Due to the wide variety of filter media, filter designs, suspension properties, conditions for separation and cost, selection of the optimum filter medium is complex. Filter media selection should be guided by the following rule: a filter medium must incorporate a maximum size of pores while at the same time providing a sufficiently pure filtrate. Fulfilment of this rule invokes difficulties because the increase or decrease in pore size acts in opposite ways on the filtration rate and solids retention capacity.
The difficulty becomes accentuated by several other requirements that cannot be achieved through the selection of a single filter medium. Therefore, selection is often reduced to determining the most reasonable compromise between different, mutually contradictory requirements as applied to the filter medium at a specified set of filtration conditions. Because of this, some problems should be solved before final medium selection. For example, should attempts be made to increase filtration rate or filtrate purity? Is cost or medium life more important? In some cases a relatively more expensive filter medium, such as a synthetic cloth, is only suitable under certain filtration conditions, which practically eliminates any cost consideration in the selection process.
Thus, the choice may only be made after consideration of all requirements. It is, however, not practical to analyze and compare each requirement with the hope of logically deducing the best choice. There is, unfortunately, no generalized formula for selection that is independent of the details of the intended application. Each cake requires study of the specific considerations, which are determined by the details or the separation process.
One can to outline a general approach for medium selection along with a test sequence applicable to a large group of filter media of the same type. There are three methods of filter media tests: laboratory- or bench-scale pilot-unit, and plant tests. The laboratory-scale test is especially rapid and economical, but the results obtained are often not entirely reliable and should only be considered preliminary. Pilot-unit tests provide results that approach plant data. The most reliable results are often obtained from plant trials.
Different filter media, regardless of the specific application, are distinguished by a number of properties. The principal properties of interest are the permeability of the medium relative to a pure liquid, its retention capacity relative to solid particles of known size and the pore size distribution. These properties are examined in a laboratory environment and are critical for comparing different filter media.
The permeability relative to a pure liquid, usually water, may be determined with the help of different devices that operate on the principle of measurement of filtrate volume obtained over a definite time interval at known pressure drop and filtration area. The permeability is usually expressed in terms of the hydraulic resistance of the filter medium. This value is found from:
When the cake thickness is 0, we may write the equation as:
Ap=Apt -Apf where Apt = pressure difference accounting for the hydrostatic pressure of a liquid column at its flow through the filter medium, supporting structure and device channels, and Apf = same pressure drop when the flow of liquid is through the supporting structure and device channels
Analytical determination of the hydraulic resistance of the medium is difficult. However, for the simplest filter medium structures, certain empirical relationships are available to estimate hydraulic resistance. The relationship of hydraulic resistance of a cloth of monofilament fiber versus fiber diameter and cloth porosity can be based on a fixed-bed model.
In evaluation and selection of a filter medium, one should account for the fact that hydraulic resistance increases gradually with time. In particular, the relationship between cloth resistance and the number of filter cycles is defined by:
The retentivity relative to solid particles (e.g., spherical particles of polystyrene of definite size) is found from experiments determining the amount of these particles in the suspension to be filtered before and after the filter media. The retentivity K is determined as follows: where g', g" = amounts of solid particles in liquid sample before and after the medium, respectively.
The pore sizes distribution, as well as the average pore size, is determined by the "bubble" method. The filter medium to be investigated is located over a supporting device under a liquid surface that completely wets the medium material. Air is introduced to the lower surface of the medium. Its pressure is gradually increased, resulting in the formation of single chains of bubbles. This corresponds to air passages through the largest-diameter pores. As pressure is increased, the number of bubble chains increases due to air passing through the smaller pores. In many cases a critical pressure is achieved where the liquid begins to "boil." This means that the filter medium under investigation is characterized by sufficiently uniform pores. If there is no "boiling," the filter medium has pores of widely different sizes. The pore size through which air passes is calculated from known relations. For those pores whose cross section may be assumed close to a triangle, the determining size should be the diameter of a circle that may be inscribed inside the triangle.
