Pub Date : 2025-02-06DOI: 10.1016/j.compchemeng.2025.109026
Niklas Nolzen , Alissa Ganter , Nils Baumgärtner , Florian Joseph Baader , Ludger Leenders , André Bardow
The rising share of volatile renewable electricity generation increases the demand for flexibility. Flexibility can be offered by industrial multi-energy systems and marketed either on the continuous intraday, day-ahead, or balancing-power markets. Thus, industrial multi-energy systems face the question where to market their flexibility. We propose a two-step method to integrate trading on the continuous intraday market into a multi-market optimization for flexible industrial multi-energy systems. First, we estimate revenues from continuous trading in the intraday market, employing option-price theory. Second, a multi-stage stochastic optimization allocates the flexibility to the three markets. The case study of an industrial multi-energy system demonstrates that coordinated bidding in all three markets reduces costs the most. A sensitivity analysis reveals that the optimal split between the different markets strongly depends on the intraday market volatility. Overall, the proposed method provides a practical decision-support tool for multi-energy systems participating in short-term electricity and balancing-power markets.
{"title":"Where to market flexibility? Integrating continuous intraday trading into multi-market participation of industrial multi-energy systems","authors":"Niklas Nolzen , Alissa Ganter , Nils Baumgärtner , Florian Joseph Baader , Ludger Leenders , André Bardow","doi":"10.1016/j.compchemeng.2025.109026","DOIUrl":"10.1016/j.compchemeng.2025.109026","url":null,"abstract":"<div><div>The rising share of volatile renewable electricity generation increases the demand for flexibility. Flexibility can be offered by industrial multi-energy systems and marketed either on the continuous intraday, day-ahead, or balancing-power markets. Thus, industrial multi-energy systems face the question where to market their flexibility. We propose a two-step method to integrate trading on the continuous intraday market into a multi-market optimization for flexible industrial multi-energy systems. First, we estimate revenues from continuous trading in the intraday market, employing option-price theory. Second, a multi-stage stochastic optimization allocates the flexibility to the three markets. The case study of an industrial multi-energy system demonstrates that coordinated bidding in all three markets reduces costs the most. A sensitivity analysis reveals that the optimal split between the different markets strongly depends on the intraday market volatility. Overall, the proposed method provides a practical decision-support tool for multi-energy systems participating in short-term electricity and balancing-power markets.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 109026"},"PeriodicalIF":3.9,"publicationDate":"2025-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143377067","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-06DOI: 10.1016/j.compchemeng.2025.109032
Lucky E. Yerimah, Christian Jorgensen, B. Wayne Bequette
Model-free Reinforcement learning (RL) has been successfully used in benchmark systems such as the Cart-Pole, Inverted-Pendulum, and Robotic arms. However, model-free RL algorithms have several limitations, including large data requirements and handling of state constraints. Model-based and hybrid RL algorithms offer opportunities to tackle these limitations. This research investigated the application of a model-based policy optimization algorithm (MBPO) for feedback control of the Van de Vusse reaction and the Quadruple tank system. MBPO-trained agents suffer from inaccuracies of the learned model and the computational burden of the online optimization neural network models and policy parameters. We propose a modified model-based policy optimization (MMBPO) algorithm that uses linear dynamic system models. This minimizes a learned model’s inaccuracies and eliminates the computational requirements of training the neural network models. Simulation results show that model-based policy optimization algorithms can track the setpoints of the dynamic systems studied.
