(636d) Superstructure Optimization of Integrated Thermochemical Process for Co-Producing Liquid Fuels and Chemicals

Although the current crude oil price is quite low, it is still necessary to develop reliable biorenewable conversion technologies from a long-term perspective due to diminishing petroleum reserves and increasing market requirements. As a very versatile energy resource, biomass can not only be converted to liquid transportation fuels for vehicular use but can also be transformed to aromatics, olefins, lubricants, industrial chemicals, and many other products currently derived from petroleum or natural gas. Indeed, co-producing these value-added chemicals in an integrated biorefinery can incentivize the de-risking of the â??front-endâ? processes that are necessary for liquid fuels production.

An integrated thermochemical conversion based biorefinery plant for coproducing premium quality liquid transportation fuels, olefins, and aromatics is systematically designed and synthesized in this presentation. The central elements contain gasification, fast pyrolysis, hydrocarbon upgrading, methanol synthesis, methanol-to-olefins, methanol-to-propylene, and methanol-to-aromatics. In general, the pretreated biomass is directed to the gasifier for syngas production and to the fast pyrolyzer for crude bio-oil production, respectively. The major portion of the hot syngas from the gasifier is used to provide heat for the fast pyrolyzer and to promote the fluidization. The crude bio-oil is converted to liquid fuels through sequential hydrotreating and downstream separation. The fluidizing gas along with the syngas from the fast pyrolyzer will finally serve as the feedstocks for methanol production. The produced methanol will be converted to olefins and aromatics through alternative routes. In the developed process superstructure, which considers multiple catalysts and reactor types for certain processing units, is formulated as a generalized disjunctive programming model. To reduce the computational complexity of optimization model, a symbolic regression based approach is proposed for generating surrogate models to describe those complex processing units. The bilinear, trilinear, quadrilinear, and concave terms are reformulated for enhancing the optimization model to be solved to global optimality. Water, power and heat integration studies are simultaneously carried out to reduce the utility cost. Trade-offs between process economics and hydrogen source selection are made. The top adjustable factors on overall economic performance are identified through ranking relevant candidates. Through this analysis, the designed integrated process exhibits high flexibility to cope with the uncertainties on market demands and feedstock sources.