PRE-PUBLICATION COPY
NATIONAL OR6ANICS RECONNAISSANCE SURVEY FOR
HALOGENATED ORGANICS IN DRINKING WATER
WATER SUPPLY RESEARCH LABORATORY
METHODS DEVELOPMENT AND QUALITY ASSURANCE LABORATORY
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U,S, ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO
APRIL 1975
-------
PRE-PUBLICATION COPY
NATIONAL ORGANICS RECONNAISSANCE SURVEY FOR HALOGENATED ORGANICS IN DRINKING WATER
by
James M. Symons
Thomas A. Beliar
J. Keith Carswe11
Jack DeMarco
Kenneth L. Kropp
Gordon G. Robeck
Dennis R. Seeger
Clois J. Slocum
Bradford L. Smith
Alan A. Stevens
Water Supply Research Laboratory
Methods Development and Quality Assurance Laboratory
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio
April 1975
-------
NATIONAL ORGANICS RECONNAISSANCE SURVEY FOR HALOGENATED ORG.ANICS IN DRINKING WATER*
Introduction
On Friday, November 8, 1974, Russell E. Train, Administrator of the U.S.
Environmental Protection Agency announced that he was ordering an immediate
nationwide survey to determine the concentration and potential effects of
certain organic chemicals in drinking water. On that date Train said, "What
we learn from this National Reconnaissance Survey will tell us how widespread
and serious the situation is that we found in the study of the New Orleans
drinking water supply."
On December 16, 1974, President Ford signed into law Public Law 93-523,
"The Safe Drinking Water Act." Section 1442(a)(9) of this Act states "The
Administrator shall conduct a comprehensive study of public water supplies and
drinking water sources to determine the nature, extent, sources of, and means
of control of contamination by chemicals or other substances suspected of being
carcinogenic. Not later than 6 months after the enactment of this title, he
shall transmit to the Congress the initial results of such study, together with
such recommendations for further review and corrective action as he deems
appropriate."
Finally, on December 18, 1974, Administrator Train named the 80 cities
to be included in the National Organics Reconnaissance Survey (NORS).
*Submitted to the Journal American Water Works Association for publication.
-------
- 2 -
Objectives
NORS had three major objectives. One, was to determine the extent of
*+@
the presence of the four trihalomethanes -- chloroform (trichloromethane),
*@ *g *@
bromodichloromethane , dibromochloromethane , bromoform (tribromomethane),
and 1,2-dichloroethane , and carbon tetrachloride in finished water. A
second objective was to determine what effect raw water source and
water treatment practices have on the formation of these compounds. The
third was to characterize, as completely as possible using existing
analytic technology, the organic content of finished drinking water produced
from raw water sources representing the major categories in use in the United
States today. This paper discusses the results of NORS that are related
123
to the first two objectives. Future papers will discuss objective three ' '
Selection of Cities
For the study of the formation of chlorination by-products, 80 water
supplies (Table 1) were chosen to participate in the NORS in consultation with
State water supply officials. These 80 supplies were geographically distributed,
some in each of the USEPA's 10 Regions, see Figure 1. The supplies were chosen
to represent as wide a variety of raw water sources and treatment techniques
as possible.
* - Selected as a possible chlorination by-product,
+ - Selected because of suspected effect on health.
@ - Selected because of presence in previously sampled finished waters.
-------
- 3 -
d
>-
u
o
<
z
o
u
LU
o
ex
O
I!
CD
to y
O
-------
- 4 -
TABLE 1
Water Utilities Studied
Region I
1.** Lawrence Water Works
Lawrence, Massachusetts
Merrimack River*
2. Waterbury Bureau of Water
Waterbury, Connecticut
Wigwam and Morris Reservoirs
Morris Treatment Station
3. Metropolitan District Commission
Boston, Massachusetts
Quabbin and Wachusett Reservoirs
Norumbego Treatment Station
4. Newport Department of Water
Newport, Rhode Island
Reservoirs
South Pond Reservoir Treatment
Plant #1
Region II
5. Department of Water Resources
New York, New York
Croton Reservoir
6. Puerto Rico Aqueduct and Sewer
Authority
San Juan, Puerto Rico
Lake Carraizo
Sergio Cuevas Water Treatment Plant
7. Passaic Valley Water Commission
Little Falls, New Jersey
Passaic River
8. Toms River Water Company
Toms River, New Jersey
Ground
Well #20
9. Buffalo Water Department
Buffalo, New York
Lake Erie
10. Village of Rhinebeck Water Dept.
Rhinebeck, New York
Hudson River
Region III
11. Philadelphia Water Department
Philadelphia, Pennsylvania
Delaware River
Torresdale Plant
12. Wilmington Suburban Water Corp.
C1aymont, Delaware
Red Clay and White Clay Creek
Stanton Plant
13. Artesian Water Company
Newark, Delaware
Ground
Llangollen Well Field Plant
14. Washington Aqueduct
Washington, D.C.
Potomac River
Delacarlia Plant
15. Baltimore City - Bureau of
Water Supply
Baltimore, Maryland
Loch Raven Reservoir
Montbello Plant #1
16. Western Pennsylvania Water Company
Pittsburgh, Pennsylvania
Monongahela River
Hays Mine Plant
17. Strasburg Borough Water System
Strasburg, Pennsylvania
Ground
18. Fairfax County Water Authority
Annandale, Virginia
Occoquan River Impoundment
New Lorton Plant
*The name of the utility is listed first, followed by the city name, the name
of the raw water source, and the name of the treatment plant sampled, if the
utility has more than one treatment plant.
**The same location numbers will be used throughout the paper.
-------
- 5 -
19. Virginia American Water Co. -
Hopewell District
Hopewell, Virginia
Appamatox River
20. Huntington Water Corp.
Huntington, West Virginia
Ohio River
21. Wheeling Water Department
Wheeling, West Virginia
Ohio River
Revion IV
22. Miami-Dade Water and Sewer
Authority
Miami, Florida
Ground
Preston Plant
23. Jacksonville Dept. of Public
Works
Jacksonville, Florida
Ground
Highlands Pumping Station
24. Atlanta Waterworks
Atlanta, Georgia
Chattahoochee River
Chattahoochee Plant
25. Owensboro Municipal Utilities
Owensboro, Kentucky
Ground
26. Greenville Water Department
Greenville, Mississippi
Ground
Water Plant Well #2
27. Tennessee American Water Company
Chattanooga, Tennessee
Tennessee River
28. Memphis Light, Gas and Water Div.
Memphis, Tennessee
Ground
Malloy Plant
29. Metropolitan Water and Sewerage Dept.
Nashville, Tennessee
Cumberland River
Lawrence Plant
*Resampled after GAC changed.
30., Commissioner of Public. Works
Charleston, South Carolina
Edisto River
Stoney Plant
Region V
31. Cincinnati Water Works
Cincinnati, Ohio
Ohio River
32. Chicago Dept. of Water and Sewers
Chicago, Illinois
Lake Michigan
South District Water Filtration Plant
33, Clinton Public Water Supply
Clinton, Illinois
Ground
34. Indianapolis Water Company
Indianapolis, Indiana
White River and wells
White River Plant
35. Whiting Water Department
Whiting, Indiana
Lake Michigan
36. Detroit Metro Water Department
Detroit, Michigan
Detroit River Intake at head of
Belle Isle
Waterworks Park Plant
37a. Mt. Clemens Water Purification
Mt. Clemens, Michigan
Lake St. Clair
37b.* Mt. Clemens Water Purification
Mt. Clemens, Michigan
Lake St. Clair
38. St. Paul Water Department
St. Paul, Minnesota
Mississippi River
39. Cleveland Division of Water
Cleveland, Ohio
Lake Erie
Division Filtration Plant.
40. City of Columbus
Columbus, Ohio
Scioto River
Dublin Road Plant
-------
- 6 -
41. Dayton Water Works
Dayton, Ohio
Ground
Ottawa Plant
42. Indian Hill Water Supply
Cincinnati, Ohio
Ground
43. Piqua Water Supply
Piqua, Ohio
Swift Run Lake
44. Mahoning Valley Sanitary District
Youngstown, Ohio
Meander Creek Reservoir
45. Milwaukee Water Works
Milwaukee, Wisconsin
Lake Michigan
Howard Avenue Purification Plant
46. Oshkosh Water Utility
Oshkosh, Wisconsin
Lake Winnebago
Region VI
47. Terrebonne Parish Waterworks
District #1
Houma, Louisiana
Bayoulafourche
48. Camden Municipal Water Works
Camden, Arkansas
Ouachita River
49. Town of Logansport Water System
Logansport, Louisiana
Sabine River
50. City of Albuquerque
Albuquerque, New Mexico
Ground
51. Oklahoma City Water Dept.
Oklahoma City, Oklahoma
Lake Hefner
Hefner Plant
52. Brownsville Public Utility Board
Brownsville, Texas
Rio Grande River
Plant #2
*
Resampled
53. Dallas Water Utilities
Dallas, Texas
Elm Fork, Trinity River
Bachman Plant
54. San Antonio City Water Board
San Antonio, Texas
Ground
Region VII
55a. Ottumwa Water Works
Ottumwa, Iowa
Des Moines River
55b.* Ottumwa Water Works
Ottumwa, Iowa
Des Moines River
56. Clarinda Iowa Water Works
Clarinda, Iowa
Nodaway River
57. Davenport Water Company
Davenport, Iowa
Mississippi River
58. Topeka Public Water iupply
Topeka, Kansas
Kansas River
South Plant
59. Missouri Utility Company
Cape Girardeau, Missouri
Mississippi River
60. Kansas City Missouri Water Dept,
Kansas City, Missouri
Missouri River
61. St. Louis County Water Company
St. Louis, Missouri
Missouri River
Central Plant
62. Lincoln Municipal Water Supply
Lincoln, Nebraska
Ground
Region VIII
63. City Water Department
Grand Forks, North Dakota
Red Lake
-------
- 7 -
64. Denver Water Board
Denver, Colorado
Marston Lake
Marston Plant
65. Pueblo Board of Waterworks
Pueblo, Colorado
Arkansas River
Gardner Plant
66. Huron Water Department
Huron, South Dakota
James River
67. Salt Lake Water Department
Salt Lake, Utah
Mountain Dell Reservoir
Region IX
68. City of Tucson Water and Sewers
Dept.
Tucson, Arizona
Ground
Plant #1
69. City of Phoenix Water and Sewers
Department
Phoenix, Arizona
Salt and Verde Rivers
Verde Plant
70. Department of Supply and Purification
Coalinga, California
California Aqueduct
74. San Diego Water Utilities Dept.
San Diego, California
Colorado River Aqueduct
Miramar Plant
75. San Francisco Water Department
San Francisco, California
San Andreas Reservoir
San Andreas Treatment Plant
Region X
76. Seattle Water Department
Seattle, Washington
Cedar River Impoundment
Cedar River System
77. Douglas Water System
Douglas, Alaska
Douglas Reservoir
78. Idaho Falls Water Dept.
Idaho Falls, Idaho
Ground
79. City of Corvallis Utilities Div.
Corvallis, Oregon
Willamette River
Taylor Plant
80. Ilwaco Municipal Water Dept.
Ilwaco, Washington
Black Lake
71. Contra Costa County Water Department
Concord, California
Contra Costa Canal and San Joaquin River
Bollman Plant
72. City of Dos Palos Water Dept.
Dos Palos, California
Delta-Mendota Canal
73. Los Angeles Department of Water and
Power
Los Angeles, California
Van Norman Reservoir
-------
- 8 -
Procedure
Engineering evaluation of treatment facilities. At each of the 80 sites
chosen for study, engineers from the USEPA Regional Office visited the water
treatment plant and evaluated the facilities. They collected basic information
on the raw water source and treatment facilities. In addition to this
information, these engineers also determined the dosage of various water
treatment chemicals used and their points of application.
Sampling for trihalomethanes, carbon tetrachloride, 1,2-dichloroethane,
Because the six compounds chosen for study were known to be volatile, a sampling
procedure was chosen that would provide for minimum loss of the six compounds
from the water to the atmosphere while the sample was in shipment or awaiting
analysis by the purging technique (see below).
The containers chosen were glass 50-ml "Hypo-Vials"* sealed with Teflon-
faced "Tuf-Bond" discs, both available from Pierce Chemical Co., Rockford, 111.
Before use, the glass vials were capped with aluminum foil and muffled at
400ฐC for at least 1 hour to destroy or remove any organic matter interfering
with analysis. With aluminum foil still in place, the bottles were packed
(along with sufficient discs and aluminum seals to secure the discs in place,
labels, and re-usable ice packs) in an insulated container and shipped to the
appropriate regional office for sampling. Sufficient materials were provided
for taking three raw- and three finished- water samples.
In the field, the vials were filled, bubble-free, to overflowing so
that a convex meniscus formed at the top. The excess water was displaced
as the disc was carefully placed, Teflon side down, on the opening of the vial.
The aluminum seal was then placed over the disc and the neck of the vial
and crimped into place. A sample taken and sealed in this manner was
*Mention of commerical products does not constitute endorsement by USEPA.
-------
- 9 -
completlyheadspace-free at the time of sampling. Usually a small bubble would
form during shipping and storage, however.
The samples collected from the 80 locations from late January to end of
April 1975, werelabeled appropriately, repacked with the frozen ice packs
in the original insulated container, and returned via air mail to the EPA's
Water Supply Research Laboratory in Cincinnati. Upon receipt at the
laboratory, the samples were refrigerated until analyzed.
