US EPA

8167500 

Meetings & Symposia 

Profile of manganese breakthrough in filter effluent after chlorine loss 

Charlton, ND; Kohl, PM 

2005 

Quebec City, QC 

2005 Water Quality Technology Conference, WQTC 2005 

English 

Manganese control can be achieved by maintaining a chlorine residual across oxide-coated filter media. But what can one expect if chlorine feed is lost? To answer this question, the Philadelphia Water Department (PWD) conducted pilot studies focusing on how long manganese control may be maintained in the event of a chlorine loss. The main variables were pH and temperature, although the influent manganese concentration did vary slightly as well. PWD also investigated the use of oxidationreduction potential (ORP), presented as EH, as a predictive tool for manganese breakthrough. Testing was performed at the Baxter and Belmont Pilot Plants. Two sets of experiments (one in the spring and one in the fall) were performed. The primary objective was to evaluate how long it would take before manganese break through after a sudden loss of chlorine occurred. A secondary objective was to determine whether ORP could be used to predict filter effluent manganese breakthrough. Breakthrough was defined as filter effluent containing more than 0.015 mg/L total manganese. Conventional treatment at coagulation pH 6.5 was employed, with two manganese oxide-coated dual media filters in operation at each plant. Filtration pH of 6.5, 7.5 and 8.5 were tested. Sodium hypochlorite (NaOCl) was applied to the influent of each filter with the goal of achieving a 0.5 mg/L free chlorine residual at filter effluent. Raw, settled and filter manganese were sampled after steady state conditions were achieved (but just prior to chlorine shutoff) to establish a baseline. Filter effluent total and dissolved manganese and ORP were sampled just prior to chlorine shutoff, and at defined intervals after chlorine shutoff until the filter run terminated. If time permitted, the filters were backwashed and restarted (without chlorination) for a subsequent filter run. The filter effluent manganese and ORP data were plotted against a triple-axis of actual sample date/time, elapsed filter run time, and elapsed time since chlorine shutoff. Results from both the Baxter and Belmont pilot plants indicated that the filters held manganese longer at higher pH and in warmer water. Manganese breakthrough was not, however, dependent on influent manganese concentration at the low levels used in these experiments (50-100 μg/L). The Belmont plant operated with higher manganese loading yet controlled manganese for as long or longer than the Baxter filters. ORP dropped significantly within the first hour after chlorine termination and became essentially constant after that. ORP did not appear to be a predictive tool for manganese breakthrough from the data collected. © 2005 American Water Works Association WQTC Conference All Rights Reserved.