Key Takeaways & Executive Findings
- •• • Slag foaming prediction is crucial for EAF efficiency and stability; the review identifies that single models fail to balance real-time capability and accuracy under multi-variable coupling, necessitating hybrid frameworks. • • Five prediction methods (empirical formulas, dimensionless modeling, thermodynamic calculations, CFD, machine learning) are systematically compared; thermodynamic tools like FactSage are used for phase equilibria, while CFD handles gas-liquid two-phase flow, but each has limitations in dynamic industrial conditions. • • The review proposes a hybrid prediction framework integrating mechanism-based and data-driven models, emphasizing physical constraints, multi-scale coupling, and multi-source data fusion to achieve 'controllable and adjustable' foaming. • • The study highlights the importance of green steelmaking, with slag foaming prediction directly impacting energy efficiency and process stability, aligning with national research funding (National Natural Science Foundation of China, No. 52074001).
Abstract
Slag foaming is a critical phenomenon in electric arc furnace (EAF) steelmaking, enhancing thermal efficiency, suppressing metal splashing, and stabilizing the refining process. Accurate prediction and control of slag foaming are essential for green and efficient steelmaking. This review systematically examines research progress on slag foaming prediction, clarifying the applicability, advantages, and limitations of different predictive methods to support intelligent control of foamy slags. Following the framework of 'influencing factors-prediction methods-development trends', the study summarizes the coupling effects of multiple variables such as basicity, viscosity, surface tension, suspended particles, gas parameters, and temperature on foam formation and stability. It compares five major prediction approaches: empirical formulas, dimensionless modeling, thermodynamic calculations, computational fluid dynamics (CFD) simulations, and machine learning models, analyzing their core concepts, merits, and constraints. Results indicate that single models often struggle to balance real-time capability and accuracy, particularly under multi-variable coupling and complex operating conditions. Therefore, a hybrid prediction framework combining mechanism-based and data-driven models is proposed, emphasizing physical constraints, multi-scale coupling, and multi-source data fusion. This integrated approach is expected to advance slag foaming prediction from 'computable' to 'controllable and adjustable', offering methodological insights for the development of green and intelligent EAF steelmaking.
1. Introduction
Slag foaming in electric arc furnace (EAF) steelmaking is a double-edged sword: while it improves thermal efficiency and stabilizes the arc, uncontrolled foaming leads to slag overflow, metal splashing, and refractory wear. Existing prediction methods—empirical formulas, dimensionless models, thermodynamic calculations, CFD simulations, and machine learning—have been developed independently, yet each suffers from critical bottlenecks. Empirical formulas often lack generalizability across varying slag compositions and operating conditions; thermodynamic models assume equilibrium, ignoring dynamic gas evolution; CFD simulations are computationally intensive and require precise boundary conditions; and machine learning models, though flexible, are data-hungry and often lack physical interpretability. The result is that no single approach can deliver both real-time prediction and accuracy in the highly coupled, transient environment of an industrial EAF.
This review addresses this gap by systematically analyzing the state-of-the-art and proposing a hybrid framework that synergizes mechanistic models with data-driven techniques. By embedding physical constraints and enabling multi-scale coupling, the framework aims to bridge the gap between laboratory-scale understanding and industrial-scale control. The authors argue that such integration is essential to transition slag foaming prediction from a purely computational exercise to a controllable and adjustable process parameter, ultimately supporting the green and intelligent transformation of EAF steelmaking.
Loading authentic research manuscript (Pages 1–5)...
Xinggan ZHANG, Yujie LIU, Mengting SHANG, Haichuan WANG, Yunjin XIA, Guilin SUN (2026). Research Progress and Intelligent Trend of Slag Foaming Prediction. The Chinese Journal of Process Engineering. https://doi.org/10.12034/j.issn.1009-606X.225250
Research & Educational Purpose Only: The translations, structured abstracts, analytical annotations, and data reports provided by SinoGreenTechare intended exclusively for academic research, internal corporate R&D, and educational benchmarking. They do not constitute formal engineering, chemical safety, legal, or professional advice.
