The Data Exchange in the Process Industry (DEXPI) standardization group has recently released a new specification addressing early process design information, particularly information found in block flow diagrams (BFDs) and process flow diagrams (PFDs). The specification provides an information model and exchange format for such design information, but it does not yet define a graphical notation or modeling approach for creating and exchanging compliant BFDs or PFDs. Our work investigates the structural and semantic correspondence between the DEXPI Process information model and the Business Process Model and Notation 2.0 (BPMN 2.0) standard and proposes a BPMN 2.0-based representation for DEXPI’s BFDs and PFDs. The approach is validated using the Tennessee Eastman process as a case study. The results show that BPMN 2.0 can represent and exchange DEXPI Process-relevant design information in a semantic and standardized way, that the resulting models can be transformed into schema-valid DEXPI 2.0 XML, and that it provides a practical and tool-supported basis for graphical modeling of BFDs and PFDs. To support adoption, we provide an open-source reference implementation, bpmn2dexpi, that realizes the proposed representation, transforms BPMN 2.0 models into DEXPI 2.0-compliant XML, and validates the output against the DEXPI 2.0 XML Schema and information model.

Modeling dynamic systems with a variable number of liquid phases is a challenging task, especially in scenarios where the model is designed for optimization tasks such as parameter estimation. Although there exist methods to model the appearance and disappearance of liquid phases in dynamic systems, they usually require integer variables. In this work, the smoothed continuous approach (SCA) is developed for use with a large number of solvers, since it relies only on continuous variables. To demonstrate the applicability of the new method, the SCA is then applied to model the batch esterification of acetic acid with 1-propanol to water and propyl acetate, and to estimate the reaction parameters. Since the mixture may separate into two liquid phases during the course of the reaction, the parameters are estimated with information on the liquid compositions of both separated liquid phases, which improves the accuracy of the parameter estimates and opens new possibilities for optimal experimental design.

An existing approach for optimization-based process synthesis with abstracted phenomena-based building blocks (PBB) is extended by implementing it into a novel MINLP framework with structural screening. Consistency across the multilayer MINLP framework is guaranteed by creating a MathML/XML data model and subsequently exporting the code to the different program parts. The novel framework focuses both on fidelity by implementing thermodynamically sound models and on generality by employing a state-space superstructure that spans a large search space. In order to retain tractability, we insert a structural screening layer which pre-screens based on binary decision variables of the superstructure by graph- and rule-based analyses, penalizing non-physical instances without solution of the underlying MINLP. The MINLP framework is successfully applied on two challenging synthesis tasks to determine the separation of the feed streams of benzene and toluene, as well as of n-pentane, n-hexane, and n-heptane utilizing superstructures with two, respectively four PBB.

New vapor-liquid equilibrium (VLE) data are continuously being measured and new parameter values, e.g., for the nonrandom two-liquid (NRTL) model are estimated and published. The parameter a , the nonrandomness parameter of NRTL, is often not estimated but is heuristically fixed to a constant value based on the involved components. This can be seen as a manual application of a (subset selection) regularization method. In this work, the practical parameter identifiability of the NRTL model for describing the VLE is analyzed. It is shown that fixing a is not always a good decision and sometimes leads to worse prediction properties of the final parameter estimates. Popular regularization techniques are compared and Generalized Orthogonalization is proposed as an alternative to this heuristic. In addition, the sequential Optimal Experimental Design and Parameter Estimation (sOED-PE) method is applied to study the influence of the regularization methods on the performance of the sOED-PE loop.

Batch processes are usually operated following recipes, which are based on experience and expert knowledge. This ensures feasible and safe operation, because process constraints are indirectly included in the recipe. However, the recipe structure itself constrains the solution space and might exclude other more efficient trajectories. Therefore, the hidden constraints are explicitly formulated, and the arising optimization problem is solved without using prior knowledge in the form of recipes. Case studies are performed on rigorous models of a batch reactor and a batch distillation column. It is demonstrated that the optimization problem formulated as a smoothed dynamic nonlinear programming problem outperforms a mixed-integer formulation. Finally, a multi-objective case is investigated that strongly outperforms a recipe-based benchmark.

Calibrating model parameters to measured data by minimizing loss functions is an important step in obtaining realistic predictions from model-based approaches, e.g., for process optimization. This is applicable to both knowledge-driven and data-driven model setups. Due to measurement errors, the calibrated model parameters also carry uncertainty. In this contribution, we use cubature formulas based on sparse grids to calculate the variance of the regression results. The number of cubature points is close to the theoretical minimum required for a given level of exactness. We present exact benchmark results, which we also compare to other cubatures. This scheme is then applied to estimate the prediction uncertainty of the NRTL model, calibrated to observations from different experimental designs.

