Flowsheet Simulation Research Articles

Improved prediction of biomass gasification models through machine learning

Gasification of lignocellulosic biomass can be used to produce syngas used as a biorefinery feedstock. To facilitate the commercialisation of the gasification process, models are used to predict the outputs, simulate the impacts of irregular circumstances, and analyse process feasibility. This paper presents a hybrid model combining Aspen Plus and machine learning (ML) algorithms to enhance the prediction of gasification outputs. A base case gasification process flowsheet simulation was implemented in Aspen Plus based on assumed thermodynamic equilibrium conditions which can lead to inaccurate results. To address this, six ML algorithms were applied to collected experimental data and analysed for accuracy and efficiency. The feature importance, accuracy improvement, and the effect of implementing the ML predictions in the gasification block on the rest of the flowsheet were investigated. This paper emphasises the need of higher accuracy models and the great potential of ML approaches to offer high accurate predictions.

The future of Ni-Cu smelting in Botswana: the choice between flash-smelting and top-submerged lance furnaces

The BCL furnace is scheduled to restart operations after being placed on care and maintenance in 2016 due to depressed nickel prices. The decision to start operations after such a stoppage ought to be led by a techno-economic evaluation of process options to make operations more resilient. For BCL, the decision lies between using the existing flash-smelting furnace (FSF) or a top-submerged lance (TSL) furnace. Available studies show that the FSF combined with other converting technologies is cheaper to operate; however, such studies are based on flowsheet simulation results of single copper concentrates and therefore do not provide a full scope of practical capabilities based on smelter operator skills. Using a t-test on normalized prior operational data from three Cu FSF and two Cu TSL, it was found that the only statistical differences between the operating costs of the FSF and TSL technologies are in the coal use for heating the feed blow and number of rebuilds: the FSF consumes coal to heat the feed blow and the TSL requires 2.4 rebuilds during a course of a single FSF campaign. A summative operating cost comparison over the existing BCL FSF 11-year life shows that BCL can decrease operational costs by some BWP 55.5 million if the operation were to change to the newer TSL technology. The savings derive from reduced coal use, which is normally associated with heating the blow when using the FSF. Despite the TSL carrying added costs of 2.4 rebuilds during a life time of one FSF furnace campaign, the cost associated with the TSL rebuilds of approximately BWP 1.3 million is minimal compared with the overall operational expenditure decrease that BCL will incur on using the TSL technology. In terms of the technology, BCL can have higher Ni-Cu recoveries at lower operational costs by elimination of Ni-Cu losses and eliminating the slag-cleaning furnace by converting to a TSL furnace.

Parallel Genetic Algorithm Interface II: A novel computational tool for accelerated simulation-based optimization

The ever increasing power of computational tools encouraged by the general need for development of more sustainable technologies fuels the interest in modern optimization approaches. While simulation-based optimization has been receiving considerable attention in the past decades, it still struggles to overcome some challenges, namely excessive computation time. This study proposes a novel optimization interface, the Parallel Genetic Algorithm Interface II (PAGAN-II), which utilizes parallelization of flowsheet simulations to drastically reduce the optimization time without the need to use clustered CPUs and/or modified optimization algorithms. Results of a detailed performance study showed up to 2100% increase in computation rate when optimizing demanding process flowsheets; and approximately 300% increase when optimizing simple ones. Capabilities of the proposed interface were demonstrated by optimization of a 5 MTPA C3MR LNG technology processing 12 different feedstocks, where a 15–30% decrease in the specific energy consumption was achieved. At the same time, the algorithm increased the optimization speed 13-fold compared to the traditional approach. This translates into a reduction of optimization time from 69 days of non-stop computation to approximately 7 days.

Techno-economic assessment of renewable dimethyl ether production pathways from hydrogen and carbon dioxide in the context of power-to-X

Dimethyl ether (DME) is an emerging alternative to traditional diesel and LPG as a power-to-X product. This study employs kinetic models to simulate potential DME production routes, assessing key technical and economic performance indicators for comparison. Aspen Plus, a commercial process flowsheet simulation tool, is used alongside Pinch methodology for heat integration. Four production routes are examined: two direct, converting feed gas directly to DME, and two indirect, involving methanol as an intermediate step. For one case in each, CO2 is initially converted to CO via the reverse water gas shift process (rWGS). Post heat integration, all routes exhibit higher cooling demand, with significant energy savings, notably in the indirect pathways. Indirect routes achieve power-to-fuel efficiencies of up to 39.6%, considering electrolysis and carbon capture. Direct routes excel in carbon conversion efficiencies, reaching 92.6%. Economically, indirect routes have higher upfront costs but lower operational expenses due to better energy efficiencies. Direct CO2 conversion yields the lowest levelized manufacturing costs at 2.45 €/kg, closely followed by the indirect route with rWGS. Further research is needed to enhance kinetic models and reduce simulation result uncertainty for direct conversion.

