7

.. . rr.‘ 5&3 . ‘ .
. to. gagififlw
. E4. ........ "HHS“. .733, link? «:3

 

latx

- .->...€1t..v»..X ‘
P.-v.‘?v.§. lulu .b-fi‘ks 7.19:}.

¢ CHOU. . Killtneoficnbl‘gu‘vfivt 2‘91! .

, I V ...§n.?t!,'£1§.’v7!ub|§.

5
p in." g: c u
. a. $77 T..Vs.ui1t¢. . I 51“.! o .
. c r... A {daunfiufigfivfifiu’fifflkfl
...vi..1‘v§..oé- tat...“ art? a {“14an
. _ » 1.1.1 6.97: 731 9...! e 2.99 {4.0 r. w
n vzyvg. .0 nyflgfimlfla a bi. .1
). . 9.1.1.315. I? .fifit ......2! v I
v o 3.5? .79 m . ‘ 5. i.

.H I‘ e v‘
Velma-33 1..

... ;
r h
. A U ~
v V vi 6 ‘ <
x 1 u? .. u. u
v t . uni’ :6 L
£9949... Gelx .96... v.
. . h f“; tig,

’ .. 73%. V.

. «Pain»; :1... . .V In. .
I iii-£19Ikhuléz.‘ $1) :1!
, 13"“ Saul... «‘4
.wfiquJv. QIVUWJWAN‘FQVOMQI ink. in: u‘
A L v .zt‘ . il... ‘nliu‘XSYMA .1?
a ‘Q’kutiefic: L #5.;temgr . my...
3 L5 q 57.“: n
c 1 a1.nu\a'95fu.‘u Q a»! a
, - 3. 325.33%in We...
italivomulvvlnln ; 0? n that

, , . .2
‘ _. x
5.! , SI!) m
i .in .

7 153... is

3;
_ :.

Igiafifld‘wRPH
fin... £18.. »
171‘. 3.. ”£510.“;
a)..1\llu§\hium , i. a:

.Ivtol‘unv . .w « , ‘ . A A... ‘ ,w l
a y .u. ‘a_ h.:..§

\ f- . . _
. .73.! .1}: v1

iiiv‘té‘lnhllxm..\lvt1éé

Cliniit agi‘ . n! v .. . .

EV... . , . a. y w a: ‘ ..

_ . ‘1... h . r

‘ . ~ ‘ a. 3 .fliflw&§g$hflskk h A

V ,iuq‘haflg h s}

.. . 965W?! “own!

‘13 .

‘ [9‘11“ ’0'? \XL.
EVIL? i!
Sinai:
x8;

. .
§3¢5 i..t.z.it‘1t\$if
iizkf‘rul .v MFA-211:? .\%5 ihAtlti’tifsEt
T... faincuwg.lfi. Na
kiisll': . 93.5%,? .
1.1;}! ‘1’... . v ‘ .
“Rf O « [Fright .M£5.v~
ibznhvlu). u«.3v.,f..vg H.31qn421‘025‘. . ;
. . . (cg-{ilk ‘
. p 910534.? tt&» , ‘
.AL&§.%§E (1.5!. lift...“ a
it“ 1!?! A .é.» - .. 31. e A
.

. K » t. . a. J A _ | . . ‘ v .

.y» , .I .. 7. {W P 15"), .111! +5 ~ ‘

nit ifiu‘VYVnux,‘~uhrt; rbl.:l..u(.1 :YM29BUME}: x! iXMuNMPVUVLleLHNQtiVV. “WEWDvfllg‘hngipn £85., . n

H...» .\.§|.ollr£l ill! fl‘bz .l‘imuflmVHs-i5vyoflufl;\nrnb! Sufi-”but. Igz‘toagvfiflwflm’hflmfifiwmm‘tfifldil

f , : . .ww‘ltilbfl it! I. 1-...I. .911J.|§§u.0 .tll‘clt «1?:
_ .. “23??“ ,+ . 1.. 1:4. ”fin. .

l/lI/l/llll/l/Ill/lI/lll/I/l/lII/l/l/l ////l////l/l///I//////I ‘

_ . 3 1293 10525 4159

' -:"""_’ Sréjf...; --

This is to certify that the

thesis entitled

Survey of Trihalomethane Levels in
Chlorinated Drinking Water of Michigan
Treatment Plants

presented by

Eileen Antoinette-Nickerson Furlong

has been accepted towards fulfillment
of the requirements for

M.S. degree in Fisheries and
Wildlife

 

625.12% [MA-

Major professor

Date 5_////€°y~

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

)V1531_J RETURNING MATERIALS:
Place in book drop to
LIBRARIES remove this checkout from
JI-uzs-unL your record. FINES will

be charged if book is
returned after the date
stamped below.

 

 

 

 

 

re.

”"*a-

are 6 raw-v

 

 

G //’743/

SURVEY OF TRIHALOMETHANE LEVELS IN
CHLORINATED DRINKING WATER OF MICHIGAN
TREATMENT PLANTS

By

Eileen Antoinette-Nickerson Furlong

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

MASTER OF SCIENCE

Department of Fisheries and Wildlife

T982

ABSTRACT
Survey of Trihalomethane Levels in Chlorinated Drinking
Water of Michigan Treatment Plants
By

Eileen Antoinette-Nickerson Furlong

Raw and finished water of 40 Michigan treatment plants were
analyzed for trihalomethane presence by the liquid-liquid extraction
method.

Total trihalomethane (TTHM) levels in the finished water of the 40
treatment plants ranged from undetectable to 281.8 ug/L with a mean of
44.0 ug/L and a median of 20.5 pg/L. Range and median values are
lower than those of the NOR and NOM surveys. Chloroform was usually
present at the highest concentrations followed by bromodichloromethane,
dibromochloromethane and bromoform.

Chlorine dose influences the amount of TTHM formed. After
adjusting for this influence, treated drinking water supplies derived
from surface water had significantly higher TTHM levels than either
Great Lakes or groundwater. TTHM levels of Great Lakes and groundwater
were significantly different only when chlorine dose effect was not
accounted for, perhaps because the organic levels are low compared to
surface water. However, when chlorinated, Great Lakes water tended to

have higher TTHM levels than did groundwater.

ACKNOWLEDGMENTS

I gratefully thank my major professor, Dr. Frank D'Itri, and
the members of my committee, Drs. Matthew Zabik and Charles Liston,
for their guidance and support.

In addition I thank the following people: Drs. J. L. Gill
and J. P. Geisy for aiding with the statistical tests and computer
program, respectively, Mr. Stephen Furlong for photographing the
map, the typists Mrs. Margaret Beaver and Ms. Terry Waters for their
job well done, and Mr. Swat Kaczmar for his instructions on the use

of the gas chromatograph.

ii

TABLE OF CONTENTS

Page

LIST OF TABLES .......................... iv
LIST OF FIGURES ......................... v
INTRODUCTION ...... ' ..................... 1
Discovery and Prevalence of THM's in Drinking Water ..... 1
Human Health Effects ..................... 2
Animal Metabolism and Toxicity ................ 3
Animal Carcinogenicity .................... 7
Trihalomethane Formation ................... 10

THM Level Regulated ..................... 23

The Study .......................... 24
METHODS AND MATERIALS ...................... 25
Treatment Plant Selection .................. 25

THM Sampling and Extraction ................. 25
Analytical Procedure ..................... 26
Statistical Analyses ..................... 27
RESULTS ............................. 28
General Results of Survey .................. 28
Geographic Distribution ................... 29
Effect of Water Source and Chlorine Dose on TTHM Levels . . . 32
Predicting Trihalomethane Levels ............... 33
DISCUSSION ............................ 34
SUMMARY ............................. 40
APPENDICES ............................ 41
Appendix A - Methods and Materials .............. 41
Appendix B - Survey of Michigan Water Treatment Plants . . . . 55
Appendix C - Statistical Analyses .............. 66
Appendix D - TTHM Predictive Models ............. 72
LITERATURE CITED ......................... 73

Table

B-1

C-l
D-l

LIST OF TABLES

Median and range concentrations of four trihalo-
methanes found in chlorinated drinking water by

NORS and NOMS (phase I) ................

Summarization of chloroform carcinogenicity studies

in animals ......................

Mean, median and range of trihalomethane levels found
in finished water of forty Michigan drinking water

plants ........................

The thirteen surveyed treatment plants which are ‘
affected by the EPA regulation, and their population

size served and TTHM levels found in finished water . . .

Survey of Michigan water treatment plants: total
trihalomethane content of chlorinated drinking water

(Part I) .......................
(Part II) .......... . .............

Data used for statistical analyses ..........

Models developed using multiple regression to predict

TTHM levels following water chlorination .......

iv

Page

38

LIST OF FIGURES

Figure Page
1 Proposed metabolism of chloroform having as an inter-
mediate phosgene (modified from Mansuy et al. (1977)
and Pohl et al. (1977)) .................. 5

2 The classical haloform reaction (from Morris, 1976;
Morris and Baum, 1978) .................. 12

3 Proposed degradation pathway of fulvic acids and
resorcinol. Cl+ represents in a simplified way any
electrophilic halogenating species of the series XOHZ,
H2, HOX, X20 where X = 1, Br or Cl (from ROOK, 1977) . . . 18

4 Proposed mechanism of CHCl3 production from citric
acid (from Larson and Rockwell, 1978) ........... 21

5 Geographic distribution of levels of four trihalo-
methanes found in finished water of 40 Michigan water
treatment plants ..................... 31

A-I Cover letter for preliminary questionnaire ........ 42
A-2 Preliminary questionnaire for determining which treat-

ment plants may be of value for THM analysis and

individual treatment conditions .............. 43

A43 Sampling and extraction instructions sent to water
treatment facilities ................... 47

A-4 Questionnaire used at time of sampling .......... 50
A-5 Precision curve of chloroform from injection of a 2 pl

mixed standard into gas chromatograph (mean i standard

deviation of 3 injections) ................ 52
A-6 Elution order of organohalide mixed standard ....... 54
C-1 One-way analysis of covariance (unequal replication) . . . 68
C-2 Statistical test to determine if quantity of chlorine

applied to raw water has an influence on TTHM levels

in finished drinking water ................ 69

C-3 Mean total trihalomethane values adjusted for chlorine
concentration effects ................... 69

V

Figure
C-4

C-5

LIST OF FIGURES (Continued)

Statistical difference between mean TTHM levels in the
finished waters when Great Lakes, groundwater and

surface water were used as water sources after chlorine
concentration effect is corrected for ...........

Statistical difference between mean TTHM levels in the
finished water when Great Lakes and groundwater were
used as water sources assuming no chlorine concentra-
tion effect .....................

vi

7O

71

INTRODUCTION

 

Discovery and Prevalence of THM's in Drinking Water

In most instances, the chlorination of drinking water results in
the presence of trihalomethanes (THM's) where none were previously
found. Though other researchers (Kleopfer and Fairless, I972; Novak et
al, l973; Tardiff and Deinzer, l973) previously noted the presence of
THM's in drinking water, the reaction of chlorine with naturally
occurring organics to form THM's was first noted by Rook in l974, and
then, later the same year, reported also by Bellar, Lichtenberg and
Kroner (l974).

Following this discovery, the Environmental Protection Agency
(EPA) undertook a nationwide survey called the National Organics
Reconnaissance Survey (NORS) to determine the extent of contamination
of drinking water by halogenated organics includingthe THM's (Symons
et al., l975). Of the 79 cities surveyed which chlorinated drinking
water, all had THM's in the finished water. When THM's were located in
raw water, it was at very low levels.

Later an additional survey, the National Organics Monitoring
Survey (NOMS), of 113 public water systems was undertaken by the EPA
(U.S.EPA, l978; U.S. EPA, l979). The results of the Phase 1 portion of
NOMS are similar to those of NORS. Median and range concentrations of

both surveys are given in Table l.

Table l. Median and range concentrations of four trihalomethanes found
in chlorinated drinking water by NORS and NOMS (phase I).

NORS NOMS

 

 

Concentration, ug/L

 

 

Trihalomethane Median Range Median Range

CHCl3 21 LD*- 311 27 L0 - 271
CHClZBr 6 LD - 116 10 L0 - 183
CHClBr2 1 LD - 100 L0 L0 - 190
CHBr3 5 LD - 92 L0 L0 - 39
Total THM 27 L0 - 482 45 L0 - 457

 

*LD, less than detection limit.

Human Health Effects

 

No chloroform was detected in the plasma of New Orleans residents
(Dowty et al., l975). However, in later studies, trihalomethanes have
been detected in blood and adipose tissues of individuals who drink
chlorinated water. In a 1976-l979 survey of Miami, Florida area
residents, all participants were found to have chloroform levels of ID
to 60 ug/L in blood (Enos, 1979). In a later study of the same city
trihalomethane levels of 2 to 400 ug/L were found in the blood and
adipose tissue (Pfaffenberger, 1980).

Many epidemiological studies have been done which link drinking
water to cancer mortality. Two basic approaches of analysis have been

attempted. The first compares cancer mortality rates between people

3

who drink groundwater and those who drink surface water (Kuzma et al.,
1977; Wilkins et al., 1979). By the second approach, cancer mortality
rates are compared between individuals drinking chlorinated and
nonchlorinated water, or cancer mortality rates are compared according
to THM levels in drinking water (Alavanja et al., 1978; Cantor et al.,
1977; Hogan et al., 1979; Wilkins et al., 1979).

From the accumulated studies, there appears to be a relationship
between drinking water and mortality incidences from total cancers, and
various urinary and gastrointestinal cancers. Many criticisms,
however, of these preliminary epidemiological studies do exist (Wilkins
et al., 1979; Maugh, l981a) which do not allow a causative ruling of
drinking water's effect on cancer mortality. Presently, more
comprehensive epidemiological studies are being done to clarify this

possibility (Maugh, 1981a).

Animal Metabolism and Toxicity

 

After exposure to chloroform, chloroform accumulates mostly in
fat, then in the liver, and to a lesser extent in blood, brain, kidney,
and muscle (Cohen and Hood, 1969). Metabolism appears to occur in both
the liver and the kidney, though more is known of chloroform metabolsim
in the liver. Metabolism occurs in the microsomes isolated from the
liver by cytochrome P-450 (a co-enzyme) mediated mixed function oxidase
(Pohl and Krishna, 1978, McMartin et al., 198l). Factors which induce
cyctochrome P-450, such as phenobarbital (Gopinath and Ford, 1975;
McMartin et al., 1981), polybrominated biphenyls (Kluwe et al., 1978a;

Kluwe et al., l978b) and fasting (Nakajima and Sato, 1979; McMartin et

al., 1981) increase the rate of metabolism. The final metabolites
appear to be carbon dioxide and chloride (Rubenstein and Kanics, 1964;
Van Dyke et al., 1964; Fry et al., 1972; Brown et al., 1974; Taylor et
al., 1974).

