(98h) Integration of Temporal Factors into Spatially Explicit Optimization of Second Generation Biofuels Supply Chains

Motivated by the need to alleviate the heavy dependence on fossil fuels and address the climate change crisis while maintaining energy security, biomass-derived liquid transportation fuels and energy products have received attention as an alternative. Particularly, there has been great interest for the development of biofuels produced from non-food sources, commonly known as second-generation biofuels. These products require ligno-cellulosic feedstock materials which are less land and water intensive, and do not contribute to the food-versus-fuels debate. Ligno-cellulosic crops are great feedstocks for biofuels production due to their low costs, net carbon emission, and ability to be grown on marginal or non-arable lands without displacing food crops [1][2][3]. With the goal of better understanding the potential of this sector, a mixed-integer linear programming (MILP) optimization framework is developed for the design and operation of the supply chain (SC) of ligno-cellulosic biofuels over a given time horizon. This work aims to integrate temporal factors into traditional SC models in order to identify key decision variables regarding the SC network, obtaining not only the lowest cost, but also environmental impact.

A typical biofuels supply chain network design (SCND) problem consists of an underlying network of harvesting sites, plants, distribution centers, and transportation modes and infrastructure linking those entities [4][5]. While the cultivation of biomass closely relates to landscaping decisions and the agricultural sector, the production of fuels depends on the technology and operation of a conversion plant. The distribution network serves to establish connections between facilities and harvesting sites and delivering materials from one place to another. Mathematical optimization models can identify the strategic and operational planning of the entire SC network. Appropriate optimization models for the design and operation of biofuel SC are critical to increasing the economic viability and decreasing the environmental impact of the system [6].

While many optimization approaches have been developed and applied to the design of second-generation biofuels SCs, there are no approaches integrating the spatial and temporal dimensions. Most spatially explicit studies look into the system for one representative year and do not account for changes in, for example, energy demands. Thus, the predicted designs may not be suitable in the future [7][8]. On the other hand, models that account for temporal aspects mainly focus on aggregate capacity and production planning, overlooking the spatial dimension [9].

To address this limitation, we develop an integrated optimization model that considers simultaneously time-varying and spatially explicit aspects, thereby ensuring that the proposed networks are robust to the ever-changing energy market. Facilities can be established at any time point throughout the horizon and once established, they can be expanded to accommodate more production. In addition to the system configuration, time-specific operational decisions, such as harvesting and production planning, are obtained.

The regional area of study includes the eight states in the US Midwest: Ohio, Michigan, Indiana, Illinois, Wisconsin, Missouri, Iowa, and Minnesota. This region is the topic of interest since switchgrass, a cellulosic feedstock, and other native grasses grown on marginal lands in the region can facilitate large amounts of carbon-negative biofuels. Marginal lands are lands less suitable for food production. They also have varying crop yields and sequestration potentials of soil-organic carbon (SOC). The availability and distribution of marginal lands depend significantly on the geographical location and the definition of marginal lands used. In this study, we specifically chose “historically abandoned” lands, which are defined as lands that have been taken out of food production for at least 5 years, and not converted to urban or water use. Crop simulation models are used to attain realistic, field-specific, high-resolution data on biomass yield and soil carbon potential, using land quality, local weather, and management options [10] [11].

In terms of energy demand evolution, the Energy Information Administration (EIA) has come up with 8 different scenarios for the biofuels market [12]. We use the “High Oil Price” projection case. The projection assumes higher global crude oil prices, making biofuels more price competitive with petroleum-based fuels. As a result, a sharp rise in biofuels domestic demands and consumption is predicted. Furthermore, we consider the

greenhouse gasses (GHG) balance of the entire system. We adopt the concept of “social cost of carbon” (SCC) so that both environmental and economic metrics can be considered. The term is defined as the economic cost by an additional ton of carbon dioxide emission according to the Environmental Protection Agency (EPA) and climate scientists [13]. By utilizing SCC.

Our results show that the inclusion of temporal aspects into spatially explicit optimization models is feasible. We show that the optimal solution varies significantly with the weight placed to environmental outcomes. The developed model helps shed light into how the integrated land-crop-logistics-biorefinery systems should be developed over time.

References:

[1] G. P. Robertson, S. K. Hamilton, B. L. Barham, B. E. Dale, R. C. Izaurralde, R. D. Jackson, D. A. Landis, S. M. Swinton, K. D. Thelen, and J. M. Tiedje, “Cellulosic biofuel contributions to a sustainable energy future: Choices and outcomes,” Science, vol. 356, 6 2017.

[2] S. N. Naik, V. V. Goud, P. K. Rout, and A. K. Dalai, “Production of first and second generation biofuels: A comprehensive review,” Renewable and Sustainable Energy Reviews, vol. 14, pp. 578–597, 2 2010.

[3] R. E. Sims, W. Mabee, J. N. Saddler, and M. Taylor, “An overview of second generation biofuel technologies,” Bioresource Technology, vol. 101, pp. 1570–1580, 3 2010.

[4] M. Eskandarpour, P. Dejax, J. Miemczyk, and O. P´eton, “Sustainable supply chain network design: An optimization-oriented review,” Omega, vol. 54, pp. 11–32, 7 2015.

[5] H. Yildiz, J. Yoon, S. Talluri, and W. Ho, “Reliable Supply Chain Network Design,” Decision Sciences, vol. 47, pp. 661–698, 8 2016.

[6] C. Papapostolou, E. Kondili, and J. K. Kaldellis, “Development and implementation of an optimisation model for biofuels supply chain,” Energy, vol. 36, pp. 6019–6026, 10 2011.

[7] R. T. Ng, D. Kurniawan, H.Wang, B. Mariska, W.Wu, and C. T. Maravelias, “Integrated framework for designing spatially explicit biofuel supply chains,” Applied Energy, vol. 216, pp. 116–131, 4 2018.

[8] L. He-Lambert, B. C. English, D. M. Lambert, O. Shylo, J. A. Larson, T. E. Yu, and B. Wilson, “Determining a geographic high resolution supply chain network for a large scale biofuel industry,” Applied Energy, vol. 218, pp. 266–281, 5 2018.

[9] A. Azadeh, H. Vafa Arani, and H. Dashti, “A stochastic programming approach towards optimization of biofuel supply chain,” Energy, vol. 76, pp. 513–525, 11 2014.

[10] B. Maestrini and B. Basso, “Subfield crop yields and temporal stability in thousands of US Midwest fields,” Precision Agriculture, vol. 22, pp. 1749–1767, 2021.

[11] E. G. O’Neill, R. A. Martinez-Feria, B. Basso, and C. T. Maravelias, “Integrated spatially explicit landscape and cellulosic biofuel supply chain optimization under biomass yield uncertainty,” Computers & Chemical Engineering, vol. 160, p. 107724, 4 2022.

[12] EIA, “Annual Energy Outlook 2023.”

[13] EPA, “The Social Cost of Carbon: Estimating the Benefits of Reducing Greenhouse Gas Emissions,” tech. rep., EPA, 2017.