Halocarbon Production from Oxidative
Biocides in Estuarine Waters
Maryland Univ.
College Park
Prepared for
Environmental Research Lab.
Gulf Breeze, FL
Feb 81
-------
P8*l-t57977
EPA 600/3-81-010
February 1981
HALOCARBOK PRODUCTION FROM OXIDATIVE
BIOCIDES IN ESTUARINE WATERS
by
George R. Helz
Rong Yew Hsu
Richard Sugam.
University of Maryland
College Park, Maryland 20742
Grant R803839-01/02
Project Officer
William P. Davis
Bears Bluff Field Station
Wadmalav Island, South Carolina 29487
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S.. ENVIRONMENTAL PROTECTION AGENCY
GULF BREEZE, FLORIDA
32561
•rmwrt* IT
-------
TECHNICAL REPORT DATA
/Please read Instructions an the rri-enc be/orr
1 REPORT NO
EPA-600/3-81-010
3. RECIPIENT'S ACCESSION NO.
1579?7
4. TITLE ANO SUBTITLE
Halocarbon Production from Oxidative Biocide?
in Estuarine Waters
5. REPORT DATE
FEBRUARY 1981 _ISS:JING_SAT£;
6. PERFORMING ORGAN NATION CODE
7. AUTHORISI
G.R. Helz, R.Y. Hsu and R. Sugam
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME ANO AOORESS
10. PROGRAM ELEMENT NO.
University of Maryland
College Park, Maryland
20742
11. CONTRACT/GRANT NO.
R803839-01/02
12. SPONSORING AGENCY NAME ANO AOORESS
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Gulf Breeze, Florida, 32561
13. TYPE OF REPORT ANO PERIOD COVERED
Final 8/1/75-7/31/77
14. SPONSORING AGENCY COOE
EPA/600/04
IS. SUPPLEMENTARY NOTES
| ,6. ABSTRACT
The formation of halo-organic compounds by chlorination of estuarine
waters has been investigated under both laboratory and field conditions.
Haloforms are readily generated in the laboratory with chlorine doses of
1 to 10 mg/1, the range employed by many coastal power plants. At sali-
nities above 1 g/kg, Br is the principal halogen in the haloform products
On a molar basis, more than 4% of the chlorine was converted to haloforms
in some tests. Ozone in the laboratory also generated haloforms in
estuarine water; the yields were similar to those obtained from chlorine.
However, only traces of haloforms were found in a power plant field site,
where apparently haloform-bypassing reactions consume free chlorine much
faster than in the laboratory. Identification of these reactions is un-
certain, but they may involve formation of stable halogenated macromole-
cules.
A large sewage treatment plant served as a volatile halocarbon source
to study the fate of these compounds. The major loss mechanism appears
to be volatilization to the atmosphere. Rates for this process are esti-
mated. However, there appears tcr be some loss under winter i&£e cover,
perhaps because of chemical or biological degradation.
17.
KEY WORDS ANO DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN £NO£D TERMS C. COSATI [• Icld/Cfoup
Halogens
Halogen organic compounds
Halogenation
Chlroination
Estuaries
Ozone
Volitization
Haloform Products
Sewage Treatment Plant
Chemical Biological Degra-
dation
06 /A
07/C
is. DISTRIBUTION
19. SECURITY CLASS i
21. NO. OF PAGES
34
Release Unlimited
20 SECURITY CLASS/ 7.
J2. PRICE
-------
DISCLAIMER
This report has been reviewed and approved for publication by the Gulf
Breeze Environmental Research L^'ucratoi /, U.S. Environmental Protection
Agency. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trades names or commercial products constitute endorsement or
recommendation for use.
11
-------
FOREWORD
The protection of our estuarine and coastal areas from damage caused
by toxic organic pollutants requires that regulations restricting the
introduction of these compounds into the environment be formulated on a
sound scientific basis. Accurate information describing dose-response
relationships for organisms and ecosystems under varying conditions is
required. The Environmental Research Laboratory, Rulr Breeze, contributes
to this information through research programs aimed at determining:
"the effects of toxic organic pollutants on individual
species and communities of organisms;
"the effects of toxic organics on ecosystem processes
and components;
"the significance of chemical carcinogens in the estuarine
and marine environments.
The production and fate of halo-organic compounds that are by-products
of .disinfection/biofouling control processes are investigated in the ra-
search project described in this report.
r (I
Henry F.^Enos
Director
Environmental Research Laboratory
Gulf Breeze, Florida
-------
ABSTRACT
The formation of halo-organic compounds by chlorination of estuarine
waters has been investigated under both laboratory and field conditions.
Haloforms are readily generated in the laboratory with chlorine doses of 1 to
10 mg/1, the range employed by many coastal power plants. At salinities
above 1 g/kg (i.e., one part per thousand), Sr is the principal halogen in
the haloform products. On a molar basis, more than 4% of the chlorine was
converted to haloforms in some tests. In the laboratory, ozone also generated
haloforms in estuarine water; the yields were similar to those obtained from
chlorine. However, only traces of haloforms were found at a power plant field
site, where halofonn-bypassing reactions apparently consume free chlorine much
fatter than in the laboratory. Identification of these reactions is uncertain,
but they may involve formation of stable halogenated macromolecules.
A large sewage treatment plant served as a volatile halocarbon source to
study the fate of these compounds. The major loss mechanism appears to be
volatilization to the atmosphere. Rates for this process are estimated. How-
ever, there appears to be.some loss under winter ice cover, perhaps because
of chemical or biological degradation.
This report was submitted in fulfillment of U.S.. Environmental Protection
Agency Grant No. R803839-01/02 to the University of Maryland. It covers the
period August 1, 1975 to July 31, 1977; the work was completed December 14,
1977.
iv
-------
CONTENTS
Foreword ill
Abstract iv
Figures vi
Tables vii
Acknowledgments viii
1. Introduction ..... 1
2. Conclusions 2
3. Laboratory Experiments on Volatile Halocarbon
Production 3
Analytical Method for Volatile Halocarbons 3
Laboratc—y Chlorination Experiments 5
Comparison of Clilorination and Ozonation 11
A. Field Studies 14
Chalk Point Power Plant 14
Back River Wastewater Treatment Plant 21
5. Other Activities 28
Workshop 28
Publications 29
References 30
Appendix 34
-------
FIGURES
Number Pa8e
1 Map showing sample locations 5
2 Bromine as a percentage of total halogen in haloform
products 1
3 Total haloform yield as a percentage of chlorine dose .... 8
4 Haloform yield as a percentage of chlorine dose vs. total
chlorine consumed in 10 minutes as a percentage of
chlorine dose 9
5 Calculated time for 99% consumption of Tree chlorine
by reaction 1 .10
6 Decay of chlorine produced oxidants in Patuxent River
water 15
7" ' Effect of temperature on decay in estuarine water chlori-
nated in the laboratory 17
8 Effect of removing organic matter on decay .......... 19
9 Back River estuary, Md., showing locations of sampling
points 21
10 Halocarbon patterns in Back River 24
vi
-------
TABLES
?age
Concentration (yg/1) of 100 ug/1 halofora solutions
one day after addition of adsorbing agents ......... 4
2 Halo forms produced in selected estuarine and coastal
water samples ....................... 6
3 Yields of bromoform in estuarine water treated with
chlorine (as OC1~) and ozone ................ 12
4 Exanples of possible oxldant consumption reactions in
estuarine water ............... . ...... 18
5 Macromolecular organic matter .... ........... 20
6 Back River halocarbons ................... 22
7 Major industrially produced halocarbons ranked by release
rate ............................ 25
8 Loss rates for volatile halocarbons ............ 26
vii
-------
ACKNOWLEDGMENTS
We are grateful to a number of our colleagues in the Department of
Chemistry at the University of Maryland, including David Anderson, Michael
Failey, Kenneth Ferri and Franz Kasler. We owe a special debt to Ronald
Block for his assistance and inspiring enthusiasm. Potomac Electric Power
Co. personnel were extremely helpful. Finally, Dr. R. A. Saunders, of the
Naval Research Laboratory, graciously lent time on his GC-MS instrument and
shared his expertise.
viii
-------
SECTION 1
INTRODUCTION
Chlorine is injected directly into marine waters by large coastal power*.
plants that use chlorine to prevent biofouling of their heat exchangers.
