The design of geometric parameters in heat exchangers exerts a notable influence on their heat transfer capabilities and fluid dynamic behaviors. Conventional optimization approaches usually depend on empirical formulas or numerical simulation methods, which are plagued by excessive computational expenses and low operational efficiency. To tackle this issue, this research puts forward a machine learning-driven optimization algorithm tailored for the distributed geometric parameters of heat exchangers. First, computational fluid dynamics (CFD) simulation technology is used to produce a large volume of sample data, which includes both the geometric parameters of heat exchangers and their corresponding performance metrics. This sample data is then used to build a comprehensive dataset. Next, machine learning tools—such as neural networks and random forests—are applied to develop prediction models that establish relationships between the geometric parameters of heat exchangers and their performance indexes. Finally, the study integrates intelligent optimization algorithms (e.g., genetic algorithms and particle swarm optimization) to realize the multi-objective optimization design of heat exchanger geometric parameters. The research findings indicate that this proposed method can remarkably boost optimization efficiency. Moreover, the optimized heat exchanger design not only enhances heat transfer performance but also effectively controls flow resistance. This provides a novel technical route for heat exchanger design, endowing the research with substantial theoretical significance and practical value in engineering applications.

Spiral-wound heat exchangers (SWHEs) offer high heat transfer efficiency and compact design advantages, making them well-suited for services in process industries. Accelerating the application of SWHEs demands design methodologies that avoid extensive user manipulations and complex solution procedures. This study develops a novel incremental-based heat transfer framework for the automated design of single-phase SWHEs, which simultaneously optimizes multi-stream allocation across activated tube layers and exchanger geometries. At each increment, energy balances are enforced for all streams using local heat transfer coefficients and areas. On the tube-side, flow distribution is optimized by permitting variable split heat capacities and mass flow rates within tube layers while ensuring pressure balance for each stream at the bundle outlet. New correlations for shell-side flow regimes are introduced into the proposed sizing model to link discrete tube-layer selections with their corresponding cross-sectional areas throughout the optimization process. The capability of the proposed framework is demonstrated through four case studies, including model validation, two-stream and multi-stream SWHE design, and application to an industrial-scale heat exchanger network (HEN). Rigorous Aspen EDR-CoilWound simulations validate the proposed model and design results, with the HEN case exhibiting only a 2.95 % deviation from the target duty. In Case Study 2, SWHE results in a 24.29 % reduction in required heat transfer area. Case Studies 3 and 4 demonstrate that SWHE configurations can achieve 31.8 %-40.7 % reductions in exchanger volume, attributable to their superior compactness relative to conventional shell-and-tube heat exchangers (STHEs). Benchmarking against detailed STHE designs further clarifies optimal deployment strategies and highlights residual limitations of SWHE technology.

Vertical thermosyphon reboilers (VTRs) are widely used in distillation due to their high heat transfer efficiency and low fouling tendency. However, conventional design approaches and commercial tools often treat two-phase heat transfer, geometry, and pressure balance in isolation and depend on extensive manual tuning, which may yield low solution quality. To address this gap, a rigorous mixed integer nonlinear programming (MINLP) optimization framework is established for VTR design that couples thermodynamic, hydraulic, and geometric decisions simultaneously. A two-stage heat transfer modelling approach is proposed to capture the convective boiling transition by integrating sensible heating and boiling heat transfer mechanisms. Within this approach, a dynamic switching scheme selects heat transfer correlations according to the vapor fraction, which improves accuracy across different operating conditions. Moreover, a comprehensive pressure balance model is established that spans the external piping, the reboiler, and the column sump liquid level to ensure the resulting natural circulation is stable. Geometric and operating variables are optimized simultaneously to deliver coordinated design choices under duty and layout constraints. The optimization is implemented in GAMS/48 and solved using SCIP solver. To demonstrate engineering applicability, the thermodynamic modelling results are compared with Aspen EDR simulations, revealing relative errors ranging from 3.8 % to 15.6 %. The optimized design increases the overall heat transfer coefficient by 16.60 % and reduces the required area by 13.86 %. Furthermore, stability analysis under varying heat duties reveals clear operational boundaries, highlighting the importance of coordinating liquid level and skirt height to maintain feasible natural circulation.

