United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-84-005 Jan. 1984
4MER& Project Summary
Alternative Water Disinfection
Schemes for Reduced
Trihalomethane Formation
Volume II. Algae as
Precursors for
Triha/omethanes in
Chlorinated Drinking Water
Kathryn F. Briley, Robert F. Williams, and Charles A. Sorber
Three species of algae (Anabaena
cylindrica, Scenedesmus quadricauda,
and Pediastrum boryanum) were inves-
tigated for their trihalomethane (THM)
formation potential in water treated
with chlorine. Algae were cultured and
the cells (algal biomass) were separated
from the extracellular products (ECPs)
at several points along the normal
growth curves of each species for a
separate study of their contributions as
THM precursors. The cells were resus-
pended in organic- and chlorine-demand-
free water, and the cells and ECPs were
then separately dosed with three chlo-
rine concentrations. The THMs formed
after 1 and 24 hr of chlorine contact
time were analyzed by the gas-sparging
technique and gas chromatography.
For each point examined along the
growth curve, growth was monitored
by both cell counts and fluorometric
assay of chlorophyll-a. Correlation of
the algae growth period and THM
production was observed. Furthermore,
significant levels of THM were produced
from both the ECPs and the isolated
algal cells of all three species when
dosed with chlorine. As expected, the
THM levels formed were related to the
free chlorine residual and to the TOC
levels observed. These findings suggest
that THMs may be partially reduced by
observation and control of the natural
phytoplankton communities in the wa-
ter sources for domestic water supplies.
This Project Summary was developed
by EPA's Municipal Environmental Re-
search Laboratory. Cincinnati. OH. to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
Chlorination of surface waters gener-
ally results in the formation of trihalo-
methanes (THMs). These compounds
result from a reaction or series of reac-
tions of chlorine with natural organic
precursor material present in the source
water. The exact nature of the precursor
material is not clear, though humic and
fulvic acids and algae have been impli-
cated as contributors. Chloroform (CHCI3)
is the THM found in greatest quantities in
finished waters, but dibromochlorometh-
ane (CHBr2CI), bromodichloromethane
(CHBrCI2), and bromoform (CHBr3) are
commonly found in varying quantities as
-------
well. The U.S. Environmental Protection
Agency (EPA) has established a maximum
contaminant level of 0.10 mg/Lfor THMs
in finished waters for communities with
populations of 10,000 or more. Attempts
to reduce total trihalomethanes (TTHMs)
below the established levels in an eco-
nomical way are hampered by the lack of
a complete understanding of the nature
of the THM precursors. Identification of
THM precursors would allow establish-
ment of techniques to remove them
selectively.
The purpose of this study was to
examine a class of potential THM pre-
cursors in surface waters. Specifically,
investigations were made of the contribu-
tion of algal cells and extracellular meta-
bolic products (ECPs) from algae to the
production of THMs. This information
may be useful for developing alternative
disinfection procedures that can allow
reduced levels of THMs in finished water
and still maintain effective water disin-
fection.
Experimental Procedures
To learn the extent to which algal cells
and ECPs might contribute to THM forma-
tion, three algal species were grown in
the laboratory under controlled condi-
tions. The cells were separated from the
ECPs, each fraction was chlorinated, and
THM concentrations were measured. The
latter were compared with algal cell
concentrations in the original suspension
to indicate the ability of algae to cause
THM problems.
The algal species chosen for study
were nonaxenic Anabaena cylindrica,
Pediastrumboryanum, and Scenedesmus
quadricauda. These species were select-
ed for study because they are easy to
grow in the laboratory and because they
are abundant in South Central Texas
lakes. The algae were grown using Bold*
basal pH 6.6 medium cultured in Erle-
meyer flasks with metal snap-on caps
that permitted some air flow to the
cultures without allowing contamination.
The cultures were further aerated by
frequent shaking. The algae were sub-
jected to 16 hr of light and 8 hr of dark
each day on a constant schedule with a
light source that ranged from 275 to 450
foot-candles. The temperature for algal
growth was maintained between 23° and
25°C.
