This article was downloaded by: [University of New Mexico]On: 27 November 2014, At: 01:05Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Environmental Technology LettersPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/tent19

Ozonation of aquatic organic matter and humicsubstances: An analysis of surrogate parameters forpredicting effects on trihalomethane formation potentialGary L. Amy a , Paul A. Chadik a , Raymond A. Sierka a & William J. Cooper ba Environmental Engineering Program, Department of Civil Engineering , University of Arizona ,Tucson, Arizona, 85721, U.S.A.b Drinking Water Research Center , Florida International University , Arisona, Miami, Florida,33199, U.S.A.Published online: 17 Dec 2008.

To cite this article: Gary L. Amy , Paul A. Chadik , Raymond A. Sierka & William J. Cooper (1986) Ozonation of aquatic organicmatter and humic substances: An analysis of surrogate parameters for predicting effects on trihalomethane formation potential,Environmental Technology Letters, 7:1-12, 99-108, DOI: 10.1080/09593338609384395

To link to this article: http://dx.doi.org/10.1080/09593338609384395

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in thepublications on our platform. However, Taylor & Francis, our agents, and our licensors make no representationsor warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Anyopinions and views expressed in this publication are the opinions and views of the authors, and are not theviews of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should beindependently verified with primary sources of information. Taylor and Francis shall not be liable for any losses,actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoevercaused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyoneis expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

1

Environmental Technology Letters, Vol. 7, pp. 99-108© Science & Technology Letters, 1986

OZONATION OF AQUATIC ORGANIC MATTER AND HUMICSUBSTANCES: AN ANALYSIS OF SURROGATE PARAMETERS

FOR PREDICTING EFFECTS ON TRIHALOMETHANEFORMATION POTENTIAL

Gary L. Amy* Paul A Chadik, and Raymond A. Sierka, Environmental Engineering Program, Department of CivilEngineering, University of Arizona, Tucson, Arizona 85721 U.S.A.

William J. Cooper, Drinking Water Research Center, Florida International University, Miami, Florida 33199 Arisona, U.S.A.

(Received 28 November 1985, in final form 18 December 1985)

ABSTRACT

Ozone represents a potential oxidant for controlling trihalomethanes (THMs)during water treatment. The results presented herein indicate that partial oxidationof THM precursors by ozone produces by-products which are lesa reactive in formingTHMs upon chlorination. It is demonstrated that surrogate parameters such as non-volatile total organic carbon, UV absorbance, and fluorescence are capable ofmonitoring the performance of the ozonation process in reducing the THM formationpotential of waters containing THM precursors.

INTRODUCTION

Humic substances, comprised of fulvic and humic acids, occur ubiquitously insurface waters and in some groundwater. An extensive survey of surface waters in theUnited States revealed an average dissolved organic carbon (DOC) concentration of4.7 mg H "1 (1). The organic content of natural surface waters is generallycomprised of approximately 50% aquatic humic substances (2), with fulvic acid beingthe predominant fraction (3). This paper describes the ozonation of eight differentwaters, six waters obtained from natural sources and two synthetic waters producedfrom soil-derived humic and fulvic acids. These natural waters contrast the organiccontent of synthetic waters, developed for highly controlled laboratory studies.

It has been well established that the chlorination of waters containing humicsubstances results in the formation of trihalomethanes (THMs) (4). Work by others(5,6) indicates that ozone may be effective as a THM control strategy based on itsability to oxidize THM precursors such as humic substances to partial oxidation pro-ducts which are much less reactive in forming THMs upon chlorination. Much of theprevious work has focused on ozonation of "model compounds" or has emphasized the

99

Dow

nloa

ded

by [

Uni

vers

ity o

f N

ew M

exic

o] a

t 01:

05 2

7 N

ovem

ber

2014

response of a specific water source to ozonation. The research described herein isintended to delineate the response of a broad spectrum of waters to ozonation.

The specific objectives of this paper are (1) to compare the general response ofaquatic organic matter and humic substances present in a wide range of natural andsynthetic waters to ozonation applied for THM control and (2) to evaluate thepotential use of several surrogate parameters for monitoring the effects of ozonationon trihalomethane formation potential (THMFP).

