DSpace at VNU: Trihalomethane formation by chlorination of ammonium- and bromide-containing groundwater in water supplies of Hanoi, Vietnam

Under typical treatment conditions total chlorine residual 0.5– 0.8 mg/L, the total THM formation was always below WHO, EU, and USEPA drinking water standards and decreased in the order

Trihalomethane formation by chlorination

of ammonium- and bromide-containing groundwater in water

supplies of Hanoi, Vietnam

a Centre for Environmental Technology and Sustainable Development (CETASD), Hanoi University of Science, Nguyen Trai Street 334,

Hanoi, Viet Nam

b EAWAG, Swiss Federal Institute for Environmental Science and Technology, Ueberlandstrasse 133, CH-8600 D ubendorf, Switzerland

Received 16 October 2002; received in revised form 10 February 2003; accepted 12 February 2003

Abstract

The occurrence and the fate of trihalomethanes (THMs) in the water supply system of Hanoi City, Vietnam was investigated from 1998 to 2001 The chlorination efficiency, THM speciation, and, THM formation potential (THMFP) was determined in the water works and in tap water With regard to THM formation, three types of groundwater resources were identified: (I) high bromide, (II) low bromide, and (III) high bromide combined with high ammonia and high dissolved organic carbon (DOC) concentrations Under typical treatment conditions (total chlorine residual 0.5– 0.8 mg/L), the total THM formation was always below WHO, EU, and USEPA drinking water standards and decreased in the order type I>type II>type III, although the THMFP was >400 mg/L for type III water The

type II water, the bromo-THMs still accounted for some 40% although the bromide concentration is significantly lower

bromide Based on chlorine exposure (CT) estimations, it was concluded that the current chlorination practice for type I and II waters is sufficient for X2-log inactivation of Giardia lamblia cysts However, in type III water the applied chlorine is masked as chloramine with a much lower disinfection efficiency In addition to high levels of ammonia, type III groundwater is also contaminated by arsenic that is not satisfactory removed during treatment N-nitrosodimethylamine, a potential carcinogen suspected to be formed during chloramination processes, was below the detection limit of 0.02 mg/L in type III water

Keywords: Disinfection by-products; Water distribution system; Trihalomethane formation potential; Competition kinetics; Chlorine exposure; N-nitrosodimethylamine (NDMA); Hanoi; Vietnam; Bromide; Ammonium

1 Introduction

Hanoi has a strongly increasing water demand due to

the rapid growth of the urban population (3.5 Mio

resources for Hanoi are reduced groundwaters contain-ing variable levels of dissolved iron(II) and

respectively Moreover, Hanoi’s groundwater contains excessive concentrations of arsenic that are partly removed in the WTPs to lower but not fully acceptable

*Corresponding author Tel.: 823-5078; fax:

+41-1-823-5058.

E-mail address: michael.berg@eawag.ch (M Berg).

0043-1354/03/$ - see front matter r 2003 Elsevier Science Ltd All rights reserved.

doi:10.1016/S0043-1354(03)00138-6

several locations have significantly differing chemical

compositions that are influenced by varying proportions

Since the beginning of the 20th century chlorination

has been a key-treatment for improving the

microbio-logical safety in drinking waters However, an undesired

formation of disinfection by-products (DBPs) results

from the reaction of chlorine with natural organic

matter (NOM) and includes products such as

The total concentration of trihalomethanes (THMs)

and the formation of individual THM species in

chlorinated water strongly depend on the composition

water treatment and on the residual chlorine in the

cancer risk, the United States Environmental Protection

Agency (USEPA), the World Health Organization

(WHO), and the European Union (EU) introduced

regulations for THMs in drinking water Whereas the

USEPA and the EU regulate total THM concentrations

as 80 or 100 mg/L, respectively, the WHO provides

In Vietnam, chlorination is generally applied for

drinking water disinfection The requirement for the

minimum residual chlorine concentration in water at the

outlet of water treatment plants (WTPs) and in the

So far, no information on THM concentrations in drinking water was available To determine THM levels

