|
Chloramines have been used in the U.S. for drinking water disinfection for over 70 years.
Chloramine use was prevalent in the 1930s (Spaulding 1929, Braidech 1931, Lyles 1931,
Goehring 1931, and Ely 1933). A nationwide survey conducted in 1938 showed that 16 percent
of municipal utilities used chloramines at some point in their treatment process (AWWA 1941).
Chloramine use declined during WWII because of the inability to obtain ammonia (White 1986),
and by 1959, the number of utilities using chloramines had decreased to 6 percent (Baker 1959).
By the early 1960s, chloramine use had dropped to less than 3 percent (Durfor and Becker
1962).
In the early 1970s, researchers discovered that halogenated organics were formed during
chlorination (Rook 1974, Bellar and Lichtenberg 1974). Subsequent toxicological studies
classified some of these disinfection byproducts (DBPs) as possible human carcinogens (Kimura
Ebert and Dodge 1971, Page, Harris, and Epstein 1976, Kuzma, Kuzma, and Buncher 1977).
There has been resurgence in interest in chloramination over the last 30 years as utilities seek a
means to provide secondary disinfection while limiting disinfection byproduct (DBP) formation.
In 1979, EPA published the first regulations limiting DBPs in the National Interim Primary
Drinking Water Regulations for Control of Trihalomethanes. A maximum contaminant level
(MCL) for total trihalomethanes (TTHMs) was set at 100 ΅g/L. As researchers began to
evaluate alternative disinfectants to limit DBP formation, it was discovered that chloramines
resulted in significantly less TTHM formation than free chlorine (Brodtman and Russo 1979 and
Norman, Harms, and Looyenga 1980).
Laboratory results, in which chloramines required much higher concentrations and longer
contact time to achieve microbiological inaction levels comparable to chlorine, led EPA to
impose a ban on use of chloramines (National Interim Primary Drinking Water Regulations,
USEPA 1978). Field results, however, indicated a much higher chloramine biocidal efficiency
than predicted by laboratory research (Shull 1981, Brodtman and Russo 1979, and Mitcham,
Shelley, and Wheadon 1983). These successful applications of chloramination in field tests
convinced EPA to rescind the ban a year later (USEPA 1979). The discrepancy between lab and
field results for chloramine biocidal efficiency has been theorized to result from the relative
resistance of in situ and laboratory grown organisms, the method of chloramine application (lab
tests were typically conducted with preformed chloramines), and the criteria for evaluating the
effectiveness of a disinfectant.
The Disinfection Systems Committee of AWWA conducted a survey of disinfection
practices of large and medium sized utilities (serving more than 10,000 people) in 1998 (AWWA
Disinfection Systems Committee 2000). This survey extended the database collected in similar
surveys conducted in 1979 (Hoehn and Johnson 1983) and 1989 (Bishop 1992). The 1998
survey responses showed that while chlorine was still the most widely used disinfectant,
chloramine use increased from 20% in 1989 to 29.4% in 1998. In addition, 50 percent of the
respondents indicated that they were considering changing their disinfection practices. The most
frequent reason stated for contemplating change was to meet D/DBP regulations. The most
frequent change under consideration was use of alternative disinfectants.
Currently Stage 1 of the D/DBP rule sets the MCL for TTHMs at 80 ΅g/L and for HAA5
at 60 ΅g/L based on a systemwide running annual average. In the proposed Stage 2, MCLs will
remain at 80 ΅g/L for TTHMs and at 60 ΅g/L for HAA5, but utilities must meet these targets
calculated as locational running averages.
CHLORAMINATION CHEMISTRY
Inorganic chloramines are produced for water treatment disinfection in reactions between
ammonia (dosed and/or naturally occurring) and hypochlorous acid. When chlorine is added to
water, organic chloramines are also produced in reactions between organic nitrogen compounds
and chlorine. Formation of organic chloramines is undesirable because they have no biocidal
potency (Feng 1966). Nitrogen organics also reduce the formation of inorganic chloramines by
binding to chlorine or by the transfer of chlorine from inorganic chloramines to nitrogen organics
such as amino acids (Margerum, Gray, and Huffman 1978). Another problem is that organic
chloramines may react as free chlorine or ammoniacal chloramines in several of the commonly
used tests to monitor these residuals. Therefore, the true nature of the disinfectant residual can
be misinterpreted.
Ammonia is a highly soluble weak base. In reaction with water, ammonia forms
ammonium and hydroxyl ions. The deprotonation of ammonium is shown in the following
acid/base reaction:
NH4+ = NH3 + H+ (pKa = 9.3 at 25 °C)
When elemental chlorine gas is added to water, the chlorine is rapidly hydrolyzed to form
hypochlorous acid and chloride ion in the following reaction:
Cl2 + H2O = HOCl + H+ + Cl- (Keq = 5 x 104 M-1 at 25 °C)
Except for conditions of very low pH or high chloride concentration, the reaction proceeds far to
the right and only a very small percentage of elemental chlorine remains.
