(23e) Integrated Biorefinery Design Under Multi-Scale Uncertainties
AIChE Annual Meeting
Sunday, November 7, 2021 - 4:30pm to 4:45pm
Many parameters used in the biorefinery design come with considerable uncertainties. For instance, experimental yields and separation efficiencies are often utilized for industrial-scale plant design. However, there are inevitably subject to intrinsic variations and inaccurate measurement. Moreover, the performance of scaled-up processes may differ from its bench-scale counterpart.7 The feedstock supply and composition provide another source of uncertainties to be considered in process design. Meanwhile, the prices of chemicals are volatile and unpredictable, affecting the plant's profitability.8 Incorporating uncertainties in optimization has gained growing attention over recent years.9 Currently, the number of uncertain parameters considered in the process design is limited by computational complexity. Two commonly used optimization methods under uncertainty are robust optimization and stochastic programming. Stochastic programming uses probability distributions of uncertain parameters to build the scenario tree and calculate the expected values as objective functions. Proper sampling and scenario reduction techniques are crucial to its implementation.10 When ensuring process feasibility is the top priority, or only ranges of parameters are known, the robust optimization problem should be adopted to provide a conservative process design that is feasible even in the worst case.11
In practice, uncertainties from various sources may have different impacts on the process outcome.12 In the multistage optimizations under uncertainty, wait-and-see decisions are taken after the uncertaintyâs realization, which are captured by the affine decision rule in the adaptive robust optimization.13 On the other hand, decisions may affect the endogenous uncertaintyâs distribution or the time it realizes.14 Furthermore, uncertainties associated with process operation or product quality need to be treated with more care to ensure the process's feasibility in all scenarios. Other uncertainties, including market prices, are less restricted, for which optimizing expected profit is more appropriate.12 Yue and You also formulated a stochastic robust optimization that distinguishes uncertainties associated with strategic and operational decisions.7
In this work, we consider a unified framework to include different levels of uncertainties in the biorefinery design. Various feedstocks, including lignocellulosic biomass, food waste and technical lignin, introduce the flexibility for supply but also composition uncertainty. Based on experimental data, processing block flowsheets are developed in Aspen Plus to synthesize a variety of chemicals from biomass.15 The input, output, operating and capital cost information of each block is extracted and scaled to formulate the optimization problemâs objective functions and constraints. Both the profitability and environmental benefits are main driven forces for biorefinery development. Thus, a multi-objective optimization problem is formulated to simultaneously minimize the greenhouse gas emissions based on life cycle assessment (LCA) and maximize profit. We first identify the primary sources of uncertainties and classify them based on their main effects of various decision making including process design, operation and LCA. The LCA calculation is independent of the previous stages, meaning that its uncertainty analysis could be performed after finalizing the plant configuration. Its uncertainty provides a reference for the accuracy of our objective values obtained from the optimization problem. Next, uncertainties on the prices and demands are considered in the strategic decisions that maximize the expected profit, while those associated with feedstock and operating conditions are more restrictive so that the feasibility is satisfied throughout the entire uncertainty range. This biorefinery design problem is solved by a combination of robust optimization and stochastic programming techniques.
1. Athaley, A.; Annam, P.; Saha, B.; Ierapetritou, M., Techno-economic and life cycle analysis of different types of hydrolysis process for the production of p-Xylene. Comput. Chem. Eng. 2019, 121, 685-695.
2. Alonso, D. M.; Hakim, S. H.; Zhou, S.; Won, W.; Hosseinaei, O.; Tao, J.; Garcia-Negron, V.; Motagamwala, A. H.; Mellmer, M. A.; Huang, K.; Houtman, C. J.; Labbé, N.; Harper, D. P.; Maravelias, C. T.; Runge, T.; Dumesic, J. A., Increasing the revenue from lignocellulosic biomass: Maximizing feedstock utilization. Science Advances 2017, 3 (5), e1603301.
3. Athaley, A.; Saha, B.; Ierapetritou, M., Biomass-based chemical production using techno-economic and life cycle analysis. AIChE J. 2019, 65 (9), e16660.
4. Liao, Y.; Koelewijn, S.-F.; Van den Bossche, G.; Van Aelst, J.; Van den Bosch, S.; Renders, T.; Navare, K.; NicolaÃ¯, T.; Van Aelst, K.; Maesen, M.; Matsushima, H.; Thevelein, J.; Van Acker, K.; Lagrain, B.; Verboekend, D.; Sels, B. F., A sustainable wood biorefinery for lowâcarbon footprint chemicals production. Science 2020, eaau1567.
5. Yuan, Z.; Eden, M. R., Superstructure optimization of integrated fast pyrolysis-gasification for production of liquid fuels and propylene. AIChE J. 2016, 62 (9), 3155-3176.
6. Gong, J.; You, F., Optimal Design and Synthesis of Algal Biorefinery Processes for Biological Carbon Sequestration and Utilization with Zero Direct Greenhouse Gas Emissions: MINLP Model and Global Optimization Algorithm. Ind. Eng. Chem. Res. 2014, 53 (4), 1563-1579.
7. Yue, D.; You, F., Optimal supply chain design and operations under multi-scale uncertainties: Nested stochastic robust optimization modeling framework and solution algorithm. AIChE J. 2016, 62 (9), 3041-3055.
8. Noh, N. M.; Bahar, A.; Zainuddin, Z. M., Scenario Based Two-Stage Stochastic Programming Approach for the Midterm Production Planning of Oil Refinery. MATEMATIKA: Malaysian Journal of Industrial and Applied Mathematics 2018, 34 (3), 45-55.
9. Sahinidis, N. V., Optimization under uncertainty: state-of-the-art and opportunities. Comput. Chem. Eng. 2004, 28 (6), 971-983.
10. Martín, M.; Martínez, A., Addressing Uncertainty in Formulated Products and Process Design. Ind. Eng. Chem. Res. 2015, 54 (22), 5990-6001.
11. Grossmann, I. E.; Apap, R. M.; Calfa, B. A.; Garcia-Herreros, P.; Zhang, Q., Mathematical Programming Techniques for Optimization under Uncertainty and Their Application in Process Systems Engineering. Theor. Found. Chem. Eng. 2017, 51 (6), 893-909.
12. Chen, Y.; Lin, M.; Jiang, H.; Yuan, Z.; Chen, B., Optimal design and operation of refinery hydrogen systems under multi-scale uncertainties. Comput. Chem. Eng. 2020, 138, 106822.
13. Zhao, L.; Ning, C.; You, F., Operational optimization of industrial steam systems under uncertainty using data-Driven adaptive robust optimization. AIChE J. 2018, 65 (7), e16500.
14. Zhang, Q.; Feng, W., A unified framework for adjustable robust optimization with endogenous uncertainty. AIChE J. 2020, 66 (12), e17047.
15. Ibarra-Gonzalez, P.; Rong, B.-G., Integrated Methodology for the Optimal Synthesis of Lignocellulosic Biomass-to-Liquid Fuel Production Processes: 1. Simulation-Based Superstructure Synthesis and Development. Ind. Eng. Chem. Res. 2020.