Turk J Chem 35 (2011) , 245 253. c TUBITAK doi:10.

3906/kim-0910-242

Assimilable organic carbon generation from algogenic organic matter in drinking water
Ji-Hoon KIM1,, Young-Ju KIM1 , Tahir Irman QURESHI2 1 Department of Environmental Engineering, Kyungpook National University, Daegu-KOREA
2

e-mail: kimjihoon@knu.ac.kr Department of Chemical Engineering, NFC-Institute of Engineering and Technological Training, Multan-PAKISTAN

Received 26.10.2009

A study of assimilable organic carbon (AOC) derived from algogenic organic matter in dierent drinking water resources was conducted. Seasonal changes in the concentration of AOC at Kamafusa Dam were dependent on the algal cell numbers. Variations in AOC concentrations were found to be more reliant on the AOC-P17 component than the AOC-NOX component. The algal culture experiment showed that extracellular organic matter (EOM) released during the growth of Phormidium tenue in M-11 culture medium led to a signicant increase in AOC concentration; however, no signicant variations of AOC concentration were observed in CT culture medium containing high amounts of dissolved organic matter. The AOC included in EOM was not easily removed by chlorination; however, the AOC included in intercellular organic matter (IOM), which was released from algal cells, was easily removed under conditions where residual chlorine was detected. Key Words: Dissolved organic compounds, bacterial regrowth, chlorination, algal material

Introduction
Maintenance of drinking water quality in water supply systems beyond water treatment plants is very important. In addition to other factors, the quality of drinking water is primarily aected by bacterial regrowth in the water supply piping network. To avoid this problem, residual chlorine has been used in water to suppress the multiplication of bacteria. However, it was learned that heterotrophic bacteria adhere to the inner surface of the
Corresponding

author

245

Assimilable organic carbon generation from algogenic..., J. H. KIM, et al., pipes, proliferating and forming a biolm under conditions with residual chlorine. 12 Regrowth of bacteria and detachment of the biolm continue over time, causing deterioration of water quality in the water distribution system. 34 In general, bacterial regrowth is associated with the presence of assimilable organic carbon (AOC). Van der Kooij 57 suggested that AOC should be less than 10 g/L to limit the growth of heterotrophic plate counts (HPCs) of bacteria in unchlorinated systems. Therefore, reduction of AOC is also an important measure for controlling bacterial regrowth. In Japan, interest has not been turned toward reduction of AOC, since the maintenance of residual chlorine has successfully controlled bacterial regrowth. However, reduction of AOC may be eective both in reducing chlorine doses and controlling bacterial regrowth. The amount of AOC is dicult to control by conventional water treatment techniques. 810 Therefore, the behavior of AOC in water resources needs to be studied. In particular, the eects of eutrophication on the behavior of AOC are not well determined, even though eutrophication signicantly aects the water purication process. Algogenic organic material (AOM) is not easily degraded, chemically or biologically. Algal blooms of blue-green algae or diatoms cause diculties in controlling dissolved organic carbon (DOC) because of the release of extracellular organic matter (EOM) and intracellular organic matter (IOM) into the water by cell lysis. Algae and extracellular organic matter can be precursors of disinfection byproducts (DBPs). The existence of AOM composed of such substances as glycolic acid, carbohydrate, amino acids, and organic phosphorous might be able to cause an increase in AOC concentration. 1114 This study was conducted in order to examine variations in the AOC concentrations of drinking water and to learn characteristics of AOC derived from algae. The seasonal changes in the AOC concentration of Kamafusa Dam, the main reservoir of Sendai, Japan, and of 3 inow rivers were investigated. A series of experiments was conducted to see the inuence of algae on AOC concentrations. Generation of AOC derived from the growth of blue-green alga, Phormidium tenue, in dierent culture medium conditions was studied. Assessment of the formation and removal of AOC derived from EOM and IOM in the course of chlorination was also conducted.

