EPA 600/5-76-004
July  1976
Socioeconomic  Environmental Studies  Series

-------
                 EVALUATION OF WATER QUALITY MODELS

                   A MANAGEMENT GUIDE FOR PLANNERS
                                  By

                         G. Paul  Grlmsrud
                         E. John  Finnemore
                           H. James Owen
                     Contract No.  68-01-2641
                        Project Officer

                        Donald H.  Lewis

            Office of Air, Land,  and Water Use
            Office of Research and Development
            U.S.  Environmental Protection Agency
                   Washington, D.C.   20460
                         Prepared  for
            U.S.  ENVIRONMENTAL PROTECTION AGENCY
              OFFICE OF RESEARCH AND DEVELOPMENT
                    WASHINGTON, D.C.   20460
For sate by the Superintendent ol Documents, U.S. Government Printing Office, Washington, D.C. 2M02 - Price $2.90

-------
                       DISCLAIMER
This report has been reviewed by the Office of Research and
Development, U.S. Environmental Protection Agency,  and approved
for publication.  Approval does not signify that the contents
necessarily reflect the views and policies of the U.S. Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
                                    ii

-------
                            ABSTRACT
     This report is designed as a handbook specifically oriented to
water quality and water resources planners and managers.  It presents
a large amount of basic information concerning water quality modeling
including procedures for:  model evaluation, model selection, inte-
gration of modeling with planning activities, and contracting modeling
projects.

     Planners without previous experience in water quality modeling may
use the information and procedures included in the handbook to determine
whether a water quality model could and should be used in a particular
planning program, and which specific model would be cost effective.  This
includes a step-by-step procedure leading to the rejection or selection
of models according to specific project needs.

     The handbook discusses the implications which accompany the decision
to model, including the needs for additional labor and specialized tech-
nical expertise which are generated.  Methods and procedures for inte-
grating the use and results of water quality models with other activities
of the planning process are described as well as the respective merits of
in-house and contracted modeling.  The handbook also deals with the pro-
cedures for obtaining and using contractual services for water quality
modeling.  Step by step instructions are provided for the preparation of
solicitations, evaluation of proposals and selection of contractors.
U. S. Government Employees are cautioned that the guidance on obtaining
contractual services contained herein does not supplant Federal Procurement
Regulations (FPR) or their own Agency contracting procedures.

     This report is submitted in fulfillment of Contract Number 68-01-2641,
under the sponsorship of the office of Research and Development, Environ-
mental protection Agency.
                                    iii

-------
                        TABLE OF CONTENTS
Chapter

   1.       INTRODUCTION
           1.1  PROJECT BACKGROUND                                  1
           1.2  PURPOSE OF HANDBOOK                                 3
           1.3  PROJECT SCOPE                                       4
   2.      MATHEMATICAL MODELING                                    7

           2.1  INTRODUCTION                                        7
           2.2  MATHEMATICAL MODELS VS.  OTHER TYPES OF MODELS        9
           2.3  TYPES OF MATHEMATICAL MODELS                       10
           2.4  USE OF MODELS                                      12
           2.5  ADVANTAGES AND LIMITATIONS OF MODELS               15
           2.6  CONCLUSIONS                                        19

   3.      EVALUATION AND COMPARISON OF  MODELS                     21

           3.1  INTRODUCTION                                       21
           3.2  GENERAL DESCRIPTION OF SELECTED MODELS             21
           3.3  DETAILED MODEL EVALUATION                          30
           3.4  OTHER MODELS                                       51
   4.      MODEL SELECTION AND COST EFFECTIVENESS EVALUATION       54
           SYSTEM

           4.1  INTRODUCTION                                       54
           4.2  DECISION TO USE A MODEL                            55
           4.3  SELECTION OF CANDIDATE MODELS                      56
           4.4  MODEL SELECTION PROCESS                            57
           4.5  PHASE I - APPLICABILITY TESTS                      57
           4.6  PHASE II - COST ESTIMATION                         64
           4.7  PHASE III - PERFORMANCE INDEX RATING:  SIMPLIFIED  68
           4.8  PHASE IV - PERFORMANCE INDEX RATING:  ADVANCED     73
           4.9  SELECTION OF ATTRIBUTE WEIGHTS                     79
           4.10 COST EFFECTIVENESS EVALUATIONS                     79
           4.11 DEMONSTRATION OF COST EFFECTIVENESS AND MODEL
                SELECTION METHODOLOGY                              81

   5.      MANAGEMENT OF MODELING                                  94

           5.1  INTRODUCTION                                       94
           5.2  NEED FOR ASSISTANCE                                95
           5.3  IN-HOUSE VS. CONTRACTUAL SERVICES                  98
           5.4  INTEGRATION OF MODELING AND PLANNING              105

-------
              TABLE  OF  CONTENTS     	 Continued.


Chapter                                                           Page

   6.      USE OF CONTRACTUAL SERVICES -  ADMINISTRATIVE, LEGAL
           AND PLANNING CONSIDERATIONS                             112

           6.1  INTRODUCTION                                       112
           6.2  PLANNING THE PROCUREMENT                           113
           6.3  PREPARATION OF THE REQUEST FOR PROPOSAL            H6
           6.4  PREPARING THE BIDDERS LIST                         136
           6.5  SOLICITATION                                       142
           6.6  EVALUUATION OF PROPOSALS                           146
           6.7  CONTRACTOR SELECTION                               153
           6.8  CONTRACTING                                         155
           6.9  PROJECT ADMINISTRATION                              160

 ACKNOWLEDGEMENTS                                                    164

 REFERENCES                                                         166

 GLOSSARY                                                           160

 ABBREVIATIONS                                                      176
                                   vi

-------
                          ILLUSTRATIONS
Figure                                                            Page

 2.1       MODEL RELATIONSHIP WITH INPUTS AND OUTPUTS              17

 4.1       FLOWCHART OF PROCEDURE FOR MODEL SELECTION AND
           COST EFFECTIVENESS EVALUATION:  PHASE I -
           APPLICABILITY TESTS                                     58

 4.2       FLOWCHART OF PROCEDURE OF MODEL SELECTION AND
           COST EFFECTIVENESS EVALUATION:  PHASE II -
           COST AND TIME CONSTRAINT TESTS                          65

 4.3       FLOWCHART OF PROCEDURE FOR MODEL SELECTION AND
           COST EFFECTIVENESS EVALUATION:  PHASE III
           PERFORMANCE INDEX RATING, SIMPLIFIED                    70

 4.4       FLOWCHART OF PROCEDURE FOR MODEL SELECTION AND
           COST EFFECTIVENESS EVALUATION:  PHASE IV -
           PERFORMANCE INDEX RATING, ADVANCED                      74
                                 vii

-------
                             TABLES
Table                                                           Page

 3.0       MODEL EVALUATION CATEGORIES                           30

 3.1       MODEL SUMMARY:   MODEL CAPABILITIES,  APPLICABLE
           SITUATIONS                                            32

 3.2       MODEL SUMMARY:   MODEL CAPABILITIES,  CONSTITUENTS
           MODELED                                               33

 3.3       MODEL SUMMARY:   MODEL FACTORS  ACCOUNTED FOR           34

 3.4       MODEL SUMMARY:   DATA REQUIRED,  FOR MODEL INPUTS        35

 3.5       MODEL SUMMARY:   ADDITIONAL DATA REQUIRED,  FOR
           CALIBRATION AND VERIFICATION                          36

 3.6       MODEL SUMMARY:   MODEL COSTS, INITIATION COSTS         37

 3.7       MODEL SUMMARY:   MODEL COSTS, UTILIZATION COSTS         38

 3.8       MODEL SUMMARY:   MODEL ACCURACY,  MODEL
           REPRESENTATION  INACCURACIES                            39

 3.9       MODEL SUMMARY:   MODEL ACCURACY,  NUMERICAL  ACCURACY     40

 3.10       MODEL SUMMARY:   MODEL ACCURACY,  SENSITIVITY TO
           INPUT ERRORS                                           41

 3.11       MODEL SUMMARY:   EASE  OF APPLICATION, SUFFICIENCY
           OF AVAILABLE DOCUMENTATION                            42

 3.12       MODEL SUMMARY:   EASE  OF APPLICATION, OUTPUT FORM
           AND CONTENT                                            43

 3.13       MODEL SUMMARY:   EASE  OF APPLICATION, UNDATEABILITY
           OF DATA DECKS                                          44

 3.14       MODEL SUMMARY:   EASE  OF APPLICATION, MODIFICATION
           OF SOURCE DECKS                                        45

 3.15       OTHER MODELS                                           52

 4.1       ATTRIBUTE WEIGHT RANGES                                80
                              vill

-------
                     TABLE     — Continued.






Chapter                                                         Page




 4.2       APPLICABLE SITUATIONS                                 83




 4.3       CONSTITUENTS MODELED                                  84




 4.4       DATA REQUIEMENTS FOR MODEL INPUTS                     85




 4.5       DATA REQUIREMENTS FOR CALIBRATION AND VERIFICATION    87




 4.6       INITIATION COSTS (PHASE II)                           88




 4.7       UTILIZATION COSTS (PHASE II)                          89




 4.8       PERFORMANCE INDEX RATING:  ADVANCED (PHASE IV)        90




 4.9       COST EFFECTIVENESS COMPARISON                         93
                                 Ix

-------
                                CHAPTER 1
                              INTRODUCTION

1.1   PROJECT BACKGROUND

An explicit goal of maintaining and enhancing water quality has been
established by the Congress and most states.  Many interstate authorities,
regional organizations, and local governments have followed suit.   The
extent of the nationwide commitment to the goal of clean water, and the
priority given its accomplishment, are evidenced by the magnitude of the
investment being made for prevention and abatement of water pollution.

The Federal Water Pollution Control Act of 1972, Public Law 92-500,
authorizes federal expenditures totaling $27 billion.  Additional large,
non-federal expenditures are encouraged by the Act's cost-sharing provi-
sions, and by its requirements for industry wastewater control expenditures
to meet effluent standards.  An undetermined further investment is being
made by governments at all levels, industries, and other entities through
diverse water quality control programs other than those mandated by the Act.

In addition to the measurable financial costs associated with the main-
tenance and enhancement of water quality, less easily identified costs are
also incurred,  Among others, these include the social costs and the
opportunity costs attendant to various water pollution control programs.
While the measurement of these and other less visible costs cannot be made
explicitly, they are undoubtedly large.

The attention focused on achieving water quality goals and the massive
investment at stake make management decisions affecting water quality of
the utmost importance.

Major decisions regarding water quality management are being made daily as
a part of numerous planning programs.  It is incumbent on planners to assure
that the expenditures associated with their recommendations be justified and
that actions  taken will fully achieve all of the expected results.

-------
 The principal types of planning programs now underway affecting water quality
 management are those which result in the preparation of one or more of the
 following:
        1)    Long-range framework plans for the comprehensive and
              coordinated management of water and/or land and related re-
              sources over large areas, such as:
                  •    state water plans;
                  •    state land-use plans, and
                  •    framework studies pursuant to Section 102 of the
                       Water Resources Planning Act of 1965, Public Law
                       89-80.

        2)    Medium-range plans proposing or identifying specific actions
              for the development,  protection, conservation, and use of
              water,  land,  and/or other resources,  such as:
                  •    basin plans  for water quality management pursuant
                       to Sec.  303(e)  of Public Law 92-500;
                  •    areawide waste  treatment management plans pursuant
                       to Sec.  208  of  Public Law 92-500;
                  •    Level B  studies pursuant to  Sec.  209  of  Public
                       Law  92-500 or Public Law 89-80,  and
                  •    specific functional plans for water resources
                       development, energy,  transportation,  land use and
                       other purposes.

        3)     Detailed  site-specific studies such as those related  to:
                  •     facilities planning pursuant  to  Sec.  201  of  Public
                       Law  92-500, and
                  •     evaluation of discharge  permit applications.

Planners and  managers  responsible for the performance  of  the above  and
other types of planning  programs must usually  deal  with  the engineering,
economic, financial, legal, institutional and  environmental aspects of
water quality.  The detail required in  the  analysis of these aspects
depends upon  the  study purpose.  Broad  framework plans seldom require the
detailed analysis appropriate  to an areawide waste-treatment management
plan or determination of the best location  for an industrial outfall.

One of the first  crucial decisions encountered by planners and by managers
of planning activities is selection of planning methodologies which are

-------
appropriate to the situation and capable of execution within the limits of
time and funds.  The selection of planning methodologies must take into
account the types of analyses which are needed and the number of analyses
to be made.

Difficult technical problems are frequently encountered in making certain
evaluations.  These may include:

      •     the location, nature, extent, and impact of non-point
            sources of waste;
      •     the complex biological, chemical and physical interactions
            which take place over time between effluents and receiving
            waters;
      •     effects of various types and amounts of waste treatment
            or of the generation and release of various pollutants
            which directly affect water quality; and
      •     water quality impacts resulting indirectly from modifica-
            tion of hydraulic regimes, land use, or other actions.

In addition to their technical difficulty, the number of waste-water manage-
ment alternatives which exist and require evaluation may be large.  The com-
plexity of water quality analysis and the need for investigation of large
numbers of management alternatives has stimulated the development of a
variety of tools to assist planning.  These tools range from simple graph-
ical techniques to sophisticated computerized models.

Water quality models have assumed an important role in the decision-making
process.  They enable types and numbers of analyses which would otherwise
be Impractical in many cases.  However, the use of water quality models can
be costly and time consuming.  It is essential that planners give careful
attention to insuring that their use is cost-effective.

1.2   PURPOSE OF HANDBOOK

The purpose of the handbook is to assist planners in selecting and using
techniques of water quality analysis which are cost-effectively matched to
their planning responsibility.  More specifically, the handbook is intended

-------
 to provide planners a sufficient introduction to water quality modeling to
 enable effective communication with systems analysts and administrators
 regarding water quality modeling.

 1.3   PROJECT SCOPE

 Water quality modeling is a highly specialized technology.   Detailed de-
 scription of every aspect of that  technology would be impractical.   How-
 ever, the overview presented in this handbook provides planners the
 information needed to answer basic questions regarding water quality
 modeling and recognize when more expert assistance is required.   References
 are included to facilitate further investigation by those desiring  to do so.

 The number of water quality models described in the handbook is limited.
 Those chosen for inclusion are readily available and likely candidates for
 frequent future use.   They also demonstrate a range of model types.

 The information regarding the included models is based on the experience of
 the authors in developing,  modifying or using a number of them and  on review
 of documentation reports  and other available information.   All of the model
 codes were examined in detail as a part of  handbook preparation.  However,
 it was not within the scope of that effort  to test run the  models.   It should
 therefore  be  recognized that minor deviations from execution times  and other
 stated characteristics may  occur in particular applications.

 The methodology for selection of the water  quality model  best, fitting  a
 particular  need  was developed especially for inclusion in the handbook.
 Though new, the  methodology  has been tested  and  progressively improved  in a
 series of  case  studies carried  out  in  cooperation  with federal, state,  and
 local planners.  The  case studies  included critical  review  of  the method-
ology by potential users and  demonstration of  its  application  to water
quality planning for  the Shohomish Estuary at  Everett, Washington.

The information provided herein which deals with the technical and admini-
strative aspects of acquiring, using, and managing consulting services  is

-------
necessarily general.  Each case is likely to be more or less unique depending
on the nature and urgency of any need for assistance, the prevailing legal
framework and other factors.   U. S. Government personnel should realize that
the guidance on obtaining contractual services contained herein does not
supplant Federal Procurement Regulations (FPR) or their own Agency contracting
procedures.

-------
                                 CHAPTER 2
                           MATHEMATICAL MODELING

2.1   INTRODUCTION

The water resources planner, or manager, must concern himself in part
with understanding the behavior of the hydrologic system with which he
works.  Understanding is gained through use of existing scientific know-
ledge, by employing tools of analysis such as mathematics, and through
the design and evaluation of experiments.  Beyond understanding the
existing system, the planner or manager must also be concerned with how
the systems would behave as a result of decisions or actions that may be
taken now or occur in the future.  Planning, in particular, must assess
the extent to which specific objectives may be achieved through the
creation of facilities that did not exist before; as, for example, the
planning of a sewage treatment plant to satisfy new water quality stan-
dards.  In addition, the planner or manager must be able to determine
more efficient resource management strategies that would improve the per-
formance of the existing system.

One way to plan, analyze, and decide on an appropriate course of action
is to make the considered changes in the actual system's operation and
observe the effects directly.  Whenever practical, this kind of controlled
experimentation is highly desirable.  However, such direct experimentation
on a river system is usually impractical.  Few planners would propose
constructing treatment plants at all proposed locations in order to select
the most effective one, nor propose the severe disruption of service to
test emergency strategies or contingency plans.  Instead, the planner may
find it necessary to model  the physical system in a way that will permit
him to analyze the essential cause and effect relationships.  The model
may be a scaled down replica of  the actual system, or it may be a
mathematical description based on known physical laws and empirical
formulas.  A water quality model, as defined in this project, is a

-------
 representation of  receiving water quality  (the system) as a function of
 various man-induced and natural  inputs  (system elements) such as effluent
 discharges, storm  runoff and solar radiation.  In a mathematical model,
 the relationships  that exist between the various system elements are
 represented by a set of mathematical equations.  Depending on the type of
 questions to be answered about the behavior of the actual system, the
 mathematical model may range from a few simple equations that can be
 solved by hand computation to hundreds of complex equations that can be
 solved only through the use of a digital computer.  No matter what level
 of complexity the equations take, when these models simulate the behavior
 of the physical system (water quality) under various conditions, the model
 is called a mathematical simulation model.

 The objective of this Chapter is to briefly describe the merits and limit-
 ations of mathematical simulation models, with particular attention placed
 on those for prediction of water quality in rivers, lakes, and estuaries.
 Various types of models exist and each varies in its applicability to
 specific situations.  The information contained in the following chapters
 is intended to aid the planner in deciding whether, in a particular situa-
 tion, the use of a simulation model is warranted and to suggest guidelines
 for the selection of a specific model.

Mathematical simulation models, when used properly and with an understand-
 ing of their limitations, can greatly expand the range of alternatives a
planner may consider and assist in providing information in an organized
form.  It should be stressed, however, that such models are nothing but
 tools to assist planners in the difficult task of evaluating alternatives.
They are not a substitute for experience and good judgment, but a means for
permitting these qualities to be used more effectively.

-------
2.2   MATHEMATICAL MODELS VS. OTHER TYPES OF MODELS

There are a multitude of ways in which a particular system can be
modeled.  Models can be classified by their structuring characteristics
as follows:

1)    Physical (Iconic) Models - These models are intended to "look like"
      the subject of inquiry.  They are usually characterized by the use
      of scaling techniques.  Examples of this type of model include the
      globe, still pictures, etc.  Very few models of this type exist
      for water quality modeling, although physical hydraulic models of
      certain water bodies have been extensively used.

2)    Analog Models - These are characterized by the use of a convenient
      transformation of one set of properties for another in accordance
      with specified rules.  For example, certain mechanical systems may
      be represented by an electrical equivalent.  Other examples of
      analog models include:  block diagrams, flowcharts, plant layouts,
      etc.

3)    Mathematical (Symbolic) Models - These are characterized by the
      fact that the components of the subject of inquiry and the inter-
      relationships among them are represented by symbols, both mathe-
      matical and logical.

Of the model types listed above, only the symbolic models appear deserving
of consideration in most water quality planning activities.  The physical
models are often useful in studying the physical characteristics of cer-
tain water bodies; however, in order to apply the model on new prototypes
an entirely new model must be constructed.  The construction of a useful
physical water quality model for a single water body is very expensive.
Also, its usefulness is quite limited.  The analog models in the class of
logical flowcharts are often useful for very general planning, but only
in describing qualitative relationships.  Electrical and mechanical
analogue models are rarely useful in water quality problems.

-------
 2.3   TYPES OF MATHEMATICAL MODELS

 No single general model exists that can simulate all aspects of water
 quality.   Furthermore,  it is not clear that such a model should be
 developed.   Its resulting complexity,  assuming that it  could be created,
 the amount  of data needed to validate  it,  and  the cost  of operation would
 likely make its use impractical.  Instead,  a wide variety of specialized
 models have been developed to efficiently  handle particular  aspects of
 water quality of interest to planners.   From the planner's viewpoint,
 models can  be classified by their applicability to various parts of a
 hydrologic  system,  the  effects simulated by the model,  and the  method of
 analysis.

 Applicability to Type of System

 Models may  be applicable to streams, estuaries,  lakes,  or impoundments.
 Some  may be used to simulate more than  one  type of water  body,  e.g.,  the
 RECEIV portion of Storm Water Management Model  is  applicable to both
 rivers and  estuaries.

 Water Quality Constituents

 Models  can  be classified  according  to the water  quality constituents  that
 they  can simulate and also  according to  methods  of  their  simulation.
 Models  have been developed  for the  concentrations  of various  subsets  of
 numerous conservative* and  nonconservative* physical, chemical  and biolog-
 ical  substances  such as  temperature, dissolved oxygen, zinc,  coliforms,
 etc.

Method of Analysis

This classification is a broad one and can be subdivided into several
others according to the level of complexity and  the method of approach:
  Words  with asterisks  are defined in the glossary.
                                      10

-------
A.    Static - Dynamic;  Static models are used to evaluate steady-state
      conditions in which the values of the variables do not change with
      time.  When some parameters are time varying or the effects of
      transient phenomena, such as storm runoff, must be evaluated, then
      dynamic models are used.  Static models tend to be simpler and
      require less computational effort.  Static solutions can be attained
      from dynamic models if the time horizon is sufficiently long so
      that steady-state conditions are achieved.  However, this is gener-
      ally an inefficient use of a dynamic model if only information
      about steady-state conditions is needed.

B.    Spatial Dimensionality:  Although real systems are three dimensional,
      sufficiently accurate results can quite often be obtained by modeling
      a system using only one or two dimensions.  For example, one-
      dimensional models would often be adequate to describe the gross
      nature of water flows in river systems, most lakes or segments of
      lakes, and shallow, well-mixed estuaries through representation
      by a network of channels.  Two-dimensional models would be appropriate
      for stratified estuaries, or for harbors and estuaries where one-
      dimensional representations too severely constrain the flow and
      transport directions.

C.    Deterministic - Stochastic;  Deterministic models are based on phy-
      sical laws of classical physics, such as mass and momentum conserv-
      ation, and on empirical formulas; they are frequently regarded as
      expected-mean-value models.  Stochastic  (or probabilistic) models
      take into account the randomness in many phenomena.  Although sto-
      chastic models (which occur in tremendous variety, depending upon
      the assumptions about the physical processes and the type of mathe-
      matics used) may be a more realistic representation of physical
      processes such as diffusion, their validation in water quality
      modeling is difficult since excessive prototype data are necessary
      to establish the various probabilities.  Most of the simulation
      models in use are deterministic.
                                      11

-------
 Method  of  Computation

 This  type  of  classification separates  the  various  techniques used  for
 solving mathematical  equations.   These techniques  range  from hand  compu-
 tation  and nomograph  solutions  to sophisticated  computer simulation.

 The principal advantages  of hand  calculation  and nomograph  solution tech-
 niques  are that  they  are  inexpensive to use for  simple,  non-extensive
 problems,  and they  offer  a  planner more opportunity to get  a quantitative
 feeling of the system processes.  The  main disadvantage  of  this  solution
 technique  is  that they are  feasible when only a  limited  number of  computa-
 tions are  needed.   For applications involving thousands  of  computations,
 this  approach becomes tedious, human errors often  arise,  and the project
 costs become  large.

 A basic argument for  using  computer simulation models is  that they can
 yield simulation results  for even complex  planning problems quickly and
 accurately.   The execution  of the computations required  for a water quality
 simulation may only take  a  few seconds or  minutes and the corresponding
 machine time  cost may be  of the order  of $5-200, depending on the  specific
 program and the cost  of computer  time.  The main objection  to computer
 solutions  is  the high cost  of model development  and set-up needed  before
 actual  computation can be made.

 Other classifications of  simulation models could be made  (e.g.,  size of
 computer memory required, computational methods, used, etc.) but these
 are generally of more interest to designers of models and system analysts
 than  to planners.

 2.4   USES  OF MODELS

 In addition to being responsible  for preparing specific plans, the planner
must  also  identify and analyze management  alternatives and make recommen-
dations  on the basis of his experience and such  technical analyses as he
                                     12

-------
may be able to accomplish.  Frequently these several tasks are subject
to severe constraints of available time, personnel, and funds.  A pri-
mary consideration should then be whether use of a model will assist him
in water quality management planning.

Mathematical simulation models cannot provide answers to all questions
related to water quality planning.  Since the basis of the models is the
mathematical representation of pertinent relationships between physical
variables such as streamflow, temperature and concentration, the simula-
tions can provide quantitative estimates of the quality effects due to
changes in either the physical aspects of the hydrologic system or the
waste loads discharged into the system.  Although ingenuity in the use
of models and interpretation of simulation results can provide the planner
with additional insights into the behavior of a particular river system,
caution should be exercised so as not to exceed the limitations imposed
by the model's underlying assumptions.  For example, the magnitude of the
peak flow that can possibly occur in a particular river reach cannot
generally be inferred from static analyses that estimate maximum steady-
state flows.

The basic value of a simulation model lies in its use to study the behavior
of real or proposed systems without the need for making observations on the
physical system itself.  Some important uses of models can be classified
into the three broad categories of system simulation, prediction of
performance, and model calibration.

Systems simulation is the most common use of models.  Simulation of
existing systems enables determining their behavior under a variety of
conditions, aids in understanding of the interrelationships between
elements of the system, and indicates the existence of possible trouble
areas.  The simulation can range from evaluation of simple steady-state
stream flows to the dynamic behavior of flows, temperatures, and concen-
trations of many water quality constituents.  Although occasionally
observations on the actual system may be made for  these purposes, this  is
often impossible because of inability to control certain natural phenomena.
                                13

-------
 Once  verified,  simulation models  can  be used  to  assess  the  effect  on
 water quality of possible changes in  the hydrologic  system  such  as the
 addition of  new waste  loads  or  treatment plants  or changes  in the  nature
 of  effluents from  such sources.   This ability to simulate the performance
 of  proposed  or  planned systems  by using mathematical simulation  models
 where no direct observation  on  the actual system would  be possible enables
 the comparison  of  alternative plans for water quality management.   The
 use of criteria such as water quality standards  as a basis  for evaluating
 the acceptability  of various plans in conjunction with  the  models'  capabil-
 ity to predict  water quality levels resulting from alternative plans can
 provide  the  planner considerable  insight into plan selection.  Furthermore,
 it  is often  possible to infer from the simulation results the  direction a
 plan  should  take to satisfy  certain water quality criteria.   For example,
 in  trying to determine  the optimum location of a treatment  facility, a
 simulation model may make it readily  apparent  that the  new  plant could not
 improve  the  water  quality to the  desired extent  regardless  of  its  location,
 and that  another approach should  be taken.

 Simulation models  are also commonly used to achieve  their own  calibration
 on  specific  prototypes.  Certain  input-output prototype data are in this
 case  used to  obtain values for unknown model parameters or  inputs.  From
 these values, it is also possible  to  estimate the relative  importance to
water  quality of modeled component  processes and relationships in  the
 corresponding prototype.  An important use of this technique is  in the
 estimation of non-point effluent  sources, which  are  difficult  to measure.
 In  this case  the data on prototype water  quality are used in conjunction
with  the model  to  calibrate  its non-point source inputs and thus estimate
the non-point sources.

These several applications of computer simulation models make them useful
tools for the technical analyses planners must make  to obtain information
about the performance of existing or proposed systems.  Furthermore, the
information obtained from the simulations can be used to evaluate and
                                      14

-------
compare different water quality plans.   However, it is again stressed
that the simulations provide only useful information.   The critical
evaluation of that information remains the responsibility of the planner.

2.5   ADVANTAGES AND LIMITATIONS OF MODELS

Mathematical simulation models can be used in the planning process in a
variety of ways as described in the previous section.   Whether they should
be used is a separate and important question.  Unfortunately, there is no
direct answer to this question that would apply to all or even most situ-
ations.  Instead, the planner needs to carefully assess the requirements
of each case and make this decision accordingly.  This decision must take
into account whether a suitable model is available or whether an existing
model should be modified, or even whether a new one meeting specifications
should be developed.  Time constraints, data availability and data collec-
tion requirements all exert a strong influence on decisions to model and
may differ widely from case to case.  Detailed guidelines for consideration
of these factors are presented in Chapter 4.  However, by considering the
advantages and the inherent limitations of computer models, some broad
guidelines regarding the utilization of models can be suggested.

The basic advantage of using mathematical simulation models is their
ability to give quantitative answers to complex planning problems, such
as the cause-effect relationships of water pollution.  Depending upon the
solution technique used, these quantitative answers can be derived quickly
and inexpensively.  For example, the execution of the computer computa-
tions required for a water quality simulation generally takes no more
than a few minutes, and incurs only nominal costs.

This is, however, only a portion of the time and cost necessary to carry
out a successful simulation.  It assumes, for example, that a suitable
model is already  operational,  that the verification with  actual data has
been accomplished, and that all data needed to carry out the simulation
are available in the prescribed format and ready for use.  The validity
                                    15

-------
of these assumptions and their implications regarding costs are discussed
below.

Model Availability

A wide variety of operational water quality models are already available.
In addition, new and more accurate models are being developed and existing
ones are being extended to include the behavior of more water quality consti-
tuents.  As the library of mathematical models becomes larger, the assumption
that an appropriate model is available becomes more valid.  The use of exist-
ing models eliminates the costs of developing new ones or modifying the existing
ones, provided, of course, that the existing models can yield the required
information.  It is also conceivable that as planners are asked in the future
to produce even more detailed plans and to consider more alternatives, water
quality models will have to be constantly expanded and made more accurate.

Data Availability

A major consideration in the use of any model is the type, amount, and accuracy
of data needed to carry out reliable simulation experiments.  Data are needed
for calibration and verification of the model, as well as for the simulation
experiments (application).

Model calibration is considered to be a separate activity from model verifi-
cation.  Calibration is performed using one or more synchronous data sets
on model inputs (effluents, stream flow, etc.) and outputs (stream quality)
to adjust and tune the model itself (see Figure 2.1).  Verification is performed
using an independent set of input and output data to test the calibrated model.
Verification can occur only when the data set used is independent of that used
for calibration.  Preferably, the conditions for verification should differ
from those used for calibration.   One verifies a calibrated model by comparing
model predictions for a given input condition with the corresponding field
data.
                                     16

-------
MODEL INPUTS
EFFLUENT LOADS
STREAM FLOW, ETC.


MODEL
REACTION
COEFFICIENTS, ETC.


MODEL OUTPUTS
STREAM QUALITY
(DO, COLIFORMS, ETC.)
        FIGURE 2.1  MODEL RELATIONSHIP WITH INPUTS AND OUTPUTS

The collection of the field data for calibration and verification or the
assembly of historical data may be a costly and time consuming procedure.
For the application of a model to a particular system, the model normally
need only be verified a single time.  Once this is accomplished, repeated
simulations may be carried out at a relatively small extra cost.  Once a
verified model is available, data describing the state of the system are
necessary in order to carry out simulation experiments.  Sometimes these
data are available directly, or can be extracted from existing data that
are expressed in different form.  Sometimes they have to be inferred from
past data or determined from observations of the prototype, with or with-
out operational experiments.  Since operational experiments are costly and
for water systems often impossible to perform, much of the experimentation
and data collection has to be done indirectly at considerable cost and
effort.  This is especially true when the performance of planned systems
is to be predicted.

Model Complexity and Accuracy

The more complex and detailed a model is, the more data it can use to des-
cribe the system and the state of the quality constituents.  However, since
data collection and reduction is time consuming and costly, the planner must
carefully evaluate the tradeoff between the level of detail provided by a
particular model and the cost of the data necessary to fully use its capabilities.
Increased complexity may also have an adverse effect on the efficiency of the
computational techniques used in the simulation and may thus increase the
cost of model operation.  The introduction of additional detail does not
necessarily increase the accuracy of the simulation.  Although increased
                                     17

-------
accuracy does usually  imply  increased complexity, the reverse is not
true and a common fallacy  is to mistake  complexity for accuracy.  The
accuracy of the  simulation results depends on the accuracy of the model,
the accuracy of  the values of  the model  parameters determined during the
calibration process, and the accuracy of the input data.  One of the
common pitfalls  in using mathematical simulation models is to attribute
much greater accuracy  to the simulation  results than is warranted by the
data used to evaluate  the model parameters and verify the model.  Such
over-reliance on the simulation results may be counter-productive and
lead to misinterpretation of the data.  A final consideration is the
tradeoff between accuracy and  time.  A plan based on simulations that
yielded results  within known error bounds but available in time for use
in decision making may be far  more preferable than one based on a highly
accurate or complex model but  not available until a later time.

In spite of their limitations, computer simulation models have certain
distinct advantages over hand  calculation approaches.  The use of nomo-
graphs and hand  calculations is limited to relatively simple problems
because more complicated ones  would require an excessive amount of time.
Furthermore, these methods are not conducive to the study of a large
number of alternatives because the analysis of each alternative would
require an approximately equal effort.  The use of computer simulation
models, on the other hand, is  characterized by an extensive effort neces-
sary for carrying out  the first simulation while only marginal effort,
usually changing a few input values, is needed to carry out all other
simulations.  Therefore, numerous alternative plans can be evaluated
quickly and at low cost.  The  actual cost of machine time is sufficiently
low that this feature occurs whether the problem is complex and requires
extensive computation or is relatively simple.  This ease with which
alternative plans can be prepared, or "what-if" questions can be answered,
is one of the strongest arguments for investing time and effort to develop
or adapt, verify, and use computer simulation models.
                                  18

-------
The decision to use a computer simulation model is not the only alter-
native to the use of simple nomographs or other techniques designed
for hand computation.  The strengths and advantages of the computer
can be used in a wide range of activities between these extremes.   The
repetitious solution of even simple equations can warrant consideration
of computerization.  In those cases where emphasis is to be placed on
developing a continuing planning process for further use, even these
more limited models may have great value.

2.6   CONCLUSIONS

Mathematical simulation based on equations of physical processes is a
relatively new tool that has become available to the water quality plan-
ner.  Its usefulness depends on his ability to obtain sufficient data
that can be used to calibrate and verify the model and carry out the
desired simulation experiments.  The planner may have to decide whether
an existing model would be appropriate or whether a new one should be
developed.  He should estimate the cost and effort associated with the
use of computer models, make comparisons with all other suitable methods
of approach, and take into account the quality and extent of information
he can obtain from alternative approaches.  If he determines that a model
should be used, then his task is to determine which model is most appro-
priate for his planning problem.  Although the use of computer simula-
tions requires a highly organized approach to planning, it enables the
study of a range of alternatives that might otherwise be beyond the scope
of study by other methods due to cost.

While no clear-cut rule can be formulated regarding the use of simula-
tion models as planning tools, a few simple guidelines can be stated on
the basis of the analysis presented so far in this manual.  The guide-
lines should not be interpreted as strict rules, but rather as broad
statements that are meant to serve as reminders.  Their validity as
guidelines will probably be demonstrated by the frequency with which
they are violated.  The list is left open ended; additions, deletions,
                                     19

-------
or modifications will be necessary as experience is gained in the use

of models for planning.


      1.    The first step is to define the problem and determine
            both what information is needed and what questions need
            to be answered.

