United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
EPA-6 00/2-78-165
September 1978
Research and Development
Breakpoint
Chlorination/ Activated
Carbon Treatment:
Effect on Volatile
Halogenated Organics
-------
EPA-600/2-78-165
September 1978
BREAKPOINT CHLORINATION/ACTIVATED CARBON TREATMENT:
EFFECT ON VOLATILE HALOGENATED ORGANICS
by
James J. Westrick
Michael D. Cummins
Jesse M. Cohen
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publi-
cation. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
11
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled
land are tragic testimonies to the deterioration of our natural environment.
The complexity of that environment and the interplay of its components
require a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem solu-
tion; and it involves defining the problem, measuring its impact, and search-
ing for solutions. The Municipal Environmental Research Laboratory develops
new and improved technology and systems to prevent, treat, and manage waste-
water and solid and hazardous waste pollutant discharges from municipal and
community sources, to preserve and treat public drinking water supplies,
and to minimize the adverse economic, social, health, and aesthetic effects
of pollution. This publication is one of the products of that research, a
vital communications link between the researcher and the user community.
This report summarizes the results of a pilot plant study investigating
the formation and removal of six volatile, halogenated organic compounds in
a tertiary breakpoint chlorination/activated carbon system.
Francis T. Mayo
Director
Municipal Environmental Research
Laboratory
111
-------
ABSTRACT
A continuous-flow pilot plant project was conducted with two major
objectives: (1) to monitor the production of six volatile halogenated
organic compounds during breakpoint chlorination of wastewater for ammonia
removal and (2) to determine the efficiency of removal of these compounds
by activated carbon.
The chlorination/activated carbon systems were operated on lime-
clarified and filtered trickling filter effluent. One chlorination system
was located before a two-stage carbon adsorber, and a second chlorination
system was between stages of a second two-stage adsorber. Breakpoint
chlorination increased the chloroform concentration from a median value
of 11 yg/1 in the tertiary filter effluent to a median value of 61 yg/1.
Chloroform production in that system was about four times greater than
that of the chlorination system located after one stage of carbon. Bromodi-
chloromethane was formed in amounts usually less than 10 yg/1 by chlorin-
ating the filter effluent. Chlorination of the first stage carbon effluent
produced lower concentrations of bromodichloromethane during the first
month of operation, but these concentrations were later equal to or greater
than those produced by chlorinating the filter effluent. Dibromochloro-
methane, bromoform, carbon tetrachloride and 1,2-dichloroethane were formed
at very low concentrations, if at all.
A 20-min empty bed contact time adsorber was much more efficient for
chloroform removal than a 10-min contact time adsorber, probably because
of competitive effects of more strongly adsorbed organics. The removal of
chloroform by the second stage carbon columns conformed to an equilibrium
expression of the Freundlich type.
This report covers the period of operation from June 3, 1976, to
August 9, 1976, and work was completed as of January 13, 1978.
IV
-------
CONTENTS
Foreword „ iii
Abstract . iv
Figures vi
Tables '• vii
Acknowledgment „ viii
1. Introduction 1
2. Conclusions . 2
3. Recommendations 4
4. Background . 5
Ammonia in Wastewater ......... 5
Ammonia Removal 5
Breakpoint Chlorination for Nitrogen Removal ....... 6
The Chlorinated Organics Problem 9
Study Objectives 11
5. Description of Facilities ... ........ 12
Process Equipment .......... . . 12
Sampling Systems 17
System Operation ....... ..... 19
6. Analytical Methods 24
Laboratory Analyses 24
Pilot Plant Analytical Methods 25
7. Results and Discussion . 27
General Process Performance 27
Formation of Volatile Halogenated Organics ..... 35
Breakpoint Chlorination of Granular Media .
Filter Effluent ...... 35
Breakpoint Chlorination of First-Stage
Carbon Effluent ............. 39
Batch Chlorination Tests 40
Granular Activated Carbon Performance . 44
Removal of Volatile Halogenated Organics ..... 44
Removal of Extractable Organic Chlorine ..... 55
References ... ....... 57
Appendix ............................. 61
-------
FIGURES
Number Page
1 Breakpoint curve ... 7
2 Flow diagram, tertiary lime clarification-filtration ...... 12
3 Sludge blanket level detector ..... 14
4 Carbon/chlorination system ................... 16
5 Breakpoint chlorination system ..... . . 16
6 Dipper sampler 18
7 Diverter sampler in the sampling position ............ 18
8 Automatic composite sampler for volatile halogenated organics , . 20
9 VHO sample programmer . .............. 21
10 VHO sampler 22
11 Ammonia removal - system 2 .............. 30
12 Ammonia removal - system 3 ................... 31
13 Residual chlorine after breakpoint chlorination .... 32
14 SOC removal by granular activated carbon ............ 34
J5 CHC13 and BrCHCl2 formed during chlorination of first stage
carbon effluent 41
16 Difference in formation of CHClv in system 2 and system 3 .... 42
17 Difference in formation of BrCHCl2 in system 2 and system 3 ... 43
18 Short-term batch chlorination test results 46
19 CHC13 in effluents from GAG system 1 . 49
20 CHC13 breakthrough, column 3A ............. 49
21 CHC13 breakthrough, column 3B . . . ....... 50
22 CHC13 breakthrough, column 2B . . 50
23 CHC13 breakthrough, column IB ... .......... 51
24 CHClg breakthrough curves . 51
25 CHC13 breakthrough as a function of applied CHC13 . „ 52
26 Capacity of 10-minute contactors for CHC13 removal ....... 54
VI
-------
TABLES
Number Page
1 Precision and Accurary of Volatile Halogenated Organics Analysis.. 26
2 Performance of Lime Clarification-Filtration of Secondary Effluent.28
3 Average Operating Conditions of Breakpoint Chlorination Systems ...28
4 Nitrogen Removal by Breakpoint Chlorination 29
5 Final Effluent Characteristics 33
6 Concentration of Halogenated Organics in Filter Effluent 36
7 Concentration of Halogenated Organics after Breakpoint
Chlorination of Filter Effluent 36
8 Formation of Halogenated Organics During Breakpoint
Chlorination of Filter Effluent 38
9 Formation of Solvent-Extractable Organic Chlorine Compounds
During Breakpoint Chlorination of Filter Effluents 38
10 Formation of Halogenated Organics during Breakpoint Chlorination
of First-Stage Carbon Effluent 39
11 Formation of Solvent-Extractable Organic Chlorine Compounds
During Breakpoint Chlorination of First-Stage Carbon Effluent .... 40
12 Batch Chlorination Results . 45
13 Carbon Column Designations 48
14 Capacity of Carbon for CHCl^ Removal 53
15 Extractable Organic Chlorine 56
VII
-------
ACKNOWLEDGMENTS
The authors gratefully acknowledge the work of Patricia Kiraly,
Arthur Turner and Michael Jelus who participated in the assembly and/or
operation of the system. The efforts of E. Marco Aieta, Joseph Carvitti
and Dirk Krouskop in pilot plant operation, data handling and special
projects were instrumental to the success of this project. The contri-
butions of the staff of the Waste Identification and Analysis Section,
WRD, of the Municipal Environmental Research Laboratory, especially
B. M. Austern, Caroline Madding and R. H. Wise are acknowledged with
sincere thanks. Lawrence J. Kamphake was of invaluable assistance in
developing and refining the pilot plant automated monitoring systems
for ammonia and residual chlorine.
VI 1 1
-------
SECTION 1
INTRODUCTION
Breakpoint chloxination is a process that can be used to oxidize
ammonia nitrogen to nitrogen gas and thus remove it from wastewaters. If
breakpoint chlorination is followed by an activated carbon adsorption system,,
the carbon can accomplish both organic removal and dechlorination. The
formation of halogenated methanes during the disinfection of drinking water
supplies by application of small doses of chlorine to waters of relatively
low organic strength is a recently documented problem facing the water
supply industry. When ammonia is removed by breakpoint chlorination, large
doses of chlorine are applied to wastewater, with its relatively high con-
centration of organic material. The possibility that high concentrations
of halomethanes might be formed has caused some concern among regulatory
agencies over the impact of breakpoint chlorination on downstream water uses,
including drinking water supply.
To provide data on the potential problems involved in halomethane for-
mation during breakpoint chlorination of wastewater, a continuous-flow pilot
plant project was conducted for a period of 9 weeks. The major objectives
were to monitor the production of six volatile halogenated organic compounds
during breakpoint chlorination of wastewater and to determine the removal
efficiency of activated carbon with regard to these compounds. This report
describes the facilities used for the project and summarizes the results
obtained.
-------
SECTION 2
CONCLUSIONS
After secondary treatment by a trickling filter system and tertiary
treatment by lime clarification and filtration, municipal wastewater from
a City of Cincinnati interceptor sewer contained small but measurable
quantities of chloroform. The other five volatile halogenated organic
compounds were detected less than half the time.
Breakpoint chlorination of the tertiary filter effluent caused the for-
mation of 43 yg CHC13/1 and 4.4 yg BrCHCl2/l (median values). The other
four compounds were formed infrequently and/or in small amounts.
No correlations were observed between the amount of chloroform formed
and any other measured parameter.
Breakpoint chlorination of the tertiary filter effluent increased the
concentration of extractable organic chlorine in three sets of samples by
an average of 11.0 yg/1.
Treatment of the tertiary filter effluent with a 10-min contact time,
granular activated carbon column before breakpoint chlorination reduced the
median values of CHCl? and BrCHClo formed during chlorination to 12 yg/1 and
2.5 yg/1, respectively. There was no net production of the other four com-
pounds in this system.
The carbon lost its ability to reduce the formation of bromodichloro-
methane after treating approximately 4500 bed volumes of wastewater; the
carbon continued to reduce the amount of chloroform produced up to the
termination of the study (8400 bed volumes).
Carbon treatment reduced the amount of extractable organic chlorine
formed during chlorination by an average of 8.1 yg/1 in three sets of
samples.
Batch chlorination test results with chlorine doses in the wastewater
disinfection range showed little or no chloroform production over a period
of 40 hr of contact time.
The service life of the 20-min contactor treating chlorinated tertiary
filter effluent was much more than twice that of a 10-min contactor. The
carbon utilization rate of the 20-min contactor treated as a single-stage
-------
contactor was 200 g/cu m, compared to a utilization rate of 340 g/cu m for
the 10-min contactor, with 5 ug CHC1-/1 in the effluent as the exhaustion
criterion.
The capacity of the second-stage columns to remove chloroform was a
precise function of the average influent concentration. The first-stage
column removed CHC1, less efficiently than the second-stage column, pre-
sumably because more strongly adsorbed organics were preferentially adsorbed
in the first stage; thereby reducing the competition for adsorption sites
in the second stage.
The removal of chloroform by carbon was not affected by the presence
of 12 mg/1 of total residual chlorine.
-------
SECTION 3
RECOMMENDATIONS
More detailed information on the formation of halogenated organics by
chlorination of wastewater should be obtained for a broader range of com-
pounds under closely controlled conditions. Identification and quantifi-
cation of halogenated compounds formed during breakpoint chlorination would
be of interest. Effects of physical and chemical factors on yield and rate
of formation should be examined. Toxicity testing could be a valuable aid
in determining the necessity of minimizing the release of halogenated or-
ganics to the aquatic environment.
Studies on optimization of trihalomethane and precursor removal by
activated carbon should be undertaken, particularly with respect to drink-
ing water treatment. The influence of contact time on carbon utilization
rate should be carefully examined on a variety of waters. Contacting modes
should be compared (multistaged series, countercurrent, parallel single
stage). Batch equilibrium tests should be conducted for comparison with
continuous-flow exhaustion data.
If the discharge of chloroform is to be limited in a system utilizing
breakpoint chlorination and activated carbon, locating the chlorination
reactor between stages of a two-stage system would be more desirable than
locating it ahead of carbon treatment. This arrangement would take advan-
tage of precursor removal in the first stage and chloroform removal and
dechlorination in the second stage.
-------
SECTION 4
BACKGROUND
AMMONIA IN WASTEWATER
Ammonia nitrogen is found in all municipal wastewaters. Human fecal
matter contains organic nitrogen, some of which is converted to inorganic
ammonia by natural biological decomposition. Likewise, the urea in urine
is hydrolyzed to ammonia during the time of collection, transport and
storage in a municipal system.
Ammonia nitrogen can exist in wastewater as the ion NH. or the dis-
solved gas NH . The equilibrium expression:
O
NH3 + H+ -f >- NH* K = S.OxlO"10 @ 20°C (1)
indicates that at neutral pH the ammonium ion is the predominant species.
In this report the terms ammonia or ammonia nitrogen or NH_-N will refer
to the sum of the concentrations of NH and NH expressed as nitrogen.
The requirements for removal of ammonia from wastewater depend upon
a number of factors. Nitrogen discharged to lakes, estuaries or bays can
accelerate the eutrophication of such bodies of water. Ammonia nitrogen
exerts a demand upon the oxygen resources of either standing or flowing
bodies of water due to bacterial oxidation of ammonia to nitrite and
nitrate. The stoichiometric requirement is 4.6 mg oxygen per nig ammonia
oxidized. Other considerations, such as fish toxicity and impairment of
disinfection efficiency could also result in a requirement for ammonia
removal from wastewater.
AMMONIA REMOVAL
The most widely practiced method for reducing the concentration of
ammonia in wastewater has been biological oxidation to nitrate (nitrifica-
tion) in activated sludge or fixed film reactors. Nitrification can be
followed by biological denitrification, wherein the nitrate is reduced by
biological action to nitrogen gas. Nitrification can be used alone to
control the problems of ammonia discharge, e.g., nitrogenous oxygen demand,
ammonia toxicity, etc. However, nitrate nitrogen is still present to act
as an algal stimulant. Thus, nitrification must be followed by denitrifi-
cation to accomplish more or less complete removal of inorganic nitrogen.
-------
In contrast to the biological mechanism of ammonia control which
results in conversion of ammonia to nitrate, several physical-chemical
processes are available that eliminate the ammonia nitrogen from the waste-
water.
Air-stripping involves providing intimate contact of air with wastewater
at pH > 11. Ammonia is in the form of dissolved gas at this pH and thus is
available for transfer to the stripping air. This process can be accomplish-
ed in conventional cooling towers, but problems of scale formation on the
packing and inefficient stripping and freezing in cold weather require spe-
cial considerations. A modification of the process passes the ammonia laden
exhaust air through an absorption tower where an acid solution absorbs the
ammonia and eventually produces a fairly concentrated solution of ammonium
salt with some commercial value. The air is continuously cycled through
stripping and absorption towers in a closed system so that the CC>2 quickly
disappears and calcium carbonate scale formation is eliminated.
