Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/286234884

Life Cycle Assessments of LanzaTech Ethanol Production: Anticipated


Greenhouse Gas Emissions for Cellulosic and Waste Gas Feedstocks

Article in Industrial & Engineering Chemistry Research · December 2015


DOI: 10.1021/acs.iecr.5b03215

CITATIONS READS

101 9,792

5 authors, including:

Robert Handler David R Shonnard


Michigan Technological University Michigan Technological University
77 PUBLICATIONS 2,344 CITATIONS 192 PUBLICATIONS 5,503 CITATIONS

SEE PROFILE SEE PROFILE

Evan Griffing Andrea Lai


Wichita State University 5 PUBLICATIONS 153 CITATIONS
29 PUBLICATIONS 523 CITATIONS
SEE PROFILE
SEE PROFILE

All content following this page was uploaded by Robert Handler on 14 March 2019.

The user has requested enhancement of the downloaded file.


Article

pubs.acs.org/IECR

Life Cycle Assessments of Ethanol Production via Gas Fermentation:


Anticipated Greenhouse Gas Emissions for Cellulosic and Waste Gas
Feedstocks
Robert M. Handler,*,† David R. Shonnard,† Evan M. Griffing,‡ Andrea Lai,§ and Ignasi Palou-Rivera§

Sustainable Futures Institute, Department of Chemical Engineering, Michigan Technological University, Houghton, Michigan 49931,
United States

Environmental Clarity, 2505 Fauquier Lane, Reston, Virginia 20191, United States
§
LanzaTech, 8045 Lamon Avenue Suite 400, Skokie, Illinois 60077, United States

ABSTRACT: LanzaTech has developed novel microbial bioreactor systems capable of direct gas fermentation to produce
ethanol from carbon-containing gases. In this study, a life-cycle assessment method is used to quantify the global warming
potential of several scenarios for producing renewable ethanol with the LanzaTech process. Scenarios considering ethanol
produced from steel mill waste gases or biomass (corn stover, forest residue, or switchgrass, via gasification) have been
considered, using input data from peer-reviewed literature, government reports, life cycle inventory databases, and LanzaTech
process engineering estimates. Using standardized life-cycle assessment methods, ethanol produced via LanzaTech fermentation
appears to result in greenhouse gas emissions that are at least 60% lower than that of conventional fossil gasoline, with biomass-
based ethanol achieving close to 90% emission reductions. Results indicate that the LanzaTech gas fermentation technology can
be a viable alternative for producing next-generation biofuels that satisfy United States Renewable Fuels Standard policies
concerning fuels with a reduced greenhouse gas emissions footprint.

■ INTRODUCTION
The United States continues to depend on petroleum for
such as competition for land and water resources, especially for
irrigated crops, the impact on the price of food; and the
resulting decline in genetic diversity, among others.5−8 Biofuels
transportation fuels, which accounted for roughly 28% of the
relying on nonfood crops, agricultural residues, or other wastes
country’s energy-related CO2 emissions in 2012.1 Roughly one-
would alleviate some of these concerns. An increasing portion
third of petroleum fuels used in the United States are imported,
of the renewable fuels requirement in the United States is due
warranting a continued focus on domestic energy production as
to come from nonstarch sources and qualify as advanced (50%
a means to improve domestic economic and energy security.2
reduction) or cellulosic (60% reduction) biofuels, depending
Several solutions are being developed to deal with the
on their level of greenhouse gas (GHG) emissions reductions
environmental, economic, and social challenges caused by
compared to petroleum sources.9 These “second-generation”
continued use of polluting fossil transportation fuels from
biofuels based on agricultural residues promise to lower GHG
imported and domestic sources. The most widely adopted emissions associated with liquid transportation fuels, but
emissions reduction strategy is the production of alternative ongoing life cycle studies will be required to ensure that
liquid transportation fuels (ethanol and biodiesel, among environmental concerns associated with expansion of conven-
others) made from renewable feedstocks. These fuels are tional agriculture are also mitigated.10
compatible as blendstock with existing transportation infra- One example of an advanced, second-generation biofuel is
structure as minor components of the final fuel mix, generally that produced by LanzaTech. LanzaTech has developed novel
10 or 15% for ethanol, and have benefitted from existing fermentation processes to convert carbon monoxide and
agricultural production systems to generate large quantities of hydrogen-containing gases into valuable fuel and chemical
feedstock. products, including ethanol, 2,3-butanediol, acetic acid,
Ethanol is the predominant alternative liquid transportation isopropanol, acetone, butanol, succinic acid, and isoprene.
fuel. It has successfully been integrated into the national fuel Process inputs can be low-value or waste gases from industries
system and is offered in a range of blend ratios with petroleum such as steel manufacturing, oil refining, and chemical
gasoline. Corn-based ethanol has been touted as a domestic production, as well as gases generated by gasification of forestry
energy success story, with measurable impacts on environ- and agricultural residues, municipal waste, natural gas, and coal.
mental metrics and rural economies.3,4 While corn-based
ethanol currently comprises the large majority of domestic
Special Issue: Sustainable Manufacturing
renewable fuel production, scientists and policy makers
continue to develop plans for transitioning to a new array of Received: September 1, 2015
renewable transportation fuels that improve upon the environ- Revised: December 3, 2015
mental benefits of corn ethanol while reducing concerns about Accepted: December 7, 2015
the environmental impact of large-scale biofuels production,

© XXXX American Chemical Society A DOI: 10.1021/acs.iecr.5b03215


Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Industrial & Engineering Chemistry Research Article

Figure 1. LanzaTech process block diagram. Light gray areas indicate potential coproduct unit operations, which are possible but not included in this
LCA study.

