Repository Material
Permanent Collection
SIMULATION OF COMPLIANCE CHOICES FOR THE
DISINFECTION BY-PRODUCTS REGULATORY IMPACT ANALYSIS
Allen B. Gelderloos1, Gregory W. Harrington2, Douglas M. Owen1, Stig Regli3,
James K. Schaefer1, John E. Cromwell III4, Xin Zhang4
INTRODUCTION
The United States Environmental Protection Agency (USEPA) is in the process of
developing regulations designed to limit the concentrations of disinfectants and their
by-products in United States drinking water systems. The feasibility of complying with such
limits is affected by other regulations, such as the Surface Water Treatment Rule (SWTR),
which specifies minimum levels of disinfection to protect against human exposure from
pathogens. Also, despite there being microbiological standards in place, different limits for
DBFs may significantly affect changes in exposure from pathogens. In developing
regulations for disinfection by-products (DBFs), USEPA wants to ensure that drinking water
utilities can effectively provide treatment that controls concentrations of both DBFs and"
pathogenic microorganisms.
As described in a companion paper (Cromwell et. al.. 1992), the objective of
regulatory analysis is to determine the potential impacts of implementing different
regulatory options. The objective of this paper is to describe one aspect of this analysis, the
methods used to estimate the following:
• The extent to which utilities might need to change from present treatment
practices to other treatment practices in order to meet different regulatory
options under consideration;
• The national occurrence of DBF concentrations (currently only for
trihalomethanes and haloacetic acids) after the treatment changes are
implemented; and
• The national occurrence of pathogens (currently only for Giardia cysts) after
treatment changes are implemented.
Trihalomethanes (THMs) and haloacetic acids (HAAs) are targeted because they are
among the DBFs identified to date of greatest health concern and models are available that
1 - Malcolm Pirnie, Inc.
2 - University of North Carolina/Malcolm Pirnie, Inc.
3 - United States Environmental Protection Agency
4 - Wade Miller Associates
l
RECYCLED PAPER
-------
predict their formation. Giardia cysts are targeted because: 1) more data are available on
their occurrence in source waters than other pathogens; 2) they are much more resistant to
disinfection than most other waterborne pathogens; 3) changes in treatment to control for
DBFs, assuming that systems must still meet the minimal requirements of the SWTR, are
likely to result in more significant changes in exposure from Giardia than from other
waterborne pathogens identified to date and 4) dose response data are available for
estimating risk from exposure. A companion paper (Grubbs, et. al.. 1992) describes in
greater detail the rational'and some of the shortcomings in selecting Giardia as the target
organism.
This paper presents a revised version of a similar modeling approach and analysis
(Gelderloos, et. al.. 1991).
MODELING CONCEPT
A Monte Carlo simulation model is being used by USEPA to support the regulatory
impact analysis of the disinfection and disinfection by-products regulations. An overview
of the simulation model is shown in Figure 1 and discussed in the following paragraphs.
This figure focuses on the Malcolm Pirnie/USEPA WTP model (Harrington, et. al.. 199la)
which is used to predict the extent of treatment changes needed by utilities to meet the
regulation based on a prediction of the occurrence of DBF and Giardia concentrations after
the treatment changes are implemented.
Treatment Model Input
As shown in Figure 1, there are three general inputs to the treatment model:
1) process characteristics, 2) raw water quality and chemical dosages for each plant, and
3) treatment constraints. Each of these items is discussed in the following paragraphs.
The overall modelling approach assumes that there are five basic types of treatment
practiced in the United States at the present time. The five treatment categories are
assumed to be the following:
Surface Waters
• Coagulation/filtration systems;
• Precipitative softening systems; and
• Unfiltered systems (including those that will be required to filter under the
SWTR).
Groundwaters
• Unfiltered systems; and
• Precipitative softening systems.
-------
This paper focuses on the approach used for the category of surface water
coagulation/filtration. This category represents 103,000,000 people served by systems
greater than 10,000 people and 12,000,000 people served by systems less than 10,000 people
(AWWA, 1991). (The four remaining categories, depending upon the availability of data,
may be analyzed under a similar conceptual approach). The assumed process characteristics
for this category are described in the subsequent "Modeling Approach" section.
For the raw water quality and treatment data, the Monte Carlo simulation model
uses assumed statistical distributions based on data from existing plants in this category
(Letkiewicz et. al.. 1992). From these distributions, 100 input data sets were generated, with
each set representing a single, hypothetical water treatment plant and its associated water
source. For this effort, it is assumed that this large number of these hypothetical water
treatment plants and their water sources is considered to be representative of the
distribution of water sources and treatment practices in use throughout the United States.
The treatment constraints for this category are designed to simulate the
requirements of various regulations, such as the SWTR and the Lead and Copper Rule.
The taste and odor constraint helps to define the treatment practices within the plant. Each
of these constraints is described in the subsequent "Modeling Approach" section.
WTP Treatment Model and Compliance Sorting Routine
The 100 hypothetical treatment plants are individually evaluated through the
treatment model to predict plant effluent and distributed water quality and to determine
whether a system meets the treatment constraints and DBF MCLs. The modules contained
within the WTP model are briefly described in the "Modeling Approach" section, while a
complete description is provided in the WTP User's Manual (USEPA, 1992).
The approach assumes that all treatment plants initially use chlorine as the only
disinfectant. A review of existing disinfection practices within this category shows that a
known percentage of systems use disinfectants other than chlorine. The majority of such
plants have switched from chlorine to alternate disinfectants presumably because of source
water quality that compromises compliance with the current THM Rule. Because the data
from all the plants within this category are combined to produce input distributions, the
model can be evaluated for its accuracy to predict the number of plants which require
alternate disinfectants or other treatment modifications to meet the current THM Rule.
The predicted water quality of the 100 plants is evaluated through a "compliance
sorting routine". This routine sorts the treatment model output according to alternative
DBF MCLs. If a system does not meet the constraints established by the SWTR or the
MCL of interest, the system implements the lowest cost modification necessary to meet the
constraints. If the new treatment method is not sufficient to.achieve compliance with the
treatment constraints or the DBF MCL of interest, the next lowest cost modification is
added to the treatment plant. This process is repeated for each system until it complies
with the treatment constraints and the DBF MCL of interest.
RECYCLED PAPER
-------
Order of Least Cost Alternatives
For coagulation/filtration plants using chlorine as the only disinfectant, the approach
assumes the following order of alternatives, from the lowest cost alternative to the highest
cost alternative:
• Eliminate pre-chlorination, if practiced;
• Install ammonia feed system to provide chloramines as secondary
disinfectant;
• Increase coagulant dose to improve DBF precursor removal;
• Switch to ozone/chloramines disinfection strategy; and
• Add GAC or membranes.
The implementation of a chlorine/chloramine or an ozone/chloramine disinfection
strategy may be effective for complying with both the SWTR and TTHM limits. However,
these alternative disinfection strategies may not be appropriate for other DBFs. Therefore,
under some possible regulatory scenarios, the implementation of either of these two
disinfection alternatives may not be possible. Therefore, the impacts of alternative DBF
MCLs were evaluated under two different compliance approaches as shown in Figure 2.
One compliance approach assumes that alternative disinfection strategies are available while
the other compliance approach assumes that alternative disinfection strategies are not
available. Analysis of these two scenarios also allows for national cost comparisons to be
made between: 1) use of alternate disinfectants, versus 2) sole use of precursor removal
strategies, to achieve different regulatory targets. Such cost analysis is discussed in a
companion paper (Cromwell et. al.. 1992).
Interaction Between the Treatment Model and Compliance Sorting Routine
If a treatment plant in this coagulation/filtration category used pre-chlorination and
could not meet the appropriate constraints, the treatment plant would first eliminate
pre-chlorination (if practiced). If this change were sufficient to meet the appropriate
constraints, no other modifications would be necessary. If this change were not sufficient
to meet the appropriate constraints, the treatment plant would eliminate pre-chlorination
but maintain primary disinfection with chlorine following filtration, and install chloramines
for secondary disinfection. However, if chloramines were not adequate as an alternative to
meet the MCL, the treatment plant would eliminate pre-chlorination and increase its alum
dose.