Ammonia can also be removed from wastewater by ion exchange using an
ammonium selective zeolite, clinoptilolite. Regenerant brine can be re-
covered and reused by stripping the ammonia at elevated pH. The closed
stripping adsorption system described above will be used at several full-
scale installations to remove ammonia from the regenerant brine.
BREAKPOINT CHLORINATION FOR NITROGEN REMOVAL
Chlorine is a strong oxidant which, when added to wastewater in suf-
ficient concentration, can oxidize inorganic ammonia to nitrogen gas. The
chemistry of the process is still somewhat undefined and will be discussed
only briefly here. Thorough discussions can be found in the literature
(1-13).
Chlorine gas dissolves in water according to the hydrolysis reaction:
HOC] + H+ + Cl~ (2)
Hypochlorous acid (HOC1) dissociates to maintain an equilibrium with the
hypochlorite ion:
HOC1 -« > H+ + OC1" K = 3.3xlO"8 @ 20°C (3)
Both HOC1 and OC1~ are active forms of aqueous chlorine. In this report
the sum of the concentrations of HOC1 and OC1~ will be designated as "free"
chlorine without regard to their proportions. Chlorination systems may use
either gaseous chlorine or hypochlorite (commonly sodium hypochlorite); the
resulting form of free chlorine depends upon the operating pH and tempera-
ture.
Aqueous chlorine reacts with ammonium ion to form a number of N-chloro
compounds known as chloramines. Monochloramine (N^Cl) is usually formed
if the pH is near neutral and the C1:NH3-N weight ratio is less than 5:1.
At low pH or higher C1:NH3-N ratios, dichloramine (NHC12) can be formed
-------
and under certain conditions (e.g., very high C1:NH3-N ratioj trichloramine
or nitrogen trichloride (NCI3)is produced. The total chloramine concentra-
tion is referred to as "combined" chlorine.
The concept of breakpoint chlorination (BPC) is illustrated in Figure 1,
a theoretical breakpoint curve. As chlorine is added to a dilute solution of
ammonia, the total chlorine residual increases at nearly the same rate as the
chlorine addition up to a C1:NH3-N weight ratio of about 5:1. Then as more
chlorine is added, the measured residual begins to decline until some point
is reached at which the chlorine residual is at a minimum. Beyond this point
the measured residual increases at the same rate as the applied chlorine. The
minimum point is known as the breakpoint. Beyond the breakpoint the chlorine
residual is nearly all "free" residual chlorine, while prior to reaching the
breakpoint the measured total residual is "combined" residual or chloramines.
The breakpoint is shown in Figure 1 as occurring at a C1:NH3-N weight ratio
of 7.6:1. This is the minimum stoichiometric ratio possible via a reasonable
sequence of reaction steps with N2 as the only nitrogenous end product.
TOTAL
RESIDUAL
CHLORINE
FREE
RESIDUAL
CHLORINE
5 7.6
Cl: NH3 -N WEIGHT RATIO
Figure 1. Breakpoint Curve.
The ammonia concentration begins to decline at about the 5:1 C1:NH,
ratio until ammonia is nearly completely absent at the breakpoint, It is
generally accepted that as the applied chlorine increases from 0 up to the
hump in the curve, monochloramine is being formed. The declining portion of
the breakpoint curve is where dichloramine is formed and decomposes to yield
nitrogen gas resulting in a decline of the ammonia nitrogen concentration.
Stone (12) summarizes the stoichiometry of the most reasonable set of re-
actions leading to breakpoint with the following expression, modified
slightly:
NH
+ 1.5 OC1'
->• .5 N2 + 1.5 H20 + H+ + 1.5 Cl"
(4)
-------
This accounts for the 7.6:1 minimum weight ratio (1-5:1 molar ratio) required
for ammonia oxidation.
Other end products from the chlorination of ammonia, notably nitrate
(NC>3) and nitrogen trichloride have been observed (1, 4, 10, 12). The
production of these substances is to be avoided for a number of reasons.
The formation of nitrogen trichloride results in a chlorine consumption of
15 mg for each mg of NClj-N produced. Nitrogen trichloride is a slightly
soluble gas with an obnoxious odor. Its presence also contributes to the
total nitrogen of the effluent, as it probably reverts to ammonia upon
dechlorination (1). The formation of each mg of nitrate-N requires 20 mg
of chlorine. The presence of nitrate also, of course, contributes to the
total effluent nitrogen and should be minimized.
Equation 4 shows a net increase of one mole of hydrogen ions as one
mole of ammonia is oxidized to N2- In addition, if gaseous chlorine is
used, the hydrolysis of sufficient chlorine to oxidize one mole of ammonia
liberates three moles of hydrogen ion. These hydrogen ion additions could
result in serious pH depression unless the we.stewater alkalinity is suf-
ficiently high or unless they are neutralized by addition of base. Palin
(11) and Pressley et al. (4, 10) observed increased NCl^ production at
lower pH although Stone (12) and Saunier and Sellick (1) did not observe
this effect. In any event, the pH of the reaction should be controlled as
near neutral as possible. In order to neutralize the hydrogen ions produced,
a source of hydroxide ions can be supplied in the form of, for example, lime
or sodium hydroxide. Four equivalents of lime or sodium hydroxide translate
to 1.05 or 1.50 parts of CaO and NaOH, respectively, needed for each part of
chlorine added. In practice, the amount of lime or sodium hydroxide added
would be controlled to maintain a constant pH in the reactor effluent. In
a full-scale breakpoint study, Stone used 1.53 g NaOH/g Cl- using pH con-
trolled sodium hydroxide feed (12).
Both the chlorine added and the alkalinity supplement contribute to a
substantial increase in total dissolved solids (TDS) in the effluent. Gas
chlorination with lime neutralization of all acidity would result in a TDS
increase of 12 parts per part of ammonia destroyed, while the use of sodium
hydroxide would result in a 15:1 increase in TDS (13). Sodium hypochlorite
is purchased from chemical suppliers or manufactured on site in the form of
a caustic solution. Thus the TDS and the pH adjustment chemical require-
ments would depend upon the caustic content of the solution fed.
Saunier and Sellick (1) conclude that the ammonia concentration reaches
its lower limit in 60-90 seconds. Stone's data (12) on a full-scale pipe-
line reactor showed minimum ammonia concentration occurring at 100-200 sec-
onds. Thus, small reaction vessels may be used for the breakpoint chlorin-
ation process.
During three months of operation of a 25 gpm (2.2 I/sec) BPC pilot
plant with good process control, Pressley, et al. reduced the ammonia
content of lime-clarified and filtered raw wastewater from 10.0 to 0.4 mg
N/l (10). The chlorine to ammonia nitrogen weight ratio was 9:1 and the
-------
pH was 7.1-7.3. The alkalinity requirement was 1.8 parts NaOH/part Cl2.
Little or no organic nitrogen removal was observed in the pilot operation,
resulting in an effluent total nitrogen content of 2.5 mg N/l.
A large-scale demonstration of breakpoint chlorination was conducted
at Rancho Cordova, California (12). A C1:NH,-N ratio of 10:1 was required
to insure a controllable free residual chlorine (8 mg/1). In a series of
experimental runs conducted in the pH range 7.0-8.0, the ammonia content
of secondary effluent was reduced from a mean value of 17.7 mg N/l to
0.55 mg N/l. Oxidation of NH3-N to NO^-N produced a mean value of 1.06 mg
NOj-N/l, with no apparent influence of pH in that narrow range. Organic
nitrogen concentration was unchanged through the breakpoint process. All
but 0-8% of the chlorine added could be accounted for in the oxidation of
influent ammonia to N2, NO, and NClj (assuming all residual NH3 was the
product of dechlorination of NClj), the oxidation of N02 to NOj and the
free residual remaining.
A variation of the breakpoint chlorination process using activated
carbon has been examined by several investigators (14,15,16,17,18,19).
During a pilot study at Owosso, Michigan, effluent from a chlorination
system was treated with activated carbon for organic removal (14,15).
It was observed there and subsequently confirmed by others that activated
carbon can react with chloramines to form nitrogen gas and chloride. Thus,
at chlorine dosages below the breakpoint, partial ammonia removal could be
achieved if activated carbon followed the BPC system. Activated carbon
following BPC also provides removal of the free-residual chlorine by
reduction of the active chlorine species to chloride (20).
THE CHLORINATED ORGANICS PROBLEM
With the development of more sophisticated analytical tools came the
awareness that a vast number of organic substances contaminate our nation's
surface waters and can thus be found in drinking water supplies. In
addition, it was observed both in the United States (21) and Europe (22)
that some finished drinking waters contained trihalomethanes that were not
found in the raw waters, indicating their formation during the process of
chlorination.
A survey of the drinking water in 80 U. S. cities was made by EPA
during 1975 (23). The concentration of six volatile halogenated organic
compounds was determined by the gas chromatographic technique described
by Bellar and Lichtenberg (24). Chloroform (CHC13) was found in the
finished water of all 80 ci.ties surveyed, ranging in concentration from
barely detectable (>0.1 yg/1) to over 300 ug/1. Bromodichloromethane
(BrCHCl2), dibromochloromethane (Br7CHCl), and bromoform (CHBr3) were
also found, although less frequently and in lower concentration. Carbon
tetrachloride (CC14) and 1,2-dichloroethane (C2H4C12) were detected in
88% and 68% of the finished waters, respectively, although there was no
evidence of their formation during chlorination.
-------
Chloroform has been named a suspected carcinogen, and its presence in
drinking waters should be minimized. Stevens et al. (25) and Love et al.
(26) have investigated the formation and removal of trihalomethanes in the
treatment of surface waters with a variety of processes. Stevens et al. (25)
show that applying chlorine after coagulation and settling rather than to
raw water greatly reduced the amount of chloroform produced. Chlorination
of solutions of humic acid gave similar results as chlorination of Ohio River
water at similar levels of total organic carbon (TOG). Formation of tri-
halomethanes by chlorination of raw water, humic acid and acetone was en-
hanced at high pH, but acetone produced very little trihalomethane when
chlorinated at neutral pH. The formation of chloroform in raw water increas-
ed with temperature. Very little trihalomethane was formed when combined
chlorine (chloramine) was used rather than free chlorine. The formation of
trihalomethane can proceed for days, until either the free chlorine resjdual
or the precursor substance is exhausted.
Love et al. (26) examined the removal of trihalomethanes and trihalo-
methane formation potential (THMFP) by a variety of processes, including
aeration, powdered and granular carbon, and ozonation. Trihalomethane
formation potential was defined as the amount of trihalomethanes formed
during contact with free chlorine for a specified time (several days) at a
specified pH and temperature (27). It is an empirical designation that is
analogous to the trihalomethane precursor concentration.
Variable results were obtained with granular activated carbon (GAG)
(26). Treatment of chlorinated surface water with GAG in 5-min empty bed
contact time pilot filters resulted in complete breakthrough of chloroform
in 6 to 12 weeks. Brominated trihalomethanes were retained by the carbon
for a longer period before breakthrough. After the GAG beds were exhausted,
there was often desorption of trihalomethanes, indicating either true equi-
librium (reversible adsorption) or displacement of trihalomethanes by more
strongly sorbed substances. In one test, a 10-min contact time carbon
contactor gave twice the service life as a 5-min contact time contactor.
With an average influent chloroform concentration of 24 yg/1, the 10-min
bed was completely exhausted for chloroform removal (0% removal) in 21
weeks. Assuming continuous operation, this amounted to an applied chloro-
form loading of 1.2 g CHCl3/kg carbon at complete breakthrough.
A 5-min GAG adsorber removed chloroform formation potential for approxi-
mately 13 weeks, while a 10-min GAG adsorber was removing greater than1 50%
chloroform formation potential at the termination of the experiment at 30
weeks' service time. GAG was less effective in reducing the potential for
formation of the bromodichloromethane and dibromochloromethane.
Morris, in a review of available literature, concluded that chlorination
of natural waters "... does not lead to the formation of all sorts of
chlorinated derivatives with any and all pollutants. Rather it proceeds by
a limited number of well-defined reactions on a few specific types of organic
structures" (28). It is well known chlorine reacts with phenols to produce
chlorophenols, causing taste and odor problems (2,28). In wastewater, the
number and concentration of organic compounds are such that it would be
surprising if some chloro-organics were not formed during chlorination.
10
-------
Jolley (29)found more than 60 chlorine-containing constituents in a
series of experiments involving chlorination of primary and secondary efflu-
ents at relatively low (disinfection) doses. He estimated that approximately
l°o of the chlorine added was associated with stable chlorine-containing or-
ganic compounds when chlorinating secondary effluent. Other investigators
have also determined that reactions can indeed occur between added chlorine
and organic constituents in sewage (30,31). Evans (32) found yg/1 quantities
of organic chlorine (extractable in 15% ethyl ether in hexane) were produced
when breakpoint chlorinating a wastewater that had been treated with lime
clarification, filtration, and activated carbon.
STUDY OBJECTIVES
Breakpoint chlorination is a candidate process for nitrogen removal in
some circumstances. This study was instigated because of the very evident
possibility of the formation of chlorinated organics when adding large doses
of chlorine to wastewaters containing high concentrations of organic sub-
stances (relative to surface waters). The objective was to operate a pilot
plant system simulating as nearly as possible a treatment train producing
high quality (reuse quality) effluent and using breakpoint chlorination for
nitrogen control. The magnitude of the chlorinated organics formation prob-
lem would be assessed together with the capability of activated carbon for
removing the chlorinated organics and for preventing their formation. This
was to be done on a continuous-flow pilot plant that would be subject to
typical changes in wastewater characteristics. The process train consisted of:
1. Secondary treatment by trickling filter
2. Tertiary lime clarification
3. Neutralization
-1. Granular media filtration (GMF)
5. Breakpoint chlorination (BPC)
ti. Granular activated carbon adsorption (GAG)
Details of the treatment processes are provided later.
Because of the difficulty in measuring a large number of chloro-organics
the trihalomethanes, CHCl-, BrCHCL-,.Br^CHCl, and CHBr3 together with CC14 and
C->H,Cl->, were the halogenated organics measured routinely. There was a grow-
ing body of data on these compounds in relation to water treatment practice,
and it would be possible to compare results of wastewater chlorination with
water treatment. The analytical procedure was not overly time comsuming,
therefore, the large number of samples would not impose an impossible burden
on the analytical staff. To supplement the data provided by measurement of
the six volatile halogenated organics (VHO), several sets of samples would be
analyzed for extractable organic chlorine. Conventional pollution parameters
would be measured to document the general efficiency of the system. Two
chlorination systems would be installed, one after the GMF and before a two-
stage GAC adsorber, and the other between stages of a second two-stage ad-
sorber. A third adsorber without any chlorination would be run in parallel
to the other two. With this arrangement, VHO removal, VHO precursor removal,
general organics removal, and chlorine removal could be examined in relation
to each other.