Materials and methods

Study sites
Samples of water were taken from Kamafusa Dam and 3 rivers, the Taro River, Kita River and Mae River. Kamafusa Dam is a main drinking water resource in Sendai, Japan and the 3 rivers ow into the dam. Water from Kamafusa Dam treated at the water treatment plant was also sampled. Sampling was carried out once bimonthly for a period of 14 months. AOC determination AOC determination is based on the measurement of the maximum extent of growth of a selected pure bacterial culture in representative water samples, in which indigenous bacteria have been killed or inactivated by heat treatment. 1516 Colony counts are used for determining bacterial densities. The analytical method developed by van der Kooij was adopted in the present study. The method employs Pseudomonas uorescens strain P17 and Spirillum strain NOX. These 2 microorganisms prefer dierent groups of compounds. P17 has a great 246

Assimilable organic carbon generation from algogenic..., J. H. KIM, et al.,

nutritional versatility and may grow on a variety of carbohydrates, aromatic acids, and amino acids. It can also grow on carboxylic acids, with the exception of formic, glyoxylic, and oxalic acids as typical byproducts of ozonation. Therefore, the growth of strain P17 was used for the determination of the concentration of aromatic and amino acids, etc. Meanwhile, NOX cannot utilize carbohydrates, alcohols, or aromatic acids, but rather a wide range of carboxylic acids. NOX can also utilize a few amino acids; however, this organism cannot assimilate amino acids when growing on mixtures of compounds. Therefore, the growth of strain NOX was used for the determination of the concentration of carboxylic acids in water. All glassware was thoroughly cleaned and rendered free of organic carbon by combustion at 550 C for 1 h. Samples were incubated at 15 0.5 C and the number of colonies was measured with the plate count technique after 72 3 h of incubation at 28 1 C on R2A agar. The total AOC from microorganisms per liter of equivalent acetate-carbon is the sum of AOC-P17 and AOC-NOX. In the present study, the yield coecients of P17 and NOX were measured as 4.1 10 7 g acetate-C/L and 1.2 10 7 g acetate-C/L, respectively. Algal culture The blue-green algae Phormidium tenue (P. tenue, ATCC 49642) was selected as test algae to form AOM. It was cultivated under an illumination condition of 3000 lx with a 12:12 photoperiod in an incubator. The temperature was maintained at 22 C P. tenue was cultured with 2 dierent types of culture media, M-11 and CT. Table 1 shows the components of each culture medium.
Table 1. Components of culture media.

M-11 culture medium NaNO3 K2 HPO4 MgSO4 7H2 O CaCl2 2H2 O Na2 CO3 FeSO4 7H2 O Na2 EDTA2H2 O Distilled water pH 10 mg 1.0 mg 7.5 mg 4.0 mg 3.0 mg 0.1mg 0.1 mg 100 mL 8.0 KNO3

CT culture medium Ca(NO3 )2 4H2 O Na2 glycerophosphate5H2 O MgSO4 7H2 O Vitamin B12 Biotin Thiamine HCl PIV metals TAPS Distilled water pH 15 mg 10 mg 5.0 mg 4.0 mg 0.01 g 0.01 g 1.0 g 0.3 mL 40 mg 99.7 mL 8.2

Chlorine oxidation Sodium hypochlorite (NaOCl) solution was standardized by titration and then diluted with distilled water. Chlorine doses were set at 0, 2, 5, 10, 20, and 30 mg/L. Sealed amber glass bottles, each 3 L, were used for the chlorine oxidation experiment. The amber glass bottles were also covered with aluminum foil to prevent photodegradation of the solutions. The contents of the bottles were stirred continuously with a magnetic 247

Assimilable organic carbon generation from algogenic..., J. H. KIM, et al.,

stirrer. After chlorine oxidation, the sample was withdrawn from each bottle, then ltered with a GF/C lter and terminated by addition of sodium thiosulfate immediately after collection. Ultraltration The sample was initially ltered through Whatman GF/F lters and then submitted to ultraltration (regenerated cellulose PL, ultraltration membranes, Millipore). The sample was fractionated into 4 fractions by molecular weight: lower than 1000 Da, 10005000 Da, 500010,000 Da and higher than 10,000 Da.

Results and discussion

Seasonal variations of AOC in dam and river water
Concentrations of AOC at each sampling point throughout the year were conducted (Table 2). At Kamafusa Dam, the concentration ranged from 111 to 305 g/L. For the 3 rivers owing into the dam, it ranged from 34 to 293 g/L. The average AOC concentration of Kamafusa Dam, 178 g/L, was approximately 1.6 times higher than that of the rivers. This indicated that the AOC concentration of Kamafusa Dam tended to be higher than that of each river (t-test, P < 0.01). Treated water had an average AOC concentration of 107 g/L. Unexpectedly, the AOC concentration of the treated water on 28 April 2007 was higher than that of Kamafusa Dam. It was thought that the sand in the rapid sand ltration unit was contaminated by organic substances. Except for the sample taken on 28 April 2007, the AOC concentration of the treated water was lower than that of Kamafusa Dam averaging about 50 g/L. A maximum AOC concentration was observed in the spring season in March and April, at all survey points. This might be due to the increased growth of aquatic plants in spring.
Table 2. Concentration of AOC at various sampling points throughout the year.

