins of micro-algae. Biotechnology Techniques, 12(5), 353–358.
Mendu, V., Harman-Ware, A. E., Crocker, M., Jae, J., Stork, J., Morton, S., DeBolt, S. (2011). Identification and thermochemical analysis of high-lignin feedstocks for biofuel and biochemical production. Biotechnology for Biofuels, 4, 43.
Mila i Canals, L., Azapagic, A., Doka, G., Jefferies, D., King, H., Mutel, C., Williams, A. (2011). Approaches for Addressing Life Cycle Assessment Data Gaps for Bio-based Products. Journal of Industrial Ecology, 15.
Miller, S. A. (2010). Minimizing land use and nitrogen intensity of bioenergy. Environmental Science & Technology, 44(10), 3932–3939.
Mireles, J., Adame, A., Espalin, D., Medina, F., Winker, R., Hoppe, T., Wicker, R. (2011). Analysis of sealing methods for FDM-fabricated parts. University Of TEXAS-El PASO.
Modahl, I., Brekke, A., & Raadal, H. (2009). Life cycle assessment of cellulose, ethanol, lignin and vanillin from Borregaard, Sarpsborg. Østfoldforskning AS, Kråkerøy.
Mohan, D., Pittman, C. U., & Steele, P. H. (2006). Pyrolysis of wood/biomass for bio-oil: a critical review. Energy & Fuels, 20(3), 848–889.
Mohan, S. V., Modestra, J. A., Amulya, K., Butti, S. K., & Velvizhi, G. (2016). A Circular Bioeconomy with Biobased Products from CO 2 Sequestration. Trends in Biotechnology, 34(6), 506–519.
Mohanty, A., Misra, M., & Drzal, L. (2002). Sustainable bio-composites from renewable resources: opportunities and challenges in the green materials world. Journal of Polymers and the Environment, 10(1–2), 19–26.
Moncada, J., Posada, J. A., & Ramírez, A. (2015). Early sustainability assessment for potential configurations of integrated biorefineries. Screening of bio�based derivatives from platform chemicals. Biofuels, Bioproducts and Biorefining, 9(6), 722–748.
Montazeri, M., Zaimes, G. G., Khanna, V., & Eckelman, M. J. (2016). Meta-Analysis of Life Cycle Energy and Greenhouse Gas Emissions for Priority Bio-based Chemicals. ACS Sustainable Chemistry & Engineering.
149
Moore, J. E. (1990). Photocurable acrylic coating composition. Moser, B. R., & Vaughn, S. F. (2012). Efficacy of fatty acid profile as a tool for
screening feedstocks for biodiesel production. Biomass and Bioenergy, 37, 31–41. Mu, D., Seager, T., Rao, P. S., & Zhao, F. (2010). Comparative life cycle assessment of
lignocellulosic ethanol production: biochemical versus thermochemical conversion. Environmental Management, 46(4), 565–578.
Mueller, S. (2010). Life Cycle Analysis of Ethyl Lactate Production and Controlled Flow Cavitation at Corn Ethanol Plants. University of Illinois at Chicago, Energy Resources Center.
Munasinghe, P. C., & Khanal, S. K. (2010). Biomass-derived syngas fermentation into biofuels: opportunities and challenges. Bioresource Technology, 101(13), 5013–5022.
National Research Council. (2000). Biobased Industrial Products: Research and Commercialization Priorities. National Academy Press. Retrieved from https://www.nap.edu/catalog/5295/biobased-industrial-products-research-and-commercialization-priorities
Nautiyal, P. (2002). Groundnut: Post-harvest operations. (R. C. for G. (ICAR), Trans.) (p. 2013).
Ndazi, B. S., Nyahumwa, C. W., & Tesha, J. (2008). Chemical and thermal stability of rice husks against alkali treatment. BioResources, 3(4), 1267–1277.
Necpálová, M., Anex, R. P., Fienen, M. N., Del Grosso, S. J., Castellano, M. J., Sawyer, J. E., … Barker, D. W. (2015). Understanding the DayCent model: Calibration, sensitivity, and identifiability through inverse modeling. Environmental Modelling & Software, 66, 110–130.
Niu, J., & Burnett, J. (2001). Setting up the criteria and credit-awarding scheme for building interior material selection to achieve better indoor air quality. Environment International, 26(7), 573–580.
NNFC. (2014). Bio-based Chemicals, Industrial Sugars and Development of Biorefineries. Retrieved from http://www.slideshare.net/AHigson/biobased-chemicals-industrial-sugar-and-the-development-of-biorefineries
Noble, I., Bolin, B., Ravindranath, N., Verardo, D., & Dokken, D. (2000). Land use, land use change, and forestry. Cambridge University Press.
Norman, A. (1969). Constitution and Biosynthesis of Lignin. K. Freudenberg and AC Neish. Springer-Verlag, New York, 1968. Molecular Biology, Biochemistry and Biophysics, vol. 2.
NREL. (1999). The technology roadmap for plant/crop-based renewable resources 2020. Nuss, P., & Gardner, K. H. (2013). Attributional life cycle assessment (ALCA) of
polyitaconic acid production from northeast US softwood biomass. The International Journal of Life Cycle Assessment, 18(3), 603–612.
150
Octave, S., & Thomas, D. (2009). Biorefinery: toward an industrial metabolism. Biochimie, 91(6), 659–664.
Omni Tech International. (2011). United Soybean Board, Life cycle impact of soybean production and soy industrial products. The United Soybean Board.
Owens, J. (1997). Life cycle assessment. Journal of Industrial Ecology, 1(1), 37–49. Papasavva, S., Kia, S., Claya, J., & Gunther, R. (2001). Characterization of automotive
paints: an environmental impact analysis. Progress in Organic Coatings, 43(1–3), 193–206.
Paster, M., Pellegrino, J. L., & Carole, T. M. (2003). Industrial bioproducts: today and tomorrow. DOE-EERE.
Patel, M., Crank, M., Dornburg, V., Hermann, B., Roes, L., Huesing, B., Recchia, E. (2006). Medium and Long-term Opportunities and Risks of the Biotechnological Production of Bulk Chemical from Renewable Resources- the Potential of White Biotechnology: the BREW Project. Utrecht University.
Patel, M., Marscheider-Weidemann, F., Schleich, J., Hüsing, B., & Angerer, G. (2005). Techno-economic feasibility of large-scale production of bio-based polymers in Europe. Techncial Report EUR, 22103.
Pennington, D., Potting, J., Finnveden, G., Lindeijer, E., Jolliet, O., Rydberg, T., & Rebitzer, G. (2004). Life cycle assessment Part 2: Current impact assessment practice. Environment International, 30(5), 721–739.
Petrou, E. C., & Pappis, C. P. (2009). Biofuels: a survey on pros and cons. Energy & Fuels, 23(2), 1055–1066.
Philp, J. (2015). Balancing the bioeconomy: supporting biofuels and bio-based materials in public policy. Energy & Environmental Science, 8(11), 3063–3068.
Pienkos, P. T., & Darzins, A. L. (2009). The promise and challenges of microalgal-derived biofuels. Biofuels, Bioproducts and Biorefining, 3(4), 431–440.
Popp, J., Lakner, Z., Harangi-Rákos, M., & Fári, M. (2014). The effect of bioenergy expansion: food, energy, and environment. Renewable and Sustainable Energy Reviews, 32, 559–578.
Posen, I. D., Griffin, W. M., Matthews, H. S., & Azevedo, I. L. (2014). Changing the Renewable Fuel Standard to a Renewable Material Standard: Bioethylene Case Study. Environmental Science & Technology, 49(1), 93–102.
Posen, I. D., Jaramillo, P., & Griffin, W. M. (2016). Uncertainty in the Life Cycle Greenhouse Gas Emissions from US Production of Three Bio-based Polymer Families. Environmental Science & Technology, 50(6), 2846–2858.
Prado, R., Erdocia, X., Serrano, L., & Labidi, J. (2012). Lignin purification with green solvents. Cellulose Chemistry and Technology, 46(3), 221.
Punter, G., Rickeard, D., Larivé, J., Edwards, R., Mortimer, N., Horne, R., Woods, J. (2004). Well-to-wheel evaluation for production of ethanol from wheat. Report by the LowCVP Fuels Working Group, WTW Sub-Group.
151
Rajadhyaksha, R. A., & Chaudhari, D. D. (1987). Alkylation of phenol and pyrocatechol by isobutyl alcohol using superacid catalysts. Industrial & Engineering Chemistry Research, 26(7), 1276–1280.
Randić, M., & Plavs̆ić, D. (2003). Characterization of molecular complexity. International Journal of Quantum Chemistry, 91(1), 20–31.
Rebitzer, G. (2002). Integrating life cycle costing and life cycle assessment for managing costs and environmental impacts in supply chains. In Cost management in supply chains (pp. 127–146). Springer.
Rebitzer, G., Ekvall, T., Frischknecht, R., Hunkeler, D., Norris, G., Rydberg, T., Pennington, D. W. (2004). Life cycle assessment: Part 1: Framework, goal and scope definition, inventory analysis, and applications. Environment International, 30(5), 701–720.
Reddy, N., & Yang, Y. (2005). Biofibers from agricultural byproducts for industrial applications. Trends in Biotechnology, 23(1), 22–27.
Rodolfi, L., Chini Zittelli, G., Bassi, N., Padovani, G., Biondi, N., Bonini, G., & Tredici, M. R. (2009). Microalgae for oil: Strain selection, induction of lipid synthesis and outdoor mass cultivation in a low-cost photobioreactor. Biotechnology and Bioengineering, 102(1), 100–112.
Rosenbaum, R. K., Bachmann, T. M., Gold, L. S., Huijbregts, M. A., Jolliet, O., Juraske, R., Margni, M. (2008). USEtox—the UNEP-SETAC toxicity model: recommended characterisation factors for human toxicity and freshwater ecotoxicity in life cycle impact assessment. The International Journal of Life Cycle Assessment, 13(7), 532–546.
RSB. (2015). Standard for Certification of Bio-Products (No. RSB-STD-02-001 v 1.4). Roundtable on Sustainable Biomaterials.
Safety, O., & Administration, H. (2015). Indoor Air Quality in Commercial and Institutional Buildings. Maroon Ebooks.
Sannigrahi, P., Ragauskas, A. J., & Tuskan, G. A. (2010). Poplar as a feedstock for biofuels: a review of compositional characteristics. Biofuels, Bioproducts and Biorefining, 4(2), 209–226.