in the urban area of Hanoi, the eight major WTPs of this

Based on the differences in dissolved organic carbon (DOC), UV254 absorption, bromide concentration and chlorine demand (mainly ammonium and ferrous iron content), it was inferred that substantial variations in THM formation can be expected in the eight Hanoi WTPs A direct comparison of the eight WTPs is favored

by the fact that very similar treatment trains are applied

In addition to full-scale data, chlorination experi-ments with natural groundwater were carried out in the laboratory to evaluate the maximum THM formation potential (THMFP) and THM yields for various chlorine doses Based on these data, a relation between chlorine exposure and THM formation was established These results may be used as a tool to adjust the chlorine doses, which is needed to achieve efficient disinfection and acceptable levels of THMs in the drinking waters of Hanoi

2 Materials and methods 2.1 Reagents

If not specified, chemicals were reagent grade from Fluka (Buchs, Switzerland) or Merck (Darmstadt,

Fig 1 Map illustrating northern Vietnam and the locations of the eight studied Hanoi WTPs The areas of differing groundwater quality (types I, II and III) are indicated The numbers 1–8 refer to the following WTPs: 1, Mai Dich; 2, Ngoc Ha; 3, Ngo Si Lien; 4, Yen Phu; 5, Luong Yen; 6, Ha Dinh; 7, Tuong Mai; and 8, Phap Van.

Germany) A standard stock solution containing the

four THMs (1 mg/mL of each in methanol) and the

internal standard p-bromofluorobenzene (1 mg/mL in

methanol) was supplied by Tokyo Kasei Kogyo

(Chuo-Ku, Tokyo, Japan) Working standard solutions of

THMs (1, 10, and 100 mg/L) were prepared from the

stock solution Calibrations were conducted with five

concentration levels in de-ionized and boiled water

Chlorine solutions of 0.4% and 0.04% were prepared

from NaOCl 4% (Aldrich, Steinheim, Germany) by

dilution with water

2.2 Water treatment plants in Hanoi

Groundwater from 30 to 70 m depth is used as raw

water All WTPs operate with similar treatment trains

including aeration, settling, sand filtration, chlorine

disinfection, and storage in a reservoir Chlorine is

applied as gas and the applied disinfectant dose is

adjusted manually to maintain a total chlorine residual

of 0.5–0.8 mg/L after the reservoir

2.3 Sampling campaigns

2.3.1 Water treatment plants

treated water before disinfection (after Fe removal), and

finished water at the outlet of the reservoir were

collected Nine sampling campaigns were conducted

between April 1998 and May 2001

2.3.2 Distribution system Two sampling campaigns (May and September 2001) were performed in three water distribution systems to investigate THMs in household tap waters These water distribution systems primarily provide drinking water from WTPs 3, 4 and 6 representing the three different groundwater types of Hanoi Finished water of the WTPs and tap water samples of the corresponding distribution system were collected at the same day 2.4 Analytical methods

Water temperature, pH, dissolved oxygen (detection limit 0.1 mg/L) (Paqualab instrument ELLE, Leighton Buzzard, UK) and residual chlorine (detection limit 0.01 mg/L) were measured at the sample collection sites Samples for DOC, cations, anions and coliforms were

concentrations were measured by a Shimadzu TOC– 5000A analyzer Ammonium and bromide were deter-mined by ion chromatography with conductivity detec-tion (HPLC 10A, Shimadzu, Japan) UV absorbance was measured with a UV/Vis spectrophotometer (Shi-madzu 1201, Japan) at 254 nm in 10 mm quartz cells Total coliforms in water were determined by the

chloramine concentrations were determined by the

2.5 Analysis of THMs Duplicate samples for THM analysis were collected in 44-mL glass vials and were capped with PTFE-faced

Table 1

Individual THM concentrations and chemical parameters in treated and finished waters of Hanoi water treatment plants

treated waterc, concentration range (classification)

finished water d , concentration range (average) e

a The areas of differing groundwater quality are indicated in Figure 1

b The numbers 1–8 of the water treatment plants refer to Figure 1

c Treated waters were collected before chlorination Number of samples: 9–15.