RELATED WORK

Various researchers have studied the ozonation of humic substances and trihalo-methane precursors (6 - 16). Some of these researchers have studied the use of ozoneas a sole process for controlling THMs, while others have examined the application oflow doses of ozone to promote subsequent flocculation (i.e. preozonation).

Related work has also been conducted in the area of statistical relationshipsbetween THMFP and selected surrogate parameters. The USEPA NORS survey (1) found thatnonvolatile total organic carbon (NVTOC) was a reasonable indicator for THM levelsobserved in 80 cities. Singer (17), in a study of North Carolina surface waters foundgood correlations between THMFP and either NVTOC, DV absorbance, or color. Edzwald(18) demonstrated that DV absorbance functioned well as a predictor for THMFP inpilot-scale direct filtration studies, while both UV absorbance and apparent colorproved effective in monitoring the performance of a conventional water treatment plantin reducing THMFP. Chrostowski (14) stated that DV absorbance at 254 nm is the mostwidely used parameter for following the course of ozonation of humic substances.Feissinger (19) indicated that NVTOC, DV absorbance, and fluorescence have been usedto monitor the breakthrough of organics from activated carbon columns.

EXPERIMENTAL METHODS AND PROCEDDRES

Ozonation experiments were conducted in a semi-batch mode using an OREC Model03B1-0 Ozonator. Ozone was applied through a fritted glass bubbler in a small plexi-glass reactor with the following cylindrical dimensions: an inner diameter of 7.5 cmand 50 cm in height. This small-scale reactor was employed due to the limited amount ',.of each natural, water source available: typically 1.5 liter batches of water were |ozonated during a given experimental run. Ozone was applied at a rate of either 112 %or 214 mg/min for times ranging from 0.5 to 10 min. Applied ozone doses were deter- js;mined by a series of KI traps and iodometric titration (20). ;::

IOzonation performance was monitored by the following parameters: nonvolatile £;

total organic carbon (NVTOC), OV absorbance (one cm path length, 254 nm, and pH 7), |'fluorescence (365 nm excitation wavelength and 415 nm emission wavelength), color |:(pH 7), and trihalomethane formation potential (THMFP). A Dohrmann DC-80 Total f;Organic Carbon Analyzer was used to measure NVTOC while DV absorbance was determinedwith a Perkin Elmer Model 200 DV-Visible Spectrophotometer. A series of 50 m£platinum-cobalt color comparison tubes were used to measure color while fluorescencewas determined with a Turner Model 111 Fluorometer. Bromide ion concentrations weredetermined using a Dionex Model 10 Ion Chromatograph. The THMFP of each untreated andozonated water was based on the following conditions: 168 hours, 20*C, ambient pHmaintained by a phosphate buffer (ranging from 6.1 to 8.0 for the natural waters and |established at 7.0 for the synthetic waters), and an applied CI2 to NVTOC ratio of .£-3:1 (mass basis). In addition to conducting a "standard" THMFP based on 168 hours, :•;additional THMFP experiments were conducted for each untreated and ozonated water at ireaction times of 96 and 2 hours, considered to be representative of relatively long-term and short-term THM formation, respectively. The concentration of totaltrihalomethanes (TTHM) was determined by measuring each of the THM species by gaschromatography using liquid/liquid extraction. Aliquots of each untreated andozonated water were centrifuged prior to each of the above analyses to removeparticulate matter. In all cases, samples after centrifugation contained turbiditiesof <1.0 NTD.

100

Dow

nloa

ded

by [

Uni

vers

ity o

f N

ew M

exic

o] a

t 01:

05 2

7 N

ovem

ber

2014

The natural waters were ozonated "as received" in the presence of turbiditylevels ranging from low (<5 NTU) to moderate (<25 NTU), while synthetic waters con-tained no externally introduced source of turbidity. All natural waters were ozonatedat ambient pH conditions with carbonate alkalinity levels ranging from less than 20 toalmost 400 mg !i ~* as CaCC>3, while synthetic waters were ozonated at an initialpH of 7.0 with a carbonate alkalinity of 60 mg &~^ as CaCO3 added in form ofNaHC03. For each water source, the goal was to produce two unique ozonated watersfor subsequent characterization and evaluation. For those particular waters con-taining aquatic organic matter amenable to ozonation, the objective was to produce onewater with a moderate (20 to 402) and another water with a substantial (40 to 60%)reduction in THMFP. For waters containing aquatic organic matter not particularlyamenable to ozone oxidation, more pragmatic considerations were used as a basis foridentifying doses for producing two unique waters. Due to the limited availablesample volume, a small ozone reactor was employed. The inherent disadvantages of thissmall-sized reactor were inefficient mass transfer and incomplete mixing. However,these limitations were not crucial to achieving the goal of this research, which wasmodelling THMFP as a function of measured surrogate parameter levels.

