Pope - Reid Associates Inc.
Environmental Engineering
245 E. 6th St., Suite 813 . St. Paul, MN 55101 . 612/227/6500
January 4, 1983
Mr. Harold D. Lester
U.S. Environmental Protection Agency
Office of Analysis and Evaluation
401 M Street, S.W. (WH-586)
Washington, D.C. 20460
Subject: Final Report for EPA Contract No. 68-01-6404,
Work Assignment No. 12
Dear Chip:
Enclosed are five copies of the final report for the above-
referenced contract and Work Assignment. The report is en-
, titled "Review .of Two NPDES Permit Assistance Manuals."
As we discussed on the telephone last week, we will not do
the case study until after you have reviewed this report.
If you have any questions concerning the report, please give
me a call.
Sincerely,
. U)(|jxawi>
Jean L. Williams
Project Manager
JLW/bl
cc: Negotiator
Patent Advisor
Library
Project File - 105.04 Wa 12


-------
ffA- Oil^
fD. Lester
onmental Protection Agency
Analysis and Evaluation
et, S.W. (WH-586)
, D.C. 20460
rinal Report for EPA Contract No. 68-01-6404,
Work Assignment No. 12
Property Cf u c p.. •
Agency fCtT-
y 0f<2r>' Ml-,08
ftn
5n i i »« >r\
earns, wa 93ioi
-CD 2 6 1991
1200 Sixth Avenue,'c
Pope-Reid Associates Inc. • 245 East 6th Street. Suite 813 • St Paul. Minnesota 55101

-------
TABLE OF CONTENTS
Page
INTRODUCTION	1
RECOMMENDATIONS	2
Cost Analysis Manual	2
Economic Achievability Protocol	2
REVIEW OF COST ANALYSIS MANUAL	5
Introduction	5
Summary of the Cost Analysis Manual	5
Capital Cost Estimates	7
Annual Expense Estimates	14
Data Sources	21
Annotated Bibliography	21
REVIEW OF ECONOMIC ACHIEVABILITY PROTOCOL	24
Introduction	24
Summary of the Economic Achievability Protocol	24
Firm Level Analysis	26
Plant Level Analysis	38
Analysis of the Economic Achievability Tests	44
INTERFACING OF THE MANUALS	50
REFERENCES	51
APPENDIX - Preparation of an Annotated Bibliography	52

-------
LIST OF TABLES
Page
Table
1.
Lang Method
8
Table
2.
Chilton Method
0
«/
Table
3.
Guthrie Method
10
Table
4.
ICARUS Method
11
Tab! e
5.
Unit Process Method
13
Table
6.
Recommended Modifications to the
Annual Expense Estimate Methodology
15
Table
7.
Data Needs for the Plant Level Analysis
43

-------
INTRODUCTION
The U.S. Environmental Protection Agency (EPA) is currently in the process
of preparing "second-round" NPDES permits. These include both BAT permits
for which effluent guidelines have been proposed or promulgated, and BEJ
(Best Engineering Judgement) permits for point sources that have no appli-
cable national effluent guidelines. The wastewater treatment technology
on which the BEJ permit is based must be shown by the permit writer to be
economically achievable for the facility receiving the permit. To do this,
the permit writer must first calculate the cost of the appropriate waste-
water treatment technology and then determine if it is economically achiev-
able .
Other offices within EPA have prepared two analytical manuals which are in-
tended as guidance to those state and regional office staff who are pre-
paring "second-round" NPDES permits. These manuals are:
e A Standard Procedure of Cost Analysis of Pollution Control Operations;
and
e Protocol for Determining Economic Achievabi1ity for NPDES Permits.
This report summarizes the results of a review by Pope-Reid Associates, Inc.
(PRA) of these two manuals. The review was based on a two-fold approach that
both evaluated the documents individually and assessed the feasibility of us-
ing the documents in conjunction with each other. Each of the manuals was re-
viewed with respect to:
•	the soundness of the methodology;
e the type(s) of information and data needed;
•	the assumptions employed; and
0 ease of use.
The manuals were also reviewed in terms of their compatibility with each other.
The format of this report reflects this two-fold approach. The two sections
which follow the recommendations present the results of PRA's review of the
Cost Analysis Manual and the Economic Achievabi1ity Protocol. The next
section discusses the use of the two manuals together. Following it is a
reference section and an appendix which contains a description of the prepara-
tion of an annotated bibliography.

-------
RECOMMENDATIONS
This section contains a summary of PRA's recommendations regarding the Cost
Analysis Manual and the Economic Achievabi1ity Protocol. The recommenations
are described in more detail in the next three sections of the report.
Cost Analysis Manual
1.	Some of the values/costs of factors (e.g., labor, land, laboratory, and
waste disposal costs) have changed since the Manual was written in 1977.
These factors should be reviewed and updated to reflect more current
values if necessary.
2.	The Cost Analysis Manual seems to be better suited to calculate capital
and 0 & M costs for chemical processing facilities in general. It should
be revised to reflect cost estimating for wastewater treatment systems.
This will include deleting those sections which are not applicable to
wastewater treatment facilities and orienting the focus of the Manual to
the needs of BEJ permit writers.
3.	The section on estimating costs to retrofit a facility should be expanded.
4.	New examples should be provided which reflect cost estimates for waste-
water treatment facilities.
5.	Information on a given topic in the present Cost Estimating Manual is
scattered throughout the two volumes. The format of the Manual should
be revised so that step-by-step calculations are shown for each relevant
cost estimating methodology.
6.	The bibliography/reference sections should be updated, expanded, and
changed in format to an annotated bibliography.
Economic Achievabi1ity Protocol
Firm Level Analysis
1.	EPA's policy on economic achievabi1ity should be clearly defined in the
Protocol.
2.	Procedures should be added to evaluate the effect of BEJ investments and
costs on the profitability of a firm.
3. Prcdedures/should be added to evaluate the effects of BEJ investments and
costs on competition with substitute products and products produced in
other countries.

-------
4.	Procedures should be adapted or supplemented to provide insight on sources
of capital available to a firm.
5.	The assumption used in several of the tests (which assumes that BEJ costs
would be paid for from current assets, earnings, or cash flow) should be
reconsidered. This assumption may be more conservative than necessary.
6.	Each test should incorporate data for three to five years. Clear guide-
lines should be provided for evaluating historical trend data.
7.	Each test should be adapted to consider the differential effects on small
firms. This is one of the most critical problems with this analysis.
8.	Clear explanations should be provided for each test. Explanations should
include definitions of all relevant terms, statements of purpose for
each test, critical values to indicate whether a firm passes or fails a
test, and step-by-step procedures for locating data and performing cal-
culations .
9.	Data items such as interest rates and investment tax credits should be
specified as variables to be supplied by the permit writer. Clear in-
structions for selecting the correct data should be provided.
10.	Errors in the explanations and sample calculations (identified in the
description of each test) should be corrected.
11.	The fixed charge coverage ratio and Beaver's ratio should be modified or
replaced. These tests do not provide adequate insight into the ability
of a firm to meet its long-term obligations.
12.	Beaver's ratio should not be relied on as a predictor of bankruptcy. If
bankruptcy prediction is an integral part of the economic achievability
analysis, a more effective procedure should be chosen.
13.	The market value analysis should be eliminated from the methodology.
Plant Level Analysis
1.	Additional guidance should be given to the permit writer for purposes of
determining the appropriate SIC code for a plant when the plant belongs to
more than one SIC.
2.	The Protocol should be modified to address circumstances in which part of
a plant's wastewater is regulated by effluent limitations guidelines and the
remaining wastewater is to be controlled by a BEJ permit.
3.	Consideration of BCT cost-effectiveness criteria should be included in the
Protocol to facilitate estimating costs for plants that also discharge con-
ventional pollutants.
4.	The three plant level tests rely on financial data from the most recent
fiscal year of a plant. The tests should be applied for the three most
recent fiscal years.
-3-

-------
5.	More guidance should be given when the three tests do not provide
definite conclusions concerning the economic achievabi1ity of BEJ
technology.
6.	The utility of the Gross Margin and Revenues tests should be recon-
sidered to determine if they provide additional information about the
economic achievability of BEJ technology.
7.	The Protocol should be revised to identify a source of the interest
rate, (or discount factor). The interest rate is needed to calculate
the total annual costs of a BEJ technology.
-4-

-------
REVIEW OF COST ANALYSIS MANUAL
Introducti on
This section contains the results of PRA's review of A Standard Procedure
for Cost Analysis of Pollution Control Operations. In reviewing this docu-
ment, PRA concentrated on evaluating the following:
•	the timeliness of the engineering assumptions used;
•	the soundness of the cost estimating methodology,
o the ease with which the Manual can be used by permit writers;
o the flexibility and accuracy of the Manual when applied to a variety
of plant situations; and
t the compatibility of the Manual with the Economic Achievabil it.y
Protocol.
Summary of the Cost Analysis Manual
The document A Standard Procedure for Cost Analysis of Pollution Control
Operations presents a standardized procedure for preparing engineering cost
estimates and economic evaluations of chemical processing facilities. This
would include pollution control operations, although the Cost Estimating
Manual is much more generalized in scope. The procedure is applicable to
projects in various economic sectors (i.e., private, regulated, and public).
The Manual presents a recommended format, termed the specification, which is
used to organize the information for the economic evaluation into three seg-
ments: the descriptive segment, the cost analysis segment, and the relia-
bility assessment. The Manual also contains guidelines to aid in the selection
of financial and operating factors and to establish the level of detail re-
qui red.
The descriptive segment contains brief descriptions of five items: facility
description; capacity rating; abstract of the scope of the project; perfor-
mance specification; and stage of development.
The cost analysis segment is made up of the three elements -- the specified
parameters, the cost estimate, and the feasibility evaluation. The specified
parameters include:
-5-