{"title":"Model-based policy optimization algorithms for feedback control of complex dynamic systems","authors":"Lucky E. Yerimah, Christian Jorgensen, B. Wayne Bequette","doi":"10.1016/j.compchemeng.2025.109032","DOIUrl":"10.1016/j.compchemeng.2025.109032","url":null,"abstract":"<div><div>Model-free Reinforcement learning (RL) has been successfully used in benchmark systems such as the Cart-Pole, Inverted-Pendulum, and Robotic arms. However, model-free RL algorithms have several limitations, including large data requirements and handling of state constraints. Model-based and hybrid RL algorithms offer opportunities to tackle these limitations. This research investigated the application of a model-based policy optimization algorithm (MBPO) for feedback control of the Van de Vusse reaction and the Quadruple tank system. MBPO-trained agents suffer from inaccuracies of the learned model and the computational burden of the online optimization neural network models and policy parameters. We propose a modified model-based policy optimization (MMBPO) algorithm that uses linear dynamic system models. This minimizes a learned model’s inaccuracies and eliminates the computational requirements of training the neural network models. Simulation results show that model-based policy optimization algorithms can track the setpoints of the dynamic systems studied.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 109032"},"PeriodicalIF":3.9,"publicationDate":"2025-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143387771","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-05DOI: 10.1016/j.compchemeng.2025.109029
Lingjian Ye , Zeyu Yang , Feifan Shen , Xiaofeng Yuan
In this paper, we design two-dimensional self-optimizing control (2D-SOC) systems for batch processes. In the framework of 2D-SOC, linear combinations of measurements are controlled along the time and batch axis, respectively, which work jointly to achieve near-optimal operation of batch process. Firstly, the global SOC approach is extended to enhance the self-optimizing performance in a wider range of disturbance space. In the presence of active-set changes, an improved solution method is presented to meet the constraint satisfactions. Then, a novel compensation algorithm is proposed to adjust the setpoints of within-batch controlled variables, which can efficiently improve the process optimality in the presence of tracking errors of batch-to-batch controlled variables and active constraint back-offs. A simple linear compensation law is optimally derived. Finally, the enhanced 2D-SOC design approach is systematically applied to two simulated batch processes, where its enhanced performances are verified.
{"title":"Design of enhanced two-dimensional self-optimizing control system for batch process","authors":"Lingjian Ye , Zeyu Yang , Feifan Shen , Xiaofeng Yuan","doi":"10.1016/j.compchemeng.2025.109029","DOIUrl":"10.1016/j.compchemeng.2025.109029","url":null,"abstract":"<div><div>In this paper, we design two-dimensional self-optimizing control (2D-SOC) systems for batch processes. In the framework of 2D-SOC, linear combinations of measurements are controlled along the time and batch axis, respectively, which work jointly to achieve near-optimal operation of batch process. Firstly, the global SOC approach is extended to enhance the self-optimizing performance in a wider range of disturbance space. In the presence of active-set changes, an improved solution method is presented to meet the constraint satisfactions. Then, a novel compensation algorithm is proposed to adjust the setpoints of within-batch controlled variables, which can efficiently improve the process optimality in the presence of tracking errors of batch-to-batch controlled variables and active constraint back-offs. A simple linear compensation law is optimally derived. Finally, the enhanced 2D-SOC design approach is systematically applied to two simulated batch processes, where its enhanced performances are verified.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 109029"},"PeriodicalIF":3.9,"publicationDate":"2025-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143360859","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-04DOI: 10.1016/j.compchemeng.2025.109034
Feilong Zhang , Liangchao Chen , Jianfeng Yang , Pengchao Wang , Jianwen Zhang , Qianlin Wang , Xu Diao , Zhan Dou
In the realm of process plants, the imperative to avert accidental incidents is compounded by the escalating specter of deliberate attacks, predominantly in the form of cyber intrusions. These cyber threats, with their attendant physical risks, are notoriously elusive to quantify, thereby impeding the plants’ ability to adapt swiftly to evolving risk profiles. This paper introduces a game-theoretic framework that translates cyber-assaults on industrial processes into process deviations induced by anomalous control actions, enabling the quantification of risk and the assessment of the cyberattacks’ impact on operational processes. Risk quantification serves as the foundation for the payoffs of both the attackers and the defenders, and it is used to address the probability and severity of incidents through static games characterized by incomplete information. Subsequently, complete information static game theory is employed to calculate the payoffs for both the attacker and the defender. This approach encompasses a spectrum of potential attacks and defenses, yielding optimal economic strategies for the defender across various temporal junctures. Furthermore, a risk tolerance model is integrated to refine the payoff calculation, offering a blueprint for the defender to execute enhanced defensive strategies. The efficacy of the proposed methodology in managing the physical risks emanating from cyberattacks is substantiated through a case study, which scrutinizes a steam stripper and its control system within a catalytic cracking unit.