Process integration and systems optimization for the hydrodeoxygenation of 5-hydroxymethylfurfural to dimethylfuran

Most biomass reaction studies have focused on optimizing product yield without considering cost and emissions as key metrics; they consider pure feeds without accounting for process integration resulting from separation and recycling and often change one parameter at a time to maximize yield. Here, we propose a framework to overcome the above issues. We demonstrate it for the hydrodeoxygenation (HDO) of 5-hydroxymethylfurfural (HMF) to produce 2,5-dimethylfuran (DMF), a crucial reaction in making lignocellulosic biomass-based platform chemicals. We consider the impact of water in the feed stemming from the sugar dehydration reactor on the HDO in 2-pentanol over a Ru/C catalyst, guided by an active learning experimental design (NEXTorch toolkit), combined with Aspen Plus process flowsheet simulation, techno-economic analysis and life cycle assessment. We demonstrate that Bayesian optimization of process flowsheets significantly reduces production costs by 26% and greenhouse gas emissions by 15% after striking a balance between raw material usage, solvent loss, and utility consumption. Such reductions are strongly correlated to the product yield due to the dominant cost of the feed. Interestingly, a slight amount of water negatively impacts greenhouse gas emissions more than production costs, requiring relatively high purity in process integration.

Accelerated design of Power-to-X process chains for transient operation using Recurrent Neural Networks

Climate change requires a reorientation towards renewable energy generation and storage technologies, especially from solar and wind-based sources. Power-to-X technologies offer a promising way to convert renewable energy into chemical energy carriers based on green hydrogen. A major obstacle to the implementation of such systems is the need to operate the plants dynamically to account for fluctuations. The conventional design of these process chains is usually done through complex dynamic flowsheet simulations, which require specialized engineers and are therefore time-consuming and costly.This study presents an innovative approach to build surrogate models of process chains by using Recurrent Neural Networks and in detail Long Short-Term Memory. To establish a basis for the training data of these models, process simulations were performed with respect to critical parameters for plant dynamics and capacity adjustment. The developed models can represent the dynamics of the process and allow the estimation of annual production volumes for different geographical locations, based on local wind speeds. In addition, plant-specific variables, such as the number of cells in hydrogen-producing electrolysis, can be modified and optimized. These models open the possibility of making informed decisions at an early project stage and reduce the risk of implementing dynamic Power-to-X plants.

Calculation and Simulation of Flowsheets for Processing Flue Gases from Thermal Power Plants to Methanol

Several flowsheets for processing flue gases from thermal power plants to commercial-quality methanol, differing in the steps of producing the feedstock for the methanol synthesis, were considered. The dependence of the methanol yield, energy efficiency, and carbon footprint of the technology as a whole (from the step of СО2 capture from flue gases to the step of methanol extraction) on the composition of flue gases (СО2 concentration) and temperature of their flow was studied. The option of processing flue gases with high initial flow temperature using the flowsheet involving high-temperature steps of СО2 separation and electrolysis was considered. Taking into account the recovery of the heat from flows and steam generation, this flowsheet has high efficiency.

Optimizing raw material composition to increase sustainability in porcelain tile production: A simulation‐based approach

AbstractReducing the environmental impact of porcelain tile production while maintaining cost‐effectiveness is challenging. This work introduced a novel modeling approach for optimizing a standard composition range comprising kaolinite (15–38 wt.%), illite (0–20 wt.%), quartz (20–40 wt.%), and feldspar (20–45 wt.%) to establish a robust composition interval for porcelain stoneware tiles. The proposed study considers several factors, such as composition impact on the manufacturing sequence, production costs, and CO2 emission. A flowsheet simulation database was generated by coupling the Dyssol framework with MATLAB. This study investigated the influence of raw material composition within the process sequence, the total CO2 emissions, and production costs within the contexts of Spain and Brazil, two of the top five global producers. Granules with a higher proportion of talc and illite exhibit reduced moisture content after spray drying, and these combinations have lower green body porosity after compaction. The addition of talc allowed for decreased porosity content after compaction reduced firing temperature, and lowered costs and CO2 emissions despite the higher prices associated with talc. The proposed simulation methodology offers a powerful decision‐making tool for optimizing raw material composition to minimize cost and CO2 emissions in the porcelain tile production. This methodology represents an early stride toward integrating digital twin methodologies within the ceramic tile sector, facilitating improved process regulation, and promoting the adoption of digital technologies.

