Bioresource Technology 328 (2021) 124833

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Integrated process simulation for bioethanol production: Effects of varying

lignocellulosic feedstocks on technical performance
Marjorie Morales *, Anders Arvesen , Francesco Cherubini
Industrial Ecology Programme, Department of Energy and Process Engineering, Norwegian University of Science and Technology (NTNU), Høgskoleringen 5, Trondheim
7034, Norway

H I G H L I G H T S

• Feedstock-specific data for bioethanol production support more accurate assessments.

• Seven lignocellulosic biomasses were analyzed.
• Detailed process simulation of bioethanol production was performed.
• Feedstock type affects conversion performances and mass/energy balances.
• Feedstock carbohydrate content is the main driver of energetic efficiency.

A R T I C L E I N F O A B S T R A C T

Keywords: Variations in lignocellulosic feedstock composition can influence conversion performance of bioethanol pro­
Agricultural residues duction, but such effects are overlooked in several studies that rely on standard conversion factors. This study
Energy crops investigates the effects of seven lignocellulosic feedstocks (belonging to the categories energy crops, forest and
Forest residues
agricultural residues) on mass, carbon, water and energy balances for biochemical bioethanol production,
Inventory data analysis
Second generation bioethanol
including a comparison of individual process step yields. We find that overall bioethanol yields vary consider­
ably, ranging between 19.0 and 29.0%, 27.3 and 46.2%, and 19.0 and 31.0%, for energy and carbon efficiency,
respectively. The highest yields are found for switchgrass, which has the largest carbohydrate content, and the
lowest for forest residues (spruce). Feedstock composition also affects water and carbon balances. Overall, the
type of biomass influences conversion performances, thereby calling for explicit representation of the effects of
biomass types in technical, economic and environmental assessment studies of bioethanol production.

1. Introduction The economic viability of the bioethanol industry is dependent on

sufficient year-round supply of high-quality and low-cost bio-based
Bioethanol is one of the most promising alternative fuels to reduce renewable feedstocks, and on efficient biomass-to-bioethanol conver­
consumption of fossil fuels in the transport sector (Wang et al., 2013). sion technologies (Li et al., 2016). Currently, the main feedstocks for
The world production of bioethanol reached 115 billion L in 2019 (IEA, bioethanol production in USA, Brazil and EU are different types of food
2020), which is about 7.7% of the global gasoline demand (OECD, crop substrates: corn starch in USA (94% of total production) (RFA,
2019). USA and Brazil produced 84% of the bioethanol worldwide in 2019), sugarcane in Brazil (99% of total production) (Barros, 2019) and
2019, followed by European Union (5%), China (3%), Canada (2%) and sugar beet in EU (40% of total production) (Sharma et al., 2020).
India (2%) (RFA, 2020). Before the COVID-19 pandemic, global bio­ However, lignocellulosic feedstocks appear to be the most promising
ethanol production was expected to increase to 130 billion L by 2024 and those with the largest deployment potentials (Sharma et al., 2020;
(IEA, 2019). A contraction of 13% in 2020 is now expected for biofuels Slade et al., 2014), and constitute an important part of most future
(bioethanol and biodiesel) owing to the pandemic (IEA, 2020), but emission scenarios consistent with stringent climate change mitigation
growth can be expected to resume once the pandemic abates. objectives (Eisentraut, 2010; Shukla et al., 2019). Lignocellulosic

* Corresponding author.
E-mail address: marjorie.morales@ntnu.no (M. Morales).

https://doi.org/10.1016/j.biortech.2021.124833
Received 17 December 2020; Received in revised form 5 February 2021; Accepted 5 February 2021
Available online 12 February 2021
0960-8524/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
M. Morales et al. Bioresource Technology 328 (2021) 124833

feedstocks include woody (softwood and hardwood) and herbaceous costly and complicated step (Zabed et al., 2017), and, depending on the
(perennial grasses) energy crops, agricultural residues (cereal straw, method used, may require detoxification treatment to remove unwanted
stover and bagasse), forest residues (sawdust, pruning and bark thinning substances (Robak & Balcerek, 2020). As the performances of each
residues), and organic portions of municipal solid wastes. Bioethanol process area are strictly interrelated, there are cascading impacts of the
production from lignocellulosic biomass materials typically has lower pretreatment yield on downstream processes, including enzymatic hy­
life-cycle greenhouse gas (GHG) emissions and lower risks to compete drolysis and fermentation (Gerbrandt et al., 2016).
with food security than bioethanol production from food and feed crops The use of process simulation offers opportunities to study specific
(Padella et al., 2019; Su et al., 2020; Zabed et al., 2017). combinations of feedstocks and conversion technologies, and to achieve
The production of lignocellulosic bioethanol (including pilot plant a detailed understanding of the conversion to bioethanol through
facilities) reached 3 BL in 2018 (Sharma et al., 2020), representing less flowsheets where all process units and streams are interconnected. It
than 1% of the total bioethanol produced worldwide. However, as determines all mass and energy balances, thus allowing for the fate of
already noted, this type of biofuel has promising potential for a future elements, chemical and energy to be identified (Humbird et al., 2011).
large-scale deployment. Several companies are demonstrating pilot and In addition, process simulation gives insights into technical performance
commercial scale plants using different technological alternatives of the reactors and other elements that can help to identify optimal
(Chandel et al., 2018; Kudakasseril Kurian et al., 2013). Commercial parameters and conditions. Detailed process modelling and optimization
plants have been constructed in US, Brazil, Europe and China, but only a studies can be valuable in demonstrating effective pathways for
few are reported as operational in US, Brazil and Norway (Padella et al., exploitation of lignocellulosic biomass (Sharma et al., 2020) and in
2019). The global production of lignocellulosic biomass is roughly 170 informing environmental assessments of bioethanol production. How­
billion metric tons per year (Su et al., 2020), and the IEA estimates that ever, partly due to a lack of availability of transparent process simula­
the potential use of 10% of global forestry and agricultural residues in tions, most existing bioethanol Life-Cycle Assessment (LCA) studies use
2030 can provide 233 billion L of ethanol globally (155 billion liter simplified mass and energy balances or take outcomes of process sim­
gasoline equivalent or 5.2 EJ) (Eisentraut, 2010). ulations for a specific feedstock as representative also for other feed­
Lignocellulosic biomass consists of three major components: cellu­ stocks (Chrysikou et al., 2018; Daylan & Ciliz, 2016; González-García
lose (40–60%), hemicellulose (20–40%) and lignin (10–25%) (Su et al., et al., 2010). Many LCA studies are based on the US Department of
2020). Due to heterogeneity of lignocellulosic feedstocks in terms of Energy’s National Renewable Energy Laboratory’s (NREL) (Humbird
composition and moisture, identifying uniform conversion methods and et al., 2011) process design for corn stover as feedstock, assuming that
generally optimal conditions is challenging. In practice, a variety of the bioethanol yield obtained for corn stover is representative also for
different conversion pathways and upgrading routes have been imple­ other biomass types. Such simplifications ignore variations in yields and
mented to convert biomass into bioethanol (Gaurav et al., 2017). There process performances when different types of lignocellulosic biomass
are two main categories of conversion technologies for bioethanol pro­ are used as feedstocks.
duction from lignocellulosic biomass: biochemical and thermochemical. Considering the significance of bioethanol being produced today in
Grassy biomass with high ash content is typically more favored by the world, and the expected growth in production volumes of lignocel­
biochemical conversion, because biochemical conversion is strongly lulosic bioethanol in future climate change mitigation scenarios,
dependent on cellulose and hemicellulose content, while the low ash and detailed analyses showing the effects of different feedstocks on con­
high lignin content of woody biomass make it more suitable for ther­ version yields, emission factors, and mass and energy balances are
mochemical processes (Li et al., 2016). This study specifically analyses needed. Such analyses can offer an understanding of the dependencies of
the biochemical alternative for bioethanol production. process conditions on key assumptions and make available a variety of
Biochemical conversion requires that the biomass is first grinded into data to improve the modelling (and transparency) of bioethanol pro­
(smaller) particles. Then, the lignocellulosic structure needs to be duction plants in environmental impact studies or in estimates of bio­
broken down into chemical fractions that include cellulose, hemicellu­ energy potentials. The aim of this work is to design a biochemical
lose and lignin polymer fractions, using a suitable pretreatment method. conversion process for bioethanol production from lignocellulosic
The released cellulose and hemicellulose molecules in the pretreated biomass and explore the effects of alternative feedstocks on process
biomass are then hydrolyzed into soluble sugars via chemical or enzy­ performance, bioethanol and co-product yields, conversion factors, and
matic method, which are finally converted into bioethanol during mi­ mass, energy, carbon and water balances. We make detailed results
crobial fermentation. Among other factors, the type of pre-treatment can available for future use by analysts. Increasing the availability of data for
have an important role in affecting the overall system performances of different biofuel conversion technologies and feedstocks can lead to
bioethanol production (Maurya et al., 2015; Talebnia et al., 2010; more accurate representations of biofuel systems in multiple applica­
Tomás-Pejó et al., 2011). For example, based on the type of pretreatment tions, including life-cycle assessments, integrated assessment models,
method applied, a sugar yield of 74–99.6% of maximum theoretical was and techno-economic analyses.
achieved after enzymatic hydrolysis of wheat straw (Talebnia et al.,
2010). 2. Material and methods
Bioethanol produced in the fermentation broth is recovered and
purified to obtain a fuel grade bioethanol. The recovery process is usu­ 2.1. Lignocellulosic feedstocks
ally done via distillation, which produces bioethanol, lignin residues and
wastewater. A portion of water is recirculated as backset, while lignin Seven types of lignocellulosic biomass feedstocks are considered in
can be combusted and converted into electricity and heat. The pre­ our analysis: energy crops (switchgrass, miscanthus and eucalyptus),
treatment before hydrolysis is necessary for lignocellulosic biomass in agricultural residues (corn stover and wheat straw), and forest residues
order to alter cellulose structures for enzyme accessibility. This is unlike (from birch and spruce).
for sugar and starch-based biomass, which only requires extraction and
hydrolysis to get fermentable sugars. - Non-food energy crops: Energy crops are typically perennial crops
The final bioethanol yield depends on the type of biomass feedstock with a short growth period, with lower requirements for irrigation
and process conversion efficiency, which further varies with plant size, and fertilizers than most food crops. In our analysis, we consider the
process conditions and microorganisms used in fermentation (Tye et al., most typical species of energy crops, a woody species and two
2016). Effective pretreatment methods to degrade the lignocellulosic perennial grasses. Eucalyptus globulus is a hardwood forest crop,
matrix into individual components are key to achieving high overall which can be cultivated at high yields in tropical and subtropical
conversion performance (Anu et al., 2020). Pretreatment is also the most climates over rotation periods that range from 5 to 10 years (Morales

