Pub Date : 2024-12-04DOI: 10.1016/j.dche.2024.100205
Kaleem Ullah , Sara Maen Asaad , Abrar Inayat
Hydrogen production from biogas presents a significant opportunity to address major sustainability challenges by providing an economically viable replacement of fossil fuels and reducing greenhouse gas emissions. However, the conversion of biogas into hydrogen using steam reforming is affected by several process parameters. Therefore, this study aims to use a combined approach of DWSIM chemical process simulator and response surface methodology (RSM) as an optimization technique to enhance the effectiveness of the hydrogen production process. The process was modeled with the help of DWSIM software and then validated. Additionally, sensitivity analysis was performed to assess the impact of varying raw material flow rates and reactor conditions on the hydrogen yield as well as investigate the effect of varying biogas composition on the hydrogen yield. Design Expert software was used to optimize the hydrogen production using the Central composite design and a quadratic model. Four input parameters were considered: biogas flow rate, steam flow rate, inlet temperature, and pressure of reformer reactor, with hydrogen yield at the outlet of the last reactor considered as the response. The model and the independent parameters were found to be significant with p-values< 0.0001. The interactions of parameters showed that pressure had the least impact on the hydrogen yield. The optimal parameters identified were 57 kg/hr biogas flow rate, 33.97 kg/hr steam flow rate, 954.38 °C reformer inlet temperature, and 12.52 bar pressure, ultimately achieving a maximum hydrogen yield of 65.992 %. Validation of optimal conditions in DWSIM simulation tool yielded a hydrogen yield of 64.874 % with an error margin of <2.0 %. Overall, this study demonstrates the effect of each parameter and optimizes the hydrogen production process to increase the yield.
{"title":"Process modelling and optimization of hydrogen production from biogas by integrating DWSIM with response surface methodology","authors":"Kaleem Ullah , Sara Maen Asaad , Abrar Inayat","doi":"10.1016/j.dche.2024.100205","DOIUrl":"10.1016/j.dche.2024.100205","url":null,"abstract":"<div><div>Hydrogen production from biogas presents a significant opportunity to address major sustainability challenges by providing an economically viable replacement of fossil fuels and reducing greenhouse gas emissions. However, the conversion of biogas into hydrogen using steam reforming is affected by several process parameters. Therefore, this study aims to use a combined approach of DWSIM chemical process simulator and response surface methodology (RSM) as an optimization technique to enhance the effectiveness of the hydrogen production process. The process was modeled with the help of DWSIM software and then validated. Additionally, sensitivity analysis was performed to assess the impact of varying raw material flow rates and reactor conditions on the hydrogen yield as well as investigate the effect of varying biogas composition on the hydrogen yield. Design Expert software was used to optimize the hydrogen production using the Central composite design and a quadratic model. Four input parameters were considered: biogas flow rate, steam flow rate, inlet temperature, and pressure of reformer reactor, with hydrogen yield at the outlet of the last reactor considered as the response. The model and the independent parameters were found to be significant with p-values< 0.0001. The interactions of parameters showed that pressure had the least impact on the hydrogen yield. The optimal parameters identified were 57 kg/hr biogas flow rate, 33.97 kg/hr steam flow rate, 954.38 °C reformer inlet temperature, and 12.52 bar pressure, ultimately achieving a maximum hydrogen yield of 65.992 %. Validation of optimal conditions in DWSIM simulation tool yielded a hydrogen yield of 64.874 % with an error margin of <2.0 %. Overall, this study demonstrates the effect of each parameter and optimizes the hydrogen production process to increase the yield.</div></div>","PeriodicalId":72815,"journal":{"name":"Digital Chemical Engineering","volume":"14 ","pages":"Article 100205"},"PeriodicalIF":3.0,"publicationDate":"2024-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143159034","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-04DOI: 10.1016/j.dche.2024.100204
Muhammad Shahbaz , Muhammad Ammar , Sukarni Sukarni
The conversion of bioresources like Spirulina platensis (SP) into value-added chemicals, such as methanol, offers a sustainable replacement of fossil fuels and contributes to greenhouse gas mitigation. This study presents an integrated process simulation model, developed using Aspen Plus v10®, for the steam gasification of SP and subsequent methanol production. Process parameters, including temperature range from 650-950 °C, steam/feed ratio from 0.5–2, and recycle ratio from 0–9, were investigated to optimize syngas composition and methanol yield. Results demonstrated that increasing temperature enhances H2 and CO production while reducing CO2 and CH4, significantly increasing methanol production from 6500 to 9500 kg/h. The steam/feed ratio also influences syngas composition and methanol yield, with higher ratios promoting H2 and CO2 production and reducing CO and CH4. The economic evaluation of two scenarios, a base case and an optimum case, shows that the capital expenditure (Capex) and operating expenditure (Opex) are 19.3M$ and 9.07M$ for the base case, and 20.018M$ and 10.21M$ for the optimum case. The analysis also reveals that the optimum case, with higher methanol production (7.2 tonnes/h compared to 6.7 tonnes/h in the base case), generates a higher net income (9.76 M$/y) and reduces CO2 emissions (4.918 tonnes CO2-e/y compared to 5.72 tonnes CO2-e/y). The energy flow indicates the input energy requirement, the energy carried by methanol, and the surplus energy, totalling 26740 kW to meet the major system's energy demands. This study provides valuable insights for researchers, policymakers, and commercial entities seeking to develop sustainable and economically viable biofuel production processes.
