Cross flow microfiltration of oil–water emulsions using kaolin based low cost ceramic membranes

Emphasizing upon the need to develop and target applications of low cost ceramic membranes, this work addresses the cross flow microfiltration (MF) of oil–water emulsions. Membranes prepared using uniaxial dry compaction method and possessing higher average pore sizes (3.06–2.16 μm) and porosities (30.1–37.4%) have been used for cross flow MF studies. Cross flow runs were conducted using a feed oil concentration of 400 mg/L of for various combinations of trans membrane pressure differentials (138–207 kPa) and cross flow velocities (Reynolds number = 3417 to 6835). The fitness of various fouling models indicates that combination fouling models could only represent the observed flux decline data which refers to existence of one of the pore blocking phenomena (complete/intermediate/standard pore blocking) during the first 10 min of experimental run followed with cake filtration for the data obtained for later time periods. Experimental investigations inferred that the membrane performance is optimal for maximum combinations of trans membrane pressure (ΔP = 207 kPa), fabrication pressure (73 MPa) and minimal cross flow velocity (Re = 3417) at which conditions, an optimal combination of fouling index, steady state flux and rejection were obtained as 11.58%, 22.14 × 10−6 m3 m−2 s and 98.52%, respectively. © 2014 Elsevier B.V. All rights reserved.

With limited literature data, this work emphasized upon the need to develop low cost ceramic membranes for various microfiltration applications. Membranes prepared with the uniaxial dry compaction method possessed higher combinations of average membrane pore size and porosity and all membranes upon cross flow microfiltration using oil water emulsions indicated that there existed significant pore blocking during the first 10 min of the experimental run. Thus, it is apparent that the prepared membranes do have the tendency to undergo irreversible fouling, which is also apparent from the fairly high fouling index evaluated in this work. Overall, the optimal combinations of membrane morphology and operating parameters refers to the M3 membrane (with average membrane pore size and porosity of 2.16 μm and 37.4% respectively) which provided a steady state flux and rejection of 22.14 × 10−6 m3 m−2 s and 98.52% respectively along with a fouling index of 11.58% at a ΔP of 207 kPa and Reynolds number (cross flow velocity) of 3417. Inferentially, this work further highlighted upon the need to prepare low cost ceramic membranes within the submicron range of microfiltration using uniaxial dry compaction method/ paste method and their subsequent cross flow microfiltration. Thus in conclusion, the low cost MF range ceramic membranes provided low fouling index during cross flow microfiltration, which indicates upon longer shelf life of the membrane and thus cost effective processing of industrial oil–water emulsions. The experimental investigations were carried out using synthetic oil-in-water emulsions and hence the evaluated flux and rejection characteristics need not represent real time situations. With real wastewater systems, the interaction of the multi-component feed with the membrane matrix could be significant and flux decline profiles could be quite different. Thus, in the near future, research shall target upon the treatment of real oily wastewater samples. Nomenclature ΔP trans membrane pressure diffrential (kPa) Δt sampling time (min) J permeate flux (m3 m−2 s) V volume of permeate (m3) A effective membrane area (m2) C concentration of oil in the feed (mg/L) Cp concentration of oil in the permeate (mg/L) R FI PWc PWf Jo kb kc ki ks R2 RMS CPB SPB IPB CF Re rejection (dimension less) fouling index (dimension less) correspond to the pure water hydraulic permeability values for cleaned membrane correspond to the pure water hydraulic permeability values for fresh membrane initial permeate flux (m3 m−2 s) complete pore blocking model constant (s−1) cake filtration model constant (s.m−2) intermediate pore blocking model constant (m−1) standard pore blocking model constant (m−0.5 s−0.5) square of correlation coefficient (dimension less) root mean square (dimension less) complete pore blocking standard pore blocking intermediate pore blocking cake filtration Reynolds number