Analysis of Energy Sustainability and Problems of Technological Process of Primary Aluminum Production

Abstract

This paper is devoted to the problem of magnetohydrodynamic stability (MHDS) in the energy-intensive process of primary aluminum production by electrolysis. Improving MHDS control is important because of the high costs and reduced efficiency caused by the instability of magnetic and current fields. In this work, a methodological analysis of modern theoretical and numerical methods for studying MHDS was carried out, and approaches to optimizing magnetic fields and control algorithms aimed at stabilizing the process and reducing energy costs were considered. This review identified key challenges and proposed promising directions, including the application of computational methods and artificial intelligence to monitor and control electrolysis in real time. In this paper, it was revealed that wave MHD instability at the metal–electrolyte phase boundary is a key physical obstacle to further reducing specific energy costs and increasing energy stability. The novelty of this paper lies in an integrated approach that combines modeling and practical recommendations. The purpose of this study is to systematically summarize scientific data, analyze the key physical factors affecting the energy stability of electrolyzers, and determine promising directions for their solution. The results of this study can be used to improve the energy efficiency and environmental friendliness of aluminum production.

1. Introduction

The production of primary aluminum by electrolysis is one of the most energy-intensive processes in the modern metallurgical industry. The electrolysis of aluminum melts takes place in a complex interaction of electric and magnetic fields, forming a magnetohydrodynamic (MHD) system. The stability of this system significantly affects energy efficiency, equipment reliability, and the final product’s quality [1].

Over the last decades, considerable efforts have been made to study and improve the magnetohydrodynamic stability of electrolysis processes. Numerous studies in the field of mathematical and numerical modeling have been carried out to predict the melt behavior in electrolyzers. Algorithms for controlling magnetic fields have been developed, and approaches to optimizing their configuration have been proposed. However, despite the progress made, the available methods often do not fully meet the requirements of industrial applications, especially under conditions of high production loads and power consumption [2,3].

Despite the widespread use of electrolysis plants, the problem of energy stability related to the control of MHDS remains relevant [4]. Insufficient stability leads to an uneven distribution of temperature and current fields, the development of turbulent flows, and electrode damage, increasing operating costs and reducing the overall production efficiency [5,6]. That is why the search for better methods and technologies to achieve the required control accuracy and process stability continues.

The aim of this study is to systematically summarize the existing scientific data, analyze the key physical factors affecting the energy stability of electrolyzers, and identify promising directions for their solution. This goal is reached by carefully examining current methods and technologies that focus on enhancing the energy stability of the primary aluminum production process. Among the tasks of this article is a review of the current state of the problem, the identification of promising directions, and the formulation of recommendations to improve the management of MHDS in the conditions of industrial electrolysis.

The scientific novelty of this article lies in the complex approach combining the latest mathematical methods of modeling, modern approaches to the optimization of magnetic fields, and the integration of theoretical data with practical technological solutions. This paper proposes directions for improving the accuracy of predicting the behavior of electrolyzers and the overall energy efficiency of the technological process.

The practical significance of this study is due to the need to reduce energy and economic costs, increase reliability, and improve the product quality at existing aluminum smelters.

This work is based on these equations: Maxwell’s equation (in a steady state), the generalized Ohm’s law for a moving conductor, the Navier–Stokes equation with Lorentz force, the condition that the melt is incompressible, and energy balance (heat transfer). In addition, other factors were also considered and are summarized in

Table 1. Analysis of additional sources.

№	Problem Type	Problem Literary
1	Temperature problems	[7,8,9,10,11,12,13,14,15]
2	Chemical problems	[16,17,18,19,20,21,22]
3	Physical problems	[23,24,25,26,27]
4	Economic problems	[28,29,30,31,32,33,34,35,36]
5	Electrical problems	[11,37,38,39,40,41,42,43,44,45,46,47,48]

for convenience.

This article includes an introduction that outlines the problem, purpose, and objectives of the research; a review of existing works and the relevance of the topic; a main section devoted to the analysis of existing technologies and approaches to MHDS management; and a conclusion containing conclusions and recommendations for further research in this area.

2. Research Methodology

The research methodology is based on a step-by-step study of the scientific literature, technological aspects, and existing engineering solutions. The research is conducted in several consecutive stages:

• Systematic analysis of the scientific and technical literature. The purpose of this stage is to determine the current state of research in the field of the primary production process of aluminum and MHD instability. Methods used include the following: analysis of publications in scientific journals (Scopus, Web of Science, Google Scholar, etc.) and study of patents and technical reports containing engineering solutions for process stabilization. Source selection criteria include the following: relevance (publications not older than 10 years, except for fundamental research) and applicability to industrial conditions.
• Classification of process challenges and factors of MHD instability. The aim is to systematize and structure the main causes of MHD instability in electrolyzers. Methods include the following: analysis of the collected literature; identification of the main factors affecting the stability of the electrolytic process; and classification of problem areas, including electrolyzer parameters (bath shape, magnetic fields, and electrode processes).
• Review of methods of identification and modeling of MHD-unsteadiness. The aim of this stage is to determine the existing approaches to the identification and analysis of MHD oscillations. Methods include the following: comparative analysis of mathematical models (hydrodynamic, electromagnetic, and complex); study of numerical modeling methods; and review of experimental diagnostic methods (current, voltage, and vibration sensors).
• Analysis of technical solutions to reduce MHD-unsteadiness. The purpose of this stage is to identify and systematize methods of MHD effects minimization in industrial production. Methods include the following: analysis of engineering solutions in patents and scientific articles; classification of ways to control MHD instability (constructive changes in electrolyzers, change in power supply modes, and active control of magnetic fields).
• Formation of conclusions and recommendations. The purpose is as follows: systematization of the analysis results and suggesting directions for further research. Methods are as follows: formation of the final table for visualization of the challenge and development of recommendations for optimization of the technological process.

The research methodology is based on a consistent study of the literature sources, analysis of existing solutions, and classification of MHD-unsteadiness control methods. The results obtained will make it possible to form scientifically sound recommendations for improving the process of primary aluminum production.

3. Primary Aluminum Production Process

3.1. A Description of the Steps of the Technological Process of Aluminum Production

Primary aluminum is produced through the electrolytic reduction of aluminum oxide (alumina) dissolved in a molten electrolyte [49]. The electrolysis plant is based on a carbon cathode placed in a steel casing lined with refractory bricks and carbon anodes lowered into the electrolyte and connected to an anode beam.

The production of primary aluminum is a complex process involving several key steps. The main stages of primary aluminum production from bauxite (the main raw material) include the following [50,51]:

• The extraction of bauxite: bauxite is the main raw material containing aluminum oxides. Extraction is carried out by open-pit mining, which allows large volumes of ore to be extracted.
• Enrichment: after mining, the ore is crushed, pulverized, and categorized to separate valuable components from waste rock.
• Bayer process: In this stage, bauxite is dissolved in a hot alkaline solution (NaOH) under pressure. Aluminum passes into solution in the form of sodium aluminate, and insoluble impurities are removed. From the cooled solution with the addition of CO₂, aluminum hydroxide (Al(OH)₃) precipitates, which, after filtration, washing, and calcination, turns into pure alumina.
• Electrolysis: Alumina is dissolved in molten cryolite (Na₃AlF₆) with the additions of AlF₃ and CaF₂ to lower the melting point. At a temperature of about 950–980 °C, the alumina decomposes in the electrolysis bath: aluminum ions are deposited on the cathode as a metal, and oxygen is released at the anode, reacting with the carbon of the anode to form CO₂. The process is accompanied by anode combustion.
• The extraction and casting of aluminum: liquid aluminum is regularly extracted from the bath with ladles and transported to the foundry, where it is cast into molds—ingots, bars, plates, and other products.

These stages require significant energy and material inputs, so the process of primary aluminum production is expensive and accompanied by environmental pollution [52]. About 95% of all the alumina produced is derived from bauxite, although nepheline, kaolin, and alunite are also used [53,54]. The production of 1 ton of aluminum requires about 2 tons of alumina.

The electrolysis of alumina is the key link in the whole process. The electrochemical reaction takes place in a bath where a current flows from the anodes through the electrolyte and aluminum layer to the cathode. The electric current causes the decomposition of Al₂O₃, resulting in the aluminum being released in liquid form and deposited at the bottom of the bath, from where it is extracted [55,56]. This technology remains the mainstream in the world practice of aluminum production due to its efficiency and sophistication.

That is why this electrochemical process is truly complex—representing a complex set of interrelated chemical, physical, and physicochemical processes.

In practice, electrolyzers are connected in series, thus forming a series of electrolyzers. A DC current is supplied from the anodes through the electrolyte and the metal thickness to the cathode, and then the current reaches the next electrolyzer in the series through the bushing (a set of conductors) [10]. Figure 1 shows the scheme of the inclusion of electrolyzers in the series.

3.2. Classification of Aluminum Electrolyzers

Electrolytic plants for the production of primary aluminum are classified according to various characteristics reflecting their technological and operational features. One of the main criteria is the scale of production: small (for laboratory purposes and small enterprises), medium (focused on specialized or regional markets), and large plants (mass production of tens of thousands of tons of aluminum per year at smelters).

There are two types of electrolysis: traditional Hall–Héroult plants that use direct current and cryolite solutions at high temperatures and newer induction plants that are more energy-efficient because they use induction currents.

Classification is possible according to the type of electrolyte: systems with liquid cryolite (950–1000 °C) are widely used, while solid-state and environmentally optimized systems with regeneration and recycling are being developed to reduce emissions and improve energy efficiency.

Depending on the level of automation, a distinction is made between manually controlled plants (mostly obsolete or small-scale) and fully automated lines, which include temperature, electrolyte composition, and product quality control systems. These plants are equipped with energy-efficient current converters.

According to the type of electrolyzer used, the plants are divided into Hall–Héroult cells, low-temperature cells, and regenerative cells. The latter are focused on energy saving and thermal optimization.