      2.    Use the simplest method that can provide the answers
            to your questions.

      3.    Use the simplest model that will yield adequate accuracy.

      4.    Do not try to fit the problem to a model, but select a
            model that fits the problem.

      5.    Do not confuse complexity with accuracy.

      6.    Always question whether increased accuracy is worth the
            increased effort and cost.

      7.    Do not forget the assumptions underlying the model used,
            and do not read more significance into the simulation
            results than is actually there.

The following two chapters provide more detailed discussions on whether
to use a mathematical model, how to evaluate existing water quality
models, and how to select a model for specific planning applications.
                                         20

-------
                            CHAPTER 3
                EVALUATION AND COMPARISON OF MODELS
3.1  INTRODUCTION

This chapter demonstrates a systematic way of evaluating water quality
models.  The evaluation procedure consists of answering many specific
questions about them;  the questions, and their answers, have been organized
for convenience into tabular form (Tables 3.1 through 3.14) within
this chapter.  The tables represent the condensation of large quantities of
descriptive text, enabling far more rapid information retrieval, and
greatly facilitating the comparison of different models.

This method of evaluation has been used directly in the procedures devel-
oped for model cost-effectiveness analysis and for final model selection
(see Chapter 4).  In some areas of evaluation, the answers are quite gen-
eral and dependent upon the application.  These areas must be looked into
carefully, but only used for evaluation on a relative scale.

The 14 models described in this handbook are all useful for the prediction
of water quality and provide a wide range of capability and applicability.
They seem to represent a large portion of the models expected to be in use
in the near future.  This does not imply that another model, not described
here, might not be preferable in a particular case.  Other models, not
selected for evaluation here, are discussed in Section 3.4 below.  In the
event models not included here are considered, the user is advised to
consider them carefully in the light of the evaluation procedure presented
in this chapter.

3.2  GENERAL DESCRIPTION OF SELECTED MODELS

The models selected for evaluation in this handbook are all deterministic
simulation models of varying complexity.  Their names and  the six groups
they have been arranged into, in accordance with their areas of applicability,
are:

                                21

-------
 Group I
 Group II
 Group III
 Group  IV
Group V
Group VI
 Steady-state Stream Models
     DOSAG-I
     SNOSCI
     Simplified Stream Models  (SSM)
 Steady-state Estuary Models
     ES001
     Simplified Estuary Models  (SEM)
 Quasi-dynamic Stream Models
     QUAL-I
     QUAL-II
 Dynamic Estuary and Stream Models
     Dynamic Estuary Model (DEM)
     Tidal Temperature Model (TTM)
     RECEIV
     SRMSCI
Dynamic Lake Models
     Deep Reservoir Model (DEM)
     LAKSCI
Near-field Models
     Outfall PLUME
The choice of these six groupings is governed by the model characteristics
summarized in Columns 1-3 of Table 3.1.

The models used to simulate only stream conditions are least complex due
to the one-dimensional characteristic of flow.  Models for simulating
stratified lakes and reservoirs fall next in line of complexity, followed
by estuarine models.  Estuary models are more complex because the prototype
flow is usually in at least two dimensions, and the boundary conditions,
such as tides, vary rapidly compared with those in lakes.  The costs of
model application tend to be proportional to their complexity.

With regard to the time domain, the models may be divided into the following
three major categories:  dynamic, dynamic-equilibrium, and steady-state.
                                22

-------
In the dynamic models, both the major inputs (drivers) and the outputs
(solution) may vary freely with time.  In dynamic-equilibrium models they
may vary only in a cyclically repetitious manner.  In steady-state models
they are unchanging with time.  The dynamic models are generally the most
expensive to implement, and the steady-state models are the least expensive.
These relative costs are due to the cost of data acquisition and manipula-
tion, as well as programmer/analyst and computer costs to run the models.

An intermediate time category of models (Group III of Tables 3.1-3.14)
have been named "quasi-dynamic," since only their weather (meteorological)
inputs may be dynamic.  These directly affect water temperature and
algae, and indirectly affect most other constituents.  The resulting solutions
have steady-state hydraulics, but dynamic water quality.

With regard to the space domain, the models summarized here fall into
three categories.  The outfall PLUME model is the only member of the near-
field category, in which very localized effects are simulated.  The one-
dimensional far-field category includes the stream, simplified estuary,
and deep* lake models; in the case of these lake models, the one dimension
is vertical, since the lake is simulated as a series of horizontal layers.
The quasi-two-dimensional far-field category consists of the dynamic
estuary models (Group IV of Tables 3.1-3.14), in which two horizontal
dimensions are simulated by a branched network or  system of one-
dimensional flow paths.

Two of these model categories  (Groups IV and V)  differ  from the  rest  in
their structure.  The  dynamic estuary, stream, and lake models are
composed  of definitely separated quantity  and quality submodels.  The
quantity  submodels first solve  for the hydrodynamics  (velocities, water
surface elevations,  etc).  These solutions are then provided  as  inputs
to the quality submodels, which use  the  then known hydrodynamics  to  solve
for the movements of  and changes in  the  quality  constituents.  One  variation
 For  definition  see Ref.  23,  page 16.

                                 23

-------
 upon this  theme  is  that  the  lake  models  must  include  the  temperature  solution
 in the  quantity  submodel,  since temperature is  the  principal  cause  of the
 stratification which  governs a lake's  behavior.

 It should  be noted  that  none of the models  evaluated  here are applicable
 to frozen  conditions.

 The documentation corresponding to most  of  the models listed  is, or soon
 will be, available  to  the  public  through government agencies.  A very
 nominal fee is usually charged for such  documentation.  At present  some
 of this documentation  probably does not  give  adequate directions to
 potential  users  (see Table 3.11).  In  such  cases it is necessary to secure
 the aid of experienced experts familiar  with  those models  to  help initiate
 the model  implementation.  In  any case,  depending upon the circumstances,
 the use of specialized and knowledgeable assistance could  greatly improve
 the efficiency of modeling efforts.

 The following sections review  the origins of  the selected models, and
 briefly compare  them within  each of the  six groups.

 Steady-state Stream Models (Group I)

 The  three models evaluated in  this category are DOSAG-I, SNOSCI, and the
 Simplified Stream Model.

 DOSAG-I is a computer nrogram which uses  the  classical Streeter-Phelps
 dissolved  oxygen sag equation  to simulate BOD and DO  variations.  It was
 prepared by the Texas Water Development  Board [1]* by improving a basic
 code originally developed, and provided  to them, by the Federal Water
Pollution Control Administration (subsequently the E.P.A.  Water Quality
Office).   It is particularly useful for  the rapid evaluation of a
number of varying stream conditions.
 Numbers in square brackets [ ] refer to the References listed in the
 back of this handbook.
                                 24

-------
SNOSGI is a modification of DOSAG-I, prepared by Systems Control,  Inc.,  and
first applied to the rivers of the Snohomish and Stillaguamish River basins
in Washington [2, 3].  The principal modification was the added capability
to simulate many more water quality constituents than just BOD and DO (see
Table 3.2).

The Simplified Stream Model (SSM) refers to those portions of the Simplified
Mathematical Modeling Methodology which pertain to non-tidal streams; it was
developed by Hydroscience, Inc. [4, 5] for the E.P.A. Water Programs Office.
Once the prospective user is familiar with the User Guide and the simplifying
assumptions made in their development, the various tables, charts, nomographs,
figures and technical data incorporated in the Simplified Mathematical Models
may be used to analyze water quality and to estimate treatment levels needed
to meet specific receiving water quality standards.  Normally only hand cal-
culations , with the help of a slide rule or possibly a technical desk cal-
culator, are required; no computer programs are included.

The Simplified Mathematical Models  (SSM and SEM) are intended to assist and
facilitate only interim planning, and should be used with discretion.  Com-
plex river systems or complex water quality problems, such as eutrophication,
are not covered by the simplified analysis.

Steady-state Estuary Models (Group  II)

The two models evaluated in this category are ES001  and  the  Simplified
Estuary Model.  Since the behavior  of estuaries is clearly not steady state,
but more cyclical due to the dominating influence of tides,  these models
simulate only the net flow or tidally averaged effects.  They do simulate,
however, longitudinal dispersion, which is generally negligible in  streams.

ES001 is a computer model which  simulates BOD and DO variations.  It was  pre-
pared by the EPA  [6, 7] tp improve  upon and document some water quality
                                      25

-------
models previously developed for them by Hydroscience, Inc.  [ 8 ]•  It is
particularly useful for the rapid evaluation of a number of varying estuary
and wasteload conditions.

The Simplified Estuary Model  (SEM) refers to those portions of the Simpli-
fied Mathematical Modeling Methodology [4, 5] which pertain to estuaries and
tidal rivers; it has the same features, purposes, and provisos as mentioned
above for  the SSM.

Quasi-dynamic Stream Models (Group III)

The two models evaluated in this category are QUAL-I and QUAL-II.  They are
both computer programs.

QUAL-I was developed during September 1969 - September 1970 by W. A. White,
R. J. Brandes, and Dr. F. D. Masch, in collaboration with the Texas Water
Development Board [9, 10].  It is more accurate and provides a more precise
definition of the stream conditions than does DOSAG-I described above  (in
Group I).  QUAL-I is designed to simulate the spatial and temporal variations
in water temperature and conservative mineral concentration as well as
BOD/DO, and is thus a more flexible model.  These added features are obtained
with a substantial increase in computational time.  The two programs are
designed to be used as complements to each other.  DOSAG-I can provide the
user with a rapid evaluation of a number of alternative conditions, whereas
QUAL-I permits a more detailed analysis of the meteorologically-affected
physical phenomena in the stream system.

QUAL-II is a modification of QUAL-I, prepared by Water Resources Engineers,
Inc., during June 1972 - May 1973 for the EPA Systems Development Branch
[11], and first applied to rivers of the Chattahoochee-Flint, Upper Missis-
sippi, Iowa and Cedar, and Santee River Basins.  The principal modification
was the added capability to simulate eight more water quality constituents
(see Table 3.2).
                                      26

-------
Dynamic Estuary and Stream Models (Group IV)

The four models evaluated in this category are the Dynamic Estuary Model
(DEM), the Tidal Temperature Model (TTM), RECEIV and SRMSCI.

The DEM was originally developed by Water Resources Engineers, Inc. (WRE)
for the Public Health Service, Division of Water Supply and Pollution Con-
trol  [12], and was then developed further for the FWPCA [13] and for the
State of California [14].  The FWQA completed its development and refinements
for use in studies of the San Francisco Bay-Delta estuary and the San Diego
Bay,  resulting in the FWQA version of the DEM [15] evaluated here.

The TTM [16, 17], also known as  the Columbia River estuary model, was
developed by the Pacific Northwest Water Laboratory of the FWPCA by incorp-
orating meteorological inputs and dynamic water temperature simulation  into
a  similar WRE version as that used to develop the DEM.

RECEIV is the name of the receiving water module  of the Storm Water Manage-
ment  Model  [18] developed by Metcalf ft  Eddy, Water Resources  Engineers, and
the University of Florida during 1969-1970.  RECEIV was developed, princi-
pally by WRE, by incorporating  into a previous dynamic equilibrium model the
capability  to simulate the  transient behavior  (toward a dynamic equilibrium)
and associated problems  caused  by dynamic storm water inflows.

SRMSCI is a modification of RECEIV, prepared by Systems Control,  Inc.,  and
first applied to the  Snohomish  and Stillaguamish  River estuaries  of Wash-
ington  [2,  3].  The principal modification  was  the  added  capability to
simulate many more water quality constituents  (see  Table  3.2).

None  of these  four models  are applicable to a  strongly  stratified estuary.
This  should be kept  in mind when modeling an estuary,  since it  is a common
occurrence  for an  estuary  to  be effectively mixed during  the  low  flow period
of the year and  stratified into two  distinct layers during the  high flow
period of  the  year.   The principal factors  contributing  to mixing are:   low
                                     27

-------
 river  flow, high  tidal velocity, large estuary width, and shallow estuary
 depth.   Several good  general descriptions of estuarine modeling techniques
 and  problems  are  available  [19, 20,  21].  It is recommended that the user of
 any  estuarine model familiarize himself at least briefly with  the general
 concepts involved before making any  decisions.

 While  all the estuary models may be  used to simulate streams,  the steeper
 the  stream the less appropriate and  convenient they become  (see, for example,
 Ref. 27, p. 26).

 All  these four estuary models utilize a chosen tidal cycle which repeats
 itself,  resulting in  a quantity (hydrodynamic) solution which  also repeats
 itself every  tidal period (dynamic equilibrium).  The DEM and  TTM are  truly
 dynamic  equilibrium models, since they accept only steady-state wasteload
 inputs.   However, RECEIV and SRMSCI  have been categorized as dynamic models,
 since  they accept transient inputs,  such as dynamic (non-steady and non-
 cyclic)  storm water inflow  (quantity and quality), resulting in a dynamic,
 transient solution which tends back  to the pre-storm dynamic equilibrium.

Another  difference between  these estuary models is that only RECEIV and
 SRMSCI can simulate tidal flats, i.e., areas which go dry at low tide.  DEM
and TTM  represent the water surface  area as remaining constant, which may
not be very appropriate for some estuaries which have extensive tidal  flats.

Differences of simulation method and convenience at the seaward boundary of
 the estuary models are that constituent concentrations in incoming tides
must be  specified for DEM and TTM, while RECEIV and SRMSCI compute them from
a specified ocean exchange or transfer coefficient.  This governs the
fractions  of  the  departing constituents which return, after dilution and
decay in  the  adjacent ocean.  These  alternative boundary conditions must be
specified  at  only one seaward boundary location for DEM, TTM, and RECEIV.
For SRMSCI, they may be specified differently at two locations, which is
convenient for a branched estuary with sloughs.
                                    28

-------
Dynamic Lake Models (Group V)

The two models evaluated in this category are the Deep Reservoir Model (DBM)
and LAKSCI.

The Dworshak Reservoir version [22] of the DRM was selected for evaluation
because it was determined to contain improvements over the original version,
and yet was not as site-specific as subsequent versions.  It simulates the
thermal behavior of deep, or strongly stratified [23, p. 16], impoundments.

A weakly stratified version of the same impoundment model [23, 24] was not
evaluated since it has been far more difficult to apply and verify on account
of the far greater amounts of data it needs.

DRM was first developed by WRE in 1967 for the California Department of Fish
and Game [25].  WRE added considerable detail and refinement in 1968-1969,
for the EPA Water Quality Office [23, 24].  They prepared the Dworshak
Reservoir version [22] by September 1969 for the Walla Walla District Office
of the Army Corps of Engineers.

LAKSCI is a modification of the DRM, prepared by Systems Control, Inc.,
during 1973-1974 for the EPA [26, 27].  It was developed for and first
applied to two lakes of the Spokane River Basin in Washington and Idaho.
The principal modification was the addition of a capability to simulate many
more water quality constituents than just water temperature.

Near-Field Models (Group VI)

The only model evaluated in this category is the Outfall PLUME model  [28],
developed in 1971 by the Pacific Northwest Laboratory of the EPA, Region X.
It is based on earlier ocean outfall design development work by the FWPCA
[29], and it solves for the geometric and dynamic behavior of a buoyant
round plume of sewage or industrial waste issuing from a port into stagnant,
density-stratified surroundings.
                                     29

-------
3.3  DETAILED MODEL EVALUATION


The model evaluation tables have been categorized according to their pur-

poses and contents as shown in Table 3.0.
                  TABLE 3.0  MODEL EVALUATION CATEGORIES
                   SUBJECT
                                                          TABLE
           MODEL CAPABILITIES

                Applicable  Situations
                Constituents Modeled
                Model Factors Accounted For

           DATA  REQUIRED

                For Model Inputs
                Additional, for Calibration and
                 Verification
           MODEL COSTS

                Initiation Costs
                Utilization Costs

           MODEL ACCURACY

               Representation
               Numerical Accuracy
               Sensitivity to Input Errors

          EASE OF APPLICATION

               Sufficiency of Available Documentation
               Output Form and Content
               Updateability of Data Decks
               Modification of Source Decks
 3.1
 3.2
 3.3
3.4

3.5
3.6
3.7
3.8
3.9
3.10
3.11
3.12
3.13
3.14
                                    30

-------
 The evaluation scheme and the detailed  evaluation of  the selected models
 is provided in Tables 3.1-3.14.   The answers provided in these  tables  are
 organized into a column for each question.   The columns  are  numbered for the
 convenience of reference in Chapter 4,  where their contents  are used in the
 model selection procedure.

 The following  explanations  of the contents  of these tables are  organized into
 numbered  Notes,  for  convenience  of reference in the tables.   The numerical portion
 of the note number identifies the table of  concern, and  the  alphabetical
 portion sequences the notes pertaining  to that  table.  Terms have not
 generally been defined in these  notes;  definitions  of terms  are provided
 in the Glossary  and  the Abbreviations,  at the end of  this handbook.

 Note  3.1.A;  Distributed Loads

 While  the documentation for the  Simplified  Stream and Estuary Models [4,5]
 describes  applications to  point  sources only,  the  technique may frequently
 be  extended  to uniformly distributed  loads,  such  as the  average effects of
 photosynthesis,  respiration,  benthal  oxygen  demand, and  nutrients released
 from decaying  organic  materials on the  bottom.  Such  an  extended technique
 is  discussed in  the ES001 documentation [6,  p.27].

 Note 3.1_.B;  Discretization

 For the purposes of simulation by  computer models,  the area  to  be simulated
must be broken up into  a number of  discrete  elements.  The means of this
 division, and  the naming of  these  elements,  varies  from model to model.  As
 a result, the  same names are  sometimes  given to different types  of elements
 in different models, and vice versa.  Here the element names  given in  the
original  documentation  (see Table  3.11) are  used.

The steady-state and quasi-dynamic  stream models  (Groups I and  III) are
representated by branched,  one-dimensional networks.  These  are  quite
similar in nature, no doubt on account  of the similarity in  their origins.
The streams are divided up  into series  of "reaches",  each having fairly
                                    31

-------
                                                                TABLE 3.1      MODEL SUMMARY:  HOTEL CAPABILITIES, APPLICABLE  SITUATIONS
bo
NJ
                                                                                                                        DISCRETIZATION LIMITATIONS1
                                                                                                                                   (Column 4)
                                                                                                                                          SPECIAL FEATURES
                                                                                                                                         AND/OR LIMITATIONS
                                                                                                                                              Column 5)
TIME VARIABILITY
   (Column 3)
CHARACTERISTICS
  (Column 2)
                                                             Far field. I.e., only longitudlna
                                                                                                                        SIO headwaters, S20 stretches.
                                                                                                                                                        Flov augmentation option
                                                                                                                                                        Both point and distributed
                                                                                                                                                        non-poini
                                                                                                    » total number of point sources In the model limits
                         See Ref.  23, p.  16  for definitions.
                         May conveniently repeat simulation,  when  aeteorological Input* are unchanged.
                         The repeat capability  (,e. Footnote  8) could  easily  be  restored  (add  tap. read for restart,  only)  by a programmer.

-------
                                         TABLE 3.2
                                                        MODEL SUMMARY:   MODEL CAPABILITIES, CONSTITUENTS MODELED
MODEL
I
II
III
IV



V

VI
DOSAG-I
SNOSCI
Simplified Stream
(SSM)
ES001
Simplified Estuary
(SEM)


QUAL-II
Dynamic Estuary
(DEM)
Tidal Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRft)
LAKSCI
Outfall PLUME
ORIGINAL MODEL
(Column 1)
DO, BOD (carbonaceous and nitrogenous).
DO, BOD, T. coli, F. coll, algae, NH3, N02, N03, OP04 , Cu,
Pb, temperature excess, four conservatives.
Conservatives, singular non-conservatives with first order
decay (coli, BOD, nutrients), coupled* BOD-DO deficit.
BOD, coupled BOD-DO deficit; both as non-conservatives with
first order kinetics. Uniform BOD/DO loads and demands of
algae and benthos.
Conservatives, singular non-conservatives with first order
decay (coli, BOD, nutrients), coupled with BOD-DO deficit.
BOD, DO, temperature and any 3 conservative constituents.
BOD, DO, temperature, NH, , NO , NO , algae, phosphorus,
benthic demand, coliforms, radioactive materials, 3
conservative constituents.
Up to 5 constituents with the following properties:
1) any can be conservative
2) any can have 1st order decay
3) any can be linked to one other-*
As above, with added feature that one of Che constituents
can be water temperature.
Any six constituents including DO, BOD, conservative con-
stituents, and non— conservative with 1st order decay. All
constituents must be in concentration units of mg/L.
Excess temperature, DO, BOD, T. coli, fecal coli, NH3, N02,
NO}, OPO^, Cu, Pb, and two conservatives.
Temperature.
The following Interlinked non-conservatives: Carbonaceous
BOD, coliforms, and three heavy metals with first order
kinetics; DO, algae, temperature; NH3, N02, NOj, and P04 with
first or second order kinetics. Also total N, chlorides, and
three heavy metal ions as conservatives.
Conservative substances only.
POSSIBLE, WITH MINOR MODIFICATIONS1
(Column 2)


Any 4 conservatives, and any 4 non-conservatives2 with first
order-* decay.


£
Conservatives; any non-conservatives or coupled non-
conservatives with first order kinetics.


(See Column 1)
(See Column 1)
(See Column 1)
(See Column 1)
(See Column 1)
4
Any 2 singular non-conservatives with 1st order decay, or 2
more conservatives, in place of Cu and Pb.


Other conservatives, and non-conservatives with first
order kinetics.
(See Column 1)
1.  Change of format statement listing constituent names.  See also Note 3.2.D of Section 3.3.
2.  See Note 3.2.A of Section 3.3, regarding mass conservation.
3.  See Note 3.2.C of Section 3.3, regarding reaction kinetics.
4.  See Note 3.2.B of Section 3.3, regarding constituent linkages.
5.  Linkage requires special attention.

-------
                                                 TABLE 3.3
                                                                MODEL  SUMMARY:   MODEL FACTORS  ACCOUNTED  FOR


I


II
___
III

IV



V


VI



DOSAG-I
SNOSCI
Simplified Stream
(SSM)
ESOO1
Simplified Estuary
(SEM)
=^===
QUAL-I
QUAL-II
=^ 	
Dynamic Estuary
(DEM)
Tidal Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)

LAKSCI
Outfall PLUME

PRINCIPAL
(Column 1)

Net flows.
Net velocity and
flow.

Net velocity and
flow.

Net flows.
Tides, net river
flow.
As above, plus
weather.
Tides, net river flow

Net heat exchange,


Discharge momentum,

BOUNDARY FACTORS
(Column 2)
Upstream rivers, point and non-point
waste loads.

River Inflows, single or multiple point
wasteload(s) .
River inflows, point and non-point
waste loads. Effects of completely
mixed bavs.
Point inflows, single or multiple point
wasteload(s) .
Upstream rivers and tributaries, point
wasteloads, bottom friction, weather.

Tides, constant upstream river flows and
concentrations. Constant wasteloads,
bottom friction, transient seaward bound
flow and concentration.
As above, plus heat exchange at surface.
Tides, constant or dynamic upstream
flows and concentrations. Constant or
dynamic effluent wasteloads, bottom
friction. Seaward bound flow and
concentration.2 Wind stress and rain.
Inflow rates and temperatures, outlet
positions and outflow rates, weather
(meteorology).
Inflow rates and quality, outlet
positions and outflow rates, weather
(meteorology) .
Water surface position, plume discharge.

INTERNAL
PHYSICAL
Dilution, advection.

11
Dilution, advection, longi-
tudinal dispersion.
"
Dilution, advection.

Dilution, advection, eddy
diffusion.
"
Dilution, advection.

Advection, diffusion, heat
budget.

Dilution, turbulent mixing,
density.
PROCESSES SIMULATED
DECAY AND/OR GROWTH
1st order decay, BOD-DO coupling,
1st and 2nd order decay, reaeration.
Linking of many constituents. Temp,
effects. Benthal releases & demands,
1st order decay, BOD-DO coupling,
As above, plus demands and releases
of algae and benthos.
First order decay, BOD-DO coupling,
BOD 1st order decay. BOD-DO
coupling, reaeration. Temperature
effects.
Numerous algae-nutrient inter-
actions, BOD-DO coupling, 1st order
decay, reaeration, temp, effects.
1st order decay, BOD-DO coupling,
reaeration for DO.
"
"
As above, plus 2nd order decay,
nutrient-algae cycle, and temp-
erature effects.


1st and 2nd order decay .reaeration;
linking of many constituents;
temperature effects; benthal


1.  See Note 3.3.A of Section 3.3 regarding temperature effects.
2.  Both the RECEIV and SRMSCI models compute the concentrations on incoming tides using a prescribed seaward exchange coefficient, and considering
    dilution processes outside the boundary.  R£CEIV models the seaward boundary concentrations as a constant value over the incoming tide   SRMSCI
    more realistically, computes a variable concentration on incoming tides.

-------
                                                                         TABLE 3.*
                                                                                       MOB. SOWARY: DATA UQDIIB),  FOR MODEL IHFOTS1
1 —
I
II
III
IV



V

IV
MODEL
DOSAG-I
S HOSC I
Simplified Strean
(SSH)
ES001
SlmplifUd Estuary
(SEM)
QUAL-I
QUAL-II
Dynamic Estuary
(DEM)
Tidal T cape r> Cure
(TtM)
RECEIV
SRMSCI
Deep Reaervolr
(DRM)
LAKSCI
Outfall FLUME
HYDROLOGIC
(Column 1)
Headwater flows, tribu-
tary and discharge flow*,
withdrawal f Iowa, ground-
water flow* (all con-
scant).

Net river flow.
Net river flow* (con-
stant), flow over damt
Net non-tidal flow.
Headwater flowav tribu-
tary flows, withdrawal
flows, groundwater flowav
(all constant).

AB above, plua (if in
estuary) the seaward
tides.

Conatant headwater flows;
constant or dynamic trib-
utary and effluent dia-
charge flows; conatant
groundvater flows; sea-
ward tides In estuary.
evaporation coefficients.

None.
HYDROOYNAMIC2 ,
(INCLUDING GEOMETRY )
Reach lengths.
Reach lengths.
Depths, velocities; dis-
tance from outfalls.
Csa's ; lengths; water
depth; bay volume*
Average water depths,
csa's, velocities.
Depths and widths, bottom
roughness (Manning's n)
all throughout stream.

Channel depths and widths,
bottom roughness, initial
velocities.


"
area relation, outlet ele-
vations; dam wldth(s) .
Initial and maximum water

Ambient water temperatures
at varloua deptha.
WATER QUALITY
(Column 3)
Constituent concentration
(constant) at headwaters
and tributaries; water
temperature.

As above, plua background
DO deficit.
As above plus bottom oxy-
coeff.. algal photosynthe-
• 1a »wrt i*ttan1rar1nn *
As SSM, plus dispersion
coefficient.
Constituent concentration
and tributaries.

Constituent concentrations
of tributary/river Inputs
(constant). Varying con-
stituent cone, at tidal
bound. Initial cone.
throughout- ^«fol»d area.

Constant or dynamic con-
stituent wasteloads from
tributaries & headwaters.
fl
Dally Inflow temperatures;
initial temperature con-
diffusion parameters.
Ditto, plus initial and
dally inflow constituent
concentrations.
None.
EFFLUENT
(Column 4)
Flow rates and con-
stituent concentra-
tions.
As above, plus cooling
water temperature rise
UOD loading rate.4
Uniform waste input*;

UOD loading rate. 4
Constant flow and




Constant or dynamic
constituent loads from
effluents.
"




P»rt diameter, density
of effluent, effluent
flow.
DECAY RATES
(Column 5)
Reaeration and two
deoxygenation coeffi-
cients, temperature

Rate coefficients.
temperature correction
factors.
Deoxygenation coeffi-
c lent s .
Rate coefficients,
factors.
Reaeration and deoxy-
genation coefficients.
BOD decay rate.
Numerous parameter
tllng coefficients.
rates or all non-
stituents.


1st and/or 2nd order
decay rates.
None.
Reaeration, decay and
f ic lent s ; cemperat ure
correction factors.
None.
OTHER
(Co loan 6}
Treatment factors.



Tidal exchange coefficient;

Salinities at boundaries*
Weather, lat-long of
oration coefficients.



Weather at specified inter-
wet & dry bulb temperature
Seaward exchange coef-
ficient. * Uindapeed and
rain.
"
Simulation dates; reservoir
elevation , latitude and
longitude; dry bulb air
temperature; wet bulb or dew
point temperature; short
wave solar radiation; sky
wind speed.
Number of discharge points,
horizontal and depth.
u>
Ui
            1.   All measurements must be for selected "simulation periods."
            2.   Channel bed conditions (Groups I-IV) may be used as a guide In selecting benthal  oxidation and reaeration coefficients  [Ref. 4, p. 62).
            3.   Data defined by the user in the discretization processes (such as reach and channel lengths and connection schemes) are not included here.
            4.   This may be obtained from charts, knowing the design population and the level  of  treatment.
            5.   Cross-sectional area.
            6.   See Note 3.4.a of Section 3.3 regarding seaward exchange coefficient.
            *   Optional, depending upon application.

-------
                                                   TABLE 3.5
                                                                  MODEL SUMMARY:  ADDITIONAL DATA REQUIRED,  FOR CALIBRATION AND VERIFICATION

I





LI



III



IV



V



VI

MODEL
DOSAG- I

SNOSCI

Simplified Strean
(SSM)
ESO01

Simplified Est-
uary (SEM)
QUAL-I

QUAL-II

Dynamic Estuary
(DEM)
Tidal Temperaturf
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DSM)
LAKSCI

Outfall PLUME

HYDROLOGIC
Streamflows within modeled area.

„







Streamflows within modeled area.

»

Net flows In channels of the
modeled area.
"
ti
"
Water surface elevation history.

"



HYDRODYNAMIC
Stream velocities within modeled
area.
„







Stream velocities within
modeled area.
"

"
"
"
It






WATER QUALITY
(Column 3)
Constituent concentrations
within modeled area.
u

»

"

«

1)

»

"
11
11
"
Time-varying temperature pro-
files and outflow temperatures
Time variations of constituent
concentrations , including
temperature, in the lake
profile and in the outflows.
Conservative constituent con-
centration within initial
Plume- 	 .
OTHER
(Column 4)






Salinity concentration dis-
tribution.
n





Salinity or dye data for
calibration of seaward
11
"
"






0V
            1.   All  these  measurements,  required to assess the accuracy of the model outputs,  must be taken during "simulation periods'"  (at  least  two  different
                periods,  representative  of different flow regimes).   See Note  3.5.A of  Section 3.3.
            2.   Because  of their  conservative nature,  salt and dye are often used as control constituents for the  determination of  the seaward  boundary  exchange
                coefficient.   See "ote  3.4.A of Section 3.3,  and Ref.  15 for results of dye studies.

-------
                                             TABLE 3.6
                                                            MODEL SUMMARY:  MODEL COSTS, INITIATION COSTS
1
MODEL
I


II

III

IV



V

DOSAG-I
SN'OSCI
Simplified Stream
(SSM)
ES001
Simplified Estuar
(SEM)
QUAL-I
QUAL-II
Dynamic Estuary
(DEM)
Tidal Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)
LAKSCI
• •• 	 !•
Outfall PLUME
COMPUTATION
MODE
(Column 1)
Computer
Computer
Hand
Computer
Hand
Computer
Computer
Computer
Computer
Computer
Computer
Computer
Computer
Computer
MODEL ACQUISITION
(Column 2)
Nominal cost, from EPA Planning
Assistance Branch, D.C.
Komiaal cost from Snonomish
County Planning Dept., Wa . , or
SCI, Calif.
Handbook from the EPA Planning
Assistance Branch
Nominal Cost, from EPA Region II,
N.V.
Handbook from the EPA Planning
Assistance Branch
Nominal cost, from EPA Planning
Assistance Branch, D.C.
"
11
EPA, Environmental Research
Laboratory, Corvalis, Oregon
tt
Nominal cost from Snohomish
County Planning Dept., Wash., or
SCI Calif
Nominal cost, from EPA Planning
Assistance Branch, D.C.
Nominal cost, from EPA Region X,
EPA, Environmental Research
Laboratory, Corvalis, '"gon
TYPE OF STAFF REQUIREMENTS
(Column 3)
At least one programmer or
engineer familiar with pro-
gramming .
ti
At least one junior level
engineer with some experience
in modeling.
At least one programmer or eng-
ineer familiar with programming
At least one junior level
engineer with some experience
in modeling.
At least one programmer or
engineer familiar with pro-
gramming .
"
At least one engineer experien-
ced in water quality modeling
and one experienced progra'.r-er ,
it
11

ii
11
One programmer.
EQUIPMENT REQUIREMENTS
(Column 4)
Requires any computer with -^27,000 word
storage and a FORTRAN IV (level C) compiler.
No tapes or disks are needed.

Hand calculator which computes logs and
exponentials would be helpful.
IBM 370 or equivalent.
Hand calculator which computes logs and
exponentials would be helpful.
Any computer with ^35,000 words of storage
and FORTRAN IV compiler. No tapes or disks.
As above, but with >45,000 words of storage.
Requires :
• ^50,000 words of storage
• Two tapes and/or disks
• Any FORTRAN IV compiler
11
As above, but FORTRAN IV (G level) compiler,

Any computer with > 35,000 words of storage
and a FOKTRAK IV comDiler.
Any computer with -. 50,000 words of storage
janH a FDRTRAV TU CC leVfel) compiler.
Any computer with ^ 10,000 words of storage
and a FORTRAN IV compiler.
1.   The ES002 model is ES001 modified to run on an IBM 1130 system.

-------
         MODEL
    I  DOSAG-I
   II
 III
  IV
      SNOSCI
      Simplified Stream
      (SSM)
      ES001
      Simplified Estuary
      (SEM)
      QUAL-I
      QUAL-H
      Dynamic  Estuary
      (DEM)
      Tidal Temperature
      (TTM)
      RECEIV
      SRMSCI
                                             TABLE 3.7      MODEL SUMMARY:  MODEL COSTS, UTILIZATION COSTS1
                                 MACHINE COSTS
                          $1 -  $5
                          $2 -  $10
                         Approximately $2 - $10
Somewhat more  than a comparable
DEM run if  temperature is
simulated.
                         Quantity $15 -  $100.
                         Quality $10 - $50.
                         similar to RECEIV with increase
                         of quality expense proportional
                         to increase of number of
                         constituents.
                                                            SET
                                                          (Column 2)
                                                         2-6 manweeks.
                                                         4-8 manweeks.
                                1-2 manweeks.
                                                         2-6 manweeks.
5-20 manweeks.
                                5-20 manweeks.
                                5-20 manweeks.
                                                                                              MANPOWER COSTS
                                                      RUNNING
                                                    (Column 3)
                                                                        Negligible.
                                                                        Negligible
                                               From a few man hours to
                                               several man days, de-
                                               pending upon application
                                                                        Negligible
                                                                        several man days, de-
                                                                                                Small - at most a few hours
                         Likely to be a little greater
                                                                                                As above,  except for runs of  nutrient-algae
                         than comparable QUAL-I run,
                                                                                                behaviour which require at  least several hours
Quantity run $20 - $300.
                                               Routine  runs take only
                                       The complexity of the model (particularly in the
                         Quality run $10 -  $100,
                                                                                                estuary) requires at least  several hours of
                                                                                                analysis and interpretation  for each run to be
                                                                                                Small - at most a few hours
                                                           ANALYSIS
                                                           (Column A)
                                                                                                Small,
                                                                                                Small,
                                       Small.
                                                                                                Small.
1.  For all  runs, including calibration, verification,  and subsequent use.

