Monitoring and modeling chlorine levels in water distribution systems

This paper describes an ongoing research project aimed at developing a reliable computer model to forecast chlorine residuals throughout the 20 km long Myponga pipeline, south of Adelaide. In spite of the high chlorine dose at the treatment plant (~6mg/l), the chlorine residual is virtually zero at the end of the pipe. By understanding the chlorine decay kinetics in this pipe and the parameters influencing this decay, a model forecasting the free chlorine residual will be developed. It will provide the basis for developing feedback systems that can be used to adjust chlorine dosing levels at the treatment plant in response to residual levels at the extremities of the distribution system. The research involves laboratory measurements as well as a field study. The laboratory measurements are undertaken in order to select the most appropriate chlorine decay kinetic model and to understand the parameters involved in this decay. Of particular interest are the four fractions of natural organic matter (NOM) extracted using ionexchange resins. The field study involves chlorine and flow monitoring by on-line sensors along the pipe. Instead of using sparse data collected along the pipe, this extended period of monitored data will enable a good calibration of the computer model. This paper describes the methodology used in the study as well as the difficulties that the fieldwork imposes