Geração de EPS no tratamento de água residuária de refinaria de petróleo em MBR e sua influência em micro e ultrafiltração

Abstract

The production of extracellular polymeric substances (EPS) in membranes bioreactor (MBR) for the removal of organic matter and nitrogen compounds present in wastewater from oil refinery as well as a possible influence of EPS in the hydraulic performance of the membranes was studied under simulated transient and steady states conditions for organic matter/biomass (F/M), volumetric organic load (VOL), nitrogen compounds/biomass (N/M) and volumetric nitrogen load (VNL) in an experimental unit for treatment of these waters that was operated for 2,387 hours and 14 minutes with modules of submerged hollow fiber membranes in PVDF, two of ultrafiltration (W21 and 28) and one of microfiltration (M0). During this period, hydraulic detention time was deliberately varied between 7.8 and 20.1 hours, however, two peaks of 45.7 and 66.4 hours were observed due to the reduction of permeate flow caused by unpremeditated disturbances in the MBR during which it became clear the importance of the mechanisms of backwash, relaxation and membrane aeration to maintain the permeate flux and transmembrane pressure at stable levels that were 17,5 L⋅m-2⋅h-1 and 5,500 to 9,000 Pa for W21 and W28, 14 L⋅m-2⋅h-1 and 17,000 to 20,000 Pa for M0. The performance of the MBR in removing organic matter was above 90% when F/M and VOL was higher than 0.01 gCOD⋅gVSS-1⋅h-1 and 25.0 gCOD⋅m-3⋅h-1, respectively. The oxidation of nitrogenous matter varied between 60 and 98% without clearly defined trend. The concentration of total suspended solids ranged from 10,000 and 4,000 g⋅m-3 and 53% of the analysis this concentration was greater than 7,000 g⋅m-3, marked decreases were observed when the VOL was less than 30.0 gCOD⋅m-3⋅h-1. The concentration of soluble EPS ranged from 4.5 to 250.3 g⋅m-3, from 4.2 to 36.5 g⋅m-3 to the loosely bound EPS and from 60.4 to 461.9 g⋅m-3 to the tightly EPS. Based on these concentrations it is possible to conclude that the loosely bound fraction is not very representative, not justifying it separate extraction from the tightly bound EPS. The centrifugation speed adopted in the extraction of EPS did not provide adequate separation EPS/sludge. The analysis of proteins and humic acids by the method of Lowry et al. (1951) modified by Frolund et al. (1995), based on the results of the standard curve concentration/absorbance, seemed appropriate. It was no possible to establish a clear cause and effect relationship between performance of MBR, EPS and hydraulic performance of the membranes, probably due to disturbances during the operation of the MBR