When chlorine is added to water as liquid sodium hypochlorite, the following reaction
occurs:
NaOCl ? Na+ + OCl-
When added as hypochlorite granules, hypochlorite is formed in the following reaction:
Ca(OCl)2 ? Ca2+ + 2OCl-
Regardless of the form of in which it is added, chlorine speciation is controlled by pH,
chlorine concentration and temperature according to the following acid/base reaction:
HOCl = H+ + OCl- (pKa = 7.5 at 25 °C)
Inorganic chloramines are formed when free chlorine reacts with ammonia in the
following series of reversible stepwise reactions (Morris and Isaac1983):
HOCl + NH3 = NH2Cl + H20 (monochloramine)
HOCl + NH2Cl = NHCl2 + H2O (dichloramine)
NH2Cl + NH3 = NCl3 + H20 (trichloramine)
The relative amounts of the three chloramine species initially formed is dependent on pH,
the Cl2:N ratio, and temperature. Monochloramine is the desired secondary disinfectant species
because of its greater stability and biocidal potency and because dichloramine and trichloramine
have much lower odor thresholds. The rate of formation of monochloramine in the above
equation is first order with respect to hypochlorite and ammonia and is highly pH dependent. The
rate expression was reported by Morris and Isaac (1983) as follows:
r (mol / L-s) = 6.6 x 108 e(-1510/T) [HOCl] [NH3]
From the above equilibrium equation, it can be seen that the rate of monochloramine formation is
maximized where the product of HOCl and NH3 is greatest. The maximum NH3 and HOCl
product occurs at pH 8.4, midway between the respective pKa values of 9.3 and 7.5. At pH 8.4,
the formation of monochloramine is complete in seconds to 1 minute. At higher or lower pHs the
reaction is slower. Between pH 7 and 9, and at Cl2:N weight ratios less than 5:1 (1:1 molar ratio),
monochloramine is the predominant species.
As the Cl2:N ratio increases or pH decreases, dichloramine formation occurs either in
reactions between HOCl and NH2Cl above or in the following pH dependent acid catalysis
reaction in which ammonia is released:
2 NH2Cl + H+ = NHCl2 + NH3 + H+
The subsequent decomposition of dichloramine is primarily responsible for loss of
residual. The rate of the acid catalysis reaction is increased with increasing proton donors and as
pH decreases. It was reported by Valentine and Jafvert (1988) that carbonic acid and bicarbonate
may catalyze to this reaction and therefore the decomposition rate of monchloramine and
dichloramine may increase as alkalinity increases. Dichloramine formation is maximized in the
pH range of 4.4 to 6 and at Cl2:N ratios of 5:1 to 7.6:1. Under normal water treatment pH
conditions, dichloramine formation is several orders of magnitude slower than that for
monochloramine.
Breakpoint chlorination occurs when all of the ammonia present has been oxidized by
chlorine. The theoretical breakpoint Cl2:N weight ratio of 7.6:1, or molar ratio of 1.5:1, comes
from the following reaction:
2 NH3 + 3 HOCl = N2 + 3 H+ + 3 Cl- + 3 H2O
Addition of chlorine beyond breakpoint results in a stable free chlorine residual. In
practice, the Cl2:N ratio required for breakpoint chlorination is typically higher than 7.6:1 due to
other chlorine demands. It is thought that trace levels of trichloramines can be present at Cl2:N
ratios beyond breakpoint chlorination.
MICROBIOLOGICAL WATER QUALITY
One of the most important functions of a disinfectant residual is the protection of
biological water quality in the distribution system. The Surface Water Treatment Rule states that
all systems must maintain a minimum disinfectant residual of 0.2 mg/L leaving the plant and a
detectable residual in at least 95 percent of measurements throughout the distribution system.
The Total Coliform Rule requires testing for total coliforms with subsequent testing for E.coli for
positive samples. A system is in violation if more than 5 percent of samples are positive for total
coliforms or if fecal or E.coli is detected at a site with a second positive total coliform.
Inactivation of microorganisms and prevention of regrowth in distribution systems are
affected by several factors including attachment of microorganisms to pipe surfaces to form a
biofilm, penetration of the disinfectant into the biofilm, biodegradable carbon concentration, and
corrosion of pipe surfaces. Recent research has suggested that chloramines are at least as
effective as free chlorine in protecting microbiological water quality in the distribution system.
Several researchers have concluded that free chlorine can be ineffective in controlling
bacteria in distribution systems (Tracy, Camarena, and Wing 1966, Reilly and Kippin 1983,
Wierenga 1985, LeChevallier, Babcock, and Lee 1987). Conversely, several studies have
reported improvements in coliform and HPC bacteria concentrations after utilities switch from
free chlorine to chloramine secondary disinfection. When Huron, South Dakota switched from
free chlorine to chloramines, HPC counts decreased 52 percent in the first month (Norman,
Harms, and Looyega 1980).
Norton and LeChevallier (1997) monitored the water quality from two systems for one
year after the systems switched from chlorination to chloramination postdisinfection. The two
systems, Indiana American Water Company and Virginia American Water Company both
switched to chloramines in 1993 because of problems controlling coliform bacteria.