Chlorinated effluents are also injected into marine waters by wastewater
treatment plants and certain industrial enterprises. Since the discoveries
of haloforms in chlorinated drinking waters by Rook (1976) and a number of
halo-organics in chlorinated sewage treatment plant effluents by Jolley (1974) .
there has been concern about whether such compounds adversely affect aquatic
environments because large volumes of chlorinated discharges are entering
coastal environments. For example, in the northern Chesapeake Bay, 2Z of the
non-saline throughput is now treated wastewater effluent (Brush, 1972).
Based on past rates of growth, this fraction will reach 30Z or more in another
century. Power plants currently circulate a volume greater than 10Z of the
non-saline throughput; if past technology and trends continue, a volume
approaching this throughput will be circulated by the end of this century
(Bongers, 1973). It is likely that similar statistics could be cited for
many other major estuaries. Thus, any persistent, bioactive compound in-
jected into the aquatic environment from these sources may be capable, in
principle, of producing massive alterations in our coastal waters.
When this project began, no previous work had been done on the formation
of halo-organics through chlorination in marine waters. Yet there was sub-
stantial evidence that the behavior of chlorine in marine waters differs
from- its behavior in fresh waters because of the involvement of Br~ (which
is ubiquitous in seasalt) in the halogen chemistry. Similarly, some
differences were known ';o exist between the organic matter in marine and
fresh water systems. Thus, research on halo-organic formation in marine
waters seemed warranted. A distinctive feature of this project is the in-
clusion of a major field study to explore the production and fate of halo-
carbons in the natural environment.
-------
SECTION 2
CONCLUSIONS
1. Trihalonethanes (CUClj, CHBrCl2r CHBr2Cl, and CHBr3) were generated at^
the 10 to 1000 pM (~3 to 300 ppb) level when fluvial arvl estuarinc water
samples were chlorinated under laboratory conditions.
2. When salinity exceeded 5 g/kg (%o) bromoform was virtually the only
trihalomethane product.
3. No volatile halocarbons other than trihalotnethanes in estuarine water
of 14 g/kg salinity were found at the pM level by headspace analysis.
4. When both ozone and chlorine were rpplied in estuarine water of 14 g/kg
salinity to produce after 10 minutes a residual of 22 pequiv/liter (0.78 ppm
as chlorine), they produced identical trihalomethene yields.
5. Field studies at an estuarine power plant that employs continuous
chlorination revealed only traces (-Ipg/liter) of volatile halocarbon pro-
ducts.
6. No "free chlorine" (as measured amperometrically at pH 7) was observed
beyond the condensers at this plant, and the negligible yields of trihalo-
methanes are tentatively ascribed to the extremely rapid consumption of
free chlorine in the condensers. More work should be done to test this
hypothesis.
7. It was possible to trace the halocarbon plume from a wastewater treat-
ment plant more than 5 miles seaward under winter ice, but only a fraction
of this distance in spring when the ice was gone.
-------
SECTION 3
LABORATORY EXPERIMENTS ON VOLATILE
HALOCARBON PRODUCTION
Haloforms, the most abundant known halo—organic products from chlori-
nation of freshwaters, were a major concern in this project. In this section,
we describe our analytical method and provide some basic experimental data on
the nature of the halofonns produced by chlorinating estuarine water!?.
ANALYTICAL METHOD FOR VOLATILE HALOCARBONS
In the field, 40.0 ml of sample was placed in a Hypovial (nominal size
50 ml; actual volume when sealed cs described, 60.19 + 0.72 ml; Pierce
Chemical Co.). If oxidant was present, it was quenched with a drop of
concentrated Na2S202 solution. Then, a Teflo..-to-rubber laminated septum was
placed over the mouth and sealed in place wit'i a crimped aluminum ring. The
sample was stored on ice and normally analyzed within 48 hours. In labora-
tory chlorination studies, the samples were first treated with the desired
amount of NaOCl and then sealed as above. For analysis, the vials were
equilibrated at 60°C in a shaking water bath; 1 ml of the vapor was extracted
through the septum cap with a gas-tight hypodermic syringe, and injected into
a gas chromatograph equipped with a Hall Electrolytic Conductivity Detector
(Tracor) . The following conditions were normally used: carrier gas, 30 ml
He min~l; column packing, 60-80 mesh Tenax GC (Applied Science Labs, Inc.);
temperature in injection port, 200°, in column, 120°-200° programmed at 8°
min~^, in transfer line, 250°, in detector furnance, 850°; detector reaction
gas, 50 ml H2 min~l; detector electrolyte, 50Z n-propanol-water. Standards
were prepared in distilled water using aliquots of a primary standard that
contained 1 gram each of the compounds of interest, then brought to 100 ml
with methanol. Coefficients of variation, based on repeated analysis of
standards in the 10^ to 10^ nM range, were about 42. In early stages of the
work, some identifications veve checked by mass spectrpmetry, but the high
selectivity of the method for volatile chloro- and bromocarbons minimizes
the danger of nisidentifying gas chromatograph peaks.
When saline samples are analyzed, using a standard curve prepared from
solutions of the analyte in distilled water, the apparent concentration C^,
obtained from the standard curve must be corrected for salting cut in the~
sample. If CT and CA are the true and apparent original concentrations of a
compound in ~tne tes~t~solution, Vy and V^ are the volumes of vapor and liquid,
Hf is the Henry's Law constant "expresses' as a concentration ratio, ^ is the
"salinity, and £ is the Setchenow coefficient of the compound in water, then
the relationship between the true and apparent concentration is:
-------
c,
'T
-as , —
e + V.
L
1 +
VL
x C
(1)
'A
For the haloforms, Hep is 0.45, 0.25,' 0.17 and 0.02 for CHCls, CHBrCl2,
CHBr2Cl and CHBr3, respectively. A reasonable numerical value for £ is
0.008 when salinity is expressed in g/kg. Using these figures in
Equation (1), we found that CT is less than C^ in 35 g/kg seawater by
about 25% for CHC13 and by aBoiit 30% for CHBr3- The correction factor
diminishes with salinity, and is within the range of analytical uncertainty
for salinities below about 5 g/kg.
Our headspace method provides a measure of the amount of analyte that
is free and unassociated with molecules other than water. If part of the
analyte is adsorbed or bound in some non-volatile form, that fraction will
go effectively undetected. Table 1 presents some data on potential adsorb-
ing agents. These were added at concentrations higher than would be
TABLE 1. CONCENTRATION (yg/1) OF 100 ug/1 HALOFORM SOLUTIONS
ONE DAY AFTER ADDITION OF ADSORBING AGENTS
Haloform Adsoroing Agent
Huoic Acid (a) Fe(OH)3 (b) Clay (c)
CHC13 98 101 99
CHBrCl2 99 93 96
CHBr2Cl 101 93 95
CHBr3 100 , 89 92
luAldrich Chem. Co. lot no. 121137; 10 mg/1.
b. Precipitated from FeCl3 and aged 1 hr before exposure to
halofonrs; 10 mg/1.
c. Wyoming Na-Montmorillonite, Clay Minerals Repository,
University of Missouri; 20 ug/1.
encountered typically in coastal waters. Inasmuch ~s the measurement un-
certainty is about 4%, the data for humic acid do not show significant
absorption. In the presence of Fe(OH>3 and montmorillonite, there is
significant absorption for the heavier haloforms. However, in most
estuarine and marine waters where the concentrations of these adsorbing
agents would be many times less, the small effects seen in Table 1 would
-------
become negligible. Thus, we conclude that volatile halocarbons are likely
to be mainly in a free state in natural waters; therefore, concentrations
obtained by a headspace procedure, when corrected by Equation 1, approximate
total concentrations.
In concluding this section, it should be noted that Kopfler et al. (1976)
and Nicholson et al. (1977) have provided evidence for the existence of a
long-lived halo-organic intermediate in chlorinated natural waters that
decays to release haloforms upon heating. Thus, in the case of decay of such
an intermediate, the quantity of haloforms detected by our technique may be
greater than the quantity in the sample at the time it was collected.
LABORATORY CHLORINAT10N EXPERIMENTS
Sampling Sites
To study halocarbon production by chlorination of coastal waters, a
group of seven samples was collected from coastal, estuarine and fluvial
sites in Maryland (Figure 1). The sites were selected to avoid local sources
of either haloforms or chlorine. The samples were filtered with 0.45 urn
DELAWARE
BAY (:
Figure 1. Map showing sample locations. See Appendix
for latitude and longitude.