If the MCL of interest is a TTHM limit of 50 fig/L, the treatment model and sorting
routine would first evaluate each system under initial conditions. As the necessary
treatment modifications are applied to each system to comply with the treatment constraints
and the TTHM limit of 50 /ig/L, the sorting routine produces distributions of calculated
DBF concentrations and Giardia concentrations. Therefore, the distribution of water quality
for the 50 /ig/L limit may contain results from systems that do not require any treatment
modifications to systems which require the most extensive treatment modifications.
-------
Risk and Cost Models
The distributions of water quality for alternative DBF MCLs and the extent of
treatment modifications are applied to the appropriate models. The cancer risk model and
national cost model are described by Cromwell et. al.. (1992) and the Giardia risk model is
described by Grubbs et. al. (1992).
BASIS FOR TREATMENT IMPROVEMENT ASSUMPTIONS
The following paragraphs briefly describe the basis for modeling each of the
treatment improvements.
Eliminate Pre-Chlorination: This step involves the removal of raw water chlorine
feed system for those systems practicing pre-chlorination. Any disinfection credit achieved
by the system in the sedimentation basin prior to this step is maintained by increasing the
size of the contact basin following filtration.
Increase Alum Dose: This step involves increasing the input alum dose by 40 mg/L.
Addition of GAC: GAC contactors with 30-minute empty bed contact time and
6-month regeneration frequency are added to the model system following filtration. In
addition, the addition of post-chlorine was relocated to a point after the GAC
Use of Criloramines: Chloramines were formed by adding ammonia at a chlorine
residual to ammonia ratio of 4:1 after primary disinfection with chlorine in the contact
basin. This ratio was used to reduce the amount of excess ammonia entering the
distribution system to reduce the potential for growth of nitrifying bacteria. This ratio was
also used to reduce the formation of dichloramine and trichloramine which lead to taste and
odor complaints. For the use of monochloramine as a secondary disinfectant, the required
chloramine concentration was calculated to meet the following constraints:
• Achieve a chloramine residual of 1.0 mg/L at the maximum residence time
in the distribution system to control the growth of nitrifying bacteria in the
distribution system; and
• Maintain a chloramine residual less than 3.0 mg/L at the first customer to
avoid concern from possible health effects (USEPA is considering proposing
an MCLG of 3 mg/L for chloramines [Orme, 1992]).
Ozone/Chloramine: In this strategy, ozone replaces chlorine as the primary
disinfectant. The location of primary disinfection is between sedimentation and filtration.
The purpose of providing filtration after ozonation is to augment biological removal of
assimilable organic carbon (AOC) within the filters. The model, however, does not assume
any reduction in chlorine demand or TOC by biological removal through filtration. Also,
chlorine demand and TTHM formation is based on TOC concentrations following
sedimentation, which is a conservative assumption for this analysis. An ozone dose of 1 mg
O3/mg TOC is assumed to meet CT requirements and the ozone demand of the natural
RECYCLED PAPER
-------
organic material (NOM) remaining after sedimentation. At this time, it is assumed that
ozone does not affect THM formation.
The secondary disinfectant for this strategy is monochloramine. Prior to the addition
of ammonia, however, a short period of free chlorine contact (one minute at average flow)
is provided following filtration for added protection against bacteria sloughing from the
biologically active filters.
MODELING APPROACH
A schematic of the baseline model treatment plant is shown in Figure 3. The
treatment process consists of a rapid mix, flocculation and sedimentation basin, followed by
filtration and a contact basin. A storage basin is provided after the contact basin; however,
the residence time of this basin is incorporated into the distribution system residence time.
Alum is added prior to the rapid mix basin. Although some of the systems in the
coagulation/filtration category use iron salts or other coagulants, the percentage is small
(approximately 12 percent). For the purpose of this analysis, the approach assumed that
the NOM removal performance of ferric coagulants is similar to alum. Based on a recent
survey, 82 percent of systems used a disinfectant prior to filtration (AWWA, 1991;
Letkiewicz et. al.. 1992;). The survey did not distinguish the precise point of
prechlorination, and in the absence of such data, the prechlorination point was assumed
to precede the rapid mix.
Caustic is also added before the rapid mix basin if the sedimentation basin pH is
predicted to be below 5.5. This pH represents the lower bound used to develop the TOC
removal equation and also helps to control residual aluminum. Post-chlorine is added
before filtration and the dose is determined by the model (see "Calculation of Chlorine
Dose and Contact Basin Size" below). Lime addition, if applicable, is also before filtration
to minimize post-precipitation problems in the contact basin. Finally, caustic is added after
the contact basin for corrosion control in the distribution system. The caustic dose is a
function of the raw water hardness (discussed in "Corrosion Constraints", below).
Figure 4 presents a list of input raw water quality parameters and treatment data
used in the model. Letkiewicz et. al.. (1992) discusses the approach used to simulate the
statistical distributions for these parameters. From these distributions, 100 sets of input
data were generated with each data set representing a unique treatment plant. One
distribution absent from this list but needed for the model is the pre-chlorine dose.
Therefore, for those systems using pre-chlorine, the dose was set by the model to achieve
a 0.2 mg/L residual at the end of the sedimentation basin.
The WTP treatment model is used to predict the total trihalomethane (TTHM) and
total haloacetic acid (THAA) formation, inactivation ratios and disinfectant residuals
associated with each hypothetical treatment plant. The following paragraphs provide an
overview of the primary equations used in the model.
Formation of TTHMs is calculated in accordance with equations presented by
Harrington, et. al. (1991b), and formation of individual THMs is calculated in accordance
with equations presented by Chowdhury and Amy (1991). These equations predict the
-------
dynamic behavior of the THM concentrations through the treatment plant and distribution
system based on water quality characteristics, such as TOC, UV-254, temperature, pH and
bromide. THM formation in the presence of chloramines is assumed to be 20 percent of
the rate of THM formation in the presence of chlorine. This 20 percent factor is based on
an estimate developed by Amy et. al. (1990).
Formation of individual HAAs is calculated in accordance with equations developed
by Haas (Patania, 1991). These equations predict the concentration of individual HAAs
(currently limited to mono-, di-, and tri- chloroacteic acid; and mono- and di- bromoacetic
acid) at a given location in the treatment plant or distribution system. The HAA equations
are based on statistical correlations between THMs and HAAs and therefore, do not
account for the cumulative effect of changing water quality parameters throughout the
treatment plant and the distribution system.
Dynamic HAA equations are currently being developed by the D/DBP Technical
Advisory Workgroup (TAW) but these equations were not used in this analysis. Additional
work will consider incorporation of these equations into this analysis. For the analysis
presented in this paper, the equations developed by Haas (Patania, 1991) are used.
Finally, chlorine decay is calculated in accordance with the equations presented by
Dharmarajah, et. al. (1991). Chloramine decay equations were developed by Malcolm
Pirnie, Inc. based on data developed by Dharmarajah, et. al.. (1991). A description of these
equations is provided in the WTP User's Manual (USEPA, 1992).
To model the compliance behavior of plants following the SWTR, Lead Rule, and
Coliform Rule, but prior to the DBP Rule, certain constraints must be applied to the model.
The constraints attempt to restrict the operation of treatment systems to a region of feasible
operation with respect to the regulations. The following paragraphs list the constraints
applied to the model.
SWTR Constraints
The SWTR constraints include disinfection requirements through the treatment plant
and in the distribution system. The modeling approach assumes that plants will operate with
a safety factor to ensure continuous compliance during abnormal conditions.
For disinfection through the treatment plant with chlorine, the chlorine dose is
calculated to achieve at least 20 percent greater inactivation than required in the SWTR.
