Real-time chlorophyll-a forecasting using machine learning framework with dimension reduction and hyperspectral data.

Since water is an essential resource in various fields, it requires constant monitoring. Chlorophyll-a concentration is a crucial indicator of water quality and can be used to monitor water quality. In this study, we developed methods to forecast chlorophyll-a concentrations in real-time using hyperspectral data on IoT platform and various machine learning algorithms. Compared to regular cameras that record information only in the three broad color bands of red, green, and blue, the hyperspectral images of drinking water sources record the data in dozens or even hundreds of distinct small wavelength bands, providing each pixel in an image with a full spectrum. Different machine learning algorithms have been developed using hyperspectral data and field observations of water quality and weather conditions. Previous studies have predicted chlorophyll concentrations using either partial least squares (PLS), which is a dimensionality reduction method, or machine learning. In contrast, our study employed the PLS technique as a preprocessing step to diminish the dimensionality of the hyperspectral data, followed by the application of the machine learning techniques with optimized hyperparameters to improve the precision of the predictions, thereby introducing a real-time mechanism for chlorophyll-a prediction. Consequently, a machine learning algorithm with R2 values of 0.9 or above and sufficiently small RMSE was developed for real-time chlorophyll-a forecasting. Real-time chlorophyll-a forecasting using LightGBM has the best performance, with a mean R2 of 0.963 and a mean RMSE of 2.679. This paper is expected to have applications in algal bloom early detection on monitoring systems.

The effects of ammonia acclimation on biogas recovery and the microbial population in continuous anaerobic digestion of swine manure.

This study investigated the ammonia toxicity and the acclimation of anaerobic microbiome in continuous anaerobic digestion of swine manure using unacclimated inoculum. When the total ammonia nitrogen concentration (TAN) reached 2.5 g N/L, the methane yield decreased from 254.1 ± 9.6 to 154.6 ± 9.9 mL/g COD. The free ammonia nitrogen concentration of the inhibited condition was 190 mg N/L. The methane yield was eventually recovered as 269.6 ± 3.6 mL/g COD with a further operation. Anaerobic toxicity assay (ATA) showed that mixed liquor from the recovered phase possessed enhanced tolerance to ammonia, not only within the exposed level in continuous operation (<2.5 g NH3/L) but also over the range (>2.5 g NH3/L). Microbial analysis revealed that continuous operation under ammonia stress resulted in the change of both bacterial and archaeal populations. The ammonia adaptation was concurrent with the archaeal population shift from Methanosaeta to Methanosarcina and Methanobacterium. The dominancy of Clostridia in bacterial population was found in the recovered phase. It is highly recommended to use an inoculum acclimated to a target ammonia level which can be pre-checked by ATA and to secure a start-up period for ammonia adaptation in the field application of anaerobic digestion for swine manure.

Silica-quaternary ammonium "Fixed-Quat" nanofilm coated fiberglass mesh for water disinfection and harmful algal blooms control.

Intensification of pollution loading worldwide has promoted an escalation of different types of disease-causing microorganisms, such as harmful algal blooms (HABs), instigating detrimental impacts on the quality of receiving surface waters. Formation of unwanted disinfection by-products (DBPs) resulting from conventional disinfection technologies reveals the need for the development of new sustainable alternatives. Quaternary Ammonium Compounds (QACs) are cationic surfactants widely known for their effective biocidal properties at the ppm level. In this study, a novel silica-based antimicrobial nanofilm was developed using a composite of silica-modified QAC (Fixed-Quat) and applied to a fiberglass mesh as an active surface via sol-gel technique. The synthesized Fixed-Quat nanocoating was found to be effective against E. coli with an inactivation rate of 1.3 × 10-3 log reduction/cm min. The Fixed-Quat coated fiberglass mesh also demonstrated successful control of Microcystis aeruginosa with more than 99% inactivation after 10 hr of exposure. The developed antimicrobial mesh was also evaluated with wild-type microalgal species collected in a water body experiencing HABs, obtaining a 97% removal efficiency. Overall, the silica-functionalized Fixed-Quat nanocoating showed promising antimicrobial properties for water disinfection and HABs control, while decreasing concerns related to DBPs formation and the possible release of toxic nanomaterials into the environment.