RESULTS AND DISCUSSION

Characterization of Untreated and Ozonated Waters

Samples were obtained from six natural water sources including the Edisto River(SC), Ilwaco Reservoir (WA), Kaw Reservoir (OK), Grasse River (NY), Pearl River (MS),and James River (SD). A soil-derived fulvic acid was obtained from a Canadian source(Contech Ltd., Ottawa) while a purified commercially available humic acid (AldrichChemical Co., Milwaukee) was used as a soil-derived humic acid.

Important characteristics of the untreated waters are summarized in Table 1.With the exception of the Pearl River and Kaw Reservoir, the natural waters studiedgenerally contained low levels of turbidity while pH levels for most of the waterswere within +/- one pH unit of neutral (7.0). Levels of aquatic organic matter andhumic substances varied significantly with NVTOC concentrations ranging from 3.94 to11.3 mg i" , UV absorbance ranging from 0.136 to 0.489, relative fluorescenceintensitites ranging from 20.2 to 94.2X, color ranging from 17 to 125 p.c.u., and 168hour THMFP's ranging from 2.33 to 9.83 mol I'*-. Relative fluorescence inten-sities were based on an absolute fluorescence level of 276 determined for a sampleobtained from the Biscayne Aquifer, another highly colored water investigated in arelated study. Bromide ion levels ranged from 0.010 to 0.245 mg £~*, NH3-Nlevels ranged from 0.04 to 0.25 mg £ ~ % total iron was found at levels rangingfrom below detection to 0.9 mg £~*, and manganese was not detected in any of thesamples. The high alkalinity levels observed in the James River and Kaw Reservoir

TABLE 1. UNTREATED WATER CHARACTERISTICS.

Water Source

Edisto River

Ilwaco Res.

Kaw Res.

Grasse River

Pearl River

Janes River

Huaic Acid

Fulvic Acid

168-hr THMFP

(ymol f l )

9.83

2.80

2.39

3.98

2.33

6.06

2.45

5.08

NVTOC

(mg i-l)

11.3

6.00

5.22

6.56

S.62

13.8

3.94

S.22

UV absorbance

(254 nm a pH 7)

.489

.329

.153

.288

.136

.296

.320

.33*

Rel. Fluor.Intensity (X)

(PH 7)

94.2

31. S

28.3

40.9

21.4

57.2

23.9

28.3

Color(p.c.u.)(PH 7)

93

70

23

55

17

28

123

124

Ambient

PH

7.3

6.1

7.7

e.a

6.6

8.0

7.0

7.0

Turbidity

(HTO)

4.4

3.4

23

1.5

IS

3.2

1.4

0.15

Alkalinity(mg I -1as CaC03)

20

20

140

30

20

390

60

60

101

Dow

nloa

ded

by [

Uni

vers

ity o

f N

ew M

exic

o] a

t 01:

05 2

7 N

ovem

ber

2014

samples indicates a substantial potential for pronounced ozone scavenger effectsexerted by bicarbonate ion.

The applied ozone doses studied during this research are summarized in Table 2,portrayed in terms of mg O3 £~* as well as mg O3 per mg NVTOC. In an ozonereactor, the amount of ozone utilized is equal to the applied O3 minus the O3leaving the reactor in the vent gas, and includes reaction, decomposition, andscavenger effects. The applied ozone doses are high when compared to economicallyreasonable doses commonly used in water treatment. However, only a small portion ofthe applied gaseous ozone was actually absorbed in solution. Mass transfer from thegaseous to liquid phase was limited by the use of a small reactor, the absence ofmechanical mixing, a short contact time, and the creation of relatively large bubblesfrom the glass frit. In short, the experimental design was not intended to optimizeozone transfer but rather to produce unique ozonated waters by achieving certainreductions in THKFP as well as surrogate parameters for subsequent statisticalanalysis. Based on expected low mass transfer efficiency, it would seem reasonable toconclude that the amounts of ozone utilized were substantially less than applieddoses.

