Smelting-Aluminothermic Reduction of Hydrogen Pre-Reduced Manganese Ores in a 200 kW DC Arc Furnace

1. Introduction

Manganese is a critical element in the steel industry, primarily used as a deoxidizer, desulfurizer, and alloying agent to enhance steel’s strength and abrasion resistance [1,2]. The global demand for Mn alloys is substantial, with millions of tonnes of high-carbon ferromanganese and silicomanganese produced annually [3]. Traditionally, ferromanganese alloys are produced through carbothermic reduction of manganese ores in submerged arc furnaces[4,5,6]. This conventional method is highly energy-intensive, consuming 2000–3500 kWh per tonne of alloy, and is a significant contributor to greenhouse gas emissions, typically releasing 1–1.4 tonnes of CO₂ per tonne of metal produced due to the reliance on metallurgical coke as both a reductant and energy source [4,5,6,7].

The urgent need to align with global climate targets, such as the Paris Agreement’s goal of decreasing CO₂ emissions by 45% by 2030 and achieving net-zero by 2050, necessitates the development of innovative and sustainable manganese production processes [6].

Current pyrometallurgical processes face challenges related to high energy consumption, substantial carbon emissions, and the generation of large volumes of by-products like slag [7,8]. These challenges highlight a critical research gap: the need for cleaner, more efficient, and environmentally benign technologies for manganese extraction.

Smelting-aluminothermic reduction presents a promising alternative to carbothermic processes. This method utilizes aluminum as a reducing agent, which is a highly exothermic reaction, potentially reducing external energy input and offering a pathway for valorizing aluminum-containing industrial wastes like Al dross [3]. Furthermore, the integration of hydrogen (H₂) pre-reduction can significantly improve the overall sustainability of the process. Hydrogen pre-reduction has been shown to accelerate the decomposition of carbonates and facilitate the metallization of manganese oxides, offering a low-emission pathway by reducing reliance on carbon-based reductants and lessening atmospheric gaseous carbon release [2,6,9]. Studies have investigated the effects of various factors, including temperature, gas composition, and ore characteristics, on the H₂ reduction process, noting that H₂ addition can increase the rate of reduction [6].

The use of direct current arc furnaces in metallurgical applications has gained traction due to several advantages over traditional alternating current furnaces, including better power efficiency, reduced electrode consumption, lower noise levels, and improved heat distribution [10]. These characteristics make DC arc furnaces particularly suitable for processes aiming for high energy efficiency and operational stability.

This study aims to provide a comprehensive comparison of the smelting-aluminothermic reduction of hydrogen pre-reduced manganese ores in a 200 kW DC arc furnace. Specifically, we investigate the comparative performance of two different H₂ pre-reduction approaches (packed-bed vertical retort and plasma rotary furnace) in conjunction with smelting-aluminothermic reduction. The assessment integrates a wide array of technical, environmental, and operational metrics, including energy and resource efficiencies, environmental burdens (e.g. GHG intensity, waste generation), eco-efficiency ratios, and operational sustainability indicators. By doing so, this research seeks to demonstrate the potential of this integrated process for achieving resilient, sustainable, and optimized manganese recovery with minimal environmental impact. The findings contribute to the ongoing efforts towards decarbonizing the metallurgical industry and advancing circular economy principles in metal production.This study is further seeks to demonstrate the HAlMan process on higher technology readiness level (TRL 6) level. For the background and description of the HAlMan process, the reader is referred else where [11,12]

2. Materials and Methods

2.1. Materials

Two manganese ores were used in this study, namely, United Manganese of Kalahari (UMK) ore and Nchwaneng ore. Both ores originates from the Kalahari Manganese Field (KMF), located in the Northern Cape Province of South Africa. The two ores differ in mineralogy and grades, and are mined by different companies (e.g. UMK ore is mined by United Manganese of Kalahari (UMK) (Pty) Ltd, whereas Nchwaning ore is mined by Hotazel Manganese Mines (HMM)). The Nchwaning ore employed in the was procured through Assmang (Pty) Ltd., Cato Ridge, South Africa. UMK ore was supplied by Transalloys (Pty) Ltd, South Africa.The as-received lumpy ores were subjected to comminution and granulometric classification into the –20 mm + 6 mm fraction to conform with the stringent operational specifications of Mintek’s feed system.

Two distinct pre-reduction approaches were deployed for the hydrogen-mediated conditioning of these ore fragments at elevated temperatures (700–900 °C): (i) a vertical retort packed-bed furnace (Figure 1) and (ii) rotary plasma furnace (sample pre-reduced by SINTEF).

Complementary raw materials incorporated into this study includes flux (burnt lime) which was supplied by local supplier and reductant (metallic aluminium), procured from Alitrop (Pty) Ltd. Burnt lime was also subjected to crushing and granulometric refinement (–20 mm + 6 mm) to ensure feed-system compatibility, while the aluminium required only screening to the prescribed size interval for operational integration. The images showing the appearance of pre-reduced Nchwaning ore, pre-reduced UMK ore, burnt lime and aluminium are presented in Figure 2.

2.1.1. Chemical Analyses

UMK pre-reduced ore, Nchwaning pre-reduced ore (both from Mintek and from SINTEF) and lime were subjected to chemical analyses which employed Inductively Coupled Plasma – Optical Emission Spectrometry (ICP-OES). The Mintek internal reference material, SARM16, was used. These methods The obtained results are summarized in Table 1 and Table 2. Species with concentrations of parts per million (as well as loss on ignition) are not reported in Table 1,

and that thus the percentages will not add to 100 percent. For mineral-assay reconciliations, the inductive coupled plasma optical emission spectroscopy (ICP-OES) results were used to validate the mineral abundance data obtained from X-ray diffraction (X-Ray Diffraction).

2.1.2. Mineralogical Analyses

For XRD mineralogical analyses, pre-reduced manganese ores we first pulverised samples and then micronized to achieve particle size below $20$m to ensure the acquisition of more representative data for minerals and elements quantification. The samples were subjected to the Bruker D8 diffractometer equipped with Fe-filtered Co K$\alpha$ radiation, and ran over a 2 theta range of 5-80 degrees with a step size of 0.02 degrees 2 theta. Minerals were initially identified using Bruker EVA^® software, while the quantification of minerals was done using TOPAS^® software. Since pre-reduction was done batch-wise, XRD results of the successive tests were used to inform on optimizing the operating conditions (e.g. temperature, holding time, gas-to-solid ratio and so forth) for the next batch.

2.2. Smelting Operations

The 200 DC arc furnace testing facility housed at the Bay 1 area of Mintek was used in this study. A schematic representation of the facility layout is presented in Figure 3, highlighting the major operational domains. For brevity, auxiliary systems (e.g., the cooling-water circuit, gas pipelines and so forth) are not depicted in the diagram. The facility extends over two floors, and consists of three primary operational domains, namely:

• Feed section: this domain is endowed with raw material hoppers, conveyor systems, and a mixing bin for charge preparation.
• Furnace section: this zone incorporates a hydraulic electrode-control system, a robust furnace shell lined with refractories, and integrated cooling-water circuits.
• Off-gas handling section:this section is furnished with an off-gas port, dedicated extraction lines, and a baghouse system for effective dust capture and emissions control.

The feed and furnace sections are equipped with instrumentation network, enabling real-time, remote monitoring and precise operational control. Conversely, the off-gas realm is not endowed with instrumentation, compelling persistent on-site vigilance and physical inspections to ensure reliability and continuity of operation. Thermal management was ensured through a dedicated cooling-water circuit that continuously cycled water through both furnace and off-gas ducting components. The facility is equipped with gas pipeline infrastructure, ensuring reliable supply of essential utilities such as compressed air and oxygen.

2.2.1. Experimental Framework

Part of the experimental framework of HAlMan Campaign 1 (Table 3) was systematically configured to investigate four critical process and operational conditions during the smelting of Nchwaning samples pre-reduced via packed-bed vertical retort furnace (Mintek) and rotary plasma furnace (SINTEF), respectively. It is important to highlight that the first part of Campaign 1 was dedicated to smelting-aluminothermic reduction of high carbon ferromanganese slag (HCFeMn slag).

A pivotal transition from HCFeMn slag to Nchwaneng ore pre-reduced using packed-bed vertical retort furnace induced reduced slag volumes and elevated bath temperatures, necessitating specific energy requirement (SER) modulation (0.95 to 0.85, in taps 24 through 26) to re-anchor the system at 1550 °C. For the Nchwaning ore pre-reduced using the rotary plasma furnace, MnO suppression of MnO in slag (<4 mass%) was again realized by increasing Al additions (26.0 to 28.6 kg, in taps 27–29). The final condition (taps 30–36) explored the flux–reductant synergy, wherein lime addition was increased from 26.7 to 30.0 kg and Al additions were increased from 28.6 to 32.0 kg. Overall, these trials provided information towards of Nchwaning ore pre-reduced to employing two different approaches, namely, packed-bed vertical retort furnace and rotary plasma furnace.

