Design and Application of Rapid Dewatering for Tailings Containing Sodium Silicate: Laboratory and Industrial Test Results

Focusing on the tailings containing high sodium silicate (water glass), this study conducted laboratory experiments and semiindustrial tests of flocculation and sedimentation for rapid tailings dewatering. Various reagent combinations, lime and anionic flocculant (APAM) (L&A), sulfuric acid and APAM (S&A), and calcium chloride and APAM (C&A), were tested.,e results show that (i) all the three combinations presented a good effect in the laboratory experiments, and the relatively optimum dosages were L&A of 2 kg/t & 30 g/t, S&A of 2.6 kg/t & 20 g/t, and C&A of 3 kg/t & 30 g/t. (ii) ,e overflow water of the L&A group became turbid after 10 hours, and the pH increased to 14. While, for the S&A and C&A groups, the solid content in overflow wastewater (SCOW) fluctuated over time but could bemostly controlled in 300 ppm.,e pH declined from 7 to 4 with time in the S&A group, and it was 7-8 in the C&A group. (iv) ,us, a backfill system with dewatering technology based on the C&A combination was designed. In the field application, the overflowwater was recycled with the SCOWof less than 250 ppm and the pH of 7.5, while the underflow concentration was greater than 70 wt.%. ,e achievement provides a way to utilize the tailings containing sodium silicate effectively and environmentally, which will reduce the tailings dam on the Earth.


Introduction
Tailings are solid wastes generated from mineral processing. With the increasing demand of economic development for mineral resources, an astounding amount of tailings, approximately 1.2 billion tonnes per year in China, have been produced that have to be managed [1]. As a detrimental industrial waste, the improper disposal of tailings leads to some issues, such as floods, the proliferation of metal elements harmful to human health, the occupation of land, and the degradation of natural environment. In this case, the utilization of tailings for cemented tailings backfill (CTB) transported to underground has become widely used in many mines [2][3][4]. It has been also proven that CTB can not only reduce the cost of operation in a mine, but also provide an environmental way to treat this solid waste [5,6]. e quality and performance of cemented tailings backfill are affected by many factors [3,[6][7][8][9][10][11][12][13][14], including the mass fraction. In industrial application, it is necessary to raise the mass fraction of CTB so that the setting time of the backfill can be shortened and the strength of backfill can be improved significantly [15]. In this case, tailings dewatering has become a crucial issue affecting the application of backfill technology in a mine [16]. For the requirement of continuous backfill technology, rapid dewatering by flocculation and sedimentation (FS), agglomerating the fine particles with polymeric flocculants, has been widely used in backfill mines [17][18][19][20]. Despite the rapid dewatering function, FS also plays a role in purifying the solid content in overflow wastewater (SCOW) [21], which makes it possible to recycle the overflow water and protect the environment.
Xianglushan Tungsten Mine (XTM), located in Jiangxi Province of China, is the second largest tungsten-producing mine in China, with a production capacity of 800,000 ton per year. A backfill system with a capacity of 80 m 3 /h was constructed in 2011. However, many muddy foams were produced ( Figure 1) on the liquid surface even after adding different flocculants, leading to high SCOW. e overflow water contained high solid that could not be recycled and then transported to the river nearby, which contaminated the environment seriously.
us, as a result, the backfill system was banned by the local environmental protection department. In this case, tailings dewatering has been a troublesome issue for XTM to realize the backfill technology and sustainable mining. Actually, this issue occurred in other tungsten mines nearby, such as Dajishan Tungsten Mine (Jiangxi Province of China) and Pangushan Tungsten Mine (Jiangxi Province of China), which have severely restricted sustainable enterprise development.
rough surveys, it was found that in the flotation plant in these mines, a large amount of dispersants and sodium silicate (water glass) [22] are added, as shown in Table 1. Sodium silicate is a dispersant widely selected in the traditionally fatty acid flotation process of tungsten minerals [23]. It also plays an important role in the flotation of scheelite and calcite [24]. As a dispersant, sodium silicate makes the solution a kind of colloid that has adhesion stress. en, the fine particles in the tailings slurry will form a very stable colloidal dispersion to prevent the adsorption between particles and flocculants [25,26], which should be the key issue affecting the effect of FS.
Setting time is a key parameter for sustainable tailing operations and thereby paste backfill production. Larger underground stopes need greater filing rates for paste backfill production.
us, chemicals that accelerate the settlement time of tailings must be chosen for nonstop backfilling. Previous researches presented effective schemes to treat mineral processing wastewater containing sodium silicate. Among them, lime is the most widely used for purifying mineral processing wastewater [27]. Additionally, as substitutes for lime, the usage of industrial by-products and cement kiln dust (CKD) was tested to treat the wastewater [28]. e chemical equation is as follows: Na 2 SiO 3 + Ca(OH) 2 = CaSiO3↓ + 2NaOH. In addition, the strong acid was also selected as an agent in the treatment of mine drainage and achieved a good effect [29]. e chemical equation is as follows: Na 2 SiO 3 + 2H + ⟶ 2Na + + H 2 SiO 3 ↓. However, they are both used in a concrete container with a large diameter for a static process, and the requirement of underflow concentration is about 38-45 wt.%, lower than that of 65-75 wt.% in CTB. Also, the setting time used was more than 8 hours, which could not satisfy the continuous operation of CTB. Anyway, the achievements above guide the research about treatment of sodium silicate in this paper. Furthermore, inspired by the previous outcomes, another strong electrolyte, CaCl 2 , was taken consideration to treat the sodium silicate in this study. erefore, in this paper, focusing on the tailings slurry containing high sodium silicate in XTM, some laboratory experiments and semi-industrial tests were conducted. Lime, sulfuric acid, and calcium chloride were selected to remove the sodium silicate before FS with polymeric flocculant. rough evaluating the indexes, such as the qualitative settling velocity, the handling capacity, SCOW, and the underflow concentration, a feasible and reliable scheme for tailings dewatering was proposed. Finally, the industrial application of this craft has been presented and evaluated.

