Enhanced Chromite Recovery from Tailings via a Custom-Designed Shaking Table: Optimization and Performance

Abstract

Significant chromite losses to tailings in gravity separation plants arise from both suboptimal separator design and inefficient beneficiation processes, posing major challenges to resource utilization, energy efficiency, and environmental sustainability. These losses are particularly critical because the material, already comminuted to liberation size, is discarded, leading to reduced concentrate yield, wasted energy input, and increased environmental pollution. To address this issue, an industrial-scale custom-designed shaking table was developed and tested to recover marketable-grade chromite concentrate (≥42% Cr₂O₃) from processing plant tailings containing 3.25%–4.25% Cr₂O₃, which had accumulated over years of chromite beneficiation. Experimental results showed that, under optimized operating parameters (320 rpm stroke frequency, 13 mm stroke length, 1° deck slope, 1300 g/L pulp density, 800 kg/h feed rate, and 7 tph wash water flow rate), Cr₂O₃ recovery increased from 8% to 27% for the first and second floor operations and from approximately 17% to 41% for the third and fourth floor operations compared with existing plant performance. The results revealed a strong interdependence between Cr₂O₃ recovery and concentrate grade, both of which are critical indicators of process efficiency. Intermediate particle sizes (−0.250 + 0.150 mm) provided the most favorable balance, yielding high recovery rates without substantially compromising the concentrated grade.

1. Introduction

Chromium is a naturally occurring element in the Earth’s crust and is present in approximately 25 different minerals. Among these, chromite is the most common and contains the highest chromium content [1]. It is a critical mineral for several key industries essential to modern life, including the chemical, metallurgical, refractory, and foundry sectors. The production of stainless steel is the primary driver of ferrochrome demand, accounting for approximately 94% of global chromium consumption [2,3].

Globally, there are over 12 billion tons of marketable chromite reserves, with the majority hosted in stratiform-type deposits. Podiform deposits constitute the second largest source of chromite reserves, while laterites—weathering products of peridotite—contribute minimally to global chromite production [4,5,6,7,8]. Chromite-bearing ores are commonly associated with gangue minerals like olivine, serpentine, chlorite-group minerals, feldspars, and pyroxenes, with specific gravities ranging from approximately 2.6 to 4 g/cm³ [9]. Processing generally involves comminution to liberate chromite particles, followed by concentration to achieve marketable Cr₂O₃ grades exceeding 34%, which can be directly used in industrial applications.

Numerous studies have employed gravity separation techniques for the beneficiation of chromite ores. These studies highlight the effectiveness of various gravity methods, such as shaking tables, spirals, jigs, and dense media separation, in concentrating chromite [10,11,12,13,14]. The significant density difference between chromite (3.58–5.09 g/cm³) [15,16] and its associated gangue minerals makes gravity concentration the most widely employed and well-established beneficiation method. This approach enables the economical separation of valuable chromite particles from gangue while allowing precise control over operational parameters to achieve higher concentrate grades [17,18,19,20,21]. Among various gravity separation techniques, shaking tables are particularly preferred for chromite beneficiation because they combine relatively low operating costs with operational simplicity and minimal maintenance requirements compared to alternative methods [22]. They operate on the flowing film principle, where a thin slurry layer moves across a riffled deck, enabling density-driven stratification and differential motion of particles for effective separation. Although they offer high separation efficiency, particularly for fine- and medium-sized particles, the process, like other flowing film-based methods, is highly sensitive to feed rates. Mono-layer feeding, where particles are distributed in a single layer, enhances separation performance and enrichment efficiency by ensuring direct interaction between particles and the flowing film, allowing gravity to act more selectively [23]. In contrast, industrial operations often prioritize throughput over selectivity, using multi-layer feeding in which particles interact with one another rather than the separating medium, thereby reducing concentrate purity [2,24].