As a general rule, pH levels did not significantly vary during the course ofozonation: final pH levels after ozonation were normally within +/- 0.5 pH units ofthe initial ambient level.

Effects of Ozonation on Aquatic Organic Matter and Humic Substances

The untreated and ozonated water characteristics presented in Tables 1 and 2provide insight into the effects of ozonation on various surrogate parameters used toindirectly measure aquatic organic matter and humic substances. Examination of UVabsorbance data indicates that the destruction of material absorbing UV light is moreapparent in the Edisto River, Grasse River, and Kaw Reservoir when compared to theother water sources. Moreover, the synthetic water results come close to "bracketing"the behavior of the natural waters. It is important to note that aromatic structureseffectively absorb UV light and that humic and fulvic acid molecules are characterizedby a largely aromatic infrastructure. NVTOC data indicate that relatively littlecomplete oxidation of aquatic organic matter occurs; however, ozone creates partialoxidation by-products with significantly different chemical characteristics. Whilethese by-products still appear as part of NVTOC measurements, they are characterized

TABLE 2. CONDITIONS AND RZSULTS OF OZONATION EXPERIMENTS.

Water Source

EdistoEdisto

11 waco11 waco

KawKaw

QrasseGrasae

PearlPearl

JaaesJaaes

Bunlc AcidHumic Acid

Fulvlc AcidFulvic Acid

ADD!led

»g t-i

74.81430

74.81430

37.4187

37.474.8

187748

150748

75.2150

37.474.8

Oq

mg C

6.62127

12.5238

7.1635.8

5.7011.4

33.3133

10.954.0

19.639.2

7.1614.3

168-HR THMFP

6.183.16

2.02. 1.23

2.071.90

3.342.57

2.231.81

5.835.75

2.431.50

3.302.08

NVTOC(ng £~1)

10.18.67

5.275.20

5.224.81

6.546.33

5.223.88

13.412.1

3.6S3.45

4.744.27

UV absorbance(254 nm & pH 7)

.172

.070

.170

.168

.066

.045

.166

.120

.058

.053

.152

.095

.171

.090

.171

.083

Rel. Fluor.Intensity (X)

(PH 7)

23.612.7

11.25.1

8.74.0

14.99.1

3.31.4

26.19.8

19.910.5

27.214.1

Color(p.c.u.)(pH 7)

244.0

1010

3.03.0

3025

3.02.0

6.01.0

4512

3813

102

Dow

nloa

ded

by [

Uni

vers

ity o

f N

ew M

exic

o] a

t 01:

05 2

7 N

ovem

ber

2014

by a reduced reactivity in forming THMs. This behavior can be evidenced by simplecalculations of THM yield (ymol THM per mg NVTQC) for a given untreated and a corre-sponding ozonated water. The data indicate that ozone significantly reduces thefluorescence of aquatic organic matter and humic substances. The effects of ozonationon THMFP are graphically shown in Figure 1 for data derived from the various watersources. This figure portrays normalized 96-hour THMFP levels (i.e., ratios of THMFPin ozonated waters to untreated waters) as a function of normalized applied ozonedose, expressed as the ratio of mg O3 per mg NVTOC. There appears to be consider-able data scatter in Figure 1, however, careful perusal of data subsets correspondingto each water source indicates some trends. These THMFP data indicate that the GrasseRiver, Edisto River, Ilwaco Reservoir, Raw Reservoir, and fulvic acid are more

1.0

1 0.4

0.2

'5 10 20- 30 40 50 100Applied Og/NVTOC (mg per mg)

Figure 1. • Normalized THMFP Reductions Observed in Ozonated Waters.

200 300

amenable to ozonation for THM control when compared to the Pearl River, James River,and humic acid. It is interesting to note that, even for some of the waters amenableto ozonation for THM control, there appeared to be a fraction of THM precursors re-fractory to ozone oxidation. This behavior is most evident in results derived fromthe Edisto River in which a lower O3 dose achieves a significant THMFP reduction,while a much higher dose accomplishes a relatively small additional incremental reduc-tion. This implies that a portion of the precursor that produces THMs is quicklydestroyed by ozone while other portions of the precursor are not significantlyaffected by the ozonation process under the given conditions.

