Development of image analysis techniques as a tool to detect and quantify morphological changes in anaerobic sludge:...

Image analysis techniques are applied to monitor the morphological changes in granular sludge present in an expanded granular sludge blanket (EGSB) reactor fed with oleic acid. Deterioration of granular sludge was monitored along the trial period by measuring the percentage of aggregates smaller than 1 mm (in terms of Feret diameter) either in terms of projected area or in terms of number of aggregates. A good correlation was obtained between these values and the percentage of aggregates smaller than 1 mm were physically sorted and quantified by the volatile suspended solid content. The ratio of total filaments length to cross-sectional area of aggregates defined as LfA, was applied to quantify the dispersion level of the granular sludge, which increased until day 141 and remained almost invariant afterwards. LfA was sensitive to the sludge deterioration process and was able to indicate, with the anticipation of about 1 month, the most significant biomass washout episode that occurred in the trial period. A mechanism of filaments' release, detachment and selective washout was proposed to explain the action of LfA from this viewpoint. The equivalent diameter of the bottom aggregates larger than 1 mm increased with the increase on the amount of long chain fatty acids associated with the biomass by mechanisms of adsorption, precipitation, or entrapment. After a threshold value of about 200 mg COD-LCFA gVSS (COD = chemical oxygen demand; LCFA = long chain fatty acids; VSS = volatile suspended solids), a migration of granular sludge from the bottom to a top-floating layer was evident.