Recent advances in algal bloom detection and prediction technology using machine learning.

Harmful algal blooms (HAB) including red tides and cyanobacteria are a significant environmental issue that can have harmful effects on aquatic ecosystems and human health. Traditional methods of detecting and managing algal blooms have been limited by their reliance on manual observation and analysis, which can be time-consuming and costly. Recent advances in machine learning (ML) technology have shown promise in improving the accuracy and efficiency of algal bloom detection and prediction. This paper provides an overview of the latest developments in using ML for algal bloom detection and prediction using various water quality parameters and environmental factors. First, we introduced ML for algal bloom prediction using regression and classification models. Then we explored image-based ML for algae detection by utilizing satellite images, surveillance cameras, and microscopic images. This study also highlights several real-world examples of successful implementation of ML for algal bloom detection and prediction. These examples show how ML can enhance the accuracy and efficiency of detecting and predicting algal blooms, contributing to the protection of aquatic ecosystems and human health. The study also outlines recent efforts to enhance the field applicability of ML models and suggests future research directions. A recent interest in explainable artificial intelligence (XAI) was discussed in an effort to understand the most influencing environmental factors on algal blooms. XAI facilitates interpretations of ML model results, thereby enhancing the models' usability for decision-making in field management and improving their overall applicability in real-world settings. We also emphasize the significance of obtaining high-quality, field-representative data to enhance the efficiency of ML applications. The effectiveness of ML models in detecting and predicting algal blooms can be improved through management strategies for data quality, such as pre-treating missing data and integrating diverse datasets into a unified database. Overall, this paper presents a comprehensive review of the latest advancements in managing algal blooms using ML technology and proposes future research directions to enhance the utilization of ML techniques.

Explainable artificial intelligence for the interpretation of ensemble learning performance in algal bloom estimation.

Chlorophyll-a (Chl-a) concentrations, a key indicator of algal blooms, were estimated using the XGBoost machine learning model with 23 variables, including water quality and meteorological factors. The model performance was evaluated using three indices: root mean square error (RMSE), RMSE-observation standard deviation ratio (RSR), and Nash-Sutcliffe efficiency. Nine datasets were created by averaging 1 hour data to cover time frequencies ranging from 1 hour to 1 month. The dataset with relatively high observation frequencies (1-24 h) maintained stability, with an RSR ranging between 0.61 and 0.65. However, the model's performance declined significantly for datasets with weekly and monthly intervals. The Shapley value (SHAP) analysis, an explainable artificial intelligence method, was further applied to provide a quantitative understanding of how environmental factors in the watershed impact the model's performance and is also utilized to enhance the practical applicability of the model in the field. The number of input variables for model construction increased sequentially from 1 to 23, starting from the variable with the highest SHAP value to that with the lowest. The model's performance plateaued after considering five or more variables, demonstrating that stable performance could be achieved using only a small number of variables, including relatively easily measured data collected by real-time sensors, such as pH, dissolved oxygen, and turbidity. This result highlights the practicality of employing machine learning models and real-time sensor-based measurements for effective on-site water quality management. PRACTITIONER POINTS: XAI quantifies the effects of environmental factors on algal bloom prediction models The effects of input variable frequency and seasonality were analyzed using XAI XAI analysis on key variables ensures cost-effective model development.

Predicting the impact of hydraulic retention time and biodegradability on the performance of sludge acidogenesis using an artificial neural network.

This study aimed to predict volatile fatty acids (VFAs) production from SDBS-pretreated waste-activated sludge (WAS). A lab-scale continuous experiment was conducted at varying hydraulic retention times (HRTs) of 7 d to 1 d. The highest VFA yield considering the WAS biodegradability was 86.8 % based on COD at an HRT of 2 d, where the hydrolysis and acidogenesis showed the highest microbial activities. According to 16S rRNA gene analysis, the most abundant bacterial class and genus at an HRT of 2 d were Synergistia and Aminobacterium, respectively. Training regression (R) for TVFA and VFA yield was 0.9321 and 0.9679, respectively, verifying the efficiency of the ANN model in learning the relationship between the input variables and reactor performance. The prediction outcome was verified with R2 values of 0.9416 and 0.8906 for TVFA and VFA yield, respectively. These results would be useful in designing, operating, and controlling WAS treatment processes.