The applied ozone doses employed in this research were admittedly high. However,the goal of this research was not to optimize the ozonation of THM precursors butrather to make a preliminary attempt at defining correlations between THMFP and vari-ous surrogate parameters for a diverse array of natural and synthetic water sources.Quite simply, high ozone doses were.applied to small sample aliquots (1.5 £) to pro- .duce readily measureable responses in terms of THMFP and surrogate parameters.Although utilized ozone doses were not measured, it is felt that the results presentedherein are representative of the high extreme of the range of viable ozonationconditions.

Evaluation of Surrogate Parameters for Predicting THMFP

Amy, et. al. (21) previously reported on a preliminary surrogate parameter

103

Dow

nloa

ded

by [

Uni

vers

ity o

f N

ew M

exic

o] a

t 01:

05 2

7 N

ovem

ber

2014

analysis of these data in conjunction with other data. The following discussion'summarizes the results of a subsequent, more rigorous analysis.

The data base subjected to statistical analysis consisted of a total of 48 cases(i.e., unique data entries) based on eight sources, three unique waters derived fromeach source (one untreated and two ozonated waters), and two reaction times (96 hoursand 2 hours). Data were entered into a CDC Cyber 175 computer system and used tocreate a data base with the Statistical Package for the Social Sciences (SPSS) soft-ware package (22, 23). The SPSS data base was analyzed using the SPSS subroutine"Regression." Simple linear regressions provided values of r^ (simple coefficientof determination), n (number of cases), overall F value (F distribution statistic),and a (level of significance or probability of the null hypothesis that thecorrelation is zero in the population from which the sample was drawn).

A series of simple regressions were conducted using selected portions of the database in order to evaluate the applicability of various surrogate parameters for pre-dicting THMFP. The first set of regressions used data derived from a reaction time of96 hours, designated as being representative of "long-term" THMFP, while a second setof regressions repeated the statistical analyses employing data derived from areaction time of 2 hours, designated as being representative of "short-term" THMFP.Each set of regressions focused on several different data subsets including (1) alldata, (2) untreated water data only, and (3) ozonated water data only. Eachregression attempted to discern the correlation between THMFP and each surrogateparameter using relevant portions of the data base. In addition to individualsurrogate parameters, the multiplicative parameters UV-TOC, Fluor-TOC, and Color-TOC,representing the products of UV absorbance and NVTOC, fluorescence and NVTOC, and v

color and NVTOC, respectively, were also evaluated. Of these, UV-TOC proved to be thebest multiplicative surrogate.

Results of these regressions are presented in Table 3, showing a summary of thesimple coefficient of determination (r*), the number of cases (n), the F-sta'tistic,and the corresponding significance «• These results suggest that color is a poorsurrogate for THMFP. Other surrogates appear to have potential merit depending on thespecific data set of concern. Considering untreated water data, fluorescence appearedto be the best single surrogate parameter followed by either NVTOC or UV absorbance,depending on whether a long-term or short-term timeframe is considered. However, themultiplicative parameter, UV-TOC, provided the best overall correlation. Considering

TABLE 3. ANALYSIS OF SURROGATE PARAMETERS FOR 96-HOOR S 2-HOOR THMFP USING DIFFERENT DATA SETS.

SurrogateParameter

NVTOCUV Aba.Fluor.ColorOV-TOC

NVTOCUV Aba.Fluor.ColorUV-TOC

NVTOCUV Abs.Fluor.ColorUV-TOC

n

2424242424

88

a88

1616161616

A 1 196-Hour

'2

.676

.414

.673

.136

.786

A l l.637.621.914.053.925

A l l.828.077.395.002.679

R a x &THMFP

F

49.915.645.43.48

80.9

R a w H a t10.59.84

63.70.34

73.6

T r e a t e d67.51.179.140.02

29.6

a

<.001.001

<.001.076

•C.001

e r D.018.020

<.001.534

<.001

H a t<.001.297.009.884<.001

n

2424242424

a t aaa8

a8

e r D1616161616

D a t a2-Hour THMFP

r2

.361

.608

.790

.357

.770

.468

.737

.807

.170

.837

a t a.384.131.622.139.404

F

12.433.882.a12.273.5

5.2816.825.21.23

30.8

S. 742.1223.12.269.48

a

.002<.001<.001.002

<.001

.061

.006

.002

.310

.001

.010

.168<.001.155.003

treated water data, the best correlation for long-term data was provided by NVTOCwhile fluorescence yielded the best correlation for short-term data. For bothlong-term and short-term data, the multiplicative term UV'TOC yielded the second best