In other words, the disinfection requirements are based on a sum of CTcak/CTnt[.i of 1.2 at
minimum temperature and peak hourly flow. Use of the ratio, CT^/CT^-a, also referred
to as the inactivation ratio, is described elsewhere (USEPA, 1991). CT^ is the calculated
CT determined from t10 values and chlorine residuals from each treatment process in the
plant. CT^-a is the required CT value for a specific temperature, pH, and chlorine residual
to achieve a specified level of inactivation. CTRe<1.d can be estimated from the following
equation (Clark and Regli, 1991):
RECYCLED PAPER
-------
-o.is
(i)
where C12 is the chlorine residual at the end of the contact time (measured as free chlorine
in mg/L), T is the temperature of the disinfected water in degrees Centigrade throughout
the contact time (within 0 to 5 degrees), pH is the pH throughout the contact time (within
the pH range of 6 to 9), N is the concentration of Giardia cysts remaining after the contact
time and N0 is the concentration of Giardia cysts prior to the contact time. For
temperatures above 5 degrees the predicted required CT is multiplied by one-half for every
10 degree increase. For example, the equation can be used to estimate the required CT at
15 degrees by applying the equation at 5 degree and multiplying the result by one-half. The
temperature used in this equation is the minimum temperature because lower temperatures
require greater disinfection. Equation 1 calculates values of CT^.j that are within 10
percent of the values listed by USEPA (1991). Therefore, the safety factor used may
actually be less than 20 percent but is at least greater than 10 percent.
The required log inactivation is given as follows:
log inactivation = -Iog10 — (2)
w
Because the model assumes a 2.5-log removal by filtration, the required level of inactivation
for a plant with a 3-log removal/inactivation requirement would be 0.5 times the safety
factor of 1.2, or 0.6. Therefore, the total log removal/inactivation would be 3.1 (2.5 plus
0.6). For an "Enhanced SWTR", the log removal/inactivation requirement depends on the
level of contamination in the source water. (The Enhanced SWTR is described in the
"Results of Enhanced SWTR Scenario" in a subsequent section). Because the amount of
removal by filtration is assumed constant, the level of required inactivation depends on the
Enhanced SWTR guidelines.
For disinfection in the distribution system, the chlorine dose is calculated to achieve
at least a 0.2 mg/L residual at a point representing the maximum residence time in the
distribution system. This constraint is beyond the minimum requirement of the SWTR that
at least a detectable residual (which may be total residual) be maintained at 95 percent of
the sampling points in the distribution system. The higher residual concentration is assumed
necessary for maintaining detectable residuals throughout most of the distribution system
and for ensuring high probability of compliance with the Coliform Rule.
Taste and Odor Constraint
To minimize chlorinous tastes and odors, the chlorine dose is calculated to maintain
a chlorine residual entering the distribution system of less than 2.0 mg/L (Krasner, et. al..
1984). Although some treatment plants currently maintain higher chlorine residuals without
-------
any taste and odor concerns, this conservative constraint is used for the purposes of this
modeling approach. Because some systems will not be able to meet the minimum residual
requirement at the end of the distribution system without exceeding the taste and odor
constraint, the modeling effort requires these systems to modify their treatment to meet
the constraint.
DBF Constraints
Without making any significant changes to treatment practices, a utility can observe
significant variability in running annual average TTHM concentrations. As shown in Figure
5, the peak running annual average TTHM concentration can be 20 percent higher than the
mean running annual average TTHM concentration. For this analysis, the other DBFs are
assumed to exhibit similar variability. To account for this variability, the percentage of
systems meeting a specific DBF MCL was based on maintaining a mean running annual
average DBF concentration of 80 percent of the specific MCL. For example, systems with
a TTHM limit of 50 /*g/L would be required to maintain a simulated TTHM average
concentration of less than 40 /ig/L (80 percent of 50 /ig/L). A similar approach was used
for each total THM and total HAA limit under consideration.
Corrosion Constraints for Lead Rule
Corrosion control is modeled in all systems by maintaining at least a minimum pH
level based on raw water total hardness values. Other expected corrosion control measures,
such as use of orthophosphate inhibitors, are assumed to not significantly affect DBF levels
and are thus not assigned any other constraint condition. For those systems with raw water
total hardness values greater than 100 mg/L, the distribution system pH was adjusted
upward, if necessary, to achieve at least 7.6. A higher pH is not considered for these
systems out of concern for precipitation of calcium carbonate. For those systems with raw
water total hardness values less than 100 mg/L, the distribution system pH was adjusted
upward, if necessary, to at least 8.0. These minimum pH values are simplifying assumptions
to address the lead and copper rule limits promulgated in May 1991.
Distribution System Chlorine Demand
The treatment model was developed to predict chlorine decay in the distribution
system based on the amount of chlorine demand present at the plant effluent. Currently,
there are no models for predicting chlorine decay specifically based upon demand exerted
by the distribution system characteristics (e.g., corrosion by-products and biofilms). In an
attempt to account for the distribution system chlorine demand, the parameters contributing
to chlorine demand (TOC and UV-254) were increased by an arbitrary percentage upon
entering the distribution system. The 20 percent increase in.TOC and UV-254 assumed in
this analysis caused an additional 10 percent of systems to exceed the chlorine constraints
listed above (chlorine residual and taste and odor constraints).
The 20 percent assumed increase in TOC and UV-254 was modeled to only affect
the chlorine decay prediction and not the predictions of DBF formation. Chlorine and
chloramine decay models for the distribution system are needed to more accurately model
this aspect of the analysis, but none are currently available.
RECYCLED PAPER
-------
Calculation of Chlorine Dose and Contact Basin Size
To simulate the post-chlorine dosages and contact basin sizes required for
compliance with the SWTR and other regulations, the model was programmed to calculate
the most efficient combination of design variables for a given system to meet all the
constraints listed above. This most efficient combination is expected to occur with the
smallest contact basin size and maximum feasible chlorine dose, because the contact basin
is expected to control cost (Harrington, 199la). To illustrate the process used, the following
example considers a conventional surface water treatment plant using a chlorine/chlorine
disinfection strategy and assumed raw water qualities.
The first constraints considered for each system are a distribution system pH for
maintaining corrosion control (see discussion above) and the SWTR requirement that a
detectable residual be present throughout the distribution system. This latter constraint was
simulated by requiring a minimum chlorine residual of 0.2 mg/L at the maximum
distribution residence time. Figure 6 shows the minimum chlorine dose required for a range
of contact times in the chlorine contact basin at average flow. This system could meet these
two constraints by operating at any point above the curve for this constraint.
The next constraint considered for each system is the SWTR requirement to achieve
primary disinfection, or CT. For each system, the objective was to achieve an inactivation
ratio of 1.2 at minimum temperature and at peak hourly flow. The chlorine dose required
to achieve this constraint is also plotted in Figure 6 (contact times at peak hour flow were
converted to average flow and a log removal of 2.5 is assumed through filtration). Again,
the selection of any chlorine dose above this curve allows the system to meet the CT
constraint.
Finally, for each system, the amount of chlorine entering the distribution system is
limited to 2.0 mg/L in an attempt to limit taste and odor concerns in the distribution
system. The chlorine dose versus contact time curve established for this constraint is shown
in Figure 6. In this case, any operation below this curve allows compliance with this
constraint. However, the onty region where this system meets all four constraints is shaded in
the figure and labeled as the "Region of Feasible Operation".
The model was programmed to determine the furthest left point in the region of
feasible operation. This point corresponds'to the minimum feasible contact basin size. The
DBF concentrations at the average distribution system and Giardia concentrations at the
plant effluent were then computed with this contact basin size and chlorine dose.
RESULTS OF SWTR SCENARIO
This section discusses the results of the analysis using the SWTR requirement of a
minimum 3-log removal/inactivation through the treatment plant. The treatment model
described in this paper generated the following results:
• Percentage of systems meeting the SWTR and existing THM Rule;
10
-------
• Percentages of systems requiring treatment modifications to meet the SWTR
and DBF goals;
• Mean TTHM and THAA concentrations at the average customer for
selected DBF goals; and
• Mean Giardia concentrations at the first customer for selected DBF goals.
Figures 7 through 10 illustrate the treatment model output under the SWTR minimum 3-log
requirement. Figures 7 and 8 indicate predicted compliance choices for systems to meet the
selected TTHM goals of 25, 50, 75, and 100 /ig/L under two different compliance
approaches: One approach assumes alternate disinfectants are not available (Figure 7), and
the other approach assumes alternate disinfectants are available (Figure 8). A similar
evaluation of the THAAs using goals of 10, 20, 30, 40, 50, and 60 /ig/L is presented in
Figures 9 and 10 for the two different compliance approaches. Note that in some cases,
especially when alternate disinfectants are not available as a treatment option, some
systems are unable to meet the target MCL even with the use of GAC. The percentage of
such systems, which would need to seek a variance, is indicated in the small upper left hand
figure below the table in Figures 7, 9, and 10.
