Optimization of process parameters for remediation performance on mass removal of tetrachloroethylene source zones by using Taguchi design of experiment and artificial neural network

Abstract

This study tested the remediation performance of various conditions for pool‐dominated tetrachloroethylene resided in the soil at three different levels of type of flushing agent, flushing agent concentration and flushing rate. The experimental data obtained according to the Taguchi orthogonal array L9 were analysed using the Minitab 17 program. Results of the experimental design analysis indicated that the most effective factor is the flushing agent type, followed by flushing agent concentration and flushing rate, respectively. All the parameters used in the system have a significant effect on the experimental results, and the Taguchi method can explain 97.94% of the experimental data. As a result of the ANN analysis, the R2 values for training, testing and validation were calculated as 0.99942, 0.99795 and 0.99964, respectively. This situation shows the importance of experimental design and selection of appropriate conditions, which is very important for rapid and cost‐effective remediation of contaminated sites.

Multi‐response optimization of process parameters for remediation of tetrachloroethylene pools by surfactants: Application of Taguchi design of experiment and Artificial Neural Network

“…Optimization techniques in groundwater studies have mostly been used in the removal of pollutants and groundwater remediation studies (Akyol, 2018; Akyol & Turkkan, 2018; Akyol et al, 2013; Difilippo et al, 2010; Mohammed et al, 2019; Şahin et al, 2022). The experimental design method used in studies on the purification of such compounds is quite limited, and two studies have been conducted so far.…”

“…In particular, it has been demonstrated by the experimental design method that the surfactant concentration is highly effective in the remediation performance. In another study by Şahin et al (2022), optimization of improvement performance with Taguchi experimental design using surfactants Tween 80, MCD and SDS in low density soils contaminated with tetrachloroethylene (PCE) and prediction of improvement performance as a function of different parametric conditions with Artificial Neural Network (ANN) method were investigated. Consequently, it will be possible to predict the results using Taguchi and ANN methods without the need for experiments related to different experimental conditions.…”

“…Consequently, it will be possible to predict the results using Taguchi and ANN methods without the need for experiments related to different experimental conditions. Şahin et al (2022), in their study, revealed that the most effective parameter is the Surfactant Type, followed by the Surfactant Concentration and Flushing Rate parameters, respectively.…”

“…In addition, when the studies in the literature were examined, it was noted that the use of ANN method within the scope of the researched decision problem was carried out only with Şahin et al, 2022. Therefore, it is considered that a significant contribution is made to the existing literature in this regard.…”

“…Since the mathematical modelling of ANNs by McCulloch and Walter Pitts in 1943, it has been used in many fields such as engineering, medicine, agriculture, sociology and psychology for learning, association, classification, generalization, estimation, feature determination and optimization purposes similar to human intelligence (Elizonda et al, 1994; Fang & Wang, 2017; Harirchian et al, 2022; Kara & Cemek, 2019; Park et al, 2005; Rasam et al, 2019; Rumelhart et al, 1986; Schmidhuber, 2015; Şahin et al, 2018, 2022; Taoufik et al, 2021; Wang et al, 2021; Wen et al, 2021).…”