Constant Pressure Filtration Research Articles

Multiphase flow models in quantifying constant pressure dead-end filtration and subsequent cake compression: 1. Dilute slurry filtration

Apart from the empirical conventional filtration model, a number of rigorous multiphase flow models are available for the description of the dead-end cake filtration of compressible slurries. In this study, Tiller's and Smiles’ models are compared with regard to their quantification of the dynamic filtration behavior of “dilute” flocculated yeast slurry during dead-end constant pressure filtration. Steady-state filtration is employed to obtain the compressive yield stress and specific resistance of the cake as functions of solid fraction. It is found that, by virtue of these cake properties, the governing equations of Smiles’ and Tiller's model can be numerically solved. The results show that Smiles’ and Tiller's models are equivalent in quantifying filterability and specific resistance, as well as solid fraction, superficial liquid velocity and solid pressure profiles. The compressible property of the cake is demonstrated by the dependence of either filterability or average specific resistance on the applied pressure. For dilute slurries, the applied pressure has a significant influence on solid fraction profile but has little influence on superficial liquid velocity profile with the maximum variance in superficial liquid velocity in the cake being determined by the solid fraction of the slurry. In the dead-end filtration of dilute slurry, the superficial liquid velocity through the cake is almost uniform and the specific resistance can be approximately obtained from correlation of filtration data by the conventional model.

Compressible cake characterization from steady‐state filtration analysis

AbstractA numerical technique for quantifying the key material properties that describe how a flocculated suspension behaves under constant pressure filtration is presented with unrestricted compressive yield stress and permeability models used to describe these material properties. Using an iterative procedure, the optimal parameters for these models are calculated as are pressure and solid fraction distribution profiles. Input parameters to the numerical analysis are flux and final cake height data obtained from batch filtration experiments, which are driven to steady‐state. This technique is validated against piston driven filtration (odeometer) and centrifuge experiments for zirconia, soil particles, and yeast assemblages. The compressive yield stress calculated from filtration experiments agrees well with values obtained by odeometer and centrifuge studies for all particle systems studied at both the low and high solid pressure regions. Similarly, the calculated permeability agrees well with the measured permeability. © 2007 American Institute of Chemical Engineers AIChE J, 2007

Creep effects in activated sludge filter cakes

Constant-pressure and stepped-pressure filtrations have been used to study the creep behaviour of filter cakes formed of activated sludge, anatase, or poly(styrene- co-acrylic acid) particles, respectively. No creep is observed in the case of the anatase filter cake, whereas the poly(styrene- co-acrylic acid) and activated sludge filter cakes exhibit pronounced creep behaviour. The consolidation stage for the poly(styrene- co-acrylic acid) filter cake can be divided into a primary consolidation stage controlled by hydrodynamic effects and a secondary consolidation stage controlled by creep effects. Up to 70% of the consolidation stage was secondary consolidation. Furthermore, the retardation time due to creep deformation was found to be much longer than the filtration time, which suggests that creep does not influence the filtration stage. The creep behaviour observed in activated sludge is pronounced, however no clear transition between primary and secondary consolidation was found. The retardation time is of the same magnitude as the filtration time. This suggests that activated sludge filtration is strongly influenced by creep, and may explain the difficulties in modelling activated sludge filtration and dewatering.

Dewatering of activated sludge by thermosensitive polymers

Thermosensitive polymers such as poly( N-isopropylacrylamide) (poly(NIPAM)) are soluble in water at low temperatures; however, they are insoluble above the intrinsic temperature of the polymer due to its transition to hydrophobicity. In this study, a novel dewatering method of activated sludge by utilizing the hydrophilic/hydrophobic transitions of thermosensitive polymers was investigated. The cationic thermosensitive copolymers of NIPAM and N, N-dimethylaminopropylacrylamide (DMAPAA) synthesized with various DMAPAA contents were used. The activated sludge and polymer solution were mixed well at room temperature. Then, the mixture was heated to the desired temperature. Subsequently, constant pressure filtration and expression processes were carried out. The filtration/expression apparatus was maintained at the desired temperature during these processes. When dewatering was performed at room temperature, the moisture content of the dewatered cake increased with the polymer dosage. This is because the sludge that was sufficiently covered with the adsorbed hydrophilic polymer molecules dispersed stably. The adsorbed polymer acts as a steric stabilizer. On the other hand, by carrying out dewatering above the transition temperature of the polymer, the moisture content of the dewatered cake decreased remarkably with an increase in the polymer dosage. The dewatering performance depended on the temperature and the DMAPAA content of the copolymer. Such high dewatering performance with thermosensitive polymers can be attributed to the hydrophobic interaction among the thermosensitive polymer molecules adsorbed on the sludge.