d Finished waters collected at the outlets of the WTPs Number of samples: 10–17, chlorine dose o1.5 mg/L, contact time 30 min.

e Average concentration from finished water samples collected between April 1998 and May 2001.

silica septa Household taps were flushed for 15 min

prior to sampling The 44-mL sample vials contained

solution) to quench residual chlorine The samples were

Head-space gas chromatography with mass

spectro-metry detection was used to analyze THMs in water

Sample volumes of 10 mL were filled in 20 mL glass vials

and spiked with 50 mL of the internal standard (10 mg/L

p-bromofluorobenzene) The vials were immediately

sealed with Teflon coated septa and aluminum

1 mL of the head-space was injected splitless into a gas

0.32 mm film, J&W Scientific, CA) coupled to a mass

spectrometer (GC/MS QP5000 Shimadzu, Japan) The

analytes were recorded in the selective ion monitoring

(SIM) mode and quantified by internal standard

calibration Recoveries of the four THMs, determined

in spiked samples at levels of 1, 10 and 20 mg/L, were in

the range of 94–114% (n ¼ 7) The method detection

limits (MDLs) were determined from the standard

deviation of 1-mg/L spiked samples (n ¼ 7; 3 sigma)

N-nitrosodimethylamine (NDMA) was analyzed with gas

chromatography and thermal energy analysis detection

(GC-TEA, detection limit 0.02 mg/L), following the

official method 982.12 of the Association of Analytical

Communities (AOAC, Gaithersburg, MD)

2.6 Laboratory chlorination experiments

Laboratory chlorination experiments were carried out

with treated waters collected before disinfection from

the WTPs 1–4 and 6–8 Samples were collected in 5-L

experiments The water samples were buffered with

1 mM phosphate and adjusted to the desired pH with

NaOH For chlorination experiments, the required

chlorine dose was added to a laboratory batch system

and the solution was well mixed for 30 s Then the water

was immediately portioned into 44-mL glass vials that

were sealed with Teflon-lined screw caps After filling

and sealing the head-space-free samples, the vials were

desired reaction time, a sample was removed, and,

residual total chlorine and chloramine were determined

above

The THMFP of the treated waters was determined as

total THM formed during a reaction time of 7 days at

residual throughout the experiment A chlorine dose of

9 mg/L was used for treated water from WTPs 2 and 4, and, of 160 mg/L for treated water from WTP 8, respectively

3 Results and discussions 3.1 Water quality parameters and THM formation (water types I, II, and III)

Table 1summarizes the water quality parameters of

the THM formation during chlorination No THMs were detectable before chlorination

3.1.1 DOC and UV254 Water of types I and II had lowlevels of DOC

these waters can be expected to be low In water type III, DOC and UV254 were relatively high The highest values were measured in treated water of WTP 8 (DOC

of NOM and a high potential for THM formation 3.1.2 Bromide

Relatively high levels of bromide in the range of 50–

140 and 70–240 mg/L were observed in water types I and

near the detection limit of 20 mg/L Bromide is important for THM formation because it is oxidized by chlorine to hypobromous acid, which contributes to the formation

3.1.3 Ammonium The fast reaction of chlorine with ammonia leads to masking of chlorine in excess of ammonia by formation

of chloramine If chloramine is the dominant species, the THM yields in chlorinated water is reduced because of

a situation was observed for water type III (WTPs 6, 7 and 8) where extremely high ammonium concentrations

of 5–25 mg/L are present Ammonium is believed to originate from mineralization of peat which is abundant

3.2 Occurrence of THM in finished waters at the WTP outlets

Fig 2shows the concentrations of THMs in finished waters at the outlet (reservoir, 1 h contact time) of WTPs using groundwater types I, II and III They are in the range of 5–56, 2–18 and 0.3–22 mg/L, respectively The high variability of the total THM concentrations can be

attributed to unsteady chlorine doses, variations in

ammonium, DOC and Fe(II) concentrations The

groundwater composition is not constant because the

10–20 wells belonging to each Hanoi WTP are operated

intermittently The THM formation in finished water

was generally in the following order of water sources:

type I>type II>type III Due to the relatively lowDOC

levels in water types I and II, the absolute levels of

THMs are quite low Higher levels of THMs would be

expected in water type III (high DOC) However,

ammonium is dominating this system (see below)