Superstructure optimization for process synthesis is a challenging endeavour typically leading to large scale MINLP formulations. By the combination of phenomena-based building blocks, accurate thermodynamics, and structural screening we obtain a new framework for optimal process synthesis, which overcomes prior limitations regarding solution by deterministic MINLP solvers in combination with accurate thermodynamics. This is facilitated by MOSAICmodeling�s generic formulation of models in MathML / XML and subsequent decomposition and code export to GAMS and C++. A branch & bound algorithm is implemented to solve the overall MINLP problem, wherein the structural screening penalizes instances, which are deemed nonsensical and should not be further pursued. The general capabilities of this approach are shown for the distillation-based separation of a ternary system.

To solve large nonlinear equation systems, we present a hybrid method based on interval arithmetic and real-valued root-finding. The proposed approach includes two new interval arithmetic based methods, the so-called cutting and a special kind of bisection of consistent variable spaces. Applied to three examples from chemical engineering, it is shown that the approach is now able to find all system solutions within a variable space as well as only one process-relevant solution in much less time thanks to the built-in root-finding step. For the latter, a conventional Newton method as well as IPOPT have been examined. The hybrid approach no longer needs a well-estimated initial point to converge to a solution, only rough initial variable bounds are required.

Optimization-based process design is a central task of process systems engineering. However, solely relying on steady-state models may potentially lead to dynamic constraint violations, hinder robust performance, or simply reduce the controllability of a process. This has led to the consideration of process dynamics in the design phase, which is commonly termed integration of design and operation / control. Recently, we proposed a framework to carry out this integrative task by formulating a large-scale nonlinear programming problem that is solved simultaneously. To this end, the dynamic process model was discretized, and dynamic variability and parametric uncertainty were included. However, the proposed framework only operates on constant lengths of the finite elements. The discretization error was not assessed. Within this contribution, a method for quantifying this discretization error and adapting the number of finite elements accordingly is incorporated into the recently proposed framework and applied on the case study of a continuous tank reactor. The obtained results with and without discretization error control are compared and, based thereon, a more suitable way to apply the control variables on the process is proposed.

We propose a general methodology of sequential locally optimal design of experiments for explicit or implicit nonlinear models, as they abound in chemical engineering and, in particular, in vapor-liquid equilibrium modeling. As a sequential design method, our method iteratively alternates between performing experiments, updating parameter estimates, and computing new experiments. Specifically, our sequential design method computes a whole batch of new experiments in each iteration and this batch of new experiments is designed in a two-stage locally optimal manner. In essence, this means that in every iteration the combined information content of the newly proposed experiments and of the already performed experiments is maximized. In order to solve these two-stage locally optimal design problems, a recent and efficient adaptive discretization algorithm is used. We demonstrate the benefits of the proposed methodology on the example of the parameter estimation for the non-random two-liquid model for narrow azeotropic vapor-liquid equilibria. As it turns out, our sequential optimal design method requires substantially fewer experiments than traditional factorial design to achieve the same model precision and prediction quality. Consequently, our method can contribute to a substantially reduced experimental effort in vapor-liquid equilibrium modeling and beyond.

Model-based optimal experimental design (OED) is a well known tool for efficient model development. However, it is not used very often. A few reasons for that are: a lack of understanding on how to work with complex OED methods and a small amount of ready-to-use tools available to directly apply OED methods. In the presented contribution OED and sampling strategies are used to categorize OED formulations as nonlinear programs. Different strategies and their combination are analyzed based on performance and robustness. Depending on the availability of measurements, control flexibility of the experimental setup, and model accuracy some strategies are more efficient than others. Based on the proposed guidelines, engineers will have a better understanding about which NLP formulation should be used for their specific task. The methods described are available to the community as a part of open-source code developed in Python.

The complexity of dynamic phenomena present in chemical processes often results in high evaluation costs of accurate first-principles models. This limits their real-time applicability, e.g. for advanced process control. A common solution is the derivation of simpler but faster data-driven surrogate models trained on simulated time series generated from dynamic samplings of a mechanistic model. For batch processes, known non-adaptive dynamic sampling methods lead to unrealistic or even infeasible operation cycles, raising the cost of gener-ating simulated datasets with sufficient information content to train accurate surrogate models. An alternative sampling strategy is developed and analyzed, where sampled input trajectories are constrained to process knowledge in the form of parametrized operation recipes. The proposed methodology is tested for the case studies of full batch cycles of a crystallizer and a batch distillation column, showing that it is more efficient in terms of convergent simulations compared to an established dynamic sampling strategy.

In the absence of knowledge about challenging dynamic phenomena involved in batch distillation processes, e.g., complex flow regimes or appearing and vanishing phases, generation of accurate mechanistic models is limited. Real plant data containing this missing information is scarce, also limiting the use of data-driven models. To exploit the information contained in measurement data and a related but inaccurate first-principles model, transfer learning from simulated to real plant data is analyzed. For the use case of a batch distillation column, the adapted model provides more accurate predictions than a data-driven model trained exclusively on scarce real plant data or simulated data. Its enhanced convergence and lower computational cost make it suitable for optimization in real-time.

