Full-scale Wastewater Treatment Plant Research Articles

Machine learning facilitated the conceptual design of an alum dosing system for phosphorus removal in a wastewater treatment plant

Wastewater treatment plants (WWTPs) face challenges in controlling total phosphorus (TP), given more stringent regulations on TP discharging. In particular, WWTPs that operate at a small scale lack resources for real-time monitoring of effluent quality. This study aimed to develop a conceptual alum dosing system for reducing TP concentration, leveraging machine learning (ML) techniques and data from a full-scale WWTP containing incomplete TP information. The proposed system comprises two ML models in series: an Alert model based on LightGBM with an accuracy of 0.92, and a Dosage model employing a voting algorithm through combining three ML algorithms (LightGBM, SGD, and SVC) with an accuracy of 0.76. The proposed system has demonstrated the potential to ensure that 88.1% of the effluent remains below the TP discharge limit, which outperforms traditional dosing methods and could reduce overdosing from 61.3 to 12.1%. Furthermore, the SHapley Additive exPlanations (SHAP) analysis revealed that incorporating the output features from the previous cycle and utilizing the results of the Alert model as the input features for dosage prediction could be an effective method for data with limited information. The findings of this study have practical applications in improving the efficiency and effectiveness of TP control in small-scale WWTPs, providing a valuable solution for complying with stringent regulations and enhancing environmental sustainability.

Unveiling the hidden diversity and functional role of Chloroflexota in full-scale wastewater treatment plants through genome-centric analyses

Abstract The phylum Chloroflexota has been found to exhibit high abundance in the microbial communities from wastewater treatment plants (WWTPs) in both aerobic and anaerobic systems. However, its metabolic role has not been fully explored due to the lack of cultured isolates. To address this gap, we use publicly available metagenome datasets from both activated sludge (AS) and methanogenic (MET) full-scale wastewater treatment reactors to assembled genomes. Using this strategy, 264 dereplicated, medium- and high-quality metagenome-assembled genomes (MAGs) classified within Chloroflexota were obtained. Taxonomic classification revealed that AS and MET reactors harbored distinct Chloroflexota families. Nonetheless, the majority of the annotated MAGs (166 MAGs with >85% completeness and < 5% contamination) shared most of the metabolic potential features, including the ability to degrade simple sugars and complex polysaccharides, fatty acids and amino acids, as well as perform fermentation of different products. While Chloroflexota MAGs from MET reactors showed the potential for strict fermentation, MAGs from AS harbored the potential for facultatively aerobic metabolism. Metabolic reconstruction of Chloroflexota members from AS unveiled their versatile metabolism and suggested a primary role in hydrolysis, carbon removal and involvement in nitrogen cycling, thus establishing them as fundamental components of the ecosystem. Microbial reference genomes are essential resources for understanding the potential functional role of uncultured organisms in WWTPs. Our study provides a comprehensive genome catalog of Chloroflexota for future analyses aimed at elucidating their role in these ecosystems.

Niche differentiation in microbial communities with stable genomic traits over time in engineered systems.

Microbial communities in full-scale engineered systems undergo dynamic compositional changes. However, mechanisms governing assembly of such microbes and succession of their functioning and genomic traits under various environmental conditions are unclear. In this study, we used the activated sludge and anaerobic treatment systems of four full-scale industrial wastewater treatment plants as models to investigate the niches of microbes in communities and the temporal succession patterns of community compositions. High-quality representative metagenome-assembled genomes revealed that taxonomic, functional, and trait-based compositions were strongly shaped by environmental selection, with replacement processes primarily driving variations in taxonomic and functional compositions. Plant-specific indicators were associated with system environmental conditions and exhibited strong determinism and trajectory directionality over time. The partitioning of microbes in a co-abundance network according to groups of plant-specific indicators, together with significant between-group differences in genomic traits, indicated the occurrence of niche differentiation. The indicators of the treatment plant with rich nutrient input and high substrate removal efficiency exhibited a faster predicted growth rate, lower guanine-cytosine content, smaller genome size, and higher codon usage bias than the indicators of the other plants. In individual plants, taxonomic composition displayed a more rapid temporal succession than functional and trait-based compositions. The succession of taxonomic, functional, and trait-based compositions was correlated with the kinetics of treatment processes in the activated sludge systems. This study provides insights into ecological niches of microbes in engineered systems and succession patterns of their functions and traits, which will aid microbial community management to improve treatment performance.