Analytic methods for chloroform, bromodichloromethane, dibromochloromethane,
bromoform, carbon tetrachloride, and 1,2-dichloroethane. The sample concentration
procedure chosen for the initial step of identification and measurement of the
4
six volatile halogenated organics was essentially that of Bellar and Lichtenberg .
In this procedure, the sample is purged with an inert gas that is passed, in
series, through an adsorbent material that traps and concentrates the organic
materials of interest. The organics are then desorbed from the trapping
material by heating under a gas flow and transferred, thusly, to the first
few millimeters of a cold gas chromatography (GC) column. Separation
(chromatography) is then carried out with temperature programming.
During this survey, only single column GC was routinely performed, mostly
because of the shortness of time for completion of the NORS. A high level of
confidence that proper identifications were made was attained by use of the
Hall Electrolytic Conductivity Detector operated in the specific halogen mode.
Further assurance of proper identifications was given by supplementary analysis
of nine raw and nine finished duplicate water pairs (from selected locations)
on a second column using a microcoulometric detector operated in the oxidative
halogen mode. Finally, gas chromatographic-mass spectrometric (GC/MS)
analysis confirmed analysis of 15 of the finished water samples.
-------
- 10 -
The glass purging device and stainless-steel traps used in the analyses
were fabricated exactly according to Bellar and Lichtenberg . The adsorbent
material used in the trap was Tenax-GC, 60/80 mesh (Applied Science, State
College, Pa. or Alltech Associates, Arlington Heights, 111.).
The chromatograph used for analysis was a Varian Model 2100 with one inlet
modified to the general configuration of Bellar and Lichtenberg's desorber
Number 1. The column used for separating the six compounds was 12-ft x 2-mm
I.D. glass, paJcked with Tenax GC, 60/80 mesh. The column effluent was connected
via a stainless-steel transfer line to a Tracer Model 310 Hall Electrolytic
Conductivity Detector (Tracer, Inc., Austin, Texas)to detect and measure the
compounds. This detector was chosen as the most suitable for the immediate
needs of the survey.
Blank water and water used for dilution of standards was
prepared by purging distilled water with helium until no interfering peaks
could be detected by use of the complete analytical procedure. Stock
standards were prepared, with dilutions, in 95% ethanol of the test compounds.
The appropriate final aqueous dilution was made by injecting 1 to 10 yl
of an appropriate stock standard directly through the valve on the 5-ml sampling
syringe (see below) into a blank water sample contained therein.
The sealed water sample, as received from the field, was heated to 25 C
in a water bath. Just before the actual analysis, the entire disc-seal
combination cap was removed with a "Dekapitator" (Pierce Chemical Co.).
Duplicate aliquots from the sample were taken as follows: A glass 5-ml Luer-Lok
syringe (plunger removed) was fitted at the tip with a closed Luer-Lok one-way
brass stopcock. The water sample was poured into the back of the barrel
-------
- 11 -
of the syringe until the barrel was completely full. The plunger was then
quickly inserted into the barrel in such a way as to eliminate air bubbles.
The valve was opened momentarily. The plunger was depressed to the 5-ml mark to
expel excess sample, whereupon the valve was again closed. Only one of these
aliquots was routinely analyzed; the duplicate was simply stored in this
configuration until the success of the first analysis was assured.
The syringe assembly containing the aliquot to be analyzed was connected
to the Luer-Lok needle that was inserted into the sample inlet of the purging
device (the needle was never withdrawn from the septum). At ther. time of
analysis, the valve was opened and the sample was expelled from the syringe
b> depressing the plunger. After this, the valve was closed until purging
was complete. After purging, the water (to be discarded) was removed by
reversing the above procedure.
The technique of purging the sample and desorbing the trap contents onto
4
the GC column were carried out exactly as described by Bellar and Lichtenberg.
Purging was for 11 minutes with a helium gas flow of 20 ml per minute.
Desorption was for 3 minutes at 180ฐC with a flow of helium through the trap
onto the GC column of 20 ml per minute (in addition to the carrier gas flow).
At this time, the GC column was at room temperature.
To separate the compounds, the column was first quickly heated to 95ฐC,
followed by a 15-minute hold, and then programmed at 2ฐC per minute to a final
temperature of 180 C with a helium carrier flow of 20 ml per minute. Conditions
for operation of the detector were those recommended by the manufacturer for
optimum performance in the halogen mode.
Compounds were identified according to retention time (measured from
beginning of the hold at 95ฐC) and quantified by comparing peak heights with those
of standards prepared at similar concentrations.
-------
- 12 -
Retention data and the range of minimum quantifiable concentrations (MQC)
encountered for the six compounds during the survey are summarized in Table 2.
TABLE 2
Chromatographic Retention and Sensitivity Data
Minimum Quantifiable
Typical Concentration (MQO)*, yg/1
Compound Retention Time (min.) Range Observed During Survey
Chloroform
1 , 2-Dichloroethane
Carbon Tetra chloride
Bromodi ch 1 oromethane
Dibromochloromethane
Bromoform
20.3
25.8
27. 7+
31.8
41.2
49.7
0.1
0.2
1.
0.2
0.4
1.
- 0.2
- 0.4
- 2.
- 0.8
- 2.
- 4.
*2% scale deflection.
+ Broad peak not completely resolved from 1,2-dichloroethane.
These retention times were typical; they varied slightly with aging of the
columns and significantly with installation of a replacement column. The
MQC was not constant throughout the study because of various changes in normal
operating parameters. No attempt was made to standardize the MQC; operational
parameters were simply adjusted to the optimum for any given day.
Confirmation analysis. As noted above, to add confidence to the routine
analysis for the six chosen volatile halogen containing organics, replicate
samples from selected locations were subjected to reanalysis by the Methods
Development and Quality Assurance Research Laboratory for quantitation on a
second GC-Detector system and for qualitative analysis with a GC/MS
system. Table 3 shows the sampling cities for these confirmation samples.
-------
- 13 -
Table 3
Cities Whose Samples Received Quantitative and Qualitative Confirmation
Quantitative Confirmation
2. Waterbury, Connecticut
7. Little Falls, New Jersey
16. South Pittsburgh, Pennsylvania
30. Charleston, South Carolina
51. Oklahoma City, Oklahoma
60. Kansas City, Missouri
65. Pueblo, Colorado
71. Concord, California
79. Corvallis, Oregon
Qualitative Confirmation
11. Philadelphia, Pennsylvania
22. Miami, Florida
30. Charleston, South Carolina
31. Cincinnati, Ohio
41, Dayton, Ohio
51. Oklahoma City, Oklahoma
55. Ottumwa, Iowa
58. Topeka, Kansas
60. Kansas City, Missouri
71. Concord, California
72. Dos Palos, California
76, Seattle, Washington
79. Corvallis, Oregon
The quantitative analysis was similar to that described by Bellar and
Lichtenberg. The following details describe the specific procedure. All samples
were stored at 4ฐC until just before analysis. Five ml of each sample was
purged for 11 minutes with nitrogen flowing at 20 ml/minute. The purging device
was maintained at 19 C. The sample was introduced into the purging device at
4 C. Therefore, as the sample was purged it warmed up to 19ฐ C at an unknown
rate. The sample was concentrated using a trap packed with 18 cm of Davison
silica gel, grade 15, 35-60 mesh. Desorption took place for 4.0 minutes at
200ฐC.
-------
- 14 -
An Infotronics Model 2400 gas chromatograph equipped with a Dohrmann
microcoulometric detector (halide specific mode, oxidative) was used to
perform the analyses. A stainless steel column packed with Porasil-C coated
with Carbowax-400, 100/120 mesh, 6-feet long, 0.1-inch I.D. was used to
perform the separations. Nitrogen flowing at 50 ml/minute was employed as the
carrier gas. The column was programmed over the following conditions: Desorb
into column for 4 minutes at <30ฐC; heat column to 50ฐC and hold 1 minute;
and program column to 175 C at 8 /minute.
Under these conditions the limit of detection for the materials of
interest were: chloroform, 0.05 yg/1; bromodichloromethane, 0.1 yg/1;
dibromochloromethane, 0.1 yg/1; bromoform, ^5 yg/1; 1,2-dichloroethane, 0.1 yg/1;
and carbon tetrachloride, 0.05 yg/1. Methylene chloride was routinely detected;
the limit of detection was 0.05 yg/1. Although other unknown halogenated organics
were detected, their concentrations were below the limit of detection for GC/MS
identification. By calculating relative retention times, the same unknown
organohalides were found to be present in many of the water supplies tested.
The qualitative analyses were performed on samples treated as described above
for quantitative confirmation with a Varian aerograph 1400 gas chromatograph
interfaced with a Finnigan 1015C quadrupole mass spectrometer controlled by
a System Industries 150 data acquisition system. A glass column packed with
Porasil-C coated with Carbowax-400, 100/120 mesh, 6 feet long x 2 mm I.D.
was used to perform the separations. Helium at 30 ml/minute was used as the
carrier gas. The column was programmed under the following conditions;
Desorb into the column for 4 minutes at <30ฐC; hold at <30 C for 1 minute;
heat column to 100ฐC and hold for 3 minutes; and program to 200ฐC at 8 /minute.
The mass ispectrometer was operated in the following mode:
-------
- 15 -
Mass range scan 20-350
Integration time 12
Samples/AMU 1
Total run 30 minutes
Some of the qualitative analyses were made on a different spectrometer
operated under slightly different conditions , Because a larger sample
volume was purged in the latter case, these analyses had better sensitivity.
Analytical methods for general organic parameters. Nonvolatile total
organic carbon (NVTOC) vas determined on an instrument made by Phase
Separations Ltd., United Kingdom. Samples are acidified with nitric acid,
purged with nitrogen gas for about 10 minutes to remove carbon dioxide, then
pumped into the instrument at a constant rate of 0.6 ml/minute for about 10
minutes. During this time the nonvolatile oranic carbon is thermally oxidized
to carbon carioxide (CCL) at 920ฐC on copper oxide, then reduced to methane (CH.)
at 450ฐC on nickle in a hydrogen atmosphere. The methane produced is measured
continuously with a flame ionization detector.
6-'
Ultraviolet absorption. See the method of Dobbs, Wise and Dean '.
Fluorescence. The rapid fluoremetric method (RFM), as described by
Sylvia , and a fluorescence emission scan (EmFS) were performed. In this
latter determination, the excitation and emission slit widths are 12 nm and
16 nm, respectively, and the aqueous sample is excited at 310 nm, with the
fluorescence emission recorded between 370 nm and 580 nm.
Quality control. Accuracy. To test the accuracy of the method used by
the Water Supply Research Laboratory during the survey, the Methods Development
and Quality Assurance Research Laboratory prepared a pair of "unknown" standard
mixtures in the following manner: Two different stock solutions, each
containing all of the compounds of interest, were prepared by injecting a
known volume of each material into a volumetric flask containing 90 ml of
methyl alcohol. After all of the compounds were injected into the flask,
-------
- 16 -
the mixture was diluted to volume (100.0 ml) and mixed by inverting. Two-
hundred microliters of the stock solution was then dosed into 1.0 liter
of super-Q water (Millipore Filter Company) and mixed by inverting two
times. One-half of the dosed water was then transferred into a 500-ml
separatory funnel. Several, 60-ml vials were then filled with the mixture
and promptly sealed with Teflon septate. The samples were stored at 4 C
until delivery to Water Supply Research Laboratory. The blank, Sample D-4,
contained only super-Q water. The calculated concentrations of the dosed
mixtures, D-2 and D-3, are listed with the analytical results in Table 4a.
Precision. To test variability of results during a typical day of analysis,
two series of 5 replicate samples were preapred as 10 discrete samples in the
same manner as standards were prepared throughout the survey. One series
was at low concentrations, the other at high. All of the samples were
analyzed exactly as described above to determine the concentration of the
six halogenated organic compounds. Spiked concentrations and relative standard
deviations (o/X ) are listed in Table 4b.
avg.
-------
- 17 -
Table 4 a
Determination of Accuracy
Concentration (yg/1)
1,2- Carbon Bromo- Dibrorao-
Sample
D-2 (True value)
D-2a (WSRL lab)
D-2b (WSRL lab)
D-2a (MDQARL lab)
D-2b (MDQARL lab)
D-3 (True value)
D-3a (WSRL lab)
D-3b (WSRL lab)
D-3a (MDQARL lab)
D-3b (MDQAHL lab)
D-4 (Blank - WSRL
D-4 (Blank - MDQA
Chloroform
74.6
63
65
* 60.8
76.3
59.6
46
46
* 53.6
58.6
lab) 0.2
RL lab) 0.1
Dichloro
ethane
10.1
9.
10
9.5
9.8
5.0
6.
5
4.8
3.6
1.
0.2
Tetra-
chloride
9.5
9.
8
7.9
5.9
6.4
5.
6
6.2
3.6
NF+
NF
dichloro-
methane
39.6
39
40
35.3
37.6
23.8
22
23
21.1
19.1
NF
NF
chloro-
methane
23.8
23
23
17
15.2
19.0
14
18
13.3
11.5
NF
NF
Bromo-
form
40.4
40
38
48.5
44.9
23.2
18
24
24.1
29.4
NF
NF
*Two weeks elapsed between the duplicate analyses at the Methods Development
and Quality Assurance Research Laboratory.
+ - None found.
-------
- 18 -
Table 4b
Determination of Precision
Low Concentration Hi^h Concentration
Spiked
Cone.
Compound (pg/1)
Chloroform
1 , 2 - Di chl or oethane
Carbon tetrachloride
Bromodichloromethane
Dibromo chl oromethane
Bromoform
2
1
2
2
2
4
Rel.
a(%)
6
5
14
5
10
20
Spiked
Cone. Rel.