For each experiment in this study,
algae from a stock culture with a known
number of cells/mL were inoculated into
"Mention of trade names or commercial products
does not constitute endorsement for use.
2 L of freshly prepared medium to make a
final dilution of approximately 103 cells/
mL. The algae were then cultured until
the growth phase desired for THM analy-
sis was reached.
The growth phases desired for THM
precursor studies were determined by
cell count techniques and chlorophyll-a
measurements. These growth-monitor-
ing techniques allowed algal growth
curves to be generated. From these
curves, the times chosen for THM pre-
cursor examination were selected with
the intention of encompassing the five
basic growth phases (lag, early expo-
nential, late exponential, early stationary,
and late stationary) of each species
employed. The death phase was not
examined. The times chosen for experi-
mentation along the growth curve of
Anabaena were days 0, 2, 4, 7, 10, 14,
and 21. Day 0 encompassed the brief, if
any, lag phase; days 2, 4, and 7 were
during the exponential growth phase, and
day 10 was transitional between the late
exponential and early stationary phases.
Days 14 and 21 were during the station-
ary phase. Similarly, for Pediastrum, days
0,2,7,10,15,21, and 28 were chosen for
study, and Scenedesmus experiments
were conducted on days 0, 2,4, 8,11,15,
and 21 of the growth cycle.
At the chosen time, a 270-mL aliquot
was removed from the 2-L experimental
culture. From this aliquot, 20 mL was
used for growth-monitoring measure-
ments done by cell enumeration and
fluorometric assay of the chlorophyll-a.
Duplicate fluorometric chlorophyll-a meas-
urements and cell counts were made on
the algae each time THM experiments
were performed.
The chlorophyll-a measurement was
•made using a fluorometer with a 5-60
excitation filter and a 2-28 emission filter.
Chlorophyll-a was extracted as described
in Standard Methods (Standard Methods
for the Examination of Water and Waste-
water, 14th edition; 1975; Amer. Public
Health Assoc., Amer. Water Works Assoc.,
Water Pollution Control Federation; pp.
1030-1032) for the spectrophotometric
procedure. A calibration curve for the
fluorometer was constructed using com-
mercially obtained chlorophyll-a. The
calibration curve obtained had a correla-
tion coefficient (R2) of 0.993 determined
by regression analysis. Reference sam-
ples were prepared by extraction in the
same manner without algae, and these
readings were subtracted from the fluo-
rescence readings before chlorophyll-a
calculations. Replicate measurements
were performed on each organism.
Cell count measurements were made
along the algal growth curves using a •
mechanical counter with an aperture
tube of 100-jUm diameter. The instrument
was calibrated using pollen and glass
spheres of known volumes, and the
optimal sensitivity settings for measuring
cell counts for each algal species were
determined. Commercially obtained elec-
trolyte solution was used for making the
cell dilutions and corresponding control
dilutions. The reference reading values
made using medium and diluent alone
were subtracted before cell count calcula-
tion. The algae were dispersed by vigorous
shaking with glass beads before cell
counts were performed.
After the 20-mL volume had been
removed for the growth measurement
procedures, the remaining 250 mL was
used for the THM precursor experiments.