Santiaguel, A. F. (2013). A second life for rice husk. Retrieved May 31, 2014, from http://irri.org/rice-today/a-second-life-for-rice-husk
Schmidt, L. D., & Dauenhauer, P. J. (2007). Chemical engineering: hybrid routes to biofuels. Nature, 447(7147), 914–915.
Scown, C. D., Gokhale, A. A., Willems, P. A., Horvath, A., & McKone, T. E. (2014). Role of Lignin in Reducing Life-Cycle Carbon Emissions, Water Use, and Cost for United States Cellulosic Biofuels. Environmental Science & Technology, 48(15), 8446–8455.
152
Searchinger, T., Heimlich, R., Houghton, R. A., Dong, F., Elobeid, A., Fabiosa, J., Yu, T.-H. (2008). Use of US croplands for biofuels increases greenhouse gases through emissions from land-use change. Science, 319(5867), 1238–1240.
Sekyere, D. (1994). Potential of bamboo (Bambusa vulgaris) as a source of raw material for pulp and paper in Ghana. Ghana Journal of Forestry, 1(199), 50.
Shakina, J., Lekshmi, K. S., & Raj, G. (2012). Synthesis and Characterisation of Novel Cross Linked Biopolyesters from Olive Oil as Eco-friendly Biodegradable Materials. Journal of Chemistry, 9(1), 181–192.
Shen, L., Haufe, J., & Patel, M. K. (2009). Product overview and market projection of emerging bio-based plastics PRO-BIP 2009. Report for European Polysaccharide Network of Excellence (EPNOE) and European Bioplastics, 243.
Shen, T. (2012). Life Cycle Modelling of Multi-Product Lignocellulosic Ethanol Systems. University of Toronto.
Sherman, S. R., & Gorensek, M. B. (2011). Separation of Lignin From Lignocellulosic Materials. Retrieved from http://www.google.com/patents/US20110253326
Shirvani, T., Yan, X., Inderwildi, O. R., Edwards, P. P., & King, D. A. (2011). Life cycle energy and greenhouse gas analysis for algae-derived biodiesel. Energy & Environmental Science, 4(10), 3773–3778.
Sialve, B., Bernet, N., & Bernard, O. (2009). Anaerobic digestion of microalgae as a necessary step to make microalgal biodiesel sustainable. Biotechnology Advances, 27(4), 409–416. https://doi.org/10.1016/j.biotechadv.2009.03.001
Sienel, G., Rieth, R., & Rowbottom, K. (2000). Ullmann’s encyclopedia of industrial chemistry. Wiley-VCH, Weinheim.
Sills, D. L., Paramita, V., Franke, M. J., Johnson, M. C., Akabas, T. M., Greene, C. H., & Tester, J. W. (2012). Quantitative uncertainty analysis of life cycle assessment for algal biofuel production. Environmental Science & Technology, 47(2), 687–694.
Sims, R. E., Mabee, W., Saddler, J. N., & Taylor, M. (2010). An overview of second generation biofuel technologies. Bioresource Technology, 101(6), 1570–1580.
Smith, W., & Consultancy, T. (2007). Mapping the development of UK biorefinery complexes. UK National Non Food Crops Centre (NNFCC).
Smolarski, N. (2012). High-value opportunities for lignin: unlocking its potential. Frost & Sullivan, Paris, 15.
Soh, L., & Zimmerman, J. (2011). Biodiesel production: the potential of algal lipids extracted with supercritical carbon dioxide. Green Chemistry, 13(6), 1422–1429.
Spolaore, P., Joannis-Cassan, C., Duran, E., & Isambert, A. (2006). Commercial applications of microalgae. Journal of Bioscience and Bioengineering, 101(2), 87–96.
Stephens, E., Ross, I. L., King, Z., Mussgnug, J. H., Kruse, O., Posten, C., Hankamer, B. (2010). An economic and technical evaluation of microalgal biofuels. Nature Biotechnology, 28(2), 126–128.
153
Subhadra, B. G., & Edwards, M. (2011). Coproduct market analysis and water footprint of simulated commercial algal biorefineries. Applied Energy, 88(10), 3515–3523.
Sun, Y., & Cheng, J. (2002). Hydrolysis of lignocellulosic materials for ethanol production: a review. Bioresource Technology, 83(1), 1–11.
Sustainable Tarde and Consulting. (2009). Biomass Products. Retrieved May 19, 2014, from http://www.stcresources.com/full/sustainable-feedstocks/
Tabone, M. D., Cregg, J. J., Beckman, E. J., & Landis, A. E. (2010). Sustainability metrics: life cycle assessment and green design in polymers. Environmental Science & Technology, 44(21), 8264–8269.
Toledano, A., García, A., Mondragon, I., & Labidi, J. (2010). Lignin separation and fractionation by ultrafiltration. Separation and Purification Technology, 71(1), 38–43.
Tsiropoulos, I., Faaij, A., Lundquist, L., Schenker, U., Briois, J., & Patel, M. (2015). Life cycle impact assessment of bio-based plastics from sugarcane ethanol. Journal of Cleaner Production, 90, 114–127.
Tufvesson, P., Ekman, A., Sardari, R. R., Engdahl, K., & Tufvesson, L. (2013). Economic and environmental assessment of propionic acid production by fermentation using different renewable raw materials. Bioresource Technology, 149, 556–564.
Ugwu, C. U., Aoyagi, H., & Uchiyama, H. (2008). Photobioreactors for mass cultivation of algae. Bioresource Technology, 99(10), 4021–4028.
Urban, R. A., & Bakshi, B. R. (2009). 1, 3-Propanediol from fossils versus biomass: a life cycle evaluation of emissions and ecological resources. Industrial & Engineering Chemistry Research, 48(17), 8068–8082.
USDA. (n.d.). BioPreferred. Retrieved July 16, 2016, from http://www.biopreferred.gov/BioPreferred/faces/pages/AboutBioPreferred.xhtml
USDOE. (2015, August 31). Biomass Feedstock Composition and Property Database. Retrieved September 1, 2015, from http://www.afdc.energy.gov/biomass/progs/search1.cgi
USEIA. (2011). Annual Energy Review, Primary energy consumption by source and sector. Retrieved from http://www.eia.gov/totalenergy/data/monthly/pdf/flow/primary_energy.pdf
USEIA. (2015a). Biodiesel and Other Renewable Fuels Overview. Retrieved from http://www.eia.gov/totalenergy/data/monthly/pdf/sec10_8.pdf
USEIA. (2015b). Fuel Ethanol Overview. Retrieved from http://www.eia.gov/totalenergy/data/monthly/pdf/sec10_7.pdf
USEIA. (2016). Statistics for biodiesel production. Retrieved October 19, 2016, from http://www.eia.gov/biofuels/biodiesel/production/table1.pdf
154
USEPA. (2007). Regulatory Impact Analysis: Renewable Fuel Standard Program- Chapter 6: Lifecycle Impacts on Fossil Energy and Greenhouse Gases. http://www3.epa.gov/otaq/renewablefuels/420r07004chap6.pdf: Assessment and Standards Division Office of Transportation and Air Quality.
USEPA. (2008). EPA’s Report on the Environment, Indicators Presenting Data for EPA Region 5.
USEPA. (2015). Program Overview for Renewable Fuel Standard Program. http://www2.epa.gov/renewable-fuel-standard-program/program-overview-renewable-fuel-standard-program (accessed 11/24/2015).
USEPA. (2016a). Climate Change: Basic Information. Retrieved October 12, 2016, from https://www.epa.gov/climatechange/climate-change-basic-information
USEPA. (2016b). Volatile Organic Compounds’ Impact on Indoor Air Quality. Retrieved November 18, 2016, from https://www.epa.gov/indoor-air-quality-iaq/volatile-organic-compounds-impact-indoor-air-quality
(USGBC), U. G. B. C. (2006). LEED® Green Building Rating SystemTM For Core & Shell Development. US Green Building Council (USGBC).
UTEX. (2011). The Culture Collection of Algae. University of Texas. Van Dam, J. E., De Klerk-Engels, B., Struik, P. C., & Rabbinge, R. (2005). Securing
renewable resource supplies for changing market demands in a bio-based economy. Industrial Crops and Products, 21(1), 129–144.
Van Duuren, J., Brehmer, B., Mars, A., Eggink, G., Dos Santos, V., & Sanders, J. (2011). A limited LCA of bio�adipic acid: Manufacturing the nylon�6, 6 precursor adipic acid using the benzoic acid degradation pathway from different feedstocks. Biotechnology and Bioengineering, 108(6), 1298–1306.
Van Haveren, J., Oostveen, E., Micciche, F., Noordover, B., Koning, C., Van Benthem, R., Weijnen, J. (2007). Resins and additives for powder coatings and alkyd paints, based on renewable resources. Journal of Coatings Technology and Research, 4(2), 177–186.
Vennestrøm, P., Osmundsen, C. M., Christensen, C., & Taarning, E. (2011). Beyond petrochemicals: the renewable chemicals industry. Angewandte Chemie International Edition, 50(45), 10502–10509.
Vink, E. T., Rabago, K. R., Glassner, D. A., & Gruber, P. R. (2003). Applications of life cycle assessment to NatureWorksTM polylactide (PLA) production. Polymer Degradation and Stability, 80(3), 403–419.
Vishtal, A. G., & Kraslawski, A. (2011). Challenges in industrial applications of technical lignins. BioResources, 6(3), 3547–3568.
Von Blottnitz, H., & Curran, M. A. (2007). A review of assessments conducted on bio-ethanol as a transportation fuel from a net energy, greenhouse gas, and environmental life cycle perspective. Journal of Cleaner Production, 15(7), 607–619.
155
Wang, M., Huo, H., & Arora, S. (10). Methods of dealing with co-products of biofuels in life-cycle analysis and consequent results within the U.S. context. Energy Policy, 39(10), 5726–5736. https://doi.org/10.1016/j.enpol.2010.03.052
WBCSD. (2011). Greenhouse Gas Protocol, Product life cycle accounting and reporting standard. World Business Council for Sustainable Development and World Resource Institute.
Werpy, T., Petersen, G., Aden, A., Bozell, J., Holladay, J., White, J., Jones, S. (2004). Top value added chemicals from biomass. Volume 1-Results of screening for potential candidates from sugars and synthesis gas. DTIC Document.
Williams, R. S. (1999). Finishing of wood. Wood Handbook : Wood as an Engineering Material. Madison, WI : USDA Forest Service, Forest Products Laboratory, 1999. General Technical Report FPL ; GTR-113: Pages 15.1-15.37.
Williamson, M. A. (2010). US biobased products market potential and projections through 2025. Nova Science Publishers.