{"title":"Game-theoretic approach to cybersecurity risk assessment and protective strategy optimization in process industry production systems","authors":"Feilong Zhang , Liangchao Chen , Jianfeng Yang , Pengchao Wang , Jianwen Zhang , Qianlin Wang , Xu Diao , Zhan Dou","doi":"10.1016/j.compchemeng.2025.109034","DOIUrl":"10.1016/j.compchemeng.2025.109034","url":null,"abstract":"<div><div>In the realm of process plants, the imperative to avert accidental incidents is compounded by the escalating specter of deliberate attacks, predominantly in the form of cyber intrusions. These cyber threats, with their attendant physical risks, are notoriously elusive to quantify, thereby impeding the plants’ ability to adapt swiftly to evolving risk profiles. This paper introduces a game-theoretic framework that translates cyber-assaults on industrial processes into process deviations induced by anomalous control actions, enabling the quantification of risk and the assessment of the cyberattacks’ impact on operational processes. Risk quantification serves as the foundation for the payoffs of both the attackers and the defenders, and it is used to address the probability and severity of incidents through static games characterized by incomplete information. Subsequently, complete information static game theory is employed to calculate the payoffs for both the attacker and the defender. This approach encompasses a spectrum of potential attacks and defenses, yielding optimal economic strategies for the defender across various temporal junctures. Furthermore, a risk tolerance model is integrated to refine the payoff calculation, offering a blueprint for the defender to execute enhanced defensive strategies. The efficacy of the proposed methodology in managing the physical risks emanating from cyberattacks is substantiated through a case study, which scrutinizes a steam stripper and its control system within a catalytic cracking unit.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 109034"},"PeriodicalIF":3.9,"publicationDate":"2025-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143377190","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This manuscript identifies, addresses and illustrates via comparisons an inconsistency and inaccuracy with the implementation of Recurrent Neural Networks (RNNs) on naturally occurring dynamical processes, particularly focusing on implementations that claim to identify input–output dynamic relationships through a state–space representation. While the RNN structure does lend itself to these types of problems, there are two major issues with how RNNs are typically structured and trained in this context. Firstly, the hidden states of the model are commonly reinitialized or discarded between each of the input–output sequences in the training data set, essentially leading to a framework where the initial state for each sequence is not trained. In contrast, in a typical state–space model identification framework, the model parameters along with the states are (and need to be) identified together. Secondly, the model structure of the RNN is different from a classic state space (SS) representation. While in state space representations the current state is defined to be a function of the state and input from the previous time step, RNNs use input from the same time step. In this paper, two changes are proposed to address these inconsistencies. The first step is to train the initial hidden states for the training sequences. To address the structural inconsistency between a state space model and the RNN, the list of hidden states retrieved from the RNN is formatted to represent the data and state pairings that a state space model would create. The effect of these corrections is demonstrated in the simplest of dynamical systems — data generated using a Linear Time-Invariant (LTI) state space model. The importance of both these corrections is demonstrated by implementing them one at a time. Interestingly, the model that performed the worst in testing was the model with only the trained hidden states. The model with no changes was slightly better, and the model with the correct input timing but no trained hidden states increased performance by a significant amount. Finally, the best results were found when both changes were implemented.
{"title":"Implementing Recurrent Neural Networks in Process Systems Engineering applications, the right way!","authors":"Aswin Chandrasekar, Tyler Wortley, Euan Bohm, Prashant Mhaskar","doi":"10.1016/j.compchemeng.2025.109027","DOIUrl":"10.1016/j.compchemeng.2025.109027","url":null,"abstract":"<div><div>This manuscript identifies, addresses and illustrates via comparisons an inconsistency and inaccuracy with the implementation of Recurrent Neural Networks (RNNs) on naturally occurring dynamical processes, particularly focusing on implementations that claim to identify input–output dynamic relationships through a state–space representation. While the RNN structure does lend itself to these types of problems, there are two major issues with how RNNs are typically structured and trained in this context. Firstly, the hidden states of the model are commonly reinitialized or discarded between each of the input–output sequences in the training data set, essentially leading to a framework where the initial state for each sequence is not trained. In contrast, in a typical state–space model identification framework, the model parameters along with the states are (and need to be) identified together. Secondly, the model structure of the RNN is different from a classic state space (SS) representation. While in state space representations the current state is defined to be a function of the state and input from the previous time step, RNNs use input from the same time step. In this paper, two changes are proposed to address these inconsistencies. The first step is to train the initial hidden states for the training sequences. To address the structural inconsistency between a state space model and the RNN, the list of hidden states retrieved from the RNN is formatted to represent the data and state pairings that a state space model would create. The effect of these corrections is demonstrated in the simplest of dynamical systems — data generated using a Linear Time-Invariant (LTI) state space model. The importance of both these corrections is demonstrated by implementing them one at a time. Interestingly, the model that performed the worst in testing was the model with only the trained hidden states. The model with no changes was slightly better, and the model with the correct input timing but no trained hidden states increased performance by a significant amount. Finally, the best results were found when both changes were implemented.