oEPA
           United States
           Environmental Protection
           Agency
           Environmental Research
           Laboratory
           Dululli MN bb804
EPA 600 3 80 079
August 1980
           Research and Development
Impacts of
Coal-Fired Power
Plants on Local
Ground-Water
Systems

Wisconsin Power
Plant  Impact Study

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology.  Elimination  of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental  Health Effects Research
      2.  Environmental  Protection Technology
      3.  Ecological Research
      4.  Environmental  Monitoring
      5.  Socioeconomic Environmental Studies
      6.  Scientific and Technical  Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8.  "Special" Reports
      9.  Miscellaneous Reports

This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed  for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia  22161.

-------
                                            EPA-600/3-80-079
                                            August 1980
     IMPACTS OF COAL-FIRED POWER PLANTS
        ON LOCAL GROUND-WATER SYSTEMS

     Wisconsin Power Plant Impact  Study
                     by
             Charles B. Andrews
              Mary P. Anderson
     Institute  for Environmental Studies
       University of Wisconsin-Madison
          Madison, Wisconsin  53706
            Grant  No. R803971
               Project Officer
                Gary E. Glass
  Environmental Research  Laboratory-Duluth
              Duluth, Minnesota
This  study was  conducted  in  cooperation with

      Wisconsin  Power and  Light Company,
      Madison Gas and Electric Company,
    Wisconsin Public  Service Corporation,
    Wisconsin Public Service Commission,
and Wisconsin Department of  Natural Resources
   ENVIRONMENTAL RESEARCH LABORATORY-DULUTH
      OFFICE OF RESEARCH AND DEVELOPMENT
     U.S.  ENVIRONMENTAL PROTECTION AGENCY
           DULUTH,  MINNESOTA  55804

-------
                                 DISCLAIMER


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

-------
                                   FOREWORD
    The U.S. Environmental Protection Agency was established  to  coordinate
our country's efforts toward protecting and improving  the  environment.
Research projects in a multitude of scientific and technical  areas are
necessary to monitor changes in the environment, to  discover  relationships
within that environment, to determine health standards, and to eliminate
potential hazards.

    One such project, which the EPA is supporting through  its Environmental
Research Laboratory in Duluth, Minnesota, is the study "The Impacts  of  Coal-
Fired Power Plants on the Environment."   This interdisciplinary  study,  based
at the Columbia Generating Station, near  Portage, Wis., and involving
investigators and experiments from many academic departments  at  the
University of Wisconsin, is being carried out by the Environmental
Monitoring and Data Acquisition Group of  the Institute for Environmental
Studies at the University of Wisconsin-Madison.  Several utilities and  state
agencies are cooperating in the study:  Wisconsin Power and Light Company,
Madison Gas and Electric Company, Wisconsin Public Service Corporation,
Wisconsin Public Service Commission, and  Wisconsin Department of Natural
Resources.

    Reports from this study will be published as a series  within the EPA
Ecological Research Series.  These reports will include topics related  to
chemical constituents, chemical transport mechanisms,  biological effects,
social and economic effects, and integration and synthesis.

    This report describes the research undertaken by the Hydrogeology
Subproject of the Columbia project.  This research includes monitoring  of
ground-water flows and temperatures at the Columbia  site,  development of
mathematical models for predicting such flows, and applications  of the
models to several problems.
                                      Norbert A. Jaworski
                                      Director
                                      Environmental  Research  Laboratory
                                      Duluth, Minnesota
                                     iii

-------
                                   ABSTRACT
    Quantitative techniques  for  simulating  the  impacts  of a coal-fired power
plant on  the ground-water  system of  river flood-plain wetland in central
Wisconsin were developed and  tested  by  using  field data collected at the
site of the 500-MW  Columbia  Generating  Station.  The most important effects
were those related  to  the  construction  and  operation of the 200-ha cooling
lake and  the 28-ha  ashpit.

    Several two-dimensional  vertically  oriented steady-state models of the
ground-water flow system were used to simulate  ground-water flows before and
after the filling of the cooling lake and ashpit.   The simulations and
supporting field evidence  indicated  that  the  creation of the cooling lake
greatly altered the configuration of the  local  flow systems and Increased
the discharge of ground water to the wetland  west  of the site by a factor
of 6.

    Chemical changes in the  ground-water  system were minor.  The plume of
contaminated ground water  originating from  the  ashpit was confined to a
relatively small area  near the ashpit.  Thermal changes in the ground-water
system are a major  impact  of  the operation  of the  cooling lake inasmuch as
the lake loses water to the  ground-water  system at a rate of 2 x 10^ m^ per
day.  The wetland,  which is a major  ground-water discharge area,  has
undergone a rapid and  dramatic change in  vegetation as  a result of changes
in both water temperature  and water  levels.

    Ground-water temperatures in the vicinity of the cooling lake were
monitored in detail for 1  1/2 yr.  The  response of subsurface temperatures
temperature was simulated  by means of a mathematical model, and predictions
were made of the long-term changes expected in  substrate temperatures in the
wetland adjacent to the power plant.  In  addition, the  use of ground-water
estimate ground-water  flow rates  away from  the  cooling  lake was
investigated.

    The model, which couples  equations  describing  ground-water flow with
those describing heat  transport  in the  subsurface,  was  used to simulate the
seasonal temperature fluctuation within seven cross sections oriented
parallel to the direction  of ground-water flow  and downgradient from the
cooling lake.  Simulated temperature patterns agreed well with field data,
but were very sensitive to the distribution of  subsurface lithologies.  The
predictive simulations suggest that  by  1987 temperatures at a depth of 0.6 ra
will not fall below 8°C within 200 m of the dike,  and that peak temperatures
near the dike will  be  10-15°C above  normal and  will occur in October and
November rather than in August.   The increase in ground-water temperatures
by 1987 will also result in a 24% increase in ground-water flow.   The

                                    iv

-------
subsurface stratigraphy of  the  site  is  such  that major  changes  in
near-surface temperatures will  only  occur within 340  m  of  the dikes,  but,  if
the different, impacts could extend  to  much  greater distances from the
cooling-lake dikes.

    This report was prepared with  the cooperation  of  faculty and graduate
students in the Department  of Geology at the  University of Wisconsin-
Madison.

    Most of the funding for the research reported  here  was provided by the
U.S. Environmental Protection Agency, but funds were  also  granted by  the
University of Wisconsin-Madison, Wisconsin Power and  Light Company, Madison
Gas and Electric Company, Wisconsin  Public Service Corporation,  and
Wisconsin Public Service Commission.  This report  was submitted  in
fulfillment of Grant No. R803971 by  the Environmental Monitoring and  Data
Acquisition Group, Institute for Environmental  Studies, University of
Wisconsin-Madison  under the partial sponsorship of the U.S. Environmental
Protection Agency.

-------
                                   CONTENTS
Foreword	
Abstract	iv
Figures	viii
Tables	xii

    1.  Introduction 	    1
    2.  Conclusions and Recommendations	    4
             Recommendations for siting future power plants   	    6
             Recommendations for future research  	    6
    3.  The Site of the Columbia Generating  Station	    8
    4.  The Impact of a Power Plant on the Ground-Water  System
          of a Wetland	19
             Configuration of the ground-water flow system 	   19
             Wetland water levels	•	23
             Ground-water temperature  	   24
             Relationship between ground-water temperatures .and
               flow rates  . .  . .  •	26
             Changes in water chemistry   	   26
    5.  Thermal Alteration of Ground Water Caused by Seepage
          from the Cooling Lake	•   30
             Mathematical model  	   30
             Results	35
    6.  Long-Term Temperature Changes in  the Ground Water
          of the Wetland	41
             Simulation studies  	   41
             Results and discussion	43

References	51
Appendices

    A.  Data specifications  for  the water-flow model  •  . *	57
    B.  Techniques for determining  boundary  conditions and
          parameters in the  heat-flow model   	   61
    C.  A finite element program to simulate single-phase heat
          flow and conservative  mass  transport in a aquifer	73
    D.  Estimation of aquifer parameters  by  using subsurface
          temperature data	•  •  195
                                     vii

-------
                                   FIGURES
Number                                                                   Page

  1  The site of the Columbia Generating  Station  during
       construction of the first unit  (24 May  1971)	   9

  2  The site of the Columbia Generating  Station  after
       construction of both units  (12  August 1977)  ...........  10

  3  Main features of the site of  the  Columbia Generating Station  ...  11

  4  Water flows and energy flows  at the  site  of  the  Columbia
       Generating Station  	 .............  12

  5  Potentiometric surface contours in the area  of the
       Columbia Generating Station 	  13

  6  Representative stratigraphic  cross sections  of the  subsurface
       at the site of the Columbia Generating  Station	14

  7  Topographic map of the site of the Columbia  Generating  Station  .  .  16

  8  Peat thicknesses in the subsurface at the site of  the Columbia
       Generating Station	•	18

  9  Simulated head distribution in cross-secion  A-A1 of Figure 6 before
       filling of the Columbia cooling lake	20

 10  Simulated head distribution in cross-section A-A1  of Figure 6 after
       filling of the Columbia cooling lake	22

 11  Water levels in the wetland west  of  the Columbia cooling  lake
       before and after filling of  the lake	24

 12  Ground-water temperature variations  in wells west  of the
       Columbia cooling lake	25

 13  Stiff diagrams of water in the cooling lake  and  ground  water  ...  27

 14  Stiff diagrams of water in the ashpit and in a well cased  to 1.5 m
       below the surface and located 2 m  west  of  the  ashpit   ......  28

 15  Location of seven cross sections  of  the Columbia site for  which
       ground-water temperature distributions  were  simulated ..*••*  31

                                     vlii

-------
16  Observed temperatures in the subsurface west of the Columbia
      cooling lake and temperatures simulated by the
      mathematical model 	 ........tit.  36

17  Observed (solid line) and simulated (dashed line) ground-water
      temperature distributions in cross-section A-A1 of Figure 15 ..  37

18  Observed (solid line) and simulated (dashed line) ground-water
      temperature distributions in cross-section B-B1 of Figure 15 ..  38

19  Seasonal fluctuations of ground-water temperature in cross-section
      A-A1 of Figure 15 at a depth of 4.5 ra at distances of 2 m,  15 m,
      50 m, and 84 m west of the cooling-lake dike ..... 	  39

20  Simulated temperatures (solid lines) and observed temperatures
      (open circles) in cross-section A-A1 of Figure 15 at a depth
      of 3*28 m at various distances west of the cooling lake   ....  40

21  Cooling lake inlet temperatures for 1975-87 used in the
      simulations of long-term temperature change in the
      ground-water system  	  42

22  Predicted ground-water temperatures from 1975 to 1987 at a  distance
      of 2 m west of the cooling-lake dike in cross-section A-A'
      of Figure 15 at a depth of 0.6 m	44

23  Pre-lake temperatures and simulated temperatures 2 m west of  the
      cooling-lake dike in cross-section A-A1 of Figure 15 at a
      depth of 0.6 m	45

24  Temperatures 150 m west of the cooling-lake dike in cross-section
      A-A' of Figure 15 at depths of 0.6 m and 3 m for the simulated
      period 1975-87 	  46

25  Temperatures 150 m west of the cooling-lake dike in cross-section
      A-A' of Figure 15 at a depth of 0.6 m	47

26  Temperatures in cross-section B-B' of Figure 15 at a depth  of
      0.6 m and at (a) 2 m west of the dike, (b) 50 m
      west of the dike, and (c) 200 m west of the cooling-lake
      dike	48-50

A-l Location of monitoring wells at the Columbia site	58

B-l Location of temperature sampling points at the Columbia
      Generating Station site  	  62

B-2 A simple circuit for measuring temperature 	  ...  65

B-3 Temperature-resistance characteristics of Yellow Springs Series
      400 thermistors and the change in resistance with
      a 0.1° change in temperature  . *	67

                                   ix

-------
 B-4 Potted thermistor assembly  showing  the  details  of  the
       thermistor mounts   	   67

 B-5 Schematic of thermistor  placement in  subsurface wells at the
       Columbia Generating  Station site  	   69

 B-6 Schematic of digital  bridge circuit used  as  a portable field
       meter for measuring  ground-water temperatures  	   71

 B-7 Details of (a) the needle probe and (b) the  experimental
       arrangement to measure thermal conductivities of
       unconsolidated materials  	   72

 B-8 Thermal conductivities of three ground-water samples  from the
       Columbia Generating  Station site which  were analyzed with
       the needle probe	'	   72

 C-l The basic procedure for  linking the ground-water flow and the
       transport equations  of the model	73

 C-2 Examples of refinement of a finite element grid	81

 C-3 Examples of numbering  of nodes in a structure   ...*.  	   93

 C-4 Cartesian orientation  of finite element grid 	   94

 C-5 Typical elements of finite element grid and correct method  of
       numbering the elements 	   95

 C-6 Numbering of nodes and elements in rectangular  grid   	  .  .   95

 C-7 Two variants of the rectangular grid  that can be generated  ....   96

 C-8 Numbering of sides of an element	97

 C-9 Finite element grids used to discretize a linear heat
       transport problem	••	104

C-10 Input data used to model one-dimensional  heat transport
       with all linear elements .	•	105

C-ll Program output for one-dimensional heat transport problem with
       linear elements	•	107

C-12 Data deck used to model one-dimensional heat transport
       with mixed elements	Ill

C-l3 Program output for the one-dimensional heat transport problem
       with mixed elements	113

-------
C-14 Analytical solutions (solid lines) and numerical  solutions
       (dots) for the linear heat transport problem  for t =  25,  50,
       and 75 days	118

C-12 Areal view of the Mohawk River Valley showing location  of the
       Schenectady and Rotterdam well fields   	  121

C-16 Cross-sectional view of the Mohawk River  alluvial aquifer
       along section A-A1 of Figure C-15	121

C-17 Grid used to discretlze the Mohawk River  problem	122

C-18 Data deck used to model heat flow in  the  Mohawk River
       alluvial aquifer  	  123

C-19 Program output for  Mohawk River problem   	  125

C-20 Schematic cross section of the Columbia Generating Station  site
       along an east-west line	131

C-21 The grid used to dlscretize the cross section simulated at  the
       Columbia Generating Station site 	  131

C-22 Data deck used to model heat flow at  the  Columbia Generating
       Station site	132

C-23 Program output for  Columbia Generating Station  problem  	  134

C-24 Grid used to discretize the aquifer simulated in  the
       heat pump problem .	138

C-25 Data deck used to model the heat pump simulation  with no
       regional ground-water flow 	  139

C-26 Program output for  the heat pump problem	141

C-27 Program flow chart	147

D-l  Temperatures (°C) in a section of the marsh adjacent to the
       cooling lake at the Columbia Generating Station site  on
       7 October 1977	200

D-2  Temperatures in a 60-m portion of cross-section B-B1 of Figure 6
       adjacent to the drainage ditch east of  the cooling lake on
       9 June 1977 and 24 October 1977	201
                                    xi

-------
                                    TABLES
Number                                                                   Page
  1   Horizontal Hydraulic Conductivities  for  the  Lithologies in
        the Subsurface of the Columbia  Generating  Station  	  15

  2   Selected Chemical Concentrations  in  Ground Water  Near the
        Ashpit of the Columbia Generating  Station   .  .  *	29

  3   Parameters Used in the Simulations for the 1 to del  Describing
        Thermal Alteration of Ground-water	35

 B-l  Temperature Sampling Points:  Locations,  Depths,  liethods, and
        Frequency of Readings	63

 B-2  Heat Capacities of the Common Components  of  Unconsolidated
        Glacial Materials  	  72

 C-l  Parameters for Both Mass and Heat Transport  Problems	76

 C-2  Parameters, Initial Conditions, and  Boundary  Conditions Used  for
        the Heat-Flow Equation in the Linear Heat  Transport Problem  . . 106

 C-3  Analytical Solution at t - 50 Days for the Problem  Posed in
        Figure C-9 and Finite Element Numerical Solutions
        at t « 50 Days for Two Grid Configuations and Two Types of
        Boundary Conditions  	 117

 C-4  Parameters, Initial Conditions, and  Boundary  Conditions Used  for
        the Simulation of Temperatures in  the Mohawk River  Alluvial
        Aquifer	120

 C-5  Parameters, Initial Conditions, and  Boundary  Conditions Used  in
        the Columbia Generating Station Problem  	  .  »  i  « » 130
                                    xli

-------
                                  SECTION 1

                                 INTRODUCTION
     Large industrial facilities are common features of the American
landscape, but knowledge of the mechanisms by which they change natural
systems is inadequate.  The Hydrogeology Subprojeot of the interdisciplinary
research project "The Impacts of Coal-Fired Power Plants on the Environment"
quantified the effects of construction and operation of the Columbia
Generating Station on the hydrogeologic system of an adjacent wetland.  The
investigators developed techniques for simulating the observed changes and for
predicting potential impacts at other sites.  Impacts on the ground-water
system of the wetland resulted from the construction of a 200-ha cooling lake
and a 28-ha ashpit.  Specifically, this study assesses the alteration of three
characteristics of the ground-water system:  (1) the quantity of ground-water
flow from the cooling lake and the ashpit into the wetland and surface-water
levels in the wetland, (2) the quality of ground water and surface water as
expressed by concentrations of the common cations and anions, and (3) the
temperatures of ground water and surface water at the site.

     Data on ground-water temperatures were collected from 1971, 4 yr before
the operation of the Columbia Generating Station, through 1977.  The basic
monitoring network consisted of 100 small-diameter wells and a subsurface
network of 64 temperature monitoring points.

    During the initial phase of the study the ground-water flow systems in the
vicinity of the generating station were delineated and the way in which the
flow systems were altered by the construction and operation of a cooling lake
and ashpit were documented.  A finite difference model was used to simulate
the alterations in the configuration of the ground-water flow system.
Analyses of changes in the chemical quality and temperature of ground water
were descriptive during this phase of the study.

    The second phase of the study dealt with the monitoring and simulation of
the transfer of one byproduct of the generating process, specifically heat,
away from the site via the ground-water system.  The study of the transfer of
heat was chosen rather than the transfer of chemical byproducts because:  (1)
temperatures can be monitored inexpensively and rapidly at a large number of
points, (2) the processes of heat transfer are well understood, and (3) the
initial phase of the study had shown that alterations in the temperature of
ground water were large, whereas alterations in the chemical characteristics
of ground water were small.  The development of the capability to simulate the
transfer of heat is a prerequisite to studying the flow of a chemical away
from the site.  The processes determining the chemical composition of ground


                                     -1-

-------
water are poorly understood, and it is difficult to obtain representative
ground-water samples.

    Data on ground-water temperatures were collected for an 18-mo period
during  1976-77. in the vicinity of the cooling lake of the Columbia Generating
Station.  The  finite element method, using isoparametric quadrilateral
elements, was  used to solve the partial differential equations describing
ground-water flow and heat transport in the subsurface.  The model was tested
by comparing the observed ground-water temperatures in the vicinity of the
cooling lake with predicted temperatures.  The simulated temperatures were in
close agreement with the observed temperatures.

    The model  was then used to simulate the long-term (12-yr) effects of
seepage from the cooling lake on ground-water temperatures near the lake.  The
temperature data collected at the Columbia Generating Station site and the
model were also used to refine estimates of ground-water' flow from the cooling
lake calculated by using potentiometric data.  A third application of the
model was the  simulation of the impacts on ground-water temperatures on the
use of heat pumps for residential heating and cooling (Andrews 1978).

    Although ground-water systems have been explored for possible energy
storage or as  a source of energy, the published literature contains no
previous work  documenting the effects of a power plant on a ground-water
system.  The possibility of temporary storage of energy in the form of heated
water in aquifers has been explored by Meyer and Todd (1973), Hausz and Meyer
(1975), Kley and Nieskens (1975), Molz et al. (1976), Tsang et al. (1976), and
Werner and Kley (1977).  Gass and Lehr (1977) advocated the use of ground
water itself as an energy source.  Gringarten and Sauty (1975), Intercom?
Resource Development and Engineering (1976), and Tsang et al. (1976) developed
models of the  transport of heat in ground-water systems to study the
feasibility of temporary storage of heated water in aquifers.  Mercer et al.
(1975) modeled the movement of heat and water in a hydrothermal system.

    Alterations in the ground-water system induced by activities at a
power-plant site can bring about significant changes in nearby flora, fauna,
and surface waters.  For example, in a wetland environment, here defined as an
area where the water table is at or near the surface at all times, alterations
in ground-water quality and ground-water discharge rates have been shown to
alter wetland ecosystems (Bay 1967, Dix and Smeins 1967, Walker and Coupland
1968, Millar 1973).  In addition, Vadas et al. (1976) and Gibbons (1976) found
that changes in wetland water temperatures led to changes in wetland
ecosystems.

    Boulter (1972) studied the extent of water-table drawdown after a wetland
is ditched.  But no one has investigated the changes that occur in water
levels, water quality, or water and substrate temperatures in a wetland when a
ground-wafer system discharging into a wetland is altered.   In the wetland
adjacent to the Columbia Generating Station, wetland vegetation changed
quickly and markedly following changes in water temperatures, water levels,
and water flows caused by the presence of the cooling lake (Bedford 1977).
During the first 2 yr of operation of the Columbia Generating Station,  a
                                     -2-

-------
community previously dominated by sedge meadow species was replaced by
emergent aquatic species and annuals (Bedford 1978).

    This report deals with the hydrogeologic environment of the site of the
Columbia Generating Station and the major effects of power-plant construction
and operation; the impacts of the generating station on the ground-water
system; the simulation of heat flow in the subsurface near the site; the
expected long-term changes in ground water at the site; and the major
conclusions and recommendations of the study.  The appendices contain a
description of the field methods used in the study (appendix A and appendix
8), the computer program for solving the model of heat and water flow in
shallow ground-water systems (appendix C), and a discussion of the use of the
heat- and water-flow model to refine estimates of flow using temperature data
(appendix D).   An additional application of the model is reported by Andrews
(1978).
                                     -3-

-------
                                  SECTION 2

                       CONCLUSIONS AND RECOMMENDATIONS
    The cooling lake at the Columbia Generating Station has dramatically
altered the water supply to the marsh west of the lake.  Before the filling of
the lake, ground water, the major source of water to this area, discharged at
a rate of less than 0.03 m3/s annually.  The discharge rate increased by a
factor of 6 after the filling of the cooling lake.  The increased water supply
to the marsh has raised surface-water levels in the sedge meadow approximately
10 cm above pre-lake levels, and the water levels now have much less seasonal
variability.

    The changes in ground-water patterns have been confined to the area
between the station's cooling lake and ashpit and the Wisconsin River.  To the
east of the cooling lake, a drainage ditch limits the effects of the lake on
the ground-water system to a narrow band east of the lake, except in the
northeast corner of the lake where there is no ditch.  In this area the water
table has risen several feet, but flow in this area is small.

    Thermal patterns in the marsh have been significantly altered by the
cooling lake.  Temperatures of the ground water that discharges in the marsh
now average several degrees above prior average ground-water temperatures, and
temperatures in the marsh are out of phase with seasonal air-temperature
patterns.  Much of the sedge meadow no longer freezes over in winter, and
temperatures Just below the surface are as high as 25° C prolong the growing
season.  Winter temperatures as high as 21° C and summer temperatures as low
as 8° C were observed.  In some areas the peak ground-water temperature
occurred during the winter months.  Changes in vegetation within the thermally
altered zone were documented by Willard et al. (1976) and are thought to occur
in response to changes in ground-water temperatures.

    As with changes in ground-water flows, the thermal alteration of ground
water was confined to the area west of the cooling lake.  Moreover,
significant changes in temperatures extended only to about 100 m west from the
dike.  The seasonal maximum ground-water temperature at any depth is a
function of distance from the cooling lake and the distribution of subsurface
lithologies.

    The movement of heat in the subsurface was simulated by using the finite
element method to approximate the differential equations for water and heat
flow.  The simulated temperature patterns agreed well with field data, but
were very sensitive to the distribution of subsurface lithologies.  Detailed
stratigraphic information was necessary to obtain reliable results.  The
problem of simulating heat flow in the subsurface is further complicated by a

                                     -4-

-------
poor understanding of how to estimate dispersivity.  Results presented in this
study show that the amplitude of the seasonal temperature wave is best
simulated when small dispersivities are used.

    Long-term simulations of thermal alterations in the ground-water system of
the marsh indicate that the cooling lake will significantly alter substrate
temperatures but only within 350 m of the cooling lake dike.  By 1987 peak
temperatures near the dike were elevated 10-15° C above normal levels and
lagged behind seasonal temperature changes by 1-2 months.  At 150 m from the
dike, peak temperatures at a depth of 0.6 m were only elevated 2,3° C above
normal levels in the summer, but winter temperatures were elevated 6-10° C
above normal levels.  Temperatures within the wetland varied spatially,  but
trends were expected to be similar to those simulated in the two cross
sections modeled.  Seepage rates will increase 20$ as the result of the
increase in temperature of seepage waters.

