c/EPA
          United States
          Environmental Protection
          Agency
           Office of Research and
           Development
           Washington, DC 20460
EPA/600/R-92/223
November 1992
3DFEMWATER/
3DLEWASTE: Numerical
Codes for Delineating
Wellhead Protection
Areas in
Agricultural Regions
Based on the Assimilative
Capacity Criterion

-------
                                             EPA/600/R-92/223
                                             November 1992
         3DFEMWATER/3DLEWASTE:
 NUMERICAL CODES FOR DELINEATING
    WELLHEAD PROTECTION AREAS IN
AGRICULTURAL REGIONS BASED ON THE
   ASSIMILATIVE CAPACITY CRITERION
                        by
           G.T. (George) Yeh,1 Susan  Sharp-Hansen,2
      Barry Lester,'Robert Strobl,'and Jeffrey Scarbrough4

             The Pennsylvania State University1
                University Park, PA 16802

                AQUA TERRA Consultants2
                Mountain View, CA 94043

                    GeoTrans, Inc.3
                   Sterling, VA 22170

                   AScI  Corporation4
                   Athens, GA 30613
                    Project Officer

                     Robert Carsel
             Environmental Research Laboratory
            U.S. Environmental Protection Agency
                   Athens, GA 30613

        ENVIRONMENTAL RESEARCH LABORATORY
        OFFICE OF RESEARCH AND DEVELOPMENT
        U.S. ENVIRONMENTAL PROTECTION AGENCY
                ATHENS, GEORGIA 30613
                                        Printed on Recycled Paper

-------
                                 DISCLAIMER
The work presented in this document has been funded by the United States
Environmental Protection Agency. It has been subject to the Agency's peer and
administrative review, and has been approved as an EPA  document. Mention of trade
names or commercial products does not constitute endorsement or recommendation for
use by the U.S. Environmental Protection Agency.
                                       11

-------
                                   FOREWORD
As environmental controls become more costly to implement and the penalties of
judgment errors become more severe, environmental quality management requires
more efficient analytical tools based on greater knowledge of the environmental
phenomena to be managed. As part of this Laboratory's research on the occurrence,
movement, transformation, impact, and control of environmental contaminants, the
Assessment Branch is developing management or engineering tools that can be used by
States to protect public drinking water wells from possible contamination.

The 1986 Amendments to the Safe Drinking Water Act require each State  to develop
and submit to the  U.S. EPA a wellhead protection program. As part of the program,
States must establish procedures for delineating  wellhead protection areas around each
water well or well  field which supplies  a public water system. In order to delineate
wellhead protection areas in agricultural regions using the assimilative capacity
criterion, the 3DFEMWATER/3DLEWASTE model has been developed.  These finite
element numerical codes simulate 1) flow and  transport in three-dimensional variably-
saturated porous media under transient conditions, 2) multiple distributed and point
sources/sinks, and  3)  processes which retard the transport of contaminants.

                            Rosemarie C. Russo, Ph.D.
                            Director
                            Environmental Research  Laboratory
                            Athens, Georgia
                                        in

-------
                                    ABSTRACT
The 1986 Amendments to the Safe Drinking Water Act require each State to develop
and submit to the U.S. EPA a wellhead protection program. As part of the program,
States must establish procedures for delineating wellhead protection areas around each
water well or well field which supplies a public water system. Of the five criteria that
have been suggested by the U.S. EPA for delineating wellhead protection areas, the
assimilative capacity criterion is potentially the  most accurate. It takes  into account
the reduction in concentration of contaminants being transported toward a well caused
by chemical and environmental  processes at the  land surface and in the  vadose and
saturated zones.

Nationwide,  agricultural areas are located in many diverse hydrogeologic environ-
ments. Recharge and pumping rates can vary widely within an area because of
irrigation practices and/or climate. In addition, contamination scenarios  must consider
multiple point and nonpoint source loadings of pesticides which vary both spatially and
temporally. In order  to delineate wellhead protection areas in agricultural regions
using the assimilative capacity criterion, the use of a numerical model is needed that
accounts for 1) flow  and transport in three-dimensional variably-saturated porous
media under transient conditions, 2) multiple distributed and point sources/sinks, and
3) processes which retard the transport of contaminants.

This document describes two related numerical codes, 3DFEMWATER and
3DLEWASTE, which can be used to delineate wellhead protection areas  in agricultural
regions  using the assimilative capacity criterion.  3DFEMWATER  (Three-dimensional
Finite Element Model of Water Flow Through Saturated-Unsaturated Media)
simulates subsurface flows, whereas 3DLEWASTE (Hybrid Three-Dimensional
Lagrangian-Eulerian Finite Element Model of Waste Transport Through Saturated-
Unsaturated Media) Models contaminant transport. Both codes treat heterogeneous
and anisotropic media consisting of as many geologic formations as  desired,  consider
both distributed and  point sources/sinks that are spatially and temporally dependent,
and accept four types of boundary conditions  (i.e., Dirichlet (fixed-head or
concentration), specified-flux, Neumann  (specified-pressure-head gradient or specified-
dispersive  flux), and  variable). The variable boundary  condition in  3DFEMWATER
simulates evaporation/infiltration/seepage at the  soil-air interface and, in
3DLEWASTE, simulates mass infiltration into or advection out of the system.
3DLEWASTE contains options to model  adsorption using a linear, Freundlich,  or
Langmuir isotherm,  plus dispersion, and first-order decay.

This report was submitted in partial fulfillment of Work Assignment Number 1,
Contract Number 68-CO-0019 by AQUA TERRA Consultants, under the  sponsorship of
the U.S. Environmental Protection Agency.  This  report covers the period May  1991  to
July 1992,  and work was completed as of August 1992.
                                         IV

-------
                            TABLE OF CONTENTS

                                                                          Page
Disclaimer	ii
Foreword	iii
Abstract	iv
Figures	viii
Tables	x
Acknowledgments	xi

1.0 INTRODUCTION	1
     1.1    Wellhead Protection Area Delineation	1
           1.1.1 Issues Related to Agricultural Regions	3
     1.2    The  3DFEMWATER/3DLEWASTE WHPA Model 4
           1.2.1  Experience  Required  to  Apply 3DFEMWATER/3DLEWASTE  .  . 6
           1.2.2  Implementing a 3DFEMWATER/3DLEWASTE Modeling
                  Study	6
     1.3    Organization of the Document	8

2.0 MODEL DEVELOPMENT, DISTRIBUTION, AND SUPPORT	9
     2.1    Development and Testing	9
     2.2    Distribution	10
     2.3    Obtaining a Copy of the 3DFEMWATER/3DLEWASTE Model	11
           2.3.1 Diskette	11
           2.3.2 Electronic Bulletin Board System (BBS)	11
     2.4    General/Minimum Hardware and Software Installation and Run
           Time Requirements	12
           2.4.1 Installation Requirements	12
           2.4.2 Run Time Requirements	12
     2.5    Installation	12
     2.6    Installation Verification and Routine Execution	13
     2.7    Code Modification	13
     2.8    Technical Help	14
           2.8.1 Electronic Bulletin Board System (BBS)	14
     2.9    Disclaimer	15
     2.10    Trademarks	16

3.0 BACKGROUND INFORMATION	17
     3.1    3DFEMWATER	17
           3.1.1 Governing Equations	17
           3.1.2 Boundary Conditions  and Transient  Source/Sink Terms	21
           3.1.3 Initial Conditions	28
           3.1.4 Steady-State	29

-------
     3.2    Numerical Approximation in 3DFEMWATER	30
           3.2.1 Galerkin Formulation	30
           3.2.2 Solution Techniques	32
     3.3    3DLEWASTE	34
           3.3.1 Governing Equations	37
           3.3.2 Boundary Conditions  and Transient Source/Sink Terms	42
           3.3.3 Initial Conditions	45
           3.3.4 Steady-State	45
     3.4    Numerical Approximation in 3DLEWASTE	45
           3.4.1 Galerkin Formulation	46
           3.4.2 Solution Techniques	47

4.0 DATA INPUT REQUIREMENTS	49
     4.1    3DFEMWATER Input Sequence	49
           4.1.1 Data Set 1: Title of the Simulation Run	49
           4.1.2 Data Set 2: Basic Integer Parameters	50
           4.1.3 Data Set 3: Basic Real Parameters	51
           4.1.4 Data Set 4: Printer and Disk Storage Control and Times for
                  Step Size Resetting	52
           4.1.5 Data Set 5: Material Properties	53
           4.1.6 Data Set 6: Soil Property Parameters	53
           4.1.7 Data Set 7: Nodal Point Coordinates	56
           4.1.8 Data Set 8: Subregional Data	56
           4.1.9 Data Set 9: Element Incidences	57
           4.1.10 Data Set 10: Material Type Correction	59
           4.1.11 Data Set 11:  Card Input for Initial or Pre-Initial Conditions .  . 59
           4.1.12 Data Set 12: Integer Parameters for Source and Boundary
                  Conditions	60
           4.1.13 Data Set 13: Distributed and Point Sources/Sinks	61
           4.1.14 Data Set 14: Variable Composite (Rainfall/Evaporation-
                  Seepage) Boundary Condition	64
           4.1.15 Data Set 15:  Fixed-Head (Dirichlet) Boundary Condition	68
           4.1.16 Data Set 16:  Specified-Flux  (Cauchy) Boundary Condition .... 69
           4.1.17 Data Set 17:  Specified-Pressure-Head Gradient (Neumann)
                  Boundary Condition	71
           4.1.18 Data Set 18: End of Job	73
     4.2    3DLEWASTE Input Sequence	74
           4.2.1 Data Set 1: Title of the Simulation  Run	74
           4.2.2 Data Set 2: Basic Integer Parameters	75
           4.2.3 Data Set 3: Basic Real Parameters	76
           4.2.4 Data Set 4: Printer and Disk Storage Control and Times for
                  Step Size Resetting	77
           4.2.5 Data Set 5: Material Properties	78
           4.2.6 Data Set 6: Nodal Point Coordinates	78
           4.2.7 Data Set 7: Element Incidences	79
           4.2.8 Data Set 8: Subregional Data	80
           4.2.9 Data Set 9: Material Type Correction	81
           4.2.10 Data Set 10:  Card Input for Initial or Pre-Initial Conditions .  . 82
           4.2.11 Data Set 11: Integer Parameters for Sources and Boundary
                  Conditions	82

                                        vi

-------
             4.2.12 Data Set 12:  Distributed and Point Sources/Sinks	84
             4.2.13 Data Set 13:  Variable Composite Boundary Condition	87
             4.2.14 Data Set 14:  Prescribed-Concentration (Dirichlet) Boundary
                    Condition	89
             4.2.15 Data Set 15:  Specified-Flux (Cauchy) Boundary  Condition .... 90
             4.2.16 Data Set 16:  Specified-Dispersive-Flux (Neumann) Boundary
                    Condition	93
             4.2.17 Data Set 17:  Hydrological Variables	95
             4.2.18 Data SetlS: End of Job	96

  5.0 PARAMETER SELECTION	97
       5.1    3DFEMWATER	97
             5.1.1 Data Set 1: Title of the Simulation Run	97
             5.1.2 Data Set 2: Basic Integer Parameters	98
             5.1.3 Data Set 3: Basic Real Parameters	99
             5.1.4 Data Set 5: Material Properties	102
             5.1.5 Data Set 6: Soil Property Parameters	106
       5.2    3DLEWASTE	110
             5.2.1 Data Set 1: Title	110
             5.2.2 Data Set 2: Basic Integer Parameters	     Ill
             5.2.3 Data Set 3: Basic Real Parameters	      113
             5.2.4 Data Set 5: Material Properties	     114
             5.2.5 Data Set 17: Hydrological Variables	      119

  6.0 EXAMPLE PROBLEMS	      121
       6.1    3DFEMWATER	      121
             6.1.1 One-Dimensional Column	      121
             6.1.2 Two-dimensional Drainage Problem	      122
             6.1.3 Three-Dimensional Pumping Problem	      125
       6.2    3DLEWASTE	      144
             6.2.1 One-Dimensional Transport Problem	      144
             6.2.2 Two-Dimensional Transport in a Rectangular Region ....      147

  7.0 REFERENCES	      150

APPENDIX A	      153
APPENDIX B	      180
APPENDIX C	      182
APPENDIX D	      187
                                        VII

-------
                                     FIGURES
                                                                               Page

3.1.    Uncofined aquifer to be approximated as a variably-saturated
       porous medium	18

3.2.    Variable pore spacing in soil under saturated flow conditions	20

3.3.    Logarithmic plot of constitutive relations for sand, clay, silty
       loam, and sandy loam: (a) moisture content vs. pressure head
       and (b) relative permeability vs. moisture content	22

3.4.    Conceptual model and mathematical approximation for variably-
       saturated flow system. Within the modeled system, transient
       source/sink terms may be applied as point sources/sinks or as
       distributed sources/sinks	23

3.5.    Use of a pressure-head gradient boundary condition to simulate a
       portion of the unsaturated zone	25

3.6.    Using a series of nodes to represent a screened well interval	27

3.7.    Pressure head versus time at a nodal point on the finite element
       grid	28

3.8.    Pressure head versus time at nodal point where steady-state
       solution is being approached	29

3.9.    Finite element grid for production from a single well in a
       variably-saturated porous medium	30

3.10.   Solution scheme for unsaturated flow analysis	33

3.11.   Use of vertical or horizontal nodal slices in the block
       iterative method	35

3.12.   Migration of dissolved contaminants through the  unsaturated zone
       into unconfined aquifer system	36

3.13.   Diagram showing the effect of scale on hydrodynamic dispersion
       processes	38
                                        Vlll

-------
3.14.   Backward particle tracking to determine the starting point of an
       advected particle	42

4.1.    Node numbering convention for the elements	58

4.2    Node numbering convention for the elements	80

5.1    Longitudinal dispersivity versus scale with data classified
       by reliability	   117

5.2.    Nomograph for determining Darcy velocity	   120

6.1.    One-dimensional transient flow through a soil column	   122

6.2.    Two-dimensional steady-state flow to parallel drains	   124

6.3.    Three-dimensional steady-state  flow to a pumping well	   128

6.4.    One-dimensional transient transport through a horizontal column	   144

6.5.    Two-dimensional transient transport in a rectangular region	   147
                                         IX

-------
                                     TABLES

                                                                              Page

5-1.   Water Density as a Function of Temperature	101

5-2.   Dynamic Viscosity of Water as a Function of Temperature	   102

5-3.   Range of Hydraulic Conductivity Values for Various Geologic
      Materials	   104

5-4.   Variability in Horizontal and Vertical Hydraulic Conductivities	   105

5-5.   Permeability of Porous Materials	   105

5-6.   Descriptive Statistics for Saturation Water Content (0S)
      and Residual Water  Content (9r)	   107

5-7.   Descriptive Statistics for van Genuchten Water Retention Model
      Parameters, a, p, and y	   108

5-8.   Mean Bulk Density (g/cm3) for Five Soil Textural
      Classifications	115

5-9.   Range and Mean Values of Dry Bulk Density for Various  Geologic
      Materials	   116

6-1.   Input Data Set for the One-Dimensional 3DFEMWATER  Problem	   123

6-2.   Input Data Set for the Two-Dimensional 3DFEMWATER  Problem	   126

6-3.   Input Data Set for the Three-Dimensional 3DFEMWATER Problem	   130

6-4.   Input Data Set for the One-Dimensional 3DLEWASTE Problem	   145

6-5.   Input Data Set for the Two-Dimensional 3DLEWASTE Problem	   148
                                        x

-------
                             ACKNOWLEDGEMENTS
This document was prepared under Work Assignment No. 1 of Contract No.  68-CO-0019
by AQUA TERRA  Consultants for the U.S. Environmental Protection Agency Office of
Research and  Development. Robert Carsel of the Environmental Research Laboratory
in Athens,  Georgia was the Technical Project Monitor and the Project Officer. We thank
him for his continuous technical and  management support throughout the course of this
project.

The 3DFEMWATER/3DLEWASTE  code was developed by G.T. (George)  Yeh of The
Pennsylvania State University. Robert Strobl, also at The Pennsylvania State
University, upgraded the code to meet U.S. Environmental Protection Agency coding
conventions. John  Kittle at AQUA  TERRA Consultants reviewed the code and
suggested modifications.

The original documentation of the 3DFEMWATER/3DLEWASTE code was prepared by
G.T. Yeh. That documentation  was substantially expanded and rewritten during the
course of this project.  At AQUA TERRA Consultants, Susan Sharp-Hansen was
responsible for rewriting the documentation.  She was assisted by Barry Lester of
GeoTrans, Inc., who wrote Section 2 and part of the introduction; by Robert Strobl, who
prepared some of the tables in the appendices; and by Jeff Scarborough of AScI
Corporation, who applied the code to  the example problems. John Imhoff,  the Project
Manager, supplied  administrative guidance and  he, Anthony Donigian, and John Kittle
reviewed the document. Technical reviewers  also included David Ward  and  Jeff
Benegar of GeoTrans,  Inc. Word processing was performed by Dorothy Inahara.
                                        XI

-------
                                    SECTION 1

                                  INTRODUCTION
This document describes two related numerical codes, 3DFEMWATER and
3DLEWASTE. Together these codes can model flow and transport in three-dimensional
variably-saturated porous media under transient conditions, with multiple distributed
and point sources/sinks, and considering processes which retard the transport of
contaminants (i.e., dispersion, decay and adsorption). Thus, they can be used to apply
the assimilative capacity criterion to the development of wellhead protection areas.
Background information about wellhead protection area  delineation criteria and
methods is provided in Section 1.1. The features and implementation of the
3DFEMWATER/3DLEWASTE codes are discussed in Section 1.2 and the contents of this
document are summarized in Section  1.3.

It is important to note that the version of 3DFEMWATER/3DLEWASTE documented in
this user's manual has substantial CPU time  requirements. A faster version of the
model is currently being developed.

1.1  WELLHEAD  PROTECTION  AREA  DELINATION

The 1986 Amendments to the Safe Drinking Water Act require each State to develop
and submit  to the U.S. EPA a wellhead protection program. As part of the program,
States must establish procedures for delineating wellhead protection areas around each
water well or well field which supplies a public  water system.  A wellhead protection
area (WHPA) is defined as the surface and subsurface area surrounding a water well or
well field through which contaminants are likely to be transported and  reach the well or
wellfield. Within the WHPA, contaminant sources need to be assessed and managed to
prevent pollution of public drinking water supplies. Existing WHP  programs are
generally aimed at one of the  following overall protection goals:

       •  Provide a remedial  action zone to protect wells from unexpected contaminant
          releases.

       •  Provide an attenuation zone to bring concentrations  of specific contaminants
          to desired levels at  the time they reach the wellhead.

       •  Provide a well-field management zone in  all or part of a well's present or
          future recharge area.

Five criteria have been suggested by the U.S. EPA (U.S.  EPA,  1987) for delineating
wellhead protection areas that will adequately protect public water supplies. The
criteria are:

      •   Distance, which considers a radial distance from the pumping well.

-------
      •  Drawdown, which considers an area within which an aquifer's potentiometric
         surface has been lowered by pumping.

      •  Time of travel, which considers the time required for a contaminant to move
         through the subsurface to a well (often only considering advection).

      *  Flow system boundaries, which consider the geographic or hydrologic fea-
         tures that control groundwater flow.

      •  Assimilative capacity, which considers environmental factors which reduce
         the concentration  of contaminants transported to a well.

One or more of the criteria may be used. The most technically demanding, but also
potentially the most accurate, is the  assimilative capacity criterion. The assimilative
capacity  criterion takes into account the reduction in concentration of contaminants
being transported toward a well caused by chemical and environmental  processes at  the
land surface and in the vadose and saturated zones.

The U.S. EPA has described six methods for applying the criteria to the delineation of
WHPAs.  Listed in order of  difficulty, the methods are:

      Arbitrary fixed  radius, which involves  drawing  a circle around  a  well. The
         radius of the circle can be based on professional judgment or an established
         distance criterion.

      •  Calculated fixed radius, in which the radius of a circle around  the well is
         determined from an  equation that considers the volume of water pumped from
         a well over a specified time.

      •  Simplified variable shapes, which makes use of "standardized forms" repre-
         senting various hydrogeologic and pumping conditions. The  set of
         standardized forms are initially prepared using an analytical model.
         Subsequent application involves selecting the  most appropriate shape for a
         given well.

      •  Analytical methods, which involve  the application of analytical  groundwater
         flow and transport models.

      •  Hydrogeologic mapping,  which makes use of geologic, geophysical, and dye
         tracing techniques to map a WHPA.

      •  Numerical models,  which involve the application of numerical models of flow
         and solute transport in the subsurface.

Application  of the first three methods is suitable for only a very limited  number of sites,
such as extensive, homogeneous, single aquifers with a relatively flat  potentiometric
surface. While analytical methods are usually  more technically  accurate than the first
three methods, their application is  still restricted to relatively simple  hydrogeologic
environments. Hydrogeologic mapping may be the only reasonable method under some
hydrogeologic conditions, such as karst or fractured aquifers. However, it will also be

                                          2

-------
necessary for some hydrogeologic mapping to be performed for application of either
analytical or numerical models.

Numerical models provide the greatest flexibility and accuracy in representing complex
environments and can be applied to nearly all types of hydrogeologic settings. The
models can also be used to predict the dynamic aspects of the WHPA, such as changes
in the size of the WHPA resulting from natural or man-made effects. Disadvantages for
this method include costs that are high relative to other methods and the need for
considerable technical expertise  in hydrogeology and modeling. The cost may be
warranted in areas where a high degree  of accuracy is desired, however.  Also, due to
limitations on model grid spacing and density, numerical models are sometimes less
suitable than analytical methods for assessing drawdowns close to pumping wells.

The more rigorous the method used for WHPA delineation, the smaller the WHPA can
be without risking underprotection and the associated potential for water quality
degradation. When a smaller WHPA  can  be defined without generating unacceptable
risk, land use restrictions can be kept to  a minimum along with the potential economic
hardships associated with such land use restrictions. The choice of WHPA delineation
methodology becomes a decision  based on generating an acceptable margin of safety,
while balancing the economic hardships to affected parties with the technical and
economic feasibility of minimizing the WHPA.

1.1.1 Issues  Related  to Agricultural Regions

Nationwide, agricultural areas are located in many diverse hydrogeologic environments
(e.g., multiple aquifer systems, fractured  and/or karst systems, and systems with wide
variations in depth to the water table). In addition, recharge can vary widely because of
irrigation practices and/or climate. Also,  domestic and irrigation wells, which pump at
different and varying rates, are commonly located throughout agricultural regions.
Therefore, the ability to model transient flow conditions  (i.e., transient recharge, a
fluctuating water table, and transient pumping from a variety of points in x,y,z space)
for a wide variety of  hydrogeologic conditions is important.

Contamination scenarios in agricultural regions must consider multiple point and
nonpoint source loadings that vary both spatially  and temporally. For  example, spills,
leaks, or the direct introduction of chemicals into well casings can result in point sources
of contamination, whereas chemical application to fields  can  result in nonpoint sources
of contamination.

Pesticide loadings to  the subsurface are affected by both surface processes and
agricultural management practices.  Examples include runoff, erosion,  chemical
volatilization, evapotranspiration, tillage  practices, and the method, amount, and timing
of pesticide application. Most of  these processes require  detailed modeling of the  surface
environment or zone  and are not addressed in models of subsurface flow and transport.
Therefore, it is suggested that text or matrix ranking or the  separate application  of an
existing model be used to estimate recharge and solute loading from the surface zone to
the subsurface (e.g., PRZM-2, see Mullins et al., 1992).

The contaminants of concern in agricultural regions are  predominantly organic
pesticides and nitrates. Pesticides are typically present in the subsurface in dilute

                                          3

-------
 concentrations. Because interest in agricultural areas is likely to focus on dilute organic
 pesticides, issues such as the transport of metals, the interactions of complex mixtures,
 or immiscible flow are not addressed by this model. Also,  because of the complexity of
 the processes associated with the transport of nitrates, nitrate contamination can not  be
 adequately modeled using this version of 3DFEMWATER/3DLEWASTE.

 1.2 THE 3DFEMWATER/3DLEWASTE  WHPA MODEL

 3DFEMWATER (Three-dimensional Finite Element Model of Water Flow Through
 Saturated-Unsaturated Media) can be used to investigate subsurface flows as a stand-
 alone model, or it can be used to provide the hydrologic flow variables required by
 3DLEWASTE. The special features of 3DFEMWATER are its flexibility and versatility
 in modeling a wide range of real-world problems. The model is designed to:

      •   Treat heterogeneous and anisotropic media consisting of as many geologic
          formations as desired.

      •   Consider both distributed and point sources/sinks that are spatially and
          temporally dependent.

      •   Accept prescribed initial conditions or obtain them by simulating a steady-
          state version of the system under consideration.

      •  Deal with a transient head variation over a freed-head (Dirichlet)  boundary.

      •   Handle time-dependent fluxes due to a varying pressure gradient along a
          specified-pressure-head gradient (Neumann) boundary.

      •  Treat time-dependent total fluxes distributed over a specified-flux  (Cauchy)
          boundary.

      •  Automatically determine variable  boundary conditions of evaporation, infiltra-
          tion, or seepage at  the soil-air interface.

      •   Include the off-diagonal hydraulic conductivity components in the modified
          Richard's equation  in order to deal with cases when the coordinate system does
          not coincide with the principal directions of the hydraulic conductivity tensor.

      •  Provide three  options (exact, under-, and over-relaxation) for estimating the
          nonlinear matrix.

      •  Include two options (successive subregion block iterations and  successive point
         iterations) for solving the linearized matrix equations.

      •  Automatically reset the time-step  size when boundary conditions or
          sources/sinks change abruptly.

3DLEWASTE (Hybrid Three-Dimensional Lagrangian-Eulerian Finite Element Model of
Waste Transport  Through Saturated-Unsaturated Media) uses a hybrid  Lagrangian-
Eulerian approach. In comparison to conventional finite element  (including  both

-------
Galerkin and upstream finite element) or finite difference (including both central and
upstream finite difference)  models,  3DLEWASTE offers several advantages. First, it
completely eliminates numerical oscillation due to advection terms. Second, it can be
applied to mesh Peclet numbers ranging from zero to infinity. (Conventional finite
element or finite difference models typically impose severe restrictions on the mesh
Peclet number.) Third, it can use very large time-step sizes to greatly reduce numerical
dispersion. In fact, the larger the time step, the better is the solution with respect to
advective transport. The time-step size is only limited by the accuracy requirement with
respect to diffusive/dispersive transport, which is normally not a very severe restriction.
Finally, the hybrid Lagrangian-Eulerian finite element approach is always superior to,
and will never be worse than, its corresponding upstream finite element method.

The 3DLEWASTE model is designed to:

      •  Treat heterogeneous and anisotropic media.

      •  Consider spatial and temporal distribution, as well as point sources/sinks.

      •  Accept prescribed  initial conditions or obtain them by simulating a steady-
         state version of the system under consideration.

      •  Deal  with transient concentrations distributed over prescribed concentration
         (Dirichlet) boundaries.

      •  Handle time-dependent fluxes over variable boundaries.

      •  Deal with time-dependent total fluxes over specified-flux  (Cauchy) boundaries.

      •  Handle time-dependent fluxes over specified  dispersive-flux (Neumann)  bound-
         aries.

      •  Include the off-diagonal dispersion coefficient tensor components in the govern-
         ing equation for dealing with cases when the coordinate system does not
         coincide with the  principal directions of the dispersion coefficient tensor.

      •  Provide two options of treating the mass matrix—consistent and lumping.

      •  Provide three options  (exact, under-  and over-relaxation)  for estimating  the
         nonlinear matrix.

      •  Include a block iteration method to solve the linearized matrix equations to
         eliminate the excessive storage demands of a direct band  matrix solution.

      •  Automatically reset the time-step size when boundary conditions or
         sources/sinks change abruptly.

      •  Simulate first-order contaminant decay.

      •  Include three adsorption models—a linear isotherm and a nonlinear Freundlich
         or Langmuir isotherm.

-------
 1.2.1 Experience Required to Apply 3DFEMWATER/3DLEWASTE

 The complexity and sophistication of the 3DFEMWATER/3DLEWASTE numerical codes
 limits the number of people who can successfully use the codes to apply the assimilative
 capacity criterion in wellhead protection area delineation. The user community is
 expected  to be State personnel,  as well as personnel at the U.S. EPA headquarters and
 regional offices, who are experienced numerical modelers with a strong background in
 hydrogeology.

 1.2.2 Implementing a 3DFEMWATER/3DLEWASTE Modeling Study

 Implementation of a  3DFEMWATER/3DLEWASTE modeling study represents a highly
 rigorous evaluation of a wellhead site. The study is generally aimed at delineating the
 WHPA with a high degree of certainty.  The project team can take into consideration the
 specific nature of present and future wellfields, the physical and chemical nature of
 potential  contaminant sources, the effect of human activities, as well as the complexity
 of the groundwater flow system through which the contaminants travel.

 Although 3DFEMWATER/3DLEWASTE  studies can provide flexibility  in defining  the
 hydrogeologic environment and  contaminant sources, they are limited by the quantity
 and quality of physical and chemical data available to define the system. When seeking
 to define  the zone of contribution in a WHPA using a 3DFEMWATER/3DLEWASTE
 analysis,  there is a law of diminishing returns. The economic benefits  gained from being
 able to minimize land use restrictions must be weighed against the costs of generating
 the necessary data and applying the model.

 Wellfield  geometry and the spatial distribution of wells within a field can strongly affect
 subsurface flow at regional and local scales. Using the 3DFEMWATER/3DLEWASTE
 model, an investigator can consider the influence of a wellfield on the regional flow
 system. On the local scale, the effects of partial penetration associated with well
 screening intervals can also be  considered. Localized flow patterns, which result from
 perturbations to the flow field and the heterogeneous nature of the geologic medium,
 influence  the movement of dissolved contaminants and determine 1) the amount of time
 required for a dissolved species  to reach the wellfield and 2) the degree of attenuation of
 the species as it approaches the field.

 The 3DFEMWATER/3DLEWASTE model also allows  the user to examine the influence
 of temporal changes in well production  on contaminant mobility. The influence of
 seasonal variations in well production and other periodic  variations (i.e.,  drought
 conditions, unseasonably warm summers, etc.), can strongly affect the potential for a
 contaminant to reach a wellfield at unacceptable levels or in an unacceptable amount of
 time. The temporal variations in well production can be considered in  conjunction with
 associated temporal changes in recharge and evapotranspiration rates.

As implied above, the  3DFEMWATER/3DLEWASTE model is not limited to
 discretization of the flow field into regularly shaped prismatic blocks (i.e. triangular and
 rectangular prisms). Therefore,  consideration of the heterogeneous nature of a  modeled
system is mainly limited by either the availability of data or the computational power of
the computer utilized. There is a practical limitation on the degree of heterogeneity
which can be simulated, based on the conflict between the grid block-size restrictions

                                          6

-------
needed to circumvent convergence problems and the number of blocks that a computer
can handle in a time-efficient manner. Within these restrictions, it is the model user's
goal to maximize the extent to which the influence of soil and rock-type heterogeneities
affect the flow system.

In nature, heterogeneities generate a strong control on the local pathways that the
dissolved chemicals will follow. The tendency for water to flow through low resistance
(high conductivity) pathways provides a  short-circuiting effect that can accelerate the
movement of chemicals to a wellfield. In contrast, occurrences of high resistance (low
conductivity) media between the source and the screened intervals of wells can inhibit
the contaminant from reaching the water supply or attenuate the contaminant to safe
concentration levels  before it reaches the water supply. The uncertainty associated with
a WHPA analysis is directly related to the presence of heterogeneity in the aquifer
properties. As the degree of heterogeneity decreases, the possibility of underestimating
or overestimating the chemical migration is reduced. On  the  other  hand, the potential
for contamination is  most uncertain when using bulk properties or using ad hoc
variances in the values of effective porosity, dispersivity and hydraulic conductivity.

Since the flow portion of 3DFEMWATER/3DLEWASTE simulates variably-saturated
conditions, a more accurate model of water storage in unconfined or partially confined
systems  can be generated.  The user can consider draining (and filling) of pore  spaces
above  the water table, which can damp the effect of time-variant changes in well produc-
tion, recharge and evapotranspiration on the flow system. Rigorous representation of
the unsaturated zone also permits examination of the influence of variable saturation on
the mobility of contaminants. Vertical infiltration through the unsaturated zone and
the associated lateral spreading of contaminants, due to the occurrence of sediment
lenses of various grain  sizes, can be considered.  Explicit simulation of the  unsaturated
zone also allows for direct consideration  of the contaminant storage capacity of the
unsaturated zone. This more accurately  depicts the role of the unsaturated zone as  a
source of contaminant infiltration into the saturated zone. The availability of different
adsorption models (linear, Freundlich and Langmuir) allows the user to choose a
contaminant storage capacity appropriate for  the waste being modeled.

The  3DFEMWATER/3DLEWASTE model includes a relatively rigorous representation of
contaminant sources by using a variety of time-dependent boundary conditions.
Contaminant sources may be represented not only as point sources  or sources of simple
geometry, as assumed in analytical solutions, but also as  sources of variable geometry.
Where applicable, contaminants already present in the subsurface  water and solid
matrix at the start of a modeling study can also be simulated. The use of infiltration or
recharge options available in the 3DFEMWATER/3DLEWASTE model provides a good
method of simulating contaminant sources such  as spatially-  and temporally-variant
pesticide or  fertilizer applications to agricultural areas.

The interaction of the regional flow field and  local wellfield perturbations can be
handled  in two ways using 3DFEMWATER/3DLEWASTE. The localized flow field may
be implemented as a freely discretized portion of the larger system where the boundary
conditions are generally associated with the regional flow field. The problem can also be
broken up into two problems of different scales, where  the regional  flow system is
modeled  for flow only and the local system is  modeled for flow and transport with the
boundary conditions  generated from the regional flow model.  The degree of interaction

                                          7

-------
between the two models is dictated by the degree of accuracy desired, and the placement
of local system boundaries.

1.3 ORGANIZATION OF THE DOCUMENT

This documentation contains the information needed to understand and apply the
3DFEMWATER/3DLEWASTE  codes to wellhead protection area delineation problems.
Section 2 contains information on model distribution and support. In Section 3, back-
ground related to the model equations, features, and numerical approximation
techniques  is presented. Section 4 is a guide to the construction of input  data sets for
the code. Assistance  in explaining and estimating some of the input  parameters is
provided in Section 5. Five simple example problems, including the corresponding input
data files, are given  in Section 6. The appendices contain  more detailed information
about the numerical  codes, including descriptions of the subroutines,  and listings of the
maximum control parameters and program variables.

-------
                                   SECTION 2

             MODEL DEVELOPMENT,  DISTRIBUTION, AND SUPPORT
NOTE: Refer to the READ.ME file for the latest supplemental information, changes,
and/or additions to the 3DFEMWATER/3DLEWASTE model documentation. A copy of
the READ.ME file is included on each distribution diskette set or it can be downloaded
from the Center for Exposure Assessment Modeling  (CEAM)  electronic bulletin board
system (BBS). It can be installed on a hard disk using the INSTALL  (diskette) or
INSTALPZ (BBS) program. It is an ASCII (non-binary) text file that can be displayed
on the monitor screen by using the DOS TYPE command (e.g., TYPE READ. ME) or
printed using the DOS  PRINT command (e.g.,  PRINT READ.ME).

The READ.ME file contains a section entitled File Name and Content that provides a
brief functional description of each 3 DFEM WATER/3 OLE WASTE file by name or file
name extension type. Other sections in this document contain further information
about:

      •   System development tools used to build the microcomputer  release of the
          3DFEMWATER/3DLEWASTE  model.

      •   Recommended hardware and software configuration for execution of the model
          and all support programs.

      •   Program execution.

      •   Minimum file configuration.

      •   Sample run times.

      •   Program modification.

      •   Technical support.

2.1  DEVELOPMENT AND TESTING

The 3DFEMWATER/3DLEWASTE model was developed and tested on a Digital Equip-
ment Corporation (DEC) VAX6310 running under version 5.4-2 of the VMS operating
system (OS)  and version 5.5-98 of VAX VMS FORTRAN- 77, and an Advanced Logic
Research (ALR) 486/25  microcomputer running under version 4.00 of IBM PC DOS and
version 2.51  of Salford  FORTRAN (FTN77/486). The following FORTRAN tools were
also used to  perform static evaluations of the 3DFEMWATER/3DLEWASTE FORTRAN
code on an IBM PS/2 Model 80-071 running under version 3.3 of IBM  PC! DOS, MICRO
EXPRESS (ME) 486/25  and 486/33 systems running  under version 5.00 of Microsoft
(MS) DOS, and a Sun SPARCstation 1+GX running  version  4.1.1  of UNIX/SunOS:

-------
       •  Ryan-McFarland FORTRAN versions 2.45, 3.10.01 (RMFORT).

       •  Microsoft FORTRAN  version 5.00 (MSFORT).

       •  Lahey FORTRAN versions 5.01, 4.02 (F77L, F77L-EM/32).

       •  Waterloo FORTRAN  version 8.5E (WATCOM  FORTRAN-77/386).

       •  Sun FORTRAN version  1.4.

       •  Silicon Valley FORTRAN version 2.81 (SVS FORTRAN-77/386).

In addition to the VAX and ALR systems, 3DFEMWATER/3DLEWASTE has also been
successfully executed on a PRIME 50 Series minicomputer running under PRIMOS, the
Sun SPARCstation, and the IBM PS/2 Model 80-071.

2.2 DISTRIBUTION

The 3DFEMWATER/3DLEWASTE model and all support files and programs are
available on diskette from the Center for Exposure Assessment Modeling, located at the
U.S. EPA Athens  Environmental Research Laboratory, Athens, Georgia, at no charge.
The CEAM has an exchange diskette policy. It is preferred that diskettes be received
before sending a copy of the model system (refer to Section 2.3, Obtaining a Copy of the
3DFEMWATER/3DLEWASTE  Model).

Included in a distribution diskette set are:

       •   3DFEMWATER/3DLEWASTE general execution and user support guide
          (READ.  ME) file.

       •   Interactive installation program (refer to  Section 2.5).

       •   Test input and output files for installation verification.

       •   Executable task image file for the 3DFEMWATER/3DLEWASTE model.

       •   FORTRAN source code files.

       •   Command and/or "make" files to compile, link, and run the task image file
          (*. EXE).

A FORTRAN compiler and link editor are NOT required to execute any portion of the
model. If the user wishes to modify the model, it will be up to the user to supply and/or
obtain:

       •   An appropriate text editor that saves files in ASCII (non-binary) text format.

       •   FORTRAN development tools to recompile and link edit any portion of the
          model.
                                        10

-------
CEAM cannot support, maintain, and/or be responsible for modifications that change the
function of the executable task image, MAKE, or DOS command files supplied with this
model package.

The microcomputer release  of the 3DFEMWATER/3DLEWASTE model is a full
implementation of the VAX/VMS version.  The microcomputer implementation of this
model performs the same function as the U.S. EPA mainframe/minicomputer version.

2.3 OBTAINING A COPY OF THE 3DFEMWATER/3DLEWASTE MODEL

      NOTE:      The following abbreviations are used below to represent different
                   quantities of computer memory:

                    1 k = 1 kilobyte = 1,024 bytes
                    1 m =  1 megabyte = 1 "meg" = 1,048,576 bytes
                   lb=lbyte

2.3.1  Diskette

To obtain  a copy of the 3DFEMWATER/3DLEWASTE distribution model package  on
diskette,  send:

      •   The appropriate number of double-sided, double-density (DS/DD 360kb) 5.25
          inch, or double-sided, high-density (DS/HD 1.44mb)  3.5 inch error-free dis-
          kettes.

          NOTE: To obtain the correct number of diskettes, contact CEAM  at
          706/546-3549.

      •   A cover letter, with a complete return address requesting the
          3DFEMWATER/3DLEWASTE model to:

          Model Distribution Coordinator
          (ATTN:  Catherine E. Green, CSC)
          Center for Exposure Assessment Modeling
          Environmental Research Laboratory
          U.S. Environmental Protection Agency
          Athens,  GA  30613-0801

Program and/or user documentation,  or instructions on how to order documentation, will
accompany each response.

2.3.2 Electronic Bulletin Board System (BBS)

To download a copy of the 3DFEMWATER/3DLEWASTE model, or to check  the status of
the latest  release of this model or any other CEAM software product, call the CEAM
BBS 24  hours a day, 7 days a week. To access the BBS, a computer with a modem and
communication software are needed. The phone number for the BBS is 706/546-3402.
Communication parameters for the BBS are:
                                        11

-------
       •300/1200/2400/9600 baud rate.

       • 8 data  bits.

       • No  parity.

       •  1 stop  bit.

In order to access the BBS at 9600 baud, a USRobotics Courier HST modem must be
used.

2.4 GENERAL/MINIMUM HARDWARE AND SOFTWARE  INSTALLATION AND RUN
TIME  REQUIREMENTS

NOTE: Refer to the READ.ME file for the latest supplemental and more complete
information, changes, and/or additions concerning specific hardware and software
installation and run time requirements.

2.4.1  Installation Requirements

       •   3.5 inch, 1.44mb  diskette drive, or 5.25 inch, 360kb diskette drive.

       •   Hard disk  drive.

       •   Approximately 8mb free hard disk storage,

2.4.2  Run Time  Requirements

       •  386 or  486 compatible  microcomputer.

       •  MS or  PC DOS version 3.30 or higher.

       •  640k base  memory.

Extended memory and hard disk storage requirements will vary with the size of the
problem being simulated. Requirements for  problems  similar to those found in Section 6
are:

       •  2mb of extended  (XMS) memory.

       •  4mb free hard disk storage.

Refer to READ.ME file for suggested modification of the  CONFIG.SYS and/or
AUTOEXEC.BAT DOS system configuration and start-up files.

2.5 INSTALLATION

To install the 3DFEMWATER/3DLEWASTE model and/or related support files on a
hard disk, insert the first distribution diskette in a compatible diskette drive (refer to
Section 2.4). Then type:

                                        12

-------
                            A:\INSTALL or B:\INSTALL

at the DOS system prompt and press the  key. Then follow instructions and
respond to prompts presented on the monitor screen by the interactive installation
program.  Complete installation instructions are also printed on each  external diskette
label. The 3DFEMWATER/3DLEWASTE distribution diskette sets implement software
product installation standards to insure the most error-free, maintainable, and user-
acceptable distribution of CEAM products. It has a "unique menu option, command, full-
screen (interactive), diagnostic, error-recovery, help, and selective installation
capabilities using state-of-the-art human-factors engineering practices and principles.

NOTE: The contents of the distribution diskettes can be copied  to another set of
"backup" diskettes using the DOS DISKCOPY command. Refer  to the DOS Reference
Manual for command application and use. The "backup" diskettes must be the same
size and storage density as the original, source diskettes.

2.6 INSTALLATION VERIFICATION  AND  ROUTINE EXECUTION

Refer to the following sections in the READ.ME file for complete instructions concerning
installation verification and routine  execution of the 3DFEMWATER/3DLEWASTE
model:

       •   File name and content.

       •   Routine  execution.

       •  Run time and performance.

       •   Minimum file configuration.

2.7 CODE MODIFICATION

Included in the diskette set are:

       •   An executable task image file for the  3DFEMWATER/3DLEWASTE model.

       •   FORTRAN source code files.

       •   Command and/or "make" files to compile, link, and run the  task image file
          (*.EXE).

If the user wishes to modify the model  or any other program, it will be up to the user to
supply and/or obtain:

       •   An appropriate text editor that saves files in ASCII  (non-binary) text format.

       •   FORTRAN development tools to recompile and link edit any portion of the
          model.
                                         13

-------
CEAM cannot support, maintain, and/or be responsible for modifications that change the
function of any executable task image (*. EXE), DOS batch command (*. BAT), and/or
"make" utility file(s) supplied with this model package.

2.8 TECHNICAL  HELP

For questions and/or information concerning:

      •   Installation and/or testing of the 3DFEMWATER/3DLEWASTE model and/or
          support programs or files, call 706/546-3590, 3548 for assistance.

      •   3DFEMWATER/3DLEWASTE model and/or program content,  application,
          and/or theory, call 706/546-3 171 for assistance.

      •   Use of the CEAM electronic bulletin board system (BBS), contact the BBS
          system operator (SYSOP) at 706/546-3548.

      *   CEAM software and distribution Quality Assurance and Control, call 706/546-
          3634.

      •   Other environmental software and documentation distributed through CEAM,
          contact the Model Distribution  Coordinator at 706/546-3549.

      •   Other support available through CEAM, contact Mr. Robert B. Ambrose,  Jr.,
          CEAM Manager:

             By mail at the following address:

                Center  for Exposure Assessment Modeling (CEAM)
                Environmental Research Laboratory
                U.S. Environmental Protection Agency
                Athens,  Georgia 30613-0801

          -   By telephone at  706/546-3130.

             By fax at 706/546-2018.

             Through the CEAM BBS message menu and commands. The CEAM  BBS
             communication parameters and telephone number are listed above.

2.8.1 Electronic  Bulletin  Board System (BBS)

To help  technical staff provide better assistance, write down a response to the following
topics before calling or writing. If calling,  be at the computer,  with the  computer on,
and in the proper sub-directory when the call is placed.
                                         14

-------
       Program information:

       Describe the problem, including the exact wording  of any  error and/or
          warning  message (s).

       List  the exact steps,  command (s), and/or keyboard  key  sequence that will
          reproduce the problem.

       Machine information:

       List  computer  brand  and  model.

       • List available RAM  (as reported by DOS CHKDSK command).

       • List available extended memory (XMS).

       •  List  name and version of extended memory (XMS) manager (i.e., HIMEM,
          VDISK, RAMDRIVE,  etc.).

       •  List  available hard  disk space (as reported by DOS CHKDSK command).

       •  List  the brand and  version  of DOS (as reported by DOS VER command).

       •  List  the name of any memory resident program(s) installed.

       •  Printer brand and model.

       •  Monitor brand and model.

NOTE:  If contacting CEAM by mail, fax, or BBS, include  responses to the  above
information  in your correspondence.

2.9 DISCLAIMER

Mention of trade  names or use of commercial products does  not constitute endorsement
or recommendation for use by the United States Environmental Protection  Agency.

Execution of the 3DFEMWATER/3DLEWASTE model,  and modifications to the DOS
system  configuration files (i.e., /CONFIG.SYS and /AUTOEXEC.BAT) must be made at
the user's own  risk. Neither the U.S.  EPA nor the program  authors can assume
responsibility for  model and/or program modification, content, output, interpretation, or
usage.

CEAM software products are built using FORTRAN-77, assembler, and operating
system  interface command languages. The code structure  and logic  of these products is
designed for single-user, single-tasking, non-LAN environment and operating platform
for microcomputer installations (i.e., single user on a dedicated system).

Users will be on their own if they attempt to install a CEAM product on a multi-user,
multi-tasking, and/or LAN based system (i.e., Windows, DESQview, any LAN). CEAM

                                         15

-------
cannot provide installation, operation, and/or general user support under any
combination of these configurations.  Instructions and conditions for proper installation
and testing are provided with the product in a READ.ME file. While
multiuser/multitasking/LAN installations could work, none of the CEAM products have
been thoroughly tested under all possible conditions. CEAM can provide scientific
and/or application support for selected products if the user proves that a given product
is installed and working correctly.

2.10  TRADEMARKS

       •   IBM, Personal Computer/XT (PC/XT), Personal  Computer/AT (PC/AT), PC
          DOS, VDISK, and  Personal System/2  (PS/2)  are registered trademarks of
          International Business Machines Corporation.

       •   DESQview is a trademark of Quarterdeck Office Systems, Inc.

       •   Sun and SunOS are registered trademarks of Sun Microsystems, Inc.

       •   SPARC is  a registered trademark of SPARC International, Inc.

       •   UNIX is a registered trademark of American Telephone and Telegraph.

       •   SVS FORTRAN-77 is a trademark of Silicon Valley Software.

       •   PRIME and PRIMOS are trademarks  of Prime Computers, Inc.

       •   Microsoft,  RAMDRIVE, HIMEM, MS,  and MS-DOS are registered trademarks
          of Microsoft Corporation.

       •   Windows is a trademark of Microsoft Corporation.

       •   RM/FORTRAN is a trademark of Language Processors, Inc.

       •   DEC, VAX, VMS, and DCL are trademarks of Digital Equipment Corporation.

       •   386 and 486 are trademarks of Intel Corporation.

       •   U.S. Robotics is a registered trademark and Courier HST is  a trademark of
          U.S. Robotics, Inc.
                                         16

-------
                                    SECTION 3

                          BACKGROUND INFORMATION
3.1 3DFEMWATER

3DFEMWATER is designed to simulate the movement of moisture through variably-
saturated porous media. Typical applications include: 1) studying the influence of
transient stresses, such as well production schemes or the onset of drought conditions,
on water table elevations, and 2)  generating flow fields for use in examining the
influence of physical processes such as rainfall and evapotranspiration on  the movement
of dissolved contaminants through the vadose zone and into aquifers (Figure 3.1). The
complementary 3DLEWASTE model is designed  to utilize  the flow data generated by
3DFEMWATER simulations in order to evaluate the associated movement of dissolved
contaminants  through the  modeled system. The  model 3DLEWASTE is described in
Section 3.3.

3.1.1  Governing Equations

The governing equation for flow of water through a variably-saturated porous medium,
as derived  from mass and momentum conservation constraints, can be written:


                          F(h) ^L = V |K(h) 
-------
Figure 3.1 UnconfLned aquifer to be approximated as a variably-saturated porous medium.

-------
                                     K(h) = krKs                                (3-2)

where kris the relative permeability, ranging in value from 0.0 to 1.0, and Ksis the
saturated hydraulic conductivity (L/T). The saturated hydraulic conductivity is a flow
property of both the porous medium and fluid, and is determined by tests performed
under saturated conditions. It represents a maximum possible value of effective
hydraulic conductivity. The relative permeability  term describes the influence of water
content on the magnitude of the effective hydraulic conductivity. Values of relative
permeability range from a minimum value reflecting the reduction of effective
conductivity at residual water content to a maximum of 1.0 reflecting saturated
conditions.

The change in relative permeability is caused by changes in moisture content, resulting
in the preferential movement of water through  certain pathways, due to the influence of
capillary forces. As the soil becomes less saturated, water drains more readily from the
large radius pore structures, the water flow becomes restricted  to pore sequences of
smaller radii (Figure 3.2) as well as those held  in layers close to the soil particles. The
result of water becoming  increasingly restricted to the smaller radius pathways is a
reduction in the spatially-averaged effective hydraulic conductivity.