Tailings Slurry.
In the laboratory experiments, the tailings slurry used were received from the tailings discharge outlet in XTM.
e tailings slurry samples with concentrations of approximately 20 wt.% were collected every half an hour for 8 hours when the ore-dressing plant was in operation.
e properties (especially the sodium silicate content) of the original tailings slurry could thus be maintained. In semi-industrial tests, the tailings slurry was directly pumped into a mixer located at the semi-industrial test site through a pipeline. e particle size distribution and other physical properties of tailings are presented in Figure 2 and Table 2. As described in Figure 2, the particles smaller than 19 μm account for 30.40 wt.%, which makes it impossible to realize tailing dewatering by self-gravity, even if ignoring the sodium silicate [30,31].

Reagents for the Treatment of Sodium Silicate (RTS).
Considering the convenience and cost of obtaining the materials, lime, sulfuric acid, and calcium chloride were selected to treat the sodium silicate in this study. Lime was provided by XTM itself. e sulfuric acid and calcium chloride were manufactured by YiCheng Co., Ltd., Hunan Province of China, and their prices are 50 Yuan/L (98%) and 3000 Yuan/t, separately.

Polymeric Flocculant.
Generally, APAM (anionic polyacrylamide) is a commonly used flocculant for tailings dewatering [32][33][34]. en, after some trial tests comparing the effect of different flocculants, an APAM with a molecular weight of 18 million (manufactured by BASF SE, Germany) was selected in this paper. e price of APAM is 7000 Yuan/t.

Experimental Methods.
e experiments in this study consisted of a laboratory experiment and a semi-industrial test. e laboratory experiment was a static test, and the purpose was to determine the optimum dosages of different reagents through observation of the qualitative settling velocity. en, the semi-industrial test was aimed at verifying the effect by detecting various crucial indexes.