In the Indiana American Water Company system, average HPCs for the two years before
conversion were 412 cfu/mL in 1991 and 428 cfu/mL in 1992. After conversion, the yearly
average HPC decreased to 31 cfu/mL. In the two years prior to conversion, 0.60 percent of 4046
samples were positive for total coliforms. After conversion to chloramines, 0.16 percent of 1868
samples were positive, almost a fourfold decrease. An increase in residual concentrations at ends
of the system was also observed. In 1991 and 1992, 22 to 24 percent of the sample sites had
average free chlorine residuals less than 1.0 mg/L as Cl2. In 1993, 5.5 percent of the sites had
average total chlorine residuals less than 1.0 mg/L as Cl2.
Similar improvements in microbiological water quality were observed at the Virginia
American Company. Using the less sensitive SPC agar (R2A agar was used in the Indiana
measurements), HPCs averaged 7.4 cfu/mL from 1991 to 1992 and 2.6 cfu/mL in 1993 after
conversion to chloramines. Prior to conversion, 22 percent of the samples were positive for
coliforms. With chloramines, only 2 samples were positive for coliforms and these were both
collected in the first week after conversion.
Volk and LeChevallier (2000) analyzed water quality data from 91 systems and an
association between coliform counts and temperature, AOC, and type and concentration of
residual. Coliform occurrences were noted when the following parameter thresholds were
exceeded simultaneously: temperature > 15 °C, AOC > 100 ΅g/L and residuals below 0.5 mg/L
free chlorine or 1.0 mg/L chloramines.
LeChevallier and colleagues (1996) performed an 18 month study of 31 systems for
factors that limit microbial growth in distribution systems. Coliforms were detected in 0.97
percent of samples with free chlorine and 0.51 percent of samples with chloramine residuals.
DiGiano and coworkers (2000) performed a 12 month study to compare bacterial
regrowth in a chloraminated system to that in a chlorinated system in an adjacent city. The
systems made a good pair for comparison because their source water quality and distribution
system temperature conditions were similar. Based on HPC data, the researchers concluded that
the chloraminated and chlorinated systems provided equivalent control of bacterial regrowth. In
each system, high HPC bacteria concentrations were associated with low disinfectant residual,
and there was a strong correlation between low disinfectant residual and long residence time or
water age.
DISINFECTION BYPRODUCT FORMATION
In research sponsored by AwwaRF, Symons et al. (1998) conducted studies on three
source waters to identify the factors that affect disinfectant byproduct (DBP) formation during
chloramination. Specific research goals were as follows:
Evaluate the chemical and operating factors that influence DBP formation.
Assess the known and unknown DBPs that are formed.
Identify treatment steps that utilities can implement to reduce DBP formation.
The interrelationships of total chlorine residual, chloramine speciation, Cl2:N ratio, pH,
bromide concentration, and variable mixing conditions were examined for their effect of the
formation of 12 monitored DBPs and total dissolved organic halogens (DOX). The 12 DBPs
were the four THMs (chloroform, bromodichloromethane, dibromomochloromethane, and
bromoform), HAA6 (mono-, di-, and tri- haloacetic acid, mono- and di- bromoacetic acid, and
bromochloracetic acid) and the cyanogen halogens (cyanogen chloride and cyanogen bromide).
The study included source waters from the three primary utility participants, City of
Austin, City of Houston, and the Metropolitan Water District of Southern California (MWDSC),
and from an additional five utilities selected to the range of geographical location, water quality,
and operational parameters. The project was divided into four tasks. In Task 1A, bench-scale
experiments were conducted on the three primary water sources to determine effects of variable
water chemistry conditions on DBP formation. Combinations of the following water chemistry
conditions tested:
| | | pH 6, 8, and 10, |
| | | total chlorine residual after 48 hours: 1, 2, and 4 mg/L, and |
| | | Cl2:N ratios of 3/1, 5/1, and 7/1. |
In Task 1B, variable mixing energies and sequential chlorine addition were tested as
follows:
| | | Low (G = 60 sec-1), medium (G = 500 sec-1), and high (G = 1000 sec-1) mixing energies
with simultaneous addition of chlorine and ammonia, and |
| | | Chlorine addition followed by ammonia addition after a 30-second delay at low and
medium mixing energies. |
In Task 2, pilot studies were conducted with the three primary source waters to verify the
findings off the bench-scale tests. In Task 3, five additional source waters were selected for
evaluation to cover a variety of geographical locations, source water qualities, and operational
variables. Full-scale operational data and selected bench-scale tests were conducted on the
secondary study waters. In Task 4, analytical methods for identifying unknown DBPs were
tested.
In Tasks 1A, 1B, 2, and 3, DBP residuals were measured after 48 hours incubation in
simulated distribution system (SDS) tests.