The distribution of the four THMs in finished waters

type I), brominated THMs (>85%) were the dominant

and most abundant species In finished water from WTP

4 and 5 (water type II), chloroform was formed in

similar concentrations as the brominated THMs In

comparison to water type II, the higher ratio of

brominated THMs in water type I can be explained by

highest bromide levels were present in groundwater type

III, yet only small amounts of brominated THMs were

detected in the treated waters (chloroform 90% of the

total THM) This is due to masking of chlorine by

ammonia which prevents bromide oxidation (see below)

3.3 Influence of ammonium on THM formation and

speciation

The THMFP determined for the three waters type,

namely WTP 2 (type I), WTP 4 (type II) and WTP 8

(type III), were 103, 59 and 406 mg/L, respectively (see

Table 2) The high THMFP found for water type III

corresponds to its high content of NOM represented by

inFig 2, the total THM concentrations in the finished

water type III are 20–200 times lower than the THMFP This can be explained by the formation of THM along

only traces of THMs are formed for chlorine doses in the far pre-peakpoint region, whereas a noticeable increase of THMs is observed near the peakpoint Between the peakpoint and the breakpoint, chloroform formation sharply increases with increasing chlorine dose

A comparison of breakpoint chlorination curves is

breakpoint chlorination is applied in WTPs of types I

that the high ammonium concentration (5–25 mg/L) present in the investigated waters of WTP 6–8 consume 40–130 mg/L chlorine For water type III it can therefore

0

10

20

30

40

50

60

WTP

groundwater type

maximum average minimum

(n = 4 to 8)

Fig 2 Average concentration and range of total THM

concentrations measured in finished waters at the outlets of

the Hanoi WTPs.

20 40 60 80 100

CHBr3 CHClBr2 CHCl2Br CHCl3

I groundwater type

0

Fig 3 Molar distribution of THM species in finished waters of the Hanoi water types I, II and III.

Table 2 Trihalomethane formation potential (THMFP) for treated Hanoi groundwaters

WTP 2 WTP 4 WTP 8 Type Ia Type IIa Type IIIa Treated waterb

Br

Laboratory conditions

a The areas of differing groundwater quality are indicated in

Fig 1 b Treated waters were collected before chlorination.

be inferred that the far pre-breakpoint chlorination

leads to the formation of monochloramine, and

conse-quently, lowTHM concentrations are present in the

chlorine dose (0.8–1.5 mg/L) applied is critical for WTP

4 (only slightly above breakpoint) and insufficient for

water type III

The THM formation in the investigated waters is

controlled by the following competition kinetics:

The kinetics of reactions (1) and (5) are not known in absolute terms, however, reaction (5) is much faster than reaction (1) Therefore, it is important to knowto which extent bromide is oxidized to assess the potential for the formation of bromo-THMs Because ammonia is the

pre-break-point region, the extent of bromide oxidation can be estimated from a competition kinetics calculation involving reactions (2)–(4) The fraction of bromide being oxidized during the phase where ammonia is in excess is in the range of a fewpercent However, near the peakpoint or from the peakpoint to the breakpoint, the fraction of bromide being oxidized becomes larger This

is due to the slower kinetics of reaction (3) as compared

can be calculated as follows:

pre-peakpoint:

0.0

0.1

0.2

0.3

0.4

0.5

0 5 10 15 20

CHBr3 CHCl3

residual chlorine

0.0

0.1

0.2

0.3

0.4

0.5

0 5 10 15 20 25 30

CHBr3 CHClBr2 CHCl2Br

residual chlorine

WTP 1

WTP 4

total THM as CHCl3

total THM as CHCl3

chlorine dose (mg(L)

chlorine dose (mg(L)

CHCl3 (a)

(b)

Fig 4 Breakpoint curves and formation of THMs (contact

time 24 h, pH 8.1) Laboratory experiments with treated waters

collected before chlorination in the Hanoi WTPs Mai Dich

(WTP 1, type I) and Yen Phu (WTP 4, type II).

