419328 Plantwide Model-Based Optimization of a Large Scale Second Generation Biorefinery

Tuesday, November 10, 2015: 4:55 PM
Salon E (Salt Lake Marriott Downtown at City Creek)
Remus M. Prunescu, Department of Electrical Engineering, Technical University of Denmark, Kgs. Lyngby, Denmark, Mogens Blanke, Electrical Engineering, Technical University of Denmark, Kgs. Lyngby, Denmark, Jon G. Jakobsen, DONG Energy and Gürkan Sin, Department of Chemical and Biochemical Engineering, Technical University of Denmark, Kgs. Lyngby, Denmark

Second generation biorefineries reached commercial reality in 2012 [1] leading to the first commission of a large scale plant in October 2013 by Beta Renewables in Crescentino, Italy [2]. Other commercial size second generation bioethanol plants followed and are already operating or under construction: project Liberty by POET-DSM (USA), Abengoa Bioenergy (USA), DuPont (USA), Måbjerg Energy Concept (Denmark) etc. [3].

Second generation biorefineries transform lignocellulosic agricultural wastes into products with higher added value. In biorefineries employing biochemical routes, this is achieved by the following four major conversion and separation technologies: biomass pretreatment, enzymatic hydrolysis or liquefaction, fermentation and purification [4]. Lignocellulosic biomass consists of cellulose, hemicellulose (xylan and arabinan), lignin, ash and other residues in negligible amount [5]. Cellulose is protected by layers of hemicellulose and lignin, and the scope of the pretreatment phase is to expose the cellulosic fibers by relocating lignin and partially hydrolyzing the hemicellulose. Hemicellulose should only be partially hydrolyzed because most of the pretreatment by-products inhibit the downstream processes. E.g. acetic acid influences the pH of medium, which affects the enzymatic activity [6], furfural is a fermentation inhibitor, and xylose and xylooligomers strongly inhibit the enzymatic activity in the hydrolysis phase [7]. Therefore there is a trade-off between biomass pretreatment and the efficiency of the subsequent enzymatic hydrolysis and fermentation processes.

In the enzymatic hydrolysis phase, enzymes hydrolyze both cellulosic and the remaining hemicellulosic fibers. The conversion route of fibers to sugars is a competitive product inhibition process thoroughly described and analyzed in [8]. Xylose and xylooligomers are the strongest inhibitors of liquefaction, even greater than cellobiose and glucose [7]. Too little biomass pretreatment would increase the amount of hemicellulose for hydrolysis, which would eventually decrease the glucose yield due to xylose and xylooligomers inhibition. On the other hand, too much biomass pretreatment would increase the amount of fermentation inhibitors. Increasing pretreatment temperature is positive for the performance of enzymatic hydrolysis while negative for fermentation yield of ethanol.

Optimal operation depends on pretreatment temperature, enzyme dosage, and yeast seed in fermentation. The first scope of this study is to formulate optimal operation of large-scale biorefineries as a mathematical programming problem. In this formulation, the performance trade-offs between pretreatment, enzymatic hydrolysis and fermentation are formulated as objective functions subject to system dynamics taking into account the above mentioned phenomena using previously developed and validated first principles dynamic models [8, 9, 10]. E.g., too much xylan in the liquefaction phase could be overcome by significantly increasing the enzyme dosage. However, enzymes are very expensive and operation costs would increase. Therefore, price tags need to be attached to raw materials, such as: feedstock, steam for pretreatment, enzymes for hydrolysis, and yeast in the fermentation phase. Economic cost functions are formulated for maximizing the profit of the biorefinery. The final outcome would be to use the economic cost functions in order to find the best possible trade-off between the optimal operation of pretreatment, liquefaction and fermentation processes in integrated manner.

The study also includes a comprehensive sensitivity and uncertainty analysis of the optimization problem with respect to feedstock composition and kinetic parameters. A Monte Carlo technique with Latin Hypercube Sampling and correlation control is used for the uncertainty analysis [11, 12]. Uncertainties in kinetics and yield of pretreatment, hydrolysis and fermentation are found to be negligible on the economic objective function. On the other hand, inflow feed composition influences significantly the objective function.


[1] J. Larsen, M. Ø. Haven, and L. Thirup. “Inbicon makes lignocellulosic ethanol a commercial reality”. In: Biomass and Bioenergy 46 (Nov. 2012), pp. 36–45. DOI: 10.1016/j.biombioe.2012.03.033.

[2] Novozymes A/S. “Beta Renewables opens biofuels plant in Italy”. In: Focus on Catalysts 2013.12 (Dec. 2013), p. 6.DOI: 10.1016/S1351-4180(13)70461-6.

[3] AEC. Cellulosic biofuels. Tech. rep. Advanced Ethanol Council, 2012.

[4] J. Larsen, M. Østergaard Petersen, L. Thirup, H. Wen Li, and F. Krogh Iversen. “The IBUS Process – Lignocellulosic Bioethanol Close to a Commercial Reality”. In: Chemical Engineering & Technology 31.5 (May 2008), pp. 765–772. DOI: 10.1002/ceat.200800048.

[5] J. B. Kristensen, L. G. Thygesen, C. Felby, H. Jørgensen, and T. Elder. “Cell-wall structural changes in wheat straw pretreated for bioethanol production.” In: Biotechnology for biofuels 1.1 (Jan. 2008), p. 5. DOI: 10.1186/1754-6834-1-5.

[6] R. M. Prunescu, M. Blanke, and G. Sin. “Modelling and L1 adaptive control of pH in bioethanol enzymatic process”. In: Proceedings of the 2013 American Control Conference. 2013, pp. 2–9.

[7] Q. Qing, B. Yang, and C. E. Wyman. “Xylooligomers are strong inhibitors of cellulose hydrolysis by enzymes.” In: Bioresource technology 101.24 (Dec. 2010), pp. 9624–9630. DOI: 10.1016/j.biortech.2010.06.137.

[8] R. M. Prunescu and G. Sin. “Dynamic modeling and validation of a lignocellulosic enzymatic hydrolysis process–a demonstration scale study.” In: Bioresource technology 150 (Dec. 2013), pp. 393–403. DOI: 10.1016/j.biortech.2013.10.029.

[9] B. Lavarack, G. Griffin, and D. Rodman. “The acid hydrolysis of sugarcane bagasse hemicellulose to produce xylose, arabinose, glucose and other products”. In: Biomass and Bioenergy 23 (2002), pp. 367–380.

[10] M. Krishnan, N. Ho, and G. Tsao. “Fermentation kinetics of ethanol production from glucose and xylose by recombinant Saccharomyces 1400 (pLNH33)”. In: Applied Biochemistry and Biotechnology 77-79 (1999), pp. 373–388.

[11] R. Morales-Rodriguez, A. S. Meyer, K. V. Gernaey, and G. Sin. “Dynamic model-based evaluation of process configurations for integrated operation of hydrolysis and co-fermentation for bioethanol production from lignocellulose”. In: Bioresource technology. 2011; 102.2:1174–84. DOI: 10.1016/j.biortech.2010.09.045.

[12] G. Sin, K. V. Gernaey, M. B. Neumann, M. C. M. van Loosdrecht, and W. Gujer. “Global sensitivity analysis in wastewater treatment plant model applications: prioritizing sources of uncertainty”. In: Water research. 2011; 45.2:639–51. DOI: 10.1016/j.watres.2010.08.025.

Extended Abstract: File Not Uploaded
See more of this Session: Dynamic Simulation and Optimization II
See more of this Group/Topical: Computing and Systems Technology Division