Predicting THM Formation and Revealing Its Contributors in Drinking Water Treatment Using Machine Learning

Abstract

Trihalomethanes (THMs) are disinfection byproducts (DBPs) that are formed during chemical disinfection of drinking water. However, a variety of factors, including water qualities and treatment conditions, can influence THM formation, making it challenging to predict and highlight the underlying mechanisms. Here, we used machine learning (ML) algorithms to analyze the complex relations between influent characteristics and operational parameters to predict THM formation. Five ML algorithms were used to create predictive models for THM formation using five input parameters (i.e., chlorine dose/DOC, reaction time, pH, bromide concentration, and temperature). The least-squares boosting (LSBoost) model demonstrated its superiority in predicting THM concentrations on the training and test sets with both R 2 values of 0.92. We also established a variant of the LSBoost model with an additional parameter, the specific ultraviolet absorbance (SUVA). Results showed that incorporating SUVA into model development gave a higher prediction accuracy on the training set with a reasonable R 2 value of 0.97. The partial dependence plots (PDPs) and Shapley additive explanation (SHAP) models were employed for in-depth analysis to identify key drivers for THM formation. Our results showed that chlorine dose/DOC was the major driver for THM formation and SUVA had the least impact on THM formation.

Applications of Artificial Intelligence, Machine Learning, and Data Analytics in Water Environments

“…The special issue includes several review articles encompassing a wide spectrum, ranging from a historical perspective of water data to computational modeling in wastewater treatment to ML modeling of environmental chemical reactions, environmental toxicology, heavy metal removal, and cyanobacterial harmful algal blooms (HABs) . One significant application of these innovative tools is ML-assisted environmental monitoring, which can address diverse problems, such as predicting effluent nutrients or influent flow rates and nutrient loads at wastewater treatment plants, , formation of disinfection byproducts, drivers of the accumulation of potentially toxic elements in sediments, greenhouse gas emissions, , occurrence of PFAS, water quality assessment, microplastics, microcystins, and differentiation of landfill leachate and domestic sludge . ML has also been extensively employed to model environmental chemical reactions and processes, including adsorption onto various materials, , biodegradation, photodegradation, and the physicochemical and meteorological variables that affect the seasonal growth and decline of HABs .…”