WATER POLLUTION CONTROL RESEARCH SERIES 12000 FIX 02/71
A System for Industrial Waste
Treatment RD & D Project
Priority Assignment
ENVIRONMENTAL. PROTECTION AGENCY WATER QUALITY OFFICE
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes
the results and progress in the control and abatement
of pollution in our Nation's waters. They provide a
central source of information on the research, develop-
ment, and demonstration activities in the Water Quality-
Office, Environmental Protection Agency, through inhouse
research and grants and contracts with Federal, State,
and local agencies, research institutions, and industrial
organizations.
Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Head, Project Reports
System, Office of Research and Monitoring, Environmental
Protection Agency, Washington, DC 202^2.
-------
A SYSTEM FOR INDUSTRIAL WASTE
TREATMENT RD&D PROJECT PRIORITY ASSIGNMENT
by
Synectics Corporation
4790 William Flynn Highway
Allison Park, Pennsylvania 15101
for the
ENVIRONMENTAL PROTECTION AGENCY
Project Number 12000 FLX
Contract Number 14-12-840
February 1971
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20102 - Price $1
-------
EPA REVIEW NOTICE
This report has been reviewed by the Water
Quality Office, EPA, and approved for publi-
cation. Approval does not signify that the
contents necessarily reflect the views and
policies of the Environmental Protection
Agency, nor does mention of trade names or
commercial products constitute endorsement
or recommendation for use.
-------
ABSTRACT
The Environmental Protection Agency faces the difficult problem of deter-
mining priorities for research and development expenditures which will
yield maximum overall benefits reflected in water quality improvements
at minimum total costs. The computerized management information system
described herein rapidly and efficiently determines such RD&D expenditure
priorities by maximizing the expected returns on investment.
The modeling system developed as a result of this effort is unique inso-
far as information management systems are concerned in the degree to
which user interaction is allowed. At any point during operation of the
system, the user may insert judgmental factors. The system also has been
designed to function with readily available data, such as that from the
Bureau of the Census. Systems which incorporate theoretically desirable,
but virtually unattainable, data have little operational utility. The
mathematical and statistical methods employed in the development of the
system focused on and supported the structuring, testing, and partial
programming of three fixed-X regression modules.
This report is submitted in fulfillment of Contract Number 14-12-840,
Program Number 12000 FLX between the Environmental Protection Agency and
Synectics Corporation.
111
-------
TABLE OF CONTENTS
ABSTRACT
ill
SUMMARY AND CONCLUSIONS 1
Model Form 1
RECOMMENDATIONS 9
INTRODUCTION 11
MODEL DEVELOPMENT 13
Module I. Location 24
Module II. Effluent Constituent 37
Module III. Industry 44
Module IV. Statistical 51
MODEL OPERATION 53
Mode I. MAP 53
Mode II. POA 58
DETERMINATION OF MODEL PRACTICABILITY 59
Functional Integrity and Fidelity 59
Operational Feasibility and Suitability 60
REFERENCES 63
APPENDICES
A. Data Element Sources 65
B. Data Collection Form 0100 73
C. Data Collection Form 6100 77
D. Data Collection Form 4100 81
E. Data Collection Form 2100 83
F. Data Collection Form 8100 89
-------
LIST OF FIGURES
Figure 1. MAPGeneral Summary Level I 3
Figure 2. MAPAction Matrix 6
Figure 3. MAPSubaction Strategy 7
Figure 4. PPBS Categories Ranked According
to Federal Funding 18
Figure 5. Locations by State of Projects
Ranked According to Federal Funding 19
Figure 6. Effluent Constituents Ranked
According to Federal Funding 20
-------
LIST OF TABLES
Table I Potential Dimensions for Ranking
Industrial Wastewater R&D Projects . . . .
Table II Types of Projects Ranked
According to Federal Funding
Table III Project Implementation Ranked
According to Federal Funding
Table IV Project Objective or Applicability
Ranked According to Federal Funding. . . .
Table V PPBS Ranks Based upon Policy
Statements and Funding
Table VI Geographic Parameters
Table VII Location Matrix
Table VIII Period-of-Record Average Water
Quality Measurements by State
Table IX Ranks of States According to
Average Water Quality Measurements . . . .
Table X Common Net Ranks of States
According to Average Water Quality Data. .
Table XI Effluent Constituent Matrix
Table XII Values of Wastewater Constituents
and Related Chemicals
Table XIII Industry Matrix
Table XIV Industrial Classification Cross Reference.
14
21
21
22
25
26
27
30
32
33
39
42
47
50
vii
-------
LIST OF FLOW CHARTS
Flow Chart 1. Location Module 35
Flow Chart 2. Mode I and II Module Control Program 36
Flow Chart 3. Effluent Constituent Module 45
Flow Chart 4. Industry Module III 49
Flow Chart 5. Statistical Module 52
Flow Chart 6. Composite Module 54
Flow Chart 7- Subroutine MAP 56
Vlll
-------
SUMMARY AND CONCLUSIONS
In a dynamic, highly technical society, new products, increasing produc-
tion rates, larger plants and production units, new processes with at-
tendant obsolescence problems, continuing economic pressures, and changing
manpower requirements throughout the society all contribute toward the
need for new effluent treatment and control technology- Hence, yester-
day's rather simplistic techniques are no longer adequate for today's
complex industrial pollution abatement problems. It is a mission of the
Environmental Protection Agency to support research and development on
techniques which will yield maximum overall benefits reflected in water
quality improvements at minimum total cost. In an effort to aid EPA in
this task, a computerizable system for determining industrial waste treat-
ment RD&D priorities for PPBS 1200 projects on a national basis has been
developed and analyzed.
Model Form
The algorithm which has been formulated to establish priorities for R&D
expenditures in the 1200 program series has been built upon five basic
data subsets, each of which considers a variety of independent variables,
i.e., variables which are defined outside of immediate EPA policy con-
siderations at the Headquarters level. Provisions are incorporated in
the management model for definition of and alterations in top-level man-
agement decisions. The overall result is a model which can be formulated
and maintained at a readiness level as an assigned administrative task
and can be readily utilized by policy makers as a dynamic management tool.
The five main subsets are listed below and the potential independent var-
iables associated with each are given; the variables are entered in terms
of relative rankings rather than as absolute values. See Appendix A for
data element sources. Twenty-four information sources are required to
update the modeling system's source data files. Table A-l identifies
the source or sources of each data element described in Table A-2.
1. Effluent Constituents
/ Associated industrial effluent volumes.
/ Concentrations in receiving waters permitting all water uses.
/ Concentrations in receiving water permitting all but the "most
sensitive" water uses.
/ Frequency of mention in State Water Quality Standards.
/ Economic effects on water uses in receiving waters.
/ EPA Regional Office appraisals of relative pollution severities,
/ Degree of public notice.
/ Target treatment costs as determined by maximum values of
recovered materials.
-------
2. Industrial Groups
/ Size of industry.
/ Geographical distribution of industry.
/Water use practices.
/ Economic status.
/ Effluent treatment facilities.
/ Production parameters.
/Effluent constituents.
3. State Dimensions
/ Industrial effluent volumes.
/ Population.
/ Land area.
/ Value added by manufacture.
/ Number of manufacturing establishments.
/ Capital expenditures by manufacturers.
/ Industrial water use.
/ Land area in farms and value of farm products.
/ Population using public water supplies.
/ Annual precipitation.
/ Recreational areas and annual use.
/ Fishing licenses issued.
/ Metropolitan area population.
/ Electrical energy production.
/ Annual water runoff and withdrawals.
/ Scientific population.
4. Project Descriptions
/ Industry involved.
/ Location.
/ Project dates.
/ Sources of funds.
/ Effluent constituents involved.
/ Type of project.
/ Project implementation.
/ Objectives of project.
5. General Statistics
/ Federal/industry R&D funds by region.
/ Federal/industry R&D funds by industry group.
The output of the model is a funding "map" which depicts the priorities
for future R&D expenditures in terms of type of industry, location, ef-
fluent constituent, and level of funding as illustrated in Figures la,
Ib, Ic, 2, and 3.
-------
u>
PPBS PRIORITY STATES % BUDGET % 1206 EFFLUENT CONSTITUENT % BUDGET % 1206
1206 (.3500) NEW YORK (8.34
ILLINOIS 5.92
* IOWA
, MINNESOTA
CALIFORNIA
FLORIDA
WISCONSIN
NEW JERSEY
* IDAHO
0 INDIANA
COLORADO
PENNSYLVANIA
4 MISSOURI
NEBRASKA
4.03
3.92
3.31
2.83
2.65
1.76
.68
.57
.51
.39
.39
.31
23.83
16.94
11.54
11.22
9.47
8.09
7.59
5.04
1.96
1.64
1.47
1.12
1.14
.89
. __ m
BOD (12.80) (36.60) ,
SUSPENDED SOLIDS (11,40) (32.60)
TOTAL DISSOLVED SOLIDS (10.78) (30.80)
.
t
o
e
4
t
,
PPBS PRIORITY STATES % BUDGET % 1204 EFFLUENT CONSTITUENT % BUDGET % 1204 ,
1204 (.2566) NEW YORK
WASHINGTON
FLORIDA
WISCONSIN
GEORGIA
MAINE
* OREGON
o MICHIGAN
ALABAMA
6.97
3.94
3.17
2.92
2.88
1.97
1.67
.79
.75
27.18
15.39
14.48
11.40
TOTAL DISSOLVED SOLIDS (16.29) 63.
BOD 5.16) 19.
SUSPENDED SOLIDS (4.21) 16.'
11.26)
7.69
6.53
3.08
2.94
51
39
10
,
t
t
PPBS PRIORITY STATES % BUDGET % 1202 EFFLUENT CONSTITUENT % BUDGET % 1202 a
1202 (.1324) LOUISIANA
TEXAS
NEW YORK
OHIO
, NEW JERSEY
WEST VIRGINIA
TENNESSEE
. MICHIGAN
PENNSYLVANIA
' VIRGINIA
3.20)
2.53
2.28
1.48
1.19
1.09
.76
.54
.14
.02
)
24.23
19.10
117.26
BOD (6.06) (45.J
COD (4.70) (35.!
SUSPENDED SOLIDS (2.48) (18. '
11.20)
9.00)
8.23)
5.71
4.15
1.09
.12
10
iO .
'0
%
I
Figure la. MAPGeneral Summary Level I
-------
t
PPBS PRIORITY STATES % BUDGET % 1201 EFFLUENT CONSTITUENT % BUDGET % 1201 .
' 1201 (.0908) OHIO (2.83) (31.18) SUSPENDED SOLIDS (7.7
o NEW YORK 2.09
ILLINOIS (1.85
" INDIANA .94)
« MICHIGAN ( .72)
PENNSYLVANIA ( .62)
23.03 IRON ( .5
20.39 ACIDITY ( .'
10.44 OIL AND GREASE .1
7.94 AMMONIA .C
6.98 PHENOLS .C
' CHROMIUM .C
. FLUORIDES .C
CYANIDE .C
TIN .C
o ZINC .t
5 (85.40 °
8 ( 6.41 .
7 ( 1.93
5 ( 1.72)
8 ( .90) ,
8 ( .90
8 ( .90
)8) ( .90) 0
)8 ( .90)
18 ( .90)
)8 ( .90)
* PPBS PRIORITY STATES % BUDGET % 1209 EFFLUENT CONSTITUENT % BUDGET % 1209 '
1209 (.0820) NORTH CAROLINA 7.87)
a TENNESSEE .18)
e MISSOURI .06
MASSACHUSETTS .05
VIRGINIA .05
95.99) TOTAL DISSOLVED SOLIDS (3.
2.26) BOD (2.
.79 SUSPENDED SOLIDS (1 .
.66
.59
95 (48.20)
38 (29.01) *
36 (22.79) ,
t
PPBS PRIORITY STATES % BUDGET % 1212 EFFLUENT CONSTITUENT % BUDGET % 1212
. 1212 (.0313) NEW YORK (2.70)
NEW JERSEY .41)
CONNECTICUT .02)
o MASSACHUSETTS .01)
86.27) TOTAL DISSOLVED SOLIDS (1.
13.03} SUSPENDED SOLIDS (1 .
.51) BOD ( .
.17)
57) (50.26) .
31) (36.16)
45) (13.56)
*
0 PPBS PRIORITY STATES % BUDGET % 1205 EFFLUENT CONSTITUENT % BUDGET % 1205
1205 (.0282) LOUISIANA (1.19)
NEW JERSEY [ .57)
. ILLINOIS .40
CALIFORNIA .30
« TEXAS .28
, PENNSYLVANIA ( .06)
42.52 BOD (2.
20.33 PHENOLS ( .
14.33 SULFIDES ( .
10.98
9.83
1.98)
e
»
e
55) (90.39)
18 ( 6.41)
09) ( 3.19) .
«
t
e
t
a
Figure Ib. MAPGeneral Summary Level I
-------
PPBS PRIORITY STATES % BUDGET % 1207 EFFLUENT CONSTITUENT % BUDGET
1207 (.0167) NEW YORK
MICHIGAN
ILLINOIS
NEW JERSEY
OHIO
WISCONSIN
INDIANA
TEXAS
, PENNSYLVANIA
.68
.27
21
.17
.16
.10
.04
.03
.02
40.61
16.03
12.77
10.03
9.32
5.80
2.49
1.76
1.15
OIL .84
SUSPENDED SOLIDS .43
METALS .40
PPBS PRIORITY STATES % BUDGET % 1210 EFFLUENT CONSTITUENT % BUDGET
* 1210 (.0080) OREGON
WASHINGTON
CALIFORNIA
.49)
.21)
.10)
61.75
26.05
12.17
SUSPENDED SOLIDS .38
BOD .23
COLOR .19)
6 PPBS PRIORITY STATES % BUDGET % 1208 EFFLUENT CONSTITUENT % BUDGET
* 1208 (.0034) ILLINOIS
« OHIO
NEW JERSEY
* CALIFORNIA
. MICHIGAN
PENNSYLVANIA
.13
.06
.06
.04
.03
.01
38.74
18.61
17.86
11.15
9.49
4.13
SUSPENDED SOLIDS ( .26)
FLUORIDES ( .08)
» PPBS PRIORITY STATES % BUDGET % 1211 EFFLUENT CONSTITUENT % BUDGET
* 1211 (.0025) OHIO
» ILLINOIS
CONNECTICUT
* MASSACHUSETTS
.17
.08
.01
.01
66.08
33.00
.59
BOD ,16
COD .07
SUSPENDED SOLIDS .02
.31)
a
o
*
% 1207 t
50.10 *
25.78 .
24.10
*
t
*
t
,
% 1210
47.10 «
28.92 .
23.96
A
* 1208 .
76.26 *
23.73 o
,
%
% 1211 e
(62.86) *
29.04
( 8.08
Figure Ic. MAPGeneral Summary Level I
-------
*
PPBS 1206 (.3500)
*
O
4
4
*
O
*
.
«
1206
NEW YORK
ILLINOIS
IOWA
MINNESOTA
CALIFORNIA
FLORIDA
WISCONSIN
NEW JERSEY
IDAHO
INDIANA
COLORADO
PENNSYLVANIA
MISSOURI
NEBRASKA
ACTION
BOD
1.07
.76
.52
.50
.42
.36
.34
.23
.09
.07
.07
.05
.05
.04
MATRIX
SS
.95
.67
.46
.45
.38
.32
.30
.20
.08
.07
.06
.04
.04
.04
'
DS
.90
*
.64 ,
.43 «
.42 '
.36 ,
.31 «
i
.29
42 ACTIONS
.19 «
.07 «
.06
.06 .
