Abstract This study addresses the productivity limitations of batch ethanol fermentation (low operational efficiency, prolonged downtime) and operational instabilities in continuous systems (substrate loss, process fluctuations) by investigating two innovative configurations for long‐term stable fermentation: the self‐cycling fermentation (SCF) and an in‐situ product removal (ISPR)‐integrated membrane bioreactor. A unified mathematical model is developed to integrate batch and continuous experimental data, which quantifies the impact of dilution rate on microorganisms as a lag effect. The lag effects of SCF and ISPR‐integrated systems are quantified, and their stability and performance superiority are analyzed using the proposed model. Results show that: (1) SCF achieves an overall productivity of 1.22 g/(L · h) with a yield of 0.408 g/g, guarantees period‐1 stability, and outperforms conventional continuous systems; (2) The ISPR‐integrated continuous system enhances productivity to 2.0 g/(L · h) with a yield of 0.425 g/g, maintains stability in delayed systems, despite one‐third of the substrate being lost in the effluent. This work establishes a theoretical framework for the design of sustainable and stable ethanol fermentation processes.
Abstract In the development of high‐sour gas reservoirs, sulphur deposition is a major cause of reservoir damage and gas well productivity decline. This study adopts the lattice Boltzmann method (LBM) to construct a multi‐physics coupled model incorporating flow, temperature and sulphur transport fields. Mesoscopic simulations are carried out to explore seepage, sulphur precipitation and deposition characteristics in porous media. The results reveal that low pressure difference lowers local flow velocity below the critical value and promotes in‐situ sulphur deposition, leading to significant losses in porosity and permeability. The permeability damage reaches 98.4% at a pressure difference of 2 MPa. Larger temperature difference further decreases sulphur solubility and accelerates deposition, with permeability dropping to 0.14 mD and a damage rate of 98.6% at 15 K temperature difference. Sulphur tends to accumulate near pore walls, narrow throats, and outlet zones, presenting obvious spatial selectivity.
Abstract Industrial brewery wastewater sludge (BWS) is generated in large volumes, causing environmental contamination issues due to its high organic matter content and, in some cases, the presence of heavy metals, toxic substances, pathogens, and emerging contaminants. Vermicomposting is a promising alternative for stabilizing organic waste. Once the process is completed, the end product (vermicompost) can be applied to soils, improving their structural, physicochemical, and fertility properties. The objective of this study was to investigate vermicomposting as a strategy for stabilizing wastewater sludge from the brewing industry and to determine the final quality of the product. To achieve the objectives, eight vermicomposting units were established with different proportions (25%, 50%, 75%, and 100% w/w) of sludge and stabilized material. Over 10 weeks, various parameters such as moisture, pH, and organic matter were monitored and analyzed. Subsequently, the products underwent a 4‐week maturation process. To determine if the mature products comply with the minimum requirements of the corresponding regulations, parameters such as C/N ratio, organic matter, cation exchange capacity (CEC), faecal coliforms, heavy metals, pH, germination index (GI), and moisture were analyzed. The 25%, 50%, and 75% brewery sludge units met the established requirements and can therefore be used for agricultural, forestry, and commercial vermicompost applications. The 100% sludge ratio did not comply; however, the maturity stage time can be extended to meet the standards. In conclusion, the vermicomposting stabilization of brewery industry sludge is a viable alternative for obtaining value‐added products.
Abstract Current oil production in Mexico is facing a severe decline, which can be reverted either by exploration of new fields or exploitation of mature ones. The latter are the subject of this study, since gigantic volumes of residual oil are still inside the reservoirs after primary and secondary production. This volume can be accessed through application of enhanced oil recovery (EOR) techniques. The objective of this paper is to set a perspective of EOR processes and to identify their potential for adding oil reserves by their application in Mexico. This potential is expressed as incremental oil volume, which can be recovered by means of innovative techniques such as the one introduced in this work: in situ upgrading technology (ISUT). ISUT process uses the reservoir as a high temperature reactor, in which heavy oil fractions, ultra‐dispersed nano catalyst, and hydrogen are injected into the reservoir. This injection mixture favours permanent oil upgrading. In this work, ISUT process is evaluated at a laboratory scale through reactivity tests and experiments conducted in a Catalytic Reactivity Process unit using a heavy oil from a Mexican naturally fractured reservoir located in the Gulf of Mexico. Based on the experimental work conducted in this study, it is concluded that oil viscosity and American Petroleum Institute (API) gravity can be permanently improved, with the advantage of no coke generation or solids precipitation. Results showed that it is possible not only to upgrade heavy oil but also to contact the oil that resides within the matrix with the application of this novel EOR process.