ESFILES: Intelligent process flowsheet synthesis using process knowledge, symbolic AI, and machine learning

Process flowsheet synthesis, design, and simulation require integrated approaches that combine domain knowledge and data-driven methods for fast, efficient, and reliable solutions. However, due to the recent surge in data and machine learning capabilities, there has been a shift towards building purely data-driven systems for process flowsheet synthesis and related problems. Such approaches have certain drawbacks. Here, we present a hybrid method that combines data-driven approaches with domain knowledge to represent process flowsheets and solve problems related to process synthesis, design, and simulation. We present an extended SFILES (or eSFILES) representation, a multi-level hierarchical flowsheet representation with varying degrees of process knowledge. At level 0, flow diagrams are represented as purely text-based SFILES strings. At level 1, the SFILES grammar, along with inferencing algorithms, is used to construct a flowsheet hypergraph explicitly representing flow diagram connectivity. At level 2, specifications needed for material and energy balance calculations are introduced, and, after simulation, the results are also added using annotated flowsheet hypergraphs. Finally, at level 3, a process ontology is connected with the annotated flowsheet hypergraph to include design and operation parameters as well as the detailed simulation results. We discuss this hierarchical framework using several case studies.

Novel Corncob‐Based Catalytic Biodiesel Production Process: Experiments, Modeling, and Simulation

AbstractA functionalized catalyst for catalyzed biodiesel production via a heterogeneous route is a highly focused area to lower the cost of production and mitigate the drawbacks of homogeneously catalyzed reactions. Production aspects such as parameter study, kinetics modeling, and simulation of continuous process flowsheets incorporating kinetic parameters are scarce in the literature. In the current work, a sulfonic group‐functionalized porous carbonaceous catalyst based on corncob was used for the esterification of oleic acid. The Langmuir‐Hinshelwood‐Hougen‐Watson (LHHW) kinetic model was found to best fit to correlate the experimental data and thus applied to deduce the kinetic parameters. The obtained kinetic parameters were incorporated into the Aspen Plus simulator to simulate the continuous biodiesel production process. The catalyst showed a strong affinity for oleic acid which enhances the reaction rate.

Aviation fuel production pathways from lignocellulosic biomass via alcohol intermediates – A technical analysis

In order to achieve the goal of limiting global warming to 1.5 °C, also the growing aviation industry needs to take measures to reduce their greenhouse gas emissions. Various renewably sourced aviation fuels can significantly reduce greenhouse gas emissions and most of them, except for example liquid hydrogen or LNG, can be used in the existing infrastructure without airport or aircraft modifications. As most of these renewably sourced fuel types are not (yet) produced at commercial scale, many technological assessment parameter (e.g. carbon or energy efficiency) are uncertain. Thus, the goal of this study is to compare two different process routes, both being based on biochemical and thermochemical conversion steps. The processes evaluated against conversion efficiency of the available raw feedstock and process energy requirements. The evaluation uses theoretical and biochemical carbon efficiency as well as energy efficiency as indicators. A steady-state flowsheet simulation for two biogenic process paths via biogas and bioethanol as intermediate products is carried out on the basis of literature data. In addition, the optional use of solid residue from the biotechnological process step by combustion for direct heat supply cases are studied. In the ethanol-based route, about 23% of the carbon in the feed can be recovered as kerosene, whereas this is only about 19% in the biogas route. Simultaneously, the ethanol-based route without the combustion of the residue has an energy efficiency of 28%, while the biogas route has an efficiency of 24%.

Simulation of technetium rejection in an advanced PUREX flowsheet

Controlling the routing of technetium to keep it away from the uranium product streams is essential in order to deliver efficient and safe recycling of spent nuclear fuel in future processes such as Advanced PUREX. In this paper, a flowsheet simulation code is developed for the simulation of technetium and zirconium routing in an Advanced PUREX flowsheet. The essence of this code is its consideration of the slow conversion of inextractable forms of zirconium to extractable forms in the modelling of mass transfer rates. The code is applied to model a flowsheet tested experimentally and is validated by experimental results. Sensitivity of the flowsheet to operating parameters, including temperature, organic-to-aqueous (O/A) flow ratio, and scrub nitric acid concentration is investigated; an optimized set of operating conditions is suggested that could result in lower amounts of technetium being routed to the uranium product.