Retrofitting industrial heat recovery systems is a key pathway toward low-carbon and resource-efficient manufacturing. Integrating waste heat recovery (WR) with multiple utility (MU) networks can significantly improve energy efficiency, but also creates strong coupling between process heat recovery, waste heat utilisation, and utility allocation. However, existing approaches often address these elements separately or sequentially, limiting the ability to identify optimal system-level retrofit solutions. This study aims to develop a system-level optimisation framework to resolve trade-offs between waste heat recovery and multiple utility integration in industrial retrofit. The framework allows for (i) modifying existing exchangers, (ii) installing new process-to-process units to enhance heat recovery, (iii) integrating waste heat recovery systems for generating high-pressure steam, low-pressure steam, and hot water, and (iv) incorporating optimal utilisation of multiple utilities. The framework is validated through two large-scale industrial case studies. In Case 1, WR and MU integration reduce utility cost to 9.8E+06 $/y (64.6% below the original design) and GHG emissions from 1.7E+08 to 1.2E+08 kg/y CO2 (30.3% reduction), with hot water generation contributing 82.7% of savings. In Case 2, the same strategy achieves a 45.6% GHG reduction relative to the baseline and lowers operation costs by 28.9%. This work provides a practical, system-level retrofit framework to support sustainable production and low-carbon transition in industrial processes.

Integrating heat pumps into large-scale electricity-to-heat industrial processes has proven highly successful in enhancing the utilisation of renewable energy and contributing to carbon emission reductions. However, most studies focus on overall system performance, overlooking the detailed thermal behaviour of the heat pump itself. This limits the adaptability of heat pumps in dynamic industrial settings. This work proposes an equation-oriented framework that enables flexible integration of thermodynamically detailed heat pump models into industrial heat recovery systems. A superstructure-based optimisation model is developed to minimise energy costs and enhance efficiency, considering process constraints, network layout, and heat pump performance. The model dynamically optimises heat pump operation and placement to enhance waste heat recovery and overall system integration. Moreover, the approach supports the integration of low-grade utilities to further improve the energy efficiency. The proposed framework is validated through an industrial-scale case study of a crude oil distillation process. Life cycle assessment is conducted to quantify potential environmental and economic benefits. Results show that integrating heat pumps into the system recovered 50.52 % of low-pressure steam, reducing the total operating cost and annual cost by 12.88 % and 12.42 %. Additionally, total net carbon emissions decreased by 28.70 %. Lower electricity prices increase heat pump use and economic benefits but also amplify rebound effects. Furthermore, although high-temperature heat pumps operating above 150 degrees C tend to increase capital expenditures, they unlock greater energy efficiency, thereby accelerating the industrial decarbonisation process.

Transitioning heat exchanger network (HEN) synthesis designs to industrial application involves operational, environmental, and cost considerations, posing computational challenges. This study proposes a systematic optimization approach integrating multi-objective, multi-period optimization HEN synthesis with waste heat recovery and multiple utilities. The proposed methodology incorporates a novel two-step unit reduction strategy to overcome the increase of model combinational complexities arisen from the multi-period features, thereby facilitating the solving of large-scale problems. Meanwhile, environmental impacts are concerned by using the technique for order preference by similarity to ideal solution approach. A new optimization route, Enhanced Pinch-assisted Multi-Objective Optimization is proposed to obtain the final decision in this multi-objective problem time-efficiently. The case study includes a 15 streams problem, and a real industrial-scale crude oil distillation preheat system. The results showed that assigning carbon compensation to the waste heat recovery option can significantly reduce carbon emissions and change energy distribution.

Achieving Net Zero emissions is a crucial goal for mitigating climate change, with chemical production processes being major contributors to greenhouse gas emissions. Conventional ethylene production plants, primarily fueled by cracked hydrogen and methane, are significant sources of carbon emissions from chemical plants. These emissions present not only environmental challenges but also economic risks due to impending carbon taxes and stricter regulations. Introducing carbon capture in ethylene production is therefore essential for reducing the carbon footprint, aligning with global sustainability targets, and ensuring economic resilience in a future where carbon emissions are increasingly penalized. This work aims to decarbonize and expand the existing ethylene plant by incorporating solar power, carbon capture process and methanol to olefin (MTO) technology. The raw materials of the MTO process include CO2 captured from flue gas, cracked hydrogen from the ethylene cracker and green hydrogen from proton exchange membrane (PEM) electrolyzer driven by solar power. Hybrid modelling is utilized to develop a mathematical model of the solar-driven PEM water electrolyzer in MATLAB, and simulate the rate-based carbon capture process, methanol production process and MTO process in Aspen Plus. The results indicate that 85% of CO2 is converted using cracked hydrogen and green hydrogen, leading to 5.4% increase in ethylene production and 53.6% increase in propylene production, respectively. The levelized revenues for ethylene in the original plant, the 70% carbon mitigation scenario, and the 85 % carbon mitigation scenario are 813 USD/tC2H4, 525 USD/tC2H4, and 258 USD/tC2H4, respectively. Sensitivity analysis reveals that mitigating 85% of CO2 becomes economically unviable if solar power exceeds 100 USD/MWh.

