The impact of mixing on the chloramination process

Abstract

Interest in chloramination treatment for potable water has increased in recent years with the advent of more stringent drinking water requirements as a result of chloramines ability to form stable chlorine residuals and minimal disinfection byproducts. However, many utilities which have switched to chloramination for secondary disinfection have reported problems upon the start-up of their new systems. It is believed that these issues may be the result of inadequate mixing. A previous study by Jain (2007) identified an optimum mixing intensity for adequate chloramination by completing bench scale experiments on a standard jar test apparatus. This study was limited to three mixing speeds and further analysis was required to confirm Jain's findings. The purpose of this study was to continue Jain's study and further investigate the impact of mixing on the chloramination process. A series of nine experiments were conducted at mixing speeds of 200 rpm and 300 rpm using a jar test apparatus. Ten dosing molar ratios of chlorine to ammonia were investigated in each experiment and sampling times of 15 and 45 minutes were utilized. Experimental conditions were set to correlate to those commonly present in drinking water treatment facilities and sodium bicarbonate was added to provide alkalinity. Experimental results were compared against model simulations produced by the Unified Plus Model system. Experimentation indicated that monochloramine production approached the theoretical maximum at a mixing speed of 200 rpm (G = 300 s-1). Similar results were obtained when mixing speed was increased to 300 rpm (G = 500 s-1). As such, the increase in mixing speed did not produce further benefits. Additionally, the breakpoint curve exhibited a left shift from the ideal breakpoint curve. As a result, further investigation was conducted into the application of carbonate to the experimental system. It was determined that the chloramination process was impacted by increasing the Ct. The mechanism for this impact is not completely known. Model comparisons with these results indicated that the system did not definitively exhibit the characteristics of either an open or closed system, but appeared to be closer to a closed system.