Anaerobic digestion of sewage sludge using anaerobic dynamic membrane bioreactor under various sludge composition and organic loading rates.

This study investigates the effects of sludge compositions and organic loading rates (OLRs) on stable biogas production during sludge digestion. Batch digestion experiments evaluate the effects of alkaline-thermal pretreatment and waste activated sludge (WAS) fractions on the biochemical methane potential (BMP) of sludge. A lab-scale anaerobic dynamic membrane bioreactor (AnDMBR) is fed with a mixture of primary sludge and pretreated WAS. Monitoring of volatile fatty acid to total alkalinity (FOS/TAC) helps maintain operational stability. The highest average methane production rate of 0.7 L/L·d is achieved when the OLR, hydraulic retention time, WAS volume fraction, and FOS/TAC ratio are 5.0 g COD/L·d, 12 days, 0.75, and 0.32, respectively. This study finds functional redundancy in two pathways: hydrogenotrophic and acetolactic. An increase in OLR promotes bacterial and archaeal abundance and specific methanogenic activity. These results can be applied to the design and operation of sludge digestion for stable, high-rate biogas recovery.

Machine learning in fermentative biohydrogen production: Advantages, challenges, and applications.

Hydrogen can be produced in an environmentally friendly manner through biological processes using a variety of organic waste and biomass as feedstock. However, the complexity of biological processes limits their predictability and reliability, which hinders the scale-up and dissemination. This article reviews contemporary research and perspectives on the application of machine learning in biohydrogen production technology. Several machine learning algorithems have recently been implemented for modeling the nonlinear and complex relationships among operational and performance parameters in biohydrogen production as well as predicting the process performance and microbial population dynamics. Reinforced machine learning methods exhibited precise state prediction and retrieved the underlying kinetics effectively. Machine-learning based prediction was also improved by using microbial sequencing data as input parameters. Further research on machine learning could be instrumental in designing a process control tool to maintain reliable hydrogen production performance and identify connection between the process performance and the microbial population.

Valorization of pretreated waste activated sludge to organic acids and biopolymer.

Polyhydroxybutyrate (PHB) is a natural polyester that may be made by utilizing volatile fatty acids (VFAs) as a substrate. VFA generated by continuous anaerobic fermentation of waste activated sludge (WAS) was fed into bioreactors for PHB synthesis in this work. Series of optimization tests were conducted to increase the biodegradability and hydrolysis of waste activated sludge. It was found out that 0.05 g/g TS of SDBS (sodium dodecylbenzene sulfonate), 70 °C (heat treatment) and 2hr (time) as pretreatment condition would give the highest solubilization. Impact of pH adjustment on the acidogenesis of pretreated WAS was evaluated in batch experiments at varying initial pH (4-10). The result indicated that when operational pH was between 7.5 and 8, the VFA yield was increased by 5.3-18.1%. Continuous acidogenic operation validated the SDBS pretreatment and pH adjustment warranted stable VFA conversion from WAS at a yield of 47% in COD basis. Firmicutes, Actinobacteria and Proteobacteria were affiliated as dominant bacterial phyla in the continuous acidogenesis. The effluent of the continuous acidogenesis was converted to biopolymer with the average yields of 0.23 g PHB-COD/g VFAadded-COD in the feast mode and 0.34 g PHB-COD/g VFAadded-COD in the famine mode. In feast and famine cycle, the average VFA utilization was 55% and 60% respectively. The sequential SDBS pretreatment, acidogenesis and PHB production would produce 162 g of PHB from 1 kg of WAS as COD basis.

Interpretation of ensemble learning to predict water quality using explainable artificial intelligence.