The decrease in effective  hydraulic conductivity, as reflected in  the relative permeability
term, is described by pairs of empirical soil-moisture curves,  These curves detail the
relationships between water content and pressure head, and  between hydraulic
conductivity and water content. Soil-moisture curves are often described as coefficients
and exponents of standard analytical functions  (Brooks and Corey, 1966; Mualem, 1976;
van Genuchten, 1980).  The 3DFEMWATER code allows the user to define the curves
using the van Genuchten functions (1980) or as sets of paired values of relative
permeability versus  moisture content and moisture content versus pressure head given
in lookup table format. The van Genuchten relationships found  in 3DFEMWATER are:
                               kr = er U - (i

and
                                [1 + (h.
                                          19

-------
   Figure 3.2.   Variable pore spacing in soil under saturated flow conditions.
where
                                           (b - 9
                                           T   "
                                                                                 (3-3c)
                                      Y = 1 - 1/p
(3-3d)
and
       9W     = moisture  content (dimensionless)
       <}>      = porosity (dimensionless)
       9wr     = residual moisture content (dimensionless)
       P, y  =  soil-specific exponents  (dimensionless)
       Ot      = soil-specific coefficient (1/L)
       ha     = air entry pressure head (L)
       9e     = effective moisture content (dimensionless)

Note that the soil-moisture content is defined as the porosity multiplied by the degree of
saturation. Typical soil-moisture  curves generated from Equations 3-3a and 3-3b are
presented in Figures 3.3a and 3.3b.
                                          20

-------
                                  NEGATIVE PRESSURE HEAD (cm)
                  1.0


                  10'

                  10"

                  i

                  10

                  10'

                  10

                  10'

                  10
                                                                I I
                    101
                                            10-'
                                    MOISTURE CONTENT
                                                                    1.0
Figure 3.3.    Logarithmic plot of constitutive relations for sand, clay, silty
              loam, and sandy loam: (a) moisture content vs. pressure head,  and
              (b) relative permeability vs. moisture content (based on data
              presented in Carsel and Parrish,  1988).
                                           21

-------
The water capacity term or storage term used in 3DFEMWATER can be written in the
form:

                                    F(h) = deydh                               (3-4)
 It should be noted that, due to the relatively small influence of compressibility on water
 capacity in the unsaturated zone (with respect to the drainage potential), soil and water
 compressibility have been ignored in the storage term. When analytical functions are
 used to describe the nonlinearity of the relative  conductivity, the derivative with respect
 to pressure head of the water content versus pressure head function must also be
 defined analytically.

 The equation governing saturated flow represents a limiting case of Richard's equation
 where the relative permeability is a constant of  1.0 and the water capacity is a constant
 equal to the specific yield for an unconfined aquifer or specific storage for a confined
 aquifer.

 3.1.2 Boundary  Conditions and Transient Source/Sink Terms

 Unique solutions to  variably-saturated flow problems are generated  by solving Richard's
 equation in conjunction with 1)  a set of boundary conditions defined at the physical
 edges of the modeled system and 2) where appropriate, source/sink terms applied within
 the system (Figure 3.4). Boundary conditions available in the 3DFEMWATER model
 include fixed-head (Dirichlet) boundaries, specified-flux  (Cauchy) boundaries, specified
 pressure-head gradient (Neumann) boundaries and variable (head-dependent flow)
 boundaries.

 Fixed-head or Dirichlet boundaries  are boundaries defined by prescribing pressure heads
 at specified boundary nodes so that:


                                h = hd(xb)yb,zb)t) on Bd                           (3-5)

 where
       hd         = specified pressure head (L)
       Bd         = portion of the system boundary subject to a Dirichlet
                       boundary condition
       XbiYb^b     = spatial coordinates on the boundary (L)

 Dirichlet boundaries are typically used to define  the perimeters of bodies of water, the
water table location, and leaking surface impoundments or other waste disposal
 facilities containing specified levels of water.  Specified pressure heads may be constant
 or allowed to vary with time reflecting physical processes such as water  level
 fluctuations associated with seasonal changes in  rainfall and evapotranspiration rates.

 The specified-flux  (Cauchy) boundary represents  the portions of the system boundary
where infiltration  or evapotranspiration rates can be quantified. The specified-flux
boundary condition can be written:
                                         22

-------
             CONCEPTUAL
AGRICULTURAL
AREA
                                                          WELL
             MATHEMATICAL
Figure 3.4.   Conceptual model and mathematical approximation for a
             variably-saturated  flow system. Within the modeled
             system, transient source/sink terms may be applied as
             point sources/sinks or as distributed sources/sinks.
                                    23

-------
                        -n-krK8 
-------
1

-,

ARID
J)
2
1












































I











rlHl

V,
s»













1

i
J)
i

































J)
^ Groun<

































JUUJ
5urtac<
r
                              4c  s
            Water Table     y
Figure 3.5   Use  of a  pressure-head gradient boundary  condition  to simulate a
            portion of the unsaturated zone.
                                    25

-------
                                h = hp(xb,yb,zb,t) on Bv                         (3-8a)

or

                         -n-krKB 
-------
WELL
• f. /-"^N f A
r
1 4
) I
ft »
. 4
P ~
»(
*
ft |
. t
9 '
»^
i
»f
»4

»4
1 1
t
l
^

(
>
^

i
i
i

i
k
T
t
r
t
f

f
t

1
1
I
I
Is
/

t
4
4
4
4

<
• 4
j
1
f
L- ELEMENT
THICKNESS
t;
i



>
r '
> '
> '

^ "
^ ^


> '
h 	 1


1

LEGEND

• FINITE ELEMENT
NODE
(•) POINT SOURCE/
SINK NODE


	 1

'^^Screened
r-^Interval
	
Figure 3.6. Using a series of nodes to represent a screened well interval.
                                  27

-------
vertically adjacent nodes are used to represent the screened interval of a well, the
volumetric flux must be distributed among the nodes. The most appropriate
distribution of the total flux is in proportion to the effective conductance, Ce, of the
individual nodes where the effective conductance of each node is defined as:
Ce =
                                                  (Ka)nLn
where n-1 and n are indices referring to the element below the node and the element
above the node, respectively, and 0.5L is half the thickness of an element.

Time-variant boundary conditions and source/sink flux or flux  intensity rates are defined
by a series of paired time and value points. This paired  data is used to assemble a look-
up table from which appropriate values are obtained using linear interpolation at
specified times of analysis.  Constant values can be  specified by assigning the same
value to a set of two time/data point pairs, making sure  that the simulation time is fully
spanned.

3.1.3  Initial  Conditions

The solution of Richard's equation also requires the initialization of pressure head
values such that:
                                 h = hj(x,y,z,t=0) in R
                                                  (3-10)
where hj is the initial pressure head distribution (L), and R is the region of interest
(Figure 3.7).  Besides providing a frame of reference for transient analyses, the initial
conditions are used to set the nonlinear parameters at the beginning of a simulation.
For transient problems, an  appropriate set of initial pressure head values may either be
input directly or derived from a steady-state simulation. For more information on these
options see Section 4.1.11.
                                              CONTINUOUS
                        INITIAL CONDITION
                                                   P1ECEWISB
                                       TIME
       Figure 3.7. Pressure head versus time at a nodal point on the finite
                  element grid.
                                          28

-------
3.1.4 Steady-State

When analyzing the influence of transient stresses, such as well production schemes and
drought conditions, on the flow system, a starting point must be assumed. The user
defines boundary conditions and flow parameters as best he/she can, then does an initial
simulation to allow the system to reach an equilibrium or steady-state (Figure 3.8). The
steady-state simulation then defines the pressure head at all points in the system and it
is from this initial condition that a transient simulation is started. Although the actual
system is never really in steady-state, by using averaged conditions (i.e., rainfall, etc.) a
reasonable starting point is generated. If the steady-state simulation fails to converge
or the results poorly match field data, flow parameters and/or boundary conditions
should be adjusted to improve the starting conditions. The steady-state or equilibrium
condition  is generated  by removing  the temporal  term from Equation 3-1. The system  is
then defined as the equilibrium reached under the average conditions.

Besides being used for initial conditions for a transient simulation, the steady-state flow
option can also be used in conjunction  with a transient transport simulation. Since the
flow system will generally reach equilibrium under non-changing stresses faster than an
associated solute transport problem, using a steady-state flow field and average
conditions to define the advective portion of solute transport will often give a good
approximation of the change of solute distribution over time. The savings in
computational effort can be considerable and, given the uncertainty of parameters in the
system, an acceptable approximation may be reached.
                  STEADY-STATE
                  SOLUTION
                     INITIAL CONDITION
                      0
                                   TIME
      Figure 3.8    Pressure head versus time at a nodal point where a steady-state
                   solution is being approached.
                                         29

-------
 3.2 NUMERICAL APPROXIMATION IN 3DFEMWATER

 The 3DFEMWATER model was developed to solve the variably-saturated flow equation
 described in Section 3.1. In the model, Richard's equation (Equation 3-1)  is
 approximated using the Galerkin finite  element technique. The time integral term in
 Equation 3-1 is approximated using backwards or central (Crank-Nicholson)  difference
 in time.  The nonlinearity of the system  is treated using Picard iteration and the
 generated set of linearized equations is solved using a block iterative method.

 3.2.1  Galerkin  Formulation

 In 3DFEMWATER,  Richard's equation is approximated using the Galerkin finite
 element  method (Finder and Gray, 1977) where the dependent variable, pressure head,
 is approximated by a trial function of the form:
                             h =  N/x^h/t)  j  = l,2,-.,n
(3-11)
where Nj(Xj,t) are the three-dimensional shape functions and hj(t) are nodal values of
pressure head at time t for the n nodes of which the finite element grid is comprised
(Figure 3.9).
70
66
60
35
50
45
40
35
30
15
£
20
3
300
502
400
'"y4! S / If / / f S 7 .
f. f Iff f f f f f .
\f / • f f / f t f f f f /
«A^ f f ' / f f J
S f 7 / / 7 /


Y






/








t
f







1
/'













































J J J J J











































>










4
1
/









t

-*









1


t
*-






5



L

j^ TT *xll
c~
' S 7 / S S / S S
/ f f f / S S
f f f f f f
S /
















































Y








/ ,


A\,


t

y
7f
t
^
• t '
/
/
' '/
            0 *• TO  120  100 200 275 350 400 450 500540 570«00«50  750 800150 900 950 1000
Figure 3.9.   Finite element grid for production from a single well in a variably-
             saturated porous medium.
                                         30

-------
 Substituting the trial functions into Equation 3-1 and applying the Galerkin criterion,
 we generate a set of weighted residual minimization equations:
                      Jw,
:F(h)_f± - V -[krKB
-------
                              Bu = £ JvNi'krK,-VNjadR                      (3-15C)


 and

            C, = £ - fkrK8-VNie-VzdR +  fN;eqdR +  f^'n^K/h^N/ + VzflB    (3-15d)
                 t. i *                 ~         *^                     I
                   L R-                R-        B-                     J

 where m is the number of elements into which the system is discretized and Ne denotes
 elemental shape functions.

 3.2.2 Solution Techniques

 To solve the series of linearized ordinary differential equations presented in Equation 3-
 15a, the time differential is replaced by a finite difference formulation, resulting in
 working equations for 3DFEMWATER of the form:
                  Atk
                           - h,k + wB^V*1 + (l-w)B^V = C;kiw           (3-16)
where k+1 represents the current time level, k the previous time level, At the length of
the current  time step and w the time weighting function (1.0=backwards in time;
0.5= Crank  Nicholson or centered in time). Note that the associated transport code,
3DLEWASTE, utilizes a backwards-in-time scheme: Therefore, when  using
3DFEMWATER to generate a flow  field for a 3DLEWASTE simulation, the backwards-
in-time option should be used. This prevents the possibility of a mismatch in the
interpolation of  time-variant boundary condition and source/sink flux values.

For each time step, the solution method involves an outer and inner iterative scheme
(Figure 3.10) where the outer iterations control convergence of the nonlinear terms in
the equations and the inner iterative scheme controls  the block-iterative  method of
solving the linearized set of equations.  For each nonlinear iteration, the  linearized set
of equations is solved using relative permeability and  storage terms updated using
pressure head values generated during the previous nonlinear (outer) iteration.  Relative
permeability and storage terms for  the first iteration in a time step are based on
pressure head values from the previous time step, or for the first time step, from the
initial conditions.

Because of the strong nonlinear nature of the soil moisture curves, the outer iterative
scheme may become unstable. To help circumvent this problem it is often helpful to
damp the iterative changes in the pressure head. One method of damping the iterative
changes is through the  use of an under-relaxation factor. Implementation of the under-
relaxation factor for the outer iterations in 3DFEMWATER is as follows:

where u is the outer under-relaxation factor and r is the iteration  number. If damping
is needed, values below one should  be used. Acceleration or over-relaxation ( 1.0
-------
            I
      Time Step Loop
           I
 Assemble stiffness matrix
      and load vector
  Solve linearized set of
   equations one block
         at a time
              Update load vector
              using new pressure
                  heads
Update nonlinear
   parameters
  Check for convergence of
  block iterative scheme
              NO
                                                           t
               YES
Check for convergence of
nonlinear terms if needed
           I
             NO
YES
 Is this the last timestep?
                            NO
               YES
     End analysis
       Figure 3.10. Solution scheme for unsaturated flow analysis.
                                33

-------
                                               + uhr1                         (3-17)

is generally not recommended for the nonlinear iterations as it may make the solution
become unstable. For transient simulations, reduction of the time-step size can also
help increase the stability of the solution scheme. Note that sometimes steady-state
problems will be difficult to solve. In this case, it is often worth trying a transient
solution approach, using expanding  time steps to approach the steady-state solution.

For each nonlinear iteration,  a set of linearized simultaneous equations is solved using a
block iterative scheme.  The user defines a set of subregions (or blocks) by  prescribing
the nodes contained in  each subregion  (Figure 3.11). The  code  then generates a series of
connectivity arrays indicating: 1) the nodes contained in each subregion,  2) for each
node, all other nodes found in elements it is part of, and 3) which of these  adjacent
nodes are located in the same subregion. The nodal equations for each subregion are
solved directly using a Gaussian solver. For each nodal equation defined in Equation 3-
15a, contributions from adjacent nodes falling outside the subregion being solved for are
generated by multiplying the matrix terms with the appropriate nodal pressure heads.
These pressure heads are generated during the last direct solution  for the subregion
containing the adjacent nodes.

Subregions are generally defined as nodal planes (Figure 3.11)  allowing the user to work
with a minimal half-bandwidth when the direct solver is  invoked. The half-bandwidth
is defined as one plus the largest difference  between the node number associated with
the nodal equation and the other nodes found in elements the node is part  of and which
are in the same block as the  node. As a general rule, subregions  comprised of vertical or
sub-vertical nodal slices provide the smallest half-bandwidth and will perform well in
the block iterative method, although this may not always be the  case. For  some
problems, horizontal slicing may be advantageous. The block iterative logic contains a
relaxation factor which can be used  to over-relax the solution and help accelerate the
rate of convergence.  Implementation of the inner over-relaxation scheme  is as follows:


                                               + ohr1                          (3-18)

where s denotes the inner iteration  number  and o is the over-relaxation  factor. The
optimal value of the over-relaxation factor usually falls between 1.5 and 1.9. A good
starting point is o =  1.72.

3.3 3DLEWASTE

3DLEWASTE is designed to simulate the movement of dissolved species  through a
variably-saturated porous medium.  Typical  applications for 3DLEWASTE include the
examination of:  1)  leachate migration from landfills  and surface impoundments, 2) the
influence on water quality of pesticide and fertilizer applications,  and 3) the
environmental impact of leaky containment structures such as  underground and above
ground storage tanks (Figure 3.12).  Velocity fields needed to define the advective
pathways of water bearing the chemicals are provided by  associated 3DFEMWATER
simulations.
                                         34

-------
                                              HORIZONTAL SLICE
CO
01
                                 •   •   '    •   ,,,,,*
                     70  120  160 200  275  350  400 450  500540 570600650   750  800850  900 950 1000
                        Figure 3.11.  Use of vertical or horizontal nodal slices in the block interative method.

-------
CO
CD
        Figure 3.12. Migration of dissolved contaminants through the unsaturated zone into an unconfined aquifer system.

-------
3.3.1  Governing Equations

The governing equation for advective-dispersive solute transport through variably-
saturated porous media, based on the laws of conservation of mass and flux, can be
written in the form:
                 9.1 + pb.  =V
-------
                 MEGASCOPIC SCALE       SOLUTE MIGRATION
                                     PATHWAY
Figure 3.13.     Diagram showing the effect of scale on hydrodynamic dispersion
                 processes.
                                      38

-------
molecular diffusion coefficient in Equation 3-20 quantifies the spreading due to
molecular diffusion.

In order to solve Equation 3-19 for a single dependent variable, the constitutive relation-
ship between the species concentrations in the dissolved and adsorbed phases must be
defined.  The 3DLEWASTE code allows the user to choose from three relationships: 1)
alinear isotherm, 2) Freundlich isotherm, or  3) Langmuir isotherm. The isotherms, as
determined in laboratory partitioning experiments, can be plotted in log-log form to
derive:


                               log S = n log C + log K                        (3-21)

or

                                       S = KG n                                (3-22)

where n is slope of the plot of log S versus log C and K is the S-axis intercept (Freeze
and Cherry,  1979). Equation 3-22 defines the Freundlich isotherm, which is often used
to describe the partitioning between the dissolved and absorbed phases. When the
isotherm has a slope n=  1, the isotherm is linear and the relationship can be defined as:
                                       -    = Kd                                (3-23)
                                       dC

where K,, is called the distribution coefficient (LVM). Linear isotherms are often used to
describe the adsorption of hydrophobic organic compounds to organic matter in soils.
The distribution coefficient is described as  a function of the organic carbon content  of
the soil as:


                                      Kd  = US*                               (3-24)

where f^ is the fractional organic carbon content and K,,,, is the normalized distribution
coefficient. There are many published lists of values for K^, (e.g., Lyman et al., 1982;
U.S. EPA, 1986; Verschueren, 1983). Data are available primarily  for pesticides and, to
a lesser degree, aromatic and polycyclic aromatic compounds. If data on K^ are not
available  for a particular chemical, a value can be estimated from empirical
relationships between K^. and some other property of the chemical  such as the water
volubility, S, the octanol-water partition coefficient, KDW, or the bioconcentration factor
for aquatic life, BCF. Lyman  et al. (1982) tabulate  12  such regression equations
obtained from data sets of different classes of chemicals.  One commonly-used
relationship (Karickhoff et al., 1979) takes  the form:


                                    Koc = 0.41KOW                              (3-25)


The Langmuir isotherm takes the form:
                                          39

-------
                                     S =  Sm"KC                               (3-26)
                                          1 + KG
where S_ is the maximum concentration allowed in the medium.
       •'max
The effective decay constant, A,, is a degradation constant that can be used to quantify
the effects of radioactive decay, or the composite effects of hydrolysis and
biodegradation. When used to quantify the effects of hydrolysis and biodegradation, the
effective decay coefficient (for a linear isotherm) takes the form:
                                A =   l      2"b + JL                          (3-27)
                                      9 + KdPb

where ^ is the first-order hydrolysis rate constant for the dissolved species, Aj is the
first-order hydrolysis rate constant for the sorbed species, and \ is the first-order
biodegradation rate constant. The dissolved species first-order hydrolysis rate can be
written in terms of the acid-catalyzed (Kg), base-catalyzed (K,,), and neutral (K,,)
hydrolysis rate constants as:


                            A! = KJH =  KJH+] + Kn+ KJOH]                      (3-28)

where [H+] is the hydrogen ion concentration and [OH-] is the hydroxyl ion
concentration. The sorbed phase first-order hydrolysis rate is considered to be a
function of the acid and neutral hydrolysis rates and is usually written in  the form:

                                 A2 = ccKa[H + ] + K,,                           (3-29)

where a is the acid-catalyst hydrolysis rate enhancement factor for the sorbed phase
with a typical value of 10.0. Note that for a nonlinear isotherm the formulation in
3DLEWASTE is valid only if \ = A,.

The governing equation for advective-dispersive solute transport in a porous medium, as
presented in Equation 3-19, describes the transport from an  Eulerian or fixed
framework. The numerical algorithm may begin to oscillate and fail to converge to a
solution of this equation when the advective term  starts to dominate over  the dispersive
term and  the equation takes on a hyperbolic nature.  Dominance of the advective term
over the dispersive term is reflected in the non-dimensional Peclet number, which is
defined as the ratio of the product of the velocity magnitude  and distance advected to
the dispersion coefficient.  In finite element analysis the critical Peclet number is the
local Peclet number of an element, where the local Peclet number is defined as:

                                      P = L/O                                 (3-30)
where L is the element length.

One method of circumventing the numerical problems  (i.e., oscillation and failure to
converge) associated with Peclet numbers greater than 2 is to address the system
through  a moving (i.e., Lagrangian) coordinate system. In the Lagrangian formulation

                                          40

-------
for solute transport in a porous medium, the temporal term is defined as a material
derivative of the form:
                      fe + Pb *i VDC = e  9C +    9s + v.vc
                            b dcDF      at    "at
where D denotes the material derivative.

The advective term, V-VC, written in index notation becomes:
                                   V-VC -     .                                (3-32)
                                           dt 9x.

where the repeated indices indicate summation. Substituting Equation 3-31 into
Equation 3-19, the governing equation for a Lagrangian framework becomes:
for a linear isotherm. The average linear velocity, V*, for a linear isotherm becomes:
                                V =    e + pb                              (3-33b)


For a non-linear isotherm, the Lagrangian equation becomes:
                        pb--   = V < 0D -VC) - W6C + PbS) + QCin - QC          (3-34a)
                   Dt    dC at

where

                                     V* = V/0                              (3-34b)

Full implementation of the Lagrangian approach implies the solution of Equation 3-33
using a moving coordinate system. Another method of circumventing the instability
problem is to utilize a hybrid Eulerian-Lagrangian approach. Such an approach is
implemented in 3DLEWASTE. In the hybrid approach, the  advective term of the
material derivative is evaluated in a Lagrangian manner by a backwards particle
tracking scheme (Figure  3.14). The particle tracking scheme generates a particle
starting location and an associated concentration, C*. This concentration, C*, is the
starting concentration  of each particle which reaches a nodal point at the end of that
particular time step. The  material term of Equation 3-33 is then approximated by:
                                         41

-------
                                 o
nodal point
initial coordinate of
particle

flowline
computed path of
particle
       Figure 3.14.     Backward particle tracking to determine the starting
                       point of an advected particle.
                                    DC
                                    Dt
  C - C*
     At
(3-35)
The diffusion-type equation is then solved using a fixed coordinate system. Note that
for a steady-state simulation, where At—>°°, the logic is implemented by multiplying the
transient storage terms by zero and evaluating the advection term in a fixed coordinate
system.

3.3.2 Boundary Conditions  and Transient Source/Sink Terms

Unique solutions to advective-dispersive solute transport problems are generated by
solving the governing equation (Equation 3-19) in conjunction with 1) a set of boundary
conditions, defined at the physical edges of the modeled system, and where appropriate,
2) source/sink terms applied within the system (see Figure 3.4).

Boundary conditions and source/sink terms available in the 3DLEWASTE model include:
                                          42

-------
       •  Prescribed-concentration (Dirichlet) boundaries

       •  Specified-flux (Cauchy)  boundaries

       •  Specified-dispersive-flux (Neumann) boundaries

       •  Variable  boundaries

       •  Point sources

       •  Distributed sources

Prescribed-concentration or Dirichlet boundaries are defined by prescribing dissolved
species concentrations at specified boundary nodes as:


                              C  = Cd (xb,yb)zb,t)   on Bd                        (3-36)

where Cd is the specified  solute concentration, Bd is the portion of the system boundary
subject to a Dirichlet boundary condition, and (xb,yb,zb) is the spatial  coordinate on the
boundary. Dirichlet boundaries are typically used to  test computer programs by
allowing comparisons with analytical solutions.  Unlike the analogous constant-head
boundaries of flow  models, constant-concentration boundaries are generally poor
approximations of contaminant source  terms.  Bodies of fresh water located upgradient
from  contaminant sources can be  approximated using constant concentration nodes.
When used to define sources, specified concentrations may be constant or allowed to
vary with time, reflecting physical processes such as degradation of the source due to
radioactive decay, hydrolysis, biodegradation, or physical removal. Concentration versus
time profiles can be defined to account for seasonal or other time-variant changes in
dissolved species levels.

The specified-flux (Cauchy) boundary represents the portions of the system boundary
where infiltration can be quantified. The specified-flux  boundary has many representa-
tions  including: 1) infiltration due to leachate migration  from a landfill or surface
impoundment, 2) application of pesticides or fertilizer to fields, and 3) the dilution
effects of rainfall or irrigation on previously applied  constituents. The specified-flux
boundary condition can be written:


                       n
-------
 of fresh water is simulated by applying the specified-flux boundary condition and setting
 the mass flux rate to zero. The automatically generated term accounting for water flow
 normal to the boundary will simulate the dilution due to infiltration.

 Also available in 3DLEWASTE is a specified-dispersive-flux or Neumann boundary
 condition of the form:


                           n< -9D -VC) = qn (xb>yb>zb, t)  on Bn                    (3-38)

 where q,, (M/T/L2) is the portion of the boundary flux attributable to the concentration
 and Bn is the portion of the system boundary subject to a specified-dispersive-flux
 boundary condition. Note  that exit boundaries  can be declared using this boundary
 condition and letting q^O. This physically simulates mass being advected out of the
 system.

 For solute transport, the variable composite boundary condition represents a combined
 specified-flux/dispersive-flux boundary which allows for time-variant itilltratiorjwater-
 loss rates. The boundary condition during infiltration is:


                   n 0                   (3-39b)

 and mass is advected  out  of the system. Like the  specified-flux boundary condition, the
 variable boundary can represent:  1) infiltration due to leachate migration from a landfill
 or surface impoundment, 2) application of pesticides or fertilizer to fields, and 3) the
 dilution effects of rainfall  or irrigation on previously applied constituents. When the
 boundary being modeled may be either an exit  or an infiltration boundary, such as a
 precipitation/ evapotranspiration boundary or a seepage face, the  variable boundary
 condition is the proper choice. The variable boundary condition can also  be used in a
 manner similar to the dispersive-flux condition to  simulate strictly exit nodes.

Internal  source/sink terms,  as  represented by the  term QCin in Equation  3-19 are also
 accounted for in  3DLEWASTE. As with  the boundary conditions, the source/sink terms
 can be constant or allowed to vary with time.  Both the fluid flux  rate, Q, and the
 injected fluid species concentration, C^, are allowed to  vary with  time. Two source/sink
 options are available in the code. The first is a point source/sink option and the second,
 a distributed source/sink option. The first option  is generally used to represent
 production or injection wells.  The fluid fluxes in wells are represented  as volumetric
 water fluxes, q^ (LVT), applied at a nodal point or  to better represent a screened
 interval, a column of nodal points  (see Figure 3.6). If vertically adjacent nodes are used
 to represent the screened interval of a well, the volumetric flux must be distributed

                                          44

-------
 among the nodes. The most appropriate method of doing this is discussed in Section
 3.1.2. Note that the applied fluid fluxes must match those used in the associated flow
 simulation.

 The distributed source option is a source intensity that is integrated over the volume of
 an element. For a distributed source element, the user defines a fluid source intensity,
 02 (LVT/L3), or fluid flux rate per unit volume for each distributed source element. This
 option allows a user modelling a large area to approximate the influence of a well field
 within an element.

 Time-variant  boundary conditions  and source/sink flux or flux intensity rates are defined
 by a series of paired time and value points. The paired data are used to assemble a
 look-up table  from  which appropriate values are obtained using linear interpolation  at
 specified times of analysis.  Constant values can be specified by assigning the same
 value to a set of two time/data point pairs, making sure that the simulation time is fully
 spanned.

 3.3.3 Initial  Conditions

 The solution  of the governing equation for solute transport in a porous medium also
 requires the initialization of concentration values such that:


                               C = C1(x,y,z,t= 0) in  R                          (3-40)

 where C; is the initial concentration distribution and R is the region of interest. The
 initial conditions are used to define the starting water quality and soil concentration
 levels for determining the fate of the dissolved  constituents.  Besides providing a frame
 of reference for transient analyses, the initial conditions are used to set the storage
 parameters for Freundlich and Langmuir isotherms at the beginning of nonlinear
 simulations. For transient problems, an appropriate set of initial concentration values
 may either be input directly or derived from  a  steady-state simulation. For more
 information on these options see Section 4.2.10.

 3.3.4  Steady-State

 When looking for a bounding solution to determine the maximum possible concentration
 levels that may be  reached in a solute transport problem,  a steady-state option may be
 employed. In the steady-state case, the time derivatives in Equation 3-19 are discarded
 and the equation, including the advective term, is solved  in an Eulerian or fixed-
 coordinate framework. Note that any solute source prescribed as a boundary condition
 or source term becomes modeled  as an infinite  source. For many systems this upper
bound may be highly conservative. The steady-state option is of no use if the source is
 solely defined by initial conditions.

 3.4 NUMERICAL APPROXIMATION IN  3DLEWASTE

The 3DLEWASTE model  was developed  to simulate advective-dispersive solute
transport in variably-saturated porous media. In the model, the hybrid Eulerian-
 Lagrangian governing equation (Equation 3-33)  is approximated using the Galerkin

                                         45

-------
finite element technique. The time integral term in Equation 3-33 is approximated
using backwards differencing in time. The nonlinearity of the system is treated using
Picard iteration and the generated set of linearized equations is solved using a block
iterative method.

3.4.1  Galerkin Formulation

In 3DLEWASTE, the diffusion equation is approximated using the Galerkin finite
element method where the dependent variable, concentration, is approximated by a trial
function of the form:


                             C = NtJCt)   = 1,2,. ..,n                      (3-41)
where Nj(Xj,t) are the three-dimensional shape functions and Cj(t) are nodal values of
concentration at time t for the n nodes of which the finite element grid is comprised  (see
Figure 3.9). Substituting the trial  functions into Equation 3-33 and applying the
Galerkin criterion, we generate a set of weighted residual minimization equations of the
form:
                          Dt
                              -VOD-VC) -X(6 +PbKd)C +QCin - QR = 0       (3-42)
for the linear isotherm case,  where W,are the weighting functions. For the Galerkin
method, the weighting functions are the same as the shape functions and, therefore,
Equation 3-42 can be written in the form:


                     PbKd):B£- VJ /)t          J            J      3    J
           R.                    R. Ot          R.           R.


where BB is the entire region boundary. The integrals given in  Equation 3-44, which are
taken over the entire region being modeled, can  be replaced by the summation  of
integrals taken over the volumes  and surfaces of individual elements of the finite
element grid. This finite element approximation generates a set of n nodal equations of
the form:

                                         46

-------
                                                     j = l,2,...n
where
                                = E  f(
                                                                             (3-45a)
(3-45b)
B1J =  f) JvN.'-eD-VN,"
                                                     dR
(3-45c)
                                                                             (3-45d)
and
                                    in N(edR + J
(3-45e)
where m is the number of elements into which the system is discretized and Ne are the
elemental shape functions.  Note that for a steady-state simulation, the full Eulerian
approach is used.  The Lagrangian  term DC/Dt is replaced by 3C/<5t and the Eulerian
term:
     E/N.T-
                                           -VNedR
 (3-46)
is added to By.

3.4.2  Solution Techniques

To solve the series of linearized ordinary differential equations represented by Equation
3-45a, the time differential is replaced by a finite difference formulation, resulting in
working equations for 3DLEWASTE of the form:
                                         47

-------
                      At

where k+1 represents the current time step, k the previous time step, and At the length
of the current time step. Note that since the transient solution scheme allows only for a
backwards difference approximation the associated flow runs should also  be solved using
backwards-in-time approximation.

For each time step, the solution method involves  an inner iterative scheme (see Figure
3.10)  which controls the block iterative method of solving the linear equations.  For
simulations where the nonlinear Freundlich or Langmuir isotherms are used, the
solution method also  involves an outer iterative scheme where the iterations control
convergence of the nonlinear terms in  the linearized set of equations.  For each
nonlinear iteration, the linearized set of equations is solved using storage terms updated
using concentration values  from  the  previous  nonlinear (outer) iteration.  Storage terms
for the  first iteration in a time step are based on  concentration values  from the  previous
time step, or for the first time step, from the initial conditions. If the outer iterative
scheme becomes unstable, it may be helpful to damp the iterative changes in the
concentration. One method  of damping the iterative changes is through the use of an
under-relaxation factor. Implementation of the under-relaxation factor for the  outer
iterations in 3DLEWASTE is:


                              c;+1=  (i  - u0)c;+ u0cr                        (3-48)

where u0 is the outer under-relaxation  factor and r is the iteration number.  If damping
is needed, values between 0.5 and 0.9  should  suffice.  Acceleration or over-relaxation
(1.0
-------
                                    SECTION 4

                           DATA INPUT REQUIREMENTS
4.1 3DFEMWATER INPUT  SEQUENCE

This section describes how to construct a data input file for 3DFEMWATER, the
variably-saturated flow code. Background information about the code that will aid in
building an input file, such as construction of a grid or selection of boundary condition
types, is provided in Section  3.1. In addition, help in selecting values for some of the
input parameters is given in  Section 5.1.

Note that maximum control parameters are associated with a number of the input
variables. These control parameters are used in the code to specify array dimensions.
For some  problems, the default values set for these parameters may be too small. If so,
they can be easily changed. The  maximum control parameters and their default values
are listed in Appendix C. Note also that the logical units used by 3DFEMWATER are
defined in Appendix B.

A complete input file consists of information supplied in 18 data sets. The contents and
format of each data set are listed below. When constructing an input sequence, it is
important to note that data sets 2 through 17 must be preceded by a record which
contains a description of the  data set.  This can be seen in the example input sequences
provided in Section 6.1. Most of the input is entered in free-format, which means that
the spacing of the input data in a record does not need to follow a set pattern. Note
that a record can consist of multiple lines, with a line defined as up to 80 columns.

The user may choose to run the model using any set of units as long as they are consis-
tently maintained in all the input. Units of mass (M), length (L), and time (T) are
indicated in the input descriptions.

4.1.1 Data Set 1: Title of the Simulation Run

      One record with FORMAT(I5,A70,2X,I2,2I1) per problem.  This record  contains
      the following variables:

        1.    NPROB = Problem number (columns  1-5).

        2.    TITLE = Array for the title of the problem. It may contain up to 70 charac-
            ters  (columns 6- 75).

        3.    IGEOM = Integer indicating if (1) the geometry, boundary and pointer
            arrays are to be printed and if (2) the boundary and pointer arrays are to
            be computed or  read via logical units (column 78). If  IGEOM is an even
            number, geometry,  boundary and pointer arrays will not be printed. If

                                         49

-------
            IGEOM is an odd number, they will be printed. If IGEOM is less than or
            equal to  1, boundary arrays will be computed and written on logical unit
            LUBAR,  but if IGEOM is greater than 1, boundary arrays will be read via
            logical unit LUBAR. If IGEOM is less than or equal to 3, pointer arrays
            will be computed and written on logical unit LUPAR, but if IGEOM is
            greater than 3, pointer arrays will be read via logical unit LUPAR. In
            summary:

            IGEOM = Even No.   Print the geometry, boundary,  and pointer arrays.
            IGEOM = Odd No.   Do not print the arrays.
            IGEOM <. 1          Compute and write boundary and pointer arrays.
            1 < IGEOM <. 3      Read boundary arrays,  compute and write pointer
                                 arrays (not used under normal conditions).
            IGEOM > 3          Read boundary and pointer arrays.

       4.   IBUG = Integer indicating if diagnostic output is desired to help determine
            problems encountered while executing the code (column 79);
            0 = no,
            1 = yes.

       5.   ICHNG = Integer control number indicating if the  cyclic change of rain-
            fall-seepage nodes is to be printed (column 80);
            0 = no,
            1 = yes.

4.1.2 Data Set 2: Basic Integer Parameters

       One record with FREE-FORMAT per problem.  It contains:

       1.   NNP =  Number of  nodal  points. .

       2.   NEL = Number of elements.

       3.   NMAT = Number of material types.

       4.   NCM = Number of elements with material property correction.

       5.   NTI = Number of time  steps or time increments (see notes at the end of
            Data Set 2).

       6.   KSS = Steady-state control;
            0 = steady-state solution,
            1 = transient-state solution (see note at the end of Data Set 2).

       7.   NMPPM = Number  of material properties per material;  this parameter
            should be set equal to 6 in the present version of the code (see Data Set 5).

       8.   KGRAV =  Gravity term control;
            0 = no gravity term,
            1 = gravity term included.

                                          50

-------
       9.   ILUMP = Mass lumping control;
            0 = no,
            1 = yes.

       10.   IMID = Mid-difference control;
            0 = no,
            1 = yes.

       11. NITER = Number of iterations allowed  for solving the  non-linear equation.

       12. NCYL = Number of cycles permitted  for iterating rainfall-seepage boundary
            conditions per time step.

       13.   NDTCHG = Number of times the time-step size will be reset to the initial
            time-step size;  NDTCHG should be >.  1 (see Section 5.1.2.10).

       14. NPITER =  Number  of iterations for a pointwise solution.

       ****  NOTE: NTI can be computed by NTI =11 + 1  + 12 + 1, where 11 is the
                    largest  integer not exceeding  Log(DELMAX/DELT)/Log(l+CHNG),
                    12 is the largest integer not exceeding (RTIME-DELT*((1+CHNG)
                    **(I1 + 1)-1)/CHNG)/DELMAX, RTIME is the real simulation time,
                    and  DELMAX, DELT.and  CHNG are defined in data set 3.

       ****  NOTE: A steady-state option may be used to provide either the final state
                    of a  system under study or the initial condition for a transient-state
                    calculation. In the former  case, KSS  = 0 and NTI  = 0 in this data
                 •   set.  In the latter case, KSS =  0 and NTI > 0.  If KSS > 0, there
                    will  be no steady-state calculation.

4.1.3 Data Set 3: Basic Real Parameters

       One  record with FREE-FORMAT per problem. It contains:

       1.   BELT =  Initial time step size, (T).

       2.   CHNG = Fractional change in the time-step size in each subsequent time
            increment, (dimensionless  decimal-point value).

       3.   DELMAX = Maximum value of BELT, (T).

       4.   TMAX = Maximum simulation time,  (T).

       5.   TOLA =  Steady-state  convergence  criterion, (L).

       6.   TOLB =  Transient-state convergence criterion, (L).

       7.   RHO = Density of water, (M/L3).

       8.   GRAY = Acceleration of gravity,  (L/T2); (e.g.,  32.17  ft/s2or 9.81 m/s2).

                                          51

-------
        9.   VISC = Dynamic viscosity of water, (M/UT).

       10.   W = Time derivative weighting factor;
            0.5 = Crank-Nicolson central and/or mid-difference,
            1.0 = backward difference.

       11.   OME = Iteration parameter for solving the nonlinear matrix equation,
            0.0 < OME < 1.0 = under-relaxation,
            1.0 = exact relaxation,
            1.0 < OME < 2.0 = over-relaxation.

       12.   OMI = Relaxation parameter for solving the linearized matrix equation
            pointwise;
            0.0 < OMI <  1.0 = under relaxation,
            1.0 = exact relaxation,
            1.0 < OMI <  2.0 = over relaxation.

4.1.4 Data Set 4: Printer and Disk  Storage Control and
       Times for Step Size Resetting

       Three records are needed per problem. The first two records are formatted input
       with FORMAT (211).  The third  record is a FREE-FORMAT input. The number of
       lines for the first two records depends on the value of NTI, the number of time
       increments. The number of lines for the third record depends on  the value of
       NDTCHG, the number of times to reset the time-step size.

       Record 1 - FORMAT(211): This record  contain the following variables:

        1.   KPRO = Printer control for steady-state and initial conditions;
            0 = print nothing,
            1 = print the values for the variables FLOW, FRATE, and TFLOW,
            2 = print values above plus pressure head H,
            3 = print values above plus total head,
            4 = print values above plus moisture content,
            5 = print values above plus Darcy velocity.

        2.   KPR(I) = Printer control for the I-th (1=1,2,.... NTI) time step;
            0 = print nothing,
            1 = print the values for the variables FLOW, FRATE, and TFLOW,
            2 = print values above plus pressure head H,
            3 = print values above plus total head,
            4 = print values above plus moisture content,
            5 = print values above plus Darcy velocity.

       Record 2 - FORMAT(211): This record  can  be used to store 3DFEMWATER
       output  in a binary file for use in plotting or as  input to  3DLEWASTE. It
       contains the following variables:
                                          52

-------
        1.    KDSKO = Auxiliary storage control for the steady-state or initial condition;

             0 = no storage,
             1 = store on logical unit LUSTO.

        2.    KDSK(I) = Auxiliary storage control for the I-th (1=1,2,.... NTI)  time step;

             0 = no storage,
             1 = store on logical unit LUSTO.

       Record 3- FREE-FORMAT: This record contains the following variables:

        1.    TDTCH(I) = Time when the I-th  (1 =  1,2,	NDTCHG) time-step-size
             resetting is needed.

4.1.5  Data Set 5: Material Properties

       Either hydraulic conductivity or permeability can be input in this data  set.  The
       flag KCP in data set 6A is used to indicate which of the two is being  used. A
       total of NMAT records are needed per problem, one for each material.

       Record I (1 = 1, 2, .... NMAT) - FREE-FORMAT: Each record  contains following
       variables:

        1.    PROP (1,1) = Saturated xx-hydraulic  conductivity or permeability of the
             medium I, (L/T or L2).

        2.    PROP(2,1) = Saturated yy-hydraulic  conductivity or permeability of the
             medium I, (L/T or L2).

        3.    PROP(3,1) = Saturated zz-hydraulic  conductivity or permeability of the
             medium I, (L/T or L2).

        4.    PROP(4,1) = Saturated xy-hydraulic  conductivity or permeability of the
             medium I, (L/T or L2).

        5.    PROP(5,1) = Saturated xz-hydraulic  conductivity or permeability of the
             medium I, (L/T or L2).

        6.    PROP(6,I) = Saturated yz-hydraulic  conductivity or permeability of the
             medium I, (L/T or L2).

4.1.6  Data Set 6:  Soil Property Parameters

       6A. Soil  Property Control  Integers

       One record per problem. This record is  FREE-FORMATTED and contains the
       following variables:
                                          53

-------
      1.   KSP = Soil property input control;
          0 = analytical input,
          1 = tabular data input.

      2.   NSPPM = Number of points in the tabular soil property functions
          when KSP = 1.  The number of parameters needed to specify the
          analytical soil functions per material when KSP = 0. (For analytical
          soil functions, NSPPM = 5 in the current version of the code.)

      3.   KCP = Permeability input control;
          0 = input saturated hydraulic conductivity,
          1 = input saturated permeability.

6B.  Analytical Soil Parameters

This subdata set is needed if and only if KSP = 0. NMAT records are needed,
one for each material type.

Record I (I = 1, 2,	NMAT)  - FREE-FORMAT:  Each record contains the
following variables:

      1.   THPROP(1,I) =  Residual moisture (water) content for material I, (—).

      2.   THPROP(2,I) =  Saturated  moisture (water) content for material I, (—).

      3.   THPROP(3,I) = Air entry pressure head for material I, (L).

      4.   THPROP(4,I) = Van Genuchten empirical  coefficient alpha for material
          I, (1/L).

      5. THPROP(5,I)  = Van Genuchten empirical coefficient beta for material
          I, (")-

6C.  Soil Properties  in  Tabular Form

This subdata set is needed if and only if KSP = 1. Four sets of records are
needed — one each for pressure, water-content, relative conductivity (or relative
permeability), and water capacity, respectively. Each set contains NMAT  records,
one for each material type. Thus the total number of records for this subdata set
is 4*NMAT. The number of lines in each record is determined by the input
parameter NSPPM, defined in data set 6A.

Record I (I = 1, 2,	NMAT)  - FREE-FORMAT:  Each record contains the
following variables:

      1.   HPROP(1,I) = Tabular value of pressure head for the first data point of
          material I, (L).
                                   54

-------
      2.   HPROP(2,I) = Tabular value of pressure head for the second data point
          of material I, (L).
 NSPPM.  HPROP(NSPPMJ) = Tabular value of pressure head for the NSPPM-th
          data point of material I, (L).

Record (NMAT + I) (I =  1, 2,	NMAT) - FREE-FORMAT:  Each record contains
the following variables:

      1.   THPROP(1,I) = Tabular value of moisture-content for the first data
          point in material I, (—).

      2.   THPROP(2,I) = Tabular value of moisture-content for the second data
          point in material 1, (—).
 NSPPM.  THPROP(NSPPMJ) = Tabular value of moisture-content for the
          NSPPM-th data point in material 1, (-).
Record (2*NMAT + I) (1 = 1, 2,	NMAT)  - FREE-FORMAT: Each record
contains the following variables:

      1.   AKPROP(IJ) = Tabular value of relative conductivity for the first data
          point in material I, (--).

      2.   AKPROP(2,I) = Tabular value of relative conductivity for second data
          point in material I, (—).
 NSPPM. AKPROP(NSPPMJ) = Tabular value of relative conductivity for the
          NSPPM-th data point in material I, (-).
Record (3*NMAT + I) (1 = 1, 2,	NMAT)  - FREE-FORMAT: Each record
contains the following variables:

      1.   CAPROP(IJ) = Tabular value of moisture-content capacity for the first
          data point in material  I,  (1/L).

      2.   CAPROP(2,I) = Tabular value of moisture-content capacity for the
          second data point in material I,  (1/L).
                                   55

-------
       NSPPM. CAPROP(NSPPMJ) = Tabular value of moisture content capacity for
                the NSPPM-th data point in material I, (1/L).

4.1.7 Data Set 7: Nodal Point  Coordinates

       Coordinates for NNP nodes, specified in data set 2, are needed. Usually a total of
       NNP records are required. However, if a group of subsequent nodes appears in
       regular pattern, an automatic generation input option can be used.

       Each record is FREE-FORMATTED and contains the following variables:

       1.    NI = Node number of the first node in the sequence.

       2.    NSEQ =  NSEQ subsequent nodes will be automatically generated.

       3.    NAD = Increment  of node number  for each of the NSEQ subsequent nodes.
       4.    XNI = X-coordinate of node NI, (L).

       5.    YNI = Y-coordinate of node NI, (L).

       6.    ZNI = Z-coordinate of node NI, (L).

       7.    XAD = Increment of x-coordinate for each of the NSEQ subsequent nodes,
            (L).

       8.    YAD = Increment of y-coordinate for each of the NSEQ subsequent nodes,
            (L).

       9.    ZAD = Increment of z-coordinate for  each of the NSEQ subsequent nodes,
            (L).

       **** NOTE: A record with nine zeroes must be used to signal  the end of this
                   data set.

4.1.8 Data Set 8: Subregional  Data

       8A. Subregion  Control  Integer

       One FREE-FORMATTED record is needed for this subdata set. It contains the
       following variable:

            1.   NREGN = Number of subregions.

       8B. Number of Nodes in  Each Subregion

       Normally, NREGN records are required. However, if the sequence of node num-
       bers follows a regular pattern between sequential subregions, the automatic
       generation input option can be used.
                                         56

-------
       Each record is FREE-FORMATTED and contains the following five variables:

            1.   NK = Subregion number of the first subregion in a sequence.

            2.   NSEQ = Number of subsequent subregions which will be  automatically
                generated.

            3.   NKAD = Increment of NK in each of the NSEQ subsequent subregions.

            4.  NODES = Number of nodes in the subregion NK.

            5.  NOAD  = Increment of NODES in each  of the NSEQ subsequent subre-
                gions.

                NO IE:   A record with five zeroes must be used  to end the input of
                          this subdata set.

       8C.  Mapping between Global Nodes and  Subregion Nodes

       This subdata set should be repeated NREGN times, once for each subregion.  For
       each subregion, normally, the number of records  equals the number of nodal
       points in the subregion.  Automatic generation  can be used, however, if the
       sequence of subregional  node numbers follows a regular pattern.

       Each record is FREE-FORMATTED and contains the following five variables:

            1.   LI = Local node number of the first node in a sequence.

            2.   NSEQ = Number of subsequent local nodes which will be generated
                automatically.

            3.   LIAD = Increment of LI for each of the NSEQ subsequent nodes.

            4.   NI = Global node number of local node LI.

            5.   NIAD = Increment of NI for each of the NSEQ subsequent nodes.

                NO IE.   A record with five zeroes must be used  to signal the end of
                          this subdata set.

                NO IE:   Local node numbers have values between one and the total
                          number of nodes in a subregion (i.e., 1,2,.. ., NODES). Global
                          node numbers are associated with the entire grid and are
                          entered using data set 7.

4.1.9  Data Set  9:   Element Incidence

       Element incidence for NEL elements, specified in data set 2, are needed.
       Usually, a total of NEL  records are needed. However,  if a sequence  of element
       numbers  follows a regular pattern, the automatic generation input option can be used.

                                         57

-------
Each record is FREE-FORMATTED and contains the following variables:

 1.   MI = Global element number of the first element in a sequence.

 2.   NSEQ = Number of subsequent elements which will be automatically
     generated.

 3.   MIAD = Increment of MI for each of the NSEQ subsequent elements.

 4.   IE(MI,1) = Global node number of the first node of element MI.

 5.   IE(MI,2) = Global node number of the second node of element MI.

 6.   IE (MI, 3) = Global node number of the third node of element MI.

 7.   IE(MI,4) = Global node number of the fourth node of element MI.

 8.   IE(MI,5) = Global node number of the fifth node of element MI.

 9.   IE (MI, 6) = Global node number of the sixth node of element MI.

10.   IE (MI, 7) = Global node number of the seventh node of element MI.

11.   IE(MI,8) = Global node number of the eighth node of element MI.

12.   IEMAD = Increment of IE(MI,1) through IE(MI,8) for each of the NSEQ
     elements.

**** NOTE: IE(MI,1)  - IE(MI,8)  are numbered according to the convention
             shown in Figure 4.1. The first four nodes start from the front,
             lower, left corner and progress around the bottom element surface
             in a counterclockwise direction. The other four nodes begin from
             the front, upper, left corner and progress around the top element
             surface in a counterclockwise direction.
                    z
                    i,
        Figure 4.1. Node numbering convention for the elements.
                                  58

-------
4.1.10 Data Set 10:  Material Type Correction

      This data set  is required only if NCM, defined in data set 2, is greater than zero.
      Normally, NCM records are required. However, if a group of element numbers
      follow a regular pattern, the automatic generation input option may be used.

      Each record is FREE-FORMATTED and contains the following variables:

        1.   MI =  Global element number of the first element in the sequence.

        2.   NSEQ = Number of subsequent elements which will be  generated automati-
            cally.

        3.   MAD  =  Increment of element number for each of the  NSEQ subsequent
            elements.

        4.   MITYP  = Type of material for element MI.

        5.   MTYPAD = Increment of MITYP for each of the NSEQ  subsequent
            elements.

      **** NOTE: A record with five zeroes must be used to signal the end of this
                    data set.

4.1.11 Data Set 11:  Card Input for Initial  or Pre-Initial Conditions

      NNP records  (i.e., one record for each node)  are normally needed. However, if a
      sequence of node numbers follows a regular pattern, automatic generation can be
      used.