{"title":"Conversion of Spirulina platensis into methanol via gasification: Process simulation modeling and economic evaluation","authors":"Muhammad Shahbaz , Muhammad Ammar , Sukarni Sukarni","doi":"10.1016/j.dche.2024.100204","DOIUrl":"10.1016/j.dche.2024.100204","url":null,"abstract":"<div><div>The conversion of bioresources like Spirulina platensis (SP) into value-added chemicals, such as methanol, offers a sustainable replacement of fossil fuels and contributes to greenhouse gas mitigation. This study presents an integrated process simulation model, developed using Aspen Plus v10®, for the steam gasification of SP and subsequent methanol production. Process parameters, including temperature range from 650-950 °C, steam/feed ratio from 0.5–2, and recycle ratio from 0–9, were investigated to optimize syngas composition and methanol yield. Results demonstrated that increasing temperature enhances H<sub>2</sub> and CO production while reducing CO<sub>2</sub> and CH<sub>4</sub>, significantly increasing methanol production from 6500 to 9500 kg/h. The steam/feed ratio also influences syngas composition and methanol yield, with higher ratios promoting H<sub>2</sub> and CO<sub>2</sub> production and reducing CO and CH<sub>4</sub>. The economic evaluation of two scenarios, a base case and an optimum case, shows that the capital expenditure (Capex) and operating expenditure (Opex) are 19.3M$ and 9.07M$ for the base case, and 20.018M$ and 10.21M$ for the optimum case. The analysis also reveals that the optimum case, with higher methanol production (7.2 tonnes/h compared to 6.7 tonnes/h in the base case), generates a higher net income (9.76 M$/y) and reduces CO<sub>2</sub> emissions (4.918 tonnes CO<sub>2</sub>-e/y compared to 5.72 tonnes CO<sub>2</sub>-e/y). The energy flow indicates the input energy requirement, the energy carried by methanol, and the surplus energy, totalling 26740 kW to meet the major system's energy demands. This study provides valuable insights for researchers, policymakers, and commercial entities seeking to develop sustainable and economically viable biofuel production processes.</div></div>","PeriodicalId":72815,"journal":{"name":"Digital Chemical Engineering","volume":"14 ","pages":"Article 100204"},"PeriodicalIF":3.0,"publicationDate":"2024-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143159035","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01DOI: 10.1016/j.dche.2024.100199
Oliver J. Fisher , Jhuma Sadhukhan , Thorin Daniel , Jin Xuan
Successfully transitioning to a net-zero and circular carbon economy requires adopting innovative technologies and business models to capture CO2 and convert it into valuable chemicals and materials. Given the high economic costs and limited funding available for this transition, robust economic modelling of potential circular carbon pathways is essential to identify economically viable routes. This study introduces a novel techno-economic analysis (TEA) of producing alcohol ethoxylate (AE7), a valuable surfactant, from industrial flue gas. Traditionally, AE7 is produced by reacting fatty alcohols with ethylene oxide derived from fossil or bio-based sources. This research explores a method using CO2 captured from steel industry flue gas to produce AE7, addressing a notable gap in the literature. It evaluates a thermo-catalytic pathway involving Fischer-Tropsch (FT) synthesis with syngas generated by the reverse-water gas-shift reaction, where CO2 reacts with H2. CO2 conversion rates range around 3% across processing capacities of 25 kt/a, 100 kt/a, and 1000 kt/a. The study finds that the CO2 mass fraction concentration in the process emission is 2.47 × 10–5, compared to 0.13 in the incoming flue gas, highlighting the system's positive environmental impact. A radial basis function neural network was built to forecast the long-term average price of fossil-based and bio-based surfactants to benchmark the results against. Economic analysis reveals that the cost of green hydrogen significantly impacts the minimum selling price (MSP), making cost parity with existing fossil-based surfactants challenging. The lowest MSP of $8.77/kg remains above the long-term forecasted price of $3.75/kg for fossil-based C12–14 AE7. However, Monte Carlo simulations show a 21% probability of achieving a positive net present value (NPV) compared to leading bio-based surfactant alternatives. Sensitivity analyses identify capital costs, the price of low-carbon hydrogen (LCOH), and diesel prices as the most influential factors affecting the MSP. Continued advancements in Fischer-Tropsch catalyst technologies, reductions in green hydrogen costs and growing consumer demand for environmentally friendly products could significantly enhance the economic feasibility of this sustainable approach, paving the way for broader adoption and contributing to a circular carbon economy.