(ng/1) a(%)
18 7
* *
* *
20 7
30 13
30 12
*Not determined at high concentrations.
In summary, these precision and accuracy data indicate these analyses were
satisfactory.
-------
- 19 -
Results
Source and treatment information. The following percentages indicate
the different categories of sources studied in this investigation:
Source %_
Ground 20
Lake or reservoir 33
River 47
Mixed 0
The percentages for the various types of treatment practiced by the utilities
in the Survey were:
Treatment _%_
100
Disinfection
Chlorination
Ozonation
Raw water chlorination
Raw water ozonation
Polyelectrolyte used
Powdered activated carbon used
Granular activated carbon
Softening
Precipitative
Zeolite
Taste and odor control practiced
99 (At some place in the
treatment system)
1 (The only treatment
practiced)
75
1
22
25
10
25
22
3
38
-------
- 20 -
A study population of about 36 million was covered in the Survey. At 60
locations practicing raw water chlorination the following percentages of
systems employed the indicated dosages:
Dose, mg/1 %
0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
>10
Unknown
15
19
12
14
14
12
4
2
2
2
4
In 86% of these locations, the raw water chlorination dose was between 1
and 6 mg/1.
The percentages for chlorine residual (free, combined, and both) at all 80
locations were:
Combined residual, mg/1
0-0.4
0.4-0.8
0.8-1.2
1.2-1.6
1.6-2.0
2.0-2.4
2.4-2.8
Unknown
60
20
6
4
4
1
4
1
-------
- 21 -
Free Residual - mg/1 %
0-0.4 41
0.4-0.8 19
0.8-1.2 4
1.2-1.6 20
1.6-2.0 4
2.0-2.4 8
2.4-2.8 3
Unknown 1
Free and Unknown Combined
Residual
Total less than 0.8 mg/1 16
In general, rather low residuals were present in the finished waters studied,
and at 16% of the locations, less than 0.8 mg/1 of free, plus combined chlorine
residual was recorded.
Raw and finished water data. From the data summarized in Table 5, the six
selected compounds measured in the raw water at the 80 locations were seen to
be nondetected (30 locations) or present in very low concentrations. One
location. (#35) was receiving water pre-chlorinated by others and this water
did contain some chloroform, bromodichloromethane, and dibromochloromethane.
The water before chlorinationw as typical of other raw waters, however. The
nonvolatile total organic carbon determination was made on each sample and these
data are listed in Table 5. Some of these data may be in question because of
suspended solids rir.some of the raw water samples. The ultra-violet absorption
and fluorescence data were considered unreliable because of the presence of
suspended solids in the samples, and were not reported.
Table 5 also summarizes all of the data on finished water quality from
the 80 locations. The ultraviolet absorption and fluorescence data for finished
waters are not presented at this time, but will be discussed later in this
paper. The -..range of each measurement is also noted at the end of Table 5.
-------
- 22 -
a
P
0
X
in p
-H
O rH
rS 3
c3 cy
E-
p
O -d
-ป o
ซ J3
Cs - X! ,
..H P 00
rH Q CD 3.
r< 0 00
03 <4H 3.
O 1 i)
S 0 C
O h cd
!H O XI i-l
XI rH -P ^^
rt XI 0> 00
a o B 3.
I
O
I o C
O rH CS
1 fj X ^
(H -H
H C
4-* &
^ i i
C^ \O CNJ %y\ T~H O
CO --( CX] (M C\a CM
tu
Ctj 2; Cti CM Cti PH
S5 fe; v fe 2
CM
Ctj tU &J O fej H,
te; 2: t^ v s* ^
fez fe ^ feS
+ 0 ^0
fcj Ct, P<4 ft,
So ft o ft z
00
^ fcj O fej *
V Oi fe; I-H ft O
Ci rH li| tO Ct4 Tt
V CT1 ft CTV ft
c
0
H
Wl r-v
[/) C
H 5
g -H fi
?-i O c/3 f-l 0)
0 P g
S 3 P Q (1)
S: o
rH
/ ^
r 1
M
S
rH
OH
' 4-*
O
S'
' O
f-f
^Qf-^ate;
servo ir '
f JJt'cd '
1 ^Jt^Q '
Q' ' C '
4^>CL
' ll
' 21 ' U5 '
1 v_^'
,
^-
d in
N5 CM
fcj tlซ
ft z
fe S5
feg
0>
fed
**
fc, CM
ft CM
tn
(U
o
0
to
05
V
1
o
p
c
s
tt
CO
PH
Q
.
trt
d O
CM CM
fej tL>
ft Z
fe z
fej CM
ft
fe s
Ptj C7)
ft CM
Ci I?
V
p
c
xj a
ง ^
'j.
fH ^
S g
W c3
*O' e.
fl c_3
queduct
s Water
ฃป
o H
p 00
r< r<
&4 '"'
.
^D
^0 OI
t*O r-(
C^l CM
V V
00 CM
C5 O
V V
fez
fcj CM
E5
fei \O
fe; rH
C) LD
c
0
H
E/)
t/1
S
o
U
h
0
H
CO
t/)
o
IVJ rH
t, M.
ซ ^
fez
fe iH
ft l-H
o>
^
i
ti
cd
o.
0)
Q
0)
p
cd
inebeck
S
m
o
Oj
rH
rH
H
^
.
O
-------
- 23 -
a
ctf
Q
rH
0)
4->
3
O 13
4i O
ซ 43
Q to
< > rt *i> *H
13 cij CJ
P Q a ซ
G -H
o x a
- ' -H 0? PL,
in co to H
3 ซป
<1> Cy rH
rH tO Hi
43 rH O rH
0)
(3
li
(3 I
o a
a O/
E 3.
E
o
o
rH
O I-H
I-H ^*~
43 0
'"" *
cซ
S
1 '
43 P
3 C
CO n)
oo O
C -P
rH 3
S '-'
CN
rH
O CN 00 CM
O O ^ l-H
ct, tu (t, u<
ft Z ft Z
CN tO
It, O fe, O
ft V ft V
fJH rH fe, r^
ft v fe: Z
fe. fc.
ft rH ft CM
fe, fe,
ft 0 ft 00
eg
" "I U
3- rH
O CS
Cd rH Cd
CD *t -H
rH 13
3 p rt ^^
CO C > 4->
CD I-H C
XI-H X ct)
p ex. tn I-H
H CO,
U 0 G
rH
rH rH
>
eg O
Ci O
fe z
fe, 0,
ft Z
fe, B,
ft Z
fez
PCJ tU
-5; 2
i-H
PJ^I o
85 v
g
tn
X
ฃ_,
4-)
n)
43
OO
I
S
OO
43
E
CO
i-H
CV
SH
fe
By
fe*
fe
fe
ci
V
e\]
^
V
X
4->
H
O
43
3
!H
n)
X
p
0
rt
'rt
n,
00
l-H
r-~
CM
tu
z
B,
z
ง
VD
O
V
vO
^
<0
( ^
rH
O-
orton
j
1
CXl CN
\H O
k~, tu
ft: 2
CM b
ft: ^
fe^
00
fed
fern
ฐฐ.
0 0
X
G
n!
&
O
u
rH
4->
CD
4J
ง0
H
U rH
H P
rH U)
IO .H
1ฐ
1 r-H
Cti rH
H 0>
C 2
H 0>
bO g,
fn O
H X
>
rH
c^l o
C\0 rH
^Ji tO
t^Q ^~
Qj O
V V
fez-
fet.
fcป \o
jS ^
ro
*~i CN
B1
o
u
rH
IU
g
P
00
K
H
P
X
O
CN
-------
- 24 -
CD
CD
a
!H
CD
P
cD
S
(3 'O
H^ CD
!i C
f-^ CD CB -rt
V CD ,3
p Q ^ to
ง X 3"H
U 4-> Q O
in o] to H
O CX JH
rH (0 OJ
rCl .M S) 'H
o3 1) *ti P
ฃ-, 4-ป *--l DO
rt t3 CD
CO G
a: C -H
o *ป
Z CD
s.s
JS
TJ
4i C
CO O
A o
TJ CD
&4 CO
*
CD
O G
H 0
CD rl
rH CD
H U
CD O "^
rH ซH CM
o c e
> cD
G 00
O rl
Z O
CD
o
C 1 -H
O CD rl
43 rl 0 rH
rl P rH ^
CD CD X M
U H 0 3.
t
g
O CD
1 rH G
X CD rH
CN O ,C -ซ.
-.H P OO
rH Q CD 3.
1
0
S 6 rH
rl O Ot
ca ID d,
CD rt C ฐ ง
4J ^ cd CD CX
CD rH rH
^ 0 CX i-H tO
-n .H TI
oo CD G > G
C Q 0 G CD
.H IP O rH
"^ S 0 ^ 00
(D CD rl O -rl
| g &, r? ฃ
^ ,
rH CN tO
CN ซCN CN
CVJ 01
1-1 O
1*4 B,
fe: Z
PO
d p-1
v 2:
^ Z
fe: CN
feS
CXI
v to
c
cD
t/) i-H
JH
O 0
rl .*~!
0 O
P O
CD O
S 43
U
CD cD
P P
C P
^ ^
zf
CN
CV O
T-1 CN
1*4 CU
E5 Z
1*4 B.
fe; Z
fe; to
fe! rH
I*, 0
fe! CN
f*. tO
S: rH
to
0
H
P
H
rH
H
3
1
CD
H
CJ
H
C
3
S
O
rl
O
to
c
CD
a
.
LO
CN
^0 T-) vJD
CO rt r-H O
Pt( [it 64 [i,
^ rS ^ 2
CN -^t
fJL| O IJL| O
fe; V r^; V
C^ rH fcj [i,
S5 V ^ 2
r-
rXj Ct( *
J3 to S; o
S; vO fe: Oi
to 05
. I-~ .0
O rH Q) tO
X
CD
g
P O
G U
rl
rl P
0
ง
M O
H (/)
> t/1
C OJ
OJ C
0) C
rH 0>
U E-
*
vO t^*-
CN CN
C^l M
Q) O
tt4 U~
fe: Z
fez"
feg
fen
1*4
fe! CN
CXI
V O
ฃ>
-H
Q
r4
0
P
CD
S
T3
g
to
CD
P P
00 cD
H rH
iJ CX
-------
- 25 -
nt
4->
cd
Q
(U
4-"
*
ซ "0
4i ID
/-^ nt 01 ซH
4-> -W B,
t3 cd O
4-> O S t/>
C -H
o x a
CJ 4-> Cj ID
in cd to H
3 '^
ID O1 rl
I-H to nt
JO M O rH
cd o '^3
H 4-> r~i bO
cd C3 fl>
eg c
Di K .H
0 -fi
Z fl>
i j
kq
ซ
+ป e
to O
P. 0
"pi Q)
rH
cd
O C
t~ 0
rH Cd
H U
4J i-H
cd o -^
0 e E
> cd
R bo
O to
Z 0
U
C 1 -H
O Cd rl
& to O I-H
to 4-1 rH \
ni (U J3 M
U H 0 3
l
O
rl
rH C
' .C at I-H
CM U ft --.
-> .H 4-ป bO
rH Q <0 3.
O
E E rH
rl O b
1
E O C4
grl Cd i-H
0 rC -^
H f. CD 3
Q 0 E
1
go
1 0 C
O rH (3
0 0 4-> ~-
to -H ID M
oa *o 6 3.
B
0
S
0 rH
l-H \
CJ 3.
'v
T
0
c
'c
p.
a
^
e CD
Z B
X
4-> 4->
-H C
r-H C
H rH
4-> (X
O * '
^^
eo
ซ
CO
o
fi
^
r-H
t>O rH O> LO IN. t^
C<] r- 1 T-H iH Cs. \Q
fcj
fe
fc^ (X PC, Ei,
oo CM fe; SE fe: 2:
V V
^ r^ ^"
Pt| O ft| O Cti U,
^: v fe v fe: 2:
fcj
fe
Cti Pu fej [iซ fei Pu
S Z S3 Z S3 Z
R^
fej Cij fej B,
E; rt Es t S3 Z
10
S3 rH Ej rH ES 0
Qj
C^" c\3
UQ ป .
Q> ** V rH V "*
t/1 fj
*-ป c
g), rt
> rH
$ ฐ"'
T3' * ^
งs 1
"I U to W
ฐ Ji .B
S 'H
T) i
c S
^f
to
-^ CM
0ปJ
t-J Co
Co
r?g
5 o
4i
^c?(ง
Q 0) i/> to CM
CM l-H rH CM rH
fel ttj B, Pt, B,
fe: S3 Z S; Z
E; E; Z ki o
fefefe feง
fiq UH fej
co 65 2 S: tO
CO
. to
i-t V O ^ Oป
CQ
T-H Lfl
to . d O V rH
jj
C
E
P
fH
+J cd
4-J fH HO
H 4J Cd
ฃ <" *
LO VO
to to
Q> CO
1
J *
itj
Q) S
e, ?