In doing so, waste carbon streams are captured and functionally


recycled into new, useable products. Fermentation products can
■ GOAL AND SCOPE
The goal of this study is to determine overall greenhouse gas
be used directly or thermochemically converted to drop-in fuels emissions from fuel ethanol production using the LanzaTech
and chemical products, such as jet fuel, olefins, and other process and to compare its environmental burdens to the
commodity chemical intermediates. These processes have the petroleum-derived gasoline life cycle GHG emissions. Our
potential to produce fuels meeting the EPA emissions criteria functional unit for this analysis will be one megajoule (MJ,
for next-generation biofuels, while utilizing a diverse range of lower heating value) of final ethanol (EtOH) fuel product,
feedstocks. LanzaTech has successfully demonstrated its gas because energy content is a reasonable predictor of perform-
fermentation technology at a 300 TPA demonstration plant ance in engines and this is a commonly used basis of
with Baosteel in Shanghai, China, and is currently operating a comparison for different fuels.
second demonstration plant with Shougang Steel at Caofeidian, To arrive at a comprehensive GHG emissions assessment of
ethanol production by the LanzaTech process, our analysis
China. These plants can utilize a range of mill gases including
takes a cradle-to-grave approach. A detailed inventory of
basic oxygen furnace (BOF), blast furnace, COREX, and coke processing inputs was developed from LanzaTech’s commercial
oven gas. Both demonstration plants have exceeded nameplate plant design in collaboration with LanzaTech process
capacity and met performance targets. LanzaTech also owns engineers. Production of all required inputs, including
and operates the Freedom Pines Biorefinery in Soperton, chemicals and energy, are included in the analysis, in addition
Georgia. The facility has biomass handling and gasification to transport of ethanol product prior to use. All emissions of
equipment, and pilot-scale testing of biomass-based fuel and gases from the bioreactor and anaerobic digestion of settled
chemical production is expected by the end of 2015. solids are included, assuming that any methane in the emissions
Initial life-cycle assessment (LCA) studies on the environ- is converted to CO2 through flaring. Final combustion of the
mental impact of steel mill waste gas conversion to ethanol in ethanol product is also included.
China have been described in the literature,11 indicating the Scenarios. Four different fuel ethanol production pathways
potential for a low-emission fuel product. This article updates are examined as part of this study. The BOF gas scenario uses
data for steel mill gas conversion to ethanol, based on design carbon-containing BOF gas as the carbon-containing fermenta-
tion feed gas. Three biomass-based scenarios are also included,
improvements from LanzaTech’s scale-up work and application
which use corn stover, switchgrass, and forest residue as
of the technology to a domestic United States market, and carbon-containing feedstocks, and are labeled accordingly. In
presents initial LCA results for ethanol produced from biogenic the biomass-based scenarios, solid biomass feedstock materials
(corn stover, forestry residues, and switchgrass) feedstocks. are converted to syngas via a gasification process, and the
GHG emissions are tabulated for each stage of the ethanol life syngas is used as the fermentation feed gas. The resulting
cycle and are compared to conventional transportation fuels different fermentation feed gases in each scenario are utilized in
and other ethanol production pathways. the LanzaTech process for the production of ethanol.
B DOI: 10.1021/acs.iecr.5b03215
Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Industrial & Engineering Chemistry Research Article

LanzaTech Process Overview. Key steps in the percentage is flared, and this percentage is highly variable by
LanzaTech process are gasification, gas handling, fermentation, region. In the United States, recent surveys indicate that BOF
and product recovery. A graphical depiction of the LanzaTech steel mill exhaust gases are not currently utilized by any United
process is shown in Figure 1. First, in the case of biomass States mills, but they may represent a significant opportunity
feedstocks, gasification produces an input gas and excess heat for heat and fuel recovery, offsetting significant portions of the
is used to generate steam and electricity. Next, in the gas cumulative GHG emissions burden for steelmaking.15 Lanza-
handling step, carbon monoxide-containing gas streams, such Tech is focusing commercial development of this technology
as waste or byproduct gases from industrial processes or on situations where BOF gas is underutilized because these
gasified biomass residues, are deoxygenated and compressed for scenarios present the greatest opportunity for adding value to
use as the primary input to the fermentation. The treated gas is current steel mill operations by turning this low-value gas
fed to a biological reactor for fermentation in which stream with variable composition and flow rates into a
proprietary microorganisms suspended in a liquid nutrient consistent source of high-value fuels and chemicals. For the
solution utilize CO as both a carbon and energy source. If purposes of this study, the prior fate of BOF exhaust gas is
present in the input gas blend, H2 can also be utilized by the assumed to be flaring, where all carbon contained in the BOF
microorganisms as a supplemental energy source. The microbes gas is converted to CO2 and released to the environment. Due
secrete fermentation products such as ethanol, 2-3 butanediol, to this assumption, the input gas stream appears in the cradle-
and acetic acid into the surrounding broth. Fermentation broth to-gate LCA as a GHG emissions credit in the LanzaTech
continuously withdrawn from the bioreactor is sent through a process, as it is being removed from this prior fate of immediate
steadfast distillation-based separation system for product and flaring release to the environment. Quantitative results of this
coproduct recovery. The fermentation can be tailored to assumption are illustrated in Table 2. Emissions of all carbon-
produce a range of valuable chemical coproducts which, if containing gases are counted in the GHG emissions profile of
recovered, processed, and transported to market, would be the fuel at every downstream processing stage. In this scenario,
considered in the overall life cycle as well. For the purposes of the LanzaTech process is envisioned to be an independent
this study, the process under analysis is one in which plant colocated with an existing steel mill. This colocation
production of ethanol is maximized, with minimal coproduct eliminates any need for transporting feedstock gas to the
creation and no coproduct recovery. LanzaTech bioreactor.
Waste streams are minimized and recycled internally within The other three scenarios involve the production and use of
the process as much as possible. Biological solids (spent carbon-containing gases from biomass sources (corn stover,
microbial biomass) are filtered out of the distillation waste switchgrass, and forest residue). Data surrounding the
stream and sent to anaerobic digestion, and a portion of the production of corn stover comes from the GREET model
filtered stream is used to make up fresh media (a water and developed by Argonne National Laboratory16 and includes
nutrient mixture) that is fed continuously to the LanzaTech inputs of materials and energy related to fertilizer use,
microbes. In this way, process water within the system is collection, 50 miles of truck transport, and carbon stock
recycled back to the fermentation bioreactor. The remaining changes associated with direct and indirect land-use change. In
liquid wastes are treated on-site via anaerobic digestion. The this case, the GREET model assumes that using corn stover as a
biogas from anaerobic digestion is mixed with a portion of the feedstock for transportation fuels would lead to small increases
reactor vent gas to increase the gas energy density, enabling the in carbon stocks on the landscape (negative CO2 emission, or
use of the mixed gas for internal steam or power generation. CO2 credit) as corn-growing land becomes more productive
Vented gas from the fermentation bioreactor is scrubbed, and less land is utilized for corn production.17
oxidized, and released to the atmosphere with a GHG Switchgrass production data also comes from the GREET
emissions penalty. Generated utilities are mostly used to offset model16 and includes inputs related to cultivation, harvesting,
utility needs; however, in the biomass-based scenarios involving 50 miles of transport, and any land-use change effects.
biomass gasification, excess electricity is exported to the grid. Switchgrass involves larger assumptions of fertilizer use
Fermentation Gas Generation. The scenario using compared to corn stover and also contains a small CO2
industrial waste gas (BOF gas scenario) is based on steel mill emission factor for land-use change as a result of increased
basic oxygen furnace gas. In 2011, global production of steel switchgrass cultivation.
was 1.5 billion tons/yr. Of this, 1.1 billion tons were from the Life-cycle input data for forest biomass has been incorpo-
blast furnace/BOF route.12 In steel-making, carbon from coal, rated from a detailed study of the Consortium for Research on
natural gas, and/or oil is used to reduce iron ore into iron metal Renewable Industrial Materials (CORRIM) group, which
according to the following reaction: focuses on whole tree thinning operations in the Southeast
Fe2O3 + 3C → Fe + 3CO region of the United States and residue grinding data from the
Pacific Northwest.18 Biomass from forest residues is generally
This reaction takes place in the blast furnace (BF) and produced from thinning operations in managed forests or when
produces molten iron or pig iron with high carbon content nonmerchantable timber (tops and branches) are separated in a
(typically 3.5−4.5%). This iron is then processed in a basic commercial harvest. Unlike the agricultural biomass scenarios,
oxygen furnace. The BOF controls the amount of carbon left in no land-use change effects or additions of fertilizer to increase
the final steel product by blowing pure oxygen over the hot biomass productivity are assumed in this study for forest
metal.13 The oxygen reacts with carbon in the pig iron and biomass. Biomass collection, processing, and 90 miles of truck
carries it away as a carbon-rich gas residue that is 50−60% CO, transport are assumed to occur using machinery typical to the
10−20% CO2, and 20−30% N2. A portion of these gases may increasingly mechanized forest products industry in the specific
be used on-site to generate heat. In Europe, about 25% of all region.
BOF steel mill gases are flared in lieu of being used for heat or Gasification of biomass streams in all cases was modeled
power generation.14 On a global basis, a much higher according to parameters outlined in a National Renewable
C DOI: 10.1021/acs.iecr.5b03215
Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Industrial & Engineering Chemistry Research Article

