y-p itipf
United States
Environmental Protection
Agency
Office of
Drinking Water
Washington D.C. 20460
EPA-520/9-79-022
Water
vvEPA
Economic Impact Analysis
Of The Promulgated
Trihalomethane Regulation
For Drinking Water
September 1979
-------
EPA-570/9-79-022
September 1979
ECONOMIC IMPACT ANALYSIS
OF THE
PROMULGATED TRIHALOMETHANE REGULATION
FOR
DRINKING WATER
by
Temple, Barker & Sloane, Inc.
33 Hayden Avenue
Lexington, Massachusetts 02173
Contract No. 68-01-4778
Project Officer
David W. Schnare
Office of Drinking Water
U.S. Environmental Protection Agency
Washington, D.C.
OFFICE OF DRINKING WATER
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C.
. -
CHicago,. 1L 60604
-------
This report has been reviewed by Temple,
Barker & Sloane, Inc. (TBS) and EPA, and
approved for publication. Approval does
not signify that the contents necessarily
reflect the views and policies of the
Environmental Protection Agency, nor
does mention of trade names or commer-
cial products constitute endorsement or
recommendation for use.
-------
CONTENTS
Page
PREFACE i
I. SUMMARY AND INTRODUCTION 1-1
II. ANALYTIC STRUCTURE AND PROCEDURE II-l
Phase I Analysis II-l
Phase II Analysis II-5
Modeling Approach 11-10
III. THE ECONOMIC IMPACT OF THE REGULATION III-l
National Cost Estimates III-l
Costs to an Individual System III-5
Monitoring Costs III-7
Summary of Demand on Supplying Industries III-8
IV. SENSITIVITY ANALYSES IV-1
Alternative Distribution of
Treatment Selection IV-1
Alternative MCLs IV-2
Alternative System Sizes Included in
Regulatory Coverage IV-4
APPENDICES
A Water Utilities Policy Testing Model
B Water Quality Data
C Individual System Treatment Costs
D References
-------
LIST OF TABLES AND EXHIBITS
Page
THM Regulation Economic Impact Analysis
Procedure (Exhibit II-l) II-2
Cumulative Percent of Population Served by and
Number of Community Water Systems (Exhibit II-2) II-4
Decision Tree for a THM Regulation of 0.10
Milligrams Per Liter MCL (Exhibit II-3) II-9
Most Probable Treatment Selection by Water
Systems Affected by MCL Regulation of THM
at 0.10 Milligrams Per Liter (Table II-l) 11-10
Summary of Total Costs for an MCL Regulation
of THM at 0.10 Milligrams Per Liter (Table III-l) III-2
Summary of Costs by Treatment Category for an
MCL Regulation of THM at 0.10 Milligrams Per Liter
(Table III-2) III-3
Economic Impact of an MCL Regulation of
THM at 0.10 Milligrams Per Liter: Comparison with
the 1977 THM Economic Impact Report (Table III-3) III-3
Number of Systems Choosing Each Treatment
for an MCL Regulation of 0.10 Milligrams Per
Liter: Comparison with 1977 THM Economic
Impact Report (Table III-4) III-4
Compliance Costs for a Typical Water System
Under an MCL Regulation of THM at 0.10 Milli-
grams Per Liter (Table III-5) III-6
Materials Requirements for Proposed THM
Regulation (Table III-6) 111-12
Sensitivity of Costs to Mix of Compliance Treat-
ments for an MCL of THM at 0.10 Milligrams Per
Liter (Table IV-1) IV-2
Summary of Total Costs Under Alternative
MCLs for THM (Table IV-2) IV-3
-------
LIST OF TABLES AND EXHIBITS
(continued)
Page
Summary of Treatment Selections by Systems
Exceeding Alternative MCLs for THM (Table IV-3) IV-4
Costs of Alternative Size Limitations for an MCL
of THM at 0.10 Milligrams Per Liter (Table IV-4) IV-5
Financial Model of Community Water Systems:
Summary Report (Exhibit A-l) A-7
THM Concentration Based on NOMS Data (Exhibit B-l) B-2
Distribution of Water Systems by Water Source,
Disinfection Practice, and THM Level (Exhibit B-2) B-3
Costs for a Typical Water System for Selected
Treatments (Exhibit C-l) C-2
Cost Summary for An Example Water Treatment Process
(Exhibit C-2) C-3
-------
PREFACE
This report has been submitted to the United States
Environmental Protection Agency in partial fulfillment of
Contract Number 68-01-4778 by Temple, Barker & Sloane, Inc.
This report supercedes the report "Economic Impact Analysis of
a Trihalomethane Regulation for Drinking Water," submitted
August 1977. The current version has been prepared in support
of the promulgation of a trihalomethane regulation.
TBS appreciates the contributions to this effort made by
several members of EPA's Office of Drinking Water, including
David Schnare, Joseph Cotruro, Arnold Kuzmak, Craig Vogt and
Victor Kimm. Also appreciated is the assistance provided by
staff members of EPA's Municipal Environmental Research Labora-
tory on technical issues related to the methods and costs of
complying with a THM regulation. These individuals include Jim
Symons, Tom Love, Bob Clark and Gordon Robeck. Finally, the
consulting firm of Culp/Wesner/Gulp, and especially Bob Gumerman,
were particularly helpful in providing information on the use
of their database containing individual system treatment costs.
-------
I. SUMMARY AND INTRODUCTION
SUMMARY
This report presents the economic impact of the promulgated
regulations limiting trihalomethane(s) (THM) in drinking water.
The regulation applies to the estimated 2,685 community water
systems serving more than 10,000 people. Some 167 million people
are covered by the regulation, approximately 80 percent of the
public served by community water systems.
Based on preliminary monitoring, some 515 (of these 2,685
water systems covered) are expected to exceed the promulgated
maximum contaminant level (MCL) of 0.10 milligrams per liter of
THM. Based on an estimate of which treatments will be used by
these utilities, total one-time capital expenditures of $85
million (1980 dollars) will be required. The combination of
increased operating and maintenance costs ($10 million) and the
annualized capital costs will require utilities exceeding the
MCL to increase annual revenues by $19 million. This represents
an average annual cost of $.70 per person served by those systems
which exceed the MCL. For those same systems, the average
increase in a typical residential customer bill for annual
water service is projected to be $1.40.
Since the costs of the treatments which utilities may select
range broadly, some residential customer bills will increase
by several times this average, while others will not experience
any increase. For customers served by very large utilities,
annual residential bills will increase by $.00, $.30, $.90,
$2.10, $3.60, or $4.80 depending on the selection of treatments.
In utilities serving 10,000 to 25,000 people, residential bill
increases for these same treatments will range from $.00 to
$12.00 per year.
INTRODUCTION
This report presents the economic impact of the promulgated
regulations limiting THM in drinking water. A prior report,
Economic Impact Analysis of a Trihalomethane Regulation for
Drinking Water (hereafter referred t~o as the 1977 THM Economic
Impact Report; see Appendix D for full citation), was published
in August 1977 based on regulations which were proposed early
in 1978. The present report incorporates several revisions
-------
1-2
since the previous study and also shows-the economic effects
of changes in the regulation itself from the proposal to the
promulgation.
The major changes in the regulation which affect costs are
listed below and each has been incorporated in the analysis:
• Coverage of smaller systems, those serving be-
tween 10,000 and 75,000 people, which were
formerly excluded,
• Reduced monitoring frequency for groundwater
systems, and
• Relaxation of limits on the use of alternate
disinfectants.
In addition, several other inputs to the analysis have
been refined in this report. These include:
• Updated engineering cost estimates for each
treatment from the August 1978 EPA study by
the consulting firm of Culp/Wesner/Culp (C/W/C),
entitled "Estimating Costs for Water Treatment
as .a Function of Size and Treatment Plant Effi-
ciency" (hereafter referred to as the EPA Unit
Treatment Cost Report),
• Revised sets of compliance choices by water
systems exceeding the MCL based upon recent
experience by utilities in the control of THM.
The resultant estimates show fewer systems
selecting adsorbents and more choosing alter-
nate disinfectants, and
• More refined estimates of the present op-
erating and financial characteristics of
water systems.
In combination, these modifications have substantially
lowered the estimate of the national economic impact of this
regulation. The balance of this report presents these revised
estimates in the three chapters and four appendices which
follow:
• Chapter II, entitled "Analytic Structure and
Procedure," describes the seven components of
the development of a national cost estimate.
Phase I of the analysis, which consists of the
-------
1-3
first three elements below, determines the num-
ber and characteristics of systems exceeding
the MCL. Phase II builds on these results
and includes the remaining four points below.
These seven components are shown in Exhibit II-l
and are listed below:
—The regulatory criteria. These are the
parameters defined by the regulation;
they determine which water systems are
covered.
—The number and characteristics of com-
munity water systems and the populations
they serve. Those systems exceeding the
MCL are divided into several size cate-
gories for analytic and presentation
purposes.
—The water quality data analysis based on
the National Organics Monitoring Survey
(NOMS). This data is used to estimate
the number of systems exceeding and the
extent to which they exceed the MCL.
—Available treatment alternatives. These
are the treatments and procedures which
water systems could implement to comply
with the regulation.
—The derivation of selected unit treat-
ment costs for individual systems using
the information from the EPA Unit Treat-
ment Cost Report.
—Estimates of which treatment strategies
utilities will adopt based on available
treatments, cost of treatment, the degree
to which the MCL is exceeded, and exist-
ing treatment practices.
