Estimation of Energy Consumption for Concentrate Process of Tungsten Ore towards the Integration of Renewable Energy Sources in Mongolia

Abstract

It is important to estimate the energy required in ore processing to select the most affordable and efficient energy system for the integration of renewable resources into the mining industry. In the present work, the energy consumption for the concentrate of tungsten ore in Mongolia was theoretically predicted based on operational variations (particle size and the hardness of the tungsten ore) and different equipment. The energy was in the range from 0.48 to 1.32 kWh/t for the crushing stage, and a cone crusher was more suitable than a jaw crusher due to the particle size of feed material and product. The required energy in the grinding stage was from 6.22 to 11.88 kWh/t using a SAG mill or from 3.04 to 7.39 kWh/t using a ball mill. The further separation by a flotation consumed 4.83 kWh/t or by a shaking table consumed 1.29 kWh/t. The maximum energy consumption per hour for the whole process was estimated to be 2–3 MW, which was better to integrate with a hybrid renewable energy system. The sizing method Power Pinch Analysis was used to estimate the electric supply based on the combination of wind, biomass and solar resources, which was sufficient for the demand from the predicted range of energy.

1. Introduction

Tungsten is a critical metal with a wide range of use in industrial fields such as thermo-emission technology, machine and engine construction, steel and superalloys chemical industry, laser technology, defense and the aerospace industry [1,2]. The most significant application of tungsten is the essential component in hard metals (cemented carbides), which accounts for 54%–72% of global tungsten production [1]. It is challenging to increase the production of tungsten to afford the growth of global demand due to its high supply risk and heavy dependency on distribution by China. It should consider production from resources outside of China, and Mongolia is one of the potential countries, which is the world’s fourth largest producer of tungsten in 2020 [3]. However, the mining operation in Mongolia is limited due to the remoteness of mine locations and instability of conventional energy resources. The solution is to develop a suitable concentrate process of tungsten ores based on the power supply from renewable energy resources, which is an attractive option for the remote and off-grid areas with available resources (sun, wind or geothermal) [4].

Mining is a complex and energy-intensive process, which consumes ~38% of global industrial energy use and 11% of global energy use [4]. The largest power demand in mining comes from the concentrate process of ores (accounting for 44% of total power), and the grinding spends 40% while other crushing and separation requires 4% of the total power [5]. The significant energy consumption and dependence on the fossil fuel in mining leads to high emission of greenhouse gases, climate change and other harmful impacts on the environment. Renewable energy is a potential and sustainable replacement to integrate into the mining operation with lower carbon emissions and pollution issues. It is important to design the ore processing regarding related operational variables and to estimate the energy consumption to select the most affordable and efficient renewable energy system. The specific energy required for the concentrate process is possibly determined using the theoretical approach or simulated method. For example, the specific energy for concentration of copper ore in Thanatia was 157.5 MW/t-Cu, based on the relationship between the energy consumption and the hardness of the ore (represented by the Bond index, 14 kWh/t) and the reduction ratio of the process (480, feed and final size of product 26,403 and 34 μm, respectively) [6]. In the case of gold ore with the grade 2.72 g/t-Au, the energy consumption for process of comminution, flotation and gravity concentration was found to be 18 kWh/t-ore with the Bond index 15 kWh/t, the feed size 245,631 μm, and final size of product 75 μm [7]. The specific energy for concentration of iron ore was estimated to be ~889 kWh/t-Fe for processes including crushing, grinding, re-grinding, and reverse flotation with the Bond index 14 kWh/t and the reduction ratio of the process 436 [8]. The energy consumption for concentration of tungsten ore can be evaluated using the similar method; however, it is more appropriate to use the theoretical approach in the present study since there are difficulties in collecting data from the actual processes in Mongolia. The calculation can be obtained from the suitable energy equations regarding the design of the concentrate process, the type of equipment, and the operational variations in the ores.