The experimental framework of HAlMan Campaign 2 (Table 4) is comprised of nine conditions designed to investigate the furnace stability, compositional control, ore transitions and power scaling. Condition 1 (taps 1–5) sought to stabilize the furnace with regard to energy and mass, yet arc instability at 100 V necessitated reduction to 70 V, alongside increases in SER (0.85 to 0.95). In pursuit of alloy and slag compositions targets, both the reductant (Al) and the flux (lime) were increased, namely aluminium (9 to 30 kg), and lime (20 to 28 kg). With targets unmet, Condition 2 (taps 6–8) further intensified reagent inputs (Al: 30 to 34 kg; CaO: 28 to 32 kg) and raised SER (0.95 to 1.0). Condition 3 (taps 9–11) explored lime fineness under 90 V, while Condition 4 (taps 12–15) saw the increase in power from 130 kW to 150 kW at 100 V.

Condition 5 (taps 16–18) marked the transition from pre-reduced UMK ore to pre-reduced Nchwaning ore, with slag withheld at tap 16 and voltage reduced to 90 V. Analogous to earlier UMK ore, Condition 6 (taps 19–21) increased SER, Al, and lime modestly (to 1.05, 36 kg, 34 kg, respectively), while Condition 7 (taps 22–24) advanced these inputs further (SER 1.15, Al 40 kg, CaO 38 kg). Condition 8 (taps 25–27) revisited high-power operation (150 kW, 100 V), which was carried forward as Condition 9 (taps 28–30).

Together, these conditions established a systematic progression of operational perturbations, elucidating the coupled roles of voltage, SER, fluxing additions, and ore transition in stabilizing furnace operations and aligning slag–alloy compositions. Campaigns 1 and 2 generated data towards that integrates technical, environmental and operational metrics of smelting-aluminothermic reduction. In this study, a detailed energy and material balances, quantification of environmental burdens, and indicators of process sustainability are covered, thereby extending beyond conventional eco-efficiency analyses.

3. Results

3.1. Effect of Pre-Reduced Ores Properties

3.1.1. Energy

A. Power Stability Metrics: Arc Stability, Voltage/Current Fluctuations

The stability of the electric arc, along with the degree of voltage and current fluctuations, are critical operational indicators in electric arc furnaces. These metrics directly influence energy efficiency, electrode and refractory life, and overall process performance [13,14].

Figure 4 shows the comparison of the set-points (current, voltage and power) against experimental observed values for the (a) UMK ore pre-reduced in a packed-bed retort furnace, and (b) Nchwaning ore pre-reduced in a packed-bed retort furnace. The voltage instabilities and voluntary voltage reduction directly affects the power stability and reduces delivered power. In many EAF operations, control systems aim to minimize variations in arc voltage and current to maximize power input and stabilize the process [15].

B. Specific Energy Requirement

The specific energy requirement (in kWh/kg), is a critical performance indicator in metallurgical processes. In this study, eq:SER was employed to calculate the specific energy requirement, SER, where the obtained results are plotted in Figure 5.

$$\mathrm{SER}={\displaystyle \frac{\mathrm{Total}\mathrm{Energy}\mathrm{In}-\mathrm{Total}\mathrm{Energy}\mathrm{Out}}{\mathrm{Total}\mathrm{Feed}\mathrm{In}}}$$

(1)

Changes in target SER setpoints, for instance, increases from 0.85 to 0.95, then to 1.05, and 1.15 (Conditions 1, 6, and 7, respectively), guide the operational modifications of other parameters like aluminum and lime additions and feed rate, all aimed at achieving the desired energy input per unit of processed material [16]. Increasing Al and lime additions (e.g., in Condition 1, 2, 6, 7) affects the chemical energy balance of the furnace.

The specific energy requirement of the process can also be estimated from the slope of a plot of average power against average feed rate (Figure 6), independently of heat losses. The y-intercept of this plot provides an estimation of the heat losses.

C. Furnace Thermal Efficiency

The energy losses to the environment are a critical factor in determining the thermal efficiency of pyrometallurgical equipment. The energy losses to the environment, used in the energy balance calculations, relate directly to the thermal efficiency of the equipment. Overall energetic efficiency in pyrometallurgical operations can be defined as the ratio of theoretical energy requirement to real energy output [8]. In this study, eq:thermalefficiency was used to define the thermal efficiency.

$$\mathrm{Thermal}\mathrm{Efficiency}={\displaystyle \frac{\mathrm{Total}\mathrm{Energy}\mathrm{In}-\mathrm{Total}\mathrm{Energy}\mathrm{Losses}\mathrm{Out}}{\mathrm{Total}\mathrm{Energy}\mathrm{In}}}\times 100$$

(2)

This method of calculating thermal efficiency usually provides a very good correlation for this facility. However, in this process, with the addition of energy from the smelting-aluminothermic reduction reaction, a low thermal efficiency is sometimes calculated, as shown in Figure 7, that gives the furnace power vs thermal efficiency.

The furnace furnace power vs thermal efficiency, calculated on a tap-by-tap basis, is presented in Figure 7. Understanding these energy balances is crucial for process improvement in electric arc furnaces and other high-temperature metallurgical processes [13,17].

C. Load Factor/Power-on Time

The operational efficiency of a smelting process is critically influenced by its load factor (${\eta }_{\mathrm{LF}}$), which represents the proportion of effective smelting time ($t_{\mathrm{smelt}}$) relative to the total operational cycle, including downtime ($t_{\mathrm{down}}$). A high load factor signifies optimized furnace utilization, crucial for economic viability and increased productivity in industrial applications. This section details the calculated load factors for the three experimental campaigns, providing insight into the operational performance of the respective pre-reduction routes and smelting.

The $t_{\mathrm{smelt}}$ is defined as the sum of all tap durations, and $t_{\mathrm{down}}$ is the sum of all downtimes, calculated using the following formulas:

$$t_{\mathrm{smelt}}=\sum \mathrm{Tap}\mathrm{duration}$$

(3)

$$t_{\mathrm{down}}=\sum \mathrm{Down}-\mathrm{time}$$

(4)

$${\eta }_{\mathrm{LF}}={\displaystyle \frac{t_{\mathrm{smelt}}}{t_{\mathrm{smelt}}+t_{\mathrm{down}}}}\times 100$$

(5)

The calculated operational metrics for each pre-reduced manganese ore are summarized in Table 5.

3.1.2. Materials

A. Reductant Efficiency

The efficiency of aluminum as a reductant is a pivotal metric in pyrometallurgical operations, particularly when leveraging secondary aluminum sources. It quantifies the effectiveness with which supplied aluminum is consumed for the intended reduction of target metal oxides, specifically manganese, iron, and silicon. Achieving high reductant efficiency is paramount for both economic viability and minimizing the environmental footprint of such processes [3,18]. As defined in Reductant efficiency1, ${\eta }_{\mathrm{Al}}$ represents the ratio of aluminum stoichiometrically consumed to produce Mn, Fe, and Si to the total aluminum fed into the furnace. Attaining high reductant efficiency is essential for both economic viability and minimizing the environmental footprint of such pyrometallurgical operations.

$${\eta }_{\mathrm{Al}}={\displaystyle \frac{\mathrm{Al}\mathrm{consumed}\mathrm{to}\mathrm{produce}\mathrm{Mn}+\mathrm{Al}\mathrm{consumed}\mathrm{to}\mathrm{produce}\mathrm{Fe}+\mathrm{Al}\mathrm{consumed}\mathrm{to}\mathrm{produce}\mathrm{Si}}{\mathrm{Total}\mathrm{Al}\mathrm{consumed}}}\times 100$$

(6)

The reductant efficiencies observed in this study (as shown in Figure 8), demonstrate variability depending on the properties of the respective pre-reduced ores. Nchwaning manganese ore pre-reduced in a retort packed bed furnace exhibited efficiencies between 30.89% and 55.32%. This can be attributed to the pre-reduction extent of the respective ores via respective pre-reduction approaches, whereas for UMK manganese ore, also pre-reduced in a retort packed bed furnace, the efficiencies spanned from 19.46% to 56.67%.

B. Elemental Accountability

The overall mass accountability provides a measure of the closure of the elemental balance across the smelting system. It compares the total elemental input (from feeds such as dross, lime, and pre-reduced ores) with the total output (in metal, slag, and dust streams).

