possible with the above methods. The most effective testing is done with
a commercially available apparatus called the crown press (Fig. 18-194).
This device can simulate the roller actions on the actual belt press and can provide very accurate cake dryness predictions.
Screw Presses A typical screw press is shown in Fig. 18-195, where the slurry is fed into the feed tank at the left-hand side. The core of a screw press is a screw conveyor turning inside a perforated or slotted cylinder. The screw has a smaller diameter at the feed end, and the diameter gradually increases and the screw pitch is shortened toward the discharge end. This design allows gradually decreasing space for slurry/cake and also increasing squeezing pressure on the
cake. As the cake moves toward the outlet, the water is squeezed out through the perforated cylinder.
Screw presses also have the advantage of continuous and automatic operation. Screw presses are primarily used in the pulp and paper, cit- rus, and dairy industries. Applications also exist in many other industries such as dewatering of synthetic rubbers and wastewater sludge. Three pressure (high, medium, and low) ranges are used. High-pressure screw presses are used for vegetable and animal oil; the capacities are rela- tively smaller. Medium-pressure units are used to dewater deformable particles (such as plastic pellet and synthetic rubber) and paper pulp.
Wastewater sludge applications normally use low-pressure options.
SELECTION OF A SOLIDS-LIQUID SEPARATOR
A good solids-liquid separator performs well in service, both initially and over time. It operates reliably day after day, with enough flexibility to accommodate to normal fluctuations in process conditions, and does not require frequent maintenance and repair. Selection of such a sepa- rator begins with a preliminary listing of a number of possible devices, which may solve the problem at hand, and usually ends with the pur- chase and installation of one or more commercially available machines of a specific type, size, and material of construction. Rarely is it worth- while to develop a new kind of separator to fill a particular need.
In selecting a solids-liquid separator, it is important to keep in mind the capabilities and limitations of commercially available devices.
Among the multiplicity of types on the market, many are designed for fairly specific applications, and unthinking attempts to apply them to other situations are likely to meet with failure. The danger is the more insidious because failure often is not of the clean no-go type;
rather it is likely to be in the character of underproduction, subspeci- fication product, or excessively costly operation—the kinds of limping failure that may be slowly detected and difficult to analyze for cause.
In addition, it should be recognized that the performance of mechan- ical separators—more, perhaps, than most chemical-processing equipment—strongly depends on preceding steps in the process. A relatively minor upstream process change, one that might be inadver- tent, can alter the optimal separator choice.
PRELIMINARY DEFINITION AND SELECTION
The steps in solving a solids-liquid separation problem, in general, are:
1. Define the overall problem, with expert assistance if necessary.
2. Establish process conditions.
3. Identify appropriate separator types; make preliminary selec- tions.
4. Develop a test program.
5. Take representative samples.
6. Make simple tests.
7. Modify process conditions if necessary.
8. Consult equipment manufacturers.
9. Make final selection; obtain quotations.
Problem Definition Intelligent selection of a separator requires a careful and complete statement of the nature of the separation prob- lem. Focusing narrowly on the specific problem, however, is not suffi- cient, especially if the separation is to be one of the steps in a new process. Instead, the problem must be defined as broadly as possible, beginning with the chemical reactor or other source of material to be separated and ending with the separated materials in their desired final form. In this way the influence of preceding and subsequent process steps on the separation step will be illuminated. Sometimes, of course, the new separator is proposed to replace an existing unit;
the new separator must then fit into the current process and accept feed materials of more or less fixed characteristics. At other times the separator is only one item in a train of new equipment, all parts of which must work in harmony if the separator is to be effective.
Assistance in problem definition and in developing a test program should be sought from persons experienced in the field. If your orga- nization has a consultant in separations of this kind, by all means make
use of the expertise available. If not, it may be wise to employ an out- side consultant, whose special knowledge and guidance can save time, money, and headaches. It is important to do this early; after the sepa- ration equipment has been installed, there is little a consultant can do to remedy the sometimes disastrous effects of a poor selection. Often it is best to work with established equipment manufacturers through- out the selection process, unless the problem is unusually sensitive or confidential. Their experience with problems similar to yours may be most helpful and avoid many false starts.
Preliminary Selections Assembling background information per- mits tentative selection of promising equipment and rules out clearly unsuitable types. If the material to be processed is a slurry or pumpable suspension of solids in a liquid, several methods of mechanical separa- tion may be suitable, and these are classified into settling and filtration methods as shown in Fig. 18-196. If the material is a wet solid, removal of liquid by various methods of expression should be considered.
