Flowsheet Simulation Research Articles

Data‐Driven Process Simulation Using Connected Surrogate Unit Models Exemplified on a Steam Methane Reforming Process

AbstractProcess simulation based on physical models often faces computational problems with respect to convergence, especially if the underlying flowsheets are complex. The use of data‐driven surrogate models connected to flowsheets promises to overcome these challenges. Using the steam methane reforming process, this paper presents the development of surrogate models – artificial neural networks – for the key units of the process that are subsequently connected to form the entire flowsheet. The accuracy of the individual surrogate models is analyzed based on the test error; the accuracy of the flowsheet is evaluated by a benchmark process simulation performed in Aspen Plus®. Therefore, the predicted key variables, here outlet temperatures and compositions, are compared to the benchmark. It is shown that their maximum error is below the typical measurement error. The comparison of the accuracy of the surrogate‐based flowsheet simulation with the Aspen Plus® simulation proves to match very well, as long as the training ranges of the underlying surrogate models are not violated. The promising results of this paper pave the way for future work, such as the optimization of process parameters or superstructure optimization.

Explaining and Integrating Machine Learning Models with Rigorous Simulation

AbstractFirst‐principle flowsheet simulation is a reliable data resource for training machine learning (ML) models for process industry, especially when plant data is not available. Process simulators play an even more important role for evaluation and validation of ML models. In this work, we present a workflow for building and evaluating ML models based on data generated by a commercial flowsheet simulator. The resulting hybrid models, combining data‐driven predictions with mass and energy balances, have much lower calculation times than the rigorous models. The implementation of such models shows great potential for solving more complex process engineering problems on the high‐dimensional space in the future, while saving the process engineer's time in the present.

Investigating the processing capacity of batch distillation by applying a second, smaller column

Increasing the processing capacity is important, especially in the pharmaceutical industry, where a given amount of waste solvent must be processed in a specific time period, else the remaining amount of waste solvent must be incinerated, which results in organic solvent loss and considerable environmental impact. In this paper, the possibility of increasing the processing capacity of batch distillation by using a second, already available, smaller batch column in addition to the original one is studied. The size of the columns is not changed during the calculations performed. Two case studies are presented. In the first one (Mixture 1), acetone, in the second one (Mixture 2), acetic acid must be recovered in high purity from their aqueous mixtures by operating either the larger column only or both columns simultaneously. The smaller column serves for preliminary separation, and its main-cut (Mixture 1) or residue (Mixture 2) is further processed in the larger one. The goal of this work is to maximise the processing capacity of the two-column process with the same recovery as in the single-column process. Both processes are modelled with a professional flow-sheet simulator. For the single-column process, the only independent variable is the reflux ratio. For the two-column process, the independent variables are the reflux ratios of both columns and the amount of intermediate product (distillate or residue) of the first column transferred to the second one. The ranges of the independent variables are determined by sensitivity studies for different charge compositions (20–50-80 mass% acetone or acetic acid), and each case is optimised by a genetic algorithm coupled to the flow-sheet simulator. It is also shown in this work that the processing capacity maximum of the two-column process is not always at the equality of the processing times of the two columns.

A Modular Approach for Surrogate Modeling of Flowsheets

AbstractFlowsheet simulation, an important building block in chemical process development and design, generally requires the solution of a nonlinear system of equations. Such a simulation can fail if either no solution exists for the chosen specifications or if the initial values for the solver are chosen unfavorably. In this work, to overcome such convergence failures, AI‐based surrogate models are trained for unit operations of a flowsheet and then interconnected. For a pressure swing distillation with recycle, it is shown that such an interconnection of surrogate models yields more accurate results compared to a surrogate model for the whole flowsheet at once, and that adequate starting points for the flowsheet simulation can be obtained from the interconnected surrogates.

Techno-economically-driven identification of ideal plant configurations for a new biomass-to-liquid process – A case study for Central-Europe

The conversion of agricultural waste materials such as bark or straw into 2nd generation biofuels constitutes an auspicious way to meet part of the future fuel demand in a sustainable way. The number of possible production routes is diverse, and the techno-economic analyses of these routes have been conducted in very different ways. The route involving gasification, gas purification, and a subsequent Fischer-Tropsch synthesis enables the production of hydrocarbons that achieve current fuel standards after upgrading in the existing refinery infrastructure. To evaluate a promising biomass-to-liquid process, a methodology is presented that incorporates economic constraints into the process design, allowing identification of a regionally optimal process design. The production costs of the new concepts are estimated by setting up a detailed flowsheet simulation in AspenPlus® based on experimental data from the successful demonstration runs of the EU-Project COMSYN. In addition, an existing techno-economic evaluation methodology incorporated into the in-house software tool TEPET (Techno-Economic Process Evaluation Tool) has been extended to evaluate the processes’ performance in different European regions and to include transportation and refining costs. The approach enables the identification of regional sweet spots shown on a map for Central-Europe, indicating production costs and favorable process design for each region. Furthermore, the results of an automated mapping for the optimal process design depending on the heat and electricity market are presented. The performed analyses show that the techno-economic evaluation tends to expand the technology in regions with low feedstock costs, while the optimal process design is defined by the regional heat and electricity market. In this work, net production costs of less than 1.12 €2019/lbiofuel were determined for regions in Hungary, Poland and Slovakia.

