Elsevier

Journal of Cleaner Production

Volume 182, 1 May 2018, Pages 313-330
Journal of Cleaner Production

Addressing global environmental impacts including land use change in life cycle optimization: Studies on biofuels

https://doi.org/10.1016/j.jclepro.2018.02.012Get rights and content

Highlights

  • Computable general equilibrium models and life cycle optimization are integrated.

  • Land use change is considered for first time in life cycle optimization.

  • Novel modeling framework is applied to global bioethanol production.

  • Results inform global cleaner production strategies and policy.

  • Proposed integrated framework can analyze variety of global production systems.

Abstract

Life cycle environmental impacts of a product or process may be global and/or spatially-explicit, such as land use change (LUC) and LUC greenhouse gas (GHG) emissions. Life cycle optimization (LCO) usually does not account for these impacts. However, for a product or process to be truly sustainable, they must be considered. We integrate computable general equilibrium (CGE)-based LUC modeling and LCO to create a novel multiobjective CGE-LUC-LCO framework to account for global environmental impacts and production costs. The framework is then applied to case studies on life cycle GHG emissions throughout the bioethanol life cycle, considering regional and global agricultural practices, land use, technological impacts, and global economic forces. The framework considers emissions from feedstock production, transportation, direct/indirect processing emissions, end use, and LUC. The model allows for selection of 16 bioethanol production pathways from 5 feedstocks, and 3 case studies with US and EU bioethanol demands are examined. The methodology identifies cleaner production strategies by considering global, spatially-explicit life cycle environmental impacts like LUC and LUC GHG emissions for the first time in LCO.

Introduction

Environmental impacts of a product or process can be global. Greenhouse gas (GHG) emissions impact climate change regardless of where they are produced. Other impacts are spatially-explicit, such as land use change (LUC) induced from corn ethanol production (Dunn et al., 2013), and cellulosic ethanol production (Qin et al., 2016). Life cycle optimization (LCO) integrates life cycle assessment (LCA) with mathematical optimization to identify more sustainable production strategies by minimizing life cycle impacts (You and Wang, 2011). However, few LCO studies holistically consider global impacts like LUC and LUC GHG emissions (Garcia and You, 2015). To the best of our knowledge, no LCO studies utilize detailed global models like computable general equilibrium (CGE) models from economics to study such impacts. While some LCO studies have looked at land use footprints (Čuček et al., 2012), land use intensity (Mata et al., 2011) or consider specific land use cases, such as open ponds for algae processing networks (Gong and You, 2014), LUC is not treated with detailed large-scale models in LCO. This knowledge gap is partly due to complexities of global land use modeling, often based on complicated, nonlinear, nonconvex, and large-scale global economic equilibrium models such as CGE models (Taheripour et al., 2007). first introduced liquid biofuels into the well-known Global Trade Analysis Project (GTAP) data base and CGE model, later expanded by (Taheripour et al., 2011) to include cellulosic biofuels. Other researchers have used these tools to estimate LUC GHG emissions of certain biofuel production scenarios (Qin et al., 2016) and understand the impact of uncertainties in the model and data (Plevin et al., 2015). If LCO is to find truly sustainable alternatives, impacts like LUC and LUC GHG emissions as part of overall life cycle GHG emissions should be accounted for with appropriate detail. Doing so increases the detail in life cycle GHG emissions accounting and allows for specific analysis of LUC in LCO models.

LCAs determine the environmental impacts of some product or process (ISO, 1997). A typical LCA has four stages: goal and scope definition, life cycle inventory analysis, life cycle impact assessment, and interpretation. Functional units and system boundaries are defined in goal and scope definition. The functional unit should quantitatively capture the product's or process's function. For example, the functional unit for LCA of biofuel production might be the energy produced, such as “1 gasoline-equivalent gallon.” (Yue et al., 2013) Next, the detailed input-output flows to or from the environment are tallied in the life cycle inventory analysis. The environmental impacts of these flows are calculated in the life cycle impact assessment stage. Finally, results are interpreted, and ideas for improved environmental performance might be identified.