-------
Hillipore filters, and then stored in a cold room (4°C) prior to use in exper-
iments. Cblorination in the laboratory was conducted with calibrated NaOCl
stock solutions, rather than with Cl2 gas; under the conditions of the tests,
Cl2 hydrolyzes almost instantly to form OC1~, so our results should be the
same as would be obtained with Cl2- Total nitrogen data for these samples
were obtained by a UV oxidation-Cd reduction method (Sugam, 1977) . Total
organic carbon determinations were done by a commercial laboratory (Martel
Laboratories, Inc.), using an Oceanography International organic carbon
analyzer.
Results
Table 2 presents the results of a series of chlorination experiments in
which samples selected to represent a full range of water conditions, from
fresh water to ocean water, were treated with two concentrations of NaOCl in
TABLE 2. HALOFORMS PRODUCED IN SELECTED ESTUARINE AND COASTAL" WATER
SAMPLES (1J~9 moles/l)a
Sample Number
Salinity (g/kg)
PH
Org-C (UM)
Tot-N (uM)b
0
7
350
185
.00
.21
1.06
7.94
270
68
3
7
270
58
.68
.94
7
7
280
34
.90
.83
10
7
270
31
.27
.82
30.35
7.36
240
13
31.54
8.10
510
NAC
NaOCl dose 14 pM (=1 mg/l):d
CHC13 20 15 ND 3 ND NDC NA
CHBrCl2 ND 17 7 1 ND ND NA
CHBr2Cl ND 15 28 13 7 ND NA
CHBr3 ND 22 183 211 122 54 NA
Yield (%)& 0.43 1.48 4.67 4.88 2.76 1.15
Cl Demand (%)* 18 22 74 78 77 55 NA
NaOCl dose 140 pM (=10 mg/1) :
CHC13
CHBrCl2
CHBr2Cl
CHBr3
Yield (%)
763
287
87
ND
2.43
41
163
564
753
3.25
10
11
126
970
2.39
4
3
42
1160
2.59
8
1
27
900
2.00
6
ND
9
245
0.56
ND
ND
16
880
1.92
. . . ,
sample.
b, Tot-N - N0~ + N0~ 4- NH- + Org-N.
c. NA = not analyzed; ND - not detected.
d. 1 yM NaOCl has the same oxidizing capacity as 1 yM Cl?.
e. Yield = moles of halogen in haloforms as percent of moles of hypochlorite
added. Note that carbon is in excess in these tests.
f. Cl Demand = percentage of initial dose detectable after 10 min by ampero-
metric titration at pH 4.0 with excess KI added.
..6..
-------
the laboratory. These levels span the range of doses typically applied by
power plants and treatment plants. Haloforms were produced in each case; no
volatile halocarbons other than halofonns were produced at-the nM level in
any of the experiments. In the fresh water sample (No. 1), CHC13 was the
dominant haloform, as has been found in most drinking waters. Conversely,
CHBr3 overwhelmingly dominated in the ocean samples. There was a systematic
shift, as the salinity increased (i.e., left to right in Table 1), in the
proportion of bromine present in the haloform products. Figure 2 shows that
bromine dominated the products at salinities above about 5 g/kg.
100-
80
IO
X
geo
^40
i_
m
55 20
0
V
o 14 pM DOSE
• 140 pM DOSE
0
IO 15 20 25
SALINITY (g/kg)
30 35
Figure 2. Bromine as a percentage of total halogen in
haloform products.
The total haloform yield, expressed as a percentage of the chlorine
dose (i.e., the amount added initially), varied considerably (Figure 3), but
generally seemed to be higher at low-to-intermediate salinities. Absolute
haloform yields were roughly proportional to chlorine dose, so that percentage
yields were of similar magnitude for both dosages tested. Although on a
molar basis, organic carbon graatly exceeded chlorine dosage in both sets of
tests, its availability appeared nevertheless to exert some minor influence
on yield. (Samples 6 and 7 are very similar, except for organic carbon con-
tent; sample 7, with higher organic carbon, yielded much more haloform).
-------
o 14 jjM DOSE
o 140 JJM DOSE
10 15 20 25
SALINITY (g/kg)
30
Figure 3. Total halofonn yield as a percentage of chlorine
dose; chlorine added as NaOCl.
It is particularly interesting to note (Figure 4) the correlation
between the percentage of the chlorine dose that disappeared within 10
minutes and the yield of halofonr.s produced. This suggests that organic
matter plays a major role in chlorine consumption, even though only a small
fraction of the consumed chlorine is recovered in the form of volatile
halocarbons.
Discussion
When chlorine is added to seawater, bromide which is present at a
concentration of about 840 yM is rapidly oxidized to HOBr (e.g. Duursma
and Parsi, 1976; Eppley et al., 1976):
HOC1 + Br
HOBr + Cl
(2)
-------
Q
_J
LJ
ro
X
o I
0 20 40 60 80 100
CHLORINE DEMAND
Figure 4. Haloform yield as a percentage of chlorine dose
vs. total chlorine consumed in 10 min, as a
percentage of chlorine dose; ilose was 14 yM.
Using the ionization constant data of Sugara and Helz (1976) and the rate con-
stant data of Farkas et al. (1949), wi calculated the time needed for HOC1 to
be consumed by reaction 2 (see Figure 5). It was assumed in these calcula-
tions that the chlorine dose was small compared to the available Br~ so chlo-
rine consumption could indeed reach 99% completion. For typical doses, this
assumption must break down at low salinities. Also at pH of 9 and higher,
equilibrium between OC1~ and OBr~ is reached before 99Z of the hypochlorite
has been consumed. Nevertheless, the figure clearly shows that under near-
neutral conditions at salinities abave about 5 g/kg, consumption of HOC1 by
•Br~ will occur within less than l.min; in many cases, the conversion time
will actually be less than 10 sec. Thus, the dominance of Br in haloform
products formed in waters that contain more than 5 g/kg seasalt is readily
explained. As pointed out in a later section, these results imply that the
organic substitution reactions which give rise r.o haloforms are slow compared
to reaction 2.
The fact that haloforros are the only major volatile halocarbon product
of the direct chlorination of marine waters (as with drinking waters) implies
that a unique mechanism exists for their synthesis. Recently, Rook (1977)
suggested that halogen attack begins at the 2nd carbon position in a 1,3-
-------
1000
Figure 5.
1C 20 30
SALINITY (g/kg)
Calculated time for 99% consumption of free
chlorine by reaction 2 (see text).
dihydroxybenzene (resorcinol)-type unit in fulvic acid, thus leading to ring
cleavage and release of a haloform molecule. In freshwaters, 60-80Z of the
dissolved and particulate organic carbon typically consists of humic sub-
stances, including fulvic acid (Reuter and Perdue, 1977). Although a number
of workers have pointed out differences between organic matter in freshwater
and seawater (Kalle, 1966; Khaylov, 1968; Sieburth and Jensen, 1968; Stuenner
and Harvey, 1974; Kerr and Quinn, 1975), sufficient similarity apparently
exists, nevertheless, to justify common use of the term, fulvic acid, to
describe marine dissolved-organic matter. Our results indicate that, fluvial
and marine organic matter share the ability to produce haloforms.
10
-------
It is clear from ttese and other results (Rook, 197A) that the organic
nwtter that supplies carbon for the halofonns reacts preferentially with Br,
rather than Cl. For example, sample 3 which contained 86 yM Br~ (estimated
from salinity) could have consumed a maximum of only about 60% of the 140 uM
chlorine dose via reaction 2. Yet the excess chlorine is scarcely represented
in the haloform products (Table 2).
One likely reason why chlorine in haloforms is disproportionately low
is that reactions involving ammonia, organic amines and amino acids rapidly
tie up chlorine in chloramine compounds:
HOC1 + k-NH2 -»• R-NHC1 + H.O
HOC1 + R-NHC1 -* R-NC12 + H^O
These reactions are extremely fast (Morris, 1967). The products are oxidants,
but probably cannot react vith dissolved organic matter to produce haloforms.
Analogous bromamines also can form in marine waters, but, at least in the
case of the inorganic bromamines, their formation is reversible (Johnson and
Overby, 1971). Thus, chlorine that becomes nitrogen-bound may be permanently
removed from the haloform generation process, whereas nitrogen-bound bromine
may continue to be available.