2 ^
0
Q> ^ EN, tS" O^ ^" C\l OO
CX]rH toปH IN.1^ C<]rH
fcj tl, fci tt, C^PU 64 PU
ง2: fe:^: fejrZ fe;Z
Tf ^3
ti^O Q>IX P^Un C^tLi
^v v.2 ^2 te: s?:
fe(tU fejtL, fej fe f*4.X,
S S ^ SI งrZ rS2
fej Et| Ctj Pt|
I^fN fe:(N r^CN r^^J"
V V
PCJ fej fej Cti
^;yD fe;tO E^r- fe;Cn
CXI
O^ r>
C^ rH O ^t Ctj OO
^rH d^O Vrf fe; i I
C C
0 0
H *H
4J 4J r4 ' *
cd Cd ID 4-1
00 4J C
H -H X Cd rf
>H 'H >H CU
S 3 0 0 C
a, a, x o
4-> C -H
rH rH 3 O 4-1
cu n> rH
4-1 -H -H
I/I VI Cd Q BH
c c *
d)O TJ C
E E I-H CO
m 4-1 4-1 rH Q
S S CO CJ ^
. .
cd J3 >
r-- f-- oo 01
to to to to
Co tO
to CM
CtB,
feS
feg
feo'
V
feoo
r-H^)-
.to
OrH
, \
c
1/1 cd
3 rH
3 -a
l-H Cd
0 0
cj oi
O -H
i-H
X r^
4-> 3
H O
CJ '-'
O
-------
- 26 -
CD
4-"
m
ง O1 IH
rH CO CD
,Q r< ^ rH
CD CD -t* 3
E-. p ซS M
CO ซ CD
S H-i f!
to d
PCI S *H
0 -S
z CD
C 60
O IH
Z 0
C 1 -rt
0 CD rl
,0 f-l O rH
IH P rH -V,
cD fl> ,C W
CJ E-i 0 3.
g
O O
' SSrH
CN 0 J3 -~-
ซ.H (-> M
H CI *^
0
O
O rH
rH "^
U =
l^t
1
r"
Ct
r
|
ฃ
i]
x
M 4-*
H C
rH 0
H rH
4-> O.
.D ^~-
Oif^- COO) Q)C-J C\,rH ^[^
^ O Q O CO ^" ^J* *O C\J i-H
f-C^ Ol Pt^tU Ptj PH rJ^U-. Ctj U-
(N tN
E^O I^Z fe; V ^rZ fe;V
V
^^t fefe S^Z lงง Sง fe
^^ ^^ ^.0 ^ ^ro
V
l^oo t^r^ fejTH ,^LO i^r--
Csa CX]
fcj Ci Pt, to fcj o d
t^OO VIO f^-H fejOO VOl
M ' *
O 4->
*ซH S
f-l TO
4-> rH
VI (X
'H
Q C
- x .2
& ctj cD
& 4-> en O
3 ^ -g -gifl
tf> X CD O -H
"r( S "p. W ^
o ซ-> p. x mo.
S: ^ % cti 3 Q> CO
p ^ C/l rH 4J 0)
IH (3 --1 cD 3
ci) CD I-H h CD S C
p rH rH Cy ^> CD
CD (X .H 4J O >
^ X cc) bo C)<
CD S C X
gs in -H 3-p
to 03 Co C Oj M
P -M -H 13 O S CtJ
X -P TJ CT rC ซH S
nj O ฃ !-< co -HO
Q N / hH a- s SK
ป
rH (N tO ^ Lf)
tj-j ฃfj
Qj O
*O tO ^J1 fNJ t^ I/) Cvi LO
'yitO LQtO fOiH ^tO ^^
V V
r*X|ฃ^ ^ fe ^ P- fej tL, fJH [I-.
W t/1
r^ rX X
o o
S ฃ rH
rH
-------
- 27 -
ca
cd
a
fH
CD
ca
3=
O T)
+i CD
H
/ ^ ca cjj ซH
t 4J -4i [L,
a ca o
4-ป Q S yi
C -H
o x a
u *j 5 CD
'~' rH * X*
i/j ca eo H
3 -ti
00- fn
rH CO ca
J3 f-l O rH
ca o M 3
H 4-> r-i bO
ca Q QJ
c/) C
Cฃ C -H
O -ri
Z CD
rl ij
^
CO O
_f, O
&, I/)
*
"ca
^j
0
t-
CD
H
4->
ca
o
c
o
C3 i
o ca
8 Z
a CD
u H
i
o
o
CN O
ซ-H
*- Q
1
i
S
fH
CO
O 1
& o
0 fH
fH O
rQ H
s-s
1
o
fH
' O
O rH
O CJ
fH -H
ca -o
e
0
fH
U
0
'c
bo
f-l
0
CD
H
O
rH
0
Q>
ca
rC
4J
CD
g
O
S
ca
.ฃ
4-1
i
i
(-<
CD
e
g
o
->
IH -M
rt ' -H
(X rH
0 -H O ' -H
4-> -H --H
ClJ rH -H
x -M a, c
4-> C H cti
H ni <i
U rH rH /-% 4^ CL,
D- rH CN rt
rt -H =tt 3: C
S r-* > OJ
O0 ui -P t/i S
^J C , c e nj rC
" * 1> O < ' rH Oj
XX rH DH flJ CQ
O ^ OQ ^^ Q ^
i t CN tO
LO 4/> LO
, \
OJ /-N
rH D
t/j rt
tn
LO
t--- LO
t-- \
f-n r-^
fN ^f
'**"' ^~^
LQ LO T-~1 fO Oj "^f" t-O O ^3 *^" Mt CN
C^ O ^Ji CN MI CN fO fO Co ^" ^Q CN
&4 tL, ft)U- fejtii fcjLL, Ct, tU fej
Ct) [In fej PH i<4 Eli fe| tU tt-( O fe( fc
fei t^tLi fJ^Cii Ct,eu t^tu 64
^0
Ctj ^fe Ptj PU fcj &jO fe|d
ฐ0
CTi
fcj Ptj PH Ptj tU Ccj O,t Etj Ol CO
,5:0 ^ 5c 5:2 fe^rH ^oo vto
Cx] CO
CN . OO ** \H
Ctj ป O Ctj Qi QO 00 * OO
!^;O VO l^rH V^t CiOO CbOO
T3
cti X
O tn rH
CQ X OH
rH X QJ
H O p .3
a) 3: rt co
+* D-i
CS rH e ^
3: v, (n CD O I)
X f-i rH rt rt
4-ป O O S J-< S
H S 3
0 4-> P 0 3= -H C
H aj ni rH ^ rt
C S 3: -P 3 *-i
O cti M CX D-
4-* oJ flj T3 O
C S 3 C p^ rt X
< ง g -H C ^ P
3 3 rH -M Oj > P, O
+-ป 4-> i-H OJ O W
CO O O U C3 H " '
B 4
cd ,0
^j- to Ln \o r- oo '
l/J LO LO LO t/) LO
LO \ฃJ Mi 01
Ml tO fO rH
Piu tij ULi
^ CN fe; 2
So' !ฃ
feง feg
fecN fe X
H3 4-J
-H
H U
3
o ca
Ul t/1
I/I (5
H ca
S X
Cl O
1/3 \O
-------
- 28 -
*"^* rt
4-1
T) rt
4-> Q
c
0 X
CJ p
rH
LO rt
3
CD O"
rH
42 fH
rt o
H P
CO
CO
01
o
z
rt
p
Q
^
CD
rt
S
O -O
13 43
H
S, 13
Q) -H
fi FTI
Q
^ in
.H
y CD
Qq a,
.cot?
(^
CO Ctf
?) T f
ซ* ?
t-J oO
ti <1>
G
o
Q) p;
.C 'H
ij
13
4i ฃ
CO O
.5 %
fi H)
BH CO
*
0)
p
0 G
E- 0
rt
Pi bo
O rH
Z O
T)
C 1 -H
O rt rH
42 rH O rH
rH 4-> rH -^.
rt -^
fH -H CD b<
03 T) E 3.
B
<2
S
O rH
rH \
45 M
U 3
S
0
<
!
t
Q)
2
X
p P
H
rH
H rH
P Cu
3 ^_
Ml O Mi rt ซJ CM
(0 CM r-f rH Oj LO
fei pLi &j tU p,.. [T
fe r2 fe ป2 fe *Z
to rt
ซ [j^ B. [i^ B*
O O ES Z 85 Z
S3 "v E; *v S- ง
^ to B; rt fe fe
|2 fe^ ^rH
l-tj LO Ctj Ctj
X X
SrH
a
& ง
o co
u h
rH IU
0 P P
4J rt C
rt ^ a>
s e
rH 4J
X rt rH
P /-v & rt
C P -H P,
3 C O m
o rt -H Q
(X 3 rH
H rH P
3 rt B rt
O rH rH ^
J 4J O
13 0 X
S F3 p
p CJ -H -H
03 v-^ J CJ
rH CM tO
vฃ) ^ VO
LO LTJ
C\3 (Nl Ci O
OI-- co^o .. . OO
ซlrH ^rH v-|rH V~(O yy I-H iH
Ct. PL, Ctj PL, fcj tU Cti Lu fej PL, fcj [I*
r^ r^ fe -2 fe S S IZ fe ^Z fe ^
^1
fej FTi CJU| PM Pt| tU Q pi, Ct| UL| fcj PL|
r^ 2 ^ 2 "^ <2 V S "^ t*Z *^ I"S
fejtL, r-XlUl fc, CtjtU r-t^tO fej^
^2 -5:2 fe;00 1^ Z ^I-H ^v
ftj Ct( (N Etj CTi Ctj fe) Ctj t^"-
^tO !^v -~5^f l^CO ^C^J t^i-H
CO
fi^ O [<| rij I-H titj ^J" fej O Ct| LO
fe;rH fe;(N l5;i-H S;nH fe;v .5:^H
OJ CS] T~S CM C\3
Oi -< VCM fe;to Cioi vv vc.n
**
I ฐ
rH "
* S
m W
CO P ryi
S 1 -3 ง
s s t: ง
CU Ci) rt t, CD
o S B & B S
rH a rH 0 tj *
a a S
O / -
CDrHfnrHfH S n OG
pa. rt a. iu H^ 43 rt
rt O -P CD r^rHO-rH
3e c ca n rt x ^* a.
o (US rt tij SH .
fH4JOC ^J a-POCD
CDCOrHTJG C T)
>fH43fH O P K!rt XfH
G icl 0) rt fH rH H-l-H p CD
OS 3CJ 3 rt .*Jo..H>
Q ^-- CL. -s X CO Q*-^ CJ -
rt' w VO 1--' 00 0>'
vO \O vD \O vO \O
IN. rt
^3 CM
fej Bi
^ Z
fcj Bj
^ Z
fe; CM
Ctj L/l
S3 rH
!g r^
CM
V rH
C
o
H
p
rt
o
-H
H
d,
T!
ง
X
rH
OH
CO
MH
O
p
-------
- 29 -
/*ป rt
4->
"O 03
4J 0
ง X
U P
- ' -H
rH
in cd
3
0 cx
rH
.0 rl
cd 0
H P
cd
in
OS
O
Z
cd
p
a
rl
0
p
cd
s
ซ -o
4-i 0
a x;
Ci in
H
rl C
Q) -H
-ri B.
^ CO
H
ซ 0
ft; p.
X
_co H
"*1 rl
co cd
O rH
s a
ซ 0
ซ-
C -H
"ti
CD
CO E
C -rl
ti ปJ
O
-a
w B
to O
r, 0
ti 0
CM CO
*
rH
cd
) i
0 B
H 0
X>
0 rl
rH Cd
H CJ
P rH
cd O -~.
rH *H W
ง ง B
B CO
0 rl
Z 0
0
a
B 1 -H
0 Cd rl
ฃ1 rl O rH
Tl P rH \
cd 0 X bo
U H U 3
I
o
rl
O 0
1 3 Cd rH
CM O pE ^^
--H p bo
rH Q O 3.
1
o
B B rH
rl O M
CQ <4H 3
1
O 1 CD
6 o c
O rl Cd rH
ri 0 ,C -^
,h rH 4J bO
H .C 0 3
o u B
1
o
, rl 0
1 o 3
o rH a
e X X rH
O O 4-> ~~.
rl 'H 0 (X
CO 13 E 3
B
0
cfl
S
O rH
rH \
X -
Ml O>
to rH
fel B,
ft Z
ft Z
fe -1
ft V
ttj
ง vO
fo
Ci 00
V r~*
OQ
. rH
o to
p
1
p
rl
cd
OH .
0
Q
rl
0
P
cd
S
X
p
s
3
O '*
c_> P
c
cd cd
P rH
in cx
o
CJ E
cd
cd B
rl rH
P rH
E O
O CQ
tj ^
rH
t-^
MiCTl cxito O^OO tO^O
MirM T-HrH C\1CN1 r-)rH
ttjU. fejtl- CtjB. Pi,B,
ftZ ftZ ftZ ftZ
fez fez fez fez
00
PL fcg BH fcj PM O
งt^ Sz gto Sv
Ctj ^ Ptj Cti <7l fej
ftto ftto fern S;^-
fcj to Uj fej 0 fe, tn
ftLO ft^O fttO ftrH
1^-1
CtjrH Cif\J fj^CM fcjt-H
ft\o vto ftin ft-*
rl
0
P 3
E ฐ
0 (X
e
P t3 p p
rl B OH E ^
cd cd 0 0 p
PH . Q BE
0 rl p cd
Q 0 Irt rl rH
P :b p cd
OP cd 0
rH (1 rl ^^ 3 rl
cd. 4> 0 P H
(X Q P E O
cd cd o to
Ul CO S --H IO Cd
O 0 (X -H 0
Q *-H Q O rl
0 00 rl E T3
<4H Of 0 Cd Cd B
O fi 'H E rl <
< Q cd B.