.04 o
.04 *
.03
o
Figure 2. MAPAction Matrix
-------
t
. SUBACTION 1
SUBACTION 2
SUBACTION 3
. SUBACTION 4
. SUBACTION 5
. SUBACTION 6
. SUBACTION 7
SUBACTION 8
. SUBACTION 9
. SUBACTION 10
t SUBACTION 11
. SUBACTION 12
. SUBACTION 13
SUBACTION 14
. ;
SUBACTION 270
e
t
INDUSTRY
INDUSTRY
INDUSTRY
INDUSTRY
INDUSTRY
INDUSTRY
INDUSTRY
INDUSTRY
INDUSTRY
INDUSTRY
INDUSTRY
INDUSTRY
INDUSTRY
INDUSTRY
1206
1204
1206
1206
1204
1206
1206
1206
1204
1206
1206
1206
1206
1204
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
NEW YORK
NEW YORK
NEW YORK
NEW YORK
NEW YORK
ILLINOIS
ILLINOIS
ILLINOIS
WASHINGTON
IOWA
MINNESOTA
IOWA
MINNESOTA
WASHINGTON
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
MAJOR EFFLUENT CONSTITUENT
BOD
DISSOLVED SOLIDS
SUSPENDED SOLIDS
DISSOLVED SOLIDS
BOD
BOD
SUSPENDED SOLIDS
DISSOLVED- SOLIDS
DISSOLVED SOLIDS
BOD
BOD
SUSPENDED SOLIDS
SUSPENDED SOLIDS
BOD
PRIORITY
PRIORITY
PRIORITY
PRIORITY
PRIORITY
PRIORITY
PRIORITY
PRIORITY
PRIORITY
PRIORITY
PRIORITY
PRIORITY
PRIORITY
PRIORITY
.107
.097
.095
.090
.085
.076
.067
.064
.058
.052
.050
.046
.045
.044
% OF BUDGET
% OF BUDGET
% OF BUDGET
% OF BUDGET
% OF BUDGET
% OF BUDGET
% OF BUDGET
% OF BUDGET
% OF BUDGET
% OF BUDGET
% OF BUDGET
% OF BUDGET
% OF BUDGET
% OF BUDGET
1.07
.97
.95
.90
.85
.76
.67
.64
.58
.52
.50
.46
.45
.44
.
.
.
,
.
.
,
.
t
,
,
.
*
t
«
a
a
*
,
;
Figure 3. MAPSubaction Strategy
-------
This study has resulted in the development of a computerizable system for
determining priorities for expenditures in industrial effluent treatment
and control research, development, and demonstration on a national basis,
i.e., for the PPBS 1200 programs. The system takes into account past and
current funding of projects, allows policy decisions on the part of EPA
management to be superimposed at several levels, and permits analysis of
the effects of various policy decisions.
The system has been shown to be practical and workable and can be readily
implemented by machine computation. The initial software requirements,
software maintenance, and hardware requirements are not excessive and
are well within the limits of available resources.
The data requirements of the system are such that maximum use can be made
of public industrial data, the various EPA publications including The Cost
of Clean Water series and the various industry profiles, current effluent
treatment technology information, current compilations of effluent and
water quality data, and the readily available data from the Bureau of
the Census. The system is such that input data format or availability
is not a limiting constraint.
The development of the system has included a means for ranking the sever-
ity of pollution by various industrial effluent constituents in varying
concentrations. Ranking is such that all significant factors involved
in pollution severity for particular constituents can be individually
evaluated and readily altered as additional information becomes avail-
able or as the relative importance of the several factors change due
to hard data acquisition or to policy decisions.
The system has been developed for implementation on the basis of aggrega-
tion of data on a state-by-state basis. The nature and detail of the
basic data are such that extension to a more detailed geographical basis
is not warranted for initial operation. Data have been utilized which
are available on a county-by-county basis. However, the system can be
readily exercised on the basis of geographical units such as census
regions or river basins, if future needs so require. The states as basic
geographical units are felt to be optimum at this time.
The system development has additionally produced a means by which past
EPA policy can be quantitatively depicted and analyzed in detail. It
has also indicated some desirable modifications in the format and con-
tent of basic input information, particularly in the "Needs Statements"
as inputed to the EPA Technical Information and Management Planning
System (TIMPS).
-------
RECOMMENDATIONS
1. Based upon favorable design and practicability results, an RD&D pri-
ority and fund allocation model should be implemented for use in 1200
PPBS category management. Owing to the current developmental state
of the model, two further steps are necessary to bring the model to
operational readiness in a form which generates a funding allocation
"map" (MODE I). These are:
Step 1. To specify operating file structures, finalize report
formats, adapt programs for each module (4) and intermodule
linkages to the EPA machine configuration, and develop data-
base maintenance and update procedures.
Step 2. To conduct critical and volume tests on the model to
demonstrate satisfactory performance in support of the Indus-
trial Pollution Control Branch technical management staff.
(1200 PPBS category "Need Statements" should provide a suffi-
cient data base for this purpose.)
2. The RD&D priority and fund allocation modeling technique could readily
be extended to include the entire PPBS category set. Each category
should be examined, however, to determine the optimum form and prac-
ticability of model functions for priority determination and fund
allocation within the specific category.
3. A most important application of the priority and fund allocation
model would be its incorporation into the Technical Information and
Management Planning System (TIMPS) as a user accessible subsystem
on an interactive basis. Since the key to successful operation of
TIMPS is some form of RSD priority list generator, it appears highly
desirable to incorporate the modeling subsystem in TIMPS as early
as possible. In this connection, a vital consideration should be
the optimization of the hard/software (console and query subroutines)
which bridges the man-machine interface in order to permit facile
and responsive information transfer between manager and data base.
As a consequence of incorporating the model within TIMPS, a somewhat
modified information flow would result which, in general, should
proceed as follows:
a. A need is generated and an area priority will be assigned.
b. The need is coded and entered on the need file.
c. The model is exercised and all existing needs assigned priorities
and ranked.
-------
d. The model-generated priority "map" is generated as a function of:
(1) Priorities at the beginning of the fiscal period.
(2) Current priorities remaining to be funded.
(3) Where and how funds have been allocated to date in the
fiscal period.
e. The "map" is submitted for executive committee review.
f. A new "map" is generated to reflect the decisions reached in
executive review.
g. Priorities are stored for submittal to central data processing.
4. To enhance the ease and precision of requirements entries into the
data base (in conjunction with the preceding recommendation) , the
"Statements of Need" should be modified and restructured to include
model significant data. The latter changes should afford those re-
sponsible for initiating requests an improved, more incisive means
for describing their requirements.
5. The STORET system should be modified so as to incorporate Water Quality
Standards and to generate a yearly measure of observed deviations from
standards at each location.
6. An effort should be made to develop the present costs of treatment
of specific constituents from such EPA programs as the Industrial
Waste Inventory, Effluent Criteria Development Projects, and Dis-
charge Permit Program.
10
-------
INTRODUCTION
As the complexity of manufacturing technology increases force changes in
pollution abatement requirements to achieve adequate protection of the
Nation's water uses, industrial waste treatment and control requires new,
improved methods. That is, new products, increasing production rates,
larger plants and production units, new processes with attendant obso-
lescence problems, continuing economic pressures, and changing manpower
requirements throughout industry all contribute toward the need for new
effluent treatment and control technology. Hence, the relatively sim-
plistic means of the past are no longer adequate for today's complex
industrial pollution abatement problems.
The Environmental Protection Agency faces the difficult problem of deter-
mining priorities for research and development fund allocations which will
yield maximum overall benefits as reflected in water quality improvements
at minimum total cost. Fiscal constraints, existing suitable research
and development facilities, and manpower insist that R&D expenditures
return the greatest possible benefits. A computer-aided management in-
formation system which rapidly and efficiently determines such R&D ex-
penditure priorities on a realistic cost/benefit basis will be of great
value. The purpose of this study was to design such a system and to
determine its practicability.
R&D expenditures applicable to the industries included in PPBS Subprogram
1200 of the EPA R&D Program were selected as the focus for this study,
which made maximum use of current data compilations. Existing data com-
pilations did not properly conform to the above industrial classifica-
tions, nor were the costs and benefits of most R&D expenditures strictly
assignable within specific industrial groups; additionally, plants and
processes, as well as the nature and magnitude of pollution, varied within
groups. Such factors, however, did not preclude the use of available data;
rather, the data merely required recognition and proper treatment in their
use.
The cost side of the economic evaluation generally presents relatively
little difficulty in cost/benefit analysis as compared to the benefit
side; the cost/benefit analysis of R&D expenditures was no exception.
The literature is replete with dissertations on various methods by which
benefits, both economic and non-economic, may be determined and few spe-
cialists in the field are in complete agreement on methodology or even
on some definitions of benefits. However, these problems were reduced
in this study, since relative benefits were of primary interest. While
absolute dollar-expressible benefits vary with methodologically different
approaches, the benefits may still be in relative agreement. For example,
the methodologically different analyses of Kneese (1964) and Bramer (1966)
reach the same conclusions as to the relative values of the various uses
of the surface waters, placing the highest values on recreational uses.
11
-------
A computerized system for determining priorities in R&D expenditures was
known to be technically within the state-of-the-art of systems analysis,
mathematical modeling techniques, economic theory, information handling,
and machine capabilities. The desirability of such a system as a manage-
ment information tool for the optimum utilization of R&D resources was
also clear. The feasibility and practicability of such a system would
be determined by the quality of the system design and the mode chosen
for eventual operation. In order to design and determine the practica-
bility of the desired computerized modeling system, consideration was
given to the various critical components, i.e., the user requirements,
the data base, and the component requirements of the system itself. A
practical system must be able to operate within existing resources and
constraints, particularly available data; an otherwise optimum system
would be of little utility. A practical system must also meet the needs
of the users. The system itself must show a cost/benefit advantage to
the user over other alternatives for determining R&D expenditure priori-
ties such as manual computations or individual judgments.
The developmental system has two basic operating modes:
Mode I - operates on specific input parameters to generate a funding
"map" which describes an optimum procedure for allocating available
monies to industrial effluent treatment RD&D as functions of Indus-
try, Location, Effluent Constituent, and Percent of Budget, at three
levels of detail.
Mode II - ranks proposed projects or "needs" as a function of speci-
fic parameters such as industry, location, etc,, and assigns a pri-
ority to each project.
The succeeding sections describe the preliminary design, development, and
practicability assessment of the model.
12
-------
MODEL DEVELOPMENT
As is the case with most mathematical model development, the basic ap-
proach was: Identify and define measurable variables, collect and anal-
yze a set of relevant "real-world" data, develop and test hypotheses
regarding the nature of apparent relationships, and refine the mathe-
matical statements of relationships by determining their predictive
accuracy on fresh "real-world" data. The data drawn from the "real
world" was in this case comprised of information on funded projects,
described in the FWQA Water Pollution Control Research Series publica-
tion DAST-38 entitled "Projects of the Industrial Pollution Control
Branch" (1970). Table I gives a list of potential dimensions for de-
scribing these projects. This list was re-evaluated with respect to the
availability of data and the phenomenon of funding industrial waste treat-
ment aad control research, development, and demonstration projects in
terns of six possible criterion variables and 83 possible predictor vari-
ables. Criterion variables are those dependent variables that best de-
scribe or measure the phenomenon which the model developer is trying to
predict or model. Predictor variables, on the other hand, are those in-
dependent er dependent variables that best "predict" the criterion vari-
ables. See Appendix B for the sample data collection From No. 0100.
Each of the 126 projects described in "Projects of the Industrial Pollu-
tion Control Branch" was coded on Form No. 0100 and entered into the
computerized data base. The data base was then subjected to associative
statistical analysis and rank ordering, resulting in base level ranks for
each variable and the degree of covariance associated with all variable
combinations. Based on these results, it was concluded that Federal funds
per project showed the most promise as a criterion variable. That is, it
was initially felt that Federal funds per project, Federal funds per month
per project, total funds per project, total funds per month per project,
other funds per project, and other funds per month per project each might
prove to be rational criterion variables. However, early correlation
analysis showed Federal funds per project to be most closely associated
with the 83 predictor variables. Federal funding for all the projects
examined was sorted on the basis of each of the following predictor
variables:
/ PPBS category
/ State in which project is located
/ Effluent constituents to be removed or reduced
/ Type of project
/ Organization implementing the project
/ Objectives of the project
Rankings of these variables and their respective predictor variables as
a function of Federal funding constitute a succinct statement of prior
EPA policy in RD&D project funding. The first three variables were shown
13
-------
Table I
Potential Dimensions for Ranking Industrial Wastewater R&D Projects
I. Industry Involved
A. Description
1. PPBS number
2. 2-digit SIC code
3. 3-digit SIC code
4. 4-digit SIC code
B. Size of Industry
1. Number of plants
a. Number of small plants
b. Number of large plants
2. Number of employees
3. Value added by manufacture
4. Total water use
a. Intake for process
b. Intake for cooling
c. Intake for sanitary and other uses
C. Geographical distribution
1. Number of states with plants
2. Number of plants per state with plants
3. Water discharged per state with plants
4. Number of employees per state with plants
5. Population in states with plants
D. Water use practices
1. Rate of water reuse
2. Water intake per dollar added by manufacture
3. % purchased water
4. % ground water
5. % brackish water
E. Economic status
1. Age of average plant
2. Return on investment
3. Return on sales
4. Unemployment rate
5. Projected growth rate
F. Wastewater treatment practices
1. % treatment facility investment of total investment
2. % treatment facility investment of projected requirements
3. % wastewater discharged to municipal sewers
G. Production parameters
1. Operational days per year
2. Production as % of capacity
3. Level of technology
a. % old technology
b. % averaged technology
c. % advanced technology
4. Average life of process equipment
14
-------
II. Objectives of RSD Project
A. Reduction of wastewater constituents
1. Type of constituent
2. % reduction
B. Wastewater discharge reduction
1. % reduction
2. % to municipal treatment facilities
C. Water use reduction
1. Process water
2. Cooling water
3. Sanitary and other uses
4. % reduction
D. App li cabi li ty
1. Large plants
2. Small plants
3. Old technology
4. Average technology
5. Advanced technology
E. Applications
1. Treatment processes
a. New process
b. Process improvement
2. Process equipment
a. New equipment
b. Equipment improvement
3. Cost reductions
a. Capital costs
b. Operating costs
4. Treatment control
a. Measurement
i. Methods of analysis
ii. Methods of sampling
iii. Instrumentation
b. Automation
5. Engineering
a. Process design methods
i. conventional
ii. Mathematical models
iii. Cost calculations
b. Treatment plant design methods
i. Conventional
ii. Mathematical models
iii. Cost calculations
c. Information handling
6. Production process modifications
7. By-product recovery
8. Manpower factors
a. Training
b. Requirements determinations
c. Requirements reductions
i. Man-hours
ii. Level of skills
15
-------
III. Type of R&D Project
A. Research
1. Desk-top study
a. Technology
i. State-of-the-art report
ii. Conceptual study
b. Economic study
c. Manpower study
2. Hands-on research
a. Laboratory
b. Bench-scale pilot plant
B. Development
1. Special facility
2. Industrial in-plant
C. Demonstration
1. Industrial in-plant
2. Joint industrial-municipal
3. Residual pollution abatement
D. Contract research
E. Grant
IV. Project Implementation
A. Profit-making R&D organization
B. Not-for-profit R&D institution
C. Manufacturing industry
D. Government unit
1. Federal
2. State
3. Regional
4. Local
E. Joint effort
F. University
G. Organizational qualifications
1. Previous performance
a. Organization
b. Research staff
c. EPA projects
d. Other federal agency projects
2. Staff qualifications
a. Professional staff
b. Support staff
3. Facility resources
a. Technical resources
b. Support resources
4. Financial resources
a. Matching funds
b. Operating funds
c. Accounting procedures
16
-------
V. Project Initiation
A. Request for proposal
B. Unsolicited proposal
C. Public notice
1. Local
2. Regional
3. National
4. Public press
5. Technical literature
6. Legislative
VI. Spin-Off Benefits of R&D Project
A. Training
1. Academic
2. On-the-job
B. Stimulation of non-federally funded R&D
C. Other applications of technology developed
D. Applications of other technology
1. Military
2. Aerospace
3. Other
VII. EPA Administrative Considerations
A. R&D budget
1. Total
2. Industrial wastewater control
3. Administrative industrial allocations
a. Contract research
b. Grants
c. Demonstrations
B. Fiscal year period
C. Prior year allocations
1. Industry
2. Type of R&D
D. Previous similar projects by others
1. Technical similarity
2. Level of funding
E. Current similar projects by others
1. Technical similarity
2. Level of funding
F. Utilization of prior base R&D
G. Projected time to implementation of R&D results
to be non-linear and their distributions are given in Figures 4-6; the
latter three are considered to be linear in accordance with the distribu-
tions shown in Tables II-IV. Note that the entries in these tables are
not mutually exclusive; that is, a project can be characterized by more
than one "type." These data may be interesting in themselves insofar as
they indicate prior EPA funding policy.