Wooley, R. J., & Putsche, V. (1996). Development of an ASPEN PLUS physical property database for biofuels components. National Renewable Energy Laboratory Golden, CO.
XIVIA. (2010). XIVIATM White Paper, Sustainable and substantiated. Retrieved from http://www.feingold.org/Research/PDFstudies/Xylitolpaper.pdf
Yang, J., Xu, M., Zhang, X., Hu, Q., Sommerfeld, M., & Chen, Y. (2011). Life-cycle analysis on biodiesel production from microalgae: water footprint and nutrients balance. Bioresource Technology, 102(1), 159–165.
Yu, J., & Chen, L. X. (2008). The greenhouse gas emissions and fossil energy requirement of bioplastics from cradle to gate of a biomass refinery. Environmental Science & Technology, 42(18), 6961–6966.
Zaimes, G. G., & Khanna, V. (2014). The role of allocation and coproducts in environmental evaluation of microalgal biofuels: How important? Sustainable Energy Technologies and Assessments, 7, 247–256.
Zaimes, G. G., Soratana, K., Harden, C. L., Landis, A. E., & Khanna, V. (2015). Biofuels via fast pyrolysis of perennial grasses: A life cycle evaluation of energy consumption and greenhouse gas emissions. Environmental Science & Technology, 49(16), 10007–10018.
Zakzeski, J., Bruijnincx, P. C., Jongerius, A. L., & Weckhuysen, B. M. (2010). The catalytic valorization of lignin for the production of renewable chemicals. Chemical Reviews, 110(6), 3552–3599.
Zhang, Y., Hu, G., & Brown, R. C. (2014). Life cycle assessment of commodity chemical production from forest residue via fast pyrolysis. The International Journal of Life Cycle Assessment, 19(7), 1371–1381.
156
Zhang, Y.-H. P. (2008). Reviving the carbohydrate economy via multi-product lignocellulose biorefineries. Journal of Industrial Microbiology & Biotechnology, 35(5), 367–375.
Zhao, X., Cheng, K., & Liu, D. (2009). Organosolv pretreatment of lignocellulosic biomass for enzymatic hydrolysis. Applied Microbiology and Biotechnology, 82(5), 815–827.
Zhu, J., & Pan, X. (2010). Woody biomass pretreatment for cellulosic ethanol production: technology and energy consumption evaluation. Bioresource Technology, 101(13), 4992–5002.
157
APPENDIX A Meta-Analysis of Life Cycle Energy and Greenhouse Gas Emissions for
Priority Bio-based Chemicals
These data have been sent to the journal of ACS Green Chemistry and Engineering
Method: Table A1- Life cycle GHG emission and energy use values for studied cases (including carbon sequestration)
Chemical Feedstock GHG (kg CO2 eq./kg)
Energy Use (MJ/kg)
Note
Succinic acid Corn Corn stover Lignocellulose sugarcane
-0.18 0.88 0.83-3.13 0.8-3.1 0.43 (-0.16)-2.13 (-0.2)-2.1 0.2-2.5
CED: 34.7 NREU: 32.7 NREU: 27-67 NREU: 28-66.5 Fossil fuel input: 28 NREU: 5-45 NREU: 5.4-44.9 NREU: 15-54.5
4 cases for corn grain, 1 case for corn stover, 2 cases for sugarcane and 1 for lignocellulosic source
Polyethylene furan dicarboxylate (PEF)
Corn starch 2.05 NREU: 33.8 Another case was also found for PEF from corn starch with just relative GHG reduction values
Propionic acid
Rapeseed meal Potato juice Sugarbeet Potato molasses artichoke
1.35 3.2 3.6 3.1 3.8
CED: 39 - - - -
Life cycle energy use was found for the first case from rapeseed meal
Itaconic acid Softwood Corn
-0.36 0.75
CED: 15 CED: 24.8
158
Polyjydroxy butyric acid (PHB)
Eucalyptus Poplar Sugarcane Corn
2.44 2.08 2.6 (-4)-(-1.6)
- - CED: 44.7 Fossil fuel input: (-6)-8
Life cycle energy use were reported for the last two cases
Xylitol Corn stover Corn cobs Pulp and paper stream
-0.93 37.16 2.15
Fossil fuel input: 3 - -
Another case was found for pulp and paper stream which had the relative GHG reduction
Phenol Poplar Eucalyptus
3.38 4.19
- -
Relative energy reductions were reported
Methanol Waste wood 0.08 - Relative energy reduction was reported
Vanillin Wood Timber and wood chips
1.6 -0.09
NREU: 44.1 CED: 36.5
Styrene Forest residue -0.002 Fossil fuel input: 0.14
Adipic acid Corn Lignin-based phenol Lignocellulosics sugarcane
0.20 9.2 9.2 -1.29 6 5.6 3.8
- NREU: 195 NREU: 44.5-195.4 CED: 35.5 NREU: 21.5-134.4 NREU: 86 NREU: 3.2-85.7
3 cases for corn grain, 1 case for lignin-based phenol, 1 case for lignocellulosics and 2 cases for sugarcane
Polylactic acid (PLA)
Corn Poplar Eucalyptus Sugarcane lignocellulosics
1.8 0.4-2.4 1.16-2.36 3.16 3.77 (-0.9)-1 (-0.13)-0.96 0.5 (-0.4)-1.5
Fossil fuel input:54 NREU: 40.1-60.8 NREU: 49-61 - - NREU: 13.2-32.9 NREU: 21-33 NREU: 30.45 NREU: 25.1-45.3
3 cases for corn, 1 case for poplar, 1 case for eucalyptus, 3 cases for sugarcane and 1 case for lignocellulosics
159
Polyhydroxy alkanoate (PHA)
Corn Corn stover Sugarcane lignocellulosics
0.48-4.48 (-0.22)-4.27 (-0.7)-6.9 0.25-0.5 (-3.7)-6.9 (-2.5)-6.9
NREU: 59.17-88 NREU: 38-112 NREU: 33.3- 111.6 NREU: 44-60 NREU: (-23.5)-109 NREU: 3.4-111.5
3 cases for corn, 1 case for corn stover, 1 case for sugarcane and 1 case for lignocellulosics
1,3-butadiene Corn Wheat/rye/sugar beet sugarcane
2.3-4 1.04-2.18 2.04-3.62
NREU: 30-40 NREU: 90-115 NREU: 60-85
Ethyl lactate Corn 0.11-0.75 - Energy use was not reported
p-Xylene Corn Red oak
5.5-9.86 1.13-2.49
- -
Energy use was not reported
Low density polyethylene (LDPE)
Corn Switchgrass Sugarcane
2.6 -2.9 -1.3 0.3-2.6
- - - CED: 102
1 case for corn grain, 1 case for switchgrass and two cases for sugarcane
Polyethylene (PE)
Corn stover -0.75 Fossil fuel input: 32
High density polyethylene (HDPE)
Sugarcane (1.5)-(-0.3) NREU: 18
1,3-propanediol (PDO)
Corn Algae Sugarcane lignocellulosics
2.7 0.57-1.17 0.5-1.8 6.67 (-1.5)-(-0.5) (-1.7)-1.8 (-0.8)-1.8
Fossil fuel input: 43 NREU: 38-52 NREU: 19.8-91.5 Fossil fuel input: 120 NREU: (-9)-7 NREU: (-16.5)-63.5 NREU: (-16.5)-63.5
3 cases for corn grain, 1 case for algae (hydroxypropioni c acid) , 2 cases for sugarcane an d1 case for lignocellulosics
1,4-butanediol
Corn stover 1.05 Fossil fuel input: 45
Acrylic acid Algae 2.26 Fossil fuel input: 49 From hydroxypropionic acid derived from algae
iso-Butanol Corn stover 0.31 Fossil fuel input: 38
160
n-Butanol Corn Sugarcane Lignocellulosics
0.7-1.1 0.61-1.11 (-2.1)-(-1.7) (-2.18)-(-1.68) (-0.5)-(-1)
NREU: 6.6-63.9 NREU: 27-67 NREU: (-40.9)- 7.4 NREU: 27-67 NREU: (-19.8)- 32.5
2 cases from corn, 2 cases from sugarcane and 1 case from lignocellulosics
Acetic acid Corn Sugarcane Lignocellulosics
4.23-6.63 4.2-6 2.33-4.72 2.3-4.7 3.1-5
NREU: 109-145 NREU: 38.9-144.9 NREU: 71-106 NREU: 17.6-106.3 NREU:27-123.4
2 cases for corn grain, 2 cases for sugarcane and 1 case for lignocellulosics
Table A2- Bio-based chemicals and their fossil-based counterpart- sourced from ecoinvent unit processes
Bio-based chemical Fossil-based equivalent Polyethylene furandicarboxylate (PEF) Polyethylene terephthalate resin, at plant/kg NREL/RNA Polyhydroxy alkanoate (PHA) High density polyethylene resin, at plant/NREL/RNA p-Xylene p-xylene, production, at plant/RER Styrene Styrene, at plant/ RER Table A3-Bio-based chemicals and their fossil-based counterparts- sourced from literature
Bio-based chemical Fossil-based equivalents Succinic acid Maleic anhydride, succinic acid Polyethylene furandicarboxylate (PEF) PET resin Propionic acid Propionic acid Itaconic acid Polyacrylic acid Polyhydroxy butyric acid (PHB) Polyethylene terephthalate (PET), low density polyethylene
(LDPE) Xylitol* - Phenol Phenol (from Cumene) Methanol Methanol Vanillin Formaldehyde resin Styrene Styrene Adipic acid Adipic acid Polylactic acid (PLA) Propylene resin (PP), polyethylene terephthalate (PET),
polystyrene (PS) Polyhydroxy alkanoate (PHA) High density polyethylene (HDPE), polystyrene (PS) 1,3-Butadiene 1,3-Butadiene Ethyl lactate Polytrimethylene terephthalate (PTT) p-Xylene p-Xylene
161
Low density polyethylene (LDPE) Low density polyethylene (LDPE) Polyethylene Polyethylene High density polyethylene (HDPE) High density polyethylene (HDPE) 1,3-Propanediol (1,3-PDO) 1,3-Propanediol (1,3-PDO) 1,4-butanediol 1,4-butanediol Acrylic acid Acrylic acid iso-Butanol iso-Butanol n-Butanol n-Butanol, maleic anhydride Acetic acid Acetic acid * No data were found for the fossil-based equivalent of xylitol Note: For each building block, several counterparts are considered sourcing from the case studies, second column in this table list all the counterparts considered for various cases
Table A4- Molecular complexities for select bio-based compounds
CAS # Biochemical Molecular Complexity
Note