10 20 30 40 50 60 70

0.2 0.4 0.6 0.8 1.0 1.2 1.4

chlorine dose (mg/L)

type I and II

breakpoint WTP 1

WTP 1

0.03 mg/L NH4

WTP 4

0.13 mg/L NH4

range of applied chlorine dose

breakpoint WTP 4

chlorine dose (mg/L)

applied chlorine dose ≤ 1.5 mg/L

type III

WTP 8

18 mg/L NH4

WTP 7

7 mg/L NH4

WTP 6

6 mg/L NH4 0

0

(a)

(b)

Fig 5 Breakpoint curves derived from contact times of 24 h at

pH 8.0 for: (a) WTP 1 and 4 and (b) WTP 6–8 The range of the chlorine dose applied in the WTPs is indicated.

peak- to breakpoint:

If the residual concentration of HOCl is known

can be estimated as

Because reaction (3) is a relatively slowprocess, HOCl

at the peakpoint can be determined as

is the initial ammonium concentration

Based on these considerations, estimates of the

bromide oxidation to HOBr were made for water types

showthat in the case of water type I up to 80% of the

bromide can be oxidized to HOBr which then further

reacts with NOM according to reaction (5) This

explains the high fraction of bromo-THMs found in

quality, however, the bromide levels are considerably

lower which explains the lower formation of

very high (up to 240 mg/L)

3.4 Estimation of the disinfection efficiency via the THM

formation

Fig 6 shows the relationship between THM

various chlorine doses derived from laboratory

experi-ments with water types I and II The chlorine exposure

(CT) is calculated as the integral under a chlorine

used to estimate the inactivation efficiency of

a 2-log inactivation of E coli, a 3-log inactivation of viruses and a 2-log inactivation of Giardia lamblia cysts can be achieved while the total THM concentration remains below10 mg/L which is significantly lower than the typical drinking water standard To reach the same germ inactivation in waters of type III where active chlorine is mainly present as chloramines, CT values of

360 and 500 mg/L min are required for virus and G

significantly higher chlorine doses of 10–30 mg/L would

be necessary for the WTPs 6, 7 and 8 However, such high chlorine doses might lead to substantially higher

the formation of NDMA (see below)

3.5 Chlorine residual concentrations in tap water samples

of the distribution systems Results from a sampling campaign in three

taken at increasing distance from the WTP Using the distance from the WTP to the sampling point and the guideline limits for the flowrate in the distribution

the average residence times in the distribution system were estimated to be less than 1 h for all tap water sampling points

Active residual chlorine concentrations generally decreased with increasing distance from the WTPs The active chlorine residual in the distribution systems

of WTPs 3 and 4 was free chlorine It was maintained at concentrations of 0.5–1.4 mg/L However, in the dis-tribution system of WTP 6 (groundwater type III), the disinfectant is chloramine and its concentration did not meet the Vietnamese requirement for residual disin-fectant concentrations (X0.3 mg/L) The concentration

of chloramine residual decreased quickly after a distance

of 500 m from WTP 6 The fast consumption of

Table 3

Estimated formation of HOBr during chlorination of pretreated raw water from groundwater resources in Hanoi

chloramine near WTP 6 may be due to the oxidation of

remaining Fe(II) and Mn(II) or a biological reduction of

3.6 Trihalomethane occurrence in tap water

analyzed in the tap water samples were below the

drinking water standards of the EU and the USEPA

The speciation of THMs in the finished waters and in the

tap waters was the same In the distribution systems of

WTPs 3 and 4 having relatively high free chlorine

residuals and dissolved oxygen of more than 2 mg/L, the

total THM concentrations increased with increasing

distance from the plant No coliforms were present in

the tap waters investigated Based on the estimates on

indicate CT values which possibly guarantee a good

inactivation, even for G lamblia cysts Interestingly, the

THM concentrations decreased in the distribution system of WTP 6 after 540 m distance There are several possible hypothesis for this observation: (i) leaking of pipelines and dilution, (ii) volatilization of THMs, and (iii) reductive degradation of THMs Hypothesis (i) is not very likely because the decrease of THMs and oxygen are different A massive volatilization of THMs would probably need a longer residence time and smaller pipeline diameters Therefore, it is most likely that a biotic or abiotic degradation of THMs occurs under low oxygen conditions