Energy Laboratory study.19 Gasification of biomass results in a integration within the host site and opportunities to take
stream of hot syngas (CO and H2 gases), which is cooled via advantage of internally produced steam and electricity.
heat integration with downstream unit operations that require Within the context of steel mill operations, there will likely
heat inputs. After cooling, the syngas can be fed into the be an excess of low-quality steam suitable for the LanzaTech
LanzaTech fermentation system. This syngas stream typically process available onsite, from coke oven gases, blast furnace
has a composition of 30−35% CO, 40−45% H2, 10% CO2, and gases, or BOF gases.23 In the BOF gas scenario, biogas from the
other trace gases. Gasification may require a small amount of anaerobic digestion of bioreactor soluble and solid metabolites
input natural gas to heat incoming biomass to gasification and vented fermentation reactor gas are combined and
temperatures as the process begins, but at steady-state the combusted to produce steam, which supplies steam in sufficient
process is self-sustaining, where gasification of complex quantities to satisfy the needs of ethanol distillation and gas
molecules into smaller molecules liberates enough energy to treatment. Required electricity comes from the off-site
heat incoming biomass and maintain the process. electricity grid. The mix of generation sources for the United
Gas Fermentation. The LanzaTech process considered States electricity grid is variable between regions and over time,
here converts carbon-containing input gas into ethanol. The and for this assessment the default United States grid
fermentation process can accommodate a range of input gas assumptions from Ecoinvent life cycle inventory database
compositions and is tolerant of typical gas contaminants such as version 2 were used,24 which includes a combination of coal
sulfur, which minimizes pretreatment requirements. The (49%), nuclear (19.6%), natural gas (17.3%), oil (3.3%), and
LanzaTech microbes utilize CO as both a carbon and energy renewables (9.6% including hydropower) as the primary
source. If present in the input gas blend, H2 can be utilized by sources of electricity generation. This grid mix results in a
the microorganisms as a supplemental energy source. conservative electricity generation profile, with emissions that
In addition to the input carbon as gas, a fermentation media could be improved or worsened based on changing the grid mix
including macro- and micronutrients for the organism is fed to reflect the most current situation in regions of the United
into the bioreactor. The LanzaTech process is a continuous States or other countries, as discussed in Results and
Discussion.
fermentation, meaning that media is continuously fed into the
Biomass gasification produces a stream of hot syngas (∼900
bioreactor while fermentation broth (containing ethanol,
°C).19 The sensible heat from this gas stream can generate
fermentation coproducts, and spent biomass) is removed at
steam, ∼0.5 tons steam per ton syngas, which can be used
an equal rate. Trace amounts of other coproducts, besides
directly in industrial processes or used to make electricity. This
ethanol, are also produced in the fermentation process. The source of utilities is a significant benefit to the biomass-based
ratio of ethanol to coproduct and identity of the coproducts can scenarios, where all utilities can be generated internally with the
be varied substantially by modifications to the process. productive use of sensible heat, including both steam heat
Product Separation. Product separation proceeds through (gasifier, gas treatment, ethanol distillation units) and power
a distillation process that has been modified to suit specific (gasifier, fermentation reactor units). The LanzaTech process is
LanzaTech operating conditions. Distillation energy require- currently modeled assuming conversion efficiencies in a
ments are in good agreement with standard models of the combined heat and power (CHP) system of 85% from syngas
conventional distillation process in an integrated distillation to steam and 35% from steam to electricity. Of the total
sequence followed by dehydration, which require 2.0−2.5 tons amount of steam produced in the biomass-based scenarios from
of steam per ton of ethanol, given the specific titer of the biomass gasification, 30% of the steam is used as process heat,
LanzaTech fermentation broth.20−22 Settled solids, composed and the rest is used to generate power for internal process use
of biomass from the fermentation organism, and other organics and to export as excess electricity. Exported electricity is
from the fermenter are separated from the product stream and assumed to displace the need for electricity from the United
sent to an anaerobic digestion unit. The LanzaTech process States grid and a credit for displaced electricity use is attributed
considered here does not make any attempt to recover the trace to the process. Roughly 20% of the power that is produced is in
amount of chemical coproduct, although modified product excess of internal use and can be exported to the grid in the
recovery units designed to separate ethanol from representative biomass-based scenarios. Electricity requirements for gas
coproducts such as butanediol have been utilized successfully in compression are roughly 50% lower in the biomass feedstock
separate LanzaTech trials. cases because biomass syngas is produced at pressure, while the
Product Transport and Use. Ethanol is assumed to be BOF exhaust gas is assumed to enter the system close to
transported 100 km by truck prior to use, based on prevalence atmospheric pressure.
of steel mill facilities and gasoline blending terminals in the In all scenarios, the combustion of biogas produced through
eastern half of the United States. Ethanol will be used as a anaerobic digestion and bioreactor vent gas is assumed in this
blended fuel component, with CO2 emissions on combustion as study to be used on-site to generate steam and/or electricity,
predicted by stoichiometry (1.91 kg CO2/kg EtOH, 71.4 g using the same CHP efficiency assumptions as stated previously
CO2eq/MJ EtOH). for biomass gasification-derived steam and electricity. This gas-
Utility Consumption. Specific inputs of electricity and based generation also results in a small displacement credit for
steam are withheld here to protect LanzaTech confidentiality, excess electricity in all analyzed scenarios.
but resulting GHG emissions are aggregated together as
“utilities” in the LCA results. Utility inputs for all scenarios are
electricity and medium-pressure steam, though specific require-
■ METHODS AND INVENTORY
Input Data. Inputs for the model LanzaTech facility have
ments are slightly different in the BOF gas scenario versus been tabulated on the basis of 1000 kg of EtOH production,
biomass feedstock scenarios. Though utility inputs are which is then normalized by the ethanol production from the
important sources of environmental impacts, the LanzaTech facility and the energy content per unit of ethanol (26.8 MJ/
process takes advantage of many opportunities for utility kg). Input data is summarized below, along with comments on
D DOI: 10.1021/acs.iecr.5b03215
Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Table 1. Input LCA Data, on the Basis of 1000 kg/h of EtOH Production
requirement item used for LCAa BOF gas scenario corn stover scenario switchgrass scenario forest residue scenario comments
utilities (steam electricity, medium voltage, United States Grid, Steam, for provided to account for generated internally generated internally generated internally generated from cooling hot syngas in biomass
and electricity) chemical processes process requirements scenarios; accounting for internal credit from
biogas produced through anaerobic digestion
Gasification Inputs
biomass input 2870 kg 2870 kg 2670 kg Dry weight basis
process water 2.19 m3 2.19 m3 2.03 m3
Fermentation Inputs
net gas inputb CO2 equivalents (CO2eq) 2.12 × 103 kg 2.15 × 103 kg 2.15 × 103 kg 2.15 × 103 kg Assuming all CO and CH4 converted to CO2 via
flaring
bioreactor nu- calcium chloride, iron(III) chloride, superphosphate, provided to account for provided to account for provided to account for provided to account for stand-ins for chloride salts, metal chlorides, P
trient inputsc ammonia, organic chemicals stoichiometric require- stoichiometric require- stoichiometric require- stoichiometric require- requirement, N requirement, vitamins
ments of bioreactors ments of bioreactors ments of bioreactors ments of bioreactors
cooling water water, decarbonised, at plant 6453 kg 7340 kg 7340 kg 7280 kg makeup water assuming 1% loss due to evaporation
and blowdown
fermentation tap water, at user 2554 kg 1250 kg 1250 kg 1250 kg makeup water due to reaction losses, media outflow
process water with solids
Industrial & Engineering Chemistry Research