COMPARISON OF CHLORINATION AND CZONATION OF ESTUARINE WATER
A few exploratory experiments were conducted in cooperation with Dr.
Ronald Block, of the Chesapeake Biological Laboratory, to compare the halo-
form-generating performance of chlorine and ozone. Those experiments employed
his bioassay system and natural estuarine water.
Methods and Results
Data on the water chemistry are given in Table 3. Chlorinated, ozonated,
and untreated control solutions were examined in parallel experiments run
simultaneously to insure that prior to treatment the water quality was the
same. For the chlorine experiments, Ca(OCl)2 stock solution (10.6 equiv/1)
was metered at 0.5 ml/min into a 3.8 1/min stream of estuarine water that
flowed continuously through an 80-1 glass tank; thus, the dosage was 139
Hequiv/1, or about 5 ppm Cl. Turnover time of water in this tank was roughly
20 min. The system was allowed to run for several hours until a steady-state
level of oxidant in the tank was observed. Then, a sample was collected in
a glass bottle and sealed so that no air-space remained in the bottle.
A second sample was obtained at the same time, but dechlorinated with
^28703 before sealing. Next, a dozen white perch (Morone americana) were
introduced into th« tank; after 30 min, two additional water samples (one of
which was quenched with Na2S203) were taken and sealed. In the ozone experi-
ments, the above procedure was conducted in a similar fashion, except that
the flowing water was treated with the oxygen-ozone gas effluent from a
Welsbach T-816 ozonator; dose was adjusted to give approximately the same
11
-------
oxidant level as in the chlorine testing tank. Turnover time of water in the
ozone tank was 8 rain. With the controls, all procedures were followed (i.e.
taking of a quenched and an unquenched sample and testing both in the pres-
ence and absence of fish). The samples were analyzed for volatile halo-
organics 18-24 hours after collection by transferring them to septum vials
and then following the method described previously.
Results
As shown in Table 3, bromoform concentrations in the neighborhood of
20 ug/1 were found in all the samples that had been Created with hypo-
chlorite or oznne. Furthermore, bromoform was not detectable in the con-
trols. This indicates that bromoform is produced by the oxidants and is
not introduced by sample handling or contamination.
TABLE 3. YIELDS OF BROMOFORM IN ESTUARINE WATER* TREATED WITH
CHLORINE (AS PCI") AND OZONE.
Fish Absent Fish Present
Hypochlorite (u-equiv/1) 23.8 21.2
Bromoform Yield (ug/1)
Na2S203 Quench 24.2 20.1
No Quench 25.8 19.8
Ozone (u-equiv/1) 23.2 21.8
Bromoform Yield (ug/1)
Na2S?03 Quench 19.0 17.0
No Quench 25.6 19.0
Controls
Na2S203 Quench <1 <1
No Quench <1 <1
*Cheraical characteristics of the water are: salinity «* 14 g/kg, pH = 8.0,
T = 15°C, NH3 < 0.1 mg/1, initial dissolved oxygen * 10 mg/1 (dissolved
oxygen in ozonated tank = 17.7 mg/1). All data are based on a single
analysis for each parameter in each sample.
In both the hypochlorite and the ozone tests, introduction of fish
lowered the steady-scate oxidant concentration and the amount of bromoform
observed. In the unquenched samples, the yield of bromoform is essentially
the same for both hypochlorite and ozone. Furthermore, in the hypochlorite
experiments, quenching with Na2S203 had little effect upon bromoform yield.
Discussion
The virtually identical haloform yields in the quenched and unquenched
chlorine samples indicate that the substitution reactions must be fast
compared to turnover time of water in the experimental tank (i.e., 20
minutes, or roughly 1000 seconds) because this is the average time between
addition of chlorine and addition of Na2S203 to the quenched sample. On
the other hand, the substitution reactions must be slow compared to reaction
12
-------
"(I) because bromoform is the only significant product. Based on Figure 5,
reaction (1) would reach completion in roughly 20 seconds at the salinity
and pH of these experiments. Thus, these observations bracket the order of
magnitude of the time needed to complete the halogen substitution.
It Is important to distinguish between the substitution reactions dis-
cussed above and the haloforro production reactions. Laboratory studies of
halofonn production from chlorinated solutions of fulvic or huraic acids,
believed to be the sources of carbon in the haloforms, suggest that this
production requires hours to reach completion (Stevens et al., 1976; Rook,
1976 and 1977). Further, Kopfler et al. (1976) and Nicholson et al. (1977)
have provided evidence for the existence of a long-lived halo-organic inter-
mediate that decays to release haloforms upon heating. Thus, the substitu-
tion of halogen atoms into organic structures in water, and the release
of haloform molecules from those organic structures, appear to be separate
processes with quite different rates. The release process apparently does
not require the presence of an oxidant residual, implying that in seawater,
it would be impossible to limit haloform production by instituting dechlori-
nation practice ut power plants unless the entire chlorination-dechlorination
cycle can be eccoraplished in a very short time span.
The excellent agreement in bromoform yield from ozone and chlorine is
somewhat surprising. In the case of the unquenched samples, the yield of
bromoform is the same within experimental error. For the quenched samples,
slightly less bromoform was obtained from ozone than from chlorine, probably
reflecting simply the shorter turnover time of water in the ozone-treated
tank. It would seem from these data that replacement of chlorine by ozone
as an antifoulant in marine power plants offers little advantage for mini-
mizing volatile halo-organics, unless it can be shown that ozone is basically
more effective than chlorine and can be used at lower doses.
The chief purpose of placing fish in the test tanks after the first set
of samples had been collected was to determine if reaction of the oxidants
with excretory products, body, or tissues produced a large increase
, in the haloform yield or in the Cl/Br ratio in the products. A large en-
hancement in yield could make haloforms potential toxic factors in bioassay
experiments intended to assess the acute toxicity of chlorine and ozone to
fish. However, the data indicate small decreases in bromoform content in
the tanks when fish are present. A small decline in the residual oxidant is
also observed. One possible explanation is that fish produce compounds
which react rapidly with oxidative biocides, but which generate no volatile
halo-organic compounds. Ammonia, urea, an! other nitrogenous products are
likely candidates. Probably they compete with the naturally occurring
bromoform precursors for the available halogen and thus lower bromoform
yield. An alternate hypothesis for the decrease in bromoform may be uptake
of this compound by the fish, either through absorption into body slimes or
transport into the body via the oouth or gills.
13
-------
SECTION 4
FIELD STUDIES
This section extends the previous work to the field and includes some
evidence regarding the fate of volatile halocarbons in the natural
environment.
CHALK POINT POWER PLANT
Site Description
Field work was done at a coal-fired power plant located on the Patuxent
River near Benedict, Maryland. Although the upper part of the Patuxent
drainage basin contains numerous residential suburbs and has a population
density of approximately 500 persons/tan^ (data from Maryland Environmental
Service, 1974), the area within 30 km of the plant, both upstream and
downstream, is dominantly rural with a population density closer to.50
persons/km^. The power plant is the only major industrial installation in
the lower part of the basin.
The salinity in the vicinity of the plant ranges with season from
virtually zero to about 15 g/kg (Cory and Nauman, 1971). Ammonia is
typically less than 0.1 mg N/l (Sugam, 1977).
The power plant contains two 355 megawatt generating units that each
circulate 16,000 I/sec (560 cfs) of estuarine water through twin single-pass
condensers. This water, which is continuously chlorinated, is discharged
into a 2.6-km-long cooling canal. At the head of the canal, the condenser
effluent is sometimes diluted with unchlorinated estuarine water by as much
as a factor of two to meet regulations limiting thermal loading.
Results
A large amount of data was collected on the inorganic aqueous chemistry
at this site (mostly under other sponsorship). Data are reported elsewhere
(Sugam, 1977) and only will be summarized here. As illustrated by the lower
two curves in Figure 6, more than 90% of the oxidant dose disappears by
the time the cooling water emerges from the plant. Comparable initial
decay has been observed by others CSnoeyink and Markus, 1974; Eppley, 1976;
Bongers et al., 1977), but rarely results in loss of as much of the initial
dose a* we observed at this site. We will refer to this phase of the decay
process as the fast phase. \t the pover plant, it occurs mostly under non-
photic conditions inside conduits and condenser plumbing. We could not
study this phase in detail because of inadequate access to the water inside
the plant.
