By W. Schultz R. Jahnig, R. Bratton, and G. Luttrell
Dewatering processes are required in coal preparation plants to remove excess surface moisture from the clean coal products. Coarser particles can be readily dewatered using simple screening systems, while finer particles require more complex units such as centrifuges and filters.
Moisture that is not removed by these processes reduces the heating value of the coal and increases the cost of transporting the cleaned products. Excess moisture can also create unacceptable handling problems for both the coal producer and downstream consumers by plugging chutes, bins, and rail cars. This problem can be particularly severe in colder regions during winter months due to coal freezing.
Screenbowl centrifuges are now the most popular method for mechanically dewatering fine coal in the U.S. These units, which were first introduced to the U.S. coal industry in 1969 (Parekh and Matoney, 1991), provide low cake moistures at high throughput capacities. For coarser feeds (1 mm topsize), modern machines can provide capacities of up to 100 tons per hour (tph) of dry solids and handle up to 800 to 900 gallons per minute (gpm) of feed slurry flow. Commercial machines typically provide cake moistures in the 14% to 20% range, although increases in the amount of ultrafine solids present in the feed can dramatically increase this value.
Screenbowl centrifuges have a large advantage over competitive processes such as filters since they provide lower product moistures and can reject a significant proportion of low-value ultrafines (minus 325 mesh) that tend to be high in ash and moisture. The machines typically achieve moistures 4% to 6% lower than filters (Gallagher et al., 1981). On the other hand, screenbowl centrifuges tend to degrade the size of the coal more than filters due to the raking action of the internal conveying system.
Operating & Design Guidelines
A screenbowl centrifuge is a continuous two-stage solid-liquid separator that combines a centrifugal solid clarifier (bowl section) together with a centrifugal filter (screen section). Slurry fed to a screenbowl is typically introduced through one end of the unit using a stationary feed pipe. The feed slurry is brought up to rotational speed in an acceleration chamber and is distributed through feed ports into the solid bowl. Solids in the feed slurry settle against the solid bowl wall due to the large centrifugal force field. This thickening action serves as the first stage of dewatering in a screenbowl centrifuge. The clarified water from the bowl section is typically discharged as a waste effluent via an adjustable overflow weir located at the feed end of the machine.
The thickened solids that settle in the solid bowl section are conveyed from the bowl section (opposite in direction to the flow of the main effluent) using a helical scroll that rotates at a slightly slower speed than the bowl. The scroll carries the solids from the solid bowl up a beach and onto the screen section where additional moisture is removed prior to discharge of the final cake. The draining action serves as the second stage of dewatering in a screenbowl centrifuge. In most cases, the material that drains through the screen section contains valuable coal and is recycled back to the feed inlet so that it can be recovered.
There are many operating parameters that influence the dewatering performance of a screenbowl centrifuge (Records and Sutherland, 2001). The most important “feed variables” are particle size, dry solids feed rate, and slurry flow rate. The most important “machine variables” include pool depth, rotational speed, and gearbox ratio.
Effect of Feed Particle Size
The final product moisture generated by a screenbowl centrifuge is highly dependent on the amount of ultrafine particles present in the feed slurry. In the coal industry, the amount of ultrafines is normally reported as the percentage of minus 325 mesh solids contained in the feed slurry. Figure 1 shows the typical moistures produced by a screenbowl centrifuge as a function of the weight percentage of minus 325 mesh solids in the dry feed. As shown, the moisture content increases rapidly as the content of minus 325 mesh solids increases. As a rule-of-thumb, the expected moisture content can be estimated using an empirical expression given by:
Moisture (%) = A + B (S)2
in which S is the percentage of minus 325 mesh solids in the feed. For a 28 mesh x 0 feed, the fitting parameters A and B are 9.2 and 0.01, respectively. Thus, a coal of this type containing 30% by weight of minus 325 mesh solids would give a moisture content of just over 18% (i.e., moisture = 9.2 + 0.01 x 302 = 18.2%). Unfortunately, this site specific value is not under the control of the plant operator and, as such, largely defines the range of potential cake moistures that can be generated for a given application. It should also be noted that many operators place a protection device (e.g., 3/8 inch screen) in the feed line to the unit to keep oversize trash from inadvertently entering (and potentially damaging) the machine.
