3.2. Chlorine modeling
Chlorine modeling using estimated parameters for bulk decay models was tested with EPANET MSX. In a first
approach, a kw equal to zero was tested. Modeling bulk decay alone led to poor simulations, mostly with
overestimated chlorine concentrations (correlation plot in Figure 2a). These results were observed for all tested
bulk decay models. Similarly to what was observed for laboratory decay tests at 15ºC, 1st order kinetics led to the
higher RMSE than 1.2 order model and the 2R model was the one which better described bulk chlorine decay in
the water supply system (Table 3).