Predicting the Formation of Disinfection Byproducts Using Environmental Parameters in Chlorinated Drinking Water

Safe drinking water is crucial, yet 2.2 billion people lacked access to it in 2022. Chlorination, a common disinfection method, can produce harmful disinfection by-products (DBPs) like trihalomethanes (THMs) and halo-acetic acids (HAAs). These DBPs pose health risks, but only 30% of over 700 identified DBPs have been quantified, emphasizing the need for better predictive models. This project aims to further understand the relationship between water parameters such as, pH, alkalinity, DOC/TOC and SUVA/UVA, and the concentration of DBPs after chlorination; by doing a data analysis on publicly accessible data from the EPA and trying to fit various regression models for it. The best fit model achieved an R2 of 0.5 which was a Ridge Regression model for the bromoform DBP. The findings reveal the challenges to making accurate predictive models without enough good quality data.