 '  SH-^JF:Ji$v^^
3.  Set up time for each »odel  depend, upon the complexity of application,  the form of available data, and staff capabilities.

-------
                                       TABLE 3.8      MODEL  SUMMARY:  MODEL  ACCURACY,  MODEL REPRESENTATION INACCURACIES1

I


II
— 1
III

IV



V

VI
MODEL
DOSAG-1
SNOSCI
Simplified Stream
(SSM)
ES001
Simplified Estuary
(SEH) 	
^^™"^™"*"^™^^***"™"
QUAL-I
QUAL-II
Dynamic Estuary
(DEM)
Tidal Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
DRMl
LAKSCI
utfall PLUME
SIMPLIFYING ASSUMPTIONS - QUANTITY
(Column 1)
Assumes stream velocity Is constant throughout a
reach (or section). No vertical and lateral
velocity variations.




11

velocity variation within channels, changes in
channel cross-sections with tides. Cannot
account for tidal flats which dry up at low tide.

As above, except can account for cross section
changes through tidal flat option.

Neglects horizontal velocities.

Assumes no ambient water flow.
SIMPLIFYING ASSUMPTIONS - QUALITY
(Column 2)
Assumes BOD 1st order decay.
Assumes 1st order decay, except NHo , NO-,; NOo
and POi may have 2nd order.
Assumes 1st order decay.
M

Assumes BOD 1st order decay. Uses empirical
evaporation equation.
Somewhat simplified nutrient-algae cycle. 1st
order decay only.
Assumes: 1st order decay, vertical homogeneity,
source immediately mixed throughout junction. -*

Assumes: 1st order decay, vertical homogeneity,
source immediately mixed throughout junction.
As above, except NH3, N02> NO, and PO, may have
st or 2nd order decay.
Neglects horizontal temperature variations.
Assumes horizontal homogeneity and 1st order de-
ay, except NHVN02, N03 & P04 may have 2nd orde
reats conservative constituents only.
OTHER ALGORITHM INACCURACIES
Errors in Table 1 of [1] .






Use of connected 1 dimen-
sional channels to simulate
2 dimensional flow and
transport.
"
ii
it



1.  Factors mentioned in this Table are limited to those considered to be o
2.  See Note 3.2.C in Section 3.3.
3.  See Footnote 1 of Table 3.1, and Note 3.1.B in Section 3.3.
f potential significance to model applications.

-------
                               TABLE 3.9
MODEL SUMMARY:  MODEL ACCURACY, NUMERICAL ACCURACY

I


II

III
IV
V
VI
MODEL
DOSAG-I
SNOSCI
Simplified Stream
(SSM)
ES001
Simplified Estuary
(SEM)
QUAL-I
QUAL-II
Dynamic Estuary
(DEM)
Tidal Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)
LAKSCI
Outfall PLUME
SOLUTION TECHNIQUE1
(Column 1)
Evaluation of integrated
equations (LaGrangian) , mass
balance .
ii

lution of analvtical eauation
Solve simultaneous equations
by matrix inversion.
Hand calculation or chart so-
lution of analytical equation
Finite difference implicit
solution.
1 "
1 Finite difference explicit
solution.
ti
1 ii
ii
Analytical and finite
difference solution.
"
Similarity solution.
STABILITY BEHAVIOR
(IF APPLICABLE)
(CoJ,unn^ 2)
N/A
N/A
N/A
N/A
N/A
i
Unconditionally
stable.
it
Conditionally
stable.
ii
it
ii
ii
ii
N/A
TIME STEP CONSTRAINT
FOR STABILITY
(C.nlitmrt TI
N/A
N/A
N/A
N/A
N/A
None.
None.
"
-------
                                         TABLE 3.10
MODEL SUMMARY:  MODEL ACCURACY, SENSITIVITY TO INPUT ERRORS*

I
II
III
IV



V

VI
MODEL
DOSAG-I
SNOSCI
Simplified Scream
(SSM)
ES001
Simplified Estuary
(SEM)
QUAL-I
QUAL-II
Dynamic Estuary
(DEM)
Tidal Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)
LAKSCI
Outfall PLUME
KNOWN SENSITIVITIES (Column 1)
HIGH


NH3 volitlzation
coefficient.






Depth, weather. d
it



NH3 volitallzation
coef f Icieat .


Time step size.
BOD decay
coefficient.

MEDIUM


Stream flow, decay
coefficients.






Streamf low, evapo- (1
ration coefficient.
source loads.
11
Quality time step size.
M
Quality time step size,
decav coefficients.

Eddy conductivity
coefficient.
As above, plus NHj
decay & volitization
coefficients, zinc
settling coefficients.

LOW









Initial conditions, (•
dust attenuation coef-
ficient, headwater temp
i friction coefficient.
it
Quantity tine step size
diffusion coefficient,
channel length.
ii




Reaeration coefficient.

ESTIMATED SENSITIVITIES (Column 2)
HIGH
Waste loads,
velocities.

"
Reaeration
coefficient.

Wasteloads, flow
velocities.
ii
Net flow, waste
loading rates.
As above, plus
depth. weather input
Ne t E low , wa s t e
loading rates.



Port depth, ambient
stratification,
effluent density,
number of ports.
MEDIUM
Streamf low t decay
coefficients.
Reaeration
coefficients.
Streamf low, decay
coefficients.
Deoxygenation coeffi-
cient; net velocity;
dispersion coefficient.
Strearaflow, decay
coefficients, waste-
loads velocities
Decay coefficients,
Streamflow.
As above, plus settling
coefficients .
Tides, decay coeffi-
cients, headwater
concentrations.
As above, plus evapo-
ration coefficient.
Tides, headwater
concentrations.
"


Port diameter, effluent
flow.
LOW





Dispersion
coefficient.
Depth, friction
and reaeration
coefficients
M
Friction coeffi-
cient, initial
conditions.
"
As above, plus
channel lengths.
M



1.  Known sensitivities for (JUAL-l are for temperature simulation only.
2.  For conditionally stable models (see Column 2 of Table 3.9), if the time step  size required  for stability  is exceeded, the  sensitivity may becone so
    high that model results are completely unreliable.

-------
                            TABLE 3.11
                                            MODEL SUMMARY:  EASE OF APPLICATION,  SUFFICIENCY OF AVAILABLE  DOCUMENTATION

I


II

ni
IV
V
VI
MODEL
DOSAG-I
SNOSCI
Simplified Stream
(SSM)
ES001
Simplified Estuary
(SEM)
QUAL-I
QUAL-II
Dynamic Estuary
(DEM)
Tidal Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)
LAKSCI
Outfall PLUME
AVAILABLE
DOCUMENT *
(Column 1)
[1]
[2, 3]
[*. 5]
[6. 7]
[4, 5]
[9, 10]
[9, 11]
[15]
[16. 17]
[18]
[2, 3]
[22, 23, 24]
[26, 27]
[28]
INDIVIDUAL DOCUMENT (S)
(Column 2)
Good theoretical background. Generally adequate for use by pro-
grammers and engineers.
[2] Adequate for use by engineers and programmers, but no pro-
gram listing.
[3] Good theoretical background.
[4] Adequate for use by engineers; but meager discussion para-
meter sensitivities and example applications.
[S] A helpful addendum.
[6] Good theoretical background. Generally adequate for use
by engineers and programmers, but poorly finished.
[7] Verification Report, incomplete and poorly organized
and finished.
[4] Adequate for use by engineers; but little discussion of
sensitivities and no example applications.
[5] A helpful addendum.
[9] Good theoretical background.
[10] Adequate for use by programmers.
[11] Adequate for programmers.
[11] & [9] together give good theoretical background.
Adequate for use by engineers And programmers not familiar with
the model's theory and use.
11
General program description. Very little detail.
Adequate for use by engineers and programmers not familiar
with the model's theory and use.
[22] Adequate but Imperfect User's Manual for the Dworshak
Reservoir version.
[23] Good theoretical background and description of model
studies*
f241 Adequate User's Manual for programmers (with [221).
[26] Adequate User's Manual for use by programmers.
[271 Good theoretical background, results of model applications
Reasonably good user documentation. Very little theoretical
discussion.
OVERALL SUMMARY
(Column 3)
Generally good.
Good.
Good.
Generally good; [7] provides little
additional help.
Good.
Good.
Good.
Good.
Good.
Poor.
Good.
Generally good.
Good.
Adequate.
1.  The numbers in brackets correspond to the References at the  back of  this  handbook.

-------
                                    TABLE 3.12      MODEL SUMMARY:   USE OF APPLICATION, OUTPUT FORM AND CONTENT
  III
   IV
   VI
            MODEL
         DOSAG-I
         SNOSCi
        Simplified Stream
        (SSM)

        E.SOU1
Simplified Estuary
(SEM)

QUAL-I
        QUAL-I1
        Dynamic  Estuary
        (DEM)
       Tidal  Temperature
       (TTM)
       RECEIV
       SRMSCI
        )eep Keservoir
        (DRW)
       LAKSCI
       Outfall PLUME
                                         OUTPUT KORN
                                          (Column 1)
                              Computer printout.
                              Computer  printout.
                      Hand  calculations.
                              Computer  printout.
                             Hand calculations.
                             Computer printout.
                             Computer printout.
                     Computer printout.   Channel  velo-
                     cities also written on tape  by
                     quantity model.
a)  magnetic tape
b)  computer printout
                            Computer printout.
                                                                                             OUTPUT CONTENT
                                                                                               (Column 2)
                                                                                    listing of input data; b)  DO, CBOD & NBOD concentratlone; at
                                                                     start and end of each reach,  plus size and location of minim'ura DO
                                                                     concentration in each reach.
                                                             DO deficit  and DO concentrations.
                                                                     a)  Listing  of  input  data  cards,  without  headings;  b)  BOD  t,  DO deficits  at
                                                                     one-tenth points  of  each  section.^
                                        Maximum DO deficit and minimum DO concentrations.

                                                          ===	
                                        a)  Constituent concent rat ions in eacli element at  specified timesteps.
                                        b)  Maximum, minimum and average concentrations durine simulation neri
                                                                    for each  reach.
                                                                                                      ncentrations during simulation period
                                                                    a) Constituent concentrations in each element at specified timesteps.
                                                                    b) Fina] constituent concentrations at end of simulation.
                                                                    c) Average reach coefficient used in simulation.	
                                        Stimma
                                        and
    lary of each tidal cycle including max. and min. flows, velocities, heads
    net flow; and max.,  min. and average concentrations at each junction   At
prescribed time intervals, also:  channel flows and velocities, junction
dentl.s. and constituent  concentrations.	
                                                            a)  Dates,  meteorology,  inflows and  outflows,  water  surface elevations,  out-
                                                            'low temperature objective.
                                                             )  See  Footnote  No.  2.
                                                                     ) As  DRM,  plus  daily  outflow  constituent  concentrations.
                                                                     ) See Footnote  No.  3.
                                                                    i)  Labeled  input  values.
                                                                    ')  Dilution  levels  along  the  effluent plume  centerline.
                                                                    _)  Elevation at which  plume stabilizes.	
1.  See Footnote No. 1 of Table 1.1, and Note 3.1.B in Section 3.3.

2'  loss 'evaporation rates "ar™ ^.^ .variaiions °{ «ater surface'and thermocline elevations, net  short and long-wave radiation  rates, net heat
    loss, evaporation rates, atmospheric vapor pressure, air and reservoir surface temperatures, outlet and downstream temperatures, depth
    variations of water temperature, (horizontal and vertical) flows, diffusion coefficient.
3.  As DRM (see Footnote No. 2 above),  plus daily constituent concentrations of the outflow, and depth variations of the constituent concentrations.

-------
TABLE 3.13
MODEL SUMMARY:  EASE OF APPLICATION, UPDATEABILITY OF DATA DECKS
MODEL
I


II
'"
IV
V
VI
DOSAG-I
SNOSCI
Simplified Stream
(SSM)
ES001
Simplified Estuary
(SEM)
QUAL-I
QUAL-II
Dynamic Estuary
(DEM)
Tidal Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)
LAKSCI
Outfall PLUME
CARD CHANGES
(Column 1)
Few.
Few.
None.
FFew.
None.
Very small, except for changes of
weather data which may involve
many cards.
ii
Few in most cases.
Small, except for weather inputs.
Small, except for transient waste
inputs.
it
Small, except for weather data.
Small, except for weather and
inflow quality data.
Minor changes, of at most a few
cards.
RECOMPUTATION
TIME
fr.nl iimn ?1
Very small.
Very small.
Relatively large.
Small.
Relatively large.
Small.
Small.
Small.
Small.
Small.
Small.
Small.
Small.
Very small.
HELPFULNESS OF
AVAILABLE DOCUMENTATION
(Column 1}
Good.
Good.
Good. Needs thorough study
and good understanding, before
using charts.
Good.
Good. Needs thorough study
and good understanding,
before using charts.
Good.
Good.
Good.
Good.
Poor.
Good.
Generally good. The three
comprising documents are less
convenient to use.
Good.
Adequate.

-------
                                              TABLE 3.14
                                                              MODEL SUMMARY:  EASE OF APPLICATION, MODIFICATION OF  SOURCE DECKS
MODEL
I
II
III
IV
V
VI
DOSAG-I
SNOSCI
Simplified Stream
(SSM)
ES001
Simplified Estuary
(SEM)
QUAL-I
QUAL- I I
Dynamic Estuary
(DEM)
Tidal Temperature
(TTM)
RECEIV
SKKSCl
Deep Reservoir
(DRM)
LAKSCI
Outfall PLUME
CODE LANGUAGE
(Column 1)
FORTRAN IV
FORTRAN IV
N/A
FORTRAN IV
N/A
FORTRAN IV (G level)
FORTRAN IV (C level)
FORTRAN II
FORTRAN IV
FORTRAN IV (C level)
FORTRAN IV (G level)
FORTRAN IV
FORTRAN IV
FORTRAN IV
HELPFUL
COMMENT
STATEMENTS
(Column 2)
Many
Many
N/A
Few
N/A
Many
Many
Some
Some
Some
Some
Many
Many
Some
ARZ THESE HELPFUL SUBROUTINES?
(Column 3)
Many
Many
N/A
Many
N./A
Many
Many
Some, but the main programs are
still very long and complex.
"
Some, but could be more.
»
Many
Many
No, but they are not needed in
such a short program.
DOES DOCUMENTATION HELP IN
PROGRAM MODIFICATION?
(Column 4)
Yes
Yes
Yes1
Yes, generally adequate.
Yes1
Yes
Yes
Yes
Yes
Very little.
Yes
Yes
Yes
Little
Ui
             1.   Helps  to modify the procedure rather  than the program.

-------
 uniform conditions within it.   Different streams  join at  "junctions", where
 three or more reaches meet.   The most upstream reach on each branch  is  a
 "headwater",  where considerable inflows may occur if the  smaller,  upstream
 streams are not being simulated.  The steady-state (Group I) stream
 models have one additional,  unique category of element, named "stretches,"
 which consist of groups  of adjacent reaches between headwaters and
 junctions.

 ES001,  the  only computer model  in Group II:  Steady-state  Estuary Models,
 is  comprised  of "sections" and  "junctions".  These  sections are similar to
 the reaches of  the Groups  I  and III models;  the junctions  differ from
 Groups  I  and  III  type, but instead  are  points  of  connection of the sections.
 ES001 offers  a  choice  of four types of  section and  14  types of junctions,
 including the effects  of dams and  completely mixed  bays.

 The dynamic estuary and  stream  models  (Group IV)  represent water bodies by
 a grid  or branched network of one-dimensional  flow  paths called "channels";
 at  the  ends of  all channels, where  they  are connected  together, are "junctions",
 These junctions are of a third  type, differing from those  in the Group I,
 II  &  III models in that  these contain between  them the entire volume of
 water in the  simulated system.   The volumes and levels of water in each
 junction are  computed, together with  the flow  rates and velocities in the
 channels.  The  channels  are  similar to  the  Group  II sections, except that for
 the channels  the  flow rates  and velocities  are not  prescribed but  instead
 are computed  from the  channel characteristics.

 The dynamic lake  models  (Group  V)  represent  impoundments by a series of
 fixed,  horizontal layers or  volumes,  through which  water  and constituents
 pass.   Only the top layer  may have  a  changing  thickness,  to account for water
 surface level changes.

In  summary, all the names  of the discrete elements  included in the models
evaluated here have unique meanings, except for the term "junction", which
has the three different meanings described above.   More details of the
discretization scheme employed by each model are provided in the documenta-
tion  (see Table 3.11).

                                  46

-------
Note 3.2.A;  Mass Conservation

Water quality constituents may be divided into two broad categories, conserva-
tive and non-conservative.  With conservative constituents, the total mass
or quantity of the constituent is conserved, even though its concentration
may change as a result of dilution.  Non-conservative constituents experience
an actual loss or gain in total constituent quantity (in addition to
change in concentration), as a result of such processes as decay, consumption,
settling out, growth, and death.

Note 3^2.B; Consj^ituent Linkages

The presence of many water quality constituents in receiving waters may have
effects on the behavior of other constituents, in a variety of ways and to
various extents.  For example, decaying organic materials consume
oxygen, and toxic materials may kill or slow the growth rates of aquatic
plants and animals.

Singular constituents are those not affected by the presence of other
constituents.

Coupled  constituents are defined as those whose behavior  is affected  (in
each case) by the presence of only one other constituent.

Linked constituents are  defined as those whose behavior is  affected by
the presence of  one or more  other  constituents.   Thus,  coupling  is  a  special
case of  linking.

Note 3.2.C:  Reaction Kinetics

The most common  type of  change  of  total  constituent mass  (see Note 3.2.A)
is that  in which the rate of change with time  (t)  is proportional to  the
amount  (m) present, namely

-------
K Is known as the rate or decay coefficient; negative values of K represent
growth processes.  This equation represents what is known as first order
kinetics.  It may be integrated to yield the exponential relationship

                      ~Kt
          m(t) = m(o)e

where m(o) is the initial condition.

Occasionally, the behavior of some constituents are found to be better
represented by second order kinetics, the corresponding equation being
          dt
Thus the order of the kinetics is seen to be the power to which m is raised
in the right-hand side of the equation.

Note  3.2.D;  Modifications  to Constituents  Simulated
                    

Table 3.2  indicates  thats for certain  models,  the  list of  constituents
which may  be modeled can be somewhat changed by making minor modifications  to
the computer programs.  For the  examples  cited, only minor modifications
are required; namely, changes in the computer  code governing the printing
of  the constituent  names, a relatively simple  task.

The list of alternative constituents could  clearly be extended by making
more  extensive changes to the code.  However,  these are not mentioned in
Table 3.2, since  they depend upon the  availability of far  more advanced
expertise  and time  to deal  correctly with the  effects of constituent
interlinkages (see Note 3.2.B) and changes  in  units.

Note  3.3.A;  Temperature Effects

The rates  of most kinetic reactions are significantly affected by temperature.
This  is generally accounted for  in water  quality models by making the rate co-
                                   48

-------
efficient K, defined in Note 3.2.C above,  a function of temperature.   This
function is typically of the form
where       T = water temperature,  C
            8 = a constant, commonly in the range 1.02 - 1.12

These relationships are the temperature effects referred to in Table 3.3.

Note 3. 4. A;  Seaward Exchange Coefficient

This coefficient (also known as the ocean exchange or transfer coefficient)
has been incorporated into many estuary models as a simple alternative to
specifying all constituent concentrations in incoming tides at the seaward
boundary.  It prescribes the fraction of the (computed) departing consti-
tuents which return, after dilution and decay in the adjacent ocean.  While
simple in concept, it should be used with some caution, since the amount of
discharged water which returns is strongly dependent upon  the tidal wave
form.  Further, this fraction should really be a time-varying function,  since
its value for returning water will strictly depend upon the length of time it
has been outside the boundary.

Note 3. 5. A;  Calibration and Verification

A model is considered to be calibrated and verified when the uncertain or
unknown values of  the various model parameters have been adjusted until  the
model predictions  correspond acceptably closely to the observed prototype
behavior.

Calibration is the first state of parameter adjustment, with a  first  set of
prototype  input and output data for a  first simulation period.   Verification
then involves modelling at least  one different simulation  period  (with a
different  set of input and output data) using  the originally calibrated

-------
parameters.  If the model predictions for  the subsequent simulation
period(s) do not also agree sufficiently closely with the corresponding
prototype behavior, then the model is not  yet verified and further calibration,
or recalibration, is necessary.  It is important that the various simulation
periods used for verification should correspond to significantly different
prototype conditions.

It should be remembered by the calibrator  that, when all calibration attempts
seem unable to attain the required accuracy of predictions, the input and/or
output data may contain errors and at such times they must be reviewed
accordingly, or the model may be inappropriate for the physical situation
being simulated.

Note 3.7.A;  Cost Impact of Discretization and Complexity

Since the cost of computer runs obviously increases with the number of segments
the model is divided into, the number of time steps the simulation period
is divided into (where appropriate),  and the number of constituents simulated,
the run cost for a given problem will vary across the range reported in
Table 3.7 depending upon the level of discretization and complexity specified
by the user.

Here, the number of segments would be the  number of reaches (model Groups I &
III), sections (ES001 model), junctions/channels (Group IV), or layers
(Group V) which comprised the model (see also Note 3.1.B).

Note 3.9.A:  Mathematical Terms

Although the mathematical terms and qualities mentioned in Table 3.9
probably have more meaning to analysts familiar with advanced mathematics
and numerical methods, they can indicate to the planner factors which should
be considered.  The fact that various models employ different solution techniques
and/or constraints does not necessarily imply any kind of superiority or
                                   50

-------
inferiority; the most appropriate technique will vary with the model
structure.  It is here presumed that the model designers and builders
considered such matters when developing their models, and incorporated the
most appropriate methods into them.  However, these qualities of the models
may serve the planner as a guide to their suitability for unusual or special
applications.

3.4 OTHER MODELS

The models evaluated in this handbook are all deterministic simulation
models.  There exist many other models different from those selected for this
evaluation, as well as other versions of those evaluated here.  Several of
these, which were evaluated in a previous study by the authors [30], along
with others too recent to have been evaluated in this study are listed in
Table 3.15.  It is expected that much the same questions would be used in
the evaluation of these other models, and that only different answers
would be obtained in some areas.

One area of modeling notably different from  those evaluated herein  is that
of ecologic modeling of receiving waters [31,32; see also 30], in which
the life forms are of prime interest.  Another different area is that of
the truly two-dimensional models [33,34], in which velocity components
are determined in two perpendicular directions over a grid of points  covering
the water body.  These two-dimensional models are more  complex than any
evaluated in this handbook.

Three-dimensional models are still more  complex.  Their development
is being approached on a number  of fronts, but presently  they are far
from being  ready for wide use  in planning.   One  of  these  approaches is  the
quasi-three-dimensional model, an  example of which  is the Puget  Sound
Model presently undre development  by Water Resources Engineers,  Inc., for
the EPA.  It will represent  the  Puget  Sound  as  three interconnected layers,
each  layer  being simulated by  a  quasi-two-dimensional model.
                                    51

-------
                                                        Table 3.15
                                                                       OTHER MODELS
Ui
                    MODEL
                 Hydrocomp
                 Simulation
                 Model
                 Estuary
                 Eco logic
                 Model
                 (ECOMOD)
                 Lake
                 Ecologic
                 Model
                 ( LAKF.CO)
                 RECEIV-II
                 WRECF.v
                                     AREA OF
                                 APPLICABILITY
Lake,
5tream
Stream,
Estuary
Lake
Stream,
Estuary
S tream,
Estuary
                                                       PARAMETERS MODELED
Temperature, BQt), conforms,
algae, zooplankton, sediment,
organic nitrogen, PO, IDS,
nutrients and conservative
constituents.
Zooplankton, benthlc animals,
fish, pH, nutrients, conser-
vative constituents, non-
conservative constituents
with 1st order decay.
Zooplankton, benthlc animals,
fish, pH, nutrients, conser-
vative constituents, non-
conservative constituents
with 1st order decay.
BOT), coliforns, nutrients, TX>,
salinity, conservative con-
stituents, non-conservative
constituents with  1st order
decay,  chlorophyll-a.
DO-BOD (linked), any four
conservative or 1st order
non-conservative.
This model is particularly useful
in its capability to predict water
runoff and the resulting stream
quality as a function of varying
weather conditions. The quantity
portion is well established; how-
ever, the quality part has been
applied to only 2 streams.
Hodels estuarlne systems by linking
One-dlmenslonal channels in a two-
dimensional fashion. For this reason
the schematiKation process requires
expert guidance. ECOMOP has had
only limited use.  Simulates tidal
flats.
LAKECO Is a derivative of the Heep
"eservotr Model. Most useful In
simulating well-stratified lakes
and reservoirs. Has been used success-
fully on numerous  prototypes.
 A modified  version  of  the EPA SWM»*
 PF.rTF.V-IT model.  Handles multiple
 tidal  Inlets,  upstreams  dans. Well
 calibrated  and tested.
 Compatible with EPA  Stormwater
 Management Model  (SWTM). Well
 calibrated and tested,
                                                                                                  AVAILABILITY
Proprietary: Available from:
  Hydrocomp, International
  1502 Page Mill Road
  Palo Alto, California 90430
At Nominal Cost From:
  National Technical Infor-
  mation Service
  5285 Port Royal Road
  Springfield, Va.  22151
At Nominal Cost Iron:
  National Technical Infor-
  mation Service
  5285 Port Royal Road
  Springteld, Va.  22151
 At nominal cost  from:
   EPA,  Planning  Assistance
    Branch
   Washington,  T\C.  21460
                                                                                                                             'aytheon
                                                                                                                             P.O.  Box 360
At nominal cost fro*i:
  EPA, Planning Assistance
   Branch
  Washington, P.C. 2046"

             or

  Water resources  Engineers
   3445 Executive Tenter Dr.
   "edlna Bide, Suit*  22T
   Austin, Texas  78731

-------
  The deterministic  models  discussed  thus  far  are  all  "event"  models,  with
  which a short  or relatively  short simulation period  is modeled.   There  is
  a growing  awareness  of  the limited  capability of  such models to provide
  sufficient information  to enable the long  term probabilistic effects, and
  consequently benefits,  of alternative management  schemes  to  be determined
  The  use  of "continuous" deterministic models  [35,36  (see  also  [30]), 37-41]
  whose  output can be  statistically analyzed,  is recommended for such  purposes,

  While  considerably more input data  is required for continuous simulation,
  the  types  of data required are identical with those required for event
 models, and they are mostly available in the U.S. in computer compatible
 form from  the National Weather Service and the U.S. Geological Survey.   But
 the much longer record (typically about a year's) Of water quality data
 needed for calibration and verification,  and the far longer computer run
 times, still remain as disadvantages to continuous modeling.   The resulting
 increased costs may be justified in  certain cases where the planned facil-
 ities are large and costly.

 An advantage of continuous models is that their  long  simulatlon periods
 prevent "false  calibration" by the adjustment of  the  initial  conditions
 as may be possible  with  the simulation  of shorter events  (up  to 2  or  3 weeks
 long,  depending on  the flow regime). On  the  other hand, when there are
 insufficient records  at  the site of  interest,  such as for  streamflows, the
 uncertainty introduced by  extending  or synthesizing records for  continuous
 simulation  may  equal  that obtained with far less  labor by  using  event models.

 Stochastic  models [e.g., 42] provide an alternative approach  to  the question
 of probabilities.  They attempt to take into  account the randomness in many
 observed phenomena.   They exist in great variety,  depending upon the
 assumptions  they make  about the physical processes, and upon  the type of
mathematics  they employ.  They also  have the disadvantage  of  requiring
very large  quantities  of prototype data in order to establish the  various
probabilities.  As a consequence, most of the  simulations  models in present
use are deterministic.
                                      53

-------
                               CHAPTER 4
                MODEL SELECTION AND COST EFFECTIVENESS
                           EVALUATION SYSTEM

4.1  INTRODUCTION

The process of selecting a water quality model for any  wastewater management
planning project may involve numerous complex considerations.   One of the
objectives of this project was to identify the important factors which in-
fluence the selection of models by planners, and to structure  the consideration
of these factors in a manner such that confusion is minimized.   This chapter
presents a methodology which can be used for model selection and cost effective-
ness evaluation, and detailed instructions on how to use it.  The methodology
does not tell planners what decisions on models to make, but provides them
with the essential questions and thought structure upon which  they or their
contractors can make the decisions.  If a consultant is used,  the procedure
also gives guidance to planners as to the types of evaluations to request
and expect from consultants.

The model selection process is designed to give users a choice of several
levels of detail they may want to consider.  The process is divided into
four phases, each going into progressively more detail and requiring pro-
gressively more effort.  These phases are:

     Phase I:  Model Applicability Tests
     Phase II:  Cost Constraint Tests
     Phase III:  Performance Index Rating - Simplified
     Phase IV:  Performance Index Rating - Advanced

The rejection of candidate models in one phase reduces the number of models to
be evaluated in the next phase.  The phases are designed accordingly.  All con-
siderations in the selection process are based upon model evaluations of candi-
date models as presented in Chapter 3.
                                     54

-------
 The amount of effort required to select an appropriate model depends upon
 how many phases the planner uses and how well he understands the problem
 at hand.  Once the planning problems are clearly defined and modeling objec-
 tives identified, model selection based upon Phase I considerations alone
 should require only a few hours of planner effort.   Completion of additional
 phases may require a substantial addition in effort.  Planners may estimate
 the amount of effort required for each phase by examining the appropriate
 evaluations described in Chapter 3 which are the basis of final model selec-
 tion.   These model evaluations are presented in Tables 3.1-3.14 of Chapter  3
 and are associated with the Phases as listed below:

      Phase I - Tables 3.1,  3.2,  3.3.C1-C2? 3.4,  3.5
      Phase II - Tables 3.6,  3.7
      Phase III - Tables 3.1,  3.2,  3.3.C1-C2,  3.4,  3.5,  3.6.C3
      Phase IV - Tables 3.3.C3-C4,  3.8,  3.9,  3.10,  3.11,  3.12,  3.13,  3.14

 The various  required evaluations also guide  the  planner  as  to  whether he  should
 consider the use and selection of  a water  quality model  in  the overall plan-
 ning program.   In most cases  the model  selection will need  to  be  preceded by
 the problem  identification  and data  inventory portions of a planning program.
 Any monitoring that  can be accomplished within the program  should  be post-
 poned  until  after  the  models  are selected.

 4.2  DECISION  TO USE A MODEL

 Prior  to model selection, the  more basic decision whether any water  quality
 model  should be used must be made.  This decision must take  into account
 whether a model would  be helpful in plan formulation and whether a suitable
 model  and sufficient data are  available.  Generally, water quality models
 are useful in any area where the quantitative relationship between varying
wasteloads and resulting water quality must be known.  This relationship
between wasteloads and receiving water quality will be of prime importance
 in all "Water Quality Limited" areas of the country  [43].  In  "Effluent
Limited" areas, waste treatment alternatives will often be fixed by  Federal
* The number 3.3.C1-C2 refers to Table 3.3, columns 1 and 2.
                                      55

-------
  Effluent Standards,  thus eliminating the necessity of a water quality model.
  The recently completed studies under Section 303e of PL 92-500 specify
  all "Water Quality Limited"  areas of the country.   If the  quantitative relation-
  ships  between water  quality  and varying waste-loads are needed  by a planner,
  he  should go ahead with the  model selection  process presented in this chapter.
  If  water quality modeling is inappropriate or  inefficient  in  a  particular plan-
  ning application,  the  model  selection process  will  make this  fact apparent,
  as  the candidate models will not  pass various  applicability and constraint tests.
  In  this case the planner must consider  other quantitative  analysis techniques
  such as data acquisition and interpretation.

  4.3  SELECTION OF  CANDIDATE  MODELS

  A set  of  candidate models must be identified before the model selection pro-
  cess is initiated.  Those models evaluated in Chapter 3 could be used as
  the  candidate  set.  However,  many other models are available which dould be
  considered.  Most available water quality models can be located at the following
 agencies:

      •    Environmental Protection Agency, Planning Assistance Branch  (WH-454)
          Waterside Mall, Washington,  D.  C.  20460.
      •    Army Corps of Engineers,  Methods Branch,  609 Second  Street, Davis,
          California,  95616.
      •    U.S.  Geological Survey,  12201 Sunrise  Valley Drive, Reston, Virginia,
          22092.
      •    State  Water Quality  Planning Offices.
      •   Colleges and Universities dealing with water  quality  problems.

The planner should  select his  set of candidate models using as many sources
as possible.  Since many model titles describe  the type of receiving water
they are applicable to, the planner should review the titles and prescreen
those obviously not applicable to his particular problem.  In some cases
planners will be interested in several types of receiving waters, such as
lakes, rivers and/or esturaries, and they should deal with them separately in
the model selection process.
                                   56

-------
4.4  MODEL SELECTION PROCESS

The model selection and cost effectiveness evaluation procedure is structured
to avoid unnecessary effort by screening out inappropriate models  early in
the process.  The evaluations of candidate models,  as shown in Chapter 3,
should be accomplished concurrently with the appropriate tests and ratings in
each phase.  For example, the evaluations of various models' abilities to
handle water body types and characteristics (see Table 3.1, Columns 1-2)
should be performed for all models while performing the water body applica-
bility tests in Phase I.  As models are rejected in various applicability  and
constraint  tests, later evaluations of these rejected models can be avoided.

The following sections give detailed instructions on how to select a water
quality model.  A  flowchart is presented for each phase so that the user
can keep  track of where he  is in the model  selection process.  The more phases
the planner uses in  the model selection,  the more confidence he can have  in
his selected model.   In many cases, however, adequate confidence in the model
selection can be attained  after the first one or two phases of the process
are completed.

4.5   PHASE I - APPLICABILITY TESTS

The first phase  of the selection  process is important:,  because many of the pre-
 selected models  can be rejected at this stage.   The applicability tests are
 intended to ask basic questions about the appropriateness of  the  models with
 respect to the problem at hand.   Those models  that are inappropriate  for  the
 case at hand are then rejected from further study.   A schematic flow  chart
 illustrating the methodology used by the planner during Phase I is presented
 in Figure 4.1.  The user may take all candidate models through the steps of
 a phase simultaneously, or one at a time, but he should not proceed to any
 subsequent phase until all aspects of the preceeding phase are complete.

