Removal of Escherichia coli as indicator of wastewater pathogens: a comparison between Aerobic Granular Sludge and Activated Sludge laboratory reactors

Abstract

An Aerobic Granular Sludge (AGS) and an Activated Sludge (AS) laboratory-scale reactors started-up with seed sludge originated from the same wastewater, and fed with the same synthetic wastewater containing Escherichia coli, were operated and weekly sampled during four months in order to compare the removal of Escherichia coli between both set-ups. Both reactors were aimed to remove organic matter, nitrogen, and phosphate. The removal of organic matter and phosphate was successful, whereas nitrogen removal efficiency fluctuated in the AGS reactor, and was never achieved in the AS reactor. In order to enhance nitrification, several changes were introduced in the AS cycle, whilst the AGS cycle was maintained unchanged during the entire research period, only varying DO concentrations. E. coli was added to the reactors after 1.5 months of operation, and weekly samples were analysed with the spread-plate technique. The log removal of E. coli for the AGS reactor was above 1.5 for the first five weeks of E. coli addition, presenting a maximum of 3.5 log removal and an average of 2.5 log removal. After the fifth week, the removal dropped to zero, and started to slowly increase again, achieving 2.5 log removal by the end of the research period. The removal of E. coli was mainly attributed to predation of E. coli cells by stalked ciliates of the genus Vorticella during aeration. Vorticella were observed by optical microscopy attached to granules, in high amounts during the periods of high E. coli removal, and in lower amounts when E. coli removal decreased. The factors affecting the fluctuations in the presence of ciliates were analysed, finding as the most relevant an event of high copper concentration and the possible depletion of particulate food eroded from the granules. Analysing the concentration of E. coli in the sludge and supernatant, it was deducted that there was attachment of some E. coli cells to the granules. When E. coli removal was significant, on average, 26% of the influent E. coli ended-up in the sludge and 0.2% in the effluent. E. coli removal in the AS reactor was never achieved. Despite AGS and AS reactors were started with seed sludge originated from the same influent wastewater, ciliates were not detected in the AS reactor on samples taken after two months of operation onwards. It was assumed that the main cause of the absence of ciliates was the lack of particulate food. Attachment of E. coli cells to bacterial flocs was not detected neither, this might be attributed to the high SRT of the reactor. A basic analysis of the applicability of AGS treatment in two case studies in Uruguay was performed, finding that the technology would be in principle suitable for the treatment of the wastewater of both localities aiming to comply the Uruguayan standards. In the case of Mercedes, it seems that an AGS WWTP would result in a more cost-effective solution than a conventional treatment. However, a pre-design and cost analysis should be performed to confirm this hypothesis. Montevideo’s wastewater has a high conductivity, and despite according to lab-scale research, AGS performance would not be affected by its level of salinity, the suitability of the technology for Montevideo’s wastewater would only be confirmed after AGS is proven to be successful in treating wastewater with these characteristics.