Recuperação de biopolímeros de lodo biológico e seu uso como biossorvente de fósforo

ABSTRACT Biopolymers can be recovered from aerobic sludge and used for environmental applications, such as phosphorus adsorbent material, instead being sent to sanitary landfills. In this resource recovery perspective, this work aimed to study the recovery of alginate-like exopolymer (ALE) from activated sludge (AS) compared to aerobic granular sludge (AGS). ALE-based biosorbent was prepared and tested to remove phosphorus from aqueous solutions, and the adsorption kinetics and isotherm models were studied. The recovery yield of ALE from a full-scale wastewater treatment plant (WWTP) AS (18.7%) was close to that obtained from a pilot-scale AGS (22%). ALE recovered from AS presented hydrogel properties and humic substances in its composition, which are important features for future applications. The equilibrium of the adsorption was reached after 10 minutes. The Langmuir isotherm model and the PFO kinetic model best fitted to the experimental data, resulting in maximum adsorption capacity of 8.164 mgP·gALE. Thus, ALE recovered from AS has the potential to be used as a phosphorus biosorbent from effluents and further used as a nutrient delivery system with hydrogel properties.

An advanced control system for nitrogen removal and energy consumption optimization in full-scale wastewater treatment plants

The development and implementation of advanced control systems enable the successful and efficient conversion of wastewater treatment plants into water resource recovery facilities. This work presents an advanced control system based on low-cost sensors (pH, ORP and DO) which enhances biological nitrogen removal and improves energy efficiency. It consists of three independent fuzzy-logic-based control loops (for nitrification, simultaneous nitrification-denitrification, and denitrification, respectively) combined with a set of knowledge-based rules. The control system was implemented in three facilities and test-run for over two years. During this time, proper process performances were achieved both in terms of water quality and of operating cost savings. Energy consumption, calculated as kWh per kg COD equivalent removed (including nitrogen removal), was reduced between 14 % and 33 %.

A comprehensive investigation to the fate of phosphorus in full-scale wastewater treatment plants using aluminum salts for enhanced phosphorus removal

This study investigated the fate of phosphorus (P) in 8 full-scale municipal wastewater treatment plants (WWTPs) in Shanghai, China, in which both biological nutrient removal and aluminum-based chemical phosphorus removal were used. The results showed that 83.8–98.9 % P was transferred to the sludge in the 8 WWTPs by both chemical and biological reactions. P speciation analysis indicated that chemical P precipitates accounted for 84.3 % in the activated sludge, of which crystalline AlPO4 and amorphous iron‑phosphorus compounds (FePs) were the main components. Sludge with more water-soluble and weakly adsorbed P was generated in the anaerobic-anoxic-oxic (A/A/O) process than in other processes. Among the 8 WWTPs, the one with the largest flow rate and relatively short sludge retention time (SRT) had the best potential to release P from all types of sludge. The recovery potential of P from thickened sludge can be improved by separately thickening the sludge produced in the high-efficiency sedimentation tank or feeding it into the dewatering process directly. Different P removal chemicals and dosing points changed the amount of chemical precipitate formed but had little effect on the composition of P accumulating organisms (PAOs) at the genus level. Although aluminum-based coagulants were applied in the investigated WWTPs, Fe in wastewater had the most positive effect on the proliferation of PAOs. The synthesis of polyphosphate was also related to the metabolism of PAOs as it affected transmembrane inorganic phosphate (Pi) transport and polyhydroxybutyrate (PHB) synthesis. The in-depth understanding of the fate of P is beneficial to improve P recovery efficiency in WWTPs.

