Chang He
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Chang He has a total of 114 co-authors. Here are the top co-authors by number of shared publications:
Bingjian Zhang
20 shared publicationsQinglin Chen
20 shared publicationsJingzheng Ren
16 shared publications- Jingzheng Ren
Department of Industrial and Systems Engineering The Hong Kong Polytechnic University Hong Kong SAR China
16 shared publications Haoshui Yu
7 shared publications- Bing J. Zhang
The Key Laboratory of Low-Carbon Chemistry & Energy Conservation of Guangdong Province, Guangdong Engineering Center for Petrochemical Energy Conservation, Sun Yat-sen University, Guangzhou 510006, China
6 shared publications
Latest Publications
- 10 November 2026
- Published by American Chemical Society (ACS) in ACS Sustainable Chemistry & Engineering
Converting Waste Tires into Blue Ammonia: Safe and Sustainable Process Design, Multiobjective Optimization and Comprehensive Assessment
Ammonia synthesis based on steam methane reforming (SMR) is characterized by intensive energy consumption and high carbon emissions. To address these limitations, novel processes integrating gasification and carbon capture technologies were proposed for converting waste tires into blue ammonia. Detailed safety, techno-economic, and life cycle assessments were conducted to evaluate comprehensive performance. Machine learning-driven multiobjective operation optimization was carried out for reducing levelized cost of ammonia (LCOA) while enhancing safety and sustainability scores based on safe-and-sustainable-by-design framework. Our results highlight the critical role of artificial neural networks in achieving efficient optimization. Additionally, the quantitative impacts of integrating various renewable energy sources were investigated. The results showed the conventional gasification (CG) integrating methyldiethanolamine (MDEA)-based carbon capture alternative achieved the optimal LCOA at 358.1 $/ton in current electricity scenario. A peak environmental sustainability score of 0.443 was achieved, outperforming traditional SMR integrating MDEA-based carbon capture pathway by 13.04%. The safest process is CG integrating monoethanolamine-based carbon capture. Under the renewable scenario, the LCA results highlighted the critical role of renewable energy integration in enhancing the environmental sustainability of waste tires-to-ammonia processes, especially achieving a 0.344 increase in environmental sustainability score with onshore wind electricity at a lower LCOA of 406.0 $/ton in plasma gasification integrating MDEA-based alternative.
- 10 November 2026
- Published by American Chemical Society (ACS) in ACS Sustainable Chemistry & Engineering
Converting Waste Tires into Blue Ammonia: Safe and Sustainable Process Design, Multiobjective Optimization and Comprehensive Assessment
Ammonia synthesis based on steam methane reforming (SMR) is characterized by intensive energy consumption and high carbon emissions. To address these limitations, novel processes integrating gasification and carbon capture technologies were proposed for converting waste tires into blue ammonia. Detailed safety, techno-economic, and life cycle assessments were conducted to evaluate comprehensive performance. Machine learning-driven multiobjective operation optimization was carried out for reducing levelized cost of ammonia (LCOA) while enhancing safety and sustainability scores based on safe-and-sustainable-by-design framework. Our results highlight the critical role of artificial neural networks in achieving efficient optimization. Additionally, the quantitative impacts of integrating various renewable energy sources were investigated. The results showed the conventional gasification (CG) integrating methyldiethanolamine (MDEA)-based carbon capture alternative achieved the optimal LCOA at 358.1 $/ton in current electricity scenario. A peak environmental sustainability score of 0.443 was achieved, outperforming traditional SMR integrating MDEA-based carbon capture pathway by 13.04%. The safest process is CG integrating monoethanolamine-based carbon capture. Under the renewable scenario, the LCA results highlighted the critical role of renewable energy integration in enhancing the environmental sustainability of waste tires-to-ammonia processes, especially achieving a 0.344 increase in environmental sustainability score with onshore wind electricity at a lower LCOA of 406.0 $/ton in plasma gasification integrating MDEA-based alternative.