Copyright & Intellectual Property Notice: Original copyright of the underlying source articles and experimental data remains with the respective authors, institutions, and original publishing journals. SinoGreenTech claims intellectual property only over its proprietary translations, analytical syntheses, and AEO structured enhancements in accordance with international fair use and academic citation principles.
Frequently Asked Questions
What are the key limitations of existing empirical formulas for slag foaming prediction in industrial EAF operations?
Empirical formulas, such as those based on slag basicity and viscosity, are often derived from specific laboratory conditions and may not capture the dynamic coupling of variables like gas flow rate, temperature fluctuations, and suspended particle effects. They typically lack real-time adaptability and can yield significant errors when extrapolated to different slag compositions or operating regimes, as noted in the review's comparison of prediction methods.
How does the proposed hybrid prediction framework address the trade-off between real-time capability and accuracy?
The hybrid framework integrates mechanistic models (e.g., thermodynamic and CFD) with data-driven machine learning models. By using physical constraints to guide the data-driven component, it reduces the need for extensive training data while improving extrapolation. Multi-scale coupling allows capturing both micro-level phenomena (e.g., bubble coalescence) and macro-level furnace dynamics, enabling real-time predictions without sacrificing accuracy, as emphasized in the review's conclusions.
What specific roles do thermodynamic calculations and CFD simulations play in slag foaming prediction, and what are their computational costs?
Thermodynamic calculations, using tools like FactSage, predict equilibrium phases and slag properties (e.g., viscosity, surface tension) under given compositions and temperatures. They are relatively fast but assume equilibrium, which may not hold during rapid gas evolution. CFD simulations model gas-liquid two-phase flow and foam dynamics in detail, providing high accuracy but requiring substantial computational resources and precise boundary conditions, making them less suitable for real-time control.
What are the main challenges in applying machine learning to slag foaming prediction, and how does the review suggest overcoming them?
Machine learning models require large, high-quality datasets that are often scarce in industrial settings. They also lack physical interpretability, making it difficult to trust predictions under novel conditions. The review suggests incorporating physical constraints into the model architecture (physics-informed neural networks) and fusing multi-source data (e.g., operational logs, sensor data, and simulation results) to improve robustness and generalizability.
How does slag foaming prediction contribute to green steelmaking, and what are the quantified benefits?
Accurate prediction and control of slag foaming can reduce energy consumption by stabilizing the arc and improving thermal efficiency, potentially lowering electricity usage by several percent. It also minimizes metal splashing and slag overflow, reducing material losses and refractory wear, thereby decreasing overall environmental impact. The review highlights that achieving 'controllable and adjustable' foaming is key to meeting green steelmaking targets.
Related Chinese Research & Cross-Citations
Activation of Peroxymonosulfate-Based Advanced Oxidation via Co@Si-A for Tetracycline Degradation: Performance and Mechanism
Cobalt-doped silica aerogel (Co@Si-A) catalysts were synthesized via a one-step sol-gel method and applied for peroxymonosulfate (PMS) activation to degrade tetracycline (TC). The catalyst with 25 wt% cobalt doping (25Co@Si-A) exhibited superior catalytic performance, achieving 98.97% TC degradation within 30 min under specified conditions (TC 10 mg/L, 100 mL). Brunauer-Emmett-Teller (BET) analysis revealed a high specific surface area and well-developed porous architecture with nano-confined spaces. The 25Co@Si-A/PMS system demonstrated outstanding adaptability across a broad pH range (5–9), maintaining >95% degradation efficiency, and showed strong resistance to sulfate and nitrate ions. In real water matrices, degradation efficiency remained around 80%. After five consecutive cycles, the system retained 82.33% degradation efficiency, with cobalt ion leaching of only 23.7 μg/L in the first cycle, indicating excellent stability. Mechanistic studies using electron paramagnetic resonance (EPR), radical quenching, and probe compound tests confirmed a synergistic radical and non-radical pathway. The primary reactive species were sulfate radicals (SO4•−), hydroxyl radicals (•OH), and singlet oxygen (1O2), with contributions of 58.53%, 9.79%, and 31.68%, respectively. Electrochemical tests indicated that 25Co@Si-A exhibited superior charge transfer compared to Co3O4, attributed to the nano-confined effect of the silica aerogel, which enhanced Co(II)/Co(III) redox cycling and PMS activation. This research provides a promising strategy for utilizing silica aerogel-based catalysts in advanced oxidation processes for water treatment.