Using machine learning models to explore the solution space of large nonlinear systems underlying flowsheet simulations with constraints

Using machine learning models to explore the solution space of large nonlinear systems underlying flowsheet simulations with constraints

PharmaPy: An object-oriented tool for the development of hybrid pharmaceutical flowsheets

Process design and optimization continue to provide computational challenges as the chemical engineering and process optimization communities seek to address more complex and larger scale applications. Software tools for digital design and flowsheet simulation are readily available for traditional chemical processing applications such as in commodity chemicals and hydrocarbon processing; however, tools for pharmaceutical manufacturing are much less well developed. This paper introduces, PharmaPy, a Python-based modelling platform for pharmaceutical manufacturing systems design and optimization. The versatility of the platform is demonstrated in simulation and optimization of both continuous and batch processes. The structure and features of a Python-based modeling platform, PharmaPy are presented. Illustrative examples are shown to highlight key features of the platform and framework.

Bridging the Gaps: from Particle-resolved to Multi-tubular Reactor Simulation

Packed bed reactors have been widely used in the chemical and process industry for several decades. They can be difficult to design and operate due to their size and complexity. Therefore, there is still room for improvement of the performance of these reactors and rigorous simulations can help here to achieve desired goals with lower upfront investment. Recent advances in modeling particle-resolved packed beds allows a detailed inside in the flow, species and temperature distribution in the beds and therefore also into the conversion of the surface reactions. Based on these simulations, model parameters for 1D-models can be estimated. These 1-D models can then be used to calculate performance of multi-tubular reactors, either by running a computationally expensive simulation with resolved tubes or by coupling the CFD simulation to an advanced process modeling tool like gPROMS. In the latter case, the flow non-uniformity as well as certain flow properties like coolant velocity and temperature is taken from the CFD simulations, while on the process modeling side heat transfer and reactions in the packed bed are calculated based on the simplified 1D models. This modeling approach is fully 2-way coupled and highly efficient in terms of accuracy and especially runtime and it can be embedded into a flow sheet simulation.In this contribution, we will present the whole simulation process and how the different steps intertwine with each other starting from the detailed particle resolved simulation all the way down to the flow sheet simulation. The benefit of this approach will be demonstrated based on several examples.

Rigorous and Customizable 1D Simulation Framework for Membrane Reactors to, in Principle, Enhance Synthetic Methanol Production

Power-to-liquid production via methanol synthesis has a high potential for emission reduction and carbon-neutral fuel production. However, the low equilibrium conversion of methanol synthesis via CO2 hydrogenation is identified as an important impediment in the further development of the technology. The latter necessitates a more innovative reactor design like membrane reactors to enhance the reaction conversion. In this article, a rigorous and customizable model for membrane reactors is developed using an equation-oriented flowsheet approach. The module requires no analytical correlations for thermodynamic properties, which often fail to reflect the system behavior accurately, or simplifying assumptions such as isobaric and isothermal operation conditions. The model was applied to the synthetic methanol production to determine to what extent a conceptual water-selective membrane reactor improves the reaction’s conversion and selectivity. We propose the process conditions at which a membrane reactor enhances these two key metrics, given the module’s heat transfer mode. This conceptual modeling serves as a guiding benchmark for future innovative reactor designs and facilitates the prospective process development and optimization due to the possible exportation and incorporation into the standard flowsheet simulators such as HYSYS or Aspen Plus.

Automated synthesis of steady-state continuous processes using reinforcement learning

Automated flowsheet synthesis is an important field in computer-aided process engineering. The present work demonstrates how reinforcement learning can be used for automated flowsheet synthesis without any heuristics or prior knowledge of conceptual design. The environment consists of a steady-state flowsheet simulator that contains all physical knowledge. An agent is trained to take discrete actions and sequentially build up flowsheets that solve a given process problem. A novel method named SynGameZero is developed to ensure good exploration schemes in the complex problem. Therein, flowsheet synthesis is modelled as a game of two competing players. The agent plays this game against itself during training and consists of an artificial neural network and a tree search for forward planning. The method is applied successfully to a reaction-distillation process in a quaternary system.

