Solar nanophotocatalytic pretreatment of seawater: process optimization and performance evaluation using response surface methodology and genetic algorithm

Abstract

In reverse osmosis seawater treatment process, membrane fouling can be mitigated by degrading organic pollutants present in the feed seawater. The present study evaluates the effectiveness of employing solar photocatalysis using TiO2/ZnO/H2O2 to pretreat reverse osmosis (RO) feed seawater under solar irradiation. Process optimisation and performance evaluation were undertaken using response surface methodology-desirability function and RSM integrated with genetic algorithm (RSM-GA). Statistical analysis was performed to determine the interactive relationships and main effects of input factors such as TiO2 dosage, H2O2 dosage, pH, reaction time and ZnO dosage. The performance evaluation was determined in terms of percentage removal of total organic carbon (TOC) and chemical oxygen demand (COD). The obtained optimum values using RSM-GA evaluation for TOC and COD removal were found to be 76.5% and 63.9%, respectively. The predicted RSM-GA results correspond well with the experimental results (TOC removal = 73.3%, COD removal = 61.2%). Utilization of renewable solar energy coupled with optimum utilisation of nanophotocatalysts enables this technique to be a unique treatment process for RO pretreatment of seawater and membrane fouling mitigation.

Experimental and Computational Analysis for Optimization of Seawater Biodegradability Using Photo Catalysis

“…Table 8 shows the responses of percentage elimination efficiencies of dependent variables. The RSM-BBD coded and actual components in terms of model equations were quadratic models [27], where R stands for dosage, S for reaction time, and T for pH level. The other terms, such as RS, RT, and TS, are the interaction terms.…”

“…These equations can be used to anticipate responses for a given amount of each component, and coded factor equations help to determine the relative https://doi.org/10.31436/iiumej.v24i2.2650 effect factors by comparing factor coefficients. The actual coded equations, however, cannot be used for determination since the coefficients are scaled to accommodate the units of each factor rather than being at the center of the design space [28]. The plot of the predicted response against the internally studentized residuals in Fig.…”

“…Response surface approach and a genetic algorithm tool were used to undertake statistical modeling in order to obtain the desirability function and forecast values. It was found that the RSM-GA predicted the experimental values accurately [27,28]. Utilizing RSM (Design Expert) and ANN-LM, a statistical model with three independent variables was created.…”

Artificial intelligence‐based multiobjective optimization of reverse osmosis desalination pretreatment using a hybrid ZnO‐immobilized/photo‐Fenton process

“…Based on foulant types, fouling in the SWRO system can be classified as organic fouling, inorganic scaling, colloidal fouling and biofouling. The biofilm that affects the RO membrane comprises 50% organic matter, among which 40% consists of organic acids 3 . Pretreatment procedures commonly employed in SWRO plants include filtration, disinfection, coagulation‐flocculation and anti‐scaling operations.…”

“…The biofilm that affects the RO membrane comprises 50% organic matter, among which 40% consists of organic acids. 3 Pretreatment procedures commonly employed in SWRO plants include filtration, disinfection, coagulation-flocculation and anti-scaling operations. The high costs involved in pretreatment processes could be minimized by incorporating renewable energy-based technologies like solar nanophotocatalysis.…”