LCO merges LCA with mathematical optimization to systematically identify the most sustainable production strategies by minimizing some life cycle impact, such as GHG emissions (You and Wang, 2011). Some studies minimized GHG emissions from algal biodiesel production (Gutierrez-Arriaga et al., 2014) and value-added chemicals from algae (Gong and You, 2015b), while others minimized CO2 emissions for integrated biomass-fossil fuel processes (He et al., 2014), minimized water consumption in the biofuels life cycle (Garcia and You, 2015a), considered social indicators (You et al., 2012), hedged against uncertainty (Gao and You, 2017), or used the eco-indicator-99 as an environmental objective (Santibanez-Aguilar et al., 2011). Many works optimized the bioenergy life cycle to minimize GHG emissions at the process and supply chain scales (Yue et al., 2014b), including optimizing a UK biofuel supply chain (Akgul et al., 2012), optimization of heat-integrated biorefinery's supply network (Čuček et al., 2014), a large-scale bioconversion process and product network (Garcia and You, 2015b), or bioelectricity supply chains (Yue et al., 2014a). However, LCO studies do not usually consider global environmental impacts, like LUC. The few that do focus on land area converted (Mata et al., 2011) or location of facilities (Yue and You, 2014). Without integrating detailed, land use modeling methodology with LCO, LUC cannot be satisfactorily considered. Furthermore, typical LCO approaches are attributional, focusing on the process/processes of interest (Gong and You, 2015a). However, changes in output from the studied process affect external markets, causing further environmental impacts and indirect LUC.

Consequential LCA could handle these impacts (Reinhard and Zah, 2009), but often no more than two levels of markets and one process are investigated (Tonini et al., 2012). (Weidema et al., 2018) make a strong argument for consequential LCA over attributional LCA when analyzing the impacts of supply chains and/or value chains. Gong and You first combined consequential LCA with LCO and applied a consequential LCO framework to renewable algal diesel production (Gong and You, 2017). However, their work was based on a tailored, partial equilibrium model that could not consider global impacts like LUC. A key novelty of this work is to expand consequential LCO to the global level and consider impacts like LUC.

LUC results when one type of land (e.g. pastureland) is converted to another (e.g. cropland), driven by socioeconomic changes and/or climate forcings. LUC is associated with undesirable environmental impacts, such as loss of biodiversity (de Baan et al., 2013) and GHG emissions (Plevin et al., 2014). Thus, LUC is an important midpoint indicator in LCA (Teixeira et al., 2016). LUC has received significant attention over recent years in LCA research (Schmidt et al., 2015) as well as LCA-based studies on biofuels including the EPA RFS2 program (Sissine, 2010), the DOE's billion ton study (Langholtz et al., 2016), the California Air Resources Board's (CARB) Low Carbon Fuel Standard (CARB, 2011). The topic has been studied in numerous research works, including a work on LUC implications for algae production (Handler et al., 2017), LUC GHG emissions estimations for oil palm expansion in Thailand for biodiesel (Permpool et al., 2016), and analyzing Thai bioenergy policy (Prapaspongsa and Gheewala, 2016). Biofuels and bioenergy production induce direct (Hertel et al., 2010) and indirect LUC (Fajardy and Mac Dowell, 2017), and care must be taken to model both (Taheripour and Tyner, 2013). Direct LUC refers to land changed for production of the biomass feedstocks. Indirect LUC can describe LUC induced by changes in the global economy from increased biofuels production, capturing LUC induced by changes in land values across the globe(Lapola et al., 2010). While the concept of LUC is intuitive, quantitatively modeling LUC and its impacts is challenging(De Rosa et al., 2016). As a result, it is difficult to plan feedstock production while considering potential environmental impacts derived from induced LUC around the globe.