Integration of Renewable Hydrogen Production in Steelworks Off-Gases for the Synthesis of Methanol and Methane

The steel industry is among the highest carbon-emitting industrial sectors. Since the steel production process is already exhaustively optimized, alternative routes are sought in order to increase carbon efficiency and reduce these emissions. During steel production, three main carbon-containing off-gases are generated: blast furnace gas, coke oven gas and basic oxygen furnace gas. In the present work, the addition of renewable hydrogen by electrolysis to those steelworks off-gases is studied for the production of methane and methanol. Different case scenarios are investigated using AspenPlusTM flowsheet simulations, which differ on the end-product, the feedstock flowrates and on the production of power. Each case study is evaluated in terms of hydrogen and electrolysis requirements, carbon conversion, hydrogen consumption, and product yields. The findings of this study showed that the electrolysis requirements surpass the energy content of the steelwork’s feedstock. However, for the methanol synthesis cases, substantial improvements can be achieved if recycling a significant amount of the residual hydrogen.

Life Cycle Analysis: Assessment of Technologies for Droplet Separation – A Case Study

AbstractFor the assessment of separation technologies from an ecological and economic perspective, a case study was performed to investigate the influence of entrainment on energy and resource demand of a single apparatus and on a single production process. Taking the example of the Rectisol process, the influence of droplet entrainment on individual separation equipment as well as on the process as a whole was identified via flowsheet simulation. Based on the calculated mass and energy balances, an LCA and LCC approach was used for the quantification of the entrainment influence.

Integrated process simulation of porcelain stoneware manufacturing using flowsheet simulation

Integrated simulation of manufacturing plants plays crucial role in development of sustainable processes and optimized products. However, no investigation has been reported for the whole process chain of porcelain stoneware manufacturing. In this contribution, semi-empirical models of porcelain stoneware processing steps have been developed and implemented into Dyssol framework. By validation based on industrial and lab-scale experiments, the simulation has been applied for modeling and sensitivity analysis of the entire process. The simulation results are in a good agreement with experiments showing that the proposed modeling strategy has high potential to be used for future digital transformation in this sector.

Dynamic flowsheet model development and digital design of continuous pharmaceutical manufacturing with dissolution modeling of the final product

The integration of continuous unit operations imposes a challenge on the pharmaceutical companies aspiring to achieve plant-wide continuous manufacturing due to the additional complexity of the dynamic interactions, process control and quality assurance. To overcome this challenge, flowsheet modeling emerged as a viable tool to gain deeper process knowledge. In this work, the dynamic model of the integrated continuous manufacturing of acetylsalicylic acid (ASA) was formed based on real continuous unit operations. Furthermore, the surrogate model of the in vitro dissolution test of ASA capsules was integrated into the flowsheet model the first time to analyze the integrated manufacturing in light of the dissolution specifications of an immediate-release formulation. Systematic optimization studies were performed on the integrated process and on the continuous unit operations step-by-step, which clearly revealed the impact of the plant-wide dynamic modeling, as the integrated approach resulted in the threefold increase in the overall productivity and the parallel decrease in the required reactant excess. Sensitivity analyses of the operating conditions and the kinetic parameters were also performed. The crystallization temperature emerged as one of the most critical parameters, which variation could even result in the failure of the dissolution specification. Moreover, by calculating the time-varying sensitivity indices, the dynamics of the error propagation could be studied. These results could facilitate the integration of existing continuous units, the development of the control strategy, and encourage the implementation of dissolution surrogate models into process flowsheet simulations.

Simulation-based life cycle assessment for hydrometallurgical recycling of mixed LIB and NiMH waste

The sustainable and competitive recovery of materials from spent batteries is becoming an urgent issue due to the increasing popularity of hybrid and fully electric vehicles. In this present study, flowsheet simulation was combined with life cycle assessment to investigate the environmental impacts of a conceptual, experimentally proven hydrometallurgical battery recycling process, where nickel metal hydride (NiMH) battery waste is used as reductant for lithium-ion battery (LIB) waste, synergistically improving the extraction of valuable metals from both waste types. Various options for sodium circulation in the precipitation of rare earths were considered in different scenarios. The results showed that the main benefit of the process was the reduced consumption of leaching chemicals. Of the investigated sodium management scenarios, crystallization of sodium sulfate was found to be the most environmentally feasible option, allowing the use of Na as a precipitation chemical for rare earth recovery. Significant potential reductions in climate change, acidification, freshwater eutrophication, and human toxicity were also achievable when compared to the life cycle impacts of the primary production of battery metals, mainly due to the high environmental footprint of primary nickel and cobalt sulfate production. Although further improvements to the process were found to be possible, the future availability of waste NiMH batteries may limit the application of the described process concept on an industrial scale.