- 27 August 2026
- Published by American Chemical Society (ACS) in ACS Sustainable Chemistry & Engineering
- Vol. 14 (35) , 16009-16025
CAPE-OPEN-Enabled Physics-Informed Surrogate Modeling and Techno-Economic Assessment of Membrane-Based Methane Dehydroaromatization
Methane dehydroaromatization (MDA) directly converts methane into aromatics, but its process-level assessment remains challenging because of coupled reaction–permeation behavior and limited interoperability with commercial flowsheet simulators. Here, a CAPE-OPEN-compliant, physics-informed surrogate model was developed for an MDA Pd-membrane reactor and deployed in Aspen Plus as a callable unit operation for recycle-integrated flowsheet simulation and techno-economic assessment. The framework integrated physics-informed learning, transfer-learning-assisted parameterization, generalized polynomial chaos surrogate modeling, and a first-of-a-kind (FOAK) techno-economic assessment for an early-deployment scenario under conservative assumptions. Across four hold-out validation cases, the generalized polynomial chaos surrogate achieved a mean relative error of 1.75% and a maximum relative error of 4.40%. Three representative schemes were evaluated to quantify the propagation of reactor-level trade-offs to plant-level production and economics. The high-yield scheme achieved the highest single-pass CH4 conversion, whereas the low-coking scheme provided the highest plant-level benzene capacity but required the largest utility operating expenditure because of its higher recycle and separation burden. Under the base-case assumptions, the three schemes yielded operating-level annual profits of 61.93–63.43 MM$ year–1, with a FOAK-adjusted total capital investment of 229.35 MM$. Sensitivity analysis identified methane price, membrane-event cost, and membrane service interval as the dominant economic drivers.
- 27 August 2026
- Published by American Chemical Society (ACS) in ACS Sustainable Chemistry & Engineering
- Vol. 14 (35) , 16009-16025
CAPE-OPEN-Enabled Physics-Informed Surrogate Modeling and Techno-Economic Assessment of Membrane-Based Methane Dehydroaromatization
Methane dehydroaromatization (MDA) directly converts methane into aromatics, but its process-level assessment remains challenging because of coupled reaction–permeation behavior and limited interoperability with commercial flowsheet simulators. Here, a CAPE-OPEN-compliant, physics-informed surrogate model was developed for an MDA Pd-membrane reactor and deployed in Aspen Plus as a callable unit operation for recycle-integrated flowsheet simulation and techno-economic assessment. The framework integrated physics-informed learning, transfer-learning-assisted parameterization, generalized polynomial chaos surrogate modeling, and a first-of-a-kind (FOAK) techno-economic assessment for an early-deployment scenario under conservative assumptions. Across four hold-out validation cases, the generalized polynomial chaos surrogate achieved a mean relative error of 1.75% and a maximum relative error of 4.40%. Three representative schemes were evaluated to quantify the propagation of reactor-level trade-offs to plant-level production and economics. The high-yield scheme achieved the highest single-pass CH4 conversion, whereas the low-coking scheme provided the highest plant-level benzene capacity but required the largest utility operating expenditure because of its higher recycle and separation burden. Under the base-case assumptions, the three schemes yielded operating-level annual profits of 61.93–63.43 MM$ year–1, with a FOAK-adjusted total capital investment of 229.35 MM$. Sensitivity analysis identified methane price, membrane-event cost, and membrane service interval as the dominant economic drivers.