Computable general equilibrium (CGE) models from economics model international trade and how it changes with changes in the economy (Hertel, 1997). Equilibrium in this work refers is economic equilibrium: economic equilibrium occurs when, for all markets of goods, capital, labor, and land, supply equals demand. CGE models are complex, highly nonlinear economic models that aim to describe the global economy and how each economic sector in each country is influenced by every other sector in every other country. Partial equilibrium models, such as the Global Change Assessment Model (GCAM) (Edmonds et al., 1997), also utilize economic equilibrium to study global impacts of trade (Kim et al., 2006), but general equilibrium models allow for broader studies. CGEs are often used in the broadest of consequential LCAs, and are considered to be more accurate when analyzing impacts felt over medium-term timescales – such as LUC (Vázquez-Rowe et al., 2013). The Global Trade Assessment Project (GTAP) CGE model is commonly used to estimate LUC (Taheripour and Tyner, 2013) and includes detailed country-level economic data (Hertel, 1997). The GTAP model has been used to analyze economic issues ranging from EU bioenergy policies (Dandres et al., 2012), national foreign policies (Engelbert et al., 2014), and trade liberalization of forestry products (Stenberg and Siriwardana, 2015) to climate change (Nijkamp et al., 2005), water scarcity (Berrittella et al., 2007), and outbreaks of livestock diseases (Boisvert et al., 2012). An important drawback of CGE models is their coarse resolution (Igos et al., 2015). However, inputs/outputs that are often aggregated, such as land use, can be seamlessly integrated into CGE models to analyze global impacts with a breadth that other methods cannot.

Many LUC studies use CGEs to model LUC, as CGE models can accommodate land use and land valuation by treating land as a factor of production (Plevin and Mishra, 2015). Changes in demand for different types of land by different economic sectors drive LUC in CGE-based LUC modeling. CGEs can be used to model LUC driven by changes in land values resulting from pre-defined shifts in the global economy. LUC causes GHG emissions from burning biomass, foregone CO2 sequestration of cleared forests, and loss of soil organic carbon (SOC). GHG emissions estimates from LUC induced by biofuels production for corn ethanol produced in the USA range from 2 g CO2-eq/MJ (Qin et al., 2016) to 55 g CO2-eq/MJ (Plevin et al., 2015). Dunn et al. used the CENTURY SOC model and data from the US Forest Service to estimate LUC GHG emissions at 7.6 g CO2-eq/MJ (Dunn et al., 2013). A follow-up study found a range of 2.1–9.3 g CO2-eq/MJ (Qin et al., 2016). Plevin et al. used an IPCC dataset and GHG emissions factors to estimate a range of 13–55 g CO2-eq/MJ (Plevin et al., 2015). However, estimating LUC GHG emissions from pre-defined bioethanol production scenarios cannot automatically identify alternative scenarios where different regions producing bioethanol using different technologies could be more sustainable, as noted by different LUC impacts when producing bioethanol in the US compared to China (Yan et al., 2010). The DOE's Billion-Ton Report identifies where in the USA to sustainably grow feedstocks, but the study does not address feedstock sourcing in other regions or countries (Langholtz et al., 2016). Few have sought to identify optimal global feedstock sourcing and bioethanol production strategies that minimize global environmental impacts, but identifying such strategies can help improve the environmental performance of bioethanol, making it an even more attractive alternative liquid fuel.

Global environmental impacts including LUC are often characterized with methods based on CGE models that cannot inherently identify ways to mitigate these impacts without user input and trial-and-error model runs to determine which production strategy has the best global environmental performance. LCO identifies sustainable production strategies but typically does not consider global, spatially-explicit impacts and does not consider impacts from supply and demand shifts in external markets. The novel, integrated CGE-LUC-LCO modeling framework proposed in this work bridges the gap between these techniques by simultaneously considering LUC and LUC GHG emissions with other life cycle impacts in LCO (Fig. 1). The framework is then applied to case studies on global bioethanol production.

Primary novelties of this work include:

  • A new CGE-LUC-LCO modeling framework that allows for detailed optimization of a product/process's global environmental impacts including LUC;

  • Application of this framework to consider global life cycle GHG emissions including those from LUC of bioethanol production for the first time in LCO;

  • New methodology to identify sustainable production decisions in the context of the global economy.

The remainder of this article is as follows. The next section poses a general problem statement and propose a general CGE-LUC-LCO model. Next, a specific multiobjective CGE-LUC-LCO model for global bioethanol production is formulated. The framework is then applied to case studies on EU and US bioethanol demand, followed by a discussion of the results.