In this study, the accountability of each element was determined to evaluate the degree of mass balance closure within the reduction process. It is defined as the ratio of the total mass of each element exiting the system through all product and by-product streams to the total mass entering the system through feed materials, as given in Equation 7.

$$\mathrm{Accountability}(\%)=\left({\displaystyle \frac{\sum m_{i,\mathrm{out}}}{\sum m_{i,\mathrm{in}}}}\right)\times 100$$

(7)

where $m_{i,\mathrm{out}}$ is the mass of element i in all output streams (metal, slag, and dust), and $m_{i,\mathrm{in}}$ is the corresponding mass of element i in all input materials.

Figure 9 presents the overall calculated elemental accountabilities for Al, Ca, Fe, Mg, Mn, and Si for Campaign 2 (Taps 1C2 to 30C2). The overall accountability values ranged from approximately 95.40% for Mn to 103.46% for Fe. Specifically, Al accountability was 97.93%, Fe was 103.46%, Mn was 95.40%, and Si was 96.34%. These values generally exceed 95%, indicating a satisfactory overall mass balance.

C. Elemental Recovery

Elemental recovery quantifies the extent to which target species were transferred from the feed materials into the metallic phase. It is a key indicator of the process efficiency and reflects both thermodynamic and kinetic factors governing reduction, phase separation, and alloy formation. The recovery of each element was calculated using eq:recovery1.

$$\mathrm{Recovery}(\%)=\left({\displaystyle \frac{m_{i,\mathrm{metal}}}{\sum m_{i,\mathrm{in}}}}\right)\times 100$$

(8)

where $m_{i,\mathrm{metal}}$ is the mass of element i recovered in the metallic product, and $\sum m_{i,\mathrm{in}}$ is the total mass of that element charged to the system via all input materials.

The overall elemental recovery values into the metal phase are shown in Figure 10. In the metal stream, Fe exhibited a high recovery of 96.44%, and Mn recovery was 65.92%. Si recovery was 42.28%, while Al showed a minor recovery of 3.87%. The elements that were recovered in the slag stream are Al: 94.02%, Fe: 6.61%, Mn: 29.20%, Si: 53.85%, whereas those in the dust stream are Al: 0.04%, Fe: 0.40%, Mn: 0.28%, Si: 0.21%.

D. Elemental Deportment

Deportment describes the distribution of each element among the various product phases, viz:metal, slag, and dust, and provides insight into the selectivity and completeness of reduction reactions. It complements the recovery analysis by indicating how effectively elements were partitioned into their thermodynamically preferred phases [19,20]. The elemental deportment (%) to each phase was determined using eq:deportment1.

$$\mathrm{Deportment}i,j(\left({\displaystyle \frac{mi,j}{\sum m_{i,\mathrm{out}}}}\right)\times 100$$

(9)

where $m_{i,j}$ is the mass of element i in phase j (metal, slag, or dust), and $\sum m_{i,\mathrm{out}}$ is the total mass of that element in all output streams.

Figure 11 illustrates the elemental deportment behavior across the process. Fe and Mn predominantly reported to the metallic phase, with 93.22% of Fe and 69.09% of Mn deporting to metal. Si showed a significant deportment to metal at 43.88%. Al, however, was primarily concentrated in the slag (96.01%). Minor amounts of elements were found in the dust stream.

E. Slag-to-Metal Ratio

The slag-to-metal ratio (SMR), defined as the mass of slag produced per unit mass of metal tapped (kg slag / kg metal), is a key metric for assessing material efficiency, energy demand, and downstream waste management in ferroalloy production. A lower SMR generally indicates more efficient utilization of raw materials and energy, reduced slag handling requirements, and a lower environmental burden. The experimental slag and metal masses obtained in this study are summarized, and the calculated overall SMRs for each campaign are presented in Table 6, derived from the aggregated mass of slag and metal from all taps where both products were successfully obtained, providing a more representative average for each campaign. The overall SMRs ranged from 1.34 kg/kg to 1.87 kg/kg. The individual tap-wise SMRs are also depicted in Figure 12.

3.1.3. Process Sustainability

A. Refractory Consumption

Refractory lining degradation is a critical operational and economic factor in high-temperature metallurgical processes, including ferroalloy production. The wear rate of refractory materials directly influences furnace longevity, maintenance schedules, energy efficiency, and overall production costs [21]. In electric arc furnaces, refractory consumption is primarily driven by thermomechanical stress, chemical corrosion by molten slag and metal, and erosion [22]. This section quantifies the refractory consumption during the experimental campaigns and discusses its implications.

Refractory consumption, specifically the loss of magnesia from the furnace lining, was estimated by calculating the difference between the cumulative MgO leaving the system in the slag and the cumulative MgO introduced with the feed materials over each campaign. The results, expressed as kilograms of MgO consumed per kilogram of metal produced, are summarized in Table 7 and in Figure 13.

The refractory consumption rates varied between 0.1051 kg MgO/kg metal and 0.1828 kg MgO/kg metal across the two pre-reduced ores. Refractory consumption rates are highly dependent on specific furnace design, operating parameters (e.g., temperature, slag chemistry, specific energy input), and the type of refractory material used.

B. Electrode Consumption

Electrode consumption is a significant operational cost and an important environmental consideration in electric arc furnace metallurgical processes. The rate of electrode wear is influenced by various factors including electrical parameters (current, voltage), furnace atmosphere, slag chemistry, operating temperature, and the specific composition of the electrode material itself [22]. Minimizing electrode consumption is crucial for improving the overall economic efficiency and reducing the carbon footprint of ferroalloy production. This section quantifies the electrode consumption during the experimental campaigns and discusses its implications.

The total mass of electrode consumed during each campaign was measured and then normalized by the total mass of metal produced to derive a specific electrode consumption rate, expressed in kilograms of electrode per kilogram of metal. The results are summarized in Table 8.

The specific electrode consumption rates varied considerably across the three experimental campaigns, ranging from a notably low 0.0014 kg electrode/kg metal for the Nchwaning ore pre-reduced in a retort packed-bed furnace to 0.1105 kg electrode/kg metal for the UMK ore pre-reduced in a retort packed-bed furnace. The Nchwaning ore pre-reduced in a plasma rotary furnace showed an intermediate consumption rate of 0.0596 kg electrode/kg metal.

The exceptionally low consumption rate observed for the Nchwaning ore pre-reduced in a retort packed-bed furnace (0.0014 kg/kg) is striking and warrants further investigation into the specific operational conditions, slag chemistry, and/or metal tapping procedures during this campaign that might have contributed to such minimal electrode degradation. This value, if representative, signifies a major advantage for this particular processing route.

3.1.4. Environmental Indicators: CO₂-Equivalent Emissions per Ton of Product, Dust and Particulate Emissions, NO_x/SO_x Emissions

Environmental impacts were quantified using attributional life-cycle assessment principles in accordance with ISO 14040 and ISO 14044 standards [23]. The system boundary for this assessment included direct process emissions, as well as indirect emissions associated with electricity generation and reductant production. The functional unit for all impact assessments is defined as one tonne of final alloy product. This evaluation combined direct process emissions measured by an accredited environmental monitoring laboratory with indirect emissions from literature-based estimates.

A. Stack Emissions and Atmospheric Pollutants

Stack monitoring of the 200 kW DC arc furnace was performed by Rayten Engineering Solutions Ltd., an independent laboratory accredited to ISO/IEC 17025:2017 standards by SANAS (T0894) [24]. The measurements were conducted during the June 2025 Mintek Campaign 2, utilizing internationally accepted reference methods that comply with the National Environmental Management: Air Quality Act of 2004 (Act 39 of 2004), specifically US EPA Methods 1, 2, 3A, 4, 5, 6C, and 7E for the respective pollutants [24]. Sample ports on the stack were utilized for volumetric flow rate and isokinetic measurements [24].

Table 9. Average measured gaseous emissions from the 200 kW DC arc furnace.

Parameter	Units	Value	Uncertainty
Average Particulate Matter	mg/Nm3	3.5	±1.14 mg
Average Sulphur Dioxide (SO2)	mg/Nm3	0	±18.87 ppm (detection limit)
Average Oxides of Nitrogen (NOx)	mg/Nm3	3.49	±6.67 ppm
Average Carbon Dioxide (CO2)	mg/Nm3	775	N/A (ambient background)
Average Oxygen (in stack)	%v/v	20.8	±0.22%

Table 9. Average measured gaseous emissions from the 200 kW DC arc furnace.