Correlations between microbial taxonomies and wastewater quality parameters in a full-scale petroleum refinery wastewater treatment plant

In this study, we investigated the correlations between microbial taxonomies and wastewater properties in a full-scale petroleum refinery wastewater biotreatment system, comprising one anoxic tank and two aerobic tanks. Over one hundred organic compounds, mostly aromatics, were detected in the anoxic tank influent, with dominants being cresols and anilines. While the biotreatment process effectively removed most organic components, n-alkanes remained in the effluent of the second aerobic tank. Thauera and Ignavibacterium were the dominant genera across all tanks. Betaproteobacteria substantially decreased in the second aerobic tank where the aromatic compounds were depleted. The correlation network displayed two distinct microbial clusters. Genera in cluster I were generally positively correlated with each other and tended to co-occur, while genera in cluster II were subdivided into two small groups that were negatively correlated with the other group members. Mantel test showed that pH, nitrite and n-alkanes were potential determining factors influencing the overall community assembly. This study provides valuable insights into the chemical changes and microbial taxonomy in a complex petroleum refinery wastewater treatment system, establishing a foundation for further studies on the community regulation strategies within similar biotreatment systems.

Denitrification Capacity of Volatile Fatty Acids from Sludge Fermentation: Lab-Scale Testing and Full-Scale Assessment

This work provides insights into the possibility of integrating recovered volatile fatty acids (VFAs) into biological nitrogen removal processes. VFAs are the main products of the acidogenic fermentation of waste sludge and are an effective carbon source for denitrification in activated sludge processes. The assessment of denitrification rates and the utilisation hierarchy of different VFAs are relevant to evaluating the possibility of replacing external carbon sources with the fermented liquid, FL, from acidogenic fermentation. To this scope, single VFAs, FL collected from a full-scale waste sludge fermenter, and commercial hydroalcoholic solutions have been tested with manometric lab-scale tests. Regarding single acids, acetic acid showed the highest denitrification rates, up to 4 mg N-NO3 g VSS−1 h−1, while more complex acids usually showed a lower denitrification rate. The synthetic VFA mixture and FL showed a higher denitrification rate than the sole acetate (up to 134% of the acetate denitrification rate). Mass balances across the full-scale wastewater treatment plant demonstrated the positive role of FL dosage in enhancing the denitrification process in the activated sludge treatment, with an average nitrogen removal equal to 57% and 78% without and with FL dosage, respectively. Batch manometric tests proved to be an efficient and reliable tool to assess the quality of the carbon sources as well as the activity of denitrifying bacteria in activated sludge samples.

Seasonality of nitrous oxide emissions at six full-scale wastewater treatment plants.

Nitrous oxide (N2O) is an ozone-depleting greenhouse gas that contributes significantly to the carbon footprint of a wastewater treatment plant (WWTP). Plant-specific measurement campaigns are required to reliably quantify the emission level that has been found to significantly vary between WWTPs. In this study, the N2O emissions were quantified from five full-scale WWTPs during 4-19-day measurement campaigns conducted under both cold period conditions (water temperature below 12 °C) and warm period conditions (water temperature from 12 to 20 °C). The measurement data were studied alongside long-term monitoring data from a sixth WWTP. The calculated emission factors (EFs) varied from near 0 to 1.8% relative to the influent total nitrogen load. The results confirmed a significant seasonality of N2O emissions as well as a notable variation between WWTPs in the emission level, which a single fixed EF cannot represent. Wastewater temperature was one explanatory factor for the emission seasonality. Both low and high emissions were measured from denitrifying-nitrifying activated sludge (AS) processes, while the emissions from only nitrifying AS processes were consistently high. Nitrite (NO2-) at the end of the aerobic zones of the AS process was linked to the variability in N2O emissions during the cold period.

Tracking the formation potential of vivianite within the treatment train of full-scale wastewater treatment plants

Phosphorus recovery is a vital element for the circular economy. Wastewater, especially sewage sludge, shows great potential for recovering phosphate in the form of vivianite. This work focuses on studying the iron, phosphorus, and sulfur interactions at full-scale wastewater treatment plants (Viikinmäki, Finland and Seine Aval, France) with the goal of identifying unit processes with a potential for vivianite formation. Concentrations of iron(III) and iron(II), phosphorus, and sulfur were used to evaluate the reduction of iron and the formation potential of vivianite. Mössbauer spectroscopy and X-ray diffraction (XRD) analysis were used to confirm the presence of vivianite in various locations on sludge lines. The results show that the vivianite formation potential increases as the molar Fe:P ratio increases, the anaerobic sludge retention time increases, and the sulfate concentration decreases. The digester is a prominent location for vivianite recovery, but not the only one. This work gives valuable insights into the dynamic interrelations of iron, phosphorus, and sulfur in full-scale conditions. These results will support the understanding of vivianite formation and pave the way for an alternative solution for vivianite recovery for example in plants that do not have an anaerobic digester.