The study produced the following results regarding the factors that affect DBP formation.
| | | There was good agreement among the bench-scale, pilot-scale, and full-scale data across
the range of geographical locations indicating that the results of the study were applicable
to a wide range of source water types. |
| | | DBP formation generally increased with decreasing pH and increasing Cl2:N ratios.
Higher DBP formation at pH 6 is consistent with the theory that dichloramine, or its
decomposition product, free chlorine, is the active halogenating agent. |
| | | DBP formation was sensitive to both pH and Cl2:N ratio. Utilities should test
combinations of these two variables to formulate a chloramination strategy that balances
DBP control with maintenance of adequate residual and prevention of nitrification. DBP
formation was lowest at Cl2:N ratio of 3/1, but free ammonia may be formed at ratios this
low with subsequent nitrification and dissipation of residual. |
| | | Dichloramine was present after the 48 hour SDS hold time only at pH 6, and the relative
percentage of dichloramine increased as Cl2:N ratio increased. |
| | | Disinfectant residual was found to have little effect on DBP formation over the range of
residual concentrations tested (1, 2, and 4 mg/L). |
| | | The predominant species of HAA6 formed were the dihalogen-substituted halogens
suggesting that these species were not well controlled by chloramines. |
| | | Substantial concentrations of DOX were formed in the SDS tests under some conditions.
In Task 1 B, the maximum SDS DOX concentration was 258 ΅g Cl-/L in the Houston
source water at pH 6 and a Cl2:N ratio of 5/1. In Task 1B, the maximum SDS DOX
concentration was 320 ΅g Cl-/L in the Houston water at pH 8, a Cl2:N ratio of 7/1, and
low mixing intensity. In Task 2, the maximum SDS DOX concentration was 203 ΅g Cl-
/L in the Houston water at a SDS incubation pH of 6, a Cl2:N ratio of 3/1, with
prechloramination and conventional coagulation. |
| | | The 12 specific DBPs measured accounted for a relatively small proportion of the total
DOX concentration in all tests. In Task 1A, the 12 DBPs accounted for less than 35
percent of DOX on a molar basis, in Task 1B, less than 32 percent, and in Task 2, less
than 45 percent. Although present and near term regulations only govern TTHMs and
HAA5, utilities may elect to consider DOX formation when developing a chloramination
strategy. |
| | | Bromide addition increased the formation of bromine substituted THMs, HAA6, and
CNX. The increased formation of DOX with bromide addition indicated increased
degree of halogen substitution during chloramination of bromide ion containing waters.
Control of DBP formation was complicated by the presence of bromide and the
complexity of bromamine chemistry. |
| | | Mixing was not as important as chemistry in controlling DBP formation. THM formation
was higher at low mixing energy than at medium or high energies. THM formation was
also higher for delayed ammonia addition. No correlation between DOX formation and
mixing energy was observed, but DOX formation was increased by delayed ammonia
addition. HAA6 and CNX formation were not affected by variation in mixing energy. It
was concluded that the kinetics for THM formation might be more rapid than the kinetics
for DOX formation in general. |
A nonhalogenated DBP, N-nitrosodimethylamine (NDMA) has been identified as a
possible formation product of chloramination. NDMA is a semi-volatile organic that has been
manufactured and used commercially for decades. NDMA has been used as a soil additive to
prevent nitrification, a plasticizer for rubber and polymers, and as a solvent in the fiber and
plastics industry. NDMA has also been detected in a number of food products including, cheese,
canned fruit, various meat products, fish and fish products, and alcoholic beverages including
beer. Tobacco smoke contains high levels of NDMA. NDMA is a known carcinogen, and EPA
has established a 10-6 cancer risk level for NDMA of 0.7 ng/L. No MCL or MCLG has been set
yet by EPA for NDMA in drinking water, and the levels of exposure to NDMA as a contaminant
in air and food far exceed the exposure level in drinking water.
Following discovery of NDMA in a drinking water well in 1998, California Department
of Health Services (CDHS) set an action level of 2 ng/L. The action level was subsequently
raised to its current level of 10 ng/L because few laboratories were capable of quantifying
NDMA at the 2 ng/L level (CDHS 2002).
In bench tests simulating drinking water conditions, Eaton and Briggs (2000) found that
NDMA formation was not increased by chlorine dose and was not dependent on pH.
Chloramines were not tested. Najm and Trussell (2001) reported a direct correlation between
chloramine dose and NDMA formation for some waters but not others. NDMA formation was
not influenced by the amount of natural organic matter or organic nitrogen.
Choi and Valentine (2002) conclusively showed that NDMA can be formed by a reaction
of monochloramine with dimethylamine (DMA). Dimethylamine is ubiquitous in drinking and
wastewaters and was believed to be a likely NDMA precursor. A mechanism was hypothesized
in which monochloramine reacts with DMA to form hydrazine that is then subsequently oxidized
to form NDMA. Choi and Valentine (2002) also developed a simple kinetic model to support the
hypothesis that NDMA can be slowly formed in drinking water. Formation from reactions
involving other nitrogenous precursors was also suspected.