Sampling period 25.04.2006 14.06.2006 29.07.2006 10.10.2006 16.01.2007 13.03.2007 28.04.2007 11.06.2007 Average

AOC (g/L) Taro River Kita River 63 67 103 54 39 293 101 165 103 109 75 89 115 173 46 240 65 121

Mae River 80 44 94 34 219 226 171 50 124

Kamafusa Dam 193 174 133 111 142 305 188 167 178

Treated water 31 53 54 44 48 192 325 64 107

Figure 1 shows the average ratio of AOC-P17 and AOC-NOX at each survey point. In the case of untreated water from the rivers and dam, the AOC-P17 component accounted for 82%-93% of the total AOC 248

Assimilable organic carbon generation from algogenic..., J. H. KIM, et al.,

concentration, while in the treated water, the AOC-P17 component was 57%. It was obvious that the increase of the AOC-NOX component was due to the production of AOC-NOX by chlorine oxidation.

Relationship between AOC and algal cell number

In order to develop a relationship between increased concentrations of AOC and the algal cell numbers at Kamafusa Dam, the seasonal variations in algal cell numbers at the dam were studied during 2002-2006 (Figure 2). Data revealed that during the spring season (FebruaryApril), algal cell numbers were higher than in the rest of the months of the year. This tendency was similar with that of increased concentrations of AOC at Kamafusa Dam in the spring (Table 2); in other words, the algal cell numbers would increase in spring and decrease in autumn. The average cell number of about 4700 cells/mL in spring was twice that of other seasons. The results suggested that an increase in algal growth could cause an increase in the AOC concentration of the water.
8000
2002 2003 2004 2005 2006

7000

Ratio AOC-P17/AOC-NOX

P17

80 60 40 20 0 Taro Rive Kita River Mae River Kamafusa Dam Treatment Water

Cell numbers (cells/mL)

100

NOX

6000 5000 4000 3000 2000 1000 0 Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. Jan. Feb. Mar.

Figure 1.

Ratio of AOC-P17/AOC-NOX at dierent

Figure 2. Seasonal variations in algal cell numbers at Kamafusa Dam.

survey points.

Eects of algal growth on AOC

P. tenue, one of the dominant species at Kamafusa Dam was selected and cultured in M-11 and CT culture media. Variations in P. tenue cell numbers in each culture medium are shown in Figure 3. The initial cell numbers of P. tenue in M-11 and CT culture media were 0.8 10 4 and 1 10 4 cells/mL, respectively. The cell number increased rapidly in each culture medium, and on the 39th day, the cell numbers of the M-11 and CT media had increased to 67 10 4 and 115 10 4 cells/mL, respectively. On the 68th day, the cell number in each culture medium was 90 10 4 and 102 10 4 cells/mL, respectively. Figure 4 shows variations in the DOC and AOC concentrations produced by P. tenue in M-11 culture medium. The DOC concentration of M-11 culture medium before inoculation of P. tenue was about 1.5 mg/L. The DOC concentration had a tendency to increase with elapsed time. In the stationary phase, after the 39th day, the DOC concentration had increased to 9.2 mg/L. The AOC concentration of M-11 culture medium before inoculation of P. tenue was 39 g/L. The AOC concentration increased signicantly in the latter half of the 249

Assimilable organic carbon generation from algogenic..., J. H. KIM, et al.,

logarithmic phase and reached about 4000 g/L during the stationary phase. The AOC-P17 component was about 90% or more. In other words, variations in the AOC concentration depended on the AOC-P17 component.
1.E+07

6000
Cell numbers (cells/mL) 1.E+06
P17 NOX DOC

12 10 DOC (mg/L) 8 6

1.E+05

AOC (g/L)

4000

1.E+04 M-11(P.tenue) CT(P.tenue) 1.E+03 0 20 40 60 80

2000

4 2

0 0 12 25 39 Elapsed time (days) 58

Elapsed time (days)

Figure 3. Prole of P. tenue cell numbers in dierent culture media with respect to time.