Abstract In copolymerization, chemical composition (CC) strongly influences rheological and end‐use properties. CC is commonly analyzed using the Mayo–Lewis equation (MLE) and derived models, where monomer reactivity ratios (RRs) are key parameters. Common RR estimation methodologies include the Meyer–Lowry equations (MeLE), Skeist's equations—based on residual monomer composition ( f vs. X ) and copolymer composition ( F vs. X ) versus conversion—and monomer balance equations (MOBE). However, a rigorous comparison of these methodologies has not been previously reported. In this work, these methods were applied to estimate RRs for three methacrylate‐based systems: methyl methacrylate (MMA)/vinyl acetate (VAc), MMA/2‐(dimethylamino)ethyl methacrylate (DEAEMA), and DEAEMA/styrene (Sty). Copolymerization kinetics and CC were determined in situ via 1 H‐NMR spectroscopy. A particle swarm optimization (PSO) algorithm was implemented alongside these methodologies and compared with the Levenberg–Marquardt optimization (LMO) algorithm. Results showed that estimated RRs for MMA/DEAEMA and Sty/DEAEMA were largely insensitive to the methodology, depending instead on intrinsic system characteristics. Conversely, the MMA/VAc system exhibited more pronounced deviations, highlighting that methodological choices matter depending on the system. While all data were satisfactorily described by the dynamic approach, the versus X formulation yielded the lowest Akaike information criterion (AIC) and Bayesian information criterion (BIC) values. Finally, both optimization algorithms produced identical RR estimates, confirming the numerical strategy did not significantly affect parameter estimation.
Abstract Elemental sulphur deposition poses a significant operational challenge in natural gas transmission systems, particularly downstream of pressure‐reducing devices where rapid pressure and temperature changes occur. These thermodynamic shifts reduce sulphur solubility, triggering nucleation and leading to issues such as valve malfunction, flow restriction, corrosion, and maintenance costs. This study presents an integrated physics‐informed framework to predict sulphur dropout within a pressure‐control valve. Unlike previous sulphur‐deposition models, the proposed framework integrates three‐dimensional computational fluid dynamics (CFD), equilibrium chemistry, sulphur solubility estimation, and classical nucleation theory within a spatially resolved workflow. CFD simulations resolve spatial pressure and temperature fields under varying operating conditions and valve openings. These fields are coupled with a chemistry module that adjusts equilibrium constants and solves a constrained reaction‐extent problem to estimate sulphur formation. Local sulphur solubility is determined through interpolation of literature data, while supersaturation‐driven nucleation rates are computed at the CFD cell level. The framework enables spatial prediction of sulphur supersaturation and precipitation within the valve. To support rapid screening, a random forest surrogate model was trained on simulation outputs, achieving a mean absolute percentage error of 15.4%, a root‐mean‐square percentage error of 20.4%, and an R 2 of 0.977. The framework provides a practical tool for identifying sulphur‐deposition hot spots, developing sulphur‐deposition operating envelopes, and enabling rapid risk screening through a machine‐learning surrogate.