      Each record is FREE-FORMATTED and contains the following variables:

        1.   NI =  Global node number of the first node in the sequence.

        2.   NSEQ = Number of subsequent nodes which  will be generated
            automatically.

        3.   NAD  =  Increment of node number for each of the NSEQ nodes.

        4.   HNI = Initial or pre-initial pressure head of node NI,  (L).

        5.   HAD  =  Increment of initial or pre-initial head for each of the NSEQ nodes,
            (L).

        6.   HRD  =  Geometrical increment  of HNI  for each of the NSEQ subsequent
            nodes; (i.e, HNI**HRD).

      **** NOTE: A record with six zeroes must be used to signal  the end of this data
                    set.
                                         59

-------
       **** NOTE: The initial condition for a transient calculation may be obtained in
                    two different ways:   1) it can be read directly from data set 11, or
                    2) the code  can perform a steady-state simulation using
                    time-invariant boundary conditions before automatically beginning
                    the transient  computations. For the first case, both KSS and NTI
                    in data set  2 should be greater than zero. In  the latter case, KSS =
                    0 and NTI > 0 and data set 1 1 is used to input the pre-initial
                    condition, which is required as the starting condition for the
                    steady-state iteration. In order to obtain a steady-state solution,
                    both KSS and NTI are set equal to zero  and data set 1 1 supplies
                    the starting condition for the steady-state solution.

4.1.12  Data Set 12: Integer Parameters  for Source and Boundary Conditions

       One record per problem is needed. This record is FREE-FORMATTED and
       contains the following variables:

        1.   NSEL = Number of distributed source/sink elements.

        2.   NSPR = Number of distributed source/sink profiles  (i.e., time histories).

        3.   NSDP = Number of data points in each of the NSPR source/sink profiles.

        4.   KSAI = Option for the distributed source/sink profiles to be input
            analytically. This variable should be set equal to zero in the current
            version of the code.

        5.   NWNP = Number of well or point source/sink nodes.

        6.   NWPR = Number of well or point source/sink profiles (i.e.,  time histories).

        7.   NWDP = Number of data points in each of the NWPR profiles.

        8.   KWAI = Option for  the  well source/sink profiles  to be input analytically.
            This variable should be set equal to zero in the current version of the code.

        9.   NDNP = Number of fixed-head (Dirichlet)  nodes  (NDNP should be > 1).

       10.   NDPR = Number of fixed-head profiles (i.e., time histories) (NDPR should
            be > 1).

       11.   NDDP = Number of data points in each freed-head profile (NDDP should be
       12.   KDAI = Option for the freed-head boundary value profiles to be input
            analytically. This variable should be set equal to zero in the current
            version of the code.

       13.   NVES = Number of variable composite (rainfall/evaporation-seepage)
            boundary element sides.

                                          60

-------
       14.    NVNP = Number of variable composite boundary nodal points.

       15.    NRPR = Number of variable composite profiles (i.e., time histories).

       16.    NRDP = Number of data points in each of the NRPR profiles.

       17.    KRAI = Option for  the variable composite profiles to be input analytically.
             This variable should be set equal  to zero  in the current version of the code.

       18.    NCES = Number of specified-flux (Cauchy) boundary element sides.

       19.    NCNP = Number of specified-flux nodal points.

       20.    NCPR = Number of specfled-flux  profiles (i.e., time histories).

       21.    NCDP = Number of data points in each of the NCPR profiles.

       22.    KCAI = Option for  the  specified-flux profiles to be input analytically. This
             variable should be set equal to zero in the current version of the code.

       23.    NNES = Number of specified-pressure-head gradient (Neumann) boundary
             element sides.

       24.    NNNP = Number of specified-pressure-head gradient nodal points.

       25.    NNPR = Number of specified-pressure-head gradient flux profiles (i.e., time
             histories).

       26.    NNDP = Number of data points in each of the NNPR profiles.

       27.    KNAI = Option for  the specified-pressure-head gradient profiles to be input
             analytically. This variable should be set equal to zero in the current
             version of the code.

4.1.13  Data  Set  13:  Distributed  and Point Sources/Sinks

       This data set is used to supply data for both distributed sources/sinks and well
       (point) sources/sinks.

       13A.  Distributed Sources/Sinks

       The following three subdata sets are needed if  and only if NSEL in data set 12 is
       greater than zero. The first subdata set is used to specify the distributed
       source/sink profiles. The second subdata set is  used to  read the global element
       numbers of the distributed source/sink  elements. The third subdata set is used
       to assign a source/sink profile to each distributed source/sink element.

       (a)     Source/Sink Profiles
                                          61

-------
There will be NSPR (see data set 12) records in this subdata set. The number of
lines in each record depends on the value of NSDP, defined in data set 12.

Record I (1 = 1, 2,	NSPR) - FREE-FORMAT:  Each record  contains the follow-
ing variables:

      1.   TSOSF(U) = Time of the first data point in the I-th profile, (T).

      2.   SOSF(1,I) = Source/sink value (as  flux rate per unit volume of
           element) of the first  data point in  the I-th profile, (LVT/LVL);  positive
           for a source and negative  for a sink.

      3.   TSOSF(2,I) = Time of the second data point in  the I-th profile, (T).

      4.   SOSF(2,I) = Source/sink value of the second data point in the  I-th
           profile,  (LVT/LVL); positive for a source and negative for a sink.
           Up to NSDP data points.

(b)     Global Element Number of All Distributed Source/Sink Elements

One record is needed for this subdata set. The number of lines in this record
depends on NSEL, defined in data set 12. The record is FREE-FORMATTED
and contains the following variables:

       1.   MSEL(l) = Global element number of the first distributed source/sink
           element.

       2.   MSEL(2) = Global element number of the second distributed
           source/sink element.
           Up to NSEL numbers.

(c)     Source/Sink Profile Type Assigned to Each Element

Usually NSEL records are needed.  However, automatic generation can be used.
Each record is FREE-FORMATTED and contains the following variables:

       1.   MI = Compressed element number of the first element in the
           sequence.

       2.   NSEQ =  Number of elements which  will be generated  automatically.

       3.   MAD = Increment of element number for each of the NSEQ subse-
           quent elements.
                                  62

-------
       4.   MITYP = Source/sink profile type associated with element ML

       5.  MTYPAD = Increment of MITYP for each of the NSEQ subsequent
           elements.

       **** NOTE: A record with five zeroes  must be used to signal the end of
                    this subdata set.

       **** NOTE: Compressed element numbers have values between one and
                    the total number of distributed source/sink elements. Com-
                    pressed element one corresponds to the first element listed
                    in  13A(b), compressed element two corresponds to the second
                    global element, etc.

 13B. Point (Well) Source/Sink

The following three subdata sets are needed if and only if NWNP in data set 12
is greater than zero. The first subdata set is used to specify the point source/sink
profiles. The  second subdata set is  used to read the global node numbers of the
point source/sink nodes. The third subdata set is used to assign a source/sink
profile to each point source/sink node.

(a)     Source/Sink Profiles

There will be NWPR (see data set  12) records in this subdata set. The number of
lines in each  record depends on NWDP, defined in data set  12.

Record  I (I = 1, 2,	NWPR) - FREE-FORMAT: Each record contains the
following variables:

       1.   TWSSF(U)  = Time of the first data point in the I-th profile, (T).

       2.   WSSF(1,I) = Source/sink flow rate of the first data point in the I-th
           profile, (LVT); positive for a source and negative for a sink.

       3.   TWSSF(2,I)  = Time of the second data point in the I-th profile, (T).

       4.   WSSF(2,I) = Source/sink flow rate of the second data point in the I-th
           profile, (LVT); positive for  a  source and negative for a sink.
           Up to NWDP data points.

(b)     Global Node Number of All Point (Well) Source/Sink Nodes

One record is needed for this subdata set. The number of lines in this record
depends on NWNP, defined in data  set 12. The record is FREE-FORMATTED
and contains the following variables:
                                  63

-------
              1.   NPW(l) = Global node number of the first point source/sink node.

             2.   NPW(2)  = Global node number of the second point source/sink node.



                  Up to NWNP numbers.

       (c)     Source/Sink Profile Type for Each Node

       Usually NWNP records are needed. However, automatic generation can be used.
       Each record is FREE-FORMATTED and contains the following variables:

              1.   NI = Compressed point source/sink node number of the first node in
                  the sequence.

             2.   NSEQ = Number of subsequent nodes which will be generated auto-
                  matically.

             3.   NAD = Increment of NI for each of the NSEQ subsequent nodes.

             4.   NITYP = Source/sink profile type associated with  node NI.

             5.   NTYPAD = Increment of NITYP for each of the NSEQ subsequent
                  nodes.

                  NOTE: A record with five zeroes must be used to signal the end of
                          this subdata set.

4.1.14  Data Set  14: Variable Composite  (Rainfall/Evaporation-Seepage) Boundary
        Condition

        The following six subdata sets are required if and only if NVES in data set 12 is
        greater than zero. The  first subdata set is used to specify the rainfall/evap-
        oration profiles. The second subdata set is used to assign the type of rain-
        fall/evaporation profile to each of the variable composite boundary element sides.
        The third subdata set is used to specify the variable composite boundary
        element sides.  The fourth subdata set is used to read the global nodal numbers
        of all the variable composite boundary nodes. The fifth subdata set  is used to
        read the pending depth  for each of the nodes. The sixth subdata set is used to
        read the allowed minimum pressure for each of the  nodes.

        14A. Rainfall/Evaporation-Seepage Profiles

        There  will be NRPR records (see data set  12) in this subdata set. The number
        of lines in each record depends on NRDP, defined in data set 12.

        Record I (I = 1, 2,	NRPR) - FREE-FORMAT: Each record contains  the  follow-
        ing variables:
                                         64

-------
     1.   TRF(l.I) = Time of the first data point in the I-th profile, (T).

     2.   RF(1,I)  = Rainfall/evaporation rate of the first data point in the I-th
         profile,  (L/T).

     3. TRF(2,I) = Time of the second data point in the I-th profile, (T).

     4.   RF(2,I)  = Rainfall/evaporation rate of the second  data point in the I-th
         profile,  (L/T).
         Up to NRDP data points.

 14B. Rainfall/Evaporation-Seepage Profile Type Assigned to Each Boundary
      Element Side

At most, NVES  (see data set  12) records are needed. However, automatic
generation can be used.

Record I (I = 1, 2,	) - FREE-FORMAT: Each record  contains the following
variables:

     1.   MI = Compressed variable boundary element  side number of the first
         element side in a sequence.

     2.   NSEQ  = Number of subsequent variable boundary element sides which
         will be generated automatically.

     3.   MIAD  = Increment of MI for each of the NSEQ subsequent variable
         boundary element sides.

     4.   MITYP = Type of rainfall/evaporation-seepage profile assigned to side
         MI.

     5.   MTYPAD = Increment of MITYP for each of the NSEQ subsequent
         sides.

         NO IE:   A record with five zeroes must be used  to signal the end of
                   this subdata set.

 14C. Specification of Variable Composite Boundary  Element Sides

Normally, NVES records are required, one each for a variable  boundary element
side. However, if a sequence of variable composite boundary element side
numbers follows a regular pattern, automatic generation may be used.

Record I (I = 1, 2,  ....)- FREE-FORMAT: Each record  contains the following
variables:
                                  65

-------
     1.   MI = Compressed variable boundary element side number of the first
         element side in a sequence.

     2.   NSEQ = Number of subsequent variable boundary element sides which
         will be generated automatically.

     3.   MIAD = Increment of MI for each of the NSEQ subsequent variable
         boundary element sides.

     4.   II = Global node number of the first node of element side ML

     5.   12 = Global node number of the second node of element side MI.

     6.   13 = Global node number of the third node of element side MI.

     7.   14 = Global node number of the fourth node of element side MI.

     8.   HAD = Increment of II for each of the NSEQ subsequent variable
         boundary element sides.

     9.   I2AD = Increment of 12 for each of the NSEQ subsequent variable
         boundary element sides.

   10.    13AD = Increment of 13 for each of the NSEQ subsequent variable
         boundary element sides.

   11.    WAD = Increment of 14 for each of the NSEQ subsequent variable
         boundary element sides.

        NO IE.    A record with 11 zeroes must be used to signal the end of
                  this subdata set.

14D. Global Node Number of All Variable Composite Boundary Nodes

At most, NVNP records  (see data set 12) are needed for this subdata set, one for
each variable boundary node.

Record I (I = 1, 2, . . . . ) - FREE-FORMAT: Each record contains  five variables:

     1.   NI = Compressed variable boundary node number of the  first node in
         the sequence.

     2.   NSEQ = Number of subsequent nodes which will be generated
         automatically.

     3.   NIAD = Increment of NI for each of the NSEQ nodes.

     4.   NODE = Global node number of node NI.

     5.   NODEAD = Increment of NODE for each of the NSEQ nodes.

                                 66

-------
          NO IE:   A record with five zeroes must be used to signal the end of
                   this subdata set.

 14E. Pending Depth Allowed for Each Variable Composite Boundary Node

 Normally, NVNP  records (see data set 12) are needed.  However, if a sequence of
 node numbers follows a regular pattern of pending depth,  automatic generation
 is used.

 Record I (I = 1, 2, ....)- FREE-FORMAT: Each record  contains the following
 variables:

      1.   NI = Compressed variable boundary node number of the first node in a
          sequence.

      2.   NSEQ = Number of subsequent nodes which will be generated
          automatically.

      3.   NIAD = Increment of NI for each of the NSEQ subsequent nodes.

      4.   HCONNI = Ponding depth of node NI, (L).

      5.   HCONAD = Increment of HCONNI for each of the NSEQ nodes,  (L).

   **** NOTE: A  record with five zeroes must be used  to signal the end of this
                subdata set.

14F. Minimum Pressure Head Allowed for Each Variable Composite Boundary
      Node

Normally, NVNP records are needed. However, if a sequence of node numbers
follows a regular pattern of minimum pressure head, automatic generation  is
used.

Record I (I = 1, 2, ....)- FREE-FORMAT: Each record contains the following
variables:

      1.   NI = Compressed variable boundary node number of the first node in a
          sequence.

      2.   NSEQ = Number of subsequent nodes which will be generated
          automatically.

      3.   NIAD = Increment of NI for each of the NSEQ subsequent nodes.

      4.   HMINNI = Minimum pressure head allowed for node NI, (L).
                                  67

-------
            5.    HMINAD = Increment of HMINNI for each of the NSEQ nodes, (L).

            **** NOTE: A record with five zeroes must be used to signal the end of
                         this subdata set.

4.1.15  Data Set  15: Fixed-Head  (Dirichlet) Boundary Condition

       This data  set is required only if NDNP in data set  12 is greater than zero. It
       consists  of three subdata sets. The first subdata set is used to specify the fixed-
       head profiles. The second subdata set is used to read the global node numbers of
       the freed-head boundary nodes. The third subdata set is used to assign a head
       profile to each Dirichlet boundary node.

       ISA. Fixed-Head Profiles

       There will be NDPR (see data set 12) records in this subdata set.  The number of
       lines in each record depends on NDDP, the number of data points  in each profile.

       Record I (I  = 1, 2,	NDPR)  - FREE-FORMAT: Each record contains the
       following variables:

            1.  THDBF(IJ) =  Time  of the  first data point in the I-th profile,  (T),

            2.    HDBF(1,I) = Total  head of the first data point in the I-th profile, (L).

            3.    THDBF(2,I) = Time of the second data point in the I-th  profile, (T).

            4.    HDBF(2,I) = Total  head of the second data point in the  I-th profile, (L).
                Up to NDDP data points.

       15B. Global Node Number of All the Dirichlet Nodes

      One FREE-FORMATTED record is needed for this subdata set. The number of
      lines in this record depends on NDNP, defined in data set 12.

            1.   NPDB(l) = Global node number of the first compressed Dirichlet node.

            2.   NPDB(2) = Global node number of the second compressed Dirichlet
                node.
                Up to NDNP numbers.

       15C. Type of Head Profile Assigned to Each Fixed-Head Node


                                         68

-------
       Normally one record per Dirichlet node (i.e., a total of NDNP records) is needed.
       However, if the Dirichlet node numbers follow a regular pattern, automatic
       generation may be used.

       Record I (1 = 1, 2,	)  - FREE-FORMAT: Each record contains the following
       variables:

            1.  NI = Compressed Dirichlet node number of the first  node in the se-
                quence.

            2.  NSEQ = Number of subsequent Dirichlet nodes which will be
                generated automatically.

            3.  NIAD = Increment of NI for each of the NSEQ nodes.

            4.  NITYP  = Type of total head profile assigned to node NI and NSEQ
                subsequent nodes.

            5.  NTYPAD = Increment of NITYP for each of the NSEQ subsequent
                nodes.

            **** NOTE: A record with five zeroes must be used to signal the end of
                         this subdata set.

4.1.16  Data Set 16: Specified-Flux (Cauchy) Boundary Condition

       This data set is required only if NCES in data set 12 is greater than zero. Four
       subdata sets are required. The  first subdata set is used to read the specified-flux
       profiles. The second subdata set is used to read the type of specified-flux profile
       assigned to each of the specified-flux boundary element  sides.  The third subdata
       set is used to read the specified-flux boundary element sides.  The fourth subdata
       set is used to read the global nodes associated with the specified-flux boundaries.

       16A. Specified-Flux  Profiles

       There will be NCPR records (see data set 12)  in this  subdata set. The number of
       lines in each record depends on NCDP, defined in data set 12.

       Record I (I = 1, 2, .... NCPR) - FREE-FORMAT: Each record contains the follow-
       ing variables:

            1.  TQCBF(1,I) = Time of the  first data point in the I-th profile,  (T).

            2.  QCBF(1,I) = Normal specified-flux of the first  data point in the I-th
                profile,  (LVT/L2); positive out from the region, negative into the region.
            3.  TQCBF(2,I) = Time of the second data point in the I-th profile, (T).
                                         69

-------
      4.   QCBF(2,I) = Normal specified-flux of the second data point in the I-th
          profile,  (LVT/L2); positive out from the region, negative into the region.
          Up to NCDP data points.

 16B. Type of Specified-Flux Profile Assigned to Each Boundary Element Side

At most, NCES records (see data set 12) are needed. However, automatic genera-
tion can be used.

Record I (I = 1, 2, . . . .  ) - FREE-FORMAT: Each record contains the following
variables:

      1.   MI = Compressed specified-flux boundary element side number of the
          first side in the sequence.

      2.   NSEQ  = Number of sides which will be generated automatically.

      3.   MIAD  = Increment of MI for each of the NSEQ sides.

      4.   MITYP = Type of specified-flux profile assigned to side MI.

      5.   MTYPAD  = Increment of MITYP for each of the NSEQ sides.

      **** NOTE: A record  with five zeroes must be used to signal the end of
                  this subdata set.

 16C. Specified-Flux Boundary Element Sides

Normally, NCES  records are required, one for each specified-flux boundary
element side. However, if a group of specified-flux boundary element side
numbers follows a regular pattern, automatic generation can be used.

Record I (I = 1, 2, ....)- FREE-FORMAT: Each record contains the following
variables:

      1.   MI = Compressed specified-flux boundary element side number of the
          first element side in a sequence.

      2.   NSEQ  = Number of subsequent specified-flux element sides which will
          be generated automatically.

      3.   MIAD  = Increment of MI for each of the NSEQ subsequent sides.

      4.   II = Global node number of the first node of element side MI.

      5.   12 = Global node number of the second node of element side MI.

                                   70

-------
            6.   I3=Global node number of the third node of element side MI.

            7.   14 = Global node number of the fourth node of element side MI.

            8.   HAD = Increment of II for each of the NSEQ subsequent element
                sides.

            9.   I2AD = Increment of 12 for each of the NSEQ subsequent element
                sides.

          10.  I3AD = Increment of 13 for each of the NSEQ subsequent element sides.

          11.  I4AD = Increment of 14 for each of the NSEQ subsequent element sides.

            **** NOTE: A record with 11 zeroes must be used to end this subdata set.

       16D. Global Node Number of All Compressed Specified-Flux Nodes

       One FREE-FORMATTED record is needed for this subdata set. The number of
       lines in this record depends on NCNP, defined  in data set 12.

            1.   NPCB(l) = Global node number of the first compressed specified-flux
                node.

            2.   NPCB(2) = Global node number of the second compressed specified-flux
                node.
                Up to NCNP numbers.

4.1.17 Data Set 17:  Specified-Pressure-Head Gradient (Neumann)
      Boundary Condition

      This data set is required if and only if NNNP in data set  12 is greater than zero.
      It consists of four subdata sets. The first subdata set is used to specify the speci-
      fied-pressure-head gradient flux profiles. The second subdata set is used to
      assign a profile to each boundary element side. The third subdata set is used to
      read the global element sides of the specified-pressure-head gradient  boundary
      elements. The fourth subdata set is used to read the global node numbers
      associated with the specified-pressure-head gradient boundaries.

       17A. Prescribed Pressure-Head Gradient Flux Profiles

      There will be NNPR  records (see data set 12) in  this subdata set. The number of
      lines in each record depends on NNDP, defined in data set 12.

      Record I (I = 1, 2,	NNPR) - FREE-FORMAT: Each  record contains the
      following variables:
                                         71

-------
      1.    TQNBF(IJ) = Time of the first data point in the I-th profile, (T).

      2.    QNBF(1,I) = Normal specified-pressure-head gradient flux of the  first
           data point in the I-th profile, (LVT/L2); positive out from the region,
           negative into the region.

      3.    TQNBF(2,I) = Time of the second data point in the I-th profile, (T).

      4.    QNBF(2,I) = Normal specified-pressure-head gradient flux of the
           second data point in the I-th profile, (LVT/L2); positive out from the
           region, negative into the region.
          Up to NNDP data points.

 17B. Type of Specified-Pressure-Head Gradient Flux Profile Assigned to Each
      Boundary Element Side

At most, NNES records are needed (see data set  12). However, automatic
generation can be used.

Record I (I = 1, 2, . . . .  ) - FREE-FORMAT:  Each record contains the following
variables:

      1.   MI = Compressed specified-pressure-head gradient element side
          number of the first side in the sequence.

      2.   NSEQ = Number of subsequent sides which will be generated
          automatically.

      3.   MIAD = Increment of MI for each of the NSEQ sides.

      4.   MITYP = Type of specified-pressure-head gradient flux profile assigned
          to side MI.

      5.   MTYPAD = Increment of MITYP for each of the NSEQ sides.

      **** NOTE: A record with five zeroes must be used to signal the  end of
                  this subdata set.

17C. Specified-Pressure-Head Gradient  Boundary Element Sides

Normally, NNES records are required, one for each specified-pressure-head
gradient boundary element side. However, if a group of specified-pressure-head
gradient boundary element side numbers follow a regular pattern, automatic
generation may be used.

Record I (I =  1, 2, ....)- FREE-FORMAT: Each record contains the following
variables:

                                   72

-------
             1.    MI = Compressed specified-pressure-head gradient boundary element
                 side number of the frost side in sequence.

            2.    NSEQ = Number of subsequent sides which will be generated
                 automatically.

            3.    MIAD = Increment of MI for each of the NSEQ subsequent sides.

            4.    II = Global node number of the  first node of element side MI.

            5,0    12 = Global node number of the  second node of element side MI.

            6.    13 = Global node number of the  third node of element side MI.

            7.    14 = Global node number of the  fourth node of element side MI.

            8.    HAD = Increment of II for each of the NSEQ subsequent element
                 sides.

            9.    12AD = Increment of 12 for each of the NSEQ subsequent element
                 sides.

          10.   I3AD = Increment of 13 for each of the NSEQ subsequent element sides.

          11.   WAD = Increment of 14 for each of the NSEQ subsequent element sides.

            **** NOTE: A record with 11  zeroes must be used to end this subdata set.

       17D. Global Node Number of All  Compressed Specified-Pressure-Head Gradient
            Nodes

       One FREE-FORMATTED record is needed for this subdata set. The number of
       lines in this record depends on NNNP, defined in data set 12.

            1.    NPNB(l) =  Global node number of the first compressed specified-
                 pressure-head gradient  node.

            2.    NPNB(2) =  Global node number of the second compressed specified-
                 pressure-head gradient  node.
                Up to NNNP numbers.

4.1.18 Data Set 18: End of Job

      If another problem is to be run, then input begins again with input data set 1. If
      termination of the job is desired, a blank line must be inserted at the end of the
      data set.
                                         73

-------
4.2 3DLEWASTE INPUT  SEQUENCE

This section describes how to construct a data input file for 3DLEWASTE, the transport
code.  Background information about the code that will aid in building an input file,
such as the types of adsorption isotherms allowed, is provided in Section 3.3. In
addition, help in selecting values for some of the input parameters is given in Section
5.2.

Note that maximum control parameters are  associated with a number of the input
variables. These control parameters are used in the code to specify array dimensions.
For some problems, the default values set for these parameters may be  too small. If so,
they can be easily changed. The maximum control parameters and their default values
are listed in Appendix C. Note also that the logical units used by 3DLEWASTE are
defined in Appendix B.

A complete input file consists of information supplied  in 18 data sets. The contents and
format of each data set are listed  below. When constructing an input sequence, it is
important to note that data sets 2 through 17 must be preceded by a record which
contains a description of the data  set.  This can be seen in the example input sequences
provided in Section 6.2. Most of the input is entered in free-format, which means that
the spacing of the input data in a record does not need to follow a set pattern. Note
that a record can consist of multiple lines, with a line  defined as up to 80 columns.

The user may choose to run the model using any set of units as long as  they are consis-
tently maintained in all the input. Units of mass (M),  length  (L),  and time (T) are
indicated in the input descriptions.

4.2.1  Data Set 1: Title of the Simulation Run

       One record with FORNLAT(I5,A70,3X,2I1) per problem. This record contains the
       following variables:

        1.   NPROB = Problem number (columns 1-5).

        2.   TITLE = Array for the title of the problem. It may contain  up to 70 charac-
            ters (columns  6- 75).

        3.   IGEOM = Integer  indicating if (1) the geometry,  boundary and pointer
            arrays are to be printed and if (2) the boundary and pointer arrays are to
            be computed or read via logical  units (column 79). If IGEOM  is an even
            number, geometry, boundary and pointer arrays will not be printed. If
            IGEOM is an  odd  number, they will be printed. If IGEOM  is  less than or
            equal to 1, boundary  arrays will be computed and written on logical unit
            LUBAR, but if IGEOM is greater than 1,  boundary arrays will be read via
            logical unit LUBAR. If IGEOM  is less than or equal  to 3, pointer arrays
            will be computed and written on logical unit LUPAR, but if IGEOM is
            greater than 3, pointer arrays will be  read via logical unit  LUPAR. In
            summary:
                                         74

-------
            IGEOM = Even No.    Print the geometry, boundary, and pointer arrays.
            IGEOM = Odd No.     Do not print the arrays.
            IGEOM <. 1           Compute and write boundary and pointer arrays.
            1 < IGEOM <. 3       Read boundary arrays, compute and write pointer
                                  arrays.
            IGEOM > 3           Read boundary and pointer arrays.

       4.    IBUG = Integer indicating if diagnostic output is desired (column 80);
            0 = no,
            1 = yes.

4.2.2 Data Set 2: Basic Integer Parameters

       One record with FREE-FORMAT per problem. It contains the following
       variables:

       1.    NNP = Number of nodal points.

       2.    NEL = Number of elements.

       3.    NMAT = Number of material types.

       4.    NCM = Number of elements with material property correction.

       5.    NTI = Number of time steps or time increments (see notes at the end of
            Data Set 2).

       6.    KSS = Steady-state control;
            0 = steady-state solution,
            1 = transient-state solution (see note at the end of Data Set 2).

       7.    NMPPM = Number  of material properties  per material; this  parameter
            should be set equal to 8  in the present version of the code (see Data Set 5).

       8.    KVI = Velocity input control;
            -1 = velocity and moisture content read from data set 17,
             1 = steady-state velocity and moisture content input read from
                FEMWATER binary file,
             2 = transient velocity and moisture content input read from  FEMWATER
                binary file.

       9.    ILUMP = Mass lumping control;
            0 = no,
            1 = yes.

       10.   IWET = Weighting function control;
            0 = Galerkin weighting,
            1 = upstream weighting.
                                          75

-------
       11.   IOPTIM = Optimization control;
            1 = upstream weighting optimization factor is to be computed,
            0 = factor is to be set equal to 1.0.

       12.   NITER = Number of iterations allowed for solving the non-linear equation.

       13.   NDTCHG = Number of times the time-step size will be reset to the initial
            time-step size; NDTCHG should be >. 1  (see Section 5.2.2.9).

       14.   NPITER = Number of iterations for a block or pointwise solution.

       15.   KSORP = Sorption model control;
            1 = linear isotherm,
            2 = Freundlich isotherm,
            3 = Langmuir isotherm.

       **** NOTE: NTI can be computed by NTI = 11 + 1+12 + 1, where II  is the
                    largest integer not exceeding Log(DELMAX/DELT)/Log(l+CHNG),
                    12 is the largest integer not exceeding (RTIME-BELT*((1 + CHUG)**
                    (I1 + 1)-1)/CHNG)/DELMAX, RTIME is the real simulation time, and
                    DELMAX, BELT, and CHNG are defined in data set 3.

       **** NOTE: A steady-state option may be used to provide either  the final state
                    of a system under study or the initial condition for a transient-state
                    calculation.  In the former case, KSS = 0 and NTI = 0 in  this data
                    set. In the latter case, KSS = 0 and NTI > 0. If KSS > 0, there
                    will be no steady-state calculation.

4.2.3 Data Set 3:  Basic Real  Parameters

       One record with FREE-FORMAT per problem. It contains the following
       variables:

        1.   BELT = Initial time step size, (T).

        2.   CHNG = Fractional change in the  time-step size in each subsequent time
            increment, (dimensionless decimal-point value).

        3.   BELMAX = Maximum value of BELT, (T).

        4.   TMAX = Maximum simulation time, (T).

        5.   OME = Iteration parameter for solving the nonlinear matrix equation;
            0.0 < OME <  1.0 = under-relaxation,
            1.0 = exact relaxation,
            1.0 < OME < 2.0 = over-relaxation.

        6.   OMI = Relaxation parameter for solving the linearized matrix equation
            pointwise;
            0.0 < OMI <  1.0 = under relaxation,

                                         76

-------
             1.0 = exact relaxation,
             1.0 < OMI < 2.0 = over relaxation.

       7.   TOLB = Transient-state convergence criterion, (L).

       8.   TOLA = Steady-state convergence criterion, (L).

4.2.4 Data Set 4:  Printer and Disk Storage Control  and Times for
       Step Size Resetting

       Three records are needed per problem. The first two records are formatted input
       with FORMAT(2I1). The third record is  a FREE-FORMAT input. The number of
       lines for the first two records depends on the value of NTI, the number of time
       increments. The number of lines for the  third record depends on the value of
       NDTCHG, the number of times to reset the time-step size.

       Record 1 -  FORMAT(2I1): This record contain the following variables:

       1.   KPRO = Printer control for steady-state and initial conditions;
            0 = print nothing,
             1 = print values for the variables FLOW, FRATE, and TFLOW,
            2 = print values above plus concentration,
            3 = print values above plus material  fluxes.

       2.   KFR(I)  = Printer control for the I-th (1=1,2,	NTI) time step; O = print
            nothing,
             1 = print values for the variables FLOW, FRATE, and TFLOW,
            2 = print values above plus concentration,
            3 = print values above plus material  fluxes.

       Record 2 -  FORMAT(2I1): This record can  be used  to store 3DLEWASTE output
       in a binary file for use in plotting results.  It contains the following variables:

       1.   KDSKO = Auxiliary storage control for the steady-state or initial condition;

            0 = no storage,
             1 = store on logical unit LUSTO.

       2.   KDSK(I) = Auxiliary storage control for the I-th (1=1,2,.... NTI) time step;

            0 = no storage,
             1 = store on logical unit LUSTO.

       Record 3- FREE-FORMAT: This record contains the following variables:

       1.   TDTCH(I) = Time when the I-th (1=1,2,....  NDTCHG) time-step-size
            resetting is needed.
                                         77

-------
4.2.5  Data Set 5: Material Properties

       A total of NMAT records are required for this data set, one for each material.

       Record I (I = 1, 2,	NMAT) - FREE-FORMAT:  Each record contains the
       following variables:

        1.   PROP(1,1) = Distribution coefficient (L3/M) or Freundlich K or Langmuir K
            for medium I, depending on the value of KSORP in data set 2.

        2.   PROP(2,1) = Bulk density for medium I, (M/L3).

        3.   PROP(3,1) = Longitudinal dispersivity for medium I, (L).

        4.   PROP(4,I) = Transverse dispersivity for medium I,  (L).

        5.   PROP(5,I) = Molecular diffusion coefficient for medium  I,  (LVT).

        6.   PROP(6,I) = Tortuosity for medium I, (Dimensionless).

        7.   PROP(7,1) = Decay constant in medium I,  (1/L).

        8.   PROP(8,1) = Freundlich N or  Langmuir SMAX for medium I.

4.2.6  Data Set 6:  Nodal Point Coordinates

       Coordinates for NNP nodes  are needed  only if KVI <. 0, where NNP and KVI are
       defined in data set 2. Usually  a total of NNP records are required. However, if a
       group of subsequent node numbers  follows a regular pattern,  an automatic
       generation input option can be used.

       Each record contains the following  variables and is FREE-FORMATTED.

        1.   NI = Node number of  the first node in the sequence.

        2.   NSEQ = Number of subsequent nodes which will be automatically
            generated.

        3.   NAD  =  Increment of node number for each of the NSEQ subsequent nodes.
        4.   XNI = X-coordinate of node NI, (L).

        5.   YNI = Y-coordinate of node NI, (L).

        6.   ZNI = Z-coordinate of  node NI, (L).

        7.   XAD =  Increment of x-coordinate  for each of the NSEQ  subsequent nodes,
            (L).

        8.   YAD =  Increment of y-coordinate  for each of the NSEQ  subsequent nodes,
            (L).

                                          78

-------
       9.    ZAD = Increment of z-coordinate for each of the NSEQ subsequent nodes,
            (L).

       **** NOTE: A record with nine zeroes must be used to signal the end of this
                    data  set.

4.2.7  Data Set 7: Element Incidence

       Element incidence for  NEL elements, specified in data set 2, are needed if <
       O. Usually, a total of NEL records are needed. However, if a group of element
       numbers follows a regular pattern, the automatic generation input option can be
       used.

       Each record is FREE-FORMATTED and contains the  following variables:

       1.    MI =  Global element number of the first element in a sequence.

       2.    NSEQ = Number of subsequent elements which will be automatically
            generated.

       3.    MIAD = Increment of MI for each of the NSEQ subsequent elements.

       4.    IE(MI,1) = Global node number of the first node of element MI.

       5.    IE(MI,2) = Global node number of the second node of element MI.

       6.    IE(MI,3) = Global node number of the third node of element MI.

       7.    IE(MI,4) = Global node number of the fourth  node of element MI.

       8.    IE(MI,5) = Global node number of the fifth node of element MI.

       9.    IE(MI,6) = Global node number of the sixth node of element MI.

       10.   IE(MI,7) = Global node number of the seventh node of element MI.

       11.   IE(MI,8) = Global node number of the eighth node of element MI.

       12.   IEMAD = Increment of lE(MI.l) through IE(MI,8) for each of the NSEQ
            elements.

       **** NOTE: IE(MI,1) -  IE(MI,8) are numbered according to  the convention
                    shown in Figure 4.2. The first four nodes start from the front,
                    lower, left corner and progress around the bottom element surface
                    in a  counterclockwise direction.  The other four nodes begin from
                    the front, upper, left corner and progress around the top element
                    surface in a counterclockwise direction.
                                         79

-------
                        z
                        i,
               Figure 4.2.  Node numbering convention for the elements.
4.2.8 Data Set 8:  Subregional Data

      This data set is needed only if KVI <. 0, where KVI is defined in data set 2.

      8A. Subregion  Control Integer

      One FREE-FORMATTED record is needed for this subdata set. It contains the
      following variable:

           1. NREGN = Number of subregions.

      8B. Number of Nodes in Each Subregion

      Normally, NREGN records are required. However, if the sequence of node num-
      bers follows a regular pattern between sequential subregions, the automatic
      generation input option can be used.

      Each record is FREE-FORMATTED and contains the following five variables:

           1.  NK = Subregion number of the first subregion in a sequence.

          2.  NSEQ  = Number of subsequent subregions which will be automatically
              generated.

          3.  NKAD = Increment of NK in each of the NSEQ subsequent subregions.

          4.  NODES = Number of nodes in the subregion NK.

          5.  NOAD = Increment of NODES in each of the NSEQ  subsequent subre-
              gions.

                                         80

-------
               NOTE:   A record with five zeroes must be used to end the input of this
                        subdata set.

       8C. Mapping between  Global Nodes and Subregion Nodes

       This subdata set should be repeated NREGN times, once for each subregion. For
       each subregion, normally, the number of records equals the number of nodal
       points in the subregion. Automatic generation can be used, however, if the
       subregional node numbers follow a regular pattern.

       Each record contains five variables and is FREE-FORMATTED.

           1. LI = Local node number of the first node in a sequence.

          2. NSEQ = Number of subsequent local  nodes which will  be generated
              automatically.

          3. LIAD =  Increment of LI for each of the NSEQ subsequent nodes.

          4. NI = Global node number of local  node LI.

          5. NIAD = Increment  of NI for each  of the  NSEQ subsequent nodes.

               NO IE:   A record with five zeroes must be used to signal the end of
                        this subdata set.

               NO IE:   Local node numbers have values between one and the total
                        number of nodes in a subregion (i.e., 1,2,. ... NODES). Global
                        node numbers are associated with the entire grid and  are en-
                        tered using  data set 6.

4.2.9 Data Set 9: Material Type  Correction

       This data set is required only if NCM > 0 and KVI <. 0, where  NCM and KVI are
       defined in data set 2.  Normally, NCM records are required. However, if  a group
       of element numbers follows a regular pattern, automatic generation may  be used.


       Each record is  FREE-FORMATTED and contains the following variables:

        1. MI = Global element number of the first element in the sequence.

        2. NSEQ = Number of subsequent elements  which will be generated
          automatically.

        3. MAD = Increment  of element number for  each of the NSEQ subsequent
          elements.

        4. MITYP = Type of material  for element MI.
                                         81

-------
        5. MTYPAD = Increment of MITYP for each of the NSEQ subsequent elements.

       **** NOTE: A record with five zeroes must be used to signal the end of this data
                    set.

4.2.10  Data Set 10: Card Input for Initial or Pre-Initial  Conditions

       NNP records (i.e., one record for each node) are normally needed. However, if a
       group of node numbers follow a regular pattern, automatic generation can be
       used.

       Each record is FREE-FORMATTED and contains the following variables:

        1. NI = Global node  number of the first node in the sequence.

        2. NSEQ  = Number  of subsequent nodes which will be generated automatically.


        3. NAD = Increment of node number for each of the NSEQ nodes.

        4. CNI =  Initial or pre-initial concentration of node NI, (M/L3).

        5. CAD = Increment of CNI for each of the NSEQ nodes,  (M/L3).

        6. CRD = Geometrical increment of CNI for each of the NSEQ subsequent nodes
           (i.e, CNI**CRD).

       **** NOTE: A record with six zeroes must be used to signal the end of this data
                    set.

       **** NOTE: The initial condition for a transient calculation may be obtained in
                    two different ways:  1) it can be read directly from data set  10, or
                    2) the code  can perform a steady-state simulation using
                    time-invariant boundary conditions before beginning the transient
                    computations.  For the first case, both KSS and NTI in data set 2
                    should be greater than zero.  In the latter case, KSS = 0 and NTI >
                    0 and data set 10 is used to input the pre-initial condition, which is
                    required as the starting condition for the steady-state iteration.  In
                    order to obtain a steady-state solution, both KSS and NTI are set
                    equal to zero and data set 11 supplies the starting condition for the
                    steady-state solution.

4.2.11  Data Set 11: Integer  Parameters  for Sources  and  Boundary Conditions

       One record per problem is needed. This  record is  FREE-FORMATTED and
       contains the following variables:

        1. NSEL  = Number  of distributed source/sink elements.
                                         82

-------
 2. NSPR = Number of distributed source/sink  profiles (i.e., time histories)
    (NSPR should be > 1).

 3. NSDP = Number of data points in each of the NSPR source/sink profiles
    (NSDP should be > 2).

 4. KSAI = Option for the distributed source/sink profiles to be input
    analytically.  This  variable  should be set equal to zero in the current version
    of the code.

 5. NWNP = Number of well or point source/sink nodes.

 6. NWPR = Number of well  or point source/sink profiles (i.e.,  time histories).

 7. NWDP = Number of data points in each of the NWPR profiles.

 8. KWAI = Option  for the well source/sink profiles to be input analytically.
    This variable should be set equal to zero in the  current version of the code.

 9. NDNP = Number of prescribed-concentration (Dirichlet)  nodes (NDNP should
    be > 1).

10. NDPR = Number of prescribed-concentration profiles (i.e., time histories)
    (NDPR should be >  1).

11. NDDP = Number of data points in each prescribed-concentration profile
    (NDDP should be >  2).

12. KDAI = Option  for the prescribed-concentration  boundary  profiles to be input
    analytically.  This variable should be set equal  to zero in the current version
    of the code.

13. NVES = Number of variable composite  boundary element sides.

14. NVNP = Number of variable composite boundary nodal  points.

15. NRPR = Number of variable composite  profiles (i.e., time histories).

16. NRDP = Number of data points in each of the NRPR profiles.

17. KRAI = Option  for the variable composite profiles to be  input analytically.
    This variable should be set equal to zero in the current version of the code.

18. NCES = Number of specified-flux  (Cauchy) boundary element sides.

19. NCNP = Number of specified-flux boundary nodal points.

20. NCPR = Number of specified-flux profiles (i.e., time histories).

21. NCDP = Number of data points in each of the NCPR profiles.

                                    83

-------
       22.  KCAI = Option for the specified-flux profiles to be input analytically. This
           variable should be set equal to zero in the current version  of the code.

       23.  NNES = Number of specified-dispersive-flux (Neumann) boundary element
           sides.

       24.  NNNP = Number of specified-dispersive-flux boundary nodal points.

       25.  NNPR = Number of specified-dispersive-flux profiles (i.e., time histories).

       26.  NNDP = Number of data points in each of the NNPR profiles.

       27.  KNAI = Option for the specified-dispersive-flux profiles to  be input
           analytically. This variable should be set equal  to zero in the current version
           of the code.

4.2.12  Data Set 12: Distributed and Point Sources/Sinks

       This data set is used to supply data for both distributed sources/sinks, and point
       (well) sources/sinks.

       12A.  Distributed Sources/Sinks

       The following three subdata sets are needed if and only if NSEL in data set 11  is
       greater than zero. The first subdata set is used  to specify the  distributed
       source/sink profiles. The second subdata set is used to  read  the global element
       numbers of the distributed source/sink elements. The third subdata set is used
       to assign a source/sink profile to each distributed source/sink element.

       (a) Sources/Sink  Profiles

       NSPR records (see data set 11) are needed.  Each record contains NSDP  data
       points, defined in data set 11. Three numbers, representing  the time, source flow
       rate, and source concentration, respectively,  are  associated with each data point.

       Record I  (I = 1, 2,	NSPR) -  FREE FORMAT:  Each record contains the follow-
       ing variables:

           1.  TSOSF(J,I) = Time of J-th data point in  I-th profile,  (T).

           2.  SOSF(J,I,1) = Source/sink flow rate of the J-th data point in the I-th
              profile,  (LVWL3); positive for source  and negative for sink.

           3.  SOSF(J,I,2) = Source/sink concentration  of the J-th data point in the I-th
              profile, (M/L3).
              Up to NSDP data points.
                                          84

-------
(b)  Global Element Number of All Distributed Source/Sink Elements

One record is needed for this subdata set. The number of lines in the record
depends on NSEL, defined in data set 11. The record is FREE-FORMATTED
and contains the following variables:

     1. LES(l)  = Global  element  number of the  first distributed source/sink
       element.

    2. LES(2)  = Global  element  number of the  second distributed source/sink
       element.
       Up to NSEL numbers.

(c) Source/Sink Profile Type Assigned to Each Element

Usually NSEL records are needed. However, automatic generation can be used.
Each record is FREE-FORMATTED and contains the following variables:

    1.  MI = Compressed element number of the first element in the sequence.

    2.  NSEQ = Number of subsequent elements which will be automatically
       generated.

    3.  MAD = Increment of element number for each of the NSEQ elements.

    4.  MITYP = Source/sink profile associated with element MI.

    5.  MTYPAD = Increment of MITYP for each of the NSEQ subsequent
       elements.

                  A record with five zeroes must be used to signal the end of
                  this subdata set.

                  Compressed element numbers have values between one and
                  the total number of distributed source/sink elements.
                  Compressed element one corresponds to  the first element
                  listed in 12A(b), compressed element two corresponds to the
                  second global element, etc.

12B. Point (Well)  Source/Sink

The following three subdata sets are needed only if NWNP in data set 11 is
greater than zero.  The first subdata set is used to specify the point source/sink
profiles. The second subdata set reads the source/sink global node numbers, and
the third assigns a source/sink profile type to each node.

(a) Source/Sink Profiles

                                  85

-------
NWPR records (see data set 11) are needed. Each record contains NWDP data
points, defined in data set 11. Three numbers, representing the time, source flow
rate, and source concentration, respectively, are associated with each data point.

Record I (I = 1, 2, .... NWPR) - FREE FORMAT:  Each record  contains the
following variables:

    1. TWSSF(J,I) = Time of J-th data point in I-th profile, (T).

    2. WSSF(J,I,1) = Source/sink flow rate of the J-th data point in the I-th
       profile, (L3/T); positive for source and negative for sink.

    3. WSSF(J,I,2) = Source/sink concentration of the J-th  data point in the I-th
       profile, (M/L3).
        Up to NWDP numbers.

(b) Global Node Number of All Point (Well) Source/Sink Nodes

One  record is needed for this subdata set. The number of lines in this record
depends on  NWNP, defined  in  data  set  11.  The  record is  FREE-FORMATTED
and contains the following variables:

    1.   NPW(l) = Global node number of the first point source/sink node.

    2.   NPW(2) = Global node number of the second point source/sink node.



        Up to NWNP numbers.

(c) Source/Sink Profile Type for Each Node

Usually one record per node (i.e., NWNP records) are  needed.  However,
automatic generation can be used.  Each record is FREE-FORMATTED and
contains the following variables:

    1.   NI = Compressed point source/sink node number of the first node in a
        sequence.

    2.   NSEQ = Number of subsequent nodes which will be  automatically
        generated.

    3.   NIAD = Increment of NI for each of the NSEQ nodes.

    4.   NITYP = Source/sink profile associated with node NI.

    5.   NTYPAD = Increment of NITYP for each of the NSEQ  subsequent nodes.

                                  86

-------
           -J, -J, -J, -J, TV T t-^i r-p -i—i
               IN u it:    A record with five zeroes must be used to signal the end of
                         this subdata set.

4.2.13 Data Set 13: Variable  Composite Boundary Condition

      The following four subdata sets are required only if NVES in data set  11 is
      greater than zero. The first subdata set is used to specify the concentration
      profiles. The second subdata set is used to assign a concentration profile type to
      each of the variable composite boundary element sides. The third subdata set is
      used to specify the variable composite  boundary element sides. The fourth
      subdata set is  used to read the global nodal number of all the variable composite
      boundary nodes.

       ISA. Concentration Profiles

      There will be NRPR records (see data  set 11)  in this subdata set. The number of
      lines in each record depends on NRDP, defined in data set 11.

      Record I  (I = 1, 2,	NRPR) - FREE-FORMAT: Each record contains  the  follow-
      ing variables:

           1. TCRSF(IJ) = Time of the first data point in the I-th profile, (T).

           2. CRSF(1,I)  = Concentration of the first data point in the I-th profile,
              (M/L3).

           3. TCRSF(2,I) = Time of the second data point in the I-th profile, (T).

           4. CRSF(2,I)  = Concentration of the second data point in the I-th profile,
              (M/L3).
              Up to NRDP data points.

       13B. Concentration Profile Type Assigned to Each Boundary Element Side

       Usually one record per variable composite boundary element side is needed.
       However, automatic generation can be used.

       Record I (I = 1, 2, ....)- FREE-FORMAT: Each record contains the following
       variables:

           1.  MI = Compressed variable boundary element side of the first side in  a
              sequence.

           2.  NSEQ = Number of subsequent sides which will be generated
              automatically.

           3.  MJAD = Increment of MI for each of the NSEQ subsequent sides.

                                         87

-------
    4. MITYP = Type of concentration profile assigned to side MI.

    5. MTYPAD = Increment of MITYP  for each of the NSEQ subsequent sides.
    -J, -J, -J, -J, TV T /^'"PT"?
        INuit:   A  record with five zeroes must be used to signal the end of
                  this subdata set.

 13C.  Specification of  Variable Composite Boundary Element Sides

Normally, NVES records are required, one each for a variable boundary element
side.  However, if a group of variable composite boundary element sides appears
in a regular pattern,  automatic generation may be used.

Record I (I = 1, 2, ....)-  FREE-FORMAT: Each  record contains the following
variables:

     1. MI = Compressed variable composite boundary element side  number of
        the first side in  a sequence.

     2. NSEQ = Number of subsequent element sides which will be generated
        automatically.

     3. MIAD = Increment of MI for each of the NSEQ element sides.

     4. 11 = Global node number of  the first node of element side MI.

     5. 12 = Global node number of  the second node of element side MI.

     6. 13 = Global node number of  the third node of element side MI.

     7. 14 = Global node number of  the fourth  node of element side MI.

     8. HAD = Increment of 11 for each of the NSEQ subsequent element sides.

     9. I2AD = Increment of 12 for each of the NSEQ subsequent element sides.

    10. 13AD = Increment of 13 for  each of the NSEQ subsequent element sides.

    11. WAD = Increment of 14 for each of the NSEQ subsequent element sides.
    -j, -j, -j, -j, TV T /^'"pT"?
        IN u it:   A  record with 11  zeroes must be used to signal the end of this
                  subdata set.

 13D.  Global Nodal Number of All Variable Composite Boundary Nodes

At most, NVNP records (see data set 11)  are needed for this subdata set.
Record I (1 = 1, 2, ....)-  FREE-FORMAT: Each record contains the following
variables:
                                  88

-------
           1.   NI = Compressed variable boundary node number of the first node in the
               sequence.

           2.   NSEQ = Number of subsequent nodes which will be generated automati-
               cally.

           3.   NIAD = Increment for NI for each of the NSEQ nodes.

           4.   NODE = Global nodal number of the node NI.

           5.   NODEAD = Increment of NODE for each of the NSEQ subsequent
               nodes.

               NO IE:    A record with  five zeroes must be used to signal end of this
                         subdata set.

4.2.14  Data  Set  14: Prescribed-Concentration (Dirichlet) Boundary Condition

      This data set is required if and only if NDNP in data set  11  is greater than zero.
      It consists of three subdata sets.  The first subdata set is  used to specify the
      prescribed-concentration profiles,  the second is used to read  the prescribed-
      concentration boundary nodes, and the third is used to assign a  concentration
      profile type to each of the Dirichlet nodes.