Settling does not give a complete separation: one product is a con- centrated suspension and the other is a liquid which may contain fine particles of suspended solids. However, settling is often the best way to process very large volumes of a dilute suspension and remove most of the liquid. The concentrated suspension can then be filtered with smaller equipment than would be needed to filter the original dilute suspension, and the cloudy liquid can be clarified if necessary. Settlers can also be used for classifying particles by size or density, which is usually not possible with filtration.
Screens may sometimes be used to separate suspensions of coarse particles, but are not widely applicable. For separating fine solids from liquids, cake filtration or the newer systems of crossflow filtra- tion should be considered. Crossflow filtration includes ultrafiltration, where the solids are macromolecules or very fine solids (Dp≤ 0.1µm), and microfiltration, where the particle size generally ranges from 0.1 to 5 µm. In microfiltration a suspension is passed at high velocity of 1 to 3 m/s (3 to 10 ft/s) and moderate pressure (10 to 30 lbf/in2gauge) parallel to a semipermeable membrane in sheet or tubular form.
Organic membranes are made of various polymers including cellulose acetate, polysulfone, and polyamide; and they are usually asymmetric, with a thin selective skin supported on a thicker layer that has larger pores. Inorganic membranes of sintered metal or porous alumina are also available in various shapes, with a range of average pore sizes and permeabilities. Most membranes have a wide distribution of pore sizes and do not give complete rejection unless the average pore size is much smaller than the average particle size in the suspension.
In microfiltration, particles too large to enter the pores of the mem- brane accumulate at the membrane surface as the liquid passes through. They form a layer of increasing thickness that may have appreciable hydraulic resistance and cause a gradual decrease in per- meate flow. A decline in liquid flow may also result from small parti- cles becoming embedded in the membrane or plugging some of the pore mouths. The particle layer may reach a steady-state thickness because of shear-induced migration of particles back into the main- stream, or the liquid flux may continue to decline, requiring frequent backwashing or other cleaning procedures. Because of the high veloc- ities the change in solids concentration per pass is small, and the sus- pension is either recycled to the feed tank or sent through several
units in series to achieve the desired concentration. The products are a clear liquid and a concentrated suspension similar to those produced in a settling device, but the microfiltration equipment is much smaller for the same production rate.
SAMPLES AND TESTS
Once the initial choice of promising separator types is made, repre- sentative liquid-solid samples should be obtained for preliminary tests. At this point, a detailed test program should be developed, preferably with the advice of a specialist.
Establishing Process Conditions Step 2 is taken by defining the problem in detail. Properties of the materials to be separated, the quantities of feed and products required, the range of operating vari- ables, and any restrictions on materials of construction must be accu- rately fixed, or reasonable assumptions must be made. Accurate data on the concentration of solids, the average particle size or size distrib- ution, the solids and liquid densities, and the suspension viscosity should be obtained beforeselection is made, not after an installed sep- arator fails to perform. The required quantity of the liquid and solid may also influence separator selection. If the solid is the valuable product and crystal size and appearance are important, separators that minimize particle breakage and permit nearly complete removal of fluid may be required. If the liquid is the more valuable product, can minor amounts of solid be tolerated, or must the liquid be sparkling clear? In some cases, partial or incomplete separation is acceptable and can be accomplished simply by settling or by crossflow filtration.
Where clarity of the liquid is a key requirement, the liquid may have to be passed through a cartridge-type clarifying filter after most of the solid has been removed by the primary separator.
Table 18-17 lists the pertinent background information that should be assembled. It is typical of data requested by manufacturers when they are asked to recommend and quote on a solid-liquid separator.
The more accurately and thoroughly these questions can be answered, the better the final choice is likely to be.
Representative Samples For meaningful results, tests must be run on representative samples. In liquid-solids systems good samples
are hard to get. Frequently a liquid-solid mixture from a chemical process varies significantly from hour to hour, from batch to batch, or from week to week. A well-thought-out sampling program over a pro- longed period, with samples spaced randomly and sufficiently far apart, under the most widely varying process conditions possible, should be formulated. Samples should be taken from all shifts in a continuous process and from many successive batches in a batch process. The influence of variations in raw materials on the separating characteristics should be investigated, as should the effect of reactor or crystallizer temperature, intensity of agitation, or other process variables.