X rl E
P <5 fi -H E Cd
H a cd s cd to
CJ i-J CO ^-^ CO ^-^
CM to Tt m
it- r^ r-^ r-
B
0
p
t/l
X
co
p
-H
Q
<4H
0
TJ
a
PJ
o^ 01 MI oo
C> O to CNl
Ct, B, Ct. B.
ft Z ft Z
fsx fe z
fc, B, fc, B.
ft Z ft Z
^f
fc, B, tt, 0
ft Z 6> V
a> oo
Cu fe, .
ft o go
ซ
d in fcj o
V rH ft *
4J
B
0
B
p e
rl 0
cd p
ฃX P
cd cd
s s
0 CO
rH Cd
P rH
P bo
cd 3
0 O
CO Q
^o r--
i-^ t^
w
Oi
t-H
LO tO C^ ^f LO rH '
Qi O T-H O CV tO ^
Q!
V
.
M
Pi, B. Pt, BH PL B,
ง Z ft Z S Z '
<^
V
to
,
fe z fe z fe z w
^-
fe, B. PL B, fc, B, fe
ft Z ft Z ง Z *=
fe, fe, B. fe, fe
ft to ft z ft in ft
CQ
Q
fe, fc, fe, m '
ft to ft to ft to _
d
V
Oi
CXI O
t-H r-^
Q fe, <ฃ> . *O '
V CM ft CM Cl rH
r
ci
V
4->
. B
> 0
H B
Q P
P rl
B "> cd
0 0 CX
B *H 0
P -P Q
rl -H .,,
Cd '-' rl ":
0. -H 0 S
0 P P g
0 = g a
rl CO
0 -rl rH
P rH cd
Cd rH ^ CX
S Cd P TH
> B CJ
tO rl Cd -H
rH O rH B
rH CJ CX 3
cd S
B, HH rl
O O O
O rH O
X X X cd
cd 4J cd 2
TJ -H H rH
rH CJ ^ ' rH
00 Cft O
t-~ r- oo
CM
CM
rH
1
in
o
o
V
.
& to
Z rl
4->
cd
3
^0 |
^
B. e
Z ^
^ O
10
r.1 C
S -H
i s
z ii
ts
ง ^
s .a
; S
^4-1
O
g
O
s^'S
^j in to
rt E 3
, ฐ -ฐ
1 -H B
B, y R
K cd o
^- rl
4-> 0
E P
i 0 0
O rH
S &
-H ฐ B
Z ฐ 8
"> 0 B
. T-H -H
J.-S'"
'"'HO
' t4H
? -H 0
v p to
E 3
cd cd
3 0
cr cu
XI
3 rl
B o
H rl
E rl
H 0
B
rl -H
O
t4H 0
XI
CM
X
0 cd
X3
a) cd
H p
cd
CD T3
0
CO rl
^-^ 0
p
TJ cd
ง *
O 3
tp cd
rl
0
E 0
Z 0
CO
1
B.
Z *
-------
- 30 -
Confirmation samples. Quantitative. The data presented in Tables 6 and
7 show good quantitative confirmation of the results of the routine analysis
of the six selected compounds in the raw and finished waters in the 80
locations. Because of the increased sensitivity of the confirmation
method used, analysis by that technique often produced a low measurable
concentration where the routine method did not find the compound. This is
not an inconsistency. The differences between the concentrations of the
routine and confirmation analyses in a few cases are not considered to be
significant.
Qualitative. The data in Table 7 show that the compounds quantitated
by the routine analysis were the correct compounds. In no case did the
routine analysis ever quantitate a given compound that later had a negative
confirmation by GC/MS when the most sensitive GC/MS method was used.
This did happen occasionally when the less sensitive GC/MS method as
described in this paper was used. In a few cases, because the sensitive
GC/MS method used a larger sample for purging, this technique would
detect the presence of a compound when none was found by the routine
procedure. These are not inconsistencies.
-------
- 31 -
^o
to CM r-^ to
[i, . O ...
2 O
PU| PU fT, ["] ( fT4
222 B, 2 2
^-* ' ^-_^ ^^ V^^ / \^_ /
/*~*\ /> t~\ /"%
PU UH PL, PL, PL,
^ rS r2I PH ^ ^
v^_^ ^ ^/ ^^ ^_/ \^_r/
/-^
/ -\ / \ CN /-^
J"T l prH f_Tt |*T |
222 B, o 2
x^-' ^ ' p?^ v^_x \ป^
"H **"** /^"N /"\ <*~^
* CM I-H cn -^J-
p^, to
52: o o o o
vป ' ^/ O ^ ' '" '
-p
r^
o
H
+J
O S
0 0
C 2
c
0
u to
rH
> rH
X rt
?H PM
3 X
Xt 0 0
0 P rH
P -P 0
rt -H I-}
s J
, ป
,CJ t-
f S
u7
UH 0 2
^^ V^ ^-^
r \
rH / \
Pt,
tw o 2;
^ vซ/ v t
/ \ / \
pLi PU
pLt ^ ^
ฃS *" ' v^-/
/"N /"*\
ry^ PL
^ S 2/
,r > / ^
fT ( p^
UH 2 2
55 v-/ v^
^ ^ / ^
CM rt
to
rH O
O ^ ' v '
ซ\
^f^J
W)
ฃ_t
3 rt
XI -H
w C
p rt
+J >
H rH
OH X
(/)
ฃg
3 0
O OH
CO
.
\o
rH
BH
|y|
B.
2
B,
2
B,
2
UH
2
CM
O
V
Xi
p
3
o
^0
r,
C
O
P
0
rH
rt
Xi
u
.
o
to
/ \
UH
2
v <
/ \
UH
2
^ '
f \
P-,
2
V '
ff*^
B.
2
V /
f~~^
UH
2
^ '
/ \
to
o
^ /
rt
r^
H
rH
O
rt
UH
2
B.
2
UH
2
UH
2
UH
2
UH
2
rt
rH
^
O
\
X
p
H
u
rt
g
o
s
rH
f^l
o
.
rH
LO
B,
2
V /
/ \
LO
.
i i
^ i
i >
UH
2
**-S
/^
B,
2
** . '
/ V
B,
2
' ^
/_ ^
LO
o
< '
f \
UH
2
N '
f^^
UH
2
V X
/ \
UH
2
< i
t \
UH
2
V 1
f-~\
UH
\ t
f N
UH
2
^ t
-------
- 32 -
r?
/ >
*0
p
c
o
U
v_*ป
^Q
0
rH
XI
rt
H
O
E~
i
o
o
r*
<3
i
o
rH
O
r-H
6
H
T3
0
rt
0
H
^
o
r-H
rC
0
o
MH
O
S
0
rH
CQ
fl^
C
rt
4->
0
G
0
C
rt
+->
Q>
O
O
O
rC
C
O
H
4->
rt
O
o
ป_3
/-^ f-^
tU UH
2^5 ง
U* UH
fc 2 2 UH
r-~\ f~\
PL, PL,
UH 2, 2, UH
/\ *"*
UH U,
U. 2 2 U-
Hป S^^y \^ / *jgป
/""N /"""*
UH tU
PL, Z 2 U.
l-jy. y ^/ t^^/ ^J
_ _
UH UH
tu 2 2 0
2 ^^ ^-^ V
O
O "0
S. rt
^ o
X -H
P O
H U
U
\
l/> O
rt TH
W XI
pj fl>
rt r*
^ DH
o to
vO ^O
,_^
1
/" \
UH
S
/"N
UH
2
^^
UH
2^
/-^\
UH
\W*
UH
S
/ v
1
UH
S
/"^
UH
^r
VHM/
^-^
UH
S
/"v
UH
V.^
-^*s
UH
^M/
/ % /-N
UH UH
UH 2 2
UH UH
UH 2 2
2 **-> ^
f \ /-~\
UH tU
UH 2 2
2 v ' v '
.T*"^ ป^*^
UH UH
UH 2 2
2,1 ^"^ v*/
tO /"^ /^N
UH UH
O 2 2
v * .^ ^-^
/ v '^
tO UH UH
5, S
t(H
H
TH
rt
U
\
d
0
o
C
o
u
rH
t^ซ
UH
UH
r^-t"
UH
2
tu
2
UH
2
UH
2
C
0
bO
V
f-H
o
ป*
in
*H
i-H
r-H
rt
^
0
U
CTi
t^-
^^
UH
^-N
i-H
o
v^_/
/^-\
UH
Z
V /
/^-\
tu
2
V /
/ \
2
*^ '
in
o
^ ^
,^-N
UH
S
i-H
O
\ป '
/ -x
UH
J2J
v_/
/ -\
UH
^
^ '
/ \
UH
V^
(N
o
V '
ฃ
C
o
rt
o
8
_] 0
x; P
o
tH X
rt X)
s I
x.o
^ri
C/3
rH -H
a> w
t^^
S rt
oS
x;
p c
x.S
XI P
rt
a>
x
rH
O
o C rt
4-1 O rH
rH O O
i-J
V) -H
H P X!
w rt o
X P
rH
rt
a>
t/>
rt' rt 0)
rt P
c C
X
rH
rt
S .
O
6 3
O w
O
4-1
0 X
tH P
rt -H
rH
rt
0 >' o*
tH
rt t/i *"C3
H rt
O wi nj
O 0
p x: -P
w p rt
H C CU
TH 0 S
rt rt O
P PHTH
rt 0
*X3 Cซ ^
H 0
P Q
!/> rt
fH P W
H rt t3
4-1 T3 O
0 0 -P
-------
- 33 -
CD
iI
ง
H
1 t
CD
H
4->
cd
H
pj
cd
^j
o"
CD
H
1 >
cd
H
i-H
3
o;
5-4
0%
,p
cd
5:
TJ
CD
I/)
H
(3
H
PL,
(4.4
O
_i
Co
4->
cd
Q
t/)
H
X
r- 1
cd
Ci
**
o
H
4->
cd
S
H
^4-4
s
o
u
i
o
O CD
i rH f3
1 43 cd
CN CJ 43
r-l Q CD
CD
f3 1 -H
O Cd rH
43 (H O
M 4-> iH
cd O 43
U t- 0
P
O
o
6
o
H
CO
/ s
1 1 1
~~^ O 1 CD
bfl S O P5
3- O in cd
f-i O 43
f3 43 rH 4->
H -H 43 CD
Q 0 S
cd
^J
cd i
13 O
ฃH (U
r-l IOC
r-l O r-l Cd
< E 43 43
v-/ O 0 4->
rH "H CD
CQ -O g
S
O
O
rH
O
1 1
u
c
o
H
4_)
cd
u
o
CN ^ \ / X
tin PL,
0 Z Z
1 v 1 V
LO LO
CN O O
W ^v>/ V ^
/"^v ("""N
[T^ rr t
rH )x^ ^
V *ป ' v* '
(gj
JC f~*\ f~~\
\O [XH ft,
^ 2H
O v ' ^ '
/ ^ ^ ^
LO \O
Jc
O LO LO
+ C? c7
K
^ rH rH
O^ vQ vO
^^-/ v-^
P
O
H
40
u
CD
G
o
U
ft
X
fn
3
J^
O
4->
cd
.
CN
y*-^ / %
/ \ / x tO CN
CN PL, PL,
Z Z * PL, 0 0
O ^^ ' v ' ^O Z v^ ^ J
V
/^"^ /"~s /* \ /^ ^
rt tO tO i-H
*
CN O O BH U, O O
v * ' ' ' Z Z ' ' % '
/ N /~~"t /* N /^"*V
PL, PL, * PL, PL,
PL, Z Z PL, PL, Z Z
Z ** J ** ' Z Z v~*/ "^^ '
/""^ /^^
tO ^-N CM /%
fTi *^ ' OH
0 Z * 0 Z
CN ' ' ' ' LO O ' ' '
t \ i \ ^ s r *\
>O t LO LO
\O G^ r^ * rH rH
*^^\ /~^ /^\
i 1 O l-ป i *
. CN
rH LO * O
O"> LO tO \O rH C?i
LO v ' > ' OO OO v ' v '
i-
X
CD
W +-
CD cd
>-3 * ft,
^ n
CD cd 43
Z a. oo
3
cd 43
i 1 -H (ft
rH 43 4J
Cd PH 4J
PL, rH .H
0 CL,
CD T)
rH Cd 43
4-> -H 3
H 43 O
j a, en
t^. rH \O
rH rH
*
CN
O
V
*
ft l
^
*
to
*
LO
to
*
oo
^
*
rH
rH
to
*
*
cd
H
(H
O
rH
PL,
ft
*H
S
rt
H
IS
ป
CM
CM
*
*sj"
tu *
^^ ^^
V
PL, *
Z CN
V
OO *
PL,
CO p^|
CO
*
O *
*
* IO
Ol i-H
*
LO *
Cn LO
rH *?!
*
*
O
U -H
43
CO O
A ft
e -H
O 4J
4-> Cd
w c
-------
- 34 -
t \
13
P
O
U
V /
r-~
rH
O
rt
H
i
O
O !H O
H H-> f-l
Cti 0) ,C
U H 0
O
S
O
rH
CQ
1
O 1 O
6 o c
O rH CIS
(H 0 43
fl rH +J
H rC
1 O C
O rH OS
6 X rC
O U -P
rH -H 0)
CO T3 E
O
o
rH
O
rH
|r^
U
r^
o
H
P
rt
o
o
!-]
CM * / v ^ \
["T | pr |
0 0 Z Z IX, PH
V V v_ / v_ / g; 2;
/"^i /^^*
\O OO
*
CM CM O O PL, *
y y v_/ v ' g to
/ v
pi
* Z PH *
Tf \O ' ' * ' Z LO
/~"^ ***\
LO Oป
* to O * *
rH O rH rH P, C7>
rH CM ' ' ' ' Z rH
/^~N /^%
OO LO
* to 00 * *
* OO CM CM PH OO
OO CM ^ ' ' ' Z "*
f ~N /^~\
rH LO
*
* O LO OO *
* ^f ^t "^ OO
00 <* > ' '' O 00
ci
o
*
> * I/)
X nS ni
O 4-1 S ")
H -H O CS
4S U rH Cj
O X
rt
^6 rt
CO S nj
ox i AJ
P nj 3 0)
X rH 4-> (^
CS AS H-> O
Q 0 OH
rH i 1 LO 00
ซJ LO LO LO
(X,
Z
f T |
PL,
Z
jc
CM
*
OO
*
^f
CM
^
O
rS
\
X
4_>
H
U
10
eS
t/i
C
rt
O
^o
^-^
[Xi
z
V 1
/ ป
to
O
/ >
PL,
Z
V /
/ \
CM
LO
O
v /
^ \
\Q
LO
v t
i >
rH
IO
CM
UH
Z
-
Z Z
to to
o o
IX, PL,
W. Z Z
Z '' ''
IX,
Z
(x,
Z
UH
PL, PL,
Z Z
PL,
Z
to
tL.
