Removal of Oil from Water by Inverse Fluidization of Aerogels

Jose A. Quevedo†, Gaurav Patel and Robert Pfeffer*‡
Otto H. York Department of Chemical Engineering, New Jersey Institute of Technology, Newark, New Jersey 07102
Ind. Eng. Chem. Res., 2009, 48 (1), pp 191–201
DOI: 10.1021/ie800022e
Publication Date (Web): August 16, 2008
Copyright © 2008 American Chemical Society
* To whom correspondence should be address. Tel.: 1-(480) 965-0362. Fax: 1-(480) 965-0037. E-mail: robert.pfeffer@asu.edu., †

Current address: Shell Global Solutions, P.O. Box 38000, 1030BN Amsterdam, The Netherlands.

, ‡

Current address: Department of Chemical Engineering, Arizona State University, Tempe, AZ 85287.

This article is part of the Fan Issue special issue.

Abstract

Surface-treated hydrophobic aerogel (Nanogel) granules of sizes between 500 and 850 μm, 1.7 and 2.3 mm, and 0.5 and 2.3 mm are fluidized by a downward flow of oil-contaminated water in an inverse fluidization mode. Aerogel particles are nanostructured, extremely light and porous, have a very large surface area per unit mass, and are sufficiently robust to be fluidized. Their hydrophobic surface gives them a strong affinity for oil and other organic compounds, with the exclusion of water. These desirable properties make them an ideal sorbent or filter media for the removal of oil from wastewater. The hydrodynamic characteristics of inverse fluidized beds of aerogel granules of different size ranges were studied by measuring the pressure drop and bed expansion as a function of superficial velocity. The oil removal efficiency and capacity of the aerogel granules in the inverse fluidized bed were found to depend mainly on the size of the granules, the initial height of the bed (amount of powder used), the void fraction, and the fluid velocity. Among the advantages of the process are the extremely low energy consumption (low pressure drop) during oil removal and the large absorption capacity. Oil concentrations of about 2000 mg/L in water could be reduced to less than 10 mg/L by the inverse fluidization process.

Introduction

One of the most challenging environmental problems today is the removal of oil from wastewater and stormwater. A large amount of wastewater is generated by industrial companies that produce or handle oil and other organic compounds, both immiscible and miscible in water. Some of these organic materials are discharged into the environment, for example, offshore oil spills and oil released during oil well extraction. Oily wastewater discharged into the environment causes serious pollution problems since the biodegradability of oil is very low and oily wastewater hinders biological processing at sewage treatment plants.
Petroleum and petrochemical plants are potential oil sources for polluting inland water caused by runoff from oil fields, refineries and process effluents.(1) Steel manufacturing and metal working are also major sources of oily wastes.(2) Municipal wastewater contains up to 36% of oily substances derived from vegetable oils and animal fats,(3) and the American Petroleum Institute (API) reports that over 200 million of the 1.3 billion gallons of used oil generated in the U.S. yearly are not collected, but rather dumped into sewers, streams, drains, landfills, and backyards.
Current technologies for oil removal include filtration, gravity separation, induced flotation, ultrafiltration, adsorption, and biological treatment. An oil and water mixture can be classified as free oil(4) for oil droplets larger than 150 μm, dispersed oil with oil droplets in the range of 20−150 μm, and emulsified oil with oil droplets smaller than 20 μm.(5) A wastewater, where the oil in the oil−water mixture is not present in the form of droplets, is said to be soluble.(2) Biological treatment of oil−water mixtures is limited only to low concentrations of oil in water since the microorganisms do not tolerate high oil concentrations and therefore its use is limited to very specific applications.
API separators(4) are used to remove free oil from wastewater and are designed to provide a sufficiently large residence time for the droplets to coalesce and to form an oil layer that can be skimmed off. Dissolved air flotation (DAF) devices are more efficient than API separators because the buoyancy of oil is enhanced by injecting small air bubbles(6, 7) and are very effective in removal of dispersed oil; however, they require the injection of air and addition of pH regulators and coagulants which contribute to the operating costs. DAF units as well as induced air flotation (IAF) units, when properly designed, can achieve an efficiency of 98% and 95% of oil removal, respectively.(8-10)
Removal of dissolved and emulsified oils is achieved by using activated carbon adsorption or membrane filtration.(11) The effluent from those operations contains almost no oil; however, due to the limited removal capacity of the activated carbon and the very high pressures and high quality feed required by membrane filtration, large capital and operating costs are associated with these devices.
Filtration provides good removal, but capacity and energy consumption have to be considered when designing these systems. All filter media have a certain permeability which determines the resistance of the media to water flowing through it. The permeability is commonly monitored by the pressure drop across the filter media and generally increases as the filter media gets saturated with the contaminants. As a result, either the amount of water passing through the filter has to be reduced or the pumping power has to be increased leading to a reduction in efficiency from an energy standpoint.
Several types of filter media (sorbents) have been studied for the removal of oil from water by a packed bed filter such as sawdust,(12) activated carbon and peat,(13) bentonite,(14) and organoclay.(15) Mysore et al.(16) studied the efficacy of vermiculite, alumina-silicate resembling mica, as a filter media for oil-in-water emulsions; they found 30−80% removal of oil depending on operating conditions. Filter media based on reed canary grass, flax, or hemp fiber were studied by Pasila(17) showing that in some cases the media adsorbed 2−4 times its weight in oil. Ribeiro et al.(18) studied dried hydrophobic aquatic plants as a filter media showing that Salvinia sp. adsorbed 1.33 g of oil/g of biomass which is much higher than that adsorbed by Peat Sorb (0.26 g of oil/g Peat Sorb), also used as an oil adsorbent. A discussion about the use of different commercially available sorbent materials to remove oil from stormwater is presented in an Environmental Protection Agency (EPA) report.(19)
The most commonly used sorbent for removal of organic compounds from water is granulated activated carbon (GAC) because of its highly porous structure that provides large internal surfaces for adsorbed molecules to reside. However, it has been reported that hydrophobic CF3 silica aerogels have a much larger absorption capacity than GAC for removing oil and other organic contaminants.(20) For example, it was found that silica aerogels adsorbed contaminants such as toluene, ethanol, chlorobenzene and trichloroethylene by 1−2 orders of magnitude more by weight than activated carbon.(21, 22) This high oil-adsorption capacity makes hydrophobic aerogels, such as Nanogel, an ideal material for removing organic compounds from water. An immediate application of Nanogel would be in the removal of immiscible organic compounds such as oils and chlorinated hydrocarbons that will readily wet the surface of the Nanogel, adsorb onto its surface, and subsequently be absorbed into its porous structure. Extensive descriptions of the use of aerogels can be found in government research laboratory websites and reports, such as the Lawrence Berkeley National Laboratory(23) and the Lawrence Livermore National Laboratory.(24)
The principles of liquid−solid fluidization have been extensively studied. A key work in the field was presented by Richardson and Zaki(25) more than 50 years ago. They found that the superficial velocity divided by the terminal velocity of a single particle is an exponential function of the void fraction in the bed. When the density of the particulate material is less than the density of the fluid, inverse liquid−solid fluidization can be applied to fluidize the solid particles in liquids. For example, aerogels, such as Nanogel, have a density much lower than water and are robust enough to be inversely fluidized. The benefits of inverse fluidization are similar to those found in conventional liquid−solid fluidization: low and constant pressure drop when operating above the minimum fluidization velocity, optimal mixing (contacting) between the solid particles and the liquid, good heat and mass transfer rates, and an adjustable voidage of the fluidized bed by changing the fluid velocity. These advantages should result in an increased adsorption/absorption capacity when removing oil or other organic compounds from water. On the contrary, filtration carried out using packed beds of granular materials has the following disadvantages: pressure drop increases proportionally to the flow rate, and the bed voidage reduces as the filter saturates with the contaminants. Both of these lead to a lower removal capacity.
Previous research using inverse fluidized beds is well-documented in the literature. For example, Fan et al.(26) studied the hydrodynamic characteristics of inverse fluidization in liquid−solid and three phase gas−liquid−solid systems. Nikov et al.(27) used an electrochemical method for liquid−solid mass transfer measurements and proposed an equation that describes the mass transfer in an inverse fluidized bed. Karamanev et al.(28) studied bed expansion characteristics of two-phase inverse fluidization using polystyrene and polyethylene spheres of varied sizes and densities and found that their experimental data showed good agreement with the Richardson and Zaki correlation. Garcia−Calderon et al.(29) studied the hydrodynamics of an inverse fluidized bed of ground Perlite particles as support in an anaerobic reactor focusing on the effects of the Biofilm thickness on bed expansion and terminal velocity; they also used the Richardson and Zaki equation and drag force models to correlate their experimental data. Other interesting studies of inverse fluidization fundamentals and applications can be found in the works of Ibrahim et al.,(30) Bendict et al.,(31) Lee et al.,(32) Cho et al.,(33) and a series of papers by Renganathan et al.(34-37)

