APPLICATIONS OF THE CROSS FLOW TEETER-BED SEPARATOR IN THE U.S. COAL INDUSTRY
Jaisen N. Kohmuench & Michael J. Mankosa, Eriez Manufacturing, Erie, PA. Rick Q. Honaker, University of Kentucky, Dept. of Mining Engineering, Lexington, KY. Robert C. Bratton, Virginia Tech, Dept. of Mining and Minerals Engineering, Blacksburg, VA.
Abstract Hindered-bed separators are recognized as low-cost, high-capacity devices for both classification and density separation; however, since their inception, there have been few significant advances in the fundamental technology. Recently, Eriez has shown through modeling and pilotscale testing that the innovative approach to feed presentation offered in the CrossFlow teeter-bed separator provides improved metallurgy when compared to traditional hindered-bed classifiers or single-stage coal spirals. This design feature prevents excess water from entering the separation chamber and disrupting the overall fluidization flow rate within the teeter zone. Most recently, a side-by-side industrial scale evaluation has verified that this technology improves overall efficiency and simultaneously reduces the separation cut-point. With regards to coal processing, data from full-scale units indicate that the CrossFlow offers good separation efficiency, high unit capacity, and metallurgical results consistent with laboratory- and pilot-scale separators.
Introduction Teeter-Bed Separators: General Hydraulic separators are frequently used in the minerals processing industry to classify fine particles according to size, shape or density. Although many types of equipment exist, a device that has been gaining popularity in recent years is the teeter-bed or hindered-bed separator. The traditional design consists of an open top vessel into which elutriation water is introduced through a series of distribution pipes evenly spaced across the base of the cell. During operation, feed solids are injected into the upper section of the separator and are permitted to settle. The upward flow of elutriation water creates a fluidized “teeter-bed” of suspended particles. The small interstices within the bed create high interstitial liquid velocities that resist the penetration of the slow settling particles. As a result, small/light particles accumulate in the upper section of the separator and are eventually carried over the top of the device into a collection launder. Large/heavy particles, which settle at a rate faster than the upward current of rising water, eventually pass through the fluidized bed and are discharged out one or more restricted ports through the bottom of the separator. Hydrodynamic studies indicate that quiescent flow/non-turbulent conditions must exist in a teeter-bed separator to maintain a high efficiency. Excessive turbulence or changes in flow conditions can result in the unwanted misplacement of particles and a corresponding reduction in separation efficiency. Unfortunately, hydraulic separators typically utilize a feed injection system that discharges directly into the main separation chamber. These simplistic feed systems consist of a vertical pipe that terminates approximately one-third of the way into the main separator body. The pipe discharge is usually equipped with a distribution plate to reduce the flow velocity and disperse the feed slurry, but this approach creates turbulence within the separator that is detrimental to an efficient separation. Another problem with the feed injection system is the discontinuity in flow velocity created by the additional water that enters with the feed solids and reports to the overflow launder. Below the feed point, the flow rate of water is dictated only by the fluidization water rate. This situation is desirable since it allows the operator to accurately control the separation size by adjusting the fluidization flow rate. However, above the feed injection point, the flow rate is the sum of both the feed water and fluidization water flow rates. As a result, the total upward velocity of water is higher above the feed injection point. In fact, at higher feed rates, the volume of water entering with the feed slurry may be greater than the volume flow of fluidization water. The discontinuity created by the feed water often results in a secondary interface of fluidized solids, which varies uncontrollably as the solids content of the feed varies. The increased/variable flow severely impacts the separation performance by increasing cut size, reducing efficiency, and limiting throughput capacity. Equipment maintenance is also an important issue in the design of a hydraulic separator. Conventional teeterbed designs use a series of lateral pipes or a steel plate located at the base of the separation zone. These pipes and plates are perforated at regular intervals with large numbers of small diameter holes. Elutriation water is injected through these holes over the entire cross-section of the separator. The large water flow rates combined with the small injection hole diameters leave the device susceptible to blockage/plugging due to contaminants in the process water. When several orifices become blocked, a dead zone occurs in the fluidization chamber resulting in a loss of performance in this area. As a result, conventional teeterbed separators have an inherent design flaw that limits both the capacity and efficiency of the unit.
Particle Settling Theory Several expressions have been developed to describe the characteristics of particles settling within a hindered state. Empirical methods are normally used to estimate the maximum concentration of solids. In fact, tests conducted using the CrossFlow separator suggested that changes to the cut point (d50) had a large impact on the maximum particle concentration of the underflow. This should be expected since fine particles tend to fill voids that occur between coarser particles; however, as more fines report to the overflow (i.e., higher cut point), these voids remain proportionally empty. To quantify this effect, tests were conducted in which the cut point and maximum packing were determined experimentally. Test data indicate that there is a linear fit between the maximum concentration of solids and d50.
Teeter-Bed Separators: Gravity Concentration The settling rate of any particle within a hindered state is a function of both particle size and density. Because of this inherent interdependency, these devices are typically used for the classification of a like species (i.e., silica sand). However, if the feed size distribution is within acceptable limits, hindered-bed separators can be used for the concentration of particles based on differences in density. Typical density applications include the concentration of heavy minerals or coal. While heavy minerals have a naturally tight size distribution, it is generally accepted for coal applications that the treated size range should have a top to bottom size ratio of no more than 6:1 in order to minimize the classification effects. Plant data suggests that efficient concentration can only be achieved if the particles are in the size range of 75 microns (200 mesh) to several millimeters in diameter. For density separations, the high-density particles settle against the rising flow of water and build a bed of teetering solids segregated according to mass. This bed of solids has an apparent density much higher than the elutriation water. Since particle settling velocity is driven by the density difference between the solid and liquid phase, the settling velocity of the particles is reduced by the increase in apparent density of the teetering bed. This artificial density forces low-density particles to report to the overflow of the separator, and high-density particles to report to the underflow. A population balance model was developed to evaluate the operating behavior of the CrossFlow teeterbed separator. The model is used to predict overflow and underflow partitions, particle size distributions, and the recovery of various density components. Input data to the model include feed rate, percent feed solids by mass, feed size distribution, fluidization water rate, and up to two density components. The discrete model is constructed using a series of wellmixed zones. Three different sections are employed to represent regions in the separator with similar mixing patterns and flow regimes (i.e., feed section, teeter-bed section, and underflow section). The development and validation of this model has been described elsewhere . To illustrate the separation characteristics of a hindered bed for density applications, the population balance model was used to investigate the effect of density difference with regards to recovery. In this effort, a feed, having two density components, was investigated. Simulations were conducted while varying the density ratio of the two components. Component 1 was varied from an SG of 3.0 to 1.25 while maintaining the density of the second component constant at 3.0 SG. For the purpose of this exercise, the density distribution of the feed was maintained equivalent for nine size classes between 1.20 and 0.100-mm . Given a constant set of operating parameters, both component 1 and 2 have an equal chance (~50%) of reporting to the separator overflow when the density (R) of both components are identical. As the SG of component 1 decreases relative to component 2, the recovery of component 1 increases substantially.
Âåðíóòüñÿ â áèáëèîòåêó