11
-------
SECTION 5
DESCRIPTION OF FACILITIES
PROCESS EQUIPMENT
Raw wastewater was pumped continuously from a 72-inch (1.8 m) inter-
ceptor serving residential and industrial areas of Cincinnati. After
primary clarification, the wastewater was directed to two parallel trick-
ling filters, each 10 ft (3 m) diameter containing 8 ft (2.4 m) of 4 in
(0.1 m) blast furnace slag as the attachment medium. The effluents from
the trickling filters were directed to two 38-in (0.97 m) diameter second-
ary clarifiers. The flow rate through the primary-secondary system was
10 gpm (0.6 I/sec).
A schematic diagram showing the major elements of the tertiary
clarification-filtration system is shown in Figure 2. Secondary effluent
from the trickling filter system flowed by gravity to a collection sump.
This sump contained a level detector which, at loss of flow, provided a
signal to shut down all tertiary system pumps, feeders, etc., and, coupled
with appropriate time-delays, to restart the system upon restoration of
influent flow. Only one shift/day attention was normally provided; the
automatic stop-start capability allowed the system to cope with periodic
loss of flow due to pump station problems or occasional shut-down of the
trickling filter system during odd hours.
TRICKLING
FILTER LIME!
EFFLUENT I •
))
Ry
1
r/\ ,-, i SLUDGE
>S Y ! DETECTOR
/
HPiD MIX
\
i
I
I
I
i
FLOCCULATORi
i
v— —
C
*v
3
/
SLUDGE RECYCLE
CLARiFIER
TO WASTE
CO2
® TO CARBON
-"— SYSTEM
NEUTRALIZATION TRI-MEDIA
TANK FILTER
(§)- DENOTES SAMPLE POINT
--- DENOTES ELECTRONIC CONTROL LOOP
Figure 2. Flow diagram, tertiary lime clarification-filtration.
12
-------
Secondary effluent was pumped at 5 gpm (0.3 I/sec) from the collection
sump to the rapid mix unit by a Moyno pump (Robbins § Meyers, Inc., Spring-
field, Ohio). Flow was monitored by a 0.5 in (13 mm) magnetic flow meter
(Fischer § Porter Co., Warminster, Pennsylvania) with remote indicator and
recorder. The wastewater was mixed with an average dose of 414 rag Ca(OH2)/l
for a period of approximately 3 min in the baffled, 16 gal (0.06 cu m) rapid
mix unit. Mixing energy was supplied by a 68 rpm turbine mixer.. The lime
dose was controlled automatically by a pH control system (Leeds £ Northrup,
North Wales, Pennsylvania). The pH was monitored by submersible probe in
the flocculation unit downstream of the rapid mix tank. The probe was placed
there in order to smooth out pH fluctuations caused by the pulsating action
of the slurry pump and the automatic flushing of the pump head with tap water.
The probe output was directed to a remote monitor and recorder and also to
an electronic controller. The controller output regulated the lime slurry
pump rate to maintain a constant pH of 11.2-11.5 in the clarification system.
Lime slurry was prepared by adding preweighed 10 Ib (4.5 kg) bags of Ca(OH2)
to water in the amount of 1 bag/10 gal (1 bag/0.4 cu m). The slurry pump
was a pH controlled diaphragm solution pump (Precision Control Products
Corp., Meriden, Connecticut) that had been modified by replacing the pump
head with that of a slurry pump head of the same manufacture. The slurry
pump head was then equipped with a timed tap water flush to clear the head
and ball checks of debris automatically. It was also necessary to acid
clean the lime pump once a day with 50% HC1 in order to dissolve scale for-
mations that had accumulated on the pump head and check valves.
The flocculation tank was a 43 gal (0.16 cu m) baffled tank, providing
a residence time of 9 min. Flocculation energy was provided by a 12 rpm
mixer with two vertical 1-in wide (25 mm) paddles located 7.5 in (190 mm)
from the center of rotation.
The flocculated wastewater entered the clarifier at a 6-in (150 mm)
diameter center well which extended to a depth 60 in (1.5 m) below the
water surface. The clarifier was 38 in (0.97 m) in diameter with a straight
side depth of 56 in (1.4 m) and a 60° hopper bottom. Clarified wastewater
flowed upward and over a peripheral effluent weir. Sludge settled to the
hopper bottom where it was continuously pumped back to the rapid mix or
to waste. The level of sludge in the clarifier was automatically maintained
at or below a specified depth by a photoelectric sludge blanket level de-
tector fabricated in our laboratory (Figure 3). When the sludge blanket
level was below the elevation of the light source/photocell, no action was
taken. When the sludge blanket rose to that point, the sludge blocked the
light to the photocell. This break in photocell output triggered a control
system which energized an electrically actuated ball valve to divert the
sludge recycle stream to waste for a preset time interval. The clarification
system could also be operated without continuously recycling sludge. Under
this mode of operation the sludge level detector-controller turned on the
sludge pump to waste for a preselected time. In either mode of operation,
sludge wasting was followed by an automatic tap water flush of the sludge
pipeline to prevent deposition in the line.
1 3
-------
Figure 3. Sludge blanket level detector.
Effluent from the lime clarifier flowed by gravity to a rectangular
tank for neutralization. The tank was 16 in (0.41 m) by 24 in (0.61 m) by
56 in (1.4 m) deep. The detention time was 16 min. The high pH lime
clarifier effluent was neutralized to pH 7.0-7.2 by adding sulfuric acid
and a manually controlled amount of carbon dioxide gas. The sulfuric acid
addition was controlled by a Leeds f, Northrup pH control system similar to
that used to control the lime feed rate. Carbon dioxide was added to pro-
vide bicarbonate alkalinity (buffer capacity). This allowed much smoother
neutralization pH control and provided buffer capacity required for down-
stream processes. A constant speed Moyno pump transferred neutralized
effluent to the next process, gravity filtration. The transfer pump rate
was slightly higher than the flow into the neutralization tank. A recycle
line off the pump discharge line returned sufficient flow to the neutrali-
zation tank so that the outflow from the neutralization tank was paced to
the inflow. The float operated valve that controlled the recycle flow also
maintained the proper water level in the neutralization tank for efficient
CC>2 dissolution.
•
The gravity filter was 11 in (0.3 m) square cross section. The media
consisted of 3 in (80 mm) of 40x80 US mesh garnet sand, 9 in (230 mm) of
20x40 US mesh silica sand and 9 in (230 mm) of 10x20 US mesh anthracite
filter media. After initial backwashing, the garnet sand layer became mixed
with the silica sand layer and lost its identity. Flow was in a downward
direction with gravity as the driving force. The flow rate through the fil-
ter was controlled at 3.3 gpm (0.21 I/sec) by three Moyno pumps on the efflu-
ent line. The resulting hydraulic loading was 3.9 gal/min/sq ft (9.5 m/hr).
These pumps supplied filter effluent to the downstream processes. Any excess
14
-------
filter influent overflowed the top of the 10 ft (3 m) high filter and was
directed to an automatic sampler. The filter media support was a #80 US mesh
stainless steel screen sandwiched between two PVC perforated plates supported
by an open-grid false bottom.
The head loss in the filter operation was monitored by a piezometer-
bubbler tube-pressure switch arrangement. When the filter clogged to the
extent that the low pressure limit was reached, the monitor automatically
shut down the transfer pump that supplied water to the filter and all down-
stream equipment. An alarm was sounded, signaling the high head loss shut-
down condition. Backwash by an air-scour-water wash system was then initi-
ated by the operator. After the water was drained to a level just above
the top of the filter media, the bed was scrubbed by introducing air into
the underdrain plenum. Two to three minutes of air scour at approximately
8 scfm/sq ft (150 m/hr) was followed by a five-minute water wash at approxi-
mately 90% expansion. After backwash was completed, the operator initiated
filter and downstream process operation by means of a pushbutton on the
filter pressure monitor panel. The filter supply pump started and when the
wastewater above the filter was at the proper level, the filter effluent
pumps and all downstream processes were turned on automatically.
A schematic diagram of the breakpoint chlorination (BPC) activated
carbon systems is shown in Figure 4. Three activated carbon systems were
operated in parallel. All were two-stage series expanded bed systems oper-
ating on filter effluent. In System 3, the filter effluent underwent BPC
before application to the first stage carbon adsorber. In System 2, the
BPC system was located between the first and second stage adsorbers. System
1 had no BPC and was used as a control.
Each activated carbon system contained 8 ft (2.4 m) (at repose) of
Filtrasorb 400 (Calgon Corporation, Pittsburgh, Pennsylvania] granular
activated carbon per stage for a total depth of 16 ft (4.9 m). The hydraulic
application rate to each adsorber was 6 gal/min/sq ft (15 m/hr) for a result-
ant contact time of 10 min/stage (based on empty bed volume, carbon at repose)
The rate of flow to each carbon system was controlled by the filter effluent
pumps and monitored by rotometers. The diameter of each adsorber was 5.8 in
(15 cm). The carbon was cleaned weekly by a separate air scour-upward flow
water wash. The air scour rate was 8 cu ft/min/sq ft (150 m/hr), and the
water backwash rate was 15 gal/min/sq ft (37 m/hr).
The breakpoint chlorination systems are illustrated in more detail in
Figure 5. The major components were one automatic ammonia analyzer (Technicon
Instruments Corporation, Tarrytown, New York) common to both systems, one
chlorination control unit, two sodium hypochlorite pump controllers, two
sodium hypochlorite pumps (Precision Control Products Corporation, Meriden,
Connecticut), two in-line static mixers, two cylindrical glass reactors and
an automatic chlorine analyzer (Technicon) measuring both total and free
chlorine. The ammonia monitor continuously measured the concentration of
ammonia nitrogen in the filter effluent. An electrical signal proportional
to the ammonia concentration was transmitted to the chlorination control
unit, a special unit designed and fabricated in-house. As finally used, the
major function of the control unit was to convert the 0-5 VDC signal from
the autoanalyzer to a 12-20 .mA signal to be supplied to the sodium
15
-------
1A
FROM TRI-
1B
MEDIA
FILTER
(D-DENOTES SAMPLE POINT
BK3
3A
3B
BK2, BK3 - BREAKPOINT
CHLORINATION
REACTORS
1A, 2A, 3A - 1st STAGE ACTIVATED
CARBON COLUMNS
IB, 2B, 3B - 2nd STAGE ACTIVATED
CARBON COLUMNS
Figure 4. Carbon/chlorination system.
PH
ANALYZER
on
i i -
j[ ':
ESIDUAL
HLORINE
^ALYZER
1
R R] IN-LINE
E m MIXER
A liJ PUMP
^ XL)- ' " ~ CONTF
0 FROM
R FILTER '
— i NaOCI
T- —TO CARBON 3A , >-
CHLORINAT
U TIMER I CONT
j L
r— -,
R ft] IN-LINE
| 5 MIXER
T 'Q- Ey»
o FROM T cor
R CARBON 1
r] — 2A
~X " ' NaOCI
w
^^ TO CARBON 2B
ROLLER
FILTER
EFFLUENT
SAMPLE
ION ,-,-,
ROL — )
NIT L1-J
NH3
ANALYZER
AP
STROLLER
Figure 5. Breakpoint chlorinatiort system.
16
-------
hypochlorite feed pumps. The autoanalyzer signal could also be overridden
by a manual input for use when standardizing the ammonia analyzer or correct-
ing malfunctions.
Thus, the input signal to the sodium hypochlorite feed pump controllers
was a function of the concentration of ammonia nitrogen in the filter efflu-
ent; this was the automatic feed forward control. The output:input ratio of
each pump controller was adjusted manually according to the operator's obser-
vation of the free residual chlorine in each of the breakpoint reactor efflu-
ents; this was the manual feedback control.
Sodium hypochlorite was diluted with distilled water to a concentration
of 7-9 g available chlorine per liter for use on the pilot system. The
available chlorine concentration was checked daily. The sodium hypochlorite
solution was pumped by the two diaphragm metering pumps through pulse elimi-
nators, rotometers and back pressure regulators to mix with the breakpoint
influents iust uostream of two Kenics Static mixers (Kenics Corporation,
Danvers, Massachusetts). Each chlorinated stream then entered a 3.8 in
(95 mm) I.D., 60 in (1.5 m) long glass reactor with a detention time of
2.7 min. Sequenced samples of each reactor effluent were directed to the
chlorine monitor, an autoanalyzer (Technicon) for measurement of free and
total residual chlorine. The operator could then observe the state of the
breakpoint chlorination process in each reactor and adjust the output:input
ratios of sodium hypochlorite pump controllers to maintain the proper con-
ditions. Additional information was available from the continuous record-
ing of the pH of each reactor effluent. Reactor pH was controlled only by
adjustment of the pH and alkalinity of the lime clarifier neutralization
tank contents. When all systems were under control, the reactor effluents
were at a pH of 7.0 ± 0.5.
A second residual chlorine analyzer accepted sequenced samples from the
final effluent from carbon column 2 and from the first and second stage
effluents from carbon column 3. The purpose of monitoring these streams
for residual chlorine was to obtain data on the efficacy of activated carbon
for removal of chlorine residual. A great deal of difficulty was encounter-
ed with this analyzer, and thus the amount of data obtained on this aspect
of the system was minimal.
SAMPLING SYSTEMS
Effluents from the trickling filter,clarifier, neutralization tank,
filter, breakpoint reactors and first and second stage carbon columns were
sampled by the automatic sampling systems shown in Figures 6 and 7. The
dipper samplers illustrated in Figure 6 were used to obtain automatic com-
posite samples from all process streams after the filter. Small volumes of
the process streams were directed continuously to the sample chambers to
keep them full. The dippers rotated once each 15 min, collecting 8 ml
aliquots and discharging them to their respective collection troughs, where
they were drained to refrigerated sample bottles. The diverter samplers
shown in Figure 7 were used for secondary effluent and neutralized lime
clarifier effluent since suspended solids tended to settle in the dipper
sample chambers.
1 7
-------
TO REFRIGERATED
COMPOSITE
SAMPLE JUG
ROTOMEIER
FROM
^
vx
SAMPLE ;
POINT
\-
OVERFLOW
^>
-TO DRAIN
Figure 6. Dipper sampler.
FROM SAMPLE POINT
I,
~y~~a~~^r~y
^
TO DRAIN
ELEQRONIC
SAMPLE
TIMER
I
i
TO REFRIGERATED COMPOSITE
SAMPLE JUG
Figure 7. Diverter sampler in the sampling position.