{"title":"Techno-economic analysis and process simulation of alkoxylated surfactant production in a circular carbon economy framework","authors":"Oliver J. Fisher , Jhuma Sadhukhan , Thorin Daniel , Jin Xuan","doi":"10.1016/j.dche.2024.100199","DOIUrl":"10.1016/j.dche.2024.100199","url":null,"abstract":"<div><div>Successfully transitioning to a net-zero and circular carbon economy requires adopting innovative technologies and business models to capture CO<sub>2</sub> and convert it into valuable chemicals and materials. Given the high economic costs and limited funding available for this transition, robust economic modelling of potential circular carbon pathways is essential to identify economically viable routes. This study introduces a novel techno-economic analysis (TEA) of producing alcohol ethoxylate (AE7), a valuable surfactant, from industrial flue gas. Traditionally, AE7 is produced by reacting fatty alcohols with ethylene oxide derived from fossil or bio-based sources. This research explores a method using CO<sub>2</sub> captured from steel industry flue gas to produce AE7, addressing a notable gap in the literature. It evaluates a thermo-catalytic pathway involving Fischer-Tropsch (FT) synthesis with syngas generated by the reverse-water gas-shift reaction, where CO<sub>2</sub> reacts with H<sub>2</sub>. CO<sub>2</sub> conversion rates range around 3% across processing capacities of 25 kt/a, 100 kt/a, and 1000 kt/a. The study finds that the CO<sub>2</sub> mass fraction concentration in the process emission is 2.47 × 10<sup>–5</sup>, compared to 0.13 in the incoming flue gas, highlighting the system's positive environmental impact. A radial basis function neural network was built to forecast the long-term average price of fossil-based and bio-based surfactants to benchmark the results against. Economic analysis reveals that the cost of green hydrogen significantly impacts the minimum selling price (MSP), making cost parity with existing fossil-based surfactants challenging. The lowest MSP of $8.77/kg remains above the long-term forecasted price of $3.75/kg for fossil-based C<sub>12–14</sub> AE7. However, Monte Carlo simulations show a 21% probability of achieving a positive net present value (NPV) compared to leading bio-based surfactant alternatives. Sensitivity analyses identify capital costs, the price of low-carbon hydrogen (LCOH), and diesel prices as the most influential factors affecting the MSP. Continued advancements in Fischer-Tropsch catalyst technologies, reductions in green hydrogen costs and growing consumer demand for environmentally friendly products could significantly enhance the economic feasibility of this sustainable approach, paving the way for broader adoption and contributing to a circular carbon economy.</div></div>","PeriodicalId":72815,"journal":{"name":"Digital Chemical Engineering","volume":"13 ","pages":"Article 100199"},"PeriodicalIF":3.0,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142744828","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01DOI: 10.1016/j.dche.2024.100198
Madhav Muthyala, Farshud Sorourifar, Joel A. Paulson
Symbolic regression (SR) is a powerful machine learning approach that searches for both the structure and parameters of algebraic models, offering interpretable and compact representations of complex data. Unlike traditional regression methods, SR explores progressively complex feature spaces, which can uncover simple models that generalize well, even from small datasets. Among SR algorithms, the Sure Independence Screening and Sparsifying Operator (SISSO) has proven particularly effective in the natural sciences, helping to rediscover fundamental physical laws as well as discover new interpretable equations for materials property modeling. However, its widespread adoption has been limited by performance inefficiencies and the challenges posed by its FORTRAN-based implementation, especially in modern computing environments. In this work, we introduce TorchSISSO, a native Python implementation built in the PyTorch framework. TorchSISSO leverages GPU acceleration, easy integration, and extensibility, offering a significant speed-up and improved accuracy over the original. We demonstrate that TorchSISSO matches or exceeds the performance of the original SISSO across a range of tasks, while dramatically reducing computational time and improving accessibility for broader scientific applications.
