Filter Cycle

The vertical recessed plate automatic press, shown schematically in Figure 15 and described previously, is another example of a horizontal belt pressure filter. Cycle times ate short, typically between 10 and 30 minutes, and the operation is fully automated. The maximum cake thickness is about 35 mm washing and dewatering (by air displacement) of cakes is possible. Apphcations include treatment of mineral slurries, sugar, sewage sludge, and fillers like talc, clay, and whiting. 

Typically, a filter cake or precoat is built up on the filter septa to prevent blinding, short filter cycle times, and cosdy cleaning of the septa. Then diatomite is added as body feed to the Hquid to be filtered so that the permeabiHty of the filter cake may be maintained. Filler aid permeabiHty of diatomite ranges from 0.06 to 30 lnF. At the end of the filter cycle the filtrate is clear and the soHds are retained in the soHd or semisoHd diatomite filter cake. The type and amount of diatomite for precoat and body feed are normally deterrnined by pilot studies (18,19). 

It is both convenient and reasonable in continuous filtration, except for precoat filters, to assume that the resistance of the filter cloth plus filtrate drainage is neghgible compared to the resistance of the filter cake and to assume that both pressure drop and specific cake resistance remain constant throughout the filter cycle. Equation (18-51), integrated under these conditions, may then be manipulated to give the following relationships ... 

Generahzed correlations are available for each of the operations which make up the full filter cycle. This means that simulated operating conditions can be varied to obtain a maximum of information without requiring an excessive number of test runs. The minimum number of test runs required for a given feed will, of course, vary with the expertise of the experimenter and the number of operations performed during the filter cycle. If, for example, the operation invdves only the dewatering of a slurry which forms a cake of relatively low to moderate porosity, frequently sufficient data can be obtained in as little as six runs. For more difficult tests, more runs are usually advisable, and the novice certainly should make a larger number of runs as there is likely to be more data scatter. 

Example 1 Sizing a Dish Filter Eqiiipmeut physical factors, selected from Table 18-9 Maximum effective siihmergeuce = 28% maximum portion of filter cycle available for dewatering = 45%. (High submergence versions require tninnion seals, and their use is hmited to specific apphcations.)... 

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 ... 

A suspension of aluminum hydroxide in water is to be filtered imder constant pressure in a batch Nutsch filter having a filtering area of 1 m. Each filter cycle is estimated to separate out 0.5 m of suspension. The operating temperature is 25° C. The following expression for the cake resistance was empirically determined from pilot tests ... 

Batch esterification, 10 478—480 Batch experimental reactor, 21 352 Batch extraction, 10 756 Batch extractor, holdup in, 10 764 Batch fermentation, 10 267 Batch filter cycles, 11 344, 345-346 Batch furnaces, 12 288—289 Batch gasoline blending, 12 413 Batch hydrogenation, 10 811 Batching, ceramics processing, 5 648 Batch injection analysis (BLA) technique, 9 586-587... 

Filter cycles, 11 344-346 batch, 11 344, 345-346 continuous, 11 344-345 Filter drag model, 26 712 Filter housing, 11 322 Filter media, 11 322, 325-326. See also Media... 

Filtration, 11 321-397 75 824—825. See also Filter cycles Filter performance Filters Microfiltration Nanofiltration membranes Ultrafiltration as advanced wastewater treatment, 25 908-909... 

Each filter has demonstrated the capacity to filter the full brine flow of 195 m3 h 1. The pressure drop through the filter medium is measured and monitored continuously. Typically, it is nearly constant over a 2-h filtration at 195 m3 h 1. Back-pulse cleaning restores the initial pressure drop from cycle to cycle, with only a slow increase over time. After 12 months running time, the initial pressure drop at the beginning of the filter cycle had increased by 0.6 bar. The filter membranes were chemically cleaned with 5 % hydrochloric acid. After a cleaning time of 2 h the filter was started again and the pressure drop was less than 0.1 bar greater than that of new filter socks. 

The material to be filtered is fed into the vessel under pressure, and separation takes place with the solids being deposited on the leaf surface, and the liquid passing through the drainage system and out of the filter. Cycle times are determined by pressure, cake capacity or batch quantity. Where particularly fine solids must be removed, a layer of precoat material may be deposited on the leaves prior to filtration, using diatomaceous earth, Perlite, or other suitable precoat materials. 

Fia. 1.—Color of Liquors during a Typical Filter Cycle. 

Density of the flakes (compared with how fast the solvent rose up the column) put constraints on the efficiency of these columns. Under perfect conditions, all the solids would sink, and only solids-free miscella would flow out the top. One never reaches perfect conditions. Reasonable capacity requires appreciable rates of vertical ascent. Fines sink slower than flakes. Agitation creates fines and causes turbulence in the miscella. A major shortcoming of vertical immersion columns is excessive fines in the miscella. The miscella had to be filtered through the solvent-tight plate-and-frame filters, two filters in paralle, one on stream, the other being cleaned. Poor flake preparation caused short filter cycles. Cleaning the filters allowed solvent to escape and created a potential fire hazard. 

A continued cost is associated with purchase and disposal of the filter medium, which is usually discarded at the end of each filter cycle. 

Overall Scale-up Factor The final design filtration rate is determined by multiplying the bench-scale filtration rate by each of the scale-up factors discussed above. While this approach may seem to be ultraconservative, one must realize that the experimenter maintains careful control over the various steps during the filter cycle while running a bench-scale test, whereas a commercial filter operates with a minimum of attendance and at average conditions which are chosen to provide a satisfactory result in a production context. 

---