Integrated Process and Molecular Design with Modelica Using Continuous-Molecular Targeting

Abstract

The performance of many chemical and energy conversion processes depends on the choice of the molecules used, e.g. as solvents or working fluids. To capture the complex relations between the properties of the molecules used and the process conditions, the selection of suitable molecules should be directly integrated into process design. Solving the resulting challenging integrated design problem is enabled by the Continous-Molecular Targeting-Computer-Aided Molecular Design (CoMT-CAMD) approach. Here, the combinatorial complexity of the molecular decisions is avoided by relaxing molecular parameters in a physically-based thermodynamic model. So far, implementations of CoMT-CAMD were based on procedural programming languages. This impedes reusability and the investigation of process variants as well as the design of complex processes. In order to overcome these shortcomings, we implement the CoMT-CAMD approach based on objectoriented process modeling and thus enable the integrated process and molecular design with Modelica. The resulting approach is demonstrated for the design of a process and the working fluid for a geothermal Organic Rankine Cycle application.

Simultaneous Optimization of Process Design and Solvent in a Flowsheet Simulator

“…To solve the simultaneous process and molecular optimization problem for the large and complex topologies of industrial processes within reasonable time and effort, generic flowsheet simulators are required. Schilling et al [14] and Gertig et al [15] have already presented examples for the implementation of the 1-stage CoMT-CAMD in gPROMS and Modelica respectively. Harraz et al [16], van Kleef et al [17] and Lee et al [18] used gPROMS with SAFT-γ Mie to solve the process model in Computer Aided Molecular and Process Design (CAMPD) problems for ORC and for CO 2 chemisorption processes.…”

“…14 and Gertig et al. 15 have already presented examples for the implementation of the 1‐stage CoMT‐CAMD in gPROMS and Modelica respectively. Harraz et al.…”