符号回归（SR）是一种强大的机器学习方法，它搜索代数模型的结构和参数，提供复杂数据的可解释和紧凑表示。与传统的回归方法不同，SR探索逐渐复杂的特征空间，这可以发现简单的模型，即使从小数据集也可以很好地推广。在SR算法中，Sure Independence Screening and Sparsifying Operator （SISSO）在自然科学中被证明是特别有效的，它有助于重新发现基本的物理定律，并为材料属性建模发现新的可解释方程。然而，它的广泛采用受到性能低下和基于fortran的实现所带来的挑战的限制，特别是在现代计算环境中。在这项工作中，我们介绍了TorchSISSO，一个内置在PyTorch框架中的原生Python实现。TorchSISSO利用GPU加速，易于集成和可扩展性，提供显着的加速和提高的准确性。我们证明了TorchSISSO在一系列任务中匹配或超过原始SISSO的性能，同时显着减少了计算时间并提高了更广泛的科学应用的可访问性。
{"title":"TorchSISSO: A PyTorch-based implementation of the sure independence screening and sparsifying operator for efficient and interpretable model discovery","authors":"Madhav Muthyala, Farshud Sorourifar, Joel A. Paulson","doi":"10.1016/j.dche.2024.100198","DOIUrl":"10.1016/j.dche.2024.100198","url":null,"abstract":"<div><div>Symbolic regression (SR) is a powerful machine learning approach that searches for both the structure and parameters of algebraic models, offering interpretable and compact representations of complex data. Unlike traditional regression methods, SR explores progressively complex feature spaces, which can uncover simple models that generalize well, even from small datasets. Among SR algorithms, the Sure Independence Screening and Sparsifying Operator (SISSO) has proven particularly effective in the natural sciences, helping to rediscover fundamental physical laws as well as discover new interpretable equations for materials property modeling. However, its widespread adoption has been limited by performance inefficiencies and the challenges posed by its FORTRAN-based implementation, especially in modern computing environments. In this work, we introduce TorchSISSO, a native Python implementation built in the PyTorch framework. TorchSISSO leverages GPU acceleration, easy integration, and extensibility, offering a significant speed-up and improved accuracy over the original. We demonstrate that TorchSISSO matches or exceeds the performance of the original SISSO across a range of tasks, while dramatically reducing computational time and improving accessibility for broader scientific applications.</div></div>","PeriodicalId":72815,"journal":{"name":"Digital Chemical Engineering","volume":"13 ","pages":"Article 100198"},"PeriodicalIF":3.0,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142759667","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-30DOI: 10.1016/j.dche.2024.100201
Sergio M.S. Neiro , Tarun Madan , Christos T. Maravelias , José M. Pinto
In this work, we address a nationwide tactical planning for industrial gas supply chains, particularly argon. The proposed approaches follow as extensions of our previous work (Comp. & Chem. Eng., 161 (2022) 107778) in which a regional argon supply chain problem is addressed; in that work, both production and distribution could be represented in detail. Two different types of deliveries from the Air Separating Units (ASU) to customers, which involve single driver deliveries for short distance trips and sleeper team that require multiple days. The nationwide problem requires simplifications to keep the problem mathematically tractable, primarily the representation of production sites with different tier costs and the aggregation of customers in clusters. The regional problem addressed in our previous work is used as a benchmark case study for benchmarking. We then focus on a real-world problem that represents a nationwide argon supply chain. Despite the size of the models, near optimal solutions could be found in reasonable times. Finally, we highlight important features of the proposed approaches.