sulfatreat magnetite, at plant 1.86 kg 1.86 kg 1.86 kg 1.86 kg stand-in for mixed-metal oxide sorbent (sulfur
removal)
wastewater treat- treatment, sewage whey digestion, class 4 1293 kg 1470 kg 1470 kg 1450 kg stand in for treatment of waste from anaerobic
ment digestion, comparable organic carbon content
treatment of bio- anaerobic treatment of solid, liquid waste (biomass, ethanol, 49.9 kg C converted to 54.8 kg C converted to 54.8 kg C converted to 54.8 kg C converted to LanzaTech internal process model to account for
solids etc.), conversion of biogas to CO2 upon combustion; CO2 and CH4; CO2 and CH4; CO2 and CH4; CO2 and CH4; anaerobic digestion of settled solids, and power
released in anaerobic digestion emissions 183.1 kg CO2eq 200.9 kg CO2eq 200.9 kg CO2eq 200.9 kg CO2eq generation from anaerobic digestion

E
transport of transport, lorry >32 t, EURO5 100 km 100 km 100 km 100 km transport of EtOH 100 km to blender
EtOH
combustion of CO2 equivalents (CO2eq) 1.91 × 103 kg 1.91 × 103 kg 1.91 × 103 kg 1.91 × 103 kg based on stoichiometry
Ethanol
a
Ecoinvent database24 unless otherwise noted. bRefers to BOF gas or gasified biomass input stream, minus bioreactor vent gas emissions. cNutrient inputs are used to create optimal production in the
reactor. These are used to control environmental conditions or used by the microorganism as nutrients.
Article

DOI: 10.1021/acs.iecr.5b03215
Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Industrial & Engineering Chemistry Research Article