For orientation in cloth selection for a given process, the following information is essential: filtration objectives (obtaining cake, filtrate or both), and complete data (if possible) on the properties of solid particles (size, shape and density), liquid (acid, alkali or neutral, temperature, viscosity, and density), suspension (ratio of solids to liquid, particle aggregation and viscosity), and cake (specific resistance, compressibility, crystalline, friable, plastic, sticky or slimy). Also, the required capacity must be known as well as what constitutes the driving force for the process (e.g., gravity force, vacuum or pressure). Based on such information, an appropriate cloth that is resistant to chemical, thermal and mechanical aggression may be selected. In selecting a cloth based on specific mechanical properties, the process driving force and filter type must be accounted for. The filter design may determine one or more of the following characteristics of the filter cloth: tensile strength, stability in bending, stability in abrasion, and/or ability of taking the form of a filter-supporting structure. Tensile strength is important, for example, in belt filters. Bending stability is important in applications of metallic woven cloths or synthetic monofilament cloths. If the cloth is subjected to abrasion, then glass cloth cannot be used even though it has good tensile strength.
From the viewpoint of accommodation to the filter-supporting structure, some cloths cannot be used, even though the filtering characteristics are excellent. For rotary drum filters, for example, the cloth is pressed onto the drum by the "caulking" method, which uses cords that pass over the drum. In this case, the closely woven cloths manufactured from monofilament polyethylene or polypropylene fiber are less desirable than more flexible cloths of polyfilament
Depending on the type of filter device, additional requirements may be made of the cloth. For example, in a plate-and-frame press, the sealing properties of cloths are very important. In this case, synthetic cloths are more applicable staple cloths, followed by polyfilament and monofilament cloths. In leaf filters operating under vacuum and pressure, the cloth is pulled up onto rigid frames. Since the size of a cloth changes when in contact with the suspension, it should be pretreated to
In selecting cloths made from synthetic materials, one must account for the fact that staple cloths provide a good retentivity of solid particles due to the short hairs on their surface. However, cake removal is often difficult from these cloths - more than from cloths of polyfilament and, especially, monofilament fibers. The type of fiber weave and pore size determine the degree of retentivity and permeability. The objective of the process, and the properties of particles, suspension and cake should be accounted for. The cloth selected in this manner should be confirmed or corrected by laboratory tests. Such tests can be performed on a single filter. These tests, however, provide no information on progressive pore plugging and cloth wear. However, they do provide indications of expected filtrate pureness, capacity
A single-plate filter consists of a hollow flat plate, one side of which is covered by cloth. The unit is connected to a vacuum source and submerged into the suspension (filtration), then suspended in air to remove filtrate, or irrigated by a dispersed liquid (washing). The filter cloth is directed downward or upward or located vertically, depending on the type of filter that is being modeled in the study.
The following is a recommended sequence of tests that can assist in cloth selection for continuous vacuum filters. If the cycle consists of only two operations (filtration and dewatering), tests should be conducted to determine the suspension weight concentration after 60 sec of filtration and 120 seconds of dewatering. The cake thickness should be measured and the cake should be removed to determine the weight of wet cake and the amount of liquid in it. The weight of filtrate and its purity are also determined. If the cake is poorly removed by the device, it is advisable to increase the dewatering time, vacuum or both. If the cake is poorly removed after an operating regime change, it should be tested with another cloth. If the cake is removed satisfactorily, filtration time should be decreased under increased or decreased vacuum. Note that compressible cakes sometimes plug pores faster at higher vacuum. After the filtration test for a certain filter cycle (which is based on the type of the filter being modeled), the suspension's properties should be examined. Based on the assumed cycle, a new filtration test should be conducted and the characteristics of the process noted. Capacity (N/m2-hr), filtration rate (m3/m2-hr) and cake wetness can then be evaluated. Also, if possible, the air rate and dewatering time should be computed. The results of the first two or three tests should not be taken into consideration because they cannot exactly characterize the properties of the cloth. A minimum of four or five tests is generally needed to achieve reproducible results of the filtration rate and cake wetness to within 3 to 5 %. When the cycle consists of filtration, washing and dewatering, the tests are considered principally in the same manner. The economic aspects of cloth selection should be considered after complete determination of cloth characteristics.
Was this article helpful?