United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-82-055 Apr. 1983
Project Summary
Drinking Water
Cost Equations
Paul Dorsey and Robert M. Clark
The Safe Drinking Water Act of
1974 emphasizes the need to consider
costs when promulgating regulations.
Much concern has been expressed
over these costs, particularly as they
relate to technological requirements.
To respond to some of these concerns,
the Drinking Water Research Division
of EPA's Municipal Environmental
Research Laboratory sponsored a re-
search effort to develop cost data (a
base of 99 unit processes) associated
with water supply unit processes.
Before being considered acceptable,
the costs developed by a design con-
sultant were reviewed by a cost-
estimating specialist.
With the use of the cost data devel-
oped in the original report, a general
equation was developed to calculate
regression estimates. With the use of
this form of equation, operating and
maintenance and capital costs (nor-
malized at 8% over a 20-yr amortiza-
tion period) were estimated. In addi-
tion, equations based on operating
and maintenance components (num-
ber of kilowatt hours, labor hours,
etc.) and construction cost compon-
ents (housing, instrumentation, etc.)
were developed.
This Project Summary was devel-
oped by EPA's Municipal Environmen-
tal Research Laboratory, Cincinnati,
OH, to announce key findings of the
research project that is fully docu-
mented in a separate report cf the
same title (see Project Report ordering
information at back).
Introduction
The Safe Drinking Water Act of 1974
emphasizes the need to consider costs
when promulgating regulations. Fre-
quently these regulations become trans-
lated into generalized design requiremnts
for technology rather than into specific
unit process designs. Concern over the
cost of regulatory requirements has
increased general interest in efficient
unit process design, as have concerns
over continuing trends in higher costs
for labor, energy, materials, and other
important input factors.
To respond to some of this concern
over the cost of water supply, the
Drinking Water Research Division of
EPA's Municipal Environmental Re-
search Laboratory sponsored a research
effort to develop cost data associated
with water supply unit processes. The
purpose of the project was to develop
operating and capital costs for water
treatment processes that are sufficiently
accurate for preliminary planning and
flexible enough for cost effectiveness
studies. Before being considered accept-
able, the costs developed for 99 unit
processes by a design consultant were
reviewed by a cost-estimating specialist.
The full report presents equations
derived from the original data and shows
how these equations can be used to
estimate costs for unit processes. Exam-
ples based on a standard layout for a
water supply treatment plant and stan-
dard values for the input factors to the
cost equations are presented.
-------
Need for Predesign Estimates
Although seldom, if ever, are any two
plants exactly the same, two plants with
similar water treatment goals can be
similar and can often be meaningfully
compared if adjusted to a common base.
When considering concepts or de-
signs, frequent economic appraisals are
necessary to eliminate noncost-effective
alternatives and to concentrate research
and engineering efforts into channels
leading to the most attractive end re-
sults. These economic appraisals must
be made without the benefit of firm
estimates based upon detailed design.
When some indication of the cost of
constructing a plant is required, an
individual closely concerned with the
development or a specialist with pre-
vious estimating experience and a refer-
ence file of cost data often prepares an
estimate. The dependability of the result
and the amount of time spent in devel-
oping the necessary data vary widely.
These estimates, broadly termed the
"predesign" estimate, are useful for
guiding research into examining the
most desirable of several process or
design alternatives. Only when the
choice between alternatives is difficult
are complete designs and price quota-
tion needed to select the proper tech-
nology.
Cost Estimation for Research
Many water treatment processes orig-
inate in the laboratory and are tested
through field-scale pilot plant studies.
When a new process is considered for
field-scale evaluation, the question of
the cost of this process in full-scale
application arises. The data presented
in this paper can be used to answer this
question. By disclosing the most costly
features of the process, a cost estimate
may suggest some specific areas for
further study.
Preliminary Engineering
Preliminary engineering to translate
laboratory and pilot plant data into
equipment design, piping, layout, build-
ings, etc., can start as soon as a project
appears attractive. The first step should
be a detailed process flow sheet with
complete material and energy balances.
Preliminary selection of types and
sizes of equipment can be made from
the flow sheet. In some cases the most
favorable equipment can be chosen by
comparing purchase prices of equipment
performing the same function. In other
cases the direct operating and installed
costs of competitive equipment must be
considered. Simple formulas can be
developed to assist in making these
comparisons. Such formulas and their
applications are given in the full report.