-------
A special sampling apparatus was fabricated for use in obtaining com-
posite samples for subsequent analysis of volatile organic compounds. A
schematic of the apparatus is shown in Figure 8. Effluents from the filter,
breakpoint reactors and first and second stage carbon columns were sampled
with this system. An in-house designed and fabricated electronic timer,
shown in Figure 9, opened the solenoid valve A to allow the sample to flush
the sample line for a selected time interval through a three-way solenoid
valve B to waste. The timer then energized the three-way valve for an
instant so that the sample was diverted into the sample chamber. The sam-
pling interval, the flush time and the sample impulse time were all adjust-
able to allow for the proper sample volume and to insure complete flushing
of the sample line. The electronic timer operated a total of nine samplers.
The composite sample chamber consisted of a 2-liter graduated cylinder
equipped with a flat bottom and a glass tube inlet-outlet port. A Teflon
float was machined to fit each sample chamber so that, as the sample volume
increased, the float would ride up on the sample surface. This float appa-
ratus permitted the collection of composite samples in closed containers
and minimized the loss of volatile organics during the composite period.
A few grains of sodium thiosulfate were added to each chamber at the begin-
ning of each sample period to prevent further reaction of chlorine and pre-
cursors during the composite period. Each sample chamber contained a mag-
netic stirring bar which was used to mix the sample at the end of the com-
posite period. An air driven magnetic mixer was located under the sample
chamber to drive the stirring bar. The photograph in Figure 10 shows most
of the components. The closed composite sampler is described in more detail
elsewhere (33).
SYSTEM OPERATION
System operation and data collection commenced on June 3, 1976, and
continued until August 9, 1976. This equipment ran 24 hours a day, 7 days
a week with only minor periods of shutdown as a result of equipment mal-
functions or interruptions in the supply of secondary effluent. Operator
surveillance was provided eight hours per day during the week and four hours
per day on weekends during the first half of the study. During the last
five weeks, technicians were in attendance 20 hours per day on weekdays and
4 hours per day on weekends, resulting in much improved system operation.
The pH of the secondary effluent, flocculation tank, neutralization
tank, and each breakpoint reactor were continuously recorded. There were
high pH excursions (up to ^ 11) in the breakpoint systems immediately upon
startup of the chlorination units following a shutdown. These excursions
were caused by the pulse eliminators unloading pressure upon cessation of
process flow, and injecting excessive sodium hypochlorite into the reactor.
When the system was restarted, that excess sodium hypochlorite resulted in
higher than normal pH (and high free chlorine) exiting the reactors for a
period of approximately 5-10 minutes.
The ammonia analyzer functioned very well with very little attention.
The baseline and a standard were checked daily and very little drift was
observed. The only problem in the use of the instrument was the time
required to pump the sample from the process stream through the various
1 9
-------
FROM SYSTEM-^;
SAMPLING POINT
A
X
SAMPLE TO OTHER SAMPLERS
TIMER
B
. '.. . — C
fir/
SAMPLE
CHAMBER
/ TEFLON
FLOAT
*
STIRRING
/BAR
D
II i H-*- AIH
fT ^ MAGNETIC
| MIXER
TO TO SAMPLE
WASTE VIAL
VALVE
TYPE
FUNCTION
A
B
C
2-WAY
SOLENOID
3-WAY
SOLENOID
MANUAL
TOGGLE
MANUAL
NEEDLE
DIRECT SAMPLE
TO VALVE B
DIRECT SAMPLE
TO CHAMBER
*
TRANSFER COMPOSITE
SAMPLE TO VIAL
& DRAIN REMAINDER
CONTROL MIXER
SPEED
Figure 8. Automatic composite sampler for volatile halogenated organics
20
-------
Figure 9. VHO sample programmer.
-------
Figure 10. VHO sampler.
22
-------
mixing coils and heating bath to the colorimeter. This resulted in a lag
time of 15 minutes between a change in the ammonia concentration of the
filter effluent and a corresponding change in the analyzer output - hence
sodium hypochlorite feed rate. The normal diurnal variation of the waste-
water resulted in a rather sharp increase in ammonia concentration of the
filter effluent beginning early afternoon and ending in the early evening.
This increase ranged from 1 mg/l/hr to as much as 6 mg/l/hr. This presented
difficulties in maintaining breakpoint conditions. Although the influent
ammonia increase could usually be accommodated by careful operator attention,
there were occasions when the breakpoint was not reached. These occasions
were indicated by high total chlorine residual but no free chlorine residual
in the BPC reactor effluents.
The residual chlorine analyzers were usually reliable in indicating
whether sufficient chlorine was being fed to reach the breakpoint. This was
evident when the recorder trace of free and total chlorine residual were
parallel, with total chlorine about 2-4 mg/1 higher than free chlorine. At
a dosage less than that required for breakpoint, the free chlorine trace was
a straight line at or near zero, while the total chlorine read high, depend-
ing upon the actual chlorine to ammonia ratio being fed. There was usually
1-5 mg/1 baseline drift in 24 hours.
Composite samples were taken from the refrigerated sample jugs three
times per week, acidified, and placed into various containers for transfer
to the analytical service laboratory. All operating data were summarized
on the basis of each sample period,e.g., residual chlorine and ammonia nitro-
gen values were averaged over the operating time for each sample period.
Lime and acid dosages were calculated from volume measurements and specified
feedstock concentrations. Carbon dioxide feed rate was estimated from the
rotometer readings. Sodium hypochlorite feed rates were determined by volume
measurements and expressed as available chlorine. The chlorine concentration
in each sodium hypochlorite stock tank was measured daily by the starch-
iodine method.
At the end of the composite period, the contents of each of the closed
composite samples were mixed with the stirring bars for a short period of
time to assure that a representative sample would be withdrawn. Then valve
"C" (Figure 8) was opened manually to allow the contents of the chamber to
flow into the bottom of a 20 ml serum vial. The vial contents were dis-
placed several times to minimize air contact and a convex meniscus was left
at the top. The vials were then sealed with a Teflon disc held in place by
a crimped aluminum cap. After the samples were transferred to the sealed
vials, the remaining contents of the sample chambers were drained, and the
floats were lifted out. The sample chambers were rinsed with distilled water,
if necessary, and a few grains of sodium thiosulfate were placed in each
chamber for dechlorination of the next composite sample. The floats were
then reinserted, and the sampler was ready for another composite period.
-------
SECTION 6
ANALYTICAL METHODS
LABORATORY ANALYSES
Composite samples were submitted to the Analytical Service Laboratory
for analysis on Monday, Wednesday and Friday of each week. Total organic
carbon (TOC) was measured on a Beckman Carbon Analyzer (Beckman Instruments,
Inc., Fullerton, California) (34). Soluble organic carbon (SOC) was deter-
mined by filtering the samples through a 0.45 x 10~3 mm pore size membrane
filter prior to analysis by the carbon analyzer. The TOC procedure removes
carbonate interference by purging with nitrogen at pH 2. Thus, some vola-
tile organic substances can be lost. Suspended solids and total dissolved
solids (TDS) were determined by the procedures in Standard Methods (34).
The automated procedure described by Kamphake et al. was used to measure
nitrite and nitrate concentrations (35). Ammonia nitrogen was determined by
an automated phenate method (36,37). Total Kjeldahl nitrogen (TKN) was
determined by manual digestion (34) followed by determination of the result-
ing ammonia by the automated phenate procedure.
The procedure of Bellar and Lichtenberg was used to measure the six
volatile halogenated organic compounds (24,38,39). The volatile substances
were purged from the sample with helium and collected on an adsorbent trap
(Tenax). The trapped organics were then introduced into a gas chromato-
graphic column (1.8 m x 4 mm glass column packed with 60/80 mesh Tenax) by
helium purging at elevated temperature. The gas chromatograph was equipped
with an electrolytic conductivity detector operated in the halide mode.
Ammonia interference was overcome by acidifying the samples before purging
to avoid ammonia volatilization. Peaks obtained as the output of the de-
tector were identified by comparison with retention times of known standards.
Concentrations were determined by comparison of peak heights with a*standard
curve.
A measure of the precision and accuracy of the analyses was obtained in
the following way. A sample of breakpoint chlorinated, lime-treated second-
ary effluent was analyzed in replicate seven times. Another portion of the
same sample was spiked with 10 yg/1 each of the six VHO's, except bromodi-
chloromethane which was added at 20 yg/1. At the time this test was con-
ducted, the peaks of 1,2-dichloroethane and carbon tetrachloride could not
be resolved. The peak at this retention time was arbitrarily assigned to
1,2-dichloroethane. The procedure was modified later to permit differenti-
ation of the two compounds. The results of the evaluation of the method are
24
-------
shown in Table 1. Both precision and accuracy were good. Recovery of the
combined 1,2-dichloroethane/carbon tetrachloride was somewhat low since a
single peak representing two compounds with differing instrument responses
was measured.
Three sets of samples were analyzed for extractable organic chlorine.
The method used was described by Evans (32). Each sample was extracted with
a 15% solution of ethyl ether in hexane, the extracts were desiccated with
sodium sulfate and the volume reduced to a few ml. The determination of
organic chlorine was made with a Dohrmann Microcoulometer (Dohrmann, Division
of Envirotech, Santa Clara, California). The instrument employs a micro-
coulometric halide detection cell in conjunction with a combustion furnace.
In the determination of chlorine containing compounds, the sample is burned
in the combustion zone at about 800°C in excess oxygen. Organic chlorine
compounds are converted to gaseous hydrogen chloride which is measured by
coulometric titration.
PILOT PLANT ANALYTICAL METHODS
The ammonia nitrogen in the filter effluent was continuously determined
by a Technicon Monitor IV autoanalyzer. The method was the same as that used
for the laboratory analysis. Free and total available chlorine residuals
were continuously determined by a dual Technicon I system with a two-pen
recorder. A modification of Palin's method (40), using diethyl-p-phenylene-
diamine indicator was developed (36). The analyzers were standardized at
least once a day using the starch-iodine method (34). The operation of the
chlorine analyzer was occasionally checked by feeding a solution of sodium
hypochlorite to a buffered 5 mg N/l ammonia standard, while simultaneously
measuring the total and free chlorine residuals. If the monitor were
functioning properly, it would trace a typical breakpoint curve.
The concentration of chlorine in the sodium hypochlorite feed solution
was determined daily by the starch-iodine method. Pilot plant staff measured
composite sample alkalinity and turbidity; alkalinity was measured by titra-
tion with standard acid to pH 4.5 and turbidity with a Hach 2100 Turbidimeter
(Hach Chemical Company, Ames, Iowa). All pilot plant continuous pH monitors
were checked against the bench-top meter daily for evidence of drift. Occa-
sional cleaning of the lime treatment and recarbonation pH probes prevented
any serious difficulties in response time.
25
-------
TABLE 1 - PRECISION AND ACCURACY OF VOLATILE HALOGENATED ORGANICS ANALYSIS
CHC13
Replicate
Sample
Avg.
S.D.
45
47
41
45
42
40
44
43.4
2.5
10 yg/1
Spike
50
58
57
54
55
48
52
53.4
3.6
BrCHCl2 Br2CHCl CHBr3 C1CH2-CH2C1 + CC14 +
Replicate 20 yg/1 Replicate 10 yg/1 Replicate 10 yg/1
Sample Spike Sample Spike Sample Spike
1.6 22 N1"
1.3 22 N
1.6 24 N
1.3 22 N
1.4 22 N
1.5 22 N
1.3 22 N
1.43 22.3 N
0.138 0.76
10 N
10 N
10 N
10 N
11 N
12 N
12 N
10.7 N
0.95
10
11
10
10
12
13
13
11.3
1.4
Replicate 10 yg/1
Sample Spike
N
N
N
N
N
N
N
N
_
16
18
17
17
17
15
15
16.4
1.1
Recov-
ery
~
100%
104%
107%
113%
—
82%
* All values - yg/1
+ Spiked with 10 yg/1 each - see explanation in text.
t N - No peak observed.
-------
SECTION 7
RESULTS AND DISCUSSION
GENERAL PROCESS PERFORMANCE
The lime clarification-filtration system received secondary effluent
from the trickling filter system. The pH of the lime clarification was held
in the range 11.2-11.5, depending upon observations of effluent clarity.
The average lime dosage requirement was 414 mg Ca(OH)2/l- Neutralization
was achieved with a sulfuric acid dosage of 1.2 meq/1 and sufficient C02 to
bring the alkalinity to approximately 100 mg/1 as CaCOg. The neutralization
pH set point was normally maintained at 7.0-7.2 in order that the pH of the
breakpoint reaction would be 7.0.
Table 2 shows the effect of the lime clarification-filtration system on
the trickling filter system effluent. The TOC was reduced by 63% with a 20%
reduction in SOC. The secondary effluent was fairly low in nitrogen and was
partially nitrified. The increase in ammonia and TKN which occurred during
the lime clarification and neutralization step was the result of deliberate
addition of 3-5 mg N/l of ammonium sulfate to the neutralization tank. This
was done to ensure that the ammonia content of the waste would remain above
the lower limit of control for the breakpoint chlorination control system.
Table 2 shows a decline in nitrate-nitrogen concentration from 5.5 mg/1
to 2.3 mg/1 in the filter effluent. This decline was probably the result of
the proliferation of biological slimes which developed in the neutralization
tank, in the pipeline leading to the filter and, possibly, within the filter
itself. The slime development could also have resulted in some SOC removal.
The growths became so profuse that sloughing biomass caused severe filter
head loss buildup and very short filter runs. Therefore, twice a week, while
the system was shut down, the neutralization tank, the transfer pipeline and
the filter were chlorinated. This was done carefully to prevent contamination
of any sample or contact of the highly chlorinated neutralization tank con-
tents with the activated carbon.
The suspended solids and turbidity removal were what might be expected
from a system of this kind. There was a net increase in TDS and a net de-
crease in alkalinity as a result of the lime addition and neutralization.
Table 3 shows the operating conditions of the breakpoint chlorination
systems. As seen in Figure 4, reactor 2 was located between carbon con-
tactors 2A and 2B. Reactor 3 treated filter effluent and discharged to
carbon contactor 3A. The ammonia concentrations shown in Table 3 are
2 7
-------
derived from the laboratory analyses of the composite samples.
data shown here were taken from pilot plant operating records.
All other
TABLE 2 - PERFORMANCE OF LIME CLARIFICATION-FILTRATION
OF SECONDARY EFFLUENT
Parameter
* S -
+ L -
t F -
Average Concentration
~S*ETFt
TOC
SOC
TKN §
NH3-N §
N02-N
N03-N
SS
TDS
Turbidity, JTU
Alkalinity as CaCOg
32.9
13.8
7.7
3.8
0.5
5.5
35
504
26
152
13.8
10.8
9.5
7.7
1.1
2.6
6
649
2.4
103
12.1
10.2
9.5
7.5
1.2
2.3
2
648
1.9
107
Secondary effluent.
