Filtration filter sizing

Normally the filter strueture consists of a stack of plates attached to a hollow shaft which are mounted inside a pressure vessel with eaeh plate eovered with a suitable filter medium. The slurry is fed under pressure into the vessel and the eake, which is retained by the filter medium, forms on the top of eaeh plate whilst the filtrate passes through the hollow shaft further to the proeess. Filter sizes may vary but generally the maximum is 60 m area and designed for a 6 bar operating pressure. Each circular plate in the staek is eonstructed with radial ribs that are welded to the bottom and support a horizontal eoarse mesh screen whieh is eovered with a fmer woven metal screen or filter cloth to retain the cake. The bottom of the plate slopes towards the hollow eentral shaft whieh lets the filtrate flow freely through circumferential holes and further down the shaft to the filtrate outlet. The elearanee between the plates is maintained by speeial spacers... 

Some designers push the RW flow rate through MM filters to as high as 15 gal/sq fit/min. This generally is too high and may lead to poor filtration. Also, the resultant cost savings due to a reduced filter size is seldom warranted because the filter is a relatively minor cost as a proportion of the entire RO scheme. A better maximum is perhaps 9 to 10 gal/sq fit/min, with 5 to 6 gal/sq ft/min being preferred. 

Fig. 11. Evaluation of particle post-CMP performance for commercial oxide slurries with fumed and colloidal silica with and without point-of-use filtration. The filter size is 0.3 /im. The bottom denotes that the slurry used is from the bottom of the drum.

Fig- 2. Radioelement concentrations in solutions contacted with powdered spent fuel (UO2 bum-up 50 MW d/kg U -y-dose rate 10 Mrad/h solid surface/solution volume ratio 1000/m at 25 °C) after sequential filtration (filter pore size 450 nm - white bars filter pore size 1.8 nm -> grey bars) solutions consist of concentrated brine (5 mol/kg NaCl) and simulated granitic groundwater (I = 2.8 x 10 1 mol/L, pH 8) (Geckeis et at. 1998). 

An example of a solid-liquid phase separation - often referred to as a mechanical separation - is filtration. Filters are also used in gas-sohd separation. Filtration may be used to recover liquid or sohd or both. Also, it can be used in waste-treatment processes. Walas [6] describes many solid-hquid separators, but we will only consider the rotary-drum filter. Reliable sizing of rotary-drum filters requires bench and pilot-scale testing with the slurry. Nevertheless, a model of the filtering process will show some of the physical factors that influence filtration and will give a preliminary estimate of the filter size in those cases where data are available. 

The honeycomb configuration of ceramic filters offers a high surface area per unit volume, thereby permitting a compact filter size [12]. The absolute filtration surface area depends on cell size, filter volume, and the plugging pattern, all of which are design parameters whose optimization, as will be shown shortly, calls for trade-offs in pressure drop, filtration efficiency, mechanical durability, thermal integrity, and space availability. 

In cross-flow flltration, the wastewater flows under pressure at a fairly high velocity tangentially or across the filter medium. A thin layer of solids form on the surface of the medium, but the high liquid velocity keeps the layer from building up. At the same time, the liquid permeates the membrane producing a clear filtrate. Filter media may be ceramic, metal (e.g., sintered stainless steel or porous alumina), or a polymer membrane (cellulose acetate, polyamide, and polyacrylonitrile) with pores small enough to exclude most suspended particles. Examples of cross filtration are microfiltration with pore sizes ranging from 0.1 to 5 pm and ultrafiltration with pore sizes from 1 pm down to about 0,001 pm. 

The chambers are then opened hydraulically so that the belt may be moved a distance somewhat greater than the length of one chamber. This action discharges cake from both sides of the filter as shown in Fig. 30.7. At the same time, part of the belt passes between spray no22les for washing. After all the cake has been discharged, the belt is stopped, the chambers are closed, and the filtration cycle is repeated. All the steps are actuated automatically by impulses from a control panel. Filter sizes range from 0.8 m (8.6 ft ) to 31.5 m (339 ft ). The overall cycle is relatively short, typically 10 to 30 min, so that these filters can be used in continuous processes. 

MF may be used to remove these heavy metals provided pretreatment chemicals are added to precipitate the metals to particles of filterable size. The chemical pretreatment step is crucial since it will affect the performance of the membrane and the resultant sludge volume as well as the contaminant removal efficiency. Reduction/oxidation, absorption/oxidation, and/or catalytic reactions are utilized along with pH adjustment to provide the optimum precipitation. Although conventional methods of waste water treatment may use a similar pretreatment chemistry, the final solid/liquid separation by gravity settling is usually not as effective as membrane filtration. 

