(544f) Computationally Efficient Distillation Energy Targeting Model for Superstructure-Based Process Synthesis

Separation processes account for a significant fraction of the total energy consumption in chemicals and fuels facilities¹. Thus, to design energy efficient systems, especially during the early design phase, it is important to develop methods that allow us to easily estimate separations-related energy requirements. Accordingly, given the importance of distillation in the process industries, the goal of this work is to develop an improved method for the calculation of energy demand of complex distillation networks. Shortcut methods, which do not rely on rigorous mass-energy balances, are particularly useful for this purpose^2,3. In particular, the Underwood equations² have been adopted in many works because of their simplicity, that is, the computationally cheap calculation of minimum vapor flow rates to achieve the separation task and from the flow the calculation of the minimum energy requirement. Notably, to solve the Underwood equations, components that are present in the stream to be separated need to be known a priori so that all relevant roots of the Underwood equations can be calculated.

However, when synthesizing an integrated reactor-separation system, component flow rates of the stream to be separated can vary, and some components can have even zero flow rates, depending on decisions in the upstream reactor subsystem(s). Thus, components in the stream to be separated may not be known a priori. Such streams are referred to as undetermined⁴. The presence of undetermined streams complicates the use of the Underwood equations; for instance, existing approaches disaggregate the stream so that each disaggregated stream can have a known set of components. Then, each disaggregated stream is connected to a separate distillation network/column. This approach can easily lead to a large optimization problem and therefore prevent the use of Underwood equations (and other shortcut models) when multiple undetermined streams are present.

To address this limitation, we propose a novel reformulation of the Underwood equations, which allows the automatic calculation of all relevant roots efficiently even for undetermined streams. We introduce binary variables to detect components in the stream to be separated, and these binary variables are then used to constrain the Underwood roots accordingly. Also, we introduce strong valid constraints which significantly enhance computational efficiency. Then, utilizing this reformulation, we propose a distillation energy targeting model inspired by the Fully-Thermally-Coupled (FTC) configuration⁵, which has the minimum energy demand among all configurations for the distillation of a zeotropic mixture. The proposed model can be used to calculate the energy target to separate a stream, potentially undetermined, into pure components without finding detailed distillation network configurations. Thus, it is particularly useful for preliminary process synthesis with multiple candidate reactions, where many undetermined streams can naturally appear. Also, we introduce some extensions to consider cases where some components do not need to be separated from each other. For example, for fuel production, some components do not need to be separated from each other if all property constraints can be met.

We present a number of case studies. First, we present the preliminary synthesis of a multi-stage itaconic acid production process. Then, we present the preliminary synthesis of ethanol upgrading bio-refinery, which has more complexity. These case studies illustrate how the proposed energy targeting model can be used to formulate and solve interesting and challenging process synthesis problems.

References

1. Sholl DS, Lively RP. Seven chemical separations to change the world. Nature News. 2016;532(7600):435
2. Bausa J, Watzdorf RV, Marquardt W. Shortcut methods for nonideal multicomponent distillation: I. Simple columns. AIChE journal. 1988;44(10), 2181-2198.
3. Underwood AJ V. Fractional distillation of multicomponent mixtures. Ind Eng Chem. 1949;41(12):2844-2847.
4. Ryu J, Maravelias CT. Computationally efficient optimization models for preliminary distillation column design and separation energy targeting. Computers & Chemical Engineering. 2020;143:107072.
5. Halvorsen IJ, Skogestad S. Minimum energy consumption in multicomponent distillation. 3. More than three products and generalized Petlyuk arrangements. Ind Eng Chem Res. 2003;42(3):616-629.