Determination of energy consumption for mining processes in advance is important to evaluate the balance between the electric demand and supply before installation of required equipment and energy systems for the concentrate process of tungsten ore in Mongolia. In the present work, the specific energy for the concentrate of tungsten ore was theoretically predicted based on a proposed process including comminution stage and separation stage. The energy consumption of each stage was calculated following the corresponding equation for the beneficiation equipment. The variation in operating parameters (particle size and hardness of the tungsten ore) was studied to estimate the range of required energy. Finally, the total energy consumption was used to propose the possible process for the concentrate stage and the affordable microgrid system of renewable energy, which is a preparation for the integration of efficient and sustainable resources into mining of tungsten in Mongolia.

2. Methodology

2.1. Tungsten Separation Process

The methodology to determine the energy consumption in this study is summarized in Figure 1, which includes comminution stage and separation stage. The comminution was assembled into two processes, crushing and grinding. The crushing is normally carried out in a primary crusher and secondary crusher; however, only one crusher can be applied in the case of tungsten ore, and it can be selected between using Jaw crusher or Cone crusher. Similarly, the grinding is achieved by either a semi-autogenous (SAG) mill or a ball mill. Further separation after comminution was obtained by gravity or flotation. The energy estimation was conducted based on a literature review of copper concentration processes due to a lack of information on the tungsten process in Mongolia. However, the input data considering the operating process of tungsten ore was employed since the particle sizes of grinding products and the feeding of raw materials are different. The particle size of the comminution process referring to the tungsten concentration process is given in

Table 1. Particle size of feed and product in comminution stage.

Stage	Equipment	Feed Particle Size F80 (μm)	Product Particle Size P80 (μm)
Crushing	Jaw crusher Cone crusher	150,000	10,000
Grinding	SAG mill Ball mill	10,000	350

.

2.2. Calculation of Energy Consumption

In this study, the energy consumption of the Jaw crusher, Cone crusher, and Ball mill is calculated based on the Bond’s equation (Equation (1)) [9].

$$W=10\times Wi\left(\frac{1}{\sqrt{P}}-\frac{1}{\sqrt{F}}\right)$$

(1)

where W is specific energy consumption (kWh/t), Wi is the Bond work index (kWh/t), P and F are 80% of the product and feed passing size (P₈₀ and F₈₀, μm)

The energy consumption of SAG mill is estimated using Starkey’s equation (Equation (2)) [10].

$$P={\left(P_{80}\right)}^{-0.33}\left(2.2+0.1T\right)$$

(2)

where P is the energy consumption of SAG mill (kWh/t), T is the grind time (min), and P₈₀ is the size of SAG mill ground products (µm).

The energy consumption of the flotation cell is predicted based on the conversion of the power of motor (kW, Equation (3)) to the energy (kWh, Equation (4)) [11].

$$P=\frac{2\pi Mn}{60}$$

(3)

where P is power of motor (kW), n is the rotation speed of the stirrer (rpm), M is the torque of the impeller (Nm).

$$kWh=\frac{kW\times T}{\eta }$$

(4)

where η is the efficiency of the motor, T is the running time (η is assumed to be 0.85 and T is assumed to be 1 h).

3. Results and Discussion

3.1. Crushing

Crushing can be performed using a jaw crusher or a cone crusher, and the required energy is a function of the parameters of the comminution process (the particle size of feeding material and the product) and the possible hardness of the ores (regarding Equation (1)). Generally, it is not accurate to determine a certain value of Bond work index due to the complex mineralogical compositions of the primary ores; hence, it should consider a wide range of Bond work index, as well as the parameters of the process. It is assumed that the Bond work index of tungsten ore is varied from 7 to 17 [12], the particle size of the raw material is from 100 to 200 mm, and the product size after crushing is 10 mm. The results calculated from Equation (1) interpreted the dependence of consumed energy in the crushing stage with the change in hardness (Bond work index) and the size of the feeding ore, presenting in Figure 2. The value of energy required per ton of ore for the crushing stage was in the range from 0.48 to 1.32 kWh/t, and it gradually increases with an increase in hardness of the tungsten ore and particle size of the feed material.