17
-------
9000-
8000"
7000'
2 6000-
H
rH
0
T3
O
w 5000-
5
§
0
^
to 4000-
T3
^j
Pn
H
05
lu 3000-
d
i -d Ci o o
ff C 0) S-
rH HJ fl 4-
o -na
rl X! P.
0) O
Q) n (0
5 3
co
H
i a
d --
<1) O
M 0
C m
d co
m 5 -P
CM <9 u
*^^ ^J
d -d
10 O O
V on
CJ P4 ft
t>
o
k
cu
d
(D
rl
H
i-H
d
^ 3
CM U
d,
s
-d
o
rl
co -d
0 Q)
H
*-^ rH
CO
-P >d
1 §
o to
ft JO
o
H -H
TO* S
P aj
(D ^3
i
rH
(0
-P
1
~1
1
1208 1210 1207 1205 1212 1209 1201 1202 1204 1206
PPBS Category
Figure 4. PPBS Categories Ranked According to Federal Funding
18
-------
3000 -I
CO
vo
^2500 -
01
H.
r-l
TlJ
4,2000-
O
rn
S
c
|9 1500-
0
A
^Tj.
3
C 1000 ~
£3
P4
id
n
^ 500-
0)
^^
PI ^^
00 C*
o
^~ * CM
in ^» H
in id oo
(3 cr* id ^j1
*-^ PI *rH *-^ *-^ I 1 «fc^
^ O ^rHPIMPlO
-^ rHt- O O O ^ C
r~ ^^m- £M id -H
^" ^^ QO-CN^^Oldld^^QO)
» rHCN rHJJCJ-r)-'-' C
HO'-'cntPtT" A O
ID ^ .^CM ld'«j3r|ldO43O
CM ^^rH ^»^,^ CMVD 0>-HCM-iH4JOlS-HrlO]
^ o»vo cnpi^i'id'^^o 0)C-'^3r(0)O.c!O-rH
VO ^.^, PI. ^O-*^«-rlP) 'COlrl 01OOHO3S
o mo --'^^vO'''fiin'acMi3oioo)id3
in f^^ *-'-HPlTJ-H«» M' rHQ^SCrHld |-^
O MOrH rH'dfj. >lldld-H >t.£UO)iidoi^i4Jid'drao) ""
rH .jsBcoojSfloigidraHoirtHB [~1 1~|
id ido-ri ido)-iJ.Pid'33rHxo« __ ___ i
M*O idbi-PGridrHOoioJwu i"n n F| I 1
JJ (ri r; 01 h C r* Ql Q iy r* n. Q rf ^H^K 1 II II II 1
01 ci3 01 -H o o swoSnnilllllll
3(d4-i(d>sua) i 1 1 i r~l 1 II II II II II 1
< o D s _,-,i-innnnni n i| n n n i
o
in
C
O
d)
rl
O
00
CM
in
01
01
01
rj
^
Q)
3
^^
CM
in
in
01
H
rl
rH
rH
H
~
^
VO
a\
in
H
01
H
a
o
^^^
PI
o
VO
id
3
rH
CK
^^
cJo
t--
VO
id
4J
0
0]
01
5
^
1
u
z
10
15 26 2^
Project Sites - States
Figure 5. Locations by State of Projects Ranked According to Federal Funding
-------
NJ
O
13000 n
12000 -
11000 -
CO
T)
10000
I
I
j 9000
I
I
I 8000
I
i 7000 -
I
6000 -
5000 -
<* 4000
0)
T3
-------
Table II
Types of Projects Ranked According to Federal Funding
Rank
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
Priority
.324
.283
.243
.614
.066
.041
.040
.100
.088
.079
.042
.030
.014
.014
.012
.002
.001
Type of Project
Grant
Demonstration
Full-scale
Industrial in-plant demonstration
Pilot plant
Research
Development
Laboratory
Hands-on research
Industrial in-plant development
Bench-scale pilot plant
Special facility
Joint industrial-municipal treatment
(except 1106)
State-of-the-art report
Desk-top study
Conceptual study
Contract
Table III
Project Implementation Ranked According to Federal Funding
Rank
9
8
7
6
5
4
3
2
1
Priority
.592
.164
.124
.041
.026
.023
.015
.011
.001
Project Implementation
Manufacturing industry ( . 708)
Government unit ( . 196)
Local government
University (.049)
Profit-making R&D organization (
Regional government
State government
Not-for-profit R&D organization
Federal government
.031)
(.013)
21
-------
Table IV
Project Objective or Applicability Ranked According to Federal Funding
Rank
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
5
5
4
3
2
1
1
1
1
Priority
.200
.181
.100
.077
.077
.072
.042
.040
.032
.019
.019
.018
.018
.016
.010
.009
.007
.006
.006
.005
.005
.004
.004
.004
.003
.002
.001
.001
.001
.001
.001
.001
Objective or Application
Treatment process study
Treatment process improvement
Economic feasibility determination
Engineering
Sludge disposal
Process design
By-product recovery
Water reuse
Plant design
Treatment control
Large plants
Effluent characterization
New treatment process
Measurement methods
Manpower factors
Manpower requirements
Production process modifications
Effluent volume reduction
Cost reductions
Information systems
Small plants
Conventional engineering techniques
Old technology
Average technology
New technology
Effluent effects
Process equipment
Mathematical models
New process equipment
Automation
Training
Manpower reduction
22
-------
From the viewpoint of developing a computer-aided priority determination
system, the nature and derivation of each of the predictor variables is
of great importance. The independent predictor variables are dimension-
able within themselves and are simple in nature, i.e., they are not func-
tions of other factors. The dependent predictor variables are complex,
i.e., they are functions of factors which lie outside of current EPA
policy control at the Headquarters level.
The simple variablesthose dimensionable within themselvescan, of
course, be aggregated as one function, which predicts Federal funding to
some low degree. However, it is essential to treat each complex variable
as a separate predictor function, thus segregating the influence of each
on the criterion variable. Three such variables were identified:
/ Industry
/ Severity of pollutant
/ Location
Furthermore, it was desirable to modularize the model system according to
these three variables, since each assumed a different hypothesis regard-
ing the amount of Federal funding per project. That is, each module re-
quired a basis for grouping variables in terms of the homogeneity of
associated prediction functions, in order to identify and describe a po-
tential rating policy for determining KD&D priorities. The point of
departure selected for deriving hypotheses was recent EPA policy. This
choice presented an additional problem, however, in that sufficiently
precise measures of policy had to be identified to permit quanitative
analysis. Although structured interviews with key EPA personnel pro-
vided considerable insights into agency goals and policies, it was also
necessary to identify an additional measure to augment these results.
The method utilized a fixed-X multiple linear regression model to gen-
erate mathematical equations defining and quantifying recent policy
(Christal, 1968). Simply put, the procedure for each module was as
follows:
/ Criterion variable and all reasonable, possible, pre-
dictor variables were defined.
/ Each variable was then scaled according to the appro-
priate properties possessed by that variable. Since
the policy capturing model relies on fixed-X assump-
tions, variables need not be normally distributed.
/ The model was then exercised over the data base and
the resultant equation examined for significance.
/ Where appropriate, additional predictor variables were
identified and included in the analysis, terms were
altered to approximate curvilinear functions, and in-
significant terms were dropped to maintain simplicity.
After each alteration, the model was exercised to gen-
erate a new function.
23
-------
To assess the approach used, prior EPA policy as measured by the Federal
funding in the various PPBS categories, was compared with announced
policy as expressed in the listings "Sources of National Pollution Pri-
orities" and "1200 Program Schedule of Milestones." In Table V, a com-
posite net ranking (policy rank) is shown for the 1200 categories and
compared with the ranking based upon prior funding; the differences be-
tween these ranks (d) are tabulated.
Comparing these rankings, using Spearman's rank correlation analysis:
6 (Ed2) _
S N (N2 - 1)
_ 6 (54) =
rs 12 (144 - 1)
The rank correlation coefficient is significant at the .01 level and in-
dicates that prior policy as measured by project funding is reliably re-
lated to stated policy. Similar procedures were used to enhance the
validity of the effluent constituent and location modules until the cost
of this incremental increase in accuracy surpassed the value of that in-
crement .
Module I. Location
Although various descriptors of geographic location are available, a
state basis proved to be most realistic, since it is a common denomina-
tor of the others and is the basis on which most data are readily avail-
able. Table VI indicates the relationships associating each state with
the following other descriptors:
/ Cost of Clean Water Series Drainage Regions
/ Water Resources Council Drainage Areas
/ Census Bureau Industrial Water Use Regions
/ Census Bureau Standard Regions and Divisions
Appendix C contains the data collection form for location dimensions.
Each state was ranked as a function of Federal funding (Figure 5, page
19) . This ranking reflects the relative importance of the location
dimensions associated with each state and is referred to as the rank
order in Table VII. For example, project location in New York was at-
tributed 1.72 times the significance of location in Minnesota and 166
times the significance of location in Virginia. Associative statistical
analysis indicated that the most important predictor variables associated
with location of projects by state are:
/ Industrial effluent volume per state, in bgy, a = 447.8,
3 = -.1156
/ Population per state, in thousands, a = 6507, 0 = -.1426
24
-------
Table V
PPBS Ranks Based upon Policy Statements and Funding
PPBS Number
Policy Rank
Funding Rank
d
1206
1204
1205
1202
1201
1209
1210
1203
1207
1211
1212
1208
11
10
9
8
7
6
5
4
3
3
2
1
11
10
5
9
8
7
3
1
4
1
6
2
0
0
4
-1
-1
T_
2
3
-1
2
-4
-1
/ Value added by manufacture per state, in millions of
dollars, a = 2418, $ = .4173
/ Annual runoff per state, in thousands acre ft/yr,
a = 21830, 3 = -1974
/ Water area in each state, in square miles, a = 1517,
3 = -.3616
/ Population density per state, in square miles,
0 = 2079, 3 = -.1469
/ Industrial water used per state, in bgy, a = 1581,
3 = .0800
25
-------
Table VI
Geographic Parameters
Cost of Clean
Water Series
Drainage Regions
North
Atlantic
Ohio
Western
Gulf
Missouri
Great Lakes
Great
Basin and
Colorado
Southeast
Upper
Mississippi
Lower
Mississippi
Arkansas-
White-
Red
Pacific
Northwest
California
Water Resources
Council
Drainage
Areas
North
Atlantic
Ohio
Rio Grande
Texas -Gulf
Sour is -Red-
Rainy
Missouri
Great Lakes
Upper
Colorado,
Lower
Colorado,
Great Basin
south
Atlantic-
Gulf
Upper
Mississippi
Lower
Mississippi
Arkansas-
Red-
White
Columbia-
North
Pacific
California
0 All
O Little
Most
Census Bureau
Industrial
Water Use Regions
New England,
De laware anc
Hudson,
Chesapeake
Bay
Ohio
Cumberland
Western
Gulf
Missouri
W. Great
Lakes, E.
Great
Lakes, St.
Lawrence
Great
Basin and
Colorado
Southeast
Upper
Mississippi
Lower
Mississippi
Arkansas-
Red-
White
Pacific
Northwest
California
Census Standard
Regions and
Divisions
Divi-
sion
Region
0)
c
1
e
New Hampshire |
e
Vermont |
O
o
Massachusetts |
e
d
rH
«
£
e
3
-H
Conned
e
New England
ji
s
I
0
o
>,
-------
Table VII
Location Matrix
State
New York
Minnesota
Florida
Louisiana
Illinois
New Jersey
Oregon
Wisconsin
North Dakota
Ohio
Iowa
Georgia
North Carolina
Washington
Maine
California
Tennessee
Michigan
Idaho
Maryland
Texas
West Virginia
Indiana
Alabama
Kentucky
Colorado
Arkansas
Delaware
Pennsylvania
Rhode Island
Rank
Number
40
39
38
37
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
Rank
Order
11.68
6.78
6.03
5.96
5.52
5.28
5.02
4.12
4.09
4.03
4.02
3.98
3.83
3.55
2.91
2.70
2.51
2.18
2.06
2.04
1.53
1.47
1.44
1.43
1.09
.86
.65
.64
.61
.39
Industrial
Wastewater
Volume
569.
87
230
843
591
395
151
236
1
1,115
103
213
146
341
163
313
287
739
47
401
1,455
67~0
830
242
117
54
42
164
1,475
16
Population
17,894
3,529
5,654
3,493
10,538
6,680
1,886
4,100
650
10,124
2,763
4,304
4,861
2,971
984
18,003
3,805
8,161
687
3,442
10,401
1,823
4,832
3,431
3,163
1,941
1,939
494
11,505
884
Value Added
by
Manufacture
7,010
1,166
874
1,168
6,465
4,731
488
2,853
17
8,821
1,165
1,390
2,072
2,041
326
6,879
1,499
8,368
197
1,674
4,196
1,433
5,097
1,311
1,585
708
373
508
6,926
291
Annual
Runoff
58,168
16,141
40,600
43,995
30,079
8,359
67,238
29,948
1,886
26,382
13,809
47,101
47,793
76,375
44,286
71,937
45,061
37,258
37,877
9,025
49,909
24,503
23,226
57,802
36,626
22,798
48,149
1,646
48,356
1,424
Water
Area
1,707
4,779
4,424
3,368
523
304
772
1,690
1,385
204
247
679
3,706
1,529
2,282
2,156
878
1,398
880
686
4,369
97
102
758
544
453
929
75
308
165
Population
Density
350.1
42.7
91.5
72.2
180.2
806.6
18.4
72.2
9.1
236.9
49.2
67.8
92.9
42.8
31.3
100.4
85.4
137.2
8.1
314.0
36.4
77.3
128.9
64.0
76.2
16.9
34.0
225.6
251.5
812.4
Industrial
Water
Use
1,054
208
690
2,141
1,473
814
270
495
23
1,935
149
655
350
996
393
1,273
649
1,299
141
531
5,069
817
1,283
583
414
122
451
286
2,903
19
to
-------
(Table VII (Continued)
State
Oklahoma
South Carolina
Puerto Rico
Connecticut
Montana
Virginia
Massachusetts
Utah
Canada
Australia
Alaska
Arizona
District of Columbia
Hawaii
Kansas
Mississippi
Missouri
Nebraska
Nevada
New Hampshire
New Mexico
South Dakota
Vermont
Wyoming
American Somoa
Canal Zone
Guam
U. S. Virgin Islands
Pacific Island Trust
Territories
Rank
Number
10
9
8
7
6
5
4
3
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Rank
Order
.38
.26
.15
.11
.07
.06
.05
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
Industrial
Wastewater
Volume
10
101
-
118
26
275
144
27
-
-
34
10
-
102
43
65
82
24
4
35
1
5
7
7
-
-
-
-
~
Population
2,461
2,528
2,578
2,784
703
4,371
5,287
977
-
-
256
1,549
795
712
2,237
2,304
4,471
1,471
418
659
1,008
700
399
338
-
-
-
-
Value Added
by
Manufacture
369
1,154
-
2,153
89
1,741
2,360
463
-
-
29
318
"
117
900
345
2,260
285
50
158
26
67
88
45
-
-
-
-
-
Annual
Runoff
19,018
24,844
-
5,877
29,823
32,652
10,129
8,152
-
-
-
4,250
-
-
12,285
53,442
40,877
7,414
3,537
12,405
3,893
2,875
12,812
20,366
-
-
-
-
Water
Area
935
775
14
139
1,535
976
424
2,535
-
-
19,980
346
6
25
208
358
640
705
651
271
221
1,091
335
633
-
-
-
-
-
Population
Density
33.8
78.7
686.8
517.5
4.6
99.6
654.5
10.8
-
-
0.4
11.5
12523.9
98.6
26.6
46.1
62.5
18.4
2.6
67.3
7.8
8.9
42.0
3.4
-
-
-
-
-
industrial
Water
Use
290
332
-
205
60
459
207
176
-
-
95
125
-
245
231
231
221
46
11
79
11
15
9
60
-
-
-
-
-
00
-------
B, the standard partial regression coefficient, is a measure of the de-
gree of importance EPA has associated with each predictor variable with
respect to past policy and Federal funding. Of course, 3 coefficients
will continue to serve as important measures of policy in the future by
constantly keeping EPA management abreast of the relative importance they
are placing on predictor variables. 3 times 100 represents the percentage
change that will occur in the criterion variable for a given change in
the predictor. For example, value added by manufacture has a 3 of .4173.