New aspects of an interval-arithmetic (IA) based automatic initialization scheme for root-finding algorithms to efficiently solve nonlinear algebraic equation systems (NLEs) are presented. Using the model of a partial condenser, it is demonstrated how additional constraints may eliminate non-physical solutions by the help of interval arithmetic for initialization. Finally, the number of stages of a multicomponent distillation column is varied to investigate the required time of the approach to obtain initial values and solutions by root-finding as a function of the system size. The overall method finds a physically feasible solution for the distillation column with 20 stages without any initial values required.

Model predictive control (MPC) and reinforcement learning (RL) are two powerful optimal control methods. However, the performance of MPC depends mainly on the accuracy of the underlying model and the prediction horizon. Classic RL needs an excessive amount of data and cannot consider constraints explicitly. This work combines both approaches and uses Q-learning to improve the closed-loop performance of a parameterized MPC structure with a surrogate model and a short prediction horizon. The parameterized MPC structure provides a suitable starting point for RL training, which keeps the required data in a reasonable amount. Moreover, constraints are considered explicitly. The solution can be obtained in real-time due to the surrogate model and the short prediction horizon. The method is applied for control of a flash separation unit and compared to a MPC structure that uses a rigorous model and a large prediction horizon.

A significant share of the energy consumption by wastewater treatment plants stems from the air compressors in the aerobic processing steps. To minimize their energy consumption, dynamic optimization is a viable option. Modeling the complex (bio)chemical processes within wastewater treatment plants is a complex task and the simulation based on those models requires significant computational effort and is hard to initialize. An alternative to this approach is data-driven modeling of these plants. Therefore, this work features recurrent neural networks. These networks are trained and tested with real data from a German wastewater treatment plant. The identified models are used to obtain optimal control trajectories. The obtained trajectories are compared to the control actions applied in reality to assess the economic benefit of this approach.

Despite the overarching necessity for sustainability, the realization of novel process concepts using innovative green reaction media, such as microemulsions is still hindered by their inherent complexity. To overcome this, a holistic guideline for the fast-track realization of microemulsion systems is presented and applied on a mini-plant for the hydroformylation of 1-dodecene in microemulsions. It combines a rigorous system analysis and identification of challenges for process design and operation. Here, critical unmeasurability of concentrations, small feasible operation regions, and a highly dynamic system behavior are identified. This is encountered by tailored PSE methods: a dynamic model for the complex three-phasic separation of microemulsions, the formulation of a concentration soft-sensor, as well as multi-rate moving horizon state estimation and subsequent dynamic real-time optimization. Applying those, stable operation of the mini-plant for up to 200 h is ensured together with a steady-state product yield of 38% and a product selectivity of 92%. (C) 2021 Elsevier Ltd. All rights reserved.

Chemie Ingenieur TechnikVolume 94, Issue 9 p. 1313-1313 Vortrag Recipe optimization of batch distillation trajectories based on a data-driven model G. Brand Rihm, Corresponding Author G. Brand Rihm g.brandrihm@tu-berlin.de Technische Universität Berlin, Process Dynamics and Operations Group, Straße des 17. Juni 135, 10623 Berlin, GermanyCorrespondence: G. Brand Rihm (g.brandrihm@tu-berlin.de), Technische Universität Berlin, Process Dynamics and Operations Group, Straße des 17. Juni 135, 10623 Berlin, GermanySearch for more papers by this authorE. Esche, E. Esche Technische Universität Berlin, Process Dynamics and Operations Group, Straße des 17. Juni 135, 10623 Berlin, GermanySearch for more papers by this authorJ.-U. Repke, J.-U. Repke Technische Universität Berlin, Process Dynamics and Operations Group, Straße des 17. Juni 135, 10623 Berlin, GermanySearch for more papers by this author G. Brand Rihm, Corresponding Author G. Brand Rihm g.brandrihm@tu-berlin.de Technische Universität Berlin, Process Dynamics and Operations Group, Straße des 17. Juni 135, 10623 Berlin, GermanyCorrespondence: G. Brand Rihm (g.brandrihm@tu-berlin.de), Technische Universität Berlin, Process Dynamics and Operations Group, Straße des 17. Juni 135, 10623 Berlin, GermanySearch for more papers by this authorE. Esche, E. Esche Technische Universität Berlin, Process Dynamics and Operations Group, Straße des 17. Juni 135, 10623 Berlin, GermanySearch for more papers by this authorJ.-U. Repke, J.-U. Repke Technische Universität Berlin, Process Dynamics and Operations Group, Straße des 17. Juni 135, 10623 Berlin, GermanySearch for more papers by this author First published: 25 August 2022 https://doi.org/10.1002/cite.202255286AboutPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat No abstract is available for this article. Volume94, Issue9Special Issue: (Bio)Process Engineering – a Key to Sustainable Development: ProcessNet and DECHEMA-BioTechNet Jahrestagungen 2022 together with 13th ESBES SymposiumSeptember 2022Pages 1313-1313 RelatedInformation