14
-------
10
<
Q
O.i
1.0
CM
G
0.1
o>
E
0.01
0 30 60 90 120 150
TIME (min.)
Figure 6. Decay of chlorine-produced oxidants in Patuxent
estuarine water. Curve A - intake water chlori-
nated with NaOCl in the laboratory at room
temperature. Curve B - samples collected from
discharge canal at night. Curve C - samples col-
lected from discharge canal during the day. Analy-
ses by amperometric titration at pH 4.0 with excess
KI.
When samples from the intake canal were chlorinated in the laboratory
with doses comparable to those used in the plant (upper curve, Figure 6),
the amount of chlorine lost during the fast phase was much less than observed
at the power plant canal. Samples extracted from the plumbing upstream of
the condensers inside the plant resembled those chlorinated in the labora-
tory; that is, they did not yet display the large loss of chlorine subse-
quently observed in the canal.
15
-------
The fast phase is followed by a slow decay phase. In the discharge
canal, this always appeared to be first order in total oxidant concentration
with tk - 30-100 min. In figure 6, the dayime decay rate appear:, to be
slightly faster than the nighttime decay rate, suggesting that photochemical
processes might play a minor role during this phase. However, in other runs,
the night decay rates were faster so the general- importance of photochemical
processes is questionable.
.No free chlorine (determined by amperoroetric procedures) was ever
detected in the canal. On the other hand, in laboratory-dosed or precon-
denser samples, for which the fast dec^y of total oxidant was much less than
in the field, free chlorine persisted for substantial periods of time.
Probably because of the rapid loss of free chlorine, only traces of
haloforms (~1 pg/1) were found in discharge canal wccers thnt were collected
in gas-tight septum bottles and analyzed 24 to 48 hours later. Yet, 10-100
yg CEEr-j/1 was readily generated in intake water by chlorinating it tilth the
same dose used by the plant and then processing it as the discharge canal
samples were processed. Precondenser samples also yielded haloforms. For
example, one sample extracted from the conduit between the chlorine injection
point and the condenser inlet was found to contain 40 yg CHBT3/1 £.nd
traces of CHB^Cl when analyzed the next day. However, its total oxidant
residual after 1 hr was 11 yN, much higher than that observed in the canal
after a comparable period.
Discussion
The finding of only traces of haloforms in the discharge canal at Chalk
Point was a very surprising result in view of the -ease rith which haloforms
could be generated under laboratory conditions. Thus, some explanation is
required.
One suggestion might be that haloforms actually uid form, but were lost
(possibly by volatilization) before the water reached the sampling point.
In this report, some data are presented on the rates and mechanisms of loss
of volatile halocarbons from estuarine waters. These studies suggest no
mechanism that could remove halocarbons sc rapidly as tc result in their
absence from the canal only minutes downstream of the plant.
Instead, the negligible haloform yields are probably a consequence of
the remarkably rapid oxidant loss in the condensers. This then raises the
question, what controls the oxidant loss? Since the fast phase of decay
consumed more oxidant witnin the. Chalk Point plant than in experiments that
matched the plant's oxidant dose, one might suspect initially that some of
the oxidant consumption under the field conditions was caused by reaction
with the condensers and plumbing, or with accumulated slimes within them.
However, a thorough review of the replacement history of both the condensers
and the sacrificial anodes, along with estimates from metallurgical data of
the likely corrosion rate of Cu-Ni condensers, suggests that metallic
corrosion probably consumes at most only a small percentage of the chlorine
used annually. It is also possible to dismiss reduction by slimes or by-
products of slime organisms as a major factor with the following argument:
16
-------
Because this plant chlorinates continuously 5 months a year, biosynthetic
production of a hypothetical reducing substance per unit volume of water
would have to equal at least the rate of chlorine consumption per unit
volume. This exceeds 1 mg C12/1 during the ~5 min period that the water is
Inside the plant. If reductant production did not match oxidant consumption,
the store of the reducing substance would be depleted rapidly. Biosynthesis
of enough material from each liter of water to accomplish this much reduction
is inconceivable in the short time available.
It is more reasonable to argue that the greater fast oxidant consumption
observed in the power plant can be explained by thermal and hydrodynamic
effects in the water itself. The rate of chlorine decay in estuarine water
is strongly temperature-dependent, as seen in Figure 7. Thus, the elevated
temperatures in the condensers should enhance decay processes. Further, it
0.01
30 60 90 120
TIME (minj
Figure 7. Effect of temperature on decay in estuarine water
chlorinated in the laboratory.
will be shown that colloidal organic matter undergoes substitution and
deamination reactions with chlorine. It is quite likely that these reactions
are partly diffusion-controlled and thus subject to acceleration by the
intense agitation occurring inside the condensers.
17
-------
The search for reducing agents to consume the oxidant probably should
be directed toward the water itself. For most well-oxygenated estuarine
waters, it is possible to reject common inorganic reducing agents (e.g.
Fe2+, Mn2+, S2~, N02~) as major chlorine consumers based upon one or both of
the following grounds (Sugam, 1977): either the concentration of the agent
in its reduced form is small compared to typical chlorine doses, or
available kinetic data suggest that the redox reaction is too slow to explain
the large chlorine decay that occurs in the fast phase. The second argument
is always subject to the objection that there may be unrecognized catalytic
processes occurring under field conditions. Nevertheless, we have come to
believe that most chlorine consumption in estuarine waters (especially con-
sumption during the fast phase) must involve reactions mainly of the types
listed in Table 4; this list is intended only to provide examples of reaction
types, not a definitive list of possible reactions.
TABLE 4: EXAMPLES OK POSSIBLE OXIDANT CONSUMPTION REACTIONS IN
ESTUARIES. (X - Cl or Br)
Carbon Reactions (Substitution, Oxidation)
R_
3H + nHOX -»•
2. R-C=0 + HOX -»• R - C=0 + HX
H H
3. R-i-COOH + HOX -»• R-C=0 + NH_ + CO- +
Nitrogen Oxidation;
4. 2NH, + 3 HOX •*• N, + 3 HX + 3H20
Self-Decomposition;
5. HOX -»• HX + %0,
6. nHOX -•• (n-1) HX + HXO
n
HX
18
-------
Reactions with Organic Carbon
That organic carbon plays a major role in the fast decay processes seems
Inescapable. As seen in Figure 8, the rate of chlorine decay in an estuarine
water sample can be significantly slowed if the sample is first subjected
to ultrafiltration (Amicon UM 2) before chlorination. Analysis shows that
the material removed by this treatment is mostly organic. The rate is more
dramatically reduced if the sample is first passed through an XAD column to
remove organic molecules. Eppley et al. (1976) reported that most of the"'
fast decay in Southern California coastal water could be eliminated if the
water was treated by a UV oxidation procedure to remove organic carbon
prior to chlorination. Crane (personal communication, EPA Bears Bluff
Field Station) has found an excellent correlation between water color (pre-
sumably a measure of dissolved humic compounds) and chlorine demand in a
South Carolina estuary.
ULTRAF1LTERED
30 60 90 120
TIME (min.)
Figure 8. Effect of removing organic matter on decay.
Ultrafiltration employed Amicon UM-2 membrane
filters.
19
-------
However! it is inconsistent to argue on the one hand, that oxidant is
consumed mostly by reaction with organic carbon while also arguing that
negligible haloform production occurred because of the rapid disappearance
of oxidant. One way to resolve this conflict is to hypothesize that the
combination of elevated temperature and intense agitation in the condensers
increases the availability of reactive sites on the entrained organic floes
and particles. Thus, instead of getting multiple halogen substitution at
relatively few carbon atoms, a necessary condition for haloform production,
one gets single halogen substitution at many different carbon atoms. Accord-
ing to this hypothesis, the resulting halogenated macromolecules are re-
latively stable and fail to split-off low molecular weight fragments which
would be detectable by methods based on gas chromatography.
Support for this hypothesis was obtained by isolating (by ultrafiltra-
tion) the macromolecular organic matter from samples collected within a few
minutes of each other in the intake and the discharge canals at Chalk Point.