This means that when there is a 1 standard deviation change in value
added by manufacture, there has been a .42 standard deviation change in
EPA funding. Noting that the standard deviation (a) for value added is
2418 and the standard deviation for rank order is 2.075, given a 2418
increase in value added, we would find a (.4173) (2.075) = .8659 increase
in rank order. Thus, we can see that value added was an important vari-
able in the past. On the other hand, industrial effluent volume has had
a slight negative effect in the past, that is to say, the higher the in-
dustrial effluent volume per state, the less money EPA has allocated to
that state. Again, this is a statement of past policy and now that it has
been measured, it can be adjusted to meet current thinking through the
model discussed herein. Thus, an increase of 447.8 of industrial effluent
would produce a rank order reduction of (-.1156) (2.075) = -.2399. Of
course, a similar exercise can be carried out for each predictor variable.
We are only interested in past policy as a starting point for future model-
ing, therefore we will not dwell on the subject, but will leave any fur-
ther exercises up to the reader.
In general, EPA has given higher priorities to states with:
/ Lower industrial effluent volumes
/ Lower populations
/ Higher values added by manufacture
/ Higher annual runoffs
/ Higher water areas
/ Lower population densities
/ Any volume of water used
Table VII shows the matrix of values for the predictor variables with re-
spect to the criterion. In addition, data were obtained from STORET for
some 400 stations. Station data were aggregated by state for measurements
of temperature, conductivity, dissolved oxygen, and pH. For each station,
data were available for these parameters in terms of averages for the
period of record which was generally 6-12 months. A tabulation was made
of the median, minimum, and maximum observed average values in each state.
In Table VIII, the maximum average values of temperature and conductivity
and the minimum average values of dissolved oxygen and pH are tabulated.
The only data discarded were conductivity values in Maine, Rhode Island,
and South Carolina which were obviously measurements in seawater.
Water quality data may be used in a number of ways as indicators of pri-
orities for pollution abatement needs; indeed, such data may well be of
prime overall importance. It has been the intent of the present study,
29
-------
Table VIII. Period-of-Record Average Water Quality Measurements by State
State
Alabama
Arizona
Arkansas
California
Colorado
Connecticut
Florida
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Puerto Rico
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Wisconsin
Wyoming
Alaska
Delaware
District of Columbia
Georgia
Hawaii
Washington
West Virginia
Max . Tem-
perature , °C
19.3
29.8
18.9
21.4
13.4
8.3
28.0
16.5
16.1
12.4
11.0
17.7
13.1
20.6
14.6
16.1
13.3
9.9
12.6
19.7
15.6
14.7
30.3
17.3
16.3
15.7
15.1
23.6
15.2
19.1
19.3
24.7
19.6
17.9
12.7
17.7
9.8
17.0
19.0
14.0
9.0
15.1
3.4
10.5
-
-
-
-
-
-
Max . Con-
ductivity, mhos
155
6334
803
1399
6334
109
669
4823
562
748
820
2405
407
409
78 (1)
719
269
222
525
7808
820
2022
18413
1716
5525
120
5823
13000
729
8940
896
748
2178
434
- (1)
5158 (1)
2680
358
812
3553
96
440
232
3227
_
_
_
_
_
_
-
Min.
D . 0 . , ppm
7.2
8.0
5.1
8.9
7.5
9.4
3.4
8.7
9.3
9.5
8.6
7.6
9.6
5.1
7.3
8.1
8.9
11.9
8.5
3.1
6.5
4.5
0.8
9.6
9.6
7.1
4.3
0.4
0.8
0.7
1.8
2.1
5.6
0.2
0.1
0.8
9.5
9.5
8.2
6.8
10.0
8.2
13.4
8.5
_
_
_
L
-
Min. pH
6.9
7.0
7.8
8.1
6.8
7.2
6.7
7.7
7.5
7.5
7.6
7.6
6.7
6.9
6.4
3.4
6.9
7.1
7.9
6.2
7.6
7.3
6.2
7.8
7.8
5.0
4.9
7.4
6.8
6.4
7.7
6.8
8.1
4.8
7.7
6.8
7.6
4.6
7.2
7.4
7.1
6.4
8.2
7.2
-
(1) Measurements in seawater discarded.
30
-------
however, to utilize data that are readily available; little suitable data
are presently available in sufficient detail or in a satisfactory form.
The data used here to illustrate a method of ranking locations by water
quality were those available from the STORET system. It is intended only
to illustrate a need; the data were not judged to be of sufficient quality
for incorporation into the model. Some states yielded complete data for
dozens of locations; others had sparse data for only a few locations.
Little basis was indicated for statistical adjustment in terms of confi-
dence levels or otherwise.
The STORET system itself should be modified to supply the needed data
with little manipulation required within the priority-determining model.
If the Water Quality Standards were incorporated into the STORET system,
the system might compare each recorded water quality measurement with the
applicable Standards and report deviations, weighted as to magnitude for
each location. The summation of such weighted deviations per year per
state might then be used as a relative water quality indicator. In each
case, of course, corrections would have to be made for frequency of mea-
surement, numbers of locations measured, proportions of each state's
streams monitored, and similar factors which might bias the indicator
function.
The most reasonable interpretation of the available water quality data
for the present purposes would be to consider them as measures of the
quality of the surface waters, making no assumptions as to origins of
constituents. We might then conclude that poorer quality is related to
higher priorities for pollution abatement, either because pollution loads
are greater or because pollution cannot be tolerated due to already poor
available water quality.
On this basis, the states are ranked according to each of the four water
quality parameters with a rank of 1 equivalent to best quality, i.e., the
lowest priority for pollution abatement measures as shown in Table IX.
In Table X, each series of ranks is reduced to a common basis to yield
ranking factors which were then summed to yield composite net ranks. On
the basis of the data used, these composite net ranks show the relative
status of the states insofar as prevailing water quality is concerned,
and may be interpreted as discussed above, as one measure of the need
for pollution abatement.
The policy capturing model produced the basic structure for equation (1)
as a result of analyzing the data in Table VII. However, interviews with
potential users suggested that the user should have an opportunity to
control the influence of past policy; thus, the policy capturing equation
was augmented to provide such an opportunity. Notice that the regression
coefficients (constants) in equation (1) are consistent with the previously
discussed 3 coefficients. For example, -.96V1 accounts for the -.1156
standard deviation change in PJJ.
31
-------
Table IX. Ranks of States According to Average Water Quality Measurements
State
Alabama
Arizona
Arkansas
Call f ornia
Colorado
Connecticut
Florida
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Mas s achus e t ts
Michigan
Minnesota
Mississippi
Missouri
Montana
Puerto Rico
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Wisconsin
Wyoming
Alaska
Delaware
District of Columbia
Georgia
Hawaii
Washington
West Virginia
Temperature
31
39
28
35
13
2
38
23
21
8
7
26
11
34
15
21
12
5
9
33
19
16
40
25
22
20
17
36
18
30
31
37
32
27
10
26
4
24
29
14
3
17
1
6
20.5
20.5
20.5
20.5
20.5
20.5
20.5
Conductivity
5
36
20
24
36
3
16
32
15
19
22
28
10
11
1
17
8
6
14
37
22
26
40
25
34
4
35
39
18
38
23
19
27
12
20.5
33
29
9
21
31
2
13
7
30
20.5
20.5
20.5
20.5
20.5
20.5
20.5
Dissolved
Oxygen
17
12
22
6
14
4
25
7
5
3
8
13
2
22
16
11
6
1
9
26
19
23
29
2
2
18
24
31
29
30
28
27
21
32
33
29
3
3
17
15
18
20
10
17
17
17
17
17
17
17
17
pH
13
12
4
2
14
10
15
5
7
7
6
6
15
13
16
22
13
11
3
17
6
9
17
4
4
18
19
8
14
16
5
14
2
20
5
14
6
21
10
8
11
16
1
10
11.5
11.5
11.5
11.5
11.5
11.5
11.5
32
-------
State
Alabama
Arizona
Arkansas
California
Colorado
Connecticut
Florida
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Puerto Rico
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Pennsy Ivani a
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Wisconsin
Wyoming
Alaska
Delaware
District of Columbia
Georgia
Hawaii
Washington
West Virginia
Temper-
ature
31
39
28
35
13
2
38
23
21
8
7
26
11
34
15
21
12
5
9
33
19
16
40
25
22
20
17
36
18
30
31
37
32
27
10
26
4
24
29
14
3
17
1
6
20.5
20.5
20.5
20.5
20.5
20.5
20.5
Conduc-
tivity
5
36
20
24
36
3
16
32
15
19
22
28
10
11
1
17
8
6
14
37
22
26
40
25
34
4
35
39
18
38
23
19
27
12
20.5
33
29
9
21
31
2
13
7
30
20.5
20.5
20.5
20.5
20.5
20.5
20.5
Dissolved
Oxygen
20
14
25
7
17
5
30
8
6
3
9
16
2
26
19
13
7
1
11
31
23
28
35
2
2
22
29
37
35
36
34
33
25
39
40
35
3
3
20
18
22
24
12
20
21
21
21
21
21
21
21
PS
23
21
6
3
25
17
26
8
12
12
10
10
26
23
28
39
23
19
5
30
10
16
30
6
6
32
34
14
25
28
8
25
3
36
8
25
10
37
17
14
19
28
1
17
21
21
21
21
21
21
21
Net
Ranking
Factor
79
110
79
69
91
27
110
71
54
42
48
80
49
94
63
90
50
31
39
131
74
86
145
58
64
78
115
126
96
132
96
114
87
114
78.5
119
46
73
87
77
46
82
21
73
83
83
83
83
83
83
83
Composite
Net
Rank
21
31
21
14
28
2
31
15
10
5
7
22
8
29
12
27
9
3
4
36
17
25
38
11
13
19
33
35
30
37
30
32
26
32
20
34
6
16
26
18
6
23
1
16
24
24
24
24
24
24
24
33
-------
(58.62 - .96VlfN
Where V = Industrial effluent volume per state N
1,N
V = Population of state N
2,N
V = Value added by manufacture in state N
3,N
V_ = Annual runoff for state N
4,N
V,. = Water area in state N
5,N
V = Population density of state N
6 ,N
V = Industrial water use in state N
7,N
W = User weighted policy influence 0 £ W <^ 10
V = Most recent policy priority accorded to state N
O f W
P = Priority for each state
N
V8 N represents the most recent policy priority accorded to state N.
Table VII refers to this initial policy priority as "rank order." It
is anticipated that the model will be exercised as policy changes occur
within EPA, thus equation (1) is concerned with the budget allocation
percentages (priorities) from the previous period. Further, equation
(1) has a "built-in" user-varied damper (W) on periodic policy changes
which allows the user to specify the magnitude of importance given to Vg/N.
Flow Chart 1 represents the functional flow of operations in the location
module. Files V, N through V~ N should be stored on magnetic disk. Al-
though random access capabilities are not necessary for the operation of
this module under Mode I , it may prove valuable under Mode II and is
clearly advantageous during update and file maintenance procedures .
These procedures are primarily concerned with updating the information
on file to insure the highest possible degree of file integrity. VQ N
represents the most recent EPA policy decisions. Upon acceptance of 'model
results by EPA, the system automatically updates this table. Operation-
ally, if access to this table is for update, control goes to "E"; if it
is to fetch, control is returned to the Linkage Coordinator (LC) . Of
course, the module may be requested either by Mode I and II Module Con-
trol Program, Flow Chart 2, or manually from console keyboard. The
module is initiated through the Linkage Coordinator. The LC retrieves
V1,N through V7/N for state N. % is calculated for that state, and if
it is not the last state, control is returned to LC. After the last state
is processed through LN a user weight is requested from the console or
batch input device. This weight determines the amount of influence ac-
corded to past policy. If a weight is not received, a value of 1.0 will
34
-------
w
Ul
/ Module
/ Request
*-/ from
Main or
er
Flow Chart 1. Location Module
-------
Terminal
Keypunch
U)
cn
Flow Chart 2. Mode I and II Module Controls Program
-------
be assigned to W. RJJ and PN are then determined and output to the con-
sole for review. If array PN is approved from the console, control is
transferred to "A"; if not, control goes to "LC" which queries the opera-
tor to ascertain why array PN was not approved and where to direct con-
trol. This process is continued until array PN is approved. In both
Mode I and Mode II operation, array PN is stored in working file Fl and
control is transferred to the Effluent Module.
Module II. Effluent Constituent
The first step in developing the Effluent Constituent Module was to deter-
mine the relative importance attributed to each effluent constituent in
terms of the criterion variable. Figure 6, page 20, depicts the result of
this analysis where, for example, BOD has an importance of 3.40 times that
of color and 532 times that of detergents. Data were collected on Data
Form 4100 (Appendix D) and from sources indicated in Table A-l and submitted
to associative statistical analysis for the purpose of identifying possible
predictor variables. The most highly correlated predictor variables were:
/ Effluent Volume, dimensionless, a = 11.18, 3 = .2029
/ State Standards, number of times mentioned, a = 1.281, 3 = .2726
/ Economic Effects, dimensionless, a = 1.723, 3 = -.1001
/ EPA Regional Standards, dimensionless, a = 13.99, 3 = .1691
/ Public Notice, dimensionless, a = .3360, 3 = -.1723
/ Low Concentration Limit, dimensionless, a = 4.197, 3 = -.2868
/ High Concentration Limit, dimensionless, a = 3.867, 3 - .3213
Six of the seven interesting predictor variables are ranks and dimension-
less; therefore, to preserve consistency effluent volume was transformed
into a dimensionless rank. (See Data Form 4100 [Appendix D]) . Consider-
able attention was given to the quantification of public notice inasmuch
as it was generally agreed that this variable has much significance as a
predictor of priority assignment and fund allocation. It was determined,
however, that a separate subsystem in the model is required for inclusion,
due to the variety of sources of appropriate raw data, variable confidence
of such, and alternative mechanisms for inputting same to the data base.