With the depletion of high-grade ores that can be directly used after mining, the enrichment of fine-grained and low-grade ore reserves has become increasingly important to meet industrial demand [25]. Consequently, lower-grade ores can be upgraded for industrial applications using suitable beneficiation techniques, depending on the degree of particle liberation and the nature of the associated gangue minerals [2,26]. At the same time, critical design parameters of shaking tables—such as stroke, stroke length, and inclination—are often fixed or difficult to adjust in industrial settings. Inadequate control over these variables reduces separation efficiency and leads to substantial chromite losses to tailings, even though these particles incur almost no comminution cost. Such losses not only represent a loss of valuable material but also contribute to environmental risks and storage challenges [3], prompting both industrial practice and academic research to increasingly focus on developing strategies for minimizing chromite losses in tailings during gravity separation [5,13,14,21,27].

This study focuses on recovering chromite from plant tailings using a custom-designed, industrial-scale shaking table that was both modified and operationally optimized for enhanced performance. The novelty lies in combining the valorization of low-grade tailings with the development of a shaking table offering on-site adjustment of critical parameters, such as stroke, stroke length, and slope, features typically absent or highly constrained in conventional systems. This adaptability enables more efficient chromite recovery under challenging conditions, particularly at higher strokes and shorter stroke lengths, which exceed the capabilities of existing plant tables. The research specifically aimed to produce marketable chromite concentrate (>42% Cr₂O₃) from tailings, identify optimal operating parameters, and benchmark the new design against current plant equipment. By targeting the recovery of approximately 10 million tons of accumulated tailings (3.25%–4.25% Cr₂O₃, comparable to run-of-mine ore) from a long-standing mining operation in Bursa, the study provides a practical, cost-effective pathway for improving chromite beneficiation efficiency.

2. Materials and Methods

2.1. Taking Samples from the Tailings Storage Area

The physical conditions of the tailings storage area were first evaluated to establish an optimal starting point for reprocessing, as the tailings originated from previous plant operations, including grinding, screening, classification, and gravity separation. Samples were systematically collected from storage areas 1–3 (Figure 1) using a 5 m × 5 m grid sampling method with JCB-type backhoe loaders (Rochester, UK). Before testing, samples from different storage areas were combined using a Bobcat mini loader, then subdivided into representative samples applying alternate shovel sampling and coning–quartering for laboratory analysis.

For particle size adjustment, the mills previously employed for grinding were operated at low speeds (10 rpm) to disperse and disintegrate any agglomerated particles before the tests. In some cases, a small proportion of grinding balls (<5% by volume) was added to break down larger ore fragments, ensuring a uniform particle size distribution for subsequent processing.

2.2. Petrographic and XRD (X-Ray Diffraction) Analyses

The investigated chrome spinel-bearing ultramafic rock is identified as serpentinized dunite. Petrographic and XRD analyses were conducted to determine the mineralogical composition of the rock. The XRD results reveal that the rock consists of forsterite (olivine), serpentine (antigorite, chrysotile), chromite, and magnetite minerals (Figure 2). Thin-section analysis (Figure 3) indicates that olivine has undergone extensive serpentinization, with more than 50% of the original olivine minerals altered to serpentine. However, tiny relict olivine minerals (from 0.300 to 0.020 mm) are still observable. Under crossed polarized light, olivine exhibits third- to fourth-order interference colors, while appearing colorless under plane polarized light. Disseminated Cr-spinel crystals are mainly opaque and dispersed within the serpentine matrix. The chromian spinels are mainly subhedral to anhedral in shape and fine- to medium-grained (ranging in size from 1.5 mm to 0.005 mm, average < 0.350 mm). They display pull-apart structures, with fractures filled with serpentine minerals.

2.3. XRF (X-Ray Fluorescence) Analysis

Chemical analyses of the as-received tailings sample (used as the feed) and the concentrate, middling, and waste samples obtained from the experiments were conducted using a Thermo Fisher Scientific Niton XL3t XRF (Waltham, MA, USA) analyzer. Calibration with standards of known composition before measurements ensured high analytical accuracy and minimized error. The results for the as-received tailings sample are presented in

Table 1. Chemical analysis of the as-received tailings sample.

Component	%
Cr2O3	4.14
FeO	6.11
SiO2	36.82
MgO	38.3
CaO	0.95
Al2O3	1.4

.

2.4. Particle Size Analysis

The particle size distribution of the tailing sample used in the experimental tests was determined by wet sieving. The size distribution and corresponding Cr₂O₃% content of each size range are given in

Table 2. Size distribution of the as-received tailings sample.