Figure 4.

Prole of DOC and AOC concentrations

produced by P. tenue in M-11 culture medium.

Figure 5 shows variations in DOC and AOC concentrations produced by P. tenue in CT culture medium. The DOC concentration of the CT culture medium was much higher than that of the M-11 culture medium. DOC concentration, which did not change throughout the culture period, averaged 152 mg/L. In contrast to M-11 culture medium, CT culture medium did not lead to a signicant variation in AOC concentration. The maximum AOC concentration during the culture period was about 80 g/L. These results suggest that the generation of AOC from algae was greatly inuenced by culture medium conditions. It was concluded that the generation of AOC occurred actively, when the concentration of organic substances in algal water was relatively low. To investigate the prole of AOC and DOC concentrations derived from EOM in the stationary phase, EOM was fractionated by ultraltration into 4 fractions of molecular weight (Figure 6). The DOC produced by EOM was detected mainly at molecular weights under 1000 Da. The percentage of DOC concentrations under 1000 Da was over 60% of the total DOC concentration. As shown in Figure 7, the AOC produced by P tenue was also detected mainly in the molecular weight fraction lower than 1000 Da. Although AOC-NOX occupied 20% of the total AOC at this fraction, AOC-NOX was not detected in the fraction for which molecular weight was over 10,000 Da.

Eects of chlorination on AOC included in AOM

Figure 8 shows the relationship between chlorine dose and residual chlorine for EOM and IOM. The demand of chlorine for the oxidation of both EOM and IOM (P. tenue, M-11 culture medium) was about 20 mg/L for the sample containing 67 10 4 cells/mL. In the case of fresh culture medium, chlorine was not consumed. Figures 9a and 9b show the variations of DOC and UV absorbance at 260 nm (E260) for the oxidation of EOM and IOM by chlorine. In the case of fresh culture medium, the DOC concentration and E260 were not signicantly changed with the increase of chlorine doses. However, the E260 of EOM had a tendency to decrease with the 250

Assimilable organic carbon generation from algogenic..., J. H. KIM, et al.,

increase of chlorine doses. In the case of IOM, the DOC concentration and E260 increased with the increase of chlorine doses. This increase in DOC and E260 by chlorination was due to the release of dissolved AOM through the destruction of algal cells. Although E260 increased due to chlorine oxidation, the maximum value of IOM was not higher than that of EOM.
100 P17 75 AOC (g/L) NOX DOC 50 80 120 DOC (mg/L)
1000 > Molecular weight

160

5000 - 1000

25

40

10,000 - 5000

10,000 <

0 0 12 25 39 58 Elapsed time (days)

0
0.0 1.0 2.0 3.0 DOC (mg/L) 4.0 5.0 6.0

Figure 5. to time.

Prole of AOC and DOC concentrations

Figure 6.

Fraction of DOC produced by EOM (M-11

produced by P. tenue in CT culture medium with respect

culture medium).

35 28 21 14 7 0
0 700 1400 2100 AOC (g/L) 2800 3500

1000 > Molecular weight

5000 - 1000

10,000 - 5000 NOX 10,000 < P17

Residual chlorine (mg/L)

EOM IOM

10

20 Chlorine dose (mg/L)

30

40

Figure 7. Fraction of AOC produced by EOM.

Figure 8. Relationship between chlorine dose and residual chlorine.

Figure 10 shows variations in AOC concentration during chlorination of EOM. Although AOC concentration increased 18% with a low chlorine dose (2 mg/L), it decreased with the increase of the chlorine dose to over 2 mg/L. A chlorine dose of 5 mg/L resulted in a decrease in AOC concentration, from 636 to 336 g/L. In the case of 10 and 30 mg/L chlorine doses, the AOC concentration further decreased to 219 and 98 g/L, respectively. The percentage of AOC-P17 and AOC-NOX for EOM before chlorine oxidation was about 50%. With the increase of chlorine dose, the AOC-P17 component decreased signicantly and the ratio of the 251

Assimilable organic carbon generation from algogenic..., J. H. KIM, et al.,

AOC-NOX component increased. When residual chlorine was detected at a chlorine dose of 30 mg/L, an AOC concentration of 98 g/L still remained. This value of chlorine seems insucient to prevent bacterial regrowth. To treat the AOC included in EOM, a greater chlorine dose than usual was required.
50 40
DOC (mg/l) 0.25

EOM IOM
E260 (1/cm)

0.20 0.15 0.10 0.05 0.00

EOM IOM

30 20 10 0 0 7 14 21 Chlorine dose (mg/L) (a) 28 35

14 21 Chlorine dose (mg/L)

(b)

28

35

Figure 9. At each chlorine dose, prole of a) DOC and b) E260.