For the purposes of model definition and initial test, source data were
entered as dichotomies and held constant. Of course, these 3 coefficients
are interpreted in the same manner as those described on page 29. In brief,
they indicate that EPA has put higher priorities on treating effluent con-
stituents with:
/ Higher effluent volumes
/ Higher state standards
/ Lower economic effects
/ Higher EPA regional standards
/ Lower public notice
/ Lower low concentration limits
/ Higher high concentration limits
37
-------
Table XI describes each effluent constituent in terms of selected predictor
variables. In addition to these seven variables, the relative cost per
pound removal of each constituent was determined. To incorporate cost con-
siderations into the model, it seemed most logical to assume that the high-
est priority for research effort would be assigned to potential cost reduc-
tions in removing constituents of lowest value, i.e., to those constituents
which could support the lowest costs even with complete recovery as salable
products. This assumption is supported further by considering the costs
involved with no recovery; thus, a high-cost product can more easily bear
waste treatment costs as an operating expense. The relative costs of re-
moval for various wastewater constituents presented a difficult case for
inclusion in the priority model. Removal cost data are sparse at best,
and are not available for most specific wastewater constituents. Bearing
in mind that the model is concerned with priorities, i.e., relative instead
of absolute measures, and that significant pollution abatement measures in
industry include conservation of materials and by-product recovery, the
values of the wastewater constituents would seem to be valid measures of
the treatment or particularly the conservation or recovery costs which de-
fine the upper economic cost limits. This is to say that if the total cost
of recovering a material is equal to or less than its value as a product
at the plant, such recovery is a no-cost operation and defines the desirable
maximum cost.
It would, of course, be desirable if specific costs of present treatment
methods for each specific constituent were available. This would, at least
in theory, be preferable to the more hypothetical "target" costs as defined
by values of materials if recoverable. As has been pointed out previously,
the inputs to the present study have been formulated on the basis of read-
ily available data and no suitable data are presently available other than
these "target" costs.
The costs of present treatment methods might be made available as a result
of EPA's Industrial Waste Inventory, Effluent Criteria Development Projects,
and Discharge Permit Program. Unless the development of treatment cost
data by specific constituents is built into these programs, such data would
be very difficult to extract and beyond what should be considered as rea-
sonable for incorporation as a procedure within the priority-determining
model. Industrial treatment costs are typically given in terms of effluent
volumes and rarely in terms of constituents removed. Municipal treatment
costs can be expressed in pounds of BOD or solids removed only because such
effluents are reasonably uniform. The costs of suspended solids removal
from steel mill effluents, by contrast, are hardly comparable to, say,
removal from the effluents of the glass industry. Within the steel indus-
try the costs of removing suspended solids from basic oxygen furnace gas-
washer water and from cold rolling mill soluble oil emulsions are greatly
different; the values of the recovered solids per unit weight are, however,
very similar.
38
-------
Table XI
Effluent Constituent Matrix
Constituent
BOD
Color
Suspended Solids
Organics
Co Li forms
Oil and Grease
CCE
Acidity
Sulfate
Thiosulfate
Phenols
Cyanide
Thiocyanate
Sulfide
Odor
Sulfur
Sulfite
COD
Chromium
Phosphate
Detergents
Phosphorous
Lead
Mercury
Arsenic
Silver
Heavy Metals
Iron
Aluminum
Zinc
Copper
Chlorides
Total Solids
Conductance
Policy
Funding
31.95
9.31
7.43
6.80
6.80
4.38
4.38
3.86
3.86
3.86
3.03
2.62
2.62
2.57
2.57
2.57
2.25
2.21
2.11
2.10
2.10
2.10
2.00
2.00
2.00
2.00
2.00
1.99
1.99
1.99
1.73
1.42
1.42
1.42
Wastewater
Volumes
38.0
26.0
38.0
18.5
18.5
37.0
18.5
30.0
24.0
7.0
29.0
25.0
13.0
32.0
17.0
16.0
3.0
28.0
11.0
18.0
20.0
6.0
8.0
1.0
18.5
18.5
18.5
27.0
5.0
10.0
10.0
36.0
33.0
33.0
State
Standards
4
4
3
1
4
4
1
1
2
1
2
2
1
1
4
1
1
4
2
2
1
1
2
1
2
2
1
1
1
1
2
2
3
1
Economic
Effects
6
5
9
5
2
6
5
4
8
5
5
7
5
5
5
5
5
6
7
5
5
5
7
7
7
5
7
1
5
5
7
3
3
3
EPA Region
Appraisals
53.0
19.0
33.0
36.0
57.0
40.0
13.0
32.0
50.0
28.'5
44.0
22.0
28.5
31.0
15.0
25.0
23.0
43.0
51.0
55.0
7.0
28.5
27.0
51.0
11.0
28.5
51.0
14.0
28.5
26.0
24.0
38.0
43.0
18.0
Public
Notice
1
1
1
1
1
2
1
2
1
1
1
1
1
1
1
1
1
1
1
1
2
1
2
2
1
1
1
1
1
1
1
1
1
1
Concentration
Low
8.0
6.0
3.0
11.0
14.0
6.0
11.0
3.0
2.0
5.0
14.0
14.0
8.0
14.0
14.0
8.0
12.. 0
8.0
12.0
8.0
14.0
12.0
12.0
13.0
12.5
12.5
14.0
12.0
12.0
7.0
11.0
2.0
1.0
1.0
Limits
High
10.0
5.0
5.0
13.0
17.0
7.0
13.0
4.0
1.0
5.0
17.0
14.0
8.0
14.0
8.5
*8.0
13.0
10.0
9.0
8.0
13.0
13.0
14.0
16.0
13.0
15.5
13.0
13.0
13.0
9.0
13.0
2.0
1.0
1.0
u>
10
-------
Table XI (Continued)
Constituent
Corrosiveness
Calcium
Silica
Magnesium
Hardness
Temperature
Hydrocarbon
Turbidity
Ammonia
Total Organic C
Alkalinity
Organic N
Total N
pH
Fluorides
Sodium
Potassium
Manganese
Toxicity
Cadium
Nitrates
Nitrites
Radioactivity
Barium
Selenium
Boron
Settleable Solids
Mercaptans
Polysacchrid
Tannin
Lignins
Pesticides
Nickels
Policy
Funding
1.42
1.27
1.27
1.27
1.27
1.20
1.00
0.99
0.85
0.85
0.53
0.40
0.40
0.37
0.32
0.22
0.22
0.22
0.22
0.22
0.22
0.22
0.22
0.22
0.22
0.22
0.21
0.20
0.20
0.18
0.18
0.15
0.09
Wastewater
Volumes
39.0
21.0
9.0
2.0
18.5
31.0
18.5
14.0
34.0
18.5
35.0
23.0
19.0
39.0
28.0
22.0
21.0
15.0
12.0
8.0
7.0
7.0
18.5
18.5
18.5
18.5
38.0
21.0
16.0
4.0
3.0
18.5
8.0
State
Standards
1
1
1
1
1
4
1
4
1
1
1
1
1
4
2
1
1
1
4
2
2
1
3
2
2
1
4
1
1
1
1
1
1
Economic
Effects
2
8
5
5
8
4
5
9
5
5
4
5
5
4
5
5
5
1
7
5
5
5
7
7
7
7
9
5
6
5
5
7
5
EPA Region
Appraisals
47.0
28.5
28.5
28.5
28.5
56.0
42.0
35.0
1.0
45.0
29.0
36.0
54.0
47.0
4.0
6.0
6.0
11.0
51.0
28.5
49.0
37.0
46.0
28.5
51.0
5.0
39.0
31.0
28.5
36.0
36.0
52.0
28.5
Public
Notice
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
2
1
1
1
1
1
1
1
1
2
1
Concentration
Low
14.0
2.0
4.0
5.0
1.0
14.0
11.0
5.0
14.0
11.0
2.0
8.0
7.0
14.0
10.0
5.0
5.0
12.0
14.0
12.0
9.0
12.0
14.0
8.0
14.0
10.0
3.0
14.0
8.0
8.0
8.0
14.0
12.0
Limits
High
8.5
3.0
6.0
4.0
1.0
8.5
13.0
5.0
13.0
13.0
2.0
8.0
8.0
8.5
11.0
6.0
6.0
13.0
13. C
8.0
8.0
12.0
17.0
9.0
13.0
1.0
5.0
15.0
9.0
8.0
8.0
17.0
13.0
-------
If possible, present costs of the treatment of specific constituents should
be acquired from the EPA ongoing programs mentioned above. Until other
data are available, the "target" costs as defined herein should be used
since they represent at least consistent and available measures of a much-
needed criterion in the model.
In Table XII, the wastewater constituents are listed with the costs per
pound (1959-60 cost index = 1.0) of the specific constituents or of in-
dustrial chemicals which are likely sources of the specific constituents.
The wastewater constituents are then ranked on the basis of decreasing
values with the rank of one equal to the highest cost. The ranks of
Table XII were entered into the model as measures of total cost consid-
erations (VS,M) with the highest numerical rank equivalent to the highest
priority for research and development where treatment costs are concerned.
In the same manner as we did for equation (1) , the data in Table XI were
subjected to fixed-X regression analysis which produced the basic struc-
ture for equation (2) . However, in this instance the user may override
the system by replacing the most recent regression coefficient with his
own relative weights (Q^) .
Max _
2,M
P.. = -1.59 + 1.15V, , /max V, , + 2.50V,,
M 1,M/ 1,M 2,M/
+ .75V M/max V + 1.10V. M/max V + 1.00V_ ../max V
3,M/ 3,M 4,M/ 4,M 5,M/ 5,
M
V/~ »» "^ 1-90V_ /max V_ + l.wvyv_ /m^^ v
6,M 7,M/ 7,M 8,M/ 8,M
2.00V9/M/max Vg/
M (2)
For M = 1 -* Last Effluent Constituent
where
P = Priority ranking for constituent M
M
V = Effluent volume for constituent M
1,M
V = State standard for constituent M
2 ,M
V = Economic effects for constituent M
3,M
V = EPA regional standard for constituent M
4,M
V = Public notice of constituent M
5,M
V = Low concentration limit of constituent M
6,M
V = High concentration limit of constituent M
7,M
V = Relative cost of removal of constituent M
8,M
41
-------
Table XII. Values of Wastewater Constituents and Related Chemicals
Constituent
PH
BOD
Suspended Solids
Oil and Grease
Chlorides
Alkalinity
Ammonia
Total Solids
Sulfides
Temperature
Acidity
Phenols
COD
Fluorides
Iron
Color
Cyanide
Sulfate
Organic N
Sodium
Me reap tons
Potassium
Calcium
Detergents
Total N
Phosphate
Odor
Polysacchrides
Sulfur
Manganese
Turbidity
Thiocyanate
Toxicity
Chromium
Copper
Zinc
Silica
Lead
Nickel
Cadmium
Nitrites
Nitrates
Thiosulfate
Phosphorous
Aluminum
Tannin
Sulfite
Equivalent Chemical
Calcium Carbonate
Molasses
Bentonite
#6 Fuel Oil
Rock Salt
Lime
Ammonia
Rock Salt
Sodium Sulfide
Steam
Sulfuric Acid
Phenol
Miscellaneous Chemicals (6)
Sodium Fluoride
Iron
Benzenoid Dyes (3)
Sodium Cyanide
Sodium Sulfate
Nitrophenol
Rock Salt
Ethyl Mercapton
Potassium Chloride
Calcium Chloride
Surface Active Agents (2)
Average of all N Compounds (8)
Calcium Phosphate
Phenol
Saccharin
Sulfur
Manganese
Bentonite
Sodium Thiocyanate
Average of Toxic Compounds
Chromium
Copper
Zinc
Silica
Lead
Nickel
Cadmium
Sodium Nitrite
Sodium Nitrate
Sodium Thiosulfate
Sodium Phosphate
Aluminum
Lignin
Sodium Sulfite
$/lb.
0.06
0.01
0.007
0.01
0.01
0.01
0.0425
0.01
0.0525
0.0003
0.01
0.14
0.13
0.11
0.10
1.48
0.17
0.0235
0.50
0.01
0.70
0.0135
0.0135
0.18
0.032
0.0375
0.14
1.60
0.0133
0.30
0.007
0.71
1.388
1.18
0.30
0.11
0.115
0.14
0.60
1.70
0.0775
0.025
0.04
0.085
0.20
0.05
0.0725
Cost Rank
30
43
44
43
43
43
33
43
31
46
43
22
23
25
26
7
21
40
15
43
12
41
41
20
38
35
22
6
42
17
44
11
8
9
17
25
24
22
14
5
28
39
34
27
19
32
29
42
-------
Table XII (continued)
Constituent
Lignins
Magnesium
Mercury
Radioactivity
Arsenic
Borium
Selenium
Silver
Conductance
CCE
Pesticides
Boron
Hardness
Organics
Corrosiveness
Settleable Solids
Coli forms
Heavy Metals
Hydrocarbons
Total Organic C
Equivalent Chemical
Lignin
Magneisum
Mercury
Highest Cost
Arsenic
Borium Bromide
Selenium
Silver Chloride
Rock Salt
Cyclic Intermediates (4)
Pesticides & Organic Ag.
Chem. (1)
Borax
Calcium Carbonate
Cyclic Intermediates (4)
Sulfuric Acid
Iron Ore
Lowest Cost
Average of all
Aliphatic Hydrocarbons (5)
Crude Products from
Petroleum and Gas (7)
$/lb.
0.05
0.27
3.00
-
0.60
0.49
5.00
12.48
0.01
0.11
0.62
0.035
0.06
0.11
0.01
0.0045
-
0.83
0.034
0.035
Cost Rank
32
18
4
1
14
16
3
2
43
25
13
36
30
25
43
45
47
10
37
36
(1) Chemical Statistics Handbook, 1966, p. 27 (1964 data)
(2) Ibid, p. 30 (1964 data)
(3) Ibid, p. 50 (1964 data)
(4) Ibid, p. 71 (1964 data)
(5) Ibid, p. 14 (1964 data)
(6) Ibid, p. 78 (1964 data)
(7) Ibid, p. 90 (1959 data)
(8) Ibid, p.285 (1959 data)
43
-------
V = EPA policy regarding constituent M, or fraction of EPA
' industrial budget spent on effluent constituent M during
last fiscal period
O = Relative weight associated with V
"N N,M
max V, ,, = Maximum value of VXT ,.
N,M N,M
The constants QN are set equal to the correlation coefficient or user
weightsiwhich depict most recent policy. Provision has been made for
the user to enter new QJJ values if he should not agree with those most
recently used. This process is illustrated below.
Flow Chart 3 represents the logical and mathematical operations of this
Effluent Constituent Module. File integrity is insured through the file
update and maintenance program which includes the dynamic Public Notice
subsystem. Again, the module may be accessed through a terminal by an
operator or through Modes I and II Module Control Program, Flow Chart 2.
Upon initiation, control goes to the LC. LC requests user weights QJJ
and determines the maximum values of V^ M -* VgfM« VN M are then re-
trieved and WN determined and weighted by QN, where Wjj is simply
VN M/"334 VN,M- Control returns to LC and the process continues until
the last constituent is processed. After the last constituent is pro-
cessed RN, N = 1,9 are summed by high concentration only, low concen-
tration only, and both. These priorities are then displayed on the
console for approval and directions as to which of the three lists to
use. Upon receiving approval and direction, control is transferred to
"B" in the Module Control Program. If approval is not received, the
user is queried relative to reason for rejection and where to restart.
Upon approval, PM are stored in file F2 and the Industry Module is called.
Module III. Industry
This module was developed in a manner similar to those previously de-
scribed. The complex variable was ranked by describing each PPB cate-
gory as a function of the prior Federal funding in that category. See
Figure 4, page 18. This ranking reflects the relative importance attri-
buted to each category. For example, from Figure 4 it is clear that
1206 (Food and Kindred Products) was attributed 1.36 times the importance
of 1204 (Paper and Allied Products) and 103 times the importance of 1208
(Stone, Clay and Glass Products).
