Thomas L. O’Connor, BSEE, MBA, is project director, H2O’C Engineering, Columbia, Mo. e-mail: [email protected]
Bart Murphy, BSCE, is superintendent of Water, Willmar (Minn.) Municipal Utilities.
John T. O’Connor, EngD, PE, is principal, H2O’C Engineering, Columbia, Mo. Phone: 877-22-WATER; e-mail: [email protected].undefinedA study aimed at distinguishing between the comparative activity of the nitrifying organisms Nitrosomonas europea and Nitrobacter winogradskyi,1 indicated that the chlorite ion interfered with the activity of N. europea that mediates the oxidation of ammonium ion to nitrite ion. While seemingly only of academic interest, this research formed the scientific rationale for the application of chlorite ion to suppress nitrification in water distribution systems.
Subsequently, the results of water system studies have suggested that chlorite formed as a by-product of the application of chlorine dioxide may serve to inhibit nitrification during distribution.2 Calling for field verification, the authors suggested the direct addition of sodium chlorite might be an effective alternative nitrification control process.
Control of Nitrification
Since 1996, Willmar (Minn.) Municipal Utilities has used breakpoint ammonium reduction treatment (BART) to control nitrification in its water distribution system.3 Although effective, BART requires the use of large amounts of chlorine for the oxidation of ammonium ion that is naturally present in the groundwater.
This study was undertaken to assess an alternative procedure for nitrification control: the addition of sodium chlorite, a chemical used both in water treatment and the food processing industry.
Initial trials involved the addition of low concentrations of sodium chlorite to the finished water of Willmar’s southwest water treatment plant beginning on May 12, 2000. Prior to and following this addition, extensive monitoring was conducted throughout the distribution system.
Three weeks after the initiation of a chlorite feed, a significant beneficial response was observed. Nitrification in the distribution system markedly decreased. Both oxygen and chloramine residuals increased at distribution sampling sites where they were formerly depleted. The implication of these results is that the activity of nitrifying bacteria in the distribution system was curtailed by the initial addition of sodium chlorite.
This procedure was replicated at each of Willmar’s two water treatment plants serving the southwest and northeast portions of the distribution system. Comparable results were obtained at each plant and in each portion of the system.
In September 2000, distribution system monitoring results indicated an increase in nitrification. This was attributed to the increased amount of ammonium ion available in conjunction with relatively high water temperatures and seasonally-decreased demand. The chlorite feed was increased and the importance of continuously modifying the feed based on monitoring was reinforced.
Dissolved oxygen is one of the most sensitive and readily-monitored indicators of nitrification in the distribution system. At Willmar, highly-efficient tower aeration routinely results in saturation of dissolved oxygen while also increasing pH to more than 8.0 by stripping carbon dioxide. At distribution sampling locations where nitrification was evident, dissolved oxygen concentrations declined to 6 mg of oxygen per liter (6 mg O/L). Within weeks of the start of chlorite addition, DO concentrations exhibited a steady increase, converging on the saturation values observed in the treatment plant finished water.
Disinfectant residuals also may serve as a sensitive index of the progress of nitrification. However, at Willmar, applied chloramine residuals are modified regularly based on distribution system data so that chloramine levels can be maintained throughout the distribution system using the minimum chlorine addition possible. Willmar’s treatment goal is to attain conditions in the distribution system that will allow chloramine residuals to remain relatively constant throughout the entire system. Under these circumstances, lower chlorine applications will be possible, and the chloramine monitoring data will converge.
An increase in chloramine persistence occurred following the addition of chlorite ion, allowing the chlorine feed at the southwest plant to be reduced. Alternately, chloramine levels increased at the sampling sites that previously exhibited the strongest evidence of nitrification.
The addition of sodium chlorite was expected to inhibit the first stage of nitrification. If inhibition was effective, levels of ammonium ion in the distribution system would increase to match the levels in the plant effluent. This effect was evident after three weeks, when distribution system samples matched up with finished water at approximately 1 mg of nitrogen per liter (1 mg N/L). Further increases in ammonium ion concentrations were expected when BART was discontinued at both plants.
Reduction of Nitrite and Nitrate Formation
Inhibition of nitrification would be expected to be accompanied by further reductions in the already low concentrations of nitrite and nitrate. Such reductions are evident in the distribution system where nitrite ion declined to an average of less than 0.1 mg N/L, and average nitrate ion concentrations bottomed out at 0.2 mg N/L.
The second phase of the study was undertaken to observe the effect of discontinuing the BART procedure. This immediately resulted in a doubling of the ammonium ion concentration in the water entering the distribution system.
The BART process was discontinued first at the southwest plant and then, after several weeks, at the northeast plant. The accelerated monitoring program was continued as ammonium ion levels increased throughout the distribution system. The increased concentration of ammonium ion is likely a contributor to the observed increase in nitrification.
Reduction of Chlorite Feed
The initial dosage of sodium chlorite was progressively reduced to determine the minimum effective chlorite dose required to control the activity of nitrifying bacteria. Chlorite concentrations in the distribution system declined in accordance with reductions in feed rate at the treatment plant, indicating that even at lower applied concentrations, chlorite persisted in the distribution system.
Many variables affected the authors’ ability to clearly observe the effect of the chlorite addition. In Willmar, two treatment plants with significant differences in source water quality provide water from opposite sides of town. Operationally, rates of chemical feed and distribution system blowoffs were modified throughout the study. Seasonally, factors such as reduced demand (increased water age), lower temperatures and the behavior of major industrial water users all had effects on distributed water quality.
Despite the complexities involved in a real distribution system, it was clear that the addition of sodium chlorite resulted in a reduction of nitrification. It was recommended that Willmar discontinue the BART and utilize sodium chlorite for the control of nitrification. Significant cost savings and safety benefits are expected to accompany the sharply reduced use of chlorine.
It was also determined that the use of sodium chlorite to control nitrification is not a set-it-and-forget-it type of chemical feed. Ongoing distribution system monitoring is required to continuously fine-tune the dosage in order to maintain an effective feed rate.