Both jaw crusher and cone crusher are commonly used in the comminution circuit as primary crusher and secondary crusher, respectively [12]. Although the energy consumption is the same value for two type of crushers using the Bond’s equation, it should consider the key parameters (the feeding size depending on the characteristic of the ores and the desired product size) to estimate the suitability and performance of crushers. The comparison between two crushers having similar power (~200 kW) show that the maximum size of feed material and the range of product size of the jaw crusher was significantly coarser than those of the cone crusher (

Table 2. Comparison between the jaw crusher and cone crusher.

Equipment	Jaw Crusher	Cone Crusher	Operating Parameters
Model	Sandvik CJ815	Sandvik CH440	-
Power (kW)	200	220	-
Range of CSS (mm)	150–300	8–48	10
Maximum feed size (mm)	1170	250	100–200
Capacity (t/h)	480–1160	58–336	283

) [13,14]. The selected product size for further grinding stage is 10 mm, while the range of product size of Jaw crusher is from 150 to 300 mm. Therefore, the cone crusher is more preferred for the concentrate of tungsten ore in the present work since the assumed value of the feed material and the product size are in the range of the data applied for the cone crusher. Finally, the energy consumption in crushing stage per hour is in the range from 135.84 to 373.56 kWh if the capacity is assumed to be 283 t/h.

3.2. Grinding

3.2.1. SAG Mill

The energy consumption of a SAG mill can be evaluated based on the Starkey’s equation (Equation (2)) depending on the value of grind time and the product size. The grind time was chosen from 22 to 62 min due to the complexity of the ore [10], and the product size after milling was assumed to be 350 μm. The predicted energy consumption was in the range from 6.22 to 11.88 kWh/t.

3.2.2. Ball Mill

The energy consumption using a ball mill can be estimated by the Bond’s equation (Equation (1)) depending on the hardness of ore, the feed material and the product size. The predicted energy consumption was in the range from 3.04 to 7.39 kWh/t with the Bond work index from 7 to 17, the feed size 10,000 μm and the product size 350 μm following the operating parameters from crushing stage.

Both ball mills and SAG mills are typical equipment for size reduction based on the tumbling methods and are widely used in the comminution circuit [15,16]. Ball mills have versatile applications as primary, secondary, tertiary grinding or re-grinding, operates with a high ball charge of 30% and can produce very fine products [15,16]. Meanwhile, SAG mills are mostly used for primary grinding to generate a coarser size of product and are enabled to operate with higher capacity and lower cost since they require less grinding media (4%–18% ball charge) than that of ball mills [15]. Although the grinding to a finer size of product can increase the recovery for the low-grade ore, the over-ground to very fine particle can lead to the loss of tungsten to the tailing. Therefore, the product size was controlled to be 350 μm, and both millers are suitable for the comminution process and have their distinguished advantages. However, it should further consider the energy consumption based on the brittle character of the tungsten ore and the capacity of the device. For example, if the grind time of SAG mills and the Bond work index were selected at average value 42 and 11 respectively and the capacity was assumed to be 188 t/h [17], the energy consumption per hour by SAG mills (1701 kWh) was relatively higher than that of ball mills (978 kWh).

3.3. Separation

Normally, flotation and gravity separation are applied to produce concentrate of tungsten at the final stage of the beneficiation process, or it can be the combination between floatation and gravity or gravity and magnetic separation depending on the character of the ores [1]. Gravity separation is employed for the sufficient large particle while flotation is more preferred for the finer particle. In the present work, the comminution was controlled to avoid the formation of fine particle; hence, flotation or gravity method was considered for the separation of the tungsten ore after crushing and grinding.