Parameter	Units	Value	Uncertainty
Average Particulate Matter	mg/Nm3	3.5	±1.14 mg
Average Sulphur Dioxide (SO2)	mg/Nm3	0	±18.87 ppm (detection limit)
Average Oxides of Nitrogen (NOx)	mg/Nm3	3.49	±6.67 ppm
Average Carbon Dioxide (CO2)	mg/Nm3	775	N/A (ambient background)
Average Oxygen (in stack)	%v/v	20.8	±0.22%

3.2. Effect of Pre-Reduction Approach

3.2.1. Energy

3.2.1.1. A. Power Stability Metrics: Arc Stability, Voltage/Current Fluctuations

Figure 14 shows the comparison of the set-points (current, voltage and power) against experimental observed values for the (a) Nchwaning ore pre-reduced in a rotary plasma furnace, and (b) Nchwaning ore pre-reduced in a packed-bed retort furnace.

C. Load Factor/Power-on Time

The calculated operational metrics for each pre-reduced manganese ore are summarized in Table 10.

Both pre-reduction approaches demonstrated gave a load factor of 96.89 %. This indicates a remarkably efficient utilization of the experimental furnace across all tests, characterized by minimal inter-tap downtime.

3.2.2. Materials

A. Reductant Efficiency

The reductant efficiencies observed for the two approaches employed in this study are shown in Figure 15, and demonstrate variability depending on the specific pre-reduction approach. For Nchwaning manganese ore lumps pre-reduced in a plasma rotary furnace, efficiencies ranged from 28.98% to 75.34%. Nchwaning manganese ore pre-reduced in a retort packed bed furnace exhibited efficiencies between 30.89% and 55.32%. This can be attributed to the pre-reduction extent of the respective ores via respective pre-reduction approaches.

B. Slag-to-Metal Ratio

The experimental slag and metal masses obtained in this study (for the two approaches) are given in Table 11, alongside the Slag-to-metal ratios. The overall SMRs were 1.34 kg/kg and 1.79 kg/kg, for rotary plasm furnace and vertical retort furnace approaches, respectively. The individual tap-wise SMRs are also presented in Figure 16.

3.2.3. Process Sustainability

3.2.3.1. A. Refractory Consumption

The refractory consumption for the two approaches, expressed as kilograms of MgO consumed per kilogram of metal produced, are given in Table 12 and in Figure 17.

4. Discussions

4.1. Effect of Pre-Reduced Ores Properties

4.1.1. Energy

A. Power Stability Metrics: Arc Stability, Voltage/Current Fluctuations

Despite furnace electric arc stability , voltage and current being critical operational indicators, the deviations between target and experimental values are inherent to electric arc furnace operations and reflect the dynamic nature of the arc, the varying charge characteristics, and the control system’s response [25,26]. The observed fluctuations (in Figure 14) are direct indication of arc instability, which can also be influenced by factors such as arc movement and plasma jet length variations [27,28].

Critically, voltage instabilities and voluntary voltage reduction directly affect power stability and reduce delivered power [10,29]. In this experimental framework, initial arc instability at 100 V necessitated a reduction in target voltage to 70 V for some trials with UMK ore, indicating a necessary operational adjustment to maintain a stable process and avoid excessive fluctuations. Control systems in many EAF operations aim to minimize variations in arc voltage and current to maximize power input and stabilize the process [26,30].

Large and frequent deviations and transients are known to increase electrode erosion significantly [31]. This occurs through enhanced oxidation at hot spots during arc wandering, mechanical tip wear during electrode slipping and re-striking, and localized hot-spot evaporation or chemical attack of the electrode and refractory lining [31]. The impact on refractory life is also pronounced, as arc wandering and poor heat coupling increase local refractory heating, which can shorten the lining’s operational lifespan[31]. High voltage and current fluctuations are also associated with increased wear on furnace components and higher energy losses [29].

Arc stability in EAFs is not typically measured by a single universal efficiency percentage. Instead, it is often quantified using statistical indices of voltage and current signals, such as standard deviation, variance, or specific power factor ranges [32,33] . High voltage and current fluctuations indicate a less stable arc, leading to erratic heat transfer and higher energy losses [29].

Researchers have proposed various metrics and models, including the Arc Quality Index, to evaluate arc quality and stability [34]. In this study, power stability in a DC arc furnace is assessed using three complementary, quantitative metrics, namely, current-tracking error (percentage deviation between measured and set-point current), voltage deviation (percentage deviation between measured and set-point voltage), and realized power deviation (percentage deviation between measured and target active power).

These quantitative metrics can be defined by

$$\Delta I_{\%}={\displaystyle \frac{I_{\exp }-I_{\mathrm{set}}}{I_{\mathrm{set}}}}\times 100\%,$$

(10)

$$\Delta V_{\%}={\displaystyle \frac{V_{\exp }-V_{\mathrm{set}}}{V_{\mathrm{set}}}}\times 100\%,$$

(11)

$$\Delta P_{\%}={\displaystyle \frac{P_{\exp }-P_{\mathrm{set}}}{P_{\mathrm{set}}}}\times 100\%.$$

(12)

These metrics together quantify how closely the furnace follows operator setpoints and the degree of transient disturbances that can increase electrode wear and refractory attack [13]. The observed statistics, indicating a systematic shortfall of realized current and power relative to set-points (mean deviations approximately -10% to -12%) with moderate shot-to-shot variability (standard deviations 4% to 6%), can be attributed to factors such as arc instabilities, occasional electrode slipping/immersion transients that reduce effective current transfer, variations in bath impedance during feed events, and voltage reduction initiated by unstable arc conditions [35].

B. Specific Energy Requirement

The Specific energy requirement (SER) is influenced by a range of operational factors, including voltage and power adjustments [36], targeted SER setpoint changes, the addition of reductants and fluxes (such as lime, coke, and oxygen) [14,37], transitions in ore type [14,38], and variations in feed rate [16]. Therefore, operational adjustments, such as an initial arc instability at 100 V leading to a reduction to 70 V to stabilize power delivery, and subsequent increases in power (from 130 kW to 150 kW) and voltage (to 100 V), had a direct impact on the total energy input and, consequently, the calculated SER.

For high-carbon ferromanganese production in submerged arc furnaces, typical specific energy requirements range between 2.0 and 3.5 MWh per tonne of alloy [39]. The innovative HAlMan process, which leverages on smelting-aluminothermic reduction, has demonstrated the potential for substantially lower specific energy requirement values. Reported figures indicate values around 880 kWh/tonne for similar eco-friendly, low-carbon manganese ferroalloy production methods, representing a significant energy reduction compared to conventional processes [40]. This aligns with findings where pre-treated materials can lead to over 50% reduction in electricity demand for HCFeMn production [41]. The current SER values, ranging from 0.8 to 1.15 kWh/kg (equivalent to 800-1150 kWh/tonne), fall within a competitive range, particularly when considering the aluminothermic nature of the process [12,42,43].

C. Furnace Thermal Efficiency

The thermal efficiency for a small-scale furnace is expected to be significantly lower than that of a commercial or large-scale pilot plant facility. Since smelting-aluminothermic reduction reactions are exothermic, the heat losses appear greater than the electrical power input due to the additional energy released during the smelting-aluminothermic reduction reactions. The furnace thermal efficiency calculated on a tap-by-tap basis is shown in Figure 7.

As illustrated in Figure 7, the addition of energy from the smelting-aluminothermic reduction reaction can, at times, lead to a calculated low thermal efficiency. This phenomenon occurs because the significant energy released during the smelting-aluminothermic reduction reactions can cause the measured heat losses to exceed the apparent power input [3,44]. This highlights the complex energy dynamics within the furnace where internal exothermic processes contribute substantially to the overall thermal state, impacting the traditional interpretation of efficiency metrics.

D. Load Factor/Power-on Time

The observed high load factors (95.89 % and 96.89 % for Nchwaning and UMK ores respectively) aligns well with the inherent operational flexibility and process control advantages offered by DC arc furnace technology, which often translates to higher furnace availability and lower heat losses during operation compared to other furnace types. Such high operational continuity is highly advantageous in industrial settings, suggesting that these smelting routes, particularly when coupled with efficient material handling, can achieve superior productivity. Industrial electric arc furnaces, particularly in sectors such as steelmaking and ferroalloy production, target high utilization rates to maximize throughput and cost-effectiveness. For instance, the average capacity utilization factor for scrap-based EAFs in the EU has been reported around 83% [45].

For the UMK ore, it’s notable that one tap (Tap No. 1C2) registered 0.0 h tap duration, which might indicate a non-operational or aborted run. While this specific instance would contribute to downtime if it were a planned part of the cycle, its impact on the overall load factor, given the very low summed $t_{\mathrm{down}}$ values, suggests that the typical operational flow between taps was very efficient.