Validation of Predicted Local Profiles in Cake Filtration

Equations describing compressible cake formation are solved using a cake layer approach. Parameters in constitutive equations are obtained from simultaneous direct local determination of solidosity and pressure in constant pressure filtration. In validation experiments, filtrate rate, filter cake height evolution, pressure and solidosity profiles of two materials were obtained and used to validate calculations. A good agreement was obtained for cumulative filtrate volume and pressure profiles. The accuracy of the predicted solidosity profiles and cake height evolution depends to a great extent on the accuracy of the constitutive equations in the low pressure region.

Effect of Suction Pressure and Aeration on the Hollow Fibre Microfiltration of Excess Activated Sludge

Under various suction pressures (8.0–88.0 kPa), the constant pressure filtration experiment of the excess activated sludge was carried out using the hollow fibre microfiltration membrane with/without aeration. The municipal excess activated sludge was very difficult to be filtered and had the characteristics of the strong compressibility (compressibility coefficient: 1.03). The aeration stripped a large part of cake formed on the membrane surface and increased the flux of the filtrate, but the effect of the aeration reduced at high suction pressure because of the strong compressibility. The very short filtration time used in this experiment little clogged pores of the hollow fibre membrane and the filtration resistance of the formed cake was much larger than the resistance of membrane clogging.

Dynamic optimization of a dead-end filtration trajectory: Non-ideal cake filtration

A control strategy aimed at minimizing energy consumption is formulated for non-ideal dead-end cake filtration with an inside-out hollow fiber ultrafiltration membrane system. The non-ideal behavior was assumed to originate from cake compression, non-linear cake resistance and a variable pump efficiency. Constant gross power, constant flux and constant pressure filtration were considered as alternatives for the optimal operating strategy. It was found that the ratio between the initial and final total resistance determines whether a large difference between these strategies occurs. This is mainly determined by the specific cake resistance, the final state and the membrane resistance. When there is a large difference between the operating strategies, the pump characteristics determine which suboptimal strategies are attractive. For a pump with a low head and a large capacity, constant flux filtration is nearly optimal, whereas for a pump with a large head and a small capacity optimal operation is closer to constant pressure filtration. Under the investigated conditions there was no significant difference ( < 0.5%) between the constant gross power and the optimal operating strategy.

Effect of adsorbed water upon permeabilities of pressure-filtered ceramic nanopowder compacts

The permeabilities of nanocrystalline 3 mol% yttria-stabilized zirconia (3Y-TZP), silica, and boehmite powder compacts are measured during their formation by constant rate pressure filtration. The Carman–Kozeny equation with no account for the effect of adsorbed water often overestimates by a factor of 2 or more the measured permeabilities, with increasing deviation with decreasing permeability. A permeability equation from the literature and one derived here, both taking into account the effect of adsorbed water, show significant improvement over the classic Carman–Kozeny equation for predicting measured permeabilities. The equation derived here allows straightforward predictions to be made of how permeability will change as the critical point of drying (when shrinkage stops) is approached.

Dynamic optimization of a dead-end filtration trajectory: Blocking filtration laws

An operating model for dead-end membrane filtration is proposed based on the well-known blocking laws. The resulting model contains three parameters representing, the operating strategy, the fouling mechanism and the fouling potential of the feed. The optimal control strategy is determined by minimizing the energy consumption for a fixed final time and produced volume. It was found that constant power filtration leads to minimal energy consumption. Constant flux and constant pressure filtration have equal energy costs. However, compared to strategies with a non-decreasing pressure and non-increasing flux, the relative savings are small. Only if the fouling mechanism resembles standard blocking and the fouling resistance is large compared to the membrane resistance, it may be attractive to implement the optimal trajectory.

Simulation of the cake formation and growth in cake filtration

A computer model was developed to simulate the cake formation and growth in cake filtration at an individual particle level. The model was shown to be able to generate structural information and quantify the cake thickness, average cake solidosity, filtrate volume, filtrate flowrate for constant pressure filtration or pressure drop across the filter unit for constant rate filtration as a function of filtration time. The effects of particle size distribution and key operational variables such as initial filtration flowrate, maximum pressure drop and initial solidosity were examined based on the simulated results. They are qualitatively comparable to those observed in physical experiments. The need for further development in simulation was also discussed.

