Development of a model to predict bulking in full-scale wastewater treatment plants, and the impact of bulking in the receiving environment

Abstract

Sludge bulking has been a continuous operational hurdle affecting the solids-liquid separation in wastewater treatment plants (WWTPs) worldwide. Excessive growth of filamentous bacteria is the primary and common cause of sludge bulking, which can have negative impacts on the wastewater treatment efficiency. Filamentous bacteria serve as the backbone structure of flocs which assist in the sludge settling process, however, their prolific growth result in slow sedimentation due to inadequate settling of flocs. The main focus of this research was to develop a model to assist in a clearer understanding of the bulking sludge phenomenon in relation to filamentous bacterial growth and to identify predictors of bulking in different biological nutrient removal (BNR) WWTPs. The growth of filamentous bacteria and sludge bulking in different WWTPs and its association with sludge bulking incidents were evaluated using different statistical models [viz. artificial neural networks (ANN), principal component analysis (PCA), cluster analysis and Decision Trees]. In addition, the effect of bulking on pathogen discharge and its potential impact on the community was assessed using a microbial risk assessment model. A total of seven WWTPs were investigated to identify the most common and dominant filamentous bacteria during bulking and non-bulking periods. A total of ten filamentous bacterial species were identified in this study with their dominance varying across the selected WWTPS during the sampling period. Based on the filament index scale ranging from 1 (None filament) to 7 (Excessive filament), the developed ANN model predicted sludge volume index (SVI) in relation to the abundances of ten filamentous species as model inputs. Among the filamentous bacteria identified, Eikelboom Type 0041 attained the highest impact on SVI, followed by Gordonia spp., Nostocoida limicola, and Thiothrix spp. Developing a model for a WWTP, with proper calibration and validation against plant operational data, can allow for proper evaluation of filamentous bacteria associated to bulking, with effective mitigating strategies. Hence, in this study, a Decision Tree model was further implemented as a novel approach in the form of a case study to evaluate the effect of influent wastewater characteristics and plant operational parameters on the dominant filamentous bacteria and sludge bulking for prediction and control. Various factors such as pH, temperature, dissolved oxygen (DO), sludge retention time (SRT), food-to-microorganisms (F/M) ratio, soluble chemical oxygen demand (sCOD), total COD (tCOD), NH4 + -N, total Kjeldahl nitrogen (TKN), phosphorus as phosphate (PO4 3- -P), TP, and total suspended solids (TSS) were considered to have an impact on filamentous dominance. High bulking incidents were observed during long SRT and nutrient deficient (low F/M) conditions. However, a negative correlation was observed with soluble sCOD and ammoniumnitrogen (NH4-N). Type 0092 was the dominant species largely responsible for sludge bulking in the selected plants, which prevailed at low F/M (< 0.08 kg COD/kg MLSS d-1 ) conditions. The secondary filaments Candidatus Microthrix parvicella increased in their abundance at low temperature (< 15.5°C), causing an increase in SVI at lower ambient temperatures. In addition, an increase on Thiothrix spp. was linked with the unbalanced ratio between readily biodegradable COD and nutrient conditions. The last objective of this study provided an assessment from an environmental health perspective, by investigating the impacts of bulking on the receiving environment, using a quantitative microbial risk assessment (QMRA) approach. This was done by studying the difference in selected microbial pathogen abundance during bulking and non-bulking conditions using qPCR. Salmonella was the most dominant species of the investigated microorganisms, during the study period (2270– 96733 copies ng-1 of DNA) followed by E. coli (4133 – 76847 copies per ng of DNA); whereas, Mycobacterium was the least (542 – 3340 copies ng-1 of DNA). During high bulking with SVI >200 mL g-1 , positive correlations were found between the selected pathogens in the final effluent. The QMRA model was applied to investigate the safety of treated effluent for (a) children, women, and men during recreational activities, (b) farmers during irrigation practices, and (c) consumers of edible plants (vegetables). The QMRA values during all bulking events exceeded the tolerable risk of 10-4 (i.e. less than one case of infection per 10 000 people) per year, as recommended by the world health organization (WHO). In addition various disinfection scenarios such as chlorination, ultraviolet (UV) and ozonation were tested to control the risks associated with pathogenic bacteria, for further information of safe disposal and reuse of the treated effluent. The application of UV provided the most effective treatment to reduce the pathogenic bacteria, except for the case of children that were exposed to Salmonella infection. To the best of my knowledge, the probable health risks associated with the discharge or reuse of WWTPs effluents under different sludge bulking events have not yet been systematically evaluated using QMRA. This research can potentially lead to the development of appropriate model systems for bulking control in full-scale WWTPs, while highlighting some of the significant contributors, environmental impact and mitigation strategies. The outcomes of this research will contribute to the current global body of knowledge in relation to predictive models for filamentous bulking control in full-scale WWTPs.