Section snippets

General problem statement

A primary goal of this work is to construct an LCO modeling framework that directly considers life cycle impacts induced by changes in the external markets and economy by increased output of the studied production process (es), like LUC and LUC GHG emissions. Broadly, when production of a particular product changes in a certain region, supply and demand for the inputs and outputs to the production process change in response. If this change in production is large enough, regional and global

Specific problem statement and input data

We next apply the framework to global bioethanol production and detail how all components and data sources are hard-linked into one, integrated CGE-LUC-LCO model. Our goal is to identify global feedstock sourcing and bioethanol production strategies with minimum total cost and minimum life cycle GHG emissions including GHG emissions from LUC. Bioethanol demand in the US and EU is considered in three case studies by building a multiobjective framework that considers global, spatially-explicit

Results and discussion

The proposed CGE-LUC-LCO model is now applied to case studies involving US and EU bioethanol demand. First, our model is verified with a “base case” version of the model using only the CGE-based LUC equations with no optimization meant to reproduce previous estimates of LUC GHG emissions induced from US corn bioethanol production. We then show how the full form of our CGE-LUC-LCO model identifies novel corn bioethanol production solutions with improved environmental performance compared to

Conclusion

The CGE-LUC-LCO modeling framework proposed in this work addresses global impacts that LCO cannot while identifying optimal production strategies that CGE-LUC studies cannot. The CGE-LUC-LCO framework was applied to case studies on global bioethanol production, minimizing production cost and life cycle GHG emissions including GHG emissions from LUC. Our proposed approach identified novel global biomass sourcing and bioethanol production strategies that result in lower life cycle GHG emissions

Acknowledgements

F. Y. acknowledges partial financial support from the National Science Foundation (NSF) Award (CBET-1643244). D. G. acknowledges the financial support from the Institute for Sustainability and Energy at Northwestern University (ISEN) and Argonne National Laboratory via a Northwestern-Argonne Early Career Investigator Award for Energy Research.

References (92)

  • D. Garcia et al.

    Supply chain design and optimization: challenges and opportunities

    Comput. Chem. Eng.

    (2015)
  • D.L. Ginnebaugh et al.

    Examining the temperature dependence of ethanol (E85) versus gasoline emissions on air pollution with a largely-explicit chemical mechanism

    Atmos. Environ.

    (2010)
  • J. Gong et al.

    Sustainable design and synthesis of energy systems

    Curr. Opin. Chem. Eng.

    (2015)
  • K. Gubicza et al.

    Techno-economic analysis of ethanol production from sugarcane bagasse using a Liquefaction plus Simultaneous Saccharification and co-Fermentation process

    Bioresour. Technol.

    (2016)
  • R.M. Handler et al.

    Land use change implications for large-scale cultivation of algae feedstocks in the United States Gulf Coast

    J. Clean. Prod.

    (2017)
  • E. Igos et al.

    Combination of equilibrium models and hybrid life cycle-input–output analysis to predict the environmental impacts of energy policy scenarios

    Appl. Energy

    (2015)
  • P. Nijkamp et al.

    Modeling the impacts of international climate change policies in a CGE context: the use of the GTAP-E model

    Econ. Modell.

    (2005)
  • N. Permpool et al.

    Greenhouse gas emissions from land use change due to oil palm expansion in Thailand for biodiesel production

    J. Clean. Prod.

    (2016)
  • P.H. Pfromm et al.

    Bio-butanol vs. bio-ethanol: a technical and economic assessment for corn and switchgrass fermented by yeast or Clostridium acetobutylicum

    Biomass Bioenergy

    (2010)
  • F. Pianosi et al.

    A matlab toolbox for global sensitivity analysis

    Environ. Model. Software

    (2015)
  • T. Prapaspongsa et al.

    Risks of indirect land use impacts and greenhouse gas consequences: an assessment of Thailand's bioethanol policy

    J. Clean. Prod.

    (2016)
  • J. Reinhard et al.

    Global environmental consequences of increased biodiesel consumption in Switzerland: consequential life cycle assessment

    J. Clean. Prod.

    (2009)
  • J.H. Schmidt et al.

    A framework for modelling indirect land use changes in life cycle assessment

    J. Clean. Prod.

    (2015)
  • F. Taheripour et al.

    Biofuels and their by-products: global economic and environmental implications

    Biomass Bioenergy

    (2010)
  • L. Tao et al.

    Process and technoeconomic analysis of leading pretreatment technologies for lignocellulosic ethanol production using switchgrass

    Bioresour. Technol.