Synergistic optimization mechanism of microstructure and magnetic properties in M-type strontium ferrite via Ce/La co-doping and pre-sintering temperature regulation
Driven by the urgent demand for green and low-carbon technologies, the development of high-performance and cost-effective rare-earth free permanent magnets has emerged as a key research focus for sustainable energy and advanced electronic applications. Among various candidates, M-type strontium ferrites have attracted considerable attention due to their excellent thermal stability, high magnetocrystalline anisotropy, and abundant raw material availability. In this study, Sr0.41La0.36Ca0.23Fe11.8Co0.2O19 was selected as the base system, and a series of samples were synthesized via a solid-state reaction combined with high-energy ball milling. The synergistic effects of varying CeO2/La2O3 mass ratios (0:10 to 10:0) and pre-sintering temperatures (1150-1200°C) on the microstructure and magnetic properties were systematically investigated. Microstructural analyses revealed that moderate Ce substitution effectively induced controlled lattice distortion and promoted densification, which inhibited abnormal grain growth and refined the microstructure. Such structural modulation not only enhanced domain wall pinning but also improved magnetocrystalline anisotropy, leading to a remarkable increase in coercivity. Magnetic measurements confirmed that the composition with a CeO2/La2O3 mass ratio of 2:8 and pre-sintered at 1180°C achieved the most balanced magnetic performance, exhibiting enhanced coercivity, sufficient remanence, and stable saturation magnetization. This work provides new insights into the cooperative effects between rare-earth doping ratios and thermal processing parameters, clarifying how lattice defects, grain boundary characteristics, and microstructural evolution collectively govern the magnetic properties of M-type ferrites. The findings establish a practical strategy for tailoring the microstructure-property relationship in rare-earth free permanent magnets, opening an optimized processing window for scalable fabrication of environmentally friendly, high-performance ferrite materials.
CFD Simulation and Structural Optimization of a Thermal Catalytic Degradation Reactor for Sulfur Hexafluoride
Sulfur hexafluoride (SF6), widely used as an insulating gas in high-voltage electrical equipment, possesses a global warming potential (GWP) 25,200 times that of CO2, necessitating efficient degradation technologies. This study employed computational fluid dynamics (CFD) to simulate the thermal catalytic degradation of SF6 in a fixed-bed reactor, integrating models for porous media, heat transfer, turbulence, and chemical kinetics. The simulations revealed significant radial non-uniformities in pressure, velocity, temperature, and species concentration distributions, with temperature identified as the dominant factor influencing degradation efficiency. Radial temperature gradients caused uneven reaction rates, with degradation rates near the wall substantially exceeding those at the central axis, thereby reducing overall SF6 conversion. To address this, structural optimizations were implemented, including reducing the reactor tube diameter and incorporating inert porous media with high thermal conductivity at both ends of the catalytic section. These modifications enhanced radial heat transfer, homogenized the temperature field, and improved the uniformity of reaction rates and species concentrations. Parametric studies on inlet gas velocity showed that both excessively low and high flow rates were detrimental: low velocities led to underutilization of the downstream catalyst and increased energy consumption, while high velocities deteriorated heat transfer and exacerbated radial temperature gradients. The optimal inlet velocity range was determined to be 0.4–0.8 m/s for a reactor tube inner diameter of 10 mm, balancing catalyst utilization, energy consumption, and degradation efficiency. This research provides data-driven guidance for the design and scale-up of SF6 catalytic degradation reactors.