CM Z
v v'
IX,
Z
00 00
o o
CM
00
to
LO
CM
vO
CM
O
3
O
U
BJ
U
C
O
U
LO
vO
-------
- 35 -
i >
T)
4J
C
0
U
v /
O
rH
,0
rt
H
CM
ft
rH
pj
O
,0
fH
rt
U
i
o
e
0
<">
rt
Q
1
O
e
o
pa
o
o
rH
A
CJ
rt
T3
|
rt
fH
CD
H
1
O
&H
O
rH
r^
O
1
O
H
O
rH
rฃ~
O
rt
CD
C
rt
fj
4_)
CD
CD
T)
rt
O
rH
f
O
o
HH
O
e
o
fH
CQ
CD
C
rt
r<^{
p
(U
g
0)
s
o
o
!H
O
i i
^cj
U
C
O
rt
4J
rt
o
o
^-^
[i,
PL, PL, PL, Z
^ ^ j^ \ /
( s
CM
.
U, PL, PL, O
222^
fc
* PL, LX, 2
t-- 2 2 ^
/-\
* * tL,
*3- PL, fc 2
to 2 2 '-'
^-*^
K vO
* Cft
tO * rH
to o to * '
^
o
* * * IO
rH LO VO CN
VO rH CN ^-^
*
*
(3
o -
4J C
SH bo O
rt C M
rH -rt CD
rt rC C
U w (D
rt
W t/)
O ft -rt
rH 0) rH
rt rH rH
o, +-> rt
rt
&
o
rCn
o
to
rt
bo
0)
rt
+J
rt
4->
rt
rH
rt
0*
CD
^
rt
4->
rt
tn
o
a,
rt
W)
4J
rt
u
rt
*T3
C
M
1
4t
rH
"0
o
4->
i
CD
^
rt
4->
rt
C
CD
X
T3
o>
4->
O
CD
rH
rH
O
o
rt
rt
t)
C/5
S
x^
CJ
CD
1
*
*
ormed by the Water Supply Research Laboratory.
HH
fH
CD
PH
I/I
rt
to
X
rH
rt
c
rt
X
i
rt
rH
P.
CD
A
4J
B
0
^1
M-l
0)
cd
*T3
CD
4J
rt
rH
rt
4J
T3
CD
-C^
p
tive confirmation analysis performed by the Methods
oi
H
^
c
otj
ฃ3
f
(D
(p^
P
ง
4-(
CD
fn
rt
^"""N
Vw-^
I/)
H
I/)
0>
X
(^
d)
5_j
rt
cx
(3
rt
rt
4->
rt
T3
CD
f"j
1 I
oratory.
X)
rt
nJ
fjC1^
o
rH
rt
(/)
4)
OS
o
c
rt
rH
3
t/5
t/)
X
+->
rt
rH
rt
J3
ex
o
Co
^J
c
CD
a
o
rH
>
CD
Q
m
TJ
CD
4->
Q,
|f
ij
jj
ctf
IJ
o
ฃ
o
rt
4_>
rt
p;
rt
B
0)
O
TJ
CD
s
rj
CD
d)
^
rt
+J
rt
f .
rH
13
O'
I
t-
-------
- 36 -
Discussion
Occurrence of organics. Trihalomethanes. The first objective of the NORS
was to determine the extent of chlorination bypproducts in finished drinking
8 9
water as reported by Rook and Bellar, Lichtenberg and Kroner . To meet
this objective, raw and finished water from 80 locations, representing a
wide variety of raw water sources and water treatment practices, were
sampled for the four trihalomethanes -- chloroform (trichloromethane), bromodi-
choromethane, dibromochloromethane, and bromoform (tribromomethane).
In general,these four compounds were not fouid in the raw waters tested.
Chloroform was present in 49 locations in concentrations of less than 1 yg/1 with
the exception of Whiting, Indiana that was receiving chlorinated raw water.
Bromodichloromethane was present in 7 locations in concentrations of less
than 0.8 yg/1 with the exception of Whiting, Indiana. Dibromochloromethane
was found only in Whiting, Indiana's raw water. Bromoform was not found in any
of the locations tested. Therefore, the presence of any of these four
compounds in the finished water was concluded to be caused by chlorination
practices.
All of the systems investigated disinfected, but in one system, the only
treatment was ozonation. All of the finished waters tested contained some
chloroform although the system described above contained less than 0.1 yg/1.
A number of finished waters did not contain bromodichloromethane, dibromo-
chloromethane, and bromoform; however, their presence was frequent enough;
to be considered widespread throughout the finished waters of the nation.
Although the range of concentrations found for each of the four
trihalomethanes varied greatly for the type of systems surveyed, the
concentrations of each of the compounds was not evenly distributed throughout
the range, but were grouped toward the lower end of the range. Therefore
high concentrations of these parameters occurred infrequently in this study.
-------
- 37 -
Note that many groundwater supplies in the United States do not chlorinate
and, therefore, probably do not contain any trihalomethane, but that none of
these supplies were included in the Survey. To show the central tendency
of the data, Figure 2 presents the frequency distribution of the "trihalomethanes.
Based on Figure 2, the theoretical finished water with the median concentration
(one-half of the data above and below) of each compound, would contain about
21 ig/1 of chloroform, 6 yg/1 of bromodichloromethane, 1.2 yg/1 of dibromo-
chloromethane, and bromoform below the detection limit of the analytical method
used.
Most of the finished waters had concentrations of the four trihalomethanes
that became less in the same order as those in the theoretical "median" water
described above; this was not true in all cases, however. The reason for
concentrations of bromoform and dibromochlormethane being greater than the
bromodichlormethane and chloroform in some finished waters is not known at this
Q
time. Rook has postulated that if bromide was present in a water, the chlorine
will oxidize the bromide to bromine and the higher molecular weight bromo-
compounds would be formed. Whether this phenomenon occurred in some of the
fnished waters surveyed is not known, but no data was developed to refute
this theory.
1,2-dichloroethane and carbon tetrachloride. Analysis for 1,2-dichloro-
ethane and carbon tetrachloride was also made on all samples because they had
been found previously in other drinking waters and had potential health
significance. In this Survey, these two compounds were not found in 67.5 percent
and 87.5 percent of the finished waters, respectively. In about one-third
of the cases where these compounds were present in the finished water, they
were also present in the raw water, indicating they were environmental
contaminants and were not created during water treatment. The cause for the
-------
- 38 -
300
0.1
2 5 10 2030 50 70 90959899
PERCENT, EQUAL TO OR LESS THAN GIVEN
CONCENTRATION
FIGURE 2 FREQUENCY DISTRIBUTION OF
TRIHALOMETHANE DATA
-------
39 -
appearance of these compounds in the finished water when they were not detected
in the raw water is not known at this time. One possibility is that this may
have merely been an artifact caused by the varying limit of detection of the
analysis, see Table 2.
Nonvolatile Total organic carbon. In addition to studying the six
specific compounds discussed above, an attempt was made to investigate the
general organic level in finished drinking waters by measuring the
nonvolatile total organic carbon (MVTOC) concentration in all 80 locations.
The range of these data was from less than 0.05 mg/1 to 12.2 mg/1, but
again, the data were grouped toward the lower end of the range (See
Figure 3). The median NVTOC concentration (one-half of the data above
and below) was 1.5 mg/1.
-------
5.0
z _
< E
ฐZ
ซ/ ฃ,
00 2.0
^ z
5ฐ
-------
- 41 -
Influence of source type and treatment practice on trihalomethane
formation. The second objective of the Survey was to determine, if possible,
the influence that the type of source water and the treatment practiced have
on the formation of chlorination byproducts. An initial examination of the
data indicated that the dominant factor influencing the creation of
chlorination byproducts was the general organic level of the water, provided
sufficient chlorine was added to produce a chlorine residual at the time of
sampling.
To test this idea, the total trihalomethane concentration for each
finished water was first calcualted by dividing each of the four concentrations
by the appropriate molecular weight and adding the quotients together. This
yielded a total trihalomethane (TTHM) concentration in yMoles/liter*. The
advantage of this procedure is that the concentrationsof all four compounds
are reflected in a single number. Analytic techniques to determine this parameter
directly are presently being considered.
These data were then plotted against the NVTOC concentration of the raw
water. The TTHM data were divided into NVTOC cells in ascending order, each
cell having a span of 0.5 mg/1 NVTOC. The average TTHM concentration was
then calculated for each cell and plotted against the appropriate NVTOC
concentration. This analysis is appropriate based on the assumption that
each cell is sufficiently large and heterogeneous, arbitrarily taken as 4 or
more, with respect to the other variables that their influence is damped out by
the averaging process.
Note: 1 yM/1 TTHM = 119 yg/1 chloroform if only chloroform is present.
-------
- 42 -
The data plotted in Figure 4 has a correlation coefficient of 0.98, in spite
of any error because of suspended solids in some of the raw water samples.
Because of the scatter of the individual data points, the correlation coefficient
is 0.75 when all of the data are considered. This shows that because most
waters contain a chlorine residual (meaning an excess of one of the reactants
is present), the concentration of the product (TTHM) is related to the
concentrationof the other reactants (unknown precursors) and further that the
NVTOC concentration is a reasonable indication of their concentration.
To examine the data another way, the chlorine demand (total chlorine
added minus total chlorine residual) was calculated for each location. On the
basis that the formation of TTHM exerted some of the chlorine demand, chlorine
demand was plotted verses average TTHM concentration (averaged over 1 mg/1
wide chlorine demand cells) in Figure 5. The correlation coefficient of
these data is 0.85, 0.61 when all of the individual data points are considered.
Although these parameters are related, the correlation is probably influenced
by the other forms of chlorine demand. Therefore, raw water NVTOC concentration
was chosen as the dominant independent variable for the analyses that follow.
After this conclusion was reached all of the data were divided into
six NVTOC concentration cells, 0-1 mg/1, 1-2 mg/1, 2-3 mg/1, 3-4 mg/1, 4-5 mg/1
and greater than 5 mg/1, to eliminate the influence of that variable and were
then sorted so that like source types and treatment practices were together.
-------
- 43 -
0.8
u
z
o
u
LU
z.
LU
O
x
OL
0,7
0.6
0.5
0.4
0,3
0.2
0.1
(4) H
(6)
9(7) (NO.) - NUMBER OF
SUPPLIES IN NVTOC CELL
'(9)
NOTE: I fl MOLE/LITER
TTHM = 119 mg/l
CHLOROFORM IF ONLY
CHLOROFORM PRESENT
FIGURE 4,
012345
RAW WATER NON-VOLATILE TOTAL ORGANIC
CARBON CONCENTRATION, mg/l
CORRELATION OF TOTAL TRIHALOMETHANE
AND NON-VOLATILE TOTALORGANIC
CARBON CONCENTRATIONS
-------
- 44 -
z
o
1.0
0.9
0.8
0.7
0.6
Z
8 0.5
LU
Z
<
X
0.4
s ฐ-3
O
<
Of
LU
0.2
0.1
I I I I I
(7)
(20)
(7)
(5)
(NO.) = NUMBER OF LOCATIONS
IN EACH CL2 DEMAND CELL
NOTE: IjuM/liter TTHM = 119 AI9/I
CHLOROFORM IF IT WAS ALL
CHLOROFORM
I
I
I
I
I
0123456
CHLORINE DEMAND (TOTAL DOSE - TOTAL RESIDUAL), mg/l
FIGURE 5 CORRELATION OF TOTAL TRIHALOMETHANE
CONCENTRATION WITH CHLORINE DEMAND
-------
- 45 -
Source Influence. Except in the upper NVTOC cell, groundwater sources had
a lower average TTHM concentration than surface waters (See Table 8). When
all NVTOC cells are considered, not much difference existed between the
various types of surface water, although river water sources had a higher
average TTHM concentration in _4 of the 6 NVTOC cells.
Table 8
Source Influence
Raw NVTOC Range, rag/1
Category
All Locations
Ground Water
River Water
Impounded Water
0-1
Avg.
TTHM
Cone.
n* yM/1
14 0.09
10 0.06
3 0.19
1 0.13
1-2
Avg.
TTHM
Cone.
n yM/1
16 0.27
2 0.21
6 0.27
8 0.29
2-3
Avg.
TTHM
Cone
n uM/1
13 0.34
1 0.11
6 0.46
6 0.27
3-4
Avg.
TTHM
Cone.
n vM/1
15 0.52
1 0.19
12 0.55
2 0.50
4-5
Avg.
TTHM
Cone.
n uM/1
9 0.67
Q
5 0.66
4 0.68
>5
Avg.