—The above estimates result in a profile
of system responses and are presented in
the form of a decision tree. This deci-
sion tree is a principal input to the
Water Utilities Policy Testing model
(PTm) computer analysis which generates
the national costs of the regulation.
-------
1-4
• Chapter III, which deals with the economic
impacts of the regulations, presents:
—National costs of the regulation. Based
on all of the elements just noted, this
section presents estimates of the costs
of compliance with the regulation at the
national level for all systems affected
by the regulation, including customer
impacts and energy consumption.
—Costs for a typical system for each alter-
native treatment. The additional capital
and operating expenses required for each
treatment are presented on a per system
basis, including the impact on consumers.
—Costs of the monitoring requirement, in-
cluding specification of the costs for
systems whose THM concentrations fall
below the MCL.
—The changes in national costs since the
1977 THM Economic Impact Report.
—The availability of the materials and
.equipment required for adding the neces-
sary treatments.
• Finally, Chapter IV presents the sensitivity
analyses of the national costs, illustrating
the range of compliance costs possible as a
result of this regulation. These analyses
were conducted to compare the above costs with
costs of:
—Alternative assumptions of treatment
selections (decision trees).
--Alternative MCLs.
—Alternative size cut-offs for water sys-
tems covered by the regulation.
Finally, this document includes four brief appendices.
The first describes the Water Utilities Policy Testing model
(PTm) which has been used to develop the national cost esti-
mates. Water quality data from NOMS are presented in Appen-
dix B. The third appendix describes the model used for deriv-
ing per system costs for the various treatments as well as
the resultant unit cost data. The final appendix includes
references to the key documents used in carrying out this
analysis.
-------
II. ANALYTIC STRUCTURE AND PROCEDURE
This chapter identifies the basic information which was
required and the manner in which it was used to develop the
national cost estimates of the proposed THM regulation. A
schematic diagram of the analytic procedure is presented in
Exhibit II-l on the following page. Phase I of the analysis
deals primarily with estimating the number of systems exceed-
ing and the extent to which they exceed the MCL. The second
phase of the analysis considers the available methods of com-
pliance and the associated costs.
PHASE I ANALYSIS
Regulatory Criteria
Naturally occurring organics have become a regulatory con-
cern primarily because of the evidence that chlorine combines
with precursor organic matter in water to form chloroform and
other related, compounds, some of which are suspected carcinogens
The regulation to reduce the level of these contaminants in
drinking water contains the following parameters:
• Maximum Contaminant Level (MCL): 0.10 milli-
grams per liter of total trihalomethanes (THM)
(chloroform, bromoform, etc.)
—applicability: community water systems
that add disinfectant to the treatment
process.
—schedule for implementation: systems serving
populations greater than 75,000—two years
after promulgation; systems serving populations
between 10,000 and 75,000—four years after
promulgation; and systems serving less than
10,000 people—at state discretion.
• Monitoring requirements: Running annual average
of quarterly samples, four samples per quarter
taken on same day for surface systems and one
sample per quarter for ground systems.
-------
Exhibit 11-1
THM REGULATION ECONOMIC IMPACT ANALYSIS PROCEDURE
REGULATORY CRITERIA
NUMBERS & CHARACTERISTICS
OF COMMUNITY WATER
SYSTEMS
WATER QUALITY DATA (NOMS)
PHASE | ANAUYSIS
NUMBERS
AND CHARACTERISTICS
OF SYSTEMS
AFFECTED
BY THE THM
REGULATION
I
I
L
-*-
AVAILABLE TREATMENTS
TREATMENT COSTS FOR
INDIVIDUAL SYSTEMS
ESTIMATES OF TREATMENT
SELECTION BY INDIVIDUAL SYSTEMS
DECISION TREE
PHASi II
THESE ARE
THE INPUTS TO PTm.
NATIONAL COSTS
ARE DETERMINED
IN THE MODEL
ESSENTIALLY
BY SUMMING
THE COSTS
OF EACH
SYSTEM'S TREATMENT
SELECTION
-------
II-3
Numbers and Characteristics
of Community Water Systems
This analysis, and the associated THM regulation, en-
compasses all water systems serving more than 10,000 people.
These systems represent 5 percent of the total number of
community water systems and serve 79 percent of the total
population that receives water from community water systems.
To provide a perspective, Exhibit II-2 illustrates the per-
centage of water systems in each size category and the
related portion of the population which they serve.
The systems serving over 10,000 have been subdivided into
six size groupings: 10,000 to 25,000; 25,000 to 50,000; 50,000
to 75,000; 75,000 to 100,000; 100,000 to 1 million; and over
1 million. In all, nine size categories were used for estimating
national costs under varying assumptions. These size categories
permit the cost analysis to reflect such differences among sys-
tems as the economies of scale associated with the sizing of
equipment for new treatment processes. The economic analysis
is, therefore, conducted on the basis of average system charac-
teristics. While economies of scale can be factored in, site-
specific costs attributable to unusual or unique circumstances
at any particular utility are not reflected in the results.
Therefore, lacking site-specific knowledge about exactly which
water systems will be affected and what treatments they would
use, these size categories allow the best estimation possible.
Water Quality Data Analysis
National Organic Monitoring Surveys of organic contaminants
in drinking water have been conducted during 1976 and 1977 by
EPA' s Municipal Environmental Research Laboratory (MERL) and
the Office of Water Supply, Technical Support Division Labora-
tory. The information from those surveys, while not completely
representative of the industry, has been used to estimate the
type and degree of water supply contamination by organic chemi-
cals across the country. Consequently, these estimates were
used to determine the proportion of water systems likely to
exceed specified maximum contaminant levels for THM and there-
fore likely to require additional treatments. Additional de-
tail on the NOMS data is included as Appendix B.
By combining the NOMS database on water quality with the
analysis of the number and characteristics of community water
systems and the regulatory criteria, the number of systems
exceeding the MCL and the extent to which they exceed it was
-------
II-4
Exhibit 11-2
CUMULATIVE
PERCENT
SERVED AT
AND ABOVE
ANY POPULATION in
CUT-OFF
CUMULATIVE PERCENT OF POPULATION SERVED
BY COMMUNITY WATER SYSTEMS
0 10,000 25,000 50,000 75,000 100,000
SYSTEM SIZE (POPULATION SERVED)
1 MILLION
CATEGORIES OF
POPULATION
SERVED
10,000
10,000-25.000
25,000-50,000
50,000-75,000
75,000-100,000
100,000-1 Mil.
Over 1 Mil.
COMPARATIVE DISTRIBUTION OF NUMBERS OF SYSTEMS AND
POPULATION SERVED BY SIZE OF SYSTEM
NUMBERS OF SYSTEMS
95.4
.33
.62
.25
.39
.02
0 10 20 30 40 50 SO 70 80 90 100
Note: A total of 58,768 community
wmr systems as of 1981 arc used
a a bas« for these percentages.
POPULATION
21.4%
11.6%
8.5%
10.8%
6.4%
12.6%
28.7%
10
20 30
PERCENT
40
50
Note: Total population served by community water
systems equals-213 million in 1981.
-------
II-5
determined. The result, then, was a conapi-lation of the number
of systems according to: (1) the water source, (2) the extent
to which the MCL is exceeded, and (3) the population category.
The second phase of the analytic procedure, discussed
below, relates to the analysis of treatment alternatives and
compliance strategies likely to be selected by those systems
exceeding the MCL.
PHASE II ANALYSIS
Available Treatment Alternatives
There are three general categories of treatment possibili-
ties. The selection of the appropriate category for a specific
water system depends in part upon the magnitude of the system's
THM level, the system's existing treatment practices, and the
costs associated with the treatment alternative. Systems are
expected to select the alternative which is lowest in cost and
least disruptive to their current practices, and which will
bring them into compliance with the regulation.
The major treatment options which are available to meet a
THM regulation are described below:
• The first alternative consists of minor modifi-
cations to current procedures. These modifica-
tions include moving the point of disinfection,
adjusting the chlorine dosage, or improving
existing conventional coagulation and sedimen-
tation practices. This approach would enable
systems which exceed the MCL only by a small
amount to comply at minimal cost.
• The second alternative involves changing disin-
fectants. Since it is the use of chlorine
which causes the formation of THM, some
systems may choose to use other chemicals for
disinfection. The alternatives considered are:
chloramines, ozone, and chlorine dioxide.
• The third alternative, using an adsorbent,
is the most complex and costly of the op-
tions. Systems with the most serious organic
-------
II-6
contamination may select treatment techniques
which require the use of granular activated
carbon (GAC), resins, or an equivalent. This
analysis has used the costs of replacing exist-
ing filter media with GAC since this technique
is the treatment method most likely to be em-
ployed in this category. Also included were the
costs of biologically activated carbon (BAG),
based on ozone in combination with GAC.
Treatment Cost Analysis
Under contract to EPA's MERL, the firm of Culp/Wesner/Culp
developed a database for deriving individual system costs for
various water treatment processes. Appendix C provides a full
set of costs per system derived from the model as well as an
example of the model's cost printout. This model, which resulted
from the EPA Unit Treatment Cost Report, requires as inputs the
design and operating flow rates of both the treatment plant and
the process being considered, chemical costs, and operating char-
acteristics such as utilization rates. These characteristics
must then be translated into operating and design parameters
useful for input to the model, such as pounds of chlorine, square
feet of surface area of contactors, etc. In addition, the model
requires specification of unit cost factors, capital cost fac-
tors, and cost (price) indices. As output, the model yields
total costs for construction, capital (which includes construc-
tion costs plus fees and contingencies, interest, and land
costs), and the components of total annual operation and mainte-
nance expenses. The model employs cost curves with the system
capacity and average flows as the primary determining factors.