Tapping the potential of anammox upgrading in full-scale wastewater treatment plants through biofilm-enhanced anammox and denitrification

Implementing the Anammox process at a large scale remains a significant challenge. However, this study presents a potential approach for upgrading full-scale wastewater treatment plants (WWTPs) to incorporate the Anammox process in treating ammonium-rich wastewater. The research focuses on a WWTP at a duck breeding farm that utilizes an anaerobic/microaerobic/anoxic/oxic (A/M/A/O) process. With 83% removal of total nitrogen, the contribution of anammox and denitrification nitrogen removal in the entire process was 52% and 31%, respectively. The study tracks the changes in water quality and analyzes the potential of simultaneous Anammox-denitrification processes to mitigate nitrate accumulation. Porous carriers were introduced into the primary anaerobic tank (Anaerobic T1) and the primary microaerobic tank (Microaerobic T1) and were harvested on the 90th day. The results revealed that the specific Anammox activity (SAA) of the microaerobic biofilm reached 252.06 mg N/g VSS·d, which was 6.0 times and 3.7 times higher than that of suspended sludge and anaerobic biofilm, respectively. The microaerobic biofilm demonstrated the capacity for simultaneous Anammox-denitrification, with anammox and denitrification rates of 65.35 and 36.3 nmol N/g·h, respectively. These rates were significantly higher than those observed in the anaerobic biofilm and suspended sludge. Furthermore, the main partial denitrification bacteria identified were Ignavibacterales UTCHB1 (57.75%) and Trichococcus (5.12%), which contributed nitrite to Candidatus Brocadia (1.54%). The findings highlight the importance of creating a suitable microenvironment using porous carriers, which facilitates the Anammox pathway for nitrogen removal in biofilms. This approach presents a potential retrofitting and upgrading strategy for existing WWTPs.

Mobility, bacterial hosts, and risks of antibiotic resistome in submicron bioaerosols from a full-scale wastewater treatment plant

Antibiotic resistome could be loaded by bioaerosols and escape from wastewater or sludge to atmosphere environments. However, until recently, their profile, mobility, bacterial hosts, and risks in submicron bioaerosols (PM1.0) remain unclear. Here, metagenomic sequencing and assembly were employed to conduct an investigation of antibiotic resistome associated with PM1.0 within and around a full-scale wastewater treatment plant (WWTP). More subtypes of antibiotic resistant genes (ARGs) with higher total abundance were found along the upwind-downwind-WWTP transect. ARGs in WWTP-PM1.0 were mainly mediated by plasmids and transposases were the most prevalent mobile genetic elements (MGEs) co-occurring with ARGs. A contig-based analysis indicated that very small proportions (15.32%–19.74%) of ARGs in WWTP-PM1.0 were flanked by MGEs. Proteobacteria was the most dominant host of ARGs. A total of 28 kinds of potential pathogens, such as Pseudomonas aeruginosa and Escherichia coli, carried multiple ARG types. Compared to upwind, WWTP and corresponding downwind were characterized by higher PM1.0 resistome risk. This study emphasizes the vital role of WWTPs in discharging PM1.0-loaded ARGs and antibiotic resistant pathogens to air, and indicates the need for active safeguard procedures, such as that employees wear masks and work clothes, covering the main emission sites, and collecting and destroying of bioaerosols.