The cells and their corresponding extra-
cellular products were separated by centri-
fugation at 2400 to 2600 rpm (1050 to
1250 x g) for 20 min. The ECPs and
medium (supernatant) were then poured
off and vacuum filtered using a 0.45-yum
(47-mm) membrane filter. The filtered
ECPs were transferred to a clean flask
and refrigerated at 4°C in the dark
overnight or until chlorine addition. Be-
fore chlorination, the ECPs were warmed
to room temperature. The pellet formed ,
after centrifugation (algal biomass) was
washed twice with 0.9 percent sterile
sodium chloride, resuspended, and recen-
trifuged. The supernatant was then dis-
carded. The washed cells were resus-
pended in 250 mLof organic-free, chlorine-
demand-free water. Approximately 10~4
moles of phosphate buffer (filtered
through a 0.45-/jm membrane filter) were
added to the resuspended cells for pH
control. The resuspended cell or ECP
aliquots were then divided into three 80-
mL samples, measured for pH and temper-
ature, and immediately dosed with the
desired amount of chlorine (for the ECPs,
7, 24, 33 mg/L; and for the cells, 1.5 and
5.9 mg/L). After 1 hr of chlorine contact
time, the pH, temperature, and chlorine
residuals (free and total) were again
measured in duplicate. Samples were
obtained for TOC, TKN, and NH3-N deter-
minations. A THM sample (1-hr contact
time) was collected in a vial containing
0.3 to 1.0 mLof 0.01 N sodium thiosulfate
(according to chlorine dose) to quench the
reaction (instantaneous THM) and sealed
with a Teflon-faced septum that allowed
no headspace. A second THM sample
was stored at room temperature in a
Teflon-sealed vial with no headspace for
24 hr and then quenched with sodium (
-------
thiosulfate (terminal THM). After the
sodium thiosulfate addition, the THM
samples were all stored at 4°C until
analysis. All vials were warmed to room
temperature before THM analysis. THM
samples were measured using the gas-
sparging technique developed by Bellar
and Lichtenberg (Bellar, T. A., and J. J.
Lichtenberg; 1974; Determining Volatile
Organics at Microgram-per-Litre Levels
by Gas Chromatography; J. Amer. Water
Works Assoc., 66(12J:739) and all anal-
yses were performed on a Varian 3700
gas chromatograph equipped with a
flame ionization detector system and a
reporting integrator.
A microprocedure for residual chlorine
determinations (free and total) was de-
veloped because of the sample size
available for analysis. The leuco crystal
violet technique for chlorine residual was
modified to use a 5-mL sample (a total of
10 ml for both free and total chlorine) in
both the presence and absence of the
Bold basal medium. The reference tech-
nique was amperometric titration. The
absorbance of 592 nm of the leuco crystal
violet color development was monitored.
Calibration curves were constructed for
free and total available chlorine residual
determinations in both chlorine-demand-
free water and in the Bold basal medium.
Chlorine-demand-free water was used in
the resuspension of the algal cells for
THM precursor studies, and the Bold
basal medium was used for the extra-
cellular product THM studies. The free
and total chlorine residuals were meas-
ured after a 1-hr chlorine contact time for
both the water and the medium calibra-
tion curves. The calibration curves ob-
tained for the chlorine-demand-free water
had correlation coefficients (R2) of 0.995
and 0.991, respectively.
Results
Increasing cell count was directly re-
lated to the amount of chlorophyll-a
produced by the cells. Linear regression
analyses for each algal species were
obtained by averaging each day's chloro-
phyll-a replicate measurements and com-
paring these with the average of that
day's cell count replicate measurements.
These analyses yielded correlation co-
efficients (R2) for chlorophyll-a to cell
count of 0.952 for Anabaena, 0.955 for
Scenedesmus. and 0.989 forPediastrum.
The three species exhibited different
growth curve shapes and growth rates.
After chlorine was added to the cell and
ECP test samples of Anabaena, THM
production was monitored at contact
times of 1 and 24 hr (Table 1). The pH for
Table 1. Comparison of THM Production from Three Species (ECPs and Cells)
Day in Growth
Cycle (days)
7*
21
Contact Chlorine
Time (hr) Dose
1
1
24
24
1
1
24
24
Low
High
Low
Medium
Medium
High
Low
Medium
Anabaena
Cells ECPs
315.6
392.8
408.6
400.3
138.0
392.8
308.3
549.2
29.6
45.4
64.6
258.9
73.1
149.5
57.2
497.0+
THM, ug/L
Pediastrum Scenedesmus
Cells ECPs Cells ECPs
252.1
424.0
97.3
343.4
481.8
31.9
119.6
59.7
944.6 167.4
466. 1 <0. 1
120.8 333. 1
124.7+ 359.7
157.1 <0.1
7. 1 <0. 1
285.6 <0.1
333.3 <0. 1
397.9
37.8
321.5
782.8
47.3
13.9
553.0
539.7
*Eight days for Scenedesmus.