Experimental Methods

The objectives of the experiments were to study both the fluidization hydrodynamic characteristics and the oil removal ability of the aerogels during the inverse fluidization process. The former requires varying the fluid velocity in order to find the relationship between the fluidized bed pressure drop and bed height with the flow velocity and the later requires keeping a constant flow rate of water but with the addition of oil into the system. At the beginning of the experiments, aerogel granules, which have a bulk density lower than that of water, remain as a packed bed at the top of the column due to buoyancy; however, by increasing the flow of water, they are fluidized and expand in the downward direction. All experiments were carried out batchwise with respect to the inventory of aerogel granules, i.e., no fresh aerogels were added to replace any oil-saturated aerogels that exited the bottom of the bed with the clean water.
A schematic diagram of the experimental setup used for inverse fluidization of aerogel granules by water is shown in Figure 1. It consists of a fluidization column, valves and piping, flowmeters, a metering pump, a static mixer, a pressure gauge, and a differential pressure transmitter with a display. The fluidization column is made of acrylic plastic with an internal diameter (ID) of 0.089 m (3.5 in.) and an outer diameter (OD) of 0.101 m (4 in.); the total length of the column is 0.86 m. The valves and piping are made of PVC, and the pipe size is 1 in. The flow of tap water is adjusted with ball valves, and flow readings are taken by two calibrated electronic digital flowmeters: one for the range between 0−3 GPM and the other for the range between 3−50 GPM (GPI series A109).
figure

Figure 1. Schematic diagram of the inverse fluidization experimental setup.

The water flow is fed at the top of the column through a distributor made up of a packed bed of glass beads supported by a steel wire mesh. Glass beads have a size of about 3 mm and are placed in the top section of the column in a packed bed of about 5 cm in height. The water flow exits from the bottom of the column and passes through a Keystone Sediment Filter (model 2323N) in order to remove any entrained aerogel granules before being discarded. The top and bottom caps of the column are used for loading and unloading of the particles. It is important to note that no oil was added when the hydrodynamic behavior of the aerogel granules was studied.
A tap before the column allowed for the reading of static pressure, which is held constant during the runs; this measurement was taken with a WIKA pressure gauge with a range of 0−15 psig (0−103 KPa). There are two additional taps before and after the column for measuring the pressure drop using a differential pressure transmitter (model 645-1, Dwyer Instruments) with a range of 0−2 psid (0−13.8 KPa) and an accuracy of 0.1%; the transmitter is connected to a flow display panel meter (C-93284-02, Cole-Parmer) which is connected to a computer through a RJ11 serial communication port and a converter to an RS-232 port. Data are collected by using Meterview software (C-93284-26, Cole-Parmer) which takes readings every 2 s.
To measure the oil removal efficiency and capacity, vegetable oil (soybean oil) from a 1 gal container is injected into the 1 in. PVC pipe size by a diaphragm pump (Pulsatron Series A Plus, 0−6 GPD) and the oil−water mixture is passed through a static mixer made up of steel wire packing which is incorporated into the PVC pipe. There are two sampling points, one just after the static mixer and another right after the column, to take samples of the oil−water mixture flowing upstream and downstream of the column, respectively.
The solid phase in these experiments consists of two types of hydrophobic aerogel granules (Nanogel from Cabot Corp.): clear aerogel (TLD) and dark aerogel (OGD). The granule size ranges were selected by sieving to obtain granules within the size ranges 0.5−0.85 and 1.7−2.3 mm; in some cases, aerogel granules were not sieved and their size range was broader, between 0.5 and 2.3 mm. The aerogel granules have a highly nanoporous structure with a pore size of about 20 nm, a density of about 100 kg/m3, and a surface area in the range of 600−800 m2/g, and they are also extremely hydrophobic due to surface treatment during their manufacture. They have an irregular shape because they are produced by grinding large blocks of the aerogel to the size ranges of the granules used in the experiments. Their large surface area per unit mass greatly enhances the adsorption of oil and other organic molecules and subsequent absorption into the large pore volume of the aerogels. The hydrophobicity of the granules can be readily observed because they preserve their density and structure after being exposed to water. For example, their fluidization behavior does not change even though the granules are immersed in water for a long time. This indicates that water has not entered into the granules’ pore structure; on the other hand, they will readily soak up oil.
The concentration of oil in water is measured by analyzing chemical oxygen demand (COD) in the sample. COD levels of tap water are relatively constant at about 10 mg/L (10 ppm); since oil is the only organic substance added to the water it is safe to assume that any increase in COD levels are due to the addition of oil. COD is measured by using a HACH DR/890 colorimeter following the procedure indicated in the HACH manual, in particular, Method 8000: Reactor digestion method USEPA approved for COD.(38, 39) A calibration curve (not shown) correlating the COD levels with oil concentration was obtained, validating the use of COD analysis as a way to monitor concentration of miscible and immiscible oil in water.
The experimental procedure is as follows. First, the pressure drop across the empty column is measured at different flow rates in order to find a correlation that can be used to determine the pressure drop of the fluidized bed alone; this is done by subtracting the empty column pressure drop from the total fluidized bed pressure drop. Then, the particles to be fluidized are chosen, weighed, and loaded into the fluidization column. Next, the column is filled with water from the bottom and air is completely removed from the system by a vent at a high point. The particles are then inversely fluidized by sending flow from the top (downward), and the hydrodynamic parameters, bed height and pressure drop, are measured by increasing the flow of water gradually. The increase in the flow of water is stopped when the bed height approaches the entire length of the column, or if too much entrainment occurs. It is important to note that the static pressure before the column is kept constant to ensure proper readings.
In the experiments for finding the oil removal efficiency and capacity, a constant water superficial velocity above the minimum fluidization velocity is maintained through the column. Injection of oil, upstream of the column, is done by starting the metering pump and adjusting its stroke displacement and frequency. Samples of water of about 500 mL, upstream and downstream of the inverse fluidized bed, are taken at regular intervals for COD analysis by the HACH colorimeter. Samplings of water as well as the measurement of hydrodynamic parameters are stopped when the bed height approaches the entire length of the column or when oil droplets appear downstream of the fluidized bed. Water samples are mixed thoroughly by using a Hamilton Beach (model 50256MW) blender to disperse the oil droplets homogenously. An aliquot (2 mL for 0−1500 mg/L COD and 0.2 mL for 0−15000 mg/L COD) from the homogenized sample is taken and inserted into the COD digestion vial which is kept in the digestion reactor (DRB-200, Hach Co.) at 150 °C for 2 h. Once the digestion is complete, the vial was allowed to cool down and then tested for COD content by the HACH colorimeter.