{"title":"A nationwide planning model for argon supply chains with coordinated production and distribution","authors":"Sergio M.S. Neiro , Tarun Madan , Christos T. Maravelias , José M. Pinto","doi":"10.1016/j.dche.2024.100201","DOIUrl":"10.1016/j.dche.2024.100201","url":null,"abstract":"<div><div>In this work, we address a nationwide tactical planning for industrial gas supply chains, particularly argon. The proposed approaches follow as extensions of our previous work (<em>Comp. & Chem. Eng., 161 (2022) 107778</em>) in which a regional argon supply chain problem is addressed; in that work, both production and distribution could be represented in detail. Two different types of deliveries from the Air Separating Units (ASU) to customers, which involve single driver deliveries for short distance trips and sleeper team that require multiple days. The nationwide problem requires simplifications to keep the problem mathematically tractable, primarily the representation of production sites with different tier costs and the aggregation of customers in clusters. The regional problem addressed in our previous work is used as a benchmark case study for benchmarking. We then focus on a real-world problem that represents a nationwide argon supply chain. Despite the size of the models, near optimal solutions could be found in reasonable times. Finally, we highlight important features of the proposed approaches.</div></div>","PeriodicalId":72815,"journal":{"name":"Digital Chemical Engineering","volume":"14 ","pages":"Article 100201"},"PeriodicalIF":3.0,"publicationDate":"2024-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143159061","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-27DOI: 10.1016/j.dche.2024.100202
Ammar Riyadh, Nicolas M. Peleato
Ensuring safe drinking water necessitates advanced management and monitoring techniques for water quality in distribution systems. This study leverages machine learning (ML) to model chlorine decay in a water distribution system (WDS) in British Columbia, Canada. A four-layer long short term memory (LSTM) network was trained to predict chlorine concentrations at a reservoir >24,000 m from the treatment plant. Explainable AI (XAI) techniques were applied to the trained network to address critical issues, such as enhancing the transparency and reliability of ML models. Several XAI methods were used to investigate the importance of sensor placement, identify the most significant features, understand feature ranges that result in poor performance, and validate model logic. Results demonstrated that for ML-based WDS control, sensor location is not critical, with high prediction accuracy achieved (mean absolute error <0.025 mg/L) even when exclusively using data from nodes spatially distant from the prediction site. XAI techniques showed the capability of identifying essential features and demonstrated that the behaviour of the ML model conformed with the expectations of chlorine behaviour. Superfluous variables were ranked low in importance, and the model learned fundamental aspects of chemical kinetics, such as temperature dependence and decay rate. Most importantly, the XAI methods applied showed the capability to communicate the reasoning for specific predictions, even at a local or sample-specific level. This study underscores the importance of transparency and trust in ML models, especially as the field transitions towards digital twin and Internet of Things (IoT) technologies, to enhance the effective management of water quality systems.
{"title":"Exploring spatial and temporal importance of input features and the explainability of machine learning-based modelling of water distribution systems","authors":"Ammar Riyadh, Nicolas M. Peleato","doi":"10.1016/j.dche.2024.100202","DOIUrl":"10.1016/j.dche.2024.100202","url":null,"abstract":"<div><div>Ensuring safe drinking water necessitates advanced management and monitoring techniques for water quality in distribution systems. This study leverages machine learning (ML) to model chlorine decay in a water distribution system (WDS) in British Columbia, Canada. A four-layer long short term memory (LSTM) network was trained to predict chlorine concentrations at a reservoir >24,000 m from the treatment plant. Explainable AI (XAI) techniques were applied to the trained network to address critical issues, such as enhancing the transparency and reliability of ML models. Several XAI methods were used to investigate the importance of sensor placement, identify the most significant features, understand feature ranges that result in poor performance, and validate model logic. Results demonstrated that for ML-based WDS control, sensor location is not critical, with high prediction accuracy achieved (mean absolute error <0.025 mg/L) even when exclusively using data from nodes spatially distant from the prediction site. XAI techniques showed the capability of identifying essential features and demonstrated that the behaviour of the ML model conformed with the expectations of chlorine behaviour. Superfluous variables were ranked low in importance, and the model learned fundamental aspects of chemical kinetics, such as temperature dependence and decay rate. Most importantly, the XAI methods applied showed the capability to communicate the reasoning for specific predictions, even at a local or sample-specific level. This study underscores the importance of transparency and trust in ML models, especially as the field transitions towards digital twin and Internet of Things (IoT) technologies, to enhance the effective management of water quality systems.</div></div>","PeriodicalId":72815,"journal":{"name":"Digital Chemical Engineering","volume":"14 ","pages":"Article 100202"},"PeriodicalIF":3.0,"publicationDate":"2024-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143160283","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-14DOI: 10.1016/j.dche.2024.100197
Lorenz T. Biegler
Recent developments in efficient, large-scale nonlinear optimization strategies have had significants impact on the design and operation of engineering systems with equation-oriented (EO) models. On the other hand, rigorous first-principle procedural (i.e., black-box ’truth’) models may be difficult to incorporate directly within this optimization framework. Instead, black-box models are often substituted by lower fidelity surrogate models that may compromise the optimal solution. To overcome these challenges, Trust Region Filter (TRF) methods have been developed, which combine surrogate models optimization with intermittent sampling of truth models. The TRF approach combines efficient solution strategies with minimal recourse to truth models, and leads to guaranteed convergence to the truth model optimum. This survey paper provides a perspective on the conceptual development and evolution of the TRF method along with a review of applications that demonstrate the effectiveness of the TRF approach. In particular, three cases studies are presented on flowsheet optimization with embedded CFD models for advanced power plants and CO2 capture processes, as well as synthesis of heat exchanger networks with detailed finite-element equipment models.