    The simulations indicate that winter temperatures will be warm enough to
prevent formation of an ice cover in almost the entire marsh between the
cooling lake and the Wisconsin River.  Near the dikes a relatively warm
microclimate will exist during most of the winter, since the temperature of
the discharging water will be above 20° C during the early winter.

    The potential for significant thermal alteration of surface-water bodies
located in ground-water discharge zones is great.  The temperature of the
ground water may rise considerably near a cooling lake, even though the total
heat flux is small.  At the site of the Columbia Generating Station H0% of the
water pumped into the cooling lake seeps into the ground-water system, yet
less than 2% of the waste heat load is discharged to the ground-water
reservoir.  The discharge rate of water from the generating station to the
cooling lake is 1 x 10^ m3 per day, and the temperature of the surface water
is increased 10-15° C.  However, because the seepage rate is only 2 x 101* m3
Per day, much of the heat load is dissipated through evaporation.  The
percentage of the heat load that leaves a cooling lake via seepage to the
ground-water system is unlikely to be greater at other power-plant sites.

    Seepage from the ashpit averages 0.3-0.6 m3/s.  Surface water in the
wetland within 50 m of the ashpit has been contaminated by seepage through the
ashpit dikes, and total dissolved solids in the surface waters have increased
over fivefold above background levels in this area.  However, significant
chemical degradation of ground water, judged by the concentration of the
common cations and anions, has not occurred in the vicinity of the ashpit.

    A plume of contaminated ground water is slowly moving eastward from the
coal pile.  Three years after coal was piled in the area, a plume of water
with a sulfate concentration greater than 1,000 mg/liter extended to a depth
°/ greater than 30 m and outward from the limits of the coal pile more than
                                     -5-

-------
RECOMMENDATIONS FOR SITING FUTURE POWER PLANTS

    All new power plants should be required to gauge accurately (within 150
all water flows into and out of the plant and the plant site.  In addition,
precipitation into and evapotranspiration from all bodies of open water, such
as ashpits and and cooling lakes, should be monitored.  Seepage from ashpits
and cooling lakes cannot be determined accurately by direct methods; instead,
seepage should be calculated indirectly by a mass balance approach.  Data on
water flows, precipitation, and evapotranspiration are required for
calculating a mass balance.  Control rooms of power plants are now designed to
measure accurately mass flows within the generating station, and therefore
water flows between the environment and the plant can also be carefully
monitored.

    Coal-storage areas should be designed so that water infiltrating the coal
pile cannot reach the water table.  Since crushed coal is highly permeable and
leachable, an open coal pile allows most of the annual precipitation to pass
through the pile, which creates water similiar to acid mine drainage.  Unless
infiltration is controlled or the infiltrating water is captured and treated,
a volume of contaminated water equal to the product of the annual
precipitation and the area of the coal pile will flow into the subsurface.

    Estimations of impacts of generating stations should recognize that
significant thermal alterations of nearby surface waters can occur even when a
closed-cycle cooling lake is used.

    Dikes of all ashpits and cooling lakes should be lined with several inches
of a relatively impermeable clay material.  Seepage can be controlled, and
sloppy designs should not be tolerated.  More seepage can usually be allowed
from a cooling lake than from an ashpit, since cooling-lake water is usually
less threatening to the environment.

    Simulation techniques should be used to evaluate potential seepage from
ashpits, cooling lakes, and coal-storage areas in the design stage of these
structures.  Designs should minimize ground water seepage and contamination.
Competent geologic consulting firms should be retained to conduct these
studies.

    Although seepage from a cooling lake may contribute to alterations in an
adjacent wetland ecosystem, the high leakage rates are not undesirable from a
lake-management standpoint.  Seepage to the ground-water system prevents an
increase in total dissolved solids in the lake.  Such an increase would
necessitate periodic flushing of the lake and the consequent release of saline
water.

RECOMMENDATIONS FOR FUTURE RESEARCH

    More field data are needed on the distribution of temperature or
conservative chemical characteristics, or both, in small ground-water systems
subjected to stresses that are altering these characteristics from background
levels.  The flow of mass and energy in ground-water systems is much more
difficult to quantify than the flow of water, mainly because the former is

                                     -6-

-------
sensitive to the hydraulic conductivity distribution.  The development of
quantitative techniques to date has been hampered by the lack of field data to
verify the models.  As demonstrated in this study, it may be more fruitful to
monitor temperature changes than chemical changes in a ground-water system if
the goal is to determine the system's capability to transport mass and energy.
Temperature can be inexpensively and rapidly monitored at a large number of
points, which is not the case for chemical characteristics.

    Existing ground-water models will be adequate in most cases for
quantifying impacts of power plants on hydrogeologic systems.  The best
existing models for these purposes are those devised by the U.S. Geological
Survey.  Future research should not be directed towards developing new models
for these purposes, unless the investigator can show clearly that existing
models are deficient.

    More study is needed of the factors controlling dispersivity, and the
appropriateness of the concept should be evaluated.

    An evaluation is needed of the prediction ability of models used when
historical data are not available;  The magnitude of error associated with
predictions should be quantified.  It may be that transport models should not
be used for predictions if historical data are not available for calibration.

    A theoretical evaluation of the validity of using either chemical or
energy distribution in the subsurface to determine hydraulic conductivity
distributions is needed.  The question to be asked is whether or not the
inverse approach can produce useful approximations in a problem that is more
complex than those investigated by Bredehoeft and Papadopulos (1965).
                                     -7-

-------
                                  SECTION 3

                 THE SITE OF THE COLUMBIA GENERATING STATION
     The Columbia Generating Station is a 1,000-MW coal-fired electric-power-
generating complex located on the flood plain of the Wisconsin River 5 miles
south of Portage, Wis.  Construction began in 1971, and the two 500-MW units
began operation in May 1975 and May 1978.  The 1,620-ha site on which the
facility was constructed consisted of an extensive marsh, mainly sedge
meadows, and wetland forests with sandy upland knolls (Figure 1).
Construction activities have dramatically altered the site.  A 200-ha cooling
lake, a 28-ha ashpit, a 16-ha coal pile, and the generating station itself are
now the dominate features of the site (Figure 2, Figure 3).

     The power plant dynamically interacts with the environment.  On an
average day the first unit alone burns 2,000 tons of coal, producing 385 MW of
electricity, 1,000 MW of waste heat, and 552 tons of coal residue.  This rate
of coal burning is approximately 38 kg/s, which represents over 1,000 MW of
power of which only about 360 MW are converted to electricity.  The cooling
lake and the ashpit are the direct recipients of most of the byproducts.  The
thermal and chemical characteristics of the byproducts differ markedly from
the thermal and chemical characteristics of water in the adjacent wetland
environments (Figure 4).

     The site of the Columbia Generating Station is located in a regional
discharge area (Figure 5).  The discharge area includes the wetlands on the
site of the power plant, the wetlands to the east of the site, which are used
for mint farming, and extensive wetlands along the lower reaches of Rocky Run
Creek.  The marsh on which the site is located occupies a former river
channel.

     The geology beneath the site of the Columbia Generating Station consists
of Irregularly eroded Upper Cambrian quartz sandstone overlain by sediments
from glacial lake bottoms and drift deposits of the Green Bay lobe of the
Wisconsin ice sheet.  The bedrock is composed of Upper Cambrian sandstones and
Precambrian granites which occur 125 m below the surface.  The surface is
covered by a layer of peat 1.5 m thick, which overlies a thin layer of organic
clay and silt.  The peat and clay are in turn underlain by alluvial sands with
clay lenses.  Representative stratigraphic cross sections are shown in Figure
6.  The hydraulic conductivity of the clay is an order of magnitude less than
the peat and thus controls the rate of ground-water discharge (Table 1).
Vertical hydraulic conductivities were estimated to be 5 to 20 times less than
horizontal hydraulic conductivities.
                                     -8-

-------
I
\o
I
        Figure 1.   The site of the Columbia Generating Station during construction of the first unit

                   (24 May 1971).

-------

i
(—•

r
         Figure 2.  The site of the Columbia Generating STation after construction of both units


                    (12 August 1977).

-------
Figure 3.  Main features of the site of the Columbia Generating
           Station.
                               -11-

-------
Figure 4.  Water flows - annual averages 1CP m3 day lt and energy flows - approximate values 1MW =
           239 kcal/sec, at the site of the Columbia Generating Station.

-------
Figure 5.  Potentiometric surface contours in the Cambrian sandstone
           (meters above mean sea level) in the area of the Columbia
           Generating Station.  Location is shown by diamond on insert
           map of the State of Wisconsin.

                                   -13-

-------
                                        COOLING  LAKE
               UJ
                UJ
                                 A'
                                                           LEGEND
                                                      Med. -coarse sand with grovel
                                                      Fine to very fine sand
                                                      Fine sand with 10% silt
                                                      Weathered sandstone
                                                      Sandstone
                                                       Sandy silt to sandy clayey silt
                                                       Grey sandy silt with organic matter
                                                       Varved clay with sand seams
Figure 6.   Representative  stratigraphic  cross  sections  of the subsurface at
             the site  of the Columbia Generating Station.   Insert shows
             locations of  sections  at the  site.
                                          -14-

-------
      TABLE 1.  HORIZONTAL HYDRAULIC CONDUCTIVITIES FOR THE LITHOLOGIES
             IN THE SUBSURFACE OF THE COLUMBIA GENERATING STATION
                                                  Horizontal hydraulic
                    Lithology                    conductivity (m/day)a


         Medium-coarse sand with gravel                  30

         Fine to very fine sand                          10

         Fine sand with 10$ silt                          6

         Weathered sandstone                              8

         Sandstone                                       3—4

         Peat                                          0.4—3

         Sandy silt to sandy clayey silt                  0.4

         Gray sandy silt with organic matter              0.04

         Varved clay with sand seams                  0.02—0.04


Conductivities were determined from a pump test, slug tests, laboratory
 permeameter tests, and estimates based on lithology and grain-size analyses.


     The 1,900-ha site ranges in elevation from 237 to 248 m above mean sea
level.  The higher areas are well drained since the underlying soil is fine to
medium sand.  The elevation of most of the site, however, is less than 239 m,
and these low areas are usually wet throughout the year because of
ground-water discharge (Figure 7).

     A well-developed ground-water flow system exists in the Cambrian
sandstone.  The discharge areas of the aquifer include the site, the wetlands
west of the cooling lake, the wetlands east of the lake, and extensive
wetlands along the lower reaches of Rocky Run Creek (Figure 3).  The sandstone
aquifer has an areal extent of approximately 200 km^ extending from the site
to the southeast (Figure 5).  Discharge from this aquifer is nearly constant
all year and averages about 0.013 m^/s per kilometer perpendicular to the
direction of flow.

     The low areas also receive ground-water recharge from the upland areas
located on the site and adjacent to the site.  These upland areas act as
recharge areas during precipitation events, when the infiltrating water flows
toward the low areas.  Travel time for water from the upland areas to the low
                                     -15-

-------
Figure 7.  Topographic map of the site of the Columbia Generating Station.
                                  -16-

-------
areas is usually less than 3 months,  but flow rates are seasonally variable.
Before construction of the cooling lake, approximately 505t of the ground water
discharging in the low areas in the spring originated from the upland areas on
and near the site; during the summer the percentage dropped to less than 5%.

     The ground-water inflow rates to the low areas are greatest where the
peat deposits are thinnest (Figure 8), namely in areas recently reworked by
fluvial processes.  The peat deposits are generally less than 2 m thick west
of the railroad track; east of the railroad track, kettles in the outwash and
ice-contact drift have been filled in by organic deposits as thick as 10 m.
North of the ashpit alluvial deposits are more complex than in other parts of
the site; sediments range from coarse to very fine, indicating that
depositional processes characteristic of both the Wisconsin River and Duck
Creek have formed these deposits.

     The cooling lake was designed to dissipate the waste heat, which is
discharged from the plant after steam has been used to generate the
electricity.  The 200-ha lake has an effective length of 5,500 m.  The lake
and the ashpit were formed by dikes 5 m high constructed entirely of local
silty sand.  The western dikes around the cooling lake were lined with
bentonite.  Water is pumped into the cooling lake from the Wisconsin River at
an average rate of 50,000 m3/day.  Of this water 2Q% discharges into the
ashpit, 40$ is lost by ground-water seepage, and 40 J is lost by evaporation.

     Hot water is discharged at a rate of 1 x TO** n>3 into the north end of the
lake east of the central divide (Figure 3).  Water circulates in a clockwise
direction and is withdrawn from the north end of the lake west of the central
divide.  Average residence time of a water particle in the lake is 5 days.
Water temperaturess at the discharge point average 10-15° C higher than
temperatures at the intake and decrease exponentially with distance from the
discharge point.  The average annual range of lake temperature is 0-45° C.
Temperatures in the lake vary only slightly with depth.

     Cooling towers have been built to dissipate some of the additional waste
heat that is being generated by the second unit.  However, the cooling lake is
projected to receive 80% of the annual heat load;  The discharge of the
additional waste heat into the cooling lake is estimated to increase lake
temperature by approximately 8° C when the cooling towers are not operating
(Wisconsin Public Service Commission  1974).
                                     -17-

-------
                    INTERVALS
            Less  than  2.'.	J  I
            Between    2' and 5'.	FTT3
            Between    5' and 10'.	E2D
            Between   10' and 20:
            Over     20*....          ....ESi
Figure 8.  Peat thicknesses  in the  subsurface at  the site of  the Columbia
            Generating Station.
                                         -18-

-------
                                  SECTION  4

                    THE IMPACT OF-A POWER PLANT ON THE
                     GROUND-WATER SYSTEM OF A WETLAND
CONFIGURATION OF THE GROUND-WATfifl FLOW SYSTEM

    The first goal of the hydrogeologic investigation at  the  site  of the
Columbia Generating Station was to determine the impact of the  station on
ground-water flows, temperature, and chemistry in the adjacent  wetland
(Anderson and Andrews 1977, Andrews and Anderson 1978).  To determine  the
physical setting in the ground-water system of the wetland before  operation of
the plant, data were collected beginning in the summer of 1971.  The position
of the water table was monitored in 80 small-diameter observation  wells, and
vertical head gradients were measured in 19 nested observation  wells.

    Two-dimensional steady-state models of several representative  vertical
cross sections oriented perpendicular to the western dike of  the cooling lake
were generated.  (Figure 6 illustrates two of these cross sections.) These
models were then combined to construct a quasi-three-dimensional model of the
flow system.  This model was used to simulate the head distribution before  and
after filling of the lake and to compute the total ground-water discharge to
the wetland in the area affected by the lake.  The vertically oriented cross
section models were patterned after the regional ground-water model of Freeze
and Witherspoon (1966).  The finite difference equations  were solved by using
the method of successive over-relaxation.  A 25-by-50 node grid was used for
all simulations.  Horizontal nodal spacing ranged from 15m near the dike to
120 m farther from the dike, and vertical spacing ranged  from 0.6  m near the
surface to 30 m at depth.  Details of the modeling procedure  can be found  in
Andrews (1976).

    Field data were used to check the validity of the model,  and the model
then computed ground-water discharge to the wetland.  The simulated head
distribution in cross-section A-A' of Figure 6 before filling of the cooling
lake is shown in Figure 9.  The ground-water flow system  was  modeled to a
depth of 150 m, but for clarity only the upper 30 m are shown.   Horizontal
flows ranged from 0 to 4.1 cm/day.  Vertical flows were mostly  toward  the
surface at rates of 0-0.16 cm/day.  Approxmiately 300 m west  of what is now
the western dike, vertical flows were as high as 0.53 cm/day.  Total discharge
to the wetland in the cross section shown in Figure 9 was 1.00  m3/day  per
meter dike length, of which 0.43 m3/day per meter discharged  in the area of
high vertical flows shown in the figure.

    A steady-state model was used to simulate pre-lake conditions, although
the water table fluctuated about 0.5 m/yr in the lowlands and as much  as

                                     -19-

-------
                             ELEVATION (METERS)
                                                              .30-.76m/yr


                                                              .!5-.30m/yr


                                                              .76-l.52m/yr

                                                              .30-.76m/yr
                                                              0.0-.30 m/yr
                                                              .30-76 m/yr
                                                              0.0-.30m/yr
Figure 9.  Simulated head  distribution in cross-section A-A * of Figure  6
           before filling  of  the Columbia cooling lake.  Equipotential
           lines are labeled  in meters.  Vertical discharge rates are
           given across  the top of the figure.
                                   -20-

-------
2 m/yr in the uplands.   However,  the water-table gradient remained nearly
constant throughout the year,  except in the upland areas where the gradient
increased markedly during periods of recharge.  The head distribution
presented in Figure 9 approximates the average annual head distribution.

    The initial filling of the cooling lake began 13 June 1974 with water
pumped from the Wisconsin River.   Pumping stopped on 17 July, when the level
was 0.5 m below the fill line.  By 1 September the lake was almost empty, and
on 4 November 1974 pumping resumed.  The water level reached the fill line on
2 January 1975.  Water levels in observation wells monitored during the
filling of the lake showed increases in vertical gradients in the area west of
the cooling lake (Andrews 1976).

    The filling of the lake greatly altered the head distribution in the
ground-water system (Figure 10).   What had been one ground-water system became
three systems.  Ground water that formerly discharged to the wetland began to
discharge near the drainage ditch east of the cooling lake.  The lake became
the only source of water for the discharge area west of the lake.  For the
cross section in Figure 10, the simulation predicted that the ground-water
discharge to the wetland west of the dike was 3.5 m^/day per meter and that
the flow into the drainage ditch was 3.7 m^/day per meter.  As in the pre-lake
simulation, vertical discharge rates were highest approximately 240-300 m west
of the dike.  In this main discharge area vertical flow rates varied from 1.02
to 1.34 cm/day.  Elsewhere, vertical flows ranged from 0.24 to 0.94 cm/day.
The position of the main discharge area varied with the cross section modeled.
A comparison of total discharge rates to the wetland for Figure 9 and Figure
10 shows that discharge to the wetland in that cross section after filling of
the lake was 3 1/2 times greater than discharge before filling.

    The effect of the cooling lake on the head distribution in several other
cross sections was also simulated.  In all simulations the horizontal
component of flow was assumed to be perpendicular to the dike.  Changes  in
configuration of the flow system in other cross sections modeled were similar
to the changes shown in Figure 9 and Figure 10.  In some cross sections,
however, the discharge to the wetland was  15 times greater after the filling
of the cooling lake.  Total discharge to the area of the wetland affected by
the lake was estimated to be six times greater after filling.

    The assumption of steady-state conditions after filling of the cooling
lake is acceptable in view of the stabilization of ground-water levels within
a month of the filling of the lake.  However, since the water in the cooling
lake is warmer than the ground water in the system before  the lake was filled,
the ground-water system is not yet in equilibrium with the altered thermal
regime.  The model allows for variation in hydraulic conductivity as a result
of differences in ground-water temperatures.  For the simulation presented in
Figure 10, the temperatures in the ground-water system were asssumed to  vary
from 26° C directly beneath the cooling lake  to 10° C at a depth of 20 m below
the lake.  The relationship between  ground-water temperature  and ground-water
flows is discussed in section 5.

    From the simulations and  the  supporting field evidence, the following
observations can be made.  On the  east and south sides of  the cooling lake,

                                     -21-

-------
                         ELEVATION  (METERS)
                                                            .30-.76m/yr


                                                            .l5-.30m/yr

                                                            .30-.76m/yr

                                                            .76-l.52m/yr
                                                            l.52-3.04m/yr

                                                            .76-l.52m/yr

                                                            .30-l.52m/yr
                                                            INFLOW
                                                            7.6-25.4m/yr
Figure 10.  Simulated head distribution in cross  section A-A ' of Figure  6
            after  filling of the Columbia cooling lake.   Equipotential
            lines  are labeled in meters.  Vertical discharge rates are
            given  across the top of the figure.
                                     -22-

-------
the drainage ditch effectively limited the impact of the lake on the
ground-water system to a relatively narrow zone.   On the north and northeast
sides of the lake where there is no drainage ditch,  the filling of the cooling
lake resulted in a rise of ground-water levels of almost 1 m.  This rise
reversed flow in this area such that water now flows from the lake eastward
beneath the coal pile and discharges in the wetland area east of the site.
The average annual seepage from the cooling lake  is estimated to be 0.18 to
0.27 m^/s.  Flow rates fluctuate about 20$ during the year as a result of
changing ground-water temperatures.

    The average annual seepage from the cooling lake will increase by
approximately 20% with both generating units in operation.  The discharge
rates will also increase because of the formation of springs in the wetland
west of the cooling lake.  Springs have formed since the filling of the
cooling lake, because vertical hydraulic gradients in many areas near the dike
exceeded 1 m/m, which approximates the critical gradient for the onset of
heaving in unconsolidated materials.  In addition, increased temperatures in
the marsh have speeded up the process of peat decomposition, and deeper water
levels in the marsh have resulted in the floating of parts of the peat mat.
The result of both these processes is a decrease in the total load on the
confining silt-clay layer, a decrease that enhances the onset of heaving.
Heaving is not predictable, and therefore accurate estimates of its rapidity
and of its impact on flow rates are impossible.  Potentiometric data from
observation wells in the marsh indicate that the springs that have formed
during the first 2 yr since the cooling lake was filled have not significantly
altered flow rates.  Increases in flow rates caused by these processes will
probably not exceed 8% during the next 10 yr.

WETLAND WATER LEVELS

    The wetland west of the cooling lake lies in two distinct drainage basins,
each of which is connected by distinct channels to tributaries of the
Wisconsin River.  The wetland area adjacent to the cooling lake and north of a
line 350 m south of the intake channel drains to Duck Creek, and the area
south of this line drains to Rocky Run Creek via a small channel (Figure 3).
Ground-water inflow to the northern basin averages approximately 0.09+0.03
mVs, and inflow to the southern basin averages approximately 0.10+0.03 m3/s.

    The water level in the wetland is a function of the ground-water inflow
rate, the rate of water leaving the wetland via surface outflow, the rate of
precipitation and evapotranspiration, and the wetland basin shape.  The rate
at which water leaves the wetland is a function of the water level and the
shape of the channels draining the wetlands.  On a daily basis
evapotranspiration from each wetland basin seldom exceeds 0.04 m3/s.  Because
of the shape of the channels that drain the basins, a change in outflow from
0.04 to 0.20 m3/s raises water levels less than 5 cm.

    Water levels in the wetland have been almost constant since the cooling
lake was filled.  The level averages about 10 cm higher than previous levels
(Figure 11).  Wetland water levels do fluctuate during the winter when ice
clogs the drainage channels and during floods on the Wisconsin River.
                                     -23-

-------
          0)
          1C
          IU

          ui 238.05
UI

UI
_l

E
UI
            237.74
            237.44
                                 cooling lake filled
               JAN
               1972
              JAN
              1973
JAN
1974
JAN
1975
JAN
1976
JAN
1977
Figure 11.   Water levels in the wetland west of the Columbia cooling lake
            before and after filling of the lake.
    Before the cooling lake was filled, ground-water inflow rates  in the
summer to each of the wetland basins were only  0.2 to 0.3  m3/s.
Evapotranspiration often exceeded  inflow, and there was no outflow from the
wetlands via the channels.  During the late  summer water levels  were lowered
as much as 0.6 m.

    The peat substrate in  the wetland has been  altered  by  the  increased flow
rates within the wetland,  the increased water temperatures, and  the increased
water levels.  Some peat has been  eroded, and new channels have  been cut in
the wetland as the internal drainage system  adjusts to  the increased flow.
Other parts of the peat mat have floated  to  the surface and been broken down
by various physical and biological processes.   Peat decomposition  rates have
Increased because of the increased water  temperatures and  increased oxygen
content of discharging ground water.  The plant community  is rapidly
responding to the new environment  as evidenced  by the replacement  of formerly
dominant sedges, Carex laoustris and Carex stricta, by  plants more tolerant of
deeper water, primarily Sagittaria latifolia and flumex  orbiculata  (Willard et
al. 1976).

GROUND-WATER TEMPERATURE

    Before the cooling lake was filled, the  temperature of ground  water
discharging to the wetland was approxmiately equal to the  average  annual  air
temperature (10° C).  The  cooling  lake, now  the source  of  water  discharging  to
the wetland, has an average annual temperature  between  10° C and 17° C,
depending on location in the lake.