2
M. Morales et al. Bioresource Technology 328 (2021) 124833

et al., 2015). Miscanthus and switchgrass are C4-type perennial data of each biomass component (See Supplementary material, available
grasses that can grow in temperate regions and are harvested to the following open data repository: https://doi.org/10.5281/zenodo.
annually (Zabed et al., 2017). 4491824). The percentages of forest residues are obtained calculating
- Agricultural residues: Agricultural residues comprise left-over resi­ the mass of each forest residue (mannual,i, Mg⋅m− 3), which is defined by
dues after harvesting and consists of stalks, leaves, cobs and husks. the equation mannual,i = BEFi⋅fi, where BEF is the Biomass Expansion
Four major crops producing agricultural residues today are wheat, Factors, in Mg⋅m− 3 stem under bark, as reported by (Lehtonen et al.,
rice, sugarcane, and corn (Saini et al., 2015). Corn stover and wheat 2004), fi is the extraction factor (percentage collected of component i),
straw were here selected as representatives of agricultural residues. and i is the biomass component (foliage, living and dead branches). BEFs
- Forest residues: Forest residues for bioethanol production can be for each biomass component follows the formula BEFi(t) = ai + bie− 0.01t,
integrated within existing wood value chains without adding further where ai and bi are parameters, t is the stand age (in years) and i is the
pressure on terrestrial ecosystems and promote a circular economy biomass component. Rotation period is assumed to be 100 and 60 years
perspective (Cavalett & Cherubini, 2018). Two types of forest resi­ for Norway Spruce and Birch sp., respectively. Values for ai and bi are
dues are considered in this study as representatives of hardwood and defined in Supplementary material. The extraction factor (fi) for these
softwood biomass, birch (Birch sp.) and Norway spruce (Spicea compounds assume 75% of foliage for Spruce and Birch, i.e. 75% of total
Abies), respectively. Forest residues are mainly composed by stem foliage available is collected; while 75% and 50% of the branches (alive
bark, branches, foliage, needles, stumps and roots. However, only and dead) are collected for Spruce and Birch, respectively.
above-ground components, i.e. foliage and branches, are considered
as available residues to be used as feedstock for bioethanol produc­
2.2. Processes simulation
tion, excluding the bark and under-ground residues (stumps and
roots).
The bioethanol production plant is divided into seven main sections
or areas (Fig. 1): Storage and chipping; Pretreatment; Enzymatic
Table 1 shows the composition and key properties of the lignocel­
saccharification; Co-fermentation of pentoses (xylose) and hexoses
lulosic feedstocks evaluated in this study. The composition and moisture
(glucose); Bioethanol recovery (distillation and dehydration); Cogene­
of lignocellulosic biomass can vary by region, soil types, weather,
ration of heat and power; and Wastewater treatment. Storage and
fertilization practices, agricultural practices, among other factors.
chipping activities are located next to the production plant. The two
However, the compositions chosen in this study are specific for each
auxiliary activities enzyme production and yeast production are inte­
feedstock type without considering variability in each component for a
grated into the plant. By other hand, chemicals production is excluded of
given feedstock.
the plant boundaries. Note that feedstock production lies outside the
Glucan, galactan and mannan belong to C6 carbohydrates, while
scope of this work, which focuses on conversions of given feedstocks to
xylan and arabinan represent C5 carbohydrates. Glucan and xylan are
bioethanol.
the main carbohydrates in lignocellulosic feedstocks. Xylan is the main
The process design is the same for all lignocellulosic feedstocks, but
hemicellulosic component, while glucan contains both cellulose and
differences in the physicochemical characterization of the feedstocks are
hemicellulose (Vieira et al., 2020).
explicitly taken into account. Standard conditions were prioritized
Among the feedstocks considered in this study, perennial grasses
above optimization of individual pathways. The dilute acid pretreat­
have the largest fraction of carbohydrates, ranging between 63.4% and
ment was selected because it is one of the most frequently used pre­
68%; followed by eucalyptus (63.8%) and corn stover (61.5%). Euca­
treatment method for bioethanol production (Laser et al., 2009). The
lyptus also has the highest glucan content. Woody species have highest
selection of data sources for reactions, yields and conditions focused on
lignin content, and agricultural residues the smallest. The opposite
similar process conditions for the dilute acid pretreatment, e.g. similar
happens with ash fractions, for which agricultural residues have the
temperature, pressure, acid concentration, residence time and solid
highest content, while the woody biomass the lowest.
loading, among the feedstocks evaluated. Saccharification and co-
The composition of forest residues (Norway Spruce and Birch) are
fermentation were evaluated separately by means of SHCoF (separated
obtained multiplying the percentages of forest residues (branches and
saccharification and co-fermentation), as opposed to simultaneously by
foliages) by the percent of every specific average chemical composition
means of SSCoF (simultaneous saccharification and co-fermentation), so

Table 1
Composition, moisture and lower heating value of lignocellulosic feedstocks delivered to the plant gate.
Composition (% DM Eucalyptus Birch sp. Residues Spruce sp. Switchgrass Miscanthus Corn stover Wheat straw
basis) globulus Residues

Glucan 48.1 31.3a 38.5 42.6 41.0 38.1 32.6

Xylan 13.7 21.5a 5.1 23.1 19.9 20.3 19.2
Arabinan 0.2 0.5b 1.8 1.5 1.7 2.0 2.4
Galactan 0.8 0.4b 1.4 0.5 0.6 0.7 0.8
Mannan 1.0 2.4b 8.0 0.3 0.09 0.4 0.3
Lignin 29.9 22.6a 34.1 24.1 22.4 20.2 16.9
Ash 0.0 0.0b 1.1 4.0 2.9 11.0 10.4
Acetate 3.6 5.1b 0.8* 1.9* 1.75* 2.5* 4.4
Extractives 2.7 16.3* 9.2 2.0* 9.6 4.8 13.0
Carbohydrates (%) 63.8 56.1 54.8 68.0 63.4 61.5 55.3
G/C ratio** 0.75 0.56 0.70 0.63 0.65 0.62 0.59
kg Carbon/kg 0.47 0.40 0.45 0.45 0.42 0.40 0.37
feedstock
a
References (Morales et al., (Räisänen and Athanassiadis, (Frankó et al., (Zhu and Pan, (Scordia et al., (Zhu and Pan, (Zhu and Pan,
2017) 2013)b(Wang et al., 2013) 2016) 2010) 2013) 2010) 2010)
Moisture*** 60.0 47.9 45.0 20.0 15.0 15.0 10.0
References (Morales et al., (Nurmi, 1997) (Frankó et al., (Laser et al., (Scordia et al., (Humbird (Bondesson and
2017) 2016) 2009) 2013) et al., 2011) Galbe, 2016)
LHV (MJ/kg) 17.2 19.0 18.5 16.7 18.0 15.8 16.8

*Values used to close the mass balance to 100%. **Glucose/Carbohydrates ratio. *** Water content (%) at the entrance of production plant.

3
M. Morales et al. Bioresource Technology 328 (2021) 124833

Fig. 1. Simplified flowchart of lignocellulosic bioethanol production plant.

as to have a clearer idea of the effect of biomass in each stage of the considered in our model employs dilute acid pretreatment, which uses
process separately. acid at low concentration and high temperature. The dilute acid pre­
Fig. 2 shows a detailed overview of the lignocellulosic bioethanol treatment, with sulfuric acid used as pre-hydrolysis agent (Conde-Mejia
plant modelled in this study. The bioethanol refinery in the present et al., 2012), increases hemicellulose degradation-solubilization rate in
design has a processing capacity of 128,000 kg biomass⋅h− 1 (wet basis). comparison to other pretreatment methods, such as steam explosion and
The process models were developed using Aspen Plus v.10 (Aspen liquid hot water (Conde-Mejia et al., 2012), and requires shorter resi­
Technologies, Inc., USA) in order to estimate the energy and mass re­ dence time. However, this method produces some toxic inhibitor com­
quirements for each biomass feedstock. The thermodynamic model pounds, such as furfural and 5-hydroxymethyl furfural (HMF),
NRTL (Non-Random Two Liquid) was used to calculate the phase be­ negatively affecting the fermenting organisms. Thus, this pretreatment
haviors. Bioethanol production was simulated as a unit-process area method requires a washing step for pulp and neutralization step for the
model, in which the reaction conversions and operative conditions for acid dissolved into the liquor in order to remove the compounds that are
pretreatment, saccharification and fermenters reactors, were taken from inhibitory to the fermentation. This removal also helps to avoid releases
literature sources. The process parameters and efficiencies used in the of toxic compounds to the environment in subsequent process steps.
simulations are detailed in the sections below, as well as the process Supplementary material shows the reactions and conversion factors
integration techniques used to conserve and re-use water, chemicals and considered for the three main categories of biomass (classified as woody
energy (heat and power). biomass, grassy biomass and agricultural residues).
The flow diagram for the pretreatment is shown in Fig. 2. Stream
2.2.1. Storage and chipping 102L is the water used to dilute the sulfuric acid (stream 101L) to 1% w/
The storage and chipping area handles incoming biomass feedstocks w and the lignocellulosic biomass (stream 101S) to 30% wt solid con­
with moisture and composition as detailed in Table 1. The equipment for centration. The process model considers energetic integration, taking
storage and chipping is physically located at the bioethanol plant, next advantage of the heat of all currents by using them to heat colder
to the pretreatment process. In the storage and chipping area, the streams, e.g. stream 102L is pre-heated through three heat exchangers
biomass is preprocessed and homogenized to uniform particle size and HE-101, HE-102 and HE-103.
bulk density. The biomass is milled to a mean particle size of 4 mm. In Milled biomass from the chipping area (stream 101S) is fed to the
the current study, the storage and chipping area is analyzed separately pretreatment reactor (R-101). The heat exchanger HE-104 uses a high
(i.e., it is not included in the Aspen Plus simulations). The calculations pressure (HP) water steam from A600 (stream 613G-HP) to reach the
include electricity and diesel consumption associated with truck de­ temperature and pressure necessary in the pretreatment reactor. The
livery, short-time yard storage, milling, conveyor belts and for feeding pretreatment reactor is a single horizontal reaction vessel and is
biomass to the pretreatment reactor. Further details on the treatment of designed to operate at 190 ◦ C and 12 atm for 10, 5 or 2 min, for woody
storage and chipping are available in the supplementary information. biomass, agricultural residues and grassy biomass, respectively. After
the pretreatment reactor, two distinct fractions can be identified: pulp
2.2.2. Pretreatment (solid fraction, stream 104F) and liquor (liquid fraction, stream 103L).
The pretreatment (also called pre-hydrolysis) improves the enzy­ The solid fraction is mainly cellulose and other hexoses (mannose,
matic digestibility of the feedstock. Pretreatment breaks down the cell galactose) and degraded lignin, while the liquid fraction includes water,
wall structure, releasing some lignin and soluble sugars (glucose, arab­ soluble carbohydrates, such as pentoses (xylose and arabinose) and
inose, xylose, galactose and mannose) through a pre-hydrolysis of the other soluble solids.
cellulose and hemicellulose (xylan). The pre-treatment method The solid fraction (stream 104F) is sent to washing system to

