Membrane fouling during the harvesting of microalgae using static microfiltration

Abstract

A challenge in the filtration of algal suspensions is the reduction in filtrate flow rate with time, associated with membrane fouling and filtercake formation. These two fouling mechanisms may occur simultaneously and have different effects on the scaling relationship of filtration rate with time, making quantitative predictions of the filtration behaviour complex. The foulants include suspended microalgal cells and their presumed Extracellular Polymeric Substances (EPS). To investigate this problem, we perform static microfiltration experiments to harvest oil-rich marine microalgae Nannochloropsis oculata. Batch filtration experiments of microalgal suspensions using glass-fibre membranes are conducted under filtration pressures varying between 0.5 and 200 kPa. Here we investigate the relative importance of potential fouling mechanisms, including pore plugging, entrance blocking, and filtercake formation. We examine variations in filtrate flux, using a novel approach involving numerical differentiation of the filtrate volume to identify how its scaling relationship with time and compare with the traditional root-time behaviour. These results are analysed with reference to the blocking filtration laws to determine the potential fouling mechanism. Rheology tests of both filtration feeds and filtrate are performed, and both optical and scanning electron microscopy are used to observe the filtercake. Our results show a significant drop in the filtrate flux after a spurt loss phase under pressure. The scaling analysis demonstrates a power law relationship between cumulative filtrate volume and time in the post-spurt phase. We show here, for the first time, that the scaling exponent varies with time, approaching a value close to 0.5 (i.e. the root-time behaviour) after a long period of time. Using the analysis, we conclude that the filtration process can be divided into various stages; a spurt loss phase when the pressure is initially applied, followed by a stage in which both internal and external filtercake formation occurs, and a final stage which is dominated by external filtercake formation (at which point root-time behaviour is observed). Furthermore, our findings suggest that fresh microalgal suspensions experienced a larger spurt loss compared to aged suspensions in microfiltration. This difference may be attributed to the limited production of EPS and microalgal debris during shorter cultivation periods. The microscopic observations reveal the invasion of microalgal cells into the membrane, which verifies the formation of an internal filtercake suggested by our scaling analysis. The contamination of the membrane increases with higher filtration pressures. Finally, we demonstrate that Fe3+ coagulant can be used to increase the microalgal particulate size before filtration, resulting in a much higher filtrate flux compared to uncoagulated suspensions. This result provides opportunities to reduce membrane contamination to a negligible level.