       14A.  Prescribed-Concentration Profiles

      There will be NDPR records (see  data set  11) in this subdata set.  The number of
      lines in each record depends on NDDP, defined in data set 11.

      Record I (I = 1, 2,	NDPR) - FREE-FORMAT: Each  record  contains the
      following variables:

           1.   TCDBF(1,I) = Time of first data point in I-th profile, (T).

           2.   CDBF(1,I) = Concentration of first data point in  I-th profile, (M/L3).

           3.   TCDBF(2,I)  =  Time of second data point in  I-th profile, (T).

           4.   CDBF(2,I) = Concentration of second data point in I-th profile, (M/L3).
               Up to NDDP data points.

       14B. Global Node Number of All the Prescribed-Concentration Nodes

       One FREE-FORMATTED record is needed for this subdata set. The number of
       lines in this record depends on NDNP, defined in data set 11.
                                         89

-------
           1.   NPDB(l) = Global node number of the first compressed prescribed-
               concentration node.

           2.   NPDB(2) = Global node number of the second compressed prescribed-
               concentration node.
               Up to NDNP numbers.

       14C. Type of Concentration Profile Assigned to Each Dirichlet Node

       Normally one record per Dirichlet node (i.e., a total of NDNP records) is needed.
       However, if the Dirichlet node numbers follow a regular pattern, automatic
       generation may be used.

       Record I (I = 1, 2, .  . . . )  - FREE-FORMAT: Each record contains the following
       variables:

           1.   NI = Compressed Dirichlet node number of the first node in the
               sequence.

           2.   NSEQ =  Number of subsequent Dirichlet nodes which will be automati-
               cally generated.

           3.   NIAD = Increment of NI for each of the NSEQ nodes.

           4.   NITYP = Dirichlet concentration profile type assigned to node NI and
               NSEQ subsequent nodes.

           5.   NTYPAD = Increment of NITYP for each of the NSEQ subsequent
               nodes.

               NO IE:    A record with five zeroes must be used to signal the end of
                         this subdata set.

4.2.15  Data Set 15:  Specified-Flux (Cauchy) Boundary Condition

       Four subdata sets are required only if NCES in data set 11 is greater than zero.
       The first subdata set is used  to read the specified-flux profiles. The second
       subdata set is used to assign  the type of specified-flux profile to each of the
       specified-flux boundary element sides. The third subdata set is used to read the
       specified-flux boundary element sides. The fourth subdata set is used to read  the
       global nodal numbers associated with the specified-flux boundaries.

       ISA. Specified-Flux Profiles

       There will be NCPR records (see data set 11) in this  subdata set. The number of
       lines in each record depends on NCDP, defined in data set 11.
                                         90

-------
Record I (I = 1, 2,	NCPR) - FREE-FORMAT: Each record contains the follow-
ing variables:

     1.   TQCBF(U) = Time of the first data point in the I-th profile,  (T).

    2. QCBF(1,I) = Normal specified-flux of the first data point in the I-th
         profile, (M/T/L2); positive out of the region, negative into the region.

    3. TQCBF(2,I)  = Time of the second  data point  in the I-th profile, (T).

    4. QCBF(2,I) = Normal specified-flux of the second data point in the  I-th
         profile, (M/T/L2); positive out of the region, negative into the region.
         Up to NCDP data points.

 15B. Type of Specified-Flux Profile Assigned to Each Boundary Element Side

At most, NCES records (see data set 11) are needed.  However, automatic genera-
tion can be used.

Record I (I = 1, 2,	) - FREE-FORMAT: Each record contains the following
variables:

    1.    MI = Compressed specified-flux boundary element side number of the
         first side in the sequence.

    2.    NSEQ = Number of subsequent sides which will be generated automati-
         cally.

    3.    MIAD = Increment of MI for each of NSEQ subsequent sides.

    4.    MITYP = Type of specified-flux profile assigned to side MI.

    5.    MTYPAD = Increment of MITYP for each of the NSEQ subsequent sides.
        NO IE:   A record with five zeroes must be used to signal the end of
                  this subdata set.

 15C.  Specified-Flux Boundary Element Sides

Normally, NCES records are required, one for each specified-flux boundary
element side. However, if a group of specified-flux boundary element side
numbers follows a regular pattern, automatic generation may be used.

Record I (I = 1, 2, . . .  . ) - FREE-FORMAT: Each record contains the following
variables:
                                   91

-------
     1. MI = Compressed specified-flux boundary element side number of the
        first element side in a sequence.

     2. NSEQ = Number of subsequent element sides which will be generated
        automatically.

     3. MIAD  = Increment of MI for each of the NSEQ subsequent sides.

     4.11 = Global node number of the first node of element side MI.

     5. 12 = Global node number of the second node of element side MI.

     6. 13 = Global node number of the third node of element side MI.

     7. 14 = Global node number of the fourth node of element side MI.

     8. HAD = Increment of 11 for each of the NSEQ subsequent element sides.

     9. I2AD = Increment of 12 for each of the NSEQ subsequent element sides.

    10. I3AD = Increment of 13 for each of the NSEQ subsequent element sides.

    11. WAD = Increment of 14 for each of the NSEQ subsequent element sides.

        NO IE.    A record with  11 zeroes must be used to signal the end of this
                  subdata set.

15D. Global Node Number of All Compressed Specified-Flux Boundary Nodes

Usually NCNP records (see data set 11) are  needed for this subdata set.
However,  automatic generation can be used.

Record I (I = 1, 2, . . . . ) - FREE-FORMAT: Each record contains the following
variables:

    1.   NI = Compressed specified-flux boundary node number of the first node
        in a sequence.

    2.   NSEQ = Number of subsequent nodes which will be generated automatic-
        ally.

    3.   NIAD = Increment for NI for each of the NSEQ nodes.

    4.   NODE = Global nodal number of the node NI.

    5.   NODEAD = Increment of the global nodal number for each of the NSEQ
        subsequent nodes.

        NO IE.    A record with five zeroes  must be used to signal end of this
                  subdata set.

                                  92

-------
4.2.16  Data  Set 16: Specified-Dispersive-Flux (Neumann) Boundary Condition

       The following four subdata sets are required only if NNES in data set 11 is
       greater than zero. The first subdata set is used to read the specified-dispersive-
       flux profiles. The second subdata set is used to assign a specified-dispersive-flux
       profile type to each boundary element sides. The third subdata set is used to
       read the  specified-dispersive-flux boundary element side. The fourth subdata set
       is used to read the global nodal numbers associated with the specfied-dispersive-
       flux boundaries.

       16A. Prescribed Specified-Dispersive-Flux Profiles

       There will be NNPR records (see data set 11) in this subdata set. The number  of
       lines in each record depends on NNDP, defined in data set  11.

       Record I (I = 1, 2, .... NNPR) - FREE-FORMAT: Each record contains the
       following variables:

          1.    TQNBF(IJ) = Time of the first data point in the I-th profile, (T).

          2.    QNBF(1,I)  = Normal  specified-dispersive flux of the first data point in
               the I-th profile,  (M/T/L2); positive out of the region, negative into  the
               region.

          3.    TQNBF(2,I) = Time of the second data point in the I-th profile, (T).

          4.    QNBF(2,I) = Normal specified-dispersive flux of the second data point in
               the I-th profile,  (M/T/L2); positive out of the region, negative into  the
               region.
               Up to NNDP data points.

       16B. Type of Specified-Dispersive-Flux Profile Assigned to Each Boundary  Ele-
            ment Side

      At most, NNES records (see data set 11) are needed. However, automatic
      generation can be used.

      Record I (I = 1, 2, . .  . . ) - FREE-FORMAT: Each record contains the following
      variables:

          1.    MI = Compressed specified-dispersive-flux boundary element side of the
               first side in a sequence.

          2.    NSEQ  = Number of subsequent sides which will be generated automati-
               cally.

          3.    MIAD = Increment of MI for each of NSEQ subsequent sides.

                                         93

-------
    4.   MITYP = Type of specified-dispersive-flux profile assigned to side MI.

    5.   MTYPAD = Increment of MITYP for each of the NSEQ subsequent sides.


        NO IE:    A record with five zeroes must be used to signal the end of
                  this subdata set.

16C. Specified-Dispersive-Flux Boundary Element  Sides

Normally, NNES records are required, one for each specified-dispersive-flux
boundary  element side. However,  if  a group of specified-dispersive-flux element
side numbers follows a regular pattern, automatic generation may be used.

Record I (I = 1, 2, . . . . ) - FREE-FORMAT: Each record contains the following
variables:

     1.   MI = Compressed specified-dispersive-flux boundary element side
        number of the first element side in a sequence.

    2. NSEQ = Number of subsequent sides which will be generated automati-
        cally.

    3. MIAD =  Increment of MI for each of the  NSEQ subsequent sides.

    4.   II = Global node number of the first node of element side MI.

    5.  12  = Global node number of the second node of element side MI.

    6.   13 = Global node number of the third node of element  side ML

    7.  14  = Global node number of the fourth node of element side MI.

    8. HAD =  Increment of 11 for each of the NSEQ  subsequent element sides.

    9. I2AD =  Increment of 12 for each of the NSEQ  subsequent element sides.

   10.   I3AD = Increment of 13  for  each of the NSEQ subsequent element sides.

   11.   I4AD = Increment of 14  for  each of the NSEQ subsequent element sides.

        NOTE:    A record with 11  zeroes must be used to signal the end of this
                  subdata set.

16D. Global Node Number  of All Compressed Specified-Dispersive-Flux
     Boundary Nodes

Usually NNNP records (see data set  11)  are needed for this subdata set.
However,  automatic generation can be used.
                                  94

-------
       Record I (I = 1, 2, ....)- FREE-FORMAT:  Each record contains the following
       variables:

          1.    NI = Compressed specified-dispersive-flux boundary node number of the
               first node in a sequence.

          2.    NSEQ =  Number of subsequent nodes which will be generated automati-
               cally.

          3.    NIAD = Increment of NI for each of the NSEQ nodes.

          4.    NODE =  Global nodal number of the node NI.

          5.    NODEAD = Increment of the global nodal number for each of the NSEQ
               subsequent nodes.

                         A record with five zeroes must be used to signal end of this
                         subdata set.

4.2.17  Data Set 17: Hydrological Variables

       This data set is needed only if KVI in data set 2 is less than  or equal  to zero.
       When KVI <; 0, two subdata sets are needed: one for the velocity field and the
       other for the moisture content.

       17A.  Velocity  Field

       Usually  NNP records (see data set 2) are  needed. However, if the velocity values
       follow a  regular pattern, automatic generation can be used.

       Record I (I = 1, 2, ....)- FREE-FORMAT:  Each record contains the following
       variables:

          1.    NI = Node number of the first node in a sequence.

          2.    NSEQ =  Number of subsequent nodes  which will be automatically gener-
               ated.

          3.    NIAD = Increment of NI for each of the NSEQ subsequent nodes.

          4.    VXNI = X-velocity component at node NI, (L/T).

          5.    VYNI = Y-velocity component at node NI, (L/T).

          6.    VZNI = Z-velocity component at node NI, (L/T).

          7.    VXAD =  Increment of VXNI for each of the NSEQ subsequent nodes,
               (L/T).
                                         95

-------
          8.    VYAD = Increment of VYNI for each of the NSEQ subsequent nodes,
               (m).

          9.    VZAD = Increment of VZNI for each of the NSEQ subsequent nodes,
               (m).

              NOTE:    A record with nine zeroes must be used to signal the end of
                        this subdata set.

       17B. Moisture Content Field

       Usually, NEL records (see data set 2) are needed. However, if the moisture
       content values follow a regular pattern, automatic generation can be used.

       Record I (I = 1, 2, . . .  . ) - FREE-FORMAT: Each record contains the  following
       variables:

          1.  MI  =  Element number of the first element in a sequence.

          2.    NSEQ =  Number  of subsequent elements which will be automatically
               generated.

          3.    MIAD = Increment of MI for each of NSEQ subsequent elements.

          4.    THNI = Moisture content  of element MI, (decimal point).

          5.    THNIAD = Increment of THNI for the NSEQ subsequent elements,
               (decimal point).

              NO IE:    A record with five zeroes must be used to signal the end of
                        this subdata set.

4.2.18  Data Set 18:  End of Job

       If another problem is to be run, then input begins again with input data set  1. If
       termination of the job is desired, a blank line must be inserted at the end of the
       data set.
                                         96

-------
                                    SECTION 5

                             PARAMETER SELECTION
This section provides guidance in selecting values for some of the parameters required
as input to the 3DFEMWATER/3DLEWASTE codes. This guidance is not intended in
any way to be used as a substitute for data collection.  The most accurate model results
are obtained from simulations which are based on site-specific information. In  some
cases, however, it is not feasible to measure certain parameters, and satisfactory results
may be obtained  using estimated values taken from the reported ranges presented here.
For easy reference, the parameters are grouped according to the data group in which
they appear in the input data sets (see Section 4). Concepts, such as initial and
boundary conditions, isotherms, distributed and point sources and sinks, and
subregional data, were introduced in Section  3 and guidance is not provided in this
section for related parameters.

5.1 3DFEMWATER

5.1.1  Data Set 1:  Title of the Simulation Run

5.1.1.1 Geometry,  Boundary, and Pointer Array  Control, IGEOM [—]

The integer IGEOM has two functions.  It is  used to specify if geometry, boundary, and
pointer arrays should be printed so that the  user can examine them. It also controls
whether the boundary and pointer arrays are written to or read from binary files.
Boundary arrays store data related to the boundary conditions. Pointer arrays store the
global matrix in compressed form and are used to construct the subregional block
matrices.  For large problems, it takes too much time to generate these arrays for  each
computer execution of a particular scenario.  Usually, they should be generated only
once and stored in binary files using logical units LUBAR and LUPAR (see Table  B-l).

In order to compute and store the boundary and pointer arrays, the user should choose a
value for IGEOM less than or equal  to one. In subsequent runs, the boundary and
pointer arrays can be read from the binary files by changing the value of IGEOM to a
number greater than  three. Whenever changes are made to the model which involve the
geometry of the problem, the boundary conditions, and the configuration of the
subregions, the arrays must be generated and stored again. Note that the option
presented in the input to read boundary arrays and compute and write pointer arrays is
not used in 3DFEMWATER under normal conditions.

For the options explained  above, if the number chosen by the user is even, the arrays
will be printed as  output.  If the number is odd, the arrays will not be printed.
                                         97

-------
5.1.2 Data Set 2: Basic Integer Parameters

5.1.2.1 Number of Material Types, NMAT [--]

This parameter is the total number of different porous media being modeled. For
example, if the region of interest is predominantly sand with clay lenses, then the value
of NMAT should be set  equal to two.  When material properties are assigned to each
material type, using data set 5  (see Section 4.1.5), the first material type should be the
predominant porous medium  (e.g.,  for the example here, the sand).

5.1.2.2 Number of Elements with Material Property Correction, NCM [—]

In the code, all the grid elements automatically are initialized as having a material type
of one. If the region being modelled is homogeneous,  the parameter NCM is set equal  to
zero. To  model a heterogeneous porous medium, NCM and the parameters in data  set
10 of the input (see Section 4.1.10) are used to specify which elements have different
material types associated with them.  The parameter NCM is the total number of
elements which have a material type different than the first material type.

5.1.2.3 Number of Time-Steps, NTI [--]

For a constant time-step size, this  number is obtained by dividing the simulation time
by the time-step  size, BELT.  If the time-step size is variable, this number is computed
using the formula given in the note at the end of data set 2 in Section 4.1.2. If a steady-
state solution is desired, NTI should be set equal to zero.

5.1.2.4 Steady-State Control,  KSS  [-]

As noted in Section 4.1.2, a steady-state option may  be used to provide either the final
state of a system under  study or the initial condition for a transient-state calculation.
In the former case, both KSS and the  number of time steps, NTI, should be set to zero.
In the latter case (i.e., when KSS = 0  and NTI  > 0), the code performs a steady-state
calculation before beginning the transient computations. If KSS =  1, no steady-state
calculation is  performed. Rather, the  code begins transient  calculations using initial
conditions supplied in data set 11 of the input.

5.1.2.5 Gravity Term Control, KGRAV [-]

This parameter indicates if the gravity term should  be included. For most cases,
KGRAV should be equal to 1. For cases when flow due to the pressure gradient is much
greater than that due to gravity, KGRAV is set to 0.

5.1.2.6 Mass Lumping Flag, ILUMP [-]

This parameter indicates if the mass matrix is to be lumped or not. Normally, one
should set  this parameter to O. Without lumping, the solution is more accurate.
However, for occasions when negative concentrations or oscillating solutions occur, this
parameter  should be set to 1. It has been suggested that for saturated-unsaturated flow
computations, the parameter ILUMP should always be set equal to 1.
                                          98

-------
5.1.2.7 Mid-differencing  Flag, IMID [--]

This parameter indicates if the more accurate mid-difference method should be used in
flow computations. For  practical purposes, IMID = 0 should be sufficient. IMID =  1 is
used only for research purposes.

5.1.2.8 Number of Iterations for the Nonlinear Equation, NITER [—]

This parameter is the number of iterations allowed for solving the nonlinear equation.
Normally, NITER = 50 should be sufficient. If this  number is exceeded and the solution
does not converge, the program will issue a warning message. When this occurs, the
user should first  recheck the input values.  If the input is correct, the program can be
re-executed using a larger value for NITER.

5.1.2.9 Number of Cycles, NCYL [-]

This parameter indicates how many cycles  are used for iterating the boundary
conditions. A value of 20 should be adequate for most problems.

5.1.2.10 Number  of Times to Reset the Time Step, NDTCHG [-]

This parameter indicates how many times the time-step size should be reset to the
initially small time-step  size.  When we start a simulation, we normally use a small
time-step size.  However, for every consecutive time step, we may gradually increase  the
time-step size by  some amount specified by the variable CHNG in Data Set  3 in Section
4.1.3. When a steep change in boundary conditions or source/sink conditions occurs,
however, the time-step size should be reset to the initially small value. (See the
example problem in Section 6.1.1.)  NDTCHG tells us how many times we want to  reset
the time-step size during a simulation.

The value of NDTCHG must be at  least one. If the  user does not want to reset the time
step, a value of one should be entered here and a very large number, larger than the
total simulation time, should be entered for TDTCH(l) in data set 4 (see Section 4.1.4).

5.1.2.11 Number  of Iterations for Pointwise Solution, NPITER  [-]

This parameter is used to input the number of iterations allowed for solving the matrix
equations with the block iteration method.  A value of 300 should be sufficient for most
problems. If this  number is exceeded and the solution does not converge, the program
will issue a warning  message. When  this occurs, the users should re-execute the
program using a  larger value  for NPITER.

5.1.3 Data Set 3: Basic  Real  Parameters

5.1.3.1 Initial Time-Step Size, DELT [T]

This is the time-step size used for the first time-step computation if the variable CHNG
is not equal to 0.0. It is  the time-step size used for  every time step if the variable
CHNG is set equal to 0.0. It is advisable to choose the value of DELT such  that:
                                         99

-------
       (F*DELX*DELX)/(DELT*K) < 1

where

       DELX = the element size (L)
       K    = hydraulic conductivity  (L/T)
       F    = specific storage (1/L)

For example, if F = 0.001 1/m, K = 0.00001 m/sec, and an element size of 10 m is used,
then BELT should be less than  10,000 seconds.

5.1.3.2 Fractional Change in Time-Step  Size, CHNG [—]

This parameter specifies how much of an increase one would like to make to the time-
step size for each subsequent time step.  Normally, a value from 0.0 to 0.5 can be used.

5.1.3.3 Maximum Allowable Time Step,  DELMAX [T]

The maximum time-step size allowed depends on how fast the system responds to
change. Use of a value  one  to ten times  the size of the initial time step is advised.

5.1.3.4 Maximum Simulation Time, TMAX  [T]

This is the actual length of time to be simulated.  If this time is exceeded before you
have made NTI step computations, the simulation will be terminated.

5.1.3.5 Steady-State Convergence Criterion,  TOLA [L]

This is the absolute error allowed for assessing if a steady-state solution for hydraulic
head has converged. The value used for  TOLA depends on how much the system is
disturbed. Normally, setting TOLA equal to  one-ten-thousandth (0.0001)  of the
maximum disturbance should be sufficient. For example, if one is conducting a
simulation of drawdown due to pumping and one expects the  maximum drawdown at
steady-state will be 1 m, then a value of TOLA equal to 0.0001 m should be sufficient.

5.1.3.6 Transient Convergence  Criterion, TOLB [L]

This is the absolute error allowed for assessing if the solution for hydraulic heads has
converged for each transient time step. A value equal to one-hundred-thousandth
(0.00001) of the maximum disturbance should be sufficient for most problems.

5.1.3.7 Density of Water,  RHO [M/L3]

The density of water,  pw, is  the ratio of its mass to its volume and has SI units of kg/m3
Density varies with temperature (Table 5-1)  and can be computed using regression
equations presented in  CRC (1981). Density  also varies with  the concentration of
dissolved chemical species. Water density appears in the definition of specific storage
                                         100

-------
TABLE 5-1.  WATER DENSITY AS A FUNCTION OF TEMPERATURE
Temperature
(°C)
0
10
20
30
40
50
60
70
80
90
100
Density
(kg/m3)
999.87
999.73
998.23
995.67
992.24
988.07
983.24
977.81
971.83
965.34
958.38
Source: Mercer et  al.,  1982; Original Reference: CRC, 1965
and in the relationship between hydraulic conductivity and intrinsic permeability
(Section  5.1.4.2).

5.1.3.8 Dynamic Viscosity of Water, VISC [M/L/T]

The viscosity of a fluid is a measure of the forces that work against flow when a
shearing stress  is applied (Lyman et al.,  1982). The more viscous a fluid is, the greater
the shear stress needed  to maintain a given velocity gradient.  Dynamic viscosity is
often expressed in terms of poise (gram per centimeter per second) or centipoise (0.01
poise). Water has a viscosity of approximately 1 centipoise at 20°C.

Viscosity varies with  temperature, as  indicated in Table  5-2, and with concentration of
dissolved chemicals. The effect of pressure on fluid viscosity is generally unimportant
(Mercer et al.,  1982). Note that dynamic  viscosity is a term in  the relationship between
hydraulic conductivity and intrinsic permeability (Section 5.1.4.2).

5.1.3.9 Time Integration  Weighting  Factor, W [—]

A value of W equal to 1.0 should be used for most practical problems  (see Equation  3-
16). Setting W equal  to 0.5 is normally done for research purposes to assess the
accuracy of the  Crank-Nicolson scheme.

5.1.3.10 Relaxation Parameter for Solving the Nonlinear Equation, OME [—]

Normally this parameter should be set to 1.0 (see Equation 3-17). If the convergence
history shows signs of oscillation, then a  value  of 0.5 should be used. If the convergence
                                         101

-------
TABLE 5-2. DYNAMIC VISCOSITY OF WATER AS A FUNCTION OF
            TEMPERATURE
Temperature
("c)
o
10
20
30
40
50
100
Dynamic Viscosity
(centipoise)
1.7921
1.3077
1.0050
0.8007
0.6560
0.5494
0.2838
Source: CRC, 1965
history shows monotonic decreases but at a very slow rate, then OME should be set to
somewhere between 1.7 to  1.9.

5.1.3.11 Iteration Parameter to Solve the Linearized Matrix Equation, OMI [— ]

Normally this parameter should be set to 1.0 (see Equation 3-18). If the convergence
history shows signs of oscillation, then set OMI to 0.5. If the solution converges
monotonically but at a very slow rate, then set OMI to between  1.7 and 1.9.

5.1.4  Data Set 5: Material Properties

In the material properties data set, the user  must input  values for either hydraulic
conductivity or permeability for each aquifer/soil material type.  The flag that tells the
code which of these two properties is being input is the permeability input control, KCP,
located in Data Group 6.

5.1.4.1 The Saturated Hydraulic Conductivity Tensor [L/T]

Hydraulic conductivity is the coefficient of proportionality which appears in Darcy's Law.
It expresses the ease  with which a fluid can be transported  through a porous medium
and is a function of properties of both the porous medium and the fluid  (Mills et al.,
1985b). It is defined as the volume  of water that will move in unit time under a unit
hydraulic gradient through a unit area measured at right angles to the direction of flow.
For three-dimensional flow in an anisotropic  medium, hydraulic conductivity varies with
direction at any point in space and  is expressed as a symmetric  second-rank tensor:
                                               Kxz
                                           V"
                                        yx;  yy
                                         102

-------
where Ky is the hydraulic conductivity tensor and x, y, and z are the coordinate axes of
the model grid. Because of symmetry, only 6 of the 9 terms are needed (K,^, Kyy, K^,
and Kxy = Ky,,Kxz = K,,, and K,, = K,y).
If the coordinate axes coincide with the principal directions of anisotropy, then the nine
components of the tensor reduce to K^, K^, and KZI, with the other components equal to
zero. For isotropic media,  hydraulic conductivity is independent of the direction of
measurement (i.e., K^ = K^ = K^ ).
Hydraulic conductivity estimates should be based on site-specific data collection (e.g.,
pumping  tests or piezometer tests). Some  typical horizontal hydraulic conductivity
values for various materials are shown in  Table 5-3. Note that hydraulic conductivity
varies over  a very wide range. As a result, values are rarely known with more than an
order-of-magnitude accuracy. Hydraulic conductivity values for fractured rock can be
found in Mercer et al. (1982).

For many materials, the vertical hydraulic conductivity is substantially smaller than the
horizontal hydraulic  conductivity (assuming horizontal bedding and measurements made
along the principal axes) (Mercer et al.,  1982). Mills et al. (1985b) state that the ratio of
horizontal to vertical conductivity, known as  the anisotropy ratio, is from 2 to 10 for
alluvium  and glacial outwash and from 1.5 to 3  for sandstone. The variability in
horizontal and vertical conductivities for a few aquifer materials is shown in Table 5-4.

5.1.4.2 The Permeability Tensor [Lz]

Intrinsic permeability is a property of the  porous medium only. It is a measure of the
resistance to fluid flow through  the medium. The greater the permeability, the less the
resistance. Like hydraulic conductivity, permeability is a symmetrical second-rank
tensor. Permeability  is equal to  hydraulic  conductivity multiplied by a scalar value, as
is seen in the following equation:

                ^ = Ky p/(pg)                                                     (5-2)

where

       ky  =  permeability  (L')
       Ky = hydraulic conductivity  (L/T)
       p.   = dynamic viscosity  (M/L/T)
       p   = density (M/L3)
       g   = acceleration of gravity (L/T2)

As was true for hydraulic conductivity, permeability estimates should be based on site-
specific data collection.  Ranges  of values for permeability are shown in Table 5-3 and in
Table 5-5. Permeability is sometimes expressed in units  of darcies. Conversion from
darcies to other units can be done by using the conversion factors provided at the bottom
of Table 5-3.
                                          103

-------
TABLE 5-3. RANGE OF HYDRAULIC  CONDUCTIVITY VALUES FOR VARIOUS
            GEOLOGIC MATERIALS  (Freeze and Cherry, 1979)












]
3
1
1

Q

















i.
8 •

1
£







ROCKS










|
!
! 8
I-
a



!
i
J





























Uncomolldaud
dspeslB
























j» 0 i

J '
ll





J
'I
ll

















5 •
'I-
J
















5
:
r [

\















i
,
?
i
j
1















I
J
i

f














T
\
1
c

















f
















k k K K K
(



















oarcyj
r-10>
_10<

-io3
-io2

- 10

"'
_io-'
-ID-*

-10°
-10^

-io-J
-io-«

-1C'7
-io-«
(cm^;
-10-3
-10^

-10-'
-in"6

-ID'7

-m*
-iO-9
-io-10

-io-11
-io-12

-io-"
-ID'"

_10"
-10 -»
(OBIS)
r-102
- 10

- 1
- 10-1

- 10-2

- 10-3
-10-*
- 10-'

-m-»
- 10-7

_ 10-*
_io-»

_ 10-K
_ 10-1:
(m/s; tgw/oay/ti-;
l_ i
-io-1

-ID'2
-10-3

-10"4

-10-'
-10-*
-10-7

-10 •*
-10-9

- 10 -«
- 1C'11

_ 10 -i:
- 10 -i:

_10«
_ 10'

-104
- 10'

- io2

- 10
- 1
_ 10-1

- 10-2
- 10 -'

-10^
_ 10-'

- 10 -«
- 10-7
                             Conversion Factors for Permeability
                             and Hydraulic Conductivity Units


cm2
ft2
darcy
m's
ft/8
U.S. gal/
day/ft2

cm2
1
9.29x10"
9.87X10'9
1.02xlO-3
S.llxlO-4
5.42xlO-10
Permeability, k*
ft2
l.OSxlO'3
1
1.06X10-11
1.1 0x10-"
3.35x10-'
5.83X10-13
Hydraulic conductivity, K
darcy
1. Olxl O3
9.42xl010
1
1. 04x10"
3.15xl04
5.49xlO'2
m/s
9.80X102
9.11X105
9.66X10'6
1
3.05xlO-1
4.72xlO'7
ft/s
3.22X103
2.99xlOe
3.17X10-"
3.28
1
1.55x1 
-------
TABLE 5-4. VARIABILITY IN HORIZONTAL AND VERTICAL HYDRAULIC
            CONDUCTIVITIES

Rock Types              KH, m/s               KV) m/s                 KV/KH
Shale
Siltstone-shale
Siltstone-shale
Sandstone
2.0 x ID'8
2.1 x ID'6
2.8 x 10-7
3.4 x 10-7
1.0 x 10-8
2.1 x lO'7
3.0 x 10-8
3.4 x lO'7
0.5
0.1
0.107
1.0
Source: Mercer et al.,  1982; Original Reference: Colder Associates,  1977
TABLE 5-5. PERMEABILITY OF  POROUS MATERIALS
    Material                                                        k (m )


Argillaceous limestone              2% porosity                    9.87 x 10"17
Limestone                          16% porosity                   1.38 x 10"13
Sandstone, silty                    12% porosity                   2.57 x 10'15
Sandstone, coarse                   12% porosity                   1.09 x 10"12
Sandstone                          29% porosity                   2.37 x 10'12
Very fine sand                      well  sorted                     9.77 x 10'12
Medium sand                       very well sorted                 2.57 x 10'
Coarse sand                        very well sorted                 3.06 x 10"
Gravel                             very well sorted                 4.24 x 10
Montmorillonite clayb                                              10"17
Kaolinite clayb                                                     10'1B
-10
,-8
a These are provided as estimates; actual values will vary.
b For the clays, only the order of magnitude is indicated.

Source: Mercer et al., 1982; Adapted from: Davis and DeWiest, 1966
                                        105

-------
5.1.5 Data Set 6: Soil Property Parameters

As was explained in Section 3.1.1, relationships between relative permeability and water
content and between pressure head and water content must be specified in order to
solve the governing equation for unsaturated flow. The 3DFEMWATER code provides
two  options for specifying these relationships. The user can  1) input parameters for
analytical expressions of these  relationships, or 2)  input the coordinates of
characteristic curves in tabular format. The analytical parameters are discussed first,
followed by the tabular data requirements.

5.1.5.1  Analytical Parameters

Analytical equations developed by van  Genuchten  (1980)  are  used in the code to describe
the relationship between  pressure head and moisture content and the relationship
between relative  hydraulic conductivity and moisture content (see Equations 3-3a
through 3-3d). In order to solve these equations, five parameters must be specified in
the input sequence for each material type: residual moisture content, saturated moisture
content, air entry pressure head, and two soil-specific empirical parameters,  alpha and
beta.

5.1.5.1.1 Residual and Saturated Moisture (Water)  Content [—1

The  volumetric moisture  content, 9, is defined as:

                9 = VW/VT                                                        (5-3)
where

       VT  = the total unit volume of a  rock or soil (L3)
       VW = the  volume of a rock or soil occupied by water (L3)

The  saturated moisture content is equal to the porosity of the medium since  all of the
void space is filled with  fluid. Under unsaturated  conditions, however,  some of the void
space is filled with air and thus, the moisture content is less than the medium's
porosity. The  residual moisture content is that  amount which can not be removed from
a soil by gravity  drainage, even under large suction pressure, because it adheres to the
grains of the soil.

Table 5-6 lists descriptive statistics for both saturated and residual moisture content for
a variety of soil types. In addition, saturated and  residual moisture content  values for a
large number of soils can be obtained using  the  interactive computer program  DBAPE
(Imhoff et al., 1990). DBAPE, which is a soils data base  analyzer and parameter
estimator, is  available from the U.S. EPA Center for Exposure Assessment Modeling
(CEAM) at the Environmental Research Laboratory in Athens, Georgia.

5.1.5.1.2 Air  Entry Pressure Head [L]

The  air entry pressure head is the threshold at which air starts to penetrate saturated
soil.  It is typically a very small negative value  for fine-grained materials and zero for
coarser materials. Its value can be estimated from  the water retention curves  of specific
                                         106

-------
TABLE 5-6. DESCRIPTIVE STATISTICS FOR SATURATION WATER CONTENT (9.)
            AND RESIDUAL WATER CONTENT (9r)

                   Saturation Water Content (9,)      Residual Water Content (9r)
                                              Statistic*
Soil Type            x       s       CV     n        x       s       CV     n
Clay**
Clay Loam
Loam
Loamy Sand
Silt
Silt Loam
Silty Clay
Silty Clay Loam
Sand
Sandy Clay
Sandy Clay Loam
Sandy Loam
0.38
0.41
0.43
0.41
0.46
0.45
0.36
0.43
0.43
0.38
0.39
0.41
0.09
0.09
0.10
0.09
0.11
0.08
0.07
0.07
0.06
0.05
0.07
0.09
24.1
22.4
22.1
21.6
17.4
18.7
19.6
17.2
15.1
13.7
17.5
21.0
400
364
735
315
82
1093
374
641
246
46
214
1183
0.068
0.095
0.078
0.057
0.034
0.067
0.070
0.089
0.045
0.100
0.100
0.065
0.034
0.010
0.013
0.015
0.010
0.015
0.023
0.009
0.010
0.013
0.006
0.017
49.9
10.1
16.5
25.7
29.8
21.6
33.5
10.6
22.3
12.9
6.0
26.6
353
363
735
315
82
1093
371
641
246
46
214
1183
* n = Sample size, x = Mean, s = standard deviation, CV = coefficient of
     variation (percent)

** Agricultural soil, less than 60 percent clay.

Source: Sharp-Hansen et al. (1990)
      Original source Carsel and Parrish (1988)
soils (Freeze and Cherry, 1979; Sharp-Hansen et al.,  1990). In practice, it is regularly
assumed to be zero.

5.1.5.1.3 Van Genuchten Parameters, a [1/L]; p [--]

These are empirical parameters needed to solve the van Genuchten analytical  equations
which are used to model unsaturated flow (see Equations 3-3a through 3-3d).
Descriptive statistics for these parameters have  been  reported by Carsel and Parrish
(1988) for a variety of soils and are shown in Table 5-7.
                                        107

-------
o
00
     5-7.  DESCRIPTIVE STATISTICS FOR VAN GENUCHTEN WATER RETENTION MODEL PARAMETERS, a, B, and y

          (Carsel and Parrish 1988)

Soil Type
Clay'
Clay Loam
Loam
Loamy Sand
Silt
Silt Loam
Silty Clay
Silty Clay Loam
Sand
Sandy Clay
Sandy Clay Loam
Sandy Loam
Parameter
X SD
0.008 0.012
0.019 0.015
0.036 0.021
0.124 0.043
0.106 0.007
0.020 0.012
0.005 0.005
0.010 0.006
0.145 0.029
0.027 0.017
0.059 0.038
0.075 0.037
a, cm"1
CV
160.3
77.9
57.1
35.2
45.0
64.7
113.6
61.5
20.3
61.7
64.6
49.4
Parameter 6
N
400
363
735
315
88
1093
126
641
246
46
214
1183
X
1.09
1.31
1.56
2.28
1.37
1.41
1.09
1.23
2.68
1.23
1.48
1.89
SD
0.09
0.09
0.11
0.27
0.05
0.12
0.06
0.06
0.29
0.10
0.13
0.17
CV
7.9
7.2
7.3
12.0
3.3
8.5
5.0
5.0
20.3
7.9
8.7
92
N
400
364
735
315
88
1093
374
641
246
46
214
1183
X
0.08
0.24
0.36
0.56
0.27
0.29
0.09
0.19
0.62
0.18
0.32
0.47
Parameter y
SD
0.07
0.06
0.05
0.04
0.02
0.06
0.05
0.04
0.04
0.06
0.06
0.05
CV
82.7
23.5
13.5
7.7
8.6
19.9
57.1
21.5
6.3
34.7
53.0
10.1
N
400
364
735
315
88
1093
374
641
246
46
214
1183
     X = Mean, SD = Standard Deviation, CV = Coefficient of Variation (percent), N = Sample Size



     * = Agricultural Soil, Clay 60 percent

-------
 5.1.5.2 Tabular Data Parameters

 If the user chooses to supply the soil moisture relationships in tabular form, four
 parameters must be specified for each functional data point: moisture content, the
 corresponding pressure head and relative permeability, and water capacity. Sets of
 these parameters must be input for each type of material being simulated. The
 necessary tabular data for a large number of soils  can be obtained from the interactive
 computer code, DBAPE, described in Section 5.1.5.1.1, with the exception of water
 capacity.

 Moisture content was described in Section  5.1.5.1.1. Pressure head, relative
 permeability, and water capacity are briefly introduced below.

 5.1.5.2.1 Relative Permeability (or Hydraulic Conductivity)  [—]

 In an unsaturated porous medium, the permeability of the water phase in the medium
 is a function of the degree of saturation. The larger the degree of saturation, the larger
 the permeability associated with the water phase. This unsaturated permeability is also
 known as the effective permeability.

 Relative permeability  is defined as the ratio of the effective permeability to the
 permeability at saturation.   Because it is a ratio, relative permeability ranges in value
 between 0.0  and  1.0. It is generally assumed that  relative permeability is a scalar,
 dimensionless non-linear function, even for anisotropic soils. Because of the relation of
 equivalence,  relative permeability is equal to relative hydraulic conductivity (Mercer et
 al., 1982).

 Curves showing the relationship between relative permeability and moisture content are
 determined experimentally for individual soils. The tabulated data available in the
 literature or  in DBAPE (Imhoff et al., 1990) are  extracted from these experimental
 results.

 5.1.5.2.2 Pressure head [L]

 In groundwater hydrology, the total hydraulic head, H, is usually considered to be the
 sum of two components: elevation head, z,  and pressure head, h.  The contribution of
velocity to the total head is neglected because velocities are usually extremely low.
 Pressure head is measured in  gage pressure. In  the saturated zone, pressures are
greater than  atmospheric and  are thus recorded as positive pressures. The water table
is defined as the location at which pressure is equal to atmospheric. This implies that
pressure head is zero  and the  total head is equal to the elevation head. Above the water
table, pressure head is less  than atmospheric and water is held in the pore spaces under
tension or suction. Thus, pressure head values in the unsaturated zone are negative.

Pressure head in the unsaturated zone is a function of moisture content—the lower the
moisture content, the  more negative the pressure head. As moisture content increases,
the surface tension forces holding the water in place between the grains of soil are
lowered, resulting in less negative pressure heads  (Freeze and Cherry,  1979). The
characteristic curve showing the relationship between pressure head and moisture
content is determined  experimentally for each porous medium.

                                         109

-------
 5.1.5.2.3 Moisture  Content Capacity [1/L]

 In an unsaturated soil, changes in moisture content, 9, are accompanied by changes in
 pressure head, h. As discussed above, the 6(h) relationship results in a characteristic
 curve for each soil.  Example characteristic curves  are shown in Figure 3.3a. The
 inverse of the slope  of this curve is called the water capacity, C(0), or the moisture
 content capacity (Mercer et al.,  1982). It is defined as:

                                 C(9) = dO/dh                                    (5-4)

 The  water capacity has no one unique value for a  porous medium. Thus, the range of
 values of moisture content capacity is related to the nature of the water characteristic
 curve.  (Mercer et al., 1982).
5.2  3DLEWASTE

5.2.1  Data Set 1: Title of the Simulation Run

5.2.1.1 Geometry, Boundary, and Pointer Array Control,  IGEOM [—]

The integer IGEOM has two functions. It is used to specify if geometry, boundary, and
pointer arrays should be printed so that the user can examine them. It also controls
whether the boundary and pointer  arrays are written to or read  from binary files.
Boundary arrays store data related to the boundary conditions. Pointer arrays store the
global matrix in compressed form and are used to construct the subregional block
matrices. For large problems, it takes too much time to generate these arrays for each
computer execution of a particular  scenario.

If 3DLEWASTE is being executed alone (i.e.,  without using 3DFEMWATER results),
these arrays should be generated only once and stored in binary files using logical units
LUBAR and LUPAR (see Table B-2). In order to compute and store the boundary and
pointer arrays, the user should choose a value for IGEOM less than or equal to one. In
subsequent runs, the boundary and pointer arrays can be read from the binary files by
changing the value of IGEOM to a number greater  than  three. Whenever changes are
made to the model which involve the geometry of the problem, the boundary conditions,
and the configuration of the subregions, the arrays must  be generated and stored again.

If 3DLEWASTE is run after executing 3DFEMWATER for the same scenario,  the
boundary array need not be recalculated (i.e., the boundary array calculated and stored
by 3DFEMWATER  can be  used). The pointer array should be recalculated, however.
This is done by setting IGEOM to a value greater than one and less than or equal to
three.

For each of the options explained above, if the number chosen by the user is even, the
arrays will be printed as output.  If the number is odd, the arrays will not be printed.
                                         110

-------
5.2.2 Data Set 2: Basic Integer Parameters

5.2.2.1 Number of Material Types, NMAT [--]

This parameter is the total number of different material types being modeled. When
material properties are assigned to each material type, using data set 5 (see Section
4.2.5), the first material type should be the predominant type. The number of material
types used in 3DLEWASTE need not be identical to the number specified in
3DFEMWATER.

5.2.2.2 Number of Elements with Material Property Correction, NCM [—]

In the code, all the grid elements automatically are initialized as having a material type
of one. To model more than one material  type, the parameter NCM and the parameters
in data set 9 of the input (see Section 4.2.9) are used to specify which elements have a
material type other than material type one. The parameter NCM is the total number of
elements  which have a material type different than the first material type.

5.2.2.3 Number of Time-Steps,  NTI [--]

For a constant time-step size, this number is obtained by dividing the simulation time
by the time-step  size, BELT. If the time-step size  is variable, this number is computed
using the formula given in the note at  the end of data set 2 in Section 4.2.2.

5.2.2.4 Steady-State  Control,  KSS [-]

As noted in Section 4.2.2, a steady-state option may be used to provide either the final
state of a system under study or the initial condition for a transient-state  calculation.
In the former case, both KSS and the number of time steps, NTI, should be set to zero.
In the latter case (i.e., when KSS = 0 and NTI > 0), the code performs a steady-state
calculation before beginning the transient computations. If KSS = 1, no steady-state
calculation is performed. Rather, the code begins transient calculations using initial
conditions supplied in data set 10 of the input.

5.2.2.5 Mass Lumping Flag, ILUMP [-]

This parameter indicates if the mass matrix is to be lumped or not. Normally, one
should set this parameter to 0. Without lumping, the solution is more accurate.
However,  for occasions when negative concentrations or oscillating solutions occur, this
parameter should be  set to 1.

5.2.2.6 Weighting Function Control, IWET [--]

This parameter indicates if the upstream  weighting function is to be used. For the
present version of code, this parameter does not affect the solution when a transient
solution is sought. If a steady-state solution is desired, one should set this parameter to
1. Thus, it is advisable to always set this  parameter to 1  for the present version of the
computer  code.
                                         Ill

-------
 5.2.2.7  Optimization Flag, IOPIYM [--]

 This parameter specifies whether the upstream weighting factor is to be optimized. This
 parameter does not affect the solution if a transient solution is sought.  For a steady-
 state solution,  it is advisable to set IOPTIM to 1. When IOPTIM is set to O, an
 upstream weighting factor of 1.0 is assumed.

 5.2.2.8  Number of Iterations for the  Nonlinear Equation, NITER [—]

 This parameter is the number of iterations allowed for solving the nonlinear equation.
 Normally, a value of NITER equal to 40 should be sufficient. If this number is exceeded
 and the solution does not converge, the  program will issue a warning message. When
 this occurs, the users should re-execute  the program using a larger value of NITER.

 5.2.2.9  Number of Times to Reset the Time  Step, NDTCHG [-]

 This parameter indicates how many times one should reset the time step size to the
 initially small time-step size.  When  we  start a computation, we normally use a small
 time-step size.   However, for every consecutive time step, we may gradually increase the
 time-step size by some amount specified by CHNG in Data Set 3 in Section 4.2.3. When
 we have a steep change in boundary conditions or in source/sink conditions, we will need
 to reset the time-step size  to the initially small value. NDTCHG tells us how many
 times we want to reset the time-step size. The value of NDTCHG must be at  least one.
 If the user does not want to reset the time step, a value of one should be entered here
 and a very large number, larger than the total simulation time, should be entered for
 TDTCH(l) in data set 4 (see Section  4.2.4).

 5.2.2.10 Number of Iterations for Pointwise  Solution, NPITER  [-]

 This parameter is used to input the number of iterations allowed for solving the matrix
 equations with the block iteration method. NPITER = 300 should be sufficient for most
 problems. If this number is exceeded and the solution does not converge, the  program
 will issue a warning message.  When this occurs, the user should first recheck the input
 values.  If the input  is correct, the program can be re-executed using a larger value for
 NPITER.

 5.2.2.11 Sorption Model Control, KSORP [--]

Although the Freundlich isotherm option can be used to simulate a linear isotherm by
 setting  the value of the exponent, n,  equal to one, it is recommended that linear
 isotherms be simulated using only the linear isotherm  option. This is because the linear
 isotherm option makes use of retarded seepage velocities, which result in a more
 accurate solution for the particle tracking scheme used in 3DLEWASTE than the pore
velocities used  in conjunction with the nonlinear adsorption models.
                                         112

-------
5.2.3 Data Set 3: Basic Real Parameters

5.2.3.1 Initial Time-Step Size, BELT [T]

This is the time-step size used for the first time-step computation if the variable CHNG
is not equal 0.0. It is the time-step size used for every time step if the variable CHNG
is set equal to 0.0. For a steady-state computation, BELT should be chosen such that no
particle travels more than one element in  one time step. For example, if an element has
a size of 10 m and the averaged velocity over this element is 0.00001 m/see, then BELT
should be less then  1,000,000 seconds.  For transient computations, one should choose a
time-step size as large as possible with BELT less than  BELX*BELWB, where BELX is
the size of the element and B is  the dispersion coefficient. For example, if the element
size is 10 m and the dispersion coefficient  is 0.00001  m2/sec, then BELT should be  less
than 10,000,000 seconds.

5.2.3.2 Fractional Change  in Time-Step Size, CHNG [—]

This parameter specifies how much of an increase one would like to make to the time-
step size for each subsequent time step. Normally, a value from  0.0 to 0.5 can be used.

5.2.3.3 Maximum Allowable  Time Step, BELMAX [T]

The maximum time-step size allowed depends on how fast the system responds to
change. Use of a value one to ten times the size of the initial time step is advised.

5.2.3.4 Maximum Simulation Time, TMAX [T]

This is the actual length of time  to be simulated.  If this time is  exceeded before you
have made NTI step computations, the  simulation will be terminated.

5.2.3.5 Relaxation Parameter for Solving the Nonlinear  Equation, OME [—]

Normally this parameter should be set  to  1.0 (see Equation 3-48). If the convergence
history shows sign of oscillation,  then a value of 0.5 should be used. If the convergence
history shows monotonic decreases but at a very slow rate, then  OME should be set to
somewhere between  1.7 to 1.9.

5.2.3.6 Iteration Parameter to Solve the Linearized Matrix Equation,  OMI [—]

Normally this parameter should be set  to  1.0 (see Equation 3-49). If the convergence
history shows signs  of oscillation,  then  set OMI to 0.5. If the solution converges
monotonically but at a very slow  rate, then set OMI to between 1.7 and 1.9.

5.2.3.7 Transient Convergence Criterion, TOLB [—]

This is the relative error allowed for assessing if a solution has converged for each  time
step. Setting TOLB  equal to 0.000001 should be sufficient for most problems.
                                         113

-------
 5.2.3.8 Steady-State Convergence Criterion, TOLA [—]

 This is the relative error allowed for assessing if a steady-state solution has converged.
 TOLA = 0.00001 should be sufficient for most problems.

 5.2.4 Data Set 5:  Material  Properties

 5.2.4.1  Distribution Coefficient [L3/M]

 Freeze and Cherry (1979) state that adsorption/desorption reactions for contaminants in
 groundwater are normally viewed as being very rapid relative to the flow velocity and
 that the amount of contaminant adsorbed is commonly a function of concentration in the
 solution. At constant  temperature and  low-to-moderate concentrations,  the functional
 relationship between the adsorbed concentration,  S (M/L3), and the dissolved concentra-
 tion, C  (M/L3), is often approximated by the  Freundlich equilibrium isotherm (Helfferich,
 1962):

                                  S  = KCn                                         (5-5)

 where  the  coefficients K and n depend on several  factors,  including the solute species
 and the nature of the porous medium. If the isotherm is linear,  n = 1, K is known as
 the distribution coefficient, Kj. The  derivation of the distribution coefficient,  which is
 different for each constituent, is discussed briefly  in Section 3.3.1.

 5.2.4.2  Bulk Density  [M/L3]

 Bulk density can be defined as the mass of a unit volume of dry soil. The  soil bulk
 density directly influences the retardation of solutes and is related to the structure and
 texture of a soil (Mercer et al., 1982).

 The bulk density of soils  typically range between 1.3 and 2.0 g/cm3, but Mercer et al.
 (1982)  state that the bulk density can be as low as 0.3  g/cm3 for soils high in organics or
 aluminum  and iron hydroxides. Representative values  for five different types of soils
 are shown in  Table 5-8. In addition, values of bulk density for a large number of soils
 can be  obtained from  the interactive computer program DBAPE,  which was discussed in
 Section 5.1.5.1.1.

The bulk density of aquifer materials may differ significantly from that of soils. There-
 fore, data on the ranges of bulk density for various geologic material are presented in
Table 5-9.  If site-specific  data are not available, the bulk  density of the saturated zone
 can be  derived using an exact relationship between porosity, particle density  and the
 bulk density (Freeze and Cherry,  1979). Assuming the particle density to  be  2.65 g/cm3,
we  can express this relationship as:

                                 pb = 2.65(1 - 0)                                   (5-6)

where
      pb     = bulk density of the soil (g/cm3)
      0      = saturated moisture content (porosity) (—)
                                          114

-------
TABLE 5-8. MEAN BULK DENSITY (g/cm3) FOR FIVE SOIL TEXTURAL
             CLASSIFICATIONS'^

Soil Texture                       Mean Value                   Range Reported
Silt Loams                            1.32                          0.86-1.67
Clay and Clay Loams                  1.30                          0.94-1.54
Sandy Loams                         1.49                          1.25-1.76
Gravelly Silt Loams                   1.22                          1.02-1.58
Loams                                1.42                          1.16-1.58
All Soils                              1.35                          0.86-1.76
a Baes,  C. F.,  Ill and R.D. Sharp. 1983. A Proposal for Estimation of
  Soil Leaching Constants for Use in Assessment Models. J. Environ.
  Qual.  12(1): 17-28 (Original reference).