4
 M. Morales et al.
5

Bioresource Technology 328 (2021) 124833

Fig. 2. Process diagram from lignocellulosic bioethanol production.
M. Morales et al. Bioresource Technology 328 (2021) 124833

eliminate inhibitory compounds, while the liquid fraction (stream 103L) 408L) is between 1 and 7% wt. The specifications of C-402 column is the
is discharged to flash tank operating at atmospheric pressure. This flash following: 18 plates and feed to plate 3, the bioethanol product leaves
is directed to liquor overliming, applying NaOH in the neutralizer the column at 45% wt of bioethanol (stream 410L). The stream 410L is
reactor, to increase the pH from 1 to 10, and then is sent to a gypsum sent to the rectification column.
separator (SEP-102). Solid stream 104F is directed to a rotatory filter The rectification column (C-402) concentrates the bioethanol to a
(SSEP-102) to extract the remaining water and then is carried to a near azeotropic composition. This column has 36 plates and power to
washing system. The washing system consists of three vessels (SSEP-103, plate 30 and leaving the column at 93% bioethanol (stream 412G). The
SSEP-104 and SSEP-105) with water in crosscurrent to eliminate the stream 410L enter to C-402 directly to plate 30. This stream 410L is used
inhibitors components as furfural and hydroxy methyl furfural (HMF). to supply energy, which completes the energy integration of the system.
Finally, the solid stream 120S (20% wt concentration) is directed to the The stream of vinasses (411L) leaving the system can still be used to
Saccharification area, while the stream 110L-E containing pentose car­ exchange heat in some systems with low temperature requirements. For
bohydrates is directly sent to A300. concentration and rectification columns the bottom stream will have
0.5% wt of bioethanol (as maximum). Both distillation columns use
2.2.3. Enzymatic hydrolysis perforated plates, with Murphy efficiencies of 0.48 for the concentration
The diagram of the enzymatic (saccharification) process is shown in column, 0.80 for the rectification column and a pressure drop of 0.1 psi
Fig. 2. In this area, the cellulose is converted to glucose and cellobiose, per plate. The uncondensed bioethanol vapors (streams A417G and
by using cellulase enzymes. The pretreated pulp (stream 120S) at 20% 304G) and CO2 emissions from the concentration column (stream 409G)
wt is mixed with the enzyme preparation (stream 203L) from A201 are sent to an absorption column (vent scrubber). In the vent scrubber,
(enzyme production section). A commercial cellulase enzyme (20 the bioethanol is re-absorbed in freshwater and recycled back to the
FPU⋅g− 1 glucose) (Humbird et al., 2011) is added to the first enzymatic concentration column (stream 10G). Finally, the stream 413G is carried
reactor (R-201) operating at 48 ◦ C. The batch reactors (R-201 and R- to molecular sieves to obtain a dehydrated bioethanol with a bioethanol
202) operate in parallel to allow a continuous discharge of pulp and to concentration greater than 99.5% wt (stream 414G). This provides a
complete the residence time (3.5 days). The first enzymatic reactors are recovery of about 99% for the systems analyzed at 5% bioethanol in the
high-solids batch reactors, but as the saccharification time advances the feed (stream 408L). The molecular sieves are columns packed with beds
viscosity is decreasing, and the slurry can be pumped to the next parallel of adsorbent, where the water is selectively adsorbed in the beds and the
enzymatic reactor. The hydrolyzed pulp is passed through a separator bioethanol vapors flows through. The bioethanol (stream 414G) can be
(SSEP-201) to separate the lignin (stream 206S) to be sent to the cooled and stored.
cogeneration section, while the slurry (stream 205L) is directed to
fermentation, after being cooled to fermentation temperature (32 ◦ C) in 2.2.6. Cogeneration
the heat exchanger HE-201. Supplementary material summarizes the This area is outlined in Fig. 2. The purpose of the cogeneration area is
enzymatic hydrolysis reactions and the conversion efficiencies. The to produce heat (high and medium pressure steam) and electricity,
enzyme production area was based on the NREL report for cellulase through the combustion of the remaining combustible solids and biogas
enzyme production (Humbird et al., 2011). This model considers the produced in other process areas, including:
preparation of a mixture of enzymes, which are catalytic proteins
capable to break down cellulose fibers into glucose, cellobiose and sol­ - The solids separated after the distillation (stream 421S), mainly
uble gluco-oligomers (see Supplementary material for details). composed of lignin and small amounts of cellulose and
hemicellulose.
2.2.4. Fermentation - Biogas (stream 704G)
In this step, glucose and other sugar hydrolysates from pretreatment - The remaining sugars (stream 712S) produced in the water
and enzymatic saccharification are fermented to bioethanol. Fermen­ treatment.
tation process occurs in the reactor R-301 (Fig. 2). The co-fermentation - Lignin from saccharification (stream 206S).
of pentoses and hexoses are considered. Two streams rich in hexoses
(glucose, galactose and mannose) and pentoses (liquid stream rich in These streams are fed to a fluidized bed boiler with a humidity of
xylose), come in the streams 207L and 110L-E respectively. Reactor R- 40%. The technology includes the use of fluidized bed boilers and
301 is inoculated with Zymomonas mobilis, which is a recombinant co- multistage electric generation turbines with steam extraction points at
fermenting bacterium capable of simultaneously ferment hexose and different pressures (for use at different points in the process).
pentoses to bioethanol. The fermentation process is at 32 ◦ C and lasts Preheated air (stream 602G) goes into the boiler at stoichiometric
1.5 days. Supplementary material lists the reaction and conversion ef­ excess. Hot flue gases (stream 603G) are used to heat the 625L preheated
ficiencies during the co-fermentation process. Although most of the water stream. The superheated steam (stream 606G), leaving the boiler,
glucose and xylose sugars are fermented to bioethanol, some sugars may is directed to a condensing turbine for generating electricity. The turbine
be lost due to contamination by other microorganisms. This sugar loss is consists of 3 stages (T-601, T-602, T-603) (the last one operating with
modeled as a fraction of sugars converted to glycerol, succinic acid and vacuum discharge) and two extraction points of steam (V-601 and V-
xylitol. Conversion of galactose, mannose and arabinose is not consid­ 602). The Rankine power cycle employed in our process design for
ered. The small quantity of bioethanol lost in the CO2 vent stream generating electricity and heat from biomass is a consolidated technol­
(304G) is recovered in the vent scrubber. The inoculum production is ogy. Given the level of development of the technology, the simulation of
conducted in a fermenter trains reactor connected in series (See Sup­ this stage is aimed at determining the production of electricity and
plementary material) until reaching the inoculum volume required in steam, without deep detail in the design of the equipment involved. The
the fermenter reactors R-301 (assumed as 10% of total volume). products of the stage are steam and electricity in enough quantities for
the bioethanol production plant to operate independently in terms of its
2.2.5. Distillation and dehydration consumption of heat and electricity, that is, without using electricity
The recovery of the bioethanol is carried out in three steps: con­ from the grid, gas or coal.
centration column, rectification column and molecular sieves to produce
99.5% bioethanol, as shown in Fig. 2. This area separates water, anhy­ 2.2.7. Wastewater treatment
drous bioethanol and solids. Fig. 2 shows an overview of the wastewater treatment (WWT) area,
The concentration column (C-401) removes most of the water and which consists of four main process steps: anaerobic digestion, aerobic
dissolved CO2. The bioethanol concentration in the inlet stream (stream digestion, ultrafiltration membranes and evaporators trains. This area