4.1.2. Materials

A. Reductant Efficiency

The high reductant efficiency (reported in par:Reductant efficiency) can be associated with factors such as the aluminum content in the charge and critical aspects of slag chemistry [3]. Optimized slag conditions, characterized by carefully controlled basicity, viscosity, and liquidus temperature, are essential for efficient metal-slag separation and high recovery rates of manganese [46,47]. Variations in these parameters can significantly impact the effectiveness of aluminum as a reductant, affecting overall efficiency and metal yield.

Furthermore, process parameters such as pre-reduction methods, furnace design and the inherent characteristics of the manganese ore (e.g., mineralogy, particle size) play a crucial role in determining reductant efficiency [4,5,48] . Excess aluminum additions, oxidation losses, and the efficiency of recovery in the final alloy also contribute to the overall reductant utilization [49]. Higher efficiency values indicate a substantial proportion of the added aluminum actively participates in the desired reduction reactions, leading to improved recovery of valuable metals and a reduction in energy consumption [49]. Continuous optimization of these integrated process elements is vital for advancing sustainable and economically viable manganese production [12,18].

B. Elemental Accountability

The observed deviations from 100% accountability can be attributable to a combination of experimental and analytical uncertainties. These can include significant measurement errors, material heterogeneity, and sampling errors during the characterization of inputs and outputs [50,51]. Incomplete recovery of fine particles, minor losses through volatilization, or unquantified dust streams can also contribute to these discrepancies [52]. Such challenges are common in high-temperature metallurgical processes, where precise measurement of flows and concentrations can be sparse [51], and the incorporation of an uncertainty model is crucial for a realistic evaluation of mass balance reliability [52,53]. For instance, uncertainties in raw material composition and weighing, as well as element distribution factors, are known to influence mass balance closure in industrial operations [54].

The accountability results are consistent with literature observations for similar pyrometallurgical reduction systems, where acceptable mass balance closures typically fall within ±10% [3]. Industrial smelting operations often report elemental accountabilities between 90% and 110% due to inherent process heterogeneity, complex phase interactions, and practical sampling constraints [7]. Therefore, the values obtained in this study confirm the reliability of the mass balance and validate the analytical methodology used for subsequent recovery and deportment analyses.

C. Elemental Recovery

The high recoveries for Mn and Fe suggest favorable thermodynamic conditions for their reduction and subsequent capture in the metallic phase. The relatively lower Si recovery indicates that a significant portion remained in the slag phase, which is expected due to its higher stability in oxide form under certain conditions. The presence of Al recovery, even if minor, suggests its co-reduction or entrapment, influenced by the reductant used and reaction temperature.

These findings are consistent with recent studies on manganese and aluminum recovery via high-temperature smelting-aluminothermic reduction processes [3,55] . The HAlMan process, for instance, highlights efficient Mn and Al-Mn alloy production through hydrogen pre-reduction followed by smelting-aluminothermic reduction, aiming for low energy consumption and carbon footprint [12,42].

Comparatively, literature reports for smelting-aluminothermic reduction or carbothermic smelting of manganese alloys show Fe and Mn recoveries typically between 85–95% under optimized industrial conditions. The recoveries obtained in this study, especially for Fe and Mn, fall within or approach this range, confirming that the process achieved near-industrial efficiency for the primary target metals despite the use of secondary aluminum as the reductant. The overall recovery data aligns with expectations for such processes and underscores the impact of process conditions and pre-reduction methods, such as hydrogen-based pre-reduction of ferromanganese oxides [9], on overall process efficiency.

Overall, the recovery data validate the process performance and demonstrate that the reduction pathway effectively transferred target metals into the alloy phase while maintaining acceptable slag compositions for subsequent handling or recycling.

D. Elemental Deportment

The observed deportment patterns confirm the strong reducibility and affinity of Fe and Mn for alloy formation, leading to their preferential partitioning into the metal phase. The high concentration of Al in the slag indicates that this element predominantly remained in its oxide form, aligning with its thermodynamic stability under the experimental conditions.

The substantial deportment of Si to the metal phase suggests effective co-reduction of silica under these process parameters. This selective partitioning is a key aspect of element distribution in silicomanganese production processes, where pilot-scale studies confirm the distribution of minor and trace elements across product streams [56]. Minimal dust losses generally signify efficient capture within the primary phases. Moreover, in the smelting-aluminothermic reduction there is no gas generation via the chemical reactions, and hence the transport of fine particles is minimal due to more natural convection.

The deportment behavior observed here agrees with reported findings for smelting-aluminothermic reduction of manganese and iron oxides, where Fe and Mn exhibit preferential reduction and metallic enrichment, while Al, Si, and Ca (if present) remain slag-associated [3,55]. The effective partitioning of Mn and Fe into the metal phase and the confinement of Al and a significant portion of Si in the slag phase (though less for Si in this overall data) indicate effective process selectivity and minimal contamination of the metallic product. Studies on slag properties in ferroalloy production further elucidate these phase distributions [5,57] .

The deportment results demonstrate that the process achieved stable and predictable phase separation, with elemental distributions consistent with thermodynamic predictions and industrial benchmarks. This confirms the reproducibility and robustness of the reduction–tapping sequence for alloy recovery.

E. Slag-to-Metal Ratio

The ratio of slag to metal is an important measure of how efficient and sustainable manganese ferroalloy manufacturing is, especially in the HAlMan process. The SMRs for industrial high-carbon ferromanganese production using submerged arc furnaces are usually between 1.0 and 1.6 kg/kg, however these numbers might change depending on the grade of the ore and the conditions under which it is being worked. The HAlMan method, which uses smelting-aluminothermic reduction, shows competitive SMRs. For example, the Nchwaning ore (retort packed-bed) has an SMR of 1.34 kg/kg, which is in line with or even higher than the upper end of usual industrial ranges. Aluminothermic methods naturally make alumina-rich slag. The exact SMR depends on how much reductant is used and the stoichiometry of the reaction. One of the main benefits of employing pre-reduced manganese ores, as this study does, is that it lowers the overall volume of slag compared to smelting raw ores. This results in lower SMRs and better energy efficiency. It is worth noting that more optimization of the process and reaching high Mn yields will reduce the slag mass and increase the metal mass, and hence lowers the SMR.

A favorable SMR has a big impact on both the cost of the process and the way waste is handled. A lower SMR means that less raw materials are used, less energy is needed to melt and heat the slag, and the furnace works better since it can process more metal. From an environmental point of view, reducing slag production directly lowers disposal costs and makes the operation more sustainable, which is important because of increased regulatory pressure on solid waste management. Changes in pre-reduction approach and how the plant runs also affect the SMRs that are seen. For instance, Nchwaning ore that was pre-reduced in a retort packed bed furnace had a lower SMR (1.34 kg/kg) than the same ore that was pre-reduced in a plasma rotary furnace (1.79 kg/kg). To get the best metal recovery and SMR, it’s important to change operational parameters like adding aluminium and changing the chemistry of the slag. The study shows that pre-reducing manganese ore, especially Nchwaning ore in a packed-bed retort furnace, can lead to competitive SMRs. This means that there is room for additional optimisation to improve both economic and environmental performance.

4.1.3. Process Sustainability

A. Refractory Consumption

Refractory consumption rates in experimental campaigns ranged from 0.1051 to 0.1838 kg MgO/kg metal, influenced by furnace design, operating parameters like temperature and slag chemistry, and the refractory material itself. Minimizing refractory wear is crucial in industrial ferroalloy production due to its impact on costs and campaign life. While specific figures for experimental manganese ferroalloy smelting are scarce, general industry benchmarks for electric arc furnace operations, especially with aggressive slags, highlight the challenge of refractory degradation. DC arc furnaces, like the one used in this study, offer flexibility but can also contribute to refractory erosion due to intense localized heat and stirring. The observed refractory consumption rates have significant implications for the HAlMan process’s economic viability and operational stability. Lower consumption directly reduces material and labor costs associated with maintenance. Slower wear extends furnace campaign life, improving availability and productivity by reducing shutdowns. Consistent refractory wear also contributes to more stable furnace operation, mitigating risks of failures. These differences in consumption are linked to pre-reduction approach and slag characteristics, with slag chemistry and basicity being a major factor. Aggressive slag components can accelerate refractory dissolution, suggesting that a lower SMR and corresponding lower consumption, as seen with Nchwaning ore, likely resulted from a less aggressive slag composition or a more stable slag-refractory interface.