One of the primary reasons why screenbowl centrifuges are able to provide low cake moistures is the selective classification and removal of ultrafine particles by the solid bowl section. As a rough rule-of-thumb, screenbowl centrifuges typically discard about half of the minus 325 mesh solids with the main effluent. For example, Table 1 shows the size-by-size performance of a screenbowl centrifuge for a minus 1 mm feed from a combined spiral-flotation circuit. The machine recovered only 88.8% of the total solids fed to the unit. However, the machine recovered in excess of 97% of all particles larger than 325 mesh. The losses were almost entirely in the ultrafine fraction, where only 58.3% of the minus 325 mesh fraction solids were recovered.
This is not necessarily a shortcoming for the screenbowl centrifuge (Hart et al., 2005). The discarded ultrafine fraction is relatively poor in terms of quality (29.9% ash) and typically contains up to 50% moisture when recovered. Therefore, for coals sold on steam contracts, it is often better to discard this size material since it does not have sufficient value to justify its recovery (Bethell, 2004; Bethell and Luttrell, 2004).
Effect of Feed Rate
Screenbowl centrifuges have limits on both the amount of dry solids and the volumetric flow of slurry that can be fed to a given machine. Table 2 provides a list of recommended maximum values for various machine sizes used in typical coal applications. Too high of a dry solids feed rate (tph) can overload the maximum transport capacity of the scroll. This problem can also limit the dewatering performance by building excessive bed depths across the screen section. An increase in slurry flow rate (gpm) reduces the time the slurry resides within the bowl. This time is commonly referred to as the slurry residence time (reported in seconds). Typically, a short residence time resulting from an excessively high feed flow rate will (i) increase the amount of solids in the effluent, (ii) coarsen the effective particle separation size, and (iii) decrease the separation efficiency. An increase in flow rate may also increase the pool depth slightly due to an increase in head resulting from the greater volume of effluent that must overflow the weir plates.
Because of the importance of this parameter, the feed slurry flow to screenbowl centrifuges should not exceed the maximum recommended values in terms of gpm. Both the slurry feed rate and the dry solids feed rate are adjustable while the machine is in operation, although these values are usually fixed by coal production requirements at the plant. The load demand for a screenbowl centrifuge is monitored on-line using torque sensors and ammeters. The torque sensor is used to monitor the solids loading on the scrolling mechanism. An increase in torque indicates that more solid material is being fed to the unit. On the other hand, the ammeter installed on the drive is used to monitor the overall load generated by the rotating assembly. Higher amps typically indicate that more feed is going into the machine, although blinding of one of the discharge hoppers can also increase the amp reading. In either case, most operators are encouraged to install dump valves in the feed line that automatically open and divert feed from the machine whenever an overload condition is detected.
Operators are encouraged to monitor and track the moisture content of the cake and the solids contents of the main effluent and screen drain. These values can be compared with historical data compiled by the operator to determine whether problems are occurring with the machine. An increase in any of these values may indicate that the machine is flooded. In some cases, an increase in percent solids for the main effluent may indicate that a hopper has become plugged with solids. In fact, many operators add a small amount of water to the screen drain to prevent solids buildup and plugging. An increase in the percent solids (or increase in particle size) for the screen drain may indicate that the screen section is damaged and needs to be repaired/patched.
Effect of Rotation Speed
Screenbowl centrifuges use centrifugal force to induce the settling of small particles in the bowl section and to pull moisture out of the solids passing across the screen section. Centrifugal force, which is the outward force acting on a body rotating about an axis, is dependent upon the rotational speed of the machine. For convenience, centrifugal force is usually reported as the number of earth gravities (i.e., number of Gs) pulling a particle away from its rotational centerline. The number of Gs can be calculated from:
G = D x N2/6,000
N is the speed (reported as rpm) of the rotating bowl assembly and D is the bowl diameter (reported as ft). As such, rotation speed has a larger impact on the centrifugal force than does the bowl diameter. In the coal industry, most machines are designed to operate at 500 Gs, which requires rotation speeds between 900 and 1,000 rpm depending on machine size.
