A Variogram-Based Tool for Variable Selection in a Wastewater Treatment Effluent Prediction

Abstract

In this study, a variogram method was utilized as a variable selection tool for finding the optimal subsets of variables for developing predictive models for the quality of wastewater treatment effluent. The quality of effluent was here assessed by biological and chemical oxygen demand and suspended solids in biologically treated wastewater. The dataset included, in addition to traditional process measurements, results of a novel optical monitoring device which was used for imaging an activated sludge process in-situ during a period of over one year. The study showed that the variogram based method has potential in fast and computationally easy variable selection. The developed models can be used for proactive monitoring and estimating the quality of effluent in several stages hours before in comparison to laboratory analysis taken from treated wastewater.