Once samples are taken, they must be preserved unchanged until tested. Unfortunately, cooling or heating the samples or the addition of preservatives may markedly change the ease with which solids may be separated from the liquid. Sometimes they make the separation easier, sometimes harder; in either case, tests made on deteriorated samples give a false picture of the capabilities of separation equip- ment. Even shipping of the samples can have a significant effect.
Often it is so difficult to preserve liquid-solids samples without deteri- oration that accurate results can be obtained only by incorporating a test separation unit directly in the process stream.
Simple Tests It is usually profitable, however, to make simple preliminary tests, recognizing that the results may require confirma- tion through subsequent large-scale studies.
Preliminary gravity settling tests are made in a large graduated cylinder in which a well-stirred sample of slurry is allowed to settle, the height of the interface between clear supernatant liquid and con- centrated slurry being recorded as a function of settling time. Cen- trifugal settling tests are normally made in a bottle centrifuge in which the slurry sample is spun at various speeds for various periods of time, and the volume and consistency of the settled solids are noted. In gravity settling tests in particular, it is important to evaluate the effects of flocculating agents on settling rates.
Preliminary filtration tests may be made with a Büchner funnel or a small filter leaf, covered with canvas or other appropriate medium and connected to a vacuum system. Usually the suspension is poured care- fully into the vacuum-connected funnel, whereas the leaf is immersed Settling
By gravity In thickeners In classifiers By centrifugal force
By heavy media
Filtration
Expression Separation by
By flotation By magnetic force On screens
By gravity By pressure By vacuum
Tubular membranes Flat sheet membranes Rotating filter elements
Screw presses Rolls
Belt presses Crossflow units
Batch presses
Continous presses On filters
FIG. 18-196 Main paths to solids-liquid separation.
SELECTION OF A SOLIDS-LIQUID SEPARATOR 18-151 in a sample of the slurry and vacuum is applied to pull filtrate into a col- lecting flask. The time required to form each of several cakes in the range of 3 to 25 mm (1⁄8to 1 in) thick under a given vacuum is noted, as is the volume of the collected filtrate. Properly conducted tests with a Büchner or a vacuum leaf closely simulate the action of rotary vacuum filters of the top- and bottom-feed variety, respectively, and may give the experienced observer enough information for complete specifica- tion of a plant-size filter. Alternatively, they may point to pressure-filter tests or, indeed, to a search for an alternative to filtration. Centrifugal filter tests are made in a perforated basket centrifugal filter 254 or 305 mm (10 or 12 in) in diameter lined with a suitable filter medium. Slurry is poured into the rotating basket until an appropriately thick cake—
say, 25 mm (1 in)—is formed. Filtrate is recycled to the basket at such a rate that a thin layer of liquid is just visible on the surface of the cake.
The discharge rate of the liquor under these conditions is the draining rate. The test is repeated with cakes of other thicknesses to establish the productive capacity of the centrifugal filter.
Batch tests of microfiltration may be carried out in small pressur- ized cells with a porous membrane at the bottom and a magnetic stir- rer to provide high shear at the membrane surface. These tests may quickly show what type of membrane, if any, gives satisfactory separa- tion, but scaling up to large production units is difficult. Small mod- ules with hollow-fiber, tubular, or spiral-wound membranes are available from equipment vendors, so that tests can be made with con- tinuous flow at pressures and velocities likely to be used for large-scale operation. The permeate flux should be measured as a function of time for different slurry concentrations, pressure drops, and solution velocities or Reynolds numbers. Often a limiting flux will be reached as the pressure drop is increased, but operation at a lower pressure drop is often desirable since the flux decline may not be as great and the average permeation rate over a batch cycle may be greater.
More detailed descriptions of small-scale sedimentation and filtration tests are presented in other parts of this section. Interpretation of the results and their conversion into preliminary estimates of such quantities as thickener size, centrifuge capacity, filter area, sludge density, cake dry- ness, and wash requirements also are discussed. Both the tests and the data treatment must be in experienced hands if error is to be avoided.
Modification of Process Conditions Relatively small changes in process conditions often markedly affect the performance of spe- cific solids-liquid separators, making possible their application when initial test results indicated otherwise or vice versa. Flocculating agents are an example; many gravity settling operations are economi- cally feasible only when flocculants are added to the process stream.
Changes in precipitation or crystallization steps may greatly enhance or diminish filtration rates and hence filter capacity. Changes in the temperature of the process stream, the solute content, or the chemi- cal nature of the suspending liquid also influence solids-settling rates.