The plants are also differentiated by their energy source: hydroelectric (e.g., in Iceland and Canada) and thermal (in regions with cheap electricity). This approach makes it possible to take geographical specifics into account.

Among the important classification features are the technological features of metallurgical furnaces: the level of integration of waste processing and environmental safety. Modern “green” plants minimize emissions through closed-loop recycling systems.

According to the type of current supply, a distinction is made between direct current, alternating current, and combined current systems. The direct current system is predominant in classic cells, while combined solutions are used in innovative plants.

The type of current feed influences the performance. In a side current feed (SCF), the current flows through the side anodes, which is less efficient. A top current feed (TCF) provides a uniform current distribution and better process control.

The anodes are divided into annealed anode (AA) and self-annealed anode (SAA) systems. AAs require regular replacement and is accompanied by CO₂ emissions. SAAs last longer and are more efficient due to self-regulated combustion.

Thus, the division of aluminum electrolyzers according to the type of anodes is associated with different operating durations of the anodes and maintenance costs [56]. Plants with annealed anodes require regular replacement, which increases operating costs, while self-fired anodes provide a more stable operation and require less maintenance.

Currently, the electrolyzer is the dominant and most studied aluminum production technology. Figure 2 shows a schematic diagram of a TCF and SAA electrolyzer. It includes anode and cathode devices, an alumina feed system, and a current feed.

The following designations are adopted in Figure 2: 1—anode harness, 2—anode frame, 3—anode casing, 4—hopper of automatic alumina feeding system (AAFS), 5—board block, 6—cathode harness, 7—current-carrying blooms, 8—cathode casing, 9—gas collecting bell, 10—anode body, and 11—anode pin.

The anode device is a carbon anode consisting of an anode mass and covered with an anode steel casing; in the lower part of which there is a gas-collecting bell collecting waste gases. The anode rail and pins are provided for the anode suspension and current supply. The formation of the anode takes place during the electrolysis process, and as the anode burns, the anode is immersed in the solution. After complete combustion, the anode pins are removed from the anode body, and the liquid anode mass is poured into the resulting hole and coked to form a secondary anode. The cathode device is an electrolysis bath. In the lower part of the bath of the furnace blocks are cast iron blooms, diverting the current, and the joints between them are filled with furnace mass. The anode gases flow into the gas-collecting bell, where the afterburning of CO₂ and tarry substances takes place with the help of burners.

3.3. Structural Scheme of Primary Aluminum Production in TCF Electrolyzers

Aluminum production is carried out by the electrolytic reduction of aluminum oxide in the electrolyte melt at a high temperature. The structural diagram of this process in top-fed electrolyzers is shown in Figure 3. In addition to the main emission sources, physical effects are also recorded: thermal noise, vibrations, and electromagnetic radiation.

Electrolysis requires the supply of an electric current through a busbar system. The current flows from the anode risers through the electrolyte to the cathode rail connected to the next electrolyzer. The busbar must provide a uniform current distribution for the stability of the MHD system inside the bath.

The process is accompanied by the release of Joule heat, which helps to maintain the required temperature. The resulting aluminum, having a higher density, settles to the bottom of the electrolysis bath and is extracted by specialized equipment for further processing in the foundry department.

Electrolyzer designs vary, which affects the electromagnetic field distribution and, consequently, the MHDS. The main sources of the field are the currents of the anode and cathode systems distributed in the molten aluminum and electrolyte. The configuration of these systems and the magnetic properties of the casings form the secondary field.

The magnetic field is asymmetric and three-dimensional. It depends on the technological mode, control principles within the automatic process control system (APCS), and hardware. This necessitates the individual control of the electromagnetic effects in specific installations.

The aluminum industry is characterized by a high resource and energy intensity, environmental load, and labor intensity. The consumption of raw materials, energy, and dust emissions are determined by the type of plant, the quality of raw materials, and the technology used. The daily productivity of TCF electrolyzers averages 1250 ± 50 kg.

Plants with AAs are characterized by the design of an anode device consisting of a set of coal blocks mounted in anode holders connected to tires. A lifting mechanism is used to replace them.

In SCF systems, the anode frame is made of steel and suspended from ropes by a chain-link system. In these systems, the anodes are placed on the side as pins, and the height of their suspension is adjusted as they burn.

4. Analysis of Basic Aluminum Production Technology

4.1. Main Production Indicators of the Aluminum Electrolyzer

The key performance indicators of an aluminum electrolyzer play a key role in assessing its efficiency and the quality of aluminum produced. These indicators depend on a variety of factors, and these in turn affect the process. Let us look at the main ones, as well as the factors affecting the indicators and how the indicators themselves affect the electrolysis process:

1.

Current density (current per unit area of anode)

• Influence on the indicator: The current density depends on the electrolysis characteristics, e.g., the geometric size of the cells, as well as the composition and temperature of the electrolyte. The process equipment and anode design also influence this indicator.
• Effect on the process: Higher current densities may result in faster aluminum reduction, but there is an increased likelihood of the anode and cathode overheating, which can cause premature wear. Reduced current densities result in a reduced plant throughput.

2.

Electrolysis voltage

• Effect on the indicator: The voltage depends on the electrolyte composition, temperature, and condition of the anodes. Changes in the chemical composition or temperature of the electrolyte as well as the wear of the anodes and cathodes can cause voltage fluctuations.
• Impact on the process: A high voltage can lead to higher energy consumption, which increases the cost of production. Reducing the voltage improves energy efficiency but can slow down the electrolysis process, especially if the voltage is too low.

3.

Electrolyte temperature

• Influence on the figure: The temperature is directly related to the power supplied to the electrolyzer and the design of the equipment, including cooling and thermal insulation systems. The temperature is also affected by the characteristics of the anodes and cathodes.
• Effect on the process: Increasing the electrolyte temperature accelerates the electrolysis process and improves the conductivity of the solution, which reduces the energy consumption. However, too high a temperature may cause the accelerated destruction of anodes and cathodes and may result in unnecessary emissions of harmful substances.

4.

Cathode and anode quality

• Influence on the indicator: the condition of cathodes and anodes depends on their material, production technology, and operating conditions (temperature, current density, and aggressiveness of the electrolyte).
• Impact on the process: High-quality cathodes and anodes contribute to a more efficient electrolysis process, reducing energy costs and extending the equipment life. The wear of anodes and cathodes leads to higher energy costs and lower aluminum yields.

5.

Concentration of aluminum oxide in the electrolyte (Al₂O₃)

• Influence on the index: This index depends on how efficiently the process of aluminum reduction from the oxide is carried out. The concentration of Al₂O₃ in the electrolyte solution can vary depending on the temperature, composition, and state of the electrolyte.
• Process impact: An insufficient aluminum oxide concentration slows down the aluminum production process and increases the energy consumption, often causing undesirable anodic effects (AEs). Too high a concentration can also lead to poor conductivity and undesirable chemical reactions.

6.

Energy efficiency (kg of aluminum per kWh)

• Effect on the index: this index depends on all the above factors, such as current density, voltage, electrolyte temperature, and the quality and design of the equipment.
• Process impact: Energy efficiency has a direct impact on the cost of the aluminum production. An increased efficiency reduces operating costs and contributes to a greener production.

7.

Plant capacity (aluminum output per unit time)

• Influence on the index: This index depends on the size of the electrolyzer, the current density, the composition of the electrolyte, and the condition of the anodes and cathodes. The stability of all operating parameters, such as temperature and voltage, is also important.
• Impact on the process: Increasing the productivity requires increasing the capacity of the electrolyzer and improving the conditions for the stable operation of the equipment. However, if the production loads are too high, the equipment may overheat, which reduces its durability.

8.

Composition of gases emitted during electrolysis

• Impact on the indicator: The composition of the gases depends on the type of electrolyte used and the process temperature. The control of carbon dioxide emissions, which are released during the reaction with carbon anodes, is important.
• Process impact: Gas emissions can have an impact on the environmental performance of the production process. Controlling the composition of gases can minimize harmful emissions and improve the environmental safety of the process.

9.

Aluminum yield (main product)

• Influence on the figure: The aluminum yield depends on the efficiency of all electrolysis steps, the quality of the anodes and cathodes, and the composition of the electrolyte. Unstable parameters, such as a high temperature or voltage, can reduce the yield.
• Process impact: maximizing the aluminum yield increases the economic efficiency of production and reduces the cost of the final product.

All of these production parameters are interrelated, and changing one of them usually affects the others. For example, increasing the electrolyte temperature can increase the current density and accelerate electrolysis, but it can also increase the energy consumption and accelerate equipment wear. It is therefore important to maintain an optimal balance between these parameters to ensure that the entire aluminum production process is highly efficient, safe, and economically viable.

4.2. Anodic Effect

The anodic effect (AE) in aluminum electrolysis is a short-term but significant increase in the voltage in the electrolyzer due to the formation of a gas film on the anodes that prevents the normal flow of the current. It is a characteristic feature of the primary aluminum production process that largely determines both the technical and economic parameters of production [40].

The precondition (or “trigger factor”) for the anodic effect is usually a lack of aluminum oxide (Al₂O₃) in the cryolite melt. When the concentration of Al₂O₃ in the melt decreases, the anode begins to oxidize more intensively, and gas bubbles are formed on its surface (most often CO and CO₂ when carbon anodes are used) [41,42]. If these bubbles do not have time to leave the anode zone, they merge into a continuous gas film that isolates the anode from the electrolyte and sharply increases the resistance of the electric circuit. As a result, the voltage on the bath (electrolyzer) suddenly increases several times compared to the normal operating value. In this case, the overcoming of the so-called gas film by the current is accompanied by sparks.