    In areas with a layer  of relatively high permeability  near the dike,
discharge rates are high and seepage from the lake has  a travel  time of less
than 1 yr before discharging to the wetland. Consequently, the  range of
ground-water temperatures  is similar to that of the lake water,  although the
extremes are somewhat attenuated.   The temperature of water in the western

-------
     25 r
     20
     15
   u

   OT
   111
   Ul
     10
         •I  I   I  I  I  I   I  I  I  I  I  I  I
                                         I  I	I  I  I   I  I  I   I
                                                               I  I   I  I  I
      JAN MAR  MAY  JULY  SEP  MOV
                   1975
JAN  MAR  MAY  JULY  SEP
             1976
NOV  JAN  MAR  MAY
             1977
Figure 12.   Ground-water temperature variations in wells west of the Columbia
            cooling  lake.   Curve A:   data from a well cased to 4 m below the
            surface  and located 3 m west of the dike.  Curve B:  data from a
            well cased  to 2.5 m below the surface and located 60 m west of
            the- dike.
portion of tne lake from May 1975 to May 1976 ranged from 1° C to over 30° C.
Ground-water temperature in a well 3 m west of the dike varied from 2° C to
25° C (Figure 12, curve A), and in areas west of tnis well, attenuation of
extremes was even greater (Figure 12, curve B).

    To measure the heat discharge into the ground-water system, 47 thermistors
were installed in the summer of 1976 to depths of 10 m below the surface along
cross-section A-A' of Figure 6.  Temperature data were collected twice a week
to document seasonal trends in the variation of ground-water temperature.  The
temperature changes observed in the ground-water system are described in
detail in section 5.  Preliminary calculations based on these thermistors
suggest that approximately 16 MM, less than 5% of the heat released to the
cooling lake, is discharged to the ground-water system.  Ground-water
temperatures increased as much as 20° C, however, in some areas west of the
cooling lake.

    Normally, the species of sedges present in the marsh emerge early and die
back late in the growing season.  During the winter of 1975, however, sedges
in some areas of the marsh turned green in December because of the warm water.
In the spring these sedges were dead, possibly because they had used up their
stored food reserves.  In  1976 these vegetation changes occurred over a
considerably larger area (Willard et al. 1977).
                                     -25-

-------
RELATIONSHIP BETWEEN GROUND-WATER TEMPERATURES AND FLOW RATES

    The groaid-water system was not yet in equilibrium with the  altered
thermal regime.  Because ground-water flow rates depend on temperature of the
ground water, leakage is greatest in late summer when the  water  is the
warmest.  Furthermore, as ground-water temperatures rise in response  to heat
input from the cooling lake, the flow rates will increase.  The  ground-water
flow model predicted that flows in the cross section shown in Figure  10 will
increase by 24J when the system reaches equilibrium.  A model which couples
ground-water flow with heat flow is discussed in section 5.

CHANGES IN WATER CHEMISTRY

    During the preoperational period (August 1972 to September 1974)  144 water
samples were taken from observation wells on the site.  During the first 7
months of operation 134 samples were collected from 57 wells.  Some of the
types of water on the site are illustrated by Stiff diagrams in  Figure 13 and
Figure 14.       '

    Significant changes in the ionic composition of ground water and  surface
waters in the wetland since the generating station began operation were
observed only near the ashpit.  Because of the similarity  in the chemical
composition of cooling-lake water and ground water, no significant changes are
likely to be observed in ground-water quality near the cooling lake.   If
changes do occur, they are likely to be in sodium concentrations caused by
changes in ion-exchange processes associated with higher flow rates.

    Major changes are likely to occur in ground-water quality near the ashpit
because ground water and water in the ashpit differ in quality.   Prediction of
these changes is not yet possible because water levels in  the ashpit  have
fluctuated widely during the first 2 yr of operation and because the  chemical
composition of ashpit waters has varied widely.  Ground-water flow rates from
the ashpit have consistently ranged from 0.3 to 0.6 rn^/s.   The filling of the
pit with ash has reduced seepage through the bottom, but this reduction has
been offset by increased flow through the dikes caused by  elevated water
levels.  Most ground-water flow from the ashpit now occurs through the dikes,
and most of the seepage water discharges near the dikes.

    Some change has been observed in ground-water quality near the ashpit.
Calcium and sulfate concentrations have increased significantly in two wells
close to the ashpit dike (Table 2).  Changes in water quality have been noted
in other wells farther from the dike, but these changes have not been as great
(Table 2).  The plume of contaminated ground water appears to be confined to a
relatively small area near the dikes.  However, pronounced increases  in
specific conductivity in surface waters in the marsh have been measured up to
50 m from the dike, which suggests that water is leaking from the ashpit
through the ground-water system and is discharging to the  wetland where the
water then spreads outward.  Head gradients in observation wells support this
hypothesis.

    The most significant degradation of ground-water quality has occurred in
the vicinity of the 17-ha coal pile.  Although background   sulfate

                                     -26-

-------
                                                      - cr
                                                         S0=
              (a)
               COOLING  LAKE
                                                      — cr
                                                         HCO-
Mg    —
                                                      - so;
              (b) SHALLOW  GROUNDWATER
       _ -n-
       Co    —
       »
       Mg    -
                                                      — cr
                                               - so;
              (c)    DEEP GROUNDWATER
                   1
                CATIONS
                               1         2
                                 ANIONS
                         EQUIVALENTS PER MILLION
Figure 13.  Stiff  diagrams of water in the cooling lake (a) and  ground water
           (b and c),  Data were  collected in fall 1975.   Ground water (b)  is
           from a well cased to 2 m below the surface and located 50 m west
           of the north end of the cooling lake; ground water (c) is from a
           well cased to 6 m below the surface and located 60 m west of the
           south  end of the cooling lake.
                                  -27-

-------
concentrations are less than 20 mg/liter, a plume of contaminated groind water
has been created with a sulfate concentration of greater than  1,000 mg/liter.
This concentration exists to a depth of more than 30 m near the coal pile and
extends more than 100 m to the east of the pile.  The plume is gradually
spreading east, toward the wetland discharge area (Andrews 1976).
      Ca
        **-
      Ca**-
      Mg"-
                                                                  — cr
                                                                   - so;
                                                        (FALL 1975)
                                                                  — cr
                                               — HCOj * COj
                                               -so;
                      (b)    GROUNDWATER WEST OF ASH PIT  (1972-1973)
      Mg" -
                                                                   - Cl"
                                               —  so
                       (C)   GROUNDWATER WEST OF ASH PIT  (FALL 1975)
             J_
              3
  J_
    J_
_L
  2
CATIONS
     1        2       3
       ANIONS
EQUIVALENTS PER MILLION
 Figure 14.  Stiff diagrams of water in the ashpit (a) and in a well cased to
             1.5 m below the surface and located 2 m west of the ashpit
             (b and c).  Comparison of (b) and (c) shows the change in water
             quality after the partial filling of the ashpit.
                                     -28-

-------
TABLE 2.  SELECTED CHEMICAL CONCENTRATIONS IN GROUND WATER NEAR
         THE ASHPIT OF THE COLUMBIA GENERATING STATION

(a/1)
K+
Na+
Ca++
Mg++
304
ci-
(m/1)
K+
Na+
Ca++
Mg++
son
ci-

Ashpit
water
2.3
10.2
77.8
10.6
44.6
10.6

Ashpit
water
2.4
10.3
54.7
4.3
36.0
10.5


1972-73
1.2
2.2
13.9
11.1
5.4
3.4


1972-73
1.7
4.3
18.8
17.3
7.7
2.8
A. West
1 m west
1975-76
9.2
8.5
16.9
9.8
18.5
4.5
B. North
1 m north
1 975-76
0.75
2.6
23.3
22.5
9.6
4.2
of the

1977
6.9
13-2
21.6
14.7
10.5
6.5
of the

1977
1.5
5.1
37.0
33.8
22.0
8.5
dike

1972-73
1.2
3.6
31.6
20.3
0.8
4.49
dike

1972-73
2.0
5.2
22.0
18.0
9.7
7.2

75 m west
1975-76
0.75
3.5
28.8
22.0
0.2
2.64

25 m north
1975-76
0.75
3.1
22.6
18.8
9.3
3.2


1977
0.5
4.0
24.3
16.9
0.1
5.0


1977
0.5
3.7
21.5
25.8
14.0
3.5
                               -29-

-------
                                  SECTION 5

                 THERMAL ALTERATION OF GROUND WATER  CAUSED  BY
                        SEEPAGE FROM THE COOLING LAKE
    In addition to monitoring and modeling groind-water flow rates near  the
Columbia Generating Station, we investigated the  effect of the  station's
cooling lake on groind-water temperatures (Andrews and  Anderson 1979).
Ground-water temperatures in the vicinity of the  cooling lake were monitored
in detail in the field for 1.5 yr.  The presence  of the cooling lake, which
loses water to the ground-water system at a rate  of 2 x 10^  m3  per day,  has
created a zone of thermally altered gromd water, but the zone  is confined to
a relatively small area hydraulically downgradient from the  cooling  lake.

    A mathematical model was developed to simulate the  response of subsurface
temperatures to seasonal changes in lake and air  temperatures.   To create the
model, equations describing groind-water flow were coupled with equations for
heat flow in the subsurface.  An equation describing the rate of heat loss
from the marsh surface was used as one of the boundary conditions for the
heat-flow model.  The model was solved numerically by using  the finite element
technique.  The model assumed that the flow of heat and water from the cooling
lake occurred in planes normal to the plane of the dike.  The flux of water
and heat was modeled in seven of these planes (Figure 15), which represent
two-dimensional vertical cross sections of the ground-water  system.  Results
from each of the cross sections were combined to  obtain the  total flow of heat
and water outward from the west dike of the cooling lake.

    Simulated temperature patterns agreed well with field data, but  were very
sensitive to the distribution of subsurface lithologies.  Results from a
predictive simulation suggest that operation of the second 500-MW unit will
increase ground-water temperatures less than 5° C at distances  greater than
15 m from the cooling lake.  The results of this  study suggest  that  the
potential for significant thermal alteration of surface water bodies located
in ground-water discharge areas is slight.

MATHEMATICAL MODEL

Governing Equations

    The equation describing the two-dimensional flow of water through a
nonhomogeneous aquifer may be witten
                          -<>•'• J'1-2  •                    '"
                                     -30-

-------
                  CD
                                                  GENERATING
                                                    STATION
                                                                     MINT
                                                                     FARM
                                                                     DRAIN
                                                                    D1
Figure 15.   Location of  seven cross sections of the Columbia site for which
            ground-water temperature distributions were simulated.
                                   -31-

-------
wnere KJJ = hydraulic conductivity, L/t;  = head, L; and x-|,X2 = cartesian
coordinates, L.  Tne ground-water velocity or specific discharge can then be
determined from
                            qi * ~Kij Ih    '                               (2)
    The movement of heat in a ground-water system can be described
mathematically assuming that:  (1) Thermal equilibrium between the liquid and
the soil particles is achieved instantaneously, (2) the densitiy of the soil
particles is constant, (3) the heat capacity is constant, and (4) the chemical
system is inert.  Under these assumptions the heat transport equation is


                                      -P-°.l.'-1.*  '
where  T = temperature, T; Djj = coefficient of dispersion, H/TtL;  pCw = heat
capacity of the saturated media, H/L^T; pCw = heat capacity of water, h/L^T;
and Qi s specific discharge in direction x^, L/t.

    If the medium is isotropic, the coefficient of dispersion is a second-rank
tensor (Scheidegger 1961) composed of two parts:  (1) The thermal conductivity
of the medium and (2) the coefficient of mechanical dispersion, which
represents the mixing caused by the heterogeneity of the velocity field.  In
this analysis the coefficient of dispersion was assumed to have the following
form (Heddel and Sunada  1970):
                    D
                                   q2q2
'22 " *22 + "L ~~q~ + *r
                     D21

            2    2 1/2
where <1 " (ij + ^    » K11 = K22 = thermal conductivity, H/TtL (thermal
conductivity is assumed to be independent of direction); QL = longitudinal
(horizontal) dispersivity, L; and <*T = transverse (vertical) dispersivity, L.

Boundary Conditions

    for the ground-water flow model, heads were specified at the water table
and along the vertical boundary east of the cooling lake.  Tne model can be
modified to consider the time-dependent case of a moving water table, but for
the problem considered here fluctuations of the water table were negligible.
Mo-flow boundaries were specified along the lower boundary and at the
Wisconsin River.

    for the heat-flow model, a heat flux was specified along the upper
boundary and along the vertical boundary east of the cooling lake.  The fluxes
of heat across these boundaries were specified to be proportional to the water

                                     -32-

-------
flow across the boundaries.   Along that  part  of the  upper boundary
representing the marsh,  a heat flux resulting from atmospheric  exchanges was
specified.   The other two boundaries were specified as no-flow  boundaries.

    The process of heat  exchange between the  marsh surface and  the atmosphere
is complex, and mathematical  representations  are necessarily rough
approximations.  For this study the boundary  between the  marsh  and the
atmosphere  was assumed to occur in the vegetation at the  height viiere transfer
of heat and water vapor  by turbulent exchange becomes effective.   This level,
known as the zero-displacement plane, was assumed equal to 0.63 of the
vegetation height.  Heat flow from the ground surface to  the zero-displacement
plane was assumed to occur by conduction and  free convection.  The heat flow
at this boundary can be  expressed as


    IB = \ (Tm *  V + hm (em ~  e

where Tg^y = effective sky temperature,  T; Tm = temperature at  the marsh
surface, T; Ta = air temperature above the plane of zero displacement,  T;  qs =
heat flux from the marsh boundary, H/tL2; h™  = mass transfer coefficient,
H/L2Tt; hy  = heat transfer coefficient,  H/L^Tt;  Y = psychrometer constant,
assumed equal to 0.66 mbar/T, M/t2LT; em = saturation vapor pressure at the
marsh boundary temperature, M/t2L; ea =  partial vapor pressure  of water in
air, M/t2L;  ? = marsh boundary long-wave emissivity, dimension! ess;  a =
Stefan-Blotzmann constant, H/T^tL2; otg = marsh boundary solar absorptivity
dimensionless; qsun = solar flux incident on the marsh boundary, H/tL2; and a
= percentage of open water in the marsh, dimensionless.

    In addition to areas of dense vegetation, the marsh also contains some
areas of open water.  For the latter it was assumed that radiative exchange
and solar absorption occur at the surface.
    The heat and mass transfer coefficients (hy and h^) are equal if the
roughness height is the same for both.  According to Mitchell et al. (1975)
these coefficients are given by an equation of the form


                     hm =  hv = ^  PCpu/ln(Z


where k = Karman coefficient, dimensionless; PCp = heat capacity of air,
H/L^T; u = air velocity, L/t; Za = reference height, L; Zo = surface roughness
height, L; and d = height of the zero- displacement plane, L.  The surface
roughness height can be estimated from vegetation height by using the
relationship given by Tanner and Pelton (1960), log Zo - 0.997-log 0.833 ,
where ^ is the average height of the vegetation.

    Following the method of Jobson (1972), we used daily average values for
air temperature (Ta) and vapor pressure (ea) in Eq. (6).  Moreover, Eq. (6)
can be simplified when temperature data at the marsh surface are available,
and the terms involving higher order dependence on Tm are approximated by
linear relations.  Then, the heat transfer from the marsh surface is
proportional to the  temperature difference between the marsh surface and the

                                     -33-

-------
air temperature at 2 m above the surface.   The constants of proportionality
were determined for a cross section for which data  had  been collected, and
these values were used when simulating heat flow in the other cross sections.
The heat-flow rates calculated by this method agreed well with values
calculated by using Eq. (6).

Solution Procedure

    The finite element method (Zienkiewicz 1977,  Finder and Gray  1977) was
used to solve fiq. (1) and Eq. (3) subject  to the boundary conditions described
in the previous section.   Each cross section modeled was divided  into 100-150
quadrilateral elements.  The procedure used to link the equations was:   (a)
Eq. (1) was solved for head at each node and Eq. (2) was solved  for velocity;
(b) fiq. (3) was solved for temperature at  each node; (c) the solution to Eq.
(3) was stepped forward in time by the Crank -Nicolson approximation for  the
time derivative for a specified number of  time steps;  (d) the hydraulic
conductivity distribution, which is a function of temperature, was adjusted
for the new temperature distribution; and  (e) steps (a)-(d) were repeated.

Required
    The physical properties required as input parameters are  thermal
conductivity, specific heat capacity,  hydraulic conductivity,  and  porosity for
each material type in the subsurface and dispersivity.

    The values of the parameters for each of the subsurface materials used in
the simulations are listed in Table 3-  Thermal conductivities were determined
by the needle probe technique (Von Herzen and Maxwell 1959).   Heat capacities
were determined by standard additive techniques (Van Wijk  and deVries  1963).
Horizontal hydraulic conductivities were estimated from aquifer test data,
slug tests, and grain-size analyses.  Vertical hydraulic conductivity in  each
element vets assumed to be one-tenth of the horizontal hydraulic conductivity.
Porosities, which were needed to estimate pCs, were approximated from
grain-size analyses.

    Longitudinal dispersivity was assumed constant for all elements and was
estimated by a trial-and-error adjustment procedure.  Lateral dispersivity is
usually assumed to be one-quarter to one-tenth of longitudinal dispersivity
(Cherry et'al. 1975).  For the simulations reported here,  the ratio of
longitudinal to transverse dispersivity was assumed to be  4.   The field data
were best simulated by using a longitudinal dispersivity of 10 cm. When
higher values were used the amplitude of the annual temperature wave observed
in the field could not be reproduced.  In Figure 16 the observed temperatures
2 m west of the dike at a depth of 3 m are plotted along with the simulated
temperatures obtained by using various values for longitudinal dispersivity.

    Values for longitudinal dispersivity reported in the literature range from
0.1 cm for laboratory studies with homogeneous sands (Hoopes and Harleman
1967) to over 100 m for field studies (Bredehoeft et al. 1976).  The value of
dispersivity is known to be related to media inhomogeneities and the scale of
the problem, but the exact relationship is unknown.  For the problem described


                                     -34-

-------
          TABLE  3-  PARAMETERS  USED IN THE SIMULATIONS FOR  THE MODEL
                DESCRIBING  THERMAL ALTERATION OF GROUND-WATER
   ' Lithology
 Horizontal
 hydraulic          Thermal          Heat
conductivity     conductivity      capacity
  (m/day)     (1(H cal/m sec°C)  (cal/cm3<>c)  Porosity
Medium coarse sand       30
  with gravel

Fine to very fine         10
  sand

Weathered sandstone       8

Sandstone                 3.5

Peat                      3.0

Sandy silt to sandy       0.4
  clayey silt

Gray sandy silt with      0.04
  organic matter

Varbed clay with          0.03
  sand seams
                     0.45


                     0.51


                     0.51

                     0.64

                     0.12

                     0.42


                     0.42


                     0.40
0.64


0.67


0.67

0.60

0.91

0.72


0.72


0.74
0.33


0.39


0.39

0.26

0.75

0.48


0.48


0.52
RESULTS

Field Data

    Temperatires were monitored twice a week at 48 points in cross-section
A-A1 in Figure 15 for 13 months and at 18 points in cross-section B-B1 in
Figure 15 for 6 months.   Temperatures were recorded at the surface and at
depths to 10 m below the surface.  In situ thermistors, hardwired to a central
control box, were used to obtain ground-water temperatures in cross-section
A-A1.  Temperatures in cross-section B-B1 were measured by lowering a
thermistor probe into three 3.175-cm wells.  The thermistors were calibrated
to ±0.1° C.  A high-resolution digital ohmmeter was used to measure the
resistance of the thermistors.  Surface-water temperatures at two sites along
cross-section A-A' were monitored continuously with liquid expansion
thermographs.
                                     -35-

-------
                           3cm
                                                             — — o90 cm
                                                                  o30cm
                                                                  o|0cm
                                                     1978
Figure 16.   Observed temperatures  in the  subsurface west of  the Columbia
            cooling lake and temperatures simulated by  the mathematical
            model.   The ratio of longitudinal  dispersivity  (CXL) to
            transverse dispersivity (al)  is  set  at 4.   (	)  is observed
            temperatures with o£ = 3 cm.   (	) is  simulated temperatures
            with OL = 90 cm.  (o)  indicate maximum and  minimum simulated
            temperatures for cxL/ctT = 4 with  L  equal to  3, 10, 30, and  90  cm.
    Tfte  temperature distributions recorded in cross-sections A-A' and B-B| at
 3-mon?h  intervals are  shorn in Figure 17 and Figure 18.  Dissimilarities in
 the temperature  distributions recorded in the two cross sections are
 attributed  to  differences in the subsurface materials, which result in
 different distributions  of ground-water velocity   Average sround-water
 velocity is slower in  cross-section B-B' by a factor of 2.  The fluctuations
 In average  lake  temperature and the seasonal fluctuations of 8™^-^er
 temperature at several distances from the dike along cross-section A-A  at a
 depth of 4.5 m are illustrated  in  Figure 19.  The lag  time »>et«en tto
 occurrence  of-the maximum  temperature in the ground water and the maximum
 cooling-lake temperature gradually increases, and the  amplitude of tne
 fluctuation decreases  with distance from the dike.
                                      -36-

-------
               (b)  JANUARY 27, 1977
                                                        0   METERS
Figure 17.  Observed  (	)  and simulated (	) ground-water  temperature
            distributions  in  cross-section A-A' of Figure  15.
                                      -37-

-------
Simulations

    Good agreement between simulated and actual temperatures was obtained when
the model was used to simulate temperatures in cross-sections A-Af  and B-B'
(Figure 17, Figure 18, Figure 20), but only because detailed data on the
subsurface distribution of materials had been obtained.  The temperature
patterns observed in the subsurface were very sensitive to the distribution of
layers with low hydraulic conductivity.  The model was most sensitive to the
extent and depth of the clay layer under the peat (Figure 8) and to a clay
layer that is present in some areas at a depth of 6-8 m.  Over 70 borings were
made in the marsh to determine the subsurface lithology.
    (a)
      5


      10


      15
APRIL  26, 1977
                                                     -15e-
                                 100
                                                             0   METERS
Figure 18.  Observed (	) and simulated (	) ground-water temperature
            distributions in cross-section B-B1 of Figure 15.
                                     -38-

-------
    30
  O
  ui
  ff
  D

  ff
  UJ
    20
    10
  UJ
                  0
                1976
N
M     A
   1977
M
Figure 19.   Seasonal fluctuations of ground-water temperature in cross-
            section A-A'  of Figure 15 at a depth of 4.5 m at distances of
            2, 15, 50, and 84 m west of the cooling lake dike.  The
            fluctuations  of average lake temperature are also shown (	)
                                    -39-

-------
      30 II l| I l[ I I I H I I I I I I I I I I I I I I I I I I I I I I M I I I I I l| I I I I I I I I I ' I I M I "IT

                  2 METERS WEST OF DIKE
           07  \
      25 P
                                         30 M 111111111111111; 1111111111 " I " I' M 1111111111111' 11 • 11 •
                                                      5O METERS  WEST OF DIKE
                                         25-
                                                                    i.. I.. 11. i.. 111111 h , 1111	1
        JJASONDJFMAMJJASONDJFMA
        1976                1977          1978
                                           JJflSONDJFMAMJJASONDJFMA
                                            1976                 1977          1978
                   n j i ij 11 j 1111111' 11«l^n ITl »1TTT
                   15 METERS WEST OF DIKE
                                                      1 | i 11 I 11 I 11 I I | M | I 111 I | I 11 11 11 I 11 I I I I | " I'

                                                      84 METERS WEST OF DIKE
        JJASONDJFMAMJJASONDJFMA
         1976                1977          1978
                                                 Q h I I 1 1 I 1 • I I I I M I I I I I I I I I I I 1 I I I I I . I I I I I I I I I I I I I I I I I 1 I II I I I I I ! I I I I 1
                                           JJASONDJFMAMJJASONDJFMA
                                            1976                1977          1978
!


25
                  1111| • 11111111111" | M I' • I " I " I " I''
                   32 METERS WEST OF DIKE
                                        „
        JJASONDJFMAMJJASONDJFMA
         1976                1977          1978
Figure 20.   Simulated temperatures (—  —)  and  observed temperatures  (o) in
              cross-section  A-A1  of Figure 15 at a depth of  3.28 m at  2, 15,
              32, 50,  and  84 m west of the cooling lake  dike.
                                          -40-

-------
                                  SECTION 6

       LONG-TERM TEMPERATURE CHANGES IN THE GROUND WATER OF THE WETLAND
SIMULATION STUDIES

    Simulations were used to predict the long-term changes in substrate
temperature that may occur in the wetland adjacent to the Columbia Generating
Station;  Wetland water levels and ground-water flow rates into the wetland
had stabilized during the first 2 yr of operation of the generating station,
but neither ground-water temperatures nor wetland vegetation (Bedford 1977)
had yet reached equilibrium.  Prediction of the nature and magnitude of change
in vegetation was not possible without an understanding of the probable
changes in substrate temperatures.