Systems Meeting SWTR and Existing TTHM Rule
The compliance percentages associated with the TTHM MCL of 100 /ig/L, shown
in Figures 7 and 8, represent the predicted behavior of treatment systems in this category
to meet the SWTR, taste and odor, and corrosion control constraints and the existing THM
Rule. These results provide an estimate of the effects of the SWTR and Lead Rule on
existing compliance behavior. The results show that 61 percent of systems would meet all
the constraints using chlorine as the only disinfectant and without any further modifications
to existing treatment conditions. An additional 21 percent of the systems would meet all the
constraints after the elimination of pre-chlorination. For the remaining 18 percent of
systems which do not meet all the constraints, 11 percent would need to increase the alum
dose and 7 percent would need to add GAC, if alternate disinfectants are not available. If
alternate disinfectants are available, the remaining 18 percent would meet all the constraints
after switching to monochloramine as the secondary disinfectant.
According to the Water Industry Database (AWWA, 1991), 63 percent of the
coagulation/filtration systems treating surface waters currently use chlorine as the only
disinfectant, 25 percent use chloramines and 12 percent use ozone or chlorine dioxide. In
other words, 37 percent use alternate disinfectants. Assuming that all the systems currently
using alternate disinfectants are doing so to meet the existing THM Rule, the modeling
results appear to under predict the number of plants that would require alternate
disinfectants (18 percent). It is important to note, however, that the elimination of pre-
chlorination was considered prior to the use of alternate disinfectants in the modeling
analysis. In practice, some plants which pre-chlorinate or prechloraminate have switched
to chloramines to meet the 100 /ig/L MCL and continue to disinfect prior to filtration. It
is possible that some of these plants may have been able to meet the MCL by eliminating
pre-chlorination, without the use of chloramines. Additionally, the assumption that alternate
disinfectants are being solely used to meet the existing MCL may not be valid for every
system. There may be other reasons, besides THM control, which cause some plants to use
ll
RECYCLED PAPER
-------
alternate disinfectants. Therefore, the results for a TTHM MCL of 100 /zg/L may simulate
treatment behavior more closely than indicated by .the initial comparison of predicted and
actual alternative disinfection practices.
Compliance Percentages for Alternative MCLs
As expected, the percentage of systems able to meet alternative TTHM and THAA
MCLs decrease as the MCL decreases. Consequently, the percentage of systems requiring
treatment modifications increase as the MCL decreases. The compliance percentages
presented in Figures 7 through 10 provide a basis for estimating the upgrade costs for each
selected DBF MCL.
DBF Occurrence Analysis
Further examination of the modeled systems which are unable to meet all the
constraints at a TTHM MCL of 100 /xg/L shows that these systems exceed either the THM
MCL or the chlorine constraints in the distribution system (i.e., residual constraint of 0.2
mg/L at the end of the system or the taste and odor constraint of 2.0 mg/L at the first
customer). For example, of the 39 (100 minus 61) systems which are unable to meet all the
constraints with "No Further Treatment", 21 do not meet the chlorine constraints and the
remaining 18 meet the chlorine constraints, but exceed the 100 /ig/L MCL. Because of the
method used by the model to calculate post-chlorine dose and contact basin size, if the
chlorine constraints cannot be met for a particular system, the model does not predict DBF
formation for the system.
The result of the above approach is that the mean DBF concentrations at given
treatment steps are calculated using only those systems which meet the chlorine constraints.
Therefore, the mean TTHM and THAA concentrations of 55 and 24 jxg/L, respectively, for
the systems requiring no further treatment are calculated from the 79 (100 minus 21)
systems which meet the chlorine constraints. The 21 systems which do not meet the chlorine
constraints generally have poorer source water quality (higher TOC and/or UV-254) and
would most likely produce THM levels above the mean concentrations. As treatment
modifications are applied, these systems meet the chlorine constraints and are included in
the mean calculations. Therefore, evaluating the changes in the mean concentrations at
different treatment steps is somewhat misleading.
The primary focus of comparison should be on the mean concentrations after all the
treatment modifications have been applied or after all the systems meet a selected DBF
goal. The comparison of the mean values associated with different DBF MCLs gives an
indication of the average reduced DBF exposure from different DBF regulatory scenarios.
For each DBF MCL evaluated in Figures 7 to 10, these mean concentrations are
highlighted.
Microbial Occurrence Analysis
For both disinfection alternatives, the model predicts higher mean Giardia
concentrations as the MCL for a given DBF is lowered, and as treatment modifications to
improve precursor removal are added. (This trend is shown graphically on the small figures
included on Figures 7 through 10). The reason for this trend is a result of how the model
12
-------
sets the post-chlorine dose and contact basin size. As described in the "Modeling Approach"
section of this paper, the post-chlorine dose is set to meet three chlorine constraints:
distribution system residual, CT and taste and odor. The distribution system residual and
CT constraints define the minimum dose, while the taste and odor constraint defines the
maximum dose. In cases where the minimum dose is defined by the distribution system
residual constraint (high chlorine demand waters), the dose required to meet the CT
constraint is exceeded, resulting in inactivation that exceeds the minimum CT. Therefore,
a higher inactivation is calculated for waters with higher chlorine demand. As processes to
increase precursor removal are added, the chlorine demand decreases, and the minimum
dose is more often defined by the CT, not the distribution system residual. Therefore, as
the chlorine demand of the water decreases, lower CT values are calculated (approaching
the minimum requirement), thus increasing the predicted risk of microbial infection. Note
that this analysis does not quantify any benefit of disinfecting a cleaner water (with possible
improved inactivation efficiency) as a result of removing TOC or chlorine demand.
Although in practice every plant which increases precursor removal may not reduce
its chlorine dose to meet the minimum disinfection requirement, this analysis shows the
possible outcome of plants faced with lower DBF MCLs if they were only required to
minimally meet the SWTR. It is important to note that in practice many systems, some of
which have very high concentrations of Giardia cysts in the source water, currently have
much higher levels of disinfection than minimally required under the SWTR (LeChevalier
et. al. 1991). At higher DBF MCLs many of these systems would be able to maintain such
disinfection practice and still meet the DBF MCL. However, at lower DBF MCLs, many
of these systems, in the absence of regulatory constraints to prevent otherwise, might adopt
low cost DBF control technologies (e.g., shifting the point of chlorination or switching to
chloramines) and decrease the level of Giardia inactivation provided to the extent of only
minimally meeting the SWTR. Thus, greater changes in Giardia concentrations could be
expected than are predicted between the high versus low target MCL scenarios in this
modeling analysis. Also, the rates of these types of risk tradeoffs could be more significant
to the populations for individual systems than to the populations considered for the
aggregate of all systems. This issue is further discussed in the companion paper by
Cromwell et. al.. (1992). The Enhanced SWTR, discussed in the next section, eliminates the
possibility of increased risk from Giardia that might result from increasingly stringent DBF
MCLs.
RESULTS OF ENHANCED SWTR SCENARIO
Although the SWTR currently requires a minimum 3-log removal/inactivation of
Giardia to meet CT requirements, USEPA recommended that utilities achieve greater
inactivations depending on the degree of contamination within the source water (USEPA,
1991). This paper refers to this guidance as the "Enhanced SWTR". According to these
guidelines, a system should consider providing a 3-, 4-, or 5-log removal/inactivation based
on the source water Giardia concentrations to ensure that the rate of infection from Giardia
is less than 1/10,000 people per year beyond the first customer. The input distribution of
Giardia to the model, however, contains values exceeding the maximum concentrations listed
in the Guidance Manual. Therefore, an extension of the log removal/inactivation
requirements was made to include 6- and 7-logs. The following table summarizes the SWTR
guidelines including the extension of those guidelines for the purpose of this analysis:
13
RECYCLED PAPER
-------
Daily Average Giardia
Cyst Concentration/100 L
(Geometric MeanV
£ 1
1-10
10-100
100-1,000
1,000-10,000
Recommended Giardia
Removal/Inactivation
3-log
4-log
5-log
6-log
7-log
The only difference between the modeling approach for the Enhanced SWTR and
the SWTR scenario presented previously is the amount of disinfection required in the plant
depending on the raw water Giardia concentrations. All other constraints are identical to
those described in the SWTR scenario. First order kinetics are assumed in extrapolating
the CT values necessary to achieve all levels of inactivation.
