Effluent parameters prediction of a biological nutrient removal (BNR) process using different machine learning methods: A case study.

This paper proposes a novel targeted blend of machine learning (ML) based approaches for controlling wastewater treatment plant (WWTP) operation by predicting distributions of key effluent parameters of a biological nutrient removal (BNR) process. Two years of data were collected from Plajyolu wastewater treatment plant in Kocaeli, Türkiye and the effluent parameters were predicted using six machine learning algorithms to compare their performances. Based on mean absolute percentage error (MAPE) metric only, support vector regression machine (SVRM) with linear kernel method showed a good agreement for COD and BOD5, with the MAPE values of about 9% and 0.9%, respectively. Random Forest (RF) and EXtreme Gradient Boosting (XGBoost) regression were found to be the best algorithms for TN and TP effluent parameters, with the MAPE values of about 34% and 27%, respectively. Further, when the results were evaluated together according to all the performance metrics, RF, SVRM (with both linear kernel and RBF kernel), and Hybrid Regression algorithms generally made more successful predictions than Light GBM and XGBoost algorithms for all the parameters. Through this case study we demonstrated selective application of ML algorithms can be used to predict different effluent parameters more effectively. Wider implementation of this approach can potentially reduce the resource demands for active monitoring the environmental performance of WWTPs.

Application of combined EO/PMS/Me2+ process in organic matter and true color removal from paint manufacturing industry wastewater.

Treatment of paint manufacturing industry wastewater by electrooxidation (EO) process in which peroxymonosulfate (PMS) and transition metals are added was investigated. In the EO/PMS process, graphite was the cathode while different anode materials (Ti/IrO2, Ti/RuO2, and Ti/SnO2) were used. The anode with the highest chemical oxygen demand (COD) and true color removal efficiency was selected. To determine the catalyst effect on the process, different transition metals (Fe2+, Cu2+, Zn2+) were added and Fe2+ was chosen as the catalyst which provided higher removal efficiency and lower cost. The central composite design was applied for the optimization of the process variables of the EO/PMS/Fe2+ process. Current density, PMS dose, Fe2+ dose, and reaction time were process variables whereas COD and true color removal efficiency were system responses. Under optimum conditions (200 A/m2 current density, 14 mM PMS dose, 2.5 mM Fe2+ dose, 60 min reaction time), the estimated COD and true color removal efficiency by the model were 74.89% and 99.86%, respectively. The experimentally obtained COD and true color removal efficiencies as a result of validation studies were 74.28% and 99.03%, respectively. Quenching experiments showed that hydroxyl and sulfate radicals were both involved in the process.

A comprehensive techno-economic analysis of income-generating sources on the conversion of real sheep slaughterhouse waste stream into valorized by-products.

The present analysis was conducted as the first research to assess the techno-economic viability of the value-added by-products (struvite, blood meal, bone meal, and raw sheepskin) from a medium-scale sheep slaughterhouse facility with a slaughtering capacity of 300 sheep per day. For this aim, a comparative technical and economic feasibility analysis was performed to assess the synergistic use of slaughterhouse-oriented rendering wastes and struvite recovery from real sheep abattoir effluent within the framework of detailed cost breakdown, break-even point, and payback period analyses. The experimental findings clearly showed that under the optimal conditions (chemical combination of MgCl2.6H2O + NaH2PO4.2H2O, a molar ratio of Mg2+:NH4+-N:PO43--P = 1.2:1:1, a reaction pH of 9.0, an initial ammonium concentration of 240 mg NH4+-N/L, and a reaction time of 15 min), struvite precipitation could effectively remove about 73%, 64%, 59%, and 82% of NH4+-N, TCOD, SCOD, and color, respectively, from the real sheep slaughterhouse waste stream. Based on various up-to-date techno-economic items considered within the break-even point analysis, the sheep slaughterhouse facility was estimated to achieve the targeted net income (€100/day) for any selling prices of €1041.30/ton, €640.05/ton, €263.72/ton, and €1.012/hide, respectively, for struvite, blood meal, bone meal, and raw sheepskin. Steel construction and chemicals were determined as the most costly components for CAPEX (capital expenditures) and OPEX (operating expenditures), respectively, and selling prices of bone meal and raw sheepskin were found to be the most critical income items on the profitability of the slaughterhouse facility. Co-monetary assessment of the struvite process and valorized compounds corroborated the economic viability of the proposed project with the payback periods of about 6.3 and 5.5 years, respectively, for the current market and the profit-oriented conditions without subsidy. The findings of this feasibility analysis, as the first of its own, could be used as guideline for simplifying the decision-making with regards to the feasibility of similar facilities and commercialization of profitable by-products.

