Recently, many consumers are complaining about their drinking water quality, so besides water sources and treatment utilities, the maintenance of water distribution network systems is a point of new issue. In fact many researchers have investigated hard and actively about the processes from water sources to treatment utilities but they have passed over the effect of water distribution system on drinking water including pipes and distribution reservoirs. As the drinking water distribution system for wide areas is introduced, the rapid expansion of supply pipes has caused many problems including increasement of carrying capacity, difficulty of maintenance and operation due to flow conditions. So, in order to distribute potable water to the consumers' tap safely, at first it is required to develop a model with which operators can estimate and guard the drinking water in the distribution systems. This work defines secondary contamination in drinking water as the pollution which occurred between treatment utilities and consumers. We researched chlorine decay as an indicator of this pollution and developed a prediction model using experiment data.
This work consists of three experiments; A batch experiment, a simulated distribution system experiment and an application to field experiment.
The first experiment is to detect the chlorine decay in batches, where the velocity is 0. The batch vessel is made of a new pipe and a used pipe which was buried in a distribution system for 10 years. The parameters which we wanted to detect in order to develop a model are chlorine concentration, iron concentration, time, temperature and TOC concentration. We have set the total detect time to be approximately 28hours which is the general retention time in field distribution systems and initial chlorine concentration to be 1mg/L which is similar to that of a treatment utility. To simulate the breakdown of a pipe or the invasion of sewage, we have set TOC concentration using Hexose and sewage, and changed temperatures using incubators. The result of the experiment showed that the effect of iron dissolved from pipe wall is critical. Many researchers have shown that the cement mortar lining could prevent corrosion, and that the hardness and alkalinity materials could have a critical effect on water quality. But factually, the result of experiment have shown that these have little effect on drinking water for the duration time which the drinking water is in system. The behavior of iron, which can react with chlorine very rapidly, is divided into two steps. The first is the range within which the capacity of iron reacted with chlorine is larger than that of iron dissolved from pipe wall, so the total concentration is decreasing as time goes by. The other is the range within which completing consumption of chlorine, the capacity of iron dissolved from pipe wall is larger than that of the iron reacted with chlorine, so the total concentration is increasing as time goes by. In order to predict the concentration of chlorine, we should analysis the relationships between chlorine decay and each parameters because each parameters have different effect on each other, and select the parameters to use for model in the order of relationship. When we use the sewage for setting TOC concentration, the result of the experiment showed that it is ideal guard for water quality to detect both the free chlorine and the total chlorine continuously.
The second experiment is to detect the chlorine decay in a simulated distribution system(SDS), where the velocity is approximately 1m/sec. The SDS is made of the cement mortar lined pipe and PVC pipe which is generally used for most outdoor distribution systems. The main difference between the SDS and the bach experiment is flow velocity. Through this experiment, we investigated the effect of velocity on chlorine decay and the difference of decay coefficients between in cast iron and in PVC. The result of this experiment has shown that the friction due to flow accelerated the chlorine decay. The trend of concentration of iron is similar to that of the batch test. The third step is to calibrate the results of research and apply these model to field distribution systems. The site used for calibration and application has 10,090 consumers, 2 distribution reservoirs which have the capacity of 660 tons, totally 387 hydrants in 15 divisions. Hydrants could divided into three categories, 324 domestic hydrants, 39 commercial hydrants, 24 public hydrants. The distribution systems were made of cast iron and PVC pipes. And there are no chlorine injector from the distribution reservoir to consumer' tap.