the source and structure of the data (Table 1). For the biomass- nature but have a similar impact on the ethanol life cycle. As
based scenarios, slightly less process water is required for the biomass grows it removes carbon from the atmosphere, and the
forest residue scenario compared to the other scenarios because biomass feedstock is given credit for this carbon removal. Based
of differences in feed gas composition resulting from biomass on the assumed carbon content of the biomass, this translates
stoichiometry. Process water and cooling water are differ- to (461 kg C/ton corn stover) × (44 kg CO2/12 kg C) = 1690
entiated in the fermentation inputs (Table 1) because of kg CO2 per ton of corn stover, which appears in the life cycle as
different levels of pretreatment assumed for the water streams, a GHG emissions credit for removing carbon from the
as more dissolved solids are removed from the cooling water atmosphere (Table 2). As in the BOF gas scenario, subsequent
prior to use. Wastewater treatment of liquids from the emissions of carbon throughout the process are accounted for
anaerobic digestor is modeled using a template from the in the life cycle, throughout all stages of the process. In all
Ecoinvent database (sewage whey digestion) that most closely scenarios, a small amount of carbon exits the system boundary
matches the organic carbon content of the LanzaTech in wastewater treated off-site, as well as CO2 from combusted
bioreactor, in order to accurately reflect emissions from the anaerobic digestor biogas, which are both accounted for as well.
wastewater treatment process. Settled solids from the A large majority (>95%) of the carbon entering the system as
LanzaTech fermentation bioreactor are sent to anaerobic BOF gas or biomass exits the system either as carbon in ethanol
digestion, where biogas is produced with a 40:60 composition or combusted vent gas from the bioreactor, similar to current
of CO2:CH4. After biogas conversion, the biogas stream is sugar fermentation reactors. The apparent carbon-use efficiency
combusted to produce energy, and complete conversion of all of the LanzaTech process varies from the BOF scenario to the
CH4 to CO2 is assumed before the biogas is released to the biomass-based cases because of the composition of carbon
atmosphere. Transportation of ethanol is assumed to be entering the reactor, but in both cases between 20 and 40% of
performed using a 32 t truck with updated emissions controls the carbon entering the system as feedstock exits in the ethanol
(class 4 in European system used in Ecoinvent) and a payload product.
of roughly 7 t. Implications of Feedstock Choice. The LanzaTech
Impact Assessment. Environmental impacts of the fermentation process is robust and can use a variety of
production and use of material and energetic inputs listed in carbon-containing gas feedstocks, and we explore the
Table 1 were included as part of this assessment through use of implications of feedstock choice in this paper. In Table 2 we
the Ecoinvent 2.1 database24 or other literature sources as
stated in Table 1. Global warming impacts were calculated Table 2. GHG Credits and Emissions Associated with
using the IPCC 2007 GWP 100a method in SimaPro LCA Feedstock Choice
software, version 7.3.25 In this method, CO2 has a global
corn forest
warming potential (GWP) of 1, CH4 = 25, and N2O = 298. A item BOF gas stover switchgrass residue
full accounting of the GWP of all climate-active chemicals,
carbon content (percent 32.4%a 46.1%b 46.1%b 49.6%b
including solvents and refrigerants used in upstream processes, carbon by weight)
is included in the impact assessmemt method using inventory GHG emissions credit from −1188 −1690 −1690 −1819
elements from the Ecoinvent database in SimaPro 7.3, and feedstock gas incorporation
these impacts are combined into CO2-equivalents (CO2eq) as a into LanzaTech process
(kg CO2eq/ton feedstock)
single metric of greenhouse gas emission impacts. As a GHG emissions due to 0 94.3c 124.8d 59.7e
comparison standard, results are presented alongside equivalent feedstock procurement
environmental impact assessments for conventional gasoline. (kg CO2eq/ton feedstock)
a
Data for petroleum gasoline greenhouse gas emissions is based BOF gas parameters from LanzaTech customer surveys, dry weight
on national inventories from Argonne National Laboratory, basis. bBiomass carbon content from LanzaTech processing estimates.
c
which report average GHG emissions of 94 g CO2eq/MJ.26 Corn stover procurement burdens (including land-use change) from


GREET model estimates.16 dSwitchgrass procurement burdens
RESULTS AND DISCUSSION (including land-use change) from GREET model estimates.16 eForest
residue procurement burdens from CORRIM research study.18
Carbon Accounting. The LanzaTech process model
described above relies on previously accepted process engineer-
ing assumptions for gasification, product recovery, and waste compare the GHG emissions credits for incorporating each
treatment. Carbon mass flows and other material flows are fully feedstock into the LanzaTech process (on the basis of carbon
accounted for in the process model, and emissions of carbon contents) and the GHG emissions associated with supplying
are accounted for in all stages of the process. As mentioned the particular feedstock. There are clear differences among all
previously, the BOF gas scenario assumes that the BOF gas four modeled scenarios on the assumed burdens associated
utilized in this process would have been flared, with oxidation with supplying feedstock to the process. As mentioned above, it
of all carbon to CO2, before being released to the atmosphere. is assumed that LanzaTech operations colocated with steel
The quantitative effect of this assumption is that BOF gas production can acquire and incorporate BOF exhaust gas with
modeled in our process, with a carbon content of 32.4%, would no additional effort compared to current steel mill operations;
be producing a GHG emission penalty of (324 kg C/ton BOF therefore, no GHG emissions are attributed to feedstock
gas) × (44 kg CO2/12 kg C) = 1188 kg CO2/ton BOF gas collection and transportation. In the biomass feedstock
(Table 2). This level of GHG emissions is now credited to the scenarios, there are differences in the GHG emissions burden
feedstock for avoiding flaring emissions and entering the of biomass feedstocks, most notably for switchgrass which has a
LanzaTech process, with subsequent emissions attributed to the larger associated GHG emissions burden due primarily to the
process for carbon emissions that occur throughout the process, assumption of increased fertilizer application and resulting soil
up to and including ethanol combustion. The assumptions N2O emissions.16 It is clear that in all cases, however, that the
regarding biomass-based scenarios are somewhat different in embodied GHG emissions associated with feedstock procure-
F DOI: 10.1021/acs.iecr.5b03215
Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Industrial & Engineering Chemistry Research Article