Three pathways to these end products have been suggested. The
first proposed intermediate to C02 formation is a free radical (Brown et
al., 1974; U.S. EPA, 1980). Evidence for this pathway was obtained
mostly from what is known of carbon tetrachloride metabolism (Smuckler,
1976; Reynolds, 1977; Poyer et al., 1978) and from the chemical
properties of chloroform (Van Dyke, 1969).

The second proposed intermediate is methylene chloride. The
pathway entails the reduction of chloroform to methylene chloride and
its subsequent oxidation to formaldehyde, formic acid and finally
carbon dioxide. None of the intermediates have been isolated in an in
vitro or in vivo system (Rubenstein and Kanics, 1964).

In the final pathway, phosgene is the suggested intermediate to C02
formation. After administering choroform, phosgene was trapped using
cysteine to form 2-oxothiazolidene-4-carboxylic acid (Mansuy et al.,
1977; Pohl et al., 1977; Pohl and Krishna, 1978). Oxygen, microsomes,
cytochrome P-450 and NADPH appear to be necessary for phosgene
formation. Pohl and coworkers (1977) and Mansuy and coworkers (1977)
proposed that chloroform is oxidized to trichloromethanol which
spontaneously dehydrochlorinates to give phosgene.

The metabolite of chloroform may do one of two things. If
glutathione, a tripeptide containing the sulfhydryl-bearing amino acid
cysteine sensitive to electrophilic attack (Cagen and Klaassen, 1980),

is present, the metabolite binds to it (Ekstrom and Hogberg, 1980).

Glutathione is necessary for metabolism of chloroform to carbon dioxide
(Rubenstein and Kanics, 1964). When glutathione levels are depleted,
the chloroform metabolite then binds covalently to microsomal protein
and lipids (Ilett et al., 1973; Uehleke and Werner, 1975; Brown et al.,
1974; Ekstrom and Hogberg, 1980).

Pieces of evidence suggest that phosgene is a highly reactive
electrophile (Mansuy et al., 1977) capable of binding with glutathione,
and it is the reactive metabolite responsible for hepatotoxicity,
nephrotoxicity and C02 formation (Pohl et al., 1980). Removal of
phosgene with the trapping agent cysteine results in both reduced C02
formation and reduced covalent binding of phosgene to microsomal
protein (Pohl et al., 1980). Figure 1 gives the proposed pathway of

chloroform metabolism.

0

 

 

2, NADPH PH -HCl //Cl phosgene
CHCl3 :Cl-E-Cl r #0 = C
cytochrome P-450 1 \‘Cl

microsomes glutathione
binding to glutathione-
microsomal phosgene

protein complex
CD2 + Cl ,

Figure 1. Proposed metabolism of chloroform having as an intermediate
phosgene (modified from Mansuy et al. (1977) and Pohl et
al. (1977)).

Chloroform-induced toxicity occurs mostly in the liver and to a
lesser extent in the kidney (Ilett et al., 1973). Most of the current
research is on hepatotoxicity. TWO phases of metabolism led to
hepatotoxicity and possibly nephrotoxicity (Ekstrom and Hogberg, 1980).
During the first phase, chloroform metabolism occurs followed by
glutathione depletion caused by binding of metabolite to glutathione
and inhibition of glutathione synthesis, and finally binding of the
metabolite to cellular protein. The second phase entails glutathione
deficiency, lipid peroxidation and necrosis.

Phenobarbital, an inducer of cytochrome P-450, induces the
proliferation of endoplasmic reticulum in the centrilobular areas of
the liver (Burger and Herdson, 1966). Covalent binding of the
metabolite occurs mostly in this area (Ilett et al., 1973). From
autoradiographic analysis, covalent binding was found to be located
mostly in the necrotic lesions (Ilett et al., 1973).

Another mode of hepatotoxicity appears to be related to the
ability of the metabolite to inhibit the microsomal ATP-dependent
calcium pump which could lead to intracellular calcium accumulation
(Moore, 1980). Both altered calcium metabolism (Mikkelson, 1978;
Cheung, 1980) and binding of metabolites to tissue macromolecules
(Miller and Miller, I966) have been implicated in carcinogenicity.

In order for necrosis and possibly carcinogenicity to occur,
glutathione depletion must first occur (Brown et al., 1974; Ekstrom and
Hogberg, 1980). This evidence suggests a threshold level is necessary

before deleterious effects occur.

Animal Carcinogenicity

 

From available studies (Table 2), chloroform given by oral
intubation in an oil-based vehicle caused liver, thyroid and kidney
tumors in mice and rats of both sexes. The National Academy of Science
(1977) reviewed the available animal toxicity studies and concluded
that a human health risk exists from exposure to THM's in chlorinated
water. Criticism of the use of the available animal carcinogenicity
studies to determine human health risk is abundant (U.S.EPA, 1979).
Risk estimation was based mostly on the National Cancer Institute (NCI)
study (1976). The major criticisms stem from the use of an oil-based
vehicle rather than water, chloroform was given by gavage rather than
gg_libitum, and very high doses and a small range of doses were used
(Tardiff, 1977; Reitz et al., 1978; U.S.EPA, 1979; Budiansky, 1980).
The studies do not duplicate actual THM exposure from drinking water
consumption. Furthermore, Eschenbrenner (1944) noted that hepa-
tomas did not occur unless hepatic necrosis first occurred. This study
however, was relatively short-term; it was over by the fifth month.

EPA defended its approach to risk assessment by stating that from
high dose to low dose and animal to man extrapolation of data is
acceptable, and methods do not exist enabling a determination of
threshold levels for adverse effects from long-term exposUre (U.S.EPA,
1979). In dealing with human health effects, EPA takes a conservative
approach favoring safety. To establish interim regulations, EPA needs
only to suspect an adverse human health effect (U.S. EPA, 1979).
Presently a carcinogenicity study on mice and rats' exposure to
low-dose chloroform in drinking water is being performed by the EPA

(Jorgenson and Rushbrook, 1980).

mcoeza
u_ocxsu . mmymsmw
mgoszu moccwx . mmpms

masocaezp
mcoszu u_umam;

mLoE:g
xmcvwx . x—co ores
mLoE:p uwoczgu
mLoE:p Lm>__

mLoE:p Paco;
mg053p ovumam;

mgoszu chmc
mgoszu o_pmamc

camcop mxmwz NH um>cmmno momm>mm
>2 m__o cgoo cw mmxz mfi L0$ x3 Ammpmsmo ecu mmpmsv
Log mmswu m mmx\ms omH cg a: mum; Poucmzumcconmo

gmmcop mxmmz

«a um>gmmno max: mu cow xwmz

can mms_p m momm>mm an .Fwo :Loo
cw mmxxms use new mmm . mmpmsmw Amwpmsmw use
mmx\ma ANN new mmfl - ayes mm_mev mu_e Hamuom

mmm>am
An mgmmcoF .mxz mm um>cmmao
mmxmmz mm Lo» xmmz Lon was?“ m
mpwo :Lou cw mmx\me oom can ooH Ammpmsmt can mmpmsv
- umEmw mmx\me omH ecu om . mFme mum; chcmZImcconmo

Ammpmsmt
mmm>mm An mmxmmz mm com ecu mmpmev mom;
”mammazpoop cw mmx\me om xm—zmo-mzmmcam

camco_ .mxz em
vm>cmmno momm>mm >3 mmxz om Lot
mpwo mwgumcm :_ mmx\mE om wows mmv3muHuH mpmz

Kama .mecznmwaz

ommfi .Huz

ommfi .Hoz

kmmmfiumoma .mom

roxmfiumomfi .mom

 

m6_=mmm pcauwc_=mwm

Fooouoca mcvmoo Pms_c<

xuzbm

 

.mpmswcm cw mmwuzpm >y_o_cmmocwocmo Egowoco—co mo :o_pmN_LmEE:m .N mFQmH

.mmmfi .Lmnzmm Eocw co_pmscovcHs

 

museumam:

esocwocmu
gm—zmeoopmam;

mmm>mm Xn mpwo o>__o
cw mgmmco— mxmmz ¢ um>cmmno
mmxmmz ma Low mxmc Lao»
xcm>m moco mmxxms ooo oa m.~m

mecoF mxmmz NH om>cmmno
momm>mm ma mp_o :Lou :_

mmxmmz mm Low xmmz can mos?“ m
mmxxme oom op a: u mmpmsmw
mmxxms oom cu m: . mmpme

Ammpmsmw ecu mmpmsv
mo_e < :ngpm qemfi .mecmcncmgomm

Amm—m5m$ use mmesv
mu_s Hamuom Nfimfi .Lamczamvmz

 

mpysmmm pcmuwc_=mwm

Fooopocm mcwmoa

_asw=< Nuzpm

 

A.p:ouv N mpnmh

10

Trihalomethane Formation

 

Chlorine is dispersed in water as Clz, H20Cl+, Cl+, Clg,
HOCl, OCl' and Cl' (Morris, 1976). The relative amount of each

species depends upon pH. At conditions usual for drinking water
chlorination, the species of any consequence are H20Cl+, Clz, HOCl

and 0Cl‘. Based upon relative concentrations and specific reactivity
of each species to nitrogenous compounds at pH 7 and 15°C, net

relative reactivity is much higher for HOCl followed by Cl2, H20Cl+
and finally OCl'. Hypochlorous acid can dissociate to hypochlorite

ion and hydrogen ion. The dissociation constant ranges from 1.6 to 3.2

x 10-8

for the temperature range 0 to 25°C (Morris, 1966).

Hypochlorous acid concentration decreases with increasing pH with a
corresponding increase in hypochlorite ion concentration. Hypochlorite
ion is the dominant species at pH values greater than 7.8 to 7.5. By
pH 9, hypochlorite ion accounts for about 96% of aqueous chlorine
(Morris, 1976). Based upon reaction studies with nitrogenous
compounds, HOCl is 10,000 times more reactive than OCl'.

HOCl is an electrophilic agent (Morris, 1976). Reaction may occur
at either the oxygen or chlorine atom; however, because the latter is
more electropositive the reactions probably occur more often at the
chorine site.

Chlorine, when first added to water, reacts rapidly with the
reducing agents such as ferrous ion, sulfide, nitrite, and easily
oxidized organic compounds. Ammonia and chlorinated ammonia are the
next compounds to react with HOCl, followed by the organic and

inorganic compounds which are more difficult to oxidize (Johnson,

1976). Reaction of the more resistant organics with chlorine is slow

11

and depends upon chlorine concentration and pH.

The classical haloform reaction, a base catalyzed reaction of
aqueous hypohalites or hypohalous acid with simple methyl ketones to
yield trihalomethanes, has been known since 1822 (Fuson and Bull,
1934).

The reaction, as given in Figure 2, requires the enolization of
the methyl ketone by proton dissociation from the alpha-carbon
resulting in an enolate carbanion. The enolate carbanion then
undergoes electrophilic attack by HOCl or 0Cl' (Morris, 1976). After
three enolizations and electrophilic attack sequences, hydrolysis
occurs resulting in an organic carboxylic acid and a chloroform.

Bellar and coworkers (1974) proposed that ethanol is oxygenated to
acetaldehyde which subsequently reacts with free chlorine to form
chloral hydrate which then decomposes to form chloroform. The basis of
this proposal is the observed presence of ethanol in tap water.
Alcohols and amines which are oxidized to ketones are capable of
reacting in the haloform reaction (Fuson and Bull, 1934).

The rate determining step is the enolization of the ketone. The
stabilization of the enol is pH dependent, increasing with increasing
pH. At pH 7 simple methyl ketones - such as acetone, acetaldehyde and
acetophenone - do not react sufficiently to produce significant amounts
of trihalomethanes (Stevens et al., 1976). Acetone concentration in
raw water is not high enough to account for the observed THM levels.

At pH values above 8 or 9, acetone and other simple methyl ketones may
significantly contribute to THM formation (Morris,.1976; Stevens et
al., 1976). These levels are above the usual water chlorination pH

values of 6 to 8.

.AwNmH .Ezmm 8:8 m_cgoz mofimfi .mwccoz Eocev co_uommg Egowopm; Fmowmmmpo 8:8 .N mczmwd

.5 238%. A A8 I 8A 888. -8 8 + 8N=|llIIIIv:8- ..8_- a + 8828

 

 

 

 

& -18 8
A+X8~I 8mac A888 8888
+2
N - - . Sm - N - -
x28 8 a InIIIIIII.AA8mmu8- ml .-v X8-8 m 8
a :8 my 88
.8. I
R
8W8L
Amac Ax8z-III-«X8NIV
-88
zoFm N
A xz8-w-m :8-w-m VMhunnuumw 88-w-m
n88 8V -18 8
8888
A +X8N:.1IIII.X8=8 bust
+I 30pm
A N18-8-8 III-8N 8-8-8 8\ 87.8-8-8
. %: V =,

08 8-.8 -28 8

13

The presence of brominated haloforms in chlorinated water has been
attributed to two reasons. The first, as suggested by Bellar and
coworkers (1974), is the presence of bromine impurities in chlorine.
Rock (1974) noted that 100 mg chlorine gas was contaminated with less
than 0.04 mg (0.04%) bromine, and concluded that bromine contribution
by chlorine was negligible.

The second reason stems from the oxidizing potential of HOCl. The
potential is strong enough to oxidize aqueous Br' and I’, but not
F', to their respective hypohalous acids (Bunn et al., 1975). HOBr
and H01 then participate in the haloform reaction forming the
respective brominated and iodinated haloforms. After spiking raw water
with 5 mg/L bromide and 5 mg/L iodide followed by chlorination with
calcium hypochlorite (1.2 mg/ml as available chorine), the following
haloforms were detected: chloroform, bromodichloromethane,
dibromochloromethane, bromoform, dichloroiodomethane,
chlorodiiodomethane, iodoform, dibromoiodomethane, bromodiiodomethane,
and bromochloroiodomethane. Hypobromous acid reacts faster with
ammonia than does hypochlorous acid, and may also react faster with
haloform precursors (Morris, 1976). This would account for the greater
proportion of brominated haloforms present than could be accounted for
by the concentration of natural bromide in raw water (Rook, 1974).