Screening plant growth effects of sheep slaughterhouse waste-derived soil amendments in greenhouse trials.

Ameliorative effects of sheep slaughterhouse waste-derived soil amendments (struvite, blood meal, bone meal) were explored and quantified by a series of comparative greenhouse trials. A scoring matrix system was developed for 25 different test plants using 300 agricultural measurements obtained for three basic growth parameters (fresh-dry plant weights and plant heights) and four different fertilizer sources including solid vermicompost. More than 70% of NH4+-N recovery from sheep slaughterhouse wastewater was achieved using a chemical combination of MgCl2.6H2O + NaH2PO4.2H2O, a molar ratio of Mg2+:NH4+-N:PO43-P = 1.2:1:1, a reaction pH of 9.0, an initial NH4+-N concentration of 240 mg/L, and a reaction time of 15 min. According to SEM micrographs, surface morphology of struvite exhibited a highly porous structure composed of irregularly shaped crystals of various sizes (11.34-79.38 µm). FTIR spectroscopy verified the active functional groups on the proximity of all fertilizer sources within the spectral range of 500-3900 cm-1. TGA-DTG-DSC thermograms of struvite revealed that the mass loss occurred in two temperature regions and reached a maximum mass loss rate of 1.63%/min at 317 °C. The average percentages of increase (57.55-100.62%) and performance points (69-79) corroborated that the fertility value of struvite ranked first on average in cultivation of the analyzed plant species. Findings of this agro-valorization study confirmed that sheep slaughterhouse waste-derived fertilizers could be a beneficial way to promote bio-waste management and environmentally friendly agriculture.

Markovian RNN: An Adaptive Time Series Prediction Network With HMM-Based Switching for Nonstationary Environments.

We investigate nonlinear regression for nonstationary sequential data. In most real-life applications such as business domains including finance, retail, energy, and economy, time series data exhibit nonstationarity due to the temporally varying dynamics of the underlying system. We introduce a novel recurrent neural network (RNN) architecture, which adaptively switches between internal regimes in a Markovian way to model the nonstationary nature of the given data. Our model, Markovian RNN employs a hidden Markov model (HMM) for regime transitions, where each regime controls hidden state transitions of the recurrent cell independently. We jointly optimize the whole network in an end-to-end fashion. We demonstrate the significant performance gains compared to conventional methods such as Markov Switching ARIMA, RNN variants and recent statistical and deep learning-based methods through an extensive set of experiments with synthetic and real-life datasets. We also interpret the inferred parameters and regime belief values to analyze the underlying dynamics of the given sequences.

Electrochemically activated persulfate and peroxymonosulfate for furfural removal: optimization using Box-Behnken design.

Furfural removal by electrochemically activated peroxydisulfate (E-PS) and peroxymonosulfate (E-PMS) was investigated. The effect of different anodes was investigated for the electrochemical activation of oxidants. Box Behnken Design was applied to determine optimum operating conditions, which were determined as follows; PS concentration: 2.3 mM, applied current: 1.15 A, pH: 3.5, and reaction time: 118.3 min for E-PS process; PMS concentration: 1.8 mM, applied current: 1.05 A, pH: 3.3, and reaction time: 107.8 min for E-PMS process. The results of the study showed that the E-PMS process is more advantageous in terms of the chemical and electricity costs to be used.

Optimization of energy costs in the pretreatment of olive mill wastewaters by electrocoagulation.