-------
9	the interest (discount) rate;
s	the facility life and depreciation period;
•	the construction time;
o	a reference unit for process costs;
e a cost index; and
e an inflation rate.
The cost estimate includes capital investment, annual expenses, and profit
and cash flow. The following methods are presented for determining capital in-
vestmant at the study or preliminary estimate level: Lang, Chilton, Guthrie,
ICARUS, and Unit Process. The cost data and any expected revenues are used to
calculate several measures of merit which represent the criteria for assessing
economic feasibility. The Manual discusses the following measures of merit:
s return on investment (ROI);
•	internal rate of return (IROR);
« payout time;
e equivalent annual cost; and
•	unit costs.
The third segment of the Manual describes methods for assessing the reliability
of the measures of merit.
The bulk of the Cost Analysis Manual consists of 11 appendices that provide de-
tailed background material and two comprehensive examples. The appendix sub-
jects are:
•	Capital Investment Estimation;
e Annual Expense Estimate;,
o	The Cash Flow Concept;
e	Discrete and Continuous Interest Factors;
•	Measures of Merit;
•	Cost Indices and Inflation Factors;
e	Rates of Return and Interest Rates;
e	Methods of Reliability Assessment;
•	Sensitivity Analysis;
e	ExampTe I -- Cost Analysis of Flue Gas Desulfurization (FGD) Retrofit
Facility; and
o	Example II -- Cost Analysis of Chlorolysis Plant.
-6-

-------
Capital Cost Estimates
Capital cost estimates will differ in the data and other resources required to
prepare them; in the experience/knowledge required of the analyst; and in the
accuracy of the results obtained. The Cost Analysis Manual discusses five
different cost estimating methodologies, which vary in terms of the accuracy
that can be obtained. In general, estimates which are the most accurate also
require the most data. The five methodologies discussed in the manual, from
the least to the most complex, are:
e	order-of-magnitude/ratio estimate;
e	study (or factored) estimate;
o	preliminary/budget authorization estimate;
e	definitive/project control estimate; and
6	detailed/firm contractor's estimate.
The first three are also called "conceptual estimates" and are the most suit-
able to use in preparing BEJ permit cost estimates. Of these three, the study
estimate will probably be most likely to provide the information required by
writers of BEJ permits.
Study estimates can be either "factored" or "unit process" estimates. Four
methods of developing factored estimates are discussed in the Manual--those of
Lang, Chilton, Guthrie, and ICARUS. In the Lang method (Table 1), the sum of
the delivered costs of the major pieces of plant equipment is multiplied by a
single factor to obtain the plant capital cost. The factor used depends only
on whether the plant processes primarily solids or fluids. In the Chilton
method (Table 2), an installation factor is applied to the sum of the delivered
prices of major plant equipment and then additional factors are used to obtain
the costs of auxiliaries. A common variant, not discussed in the Manual, is
to apply separate installation factors to the purchase cost of each piece of
major equipment, sum these installed costs, and then apply additional factors
for the costs of auxiliaries. Guthrie's method (Table 3) increases the com-
plexity and accuracy of this variant by applying individual factors for materials
and labor for installation and auxiliaries to the purchased price of each piece
of major equipment and then summing these costs. The ICARUS method (Table 4)
is a simplification of the Guthrie method.
-7-

-------
Table 1. Lang Method
Use Equation: Ip =(( E
Where:	Ip = total plant cost (total module cost; fixed capital
investment for equipment, buildings, site development);
^E = Sum of major plant items (MPIs),£e del i vered^ ; and
L = Lang factor.
1.	Total pi ant cost
2.	Interest during construction, if applicable, and capitalized
3.	Modification of the facilities and start-up costs, if capitalized
4.	Total depreciable investment
5.	Land
6.	Working capital
7.	Total capital investment
(a) • Same as FOB job site
-8-

-------
Table 2. Chilton Method
(Factored costs of sum of major plant items (MPIs), ^E, delivered)
Operating	Cost
Item	Factor	On	of Item
1.	Sum of major plant items
(MPIs), £E, delivered'3)
2.	Installed, erected equipment cost	#1
3.	Piping (includes insulation)	#2
4.	Instrumentation	#2
5.	Buildings and site development	t?2
6.	Auxiliaries (electric, steam, etc.)	#2
7.	Other	#2
8.	Total physical cost (Direct cost), DC
(sum of 2 through 7)
9.	Indirect cost (20 to 50% of DC, avg
34%), IC
10.	Total bare module cost, BMC
11.	Contingency (10 to 50% of BMC; avg
15%)
12'. Contractor's fee (about 3% of BMC)
13.	Total plant cost (Total module cost), Ip
14.	Interest during construction, if
applicable, and capitalized
15.	Modification of the facilities and
start-up costs, if capitalized
16.	Total depreciable investment
17.	Land
18.	Working capital
19.	Total capital investment
(a) Same as FOB job site
_o_
j

-------
Table 3. Guthrie Method
Sum of each MP I^a ^ (this includes adjuncts, such as solids handling
facilities, site development, industrial buildings, off-site facilities)
1.	MPIs, purchased^9^£E
2.	Direct (field) material, m
3.	Direct.(field) labor, L
4.	Sum of direct costs, (Total physical cost) DC, for each MPI and adjunct
5.	Indirect cost (20 to 50% of DC, ave 34%), IC
6.	Total bare module cost, BMC
7.	Contingency (10 to 50% of BMC; avg 15%)
8.	Contractor's fee (about 3% of BMC)
9.	Total plant cost (Total module cost), Ip
10.	Interest during construction, if applicable, and capitalized
11.	Modification of the facilities and start-up costs, if capitalized
12.	Total depreciable investment
13.	Land
14.	Working capital
15.	Total capital investment
(a) Major plant items (MPIs)
-10-

-------
Table 4. ICARUS Method
1.	Sum of installed costs for MPIs - includes indirect costs associated with
each i tem
2.	Total of special items (solids handling facilities, site development, in-
dustrial . bui ldi ngs , off-site facilities)
3.	Base plant cost (Total bare module cost), BMC
4.	Contingency (10 to 50% of BMC; avg 15%)
5.	Contractor's fee (about 31 of BMC)
6.	Retrofit increment (if applicable)
7.	Total plant cost, Ip
8.	Interest'during construction, if applicable, and capitalized
9.	Modification of the facilities and start-up costs, if capitalalized
10.	Total depreciable investment
11.	Land
12.	Working capital
13.	Total capital investment
11-

-------
The Manual considers the Lang method, which is the simplest of the three, to
be the least reliable and useful only for checking the results calculated by
one of the other three methods. In fact, the basis of the Lang method has
(4)
been questioned in the recent literature . The Chilton method, the more
complex Guthrie method, and the ICARUS method are all recommended by the
Manual and enjoy wide acceptance. However, the sources of data that can be
used with the Guthrie and ICARUS methods are limited, consisting chiefly of
(2)
Guthrie's own works and the EPA report prepared by ICARUS . There are
(8 9)
many recent sources of data for the Chilton method ' .
Another method of preparing study estimates of capital costs is discussed in
Appendix A of the Manual. This is the Unit Process method (Table 5), which is
also satisfactory and enjoys wide acceptance. The Manual is in error, however,
in stating that "At present this method seems applicable only to liquid waste
treatment where the facilities are analogous to large sewage treatment plants."
A modification of the Unit Process method involving installed costs has been
(6)
used for chemical plant equipment by Dryden and Furlow and to some extent
by Peters and Timmerhaus^, Perry^10^, and others, although the most tyDical
(35)
applications have been to large sewage treatment plants ' . In this
modification, the costs of piping and other auxiliaries are estimated by factors
applied to the sum of the installed equipment costs as in the Chilton method.
The Chilton and Unit Process methods are the most suitable to calculate the
capital cost for BEJ permits. These methods are relatively simple and the data
needed are readily available. The Manual should be revised to emphasize these
two methods, explain the differences between them, and¦recommend sources of
data for each by means of an annotated bibliography. The information on these
two methods should be consolidated. Other methods should be deemphasized or
deleted.
In addition to the plant capital cost, there are additional capital requirements
which must be addressed, including land, funds required during construction and
startup, and working capital. The discussion in the Manual is adequate but should
be consolidated and in some cases updated.
-12-

-------
Table 5. Unit Process Method
(Sum of unit process modules)
Process Module Identification	Total Cost
1.	Process	module No.	I -
2.	Process	module No.	II -
3.	Process	module No.	Ill -
4.	Process	module No.	IV -
5.	Process	module No.	V -
6.	Process	module No.	VI -
7.	Total plant cost, Ip
8.	Interest during construction,
if applicable, and capitalized
9.	Modification of the facilities and
start-up costs, if capitalized
10.	Total depreciable investment
11.	Land
12.	Working capital
13.	Total' capital investment
-13-

-------
Many facilities will need to add on or retrofit water pollution control systems
to the basic plant in order to comply with their BEJ permit requirements.
Retrofitting a plant with pollution control equipment often costs more than in-
stalling the same equipment in a new plant or expansion. Process modifications
and structural modifications which may have been otherwise unnecessary, space
constraints, utility expansions, and lost production account for a large por-
tion of this difference. The Cost Analysis Manual contains only a brief dis-
cussion of this topic. The discussion and recommendations summarized in the
Manual were developed from the retrofit of flue gas desul furization units on
coal-fired utility boilers. The viability of extending the flue gas desulfuri-
zation retrofit factors to the situations facing a writer of BEJ permits is
questionable. The Manual should therefore include more information on retrofit
and process modification cost/factors. A section discussing these topics, with a
more extensive listing of cost factors (based perhaps on case studies of waste-
water treatment system installation), would be useful to the permit writer.
Annual Expense Estimates
The Manual summarizes two methods for estimating annual expenses:
•	adaptation of actual costs; and
•	factored expense estimates.
The first method requires records of actual costs of similar or identical opera-
tions and an understanding of the rationale which was used to allocate indirect
costs (e.g., labor additives, plant overhead, general expenses, etc.) This
method should be deleted because probability of the two preceeding conditions
being fulfilled is minimal.
The second method, factored expense estimation, should be used by permit writers
to calculate annual expenses. A very good discussion of the items to be in-
cluded in the annual expense estimate is included in the Manual. This dis-
cussion should be modified as shown in Table 6 to make it a more useful tool
in the permit writing process.
-14-

-------
Element of
Annual Operating
Expense Estimate
Raw Materials
Operating Labor
Recommended Modifications to the
Annual Expense Estimate Methodology
Information
Present in
the Manual
General sources of cost
data (i.e., suppliers,
marketing people, pub-
lished prices), freight
charges and by-products
are mentioned
Summarizes three methods
for estimating operating
needs :
•	prepare a schedule of
jobs and functions/
develop labor require-
ment from time segment
data in Haines w);
(12)
s use Wessel'sv equa-
tion which is based on
process type, process
steps, and capacity; and
•	use information from
similar operations.
A table of the 1977 average
hourly earnings of chemical
workers in 10 states
Recommended
Modi fi cati ons
to the Manual
Expand; include a table
to be used as a guide
for checking freight
charges
A source other than Haines
that addresses wastewater
treatment system work and
time segments should be
included; Wessel's equation
is based on labor require-
ments in the chemical pro-
cess industry and should be
deleted; the table of
average hourly earnings
should be updated from 1977
to thepresent; listing
of sources of labor infor-
mation should be included
(i.e., plant union contract:
Bureau of Labor Statistics,
etc.)