Figures 11 through 14 present a summary of the treatment model output under the
Enhanced SWTR requirements. Figures 11 and 12 reflect the model predictions for the
selected TTHM goals of 25, 50, 75, and 100 /xg/L under two different compliance
approaches: One approach assumes alternate disinfectants are not available (Figure 11),
and the other approach assumes alternate disinfectants are available (Figure 12). A similar
evaluation of the THAAs using goals of 10, 20, 30, 40, 50, and 60 /zg/L is presented in
Figures 13 and 14 for the two different compliance approaches.
Compliance Percentages for Alternative MCLs
The compliance percentages for the Enhanced SWTR scenario exhibit the same
trends as the SWTR scenario: The compliance percentages decrease as the MCLs decrease
and the percentage of systems requiring treatment modifications increase as the MCLs
decrease. Of greater significance, however, is that the predicted percentages do not
dramatically decrease, if at all, when the Enhanced SWTR requirements are imposed. A
comparison of the compliance percentages predicted from the SWTR scenario and
Enhanced SWTR scenario for systems requiring no further treatment is shown in the table
below for selected TTHM MCLs:
Percentage of Simulated Treatment Plants Complying with TTHM MCL
SWTR Scenario
Enhanced SWTR Scenario
TTHM MCL (/ig/L)
100
61
61
75
49
49
50
33
32
25
17
14
1 - EPA guidelines (USEPA, 1991) are based on a geometric mean but in this modeling
analysis the arithmetic mean is used for purposes of conducting statistical analysis.
14
-------
These comparable results are observed throughout most of the Enhanced SWTR scenario.
The reason for this result is that the additional contact time within the plant required by the
Enhanced SWTR (which is assumed can be added, if needed, to meet the inactivation
target) does not significantly change the predicted concentration of TTHMs or THAAs at
the average customer. Because the average residence time in the distribution system
averages about 1.6 days, the additional one to three hours in the plant does not significantly
increase the TTHM or THAA concentration at the average customer when using chlorine
as the onfy disinfectant.
Differences in compliance percentages for the two regulatory scenarios only appears
to occur slightly at the lower DBF MCLs. At lower MCLs, the additional CT required for
chlorination to achieve the inactivation requirements of an enhanced SWTR, appears to be
significant enough to prevent some systems from meeting the DBF MCL, particularly at an
MCL of 25 /xg/L.
These results imply that requiring greater inactivation of Giardia for more highly
contaminated source waters may not significantly impact the costs of compliance. The
approach assumes that the cost for increased contact basin size will be included in the
compliance costs, but will not significantly influence the total cost. However, if systems do
not have space available to increase contact basin size, which many apparently do not, other
options might then be required (e.g., a stronger disinfectant such as ozone which could
significantly increase treatment costs).
DBF and Microbial Occurrence Analysis
For reasons discussed above, the increased contact times required for the Enhanced
SWTR do not significantly increase the TTHM or THAA concentrations at the average
customer. A comparison of the mean TTHM concentrations for the SWTR scenario and
the Enhanced SWTR scenario after all the treatment modifications have been added is
shown in the table below:
Mean TTHM Concentration After Treatment Modification
SWTR Scenario
Enhanced SWTR Scenario
TTHM MCL (/tg/L)
100
40
41
75
33
36
50
25
26
25
15
15
Similar results are obtained for THAAs. These results imply that requiring greater
inactivation of Giardia for more highly contaminated source waters will not significantly
increase the predicted cancer risk if systems as a whole are required to also meet a specific
DBF MCL.
Because an enhanced SWTR would require higher levels of disinfection for more
highly contaminated source waters, all waters would be treated to approximately the same
range of effluent Giardia levels. As a result, systems with contaminated sources would not
be allowed to sacrifice their disinfection practices to meet, the DBF goals. Currently,
15
RECYCLED PAPER
-------
however, simple-to-use practical analytical methods for quantifying the occurrence of
Giardia cysts in source waters (and ideally also distinguishing which are viable and
infectious to humans) are not available for determining the appropriate level of treatment.
Another issue is whether or not Giardia would be the appropriate target organism (rather
than, e.g., Cryptosporidium) under such a regulatory framework.
CONCLUSIONS
This paper presents the general approach used to simulate compliance behavior for
one treatment category. A similar approach may be used for one or more of the remaining
treatment categories depending upon the data that become available.
The predictive equations used in the model are continually being reviewed and
developed. The accuracy of the predictive equations depends on the ability of these
equations to accurately predict treatment performance over a wide range of treatment
conditions. Although the treatment model may not accurately predict treatment
performance on a plant-by-plant basis, the model is designed to simulate the general
treatment performance for large numbers of systems. The results presented in this paper
show how modeling can be used to predict compliance behavior.
Perhaps the most important modeling result, if the predictions are valid, is that the
universe of systems that use surface water and which practice coagulation and filtration
could use almost the same decision tree of treatment technologies (but with differences in
the level of disinfection) to meet the existing SWTR and stringent DBF MCLs as they could
to meet an enhanced SWTR and stringent DBF MCLs. This implies that the costs for
complying with these two different regulatory scenarios would not be significantly different
even though much greater microbial protection could be provided under an enhanced
SWTR, if such a rule could be implemented.
REFERENCES
G. L. Amy, J. H. Greenfield, and W. J. Cooper (1990). "Organic Halide Formation During
Winter Treatment Under Free Chlorine Versus Chloramination Conditions. In R. L. Jolly
et. al.. eds., Water Chlorination: Chemistry. Environmental Impact and Health Effects.
Vol. 6. Chelsea, MI. Lewis Publishers, pp 605-621.
American Waterworks Association (AWWA) (1991), The Water Industry Database.
J. E. Cromwell, X. Zhang, F. Letkiewicz, S. Regli, and B. Macler (1992). "Preliminary
Regulatory Impact Analysis of Potential Tradeoffs In Regulation of Disinfection By-
Products." USEPA Publications in press.
R. M. Clark and S. Regli (1991). "The Basis for Giardia CT Values in the Surface Water
Treatment Rule; Inactivation by Chlorine.". In the Guidance Manual for Compliance with
the Filtration and Disinfection Requirements for Public Water Systems Using Surface Water
Supplies; USEPA, Washington, D.C.
16
-------
Z. K. Chowdhury and G. L. Amy (1991). "Modeling Effects of Bromide Ion Concentration
on the Formation of Brominated Trihalomethanes." Proceedings of the 1991 AWWA
Annual Conference, Philadelphia, PA.
H. Dharmarajah, N. L. Patania and J. G. Jacangelo (1991). "Empirical Modeling of
Chlorine and Chloramine Residual Decay." Proceedings of the 1991 AWWA Annual
Conference, Philadelphia, PA.
A. B. Gelderloos, G. W. Harrington, J. K. Schaefer and S. Regli (1991). "Simulation of
Compliance Choices to Meet Both Microbial and Disinfection By-Product Treatment
Objectives." Proceedings of the 1991 AWWA Water Quality Technology Conference,
Orlando, FL.
W. D. Grubbs, B. Macler, and S. Regli (1992). "Modelling Giardia Occurrence and Risk."
USEPA Publications in Press.
G. W. Harrington (1991a). "Developing a Disinfection By-Product Control Strategy."
Controlling Disinfection Bv-Products: Utilities Experiences and Future Research Needs.
Proceedings of the 1991 AWWA Annual Conference, Philadelphia, PA.
G. W. Harrington, Z. K. Chowdhury and D. M. Owen (1991b). "Integrated Water
Treatment Plant Model: A Computer Model to Simulate Organics Removal and DBF
Formation." Proceedings of the 1991 AWWA Annual Conference, Philadelphia, PA.
S. W. Krasner and S. E. Barrett (1984). "Aroma and Flavor Characteristics of Free
Chlorine and Chloramines." Proceedings of the 1984 AWWA Water Quality Technology
Conference, Denver, CO.
M. C. LeChevalier, W. D. Norton and R. G. Lee (1991). Occurrence of Giardia and
Cryptosporidium spp. in Surface Water Supplies. Appl. Env. Microb. 57:2610-2616.
