Predictive capability of chlorination disinfection byproducts models.

Area of Science:

• Environmental Chemistry
• Water Quality Analysis
• Predictive Modeling

Background:

• Over 100 models exist for predicting disinfection byproducts (DBPs) like trihalomethanes (THMs) and haloacetic acids (HAAs).
• No previous publication has standardized the evaluation of these THM and HAA models using a common dataset.

Purpose of the Study:

• To analyze the variability and performance of existing THM and HAA predictive models.
• To identify the most robust models for predicting THM4 and HAA6.

Main Methods:

• Evaluated 87 models from 23 publications.
• Utilized standard error (SE), Marquardt's percent standard deviation (MPSD), and linear coefficient of determination (R(2)) for analysis.

Main Results:

• The most robust models predicted THM4 with SE of 48 μg L(-1) and HAA6 with SE of 15 μg L(-1), both with R(2) > 0.90.
• The majority of developed models were for THM4.

Conclusions:

• There is a significant need for more published models predicting total HAAs, individual THM and HAA species, bromate, and unregulated DBPs.
• Model performance varies, highlighting the need for standardized evaluation and development of more comprehensive DBP prediction tools.