Neutralized lime clarifier effluent.
Granular media filter effluent.
3-5 mg NH3-N/1 added to clarifier effluent
TABLE 3 - AVERAGE OPERATING CONDITIONS OF
BREAKPOINT CHLORINATION SYSTEMS
Breakpoint system
2 3
Location
Influent NH3, mg N/l
Chlorine Dose, mgl/
C1:N ratio
Total Residual Chlorine, mg/1
Free Residual Chlorine, n£/l
Mid-carbon
8.4
71
8.4
12.1
8.5
Pre-carbon
7.5
83
11.1
13.1
8.3
28
-------
Influent and effluent ammonia concentrations and chlorine to ammonia
nitrogen ratios for the two breakpoint systems are shown in Figure 11 and
Figure 12. Figure 13 illustrates the breakpoint reactor chlorine residuals
averaged over each sample period. The nitrogen data is summarized in Table 4.
The influent to System 2 was slightly higher in ammonia concentration than
the System 3 influent. This could have been the result of hydrolysis of
organic nitrogen within the first stage carbon since the organic nitrogen in
System 2 influent averaged about 1.3 mg/1 less than that of System 3 influent.
TABLE 4 - NITROGEN REMOVAL BY BREAKPOINT CHLORINATION
(Average Values)
Breakpoint system
Influent
NH
NO
NO
3'
2>
3'
TKN,
mg
mg
mg
mg
N/l
N/l
N/l
N/l
8.
0.
1.
9.
4
2
6
1
Effluent
0.
0.
1.
1.
3
1
0
3
Influent
7.
1.
2.
9.
5
2
3
5
Effluent
0.
0.
1.
1.
3
1
6
9
The pilot system was not designed to study the breakpoint chlorination
process as such. Possible nitrogen transformations within the sampling
system and the nature of the operation (widely variable influent concentra-
tions, occasional loss of breakpoint, periods of unattended operation)
limit the interpretation of the nitrogen data. For example, both breakpoint
systems showed some removal of nitrate ion. This is most certainly erroneous
since breakpoint chlorination has been shown to produce nitrates, not destroy
them. Because the objective of this project was to document the production
and removal of halogenated organics by the processes in question, no attempts
were made to reconcile the nitrogen data. Investigations which were designed
to provide data on the nitrogen and chlorine transformations during the
breakpoint chlorination of wastewater have been conducted elsewhere (4).
The residual chlorine plots are indicators of how well the systems were
controlled during a given sample period, the greater divergence of the free
and total chlorine residual values indicating longer periods out of control.
System 2 was generally easier to control because the lag time of the ammonia
monitor approximately corresponded to the total detention time in the carbon
column, so that the sensor and the control element were functioning on the
same slug of wastewater. In System 3 the control element responded to
29
-------
z
I
I
z
u!
Z
O
*-
u
10
i 5.
12.0
10.0
8.0
6.0
G
L /V
o>
E
4.0\
n
X
Z
o
JUNE
JULY
AUG.
Figure 11. Ammonia removal system 2
30
-------
i 15
*-*-
2
O
u.
O
* CI:N
K
10
*
o>
V
en
E
CO
X
12.0
10.0
8.0
6.0
4.0
2.0
INFLUENT NH3 -N
G
+ EFFLUENT NH3 -N
A / \ X •*•
—+•
JUNE
JULY
AUG.
Figure 12. Ammonia removal - system 3,
31
-------
20
a 10
E
O 0
i
Q 30
on
LU
&L
20
10
^
/I
SYSTEM 2
Total Residual
A—A
\7\
x/
A
Free Residual
SYSTEM 3
- A/
A^XA
^-o
A
y ^A-A-S
Total Residual
Free Residual
JUNE
JULY
AU6.
Figure 13. Residual chlorine after breakpoint chlorination.
32
-------
changes in ammonia concentration 15 minutes after the changes occurred,
making control more difficult and probably resulting in part in the higher
C1:N ratio required.
Table 5 shows the quality of the final effluents from the three carbon
systems, with the filter effluent characteristics repeated for comparison.
Removal of organic carbon by the three systems shows little difference on
an average basis with average TOC of approximately 5 mg/1 in the final efflu-
ent from each carbon system. The removal of soluble organic carbon by the
GAG systems was quite erratic and showed no definable pattern of breakthrough,
other than the presence of relatively high concentrations of SOC near the end
of the study; shown in Figure 14.
TABLE 5 - FINAL EFFLUENT CHARACTERISTICS"
Effluent
Parameter
Chlorine Reactor Location
TOC
SOC
TKN
NH3-N
N02-N
N03-N
SS
TDS
Turbidity
Alkalinity as CaC03
Filter
-
12.1
10.2
9.5
7.5
1.2
2.3
3
648
1.9
107
IB
None
5.3
3.3
3.9
3.4
1.0
6.2
1
638
1.0
78
2B
Mid-carbon
4.6
3.9
1.0
0.3
0.1
2.0
2
876
1.5
116
3B
Pre-carbon
5.1
4.2
1.2
0.5
0.2
1.9
1
869
1.5
116
All values mg/1 except turbidity (JTU).
33
-------
20
10
20
0)
O 10
CO
OL
u
0
ac
O
co 20
6
to
10
20
10 -
100 200
VOLUME TREATED, cu.
300
Figure 14. SOC removal by granular activated carbon.
34
-------
The irregularity of the breakthrough data preclude use of organic carbon
as a parameter in any rational comparison of the systems. The final efflu-
ent quality of all systems was excellent, with very low organic carbon and
suspended solids. Again, the nitrate data are curious, showing an increase
in nitrate concentration in the unchlorinated carbon system. This indicates
the presence of nitrifying organisms, either in the carbon column itself or
in the sampling system. However, the first stage effluent sample from carbon
System 2 (2A) which should have been similar to the first stage effluent
sample from carbon System 1 (1A) averaged 1.2 mg N03-N/1 as compared to 6.5
mg N03-N/1 in 1A. The reason for this difference has not been determined.
The reduction in alkalinity in sample IB tends to confirm the apparent nitri-
fication of the ammonia. Alkalinity changes in the breakpoint systems were
minimal, showing only a slight increase over the filter effluent. The in-
crease in TDS was greater than expected, probably due to the somewhat un-
defined composition of the sodium hypochlorite feedstock.
FORMATION OF VOLATILE HALOGENATED ORGANICS
Breakpoint Chlorination of Granular Media Filter Effluent
Granular media filter effluent was chlorinated with an average dosage
of 83 mg chlorine/1 for ammonia oxidation in System 3, with an average free
chlorine residual of 8.3 mg/1. The volatile halogenated organic compounds,
chloroform (CHClj), bromodichloromethane (BrCHCl2), dibromochloromethane
(Br2CHCl), bromoform (CHBr3), carbon tetrachloride (CCl^, and 1,2,dichloro-
ethane (^tfyC^), were measured in composite samples of the influent and
effluent of the breakpoint reactor. The results of these measurements are
illustrated in the Appendix in Figure Al through A5 and summarized in Tables
6 and 7. The minimum concentrations reported for all samples were: BrCHCl2
and CC14 - 0.2 yg/1; CHC13, BrCHCl and CHBr3 - 0.4 yg/1 and C2H4C12 - 0.8
yg/1.
The predominant compound of those measured in the filter effluent was
chloroform which was detected in all but four of the samples. The median
value was 10 yg/1. Thus,through the sewer system, a trickling filter,
clarifier overflows, and tertiary clarification-filtration, there remained
a small but measurable residual of chloroform. Only two values were over
35 yg/1, one at 150 yg/1 and the second at 2400 yg/1. The source of the
slug of 2400 yg/1 was never determined.
Bromodichloromethane and dibromochloromethane were detected in 11 of
the 25 samples of the filter effluent, while bromoform was never detected.
Carbon tetrachloride and 1,2 dichloroethane were detected infrequently but
the maximum concentrations observed were 43 yg/1 and 44 yg/1, respectively.
Breakpoint chlorination of the filter effluent resulted in the reactor
effluent concentrations shown in the Appendix in Figures Al through A5, and
summarized in Table 7. The amounts of each compound formed are shown in
Figures A6 through All in the Appendix and summarized in Table 8. All
samples of reactor effluent contained chloroform, with an increase in con-
centration over that of the filter effluent on all days except the two days
when the filter effluent contained 150 yg/1 and 2400 yg/1. The median
35
-------
TABLE 6 - CONCENTRATION OF HALOGENATED ORGANICS IN FILTER EFFLUENT
Compound
CHC13
BrCHCl2
Br2CHCl
CHBr3
CC14
C2H4C12
Average
16+
1.1
2.1
N
3.1
4.4
Median
10
N
N
N
N
N
Maximum
2400
4.6
9.7
N
13
44
Minimum
N+
N
N
N
N
N
% of days
detected
84
44
44
0
48
32
* All concentrations yg/1.
+ Maximum value
of 2400 ug/:
L (July 2
slug) omitted from calculation
of average.
t N - not detected.
TABLE 7 - CONCENTRATION OF HALOGENATED ORGANICS AFTER
BREAKPOINT CHLORINATION OF FILTER EFFLUENT*
Compound
CHC13
BrCHCl2
Br2CHCl
CHBr3
CC14
C2H4C12
Average
80+
6.9
4.3
2.3
4.. 9
4.4
Median
60
5.3
3.9
N
2.3
N
Maximum
900
23
17
22
58
51
Minimum
20
Nt
N
N
N
N
% of days
detected
100
96
65
12
65
69
* All concentrations - yg/1.
+ Maximum value
of 900 ug/1
(July 2
slug) omitt
ed from calc
ulation
of average.
t N - not detected.
36
-------
concentration of CHCl^ in the chlorinated filter effluent was 60 yg/1 with
a range of 20 yg/1 to 900 yg/1. The 900 yg/1 value occurred on the day that
the influent concentration was 2400 yg/1. Except for the two days when the
effluent concentration was less than that of the influent, breakpoint chlori-
nation caused the formation of from 17 yg/1 to 335 yg/1 of CHC13 in the fil-
ter effluent with an overall median value of 43 yg/1 formed. This range of
CHC13 concentration is similar to that found in the 80 city survey of finish-
ed drinking water supplies (23). However, the chlorine contact time in
System 3 breakpoint reactor was only 3 min before dechlorination, while
drinking water chlorination can go on for hours within a plant and days
within a distribution system. A small number of batch chlorination tests
were made after this study was completed (41). These will be discussed in
more detail later. In two one-hour tests with frequent sampling the CHC13
concentrations after 3 min were 35% to 70% of those at 30 min, a common
chlorine contact time. In a longer duration test the CHCl^ formation poten-
tial at 48 hr was over 10 times the amount formed at 3 min. Therefore, the
filter effluent undoubtedly contained sufficient precursor compounds to form
greater quantities of chloroform than were observed, but the formation was
limited by the short chlorine contact time. Because the chlorine-nitrogen
reactions are complete in a fairly short time (1,12), there is no need for
longer contact time for ammonia oxidation. Thus the time between chlorine
addition and dechlorination can be dictated by other factors, such as the
time necessary for stable control of chlorine feed rate, the time required
for suitable destruction of pathogenic organisms, and the rate of formation
of halogenated organics.
The bromodichloromethane concentration increased after breakpoint
chlorination in all but one sample. It was formed in much lesser quantities
than chloroform, with a median effluent concentration of 5 yg/1.
Dibromochloromethane was produced in the chlorination reactor less
often and in lower concentration than BrCHCl2 The carbon tetrachloride
concentration was higher in the effluent than the influent over half the
time. The average concentration increased from 3 to 5 yg/1, hardly a sig-
nificant increase According to Morris (28) CCl^ formation is unlikely.
As is evident from Figures A9 and All and Table 8, there was essentially
no production of bromoform and 1,2-dichloroethane.
Chloroform formation did not correlate with chlorine dose, chlorine
residual, C]:N ratio, effluent pH, or influent TOC, SOC or turbidity. Of
these parameters only pH would be expected to correlate strongly, and the
pH of this system varied little from sample period to sample period. Since
chlorine residual and organic carbon are both several orders of magnitude
greater than the chloroform produced, it would be surprising if these para-
meters did correlate. There is no reason to expect the chloroform pre-
cursor fraction of the TOC, for example, to remain constant in this waste-
water. And since only a small fraction of the chloroform precursor was
reacted, in the 4 min reaction time, the lack of correlation is to be
expected.
37
-------
For an indication of formation of non-volatile chlorinated organics,
three sets of samples were assayed by the solvent extraction-microcoulo-
metric procedure. The effect of breakpoint chlorination on the amount of
these materials in the filter effluent is shown in Table 9. Quite low con-
centrations in the filter effluent were increased by 3-25 yg/1 during the
chlorination process. Thus, adding large amounts of chlorine to filter efflu-
ent did not result in large quantities of non-volatile chloro-organics being
formed. This, again, could be a result of the short time of contact before
dechlorination.
TABLE 8 - FORMATION OF HALOGENATED ORGANICS DURING
BREAKPOINT CHLORINATION OF FILTER EFFLUENT*
Compound
CHC13
BrCHCl2
Br2CHCl
CHBr3
cci4
C2H4C12
Average"1"
64+
5.7
2.3
2.4
1.9
0.2
Median"1"
43
4.4
1.0
0
0.5
0
Maximum
335
22
10
22
15
15
% of days
formed
92
96
52
12
52
16
* All concentrations - yg/1.
+ Negative values included in average and median
t Data from July 2 (extremely high influent CHClj) omitted from
calculation of average.
TABLE 9 - FORMATION OF SOLVENT-EXTRACTABLE ORGANIC CHLORINE
COMPOUNDS DURING BREAKPOINT CHLORINATION OF FILTER EFFLUENTS"
Date
Reactor
influent
Reactor
effluent
Formed
6/25/76
7/9/76
8/9/76
Avg.
3.7
6.6
2.3
4.2
6.5
11.5
27.7
15.2
2.8
4.9
25.4
11.0
All values yg organic chlorine/1,
38
-------
Breakpoint Chlorination of First-Stage Carbon Effluent
The location of System 2 breakpoint chlorination reactor was after the
first stage activated carbon column and before the second stage. Location
of the chlorination system after the first stage of carbon treatment permit-
ted comparison of the halogenated organics formed in System 2 with those of
System 3, thus allowing inferences about the efficiency of the carbon for
removing percursor organics.
Breakpoint System 2 operated with an average chlorine dose of 71 mg/1
and an average free chlorine residual of 8.4 mg/1. Of the six volatile
compounds assayed, only chloroform and bromodichloromethane were consistently
higher after chlorination as shown in Table 10.
