John E. Tobiason

Publication Date



Disinfection by-product (DBP) concentrations in drinking water distribution systems are of concern due to potential health effects. Due to the US EPA Stage 2 DisinfectantlDisinfection By-Product (DIDBP) Rule, proposed in 2003, water utilities are investigating the effects of distribution systems on water quality. Under the Stage 2 DIDBP Rule, there will be a change from use of a system-wide running annual average (RAA) to the use of locational running annual averages (LRAA) for compliance with DBP maximum contaminant levels (MCLs). The Aquarion Water Company of Connecticut (A WC) funded research into improved understanding of water quality. In the Stamford, CT distribution system, the objectives of the research were to: 1. Use fluoride tracer studies to assess hydraulic residence times; 2. Evaluate factors that affect DBP formation through field testing and laboratory experiments; 3. Understand temporal and spatial variations in DBP concentrations; and 4. In preparation for new regulations, use historical water quality data and research results to identify future water quality monitoring locations in the distribution system. The fluoride tracer studies were conducted during three separate sampling events (summer 2002, winter 2003, and summer 2003). In order to determine the spatial and temporal effects on water quality in the distribution system, field measurements of pH levels, water temperatures, and free chlorine residuals were made. In addition, samples were collected for laboratory analyses of natural organic matter (NOM), UV absorbance, and DBPs. The laboratory bottle tests used water taken from the SWTP clearwell effluent. Sterilized BOD bottles were filled at the beginning of the sampling events and held in a dark, temperature controlled climate to simulate the conditions in the distribution system. At various time intervals, bottles were opened and the free chlorine residual was measured while samples were collected for DBP analyses. For the summer 2002 sampling event, hydraulic residence times at monitoring sites ranged from 2 hours to 46 hours. Free chlorine residuals ranged from 1.14 mgIL at the - SWTP to 0.07 mgIL at the Long Ridge Fire Station (FS) #2 site, which had the longest residence time of the sites sampled. The TTHM concentrations in the distribution system increased with water age (maximum value = 88 p,gIL) while HAA concentrations increase to a maximum value (87 p,gIL) and then degraded for longer water ages. The effect of a water storage facility was observed at the Pond Road sampling site where an increase in fluoride concentration occurred during the step-down in fluoride dosage, probably caused by the discharge of older water with higher fluoride levels from the Roxbury storage tank. The winter 2003 event sampling sites were similar to those of the summer 2002 event in order to evaluate seasonal effects on water age. Compared to the summer 2002 results, water ages, which ranged from approximately 3 hours to 160 hours, increased due to lower water demands. The free chlorine residual (1.13 mg/L to 0.26 mgIL) had a much slower decay rate due to colder temperatures, which also resulted in the lack of noticeable HAA degradation. However, DBP levels were considerably lower than those found in the summer 2002 sampling event. The summer 2003 event showed similar trends as fonnd in the summer 2002 event. Water ages ranged from approximately 3 hours to 66 hours and the free chlorine residual ranged from 1.13 mgIL to 0.02 mgIL. TTHl\![ concentrations increased with increasing water age, reaching a maximum of III p.gIL. The peak HAA concentration (40 p.gIL) was fonnd at a residence time of 9.8 hours and showed signs of degradation at sites with greater water ages. As fonnd in the summer of 2002, some sampling sites had longer residence times due to the release of water from the Roxbury storage tanle Associated with these longer residence times were elevated THl\![ concentrations, indicating that the Roxbury storage tank contributed in some manner to the deterioration of water quality in the distribution system. Based on the Awe historical data and the findings of this research, recommendations were made as to potential water quality monitoring sites required nnder the proposed Stage 2 DIDBP Rule. These sites will be more indicative of high DBP levels and will factor in the effects associated with water storage facilities. iv