Algal bloom is a significant issue when managing water quality in freshwater; specifically, predicting the concentration of algae is essential to maintaining the safety of the drinking water supply system. The chlorophyll-a (Chl-a) concentration is a commonly used indicator to obtain an estimation of algal concentration. In this study, an XGBoost ensemble machine learning (ML) model was developed from eighteen input variables to predict Chl-a concentration. The composition and pretreatment of input variables to the model are important factors for improving model performance. Explainable artificial intelligence (XAI) is an emerging area of ML modeling that provides a reasonable interpretation of model performance. The effect of input variable selection on model performance was estimated, where the priority of input variable selection was determined using three indices: Shapley value (SHAP), feature importance (FI), and variance inflation factor (VIF). SHAP analysis is an XAI algorithm designed to compute the relative importance of input variables with consistency, providing an interpretable analysis for model prediction. The XGB models simulated with independent variables selected using three indices were evaluated with root mean square error (RMSE), RMSE-observation standard deviation ratio, and Nash-Sutcliffe efficiency. This study shows that the model exhibited the most stable performance when the priority of input variables was determined by SHAP. This implies that on-site monitoring can be designed to collect the selected input variables from the SHAP analysis to reduce the cost of overall water quality analysis. The independent variables were further analyzed using SHAP summary plot, force plot, target plot, and partial dependency plot to provide understandable interpretation on the performance of the XGB model. While XAI is still in the early stages of development, this study successfully demonstrated a good example of XAI application to improve the interpretation of machine learning model performance in predicting water quality.

Exploration of deep learning models for real-time monitoring of state and performance of anaerobic digestion with online sensors.

The immediate response to the state disturbances of anaerobic digestion is essential to prevent anaerobic digestion failure. However, frequent monitoring of the state and performance of anaerobic digestion is challenging. Thus, deep learning models were investigated to predict the state and performance variables from online sensor data. The online sensor data, including pH, electric conductivity, and oxidation-reduction potential, were used as the input features to build deep learning models. The state and performance data measured offline were used as the labels. The model performance was compared for several deep learning models of convolutional neural network (CNN), long short-term memory (LSTM), dense layer, and their combinations. The combined model of CNN and bidirectional LSTM was robust and well-generalized in predicting the state and performance variables (R2 = 0.978, root mean square error = 0.031). The combined model is an excellent soft sensor for monitoring the state and performance of anaerobic digestion from electrochemical sensors.

Synthesis of fatty acid-based ammonium ionic liquids and their application for extraction of Co(II) and Ni(II) metals ions from aqueous solution.

Leaching of cobalt and nickel into diverse water streams has become an environmental hazard and is continuously impacting human health through the food chain. Solvent extraction is the most widely accepted for separating these metals, but traditional extractants employed in conjunction with molecular diluents often lack selectivity and caused major environmental hurdles. Therefore, the development of cost-effective, environmentally friendly technologies for recovering these heavy metals has been strongly encouraged in recent years. Herein, two halogens free, low viscous, biocompatible fatty acid-based hydrophobic ionic liquids (ILs), i.e., methytrioctylammonium oleate, methytrioctylammonium linoleate were synthesized, analytically characterized and employed for recovery of cobalt, Co(II) and nickel, Ni(II) from their aqueous solutions. Extraction behaviour of Co(II) and Ni(II) was further evaluated by varying equilibrium time, ILs molar concentration, metal loading, and temperature. Thermodynamic parameters such as enthalpy change and Gibbs free energy change were also studied during extraction process. Slope analysis suggested that the extraction mechanism was an exothermic process that followed ion-transfer from the aqueous phase to the organic phase. Results showed that both fatty acid based-ILs were found to be capable of extracting >99% of Co(II) and Ni(II) from aqueous solutions at 298 K, in 15 min of shaking time using a 1:1 (org: aq.) ratio at low concentrations of 2.5-10 g L-1. Furthermore, for methyltrioctylammonium oleate IL, Co(II) extraction was selectively preferred over Ni(II) extraction when the metal concentration was increased to above to 10 g L-1. The stripping results showed that 2 M H2SO4, and 2 M HCl successfully stripped out >99% of Co(II) and Ni(II) from the organic phase, respectively compared to HNO3.

Biofilm formation as a method of improved treatment during anaerobic digestion of organic matter for biogas recovery.

The efficiency of anaerobic digestion could be increased by promoting microbial retention through biofilm development. The inclusion of certain types of biofilm carriers has differentiated existing AD biofilm reactors through their respective mode of biofilm growth. Bacteria and archaea engaged in methanogenesis during anaerobic processes potentially build biofilms by adhering or attaching to biofilm carriers. Meta-analyzed results depicted varying degrees of biogas enhancement within AD biofilm reactors. Furthermore, different carrier materials highly induced the dynamicity of the dominant microbial population in each system. It is suggested that the promotion of surface contact and improvement of interspecies electron transport have greatly impacted the treatment results. Modern spectroscopy techniques have been and will continue to give essential information regarding biofilm's composition and structural organization which can be useful in elucidating the added function of this special layer of microbial cells.