Methods for the rapid characterization of dewaterability of particulate suspensions

A general Darcy's law based model for batch pressure filtration of particulate suspensions is described. Based on this model, two measurement indices for quantifying dewatering performance using experimental constant pressure filtration data are formulated. Results obtained using these indices are in accordance with existing results. The proposed procedure can be particularly useful in situations where the operating pressure is known, and one is only interested in investigating the effect of chemical additives and their dosage on dewaterability of a given suspension.

A new procedure for determining specific filter cake resistance from filtration data

A new procedure has been developed for the determination of specific filter cake resistance from experimental filtration data. Unlike the conventional procedure which treats constant-pressure filtration data through the t / V vs. V plot and yields a single value of the average specific cake resistance ( α av ) from a given experiment, this new procedure allows the establishment of the relationship of α av vs. the cake compressive stress ( p s ) over a range of p s values. Results from the new procedures were compared with those obtained from the procedure based on the t / V vs. V plot as well as those from compression–permeability (C–P) measurements. Discussions on the possible improvement of the new procedure are also presented.

Process Development Strategy to Ascertain Reproducible API Polymorph Manufacture

The present example illustrates the application of a consistent process development strategy to ascertain reproducible active pharmaceutical ingredient (API) polymorph manufacture. Key methodologies are illustrated using a Novartis API and carbamazepine as model substances. In the present example, the Novartis API was synthesized in the final chemical step by hydrolysis followed by acidification. The process was investigated in four steps: First, solid−liquid equilibrium studies were carried out in several solvents and yielded two polymorphs that were characterized by X-ray powder diffraction. Reliable solubility data and the transition temperature (45 °C) were extracted from these experiments as well. Second, based on the knowledge of the enantiotropic behavior of the system, a procedure was developed that crystallized the stable polymorphwhich was chosen for developmentreproducibly by seeding at low supersaturation below 45 °C, followed by cooling at a moderate rate. The absence of metastability with respect to the undesired polymorphic form throughout the process was confirmed by applying ATR-FTIR. Third, reaction conditions were modified to obtain the API in solution at undersaturated conditions. Subsequently, the API could be crystallized in a controlled manner as described above. Fourth, filtration, washing, and drying conditions were investigated to avoid scale-up problems. Constant pressure filtration yielded low compressibility of the filter cake allowing a nutsche or centrifuge as appropriate equipment for isolation. On the basis of vacuum thermogravimetry, fast drying kinetics could be determined. Bench scale paddle dryer experiments illustrated an unacceptable increase of bulk density at permanent rotation. Therefore, intermittent rotation was successfully used at production scale. Thus, the presented process development strategy leads to a robust process scale-up from lab to pilot and production plant, yielding the desired product quality with respect to physical properties as well as chemical purity. Using solid−liquid equilibrium studies, the solubility and the transition temperature (79 °C) between forms I and III of carbamazepine could be determined in 2-propanol. These results correspond well with findings of other authors. On the basis of vacuum thermogravimetry, fast drying kinetics of 2-propanol from carbamezepine is illustrated. Near-complete drying can be achieved at any pressure below the solvent vapor pressure, which points to solvent free from interactions with the solid drug substance.

Derivation of filtration equations incorporating the effect of pressure redistribution on the cake–medium interface: A constant pressure filtration

This study presents a simplified model of cake filtration that incorporates the effect of pressure redistribution on the cake–filter medium interface. The analytical solutions arrived at allow for improved prediction of the cake growth process (including cake thickness and filtration velocity as functions of time) compared with conventional theory. The pressure and void ratio profiles calculated from the present study are compared with results obtained from the use of conventional theory and the numerical solution of dynamic filtration equations.