6
M. Morales et al. Bioresource Technology 328 (2021) 124833

treats all contaminated water effluents from different areas in order to 3. Results and discussion
produce a clean and reusable water. This study assumes that the treated
water is recycled to the production plant, which reduces the freshwater 3.1. Mass balance
requirements and wastewater discharges to the environment.
Contaminated effluents of the bioethanol plant are collected and Table 2 shows the mass balance of the bioethanol plant for the
cooled in a heat exchanger (HX-701) to the operating temperature of the different types of feedstocks considered in our analysis by process con­
anaerobic digester UASB (R-701). From the anaerobic digester, an version step.
output stream (stream 705L) is continuously withdrawn to the aerobic The values for feedstock requirements range from 0.13 to 0.20 kg
reactor (R-702), while methane emissions (stream 704G) are sent to DM-basis per MJ of bioethanol produced. The use of spruce residues has
cogeneration area. The aerobic digestion (R-702) consists of a set of the highest feedstock demand per unit of energy of bioethanol produced,
continuous stirred tank reactors with cell recirculation. Cells are recir­ while switchgrass has the lowest. These differences are attributable to
culated using an ultrafiltration membrane reactor. The recirculation differences in the carbohydrates content of the feedstocks (see Table 1).
system was assumed to be the same as that designed by NREL (Humbird The highest carbohydrates content is reported for switchgrass (68%),
et al., 2011). The aerobic treatment system produces a clean water and the lowest for spruce residues (54.8%). The effects of a feedstock’s
stream (stream 711L) and sludge stream (stream 712S) composed mainly carbohydrates content on bioethanol production are dependent on the
of cell biomass, which are sent to the cogeneration area. carbohydrate conversion efficiencies (and can be set to this value), i.e.
An ultrafiltration treatment was considered for the production of the carbohydrates (glucan + xylan) used to generate 1 MJ bioethanol.
high-quality water to be reused in the process or discharged to a natural The carbohydrates conversion efficiencies correspond to an average of
water source. The system consists of two ultrafiltration membrane re­ 2.3 ± 0.1 kg⋅MJ− 1 bioethanol for all lignocellulosic biomasses
actors producing an effluent free of suspended solids (such as sulfate and evaluated.
acetate), but not of dissolved solids. The dissolved solids are removed in Comparing the water requirement for the evaluated lignocellulosic
a reverse osmosis system (RO-701 and RO-702). After the reverse feedstocks, agricultural residues requires highest volume of water into
osmosis operation, the stream 716L (mainly salts and organic molecules) the process, while the lowest volumes were used for forest biomass.
is concentrated in a 4-effect evaporator system up to 40% of total solids. More information about water balance is available in Section 3.3.
This concentrate is assumed to be disposed as a solid (stream 735L). Different process areas release CO2 emissions: the cogeneration area
produces between 74.2% and 84.3% of the total CO2 emissions, followed
by the bioethanol recovery area with a range from 12.6% to 23.8% of
2.3. Definition of net energy balance and energy efficiency CO2 emissions and about 1% to 5.6% from the wastewater treatment
area. The CO2 emissions from cogeneration are from the combustion
There are several energy metrics to summarize the energy produced reaction in the boiler. The bioethanol recovery area generates CO2
from a process system. Two energy metrics were calculated in this study: emissions from the absorption column. This column treated the uncon­
Net energy value (NEV) and Energy efficiency (ηenergy). densed bioethanol vapors mixed with CO2 and CO2 emissions from the
The NEV is the output energy minus the input energy, by each en­ concentration column. In the wastewater treatment section CO2 emis­
ergetic output i, defined by the equation NEVi = Eoutput i - Einput i, sions are coming from the anaerobic and aerobic reactions. The forest
measured as MJ. The NEV consider only direct energy flows, and it does biomasses generate the highest CO2 emissions, ranging between 0.18
not consider indirect energy flows, e.g. from transport of feedstocks to and 0.29 kg CO2 per MJ of bioethanol produced, while the grassy
the plant, from production of consumables, etc. The energy outputs biomass and agricultural residues emit between 0.15 and 0.18 kg CO2
(Eoutput i) are bioethanol, power and heat, while the energy input (Einput per MJ of bioethanol. The high CO2 emission in forest biomasses is
i) refers to the energetic demands as heat, power and diesel. However, in mainly due to the emissions from the cogeneration area (see Table 2)
the case of steam and diesel are not considered in the NEV, due to the due to their high lignin content.
steam produced is the same required into the process, i.e. NEV is equal to The requirement of sulfuric acid in the pretreatment is related to the
zero, while diesel is consumed but not produced in the plant, i.e. NEV is content of water in the input biomass. Sulfuric acid is added at 1% w/w
equal to a negative value of the diesel consumed. The NEV results for the in the pretreatment reactor, hence, the cases where biomass is rich in
different lignocellulosic feedstocks are presented in Supplementary water, such as eucalyptus and forest residues (see Table 1), require more
material. sulfuric acid than the cases low in water, such as grassy biomasses or
To investigate the energy efficiency (at which feedstock is utilized), agricultural residues. The demand for sulfuric acid leads to a demand for
the total amounts of useful energy used and generated from the plant NaOH to neutralize the stream carried out to WWT, and a consequent
were calculated and compared to the energy contained in the feedstock. production of gypsum as residue.
The ηenergy is the ratio between the “energy out” content in energetic The enzyme doses used in the saccharification are related to the
product i (Eproduct i) to “energy in” content in the biomass (Ebiomass), glucan content in the pretreated biomass, i.e. 20 mg enzyme⋅g− 1 glucan.
defined by the equation ηenergy,i = Eproduct i/ Ebiomass⋅100, measured as a The glucan content depends on the pretreated stream, which in turns
percentage of the feedstock energy content (% feedstock LHV), as depends on two factors: i) glucan content in the feedstock (Table 1) and
described by García-Velásquez and Cardona (2019). The two main en­ ii) glucan conversion to glucose, HMF, gluco-oligosaccharides, and
ergetic products (Eproduct i) are bioethanol and surplus electricity. cellobiose carried out in the pretreatment (see Supplementary material).
Additionally, other energy outputs must be considered as energetic Grassy feedstocks have a content of around 40,000 kg glucan⋅h− 1 in the
products: internal utilities (heat and power), cooling duty and ambient pretreated stream, followed by the agricultural residues between 33,650
losses. The internal utilities are considered as products, although they to 37,066 kg glucan⋅h− 1, while the lowest glucan content is in the woody
are returned to the plant facility. The cooling duty refers to the heat biomass, ranging from 19,242 to 24,964 kg glucan⋅h− 1. However, the
removed in the cooling system. The ambient losses are assumed as the enzyme dose variations in Table 2 is relative the bioethanol produced.
difference between the total Eproducts and Ebiomass. The energy content for On the other hand, the inoculum doses of yeast used in the fermenter
the products and lignocellulosic biomasses are based on the calorific reactors is assumed constant for all biomass, i.e. 10% of total fermenter
value and the mass balance arising from the stream flows computed by reactor volume (561.6 kg⋅h− 1). Thus, inoculum doses variation showed
Aspen Plus. The calorific value for the biomass and for bioethanol are in Table 2 is mainly due to the bioethanol produced.
based on the lower heating value. Lower heating value for each biomass The biomass requirements per MJ of bioethanol produced can be
is described in Table 1, while 26.8 MJ⋅kg-1 is considered as calorific compared with those reported in the literature, summarized in Table 3.
value for the bioethanol generated (Catapano et al., 2015). Values are broadly in the same range, although our results lie on the high

7
M. Morales et al. Bioresource Technology 328 (2021) 124833

Table 2
1
Mass balance per process (kg⋅MJ− bioethanol).
kg /MJ bioethanol

Materials Eucalyptus gl. Birch sp. N. Spruce Switchgrass Miscanthus Corn stover Wheat straw

Co-product: Electricity kWh [MJ] 0.022 [0.078] 0.022 [0.079] 0.040 [0.143] 0.021 [0.076] 0.021 [0.075] 0.020 [0.073] 0.018 [0.064]
PRETREATMENT
Inputs
Feedstock (DM basis) 0.14 0.17 0.20 0.13 0.14 0.14 0.16
Water* 0.32 0.37 0.54 0.43 0.44 0.46 0.47
3 3 3 3 3 3 3
Sulfuric acid 4.8⋅10− 5.8⋅10− 6.7⋅10− 4.5⋅10− 4.8⋅10− 4.8⋅10− 5.6⋅10−
3 3 3 3 3 3 3
NaOH 4.1⋅10− 5.1⋅10− 5.6⋅10− 3.7⋅10− 4.1⋅10− 4.1⋅10− 4.5⋅10−
Outputs (emissions to soil)
3 3 3 3 3 3 3
Gypsum to landfill 5.9⋅10− 7.3⋅10− 8.4⋅10− 5.5⋅10− 6.1⋅10− 6.1⋅10− 6.8⋅10−

SACCHARIFICATION
Inputs
Water* 0.73 0.65 0.96 0.58 0.59 0.56 0.55
4 4 4 4 4 4 4
NaOH 3.7⋅10− 6.4⋅10− 1.2⋅10− 1.9⋅10− 1.8⋅10− 2.6⋅10− 5.1⋅10−
3 4 3 3 3 4 4
Enzyme 1.2⋅10− 9.9⋅10− 1.4⋅10− 1.0⋅10− 1.1⋅10− 9.7⋅10− 9.4⋅10−

CO-FERMENTATION
Inputs
3 3 3 4 4 4 4
Yeast (inoculum) 1.5⋅10− 1.3⋅10− 1.5⋅10− 6.6⋅10− 6.7⋅10− 6.8⋅10− 7.3⋅10−
3 3 3 3 3 3 3
NH4CL 1.5⋅10− 1.2⋅10− 1.7⋅10− 1.2⋅10− 1.3⋅10− 1.2⋅10− 1.1⋅10−
3 3 3 3 3 3 3
Na2SO4 4.0⋅10− 3.1⋅10− 4.4⋅10− 3.3⋅10− 3.4⋅10− 3.1⋅10− 3.0⋅10−
4 4 4 5 5 5 5
CSL 1.5⋅10− 1.3⋅10− 1.5⋅10− 6.6⋅10− 6.7⋅10− 6.8⋅10− 7.3⋅10−
Water* 0.10 0.09 0.13 0.08 0.09 0.08 0.08

BIOETHANOL RECOVERY
Inputs
Water* 0.06 0.05 0.06 0.04 0.04 0.04 0.04
Outputs (emissions to air)
CO2 0.04 0.04 0.04 0.04 0.04 0.04 0.04
3 3 3 4 3 3 3
Water (steam) 1.2⋅10− 1.2⋅10− 1.2⋅10− 9.9⋅10− 1.0⋅10− 1.0⋅10− 1.0⋅10−
5 5 5 5 5 5 5
Bioethanol losses 7.0⋅10− 6.4⋅10− 6.4⋅10− 3.2⋅10− 3.2⋅10− 3.3⋅10− 3.5⋅10−
5 4 4 5 8 8 8
Extract (VOC) 8.8⋅10− 6.5⋅10− 3.7⋅10− 4.8⋅10− 9.5⋅10− 6.8⋅10− 5.7⋅10−
5 5 5 5 5 5 5
Air 2.9⋅10− 2.9⋅10− 2.9⋅10− 2.9⋅10− 2.9⋅10− 2.9⋅10− 2.9⋅10−

