Enhanced understanding of osmotic membrane bioreactors through machine learning modeling of water flux and salinity

Enhanced nitrogen prediction and mechanistic process analysis in high-salinity wastewater treatment using interpretable machine learning approach

This study introduces an interpretable machine learning framework to predict nitrogen removal in membrane bioreactor (MBR) treating high-salinity wastewater. By integrating Shapley additive explanations (SHAP) with Categorical Boosting (CatBoost), we address the critical gap in linking predictive accuracy to operational decision-making for saline systems. CatBoost achieved the best performance, with an coefficient of determination (R2) of 0.88 and root mean square error (RMSE) of 4.27 for the effluent ammonia nitrogen (NH4+-Nout), and an R2 of 0.91 and RMSE of 4.35 for the effluent total nitrogen (TNout). SHAP analysis uniquely revealed salinity's dual role in inhibiting nitrifying enzymes and disrupting carbon metabolism, with dissolved oxygen, pH and chemical oxygen demand removal efficiency as key regulators. Temperature and carbon-to-nitrogen ratio further modulated total nitrogen dynamics through electron donor availability and microbial activity. The proposed SHAP-CatBoost model in high salinity MBR combines predictive modelling with mechanical process control.

Machine learning algorithms for predicting membrane bioreactors performance: A review

Membrane bioreactors (MBR) are recognized as a sustainable technology for treating polluted effluents. Machine learning (ML) algorithms have emerged as a modeling option to predict pollutant removal and operational variables such as membrane fouling, permeability, and energy consumption, which are significant challenges for MBR application. This review examines the use of ML algorithms in MBR-based wastewater treatment, focusing on the prediction of nitrogen and organic matter removal, and operational parameters related to membrane fouling. It presents the structures and fit quality of each model, noting that artificial neural networks (ANNs) are the most commonly used algorithm, appearing in 88 % of the 57 analyzed articles. Additionally, the review identified studies using random forests, support vector machines, k-nearest neighbors and boosting techniques, among other ML algorithms, although these were less frequently encountered. The review suggests potential in exploring less-utilized models for MBR data and identifies a gap in predicting membrane lifespan and replacement with ML models. This study aims to guide the development of new models for optimizing MBR performance by highlighting effective variables and algorithms, enhancing process control, real-time data analysis, parameter adjustment, and operational efficiency.

Machine learning-based prediction of non-aeration linear alkylbenzene sulfonate mineralization in an oxygenic microalgal-bacteria biofilm

Microalgal-bacteria biofilm shows great potential in low-cost greywater treatment. Accurately predicting treated greywater quality is of great significance for water reuse. In this work, machine learning models were developed for simulating and predicting linear alkylbenzene sulfonate (LAS) removal using 152-days collected data from a battled oxygenic microalgal-bacteria biofilm reactor (MBBfR). By using nine variables including influent LAS, hydraulic retention time (HRT), biofilm density and thickness, specific oxygen production and consumption rates, microalgae and bacteria concentrations, and dissolved oxygen (DO), the support vector machine (SVM) model enabled the accurate LAS removal prediction (training set: R2 = 0.995, (root mean square error, RMSE) = 0.076, (mean absolute error, MAE) = 0.069; testing set: R2 = 0.961, RMSE = 0.251, MAE = 0.153). SVM can be also successfully applied for MBBfR operation optimization (HRT = 4.28 h, DO = 0.25 mg/L) that achieving accurate prediction of LAS mineralization.

Integrating artificial intelligence modeling and membrane technologies for advanced wastewater treatment: Research progress and future perspectives

Membrane technologies have become proficient alternatives for advanced wastewater treatment, ensuring high contaminant removal and sustainable resource recovery. Despite significant progress, ongoing research efforts aim to further optimize treatment performance. Among the challenges faced, membrane fouling persists as a relevant obstacle in membrane technologies, necessitating the development of more effective mitigation strategies. Mathematical models, widely employed for predicting treatment performance, generally exhibit low accuracy and suffer from uncertainties due to the complex and variable nature of wastewater. To overcome these limitations, numerous studies have proposed artificial intelligence (AI) modeling to accurately predict membrane technologies' performance and fouling mechanisms. This approach aims to provide advanced simulations and predictions, thereby enhancing process control, optimization, and intensification. This literature review explores recent advancements in modeling membrane-based wastewater treatment processes through AI models. The analysis highlights the enormous potential of this research field in enhancing the efficiency of membrane technologies. The role of AI modeling in defining optimal operating conditions, developing effective strategies for membrane fouling mitigation, enhancing the performance of novel membrane-based technologies, and improving membrane fabrication techniques is discussed. These enhanced process optimization and control strategies driven by AI modeling ensure improved effluent quality, optimized resource consumption, and minimized operating costs. The potential contribution of this cutting-edge approach to a paradigm shift toward sustainable wastewater treatment is examined. Finally, this review outlines future perspectives, emphasizing the research challenges that require attention to overcome the current limitations hindering the integration of AI modeling in wastewater treatment plants.