The predictor variables, from those included on Form 2100, Appendix E,
indicated to be significant from preliminary analysis were:
/ Industrial Effluent Volume, in bgy, a = 1512, 3 = -.4233
/ Water Use, in bgy, a = 3140, 3 = .4105
/ Value Added by Manufacture, in million dollars, a = 10150,
3 = -.3620
/ Employment, in thousands, a = 725.3, 3 = .0745
44
-------
Ul
Input
Documents
Keying
p
No
* . W B,
Most
Recent
User
Weights
Table
S - 1 f 9
Output
by High,,
Low, and
/Sum ConcnsJ
/ \/
[ Y^L
Rl = "l ' 21
R2 - «2 Q2
R3 ' W3 ' 83
*9 - «9 ' Q9
N
Flow Chart 3. Effluent Constituent Module
-------
/ Number of States with Plants, a = 9.921, 3 = .3584
/ Total Number of Plants, O = 8993, 8 = .2357
/ Number of Plants Using more than 20 mgy, a = 684.2, $ = .0440
See Table XIII for the matrix depicting these quantities on a national
basis. Each of the 126 projects in the data base was ranked according
to Table XIII and then subjected to analysis by the Policy Capturing
Model which produced the basic structure of the following equation:
P = (-1230. - .43V. _ ' W. + .21V W - .12V W
I Ifl 1 * / -l * -3/-1- J
+ .35V,, * W. + 101.96V,. W + -07V W
4,1 4 5,1 5 6,1 6
+ .20V., W )/103 (3)
/ ,1 /
where
P = Priority associated with the Ith industry
V = Industrial effluent volume associated with the Ith industry
I/I
V = Water use associated with the Ith industry
2,1
V = Value added by manufacture to the Ith industry
3,1
V. = Employment for the Ith industry
4,1
V = Number of states with plants for the Ith industry
5,1
V= Total plants in the Ith industry
o,l
V = Plants using > 20 mgy in Ith industry
' r I
W = User weight for each variable V K = 1,7
K K, I
Interviews with potential users suggested that the system should allow
the user to dampen or increase the influence of any of the variables in
the basic equation by entering user weights WK. Thus, the system was
altered to allow such entries.
The nature of Pj allowed measurement of prediction reliability by two
independent means. The regression analysis which produced the above
equation indicated a coefficient of determination (R2) equal to 0.84,
thus accounting for 84% of the variance in Pj. Availability of the
"Needs Statements" for PPB 1200 categories additionally permitted com-
parison of priorities indicated by past funding with those indicated by
the collective judgments of EPA personnel as to needed future research.
PPB 1200 categories were ranked on the basis of the number of "needs"
reported per industry and compared with those ranks indicated by the
module equation. The degree of correlation was then determined by
Spearman's rs as follows:
46
-------
Table XIII
Industry Matrix
PPB
1206
1204
1202
1201
1209
1212
1205
1207
1210
1208
1211
SIC
20
26
28
33,34
22
39
29
35,36,37
24
32
30
Rank
35.00
25.66
13.24
9.08
8.20
3.13
2.82
1.67
.80
.34
.05
Billion
GalA*
Effluent
Volume
690
1900
3700
4300
140
12
1300
481
123
218
160
Billion
Gal/Yr
Water
Use
1280
6026
7577
6901
311
22
6161
1216
217
389
336
Million
$
Value
Added
10073
3856
12590
14707
2649
529
3066
34976
574
3180
2667
Thousands
Em-
ployment
626
249
480
1072
353
43
119
2537
69
218
210
No. of
States
with
Plants
37
25
30
24
17
4
15
24
6
17
13
Total
Plants
21555
4393
7511
21658
5160
6865
1390
28836
10584
8415
3312
Plants
Using
20 mgy
2405
634
1062
1270
603
76
268
1287
186
555
280
-------
Needs
1206
1204
1202
1201
1209
1212
1205
1207
1210
1208
Rank Module Rank
= 9
*7
= 10
= 6
= 4
= 4
= 8
= 1
= 5
= 2
10
9
8
7
6
5
4
3
2
1
2
a
i
4
4
1
4
1
16
4
9
1
N(N -
= .73
The correlation coefficient is significant at the .01 level which in-
dicates that the module rank is reliably related to the needs rank.
Furthermore, equation (3) suggests that previously EPA has given high
priorities to industries with:
/ Low industrial effluent volumes
/ High water use
/ Low value added by manufacture
T/ High employment
/ High number of states with plants
/ High number of total plants
/ High number of large water-using plants
The Industry Module is logically and mathematically straightforward,
as can be seen from Flow Chart 4. Similar to the other modules, it has
a file update and maintenance program to insure file integrity; however,
this module requires a much less sophisticated linkage coordinator. Es-
sentially, a request enters either from the Module Control Program or a
user console. User weights WK are requested; if none are given, unity
is assumed and WK = 1.0 for all K. LC retrieves VK,I» weights it, and
determines PI for each industry. The priorities are displayed on the
console for user approval. Further, they are updated as needed to re-
flect most recent policy decisions. If approval is received, control
goes to "C" where Pz are stored in file F3. If approval is not received,
the user is queried for cause of rejection and where to continue. Upon
completion, control is transferred to the Statistical Module.
Although PPB number was chosen as the basic unit of classification for
this module, two- or four-digit SIC numbers will also suffice. Table XIV
cross references PPBs, SICs (two and four digits) and information sources.
48
-------
input
Documents
^
1 ^^
II ,
/ 1
/
Keying
;
Pile Update
Maintenance
Program
i
JL JL j
-1.23 - .00043 V. _ + .00021 V. _ - .00012 V
('Interactive^
Console ]
.00035 V. + .10196 Vc + .00007 V, + .00020 V
4,1 5,1 6,1 /,J-
I = 1 * Last Industry
Flow Chart 4. Industry Module III
-------
Table XIV
Industrial Classification Cross Reference
CITY AND COUNTY DATA BOOK
INDUSTRIAL GROUPS SUMMARY DATA
.
ndustria Group
Primary and Intermediate
Metal Products
and Plastics Products
Food and Tobacco Products
Paper and Printing
Electrical and Non-
Electrical Machinery
Transportation and Ordnance
Stone, Clay and
Glass Products
Textile/ Apparel and
Leather Products
Lumber, Wood Products
and Furniture
SIC
33 ,34
28,29,30
20,21
26,27
35,36
37, 19
32
22,23,31
24,25
EPA "COST OF CLEAN WATER--
INDUSTRIAL GROUPS SUMMARY DATA
.
naustrxaJ. Group
Primary Metals
/ Blast Furnaces/Steel Mills
*/ All Other
Chemicals and Allied Products
/ Organic Chemicals Ind.
/ Inorganic Chemicals Ind.
Petroleum and Coal Products
Rubber and Plastics
Food and Kindred Products
/ Meat Products
/ Dairy Products
/ Canned and Frozen Food
/ Sugar Refining
/ All Other
Paper and Allied Products
Machinery Except Electrical
Electrical Machinery
Transportation Equipment
Textile Mill Products
SIC
33
3312
28
2815 ,-18,
-13, -79,
-71
2812, -13,
-16, -19,
-51, -71,
-79, -92
29
30
20
201
202
203
206
26
35
36
37
22
EPAlNDUSTRIAL WASTE PROFILES
(and/or Industrial Waste Guides)
Publication itle
Blast Furnaces/Steel Mills
Organic Chemicals Industry
Inorganic Chemicals Industry
Plastics Materials and Resins
Petroleum Refining
Meat Products
Dairies
Canned/Frozen Fruits
and Vegetables
Paper Mills Except Building
Thermal Pollution (Industrial
Waste Guide)
Motor Vehicles and Parts
Textile Mill Products
Leather Tanninq and Finishing
sic
3312
2815, -18,
-13, -79,
-71
2812, -13,
-16, -19
-51, -71
-79, -92
2821
2911
201
202
2033
2037
2621
3717
22
3111
EPA BSD PROGRAM
STRUCTURE DESIGNATIONS
PPBS
1201
1202
1205
1211
1206
1204
1203
1207
1208
1209
1210
ec no ogy roup
Metal and Metal Products
Chemicals and Allied Products
Petroleum and Coal Products
Rubber and Plastics
Pood and -Kindred Products
Paper and Allied Products
Power Production
Machinery and Transportation
Equipment
Stone, Clay and
Glass Products
Textile Mill Products
Lumber and Wood Products
SIC
33,34
28
29
30
20
26
35 36 37
32
22
24
en
o
-------
Module IV. Statistical
The statistical module retrieves priorities associated with type of
project, project organization, and project implementation. See Tables
II through IV. As previously stated, these variables are dimensionable
within themselves and are largely dependent only upon current EPA policy.
Because of their arbitrary nature and the fact that they are quite vola-
tile, the module makes no decisions dependent upon past policy with re-
spect to these variables; instead, that policy is described to the user,
thereby supplying him the information he needs to support his judgment.
In addition, this module tallies Federal and industrial RD&D expenditures
by region and industry using Form No. 8100, Appendix F. The four up-
date and maintenance programs are most important subsystems in this
module. UD]_, UD2, and UD3 support routine updating to their respective
tables with change in EPA policy while UD4 is charged with the respon-
sibility of accounting for all "actions" taken by EPA during the period
of interest. Thus, when a project is funded, the amount of Federal funds,
the amount of industrial funds, and the "action description" are stored
on disk file. When the "map" is next generated, these data are used as
a baseline for action generation. Flow Chart 5 illustrates the sequence
of retrieval operations.
The variables, in order of retrieval, are:
FF = Federal funding per project P
IF = Industrial funding per project P
IND = Industry for action A
£\
LOG, = Location of action A
A
EC = Effluent constituents associated with action A
A,I
PB = Percent of budget spent on A
A
TP_ = Type of project ranking for J types
J
PI = Project implementation for I types
PO = Project organization for K types
K
After retreival, control is transferred to "D" in the Module Control
Program where the variables are stored on file F4.
51
-------
User
Terminal
T
j Module 7
/ Request /
~/ from Main /
/ or Used /
U1
10
JNo
Read
Project
Organ.
P°K
K = 1 -* 7
Project
Organ .
Table
poi
P02
POK
I No
Read Project
Implementation
PI
I
I = 1 -* LP
*
*
Project
Imp . Table
1
PI
2
«I
Flow Chart 5. Statistical Module
-------
MODEL OPERATION
There are two modes of model operation. Mode I "maps" national prior-
ities for industrial effluent treatment and control research, development,
and demonstration. It supplies EPA with a number of specific "actions"
that should be taken, as well as an appraisal of the relative importance
of each action. (See Figure 1.) The MAP may be used as a guide to proj-
ect development and funding. Mode II reverses the role of the model.
In this mode, the model reviews and assigns priorities to project descrip-
tions such as "Statements of Need," provides suggestions as to how one
might improve the priority of specific projects, and suggests how similar
or complementary "Needs" may be combined. Although the current effort
was concerned basically with Mode I, the ultimate utility of Mode II jus-
tified considerable attention to its preliminary design. The success or
failure of TIMPS is highly dependent upon a sound means for generating
an RD&D priority list. The two subsections that follow discuss each mode
of operation separately.
Flow Chart 6 indicates the ties between the four modules and the Module
Control Program, Mode I, and Mode II. Note that the only logical dif-
ference between Mode I and Mode II is that control goes to MAP in the
former and to Project Organization Algorithm (PDA) in the latter.
Mode I. MAP
The Multiple Allocation Program (MAP) is based on a multi-dimensional
resource allocation routine. Basically, the problem is to allocate a
fixed budget "K" over three variably dimensional parameters. The ques-
tion is how resources can best be allocated with respect to project ob-
jectives in order to maximize the return on investment considering all
three parameters at any point in time. In order to respond to this ques-
tion, the model must combine the individual priorities on a_ priori logical
ground, not merely according to mathematical optimization techniques.
The "MAP" is generated at three levels of specificity:
Level I - general summary of the priority associated with each PPB
category, the percent of total budget and of PPB budget allocated to
each state associated with that PPB category, and the percent of total
budget and PPB budget allocated to each primary effluent constituent.
Figure 1, pages 3 through 5, illustrates this level of "MAP." The
values used in Figure 1 are based on past policy and should only be
used as an example of possible output from the model. Since the model
requires rather complex manipulations, adequate test results and ex-
perience from applications must await the decision to complete and
fully implement the modeling system at least for the 1200 PPBS cate-
gory.
53
-------
(Jl
Flow Chart 6. Composite Module
-------
Level II - describes the priorities associated with all "Actions" for
each PPB category. Due to the time required to calculate these manu-
ally, only the "Action" Matrix for PPB 1206 has been determined for
Figure 2. Each entry in the matrix describes the percent of total
budget to be allocated to the location/effluent constituent combina-
tion. For example, BOD in New York receives 1.07% of budget.
Level III - provides a detailed description of each "subaction" pre-
scribed by the model. Subactions are described by industry, location,
effluent constituent, priority, and percent of budget. Figure 3 shows
the first 14 subactions prescribed by the model.
Of course, depending upon the size of the RD&D budget, the subactions can
be combined into larger projects, or broken into small projects; in short,
270 subactions do not necessarily dictate 270 projects.
Consider the example in Figure 1. Inputs to MAP from the four modules
are calculated as described below and illustrated in Flow Chart 7.
Priorities associated with industry ranked from high to low
P! 1 = 1 + LAST INDUSTRY
Priorities associated with location ranked from high to low
PN N = 1 -» LAST STATE
Priorities associated with effluent constituent ranked from
high to low PM M = 1 -> LAST CONSTITUENT
Past actions during that period
PBA, INDA>j, LOCAfN, ECAfI A = 1 -> LAST ACTION
In the interest of clarity, past actions will not be considered in this
example. However, in general, the following three equations illustrate
how past actions are discounted from the system:
Pj = PI;L - INDA,i (4)
where
Pj = the priority associated with industry I at time t = A
INDA j = the percent of budget spent on industry I at time t = A
Pj = industrial priority at t = A + 1 for industry I
PN " %! - LOCA,N (5)
where
PN = the priority associated with location N at time t = A
LOCA M = the percent of budget spent in location N at time t = A
PN = geographical priority at t = A + 1 for location N
55
-------
cn
PN] [PE,I(PM,! + l>] ^ [PE,I
-------
PM = pMl - ECA,M
where
pMi = ^e priority associated with effluent constituent M at time
t = A
ECA M = the percent of budget spent on effluent M at time t = A
PM - effluent constituent priority at t = A + 1 for constituent M
After Pj, PN, and PM have been adjusted by equations (4) , (5) , and (6) ,
MAP evaluates all rational combinations of industry, location, and efflu-
ent constituent in the following equation to determine subaction priori-
ties, Figure 1.
2 - LS -
PP " PI,N,M - PI PN (VI,S/VI,S) Z n PN (VI,S/VI,S)
N = 1
LE
PE,I (PM,I +" * PE,I (PM,I + " (7)
E X
where
P = Priority of project P
P = Priority of project in industry I, location N, and
' ' primary effluent constituent M
P = Priority from industrial module for industry I
P = Priority from location module for location N
N
V = Volume of effluent in state S from industry I in bgy
LS
VI,S = I _ , VI,SLS' mean VI,S
o J- /
P = Priority of effluent constituent E in industry I
E, I
LS = Last State
LE = Last Effluent Constituent
P = Priority P from effluent constituent module for industry I
M,I M
Matrix Pj N/M is then sorted from high to low, placed into array PP ,
and output to MAP, Level III, as illustrated in Figure 3.
57
-------
Equation (7) maximizes benefit/cost ratios by determining the combined
importance of a volume of an effluent constituent within an industry
within a state. The principle of optimization is simple. Returns to
scale are maximized by funding projects that deal with large volumes of
high priority effluent constituents in high priority industries in high
priority states, and further, the relative amount of funding will be
described by the relative magnitude of those priorities.
The system first makes logical comparisons of all combinations of indus-
try, location, and effluent constituents to determine which combinations
are rational. It then looks up VI>S, the volume of effluent under con-
sideration in state "s" from industry I in order to optimize returns to
scale. Notice that effluent volume is not being used in the same manner
as it was in calculating Pj and PN, the difference being that in equation
(7) effluent volume per state per industry is being used in determining
the relative weight Vj^s which reflects the combined importance of Pj
and PN. In Pj and PN we were using effluent volume per industry and per
state respectively. Notice that Pp will equal zero only if the combina-
tion of industry, state, and effluent constituent is irrational, other-
wise PP will only approximate zero for very low priority jobs. The system
determines the absolute value of such projects as a function of the budget
size and disregards projects whose value falls below a pre-set minimum.