    The model that was described in section 5 to simulate the response of
subsurface temperatures to changes in the temperatures of the cooling lake and
the air was used to simulate seasonal temperature patterns for the 12-yr
period from November 1974 to January 1987.  The governing equations, boundary
conditions, microclimatic model, and parameters remained the same in the
long-term simulation.  Changes in temperature for the 12 yr were modeled in
the same two planes, which represent two-dimensional vertical cross sections
of the ground-water system (A-A1 and B-B1 in Figure 15).

    Cooling-lake temperatures and air temperatures for the 10-yr period
1978-87 were synthesized by repeating five times the actual temperatures
recorded from January 1976 to January 1978.  For the period after April 1978
when two generating units were assumed to be operating, the temperatures were
adjusted to account for the added heat load.  When the temperature at the
generating station outlet rose above 40° C, it was assumed that the entire
heat load from the second generating unit was dissipated in the cooling
towers.  The cooling-lake temperatures at the generating station inlet for the
period 1975-87 are shown in Figure 21.

    The simulations predict that temperatures at a depth of 0.6 m will not
fall below 8° C within 200 m of the dike by 1987 and that peak temperatures
near the dike will be 10-15° C above normal and will occur in October and
November rather than in August.  The subsurface stratigraphy of the site is
such that major changes in near-surface temperatures will only occur within
350 m of the dikes, but if the stratigraphy were different the effects could
extend to much greater distances from the cooling-lake dikes.

    The temperature changes simulated in the ground-water system for the
period 1975-87 are only an approximation of the actual temperatures that may
exist in the ground-water system during this period.  Lake and air

                                     -41-

-------
            I	1	1	1	1	1	1
                                                          III
                       78   79   80  82
                                   YEAR
83   84   85   86   87
Figure 21.   Cooling-lake inlet temperatures for 1975-87 used in the
            simulations of long-term temperature change in the ground-water
            system.  Actual temperatures are shown  from May 1975 to March
            1978.  Temperatures from March 1978 to  January 1987 were
            synthesized from the 1976-78 temperature  record.
                                    -42-

-------
temperatures may deviate widely from those assumed in this report,  and
processes not accounted for in the model,  such as substrate decomposition and
flood-induced erosion, may alter ground-water flow rates.   These simulations
are, however, a reasonable approximation of the changes to be expected  in
ground-water and substrate temperatures in the vicinity of the cooling-lake.

RESUUS AND DISCUSSION

    The simulation of temperatures in the  ground-water system near  the
cooling-lake shows that temperatures in the vegetation rooting zone in  the
wetland will reach a new steady-state condition that is much different  from
the prevailing condition before the filling of the cooling-lake.  The annual
temperature patterns near the dike will stabilize by 1980, but at 150 m from
the dike temperatures will not reach equilibrium by 1987.

    Simulated temperatures in cross-section A-A1 (Figure 15) 2 m from the dike
from 1975 to 1987 at a depth of 3 m deviated from those at 0.6 m by less than
1.0° C (Figure 22).  When pre-lake temperatures are compared with simulated
temperatures at 0.6 m depth and 2 m west of the dike in cross-section A-A1 for
the periods 1978-86 and 1984-86, the change in annual temperature patterns is
pronounced.  In the period 1984-86 the temperature does not fall below  14° C,
whereas before the cooling lake was filled, substrate temperatures  at this
depth approached freezing and in the 1976-78 period the temperature fell below
5° C (Figure 23).  The peak temperatures occur approximately 45 days later in
the period 1976-78 than in the pre-lake period and approximately 30 days later
in the 1984-86 period than in the pre-lake period.  The lag decreases as the
ground-water temperature rises because the hydraulic conductivity of the
subsurface material increases with temperature,vwhich increases the
ground-water flow velocity.

    The simulated temperatures in cross section A-A1 150 m from the dike for
the period 1975-87 are shown in Figure 24.  Pre-lake temperatures at a  depth
of 0.6 m and temperatures simulated for the periods 1976-78 and 1985-87 are
shown in Figure 25.  The attenuation in the annual temperature patterns
between the pre-lake period and the 1976-78 period is a result of the
increased ground-water flow rate in the latter period.  At a distance of  150 m
from the dikes, several years are required before the impact of the
temperature variations in the lake become pronounced at a depth of 0.6  m. By
1986 the change in temperatures is pronounced with maximum and minimum  annual
temperatures elevated several degrees centigrade above pre-lake temperatures.

    In cross-section A-A1 (Figure 15) most of the ground-water discharge
occurs within 200 m of the dike.  Consequently, the atmospheric heat flux
beyond this distance is much larger than the ground-water heat flux, and thus
substrate temperatures in 1987 at a depth of 0.6 m are only slightly altered
from pre-lake conditions.  Ground-water temperatures at greater depths  do show
increases, but the rate of increase at a depth of several meters is much
slower than shown in Figure 24 for a point at a depth of 0.6 m and  a distance
of  150 m from the dike.   (Text continues on p. 48.)
                                     -43-

-------
             i	1	1	1	1—in	1
        5   76    77   78   79  80   81   82  83   84   85   86   87
Figure 22.   Predicted ground-water temperatures from 1975-87 at a distance
            of 2 m west of the cooling-lake dike in cross-section A-A1 of
            Figure 15 at a depth of 0.6 m.
                                    -44-

-------
         JAN
MAY
                             SEPT
                      JAN
MAY
                                                             SEPT
                                                      JAN
Figure 23.   Pre-lake temperatures and simulated temperatures 2  m west  of the
            cooling-lake dike in cross-section A-A'  of Figure 15 at a  depth
            of 0.6 m.   (	)  = simulated temperatures  from 1984 to  1986;
            (	)  = actual temperatures from 1976 to 1978;  (•••) =  simulated
            temperatures from 1976 to 1978 assuming  that  the cooling lake
            had not been present.
                                    -45-

-------
     O
    o

    UJ
    oc
       30
       25
2O
    UJ
    o.
    2
    UJ  10
             I     I     I
                                                   i     i     i     r
  75   76   77
                         78   79   80   81   82

                                    YEAR
83   84  85  86   87
Figure 24.   Temperatures 150 m west of the  cooling-lake dike in cross-
            section A-A1 of Figure 15 at  depths of 0.6 m and 3 m for the
            simulated period 1975-87.  (—), 0.6 m level; (	),  3  m level.
                                    -46-

-------
                 MAY
SEPT
JAN
MAY
SEPT
JAN
Figure 25.   Temperatures 150 m west  of  the  cooling-lake dike in cross-section
            A-A'  of Figure 15 at  a depth of 0.6 m.   (	) = simulated
            temperatures for 1984-86; (	> = actual temperatures for 1976-78;
            (•••)  = simulated temperatures  for 1976-78.
                                    -47-

-------
    The temperature changes simulated for cross-section A-A'  are believed  to
be representative of the changes in temperature that  will occur  in  the
vegetation rooting zone in the wetland along the dike.   The ground-water flow
patterns, which are determined by the subsurface stratigraphy, determine that
the obvious temperature changes will occur within 350 m of the dike.   The
increase in temperatures will be somewhat greater in  the southern part  of  the
wetland since the average annual lake temperature increases from approximately
21° C at the intake to 2^° C at the south end of the  lake.  This difference is
shovm in Figure 26 (a-c), which presents simulated temperatures  in
cross-section B-B' of Figure 15.  This cross section  is 200 m south of the
cooling lake intake and 1,500 m south of cross-section A-A1.   Temperatures
were simulated for 1985-8? at a depth of 0.6 m and at 2, 50,  and 200  m west of
the dike.  Temperatures 2 m west of the dike (Figure  26a) averaged  about 3° C
higher than temperatures at a similar point on cross-section  A-Ar.
                  MAY
               SEPT
JAN
MAY
                                                             SEPT
JAN
 Figure 26a.
Temperatures in cross-section B-B1  of  Figure  15  at  a  depth of
0.6 m and at 2 m west of the cooling-lake  dike.   (	)  =
simulated temperatures for 1985-87;  (•")  = simulated tempera-
tures for 1976-78 assuming no cooling  lake was present.

                        -48-

-------
   UJ
   QC
   oc
   UJ
   Q.
      30
      25
      20
       15
       nt>
       10
        JAN
     MAY
SEPT
JAN
MAY
SEPT
JAN
Figure 26b.
Temperatures in cross-section B-B*  of  Figure  15 at a depth of

0.6 m and at 50 m west of the cooling-lake  dike.   (	) =

simulated temperatures for 1985-87;  (•••) = simulated

temperatures for 1976-78 assuming no cooling  lake was present.
                                    -49-

-------
        JAN
      MAY
SEPT
JAN
MAY
SEPT
JAN
Figure 26c.
Temperatures in cross-section B-B'  of Figure  15  at  a  depth of
0.6 m and at 200 m west of the cooling-lake dike.   (	)  =
simulated temperatures for 1985-87; (•••)  = simulated
temperatures for 1976-78 assuming no cooling  lake was present.
                                     -50-

-------
                                  R INFERENCES
Anderson,  M.P.   1979.   Using models to simulate  the movement of contaminants
    through groundwater systems.   J.  Environ.  Control,  CRC  Reviews,  in  press.

Anderson,  M.P.,  and  C.B.  Andrews.   1977.   Hydrogeology, p.  83-109.   In
    Documentation of environmental change related to the Columbia Electric
    Generating  Station.  Institute of Environmental Studies, Rep.  82, Univ.
    Wisconsin-Madison,  Madison,  Wis.

Andrews,  C.B.   1976.  An analysis  of the  impact  of a coal-fired power plant on
    the ground water  supply of a wetland in central Wisconsin.  M.S.  Thesis.
    Univ.  Wisconsin-Madison, Madison, Wis.  83 p.

Andrews,  C.  1978.   The impact of  the use of heat pumps on ground-water
    temperatures. Ground Water 16:437-443.

Andrews,  C.B.,  and M.P. Anderson.   1978.   Impact of a power plant on the
    groundwater system  of a wetland.   Ground Water 16:105-111.

Andrews,  C.B.,  and M.P. Anderson.   1979.   Thermal alteration of groundwater
    caused by seepage from a cooling lake.  Water Hesour. Res., in press.

Bay, R.  1967.   Groundwater and vegetation in  two peat bogs in  northern
    Minnesota.   Ecology 48:308-310.

Bear, J.   1972.  Dynamics of fluids in porous media.  American  Elsevier,  New
    York.   764  p.

Bedford,  B.  1977.   Changes in wetland vegetation associated with leakage from
    the cooling lake of a coal-fired power plant.  M.S. Thesis.  Univ.
    Wisconsin-Madison,  Madison, Wis.  39 p.

Bedford,  B.  1978.   Alterations in the growth and phenology of  wetland  plants
    as indicators of environmental change.  Proc. Jt. Conf. Sensing Environ.
    Pollut.  New Orleans, La.  p.   170-174,

Boulter,  D.  1972.   Water table drawdown around an open ditch in organic
    soils.  J.  Hydrol.  15:329-340.

Boyle, J.M., and Z.A. Saleem.  1978.  Determination of recharge rate using
    temperature depth profiles in wells (abstract).  EOS 59(4):280.
                                     -51-

-------
Bredehoeft, J.D.,  H.  Counts, S. Robson, and J. Robertson.   1976.   Solute
    transport in groundwater systems, p. 229-256.  In Facets of hydrology.
    John Wiley and Sons, New York.

Bredehoeft, J.D.,  and I.S. Papadopulos.  1965.  Rates of vertical groundwater
    movement estimated from the earth's thermal profile.  Water Resour.  Res,
    1:325-328.

Bredehoeft, J., and G. Pinder.  1970.  Digital model of areal flow in
    multiaquifer groundwater systems:  A quasi three-dimensional model.   Water
    Resour. Res. 6:883-888.

Bredehoeft, J.D.,  and G.F. Pinder,  1973.  Mass transport in flowing
    groundwater.  Water Resour. Res. 9:194-210.

Cartwright, K.  1970.  Ground-water discharge in the Illinois Basin as
    suggested by temperature anomalies.  Water Resour. Res. 6:912-918.

Cartwright, K.  1973.  The effect of shallow groundwater flow systems on rock
    and soil  temperatures.  Ph.D. Thesis.  Univ. Illinois-Urbana, Urbana, 111.
    128 p.

Cherry, J.A., R.W. Gillham, and J.F. Pickens.  1975.  Contaminant
    hydrogeology.  Part 1:  Physical processes.  Geosci. Can. 2(2):76-84.

Cook, R.D.   1974.   Concepts and applications of finite element analysis.  John
    Wiley and Sons, Inc., New York.  402 p.

Cooley, R.   1977.   A method of estimating parameters and assessing reliability
    for models of steady state groundwater flow, 1:  Theory and numerical
    properties.  Water Resour. Res. 13:318-324.

Cooper, H.H., J.D. Bredehoeft, I.S. Papadopulous.   1967.  Response of a
    finite-diametric well to an instantaneous charge of water.  Water Resour.
    Res. 3:263-269.

Desai, C.S., and J.F. Abel.  1972.  Introduction to the finite element method.
    Van Nostrand Reinhold Co., New York.  477 p.

Dix, fl.L., and F.E. Smeins.  1967.  The prairie, meadow, and marsh vegetation
    of Nelson County, North Dakota.  Can. J. Bot. 45:21-58.

Freeze, R.A., and  P.A. Witherspoon.  1966.  Theoretical analysis of regional
    groundwater flow, I:  Analytical and numerical  solutions to the
    mathematical model.  Water Resour. Res. 2:641-656.

Gass,  T.E., and J.H.  Lehr.  1977.  Ground water energy and the ground water
    heat pun p.  Water Well J.  31 (4): 42-47.
                                     -52-

-------
Gibbons, J.W.  1976.  Thermal alteration and the enhancement of species
    populations.  In:  Thermal Ecology II.  Proceedings of a symposium held at
    Augusta, Georgia, April 2-5.  ERDA Symposium Series 40, Report CQNF-
    750425.

Green, D.W.  1963.  Heat transfer with flowing fluid through porous media.
    Ph.D. Thesis.  Univ. Oklahoma, Norman, Okla.  312 p.

Gringarten, A.C., and J.D. Sauty.  1975.  A theoretical study of heat
    extraction from aquifers with uniform regional flow.  J. Geophys. Res.
    80:4956-4962.

Hausz, W., and C.F. Meyer.  1975.  Energy conservation:  Is the heat storage
  ,  well the key?  Public Util. Fortn.  95:34-38.

Hoopes,_J., and D. Harleman.  1967.  Dispersion in radial flow from a
    discharge well.  J. Geophys. Res. 72:3595-3607.

Intercom? Resource Development and Engineering, Inc.   1976.  A model for
    calculating effects of liquid waste disposal in deep saline aquifers.
    Vol.  1.  U.S. Geol. Surv., Water Resour. Invest.,  76-61.  p.

Jaeger, J.C.  1958.  The measurement of thermal conductivities and diffusivity
    with cylindrical probes.  Trans. Am. Geophys. Union 39:708-710.

Jobson, H.E.  1972.  Effect of using averaged data on  the computed
    evaporation.  Water Resour. Res. 8:513-518.

Keys, W.S., and R.F. Brown.   1978.  The use of temperature logs to trace  the
    movement of injected water.  Ground Water 16:32-48.

Kirge, L.J.  1939.  Borehold  temperatures in the Transvaal and Orange Free
    State.  Proc. Roy. Soc. London, Ser. A,  173:450-474.

Kley, W., and H. Nieskens.  1975.  Moglichkeiten der Warmespeicherung in  einem
    Porengrundwasserleiter und technische Probleme bei einer  Ruckgewinnung der
    Energie.  Z. Dtsch. Geol. Ges.  126:397-409.

Konikow,  L.  1976.  Modeling  chloride movement  in the  alluvial aquifer  at the
    Rocky Mountain Arsenal, Colorado.  U.S. Geol. Surv., Water-Supply Pap.
    2044., Washington, D.C.   43 p.

Konikow,  L.F., and D.B. Grove.   1977.  Derivation of solute  transport in
    ground water.  U.S. Geol. Surv., Water-Resour. Invest. 77-19.  Washington,
    D.C.   30 p.

Lee,  D.R.   1977.  A  device  for measuring  seepage  flux  in  lakes and
    estuaries.  Limnol. Oceanogr.  22:140-147.

Mercer,  J.W., G.F.  Pinder,  I.G.  Donaldson.   1975.  A Galerkin finite element
    analysis of the  hydrothermal  system at  Wairakei, New  Zealand.  J. Geophys.
    Res.  80:2608-2621.

                                     -53-

-------
Meyer, C.F.,  and D.K. Todd.   1973.   Conserving energy with heat  storage  wells.
    Environ.  Sci. Technol.  7:512-516.

Millar, J.B.   1973.   Vegetation changes in shallow marsh wetlands under
    improving moisture regime.  Can. J. Bot. 51:509-522.

Mink, J.F.  1964.  Groundwater temperatures in a tropical island environment.
    J. Geophys. Res. 69:5225-5230.

Mitchell, J., W. Beckman, R. Bailey, and W. Porter.  1975.   Microclimate
    modeling of the desert,  p. 275-286.  In J>.A. de Vries and N.H.  Afgard
    [ed.] Heat and mass transfer in the biosphere.  Script a Book Co.,
    Washington, O.C.

Molz, F.J., J.C. Warman, and I.E. Jones.  1976.  Transport of water and  heat
    in an aquifer used for hot-water storage:  experimental study (abstract).
    fiOS 57:918.

Molz, F.S., J.C. Warman, I.E. Jones, and G.E. Cook.  1976.   Experimental study
    of the subsurface transport of water and heat as related to  the storage of
    solar energy.  Proc. Int. Solar Energy Soc. (Canadian and American
    sections) 8:238-244.

Myers, G.E.   1971.  Analytical methods in conduction heat transfer.
    McGraw-Hill, New York.  508 p.

Myers, G.E.   1978.  The critical time step for finite element solutions  to
    two-dimensional heat-conduction transients.  J. Heat Transfer 100:120-127.

Neuman, S.P.   1973.  Calibration of distributed parameter groundwater flow
    models viewed as a multiple objective decision process under uncertainty.
    Water Resour. Res. 9:1006-1021.

Nightingale,  H.   1975.  Ground-water recharge rates from thermometry.  Ground
    Water 13:340-344.

Ogata, A., and R. Banks.  1961.  A solution of the differential  equation of
    longitudinal dispersion in porous media, U.S. Geol. Surv. Prof. Pap.
    411-A.  Washington, D.C.  7 p.

Parsons, M.G.  1970.  Ground-water recharge rates from thermometry.  Ground
    Water 13:340-344.

Pickens, J.,  and W. Lennox.   1977.   Numerical simulation of waste movement in
    steady groundwater flow systems.  Water flesour. Res. 12:171-180.

Pinder, G.  1973.  A Galerkin-finite element simulation of groundwater
    contamination on Long Island, New York.  Water Resour. Res.  9:1657-1669.

Pinder, G., and J. Bredehoeft.  1968.  Application of a digital  computer for
    aquifer evaluation.  Water Resour. Res. 4:1069-1093.


                                     -54-

-------
Finder,  G.F.,  and W.G.  Gray.  1977.   Finite element simulation in surface and
    subsurface hydrology.  Academic Press,  New fork.  295 p.

fleddel,  D.L.,  and D.K.  Sunada.  1970.   Numerical simulation of dispersion in
    groundwater aquifer.   Colorado State Univ.,  Hydrol. Pap.  41.   79  p.

Robertson,  J.B.  1974.   Digital modeling of radioactive and chemical  waste
    transport  in the Snake River plain aquifer of the national reactor testing
    station,  Idaho, U.S.  Geol. Surv.,  open file report.  Washington,  D.C.
    41 p.

Robson,  S.G.   1978.  Application of digital profile modeling techniques to
    ground-water solute transport at Bars tow, California.  U.S. Geol. Surv.,
    Water-Supply Pap. 2050.  Washington, D.C.  28 p.

Sammuel,.E.A.   1968.  Convective flow and its effect on temperature logging in
    small diameter wells.  Geophysics 33:1004-1012.

Scheidegger,  A.E.  1961.   General theory of dispersion in porous  media. J.
    Geophys.  Res.  66:3273-3278.

Schneider,  R.   1962.  An application of thermometry to the study  of ground
    water.   U.S. Geol. Surv., Water Supply Pap.  1566-H.  Washington,  D.C.
    16 p.

Schneider,  R.   1964.  Relation of temperature distribution of ground-water
    movement  in carbonate rocks in central Israel.  Bull. Geol. Soc.  Am.
    75:209-214.

Sorey, M.L.  1971.  Measurement of vertical groundwater velocity  from
    temperature profiles in wells.  Water Resour. Res. 7:963-970.

Stallman, R.ri.  1960.  Notes on the use of temperature data for computing
    groundwater velocity.  Soc. Hydrotech. France (Nancy, France), Rapp. 3.
    p. 1-7.  (Reprinted, 1963, in U.S. Geol. Surv., Water-Supply  Pap. 1544H.
    Washington, D.C.  p. 36-46.)

Stallman, R.W.  1965.  Steady one-dimensional fluid flow in a semi-infinite
    porous medium with sinusoidal surface temperature.  J. Geophys. Res.
    70:2821-2827.

Steinhart,  J.S., and S.R.  Hart.  1968.  Calibration curves for thermistors.
    Deep-Sea Res.  15:497-503.

Supkow,  D.J.   1971.  Subsurface heat flow as a means for determining  aquifer
    characteristics in the Tucson Basin, Pima County, Arizona. Ph.D. Thesis.
    Univ. Arizona, Tucson, Ariz.  182 p.

Tanner,  C., and W. Pelton.   1960.  Potential evapotranspiration estimates by
    the  approximate energy balance of Penman.  J. Geophys. Res. 65:  3391.
                                     -55-

-------
Tsang, C.F., M.J. Lippmann, and P.A. Witherspoon.  1976.  Numerical modeling
    of cyclic storage of hot water in aquifers.  Trans. Am. Geophys. Union,
    EOS 57:918.

Vadas, R.L., M. Keser, P.C. Rusanowski, and B.R. Larson.  1976.  The effects
    of thermal loading on the growth and ecology of a northern population of
    Spartina alterniflora.  Thermal Ecology 11, National Technical Information
    Service (CONF-750425).  Springfield, Va.  p. 54-63.

Van Wijk, W.R., and D.A. de Vries.  1963.  Periodic temperature variation.  In
    W.R. Van Wijk [ed.] Physics of plant environment.  North Holland
    Publishing Co., Amsterdam, Holland.

Von Herzen, R., and A.E. Maxwell.  1959.  The measurement of thermal
    conductivity of deep sea sediments by a needle probe.method.  J. Geophys.
    Res. 64:1557-2563.

Walker, B.H., and R.T. Coupland.  1968.  An analysis of vegetation-environment
    relationships in Saskatchewan sloughs.  Can. J. Bot. 46:509-522.

Werner, 0., and W. Kley.   1977.  Problems of heat storage In aquifers.  J.
    Hydrol. 34:35-43.

Willard, D.E., W. Jones, B. Bedford, M. Jaeger, and J. Benforado.  1976.
    Wetland ecology, p. 89-135.  In Documentation of change related to the
    Columbia Electric Generating Station, Institute of Environmental Studies,
    Rep. 69, Univ. Wisconsin-Madison, Madison, Wis.

Willard, D.E., W. Jones, B. Bedford, M. Jaeger, and J. Benforado.  1977.
    Wetland ecology, p. 110-125.  In Documentation of environmental related to
    the Columbia Electric Generating Station, Institue of Environmental
    Studies, Rep. 82, Univ. Wisconsin-Madison, Madison, Wls.

Winslow, J.D.  1962.  Effect of stream infiltration on ground-water
    temperatures near Schenectady, New York.  U.S. Geol. Surv., Prof. Pap.
    450-C, Art. III.  Washington, D.C.  p. C-125-C-128.

Wisconsin Public Service Commission.  1974.  Final environmental impact
    statement for Generating Unit No. II of the Columbia Electric Generating
    Station.  Madison,  Wis.  147 p.

Zienkiewicz, O.C.  1971.  The finite element method in engineering science.
    McGraw-Hill,  London.  521 p.

Zienkiewicz, O.C.  1977.  The finite element method in engineering science.
    McGraw-Hill,  London.
                                     -56-

-------
                                  APPENDIX A

                 DATA SPECIFICATIONS FOA THE HATER-FLOW MODEL
    This appendix details procedures used at the site of the Columbia
Generating Station to gather field data on the hydrogeologic system.   The data
were collected from 1971 to 1975.  They were used to document changes in the
system caused by construction and operation of the generating station and to
establish the boundary conditions and parameters for the model of the
ground-water flow system.