Abstract High‐viscosity pseudoplastic non‐Newtonian fluids readily form localized active‐flow caverns and stagnant regions during laminar agitation, which weakens bulk circulation and prolongs mixing time. In this work, the active‐flow cavern refers to an impeller‐induced intensified flow zone rather than a true yield cavern in a viscoplastic fluid. To address this problem, a multi‐stage fractal‐arranged perforated impeller (FAP impeller) was proposed. The effects of impeller stage number and blade pitch on the mixing characteristics of sodium carboxymethyl cellulose (CMC‐Na) solutions were investigated using computational fluid dynamics (CFD) simulations and experimental validation. Single‐ and double‐stage FAP impellers with straight blades (FAP‐SB) and pitched blades (FAP‐PB) were compared in terms of flow characteristics, power consumption, active‐flow cavern evolution, and mixing time. The results showed that the double‐stage configuration expanded the active‐flow cavern and reduced high‐viscosity stagnant regions. At Re ≈ 101–272, the mixing times of the double‐stage FAP impeller with straight blades (2S FAP‐SB) and the double‐stage FAP impeller with pitched blades (2S FAP‐PB) were reduced by approximately 15.21%–21.18% and 17.79%–27.16%, respectively, compared with the single‐stage FAP impeller with straight blades (1S FAP‐SB). Moreover, 2S FAP‐PB further shortened the mixing time by 3.03%–7.59% relative to 2S FAP‐SB. Overall, 2S FAP‐SB was more effective in enhancing local shear, enlarging the active‐flow cavern, and increasing the strain rate, whereas 2S FAP‐PB exhibited better bulk mixing performance. These findings provide guidance for impeller design and equipment retrofit in the agitation of high‐viscosity non‐Newtonian fluids.
Abstract We analyze coupled heat and mass transport of a reactive species in Newtonian plane Couette–Poiseuille flow between parallel plates, with a reactive lower wall and a non‐catalytic absorbing upper wall. The model includes viscous dissipation, heat of reaction, and Arrhenius temperature dependence in the homogeneous first‐order reaction rate. In the small‐Brinkman‐number regime Br ≪ 1 with weakly exo‐/endothermic kinetics ( Q = Br q , q = O(1)), the system admits a perturbative semi‐analytical solution. The leading order recovers the isothermal reactive Graetz problem, while the first correction in Br involves two Sturm–Liouville bases: a mass base and a thermal base sharing the parabolic velocity weight. We prove a modal‐symmetry selection rule showing that only odd‐numbered modes contribute when the velocity profile and source terms are symmetric under . The fully developed dissipation‐induced temperature field is obtained in closed form as a quartic polynomial. Validation against a fully coupled implicit finite‐difference solver confirms practical agreement in the small‐ regime, with modal validation errors of order 10⁻⁴ and coupled‐field errors below 5% for . The thermal‐reactive coupling modestly accelerates axial species conversion. Quantitatively, the effective axial decay rate increases approximately linearly with ; in the validated small‐ range, the local sensitivity is , shortening the 90%‐conversion length by about 1.4% at . The analysis is directly relevant to the design of reactive coating systems and catalytic microchannel reactors operating in the small‐Brinkman regime.
Abstract Severe slugging represents a major challenge in offshore production systems, potentially compromising operational stability and the integrity of installations. Although established simulators such as OLGA® are widely used for its analysis, there is a gap in the literature regarding the application of the one‐dimensional simulator ALFAsim® for modelling and mitigating this phenomenon. In this context, this study aims to evaluate ALFAsim's® capability to identify severe slugging and analyze mitigation strategies in a system composed of a well, flowline, and riser. The methodology was based on one‐dimensional transient numerical simulations, including domain discretization with a linear mesh (10 m segments), solution of the mass and momentum conservation equations, and identification of flow patterns using the unit cell model. Boundary conditions of flow rate, pressure, temperature, and gas–oil ratio were imposed, with parametric variation of the oil flow rate between 800 and 1400 m 3 /d. For mitigation analysis, a choke valve located at the top of the riser was modelled using a C v (flow coefficient) versus opening curve, and gas injection at the base of the riser was performed at different rates. The results indicated that ALFAsim® reproduced the main hydrodynamic characteristics of the phenomenon, including cyclic flow oscillations and pressure variations, with an approximate cycle period of 7000 s. A choke valve opening of 92.5% and a gas injection of 660,000 m 3 /d are found to be the most effective conditions for flow stabilization. These findings demonstrate that ALFAsim® is a suitable tool for the analysis and mitigation of severe slugging.