Laboratory Experiment.
e laboratory experiment was divided into the following three groups: (i) adding the lime and APAM (L&A); (ii) adding the sulfuric acid and APAM (S&A); and (iii), adding the calcium chloride and 2 Advances in Materials Science and Engineering APAM (C&A). e experimental method was conducted as follows: (1) Preparation of reagents. A 10 wt.% sulfuric acid solution was prepared in 2000 mL beakers. e lime and calcium chloride were prepared and measured with an electronic scale when used. e tailings slurry in the lifting mixer was pumped into a 1000 mL measuring cylinder by a small peristaltic pump (BT-300EA, manufactured by Hengjie, Shanghai of China). Simultaneously, the designed RTS was added. en, a stirrer mixed the mixture in the cylinder with a rubber mesh for 30 s after adding the APAM solution. Finally, the cylinder was placed on a table to monitor the height of the solid-liquid interface. e designed dosage of the reagents and flocculants can be observed in the results.

SEMI-Industrial Test
(1) Process of Semi-Industrial Test. A system of tailings dewatering based on a designed semi-industrial apparatus was constructed, as shown in Figure 3. As indicated in the figure, the tailings slurry from the ore-dressing plant of XTM was transported to a buffered mixer (XB-1000, Shicheng, Jiangxi) by self-flowing. e RTS was transported to the mixer at the same time. e RTS and APAM solutions were prepared in two flocculant preparation systems in the reagent preparation room. en, the treated tailings slurry and APAM solution were concurrently pumped to a feed tank located at the top of a deep-cone thickener (diameter × height � 1000 mm × 8000 mm, manufactured by Feny Co. Ltd., Hunan Province of China) by a hose pump (KP 300, Jieheng, Chongqing of China).
(2) Indexes Detection. SCOW and pH. e tailings dewatering is a long-term and dynamic process in the industrial applications. erefore, the variation of SCOW and pH with time should be considered. SCOW was detected through a vacuum filter (2BE, Zhongxing Co., Ltd., Beijing of China), as shown in Figure 4(a). According to the water recycling requirement studied in XTM before, the SCOW must be controlled to be less than 300 ppm [35]. Furthermore, the pH was detected with pH test pen.
Handling capacity. It is necessary to determine the best handling capacity first, which is a key parameter affecting the    Advances in Materials Science and Engineering 3 SCOW and the design of the deep-cone thickener in industrial applications [17]. e handling capacity can be calculated through the feeding rate of tailings slurry as follows: where T h denotes the HC, t·m − 2 ·h − 1 ; m t is the tailings quality in a hour, t/h; ρ is the density of the tailings slurry, 1.28 t/m 3 ; Q denotes the feeding rate of the tailings slurry, m 3 /h; C is the mass fraction of the tailings slurry, wt.%; and r depicts the radius of the deep-cone thickener, 0.5 m, so the area, A, is 0.785 m 2 . Underflow concentration. e underflow was collected and placed in a given container with certain weight and volume, as shown in Figure 4(b). After drying until the weight was stable, the weight of dry tailings was obtained. en, the underflow concentration could be determined.

Static Dewatering in Laboratory Experiments
3.1.1. Effect of L&A. Figure 5 describes the height variation of the solid-liquid interface with time for different dosage combinations with lime and APAM. It can be observed from this figure that both the lime and APAM have an optimum dosage that makes the solid-liquid interface settle the fastest. As depicted in Figure 5(a), when the APAM was kept at 30 g/ t, the FS shows the best effect when the additional amount of lime was 2 kg/t. Additionally, with the increase in lime consumption, the settling velocity decreased. is is because that the solubility of Ca(OH) 2 is not significant, and the excessive Ca(OH) 2 is suspended in the liquid, likely affecting the effect of FS.
In addition, when the lime was kept at 2 kg/t, an APAM addition of 30 g/t made the effect of the settling velocity of the solid-liquid interface the fastest, as shown in Figure 5(b). Previous publications have determined that the dosage of flocculants has a best value for a given material [36,37]. e optimum combination of L&A is lime of 2 kg/t and APAM of 30 g/t. is combination was not extremely accurate but rather used as a reference for industrial application. Figure 6 presents photographs of the FS effect. Experiences indicate that excessive dosages (more than 30 g/ t) of admixtures can create some problems: blockages and abrasiveness when there is a distance between thickener and backfill plant. us, selection of proper admixtures is vital for sustainable operations.