+Onlyhigh chlorine dose data available.
all measured data points ranged between
6.5 and 7.4 for the ECPs and 6.4 and 7.6
for the cells. The temperature ranged
between 21 ° and 24°C for all Anabaena
measurements. Typically, TTHM produc-
tion increased slightly and then fell as the
stationary phase of growth was reached.
The free chlorine residual was low during
the early portions of the growth curve,
indicating a high chlorine demand. After
6 days, however, a free residual persisted,
indicating a change in the type and/or
degree of ECP produced. THM production
increased with an ill-defined maximum
near the late exponential growth phase,
followed by a slight decrease in amount of
THM. This type of pattern was also
observed for the other chlorine doses.
The drop off in THM level was not due to
chlorine limitations alone, since a free
residual (0.4 mg/L) persisted for the
32.5-mg/L dose and a similar decrease
was noted. Results of chlorinating the
isolated Anabaena cells paralleled those
observed for the ECPs. A free chlorine
residual was maintained throughout the
growth cycle except in 1(3? case of the
lowest dose (at days in the growth cycle
greater than 10).
The Scenedesmus cells exhibited a pH
range of 6.2 to 7.4, and the pH of the ECPs
ranged from 6.4 to 7.0. The temperature
for all Scenedesmus measurements
ranged from 24° to 25°C. The Scene-
desmus cells after both 1 and 24 hr of
chlorine contact time showed high chloro-
form production immediately after initia-
tion of the experiment. After the initially
high levels, the chloroform levels dropped
until approximately day 4 of growth, at
which time a peak in chloroform produc-
tion was observed (days 4 to 10). After
this peak, the chloroform levels dropped
to less than the detection limits toward
the last day examined along the growth
curve (day 21). A comparison of 7- to 8-
day versus 21 -day THMFP values appears
in Table 2. The Scenedesmus ECPs at a
1-hr contact time produced chloroform
levels in generally the same pattern as
that described for the cells. For the
Scenedesmus ECPs after 24 hr of chlorine
contact time, maximum chloroform produc-
tion occurred somewhat later in the
growth curve (approximately days 11 to
21) than that observed for the cells of the
ECPs at the 1 -hr contact time.
The Pediastrum cells had a pH range of
6.5 to 7, and the pH of the ECPs ranged
from 6.5 to 7.9. For the Pediastrum cells
and ECPs, the temperature ranged be-
tween 23° to 25°C. After both 1 - and 24-
hr contact times, the Pediastrum cells
showed two chloroform production peaks.
The chloroform levels were initially low
during the beginning of the growth curve.
Maximum chloroform production was
reached between days 7 and 10 of
growth. After day 10, the chloroform
levels dropped and then began to increase
again until the second chloroform maxi-
mum was reached between days 15 and
21. After the second peaks in chloroform
production, the amount of chloroform
dropped to low levels toward the latter
part of the growth curve. Chloroform
production from the Pediastrum ECPs for
both chlorine contact times followed the
same general pattern as that of the
Pediastrum cells.
The chlorine doses employed for the
Scenedesmus ECPs and cells were com-
parable and generally sufficient to main-
tain a residual at 1 hr of contact time.
With both Scenedesmus cells and ECPs,
a free chlorine residual was present for
all but the lowest chlorine doses after 1 hr
of chlorine contact time, even during the
latter portions of the growth curve. Thus,
the decline in chloroform levels toward
-------
the end of the growth curves for the cells
and the ECPs was not due to chlorine
limitations. This result is general for all
three species.
Discussion
The THM levels produced with the
chlorination of algae depended on the
species (Anabaena. Pediastrum, orScene-
desmus). the age of the culture, the
chlorine dose, and the substance chlori-
nated cells versus ECPs). THM production
was quite variable. Nevertheless, the
results can be presented to facilitate
comparison by calculating the THM for-
mation potential (THMFP) of cells and
ECPs per million algal cells in suspension.