NITRIFICATION
Nitrification is a two step process in which autotrophic bacteria mediate the oxidation of
ammonia to nitrate. Ammonia is oxidized to nitrite, and nitrite is oxidized to nitrate in the
following sequential reactions:
NH4+ + 3/2 O2 = NO2- + H2O + 2H+
NO2- + H2O = NO3- + 2H+
Systems that practice chloramination are susceptible to nitrification because a small
amount of ammonia is always present in a system that contains chloramines. It is estimated that
two-thirds of medium and large systems that chloraminate in the U.S. experience some problem
with nitrification and one third of those systems have operational problems caused by
nitrification (Wilczak et al. 1996).
The occurrence of nitrification in a distribution system can have adverse operational,
regulatory, and public health effects for a utility. Since some free ammonia is typically present in
systems that chloraminate, these systems are susceptible to nitrification if the loss of disinfectant
residual permits the growth of ammonia oxidizing bacteria (AOB). Once nitrification begins, it
accelerates the loss of disinfectant residual because the rate of chloramine decomposition
increases as ammonia decreases. In the limit of very low ammonia concentrations, all residual
can be lost in reactions similar to those that occur during breakpoint chlorination As chloramine
is destroyed, ammonia is released providing AOB with substrate. Additional loss of chloramine
may also be attributed to the formation of nitrite as a product of nitrifying organisms and further
fueling the nitrification reaction.
Water quality impacts from nitrification can include loss of disinfectant residual, elevated
heterotrophic bacteria counts, decrease in dissolved oxygen, and decrease in pH. Once
nitrification begins, a utility is usually compelled to institute operational corrective measures
such as a period of breakpoint chlorination or system flushing to stop the process. The loss of
disinfection residual and increase in heterotrophic bacteria counts can affect a utilitys regulatory
compliance. The Surface Water Treatment Rule requires a detectable disinfectant residual at 95
percent of the sampling locations in the distribution system. The Total Coliform Rule states that
no more than 5 percent of monthly samples can be positive for total coliforms. Nitrate and nitrite
pose an acute health risk, particularly to infants, by inducing methemoglobinemia, in which the
oxygen carrying capacity of red blood cells is compromised. Nitrate and nitrite are regulated at
the entrance to the distribution system at MCLs of 10 mg/L and 1 mg/L, as nitrogen,
respectively.
Kirmeyer et al. (1995) prepared an AwwaRF Report to address nitrification occurrence
and control. The purpose of the study was to document the occurrence of nitrification in
chloraminated systems, identify causative factors, and develop methods for controlling
nitrification. Kirmeyer and coworkers concluded that nitrification is influenced by the following
factors:
| | | Nitrification most often occurred at a pH between 7 and 9. |
| | | Nitrification occurred at a wide range of temperatures but most often at temperatures
greater than 15 °C. |
| | | Higher chloramine residual appears to deter nitrification. Once nitrification begins
however, chloramine residual disappears quickly and increasing the residual does not
appear to be effective in stopping nitrification. |
| | | Excess ammonia promotes nitrification. Ammonia concentrations were found to be 4
times higher at Cl2:N ratio of 3/1 than at a Cl2:N of 5/1. |
| | | Organic compounds can exert a chloramine demand resulting in poor bacteria
disinfection. |
| | | Long distribution system residence times promote nitrification. |
| | | Distribution system sediment and biofilms can harbor nitrifying bacteria, reduce
disinfection efficiency, and enhance the likelihood of nitrification. |
| | | Treatment plant processes that remove organic matter reduce the potential for
nitrification. |
The following monitoring program was recommended for detection of nitrification:
| | | Monitor the chemical balance of nitrogen species by measuring free and total ammonia,
nitrite, and nitrate and identify changes that occur between the treatment plant and any
point in the distribution system. A shift in nitrogen speciation from ammonia to nitrite
and nitrate is the primary indicator of the degree of nitrification that is occurring. |
| | | Perform heterotrophic plate counts (HPCs) with R2A agar. HPC can be an effective
early warning indicator for nitrification. R2A agar was recommended over standard plate
agar as being more sensitive. |
| | | Monitor chloramine residual throughout the system regularly. A dramatic decrease in
chloramine residual can indicate the onset of nitrification. |
| | | In some utilities studied, dissolved oxygen (DO) correlated well with nitrification. DO
was recommended as a secondary monitoring parameter only for utilities competent in
conducting the test. |
The authors rated nitrification control measures for short-term and long-term efficiency
on a scale of 1 to 5, with 1 being ineffective and 5 being highly effective.
Breakpoint chlorination:
Episode control 5
Long-term control 2
Breakpoint chlorination was rated the most effective means of controlling nitrification
once an episode has started. During breakpoint chlorination, feed of ammonia, the AOB
substrate, is discontinued and disinfection efficiency is increased. Many of the utilities surveyed
practice breakpoint chlorination once or twice a year for a week to a month to control
nitrification. Breakpoint chlorination was not considered an effective long-term solution because
it provides little sustained mitigation of the factors that cause nitrification.