104

Dow

nloa

ded

by [

Uni

vers

ity o

f N

ew M

exic

o] a

t 01:

05 2

7 N

ovem

ber

2014

correlation. It is interesting to note the low correlation observed for UVabsorbance for ozonated waters. Finally, when considering both untreated and ozonatedwaters together, the multiplicative parameter UV-TOC yielded the best correlation forlong- term data while, for short-term data, 'fluorescence and UV-TOC yielded comparablecorrelations. Considering those parameters which correlated best (UV-TOC, NVTOC, andfluorescence), better correlations were generally observed for 96-hour when comparedto 2-hour data. This is likely due to greater variations in THM formation among thevarious waters over a short-term reaction timeframe compared to a long-term reactiontime frame. Finally, some of the observed data scatter can be attributed to thedifferent ambient pH conditions (6.1 - 8.9) maintained during the THMFP experiments,although most ambient pH levels were close to neutral.

The above statistical analysis indicated that NVTOC, UV absorbance, and fluor-escence showed potential merit as surrogate parameters for THMFP. It was particularlynoteworthy that the multiplicative term, UV-TOC, appeared to most accurately simulateTHMFP. The multiplicative term UV-TOC is intended to simulate the complex response ofhumic substances in forming THMs: ostensibly, NVTOC may be representative of the totalamount of precursor, while UV absorbance may be indicative of the reactivity of theprecursor in forming THMs. A graphical representation of these correlations appearsin Figures 2 through 5, representing plots of 96-hour THMFP versus NVTOC, UV absor-bance, fluorescence, and UV-TOC, respectively. Each figure portrays untreated andozonated water data as well as the regression lines derived from regressing all data,untreated water data alone, and ozonated water data alone. Although fluorescenceapparently represents the best individual surrogate parameter, the multiplicative termUV'TOC provides a better simulation, as evidenced by higher r* values.

CONCLUSIONS

Aquatic organic matter and humic substances vary in their amenability towardozonation for THM control. There appears to be both a fraction of THM precursorsamenable to ozone destruction as well as an "ozone refractory" fraction. The relativepredominance of these fractions varies from source to source. Given the range ofwater sources and applied ozone doses studied, 168-hour THMFP reductions ranged fromas little as IX to as high as 682. Both fluorescence and NVTOC demonstrated potentialmerit as nonspecific surrogate parameters for predicting THMFP although a multi-plicative term, UV-TOC (the product of UV absorbance and NVTOC), proved to be the mostaccurate.

ACKNOWLEDGEMENTS

Financial support for this research was provided by the U.S. EnvironmentalProtection Agency under Grant # R809935-01. The contents do not necessarily reflectthe views and policies of the E.P.A.

REFERENCES

1. J. M. Symons, et al., Jour. AWWA, 67, 634-647 (1975).

2. E. M. Thurman and R. L. Malcolm, in Aquatic and Terrestrial Humic Materials,R. Christman and E. Gjessing, eds., Ann Arbor Science, 1983, p. 1.

3. R. R. Trussell, et al., Jour. AWWA, 70, 604-612 (1978).

4. J. Rook, Environ. Sci. Technol., 11, 478-482 (1977).

5. R. Rice, in Water Chlorination: Environmental Impact and Health Effects, Vol. 5,R. L. Jolley, ed., Ann Arbor Science, 1985, p. 1215.

6. J. Veenstra, J. Barber, and P. Khan, Ozone Sci. and Engrg., 5, 225-244 (1983).

105

Dow

nloa

ded

by [

Uni

vers

ity o

f N

ew M

exic

o] a

t 01:

05 2

7 N

ovem

ber

2014

10

ft ~~

£7u o

O Untreated[j] Ozonated

All Data

o

o

10 12 14

NVTOC (mg l " )

Figure 2. Correlations of 96-hr THMFP versus NVTOC Based on all Data (r = 0.676),Untreated Water Data (r2 = 0.637), and Ozonated Water Data (r2 = 0.828).