Before preliminary engineering can
be completed and construction can be
started, all decisions must have been
made between competitive processes,
raw materials, plant sites, scale of pro-
duction, etc. Cost considerations are
extremely important in these choices.
Simple factors will dictate most funda-
mental decisions, but somewhere in the
selection process may be a choice
between two courses of action in which
nearly all factors must be considered in
a "full-scale" estimate.
Cost estimates have their greatest
utility for these kinds of decisions. In
other words, after all other factors are
considered, the final choice of unit
processes should be based on the most
economically attractive alternatives. The
considerations involved are nearly the
same as those for evaluating the desir-
ability of construction.
Approval of Construction
In the final analysis, the decision of
whether to build a treatment facility
involves many factors that must be
weighed judgmentally:
1. total capital outlay,
2. estimated operating costs,
3. terminal production costs,
4. amount of water to be produced
per year,
5. quality specifications required,
6. processes to be used,
7. amount of waste generated,
8. location of the plant, and
9. support facilities required.
Many times misunderstandings in
comparative costs can be eliminated by
following the steps outlined above. For
example, two treatment plants, one with
sludge handling facilities and one with-
out, can have quite different costs.
Understanding which design variables
have the greatest impact on cost is often
important. The equations presented in
this report can therefore frequently be
used to perform sensitivity tests.
Reliability of Estimates
The factors affecting the reliability of
cost estimates include availability of
basic data, state of development, defi-
nition of scope, the time expended on
analysis, and experience of the analyst.
Figure 1 illustrates the effect these
factors can have on capital costs for
projected and actual facilities. These
same factors apply to estimating opera-
ting costs as well. Table 1 summarizes
the characteristics, purpose, and reliabil-
ity of five basic types of cost estimates.
Development of Cost Data
Data for this study were derived from
a carefully designed research project
study to establish water supply unit
process cost curves on a consistent and
understandable basis. The study focused
on (1) processes for removing from
potable water those contaminants in-
cluded in the National Interim Primary
Drinking Water Regulations (NIPDWR)
and (2) development of construciton and
operation and maintenance cost curves
for these processes. The final data base
contained cost curves for 99 unit pro-
cesses and a computer program for
retrieving, updating, and combining the
cost data. Characteristics of these data
are described in the following sections.
Derivation of Cost Curves
The construction cost for each unit
process considered in the original re-
search project is presented as a function
of the process design parameter that
was determined to be the most useful
and flexible under varying conditions.
Variables such as loading rate, detention
time, or other conditions that can vary
because of designer's preference or
regulatory agency requirements were
used. For example, for granular activated
carbon (GAC), concrete contactor con-
struction cost curves are presented in
terms of cubic feet of contactor volume;
this approach allows various empty bed
contact times to be evaluated. Contactor
operation and maintenance cost curves
are presented in terms of square feet of
surface area because operation and
maintenance requirements are more
appropriately related to surface area
than to contactor volume. Reactivation
facility cost curves are presented in
terms of square feet of hearth area for
the multiple hearth furnace and pounds
per day of reactivation capacity for the
other reactivation technologies consid-
ered. This approach provides greater
flexibility in the use of the cost curves
than if the costs were related to water
flow through the treatment plant.
-------
Q
•g
1
*M
2
£5
c
o
§
ex.
s*.
O
• High Estimate -
For workable but overly complex process
loaded with extras, add-ons, excessive safety features
5* *Numbers refer to types of estimates
defined in Table 1
•Cost of later i same location
plants built \ same sc°Pe
Capital cost decreases
in accordance with
the "learning curve"
• Low Estimate -
based on insufficient information
(early stage of development)
I
• Estimated •
I
• Actual'
i
\
n+1
n+2
n+3
Figure 1.
n+4
Relative Time, years
Effect of various factors on projected and actual capital costs.
n+5
n+6
n+7
/7+S
Figures 2 and 3 show typical construc-
tion cost curves and some of the opera-
ting and maintenance cost components
for pressure carbon contactors. A fea-
ture common to many of the curves is a
decrease in slope of the curve at smaller
size ranges; Figure 3 illustrates this
effect. This feature is significant in
deriving the equations.
Methods Used to Develop Cost
Curves
The construction cost curves were
developed with the use of equipment
cost data supplied by manufacturers,
cost data from actual plant construction,
unit takeoffs from actual and conceptual
designs, and published data. An ap-
proach similar to that utilized by a
general contractor in determining a
construction bid was then used to check
these cost curves. Any discrepancies
that existed between the initial estimate
and the second estimate were then
resolved.