Increasing Cl2:N ratio:
Episode control 3
Long-term control 4
Reducing the ammonia by increasing the Cl2:N ratio can be an effective control measure,
but the study indicated that nitrification can take place when only small amounts of ammonia are
present. This was rated a better preventive measure than as a measure for episode control. Once
nitrification begins, chloramine is rapidly consumed.
Increasing residual:
Episode control 2
Long-term control 4
Increasing the chloramine residual was rated an effective means for preventing
nitrification but not for stopping nitrification once the process has started. Chloramine residuals
are consumed oxidizing the nitrite before they can inactivate nitrifying bacteria. Maintaining a
chloramine residual of 2 to 3 mg/L seemed to inhibit nitrifying bacteria.
Removing organics at the WTP:
Episode control 2
Long-term control 5
Removing organics to improve distribution system biological stability was rated as a very
effective method for preventing nitrification. This method has little effect once nitrification is in
progress.
Distribtution system cleaning:
Episode control 3
Long-term control 4
Cleaning the distribution system through flushing or pigging can provide effective longterm
control of nitrification. Some utilities found that flushing provided only temporary relief
and chloramine residuals quickly disappeared again after flushing.
Improving system residence time:
Episode control 3
Long-term control 4
The authors concluded that decreasing system residence time can be an effective method
for episode and long-term control when used in conjunction with other control measures such as
breakpoint chlorination. Methods for reducing distribution system residence time include
increasing reservoir turnover, installing recirculation and rechlorination facilities on storage
tanks, designing new reservoirs with inlet and outlet pipes and preventing short-circuiting,
looping dead-end mains, and implementing regular flushing in problem areas.
Lieu, Wolfe, and Means (1993) conducted a study to examine survival and regrowth of
nitrifying bacteria after exposure to three chloramine dosages (1.7, 2.0 and 2.5 mg/L), three
Cl2:N ratios (3/1, 4/1, and 5/1), three temperatures (10, 15, and 25 °C) and two contact times (2
and 8 days). Chloramine was neutralized after the contact time except for a non-neutralized set in
which chloramine residual was allowed to decay naturally. The authors concluded that the most
important factor affecting bacteria regrowth was chloramine dose but that the effect of
chloramine dose was influenced by temperature, Cl2:N ratio, and contact time.
At 10 °C, after 2 days of disinfectant contact time, AOB were reduced by 1.5 log at
doses of 1.7 and 2 mg/L and to nondetect level at 2.5 mg/L at all Cl2:N ratios. After 6 weeks of
incubation time, no nitrite was detected in any samples and < 10 MPN AOB were detected. It
was concluded that microbiocidal efficiency was low at 10 °C, but AOB regrowth was slow.
At 15 °C, contact time and chloramine dose controlled regrowth but Cl2:N ratio had no
effect. At the 1.7 and 2 mg/L doses, nitrite was detected at 2 weeks for the 2 day exposure
samples and at 3 to 5 weeks for the 8 day exposure samples. At a 2.5 mg/L dose, nitrite was
detected at 4 weeks for the 2 day exposure set and was not detected for the samples exposed to
chloramines for 8 days. No nitrite was detected in any of the non-neutralized samples.
At 25 °C, chloramine dose, contact time, and Cl2:N all affected regrowth. At the 1.7
mg/L dose and 2 day exposure, nitrite appeared after 1 week for all Cl2:N ratios. For the 8 day
exposure set however, nitrite appeared after 1 week for the 3:1 Cl2:N ratio but not until 2 weeks
for 4:1 and 5:1 ratios. At the 2 mg/L dose, nitrite was not detected in the 8 day exposure or in the
non-neutralized set at 4:1 and 5:1 Cl2:N ratios. At the 2.5 mg/L dose, nitrite was detected only in
the 2 day exposure sample with the 3:1 Cl2:N ratio.
Disinfection efficiencies were similar at 10 and 15 °C, but regrowth occurred in twice as
many samples at 15 °C. It was concluded that nitrification potential was higher at 15 °C. The
biocidal effectiveness of chloramines at 25 °C increased when Cl2:N ratio increased from 3:1 to
4:1 and 5:1. Consequently, there were fewer samples with regrowth at 25 °C than at 15 °C. It
was concluded that increasing the Cl2:N ratio could provide nitrification control for some
situations. Regrowth was highest in the 25 °C tests confirming the expected optimal temperature
range for regrowth of 25 to 30 °C. Finally it was concluded that when nitrite levels are first
detectable, a removal of greater than 2 log is necessary to prevent regrowth.
McGuire, Lieu, and Pearthree (1999) conducted field and lab tests using chlorite ion to
control nitrifying bacteria. In the lab tests, chlorite dosages as low as 0.05 mg/L achieved 3 to 4
log inactivation of AOB over several hours. In field studies of systems using chlorine dioxide for
primary disinfection, systems with chlorite showed lower concentrations of nitrifying bacteria
than systems with chlorite. The authors suggested that chlorite could be added to the plant
effluent in low enough concentration to avoid exceeding the 0.8 mg/L MCL for chlorite.