Unique, site-specific considerations are not explicitly accounted
for beyond the inclusion of contingency costs.
Given the treatment options available to meet the THM reg-
ulation, this model was used to generate unit treatment costs
for each option for each of the nine system size categories con-
sidered. This procedure was not used in the 1977 THM Economic
Impact Report, the model not being available at that time. The
cost data from the EPA Unit Treatment Cost Report are a key ele-
ment in determining whether a given treatment is likely to be
used by a utility in meeting the MCL.
Regulatory Compliance Strategies
As previously discussed, water systems which exceed the
THM MCL have three major options available to satisfy the reg-
ulatory standard—modifying chlorination or other treatment
-------
II-7
procedures, changing disinfectants, and-adding an adsorbent.
In order to complete the basis for estimating the total national
costs of the regulation, the number of systems likely to select
each of the treatment options must be established.
Since there is no empirical method for predetermining the
choice which will be made by each water system exceeding the MCL,
a more probabilistic and structured approach was necessary. The
approach chosen, decision-tree analysis, is a step-by-step pro-
cedure which can be tracked easily and modified as new informa-
tion becomes available. A logical sequence of decision points
was designed to distribute the systems covered by the regulation
according to the most likely path they would follow. The deci-
sion made at each point is consistent with certain criteria. The
criteria are based upon:
• The treatments currently used: if a system does
not use a disinfectant it will not be affected by
a THM regulation and therefore will require no
changes in its current treatment practices;
• Water source used: if a system uses surface
water as its primary source, it is more likely
to exceed a given level of THM contamination.
Hence, the number of water systems using water
from ground or surface sources affects the num-
ber of systems which will exceed a given level
and therefore require treatment;
• Degree to which water quality exceeds MCL: if
the presence of THM is only slightly in excess
of the standard, then minimal modifications to
existing treatment procedures will be adequate
for compliance. As the level of contamination
increases, a system must consider more signifi-
cant (and costly) modification of its existing
treatment techniques;
• Economic considerations: the presumption is
that systems will adopt the least cost treat-
ment strategy which satisfies the regulation;
• Treatment effectiveness: the presence of THM
above certain levels can probably be controlled
only by the use of adsorbents. This is because
of the likelihood that water with a high disin-
fectant demand cannot be adequately disinfected
without generating a considerable amount of by-
products of unknown risks. Consequently, those
-------
II-8
few systems with a very high level of THM are
likely to require the addition of the most
costly treatment.
The estimates presented below are the results of consider-
ing these criteria. The primary participants in the evaluation
were:
• The technical staff of EPA-MERL,
• EPA Office of Drinking Water (ODW) staff, and
• TBS staff.
The decision tree, shown in Exhibit II-3, illustrates the
paths expected to be followed for compliance with the MCL at
0.10 milligrams per liter by each of the water systems which
serve more than 10,000 people. Of the 2,685 community water
systems that serve more than 10,000 people, 416 purchase the
majority of their water from other systems that are presumed to
provide treatment. Thus, a total of 2,269 systems would be
initially affected, although 289 of these are excluded because
they do not presently use a disinfectant. Of the remaining
1,980, some 515 systems are estimated to have THM levels above
0.10 milligrams per liter and hence would require changes in
their treatment processes.
In general, of the systems estimated to be in the range of
1 to 1.5 times the MCL, 60 percent are expected to modify their
existing disinfection procedures and 40 percent are expected
to change disinfectants. Of the systems with THM levels in the
range of 1.5 to 2.5 times the MCL, 25 percent are expected to
change their disinfection procedures with 75 percent switching
to a different disinfectant. Finally, of the systems exceeding
2.5 times the MCL, 80 percent are anticipated to change disinfec-
tants and the remaining 20 percent will use an adsorbent. The
results of these treatment selections are that 318 systems
would change disinfectants and 25 would use adsorbents as a
compliance strategy. The remaining systems of the original 515
would modify their existing disinfection procedures to achieve
compliance with the regulation. Also, as Table II-l shows,
approximately 28.7 million people are served by the systems
which would be likely to exceed the standard prior to any
corrective measures.
-------
1 1-9
Exhibit 11-3
DECISION TREE FOR A THM
REGUUATJON OF 0.10 MILLIGRAMS
PER LITER MCL
75%"I CJilorsmrMi
mu
10%-
3AC
' 12-month rsacovanon 11—; 131
~o£^a 1 Modify ?rocsaures
I
I
Chlortn« Oio
Ozone
*R«Or«Mnt9 tti« total numb*' of ffn*rr\* itfvin
-------
11-10
Table II-l
MOST PROBABLE TREATMENT SELECTION BY WATER SYSTEMS
AFFECTED 8Y MCL REGULATION OF THM AT 0.10 MILLIGRAMS PER LITER
Number of
Systems
Percent of
Total Affected
Population
Affected
(millions)
Move Point of
Disinfection
and/or
Adjust Dosage
172
33
12.1
Change Use
Disinfectant Adsorbent Total
318
62
14.9
25
1.7
515
100
28.7
Source: Estimates based on inputs from EPA-MERL, EPA-ODW, and TBS.
MODELING APPROACH
The two" key intermediate outputs of the analytic procedure—
the number and characteristics of the systems exceeding the MCL
and the estimate of the particular compliance strategies selected
by each—become the principal inputs to the PTm, which calculates
the national economic impacts of the regulation. Simply de-
scribed, the PTm first determines the number of systems which
would select each new treatment as a result of the regulation
being examined and then applies the relevant treatment costs.
The model determines the financial impact of those additional
costs on the utility's overall operating statements for a speci-
fied future year. A comparison of these new financial statements
with the baseline reports yields an estimate of the economic
impact of the regulation.
These computations require a complete recalculation of
the financial flows of funds that take place in a water util-
ity during a full year. A major element of these calculations
centers on capital items. The model projects capital expendi-
tures financing through a combination of available internal
sources and external sources, which include both debt and equity
at prevailing rates of return. The revenues required in a given
future year by a system requiring a new treatment consist of
the baseline revenues (those for normal operations) plus oper-
ating and maintenance costs for the new treatment plus the an-
nualized costs (capital costs plus depreciation) of the capital
expenditures for the new treatment.
-------
11-11
Per capita costs have been calculated by dividing the
additional revenues required for a given treatment (excluding
the baseline revenues for normal operations) by the total resi-
dent population served. This method of assuming all costs
would be passed along to residential customers tends to state
per capita costs at their maximum level since a portion of the
costs would normally be billed to commercial, industrial, and
wholesale customers also served by the utility. However, the
increased costs of goods and services produced by non-residen-
tial customers may, in some cases, be passed along to the resi-
dential population.
Residential bill impacts of the proposed regulation have
been estimated to provide a closer approximation of the actual
cost to be incurred by an average family. The additional re-
quired revenues are allocated both to residential customers
and to non-residential classes of customers. That proportion
of revenues presently received from non-residential customers
varies from 50 percent of revenue requirements in the over-
one-million-population-served category to 35 percent in the
10,000-to-25,000-population-served category. This proportion
tends to decline as the size of the system decreases due to
the declining number and size of commercial, industrial, and
wholesale customers of those smaller systems. Monitoring costs
are not included in either the residential bill or per capita
cost calculations.
From the use of PTm, national and typical system cost
estimates were developed for a 0.10 milligram per liter MCL
regulation for THM. These data, along with energy impacts,
supplier impacts, and sensitivity analyses are presented in
the following chapters of this report.
-------
III. THE ECONOMIC IMPACT
OF THE REGULATION
This chapter presents the economic impacts of the THM
regulation. The cost estimates are first discussed in terms
of the national costs for all systems requiring treatment and,
second, in terms of the costs to individual systems. Changes
in the analysis and resulting cost impacts since the 1977 THM
Economic Impact Report are also presented. Monitoring costs
are then discussed separately since these will be incurred by
water systems whether or not they exceed the MCL for THM.
Estimates of the demand on supplying industries constitute the
final section of the chapter.
NATIONAL COST ESTIMATES
The economic implications of a THM regulation at 0.10 mil-
ligrams per liter, covering community water systems serving
10,000 people or more, are summarized below in terms of six key
measures:
• Capital expenditure requirements during
the period are projected to be $85 mil-
lion (1980 dollars).
• Annual operating and maintenance (O&M)
expenses for the required treatments and
for monitoring are estimated at approx-
imately $10 million.
• Annual revenue requirements, reflecting
the amortization of capital expenditures
and the O&M expenses, are expected to in-
crease by a total of $19 million for the
343 systems which are likely to have cost
impacts.
• Per capita costs, in terms of total rev-
enue impacts divided by the population
served by systems with cost impacts, are
projected to be $0.70 per year, ranging
from $.00 to $6.20 per year.
• The average residential customer bill in-
crease is projected to be $1.40 per year.
This represents a range of $.00 to $12.00
per year.
-------
111-2
• The projected annual energy Celectrical, diesel,
and natural gas) consumption as a result of this
regulation is 510 billion BTUs or 0.0007 percent
of the 1977 total U.S. energy consumption (76.3
quadrillion BTUs). Total annual energy costs in
1980 dollars are projected to be $2.3 million.
As Table III-l indicates, over 54 percent of the aggregate
costs of this regulation is expected to be borne by systems
serving 75,000 people or more.