Sludge disintegration and anaerobic digestion enhancement by alkaline-thermal pretreatment: Economic evaluation and microbial population analysis.

Alkaline-thermal pretreatment was examined for waste activated sludge (WAS) disintegration and subsequent anaerobic digestion (AD). Pretreatment at 60 °C was estimated to provide better economic benefits than higher temperature conditions. The maximum methane yield of 215.6 mL/g COD was achieved when WAS was pretreated at 60 °C and pH 10 for 24 h, which was 46.6% higher than untreated WAS. The pretreatment condition also provided the maximum net savings. The degree of sludge disintegration, considering both loosely bound-extracellular polymeric substance and soluble COD, would be a better indicator to predict anaerobic digestibility than the solubilization rate that considers soluble COD alone. Microbial analysis implied that pretreatment facilitated the growth of hydrolytic bacteria, phyla Bacteroidetes and Firmicutes. In addition, sludge pretreatment enhanced the growth of both acetoclastic and hydrogenotrophic methanogens, genera Methanosaeta and Methanobacterium. The mild AT-PT would be useful to enhance the digestion performance and economic benefit of WAS digestion.

Recent advances in black liquor valorization.

Lignocellulosic biomass is projected as a prospective renewable alternative to petroleum for the production of fuel and chemicals. Pretreatment is necessary to disrupt the lignocellulosic structure for extraction of cellulose. Biomass after pretreatment is segregated into cellulose rich solid fraction and black liquor (lignin and hemicelluloses) as a liquid stream. The plant polysaccharide-based industry primarily utilizes the cellulosic fraction as raw material, and carbon rich black liquor discarded as waste or burnt for energy recovery. This review highlights the recent advancements in the biological and chemical valorization of black liquor into fuels and chemicals. The recent research attempted for bioconversion of black liquor into Bioplastic, Biohydrogen, Biogas, and chemicals has been discussed. In addition, the efforts to replace the conventional energy recovery method with the advanced chemical process along with their modifications have been reviewed that will decide the sustainability of the lignocellulosic biomass-based industry.

Enhanced anaerobic digestion of waste-activated sludge via bioaugmentation strategy-Phylogenetic investigation of communities by reconstruction of unobserved states (PICRUSt2) analysis through hydrolytic enzymes and possible linkage to system performance.

In this study, anaerobic digestion of waste-activated sludge was bioaugmented with hydrolytic bacteria, Bacteroidetes uniformis (Bacteroidetes, B) and Clostridium sp. (Firmicutes, F) at various dosages. Bioaugmentation resulted in enhanced methane conversion of waste-activated sludge. The highest methane yield of 298.1 mL CH4/g-COD, 85.2% COD conversion efficiency was obtained when Bacteroidetes uniformis and Clostridium sp. were augmented at 100 and 900 CFU/mL, respectively. The microbial community analysis demonstrated that bioaugmentation increased the proportion of Bacteroidetes, Firmicutes, and Proteobacteria. Furthermore, at the highest methane yield, the principal methanogenic pathway was altered from acetoclastic to a mixture of hydrogenotrophic and acetoclastic; the major species shifted from Methanosaeta concilii to Methanobacterium subterraneum. Predicted gene analysis revealed that increased expression of hydrolases resulted in enhanced methane conversion through bioaugmentation.

Effect of low-thermal pretreatment on the methanogenic performance and microbiome population of continuous high-solid anaerobic digester treating dewatered sludge.

Undigested and dewatered sludge at 10% total solids was pretreated at 60 °C for 3 h and fed to a lab-scale horizontal anaerobic bioreactor for 130 days with solids retention time (SRTs) from 25 to 16 d. The low-thermal pretreatment enabled higher net energy production, improved sludge treatment efficiency, and enhanced digestion stability. The highest average biomethane yield and production rate were 138.5 mL/g VS and 0.43 L/L.d, respectively, and the economic benefit was expected to be the maximum at SRT 16 d. Pretreatment did not increase the specific methanogenic activity per unit methanogen, but resulted in higher abundance of methanogenic archaea and hydrolytic bacteria. Methanogenic population shifted from hydrogenotrophic to acetoclastic, consistent with predicted gene expression at SRT equal or below 20 d. Anaerobic digestion along with low-thermal could be a feasible management strategy for undigested dewatered sludge from small WWTPs.