Predicting micropollutant removal through nanopore-sized membranes using several machine-learning approaches based on feature engineering

Predicting the efficacy of micropollutant separation through functionalized membranes is an arduous endeavor. The challenge stems from the complex interactions between the physicochemical properties of the micropollutants and the basic principles underlying membrane filtration. This study aimed to compare the effectiveness of a modest dataset on various machine learning tools (ML) tools in predicting micropollutant removal efficiency for functionalized reverse osmosis (RO) and nanofiltration (NF) membranes. The inherent attributes of both the micropollutants and the membranes are utilized as input factors. The chosen ML tools are supervised algorithm (adaptive network-based fuzzy inference system (NF), linear regression framework (linear regression (LR)), stepwise linear regression (SLR) and multivariate linear regression (MVR)), and unsupervised algorithm (support vector machine (SVM) and ensemble boosted tree (BT)). The feature engineering and parametric dependency analysis revealed that characteristics of micropollutants, such as maximum projection diameter (MaxP), minimal projection diameter (MinP), molecular weight (MW), and compound size (CS), exhibited a notably positive impact on the correlation with removal efficiency. Model combination with key variables demonstrated high prediction accuracy in both supervised and unsupervised ML for micropollutant removal efficiency. An NF-grid partitioning (NF-GP) model achieved the highest accuracy with an R 2 value of 0.965, accompanied by low error metrics, specifically an RMSE and MAE of 3.65. It is owed to the handling of the complex spatial and temporal aspects of micropollutant data through division into consistent subsets facilitating improved identification of rejection efficiency and relationships. The inclusion of inputs with both negative and positive correlations introduces variability, amplifies the system responsiveness, and impedes the precision of predictive models. This study identified key micropollutant properties, including MaxP, MinP, MW, and CS, as crucial factors for efficient micropollutant rejection during real-time filtration applications. It also allowed the design of pore size of self-prepared membranes for the enhanced separation of micropollutants from wastewater.

Adsorption of uranyl ion on hexagonal boron nitride for remediation of real U-contaminated soil and its interpretation using random forest

Acid leaching has been widely applied to treat contaminated soil, however, it contains several inorganic pollutants. The decommissioning of nuclear power plants introduces radioactive and soluble U(VI), a substance posing chemical toxicity to humans. Our investigation sought to ascertain the efficacy of hexagonal boron nitride (h-BN), an highly efficient adsorbent, in treating U(VI) in wastewater. The adsorption equilibrium of U(VI) by h-BN reached saturation within a mere 2 h. The adsorption of U(VI) by h-BN appears to be facilitated through electrostatic attraction, as evidenced by the observed impact of pH variations, acidic agents (i.e., HCl or H2SO4), and the presence of background ions on the adsorption performance. A reusability test demonstrated the successful completion of five cycles of adsorption/desorption, relying on the surface characteristics of h-BN as influenced by solution pH. Based on the experimental variables of initial U(VI) concentration, exposure time, temperature, pH, and the presence of background ions/organic matter, a feature importance analysis using random forest (RF) was carried out to evaluate the correlation between performances and conditions. To the best of our knowledge, this study is the first attempt to conduct the adsorption of U(VI) generated from real contaminated soil by h-BN, followed by interpretation of the correlation between performance and conditions using RF. Lastly, a. plausible adsorption mechanism between U(VI) and h-BN was explained based on the experimental results, characterizations, and a. comparison with previous adsorption studies on the removal of heavy metals by h-BN.

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.

Energy balance survey for the design and auto-thermal thermophilic aerobic digestion of algal-based membrane bioreactor for Landfill Leachate Treatment(under organic loading rates): Experimental and simulation-based ANN and NSGA-II

Although algal-based membrane bioreactors (AMBRs) have been demonstrated to be effective in treating wastewater (landfill leachate), there needs to be more research into the effectiveness of these systems. This study aims to determine whether AMBR is effective in treating landfill leachate with hydraulic retention times (HRTs) of 8, 12, 14, 16, 21, and 24 h to maximize AMBR's energy efficiency, microalgal biomass production, and removal efficiency using artificial neural network (ANN) models. Experimental results and simulations indicate that biomass production in bioreactors depends heavily on HRT. A decrease in HRT increases algal (Chlorella vulgaris) biomass productivity. Results also showed that 80% of chemical oxygen demand (COD) was removed from algal biomass by bioreactors. To determine the most efficient way to process the features as mentioned above, nondominated sorting genetic algorithm II (NSGA-II) techniques were applied. A mesophilic, suspended-thermophilic, and attached-thermophilic organic loading rate (OLR) of 1.28, 1.06, and 2 kg/m3/day was obtained for each method. Compared to suspended-thermophilic growth (3.43 kg/m3.day) and mesophilic growth (1.28 kg/m3.day), attached-thermophilic growth has a critical loading rate of 10.5 kg/m3.day. An energy audit and an assessment of the system's auto-thermality were performed at the end of the calculation using the Monod equation for biomass production rate (Y) and bacteria death constant (Kd). According to the results, a high removal level of COD (at least 4000 mg COD/liter) leads to auto-thermality.