-------
El ement of
Annual Operating
Expense Estimate
Direct Supervision
Maintenance Labor a
Materi als
Operating Supplies
Labor Addi ti ves
Recommended Modification to the
Annual Expense Estimate Methodology (continued)
Informati on
Present in
the Manual
Recommends that this be
estimated at 10 to 25 per-
cent of operating labor
Recommends that this be
estimated at 4 to 10 per-
cent of the total plant
cost (35 to 50 percent
for rotating equipment)
Recommends that this be
estimated at 6 percent
of the operating labor or
15 percent of maintenance
costs; special supplies
should be added in
separately
Recommends that fringes
(vacations, disability pay,
pension funds, unemployment
taxes, social security, etc.)
be estimated at 25 to 50 per-
cent of the direct labor
cost
Recommended
Modi fi cations
to the Manual
This should be examined
in more detail and
direction should be given
to aid the permit writer
in choosing the appropri-
ate percentage
Modify this section to
address wastewater treat-
ment systems
Examine this section in
more detail to determine
which factor is more rea-
sonable in the particular
situations faced by permit
writers; provide writer
with narrative to explain
estimate development; in-
clude a list of sources of
operating supply costs
Expand this section to ex-
plain the estimation
process in more detail;
update factor; include
sources of this information
(i.e., union contracts)

-------
Table 6. Recommended Modifications to the
Annual Expense Estimate Methodology (continued)
Element of	Information
Annual Operating	Present in
Expense Estimate	the Manual
Utilities	Provides a short de-
scription of how utility
costs are generally
handled; includes a
table of typical utility
and fuel costs
Effluent Treatment
and Disposal
Provides a short de-
scription of how waste
disposal costs are
generally handled; 1976
costs of sludge ponding
in clay-lined and Hypalon-
lined ponds; 1976 costs for
treatment and disposal
Preparation for	Short description of the
Shipping	cost to prepare a product
for shipping
Recommended
Modi fi cations
to the Manual
Edit the narrative so it
pertains only to the situa-
tions that permit writers
might face; utility and
fuel costs change with
time and vary from region to
region, so the table should
be updated to the present
and expanded to include
regional trends; sources of
utility information should
be included
This section should be ex-
panded and updated to
address the current waste
disposal practices and costs
(conventional and hazardous
wastes); sources which may
be used to obtain waste
treatment and disposal cost
information should be in-
cluded
This should be deleted be-
cause it is not applicable
to wastewater treatment
systems

-------
Element of
Annual Operating
Expense Estimate
Plant Overhead
Control Laboratory
Recommended Modifications to the
Annual Expense Estimate Methodology (continued)
Information
Present in
the Manual
Description of costs that
fall under the general
heading of plant overhead
(i.e., plant management,
personnel, plant protection,
store rooms, roads, sewers,
etc.); recommends that this
cost be estimated as 50 to
100 percent of operating
and maintenance, or a per-
centage of operating labor
plus a percentage of main-
tenance, or a proportion of
both labor (45 to 50 per-
cent) and investment (1 to 5
percent)
Summarizes two methods for
determining control labora-
tory costs:
•	estimate the number of
analysts and multiply this
by $40,000 to $50,000 per
year per person; and
•	take 10 to 20 percent of
the operating labor cost
Recommended
Modifications
to the Manual
This should be examined in
more detail (with respect to
wastewater treatment systems)
and direction should be
given to aid the permit writ-
er in choosing the method and
percentages which are ap-
propriate for the particular
case
Update and expand this sec-
tion so. that it applies to
wastewater treatment systems
(e.g., small companies may
contract this work out)

-------
Table 6. Recommended Modifications to the
Annual Expense Estimate Methodology (continued)
Element of
Annual Operating
Expense Estimate
Technical and Engineering
Informati on
Present in
the Manual
Recommends that these
costs be estimated in
the same manner as con-
trol laboratory cost
($40,000 to $50,000 per
year per person)
Insurance and Taxes
I
lo
Recommends these items be
estimated at 1 and 2 percent
of the fixed capital in-
vestment, respectively
Royalties	Short description of what
royalties are
Recommended
Modifications
to the Manual
Update and expand this
section so that it applies
to wastewater treatment
systems; these costs are
a function of equipment type
size, and system complexity,
etc.
May want to include regional
variations in taxes and
differences in insurance
due to type of facility (e.g
new types of insurance re-
quired for plants genera-
ting, handling, or treating
hazardous wastes)
List some examples of waste-
water treatment equipment
and/or systems that have
royalties associated with
them and provide guidelines
for estimating typical
royalty charges

-------
Table 6. Recommended Modifications to the
Annual Expense Estimate Methodology (continued)
Element of
Annual Operating
Expense Estimate
Depreciation
Information
Present in
the Manual
Describes the estimation
of depreciation by three
methods:
• straight line method;
e sum of digits (accelerated);
and
e double declining balance
(accelerated).
Also mentions that for
pollution abatement equipment,
special rules may sometimes
apply
Recommended
Modi fi cations
to the Manual
Modify this section so it
addresses only depreciation
of wastewater treatment
systems; expand to clarify
the differences between
tax depreciation and
financial depreciation;
include examples of equip-
ment lifetimes for tax and
financial depreciation cal-
culations; update and ex-
pand based on information
in IRS regulations

-------
Data Sources
Equipment vendors, raw material suppliers, and utilities can be invaluable
sources of information in the development of both capital and annual cost
estimates. The Manual states this but does not provide any insight into
how to locate quickly the vendor of a specific type of equipment or raw ma-
terial. Nor does it describe what questions should be asked of the vendor.
The Manual should include a list of sources of vendor information (i.e.,
periodicals, buyers' guides, etc.). It should also include a checklist of
questions to be asked and points to be remembered when dealing with vendors.
Examples of these are:
1.	Ask for a detailed description of the piece of equipment for which
the estimate is being obtained.
2.	Ask if the cost estimate includes such factors as taxes, shipping,
etc.
3.	If possible, obtain a cost estimate from two or more independent
vendors.
Annotated Bibliography
The reference sections in the Manual should be updated and the format re-
vised to make the Manual more useful to permit writers. The reference lists
should be updated from 1977 to the present. In addition, a program should be
instituted to periodically provide updated reference listings to permit writers.
The following sources of cost information should be added to the references.
o Construction Cost Manuals
— Innovative and Alternative Technology Assessment Manual (EPA-430/
9-78-009)
-- Richardson (subscription service)
-- Dodge (subscription service)
-- Means
o Vendor Information
-- Pollution Equipment News
-- CEE Buyer's Guide
-21-

-------
e Chemical Costs
-- Chemical Marketing Reporter
c Cost/Index Information
--	Engineering News Record and Construction Cost Index
--	Engineering News Record and Building Cost Index
--	Marshall & Swift Equipment Cost Index
--	Chemical Engineering Plant Cost Index
--	GNP Implicit Price Deflator, U.S. Bureau of Labor Statistics
--	Consumer Price Index, U.S. Bureau of Labor Statistics
--	5 MGD Treatment Plant Cost Index, for a trickling filter, U.S. EPA
--	50 MGD Treatment Plant Cost Index, for activated sludge, U.S. EPA
--	Complete Urban Sewer System Index
The format of the reference sections should be revised by annotating each ref-
erence and listing the annotated references by topic area. This could probably
best be done by preparing an annotated bibliography. It would include the
author and title of a work on a given subject area and a short abstract summa-
rizing the contents of the work. The methodology for preparing an annotated
bibliography is described in the appendix.
Because such a wide variety of information exists for water pollution control
equipment, a literature search and annotated bibliography are useful tools for
collecting and summarizing water pollution control equipment cost and design in-
formation. An annotated bibliography would be useful to permit writers because
it would provide them with a concise listing of references that may be used to
quickly answer questions that may arise during the permit development process.
In addition, the cost estimate for major equipment items should be verified
with vendors and/or the literature. An annotated bibliography would facilitate
using the literature to verify cost-estimates.
The annotated bibliography should be structured so that the references are listed
by topic area to enable the user to find appropriate references more easily.
Examples of annotated bibliography topics that may be of interest to permit
writers are:
e wastewater treatment design and operating parameters;
o capital investment estimation;
e annual expense estimation;
« methods for determining the economic feasibility of systems; and
-22-