F. J. Letkiewicz, W. G. Grubbs, M. Lustik, J. Mosher, X. Zhang and S. Regli (1992).
"Simulation of Raw Water and Treatment Parameters In Support of the Disinfection By-
Products Regulatory Impact Analysis". USEPA Publications in Press.
J. Orme, (1992) Personal Communications.
N. L. Patania (1991). "AWWA D/DBP Database and Model Project, Contract No. 1-90",
Final Report addressed to Edward Means. August 26, 1991.
USEPA (1991). Guidance Manual for Compliance with the Filtration and Disinfection
Requirements for Public Water Systems Using Surface Water Supplies; USEPA,
Washington., D.C.
USEPA (1992). "Water Treatment Plant Simulation Program User's Manual". Developed
by Malcolm Pirnie, Inc. for USEPA. USEPA Publications in Press.
17
RECYCLED PAPER
-------
Figure 1
Overview of Treatment Model
Process Characteristics
• Surfacewater
• Coagulation/Filtration
Simulated Raw Water
Quality and Chemical
Dosages for 100
Plants
Treatment Constraints
• SWTR/ESWTR
• Corrosion Control
• Taste and Odor
TOC/UV 254
Removal Module
THM Formation
Module
Alkallnlty/pH
Module
Chlorlne/Chloramlne
Module
WTP
Treatment
Model
HAA Formation
Module
Inactlvatlon
Module
Predicted Plant Effluent
and Distributed
Water Quality
for 100 plants
Average Customer
By-Product
Concentrations
Compliance Sorting
Routine
least cost algorithm;
Treatment Vectors
of Compliance Percentages
for Alternative MCLs
Assuming:
— 20% design
factor for DBPs
Effluent (1st customer)
Glardla Concentrations
for Alternative MCLs
Cancer
Risk
Model
National
Cost
Models
Glardla
Risk
Model
-------
Figure 2
Modeled Compliance Approach
Switch to
CI2/NH2CI
\
Increase
Alum Dose
Switch to
O3/NH2CI
\
Post-SWTR
Eliminate Pre-CI2
(if applicable)
i
Add
GAC
1
Add
Nanofiltratlon
Increase
Alum Dose
Add
GAC
Add
Nanofiltratlon
-------
Figure 3
Model Treatment Plant
- Caustic (If necessary)
— Pre - Chlorine (If applicable)
r—Alum
- Ume (if applicable)
- Post - Chlorine
<+> 4> •
-A^LA- >^wAJk>OkA>o>U>t>»>w/t^>J 1 ' V LA_A>*J
Rapid Flocculatlon
Mix & Clarification
Caustic (If neccessary)
Filtration Contact Storage Distribution
Basin System
MODEL ASSUMPTIONS
SEDIMENTATION BASIN
• Theoretical Residence Time = 4.5 hours (Including Flocculation and Rapid Mix)
• t(10) at Peak Hourly Flow +• Theoretical Residence Time at Average Flow = 0.70
• t(mean) at Average Flow + Theoretical Residence Time at Average Flow = 0.90
Sedimentation basin pH restricted to values greater than 5.5 since TOC removal
equation only valid down to this pH. Caustic added to Increase raw water pH,
If necessary.
FILTRATION
• Theoretical Residence Time = 15 minutes
• 2.5 log removal of Giardia credit
• t(10) at Peak Hourly Flow +• Theoretical Residence Time at Average Flow = 0.50
- t(mean) at Average Flow + Theoretical Residence Time at Average Flow = 1.0
CONTACT BASIN
t(10) at Peak Hourly Flow + Theoretical Residence Time at Average Flow
t(mean) at Average Flow + Theoretical Residence Time at Average Flow =
Residence time calculated by model
DISTRIBUTION SYSTEM
Maximum Residence Time = 2 x (Average Residence Time)
Level of pH adjustment by caustic based on total hardness
= 0.70
0.90
-------
Figure 4
Model Inputs to Treatment Plant
Raw Water Quality Treatment Data
TOG Alum Dose
UV-254 Lime Dose
Bromide Average Dally Flow
PH Pre-CI2 (Y or N)
Alkalinity Average Distribution
Total Hardness Residence Time
Calcium Hardness
Turbidity
Average Temperature
Minimum Temperature
Ammonia
Giardia
-------
Figure 5
Running Annual Average Trihalomethanes
IUU
IT
•5)
"•^ on
c 80
,g
S; en
QJ OU
0
O
| 4°
(D
O on
15 20
.c
H
n
-^
\
— N
\
"V
/*-
/^
/• >
^
\
\
\
/
/
xx
f
'\
\
\
\
/
/
/
A
v
x\
vx^
N
\
X
Total Trihalomethanes
Chloroform
Bromodichloromethane
Dibromochloromethane
Bromoform
-------
Figure 6
Contact Basin Design
CHLORINE/CHLORINE DISINFECTION STRATEGY
10
o
0)
Minimum Chlorine DOM
Needed to Meet CT Constraint
Maximum Chlorine Dose Feasible
to Meet Taste ft Odor Constraint
Minimum Chlorine Dose Needed to Meet
Distribution System Residual Constraint
REGION OF FEASIBLE OPERATION
360 720 1,080
Theoretical Contact Time at Average Flow (mln)
1,440
-------
30-Jan-92
Figure 7
MODEL OUTPUT (surface w/o softening): SWTR W/O ALTERNATIVE DISINFECTION
Treatment Code:
1 - not requiring further treatment modification
2 - eliminate pre—chlorination
3 - eliminate pre—chlor + modify alum dose
4 — pre—chlor + alum dose + GAC
TTHM
MCL1
(P9/I)
100
100
100
100
75
75
75
75
50
50
50
50
25
25
25
25
Treatment
Code
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
% of Sys.
Ending 2
61
21
11
7
49
25
18
8
33
23
24
20
17
10
14
59
Cumulative
Percentage of Systems
< MCL3
61
82
93
99
49
74
92
99
33
56
80
99
17
27
41
95
Mean Concentrations of
By- Products at Avq.Cus. (tio/1) 4
i iHMs
55
42
40
39
55
39
35
34
55
37
29
24
55
36
26
14
THAAs
24
20
20
19
24
19
18
17
24
18
15
13
24
18
13
7
Mean Concentration of
Giardia at First Customer 4
(cysts/100 L)
0.04
0.08
0.09
0.12
0.04
0.09
0.10
0:13
0.04
0.11
0.13
0.16
0.04
0.11
0.14
0.18
Percentage of Systems Using GAC
Percentage of Systems Meeting MCL
100%
1
CO
&
0%
20 30
40 50 60 70
TTHM MCL (fjg/l)
90 100
0%
30
40 50 60 70
TTHM MCL fcjg/l)
80
90 100
Mean Concentrations of DBPs at Average Customer
501
Mean Concentration of Giardia at First Customer
I
o
»—
<
§
8
Ul
40
30
20
10
8
I
g
I
I
20 30 4O 50 60 70
TTHM MCL (fjg/1)
80
90 100
0.21
0.20
0.19
0.18
0.17!
0.16i
0.15^
0.14
0.13
0.12
0.11
0.10
Giardia at First Customer
20
30
4O SO 60 70
TTHM MCL OJg/l)
80
90
100
1 respective MCLs include a 20% safety factor
2 percent of systems installing each treatment tier
3 cumulative percent of systems able to meet MCL at each treatment tier (includes 20% safety factor for MCLs)
4 mean concentration at each treatment tier of all systems: those meeting MCL and those not meeting MCL
-------
30-Jan-92
Figure 8
MODEL OUTPUT (surface w/o softening): SWTR W/ ALTERNATIVE DISINFECTION
Treatment Code:
1 — not requiring further treatment modification
2 — eliminate pre—chlorination
3 — eliminate pre-chlor -I- add ammonia
4 — pre-chlor + ammonia + alum dose
5 — pre—chlor + ammonia + alum + ozone
6 — pre-chlor + ammonia + alum + ozone 4- GAG
TTHM
MCL1
(P9/D
100
100
100
100
100
100
75
75
75
75
75
75
50
50
50
50
50
50
25
25
25
25
25
25
Treatment
Code
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
% of Sys.