106-99-0 1,3- Butadiene 21 504-63-2 1,3- Propandiol 12.4 110-63-4 1,4- Butanediol 17.5 503-66-2 3-Hydroxypropionic acid 50 64-19-7 Acetic Acid 31 79-10-7 Acrylic acid 55.9 124-04-9 Adipic acid 114 7643-75-6 Arabinitol 76.1 *L-arabinitol 617-45-8 Aspartic acid 133 *L-aspartic acid 92-52-4 Biphenyl 100 71-36-3 Butanol 13.1 N/A Cresol/Resorcinol N/A 110-82-7 Cyclohexane 15.5 97-64-3 Ethyl Lactate 79.7 56-81-5 Glycerol 25.2 78-83-1 Iso-Butanol 17.6 97-65-4 Itaconic acid 158 N/A Low-density polyethylene (LDPE) N/A 67-56-1 Methanol 2 872-50-4 N-methylpyrrolidone 90.1 106-42-3 p-Xylene 48.8 108-95-2 Phenol 46.1 N/A Polyethylene N/A N/A Polyethylene (HDPE) N/A N/A Polyethylene furandicarboxylate (PEF) N/A N/A Polyhydroxyalkanoate (PHA) N/A N/A Polyhydroxybutyric acid (PHB) N/A N/A Polylactic acid (PLA) N/A 504-63-2 Propandiol (PDO) 12.4 *1,3-
Propanediol
162
504-63-2 Propanediol 12.4 *1,3-Propanediol
79-09-4 Propionic acid 40.2 57-55-6 Propylene glycol 20.9 50-70-4 Sorbitol 105 *D-Sorbitol 100-42-5 Styrene 68.1 110-15-6 Succinic acid 92.6 121-34-6 Vanilic acid 168 121-33-5 Vanillin 135 87-99-0 Xylitol 76.1 *Proxy Compound Values for molecular complexity were obtained from PubChem
163
Results:
Figure A1- Trend of LCA publications under review
Table A5- Grouping information using the Tukey method and 90% confidence for factor, Conversion Platform for response factor absolute greenhouse gas emissions
Factor Level N Mean Grouping Thermochemical 7 6.68 A Biochemical 57 2.024 B Chemical 10 1.907 A B Hybrid 9 0.903 B Catalytic 1 0.2032 A B
Means that do not share a letter are significantly different Table A6- Tukey simultaneous tests for differences of means, 90% confidence for factor, Conversion Platform for response factor absolute greenhouse gas emissions
Difference of Levels Difference of Means
SE of Difference
90% CI T-Value Adjusted P-Value
Catalytic - Biochemical -1.82 4.42 (-12.88, 9.24) -0.41 0.994 Chemical - Biochemical -0.12 1.5 (-3.88, 3.64) -0.08 1 Hybrid - Biochemical -1.12 1.57 (-5.05, 2.81) -0.71 0.953 Thermochemical - Biochemical 4.66 1.75 (0.27, 9.05) 2.66 0.07 Chemical - Catalytic 1.7 4.59 (-9.80, 13.20) 0.37 0.996 Hybrid - Catalytic 0.7 4.62 (-10.86, 12.26) 0.15 1 Thermochemical - Catalytic 6.48 4.68 (-5.24, 18.20) 1.38 0.64 Hybrid - Chemical -1 2.01 (-6.04, 4.03) -0.5 0.987
0
1
2
3
4
5
6
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Number of LCA studies
Year
164
Thermochemical - Chemical 4.78 2.16 (-0.63, 10.18) 2.21 0.186 Thermochemical - Hybrid 5.78 2.21 (0.25, 11.31) 2.62 0.077
Table A7- Analysis of Covariance (ANCOVA) for the response variable of absolute greenhouse gas emissions and covariate of complexity
Source DF Adj. SS Adj. MS F-Value P-Value Complexity 1 11.01 11.01 0.41 0.525 Error 58 1557.92 26.86 Lack-of-Fit 18 481.48 26.75 0.99 0.486 Pure Error 40 1076.44 26.91 Total 59 1568.93 Table A8- Analysis of Covariance (ANCOVA) for the response variable of relative greenhouse gas emissions and covariate of complexity
Source DF Adj. SS Adj. MS F-Value P-Value Complexity 1 0.05 0.04996 0.07 0.788 Error 55 37.7326 0.68605 Lack-of-Fit 17 23.0343 1.35496 3.5 0.001 Pure Error 38 14.6982 0.3868 Total 56 37.7825 Table A9- Analysis of Covariance (ANCOVA) for response variable of absolute greenhouse gas emissions and covariate of molecular weight
Source DF Adj. SS Adj. MS F-Value P-Value Molecular Weight 1 69.88 69.88 2.7 0.106 Error 58 1499.05 25.85 Lack-of-Fit 16 247.95 15.5 0.52 0.921 Pure Error 42 1251.1 29.79 Total 59 1568.93 Table A10- Analysis of Covariance (ANCOVA) for response variable of relative greenhouse gas emissions and covariate of molecular weight
Source DF Adj. SS Adj. MS F-Value P-Value Molecular Weight 1 0.0089 0.00885 0.01 0.91 Error 55 37.7737 0.68679 Lack-of-Fit 15 23.0469 1.53646 4.17 0 Pure Error 40 14.7268 0.36817 Total 56 37.7825
165
Table A11- 1-way Analysis of Variance (ANOVA) for response variable of absolute greenhouse gas emissions- 1st set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Feedstock 12 117.9 9.822 0.46 0.933 Error 73 1570.3 21.511 Total 85 1688.2 Factor: Feedstock; Levels: 13; Values: Algae, Artichoke, Corn, Lignocellulose, Mixed, Potato, Rapeseed, Residue, Sugarbeet, Sugarcane, Switchgrass, Waste, and Woody Biomass Table A12- 1-way Analysis of Variance (ANOVA) for response variable of relative greenhouse gas emissions- 1st set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Feedstock 12 10.09 0.8406 1.4 0.184 Error 71 42.5 0.5986 Total 83 52.59 Factor: Feedstock; Levels: 13; Values: Algae, Artichoke, Corn, Lignocellulose, Mixed, Phenol, Potato, Rapeseed, Residue, Sugarbeet, Sugarcane, Switchgrass, Waste, and Woody Biomass
Table A13- 1-way Analysis of Variance (ANOVA) for the response variable of absolute greenhouse gas emissions- 2nd set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Composition 1 9.2 9.204 0.46 0.499 Error 84 1678.99 19.988 Total 85 1688.2 Factor: Building Blocks; Levels: 2; Values: Sugar, Lignin
Table A14- 1-way Analysis of Variance (ANOVA) for the response variable of relative greenhouse gas emissions- 2nd set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Composition 1 0.4264 0.4264 0.67 0.415 Error 82 52.1613 0.6361 Total 83 52.5877 Factor: Building Blocks; Levels: 2; Values: Sugar, Lignin
166
Table A15-1-way Analysis of Variance (ANOVA) for the response variable of absolute greenhouse gas emissions- 3rd set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Conversion Platform 4 162.1 40.54 2.11 0.087 Error 79 1516 19.19 Total 83 1678.1 Factor: Conversion; Levels: 5; Values: Biochemical, Catalytic, Chemical, Hybrid, Thermochemical Table A16- 1-way Analysis of Variance (ANOVA) for the response variable of relative greenhouse gas emissions- 3rd set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Conversion Platform 4 1.17 0.2925 0.45 0.77 Error 78 50.34 0.6454 Total 82 51.51 Factor: Conversion; Levels: 5; Values: Biochemical, Catalytic, Chemical, Hybrid, Thermochemical Table A17- 1-way Analysis of Variance (ANOVA) for the response variable of absolute greenhouse gas emissions- 4th set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Geography 4 28.4 7.1 1.41 0.242 Error 54 271.2 5.022 Total 58 299.6 Factor: Geography; Levels: 5; Values: Thailand, Europe, USA, Canada, and Brazil Table A18- 1-way Analysis of Variance (ANOVA) for the response variable of relative greenhouse gas emissions- 4th set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Geography 4 0.5438 0.1359 0.17 0.954 Error 55 44.6105 0.8111 Total 59 45.1543 Factor: Geography; Levels: 5; Values: Thailand, Europe, USA, Canada, and Brazil Table A19- 1-way Analysis of Variance (ANOVA) for the response variable of absolute greenhouse gas emissions- 5th set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value LCA Coproduct Handling Method 3 16 5.333 0.91 0.439 Error 65 378.91 5.829 Total 68 394.9 Factor: LCA Coproduct Handling Method; Levels: 4; Values: Economic, Hybrid, Mass, System Boundary Expansion
167
Table A20- 1-way Analysis of Variance (ANOVA) for the response variable of relative greenhouse gas emissions- 5th set of parameters Source DF Adj. SS Adj. MS F-Value P-Value LCA Coproduct Handling Method 3 0.9539 0.318 0.42 0.742 Error 63 48.1231 0.7639 Total 66 49.077 Factor: LCA Coproduct Handling Method; Levels: 4; Values: Economic, Hybrid, Mass, System Boundary Expansion Table A21- 1-way Analysis of Variance (ANOVA) for the response variable of absolute greenhouse gas emissions- 6th set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Land Use Change 2 27.07 13.53 0.68 0.511 Error 83 1661.13 20.01 Total 85 1688.2 Factor: Land Use Change; Levels: 3; Values: No LUC, dLUC, and dLUC & ILUC Land Use Change (LUC); Direct Land Use Change (dLUC); Indirect Land Use Change (ILUC) Table A22- 1-way Analysis of Variance (ANOVA) for the response variable of relative greenhouse gas emissions- 6th set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Land Use Change 2 1.656 0.8282 1.32 0.274 Error 81 50.931 0.6288 Total 83 52.588 Factor: Land Use Change; Levels: 3; Values: No LUC, dLUC, and dLUC & ILUC Land Use Change (LUC); Direct Land Use Change (dLUC); Indirect Land Use Change (ILUC) Table A23- ANCOVA and ANOVA summary results for bio-based chemicals greenhouse gas emissions meta-data
Parameter Covariate or Factor
Factor Levels Response Factor P-value
Statistically Significant (α = 10%)
Complexity Covariate - GHG Absolute 0.525 No Complexity Covariate - GHG Relative 0.788 No Molecular Weight Covariate - GHG Absolute 0.106 No Molecular Weight Covariate - GHG Relative 0.91 No Feedstock Factor 13 GHG Absolute 0.933 No Feedstock Factor 13 GHG Relative 0.184 No Composition Factor 2 GHG Absolute 0.499 No Composition Factor 2 GHG Relative 0.415 No Conversion Platform Factor 5 GHG Absolute 0.087 Yes Conversion Platform Factor 5 GHG Relative 0.77 No Geography Factor 5 GHG Absolute 0.242 No