-------
c retrofit problems and how they impact cost estimates.
Following is an example of a recommended format for the reference sections.
CAPITAL INVESTMENT ESTIMATION
(1)	Culp, R.L., Wesner, G.M., and Culp, G.L. Handbook of
Advanced Wastewater Treatment. 2nd edition. Van
Nostrand Reinhold Company, New York, New York. 1978.
632 pp.
Fully describes the practical engineering design and
operation of advanced wastewater treatment plants. The
chapter on estimating costs presents cost curves for
both conventional and advanced processes; shows how to
use cost indices to adjust these costs to any time
frame-, and provides the means for adjusting labor,
material and other costs which are subject to local
variations.'
(2)	Guthrie, K.M. Process Plant Estimating, Evaluation and
Control. Craftsman Book Company of America, Sol ana Beach,
California. 1974. 606 pp.
A complex but comprehensive technique is presented to
quickly assemble capital cost estimates, prepare economic
feasibility studies, establish resource control during
construction, and maintain economic stability over the
productive life of capital projects. Summarizes labor,
material and equipment costs for every type of refinery
and fluid-phase chemical process plant. Some of these
process modules are suitable for pollution control.
-23-

-------
REVIEW OF ECONOMIC AC HI EVA BILITY PROTOCOL
Introduction
Not only will permit writers need an engineering cost manual to estimate the
costs of BEJ technology, but they will also need a manual that will aid them
in determining the economic achievability of the costs of proposed BEJ tech-
nology. Although such a manual has not yet been written, a methodology that
estimates the economic achievability of BEJ technology has been proposed.
This methodology is entitled Protocol for Determining Economic Achievability
for NPDES Permits. The results of PRA's evaluation of the Protocol are pre-
sented in this section, which contains four parts. The first is a brief
summary of the Protocol. The second and third sections review and discuss
two major sets of economic achievability tests: firm level tests and plant
level tests. The fourth section discusses problems with the Protocol.
Several issues could be raised about the scope of the review of the Protocol.
These include:
•	whether EPA has an explicit policy defining "economic achievability";
« whether EPA has decided to have permit writers implement this policy
in BEJ permits; and
•	whether EPA intends to provide permit writers with a step-by-step
manual to use in implementing this policy.
PRA's review has not addressed these issues. We have concentrated our efforts
on the changes in the Protocol that would be required to make it useful as a
manual for permit writers to use in implementing EPA's policy on economic
achi evabi1i ty.
Summary of the Economic Achievability Protocol
The objective of the Protocol is "to assist permit writers in determining the
effect of installing pollution control technologies on the financial condition of
the firms and plants." The approach that was chosen to meet this objective had
to have three characteristics. The first characteristic was that the approach had
to define economic achievability. The Protocol defines economic achievability at
-24-

-------
the firm level and the plant level as the ability to "afford to purchase
and operate treatment equipment." The second characteristic was that the
approach had to recognize the limitations on availability of firm-level and
plant-level financial data. The third characteristic was to "recognize the
limited resources available to each writer." These three characteristics were
molded into the Protocol by establishing two different sets of tests: firm
level tests and plant level tests.
The firm level tests consist of seven tests. Five of them are related to
financial statement analysis. Financial statement analysis uses data from
balance sheets and income statements to calculate financial ratios. The finan-
cial ratios are used to analyze the ability to raise the necessary capital to
buy and install BEJ technology. These ratios are: the current ratio, the
quick ratio, the fixed charge coverage ratio, Beaver's ratio, and the debt-
equity ratio. The remaining two firm level tests are related to stock market
value analysis. They are used to measure the effect of pollution control costs
on stock price and to examine trends in the market value of the stock.
The plant level tests, which according to the Protocol are intended for use only
when a firm contests the results of the firm level analysis, emphasize the re-
lationship of a plant's earnings before taxes to its annual cost.of BEJ technology.
Three plant level tests are suggested: the earnings test, the gross margin test,
and the revenue test.
The Protocol was prepared for an EPA workshop which was held in August 1982. The
document consists of copies of slides that were exhibited at the workshop". As such,
it is not in a format that lends itself to fully informed review and criticism. PRA
did not have the benefit of the discussions and explanations by the workshop modera-
tor and participants in our review of the Protocol , but we tried to be sensitive to
this problem.
The Protocol was not written to be used directly by a permit writer. It is obvious
that a substantial writing effort will be needed to transform the Protocol into a
manual which is useable by a permit writer. In .the evaluation, PRA addressed the
methodology and content of the Protocol rather than its format and presentation.
-?5-

-------
Introduction
Firm Level Analysis
The firm level analysis is designed to provide a rigorous methodology for
estimating economic achievabi1ity using publicly available data. The pro-
cedure combines financial statement analysis and market value analysis to
evaluate the ability of a company to afford a proposed investment in'BEJ
technology.
The financial statement analysis evaluates the historical performance of a
firm by calculating three types of financial ratios:
e liquidity ratios (current ratio and quick ratio), which measure
the ability of a firm to meet its short-term financial obligations;
e solvency ratios (fixed charge coverage ratio and Beaver's ratio)
which measure the ability of a firm to meet long-term financial
obligations; and
e leverage ratio (debt-equity), which indicates the extent to which
a firm's financial resources have been provided through borrowing.
The Protocol states that the five firm level tests require three types of
data:
•	the four-digit SIC code;
0 financial statements for a firm from Moody's Industrial Manual; and
•	industry averages for selected ratios from Annual Statement Studies
by Robprt Morris Associates.
The market value analysis attempts to predict the future financial performance
of a firm by evaluating the effect of investment in BEJ technology on the market
value of its common stock. The two tests in the market value analysis evaluate
the ratio of market value to book value. The tests measure:
e market to book ratios for the three preceding years (without measuring
the effect of pollution control investment); and
e market to book ratio for the most recent year, adjusted to reflect the
effects of the investment in BEJ technology.
The Protocol does not specify the sources for data on market and book values.
-26-

-------
The rest of this section presents the individual tests and problems with their
use in the Protocol.
Current Ratio
The current ratio is a commonly-used indicator of the ability of a firm to meet
its short-term obligations. The ratio, which compares current assets to current
liabilities, indicates the extent to which short-term obligations are covered by
cash and near-cash assets. The Protocol defines the current ratio as:
PR -
CR CL
where: CR = current ratio;
CA = current assets;
CL = current liabilities.
The Protocol does not clearly define the components to be included (or excluded)
in current assets or current liabilities. Other references indicate that current
assets include cash, marketable securities, accounts receivable, and inventories.
Current liabilities include accounts payable, short-term notes payable, the current
portion of long-term debt, the current portion of lease obligations, and accrued
taxes.
The Protocol also does not explain where to find the values to be used in the
test calculations. The values for current assets and current liabilities are taken
from the Comparative Consolidated Balance Sheet in Moody's Industrial Manual.
The current ratio test consists of two parts. First, the current ratio is calculated
without the cost of the BEJ technology. Then the ratio is recalculated with the
capital cost of the BEJ technology subtracted from current assets. The second part
of the test could be expressed as:
CA-CI
CR =
CL
where: CR = current ratio;
CA = current assets;
CI = capital cost of BEJ technology;
CL = current liabilities.
-27-

-------
For all of the ratio tests, the capital cost is adjusted to reflect the in-
vestment tax credit, which is assumed to be 15%. The investment tax credit
should be entered by the permit writer to assure that the calculations reflect
current tax law. Data sources and instructions should be provided to assure
that the correct value is used.
The intended use of the current ratio without the cost of pollution controls is
unclear. The example in the Protocol presents this ratio for each of the three
previous years. This information is clearly useful for indicating the trend in
liquidity position over time. However, no guidance is offered on how to incor-
porate this information into the economic achievabi1ity analysis and decision-
making process.
The Protocol presents two "critical values" for evaluating the current ratio with
the cost of pollution controls. The first is a traditional rule-of-thumb for credit
analysis which suggests a minimum current ratio of 2.0. This ratio would assure
that the firm could cover its obligations even if the value of current assets,
particularly inventories, was substantially reduced-in the event of a forced
liquidation. The second is the "average" current ratio for the industry. In actual
practice current ratios differ from industry to industry. Upper quartile, median,
and lower quartile values for current ratios in each four-digit SIC category are
available in Annual Statement Studies by Robert Morris Associates, Inc.
The Protocol presents both the traditional "2.0" rule of thumb and the industry
average as "critical values" for the current ratio. Neither value is explicitly
designated as the pass/fail point for this test.
The definition of these critical values produces two potential shortcomings
in the use of the current ratio in the Protocol. First, the test result may
be ambiguous if the current ratio for a firm falls between the industry average and
2.0. Second, if the industry average (actually the median) is used as the pass/fail1
point for this test, half of the firms in the SIC category would be expected to fail1
the test.
These problems could easily be corrected by specifying a single critical value to
indicate if a firm passes the current ratio test. EPA may also wish to consider a
different industry reference point such as the lower quartile to indicate passing
or failing this test.
-28-

-------
Although the presentation in the Protocol causes the problems described above,
the current ratio appears to be a useful indicator of the ability of a firm to
meet its short-term obligations.
Quick Ratio
The quick ratio, which is closely related to the current ratio, is also a common
indicator of ability to meet short-term obligations. The quick ratio compares
quick assets (current assets minus inventories) with current liabilities. The
Protocol defines the quick ratio as:
QR = ^r-
where: QR = quick ratio;
CA = current assets;
I = inventories;
CL = current liabilities.
The numerator includes all current assets except inventories. These current
assets are owned by the firm or legally obligated to it. The firm can reasonably
expect to convert these assets to cash at their book value. Inventories, on the
other hand, might have to be sold below book value in case of liquidation. Current
liabilities are defined exactly as they are for the current ratio.
The Protocol does not explain how to obtain the values to be used in the test
calculations. The values for current assets, inventories, and current liabilities
are all available from the Comparative Consolidated Balance Sheet in Moody's
Industrial Manual.
The quick ratio test includes two parts. First, the quick ratio is calculated with-
out the cost of the BEJ technology. Then the quick ratio is recalculated- with the
capital cost of BEJ technology subtracted from quick assets. The second test can be
expressed as:
CA-I-CI
QR =
CL
where: QR = quick ratio;
CA = current assets;
I = inventories;
CI = capital cost of BEJ technology;
CL = current liabilities.
-29-

-------
As with the current ratio, the intended use of the quick ratio without the cost
of pollution control is unclear. The Protocol presents this ratio for the three
previous years in the example calculations, but does not explain how to incorporate
this useful information into the evaluation of economic achievabi1ity.
The Protocol presents two "critical values" for the quick ratio with the cost of
pollution controls. The first is a rule-of-thumb for credit analysis which suggests
a minimum quick ratio, of 1.0. This ratio would assure that the firm could cover all
of its current obligations with cash and near-assets. That is, all current obliga-
tions could be met without liquidating inventories.
The second is the "average" quick ratio for the industry. The Protocol indicates
incorrectly that industry averages are available from Robert Morris Associates.
However, quick ratio is not included in Morris and cannot be calculated from the
information given. (Morris does include industry averages for total current assets,
inventories and current assets. However, a correct industry average would be an
average of quick ratios for individual firms, not a single calculation using industry
averages for each component of the test.) The effect of this shortcoming is to pro-
vide only a single critical value, the rule-of-thumb value of 1.0.
Despite the problems cited above, the quick ratio is a useful indicator of the
ability of a firm to meet short-term obligations. The quick ratio complements the
current ratio by indicating liquidity problems which could be masked by large but
hard-to-1iquidate inventories.
Fixed Charge Coverage Ratio
The fixed charge coverage ratio measures the ability of a firm to cover fixed finan-
cial obligations from operating earnings. The ratio indicates the extent to which
earnings can decline without causing the firm to have trouble meeting interest and
other fixed charge obligations. The Protocol defines this ratio as:
FCCR =
where: FCCR = fixed charge coverage ratio;
CEBFC = cash earnings before fixed charges;
FC = fixed charges.