Heat exchanger networks (HEN) synthesis faces challenges in addressing non-isothermal mixing and nonconstant thermal properties of streams, giving rise to computationally-hard mixed integer nonlinear programming problems. An integrated algorithm framework including the simple additive weighting method, the twolevel hybrid algorithm, and the epsilon-constraint method is proposed for multi-objective optimization (MOO) of HEN with variable heat capacity. The proposed framework that incorporates a new stream matching set and a new heat exchanger vector to improve optimization and generate feasible solutions, is capable of efficiently handling the trade-off between total annualized costs (TAC) and environmental impacts (EI). Two case studies are conducted to verify the effectiveness and superiority of the proposed framework. Optimization results demonstrate that the proposed framework can find competitive solutions with lower TACs and achieve the tradeoff solutions whose EIs are considerably reduced by 16.4% and 7.89% compared to the lowest TAC solutions reported in literature.

This work focuses on heat exchanger networks (HENs) synthesis (HENS) considering the optimal locations of multiple utilities. Based on an extended stage-wise superstructure where available heaters and coolers are placed at all stages, HENS is modeled as a computationally-hard mixed integer nonlinear programming (MINLP) problem. To obtain high-quality solutions, we propose a new hybrid algorithm framework that combines deterministic algorithm (commercial solver) and genetic algorithm (GA) without the use of penalty functions. In the outer level of the framework, GA is employed to optimize the integer variables which represent the existences of matches between process streams as well as the available heaters and coolers at intermediate stages. In the inner level, a reduced-size MINLP model is built to minimize the total annualized costs (TACs) of HENs generated in the outer level. We also propose three new sets to exclude infeasible stream matches, thereby the HENs generated in the outer level are all feasible and our GA does not need any penalty terms. Four literature examples are tested and optimal solutions with lower TACs are obtained within acceptable computing time compared to solutions reported in literature.

Approaches for Heat Exchanger Network (HEN) synthesis have become increasingly significant in recent years because of their potential role in saving energy and tackling climate change. Unlike existing methods that rely on introducing additional stages to accommodate multiple utility options, while often neglecting the potential of waste heat recovery, this work proposes a novel mathematical optimization methodology to tackle the HEN synthesis problem and bring energy benefits of waste heat recovery. The method aims to overcome the drawbacks in existing methods related to low computational efficiency resulting from numerous discrete combinations. An enhanced stage-wise superstructure is presented to automatically optimize selections of stages covering waste heat recovery or heat recovery in hot streams, and multiple hot utilities or heat recovery in cold streams, formulated as a mixed-integer nonlinear programming (MINLP) problem. Grand composite curve (GCC) is adopted to implement preliminary simplifications for the superstructure to solve large-scale HEN problems, by cutting the inappropriate utilities according to the pinch method and minimum temperature driving force to eliminate redundant combinations. The results show that the proposed approach can provide cost-efficient solutions with lower total annual cost (TAC) due to significant reductions in energy cost, compared with previous works. Specifically, the proposed approach achieves TAC savings of 14.5 %, 3.15 %, and 4.5 % for Case 1, Case 2, and Case 3.