TTHM
Cone.
n yM/1
13 1.23
2 1.65
7 1.35
4 0.82
*Number of locations.
-------
- 46 -
Treatment influence. Higher average TTHM concentrations were observed
at locations were raw water chlorination was practiced (see Table ง). An attempt
was made to relate raw water chlorine dose to average TTHM production, but the
number of locations in each cell was too small to produce meaningful data.
The trend of average TTHM production was generally higher as raw water chlorine
dose increased, but the data were quite variable.
The data on chlorine residual indicated that finished waters that did
not contain much free chlorine residual (Item 5, Table 9) had lower TTHM
concentrations than systems that had higher free chlorine residuals.
Table 9
Chlorination Practice Influence
Raw NVTOC Range, mg/1
Category
All Locations
Raw Water Chlorination
No Raw Water
Chlorination
<0.4 mg/1 Combined Cl
Residual, >0.4 mg/1
Free Cl_ Residual
<0.4 mg/1 Free C10
0-1
Avg.
TTHM
Cone.
n* yM/1
14 0.09
4 0.18
10 0.06
5 0.11
7 0.05
1-2
Avg.
TTHM
Cone.
n yM/1
16 0.27
14 0.26
2 0.32
9 0.32
5 0.15
2-3
Avg.
TTHM
Cone.
n yM/1
13 0.34
9 0.36
4 0.31
8 0.40
3 0.13
3-4
Avg.
TTHM
Cone.
n yM/1
15 0.52
13 0.55
2 0.34
6 0.48
3 0.30
4-5
Avg.
TTHM
Cone.
n yM/1
9 0.67
7 0.73
2 0.47
2 0.51
3 0.47
>5
Avg.
TTHM
Cone.
n pM/1
13 1.23
11 1.27
2 1.06
4 1.56
6 0.63
Residual
*Number of locations.
-------
All of the locations that practice filtration were sorted into NVTOC
concentration cells and then resorted based on the use of polyelectrolytes,
either as a coagulant^or filter-aid. This was to gain insight as to whether
or not polyelectrolytes could act as a precursor for TTHM formation. In the
study group, the polyelectrolyte dose varied from 0.02 mg/1 to 7.7 mg/1 on the
days of sampling. At two locations the dose was unknown. In Table 10, the
indication can be seen that using polyelectrolyte resulted in higher average TTHM
concentrations in all NVTOC cells. Additional controlled experiments must
be done to definitely establish whether the polyelectrolyte is reacting directly
with the chlorine.
Table 10
Filtration Practice Influence
Raw NVTOC Range, mg/1
Category
All Locations
All Filter Plants
w/Polyelectrolyte
w/o Polyelectrolyte
0-1
Avg.
TTHM
Cone.
n* yM/1
14 0.09
5 0.17
2 0.21
3 0.14
1-2
Avg.
TTHM
Cone.
n yM/1
16 0.27
14 0.28
4 0.32
10 0.27
2-3
Avg.
TTHM
Cone
n yM/1
13 0.34
9 0.36
2 0.53
7 0.31
3-4
Avg.
TTHM
Cone
n yM/1
15 0.52
13 0.56
3 0.68
10 0.52
4-5
Avg.
TTHM
Cone
n yM/1
9 0.67
9 0.67
3 0.94
6 0.54
>5
Avg.
TTHM
Cone.
N yM/1
13 1.23
13 1.23
2 2.56
11 1.01
*Number of locations
Based on a comparison of NVTOC concentrations in the
in the 63 locations where filtration was practiced a
NVTOC occurred, on the average.
raw and finished waters
30 percent reduction in
-------
- 48 -
In the 17 treatment plants where precipitative softening is practiced,
all but 3 had a pH of 9.0 or more in the finished water. Two plants produced
a finished water with a pH over 10.0. The data in Table 11 shows a higher
TTHM concentration in 4 of the NVTOC cells. Further, the average TTHM
concentration for all 17 precipitative softening plants was 0.84 yM/1, while
the average TTHM concentration for all 80 locations was 0.49 yM/1. This
follows the expected trend for pH dependency of the classical haloform
reaction and indicates that chlorination at higher pH will produce higher
concentrations of trihalomethanes 4 other conditions being equal.
Table 11
Influence of Precipitative Softening
Raw NVTOC Range, mg/1
Category
All Locations
Ppt Softening
0-1
Avg.
TTHM
Cone.
n* yM/1
14 0.09
2 0.13
1-2
Avg.
TTHM
Cone.
n yM/1
16 0.27
2 0.49
2-3
Avg.
TTHM
Cone.
n yM/1
13 0.34
0 -
3-4
Avg.
TTHM
Cone.
n yM/1
15 0.52
5 0.55
4-5
Avg.
TTHM
Cone.
n yM/1
9 0.67
2 0.35
>5
Avg.
TTHM
Cone.
n yM/1
13 1.23
6 1.61
*Number of locations.
-------
- 49 -
In the 19 treatment plants using powdered activated carbon (one plant
sampled twice) the dosage varied from 0.6 mg/1 to 17.5 mg/1 with a median of 2.3.mg/l,
All of these plants treated surface waters. From Table 12 it can be seen that
(except for the highest NVTOC cell) locations where powdered activated
carbon (PAC) was used had average TTHM concentrations lower than those
locations not using PAC. The PAC dosage for the 4 plants in the highest NVTOC
cell was not sufficiently different from the overall median dose of 2.3 mg/1
iftoted above to readily explain the apparent difference in treatment performance.
The NVTOC concentration in these waters may have been too high to be influenced
by the PAC or the number of locations in this cell may just be too small to
draw reliable conclusions.
Table 12
Influence of Powdered Activated Carbon
Raw NVTOC Range, mg/1
Category
All Locations
All Filter Plants
With PAC
Without PAC
0-1
Avg.
TTHM
Cone.
n* yM/1
14 0.09
5 0.17
2 0.16
3 0.18
1-2
Avg.
TTHM
Cone.
n yM/1
16 0.27
14 0.28
3 0.20
12 0.28
2-3
Avg.
TTHM
Cone.
n yM/1
13 0.34
9 0.36
1 0.13
8 0.39
3-4
Avg.
TTHM
Cone,
n uM/1
15 0.52
13 0.56
6 0.46
7 0.64
4-5
Avg.
TTHM
Cone.
n yM/1
9 0.67
9 0.67
4 0.24
5 1.02
>5
Avg.
TTHM
Cone.
n yM/1
13 1.23
13 1.23
4 1.29
9 1.21
*Number of locations
-------
- 50 -
Only eight water treatment plants used granular activated carbon (GAC)
as a combination filtration/adsorption media, and this number is too small to
make an analysis as above. All treat surface water and chlorinate raw water,
and all but one had >0.4 mg/1 free residual in the finished water, so soire of
the variables noted above were eliminated. Because all but one of the location-.
sampled were using granular activated carbon that had been in place for at
least several months, the activated carbon was exhausted for NVTOC removal in
these locations. This is shown in Table 13. The average NVTQC removal at
these locations was not much higher than the "greater Lh.iri '-> -, pe.\,en> '\ ;Tt)C
(footnote, Table 10) .
removal" previously reported for all coagulation-filtration plants/
the TTHM concentration in these finished waters being higher than the TTHM
concentration in the theoretical "median" finished water for the entire survey
in 6 out of these 7 locations using GAC exhausted for NVTOC rejfiu.'al is nut
surprising. This is also true when the data are examined on a "TTHM production
per unit of NVTOC" basis.
Table 13
Summary of Granular Activated Carbon Plants
TTHM TTHH/
1 Cone. Fin. NVTOC,
Plant
A
B
C
D
E
F
G
Avg. (A to G)
Fresh Coal-Base GAC'_ H
Fresh Lignite-base GAC
(Resample)-B
Finished Water
NVTOC
Cone.
mg/1
1.0
1.4
1.6
1.9
3.2
4.2
4.4
0.2
1.4
% Rem<
of NV'
55
30
56
53
41
30
32
42
95
79
0 . 31 0 . 31
0,14 0.10
0,82 0.51
0.60 0.31
1 , 36 0,4^
1.19 0,28
0,74 0,18
0.30
0.08
Theoretical median water* 1.5 - 0.22
*See Figures 2 and 3 for median concentrations.
-------
Shortly after the Survey samples were taken at one of these locations (B),
the granular activated carbon was removed and replaced with virgin lignite-
base material. 'Ihis location was resampled in an effort to evaluate the
performance of fresh granular activated carbon. At another location fresh
coal base granular activated carbon had been in place only 2 weeks at the
time of sampling, The summary data for the fresh materials (Table 13) show
a marked improvement in the parameters listed, indicating the effectiveness
of fresh grara'iac activated carbon for treatment.
Another attempt was made to evaluate the performance of granular
activated carbon for treating a variety of waters by monitoring the
activate^ caibor- units installed in the five locations where additional
sampling WP.S being undertaken. These samplers were 3-foot-long (90 cm)
colijnns ef COP 1-based granular activated carbon operated downflow at a
i:Iltrr'tion rate of 3.2 gallons per minute/square foot (8 meters/hour).
Finished waier v.as passed through them for 7 days. The contact time
KOS 1 *o 4 Tinuul"1'. Fresh granular activated carbon produced low NVTOC
.-jTiceni ration at first in all locations (Table 14), except Miami, Florida,
where tlie organic loซd was so heavy that a different treatment mode
would be needed to produce a lower NVTOC concentration.
-------
- 52 -
Table 14
Performance of Fresh Coal-Based Granular Activated Carbon Samplers Treating
Finished Water
Location
Day
NVTOC Concentration - mg/1
Influent to Effluent from NVTOC
Sampler Sampler Removed
Miami,
Florida
Seattle,
Washington
Ottumwa,
Iowa
Philadelphia,
Pennsylvania
Cincinnati,
Ohio
0
7
0
7
0
7
0
7
0
7
8.1
7.1
1.9
0.8
3.6
3.4
2.0
1.9
1.2
1.6
1.3
3.5
1.9*
0.05
1.6*
0.9
0.3
0.5
0.1
0.1
84%
51%
0%*
94%
56%*
73%
85%
74%
92%
94%
*Data Suspect
-------
- 53 -
The two water utilities in the United States (at Strasburg, Pennsylvania
and Whiting, Indiana) currently using ozone as a treatment unit process were
included in this Survey. At Strasburg ozonation is the only treatment. The
source in this water system is 12 springs -- goundwater of high quality.
None of the six selected compounds were found in the source water, and it
had a very low NVTOC concentration. The ozone residual immediately after
application was 0.4 mg/1. The finished water was also of good quality.
Only chloroform at a concentration of less than 0.1 yg/1 was found in the
finished water. Although this concentration was the lowest found in the
Survey, it was a definite positive value. The possible source of this
slight amount of chloroform could be the oxidation of the chloride in the
water to chlorine by the ozone, with the subsequent formation of some
chloroform. The chloride level in this water was low, but sufficient
was present to account for the chloroform formation, if this reaction occurs.
The situation was different in Whiting, Indiana. Here, Lake Michigan
is the source. The source water contained 0.1 yg/1 of chloroform and less
than 0.8 yg/1 of bromodichloromethane. None of the other selected compounds
were found in the source water. A nearby refinery takes water from Lake
Michigan, chlorinates it, uses most of the water itself and sends some to
the Whiting water treatment plant. Therefore, when the sampled water arrived
at the water plant, it contained 16 yg/1 of chloroform, 11 yg/1 of
bromodichloromethane, and 3 yg/1 of dibromochloromethane. The water was
then ozonated, with about 2 mg/1 being applied in two 21-foot (6.5m) deep
towers for a detention time of 6 to 8 minutes. Following ozonation water
was then chlorinated again, with ammonia added to produce a combined chlorine
-------
- 54 -
residual. The water is then coagulated and filtered before distribution.
The finished water contained only 0.5 yg/l of chloroform and 0.3 yg/l of
bromodichloromethane. At this time, the mechanism causing the reduction
in trihalomethane concentration is not known. These organics could have
been stripped from the water during the process of ozonation; they could
have been oxidized by the ozone, or they could merely have been lost to
the atmosphere during the passage through the open settling basin and
filters. Subsequent sampling at this plant has indicated that the latter
explanation is the most probable.
The data shown in Tables 8 and 9 indicate that the use of surface
water as a source, raw water chlroination, and the presence of greater:than
0.4 mg/1 free chlorine residual enhances the formation of trihalomethanes.
To test this indication the data were sorted on these bases. Out of the
entire survey, 39 locations met these three criteria. Of these, 13
finished waters had an NVTOC concentration equal to or less than the Survey
median concentration of 1.5 mg/1; the remaining 26 had greater than the
median NVTOC concentration. Of those 13 with a finished water NVTOC
concentration equal to or below the median concentration, 62% had a TTHM
concentration above the median TTHM concentration, indicating agreement with
the hypothesis. Of those 26 with a finished water NVTOC concentration greater
than the median concentration, only 19% had TTHM concentrations below the
median TTHM concentration, in opposition to the proposed hypothesis. This
indicates general agreement with the indications of Tables 8 and 9.
The difficulty with using these data, and those of Tables 10, 11, 12,
and 13 is that although the averages in the various NVTOC cells show some trends,
the individual data are quite scattered, for reasons unknown. Therefore,
while these analyses are useful to obtain an indication of source and
-------
- 55 -
treatment influences, carefully controlled experiments are needed in the
future to understand these reactions more exactly.
Alternate indicators of organic contaminant levels. Because organic
contaminants vary in toxicity, specific organic compounds should be monitored
in finished waters. This is the recommended procedure for monitoring
organochlorine pesticides, for example. Except for a few specific
examples, this approach is beyond the capabilities of most water utilities
and to some degree even is beyond the capabilities of researchers, given
the current state of organic analysis. For example, all individual organic
compounds present in water cannot now be identified and quantified.