    (2011)
  • R.F.M. Teixeira et al.

    Towards consensus on land use impacts on biodiversity in LCA: UNEP/SETAC Life Cycle Initiative preliminary recommendations based on expert contributions

    J. Clean. Prod.

    (2016)
  • B.P. Weidema et al.

    Attributional or consequential Life Cycle Assessment: a matter of social responsibility

    Journal of Cleaner Production

    (2018)
  • D. Yue et al.

    Biomass-to-bioenergy and biofuel supply chain optimization: overview, key issues and challenges

    Comput. Chem. Eng.

    (2014)
  • S. Berry

    Biofuels Policy and the Empirical Inputs to GTAP Models

    (2011)
  • A. Bot et al.

    Land Resource Potential and Constraints at Regional and Country Levels

    (2000)
  • A.K.D. Brooke et al.

    GAMS: a User's Guide: Course Technology

    (1988)
  • R.H. Byrd et al.

    KNITRO: an Integrated Package for Nonlinear Optimization, Large-scale Nonlinear Optimization

    (2006)
  • H. Cai et al.

    Consideration of land use change-induced surface albedo effects in life-cycle analysis of biofuels

    Energy Environ. Sci.

    (2016)
  • CARB
    (2011)
  • L. de Baan et al.

    Land use impacts on biodiversity in LCA: a global approach

    Int. J. Life Cycle Assess.

    (2013)
  • J.B. Dunn et al.

    Land-use change and greenhouse gas emissions from corn and cellulosic ethanol

    Biotechnol. Biofuels

    (2013)
  • J. Edmonds et al.

    An integrated assessment of climate change and the accelerated introduction of advanced energy technologies-an application of MiniCAM 1.0

    Mitig. Adapt. Strategies Glob. Change

    (1997)
  • EIA

    International Energy Statistics: Total Petroleum Consumption

    (2016)
  • EPA

    Emission Factors for Greenhouse Gas Inventories

    (2014)
  • EU

    Directive 2009/28/EC of the European Parliament and of the Council

    (2009)
  • M. Fajardy et al.

    Can BECCS deliver sustainable and resource efficient negative emissions?

    Energy Environ. Sci.

    (2017)
  • A.E. Farrell et al.

    A low-carbon Fuel Standard for California, Part 1: Technical analysis

    (2007)
  • R. Frischknecht et al.

    The ecoinvent database: overview and methodological framework (7 pp)

    Int. J. Life Cycle Assess.

    (2005)
  • D. Garcia et al.

    Life cycle network modeling framework and solution algorithms for systems analysis and optimization of the water-energy nexus

    Processes

    (2015)
  • D. Garcia et al.

    Multiobjective optimization of product and process networks: general modeling framework, efficient global optimization algorithm, and case studies on bioconversion

    AIChE J.

    (2015)
  • D. Garcia et al.

    Network-based life cycle optimization of the net atmospheric CO2-eq ratio (NACR) of fuels and chemicals production from biomass

    ACS Sustain. Chem. Eng.

    (2015)
  • Cited by (32)

    • Environmental impact analysis of biofuels and bioenergy: a globalperspective

      2022, Biofuels and Bioenergy: A Techno-Economic Approach
    • Concentrating photovoltaic/thermal collectors coupled with an anaerobic digestion process: Dynamic simulation and energy and economic analysis

      2021, Journal of Cleaner Production
      Citation Excerpt :

      Several novel treatment systems were proposed in the past few years (Guven et al., 2019), to limit the energy consumption of such process and to address the issue related to the management of the organic fraction of MSW (Organic Waste, OW). In this framework, the use of biogas produced by OW is becoming more and more attractive (Garcia and You, 2018), producing energy vectors (biomethane, biogas, etc.) (Shen et al., 2015) that can be further used to produce heat and power (Salman et al., 2017), for example in Combined Heat and Power systems (Wang et al., 2019). Obviously, biogas is a GHG, with a methane concentration ranging from 45% to 65% (Aguilera and Gutiérrez Ortiz, 2016) (e.g.: 60% in reference (Abdeshahian et al., 2016)) - and heating values up to 28 MJ/m3.

    View all citing articles on Scopus
    View full text