DATA FOR SPECIFYING BOUNDARIES

Side Boundaries

    The position of the site in the regional ground-water flow system was
needed to determine the size of the area to be modeled.  Water-level data and
well logs were collected from more than 100 domestic wells in the vicinity of
the Columbia Generating station.  Most of the data were obtained from the
U.S. Geological Survey in Madison, Mis.  The data were used to estimate
aquifer transmissivity, flow rates, and the extent of the regional
ground-water flow system.  Transmissivity was estimated from pump tests and
specific capacity data.  Flow rates were computed by using the estimated
transmissivity value and the water-table gradients existing in the field.  The
extent of the flow system was estimated by contouring water-level data from
the wells open to the sandstone aquifer in the vicinity of the site.

    The information on the regional ground-water system was used to justify
the assumption that the left and bottom boundaries of the cross section
modeled were no-flow boundaries and to specify the flow rates for the right
boundary of the cross sections modeled.

Upper Boundary

    Over 80 well point piezometers 3.2 cm in diameter were installed on the
site to monitor the position of the water table.  The piezometers had 46-cm,
80-gauge screens.  They were driven in place in the lowlands and augered into
place in the uplands.  They were positioned so that they were open to the
aquifer Just below the water table.  Monitoring was done with a steel
measuring tape at monthly intervals beginning in summer 1971.  The water
levels in two wells were recorded continuously.  Beginning in fall 1974 water
levels were recorded at  14 points in the wetlands on a monthly basis.  All
observation points were  surveyed to establish relative elevations.  The
locations of the wells are shown in Figure A-1.
                                     -57-

-------
             • U.W. MONITORING WELLS
             © D.N.R. WELLS
             * WELLS THAT HAVE BEEN REMOVED
                                                   61
                                            • 58
                                                MILES
                                                  Vt
                                                                    77-79
                                                                      •
                                             KILOMETERS
Figure A-l.   Location of monitoring wells at  the Columbia site.
                                      -58-

-------
    Water-level data from all monitoring wells were punched onto IBM Cards.   A
computer program was written in PL/1 (Programming Language, University of
Maryland) that graphs the water level in each well by water year, computes the
rate of change in water levels, and computes the water-table gradient between
the observation points and the rate of change in the gradients.  These
water-level data were used to specify ground-water potential along the upper
boundary of the cross sections modeled and to justify the use of a
steady-state model.

DATA FOR SPECIFYING STATE VARIABLES

Permeabilities

    The range of permeabilities for the various materials on the site was
determined by combining data from a pump test, slug tests, and laboratory
permeability tests with lithologic and grain-size information.  Permeability
distribution.in the system was then modeled by using lithologic and grain-size
information from over 100 borings on the site.  (The locations of the borings
are shown in appendix B.)  A pump test was run in the alluvial materials on
the site in winter 1972 by the Lane Engineering Firm of Chicago, 111.  The
data were analyzed by using the Dupuit-Theim equation for steady radial flow
without vertical movement.  Slug tests were run on 23 of the observation wells
on the site.  Water was withdrawn from the wells, and the recovery time was
recorded.  The data from these tests were analyzed with the techniques
developed by Cooper, Bredehoeft, and Papadopulos (1965).

    Laboratory permeabilities were run on 23 samples with soil test
permeameters under both falling and constant head conditions and under various
density conditions.  Grain-size analyses (percentages of medium sand, fine
sand, silt, and clay) were run on 43 diverse samples from the site.  The
information was then correlated with the permeability tests, and a working
model was formulated between lithology and permeability.  This correlation of
permeabilities with lithologies was used with the logs of the borings to
develop seven cross-sectional permeability models.

Temperature

    Temperatures were monitored bimonthly beginning in 1973 by lowering a
Yellow Springs Instrument thermistor into observation wells.  Before the
temperature was recorded, the observation well was pumped with a hand-operated
vacuum pump  until  the temperature of the water became nearly constant.

CALIBRATION  DATA

    In addition to the data  collected to define the boundary and state
variables  in  the  system, extensive  field data were collected to provide a
check on model simulations.
                                     -59-

-------
Piezometer Nests

    Nineteen observation wells were installed in the lowlands at the site.
They were open between 4.5 and 7.6m below the water table.   Seven wells were
installed in the uplands so that they were open between 9.1  and 18.3 m below
the water table.  These wells were placed adjacent to an observation well open
just below the water table.  Water levels in these wells were monitored at
monthly intervals.  The vertical gradients and the rate of change of vertical
gradients have been computed for each of these wells for each monitoring date.
The data from these wells were used to judge the fit of the models'
simulations of the ground-water system to the actual situation.

Direct Measurements of Flow

    Surface water in the wetland west of the cooling lake drains through
well-defined channels during low river stages.  The amount of discharge
through these channels was monitored several times during fall 1975 on
overcast days with a rod-suspended pygmy current meter.  Measurements were
taken at 0.6 of the total depth at 1-ft intervals across the streams.  By
similar methods the flow in the drainage ditch on the east side of the cooling
lake was measured several times at both the northeast corner and the southwest
corner of the cooling lake.

    Discharge into the drainage ditch on the east side of the lake from the
ground-water system was monitored with seepage collectors.  Seepage collectors
are 55-gal barrels cut in half whose open end is positioned into the substrate
at the bottom of the ditch (Lee 1977).  On the closed end, which is also
submerged, the spout is covered by a plastic bag.  The flow of water into the
plastic bag is recorded, which provides a direct measurement of the
ground-water inflow into the ditch in the area covered by the barrel.  By
means of the seepage collectors, rates of ground-water flow into the drainage
ditch on the east side of the lake were determined at various locations on
several different dates.
                                   -60-

-------
                                  APPENDIX B

              TECHNIQUES FOR DETERMINING BOUNDARY CONDITIONS AND
                      PARAMETERS IN THE HEAT-FLOW MODEL
    The model developed to simulate the transport of heat away from the
cooling lake at the Columbia Generating Station requires for its solution the
specification of initial and boundary conditions for the partial differential
equations describing heat flow and the equations describing water flow.  Five
sets of parameters are also required:  Hydraulic conductivity, storage
coefficient, thermal conductivity, heat capacity, and dispersivity.  The field
and laboratory techniques used to define the boundary conditions and the
parameters needed for the water-flow equation were described in appendix A.
The field techniques used to measure temperature and the laboratory techniques
used to measure thermal conductivity and heat capacity are described in this
appendix.  The technique used for specifying the dispersivity parameters is
described in section 6.

TEMPERATURE MEASUREMENTS

    Temperatures were recorded weekly at 40-110 points in the subsurface in
the vicinity of the Columbia Generating Station site from August 1976 to
January 1978 (Figure B-1).  The depths at which temperatures were taken, the
dates when readings were taken, and the manner in which temperatures were
measured are listed in Table B-1.

    Most temperatures in the field were measured with an electronic device
with a thermistor as the temperature sensor.  A thermistor sensor system was
used because of its accuracy and its versatility.  A relative accuracy of
0.1°C is easy to maintain in the field, and probes could be buried, lowered
down wells, or located far from an accessible point.  Liquid expansion
thermographs with an accuracy of + 0.5°C were used for continuous temperature
recording because these mechanical devices are much easier to maintain and
less expensive than a continuous recording device with a thermistor sensor.

    An electronic temperature measuring device is extremely simple in concept
(Figure B-2).  It consists of a thermistor whose electrical resistance is
almost entirely a function of its temperature as the sensing device and an
electronic measuring device to measure either directly or indirectly the
resistance of the sensor, generally an ohmmeter or a wheatstone bridge.  Two
types of thermistors and three measuring devices were used at the Columbia
Generating Station site.
                                     -61-

-------
Figure B-l.
Location of temperature sampling points at the Columbia
Generating Station site.
                                   -62-

-------
    TABLE B-l.  TEMPERATURE SAMPLING POINTS:  LOCATIONS, DEPTHS,
                PROBES, AND FREQUENCIES OF READINGS
Depth below
Sampling surface at which
point Distance temperature
number from dike (m) was measured (m)
130

131


132


133

134

135

35
2 0.
3.
15 0.
1.
6.
32.5 0.
1.
6.
51

86

131.5

15,
05,
15,
51,
10,
15,
52,
10,






3 2,3,4
0.
6.
0.
3.
9.
0.
3.
9.






91, 1.52,
10, 8.70
46, 0.91
05, 4.57
76
46, 0.91
05, 4.57
76






,5,6,7,8,9
Probe used
to take
temperature
in situ

in situ


in situ


in situ

in situ

in situ

probe

probe


probe


probe

probe

probe

probe lowered
Approximate
frequency
of readings
8/8/76-8/77 weekly
9/77-1/78 monthly
8/8/76-8/77 weekly
9/77-1/78 monthly

8/8/76-8/77 weekly
9/77-1/78 monthly

8/8/76-8/77 weekly
9/77-1/78 monthly
8/8/76-8/77 weekly
9/77-1/78 monthly
8/8/76-8/77 weekly
9/77-1/78 monthly
8/8/76-1/17/78
36
37
68
 24
 28
140
 31
 61
194
2,4,6,7,8


2,3,4,5


4,6,8,10,12


3,4,5,6,7,8,9,10


2,4,6,8,10,12,13


2,4,6,8
down 3.175 cm;
PVC pipe
probe lowered
down 3.175 cm;
PVC pipe
probe lowered
down 3.175 cm;
PVC pipe
probe lowered
down 3.175 cm;
galvanized pipe
probe lowered
down 3.175 cm;
PVC pipe
probe lowered
down 3.175 cm;
PVC pipe
probe lowered
down 3.175 cm;
galvanized pipe
weekly

8/10/77-1/18/78
weekly

8/10/77-1/18/78
weekly

8/10/77-1/18/78
weekly

5/31/77-1/18/78
weekly

8/10/77-1/18/78
weekly

4/22/77-1/18/78
weekly
                                  -63-

-------
TABLE B-l (continued).
                           Depth below
Sampling                 surface at which   Probe used
  point      Distance      temperature       to take
 number   from dike. (m)  was measured (m)  temperature
                                                   Approximate
                                                    frequency
                                                   of readings
  143
   43
    75
   57
L1/L2
    37
   126
    73     west side of
          ditch

    74     east side of
          ditch
14 m east of
ditch
Lake inlet
and outlet
Areal coverage
of marsh (not
shown in
Figure B-l)
Transect 1   0.2 m
to transect
42 (not shown
on Figure B-l)
2,4,6,8


2,4,6,8,10
                  probe lowered
                  down 3.175 cm;
                  galvanized pipe
                  probe lowered
                  down 3.175 cm;
                  galvanized pipe

1,2,3,4,5,6,7,8,9 probe lowered
                  down 2.54 cm;
                  PVC pipe

1,2,3,4           probe lowered
                  down 3.175 cm;
                  PVC pipe

4,6,8,10,12,16    probe lowered
18,20.21          down 3.175 cm;
                  galvanized pipe

10,12,14,16,18,20 probe lowered
                  down 3.175 cm;
                  galvanized pipe
               8,10,12,14
                  probe lowered
                  down 3.175 cm;
                  galvanized pipe

                  thermistor
               0.1, 0.9 m at 45  in situ
               locations in marsh
               1.0, 3.0 m
                  shallow
                  temperature
                  probe
4/22/77-1/18/78
weekly


4/22/77-1/18/78
weekly


5/31/77-1/18/78
weekly


5/31/77-1/18/78
weekly


5/31/77-1/18/78
weekly


10/1/76-12/12/78
monthly


10/1/76-12/12/78
monthly


daily average
value recorded
every day plant
is operating

10/28/76, 11/6/76
11/13/76, 12/4/76
                                  9/30/77,  10/23/77,
                                  12/03/77, 1/10/78,
                                  3/30/78
                                    -64-

-------
               THERMISTOR
       POWER
       SUPPLY
VARIABLE
RESISTOR
Figure B-2.  A simple circuit for measuring temperature.
                        -65-

-------
Thermistors

    Most of the temperature data at the Columbia Generating Station site were
taken with Yellow Springs Instrument Company Series 400 thermistors.   These
thermistors are calibrated by Yellow Springs, and all thermistors of this type
have the same calibration curve with a maximum error of 0.1°C.   The probes are
guaranteed to remain calibrated for 1 yr.  The temperature-resistance
characteristics of these probes, as well as the resistance change for a 0.1° C
temperature change, are shown in Figure B-3.  The thermistors were used in
three modes:  as sensors that were lowered down a wall, as sensors mounted on
2-m brass rods, and as sensors placed semipermanently in the subsurface.

    The easiest and least expensive means for measuring temperature in the
subsurface is to lowsr a probe down a fluid-filled well and take readings at
various depths on the way down.  The only problem with this technique is that
the temperature profile in the well will not correspond to the  temperature
profile in the suftsurface materials if the temperature gradients are  above a
critical value.  For a 3.175-cm well the critical gradient is approximately
 1°C/m at  10°C and 0.2°C at 25°C (Sammel 1968).  The critical gradient was
exceeded  in many cases at the Columbia site.  The error induced in the
temperature readings is probably less than +_ 0.5°C if the critical gradient is
exceeded by less than 1 order of magnitude.  Near the surface,  where  the
gradient can be large, the induced error was greater.  Nevertheless,  because
of tne advantages of simply lowering a probe down a well, this  technique was
used to record temperatures at many locations in the subsurface.

    The thermistors that were lowered down wells were purchased as a unit
designed for this purpose.  These thermistors were sealed by the manufacturer
in an epoxy resin and attached to a shielded cable.  The maximum diameter of
the assembly was 0.4 cm.  The time constant of these thermistors is 7 s.  A
thermistor assembly as described above was mounted inside a 0.95-cm brass rod,
2.1 m long, for use as a shallow temperature probe for accurately measuring
temperatures near the surface.  The probe could easily be pushed into the soft
sediments on the Columbia site to a depth of 2 m.  The thermistors purchased
as a unit were calibrated with respect to each other so that temperatures
between probes could be compared with an accuracy of +• 0.05°C.

    In one plane perpendicular to the dike (Figure fi-1), 42 thermistors were
mounted in six wells with six to eight thermistors in each well.  These
thermistors were purchased as unmounted thermistors.  Each thermistor was
soldered to three wires on a cable, two wires were soldered to one lead, and
the assembly was potted with silicon sealer inside a piece of tygon tubing
0.64 cm by 5 cm (Figure B-4).   The wires soldered to each thermistor were of
sufficient length to reach the surface.  All the thermistors to be placed in a
well were bound together and wired to a connector that was placed at the
surface when the  thermistors were inserted in the well.  A cable was wired
from the thermistors at each of the six wells to a control box  on the dike.
The control box on the dike was equipped with a 48-position switch to allow
access to each thermistor.   Three wires were connected to each thermistor so
that the lead-wire resistance could be measured and subtracted  from the
resistance measured for the thermistor plus lead wire.


                                     -66-

-------
    40
   o 30
   o
   U.
   O
   IE 20
   UJ
   Q.

   CO
   S
   X
   o
     10
                                                  8000
                                                              7000
                                                  6000
                                                                   CO
                                                                   2
                                                  5000
                                                       UJ

                                                  4000 z
                                                       CO

                                                       CO
                     10            20
                           TEMPERATURE °C
                                   30
                                                  3000
                                                  2000
                                                              1000
Figure B-3.  Temperature-resistance characteristics of Yellow Springs Series

            400 thermistors ( — = right axis) and the change in resistance

            with a 0.1° C change in temperature ( 	 = left axis).
      .64cm
                       SILICON  SEALER
          TYGON   THERMISTOR  32GAUGE   SOLDER   22 GAUGE
          TUBING                LEADS     JOINT    COPPER LEADS
Figure B-4.
Potted  thermistor assembly showing the details of the

thermistor mounts.
                                  -67-

-------
    The wells into which the thermistors were placed were 10-m deep,
water-filled wells constructed with 3.175-cm schedule-40 PVC pipe jetted  into
place to minimize disturbance of subsurface materials.   After the cables  were
placed in the wells, Fiberglas insulation was stuffed into the remaining  space
to minimize convection.  Thermistors at depths of 0.1,  0.4,  and 0.8 m were
installed near the well rather than in the well.  The field arrangement is
shown schematically in Figure B-5.  The thermistors on each cable were not
calibrated relative to each other, but the cables were pulled twice during the
18 months of temperature-data collection to check for deviation from the
calibrated range.

    In addition to the in situ thermistors described above,  200 additional
thermistors (Fenwall Disc Type JB3U1) were placed at 60 locations in 12  lines
perpendicular to the dike in the marsh at depths of 0.1 and 0.9 m.  Each  line
of thermistors was wired to a control box on the dike .   Four sets of readings
were  taken in October and November 1976, but in December the multicolored
22-gauge copper wire, which was used to connect the thermistors to the control
boxes on the dikes, became incorporated in the winter dwellings of the marsh
muskrats .

Meters Used for Temperature Monitoring

    Because of delays in obtaining equipment, three meters were used during
the  17 months of temperature monitoring.  All meters were frequently
calibrated against a known resistance.

    From August through December  1976 all thermistor resistances were measured
on a  Shall era ft four-dial wheatstone bridge with a 5  A full scale needle
galvanometer.  Resistances could be read accurately to ± 3ft at 4,000 n with
this meter, and it probably had an operating accuracy of +_ 10 tt.

    A Digitec digital recording ohmmeter mounted in a vehicle was used for
measuring resistances from January 1977 through January 1978.  Most of the
resistance measurements were made with this instrument.  This instrument  had
an operating accuracy at 4,000 ft of approximately +_ 4 fl and was operated  at
all times in the field with a 12 VDC to 120 VAC square wave converter and at a
temperature of at least 20°C.

    A four-dial wheatstone bridge with a digital null meter was constructed as
a backup for the digital ohmmeter and as a portable field meter (Figure B-6).
This meter had a working accuracy at 4,000 ft of approximately ±5 & and was
used primarily with the shallow temperature probe for measuring temperatures
near the surface.

Teature
    Two resistances for each location were recorded in the field, a line
resistance and a total resistance.  The resistances were then coded, punched,
cross checked for errors, and then converted to temperature by the following
equation (Steinhart and Hart 1968):

                         T"1 - A + B log R + C(log R)   ,

                                    -68-

-------
                                                    CONTROL BOX
             MARSH SURFACE
                   \
 Ill
 o
 4
 u
10
                                                    	WELL CASING,
                                                      PVC
                                                    	WELL DIAMETER'
                                                    1   3.175cm
       131.5
             86
    51
METERS
     32.5
FROM WEST EDGE
                                        OF
15
DIKE
                                                   2 O
-30
Figure  B-5.
          Schematic of thermistor placement in  subsurface wells at the
          Columbia Generating Station site.  The wells shown correspond  to
          locations 130-135 of Figure B-l.   Horizontal dimensions are
          distorted, but vertical dimensions are approximately correct.
  1.1 V-r-
         DECADE
         RESISTANCE
         BOX
                                          ANALOGIC AN 2545
                                         DIGITAL PANEL METER
                        Display=
                                                         xlOOO
:"••
•...'
•'"":
•.."
2
?
•
7
Figure B-6.
          Schematic of digital bridge circuit used as a portable field
          meter for measuring ground-water temperatures.
                                   -69-

-------
where the constants A, 8. and C were determined from the  thermistor
calibration data, and T~^ is the inverse Kelvin temperature.

    The relative accuracy of all temperature measurements taken  in  the  study
is within + 0.5°C of the actual temperature of the thermistor probe.  The main
sources of error were thermistor calibration, meter drift,  and incorrect meter
calibration.  The correspondence of probe temperatures to actual subsurface
temperatures is unknown.  The deviation between these temperatures  is a
function of the mode of measurement used and was assumed  to be small in all
cases.

THERMAL CONDUCTIVITY DETERMINATION

    The thermal conductivities of unconsolidated materials at the Columbia
Generating Station site were measured by the needle probe method (Von Herzen
and Maxwell 1959).  In the needle probe method, which was developed for
determining the thermal conductivity of deep sea sediments, a thin  needle is
inserted into a sample and is heated along its length.  The sample  can  be as
small as 3 cm in radius and 7 cm long.  From a record of  temperature increase
in the needle over time, the thermal conductivity can be  easily  calculated.
The probe and the experimental arrangement for measuring  thermal conductivity
for this study are shown in Figure B-7 .  The thermistor in the probe was
calibrated to 0.05°C.  The needle probe was not checked for accuracy against a
known standard.

    The theory for the needle probe has been worked out in detail by Jaeger
(1958).  His analysis shows that temperature increase of  the  probe  is
approximated by
                                            ),  t>a2/a,
                                   1.7811 *

where a = thermal diffusivity of the sediment sample, L2/t;  a = probe radius,
L; t = time, t; and q = heat input per unit length per unit  time,  H/tL.

    A plot of T versus In t will give a straight line, the slope of which
determines K for known values of q.  The data collected from measurements of
the thermal conductivity of three samples from the Columbia Generating Station
site are shown in Figure 8-8.   All samples were checked for reproducibility,
the standard error was less than 3>, and in all measurements thermal
conductivity was assumed to be independent of direction.

DETERMINATION OF HEAT CAPACITY

    Since heat capacity of a composite material can be determined  from a
weighted average of the heat capacities of the individual parts in the
composite, it was assumed that the heat capacity of the unconsolidated
materials at the Columbia Generating Station site could be determined from the
relation

           pCs " pCwXw + pCorgXorg + pCqtZXqtz + pCclayXclay  '
                                    -70-

-------
               NEEDLE CROSS SECTION
          2O GAUGE
          'HYPODERMIC
EPOXY

THERMISTOR
.STAINLESS
STEEL

HEATING
WIRES
              LOW RESISTANCE WIRES
              WHICH CONNECT TO HIGH
             'RESISTANCE HEATING WIRE
              AT BASE OF NEEDLE

              THERMISTOR LEADS
        4-PIN
     CONNECTOR
Figure B-7.   Details of (a) the needle probe and (b) the experimental arrange-
             ment to measure thermal  conductivities of unconsolidated
             materials.  The power  supply was used to supply a constant
             voltage to the needle  probe.  The voltage was then monitored on
             the volt meter, and  the  ohmmeter was used to record the resis-
             tance changes in  the thermistor embedded in the thermistor.
where Xw, Xorg, Xqtz, and  Xciay denote fractional parts  of the material made
up of water, organic material,  quartz and feldspar, and  clay minerals,
respectively.  The heat  capacities of these four components are  listed  in
Table B-2.  In a given sample all volatiles were assumed to be organic, all
materials greater than 0.001 mm in diameter were assumed to be quartz and
feldspars, and all materials less than 0.001 mm in diameter were assumed to be
clay minerals.  The water  content of the samples was  determined  in  the
Quarternary Laboratory at  the University of Wisconsin-Madison, and  the
organic, sand and silt,  and clay fractions were determined by the Soil  and
Plant Laboratory of the  University of Wisconsin-Madison.
                                      -71-

-------
(a)
34


u
w 32
K
3

<
tc
HI
£30
Z
u
1-
28
. . . .....J .... -v
-
0
_

• .,
^^^^^
A ^^^^^

•^^^
l^r
^JQ
- ^^**
^* +
•*^
<* , -, . 1 ....1 . . . I . ...
(b)

u
• 36
Ul
3
£T
^
a
u 34
a.
Z
u
K
32

. ...... ..| , - • , .^,.
•X
^p
— 1^
Jr
^^
,j^^
^^^

— ^^
./
Jr
^
«'
.^
, , 1 , , ..1 , ,,!,,,,
10 100 1000 10 100 I00(
SECONDS SECONDS
(C) 44
42
u
w 4°
K
2 38
|36
IX
«- 34
32

. ..,...., .
X
/
/*





\ / , 1
X . , 1 , . ..1 , ..!..,,











10 100 IOOO
Figure B-8.
                       SECONDS
Thermal conductivities of three ground-water samples from the
Columbia Generating Station site which were analyzed with the
needlfe probe.  Thermal conductivities were calculated to be
4.9 x 10~3 cal/°C cm sec for medium to fine sand (a), 5.0 K 10~3
cal/°C cm sec for fine sand (b), and 1.5 x 10~3 cal/°C cm sec
for hemic peat (c).
           TABLE B-2.   HEAT CAPACITIES OF  THE  CCMMON COMPONENTS  OF
                      UNCONSOLIDATED GLACIAL MAT£fiiALSa
                          Component
                                Heat capacity
                                  (cal/m3°C)
Quartz and feldspar
Clay minerals
Organic matter
Water e 25°C
0.51 x 106
0.51 x 106
0.60 x 106
1.00 x 106

aAdapted from Van Hijk and  de Vries (1963).
                                    -72-

-------
                                  APPENDIX C
            A FINITE ELEMENT PROGRAM TO SIMULATE SINGLE-PHASE HEAT
              FLOrt OH CONSERVATIVE MASS TRANSPORT IN AN AQUIFER

BRIEF DESCRIPTION OF THE MODEL

    The following computer code is designed to solve the two-dimensional
equations of ground-water flow and heat or mass transport for an aquifer  by
using the finite element method.  The program was written to facilitate the
analysis of ground-water contamination in shallow glacial aquifers.   The
program solves for aquifer potentials and temperatures or concentrations.
However, simultaneous simulation of the transfer of heat and mass in an
aquifer is not possible.  The aquifer to be simulated may be artesian, a  water
table, or a combination of both.  It may be heterogeneous and anisotropic  and
may have irregular boundaries.  Tne program was designed to simulate
cross-sectional problems, but it has also been used successfully to  simulate
areal problems.