168
Geography Factor 5 GHG Relative 0.954 No LCA Coproduct Handling Method
Factor 4 GHG Absolute 0.439 No
LCA Coproduct Handling Method
Factor 4 GHG Relative 0.742 No
Land Use Change Factor 3 GHG Absolute 0.511 No Land Use Change Factor 3 GHG Relative 0.274 No
Table A24- Grouping information using the Tukey method and 90% confidence for factor, Conversion Platform for response factor absolute nonrenewable energy use
Factor Level N Mean Grouping Thermochemical 1 2.600 A Hybrid 4 -0.131 B Biochemical 38 -0.2976 B Chemical 1 -0.5000 B
Means that do not share a letter are significantly different Table A25- Tukey simultaneous tests for differences of means, 90% confidence for factor, Conversion Platform for response factor absolute nonrenewable energy use
Difference of Levels Difference of Means
SE of Difference
90% CI T-Value Adjusted P-Value
Chemical - Biochemical -0.202 0.555 (-1.517, 1.112) -0.36 0.983 Hybrid - Biochemical 0.166 0.288 (-0.516, 0.848) 0.58 0.938 Thermochemical - Biochemical 2.898 0.555 (1.583, 4.212) 5.22 0 Hybrid - Chemical 0.369 0.612 (-1.082, 1.819) 0.6 0.931 Thermochemical - Chemical 3.1 0.774 (1.266, 4.934) 4 0.001 Thermochemical - Hybrid 2.731 0.612 (1.281, 4.182) 4.46 0
Table A26- Grouping information using the Tukey method and 90% confidence for factor, LCA Coproduct Handling Method for response factor absolute nonrenewable energy use
Factor Level N Mean Grouping Mass 5 0.493 A Hybrid 36 -0.2833 B Economic 1 -0.7300 AB System Expansion 1 -0.7700 AB
Means that do not share a letter are significantly different
169
Table A27- Tukey simultaneous tests for differences of means, 90% confidence for factor, LCA Coproduct Handling Method for response factor absolute nonrenewable energy use
Difference of Levels Difference of Means
SE of Difference
90% CI T-Value Adjusted P-Value
Hybrid - Economic 0.447 0.669 (-1.137, 2.030) 0.67 0.908 Mass – Economic 1.223 0.722 (-0.488, 2.934) 1.69 0.341 System Expansion - Economic -0.04 0.933 (-2.249, 2.169) -0.04 1 Mass - Hybrid 0.776 0.315 (0.031, 1.522) 2.47 0.081 System Expansion - Hybrid -0.487 0.669 (-2.070, 1.097) -0.73 0.885 System Expansion - Mass -1.263 0.722 (-2.974, 0.448) -1.75 0.313
Table A28- Grouping information using the Tukey method and 90% confidence for factor, Land Use Change for response factor absolute nonrenewable energy use
Factor Level N Mean Grouping dLUC & ILUC 2 0.92 A No LUC 39 -0.2366 B dLUC 4 -0.6112 B
Means that do not share a letter are significantly different Table A29- Tukey simultaneous tests for differences of means, 90% confidence for factor, Land Use Change for response factor absolute nonrenewable energy use
Difference of Levels Difference of Means
SE of Difference
90% CI T-Value Adjusted P-Value
dLUC – No LUC -0.375 0.335 (-1.081, 0.331) -1.12 0.508 dLUC & ILUC – No LUC 1.152 0.463 (0.177, 2.126) 2.49 0.044 dLUC & ILUC – dLUC 1.526 0.553 (0.362, 2.691) 2.76 0.023
Table A30- Analysis of Covariance (ANCOVA) for the response variable of absolute nonrenewable energy use and covariate of complexity
Source DF Adj. SS Adj. MS F-Value P-Value Complexity 1 4426 4426 2.57 0.12 Error 28 48175 1721 Lack-of-Fit 5 24825 4965 4.89 0.003 Pure Error 23 23350 1015 Total 29 52601
170
Table A31- Analysis of Covariance (ANCOVA) for the response variable of relative absolute nonrenewable energy use and covariate of complexity
Source DF Adj. SS Adj. MS F-Value P-Value Complexity 1 0.0096 0.00963 0.03 0.874 Error 28 10.4454 0.37305 Lack-of-Fit 5 6.3124 1.26248 7.03 0 Pure Error 23 4.133 0.1797 Total 29 10.455
Table A32- Analysis of Covariance (ANCOVA) for response variable of absolute nonrenewable energy use and covariate of molecular weight
Source DF Adj. SS Adj. MS F-Value P-Value Molecular Weight 1 1557 1557 0.85 0.363 Error 28 51043 1823 Lack-of-Fit 5 27693 5539 5.46 0.002 Pure Error 23 23350 1015 Total 29 52601
Table A33- Analysis of Covariance (ANCOVA) for response variable of relative nonrenewable energy use and covariate of molecular weight
Source DF Adj. SS Adj. MS F-Value P-Value Molecular Weight 1 0.4714 0.4714 1.32 0.26 Error 28 9.9836 0.3566 Lack-of-Fit 5 5.8506 1.1701 6.51 0.001 Pure Error 23 4.133 0.1797 Total 29 10.455
Table A34-1-way Analysis of Variance (ANOVA) for response variable of absolute nonrenewable energy use - 1st set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Feedstock 5 9823 1965 1.49 0.214 Error 39 51268 1315 Total 44 61091
Factor: Feedstock; Levels: 13; Values: Algae, Artichoke, Corn, Lignocellulose, Mixed, Potato, Rapeseed, Residue, Sugarbeet, Sugarcane, Switchgrass, Waste, and Woody Biomass
171
Table A35- 1-way Analysis of Variance (ANOVA) for response variable of relative nonrenewable energy use - 1st set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Feedstock 5 2.543 0.5087 1.12 0.367 Error 39 17.755 0.4553 Total 44 20.299
Factor: Feedstock; Levels: 13; Values: Algae, Artichoke, Corn, Lignocellulose, Mixed, Phenol, Potato, Rapeseed, Residue, Sugarbeet, Sugarcane, Switchgrass, Waste, and Woody Biomass
Table A36- 1-way Analysis of Variance (ANOVA) for the response variable of absolute nonrenewable energy use - 2nd set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Composition 1 66.2 66.22 0.05 0.83 Error 43 61024.3 1419.17 Total 44 61090.6
Factor: Building Blocks; Levels: 2; Values: Sugar, Lignin
Table A37-1-way Analysis of Variance (ANOVA) for the response variable of relative nonrenewable energy use - 2nd set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Composition 1 0.0809 0.08094 0.17 0.68 Error 43 20.2178 0.47018 Total 44 20.2988
Factor: Building Blocks; Levels: 2; Values: Sugar, Lignin Table A38- 1-way Analysis of Variance (ANOVA) for the response variable of absolute nonrenewable energy use - 3rd set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Conversion Platform 3 490.6 163.5 0.11 0.954 Error 40 60129.9 1503.2 Total 43 60620.5
Factor: Conversion; Levels: 5; Values: Biochemical, Catalytic, Chemical, Hybrid, Thermochemical
172
Table A39-1-way Analysis of Variance (ANOVA) for the response variable of relative nonrenewable energy use - 3rd set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Conversion Platform 3 8.291 2.7637 9.22 0 Error 40 11.995 0.2999 Total 43 20.286
Factor: Conversion; Levels: 5; Values: Biochemical, Catalytic, Chemical, Hybrid, Thermochemical Table A40- 1-way Analysis of Variance (ANOVA) for the response variable of absolute nonrenewable energy use - 4th set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Geography 4 2079 519.6 0.57 0.689 Error 26 23837 916.8 Total 30 25916
Factor: Geography; Levels: 5; Values: Thailand, Europe, USA, Canada, and Brazil Table A41-1-way Analysis of Variance (ANOVA) for the response variable of relative nonrenewable energy use - 4th set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Geography 4 0.8092 0.2023 0.4 0.809 Error 26 13.241 0.5093 Total 30 14.0502
Factor: Geography; Levels: 5; Values: Thailand, Europe, USA, Canada, and Brazil Table A42- 1-way Analysis of Variance (ANOVA) for the response variable of absolute nonrenewable energy use - 5th set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value LCA Coproduct Handling Method 3 1787 595.6 0.4 0.757 Error 39 58775 1507.1 Total 42 60562 Factor: LCA Coproduct Handling Method; Levels: 4; Values: Economic, Hybrid, Mass, System Boundary Expansion
173
Table A43-1-way Analysis of Variance (ANOVA) for the response variable of relative nonrenewable energy use - 5th set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value LCA Coproduct Handling Method 3 3.247 1.0825 2.49 0.075 Error 39 16.959 0.4348 Total 42 20.206 Factor: LCA Coproduct Handling Method; Levels: 4; Values: Economic, Hybrid, Mass, System Boundary Expansion Table A44- 1-way Analysis of Variance (ANOVA) for the response variable of absolute greenhouse gas emissions- 6th set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Land Use Change 2 1539 769.5 0.54 0.585 Error 42 59552 1417.9 Total 44 61091 Factor: Land Use Change; Levels: 3; Values: No LUC, dLUC, and dLUC & ILUC Land Use Change (LUC); Direct Land Use Change (dLUC); Indirect Land Use Change (ILUC) Table A45-1-way Analysis of Variance (ANOVA) for the response variable of relative nonrenewable energy use - 6th set of parameters