Results and Discussion

Hydrodynamics of Inverse Fluidized Beds of Aerogel Granules
The hydrodynamic characteristics of inverse fluidized beds of aerogel granules are represented by the fluidized bed pressure drop and the bed expansion. Plots of these variables against the superficial fluid velocity are used to find the minimum fluidization velocity of the granules (Umf) as shown in Figures 2 and Figure 3a and b. The minimum fluidization velocity is dependent on the granule size; for example, Figure 2 shows that the minimum fluidization velocity for small granules is about 0.007 m/s, and Figure 3a shows that the Umf for large granules (1.7−2.3 mm in diameter) is about 0.02 m/s. A broader particle size distribution (0.5−2.3 mm) results in an intermediate minimum fluidization velocity of 0.013 m/s, as shown in Figure 3b. Besides showing the Umf values, these figures also show that the plateau pressure drop increases proportionally with the amount of powder being fluidized, and that the maximum pressure drop across the fluidized bed of granules remains constant at fluid velocities larger than Umf. The experimentally measured values of Umf are compared to predicted values using correlations found in the literature (see below).
figure

Figure 2. Inverse fluidized bed pressure drop vs superficial fluid velocity of small aerogel granules.

figure

Figure 3. Inverse fluidized bed pressure drop vs superficial fluid velocity of large aerogel granules.

The fluidized bed heights corresponding to Figures 2 and 3 are shown in Figures 4 and 5, respectively. Bed height data are used to determine the average theoretical granule sizes by using the Richardson−Zaki(25) model, and these values are compared to the experimentally measured granule sizes. The minimum fluidization velocity (Umf) data and the bed height data as a function of water velocity are of significance since, to the best of our knowledge, there are no previous research papers reporting on liquid inverse fluidization of aerogel granules.
figure

Figure 4. Bed height vs superficial fluid velocity corresponding to the data in Figure 2.

figure

Figure 5. Bed height vs superficial fluid velocity corresponding to data in Figure 3.

Finding the Density of the Granules (ρp) and the Internal Porosity (εp) from Experimental Data
The density of the granules is important to determine the void fraction (porosity) of the fluidized bed and other hydrodynamic properties. The Cabot Corporation, the manufacturer of the aerogel granules used in the experiments, lists the bulk density and internal porosity of the granules to be about 64 kg/m3 and 0.95, respectively, but does not report the value of the granule density. However, its value can be calculated from the experimental data by using a force balance, i.e., fluidized aerogel granules in equilibrium are acted on by buoyancy, gravity, and drag forces. The buoyancy force is given by while the gravity force is and the drag force is represented by A force balance on the particles gives
The drag force applied on the particles during fluidization (assuming negligible wall effects) is given by the experimental pressure drop divided by the cross sectional area of the fluidization column Substituting eqs 1, 2, and 5 into eq 4 gives After simplification and rearrangement of the terms, and with the definition of the density of the granules eq 7 becomes Using eq 9, the volume of the particles is given by It is important to note that Vp is independent of the bed height.
The initial bulk density of the bed is given by The void volume can be found by subtracting the volume of the particles (Vp) from the total volume of the fluidized bed (Vb) hence, the void fraction of the fluidized bed is given by The solid fraction is given by and the density of the particles is given by eq 8. Finally, the internal porosity of the particles can be found by
The aerogel granule (or particle) density is needed to calculate the void fraction of the fluidized bed, which is used in correlations such as the Richardson−Zaki equation to estimate the average granule size and terminal velocity. The aerogel granule densities are estimated using the equations above, and are listed in Table 1. Equation 10 is of particular significance since it can be used to calculate the particle density if reliable measurements of pressure drop are available, or it can be used to calculate the pressure drop across the fluidized bed if data on the particle density are available. This equation is also useful in scale-up of the process since it can predict the pressure drop across the fluidized bed.
Table 1. Calculation of the Granule Density and the Initial Void Fraction from Experimental Data
granule size/type(mm/type)mass(kg)ΔP (Pa)particles volume(m3)ρp(estim.; kg/m3)initial bedheight (m)bulk density(kg/m3)void fraction(ε0)
0.5−0.85 TLD 1010.1061185.98.7E–041210.264650.47
0.053579.24.3E–041230.143600.52
0.068730.85.4E–041260.183600.52
1.7–2.3 TLD 3020.1972240.81.6E–031190.484660.45
0.11103.28.2E–041230.216750.39
0.131516.81.1E–031170.306680.41
0.5–2.3 TLD 3020.1331482.41.09E–031210.317670.45
Finding the Richardson−Zaki Exponent (n), the Terminal Velocity (Ut), and Estimating the Average Size of the Granules (dp)
The Richardson−Zaki (RZ) correlation(25) is among the most useful methods to estimate the terminal velocity and the size of the fluidizing particles/granules. The RZ equation is where U is the superficial velocity and Ui is the settling velocity at infinite dilution. The Richardson−Zaki exponent or index (n) is a function of the particle terminal Reynolds number (Ret) and the particle to column diameter ratio as given below where eq 17 is specifically applicable for the smaller aerogel particles (0.5−0.85 mm) and eqs 17 and 18 are applicable for the larger aerogel granules (1.7−2.3 mm) depending on the Reynolds number, Ret. In these equations, the Reynolds number at terminal velocity is defined as The RZ exponent (n) can also be obtained from experimental data by plotting the logarithm of the superficial velocity against the logarithm of the void fraction where the slope of a linear regression of the data gives the Richardson−Zaki exponent (n) and the y-intercept gives the settling velocity at infinite dilution (Ui).
After calculating the void fraction (ε) from eq 13 as the bed expands and H increases with increasing fluid velocity, the experimental data are plotted in Figure 6 for the three different aerogel granule size ranges. The settling velocity at infinite dilution (Ui) and the terminal velocity are related by the following equation as suggested by Richardson−Zaki(25) In our experiments, the value of dp/D is very small and is in the range of 0.0055−0.025. The terminal velocity is given by (see the work of Sakiadis(40)) where Cd is the drag coefficient which is a function of the particle Reynolds number,
figure