10

UntreatedOzonatedAll Data

o

0.2 0.3UV Absorbance

074 0.5

Figure 3. Correlations of 96-hr THMFP versus UV "Absorbance Based on AllData (r2 = 0.414), Untreated Water Data (r2 = 0.621), and Ozonated WaterData (r2 = 0.077) .

106

Dow

nloa

ded

by [

Uni

vers

ity o

f N

ew M

exic

o] a

t 01:

05 2

7 N

ovem

ber

2014

10

^ 6

ui

•o•

IT

UntreatedOzonated

UB ,<

—

0 20 40 • 60 80 100Relative Fluorescence Intensity (%)

Figure 4. Correlations of 96-hr THMFP versus Fluorescence Based on AllData (r2 = 0.673), Untreated Water Data (r2 = 0.914), and Ozonated WaterData (r2 =0.395).

10

O UntreatedQ Ozonated

All Data

1.0 2.0 3.0 4

UV-TOC (mg 1~ cm" )

.0 5.0 6 0

Figure 5. Correlations of.96-hr THMFP versus UV'TOC Based on All Data(r2 = 0.786) , Untreated Water Data (r2 = 0.925), and Ozonated WaterData (r2 = 0.679).

107

Dow

nloa

ded

by [

Uni

vers

ity o

f N

ew M

exic

o] a

t 01:

05 2

7 N

ovem

ber

2014

7. W. Glaze, G. Peyton, S. Huang, and J. Burleson, Environ. Sci. and Technol., 16,454-458 (1982).

8. J. Larose, A. Grasmick, P. Blondean, S. Elmaleh, and J. Legeron, Ozone Sci. andEngrg., 4, 79-89 (1982).

9. M. Umphries, et al., Ozone News, Vol. 6, No. 3, Part 2, March 1979.

10. R. Rice, Ozone Sci. and Engr., 2, 75-99 (1980).

11. J. Mallevialle, in Oxidative Techniques in Drinking Water Treatment, W. Kuhn andH. Sontheimer, eds., U.S. EPA Report No. EPA-570/9-79-020, 1979, p. 291.

12. J. Lawrence, H. Tosine, F. Onuska, and M. Comba, Ozone Sci. and Engrg., 2, 55-64(1980).

13. D. Maier, in Oxidative Techniques in Drinking Water Treatment, W. Kuhn and H.Sontheimer, eds., U.S. EPA Report No. EPA-570/9-79-020, 1979, p. 394.

14. P. Chrostowski, Physical-Chemical Mechanisms of Reactions of Ozone with AquaticHumic Substances, Ph.D. Dissertation, Drexel Universtiy, Philadelphia,Pennsylvania (1981).

15. R. Sierka and G. Amy, Ozone Sci. and Engrg., 7, 47-62 (1985).

16. D. Reckhow and P. Singer, Jour AWWA, 76, 151-157 (1984).

17. P. C. Singer. et al., Jour. AWWA, 73, 392-401 (1981).

18. J. Edzwald, Removal of Trihalomethane Precursors by Direct Filtration andConventional Treatment, USEPA Report, EPA-600/2-84-068, 1984.

19. F. Feissinger and L. Brener, in Activated Carbon Adsorption of Organics from theAqueous Phase, M. McGuire and I. Suffet, eds., Ann Arbor Science, 1980, p. 489.

20. American Public Health Association, Standard Methods for the Examination of Waterand Wastewater, 15th Edition, 1980.

21. G. Amy, P. Chadik, and P. King, in Proceedings, AWWA 1984 Annual Conference,1984, p. 1651.

22. N. Nie, et al., Statistical Package for the Social Sciences, McGraw-Hill, 1975.

23. C. Hull and N. Nie, SPSS Update 7-9, New Procedures and Facilities for Releases7-9, McGraw-Hill, 1981.

108

Dow

nloa

ded

by [

Uni

vers

ity o

f N

ew M

exic

o] a

t 01:

05 2

7 N

ovem

ber

2014