In 1950, Booth and Saunders noted that the following groups of
compounds participated in the haloform (iodoform) reaction.

1. CH3-C-R, when R=H, hydrocarbon radical, COOH and esters,
(CH2)nC00H and esters. R cannot be 0H, substituted 0H, NH2 or
certain substituted amino groups.

0H

2. CH3-CH-R, when easily oxidized to the above compound.

14

3. Oximes, when easily oxidized to a methyl ketone.

4. Compounds containing -COCH2CO-, or -CH(0H)CH2CH(OH)- if not
connected to heavily substituted carbon atoms.

5. Quinone or hydroxyquinone with at least one unsubstituted
position adjacent to the carbonyl or hydroxyl group.

6. Meta—dihydric phenols, such as resorcinol and phloroglucinol.
Tautomerism to the keto-form may account for the reaction.

The possibility that humic substances could be precursors of
trihalomethanes during water chlorination was first proposed by Rook
(1974). Humic substances are naturally occurring organics responsible
for the yellow to brown color in some natural waters (Packham, 1964),
and are resistant to microbial degradation (Kuznetsov et al., 1970).
Humic substances arise from one or a combination of three sources -
water soluble extractions of living woody substances, degradation
products of decaying wood, and soil organic matter (Christman and
Ghassemi, 1966). Aqueous humic substances can be divided into three
fractions - fulvic acid, an HCl-soluble fraction; humic acid, an
HCl-insoluble and ethanol-insoluble fraction; and hymatomelanic acid,
an HCl-insoluble and ethanol-soluble fraction (Black and Christman,
1963). Weber and Wilson (1975) suggest that the non-acidic ester group
of humic acid breaks down by hydrolysis into an aromatic acid and an
alcohol containing a carboxyl group, thus forming fulvic acids. This
suggestion could explain the higher molecular weight of humic acid,
greater number of alcohol hydroxyl groups in fulvic acids, and the
similar phenol hydroxyl values of both acids.

Both humic and fulvic acids react with free chlorine to produce

haloforms. Commercially available humic acid, when suspended in water

15

at pH 7 or dissolved at higher pH then readjusted to pH 7, had THM
formation rate curves similar to those observed during chlorination of
natural waters (Stevens et al., 1976). At both pH 6.7 and pH 9.2, the
haloform reaction was complete after 100 hours. 0f the available
carbon in humic acid, 0.7% and 1.4% react forming haloforms at pH 6.7
and pH 9.2, respectively (Stevens et al., 1976). The increase in THM
yield with increasing pH, may be explained by the presence of reactive
sites on humic acids, normally unreactive at neutral pH, becoming
reactive at higher pH values.

Yields of haloforms from chlorination of fulvic acids vary from
0.3% to 0.9% of available carbon depending upon chlorination conditions
(Rock, 1976). After chlorine (870 mg/L) was in contact with fulvic
acid (250 mg/L) as total organic carbon (TOC) for four hours at 10°C,
a large increase in THM concentration was noted at pH values between 8
and l0. The reaction was less affected by a pH change in the range
from 6 to 8 (Rook, 1976).

Oliver and Lawrence (1979) and later Oliver and Visser (1980)
reported a small difference for THM formation between humic acids and
fulvic acids. The reactions were carried out at pH ll and 20°C for 72
hr. The chlorine concentration was 15 mg/L and humic material
concentration was 1 mg/L as TOC. Under different reaction conditions
(pH 6.5, 20°C, 100 hr, chorine dose of 10 mg/L, and humic material
concentration of 2 mg/L as TOC) haloform yields from fulvic acids was
less than half that of humic acids (Babcock and Singer, 1979).

Ultimately fulvic acids may be the major source of THM precursors.
From evaluation of ten 20 or 40 liter samples of raw water, source or

sources unknown, fulvic acid was determined to represent about 87% of

16

the humic substances. Hymatomelanic acid and humic acid account for
only 12% and 1%, respectively, of the humic substances (Black and
Christman, 1963). These findings agree with Packham (1964) who found
fulvic acids to be the major organic constituent of colored waters.

The molecular structures of humic substances are unknown; however,
various models have been proposed. The first model, as described by
Kuznetsov (1970), is for humic acid. The humic acid molecule may
contain an aromatic nucleus with cyclical and chain-form organic
nitrogen and carbohydrates covalently bound to it.

The second model is for fulvic acid, and consists predominantly of
an aliphatic or alicyclic backbone (Anderson and Russell, 1976). A
polymaleic acid model was proposed with unsaturation and vicinal
carboxyl groups present. The polymer also contained phenolic
derivatives, polysaccharides, ammonium ion and amino acid residues,
but these groups were determined to be of less importance.

Khan and Schnitzer (1972) determined, contrary to the findings of
Anderson and Russell (1976) that the major oxidation products of humic
and fulvic acids were predominantly aromatic, that is, aromatic
carboxylic acids and phenolic acids. Aliphatic carboxylic acids were
also found but at a much lower concentration. Methylation of humic
substances resulted in degradation products that could be isolated.
Since methylation is known to reduce hydrogen-bonding, Khan and
Schnitzer (1971) proposed that humic substances consist of ”phenolic
and benzene carboxylic acids joined by hydrogen-bonds, on which
alkanes, fatty acids, and dialkyl phthalates are adsorbed."

Various degradation products of fulvic acid and other humic

substances have been isolated (Christman and Ghassemi, 1966).

17

Chlorination studies have been performed on many of these degradation
products. The results of some of these studies are given below.
Diketoalicyclic rings give a positive haloform reaction when the
keto groups are beta to each other (Rook, 1976). Dimedone,
1,3-indandione and 1,3-cyclohexanedione are examples of such compounds.
Like methylketone THM yield increases with increasing pH values because

alkaline conditions favor the enolization reaction

-9- c -g-+——r-c=c-c-

0 H2 0 OHli()

thus THM formation.

Resorcinol (meta-dihydroxybenzene) also reacts in the haloform
reaction (Rook, 1977). Because hydroxyl groups are ortho and para
activators, they doubly activate the carbon atom to favor formation of
a carbanion and electrophilic halogenation. Rook (1977) proposed a
reaction mechanism (Figure 3) based upon the observation by Moye (in
Rook, 1977) of pentachlororesorcinol as a product of the chlorination
of resorcinol with nonaqueous chlorine.

Resorcinol, as found in natural waters, may be substituted by R1,
R2, and R3, which could be the fulvic acid matrix once or twice, H,
OH, OCH3 or COOH. Chlorination of the activated carbon (1) is rapid.
The intermediate carbanion (II) can be rapidly protonated or
halogenated to compounds III and IV, respectively. Hydrolytic or
oxidative cleavage of compounds III and IV is responsible for most of
the known degradation products of fulvic acids. Cleavage occurs along
dotted lines a, b and c. The resulting products, depending upon the

groups present at the R positions, include chloroform, methylene

18

 

0H 0 C)
c E Zz’OH
R c/ \H HOCl R c/ bC/C]
l“ I g In I\C] I
RIC C-OH 1n H 0 ' R c c=o
2 \c// 2 2 \C/
3 R§// “\c1
oHC’

Figure 3.

 

 

 

 

c c=0
Rz/ c/
‘_ R3/ \c1 4
H+ Cl+
11
R /CO0H COOH
l‘c R1"?
R/C R/p r‘ | gl'Pl H+
I I I 2 ’1 : l C]
2 l - ' ’ _‘_ /'
z . L I
C] i 0 i 3 f 1 Cl
3 ' - b 'a
b a
111 1"

Proposed degradation pathway of fulvic acids and
resorcinol. Cl+ represents in a simplified way any
electrophilic halogenating species of the series
XOH+, H , HOX, X 0 where X = I, Br or Cl (from Rook,
197 ). 2 2

19

chloride, tetrachloroacetone, pentachloroacetone, hexachloroacetone,
chloral, dichloromaleic acid and trichloroethylene.

The presence of a stable chlorinated cyclopentene, such as
3,5,5-trichlorocyclopent-3-ene-1,2-dione, upon chlorination of
resorcinol, was detected by Christman and coworkers (1978). The
quantity of this compound increased as resorcinol decreased, and along
with chloroform appeared to be the major product of chlorination. At
molar chlorine to carbon ratios below the optimal ratio for chloroform
production (1.14), cyclopentene exists. As the reaction endpoint of
1.14 ClZ/C is approached, the cyclopentene begins to disappear and is
gone when the endpoint exists. The formation of stable cyclopentene
does not agree with the postulated reaction mechanism of Rook. The
product does, however, confirm Rook's suggestion that the carbon
between the two hydroxyl groups becomes the chloroform carbon atom.

Alkaline conditions also increase the reaction rate by favoring
phenoxide ion formation (Rook, 1977) which is a stronger activator than
is an aromatic hydroxyl group. The presence of the phenoxide ion may
be why THM formation rate increases with increasing pH values.

Substitution of the meta-dihydroxybenzene compound with various
groups can alter chloroform yield (Norwood et al., 1980) by altering
the activation of the carbon ortho to the two hydroxyl groups. At pH 7
and 25°C, chloroform yield fromresorcinol, or a molar basis, was 88%.
The addition of a methyl group at the position meta to the two hydroxyl
groups did not significantly affect chloroform yield (85%). The
replacement of the methyl group with a carboxyl group, however, reduced
chloroform yield to 45%. Meta-dimethoxybenzene, a diester, produces

much less chloroform since keto-enol stabilization reaction does not

20

occur as readily. The chloroform yield for 3,5-dimethoxybenzoic acid
was 0.9%. Unsaturated alkyl side chains tend to increase chloroform
yield. The presence of a hydroxyl group on the carbon inbetween the
two hydroxyl or methoxy groups greatly decreases chloroform yield by
interfering with keto-enol stabilization. For example, ortho- and
para-dihydroxybenzenes yield much less chloroform than
meta-dihydroxybenzene, 0.5%, 1.5%, and 85%, respectively at pH 7,
because double activation of any carbon does not occur for the ortho-
and para-compounds (Rook, 1977). Chloroform yield for both these
compounds increases somewhat at higher pH values, but not to the 100%
yield observed for meta-dihydroxybenzene.

Quinones and compounds oxidized to quinones, such as tannic acid,
could undergo degradation by a pathway similar to that of resorcinol as
proposed by Rook (in Youssefi and Zenchelsky, 1978).

Citric acid is another possible precursor which has been found in
natural water (Bjork, 1975) and in waste water (Afghan et al., 1974).
Unlike most proposed precursors, chloroform yield from citric acid
reaches a maximum at pH 7 rather than at more alkaline conditions
(Larson and Rockwell, 1978). The proposed reaction mechanism is given
in Figure 4. According to this scheme, the rate limiting step would be
the oxidative decarboxylation of citric acid. If this reaction
mechanism requires the stabilization of the citrate trianion, the pH of
maximum reaction would be halfway between the pH of maximum trianion
concentration (pKa 6.40) and hypochlorous acid concentration (pKé
7.49), or pH 6.94, a nearly neutral pH (Larson and Rockwell, 1978).

The chlorination of amino acids leads to the formation of

dichloroacetonitrile (Maugh, 198lb) which decomposes to form

21

 

 

 

 

 

 

 

_ CH2C00 HOCl
00C-——C<r—0H )
CHzcoo‘
co2
.. + _
cc1 coo HOCl CH coo
o=c/’ 2 _ -a_2. o=c" 2
\‘CH coo \CH coo‘
2 2
HOCl
If //cc13
o - c _ + co2
\‘CHZCOO
H20
" ’/,CO0H
CH2 _ + CHCl3
‘~coo

Figure 4. Proposed mechanism of CHCl production from
citric acid (from Larson and Rockwell, 1978).

22

chloroform. Amino acids occur in natural waters.usually associated
with the humic material at concentrations equivalent to about 1% of the
carbon fraction (Lytle and Perdue, 1981).

Organic compounds containing the nitrogen-bearing pyrrolic ring,
such as chlorophyll, produce chloroform under increased alkaline
conditions (Morris and Baum, 1978). Like those in phenol, the hydrogens
alpha to the nitrogen are activated and become the site of
electrophilic halogenation. Significant haloform formation occurs only
at elevated pH.

.Algae upon chlorination yield CHCl3, though not to the extent of
algal by-products (Hoehn et al., 1980). Algal extracellular products,
such as D-mannitol (Crane et al., 1980), yield THM‘s upon chlorination
(Hoehn et al., 1980). From studies with extracellular products, it was
found that as much chloroform per unit carbon was formed as from humic
and fulvic acids, and the THM yields are greater under more alkaline
conditions than at neutral pH (Hoehn et al., 1980). Due to the ease of
bacterial degradation, algal by-products may not exist very long (Hoehn
et al., 1980), thus their contribution to THM formation may be small.

Therefore, a large variety of compounds normally found in natural
waters can react with chlorine to form haloforms. Not all the
aforementioned compounds need to be associated with the humic fraction
because fulvic acid and related compounds appear to be the major THM
precursors. The contribution of each compound depends upon its
relative concentration and reactivity as well as the chlorination
conditions such as chlorine concentration and pH. These same
conditions also determine which other halogenated compounds are

produced (Morris and Baum, 1978; Norwood et al., 1980).

23

THM Level Regulated

 

Based upon the Safe Drinking Water Act (Public Law 93-523), an
amendment to the Public Health Service Act (U.S. Congress, 1974), the
EPA has the authority to establish the National Primary Drinking Water
Regulation for those chemicals the EPA administrator determines may be
harmful to human health. Causative proof of adverse effect is not
necessary to establish a regulation. Since many organic contaminants
are present at only trace concentrations, this factor is important. A
significant health risk may exist since exposure is chronic. Adverse
human health effects are determined from epidemiological and
toxicological evidence (U.S. EPA, 1978). The Safe Drinking Water Act
requires the EPA to set regulations which protect human health to the
maximum level using economically feasible treatment methods available
in I974.

On November 29, l979, EPA amended the National Interim Primary
Drinking Water Regulations by establishing a Maximum Contaminant Level
(MCL) of 100 ug/L for total trihalomethanes (U.S. EPA, 1979). This
regulation affects community water supply systems servicing 10,000 or
more people.

Within one year of the date of promulgation, water suppliers
servicing 75,000 or more people are required to monitor THM levels for
one year. The following year each system must comply with the set MCL.
Three years after promulgation water suppliers servicing 10,000 to
75,000 people must monitor THM levels for one year. The following year

the established MCL must be complied with.