Figure 6. Relationship between the superficial velocity and the void fraction of inverse fluidized beds of aerogel granules accordingly to the Richardson−Zaki equation.

For spherical particles Only eq 26 is applicable to our data since Rep in our experiments is greater than 0.1 and less than 1000. On the basis of our experimental data and the equations above, the Richardson−Zaki exponent (n), the terminal velocity (Ut), and the average particle diameter were calculated for the different experimental runs as shown in Table 2. As seen in the table, the values of the Richardson−Zaki exponent (n), for all three particle size ranges investigated, calculated from eq 17 agree very closely with the values obtained from the slopes of the three straight lines in Figure 6.
Table 2. Richardson−Zaki Bed Expansion Parameters and Calculation of the Average Particle Size from Experimental Data
granule size/type(mm/type)dp0 (m)Rep(eq 24)Cd(eq 26)Ret(eq 20)RZ (exp)(n)RZ (eq 17) (n)RZ(ln (Ui))Ut(eq 22; m/s)dp(eq 23; mm)
0.5−0.858.5E−0412.33.5336.343.203.181.560.04870.727
1.7−2.32.1E−0383.21.18316.362.492.472.660.15142.366
0.5−2.31.6E−0359.21.39207.422.832.772.510.13032.068
Relationship between the Drag Force Function and the Bed Void Fraction
According to Fan et al.,(26) correlations to predict bed expansion in an inverse fluidized bed can be developed by using a drag force function f, defined as the ratio of the drag force of fluid on particles in a multiparticle system to that in a single particle system, as a function of the Archimedes number Ar = dp3(ρl − ρp)ρlg/μl2 and the Reynolds number Rep = Uρldp/μl. Since the Reynolds numbers for all of our experiments are within the range 2 < Re < 500, the following drag force function taken from Fan et al.(26) is used
In each experimental run, the bed pressure drop and bed expansion were measured as the fluid velocity is increased so that the void fraction, ε, in the fluidized bed could be calculated from the bed expansion at each different fluid velocity. Similarly, the Archimedes and Reynolds numbers and the drag force function “f” were calculated at each different fluid velocity and the drag force function was plotted against the void fraction as shown in Figure 7 for the different sized aerogel granules. Straight lines are obtained for all three granule sizes, the average slope of −4.18 closely agrees with the value of −4.05 suggested by Fan et al.(26) The ranges of applicability of the Fan et al. equation are ε =0.4−0.88, dp/D = 0.062−0.250 and 110000 < Ar < 7650000. Since our results are in good agreement with eq 30, it appears that its range of applicability can be extended to particles with Archimedes numbers less than 100000, such as aerogel granules.
figure

Figure 7. Relationship between the drag force function “f” as defined by Fan et al.(26) and the void fraction ε.

Calculation of the Minimum Fluidization Velocity
It is important to compare the experimental minimum fluidization velocity values against values calculated from equations available in the literature. A classical well-known correlation based on the Ergun equation for predicting Umf, in a conventional fluidized bed was introduced by Wen and Yu.(41) This equation should be applicable to inverse fluidization as well assuming that the drag force of the fluid moving with superficial velocity (Umf) is equal to the buoyancy force less the weight of the particles as described by Karamanev et al.(28) In Table 3, the experimentally measured values of Umf are compared against the theoretical values of Umf calculated from eq 31. This equation correctly estimates the minimum fluidization velocity of the large aerogel granules but not the small ones; this could be due to the fact that the Archimedes number for small granules is almost 2 orders of magnitude lower than that of the large granules.
Table 3. Comparison of the Experimental and Theoretical Minimum Fluidization Velocities
granulesize (mm)Umf(exp; m/s)dp(mm)ArRemfUmf(eq 31; m/s)error(%)
0.5−0.850.0060.734692.00.00352
1.7−2.30.0182.295103370.0179
0.5−2.30.0131.634885170.01119.4
Removal of Oil from Water by Using an Inverse Fluidized Bed of Aerogel Granules
As explained in the experimental methods section, after inversely fluidizing aerogel granules at a certain flow velocity, oil was added to the water to study the oil removal efficiency and the changes in the fluidization characteristics, such as the pressure drop and the bed height, during the adsorption/absorption of oil by the aerogel granules. The concentration of oil was also monitored by analyzing the chemical oxygen demand (COD) at several time intervals during the experiments. The operating conditions and measured variables for each of the experiments are listed in Table 4. Among the variables that were changed to compare the oil removal efficiency are the fluid superficial velocity, particle size, particle type, amount of particles, initial bed height, and concentration of oil upstream of the fluidized bed. The variables that were monitored during the oil removal are the bed height and COD levels as measured by the HACH colorimeter with respect to time. Also, to evaluate the oil removal efficiency and capacity of the inverse fluidized bed, the following arbitrary criterion was adopted: oil removal by an inverse fluidized bed was acceptable only if the COD levels remained below 100 mg/L or 100 ppm. The fluidized bed height and the time at which the 100 mg/L COD level are reached are recorded and shown in Table 4. The longer the time required to reach the 100 mg/L COD level, the better the oil removal efficiency.
Table 4. Summary of Results Corresponding to the Oil Removal from Water by an Inverse Fluidized Bed of Aerogels
figurefluidvelocity(m/s)U/Umfratioparticletypeparticlesize (mm)mass ofparticles(kg)upstream oil conc(g of oil/kg H2O)COD entrance(mg/L)initial fluidbed height(m)max. bedheight (m)time atCOD = 100 (s)removalcapacity(kg oil/kg)
80.03051.5TLD 3021.7−2.30.0540.264900.160.2836603.3
90.03051.5OGD 3031.7−2.30.0490.3910000.140.2323403.5
100.0304.4OGD 3030.5−.850.0560.264900.390.4422001.9
120.0243.5OGD 3030.5−.850.0560.369000.320.3355205.3
130.0243.5TLD 1010.5−.850.0560.369000.340.3445004.3
140.0111.5TLD 1010.5−.850.0560.4517000.140.23136207.1
150.0182.6TLD 1010.5−.850.0560.4818000.220.2268406.5
170.0101.5TLD 1010.5−.850.1080.4717500.430.4415360?
Table 4 also lists the calculated (not measured) upstream oil concentrations as per the oil pump settings and the flow of water. The COD levels of the oil−water mixture before the fluidized bed are also shown. COD concentration, given in milligrams per liter, is not equivalent to the concentration of oil in water but they are proportional. The amount of oil removed is calculated by multiplying the upstream oil concentration by the water flow rate and the elapsed time to reach a COD level of 100 mg/L downstream of the fluidized bed; here we assume that this downstream oil concentration is negligible. The removal capacity is found by dividing the oil removed by the weight of aerogel granules used in the experiment. Table 4 shows that in some cases the aerogel granules removed up to 7 times their weight in oil.
A first set of experiments shows a comparison between large (1.7−2.3 mm) translucent TLD 302 (Figure 8) and large dark OGD 303 (Figure 9) Nanogel granules. The calculated oil removal capacity (shown in Table 4) indicates that OGD 303 granules have a slightly larger capacity than TLD 302 granules. A run with smaller (0.5−0.85 mm) OGD 303 granules fluidized at a higher ratio of velocity to minimum fluidization velocity, U/Umf = 4.4 (Figure 10), indicates that the oil removal efficiency is dependent on the bed voidage. When the fluid velocity is several times higher than the minimum fluidization velocity, the bed of granules is more expanded and the void fraction is higher; therefore, oil droplets can pass easily through the bed of granules. Note that the criterion established for the oil removal (elapsed time to reach a COD level of 100 mg/L) considers collection efficiency rather than capacity. Thus even though a COD level of 100 mg/L has been quickly reached downstream of the fluidized bed, the aerogel granules are not fully saturated with oil.
figure