Source DF Adj. SS Adj. MS F-Value P-Value Land Use Change 2 3.199 1.5997 3.93 0.027 Error 42 17.099 0.4071 Total 44 20.299 Factor: Land Use Change; Levels: 3; Values: No LUC, dLUC, and dLUC & ILUC Land Use Change (LUC); Direct Land Use Change (dLUC); Indirect Land Use Change (ILUC) Table A46-ANCOVA and ANOVA summary results for bio-based chemicals nonrenewable energy use meta-data
Parameter Covariate or Factor
Factor Levels Response Factor P-value
Statistically Significant (α = 10%)
Complexity Covariate - NREU Absolute 0.12 No Complexity Covariate - NREU Relative 0.874 No Molecular Weight Covariate - NREU Absolute 0.363 No Molecular Weight Covariate - NREU Relative 0.26 No Feedstock Factor 13 NREU Absolute 0.214 No Feedstock Factor 13 NREU Relative 0.367 No Composition Factor 2 NREU Absolute 0.83 No Composition Factor 2 NREU Relative 0.68 No Conversion Platform Factor 4 NREU Absolute 0.954 No Conversion Platform Factor 4 NREU Relative 0 Yes Geography Factor 5 NREU Absolute 0.689 No
174
Geography Factor 5 NREU Relative 0.809 No LCA Coproduct Handling Method
Factor 4 NREU Absolute 0.757 No
LCA Coproduct Handling Method
Factor 4 NREU Relative 0.075 Yes
Land Use Change Factor 3 NREU Absolute 0.585 No Land Use Change Factor 3 NREU Relative 0.027 Yes
Table A47-1-way Analysis of Variance (ANOVA) for the response variable of Greenhouse Gas Emissions (Absolute)
Source DF Adj. SS Adj. MS F-Value P-Value Plant Capacity 2 157.5 78.77 3.99 0.022 Error 77 1520.6 19.75 Total 79 1678.1
Factor: Plant Capacity; Levels: 3; Values: Commercial Scale, Pilot Scale, Lab Scale
Table A48-Grouping Information Using the Tukey Method and 90% Confidence for Factor Plant Capacity with response factor GHG Absolute
Factor Level N Mean Grouping Pilot Scale 5 7.44 A Lab Scale 11 2.862 A B Commercial Scale 64 1.705 B
Means that do not share a letter are significantly different
Table A49-Tukey Simultaneous Tests for Differences of Means, 90% Confidence for Factor Plant Capacity with response factor GHG Absolute
Difference of Levels Difference of Means
SE of Difference
90% CI T-Value Adjusted P-Value
Lab Scale - Commercial Scale
1.16 1.45 (-1.87, 4.18) 0.80 0.706
Pilot Scale - Commercial Scale
5.73 2.06 (1.43, 10.04) 2.78 0.019
Pilot Scale - Lab Scale 4.57 2.40 (-0.42, 9.57) 1.91 0.143
175
Table A50- 1 Way Analysis of Variance (ANOVA) for the Response Variable: Greenhouse Gas Emissions (Relative)
Source DF Adj. SS Adj. MS F-Value P-Value Plant Capacity 2 3.139 1.5695 2.43 0.095 Error 74 47.821 0.6462 Total 76 50.960
Factor: Plant Capacity; Levels: 3; Values: Commercial Scale, Pilot Scale, Lab Scale
Table A51-Grouping Information Using the Tukey Method and 90% Confidence for Factor Plant Capacity with response factor GHG Relative
Factor Level N Mean Grouping Lab Scale 11 -0.084 A Commercial Scale 64 -0.5992 A Pilot Scale 2 -1.098 A
Means that do not share a letter are significantly different
Table A52-Tukey Simultaneous Tests for Differences of Means, 90% Confidence for Factor Plant Capacity with response factor GHG Relative
Difference of Levels Difference of Means
SE of Difference
90% CI T-Value Adjusted P-Value
Lab Scale - Commercial Scale 0.516 0.262 (-0.032, 1.063) 1.96 0.128 Pilot Scale - Commercial Scale
-0.499 0.577 (-1.703, 0.705) -0.87 0.664
Pilot Scale - Lab Scale -1.015 0.618 (-2.304, 0.274) -1.64 0.235
176
Table A53- CO2 emissions from EOL phase of bio-based and fossil-based chemicals
Bio-based chemical/ Fossil-based counterpart
EOL CO2 emissions (kg CO2/ kg)
Comparative EOL results (bio-based relative to fossil-based)
Succinic acid Adipic acid
1.49 1.80
-17%
Succinic acid Maleic anhydride
1.49 1.79
-17%
PEF PET
1.92 3.14
-39%
Itaconic acid Polyacrylic acid
1.69 1.83
-7%
PHB PET
1.69 2.50
-32%
PHB LDPE
1.69 3.13
-46%
PLA PET
1.46 3.14
-53%
PLA PS
1.46 3.37
-56%
PLA PP
1.46 3.13
-53%
PHA HDPE
2.04 3.15
-35%
PHA PS
2.04 3.37
-38%
Ethyl lactate PTT
1.86 2.34
-20%
177
Table A54-Cradle-to-grave GHG emissions for bio-based chemicals relative to petrochemical counterparts
Bio-based chemical Cradle-to-grave GHG emissions (% change from petrochemical counterpart) Low value Average High value
Succinic acid -92% -70% -46% PEF - -47% - Propionic acid -40% -22% -11% Itaconic acid -71% -59% -46% PHB -138% -61% -31% Phenol -28% -23% -19% Methanol - -15% - Vanillin -54% -37% -19% Adipic acid -89% 12% 90% PLA -91% -57% -35% PHA -135% -26% 89% 1,3-Butadiene -20% 7% 35% Ethyl lactate -65% -59% -54% LDPE -95% -44% 16% PE - -114% - HDPE -68% -56% -45% Propanediol -99% -62% -42% 1,4-Butanediol - -64% - i-Butanol - -63% - n-Butanol -97% -64% -19% Acetic acid -36% 9% 70% p-Xylene 34% 143% 298% Acrylic acid -83% -61% -61%
* Note: For some of the bio-based chemicals mentioned in this table, cradle to grave GHG results are available from the reference studies and may be different from what reported here. For the values mentioned in this table, the main assumption is that, the carbon content in the composition of each building block is going to be released as CO2 during EOL scenarios.
178
APPENDIX B
Life Cycle Assessment of Catechols from Lignin Depolymerization
These data have been sent to ACS Sustainable Chemistry and Engineering
Method:
Life Cycle Inventories
Bio-based Route Inventory.
Table B1 lists all inventory inputs for production of 1 kg tert-butyl catechol from bio-based
resources. Input parameters were based on experimental data(Barta et al., 2014) scaled for an
industrial plant operating with a capacity of 75 ton lignin/day. ASPEN plus simulation output is
54 tons of TBC/day. Material inputs are linearly extrapolated from the experimental data source
(Barta et al., 2014) and scaled as shown below for four different inputs of main steps:
Cultivation:
Cutivatedamount nuts nutshells
1g nuts nutshells0.7gnutshells
∗110gnutshells
13.62gcrudelignin∗
8.39gcrudelignin8.077gpurifiedlignin
∗ 7.29gpurifiedlignin
4.98gEt. Ac. solublelignin∗ 1ton nuts nutshells10 g nuts nutshells
∗10 gEt. Ac. solublelignin1tonEt. Ac. solublelignin
∗ 75tonEt. Ac. solublelignin
day∗
day53.5tonTBC
24.6kg nuts nutshells
kgTBC
179
Organosolv Extraction:
Methanolmake upflow
500mlMeOH
13.62gcrudelignin∗0.79gMeOH1mlMeOH
∗3gMeOHused100gMeOH
∗8.39gcrudelignin
8.077gpurifiedlignin∗
7.29gpurifiedlignin4.98gEt. Ac. solublelignin
∗10 gEt. Ac. solublelignin1tonEt. Ac. solublelignin
∗1 10
∗75tonEt. Ac. solubelignin
day
∗ day
53.5tonTBC1.85
ton/kgMeOHton/kgTBC
Energy consumption for this process was sourced from ASPEN Plus simulations conducted for
several extraction methods on softwood lignin as reported in Conde-Mejia et al.(Conde-Mejía et
al., 2012) and modified for our analysis:
Heatingenergy
7.78MMBtu
tonbiomass nutshells ∗1055MJ1MMBtu
∗1tonnutshells
1000kgnutshells
∗1kg nuts nutshells
0.7kgnutshells∗ 24.6kg nuts nutshells
kgTBC
Lignin Purification:
Dichloromethanemake upflow
150mlDCM8.077gpurifiedlignin
∗1.322gDCM1mlDCM
∗3gDCMused100gDCM
∗7.29gpurifiedlignin
4.98Et. Ac. solublelignin∗10 gEt. Ac. solublelignin1tonEt. Ac. solublelignin
∗1tonDCMused10 gDCMused
∗ 75tonEt. Ac. solublelignin
day∗
day53.5tonTBC
1.5ton/kgDCMton/kgTBC
180
Lignin Depolymerization:
Sub-critical methanol is not reacting in this step since it is used as a solvent for lignin, so it can be
recovered at rates exceeding 99%. Here we model 99% recovery of methanol, in contrast to 97%
recovery rates for dichloromethane and xylene. A small amount of methanol is lost with the
unreacted solubilized lignin:
MeOHmake upflow
30mlMeOH4.98gEt. Ac. solublelignin
∗0.79gMeOH1mlMeOH
∗1gMeOHused100gMeOH
∗10 gEt. Ac. solublelignin1tonEt. Ac. solublelignin
∗1tonMeOHused10 gMeOHused
∗75tonEt. Ac. solublelignin
day
∗day
53.5tonTBC0.01
ton/kgMeOHton/kgTBC
Energy consumption for nutshell preparation, lignin extraction and catalytic depolymerization
were estimated from ecoinvent unit processes, literature(Conde-Mejía et al., 2012) and ASPEN
Plus simulations, respectively.
Catalyst Preparation:
Cu-PMO catalyst preparation was modeled separately based on experimental data by Hansen et
al.(Hansen et al., 2012) Unit processes for input metal salts aluminium nitrate, copper nitrate, and
magnesium acetate were built based on industrial chemistry description in Ullman’s Encyclopedia
of Chemical Engineering(Sienel et al., 2000) and Handbook of Inorganic Chemicals.(Patnaik,
2003) Energy consumption of the catalyst preparation process was estimated based on ASPEN
Plus simulation.
181
For LCA modeling in SimaPro, all input parameters were chosen from existing unit processes in
the ecoinvent 3.1. database, using life cycle inventory unit processes adjusted for the US energy
system (US-EI database; Earthshift, Huntington, VT).