{"title":"The trust region filter strategy: Survey of a rigorous approach for optimization with surrogate models","authors":"Lorenz T. Biegler","doi":"10.1016/j.dche.2024.100197","DOIUrl":"10.1016/j.dche.2024.100197","url":null,"abstract":"<div><div>Recent developments in efficient, large-scale nonlinear optimization strategies have had significants impact on the design and operation of engineering systems with equation-oriented (EO) models. On the other hand, rigorous first-principle procedural (i.e., black-box ’truth’) models may be difficult to incorporate directly within this optimization framework. Instead, black-box models are often substituted by lower fidelity surrogate models that may compromise the optimal solution. To overcome these challenges, Trust Region Filter (TRF) methods have been developed, which combine surrogate models optimization with intermittent sampling of truth models. The TRF approach combines efficient solution strategies with minimal recourse to truth models, and leads to guaranteed convergence to the truth model optimum. This survey paper provides a perspective on the conceptual development and evolution of the TRF method along with a review of applications that demonstrate the effectiveness of the TRF approach. In particular, three cases studies are presented on flowsheet optimization with embedded CFD models for advanced power plants and CO2 capture processes, as well as synthesis of heat exchanger networks with detailed finite-element equipment models.</div></div>","PeriodicalId":72815,"journal":{"name":"Digital Chemical Engineering","volume":"13 ","pages":"Article 100197"},"PeriodicalIF":3.0,"publicationDate":"2024-11-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142656114","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-24DOI: 10.1016/j.dche.2024.100196
AmirMohammad Ebrahimi, Davood B. Pourkargar
This paper focuses on developing an adaptive system decomposition approach for multi-agent distributed model predictive control (DMPC) of integrated process networks. The proposed system decomposition employs a refined spectral community detection method to construct an optimal distributed control framework based on the weighted graph representation of the state space process model. The resulting distributed architecture assigns controlled outputs and manipulated inputs to controller agents and delineates their interactions. The decomposition evolves as the process network undergoes various operating conditions, enabling adjustments in the distributed architecture and DMPC design. This adaptive architecture enhances the closed-loop performance and robustness of DMPC systems. The effectiveness of the multi-agent distributed control approach is investigated for a benchmark benzene alkylation process under two distinct operating conditions characterized by medium and low recycle ratios. Simulation results demonstrate that adaptive decompositions derived through spectral community detection, utilizing weighted graph representations, outperform the commonly employed unweighted hierarchical community detection-based system decompositions in terms of closed-loop performance and computational efficiency.
{"title":"Multi-agent distributed control of integrated process networks using an adaptive community detection approach","authors":"AmirMohammad Ebrahimi, Davood B. Pourkargar","doi":"10.1016/j.dche.2024.100196","DOIUrl":"10.1016/j.dche.2024.100196","url":null,"abstract":"<div><div>This paper focuses on developing an adaptive system decomposition approach for multi-agent distributed model predictive control (DMPC) of integrated process networks. The proposed system decomposition employs a refined spectral community detection method to construct an optimal distributed control framework based on the weighted graph representation of the state space process model. The resulting distributed architecture assigns controlled outputs and manipulated inputs to controller agents and delineates their interactions. The decomposition evolves as the process network undergoes various operating conditions, enabling adjustments in the distributed architecture and DMPC design. This adaptive architecture enhances the closed-loop performance and robustness of DMPC systems. The effectiveness of the multi-agent distributed control approach is investigated for a benchmark benzene alkylation process under two distinct operating conditions characterized by medium and low recycle ratios. Simulation results demonstrate that adaptive decompositions derived through spectral community detection, utilizing weighted graph representations, outperform the commonly employed unweighted hierarchical community detection-based system decompositions in terms of closed-loop performance and computational efficiency.</div></div>","PeriodicalId":72815,"journal":{"name":"Digital Chemical Engineering","volume":"13 ","pages":"Article 100196"},"PeriodicalIF":3.0,"publicationDate":"2024-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142561376","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-22DOI: 10.1016/j.dche.2024.100195
Feiyang Ou , Henrik Wang , Chao Zhang , Matthew Tom , Sthitie Bom , James F. Davis , Panagiotis D. Christofides