 Tables 4.2-4.5 (presented at end of this Chapter) present work sheets upon
 which decisions on the various models can be recorded.  Similar tables should
 be used  for record keeping of the applicability tests  and  subsequent  Phase
 III performance index ratings.  Any candidate models should be listed on the

                                     57

-------
Ui
00
                       C
                           START
                  Perform "water body"
                  evaluations  (3.1.C1-C2)
                  for  each model (see Ch.  3)
                           Pass
                        water body'
                     applicability
                     test (4.2.1)?
                        each model)
                                   Pass
                                "discreti-
                            zation &  other fe
                            ures" applicabili
                               est?(4.2.3Xe
                                     el)
Perform "tim
evaluations
e variability
(3.1.C3) for
each remaining model


                           Pass
                       time varia-
                    bility  appllcabill
                      test  (4.2.2)?
                         each mode
                             Yes
                 Perform "discretization
                 and other features"
                 evaluations (3.1.C4-C5)
                 for each remaining model
                         Perform "constituents
                         modeled" evaluations
                         (3.2.C1-C2) for each
                         remaining model
                                  Pass
                               "constitu-
                             ents modeled
                            applicability test
                               4.3.1)? (ea
                                    el)
                         Review data availability
                         with respect to evalua-
                         tions (3.4.C1-C6 and
                          3.5.C1-C4)
                                                                              10
                        Perform data availability
                        evaluations (3.4.C1-C6
                        and 3.5.C1-C4)
                                                                   T
                                                        "etemlne effect of Hata
                                                        limitations on previous
                                                        model applicability tests
                                                                                                                      12
Estimate extent of monitor-
Ing, if any, which can be
accomplished within pro-
ject to reduce effects of
data limitations.  Do for
each model not yet rejectee
                            End of Phase I.  Remain
                           ing models are those
                           appropriate for particu-
                              lar application
      ,   Pass   ^
  ^    applicabilit
  tests  4.2.1-3 and
4.3.1  based upon data
    pliability (bxs
          l, 12)'   '
                                                                                                rerform driving forces and
                                                                                                boundary factors evalua-
                                                                                                tions  (3.3.C1-C2)
                                                                                                 Reevaluate modeling
                                                                                                 objectives and/or select
                                                                                                 new candidate models
                FIGURE 4.1
FLOWCHART OF PROCEDURE FOR  MODEL SELECTION AND  COST  EFFECTIVENESS  EVALUATION-
PHASE I - APPLICABILITY  TESTS

-------
left side of these work sheets-.   The particular models included  here in
Tables 4.2-4.5 for the purposes  of an illustrative example are those evalu-
ated in Chapter 3.  Results of applicability tests should be recorded as  "yes"
for applicable models  and  "no"   for non-applicable models in the appropriate
columns as shown in Tables 4.2-4.5.  As the various tests proceed, rejected
models should be crossed out or  deleted from subsequent work sheets to avoid
unneeded work performing tests and ratings of previously rejected models.

Type of Water Body

The ability of a model to simulate the behavior of the correct type of water
body is of prime importance.  Most model documentation reports provide infor-
mation on the types of water bodies they are capable of simulating and have
previously been applied to.  For a cursory analysis, the user can simply con-
sider what the model documentation says and base his applicability test and
performance rating accordingly.   However, for a deeper understanding of the
problem the following analysis should be made.  First, the user should deter-
mine the spatial dimensions for which variability of water quality or flow
are important.  Consideration must first be given to the scale of interest,
i.e., far-field.  Most existing water quality standards are specified in terms
of concentrations outside an initial mixing zone, thus requiring a look at the
large-scale, far-field effects of water pollution.

Next, the extent  of concentration  and/or flow variability  in the prescribed
field  (i.e., scale) of concern must be considered.   If a parameter is fairly
constant in a certain dimension,  then  it can be averaged over that dimension
without excessive error in  the analysis.  This  is demonstrated  in  the far-field
view of a river.  In this case,  pollutant particles  are  usually quite evenly
distributed over  both vertical and lateral  directions  (i.e.,  over  the cross-
section).  Only in  the longitudinal  direction  do  significant  concentration
variations exist.   Similarly, in shallow, vertically well-mixed estuaries
large  concentration and flow variations may occur in any horizontal direction,
but not  in the vertical direction.   Conversely,  in strongly stratified  lakes
variations are primarily  important in  the vertical direction and not in the
horizontal directions.

                                        59

-------
  Time  Variability

  The basic  concern  in  this  test  is  to determine which model variables, if any,
  must  be  considered as time-varying.  For this the user should determine: 1)
  the type of  simulation period he wants to use, i.e., short-term low flow,
  storm flow periods, long-term;  and 2) which of the inputs, particularly flow
  effluent loads and weather, are known to vary significantly over the simu-
  lation period.

  Models should be evaluated for  applicability and effectiveness in this test
  based upon their ability to handle the appropriate time variations.  Many
  water quality models  solve for  time-varying concentrations, but only account
  for one  or two time-varying inputs.  For many so-called dynamic models,  efflu-
  ent loads can only be accepted as constant inputs.   It should be remembered
  that time varying models can be used to obtain steady-state solutions,  but
  steady-state models cannot be used to solve for  time-varying solutions.

 Descretization and Special Features

 The limitations on the number of spatial  grids, or  segments,  in those models
 presently available is not a problem in most  applications.  However, as  three-
 dimensional models  are developed and  applied,  limitations  in  computer storage
 will become a major factor.   One way  of solving this problem  is  to  divide  the
 water  body  into sections  to be modeled separately,  and  then to  connect
 or  mesh the water quality results  at  the  boundaries.  For  the models
 evaluated in  Chapter 3, and most other models  presently being used  by
 planners, spatial gtid limitations  do not represent major  constraints.
There are certain special features of the prototype receiving waters which
may be important in water quality simulations.  These may include the presence
of tidal flats, flow augmentation sources, storm loadings, etc.  In most situ-
ations these features are not vital, but they can be of useful assistance.  If
a special feature is considered vital,  the applicability test should be used
to reject inappropriate models.

                                    60

-------
Constituents Modeled

A list of water quality constituents of interest to the planner  should be
made.  This may include constituents showing existing or potential  problems
based upon historical water quality data and knowledge of present and future
point and non-point pollutant sources.  It is possible that several of  the con-
stituents which are desired for analysis are not found in any candidate models.
For example, color and odor are two parameters which are found in most  water
quality standards, but are not included in models due to limitations in the
state of the art of water quality modeling.  For this reason, planners must
compare their desired list of constituents with those offered in candidate
models.

Some models can be used for a whole class of constituents given by
a certain  type of kinetic reaction, e.g., first order decay.  This does
not  apply  to  coupled  constituents  such as dissolved oxygen and algae,
since their nature  is unique.
 Candidate models  which are capable  of  simulating  the appropriate  constituents
 pass the applicability test and  can be given Phase  III  ratings based  on  their
 relative adequacy in modeling those constituents.

 Model Input Data

 Another of the most important factors  to be accounted  for in model selection
 is the availability of data for model  inputs.   The ability for  a  model to
 simulate system characteristics is  limited by the quality of input data.

 Input data limitations will likely limit the applicability of the candidate
 models.  For example, a model accounting for transient effluent loads may
 be more appropriate in a certain case than a steady-state model,  but also
 more expensive.  If, at the present time, only single measurements of the
 effluents have been made,  the transient input mode of  the model  can not be
 accurately exercised.  In  this  case,  the transient model would  probably be
 used with constant effluent inputs, thus negating  its  increased  usefulness.

                                     61

-------
Two further input data considerations must be made with respect to the scope
of model studies:

     1)  Can additional data be acquired within the project?
     2)  Will the model be used over a number of years (i.e., the next 20
         years) in which the needed additional data will probably become
         available, or is the model only to be used at the present time,
         eliminating the possibility that future data improvements will be
         helpful?

If the scope of a planning/modeling project can accomodate a monitoring program,
the planner must determine the amount of data needed for each model which can
be obtained.  This data should then be included in the data evaluations
(evaluations 3.4.C1-C6), and model applicability tests.

Reliance on uncertain future data acquisition programs should not affect the
applicability screening tests, but data from these programs may influence
the applicability of the models in the future.  Therefore, the usefulness
of future data programs should be accounted for in the Phase III applica-
bility ratings.

Input data ratings should be based upon the general accuracy that the various
inputs can be specified.  Input data with poor accuracy should be rated low.

These considerations for data inputs may show that the data base is completely
insufficient for all candidate models.   In this case, models are probably not
useful and other, less rigorous forms of analysis such as data acquisition and
interpretation should be proposed.

Driving Forces and Boundary Factors

All important driving forces which tend to move pollutants in the real water
system should be listed.  These may simply be flows and velocities or, where
these are unknown, bed slope (gravitation), tides (gravitation), wind,
and density currents due to water temperature gradients.   The planner should
be careful in selecting only those driving forces which are important in the

                                    62

-------
movement of pollutants in the subject receiving waters.   Then  he may  compare
his list with those driving forces offered by the various models.

Also to be included in this test is the consideration of all receiving water
boundary conditions affecting the concentrations of pollutants.   These may
include:

     Headwater Inflows
     Tributary Inflows
     Goundwater Inflows
     Slope of Bed
     Bottom  Friction
     Water Depth  (Hydraulic Radius)
     Point Effluent Discharges
     Non-Point Effluent  Discharges  (Including Benthal Exchanges)
     Weather (Heat Budget, Wind,  etc.)
     Water Withdrawals

 The need for including  these  boundary factors  as time varying inputs  should
 also be determined.   This requires  consideration of  time variability  for  each
 significant  boundary  factor.

 The model applicability tests in this category should account for  each model's
 ability to handle the significant boundary conditions and driving  forces, and
 should be entered into a table similar to Table 4.3.

 Summary of Phase I

 The tests in this first phase of the model selection process are intended
 to compare hard constraints of the particular user application with the
 various characteristics of candidate models.  Those models clearly not meeting
 a user constraint  should be rejected.  However, if the  tests show that  a model
 is marginally applicable, then it may be maintained for  further consideration
 in  Phase III, where  the level  of applicability  is given a rating.

                                     63

-------
 At the end of Phase I the user has the option of Continuing the selection
 process with Phase II,  or selecting a model based only on Phase I considerations.
 If all the candidate models are rejected in phase it then the user must either
 find new candidate models,  re-evaluate his applicability constraints, or aban-
 don attempts to use a water quality model for his planning needs.

 4.6  PHASE II - COST ESTIMATION

 This phase presents a costing  system which can be used in model selection.
 Both elapsed project time and  dollar cost are considered as cost items which
 for each model,  must be compared with user constraints.   Those models that
 require too much project time,  or  far too many funds,  should be rejected at
 this point.   The dollar costs  estimated  in this section  are used in final
 cost effectiveness comparisons.  A task  flow diagram of  all planner activi-
 ties in Phase II is given in Figure  4.2.

 Tables 4.6 and 4.7 (presented  at the end  of  this Chapter)  show sample work
 sheets to be used  for time  and  cost  estimation and constraint  tests.   These
 work sheets  should be used  to  keep records of  the various  costs  for final
 effectiveness  evaluations performed  in Phase  III or  IV.

 Model  Acquisition  Costs

 Most of the  existing water quality models are  in the public domain. They
 should require only nominal materials and shipping costs, and a few weeks
 time for  acquisition.  Some models are privately owned, however, and can
 be used only under a lease or purchase agreement.  Acquisition costs may in-
 clude  a surcharge on each run made, i.e., a certain percent of the computer
 charges.  Documentation for these programs may also cost the user.  The total
 cost of acquisition, including any use surcharge should be estimated according
 to the expected amount of use.

Equipment Requirements

Consideration of equipment requirements should be based upon comparisons of
available capabilities with any hardware requirements.   This may include slide

                                    64

-------
                      20
Determine  constraints of
time and funds available
for water  quality modeling
          i
  21
Perform "cost" evaluations
(3.6.C1-C4 and 3.7.C1-C4)
on remaining models
                      22
Estimate  required time for
each model in 4.6.1-3 and
A.7.2 testa
          i
  23
Estimate  required cost for
each model in 4.6.1-3 and
4.7.1-2 tests
                                       Total time - time  (4.6.1)
                                       + time (4.6.3)  +
                                       time (4.7.2)
Total time - time (4.6.2)
+ time (4.7.2)
                                   Is
                               total time
                              time constraint
                              from box 20
                                                                                                          30
                                                              End of Phase II.  Re-
                                                           maining models are those
                                                           ^suitable which pass  time
                                                              d cost constraints
                                            Add costs estimated in
                                            box 23  to derive total
                                            cost,  for each remaining
                                            model
    FIGURE 4.2
FLOWCHART OF PROCEDURE  FOR MODEL  SELECTION AND COST EFFECTIVENESS  EVALUATION:
PHASE  II -  COST AND TIME CONSTRAINT  TESTS

-------
 rules and desk calculators as well as high speed digital computers.  If
 digital computers are needed, additional consideration must be given to
 requirements of keypunching equipment, card readers, tapes, output printers
 and other devices.  Use of large computer facilities does not necessarily
 require its physical proximity.   Many of the services needed are available
 through use of remote terminals  that may be more conveniently located.   Both
 time and cost of equipment acquisition should be included in the cost esti-
 mates.

 Data Acquisition Costs

 As a result  of  the input and  calibration data considerations discussed  in
 Phase I, monitoring programs  may  optionally be considered as part  of  each
 candidate  model  application,  and  accordingly,  should be  included in cost
 estimates.   Both the time and  cost  for implementing  these monitoring  pro-
 grams must be estimated  by the planner (see  [44], for assistance in these
 cost estimates).

 Machine Costs

All machine charges  for computation during the course of  the project  should
be estimated under this category.   These charges may be for computation
time, data storage on tapes or disks, data transfer, and output display.
Estimates can be made by determining the approximate cost of a single com-
puter run.   Model documentation reports generally provide information on
computer run times for various computers.  If the documentation does not
provide this information, then past users of the models should be contacted
for information on their run times or computer costs.  For a particular model
the computation time is about  proportional to:

     1)   The number of constituents modeled,
     2)   The number of segments (i.e.,  reaches,  junctions) modeled, and
     3)   The  number of time steps  used  (for  time varying  models).
                                   66

-------
Thus, projections of computer run times can be made using information on  the
numbers of constituents,  segments,  and time steps corresponding  to past model
run times.  From these considerations and the expected number  of runs, the
user can determine the approximate machine costs for each candidate  model.

Manpower Costs

The number of personnel and level of expertise needed in a modeling  project
is another important consideration.  The ultimate usefulness of  models  is
often limited by the ability of personnel to properly implement  the  models
and analyze the results.

The specification of needed staff for implementation of each model should
be based upon the level of model complexity, the amount of time needed  to
implement the model, and the time span allocated for the project.  Planners
may choose to use contractors for certain modeling tasks.  A detailed dis-
cussion of the advantages and disadvantages of this alternative is given in
Chapter 5.

Manpower  time for modeling must include  the times  to:

     1)   Obtain and/or train personnel for the job;
     2)   Set up the models for implementation  including  the assembling of
          input data;
     3)   Run the model and analyze  the results  for each  run.

Approximate manpower  costs can be  obtained using the  manpower times and  the
project  salaries  for  the various personnel.

Cost Constraint Tests

In this  phase  of  the  selection process,  the total estimated costs can be
obtained.  The  total  time  required for a candidate modeling project, not
                                     67

-------
 including reporting time, can be estimated from the following (refer to
 Tables 4.6 and 4.7).

      a)  If  T(4.6.1)  + T(4.6.3)  > T(4.6.2)
          Then Total Time (T)  = T(4.6.1)  + T(4.6.3)  + T(4.7.2)
      b)  If  T(4.6.1)  + T(4.6.3)  < T(4.6.2)
          Then Total Time (T)  = T(4.6.2)  + T(4.7.2)

 Model and data acquisition,  and acquiring and  setting up  the required  equip-
 ment can be accomplished simultaneously.

 The total cost required for a candidate  modeling project  is simply  the total
 of all the incremental costs

      Total Cost (C)  =  C(4.6.1)  +  C(4.6.2)  + C(4.6.3)  + C(4.7.1) + C(4.7.2)

 Since the estimates  of cost and time for  use of the various models will usually
 be very approximate, some discretion should be used in rejecting models.  For
 example,  only  those models whose  cost estimates grossly exceed the modeling
 budget  should  be rejected in  this phase.

 4.7   PHASE III  - PERFORMANCE  INDEX RATING:  SIMPLIFIED

 This  portion of the model selection process gives a method for estimating the
 effectiveness  of the candidate models.  The effectiveness is obtained  through
 a  "Performance  Index Rating," which is divided into two parts.   The first,
 "simplified" part accounts for  the more basic, and usually more important
 model attributes which have previously been discussed  in the Phase I and Phase
 II  tests.  The second  "advanced" part of Performance  Index (PI)  rating, per-
 formed  in Phase IV involves much more detailed, and usually somewhat less im-
 portant considerations of the models.   Since these additional considerations
 require a large addition of work for the planner,  he is not discouraged to make
his final model selection based only upon considerations through  the simplified
PI rating (Phase III).   In most cases,  this will give the planner a very good
idea of which model is best for his particular planning problem.   A brief

                                    68

-------
review of the contents of the second part (Phase IV) will then usually
indicate whether those further considerations are necessary.

A task flow diagram of all planner activities in Phase III is given in Figure
4.3.  Work sheets for the simplified performance index ratings are provided
in Tables 4.2-4.5 and 4.7.

Performance Index System

The estimated performances of various candidate models in a particular appli-
cation are quantified in terms of certain model attributes.   The attributes
to be examined in Phase III are given in various columns of Tables 4.2-4.5
and 4.7.  As shown in these tables, each attribute can be assigned a rating
and weight.  The attribute ratings are based upon the expected model capa-
bilities given a certain user application.  For each attribute the ratings
are given as follows:

     10 - excellent
      8 - good
      6 - fair
      4 - poor
      2 - very poor
      0 - completely inadequate.

The weights are used to adjust the impact of each attribute rating on the
overall Performance Index, based upon their relative importance.  For example,
if the "Time Variability" capability of a model is much more important than
the "Constituents Modeled" capability, then it should have a larger weight.
Attribute weights are specific to each application.  They must be assigned
by the planner based upon his judgement of the importance of each attribute
to his planning problem.  In assigning weights the most significant factor
is the relative importance of the various attributes.  For a particular appli-
cation a single set of weights should be used for all candidate models.  Further
details on the assigning of weights are given in Section 4.9.
                                    69

-------
o
                                                                   ough
                                                               information for
                                                          rating of constituents
                                                             modeled (4.3.1)?
                    Specify 4.2.1  rating
                                                           Specify 4.3.1 rating
            Enough
No   ,^'infomiation for
     rating of tine varla
       bllity (4.2.2)?
       Enough.
    information for
rating of driving forces
   d boundary tactors
      (4.3.2)7
                    Specify 4.2.2 rating
                                             Specify 4.3.2  rating
                           ough
                      information for
                   rating of discretization
                     nd other features
                         (4.3.3)?
                                                   Enough
                                               Information for
                                           rating of overall data
                                                   (4.5.3)?
                    Specify 4.3.3 rating
                                                           Specify 4.5.3 rating
                                                                            No
                                               Enough
                                           information for
                                        rating of manpower
                                               (4.7.2)?
                                                                                    Specify 4.7.2 rating
         Is
    Simplified PI
 rating sufficient
for  model selection?
                                     Assign  relative importance
                                     weights to the rated
                                     attributes
                                                i
                                                                                                                      48
                                     Compute the Simplified
                                     performance index (PI)
                                     rat ing
                                                                                                                     49
                                                                                List PI ratings and estimated
                                                                                coats from box 24 and select
                                                                                model on basis of comparison
                                                                                                          i
                                                                                                     (    END     J
            FIGURE  4.3  FLOWCHART OF PROCEDURE FOR MODEL  SELECTION  AND COST EFFECTIVENESS EVALUATION:
                           PHASE  III -  PERFORMANCE INDEX RATING,  SIMPLIFIED

-------
The overall performance index of the jth model can be computed using the
equation:
FI(J)
                        n
                           [Rating (i,j)]  [Weight (i)]

                                n
                                    Weight (i)
      where i = the attribute numbers
            n = number of attributes considered

The ratings for each attribute, based upon the appropriate model evaluations
(Chapter 3), must be specified by the user.  Ratings are determined using the
same considerations as those in the Phase I, model applicability tests.  This
includes limitations on applicability due to expected data limitations.  As
stated above, the ratings are a function of how well a model can represent
the prototype system.  They are given as either excellent, good, fair, poor,
very poor and completely inadequate.

In some cases there may not be enough information available to assign a rating.
If this is true and the attribute is of relatively minor importance, that
attribute can be completely left out of the PI rating for the model in question.
Then, that attribute will have no effect on the rating.  If the evaluator feels
that an attribute is of major importance, he should attempt to obtain the nec-
essary information on it.

Applicability Ratings

Ratings in this category include those for consideration of water body (4.2.1,
Table 4.2), time variability (4.2.2.), discretization and other features (4.2.3),
constituents modeled (4.3.1, Table 4.3), and driving forces and boundary factors
(4.3.2).   For these ratings the same subjects should be considered as in the
applicability tests discussed  in Phase  I.  This includes  limitations
on model applicability due  to  data  insufficiencies.  The  user  should
enter these ratings in the  appropriate  columns of  tables  similar  to
Tables 4.2 and 4.3.
                                   71

-------
 Data  Rating

 This  rating  should be based on the overall quality and accuracy of data needed
 for input  and  calibration of the candidate models.  The rating should not in-
 clude effects  on model applicability due to limitations in the data base, as
 these effects  are accounted for in the applicability ratings.  The quality of
 all needed input and calibration data included in tests on the Table 4.4 and
 4.5 work sheets must be accounted for.  This may be done by rating each of
 the categories of data needs on a scale of 0-10, and entering the ratings
 into  Tables  4.4 and 4.5.

 Generally, model results are only as accurate as the least accurate inputs.
 Thus,  for  overall rating of input data, the lowest rated category should be
 used.  For example, if the input water quality data (test 4.4.2, Table 4.4)
 have  the lowest rating, then that rating should be used to represent the
 quality of all input data.

 The quality  and accuracy of calibration and verification data should be used
 to alter the input data rating, thus providing an overall data rating (Table
 4.5).  If  the  calibration and verification data are excellent, they may be
 used  to tune some of the weak inputs, and the overall data rating may be im-
 proved from  the input data rating.   Conversely, if the calibration and verifi-
 cation data  are very poor, the overall data rating should be reduced to account
 for the limited ability to assure accurate model results.

Manpower Rating

The capability of the manpower proposed to operate the candidate models and
analyze their results is an important factor in the model PI ratings.   For
most project staffs this ability will be excellent in the case of very simple
models.  As models become more complex the manpower rating (see Table 4.7,
test 4.7.2) of a given staff should be reduced.   The extent of this reduction
depends upon the staff capability.
                                      72

-------
 In assigning  the manpower rating consideration should be given to the amount
 of time  available  for staff training, and the level of expertise of possible
 contractors.  Chapter 5 gives detailed discussions of these subjects.  This
 rating for  each model should be placed in the appropriate column of Table 4.7.

 Summary  of  Phase III

 This  section  presents the first part of the PI rating system.  If the planner
 decides  not to perform the more advanced, second part given in Phase IV, he
 should then select his attribute weights and make his final model selection
 as discussed  in Sections 4.9 and 4.10 below.  Otherwise, he should continue
 with  Phase  IV as presented in the next section.

 4.8  PHASE  IV - PERFORMANCE INDEX RATING; ADVANCED

 The final phase of the model selection and cost effectiveness evaluation pro-
 cess  is  provided for more intensive probing into details of various candidate
 models.  Most of the ratings in this phase require extensive user insight
 and experience in water quality analysis and modeling.  For this reason it
 is  included as optional in the overall selection process.  A task flow diagram
 of  all planner activities in Phase IV is given in Figure 4.4.  Table 4.8
 shows the work sheet for the Phase IV ratings.  All of the ratings can be
 based on the appropriate evaluations given in Chapter 3.  The guidelines for
 deriving ratings and weights are the same as those for Phase III.

 Internal Factors

 The purpose of this rating is to determine which models have the capability
 of handling appropriate internal factors including; dilution, advection,
 diffusion, biological decay,  settling, and reaeration.  To assign this rating
 the user must determine which internal processes have significant effect on
his water quality problem.   Then he should perform the pertinent evaluations
 shown in Chapter 3 (Table 3.3)  for each candidate model.  Ratings should be
based upon a comparison of modeled internal process with those expected to
be  significant in the receiving water system,  and should be entered into
Table 4.8.

                                      73

-------
                          50
   Perform "internal processes"
   evaluations (3.3.C3-C4)
  If enough information,
  specify 4.8.1 rating
  Perform "model representation
  accuracy" evaluations
  (3.8.C1-C3)
  If enough  information,
  specify 4.8.2 rating
I
  Perform "numerical accuracy"
  evaluations (3.9.C1-C4)
 If enough information,
 specify 4.8.3 rating
 Perform "available documenta-
 tion" evaluations (3.11.C1-C3
I
                        „
 If enough Information,
 specify 4.8.4 rating
                                                                        58
                                  Perform "output form and
                                  content" evaluations
                                  (3.12.C1-C2)
                                                           I

                                  If enough information,
                                  specify 4.8.5  rating
                                              I
                                                                       60
                                  Perform "updateability"
                                  evaluations  (3.13.C1-C4)
                                                           I

                                  If enough  information,
                                  specify  4.8.6 rating
                                              I
                                                                       „
                                  Perform "ease of modifica-
                                  tion"  evaluations
                                  (3.14.C1-C4)

                                  If  enough  information,
                                  specify  4.8.7 rating
                                                            *
                                  Assign  importance weights to
                                  all  rated attributes in
                                  Phases  III and IV
                                                            1
                                                          „
                                  Compute final performance
                                  index ratings
                                              I
                                                                       66
                                               List PI  ratings and estimated
                                               costs, and  select model on
                                               basis of comparisons
                                                           I
                                                      f    END    "N
FIGURE  4.4   FLOWCHART  OF PROCEDURE FOR MODEL  SELECTION  AND  COST
               EFFECTIVENESS EVALUATION:   PHASE  IV -  PERFORMANCE
               INDEX RATING, ADVANCED
                                       74

-------
 Model Representation Accuracy

 Model algorithms generally represent simplifications of events occurring
 in the real water system.  These algorithms are usually in the form of mathe-
 matical equations which are arrived at under certain simplifying assumptions.
 Therefore, the simplifying assumptions used to derive model algorithms
 should be identified using available documentation and program listings,
 and their possible impact on overall model accuracy assessed.

 The planner must rate each model on the basis of its ability to represent the
 actual water quality processes that it attempts to account for.  Thus,  he
 should examine each process modeled (generally found in theoretical descrip-
 tions in the model documentation),  determine the simplifying assumptions used
 in representing these processes,  and approximate the overall effect on the
 model outputs (water quality predictions)  due to these simplifying assumptions.
 This  is  obviously not an easy task  and requires much insight into water quality
 processes  and the candidate models.
 Numerical  Accuracy
Another  important  factor which determines  the accuracy of a mathematical model
is  its numerical characteristics.  This is especially true of the so-called
finite difference  methods which rely on numerical approximations of differential
equations to generate solutions.  Models that solve analytical equations with-
out these numerical approximations have no numerical problems, and therefore
can be given an excellent rating in this category.  For the other models two
main factors must  be considered in determination of numerical accuracy; stabil-
ity and numerical  dispersion (mathematically known as discretization error
caused by the use  of large spatial grid sizes).

Stability  Models which are unconditionally stable are generally safer to use
than others which  can develop huge errors under certain conditions of time-step
size.   Models which are conditionally stable can be successfully used, but more care
must be taken in space step (Ax)  and time step (At)  selection to assure reason-
                                   75

-------
 able  results.  A  typical stability constraint is of the form

                           Ax
     At must be less than .

     V = A characteristic velocity, such as the tidal wave velocity.

As  the desired spatial grid becomes smaller, the timestep must be made smaller
to  ensure preservation of stability in conditionally stable models.  Reasonable
costs must also be assessed, since computational costs are usually proportional
to  timestep smallness and the number of spatial grids.

Numerical Dispersion.  The other factor, numerical dispersion, is very elusive
and difficult to quantify.  It is an undesirable quality brought about by solv-
ing equations in discretized grid form.  In general, it is difficult to control
and creates inaccuracies in model results.  The extent of this problem varies
greatly between models and even between applications of a single model.  Models
which, from historical experience, have had significant problems with numerical
dispersion should be rated very low in this category, if only due to the un-
predictable accuracy of the results.

Both stability and numerical dispersion are difficult factors to precisely
quantify.  For those planning staffs not familiar with these terms, it is
wise to steer away from models which have been shown through any literature
to  have numerical stability or dispersion problems, or where information is
not available to fully explain the numerical accuracy of past applications.

Ratings for numerical accuracy should be based upon the above considerations
and entered into Table 4.8.

Sufficienty of Available Documentation

This and the following three ratings are for consideration of the relative ease
of applications.   The quality of available documentation can be very impor-
tant in the successful set up and implementation of water quality models.
                                    76

-------
 Excellent documentation would  posses  all  of  the  following:

      1)   Detailed  theoretical  description;

      2)   Discussion  of the numerical  characteristics and parameter
          sensitivities;

      3)   Detailed  discussion of how to set up the model including grid lay-
          out  (descretization);

      4)   Program description including flowcharts;

      5)   Card by card description of model inputs including the assumed
          units;

      6)   Description of how to use model options, such as a flow augmentation
          option;

      7)   Detailed  definitions of all variables used in the models;

      8)   Presentation and description of the model outputs;

      9)   Discussion of calibration and verification;
    10)   Program listings;

    11)   References.


The ratings in this category should account for the documentation coverage of
each of these areas,  and should be entered into Table 4.8.

Output Form and Content


The form and content  of water quality model outputs can be of special impor-
tance to the planner.  Some typical output forms are:

     •   Printed tables with headings,

     •   Plots of water quality and/or flow velocities,

     •   Magnetic tape or  disk  file storage,

     •   Long strings of numbers representing concentrations,  but without
         headings.

                                    77

-------
The  first  three of  these forms can each be particular benefit for certain
applications.  The  last form is usually difficult to work with, requiring
added analysis time, and may still lead to added analysis error.  Generally,
outputs  should be well structured and labled in order to minimize user con-
fusion.

The  primary output  of water quality models is a description of water quality
distribution in the subject receiving waters.  The content of this output
may  include:

     •   Statistical characteristics of water quality over simulated time
         periods,
     •   Water quality at each time step,
     •   Water quality only at end of simulation,
     •   Combinations of the above.

Other output contents which may be of importance are: printed input data (help-
ful  in checking the input files); flows and velocities computed in the model;
and  kinetic coefficients computed in the model.

The  ratings on this category should be made according to the above consider-
ations of output form and content, and entered into Table 4.8.

Data Deck Design

One  important use of mathematical models is to obtain answers quickly to various
"what-if?" questions for possible alternative future events.  Answering many
such questions requires repeated applications of the model(s) with appropriate
changes  to the inputs.  Therefore, input decks should be designed in a com-
pact and efficient manner, and available documentation should give all the
information needed to make the appropriate changes.

The rating in this category should be based upon the relative amount of effort
needed to make re-runs of the models.   This includes consideration of the num-
ber of cards needing changes, recomputation time, and the helpfulness of avail-
                                     78

-------
 able documentation.   The ratings should be  entered  into Table 4.8.

 Ease of Modification

 In many cases,  some  modifications to  available models may be  needed  for
 specific applications.   The  ease of these modifications can vary greatly
 between models.   The main factors which contribute  to easy modifications  are:
 proper  code  language,  the use  of many comment  statements in computer  code,
 the use of many subroutines  in computer models,  and the adequacy of  available
 documentation (chapter  3, Table 3.14).   Ratings  in  this category should be
 based upon the  expected modifications needed for each model and  upon  the
 above-listed factors concerning the ease of modification,  and should  be
 entered on Table 4.8.   If program modifications  are definitely not needed,
 this category should not be  rated.

 4.9  SELECTION  OF ATTRIBUTE  WEIGHTS

 The assignment  of attribute  weights should  be  made  after the  user has speci-
 fied all his  ratings  (at the end  of Phase III  or  IV, as shown in Figures  4.3
 and 4.4).  At this point he  should have  a good feel for the relative  import-
 ance of each  rated attribute.

As  stated earlier, the weights  are intended to account  for  the relative
 importance of the various attributes  in  a certain application.   Typical
 ranges of weights for each attribute  are shown in Table  4.1.  The attribute
weights should be recorded in the appropriate  table  (Tables 4.2  - 4.8) for
final calculations of the Performance Index Ratings.

4.10  COST EFFECTIVENESS EVALUATIONS

Final model selection at the end of Phase III or Phase  IV involves simple utili-
zation of information derived in earlier tasks.  This information consists of:

      1)  The total dollar costs derived in Phase II for each model:
      2)  The overall Performance Index ratings either from Phase III,
          or  from both Phases III and IV.
                                      79

-------
            TABLE 4.1  Attribute Weight Ranges
Attribute
Water Body
Time Variability
Discretization & Other
Features
Constituents Modeled
Data Accuracy
Manpower
External Processes
Accounted For
Internal Processes
Accounted For
Model Representation
Accuracy
Numerical Accuracy
Documentation Sufficiency
Output Form and Content
Updateability
Ease of Modification
Weight Range
.8-1.5
.8-1.5
.4-1.2
.5-1.2
.5-1.5
.6-1.6
.7-1.2
.5-1.0
.5-1.2
.5-1.5
.2-1.0
.2- .8
.2- .8
.2- .7
Note:
       Values are normalized so that the average
       weight of the attributes is about 1.0
                          80

-------
The dollar costs and PI ratings should be tabulated as illustrated in
Table 4.9.  Here, the expected performance and cost of each model can be
compared for cost effectiveness ranking and model selection.  In the last
column of Table 4.9, the performance index to dollar cost ratio is given.
The user can optionally use this ratio directly in model selection.  How-
ever, in the final analysis the planner should use his own best judgment in
conjunction with the computed values in this table.

4.11  DEMONSTRATION OF COST EFFECTIVENESS AND MODEL SELECTION METHODOLOGY

The major features of the methodologies presented in this chapter are demon-
strated for a particular planning problem on the estuarine portion of the
Snohomish River Basin in Washington State.  The demonstration represents a
realistic, although hypothetical water quality problem which could be ad-
dressed in an Areawide Waste Treatment Management Study  [43] as required by
Section 208, Federal Water Pollution Control Act Amendments of 1972.

Example Problem

The Snohomish Estuary is subject to municipal, and industrial wastewater,
storm drainage, and extensive agricultural runoff.  Many of these wastes
are known to have substantial time variance.  Wastes may also enter the
estuary from its outer boundary, Puget Sound, through tidal movements.  Most
of the estuary is highly channelized, with several branching sloughs.  The
problem, which may require a mathematical water quality model, is to predict
the varying levels of water quality as a result of future increasing point
and non-point wastewater discharges.

Presentation of Results

Tables 4.2 - 4.8 present the work sheets which demonstrate  results of the
various tests and ratings for the example problem.  The  tables should be
entered and read by proceeding in order of Phases, first for Phase I por-
tions, then Phase II, etc.  The numbers in these Tables  are intended to  be
realistic, but are only used for illustrative purposes.    Ratings and/or
                                     81

-------
weights in phases III and IV could differ from those shown  ( even though
the same models were being evaluated) depending on the nature of the
receiving water body studied and the particular requirements of the
individual model user.