DOM and DON transformation in full-scale wastewater treatment plants: Comparison of autotrophic and heterotrophic nitrogen removal units

There is limited knowledge regarding the differences and relationships in molecular composition between dissolved organic matters (DOM) and dissolved organic nitrogen (DON) in autotrophic and heterotrophic nitrogen removal processes. In this study, the composition of DOM and DON during two full-scale wastewater treatment plants (WWTPs), an autotrophic nitrogen removal with partial nitrification-anammox process (PN/A) and a heterotrophic nitrogen removal with anaerobic/anoxic/oxic process (A2O), were systematically investigated by using ultraviolet–visible spectroscopy, Excitation-Emission Matrix (EEM) and Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). The results indicated that both PN/A and A2O resulted a decrease in molecular diversity and bio-availability of DOM. However, there were distinct variations in DON between the two processes. In PN/A, transformation of DOM and DON was found to be coupled, while there was a lack of correlation observed between DOM and DON in A2O. Pearson correlation analysis revealed that transformation of DOM and DON in WWTPs was associated with microbial roles. During biotreatment, PN/A exhibited a higher metabolic capacity for degrading high bio-available and high aromatic compounds compared to A2O. The autotrophic nitrogen removal unit in PN/A resulted in an increase in molecular weight (MW) of DON, as well as a decrease in bio-availability and unsaturation degree of DON. Therefore, autotrophic nitrogen removal unit may have a positive influence on control of eutrophication and disinfection by-products (DBPs) formation. This study improved the understanding of molecular composition and relationship between DOM and DON, as well as microbial roles on DON transformation within autotrophic and heterotrophic nitrogen removal processes.

Microplastics shape microbial interactions and affect the dissemination of antibiotic resistance genes in different full-scale wastewater treatment plants

Wastewater treatment plants (WWTPs) pose a potential threat to the environment because of the accumulation of antibiotic resistance genes (ARGs) and microplastics (MPs). However, the interactions between ARGs and MPs, which have both indirect and direct effects on ARG dissemination in WWTPs, remain unclear. In this study, spatiotemporal variations in different types of MPs, ten ARGs (sul1, sul2, tetA, tetO, tetM, tetX, tetW, qnrS, ermB, and ermC), class 1 integron integrase (intI1) and transposon Tn916/1545 in three typical WWTPs were characterized. Sul1, tetO, and sul2 were the predominant ARGs in the targeted WWTPs, whereas the intI1 and transposon Tn916/1545 were positively correlated with most of the targeted ARGs. Saccharimonadales (4.15 %), Trichococcus (2.60 %), Nitrospira (1.96 %), Candidatus amarolinea (1.79 %), and SC-I-84 (belonging to phylum Proteobacteria) (1.78 %) were the dominant genera. Network and redundancy analyses showed that Trichococcus, Faecalibacterium, Arcobacter, and Prevotella copri were potential hosts of ARGs, whereas Candidatus campbellbacteria and Candidatus kaiserbacteria were negatively correlated with ARGs. The potential hosts of ARGs had a strong positive correlation with polyethylene terephthalate, silicone resin, and fluor rubber and a negative correlation with polyurethane. Candidatus campbellbacteria and Candidatus kaiserbacteria were positively correlated with polyurethane, whereas potential hosts of ARGs were positively correlated with polypropylene and fluor rubber. Structural equation modeling highlighted that intI1, transposon Tn916/1545 and microbial communities, particularly microbial diversity, dominated the dissemination of ARGs, whereas MPs had a significant positive correlation with microbial abundance. Our study deepens the understanding of the relationships between ARGs and MPs in WWTPs, which will be helpful in designing strategies for inhibiting ARG hosts in WWTPs.

Investigating the viral ecology and contribution to the microbial ecology in full-scale mesophilic anaerobic digesters