Occasionally it is desirable to add a heavy, finely divided solid to form a pseudo-liquid suspending medium in which the particles of the desired solid will rise to the surface. Attachment of air bubbles to solid particles in a flotation cell, using a suitable flotation agent, is another way of changing the relative densities of liquid and solid.
Consulting the Manufacturer Early in the selection campaign—
certainly no later than the time at which the preliminary tests are com- pleted—manufacturers of the more promising separators should be asked for assistance. Additional tests may be made at a manufacturer’s test center; again a major problem is to obtain and preserve representa- tive samples. As much process information as tolerable should be shared with the manufacturers to make full use of their experience with their particular equipment. Full-scale plant tests, although expensive, may well be justified before final selection is made. Such tests demon- strate operation on truly representative feed, show up long-term oper- ating problems, and give valuable operating experience.
In summary, separator selection calls for clear problem definition, in broad terms; thorough cataloging of process information; and pre- liminary and tentative equipment selection, followed by refinement of the initial selections through tests on an increasingly larger scale. Reli- ability, flexibility of operation, and ease of maintenance should be weighed heavily in the final economic evaluation; rarely is purchase price, by itself, a governing factor in determining the suitability of a liquid-solids separator.
TABLE 18-17 Data for Selecting a Solids-Liquid Separator*
1. Process
a.Describe the process briefly. Make up a flowsheet showing places where liquid-solid separators are needed.
b.What are the objections to the present process?
c.Briefly, what results are expected of the separator?
d.Is the process batch or continuous?
e.Number the following objectives in order of importance in your prob- lem: (a) separation of two different solids ; (b) removal of solids to recover valuable liquor as overflow ; (c) removal of solids to recover the solids as thickened underflow or as “dry”
cake ; (d) washing of solids ; (e) classification of solids ; (f) clarification or “polishing” of liquid ; (g) con- centration of solids .
f.List the available power and current characteristics.
2. Feed
a.Quantity of feed:
Continuous process: gal/min; h/day; lb/h of dry
solids.
Batch process: volume of batch: ; total batch cycle: h.
b.Feed properties: temp. ; pH ; viscosity . c.What maximum feed temperature is allowable?
d.Chemical analysis and specific gravity of carrying liquid.
e.Chemical analysis and specific gravity of solids.
f.Percentage of solids in feed slurry.
g.Screen analysis of solids: wet dry
h.Chemical analysis and concentration of solubles in feed.
i.Impurities: form and probable effect on separation.
j.Is there a volatile component in the feed? Should the separator be vapor-tight? Must it be under pressure? If so, how much?
3. Filtration and settling rates
a.Filtration rate on Büchner funnel: gal/(min)(ft2) of filter area under a vacuum of in Hg. Time required to form a
cake in thick: s.
b.At what rate do the solids settle by gravity?
c.What percentage of the total feed volume do the settled solids occupy after settling is complete? After how long?
4. Feed preparation
a.If the feed tends to foam, can antifoaming agents be used? If so, what type?
b.Can flocculating agents be used? If so, what agents?
c.Can a filter aid be used?
d.What are the process steps immediately preceding the separation? Can they be modified to make the separation easier?
e.Could another carrying liquid be used?
5. Washing
a.Is washing necessary?
b.What are the chemical analysis and specific gravity of wash liquid?
c.Purpose of wash liquid: to displace residual mother liquor or to dissolve soluble material from the solids?
d.Temperature of wash liquid.
e.Quantity of wash allowable, in lb/lb of solids.
6. Separated solids
a.What percentage of solids is desired in the cake or thickened under- flow?
b.Is particle breakage important?
c.Amount of residual solubles allowable in solids.
d.What further processing will have to be carried out on the solids?
7. Separated liquids
aClarity of liquor: what percentage of solids is permissible?
b.Must the filtrate and spent wash liquid be kept separate?
c.What further processing will be carried out on the filtrate and/or spent wash?
8. Materials of construction
a.What metals look most promising?
b.What metals must not be used?
c.What gasket and packing materials are suitable?
*U.S. customary engineering units have been retained in this data form. The following SI or modified-SI units might be used instead: centimeters =inches× 2.54; kilograms per kilogram =pounds per pound ×1.0; kilograms per hour = pounds per hour ×0.454; liters per minute =gallons per minute ×3.785; liters per second⋅square meter =gallons per minute⋅square foot ×0.679; and pas- cals=inches mercury ×3377.