Table 4.9 presents the final ranking of unrejected models.  In this
example RECEIV ranked highest.  Had it been a requirement to model
the full range of pollutants handled by SRMSCI, it would have been
the model of choice.  The costs, however, would have been higher
because of the added costs of computer time, and the requirement
for gathering additional data for calibration and verification.
Furthermore, models are continually being changed and updated and
the ratings could be much different for newer versions.  The reader
is cautioned, however, that as stated previously, a more complex
model is not necessarily a better one.   It is best to stick to the
simplest model that will accomplish the task at hand.
                              82

-------
                                           TABLE 4.2   APPLICABLE SITUATIONS
oo
CANDIDATE MODELS
DOSAG-1
SNOSCI
Simplified
Stream (SSM)
ES001
Simplified
Estuary (SEM)
QUAL-I
QUAL-II
Dynamic
Estuary (DEM)
Tidal
Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)
LAKSCI
Outfall
PLUME
4.2.1
WATER BODY
PHASE I
APPLICABLE?
NO
NO
NO
YES
YES
NO
NO
YES
YES
YES
YES
NO
NO
NO
PHAS1
RATING

/
/


/

8
8
8
8
/
s

S III
WEIGHT
/
'




/
1.2


J





/
4.2.2
TIME VARIABILITY
PHASE I
APPLICABLE?
/
/
/
NO
NO
/

YES
YES
YES
YES
/
/

PHAS
RATING


/


/
/
4
4
8
8

,
/
E III
WEIGHT
/
/

/
/


\



.4



/
/

4.2.3 DISCRETIZATION AND
OTHER FEATURES
PHASE I
APPLICABLE?
,
/
/

/
/
/
YES
YES
YES
YES

/
/
PHAS
RATING
'


/
X


6
6
8
8
/


E III
WEIGHT

/
/


/
/
0.6


|

/
/

-------
oo
-P"
                                                   TABLE  4.3
            CONSTITUENTS MODELED
                                  CANDIDATE MODELS
                                  DOSAG-I
                                  SNOSCI
                                                      4.3.1
                                                             CONSTITUENTS MODELED
 PHASE I
APPLICABLE?
                                  Simplified
                                  Stream  (SSM)
                                  ES001
                                   Simplified
                                   Estuary  (SEM)
                                   QUAL-I
                                   QUAL-II
                                   Dynamic Estuary
                                   (DEM)
                                   Tidal
                                   Temperature (TTM)
                                   RECEIV
                                   SRMSCI
    YES
    YES
                                                           YES
                                   Deep Reservoir
                                   (DRM)
                                   LAKSCI
                                   Outfall
                                   PLUME
                                                           YES
                                                                          PHASE  III       PHASE I
                                                                      RATING   WEIGHT   APPLICABLE?
                                 4.3.2. DRIVING FORCES AND BOUNDARY
                                          FACTORS ACCOUNTED FOR
                                     YES
                                     YES
                                                                                            YES
                                                                                            YES
     PHASE III
RATING   WEIGHT
                                                                                                                  0.9

-------
00
1
™»«..^— — — —
CANDIDATE MODELS
DOSAG-I
SNOSCI
Simplified
Stream (SSM)
ESO01
Simplified
Estuary (SEM)
QUAL-I
QUAL-II
Dynamic Estuary
(DEM)
Tidal
Temperature (TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)
LAKSCI
Outfall
PLUME
'ABLE 4.4 DATA REQUIREMENTS FOR MODEL INPUTS
4.4.1
HYDROLOGIC AND GEOMETRIC
PHA£
APPLICABLE?
/
/
/

/
/
/
YES
YES
YES
YES

/
/
E I
APPLICABILITY
LIMITATIONS

/
/
/
/
/
/
NONE
NONE
NONE
NONE
/

/
PHASE III
RATING
/

/
/
/

/
7
7
7
7
/
/
'
4.4.2
WATER QUALITY
PHASE
APPLICABLE?
/
/

/
/
/

YES
YES
YES
YES

/
/
I
APPLICABILITY
LIMITATIONS
/
/
y

/
/
/
NONE
NONE
NONE
NONE
/

/.

PHASE III
RATING


/
/


/
5
5
6
6
/
/
i^BBMBH^^

-------
                                           TABLE 4.4  (Cont.)    DATA  REQUIREMENTS FOR MODEL  INPUTS
CO

CANDIDATE MODELS
DOSAG-I
SNOSCI
Simplified
Stream (SSM)
ES001
Simplified
Estuary (SEN)
QUAL-I
QUAL-II
Dynamic Estuary
(DEM)
Tidal
Temperature (TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)
LAKSCI
Outfall
PLUME
4.4.3
EFFLUENT
PHASI
APPLICABLE ?
S
/
/
/
/
/
/
YES
YES
YES
YES

/
/
: i
APPLICABILITY
LIMITATIONS
j
/
/
/
s
/
/
Only con-
stant point
source
it
ii
it
/

/
PHASE III
RATING
X
X
/
/
y

/\
4
4
4
4
/
/

4.4.4
OTHER
PHAS
APPLICABLE ?
/
/
/
/
/
/

	






/
/
/
I
APPLICABILITY
LIMITATIONS
/
/
/
/
/
/
/
/
/
/
/
/
/
/
PHASE III
RATING
/
/
/

/
/
/
/
/
/
/
f



-------
TABLE 4.5   DATA REQUIREMENTS FOR CALIBRATION AND VERIFICATION

CANDIDATE MODELS
DOSAG-I
SNOSCI
Simplified
Stream (SSM)
ES001
Simplified
Estuary (SEM)
QUAL-I
QUAL-II
Dynamic Estuary
(DEM)
Tidal
Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)
LAKSCI
Outfall
PLUME
A. 5.1
HYDROLOGIC AND HYDRODYNAMIC
PHASE
APPLICABLE ?

/
/
/

/
/
YES
YES
YES
YES
/

/
I
APPLICABILITY
LIMITATIONS
X
/
/
/
/

/
NONE
NONE
NONE
NONE
/

/
PHASE III
RATING
/
/

/
/
/

8
8
8
8

/
/
4.5.2
WATER QUALITY
PHAS
APPLICABLE ?
/
/
/

/
/
/
YES
YES
YES
YES
/

/
E I
APPLICABILITY
LIMITATIONS

/
/
/
y
/
/
NONE
NONE
NONE
NONE
/
/
/
PHASE III
RATING
/

/
/
/
/

6
6
6
6

/
/
OVERALL DATA RATING
PHA
RATING
jt
/
/


/
/
6.0
6.0
6.2
6.2
/


SE III
WEIGHT
' /
/
/
/
/
f
/
/
0.8


1 F
/
/
/

-------
oo
oo
                                                        TABLE 4.6   INITIATION  COSTS (PHASE II)
CANDIDATE MODELS
DOSAG-I
SNOSCI
Simplified
Stream (SSM)
ES001
Simplified
Estuary (SEM)
QUAL-I
QUAL-II
Dynamic Estuary
(DEM)
Tidal Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)
LAKSCI
Outfall PLUME
4.6.1
MODEL ACQUISITION
T(4.6.1)

/
/
/
/
/
/
1 week
2 weeks
2 weeks
2 weeks
/

/
C(4.6.l)
/

/
/
/

/
$40
$40
$40
$40
/
/

4.6.2
EQUIPMENT REQUIREMENTS
TIME
T(4.6.2)
-7-
/

/
/
/

0
0
0
0
/
/
/
COST
C(4.6.2)
—T
/
/

/
/
/
0
0
0
0
/

/
4.6.3
DATA ACQUISITION
TIME
T(4.6.3)

/
/
r 	

/
/
20 weeks
20 weeks
20 weeks
22 weeks
/
/

COST
C(4.6.3)


/
/
/


$18,000
$18,000
$16,000
$20,000
/
/
/

-------
                    TABLE 4.7   UTILIZATION  COSTS  (PHASE II)
                    mtf^m^m
                    4.7.1
                                                                                                             COST CONSTP.AIM
                                                                                                          TEST!
                                                                                     TOTAL COSTS
                                                         MANPOWER  COSTS
                         MACHINE  COSTS
CANDIDATE MODELS
                                                          COST
                                                         C(4.7.2)
                                      TOTAL
                                    C(4.7.1)
 TIME
T(4.7.2)
COST PER RUN
Simplified
Stream (SSM)
Simplified
Estuary (SEM)
Dynamic Estuary
(DEM)
                                                                              1.0   38 wk
                                               17 weeks
                                                 (2  men)
Tidal
Temperature
(TTM)
                                                                                    42 wk   $56,040
Deep Reservoir
(DEM)
Outfall PLUME

-------

-------
                    TABLE 4.8 Cont.   PERFORMANCE INDEX RATING:   STAGE  2  (PHASE  IV)
                                  	                  	-^—^—•^•^^•^^^^•^••i^*
CANDIDATE MODELS
                                                  SUFFICIENCY  OF
                                                                        OUTPUT FORM AND CONTENT
AVAILABLE DOCUMENTATION
                     NUMERICAL ACCURACY
DOSAG-I
^^^^•i^^BM

SNOSCI
Simplified
Stream (SSM)
Simplified
Tidal
Temperature
(TTM)
 RECEIV
 !•——^—

 SRMSCI
 LAKSCI
 ^———•

 Outfall PLUME

-------
                                       TABLE 4.8  Cont.  PERFORMANCE INDEX RATING:  STAGE 2 (PHASE IV)
10
N)
CANDIDATE MODELS
DOSAG-I
SNOSCI
Simplified
Stream (SSM)
ESOO1
Simplified
Estuary (SEM)
QUAL-I
QUAL-II
Dynamic Estuary
Tidal
Temperature
(TTM)
RECEIV
SRMSCI
Deep Reservoir
(DRM)
LAKSCI
Outfall PLUME
4.8.6
UPDATEAB
RATING

/
/
/

'
/
6
6
6
6
/
/

ILITY
WEIGHT
/


/
/
/

0


1
.6




/
/
4.8.7
EASE OF MOD
RATING
'

/


/
/
8
7
6
7
/
/

IFICATION
WEIGHT
S

/
/
/



0


\
.3


i

/
/
OVERALL PI RATING (PHASE III & IV)
/ /
/ /
/
/
/ /
/ /
/ /
5.96
6.14
7.15
7.64
/ /
/
/

-------
U>
                                                       TABLE 4.9




                                              COST EFFECTIVENESS COMPARISON
RANK
3
4
1
2
MODEL
DEM
TTM
RECEIV
SRMSCI
PI RATING
(From Phases III &IV)
5.96
6.14
7.15
7.64
TOTAL COST OF APPLICATION
(From Phase II)
$50,740
56,040
53,040
60,540
PI*104
DOLLAR
1.17
1.09
1.35
1.26

-------
                               CHAPTER 5
                         MANAGEMENT OF MODELING

 5.1  INTRODUCTION

 The approach to  water  pollution  control has  changed  significantly in recent
 years.   Emphasis is  being  placed on  comprehensive planning for increasingly
 larger  regions than  heretofore and on  the  implementation of plans.  The
 extremely high financial and  social  costs  of plan implementation has re-
 quired  increasingly  more explicit and  detailed analyses of greater numbers
 of  alternatives.

 The foregoing and other trends have affected the manner in which water quality
 planning is  conducted.  Interdisciplinary  teams working within short time
 frames  are responsible for guiding the efficient expenditure of large amounts
 of  planning  funds and for  the use of sophisticated technologies, among which
 modeling is  eminent.  It is assured that the use of water quality models will
 become  even  more widespread in the future.

 Those persons responsible  for supervision of significant water quality
 planning programs must be prepared to deal extensively with decisions related
 to modeling.  Reluctance to use water quality simulation models where their
 analytical strengths are applicable will severely handicap planning.   Yet,
many of  the  agencies presently being assigned water quality planning respon-
 sibility do  not have an already developed capability for modeling and many
planners are not experienced in supervision of water quality modeling pro-
grams .

Adequate administration of water quality modeling programs requires  that
managers be prepared to develop or acquire appropriate technical capability
and effectively integrate this capability with the remainder of the  planning
team.  This Chapter addresses some of the management considerations  attendant
to undertaking a  water  modeling program.

-------
5.2  NEED FOR ASSISTANCE

The planning manager must assess the adequacy of the existing in-house
staff to undertake and complete any new program which is to be initiated.
The earliest possible recognition of deficiencies is beneficial in that it
allows maximum flexibility in remedial action.

Types of Assistance Required

The major reasons for securing assistance are the need for specialized
expertise and need for additional manpower.  Occasionally the two needs
may merge.  It is important that the nature of need be explicitly recognized,
as it may affect the procedures for obtaining assistance and the sources
which are considered.

Specialized Expertise.  One of the basic decisions affecting the need for
obtaining specialized expertise is selection  of the type of modeling, if
any, to be used.  While the less complex modeling procedures based on
nomographs may be readily applied by most planning staffs, the introduction
of computerized models often represents a substantial increase in complexity
and requires  an entirely new set of skills.

The particular disciplines  required in a computerized water  quality modeling
program depend on  the  type  of  model or models to be used,  the  availability
of data and  facilities, and  the detail of the  results which is  expected.   The
skills  needed may  range  from those  for preparation of data inputs and model
execution to  those  required to modify an existing model or develop a  new
model appropriate  to the situation  at hand.   Generally, the  decision  to model
implies a need  for capability  in the  selection of  the  specific type  and
version of  model  to be adapted or  constructed, data analysis,  programming,
key  punching, machine operation, and interpretation of  results.  Less obviously
related capabilities may also  be needed,  such as those for directing or
                                    95

-------
  accomplishing water quality sampling and analysis.  Assessment
  of  the  full  range of expertise which is needed to conduct a water
  quality modeling program must be based on a thorough identification of
  each step in the intended program.  This assessment requires some familiar-
  ity with both modeling and planning.

 Additional Manpower.  A need for contractual services for water quality
 modeling or  related services may arise quite apart from the need to
 obtain  specialized expertise.   Even if adequate technical capability is
 available on staff to carry out all of the modeling related activities, the
 existence of other commitments, accelerated time schedules,or the need for
 simultaneous conduct of activities at several locations may require obtaining
 assistance.   The planning manager should carefully appraise manpower needs
 with regard  to  the type and  level of experience required.  If adequate
 technical capability exists  on staff to make all decisions and to provide
 supervision,  less experienced  and less  expensive assistance can be obtained.
 Similarly, in arranging assistance primarily to obtain  additional manpower,
 some thought  should  be  given to the effectiveness of how that manpower is
 used.  While  it  might be most  convenient, for  example,  to press all  avail-
 able personnel into  service  to  conduct  a widespread  water quality sampling
 program, the  use of  contracted  professionals for  that purpose is not  likely
 to be cost-effective.

Timing of Assistance

The  stage of  the planning process at which water  quality modeling expertise
is required and  the extent of assistance needed vary according to the type
and  purpose of the plan to be prepared.   It is largely independent of whether
such expertise is available on staff or obtained  through contracting.

It would be desirable, of course, for the planning manager to maintain constant
and  detailed coordination with all members of the planning team.  Practicality
prevents this, however, particularly In the early stages of the planning
process.   At  that time, the full identification of the planning participants
                                  96

-------
may not be complete, but may include individuals from multiple federal,  state,
and local agencies and from the private sector.  In addition to defining  the
planning process in sufficient detail to proceed, the planning manager  may
be engaged in the development of organizational arrangements, preparation
of applications or requests for funds, and miscellaneous other activities.
Time is at a premium under such circumstances and crucial errors may be made
if modeling related considerations are not given adequate attention.  It
is particularly important that realistic budgets and schedules for water
quality modeling be established from the outset and that a proper decision
be made as to the general type and level of modeling to be included.
Detailed consideration of the modeling program is also important during
design of the inventory phase of the planning  process, development of
procedures for plan formulation and evaluation, and in arrangements  for
future updating of the plan.

While  some specific attention must be  given  to water quality modeling  (or
at  least  to whatever  technique of analysis is  to be used) at each of
several  stages  in  the planning process,  the  detail and  intensiveness of
consideration depend   on  the overall  importance  of water  quality and the
analysis  techniques  to meeting project purposes.  Water  quality modeling is
clearly  a more  major  part of a planning effort than one  to  guide land  use
or  general water  resources development.   The same  types  of  concerns  dis-
 cussed in Chapter 4  relating to  selection of cost  effective procedures
 of  analysis  indicate the  stress  to  be placed on water  quality modeling
 as  opposed  to  other  elements of  the plan.
 At whatever stages of the planning process decisions regarding water
 quality modeling are made,and regardless of the importance of water quality
 modeling vis-a-vis other planning elements, experienced and expert counsel
 should be available to the planning manager.  It should be recognized that
 the information contained in Chapters 2, 3 and 4 is only a rudimentary intro-
 duction to modeling.  This material is intended to assist planners  in
 communicating with modeling specialists and to assist  in some basic decisions
 regarding water quality modeling.  Some'additional points of greater complexity

                                   97

-------
 than  those  included or indicated  therein may be important in a particular case.

 5.3   IN-HOUSE vs. CONTRACTUAL SERVICES

 The decision as to whether modeling or other services related to modeling
 should be carried out by in-house staff or on a contractual basis warrants
 particular  attention.  Depending on the circumstances at hand, the appropriate
 extent of contractual services may range from none to a preponderate portion
 of any particular project.  Careful analysis of project requirements and
 objectives  as well as in-house capability is required to assure that the
 proper types and amounts of services are contracted.

 Advantages  of In-House Efforts

 There are considerable advantages to carrying out some or all of the water
 quality modeling portions of a planning effort with in-house staff.  Among
 others, these include certain aspects of the following:

                 1.  Coordination
                 2.  Ease of supervision
                 3.  Management flexibility
                 4.  Implementation capability
                 5.  Updating capability
                 6.  Cost

Whether the advantages of the above outweigh the advantages of contracting
 for services to accomplish a particular planning elements depends on the
 applicable circumstances and the importance attached to each potential
benefit.

Coordination.   The use of in-house staff considerably eases problems of
achieving a desirably high level of coordination,  particularly where only
one agency is furnishing all or the bulk of the required staff effort.  The
familiarity of staff members one with another and their prior experience in
cooperatively conducting other programs facilitates the development of
                                    98

-------
smooth and productive working relationships.   Where staff members have
responsibility for management of or participation in other agency programs,
cognizance of those programs and their relationship to the planning effort
is automatically assured.

Aside from the achievement of coordination, the mechanics of performing
coordination can frequently be made simpler,  less formal, and far less time
consuming if only staff members are involved.  Transmittals of information
may be supplemented by frequent discussions;  group meetings are less
difficult to arrange; and events may be managed on a day to day basis.

Ease of Supervision.  A far greater capability usually exists for close
technical supervision if water quality modeling activities are carried out
in-house.  When outside assistance is used, it is generally necessary to
specify in some detail the procedures to be followed, unless arrangements
are made on a time and materials basis.  In some cases, such as those
where the identification of problems to be addressed has not been com-
pleted, specification of all technical procedures may be difficult and
prone to error.  Any errors which are discovered may be difficult and
expensive to correct if  contractual modifications are required.  This is
not to suggest that in-house-modeling efforts can or should be undertaken
without detailed fore-thought  or pursued without well prepared guidance.
However, errors or deficiencies may be detected earlier and corrective
•action more readily initiated  when only staff members are  involved.
Management Flexibility.  Problems  inevitably  arise  in  any planning  program
which affect  the  intended  smooth flow  of  activities.   Disruptions may  include
sudden changes  in funding,  acceleration or  delay in schedules  and/or the
introduction  of new objectives  resulting  from federal  or state legislation
or  other  sources.   Any of  these may require changes in the  planning pro-
 gram ranging  from minor adjustments to thorough revisions.   The types  of
corrective action required to accomodate  unforeseen conditions may  be  severely
constrained by  contractual commitments or at  least  expensive to undertake.
                                     99

-------
Water quality modeling programs are particularly vulnerable to changes in
approach because of their complexity and the substantial preparation which
precedes model use.   The type and purpose of planning may, and probably
will, affect the likelihood of such changes.  For example, the extent of
water quality modeling which might be used in a framework study pursuant to
Section 102 of the Water Resources Planning Act of 1965 (Public Law 89-80) may
not  be critical to the success of the plan.  If modifications are necessary
to conserve time or funds, water quality modeling may be considered as a
minor program element and either abbreviated or totally eliminated in favor
of other techniques of analysis.  This type of modification is far less
likely to occur in those planning programs in which water quality management
is the principal concern.  In short, in-house accomplishment of water
quality modeling may offer substantial benefits where uncertainties exist in
the  future of the planning program.

Capability to Implement.  The objective of most planning programs is imple-
mentation of one sort or another.  The range of potential implementation
action which might be undertaken for water quality control purposes is large.
Framework level studies seldom result directly in project development since
they are not suitably detailed to serve as a congressional authorization
document.  They may give rise,however, to further functional planning for
water quality improvement, development of specific project plans,  and
other similar activities.   Basinwide water quality planning may
result in identification of regional treatment schemes, needs for new
facilities or modification of existing facilities, and standards of ambient
or effluent quality.  Areawide waste treatment management plans directed by
Sec.  208 (b) of Public Law 92-500 are assured of requiring a diversity of
actions which may include structural programs for point and nonpoint sources
of pollution, regulatory programs for point and nonpoint sources of pollution,
and further efforts at problem identification and planning.

Implementing water quality management plans of various types will require
knowledge of and perhaps further application of the analytical techniques
used in planning.   Benefits may correspondingly result from maintaining
                                    100

-------
the continuity of experienced individuals and organizations.   The magnitude
of any such benefits and their consequent impact on the decision to  develop
in-house capability for water quality modeling or to use contracted services
may vary.  Factors to be considered in making the decision include, among
others, the certainty of the necessary funding, authorization, and other
requirements for any anticipated future activities.

Capability to Update.  Few plans of significant scope are firmly fixed in
time,and some updating of plans is normally anticipated.  Modifications may
come about through general revision and improvement or through adjustments
to specific items of new information.  Water quality models and other
planning aids involve significant investments.  The opportunity for  their
repeated use in updating or otherwise revising the plan makes their  original
use more cost-effective.  The continued use of water quality models  may also
be useful in testing the effect of proposed developments or policies, or  assuring
their consistancy with the plan.  These or other reasons may indicate the
future need to apply the same water quality models on a repetitive basis.
Where this need is anticipated, there may be advantages to assuring  the
capability to update or modify plans or operate the models for other purposes
exists on an in-house basis.  The capability to update implies a thorough
understanding of the tools for analysis including their limitations,
applicability, and the assumptions on which they are founded.  The most
certain way of assuring this capability is to perform all original work on an
in-house basis.  When this is impractical, inclusion of adequate provisions
for staff training by the contractor may be necessary.

The importance of achieving the capability for future model use in deciding
to perform water quality modeling on an in-house basis depends on the extent,
nature, and timing of the expected future use.  Some plans, such as broad
statewide plans, Level A framework studies, and the Level B studies
pursuant to Public Law 89-80 may only be scheduled for revision or updating
every 5  to 10 years.  Considerations of future model use in such  cases must
include  the possibility of staff turnover, development  of  improved analytical
techniques and any uncertainty of need.  Similarly, updating may  not  be  a
factor in planning for specific facilities which are not expected to be
                                      101

-------
updated in the future.  However, for some types of planning, updating is
mandatory.  Areawide waste treatment management plans, for example* are to
be updated on an annual basis [Sec. 208(b)(2)(A)].  Other types of water
quality control planning directed by Public Law 92-500 such as that pursuant
to Sec. 303(e) are to be maintained on a continuing basis.  Where the near
future need to update is assured, the development or use of in-house capability
for water quality modeling has significant advantages.

Cost.  Assuming that time-consuming and expensive staff training can be avoided,
the conduct of water quality modeling programs is less expensive if done
using in-house capability.  In addition to the obvious elimination of the
profit component of a contractor's cost, savings may arise from other sources
including elimination of travel, costs of contracting and contract administr-
ation, salary differential, and allocation of overhead.  The magnitude of the
savings resulting from a decision to develop in-house capability may be
substantial under certain conditions.  Their appraisal should not overlook
the following:

                 1.   Recruitment costs (if additional staff are
                     required);
                 2.   Cost of compensation including salary, taxes,
                     and other parts thereof;
                 3.   Overhead including costs for space, utilities,
                     administration, supervision, and supporting services;
                     and
                 4,   Costs resulting from non-productive time such as
                     vacation, sick leave, periods of low workload and others.

If the appraisal of cost indicates that it is advantageous to carry out water
quality modeling activities on an in-house basis, and such an approach would
be consistent with other program needs or constraints, then the risk of
doing so should be carefully reviewed.  If experienced and expert staff are
already available or can be recruited quickly,  there is little risk in
proceeding.  However, if the water quality modeling program is to be
                                   102

-------
initiated and pursued while the development and training of staff is under-
way, the opportunity for serious errors may overshadow any cost savings.   In
this event, consulting assistance of at least an advisory nature can be very
useful.

Advantages of Contractual Services

The use of contractually arranged services has several potential advantages.
As previously noted, contracting offers one means of obtaining technical
expertise, additional labor, or both.  Contractual services also provide a
means  for  initiating and pursuing programs concurrent with development of
the staff  capability needed to implement and update plans which  are prepared.
In addition  to  those important advantages, the use of contractual services may
save time, result in a  lower overall cost, and assist in staff management.

Time Savings.  Well established consulting firms operating within their area
of special expertise can be of considerable use in the prompt initiation and
aggressive pursuit of a planning program.  Such firms often have expert staff
available who can assist in making many of the technical decisions without  the
need for lengthly research.  Frequently, if consulting staff have participated
in programs  similar to  those being undertaken, their familiarity with  state,
federal, and other requirements may help in expediting the preparation of a
detailed plan of work and  initiation of the planning process.

In some cases,  considerable time may be saved by the selective  retention of
contractors  who have particular experience.  For example,  water  quality
analyses and modeling conducted as part of areawide waste  treatment management
planning pursuant to Sec.  208  of Public Law  92-500 might be  considerably  aided
in  some of its  parts by appropriate use of contractors who are engaged in or
who have completed related work for one or more  communities  within  the desig-
nated  area.  Conversely, specific facilities planning might  be accomplished
more efficiently by use of contractors who have  had  specific experience  in
the region as a part of areawide or basinwide  planning.   In either  case,  the
                                      103

-------
 contractor's familiarity with the local situation  is  a  valuable commodity
 which enhances  his technical capability and  helps assure  coordination
 between at least  the modeling related  portions  of different planning programs.
 Similar to the  preceding,  some benefits may  be  obtained by the use of
 common contractors for  the preparation of  appropriate  portions of Level B
 studies and water quality  plans pursuant to  Sec.  303(e) of Public Law
 92-500.

 Additional time savings may result from using contractual services for water
 quality modeling  due to the contractor's access to equipment and the avail-
 ability of specific models or other needed items  which may have been prepared
 or  used previously.

 Chapter 6  includes a discussion regarding  the procedures for assuring
 potential  contractors have the requisite capability  to be of immediate
 assistance.

 Cost.   Notwithstanding  the usually higher  cost  of  contractual services on a
man-hour basis, there may  be  some financial  savings  attendant to their use.
As an example,  the wider array of specialists available through an appropriate
contractor may  increase  the efficiency with which  particular technical tasks
are accomplished.   In addition,  there may  be substantial savings of an admini-
strative nature, since  the contractor's  fee normally includes payment for the
technical and other supervision necessary  to provide whatever product was
specified and to assure  its adequacy.  In  some  cases, the schedule for
planning may not require the continual participation of modeling specialists
and, in that event, the contractor absorbs the  cost of idle time.

Management.  Difficulties are commonly encountered in balancing staff size
against workload,  particularly for smaller offices.  Initiation of new
programs for water quality planning or other purposes frequently require
expansion of the existing staff.  If the new program is to be pursued only
for a limited period of  time, managers are faced with the need to reduce
staff until a new  expansion in personnel is needed for some later project.
                                   104

-------
Even if desirable as a management approach, there are several obstacles
to such a staffing policy.

The time required to locate, interview, recruit, orient and train new staff
prior to their fully effective participation may require several months.
In addition, potential employees may be reluctant to accept employment on a
short term basis at normal compensation levels.  Difficulties can be
encountered in quickly locating sufficient numbers of persons with approp-
riate qualifications.  Wide variations in the number of technical staff
members may also adversely impact the overall management and operation of
relatively small organizations through stimulation of undesired growth in
administrative and other supporting staffs.  If staff expansions of major
scope are undertaken to accomplish projects of limited duration, administr-
ators must anticipate a difficult period during which employees are terminated.

Use of contractually arranged services to meet peak workload requirements
can relieve many of the management problems which would otherwise exist.
Staff growth and selective recruitment of individual staff members can be
oriented more specifically to the long term needs and goals of the organization.
Similarly, the supporting services for the technical assistance thus obtained
is largely or wholly provided by the contractor.
5.4  INTEGRATION OF MODELING AND PLANNING

Water quality models are tools for accomplishing one portion of the planning
process.  Their effective use demands more than the technical expertise to
select, prepare, execute and interpret the results of whatever model may
be used.  It is imperative that the use of such analytical tools and the
analyses performed be properly and thoroughly integrated with the numerous
other portions of the planning process.  Adequate integration of activities
does not occur automatically.  It requires that the analyst and planner both
recognize explicitly the different nature of their respective roles, the
                                    105

-------
relation between modeling and planning, and the means whereby coordination
can be achieved.

Nature of Planning and Modeling

Planning and modeling are distinctly different activities.   However, they
cannot be separated to the extent that planners are isolated from the
modeling-related decisions or that analysts do not participate in planning.
The close cooperation of both is needed and when secured, is mutually
beneficial.

Planning Process.  The planning process required to support resource manage-
ment decisions in the public sector is enormously complex for all but the
simplest cases.  Generally, decisions of any magnitude are arrived at only
after extensive consideration of numerous alternatives.  Participants in the
decision making process frequently include the public or some part of the
public forming the constituency of the decision makers; federal and state
officials, agencies, and departments; and various local entities including
general and special purpose governments and special interest groups.  Each of
these participants may view whatever problem is to be addressed in different
terms according to their objectives.  Each may accordingly place various
constraints on the types of solutions which are acceptable and the methods
of implementation which will be supported.

It is the planner's task to describe the problem satisfactorily.  This includes
describing its severity and urgency, assembling and interpreting the information
necessary to identify alternative solutions, and assisting in selection of
the preferred solution.  As part of the foregoing, the planner must articulate
or participate in the articulation of objectives and goals, identify
informational needs and appropriate techniques of analysis, and develop the
criteria for selection among feasible alternative plans.  The planner must
also find suitable ways of presenting the results of the planning effort in
a meaningful way which facilitates its use by other participants in the
planning process.
                                       106

-------
 The major types of planning programs  which are either  specifically under-
 taken for water quality control  purposes  or which include  consideration  of
 water quality control as one of  their principal purposes require  that
 consideration be given to all aspects of  the problems  addressed and  the
 solutions proposed.   Generally the planner must investigate the engineering,
 financial,  economic,  environmental, social and perhaps other aspects of
 alternative plans.  The number of alternative  actions  which could be taken
 may be large.   Their  evaluation  requires  that  the planner  have the
 capability  available  to systematically apply suitable  analytic techniques.
 Water quality models  are one available technique.

 Modeling.   The process  of modeling or otherwise performing the analyses
 needed for  planning requires that the analyst  participate  in the description
 of  the problem and in the selection of the analytical  techniques to be used.
 In  addition to performing the analyses, the analyst should be prepared to
 assist in the  selection of  alternative plans to be considered, evaluate  the
 effect of assumptions, and assist in the presentation and interpretation  of
 results.

 Depending on the type of planning undertaken,  analysts concerned with water
 quality modeling may represent only a  portion  of  the analytical capability
 on  the planning team.    In such cases  the water  quality analysts must be
 prepared and able to effectively cooperate with analysts having other
 responsibilities and to  achieve compatibility  among the technologies employed.

 Relation of Modeling to Other Planning Activities

As suggested in the preceding discussion of the nature of planning and
modeling activities, the planning process includes a number of elements which
can be accomplished more effectively with the assistance of the analyst and
 the availability of models.  A comprehensive and detailed discussion of the
relationship between modeling and planning would depend on discussion of a
specific case.  However, the general and broadly applicable relationships
of modeling to major portions of  the planning process merits note.
                                    107

-------
Most plans, including broad framework studies and statewide plans, Level B
studies and other basinwide plans and more specific plans for smaller areas
are approached in a similar fashion.  The general steps include description
of the problem and organization for planning, data gathering, development and
evaluation of alternatives and recommendation of the preferred solution.

Description of the Problem.  The full description of water quality problems
or the water quality component of problems requires identification of the
location, nature, and magnitude of the pollution sources as well as the
resulting impact on water quality.  Preparation of a complete problem description
may be relatively easy in the case of individual point sources of waste which
have been or can be readily monitored and where ambient water quality changes
can be directly related to specific effluents.  In other and more complex
cases, arriving at an adequate description of the problem may constitute a
major portion of the planning process.  In these more difficult cases, the
water quality analyst, using modeling or other appropriate techniques, becomes
a key member of the planning team.

Nonpoint sources of pollution occurring with variable severity throughout
an area or different portions of an area and in conjunction with one another
present a particularly difficult task for a planner.  Component parts of
such pollution loads must be identified and quantified in at least general
terms.  Otherwise, corrective actions are likely to be proposed on a more
or less arbitrary basis and without assurance or even a knowledgeable estimate
of their efficacy.  The financial and social costs of pollution abatement
measures are too large to permit such speculative planning.  It falls then
to the analyst and the planner to set forth an analytical procedure which
will remove the mask of complexity and enable clear understanding and
description of the problem.

The kinds of analytical assistance required by the planner in problem
definition varies.  In many cases, the analyst's assistance is needed to
accumulate the water quality and other data which is already available,
transform it into the most useful format and terms, and help infer its
                                     108

-------
 meaning.   More detailed  and/or more  complex  analyses may be necessary for
 areawide  waste treatment management  planning,  facilities planning, waste-
 load allocations  studies,  and other  specific types of plans.  It may be
 necessary in some instances  to use sophisticated models to help interpret
 available information.   Regardless of the level of problem definition which
 is  required,  it is generally useful  for planners to anticipate that some
 analytical assistance will be needed for proper problem definition and to
 provide for  its early availability.

 Inventory Phase.   Some  forethought  must be given to planning to insuring
 the availability  of required information in the form and at the time it will
 be  needed.  The analyst  should play  an important part in itemizing the types
 of  data and  information  to be used throughout  the planning process, comparison
 of  the data and information needed and available, and identification of needed
 data collection.  Water  quality models have specific data requirements as
 noted in  Chapters 3 and  4.  Review of the data and information to be used in
 the  modeling or other analytical procedures can be a major aid in assessing
 the  overall data requirements of the planning program.

Development and Evaluation of Alternatives.  The analyst may be of substantial
assistance to  the planner  in the development of alternative plans and plays
the  key role in evaluation of their potential impact on water quality.   Assist-
ance  in the development  of alternative plans may include identifying all of
the  factors whose modifications would favorably affect water quality, describing
the  sensitivity of water quality to various types of changes which might be
planned, and displaying alternative plans in a systematic fashion which
facilitates their study.   Based on the experience gained in testing early
alternatives and observation of the reaction of the system to various proposed
modifications, the analyst may be particularly helpful in suggesting the
scale and location of various plan features.

It is a principal responsibility of the analyst to identify the water quality
impacts likely to result  from whatever alternative plans are developed.  Given
an alternative plan,  the  analyst must translate the plan into appropriate

                                     109

-------
mathematical terms, perform the analysis using a simulation model or what-
ever other technique has been selected and report the results in an under-
standable fashion.

The analyst and the planner must work together closely throughout the plan
formulation and evaluation portion of the planning process.  The planner
must simultaneously coordinate the effort of the water quality analyst with
those analysts evaluating other aspects of each alternative to assure the
overall assessment gives appropriate attention to all important points.

Recommendation of the Preferred Solution.  When the evaluation of various
feasible alternatives has been completed, the planner is usually required to
recommend or suggest a preferred plan which best meets the original objectives
within the applicable constraints.  Often, the planner is required or desires
to present other meritorious alternatives for public and official review.

Full display of alternative plans for water quality control or for plans of
larger scope including water quality control measures is difficult.  To
promote understanding on the part of lay reviewers, displays should be brief
and relatively simple.  Yet, oversimplication can obscure important points.
The planner and the analyst must cooperatively identify the characteristics of
each plan to be displayed and the best method of display.  Importantly, the
analyst should itemize all of the assumptions on which the assessment of
water quality rests and the probable effect of errors therein, point out the
accuracy of the analytical procedures which were used, and array results in
some organized form.