"From Dean et al.  (1989)
5.2.4.3 Longitudinal and Transverse Dispersivity [L]

Hydrodynamic dispersion is a non-steady, irreversible mixing process by which a
contaminant spreads as it is transported through the subsurface. It  results from the
effects of two components: molecular diffusion and mechanical dispersion.  The larger
the hydrodynamic dispersion term is, the larger the spreading of an  initially localized
contaminant.  Molecular diffusion is discussed in  Section 5.2.4.4. Mechanical dispersion,
D, is caused by variations  in pore velocities in a soil or aquifer material. In addition,
variations in the rate of advection caused by aquifer inhomogeneity and spatially-
variable  hydraulic conductivities results in plume spreading, which is often confused
with dispersion (Keely,  1989).

Although mechanical dispersion is a second rank tensor, by assuming that  a material is
isotropic with respect to dispersion,  the dispersion tensor can be expressed in terms of
the average groundwater velocity and two constants: the longitudinal and transverse
dispersivity (see  Equation 3-20). Longitudinal dispersivity, OCL, is defined as the
characteristic mixing length in the direction of groundwater flow and lateral
dispersivityjOj., is the mixing length in the  directions  perpendicular to flow.

Values for dispersivity are difficult to determine.  Research has shown that the values
are dependent on the scale of the problem being studied (EPRI,  1985). This can be seen
in Figure 5.1. Usually,  dispersion is determined by  adjusting the dispersivity values
until modeling results match historical data (Mercer et al., 1982). Transverse
dispersivity values are commonly thought to be lower than longitudinal dispersivity
values by a factor of 3 to 20. However, recent studies suggest that transverse
dispersivity values should  be at least an order-of-magnitude smaller  than longitudinal
dispersivity values (Gelhar et al., 1992) and may even be close to zero (U.S. EPA, 1989).


                                         115

-------
TABLE 5-9. RANGE AND MEAN VALUES OF DRY BULK DENSITY FOR VARIOUS
           GEOLOGIC MATERIALS
Material
Clay
Silt
Sand, fine
Sand, medium
Sand, coarse
Gravel, fine
Gravel, medium
Gravel, coarse
Loess
Eolian sand
Till, predominantly silt
Till, predominantly sand
Till, predominantly gravel
Glacial drift, predominantly silt
Glacial drift, predominantly sand
Glacial drift, predominantly gravel
Sandstone, fine grained
Sandstone, medium grained
Siltstone
Claystone
Shale
Limestone
Dolomite
Granite, weathered
Gabbro, weathered
Basalt
Schist
Range (g/cm3)
1.18-1.72
1.01-1.79
1.13-1.99
1.27-1.93
1.42-1.94
1.60-1.99
1.47-2.09
1.69-2.08
1.25-1.62
1.33-1.70
1.61-1.91
1.69-2.12
1.72-2.12
1.11-1.66
1.36-1.83
1.47-1.78
1.34-2.32
1.50-1.86
1.35-2.12
1.37-1.60
2.20-2.72
1.21-2.69
1.83-2.20
1.21-1.78
1.67-1.77
1.99-2.89
1.42-2.69
Mean (g/cm3)
1.49
1.38
1.55
1.69
1.73
1.76
1.85
1.93
1.45
1.58
1.78
1.88
1.91
1.38
1.55
1.60
1.76
1.68
1.61
1.51
2.53
1.94
2.02
1.50
1.73
2.53
1.76
Reference: Morris and Johnson (1967); Mills et al. (1985b)
                                  116

-------
                         10'1    10"   10
                                               10'   10*    103    10c
Figure 5.1.    Longitudinal dispersivity versus scale with data classified by reliability
              (from Gelhar et al., 1992).
As initial estimates for longitudinal and transverse dispersivity, Dean et al. (19$9)
suggest the following relationships, based on values presented in the Federal Register
(1986):

                                       OL = 0.1 x,                                  (5-7 a)

                                       Op = OL/S.O                                  (5-7b)

where x, is the distance from the source to a downgradient point of interest.

5.2.4.4 Molecular Diffusion Coefficient in  Water  [L2/T]

As stated above, molecular diffusion and mechanical dispersion are both responsible for
the dispersion of solutes in groundwater systems. Molecular diffusion, which is a non-
reversible process, is typically small compared  to mechanical dispersion and is often
neglected in groundwater studies.  However, when groundwater velocities are very low,
molecular diffusion can become  significant.

The  flux of a  solute in a fluid due to molecular diffusion is described by Pick's Law,
which states that the flux is proportional  to the  concentration gradient. The coefficient
of proportionality is  called the molecular  diffusion coefficient, am. Values for the
molecular diffusion coefficient in a  fluid continuum are generally well known and are

                                           117

-------
 typically in the range of 10~9 m2/s or less at 20°C. If necessary, am, which varies with
 temperature, can be estimated from methods described in Lyman et al. (1982).

 5.2.4.5 Tortuosity [—]

 The molecular diffusion coefficient for a solute in a porous medium is smaller than the
 coefficient of diffusion in a body of water because diffusion in solids is negligible. The
 amount by which the molecular diffusion coefficient is reduced is expressed by a
 coefficient called tortuosity. Tortuosity is a second-rank  tensor which for isotropic
 conditions reduces to a scalar. It expresses the effect of  the configuration of the water
 occupying  a porous medium (Bear and Verruijt,  1987).

 De Marsily (1986)  states that a medium's tortuosity, T, can be defined as:

                                       T = 1/F<(>                                   (5-8)

 where

      F =  formation factor (the ratio  of a rock's electric  resistivity over the resistivity of
           its contained water) (—)
      (j> =  total porosity (—)

 The author states that tortuosity varies in practice  from  0.1 for clays to 0.7 for sands.
 Freeze  and Cherry (1979) state that the coefficient, which is always less than one,
 usually has a value between 0.01 and  0.5.

 Bresler (1973), as found in Dean et al. (1989), provides the following equation to
 estimate diffusion coefficient in a porous medium:

                                      Dm = Dwaebe                                  (5-9)

 where
      Dm  =  coefficient of diffusion in a porous  medium  (cm2/day)
      Dw  =  coefficient of diffusion in water (cm2/day)
      a    =  soil constant having a range of 0.001  to 0.005
      b    =  soil constant having an approximate value  of 10
      9    =  volumetric  water content (cm3/cm3)

 In the above equation, the term aebe represents an estimate of the soil's tortuosity.

 5.2.4.6  Decay Constant  [1/T]

A number  of processes, such as hydrolysis and biodegradation,  contribute to the
 disappearance of chemicals in the subsurface.  The  extent to which these processes are
 important  depends on both environmental conditions and the chemical's properties. In
 this model, the effects of individual processes on the degradation of a chemical in the
 subsurface are not considered. Instead, lumped first-order decay with respect to the
 concentration of the solute is assumed to occur, with a single first-order decay constant
 controlling the modelled rate of disappearance in each porous material.
                                          118

-------
When estimating a value for the first-order decay constant, one should determine which
processes are likely to be important at the study area. Hydrolysis is a potentially
significant elimination pathway for many organic chemicals. However, for chemicals
that readily biodegrade,  hydrolysis may be insignificant relative to biodegradation.
Methods of estimating a first-order rate constant resulting from hydrolysis are presented
in Lyman et al. (1982). Values for hydrolysis rate constants can be found in a large
number of references, including Lyman et al. (1982), Mabey et  al. (1982), and Mills et al.
(1985a).

Although biodegradation is the most significant means of removal for many organics in
the subsurface, it is a very complex and poorly understood process. Biodegradation in
the subsurface depends on a number of variable and/or unknown processes, such as the
number of microorganisms present, the availability of oxygen and other nutrients,  and
the Ph and temperature of the subsurface environment  (Sharp-Hansen et al., 1990).
Therefore it is very difficult to estimate the first-order decay coefficient resulting from
biodegradation.  Laboratory-derived biodegradation rate constants have been compiled
by Lyman et al. (1982), Mabey et al. (1982), and Mills et al. (1985a), among others.
However, these laboratory-based values may be inappropriate for field conditions.
Therefore, considerable care should be exercised if these data are used.

5.2.5  Data Set 17: Hydrological Variables

For most wellhead protection applications of this code, the velocity field and moisture
content field will not need to be specified in the input. Instead, these variables should
be calculated  and stored by the variably-saturated flow code, 3DFEMWATER. The
stored arrays of data are then accessed by 3DLEWASTE. Only when 3DLEWASTE is
executed without first running 3DFEMWATER does the user need to supply values for
these variables. Moisture content was introduced in Section  5.1.5.1.1 and will not be
discussed here.

5.2.5.1 Velocity Field [L/T]

The velocity distribution is needed to  quantify transport by  advection. Groundwater
velocities are  routinely determined indirectly using measurements of hydraulic head,
hydraulic conductivity, and Darcy's equation. For the case when  the x, y, and z axes
coincide with  the principal directions of anisotropy, Darcy's Law, in terms of the Darcy
velocity,  is written as:

                                    v^-K^dh/dx                             (5-10a)

                                    vy = -Kydh/dy                             (5-1 Ob)

                                    v^-K.dh/dz                              (5-10c)

where K,, Ky, and Kj are the hydraulic conductivity values in the x, y, and z directions,
and dh/dx, dh/dy,  and dh/dz are the hydraulic gradients in the x, y, and z directions. A
more  generalized form can be written  as:
                                         119

-------
         dh/dx - K^ dh/dy - K,z dh/dz


         dh/dX - Kyy dh/dy > Ky, dh/dZ


vz = -K,, dh/dx - K,y dh/dy - K,z dh/dz
                         Vy = -
(5-ha)

(5-llb)

(5-llc)
Since velocity depends on the gradient as well as the hydraulic conductivity, its range is
somewhat arbitrary. A range of velocities is given in Figure 5.2.



5J






c
5
3


U5
a!
< CO
c
^
5

C/J
1



g
ifl
yr

EU
H

1
O





















K q ah/it
m/s
1-

10-L

10 i

,o-l





»i

10-t-
io-7-


10-*-
10-?-
10-!°-

10-"-
10-12-

ft/d mis
I ~
- 10'
104

10 4-
- 103
10*-
_ 102

•
10 —
- 10 '
10 -ie-
_ 1

10 "U
-10-'
10-"-
_io-2

m/s
"I T <
_M4
ft/d " l0"1
_102




_ 1C'2




-- io-2

-- 10-3





= I2

ft/y
_10^






_ 10-3
-»-
_io-5
^

- 10-7
     Figure 5.2. Nomograph for determining Darcy velocity (from  Mercer et al.
                  1982)
                                         120

-------
                                     SECTION 6

                               EXAMPLE  PROBLEMS
6.1  3DFEMWATER

To demonstrate the application of 3DFEMWATER, three simple example problems are
presented. These three problems represent one-, two-, and three-dimensional
applications, respectively. For  each problem, a brief description  and a correctly-
constructed input data set are given. The corresponding output is not included in this
documentation. Rather, it is distributed along with the code by the EPA Center for
Exposure Assessment  Modeling (CEAM) at the  Environmental Research Laboratory in
Athens, Georgia. See Section 2 for information about  obtaining the code.

6.1.1 One-Dimensional Column

One-dimensional transient flow through a column is  simulated in this example. The
column is 200 cm long and  is 50 cm by 50 cm in cross-section (Figure 6.1). The soil in
the column is assumed to be a sandy clay loam  which has a saturated hydraulic
conductivity of 31.4 cm/d, a porosity of 0.39 and a residual moisture content  of 0.10. The
unsaturated characteristic hydraulic properties of the  soil  in the column are represented
by the van Genuchten analytical functions with the empirical coefficient alpha equal to
0.059 and the empirical coefficient beta equal to 1.48.

The initial conditions assumed are a pressure head of -90.0 cm imposed on the top
surface of the column, 0.0 cm on the  bottom surface of the column, and -97.0 cm
elsewhere. The boundary conditions are as follows. No flux is imposed on the  left,
front, right, and back surfaces  of the  column (this is done  automatically by the code).
Pressure head is held  at 0.0 cm on the bottom surface using a Dirichlet boundary
condition. A variable boundary condition is used on the top surface of the column with a
pending depth of zero, minimum pressure of -90.0 cm, a rainfall  of 5.0 cm/d for  the first
10 days, and a potential evaporation of 5.0 cm/d for the second 10 days.

The region of interest,  that is, the whole column, is discretized with 1 x 1 x 40 = 40
elements with the element size equal to 50 x 50 x 5 cm. This results in  2x2x41 =  164
node points.  For this simulation, each of the four vertical lines is considered  a
subregion. Thus,  a  total of four subregions, each with  41 node points, is used for the
subregional block iteration simulation.

A  variable time step size is  used. The initial time step size is 0.05 days, and  each
subsequent time step size is increased by 0.2 times with a maximum time step
                                         121

-------
                                                           200cm
           Figure 6.1. One-dimensional transient flow through a soil column.
size not greater than 1.0 d. Because there is an abrupt change in the flux value
from 5 cm/d (infiltration) to -5 cm/d (evaporation) imposed on the top surface at day 10,
the time step  size is automatically reset to 0.05 d on the tenth day. Because a 20-day
simulation is to be made, 44 time steps are needed.

A pressure head tolerance of 0.02 cm is selected for the nonlinear iteration and a
tolerance  of 0.01 cm is used for the block iteration. The relaxation factors for both the
nonlinear iteration and block  iteration are set equal  to 0.5.

The input data set for this problem,  prepared according to the instructions in Sections
4.1  and 5.1, is shown in  Table 6-1.

6.1.2 Two-dimensional Drainage  Problem

Two-dimensional steady-state flow is simulated in this problem. The region of interest
is bounded on the left and right by parallel drains which fully penetrate the medium.
The bottom is an impervious  layer and the top is an air-soil interface (Figure 6.2). The
distance between the two drains is 20 m. The medium is assumed to have a saturated
horizontal hydraulic conductivity of 0.31 m/d and vertical hydraulic conductivity of 0.12
m/d, a porosity of 0.39, and a field capacity of 0.10.  The unsaturated characteristic
                                         122

-------
 TABLE 6-1.   INPUT DATA SET FOR THE ONE-DIMENSIONAL 3DFEMWATER
              PROBLEM
  1 SIMULATION OF ONE-D COLUMN INFILTRATION-EVAPORATION; L=CM, T=DAY, M=G Oil
C ******* DATA SET 2: BASIC INTEGERS
  164    40   1     0     44     1     61     0      0    50    20     3     100
C ******* DATA SET 3: BASIC REAL PARAMETERS
   0.05DO   0.2DO   l.ODO   20.0DO   2.0D-2    2.0D-2    l.ODO    7.316D12
  1.1232D2    l.ODO   0.5DO   0.5DO
C ******* DATA SET 4: PRINTER, STORAGE CONTROL AND TIME STEP SIZE RESETTING
333030300030003003000033303030003000300300003
111010100010001001000011101010001000100100001
  1.0D01     2.0000D1  1.0D38
C ******* DATA SET 5: MATERIAL PROPERTIES
   O.ODO     O.ODO    31.40DO    O.ODO     O.ODO    O.ODO
C ******* DATA SET 6: SOIL PROPERTY PARAMETERS
  050
  0.100DO     0.390DO     O.OODO      0.059DO       1.48DO              THPROP
C ******* DATA SET 7: NODE COORDINATES
   1
   42
   83
  124
   0
40
 40
 40
 40
 0
1
1
1
1
0
 O.ODO
 O.ODO
50.0DO
50.0DO
 0.0
50.0DO
O.ODO
O.ODO
50.0DO
0.0
O.ODO
O.ODO
O.ODO
O.ODO
0.0
O.ODO
O.ODO
O.ODO
O.ODO
0.0
O.ODO
O.ODO
O.ODO
O.ODO
0.0
5.0DO
5.0DO
5.0DO
5.0DO
0.0
C ******* DATA SET 8: SUBREGIONAL DATA
   4
1
0
1
0
1
0
1
0
1
0
3
0
40
0
40
0
40
0
40
0
1
0
1
0
1
0
1
0
1
0
41
0
1
0
42
0
83
0
124
0
0
0
1
0
1
0
1
0
1
0
                                              END OF NNPLR(K)

                                              END OF GNLR(I,1)

                                              END OF GNLR(I,2)

                                              END OF GNLR(I,3)

                                              END OF GNLR(I,4)
C ******* DATA SET 9: ELEMENT INCIDENCES
   1   39    1    42    83    124   1   43
   00000     000
C ******* DATA SET 11: INITIAL CONDITIONS
                                    84
                                    0
                                   125
                                    0
                                    1
                                    0
                                   END OF IE
1
2
43
84
125
41
0
3
38
38
38
38
3
0
41
1
1
1
1
41
0
O.ODO
-9.70D1
-9.70D1
-9.70D1
-9.70D1
-9.00D1
0.0
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
0.0
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
0.0
                                                     END OF 1C
C ******* DATA SET 12: SOURCE/SINK AND B. C. CONTROL INTEGERS
   00000000412
   14140000000
                                                       0
                                                       0
                                       123

-------
TABLE 6-1. INPUT DATA SET FOR THE ONE-DIMENSIONAL 3DFEMWATER PROBLEM
            (concluded)
C ******* DATA SET 14: VARIABLE BOUNDARY CONDITIONS
O.ODO
1
0
1
0
1
0
1
0
1
0
*******
O.ODO
1
1
0
0

0
0
0
0
3
0
3
0
3
0
DATA

42
3
0

5.0DO
0
0
0
0
1
0
1
0
1
0
SET 15:
O.ODO
83
1
0

10.
1
0
82
0
41
0



0.0
ODO 5.0DO 10.001DO -5.0DO 1.0D38 -5.(
0
0
123 164
0 0
41
0
O.ODO
0.0
-90.0DO
0.0



END OF IRTYP
41 0
0 0

END
O.ODO
0.0
O.ODO
0.0
0 0
0 0

OFNPVB
0.0
0.0
0.0

0
0 END OF ISV(J,I) ,



END OF HCON

END OF HMIN
DIRICHLET BOUNDARY CONDITIONS

124
1
0
T5TSTT)
1.0D38

0
0
OF .TOR
O.ODO







END OF




IDTYP

                                                                            000
                          -10m
                                          •10m
            Figure 6.2. Two-dimensional steady-state flow to parallel drains.
                                      124

-------
hydraulic properties of the medium are given by the van Genuchten analytical
functions with the empirical coefficient alpha equal to 0.059 and the empirical
coefficient beta equal to 1.48.

Because of symmetry, the region to be simulated is 0.0< x <  10.0 m and 0.0< z <  10.0
m, with a width of 10 m assumed in the y-direction. A no flux boundary is imposed on
the left (x = 0.0), front (y = 0.0), back (y = 10.0), and bottom (z = 0.0) sides of the
region.  Pressure head is assumed to vary from zero at the water surface (z = 2.0) to
2.0 m at the bottom (z = 0.0) on the right side  (x = 10.0). Variable conditions are used
elsewhere. Pending depth is assumed to be  zero meters on the whole variable
boundary. Fluxes on the top of the variable boundary are assumed equal to 0.006 m/d
and on the right side, above the water surface,  are equal to zero. A pre-initial
condition for the steady-state solution is set as h = 10- z.

The region of interest  is discretized with lOx 1  x 10 =  100 elements with the element
size equal to  1 x 10 x 1 cm. This results in  11x2x11= 242 nodal points. Each of the
two vertical planes is  considered a subregion. Thus, a total of two subregions, each
with  121  node points,  is used for the subregional block iteration simulation.

A pressure head tolerance of 0.002 m is set for  the for nonlinear iteration and a value
of 0.001 m is  used for the block iteration. The relaxation factors for both the nonlinear
iteration and  block iteration are set equal to 0.5.

The input data set for this problem, prepared according to the instructions in Sections
4.1 and 5.1, is shown in Table 6-2.

6.1.3 Three-Dimensional Pumping Problem

Three-dimensional  steady-state flow to a pumping well  is simulated in this  problem.
The region of interest  is bounded on the left and right by hydraulically connected
rivers; on the front, back, and bottom by impervious confining beds;  and on the  top  by
an air-soil interface (Figure 6.3).  A pumping well is located at (x,y) = (540,400)  in
Figure 6.3. Initially, the water table is assumed to be horizontal and is 60 m above the
bottom  of the  aquifer.  The water level at the well is then lowered to a height of 30  m.
This height is  held until a  steady state condition is reached. The medium in the region
is assumed to be anisotropic and has saturated  hydraulic conductivity components K^
= 0.31 m/d, Kyy = 0.03  m/d,  and K^z = 0.12 m/d.  The porosity of the medium is 0.10  and
the field capacity is 0.39. The unsaturated characteristic hydraulic properties of the
medium are given by the van Genuchten analytical functions with the empirical
coefficient alpha equal to 0.059 and the empirical coefficient beta equal to 1.48.

Because of symmetry,  the region to be simulated is taken as 0 < x < 1000 m, 0 < y <
400 m,  and 0  < z < 72 m. Two types of boundary conditions are used. Pressure head
is assumed hydrostatic on two vertical planes.  The first is located at x = 0 and 0 <  z
< 60 and the second, at x = 1000  and 0 < z < 60. A no flux boundary is imposed on all
other boundaries of the  flow regime.  The pre-initial condition for the steady-state
solution is set so that  the pressure head, h = 60- z.
                                         125

-------
TABLE 6-2.    INPUT DATA SET FOR THE TWO-DIMENSIONAL 3DFEMWATER
              PROBLEM
  2 SIMULATION OF TWO-D STEADY DRAINAGE; L=M, T=DAY, M=KG                111
C ******* DATA SET 2:  BASIC INTEGERS
  242   100  1      0     00    6    1    0    0    50    20    1     100
C ******* DATA SET 3:  BASIC REAL PARAMETERS
  0.05DO      0.2DO    l.ODO    20.0DO    2.0D-3    2.0D-3    l.ODO    7.316D10
  1.1232D4    l.ODO    0.5DO    0.5DO     0.0
C ******* DATA SET 4:  PRINTER, STORAGE CONTROL AND TIME STEP SIZE RESETTING
33
11
  1.0D38
C ******* DATA SET 5:  MATERIAL PROPERTIES
  0.31DO      O.ODO    0.12DO   O.ODO     O.ODO    O.ODO
C ******* DATA SET 6:  SOIL PROPERTY PARAMETERS
  050
  0.100DO     0.390DO  O.OODO   0.059DO   1.48DO             THPROP
C ******* DATA SET 7:  NODE COORDINATES
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           O.ODO     O.ODO
                                                           0.0       0.0
1
2
3
4
5
6
7
8
9
10
11
122
123
124
125
126
127
128
129
130
131
132
0
Q *******
2
1
0
1
0
1
0
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
0
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
0
DATA SET 8:

1
0
120
0
120
0

1
0
1
0
1
0
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
0.0
SUBREGIONAL

121 0
0 0
1 1
0 0
122 1
0 0
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
10.0DO
10.0DO
10.0DO
10.0DO
10.0DO
10.0DO
10.0DO
10.0DO
10.0DO
10.0DO
10.0DO
0.0
DATA


END OF

END OF

END OF
O.ODO
l.ODO
2.0DO
3.0DO
4.0DO
5.0DO
6.0DO
7.0DO
8.0DO
9.0DO
10.0DO
O.ODO
l.ODO
2.0DO
3.0DO
4.0DO
5.0DO
6.0DO
7.0DO
8.0DO
9.0DO
10.0DO
0.0



NNPLR(K)

GNLR(I,1)

GNLR(I,2)
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
l.ODO
0.0








                                      126

-------
TABLE 6-2.    INPUT DATA SET FOR THE TWO-DIMENSIONAL 3DFEMWATER
             PROBLEM (continued)
C ******* DATA SET
1
11
21
31
41
51
61
71
81
91
0
IE
9
9
9
9
9
9
9
9
9
9
0

1
1
1
1
1
1
1
1
1
1
0

C ******* DATA SET
1
2
3
4
5
6
7
8
9
10
11
122
123
124
125
126
127
128
129
130
131
132
0
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
0
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
0
C ******* DATA SET
0
18
0
38
0
2
9: ELEMENT INCIDENCES
1
12
23
34
45
56
67
78
89
100
0

12
23
34
45
56
67
78
89
100
111
0

133
144
155
166
177
188
199
210
221
232
0

122
133
144
155
166
177
188
199
210
221
0

2
13
24
35
46
57
68
79
90
101
0

13
24
35
46
57
68
79
90
101
112
0

134 123 1
145 134 1
156 145 1
167 156 1
178 167 1
189 178 1
200 189 1
211 200 1
222 211 1
233 222 1
000 END OF

11: INITITAL CONDITIONS
10.0DO
9.0DO
8.0DO
7.0DO
6.0DO
5.0DO
4.0DO
3.0DO
2.0DO
l.ODO
O.ODO
10.0DO
9.0DO
8.0DO
7.0DO
6.0DO
5.0DO
4.0DO
3.0DO
2.0DO
l.ODO
O.ODO
O.ODO























O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
12: SOURCE/SINK AND
0 0
2 0
0
0
0
0























B.C.
0
0
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
CONTROL













































END OF 1C
INTEGERS
6 1
0 0
2 0
00000
                                  127

-------
TABLE 6-2. INPUT DATA SET FOR THE TWO-DIMENSIONAL 3DFEMWATER
            PROBLEM  (concluded)
C ******* DATA SET 14: VARIABLE BOUNDARY CONDITIONS
O.ODO
O.ODO
1
11
0
1
11
0
1
12
20
31
0
1
0
1
0
*******
9
7
0
9
7
0
10
7
10
7
0
37
0
37
0
6.0D-3
O.ODOO
1 1
1 2
0 0
1 11
1 120
0 0
1 11
1 120
1 132
1 241
0 0
1
0
1
0
DATA SET 15:
O.ODO
111
1
0
n
112
5
0
2.0DO
113 232
1 1
0 0
1.0D38
1.0D38
0
0
0
22 143
241 242
0 0
11
-1
11
-1
0
O.ODO
O.ODO
-90.0D2
O.ODO
6.0D-3
O.ODOO


132
121
0




END
O.ODO
O.ODO
O.ODO
O.ODO

END OF IRTYP
11 11 11 11
-1 -1 -1 -1
0000 END OF ISV(J,I) J=l,4




OFNPVB
O.ODO
O.ODO END OF HCON
O.ODO
O.ODO END OF HMIN
DIRICHLET BOUNDARY CONDITIONS
1.0D38
233 234
0
0 END OF
2.0DO





IDTYP
F.Nn OF .TOR 	
	 nnn
                                       *- 460m—H
          Figure 6.3. Three-dimensional steady-state flow to a pumping well.
                                    128

-------
The region of interest is discretized with 20 x 8 x  10 = 1600 elements, resulting in 21 x
9x11= 2079 nodal points. The nodes are located at x = 0, 70,  120, 160, 200,  275, 350,
400, 450, 500, 540, 570, 600, 650, 750, 800, 850, 900, 950, and 1000 in the x-direction,
and at z = 0,  15, 30,  35, 40, 45, 50, 55, 60, 66, and 72 in the z-direction.  In the y-
direction, nodes are spaced evenly at Az = 50 m. For the simulation, each of the nine
vertical planes perpendicular to the y-axis is considered a subregion. Thus, a  total of 9
subregions,  each with 231 node points, is  used for the subregional block iteration
simulation.

The pressure head tolerance is set at  .01 m for the nonlinear iteration and is .005 m
for the block iteration. The relaxation factors for the nonlinear iteration and block
iteration are set equal to 1.0 and 1.5,  respectively.

The input data set for this problem, prepared according to the instructions in  Sections
4.1 and 5.1, is shown in Table 6-3.
                                         129

-------
TABLE 6-3. INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
           PROBLEM
  3 SIMULATION OF THREE-D PUMPING WELL; L = M, T = DAY, M = KG
C ******* DATA SET 2: BASIC INTEGER PARAMETERS
 2079 1600  1  0  0  0  6   1  0  0  50 20  1 100
C ******* DATA SET 3: BASIC REAL PARAMETERS
     0.05DO   O.ODO  l.ODO 20.0DO   l.OD-2  l.OD-2 l.ODO  7.316D10
     1. 1232D4 l.ODO  l.ODO  1.5DO
C ******* DATA SET 4: PRINTER, STORAGE CONTROL AND TIME STEP SIZE RESETTING
55
11
     1.0D38
C ******* DATA SET 5: MATERIAL PROPERTIES
     0.31DO 0.03DO    0.12DO  O.ODO   O.ODO  O.ODO
C ******* DATA SET 6: SOIL PROPERTY PARAMETERS
     050
     0.1000DO    0.390DO   O.OODO   0.059DO   1.48DO         THPROP
C ******* DATA SET 7: NODE COORDINATES
Oil
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.70D+02
0.70D+02
0.70D+02
0.70D+02
0.70D+02
0.70D+02
0.70D+02
0.70D+02
0.70D+02
0.70D+02
0.70D+02
0.12D+03
0.12D+03
0.12D+03
0.12D+03
0.12D+03
0.12D+03
0.12D+03
0.12D+03
0.12D+03
0.12D+03
0.12D+03
0.16D+03
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0. OOD+00
O.OOD+00
0. OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0. OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0. OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0. 15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0. 15D+00
0.30D+00
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
O.OOD+00
0. OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0. OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0. OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0. OOD+00
                                     130

-------
TABLE 6-3. INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
          PROBLEM (continued)
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
0.16D+03
0.16D+03
0.16D+03
0.16D+03
0.16D+03
0. 16D+03
0. 16D+03
0.16D+03
0. 16D+03
0. 16D+03
0.20D+03
0.20D+03
0.20D+03
0.20D+03
0.20D+03
0.20D+03
0.20D+03
0.20D+03
0.20D+03
0.20D+03
0.20D+03
0.28D+03
0.28D+03
0.28D+03
0.28D+03
0.28D+03
0.28D+03
0.28D+03
0.28D+03
0.28D+03
0.28D+03
0.28D+03
0.35D+03
0.35D+03
0.35D+03
0.35D+03
0.35D+03
0.35D+03
0.35D+03
0.35D+03
0.35D+03
0.35D+03
0.35D+03
0.40D+03
0.40D+03
0.40D+03
0.40D+03
0.40D+03
0.40D+03
0.40D+03
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
                                  131

-------
TABLE 6-3. INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
          PROBLEM (continued)
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
0.40D+03
0.40D+03
0.40D+03
0.40D+03
0.45D+03
0.45D+03
0.45D+03
0.45D+03
0.45D+03
0.45D+03
0.45D+03
0.45D+03
0.45D+03
0.45D+03
0.45D+03
0.50D+03
0.50D+03
0.50D+03
0.50D+03
0.50D+03
0.50D+03
0.50D+03
0.50D+03
0.50D+03
0.50D+03
0.50D+03
0.54D+03
0.54D+03
0.54D+03
0.54D+03
0.54D+03
0.54D+03
0.54D+03
0.54D+03
0.54D+03
0.54D+03
0.54D+03
0.57D+03
0.57D+03
0.57D+03
0.57D+03
0.57D+03
0.57D+03
0.57D+03
0.57D+03
0.57D+03
0.57D+03
0.57D+03
0.60D+03
0.60D+03
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
                                  132

-------
TABLE 6-3. INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
          PROBLEM (continued)
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
0.60D+03
0.60D+03
0.60D+03
0.60D+03
0.60D+03
0.60D+03
0.60D+03
0.60D+03
0.60D+03
0.65D+03
0.65D+03
0.65D+03
0.65D+03
0.65D+03
0.65D+03
0.65D+03
0.65D+03
0.65D+03
0.65D+03
0.65D+03
0.70D+03
0.70D+03
0.70D+03
0.70D+03
0.70D+03
0.70D+03
0.70D+03
0.70D+03
0.70D+03
0.70D+03
0.70D+03
0.75D+03
0.75D+03
0.75D+03
0.75D+03
0.75D+03
0.75D+03
0.75D+03
0.75D+03
0.75D+03
0.75D+03
0.75D+03
0.80D+03
0.80D+03
0.80D+03
0.80D+03
0.80D+03
0.80D+03
0.80D+03
0.80D+03
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
                                  133

-------
TABLE 6-3. INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
          PROBLEM  (continued)
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
0
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
0
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
0
0.80D+03
0.80D+03
0.80D+03
0.85D+03
0.85D+03
0.85D+03
0.85D+03
0.85D+03
0.85D+03
0.85D+03
0.85D+03
0.85D+03
0.85D+03
0.85D+03
0.90D+03
0.90D+03
0.90D+03
0.90D+03
0.90D+03
0.90D+03
0.90D+03
0.90D+03
0.90D+03
0.90D+03
0.90D+03
0.95D+03
0.95D+03
0.95D+03
0.95D+03
0.95D+03
0.95D+03
0.95D+03
0.95D+03
0.95D+03
0.95D+03
0.95D+03
0.10D+04
0.10D+04
0.10D+04
0.10D+04
0.10D+04
0.10D+04
0.10D+04
0.10D+04
0.10D+04
0.10D+04
0.10D+04
0.0
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00

0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
O.OOD+00
0.15D+02
0.30D+02
0.35D+02
0.40D+02
0.45D+02
0.50D+02
0.55D+02
0.60D+02
0.66D+02
0.72D+02
0.0
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.0
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.50D+02
0.0
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.00.0
                                  134

-------
TABLE 6-3. INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
          PROBLEM (continued)
C ******* DATA SET 8: SUBREGIONAL DATA
9
1 8 1 231 0
0000 0
1 230 1 1 1
0000 0
1 230 1 232 1
0000 0
1 230 1 463 1
0000 0
1 230 1 694 1
0000 0
1 230 1 925 1
0000 0
1 230 1 1156 1
0000 0
1 230 1 1387 1
0000 0
1 230 1 1618 1
0000 0
1 230 1 1849 1
0000 0
C ******** DATA SET
1
11
21
31
41
51
61
71
81
91
101
111
121
131
141
151
161
171
181
191
201
211
221
231
241
251
261
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
END OF NNPLR(9)
END OF GNLRd.l)
END OF GNLR(I,2)
END OF GNLR(I,3)
END OF GNLR(I,4)
END OF GNLR(I,5)
END OF GNLR(I,6)
END OF GNLR(I,7)
END OF GNLR(I,8)
END OF GNLR(I,9)
10: ELEMENT INCIDENCES
1
12
23
34
45
56
67
78
89
100
111
122
133
144
155
166
177
188
199
210
232
243
254
265
276
287
298
12
23
34
45
56
67
78
89
100
111
122
133
144
155
166
177
188
199
210
221
243
254
265
276
287
298
309
243
254
265
276
287
298
309
320
331
342
353
364
375
386
397
408
419
430
441
452
474
485
496
507
518
529
540
232
243
254
265
276
287
298
309
320
331
342
353
364
375
386
397
408
419
430
441
463
474
485
496
507
518
529
2
13
24
35
46
57
68
79
90
101
112
123
134
145
156
167
178
189
200
211
233
244
255
266
277
288
299
13
24
35
46
57
68
79
90
101
112
123
134
145
156
167
178
189
200
211
222
244
255
266
277
288
299
310
244
255
266
277
288
299
310
321
332
343
354
365
376
387
398
409
420
431
442
453
475
486
497
508
519
530
541
233
244
255
266
277
288
299
310
321
332
343
354
365
376
387
398
409
420
431
442
464
475
486
497
508
519
530
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
                                  135

-------
TABLE 6-3.    INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
            PROBLEM (continued)
271
281
291
301
311
321
331
341
351
361
371
381
391
401
411
421
431
441
451
461
471
481
491
501
511
521
531
541
551
561
571
581
591
601
611
621
631
641
651
661
671
681
691
701
711
721
731
741
751
761
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
309
320
331
342
353
364
375
386
397
408
419
430
441
463
474
485
496
507
518
529
540
551
562
573
584
595
606
617
628
639
650
661
672
694
705
716
727
738
749
760
771
782
793
804
815
826
837
848
859
870
320
331
342
353
364
375
386
397
408
419
430
441
452
474
485
496
507
518
529
540
551
562
573
584
595
606
617
628
639
650
661
672
683
705
716
727
738
749
760
771
782
793
804
815
826
837
848
859
870
881
551
562
573
584
595
606
617
628
639
650
661
672
683
705
716
727
738
749
760
771
782
793
804
815
826
837
848
859
870
881
892
903
914
936
947
958
969
980
991
1002
1013
1024
1035
1046
1057
1068
1079
1090
1101
1112
540
551
562
573
584
595
606
617
628
639
650
661
672
694
705
716
727
738
749
760
771
782
793
804
815
826
837
848
859
870
881
892
903
925
936
947
958
969
980
991
1002
1013
1024
1035
1046
1057
1068
1079
1090
1101
310
321
332
343
354
365
376
387
398
409
420
431
442
464
475
486
497
508
519
530
541
552
563
574
585
596
607
618
629
640
651
662
673
695
706
717
728
739
750
761
772
783
794
805
816
827
838
849
860
871
321
332
343
354
365
376
387
398
409
420
431
442
453
475
486
497
508
519
530
541
552
563
574
585
596
607
618
629
640
651
662
673
684
706
717
728
739
750
761
772
783
794
805
816
827
838
849
860
871
882
552
563
574
585
596
607
618
629
640
651
662
673
684
706
717
728
739
750
761
772
783
794
805
816
827
838
849
860
871
882
893
904
915
937
948
959
970
981
992
1003
1014
1025
1036
1047
1058
1069
1080
1091
1102
1113
541
552
563
574
585
596
607
618
629
640
651
662
673
695
706
717
728
739
750
761
772
783
794
805
816
827
838
849
860
871
882
893
904
926
937
948
959
970
981
992
1003
1014
1025
1036
1047
1058
1069
1080
1091
1102
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
                                  136

-------
TABLE 6-3.   INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
            PROBLEM  (continued)
771
781
791
801
811
821
831
841
851
861
871
881
891
901
911
921
931
941
951
961
971
981
991
1001
1011
1021
1031
1041
1051
1061
1071
1081
1091
1101
1111
1121
1131
1141
1151
1161
1171
1181
1191
1201
1211
1221
1231
1241
1251
1261
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
881
892
903
925
936
947
958
969
980
991
1002
1013
1024
1035
1046
1057
1068
1079
1090
1101
1112
1123
1134
1156
1167
1178
1189
1200
1211
1222
1233
1244
1255
1266
1277
1288
1299
1310
1321
1332
1343
1354
1365
1387
1398
1409
1420
1431
1442
1453
892
903
914
936
947
958
969
980
991
1002
1013
1024
1035
1046
1057
1068
1079
1090
1101
1112
1123
1134
1145
1167
1178
1189
1200
1211
1222
1233
1244
1255
1266
1277
1288
1299
1310
1321
1332
1343
1354
1365
1376
1398
1409
1420
1431
1442
1453
1464
1123
1134
1145
1167
1178
1189
1200
1211
1222
1233
1244
1255
1266
1277
1288
1299
1310
1321
1332
1343
1354
1365
1376
1398
1409
1420
1431
1442
1453
1464
1475
1486
1497
1508
1519
1530
1541
1552
1563
1574
1585
1596
1607
1629
1640
1651
1662
1673
1684
1695
1112
1123
1134
1156
1167
1178
1189
1200
1211
1222
1233
1244
1255
1266
1277
1288
1299
1310
1321
1332
1343
1354
1365
1387
1398
1409
1420
1431
1442
1453
1464
1475
1486
1497
1508
1519
1530
1541
1552
1563
1574
1585
1596
1618
1629
1640
1651
1662
1673
1684
882
893
904
926
937
948
959
970
981
992
1003
1014
1025
1036
1047
1058
1069
1080
1091
1102
1113
1124
1135
1157
1168
1179
1190
1201
1212
1223
1234
1245
1256
1267
1278
1289
1300
1311
1322
1333
1344
1355
1366
1388
1399
1410
1421
1432
1443
1454
893
904
915
937
948
959
970
981
992
1003
1014
1025
1036
1047
1058
1069
1080
1091
1102
1113
1124
1135
1146
1168
1179
1190
1201
1212
1223
1234
1245
1256
1267
1278
1289
1300
1311
1322
1333
1344
1355
1366
1377
1399
1410
1421
1432
1443
1454
1465
1124
1135
1146
1168
1179
1190
1201
1212
1223
1234
1245
1256
1267
1278
1289
1300
1311
1322
1333
1344
1355
1366
1377
1399
1410
1421
1432
1443
1454
1465
1476
1487
1498
1509
1520
1531
1542
1553
1564
1575
1586
1597
1608
1630
1641
1652
1663
1674
1685
1696
1113
1124
1135
1157
1168
1179
1190
1201
1212
1223
1234
1245
1256
1267
1278
1289
1300
1311
1322
1333
1344
1355
1366
1388
1399
1410
1421
1432
1443
1454
1465
1476
1487
1498
1509
1520
1531
1542
1553
1564
1575
1586
1597
1619
1630
1641
1652
1663
1674
1685
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
                                  137

-------
TABLE 6-3.    INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
             PROBLEM  (continued)
1271
1281
1291
1301
1311
1321
1331
1341
1351
1361
1371
1381
1391
1401
1411
1421
1431
1441
1451
1461
1471
1481
1491
1501
1511
1521
1531
1541
1551
1561
1571
1581
1591
0
Q *******
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
0
DATA
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
SET 11
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
1464
1475
1486
1497
1508
1519
1530
1541
1552
1563
1574
1585
1596
1618
1629
1640
1651
1662
1673
1684
1695
1706
1717
1728
1739
1750
1761
1772
1783
1794
1805
1816
1827
0
1475
1486
1497
1508
1519
1530
1541
1552
1563
1574
1585
1596
1607
1629
1640
1651
1662
1673
1684
1695
1706
1717
1728
1739
1750
1761
1772
1783
1794
1805
1816
1827
1838
0
1706
1717
1728
1739
1750
1761
1772
1783
1794
1805
1816
1827
1838
1860
1871
1882
1893
1904
1915
1926
1937
1948
1959
1970
1981
1992
2003
2014
2025
2036
2047
2058
2069
0
1695
1706
1717
1728
1739
1750
1761
1772
1783
1794
1805
1816
1827
1849
1860
1871
1882
1893
1904
1915
1926
1937
1948
1959
1970
1981
1992
2003
2014
2025
2036
2047
2058
0
1465
1476
1487
1498
1509
1520
1531
1542
1553
1564
1575
1586
1597
1619
1630
1641
1652
1663
1674
1685
1696
1707
1718
1729
1740
1751
1762
1773
1784
1795
1806
1817
1828
0
1476
1487
1498
1509
1520
1531
1542
1553
1564
1575
1586
1597
1608
1630
1641
1652
1663
1674
1685
1696
1707
1718
1729
1740
1751
1762
1773
1784
1795
1806
1817
1828
1839
0
1707
1718
1729
1740
1751
1762
1773
1784
1795
1806
1817
1828
1839
1861
1872
1883
1894
1905
1916
1927
1938
1949
1960
1971
1982
1993
2004
2015
2026
2037
2048
2059
2070
0
1696
1707
1718
1729
1740
1751
1762
1773
1784
1795
1806
1817
1828
1850
1861
1872
1883
1894
1905
1916
1927
1938
1949
1960
1971
1982
1993
2004
2015
2026
2037
2048
2059
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0 END OF IE
: INITIAL CONDITIONS









0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02




0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.
0.
0.
OOD+00
OOD+00
OOD+00
O.OOD+00
0.
OOD+00
O.OOD+00
0.
0.
0.
0.
0.
0.
OOD+00
OOD+00
OOD+00
OOD+00
OOD+00
OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0




























































                                  138

-------
TABLE 6-3. INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
          PROBLEM (continued)
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
231
231
231
231
231
231 '
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
                                  139

-------
TABLE 6-3. INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
          PROBLEM (continued)
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
                                  140

-------
TABLE 6-3. INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
          PROBLEM  (continued)
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
                                  141

-------
TABLE 6-3. INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
          PROBLEM (continued)
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
231
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
0.10D+02
0.50D+01
O.OOD+00
-0.60D+01
-0.12D+02
0.60D+02
0.45D+02
0.30D+02
0.25D+02
0.20D+02
0.15D+02
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
O.OOD+00
0.0
0.0
0.0
0.0
0.0
0.0-
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
                                  142

-------
TABLE 6-3. INPUT DATA SET FOR THE THREE-DIMENSIONAL 3DFEMWATER
          PROBLEM (continued)
216 8 231 0.10D+02 O.OOD+00 0.0
217 8 231 0.50D+01 O.OOD+00 0.0
218 8 231 O.OOD+00 O.OOD+00 0.0
219 8 231 -0.60D+01 O.OOD+00 0.0
220 8 231 -0.12D+02 O.OOD+00 0.0
221 8 231 0.60D+02 O.OOD+00 0.0
222 8 231 0.45D+02 O.OOD+00 0.0
223 8 231 0.30D+02 O.OOD+00 0.0
224 8 231 0.25D+02 O.OOD+00 0.0
225 8 231 0.20D+02 O.OOD+00 0.0
226 8 231 0.15D+02 O.OOD+00 0.0
227 8 231 0.10D+02 O.OOD+00 0.0
228 8 231 0.50D+01 O.OOD+00 0.0
229 8 231 O.OOD+00 O.OOD+00 0.0
230 8 231 -0.60D+01 O.OOD+00 0.0
231 8 231 -0.12D+02 O.OOD+00 0.0
0 0 0 0.0 0.0 0.0
C ******* DATA SET 12: SOURCE/SINK AND B. C. CONTROL
000
000
C ******* DATA SET
O.ODO 60.0DO
O.ODO 30.0DO
123
239 240 463
699 700 701
1159 1160 1161
1619 1620 1621
1857 221 222
458 459 460
918 919 920
1378 1379 1380
1838 1839 1840
2076 2077 111
1 161 1
163 2 1
000
0
0
0
0
0 0
0 0
0
0
15: DIRICHLET BOUNDARY
1.
1.
4
464
702
OD38
OD38
5
465
925
1162 1163
1622
223
683
921
1381
1841
112
1
2
0
1623
224
684
922
1382
1842
113
0
0
0
60.0DO
30.0DO
6 7
466 467
926 927
1164 1387
1624 1625
225 226
685 686
1145 1146
1383 1384
1843 1844





8
468
928
1388
1626
227
687
1147
1607
1845



165
0

END OF 1C
INTEGERS
2
0
2
0
0
0

0

0 0
CONDITIONS


9
469
929
1389
1849
228
688
1148
1608
1846





232
470
930
1390
1850
229
689
1149
1609
2069





233
471
931
1391
1851
452
690
1150
1610
2070





234
694
932
1392
1852
453
691
1151
1611
2071





235
695
933
1393
1853
454
914
1152
1612
2072





236
696
1156
1394
1854
455
915
1153
1613
2073





237 238
697 698
1157 1158
1395 1618
1855 1856
456 457
916 917
1376 1377
1614 1615
2074 2075



END OF IDTYP
o pvwn OF .TOR 	
nno



                                  143

-------
6.2  3DLEWASTE

To demonstrate the application of 3DLEWASTE, two simple example problems are
presented. For each problem, a brief description and a correctly-constructed input data
set are given. The corresponding output is not included  in this documentation. Rather,  it
is distributed along with the code by the EPA Center for Exposure Assessment Modeling
(CEAM)  at the Environmental Research Laboratory in Athens, Georgia. See Section 2 for
information about obtaining the code.

6.2.1 One-Dimensional Transport Problem

Transient one-dimensional transport through a horizontal column is simulated in this
example  (Figure 6.4). The dimensions of the column are  identical to those of the column
in Figure 6.1. It has a length of 200 cm and a  50 cm x 50 cm cross-section. Initially, the
concentration is 0.0 g/cm3 throughout the region of interest. The concentration at x = 0.0
cm is maintained at C = CO = 1.0 g/cm3 (a Dirichlet boundary).  A variable boundary
condition is used to specify the natural condition of zero  gradient flux at x = 200.0 cm. A
bulk density of 1.2 g/cm3 and a longitudinal dispersivity of 5.0 cm are assumed. No
adsorption or decay is allowed. A specific discharge (Darcy velocity)  of 2.0 cm/d is
assumed and a moisture content of 0.4 is used.

The  region of interest, that  is,  the whole column, is discretized with  1 x 1 x 40 = 40
elements with the element size equal to  50 x  50 x 5 cm. This results in  2x2x41 = 164
node points.  For this simulation, each of the four vertical lines is considered a subregion.
Thus, a total of four subregions, each with 41 node points, is used for the subregional
block iteration simulation.  A constant time-step size of 0.5 is used and  a  40 time-step
simulation is run. For this  discretization, the mesh Peclet number is Pe = 1 and the
Courant  number is Cr = 0.5.