-------
The Protocol does not clearly explain the components of cash earnings before
fixed charges or fixed charges. However, they can be inferred from the example
given in the text. Cash earnings before fixed charges can be expressed as:
CEBFC = NPBT+FC+D
CEBFC = cash earnings before fixed charges;.
NPBT = net profit before taxes;
FC-= fixed charges;
D = depreciation.
can be expressed as:
FC = IE+OFP+CPLTD
FC =	fixed charges;
IE =	interest expense;
OFP =	other fixed payments;
CPLTD =	current portion of long-term debt.
The information required to compute the value of these components can be calculated
with some difficulty from the Comparative Consolidated Income Statement in Moody's
Industrial Manual. However, this calculation would require a clear explanation be-
cause both numerator and denominator require calculations involving several lines
in the Moody's Income Statement. (For example, calculation of the numerator would
involve ten different lines in Moody's.)
The example in the Protocol simplifies the calculation by referring to a sample
income statement presumably provided by the firm being analyzed. Use of this state-
ment does not agree with the statement of data needs at the beginning of the firm
level analysis section of the Protocol.
The test calculates the fixed charge .coverage ratio with the capital cost of BEJ
technology subtracted from the numerator. That is:
FCCR = "BRpCI
where: FCCR = fixed charge coverage ratio;
CEBFC = cash earnings before fixed charges;
CI = capital cost of BEJ technology;
FC = fixed charges.
where:
Fixed charges
where:
-31-

-------
The Protocol specifies that the capital cost should be calculated assuming that
it will be financed with proportions of debt equal to the current debt ratio for
the whole firm. This assumption requires the analyst to calculate the debt ratio
for the firm, the portion of the BEJ technology cost financed with debt> and the
interest and principal payments. Step-by-step instructions would probably be re-
quired to assure that this test could be completed without error.
In addition, the calculations require two items of data which the Protocol does
not explain how to obtain. First, interest charged on new debt is not available
from any of the data sources and would probably have to be supplied by the permit
writer or the firm. A specific source or basis should be designated to avoid
ambiguity. Second, annual operating and maintenance (O&M) expenditures would
probably be supplied by the permit writer using cost engineering data.
The completed fixed charge coverage ratio calculation is compared with the critical
values listed in the Protocol. A ratio greater than 2.0 indicates solvency, a ratio
less than 1.5 indicates insolvency, and a ratio between 1.5 and 2.0 indicates
questionable solvency. The Protocol does not explain the basis for these designations
The fixed charge coverage ratio provides useful information on the impact of capital
investment for BEJ technology on the ability of a firm to meet its fixed charge obli-
gations. However, the test procedure is poorly explained and complicated to perform.
This procedure should be clarified and simplified or an alternative test should be
provi ded.
Beaver's Ratio
Beaver's ratio measures the.extent to which a firm can cover its total liabilities
with current earnings. The Protocol states that Beaver's ratio has been identified
as the "single best predictor of bankruptcy." The ratio is presented as:
BR - IGCF
CL+LTD
where: BR = Beaver's ratio;
I6CF = internally generated cash flow;
CL = current liabilities;
LTD = long-term debt.
-32-

-------
The Protocol does not clearly explain what is included in the components of
the ratio. The reader can infer from Worksheet 4a that the "internally generated
cash flow" equals net income after taxes plus depreciation. Current liabilities
are defined as in the current ratio.
The definition of long-term debt in the Protocol is unclear. The example in
Worksheet 4a uses long-term liabilities in the calculation of Beaver's ratio.
The value for long-term liabilities, which is taken from the sample balance sheet,
includes long-term debt, deferred income taxes, minority interest, and other accrued
liabilities. (The difference between long-term liabilities and long-term debt in
the example calculation exceeds $19,000,000.) The inconsistency between the defini-
tion of Beaver's ratio and the calculation must be corrected to eliminate potential
confusion.
The data required for calculating the ratio are available from the Comparative
Consolidated Balance Sheet in Moody's Industrial Manual and from the sample balance
sheet. The sample balance sheet follows the examples from Moody's but is not re-
ferenced in the Protocol.
Values for net income after taxes, depreciation, and current liabilities are taken
directly from Moody's. The value for long-term liabilities in the example calcula-
tion is taken from the sample balance sheet. (The value for long-term liabilities
can also be calculated by adding four lines in the Moody's balance sheet.)
The Beaver's ratio test calculates the ratio with the numerator and denominator ad- •
justed to incorporate the effects of the proposed investment in BEJ technology.
The test assumes that the investment would be financed with proportions of debt
equal to the debt ratio of the firm. This assumption requires the user to calculate
the debt ratio for the firm, the portion of the BEJ technology investment financed
with debt, and interest and principal payments.
The test calculation requires data for interest charged on new debt and annual O&M
expenditures as described in the section on the fixed charge coverage ratio. In
addition, the calculation includes values for marginal income tax rate and deprecia-
tion rates. These are assumed to be constants in the Protocol, but probably should
be entered by the user to assure that they reflect current tax law.
-33-

-------
The completed Beaver's ratio calculation is compared with critical values
provided in the Protocol. A ratio greater than 0.2 indicates solvency, a ratio
less than 0.15 indicates insolvency and a ratio between 0.15 and 0.2 indicates
questionable solvency. The Protocol does not explain the basis for these
critical values.
The Protocol states that a study has identified the ratio as the single best
predictor of bankruptcy. The study in question was William H. Beaver's "Financial
Ratios as Predictors of Failure" published in 1967.^ ^ Beaver's work emphasized
empirical analysis of bankrupt and non-bankrupt firms. Beaver's approach, which
relied on single ratios as predictors of bankruptcy, has not been widely accepted.
The literature since Beaver's pioneering work has emphasized techniques for com-
bining the insights from a variety of financial ratios.
Some researchers have questioned the usefulness of the various ratios for predict-
ing business failures. While several approaches have been used to classify firms
as bankrupt or non-bankrupt after the fact, none has been demonstrated as a success
ful predictor of future business failures.
Beaver's ratio appears to have been included in the Protocol primarily for its use-
fulness in predicting bankruptcy. Because of the problems described above, however
the use of Beaver's ratio should be reconsidered. If predicting bankruptcy is an
important part of the economic achievabi1ity evaluation, a substantially different
approach may be required.
Debt-Equity Ratio
The debt-equity ratio measures the extent to which the capital resources of a firm
are financed through debt. The ratio which compares debt to stockholders' equity i
presented as:
where: DER = debt-equity ratio;
LTL = long-term liabilities;
TSE = total stockholders' equity.

-------
The Protocol does not explain what is included in the long-term liabilities
or total stockholders'equity. Long-term liabilities are defined, as in Beaver's
ratio, to include long-term debt, deferred income taxes, minority interest, and
other accrued liabilities. Total stockholders' equity includes common stock at
par, additional paid-in capital, preferred stock, and retained earnings.
The example calculation presented in Worksheet 5 uses data taken from the sample
balance sheet for both components of the ratio. Both items could be taken from
the Comparative Consolidated Balance Sheet in Moody's Industrial Manual. Long-
term liabilities would have to be calculated by adding four separate lines in
Moody's. Total stockholders' equity could be taken directly from Moody's. The
net effect would be a calculation which would be no more complicated than the
example calculation.
The debt-equity ratio test calculates the ratio without including the effect of
the proposed investment in BEJ technology. This approach is based on the assump-
tion that the investment would be financed using a proportion of debt equal to
the debt ratio of the firm. However, this requires two additional assumptions.
First, all of the debt will be funded as a long-term liability. Second, all of
the non-debt portion of the investment must be from some form of additional capital
contribution from stockholders (such as sale of additional stock). If any of the
investment is financed from existing assets the stockholders' equity (the denominator)
will not increase proportionately.
This approach may not be consistent with the other tests, because it makes the non-
conservative assumption that the firm will choose to finance the non-debt portion
of the capital investment through sale of stock. The opposite assumption, that'the
non-debt portion of the capital .investment would be financed out of current assets,
would be more consistent. This could be accomplished by revising the test to incor-
porate the effect of the proposed investment on the ratio, as the Protocol does for
all of the previous tests.
The Protocol states that no general target exists for this ratio but that industry
averages and historic ratios are important comparative indicators. The example cal-
culation in Worksheet 5 presents values for upper quartile, median, and lower quartile
in the industry. This approach causes two problems. First, the lack of target ratios
assures that for at least some firms the results of this test would be indeterminate.