In this study, the electrocoagulation process was evaluated as a pretreatment process for olive mill wastewaters. Aluminium (Al) and iron (Fe) electrodes, several contact times and 0.5, 1 and 2 A currents were used to compare chemical oxygen demand (COD) removal efficiencies for each case. The optimum contact time and current were 45 minutes and 1 A, respectively, which resulted in a COD removal of 58.7% with an Al electrode. Experimental data from distinct operational conditions were used to fit a model for COD removal efficiencies. Energy consumption was also predicted. Under optimum operational conditions, the treatment cost was approximately Euro 0.13 kg(-1) CODremoved and Euro 4.41 m(-3). The results showed that the electrocoagulation process was a cost-effective method for the pretreatment of olive mill wastewaters.

Achieving Online Regression Performance of LSTMs With Simple RNNs.

Recurrent neural networks (RNNs) are widely used for online regression due to their ability to generalize nonlinear temporal dependencies. As an RNN model, long short-term memory networks (LSTMs) are commonly preferred in practice, as these networks are capable of learning long-term dependencies while avoiding the vanishing gradient problem. However, due to their large number of parameters, training LSTMs requires considerably longer training time compared to simple RNNs (SRNNs). In this article, we achieve the online regression performance of LSTMs with SRNNs efficiently. To this end, we introduce a first-order training algorithm with a linear time complexity in the number of parameters. We show that when SRNNs are trained with our algorithm, they provide very similar regression performance with the LSTMs in two to three times shorter training time. We provide strong theoretical analysis to support our experimental results by providing regret bounds on the convergence rate of our algorithm. Through an extensive set of experiments, we verify our theoretical work and demonstrate significant performance improvements of our algorithm with respect to LSTMs and the other state-of-the-art learning models.

Degradation of oxytetracycline in aqueous solution by heat-activated peroxydisulfate and peroxymonosulfate oxidation.

Oxytetracycline (OTC) is a broad-spectrum antibiotic that resists biodegradation and poses a risk to the ecosystem. This study investigated the degradation of OTC by heat-activated peroxydisulfate (PDS) and peroxymonosulfate (PMS) processes. Response surface methodology (RSM) was used to evaluate the effect of process parameters, namely initial pH, oxidant concentration, temperature, and reaction time on the OTC removal efficiency. According to the results of the RSM models, all four independent variables were significant for both PDS and PMS processes. The optimum process parameters for the heat-activated PDS process were pH 8.9, PDS concentration 3.9 mM, temperature 72.9°C, and reaction time 26.5 min. For the heat-activated PMS process, optimum conditions were pH 9.0, PMS concentration 4.0 mM, temperature 75.0°C, and reaction time 20.0 min. The predicted OTC removal efficiencies for the PDS and PMS processes were 89.7% and 84.0%, respectively. As a result of the validation experiments conducted at optimum conditions, the obtained OTC removal efficiencies for the PDS and PMS processes were 87.6 ± 4.2 and 80.2± 4.6, respectively. PDS process has higher kinetic constants at all pH values than the PMS process. Both processes were effective in OTC removal from aqueous solution and RSM was efficient in process optimization.

Decolorization of textile wastewater by electrooxidation process using different anode materials: Statistical optimization.

The presence of reactive dyes in textile wastewater is a serious environmental concern due to their associated mutagenic and carcinogenic effects. The present study aims to analyze the effect of different anodic materials on the decolorization of a real textile wastewater effluent. For this purpose, four different anodic materials-TiO2 -coated platine, TiO2 -coated ruthenium dioxide (RuO2 ) (viz., RuO2 ), titanium dioxide (TiO2 ), and graphite-were connected, respectively, to titanium dioxide (TiO2 ) used as a cathode electrode. Color and cost optimization studies were performed using the response surface methodology and the Box-Behnken experimental design (BBD). According to ANOVA results, the R2 values for Pt/TiO2 , RuO2 /TiO2 , TiO2 /TiO2 , and graphite/TiO2 electrode pairs were found to be 97.4%, 93.8%, 92.44%, and 92.2%, respectively, indicating a good compatibility as it is close to one. The results show that color removal efficiencies at the optimal conditions were 86.3%, 90.8%, 91.5%, and 93.6% for Pt/TiO2 , graphite/TiO2 , TiO2 /TiO2 , and RuO2 /TiO2 , respectively. Furthermore, energy consumption cost at the optimum conditions was also evaluated, and the results were as follows: Pt/TiO2 (0.95 €/m3 ), graphite/TiO2 (0.74 €/m3 ), TiO2 /TiO2 (0.31 €/m3 ), and RuO2 /TiO2 (0.26 €/m3 ). Consequently, this research paper shows that all of the tested anodic materials give satisfactory color removal efficiencies higher than 86%. When energy consumption and color removal are considered together, the use of TiO2 /TiO2 and RuO2 /TiO2 pairs would be preferred. PRACTITIONER POINTS: Anodic contribution was investigated for decolorization of textile wastewater by electrooxidation process. Graphite, TiO2 -coated Pt, TiO2 -coated RuO2 , and TiO2 were used as anode materials. Highest color removal with lowest energy consumption was achieved with TiO2 -coated RuO2 anode material (93.6%).