The  input data set for this problem, prepared according to the instructions in  Sections 4.2
and  5.2, is shown in Table  6-4.
                   c-i
      Figure 6.4. One-dimensional transient transport through a horizontal column.
                                          144

-------
TABLE 6-4. INPUT DATA SET FOR THE ONE-DIMENSIONAL 3DLEWASTE PROBLEM
 1 ONE-D FIRST TYPE BOUNDARY VALUE PROBLEM WITH 3DLEWASTE L=CM,T=DAY,M=G 00
C ******* DATA SET 2: BASIC INTEGER PARAMETERS ***** ***************************
 164 40  1  0  40  1  8 -1  1  0  1  1  1 100  1
C ******* DATA SET 3: BASIC REAL PARAMETERS ***** ******************************
  0.50DO O.ODO l.ODO 20.0DO l.ODOO  l.ODO  l.OD-3  l.OD-4
C ******* DATA SET 4: PRINTER, STORAGE AND TIME RESETTING CONTROL *************
55000000005000000000500000000050000000005
11000000001000000000100000000010000000001
  1.0D38
C ******* DATA SET 5: MATERIAL PROPERTIES *************************************
  O.ODO 1.2DO 5.0DO O.ODO  O.ODO  l.ODO O.ODO O.ODO
C **** *** DATA SET 6: NODE COORDINATES ****************************************
  1 40 1 O.ODO 50.0DO O.ODO O.ODO O.ODO 5.0DO
 42 40 1 O.ODO O.ODO O.ODO  O.ODO O.ODO 5.0DO
 83 40 1 50.0DO O.ODO O.ODO O.ODO O.ODO 5.0DO
124 40 1 50.0DO 50.0DO O.ODO O.ODO O.ODO 5.0DO
  000    0.0    0.0    0.0   0.0  0.0   0.0   END OF COORDINATES
C ******* DATA SET 7: ELEMENT CONNECTIVITY ***** *******************************
  1 39 1 42 83 124  1 43 84 125 2  1
  000000000000    END OF IE
Q ******* DATA SET 8" SUBREGIONAL DATA ***** ***********************************
  4
  1 3 1 41  0
  00000                      END OF NNPLR(K)
  140 1 1 1
  00000                      END OF GNLR(I,1)
  140 142  1
  00000                      END OF GNLR(I,2)
  1 40 1 83 1
 00000                      END OF GNLR(I,3)
  1  40  1  124   1
 00000                       END OF GNLR(I,4)
£ ******* DATA SET 10: INITIAL CONDITIONS ***** ********************************
    1    3    41      l.ODO   O.ODO   0.0
    2   38      1      O.ODO   O.ODO   0.0
   43    38    1       .ODO   O.ODO   0.0
   84    38    1      O.ODO   O.ODO   0.0
  125    38    1      O.ODO   O.ODO   0.0
   41    3    41      O.ODO   O.ODO   0.0
   0 0 0 0.0 0.0 0.0                    END OF 1C
C ******* DATA SET 11: SOURCE/SINK AND B. C. CONTROL INTEGERS ***** ************
 0000 0000 4120
 14120  00000 00000
                                         145

-------
TABLE  6-4.  INPUT DATA SET FOR THE ONE-DIMENSIONAL 3DLEWASTE  PROBLEM
            (concluded)
C ******* DATA SET 13: VARIABLE BOUNDARY CONDITIONS ***** **********************
 O.ODO O.ODO 1.0D38  O.ODO
  10010
  00000             ENDOFIRTYP
  1 0  0 82 123 164 41 0 0 0 0
  000 0  0  000000       END OF ISV(J,I) J=l,4
  1   3  1  41  41
  00000             ENDOFNPVB
C ******* DATA SET 14: DIRICHLET BOUNDARY CONDITIONS **************************
   O.ODO  l.ODO  1.0D38   l.ODO
  1  42 83  124
  13110
  00000             ENDOFIDTYP
C ******* DATA SET 16: HYDROLOGICAL BOUNDARY CONDITIONS ***********************
 1 163 1 O.ODO O.ODO 2.0DO O.ODO O.ODO O.ODO
 0  00  0.0 0.0  0.0 0.0 0.0  0.0 END  OF  VELOCITY
  1 39 1  0.4DO 0.0
  0   0  0   0.0   0.0          ENDOFTH
  Q **************** END OF JOB ***** ***************************************oo
                                         146

-------
6.2.2 Two-Dimensional Transport in a Rectangular Region

This is a two-dimensional transport problem in a rectangular region 600.0 cm long, 270.0
cm high, and  1.0 cm thick (Figure 6.5). Initially, the concentration is zero g/cm3 through-
out the region of interest. A concentration of 1.0 g/cm3 is maintained at x = 0.0 cm and
180.0 cm < y  <270.0 cm by applying a Dirichlet boundary condition. A concentration of
0.0 g/cm3 is maintained at x = 0.0 cm and 0.0 cm < y < 90.0 cm and 180.0 cm < y <270.0
cm. A natural condition is imposed at x = 600 cm using a variable  boundary condition
(Equation 3-39b). A single material with  a bulk density of  1.2 g/cm3, a longitudinal
dispersivity of 10.0 cm, and a lateral dispersivity of  1.0 cm modeled. No adsorption or
decay is allowed. A specific discharge (Darcy velocity) of 2.0 cm/d is used and a moisture
content of 0.2  is assumed.

The region is divided into 9x9x1=81 elements, resulting inl0x!0x2 = 200 nodes.
The element size is 60.0 cm x 30.0 cm x 1.0 cm. Each of the two vertical planes is
considered a subregion. Thus, a total of two subregions, each with  100 nodal points, is
used for the subregional block iteration simulation. A time-step size of 4.5 is used and 40
time steps are simulated.

The input data set for this problem, prepared according to the instructions in Sections 4.2
and 5.2, is shown in Table 6-5.
         V=2.0
                           c=o.o
                          C=1.0
                          00.0
270
                                       600
        Figure 6.5. Two-dimensional transient transport in a rectangular region.
                                         147

-------
TABLE 6-5. INPUT DATA SET FOR THE TWO-DIMENSIONAL  3DLEWASTE PROBLEM
1 TWO-D FIRST TYPE BOUNDARY VALUE PROBLEM WITH 3DLEWASTE L=CM,T=DAY,M=G 00
C ******* DATA SET 2: BASIC INTEGER PARAMETERS ***** ***************************
 200 81 1 0 40 1 8 -1  1  0  1  1  1 200  1
C ******* DATA SET 3: BASIC REAL PARAMETERS ***** ******************************
 0.45D1 O.ODO 9.0DO 3.60D2 l.ODOO 1.78DO l.OD-3 l.OD-4
C ******* DATA SET 4: PRINTER, STORAGE AND TIME RESETTING CONTROL *************
55000000005000000000500000000050000000005
11000000001000000000100000000010000000001
   1.0D38
C **** *** DATA SET 5: MATERIAL PROPERTIES *************************************
 O.ODO 1.2DO 10.0DO l.ODO O.ODO l.ODO O.ODO O.ODO
C **** *** DATA SET 6: NODE COORDINATES ****************************************
1
2
3
4
5
6
7
8
9
10
101
102
103
104
105
106
107
108
109
110
0 0
9 10 O.ODO O.ODO O.ODO 6.0D1 O.ODO
9 10 O.ODO 3.0D1 O.ODO 6.0D1 O.ODO
9 10 O.ODO 6.0D1 O.ODO 6.0D1 O.ODO
9 10 O.ODO 9.0D1 O.ODO 6.0D1 O.ODO
9 10 O.ODO 12.0D1 O.ODO 6.0D1 O.ODO
9 10 O.ODO 15.0D1 O.ODO 6.0D1 O.ODO
9 10 O.ODO 18.0D1 O.ODO 6.0D1 O.ODO
9 10 O.ODO 21.0D1 O.ODO 6.0D1 O.ODO
9 10 O.ODO 24.0D1 O.ODO 6.0D1 O.ODO
9 10 O.ODO 27.0D1 O.ODO 6.0D1 O.ODO
9 10 O.ODO O.ODO l.ODO 6.0D1 O.ODO
9 10 O.ODO 3.0D1 l.ODO 6.0D1 O.ODO
9 10 O.ODO 6.0D1 l.ODO 6.0D1 O.ODO
9 10 O.ODO 9.0D1 l.ODO 6.0D1 O.ODO
9 10 O.ODO 12.0D1 l.ODO 6.0D1 O.ODO
9 10 O.ODO 15.0D1 l.ODO 6.0D1 O.ODO
9 10 O.ODO 18.0D1 l.ODO 6.0D1 O.ODO
9 10 O.ODO 21.0D1 l.ODO 6.0D1 O.ODO
9 10 O.ODO 24.0D1 l.ODO 6.0D1 O.ODO
9 10 O.ODO 27.0D1 l.ODO 6.0D1 O.ODO
0 0.0 0.0 0.0 0.0 0.0
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
O.ODO
0.0 END OF COORDINATES
C ******* DATA SET 7: ELEMENT CONNECTIVITY ***** *******************************
1
10
19
28
37
46
55
64
73
0 0
81 1 11 12 2 101 111 112 102 1
8 1 11 21 22 12 111 121 122 112 1
8 1 21 31 32 22 121 131 132 122 1
8 1 31 41 42 32 131 141 142 132 1
8 1 41 51 52 42 141 151 152 142 1
8 1 51 61 62 52 151 161 162 152 1
8 1 61 71 72 62 161 171 172 162 1
8 1 71 81 82 72 171 181 182 172 1
8 1 81 91 92 82 181 191 192 182 1
0000000000









END OF IE
C ******* DATA SET 8: SUBREGIONAL DATA ***** ***********************************
   2
  1  1  1 100 0
  00000                        END OF NNPLR(K)
  1  99  1 1  1
  00000                        END OF GNLR(I,1)
  1  99  1 101 1
  00000                        END OF GNLR(I,2)
                                        148

-------
TABLE  6-5.  INPUT DATA SET FOR THE TWO-DIMENSIONAL 3DLEWASTE PROBLEM
            (concluded)
C ******* DATA SET 10: INITIAL CONDITIONS *************************************
   1  199  1      O.ODO   O.ODO 0.0
   0 0 0 0.0 0.0 0.0               END OF 1C
C ******* DATA SET 11: SOURCE/SINK AND B. C. CONTROL INTEGERS *****************
 0  00000008120
 9  10  1200000000000
C *** **** DATA SET 13: VARIABLE BOUNDARY CONDITIONS ***************************
 0. ODO O.ODO 1.0D38 O.ODO
   18110
   00000                      ENDOFIRTYP
   1 8 1 91 92  192  191  1 1  1 1
   00000000   0 00           END OF ISV(J,I) J=l,4
   1   9   1  91   1
   00000                      ENDOFNFVB
C ******* DATA SET 14: DIRICHLET BOUNDARY CONDITIONS **************************
   O.ODO  l.ODO  1.0D38  l.ODO
   4  5  6   7  104  105   106  107
   17110
   00000                      ENDOFIDTYP
C ******* DATA SET 16: HYDROLOGICAL BOUNDARY CONDITIONS ***********************
 1  199 1 2.0DO O.ODO O.ODO O.ODO O.ODO  O.ODO
 000    0.0  0.0   0.0   0.0   0.0   0.0     END OF VELOCITY
   1  80  1 0.2DO 0.0
   0   0  0   0.0   0.0                  ENDOFTH
   Q **************** gj^D OF JOB ***** ****** ***** ***** ***** ***** *****#*******QQ
                                        149

-------
                                    SECTION 7

                                  REFERENCES
Bear, J.  and A. Vermijt.  1987. Modeling  Groundwater Flow and Pollution. D. Reidel
     Publishing Company, Dordrecht, Holland. 414p.

Bresler, E. 1973. Simultaneous Transport of Solutes  and Water Under Transient
     Unsaturated  Flow Conditions. Water Resour.  Res. 9(4):975-986.

Brooks, R.H. and A.T. Corey.  1966. Properties  of Porous Media Affecting Fluid Flow.
     ASCE J. Irrig. Drain Div. 92 (IR2):61-68.

Carsel, R.F. and R.S. Parrish.  1988. Developing Joint Probability Distributions of Soil-
     Water  Retention  Characteristics. Water Resour. Res. 24(5):755-769.

CRC. 1965. Handbook of Chemistry and  Physics. 46th Edition. CRC Press, Boca
     Raton, Florida.

CRC. 1981. Handbook of Chemistry and  Physics. 62nd Edition. CRC Press, Boca
     Raton, Florida.

Davis, S.N.,  and J.M. DeWiest. 1966. Hydrogeology.  John Wiley & Sons,  Inc.

Dean, J. D., P.S. Huyakorn, A.S. Donigian, Jr., K.A. Voos, R.W.  Schanz, and R.F.
     Carsel.  1989. Risk  of Unsaturated/Saturated  Transport and  Transformation of
     Chemical Concentrations  (RUSTIC).  Volume II: User's Guide. U.S.
     Environmental  Protection Agency, Athens, Georgia.
     EPN600/3-89-048b.

de Marsily,  G. 1986.  Quantitative Hydrogeology: Groundwater Hydrology for
     Engineers. Academic Press, San  Diego, California. 440 p.

Electric Power Research  Institute. 1985. A Review  of Field Scale Physical Solute
     Transport Processes in Saturated and Unsaturated Porous Media. EPRI EA-
     4190, Project 2485-5, Palo Alto, California.

Federal Register.  1986.  Hazardous Waste Management System:  Land  Disposal
     Restrictions. U.S. Environmental Protection Agency, Vol.  51,  No. 9.

Freeze,  R.A.  and J.A.  Cherry.  1979.  Groundwater. Prentice-Hall,  New Jersey.  604
     PP.

Gelhar,  L.W.,  C. Welty, and K.R. Rehfeldt. 1992. A Critical Review of Data on Field-
     Scale Dispersion in Aquifers. Water  Resour. Res. 28(7):1956-1974.

                                        150

-------
Colder Associates. 1977. Development of Site Suitability Criteria for a High Level
     Waste Repository. Lawrence Livermore Laboratory,  UCRL-13793.

Helfferich, F. 1962.  Ion Exchange. McGraw-Hill, New York, New York.

Imhoff, J. C., R.F. Carsel, J.L. Kittle, Jr., and P.R.  Hummel. 1990. Database Analyzer
     and Parameter Estimator,  DBAPE (User's Manual). EPM600/3-89/083.
     Environmental  Research Laboratory, U.S Environmental Protection Agency,
     Athens, Georgia.

Karickhoff, S. W., D.S. Brown, and T.A. Scott. 1979. Sorption of Hydrophobic
     Pollutants on Natural Sediments. Water Research 13:241-248.

Keely, J.F. 1989. Performance Evaluations  of Pump-and-Treat Remediations.
     EPA/540/4-89/005, U.S. Environmental Protection Agency, Ada, Oklahoma. 19 pp.

Lyman,  W.J., W.F. Reehl, and D.H. Rosenblatt. 1982. Handbook of Chemical Property
     Estimation Methods: Environmental Behavior of Organic  Compounds. McGraw-
     Hill, New York. 960 pp.

Mabey, W. R., J.H. Smith, R.T. Podoll, H.L. Johnson, T. Mill, T.W. Chou, J. Gates, I.
     Waight  Partridge, J. Jaber, and D. Vandenberg. 1982. Aquatic Fate Process
     Data for Organic Priority Pollutants. EPA/440/4-81-014. Office of Water
     Regulations  and Standards, U.S. Environmental Protection Agency, Washington,
     DC.

Meyer, C.A., R.B. McClintock, G.J. Silvestri, and R.C. Spencer. 1968.  1967 ASME
     Steam Tables. 2nd Edition. The American Society  of Mechanical Engineers, New
     York, New York.

Mercer, J. W., S.D. Thomas, and B. Ross. 1982. Parameters and Variables Appearing
     in  Repository Siting Models. NUREG/CR-3066.  U.S. Nuclear Regulatory
     Commission, Washington, D.C.

Mills, W. B., D.B. Porcella, M.J. Ungs, S.A. Gherini, K.V. Summers, L. Mok,  G.L. Rupp,
     G.L. Bowie,  and D.A.  Haith. 1985a. Water Quality  Assessment A Screening
     Procedure for Toxic and Conventional Pollutants in Surface and Ground Water:
     Part I. EPA/600/6-85/002a. U.S. Environmental Protection Agency, Athens, GA.

Mills, W. B., D.B. Porcella, M.J. Ungs, S.A. Gherini, K.V. Summers, L. Mok,  G.L. Rupp,
     G.L. Bowie,  and D.A. Haith. 1985b.  Water Quality Assessment: A Screening
     Procedure for Toxic and Conventional Pollutants in Surface and Ground Water:
     Part II. EPA/600/6-85/002b. U.S. Environmental Protection Agency,  Athens,  GA.

Morris, D.A., and A.I. Johnson.  1967.  Summary of  Hydrologic and Physical Properties
     of Rock and Soil Materials  as Analyzed by the Hydrologic Laboratory of the U.S.
     Geological Survey. U.S. Geological Survey Water Supply Paper 1839-D, 1967.

Mualem,  Y. 1976. A  New Model for Predicting the  Hydraulic Conductivity  of
     Unsaturated Porous Media. Water Resour. Res. 12(3):513-522.

                                        151

-------
Mullins, J.A., R.F. Carsel, J.E. Scarborough, and A.M. Ivery. 1992. PRZM-2 User's
     Manual, Version 1.0.  Office of Research and Development, U.S. Environmental
     Protection Agency, Athens, GA.

Finder, G. F., and W. G. Gray. 1977.  Finite Element Simulation in Surface and
     Subsurface Hydrology. Academic  Press, New York.

Sharp-Hansen, S., C. Travers, P. Hummel, and T. Allison. 1990.  A Subtitle D  Landfill
     Application Manual for the Multimedia Exposure Assessment Model
     (MULTIMED).  Office of Research and Development, Environmental Protection
     Agency, Athens, GA.

U.S. Environmental  Protection Agency. 1987. Guidelines for Delineating Wellhead
     Protection Areas. EPA No. 440/6-87-010. Office of Groundwater Protection, U.S.
     Environmental  Protection Agency, Washington, DC. NTIS No. PB88-111430-AS.

U.S. Environmental  Protection Agency. 1989. Transport and Fate of Contaminants in
     the Subsurface: Seminar Publication.  EPA/625/4-89/019.  Center for
     Environmental  Research Education,  U.S. Environmental Protection Agency,
     Cincinnati, OH.

van Genuchten, M.T. 1980.  A Closed-form Equation for  Predicting the Hydraulic
     Conductivity of Unsaturated  Soils. Soil Sci. Sot. J.  44:892-898.
                                        152

-------
                                  APPENDIX A

           PROGRAM STRUCTURE AND  SUBROUTINE DESCRIPTIONS
A.I  3DFEMWATER

3DFEMWATER consists of a main program, FEMWAT3D, and  22 subroutines.  Figure
A. 1 shows the structure of the program. The subroutines are listed in Table A-l and
the functions of these subroutines are briefly described below.

A. 1.1  Subroutine ALLFCT

This subroutine is called by subroutine  GW3D to compute values for all the source/sink
and boundary nodes and elements. It uses linear interpolation of tabular data to
simulate variations in time for these conditions.
A. 1.2 Subroutine ASEMBL

This subroutine is called by subroutine GW3D. After calling subroutine Q8 to evaluate
the element matrices, it sums over all element matrices to form a global matrix
equation governing  the pressure head at all nodes.
A. 1.3 Subroutine BASE
This subroutine is called by subroutines Q8DV and Q8 to evaluate the value of the
base function at a Gaussian point.
A. 1.4 Subroutine BC

This subroutine, which is called by subroutine GW3D, incorporates Dirichlet, variable
composite, specified-flux  (Cauchy), and specified-pressure-head gradient (Neumann)
boundary conditions.  For a Dirichlet boundary condition, an identity algebraic
equation is generated for each Dirichlet nodal point. Any other equation having this
nodal variable is modified accordingly to simplify the computation. For a specified-flux
surface, the integration of the surface source is obtained by calling subroutine Q4S and
the result is added  to the load vector. For a specified-pressure-head gradient surface,
the integrations of both the gradient and gravity fluxes are obtained by calling the
subroutine Q4S. These fluxes are added  to the load vector. The subroutine BC also
implements  the variable  composite  boundary condition. First, it checks all infiltration-
evapotranspiration-seepage points, identifying any of them that are Dirichlet
                                        153

-------
Figure A. 1. Program structure of 3DFEMWATER
                    154

-------
TABLE A-l.  SUBROUTINES  INCLUDED  IN 3DFEMWATER
Subroutine
ALLFCT
ASEMBL
Called By

GW3D


GW3D
BASE
BC
Q8DV, Q8
GW3D
BCPREP
GW3D
BLKITR


DATAIN

GW3D
GW3D


GW3D

FEMWAT3D
PAGEN
DATAIN
PRINTT
GW3D
Description

Interpolates functional values for source/sink
and boundary conditions.

Evaluates the element matrices and then
sums over all element matrices to form a
global matrix equation governing the pres-
sure head at all nodes.

Evaluates the value of the base function at a
Gaussian point.

Incorporates Dirichlet, specified-flux
(Cauchy),  specified-pressure-head gradient
(Neumann),  and variable composite boundary
conditions.

Prepares the infiltration/seepage boundary
conditions during a rainfall period or the
seepage/evapotranspiration boundary con-
ditions during non-rainfall periods.

Solves the matrix equation with  block itera-
tion methods.

Reads and prints all input information.

Controls the entire sequence of operations.
Performs either the steady-state computation
alone, or a transient-state computation using
the steady-state solution as the initial condi-
tion, or  a transient computation using user-
supplied initial conditions.

Preprocesses pointer arrays that are needed
to store  the  global matrix in compressed form
and to construct the subregional block matri-
ces.

Prints the flow variables, which include the
fluxes through variable boundary surfaces,
pressure head, total head, moisture content,
and Darcy velocity components.
                                       155

-------
TABLE A-l. SUBROUTINES  INCLUDED IN  3DFEMWATER  (continued)
Subroutine

Q4S


Q8


Q8DV


Q8TH


READN


READR


SFLOW



SOLVE


SPROP
STORE
Called By

BCPREP, BC,


ASEMBL


VELT


SFLOW


DATAIN


DATAIN


GW3D



BLKITR


GW3D
GW3D
SURF
DATAIN
Description

Evaluates the boundary surface load vector
SFLOW over a SFLOW boundary segment.

Computes the element matrices and element
load vector.

Computes the integration of N(I)*N(J) and -
N(I)* K42RAD(HT) over an element.

Evaluates the integration of moisture content
and sources/sinks over an element.

Automatically generates integer input if
required.

Automatically generates real number input if
required.

Computes the fluxes through boundaries and
the rate at which water content increases in
the region of interest.

Solves a matrix equation with the direct
band method.

Calculates the values of moisture content,
relative hydraulic conductivity, and water
capacity using van Genuchten analytical
functions.

Stores the solution in binary on logical unit
LUSTO. Information stored includes regional
geometry, subregion data, and hydrological
variables such as pressure head, total head,
moisture content,  and the Darcy velocity
components.

Identifies the boundary sides, sequences the
boundary nodes, and computes  the direction-
al cosine of the surface sides.
                                      156

-------
TABLE A-l. SUBROUTINES  INCLUDED IN 3DFEMWATER (continued)
Subroutine

VELT
Called Bv

GW3D
Description

Evaluates the element matrices and the
derivatives of the total head, and then sums
over all element matrices to form a matrix
equation  governing the velocity components
at all nodal points.
                                      157

-------
points.  If there are Dirichlet points, they are incorporated using the method described
above. If a given point is not a Dirichlet point,  the point is bypassed. Second, it
checks all rainfall-evaporation-seepage points again to see if any of them is a specified-
flux point. If it is, then the  computed flux by infiltration or potential evapotranspira-
tion is added to the load vector. If a given point is not a specified-flux point, it is
bypassed. Because the infiltration-evaporation-seepage points are either Dirichlet or
specified-flux points,  all points are taken care of in this manner.

A. 1.5 Subroutine BCPREP

This subroutine is called by GW3D to prepare the infiltration-seepage boundary
conditions during a rainfall  period or the seepage-evapotranspiration boundary
conditions during non-rainfall periods.  It decides the number of nodal points on the
variable boundary to be considered as  Dirichlet or specified-flux (Cauchy) points. It
computes the number of points that change boundary conditions from 1)  ponding depth
(Dirichlet types) to infiltration (specified-flux types), or 2) infiltration to pending depth,
or 3) minimum pressure (Dirichlet types) to infiltration during  rainfall periods.  It also
computes the number of points that change boundary conditions from potential
evapotranspiration (specified-flux types) to minimum pressure, or from ponding  depth
to potential evapotranspiration, or from minimum pressure to potential evapotrans-
piration during non-rainfall  periods. Upon completion, this subroutine  returns the
Darcy flux (DCYFLX),  infiltration/potential evapotranspiration rate (FLX), the ponding
depth nodal index (NPCON), the flux-type nodal index (NPFI.X), the minimum
pressure nodal index (NPMIN),  and  the number of nodal  points (NCHG)  that have
changed boundary conditions.

A.I.6 Subroutine  BLKITR

This subroutine is called by  subroutine GW3D to solve the matrix equation with block
iteration methods. For each subregion, a block matrix equation is  constructed based
on the global matrix  equation and two pointer arrays, GNPLR and LNOJCN (see
subroutine PAGEN),  and the resulting block matrix  equation is solved with the  direct
band matrix solver by calling subroutine SOLVE.  This is done for  all subregions for
each iteration until a convergent solution is obtained.

A. 1.7 Subroutine  DATAIN

Subroutine DATAIN  is called by subroutine GW3D.  It reads all data input described
in Section 4.1 except data  set 1.  It also calls subroutine  SURF to identify the surface
elements  and boundary nodes, and subroutines READR and READN, respectively, to
automatically generate real and integer numbers.

A. 1.8 Subroutine  GW3D

Subroutine GW3D controls the entire sequence of operations. It performs either a
steady-state computation alone (KSS = 0 and NTI = 0), or a transient- state computa-
tion  using the steady-state solution as the initial condition (KSS = 0, NTI > 0), or a
transient computation using user-supplied initial conditions (KSS = 1, NTI > 0).
                                         158

-------
GW3D calls subroutine DATAIN to read and print input data; subroutine PAGEN to
generate pointer arrays; subroutine ALLFCT to obtain source/sink and boundary
values; subroutine SPROP to obtain the relative hydraulic conductivity, water capacity,
and moisture content from the pressure head; subroutine VELT to compute Darcy
velocity, subroutine BCPREP to determine if a change of boundary conditions is
required, subroutine ASEMBL to assemble the element matrices over all elements;
subroutine BC  to implement the boundary conditions; subroutine BLKITR to form and
solve the subregional block matrix equations;  subroutine SFLOW to calculate flux
through all types of boundaries and water accumulated in the media; subroutine
PRINTT to print out the results; and subroutine STORE to store the flow variables for
input to 3DLEWASTE or for plotting.

A. 1.9 Subroutine PAGEN

This subroutine is called by subroutine DATIUN to preprocess pointer arrays that are
needed to store the global matrix in compressed form and to construct the subregional
block matrices. The pointer arrays automatically  generated in this subroutine include
the global node connectivity (stencil), GNOJCN(J,N), regional node connectivity,
LNOJCN(J,I,K), total node number for each subregion, NTNPLR(K), the bandwidth
indicator for each subregion, LMAXDF(K), and a partial fall-up for the mapping array
between global  node number and local subregion node number, GNPLR(I,K), with I =
NNPLR(K) + 1  to NTNPLR(K). Here GNOJCN(J,N) is the global node number of the
J-th node connected to the global node N; LNOJCN(J,I,K) is the local node number of
the J-th node connected to the local node I in the K-th subregion; NTNPLR(K) is the
total number of nodes in the K-th subregion, including the interior nodes, the global
boundary nodes, and intra-boundary nodes; LMAXDF(K) is the maximum difference
between any two nodes of any element in the K-th subregion; and GNPLR(I,K) is the
global node number of the I-th local-region node in the K-th subregion. These pointer
arrays are generated based on the element connectivity, IE(M,J), the number of nodes
for each subregion,  NNPLR(K), and the  mapping  between global node and local-region
node, GNLR(I,K), with 1=1, NNPLR(K).  Here IE(M,J) is the global  node number of J-
th node of element M; NNPLR(K) is  the number of nodes in the K-th subregion,
including the interior nodes and the global boundary nodes, but not the intraboundary
nodes.

A. 1.10 Subroutine  PRINTT

This subroutine, which is called by GW3D, is used to line-print the flow variables.
These include the fluxes through variable boundary surfaces, the pressure head, total
head, moisture  content, and Darcy velocity components.

A. 1.11 Subroutine O4S

This subroutine is called  by subroutines BC, BCPREP, and SFLOW to compute the
surface node flux of the type:
                                        159

-------
                                 RQ(I) = jN/qdB
                                                                       (A-l)
where q is either the specified-flux, specified-pressure-head gradient flux, or gravity
flux; B is  the global boundary of the region of interest; N;e is the basis functions for
nodal point i of element e; and RQ(I) is a 3DFEMWATER code parameter.
A.I.12 Subroutine O8

This subroutine is called by the subroutine ASEMBL to compute the element matrix
given by
                               QA(I,J) =
                                       N,ed R
                                                                      (A-2a)
                          QB(I,J) = J(VNie)-K.kr
-------
                               QRd.J)  =jNieNjedR
where QR(IJ) is a 3DFEMWATER program variable.  Subroutine Q8DV also evaluates
the element load vector:

                         QRX(I) = - fNiei-K.kr
-------
A.I.15 Subroutine  READN

This subroutine is called by subroutine DATAIN to generate integer numbers for input
data sets 8, 9, 12(c), 12(f),  14(b) through 14(d), 15(c),  16(b), 16(c), 17(b), and 17(c),
which are described in Section 4.1.
A. 1.16 Subroutine READR

This subroutine is called by subroutine DATAIN to generate real numbers input for
data sets 7, 14(e), and 14(f) (see Section 4.1). Automatic generation of regularly
patterned data is built into this subroutine.
A. 1.17 Subroutine SFLOW

This subroutine is called by subroutine GW3D.  It is used to compute the fluxes
through various types of boundaries and the rate at which water content increases in
the region of interest. In this subroutine, the function of variable FRATE(7) is  to store
the flux through the whole boundary enclosing the region of interest. It is given by:

                       FRATE(7) = f(VA + Vyny +  Vznz)dB                   (A-5)
where Vx, Vy, and Vz are Darcy velocity components, and n^, riy, and nz are the direc-
tional  cosines of the outward unit vector normal to the boundary B. FRATE(l) through
FRATE(5) store the flux through the Dirichlet boundary, BD, specified-flux boundary,
Bc, specified-pressure-head boundary, BN, the seepage/evapotranspiration boundary,
Bg, and infiltration boundary, BR, respectively, and are given by:

                       FRATE(l)  = f(VA + Vyny + Vziiz)dB                  (A-6a)
                       FRATE(2)  = f(VA + Vyny + Vznz)dB                  (A-6b)
                                    B.
                       FRATE(3) = J(Vxnx + V,n, + Vznz)dB                  (A-6C)
FRATE(6), which is related to the numerical loss, is given by:
                                        162

-------
                       FRATE(4) =  (Vxnx  + Vyny + VzigdB                 (A-6d)
                                   B.
                       FRATE(5) =  (Vxnx + Vyny + Vznz)dB                 (A-6e)
                                   B,
                                                5
                      FRATE(6) = FRATE(7) -  £ FRATE(I)                 (A-7)
                                               1=1

FRATE(8) and FRATE(9) are used to store the source/sink and increased rate of water
accumulation within the media, respectively:

                               FRATE(8) =  - fqdR                          (A-8)
and

                              FRATE(9) =  (F^idR                         (A-9)
                                          R  "v


If there is no numerical error in the computation, the following equation should be
satisfied:

                       FRATE(9)  = -[FRATE(7) +  FRATE(8)]                (A-IQ)

and FRATE(6) should be equal to zero.

A.I.18 Subroutine SOLVE

This subroutine is called by the subroutine BLKITR to solve a matrix equation of the
type:

                                    [c]{x| = {y}                             
-------
A. 1.20 Subroutine  STORE

This subroutine, which is called by GW3D, is used to store the flow variables in a
binary file. The stored data are intended for use in 3DLEWASTE  or for plotting. The
information stored  includes region geometry, subregion data, and hydrological vari-
ables such as pressure head, total head, moisture content, and Darcy velocity compo-
nents.

A. 1.21  Subroutine  SURF

Subroutine SURF is called by subroutine DATAIN.  It identifies the boundary sides,
sequences the boundary nodes, and computes the directional cosine of the surface
sides. The mappings from  boundary nodes to global nodes are  stored in NPBB(I)
(where NPBB(I) is  the global node  number of the I-th boundary  node). The boundary
node numbers of the four nodes for each boundary side are stored in ISB(I,J) (where
ISB(I,J) is the boundary node number of I-th node of J-th side, I =  1 to 4). There are
six sides for  each element. Which of these six sides  is the boundary side is determined
automatically in the subroutine SURF and is stored in ISB(5,J). The global element
number, to which the J-th boundary side belongs, is also  preprocessed in the subrou-
tine SURF and is stored in ISB(6,J). The directional cosines of the J-th boundary side
are computed and stored in DCOSB(I.J) (where DCOSB(I.J) is  the  directional cosine of
the J-th surface with I-th coordinate, I =  1 to 3). The information contained in NPBB,
ISB, and DOSB, along with the number of boundary nodes and the number of bound-
ary sides, is  returned to subroutine DATAIN for other uses.

A. 1.22 Subroutine  VELT

This subroutine is called by subroutine GW3D. It calls subroutine  Q8DV to  evaluate
the element matrices and the derivatives of the total head. It then sums over all
element matrices to form a matrix equation governing the velocity  components at all
nodal points. To save computational time, the  matrix is  diagonalized  by lumping.  The
velocity components can thus be solved point by point. The computed velocity field is
then returned to GW3D through the argument. This velocity field is also passed to
subroutine BCPREP to evaluate the Darcy flux across the seepage-infiltration-evapo-
transpiration surfaces.

A.2 3DLEWASTE

LEWASTE consists of a main program, LEWAST3D, 30 subroutines, and a function.
Figure A.2 shows the structure of the program.  The subroutines' and function are
listed in Tables A-2 and A-3, respectively, and the purposes of the subroutines are
briefly described below.

A.2.1  Subroutine ADVBC
This subroutine is called by GM3D to implement the boundary conditions. For a
Dirichlet boundary, the Lagrangian  concentration is specified. For variable bound-
aries, if the flow is directed out of the region, the fictitious particle associated with the
boundary node must come from the  interior nodes. Hence the Lagrangian concentra-
tion for the boundary node has already  been computed from subroutine ADVTRN  and

                                        164

-------
Figure A. 2. Program structure of 3DLEWASTE
                    165

-------
TABLE A-2. SUBROUTINES INCLUDED IN 3DLEWASTE
Subroutine
Called Bv
Description
ADVBC
ADVTRN
AFABTA
ALLFCT
ASEMBL
BC
BLKITR
DATMN
FLUX
GM3D
GM3D
GM3D
GM3D
GM3D
GM3D
GM3D
GM3D
GM3D
GM3D
LEWAST3D
Applies specified-flux (Cauchy), variable, and
Dirichlet boundary conditions.

Computes the Lagrangian concentrations at
all nodes and finds in which element the
fictitious particle is located.

Calculates the values of the upstream
weighting factors along the 12 sides of all
elements.

Interpolates functional values for source/sink
and boundary conditions.

Evaluates the element matrices and then
sums over all element matrices to form a
global matrix equation governing the concen-
tration distribution at all nodes.

Incorporates Dirichlet, variable composite,
specified-flux (Cauchy), and specified-dispers-
ive-flux (Neumann) boundary conditions.

Solves the matrix equations with block itera-
tion methods.

Reads and prints system parameters, geome-
try, boundary and initial conditions, and
properties of the solute and media.

Evaluates the element matrices and the
derivatives of concentrations and then sums
over all element matrices to form a matrix
equation governing the flux components at
all nodal points.

Controls the entire sequence of operations.
Performs either the steady-state computation
alone, or a transient- state computation us-
ing the steady-state solution as the initial
condition, or a transient computation using
user-supplied initial  conditions.
                                       166

-------
TABLE A-2. SUBROUTINES INCLUDED IN 3DLEWASTE (continued)
Subroutine

LELGEN

MPLOC
NDTAU
PAGEN
PRINTT



Q4ADB



Q4BB



Q4CNB



Q8


Q8DV


Q8R


READN
Called By

GM3D

NDTAU,
ADVTRN
GM3D
DATAIN
GM3D
ADVBC
SFLOW
BC
ASEMBL
FLUX
SFLOW
DATAIN
Description

Finds the elements connecting to each node.

Locates the fictitious particle associated
with a particular node. Computes the prod-
uct of the outward unit vector with the vector
from a node on the surface to the fictitious
particle.

Determines the number of subtime steps and
the subtime step size for Lagrangian integr-
ation.

Preprocesses pointer arrays that are needed
to store the global matrix in compressed form
and to construct  the subregional block matri-
ces.

Prints material flow, concentration, and
material flux output as specified by the pa-
rameter KPR.

Implements Dirichlet, specified-flux
(Cauchy), and  variable boundary conditions
in the  Lagrangian step computation.

Computes normal flow rates (M/T) by inte-
grating the normal fluxes (M/L2/T) over a
boundary surface.

Computes the  boundary-surface matrix and
the boundary-surface load vector over a boun-
dary surface.

Computes element matrices and element load
vectors.

Computes the  integration of N(I)*N(J) and -
N(I)*D>GRAD(C) over an element.

Computes the  material integration and ele-
ment source integration over an element.

Automatically  generates integer input if
required.
                                      167

-------
TABLE A-2. SUBROUTINES INCLUDED  IN 3DLEWASTE (concluded)
Subroutine
Called Bv
Description
READR


SFLOW



SHAPE



SOLVE


STORE


SURF



THNODE

XSI3D
DATAIN


GM3D



Q8DV, Q8



BLKITR


GM3D


DATAIN



GM3D

ADVTRN
Automatically generates real number input if
required.

Computes the flux rates through various
types of boundaries and the rate at which
material increases in the region of interest.

Computes the base and weighting  functions,
their derivatives with respect to X, Y, Z, and
the Jacobian at a  Gaussian point.

Solves a matrix equation with a band matrix
solver.

Stores pertinent quantities on a auxiliary
device for future uses (e.g., for plotting).

Identifies the boundary sides and sequences
of the boundary nodes, and computes the
directional cosine  of the surface sides.

Computes moisture content at a node.

Computes the local coordinate of an element
given the global coordinate within  that ele-
ment.
                                      168

-------
TABLE A-3. FUNCTIONS INCLUDED IN 3DLEWASTE
Function           Called By          Description

FCOS              MPLOC            Computes the inner product of an outward
                                      normal of the surface with a vector connect-
                                      ing a point on the surface and the fictitious
                                      particle to determine if the fictitious particle
                                      lies inside the surface.
                                      169

-------
the implementation for such a boundary segment is bypassed. For variable bound-
aries, if the flow is directed into the region, the concentration of incoming fluid is speci-
fied. The  Lagrangian concentration is then calculated as:
                              V= /N.VnCindB/jN.VndB                       (A-12)
                                             B;

where
       C;*     = Lagrangian concentration at the boundary node i
       V  n =   normal vertically integrated Darcy velocity
       C;,,     = concentration of incoming fluid
       B      = global boundary of the region of interest
       NI     = basis function for nodal point i of element e

Specified-flux (Cauchy) boundary conditions are  normally applied to the boundary
where flow is directed into the region, where the material flux of incoming fluid is
specified. The Lagrangian concentration is thus  calculate as:
                                C;=jN,qcdB/jNiVndB                         (A-13)
where C," is the Lagrangian concentration at the boundary node i and qc is the Cauchy
flux of the incoming fluid.

A.2.2 Subroution  ADVTRN

This subroutine is called  by GM3D to compute the Lagrangian concentrations at all
nodes. It calls subroutine MPLOC to find which element a fictitious particle is located
in.  It also calls subroutine XS13D to compute the local coordinate, given the global
coordinate, of the  fictitious particle. If the fictitious particle associated with  a particu-
lar node is located in the interior of the region, the Lagrangian concentration is
is obtained by finite element interpolation of the concentration at the previous  time
step. If the fictitious particle associated with a particular node is outside the region of
interest, the Lagrangian  concentration is set equal to the previous time-step concentra-
tion of the boundary node that is  closest to the fictitious particle.

A.2.3 Subroutine  AFABTA

This subroutine, which is called by subroutine GM3D, calculates the values of up-
stream weighting factors  along 12 sides of all elements.

A.2.4 Subroutine  ALLFCT

This subroutine is called  by subroutine GM3D to compute values for all the source/sink
and boundary nodes and  elements. It uses linear interpolation of tabular data to
simulate variations in time for these conditions.

                                         170

-------
A.2.5 Subroutine  ASEMBL

This subroutine is called by subroutine GM3D. After calling subroutine Q8 to evaluate
the element matrices, it sums over all element matrices to form a global matrix
equation governing the concentration distribution at all nodes.

A.2.6 Subroutine  BC

This subroutine, which is called by subroutine GM3D, incorporates Dirichlet, variable
composite, specified-flux, and specified-dispersive flux boundary conditions. For a
Dirichlet boundary condition, an identity algebraic equation is generated for each
Dirichlet nodal point. Any other equation having this  nodal variable is modified
accordingly to simplify the  computation. For a variable composite surface, the
integration of the  normal velocity times the incoming concentration is added to the
load vector and the integration of normal velocity is added to the  matrix. For the
specified-flux boundaries, the integration of flux is added to the load vector and the
integration of normal velocity is added to the matrix.  For a specified-dispersive-flux
boundary,  the integration of gradient flux is added to the load vector.

A.2.7 Subroutine  BLKITR

This subroutine is called by subroutine GM3D to solve the matrix equation with block
iteration methods. For  each subregion, a block  matrix equation  is constructed based
on the global matrix  equation and two  pointer arrays,  GNPLR and LNOJCN (see
subroutine PAGEN).  The resulting block matrix equation is solved with the direct
band matrix solver by calling subroutine SOLVE. This is done for all subregions for
each iteration until a convergent solution is obtained.

A.2.8 Subroutine  DATAIN
Subroutine DATAIN is called by subroutine GM3D. It reads and prints all data input
described in the Section 4.2 except data set 1. It also calls subroutine SURF to
identify the boundary  segments and boundary nodes and subroutines READR and
READN, respectively,  to automatically generate real and integer numbers.

A.2.9 Subroutine  FLUX

This subroutine is called by subroutine GM3D. It calls subroutine Q8DV  to evaluate
the element matrices and the derivatives of concentrations. It then sums over all
element matrices to form a matrix equation governing the flux components at all nodal
points. To save computational time, the matrix is diagonalized by lumping. The flux
components due to dispersion can thus be solved point by point. The flux due to
velocity is then added  to  the computed flux due to dispersion. The computed total flux
field is then returned to GM3D  through the argument.

A.2.1O Subroutine GM3D

The subroutine GM3D controls  the entire sequence of operations. It performs either a
steady-state computation alone  (KSS = 0 and NTI = 0), or a transient- state computa-
                                        171

-------
tion using the steady-state solution as the initial condition (KSS = 0, NTI > 0), or a
transient computation using user-supplied initial conditions (KSS =  1, NTI > 0).

GM3D calls subroutine DATAIN to read and print input data; subroutine LELGEN to
generate the pointer array element stencil that describes all elements connected to any
node; subroutine ALLFCT to obtain sources/sinks and boundary values; subroutine
AFABTA to obtain the upstream weighting factor based on velocity  and dispersivity
(the upstream weighting factor is needed for solving the steady-state option of
3DLEWASTE); subroutine FLUX to compute  material  flux;  subroutine ASEMBL  to
assemble the element matrices over all elements; subroutine BC to implement the
boundary conditions; subroutine BLKITR to solve the resulting matrix equations with
block iteration methods; subroutine SFLOW to calculate  flux through all types of
boundaries and the water accumulated in the media; subroutine PRINTT to print out
the results; subroutine STORE to store the results for plotting; subroutine THNODE to
compute the value of moisture content plus bulk density times distribution coefficient
in the case of a linear isotherm, or the moisture content in the case  of a nonlinear
isotherm at all nodes; subroutine NDTAU to compute the number of subtime steps nd
the subtime step sizes used for integration in the Lagrangian step; ADVTRN to
compute the Lagrangian concentrations at all nodes; and subroutine ADVBC to
implement boundary conditions in the Lagrangian step.

A.2.11 Subroutine LELGEN

This subroutine is called by subroutine GM3D  to preprocess the pointer array (the
global elements stencil), LRL(K,N), where LRL(K,N) is  the global element number of
the K-th element connected to the global node N. This  pointer array is generated
based on the element connectivity IE(M,J). Here IE(M,J) is the global node number of
the J-th node of element M.  This  pointer array is needed to facilitate the location of
fictitious particles.

A.2. 12  Subroutine MPLOC
This subroutine is called by NDTAU and ADVTRN to locate the fictitious particle
associated with a particular node. It uses the function FCOS to compute the product
of the outward unit vector with the vector from a node on the surface to the fictitious
particle.

A.2.13  Subroutine NDTAU

This subroutine is called by GM3D to compute the subtime-step size and the number of
subtime steps such that no fictitious particle travels over an element within a subtime
step. The subtime-step size and the number of subtime steps are used in subroutine
ADVTRN.

A.2.14  Subroutine PAGEN

This subroutine is called by subroutine DATAIN to preprocess pointer arrays that are
needed to store the global matrix in compressed form and to construct the subregional
block matrices. The pointer arrays automatically generated in this subroutine include
the global node connectivity (stencil),  GNOJCN(J.N),  regional node connectivity,

                                        172

-------
LNOJCN(J,I,K), total node number for each subregion, NTNPLR(K), the bandwidth
indicator for each subregion, LM.AXDF(K), and a partial fill-up for the mapping array
between global node number and local subregion  node number, GNPLR(I,K), with I =
NNPLR(K) + 1 to NTNPLR(K). Here GNOJCN(J,N) is the global node number of the
J-th node connected to the global node N; LNOJN(J,I,K)  is the local node number of
the J-th node connected to the local node I in K-th subregion; NTNPLR(K) is the total
number of nodes in the the K-th subregion, including the interior nodes, the global
boundary nodes, and intra-boundary nodes; LMAXDF(K) is the maximum difference
between any two nodes of any element in  the K-th subregion; and GNPLR(I,K) is the
global node number of the I-th local-region node in the K-th subregion. These pointer
arrays are generated based on the element connectivity,  IE(M,J), the number of nodes
for each subregion,  NNPLR(K), and the mapping  between global node and local-region
node, GNLR(I,K), with 1=1, NNPLR(K).  Here IE(M,J) is the global node number of
the J-th node of element M; NNPLR(K) is the number of nodes in the K-th subregion,
including  the interior nodes and the global boundary nodes, but  not the intraboundary
nodes.

A.2. 15 Subroutine  PRINTT

This subroutine, which is called by GM3D, is used to line-print the simulation results.
These include the fluxes through variable boundary surfaces, the concentration, and
vertically  integrated material flux components.

A.2.16 Subroutine  O4ADB

This subroutine is called by subroutine ADVBC and implements Dirichlet, specified-
flux, and variable boudary conditions in a Lagrangian step computation.

A.2.17 Subroutine  O4BB

This subroutine is called by subroutine SFLOW to perform surface integration of the
following type:


                                RRQ(I)=jN;eFdB                           (A-14)
where F is the normal flux and RRQ(I) is a 3DLEWASTE program variable.

A.2.18  Subroutine O4CNB

This subroutine is called by the subroutine BC to compute the surface node flux of the
type:
                                        173

-------
                                 RQ(I)=jNieqdB                          (A-15a)
where q is either the specified- (or Cauchy) flux, specified-dispersive- (or Neumann)
flux, or n'VCv; and RQ(I) is a 3DLEWASTE program variable. It also computes the
boundary  element matrices:
                               BQ(I,J)=jNieVNjedR                        (A-15b)
where A^e is the basis function for nodal point j of element e, R is the region of
interest, V is the Darcy velocity, and BQ(I,J) is a 3DLEWASTE program variable.

A.2.  19 Subroutine 08

This subroutine is called  by the subroutine ASEMBL to compute  the element matrix
given by
                               QAdlJ)=jN1e9NjedR                         (A-16a)
                            QAA(I,J)=  N/p^N/dR                      (A-16b)
                                     R      dC
                           QB(I,J)=J(VN,6)-eD 
-------
where
       Cw    =  dissolved concentration at the previous iteration
       D     =  dispersion coefficient tensor
       0     =  moisture content
       S     =  species concentration in the adsorbed phase
       Q     =  source rate of water
       pb     =  bulk density of the porous medium
       X.     =  material decay constant
       V     =  del  operator indicating gradient
       V-     =  del  operator indicating divergence

and where QA(IJ), QAA(IJ), QB(IJ), QV(IJ), and QC(IJ) are 3DLEWASTE program
variables. Note that dS/dC should be evaluated at Cw.  Subroutines Q8 also calculates
the element load vector given by:


                                 L-[-Apb(Sw-^CJ +QCin]dR                 (A-16f)
                                 i L   r o^-  w  1 s~* W7  ^  inj
where Cw and SW are the dissolved and adsorbed concentrations at the previous
iteration,  respectively, and QR(I) is a program variable.

A.2.20 Subroutine  Q8DV

Subroutine Q8DV is called by subroutine FLUX to compute the element matrices given
by:
                                QB(I,J)=fNieNjedR
(A-17a)
Subroutine Q8DV also evaluates the element load vector:

                           QRX(I)= - jNiei-9D
-------
where
       Cj     = concentration at nodal point j
       i       = unit vector along  the x-direction
       j       = unit vector along  the y-coordinate
       k      = unit vector along  the z-coordinate

and where QRX(I), QRY(I), AND QRZ(I)  are program variables.

A.2.21 Subroutine  Q8R

This subroutine, which is called by subroutine SFLOW, is used to compute contribu-
tions to FRATE(8), FRATE(9),  FRATE(l), and FRATE(14),  discussed in Section A.2.24,
by performing material integration and element source integration over an element.
                                   QRM=J(
ecdR                            (A-18a)
                                   QDM-JsdR                             (A-18b)
                   SOSM=J[QCin(l+sign( Q))+QC(1 -sign(Q))]/2dR
where QRM, QDM, and SOSM are 3DLEWASTE program variables.

A.2.22 Subroutine  READN

This subroutine is called by subroutine DATAIN to generate integer numbers for the
input data sets if required.

A.2.23 Subroutine  READR

This subroutine is called by subroutine DATAIN to automatically generate real
numbers for the input data sets if required. Automatic generation-of regularly
patterned data is built into this subroutine.

A.2.24 Subroutine  SFLOW

This subroutine is called by subroutine GM3D. It is used to compute flux rates
through various types of boundaries and the rate at which material increases in the
region of interest.  In this subroutine, the variable FRATE(7)  stores the flux through
the whole boundary. It is given by
                                        176

-------
                             FRATE(7)=(FA +Fyny)dB                       (A-19)
where B is the global boundary of the region of interest; Fx, and Fy are the vertically
integrated flux components and n,, and iiy are the directional cosines of the outward
unit vector normal to the boundary B.   FRATE( 1) stores the flux rates through a
Dirichlet boundary Bd. FRATE(2)  and FRATE(3) store the flux rate  through specified-
flux (Cauchy)  and specified-dispersive-flux (Neumann) boundaries, respectively.
                            FRATE(1)= (F^+FHdB                      (A-20a)
                            FRATE(2)= lF^n, +Fyny)dB                      (A-20b)
                                               yy
                                       B,
                            FRATE(3)=J(Fxii, +Fyny)dB                      (A-20c)
FRATE(4) and FRATE(5) store incoming flux and outgoing flux rates, respectively,
through the variable boundaries B, and Bv+, as given by:
                            FRATE(4)= J(F)tn]t+Fyny)dB
                                      B,-




                            FRATE(5)= J(FInx+Fyny)dB                      (A-20e)
where Bv- and Bv. are that part of variable boundary where the fluxes are directed
into the region and out from the region, respectively. The integration of Equations A-
20a through A-20e is carried out by the subroutine Q4BB.