The consequences of the AE can be manifold [57]. Firstly, the energy consumption increases because more energy is required to maintain the electrolysis process when the AE occurs. Secondly, the aluminum separation process is temporarily slowed down or almost stopped. Thirdly, gas emissions increase dramatically, and in addition to CO₂ and CO, fluorinated gases (e.g., CF₄) may be generated intensively, which negatively affects the environmental performance of the production. Fourthly, the equipment is subjected to additional stresses: high temperatures and surges accelerate the wear of the anodes and other structural elements.

There may be some advantages to the anodic effect in certain aspects, although these are not usually considered “desirable” [57]. Short-term and controlled anodic effects sometimes help to remove excess fluoride compounds or partially “clean” the anode surface, but such situations require precise technological control and are rare in practice. Much more importantly, the AE serves as a kind of indicator of unfavorable melt conditions (primarily Al₂O₃ deficiency), so that the operator can judge from its occurrence the need for parameter adjustments.

The disadvantages of the AE far outweigh the potential benefits. Its occurrence leads to voltage spikes, a higher specific energy consumption, the loss of production time, and more harmful emissions. Plant costs increase, and the quality and consistency of the aluminum produced may decrease.

The way to control the anodic effect is primarily to maintain the optimal concentration of Al₂O₃ in the melt [58]. For this purpose, the supply of alumina (aluminum oxide) to the bath is automatically or manually controlled. Maintaining the proper temperature, using anodes of the proper quality (structure and density), and ensuring regular electrolyte mixing and proper bath geometry are also important. Modern automated systems control the electrolyzer parameters (voltage, current, and material consumption) and feed alumina based on the signal of the approaching AE, thus minimizing its risk. If the effect does occur, it is eliminated as quickly as possible by an additional Al₂O₃ feed, increased melt circulation, or a short-term reduction in the current.

The AE in the aluminum electrolysis process occurs when the concentration of alumina in the electrolyte falls below the critical value (1.5–2.0%). This effect is accompanied by a sharp voltage spike on the electrolyzer due to the poor wetting of the anode by the electrolyte, as well as an increase in resistance at the anode–electrolyte interface [58,59,60]. The AE can be both regular (planned) and abnormal, which arises as a result of inefficient alumina feeding by the AAFS or changes in the thermal balance. However, it is worth noting that with the introduction of the AAFS, the need to cause routine AEs has decreased. Reducing the frequency of unplanned AEs, as well as their prevention, is the main task in AE management.

The anodic effect has a beneficial effect on the separation of the foam from the electrolyte, the dissolution of the sludge, and anode bottom leveling. The negative impacts of this effect are aluminum losses, power consumption (150 kW∙h), and a reduction in the series current (by 9–20 kA).

Thus, the AE is an important technological feature reflecting the state of the “melt–anode–cathode” system; its prevention or rapid elimination reduces the energy consumption, metal losses, and harmful emissions, which significantly improves the economic efficiency and environmental safety of aluminum production.

4.3. MHD Instability

In the process of electrolysis, the aluminum melt is in constant motion due to the influence of electromagnetic forces on it [61]. In addition to the metal, the cryolite is also in intensive motion—the movement of which can be explained by the influence of evacuating anode gases and temperature convection forces on it. Thus, there is a mutual influence of the metal motion and cryolite motion in the electrolyzer bath.

MHD instability in the electrolysis process is the result of the interaction of a large electric current that flows through the “liquid aluminum–electrolyte–anode” system, with the magnetic field formed by the same current and the surrounding power infrastructure (busbars, adjacent baths, etc.). The main prerequisites and factors of this challenge contributing to the development of oscillations are discussed below, and the ways of combating this phenomenon are described in more detail [62,63].

The indicator of the electrolysis process efficiency as the current yield is often negatively affected by the circulation and metal agitation in the electrolysis bath. MHD instability consists of the reduction in the plant productivity as a result of a decrease in the current yield [64].

Now let us consider in more detail the causes and nature of the oscillations. During the operation of an aluminum electrolyzer, a significant direct current (several hundred kA in modern high-performance plants) flows through the liquid aluminum layer and the electrolyte melt. This current generates a magnetic field, and the nature of the field is largely determined by the configuration of busbars, the location of neighboring baths, and other infrastructure elements. As a result, Lorentz forces appear in the molten aluminum (which is an electrically conductive liquid).

When a current flows in the metal and electrolyte [65] and a magnetic field is present, these forces create local “shocks” in the liquid. If the magnetic field distribution is non-uniform, a series of vortices and local flows and wave-like oscillations of the “metal–electrolyte” boundary appear. Lorentz forces can displace the liquid aluminum relative to the electrolyte layer, forming waves of different amplitudes at the interface between the two media (the aluminum is heavier than the electrolyte and is located at the bottom). With a certain combination of parameters (layer thickness, current, magnetic field, electrolyte density, and temperature), these waves can be amplified due to the resonance or “pumping” of energy from Lorentz forces.

It is important that when liquid aluminum is displaced the current can be redistributed and the magnetic field can be enhanced in the zones where the aluminum layer “rises” [61]. This creates positive feedback: once the movement of the metal in one point occurs, it can cause a further strengthening of the Lorentz force and an even greater deformation of the metal–electrolyte boundary.

The following can be attributed as the factors that have a significant influence on the formation of MHD instability:

• Busbar geometry and design: an unfortunate busbar configuration resulting in an asymmetric or overly strong horizontal component of the magnetic field increases the likelihood of MHD instability.
• Magnitude of the current flowing: The higher the current, the stronger the created magnetic field is and the higher the risk of inducing oscillations. In large baths with currents of 300–500 kA and more, the task of preventing MHD instabilities comes to the forefront.
• Aluminum layer thickness and electrolyte level: A too thin aluminum layer makes the metal–electrolyte medium more mobile, which makes it easier to form waveforms. If the aluminum thickness is very thick the mixing losses increase and the optimal geometry of the current distribution may change.
• Electrolyte density and temperature: Changes in the electrolyte density (e.g., with variations in the Al₂O₃ content or other additives) affect the wave processes. Higher temperatures can contribute to changes in the melt viscosity and make the motion more intense.
• Location of neighboring baths and mutual influence: in large shops, the electrical circuits of the baths are often interconnected, and the magnetic fields from neighboring electrolyzers overlap, amplifying or, less often, compensating the waves.

The main consequences of the development of MHD instability are as follows:

• Metal level fluctuations: during strong fluctuations, aluminum can locally rise to the anodes, increasing the risk of short circuits and an uncontrolled current flow through the metal.
• Increased energy consumption: changing the current distribution in the bath leads to increased losses and voltage fluctuations, which ultimately have a negative impact on energy efficiency.
• Reduced metal quality: wavy movements can promote an increased mixing of the slag, electrolyte, and metal, increasing the risk of aluminum contamination or impurities.
• Accelerated wear of equipment: high dynamic loads on anode surfaces and bath walls can shorten the service life of process components.
• As a result of MHD instability growth, the aluminum surface undulations become comparable to the interpole gap. When a critical value is reached, the continued growth of the MHD instability can provoke a sharp deterioration of the electrolyzer performance. The identification of MHD instability is one of the most difficult challenges; the solution of which directly affects the quality of the products.

Due to the significant negative effects, ways to combat and prevent MHD instability have been developed. One such way is to optimize the busbar layout and configuration of the electrolyzer [66]. The main tool is a well-thought-out design of the current-carrying busbar placement that minimizes the uneven distribution of the magnetic field. Symmetry is important, using special busbar arrangements (e.g., reverse current busbars running along a specific geometry) to reduce horizontal field components. Another method of prevention is to control the thickness of the aluminum layer. By using systems that measure metal levels in real time, technologists maintain the layer at a certain optimum thickness. This reduces the mobility of the medium at the same current values and reduces the risk of the system going into high oscillation mode.

The control of the bath supply regimes also contributes to reducing the probability of MHD instability. Maintaining a stable concentration of Al₂O₃ in the electrolyte and temperature control are important not only for avoiding AEs but also for reducing the probability of MHD disturbances: an electrolyte with the right density and viscosity is less susceptible to uncontrolled mixing [67].

Measures to eliminate MHD instability often include the installation of active vibration suppression systems. In some cases, sensors are implemented that continuously monitor the metal level and wave pattern. If signs of dangerous oscillations are detected, the control system can quickly adjust the current strength or change the feed mode.

Adjusting the current or voltage when critical fluctuations occur also avoids undesirable effects. In large electrolysis plants, if increasing instability is detected, the total current can be temporarily reduced or the load can be redistributed between the baths to reduce the amplitude of the fluctuations. Similarly, the voltage can be slightly changed for a short time to avoid operating in the dangerous “resonance” range.

It is worth noting that a pre-calibrated arrangement of anodes and cathodes has a positive effect on the stability of the MHD properties of the melt. When designing an electrolyzer, the arrangement of anodes, cathodes, and jumpers is taken into account in order to maximize the field homogeneity and eliminate large current gradients.

Thus, MHD instability is a complex phenomenon with many causes related to the physics of the electrolyte and metal, electromagnetic processes in the bath, and the current supply scheme. The fight against this instability is largely determined by the proper design of the electrolyzer and busbar system, the maintenance of an optimal aluminum layer thickness, and stable electrolyte parameters, as well as the use of monitoring and automatic control systems that allow for the prompt correction of modes at the first signs of fluctuations. The better the entire chain—from design to daily parameter control—the higher the stability of the operation and the lower the cost of maintaining the required aluminum quality.

5. Scientific Papers and the State of Scientific Research

This review presents an analysis of the current state-of-the-art industrial aluminum electrolyzers and identifies the most important challenges and their prospective solutions by synthesizing information on already existing or previously applied ideas and technological solutions in this field.

Currently, aluminum is produced by the electrolysis of alumina [68,69]. However, in addition to aluminum, other metals, both non-ferrous and ferrous, are also in demand in industry [7,70].

