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


ŞAHİN Y., Kasap S. S., Akyol G., AKYOL N. H.

Water and Environment Journal, vol.37, no.3, pp.454-469, 2023 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 37 Issue: 3
  • Publication Date: 2023
  • Doi Number: 10.1111/wej.12849
  • Journal Name: Water and Environment Journal
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Agricultural & Environmental Science Database, Applied Science & Technology Source, Aqualine, Aquatic Science & Fisheries Abstracts (ASFA), CAB Abstracts, Compendex, Computer & Applied Sciences, Environment Index, Geobase, Greenfile, INSPEC, Pollution Abstracts, Veterinary Science Database
  • Page Numbers: pp.454-469
  • Keywords: ANN, cosmetic, cyclodextrine, DNAPL, experimental design, multi-response Taguchi, sodium dodecyl sulfate, tetrachloroethylene, Tween 80
  • Kocaeli University Affiliated: Yes

Abstract

Within the scope of the study, the effectiveness of the experimental conditions was tested by performing a multi-response Taguchi experimental design for the optimization of the minimum cost remediation performance with Tween 80, Methyl beta cyclodextrine (MCD) and Sodium dodecyl sulfate (SDS) from tetrachloroethylene (PCE) contaminated porous media. Tween 80, MCD and SDS were extensively used in cosmetic industry as emulsifier. Both time of remediation and cost of remediation were studied as two separate response variables in three replicate experiments conducted according to the Taguchi L9 orthogonal experimental design. In the multi-response Taguchi analysis, the sensitivity analysis was performed by systematically changing the weights determined for two separate response variables in the calculation of total loss of quality (TNQLj). Optimum experimental conditions were determined with the help of the calculated multi-response signal/noise (S/N) ratios (MRSN). The results show that the type of Flushing Agent is the most important factor in optimizing the remediation time and remediation cost for the removal of dense non-aqueous phase liquid (DNAPL) PCE mass. Flushing rate is considered to be the least contributing factor. Furthermore, the results of the analysis of variance (ANOVA) showed that all parameters used in the system had a significant effect on the experimental results and the Taguchi method could explain 97.15% of the Remediation Time and 92.03% of the Remediation Cost. Afterwards, the data obtained from the experiments performed according to the experimental design were modelled using Artificial Neural Network (ANN) to estimate the remediation performance and remediation cost without performing new experiments.