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 109027"},"PeriodicalIF":3.9,"publicationDate":"2025-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143348537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-03DOI: 10.1016/j.compchemeng.2025.109033
Rofice Dickson , Seyed Soheil Mansouri
Feeding the growing global population sustainably while minimizing environmental impact is grand challenge for human society. Methane-based single-cell proteins through fermentation (bio-SCP) have emerged as a promising alternative to traditional protein sources (animal and crops), addressing the significant greenhouse gas emissions from livestock production. This study explores an innovative approach to bio-SCP production using synthetic natural gas derived from biogas. The process integrates biogas production via anaerobic digestion of food waste, biogas upgrading through a series of treatments, and SCP production via aerobic fermentation of methane. Detailed process modeling reveals that the proposed design consumes 25,000 kg/h (200 Mt/y) of food waste, producing 4,269.4 kg/h (34.2 Mt/y) of SCP and valuable by-products such as biofertilizer, elemental sulfur, low-pressure steam, and nitrogen. Notably, the proposed design achieves close to 100 % energy self-sufficiency. Techno-economic analysis indicates a capital investment of $733.5 million, annual operating costs of $43.96 million, and a minimum product selling price of $1.02/kg of bio-SCP, demonstrating promising economic viability, especially with nitrogen by-product sales. A cradle-to-gate life cycle assessment highlights the environmental benefits of bio-SCP, showing significant reductions in environmental impacts compared to fossil-driven SCP production. This study underscores the potential of bio-SCP in sustainable animal nutrition and greenhouse gas emission reduction.
{"title":"Sustainable production of fermentation-based novel proteins","authors":"Rofice Dickson , Seyed Soheil Mansouri","doi":"10.1016/j.compchemeng.2025.109033","DOIUrl":"10.1016/j.compchemeng.2025.109033","url":null,"abstract":"<div><div>Feeding the growing global population sustainably while minimizing environmental impact is grand challenge for human society. Methane-based single-cell proteins through fermentation (bio-SCP) have emerged as a promising alternative to traditional protein sources (animal and crops), addressing the significant greenhouse gas emissions from livestock production. This study explores an innovative approach to bio-SCP production using synthetic natural gas derived from biogas. The process integrates biogas production via anaerobic digestion of food waste, biogas upgrading through a series of treatments, and SCP production via aerobic fermentation of methane. Detailed process modeling reveals that the proposed design consumes 25,000 kg/h (200 Mt/y) of food waste, producing 4,269.4 kg/h (34.2 Mt/y) of SCP and valuable by-products such as biofertilizer, elemental sulfur, low-pressure steam, and nitrogen. Notably, the proposed design achieves close to 100 % energy self-sufficiency. Techno-economic analysis indicates a capital investment of $733.5 million, annual operating costs of $43.96 million, and a minimum product selling price of $1.02/kg of bio-SCP, demonstrating promising economic viability, especially with nitrogen by-product sales. A cradle-to-gate life cycle assessment highlights the environmental benefits of bio-SCP, showing significant reductions in environmental impacts compared to fossil-driven SCP production. This study underscores the potential of bio-SCP in sustainable animal nutrition and greenhouse gas emission reduction.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 109033"},"PeriodicalIF":3.9,"publicationDate":"2025-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143349507","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-31DOI: 10.1016/j.compchemeng.2025.109030
Sebastian Andrade-Becerra , Lisette Samarti-Rios , Jesus Alvarez , Luis Alvarez-Icaza
This paper addresses the joint design of the nominal state motion and feed-forward (FF) output-feedback (OF) tracking control for a class of exothermic batch reactors, using a hydrothermal carbonization (HTC) reactor as a case study. The objective is to design the nominal state motion to maximize profit while incorporating a control strategy that includes an event-based component to stop the batch. The reactor's sustained evolution and profit per unit time are modeled by the state motion of a nonlinear, non-autonomous ordinary differential equation (ODE) over a finite-time interval. The problem is solved within a constructive control framework, combining concepts from chemical reactor engineering and nonlinear control theory, tailored to the characteristics of batch reactors. First, the nominal optimal operation is designed with a balanced compromise between duration, robustness, and heat exchange load, through: (i) the identification of passivity and detectability solvability conditions for the OF tracking control, and (ii) a simple recursive numerical construction approach. Then, a nonlinear FF OF tracking-termination control is developed, ensuring closed-loop (CL) robust motion stability, along with a straightforward gain tuning scheme. The proposed methodology is demonstrated and validated through numerical simulation with a representative HTC reactor example.