24

The Study

Due to the health concerns and EPA's regulations, the amount of
trihalomethanes present in Michigan's public water supply is of
interest. A survey was conducted of 40 water treatment plants on a one
time basis. The results are compared to those of the National Organics
Reconnaissance and National Organics Monitoring Surveys, and are I
examined for a possible geographic distribution pattern. Since
treatment conditions affect THM yield, results were analyzed to

determine if THM yield is related to water source and chlorine dose.

 

METHODS AND MATERIALS

 

Treatment Plant Selection

An index summary and addresses of treatment plants were supplied
by the Michigan Department of Public Health (1979, 1980). From this
supplied information 173 treatment plants were selected as possible
participants because water chlorination was practiced, and when
supplied by other treatment plants, the intake water was not already
chlorinated. Each of the 173 treatment plants were sent a letter
(Figure A-l) requesting their cooperation in the survey and a
preliminary questionnaire (Figure A-2) requesting information about
their chlorination procedure. The 40 treatment plants that agreed to
participate were sent a sampling kit, sampling instructions (Figure

A-3) and a second questionnaire (Figure A-4).

THM Sampling and Extraction

Water sampling and organohalide extraction are given in Figure
A-3. Organohalide extraction was a modified liquid-liquid extraction
method of Mieure (1977) and Richard and Junk (1977) developed by
Kaczmar (1979). Fisher's certified methylcyclohexane lot number 792990

was used to extract the organohalides from the aqueous solutions.

25

26

Analytical Procedure

 

Instrument: Varian 1800 Gas Chromatograph

Detector: Tritium foil electron capture

Standing current: 1.184 x 10'8 amps

Column: 2 m x 3 mm i.d. nickel column packed with 10% squalane

on chromosorb W—AW
Column temperature: 80°C, isothermal
Injection: On-column (injection precision curve for chloroform-
Figure A-5)

Injection port temperature: 200°C

Detector temperature: 165°C

Carrier gas: nitrogen, head pressure of 32 psi

Sample size: 2 ul

Mixed standards were obtained from Environmental Monitoring and
Support Laboratory, U.S. Environmental Protection Agency and were
called Water Supply Quality Control Check Samples for Volatile
Organics. Stock solutions containing chloroform (1369.2 ug/L),
1,2-dichloroethane (544.2 ug/L), 1,1,1-trichloroethane (223.8 ug/L),
trichloroethylene (379.4 ug/L), carbon tetrachloride (25l.4 ug/L),
tetrachloroethylene (175.2 ug/L), bromodichloromethane (238 ug/L),
dibromochloromethane (343 ug/L) and bromoform (284.8 ug/L) were made
up monthly in methylcyclohexane. Working standards were made by
serially diluting stock solution with methylcyclohexane. Eluted peaks
were identified by comparing elution time with that of individual
organohalides obtained from RCK-OOI Organohalide Kit Lot Number I-139

(Analabs, Inc., New Haven, CT). Elution order is given in Figure A-6.

27

Statistical Analyses

Analysis of covariance and Student's t test were used to determine
significance of chlorine dose on total trihalomethane levels. One-way
analysis of covariance and Bonferroni's t-test were used to determine
the effect of water source on total trihalomethane levels after
chlorine dose effect was accounted for. When chlorine dose effect was
not accounted for, analysis of variance and Bonferroni's t-test were
used to determine the effect of water source on total trihalomethane
level (Gill, 1978a; Gill, 1978b; Gill, 1981).

An attempt was made to develop a model for predicting TTHM levels
in drinking water using a computer program for multiple regression (SAS
Institute Inc., 1979). Four approaches were used to analyze the data:
stepwise forward, stepwise backward, stepwise minimum, and stepwise

maximum.

RESULTS

General Results of Survey

 

Information from questionnaires - including sampling date, water
source, chlorine dose, residual chlorine concentration, water
temperature, pH, turbidity, color, and treatment procedure - along with
trihalomethane survey results are given in Table B-1 (parts I and II).
Total trihalomethane levels, in the finished water of 40 Michigan
drinking water plants, ranged from undetectable to 281.8 ug/L with a
mean of 44.0 ug/L and a median of 20.5 ug/L.

The mean, median, and range concentrations of the four individual
trihalomethanes detected in finished water are given in Table 3. Both
range, median, and mean values of individual trihalomethanes decreased
with decreasing chlorine atom presence in each trihalomethane;
chloroform was present at the highest concentrations followed by

bromodichloromethane, dibromochloromethane, and finally bromoform.

28

 

29

Table 3. Mean, median and range of trihalomethane levels found in
finished water of forty Michigan drinking water plants.

 

Concentration (ug/L)

 

 

Contaminant Mean Median Range
CHCl3 33.4 16.7 UD*-201.4
CHClZBr 6.4 2.7 U0 - 54.2
CHClBr2 4.1 2.2 U0 - 39.6
CHBr3 0.1 UD*** UD - 1.6
TTHM** . 44.0 20.5 UD -281.8

 

*UD, undetectable
**TTHM, total trihalomethane
***Only three treatment plants had bromoform detected in finished
waters.

Geographic Distribution

 

The trihalomethane levels in the finished waters of the surveyed
Michigan water treatment plants are given by location of treatment
plants on an outline map of Michigan (Figure 5). Treatment plants
located along the shores of the Great Lakes and connecting waters
tended to have higher trihalomethane levels than those located inland.
Four treatment plants located inland from the Great Lakes had
trihalomethane levels in finished water comparable to those of
treatment plants located along the Great Lakes. These four treatment

plants received river or inland lake water.

 

30

Figure 5. Geographic distribution of levels of four trihalo-
methanes found in finished water of 40 Michigan water
treatment plants.

31

 

 

 

 

   

 

 

NA81818

IX/s/I

HE83 88<8m

A8Nemz8

”filo.-

32

Effect of Water Source andChlorine Dose on TTHM Levels

 

The results of statistical analyses are given in Appendix C.

Total trihalomethane level in finished water is significantly affected
by the quantity of chlorine applied to raw water (Figure C-2). The
higher the chlorine concentration, the more the TTHM level tends to
increase.

Chlorine dose effects were accounted for before determining the
effect of water source on TTHM levels (Figure C-3). 0f 35 treatment
plants surveyed, five received water from a river or inland lake source
(surface water), 17 received water from the Great Lakes and connecting
waters, and the final 13 plants received groundwater (Table C-l). Five
treatment plants were not used for statistical analyses since chlorine
dose information was not available for these plants. Finished water
obtained from surface water had TTHM levels significantly different
from those of Great Lakes (P < 0.01) and groundwater (P < 0.01)

(Figure C-4). Treatment plants using surface water tended to have the
highest TTHM levels.

TTHM levels were not significantly different when the water source
was from the Great Lakes and/or connecting waters or from groundwater
(Figure C-4). When the quantity of chlorine applied to raw water was
assumed to have no effect on TTHM levels, mean TTHM levels in finished
water obtained from Great Lakes or connecting waters was significantly
different from that when groundwater was the source (P < 0.05) (Figure
C-5). When the Great Lakes was the water source, TTHM levels were

higher than levels from groundwater.

33

Predicting Trihalomethane Levels

 

A model for predicting trihalomethane levels in drinking water was
developed using multiple regression. The independent variables used
were those most often supplied by the treatment plant operators:
chlorine dose, hydrogen ion concentration, turbidity, and temperature.
Only 26 plants supplied data for all four variables.

The models developed are given in Table D-l. Only two variables,
chlorine dose and hydrogen ion concentration, helped explain the
variability in TTHM levels observed in the data. When chlorine dose is
the only independent variable used, the model is

[TTHM] = - 5.74 + 23.08 (chlorine dose).
The model has a Fisher's exact test value of 96.12 which is highly
significant at p < 0.0001. This model explains 80% of the variability
found in the TTHM levels.

Alone, hydrogen ion concentration does not significantly explain
the variability in TTHM levels. When added to the model with chorine
dose --

[TTHM] = - 13.5 + 22.24 (chlorine dose) + (2.38 x 108) HCON
-- the model explains an additional 3% of the variability in TTHM

levels for a total of 83%. This model is significant with a Fisher's
exact test value of 57.06 and p < 0.0001. The addition of either of

the two remaining variables, turbidity and temperature, added less than

1% to the percent variability explained.

DISCUSSION

 

The median and range TTHM values for the Michigan survey (Table 3)
are lower than both the NORS and NOMS values (Table 1). Many possible
reasons exist for this difference; two, however, are particularly
relevant. First, water sampling for the Michigan survey took place
from November through March when leSs trihalomethanes are formed due to
a combination of lower water temperature and less organics available
for reaction (Stevens et al., 1976; Young and Singer, 1979; Otson et
al., 198l). Sampling for the NOR survey took place from late January
to late April (Symons et al., 1975). The time period is about the
same, but the NORS sampling often occurred after the spring thaw.
Second, the NOR survey analyzed proportionally more surface waters and
less ground- and GreatLake5\maters than this survey. Surface waters,
as will be discussed later, contain more organics available for THM
formation than do ground- or Great Lake waters.

In addition to the trihalomethanes, two alkylhalides, carbon
tetrachloride and tetrachloroethylene, were detected in chlorinated
water (Table B-l, part II). While both compounds were infrequently
detected in raw and finished waters, in some finished waters where
total trihalomethane levels exceeded 60 ug/L, the concentrations of
these compounds were often greater than in raw water. In the NOR
survey carbon tetrachloride and 1,2-dichloroethane, but not

tetrachloroethylene, were detected in a small proportion of chlorinated

34

ll

 

 

35

waters (Symons et al., 1975); however, due to 1,2-dichloroethane and
1,1,1-trichloroethane eluting together, 1,2-dichloroethane could not be
quantified in this survey.

No one factor determines the amount of trihalomethane ultimately
formed during chlorination; however, chlorine concentration and
concentration of organics seem particularly important. Surface water,
generally, has a higher total organic carbon content than does
groundwater (Symons et al., 1975). Oliver and Lawrence found levels of
THM precursors to be "relatively low" in Great Lakes water (in Anon.,
1977). Differences in total organic carbon content between the three
sources of water could account for the observed differences in the
amount of trihalomethanes formed. According to the results, however,
the difference between THM formation in Great Lakes water and
groundwater may be due more to chlorine dose than to organic load,
because TTHM levels significantly differed only when mean TTHM values
were not adjusted for chlorine dose effects (Figures C-3, C-4, and
C-5).

The observation that chlorine dose affects THM yield agrees with
the findings of other researchers (Zogorski et al., 1978; Moore et al.,
1979; Vajdic, 1080; Otson et al, 1981). Otson and coworkers (1981)
noted that chlorine dosages and chlorine demands were the dominant
factors influencing THM yield when river water was used as the source.
A variable amount of the chlorine dose is used to satisfy the demand.
The remaining free residual chlorine is available to form
trihalomethanes. Therefore, assuming total organic carbon is not
limiting, free residual chlorine should influence THM formation more

than chlorine dose would. Residual chlorine values are given in

36

Table B-1, part I. Unfortunately water suppliers often did not specify
whether the values were free, combined, or total residual chlorine. No
statistical analysis could be made.

A general description of water treatment procedure for each
treatment plant is given in Table B-1, part 11. Treatment procedure
can influence THM formation; however, no conclusions could be drawn
from the supplied information. Treatment conditions which affect THM
yield are the presence of lime water softening by raising pH,
recarbonation of softened water which lowers pH, placement and type of
filtration, oxidation, and disinfection, and finally chlorine contact
time (Zogorski et al., 1978; Kavanaugh, 1978; U.S. EPA, 1979; McGuire
and Suffet, 1979; Brodtmann and Russo, 1979; Norman et al., 1979).

Of the models developed to predict TTHM levels, the two variable
model containing chlorine dose and hydrogen ion concentration was the
best. It explained 83% of the variability in TTHM levels generated in
this survey. However, as a predictive tool, the model is inadequate.
To be of value the model should explain nearly 100% of the variability
(Geisy, 1982).

The four variables used, chlorine dose, hydrogen ion
concentration, turbidity, and temperature, were the variables most
often supplied by treatment plant operators. A model using variables
available to plant operators would be useful. A model has been
developed using those variables known to influence THM formation (Moore
et al., 1979).

From this model alone, chlorine dose and hydrogen ion
concentration cannot be said to directly influence THM formation

because interrelationships with unmeasured variables are unknown

37

(Geisy, 1982). However, the results do support the previous
statistical test showing the importance of chlorine doSe in THM
formation. 0f the unmeasured variables, total organic carbon would
probably be the most important. Unfortunately no treatment plant
operator supplied data for this variable.

Five of the 40 plants had TTHM levels above the federally set MCL
of 100 ug/L (Table B-1, part II). Analyses were done at one point in
time, during winter when trihalomethane concentration is expected to be
at its lowest. At other times, more favorable to THM formation, such
as in spring, summer or fall, temperature and organic load increase and
chlorine dose is normally increased, more plants producing potable
water with TTHM values about 100 ug/L would be expected.

One hundred and seven Michigan plants, including 13 of the 40
plants surveyed (Michigan DPH, 1979) will be affected by the limit set
by the EPA. The population served and TTHM levels measured in the
finished water of the 13 affected plants are given in Table 4. TWO of
the thirteen plants had TTHM levels above 100 ug/L at the time of
sampling.

This survey does not statistically represent trihalomethane levels
present in finished waters of Michigan water treatment plants. To be a
statistically representative survey, a portion of those treatment
plants which did not agree to participate would need to have

trihalomethane analysis of finished water (Gill, 1981). Of the 951

38

Table 4. The thirteen surveyed treatment plants which are affected by
the EPA regulation, and their population size served and
TTHM levels found in finished water.

 

 

WSSN Township Served Population* TTHM, ug/L
0710 Big Rapids 12,000 227
0040 Adrian 20,382 151
0160 Alpena 14,000 68
0600 Benton Harbor 16,481 57
2890 Grosse Pointe Farms 11,700 49
4570 Muskegon 44,631 47
2170 Escanaba 15,100 46
4580 Muskegon Heights 17,304 35
5480 Port Huron _37,500 25
5950 Sault Sainte Marie 15,000 15
4340 Michigan State University 50,000 11
3760 Lansing 131,546 4
5520 Portage 25,000 ND**

 

*Population size obtained from Michigan DPH, 1979
**N0, none detected

39

Michigan water system suppliers, 488 chlorinate their water (Michigan
DPH, 1979). The 40 treatment plants which participated in the survey
account for 8.2% of the Michigan treatment plants which practice
chlorination. However, this study does indicate the THM levels of
those plants surveyed and could be used as a beginning point for an
expanded survey. In addition, trends were noted regarding the

influence of water source and chlorine dose on TTHM levels.