B. Electrode Consumption

The specific electrode consumption rates varied considerably across the three experimental campaigns, ranging from a notably low 0.0014 kg electrode/kg metal for the Nchwaning ore pre-reduced in a retort packed-bed furnace to 0.1105 kg electrode/kg metal for the UMK ore pre-reduced in a retort packed-bed furnace. The Nchwaning ore pre-reduced in a plasma rotary furnace showed an intermediate consumption rate of 0.0596 kg electrode/kg metal.

The exceptionally low consumption rate observed for the Nchwaning ore pre-reduced in a retort packed-bed furnace (0.0014 kg/kg) is striking and warrants further investigation into the specific operational conditions, slag chemistry, and/or metal tapping procedures during this campaign that might have contributed to such minimal electrode degradation. This value, if representative, signifies a major advantage for this particular processing route.

Electrode consumption rates in industrial ferroalloy production are typically higher than those observed in steelmaking EAFs due to the more aggressive nature of ferroalloy slags and often longer processing times. In the HAlMan process, there is no CO₂ gas formation in the smelting-aluminothermic reduction furnace, and therefore there is no electrode consumption via the Boudouard reaction at elevated temperatures, which is a mechanism of graphite electrode consumption in the ferroalloys furnaces.

While precise, universally comparable figures for electrode consumption in laboratory or pilot-scale DC EAFs for ferromanganese are not readily available, for high carbon ferromanganese production in submerged arc furnaces, energy consumption typically ranges from 2.0 to 3.5 MWh per tonne of alloy [5]. Electrode consumption is a component of overall production cost, and reducing it is a continuous goal in the industry [58]. Using manganese blends with a higher Mn/Fe ratio can reduce coke consumption, which in turn can lead to reduced electrode consumption [59].

Environmental Indicators: CO₂-Equivalent Emissions per Ton of Product, Dust and Particulate Emissions, NO_x/SO_x Emissions

4.1.4.1. A. Stack Emissions and Atmospheric Pollutants

The summary of measured stack gas parameters for the 200 kW DC arc furnace during HAlMan process Campaign 2 operation have shown a CO₂ concentration of 775 mg/Nm³, which represents ambient background levels, not a process emission from the furnace [24]. The observed high oxygen content of 20.8 %v/v indicates that the oxidizing off-gas conditions and minimal formation of reducing gases such as CO, which was measured at 0 mg/Nm³ [24]. These values are well below typical industrial thresholds, indicating a low atmospheric impact; however, no specific regulatory limits were compared against as this was the first run of the process [24].

Assuming a representative volumetric flow rate of 1,200 Nm³ h⁻¹ and an alloy production rate of 10 kg h⁻¹, the corresponding mass emission rates and emission factors were calculated. These translate to emission factors of 26 kg PM, 0.8 kg SO₂, 2.4 g NO_x, and 0.34 kg CO per tonne of alloy. Compared with typical values for conventional carbothermic HCFeMn smelting (PM 1–3 kg t⁻¹, SO₂ 0.3–0.8 kg t⁻¹, NO_x 0.2–0.6 kg t⁻¹; [7,60]), the HAlMan process exhibits a substantially lower gaseous emission intensity and a cleaner overall profile.

B. Greenhouse-Gas Emissions

Direct CO₂ emissions from the furnace are negligible owing to the absence of carbon reductants. The total CO₂-equivalent impact therefore consists of indirect contributions from electricity generation and aluminum reductant production:

$$E_{{\mathrm{CO}}_2,\mathrm{total}}=E_{\mathrm{elec}}+E_{\mathrm{Al}-\mathrm{prod}}.$$

(13)

For a specific energy consumption of 1,500 kWh t⁻¹ and a South African grid emission factor of 0.8 kg CO₂ kWh⁻¹, the electricity-related contribution ($E_{\mathrm{elec}}$) is calculated as 1.20 t CO₂-eq t⁻¹. Assuming 0.30 t Al t⁻¹ alloy and an embodied carbon intensity of 0.5 kg CO₂ kg⁻¹ for recycled aluminum , the aluminum production contribution ($E_{\mathrm{Al}-\mathrm{prod}}$) is 0.15 t CO₂-eq t⁻¹. This gives a total CO₂-equivalent impact of 1.35 t CO₂-eq t⁻¹.

This represents a 60-65% reduction compared with the 3-4 t CO₂-eq t⁻¹ typical of carbothermic HCFeMn production [7,60]. The choice of electricity source and the use of recycled aluminum significantly influence the overall carbon footprint of metallurgical processes [61,62].

C. Aggregate Environmental Load

To express a single environmental-load index, the individual impact categories were weighted following the approach of [63] and [64], employing ReCiPe midpoint normalization (eqn:Environmental load). The specific weighting factors ($w_1$, $w_2$, $w_3$, and $w_4$) correspond to the relative environmental relevance of each category (CO₂, SO₂, NO_x, and PM, respectively).

$$E_{\mathrm{load}}=w_1{E}_{{\mathrm{CO}}_2}+w_2{E}_{{\mathrm{SO}}_2}+w_3{E}_{{\mathrm{NO}}_{\mathrm{x}}}+w_4{E}_{\mathrm{PM}},$$

(14)

where weighting factors $w_1=1.0$, $w_2=0.25$, $w_3=0.30$, and $w_4=0.05$ correspond to the relative environmental relevance of each category (ReCiPe midpoint normalization). The resulting aggregate impact for the HAlMan process is dominated by indirect CO₂ emissions, while contributions from acidifying and particulate pollutants remain minor.

4.2. Effect of Pre-Reduction Approach

4.2.1. Energy

A. Power Stability Metrics: Arc Stability, Voltage/Current Fluctuations

Figure 14 shows the comparison of the set-points (current, voltage and power) against experimental observed values for the (a) Nchwaning ore pre-reduced in a rotary plasma furnace, (b) UMK ore pre-reduced in a packed-bed retort furnace, and (c) Nchwaning ore pre-reduced in a packed-bed retort furnace.

Deviations between target and experimental values are inherent to electric arc furnace operations and reflect the dynamic nature of the arc, the varying charge characteristics, and the control system’s response [25,26]. The observed fluctuations (in Figure 14) are direct indication of arc instability, which can also be influenced by factors such as arc movement and plasma jet length variations [27,28].

Critically, voltage instabilities and voluntary voltage reduction directly affect power stability and reduce delivered power [10,29]. In this experimental framework, initial arc instability at 100 V necessitated a reduction in target voltage to 70 V for some trials with UMK ore, indicating a necessary operational adjustment to maintain a stable process and avoid excessive fluctuations. Control systems in many EAF operations aim to minimize variations in arc voltage and current to maximize power input and stabilize the process [26,30].

Large and frequent deviations and transients are known to increase electrode erosion significantly [31]. This occurs through enhanced oxidation at hot spots during arc wandering, mechanical tip wear during electrode slipping and re-striking, and localized hot-spot evaporation or chemical attack of the electrode and refractory lining [31]. The impact on refractory life is also pronounced, as arc wandering and poor heat coupling increase local refractory heating, which can shorten the lining’s operational lifespan[31]. High voltage and current fluctuations are also associated with increased wear on furnace components and higher energy losses [29].

Arc stability in EAFs is not typically measured by a single universal efficiency percentage. Instead, it is often quantified using statistical indices of voltage and current signals, such as standard deviation, variance, or specific power factor ranges [32,33] . High voltage and current fluctuations indicate a less stable arc, leading to erratic heat transfer and higher energy losses [29].

Researchers have proposed various metrics and models, including the Arc Quality Index, to evaluate arc quality and stability [34]. In this study, power stability in a DC arc furnace is assessed using three complementary, quantitative metrics, namely, current-tracking error (percentage deviation between measured and set-point current), voltage deviation (percentage deviation between measured and set-point voltage), and realized power deviation (percentage deviation between measured and target active power).

These quantitative metrics can be defined by

$$\Delta I_{\%}={\displaystyle \frac{I_{\exp }-I_{\mathrm{set}}}{I_{\mathrm{set}}}}\times 100\%,$$

(15)

$$\Delta V_{\%}={\displaystyle \frac{V_{\exp }-V_{\mathrm{set}}}{V_{\mathrm{set}}}}\times 100\%,$$

(16)

$$\Delta P_{\%}={\displaystyle \frac{P_{\exp }-P_{\mathrm{set}}}{P_{\mathrm{set}}}}\times 100\%.$$

(17)

These metrics together quantify how closely the furnace follows operator setpoints and the degree of transient disturbances that can increase electrode wear and refractory attack [13]. The observed statistics, indicating a systematic shortfall of realized current and power relative to set-points (mean deviations approximately -10% to -12%) with moderate shot-to-shot variability (standard deviations 4% to 6%), can be attributed to factors such as arc instabilities, occasional electrode slipping/immersion transients that reduce effective current transfer, variations in bath impedance during feed events, and voltage reduction initiated by unstable arc conditions [35].

