Application of artificial intelligence tools in wastewater and waste gas treatment systems: Recent advances and prospects.

The non-linear complex relationships among the process variables in wastewater and waste gas treatment systems possess a significant challenge for real-time systems modelling. Data driven artificial intelligence (AI) tools are increasingly being adopted to predict the process performance, cost-effective process monitoring, and the control of different waste treatment systems, including those involving resource recovery. This review presents an in-depth analysis of the applications of emerging AI tools in physico-chemical and biological processes for the treatment of air pollutants, water and wastewater, and resource recovery processes. Additionally, the successful implementation of AI-controlled wastewater and waste gas treatment systems, along with real-time monitoring at the industrial scale are discussed.

Optimization of low-grade coal and refuse-derived fuel blends for improved co-combustion behavior in coal-fired power plants.

This study is aimed at utilizing three waste materials, i.e., solid refuse fuel (SRF), tire derived fuel (TDF), and sludge derived fuel (SDF), as eco-friendly alternatives to coal-only combustion in co-firing power plants. The contribution of waste materials is limited to ≤5% in the composition of the mixed fuel (coal + waste materials). Statistical experimental design and response surface methodology are employed to investigate the effect of mixed fuel composition (SRF, TDF, and SDF) on gross calorific value (GCV) and ash fusion temperature (AFT). A quadratic model is developed and statistically verified to apprehend mixed fuel constituents' individual and combined effects on GCV and AFT. Constrained optimization of fuel blend, i.e., GCV >1,250 kcal/kg and AFT >1,200 °C, using the polynomial models projected the fuel-blend containing 95% coal with 3.84% SRF, 0.35% TDF, and 0.81% SDF. The observed GCV of 5,307 kcal/kg and AFT of 1225 °C for the optimized blend were within 1% of the model predicted values, thereby establishing the robustness of the models. The findings from this study can foster sustainable economic development and zero CO2 emission objectives by optimizing the utilization of waste materials without compromising the GCV and AFT of the mixed fuels in coal-fired power plants.

Ethanol mediated As(III) adsorption onto Zn-loaded pinecone biochar: Experimental investigation, modeling, and optimization using hybrid artificial neural network-genetic algorithm approach.

Organic matters (OMs) and their oxidization products often influence the fate and transport of heavy metals in the subsurface aqueous systems through interaction with the mineral surfaces. This study investigates the ethanol (EtOH)-mediated As(III) adsorption onto Zn-loaded pinecone (PC) biochar through batch experiments conducted under Box-Behnken design. The effect of EtOH on As(III) adsorption mechanism was quantitatively elucidated by fitting the experimental data using artificial neural network and quadratic modeling approaches. The quadratic model could describe the limiting nature of EtOH and pH on As(III) adsorption, whereas neural network revealed the stronger influence of EtOH (64.5%) followed by pH (20.75%) and As(III) concentration (14.75%) on the adsorption phenomena. Besides, the interaction among process variables indicated that EtOH enhances As(III) adsorption over a pH range of 2 to 7, possibly due to facilitation of ligand-metal(Zn) binding complexation mechanism. Eventually, hybrid response surface model-genetic algorithm (RSM-GA) approach predicted a better optimal solution than RSM, i.e., the adsorptive removal of As(III) (10.47µg/g) is facilitated at 30.22mg C/L of EtOH with initial As(III) concentration of 196.77µg/L at pH5.8. The implication of this investigation might help in understanding the application of biochar for removal of various As(III) species in the presence of OM.

Optimization of co-digestion of various industrial sludges for biogas production and sludge treatment: methane production potential experiments and modeling.

Optimal biogas production and sludge treatment were studied by co-digestion experiments and modeling using five different wastewater sludges generated from paper, chemical, petrochemical, automobile, and food processing industries situated in Ulsan Industrial Complex, Ulsan, South Korea. The biomethane production potential test was conducted in simplex-centroid mixture design, fitted to regression equation, and some optimal co-digestion scenarios were given by combined desirability function based multi-objective optimization technique for both methane yield and the quantity of sludge digested. The co-digestion model incorporating main and interaction effects among sludges were utilized to predict the maximum possible methane yield. The optimization routine for methane production with different industrial sludges in batches were repeated with the left-over sludge of earlier cycle, till all sludges have been completely treated. Among the possible scenarios, a maximum methane yield of 1161.53 m(3) is anticipated in three batches followed by 1130.33 m(3) and 1045.65 m(3) in five and two batches, respectively. This study shows a scientific approach to find a practical solution to utilize diverse industrial sludges in both treatment and biogas production perspectives.

Characterization of co-digestion of industrial sludges for biogas production by artificial neural network and statistical regression models.