A Review on Energy-Saving and Consumption-Reducing Technologies for Thermal Power Units Based on Economic Benefit Evaluation
Thermal power units have long dominated China's energy structure due to the low cost of coal and their role in ensuring grid stability. However, under the dual pressures of climate change and national carbon peaking/neutrality goals, the environmental impact of their 'three wastes' has become critical, necessitating energy-saving retrofits. This review systematically examines mainstream energy-saving technologies for thermal power units, including boiler combustion optimization, heating surface cleaning, turbine flow path upgrades, waste heat recovery and cascade utilization, and cold-end system optimization. Using coal consumption rate as the core economic index, the study integrates case studies and operational data from typical domestic and international units to evaluate the latest progress, practical effects, advantages, and limitations of each technology. Results indicate that these technologies significantly improve energy efficiency and reduce pollution. For instance, boiler combustion optimization based on support vector machines and neural networks enhances thermal efficiency and reduces NOx emissions. Turbine flow path modifications, from full three-dimensional CFD optimization to advanced blades and combined steam seals, yield notable gains in cylinder efficiency and heat rate reduction. Low-temperature economizers reduce coal consumption and auxiliary power/water use in dust removal and desulfurization systems. Heat pump applications include absorption, compression, and hybrid types. In cold-end optimization, data-driven predictive maintenance and real-time performance tuning of condensers achieve nearly 50% energy savings in circulating water pumps and an average coal consumption reduction of 2-3 g/(kW·h). Despite these advances, gaps remain in multi-objective optimization robustness, intelligent diagnosis, and advanced materials. Future research should focus on deep reinforcement learning for adaptive control, sensor networks for real-time diagnostics and predictive maintenance, and high-temperature corrosion-resistant materials for heat exchangers, while balancing initial investment and maintenance costs.
Gas-liquid dispersion characteristics in a stirred tank equipped with porous aeration tube
Gas-liquid stirred tanks are widely used in oxidation, hydrogenation, and other chemical processes, where the gas dispersion state directly affects production efficiency. This study systematically investigated the effects of impeller type, impeller installation height, and rotational speed on gas-liquid dispersion in a stirred tank equipped with a porous tube sparger. Two typical impellers, a wide hydrofoil (WH) and a half-elliptical disk turbine (HEDT), were tested at various installation heights (L/D ratios) and gassing rates. The critical rotational speed for complete gas dispersion, agitation power consumption, and overall gas holdup were measured. Results showed that for both impellers, the critical Froude number (Fr) decreased significantly with increasing gas flow number (FlG). Under the same gassing rate, the HEDT impeller generally required a higher critical Fr and greater agitation power for complete dispersion compared to the WH impeller. Relative power demand (RPD) decreased as FlG increased, with a more pronounced decline at higher L/D ratios. At different impeller positions, the RPD of the HEDT impeller was higher than that of the WH impeller, indicating that the HEDT impeller's power was less affected by gas. Notably, the impeller installation height significantly influenced gas holdup and power consumption. When L/D = 0.75, higher gas holdup and lower power consumption were observed. This work provides crucial theoretical and data support for optimizing the design of gas-liquid stirred tanks with gas sparging, offering clear engineering value for enhancing mass transfer efficiency and energy-saving operation in chemical processes.
Efficient Recovery of Lithium and Cobalt from Spent Lithium-Ion Batteries Using a ChCl-OA-H2O Deep Eutectic Solvent
The proliferation of lithium-ion batteries (LIBs) in portable electronics and electric vehicles has generated a pressing need for sustainable recycling of spent batteries. Conventional pyrometallurgical and hydrometallurgical routes suffer from low metal recovery efficiencies or require additional precipitants. This study introduces a clean and efficient process for recovering lithium (Li) and cobalt (Co) from spent LiCoO2 cathode materials using a choline chloride-oxalic acid-water (ChCl-OA-H2O) deep eutectic solvent (DES). The method exploits selective precipitation of Co as cobalt oxalate dihydrate (CoC2O4·2H2O) followed by water-content-regulated recovery of Li as lithium oxalate (Li2C2O4) via evaporation crystallization, eliminating the need for external precipitants. Under optimized conditions (molar ratio 1:1:8, solid-liquid ratio 100 g/L, 90 °C, 6.5 h), the leaching efficiency of Li reached 99.4%, with recovery efficiencies of 88.3% for Li and 97.8% for Co. The DES system demonstrated robust cycling stability, maintaining Li and Co recoveries of 78.1% and 92.8% after six regeneration cycles. This work provides a low-pollution, economically viable pathway for LIB recycling, contributing to resource sustainability and offering significant industrial potential.