3.3.1. Flotation

The product with size of 350 μm was continuously sent to the flotation circuit, which consisted of two stages: (i) conjunction of rougher and scavenger cells and (ii) cleaner cells and a re-cleaner stage [6]. The operating variables and the number of cells were estimated referring to the flotation of the concentrate process for copper ore [17]. The torque of the motor and impeller speed were predicted by the power of each flotation cell. The total energy consumption per hour for flotation stage was 447 kWh and the total specific energy consumption of each flotation cell regarding their capacity was 4.83 kWh/t (

Table 3. Prediction of energy consumption for the flotation process.

Stage	Specific Energy Consumption (kWh/t)	Capacity (t/h)	Energy Consumption Per Hour (kWh)
Rougher	0.63	206	132
Scavenger	0.72	180	131
Cleaner 1	0.97	85	82
Cleaner Scavenger	1.15	43	49
Cleaner 2	0.7	47	33
Cleaner 3	0.61	31	20
Total energy consumption per hour (kWh)			447

).

3.3.2. Gravity

A shaking table was applied as gravity technique for the concentrate of tungsten ore. The shaking table does not depend on the particle size and physical properties of minerals, and the energy consumption is determined by the power consumption of the motor. Therefore, if the power consumption of the shaking table was estimated to be 1.1 kW [18], the energy consumption of the shaking table is 1.29 kWh/t.

3.4. Propose the Possible Process for Concentrate of Tungsten Ore

The energy consumption for the beneficiation of tungsten ore was investigated based on the proposed circuit containing of crushing, grinding and separation. The final process can be customized regarding the suitability and the performance of the device through each stage. The possible combinations are: (i) Process 1 (cone crusher + SAG mill + flotation); (ii) Process 2 (cone crusher + SAG mill + shaking table); (iii) Process 3: cone crusher + ball mill + flotation, and (iv) Process 4 (cone crusher + ball mill + shaking table). The characteristic feature of tungsten ore is complex mineralogical compositions; hence, the estimation of energy consumption present as a range of values for the crushing and grinding stages. Therefore, the maximum value of each stage was considered to evaluate the whole process and predict the scale of the microgrid system (Figure 3). The results in Figure 3 show that the most energy-intensive stage in the circuit is grinding, accounting for ≥63% for all processes. The largest and lowest amount of energy consumption per hour are attributed to Process 1 (3054 kWh) and Process 4 (2006 kWh). The concentrate process of tungsten with variation in the hardness and the particle size of feed and product materials can be operated by the circuit containing a cone crusher, a ball mill and a shaking table with the least value of energy consumption per hour. However, the concentrate process can be adjusted due to the actual tungsten ore properties and the dimension of the mines in Mongolia. For example, the appropriate miller, crusher and sizing techniques should be carefully designed in comminution stage to avoid overgrinding and formation of fine particles due to the brittle nature of tungsten ores (scheelite or wolframite) [1]. Further separation should be considered based on the mineral characterization of the tungsten ore. Gravity method is commonly applied for wolframite since it occurs in coarser mineralization, while flotation is more suitable for scheelite regarding the characteristics of fine grain size and low grade [1]. Therefore, the customization of the equipment should be flexible and later established after obtaining the proper information from the real mining site by sufficient mineralogical analysis.

3.5. Estimation of the Balance between Energy Demand and Supply for Concentrate of Tungsten Ore

The significant increase in metal demand in general and tungsten in particular with the decrease in those mineral grades have depicted the corresponding growth in the energy demand for the mining industry. Meanwhile, the evolution of the renewable energy market and the reduction in cost of renewable technologies has become a potential opportunity for the replacement of convention energy resources used in mining [4]. Microgrid technology (small power grid for a renewable energy system) is suitable for rural mining sites with a reasonable cost for operating [19]. There have been numerous projects to utilize renewable energy resources in the mining industry at different scales of capacity in the world (

Table 4. Utilization of renewable energy resources in the mining industry at different scales of capacity [19,20].