3.7 N-nitrosodimethylamine formation Samples from WTP 6 and its distribution system were also checked for a possible formation of NDMA which

is a probable human carcinogen It has been shown that NDMA is formed during chloramination processes due

to the reaction of monochloramine with dimethylamine

detection limit of 0.02 mg/L in the finished water as well

as in the five tap water samples analyzed For the

WTPs, the NDMA formation potential can therefore be considered negligible Should chlorine doses be in-creased, the risk of NDMA formation has to be considered again

4 Conclusions and recommendations This study shows that the THM concentrations in all Hanoi WTPs were below WHO, EU and USEPA drinking water standards For water type III with high DOC and bromide levels and a THMFP of >400 mg/L, higher total THM concentrations as well as a higher proportion of bromo-THMs were expected The low THM formation could be explained by competition kinetics of chlorine for the oxidation of ammonia and

in the Hanoi distribution system The following conclu-sions and recommendations can be drawn from the findings obtained through this study

1 Based on estimations of chlorine exposure (CT), it can be hypothesized that for the inactivation of coliforms, virus, and Giardia lamblia cysts, the chlorine dose of 0.8–1.5 mg/L applied for disinfection

by the Hanoi water works is inefficient for water type III, and critical for water types I and II if ammonium concentrations exceed 0.1 mg/L

2 With regard to the residual chlorine of 0.3 mg/L required in the distribution system and tap water,

it is recommended to maintain a steady chlorine dose of 1.5 mg/L for water types I and II if ammonium concentrations are below0.15 mg/L

2

4

6

8

10

12

chlorine exposure (CT), (mg/L.min)

0

5

10

15

20

25

30

chlorine exposure (CT), (mg/L.min)

type I

type II 0

Giardia lamblia cysts (CT = 25)

Virus (CT = 1), E.coli (CT < 0.05)

25

(a)

(b)

Fig 6 THM formation during chlorination of treated waters

at 25C and pH 7.5 for various chlorine doses (a) Water type I,

chlorine dose 1.1 mg/L (diamonds), 1.9 mg/L (squares); (b)

water type II, chlorine dose 0.5 mg/L (diamonds), 1.0 mg/L

(squares), 1.9 mg/L (triangles) Required CT values for E coli,

virus, and Giardia lamlia cysts are indicated.

Water treatm

fr WT

Estimated residence time

T  ( C)

O2

Coliform (MPN/ 100

a Belowdetectio

Lower chlorine doses and/or ammonium

concentra-tion >0.15 mg/L result in residual chlorine

3 Due to the high ammonium levels of X15 mg/L in

water type III, chlorine is mainly transformed to

chloramine which is a considerably less efficient

doses of 10–30 mg/L necessary to reach a 2-log

inactivation of Giardia lamblia cysts might cause

substantially higher THM concentrations and

possi-bly form NDMA A satisfactory quality of drinking

water derived from water type III therefore requires

and DOC In addition, a primary disinfection stage

should be implemented

4 Chlorine, THM and oxygen concentrations

de-creased in the distribution system of water type III

(WTP 6) In this water, oxygen is possibly consumed

by nitrification of ammonia in the reservoir and the

distribution system THMs are then possibly

de-graded in the resulting anaerobic milieu Even

though this process is desired, lowoxygen

water quality issues such as corrosion, taste and

odor

Acknowledgements

This study was funded by the Swiss Agency for

Development and Cooperation (SDC) in the

frame-work of the Swiss–Vietnamese Cooperation Project

ESTNV (Environmental Science and Technology in

Northern Vietnam) We are indebted to the Hanoi

Water Business Company for their cooperation and

sampling assistance

References

[1] Berg M, Tran HC, Nguyen TC, Pham HV, Schertenleib R,

Giger W Arsenic contamination of groundwater and

drinking water in Vietnam: a human health threat Environ

Sci Technol 2001;35:2621–6.