Wolfe et al. (1990) conducted a study to determine the factors that affect the occurrence
of AOB in chloraminated distribution systems. The authors concluded that the occurrence of
AOB was seasonally dependent with higher concentrations in the summer months. AOB
correlated highly with temperature and HPC bacteria. No AOB were found in the reservoir when
temperatures were below 16 to 18 °C. AOB were found throughout the distribution system and
were found in higher concentrations in sediment than in pipe wall biofilms. AOB were 13 times
more resistant to chloramines than chlorine.
Regan, Harrington, and Noguera (2002) also found an increase in HPC bacteria at the
onset of nitrification and suggested HPC measurement with R2A agar as an early warning
indicator.
CONVERSION
Over the last 20 years, several utilities have converted from free chlorine to chloramines
to reduce THM formation in their systems. The Metropolitan Water District of Southern
California (MWDSC) converted to chloramines in 1984 and documented their experience in
several publications. These publications provide a clear example of the steps conducted by one
large utility in converting to chloramines as well as presenting some of the problems that arose
and their solutions.
Prior to conversion, MWDSC conducted a study to evaluate the feasibility of switching to
chloramination (Kreft et al. 1985). MWDSC first tried other steps to reduce THM formation.
Preoxidation with potassium permanganate was not successful. Enhanced coagulation did not
provide sufficient THM reduction. Ozone was deemed too expensive and chlorine dioxide was
discounted because of concerns over chlorite and chlorate formation. Pilot testing with
chloramines was successful.
MWDSC conducted a feasibility study to evaluate the following factors:
| | | Chemistry and DBP formation. |
| | | Disinfection capacity. |
| | | Taste and odors. |
| | | Bacterial growth in uncovered reservoirs and effects on swimming pools. |
| | | Maintenance and effectiveness of disinfectant residual. |
| | | Health effects of chloramines. |
| | | Impact on member agencies that use chlorination, i.e. effect of blending. |
| | | Technological feasibility, how to implement. |
| | | Costs. |
MWDSC used jar testing to develop a chlorine breakpoint curve for their water to
determine the range of chlorine to ammonia feed ratio necessary to produce monochloramine.
Monochloramine residual increased to a Cl2:N ratio of about 5:1, and reached a minimum at a
ratio of 10.5:1. Free chlorine residual began to appear at a ratio of 11:1. MWDSC concluded that
the Cl2:N ratio should be maintained at less than 5:1. Some dichloramine was formed prior to
breakpoint and after breakpoint the trichloramine ratio increased.
MWDSC concluded that chloramines would provide adequate disinfection capacity.
MWDSC noted that at their operating pH of 8.1 to 8.2, OCl- was the predominant chlorine
species after chlorination. It was noted that the difference in disinfectant potency reported
between monochloramine and hypochlorite is less than that between monochloramine and
hypochlorous acid. Additionally MWDSC pointed to work by Collins and Selleck (1972) and
Kruse, Snead, and Olivieri (1981) that indicated that total chlorine is as effective as free chlorine
in inactivating bacteria at long contact times. MWDSC also relied on successful secondary
disinfection reported at over 70 utilities practicing chloramination.
To assess taste and odor effects of conversion to chloramination, MWDSC conducted a
side-by-side flavor profile analysis (FPA) of chlorinated and chloraminated water. The FPA
panel concluded that the chloraminated water had a much less offensive odor.
MWDSC conducted educational seminars with member agencies, hospitals, patients and dialysis
patients to explain the potential adverse effects of chloramines on dialysis patients. Information
was distributed to pet stores, zoos, aquariums and the public about protecting fish from
chloramines.
Many of its member agencies bought water from MWDSC to augment their own
chlorinated supplies. MWDSC met with member agencies to discuss blending issues, particularly
the avoidance of loss of disinfection residual due to breakpoint chlorination. MWDSC developed
a computer model to predict the effects of blending (Barrett, Davis, and McGuire 1985). Jar
testing was used to calibrate the model. The model was then used to predict acceptable blends of
free chlorine and chloramines for each of its member agencies to prevent breakpoint
chloramination and odors caused by dichloramine ad trichloramine.
In their cost analysis, MWDSC estimated that conversion to chloramines would cost
much less than GAC treatment, an alternative method of reducing THMs by removing
precursors.
Following conversion in 1984, MWDSC staff summarized the practical aspects of
implementing chloramination (Dennis, Rauscher, and Foust 1991) MWDSC selected aqueous
ammonia as used by Denver and Philadelphia because it is less expensive to store, less hazardous
to transport, and permits use of simple metering pumps. MWDSC initially chose concurrent
addition of chlorine and ammonia to limit THM formation and because disinfection CT was
obtained with the three hour contact time.
MWDSC conducted an extensive planning program prior to actual conversion. An
aqueous ammonia manual was prepared that included feed equipment, piping layouts, analysis
and safety and maintenance issues. A formal request was sent to the state of California to make
the change. A series of 6 workshops was conducted with the California Department of Health
Services (CDHS), member agencies, and local health departments.