Effect of S&A.
Similar to the lime, sulfuric acid is also used as a reagent to treat the sodium silicate in the tailings slurry. e chemical reaction is as follows: Na 2 SiO 3 + H 2 SO 4 � Na 2 SO 4 + H 2 SiO 3 ↓. e relative molecular masses of Ca(OH) 2 and H 2 SO 4 are 74 and 98, respectively; thus, according to the optimum dosage of lime at 2 kg/t, the sulfuric acid consumption in this study was kept as 2 kg/t × (98/74) ≈ 2.6 kg/t. Figure 7(a) depicts the height variation of the solidliquid interface with time when the addition of sulfuric acid was 2.6 kg/t and the APAM changed from 10 g/t to 40 g/t.
is figure indicates that when the dosage of APAM is 20 g/t, the settling velocity presents the best effect. en, with the increase of APAM, the settling trend variation is not obvious and even lower. e best combination of S&A is sulfuric acid of 2.6 kg/t and APAM of 20 g/t, approximately. In this case, the experimental photograph is shown in Figure 7(b).

Effect of C&A.
Likewise, the relative molecular mass of CaCl 2 is 110; thus, the calcium chloride consumption was 3 kg/t. Additionally, in this case, Figure 8(a) shows the height variation of the solid-liquid interface with time when the addition of calcium chloride was 3 kg/t and the APAM changed from 10 g/t to 40 g/t. It can be concluded that the best combination is approximately calcium chloride of 3 kg/t and APAM of 30 g/t. Figure 8(b) presents the effect at this combination.

Best Handling Capacity (BHC).
e reagent dosage was kept consistent with the results of the laboratory experiments, while the feeding rate of tailings slurry was changed from 1.6 m 3 /h to 2.6 m 3 /h to detect the SCOW variation.
at is, the handling capacity changed from 0.39 t·m − 2 ·h − 1 to 0.64 t m − 2 ·h − 1 . is process lasted for only one hour. Figure 9 describes the variation of SCOW with varying handling capacity.
is figure indicates that the SCOW increases with the handling capacity. According to the water recycling requirements, the SCOW must be controlled to be less than 300 ppm. In this case, it was concluded that the BHC values of L&A, S&A, and C&A groups were 0.54 t·m − 2 ·h − 1 , 0.49 t·m − 2 ·h − 1 , and 0.59 t·m − 2 ·h − 1 , respectively. It is obvious that the S&A group presented the worst handling capacity in FS. As we know, unlike CaSiO 3 , H 2 SiO 3 is a flocculant precipitate in the water, which may more or less affect the adsorption between tailings and flocculants [38,39]. Figure 10 comprehensively depicts the change of SCOW and pH with time under the determined reagents dosages and BHC. Figure 11 presents the FS effect of different combinations on the liquid surface after 10 hours. Analysing these figures, the following conclusions are suggested:

