In this study, a method of estimating the ultimate concentration and the rate constant in a kineticmodel for disinfection byproduct (DBP) formation was developed by using a nonlinear regression technique to fit themodel to calculated water age (from water distribution system hydraulic models) and DBP fieldsampling data from actual water distribution systems. To better understand the chemistry inreactions of HAAs with elemental iron, and perhaps incorporate associated rate constant data indistribution system water quality models, batch experiments were also conducted with varyingchlorine residual concentration and constant pH.Results indicated that the regression method is fairly accurate in predicting HAA concentrations;however, limitations were observed in the method's ability to predict peak concentrations. A bigadvantage of this fitting technique is that it can potentially account for HAA biodegradation andother distribution system reactions that are typically not characterized or quantifiedexperimentally in the traditional bottle tests. The lab results in this study indicated that DCAA(dichloroacetic acid) was converted to CAA (chloroacetic acid), and that CAA did not degradefurther. BAA (bromoacetic acid) was converted to acetic acid. Rates did not seem to bedependent upon the level of chlorine residual. There is still a lack of understanding of HAAdegradation rates under various conditions of pH, chlorine residual, and iron loading. Rateconstants extracted from more rigorous bottle tests, such as from these experiments, couldpotentially be used in distribution system models to make better predictions of the amount anddistribution of HAAs. Includes 8 references, tables, figures.