Table III-l
SUMMARY OF TOTAL COSTS FOR AN MCL REGULATION
OF THM AT 0.10 MILLIGRAMS PER LITER
(millions of 1980 dollars)
Systems Serving Populations of:
Number of Systems Exceeding MCL
National Costs
Capital Expenditures
Operating & Maintenance
Expenses
Consumer Charges
Revenue Requirements
Annual Per Capita Cost (S/year)
> 75,000
95 out of 390
S 45.6
S 5.4
S 10.1
S 0.60
Residential Bill Increase (S/year) S 1.20
Range of Residential 3ill
Increase (S/year) $0-8.90
10,000-75,000
420 out of 2,295
S 39.8
% 4.6
S 3.6
S 0.90
S 1.80
SO-12.00
Total
515 out of 2,685
S 85.4
% 10.0
S 18.7
S 0.70
S 1.40
SO-12.00
Source: Estimates based on inputs from EPA-MERL, EPA-OOW, and TBS.
These cost figures include the expenses of the 515 systems
adding or altering treatment practices plus monitoring costs for
all systems using a disinfectant. Table III-2 breaks down these
costs into those attributable to each treatment category. Ap-
proximately 53 percent of the total capital cost is attributable
to the 25 systems adding adsorbents, though these systems repre-
sent only 5 percent of the number affected by the regulation.
Table III-3 presents a comparison of the revised national
cost estimates with those outlined in the 1977 THM Economic
Impact Report.
-------
III-3
Table III-2
SUMMARY OF COSTS BY TREATMENT CATEGORY
FOR AN MCL REGULATION OF THM
AT 0.10 MILLIGRAMS PER LITER
(millions of 1980 dollars)
Change Disinfectant
Use Adsorbent
No Cost Changes
Total
* Systems
318
25
172
515
Capital
Expenditures
$85.4
Annual 0 4 M
Expense*
Does not include monitoring costs.
Source: Estimates based on inputs from EPA-MERL, EPA-ODW,
and TBS.
ECONOMIC
THM AT
COMPARISON WITH
Coverge According to
Population Served
Number of Systems
Covered
Impacted (exceed MCL)
Cost Impacted**
National Costs
Capital Expenditures
04M Expenses
Revenue Requirements
Consumer Charges
Average Cost Per Capita
(S/year)
Residential Sill Increase
(S/year)
Table III-3
IMPACT OF AN MCL REGULATION OF
0.10 MILLIGRAMS PER LITER:
THE 1977 THM ECONOMIC IMPACT REPORT*
(millions of dollars)
Results from 1977 THM Report
(1976 dollars) (1980 dollars)
> 75,000 > 75,000
390 390
86 86
65 65
S154.4 $209. 8
$ 25.9 $ 35.2
S 36.0 $ 48.9
S 2.10 S 2.80
S 3.70 $ 5.00
»
TBS report "Economic Impact Analysis of a Trihalomethane
for Drinking Water" for EPA, August 1977.
Current
Estimate
*•** (1980 dollars)
> 10,000
2,685
515
343
$ 85.4
S 10.0
S 18.7
S 0.70
S 1.40
Regulation
Cost impacted refers to those systems requiring different treatments —
not just modification to existing procedures or monitoring.
**
Inflated using the producer price index where 1976=1732 and 1980=2353.
Source: Estimates based on inputs from EPA-MERL, EPA-ODW, and TBS.
-------
III-4
As is evident from Table III-3, a^number of the analytic
inputs used to derive the national cost estimates have been
altered since the preparation of the 1977 THM Economic Impact
Report. These include:
• Lowering the population cut-off from 75,000
to 10,000 people served.
• Changes in the decision tree which resulted in
a smaller percentage of systems choosing an ad-
sorbent to comply with the MCL. Table III-4
provides a summary of the number of systems
choosing each treatment alternative, for both
the present report and the 1977 THM Economic
Impact Report.
• Changes in the assumptions regarding the use of
adsorbents. In comparison to the previous TBS
study, longer regeneration cycles were assumed
as well as replacement of sand in conventional
filters with carbon rather than post-filtration
contactors.
• Changes in the individual system treatment costs.
The 1979 estimates are derived from the EPA Unit
Treatment Cost Report.
The Survey of Community Water Systems (see ref-
erence in Appendix D) database was updated and
reevaluated, resulting in a larger number of
surface systems which use chlorine for disinfection
Table III-4
NUMBER OF SYSTEMS CHOOSING
EACH TREATMENT FOR AN MCL
REGULATION OF 0.10 MILLIGRAMS PER LITER:
COMPARISON WITH 1977 THM ECONOMIC IMPACT REPORT
Treatments
1977 Report
Systems Serving
Over 75,000 People
1979 Report
Systems Serving
Over 75,000 People
1979 Report
Systems Serving
Over 10,000 People
Chlorine
Chloramines Dioxide
21
SAC
Plus
Ozone GAC Ozone
25
234
45
39
13
12
Source: Estimates tiasea on incuts from EPA-MERL, EPA-OOW, and T3S.
-------
III-5
• Update of the monetary basis -oiN the analysis
from 1976 dollars to 1980 dollars.
A more detailed summary of the treatment costs for an indi-
vidual system in three of the six size categories over 10,000
people appears on the following page. A full description of
individual system treatment costs is provided in Appendix C.
COSTS TO AN INDIVIDUAL SYSTEM
The costs for the five types of treatments--ozonation,
chlorine dioxide, chlorination/ammoniation, GAC, and GAC plus
ozone (commonly referred to as BAG)—can best be compared on
the basis of additional per capita costs for an individual
water system. They are as follows:
• Ozonation (plus residual disinfectant) is the
most capital intensive of the three alternate
disinfectant treatments. Systems serving over
one million people would need capital expendi-
tures of about $7.2 million each. Annual
per capita costs range from $1.40 to $3.00.
• Chlorine dioxide treatment requires only minor
investment but cosiderable expense for the
purchase of sodium chlorite. Per capita costs
range from $0.60 to $1.20 per year.
• Chlorination/ammoniation is the least expensive
treatment with annual per capita costs in the
$0.20 to $0.30 range.
• Adding GAC as an adsorbent involves substantial
capital expenditures (approximately $21 million
for a typical system serving over one million
people) as well as continuing operating expenses
for reactivation. Per capita costs range from
$2.40 to $3.70.
• The use of BAG results in per capita costs
ranging from $3.20 to $6.20. BAG is the most
capital intensive treatment with capital expend-
itures of $27 million for a typical system serv-
ing over one million people.
-------
III-6
The capital expenditures, annual revenue requirements, and
per capita costs are shown in Table III-5 for each treatment a-
bove and for each of three size categories over 10,000 (of the
six size categories employed in the analysis). It is clear that
the range of costs is broad across treatments and size categories,
The use of an adsorbent (GAC or BAG) is considerably more ex-
pensive than any of the alternative disinfectants for all
Table III-5
COMPLIANCE COSTS FOR A TYPICAL WATER SYSTEM UNDER
AN MCL REGULATION OF THM AT 0.10 MILLIGRAMS PER LITER
(1980 dollars)
10,000-25,000 75.000-100,000 Over 1 Million
Average Population Served
Per System
Ozone
Capital Expenditures
Revenue Requirements/Year
Annual Per Capita Cost
Chlorine-Dioxide
17,000
341,000
51,000
3.00
93,000
S 1,471,000
210,000
2.30
1,223,000
S 7,161,000
1,672,000 j
1.40
Capital Expenditures
Revenue Requirements/Year
Annual Per Capita Cost
Chlorination/Ammoniation
Capital Expenditures
Revenue Requirements/Year
Annual Per Capita Cost
GAC
Capital Expenditures
Revenue Requirements/Year
Annual Per Capita Cost
GAC and Ozone (8AC)
Capital Expenditures
Revenue Requirements/Year
Annual Per Capita Cost
S 30,000
20,000
1.20
$ 12,000
5,000
0.30
S 435,000
64,000
3.70
$ 760,000
106,000
6:20
Source: Derived using a computer model based
Report with inputs provided by TBS in
Gulp and MERL staff members.
$ 38,000
60,000
0.70
$ 22,000
21,000
0.20
S 1,733,000
266,000
2.90
S 3,141,000
424,000
4.50
on EPA's Unit
conjunction
S 362,000
717,000
0.60
S 61,000
259,000
0.20
$21,063,000
2,902,000
2.40
S27,"259,000
3,883,000
3.20
Treatment Cost
with Culp/Wesner/
-------
III-7
size categories.1- Among the disinfectants, chlorine plus
ammonia is always the least expensive. As discussed in Chapter
II, the number of systems selecting each treatment and the
system treatment costs represent the primary inputs into the
analysis of the total national cost of the THM regulation.
MONITORING COSTS
In addition to the treatment costs which the 515 impacted
water systems serving over 10,000 will incur, there are speci-
fic monitoring requirements included in the regulation. The
costs associated with those requirements are included in the
national cost estimates presented earlier in this chapter. All
systems serving over 10,000 people and which use a disinfectant
will be required to monitor for the presence of THM.
Sampling frequencies required by the regulation vary ac-
cording to water source. Water systems drawing some or all of
their water from surface sources will monitor at a minimum fre-
quency of four samples per quarter taken on the same day. Sys-
tems with only groundwater sources which do not have high levels
of total organic carbon (TOO may reduce the frequency to one
sample per quarter. Following the first year of monitoring,
those systems not exceeding the MCL may reduce the monitoring
frequency at their State's discretion.