TABLE 10 - FORMATION OF HALOGENATED ORGANICS DURING
BREAKPOINT CHLORINATION OF FIRST-STAGE CARBON EFFLUENT"
Compound Average"1"
CHC13
BrCHCl2
Br2CHCl
CHBr3
CC14
C2H4C12
14
2.9
0.7
0
0
0.4
Median"1" Maximum
12
2.5
0
0
0
0
71
10
5.8
0
5.8
5.7
% of days
formed
85
73
23
0
38
12
* All concentrations
+ Negative values are
- ug/1.
included
in the average
and median .
A comparison of Tables 10 and 8 shows that the median values of VHO
formed in System 2 totaled 14.5 ug/1 as compared to 49 yg/1, the total of
the median values of the formation of compounds in System 3. The amount
of chloroform produced by chlorinating after carbon treatment was only
25-30% of that formed by chlorinating before carbon treatment. Bromodi-
chloromethane was formed in an amount of 50-60% of that formed in System 3.
The observation that carbon was more effective in preventing the formation
of CHC13 than BrCHCl2 corroborates the findings of Love et al. (26) in the
treatment of river water. Dibromochloromethane, carbon tetrachloride and
1,2-dichloroethane concentrations occasionally increased, but occasional
decreases were also observed, resulting in little or no net formation.
Bromoform was never detected in the influent or the effluent of System 2
chlorination reactor.
39
-------
The time plot of CHC13 and BrCHCl2 formation in System 2 breakpoint
reactor shown in Figure 15 illustrates only slight, if any, increasing
tendency with time on stream. However, when the differences between the
amounts formed in System 2 and System 3 are examined, an indication of the
precursor removal by carbon can be seen. Figures 16 and 17 are time plots
of these differences. Figure 16 shows that the tendency to form less CHClj
during the first-stage carbon effluent chlorination had not been exhausted
by the end of the study. Figure 17, on the other hand, shows that by mid-
July, the difference in formation of BrCHCl2 in the two systems was minimal.
Therefore, after treating approximately 4500 bed volumes, the first stage
of carbon treatment was no longer effective in reducing the formation of
bromodichloromethane.
The extractable chlorinated organics data for System 2 breakpoint re-
actor are shown in Table 11. In all cases, lesser amounts of chloroorganics
were formed by chlorination after carbon treatment than by chlorinating
before. The August 9 sample indicated that the carbon was still removing
precursor substances at the termination of the study, assuming all other
factors equal. At that time, the first-stage carbon column was removing
very little TOC, and thus the influents to the two chlorination reactors
were nearly the same concentration. Although the carbon was exhausted in
respect to TOC, it was still effective in reducing the amount of non-
volatile chlorinated organics formed during breakpoint chlorination.
TABLE 11 - FORMATION OF SOLVENT-EXTRACTABLE ORGANIC CHLORINE
COMPOUNDS DURING BREAKPOINT CHLORINATION OF
FIRST-STAGE CARBON EFFLUENT*
Amount less
Date Influent Effluent Formed than System 3
6/25/76
7/9/76
8/9/76
N+
1.8
1.3
2.0
3.1
6.7
2.0
1.3
5.4
0.8
3.6
20.0
Avg. 1.0 3.9 2.9 8.1
* All values yg organic chlorine/1.
+ N - denotes not detected.
Batch Chlorination Tests
A limited number of batch chlorination tests were conducted after the
pilot plant study had been terminated (41). The objectives were to obtain
an indication of the effect of chlorine contact time on VHO formation and
to examine the formation of VHO at chlorine dosages, less than breakpoint,
40
-------
80
60
40
O)
=1
20
o
BrCHCI2
JUNE
JULY
AUG.
•'igure 15. CHCl^ and BrCHClj formed during chlorination of first stage carbon effluent.
-------
100
o>
UJ
^
CO
U
I
-100
-200
co -300
u
I
u
-400
Q Q r* Q
o
o
JUNE
Q O
O
JULY
AUGUST
Figure 16. Difference in formation of CHClj in system 2 § system 3.
42
-------
00
LLJ
f—
00
to
CN
U
I
U
k.
CQ
cs":
UJ
to
U
CQ
20
10
-10
-20
o
o
G
G
JUNE
JULY
AUG.
Figure 17. Difference in formation of BrCHCl2 in system 2 and system 3.
-------
commonly employed during disinfection. Table 12 shows the results of long-
term chlorine contact with a variety of effluents under a variety of con-
ditions. The effluents listed in the table are as follows: GMF--tertiary
lime-filter effluent, the same as designated in the pilot study; GAC--
effluent from Carbon System 1 after the carbon had been replaced with fresh
carbon; secondary—effluent from the trickling filter system. The results
of the breakpoint runs indicate that chlorination of the GMF effluent with
a fairly high TOG and high free residual resulted in 600 ug CHC13/1, while
the GAG runs indicate relatively small amounts formed over 40 hours. The
use of disinfection dosages resulted in quite low CHC13 formation ranging
from zero to 13 ug/1 formed. It was observed that with disinfection dosages,
nearly all the chloroform was produced early in the runs. After 0.5 to 3
hr no additional formation was observed.
A set of short-term (60-min) runs were conducted to observe better the
CKC13 formation with shorter chlorine contact times. The results of these
tests are illustrated in Figure 18. These curves are characterized by a
very rapid initial rate of formation for the first 2-10 min followed by a
general decline in the rate of formation. In one test, chlorination of
GAG effluent produced 11 yg CHC13/1 in the first two minutes and none
thereafter. From the results of Figure 18, another factor is added to help
explain the lack of correlation of CHClj formation in the pilot plant GMF
effluent and any other measured parameter. At 4 to 5 min chlorine contact
time, the rate of formation is fairly rapid; minor deviations in delivery
time of the sample from the reactor to the sample chamber or momentary
changes in reactor detention times due to upstream sampling could result in
substantial variation in the amount formed before chlorination. A rigorous
examination of the factors influencing the formation of trihalomethanes
could possibly delineate effects of pH, chlorine residual, wastewater pre-
treatment, etc. In a real-world wastewater treatment system,however, vari-
ations in precursor concentration cannot be measured or predicted instan-
taneously. Therefore, common sense and the data at hand indicate that if
breakpoint chlorination is to be practiced, chlorination of the wastewater
at the point of lowest organic concentration with the lowest manageable
free residual at neutral pH with the shortest maneageable contact time
before dechlorination should result in the minimum production of trihalo-
methanes. The application of dosages of chlorine which result in only
combined residual produce little or no trihalomethane.
GRANULAR ACTIVATED CARBON PERFORMANCE
Removal of Volatile Halogenated Organics
Chloroform was the only compound of the six VHO's that was detected
consistently and in significant quantity in the samples taken around the
activated carbon and BPC systems. Therefore, the major emphasis in the
analysis of carbon performance is on chloroform removal.
As shown in Figure 4, each GAC system is treated as three adsorption
systems. For example, GAC System 1 has a 10-min contact time adsorber
treating GMF effluent (1A), a 20-min contact time adsorber treating GMF
effluent (1A + IB, designated here 1AB), and a 10-min contact time adsorber
44
-------
TABLE 12 - BATCH CHLORINATION RESULTS
Organic carbon
Run total/ soluble
Effluent
GMF
GAG
GAG
GMF
Secondary
Secondary
GAG
GMF
GAG
Type*
B
B
B
D
D
D
D
D
D
mg/1
20/17
8.6/8.6
5.6/4.6
10/5.8
69/16
25/13
1.4/1.2
8.6/5.8
1.4/1.0
ci2
dose
mg/1
116
88
80
30
30
10
30
10
10
Init.Res
C12 +
free/ total
mg/1
36/39
25/30
8/13
1/20
1/16
0/7
1/20
0/3.5
0/7
Final Res.
ci2 t
free/total
mg/1
8/9
10/20
2/6
1/16
0/8
0/3
1/13
0/2.4
0/4
PH
7.1
7.8
7.3
8.1
7.7
7.5
7.3
7.7
7.3
CHC13
formed §
mg/1
600
30
27
13
12
5
2
N#
N
* B denotes breakpoint chlorination;
D denotes disinfection dosages applied - less than breakpoint.
+ Initial residual chlorine measured 5 min after hypochlorite addition.
t Final residual chlorine measured 40 hr after hypochlorite addition.
§ Amount formed after 40 hr contact time.
# None formed.
-------
10
10 20 30 40 50
CHLORINE CONTACT TIME, min.
60
Figure 18. Short-term batch chlorination test results.
46
-------
treating the effluent from adsorber 1A (IB). The carbon column designations
are shown in Table 13. Adsorber 2 AB cannot be considered as a single 20-
min system from a VHO removal standpoint because of the breakpoint reactor
between stages. As Table 13 indicates, the interpretation of the chloro-
form removal data for systems 1A, 1AB, 2A and 3AB is hindered because of the
extremely large influent chloroform measurement in the samples of July 2.
The effect of the large influx of CHCl^ on July 2 on GAG System 1 is shown
in Figure 19, in which the CHCl^ concentrations after 10 min and 20 min GAG
contact are shown in a time plot. After 10 min contact, the CHC13 concen-
tration on July 2 was reduced from 2400 yg/1 to 59 yg/1, while none was
detected after 20 min contact time. After July 2, the first stage effluent
was usually higher than the influent, while the second stage maintained low
effluent levels for some time. In fact, the second stage exhibited a fairly
uniform breakthrough curve, apparently unaffected by the July 2 event.
The CHC13 concentrations in the effluents from Columns 3A, 3B, 2B and
IB are shown as a function of volume of wastewater treated in Figures 20,
21, 22, and 23. These effluents were unaffected by the July 2 slug and
illustrate the shape of the breakthrough curves and the relatively good fit
of the smooth curves to the data points. After the breakthrough of CHClj
in the effluent of column 3A occurred (Figure 20), the column was kept in
service. Its capacity for removal of CHC13 was essentially exhausted, show-
ing removal only when the influent concentration rose to much higher levels
than usual. The effluent from 3A was often higher in CHCl^ than the influ-
ent after breakthrough occurred. This phenomenon was also observed in the
carbon treatment of chlorinated river water, as described earlier (26).
The four breakthrough curves are shown together in Figure 24 for com-
parison. Several superficial observations can be made from examination of
the breakthrough curves. Locating the breakpoint chlorination system in
the mid-carbon position resulted in about 20% longer running time before
carbon broke through to a concentration of 5 yg/1. Examination of the
breakthrough curves of System 3 for influence of contact time results in a
somewhat unexpected observation. Using CHClj concentration of 5 or 10 yg/1
as the exhaustion criterion, 20-min contact time allows over 3 times the
run length to exhaustion as 10-min contact time.
Love et al. found that doubling the contact time approximately doubled
the bed life (26). Other work on the breakthrough of sewage organics showed
a 20-min contactor had less than twice the service life of a 10-min contact-
or with TOG as the effluent criterion (43). Approximately 50 cu m and 165
cu m had been treated in System 3 when the effluents from the first and
second stages, respectively, reached a CHClg level of 5 yg/1. Since the
volume of the 10-min contactor was 0.041 cu m and the volume of the 20-min
contactor was 0.082 cu m, the throughputs to an exhaustion criterion of 5
yg CHC13/1 in the 10-min and 20-min contactors were 1200 and 20-00 bed vol-
umes, respectively. Carbon utilization rate can be defined as the weight
of carbon exhausted per unit volume of wastewater treated. Using a bed
bulk density of 420 ]cg/cu m, a 20-min contactor would be exhausted at a
rate of 200 g carbon/cu m wastewater treated, while the utilization rate
of a 10-min contactor would be 340 g/cu HI. This is a significant differ-
ence in performance which suggests that for a weakly sorbed substance such
47
-------
TABLE 13 - CARBON COLUMN DESIGNATIONS
Carbon column Contact time
designation (minj
1A 10
IB 10
1AB 20
2A 10
2B 10
2AB
3A 10
3B 10
3AB 20
Influent
GMF
1A
GMF
GMF
BK2
-
BK3
3A
BK3
Remarks
1
2
1
1
2
3
4
2
5
Remarks:
1. Influent to 1A, 1AB and 2A included the 2400 ug/1 value of July 2.
Capacity and breakthrough data are not useful.
2. IB, 2B, and 3B all received wastewater treated with 10-min carbon
contact. The CHC13 concentrations in their influents are variable
but within a reasonable range. Capacity and breakthrough data from
these second-stage columns are useful.
3. 2AB cannot be treated as a single 20-min contact time adsorber because
of the chlorination reactor which altered the wastewater composition
between stages.
4. 3A was exhausted before the July 2, 900 ug/1 influent concentration
was observed. The capacity and breakthrough data to the point of
exhaustion can be used.
5. 3AB was not exhausted at the time the July 2, 900 yg/1 influent
concentration was observed. Capacity and breakthrough data are
only marginally useful.
4 8
-------
100
80
60
_PJ
u
40
20
N.D
2400 M9/I /\
in Influent / *
V / 1 1st Stage (1A)
\J
JUNE | JULY I AUG.
Figure 19. CHC13 in effluents from GAG system 1.
i i
I r
J
1
50 100 ISO 200 250 300 350
VOLUME TREATED, cu. m.
Figure 20. CHC13 breakthrough, column 3A .
49
-------
50 100 150 200 250 300 350
VOLUME TREATED cu. m.
Figure 21. CHCl^ breakthrough, column 3B .
50
100 150 200 250 300 350
VOLUME TREATED cu. m.
Figure 22. CHC13 breakthrough, column 2B.
50
-------
70
60
50
40
30
20
10
I. .. * I
0 50
100 150 200 250 300 350
VOLUME TREATED, cu. m.
Figure 23. CHC1, breakthrough, column IB.
300
100 200
VOLUME TREATED, cu. m.
Figure 24. CHCl^ breakthrough curves.
51
-------
as chloroform, longer contact times may give improved utilization of the car-
bon. A probable explanation is that because more strongly adsorbed organics
are removed in the first stage, the competition for available adsorption sites
is reduced in the second stage. As will be seen later, removal of competing
organics in the first stage apparently allowed more efficient chloroform ad-
sorption in the second stage. This could be especially important in consid-
erating treatment of polluted surface waters with activated carbon for re-
moval of trihalomethanes. One approach often considered for installation of
GAG in drinking water treatment works is replacement of filter sand with GAG.
This approach usually results in empty bed contact times of only 4-8 min.
A careful examination of the effects of contact time and removal of competing
organics could result in lower GAG regeneration requirements for trihalo-
methane removal.
Figure 25 shows the pattern of chloroform breakthrough as a function of
cumulative weight of CHC1 applied to each 10-min contactor. The cumulative
applied loading was calculated from the concentrations in the influent to
each carbon column as designated in Table 13, with columns 3A and 3B treated
as separate adsorbers. The "B" columns all were second-stage columns treat-
ing wastewater that had already been treated with 10-min carbon contact.