Particle Size (mm)	Weight (%)	∑ Undersize (%)	Cr2O3 (%)	Cr2O3 Distribution (%)
+1	0.28	100.00	1.61	0.11
−1 + 0.850	0.05	99.72	1.73	0.02
−0.850 + 0.710	0.06	99.68	1.86	0.03
−0.710 + 0.600	0.23	99.61	2.01	0.11
−0.600 + 0.500	0.48	99.38	2.46	0.28
−0.500 + 0.425	8.70	98.90	2.55	5.36
−0.425 + 0.355	15.38	90.21	2.65	9.84
−0.355 + 0.300	8.89	74.82	3.35	7.19
−0.300 + 0.250	15.32	65.93	3.70	13.68
−0.250 + 0.212	3.71	50.62	4.13	3.70
−0.212 + 0.180	6.12	46.90	4.28	6.32
−0.180 + 0.150	9.17	40.78	4.32	9.56
−0.150 + 0.125	3.71	31.62	5.49	4.91
−0.125 + 0.106	6.17	27.91	5.81	8.65
−0.106 + 0.090	4.35	21.74	5.76	6.04
−0.090 + 0.075	3.51	17.40	5.68	4.82
−0.075 + 0.063	3.17	13.88	5.87	4.50
−0.063 + 0.053	1.64	10.71	5.81	2.29
−0.053 + 0.045	2.20	9.07	5.90	3.13
−0.045 + 0.038	2.84	6.88	5.73	3.93
−0.038	4.04	4.04	5.67	5.53
Total	100		4.14

. According to the results, the Cr₂O₃% content varied between 1.61% and 5.90%. The lowest Cr₂O₃ content was obtained in larger particle sizes, while the grade increased as the particle size decreased, resulting in the highest Cr₂O₃ grade with −0.150 mm particle size. Given that the Cr₂O₃ grade for the +0.500 mm size was below 1% and accounted for less than 0.6% of the total feed, the sample finer than 0.500 mm was used throughout the study.

2.5. Shaking Table Customization and the Methods Applied

The shaking tables in the existing plant have been in operation for many years and exhibit limited flexibility in adjusting key operating parameters such as stroke, stroke length, and deck slope. Furthermore, mechanical fatigue has reduced their overall operational efficiency. To address these limitations, while preserving the principal design characteristics of the plant-scale shaking table, specifically the deck dimensions (488 cm × 183 cm) and the riffle configuration on the deck (3 mm high and spaced 6 mm apart), a customized shaking table with easily adjustable high stroke speeds, variable stroke lengths, and an adjustable slope mechanism (

Table 3. Operating variables of the customized shaking table.

Variables	Values
Strokes (rpm)	≤450 rpm
Length of strokes (mm)	3–28 mm
Deck slope (°)	>0 ± 0.05

) was designed and manufactured for this study. Several components were integrated to optimize operation, including a digital display speed driver (Figure 4a), a lever-type transmission for easy stroke adjustment (Figure 4b), and a jointed and hinged assembly enabling precise control of deck inclination (Figure 4c). The deck slope was measured by a digital display inclinometer (Figure 4d).

A schematic diagram and photographs of the custom-designed shaking table are presented in Figure 5a,b. To improve operational control, a mechanism comprising a pump and feeding cone was implemented, allowing the table to operate in a closed circuit until equilibrium was achieved under the specified variables during pulp feeding. To prevent particle settling and clogging at the cone outlet, a mixing mechanism and pinch valve system were integrated into the feeding cone, ensuring the pulp maintained the desired density before entering the table. Product collection chutes positioned around the table facilitated the transfer (recirculation) of waste, middling, and concentrate streams back to the pump through a return pipe system, thereby maintaining a continuous feed to the cone. In addition, spare outlet systems equipped with sampling valves were installed on the chutes, enabling the collection of waste, middling, and concentrate samples once steady-state operation was reached under the designated parameters.

The shaking table setup was operated in a closed circuit until the operating variables reached equilibrium. Subsequently, samples were simultaneously collected from the concentrate, middling, and waste outlets using sampling valves (concentrate/middling and waste) over an average period of 10 s. After collection, each product was oven-dried at 100 ± 10 °C, weighed after drying, and analyzed by XRF to determine its chemical composition.