Variation of SCOW and pH with Time.
(1) For the L&A group, the SCOW increased gradually with time. After 10 hours of continuous operation, the detected SCOW increased beyond the criterion higher than 300 ppm. And, many muddy foams gathered on the liquid surface at this time, as Figure 11(a) shows. In addition, the pH varied from 11 to 14, which meant that the  Advances in Materials Science and Engineering overflow water was strongly alkaline because OHwas produced after adding lime. In this alkaline environment, some of the remaining foaming agent and foam stabilizer used for flotation would produce stable foams [40]. ese foams adsorbed some tailings particles to affect the FS effect and increase the SCOW. Stable muddy foams would also affect the thickener underflow and thereby the behaviour of cemented backfill produced from these tailings  Advances in Materials Science and Engineering (2) For the S&A and C&A groups, the SCOW fluctuated over time but was mostly controlled at 300 ppm. Furthermore, the pH declined from 7 to 4 with time in the S&A group, while the pH remained at 7-8 in the C&A group. It can be observed from Figures 11(b) and 11(c) that the overflow in these groups showed high clarity. However, the low pH in the S&A group required anticorrosion treatment for devices and a neutralization system for pH. ese additions will increase both the initial investment and operating cost. Figure 12 describes the underflow concentration variation for different combinations with time. It can be observed that the underflow concentration increased with time and remained mostly higher than 70 wt.% after 10 or 12 hours for the S&A group and C&A group, except for a trend of decline in the C&A group after 12 hours. Actually, the final concentration of tailings mainly depends on its own characteristics and the parameters of the deep-cone thickener [41,42] regarding waste of time. However, in the continuous CTB backfill, the rapid dewatering is required to ensure the high underflow concentration in a limited time. As studied above, the effect became worse after 10 hours, which then affects the settling velocity and, as a result, the underflow concentration dropped. Table 3 compares the effect of different combinations on FS. It is obvious that the cost of the C&A group is the most expensive. However, the  other combinations require equipment to remove the foams or neutralize the pH of the overflow water. Specially, the acid environment in the S&A group is potentially harmful to people and requires devices to be treated for corrosion. ese additional crafts will increase the cost more than the higher dosage cost in the C&A group. Complex craft leads to increased difficulty in operation and control. In addition, the BSU of the C&A group is maximal, which will reduce the diameter of the deep-cone thickener. Synthetically, the C&A group was selected in XTM. Figure 13  e cemented paste will be formed in combination with the tailings, cement, and water (if needed). Finally, the paste is pumped to a stope underground, and the overflow water is stored in a recycling reservoir. Figure 14presents the overflow water and underflow concentration in the field application. e figure indicates that the overflow water showed good clarity. Actually, the SCOW detected in the field could be controlled at approximately 250 ppm, while the underflow concentration maintained a concentration higher than 70 wt.% and often reached a highest value of 73 wt.%. In addition, the overflow water has been recycled in the ore-processing craft successfully, which shows an economic and environmental effect.   Advances in Materials Science and Engineering

Conclusions
is study conducted laboratory experiments and semiindustrial tests for dewatering the tailings containing high sodium silicate (water glass). In addition, the industrial application was also evaluated. Based on the results, the following conclusions are obtained.
(1) Lime, sulfuric acid, and calcium chloride showed a good FS effect in the laboratory experiments when combined with APAM. (2) e selected BHC values of the L&A, S&A, and C&A groups were 0.54 t·m − 2 ·h − 1 , 0.49 t·m − 2 ·h − 1 , and 0.59 t·m − 2 ·h − 1 , respectively. (3) In the semi-industrial tests, the overflow water of the L&A group became turbid after 10 hours, and the pH increased to 14. For the S&A and C&A groups, the SCOW fluctuated over time but could be mostly controlled at 300 ppm. In addition, the pH declined from 7 to 4 with time in the S&A group, while the pH was 7-8 in the C&A group. (4) e C&A group (calcium chloride and APAM) was selected in the field. e backfill craft with new FS technology based on the result was designed and operated. It presented a good effect with an SCOW of less than 250 ppm and an underflow concentration of greater than 70 wt.%. e achievement of this paper provides a way to utilize the tailings effectively and environmentally, which will also reduce the tailings dam on the Earth. It needs to be note that, however, the FS with two steps is too complicated in engineering application. A new technology with one reagent should be thus studied in the future.

Data Availability
e photograph data used to support the findings of this study are all included within the article.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.