In an attempt to assess the diversity of viruses and their potential to modulate the metabolism of functional microorganisms in anaerobic digesters, we collected digestate from three mesophilic anaerobic digesters in full-scale wastewater treatment plants treating real municipal wastewater. The reads were analyzed using bioinformatics algorithms to elucidate viral diversity, identify their potential role in modulating the metabolism of functional microorganisms, and provide essential genomic information for the potential use of virus-mediated treatment in controlling the anaerobic digester microbiome. We found that Siphoviridae was the dominant family in mesophilic anaerobic digesters, followed by Myoviridae and Podoviridae. Lysogeny was prevalent in mesophilic anaerobic digesters as the majority of metagenome-assembled genomes contained at least one viral genome within them. One virus within the genome of an acetoclastic methanogen (Methanothrix soehngenii) was observed with a gene (fwdE) acquired via lateral transfer from hydrogenotrophic methanogens. The virus-mediated acquisition of fwdE gene enables possibility of mixotrophic methanogenesis in Methanothrix soehngenii. This evidence highlighted that lysogeny provides fitness advantage to methanogens in anaerobic digesters by adding flexibility to changing substrates. Similarly, we found auxiliary metabolic genes, such as cellulase and alpha glucosidase, of bacterial origin responsible for sludge hydrolysis in viruses. Additionally, we discovered novel viral genomes and provided genomic information on viruses infecting acidogenic, acetogenic, and pathogenic bacteria that can potentially be used for virus-mediated treatment to deal with the souring problem in anaerobic digesters and remove pathogens from biosolids before land application. Collectively, our study provides a genome-level understanding of virome in conjunction with the microbiome in anaerobic digesters that can be used to optimize the anaerobic digestion process for efficient biogas generation.

Optimization of Fe(II)/SPC Sludge Conditioning and Dewatering Process Based on Response Surface Methodology: Full-Scale Application

Sludge conditioning is a crucial step in sludge dewatering aimed at minimizing excessive sludge production. The Fenton process, which harnesses oxidative radicals to dismantle extracellular polymeric substances (EPS) and microorganisms, has been unequivocally proven to enhance sludge dewaterability. However, the widespread adoption of the Fenton process is hampered by its high costs and logistical challenges in transportation. In contrast, the Fe(II)-activated sodium percarbonate (Fe(II)/SPC) process has emerged as a promising technology for sludge conditioning due to its remarkable performance and safe operation. However, limited information is available regarding the optimization of Fe(II)/SPC for sludge conditioning and dewatering at full scale. This study conducted the sludge conditioning and dewatering process within a full-scale wastewater treatment plant, utilizing the response surface methodology (RSM) to optimize the Fe(II)/SPC process. Furthermore, this study investigated its impact on sludge structure and compared the economic benefits of the Fe(II)/SPC process with other full-scale conditioning processes. The results of bound water and LDH analysis revealed that the Fe(II)/SPC process not only degraded EPS but also disrupted microbial cells, thereby releasing intracellular water. Based on the RSM results, we successfully established a polynomial prediction model to determine the optimal capillary suction time (CST) and moisture content. The optimal parameters determined through RSM were an initial pH of 3.02, Fe(II) dosage of 0.05 g/g TSS, and SPC dosage of 0.07 g/g TSS. The validation test confirmed the accuracy of the prediction results, with the conditioned sludge exhibiting a CST of 31.6 s and a moisture content of 51.47%. Furthermore, when compared to the PFS and Fenton processes, the Fe(II)/SPC process demonstrated higher economic efficiency and safety, while maintaining good dewatering performance. Overall, the Fe(II)/SPC treatment shows promise as a prospective sludge dewatering and conditioning process.

Multi-step and multi-task learning to predict quality-related variables in wastewater treatment processes

In wastewater treatment processes, lack of hardware sensors together with unacceptable dynamics, strong nonlinearity and large time delay often leads to a large number of key variables that are difficult to measure online accurately and timely, then frustrate safe operations of the processes. To accurately and timely capture the short-term behavior changes and trend development of critical variables, a novel neural network based soft-sensing model is proposed to take full use of multi-task learning, direct multi-step prediction strategy and evolutionary algorithm to formulate a novel multi-task multi-step evolution (MTMSE) neural network. Firstly, single-output MTMSE (SO-MTMSE) neural network is used to realize the dynamic monitoring of a single variable. Moreover, by considering the spatiotemporal interaction among the data, the model is extended to multi-output MTMSE (MO-MTMSE) neural network to simultaneously realize multi-step prediction of multiple variables, thus providing a desired reference for optimizing the wastewater treatment processes. Finally, the proposed model is applied to the benchmark simulation model 2 (BSM2) and a full-scale wastewater treatment plant in Shenzhen. And the results show that the proposed dynamic soft sensor model outperforms the standard methods, such as autoregressive moving average model (ARMA) and multiple output gaussian process regression (MGPR).