Construction Cost
Components
The costs for eight principal construc-
tion components were developed and
then aggregated to give the construction
cost for each unit process. The eight
components are (1) excavation and
sitework, (2) manufactured equipment,
(3) concrete, (4) steel, (5) labor, (6) pipe
and valves, (7) electrical and instrumen-
tation, and (8) housing. These categories
also provide enough detailed informa-
tion to permit accurate cost updating.
The construction cost curves are not
the final capital cost for the unit process
because these curves do not include
costs for general contractor overhead
and profit, administration, engineering
and legal fees, fiscal determinations,
and interest during construction. These
items are more directly related to the
total cost of a project than to the cost of
the individual unit processes. Therefore,
if more than one unit process is required,
they are more appropriately added after
the costs of the individual unit processes
are totaled.
Operation and Maintenance
Cost Components
Operation and maintenance require-
ments were developed for building-
related energy, process energy, main-
tenance material, and labor. The separate
determination of building energy allows
for regional variations. Energy require-
ments are presented in kilowatt-hours
per year for electricity, standard cubic
feet per year for natural gas, and gal Ions
per year for diesel fuel. Labor is pre-
sented in hours per year, allowing local
variations to be incorporated into the
operation and maintenance cost calcu-
lations. Maintenance material cost,
given in dollars per year, does not
include the cost of chemicals. Chemical
costs are added separately as shown in
later examples.
In the original study, for the majority
of the 99 unit processes, three separate
-------
Table 1. Definition of Five Basic Types of Estimates of Total Plant Cost
Estimate Type (each has
several designations)
1. Order-of -magnitude ratio
Characteristics
Rapid. Very rough.
Purpose
Preliminary indication. Check on
result by more detailed method.
Usual
Reliability
About + 30%
to -60%
2. Study (commonly a so-called
factor estimate}
3. Preliminary budget
authorization
4. Definitive project control
5. Detailed firm contractor's
Flow diagram, material and
energy balance, type and size of
equipment.
In addition to above, surveys and
some engineering of foundations.
transportation facilities, buildings,
structures, lighting, etc.
More detailed engineering, but
usually short of complete specifica-
tions and working drawings.
Requires experienced estimating
organization and substantial outlay.
Complete site surveys, specifica-
tions, working drawings.
For generalized evaluations. ± 30%
Guidance for further investigation.
Basis for process selection.
R&D guidance.
Basis for decision to undertake ± 20%
detailed engineering. Sometimes
basis for budget authorization.
Can be for generalized evaluation,
but usually for site specific
installation.
Sometimes the basis for budget ± 10%
authorization. Provides improved
estimate of project to be built. For
site specific installations.
Made to control cost of project ± 5%
being built. For site specific
installations.
graphs were used to present construc-
tion and operation and maintenance
curves. The first graph presented the
construction cost, the second graph
presented energy (electrical, natural
gas, and diesel fuel) and maintenance
material requirements, and the third
graph presented labor requirements and
total operation and maintenance cost.
Updating Construction Cost
Curves
For many engineering purposes, a
single cost index such as the Engineer-
ing News Record (ENR) Construction
Cost Index (CCI) can be used to update
construction costs. When this approach
«j
2
-------
100.000 1,000,000
10,000 100,000
o>
u
I
J
1,000 10,000
1,000
Bui/ding Energy
Maintenance
Material
Process Energy
100
1,000 10.000
Total Contactor Area - ft2
10
100 1,000
Total Contactor Area - m2
Figure 3. Operation and maintenance requirements for pressure carbon contac-
tors: building energy, maintenance material, and process energy.
ical specificity of the BLS indices and
the additional effort involved in using
seven additional indices rather than a
single index. The approach used to
develop the cost data provides the user
with a choice of updating procedures:
(1) The individual cost indices can be
applied to the components, and the
component costs can then be added to
obtain the unit process cost. (2) As an
alternative, the total unit process cost
can be updated in a single step by
applying the ENR Construction Cost
Index to the sum of the eight cost
components.
Updating Operation and
Maintenance Costs
Operation and maintenance costs are
updated somewhat differently than are
construction costs. To update energy
and labor costs, the prevailing unit costs
for energy and labor are used. Main-
tenance material costs that do not
include chemicals are best updated with
the use of the new Producer Price Index
for Finished Goods.