Since monitoring may be carried out on a plant-by-plant
basis, some utilities may sample more extensively than indi-
cated above. However, it is not possible to make an accurate
estimate of the exact number of samples. Hence the costs pre-
sented reflect costs to be incurred by single plant utilities.
The annual national monitoring costs for the 2,396 systems
serving over 10,000 which do not use a disinfectant amount to
about $1.1 million.2 This estimate is based on a $50 per sam-
ple cost, assuming four samples per quarter for surface systems
However, the use of an adsorbent has the ancillary benefit
of generally reducing the level of other organic chemicals
in addition to THM. Its use may also result in reduced dis-
infectant demand .
2
These 2,396 systems are those which remain after subtracting
from 2,685 the 289 systems which do not use a disinfectant.
-------
III-8
and one sample per quarter for groundwater systems. This
amounts to annual costs of $800 and $200 respectively. Moni-
toring costs were computed based upon a survey of contract ana-
lytical laboratories currently performing THM analyses.3 Per
sample costs ranged from $25 to $100. After these regulations
have been promulgated, the increased volume of business and
competitive factors would be expected to reduce the analytical
costs to well below $50 per sample. Therefore, the cost of
monitoring can be expected to decline over time.
Although the monitoring costs are based on the use of com-
mercial laboratories, EPA expects that a number of community
water systems will choose to purchase the equipment and monitor
for THM on-site more frequently than the minimum, for operational
control as well as for compliance purposes. However, no capital
costs are included in the national cost estimates for such labor-
atory equipment. An additional benefit from the on-site analy-
tical capability is that the necessary equipment, which includes
a gas chromatograph, is versatile and can be used to monitor
for the presence of many other organic chemical contaminants
besides THM.
SUMMARY OF DEMAND ON
SUPPLYING INDUSTRIES
Aside from the costs of adding treatments to comply with
the organics regulations, EPA has also considered the level
of demand which would be placed upon industries supplying the
required materials and equipment. The six areas examined
include:
• Energy,
• Granular activated carbon,
• Regeneration furnaces,
• Chlorine dioxide,
• Ozonators, and
• Aqua ammonia.
3
The cost of equipping an existing laboratory with an appro-
priate gas chromatograph is dependent upon which analytical
procedure is selected and the type of instrument. The basic
(footnote continued on next page)
-------
III-9
In general, the conclusion is that undel: the regulation,
given the expected distribution of systems using each
treatment, no significant problems exist at the present time
for satisfying the demand in any of the areas listed above.
With the exception of chlorine dioxide, an industry in which
rapid expansion is possible, the estimated demands are well
within the capacity of the industries providing the materials.
It should be noted, however, that these demand projections
are based upon the needs of those systems assumed to be out of
compliance. Demand could be somewhat higher under at least two
conditions: (1) systems which do not exceed the MCL neverthe-
less decide to add a treatment which will reduce their THM
levels, and (2) systems which do exceed the MCL add more treat-
ment capacity to reduce THM levels considerably below the MCL.
If both conditions occurred, the demand projections would be
understated.
Energy
The regulation will have a negligible impact on annual
U.S. energy consumption. The total annual energy requirements
of the various treatment alternatives selected by utilities to
meet the regulation are as follows: electric power, 40 million
kilowatt-hours; diesel fuel, 64,000 gallons; and natural gas,
76 million cubic feet. In 1980 dollars these total annual
energy requirements are estimated to cost $2.3 million per year
The annual electric power demand of 40 million kilowatt-hours
is approximately 0.002 percent of 1977 total domestic electric
power sales. The annual diesel fuel demand represents 0.00002
percent of the 1977 total domestic demand for refined oil
products. At 76 million cubic feet, the annual natural gas
demand represents less than 0.004 percent of the 1977 domestic
natural gas demand. Together these energy requirements repre-
sent 0.0007 percent of total 1977 U.S. energy consumption, or
510 billion BTUs out of a total of 76 quadrillion BTUs. When
(continued from previous page)
instrumentation for the "liquid-liquid" extraction method con-
sists of a gas chromatograph with an "Electron Capture" detec-
tor and recorder; the basic cost is approximately $5,000 to
$10,000. The basic instrumentation for the "purge and trap"
method consists of a gas chromatograph, a "Hall" detector,
purge and trap sample concentrator, and recorder; the basic
cost ranges from $10,000 to $20,000. In either case, some
additional expenditures for accessories would be added.
-------
111-10
compared to estimates of 1980 energy consumption of 80 quadril-
lion BTUs, the portion of total consumption attributable to the
THM regulation decreases to 0.0006 percent.
Approximately 87 percent of the electric power demand is
due to ozone disinfection processes. GAC treatment and ozona-
tion together represent 96 percent of the total electric power
demand.
The diesel fuel and natural gas requirements are created
by the GAC regeneration process. For those water utilities
without on-site GAC regeneration, transport of GAC to a regional
processing site will require diesel fuel. The regeneration
process itself requires either oil or natural gas as an energy
source. In preparing these energy demand estimates, it was
assumed that only natural gas would be used in GAC regeneration
furnaces.
Granular Activated Carbon
A THM regulation will result in some systems treating their
water with adsorbents. The estimated demand for the initial fill
of GAC would be 3.7 million pounds for the 25 systems expected
to use adsorbents to comply with the regulation. This level of
demand, along with the demand generated by the annual replacement
of carbon (0.2 million pounds per year) lost in reactivation
cycles, could easily be met by the carbon industry. Current
available excess capacity in the GAC industry is over 100
million pounds per year.
Regeneration Furnaces
A ThM regulation at 0.10 milligrams per liter would create
a total demand of 8 to 26 regeneration furnaces for those sys-
tems employing adsorbents. The limits of this range should be
viewed as the minimum and maximum since many of the smaller
systems will share ownership and many of the larger systems,
particularly those serving over one million persons, often have
more than one treatment plant. At these larger sizes it is
more economical to purchase a furnace at each plant rather than
to transport large volumes of carbon to a central furnace loca-
tion. Even if the upper limit estimate of 26 furnaces were
increased substantially, the furnace industry could supply an
-------
III-ll
adequate number of furnaces.4 The furnace industry's current
excess capacity exceeds 100 custom-designed furnaces per year
Chlorine Dioxide
The use of chlorine dioxide treatment rather than chlori-
nation to meet a THM regulation could create an annual demand
for over four million pounds of sodium chlorite. Excess in-
dustry capacity of at least three to four million pounds pre-
sently exists. The industry claims rapid capacity expansion
is possible if required by additional demand.
Ozonators
A second alternative to chlorination as a disinfection
process is ozonation. Such a treatment will require the
purchase of an ozonating system to produce the needed ozone
with electrical energy. Since approximately 52 water systems
are expected to use the ozone disinfection process, and since
some large systems with more than one treatment plant would
purchase several ozonating systems, the number of ozonating
systems required will be higher than 52. Although ozonator
demand is larger than projected in the 1977 THM Economic Impact
Report, production capacity constraints are not likely to
affect implementation of the regulation.
Aqua Ammonia
A third alternative method of disinfection is the use of
aqua ammonia in combination with chlorine. The use of this
treatment to meet the THM standard could create an annual de-
mand of over seven million pounds (3,500 tons) of aqua ammonia.
Given the 1976 domestic consumption of aqua ammonia of 659,000
tons, this demand level is minimal and creates no constraint to
compliance.
The following table summarizes the estimated demand for
the major equipment and materials likely to be needed by water
systems exceeding the MCL.
4
A two-year lead time is generally required for the design,
construction, and start-up of custom regeneration furnaces.
-------
Table 1II-6
MATERIALS REQUIREMENTS FOR PROPOSED THM REGULATION
Regulation
THM 0 0.10
Milligrams Per
Liter
Energy
Adsorbent
Chlorine Dioxide
Ozonators
Aqua Ammonia
Number Energy
of Demand in
Systems Billions
Affected of BTUs1
515
510
Number
of
Systems
Affected
25
GAC Demand in
Million Lbs.--
Initial Fill2
3.7
Minimum
Demand for
Furnaces-*
8-26
Number
of
Systems
Affected
Annual Demand
for NaCIO?
in Million Lbs.4
4.34
Number
of Minimum
Systems Number of
Affected Ozonators5
52
52
Number
of
Systems
Affected
234
Annual Demand
for Ammonia in
Million Lbs.6
7.14
H-l
>—I
I
h-1
to
Energy demand is drawn from individual system estimates and translated into BTUS using the following conversion factors: (a) electricity--10,500
BTU/KWH; (b) diesel fuel--138,700 BTU/gallon; and (c) natural gas--l,050 BTU/cubic foot.
2GAC demand based on nine-minute contact time and shared regeneration where practical.
3This is the range of furnace demand based on the number of systems that would require furnaces. Systems with multiple plants would require multiple
furnaces.
demand based on dosage of 1.5 milligrams per liter of C102-
Each system will use one or more ozonators, with only the largest systems requiring more than one.
6Ammonia demand based on 9.5 pounds per million gallons treated.
Source: Estimates drawn from outputs of the model of EPA's Unit Treatment Cost Report.
-------
IV. SENSITIVITY ANALYSES
There are several variables in the economic analysis which
if changed, produce significant differences in the results.
The following section summarizes the effect of:
• Varying the mix of treatments which systems
would be expected to select,
• Changing the MCL to a higher or lower THM
level, and
• Including system size boundaries above and
below 10,000 people in the regulation.