SUMMARY

Total trihalomethane levels, in the finished water of 40 Michigan
drinking water plants, ranged from undetectable to 281.8 ug/L with
a mean of 44.0 ug/L and a median of 20.5 ug/L. Range and median
values are lower than those of the NOR and NOM surveys. Chloroform
was usually present at the highest concentrations followed by
bromodichloromethane, dibromochloromethane and bromoform.

At the time of sampling, five of the 40 plants had TTHM levels
above the EPA MCL of 100 ug/L, of which two would be affected

by the EPA regulation.

Chlorine dose influences the amount of TTHM formed. After
adjusting for this influence, treated drinking water supplies
derived from surface water had significantly higher TTHM levels
than either GreatLakescw'groundwater. TTHM levels of Great Lakes
and groundwater were significantly different only when chlorine
dose effect was not accounted for, perhaps because the organic
levels are low compared to surface water. However, when
chlorinated, Great Lakes water tended to have higher TTHM levels

than did groundwater.

4o

APPENDICES

APPENDIX A

Methods and Materials

41

Figure A-l. Cover letter for preliminary questionnaire.

42

MICHIGAN STATE UNIVERSITY

 

m or WATER MIC“ FAST LANSING ' MICHIGAN - 0814
NATURAL was mm

April 29, 1980

Dear Sir:

To determine the effect of water chlorination on the haloform content of
finished drinking water, I would like to enlist your cooperation to answer the
enclosed questionnaire and send water samples to our laboratory for analysis.
In return for your cooperation, I will be happy to send you your individual
results and a compilation of the total data. Obviously, the more participation
we have, the more comprehensive the study will be.

The first step is to answer the enclosed questionnaire which has a pre-
addressed, postage-paid return envelope. If you signify that you are also
willing to send the water samples, you will then be sent the implements and
simple instructions. It will require very little time to fill the six sample
vials that will be enclosed with a syringe, needle, and pre-addressed, postage-
paid return container.

The sampling procedure entails placing 5 ml of water in each vial and
shaking it for one minute. Three vials are for raw water and three for finished
drinking water.

Your assistance in this survey is much appreciated. It will give you more
information about the quality of your own drinking water and also will enable
us to learn more about the overall quality of drinking water in the state of
Michigan.

If you have further questions, you can contact me or Eileen A. Nickerson,
a graduate student who is working on this project, at either of the following
telephone numbers: (517) 353-3742 or (517) 485-8188.

Thank you for your time and cooperation.

Sincerely yours,

Frank M. D'Itri
Professor of
Water Chemistry

tjw

Figure A-I
' Enclosure

43

QUESTIONNAIRE
WATER SOURCE
1. Source of raw water? '
Ground Stream or River
Great Lake Lake or Pond _ Other

Name of Source:

 

Location:

 

2. Is this source shared with any other water treatment plants?
Yes No

Plant names(s):

 

 

 

 

bl

What type of intake valve is used? (e.g., buried intake valve):

 

4. If known, what is the organic content of the raw water?
Dissolved organics
Paiticullce crganics

Total organics

 

5 Is there any industry, agriculture, sewage outfall, chemical dump near the
"water source?

Yes No

Name(s):

 

 

 

 

PREVIOUS STUDIES

6. Has any o:ganohalide, in particular halomethane, study ever been performed at
the treatment plant before?
Ye: _____ No _____
Briefly, what were the results, or where can the results be obtained?

 

 

/. Has the organic content of the water ever been measured before?
Yes No

Brief y, what were the results, or where can they be obtained?

 

 

 

 

Figure A-2. Preliminary questionnaire for determining which treatment plants
may be of value for THM analysis and individual treatment conditions.

 

44

Questionnaire

Page 2

8. Would you be willing to send water samples to our laboratory for haloform
analysis?
Yes No

 

TREATMENT PROCESS

9. On a separate sheet of paper, question 6 is presented in the form of a chart.
The following may aid you in completing the chart.

 

A. Columnl.- Sequence the treatment steps listed in Column 2 from the time
the raw water enters the plant until it is ready for diStribution. Please
mark the first treatment step one (1), the next step two (2), and so on.

8. Fill in the other columns with the apprOpriate information or NA if it is
not applicable.

C. If the answer to a question is not known, please write "Unknown" in the
proper block.

10. What is the chlorine breaking point?

 

11. How much residual chlorine and organics remain in the finished water?
Residual chlorine:

Total organics:

 

 

12. How much water is treated per day or per year?

 

Figure A-2. (Cont.)

 

45

A.s:888

 

 

.N-< 8288AL

fiIllrllIIs--IIIIJ

fi-I-I "I I‘II In 0.1.1

 

III III g'lIl-I I

mos“:

 

mo;8c

 

 

newcouuom

 

 

~O>DEOZ —NUQZ

 

coauoob:_m_c
\zoduccmxo

 

:oA88t8AAl

 

:omaouzosmwom

 

 

 

 

 

_c::...:=< as;
:1Am oaAmozzc

 

of: v. E2559

 

 

 

 

 

I II IIIIIIIII

 

 

 

 

 

 

comma—sunou
acosueouuoc;
III Y I I .|Il'lllil .|.|l III:
7:78.588 FE:OEE< . .uuv cog—i 3:55. 0...:ch— EF. .5332: A... 28 3:8 safe: .1er 8.8.85.2
on? :_ .:omom; ou< 5.8 =cAuauu mash “AaUAEozu m. 88:8:ou Incoeeob coma: uzoeacowh oueozvom
v:e mAeuAeogu coma -couzou o_:euuo x52 x:< ace
Ioou=AmA= ;~_3 you.» oath u:oeaaoch
Icou zquz mAaUAEozu
LII-III I IYIIIIIIIIIIlIIIIIIIIIIITIIII lll'lls-lilll'IL-Illlv“-llIlll.lla llillllllll

.o no.8mozo

 

46

Figure A-3. Sampling and extraction instructions sent to water
treatment facilities.

47

SAMPLING AND EXTRACTION INSTRUCTIONS

List of Materials:

6 or 9 vials (pro-labeled and containing the extrant)
l syringe

l needle

scotch tape

1 plastic bag

1 envelope

mailing label

mailing tape

return postage

Directions: (Read carefully before you begin.)

1.

To minimize contamination, raw water should be sampled before finished
water. Select the three vials labeled Raw 1, Raw 2, and Raw 3, and
remove the protective tape from around the cap. Attach the needle to
the syringe, and then rinse the unit by drawing raw water fully into
and then expelling from the syringe three times.

Draw 5 m1 of raw water into the syringe
and place the water in vial Raw 1. Let
the water run down the side of the vial
to minimize the volatization of the
trihalomethanes (see Figure 1). Cap
the vial and shake it for 60 seconds.

Repeat step two for the vials labeled
Raw 2 and Raw 3.

In some instances, additional water sampling is necessary befbre
sampling of chlorinated water. If you have vials labled Pre-l,
Pre-Z, and Pre-S, you have been chosen to participate in this step.
The water to be sampled:

 

 

Rinse the syringe as before, except rinse it 5 times using this water.
Remove the protective tape from the vials labeled Pre-l, Pre-Z, and
Pre-S. Using this source of water, repeat step 2 for each of the
three vials.

Proceed to finished water sampling. Select the three vials labeled
Final 1, Final 2, and Final 3, and remove the protective tape. Rinse
the syringe and needle 5 times. Using finished water, repeat step 2
for each of the three vials.

In some instances, 3 control vials containing the extrant have also
been sent. These vials are labeled MC-l, MC-Z, and MC-3. If you

do 233 have these vials, proceed with step 6. Vial MC-l should not
53 uncapped, it is a contamination control. At the time of sampling
for raw water, vial MC-Z should have its protective tape removed and
then uncapped. Recap the vial and shake it for 60 seconds. Vial
MC-3 should be treated as was vial MC-Z, except during sampling of
finished water. '

Figure A-3

10.

48

When the sampling is finished, make certain the caps are tight and
wrap scotch tape around the cap end of the vial such that the cap
and some of the vial are covered.

Replace the vials into the plastic bag and the completed questionnaire
into the envelope. Place both the envelope and the bag into the box.
Reseal the box with the supplied mailing tape. Cover the old mailing
address with the new label and place the supplied postage on the box.

Discard the syringe and needle.

Thank you.

/—

 

 

 

50

QUESTIONNAIRE

Name and address of treatment plant:

 

 

 

Time of sampling: Raw: AM PM / /

 

 

Finished: AM PM

 

Retention time between chlorination and sampling of finished water:

 

pH:
Raw Water:

 

At Time(s) of Chlorination:

 

Final Water:

 

Temperature:

Raw Water:

 

At Chlorination:

 

Final Water:

 

An indication of organic lode:

Method:

 

Measure: Raw Water:

 

At Chlorination:

 

Finished Water:

 

(Examples of methods: turbidity, colorimetry, organic carbon,
Biological Oxygen Demand, Chemical Oxygen Demand.)

Chlorination:

Chlorine Dose:

 

Amount of Residual Chlorine:

 

Is any method used to reduce volatile organics following chlorination
(e.g., aeration, activated carbon)? YES NO

If Yes, method:

 

 

Figure A-4

51

Figure A—5. Precision curve of chloroform from injection of a
2 p1 mixed standard into gas chromatograph (mean
i standard deviation of 3 injections).

 

52

50 -

40 —

peak area, sq. mm
L»)
O
l

N
O
l

 

 

l I I If

0 20 4O 60 80 100
chloroform concentration, ug/L

Figure A-5

 

53

.vcmvcmpm vmxre mu__m;ocmmco co Lmvco :o_p:_m

.m-< mismca

54

cmzu

N

 

Fuouu

m

—u

Fu

N

©-< mismam

Lqu

 

_UIQHUN.U

m

 

u

 

 

 

 

 

czocx
IE:

 

 

 

 

_uo
Form

Fucmzu

 

APPENDIX B

Survey of Michigan Water Treatment Plants

 

55

 

 

-- mN.o _.a o.m Fa=_u
-- m.a a.~ o.m azaa m.F a.m coca: axaa oa\ap\__ aao: “tea oaam
o a_.o a.a o.m xFa=_a Asaa aaacaehv
N a._ F.a a.m azaa a._ mN.m coca: axaa oa\m_\_. acaa_< oafio
-- -- a.“ o.a Pagaa
-- -- M.“ o.m caaa .oa a.o
-- -- m.a o.m 3am N.F.xca.o .ta a.N ta>_a =a_a=H _a\am\m aaacaaccaz oaoa
_ amH.o a.“ o.m x_a=aa Aaoz aa
aa aa.m a.“ o.m 5.xata .oa a._ aam a_aa_av
aH a~.m m.a o.m 3am a.o .ta o.m :aaa;a_z axaa Ha\am\m acaaaaa_a oaam
-- HH.o m.a a.a xpacaa coa_aa
-- N.a a.“ o.“ 5.xaaa m.N c a_oaaaa
-- N.a a.“ 0.“ 2am m.~ a a.a aaam axaa oa\a_\__ _Eaaa-oa_t=a om_N
-- mm.o m.a m._F x_a:_a
-- N.F m.o m.oH Eaaa xaaaa c_oz
-- _.a F.a m.o_ azam _.N aa.m co :a_aa< axaa oa\mfi\ua cacta< oaoo
-- ~.o a.a o.N apac_a
-- o.~ a.a o._ 3am “.0 a.m ta>_a coaaxaaz Ha\am\m au_aaa a_a oHNc
-- a.o _.N a.“ Fa=_a
-- m.N N.“ a.m 2am o.~ o.a_ c_a_am ta>ca oa\a_\HH a_a_ataaa onafi
ca ah 2a 00 ua_asam 4\ae 4\aa aaaaam aaaa a=a_a azmmz

mczua o.mcwcopgo n.amoo mcwrasam “caspamch

ILmQEaH Pazuwmmm mcwcopgu

AH acadv .cmumz

mcwx:_cu umpmcwco_:o yo pampcoo acmsumEoFmgwcu Fapou ”mp:w_a acmEpmmcp capo: :mm_;owz we xm>c3m .Fum mpnmp

56

 

 

a_.o N.a m.__ Face;
a_.o m.a m.P_ zaa a.o a.o acaoaa _a\a_\m aaa_ccaapa Gama
o_.o a.“ o.a Pa:_a
a.~ F.a o.a 2am ma.o a._ ta>ca t_a_a .am oa\ap\__ Apia acctaz oaoa
mo.o a.a o.a _a=_a magaaaz
a.” N.a o.m 25a. F.P a.N :aa_;aaz axaa _a\a_\m coaaxaaz came
am_.o a.m O.“ ¥_acaa Acoaaz .a satay
ao.m F.a o.m azaa ao._ am._ aa>aa t_a_a .um oa\mm\__ ap__>a>gaz oafia
¢.o a.a a.m apa=_a :_aac=oz
m.o m.m a.a 2am ma.o ao.m acwoac< axaa .a\m_\m coca ooam
ao.o a.a o.~ xPa=_a Aaoz ac
s.aaaa zaa a_aaaav
_.a N.a o.~ 3am 0., N.N :aa_;a_z axaa .a\a\m aaacaaaa oa_m
mo.o m.a N.N Pagan
a._ m.a 5., 3am NF._ oa._ :aaa;a_z axaa _a\a_\m :aaaxaaz oama
ao.o M.“ o.a x_a=_a astaa acaaa
o.a m.a o.a azaa aa.o ma._ t_a_a .am axaa oa\a_\__ aaaOLa oaam
mm.o a.a m.m Pa=_a
o.m a.“ m.m 3am a.o a.m caacgaaz axaa tantaz coacaa ooao
aH.o a.a a.“ Pa=_a
om.o m.“ 0.0 3am N._ a.N caa_;acz axaa _a\am\a casaa_aa ammo
:a go aa_asam 4\as 4\as aaaaom aaaa acapa azmmz
«capo o.m:_co_;u n.mmom m:__asmm ucmauaace
Icmasmh Paauwmam mewcoPco
A.acaav .H-a a_aac

57

 

 