Mode II. POA
The Project Organization Algorithm serves three major purposes:
1. Determines priorities based on structural project descriptions.
2. Illustrates how similar or complementary projects may be com-
bined under one set of objectives.
3. Determines how priorities on a given project may be increased,
e.g., a project may be enhanced, if done in a state other than
the one proposed, and the reason.
In general, POA is a special organization of the MAP algorithm; that is,
it is mathematically the same, but logically reorganized. Upon receiving
control from Module Control, PIf PN, and PM are adjusted for past funding
as in MAP; however, under POA, the system is provided values for industry
"I," location "N," and effluent constituents "M." Based on these values,
PI' PN' PM' VI,S' and PE i are retrieved. P is then determined as
a function of these values. Once the last pr6:ject description is eval-
uated, matrix PI,N,M is sorted in descending order into array Pp and
displayed. The system then compares all combinations of projects; where
like or complementary projects are found, it suggests they be combined
and reports the new priority. Based primarily on past funding, the sys-
tem maximizes each priority by substituting higher priority locations and
effluent constituents for, or in addition to, those proposed. It is anti-
cipated that model output to POA will be similar to Level III, MAP, Figure
3.
58
-------
DETERMINATION OF MODEL PRACTICABILI1Y
Demonstrable practicability was an overriding consideration in the design
and development of a suitable RD&D priority and fund allocation model.
In general, this meant that any model design chosen for development must
not only be technically sound and potentially responsive to management
user needs, but amenable to implementation within established EPA re-
sources and constraints. Specifically, the practicability of the model
design was scrutinized from two separate, though related, stand-points:
(1) Functional integrity and fidelity to user requirements, and (2) op-
erational feasibility and suitability implications. These areas are
discussed, in turn, below.
Functional Integrity and Fidelity
The functional design of the four modules which comprise the priority
and fund allocation model in its present developmental form has been
described in detail in preceding sections and needs not be repeated here.
Based on those descriptions, however, points bearing on the model's in-
tegral nature should be apparent.
1. With the exception of the Statistical Support Module, module
design was configured on a single, largely compatible frame-
work. Only those differences which were essential to account
for unique predictor variable substrata were retained. In fact,
the basic form of the policy capturing algorithm applies equally
well across the three primary modules.
2. Although, as noted above, this contract was intended only for
initial design, sufficient preliminary programming of module
subroutines was accomplished to enable the modules to be op-
erated in sequence under operator control. Without a deter-
mination of PPBS category scope or the desired machine config-
uration, further programming at this stage would, of course,
have been potentially wasteful and inefficient.
3. Each module has been provided with a comparable elemental main-
tenance and update subroutine sufficient for entry into the next
developmental step.
4. While not a required portion of the effort, a second mode of
model operation has been developed which permits "Statement
of Need" evaluation on an integral basis with priority deter-
minations.
59
-------
A major effort was made to assure that the model design and anticipated
implementation form would be responsive to critical user information
requirements. Accordingly, a number of important provisions serve this
end. Chief among these are:
1. User access to all variables and primary operations has been
retained in order that final judgment can be exercised by the
user at any and all levels. Thus, the user is able to interact
with the priority determination and fund allocation process at
all times, which ensures that the rationale for any given set
of model outcomes is entirely explicit.
2. A wide range of user requests can be dealt with by the model
design. Generally, such requests may be grouped in three very
broad classes:
/Assessment of prior policy effects.
/ Comparison of immediate alternative policy strategies to
support current allocation demands.
/ Comparisons among and between longer range contingency
plans.
3. The model design, owing to its modular form, permits a large
degree of manual execution. Limitations on this capability are
due primarily to the facility with which the model accepts a
large number of variables for consideration, not the complexity
of operations. This latter capability is enhanced by the pro-
visions for collapsing a given module (sloughing variables) at
the user's discretion.
4. Provisions were made to support a large set of alternatives for
establishing the procedures by which final priority and fund allo-
cations will be determined, since both ready-file access and itera-
tive refinement procedures are simple and well defined.
5. Depending upon a final determination of user requirements with
respect to precision and degree of confidence, the "Public Notice"
variable is accommodated within the model. Options for public
notice source data include at least the following:
/ Congressional Record.
/ Complaints (number and type) received at regional/state
regulatory and enforcement agencies.
/ Media coverage (frequency, time, space allocated).
Operational Feasibility and Suitability
The implications of the model design can be soundly inferred from present
design characteristics and preliminary tests of model behavior. Thus,
assuming a typical International Business Machine System 360 model 50
machine configuration, the following essential operating parameters can
be specified:
60
-------
Average Maximum Run Time
Disk lo (2311): approximately 75103 sees/run (assumes 75 milli-
second/access and that files 1-4 are either held in main frame
core [lower figure] or stored on disk [higher figure]).
Central Processor: approximately 85 sees/run.
IO: variable dependent upon device (e.g., printer or console),
device software and extent of output resultant from user request.
Core Requirements (Assumes segmented, modularized instruction set)
Average resident core required during execution: 400 words,
assuming data-base matrices (or arrays) associated with given
execution available in core.
Resident core storage: 2700 words.
Total core required: 3100 words.
Program
Approximately 2000 Fortran IV instructions, where an average 300
are in resident core at any given time.
Data Reduction Manpower
Input data reduction manpower for module file data base update is
estimated at .5 mandays/month for the 1200 PPBS category. This
minimal requirement owes to the reliance upon readily available
data, much of which are already collected and reduced to machine
manipulable form by EPA on a regular basis.
It can be concluded that the entire model places a trivial demand upon ma-
chine and personnel resources and is well within existing EPA constraints.
61
-------
REFERENCES
Bramer, H. C. Economics and water pollution abatement. Water and
Sewage Works, 1966.
Christal, R. E. JAN: A technique for analyzing group judgment.
Journal of Experimental Education, 1968, 36_(4) , 24-27.
Christal, R. E. Selecting a Harem - And other applications of the
policy-capturing model. Journal of Experimental Education, 1968,
36_(4) , 35-41.
Kneese, A. V. The economics of regional water quality management.
Baltimore: Johns Hopkins Press, 1964.
63
-------
APPENDIX A
DATA ELEMENT SOURCES
65
-------
Table A-l
Data Elements and Information Sources
CFl
Effluent Constituents
Associated industrial effluent volumes
Concentrations in receiving waters permitting all water uses
Concentrations in receiving water permitting all but the
"most sensitive" water uses
Frequency of mention in State Water Quality Standards
Economic effects on water uses in receiving waters
RPR Regional office appraisals of relative pollution severities
Degree of public notice
Target treatment costs as determined by maximum values of re-
covered materials
Industrial Groups
Size of industry
Geographical distribution of industry
Water use practices
Economic status
Wastewater treatment facilities
Production parameters
Wastewater constituents
State Dimensions
Industrial Wastewater volumes
Population
Land area
Value added by manufacture
Number of manufacturing establishments
Capital expenditures by manufacturers
Industrial water use
Land area in farms and value of farm products
Population using public water supplies
Annual precipitation
Recreational areas and annual use
Fishing licenses issued
Metropolitan area population
Electrical energy production
Annual water runoff and withdrawals
Scientific population
Funded Project Descriptions
Industry involved
Location
Project dates
Sources of funds
Wastewater constituents involved
Type of project
Project implementation
Objectives of project
General Statistics
Federal/industry R&D funds by region
Federal/industry RSD funds by industry group
Source of Information Code
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
]
X
X
-------
Table A-2
Sources of Information
1. United States Department of Commerce, Bureau of the Census. Statis-
tical abstract of the United States. (91st ed.) Washington, D. C.:
Author, 1970.
2. United States Department of Commerce, Bureau of the Census. 1967
census of manufactures. Washington, D. C.: Author, 1971. (Publica-
tion Number MC67(2), Series includes 80 reports):
19 Ordnance and accessories
20 Food and kindred products
20A Meat products
20B Dairy products
20C Canned, cured, and frozen foods
20D Grain mill products
20E Bakery products
20F Sugar and confectionery products
20G Beverages
2OH Fats and oils
201 Miscellaneous foods and kindred products
21 Tobacco products
22 Textile mill products
22A Weaving mills
22B Knitting mills
22C Dyeing and finishing textiles, except wool fabrics and knit goods
22D Floor covering mills
22E Yarn and thread mills
22F Miscellaneous textile goods
23 Apparel and other textile products
23A Men's and boy's apparel
23B Women's and misses' apparel
23C Women's and children's underwear; headwear; children outerwear
23D Miscellaneous apparel and accessories
23E Miscellaneous fabricated textile products
24 Lumber and wood products
24A Logging camps, sawmills, and planing mills
24B Millwork, plywood, and prefabricated structural wood products
24C Wooden containers and miscellaneous wood products
25 Furniture and fixtures
25A Household furniture
25B Office, public building, and miscellaneous furniture; office
and store fixtures
67
-------
26 Pulp, paper, and board mills
26A Pulp, paper, and board mills
26B Converted paper and paperboard products, except containers
and boxes
26C Paperboard containers and boxes
27 Printing and publishing
27A Newspapers, periodicals, books, and miscellaneous publishing
27B Commercial printing and manifold business forms
27C Greeting cards, bookbinding, and printing trade services
28 Chemicals and allied products
28A Industrial chemicals
28B Plastic materials, synthetic rubber, and manmade fibers
28C Drugs
28D Soap, cleaners, and toilet goods
28E Paints and allied products; gum and wood chemicals
28F Agricultural chemicals
28G Miscellaneous chemical products
29 Petroleum and coal products
30 Rubber and plastics products, N.E.C.
31 Leather and leather products
31A Tanning; industrial leather goods; and shoes
31B Leather gloves; and miscellaneous leather goods
32 Stone, clay and glass products
32A Glass products
32B Cement and structural clay products
32C Pottery and related products
32D Concrete, plaster, and cut stone products
32E Abrasive, asbestos, and miscellaneous nonmetallic mineral
products
33 Primary metals industries
33A Blast furnaces, steel works, and rolling and finishing mills
33B Iron and steel foundries
33C Smelting and refining of nonferrous metals and alloys
33D Nonferrous metal mill and foundry products
33E Forging and miscellaneous primary metal products
34 Fabricated metal products
34A Metal cans, cutlery, handtools, and general hardware
34B Heating apparatus (except electric and plumbing fixtures)
34C Fabricated structural metal products
34D Screw machine products, fasteners and washers; metal stamp-
ings; and metal services
34E Miscellaneous metal products
35 Machinery, except electrical
35A Engines and turbines and farm machinery and equipment
35B Construction, mining, and materials handling machinery and
equipment
68
-------
35C Ketalworking machinery and equipment
35D Special industry machinery, except metalworking machinery
35E General industrial machinery and equipment
35F Office, computing and accounting machines
35G Service industry machines and machines shops
36 Electrical equipment and supplies
36A Electrical measurement and distribution equipment
36B Household appliances
36C Electric lighting and wiring equipment
36D Communication equipment, including radio and TV and elec-
tronic components and supplies
37 Transportation equipment
37A Motor vehicles and equipment
37B Aircraft and parts
37C Ship and boat building, railroad and miscellaneous transporta-
tion equipment
38 Instruments and related products
38A Instruments; surgical, dental, and ophthalmic equipment and
supplies
38B Photographic equipment; clocks, watches, and watchcases
39 Miscellaneous manufacturing industries
39A Jewelry, silverware, and plated ware
39B Musical instruments and parts; toys and sporting goods
39C Office supplies, costume jewelry, and notions
39D Miscellaneous manufactures
3. United States Department of the Interior, Federal Water Pollution Con-
trol Administration. The cost of clean water. Washington, D. C.:
Author, 1968. (Series includes the following reports:)
Volume 1. Summary report
Volume 2. Detailed analyses
Volume 3. Industrial waste profiles
Number 1. Blast furnaces and steel mills
Number 2. Motor vehicles and parts
Number 3. Paper mills except building
Number 4. Textile mill products
Number 5. Petroleum refining
Number 6. Canned and frozen fruits and vegetables
Number 7. Leather tanning and finishing
Number 8. Meat products
Number 9. Dairies
Number 10. Plastics materials and resins
Volume 4. State and major river basin municipal tables
69
-------
4. United States Department of the Interior, Federal Water Pollution Con-
trol Administration. The cost of clean water and its economic impact.
Washington, D. C.: Author, 1969. (Series includes the following
reports:)
Volume Number 1. The report
Volume Number 2. Appendix
Volume Number 3. Sewerage charges
Volume Number 4. Projected wastewater treatment costs in the organic
chemicals industry.
5. United States Department of the Interior, Federal Water Pollution Con-
trol Administration. The economics of clean water. Washington, D. C.:
Author, 1970. (Series includes the following reports:)
Volume Number 1. Detailed analysis
Volume Number 2. Animal waste profile
Volume Number 3. Inorganic chemicals industry profile
6. United States Department of the Interior, Federal Water Pollution Con-
trol Administration. Projects of the industrial pollution control
branch. Washington, D. C.: Author, 1970. (Publication DAST-38 of
the Water Pollution Control Research Series)
7. United States Department of Commerce, Business and Defense Services
Administration. United States industrial outlook1970.
8. United States Department of the Interior, Federal Water Pollution Con-
trol Administration. Research, development, and demonstration
projects. Washington, D. C.: Division of Applied Science and Tech-
nology, 1970.
9. Lawson, B. R. Atlas of industrial watery use. Ithaca, New York:
Cornell University Water Resources Center, 1967. (Publication 18)
10. Manufacturing Chemists Association. Toward a clean environment, 1967.
11. National Association of Manufacturers and Chamber of Commerce of the
United States. Water in industry, 1965.
12. United States Department of Commerce, Bureau of the Census. Water
use in manufacturing. Washington, D. C.: Author, 1963. (Publica-
tion Number MC63(1)-10)
13. Ackerman, E. A., & Lof, G. O. G. Technology in American water devel-
opment. Resources for the Future Inc., 1959.
14. United States Department of Commerce, Bureau of the Census. County
and city data book 1967. Washington, D. C.: Author, 1967.
70
-------
15. United States Department of Commerce, Business and Defense Services
Administration. Industry profiles 1958-1967, 1969
16. Resources Agency of California, State Water Quality Control Board.
Water quality criteria, 1963. (Publication Number 3-A)
17. United States Department of the Interior, Office of Water Resources
Research. The economic value of water in industrial uses. Washing-
ton, D. C.: Author, 1969.
18. Public notice, originating from sources such as newspapers, radio,
television, periodicals, congressional record, conservation groups,
local government, etc.
19. Reports originating within the framework of the Federal Water Quality
Administration.
20. McDermott, J. H., & Sayers, W. T. The role of water quality monitor-
ing in water pollution control. National Meeting of the American
Chemical Society, New York, 1969.
21. Bramer, H. C. Economically significant physicochemical parameters of
water quality for various uses. American Society of Testing Mate-
rials Symposium on Water Quality, Philadelphia, 1966.
22. Manufacturing Chemists Association, Inc. Chemical statistics handbook.
(6th ed.), 1966.
23. United States Department of the Interior, Office of Water Resources
Research. Water resources research catalog; Volume 5. Washington,
D. C.: Author, 1969.
24. Schnell Publishing Company. Oil, paint, and drug reporter. New York:
Author, weekly edition.
71
-------
APPENDIX B
FUNDED PROJECT DESCRIPTION FOR 1200 PROGRAM R&D EXPENDITURES PRIORITIES
Data Collection Form 0100
73
-------
FUNDED PROJECT DESCRIPTION FOR 1200 PROGRAM R&D EXPENDITURES PRIORITIES
1. INDUSTRY INVOLVED
a. PPBS number(s)
b. SIC code(s)
2. LOCATION
a. Grantee address (state)
b. Project site(s) (state)
3. DATES
] II I I
1234 5678 9 10 11 12 13 14 15 16
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 34
35 36 37 38
a. Date awarded: month
4. FUNDING
year
b. Duration (months)
39 40
41 42
43 44
a. Total project cost: amount (thousands) $
b. Federal funds: amount (thousands) $
5. WASTEWATER CONSTITUENTS TO BE REDUCED
months
months
51 52 53 54
55 56
Constituent
Rank No.