Along with the mineral and raw materials complex [71], the fuel and energy complex is of particular interest for this study, since the efficiency of its resource utilization directly affects the cost of the extraction, processing, or metallurgical processing of various types of mineral raw materials [54,72]. The main engine of progress in this area is the research and development of new technologies initiated by the state apparatus, which allow for the solving of the set tasks [53,70].

However, industrial modernization should take into consideration the issue of its impact on the environmental situation [73,74]. Any economically beneficial attempts at modernization should be consistent with the current level of the environmental culture [75]. In other words, innovations in economic activity should not have a negative impact on the environment [76].

In order to achieve the stated objective of this study, the following objectives need to be addressed:

• Analyzing the problem of the research topic;
• Analyzing existing research on the research topic;
• The identification of unresolved scientific challenges related to the object of research;
• Analyzing promising theoretical research that can be applied to solve the identified open science challenges.

5.1. Analyzing the Problem of the Research Topic

Carl Soderberg’s method played an important role in the development of the aluminum industry by increasing the productivity and reducing production costs through the use of self-calcining anodes. This simplified the electrolysis process and increased the aluminum production. As science and technological progress has developed, much has been refined and changed in this technology [69]. The improvement of the methods of aluminum production remains an urgent task to this day. Scientists around the world are studying the processes that make up the electrolyzer system. The identification of new solutions and the development and refinement of monitoring and control systems are necessary to improve the economic and environmental efficiency of the process [77,78].

To improve the control of the system, modeling is used, which allows for the qualitative and optimal control of the technological process, as it reflects the dependencies between the parameters, the influence of various factors, etc. [79]. A lot of research has been devoted to the mathematical modeling of the processes occurring in the electrolyzer [80]. The future development of technology is dependent on the improvement of the current work: mathematical models, algorithms for numerical solutions, and computing systems for modeling.

MHD challenges for aluminum electrolyzers are becoming increasingly important as the current in modern electrolyzers increases. The electric current, interacting with the associated magnetic field, creates stirring effects that limit cell performance and can cause the instability of the interface between the liquid aluminum and the electrolyte. This can lead to significant power losses, technology failures, and increased environmental pollution [81].

To minimize the harmful effect of the magnetic field, special designs of the harness are created to ensure its uniform distribution over the electrolyzer and the working space [82]. This is especially important for high-power electrolysis plants, since the magnetic field generated by the current affects the nearby electrolyzers. Another method concerns the reduction in the magnitude of the horizontal currents flowing in the bath [8].

However, there is no single design that would be able to ensure an absolutely uniform distribution of currents and, as a consequence, the distribution of the intensity and electromagnetic field over the entire area of the installation [83]. Therefore, the task of future research is to control, monitor, and analyze the parameters measured directly, i.e., the magnetic field strength.

The phenomenon of metal undulation occurs when the voltage fluctuates and the anode touches the metal and the electrolysis reaction stops [84]. The cause of this phenomenon is related to the redistribution of the current in the melt and the electromagnetic forces acting on it. This can be observed in medium- and high-capacity electrolyzers, but solutions to eliminate this challenge have been developed by practical experience, and their application is not always reasonable. The stability of magnetogravity waves is related to the oscillation of the magnetic field. The components for the horizontal direction are small because of the symmetry in the center of the bath. But the small value of the magnetic strength in the center does not guarantee the uniformity of the magnetic field [85,86].

The highest values of the magnetic flux current density are recorded in anode blocks and in plugs made of nonmagnetic materials that transmit the current to and from the electrolyzer. When modeling the magnetic field, to simplify the challenge, the assumption of the symmetry of the magnetic field with respect to the middle, transverse, and vertical planes of the electrolyzer is often made, which does not allow us to see the full model of the MHD force, which is asymmetric and becomes the object of future research.

Existing software packages for modeling electromagnetic fields in electrolyzers are constantly being improved and maintained [87]. A clear understanding of the mechanisms of their occurrence is required to determine the onset of MHD instabilities [88,89].

The above challenges are solved with a satisfactory assessment, so it is necessary to pay more attention to this area of research to achieve the desired result of systems for the modeling and monitoring of the electrolysis process [90]. The existing directions for stabilizing the magnetic field strength, developing control algorithms, and developing systems for monitoring the condition of electrolyzers are effective but not fully developed [91]. Thus, it seems reasonable and important to create complex monitoring modules for improving the efficiency of electrolyzers.

5.2. Analysis of Existing Studies

The mathematical modeling of electrolyzers and the prediction of their unstable behavior have reached a certain maturity since the first MHD principles were introduced. The application of simulation techniques has become the main tool to analyze the electrolysis process with different parameters and at different time intervals [92,93]. Recent works develop three-dimensional unsteady models considering real production data and investigate numerical methods to simulate the behavior of an industrial aluminum electrolyzer [94,95]. The study of oscillation relaxation in an electrolyzer with a free boundary between phases has shown that the smoothed particle hydrodynamics method is applicable to more realistic three-dimensional problems, such as the feedback control of aluminum electrolysis, which can provide the prompt prevention of MHD instability.

Numerous studies have been devoted to investigating the stability of the interfacial electromagnetic field wave, which is the key to achieving a high energy efficiency and production safety [96,97]. The three-dimensional model of the two-phase quasi-stationary bath–metal flow combined with the transient electromagnetic force reveals that the local cathodic electric cutoff has little influence on the overall melt flow field pattern, but the local metal velocity and interface deformation change to a certain extent. Compared with the uniform cathodic current distribution, the differential cathodic current distribution approach can help suppress the interface deformation and improve the anode–cathode spacing distribution over the cells [43].

A multiphase MHD model based on computational fluid dynamics is also designed to simulate the melt flow and deformation of the bath–metal interface. It considers the complex physics of the MHD challenge by coupling two immiscible fluids, an electromagnetic field, and the Lorentz force and flow turbulence and includes detailed cell geometry, such as the side protrusion profile and all channels. However, some aspects of aluminum reduction cell modeling are still not included in the current models, such as the bubble flow under the anode, thermal effects, the anode burnout mechanism, the dynamic protrusion profile, and MHD waves at the bath–metal interface.

The study of three-dimensional mathematical models makes it possible to understand the emergence, growth, and coalescence of bubbles under the anode and the transition of bubbles from the microlevel to the macrolevel [66]. Taking into consideration the nonlinear temperature dependence of all the physical characteristics of materials, it was found that the frequency of the bubble release increases with an increasing current density. The bubble motion, magnetic forces, and thermal convection significantly affect the overall dynamics of the electrolyte and metal.

In recent decades, a number of solutions have been proposed for the MHDS of aluminum electrolyzers [98,99]. But previous studies did not at first consider the influence of bath boundaries and later, based on the surface of the electrolyte and aluminum layers, reduced the challenge to two-dimensional models by averaging the forces over the layer thickness [100,101]. More recent works consider a fully three-dimensional structure of these layers, aiming to investigate the MHD flows observed in the electrolyzer or the bubble flow in the electrolyte [59,67].

Particular attention is paid to optimizing the energy state. The thermal balance of aluminum electrolyzers must be strictly controlled to improve the efficiency and sustainability of this industrial process. Electrolyte superheat is an important indicator of the thermal balance condition of the cells for aluminum recovery. In industrial practice, daily accurate measurements of superheat in each element is cost prohibitive. A common alternative is to calculate superheat based on additive concentrations in the electrolyte, which involves the challenges of high errors and long delays. A diagnostic method based on Bayesian networks takes into account important symptoms and factors using more useful information, not only the calculated superheat [102].

In [9], for the first time, transient thermal and electric fields are modeled and confirmed by industrial measurements over the lifetime of the anode. Transient thermoelectric modeling demonstrates the effect of increasing the operating voltage on both the anode cover and the side ledge, which leads to the wear of these elements.

The increase in productivity is associated with an increase in the current strength; in this regard, negative consequences of the effect of thermal loads on the system are possible [103]. Preliminary analyses of the thermal state of the electrolyzer show that the temperature of the structural elements of the cathode body increases significantly when the current increases up to 350 kA [104,105]. To control the thermal balance, algorithms for selecting a given voltage and maintaining process parameters within specified limits should be used [10,105].

In order to achieve the energy efficiency of the aluminum production process, accompanied by a large energy consumption, the level of energy consumption is reduced [106,107]. Using the least squares support vector method, electrolyzer voltage prediction models based on metaheuristic algorithms (Ant Lion Optimizer) are developed and optimized [108]. The results show that the combined model can accurately determine the voltage set point to provide effective theoretical guidance for actual electrolytic aluminum production.

The load characteristics have a significant influence on the static voltage stability analysis of the power grid because it has a large number of relatively centralized aluminum electrolytic loads [11]. In the current stabilization mode, accurate models of load characteristics are built, which show that the static voltage stability margin indices can describe the static voltage stability of the grid well [109].

To improve the energy and environmental parameters of the aluminum electrolyzer, a technical solution has been proposed that provides the division of the SAA into separate units in a common anode housing [60]. The use of this anode device reduces the volume of the gas layer under it, formed during electrolysis, thereby reducing the energy consumption to overcome the resistance of the layer. The composite anode conductor reduces power losses and costs because less base metal is required, and its strength is increased, and the electrical contact area remains stable [110].

In “Production Management of Aluminum Electrolysis at Super Low Voltage”, a new integrated energy-saving technology characterized by protruding cathode blocks and cathode collector rods with a high conductivity was demonstrated and tested in 300 kA electrolysis plants [111]. Technical and operational parameters were optimized to maintain normal operations at an ultra-low voltage. Under certain process conditions, the electrolyzers operated efficiently and stably, the electrolyte superheat was maintained in a moderate range, the side ledge of the cell was maintained at an appropriate thickness, no sludge accumulated on the bottom surface, and no coal dust accumulated on the electrolyte surface.