Highlighting the cathodic contribution of an electrooxidation post-treatment study on decolorization of textile wastewater effluent pre-treated with a lab-scale moving bed-membrane bioreactor.

This study is carried out to investigate the effect of the cathodic contribution in the performance of electro-oxidation process for decolorization of the textile wastewater effluent pre-treated with a lab-scale moving bed-membrane bioreactor. For this purpose, titanium dioxide (TiO2) was used as anode electrode and four different cathodic electrode materials: Graphite, TiO2, TiO2-coated Platine, and TiO2-coated ruthenium dioxide (RuO2) (namely RuO2) were tested and compared for their color removal efficiencies. Besides, the optimization parameters that affect color removal in correspondence to the electrode materials, such as applied current, electrolysis time, and pH were studied. In this context, the optimum parameters for each electrode material were selected, and the color removal percentages were found as 92.95%, 91.58%, 91.40%, and 89.17% for TiO2/Graphite, TiO2/Platine, TiO2/TiO2, and TiO2/RuO2, respectively. Finally, the operational cost for each of the tested cathodic electrode materials was calculated in each of the studied optimization parameters making it easier and practical for the selection and evaluation of the electrode materials by the readers. The correlation coefficients (R2) were 81.2%, 87.1%, 86.7%, and 88.6% respectively as a result of the optimization study using the nonlinear regression modeling.

Vacuum-assisted thermal drying of wastewater treatment sludge.

Sludge drying is one of the main problems of wastewater treatment plants. It is very important to facilitate the drying process in terms of drying efficiency, time duration, and cost of the processes, so that transportation and dumping of sludge will accordingly be realized effectively. In this study, vacuum assisted thermal drying was studied. Under vacuum conditions, the water content of the sludge was examined easier than at atmospheric pressure in the drying process. For this reason, in this study, time, temperature, and sludge mass surface area on drying efficiency were evaluated under two different pressure levels, such as 30 mbar and 1,000 mbar. To optimize these parameters, the Response Surface Methodology approach was utilized. Results showed that the effect of vacuum condition on sludge drying was remarkable at obtaining at least 65% of solid material ratio in sludge, which is the lowest limit value on landfilling legislation in Turkey. Data obtained from the study shows that, especially at high temperatures, contribution on sludge drying of vacuum condition is more effective in terms of drying time. A smaller sludge surface area is also found more significant on vacuum drying.Implications: Drying the sludge under low pressure shortens the drying time. It is possible to reach higher solid material ratio under low pressure. Time, temperature, and surface areas of sludge are effective parameters in vacuum drying.

Decolorization and COD reduction of UASB pretreated poultry manure wastewater by electrocoagulation process: a post-treatment study.

The performance of electrocoagulation (EC) technique for decolorization and chemical oxygen demand (COD) reduction of anaerobically pretreated poultry manure wastewater was investigated in a laboratory batch study. Two identical 15.7-L up-flow anaerobic sludge blanket (UASB) reactors were first run under various organic and hydraulic loading conditions for 216 days. Effects of operating parameters such as type of sacrificial electrode material, time of electrolysis, current density, initial pH, and electrolyte concentration were further studied to optimize conditions for the post-treatment of UASB pretreated poultry manure wastewater. Preliminary tests conducted with two types of sacrificial electrodes (Al and Fe) resulted that Al electrodes were found to be more effective for both COD and color removals than Fe electrodes. The subsequent EC tests performed with Al electrodes showed that optimal operating conditions were determined to be an initial pH of 5.0, a current density of 15mA/cm(2), and an electrolysis time of 20min. The results indicated that under the optimal conditions, about 90% of COD and 92% of residual color could be effectively removed from the UASB effluent with the further contribution of the EC technology used as a post-treatment unit. In this study, the possible acute toxicity of the EC effluent was also evaluated by a static bioassay test procedure using guppy fish (Lebistes reticulatus). Findings of this study clearly indicated that incorporation of a toxicological test into conventional physicochemical analyses provided a better evaluation of final discharge characteristics.