CO-GENERATION
Inputs
Air 0.64 0.82 1.07 0.49 0.56 0.50 0.61
Water* 0.38 0.43 0.54 0.25 0.28 0.26 0.30
Chemicals inorganics 0.004 0.004 0.004 0.002 0.002 0.002 0.002
Outputs (emissions to air)
Water (steam) 0.13 0.15 0.20 0.10 0.11 0.11 0.12
CO2 0.15 0.18 0.24 0.11 0.13 0.11 0.13
3 3 3 3 3 4 3
CO 1.3⋅10− 1.5⋅10− 2.1⋅10− 1.0⋅10− 1.1⋅10− 9.9⋅10− 1.1⋅10−
3 3 3 4 4 4 4
NO2 1.3⋅10− 1.1⋅10− 1.4⋅10− 8.5⋅10− 8.7⋅10− 8.5⋅10− 8.5⋅10−
5 5 5 6 6 6 6
NO 1.3⋅10− 1.1⋅10− 1.4⋅10− 8.4⋅10− 8.7⋅10− 8.5⋅10− 8.5⋅10−
Air 0.52 0.67 0.88 0.40 0.46 0.41 0.50
Outputs (emissions to soil)
3 3 3 3 3 3 3
Sludge (ashes) 1.8⋅10− 8.5⋅10− 6.2⋅10− 1.4⋅10− 4.3⋅10− 2.4⋅10− 6.3⋅10−
6 6 6 6 3 2 2
Remaining sugars (ashes) 3.5⋅10− 2.3⋅10− 9.1⋅10− 5.2⋅10− 4.1⋅10− 1.6⋅10− 1.7⋅10−
Outputs (emissions to water)
Water (blowdown) 0.013 0.012 0.013 0.006 0.006 0.006 0.007

WASTEWATER TREAT.
Inputs
3 2 3 3 3 3 3
NH3 (nutrients) 3.0⋅10− 1.1⋅10− 8.7⋅10− 2.5⋅10− 5.7⋅10− 3.4⋅10− 7.9⋅10−
Air 0.12 0.71 0.48 0.09 0.35 0.19 0.53
2 3 3 3 3 3 3
NaOH 1.1⋅10− 7.9⋅10− 8.2⋅10− 1.8⋅10− 1.4⋅10− 1.4⋅10− 1.1⋅10−
Outputs (emissions to air)
Water (steam) 3.0⋅10− 3 1.9⋅10− 2
1.3⋅10− 2
2.2⋅10− 3
9.3⋅10− 3
4.8⋅10− 3
1.4⋅10− 2

CO2 2.1⋅10− 3 1.3⋅10− 2

8.9⋅10− 3
1.6⋅10− 3
6.2⋅10− 3
3.3⋅10− 3
9.5⋅10− 3

Air 0.11 0.69 0.46 0.08 0.34 0.17 0.51

NH3 9.7⋅10− 5 2.0⋅10− 3
8.5⋅10− 4
7.4⋅10− 5
6.3⋅10− 4
2.1⋅10− 4
1.3⋅10− 3
5 5 6 5 6 5
Methane 1.1E-05 1.3⋅10− 1.7⋅10− 8.0⋅10− 1.1⋅10− 9.9⋅10− 1.1⋅10−
Outputs (emissions to water)
2 2 2 3 3 3 3
Water (waste) 2.3⋅10− 1.4⋅10− 1.7⋅10− 3.7⋅10− 5.1⋅10− 3.4⋅10− 3.9⋅10−
8 9 9 9 9 9 9
Bioethanol losses 1.4⋅10− 5.8⋅10− 6.8⋅10− 1.7⋅10− 2.2⋅10− 1.5⋅10− 1.6⋅10−
5 5 6 6 6 6 5
Acetic acid 2.1⋅10− 2.4⋅10− 9.8⋅10− 5.9⋅10− 7.9⋅10− 6.6⋅10− 1.1⋅10−
2 2 2 3 3 3 3
NaSO4 2.3⋅10− 1.7⋅10− 1.9⋅10− 6.1⋅10− 5.6⋅10− 5.4⋅10− 4.7⋅10−
3 4 3 4 4 4 4
NH4Cl 1.2⋅10− 9.1⋅10− 1.3⋅10− 9.4⋅10− 9.8⋅10− 8.9⋅10− 8.6⋅10−
5 5 5 5 5 5 5
Sodacetate 3.4⋅10− 5.9⋅10− 1.1⋅10− 1.7⋅10− 1.7⋅10− 2.4⋅10− 4.7⋅10−
5 5 5 5 5 5 5
Glycerol 1.6⋅10− 1.6⋅10− 1.6⋅10− 1.6⋅10− 1.6⋅10− 1.6⋅10− 1.6⋅10−
4 4 4 4 4 5 5
Remaining sugars 2.3⋅10− 4.3⋅10− 1.9⋅10− 2.1⋅10− 2.0⋅10− 6.6⋅10− 7.0⋅10−
Outputs (recycled into process)
Water (clean)* 1.61 1.55 2.13 1.25 1.30 1.26 1.26

Note: *water in each process area are covered by the clean water obtained from wastewater treatment area. More detail in Section 3.3.

8
M. Morales et al. Bioresource Technology 328 (2021) 124833

Table 3
Comparison of technology, biomass requirement and energy yield reported in literature.
Ref. (Laser et al., (Laser et al., (Laser et al., (Karlsson (Luo et al., (Karlsson (Cardona and (Cardona and (Morales
2009) 2009) 2009) et al., 2014) 2009) et al., 2014) Sánchez, 2006) Sánchez, 2006) et al., 2017)

BIOMASS Switchgrass Switchgrass Switchgrass Wheat straw Corn Softwood Harwood chips Harwood chips E. globulus
stover forest residues
LHV feed (MJ⋅kg 16.7 17.2 17.2 – – – 17.2 17.2 –
biomass− 1)

Technology
Pretreatment DA AFEX AFEX DA DA Steam + SO2 DA DA DA
Hydrolysis & SSF + C5F CBP CBP SSF SSCoF SSF SH + C6F + SSCoF SH + C6F
Fermentation C5F
Bioethanol recovery TCD + MS IHOSR IHOSR N.S. TCD + MS N.S. TCD + MS TCD + MS TCD + MS
Cogeneration Rankine Rankine GTCC Rankine Rankine Rankine N.I. N.I. Rankine

Biomass
requirements
kg DM 0.18* 0.13* 0.13* 0.13 0.13* 0.19 0.19* 0.15* 0.11*
biomass⋅MJ− 1
bioethanol
MJ biomass⋅MJ− 1 3.08* 2.30* 2.30* – 2.19 – 3.25* 2.62* –
bioethanol

Energy yields
L bioethanol⋅kg DM 0.318 0.44 0.44 – 0.35* – 0.25 0.31 0.422*
biomass− 1
MJ bioethanol⋅kg 5.43* 7.48* 7.48* 7.88 7.46 5.25 5.29* 6.56* 8.93*
DM biomass− 1
kg bioethanol⋅kg 0.203 0.279 0.279 0.294 0.278* 0.196 0.198* 0.245* 0.333
DM biomass− 1

*Calculated considering bioethanol: LHV: 26.8, Density: 0.79 kg/l.

DA: Dilute acid pretreatment.
AFEX: Ammonia fiber expansion.
SSF: Simultaneous saccharification and hexose fermentation.
SH: Separated hydrolysis.
C5F: Separated pentose fermentation.
C6F: Separated hexose fermentation.
SSCoF: Simultaneous saccharification and hexose and pentose cofermentation.
CBP: Consolidated Bioprocess: cellulase production, cellulose hydrolysis, hexose fermentation and pentose fermentation in a single process.
TCD: Two columns distillation.
MS: Molecular sieves.
IHOSR: Distillation in a single column with direct steam injection and an intermediate heat pump with optimal side stream return.
GTCC: Power via gas turbine combined cycle production.
N.S.: Not specified.
N.I.: Not included.

end for eucalyptus and wheat straw. Eucalyptus, in our study requires conventional technology based on Rankine power cycle. By comparison,
0.14 kg⋅MJ− 1 bioethanol versus 0.11 kg⋅MJ− 1 bioethanol reported in another study found that the surplus electricity for bioethanol produced
(Morales et al., 2017), even if the technology used in both bioethanol from switchgrass is about 0.01–0.04 kW⋅MJ− 1 bioethanol when
plants is similar. Wheat straw requires 0.14 kg⋅MJ− 1 bioethanol in our considering a Rankine cogeneration system, but this increased to 0.07
study, which is similar to 0.13 kg⋅MJ− 1 bioethanol found in (Karlsson kW⋅MJ− 1 bioethanol when using GTCC (Gas turbine combined cycle)
et al., 2014) with different technology assumptions. While in our study power technology (Laser et al., 2009).
the saccharification and fermentation steps are separated (separated Table 4 shows the energy efficiency for the different process steps
saccharification and co-fermentation, SSCoF), (Karlsson et al., 2014) and lignocellulosic feedstocks. The internal input of steam and elec­
analyze simultaneous saccharification and fermentation (SSCoF), which tricity used in the plant represents 26% (24–30%) of feedstock LHV on
is a more advanced technology option. However, in the present study average. In comparison, a previous study reported process steam and
conventional technologies were prioritized above more optimal or power requirements of 24.1% of feedstock (switchgrass) LHV (Laser
advanced options. SHCoF allows us to gain a clearer idea of the effects of et al., 2009); however, the authors significantly reduced the energetic
biomass type in each stage of the process separately. requirement in the process to 16% of feedstock by using a GTCC) instead
of a Rankine cogeneration system. Rankine power was selected as
3.2. Energy balance cogeneration technology in our study because it is the most well-
understood and studied technology for bioethanol plants. In the case
Surplus electricity is co-produced in the bioethanol plant (see of energy sent to cooling duty and ambient losses, the current results are
Table 2) in magnitudes that are very similar for the different biomass in the range of 25%-34% of feedstock LHV, while the ambient losses are
feedstocks (0.02 kWh⋅MJ− 1 bioethanol or 0.07 MJ⋅MJ− 1 bioethanol), lower than 17% of feedstock LHV (Table 4). Similar values have been
except for spruce residues, for which the surplus electricity is two times reported for a corn stover-based bioethanol plant in Sheehan et al.
higher (0.04 kWh⋅MJ− 1 bioethanol or 0.14 MJ⋅MJ− 1 bioethanol). This is (2003); here, cooling tower energy and ambient losses each represented
mainly explained by the high lignin content of the spruce residues, 34% about 22% of feedstock LHV (Sheehan et al., 2003).
in comparison to 24% on average for the other feedstocks. This residual The bioethanol efficiency ranges between 27.3% of feedstock LHV
lignin after saccharification is burned in the boiler to generate electricity (0.24 L bioethanol⋅kg feedstock− 1) for spruce and 46.2% of feedstock
and/or steam. The cogeneration process in our study considers a LHV (0.36 L bioethanol⋅kg− 1 feedstock) for switchgrass. These