Column 3A operated on chlorinated filter effluent. While the loading to each
column was low, (< 0.3 g CHCl3/kg carbon) all columns produced very low
effluent CHClj. The differences in performance are only apparent as the cum-
ulative loading increases and breakthrough occurs.
0.2
0.4
0.6
0.8
Figure 25.
CUMULATIVE APPLIED CHCI3 LOADING, g CHCI3 /kg Carbon
CHC13 breakthrough as a function of applied CHC13.
52
-------
Table 14 shows the capacity of the carbon for chloroform removal in each
of the 10-min columns. Carbon capacity is defined as the cumulative weight
of CHClj removed per unit weight of carbon up to some point in the service
time of the carbon. Values shown are for effluent breakthrough concentrations
of 5 yg/1 and 10 yg/1 and also for the maximum capacity observed before com-
plete exhaustion. As is evident from Figure 25, nearly all the applied chlo-
roform was removed up to the time of breakthrough. Therefore, the values in
Table 14 reflect the same pattern as Figure 25. Very little difference in
carbon capacity can be detected at a 5 yg/1 effluent level. At the 10 yg/1
effluent level, IB had removed less CHCls than 2B, while 3A and 3B had both
removed substantially greater amounts of CHC13 than 2B. At the point of zero
removal or complete exhaustion, column 3B had exhibited greater capacity for
chloroform removal than 3A.
TABLE 14 - CAPACITY OF CARBON FOR CHC13 REMOVAL*
g CHC1T removed/g carbon
j. L. cm
At 5 g CHC1 breakthrough
At 10 g/1 CHC1 / breakthrough
*J
Maximum capacity
3A
0.36
0.61
0.64
3B
0.38
0.62
0.74
2B
0.37
0.43
0.47
IB
0.32
0.35
0.35
*Ten minute contact time.
The maximum capacity values are further illustrated in Figure 26 where
they are plotted against the flow weighted average concentration of CHC13 in
the influents to each 10-min contractor up to the time of zero removal in an
equilibrium type plot. The average influent concentration is analogous to
the equilibrium concentration of solute, and the maximum capacity can be con-
sidered the capacity at equilibrium. The three second-stage columns fit a
Freundlich type equation very well:
FT max
= °-°22
r0.87
inf
[5]
where — max = maximum capacity at complete exhaustion, g CHCl^ removed
M kg carbon
and Cin£ = flow weighted average influent concentration, yg CHC1_/1.
The data point for column 3A lies below this line of best fit for the B
columns indicating less efficient adsorption by 3A. The B columns all treated
wastewater which had been treated with 10 min contact with GAC, while 3A
treated GMF effluent. It appears from Figure 26 that the CHC13 adsorption
characteristics of the three B columns were similar even though the loading
pattern was variable throughout the study and among the three columns.
Another point of interest is that column 2B influent contained an average of
12 mg/1 total chlorine residual of which 8.4 mg/1 was free chlorine residual.
53
-------
Figure 26 indicates no effect of chlorine on the CHC1- removal capacity of
column 2B.
<
o
1.0
0.8
c 0.6
o
o 0.4
x x
•* u
0.2
0.1
10
200
20 40 60 80 100
AVERAGE INFLUENT CHCI3
AT MAXIMUM CAPACITY, fig/\
Figure 26. Capacity of 10-min contactors for CHCl? removal
The less efficient removal of CHClv by column 3A is almost certainly
caused by competition of more strongly adsorbed sewage organics for adsorp-
tion sites. Removal of these sewage organics by the A columns reduces the
competition for sites in the B columns and permits more efficient removal of
chloroform in the second stage columns.
A laboratory batch equilibrium adsorption test was run using CHC13 in
carbon-treated distilled water to which minerals had been added to simulate
the salt content of sewage. (43) In this test the resulting equilibrium
expression resulted:
£=0.033 C°'84
M eq
[6]
^ - weight of CHC13 removed per unit weight of carbon,
g CHC1 /kg carbon
eq
= concentration of CHC13 in the liquid phase at equilibrium,
pg CHC13/1.
This isotherm line, if plotted on Figure 26, would lie above the B
column data with nearly the same slope. This could imply that the organics
discharged from the A columns did reduce the efficiency of the B columns for
chloroform removal when compared with a CHC1 solution free of sewage organics
54
-------
While a great deal should not be made of the comparison of the B column
data with the results of one batch test, it is encouraging that the capacity-
concentration relationship obtained from the column data taken over a period
of 9 1/2 weeks under highly variable conditions appears reasonable in the
light of the laboratory results. The system was not designed to obtain data
for comparisons between batch and column tests. If that were the objective,
numerous isotherms run on the wastewater actually being fed to the columns
and columns spiked with known, constant concentrations of CHC1 would have
been more useful.
The column data indicate that in a wastewater of a given quality, the
capacity of carbon for removal of chloroform can be directly related to the
influent concentration by a Freundlich-type expression. It is also apparent
that the presence of competing organics can greatly reduce the efficiency of
chloroform removal. For example, if column 3A had functioned as efficiently
as the three B columns for CHC13 removal (if it had conformed to Eq [5] at
its average influent concentration of 119 yg/1), its capacity would have been
1.41 g/kg, more than double its actual capacity. The data also indicate
that applying an average of 12 mg/1 total residual chlorine does not diminish
the efficiency of carbon for removing chloroform.
Data on the removal of BrCHCl2, Br^CHCl, CC14 and Q.2^4C12 are shown in
the Appendix. Bromoform was never detected in the effluent of any of the
carbon columns. BrCHCl2 was nearly always found in the influent to carbon
system 3. The first detectable BrCHCl2 in the effluent from 3A occurred after
2000 bed volumes of chlorinated filter effluent with an average concentration
of 12 yg/1 had been treated. At that time 0.06 g BrCHCl2/kg carbon had been
removed. If column 3AB is considered as a whole, consistent breakthrough was
not observed until 3500 bed volumes had been treated. These data show the
same effect of more efficient removal by the second stage than the first
stage as was evident from the chloroform data.
Carbon tetrachloride began appearing consistently in the effluents from
all three A columns when ^ 2500 bed volumes had been treated. The B columns
contained CC14 only in two to three samples throughout the project and then
in very low concentrations (< 2 yg/1) . The amounts of the other specific
organics in the carbon influents were too low and were present too in-
frequently to merit discussion.
Removal of Extractable Organic Chlorine
Table 15 gives the concentrations of extractable organic chlorine in the
composite samples taken on the three dates shown. With low influent levels
the carbon effluents in Systems 1 and 2 contained less than 2 yg/1 in all
samples. However, with an increase with time on stream System 3 showed some
evidence of breakthrough of organic chlorine in the effluent of each stage.
On the third sample set (the last day of the project) the first stage carbon
(3A) removed only 5% of its influent organic chlorine, while the final
effluent showed 80% removal. The data are very limited and supply only order
of magnitude indications of removal of extractable organic chlorine by
activated carbon. The more heavily loaded carbon system showed signs of
55
-------
organic chlorine breakthrough, although this was not as rapid as chloroform
breakthrough.
TABLE 15 - EXTRACTABLE ORGANIC CHLORINE*
Date
6/25
7/9
8/9
F 1A
3.7 N+
6.6 1.6
2.3 N
IB
N
1.1
N
2A
N
1.8
1.3
BK2
2.0
3.1
6.7
2B
N
1.7
N
BO
6.5
11.5
27.7
3A
0.2
3.8
26.5
3B
N
1.5
5.2
All values - yg organic chlorine/1.
Sample designations refer to Figure 4.
N- Not detected.
56
-------
REFERENCES
1. Saunier, B., and R. R. Selleck. Kinetics of Breakpoint Chlorination and
of Disinfection. SERL Report No. 76-2. University of California,
Berkeley, May 1976.
2. White, G. C. Handbook of Chlorination. Van Nostrand Reinhold, 1972.
3. Wei, I. W., and J. C. Morris. Dynamics of Breakpoint Chlorination.
Chemistry of Water Supply, Treatment and Distribution. A. J. Rubin, ed.,
Ann Arbor Sciences Publishers, Inc., 1975.
4. Pressley, T. A., D. F. Bishop, and S. G. Roan, Ammonia Nitrogen
Removal by Breakpoint Chlorination. Environ. Sci. § Tech., 6, 622,
(1972).
5. Isomura, Y., Comparative Measurements of Combined Residual Chlorine
Around the Breakpoint. Bull, of the Chemical Society of Japan, 40,
835, April 1967-
6. Morris, J. C. Chlorination and Dechlorination-State of the Art. Jour.
Amer. Water Works Assn., 63, 769, (1971).
7. Griffin, A. E, and N. S. Chamberlain. Some Chemical Aspects of Break-
point Chlorination. Jour. N.E. Water Works Assn., 55, 371 (1941).
8. Moore, E. W. Fundamentals of Chlorination of Sewage and Waste. Water
and Sewage Works, 98:130, (1951).
9. Moore, W. A., S. Megregian and C. C. Ruchoft. Some Chemical Aspects of
the Ammonia-Chlorine Treatment of Water. Jour. Amer. Water Works Assn.,
35:1330 (1943).
10. Pressley: T. A., D. F. Bishop, A. P. Pinto, and A. F. Cassell. Ammonia-
nitrogen Removal by Breakpoint Chlorination. U.S. Environmental
Protection Agency Report No. EPA 670/2-73-058, Sept. 1973.
11. Palin, A. T. A Study of the Chloro Derivatives of Ammonia and Related
Compounds, with Special Reference to Their Formation in the Chlorination
of Natural and Polluted Waters. Water and Water Engineering, 54:151,
54:189 and 54:248 (1950).
12. Stone, R. W. Full-Scale Demonstration of Nitrogen Removal by Breakpoint
Chlorination. U.S. Environmental Protection Agency Report No. EPA-600/2-
78-029,March 1978.
57
-------
13. Process Design Manual for Nitrogen Control. U.S. Environmental
Protection Agency Technology Transfer, Oct. 1975.
14. Barnes, R. A., P. F. Atkins, Jr., and D. A. Scherger. Ammonia Removal
in a Physical-Chemical Wastewater Treatment Process. U.S. Environmental
Protection Agency Report No. EPA-R2-72-123, Nov. 1972.
15. Atkins, P. F. Jr., D. A. Scherger, R. A. Barnes, and F. L. Evans III.
Ammonia Removal by Physical-Chemical Treatment. Jour. Water Poll.
Control Fed., 45:2372 (1973).
16. Stasiuk, W. N. Jr., C. J. Hetling, and W. W. Shuster. Nitrogen Removal
by Catalyst-aided Breakpoint Chlorination. Jour. Water Poll. Control
Fed., 46:1974 (1974).
17. Scaramelli, A. B.., and F. A. DiGiano. A Kinetic Analysis of Ammonia
Removal by Chloramine Conversion on Activated Carbon. University of
Massachusetts Report No. Env. E.53-76-2, June 1976.
18. Scaramelli, A. B., and F. A. DiGiano, Effect of Sorbed Organics on the
Efficiency of Ammonia Removal by Chloramine-carbon Surface Reactions.
Jour. Water Poll. Control Fed., 49:693 (1977).
19. Bauer- R. C., and V. L. Snoeyink. Reactions of Chloramines with Active
Carbon. Jour. Water Poll. Control Fed., 45:2290 (1973).
20. Suidan, M. T., and V. L Snoeyink. Reduction of Aqueous Free Chlorine
with Granulated Activated Carbon. University of Illinois Water
Resources Center Report UILU-WRD-75-0103, August 1975.
21. Bellar, T. A., J. J. Lichtenberg, and R. B. Kroner. The Occurrence
of Organohalides in Chlorinated Drinking Water. Jour. Amer. Water Works
Assn., 63:703 (1974).
22. Rook, J. J. Formation of Haloforms during Chlorination of Natural
Waters. Water Treatment and Examination, 23:234 (1974).
23. Symons, J. M., et al. Natural Organics Reconnaissance Survey for
Halogenated Organics. Jour. Amer. Water Works Assn., 67:634 (1975).
*
24. Bellar, T. A. and J. J. Lichtenberg. The Determination of Volatile
Organic Compounds at the yg/1 Level in Water by Gas Chromatography.
U.S. Environmental Protection Agency Report No. EPA-670/4-74-009,
November 1974.
25. Stevens, A. A., C. J. Slocum, D. R. Seeger and G. G. Robeck. Chlorination
of Organics in Drinking Water. Jour. Amer. Water Works Assn., 68:615
(1976).
58
-------
26. Love, 0. T. Jr., J. K. Carswell, R. J. Miltner, and J. M. Symons.
Treatment for the Prevention or Removal of Trihalomethanes in Drinking
Water. Appendix 3 to Interim Treatment Guide for the Control of
Chloroform and Other Trihalomethanes. Water Supply Research Division,
U.S. Environmental Protection Agency (June 1976).
27. Stevens, A. A., and J. M. Symons. Measurement of Trihalomethane and
Precursor Concentration Changes. Jour. Amer. Water Works Assn., 69:546
(1977).
28. Morris, J. C. Formation of Halogenated Organics by Chlorination of
Water Supplies. U.S. Environmental Protection Agency. Report No.
EPA-600/1-75-002, March 1975.
29. Jolley, R. L. Chlorine-containing Organic Constituents in Chlorinated
Effluents. Jour. Water Poll. Control Fed., 47:601 (1975).
30. Glaze, W. H. and J. E. Henderson IV. Formation of Organochlorine Com-
pounds from the Chlorination of a Municipal Secondary Effluent. Jour.
Water Poll. Fed., 47:2511 (1975).
31. Carlson, R. M., and R. Caple. Chemical/Biological Implications of Using
Chlorine and Ozone for Disinfection. U.S. Environmental Protection
Agency Report No. EPA-600/3-77-066 (June 1977).
32. Evans, F. L. III. Chlorination of Organic Compounds during Breakpoint
Chlorination of Wastewater. U.S. Environmental Protection Agency,
Cincinnati, OH., Unpublished Report (1972).
33. Westrick, J. J., and M. D. Cummins. Collection of Automatic Composite
Samples Without Atmospheric Exposure. U, S. Environmental Protection
Agency, Cincinnati, OH., Unpublished Report (1977).
34. Standard Methods for the Examination of Water and Wastewater. 14th ed.
Amer. Pub. Health Assn., Washington, D. C. (1975).
35. Kamphake, L. J., S. A. Hannah, and J. M. Cohen. Automated Analysis
for Nitrate by Hydrazine Reduction. Water Research, 1;205 (1962).
36. Kamphake, L. J. U. S. Environmental Protection Agency, Cincinnati, OH.,
Personal Communication.