    The basic procedure is to solve a ground-water flow problem for  potentials
in an aquifer and tnen to use the potential information to compute a velocity
distribution in the aquifer.  The heat or mass transport equation is then
solved for temperatures or concentrations (Figure C-1).
      Solve for ground-
      water potentials
      in the aquifer.
                                 If a transient problem
                                 repeat for specified number
                                 of time steps using the
                                 Crank-Nicolson method to
                                 step the solutions forward
                                 in time.
                    Compute ground-
                    water velocities.
Solve for nodal
temperatures or
concentrations.
Figure C-1.
The basic procedure for linking the ground-water flow and the
transport equations of the model.   In general, smaller time steps
will be used for the heat/mass transport equation than for the
water-flow equation.
                                     -73-

-------
    This flexible program includes these  features:

1) Meuman,  Dirichlet,  and mixed  boundary  conditions can  be handled.

2) The basic element shape is  a  quadrilateral.   The sides of the elements can
   be linear, quadratic,  or cubic, and  thus  element sides can  have two, three,
   or four  nodes.  Any given element  can  have  four linear sides or any mix of
   linear,  quadratic,  and cubic  sides.  This feature allows considerable
   flexibility in designing a  finite  element grid.

3) The program handles hydrodynamic dispersion.

4) Point, line, or areal  sources of water, heat, or mass can be represented.

5) The problem need not be a linked problem; if  desired, the model can be used
   to solve only for aquifer potentials.

6) hydraulic conductivity can  be represented as  a  function of  temperature.

    The program is relatively  simple  to use, but those unfamiliar with the
development of the equation describing  the convective-dispersive transport of
neat or mass in an aquifer, or those  unfamiliar  with the limitation of the
numerical approximation technique, may  experience  difficulties in using the
program.  Oscillations and instabilities  are frequently  encountered in solving
the convective-dispersive equation because (1) the equation behaves like a
hyperbolic partial differential  equation  when  convective transport dominates
over diffusive transport; (2)  in the  numerical technique used  in the  program
velocities are not continuous  everywhere; and  (3)  the numerical scheme cannot
transmit sharp contaminant fronts.  Oscillations and instabilities can
generally be dampened  by reducing element size or  by reducing  the time step.

THE EQUATIONS SOLVED BY Trie] PROGRAM

    The program was written to solve  the  differential equation describing mass
or energy transport in a ground-water aquifer.   Since the rate of transport is
a function of ground-water velocity,  the  first step in solving the transport
equation is to solve the  ground-water flow equation  so that ground-water
velocities  can be determined.

Hater-Flow Equation

    The following partial differential  equation, which can be  used to describe
ground-water flow in a confined  or water-table aquifer,  is solved by  the
program

                aJT^il-f^ + W-S|*=0  ,    i.J-1.2   ,
                3xi    ij 3xJ           8t                               (c-i)

where %j are the components of  the hydraulic  conductivity tensor, which may
be a function of temperature,  Lf1; b = aquifer  thickness, L;   = ground-water
potential,  L; w = flux of recharge per  unit  area,  L/T; and S  = storage
coefficient for a confined aquifer, or  the specific yield in  a water-table

                                   -74-

-------
aquifer, dimension! ess.  Boundary coalitions may  be some combination of the
following:

                                  -        - - R  - p    = ° '
where                             ^i^l
                        D' . . =04 - + o
                          11    L   q
                                              qlq2
                       D ' „ . — V, n ~  \. ""W ~ «m / 	I	  >
                         21     12    VT,    T7   q
                                  q2q2
                                              q   '
aL = longitudinal dispersivity, L; Ts = transverse dispersivity, L;
velocity or specific discharge, which is calculated by

 q  = mean velocity,  (q-| +

    The remainder of the  parameters are defined differently for a mass
 transport  problem and  for a heat transport problem (Table C-1).
 Boundary conditions may be some combination of the following:

                                C = constant,                            (C-6)

                                 = constant,                             (C-7)
                             -
                             9xi
                                 3C
                                 ^ a ^ * '                             (C-8)

                                     -75-

-------
       TABLE C-1.   PARAMETiSRS FOR BOTH MASS AND HEAT  TRANSPORT PROBLEMS
             Mass transport  problem
  Heat transport problem
           Components of molecular
           diffusion tensor,  L2/t;
     C     rfass concentration,  M;

     g     Unity, dimensionless;

           Effective porosity,  dimen-
           sionless;

     A     Mass recharge rate,  M/t.
Components of the thermal
conductivity tensor for the
saturated medium, H/t LT;
Temperature, T;

Heat capacity of water,
Heat capacity of the saturated
medium, H/L^T;

Heat generation or recharge
rate, H/L3t.
                                ac
                                                                         (C-9)
                                3C
                                    a C - C«
                                                                        (C-10)
where C  = concentration beyond  an exterior  boundary.

    A physical  significance can  be ascribed  to  each of the  terms in £q. (C-4),
since the equation can be thought of as a mass  balance equation which states
that flow into  and out of a differential volume sums to zero.  The processes
described by each of the terms are:   transfer by molecular  diffusion,
  3       3C
 •jjc~ ^D£4^~^ »  transfer by velocity  fluctuations, called dispersive
                                    -76-

-------
0                                       ;    source
transfer, •- (g  D'^   -)  .    convective transfer,    g q±    -
                                sr
or sink,  R:  storage changes,  p ~ .
                                at
    The use of Eq.  (C-4) to describe mass or heat transport in an  aquifer
implies the following assumptions:  (1) the system is chemically inert,  (2)
the aquifer is incompressible, (3) density is constant within the  system,  (4)
all transport occurs by molecular diffusion, mechanical dispersion, and
convective transport, (5) flow can be represented in two dimensions, and (6)
the flow of water is laminar.  Bear (1972) and Konikow and Grove (1977)
present derivations of t,q.  (C-4) to describe solute transport in ground water.
The development of Eq. (C-4) to descrioe energy transport in ground water  will
be discussed.

THE APPROXIMATION TECHNIQUE

    The finite element method is used in the program to approximate Eq.  (C-1)
and Eq, (C-4) to solve for the unknown potentials and concentrations in  an
aquifer.  The finite element method, like the finite difference method,  is a
discretization technique.  Tne continuous aquifer is represented by a number
of regions, called elements, each bounded by a set of nodes.  Integral
equations describing water flow and heat or mass flow are developed for each
region, giving two sets of simultaneous algebraic equations which  are solved
to determine the unknown potentials and concentrations.  The finite element
method generates the integral equation from the governing differential
equation and evaluates the integral over an element by a numerical integration
scheme.  Thus, the finite element method operates on a region, whereas the
finite difference technique only operates on a point.

    The finite element method has several advantages over the finite
difference method for solving Eq. (C-1) and Eq. (C-4), and, in fact, the
finite difference method may be unacceptable for solving fiq. (C-4) (binder
1973).  The advantages are:  (1) The finite element method is characterized by
less overshoot and less numerical smearing of a concentration front  (Finder
and Gray 1977); (2) velocities can have a functional representation, which
reduces oscillations in the solution to Eq. (C-4); (3) boundary conditions are
much easier to treat in the finite element method, and (4) irregular
boundaries are easier to handle with the finite element method.

    Two techniques can be used in formulating the approximating integral
equations that are the foundation of the finite element method.  The procedure
used in this study to develop the required finite element formulation is the
fiayleigh-flitz procedure, or the minimum potential energy procedure, which  is
based on the calculus of variation.  The other procedure, the Galerkin method,
is more general in application and has been in favor in the literature.  It
was used by Pinder and Gray (1977) to develop the approximating integral

                                     -77-

-------
equations for equations similar to Eq.  (C-1) and Eq.  (C-4).   Both methods
yield identical finite element formulations for Eq.  (C-1) and Eq.  (C-4).

    finite £lemejit Formulation

    The basis of the variational principle is to develop an integral equation
describing the total energy within a region and then to minimize the integral.
This principle implies that potentials  and concentrations in an aquifer will
always be at levels that minimize total energy for a given set of boundary
conditions.

    Integral equations can be developed from first principles by quantifying
potential energy in the system, but the Euler-Lagrange equation (Myers 1971)
facilitates the process by transforming the governing partial differential
equations to integral equations that describe total energy within a region.

    The variational statements for Eq.  (C-1) and Eq.  (C-4) for an element
within the aquifer are
                                                           2
        H  - 1/2 // KO--) dxdx  - 1/2 // [2W+ + s-] dx     ;  (C-11)
                 ar                      o                         o
             - p -fT-lbdx.dx- + 1/2 §  H(C  - 2CC»)bds + 1/2 § gq .  C bdS
                 o t     i  L                                    TL.

                                                                        f C—
             + 2 §  gq± C«CbdS  ;


where nw = total potential for water flow, n T = total potential for heat or
mass transport, h = convection coefficient (units depend on type of problem),
and b = aquifer thickness.

    The next step is to minimize Eq. (C-11) and Eq. (C-1 2), which is
accomplished by setting the first derivatives of Hw and n^ equal to zero.
Before this is done, it is convenient to introduce the following approximating
functions to describe the system at any interior point as a function of nodal
values:


                                * -  [N]  {N> ,                          (C-13)


                              3<>/3x  - [n,x]  {*N}  ,                       , and                        
                                    -78-

-------
                             3C/3x = [N'.x]  {CN>  ,                       (C-16)

where $ ,  C = potential and concentration respectively at any interior point;
     = ground-water potentials at nodal points in an element,  a column matrix;
     = concentrations at nodal points in an  element, a column  matrix; and  [N],
[N ' ] = shape functions which relate nodal values  of potential  or concentration
to values at any point in the element, a row matrix.  This program  allows  each
element to have a minimum of 4 and a maximum of 12 nodes.  For example,  if an
element has six nodes the shape function has the  following form:
where i = .  .  .  $5 are nodal values of potential.

    Derivation of the coefficients of the shape functions is not
straightforward,  but an understanding of how they are derived is not  necessary
for a general understanding of the finite element method.  It is sufficient to
know that if an element side has two nodes the shape function is linear,  if
the element side  has three nodes the shape function is quadratic,  and if  the
element side has  four nodes the shape function is cubic.   Therefore,  the  more
nodes in an element, the higher the order of the polynominal used to
approximate an interior point.  This program uses isoparametric  elements  of
the serendipity family.  (See Pinder and Gray 1977 for more details.)

    To solve the  flow problem Eq.  (C-11) is then minimized as follows:
            an.,
                        'ij[N,xi]i  [N,Xj.
                                   T
                   - S -       ,   [N]1  [N] 3X3x  = 0
    The integrations are performed using Gaussian quadrature (Zienkiewicz
1971).  After the equation is integrated, it is convenient to combine terms
and put the equation in the following form:

                                        8{V
                         [S]  UN> - [C]—§r- [R]  ,                    (C-19)

where [s] = bKij[M,Xi]T [w,xj] 8x18x2, which is often called the element
structure matrix; [c] = S [NJT [N] 3XT3x2, which is called the element
capacitance matrix; and [r] = W [N] 3x-|3x2, which is called the element
recharge matrix.

    An equation in the form of Eq. (C-19) is developed for each element  in an
aquifer.  These equations are then added to give the following equation:


                         [s]  UN} = [c]  -^l^1- [R]  ,                    (c-20)
                                    -79-

-------
where [S] = global structure matrix, [C] = global capacitance matrix,  and [fi]
= global recharge matrix.

    The column matrix will have one column for each node, and the  square
matrices will have a column and a row for each node.  The square matrices will
be symmetric and banded, and only half the matrix will be stored in the
program.

    The time derivative is evaluated by the Crank-Nicolson approximation:


             ([C] H-f^- [S]) {*N}1+1 - ([C] - [S]  |^) {^ + [R] At .    ((,_21)


    The final matrix equation is solved by Gaussian elimination (Cook  1974).

    To solve the mass or heat transport problem,  Eq. (C-12) is minimized as
follows:

                     [Dtj  [N,x±]T  [N.Xj] +g D'^  [N.x^^N.



                             T                  3{CKT}     T
              -   g 1  [N,x]T[N]b3x3x{C}-  p    "   [N]
                H[N]T [N]bdS{CN) -   HfNjC^dS +   q±g[N]T [N]bdS
                            ^C^ [N]bdS -
0 .             (C-22)
    This equation can be reduced to a form analogous to Eq. (C-19).  As
before, one equation is developed for each element, and these equations are
summed to give an equation analogous in form to Eq. (C-20).  The time
derivative is again evaluated by the Crank-Nicolson approximation.  The final
matrix equation, which is assymmetric due to the presence of the terms
[N,XiJT[N] in Eq. (C-22), is solved by the Gauss-Doolittle method (Desai and
Abel 1972).

CONSIDERATIONS IN DESIGNING A FINITE ELEMENT GRID

    The basic step in using a finite element scheme is to subdivide the region
of interest into an assemblage of smaller regions called elements.  The
process of discretizing is an exercise of engineering judgment in which one
must choose the number, shape, size, and configuration of elements in such a
way that the original continuum is simulated as closely as possible.  A grid
configuration that provides a greater number of nodes will generate a more
accurate solution, but at the same time will lead to more computational
effort.  In general, the mesh should be refined in the region of steep
gradients.  The element shapes are quadrilaterals having straight boundaries,
although curved boundaries are also possible.
                                    -80-

-------
    All boundary conditions described for Eq.  (C-1)  and (C-4), except for
constant flux boundaries, are handled explicitly by  the program  and are
explained in the data-input section of the appendix.   A constant flux boundary
of zero flux is assumed by the finite element  method if no other boundary
condition is specified.  A finite flux boundary is treated by assigning a line
source or sink of the appropriate magnitude.   Four sample  problems will be
presented to illustrate the discretization procedure.

    In designing the finite element grid, 10 considerations should be
observed.

1) All elements should be made nearly rectangular, since distorted element
shapes decrease the solution accuracy.

2) Elements with more than four nodes should be used sparingly.   Although an
element with more than four nodes uses a higher order approximation function
for inte'rior points, replacing an element with six nodes by two  elements with
four nodes each usually provides as good an approximation  with less
computational effort.  (This occurs because a  lower  order  Gaussian quadrature
scheme can be used for elements with only four nodes.) Elements  with more than
four nodes are best utilized for refining the  finite element grid in critical
areas (Figure C-2).
Figure C-2.
Examples of refinement of a finite element grid.   Left:   bad
use of multi-node sides; center:  the preferred way of
discretizing a region; right:  good use of multi-node  sides.
3) Nodes should be placed close together in areas where system parameters
exhibit spatial changes, in areas where velocity is relatively rapid  or  the
velocity distribution is complex, and in areas where a sharp temperature or
concentration front is expected.

4) Exterior element boundaries across which a flux of mass or energy  occurs
must have only two nodes.

5) The velocity field near an exterior element boundary across which  a flux of
mass or energy occurs must not be complex.  Preferably, all flow is normal to
the boundary.  This can be insured by setting up dummy elements along a
boundary with a high hydraulic conductivity in a direction normal to  the
boundary.  Likewise, the boundary should be oriented normal to one of the
principle directions of the hydraulic conductivity tensor.

6) Boundaries within the area to be modeled should be located accurately.
Distant boundaries can be located approximately with fewer nodes by expanding
                                    -81-

-------
the grid.  Elements with more than four nodes are very useful for expanding
the grid away from critical areas.

7) Constant concentration boundaries are generally unrealistic for the
transport equation.  Transport into the system can best be represented  by
multiplying the water flux across the boundary by the concentration exterior
to the system.

8) The grid should be oriented to coincide with the principle directions of
the hydraulic conductivity tensor.

9) The core requirements and computation time are proportional to the number
of nodes representing the aquifer and to the number of times system parameters
or time steps are changed.

10) If no boundary condition is specified along an exterior boundary, a
no-flow boundary is assumed.

DATA DECK INSTRUCTIONS

The Data Groups

    All data are read into the model with a free format, except for
alphanumeric data.  In free format, the data requested by one read statement
may be put on one input card or as many input cards as the reader desires.
Bach bit of data, however, must be followed by either a comma or a blank.   The
data for each read statement must begin on a new card.  Two read statements
cannot read from the same card.  The data and program output for four  sample
problems will be presented to illustrate the data deck preparation procedures.
Several aspects of the data preparation that may be confusing are explained in
more detail in the data notes following this section.  The read statements
that must be included on every program execution are underlined.

Group 1—

    This group of data, which is read by both the main program and the  data
input subroutine, contains data required to dimension the model and data
required to determine the type of problem.

Read Statement     Variable                     Definition

  J. and £           TITLE         Any title that the user wishes to print on
                                  two lines at the start of output.  The input
                                  must contain two cards.

     3              LM            Number of elements
                    LM            Number of nodes
                    MBAND         Bandwidth (notes 1 and 2)
                                    -82-

-------
                  KPRlNT
                  LOW
                  MQ
                  NLA
                  CFACT
  and
2 and 8
VV(I)
An integer that can have values in the range
of one to five.
1—problem to be solved is a steady-state
ground-water problem.
2—problem to be solved is a transient
ground-water problem.
3—problem to be solved is a steady-state
ground-water problem linked to a transient
heat or mass transfer problem.
4—problem to be solved is a transient
ground-water problem linked to a transient
heat or mass transfer problem.
5—problem to be solved is a transient mass
or heat transfer problem.

An integer that can have values of 0 or 1.
Set to 0 to suppress printing of input data;
otherwise set to 1.

Set to 99 for calculating coefficient
matrix stability for a transient problem
(see note 3); otherwise set to zero.

Set to 2 if one card is to be read in for
each element giving the element node
numbers.  Set to 0 if the spatial
structure for the problem is rectangular
and all elements have only four nodes
(data group II).

An integer specifying the method to be used
for calculating velocities in a linked
problem.  Set to 1 if velocities are to be
calculated at each of the Gauss points; this
is generally the better method.  Set to 2 if
velocities are to be calculated only at the
center of each element.  Set to 0 if the
problem is not linked problem.

Set as 1fi13; this value is used to maintain
the specified boundary conditions.  Set
higher if specified boundaries are not
reproduced by model.

Format statement for printing out nodal
values.  Two cards are required "(note 4).

Format statement for printing out element
values.  Two cards are required.
                                  -83-

-------
Group II—

    The data contained in this group of cards are  required  to  specify the
spatial structure of the nodal array.  The type  of data  to  be  input is
determined by the value previously specified for variable MQ.  If the
structure is rectangular with four nodes per element,  the program
automatically generates a grid and numbers the nodes and elements, and the
data input is simple.  If the structure is not rectangular,  the location of
each node must be input as must cards specifying the nodes  defining each
element (notes 5, 6,  7, and 8).
Read Statement
      1
Variable
 Q
If MQ equals 2 only:

one card for each   J
node                XLOC
                    iLOC
one card for each
element
J
NOD(K,1)
NODU.2)
NOD(K,3)
                    NODOC.5)
                    NOD(K,1)
IF MQ equals 0 only:
                    KSPACX
                    KSPAC1

                    BU)
       5

       6

Group III--
 MLS
                            Definition
              Input the alphanumeric characters—up II.
               node number
               X location of the node
               \ location of the node

               element number
               The element node numbers.   The node numbers
               for each element must be listed counter-
               clockwise, starting at the  corner nearest
               the origin.  List the nodes in sequence;
               if a node is not a corner node, place an
               asterisk after it.  Repeat  the first node as
               the last value on each card.
              Number of nodes in the X direction.
              Number of nodes in the X direction.

              KSPACX values,  listing the  X  coordinates in
              order; smallest value should  be  listed
              first.

              KSPAC1 values,  listing the  X  coordinates in
              order; .largest  value should be listed
              first.

              Optional (note  8).

              Optional (note  8).
    This group of data contains information on  the  initial conditions and on
the specified boundary conditions.   The data to be  read  in depends on the
value specified for KTXPE.
                                   -84-

-------
Read Statement     Variable

       1            W


If KTlPfi equals 1:

       2            HfiA(I)
IF KTlrti equals 2:
                    ALPH
                    ALPhM
                    KBOUN
IF KTlPt; equals 5:

       2
                    HI (I)
                    HfcA(l)
ALPHA

ALPHAM
KBOUND
                    PC*
                            Definition

              Input the alphanumeric  characters—
              111.
              Array used to specify constant  potential
              boundaries.  A zero is entered  if a
              potential is not specified at the node;
              otherwise the specified potential value is
              entered.   See note 9 for the  format for
              entering these values.
              The length of each time step.
              Multiplication factor for each time  step.
              Length of the next time step is equal  to
              the length of the old time step multiplied
              by ALPHM.  The program works most
              efficiently if ALPHM-equals 1.
              Set to 1  if subroutine BOUND is to be  called
              at each time step during execution of  water-
              flow equation.  Set to 2 if subroutine BVAL
              BVAL is to be called at each time step during
              execution of water-flow equation (note 13).
              Set to 99 when determining structure matrix
              stability for water-flow equation (note 3).
              Otherwise set to 0.

              Array used to specify the initial potential
              at each node (use format explained in  note
              9).

              See above.
The length of time step for heat/mass
equation.
Multiplication factor for the time step.
An integer then can havve values 0,  1,  2,
or 99.  See KBOUN above, except KBOUND
applies to the heat or mass flow equation.
heat capacity of water, or 1 in a mass
transport problem with units of meters,
grams, and ppm.

An array used to specify the initial
temperature or concentration at each
node (note 9).
                                    -85-

-------
                    HEA(I)

                    HEAD(I)
IF KTlPfi equals 4 or 5:
                    ALPH
                    ALPHM
                    BOUN
                    ALPHA
                    ALPttAM
                    BOUND
                    PCW
See above.  Specified potential values.

Array used to identify specified
concentration boundaries.   A zero is  entered
if temperature or concentration is not
specified at the node; otherwise the
specified value is entered (note 9).
See description above.



See description above.

See description above.
       5            A(I)          See description above.

       6            HfiA(I)        See description above.

       7            h£AD(I)       See description above.

Group IV—

    This group of data contains information on the model parameters.

Head Statement     Variable                     Definition
       1
                    MMA1A
                    FACTh

                    FACTV

                    FACTwS
                    FACTH1
                    FACTV 1
                    FACTHS
Input the alphanumeric characters—
Group IV.

An integer, less than 51,  specifying the
number of different material types in the
problem.  A homogeneous area has only one
material type.

This card contains multiplication  factors
for the parameters to be input  later.
Factor for X direction hydraulic
conductivity.
Factor for 1 direction hydraulic
conductivity.
Factor for storage coefficient.
Factor for X direction thermal  conductivity.
Factor for 1 direction thermal  conductivity.
Factor for specific heat capacity  of
saturated medium.
                                     -86-

-------
                    FACTDL
                    FACTDT

                    MA1(J)
MMATA cards,
one card for each
material type.
                    KX
                    Mi
                    STO
                    PX

                    PI

                    PCX

                    DIFF(J,1)
                    DIFF(J,2)
Factor for longitudinal dispersivity.
Factor for transverse dispersivity.

Type of material array.  Integer values
in the range of 1-50 are entered for each
element, indicating the type of material in
that element (format explained in note 9).

Integer number in the range of 1-15
identifying the material type.

X-direction hydraulic conductivity
1-direction hydraulic conductivity
Storage coefficient
X direction thermal conductivity, or
molecular diffusion coefficient.
Y direction thermal conductivity or
i direction molecular diffusion coefficient.
Heat capacity of the saturated medium, or in
a mass transfer problem the porosity.
Dispersivity coefficient, lateral.
Dispersivity coefficient, transverse.
NOTE:   Only values for J,  WX,  HI,  and  STO need  be  entered  if KT1PE equals 1 or
       2.   If KTIPE equals 5,  values for WX,  Wl, and STO must not be entered.