-- -- 0.0 0.0_ 0a000 0_.F p
-- -- 0.“ 0.0_ 3am 00.0 a 00.0 000000 0a0a_a< 0000
-- -- -- 0.00 000_a
-- -- -- 0.00 200 0.0 0.0 000000 _0\0_\0 000_0az 0000
-- -- -- -- Facaa _._ c
-- -- -- -- 300 0.0 0 000000 _0\0_\0 ax0ax_a¥ 0000
-- 00.0 0.“ 0.0_ _0000 0000ta>000
-- 00.0 0.N 0.0, 0300 0.0 0.0 000000 00\_P\__ .00 00000002 0000
-- _0.0 0.0 0a.0 Pac_u A0a0000 ta>00 a_aaz
-- “0.0 0.0 05.0 3am 0.0 0.0 0_a_taz .00 _0\NN\N .a00 a0_0a0 0000
-- -- a.“ 0.00 _0000
-- -- 0.5 0.0, 300 0.0 000000 _0\0P\0 0000000 La0a0 0000
-- _0.0 0.5 0._ .acaa A000 acaazaagv
-- 00.. 0.“ 0._ 200 0.0 00., 0000a000 axaa _0\0N\N a00<.0 0N00
0 0.0 0.0 0.0 F_a000 .00 aa0000
0_ 0.0 “.0 0.0 0300 0.0 00._ 0000a000 axaa _0\0N\N a0_a aacgz 000a
-- 00.0 0.“ 0.0 .0000
-- 00.0 a.“ 0.0 300 N._ N0._ to_aa000 axaa _0\0N\N 0000_00z 0000
-- 00.0 0.0 0.“ _a0_a .
-- _.0 0.0 “.0 300 0., 0.0 ta>_0 L_a_0 .00 00\N_\__ 0000: 0000 0000
00 a0 :0 0o 0apaaa0 0\00 0\00 a00000 a0a0 000_a a2003

acspa 0.0:0copsu .m0oo m:0_050m #:0500000

-cqumH F00000mm mcwco_;o

A.00000 .F-0 a_0ac

f_l—l—' 0? PM l—F" F-N r-l— l'_l_' 0000 0000

F

L00 CO CO

CO Or— WED

58

O
0

©0000 l\l\ [\N [\C‘ [\05 rxm I\I'\

r-r—I— \DQ‘ 00 l—"\ [\w 00 00 mm mm

CNN CO NO”) ON r—m £00 0000

00:00
30m
F0000
30m
_0:_d
30x
00:00
30m

00:00
x mac
00:00
0002

P000;
30m
F0:_d
xzam

0 .0000
a Pac_a
30m

mm.o

m.O\w.o

n w

or)
CO
LL.l-—

.>oz cow
mm.H n x

mm.o

m.o

mm.m

m.—

m¢.o

_._

N.o\m.o

n

cczocw

ncsoco

000000

eczocw

uczocw

uczoco

nczocw

nczocw

nczocu

 

a
umF 50m

0\05

.mc_co_;o

Pa0000a0

0\0e
.mmoa
0

0000000

muczom

ow\m\m_ m_00>0_050_ou 0000
_w\m\m Fpazoa
Pw\0m\m apuo_caco
_m\om\m __0300
om\m\NH
00\0\N_ 00000a0
0000 mocoa
Pm\om\m c0< ao_a;0:0¥
_m\om\m xu00 00030:
_m\—m\m 000900
0000 0:0,;
m:__asam “caspamch

 

A.0000v .0-0 a_0ac

59

 

 

000000
0000F0z
000000
00000000000020.
F0000
000a~000a3 000000 0000000 a000000 0000
0.0 0.0 _.FP F0000 0.0 H
0.0 0.0 0.00 300 0.0 0 000000 ow\mm\__ 0000000 0000 0000
u- 0.0 0.0 F0000 0._ H .000 0N
.. 0.0 0.00 30m 0.0 0 \.00_ 0_ 000000 _m\0p\m 0000302 0_00
.. 0.0 N.NH F0000
u- 0.0 N.NH HFF030:
.. 0.0 0.NH 0000000
.. 0.0 0.NH 00000 0.0, 000000 om\0N\FP 00002 00_0
a0 00 0o 0a00000 0000 0000 000000 a000 00000 02000
00000 0.00000_00 .0000 00000000 000500000
-00000» P0000000 M0000000
A.00000 ._-0 0_00h

6O

 

 

 

0.00 oz 0.0 0.00 0.N0 oz oz 00000

0_.0 .m .0 oz oz oz oz oz oz oz 300 0000
0.00 oz 0.00 0.~0 0.00 00.0 00.0 00000

0 .m .N .0 oz oz oz oz oz 00.0v 00.0 0300 0000
0 .0 0.00 oz 0.0 0.0 0.00 oz «0.0 000000

.0 .0 .N .0 oz oz 02 oz oz oz 00.0 0300 0000
0.00 oz oz 0.0 0.00 00.N 00.0 00000
0 .0 .m 0.00 oz oz 0.0 0.00 00.0 00.0 0000

.N-0-0-0 oz oz oz oz oz oz 00.0 200 0000
0.000 oz 0.0 0.00 0.0N0 oz 00.0 000000
0 .0-0 0.Nm0 oz 0.0 0.00 0.N~0 oz 00.0V 0.0000

.0.0 .N .0 oz oz oz oz oz oz 00.0V 300 000w
~.~00 oz 0.0m N.00 0.00 00.0 00. 000000
0 .0 .0 N.000 oz 0.00 0.00 0.00 00.0V 00.0 0.0000

.N0 . 00 . 0 .0 oz oz oz oz oz oz 3.0 300 0000
0.000 oz 0.0 «.00 0.000 00.0 00.0 000000
0 .0 .0 0.0 oz oz ~.0 0.0 oz N0.0 0000

.m .00 .0 oz oz oz oz oz oz 2.0 0300 0000
0 0.000 oz 0.0 0.NN 0.000 oz 0~.0 000000

.0-m-m-0 oz oz oz oz oz oz 00.0 300 0000
0.000 oz 0.00 0.00 000 oz 00.0 00000

0 .m .N-0 oz oz oz oz oz coz 00.0 300 0000

0a000a0000 02000 0000 N000000 0000000 00000 .00 u 0N 0000 0a00000 02002

000000000
0\00 .0000000000000 000000000000
000 00000 0.00002
00000000 00000000000 00 0000000 00000000000000 00000 0000000 000000000 00003 0000000: 00 00>000 .0-0 00000

 

 

0 0.00 oz 0.0 0.0 0.00 00000 00.o 00000

.0 .0 .0 .0 oz oz oz oz oz 00.0 00.0 000 o000
0-0-0 0.00 oz 0.0 0.0 0.00 oz 0o.oV 00000

.00.0 .0 oz oz oz oz oz oz 00.0 300 o000
0.00 oz 0.0 0.0 0.00 oz o~.o 00000

0 .0 .0.0 oz oz oz oz oz 00.0 00.o 000 o0o0
0.00 0.0 0.0 0.00 oz 0o.o 00:00

0 .0 .0 .0 oz oz oz oz oz oz 00.0 300 0000
0.00 oz o.0 o.0 o._N oz 00.0 000000

0 .0 .0 .0 oz oz oz oz oz oz 00.o V.300 o000
0 .0 .00 0.00 oz 0.0 0.0 0.00 oz 00.0 000000

.N .00 .0 oz oz oz oz oz oz 0o.o 200 800
0.00 oz 0.0 0.0 o.00 oz 0o.o 000000
0 .0 .0 0.00 oz o.~ 0.0 0.00 oz 0o.o 0.0000

.0 .0 .N .0 oz oz oz oz oz oz 00.0 300 0000
0.00 oz 0.0 0.00 0.00 oz oz 00000

0 .0 .0 .0 oz oz oz oz oz oz oz 300 o000
0.00 oz 0.00 0.0 0.00 oz 00.0 000000

0 .0 .0 .0 oz oz oz oz oz oz 00.0 0300 o000
0 .0 0.00 oz 0.0 o.o0 0.00 oz oz 00000

00-0-0-0 oz oz oz oz oz oz oz 300 oo0o

0003000000 00:00 00000 0000000 0000000 00000 000 u 00 0000 0000000 0zmmz

000000000

000: .0000000000000 000000000000

 

0.00000 .0.0 00000

62

 

 

 

o.o o._ o.o N.o oz oz oz o_oo_o
o.o oz o.o o.m oz oz oz o_mo_o

o .oo oz oz oz oz oz oz oz zoo oooo
z .o .m o.o oz o.o o._ o.z oz ooov _oo_o

.m .N .o_ oz oz oz oz oz oz oooV zoo oNoo
_.o oz z.o o.N o.o oz oz _oo_o

o_ .o-o-o _.z oz N.m o.~ o._ oz oz zoo oooo
o.oz oz oz o.o o.oo No.o o_.o omooo

z oz oz oz oz oz No.o oo.o zoo ooom
o_ o.o_ oz o.o o.~ _.o oz oz.o zoooo

.o-o-o-o-z z.o oz oz N.o o.N oz oo.o zzmz oooz
o.o_ oz o.o o.N o.N_ oo.o o_.o _o=oo

_ oz oz oz oz oz om.o N_.o zzmz oooo
o.oo oz o.o o.o o.o_ oz oz _oo_o

_ oz oz oz oz oz oz oz zoo oooo

o.o~ oz oz o.N z.zz oz oz .oo_o ,

. oz oz oz oz oz oz oz zoo ozoo
o-z .o o.o_ oz m.o o.N o.o_ oz mo.o __oo_o

.o .o-~ ._ oz oz oz oz oz oz oo.o zoom oooz
o._~ oz o.o o.~ o.z_ oz oN.o zoo_o

z ._ o.zo oz o.o N.o o.o~ oz mo.o 2oz oooo

mmozomuooo zzzzz ooozo Noozozo ooNPozo ozozo ozoo u oooo zso om_oEmo mzmmz

Homeuomop
4\m: .co_umgpcmucoo mu_Po:ocmmgo
z.oooov ._-o m_omz

63

 

 

 

oz oz oz oz oz oz oz zoooo
o .z .o oz oz oz oz oz oz oz 3oz oozo
oz oz oz oz oz oz N_.o _oo_o
oz oz oz oz oz oz oo.o oo_mzoz
oz oz oz oz oz N.o Nooo omzoooo
z ._ oz oz oz oz oz o.o oo.o ozooo oz_o
o.o oz oz o.o oz oz oz.o _mooo
o .o_ oz oz oz oz oz oz o_.o 3oz oooo
_.N oz oz oz z.o oz oN.o zoo_o
_ oz oz oz oz oz oz 8o 5. 88
o.o oz N.N o.o oz oz oz zoozo
z “o oz oz oz oz oz oz oz zoo ozoo
o.o oz o.~ o._ oz oz oz zoo_o
o .z .mz-o oz oz oz oz oz oz oz zoo oooo
o.m oz oz o.o o.o oz o_.o z_m=zo
oz oz oz oz oz oz o_.o moo
z .o _.o oz oz o.o z.o oz o_.o _oozo
ao_-m_-o-_ oz oz oz oz oz oz o_.o mm_z oozm
o.o o.o N.N _._ o.o oz o_.o _oozo
o-_ oz oz oz oz oz oz o_.o zoo omom
o.o o.o z.o o.o o.o oz o_.o Pmo_o
_ oz oz oz oz oz oz mo.o zzmz oooo
mmzoomoooo zzzzz ooozo Noo_ozo ooozozo m_ozo N_oo u oo_o o_oo omPoEom mzmmz
“cwspmmgh
4\mz .:o_pogp:mocou mu__m;o:omoo
z.oooov .o-o m_omz

64

.Aczocxcz cw we?“ pumucou :_ wocmgwwwwu mzpv mE_w
pumucoo mczgopnu zmmcoz o mucmmmzomg on: meow; .=o= mm: can» wcw— cowuzozgumwu mop czov gmcagoz mo: zoom

.mcocmeoFoz_La Fopop .zzhz;
.mmozz we mmoom>o mop we vmmpmc_ mPoEmm Laps: moo $0 mp_:mmo wzzm

.mawc: Lopou pzmoounazczmeQ moo muzcs .Lopou .uw
.Ampwca zuwv_ngzp ongmEozwgomcv m.az moo wwwca .zp_cwooou .Hm

.8538 mazmposom 33332 9: 98;; :25 93 05 ”:2: 3 moon: m5 ...z6~_.. DE 25me :33 3 mac: o 323v

.um_$_uwom no: mo: mcwgopso Fozcwmmg $0 moo“ mop .zmwpw:_ o: m_
moms» moms: .m:_oo_;u Poznzmmo Fmpoa . h mmcwgozgu Poouwmmo mmow . o ”co_uogu:wu:ou mowzopco Pooczmwmu
.mmoc cowpmczgozoomoo . Lo .mmoc :oozm:_gozouumoo - oo mmmou mowzopcoo
.Lmo532 Empmxm apooom Lopez .zmmzo

 

 

oz oz oz oz oz oz oz o_oooo
oz oz oz oz oz oz oz osmz_oz
oz oz oz oz oz oz oz o_mowo
oz oz oz oz oz oz oz oommoooomoooz
oz oz oz oz oz z.o oz o_mooo
o .z .o oz oz oz oz oz oz oz oozmzoommz oooo
oz oz oz oz oz oz oo.o Poooo
o_ ._ oz oz oz oz oz oz oo.o zoo oooo
mezzomoooo zzzz moozo ~.o_ozo .oN_ozo ozozo Nso u oozo o_oo om_osmm ozmmz

 

ucmsummoz ;
o\m: .cowumgpcmucou mc_Fo:o:omLo

 

A.ooooo .o-o moooz

65

$0>050$ $0005

00$00000

A00$x0$0 000000 0$00$PV 00$000000000L
00$:0u$0m

00$$pu0m . 00$000$$$00$0

A0E$Pv uc0sumon00 :0

00$$$00om 00: $0000

A500 0000V 0005000000 :0

O
O
O
O
O
0

0H
mH
0’
ma
NF
—$
op
m

00$popc0e$00m .
00$00_:0000 .
00$000$Lopgou000 .

A0pozomocoou0sox0zv $000000 00$00$$00 .m
00$000$000$$ .N

00$000$x0 .0

00$000$Lopouupmoo .m

00$00$up$$ .0

m

N

F

.00000 $0$uc0000m 00000000 005500 .0000 0500 0;“ :$ 00000 00:0
0000500000 0000$00_ 00:000 .0000000000 $0 00000 050 :$ 00>$m 000 00000 0:0Epo0oz .000000000 0005000L$m

.00$$$00om 0m$3z0guo mm0_c: .m0_osom 00003 00050 $0 0:00:00 00$Fozocom$0 $0 0moz0>0 0:0 000 00$:m0mo

. waowumt mco: .QZU
.mo$oo zo_o$oooo oomzooo .o$oo

.0000 003 00$000$00_:0-0Lo oz .:0$00:$00_;0 000$0o 0:0 .:0$0000$$$ L0p$0 00x00 003 0$osmm-0ooc
.:0$000$00_;0-0000 000$00 0:0 .:0$pogp_$$ $0a$0 c0zmp 003 0$050mu0$oe

.0005“ $0 000umo$ m0_osmm $0003 030 $0 000$0>0 05$

x

 

A.oooov .o-o 0_oo$

APPENDIX C
Statistica] Analyses.