Figure 8. Chemical oxygen demand (COD) and inverse fluidized bed expansion (squares) as a function of time of 54 g of aerogel granules (TLD 302) with sizes between 1.7 and 2.3 mm during removal of oil from water (0.26 g of oil/kg of water and fluid velocity of 0.0305 m/s).

figure

Figure 9. Chemical oxygen demand (COD) and inverse fluidized bed expansion (squares) as a function of time of 49 g of aerogel granules (OGD 303) with sizes between 1.7 to 2.3 mm during removal of oil from water (0.39 g of oil/kg of water and fluid velocity of 0.0305 m/s).

figure

Figure 10. Chemical oxygen demand (COD) and inverse fluidized bed expansion (squares) as a function of time of 56 g of aerogel granules (OGD 303) with sizes between 0.5 to 0.85 mm during removal of oil from water (0.26 g of oil/kg of water and fluid velocity of 0.0305 m/s).

The fluidized bed expansion provides clues about the way the oil droplets are captured and the saturation of the aerogel granules. It is well-known that the minimum fluidization velocity is proportional to the particle size when comparing similar materials; this also means that at similar fluid velocities, a bed of small granules will have a larger void fraction (larger expansion) than a bed of large granules. A large void fraction in a fluidized bed increases the dynamics of the solid phase (granules), i.e., increases the degree of mixing or axial dispersion between the liquid and solid phases because the granules are surrounded by more “free” space. If there is a high degree of mixing in the fluidized bed then the concentration of the contaminant oil in the suspension phase can be assumed to be uniform, as in a continuously stirred tank reactor (CSTR). However, the fluidized bed behaves more like a packed bed and the fluid flow is similar to that of a plug flow reactor (PFR) when the voids are smaller as given for fluid velocities close to the minimum fluidization velocity.
The degree of mixing in a particulate fluidized bed can be related to the ratio of the actual fluid velocity to the minimum fluidization velocity (U/Umf). A CSTR-like mixing (high axial dispersion) will occur at high U/Umf ratios and a PFR-like mixing (low axial dispersion) will occur at low U/Umf ratios. During oil removal, a CSTR-like mixing leads to a homogeneous saturation of the granules which translates into bed expansion due to the simultaneous reduction in the buoyancy of most of the granules, whereas a PFR-like flow leads to saturation of the granules at the top of the column, and those heavier granules are easily entrained by the flow due to their reduced buoyancy compared to the granules in the rest of the bed resulting in a reduction of the fluidized bed height.
Figure 11 shows the differential pressure drop of the inverse fluidized beds corresponding to the three experiments described above. This figure, and all subsequent pressure drop plots, was constructed using the average value of actual pressure drop measurements at each instant of time. The pressure drop initially increases as oil is added to the water because of the inverse fluidization of small oil droplets, i.e., the flow of water has to overcome their buoyancy force. After a short time, the pressure drop reaches a maximum when equilibrium between the oil added into the fluidization column and the oil adsorbed/absorbed by the granules is established. At this point in time a decrease in pressure drop occurs because the aerogel granules, as they adsorb oil, become heavier, reducing their buoyancy and the drag force needed to fluidize them. The loss of any oil-saturated granules entrained from the fluidization column reduces the pressure drop even further.
figure

Figure 11. Averaged pressure drop across the inverse fluidized beds of aerogel during the removal of oil corresponding to Figures 9–11. The superficial flow velocity was kept constant at 0.0305 m/s.

After saturation of the granules occurs, the average pressure drop remains constant with time (with large fluctuations) as seen in Figure 11 because some of the remaining oil-saturated granules (those that have not exited the bed with the clean water) begin to agglomerate at the top of the column. A buoyant layer of oil and the remaining aerogel granules (held by the distributor) are seen at the top of the column at the end of the experimental run. Note that this phenomenon only occurs because we have run the experiments in a batch mode with respect to the inventory of the granules, and have allowed some of the granules to become highly oil-saturated. In actual practice, clean granules would be fed to the column (to replace those that have exited from the bed) in order to keep the system far from becoming saturated with oil. The pressure drop across the fluidized bed of granules during oil removal can be used to control the amount of fresh aerogel granules to be added to the column to avoid saturation of the system.
Figures 12 and 13 show the removal efficiency of smaller aerogel granules; Figure 12 shows that for OGD 303 (dark) small granules (0.5−0.85 mm), and Figure 13 shows that for TLD 101 (clear) small granules. As can be seen from the plots and the data reported in Table 4, TLD 101 granules adsorbed less oil than OGD 303 granules. Thus, we conclude that the OGD granules have a larger oil adsorption capacity.
figure

Figure 12. Chemical oxygen demand (COD) and inverse fluidized bed expansion (squares) as a function of time of 56 g of aerogel granules (OGD 303) with sizes between 0.5 and 0.85 mm during removal of oil from water (0.36 g of oil/kg of water and 0.0244 m/s fluid velocity).

figure

Figure 13. Chemical oxygen demand (COD) and inverse fluidized bed expansion (squares) as a function of time of 56 g of aerogel granules (TLD 101) with sizes between 0.5 and 0.85 mm during removal of oil from water (0.36 g of oil/kg of water and 0.0244 m/s fluid velocity).