- 22 August 2026
- Published by American Chemical Society (ACS) in Industrial & Engineering Chemistry Research
Biobjective Optimization for Demand-Driven Green Electrification Retrofit of Petrochemical Steam Power Systems under Multiple Operating Conditions
Petrochemical plant steam power systems (SPSs) pursue higher sustainability with more green energy integration, but frequently suffer from an energy supply demand imbalance caused by production-side uncertainties, including load fluctuations, equipment changes, process optimization, and seasonal variations. Coupling green energy-enabled supply with uncertain demand while balancing system energy efficiency, economic performance, and trade-offs between electrification and steam utilization remains a critical challenge. This work proposes a demand-driven sustainable retrofit framework covering demand-side steam-electricity balance and interchange modeling, high-efficiency system structure reconstruction, and plant-wide supply demand rebalance optimization. A universal steam generation and consumption model is developed to quantify steam-electricity demand variations under disturbances. Green electrification schemes are designed, and eight operating scenarios are established. Subsequently, a biobjective mixed-integer nonlinear programming model is formulated to maximize economic benefits and green electricity penetration. Results show that under normal operating conditions, the proposed method increases green electricity penetration from roughly 24% to 70%, lifts the heat-to-power ratio from 4.2 to 18.0, and improves system energy efficiency from 60.3% to 84.9% compared with the original system, while directing SPSs toward a heat-dominated operation pattern. Additionally, coal consumption is reduced by nearly 73.3% to cut system carbon emissions, and all performance is notably improved across other scenarios as well. The synergistic supply demand optimization offsets local thermodynamic disadvantages of electrification and balances deep decarbonization with economic performance. This study provides a practical technical route for the decarbonization and sustainable retrofit of petrochemical SPSs.
- 22 August 2026
- Published by American Chemical Society (ACS) in Industrial & Engineering Chemistry Research
Biobjective Optimization for Demand-Driven Green Electrification Retrofit of Petrochemical Steam Power Systems under Multiple Operating Conditions
Petrochemical plant steam power systems (SPSs) pursue higher sustainability with more green energy integration, but frequently suffer from an energy supply demand imbalance caused by production-side uncertainties, including load fluctuations, equipment changes, process optimization, and seasonal variations. Coupling green energy-enabled supply with uncertain demand while balancing system energy efficiency, economic performance, and trade-offs between electrification and steam utilization remains a critical challenge. This work proposes a demand-driven sustainable retrofit framework covering demand-side steam-electricity balance and interchange modeling, high-efficiency system structure reconstruction, and plant-wide supply demand rebalance optimization. A universal steam generation and consumption model is developed to quantify steam-electricity demand variations under disturbances. Green electrification schemes are designed, and eight operating scenarios are established. Subsequently, a biobjective mixed-integer nonlinear programming model is formulated to maximize economic benefits and green electricity penetration. Results show that under normal operating conditions, the proposed method increases green electricity penetration from roughly 24% to 70%, lifts the heat-to-power ratio from 4.2 to 18.0, and improves system energy efficiency from 60.3% to 84.9% compared with the original system, while directing SPSs toward a heat-dominated operation pattern. Additionally, coal consumption is reduced by nearly 73.3% to cut system carbon emissions, and all performance is notably improved across other scenarios as well. The synergistic supply demand optimization offsets local thermodynamic disadvantages of electrification and balances deep decarbonization with economic performance. This study provides a practical technical route for the decarbonization and sustainable retrofit of petrochemical SPSs.
- 20 July 2026
- Published by Wiley in AIChE Journal
- Vol. 72 (10) , e70555
Hydrodynamic regulation of flow‐electrode capacitive deionization by Tesla‐type flow channels for lithium extraction
Flow‐electrode capacitive deionization (FCDI) for lithium extraction is often limited by coupled slurry hydrodynamics, particle redistribution, and electrochemical transport, especially at high solid loading. Here, Tesla‐type asymmetric flow channels were introduced to regulate slurry transport and intensify charge/ion transfer in FCDI. Computational fluid dynamics simulations showed improved flow uniformity, suppressed particle accumulation, and reduced flow resistance relative to a serpentine channel. Electrochemical impedance spectroscopy and distribution of relaxation times analyses further indicated lower transport resistance and enhanced coupled charge‐transfer/ion‐migration processes. Consequently, the Tesla‐type system achieved an average salt removal rate of 0.327 μmol cm −2 min −1 , and a charge efficiency of 81.2% at 1.2 V. After optimization, the system reached 0.868 μmol cm −2 min −1 and 93.9% charge efficiency at 1.6 V. These results demonstrate a device‐level transport intensification strategy for slurry‐electrode electrochemical separations.