ALTERNATIVE DISTRIBUTION
OF TREATMENT SELECTION
The analysis of economic impacts of the THM regulation has
assumed a specific set of treatment choices for systems exceed-
ing the MCL, as outlined in Chapter II. If the same systems
were to choose a different mix of treatments, the level of
total costs would change. Total national costs are especially
sensitive to the number of systems using an adsorbent. If only
one additional system chose BAG to comply with the regulation
and that system served over one million people, total national
costs would increase by approximately $27 million or 32 per-
cent. Since the behavior of systems is uncertain, an example
of the costs for a different mix has been presented below,
along with the costs of the most likely mix of treatments.
As indicated above, the major factor determining the total
economic impact of a change in treatment mix is the percentage
of systems which would use adsorbents instead of changing dis-
infectants. In the example given in Table IV-1, the number of
systems using adsorbents has been increased from 5 percent of
the systems which exceed the MCL to 30 percent. The projected
economic impacts of the regulation change accordingly: annual
revenue requirements increase from $18.7 million to $43.8 mil-
lion, a 134 percent increase. Capital expenditures display a
similar sensitivity to the assumed mix of compliance treatment
strategies; they increase by 219 percent to $272.7 million.
-------
IV-2
Table IV-1
SENSITIVITY OF COSTS TO MIX OF COMPLYING
TREATMENTS FOR AN MCL OF THM
AT 0.10 MILLIGRAMS PER LITER
(millions of 1980 dollars)
Number of Systems
Exceeding the MCL
Installing Adsorbents
National Costs
Capital Expenditures
01M Expenses
Revenue Requirements
Consumer Charges
Average Cost Per Capita
(S/year)
Residential Bill Increase
(S/year)
Best
Estimate
515
25
S 85.4
$ 10.0
S 18.7
$ .70
S 1.40
Higher Use
of Adsorbent
(GAC)*
515
153
S272.7
$ 16.3
$ 43.8
S 1.80
S 3.30
No change in the number of systems choosing SAC
was assumed.
Source: Estimates based on inputs from EPA-MERL,
EPA-OOW, and TBS.
ALTERNATIVE MCLs
The maximum contaminant level of THM at 0.10 milligrams
per liter was selected on the basis of the protection it would
afford through a considerable reduction of THM in water con-
sumed by a large proportion of the population. This protection
could be achieved while minimizing the negative effect on the
microbiological quality of the water. Two alternative MCLs
were examined in order to illustrate the sensitivity of total
costs to a change in the MCL. One case represents a somewhat
more stringent MCL of THM at 0.05 milligrams per liter. The
second case represents a less stringent MCL of THM at 0.15
milligrams per liter.
-------
IV-3
Given a mix of treatment selections,-the most important
variable in determining the economic impact of these alterna-
tive MCLs is the number of systems exceeding those levels. In
the first case of THM at 0.05 milligrams per liter, 35 percent--
or 944—of the systems serving over 10,000 people would exceed
that level. In the second case (THM at 0.15 milligrams per
liter), only 11 percent—or 301 systems—would exceed that MCL.
The treatment options and their associated costs are
assumed to be the same as those used in estimating the national
costs for an MCL at 0.10 milligrams per liter. However, the
mix of treatments varies for the 0.05 milligram per liter case
in that more systems would select adsorbents to comply with the
more stringent MCL. The decision criteria for selecting a par-
ticular treatment in the 0.15 milligram per liter case are the
same as those assumed for the MCL at 0.10 milligrams per liter.
Table IV-2 below compares the total costs of these alter-
native MCLs to the cost for the MCL at 0.10 milligrams per
liter. The impact in terms of capital expenditures is projec-
ted to be $314.8 million for the 0.05 milligram per liter level
and $45.0 million for the 0.15 milligram per liter level versus
the projections presented earlier of $85.4 million for a 0.10
milligram per liter regulation. The other aggregate impacts,
such as annual operating and maintenance expenses and annual
revenue requirements, vary similarly.
Table
IV-2
SUMMARY OF TOTAL COSTS UNDER ALTERNATIVE MCLs FOR
(millions of
Number of Systems Exceeding
Capital Expenditures
O&M Expenses
Revenue Requirements
1980 dollars)
MCL
(milligrams per
O.OS 0.10
the MCL 944 515
S314.8 $85.4
$ 22.0 $10.0
$ 53.8 $18.7
Residential Rate Increase (S/year) $ 2.30 $1.40
Source: Estimates based on
and TBS.
THM
liter)
0.15
301
$45.0
$ 6.0
$10. 6
$1.40
inputs from EPA-MERL, EPA-ODW,
-------
IV-4
Table IV-3 summarizes the number of ^systems estimated to
select each treatment alternative under three MCLs: 0.05,
0.10, and 0.15 milligrams per liter. In the 0.05 milligram
per liter case, it is estimated that the largest portion (48
percent) of the 944 systems will elect to change disinfectants,
At the 0.15 milligram per liter MCL, approximately 59 percent
would elect to change disinfectants.
Table IV-3
SUMMARY OF TREATMENT SELECTIONS
3Y SYSTEMS EXCEEDING ALTERNATIVE MCLs FOR
THM
Treatment
Category j
Change Disinfectant
Use Adsorbent
Modify Procedures
i
Total
9 .05 mg/1 THM 9 .10 mg/1
Systems
•153
153
338
944
* Systems
318
25
172
515
THM
THM ? .15 mg/1
4 Systems
179
12
110
301
Source: Estimates based on inputs from EPA-MERL, EPA-OOW, and TBS.
ALTERNATIVE SYSTEM SIZES INCLUDED
IN REGULATORY COVERAGE
The final example of cost sensitivity is the analysis of
extending the coverage of a THM regulation to systems smaller
than those serving 10,000 people. This section presents a
summary of the number of systems exceeding the MCL and the
related costs for four alternative system size boundaries in
addition to the population cut-off in the regulation: (1) all
community water systems serving over 1,000 people; (2) those
serving over 10,000 (the regulation); (3) those serving over
50,000; (4) those serving over 75,000; and (5) those serving
over 100,000.
This portion of the analysis assumes an MCL for THM at
0.10 milligrams per liter and determines the impacts under the
five alternative size limitations indicated above. The number
of systems which exceed the MCL increases substantially as
the population cut-off is lowered. If the lower boundary
-------
IV-5
were reduced to 1,000 people, more than three times as many
systems would be out of compliance with the 0.10 milligrams per
liter level of THM.
As shown in Table IV-4, the aggregated economic impacts
would not increase substantially if the boundary were lowered
to 1,000. In the case of those systems serving under 10,000
persons, the severity of the impact is at the individual system
level, rather than at the national level. Capital expenditures,
for example, would increase from $85.4 million with a 10,000
population cut-off to $131.4 million with a cut-off at 1,000.
Annual revenue requirements would increase similarly from $19
million with a 10,000 cut-off to $28 million with a cut-off at
1,000 people.
Table IV-4
COSTS OF ALTERNATIVE SIZE LIMITATIONS
FOR AN MCL
OF THM AT 0.10
MILLIGRAMS
PER LITtR
(millions of 1980 dollars)
Population
Number of Systems Exceeding the MCL
Capital Expenditures
O&M Expenses
Revenue Requirements
Residential Rate Increase (S/year)
Source: Estimates based on inputs
Serving
> 1,000
1,653
$131.4
$ 14.8
$ 27.9
$ 2.90
from EPA-MERL,
Serving
> 10,000
515
$85.4
$10.0
$18.7
Served
Serving Serving
> 50,000 > 75,000
165
$58.1
$ 6.7
$12.7
$1.40 $1.20
EPA-OOW, and TBS.
95
$45.6
$ 5.4
$10.1
$1.20
Serving
> 100,000
63
$37.7
$ 4.6
$ 8.5
$ 1.10
-------
Appendix A
WATER UTILITIES POLICY TESTING MODEL
The Water Utilities Policy Testing model (PTm) was the
primary computational tool used to estimate the national costs
of compliance with the proposed THM regulation. Computations
were performed by integrating the "building blocks" summarized
in Chapter II. The following sections of this appendix describe
the objectives of the model design, outline the general model
structure, and list some of the key inputs used in the model.
MODEL DESIGN OBJECTIVES
PTm was developed for evaluating the economic and financial
impact on community water systems of various proposed regula-
tions under the Safe Drinking Water Act. Given the need for
testing alternative treatments and costs and various assumptions
about the industry, the model was designed to provide maximum
flexibility and rapid turnaround for policy analysis. The
specific model characteristics which are important in meeting
this need include the following:
• All results can be available at the regional
level (EPA regions) as well as at the national
level;
• Separate projections are produced for the nine
different sizes of water systems (population
served) defined in the "Survey of Operating and
Financial Characteristics of Community Water
Systems," performed by TBS for EPA (hereafter
referred to as the Survey of Community Water
Systems);
• Added projections are available to provide more
detail on systems serving between 10,000 and
100,000 people;
• Financial and other results are reported sepa-
rately for public and private system ownership;
• Cost analyses are performed separately by system
water source (ground, surface, and purchased);
and
-------
A-2
Results are obtainable in aggregate for the
nation or a region and also for a representative
system of a given type (e.g., a privately owned
system in EPA Region V serving a population of
between 25 and 100 people).
MODEL STRUCTURE
The
shown in
program elements in the superstructure of PTm
the following diagram:
are
Demand/Capacity
Module
Decision Tree
Module
Treatment
Module
Finance
Module
Report
Generator
Each module has its own input files with each data item
identified b'y a phrase in English. This feature greatly facil-
itates updating and assumption modification for sensitivity
analyses. The input requirements and function of each of the
above model components are described below.