Smart Manufacturing, or Industry 4.0, has gained significant attention in recent decades with the integration of Internet of Things (IoT) and Information Technologies (IT). As modern production methods continue to increase in complexity, there is a greater need to consider what variables can be physically measured. This advancement necessitates the use of physical sensors to comprehensively and directly gather measurable data on industrial processes; specifically, these sensors gather data that can be recontextualized into new process information. For example, artificial intelligence (AI) machine learning-based soft sensors can increase operational productivity and machine tool performance while still ensuring that critical product specifications are met. One industry that has a high volume of labor-intensive, time-consuming, and expensive processes is the semiconductor industry. AI machine learning methods can meet these challenges by taking in operational data and extracting process-specific information needed to meet the high product specifications of the industry. However, a key challenge is the availability of high quality data that covers the full operating range, including the day-to-day variance. This paper examines the applicability of soft sensing methods to the operational data of five industrial etching machines. Data is collected from readily accessible and cost-effective physical sensors installed on the tools that manage and control the operating conditions of the tool. The operational data are then used in an intelligent data aggregation approach that increases the scope and robustness for soft sensors in general by creating larger training datasets comprised of high value data with greater operational ranges and process variation. The generalized soft sensor can then be fine-tuned and validated for a particular machine. In this paper, we test the effects of data aggregation for high performing Feedforward Neural Network (FNN) models that are constructed in two ways: first as a classifier to estimate product PASS/FAIL outcomes and second as a regressor to quantitatively estimate oxide thickness. For PASS/FAIL classification, a data aggregation method is developed to enhance model predictive performance with larger training datasets. A statistical analysis method involving point-biserial correlation and the Mean Absolute Error (MAE) difference score is introduced to select the optimal candidate datasets for aggregation, further improving the effectiveness of data aggregation. For large datasets with high quality data that enable model training for more complex tasks, regression models that predict the oxide thickness of the product are also developed. Two types of models with different loss functions are tested to compare the effects of the Mean Squared Error (MSE) and Mean Absolute Percentage Error (MAPE) loss functions on model performance. Both the classification and regression models can be applied in industrial setti
{"title":"Industrial data-driven machine learning soft sensing for optimal operation of etching tools","authors":"Feiyang Ou , Henrik Wang , Chao Zhang , Matthew Tom , Sthitie Bom , James F. Davis , Panagiotis D. Christofides","doi":"10.1016/j.dche.2024.100195","DOIUrl":"10.1016/j.dche.2024.100195","url":null,"abstract":"<div><div>Smart Manufacturing, or Industry 4.0, has gained significant attention in recent decades with the integration of Internet of Things (IoT) and Information Technologies (IT). As modern production methods continue to increase in complexity, there is a greater need to consider what variables can be physically measured. This advancement necessitates the use of physical sensors to comprehensively and directly gather measurable data on industrial processes; specifically, these sensors gather data that can be recontextualized into new process information. For example, artificial intelligence (AI) machine learning-based soft sensors can increase operational productivity and machine tool performance while still ensuring that critical product specifications are met. One industry that has a high volume of labor-intensive, time-consuming, and expensive processes is the semiconductor industry. AI machine learning methods can meet these challenges by taking in operational data and extracting process-specific information needed to meet the high product specifications of the industry. However, a key challenge is the availability of high quality data that covers the full operating range, including the day-to-day variance. This paper examines the applicability of soft sensing methods to the operational data of five industrial etching machines. Data is collected from readily accessible and cost-effective physical sensors installed on the tools that manage and control the operating conditions of the tool. The operational data are then used in an intelligent data aggregation approach that increases the scope and robustness for soft sensors in general by creating larger training datasets comprised of high value data with greater operational ranges and process variation. The generalized soft sensor can then be fine-tuned and validated for a particular machine. In this paper, we test the effects of data aggregation for high performing Feedforward Neural Network (FNN) models that are constructed in two ways: first as a classifier to estimate product PASS/FAIL outcomes and second as a regressor to quantitatively estimate oxide thickness. For PASS/FAIL classification, a data aggregation method is developed to enhance