Country	Type of Ore	Renewable Energy	Capacity
Australia	Lithium	Wind and Solar	116.4 kW
South Africa	Coal	Solar	240 kW
Australia	Nickel	Solar	300 kW
South Africa	Chromium	Solar	1 MW
Canada	Zinc	Solar	1 MW
USA	Molybdenum	Solar	1 MW
Chile	Gold	Solar	1.1 MW
Chile	Gold	Solar	1.26 MW
Chile	Zinc gold	Wind	1.5 MW
USA	Gold	Solar	1.51 MW
Australia	Bauxite	Solar	1.7 MW
Argentina	Gold	Wind	2 MW
Canada	Nickel	Wind	3 MW
Mauritania	Iron	Wind	4.4 MW
Suriname	Gold	Solar	5 MW
Canada	Diamond	Wind	9.2 MW
Chile	Gold	Solar	10 MW
Australia	Copper	Solar	10.6 MW

) [20,21].

The average scale (<5 MW) is affordable to supply for the proposed concentrate process of tungsten ore in the present work with the maximum energy consumption per hour ~3 MW. The microgrid generation can include one renewable resource (wind, solar, fuel cells, biomass and appropriate energy storage); however, the operation of only one renewable technology has certain limitations such as the intermittent nature of the resource depending on the environmental conditions or degradation of storage systems due to load power variations. Therefore, the hybrid power generation from two or more renewable technologies is considered as an attractive alternative, which can offer high reliability, high efficiency, better power quality, and low energy storage requirements [19]. The microgrid system based on solar cells, wind power or biomass can be a promising option to use in the mining process for tungsten ores in remote sites of Mongolia, and it can predict and prepare for designing the suitable scales in advance based on the theoretical estimation of energy consumptions for the proposed concentrate processes in Section 3.4. Hybrid Power Systems (HPS) containing different renewable generators can be designed using Power Pinch Analysis (PoPA) method, and the amount of generated energy from the sizing of system can be compared with the required energy [22]. The calculation of energy supply is calculated regarding the assumed value of capacity for renewable generators (listed in

Table 5. Assumption of capacity and energy generation for renewable generators.

Renewable Source		Operating Time		Time Interval (h)	Capacity (MW)	Energy Generation (MWh)
AC	DC	From	To
Wind		0	12	12	1.5	18
Biomass		0	24	24	2	48
	Solar	8	18	10	1.5	15

), which is based on the range of capacity for each resources: wind and solar (Table 4) and biomass from 2 to 1000 MW [23]. The demand energy is the value from Process 1 and 4 (with the assumption that the equipment of both processes use DC electricity for 24 h), and the possible HPS to supply for those processes are assumed as HPS1 (combination of wind, biomass and solar) and HPS2 (combination of biomass and solar) regarding the capacity and working hour.

Calculation of cumulative storage capacity [22]

$$B_t=B_{t-1}\left(1-\sigma \times T\right)+\left((C_t\times {\eta }_c)+\frac{D_t}{{\eta }_d}\right)$$

(5)

where B_t: battery capacity (MWh); B_t−1: battery capacity at previous time interval (MWh); C_t: charging quantity (MWh); D_t: discharging quantity (MWh); σ: hourly self-discharge rate (0.00004/h); t: time (h); T: time interval (h); η_c: charging efficiency (0.9); η_d: discharging efficiency (0.9).

The calculation of energy demand and supply for Period 1, 2 and 3 of working hour for Process 1 and HPS1 shows that the amount of generated energy exceeds the required quantity (

Table 6. Estimation energy demand (Process 1: 3054 kWh = 3.054 MWh) and energy generation from renewable resources (HPS1: wind, biomass and solar).

Time	Time Interval (h)	Capacity (MW)		Energy Generation (MWh)			Energy Demand (MWh)
		AC	DC	AC	DC	AC to DC 1	DC
0
8	8 (Period 1)	3.5	0	28	0	26.6	24.4
12	4 (Period 2)	3.5	1.5	14	6	13.3	12.2
18	6 (Period 3)	2	1.5	12	9	11.4	18.3
24	6 (Period 4)	2	0	12	0	11.4	18.3

¹ AC convert to DC = Amount of AC × Rectifier efficiency (0.95).