The effect of using a different fluid superficial liquid velocity was also studied, keeping all other operating variables the same, and is shown in Figures 14 and 15. At a low fluid velocity (0.0107 m/s), the collection efficiency is higher since it takes more than 14000 s to produce a level of 100 mg/L of COD downstream of the bed, as compared to 7000 s at the higher fluid velocity (0.0183 m/s). It can be seen from the data that a higher oil removal capacity is obtained at the lower flow velocity. Also, at lower fluid velocity, the drag force is lower allowing the granules to further saturate. In this case, the bed height remains almost constant for about 2 h after which the granules become saturated and groups of particles move downward expanding the bed in a short period of time (3000 s). On the other hand, when the fluid velocity is larger, the voids in the fluidized bed are larger and the drag force is also higher; hence, partially oil-saturated granules leave the bed due to entrainment. This is reflected in a reduction of the fluidized bed height over time.
figure

Figure 15. Chemical oxygen demand (COD) and inverse fluidized bed expansion (squares) as a function of time of 56 g of aerogel granules (TLD 101) with sizes between 0.5 and 0.85 mm during removal of oil from water (0.48 g of oil/kg of water and 0.0183 m/s fluid velocity).

figure

Figure 16. Averaged pressure drop across the inverse fluidized beds of aerogel during the removal of oil corresponding to Figures 14 and 15.

Figure 16 is a plot of the pressure drop across the fluidized bed for the experiments described by Figures 14 and 15, respectively, and shows a change in the rate at which the pressure drop increases/decreases due to the different superficial fluid velocity. At the higher fluid velocity the drag force is larger and entrainment is increased which translates into a faster decrease in the pressure drop across the fluidized bed.
figure

Figure 14. Chemical oxygen demand (COD) and inverse fluidized bed expansion (squares) as a function of time of 56 g of aerogel granules (TLD 101) with sizes between 0.5 and 0.85 mm during removal of oil from water (0.45 g of oil/kg of water and 0.0107 m/s fluid velocity).

Figure 17 shows the COD levels and the bed height of a fluidized bed of 108 g of TLD 101, which is almost double the amount of granules used in most of the previous experiments. As seen by the figures and Table 4, a larger amount of granules results in a taller initial bed height (at the same superficial velocity) which implies a longer residence time for the oil droplets in the fluidized bed. An immediate consequence of the taller bed height is a better oil removal efficiency. COD levels downstream of the fluidized bed containing more granules remain lower than 40 mg/L during the entire experiment (Figure 17), while in the fluidized bed with less granules but at the same operating conditions (see Figure 13), COD levels remain below 90 mg/L during the experiment before reaching the 100 mg/L COD limit. Note that in this experiment (Figure 17) the COD level downstream of the fluidized bed did not reach the 100 mg/L COD limit used as a reference for the capacity criterion. This indicates that the bed was not fully saturated with oil.
figure

Figure 17. Chemical oxygen demand (COD) and inverse fluidized bed expansion (squares) as a function of time of 108 g of aerogel granules (TLD 101) with sizes between 0.5 and 0.85 mm during removal of oil from water (0.47 g of oil/kg of water and 0.0102 m/s fluid velocity).

Regarding the fluidized bed height, as mentioned above, when less granules are used, the aerogel granules tend to saturate more uniformly because of the CSTR-like mixing, which leads to an increase in bed height. However, when we have a longer fluidized bed there is a gradient in the concentration of oil along the fluidized bed, so that granules at the top saturate with oil at a faster rate than others, and these saturated granules leave the fluidized bed due to entrainment; therefore, a reduction in the bed height is observed.