The olefins production relies on thermal cracking, which emits significant greenhouse gas. This study proposed a novel clean olefins production process, which utilizes cracked hydrogen from ethane cracking, to converted captured CO2 from flue gas into methanol and finally to produce olefins. A real industrial plant with production rates of 819200 t/y of ethylene and 77520 t/y of propylene is selected for case study. The proposed processes are simulated in Aspen Plus, with consideration of heat integration. By generating two optimal heat exchanger networks for high-capacity and low-capacity operations scenarios, increased heat recovery of 35.11 MW and 29.33 MW can be achieved compared to the base cases. This results in an improvement in process energy efficiency through effective heat integration between the ethylene, CCS, and MTO processes. The life cycle assessment shows that all cracked hydrogen can convert 70 % CO2 from flue gas. In this scenario, the global warming potential (GWP) is 1.64 kg CO2 eq/kg of olefins, slightly higher than demonstration industrial plant (1.53). If 85 % of CO2 is converted with support of electrolyzer and photovoltaic power, although the GWP during production process decreases to 1.47, the manufacture of electrolyzer leads to significant emission and which is undesirable.

To address the issue of the traditional oil vapor recovery process of adsorption and spraying being unable to achieve the desired recovery effect during the summer months, this study innovatively couples the precooling system, and proposes a multi-stage oil vapor recovery technology of precooling-adsorption-absorption. This technology overcomes the limitations of the previous single-stage technology and is of great significance for reducing environmental pollution and improving the efficiency of the oil depot. For the precooling system, both series and parallel processes of the circulating water lines were studied, considering 18 working conditions, and focusing on the design and optimization of the key equipment, precoolers 1 and 2. The case results (the capacity of 800 m3/h) showed that the maximum water saving is 1.22 t/h for series arrangement compared with the parallel water lines in the precooling system. By optimizing the structural parameters of the precooler, the water consumption of the precooler can be effectively regulated because the overall heat flux is improved. Combined with the actual situation on site, the water lines of two precoolers were finally selected for series arrangement. Precooler 1 is a U-shaped tube type with a heat transfer area of 47 m2 and water consumption of 4,660 kg/h. Precooler 2 is a fixed tube-sheet type with a heat transfer area of 8.4 m2 and 900 kg/h water consumption. The optimized system can effectively reduce the temperature of the oil vapor, and alleviate the problem of hightemperature alarms in the adsorption tanks.

Shell and tube horizontal thermosyphon reboilers (STHTs) find wide applications in the process industry particularly due to their inherent advantages in handling large duties, high viscosity fluids and low fouled susceptibility. However, current design commercial packages suffer from obvious limitations including extensive user manipulations, trial-and-error necessity, increased human factor influence. To overcome these limitations, we develop in this work a rigorous MINLP model for STHT sizing, by achieving automated selection of detailed exchanger geometries with optimal overall heat transfer coefficient. The corrected minimum cross-flow area in shell-side is incorporated to accommodate full cross-flow patterns in tube bundle, thus handling popular service of X shell without baffle construction. A dependable convective boiling heat transfer model is proposed to obtain the shell-side heat transfer coefficient, validated with Aspen EDR. The critical variable, heat flux, which affects nucleate boiling heat transfer performance, is optimized through a simultaneous equation-solving process using a global solver, reducing excessive unit overdesign. Pressure balance between STHT and distillation column is fully considered by a proposed optimization model incorporating pressure drop and pipe sizing. By employing global optimization techniques and automating the design process, this innovative approach minimizes manual adjustments and provides more accurate and cost-effective designs.

Typical solar power tower (SPT) systems employ molten salt as the heat transfer and thermal energy storage medium to facilitate stable energy output. However, these systems are constrained by their limited operating temperature, which is insufficient to supply heat for high-temperature electrolysis. In this paper, a small-scale (2.5 MW) solar-thermal-assisted energy system with a SPT, a supercritical CO 2 (SCO 2 ) Brayton cycle, and solid oxide electrolysis/fuel cells (SOEC/SOFC) is proposed. With heat supply via air, the SOEC subsystem can operate at 800 degrees C to reach high energy efficiency and reduce electric demand, which replaces conventional waste heat from flue gas produced by fossil fuel combustion. Such a system can convert excess electricity into hydrogen for storage or sale and provides stable electricity 24 h a day. The key factors determining the system performance are investigated, including the turbine inlet parameters, main compressor inlet parameters, the recompression fraction of the SCO 2 subsystem, and the operating temperature and current density of the SOEC/SOFC subsystems. The case study shows that the net hydrogen output can reach 13365.4 kWh/d in summer with the SOEC operating at 800 degrees C, which is 86.7 % higher compared to 7157.6 kWh/d when operating at 600 degrees C. To recoup the investment costs by the 20th year, hydrogen must be priced at 6.5 $/kg. After multi-objective optimization, the optimal exergy efficiency and capital investment for the SPT-SCO 2 -SOEC subsystem are determined to be 29.6 % and 3.65 M$, respectively. For the SOFC subsystem, the corresponding figures are 56.1 % for exergy efficiency and 0.23 M$ for capital investment. This study benefits the solar power generation and hydrogen production by high-temperature electrolysis.