Furthermore, in order to study the eect of chlorination on AOC for IOM, chlorine oxidation of IOM was performed. IOM was collected by centrifugation of algae cells. The results are interpreted in Figure 11. When the chlorine dose was 0 mg/L, the AOC concentration in IOM was about 2 g/L. However, AOC concentrations were increased with the increase of chlorine doses. For example, the AOC concentration increased to 92 g/L at a chlorine dose of 2 mg/L. In the case of 20 and 30 mg/L of chlorine doses, however, AOC was not detected. The ratio of AOC-P17 to AOC-NOX for IOM diered from that of EOM. AOC-P17 was higher than AOC-NOX, at above 74% for each chlorine oxidation. From this, it was found that the AOC included in IOM could be removed suciently when chlorination was carried out as usual.
200 P17 NOX

1000 AOC (g/L) 800 600 400

AOC (g/L)

150

P17 NOX

100

50

200 0 0 2 5 10 Chlorine dose (mg/L) 30

0 0 2 5 10 Chlorine dose (mg/L) 20 30

Figure 10. AOC concentration after chlorination (EOM).

Figure 11. AOC concentration after chlorination (IOM).

252

Assimilable organic carbon generation from algogenic..., J. H. KIM, et al.,

Conclusions
Higher concentrations of AOC at Kamafusa Dam, as compared with those of the rivers, indicated that eutrophication could cause changes in the concentration of AOC. Algogenic materials seemed to be an important factor in AOC variation. The AOC included in EOM could be formed largely by algae grown in fresh culture media. Generation of AOC occurred actively when the concentration of organic substances in algal material was relatively low. EOM with molecular weights lower than 1000 Da produced by P. tenue would help the formation of AOC. Chlorination increased DOC and E260 absorbance due to the release of dissolved AOM. Chlorination of EOM and IOM revealed that the AOC included in EOM was dicult to remove; however, the AOC included in IOM could be removed under conditions where residual chlorine was detected.

Acknowledgement
This work was nancially supported by the Korean Ministry of Environment (MOE) under the ET-Human Resource Development Project.

References
1. LeChevallier, M. W. NSF International/World Health Organization Symposium on HPC Bacteria in Drinking Water, 2002, 405-416. 2. Volk, C.; Dundore, E.; Schiermann, J.; LeChevallier, M. Wat. Res. 2000, 34, 1967-1974. 3. Tsai, Y. P.; Pai, T. Y.; Qiu, J. M. Journal of Biotechnology 2004, 111, 155-167. 4. Hu, J. Y.; Ng, W. J. Wat. Res. 2002, 36, 891-898. 5. van der Kooij, D. JAWWA 1992, 84, 5765. 6. Escobar I. C.; Randall A. A. Wat. Res. 2001, 35, 4444-4454. 7. van der Kooij, D; Veenendaal, H. R.; Baars-Lorist, C; van der Klift, D. W.; Drost, Y. C. Wat. Res. 1995, 29, 1655-1662 8. Lehtola, M. J.; Miettinen, I. T.; Vartianen, T.; Martikainen, P. J. Wat. Res. 2002, 36, 3681-3690. 9. Schmidt, W.; Hambsch, B.; Petzoldt, H. Wat. Sci. Tech. 1998, 37, 91-96. 10. Plummer, J. D.; Edzwald, J. K. Wat. Sci. Tech. 1998, 37, 49-55. 11. Hem, L. J.; Efraimsen, H. Wat. Res. 2001, 35, 1106-1110. 12. Charnock, C.; Kjonno, O. Wat. Res. 2000, 34, 2629-2642 13. Escobar, I. C.; Randall, A. A. Wat. Res. 2000, 34, 1680-1686. 14. Kim, H. C.; Yu, M. J. Wat. Res. 2005, 39, 4779-4784. 15. Liu, W.; Wu, Z.; Wang, S. L.; Ong, J. Y.; Ng, W. J. Wat. Res. 2002, 36, 891-894. 16. Weidong, Z.; Francis, A. D. Wat. Res. 2002, 36, 1469-1474.

253