The unclassified tailings used herein were sampled from a copper mine in Sinkiang, China. The tailings size distribution was analyzed by a laser particle analyzer; the wet-sieving method was utilized simultaneously for particles >75 μm. As shown in Fig. 4, the tailings have a reasonable fine particle content (<20 μm), which is essential for the preparation of CPB [37–38]. Generally, a proportion of at least 15wt% fine tailings is required for retaining sufficient water to produce the CPB .
The tailings’ specific gravity is 2.662. The coefficient of uniformity CU (18.36) and coefficient of curvature CC (1.62) revealed that the tailings contain a wide range of size distribution and have representations of all dimension fractions. The sampled tailings could be classified as a relatively well-graded material . The chemical characteristics of the tailings (Table 1) were determined by X-ray fluorescence. In the thickening experiments, the flocculant utilized was the synthetic anionic polyacrylamide, Magnafloc 5250, because of its superior performance in practical backfilling operation at this mine.
SiO2 CaO Al2O3 Ag MgO S As Cu Pb Au 65.68 9.30 6.28 1.61 1.42 0.40 0.06 0.05 0.04 0.03
Table 1. Main chemical composition of the sampled tailings
Conventional laboratorial settling experiments are generally conducted to measure cylinders due to its simplicity. However, the results are of limited guidance for practical thickening design. Hence, a continuous laboratorial thickening system was built, as illustrated in Fig. 5. The settling column was 100 mm in diameter and 1000 mm in height, which were designed to minimize wall effects without excessively increasing the feed slurry quantity on a bench-scale and continuous basis .
Figure 5. Schematic of the laboratorial thickening system. H1, H2, and H3 represent various circulation heights.
Initially, nine sets of sedimentation (without discharging) were performed as a simulation of the pauses in practical thickening to determine the subsequent experimental factors, namely, the feeding content, feeding flow rate, and rake speed, as shown in Table 2. The factors were selected in accordance with practical operations and the similarity criterion. As the tailings were sampled from the mineral processing plant at a content of 25wt%–30wt% and were generally diluted to 10wt%–20wt% before mixing with the flocculant, the feeding content was designed to be 10wt%, 15wt%, and 20wt%. In terms of the feeding flow rate, the calculation of the solids’ throughput per unit cross-sectional area of the experimental column proposed the range of feeding flow rate to be 156.5–2355.7 mL·min−1; this parameter was set to 400, 600, and 800 mL·min−1 in this experiment. In field practice, the rake speed is typically 0.14–0.3 r·min−1 within paste thickeners. In accordance with the speed scale, this parameter was calculated to be 0.5–11.5 r·min−1 in the experimental device herein. Therefore, it was set to 2, 6, and 10 r·min−1.
T-1 10 400 2 90 75.9 T-2 10 600 6 70 73.1 T-3 10 800 10 45 74.5 T-4 15 400 6 60 72.6 T-5 15 600 10 40 70.6 T-6 15 800 2 25 71.7 T-7 20 400 10 44 69.5 T-8 20 600 2 30 69.4 T-9 20 800 6 22 72.1 Note: Underflow concentration refers to the underflow solid content at 12 h from the beginning of feeding.
Table 2. Parameters and results in the sedimentation tests
For each test, 4 kg of tailings was fed at corresponding contents and flow rates for corresponding feeding times. Then, the feeding was ceased, and the underflow concentration at 12 h from the beginning was recorded (Table 2). Throughout each test, the rake was operated at a corresponding speed. The flocculant was diluted to a content of 0.01wt% and utilized at a dosage of 20 g·t−1 in the pilot-scale thickening process. The results show that the underflow solid content of T-1 tends to be high and likely to cause rake failure. Therefore, the parameters of T-1 were used as the basis to perform subsequent circulation experiments.
Before the circulation test, the tailings slurry was thickened in accordance with the T-1 parameters. This thickening process was performed by pumping the tailings slurry of 10wt% solid content at a pump flow rate of 400 mL·min−1. After 90 min, 4 kg of tailings was fed into the thickening system. Then, the feeding was ceased, and the circulation began by returning the underflow to higher levels, as shown in Fig. 5. In the thickening process, the rake speed was set to 2 r·min−1.
Nine sets of circulation tests were conducted. The two parameters (circulation height and flow rate) were selected in accordance with practical experience in their relationship with the mud bed and feeding and discharging rates. The heights of 140, 240, and 340 mm were set (H1, H2, and H3, respectively, as shown in Fig. 5); for each height, three flow rates, i.e., 200, 600, and 1000 mL·min−1 were tested (Q1, Q2, and Q3, respectively). As soon as each thickening process was completed, a circulation test was conducted, and each circulation was operated for 180 min. The solid content of the underflow was measured every 20 min. The rheological properties of the resulting underflow after circulation were measured.
The resulting underflow was sampled and tested by using the Brookfield R/S+ vane rheometer. Vane rheometers are widely accepted as a preferred approach for the rheological characterization of concentrated suspensions, especially for yielding materials . This method has two distinct advantages in yield stress measurement as it enables the yielding to occur within the suspension and minimizes the disturbance caused during immersion of the vane. The instrument could be used in two modes: shear stress-controlled and shear rate-controlled modes. In the latter, flow curves based on the obtained stress to rate data could be analyzed by applying the constitutive equations (linear relationship: Bingham model, nonlinear relationship: Herschel–Bulkley model and Casson model) or through extrapolation to the zero shear rate . In this study, the controlled shear rate mode was adopted and set at a shear rate of 0–120 s−1. The obtained flow curves were analyzed with the Bingham model because the thickened tailings showed a nearly linear rheological relationship, and this model was commonly utilized for paste characterization .
Changes in underflow solid fraction and yield stress in paste thickeners by circulation
12 November 2019
Revised: 1 September 2020
Accepted: 2 September 2020
Available online: 3 September 2020
Abstract: The trouble-free and efficient operation of paste thickeners requires an optimal design and the cooperation of each component. When underflow discharging is suspended, alleviating the vast torque that the remaining solids within the thickeners may place on rakes mainly lies in the circulation unit. The mechanism of this unit was analyzed, and a mathematical model was developed to describe the changes in underflow solid content and yield stress. The key parameters of the circulation unit, namely, the height and flow rate, were varied to test its performance in the experiments with a self-designed laboratorial thickening system. Results show that the circulation unit is valid in reducing underflow solid fraction and yield stress to a reasonable extent, and the model could be used to describe its efficiency at different heights and flow rates. A suitable design and application of the circulation unit contributes to a cost-effective operation of paste thickeners.