B. Load Factor/Power-on Time

All experimental campaigns, including both the Nchwaning manganese ore lumps pre-reduced in a plasma rotary furnace and the UMK and Nchwaning ores pre-reduced in retort packed bed furnaces, demonstrated exceptionally high load factors, ranging from approximately 95.89% to 96.89%. This indicates a remarkably efficient utilization of the experimental furnace across all tested pre-reduction routes, characterized by minimal inter-tap downtime. Such high operational continuity is highly advantageous in industrial settings, suggesting that these smelting routes, particularly when coupled with efficient material handling, can achieve superior productivity. This performance aligns with the inherent operational flexibility and process control advantages offered by DC arc furnace technology, which often translates to higher furnace availability and lower heat losses during operation compared to other furnace types.

For the UMK ore, it’s notable that one tap (Tap No. 1C2) registered 0.0 h tap duration, which might indicate a non-operational or aborted run. While this specific instance would contribute to downtime if it were a planned part of the cycle, its impact on the overall load factor, given the very low summed $t_{\mathrm{down}}$ values, suggests that the typical operational flow between taps was very efficient. The very high load factors observed for the retort packed bed furnace campaigns indicate that the downtime associated with cooling, material discharge, and subsequent reloading, although inherent to batch processes, was managed very effectively in these experimental setups, resulting in minimal idle periods. While pilot-scale experimental facilities typically experience more frequent interruptions for process adjustments, sampling, and maintenance compared to optimized industrial plants [21,67], the consistently high load factors observed across all campaigns suggest robust operational management during these trials.

Industrial electric arc furnaces, particularly in sectors such as steelmaking and ferroalloy production, target high utilization rates to maximize throughput and cost-effectiveness. For instance, the average capacity utilization factor for scrap-based EAFs in the EU has been reported around 83% [45].

The high load factors achieved in all experimental campaigns, notably exceeding typical industrial averages, suggest a strong potential for industrial scaling and commercial viability for these manganese production routes. This indicates that the operational aspects, at least in terms of furnace availability and minimized idle time, are highly favorable. High downtime, conversely, leads to significant economic penalties due to lost production and sustained energy losses from refractory linings in intermittent operations [67].

The presented load factors underscore that the choice of pre-reduction approach, while influencing chemical and metallurgical efficiencies, was implemented in a manner that ensured high operational continuity in the subsequent smelting process. This operational analysis provides crucial quantitative data for comprehensive techno-economic assessments and informs strategic decisions regarding the industrial scalability and commercial viability of these novel manganese production routes. Future work might focus on maintaining these high load factors during scale-up to larger industrial operations, considering the challenges associated with managing larger material volumes and more complex operational logistics.

4.2.2. Materials

A. Reductant Efficiency

$${\eta }_{\mathrm{Al}}={\displaystyle \frac{\mathrm{Al}\mathrm{consumed}\mathrm{to}\mathrm{produce}\mathrm{Mn}+\mathrm{Al}\mathrm{consumed}\mathrm{to}\mathrm{produce}\mathrm{Fe}+\mathrm{Al}\mathrm{consumed}\mathrm{to}\mathrm{produce}\mathrm{Si}}{\mathrm{Total}\mathrm{Al}\mathrm{consumed}}}\times 100$$

(18)

The reductant efficiencies observed in this study (as shown in Figure 15), demonstrate variability depending on the specific pre-reduction method and ore type. For Nchwaning manganese ore lumps pre-reduced in a plasma rotary furnace, efficiencies ranged from 28.98% to 75.34%. When UMK manganese ore was pre-reduced in a retort packed bed furnace, the efficiencies spanned from 19.46% to 56.67%.

The complete reduction of manganese oxide via the aluminothermic process is fundamentally influenced by factors such as the aluminum content in the charge and critical aspects of slag chemistry [3]. Optimized slag conditions, characterized by carefully controlled basicity, viscosity, and liquidus temperature, are essential for efficient metal-slag separation and high recovery rates of manganese [46,47]. Variations in these parameters can significantly impact the effectiveness of aluminum as a reductant, affecting overall efficiency and metal yield.

Nchwaning manganese ore pre-reduced in a retort packed bed furnace exhibited efficiencies between 30.89% and 55.32%. This can be attributed to the pre-reduction extent of the respective ores via respective pre-reduction approaches.

Furthermore, process parameters such as pre-reduction methods, furnace design and the inherent characteristics of the manganese ore (e.g., mineralogy, particle size) play a crucial role in determining reductant efficiency [4,5,48] . Excess aluminum additions, oxidation losses, and the efficiency of recovery in the final alloy also contribute to the overall reductant utilization [49]. Higher efficiency values indicate a substantial proportion of the added aluminum actively participates in the desired reduction reactions, leading to improved recovery of valuable metals and a reduction in energy consumption [49]. Continuous optimization of these integrated process elements is vital for advancing sustainable and economically viable manganese production [12,18]. The process was implemented at elevated temperatures and the charged aluminum metal was in exposure to air for some time before the start of reduction reactions, and this may oxidize a portion of the aluminum metal, and hence reduce the redundant efficiency. Hence, using a protective gas may provide higher reduction efficiency.

B. Slag-to-Metal Ratio

Comparing these values to industrial and literature benchmarks reveals insights into the efficiency of the HAlMan process.

• Industrial High Carbon Ferromanganese Production: Conventional industrial HCFeMn production via submerged arc furnaces typically operates with SMRs in the range of 1.0 to 1.6 kg/kg, although values can vary depending on ore quality, furnace type, and operating practices [69]. The HAlMan process, which utilizes smelting-aluminothermic reduction, has been benchmarked for its eco-efficiency against industrial carbothermic HCFeMn production, with positive findings regarding energy consumption and emissions. The reported SMRs in this study, particularly the 1.34 kg/kg for Nchwaning ore (retort packed-bed), fall within or are competitive with the upper end of typical industrial ranges for HCFeMn.
• Aluminothermic Reduction: Aluminothermic processes, by their nature, can generate significant amounts of alumina-rich slag. The specific SMR for smelting-aluminothermic reduction largely depends on the reductant usage, slag former additions, and overall process stoichiometry. Generally, processes aiming for high metal recovery might produce more slag to ensure efficient separation and impurity removal. It is worth noting that the obtained slag byproduct is valuable and consumable and can be utilized for smelter grade alumina production for the primary aluminum production.
• Pre-reduced Ores: The use of pre-reduced manganese ores, as in this study, is expected to reduce the overall slag volume compared to smelting raw ores, as a significant portion of oxygen has already been removed. This should inherently contribute to lower SMRs and improved energy efficiency.

Implications for Process Economics and Waste Management:

The SMR directly impacts the economic and environmental sustainability of the process:

• Process Economics: A lower SMR leads to several economic advantages:
  • Reduced Raw Material Consumption: Less slag formers are required, decreasing raw material costs.
  • Lower Energy Demand: Less slag to melt and heat reduces specific energy consumption per unit of metal produced.
  • Increased Furnace Productivity: A lower slag volume allows for a higher throughput of metal, improving furnace capacity utilization.
• Waste Management: Slag generation is a major waste stream in ferroalloy production. A lower SMR translates directly to:
  • Reduced Slag Disposal Costs: Significant cost savings on landfilling or further processing of slag.
  • Lower Environmental Impact: Minimizing solid waste contributes to a more sustainable operation and addresses increasing regulatory pressures. Studies highlight the importance of characterizing and valorizing metallurgical slags to reduce environmental impact and conserve resources.

Relationship to pre-reduction approach and operational adjustments:

The observed SMRs can be related to the choice of pre-reduction technology and operational parameters:

• Pre-reduction approach:
  • Nchwaning ore pre-reduced in a packed bed vs. Nchwaning ore pre-reduced in a rotary plasma furnace: The Nchwaning ore pre-reduced in the retort packed bed furnace yielded a lower SMR (1.34 kg/kg) compared to the same ore pre-reduced in the plasma rotary furnace (1.79 kg/kg). This difference could be attributed to variations in the degree of pre-reduction achieved by each method, or differences in the residual gangue composition and morphology after pre-reduction. A more selective or efficient removal of gangue components during pre-reduction in the retort packed-bed could lead to a cleaner feed for the smelting furnace, thus requiring less slag generation.
  • UMK: The UMK ore pre-reduced in the retort packed bed furnace resulted in the highest SMR (1.87 kg/kg) among all campaigns. This suggests that the specific characteristics of the UMK ore, even after pre-reduction in the retort, might necessitate a higher slag volume during smelting to achieve desired metal recovery or impurity removal. Ore composition, particularly the ratio of manganese oxides to silica and alumina, heavily influences the required slag volume and composition.
• Operational Adjustments: While the provided data does not directly detail SER modulation or specific Al additions per tap, these operational parameters are crucial in influencing SMR.
  • Aluminium Additions: In smelting-aluminothermic reduction, the amount of aluminium reductant added directly impacts the quantity and composition of the alumina-rich slag formed. Precise control of Al additions is vital to optimize both metal recovery and SMR.
  • Slag Chemistry and Basicity: Adjustments to slag chemistry (e.g., lime or silica additions) to achieve optimal basicity and fluidity for efficient metal-slag separation and impurity removal will directly influence the slag volume and, consequently, the SMR. These adjustments are often fine-tuned based on the specific feed material and desired product quality.