-------
Second, the ratio used in Robert Morris Associates' Annual Statement Studies
is not the same as the ratio used in the Protocol. The nearest ratio in
Morris is the debt-worth ratio which is defined as:
D ;tl
"W TNW
where: D = debt;
W = worth;
TL = total liabilities;
TNW = tangible net worth.
This ratio is different from the debt-equity ratio in two respects. First, the
numerator includes current liabilities. Second, Morris does not clearly define
tangible net worth. Thus the denominator may differ from the denominator used
in the Protocol. With these problems, this test offers no clear guidelines for
interpreting the results.
The debt-equity ratio is a useful indicator of the extent to which the capital
of the company has been obtained through borrowing. This ratio may provide in-
sight concerning the ability of a firm to raise capital through additional borrow-
ing. However, the test presented in the Protocol is not defined with sufficient
clarity to be used in evaluating the economic achievabi1ity of potential invest-
ments in BEJ technology.
Market Value Analysis
Market value analysis is intended to provide an estimate of the future financial
performance of a firm. This analysis examines trends in market value of the
firm's stock and measures the effect of pollution control costs on stock price..
Several major problems limit the usefulness of this test.
The Protocol assumes that stock market value equals the net present value of the
expected future cash flows of a firm. This assumption may agree with theory but
be infeasible in use. It should be reconsidered to assure its appropriateness and
usefulness in this context.
The procedure requires the user to calculate the net present value of the proposed
investment in BEJ technology. However, the procedure is unclear and the sources
for several types of data have not been provided.
-36-

-------
The example calculations for net present value in Worksheet 6 include several
items which are not explained. The rate of growth in operating cost is based
on cost engineering assumptions which are not stated. The user must know these
assumptions to generate the appropriate value. The definitions and values for
company beta, risk-free rate, and credits for product recovery are not explained.
The user must be provided with a source for these data to perform the analysis.
The net present value of the investment in BEJ technology is used to calculate
an adjusted stock price. The procedure calculates the effect of the investment
on stock price if the cost is spread over all outstanding shares. Data are re-
quired for the high and low values of the stock and the number of outstanding
shares. The Protocol needs to be revised to indicate the source of these data
and the specific dates from which they should be taken. Without explicit direc-
tions, many valid results are possible because of the daily fluctuations in
market price.
The market value analysis involves two tests. The first compares the market value
to book value of the stock without adjustment for the investment in BEJ technology.
As with the procedure for adjusting stock prices, sources and specifications for
market and book values have not been provided. In addition, no guidelines are
presented for interpreting the results. This test is intended to show the historic
trend of market to book ratio.
The second test calculates the market to book value after the market value has been
adjusted to incorporate the effects of the investment in BEJ technology. No speci-
fic guidelines for interpreting the results of this test have been provided. The
Protocol does state that a firm will not go bankrupt as long as the market value
remains above zero after the effects of the pollution control investment have been
incorporated. No justification is provided for this statement. In addition, it
offers little insight for interpreting the ratios for non-bankrupt firms.
The market value analysis is designed to increase the predictive power of the
economic achievability methodology. However, the problems described above would
prevent the user from developing exact and unambiguous results. In addition, the
variability of market prices suggests that the predictive capabilities offered by
this analysis are questionable. This test should probably be deleted from the
Protocol and one or more alternative tests should be substituted for it.
-37-

-------
Plant Level Analysis
Introduction
According to the Protocol, if a firm contests the firm level analysis, plant
level analysis is performed. The plant level analysis consists of three
tests: the earnings before taxes test, the gross margin test, and the revenues
test. All of the tests attempt to estimate a plant's earnings before taxes and
compare them to the annual cost of BEJ technology. If the estimated earnings
before taxes exceed the annual cost of BEJ technology, the technology is con-
sidered to be economically achievable.
Given this definition of economic achievability at the plant level, there are
two implicit assumptions that must be true if the plant level analysis is to
generate realistic results. The first is that a plant will not be able to pass
forward any of its increased pollution control costs. The validity of this
assumption will vary from industry to industry and will partly depend on the cost
of the BEJ pollution control technology that the competitors of the plant will be
required to install. In a very competitive industry, a plant will presumably be
able to pass forward few of its increased costs. The opposite will be true of a
plant in a less competitive industry, unless the competitors are issued permits
that require less costly BEJ technology.
The second assumption is that a plant would be willing to expend a significant
portion, if not all, of its earnings before taxes on BEJ pollution control, per-
haps reducing the profitabi1ity of the plant to near zero. This is an unlikely
option to be exercised.
Although the 'above two assumptions tend to have a balancing effect on each other,
their net effect is unknown. It is recommended that EPA reconsider the plant
level analysis because of the possibly biased results to which these assumptions
could lead.
The following sections of the report describe each of the three tests, analyze
some of the shortcomings of the tests, and discuss the data requirements for per-
forming the plant level tests.
-38-

-------
Earnings Before Taxes Test
The earnings before taxes (EBT) test compares the annual cost of additional
pollution control expenditures due to the BEJ permit with a plant's earnings
before taxes. If EBT exceed th.e annual costs of pollution expenditures the
technology is deemed economically achievable; if EBT are less, the technology
is not economically achievable. If EBT are equal to additional BEJ annual
costs, no concrete decision can be made.
Earnings before taxes are defined by the Protocol -as the following:
"EBT - PR - COGS - CO
where: EBT	=	earnings before-taxes;
PR	=	plant revenues;
COGS	=	cost of goods sold;
CO	=	corporate overhead.
Additional BEJ annual costs are defined as the following:
AC - ACC + OM
where: AC = annual costs of BEJ technology;
ACC = annualized capital cost = capital cost x capital
recovery factor;
OM = annual operating and maintenance costs.
The EBT test thus is defined as the following:
?
(PR - COGS - CO) - (ACC + OM) < = > 0
In practice, the test is limited by the problem of the allocation of overhead
to a specific plant. As indicated in the Protocol, corporate overhead is not
usually allocated to individual plants and biases in corporate overhead would be
difficult to detect. Although this is usually a function of size, this may not
always be the case. For a one-plant company, the distinction between plant over-
head, which* is included in cost of goods sold, and corporate overhead may not be
easy to delineate.
-39-

-------
Gross Margin Test
Gross margin (GM) is the difference between revenues and cost of goods sold.
The term is applicable at the plant level as well as the firm level. This
test compares the annual cost of BEJ technology as a percentage of gross margin
to the industry's earnings before taxes expressed as a percentage of the industry's
gross margin. (Industry is defined as the four-digit SIC code of a plant.)
The test looks like the following:
AC f/EBT\
GMP \m)l
annual costs of BEJ technology;
gross margin of the plant;
the ratio of earnings before taxes
to gross margin for the industry.
When'AC/GM is less than (EBT/GM)j, the BEJ technology is considered to be
economically achievable. If not, the test is inconclusive and plant closure
analysis is necessary. The source of (EBT/GM)j is Annual Statement Studies pub-
lished by Robert Morris Associates, Inc.
It is unclear what additional information is given by the GM test when the EBT for
a plant are known. This test gives some indication of financial conditions of a
plant vis-a-vis that of a typical plant in the industry. It does not, however, lead
the permit writer to any additional information concerning the economic achievabi1ity
of installing BEJ technology.
If the Protocol were revised such that this test was performed only when the EBT-
test could not be performed, it could serve as an adequate substitute for estimating
EBT. Multiplying the plant's GM by the industry's EBT/GM gives a good estimate if
one assumes that the EBT/GM of the plant is equal or nearly equal to the EBT/GM of
the industry. Thus, if the above assumption is true the following is true:
(1)	Plant EBT = Industry (EBT/GM) X Plant GM
Then:
(2)	Plant GMT = Industry (EBT/GM)
Note that equation (2) is very similar to the GM test inequality. The major difference
is that the EBT of the plant occurs on the left side of the equation instead of the
annual cost of BEJ technology for the plant.
where: AC =
GM =
P
'EBT>
,GM ,
-AO-

-------
When the test is viewed in this manner, it is identical to the EBT test,
except that the annual cost of BEJ technology is compared to an estimate
of the plant's EBT instead of actual EBT; however, the test relies on the
assumption that the EBT/GM for the plant is equal to the EBT/GM for the
industry.
A major disadvantage of this test is that it assumes that the balance sheet
of a single plant is typical of the industry. It can be shown that a plant
which has higher EBT than typical plants of the same size can have a lower
EBT/GM than the industry EBT/GM. Therefore, if economic achievabi1ity is de-
fined as the plant being able to cover the annual cost of BEJ technology from
its EBT, this test may lead to the wrong conclusion. The true utility of EBT/GM
is that it is an indication of how much corporate overhead is "eating into" the
plant's EBT. This is shown by the following:
EBT _ R - COGS - CO _ 1 _ ' CO
GM R - COGS	R - COGS
where: R = plant revenues;
COGS = cost of goods sold;
CO = corporate overhead;
EBT = earnings before taxes;
GM = gross margin.
Another disadvantage is that the plant may be unwilling or unable to provide GM
data, and without GM data this test cannot be performed.
Revenue Test
The revenue test compares the annual cost to a plant for BEJ technology (expressed
as a percentage of revenues) to the industry average of earnings before taxes
(expressed as an average of revenues). The comparison is the following:
?
AC < = > f EBT
Rn	V R /J
where: AC = annual cost of BEJ technology;
Rp = plant revenues;
mX) = the industry average of earnings before taxes
V /I divided by revenues.
-41-