Since both the centrifuge force and residence time have a direct impact on performance, the product of these two values (i.e., G-seconds) is an important operating and design variable for centrifuges. A higher rotation speed will typically settle more solids and put less solids in the effluent, i.e., the unit makes a finer particle cut size due to the higher G-seconds. Likewise, a lower rotational speed settles less solids and puts more solids in the effluent, i.e., the unit makes a coarser cut size due to the lower G-seconds. As such, the rotation speed of the machine can alter the size distribution of the dewatered cake, which as indicated previously can impact the moisture content of the dewatered cake product. Speed can be increased or decreased while the machine is operating if a variable speed drive is used. However, in most coal applications, rotation speed is usually fixed as dictated by the motor rpm and sheave sizes.
Effect of Pool Depth
Adjusting the weir plate to decrease pool depth reduces the residence time of slurry in the bowl (Figure 2). A shallower pool will usually (i) increase the amount of fines (solids) in the effluent, (ii) decrease the efficiency and coarsen the cut size by reducing the G-seconds, (iii) increase the length of beach not covered by the pool (this will typically decrease cake moisture), and (iv) increase gearbox torque levels. While pool depth cannot be adjusted while in operation, it is an important variable that needs to be established during equipment setup. For maximum dryness, screenbowl centrifuges should normally be run using a Deep Pool weir setting that provides a low pool depth of 2 inches or less. Likewise, for maximum recovery of solids, screenbowl centrifuges should be run using a weir setting that provides a deep pool. Deep pools are normally set to 3.03.5 inches for 36- x 72-inch machines and 4.0-4.5 inches for 44- x 132-inch machines.
In select cases, the recovery of ultrafine solids may be improved by adding chemicals such as polymeric flocculants into the screenbowl centrifuge. While attempts to add these reagents directly into the feed slurry are usually unsuccessful, the use of an injection tube to directly introduce chemicals into the low-solids clarified pool within the rotating bowl section has been shown to work in many cases. The injection pipe runs along the inside wall of the slurry feed pipe and exits just prior to end where the feed slurry is discharged. As such, this modification requires the purchase of a new feed tube for the centrifuge. The use of the injection tube will typically increase the moisture content of the dewatered coal product due to the recovery of more ultrafine solids. As a result, the technology is most appropriate for cases where the additional moisture can be tolerated without contractual penalties or when other drying options (such as thermal dryers) are available downstream. The injection tube may be particularly useful for metallurgical coal plants were high recoveries of carbonaceous solids is important.
Effect of Gearbox Ratio
The gearbox used on the centrifuge is a two stage planetary gearbox. It is mounted on the centrifuge rotating unit and is rotated at the bowl speed. The gearbox ratio affects the differential speed between the bowl (motor driven) and the scroll (gearbox driven). Commercially available gearboxes can provide ratios as low as 20:1 and as high as 140:1. For coal applications, the gearbox ratio is typically fixed at 40:1. This means that the scroll rotates 39 times for every 40 turns of the bowl. A low gearbox ratio provides a high differential speed (Nd), which is defined as the difference in rotational speed between the external bowl and the internal scroll. The differential speed controls how quickly the solids are conveyed out of the machine (i.e., the higher the differential speed, the faster solids are removed). The conveyor speed is the rotating speed (reported in RPM) of the scroll. The conveyor speed (Nc) is calculated by subtracting the differential speed (Nd) from the bowl speed (N). The conveyor velocity (V) is the linear rate at which solids are moved along the inside diameter of the bowl. This velocity can be calculated using:
V = Nd x P/60
V is the conveyor velocity (reported in inch/sec), Nd is the differential speed (reported in rpm), and P is the pitch of the screw conveyor (reported in inches). For example, a machine rotating at 900 rpm with a 40:1 gearbox will have a differential speed of 22.5 rpm. If the pitch is 12 inches, this will give a conveyance speed of 4.5 inch/sec. Since the ratio is typically fixed for coal applications, the differential speed is directly determined by the operating speed of the machine. A high differential speed will typically (i) produce a wetter cake due to shorter residence time on the beach and screen, (ii) decrease effluent solids due to lower residence time in bowl, and (iii) in some cases increase the effluent solids due to turbulence of conveyance (a stirring effect). Any change to the gearbox ratio affects classification and the particle size distribution of the products.
