On the effect of flocculation as pretreatment process and particle size distribution for membrane fouling reduction

This work deals with the feasibility of an olive vegetation wastewater purification process consisting of two consecutive steps: coagulation followed by membrane separation. The target of this work was to reach a purification grade of the wastewater compatible with the limits for the discharge into a municipal sewer that is a COD value less than 500 mg L(1. A feedstock with 55 g L(1 of COD was used for the experimental work. In a first step, the feedstock was pretreated by a flocculation process using as coagulant aluminium sulphate or aluminium hydroxide. Then, the pretreated wastewater stream was purified by four consecutive batch membrane processes consisting of microfiltration, ultrafiltration, nanofiltration and reverse osmosis, each with a recovery rate of 90%. For each membrane, at different recovery values, the critical flux values were measured. Finally, all the streams were analyzed by means of a nanosizer, capable to measure the suspended particle size distribution in the samples from 0.8 nm up to 6.5 mm. Flocculation changes the fouling outcome on the membranes by changing sensibly the particle size distributions of the solution. The membrane’s performances are strictly dependant on the coagulant type used during the pretreatment operation. In order to develop full advantages on fouling issues, the particle size shift effect of the flocculation process should be carefully studied and evaluated, and needs fine-tuned optimization.

The critical flux values, and thus fouling, are sensibly influenced by the flocculation process due to both solute concentration reductions by sedimentation and particle size shifts. The latter effect needs particular attention and optimization, and influences the fouling outcome on the membranes in relationship to the membrane’s pore size. Therefore, the flocculation process as pretreatment for membrane operation is a very important and useful practice to avoid fouling issues, as long as it is optimized and do not produce aggregates of similar dimension than the pore size. If a particle size shift is not desired, alternative pretreatment processes should be taken into account. Different coagulants perform different particle size shifts, thus should be considered as an important pretreatment process parameter. A fitting equation of the critical flux as function of the particle size permits to perform consistent value estimations without time consuming experimental measurements. This is of great advantage for all membrane process optimization methods and advanced control systems which are based on the critical flux theory.