-------
As with the GM test, this test could be used to estimate plant revenues when
actual plant revenues are unknown. However, the usefulness of this test when
actual plant revenues are known is unclear.
It can be shown that a plant that has a higher EBT than typical plants of the
same size can also have a lower EBT/R than the industry EBT/R. Thus if plant
economic achievabi1ity is defined as the plant being able to cover the annual
cost of BEJ technology from its EBT, this test may lead to the wrong conclusion.
The actual utility of this test is that it is a measure of the costs-to-revenues
ratio for a plant. This is shown by the following:
EBT _ R - COGS - CO
R	R
where: R - plant revenues;
COGS = cost of goods sold;
CO = corporate overhead;
EBT = earnings before taxes.
Data Requirements and Sources
Table 7 illustrates the data requirements for each of the three tests. A major
advantage of the plant level analysis is that the data requirements are minimal.
The source of plant financial data is the company. These data may be the most
difficult to obtain. EPA should consider whether there is any legal obligation
on the part of the company to provide these data. The definition of what is in-
cluded in the financial data will probably vary from firm to firm. The Protocol
should include precise definitions of these terms and how to manipulate the data
if the firm keeps its books on a different .basis. These definitions should be
consistent with those used in Annual Statement -Studies.
The capital and annual costs of the treatment technology are calculated by the
permit writer using the methodology given in the engineering cost analysis manual.
The Protocol is vague about the calculation of the interest rate for the capital
recovery factor; more explanation of the derivation of the interest rate is needed.
Industry ratios for the GM and revenue tests can be calculated from Annual Statement
Studies. These averages are not directly given, and the Protocol should include an
explanation of how to calculate them.
-42-

-------
Jable 7. .Data Needs for the Plant Level Analysis
Data
Tests
Source of Data
EBT
GM
Rev,
Financial Data
Plant Revenue	X X
Cost of Goods Sold X X
Corporate Overhead X
BEJ Technology Costs
Capital Cost	X X
Capital Recovery
Factor
-	Life of Equipment X X
-	Interest Rate X X
Annual 0 & M	X X
X
X
Company
Company
Company
Permit Writer/Cost Manual
Permit Writer/Cost Manual
?
Permit Writer/Cost Manual
Industry Ratios
EBT/GM
EBT/R
Plant SIC Code
X
X
Annual Statement Studies
Annual Statement Studies
Plant or Dun & Bradstreet
-43-

-------
The plant SIC code should be available from the plant. If it is not, Dun &
Bradstreet lists SIC codes for many plants.
Analysis of the Economic Achievabi1ity Tests
Introduction
In analyzing the Protocol PRA identified several problem areas where changes
or revisions should be made. These problem areas are discussed below. They
are divided into three different types: problems common to both the firm level
and plant level analyses, problems related to the firm level analysis only, and
problems related to the plant level analysis only.
Common Problems
Policy on Economic Achievabi1ity. The policy of EPA on "economic achievabi1ity"
is not clearly defined in the Protocol. A proposed BEJ technology is considered
to be economically achievable if a firm or plant "can afford to purchase and
operate" the required treatment equipment. This phrase could encompass a wide
range of interpretations ranging from limits on reduced profitability to limits
on plant closure.
The exact intent of the above definition cannot be inferred from the Protocol.
Although the methodology includes measures of ability to raise capital, potential
for bankruptcy, and other considerations, no consistent policy is evident.
Time Frame. Several procedural problems limit the usefulness of the Pro toco 1.
First, most of the tests provide a one-year view of a plant or firm. A few tests re-
quire data on three-year trends. However, the methodology does 'not provide clear
guidelines f~r interpreting these tests. As a result, the pass/fail tests all span a
single-year increment. Because of this emphasis on short-term time frames, the
methodology provides little insight into long-term historical performance of a
plant or firm.
The market value analysis provides very little insight into future performance of
a firm. The Protocol provides no indication of the effect that business cycles
may have on the performance of a firm. As a result, permit writers will be forced
to write permits to cover a long time period based on a very short-term perspective.

-------
Although not explicitly stated, it is implied that the three plant level
tests should be performed using financial data for the most recent fiscal
year of a plant. Because swings in the economy directly affect the financial
performance of a plant, it would be better to perform these tests for the
latest three years, or some type of averages for the latest three years. We
believe this would produce a clearer picture of the ability of a plant to
afford the BEJ technology.
SIC Code Definition. The GM test and the revenue test in the plant level
analysis and the current, quick, and debt-equity ratios in the firm level
analysis all require the permit writer to compare financial data to industry
averages for firms or plants in the same SIC code. No direction is given when
a plant or firm belongs to more than one SIC. This problem could be solved by
using the primary SIC of a plant as a basis for determining the SIC to which it
should be compared. However, it is likely that for large integrated shops, the
production activity of the primary SIC may not be the major wastewater-producing
activity. For example, an electroplating captive operation may be the major
contributor of toxic pollutants, yet the primary SIC code for the plant may be
that for automobile production. The permit writer must then decide whether the
proper SIC code (for purposes of the plant tests) is SIC 3471 (Electroplating) or
3711 (Motor Vehicles and Passenger Car Bodies).
This problem is potentially more serious at- the firm level. A large diversified
firm may have plants in many different SIC codes. The financial situation of a
diversified firm may be determined primarily by activities unrelated to the SIC
code of the individual plant being evaluated for a permit. As a result, comparison
of financial data for the firm with industry averages for the SIC group of a plant
may produce misleading results.
Interest Rate and Investment Tax Credit. In some instances, the Protocol does not
provide the data which the permit writers will need to perform the tests. Two
examples are the interest rate and investment tax credit.
Some of the firm-level tests use investment tax credits in the calculation of the
ratios, but no data source' is provided for the permit writer. The amount of the
investment tax credit that can be taken in any one year is a function of the current
Federal tax law. To minimize use of a permit writer's time, the Protocol should
-45-

-------
explain how to calculate the investment tax credit. Basic information
that is needed is the type of investment (building, land, rolling stock,
etc.) and the amount of tax credit that Federal law allows for each type
of investment.
To calculate the total annual costs of a BEJ technology, the capital costs,
annual O&M costs, and the appropriate interest rate must be known. Capital
costs and -annual O&M costs can be calculated directly by using the engineering
cost manual; however, the interest rate will vary among both firms and indus-
tries. The Protocol provides no guidance as to the appropriate interest rate or
the specific total annual cost calculation when the interest rate, capital costs,
and annual O&M costs are known.
Presentation. The Protocol was intended as a workshop guidebook. Accordingly,
it includes only a brief explanation and an example for each test. The level of
detail presented is not sufficient to permit effective use of the-Protocol.
The procedures for gathering the appropriate data from Moody's and Morris are not
clearly explained. In a few instances, the examples are actually calculated using
data from an alternate source.
Several of the test procedures include errors which would prevent a permit writer
from -calculating the results correctly. These problems are described in detail
for each test in the preceding sections.
Several of the critical values for the firm level analysis do not indicate clearly
whether a proposed BEJ technology would pass the test. Critical values are not
provided for some of the tests. As a result, these tests can be expected to be
inconclusive in many instances. In addition, the plant level analysis does not
contain guidance for those situations where the tests give conflicting results.
The example in the Protocol that exhibits the use of the tests in tandem show that
the technology is economically achievable under all three tests. The plant level
analysis should be expanded to address conflicting results.
The firm level analysis does include a table indicating how to interpret the tests
in case they present conflicting results. However, the table should be explained
and more detail should be provided.
-46-

-------
Firm Level Problems
Firm Size. Several of the tests may actually measure firm size, rather than
the impact of the BEJ investment or annual costs on the firm. The BEJ invest-
ment or annual costs are likely to be a larger percentage of a small firm's
assets, earnings, or cash flow than those of a larger firm. As a result, a small
firm is more likely to fail the tests.
For example, the impact of BEJ costs for one plant of a 50-plant company may be
negligible when compared to the company's financial base. For a single-plant
firm the impact, as measured by the Protocol, is much larger because the relative
amount of BEJ investment or annual costs is a much larger portion of the company's
financial base. If one assumes that the results of the financial tests should be
the same for plants with identical production technology and financial costs, then
the size of the firm should be irrelevant. The tests recommended in the Protocol
obscure this distinction. One anomalous result is that a financially weak plant
in a multi-plant firm may receive a more stringent BEJ permit than a prosperous
plant in a smaller firm. This appears to be contrary to the Protocol's intent.
The use of financial ratios may mask the differential effects which pollution
control investments may have on small firms. For example, small firms may face
difficulties in raising capital that are not predicted by the ratio tests.
The data required to conduct these tests for small firms may not be available in
Moody's for two reasons. First, firms are included in Moody's for a fee, and
companies that do not sell common stock in substantial volume or issue bonds are
not likely to be listed. Second, closely held firms will not be listed. Alternate
data sources should be specified for these firms.
Scope of the Methodology. The firm level analysis includes financial statement
analysis as an indicator of historical performance and market value analysis as a
predictor of future performance. However, several relevant issues are not addressed.
First, the methodology compares the financial performance of a firm with that of
other firms in the same industry. Firms may also face substantial competition from
industries which supply substitute products and from firms in other countries which
may not bear the same pollution control costs. These factors affect the ability of
-47-

-------
a firm to pass costs forward. The Protocol does not address these important
economic considerations.
Second, the financial statement analysis emphasizes liquidity, solvency, and
leverage. All of these relate to the ability of a firm to borrow the necessary
capital. The methodology does not evaluate sources of capital, which may be
much more limited for some firms and industries than for others. In addition,
the methodology does not evaluate the effect of the BEJ requirements on the
profitability of a firm. The adverse effect on the profitability of a healthy
firm may be substantial without causing the firm to fail any of the ratio tests.
Third, the firm level analysis primarily measures the effects of capital expen-
ditures on the firm. The effects of operation and maintenance costs, which can
exceed annualized capital costs, are generally ignored. The firm level analysis
should be revised to consider the financial effects of operation and maintenance
costs.
Fourth, the tests all rely on an unlikely assumption. For all of the tests except
the debt-equity ratio> the methodology assumes that the investment for at least
the non-debt portion of the capital investment would be financed from current
assets, current cash flow, or current earnings. This assumption may not be justi-
fied.
Finally, several of the tests involve a comparison with the median value for the
industry. As a result, half of the firms in each industry can be expected to fail
each of these tests. A comparison with the lower quartile value for the industry
may be more appropriate.
Plant Level Problems
Revenues. All of the tests in the plant level analysis rely on data for the revenues
of a plant. In an integrated corporation, a plant may serve as a "cost" center and
may not generate revenues. One example is an ore mine that feeds an ore1 smelter
which also belongs to the company. The mine incurs costs but no revenues. The smelter
generates revenues that can be allocated collectively, but not individually, to both
the smelter and the mine. Another example is a sub-assembly plant that produces
assembled parts for the main assembly plant. The corporation may assign a transfer
price to the ore or assembled parts, but this price may not reflect the market price
(if.one exists) because of tax considerations.
-48-