Achieving Integrated Planning

The frequent and extensive interaction between the planner and the analyst
described in the preceding sections can be achieved if the arrangements for
planning are made with such interaction in mind.  In particular, specific
attention should be given to the nature of the relationship between planning
and modeling while preparing the plan of work and the schedule and budget
therefore.

                                    110

-------
 Plan of Work.  The plan of work becomes a guide to performance of the planning
 effort when it is completed.  Where grant funds are involved,  the plan of
 work may also become a part of the contractual arrangement between grantee
 and grantor.   Extreme care is needed in preparing the plan of  work for
 complex planning efforts to insure the activities described and their
 sequence will result in successful completion.  Unless the planner is well
 versed in the details of the analytical procedures to be used,  the analyst
 should participate fully in the preparation of the plan of work.   The
 descriptions  of individual tasks within the plan of work should be in
 sufficient detail to identify all of the interrelationships between
 analytical and planning efforts.   Special care should be taken to insure
 that no important steps are omitted  and that the sequence of activities will
 enable the maximum integration of activities.

 Schedules  and Budgets.   Scheduling and  budgeting a plan of work is frequently
 made difficult by the number of  constraints imposed by legislation, grantor
 agencies and  perhaps other  sources.  Along  with  the obvious need  for  consistency
 with time  and funding constraints, schedules and budgets must be  prepared  so
 as  to  properly pace  the  planning  process  and to  place appropriate  emphasis
 on  each aspect of  the effort.  The time and  cost of water quality  modeling
 should  be  adjusted to provide  a  level of  detail  and accuracy which is
 commensurate  with  other  analyses  to  be performed.   Schedules and budgets
 must reflect  the  full cost  and time  of  steps preparatory to modeling as well
 as  for  the modeling  effort.  These and other types  of  considerations make
 it  mandatory  that analysts  assist  in the  preparation  of  schedules  and budgets.
 Failure to give proper attention  to  the details  of  analytical procedures
 during  schedule and budget preparation usually results in underestimation
 of  the  time and cost  required.  If this occurs,  the planner will face an
 unduly difficult problem in  integrating modeling and  the prospect  of making
major adjustments in  the planning process after work has begun.
                                     Ill

-------
                               CHAPTER 6
                     USE OF CONTRACTUAL SERVICES -
           ADMINISTRATIVE, LEGAL AND PLANNING CONSIDERATIONS
 6.1   INTRODUCTION

 Decisions to use mathematical models  to  assist  in water  quality manage-
 ment  planning should  be based on a well  defined need.  Likewise,  selection
 of the  particular model or  type of model to be used must be made  considering
 the model will be used in the planning process.  These assessments as well
 as the  carrying out the planning process requires expertise in mathematical
 modeling  and planning.  Sufficient expertise may exist in-house or may be
 obtained  through the use of contractual  services.  This Chapter discusses
 the several  steps of obtaining contractual services for purposes related
 to water  quality modeling and contains some suggestions of procedures
 therefore.

 The usual  purpose in using contractual services for the preparation or
 use of  water quality models is either to obtain access to  skills  which are
 not available on an in-house basis or to supplement the  amount of manpower
 available.   In either case, the objective is to obtain a needed product  or
 service at a fair price, on a timely  schedule, and in a  manner most compatible
 with  other parts of the planning process.

The general  steps involved in securing and using consultant services
 which are dealt with  in this chapter  include:

      •  Planning the  Procurement;
      •  Preparation of the  Request for   Proposal;
      •  Preparation of the  List of Bidders;
      •  Solicitation  of Proposals;
      •  Evaluation of Proposals;
      •  Contractor Selection;
      •  Contract Negotiation Preparation,  and Award; and
      •  Project Administration.
                                   112

-------
Many of the basic procedures which must be followed by governmental
 bodies  in carrying  out  the  above  steps are  frequently  established by
 Federal and  State statutes,  rules and regulations, ordinances and/or
 explicit  policies controlling  the business  practices of public agencies.
 No specific  effort  has  been made  to  tabulate or  identify specific legal
 requirements of  these types.   However, the  discussion  calls attention to
 those points  likely to be  subject to particular legal requirements.  The
 discussions  presented herein describe generally  beneficial practices.
 They are  not intended to replace  an  adequate legal review of procedural
 requirements.

 6.2 PLANNING THE PROCUREMENT

Consideration should be given to certain general aspects of contracting
 before  any effort is made to prepare a Request  for Proposal.  The principal
 preparatory  steps include an evaluation  of  the  type of contractural arrange-
 ment to be developed and the capability  of  the  in-house  staff to conduct
 the procurement  process satisfactorily.

 Types  of  Procurements

Procurement of assistance in water quality modeling may be done on
 either  a  competitive or non-competitive  basis.   A competitive procurement,
 in turn,  can either be  advertised or negotiated.  An  advertised procurement
 includes  bids and award of  the contract  to  the  lowest  responsive bidder,
 based  on  compliance with a  detailed  study design or plan of work.  A  nego-
 tiated  procurement  usually  involves  reaching  agreement with a  contractor
 without formal advertising  for bids, but after  soliciting proposals  from
 several qualified sources.   A  non-competitive  (sole source) procurement
 is a contract negotiated directly with a single source.   Each  of  the  several
 types  of  procurement procedures have different  characteristics which  affect
 their suitability for use in differing planning situations.

Advertised Procurements.  The publicly advertised procurement is
 probably the best known procedure in the public works area or  in general
                                     113

-------
governmental use.  This procedure may sometimes be suitable for use in
obtaining, modeling or modeling-related assistance.  However,  there are
several criteria which should be considered in determining whether an
advertised procurement should be used in searching for a suitable contractor.
Unless comparison of the situation to those criteria indicates that this
type of procurement is in order, the costs associated with the advertisement
process may not be recouped.  The principal criteria which should be met are:

     1.  The modeling or modeling-related services required can be defined
         in sufficient detail to permit qualified contractors to bid a
         fixed price for completion of the desired services on a relatively
         equivalent basis;
     2.  The modeling or other services required are such that the contract
         can be awarded on the basis of only the fixed price without a
         technical evaluation of the contractor's proposed techniques;
     3.  A number of qualified sources exist which could conceivably respond
         to the advertisement (a later section titled "Identification of
         Prospective Contractors" addresses this determination); and
     4.  The anticipated value of the procurement is large enough to warrant
         the expense of an advertised procurement.  Local regulations
         will frequently specify minimum procurement value.

Negotiated Procurement.     The more flexible negotiated procurement  is
frequently used for modeling services.  The conditions indicating use of
this procedure often fit the situation which exists at the time water
quality planning efforts are initiated.  Those conditions include:

     1.  The modeling-related services desired cannot be explicitly described
         in sufficient detail to eliminate all considerations but cost; and
     2.  The work to be performed is of sufficient complexity and depth that
         an evaluation of the approach and techniques proposed for accomplish-
         ment is warranted.

If it is determined that a negotiated procurement will be made, either a com-
petitive or non-competitive route toward negotiation may be followed.  Before
                                    114

-------
deciding on the use of non-competitive negotiation for a particular
procurement, a detailed investigation of legal matters should be made
to assure correct procedure.  Non-competitive procurements are usually
restricted to special circumstances.

Legal Considerations.   While most types of contracting arrangements are
enabled for governmental bodies, strict regulations sometimes exist as
to the circumstances under which each may be used.  The need to obtain
the most favorable conditions normally requires consideration of more
than one contractor and competition is frequently required by law unless
unusual conditions warrant the use of a non-competitive (sole source)
negotiated procurement.  While the specific requirements justifying a
sole source procurement may differ between various governmental units,
they generally include one or more of the following:

     1.  The procurement is excluded from competition due to small cost,
         procurement from another public agency, or a finding that the
         desired service is only available from a single source.  (This
         latter basis is not likely to exist with regard to water quality
         modeling services);
     2.  The procurement is a continuation of previous work;
     3.  A substantial investment already made in one Contractor's work
         would have to be duplicated by another; or
     4.  The service required is patented, copy-righted, or proprietary.

Planning Considerations.  It is important to note that in undertaking any
contracting for modeling or modeling services, the nature of the services
or products sought must be identified to the fullest degree possible.  Any
uncertainty pertaining to the technical aspects of the modeling, the type
of simulation desired, or the role of modeling within the overall planning
program, should be specifically recognized.  Planners should insure that
administrative staff and others involved in a potential procurement are
aware of the degree of uncertainty on these and any other important points.
                                     115

-------
The range of effort on the part of the contractor which might be needed
to respond to these uncertainties may play an important part in the selection
of procurement procedures and the type of contractual arrangement to be
used.  In particular, advertised and competitive procurements tend to be
most applicable to those cases in which little uncertainty exists and the
services needed can be exactly described.

Conducting the Procurement Process
Before beginning the procurement process, an assessment should be made of
the in-house capability to conduct each step of the process without assistance.
It should be particularly noted that preparation of the technical portions
of an adequate RFP requires nearly the same knowledge and skills as those
being solicited.  If these capabilities are not available in-house, the
Purchaser may be well advised to retain a consultant to assist in making a
preliminary assessment of the type and amount of assistance needed, the
relationship between modeling and planning, and in preparing the RFP.  When
proposals are received, the same consultant can perhaps render valuable
assistance in their evaluation.  However, retention of a consultant for
these purposes should exclude that consultant from consideration in
contracting for carrying out those or other closely related portions of
the work.

If it appears that a consultant will be needed to assist in any part of
the procurement process, the necessary services should be arranged at an
early date.

6.3  PREPARATION OF THE REQUEST FOR PROPOSAL

Several potential contractors are usually considered in any substantial
procurement involving a fixed scope of work.  As one basis for selection,
it is usually necessary that each contractor to be considered provide a
proposal or offer of services for review.  The basis for such offers is
the Request for Proposal (RFP) issued by the purchaser.  The purpose of  the
RFP  is to describe the work desired in sufficient detail to bring  forth
                                      116

-------
proposals which are clear, to the point, and which contain the information
needed for their comparison and analysis.  The RFP should also be
sufficiently explicit to avoid the submission of unqualified or poor
proposals due to misunderstanding of the services which are sought.

A well prepared RFP is a major step toward a satisfactory project in any
undertaking.  This is especially true in the case of a rapidly developing
technology such as water quality modeling.  The extra effort spent in the
methodical preparation of a well defined and well written RFP results in
submission of better organized and more responsive proposals from the
potential contractors.  This, in turn, simplifies the subsequent evaluation
and contractor selection tasks and helps assure ultimately satisfactory
performance by the contractor.  The proposal request and proposal evaluation
activities are so closely interrelated that each section of the RFP should
be developed with the idea in mind of evaluating the contractor's response
to that section.

Prior to writing the RFP, consideration should be given to at least the
following questions:

     1.  What is the type and level of planning, purpose and importance
         of the planning in which the Purchaser is engaged or about to
         undertake; what is the role of modeling in the overall project;
         and, what in-house resources are available to complete or
         participate in completion of the work?

     2.  Is a contractor needed to provide expertise or additional manpower;
         and, what are the specific qualifications or expertise required
         for providing the needed assistance?

     3.  On what basis will the prospective contractors be evaluated
         and a successful contractor be selected?

     4.  What type of contract (i.e., fixed price, time and materials, cost
         plus fixed fee, etc.) is expected to result from the procurement?
                                    117

-------
     5.  How will  the contractor's performance be monitored,  and what will
        constitute  satisfactory performance?

     6.  In what way will the contractor's end results aid  the Purchaser
        and  the  agencies working  with the Purchaser in achieving  their
         objectives?

These, broad questions  should be answered, at least  in a preliminary form,
at the time preparation of the RFP is begun, rather than during or after
preparation.   The answers should largely be the  result of an early analysis
of what level of modeling, if any, is required.   The first and last questions
in particular require that detailed consideration be given to the  relation-
ship between modeling and other parts of the planning process.

The remainder of this section gives some suggestions for the structuring  and
content of an RFP.  The subheadings that follow may, with minor variations
as appropriate, be the ones used in the RFP.  Administrative, legal and
technical knowledge are needed to compose a complete RFP.

Program Objectives

Before undertaking preparation of the RFP,  the originator should have at
hand a clear statement of the problem,  the  objectives of the project, and
the  desired end results.  In addition,  he should possess some knowledge of
the  procedural requirements  to be followed  in the procurement process.
Assembly of this  information requires a joint effort  of the planning,
administrative, and/or legal staff members.  At this  point,  it may not be
clear yet  what type of water quality model  is needed  or exactly what types
of data might be  required to support  their  use.  However, the RFP should
leave no doubt about what specific problems the Contractor  is  to  address,
what specific end items  he  is  to  deliver, and what  these deliverables  are
to accomplish for the Purchaser.

A good  definition of  the work to  be  contracted  is  highly desirable and
 aids the  Purchaser  in several  area including:
                                     118

-------
      1.  Preparing and organizing the entire RFP document;

      2.  Estimating the scope, duration and nature of the needed consulting
          services;

      3.  Assessing the probable data requirements and the availability of
          data;

      4.  Determining the type of contractual arrangement best suited for
          the  desired services;

      5.  Evaluation of proposals and selection of Contractors;  and

      6.  Ensuring that the entire contract effort will be aimed at  producing
          the specific end items needed to satisfy job requirements.

 A clear statement of  the problems to be  addressed by the Contractor  and
 the specific objectives  to be met also helps the potential contractors to
 Prepare responsive proposals.   An RFP which unambiguously identifies the
 Purchaser's  needs and  the  Contractor's responsibilities  is particularly
 beneficial for a number  of reasons which,  among others,  include:

      1.   Encouraging  contractors  to  concentrate proposal efforts on  meeting
          the prescribed  needs  and  responsibilities,  rather than attempting
          to  interpret the  Purchaser's  requirements;

     2.   Reducing  the need  for  the Contractor to request  clarification
          of  requirements as he prepares his proposal.

     3.  Encouraging a greater number of responses through a reduction both
          in  the  cost (and risk) of proposal preparation and in the uncer-
         tainties of estimating the cost of project performance.

The description of the technical objectives of proposed contractural services
^ seldom undertaken by administrative staff.  However, contractural arrange-
                                   119

-------
merits for water quality modeling often have specific objectives closely
related to administrative decisions.  One commonly encountered example is
that of training.  If an administrative decision has been made to develop
in-house capability in the modeling area, a contractor may be expected to
provide training in addition to the performance of selected technical tasks.
If this is the case, appropriate objectives should be identified by the
administrative staff to guide that portion of the effort.  Other areas for
administrative  consideration include the desired visibility of consultants
through participation in public meetings, their relationship to other involved
agencies and the matters of how work is performed, transmitted, and
coordinated.

The major portion of the effort required to accurately describe the technical
objectives of the services to be contracted is normally the responsibility
of the planning staff.  The description of objectives should be as specific
as possible.  As an example, a statement such as "the objectives include
selecting, verifying, and applying a model to predict water quality" in a
river, lake, or estuary leaves many questions unanswered.  The objectives
could, in addition, very usefully describe the particular constituents to
be modeled, the desired accuracy of prediction, the specific reaches of
water to be modeled and other basic information.

In those cases where the planning to be undertaken is guided by well
defined federal or state programs, quoting or referencing legislatively
established objectives or those included in published guidelines may be
useful.

Scope of Work
Once the technical and administrative objectives of the  project have been
determined, the extent and  type of consulting services which  are needed  can
be established.  It is usually advantageous to prepare a Scope of
Work statement.  If prepared, the Scope of Work should summarize
                                      120

-------
the individual tasks to be performed and the expected end items (usually
called deliverables) to be provided by the Contractor.  The Scope of Work
should also explain the specific responsibilities which the Purchaser will
assume to aid the Contractor in carrying out his work.

The Scope of Work should identify the practical limitations on the
services to be performed by the Contractor.  These limitations may be due
to the portion of the overall planning budget allocated to the modeling
related services, the available data, the desired time schedule, or other
more subjective considerations.

It is important for all Proposers to clearly understand the Purchaser's needs
and to have a realistic concept of his resources.  This information is best
conveyed to the Proposer as an integral part of the written RFP as oral
communications about the scope of work are too easily misinterpreted.  If
discussions concerning the Scope of Work take place during preproposal
interviews, it should be clearly impressed upon the Proposer that such comments
are informal and that the written Scope of Work provides the official
description of the services desired.

Background of Program

The RFP should contain a section presenting a sufficient amount of appropriate
background information to familiarize Proposers with the subject of the
solicitation.   The background statement may include historical information
and describe existing conditions, expected developments, and any major
decisions which are already made or pending.  It should also describe any
unusual complexities which might arise during the project.  The foregoing
and any other considerations should be presented which, in the Purchaser's
judgment, are necessary for a Proposer to acquire a good understanding of the
reasons for issuing the RFP and the problems addressed in the RFP.

Background information of a technical nature is particularly useful to
potential Proposers in sizing up the problem and in anticipating the manner in
                                    121

-------
which the project will be pursued.  Information on past studies and the
principal conclusions thereof, pending physical developments, and imposed
schedules supplements and assists in interpretation of the Scope of Work
and Objectives sections of the RFP.  Particular agencies which will be
participating in the project  should be identified and the procedures for
coordinating their respective efforts should be stated.  In most cases,
specific  identification of the particular program  (e.g., areawide waste
treatment management  planning pursuant to Section  208 of Public Law 92-500,
or wasteload allocations  pursuant to  Section 303 of Public Law 92-500)
 toward which the contracted  services  are  to contribute will  aid  considerably
 in conveying the expected level  of detail.

 Depending on  the nature of the  consulting services requested,  the necessary
 background information may be condensed  into  a short  paragraph or it  may
 require an extensive discussion with illustrations.  In the latter case, it
 may be convenient to provide the narrative and/or the illustrative material
 as an appendix or separate attachment to the RFP.

 It is appropriate in some cases  to provide the necessary background information
 orally to each Proposer  during a pre-qualification interview (Section 6.4.2).
 Distributing this material  in writing, however, assures that all Proposers
 receive  exactly  the  same information and, consequently, that none of them
 obtain a privileged  position.

  Technical Information  and Data

  Depending on  the nature of  the  planning  which is to  be conducted,  the water
  quality models and/or  modeling-related  services which are needed may have
  substantially different data requirements.   Some of  the needed data may be
  available from various  studies carried  out previously, while others must be
  measured, estimated or  assumed.

  All of  the types of data thought to be needed should be identified and its
   availability determined before  preparation of the RFP.  In addition to
   identifying its availability, the validity of data from various sources should

                                       122

-------
be examined. Comparison of suitable existing data to the data needed for
the various types of models (see Section 4.2.5)  will indicate the magnitude
of any data deficiencies.

Data deficiencies warrant special attention due to the high cost which can
be encountered in field investigations.  Needs for data which are identified
may be met in some cases by use of "typical" values from the literature at
a cost far less than that for field collection of data.  The suitability of
one method over another depends on the particular type of data needed, the
sensitivity of model results to factors dependent on the missing data and
expectations of model accuracy.  Care should be taken to avoid requiring a
higher degree of accuracy in data inputs than  is justified by the intended
use of the model results.

Using the results of the preliminary investigation  of data needs and avail-
ability, the Purchaser should carefully design the  portions  of  the  RFP
which deal with  the data related aspects of  the project.  It is  generally
convenient  to sub-divide this portion  of the RFP  into  sections  dealing  with
Purchaser supplied  information  and  Contractor  supplied information.

Purchaser Supplied  Information.  In this section of the RFP the Purchaser
should identify  and provide the types  of technical information which will
enable the  Contractor  to define the various tasks of his proposal in sufficient
detail for  accurate identification of  the  associated time and costs.  Data
needed for  the  Contractor's proposal preparation effort should be presented
 in the body of  the  RFP if it is not too extensive.  If the information
 cannot be  conveniently presented,  then appropriate and readily available
 references  containing such data should be  listed in the RFP.  Requiring
 contractors to  spend  an excessive amount of time and effort to secure the
 information necessary for proposal preparation could discourage presentation
 of a thoughtful and productive proposal which might otherwise be forthcoming.
                                     123

-------
The RFP should outline In detail the types of data that would be made
available to the selected Contractor for the performance of the contract
and the form and sources thereof.  Data to be used in the work effort may
include information which is already available to the Purchaser or to be
gathered and compiled by the Purchaser prior to or after award of a
contract.  The RFP should also describe or note any information available
from various sources which is subject to approval and/or modification by
the Purchaser before its use.

Finally,  the data  thought or known  to be needed which  is not  available  from
 the Purchaser  should also be identified.   Such information may already  be
 available to  the Contractor from other  sources or have to  be  collected  and
 compiled by the Contractor after award  of  a contract.   Alternatively,  some
 information may not  be available from any known  source, and must therefore
 be assumed by the  Contractor  subject to approval by the Purchaser.

 Contractor Supplied Information.  While the RFP  should identify in detail
 at least the types of data that the Purchaser is capable of  and planning to
 supply to the Contractor, the Contractor-supplied data may be identified in
 broader terms.   It may sometimes be appropriate to include a requirement in
 the RFP for the Bidder to state all data requirements  for the project,
 whether or not  the Purchaser is able to supply certain types  of information.
 In this case,  the Bidder should be requested to identify  overall data  require-
 ments,  suggest methods for acquiring all  input data,  and  discuss the import-
 ance  or value  of  each type of  information for the  purpose of  the modeling  o*
 other effort  to be contracted.  Analysis  of the  responses to  these  types  of
 requirements will give a good  indication of the  Bidder's  understanding of
  the  problem,  the  project objectives,  and the type  of  services desired.

  Estimated Time and Effort

  It is in the Purchaser's interest for the RFP to indicate his commitments and
  resources, as well as his  needs.  When preparing a proposal, a Bidder should
  have a reasonably good idea about the desired schedule and/or completion date
  and  the general level of effort desired by the Purchaser.

                                     124

-------
  The RFP may indicate the desired level of effort by stating  an  anticipated
  or approximate dollar value for consulting services.  Alternatively,  the
  RFP may indicate the Purchaser's estimate of  the magnitude of the needed
  Contractor's effort in man years or man months.   Finally, the RFP may
  specify only time schedules,  tasks, and deliverables and ask the Bidders
  to estimate the cost of  services independently.

 Availability of information concerning  the approximate number of dollars to
 be spent can partially compensate for inadequacies in the description of work
 required.   However,  the manner  and detail of information provided may be
 subject to  strict control.  A legal review of this point should be obtained
 before issuance of an  RFP containing such information or before responding to
 inquiries.

 By all means, care should be taken not to expose the amount  of money
 certified or otherwise available when the procurement is essentially a
 competition between qualified bidders based on the lowest price.  Likewise
 every effort must be made to avoid exposing such information to  one  and not
 another of the bidders.

 The various preceding considerations  are vital to determination  of the
 price which will eventually be paid for the requested services.  Whichever
 approach is chosen,  the Purchaser should keep  in mind that buying professional
 engineering services is different from  buying  equipment or non-professional
 labor,  and  differences in prices do not  tell the whole story  about the amount
 and quality of services rendered.  An unsatisfactory situation may easily
 occur  if  the price ultimately agreed upon  is not  equitably matched to the
 expected products or  services.

 Project Organization and  Work  Statement

The RFP should specify, in as much detail as practical,  any individual
tasks, steps or phases  into which the Purchaser desires the work to be
divided.  The products or deliverables which are expected to  result from
each increment of the project should be identified.  The section of the RFP
                                    125

-------
containing this description may be titled "Statement of Work",  "Organization
of Work", "Guidelines for the Contractor", "Tasks and Deliverables",  or
some other heading appropriate for the particular project and the RFP.  By
whatever name it is called, the purpose of this section of the RFP is to:

     1.  Organize  the project into major  tasks;

     2.  Define the deliverable end  items at the conclusion of each  task;
          and
      3.  Pave the way for the definition of  a contractually binding work
          statement for the selected Contractor.

 The Purchaser is well advised to describe each task and each deliverable
 in detail if adequate information and knowledge is available concerning the
 consulting work to be performed, the requirements for and availability of
 input  data, and the expected end results.  A detailed work  statement may
 enable the Bidder to  size up his tasks with  sufficient confidence  to
 quote  prices without  consideration of any additional cost  for contingencies
 due  to uncertainty.

 The  inclusion of a detailed work statement  in the RFP  is not always
 practical and may involve considerable risk because of the nature of  the
  project.  In such a case, the RFP should encourage or  require the Contractors
  to describe the tasks and deliverables in detail in their proposals.   In
  this  event, the detailed definition of tasks and the structuring of the
  project becomes one  of  the most important requirements of  a responsive
  proposal and  should  specifically be discussed in the RFP.

  The RFP should  require  each proposal  to include  a  complete Statement of
  Work  in which each major task and deliverable is individually  identified
   and described.   If possible,  each major task and deliverable should  also
   be priced out separately.   There is an advantage in avoiding a too general
   work statement with a large single lump-sum cost.   Contracts of this type
   are  difficult to manage and the price is  difficult to substantiate.
                                      126

-------
Emphasis in pricing should be placed,  however,  on major tasks (e.g.,  no
less than $5,000 to $10,000 per task), since too many small tasks
unnecessarily complicate contract management and can escalate administration
expenses.

A logical and consistent format should be used in the RFP for specification
of tasks, division of the work into phases and other matters which help
organize the technical portions of the project.  If overall study schedules
are established and flow charts prepared, they should be presented and may
provide the needed organizational concept through the time sequencing of
activities.  In the absence of a sound schedule, activities can be grouped
according to their type such as inventory, analysis, plan formulation and
evaluation, and review.  However done, the organizational concept should be
stated and instructions given as to how proposals should relate to the concept,

The difficulty of proper and timely evaluation of a number of proposals
places a premium upon obtaining some uniformity among the proposals presented.

Project Monitoring

The RFP should deal explicitly with the procedures which the Purchaser will
follow in reviewing and approving the Contractor's performance.  A section
of the RFP should be devoted to this and, at a minimum, should describe  the
following:
     1.  All major points (milestones) of review and approval;

     2.  Reports required of the Contractor including periodic progress
         reports, oral presentations,  written or published interim reports,
         papers, technical notes, deliverable items specified in the Contract
         Work Statement, and final reports;

     3.  Statement of the time periods required for Purchaser's approval of
         each Contractor supplied report or deliverable item; and

     4.  Description of the mechanics of the reporting, review and approval
         process with particular attention to who receives what report in
                                     127

-------
         how many copies, and what constitutes approval.

The clear definition of the review and approval process is very important,
especially if deliverables other than periodic progress reports are to be
provided.  Overall monitoring of project progress can be facilitated if
deliverables subject to acceptance or approval by the Purchaser mark the
conclusion of major tasks.  Such deliverables, if agreed to represent task
completion when approved, enable both parties to measure progress more
accurately.

Care should be taken in describing the arrangements for monitoring projects
to assure the contractor anticipates and commits to supply meaningful
information.  Periodic written progress reports should follow a carefully
prescribed format which facilitates comparison of consecutive reports.  In
addition, the detail of technical information to be contained in progress
reports should be specified.  Any Contractor supplied estimate of completion
should be clearly defined as to whether it is to be on the basis of dollars
expended, hours of effort, or some other measure of progress.  Written
progress reports should also require the contractor to explicitly identify
and describe any difficulties encountered in carrying out the work effort
and any actions necessary on the Purchaser's part to rectify the situation.

In addition to written progress reports, it may be useful to require the
participation of the contractor in periodic meetings of all key project
participants to enable and assure coordination of all aspects of the planning
process.

Project Staffing

As a rule, the RFP should request the Contractor to provide a comprehensive
staffing plan for the project.  Contractor supplied information should include
the names of all key personnel the Contractor expects to be involved in the
project and the extent of their availability for project work.
                                   128

-------
The RFP should require that the Contractor's Project Manager or Principal
Investigator who will be responsible for the Contractor's technical and
financial performance on the project be identified by name in the
proposal.  A statement should also be required which outlines his
responsibilities to the  Purchaser,  his  position  in the Contractor's organization
and the authority given him to represent the Contractor in prosecution of the
project.  As appropriate, other key personnel should also be identified by
name, position, related experience and the nature of their assignment to
the project.  For a major contract, other experienced personnel who are
available to the Contractor and may be assigned to the project on an "as-needed"
basis should also be identified.

A description outlining staff assignments planned by the Contractor for each
major task of the project should be requested whenever appropriate.  Such
a description may include the specific responsibilities of each key
individual on each task, and the percentage of his time to be spent on
this assignment over a specified period.  Alternatively, the RFP may ask for
identification of the number of professional hours or days anticipated to
be contributed by each individual to each task.  The availability of the
key personnel who are identified for particular assignments or whose
qualifications are particularly cogent should also be stated, and possibly
guaranteed, in the proposal.  To enable evaluation of the availability of
the staff proposed by the Contractor,  it is helpful to request identification of
any other projects for which the proposed staff members have concurrent
responsibility.

In all considerations of project staffing, the Purchaser should bear in
mind that Contractors' employees are subject to variations in their
performance, and that Contractors may  sometimes experience staffing
problems unforeseen at the time of proposal.  Reasonable flexibility in
Contractor staffing should be anticipated and allowed for in the Purchaser's
enforcement of any submitted staffing  plans.  As  a general rule, changes  in
proposed staffing should not be a source of conflict as long as  the Contractor's
performance is not detrimentally affected.
                                     129

-------
The contractor's capability to coordinate his efforts with those of other
contractors and/or in-house staff may be important if the modeling or
modeling-related services to be procured cannot be clearly separated
from other portions of the planning process.  In this event, the RFP
should specifically request the Contractor to identify the staff proposed
for this function and to describe their prior experience and other
qualifications with emphasis on planning management.

Project Schedule
One of the most important aspects of management of a planning program is
its timely execution.  The high level of interdependency between various
planning activities makes it essential that all parties responsible for an
identifiable portion of the work, including contractors, have well defined
schedules.  Adherance to schedules generally becomes increasingly important
as the complexity and scope of planning increases and as additional disciplines
are incorporated in the study program.

The RFP should require the Contractor to provide an explicit schedule for
completion of the proposed work.  The schedule should include the time
for performance of all separately identified tasks, the dates of deliver-
ables, schedules for presentations, Purchaser's approvals, report production
and publication.  It is generally unimportant whether schedules are
summarized in tabular, bar-chart or PERT-chart form.  Whether constructing
a schedule to be included in the RFP or reviewing a Bidder proposed
schedule, the reasonableness of the individual work items and deadlines of
the schedule should be very carefully examined.  For example, schedules
should be evaluated to assure adequate time is provided for data collection
after allowance for unforeseen difficulties and seasonal limitations, model
verification, report preparation, report approval, staff training, and other
project tasks.  Schedules should also contain some provisions for contingencies.
It may also be useful for the Purchaser to determine how the proposed
schedule compares with those of other similar projects, and with the
schedules of any other major projects which a Contractor may have underway
or planned.

                                      130

-------
The analysis of any requested or proposed schedule should be sufficient to
convince the Purchaser of its feasibility, any assumptions of abnormally
high level of effort or efficiency on the part of the Contractor's or
Purchaser's staff members.

Project Costs

The cost of the modeling or other services must be addressed in the RFP
regardless of the role of price in the evaluation of proposals.  It is
convenient to devote a separate section of the RFP to a discussion of
the project costs and the specific cost information expected to appear
in Contractor's proposals.

Each Proposer should be required to state the total price of his proposed
services.  This price may be a single figure of total cost in the case of
a firm fixed-price contract or a price ceiling not to be exceeded without
prior approval by the Purchaser in case of a cost-plus-fixed-fee contract.
As may be applicable, the RFP should state either that any modification of
the price wil- require renegotiation of the contract, or that requests
for price increases cannot be considered.  If permitted, any modification
of the cost after execution of a contract should always be based on a
demonstration of clear advantage to the Purchaser.

If the Statement of Work adequately identifies all major tasks, it is
often to the Purchaser's advantage to request a breakdown of the total
cost to separate major tasks.  This approach encourages good financial
management of the project by the Contractor and enables closer monitoring
of the Contractor's progress by the Purchaser.

With the notable exception of the fixed-price contract, most types of
contractual arrangements require quotation of the Contractor's labor rates
by categories of personnel, as well as identification of the type and
amount of the various burdens which will be added to labor and direct costs,
                                   131

-------
Any details concerning cost which are requested by the RFP should preferably
be submitted as a separate cost proposal accompanying the technical
proposal.  While the latter may include the total price but not the cost
breakdown, it is desirable to completely eliminate cost information from the
technical proposal in the case of competitive bidding.  The inclusion of all
cost information in the cost proposal enables easier handling of the technical
proposals during evaluation if the Purchaser desires to insure the technical
evaluation is separate from and not influenced by price considerations.  The
separation of cost and pricing data from technical proposals should
not prevent evaluation by the technical staff of the Contractor's proposed
level of effort.  Proposers should be requested to include labor hours and
other non-dollar information in the technical proposal.

Provisions for cost escalations or cost overruns should be stated in the
RFP when such escalations may reasonably be expected and will be permitted.
Similarly, the RFP should describe any penalties which will be imposed for
non-performance, unsatisfactory performance or excessive delays due to
negligence on the part of the Contractor.

Acceptance Period

The RFP should identify the time period for evaluation of the proposals
and for entering into contract negotiations with the successful Proposer.
The Contractor's proposal should be required to state an acceptance period
for which the proposal remains in effect, in conformance with the time
period requested in the RFP.

Contractor Qualifications
The RFP should request the Proposer to furnish relevant information on
specific factors that qualify the Contractor to provide the particular
services which are sought including:

     1.   General background and history of the Proposer's organization
         in providing modeling-related services and in water quality
         planning;

                                   132

-------
     2.   Staff  organization;

     3.   Office locations  and  facilities pertinent to the project;

     4.   Specific experience of  the  Proposer's organization and of
         individual staff  members  in the detailed subject matter of  the
         RFP including familiarity with the specific models to be used,
         the use of models for water quality planning,  and the type  or level
         of planning to be undertaken;

     5.   Specific experience and staff  qualifications in the  general area
         of water quality  modeling and  water quality planning;

     6.   References to similar or  analogous projects completed and/or
         underway by the Contractor;

     7.   Technical and financial references; and

     8.   Biographies, resumes  or other  appropriate  information concerning
         key staff members and consultants.

Basis For Evaluation
The procedure to be followed in the proposal evaluation process and the
main factors to be considered should be stated in the RFP.   This information
is important as it guides the Proposer in preparing a proposer responsive-to
the needs of the Purchaser and which can be easily evaluated.

A separate  section of the RFP can be devoted to proposal evaluation.  This
section  can be brief, but should include among others,  the following items
of information:

     1.  The major criteria for proposal evaluation such as  overall
         quality, understanding of  the problem, soundness of technical
         approach, data requirements, ease  of implementation, contractor
         qualifications, price, or  others  important to the Purchaser;
                                   133

-------
     2.  The approximate or relative weights to be given to each criteria
         in the evaluation  process;

The RFP should state if the evaluation of technical and cost proposals
will be done independently.  In general, it is preferable to handle the
technical evaluation completely independently from the cost evaluation
and consider prices only after the tefchnical evaluation is complete.  This
procedure eliminates the need for detailed analyses of the cost and pricing
data for technically inadequate proposals.

The role of a consultant in assisting the Purchaser to prepare the RFP or
evaluate proposals should be clearly stated in the RFP whether or not the
Consultant is identified.  In general, identification of any consultant
involved in the evaluation process is preferable.

