(743b) The Use of GVL for the Holistic Fractionation of Biomass: Process and Energy Integration for the Production of Platform Chemicals

This work deploys two different levels of engineering analysis that justify significance and necessity of PSE tools to build high efficient and sustainable biorefinery applications. Firstly, a novel biomass fractionation chemistry – recently validated in laboratory scale (Shuai et al., 2016) – has been extrapolated into industrial scale by means of process flowsheeting and simulation to test performance. Secondly, the analysis capitalizes the combinatorial synthesis and integration methodology of Pyrgakis and Kokossis (2019) to search for cost-effective and complementary chemistries that optimally integrate with the fractionation process and valorise biomass fractions into added value biochemicals.

The biomass fractionation concept is aligned to the holistic valorisation of biomass components and the effective recovery of sugars (cellulose and hemi-cellulose) and lignin monomers. Though organosolv technologies has been credited to effectively extract xylan and glucan monomers, processing and depolymerization of lignin structures are inhibited by degradation issues related with cleavage of ether bonds (β-Ο-4, 4-Ο-5 etc.) in between monolignols; spontaneous repolymerization of C-C bonds resulting in random structural recombination of the polymer matrix; and irreversible lignin condensation with reduced aromatic monomers; thus, preventing the use of lignin compartment from sustainable valorisation.

The inefficient recovery of lignin compartment as a functional intermediate usually leads to the misconception that the lignin-rich pulp is a process-waste or should be used as fuel. The organosolv technology proposed by Shuai et al. (2016) takes advantage of the “lignin-first” biorefinery concept to stabilize and protect the lignin structure from repolymerization during processing and extraction.

GVL (γ-valerolactone) has been tested and selected, against THF and 1,4-Dioxane, as the primary solvent for the depolymerization of sugars and its ability (against others) to deter degradation of biomass components; to effectively depolymerize hemicelluloses into xylan monomers (xylose and diformyl-xylose); to inhibit humins formation; and to dissolve solid lignin. The latter ability can be tuned by simple addition of water. The chemistry takes advantage of a GVL-water (80-20%) mixture to dissolve biomass components, while higher water/GVL ratios (over 1.5) trigger lignin precipitation for separation and recovery purposes. In addition, the non-azeotropic mixture of GVL-water facilitates downstream processing and separations. Considerable improvements on the yields and quality of extracted lignin are attributed to the use of formaldehyde (3% w/w) as a co-solvent to prevent lignin condensation, to hinder the formation of C-C linkages and to produce a soluble lignin fraction (Shuai et al., 2016). Formaldehyde (FA) stabilization results in high lignin monomer yields of up to 78% (after hydrogenolysis), while in the absence of FA yields are estimated to 24.5%. Based on lab-scale efficiencies, PSE tools are implemented to extrapolate organosolv performance in industrial scale.

The process design analysis has been conducted on Aspen plus (v8.8) to construct (from scratch) the integrated processing stages of fractionation process (core process) including both upstream organosolv reaction systems and downstream processing for separation and recovery of solvents and biomass fractions (sugars and lignin). Biomass fractionation is operated in a monophasic (liquid) reactor at 120 ^οC and 2 bar. GVL hydrolyses 92% of hemicellulose into xylose, while 93% of cellulose is recovered as solid pulp after solid-liquid separation. GVL dissolves lignin, while FA is reacting with the hydroxyl content (of lignin) leading to the stabilization of the dissolved lignin. Downstream processing is developed over a sequence of separation stages.

A two-stage centrifugation sequence is used for the recovery of cellulose pulp. The effluent mixture includes the remaining FA, xylose, lignin, GVL and water. FA is recovered by distillation, while lignin is precipitated by adding water. The xylose-GVL mixture is dehydrated through a four-stage evaporation system, while GVL is finally separated from xylose via distillation and recycled back to the biomass fractionation reactor. Finally, the cellulose pulp is hydrolyzed into glucose in presence of enzymes. Xylose, glucose and lignin exiting the fractionation process constitute the three fundamental intermediates for the production of added value bioproducts. Downstream the core process, a large and complex value chain is branched comprising of over 30 candidate chemistries that compete for the valorisation of the three intermediates. The process models and simulations results of all candidate chemistries have been completed in Aspen plus (v8.8) and are accessible via the Bio Modeling Platform (BIOMP, 2020), which is operated by the NTUA/IPSEN research group (IPSEN, 2020).