Fluorescence spectroscopic characterization of dissolved organic matter in the wastewater treatment plant and hybrid constructed wetlands coupling system in winter: A case study in eastern China

Dissolved organic matter (DOM) in municipal wastewater can have a detrimental effect on aquatic ecosystems, as DOM can participate in the global carbon cycle and change the transport of other pollutants. The characteristics of DOM can also affect the utilization efficiency of dissolved organic carbon by microorganisms in water. In this study, variations in DOM quantity and quality in a full-scale wastewater treatment plant (WWTP) coupled with a hybrid constructed wetland (CW) system during cold periods were characterized by using excitation-emission matrix spectroscopy combined with parallel factor analysis (EEM-PARAFAC). The results showed that the DOM concentration significantly decreased after exiting the WWTP, while a slight increase was found in the CW. The coupled system could significantly eliminate influent DOM with a removal efficiency of 92.1%. The CW system could still remove DOM even in winter. Five fluorescent components, including three humic-like components, one protein-like component and one tyrosine-like component, were identified. Although the biological treatment process in the WWTP contributed a large number of protein-like substances, the optical indices showed that the bioavailability of wastewater was gradually reduced, indicating that DOM was selectively removed by the integrated system. Moreover, there were significant correlations between water quality parameters and DOM indices. This study could contribute to a better understanding of the elimination process of dissolved organic carbon in coupled systems of WWTPs and CWs and offer novel ideas for organic carbon resource recovery and water quality monitoring in watershed environments.

Reduction of sewage sludge and N2O emissions by an Oxic Settling Anaerobic (OSA) process: The case study of Corleone (Italy) wastewater treatment plant

Biosolid management is becoming one of the most crucial issues for wastewater treatment plant (WWTP) operators. The application of the Oxic Settling Anaerobic (OSA) process allows the minimisation of excess sludge production. This study compares conventional activated sludge (CAS) and OSA layouts in a full-scale WWTP (namely, Corleone - Italy). Extensive monitoring campaigns were conducted to assess treatment performances regarding carbon and nutrient removal, greenhouse gas (GHG) emissions, excess sludge production, and biomass activity (by means of respirometric analysis). Results showed that the effluent quality consistently met the Italian discharge limits. However, with the implementation of the OSA process, there was a decrease in ammonium removal efficiency, which could be attributed to reduced nitrifier activity related to reduced biomass production and extended anaerobic conditions affecting the nitrification process. On the other hand, the OSA configuration significantly increased phosphorus removal, indicating a high phosphorus content in the resulting waste sludge. A worsening of the sludge settling properties was observed with the OSA configuration likely due to decreased EPS concentrations. The sludge production in the OSA configuration decreased by 17.3 % compared to CAS. Nitrous-oxide measurements did not show a variation between CAS and OSA configurations, confirming that the OSA process can be a suitable solution for reducing WWTP's carbon footprint.

Prediction of biological nutrients removal in full-scale wastewater treatment plants using H2O automated machine learning and back propagation artificial neural network model: Optimization and comparison

The effective control of total nitrogen (ETN) and total phosphorus (ETP) in effluent is challenging for wastewater treatment plants (WWTPs). In this work, automated machine learning (AutoML) (mean square error = 0.4200 ∼ 3.8245, R2 = 0.5699 ∼ 0.6219) and back propagation artificial neural network (BPANN) model (mean square error = 0.0012 ∼ 6.9067, R2 = 0.4326 ∼ 0.8908) were used to predict and analyze biological nutrients removal in full-scale WWTPs. Interestingly, BPANN model presented high prediction performance and general applicability for WWTPs with different biological treatment units. However, the AutoML candidate models were more interpretable, and the results showed that electricity carbon emission dominated the prediction. Meanwhile, increasing data volume and types of WWTP hardly affected the interpretable results, demonstrating its wide applicability. This study demonstrated the validity and the specific advantages of predicting ETN and ETP using H2O AutoML and BPANN model, which provided guidance on the prediction and improvement of biological nutrients removal in WWTPs.