9
M. Morales et al. Bioresource Technology 328 (2021) 124833

Table 4
Energy efficiency (% of feedstock LHV) for the bioethanol production plant among the different feedstocks.
% feed LHV Eucalyptus gl. Birch sp. N. Spruce Switchgrass Miscanthus Corn stover Wheat straw

External inputs
Feedstock
kg DM basis/h 49895.0 66654.7 70364.8 102348.8 108745.6 108745.6 115142.4
MJ/h 858194.6 1266438.4 1301748.8 1709224.9 1957420.8 1718180.4 1934392.3

Internal inputs (supplied by Cogeneration area)

Steam
Pretereatment 9.6% 6.5% 6.4% 4.8% 4.2% 4.8% 4.3%
Bioethanol recovery 12.0% 12.9% 16.5% 17.1% 15.0% 14.9% 13.5%
WWT 3.2% 0.3% 2.9% 1.7% 2.7% 2.2% 2.5%
Subtotal (% feed LHV) 24.9% 19.8% 25.7% 23.6% 21.9% 21.9% 20.3%
Power
Pretreatment 5.5⋅10− 3% 4.6⋅10− 3% 5.9⋅10− 3% 7.5⋅10− 3% 6.6⋅10− 3% 7.6⋅10− 3% 6.6⋅10− 3%
Saccharification 3.8⋅10− 3% 2.5⋅10− 3% 3.3⋅10− 3% 3.2⋅10− 3% 2.8⋅10− 3% 3.1⋅10− 3% 2.5⋅10− 3%
Fermentation 2.4⋅10− 2% 1.7⋅10− 2% 2.0⋅10− 2% 2.2⋅10− 2% 1.9⋅10− 2% 2.1⋅10− 2% 1.7⋅10− 2%
Cogeneration 1.4% 1.4% 1.6% 1.2% 1.2% 1.2% 1.2%
WWT 0.5% 1.6% 1.0% 0.4% 1.1% 0.7% 1.4%
Enzyme production 1.4% 0.8% 1.0% 1.2% 1.1% 1.1% 0.9%
Yeast production 0.5% 0.3% 0.3% 0.2% 0.2% 0.2% 0.2%
Subtotal (% feed LHV) 3.8% 4.1% 4.0% 3.2% 3.6% 3.3% 3.8%
Total internal inputs 28.7% 23.9% 29.7% 26.8% 25.5% 25.3% 24.1%

Outputs
Bioethanol
kg/h 13175.6 14497.9 13217.9 29352.4 28942.1 28418.9 26721.9
kg/kg DM feedstock 0.26 0.22 0.19 0.29 0.27 0.26 0.23
MJ/kg DM feedstock 7.1 5.9 5.1 7.7 7.2 7.0 6.2
MJ/MJ feedstock 0.41 0.31 0.27 0.46 0.40 0.44 0.37
L/kg DM feedstock 0.33 0.28 0.24 0.36 0.34 0.33 0.29
% feed LHV 41.3% 30.8% 27.3% 46.2% 39.8% 44.5% 37.2%
Net Power
MJ/h 12074.3 16570.6 33808.0 35129.3 33064.5 32145.0 24410.1
MJ/MJ Bioethanol 0.078 0.079 0.143 0.076 0.075 0.073 0.064
% feed LHV 1.4% 1.3% 2.6% 2.1% 1.7% 1.9% 1.3%
Bioethanol þ Net Power 42.7% 32.1% 29.9% 48.3% 41.5% 46.4% 38.4%
Cooling duty
% feed LHV 26.1% 27.2% 34.2% 24.9% 31.6% 27.8% 26.9%
Ambient losses
% feed LHV 2.4% 16.8% 6.2% – 1.4% 0.6% 10.6%
Total Output 71.3% 76.1% 70.3% 73.2% 74.5% 74.7% 75.9%

Note: Only energy required and produced in the bioethanol plant is included. Energy efficiency in storage and chipping preprocessing is not included. Indirect energy
flows, such as transport of feedstocks to the plant and energy required in the chemicals production are not included.

differences in the bioethanol efficiency are mainly related to the sugar bioethanol yields, reported as MJ, kilogram and liter of bioethanol by kg
content in the feedstock. Results for bioethanol efficiencies have been DM-biomass, and MJ bioethanol by MJ biomass. These quantities
presented in other studies for switchgrass (Laser et al., 2009) and corn represent the overall process performance and can be used as compar­
stover (Sheehan et al., 2003) with similar technology choices as in our ison purposes between the different types of biomass feedstocks or
process design (i.e. dilute acid pretreatment, hexose and pentose production technologies. Grassy biomasses have the highest ethanol
fermentation, conventional distillation and Rankine cogeneration). The yields, followed by agricultural residues. As previously discussed, these
bioethanol efficiency in these studies are lower for switchgrass (40.4% differences in the bioethanol yields are mainly related to the sugar
of feedstock LHV) (Laser et al., 2009) and higher in the case of corn content in the feedstock. For comparative purposes, Table 3 shows
stover (48.7% of feedstock LHV) (Sheehan et al., 2003), in comparison overall process yields that have been reported for lignocellulosic bio­
to our results for switchgrass and corn stover, respectively. ethanol in other studies. It can be seen that the current bioethanol yields
The net energy efficiency, defined as the ratio of products energy out are in the same scale as the values reported in literature.
(bioethanol and net power) to the feedstock energy, ranges between
30% and 48% of feedstock LHV, for spruce and switchgrass, respec­
tively. The net energy efficiency reported in other studies for corn stover 3.3. Water balance
and switchgrass are 53.2% (Sheehan et al., 2003), and 43.3% (Laser
et al., 2009), using process technologies similar to our study. However, The water balance can inform about the water requirements of the
the net energy efficiency is highly dependent on the technology used in bioethanol plant and is instrumental for assessing the water footprint of
the bioethanol plant as reported by (Laser et al., 2009), who finds an biofuel production systems. Fig. 3 shows the mass of water required in
increase of 18% on net energy efficiency by changing DA pretreatment each process step for each feedstock, measured per MJ of bioethanol
to AFEX pretreatment, and an additional 6.7% when using a GTCC produced. Inoculum production and enzyme production are not
system to generate electricity and steam instead of Rankine included in the water balance. Negative values represent water re­
cogeneration. quirements, while positive values refer to clean water from the waste­
Other important results from this analysis are reported in terms of water treatment area (WWT), available for other uses. In this bioethanol
mass of bioethanol produced (kg⋅h− 1) and feedstock used (kg DM plant all the dirty process water from the different process areas were
basis⋅h− 1) in Table 4. These quantities are arising from the bioethanol collected and sent to the WWT. The technology for WWT consists of
flow computed by Aspen Plus. In addition, Table 4 shows overall anaerobic digestion, aerobic digestion, ultrafiltration membranes and
evaporators trains. This technological process sequence achieves a deep

10
M. Morales et al. Bioresource Technology 328 (2021) 124833

Fig. 3. Water balance analysis for the bioethanol plant for different biomass feedstocks.

cleaning and purification of the recycled water. The high quality of the process performance is the carbon ratio (C-bioethanol per C-biomass).
water allows its reuse, i.e. it can be reintegrated into the process, Fig. 4(a) summarizes the carbon balance for the bioethanol conversion
avoiding the requirement of large volumes of water. The net water facility, defined as C-kmol of bioethanol per C-kmol of biomass. The
requirement is shown in red bars in Fig. 3, and for the case of Eucalyptus main carbon inlet is the feedstock; other minor carbon inputs are
globulus shows a positive surplus value (0.02 kg water for 1 MJ bio­ enzyme, yeast and CSL, which are provided to the facility and represent
ethanol produced). In all the other cases there is a net demand for water, a negligible fractions of total carbon inputs (<0.01C-flux⋅C-biomass− 1).
ranging between 0.04 (birch residues) and 0.18 (wheat straw) kg water Carbon leaves the facility as combustion exhaust and ash from cogene­
to produce 1 MJ. Saccharification area is the biggest demanding of water ration, scrubber and molecular sieves vents from bioethanol recovery
in the bioethanol plant, representing 42% (0.66 kg H2O⋅MJ− 1 bio­ area and aerobic gases from WWT. Supplementary information details
ethanol) of the total water required, followed by the pretreatment (28% the values for carbon ratio from inputs and outputs flows. There is a 1%
on average, 0.43 kg H2O⋅MJ− 1 bioethanol) and cogeneration area (22% of the carbon that leaves the facility that was not possible to track. This is
on average, 0.35 kg H2O⋅MJ− 1 bioethanol), while the fermentation and due to uncertainty in the account of carbon in the wastewater treatment
bioethanol recovery areas only represent 6% (0.09 kg H2O⋅MJ− 1 bio­ of the facility.
ethanol) and 3% (0.04 kg H2O⋅MJ− 1 bioethanol) on average, respec­ Carbon efficiencies for bioethanol range between 27 and 31% for
tively. Around 22% of the total water used in the woody biomasses is agricultural residues, grassy biomass and eucalyptus, while they are
demanded by the pretreatment, which increases to 32% of the water lower for forest residues (19% for spruce and 22% for birch) (Fig. 4(a)).
required in the grassy biomasses and agricultural residues. Even though The main reason for the low carbon ratios for forest residues is the low
in most of the feedstocks exist a water requirement, there is a saving of glucan and xylan contents of the forest residues, meaning that more
this input flow due to the reuse of water from the WWT, which is around carbon is in the lignin that is burned instead of being converted to
93% of the water necessary to produce 1 MJ of bioethanol. The positive bioethanol.
value for eucalyptus represents a surplus of clean water into the process, Approximately one-third of the carbon content in the biomass goes to
which can be used in other auxiliary processes not included in the water bioethanol in the case of grassy and agricultural residues feedstocks,
balance, such as inoculum production and/or enzyme production. The which is similar to values reported in a previous study for bioethanol
water required to produce 1 kg of enzyme and 1 kg of fermenting bac­ with corn stover as feedstock (0.34C-bioethanol⋅C-biomass− 1) (Sheehan
teria are 0.82 kg and 30.7 kg, respectively (more detail in Supplemen­ et al., 2003). Most of the carbon is converted to carbon dioxide (and to a
tary material). smaller extent to carbon monoxide) in the combustion exhaust from the
cogeneration area. Combustion exhaust ranges between 50 and 65% of
the original carbon content of the biomass feedstock. The rest of the
3.4. Carbon balance carbon goes to ashes (1–5%), scrubber vent (10–16%), aerobic gases
(0.7–4%) and a negligible fraction of carbon is emitted in the molecular
An important topic that contribute to the understanding of the

11
M. Morales et al. Bioresource Technology 328 (2021) 124833

Fig. 4. (a) Carbon balance of the bioethanol plant for different biomass feedstocks. (b) Main CO2 emission streams from the bioethanol plant with potential to
be captured.

