Efficacy evaluation of conventional water treatment process and THMs modeling in drinking water of five cities in India

Abstract

The presence of trihalomethanes (THMs) in drinking waters poses a high risk to a human’s health. The study aimed to monitor the level of THMs and to develop a mathematical model to predict the same in the water supply system of five cities (Varanasi, Dhanbad, Raipur, Bhubaneswar, and Kolkata) of India. The concentration range of total THMs (TTHMs) (348 and 414 μg/L) was observed approx. 2 times higher than the prescribed guideline value of Bureau of Indian standard. Raw water contained a comparatively higher amount of THMs precursors than treated water. The adopted conventional treatment technology in the existing unit was found marginal for removal of these THMs precursors (TOC 34% and DOC 37%). A machine learning predictive model viz. Support vector machine (SVM) was developed using MATLAB 9.5. Validation results confirmed the prediction precision of SVM models is good, which can provide immense help to WTP manager for the pre-knowledge of THMs in the water for its better management and control.