Dense Medium Separation Process Research Articles

Method for determining the concentration of ferrosilicon in suspension for dense medium separation process

In Russia, dumps of diamond-bearing deposits are characterized by a high content of secondary modified non-magnetic kimberlite minerals of fine size classes (-0.05 mm), the presence of which negatively affects their processing by methods of dense media separation. In the present work, the possibility of determining the concentration of ferrosilicon in a continuous process of dense medium separation by measuring the magnetic properties of a suspension was investigated As a result of experimental studies, the dependences of volumetric concentration of ferrosilicon on the magnetic moment of various grades were obtained, which made it possible to measure the concentration of ferrosilicon in a continuous process of dense media separation by measuring the magnetic properties of the suspension.

RVFLN-based online adaptive semi-supervised learning algorithm with application to product quality estimation of industrial processes

Direct online measurement on product quality of industrial processes is difficult to be realized, which leads to a large number of unlabeled samples in modeling data. Therefore, it needs to employ semi-supervised learning (SSL) method to establish the soft sensor model of product quality. Considering the slow time-varying characteristic of industrial processes, the model parameters should be updated smoothly. According to this characteristic, this paper proposes an online adaptive semi-supervised learning algorithm based on random vector functional link network (RVFLN), denoted as OAS-RVFLN. By introducing a L2-fusion term that can be seen a weight deviation constraint, the proposed algorithm unifies the offline and online learning, and achieves smoothness of model parameter update. Empirical evaluations both on benchmark testing functions and datasets reveal that the proposed OAS-RVFLN can outperform the conventional methods in learning speed and accuracy. Finally, the OAS-RVFLN is applied to the coal dense medium separation process in coal industry to estimate the ash content of coal product, which further verifies its effectiveness and potential of industrial application.

Online prediction of dense medium suspension density based on phase space reconstruction

ABSTRACTIn the dense medium separation process of the coal preparation plant, the fluctuation of raw coal quality and the lag of heavy media separation density adjustment often lead to the instability of clean coal quality. To solve this problem, phase space reconstruction with online data was carried out to derive a mathematical formula of working dense medium density with raw coal ash and clean coal ash produced. Then the data was trained by two ways, namely Least Square Support Vector Machine (LS-SVM) and LS-SVM based on time series. Simulation results show that the latter way can achieve a more accurate prediction of heavy medium suspension density.

The use of Mössbauer spectroscopy in environmental research

The impact that mining has on the environment is becoming an ever increasing problem all over the world. South Africa, Brazil and India are main producers of various valuable resources such as for example iron ore, platinum, gold and coal, of which coal and platinum mining will be discussed in this paper. Dumping of ash, waste and discards, result in the formation of acid mine drainage (AMD) due to the high sulphur content of the coal and the waste products. The main Fe-S-bearing minerals in the coals investigated were pyrite, jarosite and ferrous sulphate, a weathering product of pyrite. In the ash produced due to combustion or gasification of the coal, the main Fe-constituents are Fe2+,3+ glass (≈ 30%) and hematite (70%). The amorphous phase of the sample was composed mainly of SiO2 and Al2O3 with trace element inclusions of Hg, Ti, Cd and As. The soil, sediment and overburden in the coal mining areas contain pyrite as Fe-S-mineral and also ferrous sulphate as weathering product, with illite the main clay mineral. From laboratory leaching products of coal and ash, sulphur in the form of SO\(_{4}^{2-}\), was found to be one of the most leached ions with a concentration ranging between 100–1000 ppm. The amount of Fe leached out from the ash samples was between 2–5 ppm, but the Fe-leachability depends on the pH, with higher amounts leached out at pH ≤ 1.5. Magnetite losses, to the amount of about 1kg per tonne of magnetite used, occur during the dense medium separation process (DMS) used in cleaning the coal, which also reports in the waste product. South Africa is the largest producer of platinum and smelting of the ore can lead to various forms of pollution. Magnetite formation in the 2-stage furnace process is used as an indicator of the effectiveness of the reduction and the Fe2+ and Fe3+ ratio is used to monitor the process. In the flash furnace the ratio is 2–6, whilst in the electric furnace it is ≤0.02. If not monitored closely a large amount of nickel loss will occur if sent to the waste dump. Mossbauer spectroscopy was used to identify the Fe-species and the results were augmented by High Resolution-Transmission Electron microscopy (HR-TEM), Field Emission-Scanning Electron Microscopy (FE-SEM)/EDS, X-Ray Diffraction (XRD) and ICP-EOS results.

Research on the intelligent control of the dense medium separation process in coal preparation plant

The density of the dense medium separation used to be controlled according to the desired density given by the operators, which depended on the ash data assayed in each hour. Recently, it is indicated through research and practice that the density of the dense medium separation could be controlled intelligently. The control process is implemented through the procedure as follows. The property of the feed coal would be calculated firstly based on the scanned ash of raw coal on line, thus the process parameters, including the density of the dense medium separation, would be forecasted and optimized. The optimized desired density of circulatory medium would be transmitted to the control system. Similarly, the liquid level of the buckets and silo, the amount of the bypass flow, and the added medium would be calculated and controlled intelligently. The research result could reduce the effect of man's activity on the dense medium separation process, shorten the delay time of the process control, and enhance the efficiency of the dense medium separation. The key technology of the intelligent control of the dense medium process is the forecasting process parameters in the dense medium separation process.