The macromolecular matter was then analyzed, the halogens being determined
by neutron activation analysis. The results (Table 5) indicate that both
samples are very similar except for the bromine content in the sample from
the discharge canal. We infer that this bromine is bound to the macro-
molecules as a result of chlorination. The data also show significant
chlorine in both these samples, but we believe this is simply residual salt,
rather than carbon-bound chlorine because it does not increase from the
intake to the discharge canal.
TABLE 5. MACROMOLECULAR ORGANIC MATTER
Item
C %
H %
•K %
Cl %
Br %
I %
0 + S (difference) %
Ash %
C/N (molar)
Intake
Canal
39.72
5.80
5.69
0.81
0.04
0.0030
47.94
100.00
6.40
8.14
Discharge
Canal
39.31
6.18
4.77
0.85
1.60
0.0026
47.29
100.00
6.42
9.61
Amount Recovered
(mg 1-^
1.31
1.29
Reactions with organic carbon, although undoubtedly important, can be
over-emphasised. Other reactions listed in Table 4 probably are also im-
portant, especially in waters that contain very little organic carbon, but
since these reactions are not directly relevant to the subject of this re-
port, they will not be discussed here.
20
-------
BACK RIVER WASTEWATER TREATMENT PLANT
When first encountered, the very low halofonn yields at Chalk Point
caused concern that perhaps haloforms were being produced, but were lost
rapidly or masked by other components in the water. To evaluate this possi-
bility, we decided to study a halocarbon source for which we would not have
to depend upon halocarbon synthesis from chlorine. We chose a large sewage
treatment plant located on the Back River estuary, east of Baltimore,
Maryland, where we established that a suite of volatile halocarbons was pre-
sent in the effluent even before chlorination. Thus,'the vagaries of the
halocarbon production process were eliminated from the experimental design.
Figure 9 is a map of Back River, a shallow, 12-km-long tributary estuary
to the Chesapeake Bay with a salinity range of about 0.4 g/kg. Its mean
depth is about 1 m and it is well mixed vertically. Near its upper end,
Back River receives 1.5 to 1.9 x 108 1 of wastewater/day from Baltimore's main
sewage treatment plant; the waste discharges often exceed freshwater flow
from the watershed by a factor of 2 (Helz et al., 1975). The plant provides
1002 secondary treatment (mostly by the trickling filter process) to domestic
and commercial wastes. The effluent is chlorinated before discharge.
BACK RiVER
Figure 9
Back River estuary, Maryland, showing locations
of sampling points. See Appendix A for latitudes
and longitudes.
21
-------
The first series of samples from Back River (Numbers 8 to 12) were
collected in early February, 1977, after the northern Chesapeake Bay had
been covered with ice for more than one month. The only uncovered area was
a 0.2 km diameter patch of water immediately above the underwater diffusers
at the discharge point in mid-river. The second set of samples (numbers
13 to 23) was collected in early May, 1977, six weeks after the spring thaw.
Results
The vo]atile halocarbons foand in Back River (Table 6) are very dif-
ferent from those produced by chlorinating estuarine waters (Table 2). The
Back River assemblage of compounds must be derived from the numerous
commercial and light-industrial firms which the treatment plant serves.
Although the same assemblage of compounds was found in February and
May, the amounts of each compound were much lower in ^he latter month. This
difference may be due to a slower rate of halocarbon disposal into the
.sevage system in May or to an increased halocarbon volatilization rate during
the trickling filter process. The first seems less likely because we know
of no reason why the disposal rate should change drastically with season.
The large differences between the February and May in-plant samples (8 and
14) are reflected by all the samples from the receiving waters; this rules
out the possibility that the February in-plant sample was spurious because
a halocarbon slug was passing by chance through the plant when the sample
was taken.
TABLE 6. HALOCARBONS (uH) IN BACK RIVER
Sample* Salinity CH2C12 CHC13 CCl^ CC12=CHC1 CC12=CC12
February 1977 (Ice cover);
8 - 775 413 186 230 369
9 0.57 512
10 0.69 304
11 1.42 152
12 2.92 ND
13 3.16 ND
May 1977 (Open water);
14b - 47
15 0.34 ND
16 0.36 24
17 0.35 9
18 0.34 ND
330 97
238 60
207 57
18 ND
ND ND
106 14
3 ND
71 3
8 2
1 ND
19-23 No volatile halocarbons detected seaward of
113
88
50
2
ND
106
4
75
14
2
station 18
356
337
114
5
ND
50
ND
35
4
ND
.
a. Sample numbers refer to
Figure 10.
b. Samples 8 and 14 were taken in the treatment
final chlorination.
plant just
before
22
-------
In a seaward direction, the concentration of volatile halocarbons
decreases systematically. The dominant process causing this decrease is
simply tidal mixing of the effluent with saline Chesapeake Bay water. How-
ever, if dilution were the only process (i.e. , if halocarbons mixed conser-
vatively), then a plot of halocarbon concentration vs. specific conductance
(measured at 25°C) would be linear (Boyle et al., 1974; Helz et al., 1975).
However, such plots were non-linear in winter (Figure 10). Since Back River
was almost entirely covered with ice, the non-conservative behavior cannot
be ascribed to volatilization. When May samples were collected, there was
not sufficient contrast between the specific conductance of the effluent and
of the water off the mouth of Back River to allow these data to be treated
as in Figure 10. However, it is probably significant that the seaward de-
cline in halocarbon concentrations was much more abrupt in May. Values near
or below the detection limit wera encountered within 2 km of the outfall in
May. In contrast, in the February series, values at station 12 (about 4 km
from the outfall) were only 20-50% of the effluent values. More rapid loss
in the later samples may be due either to volatilization in the absence of
ice cover or possibly to more active biodegradation.
Discussion
Table 7 lists the major industrially produced, volatile halocarbons
ranked according to rate of release to the environment. Release rates
were determined by Nelson and Van Duuren (1975) by estimating for the
United States the fraction of production plus imports used dispersively,
plus the fraction of production lost during manufacture. Of the 11 compounds
listed, three (CC12F2, CC^F and CH2Br-CH2Br) would not be expected in waste-
waters because they are released directly to the atmosphere. The two fluoro-
carbons are used mainly as refrigerants and aerosol propellants, and CH2Br-
CH2Br is used mainly as a gasoline additive. A fourth compound, CH2=CHC1
(vinyl chloride), is probably too volatile to be retained in wastewater
during treatment, even if it were somehow present initially. Of the remaining
seven compounds in Table 7, five were found at the nM level in Back River.
It is not known whether the other two, CH2C1-CH2C1 and CH3-CC13, are absent
because they are not discharged into the sewage system in significant
quantities, or because they are degraded or adsorbed from the wastewater
before leaving the plant. Octanol-water partition coefficients, estimated
by the method of Leo (1975), have been included in Table 7. These co-
efficients are considered a good measure of the lipophilic character of a
compound and have been shown to correlate with the tendency of a compound
to be bioactive and to bioaccumulate (Neely et al., 1974). However, neither
of the two missing compounds in Table 7 is particularly distinctive with
respect to its partition coefficient or to its Volatility. Interestingly,
no clear relationship exists between the partition coefficient and the de-
gree of curvature in the concentration vs. salinity curves in Figure 10.
CH2C12 (with the greatest curvature) and CCl2=CCl2 (with the least) have
essentially the same partition coefficient.
23
-------
CH,CIa
600
400
200
0
V
CHCU
200
100'
i
CCL
200
CCLjCHCl
300
200
100
0123456
SPECIFIC CONDUCTANCE (mmho)
Figure 10. Halocarbon patterns in Back River. Vertical
scales are all In nanomolar units for the
indicated halocarbon.