model predictive performance with larger training datasets. A statistical analysis method involving point-biserial correlation and the Mean Absolute Error (MAE) difference score is introduced to select the optimal candidate datasets for aggregation, further improving the effectiveness of data aggregation. For large datasets with high quality data that enable model training for more complex tasks, regression models that predict the oxide thickness of the product are also developed. Two types of models with different loss functions are tested to compare the effects of the Mean Squared Error (MSE) and Mean Absolute Percentage Error (MAPE) loss functions on model performance. Both the classification and regression models can be applied in industrial setti","PeriodicalId":72815,"journal":{"name":"Digital Chemical Engineering","volume":"13 ","pages":"Article 100195"},"PeriodicalIF":3.0,"publicationDate":"2024-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142531729","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-20DOI: 10.1016/j.dche.2024.100192
Maria Victoria Migo-Sumagang , Kathleen B. Aviso , Raymond R. Tan , Xiaoping Jia , Zhiwei Li , Dominic C.Y. Foo
Mitigating climate change requires a portfolio of strategies and the use of carbon dioxide removal techniques or negative emissions technologies (NETs) will be necessary to achieve this goal. However, the high implementation costs of advanced NETs lead to expensive carbon credits, hindering their broad acceptance and use. One potential solution involves governmental support through subsidies, aiming to boost the availability of NET-derived carbon credits. This research uses a graphical technique based on an extension of pinch analysis to identify the ideal subsidy level for carbon dioxide removal, taking into account factors such as carbon pricing, supply, and demand. The proposed approach modifies the limiting composite curve (LCC) methodology to accurately determine the optimal subsidy and establish the baseline amount of subsidized carbon dioxide removal needed. The approach enables the convenient and efficient construction of the LCC using a composite table algorithm. To illustrate the proposed methodology, two case studies composed of different NETs and demand sectors are investigated. The results show the most advantageous subsidy levels for these technologies, providing valuable insights to guide policymakers and investors in their decarbonization efforts. This work contributes to the development of effective governance and investment strategies by optimizing NET subsidy allocation. Such optimization is crucial for facilitating the widespread implementation of these technologies, which are in-line with the global efforts to mitigate climate change.
减缓气候变化需要一系列战略，而使用二氧化碳清除技术或负排放技术（NET）将是实现这一目标的必要条件。然而，先进的负排放技术实施成本高昂，导致碳信用额度昂贵，阻碍了其被广泛接受和使用。一个潜在的解决方案是政府通过补贴提供支持，旨在提高由负向排放技术产生的碳信用额的可用性。本研究采用基于撮合分析扩展的图形技术，在考虑碳定价、供应和需求等因素的基础上，确定二氧化碳清除的理想补贴水平。所提出的方法修改了极限复合曲线（LCC）方法，以准确确定最佳补贴，并确定所需的二氧化碳减排补贴基线量。该方法采用复合表算法，可方便、高效地构建 LCC。为了说明所提出的方法，我们对由不同的净能源和需求部门组成的两个案例进行了研究。结果显示了这些技术最有利的补贴水平，为指导政策制定者和投资者的去碳化工作提供了宝贵的见解。这项工作通过优化 NET 补贴分配，有助于制定有效的治理和投资战略。这种优化对于促进这些技术的广泛实施至关重要，而这些技术与全球减缓气候变化的努力是一致的。
{"title":"Process integration technique for targeting carbon credit price subsidy","authors":"Maria Victoria Migo-Sumagang , Kathleen B. Aviso , Raymond R. Tan , Xiaoping Jia , Zhiwei Li , Dominic C.Y. Foo","doi":"10.1016/j.dche.2024.100192","DOIUrl":"10.1016/j.dche.2024.100192","url":null,"abstract":"<div><div>Mitigating climate change requires a portfolio of strategies and the use of <em>carbon dioxide removal</em> techniques or <em>negative emissions technologies</em> (NETs) will be necessary to achieve this goal. However, the high implementation costs of advanced NETs lead to expensive carbon credits, hindering their broad acceptance and use. One potential solution involves governmental support through subsidies, aiming to boost the availability of NET-derived carbon credits. This research uses a graphical technique based on an extension of <em>pinch analysis</em> to identify the ideal subsidy level for carbon dioxide removal, taking into account factors such as carbon pricing, supply, and demand. The proposed approach modifies the <em>limiting composite curve</em> (LCC) methodology to accurately determine the optimal subsidy and establish the baseline amount of subsidized carbon dioxide removal needed. The approach enables the convenient and efficient construction of the LCC using a composite table algorithm. To illustrate the proposed methodology, two case studies composed of different NETs and demand sectors are investigated. The results show the most advantageous subsidy levels for these technologies, providing valuable insights to guide policymakers and investors in their decarbonization efforts. This work contributes to the development of effective governance and investment strategies by optimizing NET subsidy allocation. Such optimization is crucial for facilitating the widespread implementation of these technologies, which are in-line with the global efforts to mitigate climate change.</div></div>","PeriodicalId":72815,"journal":{"name":"Digital Chemical Engineering","volume":"13 ","pages":"Article 100192"},"PeriodicalIF":3.0,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142656115","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号