Group V—

    This group of data contains information on  sources  and sinks.  Line,
areal, or point or sinks can be handled  directly.   Sources are positive, sinks
are negative.  If there are no or  sinks, three  cards must  be entered, each
with two zeros.
Bead Statement    Variable

       1            Q

       2            KGEfl



                    KGENH



       3            INFLOW(J)
                    HEAT(J)
              Definition

Input the alphanumeric characters—Group  V.

Set to 1 if there is a uniform source or
sink of water over at least one element in
the problem; otherwise set to zero.

Set to 1 if there is a uniform source of
energy or mass over one element in the
problem; otherwise set to 0.

Include this only if KGErt is not equal  to 0.
One value for each element specifying the
rate of water generation per unit area  per
unit time in the element (note 9).

Include this only if KGENH is greater than
0.  One value for each element specifying
                                     -87-

-------
       10
       11
       12
Group VI—
                   LWATEfl

                   LhEAT


                   NWAT£R(J)

                   AW ATM (J)



                   NHEAT(J)

                   AHEAT(J)
                   LI NEW

                   LINEH
                   NLINEW(i,1)
                   NLINEH(I,2)
ALINEU(I)
NLINEH(I,1)
NLINEH(I,2)
ALlNEh(I)
the rate of mass or heat generation per unit
area per unit time in the element  (note 9).

Integer values specifying the  number of
point  or sinks of water.
The number of point sources or sinks of
energy or mass.

Node number of a point source  or sink of
water.
Rate of water pumpage per unit time.  Enter
sufficient cards to contain a  node number
and a rate for each water point source.

Mode number of point source of energy or
mass.
flate of energy or mass generation  per unit
time at the specified source.   Enter
sufficient cards to contain a  node number
and a rate for each point source.

An integer specifying the number of line
sources or sinks of water.
An integer specifying the number of line
sources or sinks of heat.

Element number of line source  of water and
boundary code (note 10).
Enter sufficient cards to contain  an element
number and a code for each line source of
water.

Enter one value for each line source giving
the rate of water recharge or discharge per
unit length per unit time.

Element number of line source of heat or
mass and boundary code (note 8).  Enter
two values for each line source of heat or
mass.

Enter one value for each line source giving
rate of heat input or output per unit
length per unit time.
    Data specifying if the  problem  is a  cross-section problem or an areal
problem.
                                     -88-

-------
Read Statement    Variable

        1           Q

        2           KAREAL


        3           EH
                    ITER
                    BOT(I)
                             Definition

                Input  the  alphanumeric  characters—Group VI.

                Set to 0 for cross-section problems; set to
                1  for  areal problems.

                Error  criteria  on ground-water potential
                for a  steady-state  problem.  Solution is
                reached by an iterative procedure in which
                transmissivities are adjusted until the
                potential  changes by less than the error
                criteria.

                Number of  iterations permitted for a
                steady-state solution  to be reached.

                Include only if KAREAL  equals 1.  Enter the
                Dottom elevation of the aquifer at each
                node (note 9).

                Initial potential values in the aquifer
                (use format explained  in note 9).  Do not
                input  values if the initial values were
                input  in data Group III.
Group VII—
    This group of data contains information on the  location of  flow and
convective flux boundaries.  These types of boundaries  are  only used in the
mass or heat transport equations.  Skip cards 2-7 of this group if the problem
is not a linked one.  Flow and convective boundaries are discussed in data
note 11.
Read Statement

       1
Variable

  Q


  NCONV
                    NCOMI,2)

                    CONV(I)
              Definition

Input the alphanumeric characters—Group
VII.

An integer specifying the number of
convective boundaries.

Element number of the element containing
this type of boundary.
This boundary code identifying the side
across which the flux occurs.
The value of the convection (transfer)
coefficient for a convective boundary.

Include sufficient cards to contain  three
values for each boundary of this type.
                                    -89-

-------
                    TlNF(I)
                    LEL
                    NtiLU.2)
                    AEL(I,2)
The temperature at a distance from the
boundary (temperature at infinity).
List one value for each boundary identified
by card 3 of this group, and list them  in
the same order as on card 3.

An integer specifying the number of element
sides across,which a flux of water occurs.

Element number and the boundary code.
Include two values for each element  boundary
of this type.

One value for each element boundary of  this
type specifying the temperature or
concentration of the incoming fluid.
Group VIII--
    This group of data contains information needed  for calculating a mass
balance.  The data identifies the  exterior boundaries on which potential,
temperature, or concentration are  specified (refer  to data note  12).
.Head Statement    Variable
                    LFLUXw
                    LFLUXH
                    NFLUXto(I,1)

                    NFLUXW(I,2)
              Definition

Input the alphanumeric characters—
Group VIII.

An integer specifying the number of
exterior element sides on which potential
is specified.
An integer specifying the number of exterior
element sides on which temperature or
concentration is specified.

Include only if LFLUXW is greater than 0.
Use sufficient cards to list two values
for each element side.
element number of element with a specified
potential on an exterior side.
Side index number, identifying which of the
four element sides the flux is across.  An
integer in the range of 1 to 4.  See
explanation notes.
                    NFLUXH(I,1)   Include only if LFLUXh  is greater than
                    NFLUXH (1,2)  zero.  Format is the same as above, except
                                  that here boundaries must be identified
                                  on which temperature or concentration is
                                  specified.
                                     -90-

-------
Group IX—
    This group of data contains information that controls the  flow of the
program and allows access to routines for changing boundary  conditions at each
time step, and it includes a routine for a moving boundary in  a cross-section
problem (see data note 14).
Read Statement

        1

        2
Variabig

  Q

  LA

  LD

  LE
                    LF

                    LG


                    LH

                    N TIME (I)
        5

        6


If KTYPE equals 1:

        7           Ml*
              Definition

Input the alphanumeric characters—Group  IX.

Set to 0 unless ENTfil LAKE is to be called;
if so set to 1.
Set to 0 unless ENTfli CHANG or ENTR1 CHAN
is to be called; if so set to 1.
Set to 0 unless water flows are to
recompute at uneven intervals;  if so,  L£
should equal the number of intervals at
which water flows are to be recomputed
(must be less than 100).
Set to 0 unless the problem is to be posed
in radial coordinates; if so, set to 1.
Set to 0 unless the problem is to have a
moving boundary; if so, set LG equal to the
number of nodes which are to move.
Set to 0.

Include only if Lfi is greater than 0.
LE values listing the time steps at which
flows are to be recomputed; values must be
listed in increasing order.

Include only if LF is equal to 1.   Input
1,1,1,0,0,0 if problem is to be quasi-radial
with aquifer thickness being equal to  the
x coordinate.  Input -1,1,1,0,0,0 if problem
is to be quasi-radial with aquifer thickness
being equal to the y coordinate.

Input values needed by EflTRi LAKE.

Input values needed by subroutine ADJUST
(note 14).
                Set to 1  if potentials  are  to  be written
                on file 14;  otherwise set to 2.
                                     -91-

-------
If KTYPE equals  2:

        7
MA

MB

MW

MC

MD
If KTlflS equals 3,  4,  or 5:

        7           MO

                    MP

                    MV

                    MQ
                    MR

                    MS



                    MT

                    MU
                    ME
Total number of time units for the
simulation.
Number of time steps between printing  of
head values.
Number of time steps between writing heads
on file 14.
Number of time steps between printing  of
mass balance.
Number of time steps between printing  of
flows.
(MA, MB, MW, MC, and MD must not equal
zero.)
              Total number 'of time units for the
              simulation.
              Number of time steps between printing nodal
              values.
              Number of time steps between printing nodal
              values on file 14.
              Number of time steps between printing of
              mass balance; if set to a negative number,
              mass balances will  not be computed which
              results in a considerable savings in CPU
              time.
              Number of time steps between recomputing
              water flows.
              Set to 1 if initial hydraulic conductivity
              values are to be adjusted for changing
              temperature in the  aquifer; otherwise set to
              0 (note 15).
              Set to 1 if using dispersivities greater
              than zero; otherwise set to 0.
              Set to 1 if water flows are not to be
              printed when flows  are recomputed;
              otherwise set to 0.
              (MO, MP, MV, MQ, and MR must not
              equal zero.)
              Set to 1 if the program is to stop after
              water flows are computed; otherwise set
              to 0.
Data Explanation Motes

1) Program size—

    The storage  required for the  program  is approximately

                     2000 + LN(11 + MBAND x 4) + LM x 25,

                                     -92-

-------
where LN is the number of nodes, LM is the number of elements,  and MBAND is
the bandwidth.

    The second statement in the main program is used to change  the amount of
storage allocated for the program.  The parameter variable N specifies the
maximum number of nodes, the parameter variable L specifies the maximum number
of elements, and the parameter variable M specifies the maximum bandwidth.  On
a UNIVAC 1110, if the program is compiled in FORTRAN V,  the combination of 297
nodes, 260 elements, and a bandwidth of 13 is the maximum permissible.  This
arrangement uses 64K of memory.  If the program is compiled in  ASCII FORTRAN
on a UNIVAC 1110, the program size could be quadrupled.

    If the program is being used only to solve potential problems, the value
of the parameter variable M in statement of 1 of the main program need only be
equal to one-half the bandwidth plus one.  For a linked problem M must be
equal to or less than the bandwidth.

    Also, if LWATtifl, LrifcAT, NWATiiR, NhiiAT, LINEW, LlNEH (data Group V), NCONV,
LiiL (data Group VII), LFLUXfc,  or LFLUXH (data Group VIII) exceed  50, statement
1 of the main program must be changed.  The parameter variable  Z  must be
changed from Z = 50 to 2 = the maximum value of the above-listed  variables.

2) Bandwidth--

    The bandwidth depends upon the largest difference between any two nodes in
a single element and is equal to the maximum difference between any two nodes
in an element in a structure plus one (Figure C-3).
                              10
                              11
                              12
                                                   10
                                                         11
                                                               12
       (a) bandwidth equals five
(b)  bandwidth equals six
Figure C-3.  Examples of numbering of nodes in a structure.

The nodes in a structure should be numbered so as to make the bandwidth as
small as possible.

3) Calculation of Stability for a Transient Problem—

    The EIGtitt subroutine computes the maximum (critical) time step that can be
used in a transient simulation to insure that the solution computed by the
Crank-Nicolson method will be nonoscillatory.  The routine is based on a
modified Eigen value extraction technique developed by ^yers (1978).  The
scheme slightly underestimates the maximum (critical) time step.  Tne equation
used to estimate the critical time step is
                                     -93-

-------
                   At
1
-- min
                                        '1.1
                                        LN
                                   1,1
where Atc = safe estimate of the critical time  step,  S±tj  =  the jth entry in
the ith row of the structure matrix,  and C^ i = the diagonal term  in the ith
row of the capacitance matrix.

    The critical time step is a function of system parameters,  the x and y
spacing, and boundary conditions.  The routine  prints out  the maximum  time
step for each node in the structure and prints  out the critical time step.

4) Standard FORTRAN formats—

    Standard FORTRAN, formats are used for printing node and  elements values.
Values are printed consecutively starting with  the value for node or element
1.  The format input statements allow the user  to  specify  how the  values are
to be printed out.  This option can be very helpful if an  irregular grid is
used.

    Typical formats for printing out  node and element values for a grid 5
nodes by 5 nodes would be:  (1X,5(5G12.6/)) for the node values; and
(1X,4(4G12.6/)) for the element values.  (The word FORMAT  is not used; the '('
goes in column one of the data card.) The node  values would  then be printed
out on five lines, five values on each line.

5) Orientation of the Finite Element Grid—

    The finite element grid must be orientated  correctly,  since flows  and
boundary conditions will be treated incorrectly if the grid  is  orientated
incorrectly.  The program requires a standard cartesian orientation (Figure
C-4).
Figure C-4.  Cartesian orientation of finite element grid.

Values in the x direction must increase from left to the right and values in
the y direction must increase from the bottom to the top.
                                     -94-

-------
6) Element Node Numbering—

    Typical elements are shown in Figure C-5 to illustrate  the manner in which
element node numbers must be input to the program.
                                        The input data:

                                        element #      the node numbers

                                            1      2  6   5*   4*   3   1

                                            2      587    3    4*  5

                                            3      69856
Figure C-5.   Typical elements of a finite element  grid  and  the  correct method
             of numbering the elements.

7) The Rectangular Grid Generator—

    The rectangular grid generator generates a grid in  which  the nodes and
elements are numbered in the form shown  in Figure  C-6.
                   1    5	9
1
2
2
3
3
4
4
6
S
7
6
8
7
10
8
11
9
12


















Figure C-6.   Numbering of nodes and elements in a rectangular grid.

    The program assigns to node 1 the first x and y coordinates  read  in.  For
proper orientation the list of x-coordinate values read in must  be arranged so
the smallest value is read in first, and the list of y coordinates must  be
arranged so  that the largest value is read in first.

8) Selection of Variants of Rectangular Grid—

    The program can create the two variants of a rectangular grid shown  in
Figure C-7.

    To select the options shown in Figure C-7:

1) set KSPACX = -KSPACX.

2) set MLS = 0 for option (a); set MLS = 1 for option (b).
                                     -95-

-------
Figure C-7.
Two variants of a rectangular grid that can be generated.
Left:  grid with sloping upper surface; right: grid  with all
rows sloping.
3) Input KSPACX values for the D array.   Input one value  for each set of nodes
in the x direction.  These values specify how much the  y  coordinates of the
nodes are to be adjusted from the values specified by C(I).   For the examples
illustrated above the values in the D array would be:  0,  0.5,  1.0,  1.5, 2.0,
2.5.  These values need not increase or  decrease  monotonically.

9) Format for Reading in Array Data—

    Several options are provided to the  user to simplify  the often  arduous
task of initializing array value.  The user must  initially specify  one value
to be assigned to every position in the  array.  The user  then has the options
of (1) leaving the array as is, (2) reinitializing the  entire array by
inputting one value for each position in the array, or (3) selectively
changing the initialization.  The last option is  accomplished by specifying
the array position that is to be reinitialized and the  new value.   An array
position will correspond to a node or element number.  This  format  may be used
when initializing fl, R1, HfiA, HEAD, MAT, and EOT.
    The format to be used is:

Bead Statement    Variable

       1            QI

                    TL
                    FACT
                                   Definition

                     Value to be assigned initially to each
                     position in the array.
                     Three options:
                     a.  set to -1 if one value is to be input
                         for each node or element.
                     b.  set to 0 if all values in the array
                         equal QI.
                     c.  set to a positive integer specifying
                         the number of array values to be
                         different from QI.
                     Factor by which each of the array values
                     is to be multiplied.
If IZ = -1:
                                     -96-

-------
       (sufficient cards to
   contain one value for
   each node or element)
If IZ - 0:

If IZ > 0:
       (sufficient cards to
   contain IZ x 2 values)
10) Boundary Codes—
Enter the values to be assigned to each node
or element.  Values are to be entered
consecutively beginning with node or
element 1.  After one value is entered for
each node or element, enter 999 which is a
sentinel character.

No cards are needed.
Enter a node or element number and the  value
to be assigned to this node or element.
Enter two values for each array position
that is to be reinitialized.
    When specifying line fluxes and boundary fluxes, the side of an element
across which the flux is occurring must be designated.   The sides of an
element are coded as shown in Figure C-8.
Figure C-8.  Numbering of sides of an element.

11) Flow Boundaries—

    Data Group VII allows the user to specify several types of boundary
fluxes.  The type of boundary fluxes permitted across an element side  are:

             convective, where transfer is equal to AH(Ti - T°°);

                 flow out, where transfer is equal to V^A;

                 flow in, where transfer is equal to V^T»A;

where A = the length of the element side multiplied by the aquifer thickness;
H = a transfer coefficient; for heat transfer the units are, H/L2tT; T^ =
temperature or concentration on the boundary; T°° = temperature or
concentration beyond the boundary; and V^ = water velocity normal to the
boundary.

    If the boundary is to have more than one type of flux, for example, a  flow
and a convective flux, it should be treated for purposes of data input  as
though there were two separate boundaries.

                                     -97-

-------
12) Mass Balance—

    The mass balance routine computes  an  approximate mass balance.  Several
types of boundary fluxes are computed  only  approximately by the mass balance
routine, and often the mass balance  errors  calculated by this routine will be
much higher than  they really are.  The program generally has mass balance
errors of less than 1% for both the  water-flow and  the transport equations.
The user must Identify the specified head and 'temperature or concentration
boundary conditions.  The program identifies the  other boundary conditions and
sources and sinks.

13) Boundary Conditions That Vary With Time—

    Because the program could handle a variety of boundary conditions that
might change at each time step,  a program to cover  all possibilities could not
be written.  The user must write additional routines to change boundary values
at each time st^p.

    Several entry points in subroutine BOUNDA are provided for this.  Starting
addresses for all arrays in which a  user  is likely  to change values at each
time step are passed to this subroutine as  well as  the time step counter KS.
The entry points available in this subroutine are:

(a) ENTRY LAKE

    The routine is called if LA = 1  (data Group IX).  This routine is called
immediately after the data input routine  and is called only once for a given
program execution.  The routine can  be used to read in information to be used
in altering boundary conditions during execution  of the program.

(b) ENTRY BOUND

    Called during solution of the water-flow equation by subroutine LOAD if
KBOUN (data Group III) equals 1, or  called  during solution of the transport
equation by subroutine LOAD if KBOUND equals  1.

(c) ENTRY BVAL

    Called during solution of the water-flow equation by subroutine LOAD if
KBOUN equals 2, or called during solution of the  transport equation by
subroutine LOAD if KBOUND equals 2.

(d) ENTRY CHANG

    Called during solution of the transport equation by subroutine LOAD if LD
= 1 (data Group IX).
                                     -98-

-------
(e) ENTR1  CHAN

    Called during solution of the  water-flow equation by subroutine LOAD if LD
= 1.  The  program logic  is shown on the  program  flow chart, which shows the
position of the calls to the  entry points  in BOMDA by subroutine LOAD.

14) Documentation for Moving  Boundary  Routine—

    For cross-section problems the surface boundary can be programmed to move
in response to  changing  recharge rates.  Only  boundaries with boundary code 1
may move (note 10).   The moving boundary routine is called by setting the 5th
value (LG) on card 1,  Group IX, equal  to the number of nodes that will move.
The following cards are  then  added in  the  appropriate position in data group
IX.
Read Statement

       1
Variable

  NSTEP
  NPRINT

  NPR

  ERROR

  FACTOR
                    WOV (J, 1)
                    NMOV(J,2)
                    BMOV(J,3)
                    BMOV(J,6)
              Definition

Number of time periods to be stimulated.
Number of time periods between printing  of
potentials.
Number of time periods between printing
flows and mass balance.
Change allowed in node location before
structure matrix is recomputed.
Factor to be multiplied by recharge  rates.

Four values for each moving boundary:
Element number on left of node*,
Element number on right of node*,
Storage coefficient,
Initial recharge rate.
•Note:  If several nodes are to be constrained to  rise  and  fall at the same
        rate, the left and right element number for each node of this type
        must be preceded by a minus sign (-).

    After the last card of Group IX one card is added  for each pumping period.

Read Statement    Variable                      Definition

                    BKS           Identifier for this  time  period.
                    TIME          Length of this time  period.
                    FACT          Factor that relates  initial recharge rates
                                  to recharge rates at  this time step.  The
                                  current recharge rates are calculated by
                                  multiplying FACT by  initial recharge rate.

15) Hydraulic Conductivity as a function of Temperature—

    The program in subroutine F£ adjusts hydraulic conductivities for changing
temperature distributions.  The following relations are used:
                                     -99-

-------
                    (1.917 - 0.05635T + 0.007 IT2) x  10~3, and
where p = the kinnematic viscosity in centime ter-gram-second units, and T =
the temperature in degrees centigrade.

    The relationship  is only valid for temperatures between 0° and 50°C.  The
relation also assumes that the initial hydraulic conductivities are specified
at a temperature of 15°C.  The relationship could easily be reprogrammed to
cover a different temperature range or to relate hydraulic conductivities to
concentration.

DERIVATION OF A HEAT  TRANSPORT EQUATION

    The processes that control the transport of 'heat in an aquifer are in many
ways analogous to the processes that control the transport of mass in an
aquifer.  The partial differential equation describing the transport of mass
is well known, and its derivation is clearly explained in Konikow and Grove
(1977) and is rigorously derived in Bear (1972).  Rather than derive the heat
transport equation from first principles, which would in many respects repeat
Konikow and Grove (1977), we show how the mass transport equation can be
modified to apply to  the transport of heat.

    These assumptions are made in this derivation:

1) The aquifer is incompressible and chemically inert with respect to the
   fluid;

2) Fluid density is constant;

3) Hydraulic head is the only driving force; coupled processes, Onsanger
   relationships, and density driven convections are not considered;

4) Fluid flow is laminar; and

5) Divergence of velocity equals zero.

    In addition to these  five assumptions the generalized mass transport
equation
              3  / r,   3C  . _,_ 3  ,_.   3C ,      3C    _     3C   _
             ^'ijT^+T^'uT^-'i^- •-»--••   (c_23)


as derived by Bear (1972) and Konikow and Grove (1977), assumes the following
about  the processes that transport mass in an aquifer:  (1)  Velocity driven
convection, molecular diffusion, and mechanical dispersion  are  the only
transfer processes; and (2) no transfer of mass occurs within  the  solid  phase.

    In an aquifer the basic modes of heat transfer are (Lagarde  1965, quoted
in Bear  1972, p. 640) heat transfer through the solid phase  by  conduction,

                                   -100-

-------
heat transfer through the fluid phase by conduction,  heat, transfer  through the
fluid phase by convection, and heat transfer by dispersion.

    Each of the terms in Eq. (C-23) is discussed sequentially to  demonstrate
how a similar form for each term can be used to describe the transport  of heat
in an aquifer.

Molecular Diffusion

    Heat, unlike mass, is readily transmitted through the solid phase of most
porous media if a temperature gradient exists.  In the derivation of the first
term of Eq. (C-23), it was assumed that transport does not occur  in the solid
phase.  The coefficient of molecular diffusion in the mass transport equation
is defined on the basis of tortuosity (Bear 1972), a concept that is
applicable if transfer only takes place through the fluid phase.

    The transfer of heat by molecular diffusion (conduction) through the solid
phase and the fluid phase can be treated by assuming a simple parallel
conduction model in which conduction through the fluid phase occurs separately
but simultaneously with no interchange of heat between the two media.   The
real situation is more complex since heat is interchanged continuously  between
the two phases.  Experimental data (Houpert 1965, quoted in Bear  1972,  p. 646)
suggest that the simple model is sufficient, and the assumption of
simultaneousness is valid if the time period is greater than a few  minutes.

    Therefore, heat conduction can be described by

                     no   », -    3C  v  .  /,  x   9   f,     oC
                    "    (kf   ")
                                            x±       a^                (c_24)


where kfjj = coefficient of thermal molecular diffusion for the fluid phase,
ksjj = coefficient of thermal molecular diffusion for the solid phase, and  n  =
the porosity.

Dispersive Transfer of heat

    The form of the second term in Eq. (C-23) was derived from experimental
data for mass transport in porous media.  Green (1963) concludes on the basis
of experimental data that in the range of Darcian flow the influence of the
passage of heat through the solid phase will be insignificant and that the
dispersive transfer of mass and heat will be identical.  (Dispersive transfer
is a term used to describe convective transfer that occurs due to velocity
fluctuations in the fluid.)

Internal Generation of Heat

    neat will be generated by a moving fluid because of energy dissipation  by
viscous stresses.  The rate of dissipation, E, can be estimated from


                              E  = pqg
                                    -101-

-------
where g = the gravitational constant and p  =  fluid  density.

    A fluid flowing at a rate of 1  m/day,  with a gradient  of 1 m/m, will
dissipate energy at the rate of 10"'' 3 cal/m3  per day.   Therefore, heat
dissipation is considered to be negligible.

The Absorption of Heat by the Solid Phase

    The absorption of heat by the solid phase can be treated analogously to
the adsorption of mass by the solid phase  when the  process is characterized by
a linear adsorption isotherm.  The last term  of E.  (C-23)  can be modified
(Bear 1972, Pickens and Lennox 1977) to

                             3C  . ,.  . .   3C
                           n"3t + (1~n) Alt                             (C-25)

to treat the adsorption of mass by the solid  phase , where  A = the adsorption
distribution coefficient that describes the ratio of solute on the  solid phase
to solute in the fluid phase .

    If A is defined to be the ratio of heat in the  solid phase to heat  in  the
fluid phase, Eq. (C-25) can be used to describe the absorption of heat  by  the
solid phase in a porous medium.