The advantages of the tubular filter are that it uses an easily replaced filter medium, its filtration cycle can be interrupted and the shell can be emptied of prefilt at any time without loss of the cake, the cake is readily recoverable in dry form, and the inside of the filter is conveniently accessible. There is also no unfiltered heel. Disadvantages are the necessity and attendant labor requirements of emptying by hand and replacing the filter media and the tendency for heavy solids to settle out in the header chamber. Applications are as a scavenger filter to remove fines not removed in a prior-filtration stage with a different kind of equipment, to handle the runoff from other filters, and in semiworks and small-plant operations in which the filters size, versatility, and cleanliness recommend it. 

One microscale method developed for differential extraction involves a filter-based system and lysis using acoustic energy. The sample is first infused over a filter (size and material not indicated) in which the sperm cells ( 4-6 [tm diameter) pass through unimpeded and the much larger epithelial cells ( 50 [im diameter) are retained. The DNA is then extracted using ultrasonic disruption of the cells. Although it is too early to gauge the success of this method, filtration has been explored on the macroscale for this application without widespread success. 

The second approach involves the admixture of the filter aid powder to the suspension. Such material is called body aid . The object is to open iq> tire pores of the fiher cake and provide ster filtration. The size distribution may be made coarser. The object of retreating a slurry prior to a s aration process is to change tire properties of the suspension so that a desired inq>rovement resuhs to the separation process. It is useful to conrider the parameters that mi t be altered. 

Filter sizing based on flat disc tests usually consists of recording the filtration time, volume of suspension filtered and pressure drop across the filter. An estimate of the filter performance can be achieved by calculating the volmue of filtrate produced per unit time, for constant-pressure filtration, or the volume of filtrate per unit pressure, for constant-rate filtration. Inspection of these rates will show if the membrane is acceptable, and the investigator can decide on the volume filtered before the filtration becomes either too slow or excessive in pressure drop. This provides a volume filtrate per unit area value which can be used to proportion the areas to arrive at the required filter area. 

Larger filter sizes were used to retain degraded polymer with higher molecular weights and to examine the effect of the insoluble residues on filtration characteristics. 

Filtration tests can be carried out with quite simple apparatus where the objectives are principally twofold. Firstly, cake formation rate is required for preliminary equipment selection, and secondly parameters such as cake specific resistance and solids concentration values are needed, ideally as functions of the applied pressure/vacuum, for filter sizing and filtration rate calculations (Tarleton, 1998a,b Tarleton and Wakeman, 1994c, 1999 Wakeman and Tarleton, 1990, 1991a, 1994a). 

For an initial determination of filtration performance the procedures described in Steps 2-9 are adequate. If data are required for filter sizing and simulation, then Steps 2-9 need to be repeated at a range of different constant pressures/vacua to establish any variation of cake resistance and solids concentration and thus cake compressibility (see also Section 4.7). It is likely that more sophisticated equipment, such as that described in Section 4.6, will give more reliable results. 

The principal objective of an expression test is to determine the compression deliquoring characteristics of a cake. However, the nature of the test allows both filtration and compression characteristics to be determined when the starting mixture is a suspension (i.e. where the solids are not networked or they are interacting to a significant extent). Cake formation rate, specific resistance and solids volume fraction data can be determined for the filtration phase while analysis of a subsequent consolidation phase allows the calculation of parameters such as consolidation coefficient, consolidation index and ultimate solids concentration in the cake. Repeated use of the expression test over a range of constant pressures allows the evaluation of scale-up coefficients for filter sizing and simulation as described in Section 4.7. 

Isolation The wet MBP can be isolated by vacuum-assisted filtration, filter press, or a centrifugal filter. Filtration efficiency mainly depends on the particle size of the precipitate however, other factors such as type of polymer as well as drug loading can also influence the filtration efficiency. Depending on the filtration and washing mechanism used to remove the residual organic solvent, the wet solids... 

One of the expert systems/computational tools for sizing worthwhile highlighting is the filter sizing software - called FILOS - which has been on the market for some years now. It is regularly optimised and it combines the available theoretical basics with a manageable amount of laboratory tests for determination of the key parameters which characterise the filtration process. This system is a very useful additional tool to the conventional lab-scale tests for filter sizing [7]. 

---