12
M. Morales et al. Bioresource Technology 328 (2021) 124833

sieves vent (between 0.00063% and 0.4%). These large streams of car­ potential for capture occur for switchgrass and corn stover (0.15 kg CO2
bon released as CO2 offer promising opportunity for carbon capture and MJ− 1 bioethanol), which have the highest carbon efficiency for bio­
storage (or utilization) (Bello et al., 2020; Hepburn et al., 2019) to ethanol (31%).
improve climate mitigation benefits, especially for forest residues
(Cherubini et al., 2016). Three main C-emission streams have potential 3.5. Yield analysis
to be captured and stored: combustion exhaust (from cogeneration),
scrubber vents (from bioethanol recovery area) and aerobic gases (from The yield for each process step (pretreatment, saccharification and
wastewater treatment area). These streams contain only biogenic CO2, fermentation) is a measure of the efficiency of each operation. It can
since no fossil source was considered in the process. The flows (on provide useful information related to sugar consumption and products
average for all feedstocks evaluated) are 580 ton⋅h− 1, 150 ton⋅h− 1 and obtained for each step. Any losses of fermentable sugar as unused
25 ton⋅h− 1 for combustion exhaust, scrubber vent and aerobic gases, degradation product decrease the overall bioethanol yield. In these
respectively. These emissions are mainly composed by CO2: 98.6%, cases, the term “yield” is used to describe the extent of various chemical
99.1% and 99.8% of CO2 for aerobic gases, combustion exhaust and and biochemical reactions, representing a percentage of the theoretical.
scrubber vent, respectively. Hence, the BECCS applied to capture the For example, “pretreatment yield” represents the mass ratio of the sugar
main fraction of carbon emitted in the bioethanol production platform contained in the pretreated hydrolysate over the theorical sugar con­
has the potential to deliver negative emission of CO2, i.e. biogenic CO2 tained in the feedstock. Calculations for pretreatment, saccharification
uptake exceeds the biogenic CO2 emissions, through a physical removal and fermentation yields are detailed in Supplementary material.
of CO2 from the atmosphere. The yield calculations are presented in Fig. 5 separately for pre­
Fig. 4(b) shows the three main CO2 rich streams for each feedstock treatment, hydrolysis and fermentation processes. According to our re­
evaluated. Per unit of bioethanol produced, forest residues have the sults, similar pretreatment yields (from 35% to 38%) are obtained for
highest CO2 emissions with potential for CCS, 0.23 and 0.29 kg CO2 grassy biomass, agricultural residues and birch. In the case of spruce
MJ− 1 bioethanol for birch and spruce residues, respectively. As previ­ residues and Eucalyptus globulus, the pretreatment yields decrease to 24
ously discussed, forest residues have a low carbon ratio for bioethanol, and 27%, respectively. The pretreatment represents a pre-hydrolysis of
which is mirrored by a high carbon fraction lost in other outputs, e.g. the cellulose and hemicellulose, releasing some lignin and soluble
combustion exhaust or scrubber vents. Spruce residues have the largest sugars, among which the C5 (xylose and arabinan) are the main sugars
benefits from CCS. In this case, 84% of the CO2 comes from combustion released. Hence, the low pretreatment yields obtained for spruce and
exhausts, 13% from scrubber vents and 3% from aerobic gases. Spruce eucalyptus can be explained by their low xylose and arabinan content,
has high combustion exhausts because of its high lignin content (34.1%) which leads to a low sugar released in the pretreated hydrolysate with
that is directly sent to cogeneration area. The lowest CO2 emissions with respect to the total theorical sugar content. However, this low

Fig. 5. Bioethanol yields for the bioethanol plant for the different biomass feedstocks.

13
M. Morales et al. Bioresource Technology 328 (2021) 124833

performance in the pretreatment for eucalyptus is counteracted in the Appendix A. Supplementary data
saccharification stage, where a high amount of glucose is released due to
its high glucan content. Supplementary data to this article can be found online at https://doi.
Saccharification yields are 81.1% for all feedstocks evaluated, except org/10.1016/j.biortech.2021.124833.
for miscanthus (82.9%). These yields represent the mass of glucose
released during the enzymatic hydrolysis by mass of theorical glucose References
content in the glucan recovered in the pretreatment hydrolysate, (both
values, mass of glucose release in enzymatic hydrolysis and glucan Anu, Kumar, A., Rapoport, A., Kunze, G., Kumar, S., Singh, D., Singh, B., 2020.
Multifarious pretreatment strategies for the lignocellulosic substrates for the
recovered in pretreatment, are obtained from the process simulation, see generation of renewable and sustainable biofuels: a review. Renew. Energy 160,
Supplementary material). Glucose is obtained directly from glucan 1228–1252.
conversion, through cellobiose conversion, and losses as gluco-oligomers Zhu, J., Pan, X., 2010. Woody biomass pretreatment for cellulosic ethanol production:
technology and energy consumption evaluation. Bioresour. Technol. 101 (13),
(see Supplementary material). 4992–5002. https://doi.org/10.1016/j.biortech.2009.11.007.
Fermentation yields represent the mass of bioethanol produced Barros, S. 2019. Biofuel Annual Brazil 2019. USDA Foreign Agricultutal Service.
relative to the theorical mass of bioethanol content in the feedstock, Bello, S., Galán-Martín, Á., Feijoo, G., Moreira, M.T., Guillén-Gosálbez, G., 2020. BECCS
based on bioethanol from wood residues: potential towards a carbon-negative
which is depending of glucose and xylose content. The fermentation transport and side-effects. Appl. Energy 279, 115884.
yields range between 64.7% and 67.9%, except for birch biomass, where Bondesson, P., Galbe, M., 2016. Process design of SSCF for ethanol production from
the yield is 60.5%. As the amount of bioethanol obtained from the steam-pretreated, acetic-acid-impregnated wheat straw. Biotechnol. Biofuels 9 (1),
222. https://doi.org/10.1186/s13068-016-0635-6.
fermentation reactor is similar across the feedstocks (from 13.4 to 14.7
Cardona, C., Sánchez, O., 2006. Energy consumption analysis of integrated flowsheets for
t⋅h− 1), the difference for the fermentation yield can be explained by the production of fuel ethanol from lignocellulosic biomass. Energy 31 (13), 2447–2459.
sugar (glucose and xylose) content in the feedstock, mainly in the xylose https://doi.org/10.1016/j.energy.2005.10.020.
content. This is equal to 21.5% for birch and only 5.1% and 13.7% for Catapano, F., Di Iorio, S., Magno, A., Sementa, P., Vaglieco, B.M., 2015. A comprehensive
analysis of the effect of ethanol, methane and methane-hydrogen blend on the
spruce and eucalyptus, respectively. Xylose content affects the expected combustion process in a PFI (port fuel injection) engine. Energy 88, 101–110.
(theorical) bioethanol produced. About 95% and 85% of glucose and Cavalett, O., Cherubini, F., 2018. Contribution of jet fuel from forest residues to multiple
xylose are converted to bioethanol in the fermenter reactor (Supple­ sustainable development goals. Nat. Sustain. 1 (12), 799–807.
Chandel, A.K., Garlapati, V.K., Singh, A.K., Antunes, F.A.F., da Silva, S.S., 2018. The path
mentary Material), meaning that 5% of glucose and 15% of xylose are forward for lignocellulose biorefineries: bottlenecks, solutions, and perspective on
unconverted in the fermentation and treated as sugar losses. However, commercialization. Bioresour. Technol. 264, 370–381.
sugar losses that occur in the pretreatment process are largest, indicating Cherubini, F., Huijbregts, M., Kindermann, G., Van Zelm, R., Van Der Velde, M.,
Stadler, K., Strømman, A.H., 2016. Global spatially explicit CO2 emission metrics for
that it is an essential process step to obtain highest bioethanol yields forest bioenergy. Sci. Rep. 6, 20186.
through improvements in the degradation of the lignocellulosic matrix Chrysikou, L.P., Bezergianni, S., Kiparissides, C., 2018. Environmental analysis of a
into its individual components. lignocellulosic-based biorefinery producing bioethanol and high-added value
chemicals. Sustain. Energy Technol. Assess. 28, 103–109.
Conde-Mejia, C., Jimenez-Gutierrez, A., El-Halwagi, M., 2012. A comparison of
4. Conclusions pretreatment methods for bioethanol production from lignocellulosic materials.
Process Saf. Environ. Prot. 90 (3), 189–202.
Daylan, B., Ciliz, N., 2016. Life cycle assessment and environmental life cycle costing
A detailed process simulation of bioethanol production from analysis of lignocellulosic bioethanol as an alternative transportation fuel. Renew.
different lignocellulosic feedstocks shows how efficiencies, water re­ Energy 89, 578–587.
quirements, and mass/energy/carbon balances vary with the feedstock. Eisentraut, A. 2010. Sustainable Production of Second-Generation Biofuels: Potential and
perspectives in major economies and developing countries. IEA Renewable Energy
Overall energetic efficiency varies considerably (from 27.3% to 46.2%) Papers, 2010/01.
and even within classes of feedstocks (e.g. 39.8% for miscanthus and Frankó, B., Galbe, M., Wallberg,, O., 2016. Bioethanol production from forestry residues:
46.2% for switchgrass grasses). Feedstock carbohydrate content is the A comparative techno-economic analysis. Appl. Energy 184, 727–736. https://doi.
org/10.1016/j.apenergy.2016.11.011.
main determining factor for energetic efficiency, with switchgrass pre­ García-Velásquez, Carlos A., Cardona, Carlos A., 2019. Comparison of the biochemical
senting both the highest carbohydrate content and efficiency, and spruce and thermochemical routes for bioenergy production: a techno-economic (TEA),
the lowest carbohydrate content and efficiency. With the dilute acid energetic and environmental assessment. Energy 172, 232–242.
Gaurav, N., Sivasankari, S., Kiran, G.S., Ninawe, A., Selvin, J., 2017. Utilization of
pretreatment method chosen, the pretreatment phase appears key in
bioresources for sustainable biofuels: a review. Renew. Sustain. Energy Rev. 73,
achieving high overall efficiency. 205–214.
Gerbrandt, K., Chu, P.L., Simmonds, A., Mullins, K.A., MacLean, H.L., Griffin, W.M.,
Saville, B.A., 2016. Life cycle assessment of lignocellulosic ethanol: a review of key
CRediT authorship contribution statement factors and methods affecting calculated GHG emissions and energy use. Curr. Opin.
Biotechnol. 38, 63–70.
Marjorie Morales: Conceptualization, Methodology, Software, González-García, S., Moreira, M.T., Feijoo, G., 2010. Comparative environmental
performance of lignocellulosic ethanol from different feedstocks. Renew. Sustain.
Validation, Writing - original draft, Visualization. Anders Francesco Energy Rev. 14 (7), 2077–2085.
Arvesen Cherubini: Conceptualization, Writing - review & editing, Hepburn, C., Adlen, E., Beddington, J., Carter, E.A., Fuss, S., Mac Dowell, N., Minx, J.C.,
Project administration, Funding acquisition, Supervision, Conceptuali­ Smith, P., Williams, C.K., 2019. The technological and economic prospects for CO2
utilization and removal. Nature 575 (7781), 87–97.
zation, Writing - review & editing, Project administration, Funding
Humbird, D., Davis, R., Tao, L., Kinchin, C., Hsu, D., Aden, A., Schoen, P., Lukas, J.,
acquisition, Supervision. Olthof, B., Worley, M., 2011. Process design and economics for biochemical
conversion of lignocellulosic biomass to ethanol: dilute-acid pretreatment and
enzymatic hydrolysis of corn stover. Natl. Renew. Energy Labor.
Declaration of Competing Interest IEA. 2020. Renewable Energy Market Update: Outlook for 2020 and 2021.
IEA. 2019. Renewables 2019: Analysis and forecast to 2024.
Karlsson, H., Börjesson, P., Hansson, P.-A., Ahlgren, S., 2014. Ethanol production in
The authors declare that they have no known competing financial biorefineries using lignocellulosic feedstock – GHG performance, energy balance and
interests or personal relationships that could have appeared to influence implications of life cycle calculation methodology. J. Cleaner Prod. 83, 420–427.
Kudakasseril Kurian, J., Raveendran Nair, G., Hussain, A., Vijaya Raghavan, G.S., 2013.
the work reported in this paper.
Feedstocks, logistics and pre-treatment processes for sustainable lignocellulosic
biorefineries: a comprehensive review. Renew. Sustain. Energy Rev. 25, 205–219.
Acknowledgements Laser, M., Jin, H., Jayawardhana, K., Lynd, L.R., 2009. Coproduction of ethanol and
power from switchgrass. Biofuels Bioproducts Biorefining. 3 (2), 195–218.
Lehtonen, A., Mäkipää, R., Heikkinen, J., Sievänen, R., Liski, J., 2004. Biomass expansion
This study was supported by the Research Council of Norway factors (BEFs) for Scots pine, Norway spruce and birch according to stand age for
through the projects Bio4Fuels (no. 257622), BEST (no. 288047) and boreal forests. Forest Ecol. Manage. 188 (1–3), 211–224.
BioPath (no. 294534).