Ending2
61
21
18
0
0
0
49
25
23
3
0
0
33
23
34
9
1
0
17
10
30
27
13
3
Cumulative
Percentage of Systems
< MCL3
61
82
100
100
100
100
49
74
97
100
100
1OO
33
56
90
99
100
100
17
27
57
84
97
100
Mean Concentrations of
Bv- Products at Avg.Cus. (ng/l) 4
I iHMs
55
42
40
40
40
40
55
39
34
33
33
33
55
37
28
26
25
25
55
36
23
17
15
15
THAAs
24
20
20
20
20
2O
24
19
19
19
19
19
24
18
17
16
16
16
24
18
16
13
12
12
Mean Concentration of
Giardia at First Customer 4
(cysts/100 L)
0.04
0.08
0.12
0.12
0.12
' '0.12 ••:••; ;';.-: ,--.•.•.
0.04 .
0.09
0.13
0.13
0.13
. 0.13-.;... .,.,.':;.:;/.".
0.04
0.11
0.15
0.16
0.16
0.16
0.04
0.11
0.16
0.17
0.18
0.18
Percentage of Systems Using OZONE/GAC
Percentage of Systems Meeting MCL
9O%r
80%r
70% r
6O%r
so%r
40% i-
30% |-
20%r
10% i-
0%u
Using OZOME/GAC
20
40 SO 60 70
TTHM MCL ((/g/1)
80
90
Mean Concentrations of DBPs at Average Customer
30
40 SO 60 70
TTHM MCL fcig/l)
90
100
Mean Concentration of Giardia at First Customer
8
so
40
30
20J-
lOf
TTHMs
O
20 30
80 90 100
0.21
0.20
0.19
0.18
0.17
0.16
0.15
0.14
0.13
0.12
0.11
0.10
Giardia at First Customer
20 30
40 SO 60
TTHM MCL (/fl/1)
40 50 60 70
TTHM MCL (pg/l)
1 respective MCLs include a 20% safety factor
2 percent of systems installing each treatment tier
3 cumulative percent of systems able to meet MCL at each treatment tier (includes 20% safety factor for MCLs)
4 mean concentration at each treatment tier of all systems; those meeting MCL and those not meeting MCL
70
so
90 100
-------
3O-Jan-92
Figure 9
MODEL OUTPUT (surface w/o softening): SWTR W/O ALTERNATIVE DISINFECTION
Treatment Code:
1 - not requiring further treatment modification
2 — eliminate pre-chlor(nation
3 - eliminate pre-chlor + modify alum dose
4 - pre-chlor + alum dose + GAC
THAA
MCL '
-------
30- Jan -92
Figure 10
MODEL OUTPUT (surface w/o softening): SWTR W/ ALTERNATIVE DISINFECTION
Treatment Code:
1 — not requiring further treatment modification 4 — pre—chlor + ammonia •+• alum dose
2 — eliminate pre—chlorination 5 — pre—chlor + ammonia + alum + ozone
3 — eliminate pre—chlor + add ammonia 6 — pre—chlor + ammonia -t- alum + ozone + GAC
THAA
MCL '
(P9/I)
60
60
60
60
60
60
50
50
50
50
50
50
40
40
40
40
40
40
30
30
30
30
30
30
20
20
20
20
20
20
10
10
10
10
10
10
Treatment
Code
1
2
3
4
5
6
1
2
3
4
5
is
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
% of Sys.
Ending2
70
12
18
0
0
0
64
15
17
4
0
O
56
16
22
5
1
O
48
12
22
14
3
1
37
8
8
23
7
17
19
3
1
22
5
50
Cumulative
Percentage of Systems
< MCL3
70
82
100
100
100
100
64
79
96
100
100
100
56
72
94
99
100
1OO
48
60
82
96
99
100
37
45
53
76
83
100
19
22
23
45
50
89
Mean Concentrations of
By -Products at Ava.Cus. (ua/D '
1 IHMs
55
46
43
43
43
43
55
45
42
41
41
41
55
42
38
36
36
36
55
40
34
30
30
29
55
39
29
23
22
20
55
38
26
18
15
12
THAAs
24
21
21
21
21
- ' 21
24
20
20
20
20
20
24
19
18
18
17
17
24
19
17
15
15
15
24
18
16
13
12
10
24
18
16
12
10
6
Mean Concentration of
Giardia at First Customer 4
(cysts/100 L)
0.04
0.08
0.07
0.07
0.07
0.07 ' >.;:*,::;.
0.04
0.09
0.08
0.08
0.08
• 0.08 ••••.;,:;:.^
0.04
0.09
0.09
0.10
0.10
O.10
0.04
0.10
0.10
0.11
0.11
O.11
0.04
0.11
0.12
0.14
0.14
0_t4
0.04
0.11
0.16
0.17
0.20
0.20
Percentage of Systems Using OZONE/GAC
Percentage of Systems Meeting MCL
20 30 40 50
THAA MCL (^g/l)
Mean Concentrations of DBPs
I
30 40
THAA MCL (pg/1)
Mean Concentration of Giardia at First Customer
\
20
30 40
THAA MCL Oig/n
50
1 respective MCLs include a 20% safety factor
2 percent of systems installing each treatment tier
Giardia at First Customer
30 40
THAA MCL (pg/1)
50
3 cumulative percent of systems able to meet MCL at each treatment tier (includes 20% safety factor for MCLs)
4 mean concentration at each treatment tier of all systems; those meeting MCL and those not meeting MCL
-------
30-Jan-92
Figure 11
MODEL OUTPUT (surface w/o softening): ENHANCED SWTR W/O ALTERNATIVE DISINFECTION
Treatment Code:
1 - not requiring further treatment modification
2 — eliminate pre—chlorination
3 — eliminate pre-chlor + modify alum dose
4 - pre—chlor + alum dose + GAG
TTHM
MCL1
fog/")
100
100
100
100
75
75
75
75
50
50
50
50
25
25
25
25
Treatment
Code
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
% of Sys.
Ending 2
61
20
12
7
49
24
19
8
32
22
26
20
14
14
11
61
Cumulative
Percentage of Systems
< MCLJ
61
81
93
99
49
73
92
99
32
54
80
99
14
28
39
95
Mean Concentrations of
By- Products at Avg.Cus. (Kg/I) 4
1 iHMs
57
44
41
40
57
41
36
34
57
39
30
25
57
37
26
.14
THAAs
25
21
20
19
25
20
18
18
25
19
15
13
25
19
13
7
Mean Concentration of
Giardia at First Customer 4
(cysts/100 L)
8.0E-05
9.0E-05
9.2E-05
9.OE-05 :
8.0E-05
9.9E-05
1.0E-04
9.9E-05
8.0E-05
1.0E-04
1.1E-04
1.1E-04
8.0E-05
1.0E-04
1.1E-04
1.1E-04
Percentage of Systems Using GAC
Percentage of Systems Meeting MCL
40 50 SO 70
TTHM MCL fcjg/i)
d
Mean Concentrations of DBPs at Average Customer
40 50 60 70
TTHM MCL (yg/l)
100
Mean Concentration of Giardia at First Customer
_ 50
I 4°
O
| 30
O 20
8
i 1°
LU
0
20 30 4O SO 60 70 80 90 100
TTHM MCL djgfl)
8 '-
§• 1.4E-04
£ 1.3E-04
§ 1.2E-04
1.1E-04
8
1.0E-04
9.OE-05 '
< 8.0E-05
Giardia at First Customer
20
30
4O 50 60 70
TTHM MCL (£Q/I)
80
90 100
1 respective MCLs include a 20% safety factor
2 percent of systems installing each treatment tier
3 cumulative percent of systems able to meet MCL at each treatment tier (includes 20% safety factor for MCLs)
4 mean concentration at each treatment tier of all systems; those meeting MCL and those not meeting MCL
-------
3O-Jan-92
Figure 12
MODEL OUTPUT.(surface w/o softening): ENHANCED SWTR W/ ALTERNATIVE DISINFECTION
Treatment Code:
1 — not requiring further treatment modification
2 - eliminate pre-chlorination
3 - eliminate pre-chlor + add ammonia
4 — pre—chlor + ammonia + alum dose
5 — pre—chlor + ammonia + alum + ozone
6 — pre—chlor + ammonia -I- alum + ozone +• GAC
TTHM
MCL1
fog/1)
100
100
100
100
100
100
75
75
75
75
75
75
50
50
50
50
50
50
25
25
25
25
25
25
Treatment
Code
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
% of Sys.