Abstract Waste heat recovery is common in high‐temperature water electrolysis production, but the latent heat from the hydrogen–water mixture is not fully utilized, and the electric heater to generate steam consumes much energy. In this research, a cascade heat pump is proposed to recover the latent heat from electrolysis products and generate steam. The heat pump can save as much as 65% electric energy compared with a single electric heater. Although in some cases, the latent heat is not enough to generate sufficient steam, it can still save 46.1% of energy. Also, this research emphasizes the significant influence of hydrogen and water proportion in electrolysis products. Compared with 80% water proportion, 50% water proportion can save 1.67 MW energy just in the water vaporization process for a 10 MW electrolyzer. The payback period is 3.43 years, which makes it worth investing.

Due to the complexities arisen from the non-convexities in the mathematical models for the HEN synthesis incorporating detailed exchanger design, constant heat transfer coefficients and short-cut model for the calculation of exchanger capital cost are used for a majority of approaches to obtain a synthetic network topology, which causes inaccurate heat transfer areas and trade-offs between energy usage and capital investment. This paper presents an enhanced iterative-based decomposition algorithm to achieve realistic HEN synthesis with detailed heat exchanger sizing, which targets to overcome the drawbacks associated with the use of short-cut heat exchanger model in configuration synthesis, and further presents how these exchanger details can be employed to lead the HEN synthesis towards generating more cost effective solutions. Fouled individual stream heat transfer coefficients and corrected total process cost are updated iteratively between heat exchanger design (HED) and HEN superstructure (HENS) to guide HEN topology optimization. Global optimization for heat exchanger sizing is achieved in each iteration using a global solver BARON/GAMS.34 to overcome instability in the iteration process caused by local optimum issues. A case study shows that it can provide a better solution than the results in the literature with a lower total annual cost and computational time.

Heat Exchanger Network (HEN) synthesis is primarily formulated as a mixed integer non-linear programming (MINLP) problem based on method of stage-wise superstructure (SWS). Approaches to obtain an optimal HEN configuration can adopt deterministic algorithms. But, as a large scale problem, it is difficult to solve due to the complexities arisen from nonlinearities of SWS. To overcome the model nonlinearities, stochastic algorithms and meta-heuristic approaches have been proposed in the literature to tackle the problem. However, it reaches a near-optimal HEN configuration from a series of stochastic solutions, which are obtained by the execution of many computational procedures with extensive time resource, and a global optimum will not be guaranteed from only randomly generated results. In this paper, an enhanced SWS approach is proposed, in which new temperature and heat duty constraints are updated to reduce the redundant combinations and avoid conflicted calculation of non-isothermal mixing energy balance. The present model is also extended to allow flexible stream splitting for practical applications. Then, a deterministic-based global solver (GAMS/BARON) is applied in solving three case studies. The results show that the proposed approach can provide cost-efficient HEN solutions with lower TAC than using existing stochastic and deterministic algorithms.

Stage-wise superstructure (SWS) is one of the most widely used methods for Heat Exchanger Network (HEN) synthesis. Approaches to obtain optimal HEN solution can be based on the deterministic algorithm, but it is difficult for solving a large-scale problem since the complexities arisen from nonlinearities of SWS. Several methods which employed stochastic algorithms and meta-heuristic approaches have been proposed to tackle the problem. However, it determines a near-optimal HEN configuration from a series of stochastic solutions which are obtained by the execution of many computational operations, thus it is time consuming and a global optimum will not be achieved from only randomly generated results. In this study, an enhanced SWS is presented, in which new constraints and variables are added to avoid conflicted calculation of non-isothermal mixing energy balance and reduce the redundant combinations. Moreover, the present model is extended to allow a flexible requirement of stream splitting for practical application. Then, a deterministic-based global solver (GAMS/BARON) is applied in solving three case studies, operated on computer: I7-8565U. The results showed that the proposed approach can provide a cost-efficient HEN solution with a lower TAC than that obtained from existing stochastic and deterministic algorithms.