The evaluation of proposals can be faciliated by instructing Proposers in
regard to the format to be used.  The objective of such instructions is
to have the principal proposal elements relating to the evaluation criteria
clearly identified.  Proposer compliance with organizational instructions will
minimize the time and effort for the comparison of proposals to the
requirements of the RFP.

Establishment of the criteria for technical evaluation is an important
step toward eventually obtaining the most suitable services.  If criteria
are general and non-specific, they may not provide an adequate explanation
to Proposers of those points thought most important by the Purchaser.
Development of the appropriate technical criteria is largely a planning
responsibility.
                                     134

-------
Instructions To Proposers
The RFP should discuss the procedures to be followed by the Proposer in
submitting the proposal and the Purchaser's method of handling proposals
The discussion of these points should include a description of:
     1.   The names and address of the individual or department to the
         technical and cost proposals should be mailed or delivered;

     2.   The deadline for delivery of proposals and whether the date   of
         postmark or date of receipt determines compliance with the deadline;

     3.   The Purchaser's handling of proposals (e.g., complete confindential
         treatment;  non-disclosure of proprietary information only; or full
         disclosure, as well as the ownership or return of submitted
         proposals);

     A.   Notification to Proposer of receipt of proposals;

     5.   Procedures  for Proposers to follow in requesting clarification of
         points in the RFP during proposal preparation and the Purchaser's
         method of responding to such requests (e.g., questions to be
         submitted and answered in writing only; questions to be asked
         at Bidder's conference only; or any other preferred method);

     6.   Notification to Proposers concerning the return, if any, of  their
         proposals;
                                       135

-------
     7.  Any oral interviews and discussions of proposals which are
         planned, the basis on which Proposers will receive such invitations,
         and the method of conducting interviews; and

     8.  The method of initiating contract negotiation with the successful
         Proposer.

Any instructions necessary to protect the Proposer's interests such as
marking of proprietary information should be specified.  Likewise, instructions
to protect the Purchaser's interests such as formalizing any discussions
during the proposal period should be explicitly described.  Specific dates
and places of Bidders' conferences should be identified if they are
scheduled in advance.  Possible contingencies to contract execution such
as agency approval or availability of funding, should be clearly stated.

If the Purchaser represents several agencies who will participate in the
evaluation, contracting or contract monitoring procedures, all parties
should be clearly identified in the RFP.

6.4  PREPARING THE BIDDERS LIST

Before the RFP can be made available to Contractors the means of distribution
should be explicitly decided.  The adequate review of proposals for
complicated technical efforts is time consuming and expensive.  It is often
desirable to determine to whom the RFP will be provided before its first
release.  This determination generally includes the identification and
qualification of prospective contractors and establishment of a List of
Bidders.

Identification Of Prospective Contractors

There are two alternative approaches for compiling a list of firms or
individuals from which the prospective contractors (i.e., Bidders) will be
selected.  A direct solicitation of Contractor interest may be undertaken
through advertisement, announcements, press releases, or other similar
                                     136

-------
 means.  Alternatively an indirect solicitation of promising prospects may
 be made using information from other sources such as federal and state
 agencies which have used various contractors for similar purposes.

 Governmental purchasers are frequently required by law or regulations to
 award contracts on a competitive basis in certain situations.   In such
 cases, it may be expedient to solicit an expression of interest from
 prospective Bidders through some type of widespread advertisement.   This
 approach makes the Purchaser's plans known to a large cross-section of
 potential Bidders and the Purchaser  may establish contact with  some very
 capable and qualified consultants theretofore unknown to him.   The  dis-
 advantage of solicitation of  advertisement  is the effort and time required to
 receive and sort  out  the  many responses,  only some of  which may be  relevant
 to  the Purchaser's  needs.   It also entails  the  burden  of time and cost required
 to  answer the many personal or telephone contacts that may follow as a
 result of such solicitation.   If  an  advertisement is  used,  it should
 advise that its purpose is  solely that  of securing an expression  of interest
 by  potential Contractors, and that receipt  of responses will not  be acknowledged
 nor will  the Purchaser be obligated  to  issue  an RFP  to any of the respondees.
 In  addition,  the  Purchaser  may choose  to  advise that  the procedure  and
 results of  the pre-qualification  process  will not be disclosed.

 Qualification Of  Prospective  Contractors

 The objective  of  either of  the  approaches described in the preceding section
 or a combination  of them  is to  establish a  list  of prospective Bidders.
 This first  list may be expected to require  some  screening  in order  to
 arrive at the List of Bidders to whom the RFP will be  issued.

Depending upon the number of prospective contractors identified and  the
number of Bidders to be identified, several steps may be necessary  to carry
out the qualification process including reference checks, pre-qualification
and selection of Bidders.
                                    137

-------
Reference Checks.  The method of checking references is particularly
important if the pitfalls of irrelevant information are to be avoided.
Reference checks should involve contacting more than one source of
reference.  A single good recommendation may not assure the prospective
Contractor's capabilities or his suitability to the Purchaser's specific
needs.  Similarly, care should be taken that a single unfavorable comment
does not eliminate an especially well-qualified consultant.  Reference
checks should attempt to identify the reasons behind either a favorable
or unfavorable opinion and the context in which any recommendation is
made.

Pre-Qualification.  The preliminary list of prospective bidders established
after the initial screening of expressions of interest and references must
generally be further narrowed to a final List of Bidders to whom the RFP
will be issued.  While the exposure to a number of alternative approaches
and different ideas is highly desirable, the Purchaser's resources for
assimilating and evaluating proposals may be limited.  What constitutes a
reasonable number of bidders depends on the particular situation.  If a
medium range plan requiring extensive and detailed analysis of numerous
waste water management alternatives is to be prepared and modeling is
intended to be used as a principal tool, then the sizeable costs involved
may make it worthwhile to carefully study a large number of proposals.

Several procedures may be used to reduce the number of potential bidders
to the few which are best qualified.  The two methods most commonly used
for this purpose are requests for written pre-qualification statements and
the conduct of oral pre-qualification interviews.

Both approaches have their advantages and disadvantages.  Written pre-
qualif ication statements provide formal brief but well-organized material
which can be evaluated efficiently.  However, reliance upon written
statements does not provide personal contact with the contractor's key
staff members or the opportunity to ask and pursue relevant questions.
A personal meeting with the contractor often gives additional insight into
                                    138

-------
his capabilities and the potential working relations which might be
established.

The oral interview provides a great deal more flexibility and requires
less preparation on the part of the Purchaser.  However,  it requires the
participation of more Purchaser staff members in the meeting and in a post-
meeting evaluation and it is likely to lead to a more subjective evaluation
of the prospective bidders.

If a request for written "Statements of Qualifications" is issued,  the
Purchaser should summarize the background and main objective of the planned
project in the request so that the recipient can judge the applicability of
his experience and qualifications.  It is to the Purchaser's advantage to
state clearly those specific questions he wants to be answered in a summary
form.  In addition, the request for the statements should prescribe that
the statements meet any criteria important to the Purchaser.  Frequently,
such instructions require statements to do the following:

     1.  Be brief (e.g., not more than a certain number of pages specified
         by the Purchaser; omit necessary illustrations or irrelevant
         material);

     2.  Be factual (e.g., refer to accomplished facts which can be
         easily checked rather than generalities or future plans);

     3.  Answer specific questions; and

     4.  Demonstrate only relevant experience, capabilities and facilities.

If oral interviews are conducted for pre-qualification they should:

     1.  Provide equal time for each party interviewed;

     2.  Attempt to provide the same information to and ask the same
         questions of each party interviewed; and

     3.  Permit the prospective Bidder to use his own discretion in
                                    139

-------
         presenting his qualifications within the time limit set forth
         in advance.

Selection of Bidders.   After the prospective Bidders have submitted their
qualifications either in writing or via interviews,  the evaluation and
selection of Bidders may take place.  Preferably this should be performed
by more than one member of the Purchaser's staff and is frequently
accomplished by a committee effort.  Assuming a committee is assigned for
the task, the evaluation procedure might include the following sequence
of steps:

     1.  Establishment of a ranking of the prospective Bidders by each
         member of the committee including a brief justification for the
         recommendation to "retain" or "eliminate",  and for the ranks
         assigned in the group of prospective Bidders recommended to be
         retained;

     2.  Committee review and comparison of the recommendations of its
         individual members.  Should the recommendations widely differ for
         certain prospective Bidders, the members of the committee should
         jointly examine each such case and try to reconcile differences
         in opinion;

     3.  Re-evaluation by individual members of the committee of any
         unresolved cases after having heard the arguments for and against
         their inclusion in the List of Bidders; and

     4.  Decision by the person having the final authority  (e.g., the
         appointed Project Manager, Department Head, etc.) if, after an
         additional joint meeting, some cases still remain unresolved.

The number of prospective contractors to be finally retained and issued the
RFP depends on the nature of the project and the desired consulting
services as well as on the ability of the Purchaser to give a thorough
evaluation to each proposal submitted.  In some cases  the list may be
limited  to two or  three firms while for other projects inclusion of up

-------
 to  a  dozen  competitors may be warranted.

 Bidders who are considered to be only marginally qualified should not
 be  placed on  the list.  The cost of preparing a proposal for a sizeable
 complex project is substantial and only the viable candidates should be
 requested to  bear this expense.  Issuing the RFP to a candidate whom the
 Purchaser considers unqualified or unpromising is a disservice to the
 contractor  and has no benefit to the Purchaser.

 If  the Purchaser becomes convinced prior to receipt or review of proposals
 that  one of the prospective contractors has far superior qualifications
 to  do the job, there may be no reason for competitive bidding.  Unless the
 Purchaser is  prepared to make reimbursement, it is unfair to request
 Bidders to  undertake the effort and bear the cost of proposal preparation
 just  to "keep the other fellow honest".  In such cases, the Purchaser may
 be  better advised to concentrate his efforts on justification of sole-source
 procurement and negotiation rather than on the handling and review of
 numerous proposals.

 The Purchaser should avoid premature committments to any one Bidder during
 the qualification process.  Pressure by Bidders or their representatives
 should be discouraged, and the entire qualification process should be kept
 on a strictly formal basis.

 Regardless of whether written statements, oral interviews, or both are
 used to identify the best qualified potential contractors, the planner must
 be prepared to elicit responses which are factual and which will be of
 real usefulness in reaching a decision.  Questions to be used as a basis for
written statements or asked during oral interviews should be carefully
 prepared and framed  in such specific terms that evasive or incomplete
 answers are minimized.

 Both fairness and the need for information make it desirable to ask the
 same basic questions of all potential contractors.   Accumulating an
                                    HI

-------
expanding list of questions as interviews proceed may bias the interviews
to the detriment of the Purchaser or the potential contractors.   However,
a portion of each interview might be reserved for extemporaneous discussion
of particular aspects of the potential contractor's presentations and
experience, clarification of any unclear or apparently conflicting state-
ments, and for other purposes.  Planners should be particularly alert to
the opportunity to gain insight into the interviewee's understanding of
the needs and procedures of integrating modeling activities with other
planning activities.

6.5  SOLICITATION OF PROPOSALS

Subsequent to the identification of the List of Bidders by whatever
procedure is chosen, the RFP can be distributed.  To insure fairness,
care should be taken that all copies of the RFP are mailed or otherwise
distributed simultaneously.  After the distribution of the RFP, the
Purchaser must be prepared to respond to questions and to receive and
evaluate proposals.

Responding To Questions And Contractor Contracts
Proposers may have legitimate questions prior to receiving the RFP as
well as during the proposal effort.  Questions relating to the Proposers'
understanding of the Purchaser's needs deserve careful attention and full
answer.  However, due to the nature of competitive bidding, Proposers may
attempt to obtain privileged information which would strengthen their
competitive position.  To avoid future misunderstandings and potential legal
complications, the Purchaser's contacts with all prospective Bidders are
best kept at a formal level from the initial steps of contractor consider-
ation until the selection process  is complete.  The maintenance of formal
relations can be encouraged by the following:

     1.  Designation of one person  (Purchaser Representative) on the
         Purchaser's staff to be responsible for handling all contacts
         with prospective contractors.  All inquiries and unsolicited
                                     142

-------
    contacts can be referred to this Individual.  During the
    bidder-qualification phases, the name, title and phone number
    of the Purchaser Representative should be given to all
    prospective Bidders for future contacts.

2.  Directing the Purchaser Representative to make a brief note or
    memorandum about each personal contact or phone conversation where
    information of any significance to the Purchaser's planned project
    is requested or exchanged.

3,  Maintaining contacts on a formal and noncommittal basis.  The
    Purchaser Representative may listen to ideas, but should not
    solicit any suggestions.

4.  Discouraging prospective Bidders from disclosing any information
    they consider confidential, proprietary or in the nature of a
    trade secret.  Prospective Bidders should be warned against
    informally disclosing any information which they would request
    the Purchaser to keep confidential.

5.  Discouraging inclusion in proposals of proprietary information
    unless the nature of such information is clearly identified in
    the proposal and the RFP has specifically provided for confidential
    treatment of such marked proprietary information.

6.  Restricting informal discussions between Proposers and the Purchaser
    Representative after the RFP has been issued.  Questions by each
    Proposer can be sumitted in writing.  For fairness, such questions
    can be answered by the Purchaser in written form with both the
    questions and answers distributed to all Proposers (there is no
    need, however,  to identify the Proposer asking the question).  An
    alternative method of avoiding informal or unfair discussion is
    organization of a joint meeting with all Proposers to answer all
    questions (Bidder's Conference).  Minutes of such meetings which
    summarize the answers to the questions asked, should be
    distributed to  all Proposers.
                                1A3

-------
A potential conflict may arise if, in the judgment of the Purchaser,  a
particular contractor is likely to provide services far superior to those
of any other prospective contractor, but his superior qualifications  are
a result of certain proprietary ideas, techniques or trade secrets.  Where
permissible by law, the Purchaser may consider engaging such a contractor
on a sole-source procurement basis.  In a competitive situation, all  potential
Proposers should be given an opportunity to present their ideas and approaches
and demonstrate the special benefits of each to the Purchaser prior to the
selection of the List of Bidders and finalization of the RFP.

Another potential problem concerns the possibility of a Proposer submitting
an alternative proposal, or a proposal with ideas and approaches different
from those requested in the RFP.  Alternative proposals should be admitted
for detailed evaluation only under very special circumstances, including,
among others, the following:
     1.  They meet all stated objectives of the RFP;
     2.  They differ from the requested approach only in details, and
     3.  They meet, or preferably exceed, the requirements of the RFP.

There is no single "best" solution to the problem of handling proposals
with alternate approaches or with exceptions.  If an alternate proposal is
particularly attractive from both the technical and cost aspects in part or
in whole, the Purchaser may consider negotiation over the approach, rejection
of all proposals and reissuance of a revised RFP, subdivision of the project,
or encouragement of team formation among one or more Bidders.  If these
perogatives are to be considered, however, the necessary flexibility to do
so should be reserved by so stating in the RFP.

All the above types of negotiations may involve sensitive material and
conflicting interests, and must therefore be handled with great care.  If
the RFP allows for alternate proposals or exceptions, the related statement
of the RFP will govern the evaluation and Contractor selection procedures.
In case of situations not foreseen in the RFP, the Purchaser must be governed
by applicable laws and regulations supplemented by his own judgment and
established practices.

                                     144

-------
 Receipt, Acknowledgement And Control Of Proposals

 Bidders naturally have a  high  degree of interest  in  the  consideration  of
 their proposals.   To  insure fairness to all parties  and  to  insure  against
 jeopardizing  the  evaluation process, procedures for  the  receipt, acknowledge-
 ment  and control  of proposals  should be established  prior to proposal  sub-
 mittals.

 The deadline  for  the  submittal (or receipt) of proposals should be ob-
 served, in accordance with  the procedure specified in the RFP.  All proposals
 received should be acknowledged by the Purchaser  (whether hand carried or sent
 through the mail).  Use of  a standard letter format  to acknowledge proposal
 receipt prior to  the deadline  is adequate.  It may be desirable in many cases
 to request the Proposers to clearly identify the package (proposal ID number,
 or other form of  identification) to assure timely delivery  to the proper
 individual within the Purchaser's organization.

It is generally a good practice to  specify a  proposal opening date and
 time in the RFP and to refrain from opening proposals prior to the time speci-
 fied.  This insures against accidental compromise of sensitive proposal infor-
mation during the submittal period but also requires a clear identification of
 the proposal package by the Proposer.  If this practice is to be followed,  the
 Proposers should be advised as to what information should appear on the wrapping
of the proposal package to assure its immediate recognition and proper  handling.

If technical proposals and cost proposals  are to be handled  separately
by the Purchaser, the RFP should provide for their separate submittal by the
Proposers.   In this case the technical  evaluation team should preferably not
be informed about costs until their evaluation is complete.   One way of
assuring the separation of technical and cost proposals is to have them ad-
dressed to two different individuals or departments.   If technical and cost
proposals are to be sent in a single package, placing the cost proposal in
a separate sealed envelope can  be specified.
                                    145

-------
When the evaluation is completed and the ^referred Proposer identified, un-
successful Proposers should be notified.  This notification should not precede
determination that a satisfactory contract can be developed with the preferred
bidder nor be later than the date of the contract award.  At the time of
notification, proposals may also be returned to the unsuccessful Proposers if
they and the Purchaser so desire.  Any communication with an unsuccessful
Proposer about reasons for rejection of his proposal should be confined to
the discussion of the strong and weak points of his proposal.  No other
proposals, or ranking of proposals, should be discussed.  Furthermore, such
post-selection discussions should be conducted only at the Proposer's request.

6.6  EVALUATION OF PROPOSALS

Preparation for the evaluation of proposals begins at the time the Pur-
chaser organizes the contents and format of the RFP.  If well-designed,
the RFP specifies the organization of the proposal, its contents, and the
specific areas the Purchaser wants emphasized in the Proposer's response.  The
evaluation of the proposals then consists mainly of comparing each response
with the requirements specified in the RFP.  The purpose of the evaluation
is to identify the most favorable offer of services or, in general, the
evaluation must include an analysis of the technical proposal, the cost
proposed and the Proposer qualifications.

Technical Evaluation

The extent and difficulty of the technical evaluation depends very greatly
on the subject matter of the proposals.   In general, proposals relating
to substantial amounts of work in water quality modeling and water quality
planning can be expected to be complex.   A careful procedure is required
to analyze each of the technical elements of each proposal.

Technical Elements.   The more detailed  and specific the RFP  in  spelling  out
the proposal requirements,  the easier becomes  the  task  of  identifying  the
individual  technical  elements to be evaluated.   The section  titled  "Preparation
of Proposal Request"  pointed out that the major  technical  evaluation criteria
should be outlined  in the RFP.   The Proposers's  response to  the  evaluation  criteria

                                     146

-------
should provide information on the overall project plan and the proposed
technical approach.  In addition, it should include a contractually
binding Statement of Work.

The overall project plan should provide information regarding the Con-
tractor's understanding of the desired project scope, organization of the
project into logical tasks, and staffing of each task.  Deliverable items
and detailed schedules of performance should be clearly spelled out.   The
overall project plan should also describe any probability of not meeting
proposed deadlines and the contingency plans for conduct of the project in
that event.  The procedures for project management and conformance with the
monitoring and other procedures specified in the RFP should be unequivocal.

The elements of the overall project plan give a good indication of the
Contractor's understanding of Purchaser's needs and intents.  Then also indicate
the importance attached by the Contractor to this particular project, his ability
to organize and manage projects of this type, and the probability of success
in meeting the Purchaser's objectives within the allocated time and budget.

The technical approach should closely reflect understanding  the tech-
nical requirements of the RFP.  The discussion of approach should discuss
the models and techniques to be used, data requirements, procedures for data
collection, provisions for any field work, expected results  and quality of
data to be generated for the deliverables.  The proposed content and expected
quality of each deliverable end item should be explicitly described.

The proposed technical approach provides insight into the detailed
plans of the Contractor for meeting or exceeding the requirements of the RFP.
These are of great importance in assessing the Contractor's  technical capabilities,
expertise and proposed methods for performing the individual tasks of the
project.  Specific modeling techniques, data requirements, data acquisition
and management approaches, and the use of available  technologies  should be
explained by the Contractor in sufficient detail to enable the Purchaser to
compare and evaluate alternative approaches.  The expected  findings  and
results and their organization into reports and other deliverables should  be
discussed in detail.  If any proprietary elements are involved,  they should

                                    147

-------
 be clearly identified so that their relative  merits  can be  properly con-
 sidered  in the evaluation.   In summary,  the Contractor's proposed approach
 explains how he is planning  to achieve the project objectives, and the Pur-
 chaser's appraisal should  reflect  his  level of  confidence in these plans
 as well  as his assessment  of  the extent  to which  these plans will satisfy
 his requirements.

 Although the overall  project  plan  and  the technical  approach are impor-
 tant for assessing the expected performance of  the Contractor, the con-
 tractually binding Work  Statement  in the proposal is vital.  The Work
 Statement should be characterized  by its completeness, clarity and con-
 formance with requirements.

 As opposed to the  Method of Approach,  the Statement of Work simply states
 what the Contractor will complete  and  deliver in fulfillment of the con-
 tract.   While the  explanation of the approach is an important part of the
 proposal for  purposes of evaluation, only the Statement of Work generally
 becomes  a  part of  the contract binding on all parties.  A careful evaluation
 of the proposed Statement of Work  is necessary  to ascertain that it meets
 the  Purchaser's needs and fulfills  the intent and requirements of the RFP.
 The  Statement of Work should be concise and factual.  It should include
 all  the  elements the Purchaser wants to be performed but without any
unnecessary reference to the how's and the why's that tend to obscure the
Contractor's or Purchaser's responsibilities and obligations.  A proposed
Statement of Work should not be vague, or else  the Purchaser cannot enforce
 the  intent of the contract.  In some situations, the Purchaser may want to
have a "flexible" Statement of Work for his own protection.   This, however,
can lead more often to misunderstandings or misinterpretations later
during the project, with the possibility of legal problems.   In summary,
while the  technical approach of a proposal is intended to convince the
Purchaser about the advantages  of selecting that particular Contractor, the
purpose of the Statement of Work is to clearly define the Contractor's
obligations  in performing the contracted services.
                                     148

-------
Procedures for Technical Evaluation.    The overall criteria and the
specific technical elements of evaluation have been established and documented,
and a convenient and fair evaluation procedure must be implemented.  The use
of an appropriate evaluation form can be helpful to the members of an
evaluation committee in uniformly appraising proposals.  Weights reflecting
the importance of the several features evaluated can then be assigned and
a general ranking accomplished.

The importance of each technical element of the RFP to which the proposal
responds can be expressed by a weight.  The sum of each evaluator's grade of
quality for each technical element times the assigned weighting points
gives an assessment of the proposal's overall quality and can be used for an
initial ranking of overall technical quality.

This or other similar grading procedures must be handled with care.
If there are major differences between the individual evaluators' ranking
of the proposals, these differences should be discussed and, if possible,
reconciled through detailed study of the reasons for any discrepancies.
Any ranking order established should be used as a guideline only for the
elimination of the weak proposals.  The more promising proposals which are
retained for detailed evaluation should be reexamined with specific emphasis
on their differences in each important technical aspect.  The grading
technique and weighting points can be refined for the  final analysis to
emphasize important technical elements or  to consider  special attributes
of proposals not considered in the original evaluation.

After the evaluation is complete, it is desirable that the members of
the committee be in agreement with respect to the ranking.  If  two or more
proposals are very close  in quality and satisfactory  from a technical point
of view, other considerations must govern  the final ranking of  Proposers.

Use of Consultant in Evaluation.   The evaluation of  proposals  for provision
of modeling services to assist in conducting water  quality  planning  creates
a situation not unlike  that of RFP preparation.  That  is,  the  skills necessary
                                    149

-------
 to conduct the evaluation are very nearly those proposed to be obtained.
 If the Purchaser has these skills in-house and the purpose in obtaining
 contractual services is to supplement the available labor force,  then
 no problem may exist.   However, if the purpose in issuing the RFP is to
 secure services of a type not available on a staff basis then it  is
 unlikely that existing staff members can perform a fully competent evalu-
 ation of proposals.

 In the latter case described above,  it may be cost effective to
 have a consultant assist the Purchaser in the technical evaluation.   In
 this event,  it is desirable to also  involve the same consultant in the
 preparation of the RFP.   Sufficient  expertise in RFP preparation  and
 proposal evaluation phases can aid in mutual understanding between Pur-
 chaser and Contractor  and help insure against work which may be unneces-
 sarily expensive due to  misinterpretation of project requirements.

 Documentation of Technical Evaluation.     After completion of the
 technical evaluation,  it is  important that the evaluation findings  be
 documented.   This  should include  technical data and narrative which
 clearly  state the  conclusions  reached about the proposals.   This  infor-
 mation is  important  both for subsequent negotiations  and as  a basis  for
 responding to Contractor inquiries.

The documentation of the  technical evaluation  should adequately
describe  the  recommendations of the evaluation team and  explain the  reas-
 ons why  any Proposers were thought unable  or  unsuitable  to perform  the work
The information prepared  should include the  specific  reasons  for which a
 Proposer  was  rated low or high, strong points  and/or  weakness  in  the  pro-
posal, and items which require discussion or clarification during
negotiation.  The latter may include definitions of important  terms,
work proposed which is in excess of that required by  the RFP and adequ-
acy of the proposed labor for project  completion.
                                   150

-------
Co8t Evaluation

As with the case of the technical evaluation, a well defined RFP
can simplify the task of cost evaluation.   The extent of the analysis
performed will depend on the nature of the contractural arrangement to
be used as well as the cost and complexity of the overall project.

Cost Elements.     The cost information required by the RFP must be
determined in anticipation of the type of contract to be let.  The least
amount of information is required for firm fixed-price proposals, while
detailed financial data concerning the Contractor may be necessary for
cost reimbursement type arrangements.

To simplify the presentation of cost information in a uniform and
easily comparable manner, it is often useful to provide or stipulate
forms for cost summaries.  Such forms may call for showing of labor
costs including the basic rates, estimated hours and extended totals as
well as direct and indirect costs, payroll burndens, and fee.  Some
flexibility must be allowed Contractors to display the elements of their
proposal however as the accounting systems in use vary considerably.

Some information affecting cost  is not suitable for  tabular  display
and provision should be made for contractors to furnish certain infor-
mation in narrative form.  Such information might include policies for
travel and subsistance costs, material aquisition procedures and compo-
sition of overhead and other indirect costs.

Procedures for Cost Evaluation.      Thete  are multiple  objectives  in
undertaking the evaluation of proposed costs.  Completion of the evalu-
ation normally involves careful examination  of each proposal for arith-
metic correctness, analysis of rates and subtotals of  cost  for  separate
major tasks,  as well as distribution of cost among  the  various  elements.

Cost should not be the sole criterion of contractor  selection  and  ethical
conflicts can arise if such is the case.   However,  even where  cost is one
                                   151

-------
 of the major deciding factors,  far more than a simple comparison of  total
 costs offered by technically suitable contractors  is  warrented.   The purpose
 of the review of the cost proposal is insuring that the quoted  price  is
 realistic and does not depart from the estimated price so far as to  in-
 dicate misunderstanding or overoptimism regarding  accomplishment of  the
 required scope of work, and to  insure that the price  is consistent with
 available funds.

 A portion of the cost analysis  should include evaluation of  the
 level of effort proposed.   If anticipated labor hours have not  been  in-
 cluded in the technical proposal,  this information should be extracted
 from the cost proposal and furnished to those performing the technical
 evaluation so that its reasonableness can be determined.   Acceptance of
 a proposal bearing an unrealistic  price or labor input practically
 assures later difficulty.

 Firm Qualifications

 Even though an  extensive qualification  procedure may  have  been used
 to identify and assure providing the RFP  only to carefully selected con-
 tractors,  further evaluation is usually undertaken  at the  time proposals
 are  evaluated.   The  Contractor's response  to  the requirements of  the RFP and
 to the  stated criteria for selection should  include information  both on
 qualifications  and on the  Contractor's  commitment to  the project.

The  qualifications presented should  include a description  of the
 staff,  facilities  and  general background applicable to  the project.
 Particularly  relevent  experience should be detailed as well  as the capa-
bilities of any  expert consultants or specialized subcontractors.

The Contractor's commitment to the  project is evidenced through  the
availability  of key staff members,  requirements of  current contracts, and
other such  information.  The proposal should make it  convincingly clear
 that  the Contractor not only has the capability to  carry out the proposed
project but will, in fact, place an adequate priority upon doing so.
                                  152

-------
The portion of the proposal review directed toward assessment of the
qualifications of the Contractor's firm or organization should emphasize both
resources and capabilities.  Resources include facilities and equipment
which may be required as well as adequate support services for staff
which would be involved in leading and performing the proposed project.

The personnel proposed for conducting a study should either be already
available to the Contractor or suitable arrangements should exist
for acquiring them.  In addition to simply having sufficient numbers of
appropriate staff employed or accessible, their availabiltiy should be
reviewed.  Specific staff requirements depend on the nature of the pro-
ject, but generally both management and technical skills are required.

 Information on experience  in  conducting similar projects  should be
provided for both the organization and the staff.  Close review of the
organization's work, and perhaps investigation of the success with which
the work was accomplished, can be informational.  However, as in the case
of the reference check discussed previously, care should be taken to in-
sure full understanding of circumstances.  This is particularly necessary
if the investigation turns up adverse information.

6.7  CONTRACTOR SELECTION

 If the evaluation of the proposals has not been conclusive,  a face-
 to-face meeting with one or more  Contractors may be necessary to enable  the
Purchaser to ask specific  questions about the Contractor's technical approach,
obtain any necessary clarifications,  and  generally size  up the proposed
manager and key members of the project  team.  There is little to be  gain-
ed in conducting such  interviews with all Contractors  who submitted  proposals
unless their several proposals are quite  comparable in technical quality.
 If oral interviews are desired by  the Purchaser,  they  should  be arranged
 only with the few highest-ranking  Proposers from whom the Contractor will
be selected.
                                   153

-------
 The Purchaser may elect to conduct an interview solely with the highest-
 ranking Proposer if evaluation of the proposal received results in
 a clear preference pending the resolution of certain specific points.
 An interview with the next-ranked Bidder is then necessary only if no
 satisfactory arrangement can be reached with the first-ranked Bidder.

 If oral interviews are to be conducted with several top-ranked Bidders,
 notification to each of the other invitees is neither necessary nor
 desirable.   Confidentiality on this  point is difficult to  maintain.  How-
 ever,  if achieved,  confidentiality will encourage the Bidders to focus
 their  statements and answers on the  project rather than on their compe-
 tition.

 There are no fixed rules for conduct of an oral interview.  In general,
 the objective of the questions should be to clarify certain aspects
 of either the technical  or cost proposal,  and pursue the proposed  tech-
 nical  approach in such detail  as may be of interest.   Specific  questions
 may address  the contractual arrangements,  project  management, project
 costs,  terms  and conditions, guarantees,  or any  other point  of  legitimate
 interest to  the Purchaser.   It  is especially important in  such  a meeting
 to clearly establish a measure  of Contractor performance if  it  has not
 been adequately defined  in the  proposal.   This raises  the  question of what
 constitutes satisfactory performance, its  method of  measurement, and the
 means by which  the Purchaser can enforce  requests  for  the  Contractor to
 correct  any incidents of non-satisfactory  performance.

 Minutes of  the meeting  should be taken  for the record to aid in
 evaluation and  for use in  the subsequent contract negotiation.   The Contractor
 should be advised that statements or commitments made  during  the oral
 interview will be considered as  contractually binding.   The minutes should
be approved in writing by  the Contractor.

 After  the interview,  the Contractor's specific response  and general
attitude must be evaluated by the Purchaser.  Such an evaluation should
take place soon after the oral interview, and desirably should involve all
 Purchaser representatives who participated  in  the interview.
                                   154

-------
Finally, based on the evaluation of proposals, interviews, and other
efforts at selection which may be made,  a preferred Contractor must be chosen.
Since the subsequent negotiations leading to contract execution may be
unsuccessful, one or more alternate Contractors  should be  identified.   In  the
case of a firm fixed-fee contract, the selection often entails only
identification of the several Contractors who adequately meet the require-
ments, and comparison of the bid price.

6.8  CONTRACTING

The  sequence of activities  in  contracting includes negotiation,  con-
tract preparation and contract award.  Each of these steps are important
and  must be  given close attention.

Negotiation

Netotiation, assuming all factual matters and questions are  settled,
consists of  two distinct  parts.   The  first  part  includes  reaching  agree-
ment on any  differences which  exist between the  parties which can  be  sat-
isfactorily  compromised.   The  second  part,  commonly  referred to  as bar-
gaining,  involves the tradeoff of any remaining  points of difference.

As in the case of managing the solicitation of  proposals, one in-
dividual  should be  identified  as the  negotiator  for  the Purchaser  and be
responsible  for  this phase of  the project.   Similarly, confusion can  be
avoided  if each  contractor or contractor team identifies  a single  repres-
entative.

 Where negotations  are  conducted with more than  one Contractor, formal
relations  and confidentially must be  maintained.  In such multiple nego-
 tiations,  the Purchaser should keep the arrangements strictly separate.
Under no conditions should an "auction" of  services be permitted or en-
 couraged by  revealing  competing bids  or conditions.   The Purchaser should
 always attempt to negotiate the contract form most advantageous for the
 performance  of the  work at hand.  The Purchaser must keep in mind  that
                                    155

-------
that the Contractor must also be able to operate efficiently within the
framework of that arrangement for the duration of the project.   This re-
quires that the Purchaser make a realistic assessment of the funds re-
quired for the project.  Any major discrepancy in the views of  Purchaser
and Contractor as to budgets, schedules or other major points should be ex-
amined to determine if it indicates a lack of mutual understanding with
regard to the extent, scope or detail of work expected.  Resolution of
any such misunderstandings should be achieved during negotiation.  Ul-
timately, the "best" contract in each situation is the one that will help
achieve the project objectives in the most cost-effective and timely
manner.

Contract Preparation

Preparation of the contract should be begun prior to negotiation and
completed after agreement is reached.  Normally, the contract consists of
several parts, including those dealing with:

     1.   Purpose of the project;
     2.   Scope of work;
     3.   Contract period;
     A.   Reporting requirements;
     5.   Cost collection, billing, and payment; and
     6.   General provisions.

Items  (5) and  (6) above  tend to be standard statements specified for
all contracting by an agency and frequently are readily available for in-
clusion in the contract as an attachment or appendix.  If unavailable,
examples can be readily obtained from the contracting officers  of various
state or federal agencies.

Purpose and Scope.     The purpose of the project and its scope should be
relatively well defined based on the preparation of the RFP.  Commonly,
the contractual Scope of Work both reiterates the requirements  specified
in the RFP and incorporates the successful proposal by reference.
                                  156

-------
The section of the contract which deals with the time of performance
should include the dates for beginning and completing performance and
spelling out the conditions under which amendments to the schedule will
be requested or granted, as well as the procedures for such changes.  Re-
view periods for deliverables should be specified and the impact of de-
lays defined.

However described, the contract should finally achieve the Purchaser's
requirements.  To avoid later difficulty, the Purchaser should recognize
that the Contractor is obligated to perform only the minimum work which is
specified and specify the work accordingly.  It is important that respon-
sibility for each major portion of  the work be clearly assigned  if  a joint
effort of the Purchaser and Contractor are required.  The sequence  in
which activities are to be undertaken should be stated either in schedules
or  charts.

Procedures  for  approving contract  deliverables  or units  of work
should be clearly  defined.  To  improve  clarity,  tiems  or aspects which
are mutually  agreed not to  be included  in the scope  of work  should  also
be  described  where their omission  would cause doubt.

