Abstract
Foaming of activated sludge during wastewater treatment can compromise the quality of treated wastewater (effluent) and create operational difficulties. The presence of high concentrations of mycolic acid-containing bacteria (mycolata) has been linked to foam production. Specifically, the nocardioform group of mycolata is often used as an indicator of bacterial foaming potential during secondary treatment in the activated sludge process, due to their role in foaming and their ease of enumeration. One major goal of nocardioform enumeration is to discover the threshold at which higher nocardioform concentrations will likely result in foaming. Such threshold values can potentially provide an important tool for wastewater treatment plant operators to predict and manage foaming. The purpose of my study was to help treatment plant staff control activated sludge foaming in the most efficient manner possible by comparing the accuracy and precision of four nocardioform counting methods to determine which method best predicts foaming. The second part of my study was to determine the nocardioform concentration threshold for foaming of activated sludge at Sacramento Regional Wastewater Treatment Plant (SRWTP). Four types of nocardioform counts were performed: total and dispersed filament intersection counts and total and dispersed filament length counts. These counts were performed on Gram-stained slides under 1000X magnification using an eyepiece reticule (grid) of a light microscope. The accuracy of each counting method in predicting foaming was assessed through correlation with mixed liquor foam measures. Two independent measures of foam severity were used during this study: (1) foam potential tests conducted in the laboratory and (2) observations of foam coverage over one secondary sedimentation tank. Laboratory foaming tests involved the bubbling of air through samples of activated sludge and measuring the resulting foam height and percent collapse of foam (in the absence of aeration). Precision of the nocardioform counting methods was compared by calculating the relative standard deviation (RSD) of four pseudoreplicate strip counts per method each day, followed by Friedman’s test of RSD mean ranks to test for statistically significant differences in the means. The nocardioform concentration threshold for foaming was calculated for total filament intersections/g volatile suspended solids (VSS). In order to confirm and explain the nocardioform count results, isolation and identification of nocardioforms from samples of activated sludge used for enumeration were attempted during both years of my study via culturing and 16s rDNA sequencing. The majority of my study was conducted at SRWTP over the course of two nocardioform blooms (late summer and fall of 2009 and 2010). All four counting methods showed relatively weak correlations with the mixed liquor foaming measures used during my study, likely due to the contributions to foaming made by non-nocardioform mycolata and possibly surfactants during that time. The accidental isolation of Mycobacterium on two separate occasions during my study supports the notion of non-nocardioform mycolata contributing to activated sludge foaming at SRWTP. When foam stability was used as the criterion for defining the foaming threshold of mixed liquor, the corresponding nocardioform abundance fell into the range historically associated with foaming at SRWTP (1 x 106 total intersections/g VSS). The results of my study indicate that total nocardioform intersection counts most accurately predict activated sludge foaming, compared to the other three counting methods examined. Nocardioform length counts have better precision, but are less accurate at predicting foaming compared to total intersection counts.