Assessment of Chlorella sp. as a potential feedstock for biological methane production.

Microalgal biomass sequestrates CO2 and is regarded as a promising renewable feedstock for anaerobic digestion because of its adequate carbohydrate content and lignin-free structure. This study optimizes the dilute-acid pretreatment of Chlorella sp. and subsequent biomethane production using response surface methodology and central composite design with temperature, pretreatment time and solid-to-liquid ratio as variables. A temperature of 64.1 °C, pretreatment time of 1.2 h, and a solid to liquid ratio of 0.29 were the optimal pretreatment conditions and resulted in a methane yield of 302.22 mL CH4/g COD and methane production rate of 110.04 mL CH4/g VSS-d. The severity factor of 1.5-1.6 was adequate to render the Chlorella sp. bioavailable for high methane recovery. The results obtained from the experiments conformed to those predicted by the model. This study effectively utilizes algal biomass for biomethane production and enables the possibility of scaled-up studies using a closed-loop approach.

Variation of efficiencies and limits of ultrasonication for practical algal bloom control in fields.

Algal blooms are an increasing issue in managing water resources for drinking water production and recreational activities in many countries. Among various techniques, ultrasonication is known as a cost-effective method for control of harmful algal blooms (HABs) in relatively large area of water bodies. Most of engineering parameters for operating ultrasonication have been empirically determined based on laboratory scale tests, however, field or pilot tests in real environments are still rare. For field application, duration of ultrasonication is often on a monthly basis which is impractical for stream where there is flow and thus retention time is short. More realistic experimental approaches are required for practical applications of ultrasound. In this study, relatively low frequencies (36-175 kHz) of ultrasonication with low power intensity, less than 650 W, were tested for algal control in various pilot (100-750 L) and field (4 m3) tests in a short duration (<20 min). Generally, rapid decline of sound pressure (Pa) of ultrasonication was observed with distance (80% decrease even with 0.5 m difference). In a pilot test (100 L), the highest algae reduction was achieved at 36 kHz with 0.003 W mL-1 of power density within 10 min duration, but there was a noticeable increase in microcystin due to damaged algal cells by the low frequency of ultrasound. In a short-term operation without flow, distance from the ultrasound system was an important parameter for effective algae reduction, while longer exposure time ensured sufficient algae reduction. In a circulation pond (4 m3) with flow, 108 kHz-450 W showed the greatest efficiency in algal control and approximately 50-90% algal cells reduction was observed at 36-175 kHz with less than 650 W power and 60 min duration.

Cavitational activity in heterogeneous systems containing fine particles.

Ultrasound has been increasingly used in various processes containing a variety of homogeneous and heterogeneous systems. For largescale applications, a high energy efficiency of the process is required. With this view, the calorimetric energy and cavitational activity measurements were carried out in heterogeneous systems consisting of both liquid and solid phases (fine particles) in a 28-kHz double-bath sonoreactor. Ultrasonic soil washing for the remediation of clay-sized soils (∼75 µm), contaminated with metals (Cu, Pb, and Zn), was used as a case study. As the liquid height/volume in the inner vessel increased under the same input electrical power, the inner vessel calorimetric energy also increased, whereas the total calorimetric energy between the inner vessel and the outer reactor remained approximately constant. No significant differences in calorimetric energies were observed for both with and without soil conditions. The chemical activity under similar experimental conditions was evaluated using sonochemiluminescence. Different sonochemiluminescence trends were observed depending up on the presence and size of beads. The highest total sonochemiluminescence intensity with a uniform spatial distribution was obtained from fine beads (#200, 75 µm) suspended in the vessel. Ultrasound application significantly enhanced the removal efficiency of heavy metals when combined with mechanical agitation. The enhanced removal efficiency of the combined processes was attributed to a significant removal of metals from the residual (F5) fraction. It has been concluded that ultrasound has enough extraction power to be comparable to methods that employ extremely powerful acids for washing fine particles.