Treatment of leachate by electrocoagulation using aluminum and iron electrodes.

In this paper, treatment of leachate by electrocoagulation (EC) has been investigated in a batch process. The sample of leachate was supplied from Odayeri Landfill Site in Istanbul. Firstly, EC was compared with classical chemical coagulation (CC) process via COD removal. The first comparison results with 348 A/m2 current density showed that EC process has higher treatment performance than CC process. Secondly, effects of process variables such as electrode material, current density (from 348 to 631 A/m2), pH, treatment cost, and operating time for EC process are investigated on COD and NH4-N removal efficiencies. The appropriate electrode type search for EC provided that aluminum supplies more COD removal (56%) than iron electrode (35%) at the end of the 30 min operating time. Finally, EC experiments were also continued to determine the efficiency of ammonia removal, and the effects of current density, mixing, and aeration. All the findings of the study revealed that treatment of leachate by EC can be used as a step of a joint treatment.

Optimization of treatment leachates from young, middle-aged and elderly landfills with bipolar membrane electrodialysis.

In this study, a bipolar membrane electrodialysis (BMED) process, which is thought to be an effective treatment method for leachate, was evaluated for leachates of three different ages ('young', 'middle-aged' and 'elderly'). The leachates were pretreated to eliminate membrane fouling problems prior to the BMED process. Experimental studies were carried out to determine optimal operating conditions for the three differently aged leachates in the BMED process. According to the experiment results, there was a high removal efficiency of conductivity determined at 4 membrane - 25 V for young and elderly leachate and at 1 membrane - 25 V for middle-aged leachate. It was found that the operating times required to reach the optimal endpoints (at conductivity of about 2 mS/cm) of BMED process were 90, 180 and 300 min for the middle-aged, young and elderly leachates, respectively. Under the optimum operating conditions for the BMED process, removal efficiencies of conductivity and chemical oxygen demand were determined to be 89.5% and 60% for young, 82.5% and 30% for middle-aged and 91.8% and 26% for elderly leachate, respectively.

Fuzzy-logic modeling of Fenton's strong chemical oxidation process treating three types of landfill leachates.

Three multiple input and multiple output-type fuzzy-logic-based models were developed as an artificial intelligence-based approach to model a novel integrated process (UF-IER-EDBM-FO) consisted of ultrafiltration (UF), ion exchange resins (IER), electrodialysis with bipolar membrane (EDBM), and Fenton's oxidation (FO) units treating young, middle-aged, and stabilized landfill leachates. The FO unit was considered as the key process for implementation of the proposed modeling scheme. Four input components such as H(2)O(2)/chemical oxygen demand ratio, H(2)O(2)/Fe(2+) ratio, reaction pH, and reaction time were fuzzified in a Mamdani-type fuzzy inference system to predict the removal efficiencies of chemical oxygen demand, total organic carbon, color, and ammonia nitrogen. A total of 200 rules in the IF-THEN format were established within the framework of a graphical user interface for each fuzzy-logic model. The product (prod) and the center of gravity (centroid) methods were performed as the inference operator and defuzzification methods, respectively, for the proposed prognostic models. Fuzzy-logic predicted results were compared to the outputs of multiple regression models by means of various descriptive statistical indicators, and the proposed methodology was tested against the experimental data. The testing results clearly revealed that the proposed prognostic models showed a superior predictive performance with very high determination coefficients (R (2)) between 0.930 and 0.991. This study indicated a simple means of modeling and potential of a knowledge-based approach for capturing complicated inter-relationships in a highly non-linear problem. Clearly, it was shown that the proposed prognostic models provided a well-suited and cost-effective method to predict removal efficiencies of wastewater parameters prior to discharge to receiving streams.