FRATE(6),  which is related to the  numerical loss, is given by:
                                        177

-------
                                             5
                       FRATE(6)=FRATE( 7)-T* FRATE(I)                 (A-21)
FRATE(8) and FRATE(9) store the accumulate rate in the dissolved and adsorbed
phases, respectively, as given by:


                               FRATE(8)=
                                         R
                               FRATE(9)=JpbSdR
FRATE(IO) stores the rate loss due to decay and FRATE(ll) through FRATE(13) are
set to zero as given by:
                           FRATE( 10)=J?t(eC +pbS)dR
                                      R
                      FRATE(11)=FRATE( 12)= FRATE(13)=0

FRATE(14) is used to store the source/sink rate as:


                FRATE(14)=J[QCin(l +sign(Q))+QC(l -sign(Q) )]/2dR
If there is no numerical error in the computation, the following equation should be
satisfied:
                                 14
                                £ FRATE(I)=0                          (A
                                1=7


and FRATE(6)  should be equal to zero.
A.2.25  Subroutine  SHAPE

This subroutine is called by subroutines Q8DV and Q8 to evaluate the value of the
base and weighting functions and their derivatives at a Gaussian point.
                                      178

-------
A.2.26 Subroutine  SOLVE

This subroutine is called by the subroutine BLKITR to solve a matrix equation of the
type:

                                      [C]{x}={y|                                (A-28)

where [C] is the coefficient matrix and {x) and {y) are two vectors, {x) is the unknown
to be  solved, and {y) is the known load vector. The computer returns the solution {y)
and stores it in {y). The computation is a standard banded Gaussian direct elimination
procedure.

A.2.27 Subroutine  STORE

This subroutine, which is called by subroutine GM3D, stores the simulation results in
a binery file for use in plotting. The information stored includes regional geometry,
concentrations, and vertically integrated material flux components at all nodes for any
desired time step.

A.2.28 Subroutine  SURF

Subroutine  SURF  is called by subroutine DATAIN. It identifies the boundary sides,
sequences the boundary nodes, and computes the directional cosine of the surface
sides. The  mappings from boundary nodes to global nodes  are stored in  NPBB(I)
(where NPBB(I) is the global node  number of the I-th boundary node). The boundary
node numbers of the four nodes for each boundary side are  stored in ISB(I,J)  (where
ISB(I,J) is the boundary node number of the  I-th node of the J-th side, I = 1 to 4).
There are six sides for each element. Which  of these six sides is the boundary side is
determined  automatically in  the subroutine SURF and is stored in ISB(5,J).  The
global element  number, to which the J-th boundary side belongs, is also preprocessed
in the subroutine SURF and is stored in  ISB(6,J). The directional  cosines of the J-th
boundary side are  computed  and stored in DCOSB(I.J)  (where DCOSB(I.J) is the
directional cosine of the J-th surface with I-th coordinate, I = 1 to 3).  The information
contained in NPBB,  ISB, and DOSB, along with the number of boundary nodes and the
number of boundary sides  is  returned  to subroutine DATAIN for other uses.

A.2.29 Subroutine  THNODE

This subroutine is called by GM3D to  compute (9 +pbdS/dC) for the linear isotherm
model or 9 for the  Freundlich and Langmuir nonlinear isotherm models.

A.2.30 Subroutine XSI3D

This subroutine is called by ADVTRN  to compute the local coordinate  of an element
given the global coordinate within that element.  With the local coordinate, the
Lagrangian concentration can then easily be interpolated from those on the nodes  of
the element.
                                        179

-------
      APPENDIX B




INPUT AND OUTPUT DEVICES
           180

-------
TABLE B-l. LOGICAL UNITS USED IN 3DFEMWATER
Logical Unit
Number
 Purpose
LUSTO
LUBAR
LUPAR
LUINP

LUOUT
11
13
14
15

16
 Logical unit for storing binary output for
 use in 3DLEWASTE or for plotting purposes.

 Logical unit for storing binary boundary
 arrays, if they are generated in the present
job, for use in subsequent executions of
 the same scenario.

 Logical unit for storing binary pointer
 arrays, if they are generated in the present
job, for use in subsequent executions of
 the same scenario.

 Logical unit for reading input data.

 Logical unit for writing output data.
TABLE  B-2. LOGICAL UNITS USED IN 3DLEWASTE
Logical Unit
Number
 Purpose
LUFLW
LUSTO
LUBAR
LUPAR
LUINP

LUOUT
11
12
13
14
15

16
 Logical unit for reading flow data from the
 3DFEMWATER  simulation.

 Logical unit for storing binary output for
 use in 3DLEWASTE or for plotting purposes.

 Logical unit for storing binary boundary
 arrays, if they are generated in the present
job, for use in subsequent executions of
 the same scenario.

 Logical unit for storing binary pointer
 arrays, if they are generated in the present
job, for use in subsequent executions of
 the same scenario.

 Logical unit for reading input data.

 Logical unit for writing output data.
                                       181

-------
                   APPENDIX C




DEFAULT VALUES FOR THE MAXIMUM CONTROL PARAMETERS
                       182

-------
TABLE  C-l. MAXIMUM CONTROL PARAMETERS  USED IN 3DFEMWATER
Parameter
Definition
Default Value
Location
Maximum Control-Integers for the Spatial Domain

MAXNPK Maximum Number of Nodes
MAXELK   Maximum Number of Elements
MXBESK Maximum Number of Boundary-Element Surfaces
MXBNPK Maximum Number of Boundary Nodal Points
MXJBDK   Maximum Number of Nonzero Elements in
           Any Row
MXKBDK Maximum Number of Elements Connecting
           to Any Node
                                        25578
                                        22080
                                         7138
                                         7140
                                          27
                 PMXSD.INC
                 PMXSD.INC
                 PMXSD.INC
                 PMXSD.INC
                 PMXSD.INC

                 PMXSD.INC
Maximum Control-Integers for the Time Domain

MXNTIK    Maximum Number of Time Steps
MXDTCK Maximum Number of DELT Changes
                                          100
                                          10
                 PMXTD.INC
                 PMXTD.INC
Maximum Control-Integers for Subregions

LTMXNK   Maximum Number of Total Nodal Points in any      3654
           Subregion, Including Interior Nodes, Global
           Boundary Nodes, and Intraboundary Nodes
LMXNPK   Maximum Number of Nodal Points in any           1218
           Subregion Including Interior Nodes and Global
           Boundary Nodes
LMXBWK Maximum Number of the Bandwidth in any          59
           Subregion
MXRGNK Maximum  Number of Subregions                   21
                                                    PMXSR.INC


                                                    PMXSR.INC


                                                    PMXSR.INC

                                                    PMXSRJNC
Maximum Control-Integers for Source/Sinks

MXSELK    Maximum Number of Source Elements
MXSPRK    Maximum Number of Source Profiles
MXSDPK    Maximum Number of Data Points on Each Element
           Source/Sink Profile
MXWNPK Maximum Number of Point (Well)  Nodal Points
MXWPRK Maximum Number of Point (Well)  Source/Sink
           Profiles
MXWDPK Maximum Number of Data Points on Each Point
           (Well) Source/Sink Profile
1
1
1
2
2
PMXSS.INC
PMXSS.INC
PMXSS.INC
PMXSS.INC
PMXSS.INC
                                                    PMXSS.INC
                                                                     (continued)
                                     183

-------
TABLE C-l.   MAXIMUM CONTROL PARAMETERS USED IN 3DFEMWATER
              (concluded)
Parameter     Definition                            Default Value        Location
Maximum Control-Integers for Specified-Flux (Cauchy) Boundary Conditions

MXCNPK Maximum Number of Specified-Flux Nodal Points      147         PMXCB.INC
MXCESK   Maximum Number of Specified-Flux Element         120         PMXCB.INC
           Surfaces
MXCPRK   Maximum Number of Specified-Flux Profiles           1          PMXCB.INC
MXCDPK Maximum Number of Data Points on Each             2          PMXCB.INC
           Specified-Flux Profile


Maximum Control-Integers for Specified-Pressure-Head Gradient  Boundary Conditions

MXNNPK Maximum Number of Specified-Pressure-Head         1          PMXNB.INC
           Gradient Nodal Points
MXNESK Maximum Number of Specified-Pressure-Head         1          PMXNB.INC
           Gradient Element Surfaces
MXNPRK Maximum Number of Specified-Pressure-Head         1          PMXNB.INC
           Gradient Flux Profiles
MXNDPK Maximum Number of Data Points on Each             2          PMXNB.INC
           Specified-Pressure-Head Gradient  Flux Refile


Maximum Control-Integers for Variable (Rainfall/Evaporation-Seepage) Boundary
Conditions

MXVNPK Maximum Number of Variable Nodal Points          2079         PMXRSB.INC
MXVESK   Maximum Number of Variable Element Surfaces     1960         PMXRSB.INC
MXVPRK   Maximum Number of Rainfall Refiles                2          PMXRSB.INC
MXVDPK Maximum Number of Data Point  on Each Rainfall       2          PMXRSB.INC
           Profile
Maximum Control-Integers for Dirichlet Boundary Conditions

MXDNPK Maximum Number of Dirichlet Nodal Points         210         PMXDB.INC
MXDPRK   Maximum Number of Dirichlet Total Head Profiles 2          PMXDB.INC
MXDDPK Maximum Number of Data Points on Each Dirichlet 2          PMXDB.INC
           Profile

Maximum Control-Integers for Material and Soil Properties

MXMATK Maximum Number of Material Types                6          PMXMS.INC
MXSPMK Maximum Number  of Soil Parameters Per Material 5          PMXMS.INC
           to Describe Soil Characteristic Curves
MXMPMK Maximum Number of Material Properties             6          PMXMS.INC
           per Material
                                      184

-------
TABLE C-2. MAXIMUM CONTROL PARAMETERS USED IN 3DLEWWASTE
Parameter Definition Default Value
Maximum Control-Integers for the Spatial Domain
MAXNPK Maximum Number of Nodes
MAXELK Maximum Number of Elements
MXBESK Maximum Number of Boundary-Element Surfaces
MXBNPK Maximum Number of Boundary Nodal Points
MXJBDK Maximum Number of Nonzero Elements in
Any Row
MXKBDK Maximum Number of Elements Connecting
to Any Node
Maximum Control-Integers for the Time Domain
MXNTIK Maximum Number of Time Steps
MXDTCK Maximum Number of DELT Changes
Maximum Control-Integers for Subregions
LTMXNK Maximum Number of Total Nodal Points in any
Subregion, Including Interior Nodes, Global
Boundary Nodes, and Intraboundary Nodes
LMXNPK Maximum Number of Nodal Points in any
Subregion, Including Interior Nodes and
Global Boundary Nodes
LMXBWK Maximum Number of the Bandwidth in any
Subregion
MXRGNK Maximum Number of Subregions
Maximum Control-Integers for Source/Sinks
MXSELK Maximum Number of Source Elements
MXSPRK Maximum Number of Source Profiles
MXSDPK Maximum Number of Data Points on Each Element
Source/Sink Profile
MXWNPK Maximum Number of Point (Well) Nodal Points
MXWPRK Maximum Number of Point (Well) Source/Sink

25578
22080
7138
7140
27
8

500
20

3654
1218
59
21

180
1
8

1
1
Location
PMXSD.INC
PMXSD.INC
PMXSD.INC
PMXSD.INC
PMXSD.INC
PMXSD.INC

PMXTD.INC
PMXTD.INC

PMXSR.INC
PMXSR.INC
PMXSR.INC
PMXSR.INC

PMXSS.INC
PMXSS.INC
PMXSS.INC

PMXSS.INC
PMXSS.INC
          Profiles
MXWDPK Maximum Number of Data Points on Each Point
          (Well)  Source/Sink Profile
PMXSS.INC
                                                                 (continued)
                                    185

-------
TABLE C-2.  MAXIMUM CONTROL  PARAMETERS USED  IN3DLEWWASTE
             (concluded)
Parameter      Definition                            Default Value       Location
Maximum Control-Integers for Specified-Flux (Cauchy) Boundary Conditions

MXCNPK Maximum Number of Specified-Flux Nodal Points      8          PMXCB.INC
MXCESK   Maximum Number of Specified-Flux Element         2          PMXCB.INC
           Surfaces
MXCPRK   Maximum Number of Specified-Flux Profiles          2          PMXCB.INC
MXCDPK Maximum Number of Data Points on Each            4          PMXCB.INC
           Specified-Flux Profile


Maximum Control-Integers for Specified-Dispersive-Flux Boundary Conditions

MXNNPK Maximum  Number of Specified-Dispersive-Flux        8          PMXNB.INC
           Nodal Points
MXNESK Maximum  Number of Specified-Dispersive-Flux        2          PMXNB.INC
           Element  Surfaces
MXNPRK Maximum  Number of Specified-Dispersive-Flux        2          PMXNB.INC
           Profiles
MXNDPK Maximum Number of Data Points on Each Specified- 4          PMXNB.INC
           Dispersive-Flux Profile


Maximum Control-Integers for Variable (Run-In/Flow-Out) Boundary Conditions

MXVNPK Maximum Number of Variable Nodal Points           38          PMXRSB.INC
MXVESK   Maximum Number of Variable Element Surfaces       18          PMXRSB.INC
MXVPRK   Maximum Number of Rainfall Profiles               2          PMXRSB.INC
MXVDPK Maximum Number of Data Point on Each Rainfall      4          PMXRSB.INC
           Profile
Maximum Control-Integers for Dirichlet Boundary Conditions

MXDNPK Maximum Number of Dirichlet Nodal Points          81         PMXDB.INC
MXDPRK  Maximum Number  of Dirichlet Total  Head Profiles 81         PMXDB.INC
MXDDPK  Maximum Number of Data Points on  Each Dirichlet 2         PMXDB.INC
           Profile
Maximum Control-Integers for Material

MXMATK Maximum Number of Material Types                6         PMXMS.INC
MXMPMK Maximum Number of Material Properties per         8         PMXMS.INC
           Material
                                      186

-------
                                  APPENDIX D

                      PROGRAM VARIABLE DESCRIPTIONS
Information about the program variables is given in two tables in this appendix.
3DFEMWATER program variables are listed in Table D-l and 3DLEWASTE program
variables are shown in Table D-2. In the tables, the definition, type, and units of each
variable are provided. In addition, the tables indicate 1) the subroutines associated
with each variable and 2) whether a variable is an input (I), output (0), or modified
(M) variable in the subroutines. Also, if a variable is included in a COMMON block,
the COMMON block name is given.

COMMON blocks are  used in 3DFEMWATER/3DLEWASTE to minimize the use of
subroutine arguments. Each COMMON block, which contains related variables, is
stored as a file separate from the 3DFEMWATER/3DLEWASTE code and is accessed
by the use of INCLUDE statements at the beginning of the main program and each
subroutine. Only those COMMON blocks needed for the execution of a subroutine are
included in the subroutine.
                                       187

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
          VARIABLE DESIGNATION

Variable

AGRAV

AKPROP
(U)





AKR(LM)




AKXG(8)

AKXYG(8)

AKXZG(8)

AKYG(8)

AKYZG(8)

AKZG(8)

BFLX(I)


BFLXP(I)

C(MAXNP)

CAPROP
(U)


Units

-

L/T






--




L/T

L/T

L/T

L/T

L/T

L/T

L3/T


L3/T

L

1/L



Type
Scalar

Array






Array




Array

Array

Array

Array

Array

Array

Array


Array

Array

Array



Description
Gravity Term Included?
0.0 = no, 1.0 = yes
I-th Parameter to Describe
the Relative Conductivity
as a Function of Pressure
Head for the J-th Material
or the I-th Data Point of
Relative Conductivity for
the J-th Material
Relative Conductivity at
the I-th Node of the M-th
Element


XX-Hydraulic Conductivity
at Eight Gaussian Points
XY-Hydraulic Conductivity
at Eight Gaussian Points
XZ-Hydraulic Conductivity
at Eight Gaussian Points
W-Hydraulic Conductivity
at Eight Gaussian Points
YZ-Hydraulic Conductivity
at Eight Gaussian Points
ZZ-Hydraulic Conductivity
at Eight Gaussian Points
Present Time Flux at the
I-th Boundary Node

Previous Time Flux at the
I-th Boundary Node
Final Solution

I-th Data Point of Water
Capacity for the J-th
Material
Sub- Common
routine Block
Q8

GW3D
DATAIN
SPROP




GW3D
VELT
SPROP
ASEMBL
BC
Q8DV
Q8
Q8DV
Q8
Q8DV
Q8
Q8DV
Q8
Q8DV
Q8
Q8DV
Q8
GW3D
SFLOW
PRINTT
GW3D
SFLOW
BLKITR
SOLVE
GW3D
DATAIN
SPROP

I.M.O
I

M
0
I




M
I
0
I
I
I
I
I
I
I
I
I
I
I
I
I
I
M
M
I
M
M
0
M
M
0
I
                               188

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
            VARIABLE DESIGNATION (continued)
Variable
Units fVD
,e Description
Sub-
routine
Common
Block
I.M.O
CHNG
CMATRX
CMTRXG
(MAXNP,
JBAND)

CMTRXL
CW(MAXNP) L

DCOSB(IJ) -
Scalar


Array





Array



Array


Array

Array
DCOSB(2,1) -
                    Array
DCOSB(3,1) -
                    Array
DCYFLX(NP) L3/T      Array



DELMAX   T        Scalar

DELT       T        Sealer
Multiplier for Increasing
DELT

An Array to Store the
assembled Global Matrix
                              Global Matrix
                             Assembled Matrix for a
                             subregion
                              X-Directional Cosine of
                              the I-th Boundary Side
          Y-Directional Cosine of
          the I-th Boundary Side
          Z-Directional Cosine of
          the I-th Boundary Side
          Darcy Flux Through the
          NP-th Variable  Boundary
          Node

          Maximum Value of DELT

          Time Increment
GW3D
VELT
ASEMBL
BC

BLKITR
                          GW3D
                          BLKITR

                          BLKITR

                          GW3D
                          DATAIN
                          SURF
                          BCPREP
                          BC
                          SFLOW
                          STORE

                          GW3D
                          DATAIN
                          SURF
                          BCPREP
                          BC
                          SFLOW
                          STORE

                          GW3D
                          DATAIN
                          SURF
                          BCPREP
                          BC
                          SFLOW
                          STORE

                          GW3D
                          BCPREP
                          PRINTT
                                                                 CREAL
                                                                 CREAL
                                                       ASEMBL CREAL
                                                       SFLOW
                                                       PRINTT
                                                                            M
                                                                            0
                                                                            0
                                                                            M
                    M
                    M

                    M

                    M
                    0
                    0
                    I
                    I
                    I
                    I

                    M
                    0
                    0
                    I
                    I
                    I
                    I

                    M
                    0
                    0
                    I
                    I
                    I
                    I

                    M
                    0
                    I
                                      189

-------
TABLE  D-l.  3DFEMWATER  PROGRAM VARIABLES, UNITS,  LOCATION, AND
              VARLABLE DESIGNATION  (continued)

Variable
DELTO
DHQ(8)

Units
T
L

Tvoe
Scalar
Array

Description
Time Increment
Pressure Difference
Sub-
routine

Q8TH
Common
Block I.M.O
CREAL
I
DJAC        L3       Scalar


DNX(8)       IL       Array



DNY(8)       1/L       Array



DNZ(8)       1/L       Array



DTH(I,M)     1/L       Array




DTHG(8)      1/L       Array



F(MAXNOD)   -        Array



FLOW(IO)     L3       Array

FLX(NP)      L3iT      Array



FRATE(IO)    I/VT      Array

F1Q(4)        L3/11/L2    Array
F2Q(4)
L8/T/L2   Array
                    Between the Present Time
                    Step and the Previous
                    Time Step at Eight Nodes
                    of the Element

                    Determinant of the
                    Jacobian

                    Partial Derivative  of the
                    Base Function with Respect
                    to x

                    Partial Derivative  of the
                    Base Function with Respect
                    to y

                    Partial Derivative  of the
                    Base Function with Respect
                    to z

                    Water Capacity at the I-th
                    Node of the M-th Element
                    Water Capacity at Eight
                    Gaussian Points of the
                    Element

                    Array of Real Numbers
                    that are to be Read and
                    Generated Automatically

                    Increment of Flow

                    Rainfall Flux Through the
                    NP-th VB Node
Flow Rate

Specified Normal Flux at
Four Nodes of the Surface

Gravity Flux at Four Nodes
of a  Specified-Pressure-
Head Gradient (Neumann)
Surface
                             BASE


                             BASE



                             BASE



                             BASE
                             Q8
                             READR
                             GW3D
                             BCPREP
                             BC
                                                 Q4S
Q4S
                                       CFLOW
                                                           CFLOW
                      0


                      0



                      0



                      0
GW3D
SPROP
ASEMBL
SFLOW
M
0
I
I
                       M
                       0
                       I
                                           190

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
            VARIABLE DESIGNATION (continued)

Variable Units
GNLR(LK)


H(N) L

HCON(NP) L

HDB(J) L

HDBF(IJ) L

HMIN(NP) L


HP(N) L
HPROP(LJ) L

HT(N) L



Type
Array


Array

Array

Array

Arr

Array


Arr
Array

Array



Description
Global Node Number of the
I-th Node in the K-th
Subregion

Pressure Head at the
Present Time

Pending Depth of the NP-th
Variable Boundary Node

Total Head of the J-th
Profile at the Present
Time
Total Head of the I-th
Data Point in the J-th
Profile
Minimum Pressure Allowed
for the NP-th VB Node


Previous-Time Pressure
Head at the N-th Node
I-th Data Point of
Pressure for the J-th
Material
Total Head the N-th Node


Sub- Common
routine Block
GW3D
DATAIN
PAGEN
BLKITR
STORE
GW3D
DATAIN
VELT
GW3D
DATAIN
BCPREP
BC
GW3D
BC

GW3D
DATAIN

GW3D
DATAIN
BCPREP
BC
GW3D
ASEMBL
GW3D
DATAIN
SPROP
GW3D
PRIN'IT
STORE
VELT

I.M.O
M
0
M
I
I
0
0
I
M
0
I
I
M
I

M
0

M
0
I
I
M
I
M
0
I
M
I
I
0
HTQ(8)       L       Arr


HW(N)       L       Array


IBUG        "       Scalar
Total Head at Eight Nodes      Q8DV
of the Element

Nonlinear Pressure Head       GW3D
Iterate  at the N-th Node

Diagnostic Print-Out           BLKITR
Indicator
M
                                       191

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES,  UNITS,  LOCATION, AND
             VARIABLE  DESIGNATION (continued)

Variable Units
ICTYP(MP) -

IDTYP(NP) -


IE(M,I)










IGEOM

IHALFB
ILUMP
IMID
INDTYP(
MXTYP)

Type
Array

Array


Array










Scalar

Scalar
Scalar
Scalar
Array


Description
Type of Specified-Flux
(Cauchy) Profile Assigned
to the MP-th Side
Total Head Profile Type
of NP-th Dirichlet Node

Global Node Number of the
I-th Node of the M-th
Element if I is Between 1
and 8, Material Type of
the M-th Element if I = 9






Geometry Description
Output Control
Half Band with Plus 1
Lumping Indicator
Mid-Difference Indicator
Array of Integers that
are to be Read or
Sub- Common
routine Block
GW3D
DATAIN
BC
GW3D
DATAIN
BC
GW3D
DATAIN
SURF
PAGEN
VELT
SPROP
BCPREP
ASEMBL
BC
SFLOW
STORE
CINTE

SOLVE
Q8 OPTN
OPTN
READN


I.M.O
M
0
I
M
0
I
M
0
I
I
I
I
I
I
I
I
I


I
I

0

INTYP(MP)
IRTYP(MP)



ISB(IJ)
          Generated  Automatically

Array      Type of Specified-
          Pressure-Head  Gradient
          (Neumann) Flux Profile
          Assigned to the
          MP-th  Neumam Side

Array      Type of Rainfall Profile
          Assigned to the MP-th
          Variable Boundary Side

Array      Boundary Node Number of
          the First Node of the
          I-th Boundary Side
GW3D
DATAIN
BC
GW3D
DATAIN
BCPREP

GW3D
DATAIN
SURF
BCPREP
BC
SFLOW
STORE
M
0
I
M
0
I

M
0
0
I
I
I
I
                                       192

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS,  LOCATION, AND
          VARIABLE DESIGNATION  (continued)

Variable Unite

ISB(2,I)






ISB(3,I)






ISB(4,I)






ISB(5,I)






ISB(6,I)






ISC(1,MP) -


ISC(2,MP) -


ISC(3,MP) -



Type
Array






Array






Array






Array






Array






Array


Array


Array



Description
Boundary Node Number of
the Second Node of the
I-th Boundary Side




Boundary Node Number of
the Third Node of the
I-th Boundary Side




Boundary Node Number of
the Fourth Node of the
I-th Boundary Side




Element Side Index of the
I-th Boundary Side:
l=left side, 2=front side,
3=right side, 4=back side,
5=bottom side, 6=top side


Element Number to which
the I-th Boundary Side
Belongs




Global Node Number of the
First Node of the MP-th
Cauchy Side
Global Node Number of the
Second Node of the MP-th
Cauchy Side
Global Node Number of the
Third Node of the MP-th
Cauchy Side
Sub- Common
routine Block

GW3D
DATAIN
SURF
BCPREP
BC
SFLOW
STORE
GW3D
DATAIN
SURF
BCPREP
BC
SFLOW
STORE
GW3D
DATAIN
SURF
BCPREP
BC
SFLOW
STORE
GW3D
DATAIN
SURF
BCPREP
BC
SFLOW
STORE
GW3D
DATAIN
SURF
BCPREP
BC
SFLOW
STORE
GW3D
DATAIN
BC
GW3D
DATAIN
BC
GW3D
DATAIN
BC

I.M.O
M
0
0
I
I
I
I
M
0
0
I
I
I
I
M
0
0
I
I
I
I
M
0
0
I
I
I
I
M
0
0
I
I
I
I
M
0
I
M
0
I
M
0
I
                                193

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
          VARIABLE DESIGNATION (continued)
Variable Units
ISC(4,MP)
ISC(5,MP)
ISN(1,MP)
ISN(2,MP)
ISN(3,MP)
ISN(4,MP)
ISN(5,MPF) --
ISTYP(MP)
ISV(1,MP)
ISV(2,MP)
ISV(3,MP)
ISV(4,MP)
Type
Array
Array
Array
Array
Array
Array
Array
Array
Array
Array
Array
Array
Description
Global Node Number at the
Fourth Node of the MP-th
Cauchy Side
Boundary Side Number of
the MP-th Cauchy Side
Global Node Number of the
First Node of the MP-th
Neumann Side
Global Node Number of the
Second Node of the MP-th
Neumann Side
Global Node Number of the
Third Node of the MP-th
Neumann Side
Global Node Number of the
Fourth Node of the MP-th
Neurnam Side
Boundary Side Number of
the MP-th Neumann Side
Source/Sink Type Assigned
to the MP-th S/S Element
Global Node Number of the
First Node of the MP-th
Variable Side
Global Node Number of the
Second Node of the MP-th
Variable Side
Global Node Number of the
Third Node of the MP-th
Variable Side
Global Node Number of the
Fourth Node of the MP-th
Variable Side
Sub- Common
routine Block
GW3D
DATAIN
BC
GW3D
DATAIN
BC
GW3D
DATAIN
BC
GW3D
DATAIN
BC
GW3D
DATAIN
BC
GW3D
DATAIN
BC
GW3D
DATAIN
BC
GW3D
DATAIN
ASEMBL
SFLOW
GW3D
DATAIN
BCPREP
GW3D
DATAIN
BCPREP
GW3D
DATA.IN
BCPREP
GW3D
DATAIN
BCPREP
I.M.O
M
0
I
M
0
I
M
0
I
M
0
I
M
0
I
M
0
I
M
0
I
M
0
I
I
M
0
I
M
0
I
M
0
I
M
0
I
                                194

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
            VARIABLE DESIGNATION (continued)

Variable Units
ISV(5,MP) -

ITIM
IWTYP(NP) -


KANALY
KCAI
KDAI
KDLAG
KDSK(I)
KDSKO
KFLOW
KGRAV
KKK

KNAI
KOUT
KPR(I)

Type
Array

Scalar
Array


Scalar
Scalar
Scalar
Scalar
Array
Scalar
Scalar
Scalar
Scalar

Scalar
Scalar
Array

Description
Boundary Node Number of
the MP-th VB Side

Time Step Number
Source/Sink Type Assigned
to the NP-th Well Node


Analytical Input Control
Analytical Specified-Flux
(Cauchy) Input Control
Analytical Dirichlet
Input Control
Diagnostic Print-Out
Table Number
Auxiliary Output Control
for the I-th Time Step;
0 = no auxiliary output
1 = output stored
Disk Output Control
System Flow Counter
Index of Gravity Control
Decomposition or Back
Substitution Indicator
1 = decomposition,
2 = back substitution
Analytical Neumann Flux
Input Control
Prin-Out Table Number
Line-printer Control for
Sub- Common
routine Block
GW3D
DATAIN
BCPREP
PRINTT
GW3D
DATAIN
ASEMBL
SFLOW
ALLFCT
CCBC
CDBC
PRINTT
GW3D
DATAIN
CINTE
SFLOW
CGEOM
SOLVE

CNBC
PRINTT
GW3D

I.M.O
M
0
I
I
M
0
I
I
I


M
M
0

I




M
M
                             I-th Time Step:
                             0 = print nothing
                             1 = print system mass
                               balance plus above
                             2 = print pressure head
                               plus above
DATAIN
BLKITR
PRIN'M'
0
I
I
                                     195

-------
TABLE  D-l.  3DFEMWATER PROGRAM VARIABLES,  UNITS, LOCATION, AND
              VARIABLE DESIGNATION (continued)
Variable      Units
LES(MP)




LMAXDF(K)




LMXBW


LMXBWK


LMXNP




LMXNPK
                      Type
KPRO
KRAI
KSAI
KSP
KSS
KWAI
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Array




Array




Scalar


Scalar


Scalar




Scalar
Description

3 = print total head plus
   above
4 = print moisture content
   plus above
5 = print Darcy velocity
   plus above

Output Control

Analytical Rainfall Input
Control

Analytical Distributed
Source/Sink Input Control

Soil Property Tabular
Input  Control

Steady-State I.C.  Control

Analytical Well Source/
Sink Input Control

Global Element Number of
the MP-th S/S Element
Maximum Difference Between
Eight Nodes of Any Element
Maximum No. of the
Bandwidth in any Subregion

Maximum No. of the Bandwidth
in Any Subregion

Maximum No. of Nodal
Points in any Subregion,
Including Interior Nodes
and Global Boundary Nodes

Maximum No. of Nodal Points
in any Subregion, Including
Interior Nodes and Global
Boundary  Nodes
                                         Sub-
                                        routine
          Common
           Block
I.M.O
                                                                        CINTE

                                                                        CVBC


                                                                          CS


                                                              SPROP     CINTE


                                                              ASEMBL CINTE

                                                                          CW
GW3D
DATAIN
ASEMBL
SFLOW
GW3D
PAGEN
BLKITR
DATAIN
M
0
I
I
M
0
I
0
BLKITR   LGEOM
BLKITR   LGEOM
                                          196

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES,  UNITS,  LOCATION, AND
             VARIABLE DESIGNATION  (continued)
Variable
Units
Tvoe
Description
Sub-
routine
Common
Block
I,M,0
LNOJCN
(J.I.K)
LRL(I,N)
LRN(I,N)
Array
Array
Array
LTMXNK




LTMXNP




LUBAR


LUINP




LUOUT
Scalar




Scalar





Scalar


Scalar




Scalar
LUPAR


LUSTO
Scalar


Scalar
Local Node No. of the J-th
Compressed Number Connect-
ing to the I-th Local
Node in the K-th Subregion

Global Element Number of
the I-th Element Connect!
ing to the N-th Global
Node

Global Node Number of the
I-th Node Connecting to
the N-th Global Node
Maximum No. of Total Nodal
Points in any Subregion,
Including Interior  Nodes,
and Global Boundary Nodes

Maximum No. of Total Nodal
Points in any Subregion,
Including Interior  Nodes,
and Global Boundary Nodes

Logical Unit for Storing
Binary Boundary Arrays

Logical Unit for Input
Data
Logical Unit for Output
Data
Logical Unit for Storing
Binary Pointer Arrays

Logical Unit for Storing
Binary Output
GW3D
PAGEN
BLKITR
DATAIN

GW3D
DATAIN
SURF
PAGEN

GW3D
PAGEN
ASEMBL
BC
DATAIN
BLKITR   LGEOM
STORE
GW3D
DATAIN

GW3D
DATAIN
READR
READN

GW3D
DATAIN
SURF
PAGEN
ASEMBL
BLKITR
PRINTT
READR
READN

GW3D
GW3D
STORE
M
0
 1
0

M
0
 1
0

M
0
 1
 1
0
                                        197

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION  (continued)
Variable Units
MAXBES
MAXBNP
MAXBw
MAXEL
MAXELK
MAXM.AT
MAXNOD
MAXNP
MAXNPK
MAXNTI
MXBESK
MXBNPK
MXCDP
MXCDPK
MXCES
MXCESK
MXCNP
Type
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Description
Maximum No. of Boundary
Element Surfaces
Maximum No. of Boundary
Nodal Points
Maximum No. of Bandwidth
Maximum No. of Elements
Maximum No. of Elements
Maximum No. of Materials
Maximum No. of Data Points
to be Read
Maximum no. of Nodal
Points
Maximum No. of Nodes
Maximum No. of Time Steps
Maximum No. of Boundary
Element Surfaces
Maximum No. of Boundary
Nodal Points
Maximum No. of Data Points
on Each Cauchy-Flux Profile
Maximum No. of Data Points
on Each Cauchy-Flux Profile
Maximum No. of Specified-
Flux (Cauchy) Element Surfaces
Maximum No. of Specified-
Flux (Cauchy) Element Surfaces
Maximum No. of Specified-
Sub-
routine
STORE
PRINTT
STORE
SOLVE
SPROP
PRINTT
STORE

SPROP
READR
SPROP
BLKITR
SOLVE
PRINTT
STORE









Common
Block I.M.O
SGEOM I
SGEOM I
I
I
SGEOM I
I
I

SMTL I
I
SGEOM I
I
I
I
I

SGEOM


CCBC

CCBC

CCBC
                          Flux (Cauchy) Nodal Points
                                  198

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION (continued)
Variable Units
MXCNPK
MXCPR
MXCPRK
MXDDP
MXDDPK
MXDNP
MXDNPK
MXDP
MXDPR
MXDPRK
MXDTCK
MXJBD
MXJBDK
MXMATK
MXMPMK
MXMPPM
MXNDP
Type
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Sub- Common
Description routine Block I,M,0
Maximum No. of Specified-
Flux (Cauchy) Nodal Points
Maximum No. of Specified- CCBC
Flux (Cauchy) Profiles
Maximum No. of Specified-
Flux (Cauchy) Profiles
Maximum No. of Data Points CDBC
on Each Dirichlet Profile
Maximum No. of Data Points
on Each Dirichlet Profile
Maximum No. of Dirichlet CDBC
Nodal Points
Maximum No. of Dirichlet
Nodal Points
Maximum No. of Data ALLFCT I
Points in any Profile
Maximum No. of Dirichlet CDBC
Total Head Profiles
Maximum No. Dirichlet Total
Head Profiles
Maximum No. of DELT Changes
Maximum No. of Nonzero BLKITR I
Elements in any Row PRINTT I
Maximum No. of Nonzero
Elements in any Row
Maximum No. of Material
Types
Maximum No. of Material
Properties per Material
Maximum No. of Material SMTL
Properties per Material
Maximum No. of Data Points CNBC
                          on Each Neumann-Flux
                          Refile
                                  199

-------
TABLE D-l.  3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
             VARIABLE DESIGNATION  (continued)
Variable
MXNDPK
MXNDTC
MXNES
MXNESK
MXNNP
MXNNPK
MXNPR
MXNPRK
MXNTIK
MXPR
MXRDP
MXREGN
MXRGNK
MXRPR
Units Type
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Description
Maximum No. of Data Points
on Each Neumann-Flux
Profile
Maximum No. of DELT
Changes
Maximum No. of Neumann
Element Surfaces
Maximum No. of Neumann
Element Surfaces
Maximum No. of Neumann
Nodal Points
Maximum No. of Neumann
Nodal Points
Maximum No. of Neumann-
Flux Profiles
Maximum No. of Neumann-
Flux Profiles
Maximum No. of Time Steps
Maximum No. of Profiles
Maximum No. of Data Points
on Each Rainfall Profile
Maximum No. of Subregions
Maximum No. of Subregions
Maximum No. of Rainfall
Sub- Common
routine Block I.M.O

SGEOM
CNBC

CNBC

CNBC


ALLFCT I
CVBC
BLKITR LGEOM I
STORE I

CVBC
MXSDP
MXSDPK
MXSEL
          Profiles

Scalar     Maximum No. of Data Points
          on Each Element Soured
          Sink Profile

Scalar     Maximum No. of Data Points
          on Each Element Source/
          Sink Profile

Scalar     Maximum No. of Source
          Elements
CS
CS
                                      200

-------
TABLE  D-l.  3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
              VARLABLE DESIGNATION (continued)
Variable      Units

MXSELK
MXSPMK
MXSPPM
MXWDPK
MXWNP
                      Scalar
                      Scalar
                      Scalar
MXSPR
MXSPRK
MXTYP
MXVDPK
MXVES
MXVESK
MXVNP
MXVNPK
MXVPRK
MXWDP
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
                      Scalar
                      Scalar
Description

Maximum No. of Source
Elements

Maximum No. of Soil
Parameters per  Material
to Describe Soil  Charac-
teristic  Curves

Maximum No. of Soil
Parameter Per Material
to Describe Soil
Characteristic Curves

Maximum No. of Source
Profiles

Maximum No. of Source
Profiles

Maximum No. of Integers
Allowed to be Read

Maximum No. of Data Points
on Each Rainfall Profile

Maximum No. of Variable
Element  Surfaces

Maximum No. of Variable
Element  Surfaces

Maximum No. of Variable
Nodal Points

Maximum No. of Variable
Nodal Points

Maximum No. of Rainfall
Profiles

Maximum No. of Data Points
on Each Well Source/Sink
Profile

Maximum No. of Data Points
on Each Well Source/Sink
Profile

Maximum No. of Well Nodal
Points
                                                              Sub-     Common
                                                             routine     Block      I.M.O
SPROP
SMTL
                                                                         CS
                                                             READN
                                                                        CVBC
                                                             PRINT!    CVBC
                                                                         CW
           CW
                                          201

-------
TABLE D-l.  3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION  (continued)
Variable Units
MXWNPK
MXWPR
MXWPRK
N
NBES
NBNP
NCDP
NCES
NCHG
NCNP
NCPR
NCYL
NDDP
NDNP
NDP
Type
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Sub- Common
Description routine Block I,M,0
Maximum No. of Well
Nodal Points
Maximum No. of Well CW
Source/Sink Profiles
Maximum No. of Well
Source/Sink Profiles
Base Functions Associated BASE 0
with 8 Nodes of the Element
Number of Boundary STORE CGEOM I
Element Surfaces
Number of Boundary Nodal STORE CGEOM I
Points
Number of Data Points on CCBC
Specified-Flux (Cauchy)
Profiles
Number of Specified-Flux CCBC
(Cauchy) Boundary Element
Sides
Number of Variable BCPREP 0
Boundary Nodes that has
Changed Boundary Conditions
Number of Specified-Flux CW
(Cauchy) Boundary Nodal
Points
Number of Specified-Flux CCBC
(Cauchy) Profiles
Number of Cycles per CINTE
Time Step
Number of Data Points on CDBC
Dirichlet Profiles
Number of Dirichlet Nodal CDBC
Points
Number of Data Points in ALLFCT I
                         Any Profile
                                 202

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
          VARIABLE DESIGNATION  (continued)
Variable Units
NDPR
NDTCHG -
NEL
NITER
NMAT
NMPPM
NNDP
NNES
NNNP
NNP
NNPLR(K) -
NNPR
NPBB(I)
Type
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Array
Scalar
Array
Description
Number of Dirichlet
Profiles
Number of Times to Reset
Time Step Size
Number of Elements
Number of Iterations per
Cycle
Number of Materials
Number of Material
Properties per Material
Number of Data Points on
Neumann-Flux Profiles
Number of Neumann Boundary
Element Sides
Number of Neumann Boundary
Nodal Points
Number of Nodal Points
Number of Node Points in
the K-th Subregion
Number of Neumann-Flux
Profiles
Global Node Number of the
I-th Boundary Node
Sub-
r o u t


SPROP
PRINTT
STORE
BLKXTR





BLKITR
SOLVE
PRINTT
STORE
READR
GW3D
DATAIN
PAGEN
BLKITR
STORE

GW3D
DATAIN
SURF
SFLOW
STORE
Common
i Htocfe
CDBC
CGEOM
CGEOM
CINTE
CMTL
CMTL
CNBC
CNBC
CNBC
CGEOM

CNBC

I.M.O








I
I
I
I
I
M
0
I
I
I

M
0
0
I
I
                                203

-------
TABLE  D-l.  3DFEMWATER PROGRAM VARIABLES,  UNITS, LOCATION, AND
              VARIABLE DESIGNATION (continued)
Variable
Units
Type
Description
Sub-
routine
Block
I.M.O
NPCB(MP)  -
NPCNV(I)
NPCON(NP)  -
Array      Global Node Number of the       GW3D
           MP-th Cauchy Node on          DATAIN
           Input, then is Changed to        SFLOW
           Contain  the Boundary Node
           Number

Array      Global Node Number of the       GW3D
           I-th Nonconvergent Node

Array      Pending  Condition              GW3D
           Indicator of the NP-th VB        BCPREP
           Node: 0  = this is not a           BC
           Pondidng-Condition Node        PRINT!
           for the Present Time Step,
           Global Node Number = this
           is a Pending-Condition Node
           for the Present Time
NPDB(
 MXDNP)
NPFLX(NP) -
NPITER


NPMIN(NP)




NPNB(MP)
NPR

NPROB
                                       GW3D
                                       DATAIN
                                       BC
                                       SFLOW

                                       GW3D
                                       BCPREP
                                       BC
                                       PRINTT
Array      Global Node Number of the
           NP-th Dirichlet Node
Array     Flux Boundary Condition
          Indicator of the NP-th VB
          Node; 0 = this is not a
          Flux-Condition Node for
          the Present Time Step,
          Global Node Number = This
          is a Flux-Condition  Node
          for the  Present Time

Scalar     Number of Blockwise
          Iterations  Allowed
Array      Minimum-Pressure  Condition     GW3D
           Indicator of the NP-th VB        BCPREP
           Node; 0 = this is not a           BC
           Minimum-Pressure-Condition     PRINTT

Array      Global Node Number of the       GW3D
           MP-th Neumann Node on        DATA-IN
           Input, then is Changed to        SFLOW
           Contain the Boundary Node
           Number

Scalar      Number of Profiles              ALLFCT

Scalar      Problem Number                STORE
                                                 CINTE
                                                             M
                                                             0
                                                             I
                                                             M
                                                             M
                                                             0
                                                             I
                                                             I
M
0
I
I

M
0
I
I
                                                             M
                                                             0
                                                             I
                                                             I

                                                             M
                                                             0
                                                             I
                                         204

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES,  UNITS,  LOCATION, AND
             VARIABLE  DESIGNATION (continued)

Variable Units
NPVB(NP)
NPW(NP)

NRDP

NREGN
NRPR

NSDP


NSEL


NSPPM


NSPR

NTI
NTNPLR(K) --

Type
Array
Array

Scalar

Scalar
Scalar

Scalar


Scalar


Scalar


Scalar

Scalar
Array

Description
Global Node Number of the
NP-th VB Node on Input,
then is Changed to Contain
the Boundary Node Number
Global Node Number of the
NP-th S/S Well Node

Number of Data Points on
Rainfall Profiles
Number of Subregions
Number of Rainfall
Profiles
Number of Data Points on
Element-Source/Sink
Profile
Number of Element-Source/
Sink and B.C. Control
Integer
Number of Soil Parameters
per Material to Describe
Soil Characteristic Curves
Number of Element-Source/
Sink Profiles
Number of Time Increments
Total Number of Nodes for
Sub- Common
routine Block
GW3D
DATAIN
SFLOW
PRINTT
GW3D
DATAIN
ASEMBL
CVBC

BLKITR LGEOM
STORE
CVBC

CS


CS


SPROP CMTL


CS

STORE CGEOM
GW3D

I.M.O
M
0
I
I
M
0
I


I
I













I
M
NTYPE


NVES


NVNP
          the K-th Subregion
          Including Interior, Global
          Boundary, and Intra-
          boundary Nodes

Scalar      Number of Integers to be
          Read

Scalar      Number of Variable Bounda-
          Element Sides

Scalar      Number of Variable
          Boundary Nodal Points
                                                         PAGEN
                                                         DATAIN
READN


CVBC


PRINTT    CVBC
                     0
                     0
                                       205

-------
TABLE D-l.  3DFEMWATER PROGRAM VARIABLES, UNITS,  LOCATION,  AND
              VARIABLE DESIGNATION  (continued)
Variable Units
NWDP
NWNP
NWPR
OME
OMI
PR(MXPR) L or
LVL2
PRF(MXDP, L or
MXPR) LVL2
PROP(U) L/T or
L'
Type
Scalar
Scalar
Scalar
Scalar
Scalar
Array
Array
Array
Description
Number of Data Points on
Each Well Source/Sink
Profile
Number of Well Source/Sink
Nodal Points
Number of Well Source/Sink
Profiles
Iteration Parameter for a
Non-Linear Equation
Relaxation Parameter for
Pointwise Solution
Profile Values at T
Profile Value of the Data
Point on the Profile
I-th Material Property of
the J-th Material;
1 = 1= saturated xx-
hydraulic conduc-
tivity
Sub- Common
routine Block
CW
CW
CW
BLKITR CREAL
CREAL
ALLFCT
ALLFCT
GW3D
DATAIN
VELT
ASEMBL
BC
I.M.O


I

0
I
M
0
I
I
I
QA(8,8)


QB(8,8)
Array


Array
1 = 2 = saturated yy-
     hydraulic conduc-
     tivity
1 = 3 = saturated zz-
     hydraulic conduc-
     tivity
1=4 = saturated xy-
     hydraulic conduc-
     tivity
1 = 5 = saturated xz-
     hydraulic conduc-
     tivity
1 = 6 = saturated yz-
     hydraulic conduc-
     tivity

Integration of N(I)
*DTH/DH*N(J)

8x8 Element Matrix
Q8
Q8DV
Q8
0
0
                                          206

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
          VARIABLE DESIGNATION  (continued)

Variable
QCB(J)

QCBF(IJ)

QNB(J)

QNBF(IJ)

QRX(8)
QRY(8)
QRZ(8)
QSOSM
QTHM
R(MAXNP)
RF(I,J)

RFALL(JJ)

RI(N)
RL

RLD(N)

RLDG(
MAXNP)

Units Tvoe
3 r
L3/T/L2 Array

LVT/L2 Array

(L3/T) Array
/L2

LVT/L' Array

Array
Array
Array
L3 Scalar
L3 Scalar
Array
L/T Array

L/T Array

L Array
Scalar

Array

Array

Description
Cauchy Flux of the J-th
Profile at the Present
Time
Cauchy Flux of the I-th
Data Point in the J-th
Profile
Neumann Flux of the J-th
Profile at the Present
Time
Neumann Flux of the I-th
Data Point in the J-th
Profile
X-Velocity Element Vector
Y-Velocity Element Vector
Z-Velocity Element Vector
Integration of SOURCE
Integration of DHQ*THG
Load Vector
Rainfall Rate of I-th
Data Point in J-th
Profile
Rainfall Rate of J-th
Profile at the Present
Time
Pressure Head Iterate in
BLKITR
A Working Array to Contain
the Final Solution of the
Pressure Head in BLKITR
An Array to Store the
Assembled Global Load
Vector
Global Load Vector
Sub- Common
routine Block
GW3D
BC

GW3D
DATAIN

GW3D
BC

GW3D
DATAIN

Q8DV
Q8DV
Q8DV
Q8TH
Q8TH
SOLVE
GW3D
DATAIN

GW3D
BCPREP

GW3D
GW3D

GW3D
ASEMBL
BC
BLKITR

I.M.O
M
I

M
0

M
I

M
0

0
0
0
0
0
M
M
0

M
I

M
M

M
0
M
I
                                207

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION  (continued)

Variable Units
RLDL(N)
RQ(8)
R1Q(4)
R2Q(4)
SOS(J) (L3/T)
/L3

SOSF(IJ) urn
L2/L
SOSM LYT
SOURCE L3/T
SS

SUBHD
T T
TDTCH(I) T
TFLOW(IO) L3
TH(I,M)



THDBF(I,J) T
THG(8)

Jjpe
Array
Array
Array
Array
Array

Array
Scalar
Scalar
Scalar

Char.
Scalar
Array
Array
Array



Array
Array

Description
Assembled Load Vector for
a Subregion
Integration of N(I).K.
(Unit Vector in Z)
Integration of N(I)*F1Q
Over the Boundary Segment
Integration of N(I)*F2Q
Over the Boundary Segment
Value of J-th Element
Source/Sink at Present
Time
S/S Rate of the I-th Data
Point in the J-th Profile
Source/Sink Strength of
the Element
Element Source/Sink
Strength
Xsi-Coordinate of the
Gaussian Point
Subheading
Time
Time of the I-th Time to
Reset the Time Step Size
to Initial Time Step Size
Total Flow
Moisture Content, at the
I-th Node of the M-th
Element


Time of the I-th Data
Point in J-th Head Profile
Moisture Content at Eight
Sub- Common
routine Block
GW3D
BLKITR
Q8
Q4S
Q4S
GW3D
ASEMBL
SFLOW
GW3D
DATAIN
Q8
Q8TH
BASE

PRINTT
ALLFCT
GW3D
DATAIN
CFLOW
GW3D
SPROP
SFLOW
PRINTT
STORE
GW3D
DATAIN
Q8TH

I.M.O
M
M
0
0
0
M
I
I
M
0
I
I
I

I
I
M
0

M
0
I
I
I
M
0
I
                          Gaussian Points of the
                          Element
                                  208

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS,  LOCATION,  AND
          VARIABLE DESIGNATION (continued)
Variable
THPROP
(U)
TIME
TITLE
TMAX
TOLA
TOLB
TPRF(
MXDP,
MXPR)
TQCBF(IJ)
TQNBF(LJ)
TRF(IJ)
TSOSF(IJ)
TT
TWSSF(IJ)
UU
VX(N)
Units

T
--
T
L
L
T
T
T
T
T
--
T
--
L/T
Type
Array
Scalar
Char.
Scalar
Scalar
Scalar
Array
Array
Array
Array
Array
Scalar
Array
Scalar
Array
Description
I-th Parameter to Describe
the Moisture Content as a
Function of Pressure Head
for the J-th Material or
I-th Data Point of Moisture
Content for the J-th
Material
Real Simulation Time
Title of the Problem
Maximum Value of Time
Steady-State Tolerance
Transient State Tolerance
Time of the Data Point
on the Profile
Time of the I-th Data
Point in the J-th
Specified-Flux (Cauchy)
Profile
Time of the I-th Data
Point in the J-th Neumann
Profile
Time of the I-th Date
Point in J-th Rainfall
Profile
Time of the I-th Data
Point in the J-th Profile
Eta-Coordinate of the
Gaussian Point
Time of the I-th Data
Point in the J-th Profile
Zeta-Coordinate of the
Gaussian Point
X-Component Velocity at
the N-th Node
Sub- Common
routine Block
GW3D
DATAIN
SPROP
PRINTT
STORE
STORE
CREAL
CREAL
BLKITR CREAL
ALLFCT
GW3D
DATAIN
GW3D
DATAIN
GW3D
DATAIN
GW3D
DATAIN
BASE
GW3D
DATAIN
BASE
GW3D
VELT
BCPREP
I.M.O
M
0
I






M
0
M
0
M
0
M
0
I
M
0
I
0
0
I
                                209

-------
TABLE D-l.  3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
             VARIABLE  DESIGNATION (continued)

Variable Units Tvnp Description



VY(N) L/T Array Y-Component Velocity at
the N-th Node




VZ(N) UT Array Z-Component Velocity at
the N-th Node




Sub- Common
routine Block
SFLOW
PRINT!
STORE
GW3D
VELT
BCPREP
SFLOW
PRINTT
STORE
GW3D
VELT
BCPREP
SFLOW
PRINTT
STORE

I.M.O
I
I
I
0
0
I
I
I
I
0
0
I
I
I
I
W(8)         -       Array
WSS(J)       L3/T     Array
WSSF(LJ)     I/YT     Array
X(N)
Array
Weighting Function at
Eight Points of the
Element

Value of the J-th Well
Source/Sink  at Present
Time

S/S Rate of the I-th Data
Point in the J-th Profile

X-Coordinate of the N-th
Node
XQ(8)
Array
X-Coordinate at Eight
Nodes of the Element
                                    ASEMBL   CREAL
GW3D                M
SFLOW               I
ASEMBL              I

GW3D                M
DATAIN               0

GW3D                M
DATAIN               0
SURF                I
VELT                I
BCPREP              I
ASEMBL              I
BC                   I
SFLOW               I
STORE                I

Q8DV                I
Q8                   I
BASE                I
Q4S                  I
Q8TH                I
                                       210

-------
TABLE D-l. 3DFEMWATER PROGRAM VARIABLES, UNITS, LOCATION, AND
          VARIABLE DESIGNATION (continued)

Variable Units TVHP Description

Y(N) L Array Y-Coordinate of the N-th
Node







YQ(8) L Array Y-Coordinate at Eight
Nodes of the Element



Z(N) L Array Z-Coordinate of the N-th
Node







ZQ(8) L Array Z-Coordinate at Eight
Nodes of the Element



Sub- Common
routine Block

GW3D
DATAIN
SURF
VELT
BCPREP
ASEMBL
BC
SFLOW
STORE
Q8DV
Q8
BASE
Q4S
Q8TH
GW3D
DATAIN
SURF
VELT
BCPREP
ASEMBL
BC
SFLOW
STORE
Q8DV
Q8
BASE
Q4S
Q8TH

I.M.O
M
0
I
I
I
I
I
I
I
I
I
I
I
I
M
0
I
I
I
I
I
I
I
I
I
I
I
I
                                211

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES, UNITS,  LOCATION,  AND
              VARIABLE  DESIGNATION
Variable      Unite     Type

             L        Scalar
AL
AM
APHA1
APHA2
APHA3
APHA4
AT
BETA1
BETA2
BETAS
BETA4
             L'/T      Scalar
                       Scalar
                       Scalar
                       Scalar
                       Scalar
BFLX(I)       M/T      Array
BFLXP(I)     WT      Array
Description

Longitudinal  Dispersivity
Modecular  Diffusion
Coefficient

Weighting Factor for
Side 1-2 Parallel to
the X-direction

Weighting Factor for
Side 4-3 Parallel to
the X-direction

Weighting Factor for
Side 5-6 Parallel to
the X-direction

Weighting Factor for
Side 8-7 Parallel to
the X-direction
                       Scalar      Lateral Dispersivity
                       Scalar      Weighting Factor for
                                  Side  1-4 Parallel to
                                  the Y-direction

                       Scalar      Weighting Factor for
                                  Side  2-3 Parallel to
                                  the Y-direction

                       Scalar      Weighting Factor for
                                  Side  5-8 Parallel to
                                  the Y-direction

                       Scalar      Weighting Factor for
                                  Side  6-7 Parallel to
                                  the Y-direction
                                  Boundary Flux at the I-th
                                  Boundary Node

                                  Value of BFI.X(I) at the
                                  Previous Time
 Sub-
routing

Q8DV
Q8

Q8
                              Q8DV
                              Q8
                                                                           Common
                                                                            Block
           WETX
           WETX
           WETX
           WETX
                                         WETY
                                         WETY
                                         WETY
                                         WETY
I.M.O

   I
   I
GM3D
SFLOW
GM3D
SFLOW
M
M
M
M
BQ(4,4)
                       Array      A 2 by 2 Boundary
                                  Surface Matrix
                              Q4CNVB
                                            212

-------
TABLE D-2. 3DLEWASTE  PROGRAM VARIABLES, UNITS,  LOCATION, AND
             VARIABLE  DESIGNATION (continued)
Variable     Units     Tyne

C(N)         M/L3      Array
CDB(I)
M/L3     Array
CDBF(IJ)     M/L3     Array
CHNG
CMATRX
 (NJ)
CMTRXG
 (NJ)

CMTRXL
 (NJ)

CP(N)
CQ(8)
CSQ(8)
M/L3
Scalar


Array




Array


Array


Array
M/L3     Array
M/M     Array
CSTAR(N) M/L3     Array
                    Description

                    Concentration of the N-th
                    Node at the Present Time
Dirichlet Concentration
of the I-th Profile at
Present Time

Concentration of the I-th
Data Point in the J-th
Dirichlet Concentration
vs. Time Profile

Multiplier for Increasing
DELT

An Array to Store the
I-th Non-Zero Entry of
the N-th Equation of the
Assembled Global Matrix

Global Matrix
Assembled Matrix for a
Subregion

Concentration of the N-th
Node at the Previous Time
          Dissolved  Concentration
          at Eight Points of an
          Element

          Adsorbed  Concentration
          at Eight Points of an
          Element

          Lagrangian  Concentration
          at the N-th Node
                                        Sub-
                                       routine,

                                       GM3D
                                       FLUX
                                       BLKITR
                                       SOLVE
                                       SFLOW
                                       PRINTT
                                       STORE

                                       GM3D
                                       BC
                                       ADVBC

                                       GM3D
                                       DATAIN
                                                GM3D
                                                FLUX
                                                ASEMBL
                                                BC

                                                BLKITR
GM3D
BLKITR

GM3D
DATAIN
ASEMBL
ADVTRN

Q8DV
Q8R
                            Q8R
                                                GM3D
                                                ASEMBL
                                                ADVTRN
                                                ADVBC
                                      Common
                                       Block
                                                          CREAL
                    I.M.O

                      0
                       I
                      0
                      M
                       I
                       I
                       I

                      M
                       I
                       I

                      M
                      M
                      M
                      0
                      0
                      0
M
I

0
M
I
I

I
I
                                                  0
                                                  I
                                                  0
                                                  0
                                         213

-------
TABLE D-2. 3DLEWASTE PROGRAM  VARIABLES, UNITS,  LOCATION, AND
             VARIABLE  DESIGNATION  (continued)
Variable      Units

CVB(I)       M/L3



CVBF(LJ) M/L9




CW(N)       M/L3



CWQ(8)      M/L3




DCOSB(IJ) -
Type

Array



Array




Array



Array




Array
DCOSB(2,1) -
Array
DCOSB(3,1) -
Array
DD          L'PI!