Process Design of a Multistage Drying Process via Flowsheet Simulation

AbstractMultistage drying processes including spray drying and fluidized bed unit operations are challenging to design due to the dependencies of the preceding process steps. Flowsheet simulation offers the possibility to simulate a complete process by the application of suitable short‐cut models. In this contribution a novel model for a spray dryer based on a population balance approach is presented. Furthermore, a dynamic model for a fluidized bed dryer is introduced. Experimental data is used for validation and parameter optimization. The calibrated models are then used to design an industrial drying process of multiple drying stages.

A comparative study of thermodynamic models to describe the VLE of the ternary electrolytic mixture H2O–NH3–CO2

Abstract This study aims to ascertain the influence of the activity coefficient model and equation of state for predicting the vapor–liquid equilibrium (VLE) of the multi-electrolytic system H2O–NH3–CO2. The non-idealities of the liquid phase are described by the eUNIQUAC and eNRTL models. The vapor phase is modeled with the Nakamura equation, which is compared with the ideal gas assumption. The models are validated with experimental data from literature on total pressure and ammonia partial pressure. Results show that the models UNIQUAC and NRTL without dissociation can only reproduce the experimental conditions in the absence of CO2. When the electrolytic term is considered, the eUNIQUAC model is able to reproduce the experimental data with greater accuracy than the eNRTL. The equation of state which describes the vapor phase plays no major role in the accuracy of the VLE prediction in the operational conditions evaluated here. Indeed, the accuracy relies on the activity coefficient, therefore the ideal gas equation can be considered if the non-idealities of the liquid phase are described by a well-tuned model. These findings could be useful for equipment design, flowsheet simulations and large-scale simultaneous optimization problems.

Development of Anhydrous Ethanol Purification: Reduction of Acetal Content and Vapor-Liquid Equilibrium Study of the Ethanol-Acetal Binary System.

Acetaldehyde diethyl acetal (herein called acetal) is an important pollutant of anhydrous ethanol. Isobaric vapor–liquid equilibrium (VLE) of an ethanol–acetal binary system was measured using a vapor condensate and liquid circulation VLE still. The experimental data were correlated with Wilson, nonrandom two-liquid (NRTL), and universal quasichemical (UNIQUAC) activity coefficient models, which were found suitable for representing the VLE data. Proper agreements between experimental and calculated VLE data were obtained, which were then confirmed with consistency tests. The applicability of the novel VLE data was demonstrated during an investigation of an anhydrous ethanol purification column. Reduction of the concentration of acetal and other pollutants was examined and optimized in a flowsheet environment. The modeling results were verified in a laboratory with an experimental distillation column, confirming a correct agreement between the results. It must be highlighted that the developed method is suitable for the production of pharmacopeial quality anhydrous alcohol, based on reliable, verified VLE data. The results show the importance of accurate VLE data in critical compositions (low pollutant content); moreover, aiming at high product purity, experimental validation has paramount importance. The consistency between the three platforms (VLE and distillation experiments and flowsheet simulation) confirms the accuracy of the developed method.

Application of PharmaPy in the digital design of the manufacturing process of an active pharmaceutical ingredient

Flowsheet design and optimization constitute one of the key challenges in the chemical engineering and process optimization communities. Software tools for digital design and flowsheet simulation are readily available for traditional chemical processing problems such as distillation and hydrocarbon processing, however tools for pharmaceutical manufacturing are much less widely developed. This paper introduces, PharmaPy, a Python-based modelling platform for pharmaceutical facility design and optimization. The versatility of the platform is demonstrated in simulating both continuous and batch process flowsheets.

Modeling and Flowsheet Simulation of Vibrated Fluidized Bed Dryers

Drying in fluidized beds is an important step in the production of powdered materials. Especially in the food and pharmaceutical industry, fluidized bed dryers are often vibrated to improve the drying process. In the current work, a continuous fluidized bed drying model is implemented in the novel, open-source flowsheet simulation framework Dyssol. The new model accounts for the hydrodynamic characteristics of all Geldart groups as well as the impact of mechanical vibration on the drying process. Distributed particle properties are considered by the model. Comprehensive validation of the model was conducted for a wide range of process parameters, different materials, dryer geometries and dimensions as well as the impact of vibration. Particle properties are predicted accurately and represent the broad experimental data well. A sensitivity analysis of the model confirmed grid independence and the validity of underlying model assumptions.