Electrolyzer monitoring systems ensure the stability and efficiency of the electrolytic process, improve safety, and extend the life of the equipment. They monitor and automatically analyze parameters such as current, voltage, temperature, pressure, and alumina consumption, which determine the quality of the product. In case of any deviations, the system warns of dangerous or abnormal modes and provides the opportunity to react quickly to malfunctions. This is particularly important given the risk of hydrogen leakage and the associated requirements for tightness. By continuously analyzing the operating environment and adjusting process parameters in real time, the energy efficiency is improved and resource consumption is optimized. In addition, modern monitoring systems collect and process a large amount of data, which makes it possible to predict possible failures in advance, plan repairs or the replacement of worn components, and keep the electrolyzer in an optimal condition throughout its lifetime.

In modern process industries, many parameters or states can be obtained from sensors, and these parameters or states often have a close relationship with the operating conditions [112,113]. Unfortunately, the process often operates in different modes, and the readings are often unknown [114]. In practice, labeling the sampled data is expensive and time-consuming, so identifying the operating conditions of an industrial process is difficult. In addition, sampled data from an industrial system often contain emissions or noise. Therefore, a reliable monitoring method for a multimode process is particularly important and in demand [115,116,117].

The robust dictionary learning method for processes with multiple unknown modes is able to provide satisfactory monitoring results and is more suitable for real industrial system processes [118]. When a new sample is generated, it is reconstructed by different nested dictionaries, and the smallest dictionary reconstruction error is computed as a classifier for process monitoring and fault detection.

In practice, due to complex mechanical processes such as heating, volume inhomogeneity, and different chemical reaction characteristics, there is a nonlinear relationship between variables in industrial systems. The nonlinearity creates some difficulties for monitoring. To ensure that the monitoring system can work properly in nonlinear industrial processes, the nonlinear relationship between variables should be considered [119,120,121]. A new fault detection and isolation method based on kernel dictionary learning is presented in [122]. Its special feature is the use of an iterative reconstruction method that isolates faults. Thus, it is suitable for the process monitoring of nonlinear industrial processes [123,124,125].

The aluminum oxide concentration is an important parameter in the production process of aluminum electrolysis [68]. Due to the difficult industrial production conditions and complex physical and chemical reactions in the aluminum reduction electrolyzer, online measurements and real-time monitoring are still not possible at present [44,45,126]. To solve this challenge, a study [127] proposed a soft alumina concentration determination model based on deep networks, which are a kind of artificial neural network. However, this model may have some limitations for different elements and different periodic operating conditions, such as the local AE, pole change, and bus lift in the same element. But despite this, the model improves the accuracy and universality of the predictions [1,128].

Traditional aluminum electrolysis fault diagnosis methods have challenges, such as a low accuracy, small prediction lead, and high CPU utilization, which reduces their popularity in enterprises [129,130]. In order to address the above challenges, fault diagnosis methods with switchable two-level classifiers are developed [131]. The input data are first evaluated using the first-level algorithm. If it is determined that there is no fault, the result will be output directly [4]. If it is determined that there is a fault in the electrolyzer, the data will be sent to the layer two network for specific fault diagnosis. Experimental results show that this method can switch between different algorithms according to different situations and save computational resources; in addition, the accuracy and advance prediction are greatly improved [132,133,134].

5.3. Existing Unresolved Scientific Issues

An uneven distribution of currents and voltages in the aluminum electrolysis process leads to a number of negative consequences that directly affect the efficiency and economics of production. Localized overvoltage worsens energy consumption and increases the risk of intensive electrode wear and material consumption. Due to an unbalanced current, the heat generation in individual zones is unequal, which can lead to temperature irregularities, uneven electrolyte melting, and unstable aluminum formation. This, in turn, affects the product quality and can lead to additional costs for parameter adjustments, electrode replacements, and remediation work. All of these factors combine to reduce the overall production profitability and complicate the operational control in the electrolysis shop.

The emergence of electromagnetic forces in the electrolyzer melt is caused by the interaction of the magnetic field resulting from the current flow through the structural elements and the current flow through the electrolyte and metal [135,136]. The bubble motion in the space between the board and the anode causes gas-hydrodynamic forces. This is caused by a non-uniform and unstable current distribution in the cell, which leads to the instability of the magnetic field and electromagnetic forces in the melt. This challenge is still insufficiently studied and is raised in many works.

One of the main indicators of production efficiency is the uniformity of the current distribution of anode devices. To evaluate this indicator, the current distribution non-uniformity coefficient is used on annealed anode baths. In [137] the factors influencing the current distribution are analyzed in detail. Using a physical and mathematical model of the anode operation, dependences describing the occurrence of irregularities on the anode sole were obtained. It was found that one of the reasons for the appearance of cones is the formation of a passivating layer of perfluorocarbons on the anode sole. The electrothermal condition of the busbar system of electrolyzers is the main factor affecting the MHDS and current distribution [138,139,140].

The horizontal current plays an important role in the MHD instability and power consumption. To reduce the horizontal current, steel multi-section rods are proposed. A transient three-dimensional mathematical model is developed to investigate the effect of their structure on the current density and MHDS, but their mechanism is not fully understood [141,142,143]. Nevertheless, the results show that steel rods with multiple collectors effectively reduce the horizontal current by increasing the current path. Compared with steel rods with double collectors, they cause a decrease in the voltage drop across the cathode. The maximum fluctuation of the bath–metal interface can also be reduced, which contributes to the MHDS [144].

The material and shape of the cathode affect the current density in an aluminum electrolysis cell. When the electrical resistance of the material increases, a higher voltage drop across the cathode is observed, but the horizontal current in the metal decreases. Modified cathode shapes with an inclined cathode surface, a higher collector rod, and cylindrical protrusions reduce the magnitude of the horizontal current in the metal. Cylindrical protrusions lead to a local increase in the horizontal current at their tops, but the average current is less affected.

The condition of the cathode device in the aluminum electrolysis process largely determines the efficiency and stability of the entire production cycle. An aggressive environment with high temperatures, the chemical attack of the melt, and significant electromagnetic loads gradually destroys the cathode, accelerating its wear and tear and leading to increased energy costs. Uneven or excessive damage to the cathode device can lead to a distorted current distribution, degrade the quality of the metal produced, and shorten the life of the entire electrolyzer. This increases the risk of unscheduled shutdowns, increases cathode repair and replacement costs, and makes it more difficult to ensure consistent aluminum yields of the required purity and quantity.

By simulating the real conditions of electrolytic aluminum production, studies of garnish behavior depending on the chemical properties of the electrolyte, electrolyzer operation modes, and lining materials are carried out [145]. The formation of protrusions occurs due to the heat flow arising from the temperature difference between the electrolyte and the walls of the electrolyzer. The change in the shape of the working space in the electrolyzer during electrolysis is determined by the thickness of the protrusion. Thus, the lateral protrusion has a heterogeneous composition, which depends on the electrolyte composition and heat transfer coefficients [146].

The most common thermal insulation materials used in the cathode lining of aluminum electrolyzers are molar (diatomaceous earth), calcium silicate, or vermiculite-based materials. The thermal insulation layer is critical to the overall thermal stability of the cell and is vulnerable to volatile substances, such as sodium vapor, that can penetrate the carbon cathode and refractory layer [147]. Carbon linings in the form of cathode blocks with different graphite contents are subjected to mechanical wear under electrolyzer conditions due to the rapid melt movement, and this is more significant than for amorphous carbon blocks [148,149].

Due to the increasing trend of the increasing current strength in modified aluminum electrolyzers, studies are conducted at both a characteristic cathodic current density (0.45 A/cm²) and a high cathodic current density (0.7 A/cm²) [46]. The results show that the microstructures of carbon cathodes can be modified by Joule heating and electrostatic charging with a higher current density during aluminum electrolysis. The penetration of the sodium and melt causes significant stresses and deformation in the carbon cathodes, which gradually leads to performance degradation. This shows that increasing the current in aluminum plants can aggravate the deterioration of the cathode material.

In the study [47], the authors presented data on the improvement of materials for lining carbon cathodes used in the aluminum production by electrolysis. But the challenges of the sodium ion and electrolyte velocities in sole block materials depending on permeable pore sizes and the optimization of porous material structures have not been realized yet.

5.4. A Theoretical Study of Challenge Solving

The monitoring and diagnostics of the electrolysis process play a key role in ensuring the stable and cost-effective operation of aluminum productions. In conditions of high temperatures, aggressive chemical influences, and uneven electromagnetic loads any abnormalities in the operation of the electrolyzer can lead to significant financial losses and a reduced product quality. Modern monitoring systems help to identify potential malfunctions in advance, adjust process parameters, and optimize energy consumption. It is thanks to comprehensive monitoring and timely diagnostics that the prerequisites for improving the efficiency, safety, and competitiveness of the entire aluminum smelting process are created.

The most important solution to eliminate the challenge of metal agitation is to develop a system for monitoring and identifying the intensity across the entire area of the electrolysis cell [106]. The real-time monitoring mode of electromagnetic parameters allows for minimizing the causes of metal agitation and optimizing aluminum production.

Most industrial systems frequently switch their operating modes due to various factors, such as changes in raw materials, static parameter settings, and market requirements [150,151,152]. To guarantee the stable and reliable operation of complex industrial processes under different modes of operation, a monitoring strategy should adapt to different modes of operation [153]. In addition, different modes of operation usually share some common patterns. To meet these needs, a structural dictionary learning-based method of multimode process monitoring is proposed [154]. Compared with traditional methods, the proposed approach refutes the assumption that each operating mode of an industrial process should be modeled separately. Thus, it can effectively detect faulty states [155,156]. Above all, it is suitable for monitoring real industrial systems [12,13,157].

A multi-model and multi-level aluminum electrolyte fault diagnosis method is proposed to effectively predict aluminum electrolyte faults [158]. The algorithm successfully combines data analysis with image analysis. Thus, the faults can be comprehensively detected from different angles, which makes the extraction of fault characterization information more complete. Moreover, the algorithm has demonstrated a low false positive rate, high reliability, and good application prospects.