3. Results and Discussion

This study was conducted in two distinct phases. The first phase focused on utilizing existing plant facilities to recover a commercially viable chromite concentrate from tailings using a Wilfley-type shaking table operated within the mechanical constraints of the current processing equipment. The second phase aimed to improve process efficiency through the design, modification, and operation of a custom-designed shaking table. Results from the existing plant setup were systematically compared with those obtained using the custom-designed shaking table. In addition, key variables influencing chromite recovery from tailings were identified, and their optimal operating conditions were determined using the custom-designed shaking table.

3.1. Processing of the Tailings with the Existing Plant Facilities

Before conducting experimental tests with the custom-designed shaking table, the separation process was carried out under existing plant conditions using a standard Wilfley-type shaking table operated within the mechanical limitations of the facility. As shown in the chromite enrichment plant flowsheet (Figure 6), feed material with a particle size of −0.500 mm was initially processed on the first group of tables (1st–2nd floors). The middlings from this stage were subsequently directed to the next table group (3rd floor) for additional chromite recovery, while the 4th floor tables operated in a closed circuit to minimize chromite losses from the preceding stages.

The primary objective of this phase was to produce a marketable-grade chromite concentrate (>42% Cr₂O₃) while maximizing recovery and minimizing losses to waste. The operational process variables applied during testing are presented in

Table 4. Key variables for maximizing chromite recovery from tailings under existing plant conditions.

Variables	Values
Feed grade (Cr2O3%)	4.14 ± 0.05
Strokes (rpm)	260
Length of strokes (mm)	16
Deck slope (°)	1.5 ± 0.5
Pulp density (g/L)	1200
Feed rate (kg/h)	1050
Wash-water flow rate (tph)	8 ± 1

. Under optimal operating conditions, the 1st–2nd floor tables achieved a Cr₂O₃ recovery of approximately 8%, whereas the 3rd and 4th floor tables, which processed the middlings (about 65% of the 1st–2nd floor table feed), contributed an additional 17% Cr₂O₃ recovery.

3.2. Processing of the Tailings with a Custom-Designed Shaking Table

Analysis of chromite recovery from tailings using the existing plant revealed several challenges that limited the achievement of high Cr₂O₃ yields. The operational inefficiencies were attributed to constraints in the design and adjustable parameters of the existing shaking tables used during the concentration process. To overcome these limitations, an industrial-scale shaking table was designed, customized, and fabricated, featuring adjustable low-to-high stroke frequencies, variable stroke lengths, and a conveniently adjustable deck slope, all intended to improve Cr₂O₃ recovery from tailings with greater efficiency. This custom-designed unit was subsequently employed to determine the optimal operational parameters, including stroke, stroke length, slope, pulp density, and feed particle size range, that significantly influence Cr₂O₃ recovery. The results obtained from this setup were then compared with those achieved under the existing plant conditions.

3.2.1. Effect of Stroke Rate on Chromite Recovery

Particles in contact with the surface of the shaking table are transported in the direction of table motion through frictional forces. This shaking motion, which facilitates the transport of mineral particles, is characterized by its asymmetry [23], consisting of a slower forward stroke and a faster backward stroke. Under standard operating conditions, the table is operated at 250–300 rpm. In the initial phase of this study, the variables listed in

Table 5. Variables used to determine the optimum table stroke.

Variables	Values
Length of strokes (mm)	13
Deck slope (°)	1.0
Pulp density (g/L)	1300
Feed rate (kg/h)	500
Wash-water flow rate (tph)	7
Strokes (rpm)	280–360

, known to influence the separation process, were kept constant to evaluate the effect of stroke speed (rpm) on chromite recovery from processing plant tailings.

As shown in Figure 7, chromite recovery from the enrichment plant tailings exhibited an inverse relationship with the shaking table stroke rate: recovery decreased as stroke increased. The Cr₂O₃ recovery ranged from 10% to 22%, highlighting the significant influence of stroke on separation performance. At the lowest stroke rate of 280 rpm, the shaking action was insufficient to promote effective stratification of the dense chromite particles behind the riffles. Consequently, these particles were entrained into the middling zone and mixed with gangue minerals (primarily olivine) due to the scavenging effect of the wash water, resulting in the lowest recovery of approximately 10%.

