Water Journal : Water Journal July 2012
chlorine disinfection refereed paper technical features 68 JULY 2012 water Nitrogen-containing DBPs have been measured mainly in the United States (US), Canada and Europe (Richardson et al., 2007). Simpson and Hayes sampled 16 drinking waters from around Australia and measured the sum of four HANs expressed as 4HAN, chloropicrin and cyanogen chloride (Simpson and Hayes, 1998). They found values up to 36 μg/L of 4HAN in different regions of Australia compared to the median and maximum levels of 3 and 14 μg/L found in a 2000--2002 US survey (Weinberg et al., 2002; Krasner et al., 2006). In a 2006-- 2007 US survey, median values for the sum of 4HAN was slightly higher at 4 μg/L (Mitch et al., 2009). In our previous work, we measured the DBP formation potential at three WTPs in Brisbane and found the highest DBP formation from Capalaba source water (Farré et al., 2011). Therefore, we selected this region for further evaluation. In this paper we present the concentration of DBPs at Capalaba WTP (Brisbane) and its distribution system, which also includes water from North Stradbroke Island (NSI). Methodology Samples were taken at Capalaba WTP and from the distribution system (Figure 1) during five sampling events in spring-summer 2011. The sampling dates were 5/9/2011, 3/10/2011, 7/11/2011, 21/11/2011 and 6/12/2011. Samples were taken headspace free, in acid-washed glass containers with Teflon lids. Ascorbic acid was used to quench the chlorine and to protect the present DBPs. Samples were shipped to The University of Queensland (UQ), with volatile DBPs analysed within 24 hours. The following DBPs were extracted by liquid-liquid microextraction and analysed by gas chromatography with electron capture detection (GC/ECD): trichloromethane (TCM); bromodichloromethane (BDCM); dibromochloromethane (DBCM); tribromomethane (TBM); dichloroiodomethane (DCIM); bromochloroiodomethane (BCIM); dibromoiodomethane (DBIM); chlorodiiodomethane (CDIM); bromodiiodomethane (BDIM); trichloroacetonitrile (TCAN); dichloroacetonitrile (DCAN); bromochloroacetonitrile (BCAN); dibromoacetonitrile (DBAN); chloral hydrate (CH); trichloronitromethane (TCNM); 1,1-dichloropropanone (1,1-DCP); and 1,1,1-trichloropropanone (1,1,1-TCP). Additional analyses such as total organic carbon (TOC), SUVA, bromide and iodide were done. NDMA was extracted at UQ by means of solid phase extraction and concentration under nitrogen and samples were taken to QHFSS for analysis by gas chromatography coupled to mass spectrometry (GC/MS) with chemical ionisation with ammonia gas. Results Table 1 shows TOC, bromide, dissolved and organic nitrogen (DON) values and the standard deviation for the sampling points selected for this study and Capalaba WTP. Samples have been divided according to the source water used. Samples M16 and M17 provide water from Capalaba WTP and Alexandra Hill Reservoir, which is a mix of Capalaba and NSI water. Hence, a higher proportion of Capalaba water is expected for these sites in comparison to sites M19--M33, which supply water from Alexandra Hill Reservoir without further blending. Maximum average TOC values were around 5--6 mg/L and were found in sampling points with a higher percentage of water from Capalaba WTP, while lower values were found in sampling points providing water from NSI. Bromide concentration ranged between 0.025 and 0.07 mg/L across the samples. DON was clearly higher in Capalaba water in comparison to NSI. THMs in conjunction with HAAs are the most prevalent DBPs in drinking water and are formed as a result of the reaction between chlorine and NOM. Figure 2 shows the mean concentrations of tTHMs found in Capalaba WTP and the distribution system. In all instances, tTHM values were lower than the ADWG value of 250 μg/L. Elevated values were found in M8--M17 which correspond to sampling points providing water from Capalaba WTP. On the other hand, the remaining sampling points provided water with a high contribution of NSI-treated water. Average concentrations of tTHMs at Capalaba WTP were lower than at M8, M11, M13, M14 and M16, which evidenced the ability of THMs to increase during distribution within the system, mainly as a result of hydrolysis of other DBPs (Nikolaou et al., 2001). Sampling points providing water from Capalaba showed a distribution of TCM>BDCM>DBCM>TBM, which is common speciation in drinking water with a high concentration of organic carbon and low concentration of bromide. The speciation of THMs measured in sampling points providing water from NSI was DBCM>BCDM>TBM>TCM as a result of the presence of bromide in low organic carbon waters (< 2 mg/L). The rate constant of bromide with HOCl to generate HOBr is 1.5 × 103 1/M·s (Kumar and Margerum, 1987) and the rate constant of THM formation is in the range of 0.01 and 0.03 1/M·s (Gallard and Von Gunten, 2002). It is known that, once formed, hypobromous acid reacts about 10 times faster than chlorine with NOM. The reason is that the activities of electrophilic substitution for electron release to stabilise a carbocation are more favourable for the Br atom due to its higher electron density and smaller bond strength relative to the Cl atom (Westerhoff et al., 2004). Hence, the formation of Br-DBPs is limited by the initial Br concentration, whereas the Capalaba WTP M 11 M13 M 14 M 10 M 8 M 22 M 17 M 16 M 21M 23 M 19 M 29 M 33 M 52 M 51 M 50 M 38 M 42 M 45 Alexandra Hills Reservoir Figure 1. Sampling points selected for the study. NSI = North Stradbroke Island. Green = Capalaba WTP source water. Red = NSI WTP source water. Purple = Alexandra Hill Reservoir (Capalaba + NSI). Blue = Capalaba + Alexandra Hill Reservoir. White arrow represents water from Capalaba to Alexandra Hill Reservoir (Courtesy of Allconnex Water).
Water Journal August 2012
Water Journal May 2012