(120a) A Life Cycle-Based Analysis and Optimization Framework for Designing Sustainable, Multi-Product Biomass-to-Bioproducts Value Chains

In recent years there has been a marked surge in the search for alternative sources of energy that wean the world off of dependence on fossil fuels and reduce our carbon foot-print on this world. After a boom in U.S. corn-based ethanol in the early part of the 21st century, the interest has gradually shifted towards more viable sources of biofuels and biochemicals. Cellulosic ethanol, biodiesel, and syngas are examples of such fuels that are extremely attractive owing to the fact that the raw materials can be composed completely of “left-over” wastes of food crops and forest harvests that don't interfere with the human food chain and the natural ecosystem. Even with increased research emphasis on biofuel and biochemical production technologies, large scale production is still hampered by many factors. Some of these factors that have received considerable interest in recent times include, cultivating more robust feedstock, genetically engineering enzymes and microbes that are more efficient and resistant to poisoning by by-products and improving process technologies that can render large scale production of biofuels a more profitable option for private entities. One area that has not received enough research interest in recent times is value chain modeling and optimization in order to make the entire value chain more competitive in the energy sector. Modeling an enterprise that produces renewable-based fuels and chemicals can provide valuable insight into the inter-play of the bioproduct value chain. This idea had been motivated by a recent surge in the research areas concerning supply-chain optimization and how cost-cutting measures throughout a supply-chain can render an otherwise sluggish enterprise, profitable.

In the current formulation, we apply a distributed, systematic approach to model, optimize, and analyze three aspects of sustainability for a biomass-to-bioproducts value chain; business, environmental, and social sustainability. We assume a vertically integrated bio-enterprise that bears the tasks of contracting feedstock production and conversion, and final product distribution. Life cycle analyses are conducted to generate forecasts for environmental parameters, infrastructural and employment data, input costs and supplies, product yields, and output market conditions, which are integrated with integer-based stochastic framework that yield strategic decisions such as feedstock and supplier selection, transportation logistics design, technology and product selection, waste mitigation design, and capacity design under uncertainty. A two-stage optimization strategy is used to increase computational efficiency while extracting the biggest value drivers for the successful construction and operation of a biorefinery.

The first stage in optimization is termed portfolio design; a deterministic integer programming model is formulated to design the optimal superstructure of feedstock, conversion technologies, and products. Stakeholder value of the entire value chain, including suppliers, customers, capital providers, and the environment, is maximized in the model with a rolling-time horizon of 12 years. The model equations and constraints included are material, energy, capacity, cash, and debt balances, transformation equations, supply and demand constraints, and capacity and resource constraints. Carbon taxes and mitigation credits are used to represent environmental implications of biorefinery activities. The outputs of the model include integer decisions regarding feedstock, suppliers, technologies, and product selection. Furthermore, sensitivity analysis is carried out on the optimal value chain configuration to determine process, supply chain, and market parameters that are the greatest business and environmental value drivers for the biorefinery.

The second stage of optimization is formulated as a stochastic decision tree optimization model. The purpose of the model is to design a timing strategy for feedstock production and pricing, incremental technology acquisition, and capacity design given uncertainty in input costs and supplies, technology costs, and product demands and prices. The model is similar in form and structure to the deterministic MIP from the first stage with stochastic forecasts as inputs. The decision tree nodes are represented by binary integer variables and the tree is generated using a real options methodology. Two types of options are modeled; Sequential growth options to model when to expand capacity, and switching options to represent flexible biorefining systems with the ability to switch product volumes between different products. A binomial decision tree is used to discretize continuous stochastic forecasts for inputs. The final objective function is a probability weighted stakeholder value whose form is derived from the deterministic model. The screened portfolio choices that are obtained from the first-stage optimization are used as inputs. Integer decisions for the model include capacity establishment and expansion, investment in new technologies, and product switching by re-allocation of feedstock to different product groups.

The resultant framework is applied to a prospective biomass refinery in Southeastern United States. The feedstock choices include energy crops, namely, switchgrass, sorghum, and energy cane. The feedstock decisions are affected by contractual costs for land, fixed and operating costs for crop growth and harvest, transportation expenses, environmental impact, and biomass and sugar yields of the crop. The initial product portfolio includes ethanol and butanol as fuels, and succinic and lactic acid and value-added chemicals. The conversion and separation technologies are demarcated into three groups; pretreatment, hydrolysis (steam, dilute acid (DA), ammonia-fiber explosion (AFEX)), and fermentation ( (1) separate C5 and C6 fermentation or (2) co-fermentation) which employ common unit operations for each product, and separation and product recovery (distillation, molecular sieves, pervaporation, membrane separation) which are specific to each product. The product-technology set choices are affected by capital costs, operating costs, process yields, environmental impact, and market potential.

The screening model results yield energy cane as the feedstock choice owing to high biomass yields and ease of processing, and ethanol and succinic acid as the products owing to great market potential and favorable profit margins. Dilute acid pretreatment followed by simultaneous hydrolysis and co-fermentation are selected as the conversion technologies owing to favorable capital cost structure and process economics including the energy and environmental load. Butanol and lactic acid are not chosen primarily due to the high cost of product recovery while distillation and molecular sieves (ethanol) and membrane separation (succinic acid) are chosen for the products that are selected.

The results are then fed to the second stage of optimization with uncertainty being represented in the marginal returns that are generated by capacity and production decisions. The uncertainty in the return implies uncertain revenue and cost streams. The individual components of the uncertain revenue and cost streams include uncertain product prices and uncertain input costs. The uncertain returns are assumed to capture all uncertainty during optimization while the distribution of returns is generated separately using the stochastic forecasts for the individual components. The results indicate an incremental capacity design to mitigate and manage capital and market risks with the overall expected stakeholder value at $MM 3.4 indicating that positive value is being created. The incremental capacity design is compared to economies of scale based capacity design where all capacity is established during the first planning year. The stakeholder value for economies of scale based design is $MM -30.3 indicating a $MM 33.7 advantage to using incremental capacity design. Product switching capabilities are recognized in the final two time periods (4 years) where a price shock to the upside in succinic acid prices is simulated. Capacity is re-allocated to succinic acid production to exploit the favorable market conditions with a 2% increase in the stakeholder value thus depicting the advantages of designing flexible production systems for biomass refineries.