ment activity are much less than the GHG emission credits biomass-based scenarios, because feedstock GHG emissions
associated with incorporating this renewable or waste carbon in credits and bioreactor vent gas are the two largest flows of
the LanzaTech process, as described in the previous section. carbon in the process and their magnitude far outweighs other
This analysis of the feedstock procurement stage alone does not GHG emissions shown in the life cycle. In the biomass-based
tell the entire story of ethanol produced from these feedstocks, scenarios, biomass feedstocks entering the process are credited
as carbon is released at later stages of the ethanol production for removing carbon from the atmosphere, but subsequent
process. Therefore, a thorough carbon accounting of the full emissions of carbon are accounted for at later stages of the
process was completed to track inputs and emissions of carbon process. Utility consumption (electricity and steam) accounts
from the process at all stages of the life cycle. for a large portion of GHG emissions in the ethanol life cycle
LCA Results. Greenhouse gas emissions data for the entire for the BOF gas scenario, while other inputs to the process
ethanol life cycle are presented in Table 3. In the BOF gas (nutrients, chemicals, water) result in a relatively minor
contribution (Table 3). Production and combustion of biogas
Table 3. GHG Emissions for LanzaTech Scenarios from anaerobic digestion of organic bioreactor waste results in a
GHG emissions (g CO2eq/MJ ethanol)
small amount of electricity that can be used on site, but
combustion of the biogas also results in GHG emissions that
BOF corn forest are accounted for. Accounting for all inputs to the BOF gas-
scenario gas stover switchgrass residue
based life cycle in the BOF gas scenario, we estimate a total
net gas useda −79.0 −80.2 −80.2 −80.2
GHG emissions result for LanzaTech ethanol of 31.4 g CO2eq/
feedstock procurement 0.0 11.2 14.9 6.6
MJ. This emissions total is small enough relative to the life cycle
utilities (heat and power)b 28.8 −6.3 −6.3 −8.1
GHG emissions for petroleum gasoline (67% emissions
other inputsc 1.8 1.3 1.3 1.3
reduction) that this scenario would produce ethanol that
anaerobic digestor 6.8 7.5 7.5 7.5
emissions meets the same emissions reductions targets given to cellulosic
waste treatmentd 0.9 1.1 1.1 1.0 biofuel according to EPA guidelines.9 Compared to the prior
ethanol transport 0.7 0.7 0.7 0.7 LCA study results of the LanzaTech process,11 GHG emissions
ethanol combustion 71.4 71.4 71.4 71.4 for the BOF are lower in this case than in the prior study, 31.4 g
net EtOH GHG emissions 31.4 8.0 11.7 1.5 CO2eq/ MJ in this case compared to 40.7 g CO2eq/ MJ when
% reductione 67% 92% 88% 98% steam generation was provided internally. Several methodo-
a
Net gas used refers to the net GHG emissions of the credits logical differences exist between these two studies to explain the
associated with utilizing different feedstocks, combined with the GHG different LCA results, most notably the reliance of the prior
emissions of vent gas that is released in the fermentation reactor work on electricity provided from the GHG-intensive Chinese
(Figure 1). bIncludes credits for electricity and steam generated electricity grid and reliance on older LanzaTech process models
through combustion of anaerobic digestion biogas. cRefers to with lower assumed ethanol yields.
nutrients, water, chemicals, etc. outlined in Table 1 dIncludes Specific aspects of the life cycle change for the biomass-based
wastewater treatment and solid waste disposal. eCompared to scenarios, but the overall result is quite encouraging. As
petroleum gasoline standard reference defined above (94 g CO2eq/
MJ fuel)
mentioned in Table 2, GHG emissions associated with
feedstock procurement are incorporated into these scenarios,
resulting in emissions ranging from 6.6−14.9 g CO2eq/MJ
scenario, a large credit is granted for removing BOF gas from its ethanol. The burdens associated with feedstock procurement
assumed prior fate, flaring and release into the atmosphere. are more than made up for by the benefits associated with
This credit is in part negated by large emissions observed in the biomass gasification, namely, the on-site coproduction of heat
system from the bioreactor vent gas, but the combination of and power that eliminates utility consumption throughout the
these two items, presented as “net gas used” in Table 3, is still entire process and even results in a small excess of electricity
quite large. This type of carbon accounting is repeated for the which is credited as displacing the use of United States grid

Table 4. Sensitivity Analysis Results for Key Parameters in LanzaTech Process


percent change in overall life cycle GHG emissionsa
BOF gas corn stover
% change % change % change % change
parameter base case low case high case (low) (high) (low) (high)
ethanol transport 100 km 50 km 500 km −1% 9% −4% 35%
distance
electricity grid 214 g CO2eq/MJ elec. b
190 g CO2eq/MJ elec. c
290 g CO2eq/MJ elec. d
−9% 33% 8% −28%
emissions
biogas methane 0% − 2% − 5% − 23%
emissions
utility consumption current 10% reduction 10% increase −9% 9% −20% 49%
biomass procurement current 10% reduction 10% increase − − −14% 14%
impacts
media requirements current 10% reduction 10% increase −1% 1% −2% 2%

a
As compared to 31.4 g CO2eq/MJ (BOF gas) and 8.0 g CO2eq/MJ (Table 3). bCorresponds to United States average medium voltage electricity
grid in Ecoinvent v2 database. cCorresponds to most current U.S. EPA assessment of current generation mix.27 dCorresponds to Chinese electricity
grid GHG emissions as reported in 2011.11

G DOI: 10.1021/acs.iecr.5b03215
Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Industrial & Engineering Chemistry Research Article