In the absence of measuring for specific organic compounds, the next
best alternative is to measure some organic parameter that includes a large
number of organic compounds and assume that the level of this parameter is
proportional to the level of toxicity of the water. On this basis, carbon-
chloroform extract (CCE-m) was included in the proposed Interim Primary
Drinking Water Regulations.
In the NORS nonvolatile total organic carbon was the parameter chosen
to represent the concentration of organics in the water. In Figure 4 NVTOC
was shown to be generally proportional to trihalomethane formation, so a
measure such as this is probably useful, but little else is known about NVTOC.
In an effort to find an easier analytic procedure for monitoring the
organic level in water, three other measurements were made on each raw and
finished water in addition to NVTOC concentration. These were ultraviolet (UV)
absorption, emission fluorescence scan (EmFS) and the Rapid Fluorometric
Method (RFM). An attempt was made to correlate those parameters, although
different organics absorb UV to differing degrees and different organics
fluoresce to differing degrees. Therefore, although the a priori judgment
-------
- 56 -
was that those three parameters might not correlate with NVTOC concentrations
because they would be heavily influenced by the types of organics present
in the water, the} hypothesis that different waters would be sufficiently
similar to make these procedures useful was tested.
Just as particulates in some raw waters caused some error in the NVTOC
measurement, the resultant turbidity interfered with the UV, EmFS, and RFM
measurements; therefore these data were not (analyzed. Plots of NVTOC
concentration versus each parameter for finished water (Figures 6,7, and 8)
show a wide scatter of data. In Figure 6, 39 data points are included in a
one-milligram-per-liter-wide band and only 28 data points (up to an NVTOC
concentration of 3.5 mg/lj are excluded. The overall correlation, however,
is not very good. The two fluorescence techniques correlated reasonably well
with each other, but not with NVTOC concentration (see Figure 9).
At 11 Survey locations, concentrations of nonvolatile total organic
carbon and carbon chloroform extract (CCE-m) were measured simultaneously.
Figure 10 indicates that some correlation existed between these two
parameters, but the data are rather scattered.
12
Sylvia, Bancroft and Miller have proposed the RFM measurement as a
rapid indicator of CCE-m concentrations as they found good correlations
between these two parameters in the water they were studying. T.o test their
proposal on the NORS data, the 11 data pairs were plotted in Figure 11.
For these data, the RFM measurement was poorly correlated with CCE-m
concentration.
-------
- 57 -
.18
.16
.14
c
3
LU
u
Z .12
CO
O
to
OQ . ..
< .10
g
>
.08
m 0.6
Q
LU
I .04
CO
.02
Ok
.:* . ฎ
* '
.
*
01 2345
FINISHED WATER NVTOC, mg/l
FIGURE 6. CORRELATION BETWEEN ULTRA-VIOLET
ABSORBANCE AND NON-VOLATILE TOTAL
ORGANIC CARBON IN FINISHED WATER
-------
- 58 -
21
19
.r 17
c
3
O
O
X 15
u
of 13
LU
O
on
O 1
< 9
c*
0123456
FINISHED WATER NVTOC, mg/l
FIGURE 7. CORRELATION OF RAPID FLUOROMETRIC METHOD
AND NON-VOLATILE TOTAL ORGANIC COMPOUND
CARBON IN FINISHED WATER
-------
- 59 -
1900
1700
1500
Z
<
u
U
O
z
O
1100
900
700
500
300
i ซ
.
0
2345
FINISHED WATER NVTOC, mg/l
IGURE 8. CORRELATION OF EMISSION FLUORESCENCE
SCAN AND NON-VOLATILE TOTAL CARBON
IN FINISHED WATER
-------
- 60 -
2000
1800
1600
1400
S 1200
Z
LU
U
CO
Ul
O 1000
z
o
./> 800
I
LU
ee.
LU
S 600
o
UJ
I
^ 400
Z
200
I I
I I
I I I L 1 I I I I I I I I
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
FINISHED WATER RAPID FLUOROMETRIC METHOD, UNITS
FIGURE 9 CORRELATION OF THE TWO FLUORESCENCE
MEASUREMENTS PERFORMED ON FINISHED WATER
-------
- 61 -
O>
1.2
u
1.0
ป
u
0.8
2
o
X
u
O
03
CX.
0.4
0.2
01 23456
NON-VOLATILE TOTAL ORGANIC CARBON, mg/l
FIGURE 10. CORRELATION OF NON-VOLATILE TOTAL
ORGANIC CARBON WITH CARBON
CHLOROFORM EXTRACT
-------
- 62 -
20,
18
16
14
12
ง10
o
2 6
I | | | I
"0 0.2 0.4 0.6 0.8 1.0 1.2
FINISHED WATER CCE-m CONCENTRATION, mg/l
FIGURE 11 CORRELATION BETWEEN CARBON CHLOROFORM
EXTRACT CONCENTRATION AND THE RAPID
FLUOROMETRIC METHOD IN FINISHED WATER
-------
Significance of findings. Most water treatment plants are not designed
to remove soluble organic compounds from raw water, and disinfection creates
some compounds that were not originally present in the raw water. Therefore,
the finding that all finished waters in the Survey contained one type of
organic compounds or another was not surprising. The presence of an
organic compound in a finished water is not significant, however, unless
13 14
its concentration is such that it poses a health hazard ' . These
data, therefore, must be combined with health effects data before any
significance can be attached to the findings. If a health hazard is found
to exist with any contaminant, then the treatment information currently
being developed by the research program of the Water Supply Research
Laboratory ' must be applied to remove that contaminant,
-------
- 64 -
Summary and Conclusions
1, The four trihalomethanes: chloroform (trichloromethane),
bromodichloromethane, dibromochloromethane, and bromoform (tribromomethane),
are widespread in chlorinated drinking waters in the United States and result
from the chlorination treatment process.
2. The four trihalomethanes were not found or were present in low
concentrations in the raw waters tested. Carbon tetrachloride was not found
in the raw water of 95% of the locations surveyed. 1,2-Dichloroethane
was not found in the raw water of 86% of the locations investigated.
3, The median concentration of three of these compounds found in
finished water was: chloroform, 21 yg/1; bromodichloromethane, 6 yg/1; and
dibromochloromethane, 1.2 yg/1. Bromoform was not found in finished water in
68.8% of the supplies surveyed. The range of concentration of all four
trihalomethanes was: chloroform, less than 0.1 yg/1 to 311 yg/1;
bromodichloromethane, none found to 116 yg/1; dibromochloromethane, none
found to 100 yg/1; and bromoform, none found to 92 yg/1.
4. Carbon tetrachloride and 1,2-dichloroethane were not detected very
frequently nor found in high concentrations in finished water in the
Icoations studied. Specifically, carbon tetrachloride was not found in 87.5%
of the finished waters waters tested and the highest concentration found was
3 yg/1. 1,2-Dichloroethane was not found in 67.5% of the finished waters
tested, and the highest concentration found was 6 yg/1.
5. In general, total trihalomethane concentrations were related
to the organic content of the water, as measured by the nonvolatile total
organic carbon test, when sufficient chlorine was added to create a
chlorine residual.
-------
- 65 -
6. In general, when the following conditions occurred, higher
concentrations of total trihalomethanes were found: surface water as the
source water, raw water chlorination practiced, and more than 0.4 mg/1 free
chlorine residual present.
7. In general, where precipitative softening was practiced and the
finished water had a high pH, higher total trihalomethane concentrations were
found.
8. At 7 out of the 8 locations, based on the removal of nonvolatile
total organic carbon, the granular activated carbon being used as a combination
filtration/adsorption media was exhausted.
9. When fresh granular activated carbon (coal-base and lignite-base)
was monitored at two locations, removal of nonvolatile total organic carbon
was higher than average and the total trihalomethane concentration was
lower than the average values obtained from plants where the granular
activated carbon had not been replaced for some time.
10. Nonvolatile total organic carbon concentrations did not correlate
well with ultraviolet absorption, fluorescence, and carbon chloroform
extract concentration data and the Rapid Fluorometric Method did not correlate
well with carbon chloroform extract concentration data.
11. One location treated a high quality groundwater only with ozonation
for disinfection. The finished water at this location had the lowest total
trihalomethane concentration of any location surveyed. At another location,
water containing the trihalomethanes was ozonated. Through the treatment
plant the concentration of total trihalomethanes was reduced, but follow-up
sampling indicated the mechanism causing this reduction is most likely lost
to the atmosphere through the settling basins and filters.
-------
- 66 -
Acknowledgments
The authors wish to thank the many others who contributed to this paper:
The Regional Water Supply Engineers who collected all of the samples; the members
of the Methods Development and Quality Assurance Research Laboratory who made
most of the qualitative confirmation analyses; the members of the Criteria
Development Branch, Water Supply Research Laboratory who made the remainder
of the qualitative confirmation analyses; and to Mrs. Maura M. Lilly who
typed the manuscript.
-------
- 67 -
References
1. Coleman, W.E., Lingg, R.D., Melton, R.G. ง Kopfler, F,K., GC/MS
Techniques for the Identification of Volatile Organics in Tap Water.
In preparation (June 1975).
2. Lingg, R.D., Melton, R.G., Kopfler, R.K. ง Coleman, W.E., GC/MS
Techniques for the Quantitation of Volatile Organics in Tap Water.
In preparation (June 1975). ,
3. Tardiff, R.G., Budde, W.L., Coleman, W.E., DeMarco, J., Dressman, R.C.,
Eichelberger, J.W., Kaylor, W.H., Keith, L.H., Kopfler, F.K.,
Lingg, R.D., McCabe, L.J., Melton, R.G., ง Mullaney, J.L., Organic
Compounds in Drinking Water: A5-City Study. In preparation (June 1975).
4. Bellar, T.A. ง Lichtenberg, J.J., The Determination of Volatile
Organic Compounds at the yg/1 Level in Water by Gas Chromatography.
USEPA, National Environmental Research Center, Cincinnati, Ohio,
EPA-670/4-74-009, Nov. 1974. See also, Bellar, T.A. ง Lichtenberg, J.J.,
Determining Volatile Organics at the yg/1 Level in Water by Gas
Chromatography. Jour. AWWA, 66:739 (Dec. 1974).
5. Stevens, A.A. ง Symons, J.M., Analytical Considerations for Halogenated
Organic Removal Studies. In: Proceedings AWWA Water Quality
Technology Conference, December 2 $ 3, 1975, Dallas, Texas, pp. XXVI-1
(1975).
6. Dobbs, R.A., Wise, R.H. ง Dean, R.B., The Use of Ultraviolet Absorbance
for Monitoring the Total Organic Carbon Content of Water and Wastewater.
Water Research, 6:1173 (Oct. 1972).
7. Sylvia, A.E., Detection and Measurement of Microorganics in Drinking
Water. Jour. NEWWA, 87: No. 2 (Jun. 1973).
8. Rook, J.J., Formation of Haloforms During Chlorination of Natural
Waters. Water Treatment and Examination, 23: Part 2, 234 (1974).
9. Bellar, T.A., Lichtenberg, J.J., ง Kroner, R.C., The Occurrence of
Organohalides in Chlorinated Drinking Water, Jour. AWWA, 66:703
(Dec. 1974).
10. Interim Primary Drinking Water Standards. Federal Register, 40:No. 51,
Part II, 11190 (Mar. 14, 1975).
11. Buelow, R.W., Carswell, J.K. ง Symons, J.M., An Improved Method for
Determining Organics by Activated Carbon Adsorption and Solvent
Extraction (Parts I and II). Jour. AWWA, 65:57, 195 (Jan.-Marh. 1973).
-------
- 68 -
12. Sylvia, A.E., Bancroft, D.A. ฃ Miller, J.D., Detection and Measurement
of Microorganics in Drinking Water by Fluorescence. In: Proceedings
AWWA Water Quality Technology Conference, December 2 & 3, 1974,
Dallas, Texas, pp. XXVII-1 (1975).
13. Murphy, S.D., A Report - Assessment of Health Risk from Organics in
Drinking Water by an Ad Hoc Study Group to the Hazardous Materials
Advisory Committee - U.S. Environmental Protection Agency, Washington, D.C.,
(April 30, 1975). Mimeo, 59 pp, plus Attachments.
14. Tardiff, R.G., Craun, G.F., McCabe, L.J. and Bertozzi, P.E. , Preliminary
Assessment of Suspected Carcinogens in Drinking Water - Interim Report
to Congress, Appendix VII, Health Effects Caused by Exposure to
Contaminants. U.S. Environmental Protection Agency, Washington, D.C.,
(Jun. 1975). In Press.
15. Love, O.T., Jr., Carswell, J.K., Stevens, A.A., Sorg, T.J., Logsdon, G.S.,
and Symons, J.M., Preliminary Assessment of Suspected Carcinogens in
Drinking Water - Interim Report to Congress, Appendix VI, Preliminary
Results of Pilot Plants to Remove Water Contaminants, U.S. Environmental
Protection Agency, Washington, D.C. (Jun. 1975). In Press.
16. Love, O.T., Jr., Carswell, J.K., Stevens, A.A. and Symons, J.M.,
Treatment of Drinking Water for Prevention and Removal of Halogenated
Organic Compounds (An EPA Progress Report). Presented at the 95th
Annual Conference of the American Water Works Association, June 8-13,
1975, Minneapolis, Minn.
U S GOVERNMENT PRINTING OFFICE 1975 657-641/1009
------- |