With increasing shaking table stroke, Cr₂O₃ recovery improved, and at 320 rpm, chromite particles were effectively directed into the concentrate zone without intermixing with the middling product. Under these conditions, a distinct separation layer was formed, resulting in higher recovery rates. However, a further increase in stroke speed to 360 rpm led to the intrusion of gangue minerals, primarily olivine, into the concentrate zone, producing a broad separation band and a subsequent decline in Cr₂O₃ recovery from the tailings. Considering that the tailings were processed without additional grinding, one of the factors limiting further recovery improvement is the presence of unliberated particles. These particles likely reduced the efficiency of gravity separation, as they are less prone to report to the concentrate fraction.

3.2.2. Effect of Stroke Length on Chromite Recovery

In gravity concentration using a shaking table, stroke length is a key operational variable that significantly influences particle stratification and separation efficiency by affecting particle movement according to size and density differences. For coarser feed materials, longer strokes enhance the transport of heavier particles toward the concentrate end, whereas shorter strokes are more effective for finer feeds [9]. Based on this principle, the present study investigated the effect of stroke length, ranging from 7 mm to 16 mm, on the separation performance, Cr₂O₃ recovery from tailings, and resulting concentrate grade. During these experiments, all other operating parameters, including the previously determined optimum stroke rate of 320 rpm, were held constant as listed in

Table 6. Variables used to determine the optimum stroke length.

Variables	Values
Strokes (rpm)	320
Deck slope (°)	1.0
Pulp density (g/L)	1300
Feed rate (kg/h)	500
Wash water flow rate (tph)	7
Length of strokes (mm)	7–16

. The results are presented graphically in Figure 8.

The experimental results indicated that Cr₂O₃ recovery from the tailings was strongly affected by the shaking table stroke length. At short stroke lengths (7–10 mm), heavier chromite particles predominantly remained in the middling zone and failed to reach the concentrate area, resulting in Cr₂O₃ recoveries below 13% and 19%, respectively. Increasing the stroke length enhanced particle transport and broadened the collection band, promoting more efficient separation. When the stroke length was set to 13 mm, Cr₂O₃ recovery increased to approximately 22%. However, further increasing the stroke length to 16 mm caused chromite particles to bypass the riffles, designed to create a hindered settling region toward the concentrate, and shift into the middling and reject zones. Under these conditions, effective separation on the deck surface was not achieved, leading to a decline in Cr₂O₃ recovery to below 17%.

3.2.3. Effect of Shaking Table Slope on Chromite Recovery

Deck slope is a key parameter influencing Cr₂O₃ recovery, as it must be optimized according to the properties of both valuable and gangue minerals, including particle size and density. Increasing the table slope accelerates water flow, which in turn affects the lateral movement of particles across the deck. Higher slopes are generally more suitable for coarser feed materials, whereas near-horizontal slopes are preferred for finer fractions. Considering that the d₈₀ of the feed was −0.325 mm, shaking table slopes ranging from 0.5° to 2.0° were evaluated to assess their effect on chromite recovery from the tailings. The operating variables applied during these tests are summarized in

Table 7. Variables used to determine the optimum shaking table slope.

Variables	Values
Strokes (rpm)	320
Length of strokes (mm)	13
Pulp density (g/L)	1300
Feed rate (kg/h)	500
Wash-water flow rate (tph)	7
Deck slope (°)	0.5–2.0

, and the results are presented graphically in Figure 9.

The results demonstrated a clear influence of deck slope on chromite recovery from tailings. At the minimal slope of 0.5°, reduced water flow and increased water thickness hindered effective stratification, causing chromite particles to move irregularly and mix with gangue in the reject zone, resulting in Cr₂O₃ recovery of below 11.5%. Increasing the slope to 1° enhanced stratification behind the riffles, promoting more efficient transport of chromite particles to the concentrate zone and yielding recovery values exceeding 20%. However, further increases in slope to 1.5° and 2.0° accelerated water flow, displacing additional gangue minerals toward the reject zone. Although these steeper slopes produced higher concentrate grades, the separation bands became narrower, rendering the process more sensitive to minor variations during concentrate collection, which could significantly affect final concentrate quality.