electricity. Demands of other inputs including bioreactor according to the low and high case assumptions. The overall
nutrients and process water are similar compared to the BOF LCA result was then recalculated, and the new result was
scenario, reflecting the minor changes associated with different compared to the base case result from Table 3 in terms of
feedstock composition. The resulting life-cycle GHG emissions percent change to the overall GHG emissions. This process was
for ethanol produced from biomass gasification in all biomass- repeated for the BOF gas and corn stover scenarios.
based scenarios are quite low, achieving 88−98% reductions in Importantly, none of the scenarios explored resulted in life
GHG emissions compared to fossil petroleum gasoline. Many cycle GHG emissions that would make the final ethanol
LCA studies exist for ethanol production from biomass product miss qualifying as an advanced biofuel under U.S. EPA
feedstocks, which all rely on different assumptions that make guidelines. However, results of the sensitivity analysis can
direct comparison challenging. Life-cycle GHG emissions provide a few insights on the developments of this process and
results within the well-described GREET model14 for ethanol provide commercial developers with a sense of the important
produced by fermenting corn stover (33 g CO2eq/MJ), factors that should be considered. Excessive transport distance
switchgrass (36 g CO2eq/MJ), and forest residue (15 g for ethanol product can influence the overall GHG emissions
CO2eq/MJ) are all higher than corresponding LCA results for result, but only by ∼5 g CO2eq/MJ in absolute terms. Changes
scenarios described here, suggesting that LanzaTech biomass- in the electricity grid emissions profile had opposite impacts in
based ethanol can achieve a low GHG emissions profile relative the BOF and corn stover scenario, because in the BOF gas
to common ethanol production technologies assumed for these scenario the process is relying on inputs of imported electricity,
feedstocks. while in the corn stover case the excess electricity is being
Additionally, combined process and cooling water con- supplied to the grid, making it more favorable to displace the
sumption within the LanzaTech combined gasification− prior use of electricity from a dirtier grid. Small amounts of
fermentation process for corn stover (0.40 kg/MJ ethanol), fugitive methane emitted from the anaerobic digestor can have
switchgrass (0.40 kg/MJ ethanol), and forest residue (0.39 kg/ an outsized influence on overall GHG emissions due to the
MJ ethanol) are also lower than water consumption estimates relatively large global warming potential of methane compared
at the ethanol production stage reported in the GREET model to CO2, so this process should be designed and maintained with
(0.65, 0.65, and 0.50 kg/MJ ethanol, respectively).14 Water proper attention. Utility consumption had a surprising effect on
consumption within the BOF gas scenario are lower than all the corn stover case, because an increase in utility consumption
biomass-based scenarios, 0.34 kg/MJ ethanol. ultimately reduces the amount of excess electricity that may be
Sensitivity Analysis. To illustrate the extent to which the exported to the grid for a displacement credit in this process.
LCA results presented here are dependent on certain key The goal of this life cycle assessment was to understand the
assumptions, a sensitivity analysis was performed on the life life-cycle greenhouse gas emissions of ethanol produced from
cycle inputs to the BOF gas and corn stover scenarios. Key the LanzaTech process. This analysis was based on estimates
input assumptions in the BOF gas and corn stover scenarios provided by LanzaTech of a model facility utilizing either BOF
were varied to illustrate the impacts of potential variability in exhaust gas or one of several biomass streams as the incoming
the life cycle. The sensitivity analysis results are presented in feedstock. Because input feedstocks were treated as avoided
Table 4. Parameters included in the sensitivity analysis are (1) waste streams (BOF gas) or renewable sources of carbon, large
ethanol transport distance, (2) GHG emissions associated with credits were granted for carbon entering the system boundary.
the electricity grid, (3) fugitive methane emissions from biogas Gasification of biomass into syngas for use in the LanzaTech
production, (4) consumption of utilities, (5) impacts associated process has the added benefit of colocated utility production,
with biomass procurement, and (6) consumption of which results in more favorable LCA results for the biomass
fermentation reactor media inputs. Transport distance was scenarios, despite the small added burdens associated with
reduced to 50 km for the low cases but increased to 500 km in growing and procuring biomass feedstocks.
the high cases to represent long-distance transport from As a result of this analysis, several recommendations can be
relatively remote ethanol facilities. The GHG emissions profile made to guide future study and development of LanzaTech
associated with the United States electricity grid represented in fermentation technology. The large volumes of gas entering and
Ecoinvent is calculated to be 214 g CO2eq/MJ electricity using exiting the bioreactors will require careful monitoring, as these
the IPCC 100a method used in this study. The Low case for two gas streams are significantly larger than all other sources of
this parameter represents the GHG emissions for a recent greenhouse gases in this life cycle analysis. Several assumptions
electricity generation mix using data from the U.S. EPA,27 surrounding biomass procurement have been made, using
which most notably involves a reduction in coal usage (from 49 thorough yet generalized assumptions. As with any biofuels
to 37%) and an increase in natural gas (from 17% to 30%) study, site-specific feedstock studies to verify biomass yields,
usage compared to the Ecoinvent grid used in the base case growth inputs, harvest efficiency, and transport burdens will
scenarios. A high case for this parameter was established by strengthen the current analysis and the overall field of study.
using data relating to the Chinese electricity grid, which uses a Important assumptions have been made surrounding utility
larger proportion of GHG emissions-heavy fuels like coal and generation and use, and deviations from these standard
oil. A high case was created in regards to fugitive methane assumptions could affect the resulting greenhouse gas balance
emissions from biogas production by assuming that 2% of in a negative manner (e.g., using electric power exclusively from
methane produced in anaerobic digestion escapes into the coal) or a positive manner (use of cleaner alternatives such as
atmosphere without being converted to CO 2 during wind or hydropower, or system integration to reduce steam
combustion. For parameters 4−6, the low and high cases use). These variables will likely change from situation to
were created by assuming a 10% reduction or increase in the situation and will require specific LCAs for the different
parameter, respectively, as an illustration of uncertainty in these operating conditions to verify GHG savings. The LanzaTech
areas. For each scenario that was created in the sensitivity process can be tailored to produce different coproducts
analysis, the parameter under consideration was modified alongside ethanol at different ethanol:coproduct ratios, which
H DOI: 10.1021/acs.iecr.5b03215
Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Industrial & Engineering Chemistry Research Article

can be optimized according to market conditions or other (13) Stubbles, J. The Basic Oxygen Steelmaking (BOS) Process. www.
factors. This flexibility will require additional study to steel.org/Making-Steel/How-Its-Made/Processes/Processes-Info/
determine the implications of simultaneous chemical produc- The-Basic-Oxygen-Steelmaking-Process.aspx (accessed March 2015).
tion on the environmental footprint of the LanzaTech process, Steelworks, Online Resource for Steel; American Iron and Steel
Institute.
under different schemes for allocating environmental burdens (14) Boston Consulting Group (BCG) and the Steel Institute VDeh.
between products. The LanzaTech process can enable fuel Steel’s Contribution to a Low-Carbon Europe 2050: Technical and
production with substantially reduced GHG emissions from Economic Analysis of the Sector’s Abatement Potential. http://www.
either BOF exhaust gas from steel mills or gasified biomass, as eurometal.net. 2013.
this study demonstrates. This offers a substantial opportunity to (15) U.S. Environmental Protection Agency. Available and Emerging
both produce low-carbon fuels and reduce global GHG Technologies for Reducing Greenhouse Gas Emissions from the Iron
emissions. and Steel Industry. http://www3.epa.gov/nsr/ghgdocs/ironsteel.pdf.


Office of Air Quality Planning and Standards, Sector Policies and
Programs Division, 2012.
AUTHOR INFORMATION
(16) Argonne National Laboratory. GREET 2014 Life-cycle Model.
Corresponding Author Center for Transportation Research, Energy System Division. 2014.
*Tel.: 906-487-3612. E-mail: rhandler@mtu.edu. greet.es.anl.gov.
(17) Taheripour, F.; Tyner, W. E.; Wang, M. Q. Global Land Use
Notes Changes due to the U.S. Cellulosic Biofuel Program Simulated with
The authors declare no competing financial interest. the GTAP Model. https://greet.es.anl.gov/publication-luc_ethanol.