Abstract To address the problem of balancing sensitivity and robustness in incipient fault detection (IFD) for large‐scale processes, a distributed monitoring framework based on causal‐oriented kernel regularized canonical variate analysis (CO‐KRCVA) is proposed. The framework divides operational units into sub‐blocks according to the sequence of production. To reveal the information transfer relationships and the underlying mechanisms of dynamic changes among variables, a causal difference matrix (CDM) using transfer entropy estimation is constructed for each sub‐block. Based on this, a CO‐KRCVA model is established, which specifically captures targeted minor variations of process variables to enhance the sensitivity and reliability. Its regularization strategy further overcomes the effects of random noise or interference to reduce false alarms. In addition, the Wasserstein distance (WD)‐based statistic is employed to sensitively reflect distribution shifts from both probabilistic and geometric perspectives. To facilitate global detection, Bayesian inference is used to fuse the results from local models. Finally, the proposed method and monitoring framework are validated for effectiveness on different types of faults in the vinyl acetate monomer (VAM) process and Tennessee Eastman (TE) process. Experimental results show that compared with CVA, the proposed method improves the early fault detection rate by 28.47%, 46.21%, and 12.6%, respectively, for the three fault types in the TE process, while reducing the false alarm rate by 0.52%, 1.44%, and 1.36%.
Abstract The market value of natural food antioxidants will double to 8 billion USD by 2033, driven by a global consumer demand for ‘clean label’ products and proactive wellness solutions. However, the industrial application and health labelling (commercial brands) of these compounds remain severely constrained by inconsistent analytical methods. This perspective article evaluates how distinct chemical frameworks shape antioxidant performance and outlines the technical challenges involved in measuring antioxidant power. As a result, a gap persists between regulatory expectations and consumer confidence. Paradox: Traditional, well‐established analytical metrics fail to predict how an antioxidant behaves inside the human body. A single absolute ranking system is lacking because a molecule's protective power is fundamentally context‐dependent. Solution: Here we shift the discussion away from rigid chemical metrics toward a biologically relevant framework. We evaluate how solubility environments and cellular uptake influence real‐world potency. By aligning standardized mass‐capacity metrics with cellular outcomes, we provide a transparent blueprint for high‐stability food formulation and more credible health‐related claims.
Abstract Fault diagnosis plays an important role in process monitoring under closed‐set settings. However, the unknown faults may be misdiagnosed as one of the known classes with overconfidence, leading to unreliable decisions and process unsafety. Besides, to distinguish multiple unknown fault modes, existing methods need to be completely retrained with sufficient measurements under faulty conditions. However, model retraining is time‐consuming and the labelled faulty data is rare in real industrial processes. To solve the above problems, a novel uncertainty‐aware evidential prototype network (UEPN) with few‐shot class‐incremental learning (FSCIL) for open‐set fault diagnosis (OSFD) is proposed. First, the prototype network (PN) based feature extractor is proposed to learn discriminative embeddings for known classes and preserve more embedding space for unknown ones. An uncertainty‐aware evidential classifier is further developed with uncertainty quantification to identify unknown faults and avoid overconfident misclassification. Then, to continuously identify multiple unknown fault modes, the FSCIL procedure is designed, which enables UEPN to be incrementally updated with few‐shot data of an unknown fault class. Finally, the proposed method is evaluated on the Tennessee Eastman process (TEP) and the vinyl acetate monomer (VAM) plant model. The proposed method achieves up to 10.8% higher fault diagnosis rate (FDR), 5.4% lower false positive rate (FPR), and 8.5% higher true positive rate (TPR) than the comparison methods for OSFD task of continuously distinguishing multiple unknown fault modes.
Abstract Particle deposition in wall‐bounded turbulent flows is a challenge in many industries, including pneumatic conveying, filtration, and related applications. When particles accumulate near boundaries, they reduce transport efficiency, alter near‐wall flow structures, and cause blockage. Although particle deposition has been studied extensively, the mechanisms that govern near‐wall accumulation remain unclear when several effects act simultaneously. In particular, the combined influence of flow direction, particle number density, and Stokes number has not been fully resolved when gravity, Saffman lift, and particle–particle collisions are all present. In this study, we perform direct numerical simulations of particle‐laden turbulent flow in a square duct. The fluid phase is resolved in an Eulerian framework, while the particle phase is tracked in a Lagrangian framework using a point‐particle approach. We studied six cases, three with downward flow and three with upward flow, each with a different particle number density. Results show that particle deposition at the walls decreases when particles lag the fluid, as in upward flows. For and a particle number density of , the wall particle concentration decreased by a factor of 85 in upward flow compared to downward flow. In downward flows, doubling the particle number doubles the particle–particle collision rate and reduces the wall concentration by a factor of six. In upward flows, particle number density has little influence on the concentration profiles. The findings can help identify flow conditions that prevent particle deposition.