The chemistries are attributed to chemical, biochemical and thermochemical routes for the production of solvents and fuels, food-additives and cosmetics, construction materials, common and special polymers with several uses in medicine, technology and everyday life. Questions arise about the optimal chemical paths to be selected and get integrated along with the core fractionation process. The individual and distributed operation of downstream chemistries is not an option, since the high energy and feedstocks costs are not capable to support the selected chemistries. Proper selection and integration of processes are indispensable to save energy and materials. For this purpose, the Total Site Synthesis (TSS) methodology of Pyrgakis and Kokossis (2019) has been selected to screen the infinite candidate biorefining routes and search for complementary process portfolios that optimally integrate each other securing sustainable production.

The methodology combines integration thermodynamics with mathematical programming. On one hand, a Biomass Bipartite graph Representation (BBR) is employed to map all feasible connections among the core process and the candidate chemistries. BBR holds product nodes, processing events and arcs to transform the complex 33-chemistries value chain into a superstructure-like representation for process synthesis purposes. The BBR elements (nodes, events and arcs) are used for the construction of mass balances across value chains. Mass balances are used for the selection of processing paths and operation capacities letting for competitive processes to share common upstream feedstock chemicals and searching for better integration patterns. Co-existence of competitive processes is possible to improve savings by appropriately exchanging energy among them, rather than selecting one of competitors against others (Pyrgakis and Kokossis, 2019). While BBR selects process portfolios, an integration representation is employed to optimize and evaluate energy efficiencies of selected chemistries.

A new heat cascade representation – the Total Site Cascade (TSC) – is used to model both direct and indirect integration among all processes (and their hot/cold streams) selected by BBR. Direct heat source-to-sink integration is applied among streams of each process. Indirect integration is applied to share heat from process-to-process (chemistry-to-chemistry) via steam generation and reuse by utilizing available heat of involved processes. The innovation of TSC, compared to conventional integration techniques, lies on the incorporation and mathematical formulation of both integration levels along a single cascade to systematically estimate energy targets for each candidate biorefinery site selected by BBR. Otherwise, common direct and indirect integration are provided as separate procedures, while indirect is exclusively conducted by means of graphical tools; thus, process and process-to-process integration could not systematically investigate integration efficiencies of the infinite value chain combinations. In addition, TSC holds extra options to simultaneously optimize the utility levels and maximize power cogeneration of the understudied biorefinery. TSC is described by means of an updated heat transshipment model for the constructions of energy balances and heat exchange options among all candidate process streams and hot/cold utilities. Mass (BBR) and energy (TSC) balances are finally constructed as a mixed integer-continuous linear optimization model for the selection of optimal biorefinery process portfolios.

The TSS methodology was used to make decisions about the exploitation of chemical intermediates produced by the organosolv core process, which consumes 18,750 kg/hr of wheat straw. The selected paths include the production of xylitol from xylose intermediate; the production of ethylene and glucaric acid by sharing the glucose intermediate by 64-26% to each production line, respectively; and the production of resins from lignin. The integrated biorefinery achieves up to 15% energy cost savings by appropriately sharing heat among processes via steam. The biorefinery profitability was estimated to 60 Μ€/yr, while power cogeneration was maximized to 13.2 MW.

References

1. Shuai, L., Amiri, M.T., Questell-Santiago, Y.M., Héroguel, F., Li Y., Kim, H., Meilan, R., Chapple, C., Ralph, J., Luterbacher, J.S., 2016. Formaldehyde stabilization facilitates lignin monomer production during biomass depolymerization. Science, 354, 6310, 329-333.
2. Pyrgakis, K.A., Kokossis, A.C., 2019. A Total Site Synthesis approach for the selection, integration and planning of multiple-feedstock biorefineries. Computers and Chemical Engineering, 122, 326–355
3. BIO Modelling Platform (BIOMP), https://tools.ipsen.ntua.gr, (accessed: April, 2020)
4. NTUA/IPSEN research group, http://ipsen.ntua.gr/, (accessed: April, 2020)