14
M. Morales et al. Bioresource Technology 328 (2021) 124833

Li, C., Aston, J.E., Lacey, J.A., Thompson, V.S., Thompson, D.N., 2016. Impact of Sheehan, J., Aden, A., Paustian, K., Killian, K., Brenner, J., Walsh, M., Nelson, R., 2003.
feedstock quality and variation on biochemical and thermochemical conversion. Energy and environmental aspects of using corn stover for fuel ethanol. J. Ind. Ecol.
Renew. Sustain. Energy Rev. 65, 525–536. 7 (3–4), 117–146.
Luo, L., Van der Voet, E., Huppes, G., 2009. An energy analysis of ethanol from cellulosic Shukla, P., Skea, J., Calvo Buendia, E., Masson-Delmotte, V., Pörtner, H., Roberts, D.,
feedstock–Corn stover. Renew. Sustain. Energy Rev. 13 (8), 2003–2011. https://doi. Zhai, P., Slade, R., Connors, S., Van Diemen, R., Ferrat, M., Haughey, E., Luz, S.,
org/10.1016/j.rser.2009.01.016. Neogi, S., Pathak, M., Petzold, J., Portugal Pereira, J., Vyas, P., Huntley, E., Kissick,
Maurya, D.P., Singla, A., Negi, S., 2015. An overview of key pretreatment processes for K., Belkacemi, M., Malley, J. 2019. Climate Change and Land: An IPCC special report
biological conversion of lignocellulosic biomass to bioethanol. 3 Biotech 5 (5), on climate change, desertification, land degradation, sustainable land management,
597–609. food security, and greenhouse gas fluxes in terrestrial ecosystems. IPCC, In press.
Morales, M., Aroca, G., Rubilar, R., Acuña, E., Mola-Yudego, B., González-García, S., Slade, R., Bauen, A., Gross, R., 2014. Global bioenergy resources. Nat. Clim. Change 4
2015. Cradle-to-gate life cycle assessment of Eucalyptus globulus short rotation (2), 99–105.
plantations in Chile. J. Cleaner Prod. 99, 239–249. Su, T., Zhao, D., Khodadadi, M., Len, C., 2020. Lignocellulosic biomass for bioethanol:
Morales, M., Quintero, J., Aroca, G., 2017. Environmental assessment of the production recent advances, technology trends, and barriers to industrial development. Curr.
and addition of bioethanol produced from Eucalyptus globulus to gasoline in Chile. Opin. Green Sustain. Chem. 24, 56–60.
Int. J. Life Cycle Assess. 22 (4), 525–536. Talebnia, F., Karakashev, D., Angelidaki, I., 2010. Production of bioethanol from wheat
Nurmi, J., 1997. Heating values of mature trees. Acta Forestalia Fennica 256, 1–25. straw: An overview on pretreatment, hydrolysis and fermentation. Bioresour.
OECD, 2019. Global mitigation potential of biofuels in the transport sector. Enhancing Technol. 101 (13), 4744–4753.
Climate Change Mitigation through Agriculture. OECD Publishing, Paris. Tomás-Pejó, E., Alvira, P., Ballesteros, M., Negro, M.J. 2011. Chapter 7 - Pretreatment
Padella, M., O’Connell, A., Prussi, M., 2019. What is still limiting the deployment of Technologies for Lignocellulose-to-Bioethanol Conversion. in: Biofuels, (Eds.) A.
cellulosic ethanol? Analysis of the current status of the sector. Appl. Sci. 9 (21), Pandey, C. Larroche, S.C. Ricke, C.-G. Dussap, E. Gnansounou, Academic Press.
4523. Amsterdam, pp. 149-176.
Räisänen, T., Athanassiadis, D., 2013. Basic chemical composition of the biomass Tye, Y.Y., Lee, K.T., Wan Abdullah, W.N., Leh, C.P., 2016. The world availability of non-
components of Pine, Spruce and Birch. Biofuel Region. wood lignocellulosic biomass for the production of cellulosic ethanol and potential
RFA, 2019. 2019 Ethanol Industry Outlook: Powered with Renewed Energy. Renewable pretreatments for the enhancement of enzymatic saccharification. Renew. Sustain.
Fuels Association. Energy Rev. 60, 155–172.
RFA, 2020. Annual Fuel Ethanol Production. Renewable Fuels Association. Vieira, S., Barros, M.V., Sydney, A.C.N., Piekarski, C.M., de Francisco, A.C.,
Robak, K., Balcerek, M., 2020. Current state-of-the-art in ethanol production from Vandenberghe, L.P.D.S., Sydney, E.B., 2020. Sustainability of sugarcane
lignocellulosic feedstocks. Microbiol. Res. 240, 126534. lignocellulosic biomass pretreatment for the production of bioethanol. Bioresour.
Saini, Jitendra Kumar, Saini, Reetu, Tewari, Lakshmi, 2015. Lignocellulosic agriculture Technol. 299, 122635.
wastes as biomass feedstocks for second-generation bioethanol production: concepts Wang, L., Littlewood, J., Murphy, R.J., 2013. Environmental sustainability of bioethanol
and recent developments. 3 Biotech 5 (4), 337–353. production from wheat straw in the UK. Renew. Sustain. Energy Rev. 28, 715–725.
Scordia, D., Cosentino, S., Jeffries, T., 2013. Effectiveness of dilute oxalic acid Zabed, H., Sahu, J.N., Suely, A., Boyce, A.N., Faruq, G., 2017. Bioethanol production
pretreatment of Miscanthus× giganteus biomass for ethanol production. Biomass from renewable sources: current perspectives and technological progress. Renew.
Bioenergy 59, 540–548. https://doi.org/10.1016/j.biombioe.2013.09.011. Sustain. Energy Rev. 71, 475–501.
Sharma, B., Larroche, C., Dussap, C.-G., 2020. Comprehensive assessment of 2G
bioethanol production. Bioresour. Technol. 313, 123630.

15