Table B1-Life cycle inventory for production of 1 kg TBC from bio-based resource (Organosolv extraction method)
Material/Assembly Total amount
Allocated amount
Unit
Methanol, at plant/GLO 0.01 0.01 kg Hydrogen, cracking, APME, at plant/RER 0.02 0.02 kg Cu-PMO catalyst 0.56 0.56 kg Ethyl acetate, at plant/RER 0.6 0.6 kg Dichloromethane, at plant/RER 1.5 1.5 kg Methanol, at plant/GLO 1.8 0.2 kg Husked nuts harvesting, at farm/PH 24.5 2 kg Nitrogen fertilizer, production mix, at plant, NREL/ US 0.61 0.05 kg Proxy_Phosphorous Fertilizer (TSP as P2O5), at plant NREL /US 0.98 0.1 kg Proxy_Potash Fertilizer (K2O), at plant NREL /US 0.32 0.03 kg Processes Total
amount Allocated amount
Unit
Electricity, production mix US/US 10 10 kWh Heat, natural gas, at boiler modulating <100kW/RER 141.25 16.9 MJ Cooling energy, natural gas, at cogen unit with absorption chiller 100 kW/CH
57 6.8 MJ
Wood chopping, mobile chopper, in forest/RER 17.21 2 kg Table B2-Life cycle inventory for production of 1 kg Cu-PMO catalyst
Material/Assembly Total amount
Allocated amount Unit
Sodium carbonate from ammonium chloride production, at plant/GLO 0.7 7.8E-5 kg Aluminium nitrate, Al(NO3)3.9H2O 0.07 7.8E-6 kg Copper nitrate, Cu(NO3)2.2 H2O 0.03 3.4E-6 kg Magnesium acetate, Mg(CH3COO)2.4 H2O 0.2 2.2E-6 kg Tap water, at user/RER 89 9E-3 kg Sodium hydroxide, 50% in H2O, production mix, at plant/RER 0.05 5.6E-6 kg
Processes Total amount
Allocated amount Unit
Electricity mix/US 21.8 2.44E-3 kWh
182
Fossil-based Route Inventory. Fossil-based production of TBC was modeled based on a two-step
process with the life cycle inventory given in Table B3, and shown in equations 3 and 4 in the
main text. The first step was based on the catalytic (SeO2) hydroxylation of phenol with hydrogen
peroxide.(Sienel et al., 2000) The second step is butylation of catechol using
triflouromethanesulfonic acid (TFMS) as a catalyst.(Rajadhyaksha & Chaudhari, 1987) Input
chemicals were scaled up based on equations 3 and 4 and their respective conversion yields of
100% and 35%, and scaled to 1 kg TBC as the target product. Energy inputs was estimated from
ASPEN Plus simulations. TFMS was modeled as a new assembly(Siegemund et al., 2000) in the
inventory (Table B4). Selenium dioxide was modeled as Se, adjusting for molecular weights.
Table B3-Life cycle inventory for production of 1 kg TBC from fossil-based resource
Material/Assembly Total amount Allocated amount
Unit
Phenol, at plant/RER 2.5 1.6 kg Hydrogen peroxide, 50% in H2O, at plant/RER 0.9 0.6 kg Isobutanol, at plant/RER 0.6 0.6 kg Xylene, at plant/RER 1.25 1.25 kg Trifluoromethane sulfonic acid 0.003 0.003 kg Selenium, at plant/RER 0.0003 1.9E-4 kg Processes Total amount Allocated
amount Unit
Electricity, medium voltage, at grid/US 2.95 2.4 kWh Transport, freight, rail/RER 1 1 tkm Transport, lorry >16t, fleet average/RER 0.2 0.2 tkm Heat, natural gas, at boiler modulating <100kW/RER 2 2 MJ Table B4- Life cycle inventory for production of 1 kg TFMS used in fossil-based route
Material/Assembly Total amount Allocated amount
Unit
Hydrogen fluoride, at plant/GLO 0.4 0.075 kg Methanol, at plant/GLO 0.2 0.006 kg Oxygen, liquid, at plant/RER 0.3 0.009 kg Proxy_Sulfuric acid, at plant NREL /US 0.3 0.009 kg Processes Total amount Allocated
amount Unit
183
Electricity, medium voltage, at grid/US 1 0.03 kWh Transport, lorry >16t, fleet average/RER 0.2 0.006 tkm Heat, natural gas, at boiler modulating <100kW/RER 2 0.06 MJ
Alternate Lignin Extraction. Here, we considered substitution of an alternate extraction method
for organosolv extraction, the method in our base case scenario. This alternate method is based on
a US patent(Sherman & Gorensek, 2011b) for separation of lignin. We assumed loblolly pine as
an example of softwood that contains lignin fraction with approximately the same chemical
structure as that found for candlenut shells. Ammonium hydroxide and sulfuric acid are used as
solvents for extraction of lignin. Ammonium hydroxide volume to biomass weight ratio is 16:1
and the process can achieve 60% efficiency for lignin separation.
Input parameters were scaled up based on reported inputs for lab scale analysis and ASPEN Plus
simulations:
Cultivation: Cutivatedamount nuts nutshells
1g nuts nutshells0.7gnutshells
∗110gmilledshells
13.62gavailablelignin∗ 13.62gavailablelignin8.17gpurelignin
∗ 10 gpurelignin1tonpurelignin
∗1ton nuts nutshells10 g nuts nutshells
∗ 75tonpurelignin
day
∗ day
53.5tonTBC26.22
kg nuts nutshells kgTBC
Solvent Extraction: AmmoniumHydroxidemake upflow
800mlNH OH50gnutshells
∗0.88kgNH OH1000mlNH OH
∗3kgNH OHused100kgNH OH
∗1000gnutshells1kgnutshells
∗0.7kgnutshells
1kg nuts nutshells∗26.22kg nuts nutshells
kgTBC7.7
184
Electricity consumption for this process is based on ASPEN plus simulation conducted in the same
patent(Sherman & Gorensek, 2011b) for this method and we scaled based on our biomass flow
(nutshell consumption):
Electricity 217kWh
1000kgnutshells∗
0.7kgnutshells1kg nuts nutshells
∗26.22kg nuts nutshells
1kgTBC
3.98kWhkgTBC
Lignin depolymerization is assumed to proceed in an identical fashion as the base case, with
output of 1 kg TBC. Table B5 shows the inputs for bio-based route considering the alternate
extraction method.
Table B5-Life cycle inventory for production of 1 kg TBC from bio-based resource (Solvent extraction method)
Material/Assembly Total amount
Allocated amount
Unit
Methanol, at plant/GLO 0.01 0.01 kg Hydrogen, cracking, APME, at plant/RER 0.02 0.02 kg Cu-PMO catalyst 0.56 0.56 kg Proxy_Sulfuric acid, at plant NREL /US 1.13 0.08 kg Ammonia, liquid, at regional storehouse/RER 7.7 0.5 kg Husked nuts harvesting, at farm/PH 26.22 1.3 kg Nitrogen fertilizer, production mix, at plant, NREL/ US 0.65 0.03 kg Proxy_Phosphorous Fertilizer (TSP as P2O5), at plant NREL /US 1.05 0.05 kg Proxy_Potash Fertilizer (K2O), at plant NREL /US 0.34 0.02 kg Processes Total
amount Allocated amount
Unit
Electricity, production mix US/US 14 10.2 kWh Wood chopping, mobile chopper, in forest/RER 18.45 0.9 kg
185
Alternate Lignin Source. This pathway is hypothetical, considering substitution of coconut shell
lignin and solvent extraction for candlenut shell lignin and organosolv extraction method,
respectively. Input chemicals were scaled based on original experimental data available for
candlenut shell,(Barta et al., 2014) adjusting for lignin content of coconut shell (44%) and weight
percent of nutshell (0.15% for coconut).
Cultivation: Cutivatedamount nuts nutshells
1g nuts nutshells0.15gnutshells
∗110gnutshells
40.15gavailablelignin∗ 40.15gavailablelignin34.93gpurelignin
∗ 10 gpurelignin1tonpurelignin
∗1ton nuts nutshells10 g nuts nutshells
∗75tonpurelignin
day
∗ day
53.5tonTBC29.4
kg nuts nutshells kgTBC
Solvent Extraction: AmmoniumHydroxidemake upflow
800mlNH OH50gnutshells
∗0.88kgNH OH1000mlNH OH
∗3kgNH OHused100kgNH OH
∗1000gnutshells1kgnutshells
∗0.15kgnutshells
1kg nuts nutshells∗29.4kg nuts nutshells
kgTBC1.86
Electricity 217kWh
1000kgnutshells∗
0.15kgnutshells1kg nuts nutshells
∗29.4kg nuts nutshells
1kgTBC
0.95kWhkgTBC
Table B6 shows the inventory for the alternate lignin source and process. As mentioned in the
main text, while we still scale the output of the process to 1 kg of TBC, this method is hypothetical
and the actual final products should be specified experimentally.
186
Table B6-Life cycle inventory for production of 1 kg TBC from bio-based resource (Coconut shells+ Solvent extraction method)
Material/Assembly Total amount
Allocated amount
Unit
Methanol, at plant/GLO 0.01 0.01 kg Hydrogen, cracking, APME, at plant/RER 0.02 0.02 kg Cu-PMO catalyst 0.56 0.56 kg Proxy_Sulfuric acid, at plant NREL /US 0.4 0.14 kg Ammonia, liquid, at regional storehouse/RER 1.86 0.7 kg Husked nuts harvesting, at farm/PH 29.4 1.6 kg Nitrogen fertilizer, production mix, at plant, NREL/ US 0.7 0.04 kg Proxy_Phosphorous Fertilizer (TSP as P2O5), at plant NREL /US 1.18 0.06 kg Proxy_Potash Fertilizer (K2O), at plant NREL /US 0.4 0.02 kg Processes Total
amount Allocated amount
Unit
Electricity, production mix US/US 11 10.2 kWh Wood chopping, mobile chopper, in forest/RER 4.41 1.6 kg
Waste Treatment Considerations. Waste management of various solvents used for both base case
fossil-based and bio-based routes were considered as a separate analysis here. Dichloromethane,
ethyl acetate and hydrogen peroxide were treated as hazardous wastes based on EPA Best
Demonstrated Available Technology (BDAT)(Hansen et al., 2012) for waste management of
relevant group of chemicals. Table S7 and S8 show the chosen unit processes from eco-invent and
the amount of corresponding solvent for landfill and incineration, respectively.