Characterization of Slurry Compressibility by Constant Pressure Filtration

Characterization of Slurry Compressibility by Constant Pressure Filtration

Measuring Packing Fraction Distribution in Cake Filtered under Constant Pressure and Quantitative Estimation from Measured Filtrate Volume

The particle-packing structure of alumina slurry was evaluated by constant pressure filtration. The cake forming behavior were classified into four patterns by plotting the resistance of the formed cake as a function of the particle volume in the cake; the plots were (a) a straight line, (b) a convex curve, (c) a concave curve, and (d) the combination of a concave curve and a flat line. A method of a estimation of the distribution of packing fraction in the formed cake was proposed. The distribution of packing fraction in the formed cake could be calculated unless the average packing fraction of the cake increased as the filtration proceeded ((c) concave curve). It was found that the distribution of packing fraction estimated by the proposed method was in good accordance with the experimental value obtained by γ-ray transmission measurement.

Nontraditional constant pressure filtration behavior

AbstractPressure filtration is a common industrial process used for solid–liquid separation and laboratory technique used to measure the compressional rheological properties of suspensions. Traditional approaches to the modeling of constant pressure filtration behavior of particulate systems stipulate quadratic behavior of time with filtrate volume. However, this is not necessarily a fundamental attribute of pressure filtration, but rather a result of the assumptions made of the compressibility and permeability of the material. This work solves diffusion‐type constant pressure filtration governing equations using a finite‐difference technique to demonstrate that filtration, under certain circumstances, is expected to show negligible quadratic behavior. Materials that exhibit such behavior have often been classified as nontraditional and methodologies have been developed to interpret such trends. This work illustrates that such behavior is in fact covered by extant filtration models and does not require any extra interpretation of the forces involved. Furthermore, it is shown that traditional or nontraditional behavior is dependent on the initial solids concentration and the applied pressure, such that the filtration of a suspension may show relatively long cake formation times at low initial concentrations or pressures, and relatively short times at high initial concentrations or pressures. © 2005 American Institute of Chemical Engineers AIChE J, 2005

Experimental analysis of a gel layer in dead-end ultrafiltration of a silica suspension

Unstirred dead-end ultrafiltration of a silica suspension was studied using a fully retentive membrane. A constant pressure filtration approach was used to analyze steady-state flux data. The effects of operating conditions such as applied pressure and bulk concentration on specific resistance, thickness, porosity and silica concentration of gel are reported. The specific resistance was obtained from the flux data while the Kozeny—Carman equation was used to determine the silica gel porosity for subsequent calculation of concentration of silica on a deposited layer. Further, experimental verification of time-invariant flux of earlier theoretical work is also presented. Gel concentrations and thickness estimated by utilizing the procedure described in this work were found to be within 1–12% of experimentally determined values.

Development of Slurry Characterization Method by Constant Pressure Filtration-Analysis of Cake Forming Behavior-

The particle-packing structure of Al2O3 slurry was evaluated by using constant pressure filtration. The influences of initial solid concentration and dispersant amount on the cake forming behavior were discussed. It was shown that the packing fraction of formed cake is almost independent of filtration pressure under 10 vol% of initial solid concentration. However at a relatively high solid concentration (20 and 35 vol%), the packing fraction of cake increased with filtration pressure. The cake forming behavior was classified into four patterns by plotting the resistance of the formed cake as a function of particle volume in the formed cake; the plots were (a) a straight line, (b) a convex curve, (c) a concave curve, and (d) a combination of a concave curve and a flat line. The particle dispersion and flocculate degree were discussed on the basis of the results obtained by constant-pressure filtration.

Filtration resistances in non-thermal sterilization of green coconut water

Sterile green coconut water was produced in a developed two stage laboratory scale constant pressure filtration system with a pre-filtration unit by ordinary filter paper (Whatman No. 4) for removal of suspended particles, and a micro-filtration unit by cellulose nitrate membrane (0.2 μm pore opening) for removal of microorganisms. The filtration performances of the systems were tested at three pressure differentials; 5.33, 10.67 and 16.00 kPa. Resistances in both the filtration processes and the cake compressibility factors were investigated. The filter medium resistance ( R m) and the cake resistance ( α) increased with the increase in applied pressure differentials for both the filtration systems. The values of R m varied between 3.497 × 10 9 and 5.042 × 10 9 m −1, and 1.979 × 10 10 and 3.076 × 10 10 m −1 for paper and membrane filter, respectively for the entire range of pressure differentials. The corresponding variations in α were 2.162 × 10 12–4.280 × 10 12 m/kg for paper filter and 2.144 × 10 12–3.942 × 10 12 m/kg for membrane filter. Although the R m values for corresponding pressure differentials were high for membrane filter, the values of α were comparable for both the systems. The deposited cakes were compressible with a compressibility factor of 0.4.