Optimization of water reuse and modelling by saline composition with nanoparticles based on machine learning architectures

Water is a necessary resource that enables the existence of all life forms, including humans. Freshwater usage has become increasingly necessary in recent years. Facilities for treating seawater are less dependable and effective. Deep learning methods have the ability to improve salt particle analysis in saltwater's accuracy and efficiency, which will enhance the performance of water treatment plants. This research proposes a novel technique in optimization of water reuse with nanoparticle analysis based on machine learning architecture. Here, the optimization of water reuse is carried out based on nanoparticle solar cell for saline water treatment and the saline composition has been analyzed using a gradient discriminant random field. Experimental analysis is carried out in terms of specificity, computational cost, kappa coefficient, training accuracy, and mean average precision for various tunnelling electron microscope (TEM) image datasets. The bright-field TEM (BF-TEM) dataset attained a specificity of 75%, kappa coefficient of 44%, training accuracy of 81%, and mean average precision of 61%, whereas the annular dark-field scanning TEM (ADF-STEM) dataset produced specificity of 79%, kappa coefficient of 49%, training accuracy of 85%, and mean average precision of 66% as compared with the existing artificial neural network (ANN) approach.

Smart utilisation of reverse solute diffusion in forward osmosis for water treatment: A mini review

Forward osmosis (FO) has been widely studied as a promising technology in wastewater treatment, but undesirable reverse solute diffusion (RSD) is inevitable in the FO process. The RSD is generally regarded as a negative factor for the FO process, resulting in the loss of draw solutes and reduced FO efficiency. Conventional strategies to address RSD focus on reducing the amount of reverse draw solutes by fabricating high selective FO membranes and/or selecting the draw solute with low diffusion. However, since RSD is inevitable, doubts have been raised about the strategies to cope with the already occurring reverse draw solutes in the feed solution, and the feasibility to positively utilise the RSD phenomenon to improve the FO process. Herein, we review the state-of-the-art applications of RSD and their benefits such as improving selectivity and maintaining the stability of the feed solution for both independent FO processes and FO integrated processes. We also provide an outlook and discuss important considerations, including membrane fouling, membrane development and draw/feed solution properties, in RSD utilisation for water and wastewater treatment.

Predicting Extraction Selectivity of Acetic Acid in Pervaporation by Machine Learning Models with Data Leakage Management

The extraction of acetic acid and other carboxylic acids from water is an emerging separation need as they are increasingly produced from waste organics and CO₂ during carbon valorization. However, the traditional experimental approach can be slow and expensive, and machine learning (ML) may provide new insights and guidance in membrane development for organic acid extraction. In this study, we collected extensive literature data and developed the first ML models for predicting separation factors between acetic acid and water in pervaporation with polymers' properties, membrane morphology, fabrication parameters, and operating conditions. Importantly, we assessed seed randomness and data leakage problems during model development, which have been overlooked in ML studies but will result in over-optimistic results and misinterpreted variable importance. With proper data leakage management, we established a robust model and achieved a root-mean-square error of 0.515 using the CatBoost regression model. In addition, the prediction model was interpreted to elucidate the variables' importance, where the mass ratio was the topmost significant variable in predicting separation factors. In addition, polymers' concentration and membranes' effective area contributed to information leakage. These results demonstrate ML models' advances in membrane design and fabrication and the importance of vigorous model validation.

Prediction of organic contaminant rejection by nanofiltration and reverse osmosis membranes using interpretable machine learning models

Efficiency improvement in contaminant removal by nanofiltration (NF) and reverse osmosis (RO) membranes is a multidimensional process involving membrane material selection and experimental condition optimization. It is unrealistic to explore the contributions of diverse influencing factors to the removal rate by trial-and-error experimentation. However, the advanced machine learning (ML) method is a powerful tool to simulate this complex decision-making process. Here, 4 traditional learning algorithms (MLR, SVM, ANN, kNN) and 4 ensemble learning algorithms (RF, GBDT, XGBoost, LightGBM) were applied to predict the removal efficiency of contaminants. Results reported here demonstrate that ensemble models showed significantly better predictive performance than traditional models. More importantly, this study achieved a compelling tradeoff between accuracy and interpretability for ensemble models with an effective model interpretation approach, which revealed the mutual interaction mechanism between the membrane material, contaminants and experimental conditions in membrane separation. Additionally, feature selection was for the first time achieved based on the aforementioned model interpretation method to determine the most important variable influencing the contaminant removal rate. Ultimately, the four ensemble models retrained by the selected variables achieved distinguished prediction performance (R²_adj = 92.4 %-99.5 %). MWCO (membrane molecular weight cut-off), McGowan volume of solute (V) and molecular weight (MW) of the compound were demonstrated to be the most important influencing factors in contaminant removal by the NF and RO processes. Overall, the proposed methods in this study can facilitate versatile complex decision-making processes in the environmental field, particularly in contaminant removal by advanced physicochemical separation processes.