Process Control of Dense Medium Separation Based on Improved Implicit Generalized Predictive Control Algorithm

In order to solve the model mismatch between density and liquid level of heavy medium suspension during the dense medium separation process, implicit Generalized Predictive Control (GPC) was applied to realize the on-line identification for model parameters and the decoupling control for density and liquid level of heavy medium suspension. And then improving the implicit GPC algorithm, it can ensure good control performance by constraining control increment to avoid solving the inverse matrix and reduce the computational complexity. The simulation result shows that improved implicit GPC is more effective for the model characteristic of the dense medium separation process with large lag and strong coupling, it has strong anti-interference capability, and the system output is still stable in the case of model mismatch.

Coal dense medium separation dynamic and steady-state modelling for process control

Coal dense medium separation is a popular beneficiation process used for the upgrading of coal ore into power station and metallurgical coal. The control systems used in coal beneficiation are often limited to localised regulatory control of feed rate and medium density. A coal dense medium separation process can benefit substantially from process control provided that a dynamic model for this process is available as was previously developed by the authors for a fine coal dense medium cyclone (DMC) circuit. In this paper, the previous model is adapted to a coarse coal DMC circuit and validated over a wider range of operating conditions using real plant data. The model is further validated by reducing it to steady-state to form a partition curve. This curve is then compared to one derived from actual production data. The derived model is able to provide an estimate of the DMC overflow coal product that should be sufficient for process control.

Stability Study of an Air Dense Medium Fluidized Bed Separator for Beneficiation of High-Ash Indian Coal

Indian high ash noncoking coal contains a substantial quantity of near-gravity materials (NGM). The presence of NGM needs beneficiation in dense medium separation process. Air dense medium fluidized bed separator (ADMFBS) uses the magnetite medium to improve the separation efficiency of beneficiation of high NGM coal. Stability of the particulate fluidized bed in this system is the essential prerequisite for the separation of heavy and light particles. In this study, the characterization of medium and particulate fluidized bed was assessed to maintain the nonbubbling condition. Stability of the fluidized bed was characterized by different expressions like fluidization index, particulate expansion function, Froude number of particle, Reynolds number of particle, and pressure drop ratio using the minimum fluidization velocity, minimum bubbling velocity, pressure drop of distributor and bed, bed porosity, air viscosity, aspect ratio, density of air, and density of medium. It was found that the fluidization indexes of this system for different experimental runs are around 1 to 2. The system was scaled up from laboratory to pilot scale, having a feed throughput capacity of 600 kg/hr. Indian high ash noncoking coal at particular size (−25 + 6 mm) was used in continuous operation. The partition number was calculated based on washability data of product and reject. The Ep and imperfection value for the pilot-scale ADMFBS was found to be 0.12 and 0.071, respectively.

Static and Dynamic Stability in Dense Medium Separation Processes

Static and dynamic stability in dense medium separation processes are defined, and a model for dynamic DMS systems provides the density differential (a measure of dynamic stability) as a function of the static stability and parameters related to the geometry of separatory vessels and operating conditions. A pilot test rig was installed to: (a) study the behaviour of conical and cylindrical dense medium dynamic separators particularly concerning sink and float densities and density differential under different operating conditions; (b) check the correlation model between density differential and static stability. The results provide an interesting description of the different behaviour of conical and cylindrical vessels, regarding the density differential they exhibit, and help explain the differences between the performance of the two types of separators.

The Rheology of Ferrosilicon Dense Medium Suspensions

Medium rheology is important to the efficiency of dense medium separation processes. An impediment to the selection of appropriate ferrosilicon products for a given DMS process in the past has been the lack of a consistent and comprehensive set of data describing the rheology of the range of manufactured ferrosilicon products. Rheological tests have therefore been conducted using Samancor ferrosilicon dense media over a wide range of media properties and densities, with a modified Debex online viscometer. The data have been successfully fitted to the MV-model that allows apparent viscosity to be predicted as a function of particle size distribution and shape, and medium density and temperature. The results from this procedure provide the expected viscosity ranges for uncontaminated ferrosilicon dense media slurries.

Performance of conical and cylindrical separately vessels in dynamic dense-medium separation processes

To obtain a reliable comparison between conical and cylindrical separatory vessels used in dynamic dense-medium separation systems, the following two studies were made: a study of the swirling flow in conical and cylindrical vessels using laser-Doppler velocimetry (LDV) and a study of the behavior of suspensions of magnetite and ferrosilicon in water in both separators. It was found that the tangential velocity profiles differ in conical and cylindrical vessels. In a conical cyclone, free vortex conditions affect the largest part of the vessel, while in a cylindrical vessel, the motion approaches that of a forced vortex, i.e., that of a centrifuge. It was also found that the conical vessel manifests higher centrifugal acceleration and very strong densifying (the density differential) and, hence, the density gradient inside the vessel is large. It was found that the cylindrical vessel manifests lower centrifugal acceleration and low densifying, i.e., low-density differentials, and hence, the density gradient inside the vessel is small. Also, due to greater densifying, the density cut-point in cyclones is higher than the operating medium density, while, in cylindrical vessels, the density cut-point is closer to the medium density. The results provide an interesting description of the differing behaviors of conical and cylindrical vessels, and they help explain the differences in performance between the two types of separators.

Modelling and simulating dense medium separation processes — A progress report

There is an extensive literature on the performance of dense medium separation (DMS) processes in both coal (low density) and mineral (high density) applications. However, very few workers have grappled successfully with the problem of developing effective mathematical models of DMS processes for simulation, other than the trivial option of using partition curves with arbitrary parameter selection. This paper reviews the literature on the modelling of both DMS bath and cyclone separators, and identifies the strengths and weaknesses in the present simulation capabilities. Functions for partition curves are considered in the context of process modelling. Recent work at the JKMRC is summarised, including the development of models of DM drums and cyclones, and a new DMS computer simulation module, JKSimDM, is reported. It is concluded that effective simulation models are now available or within reach, and the paper concludes with a statement of objectives for future research in this area.