DELMAX   T
Description

Variable Concentration of
the I-th Profile at the
Present Time

Concentration of the I-th
Data Point in the J-th
Variable Concentration
vs. Time Profile

Nonlinear Iterate of the
Concentration  at the N-th
Node

Iterate of the Dissolved
Concentration  at Eight
Gaussian Points of the
Element

X-Directional Cosine of
the I-th Boundary Side
Y-Directional Cosine of
the I-th Boundary Side
Z-Directional Cosine of
the I-th Boundary Side
Scalar


Scalar    Maximum Value of DELT
Effective  Molecular
Diffusion Coefficient
 Sub-
routine.

GM3D
BC
ADVBC

GM3D
DATAIN
GM3D
ASEMBL
Q8
GM3D
DATAIN
SURF
BC
Q4CNVB
SFLOW
ADVBC
Q4ADB

GM3D
DATAIN
SURF
BC
Q4CNVB
SFLOW
ADVBC
Q4ADB

GM3D
DATAIN
SURF
BC
Q4CNVB
SFLOW
ADVBC
Q4ADB

Q8DV
Common
 Block    I.M.O

           M
            I
            I

           M
           M
            0
            I
           M
           M
           0
            I
            I
            I
            I
            I

           M
           M
           0
            I
            I
            I
            I
            I

           M
           M
           0
            I
            I
            I
            I
            I
                                     CREAL
                                         214

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES, UNITS, LOCATION, AND
            VARIABLE DESIGNATION (continued)

Variable Units
DELT T



DELTO T
DJAC L3

DNX(8) 1/L

DNY(8) 1/L
DNZ(8) 1/L

DSDCQ(8) L31M

DTAU T
DTH(LM) 1/L
DTHG(8) 1/T
ETA

F(MAXNOD) -

FLOW M/L
FQ(d) M/L'/T

Type
Scalar



Scalar
Scalar

Array

Array
Array

Array

Scalar
Array
Array
Scalar

Array

Scalar
Array

Description
Time Increment



Time Increment
Determinant of the
Jacobian
Partial Derivative of the
Base Function with Respect
to x
Partial Derivative of the
Base Function with Respect
to Y
Partial Dervative of the
Base Function wtih Respect
to z
The Derivative of Adsorbed
Concentration with Respect
to Dissolved Concentration
at Eight Points of the
Element
Sub-Time Step Size
(TH(I,M)-THP(LM)/DELT
dTH/dt at Eight Gaussian
Points of the Element
Local Coordinate of the
Particle
Array of Real Numbers
that are to be Read and
Generated Automatically
Increment of Flow
Normal Flux at Four Points
Sub- Common
routine Block
ASEMBL CREAL
SFLOW
PRINTT
NDTAU
CREAL
SHAPE

SHAPE

SHAPE
SHAPE

Q8

NDTAU
ADVTRN
GM3D
ASEMBL
Q8
XSI3D

READR

CFLOW
Q4BB

I.M.O
I
I
I
I

0

0

0
0



0
I
M
I
I
0

0


I
FRATE
M/T
Scalar
of the Element Surface

Flow Rate
CFLOW
                                   215

-------
TABLE  D-2.  3DLEWASTE PROGRAM VARIABLES, UNITS,  LOCATION, AND
              VARIABLE  DESIGNATION (continued)

Variable
FX(N)




N(N)




FZ(N)





Units Tvne Description
J r '
(M/L2)/T Array X-Direction Material Flux
at the N-th Node



M/LVT Array Y-Direction Material Flux
at the N-th Node



M/LVT Array Z-Direction Material Flux
at the N-th Node



Sub-
routine
GM3D
FLUX
SFLOW
PRINT!
STORE
GM3D
FLUX
SFLOW
PRINTT
STORE
GM3D
FLUX
SFLOW
PRINTT
STORE
Common
Block I.M.O
0
0
I
I
I
0
0
I
I
I
0
0
I
I
I
GAMA1
GAMA2
GAMA3
GAMA4
GNLR(I,K) -
IBC
Scalar     Weighting Factor for
          Side 1-5 Parallel to
          the Z-direction

Scalar     Weighting Factor for
          Side 2-6 Parallel to
          the Z-direction

Scalar     Weighting Factor for
          Side 4-8 Parallel to
          the Z-direction

Scalar     Weighting Factor for
          Side 3-7 Parallel to
          the Z-direction

Array     Global Nodal Number of
          the I-th Local Nodal
          Number in the K-th Sub-
          region. This Array is an
          Input for 1 = 1,2, ..,
          NNPLR(K).  For I =
          NNPLR(K)+1, . . . NTNPLR(K),
          this Array is Generated
          Based on IE(NEL,8) and
          Inputted GNLR.

Scalar     Index of Boundary Condi-
          tion Type
          WETZ
          WETZ
          WETZ
          WETZ
GM3D
DATAIN
PAGEN
BLKITR
M
M
M
 I
Q4CNVB
Q4ADB
                                         216

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES,  UNITS, LOCATION, AND
              VARIABLE  DESIGNATION  (continued)

Variable Units
IBUG
IICTYP(MP) --



IDTYP(NP)



IE(M,I)
















Tvne Description
J r 	 * 	
Scalar Debugging Indicator
Array Type of Specified-Flux
(Cauchy) Profile Assigned
to the MP-th Cauchy Side

Array Type of Dirichlet Con-
centration Profile
Assigned to the NP-th
Dirichlet Node
Array Global Node Number of the
I-th Node of the M-th
Element if I is Between
1 and 8. When I = 9,
This is an Integer to
Indicate the Material
Type of the M-th
Element.








Sub-
routine, Block
BLKITR
GM3D
DATAIN
BC
ADVBC
GM3D
DATAIN
BC
ADVBC
GM3D
DATAIN
SURF
PAGEN
LELGEN
AFABTA
ASEMBL
BC
FLUX
SFLOW
STORE
THNODE
NDTAU
ADVTRN
MPLOC
ADVBC

I.M.O

M
M
I
I
M
M
I
I
M
M
I
I
I
I
I
I
I
I
I
I
I
I
I
I
IGEOM


IHALFB

ILUMP

INDTYP(
  MXTYP)


INTYP(MP)
IOPTIM
Scalar    Geometry  Description
         Output Control

Scalar    Half Band Width Plus 1

Scalar    Lumping Indicator

Array     Army of Integers that
         are to be Read or
         Generated Automatically

Array     Type of Specified-
         Dispersive-Flux (Neumann)
         Profile Assigned to the
         MP-th Neumam Side

Scalar    Optimizing  Weighting
         Factor Indicator
LELGEN   CINTE
          OPTN
SOLVE
READN
GM3D
DATAIN
BC
AFABTA   OPTN
0
M
M
I
                                        217

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE  DESIGNATION (continued)

Variable Units
ISB(1,I)



ISB(2,I)



ISB(3,I)



ISB(4,I)



ISB(5,I)


ISB(6,I)


ISC(1,MP) -

ISC(2,MP) -


Type
Array



Array



Array



Array



Array


Array


Array

Array


Description
Boundary Node Number of
the First Node of the
I-th Boundary Side


Boundary Node Number of
the Second Node of the
I-th Boundary Side


Boundary Node Number of
the Third Node of the
I-th Boundary Side


Boundary Node Number of
the Fourth Node of the
I-th Boundary Side


Element Side Index of the
I-th Boundary Side:
l=left side, 2=front side,
3=right side, 4=back side,
5=bottom side, 6=top side
Element Number to which
the I-th Boundary Side
Belongs


Global Node Number of the
First Node of the MP-th
Specified-Flux (Cauchy)
Side
Global Node Number of the
Second Node of the MP-th
Specified-Flux (Cauchy)
Side
Sub- Common
routine Block
GM3D
DATAIN
SURF
BC
SFLOW
ADVBC
GM3D
DATAIN
SURF
BC
SFLOW
ADVBC
GM3D
DATAIN
SURF
BC
SFLOW
ADVBC
GM3D
DATAIN
SURF
BC
SFLOW
ADVBC
GM3D
DATAIN
SURF
BC
SFLOW
ADVBC
GM3D
DATAIN
SURF
BC
SFLOW
ADVBC
GM3D
DATAIN
BC
ADVBC
GM3D
DATAIN
BC
ADVBC

I.M.O
M
M
0
I
I
I
M
M
0
I
I
I
M
M
0
I
I
I
M
M
0
I
I
I
M
M
0
I
I
I
M
M
0
I
I
I
M
M
I
I
M
M
I
I
                                218

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES,  UNITS, LOCATION, AND
           VARIABLE DESIGNATION (continued)

Variable Units

ISC(3,MP) -



ISC(4,MP) -



ISC(5,MP) -



ISN(1,MP) -



ISN(2,MP) --



ISN(3,MP) -



ISN(4,MP) -



ISN(5,MP) -


ISTYP(M) -



ISV(l.MP) --




Type
Array



Array



Array



Array



Array



Array



Array



Array


Array



Array




Description
Global Node Number of the
Third Node of the MP-th
Specified-Flux (Cauchy)
Side
Global Node Number of the
Fourth Node of the MP-th
Specified-Flux (Cauchy)
Side
Boundary Side Number of
the MP-th Specified-Flux
(Cauchy) Side

Global Node Number of the
First Node of the MP-th
Specified-Dispersive-Flux
(Neumann) Side
Global Node Number of the
Second Node of the MP-th
Specified-Dispersive-Flux
(Neumann) Side
Global Node Number of the
Third Node of the MP-th
Specified-Dispersive-Flux
(Neumann) Side
Global Node Number of the
Fourth Node of the MP-th
Specified-Dispersive-Flux
(Neumann) Side
Boundary Side Number of
the MP-th Neumann Side

Type of Source Profile
Assigned to the M-th
Element

Global Node Number of the
First Node of the MP-th
Variable Side

Sub- Common
routine Block

GM3D
DATAIN
BC
ADVBC
GM3D
DATAIN
BC
ADVBC
GM3D
DATAIN
BC
ADVBC
GM3D
DATAIN
BC

GM3D
DATAIN
BC

GM3D
DATAIN
BC

GM3D
DATAIN
BC

GM3D
DATAIN
BC
(IM3D
DATAIN
ASEMBL
SFLOW
GM3D
DATAIN
BC
ADVBC

I.M.O
M
M
I
I
M
M
I
I
M
M
I
I
M
M
I

M
M
I

M
M
I

M
M
I

M
M
I
M
M
I
I
M
M
I
I
                                219

-------
TABLE D-2.  3DLEWASTE PROGRAM VARIABLES,  UNITS, LOCATION, AND
            VARIABLE  DESIGNATION  (continued)

Variable Units
ISV(2,MP) --


ISV(3,MP) -


ISV(4,MP) -


ISV(5,MP) -


ITIM
IVTYP(MP) -

IWET
IWTYP(I) -

KANALY -
KCAI
KDA1
KDIAG
KDSK(I)

Type
Array


Array


Array


Array


Scalar
Array

Scalar
Array

Scalar
Scalar
Scalar
Scalar
Array

Description
Global Node Number of the
Second Node of the MP-th
Variable Side

Global Node Number of the
Third Node of the MP-th
Variable Side

Global Node Number of the
Fourth Node of the MP-th
Variable Side

Boundary Side Number of
the MP-th Variable Side


Time-Step Index
Type of Variable Concen-
tration Profile Assigned
to the MP-th Variable Side

Upstream Weighting
Indicator
Type of Source Profile
Assigned to the I-th Node

Analytical Input Control
Analytical Cauchy-Flux
Input Control
Analytical Dirichlet Input
Control
Diagnostic Output Table
Index
Store Results on Logical
Unit 12 for the I-th
Sub- Common
routine Block
GM3D
DATAIN
BC
ADVBC
GM3D
DATAIN
BC
ADVBC
GM3D
DATAIN
BC
ADVBC
GM3D
DATAIN
BC
ADVBC
PRINTT
GM3D
DATAIN
BC
ADVBC
OPTN
GM3D
DATAIN
ASEMBL
SFLOW
ALLFCT
CCBC
CDBC
PRINTT
GM3D
DATAIN

I.M.O
M
M
I
I
M
M
I
I
M
M
I
I
M
M
I
I
I
M
M
I
I

M
M
I
I
I


0
M
M
KDSKO
        Time Step? 0=no, l=yes

Scalar    Disk Output Control
CINTE
                                   220

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES,  UNITS,  LOCATION, AND
              VARIABLE  DESIGNATION (continued)
Variable Units
3KFLOW
Type
Scalar
Description
Flow Indicator
Sub-
routine
SFLOW
Common
Block
I.M.O
I
KKK




KNAI


KOUT


KPR(I)
KPRO
KRAI
KSAI
KSORP
KSS
KVI
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
KWAI



LAMBDA

LES(I)
1/T
                     -1 = Initial or Pre-
                        initial Condition
                      0 = Steady-state
                      1. Transient

           Scalar     Decomposition or  Back
                     Substitution  Indicator
                     1 = Decomposition
                     2 = Back Substitution

           Scalar     Analytical Neumann  Flux
                     Input Control

           Scalar     Output-Table Number
                     Index

           Array     Line Printing Indicator
                     for the I-th Time Step:
                     0 = print nothing
                     1 = print fluxes through
                       all types  of boundaries
                     2 = print concentration also
                     3 = print material flux also
                                   Output Control

                                   Analytical Rainfall Input
                                   Control

                                   Elemen-source Input
                                   Control

                                   Sorption Model Indicator

                                   Steady-State I.C.  Control

                                   Flow Variable Input
                                   Control
Scalar     Well Source Input Control
          0 = Tabular Input
          1 = Analytical Input

Scalar     Decay Constant

Array     Global Element Number of
          the  I-th  Element-Source
                                        SOLVE
                                                   CNBC
                                        PRINTT
0
GM3D
DATAIN
BLKITR
PRINT!
CINTE
CVBC
CELS
THNODE OPTN
ASEMBL CINTE
CINTE
CNPS
Q8
GM3D
DATAIN
ASEMBL
SFLOW
M
M
I
I



I
I


I
M
M
I
I
                                            221

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES,  UNITS, LOCATION, AND
              VARIABLE DESIGNATION (continued)
Variable
Units TV
JTP Description
Sub-
routine
Common
Block
I.M.O
LMAXDF(K)
Array
LMXBW


LMXBWK


LMXNP




LMXNPK




LNOJCN(J,
                      Scalar


                      Scalar


                      Scalar




                      Scalar




                      Array
LOCP


LRM(I.N)
Scalar


Array
LRN(I.N)
Array
Maximum No. of Difference
Between Nodes for Any
Element in the K-th Local
Region. This Array is
Generated from the Array
LNOJCN.

Maximum No. of the Band-
width in Any Subregion

Maximum No. of the Band-
width in Any Subregion

Maximum No. of Nodal
Points in Any Subregion,
Including Interior Nodes
and Global Boundary Nodes

Maximum No. of Nodal
Points in Any Subregion,
Including Interior Nodes
and Global Boundary Nodes

Local Node Number of the
J-th Node Connecting to
I-th Local Node for the
K-th Subregion. This Array
is Generated  from GNLR  and
1 =1,2,3	NNPLR(K).

Indicator of the Location
of the  Fictitious Particle

Global Element Number of
the I-th Element Connect
ing to the N-th Global
Node
Global Node Number of the
I-th Node Connecting to
the N-th Global Node
GM3D
PAGEN
BLKITR
DATAIN
                                     GM3D
                                     PAGEN
                                     BLKITR
                                     DATAIN
MPLOC
GM3D
LELGEN
NDTAU
ADVTRN
MPLOC
DATAIN
PAGEN
SURF

GM3D
PAGEN
ASEMBL
BC
NDTAU
DATAIN
                                                                                M
                                                                                0
                                                                                 I
                                                                                M
                                               LGEOM
                                               LGEOM
                      M
                      0
                      I
                      M
                                                                                M
                                                                                0
                                                                                 I
                                                                                 I
                                                                                 I
                                                                                M
                                                                                0
                                                                                 I

                                                                                M
                                                                                0
                                                                                 I
                                                                                 I
                                                                                 I
                                                                                M
                                        222

-------
TABLE  D-2. 3DLEWASTE PROGRAM VARIABLES,  UNITS, LOCATION, AND
             VARIABLE DESIGNATION (continued)

Variable Units
LTMXNK

LTMXNP

LUBAR
LUFLW

LUINP


LUOUT












LUPAR
LUSTO

Type
J r
Scalar

Scalar

Scalar
Scalar

Scalar


Scalar












Scalar
Scalar

Description
Maximum No. of Total Nodal
Points in Any Subregion,
Including Interior Nodes,
and Global Boundary Nodes
Maximum No. of Total Nodal
Points in Any Subregion,
Including Interior Nodes,
and Global Boundary Nodes
Logical Unit for Storing
Binary Boundary Arrays
Logical Unit for Flow Data

Logical Unit for Input
Data


Logical Unit for Output
Data












Logical Unit for Storing
Binary Pointer Arrays
Logical Unit for Storing
Binary Output
Sub- Common
routine Block


LGEOM

GM3D
DATAIN
GM3D
DATAIN
GM3D
DATAIN
READR
READN
GM3D
DATAIN
SURF
PAGEN
LELGEN
ASEMBL
BLKITR
PRINTT
READR
READN
NDTAU
ADVTRN
XS13D
BC
GM3D
GM3D
STORE

I.M.O




I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I






I
I
I
M



MAXBES


MAXBNP
Scalar     Element Number where
         the Fictitious Particle
         is Located

Scalar     Maximum No. of Boundary
         Element Surfaces

Scalar     Maximum No. of Boundary
         Nodal Points
XS13D
         SGEOM
         SGEOM
                                     223

-------
TABLE D-2. 3DLEWASTE  PROGRAM VARIABLES, UNITS, LOCATION, AND
              VARIABLE DESIGNATION (continued)
Variable Units
MAXBW
MAXEL
MAXELK
MAXMAT
MAXNOD
MAXNP
MAXNPK
MP
MXBESK
MXBNPK
MXCDP
Type
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Sub- Common
Description routine Block I,M,0
Maximum No. of Band SOLVE I
Width
Maximum No. of Elements SGEOM
Maximum No. of Elements
Maximum No. of Materials MAIL
Maximum No. of Data READR I
Points to be Read
Maximum No. of Nodal SOLVE SGEOM I
Points FCOS I
Maximum No. of Nodes
Element where the MPLOC 0
Fictitious Particle
is Located
Maximum No. of Boundary-
Element Surfaces
Maximum No. of Boundary
Nodal Points
Maximum No. of Data Points CCBC
MXCDPK



MXCES


MXCESK


MXCNP


MXCNPK


MXCPR
         on Each Specified-Flux
         (Cauchy)  Profile

Scalar    Maximum No. of Data Points
         on Each Specified-Flux
         (Cauchy)  Profile

Scalar    Maximum No. of Cauchy
         Element  Surfaces

Scalar    Maximum No. of Cauchy
         Element  Surfaces

Scalar    Maximum No. of Cauchy
         Nodal Points

Scalar    Maximum No. of Cauchy
         Nodal Points

Scalar    Maximum No. of Cauchy-
         Flux Profiles
CCBC
CCBC
CCBC
                                       224

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES, UNITS, LOCATION, AND
            VARIABLE DESIGNATION (continued)
Variable Units
MXCPRK
MXDDP
MXDDPK
MXDNP
MXDNPK
MXDP
MXDPR
MXDPRK
MXDTC
MXDTCK
MXJBD
MXJBDK
MXKBD
MXKBDK
MXMATK
MXMPMK
MXMPPM
Type
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Sub- Common
Description routine Block I,M,0
Maximum No. of Cauchy-
Flux Profiles
Maximum No. of Data Points CDBC
on Each Dirichlet Profile
Maximum No. of Data Points
on Each Dirichlet Profile
Maximum No. of Dirichlet CDBC
Nodal Points
Maximum No. of Dirichlet
Nodal Points
Maximum No. of Data ALLFCT I
Points in Any Profile
Maximum No. of Dirichlet CDBC
Total Head Profiles
Maximum No. of Dirichlet
Total Head Profiles
Maximum No. of DELT SGEOM
Changes
Maximum No. of DELT
Changes
Maximum No. of Nonzero SGEOM
Elements in Any Row
Maximum No. of Nonzero
Elements in Any Row
Maximum No. of Elements SGEOM
Surrounding a Global Node
Maximum No. of Elements
Surrounding a Global Node
Maximum No. of Material
Types
Maximum No. of Material
Properties per Material
Maximum No. of Material MATE
                          Properties per Material
                                  225

-------
TABLE D-2. 3DLEWASTE  PROGRAM VARIABLES, UNITS,  LOCATION,  AND
              VARIABLE DESIGNATION  (continued)
Variable

MXNDP
MXNDPK
Units
MXSDP
MXSDPK
          Type

          Scalar
          Scalar
MXNES
MXNESK
MXNNP
MXNNPK
MXNPR
MXNPRK
MXNTI
MXNTIK
MXPR
MXRDP
MXREGN
MXRGNK
MXRPR
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
          Scalar
          Scalar
Description

Maximum No. of Data Points
on Each Specified
Dispersive-Flux  (Neumann)
Profile

Maximum No. of Data Points
on Each Neumann-Flux
Profile

Maximum No. of Neumann
Element  Surfaces

Maximum No. of Neumann
Element  Surfaces

Maximum No. of Neumann
Nodal Points

Maximum No. of Neumann
Nodal Points

Maximum No. of Neumann-
Flux Profiles

Maximum No. of Neumann-
Flux Profiles

Maximum No. of Time Steps

Maximum No. of Time Steps

Maximum No. of Profiles

Maximum No. of Data Points
on Each Rainfall Profile

Maximum No. of Subregions

Maximum No. of Subregions

Maximum No. of Rainfall
Profiles

Maximum No. of Data Points
in Any Element Source/Sink
Profile

Maximum No. of Data Points
in Any Element Source/
Sink Profile
  Sub-
routine
Common
 Block     I.M.O

 CNBC
                                                                      CNBC
                                                                      CNBC
                                                                      CNBC
                                                                     SGEOM
                                                           ALLFCT
           CVBC


          LGEOM



           CVBC


           CELS
                                         226

-------
TABLE D-2. 3DLEWASTE  PROGRAM  VARIABLES,  UNITS, LOCATION, AND
             VARIABLE DESIGNATION  (continued)
Variable Units
MXSEL
MXSELK
MXSPR
MXSPRK
MXTYP
MXVDPK
MXVES
MXVESK
MXVNP
MXVNPK
MXVPRK
MXWDP
Type
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Scalar
Sub- Common
Description routine Block I,M,0
Maximum No. of Source CELS
Elements
Maximum No. of Source
Elements
Maximum No. of Element CELS
Source Profiles
Maximum No. of Element
Source Profiles
Maximum No. of Integers READN I
Allowed to be Read
Maximum No. of Data Points
on Each Rainfall Profile
Maximum No. of Variable CVBC
Element Surfaces
Maximum No. of Variable
Element Surfaces
Maximum No. of Variable CVBC
Nodal Points
Maximum No. of Variable
Nodal Points
Maximum No. of Rainfall
Profiles
Maximum No. of Data Points CNPS
MXWDPK



MXWNP


MXWNPK


MXWPR
         on Each Well Source/Sink
         Profile

Scalar     Maximum No. of Data Points
         on Each Well Source/Sink
         Profile

Scalar     Maximum No. of Well Nodal
         Points

Scalar     Maximum No. of Well Nodal
         Points

Scalar     Maximum No. of Well
         Source/Sink Profile
CNPS
CNPS
                                      227

-------
TABLE  D-2. 3DLEWASTE PROGRAM  VARIABLES, UNITS,  LOCATION,  AND
             VARIABLE  DESIGNATION (continued)

Variable Units
MXWPRK

N(8)
NBES
NBNP
NCDP

NCES
NCM

NCNP
NCPR
NDDP

NDNP

NDP

NDPR

NDTCHG

NEL
NITER

Tj/pe
Scalar

Array
Scalar
Scalar
Scalar

Scalar
Scalar

Scalar
Scalar
Scalar

Scalar

Scalar

Scalar

Scalar

Scalar
Scalar

Description
Maximum No. of Well
Source/Sink Profiles
Base Function of Eight
Points of the Element
Number of Boundary Element
Surfaces
Number of Boundary Nodal
Points
Number of Data Points on
Specified-Flux (Cauchy)
Profiles
Number of Cauchy Boundary
Element Sides
Number of. Cycles per Time
step
Number of Cauchy Boundary
Nodal Points
Number of Specified-Flux
(Cauchy) Profiles
Number of Data Points on
Dirichlet Profiles
Number of Dirichlet Nodal
Points
Number of Data Points in
Any Profile
Number of Dirichlet
Profiles
Number of Times to Reset
Time Step Size
Number of Elements
Number of Iterations per
Sub- Common
routine Block

SHAPE
CGEOM
CGEOM
CCBC

CCBC
CINTE

CCBC
CCBC
CDBC

CDBC

ALLFCT

CDBC

CGEOM

CGEOM
BLKITR CINTE

I.M.O

0













I






I
NMAT
Scalar
Cycle

Number of Materials
MATL
                                     228

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION (continued)

Variable Units
NMPPM

NNDP

NNES
NNNP
NNP

NNPLR(K) -

NNPR
NODENP -
NP1
NP2
NP3
NPBB(I)




NPCB(NP) -

Type
Scalar

Scalar

Scalar
Scalar
Scalar

Array

Scalar
Scalar
Scalar
Scalar
Scalar
Array




Array

Description
Number of Material
Properties per Material
Number of Data Points on
Specified-Dispersive-Flux
(Neumann) Profiles
Number of Neumann Boundary
Element Sides
Number of Neumann Boundary
Nodal Points
Number of Nodal Points

Number of Nodes for the
K-th Subregion Including
Interior and Global
Boundary Nodes
Number of Specified-
Dispersive-Flux (Neumann)
Profiles
Nodal Point of Interest
First Node on the Surface
Second Node on the Surface
Third Node on the Surface
Global Node Number on the
I-th Boundary Node




Global Nodal Number of
the NP-Cauchy Node on
Input. Then it is Changed
to Contain the Boundary
Sub- Common
routine Block
MATL

CNBC

CNBC
CNBC
SOLVE CGEOM
READR
GM3D
DATAIN
PAGEN
BLKITR
CNBC
MPLOC
FCOS
FCOS
FCOS
GM3D
DATAIN
SURF
SFLOW
NDTAU
ADVTRN
ADVBC
GM3D
DATAIN
SFLOW
ADVBC

I.M.O





I
I
M
M
I
I

I
I
I
I
M
M
0
I
I
I
I
M
M
I
I
                                229

-------
TABLE D-2. 3DLEWASTE  PROGRAM VARIABLES, UNITS, LOCATION, AND
             VARIABLE DESIGNATION  (continued)

Variable Units
NPDB(NP)

NPITER

NPNB(NP)

NPR
NPROB
NPVB(NP)

NPW(I)


NRDP

NREGN
NRPR

Type
Array

Scalar

Array

Scalar
Scalar
Array

Array


Scalar

Scalar
Scalar

Description
Global Node Number of the
NP-Dirichlet Node on
Input. Then it is Changed
to Contain the Boundary
Node Number
Number of Blockwise
Iterations Allowed
Global Nodal Number of
the NP-Neumann Node on
Input. Then it is Changed
to Contain the Boundary
Node Number.
Number of Profiles
Problem Number
Global Nodal Number of
the NP-Variable Node on
Input. Then it is Changed
to Contain the Boundary
Node Number
Global Node Number of the
I-th Well Node


Number of Data Points on
Rainfall Profiles
Number of Subregions
Number of Rainfall
Sub- Common
routine Block
GM3D
DATAIN
BC
ADVBC
SFLOW
CINTE

GM3D
DATAIN
SFLOW
ALLFCT
STORE
GM3D
DATAIN
SFLOW
ADVBC
GM3D
DATAIN
ASEMBL
SFLOW
CVBC

LGEOM
CVBC

I.M.O
M
M
I
I
I


M
M
I
I
I
M
M
I
I
M
M
I
I




NSDP
Scalar
NSEL
NSPR
NTAU
Scalar
Scalar
Scalar
Profiles

Number of Data Points
in Any Element Source/
Sink Profile

Number of Source/Sink
Elemente

Number of Source/Sink
Profiles

Number of Subtime Steps
CELS



CELS


CELS
                                                        NDTAU
                                                        ADVTRN
                                                      0
                                                      I
                                      230

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION (continued)

Variable Units
NTI
NTNPLR(K) -
NTYPE
NVES
NVNP
NWDP

NWNP
NWPR

OME

OMI

PR(MXPR) L/T,L,
M/L3
PRF(MXDP, L/T,L,
MXPER) M/L8
PROP (1,1) L3/M





Type
Scalar
Array
Scalar
Scalar
Scalar
Scalar

Scalar
Scalar

Scalar

Scalar

Array
Array
Array





Description
Number of Time Increments
Total Number of Nodes for
the K-th Subregion Includ-
ing Interior, Global
Boundary, and Intraboundary
Nodes
Number of Integers to be
Read
Number of Variable Boundary
Element Sides
Number of Variable Boundary
Nodal Points
Number of Data Points in
Any Point-Source Profile
Number of Wells
Number of Well Source
Profiles
Iteration Parameter for a
Non-Linear Equation
Relaxation Parameter for
Pointwise Solution
Profile Value at Time t
Profile Value of the Data
Point on the Profile
Distribution Coefficient
or Freudlich K or Langmuir
K




Sub- Common
routine Block
CGEOM
GM3D
DATAIN
PAGEN
READN
CVBC
CVBC
CNPS

CNPS
CNPS

BLKITR CREAL

CREAL

ALLFCT
ALLFCT
GM3D
DATAIN
AFABTA
FLUX
ASEMBL
SFLOW
THNODE

I.M.O
M
M
0








I



0
I
M
M
I
I
I
I
I
                                231

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION (continued)

V a r i a b 1 dJnits TVHP Description
PROP (2, 1 M/Lg Array Bulk Density





THNODE I
PROP (3, 1) L Array Longitudinal Dispersivity






PROP (4, 1) L Array Transverse Dispersivity






PROP (5,1) L2/T Array Molecular Diffusion
Coefficient





PROP (6, 1) -- Array Tortuosity






PROP (7, 1) 1/L Array Decay Constant






Sub- Common
routine Block
GM3D
DATAIN
AFABTA
FLUX
ASEMBL
SFLOW

GM3D
DATAIN
AFABTA
FLUX
ASEMBL
SFLOW
THNODE
GM3D
DATAIN
AFABTA
FLUX
ASEMBL
SFLOW
THNODE
GM3D
DATAIN
AFABTA
FLUX
ASEMBL
SFLOW
THNODE
GM3D
DATAIN
AFABTA
FLUX
ASEMBL
SFLOW
THNODE
GM3D
DATAIN
AFABTA
FLUX
ASEMBL
SFLOW
THNODE

I.M.O
M
M
I
I
I
I

M
M
I
I
I
I
I
M
M
I
I
I
I
I
M
M
I
I
I
I
I
M
M
I
I
I
I
I
M
M
I
I
I
I
I
                               232

-------
TABLE D-2.  3DLEWASTE PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION (continued)
Variable Unite
PROP(8,I)
QA(8,8)
Q,AA(8,8)
QB(8,8)
QBMP M/LVT
QC(8,8)
QCB(I) M/L2/T
QCBF(I,J) M/T/L2
QDM
QNB(I) M/LVT
QNBF(I,J) WT/L2
QR(8)
QRM
QRX(8)
QRY(8)
QRZ(8)
Type
Array
Array
Array
Array
Scalar
Array
Array
Array
Scalar
Array
Array
Array
Scalar
Array
Array
Array
en
Freundlich N or Langmuir
SMAX
An Element Matrix
An Element Matrix
An Element Matrix
Flux or Concentration of
the Boundary Side
An Element Matrix
Value of Cauchy Flux at
the Present Time of the
I-th Cauchy Flux Profile
Flux of the I-th Data
Point in the J-th Cauchy
Flux vs. Time Profile
Integration of Local
Variable S
Value of Neumann Flux at
the Present Time of the
I-th Neumann Flux Profile
Flux of the I-th Data
Point in the J-th Neumam
Flux vs. Time Profile
An Element Load Vector
Integration of TH*C
Element Lead Vector for
x-Flux
Element Load Vector for
Y-Flux
Element Load Vector for
Sub- Common
routine Block
GM3D
DATAIN
AFABTA
FLUX
ASEMBL
SFLOW
THNODE
Q8
Q8
Q8DV
Q8
Q4CNVB
Q4ADB
Q8
GM3D
BC
ADVBC
GM3D
DATAIN
Q8R
GM3D
BC
GM3D
DATAIN
Q8
Q8R
Q8DV
Q8DV
Q8DV
I.M.O
M
M
I
I
I
I
I
0
0
0
0
I
I
0
M
I
I
M
M
0
M
I
M
M
0
0
0
0
0
                          z-Flux
                                233

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION (continued)

Variable Units
QV(8,8)
R(MAXNP)
RHOB M/L3
RI(N)

RL(N)
RLD(N)

RLDG
MAXNP)
RLDL(N)
RQ(4) M/T
RQI(4) M/L2/T
RQL(4) L3/L2/T
SOS(Ll) L3/L2/T

SOS(I,2) L7L7T

SOSC M/L3
SOSCP M/L3
SOSF L7TIL'

Type
Array
Array
Scalar
Array

Array
Array

Array
Array
Array
Array
Array
Array

Array

Scalar
Scalar
Array

Description
An Element Matrix
Load Vector
Bulk Density of the
Material in Element
Working Array Used in
Subroutines BLKITR and
ADVBC
Working Array Used in
Subroutine ADVBC
An Array to Store the
Right Hand Side of the
N-th Equation of the
Assembled Global Load
Vector
Global Load Vector
Assembled Load Vector for
a Subregion
Integrated Flux at Four
Nodes of the Element
Surface
Material-Flux at Four
Nodes of the Surface
Flow-Flux at Four Nodes
of the Surface
Source Flow Rate of the
I-th Profile at Time t

Source Concentration of
the I-th Profile at Time
t
Source Concentration
Concentration in Element-
Source
Source Flow Rate of the
I-th Data Point in the
Sub- Common
routine Block
Q8
SOLVE
Q8
GM3D
BLKITR
ADVBC
GM3D
ADVBC
GM3D
ASEMBL
BC

BLKITR
GM3D
BLKITR
Q4CNVB
Q4BB
Q4ADB
Q4ADB
GM3D
ASEMBL
SFLOW
GM3D
ASEMBL
SFLOW
Q8
Q8R
GM3D
DATAIN

I.M.O
0
M
I
M
0
M
M
M
M
0
0

I
M
I
0
0
0
0
M
I
I
M
I
I
I
I
M
M
                          J-th Profile
                                 234

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES,  UNITS, LOCATION, AND
           VARIABLE DESIGNATION (continued)
Variable Units
SOSF M/L'
(U,2)
SOSM M/T
SOSQ L3/T
SOSQP LVT
SS
SWQ(8) WT
T T
TAU
TCDBF(U) T
TCVBF(IJ) T
TDTCH(I) T
TFLOW ML
TH(I,M)
THG(8)
THN(N)
Type
Array
Scalar
Scalar
Scalar
Scalar
Array
Scalar
Scalar
Array
Array
Array
Scalar
Array
Array
Array
Description
Source Concentration of
the I-th Data Point in
the J-th Profile
Integration of Q*Cin
Element-Source Flow Rate
Element-Source Flow Rate
XSI-Coordinate of the
Gaussian Point
Iterate of the Adsorbed
Concentration at Eight
Gaussian Points of the
Element
Time
Tortuosity
Time of the I-th Data
Point in the J-th
Dirichlet Concentration
vs. Time Profile
Time of the I-th Data
Point in the J-th Vari-
able Concentration vs.
Time Profile
Time of the I-th Time
to Reset Time-Step Size
( = DELTO)
Total Flow
Moisture Content at the
I-th Node of the M-th
Element
Moisture Content at Eight
Gaussian Points of the
Element
Moisture Content at the
N-th Node
Sub- Common
routine Block
GM3D
DATAIN
Q8R
Q8
Q8R
SHAPE
Q8
ALLFCT
Q8
GM3D
DATAIN
GM3D
DATAIN
GM3D
DATAIN
CFLOW
GM3D
FLUX
ASEMBL
SFLOW
THNODE
Q8
Q8R
GM3D
THNODE
NDTAU
I.M.O
M
M
0
I
I
I
I
I
I
M
M
M
M
M
M

M
I
I
I
I
I
I
M
0
I
                                235

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION (continued)
Variable Units

THP(I,M)
THQ(8)
TIME T
TITLE
TMAX T
TOLA L
TOLB L
TPRF( T
MXDP,
MXPR)
TQCBF(LJ) T
TQNBF(LJ) T
TSOSF(IJ) T
TT
TWSSF(IJ) T
UU
VX(N) UT
Type
Array
Array
Scalar

Scalar
Scalar
Scalar
Array
bray
Array
Array
Scalar
Array
Scalar
Array
Description
Value of TH(I,M) at the
Previous Time
Moisture Content at Eight
Points of the Element
Time
Title of the Problem
Maximum Value of Time
Steady-State Tolerance
Transient-State Tolerance
Time of the Data Point on
the Profile
Time of the I-th Data
Point in the J-th Cauchy
Flux vs. Time Profile
Time of the I-th Data
Point in the J-th Neumam
Flux vs. Time Profile
Time of the I-th Data
Point in the J-th
Element Source Profile
Eta-Coordinate of the
Gaussian Point
Time of the I-th Data
Point in J-th Well
Source Profile
Zeta-Coordinate of the
Gaussian Point
X-Component Velocity at
the N-th Node
Sub- Common
routine Block
ADVTRN
ADVBC
GM3D
ASEMBL
THNODE
Q8DV
DATAIN
PRINTT
STORE
STORE
CREAL
CREAL
BLKITR CREAL
ALLFCT
GM3D
DATAIN
GM3D
DATAIN
GM3D
DATAIN
SHAPE
GM3D
DATAIN
SHAPE
GM3D
FLUX
I.M.O
I
I
M
I
I
I
M
I
I
I


I
I
M
M
M
M
M
M
I
M
M
I
M
I
                                236

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION (continued)

Variable Units






VXP(N) L/T






VXQ(8) L/T



VY(N) L/T







VYP(N) L/T






VYQ(8) L/T



VZ(N) L/T








Type Description






Array Value of VX(N) at the
Previous Time





Array X-Velocity of Eight Nodes
of the Element


Array Y-Component Velocity at
the N-th Node






Array Value of VY(N) at the
Previous Time





Array Y- Velocity of Eight Nodes
of the Element


Array Z-Component Velocity at
the N-th Node






Sub- Common
routine Block

AFABTA
ASEMBL
BC
NDTAU
ADVTRN
ADVBC
GM3D
AFABTA
ASEMBL
BC
NDTAU
ADVTRN
ADVBC
Q8DV
Q8
Q4CNVB
Q4ADB
GM3D
AFABTA
FLUX
ASEMBL
BC
NDTAU
ADVTRN
ADVBC
GM3D
AFABTA
ASEMBL
BC
NDTAU
ADVTRN
ADVBC
Q8DV
Q8
Q4CNVB
Q4ADB
GM3D
AFABTA
FLUX
ASEMBL
BC
NDTAU
ADVTRN
ADVBC

I.M.O
I
I
I
I
I
I
M
I
I
I
I
I
I
I
I
I
I
M
I
I
I
I
I
I
I
M
I
I
I
I
I
I
I
I
I
I
M
I
I
I
I
I
I
I
                                237

-------
TABLE D-2. 3DLEWASTE PROGRAM VARIABLES, UNITS, LOCATION, AND
           VARIABLE DESIGNATION (continued)

Variable Units
VZP(N) L/T






VZQ(8) L/T



W(8)


WETAB(J,
M)


WSS(Ll) LVT


WSS(I,2) M/L3


WSSF L3/T
(JJ.1)

WSSF M/L3
(J.1,2)

WWRK(N)

X(N) L










Tvoe
J r
Array






Array



Array


Array



Array


Array


Array


Array


Array

Array










Description
Value of VZ(N) at the
Previous Time





Z-Velocity of Eight Nodes
of the Element


Weighting Function at
Eight Points of the
Element
Weighting Factor for the
J-th Side of the M-th
Element

Well Source Flow Rate of
the I-th Profile at Time
t
Well Source Concentration
at the I-th Profile

Well Source Flow Rate of
the I-th Data Point in
the J-th Profile
Well Source Concentration
of the I-th Data Point in
the J-th Profile
Working Array Used in
Subroutine THNODE
X-Coordinate of the N-th
Node








Sub- Common
routine Block
GM3D
AFABTA
ASEMBL
BC
NDTAU
ADVTRN
ADVBC
Q8DV
Q8
Q4CNVB
Q4ADB
SHAPE


GM3D
AFABTA
FLUX
ASEMBL
GM3D
ASEMBL
SFLOW
GM3D
ASEMBL
SFLOW
GM3D
DATAIN

GM3D
DATAIN

GM3D
THNODE
GM3D
DATAIN
SURF
AFABTA
FLUX
ASEMBL
BC
SFLOW
STORE
THNODE

LMQ
M





I
I
I
I
I
0


M
0
I
I
M
I
I
M
I
I
M
M

M
M

M
0
M
M
I
I
I
I
I
I
I
I
                                238

-------
TABLE D-2.  3DLEWASTE PROGRAM VARIABLES, UNITS,  LOCATION,  AND
             VARIABLE DESIGNATION (continued)

Variable Units Tvnp Description





XP L Scalar X-Coordinate of the
Fictitious Particle

XQ(8) L Array X-Coordinate at Eight
Points of the Element





Sub- Common
routine Block
NDTAU
ADVTRN
MPLOC
XS13D
ADVBC
MPLOC
FCOS
XSI3D
Q8DV
Q8
SHAPE
Q4CNVB
Q4BB
Q8R
Q4ADB

I.M.O
I
I
I
I
I
I
I
I
I
0
I
I
I
I
I
XSI


Y(N)
Scalar


Array
Local Coordinate of the
Particle

Y-Coordinate of the N-th
Node
YP
YQ(8)
Scalar
Array
Y-Coordinate of the
Fictitious
Y-Coordinate at Eight
Points of the Element
XSI3D
GM3D
DATAIN
SURF
AFABTA
FLUX
ASEMBL
BC
SFLOW
STORE
THNODE
NDTAU
ADVTRN
MPLOC
FCOS
XSI 3D
ADVBC

MPLOC
FCOS
XSI3D

Q8DV
Q8
SHAPE
Q4CNVB
Q4BB
Q8R
Q4ADB
M
M
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I

 I
 I
I
0
I
I
I
I
I
                                      239

-------
       TABLE D-2. 3DLEWASTE  PROGRAM VARIABLES, UNITS, LOCATION, AND
                    VARIABLE DESIGNATION (continued)

Variable Unite TVDP Description
Z(N) L Array Z-Coordinate of the N-th
Node














ZP L Scalar Z-Coordinate of the
Fictitious Particle

ZQ(8) L Array Z-Coordinate at Eight
Points of the Element

Sub- Common
routine Block
GM3D
DATAIN
SURF
AFABTA
FLUX
ASSEMBL
BC
SFLOW
STORE
THNODE
NDTAU
ADVTRN
MPLOC
FCOS
XSI3D
ADVBC
MPLOC
FCOS
XSI3D
Q8DV
Q8
SHAPE
Q4CNVB
Q4BB
Q8R
Q4ADB

I.M.O
M
M
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
0
I
I
I
I
I
       ZTA
Scalar    Local Coordinate of the
         Particle
XSI3D
0
U.S. GOVERNMENT PRINTING OFFICE: 1992-750-002/60,103
                                             240

-------

-------
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268

Official Business
Penalty for Private Use
$300
EPA/600/R-92/223
Please make all necessary changes on the below label,
detach or copy, and return to the address in the upper
left-hand comer.

II you do not wish to receive these reports CHECK HERE
detach, or copy this cover, and return to the address in the
upper left-hand comer.

-------