Abstract
According to limitations of previous chlorine decay models, a new model named as variable rate coefficient model (VRC) was built to improve the accuracy and reliability of chlorine decay model. The VRC model includes 4 parameters (X , κ , k and α) and is based on the kinetic rate equations from concurrent bimolecular second-order reactions. In this study, 6 data sets from different water sources were used. Three models including VRC model, the first-order model and a second-order reactive species model built by Boccelli et al. were calibrated and compared. Results show that compared with other two models, the simulated values of VRC model agree very well with the experimental data. The VRC model can accurately predict chlorine decay. In order to ensure the accuracy of VRC model, it is preferable to calibrate the VRC model using data sets with a chlorine concentration that is at least as high as or even higher than that used during normal operating conditions. Furthermore, the chlorine concentrations used as model input should be adjusted to minimize the differences between model predictions and the observed chlorine concentration.