{"title":"Joint operation-control design for a class of batch exothermic reactors","authors":"Sebastian Andrade-Becerra , Lisette Samarti-Rios , Jesus Alvarez , Luis Alvarez-Icaza","doi":"10.1016/j.compchemeng.2025.109030","DOIUrl":"10.1016/j.compchemeng.2025.109030","url":null,"abstract":"<div><div>This paper addresses the joint design of the nominal state motion and feed-forward (FF) output-feedback (OF) tracking control for a class of exothermic batch reactors, using a hydrothermal carbonization (HTC) reactor as a case study. The objective is to design the nominal state motion to maximize profit while incorporating a control strategy that includes an event-based component to stop the batch. The reactor's sustained evolution and profit per unit time are modeled by the state motion of a nonlinear, non-autonomous ordinary differential equation (ODE) over a finite-time interval. The problem is solved within a constructive control framework, combining concepts from chemical reactor engineering and nonlinear control theory, tailored to the characteristics of batch reactors. First, the nominal optimal operation is designed with a balanced compromise between duration, robustness, and heat exchange load, through: (i) the identification of passivity and detectability solvability conditions for the OF tracking control, and (ii) a simple recursive numerical construction approach. Then, a nonlinear FF OF tracking-termination control is developed, ensuring closed-loop (CL) robust motion stability, along with a straightforward gain tuning scheme. The proposed methodology is demonstrated and validated through numerical simulation with a representative HTC reactor example.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"196 ","pages":"Article 109030"},"PeriodicalIF":3.9,"publicationDate":"2025-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143394354","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-30DOI: 10.1016/j.compchemeng.2025.109021
Konstantinos Katsoulas, Federico Galvanin, Luca Mazzei, Maximilian Besenhard, Eva Sorensen
Chromatography is a key purification process in the pharmaceutical industry. The process design is based on knowledge of the adsorption isotherm that describes the separation within the chromatographic column. Although obtaining the values of isotherm model parameters has traditionally been the work of experimentalists, recently design methods based on mathematical models have emerged, and for these, accurate isotherm models and model parameter values are crucial. Different methods exist for parameter estimation, all depending on experiment execution. Model-Based Design of Experiments (MBDoE) can be used to optimally design experiments that maximise the information obtained from each experiment. In this work, we propose an MBDoE-based methodology that aims to identify the most suitable isotherm model, to estimate its parameters, and to evaluate its predictive capability. The methodology is tested on an in-silico case study where the performance is compared to that of traditional factorial design of experiments.
{"title":"Model-based design of experiments for efficient and accurate isotherm model identification in High Performance Liquid Chromatography","authors":"Konstantinos Katsoulas, Federico Galvanin, Luca Mazzei, Maximilian Besenhard, Eva Sorensen","doi":"10.1016/j.compchemeng.2025.109021","DOIUrl":"10.1016/j.compchemeng.2025.109021","url":null,"abstract":"<div><div>Chromatography is a key purification process in the pharmaceutical industry. The process design is based on knowledge of the adsorption isotherm that describes the separation within the chromatographic column. Although obtaining the values of isotherm model parameters has traditionally been the work of experimentalists, recently design methods based on mathematical models have emerged, and for these, accurate isotherm models and model parameter values are crucial. Different methods exist for parameter estimation, all depending on experiment execution. Model-Based Design of Experiments (MBDoE) can be used to optimally design experiments that maximise the information obtained from each experiment. In this work, we propose an MBDoE-based methodology that aims to identify the most suitable isotherm model, to estimate its parameters, and to evaluate its predictive capability. The methodology is tested on an in-silico case study where the performance is compared to that of traditional factorial design of experiments.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 109021"},"PeriodicalIF":3.9,"publicationDate":"2025-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143348536","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-30DOI: 10.1016/j.compchemeng.2025.109013
Miriam Sarkis, Nilay Shah, Maria M. Papathanasiou
In recent years, the market boom of next-generation therapies and vaccines has pressured the pharmaceutical industry to rapidly scale up capacity to meet societal needs. Manufacturers catering for these markets reported shortages due to unforeseen demand trends and a crucial uncertainty in capabilities of platforms still under development. In this work, we present an optimization-simulation framework for the design of resilient supply chains to manufacturing uncertainty. Given previously quantified probability distributions of process parameters, we formulate stochastic optimization problems integrating process uncertainty via a sampling-based methodology. Stochastic programming results in networks of higher optimal costs compared to deterministic approaches. Furthermore, stochastic designs ensure product supply meets target demands under simulated uncertainty and result in a larger probability of achieving lower costs per dose. The optimization-simulation framework is used to test solution stability for a varying number of optimization scenarios, highlighting that the minimum number of samples to guarantee stability is problem-specific, thus motivating the investigation of scenario reduction techniques to ensure stability of scenario sets a priori. Overall, the cost-supply benefits of integrating manufacturing uncertainty are quantified, demonstrating the scope for its consideration in strategic planning problems in the sector.