3.2.4. Effect of Pulp Density on Chromite Recovery

Maintaining the appropriate solid–water balance is crucial for effective gravity concentration. Pulp density, expressed as the percentage of solids in a solid–liquid mixture, directly affects the slurry viscosity. When pulp density exceeds the optimal value, increased viscosity reduces fluidity, preventing most minerals from stratifying behind the riffles. Conversely, low pulp densities allow fine heavy particles to be carried with the lighter particles toward the reject zone. As a result, gravity concentration equipment operates most efficiently within the narrow range of pulp densities, and even slight deviations from this range can significantly reduce separation efficiency [25]. This part of the study investigated the effect of varying pulp densities on chromite recovery from tailings. Tests were conducted with pulp densities ranging from 1100 to 1400 g/L, while previously established optimum values for stroke, stroke length, and slope were maintained. The tested variables and corresponding results are summarized in

Table 8. Variables used to determine the optimum pulp density.

Variables	Values
Strokes (rpm)	320
Length of strokes (mm)	13
Deck slope (°)	1.0
Feed rate (kg/h)	500
Wash water flow rate (tph)	7
Pulp density (g/L)	1100–1400

and presented graphically in Figure 10.

The results indicate that at a low pulp density (1100 g/L), chromite fines were largely carried down the slope with the water, predominantly entering the waste zone due to excessive water content, which limited the recovery of marketable-grade concentrate. Increasing the pulp density enhanced Cr₂O₃ recovery, reaching an optimum at 1300 g/L. However, further increasing the pulp density to 1400 g/L caused chromite particles to be displaced toward the waste and middling zones due to enhanced mechanical carrying effects, ultimately reducing the overall yield. These findings highlight the critical role of pulp density in balancing mineral transport and separation efficiency, demonstrating that both under- and over-dilution of the pulp can adversely affect Cr₂O₃ recovery.

3.2.5. Effect of Feed Rate on Chromite Recovery

For mineral processing equipment to be economically viable, it must operate at optimal efficiency and capacity. Operating below the ideal capacity results in lost economic value, reduced production of marketable product per unit time, and higher operating costs, whereas exceeding optimal capacity decreases effectiveness and separation efficiency. Consequently, achieving the highest possible Cr₂O₃ yield requires operating the shaking table at its optimum feed rate. Based on the optimum variables determined earlier in this study (

Table 9. Variables used to determine the optimum feed rate.

Variables	Values
Strokes (rpm)	320
Length of strokes (mm)	13
Deck slope (°)	1.0
Pulp density (g/L)	1300
Wash water flow rate (tph)	7
Feed rate (kg/h)	500–1200

), enrichment tests were carried out with feed rates ranging from 500 to 1200 kg/h to both maximize processing capacity and identify the most effective feed rate for shaking table operation.

The optimal conditions determined in this study were initially applied at a feed rate of 500 kg/h, with the wash water flow rate maintained at 7 tph, based on existing plant conditions and preliminary tests. The results showed that increasing the feed rate from 500 kg/h to 800 kg/h substantially enhanced Cr₂O₃ recovery (Figure 11). At a 500 kg/h, Cr₂O₃ recovery was approximately 22%, whereas increasing the feed rate to 800 kg/h increased recovery to over 28%. This improvement can be attributed to the fixed wash water rate of 7 tph, which was relatively high for the lower feed rate and caused chromite particles to be washed away into the reject zone. At the higher feed rate, gangue minerals were displaced toward the middling and reject zones, resulting in more efficient separation and higher Cr₂O₃ recovery values.

However, further increasing the feed rate to 1200 kg/h caused chromite particles to be transported along with gangue minerals without effective separation, resulting in their accumulation in the middling and reject zones and preventing the production of a marketable-grade concentrate. At these higher feed rates, the shaking table riffles, designed to create hindered settling conditions, were unable to retain the chromite particles effectively, allowing them to escape to the reject zone. Combined with the insufficient wash water relative to the increased feed, this led to a sharp decline in Cr₂O₃ recovery, dropping below 10%. These findings highlight the critical balance required between feed rate and wash water flow to maintain effective separation and maximize recovery in shaking table operations.