24
-------
TABLE 7: MAJOR INDUSTRIALLY PRODUCED HALOCARBONS RANKED BY RELEASE RATE
Compound
Release Rate Boiling Point
(109 g yr-1)
log p
oct
In Back River
cci2=cci2
cn2ci-CH2ci
CHC1-CC12
CC12F2
CH2C12
CH2Br-CH2Br
CH3-CC13
CC13F
CHC13
CH2-CHC1
CCIA
267
245
195
183
167
138
129
124
76
67
48
121
83
87
-28
40
131
74
24
61
-14
77
1.27
1.48
0.96
2.08
1.26
1.76
2.50
2.52
1.96
0.34
2.96
yes
no
yes
no
yes
no
no
no
yes
no
yes
The much less rapid decrease in halocarbon concentrations when Back
River was covered with ice suggests that volatilization is an important loss
process. Dilling et al. (1975) and Dilling (1977) established that a large
variety of C].-C3 halocarbons escape from open, stirred beakers in the labor-
atory with half lives (T) on the order of 20 min. Liss and Slater (1974)
have developed a simple model by which the rates of diffusive exchange of
gases across the air-sea interface may be estimated. The model, as applied
to the volatile halocarbons, leads to the following decay equation (Mackay
and Leinonen, 1975):
(A)
C_ - C_ exp (-K, t/L)
Here, CQ_ is the initial concentration, and Ct is the mean concentration at
tisne=t£i; L_ is the mean depth of the water. "Values of K^ are presented in
Table 8. ~The half-life of ^ compound in a well mixed water body would be:
0.69 L K
-1
C5)
Be-
Values of T for two arbitrary m«an depths are also given in Table 8.
cause the validity of the simple Liss-Slater model may be subject to
challenge, especially if applied to organic-rich, contaminated water around
waste discharges, the half-lives in Table 8 must be regarded as tentative
estimates. However, the important implication, that relatively shallow,
well-mixed coastal waters will lose volatile halocarbons to the atmosphere
within a few days, is probably correct.
25
-------
It Is noteworthy that the half-lives for all the compounds in Table
8 are similar in magnitude. This implies that although volatilization will
cause fairly rapid decline in the concentration of any one compound, it will
cause the concentration ratio between two compounds to change only slowly.
Thus, these ratios could be "fingerprints" for a particular discharge plume
and might be useful as plume tracers.
TABLE 8.' LOSS RATES FOR VOLATILE HALOCARSOSS
CC1 =CC1.
2 2
CHC1=CC12
CH-C1.
2 2
CH,CC1,
3 3
CHC13
cci4
CHBrei_
CHBr2Cl
CHBr,
K. (cm hr )
10.2
11.3
13.1
11.4
11.4
10.6
8.9
6.8
4.1
T (hrs)
L = 1m
6.8
5.9
5.3
6.0
6.0
6.5
7.8
10.2
16.9
L = 5m
34
30
26
30
30
33
39
51
85
Loss of volatile halocarbons to the atmosphere Is likely to constitute
a permanent sink from the aquatic geochemist's viewpoint. Published
estimates for the lifetime of these compounds in the troposphere range from
a few days for the unsaturated compounds to a thousand years for CC14 (Yung
et al., 1975; Dilling et al., 1976; Galbally, 1976), However, even in con-
taminated urban air, concentrations of these compounds are usually less than
1 x 10~9 by volume. Since the Henry's law constants (Hilling, 1977) are
mostly less than unity, rain would contain mu"h less than 1 nM, even if
saturated. This is below the concentrations found In this study; therefore
in contaminated coastal waters, rain will be a diltttent rather than a source.
Measurements in rainwater by Pearson and McConnell (1975) support this
conclusion.
Whereas dilution ar.d volatilization are probably the major processes
that control concentration patterns of volatile hydrocarbons in transport
away from a discharge point, the.February Back River data show that these
compounds are consumed to some extent even when volatilization is precluded
by Ice cover. A number of processes might contribute to this. Hydrolysis
or similar abiotic chemical mechanisms might be involved, although limited
available information suggests that these processes will be slow. For
example, Zafiriou (1975) estimated that the 50% conversion of CH3C1 to
CH30H in seawater would take 2.5 yr at 20°C and 14 yr at 10°C. Dilling et
al. (1975) determined half lives of 9 to 18 months for 5 chlorocarbons in
oxygenated solutions. Exposure of the solutions to sunlight increased the
26
-------
"-_decay rate by a factor of 1.5 to 2.0 for olefinic compounds, but had little
effect on saturated compounds. The products in these reactions were not
: determined.
, Whether biochemical pathways offer a faster consumption mechanism for
| volatile halocarbons is not yet clear. A brief attempt by Jensen and
'-Rosenburg (1975) to answer this question yielded inconclusive results.
1 A potentially important removal mechanism in Back River involves trans-
j fer into the sediments. Although our results and those of others suggest
that volatile halocarbons are inefficiently adsorbed by such substances as
humic acid, ferric hydroxide, and clay (montmorillonite), there may be
'• biologic pathways in which the compounds become bound in fecal pellets
; or other biogenic debris. Furthermore, the sediments in Back River have the
consistency of black mayonnaise and may act somewhat like a non-polar liquid,
simply extracting the dissolved halocarbons from the overlying aqueous
| layer. However, two attempts to establish the presence of these compounds
! in Back River sediments yielded negative results. In one experiment, 5 g
j of sediment was placed in a septum vial with 40 ml of H2O; after sealing,
• Che vial was shaken for an hour at 60 C, and the headspace above the slurry
i then analyzed. In the second experiment, & Soxhlet extraction of 50 g of
sediment was performed for 1 week with 9:1 benzene-methanol. These experi-
ments indicate that if volatile halocarbous are transferred into the Back
River sediments, then they are probably degraded there, presumably by
anaerobic biologic processes. ;
-------
SECTION 5
OTHER ACTIVITIES
WORKSHOP
In the Spring of 1976, a workshop on the fate and effects of chlorine
in Coastal Waters was held at Solomons, Maryland, with financial support
from this grant. The workshop was organized jointly by the principal
investigator and Dr. Ronald Block of the University of Maryland. About 30
scientists from as far away as Great Britain participated. Althougn there
were 15 scheduled talks in the 2\ days of the workshop, emphasis was placed
on informal discussion of problems and research needs associated with the
use of chlorine in marine waters.
A full report on the workshop has been published (Chesapeake Science,
18, 96-160), and reprints of this report have been provided to EPA. Among
the research needs identified in that report are a need to understand the
reaction paths and mechanisms by which chlorine disappears in marine water,
a need to account for all the products and a need to improve analytical
procedures. Bioassay methodology was a major topic of discussion at the
Workshop and some recommendations regarding the design of bioassays have
been cade in the report.
PUBLICATIONS
The following publications have resulted from work on this project:
Sugam, R., and Helz, G. R., 1977, Speciation of chlorine produced oxidants in
marine waters: theoretical aspects. Chesapeake Science, 18:113-115.
Block, R., G. R. Helz, and W. P. Davis, 1977, The fate and effects of chlorine
in coastal waters, Chesapeake Science, 18:97-101.
Helz, G. R., R. Y. Hsu, and R. M. Block, 1977, Bromoform production by oxida-
tive biocides in marine waters. In: Ozone-Chlorine Dioxide Oxidation
Products of Organic Materials, R. G. Rice, J. A. Cotruvo and M. E.
Browning, eds. International,. Ozone Institute, pp. 69-76.
Helz, G. R., R. Sugam, and R. Y. Hsu, 1978, Chlorine degradation and volatile
halocarbon generation in estuarine waters. In: Water Chlorination:
Environmental Impact and Health Effect, R. L. Jolley, D. H. Hamilton
and, H. Gorchev, eds. 2, pp. 209-222.
Helz, G. R., and R. Y. Hsu, 1978, Volatile chloro- and bromocarbons in coastal
waters, Limnol. Oceanogr., 23:858-869.
28
-------
In addition to these publications, oral presentations have been made at
the annual meeting of the American Geophysical Union (1977), at the 2nd
Conference on Water Chlorination in Gatlinburg, Tenn. (1977) and at the
International Symposium on the Analysis of Hydrocarbons and Halogenated
Hydrocarbons in Hamilton, Ont. (1978).
29
-------
REFERENCES
Bongers, L. 1973. Assessing the impact of power plants on Maryland's water-
ways. Record of the Maryland Power Plant Siting Act. 2:1-4.
Bongers, L. H., T. T. O'Connor, and D. T. Burton. 1977. Bromine chloride -
an alternative to chlorine for fouling control in condenser cooling
systems. EPA 500/7-77-053, U.S. Envrionmental Protection Agency,
Cincinnati, Ohio.
Boyle, E., R. Collier, A. T. Denglar, J. M. Edmond, A. C. Ng, and R. F.
Stallard. 1974; On the chemical Mass-balance in estuaries. Geochim.
Cosmochim. Acta. 38:1719-1728.
Brush, L. M. 1972. Domestic and municipal waste loading to Chesapeake Bay.