The THMFP (//g/106 cells) for the cells
and ECPs of each of the algal species is
listed in Table 2 for two time points in the
growth cycle. The chlorine doses chosen
for representation were 23.7 mg/L for
the ECPs and 5.0 mg/L for the cells. The
range of values depends on the algal
species and the position of the algae in its
growth cycle, with greater THM formation
occurring in the exponential growth
phase; but the values do not vary greatly
between chlorine contact times of 1 and
24 hr. A bloom of these three species
would produce 0.2 to 120 prg/L THM per
106 cells. The ECPs produced by 108cells
of the species studied would produce an
additional 0.1 to 50 tig of THM. These
values are approximate and vary with
growth condition and chlorine dose, but
algal sources can clearly contribute sig-
nificantly to the overall THM production
in a treatment facility.
The amount of algae-related THM
precursor in water can be influenced by
factors other than the number of algae
present in the water. In an undisturbed
natural water, both cells and ECPs would
be present and would contribute to THM
formation. In a water treated with algacide
to control algal blooms, some cells would
be present along with ECPs and a quantity
of disintegrating cell material. At a water
treatment plant that practiced coagulation
and filtration for algae removal and then
chlorinated the filtrate, the cells would be
removed from the water before chlorina-
tion, but the ECPs would probably pass
through the plant and react with chlorine
to produce THMs.
The higher chlorine doses and the
longer chlorine contact time (24 hr) for
the cells and ECPs of both Pediastrum
and Scenedesmus produced lower chloroform
levels than expected (i.e., compared with
the chloroform levels produced by the
lower chlorine doses and by the 1-hr
contact time). The reason for this obser-
vation has not been established, but
possibly organohalide compounds other
than haloforms were formed, given that
these species had higher chlorine doses
and a longer chlorine contact time, and
may have different precursor compounds
from Anabaena. These factors imply
different mechanisms of action.
The concentrations of bromodichloro-
methane, dibromochloromethane, and
bromoform for the Pediastrum and
Scenedesmus cells and ECPs were gener-
ally below the detection limits, except for
a few isolated instances in which the
bromine-containing volatile concentra-
tions were disparately high when com-
pared with the chloroform levels. These
results were probably caused by contami-
nants rather than by the compounds of
interest. Thus for Scenedesmus and
Pediastrum, only the chloroform levels
were used in the data analysis.
The technique used for the Anabaena
THM analyses gave no bromoform results,
and the dibromochloromethane measure-
ments were questionable. For this reason,
the total THM data given here for
Anabaena actually reflect the concentra-
tions of chloroform and bromodichloro-
methane only.
The results presented demonstrate
that high THM concentrations are pro-
duced from both algal biomass and metab-
olites. The significance of these results
suggests that THM may be partially
reduced by observing and controlling the
natural phytoplankton communities in
the water source for domestic water
Table 2. THMFP for the Three Algal Species (ECPs and Cells) with a Medium Chlorine Dose
Time in Growth Contact
Cycle (days) Time (hr)
7* 24
21 24
21 1
Anabaena
Cells ECPs
3.6
0.8
0.2
2.3
0.7
0.1
THMFP
I tig /1W Cells)
Pediastrum
Cells ECPs
122.6
5.7
45.7
44.5
31.6
14.9
Scenedesmus
Cells ECPs
21.6
<.001
<.001
47.1
6.6
0.6
'Eight days for Scenedesmus.
supplies. Clearly, humic or fulvic pre-
cursors or both contribute to THM produc-
tion upon chlorination; but this source
can only account for a portion of the
THMs produced. To date, no complete
mass balance has been possible to deter-
mine all the precursor molecules that
produce THMs. This problem is further
complicated by the incomplete yields of
THMs that model compounds produce. A
significant amount of algal biomass is
unlikely to survive through the coagula-
tion and filtration steps of a well-operated
water treatment plant; however, ECPs
may persist if they are not destroyed by
bacteria or removed in the coagulation
process. Furthermore, algae can pass
through the coagulation and filtration
steps if they are present in large numbers
or if the treatment plant is not being
optimally managed. THM precursors
(ECPs and cells as well as humic and
fulvic acids) will be present at the point of
prechlorination (chlorination before co-
agulation), a process used at many sur-
face water treatment plants. For this
reason, the presence of such materials
may allow the THM levels to be greater
than those formed if coagulation and
filtration are performed before chlorina-
tion.