3.2.6. Effect of Wash Water Flow Rate on Chromite Recovery

Insufficient wash water flow hinders the effective removal of low-density gangue minerals from high-density valuable minerals, resulting in low-grade concentrate. Conversely, excessive wash water can displace high-density minerals toward the low-density fraction, leading to significant losses of valuable material to waste and reducing Cr₂O₃ recovery [28]. In this part of the study, the previously established optimal operational parameters were maintained (

Table 10. Variables used to determine the optimum wash water flow rate.

Variables	Values
Strokes (rpm)	320
Length of strokes (mm)	13
Deck slope (°)	1.0
Pulp density (g/L)	1300
Feed rate (kg/h)	800
Wash water flow rate (tph)	6–9

), while the effect of varying wash water flow rates (6–9 tph) on Cr₂O₃ recovery was systematically evaluated. The results (Figure 12) emphasize the critical balance between wash water flow and mineral density required to achieve optimal separation and maximize chromite recovery.

The experimental results demonstrated that the wash water flow rate has a substantial effect on the efficiency of the separation process. At the lowest flow rate of 6 tph, water was insufficient to achieve effective separation, resulting in poorly defined concentrate, middling, and reject zones on the shaking table surface and low overall recovery. Under these conditions, careful control was required to prevent mixing olivine with chromite and to produce a marketable-grade concentrate, further limiting recovery. Increasing the wash water flow rate to 7 tph improved the separation: lighter gangue minerals were effectively displaced toward the middling and reject zones, enabling easier attainment of marketable concentrate grades and higher Cr₂O₃ recoveries compared to the lower flow rate. However, further increasing the wash water flow to 8–9 tph caused chromite particles to be carried into the middling or reject zones due to excessive water, thereby reducing separation efficiency and causing Cr₂O₃ recovery to drop sharply below 15%. These results underscore the critical need to balance wash water flow with feed characteristics to optimize both recovery and concentrate quality in shaking table operations.

3.3. Comparison of the Custom-Designed Shaking Table Performance with Existing Plant Conditions

Previous results (Section 3.1) indicated that under existing plant conditions, the 1st and 2nd floor shaking tables produced a maximum chromite concentrate of less than 1% of the feed, corresponding to a maximum Cr₂O₃ recovery of 8%. Approximately 65% of the feed, classified as middlings in the first step, was subsequently reprocessed using the 3rd and 4th floor shaking tables to recover marketable-grade concentrate and minimize chromite losses. This part, therefore, aimed to compare the performance of the existing plant shaking tables with that of the custom-designed shaking table in terms of Cr₂O₃ recovery. The optimized operating parameters for the custom-designed shaking table and the corresponding results are summarized in Table 10 and presented in Figure 13, respectively.

The primary objective of the enrichment processes was to obtain marketable chromite concentrate from the plant tailings. Under existing plant conditions, the 1st and 2nd floor shaking tables achieved a maximum Cr₂O₃ recovery of only 8%, indicating substantial chromite losses. This low recovery can be attributed to mechanical fatigue from prolonged operation and design limitations of the existing tables, which restricted their separation efficiency. In contrast, the use of a custom-designed shaking table with optimized operating parameters significantly enhanced Cr₂O₃ recovery, increasing it to over 27%. This improvement highlights the importance of equipment optimization in tailings processing and demonstrates the potential economic gains without additional mining or grinding. Moreover, the proportion of middlings sent to the 3rd and 4th floor shaking tables was reduced by 40%, reflecting more efficient separation and lower material recirculation, which collectively enhance overall plant performance.

When evaluating the recovery of marketable-grade chromite concentrate from the middlings of the 1st and 2nd floor tables, the 3rd and 4th floor shaking tables under existing plant conditions achieved only approximately 17% Cr₂O₃ recovery. In contrast, reprocessing these middlings using the custom-designed shaking table substantially increased recovery to over 40%. This result underscores the critical role of tailored equipment design in capturing chromite particles that would otherwise remain unextracted and emphasizes the potential of customized solutions to improve mineral utilization and promote more sustainable processing practices.

