Norbert Asprion

Application of IR-spectroscopy in thermodynamic investigations of associating solutions

Abstract This paper presents results from a study on the use of IR-spectroscopic information in the development of thermodynamic models of H-bonding solutions. It is based on a previous comprehensive IR-spectroscopic study on H-bonding in solutions of alkanols and phenol in different solvents, which is used here to develop a physico-chemical model of the Gibbs energy of those solutions. The chemical part of the model, describing the association, is parameterized using the IR-data alone. In the physical part of the model, UNIQUAC is used with parameters determined from infinite dilution activity coefficients, i.e. from thermodynamic data. This physico-chemical model, the UNIQUAC association model, is compared to the conventional UNIQUAC model using only the same thermodynamic data to parameterize both models. The separation of the chemical and physical effects based on IR-spectroscopic data in the UNIQUAC association model leads to a clearly improved predictive power. Comparisons for vapor–liquid equilibria, activity coefficients and excess enthalpies are presented. The paper is thus a case study on the benefits that can be drawn from using IR-spectroscopic data in the development of thermodynamic models of H-bonding solutions. The procedure can be extended to other types of models like equations of state or molecular models.

Simulation of mass transfer in reactive absorption

Abstract The discretization of the column and the films plays a significant role in mass transfer calculations and changes results significantly. The use of special grid distribution for the discretization helps to reduce the computational effort and guarantees reasonable results. The performance of these grid distributions will be demonstrated for a known and solved problem. The differences in calculation result will be shown for a column simulation.

Challenges in process optimization for new feedstocks and energy sources

Abstract Current and future challenges of optimization in the process industry are discussed. The gap between academic research and industrial workflow is analyzed. Moreover, issues arising from the shift from conventional fossil fuels as both feedstock and energy source to nonconventional feedstocks (shale gas, tar sands, CO2 and biomass) and penetration of intermittent renewable energy are discussed. This manuscript focuses mainly on offline model-based optimization of design and operation, including the generation and selection of promising process alternatives for new feedstocks in conceptual design, multi-objective optimization, the estimation of thermodynamic parameters of new intermediates and the optimization of process operation under the volatile availability of the new feedstocks and energy sources. Moreover, a number of opportunities and needs for research and development are identified, including the simultaneous optimization of feedstocks, processes and products and a production able to process a variety of feedstocks and to utilize energy when it is cheap.

Decision Support by Multicriteria Optimization in Process Development: An Integrated Approach for Robust Planning and Design of Plant Experiments

Abstract In simulation-based process design, model parameters, like thermodynamic data, are affected by uncertainties. Optimized process designs should, among different other objectives, also be robust to uncertainties of the model parameters. In industrial practise, it is important to know the trade-off between an increase in robustness and the other objectives – like minimizing costs or maximizing product purities. This contribution describes a practical procedure how to incorporate robustness as an objective into a multicriteria optimization framework. The general procedure is illustrated by a concrete example. Finally, we argue that the same approach is useable for an optimal design of plant experiments.

Graybox Models - New Opportunities for the Optimization of Entire Processes

Abstract The use of process optimization is often limited due to missing models for certain process steps. The systematic combination of available operating data with previous knowledge in so called graybox models helps to bridge this gap. Following the concept of incremental model identification it will be shown how the workflow has been integrated into a simulation software environment. It will be applied to a cumene process.

INES – Interface between Experiments and Simulation

Abstract The development of chemical processes is usually based on both experiments (often in pilot plants), and process simulation. Design of experiments, data evaluation and reconciliation, model development and validation are essential steps in this procedure. Different tools and approaches are available for each of these tasks but in the process developer’s workflow, they are usually not supported in an integrated way. Therefore, in the project INES, on which this paper reports, a new interface between experiments and simulation for process design was created, and integrated in a tool box which comprehensively supports process design. It contains modules for data selection and reconciliation, sensitivity analysis, and model validation and -adjustment. Methods from the literature are suitably combined to support the overall goal. The chosen methods, their combination and implementation are described and examples are given which demonstrate the benefits of the new interactive tool in the process development workflow.

Model-based Design of Experiments Using a Flowsheet-Simulator

Abstract The quality of process optimization depends on model parameters. The use of operating data from mini-, pilot or production plants can help to improve model agreement with operation. To support the design of experiments (DoE) in plants with many operability limitations a model-based DoE was implemented into a flowsheet simulator. In this contribution, the implementation of the method is described and its use for the identification of kinetics is demonstrated for a cumene process.