Artificial neural network modeling for decolorization of textile dye effluent

Introduction Effluents from the textile industry are not only colored but also contain large amounts of color, COD, BOD and dissolved solids. Textile industry is the one producing huge amount of wastewater. Due to stringent environmental regulations, treatment is mandatory before discharge the effluent. Conventional treatment methods include activated carbon adsorption, solvent extraction, chemical and biological methods. These methods are having their own pros and cons when they applied individually [1]. Recent years advanced oxidation technique is growing much faster to treat the industrial effluents. Due to the increasing economic, social, legal and environmental pressure the best available technology to perform without producing any secondary pollution is electrochemical technology. The advantage of electrochemical treatment has environmental compatibility, versatility, energy efficiency, safety, selectivity, amenability to automation, and cost effectiveness [2]. Electrochemical methods have been successfully applied in treatment of several industrial waste waters [3-6]. Fernandes et al., [7] studied the electro oxidation of CI Acid Orange 7 using boron doped diamond electrode. Yusuf Yavuz and Savas Koparal [8] studied the electro oxidation of phenol in a parallel plate reactor using Ruthenium metal oxide electrode and Apostolos Vlyssides et al., [9] studied the degradation of methyl parathion in aqueous solution using Ti/Pt and Stainless steel as the electrodes. Kinetic modeling of electro oxidation process is very difficult to determine because of various steps involved in the electro oxidation, leading to uncertainties in the design and scale –up of chemical reactors of industrial interest. An alternative modeling technique to model the process is Artificial Neural Networks (ANN). ANN does not require any mathematical description of the phenomena involved in the process, and might therefore prove useful in simulating and upscaling complex electro oxidation systems. The network strongly depends on the process variables involved in the process as well as set of data and the domain used for training purpose [10, 11]. In the present study decolourization of textile effluent containing Acid Red 88 dye effluent by electro oxidation method has been investigated. The effects of operating parameters such as Effluent flow rate, initial effluent concentration, current density and initial pH on color removal efficiency have been investigated. An important objective was to obtain an ANN model that could make reliable prediction of the color removal efficiency of the electro oxidation process. Material and methods Experimental The commercial dye used in this investigation was purchased from Qualigens Fine Chemicals, Mumbai, India. The chemicals used in the study were of analytical grade. The chemical structure and other characteristics of this dye are shown in Table 1.Dye solution was prepared by dissolving the dye in the distilled water. The electrolyte concentration was adjusted using sodium chloride (NaCl). The electro oxidation was carried out in a Continuous Stirred Tank Electrochemical reactor. Schematic of an experimental set up is shown in Figure 1. Experiments were conducted in a Continuous Stirred Tank Electrochemical Reactor having 300 ml capacities. Oxide coated titanium metal is used as anode and stainless steel is cathode. Experiments were carried out under galvanostatic conditions using a DC-regulated power source (HIL model 3161) of 0–5A and 0–30V. Synthetically prepared Acid Red 88 dye effluent is taken in the cell and uniform mixing is achieved by using magnetic stirrer. Samples were collected periodically and analyzed. The pollutant degradations were estimated by UV/Visible spectrophotometer (Hitachi, Japan).