37. Weatherburn, M. W. Phenol-Hypochlorite Reaction for Determination of
Ammonia. Anal. Chem. 39:971 (1967).
38. Stevens, A. A., and J. M. Symons. Analytical Considerations for
Halogenated Organic Removal Studies. Proceedings AWWA Water Quality
Technology Conference, AWWA, Denver, Colorado (1974).
39. Interim Report on Montgomery Simulation. U.S. Enviromental Protection
Agency, MERL In-house Report, July 1975.
59
-------
40. Palin, A. T. The Determination of Free and Combined Chlorine in Water
by Use of Diethyl-p-phenylene Diamine. Jour. Amer. Water Works Assn.,
49:873 (1957).
41. Krouskop, D. J. The Formation of Volatile Organics in Chlorinated Waste-
water. Report for CE501, University of Cincinnati, Dec. 1977.
42. Westrick, J. J., U. S. Environmental Protection Agency; Cincinnati, OH.,
Unpublished Data.
43. Cohen, J. M., R. Middendorf, and R, A, Dobbs. Adsorption of Volatile
Halogenated Hydrocarbons: Correlation with Molecular Properties.
U. S. Environmental Protection Agency, Cincinnati, OH., Unpublished
Report.
60
-------
APPENDIX
VOLATILE HALOGENATED ORGANICS DATA
TABLE A-l - CHC1,*
Date
6/4
7
9
11
14
16
18
21
23
25
28
30
7/2
5
7
9
12
14
16
19
21
23
26
28
30
8/2
6
9
Cu m
treated F
5
19
29
37
50
61
69
83
93
104
119
12?
140
157
168
178
195
206
217
232
243
253
270
281
292
308
+325
+342
26
20
14
12
9.5
5.5
10
5.4
9.3
4.8
3.3
11
2400
15
150
19
24
33
11
N
N
N
N
2.8
4.9
13
1A
N
0.5
0.9
1.5
2.3
4.1
N
5.2
4.4
2.5
4.9
6.1
59
47
51
55
75
69
25
38
21
N
28
15
12
40
IB
N
N
N
0.4
N
N
N
N
1.9
1.8
N
2.0
N
N
2.0
0.7
2.5
3.9
4.7
4.5
8.7
9.3
8.3
9.4
32
55
2A
N
N
N
N
2.1
4.0
3.1
9.4
9.3
6.0
9.5
8.5
47
43
41
63
47
4.8
28
38
23
23
17
11
15
37
BK2
2.5
7.7
5.7
20
12
3.3
23
19
13
13
13
13
75
62
39
58
86
76
40
50
31
36
29
39
41
57
2B
N
1.7
N
1.3
N
N
N
2.3
N
2.7
N
5.4
3.4
3.6
3.6
3.6
3.7
6.7
7.3
10
12
18
21
26
40
94
BK3
110
86
104
150
170
340
96
53
39
48
20
44
900
35
45
75
52
63
37
29
27
59
28
152
62
67
3A
N
1.5
1.6
4.1
6.0
17
N
22
33
46
74
83
104
67
119
77
133
131
67
76
51
53
29
21
39
70
3B
2
I
N
1
0
0
0
3
N
3
3
6
4
4
4
2
12
13
16
17
24
25
50
48
80
150
.6
.8
.9
.8
.4
.9
.6
.3
.5
.7
.4
.2
.5
.9
* All values yg/1.
+ Problems in Analysis
CHC1 data for 8/6 and 8/9 not available.
N - denotes not detected. For sample designations refer to Figure 4.
61
-------
TABLE A-2
BrCHCl-
Date
6/4
7
9
11
14
16
18
21
23
25
28
30
7/2
5
7
9
12
14
16
19
21
23
26
28
30
8/2
6
9
Cu m
treated
5
19
29
37
50
61
69
83
93
104
119
129
140
157
168
178
195
206
217
232
243
253
270
281
292
308
325
342
F
2.7
2.8
2.8
1.5
N
1.9
1.6
1.9
N
N
N
N
N
N
N
N
2.7
3.6
4.6
N
N
N
N
N
1.0
3.5
3.9
2.2
1A
N+
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
1.3
N
0.4
IB
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
0.2
N
0.6
2A
N
N
N
N
N
N
2.0
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
0.9
1.4
0.6
BK2
N
2.8
0.3
7.4
5.1
N
3.8
2.5
N
N
N
N
N
1.8
2.2
2.5
2.1
5.2
4.8
3.3
3.7
2.9
5.5
8.2
7.5
11
26
2.8
2B
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
2.5
4.2
5.5
5.6
BK3
6.0
7.0
7.3
23
19
19
6.0
6.3
4.4
5.5
3.5
5.2
2.7
3.5
N
4.5
3.5
5.7
6.9
5.1
4.8
2.9
4.7
3.9
8.7
9.0
8.4
4.2
3A
N
N
N
N
N
N
N
1.8
3.0
2.5
4.1
N
4.0
3.3
4.3
2.4
5.5
8.1
6.7
8.5
5.7
2.7
4.8
5.9
8.9 •
27
13
8.8
3B
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
2.6
N
N
N
4.9
6.0
8.7
7.5
* All values yg/1. For sample designations refer to Figure 4.
+ N denotes not detected.
62
-------
TABLE A3
Br2CHCl
Date
6/4
7
9
11
14
16
18
21
23
25
28
30
7/2
5
7
9
12
14
16
19
21
23
26
28
30
8/2
6
9
Cu m
treated
5
19
29
37
50
61
69
83
93
104
119
129
140
157
168
178
195
206
217
232
243
253
270
281
292
308
325
342
F
3.4
6.0
8.5
1.9
N
3.0
1.4
N
N
N
N
N
N
N
3.6
N
N
2.6
8.9
9.7
N
N
N
N
4.1
1.6
1.7
1.1
1A
N+
N
N
N
8
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
.4
N
N
IB
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
2A
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
.4
N
BK2
N
2.1
1.0
4.7
4.1
N
1.0
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
5.9
6.3
0.8
2B
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
0.2
0.8
0.8
BK3
4.8
5.3
9.5
12
10
9.8
5.0
5.8
N
N
3.2
4.7
N
5.7
N
2.8
2.0
N
6.1
17
N
N
N
N
N
4.0
3.1
1.0
3A
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
2.7
1.1
2.5
N
N
N
N
N
N
N
5.0
6.2
3.9
3B
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
0.4
1.2
0.8
* All
+ N -
values
denotes
yg/i.
For sample
designations
refer to
Figure
4.
not detected.
63
-------
TABLE A-4 - CHBr3
Date
6/4
7
9
11
14
16
18
21
23
25
28
30
7/2
5
7
9
12
14
16
19
21
23
26
28
30
8/2
6
9
Cu m
treated
5
19
29
37
50
61
69
83
93
104
119
129
140
157
168
178
195
206
217
232
243
253
270
281
292
308
325
342
F
N+
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
1A
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
IB
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
2A
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
BK2
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
0.8
N
2B
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
BK3
17
N
N
22
20
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
0.4
N
N
3A
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N»
N
N
N
3B
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
* All values yg/1. For sample designations refer to Figure 4.
+ N - denotes not detected.
64
-------
TABLE A-5 - CC1,
Date
6/4
7
9
11
14
16
18
21
23
25
28
30
7/2
5
7
9
12
14
16
19
21
23
26
28
30
8/2
6
9
Cu m
treated
5
19
29
37
50
61
69
83
93
104
119
129
140
157
168
178
195
206
217
232
243
253
270
281
292
308
325
342
F
1.1
N
43
N
N
N
N
N
N
3.2
1.0
9.4
N
2.2
N
2.0
2.5
4.1
6.5
N
N
N
N
1.1
N
N
N
N
1A
N+
0.5
0.9
N
N
N
N
N
N
1.9
N
2.5
4.0
3.6
2.8
2.3
3.8
4.4
3.7
4.7
2.5
2.9
4.0
3.0
2.5
3.1
N
N
IB
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
1.0
N
N
N
0.4
N
N
2A
N
N
N
N
N
N
N
N
N
1.5
9.8
2.6
N
3.6
2.7
1.4
N
N
0.9
3.7
3.3
3.9
4.0
2.8
2.1
2.7
N
N
BK2
N
0.6
0.2
N
N
N
N
N
N
1.8
2.8
2.0
2.5
2.4
2.3
2.1
N
5.8
3.6
4.8
1.8
2.0
3.0
4.0
3.6
3.3
0.6
N
2B
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
1.0
1.7
N
N
0.9
N
N
BK3
2.3
3.5
58
N
N
N
N
N
N
4.5
2.3
11
N
1.0
N
2.5
3.0
3.3
4.0
N
4.4
14
3.1
7.8
2.0
0.8
3.4
N
3A
N
N
N
N
N
N
1.3
N
2.5
3.2
4.6
4.1
N
4.0
3.8
2.0
7.0
7.2
8.6
6.6
4.1
2.5
4.7
2.0
6.2
1.0
3.0
N
3B
N
N
N
N
N
N
1.0
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
0.9
N
N
* All values yg/1. For sample designations refer to Figure 4.
+ N - denotes not detected.
65
-------
TABLE A-6 - C H Cl
Date
Cu m
treated F
6/4
7
9
11
14
16
18
21
23
25
28
30
7/2
5
7
9
12
14
16
19
21
23
26
28
30
8/2
6
9
5
19
29
37
50
61
69
83
93
104
119
129
140
157
168
178
195
206
217
232
243
253
270
281
292
308
325
342
5.4
3.2
1.1
44
8.6
6.4
7.9
26
20
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
1A
N+
N
N
N
1.6
1.1
N
1.3
1.4
9.5
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
IB
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
2A
N
N
N
N
N
BK2
N
N
N
N
N
N 2.9
N 2
2.6 2
1.6
N
N
N
N 5
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
.9
.6
N
N
N
N
.7
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
2B
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
BK3
5.2
N
N
59
9.6
8.0
6.7
21
7.0
N
N
N
7.2
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
3A
N
N
N
0.8
1.9
1.0
N
2.7
N
N
N
N
7.1
N
N
N
N
N
N
N
N
N
N
N
N*
N
N
N
3B
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
* All
+ N -
values
denote.'
yg/i.
i not df
For sampl
it.fir.t.firl -
e designations
refer
to
Figure 4.
66
-------
io,ooor
sooo -
25 20 40 60 80 95 98
PERCENT OF OBSERVATIONS ^ STATED VALUE
Figure Al. CHC1 concentrations before and after
chlorination of G'lF effluent.
67
-------
50
~\iiiiir
20
\
o>
* 10
jw
U
X
u
* 51
Effluent
1L
Influent
5 10 20 30 40 50 60 70 80 90
PERCENT OF OBSERVATIONS < STATED VALUE
Figure A2. BrCHCl2 concentrations before and after
chlorination of GMF effluent.
95 98
68
-------
o>
u
X
u
5 10 20 30 40 50 60 70 80 90 95
PERCENT OF OBSERVATIONS < STATED VALUE
98
Figure A3. Br2CHCl concentrations before and after
chlorination of GMF effluent.
69
-------
50
20
10
Effluent
_L
Influent
5 10 20 30 40 50 60 70 80 90 • 95 98
PERCENT OF OBSERVATIONS < STATED VALUE
Figure A4. CC1.4 concentrations before and after
chlorination of Gr'F effluent.
70
-------
i i i i r
1 r
50
_ 20
\
o>
o 10
^
I
-------
1000,
500
200
100
50
o>
20
u_ io
n
U
X
u
_L
_L
_L
_l
25 20 40 60 80 95
PERCENT OF OBSERVATIONS < STATED VALUE
Figure A6. CHClj formed during chlorination of GMF effluent.
12
98
-------
O)
cT
LU
5
o
LL.
U
Cfi
100
50
20
10
1
I
I
I
I
J_
25 20 40 60 80 95 98
PERCENT OF OBSERVATIONS^ STATED VALUE
Figure A7. BrCHClo formed during chlorination of GMF effluent.
73
-------
50
o>
=*.
d
UJ
O
x
u
CN
20
10
10
20 30 40 50 60 70 80 90 95 98
PERCENT OF OBSERVATIONS < STATED VALUE
Figure A8. Br2CHCl formed during chlorination of GMF effluent.
74
-------
50
20
cT 10
UJ
O
r, 5
Cfl
u
5 10 20 30 40 50 60 70 80 90
PERCENT OF OBSERVATIONS < STATED VALUE
95 98
Figure A9. CHBr formed during chlorination of GMF effluent.
75
-------
50
O)
5.
O
01
«
O
LL.
*
O
u
20
10
1
2 5 10 20 30 40 50 60 70 80 90 95 98
PERCENT OF OBSERVATIONS < STATED VALUE
Figure AID. CC14 formed during chlorination of GMF effluent.
76
-------
50
I I I I
20
CD
Q 10
LU
O
LL.
X
cs
1
2 5 10 20 30 40 50 60 70 80 90 95 98
PERCENT OF OBSERVATIONS < STATED VALUE
Figure All. CM.£l formed during chlorination of GMF effluent,
77
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-78-165
3. RECIPIENT'S ACCESSI Of* NO.
4. TITLE AND SUBTITLE
BREAKPOINT CHLORINATION/ACTIVATED CARBON TREATMENT:
EFFECT ON VOLATILE HALOGENATED ORGANICS
5. REPORT DATE
September 1978 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
James J. Westrick, Michael D. Cummins and
Jesse M. Cohen
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Wastewater Research Division, TPDB, P-CTS
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
10. PROGRAM ELEMENT NO.
BC611, SOS#5, BE-10
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory-Gin.,OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
6/76 - 1 /7R
114. SPONSORING AGENCY CODE
EPA/600714
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The production and removal of six volatile halogenated organic compounds
during treatment of tertiary clarified and filtered wastewater by breakpoint
chlorination and activated carbon was examined in a continuous flow pilot plant.
Short contact time breakpoint chlorination of filter effluent increased chloro-
form and bromodichloromethane concentrations by 50 yg/1 and 10 yg/1, respectively.
Treatment by carbon prior to chlorination reduced the formation of chloroform by
a factor of four, but was less effective in reducing the amount of bromoform
produced.
A 20-minute empty bed contact time activated carbon adsorber was much more
efficient for chloroform removal than a 10-min adsorber, probably because of
competitive effects of more strongly adsorbed organics. The removal of chloro-
form by the second stage carbon contactors conformed to an equilibrium expression
of the Freundlich type.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Chlorination - sewage treatment
Activated carbon
Halohydrocarbons
Pilot Plants
Ammonia
b.IDENTIFIERS/OPEN ENDED TERMS
Breakpoint chlorination
c. COSATI Field/Group
13/B
3. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
86
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
78
5 COVFiA.ifflT "HINTING OFFICE 1976- - 7 5 T - 140/1436
------- |