Assessing the future of second-generation bioethanol by 2030 – A techno-economic assessment integrating technology learning curves

Lignocellulosic biomass is the most abundant source of renewable biomass and is seen as a high-potential replacement for petroleum-based resources. The conversion technologies to advanced biofuels are still at a low maturity level, thus allowing for future cost reductions through technological learning. This fact is barely considered in state-of-the-art techno-economic assessments and a structured approach to account for technological learning in techno-economic assessments is needed. In this study, a framework for techno-economic assessments of advanced biofuels, integrating learning curves, is proposed. As a validation of this framework, the economic feasibility of the valorization of corn stover for the production of second-generation bioethanol in Belgium is studied. Process flowsheet simulations in Aspen Plus are developed, with an emphasis on the comparison of four different pretreatment technologies and two plant capacities at 156 dry kt biomass/y and 667 dry kt/y. The dilute acid pretreatment model of the large-scale biorefinery required the lowest minimum learning rate to reach an economically feasible biorefinery by 2030, being 3.9%, almost half as the one calculated for the smaller scale plant. This learning rate seems to be achievable based on learning rates commonly estimated in literature. We conclude that there is a potential for advanced ethanol production in Belgium under the current state of technology for large-scale biorefineries, which require additional biomass imports, when accounting for future cost reductions through learning.

Real‐Time Interactive Navigation on Input‐Output Data Sets in Chemical Processes

AbstractInput‐output data sets are ubiquitous in chemical process engineering. We introduce a real‐time interactive navigation framework that provides several capabilities to the decision maker (DM). Once a surrogate model is trained the DM can perform what‐if analyses in both input and output spaces by manipulating sliders. An approximated convex hull spanned in the input space supports both a reliable surrogate prediction and a navigation close to the data set. The framework has been tested on data sets obtained with a flowsheet simulator modeling a real steam methane reforming process.

Surrogate model-based optimisation of a batch distillation process

A surrogate model-based method is proposed for optimising batch distillation processes and applied to the recovery of methanol from a five-component azeotropic waste solvent mixture, where pollutants are removed in two fore-cuts and an after-cut. The objective function is the profit of a single batch, while constraints are for the purity of the main cut and composition of the second fore-cut to be recycled. Simulations are performed by a flow-sheet simulator in a set of points in the space of optimisation variables (reflux ratios of steps, stopping criteria of fore-cuts). Algebraic surrogate models are fitted by ALAMO to simulation results to describe the objective function and the constraints. The resulting optimisation problem is solved numerically. The profit obtained is by 5% higher than the one previously obtained by genetic algorithm (commonly used for optimisation of batch distillation), while the number of simulations is reduced to its third. The highest profit, previously obtained by the Nelder-Mead simplex method, is approached within 1%. Although the simplex method required fewer simulations, the new method proposed here is a global one. The process is re-optimised for different prices to investigate their influence on the profit and optimal values of operational parameters.

Computer‐Aided Molecular and Process Design

Increasing energy prices, tighter environmental regulation, and global competition urge the industry to optimize its processes for converting chemicals and energy. This optimization is far from trivial. Often, the key to major breakthrough relies not only on optimal process conditions but also on selecting suitable molecules as solvents, catalysts, or heat transfer fluids. This selection is highly coupled with the process itself: Choosing the right molecule is a crucial lever for process performance. However, the molecular search space is vast: more than 200 000 000 molecules have already been reported in the CAS registry. The search for the best molecule is thus the search for the needle in the haystack. The ultimate success of a process is only ensured if a molecule is selected based on its optimal process performance. Traditionally, however, the selection of a molecule and the design of the process have been approached sequentially and independently. Computer-aided molecular and process design (CAMPD) methods have been developed to tackle this challenge and systematically tailor molecules for a given application. Thereby, CAMPD can contribute to the transition towards a more sustainable energy and chemical industry, for example, by identifying energy-efficient carbon capture materials, sustainable working fluids for heat pumps, or green solvents for production processes. What is required for a successful application of CAMPD? The basis of a CAMPD problem is a computer-aided molecular design (CAMD) formulation that allows treating the molecular structure as a degree of freedom in an optimization problem. The CAMD formulation is linked with a property model that enables predicting all molecule properties required to evaluate the constraints and objective functions. To reflect the process performance, classical CAMD methods used simplified property-based indicators as the objective function. Recent CAMPD methods integrate detailed process models to assess process performance directly. The recent development of CAMPD/CAMD includes using advanced physically-based and data-driven property models. Moreover, the scope of predicted properties moves beyond the thermodynamic properties to environmental properties such as global warming potential. After decades of CAMPD/CAMD development, these tools have become available for industrial use by implementing CAMPD in flowsheet simulators. Thereby, the effort and required expert knowledge are greatly reduced to set up and solve a CAMPD problem. However, including such new tools in the daily workflow usually takes time. The urge to develop new sustainable processes hopefully will enhance the application of these CAMPD tools. In this special issue of Chemie Ingenieur Technik, authors from industry and academia present their latest developments and applications of CAMPD in energy and chemical engineering. The special issue cannot provide a comprehensive picture of this rapidly developing field. Thus, we aim to provide the reader with an impression of the current status and possible future directions of CAMPD. We hope readers will find inspiration in this special issue and thank all authors for their highly appreciated contributions! Norbert Asprion André Bardow Jonas Mairhofer Johannes Schilling