[2] Singer PC, ReckhowDA In: Letterman RD, editor.

Water quality and treatment, a handbook for community

water supplies, 5th ed New York: McGraw-Hill, 1999 p.

12.1–12.51.

[3] Singer PC Formation and characterization of disinfection

by-products In: Craun GF, editor Safety of water

disinfection: balancing chemical and microbial risk.

Washington, DC: ILSI Press; 1993.

[4] World Health Organization (WHO) Guidelines for

drinking water quality, 2nd ed., vol 1 Recommendations,

Geneva, Switzerland, 1993.

[5] Miles AM, Singer PC, Ashley DL, Lynberg MC, Mendola

P, Langlois PH, Nuckols JR Comparison of

trihalo-methanes in tap water and blood Environ Sci Technol 2002;36:1692–8.

[6] Golfinopoulos SK The occurrence of trihalomethanes in the drinking water in Greece Chemosphere 2000;41: 1761–7.

[7] Vietnamese Ministry of Health Standards for drinking water, Hanoi, Vietnam, 1988.

[8] Standard Methods In: Eaton AD, Clesceri LS, Greenberg

AE, editors Standard methods for the examination of water and drinking water, 19th ed Washington, DC: American Public Health Association, 1995.

[9] Pinkernell U, Nowack B, Gallard H, von Gunten U Methods for the photometric determination of reactive bromine and chlorine species with ABTS Water Res 2000;34:4343–50.

[10] Berg M, M uller SR, Schwarzenbach RP Simultaneous determination of triazines including atrazine and their major metabolites hydroxyatrazine, desethylatrazine, and deisopropylatrazine in natural waters Anal Chem 1995;67:1860–5.

[11] Gallard H, von Gunten U Chlorination of natural organic matter: kinetics of chlorination and of THM formation Water Res 2002;36:65–74.

[12] Symons JM, Krasner SW, Sclimenti MJ, Simms LA, Sorensen Jr HW, Speitel Jr GE, Diehl AC Influence of bromide ion on trihalomethane and haloacetic acid formation In: Minear RA, Amy GL, editors Disinfection by-products in water treatment: the chemistry of their formation and control Boca Raton, FL: CRC Lewis Publishers; 1996 p 91–130.

[13] Wolfe RL, Ward NR, Olson BH Inorganic chloramines

as drinking-water disinfectants—a review J Am Water Works Assoc 1984;76:74–88.

[14] Morris JC, Isaac RA A critical reviewof kinetic and thermodynamic constants for the aqueous chlorine-ammonia system In: Jolley RL, Brungs WA, Cotruvo

JA, Cummings RB, Mattice JS, Jacobs VA, editors Water chlorination: environmental impact and health effects, vol 4 Ann Arbor, MI: Ann Arbor Science; 1981.

p 49–62.

[15] Margerum DW, Gray Jr ET, Huffman RP Chlorination and the formation N-chloro compounds in water treat-ment In: Brinckman FE, Bellama JM, editors Organo-metals and organomataloids: occurrence and fate in the environment Washington, DC: ACS Books; 1978.

p 278–91.

[16] Kumar K, Margerum DW Kinetics and mechanism of general-acid-assisted oxidation of bromide by hypo-chlorite and hypochlorous acid Inorg Chem 1987;26: 2706–11.

[17] von Gunten U, Driedger A, Gallard H, Salhi E By-products formation during drinking water disinfection: a tool to assess disinfection efficiency? Water Res 2001;35: 2095–9.

[18] USEPA Disinfection profiling and benchmarking gui-dance manual, Appendix C: CT values for inactivation achieved by various disinfectants, EPA-815-R-99-013, August 1999, Washington, DC, 1999.

[19] Langlais B, ReckhowDA, Brink DR, editors Ozone in water treatment: application and engineering, 2nd ed Chelsea, MI: Lewis Publishers, 1991 p 221.