3.4. Processing of the Tailings Classified into Narrow Particle Size Ranges: Effect of Feed Particle Size on Chromite Recovery Using a Custom-Designed Shaking Table

Gravity concentration equipment, including jigs, spirals, and shaking tables, typically operates on feed materials classified into narrow particle size ranges to minimize the influence of particle size on recovery [24]. Each method achieves optimal efficiency only when operated under conditions specifically adjusted for the target particle size fraction. Consequently, careful optimization of both operational parameters and particle size characteristics is essential to achieve peak separation performance. While the optimal particle size ranges for gravity separation are broadly recognized, the quantitative relationship between particle size and separation efficiency remains insufficiently explored.

In this phase of the study, tailings enrichment was conducted using a custom-designed shaking table after classifying the feed into four narrow particle size fractions: −0.500 + 0.355 mm, −0.355 + 0.250 mm, −0.250 + 0.150 mm, and −0.150 mm. The previously established optimal operational variables (

Table 11. Optimized operating variables for the custom-designed shaking table.

Variables	Values
Strokes (rpm)	320
Length of strokes (mm)	13
Deck slope (°)	1.0
Pulp density (g/L)	1300
Feed rate (kg/h)	800
Wash water flow rate (tph)	7

) were applied consistently across all fractions to evaluate the effect of particle size on Cr₂O₃ recovery. The results clearly demonstrated that size-dependent particle behavior governs the efficiency of chromite–gangue separation (Figure 14). Effective separation was driven by hindered settling and selective layering, which were most pronounced in the −0.250 + 0.150 mm size fraction. Within this optimal range, the significant specific gravity differential between chromite and the lighter gangue (e.g., olivine, serpentine) promoted efficient separation, yielding high recoveries while maintaining concentrate quality, with peak Cr₂O₃ recovery exceeding 50%. In contrast, the −0.355 + 0.250 mm fraction yielded a lower recovery of approximately 45%, attributed to incomplete stratification and localized turbulence that disrupted consistent layer formation. Recovery sharply declined to ~16% for the −0.150 mm fraction due to the process inefficiencies inherent to fine particle processing. As fine or slow-settling particles remained suspended, they increased pulp viscosity. This elevated viscosity directly impaired separation in two ways: it enhanced the entrainment of gangue minerals, preventing their clean rejection, and it reduced the visual clarity of mineral bands on the deck, hindering precise mechanical separation [24]. Similarly, the coarsest fraction (−0.500 + 0.355 mm) exhibited poor separation due to hindered particle mobility and erratic movement on the deck surface. These findings corroborate classical stratification theory [29], demonstrating that gravity separation efficiency is governed by a complex interplay between particle characteristics (size, shape, and density) and hydrodynamic conditions [30]. Ultimately, particle size was identified as the most influential parameter controlling Cr₂O₃ recovery on the custom-designed shaking table.

4. Summary and Conclusions

This study clearly demonstrates the substantial potential of a custom-designed shaking table to enhance chromite recovery from plant tailings significantly. By redesigning the conventional table for improved usability and operational flexibility, the research provides a practical solution to both mitigate environmental pollution associated with tailings while economically valorizing chromite minerals that would otherwise be lost—critically, without additional mining or grinding costs.

The optimized operating parameters—320 rpm table speed, 1° inclination, 13 mm amplitude, 1300 g/L pulp density, 800 kg/h feed rate, and 7 tph wash water flow—resulted in significant performance gains. Under these conditions, Cr₂O₃ recovery increased from approximately 8% to 27% for the 1st–2nd floor tables and from roughly 17% to over 40% for the 3rd–4th floor table groups. Consequently, a marketable-grade chromite concentrate was produced from tailings (average 4.15 ± 0.05% Cr₂O₃), effectively more than doubling the recovery rates achieved under existing plant conditions.

The study also emphasizes that separation efficiency is fundamentally governed by feed particle size. Classifying the tailings into discrete size fractions markedly improved Cr₂O₃ recovery. The −0.250 + 0.150 mm fraction exhibited optimal performance, driven by hindered settling and selective layering, achieving a recovery of ≥50% without compromising concentrate grade, followed by the −0.355 + 0.250 mm fraction (~44%).

Overall, this research demonstrates the effectiveness of a custom-designed shaking table and provides mineral processing plants with a technically viable and economically attractive approach to substantially enhance chromite recovery from challenging tailings. These findings support improved resource utilization and contribute to more sustainable mining operations.