The SMRs achieved in this study demonstrate that manganese ore pre-reduction, particularly with the Nchwaning ore in a packed-bed retort furnace, can lead to competitive SMRs when compared to industrial benchmarks. Further optimization of pre-reduction processes and smelting parameters, especially regarding aluminum utilization and slag chemistry, can further enhance these ratios, driving improved economic and environmental performance.

4.2.3. Process Sustainability

A. Refractory Consumption

Refractory consumption, specifically the loss of magnesia from the furnace lining, was estimated by calculating the difference between the cumulative MgO leaving the system in the slag and the cumulative MgO introduced with the feed materials over each campaign. The results, expressed as kilograms of MgO consumed per kilogram of metal produced, are summarized in Table 12 and in Figure 17.

The refractory consumption rates varied between 0.1051 kg MgO/kg metal and 0.1838 kg MgO/kg metal across the experimental campaigns. Refractory consumption rates are highly dependent on specific furnace design, operating parameters (e.g., temperature, slag chemistry, specific energy input), and the type of refractory material used. In industrial ferroalloy production, minimizing refractory wear is a continuous objective, given its direct impact on operational costs and campaign life.

• Ferroalloy Production: While precise, universally comparable figures for refractory consumption in experimental manganese ferroalloy smelting are scarce, general industry benchmarks for EAF operations, particularly those involving aggressive slags, often highlight the challenge of refractory degradation. For example, specific refractory consumption in steelmaking EAFs can range significantly, but typical values are often targeted below 1 kg/tonne steel. Given the nature of manganese alloys and the typically more aggressive slags, these values might be higher for ferroalloy production.
• DC Arc Furnaces: The DC arc furnace used in this study is known for its ability to operate with greater flexibility in terms of raw materials and slag chemistry compared to AC furnaces. However, the intense localized heat and strong stirring action caused by the arc can also contribute to refractory erosion if not properly managed [21]. In ferroalloy furnaces, there is usually a cold burden or frozen slag in contact with the refractory, and therefore much less refractory interaction occurs in comparison with EAF in which molten slag is in contact with molten slag.

Implications for Process Economics and Operational Stability:

The observed refractory consumption rates have significant implications for the economic viability and operational stability of the HAlMan process:

• Cost Reduction: A lower refractory consumption rate, as achieved with the Nchwaning ore pre-reduced in a retort packed-bed furnace, translates directly into reduced material costs for refractories and decreased labor costs associated with patching and relining the furnace.
• Increased Furnace Availability: Slower refractory wear extends furnace campaign life, leading to fewer planned and unplanned shutdowns for maintenance. This improves furnace availability and overall productivity, directly impacting the process’s economic performance [67].
• Operational Stability: Consistent and predictable refractory wear contributes to more stable furnace operation, reducing the risk of breakouts or other refractory-related failures.

Relationship to pre-reduction approach and operational adjustments:

The differences in refractory consumption rates can be linked to the pre-reduction approach and the resulting slag characteristics:

• Slag Chemistry and Basicity: The most significant factor influencing refractory wear in smelting is the chemical interaction between the molten slag and the refractory lining. Higher slag basicity and the presence of certain aggressive components (e.g., highly oxidizing conditions or specific oxides) can accelerate refractory dissolution. The lower SMR for the Nchwaning ore (retort packed-bed) and its corresponding lower refractory consumption suggest that this combination might have produced a less aggressive slag composition or a more stable slag-refractory interface. The specific composition of the slag (e.g., MgO content, silica content) dictates its corrosivity to MgO-based refractories [21].

B. Electrode Consumption

The specific electrode consumption rates varied considerably across the three experimental campaigns, ranging from a notably low 0.0014 kg electrode/kg metal for the Nchwaning ore pre-reduced in a retort packed-bed furnace to 0.1105 kg electrode/kg metal for the UMK ore pre-reduced in a retort packed-bed furnace. The Nchwaning ore pre-reduced in a plasma rotary furnace showed an intermediate consumption rate of 0.0596 kg electrode/kg metal.

The exceptionally low consumption rate observed for the Nchwaning ore pre-reduced in a retort packed-bed furnace (0.0014 kg/kg) is striking and warrants further investigation into the specific operational conditions, slag chemistry, and/or metal tapping procedures during this campaign that might have contributed to such minimal electrode degradation. This value, if representative, signifies a major advantage for this particular processing route.

Electrode consumption rates in industrial ferroalloy production are typically higher than those observed in steelmaking EAFs due to the more aggressive nature of ferroalloy slags and often longer processing times.

While precise, universally comparable figures for electrode consumption in laboratory or pilot-scale DC EAFs for ferromanganese are not readily available, for high carbon ferromanganese production in submerged arc furnaces, energy consumption typically ranges from 2.0 to 3.5 MWh per tonne of alloy [5]. Electrode consumption is a component of overall production cost, and reducing it is a continuous goal in the industry [58]. Using manganese blends with a higher Mn/Fe ratio can reduce coke consumption, which in turn can lead to reduced electrode consumption [59].

4.3. Comparative Benchmarking

When benchmarked against industrial carbothermic HCFeMn production, the HAlMan process exhibits substantial improvements in multiple performance indicators: (i) a marked reduction in specific energy consumption, (ii) significantly lower CO₂-equivalent emissions, (iii) precipitously lower NO_x/SO_x stack outputs under the pilot operating conditions, and (iv) superior eco-efficiency indices. These improvements collectively validate the environmental and economic promise of the HAlMan approach for valorising HCFeMn slag.

Table 13 summarises the pilot (200 kW DC arc furnace) results and places them in the context of typical industrial ranges for conventional carbothermic HCFeMn and related ferroalloy production routes.

Interpretation and Scale-Up Implications

The pilot results demonstrate a clear energy advantage for the HAlMan route: the measured SER ($\approx 0.88$ MWh t⁻¹) is 56-75% lower than the commonly reported SAF benchmarks (2.0-3.5 MWh t⁻¹), reflecting the combined effect of the aluminothermic exotherm and the DC arc furnace configuration. The estimated cradle-to-gate CO₂-intensity ($\approx 1.35$ t CO₂-eq t⁻¹) likewise compares favourably with typical carbothermic values (≈ 3–4 t CO₂-eq t⁻¹), indicating major potential for GHG mitigation when secondary (recycled) aluminium and low-carbon electricity are used.

However, several caveats must be considered for upscaling:

• Slag volumes: The SMR observed in this study was 1.34 kg/kg and 1.79 kg/kg for the Nchwaning pre-reduced ores in a retort furnace and plasma furnace, respectively. For the UMK ore pres-reduced in the retort furnace, the SMR was observed to be 1.87 kg/kg.
• Emission normalisation and abatement: Pilot stack concentrations indicate low gaseous emissions by concentration, but normalised emissions (kg t⁻¹) are sensitive to assumed flue volumes and capture efficiency. Industrial implementations would need robust abatement and continuous monitoring to ensure compliance and low net emissions after normalisation.
• Operational stability: Measured current/voltage/power deviations (mean deviations ∼10–12%) show room for process-control improvements; stability gains typically reduce electrode/refractory wear and raise delivered power, improving SEC further at scale.
• Economic balance: The pilot energy efficiency indicator (EEI) and cost assessment (CAE) indicator in Section 3.1.1 suggest favourable eco-economic outcomes under the study assumptions (product price, aluminium costs, grid factor). Full techno-economic analysis (including capex, slag conditioning and transport, and regulatory compliance) is required before firm commercial viability claims.

Overall, the pilot campaign results indicate that the HAlMan process can deliver major energy and emission advantages versus conventional carbothermic HCFeMn production, while presenting specific challenges (notably slag volumes and process stability) that must be targeted in scale-up and integration studies.