-------
Corporate Overhead. The Protocol recognizes that allocations of corporate
overhead to specific plants may be difficult to estimate and biases in cor-
porate overhead difficult to detect. (Corporate overhead is a data require-
ment only for the EBT test.) This may be a problem for both large and small
corporations alike. A one-plant company may not be able to distinguish between
plant overhead (which is included in cost of goods sold) and corporate overhead.
Value of GM and Revenue Tests. The Protocol fails to discuss the role of the GM
and revenue tests. It states that the GM and revenue tests are "designed to
provide a measure of economic achievability equivalent to the earnings test."
In light of this statement, it would seem that these two tests are redundant.
As stated earlier, these tests provide no information that is not already pro-
vided by the EBT test. All three tests attempt to relate the annual costs of
BEJ technology to the plant's earnings before taxes', yet the EBT test provides
the best measure because it relies exclusively on plant data, while the other
two tests rely on industry data as a proxy for plant data. Both the GM and
revenue tests should be reconsidered in terms of their validity and utility.
Partial Regulation. An integrated plant may have part of its wastewater genera-
tion already regulated by national BAT guidelines, but the rest of its waste-
water generation may not have BAT guidelines. A current example of such a situa-
tion is an aluminum refinery plant that also has a coil coating operation. No
BAT guidelines have been promulgated for aluminum refining, yet BAT guidelines
are already in place for coil coating on aluminum. Provisions need to be made
in the Protocol to deal with such circumstances.
_A9-

-------
INTERFACING OF THE MANUALS
The Cost Analysis Manual and the Economic Achievabi1ity Protocol cannot be
used in conjunction with each other by a permit writer unless major modifi-
cations are made to the two documents. In addition, the information con-
tained in the two manuals is not set forth in a manner that would facilitate
use of the manuals by a permit writer. A step-by-step presentation of the
methodologies in both manuals would do much to improve their utility^ either
for use separately or in conjunction with one another.
There is some overlap between the information in the Cost Analysis Manual
and the Protocol , in that the Manual addresses some aspects of the economic
feasibility of a technology. If the Protocol is to be used in conjunction
with the Manual, this overlap should be eliminated. This can be done by
deleting the following sections from the Cost Analysis Manual.
e Volume I
-	Cost Estimate -- Net Profit and Cash Flow (pp. 12, 27, 33, and
37); and
-	Feasibility Evaluation (pp. 14-16, 27, 28, 33, 37, and 38).
e Volume II
-	The Cash Flow Concept (Appendix C);
-	Discrete and Continuous Interest Factors (Appendix D);
-	Measures of Merit (Appendix E); and
-	Rates of Return and Interest Rates (Appendix G).
-50-

-------
REFERENCES
(1)	Beaver, William H.,"Financial Ratios as Predictors of Failure",
Empirical Research in Accounting. Selected Studies, 1966 Journal of
Accounting Research, 1967 , pages 71-111..
(2)	Blecker, Herbert G., and Theodore W. Cadman (ICARUS Corporation),
Capital and Operating Costs of Pollution Control Equipment Modules -
Volume I - User Guide (PB 277 804, EPA-R5-73-023a), Volume II - Data
Manual (PB 224 536, EPA-R5-73-0236), U.S. Environmental Protection
Agency, July 1973.
(3)	Burns and Roe, Consulting Engineers, Innovative and Alternative Tech-
nology Assessment Manual (publication MCD-53, EPA 430/9-78-009) U.S.
Environmental Protection Agency, February 1980.
(4)	Cran, John,"Improved Factored Method Gives Better Preliminary Estimates",
Chemical Engineering, volume 88, number 7, April 6, 1981, pages 64-79.
(5)	Culp, Russel L., George Mack Wesner, and Gordon L. Culp, Handbook of
Advanced Wastewater Treatment, Second Edition, Van Nostrand Reinhold
Company, Cincinnati, 1978.
(6)	Dryden, Charles E., and Richard H. Furlow, Chemical Engineering Costs,
1966 Edition, The Ohio State University, Columbus.
(7)	Haines, T.B. "Direct Operating Labor Requirements for Chemical Processes",
Chem. Eng. Progress, 45:556, 1957.
(8)	Hall, Richard S., Jay Matley, and Kenneth J. McNaughton, "Current Costs of
Process Equipment", Chemical Engineering, volume 89, number 7, Aoril 5,
1982, pages 80-116.
(9)	Kinkley, M.W., and R.B. Neveril (GARD, Inc.), Capital and Operating
Costs of Selected Air Pollution Control Systems (PB 258 484, EPA 430/3-
76-014), U. S. Environmental Protection Agency, May 1976. The costs
from this report have been included in series of articles by William M.
Vatavuk and Robert B. Neveril which haVe appeared in Chemical Engineer-
ina since 1980.
(10)	Perry, Robert H., and Cecil H. Chilton, Chemical Engineers Handbook,
Fifth Edition, McGraw - Hill Book Company, St. Louis, 1973.
(11)	Peters, Max S., and Klaus D. Timmerhaus, Plant Design and Economics for
Chemical Engineers, Second Edition, McGraw - Hill Book Company, St.
Louis, 1968.
(12)	Wessel, H.E. "New Graph Correlates Operating Labor Data for Chemical
Processes", Chem. Eng., 59 (7): 208-210, 1952.
-51-

-------
APPENDIX
PREPARATION OF AN ANNOTATED BIBLIOGRAPHY
Introducti on
In the text of this report, PRA has recommended that the reference sections
of the Cost Analysis Manual be updated and changed to an annotated bibliogra-
phy format. The purpose of this appendix is to provide information on ab-
stracts and indexes that deal with water pollution control and an indication
of the level of effort required to revise the reference sections in the Manual
by preparing an annotated bibliography.
There are two ways to prepare an annotated bilbiography, either by means of
a manual search of the literature or by doing a computer-based literature
search. The procedures for compiling and annotating a bibliography are de-
scribed in the following sections.
Literature Search
The first step in a literature search is to compile a list of key words which
is relevant to the subject matter. Some of the computerized data bases have a
thesaurus, for selecting the key words used in their data base. Once a list of
key words has been compiled, the. searcher begins examining pertinent journals,
abstracts, indexes, and/or computerized data bases for information using the
key words. Each citation stored in a computerized data base has a list of
descriptors or identifiers that describe the contents of the publication and
an abstract of the publication. During the search the "computer scans either
the list of descriptors or the title and the abstract for the key words or
phrases. The latter type of search, scanning the title and abstract for key
words, is called free text. If the key words or phrases which are input to
the computer appear in either the descriptors or free text, the citation is
printed. Citations are printed offline at computer centers, so depending on
the location of the searcher, a few days' time should be allowed for mail de-
livery of the printout. Each printed citation includes the following in-
formation: author, title, source of the article (i.e., journal, conference,
etc.), and the abstract. However, it should be noted that the Pollution Ab-
stracts data base does not contain abstracts for articles dated earlier than
1978.

-------
The following abstracts and indexes are recommended sources for obtaining in-
formation on water pollution control equipment. For those abstracts and in-
dexes available in computerized form, the name of the computerized data base
is listed.
e APPLIED SCIENCE AND TECHNOLOGY INDEX
Indexes 218 English-language periodicals in an alphabetical subject
index only. Orientation is toward trade literature. Issued monthly.
e CHEMICAL ABSTRACTS
Computerized data base -- Chemical Abstracts Condensates
Covers the whole field of chemistry and chemical engineering. Over
9000 journal articles, dissertations, patents, technical reports,
conference proceedings, and books are covered. Signed abstracts.
Issued weekly.
e ENGINEERING INDEX
Computerized data base -- Compendex
Major abstract publication for engineering literature. Covers
journal articles, publications of engineering societies, conference
proceedings, and selected government reports in civil, electrical,
environmental, industrial, and mechanical engineering. Published
monthly.
e GOVERNMENT REPORTS ANNOUNCEMENTS AND INDEX
Computerized data base -- NTIS
Covers reports of government-sponsored research and reprints of
articles resulting from government research in engineering, biology,
physical sciences, and life sciences.
9 POLLUTION ABSTRACTS
Computerized data base -- Pollution Abstracts
Covers journal articles, books, government reports, and conference
proceedings in air pollution, marine pollution, freshwater pollution,
sewage/wastewater treatment, solid wastes, land pollution, pesticides/
chemical contaminants, noise pollution, radiation, and environmental
action.
o SELECTED WATER RESOURCES ABSTRACTS
Covers books, journals articles, reports, and conference proceedings
on water quality management including wastewater treatment processes.
e WATER POLLUTION ABSTRACTS
Covers books, journals, proceedings, reports, and foreign literature
in the areas of conservation of water resources, analysis and examina-
tion of water and waste sewage, trade wastewaters and the effects of
pollution. Published monthly.
o U.S. GOVERNMENT PUBLICATIONS: MONTHLY CATALOG
A current bibliography of publications issued by all branches of the
government. Arranged by issuing body.

-------
After those abstracts which relate to the subject matter have been identified,
the journal articles, books, reports, and conference proceedings which the
abstracts summarize should be obtained for review.
Obtaining and Annotating Publications
A copy of as many of the publications as possible should be obtained for re-
view. Many of the journal articles can be photocopied from the serials col-
lection of a local university. Also, through Interlibrary Loan (ILL), books
and government and consultants' reports can be borrowed from any library with-
in the United States. Each publication should be reviewed and discrepancies
in the literature should be noted. When there are discrepancies in the
literature, the annotated bibliography should recommend which reference to
use. Also, if a publication would be especially useful, it should be recom-
mended in the annotated bibliography.
An annotated bibliography can be prepared either by using the abstract found
in the indexes, printed abstracts, or computer printout, or an abstract can be
written after reviewing the publication.
-54-

-------