Abstract The evaluation of the operational status of industrial processes is challenging due to multiple, often conflicting, performance dimensions, such as product quality, energy consumption, and material loss. A significant limitation of existing methods is the failure to capture that the interrelationships between process variables can differ substantially for each distinct performance indicator, or the reliance on a single, fixed process topology. This paper proposes a novel multi‐graph fusion framework for comprehensive status evaluation. First, this paper constructs a multiple indicator‐related quality‐graph‐network, each designed to capture the unique structural information relevant to a single indicator. Second, to integrate this differentiated information, this paper introduces a hybrid fusion method. This method intelligently revises node features by combining supervised revision (based on model reliability analysis) and unsupervised revision (based on node importance analysis), generating a comprehensive graph representation. Finally, a comprehensive evaluation model is built upon the fused graph features to assess the overall operational status. The effectiveness and feasibility of the proposed method are verified through application to the Tennessee Eastman benchmark process.
Abstract In response to the common issues in traditional light‐burned magnesium oxide fluidized bed calcination furnaces, such as uneven particle distribution, non‐uniform temperature fields, and low reaction efficiency, this study proposes a jet‐pulse fluidized bed roasting device with a contraction‐expansion mechanism. By establishing a coupled CFD‐DPM numerical model, the study systematically analyzes the influence mechanisms of key geometric parameters, such as jet spacing, contraction angle, throat length, and contraction ratio, on the gas–solid two‐phase flow characteristics and the decomposition behaviour of magnesite within the furnace. Based on the results of single‐factor experiments that clarified the independent effects of various parameters, the response surface methodology was used for multi‐objective collaborative optimization of the throat length, contraction ratio, and contraction angle. The results show that when the throat length is 185 mm, the contraction ratio is 0.5, and the contraction angle is 26°, the predicted decomposition rate of magnesite is 98.84%, with a corresponding predicted turbulent kinetic energy of 5 m 2 /s 2 . Pilot‐scale experimental results show that the average decomposition rate reaches 98.54%, with a prediction error of only −0.3%, validating the model's good predictive accuracy and engineering applicability. This study provides theoretical basis and methodological support for the structural design and performance optimization of jet‐pulse fluidized bed calcination furnaces.
Abstract As microfluidic chips emerge as a prominent and widely adopted technology, the multiple integrated units on the chip are progressively becoming the focal point of research. This paper investigates a T‐shaped passive micromixer based on the Cantor fractal principle. The mixing efficiency, concentration uniformity, and pressure drop are evaluated by considering baffle arrangement, fractal order, baffle height, baffle spacing, and Reynolds number. Firstly, comparison of symmetric and staggered Cantor baffles shows that the staggered structure enhances transverse splitting and recombination and therefore provides higher outlet mixing efficiency. For the two fractal orders examined in this work, namely primary and quadratic Cantor baffles, increasing the fractal order from primary to quadratic produces only a limited improvement in mixing efficiency; therefore, this conclusion should not be extrapolated to higher‐order Cantor structures without further verification. Increasing baffle height or decreasing baffle spacing enhances mixing but also increases pressure drop. Considering both mixing and hydraulic safety, the recommended geometric parameters are h = 0.15 mm and d = 0.1 mm. A quantitative performance metric, defined as the outlet mixing index divided by pressure drop, is added to evaluate the trade‐off between mixing enhancement and hydraulic penalty. At Re = 0.1 and 1, molecular diffusion and residence time dominate the mixing process, whereas at higher Re the baffles increasingly promote transverse advection and vortex‐assisted stretching. At Re = 100, the optimized micromixer achieves nearly complete outlet mixing in the numerical model.