Constituent
Rank No.
6. TYPE OF PROJECT
a. Research
65 66
I I Desk-top study j j
Laboratory
Conceptual study
67 68
69
70
72
D
Form No.
74
Hands-on research
Bench-scale pilot
State-of-the-art report
0 0
Card No.
76 77 78 79
D
71
n
73
D
75
80
74
-------
b. Development I _ Special Facility | _ | Industrial in-plant | __
. Demonstration
Industrial in-plant | | Joint industrial-municipal |_
d. Grant
e. Contract
f. Full-scale
g. Pilot plant
8
10
7. PROJECT IMPLEMENTATION
a. Manufacturing Industry
b. University, or related group
11
12
c. Not-for-profit R&D Institution
d. Profit-making R&D organization
14
e. Governmental unit
8. OBJECTIVES OF PROJECT
Federal
Regional
State
Local
15
16
17
18
19
22
New
New
a. Wastewater volume reduction
c. Treatment process
d. Process equipment
e. Cost Reductions
g. Treatment control
h. Engineering
b. Water reuse
21
Improvement
23 24
Improvement
25 26 27
f. Economic feasibility determination
28
29
Measurement
Automation
30
Conventional
31 32
Mathematical Models
33
Process design
Plant design
36 37
34 35
Information systems
38
i. Production process modifications
k. Manpower factors | | training
41 42
j. By-product recovery
40
requirement
reduction
43
44
1. Applicability: Small plants | | Large plants
Old technology
45
Average technology
46
New technology
47
48
49
n. Wastewater characterization
>. Sludge disposal methods
n. Wastewater effects
50
51
Form No.
0 0
Card
52
76 77 78 79
80
project No.
75
-------
APPENDIX C
STATE DIMENSIONS FOR 1200 PROGRAM R&D EXPENDITURES PRIORITIES
Data Collection Form 6100
77
-------
STATE DIMENSIONS FOR 1200 PROGRAM R&D EXPENDITURES PRIORITIES
1. STATE
2. CENSUS BUREAU REGION Code
3 4
Code
3. CENSUS BUREAU DIVISION Code
5 6
4. EPA (COST OF CLEAN WATER) DRAINAGE REGION(s):
a. Region Code
b . Region Code
c . Region Code
d . Region Code
5. WATER RESOURCES COUNCIL DRAINS
a. Area Code
b . Area Code
c . Area Code
d . Area Code
6. CENSUS BUREAU INDUSTRIAL WATEF
a . Region Code
b. Region Code
c . Region Code
d . Region Code
7 8
11 12
15 16
Number of Counties
Number of Counties
Number of Counties
Number of Counties
19 20
LGE AREA(s) :
23 24
27 28
31 32
_L
9 10
13 14
17 18
21 22
Number of Counties ;
Number of Counties
Number of Counties
Number of Counties
25 26
29 30
33 34
35 36
I USE REGION (s) :
39 40
43 44
47 48
51 52
7. INDUSTRIAL WASTEWATER VOLUME
Number of Counties
Number of Counties
Number of Counties
Number of Counties
(billion gallons)
Data Valid for the Year
Form No.
6
1
0
0
Card No.
1
37 38
41 42
45 46
49 50
53 54
55
56
19
57 58
59 60
76 77 78 79
80
78
-------
1 POPULATION
2. LAND AREA
(1000)
12345
(sq. mi.)
6 7 8 9 10 11
3. VALUE ADDED BY MANUFACTURE
($ million)
12 13 14 15 16
4. TOTAL NUMBER OF MANUFACTURING ESTABLISHMENTS
17 18 19 20 21
5. NUMBER OF MANUFACTURING ESTABLISHMENTS WITH MORE THAN 20 EMPLOYEES
a. Food and Tobacco Products
b. Textile, Apparel, and Leather Goods
c. Paper and Printing
d. Chemicals, Petroleum, Rubber, and Plastics
e. Lumber, Wood Products, and Furniture
f. Stone, Clay, and Glass Products
g. Primary and Intermediate Metal Products
h. Electrical and Non-Electrical Machinery
i. Transportation and Ordnance
j. Instruments and Miscellaneous Products
6. CAPITAL EXPENDITURES BY MANUFACTURERS
7. INDUSTRIAL WATER USE, ANNUAL
22 23 24 25 26
27 28 29 30
31 32 33 34
67 68 69 70 71
72 73 74 75
6
1
0
0
Form No. I ° I x I u I u I Card No.I 2
76 77 78 79 80
79
-------
1. LAND IN FARMS
2. VALUE OF FARM PRODUCTS SOLD
3. WATER AREA
4. POPULATION USING PUBLIC WATER SUPPLIES
5. ANNUAL PRECIPITATION
6. STATE PARK AND RECREATIONAL AREAS
7. STATE PARK AND RECREATIONAL AREAS
8. STATE FISHING LICENSES ISSUED
9. METROPOLITAN AREA POPULATION
10. ELECTRICAL ENERGY PRODUCTION
11. ANNUAL WATER RUNOFF
12. ANNUAL WATER WITHDRAWALS
13. AGRICULTURAL AND BIOLOGICAL SCIENTISTS
(1000 acres)
123456
($ million)
7 8 9 10
(sq. mi.)
(inches)
(1000 acres)
(1000 visits)
(1000)
31 32 33 34
(1000)
35 36 37 38 39
(10° kw.-hr.)
40 41 42 43 44 45
(1000 acre-feet)
(1000 acre-feet)
46 47 48 49 50
51 52 53 54 55
56 57 58 59 60
14. PSYCHOLOGISTS, ECONOMISTS, AND OTHER SOCIAL SCIENTISTS
61 62 63 64 65
15. ATMOSPHERIC AND EARTH SCIENTISTS
66 67 68 69 70
16. ALL OTHER SCIENTISTS
71 72 73 74 75
Form No.
6
1
0
0
Card No.
76 77 78 79
80
80
-------
APPENDIX D
POLLUTION SEVERITY OF WASTEWATER CONSTITUENT FOR 1200 PROGRAM
R&D EXPENDITURES PRIORITIES
Data Collection Form 4100
81
-------
POLLUTION SEVERITY OF WASTEWATER CONSTITUENT FOR 1200 PROGRAM
R&D EXPENDITURES PRIORITIES
1. WASTEWATER CONSTITUENT
Code
2. FREQUENCY OF MENTION IN STATE WATER QUALITY STANDARDS
456
3. INDUSTRIES PRODUCING WASTEWATER CONSTITUENT (SIC numbers):
a. Industry
b. Industry
c. Industry
d. Industry
e. Industry
4. CONSTITUENT RANKED
5. CONSTITUENT RANKED
on basis of economic effect:
23 24 25 26
highest rank
on basis of regional appraisals: highest rank
31 32
6. CONCENTRATIONS OF CONSTITUENT IN mg. PER LITER TO PERMIT:
33 34
a. "All" water uses
7. CONSTITUENT RANKED
8. CONSTITUENT RANKED
35 36 37 38 39 40
b. "Most" water uses
41 42 43 44 45 46
on basis of "All" water uses: highest rank
on basis of "Most" water uses: highest rank
51 52
9. FREQUENCY OF PUBLIC MENTION RANKED
on basis of highest rank of
55 56
10. IN THE ABSENCE OF SPECIFIC INFORMATION, RANK THIS WASTEWATER CONSTITUENT
57 58
the same as
Code
Data Valid for the Year
73 74
Form No. I 4 1 l I ° I ° I No. of Cards I 1 I End of Card Li
75 76 77 78 79 80
82
-------
APPENDIX E
INDUSTRY DIMENSIONS FOR 1200 PROGRAM R&D EXPENDITURES PRIORITIES
Data Collection Form 2100
83
-------
INDUSTRY DIMENSIONS FOR 1200 PROGRAM R&D EXPENDITURES PRIORITIES
1. INDUSTRY DESCRIPTION
a. PPBS number(s)
b. SIC code(s)
7234 5678
9 10 11 12 13 14 15 16 17 18 19 20
2. SIZE OP INDUSTRY (Plants using more than 20mgy, except for item 2a)
> 20mgy intake
b. Number of Large Plants
a. Number of Small Plants
c. Number of Employees
d. Value Added by Manufacture
e. Added Water Use
($ million)
(billion gallons) (
f. Annual Water Intake
g. Annual Water Intake for Process
3. GEOGRAPHICAL DISTRIBUTION OF INDUSTRY (Plants using more than 20mgy)
a. Number of State with Plants
b. Number of Plants per State in States with Plants
c. Water Discharged per State in States with Plants (billion gallons)
34 35 36 37 38 39
(billion gallons) j | |
58 59 60 61
d. Number of Employees per State in States with Plants
e. Population of States with Plants (1000)
(1000)
62 63 64
65 66 67 68 69 70
Data Valid for the Year
19
71 72
Form No.
2
1
0
0
Card No.
76 77 78 79
80
84
-------
4. WATER USE PRACTICES IN INDUSTRY (Plants using more than 20mgy)
a. Purchased Water Intake
b. Ground Water Intake
c. Brackish and/or Salt Water Intake
5. ECONOMIC STATUS OF INDUSTRY
a. Depreciation Period for Major Equipment
b. Return on Invested Capital
(billion gallons)
(billion gallons)
(billion gallons)
9 10 11 12
(years)
13 14
c. Return on Sales
15 16 17
d. Projected Growth Rate (%)
on the basis of
e. Value of Shipments
18 19 20
Increase I J Decrease
21 22 23 24 25
years)
from
26 27
28 29
30 31
($ million)
32 33 34 35 36
6. WASTEWATER TREATMENT FACILITIES
a. Investment Provided by Industry
($ million)
37 38 39 40
b. Investment Provided by Muncipalities
($ million)
7.
c. Treatment Facility Investment of that Required
d. Wastewater Discharged to Municipal Sewers
e. Annual Operating and Maintenance Costs
PRODUCTION PARAMETERS
($ million)
a. Operational Days per year
51 52 53
c. Level of Technology (%): old
b. Production as % Capacity
54 55 56
average
advanced
57 58
59 60
61 62
Form No.
2
1
0
"1
Card No.
76 77 78 79
80
85
-------
8. WASTEWATER CONSTITUENTS ASSOCIATED WITH INDUSTRIAL PROCESSES
Biochemical Oxygen Demand
Organics
Phenols
Chemical Oxygen Demand
Brines
Copper
Acetates
Temperature
Taste
Alkanlinity
Chlorophenols
Cadmium
Lignins
Detergents
Lead
13
16
19
D
22
Color
Oil and Grease
Cyanide
Chromium
Iron
Total Solids
Total Nitrogen
Hydrocarbons
Ammonia
Metals
PH
Toxicity
Settleable Solids
Mercury
Hardness
43
44
Other Constituents Codes
Other Constituents
58 59 60
Form No.
Suspended Solids
Acidity
Odor
Phosphate
Zinc
Slime Growths
Coliforms
Turbidity
Hexane Solubles
Organic Nitrogen
Fluorides
Nitrogen
Nickel
Arsenic
Radioactivity
Codes
61 62 63
2
1
0
0
Card No.
*s
n
12
18
D
21
45
76 77 78 79
80
86
-------
STATES WITH PLANTS OF THE INDUSTRY USING MORE THAN 20mgy:
Maine
Massachusetts
New York
Ohio
Michigan
Iowa
South Dakota
Delaware
Virginia
South Carolina
Kentucky
Mississippi
Oklahoma
Idaho
New Mexico
Nevada
California
40
n
43
New Hampshire
Rhode Island
New Jersey
Indiana
Wisconsin
Missouri
Nebraska
Maryland
West Virginia
Georgia
Tennessee
Arkansas
Texas
Wyoming
Arizona
Washington
Alaska
35
Vermont
Connecticut
Pennsylvania
Illinois
Minnesota
North Dakota
Kansas
District of Columbia
North Carolina
Florida
Alabama
Louisiana
Montana
Colorado
Utah
Oregon
Hawaii
49
50
51
. 2 I 1 I ° I 0 Card No. 4
87
-------
APPENDIX F
GENERAL STATISTICS FOR 1200 PROGRAM R&D EXPENDITURES PRIORITIES
Data Collection Form 8100
89
-------
GENERAL STATISTICS FOR 1200 PROGRAM R&D EXPENDITURES PRIORITIES
CENSUS BUREAU REGION
New England
Middle Atlantic
East North Central
West North Central
South Atlantic
East South Central
West South Central
Mountain
Pacific
Data Valid for the Year
INDUSTRIAL RSD FUNDS ($ MILLION)
FEDERAL
1 2
3 4
9 10 11 12
17 18 19 20
25 26 27 28
33 34 35 36
41 42 43 44
49
50 51 52
57 58 59 60
INDUSTRY
567
8
13 14 15 16
21 22 23 24
29 30 31 32
37 38 39
40
45 46 47 48
53 54 55 56
61 62 63 64
65 66 67 68 69 7O 71 72
19
73 74
Form No.
8
1
0
-I
Card No.
76 77 78 79
80
90
-------
INDUSTRIAL R&D FUNDS ($ MILLION)
INDUSTRIAL CATEGORY
Food and Kindred Products
Textiles and Apparel
Lumber, Wood Products, and Furniture
Paper and Allied Products
Chemicals and Allied Products
Petroleum Refining and Extraction
Rubber Products
Stone, Clay, and Glass Products
Primary Metals and Fabricated Products
Machinery and Transportation Equipment
All Other Industries
FEDERAL
| |
1
2 3
7 8
11 12
15 16 17
21 22 23
28 29 30
34
35 36
40 41
45 46 47
51 52 53
INDUSTRY
24
4
5 6
9 10
13 14
18 19 20
25 26 27
31 32 33
37 38 39
42 43 44
48 49 50
54 55 56 57
58 59 60 61 62
63 64 65 67 68
Data Valid for the Year
19
74 75
Form No.
8
1
0
0
Card No.
76 77 78 79
80
91
-------
1
5
ACORNS ion Number
2
Subject Field & Group
06A,05D
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
Syne cti cs Co roo ration
4790 Wm. Flynn Highway, Allison Park, Pennsylvania 15101
Title
A System for Industrial Waste Treatment RD&D Project Priority Assessment
10
Authors)
Bramer, Henry C.
DeHaven, Robert C.
Leavitt, Alvan W.
16
Project Designation
EPA, WQO Contract No. 14-12-840
21
22
Citation
23
Descriptors (Starred First)
*Priorities, *Resource Allocation, *Research and Development, Industrial Waste,
Cost Allocation, Federal Budgets, Expenditures, Economic Prediction, Systems
Analysis, Statistical Models, Mathematical Models, Computer Models, Information
Retrieval, Water Pollution Source
25
Identifiers (Starred First)
*Management Information System
27
Abstract
JThe Environmental Protection Agency faces the difficult problem of determining priori-
ties for research and development expenditures which will yield maximum overall bene-
fits reflected in water quality improvements at minimum total costs. The computerized
management information system described herein rapidly and efficiently determines such
RD&D expenditure priorities by maximizing the expected returns on investment.
The modeling system developed as a result of this effort is unique insofar as informa-
tion management systems are concerned in the degree to which user interaction is al-
lowed. At any point during operation of the system, the user may insert judgmental
factors. The system also has been designed to function with readily available data,
such as that from the Bureau of the Census. Systems which incorporate theoretically
desirable, but virtually unattainable, data have little operational utility. The
mathematical and statistical methods employed in the development of the system fo-
cused on and supported the structuring, testing, and partial programming of three
fixed-X regression modules.
Abstractor
Robert C. DeHaven
Institution
Synectics Corporation
WR:102 (REV. JULY 1969)
WRSIC
SEND TO:
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 20240
* GPO: 1969-359-339
------- |