Screenbowl centrifuges offer a long operating life and relatively low maintenance requirements. These benefits can be largely attributed to rigorous design and manufacturing standards and the
use of highly wear resistant materials. Scroll assemblies are fabricated with ceramic tile edges that are individually placed and ground to provide the desired clearance tolerances.
Likewise, the screen assembly incorporates special tungsten carbide bars that are placed by hand within the circular screen section. Materials such as these have greatly extended the service life of screenbowl centrifuges to more than 10,000 hours before rebuilds, which have helped these units become the dominant method for dewatering fines in the coal industry.
Although screenbowl centrifuges are relatively carefree, basic maintenance steps are necessary to keep the units operating effectively. Each shift, operators are encouraged to inspect the overall condition of the unit, listen for unusual noises and to look for abnormal vibrations. The housings and hoppers should be examined for leakage and blockage. The temperatures of the drive motor and oil pump should be monitored to ensure that they are operating within specified limits.
Likewise, oil pressure and oil level should be checked, along with the condition of the hoses and fittings to ensure no leaks exist. The backwash system should be used during each shutdown to remove any solids that may have accumulated in the unit. The level of oil in the main circulating system, sump supply, and gearboxes should be checked each week and the conveyor bushings, thrust bearings and main drive motor should be greased. On a monthly basis, the plows, screen bars and scroll should be inspected for wear, damage and broken parts. The oil in the circulating oil lubrications systems should be changed every six months. When followed, these simple maintenance steps will help to ensure that screenbowl centrifuges operate efficiently with a long service life.
Since first being introduced to the coal industry in the late 1960s, screenbowl centrifuges have become the most common method for dewatering fine coal. Since that time, the machine has developed a well-earned reputation for providing a better reliability and lower product moisture than competitive technologies such as vacuum filters. This article has provided a review of some of the most important process variables that impact the performance of screenbowl centrifuges when used in coal applications. Basic recommendations for how the machine should be operated and monitored have been provided. In addition, a brief review has been provided that describes some of the basic maintenance steps that should be followed in order to ensure that the machine provides an effective, long and trouble-free service life.
This article was adapted from a paper presented at Coal Prep 2008. Schultz and Jahnig represent Decanter Machine, which is located in Johnson City, Tenn. Bratton and Luttrell are professor with Virginia Tech’s Mining & Minerals Engineering Dept., located in Blacksburg, Va.
Gallagher, E. Lewis, J.E., Post, J.J., Swanson, A.R., and Armstrong, L.W., 1981. “Dewatering of Fine Coal by Screen-Bowl Centrifuges,” Proceedings, 1st Australian Coal Preparation Conference, Newcastle, Australia, April 6-10, 1981, pp. 135-154.
Hart, G., Townsend, P., Morgan, G. and Firth, B., 2005. “Improving Fine Coal Centrifuging Stage 3,” Australian Coal Association Research Program, Project C9047.
Parekh, B.K. and Matoney, J.P., 1991. “Mechanic Dewatering,” in Coal Preparation, 5th Edition, Chapter 8, (Leonard, J.W., Ed.), Society for Mining, Metallurgy and Exploration, Inc. (SME), Littleton, Colorado, pp. 499-580.
Bethell, P.J., 2004. “Froth Flotation–To Deslime or Not to Deslime?,” CPSA Journal, Vol. 3, No.1, Spring 2004, pp.
Bethell, P.J. and Luttrell, G.H., 2004. “Effects of Ultrafine Desliming on Coal Flotation Circuits,” Proceedings, Centenary of Flotation Symposium, Brisbane, Queensland, Australia,
Records, A. and Sutherland, K., 2001. “Decanter Centrifuge Handbook,” 1st Ed., Elsevier Science Inc., New York, N.Y., 2001, 421 pp.