{"title":"Resilient pharmaceutical supply chains: Assessment of stochastic optimization strategies for process uncertainty integration in network design problems","authors":"Miriam Sarkis, Nilay Shah, Maria M. Papathanasiou","doi":"10.1016/j.compchemeng.2025.109013","DOIUrl":"10.1016/j.compchemeng.2025.109013","url":null,"abstract":"<div><div>In recent years, the market boom of next-generation therapies and vaccines has pressured the pharmaceutical industry to rapidly scale up capacity to meet societal needs. Manufacturers catering for these markets reported shortages due to unforeseen demand trends and a crucial uncertainty in capabilities of platforms still under development. In this work, we present an optimization-simulation framework for the design of resilient supply chains to manufacturing uncertainty. Given previously quantified probability distributions of process parameters, we formulate stochastic optimization problems integrating process uncertainty via a sampling-based methodology. Stochastic programming results in networks of higher optimal costs compared to deterministic approaches. Furthermore, stochastic designs ensure product supply meets target demands under simulated uncertainty and result in a larger probability of achieving lower costs per dose. The optimization-simulation framework is used to test solution stability for a varying number of optimization scenarios, highlighting that the minimum number of samples to guarantee stability is problem-specific, thus motivating the investigation of scenario reduction techniques to ensure stability of scenario sets <em>a priori</em>. Overall, the cost-supply benefits of integrating manufacturing uncertainty are quantified, demonstrating the scope for its consideration in strategic planning problems in the sector.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 109013"},"PeriodicalIF":3.9,"publicationDate":"2025-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143349789","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-27DOI: 10.1016/j.compchemeng.2025.109011
Avan Kumar , Harshitha Chandra Jami , Bhavik R. Bakshi , Manojkumar Ramteke , Hariprasad Kodamana
Polyethylene terephthalate (PET) is valued for its durability, tensile strength, low moisture absorption, and cost-effectiveness. However, its non-biodegradability poses an environmental threat, and plastic recycling is the sole remedy. This study proposes an NLP framework for concisely extracting and summarizing key information on recycling technologies and alternatives from relevant scientific literature. This NLP framework comprises three approaches: time-series knowledge graphs, dynamic transformer-based topic modeling, and estimating popularity indices for technologies. The framework aims to streamline the extraction of qualitative and quantitative insights for sustainable and economical PET waste recycling pathways. Key findings of the study show that there is a 406% rise in pyrolysis technology use, a 278% increase in chemical conversion, and a 1353% surge in waste PET utilization for electronic device-making. It is worth noting that some of the identified recycling pathways corroborate well with the actual implementation in the industries.
{"title":"An evolutionary study on technologies for polyethylene terephthalate waste recycling using natural language processing","authors":"Avan Kumar , Harshitha Chandra Jami , Bhavik R. Bakshi , Manojkumar Ramteke , Hariprasad Kodamana","doi":"10.1016/j.compchemeng.2025.109011","DOIUrl":"10.1016/j.compchemeng.2025.109011","url":null,"abstract":"<div><div>Polyethylene terephthalate (PET) is valued for its durability, tensile strength, low moisture absorption, and cost-effectiveness. However, its non-biodegradability poses an environmental threat, and plastic recycling is the sole remedy. This study proposes an NLP framework for concisely extracting and summarizing key information on recycling technologies and alternatives from relevant scientific literature. This NLP framework comprises three approaches: time-series knowledge graphs, dynamic transformer-based topic modeling, and estimating popularity indices for technologies. The framework aims to streamline the extraction of qualitative and quantitative insights for sustainable and economical PET waste recycling pathways. Key findings of the study show that there is a 406% rise in pyrolysis technology use, a 278% increase in chemical conversion, and a 1353% surge in waste PET utilization for electronic device-making. It is worth noting that some of the identified recycling pathways corroborate well with the actual implementation in the industries.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 109011"},"PeriodicalIF":3.9,"publicationDate":"2025-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143349506","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","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