The characteristics and impact of industrial sludges of paper, chemical, petrochemical, automobile, and food industries situated in the Ulsan Industrial Complex, Ulsan, Republic of Korea in co-digestion for biogas production were assessed by artificial neural network (ANN) and statistical regression models. The regression model was based on a simplex-centroid mixture design and the ANN was based on a resilient back-propagation algorithm (topology 5-7-1). Using connection weights and bias of the trained ANN model, the impact of each sludge of co-digestion was assessed using Garsons' algorithm. Results suggested that the modelling and predictability of ANN were superior to the regression model with accuracy (A(f)) 1.01, bias (B(f)) 1.00, root mean square error 3.56, and standard error of prediction 2.51%. Sludge from the chemical industry showed the highest impact on specific methane yield (SMY(VS)) with a relative importance of 28.59% followed by sludges from paper (20.07%), food (19.59%), petrochemical (15.92%), and automobile (15.82%) industries. The interactions between diverse industrial sludges were successfully modelled and partitioned into various synergistic and antagonistic effects on SMY(VS). Synergistic interactions between the chemical industry sludge and either petrochemical or food industry sludges on SMY(VS) were detected. However, strong negative interaction between automobile sludge and other sludges was observed. This study indicates that though the ANN model performed better in prediction and impact assessments, the regression model reveals the synergistic and antagonistic interactions among sludges.

Modeling and optimization of poly(3hydroxybutyrate-co-3hydroxyvalerate) production from cane molasses by Azohydromonas lata MTCC 2311 in a stirred-tank reactor: effect of agitation and aeration regimes.

The effects of agitation and aeration rates on copolymer poly(3-hydroxybutyrate-co-3-hydroxyvalerate) [P(3HB-co-3HV)] production by Azohydromonas lata MTCC 2311 using cane molasses supplemented with propionic acid in a bioreactor were investigated. The experiments were conducted in a three-level factorial design by varying the impeller (150-500 rev min(-1)) and aeration (0.5-1.5 vvm) rates. Further, the data were fitted to mathematical models [quadratic polynomial equation and artificial neural network (ANN)] and process variables were optimized by genetic algorithm-coupled models. ANN and hybrid ANN-GA were found superior for modeling and optimization of process variables, respectively. The maximum copolymer concentration of 7.45 g l(-1) with 21.50 mol% of 3HV was predicted at process variables: agitation speed, 287 rev min(-1); and aeration rate, 0.85 vvm, which upon validation gave 7.20 g l(-1) of P(3HB-co-3HV) with 21 mol% of 3HV with the prediction error (%) of 3.38 and 2.32, respectively. Agitation speed established a relative high importance of 72.19% than of aeration rate (27.80%) for copolymer accumulation. The volumetric gas-liquid mass transfer coefficient (k (L) a) was strongly affected by agitation and aeration rates. The highest P(3HB-co-3HV) productivity of 0.163 g l(-1) h(-1) was achieved at 0.17 s(-1) of k (L) a value. During the early phase of copolymer production process, 3HB monomers were accumulated, which were shifted to 3HV units (9-21%) during the cultivation period of 24-42 h. The enhancement of 7.5 and 34% were reported for P(3HB-co-3HV) production and 3HV content, respectively, by hybrid ANN-GA paradigm, which revealed the significant utilization of cane molasses for improved copolymer production.

Artificial intelligence based modeling and optimization of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) production process by using Azohydromonas lata MTCC 2311 from cane molasses supplemented with volatile fatty acids: a genetic algorithm paradigm.

The present work describes the optimization of medium variables for the production of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) [P(3HB-co-3HV)] by Azohydromonas lata MTCC 2311 using cane molasses supplemented with propionic acid. Genetic algorithm (GA) has been used for the optimization of P(3HB-co-3HV) production through the simulation of artificial neural network (ANN) and response surface methodology (RSM). The predictions by ANN are better than those of RSM and in good agreement with experimental findings. The highest P(3HB-co-3HV) concentration and 3HV content have been reported as 7.35 g/l and 16.84 mol%, respectively by hybrid ANN-GA. Upon validation, 7.20 g/l and 16.30 mol% of P(3HB-co-3HV) concentration and 3HV content have been found in the shake flask, whereas 6.70 g/l and 16.35 mol%, have been observed in a 3 l bioreactor, respectively. The specific growth rate and P(3HB-co-3HV) accumulation rate of 0.29 per h and 0.16 g/lh determined with cane molasses are comparable to those observed on pure substrates.

Modeling studies on simultaneous adsorption of phenol and resorcinol onto granular activated carbon from simulated aqueous solution.

The modelling study on simultaneous adsorption of phenol and resorcinol onto granular activated carbon (GAC) in multicomponent solution was carried out at 303K by conducting batch experiments at initial concentration range of 100-1000 mg/l. Three equilibrium isotherm models for multicomponent adsorption studies were considered. In order to determine the parameters of multicomponent adsorption isotherms, individual adsorption studies of phenol and resorcinol on GAC were also carried out. The experimental data of single and multicomponent adsorption were fitted to these models. The parameters of multicomponent models were estimated using error minimization technique on MATLAB R2007a. It has been observed that for low initial concentration of adsorbate (100-200mg/l), modified Langmuir model represents the data very well with the adsorption constant (Q(0)), 216.1, 0.032 and average relative error (ARE) of 8.34, 8.31 for phenol and resorcinol respectively. Whereas, for high initial concentration of adsorbate (400-1000 mg/l), extended Freundlich model represents the data very well with adsorption constant (K(F)) of 25.41, 24.25 and ARE of 7.0, 6.46 for phenol and resorcinol respectively. The effect of pH of solution, adsorbent dose and initial concentrations of phenol and resorcinol on adsorption behaviour was also investigated.