For analyzing and controlling industrial electrolyzers, their various parameters are calculated using the cell voltage. In [159], the frequency segmentation of the cell voltage for filter designs is used as a basis to obtain these parameters. The proposed method is more sensitive to changes in the electrolyzer state and provides more detailed information about the electrolyzer online, thus providing a more reliable and accurate basis for monitoring the cell state and making control decisions [160,161].

The production of primary aluminum is a continuous and complex process that must be carried out in a closed loop, which limits the possibilities for experimentation to improve production. In this sense, it is important to have ways to computer simulate this process without directly affecting the plant, since such direct intervention can be dangerous, expensive, and time-consuming [162]. This challenge is addressed by combining real data, an artificial neural network method, and clustering techniques to create virtual sensors to estimate the temperature, aluminum fluoride percentage, and metal level in the electrolytic bath [163,164]. The results demonstrate the effectiveness and feasibility of the proposed virtual sensor approach in the aluminum industry, which can improve the process control and save resources [165,166].

The optimal control of the aluminum electrolysis production process has long been a difficult industrial challenge due to the inherent difficulty in creating an accurate dynamic model [167]. In [168], a new robust optimal control algorithm based on adaptive dynamic programming is proposed, where the system obeys the input constraints. To create an accurate dynamic model, a recursive neural network is used to reconstruct the dynamics of the system using the input–output production data. This controller is applied to the system and proves to be efficient and fast.

Aluminum electrolysis cells are characterized by harsh operating conditions where several measurements have to be conducted manually. Due to the operating costs associated with manual sampling, the sampling rate of these measurements is low. Thus, the information in the data can be limited, making it difficult to develop reliable data-driven methods for aluminum electrolysis. Over the years, a wide range of physics-based models have been developed to provide an excellent systematic knowledge of the dynamics in aluminum electrolyzers. However, due to very complex and interconnected sub-processes, current physics-based models are inadequate to accurately express the dynamics in the cell. The combination of inadequate prediction models and low sampling rates makes the estimation of process parameters in the aluminum electrolysis process less accurate than required for an optimal and safe cell operation. In [5], a new hybrid modeling approach is proposed that takes into account inadequate prediction models and low sampling rates. The signal estimation of the unmodeled dynamics is integrated into the extended Kalman filter as a pseudo-measurement to improve the estimation of the system states [169]. The new method is applicable to signals with stationary periodic unmodulated dynamics. A case study is conducted on simulated data from a sub-process describing the mass balance in an aluminum electrolyzer.

The developed new nonstationary three-dimensional mathematical model of the electrolyzer allows performing coupled thermoelectric and MHD calculations factoring in the formation of lateral protrusions [48]. It considers the nonlinear dependence of the electrical conductivity and thermal conductivity coefficients of materials on temperature and, for ferromagnetic materials, the nonlinear dependence of the magnetization on the magnetic field strength. The model takes into account internal heat sources due to the flow of the electric current, exothermic reactions, and additional thermal effects associated with the loading of raw materials and phase transitions. The experimental verification of the mathematical model is carried out on a top current electrolyzer. This paper presents computational and experimental data of magnetic, electric, thermal, and hydrodynamic fields. This model can be used to evaluate the performance and design parameters of new and upgraded aluminum electrolyzers. Further studies are aimed at refining the calculated results by improving the mathematical model.

In aluminum reduction technology, the current in the cell is first distributed through a series of anodes before entering the electrolytic bath and metal pad. This can be visualized as a series of parallel resistances, each with its own parameters and resulting current. In [170], a transient model of the anode current distribution is presented. This model tracks the properties of each anode (lifetime, size, and presence or absence of slots) and predicts the local current, anode-to-cathode spacing, and current efficiency throughout the anode cycle. More complex phenomena, such as the deformation of the metal pad and the formation of protrusions around new anodes, are also accounted for by simplified laws. The model has been thoroughly validated against industrial measurements and can now be used to provide information on metallurgical furnace behavior, estimate local pole spacing, and make design and operating decisions.

The current practice in the field of the anodic oxidation of aluminum and its alloys is based mainly on a set of partial empirical experiences of technologists obtained during electrolyzer operation. A deeper and more comprehensive identification of the influence of chemical and technological factors acting in the anodic oxidation process allows for the optimization of the parameters and the process [171]. With the help of electrochemical experiments, the optimal parameters of the current supply mode and current density are determined. To ensure a constant voltage on the bath, a minimum interval of change in the electrode polarity is selected, since the voltage depends on the time of the period of change in its polarity.

The aluminum electrolysis process is traditionally accompanied by high gas emissions, solid waste generation, and significant energy consumption, which raise environmental safety and sustainability issues. The reduction in harmful emissions, efficient waste management, and optimization of energy costs are becoming an integral part of technological solutions to minimize the negative impact on the environment. In order to achieve environmental neutrality and comply with industrial safety standards, enterprises have to implement modern systems of gas purification and waste treatment and improve methods of the production process control. Only an integrated approach and the continuous improvement of technologies make it possible to reduce the environmental footprint while maintaining the high efficiency and competitiveness of the aluminum industry.

Aluminum is an energy-intensive material whose environmental impact is highly dependent on the amount of electricity consumed during the smelting process [172]. Studies [92,173] simulate the estimated environmental impacts of primary aluminum production under different integrated assessment modeling scenarios based on common socioeconomic pathways and their climate change mitigation scenarios. The results project an average global carbon intensity of between 8.6 and 18.0 CO₂ eq/kg in 2100, compared to 18.3 CO₂ eq/kg at present, which can be further reduced under the mitigation scenarios. Scaling the impact of aluminum production on the global demand shows that the total emissions would be between 1250 and 1590 CO₂ eq/kg for the baseline scenarios by 2050, while absolute decoupling is only achievable with stringent climate policies that fundamentally change electricity consumption patterns. Achieving greater emission reductions will require cyclical strategies that go beyond the primary material production itself and involve other stakeholders [174,175].

Carbon-free aluminum electrolysis based on inert anodes is a breakthrough and fundamental technology to achieve the goal of carbon neutrality in the electrolytic aluminum industry, since only O₂ is released instead of the CO₂ produced by the conventional aluminum electrolysis using carbon anodes [176,177]. In recent years, the optimization of the material preparation process and advances in electrolysis have opened new possibilities for industrial applications of inert anodes [178,179,180]. Metals and metal–ceramics are the most likely alternatives to carbon anodes, possessing a satisfactory electrical conductivity, high resistance to thermal shock, and adequate corrosion resistance [181].

Carbon products, such as anodes and packing paste, must have well-defined physical, mechanical, chemical, and electrical properties to effectively perform their functions in the aluminum electrolyzer. The physical and mechanical properties of these products are set during the molding process. The optimization of the molding process is crucial to improve the properties of carbon products and, hence, to improve the energy efficiency and reduce greenhouse gas emissions in the Hall–Héroult process [182,183,184].

In [185], the description of the gas evolution in aluminum electrolysis is presented in more detail. The influence of temperature and the chemical equilibrium on steady-state conditions is also discussed. A typical chemical composition of the feedstock is presented to evaluate its contribution to the gas composition, especially with respect to its impurity levels.

The modernization of the cathode device of an aluminum electrolyzer is one of the most important areas for increasing productivity and reducing costs in the industry. The extreme conditions inside an aluminum electrolyzer, due to high temperatures and aggressive media, inevitably lead to cathode wear and the deterioration of its functional characteristics. New materials, design solutions, and improved installation technologies help to extend the life of the cathode block, reduce energy losses, and improve the quality of the produced metal. The implementation of such innovations has a direct impact on the economic efficiency and environmental safety of the entire production cycle.

The properties of sidewall blocks should be different from those of feed blocks, since they are not intended for current transfer. Currently, the main way to improve the production technology of carbon cathode blocks is to increase the proportion of artificial graphite in the formulation and add a high-temperature treatment of the blocks using electricity [186]. This provides a decrease in the specific electrical resistance and increases the thermal conductivity and thermal stability. The study of the influence of the component composition on the properties of carbonaceous blocks based on traditional materials shows that reducing the graphite content to 15% allows for the optimization of their properties, i.e., increasing the specific electrical resistance, which means excluding current losses through the walls of the electrolyzer and increasing the mechanical strength and reducing the thermal conductivity; all this compensates for the need for additional thermally insulated vermiculite plates [187,188].

The development and subsequent implementation of scientific and methodological approaches to optimize the start-up period of the electrolyzer is one of the most promising areas directly affecting the service life of aluminum electrolyzers. Thus, in [14], a simplified heat transfer model is proposed for selecting the optimal electrolyzer voltage during the start-up period of operation, when the casing is not yet formed as an additional thermal insulation layer.

6. Discussion

MHDS in the primary production process of aluminum is a critical factor [67,109,135] affecting the efficiency [172] and safety [189] of the electrolysis process. During the analysis of existing works, it was found that the challenge of MHDS in such processes remains poorly understood.

On the one hand, the stabilization of the electrolysis process with a consideration of MHDS can significantly increase productivity [149,172], reduce energy costs [183,186], and minimize environmental risks [189,190]. On the other hand, approaches to solving this challenge require a deep interdisciplinary integration of the knowledge from plasma physics, mathematical modeling, engineering, and materials science.

Figure 4 presents a diagram showing the relationship of the keywords of scientific publications on the research topic, which was compiled as a result of the methodological analysis of the literature sources. This diagram shows the relationship of the challenge of aluminum electrolysis with other challenges of a physicochemical nature.

The analysis of existing studies on the challenge of the MHDS of the primary production process of aluminum has revealed a number of significant regularities. Most modern works emphasize the role of electromagnetic perturbations in electrolysis baths [67,94], which lead to a disturbance of the equilibrium and a reduction in process efficiency [96,97]. However, despite the abundance of data on theoretical models of MHD instabilities [59,66,98] of various configurations [100,101] there is still no unified methodology for their identification [138,139,144] integrated with industrial control systems [135,141,143].