■ ACKNOWLEDGMENTS
A portion of this work was funded by a 2011 award from the
Argonne National Laboratory, Argonne, IL, 2011.
(18) Johnson, L.; Lippke, B.; Oneil, E. Modeling Biomass Collection
and Woods Processing Life-Cycle Analysis. Forest Products Journal
2012, 62 (4), 258−272.
U.S. Department of Energy to LanzaTech and Michigan (19) Dutta, A.; Talmadge, M.; Hensley, J.; Worley, M.; Dudgeon, D.;
Technological University (Award DE-EE0005-356). Barton, D.; Groenendijk, P.; Ferrari, D.; Stears, B.; Searcy, E. M.;

■ REFERENCES
(1) U.S. Environmental Protection Agency. Sources of Greenhouse
Wright, C. T.; Hess, J. R. Process Design and Economics for Conversion of
Lignocellulosic Biomass to Ethanol: Thermochemical Pathway by Indirect
Gasification and Mixed Alcohol Synthesis; NREL/TP-5100-51400;
National Renewable Energy Laboratory: Golden, CO, 2011; p 187.
Gas Emissions. http://www.epa.gov/climatechange/ghgemissions/ (20) Ingledew, W. M. The Alcohol Textbook; A Reference for the
sources/transportation.html. 2014. Beverage, Fuel and Industrial Alcohol Industries; Ethanol Technology
(2) U.S. Energy Information Administration. Annual Energy Outlook Institute, Nottingham University Press: United Kingdom, 2009.
2015. http://www.eia.gov/forecasts/aeo. U.S. Department of Energy, (21) Cardona, C. A.; Sánchez, O. J. Fuel Ethanol Production: Process
Washington, D.C, 2015. Design Trends and Integration Opportunities. Bioresour. Technol.
(3) Wang, M. Q.; Han, J.; Haq, Z.; Tyner, W. E.; Wu, M.; Elgowainy, 2007, 98 (12), 2415−57.
A. Energy and greenhouse gas emission effects of corn and cellulosic (22) Madson, P. W.; Monceaux, D. A. Fuel ethanol production. In
ethanol with technology improvements and land use changes. Biomass The Alcohol Textbook, 3rd ed.; Jacques, K. A., Lyons, T. P., Kelsall, D.
Bioenergy 2011, 35 (5), 1885−1896. R., Eds.; Nottingham University Press: Nottingham, U.K., 1999;
(4) Swenson, D.; Eathington, L. Determining the regional economic Chapter 17.
values of ethanol production in Iowa considering different levels of local (23) Johnson, I.; William, T.; Choate, W. T.; Davidson, A. Waste
investment. Part A: Developing a modeling and measurement structure; heat recovery: Technology and opportunities in US industry. http://
Iowa State University, Department of Economics: Ames, IA, 2006. www1.eere.energy.gov/manufacturing/intensiveprocesses/pdfs/
(5) Fargione, J. E.; Cooper, T. R.; Flaspohler, D. J.; Hill, J.; Lehman, waste_heat_recovery.pdf. US Department of Energy, Office of Energy
C.; Tilman, D.; McCoy, T.; McLeod, S.; Nelson, E. J.; Oberhauser, K. Efficiency and Renewable Energy, Industrial Technologies Program;
S. Bioenergy and Wildlife: Threats and Opportunities for Grassland Prepared by BCS, Inc., 2008.
Conservation. BioScience 2009, 59 (9), 767−777. (24) Frischknecht R.; Althaus H. J.; Doka, G.; Dones, R.; Heck, T.;
(6) Pimentel, D.; Marklein, A.; Toth, M. A.; Karpoff, M. N.; Paul, G. Hellweg, S.; Hischier, R.; Jungbluth N.; Nemecek, T.; Rebitzer, G.;
S.; McCormack, R.; Kyriazis, J.; Krueger, T. Food Versus Biofuels: Spielmann, M. Overview and Methodology, Final report Ecoinvent, v2.0
Environmental and Economic Costs. Human Ecology. 2009, 37 (1), 1− No. 1. http://www.ecoinvent.ch/. Swiss Centre for Life Cycle
12. Inventories: Duebendorf, Switzerland, 2004.
(7) Searchinger, T.; Heimlich, R. Avoiding Bioenergy Competition for (25) PRé - Project Ecology Consultants. SimaPro, version 7.2. http://
Food Crops and Land.Working Paper, Installment 9 of Creating a www.pre.nl.
Sustainable Food Future. World Resources Institute, Washington, DC. (26) Elgowainy, A.; Han, J.; Cai, H.; Wang, M.; Forman, G. S.;
2015. http://www.worldresourcesreport.org. DiVita, V. B. Energy efficiency and greenhouse gas emission intensity
(8) Solomon, B. D. Biofuels and sustainability. Ann. N. Y. Acad. Sci. of petroleum products at US refineries. Environ. Sci. Technol. 2014, 48
2010, 1185, 119−134. (13), 7612−7624.
(9) U.S. Environmental Protection Agency. Regulation of Fuels and (27) U.S. Environmental Protection Agency. Emissions and
Fuel Additives: Changes to Renewable Fuel Standard Program [EPA- Generation Resource Integrated Database (eGRID), eGRID 2012
HQ-OAR-2005-0161; FRL-8903-1] Fed. Reg. 2009, 74 (99), 24908. Summary Tables. http://www2.epa.gov/energy/egrid.
http://www.epa.gov/otaq/fuels/renewablefuels/regulations.htm.
(10) Lark, T. J.; Salmon, J. M.; Gibbs, H. K. Cropland expansion
outpaces agricultural and biofuel policies in the United States. Environ.
Res. Lett. 2015, 10 (4), 044003.
(11) Ou, X.; Zhang, X.; Zhang, Q.; Zhang, X. Life-cycle analysis of
energy use and greenhouse gas emissions of gas-to-liquid fuel pathway
from steel mill off-gas in China by the LanzaTech process. Frontiers in
Energy 2013, 7 (3), 263−270.
(12) World Steel. Iron Production 2011; World Steel Association,
2012. www.worldsteel.org.

I DOI: 10.1021/acs.iecr.5b03215
Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

View publication stats

You might also like