- 20 July 2026
- Published by Wiley in AIChE Journal
- Vol. 72 (10) , e70555
Hydrodynamic regulation of flow‐electrode capacitive deionization by Tesla‐type flow channels for lithium extraction
Flow‐electrode capacitive deionization (FCDI) for lithium extraction is often limited by coupled slurry hydrodynamics, particle redistribution, and electrochemical transport, especially at high solid loading. Here, Tesla‐type asymmetric flow channels were introduced to regulate slurry transport and intensify charge/ion transfer in FCDI. Computational fluid dynamics simulations showed improved flow uniformity, suppressed particle accumulation, and reduced flow resistance relative to a serpentine channel. Electrochemical impedance spectroscopy and distribution of relaxation times analyses further indicated lower transport resistance and enhanced coupled charge‐transfer/ion‐migration processes. Consequently, the Tesla‐type system achieved an average salt removal rate of 0.327 μmol cm −2 min −1 , and a charge efficiency of 81.2% at 1.2 V. After optimization, the system reached 0.868 μmol cm −2 min −1 and 93.9% charge efficiency at 1.6 V. These results demonstrate a device‐level transport intensification strategy for slurry‐electrode electrochemical separations.
- 6 May 2026
- Published by American Chemical Society (ACS) in Industrial & Engineering Chemistry Research
- Vol. 65 (19) , 10082-10097
Feasibility Study on Reactor Preliminary Design via Fine-Tuning of Large Language Models
- 6 May 2026
- Published by American Chemical Society (ACS) in Industrial & Engineering Chemistry Research
- Vol. 65 (19) , 10082-10097
Feasibility Study on Reactor Preliminary Design via Fine-Tuning of Large Language Models
- Open Access
Modelling and Optimization of Petrochemical Hybrid Renewable Energy Systems Considering Energy Interchangeability, Uncertainty and Storage for Coupling Energy Supply and Utilization Sides
Petrochemical hybrid renewable energy systems (PHRESs), integrating renewable and fossil energy sources, have garnered more and more attention for sustainable manufacturing. However, achieving concurrent optimization of energy supply reliability and carbon mitigation in these complex systems remains a critical challenge. This study proposes an innovative bilateral optimization framework coupling supply-side energy management with demand-side flexibility. On the supply side, a scenario-based two-stage stochastic programming method synergizes with energy storage systems to address renewable energy intermittency, considering a time-of-day tariff from the grid. On the utilization side, heat energy-based and shaft work-based energy interchangeability are introduced and leveraged to enable both qualitative and quantitative flexibility in process unit requirements and thus obtain energy consumption relaxation models for relaxing the design boundaries of PHRESs. These dual strategies are then coupled in a two-stage mixed-integer programming model framework for the optimal design of PHRESs. Applied to a large-scale refinery incorporating carbon taxation and dynamic electricity price, the proposed methodology demonstrates superior performance through five comparative cases. Compared to the Base Case, the Optimal Case using the proposed method can reduce the total annual cost by 14.82%, and stochastic programming reveals over a 40% probability of carbon mitigation in the uncertain space.