Figure 5 shows the world dynamics of publications in the field of aluminum electrolysis over the last 10 years. Here, we can see that the number of papers related to aluminum electrolysis is increasing throughout the period considered, but there are slight drops in 2016 and 2021. The years 2023–2024 show the highest interest in this topic. This phenomenon can be attributed to an increase in the global demand for aluminum.

The comparison of different methods of the control and diagnostics of MHD instabilities (analysis of the spectrum of voltage fluctuations, numerical modeling of current distribution, and application of machine learning systems) has shown their ambiguous efficiency. Thus, numerical methods [66,139,191] allow for the achievement of a high prediction accuracy but require significant computing power and are not always applicable in real time. At the same time, methods based on machine learning [88,118,122] demonstrate the potential for self-learning and adaptation but require the creation of extensive training samples [153,154], which limits their application to non-standard production plants.

A significant gap in the study of the challenge is the insufficient integration of MHD instability identification methods with automated electrolyzer control systems [118,120,122]. To date, most solutions are focused on ex post facto data analysis [4,126,154], whereas predictive models capable of the real-time signaling of the development of critical perturbations are needed to ensure a good process stability [59,125,128].

In addition, the influence of the design features of electrolyzers on the level of MHD instabilities has not been sufficiently studied. The works [9,121] consider some aspects of changing the electrode geometry, but there is no comprehensive analysis of the relationship between the bath design and the dynamics of electromagnetic disturbances [43,113,140].

The results of this analysis emphasize the need to develop multi-level systems for the identification of MHD instabilities [115,158], including combined approaches [172]: the integration of numerical modeling with machine learning [66,88,139] and sensing methods [153,154,191]. A promising area of research is the development of intelligent predictive control systems [59,125,128] using hybrid data analysis algorithms [74,96] for the early detection of critical states [107,182].

Additionally, a detailed study of the influence of electrode modification [81,83,97] and electrolysis bath designs [44,181,192] on the level of MHD instability [43,66,67] and related negative effects [113,140] is required. In this context, the application of multicriteria optimization methods can contribute to the formation of more stable [113,121] and energy-efficient technological processes of primary aluminum production [45,130].

One of the significant aspects identified in the methodological analysis is the need to fine-tune the magnetic field during the electrolysis process, which requires the use of complex multivariate models [83,133,134]. The application of numerical methods and high-speed computations combined with experimental data can serve as a key to creating more accurate predictive models for MHDS in aluminum production. However, despite the promise of such methods, there are a number of challenges associated with high computational costs and possible errors in modeling real-world conditions.

In modern conditions, the issue of the optimization of control systems for maintaining MHDS in different phases of the technological process remains unsolved [74,107,134]. The solution to this challenge requires the development of innovative control algorithms and technologies that allow for the implementation of theoretical achievements in practice.

In the course of the literature analysis, the key challenge of aluminum production was identified—metal agitation [5,59]; the nature of its behavior is not properly studied, but it is shown that the redistribution of currents in the electrolytic solution has a direct influence on this phenomenon. The identification of the electromagnetic field within the electrolysis bath will allow us to draw conclusions both about the course of the entire technological process and about metal undulation, since it arises as a result of the current redistribution in the electrolytic solution [105,157]. It is also worth noting that MHDS also depends on the aforementioned redistribution of currents, which can generate a distortion of the magnetic field [66,67].

At the enterprises of the metallurgical industry [3,10,101], regardless of the type of electrolysis plant for aluminum production [141,148], complex means for the production process control are used [151,164]. Nowadays, mathematical modeling methods [5,137,167] are widely used in production [48,170] for the control of process parameters [14,180]. An electrolyzer is a complex technical device, which is also based on a number of complex elements—in the process of the operation of which, for various reasons, losses of electrical energy may be allowed. The identification of such losses can often be realized only by the effects they have on the technological process.

Table 2. The generalized study scheme.

Challenge	Objectives	Methods	Planned or Achieved Results
Lack of possibility to accurately measure dynamic parameters in the bath during electrolysis	Develop or adapt sensors and systems to measure the speed and direction of molten aluminum and electrolyte circulation. Identify the most informative areas for sensor placement. Establish a methodology for calibration and verification of measuring instruments.	Use of fiber optic sensors, acoustic, or ultrasonic measurement methods. Laser Doppler anemometry for measuring flow velocities (in laboratory testing). Numerical simulation of flows to refine sensor locations (CFD packages, specialized MHD modules).	Obtaining accurate data on velocity distribution and turbulence of electrolyte flow. Improving the reliability of calculations and models in estimating MHD parameters. Development of recommendations for modernization of measuring equipment on an industrial scale.
Increase in the level of environmental pollution	Identify sources and mechanisms of emissions of harmful gases and dust particles. Develop measures to reduce emissions (e.g., optimization of current modes, use of additional filtering systems). Develop a set of measures for waste recycling.	Laboratory analysis of emission samples. Mathematical modeling of pollutant mass balance under different electrolysis modes. Environmental audit of the enterprise and development of programs to reduce harmful emissions.	Reduction in the overall level of air and waste pollution when the conditions of process mode optimization are met. Development of methodological recommendations for the operation of gas cleaning systems. Possibility to reduce the impact of negative factors of the MHD process on the environment.
Wear and destruction of metal structures of the cathode device	Investigate mechanisms of corrosion and erosion of cathode metal structures during high-temperature interaction with melts. Develop protective coatings or formulations that extend the life of cathodic structures. Optimize the shape and material of the cathode block to reduce areas of concentrated current loads.	Metallographic analysis and high-temperature oxidation tests. Numerical modeling of temperature distribution and current lines (electrodynamic analysis). Utilizing experimental testbeds to test new materials.	Increasing the life of cathode structures by using more stable materials and optimizing the shape. Reduced repair and maintenance costs. Increasing the stability of electrolyzer operation due to the prevention of emergency situations.
Electricity losses due to interaction of electric and magnetic fields	Determine in which parts of the plant the electric and magnetic fields are most intense and lead to maximum losses. Investigate ways to shield or redirect magnetic fluxes. Develop proposals to reduce inefficient circulation of currents inside metal parts.	3D modeling of fields (electromagnetic analysis, e.g., using COMSOL 5.6 or ANSYS 2024 R2 packages). Experimental measurement of induction losses and eddy currents (in some cases using Foucault current sensors). Optimization of power contacts and shunt circuits and finite element method (FEM) calculations.	Reduction in total energy losses and, as a consequence, a reduction in the cost of aluminum smelting. Development of recommendations on constructive protection of electrolyzer elements from excessive magnetic fields. Improved plant efficiency and process stability.
Disturbance of MHDS of the electrolyzer	Identify the main causes of perturbations (turbulent flows, inhomogeneity of electromagnetic fields, etc.). Develop means of active and passive control of MHD parameters (geometric and electrical changes). Create tools for operational control of the MHD state (digital twins, real-time diagnostics systems).	Multidimensional numerical modeling (MHD modules for simultaneously solving the equations of magnetic and hydrodynamic fields). Laboratory and semi-industrial experiments on smaller-scale bath models. Optimal control methods (current load control, feedback-based field tuning).	Stabilization of electrolysis processes due to a more uniform distribution of electromagnetic forces. Reducing the occurrence of waves at the metal–electrolyte interface, reducing the risk of “breakouts” and flooding. Improved product quality (due to uniformity of process conditions).
Lining wear under electrolysis conditions	Analyze thermal and mechanical stresses on lining materials. Investigate new materials (refractory mixtures, ceramics, composites) with improved resistance. Develop recommendations on temperature regime and methods of lining installation.	Laboratory tests for thermal shock and analysis of structural changes in material under cyclic loads. Thermographic studies to assess temperature fields (including industrial-scale infrared camera applications). Numerical modeling of heat distribution and mechanical stresses (FEM, thermodynamic analysis).	Increased liner life by optimizing thermal conditions and selecting more resistant materials. Reduced downtime associated with the replacement of liner blocks. Overall increase in electrolyzer efficiency and cost reduction.
Damage to anodes during slight circulation of electrolyte near the back of the unit	To study the peculiarities of electrolyte circulation and the mechanism of influence on anodes in the back wall zone. Develop recommendations for modifying the anode design or the design of the bath itself to improve flow. Develop regulations on the frequency and methods of diagnostic control of anode condition.	Simulation modeling of local electrolyte flows. Methods of non-destructive testing (ultrasonic diagnostics, X-ray fluorescence analysis). Experimental studies on industrial plants with “transparent” models.	Reduced anode failures and deformation due to more uniform circulation. Reduced costs for anode replacement and repairs to the back of the electrolyzer. Improvement of technological stability and uniformity of smelted metal quality.
Insufficient study of the nature of the mutual influence of electromagnetic and thermal fields during electrolysis	Develop a combined model (electromagnetic + thermal field) for a complex description of processes. Conduct a series of computational experiments at different current parameters, geometry, and electrolyte composition. Verify the adequacy of the model on the basis of experimental data.	Multiphysics numerical modeling (ANSYS, COMSOL, or similar packages). Experiments with varying heating regimes and current densities to study the dynamics of temperature distribution. Comparison of calculated results with thermographic and electromagnetic field measurements.	A deep understanding of the mutual correlations between thermal and electromagnetic effects. Ability to predict areas of overheating and increased electromagnetic activity. Creation of methodical recommendations on optimal control of thermal and electromagnetic modes.
Uneven heating of metal in the working area due to an uneven electromagnetic field	Determine the degree of influence of electromagnetic field distribution on local heating of metal. Develop methods of field equalization (geometric optimization of the bath, additional shielding, adjustment of current parameters). Implement real-time metal and electrolyte temperature monitoring systems.	Thermal imaging combined with electromagnetic field measurement. Use of corrective shunts, shielding devices, and current distribution control systems. CFD and MHD modeling to predict the impact of design and electrical circuit changes.	More uniform temperature distribution in the working zone and improved metal quality. Reduction in localized overheating that reduces equipment life. Improving the overall energy efficiency of electrolysis by minimizing losses.

presents a generalized scheme of this study, covering a comprehensive methodology for solving the challenges associated with MHDS in the production of primary aluminum. For each highlighted challenge specific tasks are formulated, research methods are outlined, and expected results are indicated. This approach will allow us to systematically plan the work, provide continuity between the stages (from the literature analysis and theoretical modeling to field experiments and the implementation of the results), and achieve the goals set to improve the efficiency and reliability of electrolyzers.