Effect of Water Source and Chlorine Dose on TTHM Levels

66

Table C-l. Data used for statistical analyses.

 

Surface Water Data

 

 

WSSN Number* TTHM** Level (pg/L) Chlorine Dose (mg/L)
1770 281.8 14.0
0710 226.9 5.8
0040 151.4 3.66
4040 81.5 3.7
3400 44.2 3.9
7] i S.E.M. = 157.16 i 17.294 ug/L X1 = 6.05 mg/L r] = 5

 

Great Lakes and Connecting Waters Data

 

 

WSSN Number* TTHM** Level (pg/L) Chlorine Dose (mg/L)
2150 142.2 6.8
2640 136.0 4.4
0160 68.0 3.25
5470 65.0 3.4
0850 57.6 2.8
0600 56.7 2.8
2890 48.6 1.85
4570 47.2 1.90
2170 46.4 2.2
4160 37.0 1.54
4580 34.6 2.3
4090 26.3 1.4
5480 24.6 2.0
4560 21.0 1.82
7080 19.9 1.65
3670 19.4 1.06
5950 15.3 0.8

12 i S.E.M. = 50.90 i 9.379 ug/L 72 = 2.47 mg/L r2 = 17

 

67

Table C-l. (Cont.)

 

Groundwater Data

 

WSSN Number* TTHM** Level (pg/L) Chlorine Dose (mg/L)
5370 24.9 0.8
1260 18.9 1.1
4340 10.9 1.4
6940 6.1 1.0
0120 6.3 1.4
3240 6.0 1.1
3630 5.0 0.46
3760 4.1 1.9
3760 3.8 2.35
3950 3.4 0.5
1340 2.8 0.92
1570 0.4 1.97
4170 0.0 1.0
Y3 : S.E.M. = 7.12 i 10.725 ug/L X3 = 1.22 mg/L r3 = 13

 

*NSSN, water supply system number.
**TTHM, total trihalomethane

 

 

68

 

Source of Variation d.f. _§iy spxy ‘555
Water Source (unadjusted) t-l = 2 SST(y) SPT SST(x)
Plants/Source (unadjusted) n-t = 32 SSE(y) SPE SSE(x)
Total n-l = 34 SSy SPXy SSx
t r 2 2
$3 = 2 2 vi. - (Y ../n) = 142931.91
Y i=l 3:1 3
$3 = E (v2 /r) - (Y2 /n) = 81329 66
T(y) i=1 i. on o
= - = 6 6 .
SSE(y) SSy SST(y) 1 02 25
t r 2 2
$5 = 2 2 xi. - (x ../n) = 204.42156
i=l j=l 3
$3 = E (x2 /r) - (x2 /n) = 84 21496
T(X) i=1 i. o. o
SSE(X) = SSx - SST(X) = 120.2066
n n n
Sny = iZl x1y1 - 1;] x1 1;] yi/n = 3899.159
t
SPT = Z (xi.yi./ri) - (x..y..)/n = 2613.457

SP = SP - SP = 1285.702

E xy T
= 2 =
SSR(E) SPE/SSE(X) 1375l.57l

SS = SSE(Y) - SS = 47850.68

E R(E)

MSE = SSE/(n-t) = 1495.3338

Figure C-l. One-way analysis of covariance (unequal replication).

 

 

69

H z B = 0

 

t = b/VMSE/SSE(X) = 3.0326
Student's t : i t0.025,n-l = 2.032

Since 3.0326 is greater than 2.032, the quantity of chlorine applied
to raw water has a significant effect on TTHM levels at the 95% confi-
dence limit.

Figure C-2. Statistical test to determine if quantity of chlorine
applied to raw water has an influence on TTHM levels
in finished drinking water.

- X), where b = SPE/SSE(X)

AK K K
b = 10.696
Surface water YAl = 122.4 ug/L
Great Lakes and - _
Connecting Waters YA2 - 54'43 “g/L
Groundwater 7A3 = 24.02 pg/L

Figure C-3. Mean total trihalomethane values adjusted for chlorine
concentration effects.

 

Bonferroni t-test t8 = (VA1 - VAZ)//mSEE(1/r1)+(1/r2)1

Degrees of freedom = n - t - l = 3l

l. Surface water versus Great Lakes and connecting waters

t8 = 3.4555
Since 3. 4555 is greater than 3. l8l (t “0 5,m =3), the
TTHM levels are significantly differegt (P< .0l).

2. Surface water versus groundwater

t8 = 4.834

Since 4. 834 is greatr than 3. l8l (t 0.005, m=3), the
TTHM levels are significantly diffePent (P < 0. Cl).

3. Great Lakes and connecting waters versus groundwater

t8 = 2.135

Since 2. l35 is smaller than 2. 532( 0.025, m=3), the
TTHM levels are not significantly di§ferent (P > 0.05).

Figure C-4. Statistical difference between mean TTHM levels in the
finished waters when Great Lakes,groundwater‘and surface
water were used as water sources after chlorine concen-
tration effect is corrected for.

71

Analysis of Variance

SS 61602.25

E

MS = SSE/n-t = 1925.0703

E
Degrees of freedom = n-t = 32

Bonferroni t-test

t =(V2 - V3)//mSE[(1/r2)+ (1/r3)] = 2.709

 

8
Since 2. 709 is greater than 2. 526( 0.025, m = 3), TTHM
levels would be significantly differgnt at 95% confidence
limits.

Figure C-S. Statistical difference between mean TTHM levels in the
finished waters when Great Lakes and groundwater were
used as water sources assuming no chlorine concentra-
tion effect.

APPENDIX D
TTHM Predictive Models

.30. .0¢=002mosaovoo + zohz .zo$0$oooovoo + A;
.:0$u0$pc0o:00 :0 :0moo0xsvmm + moxms .0000 00$30$00v + 0 u mooo.zzhhu . 0_:Ego$ 0:0 000$
00p=p$pmozm 000 A 0 0:0 .mm . 0 . 0v 000—0> 000Fm 000 A 0v 000030pc$ 0:0 ._0005 0;» c$0000 0» ”0002

72

0000 000x0 0.30;m$o¥

 

 

 

mo.mm Hooo.o Nm.©N 0H.O ON: moH x om.N NN m.oH1
I PmUOE anmWLM> L30;
0N.mm Hooo.o m¢.©m mN.o moH x m¢.N NN ¢.mH1
N©.mw Hooo.o ¢¢.mm ON: mod x Om.N NN N.m 1
meUOE MFDMFLG> mmLSH
®H.om Hooo.o N¢.©¢ NH: MN O.m 1
Nn.om Hooo.o mH.w¢ H¢.H1 . MN o.m
mN.mm Hooo.o ©O.Nm moH x mm.N NN m.mH1
I m—wUOE GFQMWLM> 03H
mm.o ©NNN.O mo.o ma Rm
No.0 mmom.o mp.o Nm.H1 Om
0H.H HNmo.o mo.m mofi x H¢.0 m0
No.0m Hooo.o NH.©m MN m.m 1
I WFQUOE wpnwwLw> wCO
cwcwmpaxm AHw—wn 0 m N H o
IwwLw> ZIHP N u A DOLQ *m m m m m m
.COwHMCwLOFSU

30003 mc$30$_0$ m$0>0$ Expo 00$00oo 00 00$mm03000 0_0$HF:E

mc$mz 0000$0>00 m_00oz

.Hno 0Fn0h

LITERATURE CITED

 

LITERATURE CITED

 

Afghan, B. K., Leung, R., and Ryan, J. F., 1974. Automated fluoro-
metric method for determination of citric acid in sewage and
sewage effluents. Water Res. 8: 789-795.

Alavanja, M., Goldstein, I., and Susser, M., 1978. A case control
study of gastrointestinal and urinary tract cancer mortality and
drinking water chlorination. In Water Chlorination:
Environmental Impact and Health Effects (J. L. Jolley, H.
Gorchev, and D. H. Hamilton, eds.). Ann Arbor Science Pub-
lishers Inc., Ann Arbor, Mich., Vol. 2, pp. 395-409.

 

 

Anderson, H. A. and Russell, J. D., 1976. Possible relationship be-
tween soil fulvic acid and polymaleic acid. Nature, 260: 597.

Anonymous, 1977. Chloroform precursors found in natural waters.
Chemical and Engineering News, 55 (June 6, 1977): 6-7.

Babcock, D. B. and Singer, P. C., 1979. Chlorination and coagu-
lation of humic and fulvic acids. Jour. Amer. Water Works Assoc.
71: 149-152.

Bellar, T. A., Lichtenberg, J. J., and Kroner, R. C., 1974. The occur-
ence of organohalides in chlorinated drinking waters. Jour. Amer.
Water Works Assoc. 66: 703-706.

Bjork, R. G., 1975. GLC determination of PPB levels of citrate by con-
version to bromoform. Anal. Biochem. 63: 80-86.

Black, A. P. and Christman, R. F., 1963. Chemical characteristics of
fulvic acids. Jour. Amer. Water Works Assoc. 55: 897-912.

Booth, H. and Saunders, B. C., 1950. The iodoform reaction. Chemy.
Ind. 1950: 824-825.

Brodtmann, N. V. and Russo, P. J., 1979. The use of chloramine for
reduction of trihalomethanes and disinfection of drinking water.
Jour. Amer. Water Works Assoc. 71: 40-42.

Brown, B. R., Sipes, I. G., and Saglayn, A. M., 1974. Mechanisms of
acute hepatic toxicity; chloroform, halothane, and glutathione.
Anesthesiology 41: 554-561.

Budiansky, S., 1980. The risky business of assessing risk. Environ.
Sci. Technol. 14: 1281-1282.

73

74

Bunn, W. W., Haas, B. B., Deane, E. R., and Kleopfer, R. D., 1975.
Formation of trihalomethanes by chlorination of surface water.
Environ. Letters 10: 205-213.

Burger, P. C. and Herdson, P. B., 1966. Phenobarbital-induced fine
structural changes in rat liver. Amer. J. Pathol. 48: 793-809.

Cagen, S. Z. and Klaassen, C. D., 1980. Binding of glutathione-
depleting agents to metallothionein. Toxicol. Appl. Pharmacol.
54: 229-237.

Cantor, K. P., Hoover, R., Mason, T., and McCabe, L. J., 1977. Asso-
ciation of cancer mortality rates and trihalomethane levels in
municipal drinking water supplies. Amer. J. Epidemiol. 106:
230-231.

Cheung, W. Y., 1980. Calcium and Cell Function. Academic Press,
New York, Vol. 1.

Christman, R. F. and Ghassemi, M., 1966. Chemical nature of organic
color in water. Jour. Amer. Water Works Assoc. 58: 723-741.

Christman, R. F., Johnson, J. D., Hass, J. R., Pfaender, F. K., Liao,
W. T., Norwood, P. L., and Alexander, H. J., 1978. Natural and
model aquatic humics: reactions with chlorine. In Water
Chlorination: Environmental Impact and Health Effects (J. L.
Jolley, H. Gorchev, and D. H. Hamilton, eds.), Ann Arbor Science
Publishers, Inc., Ann Arbor, Mich., Vol. 2, pp. 15-28.

 

Cohen, E. N. and Hood, N., 1969. Application of low temperature auto-
radiography to studies of the uptake and metabolism of volatile
anesthetics in the mouse. Anesthesiol. 30: 306-314.

Crane, A. M., Kovacic, P., and Kovacic, E. D., 1980. Volatile halo-
carbon production from the chlorination of marine algal by-
products, including D-mannitol. Environ. Sci. Technol. 14:
1371-1381.

Dowty, B., Carlisle, 0., Laseter, J. L., and Storer, J., 1975. Halo-
genated hydrocarbons in New Orleans drinking water and plasma.
Science 187: 75.

Ekstrom, T. and Hogberg, J., 1980. Chloroform-induced glutathione
depletion and toxicity in freshly isolated hepatocytes. Biochem.
Pharmacol. 29: 3059-3065.

Enos, H., 1979. Exposure to halogenated organic chemicals from drink-
ing water. In Directory of On-Going Research in Cancer Epide-
miology. International Agency for Research on Cancer, German
Cancer Research Center, p. 342.

Eschenbrenner, A. B., 1944. Induction of hepatomas in mice by repeated
oral administration of chloroform, with observations on sex
differences. Jour. Natl. Cancer Instit. 5: 251-255.

75

Fry, B. J., Taylor, T., and Hathway, D. E., 1972. Pulmonary elimina-
tion of chloroform and its metabolite in man. Arch. Int. Pharma-
codyn. 196: 98-111.

Fuson, R. C. and Bull, B. A. 1934. The haloform reaction. Chem. Rev.
15: 275-309.

Geisy, J. P., 1982. Personal communication. Michigan State Univer-
sity, E. Lansing, MI 48824.

Gill, J. L., 1978a. Design and Analysis of Experiments in Animal
and Medical Sciences, Vol. l. The Iowa State University
Press, Ames, Iowa. 409 pp.

 

 

Gill, J. L., 1978b. Design and Analysis of Experiments in Animal
and Medical Sciences, Vol. 3 (Appendices). The Iowa State
University Press, Ames, Iowa. 173 pp.

 

 

Gill, J. L., 1981. Personal communication. Michigan State Univer-
sity, E. Lansing, Mich.

Gopinath, C. and Ford, E. J. H., 1975. The role of microsomal hydroxy-
lases in the modification of chloroform hepatotoxicity in rats.
Br. J. Exp. Path. 56: 412-422.

Hoehn, R. C., Barres, D. B., Thompson, B. C., Randall, C. W., Grizzard,
T. J., and Shaffer, P. T. B., 1980. Algae as sources of trihalo-
methane precursors. Jour. Amer. Water Works Assoc. 72: 344-350.

Hogan, M. 0., Chi, P., Hoel, D. G. and Mitchell, T. J., 1979. Asso-
ciation between chloroform levels in finished drinking water
supplies and various site-specific.cancer mortality rates. J.
Environ. Pathol. Toxicol. 2: 873-887.