Abstract This study investigates the enhancement of CO 2 solubility in aqueous sodium glycinate (SG) solutions blended with two polyamines: piperazine (PZ) and dipropylenetriamine (DPTA). Vapour–liquid equilibrium (VLE) data were experimentally measured using an equilibrium stirred cell at 313–333 K and CO 2 partial pressures of 2–120 kPa. Six blend compositions were investigated, containing 5, 10, or 15 mass% PZ or DPTA combined with 25, 20, or 15 mass% SG, respectively, at a constant total amine concentration of 30 mass%. CO 2 loading (mol CO 2 /mol total amine) increased substantially with higher polyamine content, with the 15 mass% DPTA +15 mass% SG blend showing the highest capacity. Solvent density was correlated using excess molar volume models, yielding mean absolute errors of 0.0265% (PZ–SG) and 0.0258% (DPTA–SG). The VLE data were successfully fitted using modified Kent–Eisenberg models, with mean absolute errors (MAE) of 0.75% and 0.80%, respectively. Four machine learning models (XGBoost, ANN, random forest, SVR) were evaluated using a combined dataset of 660 experimental and literature data points. XGBoost showed the strongest predictive capability ( R 2 = 0.996), outperforming random forest ( R 2 = 0.99), ANN ( R 2 = 0.972) and SVR ( R 2 = 0.92). The isosteric heat of CO 2 absorption, estimated via the Clausius–Clapeyron equation, ranged from 25 to 55 kJ/mol CO 2 —lower than 30 wt.% MEA (~85 kJ/mol). These results demonstrate that polyamine‐activated SG solvents significantly enhance CO 2 capture and that data‐driven models effectively predict CO 2 solubility for solvent screening and process design.
Abstract Recently, new research studies on nanofluids are increasingly focusing on hybrid nanofluids with enhanced thermophysical properties to overcome the limitations imposed by single‐component nanofluids. This work experimentally investigated the thermal conductivity of a novel hybrid nanofluid consisting of a mixture of MgO–ZnO and ethylene glycol (EG). Three significant variables were tested to investigate the performance of the new hybrid nanofluid: volume fraction (0.2%–1%, at a constant mixture of 50–50 ratio), temperature (set to 25–50°C with high accuracy), and nanoparticle diameter (set to 20, 55, and 90 nm). From the experimental study, it was identified that the maximum enhancement of the thermal conductivity was 11.48%, which was at the smallest nanoparticle diameter (20 nm), highest temperature (50°C), and highest solid concentration (1%). In addition, a new and accurate correlation function for this phenomenon was established. By using linear regression and MANOVA analysis, the correlation function achieved an R 2 of 0.99 with a difference of at most 1.15%, reflecting its high accuracy of prediction. It is clear that the values of all three variables—nanoparticle size, temperature, and volume fraction—played a highly important and interdependent role in the determination of the values of the thermal conductivity of the nanofluid. By performing a sensitivity analysis on the results of interest, the significant role of thermal conductivity in the variation of the variables at the extremes of the experimental range—20 nm, 50°C, and 1%—was established.
Abstract The high‐stakes nature of chemical production drives the demand for efficient and precise fault diagnosis techniques. However, chemical process data exhibit complex temporal evolution patterns alongside spatial correlation characteristics, making it challenging to simultaneously capture time‐varying dynamic features and spatial dependencies among variables, which consequently compromises the diagnostic accuracy. This manuscript proposes an adaptive spatio‐temporal graph neural network for fault diagnosis in chemical processes (ST‐FDGNN). The framework introduces a data‐driven adaptive graph structure learning mechanism, eliminating the reliance on predefined graph structures and enabling automatic discovery of latent variable association. By integrating temporal convolutional networks (TCN) and graph neural networks (GNN) deeply, the model achieves collaborative learning of complex spatio‐temporal features among multiple process variables. Additionally, an improved spatial pooling strategy is incorporated to enhance the efficiency of feature extraction. ST‐FDGNN is applied to the Tennessee Eastman process and the continuous stirred tank reactor system, and compared with several advanced fault diagnosis methods. The results demonstrate that the proposed model outperforms existing approaches across multiple statistical metrics, exhibiting superior diagnostic accuracy and generalization capability.