The proposed generalized research scheme covering the main challenges, tasks, methods, and expected results within the framework of the methodological analysis and search for technical solutions to identify MHDS in primary aluminum production forms a holistic and systematic approach to solving technological and environmental issues. The detailed consideration of each challenge [193] (from the lack of accurate methods for measuring dynamic parameters in the bath [194,195] to the increase in the level of environmental pollution and the wear of structural elements [15,196]) allows for us to build a logic of research, starting from setting goals and identifying the key factors of influence and ending with the implementation of practical recommendations and the evaluation of their effectiveness [6]. Thus, the implementation of this scheme contributes to the improvement of the reliability, productivity, and environmental safety of electrolyzers, which not only improves the economic performance but also ensures the sustainable development of production in accordance with modern standards of quality and environmental responsibility.

To determine the countries that have made the greatest contribution to the research of the challenge of aluminum production, a graph of the most active regions was compiled (Figure 6). In the first place is China (757 papers), then Russia (342 papers), and in the third place is the USA (307 papers). With 93 papers, Canada is in the fourth place, and India completes the top five with 86 papers.

During the preparation of this review, only those publications available in open databases in Russian and English for the last 10 years were analyzed. Consequently, the results may not fully reflect new patent developments or studies published in highly specialized journals. In the future it is planned to expand the search by using additional databases and to analyze the practical experience of the implementation of MHDS identification technologies in large aluminum companies. Works for the last 10 years were considered because of the appearance in this period of more modern methods of the identification of unsteady distributed electromagnetic fields and MHD instability, as well as the introduction of artificial intelligence.

7. Conclusions

Within the framework of the performed methodological analysis, a comprehensive assessment of existing approaches to solving the challenge of MHD instability in the process of primary aluminum production was carried out. Despite significant progress in MHDS research in other areas, aluminum production technologies require further research and the development of specialized solutions for the stabilization of electrolysis.

The application of mathematical models, computational methods, and new experimental techniques can become the basis for the creation of effective tools for monitoring and managing MHDS. However, the full implementation of these solutions requires the integration of different scientific disciplines, as well as further research to optimize existing models and algorithms.

Special attention should be paid to the development of adaptive control systems, which are able to consider the dynamic change in the process state and promptly correct the parameters of the magnetic field and other technological factors. This will not only stabilize the process but also improve its economic and environmental efficiency, reducing the energy consumption and carbon footprint.

Currently, there is a tendency to intensify aluminum production by increasing the current of the electrolysis plants. The prerequisite for this is the growing demand for primary aluminum. For the development of the industry, it is necessary to use the potential of aluminum consumption growth, as this metal is used in the manufacturing of a wide range of products in many spheres of human activity, from goods that meet the household needs of the population to products used in aviation.

Modern primary aluminum production is influenced by several global trends that shape the industry’s development strategy and technology requirements:

1.

Improving energy efficiency:

• Manufacturers are striving to reduce energy costs by improving equipment and utilizing energy-saving technologies.
• New designs for electrolyzers with improved thermal insulation and optimized process modes are being introduced.

2.

Greening of production:

• Environmental regulations are becoming more stringent, so companies are looking for ways to minimize greenhouse gas emissions, toxic fluoride compounds, and other harmful substances.
• Priority is given to switching to cleaner energy sources (hydroelectric, solar, and wind power) in order to reduce aluminum’s carbon footprint.

3.

Digital integration and automation:

• The process is automated thanks to monitoring and control systems based on Big Data, machine learning, and the Internet of Things (IoT).
• Predictive algorithms are used to monitor electrolyzers, predict failures, and optimize the process in real time.

4.

Development of MHD stable technologies:

• More and more attention is being paid to the challenge of MHDS, as any perturbations have a negative impact on the energy efficiency and aluminum quality.
• Comprehensive systems for the identification and control of MHD-unsteadiness, including numerical modeling and machine learning algorithms, are developed.

5.

Consolidation of production facilities:

• To reduce unit costs and improve competitiveness, aluminum companies often go down the path of consolidating and scaling up production.
• Large enterprises implement innovative solutions faster than small ones because they have the resources for research and modernization.

6.

Application diversification and demand:

• Aluminum is in demand in the automotive, aerospace, electronics, and construction industries, which is driving the development of better and cleaner alloys.
• The growing demand for lightweight and strong materials (e.g., for electric vehicles) is pushing for further improvements in technologies that reduce costs and improve metal properties.

7.

Strengthening scientific and technical collaboration:

• Companies, research institutes, and universities are increasingly joining consortia to develop new electrolysis methods and analyze data.
• Best practices are being shared internationally, and large companies are investing in startups engaged in the digital transformation of the aluminum industry.

The combination of these trends determines the vector of the development of primary aluminum production—the industry is becoming more technological, environmentally friendly, and flexibility- and quality-oriented, as well as actively using digital tools for sustainable competition.

The modern research on aluminum electrolysis has already accumulated an extensive knowledge base on the thermodynamics and kinetics of the process, including the mechanisms of aluminum oxide reduction, thermal balances, and the peculiarities of melt interactions with electrode materials. The chemical composition of the electrolyte and various additives affecting the melting and electrical conductivity of the melt, as well as a wide range of impurities in cryolite (NaF, AlF3, CaF2, etc.), have been studied in no less detail. Considerable attention has been paid to electromagnetic and hydrodynamic perturbations, since MHDS directly affects the efficiency and final quality of the melted metal. Based on these studies, both theoretical models for describing MHD instabilities and numerical methods for calculating the current distribution in the bath have been developed. In addition, materials and designs for anodes and cathode blocks are being actively improved to reduce their wear and increase the stability of electrolysis. Such an integrated approach to traditional aspects—from thermodynamics to equipment improvement—is already yielding significant economic and technological benefits.

At the same time, there are areas that have received relatively less attention so far. Among the least studied are the methods of the integration of digital technologies in real-time modes, while the “digitalization” of enterprises and control systems has become a defining trend in recent years. So far, there are still no unified universal methods for adapting MHD instability identification solutions to a wide variety of industrial operating conditions, and no uniform criteria for testing such systems at large-scale production facilities have been developed. In addition, the design features of electrolyzers and their influence on the MHD system parameters are not sufficiently considered, especially in the case of non-standard or small-format plants. Environmental aspects related to the full life cycle and carbon footprint assessment are often left “outside” the main technological studies. A similar situation can be observed in comparative studies on the application of artificial intelligence: most works search for solutions to specific challenges without analyzing different types of algorithms from the standpoint of universality and the ease of scalability. Finally, medium-sized and small-scale enterprises are rarely the focus of research, because traditionally technological solutions are developed for large industries, and many aspects of the economic profitability and complexity of implementation in more modest conditions are insufficiently covered.

In the future, new technologies can be expected to significantly improve the stability and efficiency of the primary aluminum production process. An important trend is also the use of artificial intelligence methods to optimize control and predict unstable situations in real time, which will open new horizons for the automation and improved safety of the production process.

Based on the literature analysis conducted, the following conclusions were drawn:

• Simulation has become a major tool for designing and optimizing the performance of aluminum electrolyzers and remains a priority area of research.
• To ensure a good service life of the electrolysis bath, it is necessary to select a rational composition of lining material or to take measures to improve the efficiency of the existing ones.
• Primary aluminum production today is characterized by a high energy demand and greenhouse gas emissions, where electrolysis is one of the most energy-intensive processes.
• More energy-efficient technologies and practices have been developed over the years, and historically inefficient or environmentally unfavorable equipment is being modernized as the effects of emissions are better understood.
• There is an assumption that the most optimal electrolyzer will be one with a large number of anodes in the form of pins with the same potentials and with a minimum distance between them.
• The benefit of future developments is that they will consider and connect hydrodynamic, electromagnetic, thermal, and electrochemical processes, along with changes in the bath shape caused by heat and mass transfer.

Vertical electrolyzers are a promising technology that is available for modernization and increased capacity. Primary aluminum production will become increasingly important as this metal is used everywhere and the demand for it is only growing.

The comparative analysis of publications has shown that the key physical obstacle to further reductions in the specific energy consumption and the improvement of energy stability remains the MHD wave instability of the «metal–electrolyte» interface. This instability occurs at elevated currents. Since this work was performed to define the tasks of future research, in the future it is planned to establish the presence or absence of a relationship between the occurrence of MHD instability and the geometric shape of electrolysis baths, as well as the relationship between MHD instability and the specific current density. This will make it possible in the future to determine possible ways to increase the MHD, which, in turn, will have a positive effect on the reduction in energy costs without the deep modernization of equipment. Thus, the degree of MHD controllability should be considered the main physical criterion of the energy stability of aluminum electrolyzers and a priority area of future technological developments—for example, an APCS capable of correcting the MHD in real time.

Thus, the challenge of MHDS remains relevant and requires a comprehensive approach that combines theoretical developments with practical solutions aimed at improving the productivity and stability of the primary aluminum production process. It is expected that further research in this area will lead to the creation of more efficient and sustainable control methods, contributing to the improvement of aluminum production technologies and improving their competitiveness in the global market.