In: Annual Report, June 1, 1971 to May 31, 1972. Chesapeake Research
Consortium Inc., Baltimore, MD.
Cory, R. L., and J. W. Nauman. 1971. Water quality at Patuxent River Bridge,
MD, January 1969 through November 1969. U.S. Geological Survey Open
File Report. 71 pp.
Dilling, W. L. 1977. Interphase transfer processes. II. Evaporation rates
of chloro methanes, ethanes, ethylenes, propanes, and propylenes from
dilute aqueous solutions. Comparisons witli theoretical pr?dictions.
Environ. Sci. Technol. 11:405-9.
Dilling, W. L., C. J. Bredeweg, and N. B. Tcfertiller. 1976. Organic photo-
chemistry. Simulated atmospheric photodecomposition rates of methylene
chloride, 1,1,1-trichloroethane, trichlorethylene, tetrachloroethylene
and other compounds. Environ. Sci. Technol. 10:351-6.
Dilling, W. L., N. B. Tefertiller, and G. J. Kallos. 1975. Evaporation rates
and reactivities of methylene chloride, chloroform, 1,1,1-trichloro-
ethane, trichloroethylene, tetrachloroethylene and other chlorinated
compounds in dilute aqueous solutions. Environ. Sci. Technol. 9:833-8.
Duursma, E. K., and P. Parsi. 1976. Persistence of total and combined chlo-
rine in seawater. Neth. J. Sea Res. 10:192-214.
Eppley, R. W., E. H. Renger, and P. M. Williams. 1976. Chlorine reactions
with seawater constituents and the inhibition of photosynthesis of
natural marine phytoplankton. Estuarine Coastal Mar. Sci. 4:147-161.
Farkas, L., M. Lewin, and R. Bloch. 1949. The reactions between hypochlorite
and bromides. J. Am. Chem. Soc. 71:1988-1991.
30
-------
Galbally, I. E. 1976. Man-made carbon tetrachloride in the atmosphere.
Science. 193:573-6.
Helz, G. R., R. J. HuggeLt, and J. M. Hill. 1975. Behavior of Mn, Fe,.Cu,
Zn, Cd, and Pb discharged from a wastewater treatment plant into an
cstuarine environment. Water Res. 9:631-6.
Jensen, S., and R. Rosenberg. 1975. Degradability of some chlorinated
aliphatic hydrocarbons in seawater and sterilized water. Water Res.
9:659-61.
Johnson, J. D., and R. Overby. 1971. Bromine and bromamine disinfection
chemistry. J. Sanit. Eng. Div. Proc. Am. Soc. Civ. Eng. 97:617-28.
Jolley, R. L. 1974. Determination of chlorine-containing organics in
chlorinated sewage effluents by coupled 36C1 tracer-high resolution
chromatography. Environ. Lett. 7:321-340.
Kalle, K. 1966. The problem of the gelbstoff in the sea. Oceanogr. Mar.
Biol. Annu. Rev. 4:91-104.
Kerr, R. A., and J. G. Quinn. 1975. Chemical studies on the dissolved
organic matter in seawater: isolation and fraccionation. Deep Sea
Res. 22: 107-16.
Khaylov, K. M. 1968. Dissolved organic macromolecules in seawater. Geochem.
Int. 5:497-503.
Kopfler, F. C., R. G. Melton, R. D. Lingg, and W. E. Coleman. 1976. GC/MS
determination of volatiles for the national organics reconnaissance
survey (NORS) of drinking water. In: Identification and analysis of
organic pollutants in water, L. H. Keith, ed. Ann Arbor Science
Publishers, Ann Arbor, Michigan.
Leo, A. J. 1975. Calculation of partition coefficients useful in the evalu-
ation of relative hazard's of various chemicals in the environment. In:
Structure-activity correlations in studies of toxicity and bioconcentra-
tion with aquatic organisms, D. G. Veith, and D. E. Konasewich, eds.
Great Lakes Research Advisory Board, p. 151-176.
Liss, P. S., and P- G. Slater. 1974. Flux of gases across the air-sea
interface. Nature. 42:181-4.
31
-------
Mackay, D., and P. J. Leinonen. 1975. Rate of evaporation of low-solubility
contaminants from water bodies to atmosphere. Environ. Sci. Technol.
9:1178-1180.
Maryland Environmental Service. 1974. The Patuxent River Basin Water Quality
Management Plan. Volumes 1 and 2.
Morris, J. C. 1967. Kinetics of reactions between aqueous chlorine and
nitrogen compounds. In: Principles and applications of water chemistry,
S. D. Faust, and J. V. Hunter, eds. Wiley and Sons, New York. pp. 23-
53.
Neely, W. B., D. R. Branson, and G. E. Blau. 1974. Partition coefficient
to measure bioconcentration potential of organic chemicals in fish.
Environ. Sci. Technol. 8:1113-15.
Nelson, N., and B. Van Duuren. 1975. Final report of National Science
Foundation workshop panel to select organic compounds hazardous to the
environment. Unpublished.
Nicholson, A. A., 0. Meresz, and B. Lemyk. 1977. Determination of free and
total potential haloforms in drinking water. Anal. Chem. 49:814.
Pearson, C. R., and G. McConnell. 1975. Chlorinated Cj and C2 hydrocarbons
in the marine environment. Proc. R. Soc. London. 189:305-32.
Reuter, J. H., and E. M. Perdue. 1977. Importance of heavy metal-organic
matter interactions in natural waters. Geochim. Cosmochim. Acta.
41:325-34.
Rook, J. J. 1974. Formation of haloforras during chlorination of natural
waters. Water Treat. Exam. 23:234-43.
Rook, J. J. 1976. Haloforms in drinking water. J. Am. Water Works Assoc.
68:168.
Rook, J. J. 1977. Chlorination reactions of fulvic acids in natural waters.
Environ. Sci. Technol. 11:478-83.
Sieburth, J. M., and A. Jensen. 1968. Studies on algal substances in the
sea. I Gelbstoff (humic material in terrestrial and marine waters).
J. Exp. Mar. Biol. Ecol. 2:174-189.
32
-------
Stevens, A. A.i C. S. Slocum, P. R. Seeger, and G. G. Robeck. 1976. Chlori-
nation of organics in drinking water. In: Proceedings Conference on
the environmental impact of water chlorination, R. L. Jolley, ed. Oak
Ridge Nat. Lab. CONF-751096. p. 85.
Snoeyink, V. L., and F. I. Markus. 1974. Chlorine residuals in treated
effluents. Water Sewage Works. 121:35-45.
Stuenner, D. H., and G. R. Harvey. 1974. Huraic substances from seawater.
Nature. 250:480-1.
Sugam, R. 1977. Chlorine degradation in estuarine waters. Unpublished Ph.D.
Dissertation, University of Maryland.
Sugam, R., and G. R. Helz. 1976. Apparent ionization constant of hypo-
chlorous acid in seawater. Environ. Sci. Technol. 10:384-6.
Yung, Y. L., M. B. McElroy, and S. C. Wofsy. 1975. Atmospheric halocarbons:
A discussion with emphasis on chloroforai. Geophys. Res. Lett. 2:397-9.
Zafiriou, 0. C. 1975. Reaction of methyl halides with seawater and marine
aerosols. J. Mar. Res. 33:75-81.
33
-------
APPENDIX
LATITUDE AND LONGITUDE OF STATIONS
Station
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
IS
19
20
21
22
23
Latitude
38°
39 o
38°
38°
38°
38°
38°
39°
39 »
39°
39 o
39°
39 o
399
39 o
39 o
39°
39°
39°
39°
39°
39 c
39 o
49'
11'
33'
25 r
20'
14'
14'
17'
17'
17*
16'
14'
15'
17'
17'
17'
17'
17'
16'
16'
15'
14'
14'
40"
40"
15"
25"
00"
50"
50"
45"
50"
25"
20"
40"
35"
45"
45"
30"
10"
00"
50"
25"
35"
45"
45"
Longitude
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
76°
42'
16'
25'
35'
20'
7'
7'
29'
28'
27'
26'
24'
23'
29'
28'
28'
27'
27'
27'
26'
26'
25'
24'
25"
45"
20"
35"
50"
55"
10"
20"
15"
55"
25"
20"
55"
20"
40"
15"
55"
40"
15"
40"
45"
50"
40"
34
------- |