• Control Program. This model component accepts
as input a control file where the user has spe-
cified the parameters for the scope of the
analyses (e.g., regions, system ownership, sys-
tem size, and current or constant dollars).
The Control Program executes the specified
modules with the desired inputs.
• Demand/Capacity Module. Given the initial popu-
lation, number of systems, production, and pro-
duction capacity, this module calculates future
demand, the necessary additions to production
capacity, and the timing of those additions.
Required inputs include growth rates for popu-
lation and the number of systems in each size
category, and the sources of future additions
(surface, ground, and purchased water sources).
-------
A-3
• Decision Tree Module. The strategy which any
individual water system will select for com-
pliance with a regulation depends upon many
factors: the water source used, treatments
currently in place, water quality, economic
considerations, and treatment effectiveness.
The various combinations of these factors are
displayed in the form of a decision tree (a
sample of which is included in Chapter II). The
identification of the appropriate compliance
strategy or strategies at the end of each path-
way through the tree is determined exogenously,
based upon engineering and other professional
judgments. This program module accepts the
specification of the decision tree as input and
computes the total number of water systems
which would, according to the decision tree,
select each specific method of compliance.
• Treatment Module. Using the results of the Deci-
sion Tree Module, the Treatment Module projects
capital expenditures and operating and maintenance
expenses required to fund the compliance strate-
gies for the regulation under analysis. Input
variables include timing of compliance and costs
per unit of capacity and production for each
potential treatment or nontreatment option.
• Finance Module. Using the results of the previous
modules, the Finance Module projects all the fi-
nancial information necessary to generate pro
forma income statements, balance sheets, and
sources and uses of funds statements. Input
variables include initial values for the bal-
ance sheet accounts and income statement line
items, sources for capital addition, tax rates,
inflation rates, depreciation rates, and rates
of return on capital. Building "from the bottom
up," the module determines the required operating
revenues and the corresponding consumer charges
necessary to cover operating expenses and returns
to capital.
• Report Generator. The Report Generator makes
available the full power of PTm to the user by
providing the results of the model in detailed
and concise reports. The following reports are
-------
A-4
available for specified sets -of-years, and for
system sizes, system ownerships, and regions
which the user selects:
—Summary Report
— Income Statement
—Balance Sheet
—Sources and Uses of Funds Report
—Demand Report
—Capacity Report
—Treatment Report (number of systems selecting
each treatment)
—Treatment Capital Expenditure Report
—Treatment Operating and Maintenance Expenses
Report
An example of a Summary Report is included as Exhibit A-l at the
end of this appendix.
KEY INPUTS
An important component of the model is the set of basic
inputs used in the calculations. These are the result of inte-
grating :
• Information from the TBS Survey of Operating and
Financial Characteristics of Community Water Sys-
tems (EPA-57Q/9-77-Q03, April 1977);
• Data compiled by the National Association of
Water Companies in their annual publications,
Financial Summary for Investor-Owned Water Util-
ities and Financial & Operating Data";
• Information published- by Moody's Investor Ser-
vice, Inc. ;
• Information contained in EBASCO Business Con-
sulting Company's Analysis of Public Utility
Financing;
• Experience gained through consultation with
water industry personnel;
-------
A-5
• Information obtained in interviews with members
of the banking and investment community; and
• Professional judgment.
Some of the major inputs are summarized below.
Size and Scope of the Water Utility Industry
The baseline projections forecast a specific number of
water systems existing in each of the nine size categories
in the year 1979. The number of systems was derived from the
Federal Reporting Data System and from the Survey of Community
Water Systems conducted in 1976. These numbers appear on the
summary printout in Exhibit A-l, along with the average produc-
tion per capita per day and the average number of people served
per system.
The average population served by a water system ranges
from 52 in the smallest category to 2.4 million in the largest
category. Production is the second measure of system size
varying from over 200 gallons per capita per day for systems
serving over one million people to 110 gallons per capita per
day for those serving under 100 people.
Industry Growth
It is forecast that the industry will experience modest
growth over the 1979-1981 forecast period. The anticipated
growth in water production, ranging from 0 to 22 percent
annually for the various system sizes, is the result of two
assumptions: (1) continued growth in population and the number
of customers and (2) a small annual increase in per capita
water consumption.
Financing
The financing of capital items has an important effect on
the industry's ability to assimilate any major new requirements
for capital expenditures. In projecting capital requirements,
PTm calculates the internal flow of funds with any remaining
fund needs to be obtained from external sources.
The projections in this analysis indicate that approxi-
mately 47 percent of the funds for normal (baseline) capital
expenditures in the 1979-1981 period will come from internal
sources. The model projects, however, that additional capital
-------
A-6
expenditures to meet new regulations wflT have to come exclu-
sively from external sources because the internal sources will
be exhausted in baseline uses.
When external financing is required, it is assumed that
it will be obtained in the proportions that have prevailed for
the past five years. For private systems, the proportions are
as follows: a range of 90 percent long-term debt for the
smaller private systems to 77 percent long-term debt for the
larger private systems, with the remainder in the form of com-
mon stock or other equity capital. Public systems are assumed
to use long-term debt as the sole source of external financing
for treatment expenditures. Other assumptions with respect
to financing include the following:
• Long-Term Debt Interest Rates. The embedded
rate for existing debt is 4 percent for public
systems and ranges from 4.5 percent to 6.6 per-
cent for private systems. The interest rate on
new debt is assumed to be 7 to 8 percent for
publicly owned systems and 9 to 10 percent for
privately owned systems. Although present rates
are slightly higher, the above interest rates
are those expected to prevail during the period
of compliance.
• Return on Equity. The rate of return on common
equity for investor-owned systems is 2 percent
for the smaller size categories and 8 to 10 per-
cent for the larger size categories. Also, the
general operating surplus for publicly owned sys-
tems is expressed in the model as a return on
other capital. This return on other capital
averages between 2 and 3 percent.
Industry Structure
The overall structure of the industry—the proportion of
publicly owned and investor-owned systems and the mix of primary
water sources—has been maintained throughout the forecast pe-
riod. The current projections do not assume a trend toward re-
gionalization of water systems or any other major changes in
industry structure which are potential results of compliance
with the THM regulation.
-------
1981
REGION 11
TOTAL OWNERSHIP
CONSTANT HOLLARS
DATA FOR ALL SYSTEMS
(MILLIONS OF DOLLARS)
Exhibit A-l
FINANCIAL MODEL OF COMMUNITY WATER SYSTEMS
TEMPLE» BARKER AND SLOANE
25-99 100-499
SUMMARY REPORT
500-999 1000- 2500- 5000-
lOtOOO- lOOfOOO- > 1 MIL
TOTAL
NO. OF SYSTEMS
AVG, PRODUCTION (000 GD)
21585.0
5.6
CAPITAL EXPENDITURES (CUM.)
FOR TREATMENT .0
OTHER 49.A
SOURCES OF FUNDS
TOTAL SOURCES (CUM.) 64.3
EXTERNAL SOURCES (CUM.) 33.2
OPERATING REVENUES AND EXPANSES
TOTAL OPERATING REVENUES 63.6
0/M COSTS FOR TREATMENT .0
CONSUMER CHARGES(CENTS/1000 GALS.)
RESIDENTIAL 64.7
REQUIK'EDr ALL CUSTOMERS 166.4
CONSUMER CHARGES($/CAPITA/YEAR>
RESIDENTIAL 20.7
REQUIRED* ALL CUSTOMERS 58.9
17668.0
25.9
.0
141.7
199.8
102.6
173.1
.0
100.4
119.0
26.7
40.4
5903.0
78.5
.0
173.0
227.9
132.7
161.9
.0
144.7
110.0
31.1
38.5
6050.0
217.9
.0
447.7
556.7
369.5
364.5
.0
93.5
91.3
25.9
36.2
2809.0
532.7
.0
479.7
620.8
382.8
399.6
.0
81.9
82.2
21.7
37.4
2068.0 2442.0 232.0 11.0 58768.0
1026.5 5532.1 50557,5 514449.5 627.5
.0 .0 .0
600.5 2510.4 1523.2
739.1
485.0
465.3
.0
72.5
72.3
21.5
32.3
3283.7
1658.9
3111.2
.0
85.4
73.4
22.8
38.9
2046.7
965.7
2057.9
.0
75.9
54.0
16.7
33.6
.0 .0
463.0 6388.9
623.2
344.6
687.5
,0
48.4
36.6
11.3
25.5
8362.1
4474.9
7484.6
.0
79.0
63.5
19.8
35.2
Ifc
I
-------
Appendix B
WATER QUALITY DATA
A major element in calculating the cost of a regulation
is the determination of the number of water utilities which
will have to add to or alter their treatment practices to
comply with the regulatory standard. Data from the National
Organic Monitoring Surveys (NOMS) have been used in this
analysis for determining the number of and extent to which
systems exceed a given MCL. These surveys were conducted by
EPA's Office of Water Supply, Technical Support Division, and '
the Municipal Environmental Research Laboratory. The results
yielded data on a broad range of known and suspected organic
contaminants in drinking water, including THMs.
The data for the various rounds of NOMS at each site were
averaged for use in this analysis. The combination provides a
proxy for the annual average data likely to be found by water
systems complying with the monitoring requirements of a regula-
tion. The first exhibit shows this combined data in detail.