Open AccessModelling and Optimization of Petrochemical Hybrid Renewable Energy Systems Considering Energy Interchangeability, Uncertainty and Storage for Coupling Energy Supply and Utilization Sides
Petrochemical hybrid renewable energy systems (PHRESs), integrating renewable and fossil energy sources, have garnered more and more attention for sustainable manufacturing. However, achieving concurrent optimization of energy supply reliability and carbon mitigation in these complex systems remains a critical challenge. This study proposes an innovative bilateral optimization framework coupling supply-side energy management with demand-side flexibility. On the supply side, a scenario-based two-stage stochastic programming method synergizes with energy storage systems to address renewable energy intermittency, considering a time-of-day tariff from the grid. On the utilization side, heat energy-based and shaft work-based energy interchangeability are introduced and leveraged to enable both qualitative and quantitative flexibility in process unit requirements and thus obtain energy consumption relaxation models for relaxing the design boundaries of PHRESs. These dual strategies are then coupled in a two-stage mixed-integer programming model framework for the optimal design of PHRESs. Applied to a large-scale refinery incorporating carbon taxation and dynamic electricity price, the proposed methodology demonstrates superior performance through five comparative cases. Compared to the Base Case, the Optimal Case using the proposed method can reduce the total annual cost by 14.82%, and stochastic programming reveals over a 40% probability of carbon mitigation in the uncertain space.
- 20 November 2025
- Published by Elsevier in Chemical Engineering Science
- Vol. 321
A multimodal foundation model of multi-physics for chemical reactor design
- 20 November 2025
- Published by Elsevier in Chemical Engineering Science
- Vol. 321
A multimodal foundation model of multi-physics for chemical reactor design
- Open Access
A Comprehensive Study of Machine Learning for Waste-to-Energy Process Modeling and Optimization
This study presents a comprehensive study integrating machine learning, life cycle assessment (LCA) and heuristic optimization to achieve a low-carbon medical waste (MW)-to fuel process. A detailed process simulation coupled with cradle to gate LCA is employed to generate a dataset covering diverse process operation conditions, embodied carbon of supplying H2 and the associated carbon emission factor of MW treatment (CEF). Four machine learning techniques, including support vector machine, artificial neural network, Gaussian process regression, and XGBoost, are trained, each achieving test R2 close to 0.90 and RMSE of ~0.26. These models are integrated with heuristic algorithms to optimize operating parameters under various green hydrogen mixes (20–80%). Our results show that machine learning models outperform the detailed process model (DPM), achieving a minimum CEF of ~1.3 to ~1.1 kg CO2-eq/kg MW with higher computational stabilities. Importantly, the optimization times dropped from hours (DPM) to seconds (machine learning models) and the combination of Gaussian process regression and particle swarm optimization is highlighted, with an optimization time under one second. The optimized process holds promise in carbon reduction compared to traditional MW disposal methods. These findings show machine learning can achieve high predictive accuracy while dramatically enhancing optimization speed and stability, providing a scalable framework for extensive scenario analysis during waste-to-energy process design and further real-time optimization application.
Open AccessA Comprehensive Study of Machine Learning for Waste-to-Energy Process Modeling and Optimization
This study presents a comprehensive study integrating machine learning, life cycle assessment (LCA) and heuristic optimization to achieve a low-carbon medical waste (MW)-to fuel process. A detailed process simulation coupled with cradle to gate LCA is employed to generate a dataset covering diverse process operation conditions, embodied carbon of supplying H2 and the associated carbon emission factor of MW treatment (CEF). Four machine learning techniques, including support vector machine, artificial neural network, Gaussian process regression, and XGBoost, are trained, each achieving test R2 close to 0.90 and RMSE of ~0.26. These models are integrated with heuristic algorithms to optimize operating parameters under various green hydrogen mixes (20–80%). Our results show that machine learning models outperform the detailed process model (DPM), achieving a minimum CEF of ~1.3 to ~1.1 kg CO2-eq/kg MW with higher computational stabilities. Importantly, the optimization times dropped from hours (DPM) to seconds (machine learning models) and the combination of Gaussian process regression and particle swarm optimization is highlighted, with an optimization time under one second. The optimized process holds promise in carbon reduction compared to traditional MW disposal methods. These findings show machine learning can achieve high predictive accuracy while dramatically enhancing optimization speed and stability, providing a scalable framework for extensive scenario analysis during waste-to-energy process design and further real-time optimization application.