An extensive notification program was conducted that coordinated notification of CDHS,
5 county health departments, the Federally funded End-stage Renal Dialysis Coordinating
Council, local nephrologists, and all member agencies. Local pet shops and koi clubs were also
notified. Over 60 newspaper articles, press releases, and radio and television announcements
were made, and a telephone hotline was set up to answer questions.
Two task groups were formed to facilitate implementation. One was created to procure
equipment, monitor construction schedule, test pumping equipment, and ensure the readiness of
the WTPs and reservoir facilities to feed ammonia. The other task group was responsible for
training operators and lab personnel. Ten training seminars were conducted.
Potential problems were uncovered and remedied in the pump-testing program. The
Buna-N rubber stators failed within a week and were replaced with ethylene propylene diene
monomer material. Leakage from the packing box on the pump was detected. The packing boxes
were replaced with mechanical seals.
Immediately after conversion, some dialysis centers reported problems removing
chloramines and a few cases of hemolytic anemia were reported. MWDSC immediately switched
back to chlorination. In cooperation with CDHS and county health departments, interviews were
conducted with dialysis centers and all dialysis centers were inspected. Standards for chloramine
removal were developed and a plan was formulated for each facility. After changes had been
made, inspectors went back to each facility and conducted another inspection. This process took
approximately 6 months at which time chloramination was resumed.
In June of 1985, MWDSC began receiving complaints of fishy odors. Jar tests indicated
that the odor problem could be eliminated with 1 hour of free chlorine contact time.
Prechlorination was instituted with chlorine added 1 hour ahead of ammonia.
Ten weeks after conversion, in 1984, MWDSC observed nitrification in the reservoirs. In
1987 and 1988, nitrification occurred in the distribution system. MWDSC developed a
nitrification control program that included early warning sampling, decreasing distribution
system residence time, increasing the Cl2:N ratio from 3:1 to 5:1, and maintaining a free chlorine
residual for 30 days each year.
PROBLEMS
Conversion to chloramination can help utilities reduce THM formation and increase
residual longevity, but there are also potential problems associated with chloramination that
utilities must avoid. Chloramines pose a risk to dialysis patients in the form of
methemoglobinemia and acute hemolytic anemia. Chloramines can cause methemoglobinemia in
dialysis patients by oxidizing Fe 2+ to Fe 3+, methemoglobin, in the blood (Klein 1986).
Methemoglobin is incapable of carrying oxygen or carbon dioxide. Chloramines can also
denature proteins within the red blood cell forming aggregates (Heinz bodies).
In 1981, there was a reported outbreak of Heinz-body positive hemolytic anemia in a
Sydney, Australia hospital (Caterson et al. 1982). It was found that the activated carbon filter
system was inadequate. After increasing the filter length, the problems disappeared. Kjellstrand
and coworkers (1974) found that excess levels of chloramines led to methemoglobin formation,
Heinz-body formation, and hemolysis in dialysis patients. It was found that chloramines at a
concentration of 0.25 mg/L invariably caused hemolysis in dialysis patients.
When a utility converts to chloramines, aquarium owners must take precautions to protect
their fish. Chloramines are introduced directly to the bloodstream of fish through their gills and
pose some of the same health risks incurred by dialysis patients, including hemolysis. Aquarium
owners must protect their fish against exposure to chloramines and the buildup of ammonia. In
preparation for conversion to chloramination, utilities typically take steps to educate the public
and pet storeowners regarding methods to protect fish.
Chloramines have also been found to degrade elastomers more aggressively than
chlorine. Reiber (1993) tested elastomer coupons in an accelerated life cycle protocol with
chlorine and chloramines. Reiber reported the following:
| | | Chloramines were more aggressive in degrading elastomers than equivalent
concentrations of chlorine, and dichloramine was more aggressive than monochloramine.
It was theorized that chloramines achieved greater penetration into the elastomer material
due to less demand than exerted for chlorine at the elastomer surface. |
| | | Excess ammonia addition or the presence of free ammonia was not responsible for
degradation. Coupons exposed to solutions of ammonia at different pH values showed the
same amount of degradation as dechlorinated tap water. |
| | | Forms of elastomers varied widely in their resistance to degradation. The materials most
susceptible to attack were the natural isoprene and synthetic isoprene derivatives.
Synthetic polymers engineered for chemical resistance performed well in the life cycle
tests. |
| | | A water swell test was found a reliable indicator for material degradation. The tendency
of elastomer material to absorb water was found to correlate strongly to breakdown of
material strength properties such as maximum stress and strain. |
| | | Elastomer degradation was found to be sensitive to temperature and to occur more
rapidly with increased temperature. |
Source: A Guide for the Implementation and Use of Chloramines by Harms and Owen. © 2004 AwwaRF. Reproduced with permission. This material is presented solely for informational purposes. More details are available at http://www.awwarf.org/research/TopicsAndProjects/projectSnapshot.aspx?pn=2847.
|
|