 Contract Period.      This  section of the contract should define the
 effective date of the contract, the time when work is to begin, and
 the length of the time available or a completion date.  Contract
 periods are normally specified in calendar time without regard to
 "work days".   If interim dates are important to the Purchaser, accomp-
 lishment of certain tasks  or portions of work may be specified to be
 completed by particular dates.   This is likely to be the case when
 modeling or model results  must be integrated with the results of
 economic, engineering, financial, environmental and other studies.

 Any conditions which affect the initiation or effectiveness of  the
 contract should be enumerated.   These conditions may include approval of
 funds,  the contractor or the contract by other agencies or political
 bodies.
                                    157

-------
  Care should be taken to  avoid any  confusion  in  the project schedules
  that may  be brought  about by specifying a schedule in the contract and
  simultaneously incorporating differing schedules  (such as proposal sched-
  ules) by  reference.  Resolution should be provided by identifying the
  prededence  in  authority of statements contained in the contract and in
  other documents which are referenced.  The contract should also provide
  for  the eventuality of contract termination prior to its completion or
  the  finish  date.  Termination may be warranted under several conditions,
  and  the contract should adequately protect the interests of both parties.
 Provisions  for termination as well as contractual disputes should be
 determined  in cooperation with legal counsel.

 Reporting Requirements.      A number of types of reports may be required,
 including periodic progress  reports, technical descriptions of work
 elements,  various documentation  reports and  overall  project reports.   In
 addition,  one objective  of the work undertaken may be  to produce some
 specialized  report.   Each of the  reports which are to  be prepared during
 the course of the contracted effort should be carefully  defined and
 scheduled.  The description  of each report should  state  its purpose, con-
 tent  and  form.   If an opportunity for review  and approval is  desired,  the
 procedure  for review  should  be stated.  For major  reports,  the develop-
 ment  and  submission of an  outline should be required in  advance of actual
 report preparation.   The  content  of  progress  reports should be sufficient
 to  present a clear picture of both  the  technical accomplishments  of the
 project and  to  enable comparison of  progress with  the  time and funds
 committed  and remaining.  Reports should be scheduled with care.  Adequate
 time  must  be allowed not only for the Purchaser and Contractor  staffs  to
prepare, review, revise and publish  reports, but also for the  partici-
pation of  the public and official bodies.  In addition, report  scheduling
must  consider their use as essential information for other planning activ-
ities.  In general, the Purchaser should not anticipate that the Contractor
will provide any reports  or other written material except that called for
by the contract.  If either the Purchaser,  a grantor  agency, or some
other related agency has  specific requirements for the  content or form of
reports,  they should  be incorporated in the contract, either in full or
by reference.
                                   15 H

-------
Cost Collection. Billing and Payment.      The normal procedure for all
but relatively small value contracts provides for periodic payments
based on progress.  For such purposes, progress can be measured by con-
tractor input in labor and other costs on a time basis or by completion
of portions of work.  In either case, payments are usually limited so
that a sum remains unpaid until final project completion to assure con-
tractor performance.

If multiple sources of  funds are to be used to meet Contractor Billings,
then the billling  information required by each as a basis  for pay-
ment should be  identified.   The contract  should  give  explicit  instruc-
tions  to  the  Contractor on  the  content and  form  of  information to be sub-
mitted on invoices.

 General Provisions.      This section of  the contract covers routine ad-
 ministration matters.  Included are such things  as the use of subcon-
 tractors, prior approval of certain items,  patents and other similar
 items.

 The use of subcontractors in any significant amount should be fore-
 seen by the  contractor at the time of proposal preparation.  Therefore,
 some  limitation is warranted to avoid the later diffusion of responsibility
 for accomplishment of  the work.  However, conditions can  arise unexpectedly
 in which use of a subcontractor having specialized capabilities  can be of
 substantial  value.   Permission to  use contractors  can reasonably be made
 subject  to Purchaser approval  based  on demonstrated  benefit,  qualifications,
 cost  and  other factors.

 Prior approval of expenditures is  sometimes  retained by Purchasers
  for such items as equipment purchases,  travel and other items to be
  charged against the project.   While possible and reasonable, a situation
  could easily develop in which the Purchaser is overburdened dealing with
  approval requests, and the Contractor is unduly restricted.  Only  suffic-
  ient control is necessary to prevent impractical or clearly inefficient
  project expenditures is warranted.
                                   159

-------
  Contract Award

  When negotiation is complete and a contract drawn which fully des-
  cribes the understanding between parties,  the  stage  is  set  for  the execu-
  tion of the contract.

  The award is normally accomplished by forwarding one or more  copies
  of  the prepared  contract to  the  Contractor  for signature and  return.  The
  documents  are  then  signed by  the necessary  official(s)  for  the Purchaser
  and  a  copy returned to  the Contractor.

  If  not done earlier,  unsuccessful bidders should  be  notified  at this
  time of  the contract award.

  6.9  PROJECT ADMINISTRATION

  After  the  execution of  the contract and initiation of work, the
 Purchaser must assume a dual role.  The contract must be administered;
 and, simultaneously, the Purchaser and Contractor must develop and main-
 tain close cooperative working relations.

 Liaison With Contractor

 Contract between the Purchaser and Contractor should  be frequent
 enough to enable  close coordination of respective work efforts and the
 earliest possible identification  of problems.   Liaison can be  carried  out
 in a  number of  ways,  including  telephone, correspondence, and  personal
 conferences.  In  all but the smallest  projects, questions of procedure
 inevitably  arise  due to  unforeseen problems  or points  not covered in the
 written Work Statement.   To avoid  later problems,  resolution of such ques-
 tions should be documented in writing.

 In carrying out the  liaison portion of contract administration,  the
Purchaser should strictly observe  the Contractor's freedom to manage his
work as he sees fit within the limits of the contract.  Unless serious
                                   160

-------
problems are clearly evident,  and appear sure to affect the quantity  and
work, the Purchaser should be concerned only with results.

Evaluation of Progress

Measuring the accomplishment of the Contractor is important both to
assure project objectives are being met and as a basis of payment.  Pro-
gress is a combination of both the volume and quality of work produced
under the contract.

The volume of work can be established  on  the basis of periodic pro-
gress reports, preparation  of  reports  on  specified technical topics  and
liaison activities.   Initiation  of projects  or major  tasks of projects  is
frequently accompanied by a variable  amount  of work for which no  clear
product is produced.  While this time devoted  to "start up" is  necessary,
it should be kept within reasonable bounds.  The Work Statement should
provide for  the  sequential  completion of  tasks  to avoid the difficulties
which can occur  if the  Contractor elects  to "start up" numerous tasks
without pressing for completion of any.

 The portion of progress evaluation devoted to measuring the quality
 of the work which is accomplished is  more difficult.   If the Purchaser
 originally secured the Contractor's services as a supplement to his work
 force and has the necessary skills available,  these staff can probably
 provide an adequate evaluation.  If,  however,  the Contractor's services
 were obtained to  secure expertise not otherwise available to the Purchaser,
 he may be in a poor position to judge the technical acceptability of the
 Contractor's work.  In  this case, assistance can be sought from various
 sources, such as  federal and state agencies, but the Contractor's repu-
 tation  and demonstrated competence becomes of prime  importance.

 Receipt of  Deliverables

  Items  specified as contract deliverables should be accounted for and
 handled in  very specific fashion.  Receipt  of  such deliverables  by  the

                                  161

-------
Purchaser should be formally acknowledged to the Contractor.  Review
comments and approvals of acceptable deliverables should likewise be
documented to firmly establish the status of each task and the project.

Assessment of Performance

At the conclusion of the Contractor's effort, an assessment of per-
formance must be made as a basis of final payment.   This performance
assessment should emphasize comparison of the Contractor's accomplishment
against the contract requirements.  The Purchaser must refrain from allow-
ing the assessment to be affected by any outside considerations.   The
specific elements of the evaluation should be directed toward the account-
able items in the Work Statement, including the quality and timeliness of
their accomplishment.

-------
                        ACKNOWLEDGEMENTS
     The authors wish to acknowledge the contributions of a number of
individuals to the development of this handbook.   Particular recog-
nition must go the the Project Officer, Mr.  Donald H.  Lewis, and Dr.
Roger D. Shull both of the Systems Analysis  Branch, EPA Office of
Research and Monitoring, for their guidance  and support throughout the
project.

     Mr. William P. Somers and Mr. John Kingscott, of the EPA Planning
Assistance Branch, gave the project initial  direction based on their
experience with existing water quality models.

     Mr. David Legg of the Army Corps, of Engineers, North Pacific Divis-
ion Office, Mr. Hayden Street of the Snohomish County Planning Department,
Washington, and Dr. Bob Milhous of the Washington  State Department of
Ecology, all generously provided critical review and helpful suggestions
for the model evaluation and cost effectiveness rating systems.

     We are also grateful to our Project Manager,  Dr. Robert B.  Schainker,
for his encouragement, guidance and review of the  draft material  for  the
successful completion of this project.
                                  163

-------
164

-------
                                 REFERENCES
1.   Texas Water Development Board.   DOSAG-I,  Simulation of Water Quality
     in Streams and Canals:   Program Documentation and User's  Manual.   NTIS,
     Springfield, Virginia (PB 202 974),  September 1970.

2.   Systems Control, Inc.,  and Snohomish County Planning Department.   Water
     Quality Management Plan for the Snohomish River Basin and the Stilla-
     guamish River Basin, Volume IV - Computer Program Documentation,  Part A:
     Steady-state Stream Model (SNOSCI),  and Part B:  Dynamic Estuary  Model
     (SRMSCI).  Snohomish County Planning Dept., Everett, Washington,  1974.

3.   Systems Control, Inc.,  and Snohomish County Planning Department.   Water
     Quality Management Plan for the Snohomish River Basin and the Stilla-
     guamish River Basin:  Volume III - Methodology (Chapter 4).  Snohomish
     County Planning Department, Everett, Washington, 1974.

4.   Hydroscience, Inc.  Simplified Mathematical Modeling of Water Quality.
     Water Programs, Environmental Protection Agency, Washington, D.C.
     (U.S. Government Printing Office:  1971-444-367/392), March 1971.

5.   Hydroscience, Inc.  Addendum to Simplified Mathematical Modeling of
     Water Quality.  Water Programs, Environmental Protection Agency, Wash-
     ington, D.C.  (U.S.  Government Printing Office:   1972-484-486/291), May
     1972.

6.   Chapra,  S.  C. and  S. Gordimer.  A Steady-state,  One Dimensional,
     Estuarine Water Quality Model  (Documentation  for ES001).  Data Systems
     Branch,  EPA Region II, New York,  September  1973.

7.   Chapra,  S.  C. and  S. Gordimer.  Addendum to ES001:  Verification  of
     Model  for New York Harbor.  Data  Systems Branch, EPA Region II, New York,
     September 1973.

8.   Hydroscience,  Inc.   Mathematical  Models  for Water  Quality  for the Hud-
     son-Champlain and  Metropolitan Coastal Water  Pollution Control Project,
     FWPCA.   Hydroscience,  Inc., Westwood, N.  J.,  April 1968.

9.   Masch,  Frank D.  and Associates, and  the  Texas Water Development  Board.
     Simulation  of Water Quality  in Streams and Canals: Theory and Descrip-
     tion of the QUAL-I Mathematical Modeling System.  Texas  Water Development
     Board,  Austin,  Texas (Report  128), May 1971.

10.  Texas Water Development Board.  QUAL-I,  Simulation of Water Quality  in
     Streams and Canals: Program Documentation and User's Manual.  Texas
     Water Development  Board,  Austin,  Texas,  September 1970.

11.  Roesner, L. A., J. R.  Monser,  and D. E.  Evenson.  Computer Program Docu-
     mentation for the Stream Water Quality Model, QUAL-II.   Water Resources
     Engineers,  Inc.,  Walnut Creek, California, May 1973.
                                        165

-------
 12.  Water Resources Engineers, Inc.  A Water Quality Model of the Sacra-
      mento-San Joaquin Delta.  Report to the U.S. Public Health Service,
      Region IX, June 1965.

 13.  Water Resources Engineers, Inc.  A Hydraulic Water Quality Model of
      Suisun and San Pablo Bays.  Report to the Federal Water Pollution Con-
      trol Administration, Southwest Region, March 1966.

 14.  State of California Water Resources Control Board.  Final Report
      (Abridged Preliminary Edition):  San Francisco Bay-Delta Water Quality
      Control Program.  March 1969.

 15.  Feigner, K. D., and H.  S. Harris.  Documentation Report, FWQA Dynamic
      Estuary Model.  U.S. Department of the Interior. FWQA, 248 pp, July
      J-.7 / U •

 16.  Callaway, R.  J., K. V.  Byram and G. R. Ditsworth.   Mathematical Model
      of the Columbia River from the Pacific Ocean to Bonneville Dam, Part I:
      Theory,  Program Notes and Programs.  FWQA Pacific  Northwest Water
      Laboratory, Corvallis,  Oregon, November 1969.

 17.  Callaway, R.  J. and K.  V.  Byram.   Mathematical Model of the Columbia
      River from the Pacific  Ocean  to Bonneville Dam, Part II:  Input-Output
      and Initial Verification Procedures.   EPA,  Pacific Northwest Water
      Laboratory, Corvallis,  Oregon, April  1971.

 18.  Metcalf  & Eddy, Inc., University  of Florida, and Water Resources Engin-
      eers,  Inc.  Storm Water Management  Model,  Volumes  1-4.   EPA Report
      No.  11024DOC,  (U.S.  Government Printing Office, Stock Nos.  5501-0109
      0108,  0107, 0105),  July-October 1971.

 19.  Pritchard,  D.  W.  Estuarine Circulation Patterns.   Proceedings of Amer-
      ican  Society  of Civil Engineers,  Vol.  81,  No.  717,  1955.

 20.   Pritchard,  D.  W.  Estuarine Hydrography.  Advances  in Geophysics,  New
      York, Academic Press, 1956.

 21.   Tracer, Inc.   Estuarine Modeling  - An Assessment.   EPA Water  Quality
      Office, 16070DZV02/71 (U.S. Government Printing Office,  Stock No.  5501-
      0129), February  1971.

 22.   Water Resources Engineers, Inc.   Temperature Prediction  in Dworshak
      Reservoir by Computer Simulation, Computer Application  Supplement
      Water Resources Engineers, Inc., Walnut Creek,  California, September


23.  Water Resources Engineers, Inc.  Mathematical Models for the  Prediction
      of Thermal Energy Changes in Impoundments.  EPA Water Quality Office,
     16130EXT12/69,  (U.S. Government Printing Office:  1969-359-339)
     December 1969-

24.  Water Resources Engineers, Inc.  Mathematical Models for the Prediction
     of Thermal Energy Changes in Impoundments, Computer Application
     Supplement.  Water  Resources Engineers, Inc., Walnut Creek, California,
     Undated (after March 1969).

                                       166

-------
25.   Water Resources Engineers,  Inc.   Prediction of Thermal Energy in Streams
     and Reservoirs.  A Report to the California Department of Fish and Game,
     30 June 1967,  revised 30 August  1968.

26.   Shepherd, J. L. and E. J. Finnemore.   Spokane River Basin Model Project:
     Vol. VI - User's Manual for Stratified Reservoir Model.  Air and Water
     Programs Division, EPA Region X, Seattle, Washington,  October 1974.

27.   Finnemore, E.  J. and J. L.  Shepherd.   Spokane River Basin Model Project:
     Volume I - Final Report.  Air and Water Programs Division, EPA Region X,
     Seattle, Washington, October 1974.

28.   Baumgartner, D. J., D. S. Trent  and K. V. Byram.  User's Guide and
     Documentation for Outfall Plume Model.  EPA Region X,  Pacific Northwest
     Water Laboratory, Corvallis, Oregon (Working Paper #80), May 1971.

29.   Baumgartner, D. J. and D. S. Trent.  Ocean Outfall Design:  Part I,
     Literature Review and Theoretical Development.  FWPCA, April 1970.

30.   Systems Control, Inc.  Use of Mathematical Models for Water Quality Plan-
     ning.  State of Washington, Department of Ecology, WRIS Technical Bulletin
     No.  3, June 1974.

31.  Chen, C. W. and Orlob, G. T.  Ecologic Simulation for  Aquatic  Environ-
     ments, Final Report.  Water Resources Engineers, Inc., Walnut  Creek,
     California, December  1972.

32.  Chen, C. W. and Orlob, G. T.  Ecologic Simulation  for  Aquatic  Environ-
     ments, First Annual Report.  Water Resources  Engineers,  Inc.,  Walnut
     Creek, California, August 1971.

33.  Leendertse, Jan J.  Aspects of  a Computational  Model  for  Long-Period
     Water-Wave  Propagation.  Rand Corporation  Memorandum  RM-5294-PR,  May
     1967.

34.  Masch,  Frank D.   A Numerical Model for  the Simulation of  Tidal Hydro-
     dynamics in Shallow Irregular Estuaries.   Technical Report  HYD 12-6901,
     Hydraulic Engineering Laboratory, Department  of Civil Engineering,  The
     University  of  Texas  at Austin,  1969.

35.  Hydrocomp  International, Inc.   Hydrocomp Simulation Programming Opera-
      tions Manual  (Second Edition).   Hydrocomp International,  Inc., Palo
     Alto, California,  July 15,  1969.

36.  Lombardo,  P.  S.  and D.  D.  Franz.  Mathematical Model  of Water Quality
      Indices in  Rivers and Impoundments.   Hydrocomp Incorporated, Palo Alto,
      California, December 8,  1972.

 37.   Roesner, L. A., H. M. Nichandros, R.  P.  Shubinski, A. D. Feldman,
      J. W. Abbott,  and A.  0.  Friedland.  A Model for Evaluating Runoff
      Quality in Metropolitan Master  Planning.  ASCE Urban Water Resources
      Research Program Technical Memorandum No. 23, American Society of Civil
      Engineers,  New York, April 1974  (reprinted by the Corps of Engineers
      Hydrologic Engineering Center,  Davis, California, August 1974).
                                       167

-------
 38.   U.S.  Army Corps  of  Engineers, Hydrologic  Engineering Center.  Urban
      Storm Water  Runoff:   STORM.  Corps  of Engineers, Hydrologic Engineering
      Center Generalized  Computer Program 723-S8-L2520, Davis, California,
      January 1975 (draft User's Manual).

 39.   U.S.  Army Corps  of  Engineers, North Pacific Division.   Streamflow Syn-
      thesis & Reservoir  Regulation:  Program Description & User Manual for
      SSARR Model.  Program 724-K5-G0010,  Corps of Engineers, North Pacific
      Division,  Portland,  Oregon, September 1972  (revised December 1972).

 40.   Burnash,  R.  J. C.,  R.  L. Ferral and R. A. McGuire.  A Generalized
      Streamflow Simulation System:  Conceptual Model for Digital Computers.
      National Weather Service, River Forecast Center, Sacramento, California,
      and the California  Dept. of Water Resources, Division of Resource Devel-
      opment,  Sacramento,  California, March 1973.

 41.   Hydrologic Research Laboratory, National Weather Service.  National
      Weather  Service  River Forecast System, Forecast Procedures.  NOAA Tech-
      nical Memorandum NWS  HYDRO-14, NWS,  U.S. Department of Commerce, Silver
      Springs, Maryland, December 1972.

 42.   Stochastics,  Inc.   Stochastic Modeling for Water Quality Management.
      EPA Water  Quality Office, D.C., Report No. 16090DUH02/71 (U.S. Govern-
      ment  Printing Office,  Stock No. 5501-0104), February 1971.

 43.   92nd  Congress.  Federal Water Pollution Control Act Amendments of 1972.
      Public Law 92-500.  Washington, D.C., October 1972.

 44.   Beckers, C. V. and Stanley G.  Chamberlain.  Design of Cost-Effective
      Water  Quality Surveillance Systems.  For EPA, Office of Research and
      Development,  Washington, D.C., January 1974.

 45.   Carroll, T. E.,  Howard M. Messner and Edward T. Rhodes.  Guide for Con-
      tract Project Officers.  EPA,  Washington, D.C., November 1971.

46.  Federal Water Pollution Control Administration.  Guidelines for Grants -
     Comprehensive River Basin Planning.  August 1967.

47.  Environmental Protection Agency.   Grant Programs - Interim Regulations.
     Federal Register, Vol. 36,  No.  229, Washington, D.C.,  November  27,  1971.

48.  Environmental Protection Agency.   Water Quality Management Plans -
     Preparation Guidelines for  States.   Federal Register,  Vol.  38,  No.  99,
     Washington, D.C., May 23, 1973.
                                     168

-------
                                  GLOSSARY
advection


aeration


algae


algorithm


ambient
augmentation
of flow

basin
benthic,
benthos
biota

bit


BOD

boundary
conditions



byte



calibration
The hydraulic mechanism by which water quality constituents
are transported in the direction of the water flow.

The process or state of being supplied or impregnated with
air.

Largely aquatic nonvascular plants that grow in either sea
water or fresh water; seaweeds and pond scum are algae.

A rule or procedure for solving a logical or mathematical
problem, frequently as incorporated into computer programs.

Surrounding.  In this handbook, a representative volume of
surrounding receiving waters.

The release of water from dam-controlled reservoirs when
stream level is low,

A region in which the strata or layers of rock dip in all
directions toward a central point.  Thus, it is any hollow
or trough in the earth's crust, whether filled by water or
not.  (Also see;  drainage basin; river basin.)

Relating to the bottom underlying a body of water (for
example, mud-dwelling mollusks are benthic organisms, or
benthos).

Living things; the plant and animal life of a region.

A basic unit of computer storage, symbolically capable of
representing only a "1" or a "0."

See oxygen demand, biochemical.

The conditions around the spatial boundary of a problem
area, which govern its solution.  Here, the forces applied
at the receiving waters' boundaries, and the flows cross-
ing them.

A few bits  (typically 6 or 8, depending upon the computer)
of computer storage, required to store one character.  See
"bit."                                                 	

The procedure of assigning values  to  the uncertain or
unknown parameters in a simulation model and adjusting
them until model predictions correspond acceptably closely
with observed prototype behavior.  See also Note 3.5.A of
Section 3.3.
                                      169

-------
 channel
 coliform
 bacteria
 compiler
 confluence
 conservative
 constituent
 constituent
 continuous
 model

 coupled
 constituent
diffusion
dilution
discharge
dispersion,
longitudinal
dissolved
oxygen (DO)
 An elemental one-dimensional flow path having the usual
 properties of a water channel, which is used to construct
 certain receiving water simulation models.

 Any of a number of organisms common to the  intestinal
 tract of man and animals,  whose presence in waste water and
 receiving waters is an indicator of pollution.

 A programmed component of  a computer, which converts
 sophisticated programming  language into elementary instruc-
 tions and binary code.

 The point at which one stream flows into another or where
 two streams converge and unite.

 A constituent (see below)  whose total mass  or quantity in
 receiving waters is conserved, even though  its  concentra-
 tion may change as a result of dilution.  See Note 1.2.A of
 Section 3.3.

 A physical, chemical or biological quantity whose presence
 in water is a factor in, or indicator of, water quality.

 A model which simulates continuously varying processes over
 a long period of time, typically many years.

 A constituent whose nonconservative behavior is affected by
 the presence of  a second constituent.   See  Note 3.2.B of
 Section 3.3.

 A process by which water quality constituents are trans-
 ported,  primarily depending upon the concentration gradients.
 Therefore can occur in directions  different  from the  flow
 direction.

 The  reduction of  a constituent concentration  by mixing  in
 water  containing  a lower concentration.

 The  volume  of water  that passes  through a given cross-
 section  of  a  channel or  sewer during a unit of  time;  com-
 monly measured in cubic  feet per second.

 The  process by which prototype concentrations are  changed as
 a result  of  the non-uniform velocity distribution  at a  chan-
 nel  cross-section.

 The  oxygen  freely  available in water.  In unpolluted water,
 oxygen is usually  present in amounts of 10 ppm  or more.
Adequate dissolved oxygen is necessary for the  life of  fish
 and  other aquatic  organisms.  About 3-5 ppm is  the lowest
 limit for support  of fish life over a long period  of  time.
                                      170

-------
dissolved
solids (DS)
distributed
load
DO

DO deficit


drainage basin


driving forces


dynamic
dynamic
equilibrium
estuary


event model
exchange
coefficient

extinction
depth (of lake)

FORTRAN
groundwater
The total amount of dissolved material, organic and
inorganic, contained in water or wastes.  Excessive dis-
solved solids can make water unsuitable for industrial
uses, unpalatable for drinking, and even cathartic.  Pot-
able water supplies may have dissolved solids content
from 20 to 1000 mg/1, but sources which have more than
500 mg/1 are not recommended by the U.S. Public Health
Service.

A constituent load which enters the receiving water over
a considerable distance, as in the case of groundwater
seepage, rather than at a point as with a sewer outfall.

See "dissolved oxygen."

The extent by which the DO concentration falls below its
saturation level.

The area which contributes runoff to a stream at a given
point (an individual section of a watershed).

The forces promoting movement in the receiving water,
primarily gravity and tides.

A process which may vary freely with time.  This includes
both the inputs and the solution in a computer model.

A process which may vary with time, but only over a
limited period (e.g., one day) which repeats itself in
cycles.  Also known as dynamic steady state.

The mouth of a river, where tidal effects are evident and
where fresh water and sea water mix.

A model which simulates the processes occurring in just a
single event, typically for a near-steady-state condition
or for only one major variation during a relatively short
period of time.

The fraction of material leaving an embayment during ebb
tide, which returns on the following flood tide.

The depth below a lake surface at which the light inten-
sity is only 1% of its intensity at the surface.

A scientific language commonly used by programmers to dir-
ect computer activities.

Water in the pores and crevices of the earth's mantle rock
which has entered them chiefly as rain water percolating
down from the surface.  As opposed to  the rain water which
runs off in streams; all water below the water table.
                                     171

-------
headwater

heat budget


hydrology


impoundments


junction



kinetics
linked
constituent
numerical
dispersion
nutrients
Manning * 3 n
model
nonconservat ive
constituent
oxygen demand,
biochemical
(oxygen-
depleting
effect; BOD)
The most upstream portion of a river, stream, or creek.

The accounting of the various factors governing water
temperature.

The science of the behavior of water in the atmosphere,
on the earth's surface, and underground.

A water reservoir or lake formed by its confinement and
storage.

In rivers, the point of connection of two upstream
stretches or segments.  In some estuary models a junction
is a segment of the estuary.

The dynamics of physical, chemical and biological react-
ion processes.

A constituent whose nonconservative behavior is affected
by the presence of one or more other constituents.  See
Note 3.2.B of Section 3.3.

Error in models using numerical approximations, caused by
the use of grids of discrete size.  Also called discreti-
zation error.

Chemical compounds upon which plant life commonly feeds.
Ammonia, nitrates, nitrites and phosphates are the most
common nutrients.

A coefficient used to describe boundary (i.e., stream
bed) roughness in hydrodynamics.

A physical, analog, or mathematical system for represent-
ing a prototype.

A constituent whose total mass reduces in receiving waters
as time proceeds through certain physical, chemical or
biological interactions.

The amount of oxygen required for aerobic bacteria to
oxidize completely the organic decomposable matter in
water within a specified time and at a given temperature -
an index to the degree of organic pollution in the water.
When discharged to a watercourse, waste containing BOD
constituents will consume dissolved oxygen in the water;
the BOD indicates the amount of oxygen used up.  Waters
that receive high BOD waste undergo reduction of oxygen and
consequent damage to aquatic life.
                                    172

-------
pollutant
pollution
(of water)
reach



reaeration

river basin
river basin
concept
section
segment
simulation
Hydrogen ion concentration which reflects the balance
between acids and alkalies.  The extreme readings are 0
and 14.  The pH of most natural water falls within the
range 4 to 9.  A pH of 7.0 indicates neutral water.  A
6.5 reading is slightly acid; an 8.5 reading is alkaline.
Slight decrease in pH may greatly increase the toxicity
of pollutants such as ammonia.  Alkaline water will tend
to form a scale; acid water is corrosive; good water
should be nearly neutral.

A constituent which pollutes waters by its presence in
excessive quantities.

Contamination or other alteration of the physical, chem-
ical or biological properties of water, including changes
in temperature, taste, color, or odor of the water, or
the discharge into the water of any liquid, gaseous,
radioactive, solid, or other substance that may create a
nuisance or render such water detrimental or injurious to
public health, safety or welfare.  Broadly, pollution
means any change in water quality that impairs it for the
subsequent user.

A discrete portion of river,  stream or creek.  For modeling
purposes a reach is somewhat homogeneous in its physical
characteristics.

See aeration.

The total area  (also called a watershed) drained by a
river system.  The river basin is increasingly coming to
be regarded as a social and economic unit for community
development and conservation of water, soil, forests and
related resources.

The notion that each river system, from its headwaters to
its mouth, is a single unit and should be treated as such.
This concept recognizes the interrelationships of resource
elements in a single basin, and assumes that multiple-
purpose development can take these interrelationships into
account.  It extends the principle of ecological balance
to the whole of the area and its occupants.

A portion of a river basin, generally larger than a segment,
which is bounded by headwaters or major river junctions.

A discrete portion of a water body of somewhat homogeneous
character, as represented in mathematical models.   (Also
see;  reach, junction.)

The representation of a system by a device that imitates
the behavior of the system.
                                      173

-------
 singular
 constituent

 stability
 steady-state
 stretch

 simulation
 period
 therraocline

 tidal  averaging
 transient



 treatment
 factor

UOD

verification
watershed

word (of
computer
storage)
 A constituent whose behavior is not affected by the pre-
 sence of other constituents.

 A characteristic of numerical models.   Unstable models
 develop large numerical errors as computations proceed.
 Numerical errors reduce in stable models as computations
 proceed.

 Quantities (e.g.,  inputs and solution)  do not vary with
 time (but may vary over space).

 See section.

 A characteristic time  for which a mathematical model
 simulates a system, using data obtained during that time
 period.

 Zone of  rapid temperature change with water depth.

 Averaging of  processes such as water currents and  pol-
 lutant  transport over  an entire tidal cycle.   This averag-
 ing may  reduce or  eliminate the need to solve for  time
 variations in tidally  influenced waters.

 A temporary and brief  time-varying solution during re-
 adjustment to equilibrium or dynamic equilibrium,  result-
 ing from a sudden  change in input(s).

 Percentage by which pollutants in effluents  are  reduced  in
 wastewater treatment.

 Ultimate  oxygen demand.   Generally about  1.5  x  5-day BOD.

 The act of  testing  a model's  accuracy using a different
 simulation  period,  i.e.,  an independent set of  input and
 output data,  from that used  in calibration.   See also
 Note 3.5.A  of  Section  3.3.

 See river basin; drainage basin.

A few bytes (typically 4 or 6, depending upon the computer)
of computer storage, required  to store one variable.  See
 "byte," "bit."                                        	
                                     174

-------
                               ABBREVIATIONS
ASCE
DEM
DRM
EPA
FWPCA
FWQA
NTIS
NWS
SCI
SEM
SSM
TTM

BOD
cf
cf s
COD
coli
csa
Cu
DO
F. coli
fps
ft
hr
in.
JCL
Ib
Ib/day
mgd
mg/L
mi.
American Society of Civil Engineers
Dynamic Estuary Model
Deep Reservoir Model
Environmental Protection Agency
Federal Water Pollution Control Administration
Federal Water Quality Administration
National Technical Information Service
National Weather Service
Systems Control, Inc.
Simplified Estuary Model
Simplified Stream Model
Tidal Temperature Model

biochemical oxygen demand (5-day)
cubic feet
cubic feet per second
chemical oxygen demand
coliform bacteria
cross-sectional area
copper
dissolved oxygen
fecal coliforms
feet per second
feet
hour
inches
job control language
pounds
pounds per day
million gallons per day
milligrams per liter
miles
                                     175

-------
  2
mi /day       - square miles per day
min           - minutes
MPN           - most probable number
NH,           - ammonia
N02           - nitrite
N03           - nitrate
              - orthophosphate
Pb            - lead
pH            - hydrogen ion concentration (see Glossary)
ppm           - parts per million (weight/weight)
Ft.           - point
sec           - second
sq ft         - square feet
SS            - suspended solids
T. coli       - total coliforms
UOD           - ultimate oxygen demand
yr            - year
A             - an increment of
                                    176

-------
                                   TECHNICAL REPORT DATA
                            (Please read Jattructions on the reverse before completing}
 4. TITLE ANDSUBTITLE
   EVALUATION OF WATER QUALITY  MODELS:   A MANAGEMENT
   GUIDE FOR PLANNERS
 7. AUTHOR(S)                  ~ "           '  ~~

   G. Paul Grimsrud, E. John Finnemore,  H. James Owen
                                                           3. RECIPIENT'S ACCESSION-NO.
                                                           5. REPORT DATE
                                                           6. PERFORMING ORGANIZATION CODE
             8. PERFORMING ORGANIZATION REPORT NO.
                                   I ESS
   Systems Control, Inc.
   1801 Page Mill Road
   Palo Alto, California  94304
             10. PROGRAM ELEMENT NO."
                 1HC 619
             11 CONTRACtVGRANT"
 TJoT
                68-01-2641
  2. SPONSORING AGENCY NAME AND ADDRESS
   Office of Research  and  Development
   Office of Air, Land and Water Use
   U.S. Environmental  Protection Agency
   Washington, D.C.  20460
              13, TYPE OF REPORT AND PERIOD COVERED
                     Final	
              14, SPONSORING AGENCY CODE ' '	
                This report is designed as a handbook specifically oriented to water
  quality and water resources planners  and managers.  It presents a  large amount of
  basic information concerning water  quality modeling including procedures for-  model
  evaluation, model selection, integration of modeling with planning activities  and
  contracting modeling projects.   Planners without previous experience  in water'quality
  modeling may use the information and  procedures included in the handbook to determine
  whether a water quality model could and should be used in a particular  plannine pro-
  gram, and which specific model would  be cost effective.  This includes  a step-bv-sten
  procedure leading to the rejection  or selection of models according to  specific pro-
  ject needs.  The handbook discusses the implications which accompany  the decision to
  model, including the needs for additional labor and specialized technical expertise
  which are generated.  Methods and procedures for integrating the use  and results of
  water quality models with other  activities of the planning process are  described as
  well as the respective merits of in-house and contracted modeling.  The handbook also
  deals with the procedures for obtaining and using contractual services  for water
  quality modeling.  Step-by-step  instructions are provided for the  preparation of
  solicitations, evaluation of proposals and selection of contractors.  This report is
  submitted in fulfillment of Contract  Number 68-01-2641, under the  sponsorship of the
  Office of Research and Development. Environmental Protection Agency
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lOENTIFIERS/OPEN ENDED TERMS
                                                                        c.  COSATI Field/Group
  Water Quality Modeling, Mathematical Modeli.
  Wastewater Management Planning,  Cost
  Effectiveness, Simulation, Performance
  Index Rating, Management, Contracting,
  Assimilation Analysis, Wasteload
  Allocation
 Model Selection,  Model
 Application, Model Cost
 Effectiveness
14A
Methods and
Equipment/
Cost Effective-
                             ness
 8, DISTRIBUTION STATEMENT

       UNLIMITED
19. SECURITY CLASS (This Report)
     UNCLASSIFIED
                                                                        22. PRICE
KPA Perm 1110-1 («-73)

 *U.S. GOVERNMENT PRINTING OFFICE: 1976-210*10:149

-------