Ilett, K. F., Reid, W. D., Sipes, I. G., and Krishna, G., 1973.
Chloroform toxicity in mice: correlation of renal and hepatic
necrosis with covalent binding of metabolites to tissue macromole-
cules. Exp. Mol. Path. 19: 215-229.

Johnson, J. D., 1976. Measurement and persistence of chlorine resi-
dules in natural water. In The Environmental Impact of Water
Chlorination, Proc. Conf. (R. L. Jolley, ed.). Oak Ridge
National Laboratory, Energy Research and Development Administra-
tion, U. S. Environmental Protection Agency, Vol. 1, pp. 43-71.

 

Jorgenson, T. A. and Rushbrook, C. J., 1980. Effects of Chloroform
in the Drinking Water of Rats and Mice: Ninety-day Subacute
Toxicity Study. Health Effects Research Laboratory, Office
of Research and Development, U. S. Environmental Protection
Agency, Cincinnati, Ohio. EPA-600/1-80-03O (July 1980).

 

 

76

Kaczmar, S., 1979. The Occurrence and Behavior of Halomethanes in
the Aquatic Environment. A Master's Thesis. Department of
Fisheries and Wildlife, Michigan State University, E. Lansing,
Mich.

 

 

Kavanaugh, M. C., 1978. Modified coagulation for improved removal of
trihalomethane precursors. Jour. Amer. Water Works Assoc. 70:
613-620.

Khan, S. U. and Schnitzer, M., 1971. Further investigations on the
chemistry of fulvic acid, a soil humic fraction. Can. J. Chem.
49: 2302-2309.

Khan, S. U. and Schnitzer, M., 1972. Permanganate oxidation of humic
acids, fulvic acids, and humins extracted from Ah horizons of a
Black Chernozem, a Black Solod and a Black Solonetz soil. Can.
J. Soil Sci. 52: 43-51.

Kleopfer, R. D. and Fairless, B. J., 1972. Characterizationcfliorganic
components in a municipal water supply. Environ. Sci. Technol.
6: 1036-1037.

Kluwe, W. M. and Hook, J. B., 1978. Polybrominated biphenyl-induced
potentiation of chloroform toxicity. Toxicol. Appl. Pharmacol.
45: 861-869.

Kluwe, W. M., McCormack, K. M., and Hook, J. B., 1978a. Potentiation
of hepatic and renal toxicity of various compounds by prior

exposure to polybrominated biphenyls. Environ. Health Perspect.
23: 241-264.

Kluwe, W. M., McCormack, K. M., and Hook, J. B., 1978b. Selective
modification of the renal and hapatic toxicities of chloroform by
induction of drug metabolizing enzyme systems in kidney and liver.
J. Pharmacol. Exper. Therap. 207: 566-573.

Kuzma, R. J., Kuzma, C. M., and Buncher, C. R., 1977. Ohio drinking
water source and cancer rates. Amer. J. Publ. Health 67: 725-729.

Kuznetsov, I., 1970. The Microflora of Lakes and Its Geochemical
Activity. University of Texas Press, Austin, Texas. 503 pp.

Larson, R. A. and Rockwell, A. L., 1978. Citric acid: potential
precursor of chloroform in water chlorination. Natunwissen-
schaften 65: 490.

Lytle, C. R. and Perdue, E. M., 1981. Free, proteinaceous, and humic-
bound amino acids in river water containing high concentrations of
aquatic humus. Environ. Sci. Technol. 15: 224-228.

Mansuy, D., Beune, P., Cresteil, T., Lange, M., and Leroux, J. P.,
1977. Evidence for phosgene formation during liver microsomal
oxidatigon of chloroform. Biochem. Biophys. Res. Commun. 79:
513-517.

77

Maugh, T. H., 1981a. New study links chlorination and cancer. Science
211: 694.

Maugh, T. H., 1981b. Another potential problem in drinking water.
Science 212: 431.

McGuire, M. J. and Suffet, I. H., 1978. Adsorption of organics from
domestic water supplies. Jour. Amer. Water Works Assoc. 70:
621-636.

McMartin, D. N., O'Connor, J. A., and Kaminsky, L. S., 1981. Effects
of differential changes in rat hepatic and renal cytochrome
P-450 concentrations on hepatotoxicity and nephrotoxicity of
chloroform. Res. Commun. Chem. Pathol. Pharmacol. 31: 99-110.

Michigan Department of Public Health, 1979. Index Summary of Water
Departments in Michigan (computer printout). November 30,
1979. Lansing, Mich.

 

 

Michigan Department of Public Health, 1980. Addresses of Water
Departments in Michigan. January, 1980. Lansing, Mich.

 

 

Mieure, J. P., 1977. A rapid and Sensitive method for determining
volatile organohalides in water. Jour. Amer. Water Works Assoc.
69: 60-62.

Mikkelson, R. B., 1978. Calcium and neoplasia. Prog. Exp. Tumor
Res. 22: 123-150.

Miller, E. C. and Miller, J. A., 1966. Mechanisms of chemical carcino-
genesis: nature of proximate carcinogenesis and interactions with
macromolecules. Pharmacol. Rev. 18: 805-838.

Moore, G. S., Tuthill, R. W., and Polakoff, D. W., 1979. A statistical
model for predicting chloroform levels in chlorinated surface
water supplies. Jour. Amer. Water Works Assoc. 71: 37-39.

Moore, L., 1980. Inhibition of liver-microsome calcium pump by in
vivo administration of CCl4, CHCl3, and 1,1-dichloroethylene
(vinylidene chloride). Biochem. Pharmacol. 29: 2505:2511.

Morris, J. C., 1966. The acid ionization constant of HOCl from 5 to
35°C. J. Phys- Chem. 70: 3798-3802.

Morris, J. C., 1976. The chemistry of aqueous chlorine in relation to
water chlorination. In The Environmental Impact of Water
Chlorination, Proc. Conf. (R. L. Jolley, ed.), Oak Ridge
National Laboratory, Energy Research and Development Adminis-
tration, U. S. Environmental Protection Agency, Vol. 1, pp. 27-41.

 

78

Morris, J. C. and Baum, B. 1978. Precursors and mechanisms of halo-
form formation in the chlorination of water supplies. In Water
Chlorination: Environmental Impact and Health Effects (R. L.
Jolley, H. Gorchev, and D. H. Hamilton, eds.). Ann Arbor Science
Publishers Inc., Ann Arbor, Mich., Vol. 2, pp. 29-48.

 

Nakajima, T. and Sato, A., 1979. Enhanced activity of liver drug-
metabolizing enzymes for aromatic and chlorinated hydrocarbons
following food deprivation. Toxicol. Appl. Pharmacol. 50:
549-556.

National Academy of Sciences, 1977. Drinking Water and Health.
Vol. 1, Washington, D. C. 939 pp.

 

National Cancer Institute, 1976. Report on Carcinogenesis Bioassay
of Chloroform. Carcinogenesis Program, Division of Cancer
Cause and Prevention, National Cancer Institutes of Health,
Bethesda, Maryland. March 1, 1976. 60 pp.

 

 

Norman, T. S., Harms, L. L., and Looyenga, R. W., 1980. The use of
chloramines to prevent trihalomethane formation. Jour. Amer.
Water Works Assoc. 72: 176-180.

Norwood, D. L., Johnson, J. D., Christman, R. F., Hass, J. K., and
Bobenrieth, M. J., 1980. Reactions of chlorine with selected
aromatic models of aquatic humic material. Environ. Sci. Technol.
14: 187-190.

Novak, J., Zluticky, J., Kubelka, V., and Mostecky, J., 1973. Analysis
of organic constituents present in drinking water. J. Chromatogr.
76: 45-50.

Oliver, B. G. and Lawrence, J., 1979. Haloforms in drinking water: a
study of precursors and precursor removal. Jour. Amer. Water
Works Assoc. 71: 161-163.

Oliver, B. G. and Visser, S. A., 1980. Chloroform production from
the chlorination of aquatic humic material: the effect of mole-
cular weight, environment, and season. Water Res. 14: 1137-1141.

Otson, R., Williams, 0. T., Bothwell, P. D., and Quon, T. K., 1981.
Comparison of trihalomethane levels and other water quality para-
meters for three treatment plants on the Ottowa River. Environ.
Sci. Technol. 9: 1075-1080.

Packham, R. F., 1964. Studies of organic color in natural water. J.
Water Treat. Exam. 13: 316-334.

Pfaffenberger, C. D., 1980. Exposure to halogenated organic chemicals
from drinking water. In Directory of On-Going Research in
Cancer Epidemiology. International Agency for Research on
Cancer, German Cancer Resarch Center. p. 339.

 

 

79

Pohl, L. R., Bhoosan, B., Whittaker, N. F., and Krishna, G., 1977.
Phosgene: a metabolite of chloroform. Biochem. Biophys. Res.
Commun. 79: 684-69l.

Pohl, L. R. and Krishna, G., 1978. Deuterium isotope effect in bio-
activation and hepatotoxicity of chloroform. Life Sci. 23: 1067-
1072.

Pohl, L. R., Martin, J. L., and George, J. W., 1980. Mechanism of
metabolic activation of chloroform by rat liver microsomes. Bio-
chem. Pharmacol. 29: 3271-3276.

Poyer, J. L., Floyd, R. A., McCay, P. B., Janzen, E. G., and Davis,
E. R., 1978. Spin-trapping of the trichloromethyl radical pro-
duced during enzymic NADPH oxidation in the presence of carbon
tetrachloride of bromotrichloromethane. Biochem. Biophys. Acta
539: 402-409.

Reitz, R. H., Gehring, P. J., and Park, C. N., 1978. Carcinogenic risk
estimation for chloroform: an alternative to EPA's procedures.
Food Cosmet. Toxicol. 16: 511-514.

Reuber, M. D., 1979. Carcinogenicity of chloroform. Environ. Health
Perspect. 31: 171-182.

Reynolds, E. S., 1977. Liver endoplasmic reticulum: target site
of halomethane metabolism. Adv. Exp. Med. Biol. 84: 117-137.

Richard, J. J. and Junk, G. A., 1977. Liquid extraction for the rapid
determination of halomethanes in water. Jour. Amer. Water Works
Assoc. 69: 62-64.

Rook, J. J., 1974. Formation of haloforms during chlorination of
natural waters. Water Treat. Exam. 23: 234-243.

Rook, J. J., 1976. Haloforms in drinking water. Jour. Amer. Water
Works Assoc. 68: 168-172.

Rook, J. J., 1977. Chlorination reactions offulvic acids in natural
waters. Environ. Sci. Technol. 11: 478-482.

Rubenstein, D. and Kanics, L., 1964. The conversion of carbon tetra-
chloride and chloroform to carbon dioxide by rat liver homo-
genates. Can. J. Biochem. 42: 1577-1585.

SAS Institute Inc., 1979. Statistical Analysis System, Version 79.3.
SAS User's Guide, 1979 edition. Box 8000, Cary, N.C. 27511.

Smuckler, E. A., 1976. Structural and functional changes in acute
liver injury. Environ. Health Perspect. 15: 13-25.

Stevens, A. A., Slocum, C. J., Seeger, D. R., and Robeck, G. G.,
1976. Chlorinationcfi'organics in drinking water. Jour. Amer.
Water Works Assoc. 68: 615-62l.

80

Symons, J. M., Bellar, T. A., Carswell, J. K., DeMarco, J., Kropp, K.
L., Robeck, G. G., Seeger, D. R., Slocum, C. J., Smith, B. L., and
Stevens, A. A., 1975. National Organics Reconnaissance Survey
for halogenated organics (NORS). Jour. Amer. Water Works Assoc.
67: 634-647.

Tardiff, R. G., 1977. Health effects of organics: risk and hazard
assessment of ingested chloroform. Jour. Amer. Water Works Assoc.
69: 658-661.

Tardiff, R. G. and Deinzer, M., 1973. Toxicity of organic compounds in
drinking water. In Water Qual. Conf., Proc. 15th (V. L.
Snoeyink, ed.). pp. 23-37.

 

Taylor, D. C., Brown, D. M., Keeble, R., and Langley, P. F., l974.
Metabolism of chloroform. II. A sex difference in the metabolism
of [14C]-chloroform in mice. Xenobiotica 4: 165-174.

Uehleke, H. and Werner, T., 1975. A comparative study on the irre-
versible binding of labeled halothane trichlorofluoromethane,
chloroform, and carbon tetrachloride to hepatic protein and lipids
in vitro and in vivo. Arch. Toxicol. 34: 289-308.

United States Congress, 1974. Public Law 93-523. United States
Statutes at Large. 930 Congress, 20 Session, Vol. 88 (pt. 2):
1660-1694.

United States Environmental Protection Agency, 1978. Interim Primary
Drinking Water Regulations; control of organic chemical contami-
nants in drinking water. Federal Register 43: 5756-5780.

United States Environmental Protection Agency, 1979. National Interim
Primary Drinking Water Regulations; control of trihalomethanes in
drinking water. Federal Register 44: 68624-68707.

United States Environmental Protection Agency, 1980. Ambient Water
Quality Criteria for Chloroform. EPA 440/5-80-033 (October
1980).

 

 

Vajdic, A. H., 1980. Prechlorination and Trihalomethane Formation
at Water Treatment Plants. Research Publication No. 84.
Ministry of the Environment, Ontario, Canada. 16 pp.

 

 

Van Dyke, R. A., 1969. On the fate of chloroform. Anesthesiology
30: 257-258.

Van Dyke, R. A., Chenoweth, M. B., and Van Poznak, A., 1964. Meta-
bolism of volatile anesthetics. I. Conversion in vivo of
several anesthetics to 14002 and chloride. Biochem. Pharmacol.
13: 1239-1247.

Weber, J. H. and Wilson, S. A., 1975. The isolation and characteri-
zation of fulvic acid and humic acid from river water. Water
Res. 9: 1079-1084.

81

Weisburger, E. K., 1977. Carcinogenicity studies on halogenated
hydrocarbons. Environ. Health Perspect. 21: 7-16.

Young, J. S. and Singer, P. C., 1979. Chloroform formation in public
supplies: a case study. Jour. Amer. Water Works Assoc. 71:
87-95.

Youssefi, M., Zenchelsky, s. 1., and Faust, s. 0., 1978. Chlorination
of naturally occurring organic compounds in water. J. Environ.
Sci. Health, A13: 629-637.

Zogorski, J. S., Allgeier, G. D., and Mullins, R. L., 1978. Removal
of Chloroform From Drinking Water. University of Kentucky,
Water Resources Research Institute, Lexington, Kentucky. 86 pp.

CHIGRN STRTE UNIV

M919119911111911911911111111111111119111