). Although the supply cannot afford the demand for the last Period, the unused electricity in previous periods can be charged and can compensate for the deficit. The quantity of cumulative storage capacity is determined by the battery capacity including self-discharge, charging and discharging (Equation (5)); however, the hourly self-discharge rate is significantly small and there is no amount of discharge. The cumulative storage capacity can be simply determined by multiplying the sum of unused energy after Period 1 (2.2 MW), Period 2 (7.1 MW), and Period 3 (2.1 MW) with charging efficiency (0.9). Finally, the total of generated energy in Period 4 (11.4 MW) and the cumulative storage capacity (10.2 MW) from previous periods can sufficiently supply for the demand (18.3 MW). The design of HPS1 containing three renewable resources was therefore affordable for the required energy of Process 1, and the remaining electricity can be charged in the batteries for the next day operation or supply for other appliances.

The results in

Table 7. Estimation energy demand (Process 4: 2006 kWh ≈ 2 MW) and energy generation from renewable resources (HPS2: biomass and solar).

Time	Time Interval (h)	Power Source (MW)		Energy Generation (MWh)			Energy Demand (MWh)
		AC	DC	AC	DC	AC to DC 1	DC
0
8	8 (Period 1)	2	0	16	0	15.2	16
18	10 (Period 2)	2	1.5	20	15	19	20
24	6 (Period 3)	2	0	12	0	11.4	12

¹ AC convert to DC = Amount of AC × Rectifier efficiency (0.95).

indicate the insufficiency of energy supply from HPS2 for the first Period of Process 4 (15.2 MWh < 16 MWh), which required outsourced electricity generation. However, the next Period 2 generated a much higher amount of energy and can supply for not only Period 2 but also Period 3. The total of generated energy in Period 3 (11.4 MW) and the cumulative storage capacity (12.6 MW) from Period 2 could generate a surplus for the demand (12 MW). Therefore, the remaining energy as cumulative storage capacity after Period 3 (10.8 MW) could be used for the deficit of Period 1 the next operating day. The supply of HPS2 was not sufficient for only the first day of operating; however, it could compensate the demand for the next operating day by the cumulative storage capacity.

Generally, the sizing method PoPA indicates that the proposed HPSs of two or three renewable resources can be effective and affordable for energy consumption in the range of obtained value (2006 to 3054 kWh). The prediction of energy consumption in advance can be useful for establishing suitable energy systems into mining activities, which is a potential integration towards sustainable development and reduction in carbon emission as well as pollution issues. However, it should consider the customization of the process and the separation devices later depending on the mineralogical properties of tungsten ore and the available resources in the Mongolia mining site.

4. Conclusions

A process for concentration of tungsten ore in Mongolia was proposed and the energy consumption for each stage was theoretically predicted based on the variation in operating parameters (the hardness of the ore and the particle size). The energy required for crushing stage was in the range from 0.48 to 1.32 kWh/t. In the grinding stage, the predicted energy consumption was in the range from 6.22 to 11.88 kWh/t using a SAG mill or from 3.04 to 7.39 kWh/t using a ball mill. The separation could be operated by a flotation (consuming 4.83 kWh/t) or a shaking table (consuming 1.29 k kWh/t). The total process including crushing, grinding and separation using different equipment was used to estimate the energy consumption for the concentrate of tungsten, which required an amount of energy in the range of 2006 to 3054 kWh. The sizing method PoPA was used to design and estimate the amount of energy supply from HPSs containing two or three renewable energy resources. The estimation indicated the possible affordability between the range of energy consumption for concentrate processes and the generated energy from proposed HPSs. It is an important evaluation for further installation of necessary equipment and energy systems for beneficiation of tungsten ore in the Mongolia mining site.