Conclusions
1. Significant THM levels (10 to 950
/ug/L) were produced from both the
resuspended cells and the ECPs at
various stages of the growth curves
of Anabaena, Pediastrum, and Scene-
desmus. The Anabaena cells provided
as many or more THMs as did the
Anabaena ECPs. The Scenedesmus
ECPs yielded more chloroform than
the corresponding algal biomass. For
Pediastrum, the chloroform provided
by the ECPs upon chlorination was
comparable to the chloroform levels
provided by the Pediastrum eel Is. The
THM concentrations from all three
algal species were comparable with
THM yields obtained in other studies
evaluating humic and fulvic acids.
2. THM production from the Anabaena
ECPs increased as the stationary
growth phase was approached. The
THM production from the Anabaena
cells also followed this trend.
3. Chloroform production from the
Scenedesmus cells and ECPs was
initially somewhat high, then de-
clined until the late exponential or
early stationary growth phase, at
-------
which point maximum chloroform
production occurred.
4. The Pediastrum cells and ECPs pro-
vided two chloroform production
peaks during the growth period ex-
amined. Both chloroform production
peaks occurred during the period of
growth considered to be the late
exponential or early stationary growth
phase.
5. THM production by the chlorination
olAnabaena, Pediastrum, and Scene-
desmus cells and ECPs was related
to the total organic carbon (TOC) pres-
ent, the cell count (growth phase),
and the chlorine residual after 1 hr of
chlorine contact time.
6. The ECP chlorine residuals of all
three organisms increased with cul-
ture age, indicating a change in type
and/or degree of ECPs produced. An
alternative explanation is that a
chlorine-demanding component of
the medium may have been assim-
ilated by the cells.
7. Because cell count is related to THM
production and to chlorophyll-a con-
centration along the growth curve, a
relationship between chlorophyll-a
and THM production for the algal
cells is inferred. The ECPs do not
have significant levels of chlorophyll-
a, indicating that because THMs are
produced from both the cells and the
ECPs, more than one THM precursor
must be involved.
arbitrarily be abandoned. Monitoring
algal conditions can help minimize
THMs and simultaneously help maxi-
mize finished water quality through
the proper choice of chlorination
conditions.
Kathryn F. Briley, Robert F. Williams, and Charles A. Sorber are with the
University of Texas, San Antonio, TX 78285.
Gary S. Logsdon is the EPA Project Officer (see below).
The complete report, entitled "Alternative Water Disinfection Schemes for
Reduced Trihalomethane Formation: Volume II. Algae as Precursors for Tri-
halomethanes in Chlorinated Drinking Water," (Order No. PB84-129 006; Cost:
$11.50, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati. OH 45268
Recommendations
1. Since this study demonstrated that
both algal biomass and metabolites
from algae produce high THM con-
centrations, THM should be reduced
by proper choice of the position of
chlorination in a water treatment
plant. Coagulation and filtration of
surface waters should precede chlori-
nation to reduce THM levels in fin-
ished waters. Such treatment before
chlorination also reduces other THM
precursors (organic and humic) that
may be present.
2. Algal blooming should be monitored
and controlled so that appropriate
water treatment procedures can be
adopted for the source water condi-
tions. Prechlorination is an advan-
tageous procedure at many water
treatment plants and should not
-------
United States
Environmental Protection
Agency
Official Business
Penalty for Private Use $300
Center for Environmental Research
Information
Cincinnati OH 45260
;;--, -~Ti
;',,', ,«
IL 60601
i, U.S. GOVERNMENT PRINTING OFFICE: 1984-759-102/844
------- |