Conclusions

The hydrodynamic characteristics of inverse fluidized beds of aerogel (Nanogel) granules were studied by measuring the pressure drop and bed expansion for different operating conditions, and the experimental results are in good agreement with previous studies on liquid−solid fluidized beds. The experimentally measured fluidized bed pressure drop, at full fluidization, was used to estimate the volume occupied by the granules so that knowing the mass of granules used, the density and the void fraction of the aerogel granules could be calculated. The void fraction and the fluid velocity data were used to estimate the average particle size and terminal velocity using the Richardson−Zaki equation. It is important to note that the Ret was in the range of 200−500, so that the Richardson−Zaki index, n, is around 2.3 which agrees with our experimental data. The calculated values of the average particle size, based on equations for the terminal velocity and drag force, also agree very closely with the actual size of the granules.
A drag force approach for multiparticle systems introduced by Fan et al.(26) correlates the aerogel data quite well for the drag force function and the void fraction in the bed. In addition, the minimum fluidization velocity (Umf) can be estimated by using the correlation given by Wen and Yu(41) for the large aerogel granules, but the correlation gives a poor estimate of the Umf for the smaller sized granules, probably because the Archimedes number for the smaller granules is almost 2 orders of magnitude lower than that of the large granules.
The oil removal efficiency of aerogel granules depends mainly on the size of the granules, the initial height of the fluidized bed (amount of granules used), the void fraction of the bed, and the fluid velocity. Smaller granules will fluidize at lower fluid velocities and result in a lower bed void fraction leading to better removal efficiency. The efficiency drops if the fluid velocity is increased. A taller bed height, at the same fluid velocity, implies a longer residence time for the oil droplets in the fluidized bed; therefore, a better oil removal efficiency is achieved. Also, a lower fluid velocity allows for a higher saturation of the granules since the drag force is lower and the granules spend more time in the fluidized bed without being entrained.
With regard to the oil removal capacity, the experiments show that it is better to work with a taller fluidized bed (more granules), smaller granules, and low fluid velocities. It has been found that at low flow rates the oil adsorption capacity of aerogel granules can be up to 7 times their weight, and the removal efficiency can be as high as 99%. A typical fluid velocity during operation would be in the range of 1 to 2 cm/s for granules less than one millimeter in size. The optimum height of the bed for a given bed diameter (amount of granules used) will depend both on the removal efficiency and capacity that is desired.
A great advantage of the inverse fluidized bed over a packed bed filter is the low pressure drop which translates in low energy consumption. The highest pressure drop obtained in the experiments was around 0.2 psi (1300 Pa) when removing oil with 100 g of granules. The pressure drop across the inverse fluidized bed will not increase as the oil is adsorbed/absorbed in the granules; moreover, it decreases unless more granules are added. In a packed bed filter, the pressure drop depends on the flow rate passing through the bed; in a fluidized bed, the pressure drop is only proportional to the amount of granules fluidized. In several applications of packed bed filters, the limiting pressure drop is about 10 psi; this value can be easily reached with a very small amount of granules depending on granule size and fluid velocity. However, in order to reach a differential pressure drop of 10 psi across an inverse fluidized bed of aerogel granules in a column of 3.5 in. internal diameter, as much as 6 kg of aerogel granules could be used. These 6 kg of aerogel granules could adsorb up to 42 kg of oil, and the bed height of the fluidized bed would be about 16 m (52.5 ft.). Thus, it can be concluded that an inverse fluidized bed of aerogel granules is an excellent candidate for removing oil and other organic contaminants from wastewater.
For continuously removing oil from contaminated water, it is desirable to keep the fluidized bed height constant. This can be achieved by having oil saturated particles entrained from the fluidized bed and then separated from the clean water downstream with a filter or other solid−liquid separating device. To compensate for the loss of particles, fresh aerogel granules can be added at the bottom of the fluidization column which will float upward, countercurrent to the downward flow of contaminated water, due to their buoyancy. The behavior of such a continuous fluidized bed device with respect to a constant granule inventory is presently being investigated.
The use of aerogels as an adsorbent/absorbent for contaminant oil removal from water in an inverse fluidized bed offers high removal efficiency and absorption capacity and is based on the transfer of oil from a liquid phase (water) to a solid phase (aerogel). Most conventional sorbents currently used for removing oil from water end up in landfills or are incinerated after a single use. However, to make the process more economical and competitive, it will be necessary to recover the oil and reuse the aerogels. With regard to recovering the oil from the saturated aerogel granules, there are several options such as pressing, washing with hydrocarbons or other solvents, or use of superheated steam.
In a recent paper, Wei et al.(42) describe the use of a water solution of a biodegradable biosurfactant, rhamnolipids biosurfactant JBR215, to remove oil from used polypropylene nonwoven sorbents. They found that by using biosurfactant washing, more than 95% of the oil was removed from the used sorbents, depending on the washing conditions, and they concluded that biosurfactants have considerable potential for recycling used sorbents. While aerogels have a totally different structure and different properties compared to polypropylene nonwoven sorbents and the aerogel nanoporous structure may be permanently lost during the adsorption/absorption of oil, it may be possible to regenerate the saturated aerogels continuously by washing with biosurfactants in a second fluidized bed placed in series with the first. These alternatives will have to be evaluated in the future to find the optimal disposal/regeneration method.

Acknowledgment

We gratefully acknowledge partial financial support from the National Science Foundation through Grant CBET 0730465 during the preparation and writing of this manuscript. We also wish to thank the Cabot Corporation for providing partial financial support during the experimental phase of this research, as well as the aerogel (Nanogel) granules used in the experiments. Technical discussions and advice from Cabot personnel, in particular, Dr. Nirmalya Maity and Dr. Bart Kalkstein, are also greatly appreciated.

References

This article references 42 other publications.

  1. 1.
    Johnson, R. F.; Manjrekar, T. G.; Halligan, H. R.Removal of oil from water surfaces by sorption on unstructured fibers Environ. Sci. Technol. 1973 7 439 443
    [ACS Full Text ACS Full Text], [ChemPort]
  2. 2.
    Paterson, J. W. Industrial Wastewater Treatment Technology, 2nd ed.; Butterworth Publishers, Inc.: Stoneham, MA, 1985.
  3. 3.
    Quemeneur, M.; Marty, Y.Fatty acids and sterols in domestic wastewaters Water Res. 1994 28 1217 1226
    [CrossRef], [ChemPort]
  4. 4.
    Manual on Disposal of Refining Wastes; American Petroleum Institute: Washington, D.C., 1969; Vol. 5, pp 5− 15
  5. 5.
    Manning, F.; Snider, E. H.Assesment data base for petroleum refining wastewater and residues; U.S. Department of Commerce, NTIS: Washington, D.C., 1983; Vol. 9, pp 4− 101
  6. 6.
    Tchobanoglous, G.; Burton, L. F.; Stensel, H. D. Wastewater Engineering: Treatment and Reuse; Mc-Graw Hill: New York, 2003; pp 419− 422.
  7. 7.
    Bennett, G. F.The removal of oil from wastewater by air flotation: A review. CRC Crit. Rev. Environ. Contr.1988, 18 ( 3), 189− 253
  8. 8.
    Zouboulis, A. I.; Avranas, A.Treatment of oil-in-water emulsions by coagulation and dissolved-air flotation Colloid. Surf. A, Phys. Eng. Asp. 2000 172 1−3 153 181
    [CrossRef], [ChemPort]
  9. 9.
    Suzuki, Y.; Maruyama, T.Removal of Emulsified Oil from Water by Coagulation and Foam Separation Sep. Sci. Technol. 2005 40 3407 3418
    [CrossRef], [ChemPort]
  10. 10.
    Gaaseidnes, K.; Turbeville, J.Separation of oil and water in oil spill recovery operations Pure Appl. Chem. 1999 71 1 95 101
    [CrossRef], [ChemPort]
  11. 11.
    Goldsmith, R.; Hossian, S.Ultrafiltration concept for separating oil from water; U.S. Coast Guard: Washington, D.C., January 1973.
  12. 12.
    Cambiella, A.; Ortea, E.; Rios, G.; Benito, J. M.; Pazos, C.; Coca, J.Treatment of oil-in-water emulsions: Performance of a sawdust bed filter J. Hazard. Mater. 2006 B131 195 199
    [CrossRef]
  13. 13.
    Mathavan, G. N.; Viraraghavan, T.Use of peat in the treatment of oily waters Water, Air, Soil Pollut. 1989 45 17 26
    [ChemPort]
  14. 14.
    Viraraghavan, T.; Moazed, H.Removal of oil from water by bentonite Fresenius Environ. Bull. 2003 12 9 1092 1097
    [ChemPort]
  15. 15.
    Alther, G. R.Organically modified clay removes oil from water Waste Manage. 1995 15 8 623 628
    [CrossRef], [ChemPort]
  16. 16.
    Mysore, D.; Viraraghavan, T.; Jin, Y. C.Vermiculite Filtration for Removal of Oil from Water Pract. Period. Hazard., Toxic, Radioact. Waste Manage. 2006 156 161
    [CrossRef]
  17. 17.
    Pasila, A.A biological oil adsorption filter Mar. Pollut. Bull. 2004 49 1006 1012
    [CrossRef], [PubMed], [ChemPort]
  18. 18.
    Ribeiro, T. H.; Rubio, J.; Smith, R. W.A Dried Hydrophobic Aquaphyte as an Oil Filter for Oil/Water Emulsions Spill Sci. Technol. Bull. 2003 8 5−6 483 489
    [CrossRef], [ChemPort]
  19. 19.