Table B7-Landfill waste treatment scenario for base case bio-based and fossil-based routes
Treatment Process Treated solvent Allocated amount
Unit
Bio-based Route
Proxy_Disposal, n-butyl alcohol, to sanitary landfill NREL /US
Methanol 0.2 kg
Disposal, hazardous waste, 0% water, to underground deposit/DE
Dichloromethane 1.5 kg
Proxy_Disposal, formaldehyde, to unspecified treatment NREL /US
Ethyl acetate 0.6 kg
Fossil-based Route
187
Proxy_Disposal, light aromatic solvent naphtha, to sanitary landfill NREL /US
Phenol 1.6 kg
Disposal, hazardous waste, 0% water, to underground deposit/DE WITH US ELECTRICITY U
Hydrogen peroxide
0.6 kg
Proxy_Disposal, n-butyl alcohol, to sanitary landfill NREL /US
Isobutanol 0.6 kg
Proxy_Disposal, light aromatic solvent naphtha, to sanitary landfill NREL /US
Xylene 1.2 kg
Table B8-Incineration waste treatment scenario for base case bio-based and fossil-based routes
Treatment Process Treated solvent Allocated amount
Unit
Bio-based Route Disposal, solvents mixture, 16.5% water, to hazardous waste incineration/CH
Methanol 0.2 kg
Disposal, hazardous waste, 25% water, to hazardous waste incineration/CH
Dichloromethane 1.5 kg
Disposal, hazardous waste, 25% water, to hazardous waste incineration/CH
Ethyl acetate 0.6 kg
Fossil-based Route
Treatment Process Treated solvent Allocated amount
Unit
Disposal, solvents mixture, 16.5% water, to hazardous waste incineration/CH
Phenol 1.6 kg
Disposal, hazardous waste, 25% water, to hazardous waste incineration/CH
Hydrogen peroxide 0.6 kg
Disposal, solvents mixture, 16.5% water, to hazardous waste incineration/CH
Isobutanol 0.6 kg
Disposal, solvents mixture, 16.5% water, to hazardous waste incineration/CH
Xylene 1.2 kg
188
Results:
Figure B1- Results for process and material contribution in production of 1 kg Cu-PMO catalyst
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Electricity
Sodium hydroxide
Tap water
Magnesium acetate
Copper nitrate
Aluminium nitrate
Sodium carbonate
189
Table B9-Relative results for residual solvents treatment methods
Impact Category Ozone depletion
Global warming
Smog Acidific-ation
Eutrophi-cation
Carcino-genics
Non-carc-inogenics
Respiratory effects
Ecotoxicity Fossil fuel depletion
Unit kg CFC-11 eq
kg CO2 eq
kg O3 eq
kg SO2 eq
kg N eq CTUh CTUh kg PM2.5 eq CTUe MJ surplus
Comparative Results
13,084% -13% 65% 79% 35% 78% 144% 103% 18% -54%
Fossil-based Route + incineration
7.72E-07 2.19E+01 6.59E-01
6.05E-02
4.29E-02 8.59E-07 5.77E-07 4.31E-03 2.23E+01 4.81E+01
Bio-based Route + incineration
1.02E-04 1.90E+01 1.09E+00
1.08E-01
5.79E-02 1.53E-06 1.41E-06 8.74E-03 2.64E+01 2.21E+01
Comparative Results
17,529% -1% 75% 85% 54% 28% 184% 113% -5% -59%
Fossil-based Route + landfill
5.75E-07 1.37E+01 5.73E-01
5.42E-02
2.91E-02 6.27E-07 3.82E-07 3.93E-03 1.98E+01 4.60E+01
Bio-based Route + landfill
1.01E-04 1.35E+01 1.00E+00
1.00E-01
4.48E-02 8.00E-07 1.08E-06 8.36E-03 1.88E+01 1.91E+01
190
Figure B2-Total environmental impacts for 1 kg TBC from bio-based routes (candlenut shell and coconut shell) and fossil-based route (phenol)
0
4
8
12
16
kg CO2 eq.
0
0.4
0.8
1.2
kg O3 eq.
0
0.02
0.04
0.06kg N eq.
0.E+00
4.E‐07
8.E‐07
CTU
h
0
0.004
0.008
0.012
kg PM2.5 eq.
0
8
16
24
CTU
e
0
10
20
30
40
50
MJ surplus
TBC from candlenut shells
TBC from coconut shells
TBC from fossil source
0.E+00
5.E‐05
1.E‐04
2.E‐04
kg CFC
‐11 eq.
0
0.05
0.1
0.15
kg SO2 eq.
0.0E+00
5.0E‐07
1.0E‐06
1.5E‐06
CTU
h
Ozone Depletion Potential
Global Warming Potential Smog
Acidification Eutrophication Carcinogenics
Non‐carcinogenics Respiratory Effects Ecotoxicity
Fossil Fuel Depletion
191
APPENDIX C
Life Cycle Assessment of UV-Curable Biobased Wood Flooring Coatings
Method:
BRC and control layers are modeled in ecoinevnt, using existing unit processes or creating new
ones where the exact chemical or its approximate is not available. Life cycle inventories of these
unit processes include material inputs and energy use. Material inputs are mostly sourced from
literature(Hess et al., 1995; Sienel et al., 2000) and MSDS data. Energy use, on the other hand, is
not reported for most of these chemicals, so default specifications of existing unit processes for
organic chemicals, are used as primary estimations for energy use and chemical plant infrastructure
(Table C1). Mentioned default values are based on average values for European industrial plants,
adjusted based on US energy systems, however, the choice of production method and chemical
complexity can have significant effects on these values.
Table C1-Default values in ecoinvent for organic chemical unit processes (per kg of target chemical)
Parameter Value
Heat, unspecific, in chemical plant/RER with US electricity U (MJ) 2.0
Electricity, production mix US/US with US electricity U (kWh) 0.3
Chemical plant, organics/RER with US electricity U 4.0 × 10-10
Using above data, about forty new unit processes are added to the ecoinvent in order to encompass
all the chemicals used in coatings formulations. In some cases, the precursors of the target chemical
192
are not available in the database so the modeling include all the upstream processes up to the first
precursor available in ecoinvent database.
Results:
Alternative BRC formulation is modeled and assessed as a complementary analysis. Table C2
shows absolute and comparative LCA results of this formulation relative to the conventional
control coating.
Table C2-Absolute and relative life cycle impacts of alternative BRC wood flooring coating compared to control UV-cured coatings (per m2 of coating)
Impact Category Unit % Change
(alternative BRC relative to control)
Ozone depletion kg CFC-11 eq. -32%
Global warming kg CO2 eq. -43%
Smog kg O3 eq. -52%
Acidification kg SO2 eq. -26%
Eutrophication kg N eq. 1%
Carcinogenics CTUh -19%
Non-carcinogenics CTUh -52%
Respiratory effects kg PM2.5 eq. -57%
Ecotoxicity CTUe -41%
Fossil fuel depletion MJ surplus -53%
193
Results of Table C2 show significant reduction in environmental impacts of BRC formulation
compared to the control UV-cured coating. Synthesis of renewable building blocks from
agricultural residues, mitigates impacts of agricultural activities while providing the same function
and durability. Figure C1 shows comparative results for different layers of primary and alternative
BRC formulation, compared to the control coating counterparts. Green and gray colors are same
as before while the blue bars represent the alternative scenario for BRC formulation. Abrasion
resistant sealer, sanding sealer and topcoat are abbreviated as ARS, SS and TC. As indicated in
the figure, the alternative formulation is showing superior performance, especially in four
categories of smog formation, acidification, eutrophication and respiratory effects. Observed trend
is mainly due to the low contribution of corn stover in environmental impacts of cultivation and
milling process of corn. As mentioned in chapter 4, based on economic allocation, share of corn
stover from associated impacts is only 12%.
194
Figure C1- Life cycle comparison between layers of alternative BRC coating and control coating
195
APPENDIX D Evaluating Microalgal Integrated Biorefinery Schemes: Empirical Controlled
Growth Studies and Life Cycle Assessment
These data have been sent to the journal of Bioresource Technology Method:
Table D1- Life cycle inventory data used or created for modeling the production of freshwater and marine growth media. Mass quantities used in life cycle inventory are adjusted for levels of hydration and relative purity.
Chemical Concentration (g/L)
Data source Life cycle inventory notes
Seawater NaNO3 (replete) 6.07E-01 [stoichiometric
calculation] nitric acid + soda ash
NaNO3 (deplete) 6.07E-02 [stoichiometric calculation]
nitric acid + soda ash
Na2Glycerophosphate .5H2O
6.86E-03 [stoichiometric calculation]
Glycerol + Na2HPO4
HEPES buffer 6.48E-02 ecoinvent 2.2 acetic acid, 90% in H2O Biotin 7.35E-07 [below cut-off threshold] CoCl2·6H2O 1.53E-05 [below cut-off threshold] Fe(NH4)2(SO4)2·6H2O 2.23E-03 [stoichiometric
calculation] FeSO4 + (NH4)2SO4
FeCl3·6H2O 1.56E-04 ecoinvent 2.2 iron (III) chloride, 40% in H2O, at plant
H3BO3 3.63E-06 ecoinvent 2.2 boric acid, anhydrous, powder, at plant
MnSO4·H2O 5.23E-04 ecoinvent 2.2 manganese oxide, at plant/CN U Na2EDTA·2H2O 1.91E-03 ecoinvent 2.2 EDTA, ethylenediaminetetraacetic
acid, at plant Thiamine 3.24E-05 [below cut-off threshold] Vitamin B12 3.97E-06 [below cut-off threshold] ZnSO4·7H2O 7.01E-05 ecoinvent 2.2 zinc monosulphate, ZnSO4.H2O, at
plant
Freshwater NaNO3 (replete) 6.07E-01 [stoichiometric
calculation] nitric acid + soda ash
NaNO3 (deplete) 6.07E-02 [stoichiometric calculation]
nitric acid + soda ash
CaCl2·2H2O 2.50E-02 ecoinvent 2.2 calcium chloride, CaCl2, at plant MgSO4·7H2O 7.50E-02 ecoinvent 2.2 magnesium sulphate, at plant
196
K2HPO4 7.50E-02 [stoichiometric calculation]
KOH + phosphoric acid
KH2PO4 1.75E-01 [stoichiometric calculation]
KOH + phosphoric acid
NaCl 2.50E-02 ecoinvent 2.2 sodium chloride, powder, at plant Na2EDTA·2H2O 4.50E-03 ecoinvent 2.2 EDTA, ethylenediaminetetraacetic
acid, at plant FeCl3·6H2O 5.82E-04 ecoinvent 2.2 iron (III) chloride, 40% in H2O, at
plant MnCl2·4H2O 2.46E-04 [stoichiometric
calculation] MnO + HCl (incl. avoided Cl2)
ZnCl2 3.00E-05 [below cut-off threshold] CoCl2·6H2O 1.20E-05 [below cut-off threshold] Na2MoO4·2H2O 2.40E-05 [below cut-off threshold] CaCO3 (optional) 2.00E-04 ecoinvent 2.2 limestone, milled, packed, at plant Vitamin B12 1.35E-04 [below cut-off threshold] HEPES buffer pH 7.8 3.60E-02 ecoinvent 2.2 acetic acid, 90% in H2O Biotin 2.50E-05 [below cut-off threshold] Thiamine 1.10E-03 [no data]