Modeling and Experimental Validation for the Preparation of Concentrated Tea Extract Using a Forward Osmosis Process Using a Food-Grade Inorganic Draw Solute

In this study, a one-dimensional mathematical model was developed for the concentration of tea extract using a forward osmosis (FO) process. The membrane transport parameters were estimated by minimizing the error between the model output and the experimental results. The model equations were solved to predict the FO performance within allowable error limits. Further, a series of process flow sheet simulation studies were performed to investigate the effect of different operating conditions on the overall FO performance. Based on the simulation results, a multicriteria optimization was developed to attain the maximum permeate and minimum specific reverse solute flux (SRSF). Using the best-performing process flow sheet model, the feasibility of seawater and high-concentration reject brine as a draw solution was also investigated for the concentration of tea extract. The high-concentration reject brine (105.17 g L–1) can result in rapid dewatering with minimal SRSF (0.14 g L–1) along with simultaneous seawater dilution (81.38 g L–1).

Calculation and Simulation of Flowsheets for Processing Flue Gases from Thermal Power Plants to Methanol

Calculation and Simulation of Flowsheets for Processing Flue Gases from Thermal Power Plants to Methanol

Application of Flowsheet Simulation Methodology to Improve Productivity and Sustainability of Porcelain Tile Manufacturing

Porcelain tile manufacturing is an energy-intensive industry that is in dire need of increasing productivity, minimizing costs, and reducing CO2 emissions, while keeping the product quality intact to remain competitive in today’s environment. In this contribution, alternative processing parameters for the porcelain tile production sequence were proposed based on simulation-based process optimization. Flowsheet simulations in the Dyssol framework were used to study the impact of the milling and firing process parameters on the electrical and thermal energy consumption, final product quality, and productivity of the entire processing sequence. For this purpose, a new model of gas flow consumption in the sintering stage was proposed and implemented. During optimization, the primary condition was to maintain the product quality by keeping the final open porosity of the tile within the specified industrial range. The proposed simulation methodology proved to be effective in predicting the influence of the processing parameters on the intermediate and final products of the manufacturing sequence, as well as in estimating the production costs for the Brazilian and Spanish economic conditions. This approach has shown great potential to promote digitalization and establish digital twins in ceramic tile manufacturing for further in-line process control.

Improving the sustainability of porcelain tile manufacture by flowsheet simulation

Porcelain tiles manufacturing is energy-intensive since it contains several stages in which the product is subjected to thermal treatments. The energy used in the process mainly comes from the combustion of natural gas, pushing the industry to decrease its consumption to upkeep its competitiveness in the coming targets of modern sustainable policies. This work presents a simulation-based strategy for the holistic optimization of energy consumption in the tile manufacturing process. Mechanistic models for all involved processing units, describing the respective thermal and electrical energy consumptions, have been implemented in the Dyssol framework and validated against available experimental data. The energy consumption optimization procedure was performed using a coupled Dyssol-Matlab framework. The proposed cost function, which considered electric and thermal energies, was used as the main criterion to determine the optimal design set of process parameters. Various optimization approaches to minimize the overall cost function and the required computational time have been compared. The proposed strategy allows the computing of improved process parameters resulting in minimization of energy consumption and CO2 emissions while keeping product quality intact. Among the solutions, a reduction of 30.2% of fuel per ton of fired tile can be obtained with modification of the processing parameters.