Data Manipulation
In the original project, a computer
program was developed to allow costs to
be retrieved, updated, and analyzed. The
computer program and data file that
contain the construction costs and
operation and maintenance require-
ments are presented in the full report.
A number of other factors can be
varied in addition to the indices used for
updating: (1) the operating parameter
for the process, (2) sitework and inter-
face piping cost, (3) special subsurface
requirements, (4) standby power require-
ments, (5) general contractor's overhead
and profit percentage, (6) engineering
costs, (7) land cost, (8) legal, fiscal, and
administrative costs, and (9) interest
during construction and capital cost
amortization.
The computer program, written in
Fortran, was designed to operate on the
PDP-11/70 computer system. Although
a compact and efficiently written pro-
gram, it still requires a relatively large
amount of memory and must be com-
piled on a specific computer before it
may be used.
To improve the usefulness of the cost
data, a series of analytic equations were
developed for each of the u nit processes
considered in the full report where the
equation development is described.
Estimates of Construction and
O&M Cost Components
Often in updating or checking cost
estimates, it is useful to know the
quantities of the individual components
that makeup these estimates. In the
following sections, a methodology is
presented for providing these kinds of
estimates.
Construction Cost
Components
The report presents construction cost
distribution factors based on the follow-
ing components: excavation; manufac-
tured equipment; concrete; steel; labor;.
pipe and valves; electrical and instru-
mentation; and housing. For example,
the construction cost for Chlorine Feed-
Cylinder Storage is as follows: 0%
excavation, 50% manufactured equip-
ment, 0% concrete, 0% steel, 6% labor,
4% pipes and valves, 6% electrical and
instrumentation, and 34% housing.
These factors are useful for updating
-------
Table 2. Indices for Construction Costs
Cost Component
Type of Index
October 1978
Value of Index
Excavation and sitework
Manufactured equipment
Concrete
Steel
Labor
Pipe and valves
Electrical equipment
and instrumentation
Housing
ENR Wage Index (SkilledLabor) 247
BLS General Purpose Machinery 221.3
and Equipment - Code 114
BLS Concrete Ingredients -
Code 132 221.1
BLS Steel Mill Products -
Code 1013 262.1
ENR Wage Index (Skilled Labor) 247
BLS Valves & Fittings -
Code 114901 236.4
BLS Electrical Machinery and
Equipment - Code 117 167.5
ENR Building Cost Index 254.76
costs when various factors escalate at
different rates. They can also be used to
compare cost estimates from different
sources.
Operating and Maintenance
Components
Operating and maintenance compo-
nents used in developing these cost
curves are: process electricity in kilowatt
hours per year; maintenance material in
dollars per year; labor in hours per year;
building energy in kilowatt hours per
year; natural gas in standard cubic feet
per year; and diesel fuel in gallons per
year. Equations have been developed so
that the amount of each component is
estimated based on the magnitude of
the process variable.
Computer Program
A computer program was written to
provide the capability for easy cost
calculation, and examples have been
developed in the full report.
Summary and Conclusions
Cost estimating can serve a number
of purposes. Predesign estimates can be
used to compare the cost of two treat-
ment plants with similar water treat-
ment goals. These types of estimates
can be used to choose the most cost
effective treatment alternative. Cost
estimating for research can be used to
assist in making choices among new
and developing technologies and to
suggest specific areas for future re-
search. Reliability of cost estimates is
important to make proper judgments
with regard to the type of basic cost
estimate to be considered.
The full report presents a series of
cost equations for conventional water
treatment unit processes that are accur-
ate to approximately three significant
figures. The equations, within their level
of accuracy, can be usedfor making engi-
neering cost estimates.
The EPA authors Paul Dorsey and Robert M. Clark (also the EPA Project Officer,
see below) are with the Municipal Environmental Research Laboratory,
Cincinnati, OH 45268.
The complete report, entitled "Drinking Water Cost Equations," (Order No. PB
83-181 826: Cost. $16.00, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
•&U. S. GOVERNMENT PRINTING OFFICE: 1983/659-095/1933
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
RETURN POSTAGE GUARANTEED
_. . _.
Third-Class
Bulk Rate
LOU « TILLEY
REGION V EPA
LIBRARIAN
230 S DEARBORN ST
CHICAGO IL 60604
------- |