Ending2
61
20
16
3
0
O
49
24
22
4
1
0
32
22
25
17
4
0
14
14
17
27
25
3
Cumulative
Percentage of Systems
< MCL3
61
81
97
100
100
. : .''•': 10O '•: :•:.:
49
73
95
99
100
100
32
54
79
96
100
100
14
28
45
72
97
100
Mean Concentrations of
By- Products at Avg.Cus. (pg/l) 4
I IHMs
57
44
43
41
41
..•.'•:>:::.-^:,4i
57
41
38
36
36
36
57
39
33
27
26
26
57
37
29
19
15
15
THAAs
25
21
22
21
21
•'. ?..::- ;&:;:.r::;".2i.
25
20
20
20
20
20
25
19
19
17
16
16
25
19
18
14
12
12
Mean Concentration of
Giardia at First Customer *
(cysts/100 L)
8.0E-05
9.0E-05
9.3E-05
9.3E-05
9.3E-05
9.3E-O5
8.0E-05
9.9E-05
1.0E-04
1.0E-04
1.0E-04
1.0E-O4
8.0E-05
1.0E-04
1.1E-04
1.1E-04
1.1E-04
1.1E-04
8.0E-05
1.0E-04
1.1E-04
1.1E-04
1.1E-04
1.1E-04^.'-::i----:::'"-r,-;:
Percentage of Systems Using OZONE/GAC
Percentage of Systems Meeting MCL
in
&
40 SO 60 70
TTHM MCL 0/grt)
9O 100
O
z
I
»
&
#
40 SO 60 70
TTHM MCL (pg/l)
100
Mean Concentrations of DBPs at Average Customer
Mean Concentration of Giardia at First Customer
s.
o
\
8
o
z
50
4O
30"
20
10-
TTHMs
20 30
80 90 100
1.5OE-04
1.4OE-04
1.3OE-O4
1.2OE-04
1.10E-O4
1.00E-04
g.OOE-OS
S.OOE-OS
Giardia at First Customer
20
30
40 SO 60 70
TTHM MCL Ofl/1)
4O 50 60 70
TTHM MCL (fjg/l)
1 respective MCLs include a 20% safety factor
2 percent of systems installing each treatment tier
3 cumulative percent of systems able to meet MCL at each treatment tier (includes 20% safety factor for MCLs)
4 mean concentration at each treatment tier of all systems; those meeting MCL and those not meeting MCL
80
90 100
-------
30-Jan-92
Figure 13
MODEL OUTPUT (surface w/o softening): ENHANCED SWTR W/O ALTERNATIVE DISINFECTION
Treatment Code:
1 - not requiring further treatment modification 3 - eliminate pre-chlor + modify alum dose
2 — eliminate pre—chlorination 4 — pre—chlor + alum dose + GAC
THAA
MCL '
(P9/I)
60
60
60
60
50
50
50
50
40
40
40
40
30
30
30
30
20
20
20
20
10
10
10
10
Treatment
Code
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
% of Sys.
Ending2
70
11
12
7
62
16
15
7
56
16
15
13
47
12
21
20
35
10
20
35
17
3
22
58
Cumulative
Percentage of Systems
< MCL5
70
81
93
99
62
78
93
99
56
72
87
99
47
59
80
99
35
45
65
98
17
20
42
84
Mean Concentrations of
By— Products at Avg.Cus. ura/n 4
l iriMs
57
48
45
44
57
46
43
42
57
44
40
37
57
42
36
31
57
40
32
24
57
39
28
17
THAAs
25
21
20
20
25
21
19
19
25
20
18
16
25
19
16
13
25
19
14
9
25
19
13
8
Mean Concentration of
Giardia at First Customer '
(cysts/100 L)
8.0E-05
9.0E-05
9.0E-05
8.8E-05
8.0E-05
9.3E-05
9.3E-05
9.2E-05
8.0E-05
9.7E-05
9.8E-05
9.7E-05
8.0E-05
1.0E-04
1.1E-04
1.1E-04
8.0E-05
1.0E-04
1.1E-04
1.1E-04
8.0E-05
1.0E-04
1.1E-04
1.1E-04
Percentage of Systems Using GAC
Percentage of Systems Meeting MCL
100*1
*
BO% (•
Us ing GAC
30 40
THAA MCL 0/9/1)
50
100%
90%
80%
70%
60%
50%
40%
30%
2O%
10%
Mean Concentrations of DBPs
10 20 30 40 50 60
THAA MCL (yg/l)
Mean Concentration of Giardia at First Customer
so
20
20
30 40
THAA MCL (pg/I)
60
-J 1.56-04
8
5 1.4E-04
1.3E-04
1.26-04
1.1E-04
1.0E-04
9.06-05
8.06-05
Giardia at First Customer
30 40
THAA MCL 0/g/l)
1 respective MCLs include a 20% safety factor
2 percent of systems installing each treatment tier
3 cumulative percent of systems able to meet MCL at each treatment tier (includes 20% safety factor for MCLs)
4 mean concentration at each treatment tier of all systems; those meeting MCL and those not meeting MCL
-------
3O-Jan-92
Figure 14
MODEL OUTPUT (surface w/o softening): ENHANCED SWTR W/ ALTERNATIVE DISINFECTION
Treatment Code:
1 — not requiring further treatment modification 4 — pre-chlor + ammonia + alum dose
2 — eliminate pre—chlorination 5 — pre—chlor + ammonia + alum + ozone
3 — eliminate pre—chlor + add ammonia 6 — pre—chlor + ammonia + alum + ozone + GAC
THAA
MCL '
(M9/I)
60
60
60
60
60
60
50
50
50
50
50
SO
40
40
40
40
40
40
30
30
30
30
30
SO
20
20
20
20
20
2O
10
10
10
10
10
10
Treatment
Code
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
*, 5
••&- 6
1
2
3
4
5
6
1
2
3
4
5
6
%of Sys.
Ending2
70
11
17
2
0
0
62
16
18
3
1
0
56
16
16
10
2
0
47
12
12
18
10
1
35
10
4
18
16
17
17
3
0
15
15
50
Cumulative
Percentage of Systems
< MCL3
70
81
98
100
100
100
62
78
96
99
100
too
56
72
88
98
100
1OO
47
59
71
89
99
100
35
45
49
67
83
100
17
20
20
35
50
89
Mean Concentrations of
Bv-Products at Ava.Cus. (uq/TI 4
TTHMs
57
48
46
45
45
45
57
46
44
43
42
42
57
44
41
38
37
37
57
42
37
32
30
30
57
40
33
26
23
21
57
39
31
21
16
12
THAAs
25
21
22
22
22
22
25
21
21
20
20
20
25
20
20
18
18
18
25
19
19
16
15
:, • 15
25
19
18
14
12
to
25
19
18
13
10
6
Mean Concentration of
Giardia at First Customer '
(cysts/1 OOL)
8.0E-05
9.0E-05
9.1E-05
9.1E-05
9.1E-05
•.IE-OS
8.0E-05
9.3E-05
9.3E-05
9.3E-05
9.3E-05
S.3E-O5 .••::
8.0E-05
9.7E-05
9.7E-05
9.8E-05
9.8E-05
9.8E-O5
8.0E-05
1.0E-04
1.1E-04
1.1E-04
1.1E-04
1.1E-O4
8.0E-05
1.0E-04
1.1E-04
1.1E-04
1.1E-04
1.1E-O4
8.0E-05
1.0E-04
1.1E-04
1.1E-04
1.1E-04
1.1E-O4
Percentage of Systems Using OZONE/GAC
Percentage of Systems Meeting MCL
20 30 40
THAA MCL (p a/I)
100%
9O%
80%
70%
60%
50%
4O%
3O%
20%
10%
0%
30 40
THAA MCL 0/9/1)
SO
Mean Concentrations of DBPs
Mean Concentration of Giardia at First Customer
20
30 40
THAA MCL 0/g/l)
SO
Giardia at Rnt Customer
30 40
THAA MCL 0/9/0
SO
1 respective MCLs include a 20% safety factor
2 percent of systems installing each treatment tier
3 cumulative percent of systems able to meet MCL at each treatment tier (includes 20% safety factor for MCLs)
4 mean concentration at each treatment tier of all systems; those meeting MCL and those not meeting MCL
------- |