    USEPA.

    Sorbent Materials in Storm Water Applications.http://epa.gov/owm/mtb/mtbfact.htm (accessed July 2006).
  20. 20.
    Reynolds, J. G.; Coronado, P. R.; Hrubesh, L. W.Hydrophobic Aerogels for Oil-Spill Cleanup - Intrinsic Absorbing Properties Energy Sources 2001 23 831 843
    [CrossRef], [ChemPort]
  21. 21.
    Hrubesh, L. W.; Coronado, P. R.; Dow, J. P.Method for removing organic liquids from aqueous solutions and mixtures. U.S. Patent 6709600 B2. March 23, 2004.
  22. 22.
    Hrubish, L. W.; Coronado, P. R.; Satcher, J. H.Solvent removal from water with hydropholic aerogels J. Non-Cryst. Solids 2001 285 1−3 328 332
    [CrossRef]
  23. 23.
    Ayres, M.; Hunt, A.

    Silica aerogels.

    http://eetd.lbl.gov/ECS/aerogels/satoc.htm (updated April 2004), Ernest Orlando Lawrence Berkeley National Laboratory. Environmental Energy Technologies Division (EETD).
  24. 24.

    Aerogel/Granulated Activated Carbon composites for adsorption of contaminants in water.

    https://ipo.llnl.gov/technology/ profile/aerogel/GACComposites/index.php (accessed July 2006), Lawrence Livermore National Laboratory.
  25. 25.
    Richardson, J. F.; Zaki, W. N.Sedimentation and fluidization: Part I Trans. Instn. Chem. Eng. 1954 32 35 53
    [ChemPort]
  26. 26.
    Fan, L-S.; Muroyama, K.; Chern, S. H.Hydrodynamic Characteristics of Inverse Fluidization in Liquid-Solid and Gas-Liquid-Solid Systems Chem. Eng. J. 1982 24 143 150
    [CrossRef], [ChemPort]
  27. 27.
    Nikov, I.; Karamanev, D.Liquid-Solid Mass Transfer in Inverse Fluidized Bed AIChE J. 1991 37 5 781 784
    [CrossRef], [ChemPort]
  28. 28.
    Karamanev, D. G.; Nikolov, L. N.Bed Expansion of Liquid-Solid Inverse Fluidization. AIChE J. 1992 38 12 1916 1922
    [CrossRef], [ChemPort]
  29. 29.
    Garcia-Calderon, D.; Buffiere, P.; Moletta, R.; Elmaleh, S.Influence of Biomass Accumulation on Bed Expansion Characteristics of a Down-Flow Anaerobic Fluidized-Bed Reactor Biotechnol. Bioeng. 1998 57 2 136 144
    [CrossRef], [PubMed], [ChemPort]
  30. 30.
    Ibrahm, Y. A. A.; Briens, C. L.; Margaritis, A.; Bergongnou, M. A.Hydrodynamic Characteristics of a Three-Phase Inverse Fluidized-Bed Column AIChE J. 1996 42 7 1889 1900
    [CrossRef]
  31. 31.
    Femin Bendict, R. J.; Kumaresan, G.; Velan, M.Bed expansion and pressure drop studies in a liquid-solid inverse-fluidised bed reactor Bioprocess Eng. 1998 19 137 142
    [CrossRef], [ChemPort]
  32. 32.
    Lee, D. H.; Epstein, N.; Grace, J. R.Hydrodynamic transition from fixed to fully fluidized beds for three-phase inverse fluidization Korean J. Chem. Eng. 2000 17 6 684 690
    [CrossRef], [ChemPort]
  33. 33.
    Cho, Y. J.; Park, H. Y.; Kim, S. W.; Kang, Y.; Kim, S. D.Heat Transfer and Hydrodynamics in Two-and Three-Phase Inverse Fluidized Beds Ind. Eng. Chem. Res. 2002 41 2058 2063
    [ACS Full Text ACS Full Text], [ChemPort]
  34. 34.
    Renganathan, T.; Krishnaiah, K.Prediction of Minimum Fluidization Velocity in Two and Three Phase Inverse Fluidized Beds Can. J. Chem. Eng. 2003 81 853 860
    [ChemPort]
  35. 35.
    Renganathan, T.; Krishnaiah, K.Stochastic Simulation of Hydrodynamics of a Liquid-Solid Inverse Fluidized Bed Ind. Eng. Chem. Res. 2004 43 4405 4412
    [ACS Full Text ACS Full Text], [ChemPort]
  36. 36.
    Renganathan, T.; Krishnaiah, K.Liquid phase mixing in 2-phase liquid-solid inverse fluidized bed Chem. Eng. J. 2004 98 213 218
    [CrossRef], [ChemPort]
  37. 37.
    Renganathan, T.; Krishnaiah, K.Voidage characteristics and prediction of bed expansion in liquid-solid inverse fluidized bed Chem. Eng. Sci. 2005 60 2545 2555
    [CrossRef], [ChemPort]
  38. 38.
    Jirka, A. M.; Carter, M. J.Reactor digestion method for COD analysis Anal. Chem. 1975 47 8 1397
    [ACS Full Text ACS Full Text], [PubMed], [ChemPort]
  39. 39.
    Method 8000: Reactor digestion method USEPA approved for oxygen chemical demand wastewater analysis. DR/890 Datalogging Colorimeter Handbook; Hach Co.: Loveland, CO, 2004; pp 427− 436
  40. 40.
    Sakiadis, B. C.Fluid and particle mechanics. Perry’s Chemical Engineer’s Handbook; McGraw-Hill: New York, 1984
  41. 41.
    Wen, C. Y.; Yu, Y. H.Mechanics of fluidization Chem. Eng. Prog. Symp. Ser. 1966 62 2 100
    [ChemPort]
  42. 42.
    Wei, Q. F.; Mather, R. R.; Fotheringham, A. F.Oil removal from used sorbents using a biosurfactant Bioresour. Technol. 2005 96 331 334
    [CrossRef], [PubMed], [ChemPort]

Tools

SciFinder Links

SciFinder subscribers: Click to sign in | Not a SciFinder subscriber? Learn more at www.cas.org

Explore by:

History

  • Published In Issue January 07, 2009
  • Article ASAPAugust 16, 2008
  • Received: January 4, 2008
    Accepted: May 22, 2008
    Revised: May 19, 2008

Recommend & Share

Related Content

Other ACS content by these authors: