Experimental study of mixing phenomenon in water distribution networks under real-world conditions

Abstract

Various studies have shown that mixing at the junction of water distribution networks (WDNs) can be neither complete nor instantaneous. Many studies have also been carried out to find the parameters that have a significant effect on the mixing phenomenon by experimental and numerical investigations. The Reynolds ratio of inlet flows, the Reynold ratio of outlet flows, the pipe size of junction's legs (junction configurations), and the type of junction (cross or double-t) are among the most important factors mentioned in these studies. Other studies also focused on developing mixing models. However, these studies and models were based on experimental conditions in which the pipe size was about 25 mm, and the pressure was about 30 kPa. Despite the fact that these models provide acceptable results on the laboratory scale and all of them have been validated based on laboratory results, many researchers still acknowledge that studies on the effects of real conditions of urban WDNs on the mixing phenomenon are lacking [1], [2] and [3]. The pipe size of junctions and the pressure inside the network are distinguishing features of real urban WDNs and laboratory setups that were used in previous studies. Therefore, this research investigates the effect of pipe size and pressure on the mixing phenomenon in a cross junction with the same pipe sizes in all four legs under real-world conditions. Flow rates and pressures were selected based on a statistical study conducted on the Quebec City WDN and previous research work, in which the pressures were 5, 140, 320 and 430 kPa, and the flow rates were a combination of 1.50, 2.00, 2.25, 2.50 and 3.00 LPS in each inlet and outlet pipes. In other words, the Reynolds number in the experiments was between 21,000 and 43,000, and the Reynolds ratio of inlet flows as well as the Reynolds ratio of outlet flows were 0.5, 0.8, 1, 1.25 and 2. In this network, two cross junctions with the same pipe size in each junction's legs of 100 and 150 mm were investigated. Salt was used as a traceable solute, and conductivity meters were used to measure the salt concentration in pipes. In this series of experiments, only one inlet had salty water. The mixing was characterized by the dimensionless concentration in outlets, defined as the observed conductivity in one of the outlet flow divided by the inlet salty water conductivity (after subtracting the conductivity of tap water for both measurements). The results showed that the pressure had only a little effect on mixing, since the dimensionless concentration changed by about 0.05 when considering the low pressure of 5 kPa as compared to the experiments with the other pressure values, while this variation dropped to about 0.02 for all pressures above 140 kPa. Between 100 and 150 mm, the pipe size modified the dimensionless concentration by about 0.035. Taking into account the uncertainty of the experiments, it can be concluded that pressure and pipe diameters (100 and 150 mm) have an insignificant impact on the mixing phenomenon.