For example, on that basis, 64.2 percent of the surface water
systems and "81.7 percent of the groundwater systems showed THM
levels of less than 0.10 milligrams per liter.
The combined data were analyzed to estimate the number of
systems exceeding certain levels of THM contamination. The
second exhibit in this chapter displays the number of systems
in several system size categories which are estimated to exceed
the THM level of 0.10 milligrams per liter. Of the 515 larger
systems also exceeding this level, 420 are in the 10,000 to
75,000 size category and 95 serve more than 75,000 people.
-------
B-2
Exhibit 3-1
THM CONCENTRATION BASED ON MOMS DATA
SURFACE WATER
Percent shows cumulative portion
of water systems with THM levels
at or below number indicated.
79.6%
92.1%
2.3%
34.1%
0.01 0.05 0.10 0.15
THM CONCENTRATION IN MILLIGRAMS PER LITER
-
0.25
GROUNDWATER
Percent shows cumulative portion
of water systems with THM levels
at or below number indicated.
68.4%
28.3%
81.7%
88.3%
O.OI 0.05 0.10 0.15 0.25
THM CONCENTRATION IN MILLIGRAMS PER LITER .
SOURCE: THE NATIONAL ORGANIC MONITORING SURVEY PROVIDED 3Y EPA-MERL.
-------
Exhibit B-2
DISTRIBUTION OF WATER SYSTEMS BY WATER SOURCE,
DISINFECTION PRACTICE, AND THM LEVEL*
System Size by
Population Served
Over 1 mil lion
100,000-1 million
75,000-100,000
10,000-75,000
1,000-10,000
Total Over 10,000
All Systems
S
10
135
59
666
2,181
870
G
1
66
59
1,273
7,313
1,399
P2
0
31
29
356
1,433
416
Total
11
232
147
2,295
10,927
2,685
Systems Which Use
a Disinfectant
S
10
134
59
612
2,138
815
G
1
60
59
1,045
4,865
1,165
Total
11
194
118
1,657
7,003
1,980
Systems Which Disinfect
and Have THM Levels >
0.10 Milligrams Per Liter
S
4
48
21
219
765
292
G
0
11
11
201
934
223
Total
4
59
32
420
1,699
515
CO
1
Figures include anticipated treatment changes due to the Interim Primary Drinking Water Regulations and
growth in the number of systems which serve fewer than 10,000 people.
"Water sold wholesale is assumed to have been treated by its seller. Hence systems which purchase the
majority of their water would not require additional treatment. P = more than 50 percent of the water
distributed by the system is purchased. Similarly, S = > 50 percent is surface water, and G = > 50 percent
groundwater.
Source: Estimates based on inputs from EPA-MERL, EPA-ODW, and TBS.
-------
Appendix C
INDIVIDUAL SYSTEM TREATMENT COSTS
The unit treatment costs used in this analysis were derived
from a computer model containing the data presented in an EPA
publication entitled "Estimating Costs for Water Treatment as a
Function of Size and Treatment Plant Efficiency" (EPA-600/2-78-
182). This document, written by the consulting firm of Gulp/
Wesner/Culp, provides detailed cost data for most major drinking
water treatment processes. The costs for five treatments are
shown in Exhibit C-l.
The model itself requires the specification of plant design
and operating flows, treatment design and operating flows, fac-
tor costs, fees, and cost indices. The model outputs yield to-
tal capital costs before and after fees are added, total oper-
ation and maintenance costs, and costs per thousand gallons for
both capital and 0 & M. A sample output is presented in Exhibit
C-2 for the chlorine/ammoniation treatment for a system serving
100,000 to 1 million people.
To determine the necessary model inputs, meetings were held
with staff members from EPA-MERL, EPA-ODW, C/W/C, and TBS. Based
on these discussions, the individual system costs presented in
Table C-l were derived for the nine system sizes serving popula-
tions greater than 1,000 and five treatments (only the six sizes
representing systems serving more than 10,000 people are pre-
sented) .
The following represent the primary technical assumptions
used in deriving these costs: (a) chlorine/ammoniation dose at
4 milligrams per liter; (b) chlorine dioxide doses at 1.5 milli-
grams per liter; (c) ozone dose at 2 milligrams per liter with a
five-minute detention time; (d) GAC with a nine-minute empty bed
contact time, conversion of existing filter beds, 360-day regen-
eration cycle, and multiple hearth regeneration furnaces with
some of the smaller systems using regional regeneration facili-
ties; and (e) GAC plus ozone which combines the assumptions for
GAC and ozone with the exception that a 720-day regeneration
cycle is used.
-------
Exhibit C-l
COSTS FOR A TYPICAL HATER SYSTEM
FOR SELECTED TREATMENTS
(1980 dollars)
Average Population Served
Per System
Ozone
Capital Expenditures
Annual 0 & M Expense
Annual Per Capita Cost
Chlorine Dioxide
Capital Expenditures
Annual 0 & M Expense
Annual Per Capita Cost
Chlorination/Ammon latlon
Capital Expenditures
Annual 0 & M Expense
Annual Per Capita Cost
GAC
Capital Expenditures
Annual 0 & M Expense
Annual Per Capita Cost
GAC and Ozone (BAC)
Capital Expenditures
Annual 0 & M Expense
Annual Per Capita Cost
10,000-25,000
17,000
$341,000
17,000
3.00
$ 30,000
17,000
1.20
$ 12,000
4,000
0.30
$435,000
20,000
3.70
$760,000
29,000
6.20
25,000-50,000
37.000
$ 625,000
27,000
2.40
$ 31,000
26,000
0.80
$ 15,000
8,000
0.20
$ 760,000
36,000
3.10
$1,357,000
48,000
5.00
50,000-75,000
63,000
$1,080,000
44,000
2.40
$ 34,000
41,000
0.70
$ 19,000
13,000
0.20
$1,264,000
63,000
3.00
$2.299,000
79,000
5.00
75,000-100,000
93,000
$1,471,000
60,000
2.30
$ 38,000
56.000
0.70
$ 22,000
18,000
0.20
$1,733,000
90,000
2.90
$3,141,000
104,000
4.50
100,000-1 Million
264,000
$2,729,000
148,000
1.60
$ 76,000
148,000
0.60
$ 31,000
51,000
0.20
$5,240,000
269,000
3.00
$7,753,000
307,000
4.20
Over 1 Million
1.223,000
$ 7,161,000
943,000
1.40
$ 362,000
680,000
0.60
$ 61,000
253,000
0.20'
$21,063.000
756,000
2.40
$27,259,000
1,107,000
3.20
o
I
to
Source: Derived using EPA's computer model of unit
EPA-OOW, C/W/C, and TBS.
treatment costs (described 1n EPA's Unit Treatment Cost Report) With Inputs from EPA-MERL,
-------
Exhibit C-2
OOiil iiltMMAKY f"OF< J WAI IK I Kt AI MINI I'KOCF-iili
I'FvOCFItiH
1 AUIIA AHMONEA Ft lilt FACHIIY
* I'JA.OO/MIN F OR AMMONIA
F'l OW MtiH
HE HKiN ACIIIAl
F'KOCE tiii I'AkAME IKE< COi.ilii Hill I AKii
HtiiiON OfEKAILNfi CONiilK CAM IAI
;i(>4.4 I li/IlAY
i H/HAY
JOV40.
COiilii CliJ/1000 OAI
OSM nK.ui io IAI.
:>.:i7o o.o i/
IOIAL
JOV'10.
O.i'/lt <).()!/
iil IL'UOIcKr iilJUiiimF ACf »!ilNHY F'I)WU<
(j[ N CON OH X I'r Jll.OX.
t N(ilN(.tF.()()
iil IF WOKKr [HIE FE I F IIE I F t/UAL = 0.040
NAIDkAl HAi'JF 4>/CU F I --O.OOJO
HI UNI) F.NF kOY UfiE'rKWH/HU F I/Yk - 101?.A
C0!.i I JNUfXFJJ
E.XCAVAIION (E'NFJ KK1LL.FH l.AHI)F!>= D90.2
MANOf ACUIF -- ,<:M.:.'
t AFtl)F<(f:NF< BKI1LEU l.AHOK) •-•- :>VO. i!
firtS « VALVEli (HIS tll4.y()l) - l'yL',4
E."l E'CIK'ICAI. X INSIF< (HIS 111/1) = ty4.('>
HIJIiSINU (ENFJ HlltllUNI) (70KD ' .501.1
K fEJECF. TNHE.:X - :Mr»..i
-------
Appendix D
REFERENCES
Further documentation of previous efforts related to this
economic analysis of the THM regulation can be found in the
references listed below.
1. "Economic Impact Analysis of a Trihalomethane
Regulation for Drinking Water," August 1977,
U.S. EPA Office of Water Supply.
2. "Estimating Costs for Water Treatment as a
Function of Size and Treatment Plant Efficiency,"
August 1978, EPA-600/2-78-182.
3. "Policy Testing Model for Water Utilities—Final
Report" (submitted to U.S. EPA Office of Water
Supply), August 1976, Temple, Barker & Sloane, Inc.
4. "Survey of Operating and Financial Characteristics'
of Community Water Systems," April 1977, U.S. EPA
Office of Water Supply, EPA-570/9-77-003.
5. "Interim Treatment Guide for the Control of Chloro-
form and Other Trihalomethanes," June 1976, EPA-MERL,
Water Supply Research Division.
U.S. GOVERNMENT PRINTING OFFICE: 1979 -Z81-147/141
------- |