    The heat transport equation can then be obtained by substituting  the terms
developed above into Eq. (C-23):
    By convention the concentration of heat is generally expressed  as a
temperature, where concentration equals temperature multiplied by the heat
capacity.  The heat capacities of a fluid and solid phase are generally not
the same.  Each term in Eq. (C-26) refers to heat transfer in either the solid
or the fluid phase; no term refers to a combined heat transfer.  Making the
appropriate substitutions, Eq. (C-26) can be written as:
                           - R -  «      -  >c   (1-) *   - ° •          (c-27)
where pCy, = heat capacity of the fluid, H/L3°C; and pCsoi = heat capacity of
the solid phase, H/L^OC.  The equation can be simplified by introducing the
following conventions.

1) %j = (1-n) pCgoikSij + npC^kfij,


                                    -102-

-------
    where K^j are the components of the thermal conductivity tensor for the
saturated media.  It is much simpler to determine experimentally the  combined
coefficient than the individual ones, which are a function not only of phase
composition but also of phase geometry.

2} The absorption distribution coefficient A = pCgoi/pC^.  This relationship
holds by definition.

3) The heat capacity of a whole is equal to the sum of the heat capacities of
its parts:  pCs = npCw + (1-n) pCsol,

    Equation (C-27) then reduces to the following form when the substitutions
are made which is in the same form as Eq.  (C-23):
SAMPLE PROBLEMS

Linear Heat Transport

    This example problem demonstrates the use of all linear elements and the
use of mixed higher order elements to solve a one-dimensional convective heat
transport problem.  The problem analyzed and the finite element grids used to
discretize the problem are show in Figure C-9.   The parameters used in the
problem are listed in Table C-2.  The input data and the program-generated
output for both sample simulations are listed in Figures C-10,  C-11, C-12, and
C-13.  The simulated temperatures are nearly identical for both the grid with
mixed elements and the one with all linear elements (Table C-3).  Simulated
temperatures differ by less than 0.5°C.

    In analogy to Ogata and Banks (1961), an analytical solution to the linear
convective heat transport problem is:

                        x-q't            XqpCw           x+q't
     T = T /Z  [erfc(	r^) + exp(-
                    Z(K't/pCg)                       Z(K't/pCs)

where q1 = qpCw/pCs, and K.' = Kfc + qaj pCw, in which the following boundary
conditions are used:

                              T(x,0) = 0, x > 0;
                           T(0,t) = T0, t >_ 0; and
                              T(»,t) =0, t >_ 0.

    The analytical solutions at t = 25, 50, 75 days are shown with the
simulated temperatures in Figure C-14.  The analytical solution at t = 50 days
is compared to the simulated temperatures at  t = 50 days in Table C-3.  In
addition, the simulated temperatures for the  all linear element grid with
nodes 21 and 22 specified at 15°C at t = 50 days are also presented in Table

                                    -103-

-------
1 3 57 91
1
2
3
4
5
(3 15 17 19 2
6
7
8
9
10
                   68      1012
                         14  16   18
20
22
                                                                    (a)
             36   8 10 14   16 18
                         22  24   26
29
31




1 '

<
>4 2
1
II 1
3
12
19
4
'20

5 '


'27 6


7



, (b)
2 57 9 13 15 17 21 23 25 28 30 32
t
*
1
Vio-c


V = . 1786m /day


T0=I5"C
(c)
Figure C-9.
Finite element grids used to discretize a linear heat
transport problem.  The problem is depicted in  (c).  The
grid with all linear elements is shown in (a), and the grid
with mixed higher order elements is shown in  (b).  Both
node numbers and element numbers are shown in (a) and (b).
                                  -104-

-------
         Convective  test  problem  rectangular  matrix
         Units used  are meters, days,  celsius,  calories

         10 22 4
         30001  1E13
         (lx,2G12.6)

         (lx,7G12.6)

         GROUP II
         11 2
         0 4 7 8  11  12  16  18  20 24  28
         4 0
         GROUP III
         .510  1E6
         10 0 1
         0 4 1
         1 100 2  100  21 105  22  105
         0 0 1
         GROUP IV
         1
         11111111
         1 0 1
         1110  55000  55000  700000 .1  .1
         GROUP V
         0 0
         0 0
         0 0
         GROUP VI
         0
         GROUP VII
         1
         1 3 0
         10
         1
         10 4 0
         15
         GROUP VIII
         2 0
         1 3 10  4
         GROUP IX
         00000000
         75 50 150 50 150  0  1 0
Figure C-10.  The input data used to model one-dimensional heat
              transport with all linear elements.
                             -105-

-------
     TABLE C-2.   PARAMETERS,  INITIAL CONDITIONS,  AND BOUNDAR1 CONDITIONS
    USED FOh THE HEAT-FLOW EQUATION IN ThE LINEAR HEAT TRANSPORT PROBLEM3


V

Kt
PCy

PCS

time step
Parameters
0. 1786 m/day

0.55fx 105 cal/m day °C
0.1 x 10? cal/m2 °C

0.7 x 106 cal/m2 °C

0.5 day
Initial
conditions
TA = 10°C
at all
nodes





Boundary
conditions
Dirichlet (convective flow)
boundaries at x = 0 and
x = 28
Temperature of incoming
fluid: 15°C

No flow at all other
boundaries

aThe water-flow equation was solved for head gradient of 0.1786 m/m with
 hydraisic conductivity and porosity set to unity to obtain the specified
 velocity.
                                    -106-

-------
                  CONVECTIVE  TEST  PROBLEM RECTANGULAR MATRIX
                  UNITS  USED  ARE METERS,  DAYS,  CELSIUS, CALORIES

                                                             04//07/78
                                                             17:01:09

    #OF NODES  22  #  OF  ELEMENTS  10  BANDWIDTH  4

    VELOCITIES ARE  BEING CALCULATED AT CURRENT TIME STEP    1

    THE X-SPACING IS

       0.0        4.0000     7.0000     8.0000     11.0000      12.0000      16.0000      18.0000
       20.0000     24.0000    28.0000

    THE Y-SPACING IS
        4.0000     0.0

    EQUATION   2 TIME  STEP  .500000           HEAT OR MASS CAPACITY COEFFICIENT   .100000+07

    PARAMETER  FACTORS--IN ORDER   1.00000  1.00000  1.00000   1.00000  1.00000  1.00000   1.00000   1.00000

§      MATERIAL  HOR PERM  VER PERM    STORAGE   HOR THERM COND    VERT THERM COND  SPECIFIC DISPERSION COEFS
I                                                                                     H E AT
                  1  1.000   1.000       .0000       .5500+05            .5500+05     .7000+06.1000+00.1000+00

    THE VALUES IN THE TYPE OF MATERIAL MATRIX ARE
       1.00000     1.00000    1.00000    1.00000    1.00000     1.00000     1.00000
       1.00000     1.00000    1.00000

     INFORMATION FOR MASS OR HEAT TRANSFER ACROSS A SPECIFIED  HEAD BOUNDARY
     ELEMENT NUMBER--BOUNDARY CODE—TEMPERATURE OR CONCENTRATION  OF INCOMING FLUID

     CONVECTIVE OUT  AND CONDUCTIVE BOUNDARY INFORMATION         ELEM  //--BOUNDARY  CODE —TRANSFER COEFFICIENT
          1  3   .000000
     Figure C-ll.   Program output  for  one-dimensional heat transport problem with linear elements.(continued)

-------
     WATER FLOW BOUNDARIES—ELEM « AND BOUNDARY CODE
       1  3       10   4
     POTENTIAL DISTRIBUTION AT TIME STEP
                                       .000000
      100.0000
       100.714
       101.250
       101.429
       101.964
       102.143
       102.857
       103.214
       103.571
       104.286
       105.000
            100.0000
             100.714
             101.250
             101.429
             101.964
             102.143
             102.857
             103.214
             103.571
             104.286
             105.000
     B. FLUX RECHARGE
     B. FLUX DISCHARGE
     CHANGE IN STORAGE
     QUANTITY PUMPED
     DIFFERENCE
                            CUMULATIVE MASS BLANCE
                                  .714288
                                  .714281
                                  .000000
                                  .000000
                                  .715256-05
RATES FOR THIS
    .714288
    .714281
    .000000
    .000000
    .715256-05
                                                                                STEP
O
oo
TIME STEP   .000000
FLOWS IN THE X DIRECTION
  .714281   .714284  .714285  .714286  .714287  .714286  .714287
  .714288   .714288  .714288

FLOWS IN THE Y DIRECTION
  .000000  .715256-06  .119209-06  .357628-06  .000000  .476837-06  .000000
  .000000 -.476837-06  .000000

THE DISPERSION ROUTINE IS BEING USED
                                               Figure  C-ll.   (continued)

-------
     TEMPERATURE DISTRIBUTION AT TIME STEP
                                        25.0000
10.0001
9.99935
10.0014
10.0074
9.99971
9.96143
9.90413
10.3286
11.4248
14.1568
15.3485
10.0002
9.99949
10.0016
10.0074
9.99992
9.96156
9.90422
10. 3286
11.4248
14.1569
15. 3486
l
»-•
s
CONDUCTIVE TRANSFER
CONVECTIVE TRANSFER--OUT
CONVECTIVE TRANSFER—IN
B. FLUX RECHARGE
B. FLUX DISCHARGE
CHANGE IN STORAGE
QUANTITY PUMPED
DIFFERENCE
CUMULATIVE MASS BLANCE
      .000000
      .178571+09
      .267858+09
      .000000
      .000000
      .892868+08
      .000000
      429.500
     TEMPERATURE DISTRIBUTION AT TIME STEP
       10.0037
       9.97586
       9.94439
       10.0029
       10.4781
       10.9284
       12.8394
       14.0114
       14.5750
       15.1468
       14.9093
           10.0039
           9.97601
           9.94460
           10.0029
           10.4783
           10.9285
           12.8395
           14.0115
           14.5750
           15.1468
           14.9093
                                        50.0000
RATES FOR THIS TIME STEP
        .000000
        .357147+07
        .535716+07
        .000000
        .000000
        .178568+07
        .000000
        8.37500
                                   Figure C-ll.  (continued)

-------
    CONDUCTIVE TRANSFER
    CONVECTIVE TRANSFER-OUT
    CONVECTIVE TRANSFER-IN
    B.  FLUX RECHARGE
    B.  FLUX DISCHARGE
    CHANGE IN STORAGE
    QUANTITY PUMPED
    DIFFERENCE
CUMULATIVE MASS BLANCE
        .000000
        .357165+09
        .535716+09
        .000000
        .000000
        .178550+09
        .000000
        896.781
                                                        RATES FOR
          THIS TIME STEP
          .000000
          .357283+07
          .535716+07
          .000000
          .000000
          .178431+07
          .000000
          14.2500
o
    TEMPERATURE DISTRIBUTION AT TIME STEP
      10.0302   10.0304
      10.5080   10.5082
      11.6597   11.6599
      12.1306   12.1306
      13.4834   13.4837
      13.8607   13.8608
      14.9332   14.9333
      14.9278   14.9278
      15.0309   15.0309
      15.0045   15.0045
      14.9912   14.9913
                  75.0000
    CONDUCTIVE TRANSFER
    CONVECTIVE TRANSFER-OUT
    CONVECTIVE TRANSFER-IN
    B. FLUX RECHARGE
    B. FLUX DISCHARGE
    CHANGE IN STORAGE
    QUANTITY PUMPED
    DIFFERENCE
CUMULATIVE MASS BLANCE
        .000000
        .535367+09
        .803574+09
        .000000
        .000000
        .268205+09
        .000000
        1372.56
RATES FOR THIS TIME STEP
          .000000
          .358047+07
          .535716+07
          .000000
          .000000
          .177668+07
          .000000
          7.93750
   Figure C-ll.  Program output for one-dimensional heat transport problem with linear elements,

-------
     CONVECTIVE TEST PROGRAM
     UNITS USED ARE METERS, DAYS,  CELCIUS,  CALORIES

     7  32  12
     30021 1E13
     (lx,2G12.6/3G12.6/2G12.6/2G12.6/4G12.6/2G12.6/2G12.6/4G12.6/2G12.6/2G12.6/
     3G12.6/2G12.6/2G12.6)
     (lx,7G12.6)

     GROUP II

     104
     200
     344
     442
     540
     654
     750
     8  7. 4
     970
     10   8  4
     11   8  3
     12   8  1
     13   8  0
     14   9  4
     15   9  0
     16   11  4
     17   11  0
     18   12  4
     19   12  3
     20   12  1
     21   12  0
     22   16  4
     23   16  0
     24   18  4
     25   18  0
     26   20  4
     27   20  2
     28   20  0
     29   24  4
     30   24  0
     31   28  4
     32   28  0
     1  2 5  4*  3  1  2
     2  5 7*  9* 13  12*  11*  10  8*  6*  3  4*  5
     3  13  15*  17*  21  20*  19* 18  16*  14*  10  11*  12*  13
     4  21  23  22 18  19*  20*  21
     5  23  25*  28  27*  26  24* 22  23
     6  28  30  29 26  27*  28
     7  30  32  31 29  30


                                                                (continued)


Figure C-12.  The data deck used  to model one-dimensional heat transport
               with mixed elements.
                                       -Ill-

-------
                  GROUP III

                  .5  1  0  1E6
                  10  0  1
                  041
                  1  100  2  100  31  105  32  105
                  001

                  GROUP IV

                  1
                  11111111
                  101
                  1110  55000  55000  700000  .1  .1

                  GROUP V

                  0  0
                  0  0
                  0  0

                  GROUP VI

                  0

                  GROUP VII

                  1
                  130
                  10
                  1
                  7  A 0
                  15
                  GROUP VIII

                  2  0
                  1374

                  GROUP IX

                  00000000
                  75 50  150  50  150  0  1  0
Figure C-12.   The data deck used  to model one-dimensional heat transport
               with mixed elements.
                                      -112-

-------
                 CONVECTIVE  TEST PROGRAM
                 UNITS  USED  ARE METERS, DAYS, CELCIUS, CALORIES
                                                                               04/07/76
                                                                               1,7:00:46
   t OR NODES   32  // OF ELEMENTS    7 BANDWIDTH   12

   VELOCITIES ARE BEING CALCULATED AT CURRENT TIME STEP      1

   EQUATION 2 TIME STEP   .500000            HEAT OR MASS CAPACITY COEFFICIENT  .100000+07

   PARAMETER FACTORS—IN ORDER   1.00000     1.00000   1.00000   1.00000   1.00000   1.00000   1.00000    1.00000

      MATERIAL  HOR PERM    VER PERM     STORAGE     HOR THERN COND
                                            VERT THERM COND   SPECIFIC
                                                              HEAT
                    1.000
   1.000
.0000
.5500+05
                                                                          .5500+05
                                                               .7000+06
   THE VALUES IN THE TYPE  OF MATERIAL MATRIX ARE
      1.00000   1.00000    1.00000   1.00000   1.00000   1.00000 1.00000

   INFORMATION FOR MASS  OR HEAT TRANSFER ACROSS A SPECIFIED HEAD BOUNDARY
   ELEMENT  UNSER—BOUNDARY CODE—TEMPERATURE OR CONCENTRATION OF INCOMING FLUID
        7   4   15.000
   CONVECTIVE OUT AND CONDUCTIVE BOUNDARY INFORMATION   ELEM //—BOUNDARY CODE—TRANSFER COEFFICIENT
        1   3       .000000

   WATER FLOW BOUNDARIES—ELEM # AND BOUNDARY CODE
     13        74
   POTENTIAL DISTRIBUTION AT TIME  STEP
     100.0000    100.0000
     100.714     100.714
     100.893     100.893
     101.253     101.253
     101.429     101.429
     101.657     101.657
     101.964     101.964
     102.143     102.143
     102.857     102.857
               .000000
100.714


101.429   101.429


102.143   102.143
                                                            DISPERSION
                                                            COEFS
                                                            .1000+00  .1000+00
                                                                                                  (continued)
Figure C-13.   Program output for the  one-dimensional heat  transport problem with mixed elements.

-------
  103.214
  103.571
  104.236
  103.000
103.214
103.571
104.236
105.000
103.571
B. FLUX RECHARGE
B. FLUX DISCHARGE
CHANGE IN STORAGE
QUANTITY PUMPED
DIFFERENCE

TIME STEP     .000000
FLOWS IN THE X DIRECTION
   .714262      1.42996

FLOWS IN THE Y DIRECTION
  -.476837-56   .558137-05
                 CUMULATIVE MASS BLANCE
                       .714289
                       .714202
                       .000000
                       .000000
                       .267029-04
                                    RATES FOR THIS TIME STEP
                                         .714289
                                         .714262
                                         .000000
                                         .000000
                                         .267029-04
              .1,74388
                   .915181
.714290
.714292
               .540157-05   -.548363-05
                                .00-000  -.953674-06
                                                                 .714289
                       .000000
THE DISPERSION ROUTINE IS BEING USED

TEMPERATURE DISTRIBUTION AT TIME STEP     25.0000
                        10.0002
                        10.0058   10.0076
                        9.97589   9.96879
                        11.4717
10.0005
10.0001
9.99925
10.0025
10.007?
10.0073
9.99686
9.90961
9.7985?
10.3231
11.4719
14.1362
13.3683
10.0006
9.99989
9.99977
10.0026
10.0058
10.0079
9.9551
9.97127
9.8972?
10.373?
11.4715
14.1365
15.3502
 CONDUCTIVE TRANSFER
 CONVECTIVE TRANSFER-OUT
 CONVECTIVE TRANSFER-IN
 B.  FLUX  RECHARGE
 B.  FLUX DISCHARGE
 CHANGE  IN STORAGE
 QUANTITY PUMPED
 DIFFERENCE
              CUMULATIVE MASS BLANCE
                   .000000
                   .178569-K)9
                   .267858+09
                   .000000
                   .000000
                   .589193+08
                   .000000
                   370063.
                                RATES FOR THIS TIME STEP
                                         .000000
                                         .357150+07
                                         .535716+07
                                         .000000
                                         .000000
                                         .180833+07
                                         .000000
                                         -22663.4
                                        Figure C-13.   (continued)

-------
TEMPERATURE DISTRIBUTION AT TIME STEP
                        9.98017
                        9.99911 10.0018
                        10.8995 10.8954
                        14.5313
           50.0000
10.0049
9.96511
9.96235
9.96261
10.0013
17.1391
17.547?
17.773
12.7773
13.9537
14.5316
15.1661
14.8985
10.0050
9.93723
9.96254
9.95263
9.95-12
17.1396
17.5474
17.3997
12.9877
13.9556
14.5311
15.1663
14.5984
CONDUCTIVE TRANSFER
CONVECTIVE TRANSFER-OUT
CONVECTIVE TRANSFER-IN
B. FLUX RECHARGE
B. FLUX DISCHARGE
CHANGE IN STORAGE
QUANTITY PUMPED
Ul
1














DIFFERENCE
TEMPERATURE
10.8259
10.5035
10.6779
11.6724
12.1127
12.5753
13.4747
13.7692
14.9299
14.9297
15.0453
15.771?
14.9917

DISTRIBUTION
10.0261
17.5926
17.9895
11.6724
12.1154
12.5766
13.4742
13.5693
14.9374
14.9296
15.044?
15.7722
14.9917
.331
AT TIME STEP

10.5035


12.1154


13.8603


15.0451


CUMULATIVE MASS BLANCE
   .000000
   .357160-H)9
   .535716+09
   .000000
   .000000
   .17257+09
   .000000
                                         75.0000
                                      12.1135
                                      13.8675
RATES FOR THIS TIME STEP
     .000000
     .357313+07
     .535716+07
     .000000
     .000000
     .179040+07
     .000000
    -6372.50
                                       Figure C-13.   (continued)

-------
          CONDUCTIVE TRANSFER
          COSVECTIVE TRANSFER-OUT
          COHVECTIVE TRANSFER-IN
          B. FLUX RECHARGE
          B. FLUX DISCHARGE
          CHANGE IN STORAGE
          QUANNTITY PUMPED
          DIFFERENCE
          EXGT
CUMULATIVE MASS BLANCE
      .000000
      .635375409
      .000000
      .000000
      .000000
      .567842409
      .000000
      356718.
RATES FOR THIS TIME STEP
       .000000
       .357859+07
       .000000
       .000000
       .000000
       .183896+07
       .000000
      -60682.3
Figure C-13.   Program output  for the  one-dimensional  heat transport problem with mixed elements.

-------
    TABLE C-3.  ANALYTICAL SOLUTION AT t= 50 DAYS FOR THE PROBLEM POSED IN
    FIGURE C-9 AND FINITE ELEMENT NUMERICAL SOLUTIONS AT t = 50 DAIS FOR
        TtoO GRID CONFIGURATIONS AND TWO T1PES OF BOUNDAfll CONDITIONS

-
x = 0
x = 4
x = 8
x = 10
x = 12
x = 16
x = 20
x = 24
Analytical
15. Oa
14.982
14.649
14.017
12.962
10.787
10.062
10.002
Linear
elements (a)
14.901
15.147
14.575
14.011
12.840
10.929
10.003
9.976
Mixed
elements
14.899
15.166
14.531
13.945
12.867
10.900
9.999
9.976
Linear elements
with nodes 21 & 22
specified at 15°C
15.0
15.061
14.484
13.981
13.036
10.724
10.095
9.763
.
aAll values listed are temperatures in degrees centigrade,  x = 0 is at the
 right in Figure C-10.
                                    -117-

-------
                                                                 0
                       8       12       16      20
                      DISTANCE  (METERS)
                                          24
28
Figure C-14.
Analytical solutions (solid  lines) and numerical solutions
(dots)  for the linear heat transport problem for t = 25, 50
and 75  days.
                                 -118-

-------
C-3.   Specification of the boundary temperature,  rather than use of a
convective boundary, constrains the solution from oscillating as much near the
origin and is the exact boundary condition used by Ogata and Banks (1961).

The Mohawk River

    This problem demonstrates the use of the program to solve an areal problem
to investigate the effect of stream infiltration on ground-water temperatures
in an alluvial aquifer along the Mohawk River near Schenectady, N.Y.  (Figure
C-15).

    The flood plain of the Mohawk River in the vicinity of Lock and Dam 8,
about 2 miles west of the city of Schenectady, is underlain by more than  100
ft of unconsolidated deposits.  From the surface downward the deposits consist
of 30 ft of flood-plain alluvium, 20-100 ft of sandy gravel and sand, and,
immediately above the bedrock, a layer of glacial till 25-50 ft thick (Figure
C-16).  The sandy gravel and sand deposit is tapped by the well fields of the
city of Schenectady and torn of Rotterdam,  kvinslow (1962) monitored
temperatures in this aquifer and concluded that most of the water pumped  from
the well fields originated as infiltration from the Mohawk River.  This model
was programmed to determine if the temperature patterns observed by Winslow
could be simulated.

    The grid used to model the aquifer is shown in Figure C-17.  The boundary
conditons and parameters used in the simulation are listed in Table C-4.  All
water flow was assumed to occur within the principal aquifer, and  the
principal aquifer was assumed to be thermally insulated from the overlying
flood-plain deposits and the underlying glacial tills.  (This assumption
greatly simplified  the data input, but it could be relaxed if a better
approximation is desired.)

    The temperature of the river water was changed at each time step.   River
temperatures for a  1-yr period were read  in at the beginning of program
execution by a call to EwTRl LAKE and were then altered at each time  step by a
call to fiNTRi BOUND.  ENTRi CHANG was called  at each time step to  compute heat
flow out of the system with the pumped water.  Heat flow at  the wells for each
time step was set equal to the temperature at the wells multiplied by the heat
capacity and the rate of pumpage.

    The data and the program  output  for a 10-day simulation, with pumpage
rates only  10* of actual rates, are  listed in Figure C-18 and Figure C-19.   If
actual pumpage rates are used, a time step of 0.1 day must  be used to insure
solution  stability, unless a  corrector is applied near  the wells to force
stability.
                                    -119-

-------
   TABLE C-4.   PARAMETERS,  INITIAL  CONDITIONS,  AND BOUNDARY CONDITIONS USED

   FOR THE SIMULATION OF TEMPERATURES IN TI£ MOHAWK  RIVER ALLUVIAL AQUIFER
Hater flow


     Boundary conditions


          Mohawk River


          All other boundaries


     Parameters (principle aquifer)



          K11 - K22


     Wells—discharge rates


          Node 70


          Nodes 130, 131, 145, 146
   Specified head of 100 fta


   No flow





   100,000 gal/day ft2





   2,000,000 gal/day


   4,500,000 gal/day at each node
Heat flow


     Initial conditions


     Boundary conditions


          River


          All other boundaries


     Parameters

          vt    rt
          *11 = K22
     Time step
    I±  =  10°C





    Connective  flux  in


    No  flow





1.3 x 101* cal/day ft°C


  2.1 x 10* cal/ft3  oc


  10 ft


  1 ft


    1 day
aUnits of feet, gallons, calories, and days were used in this simulation.
                                    -120-

-------
                                              •ROTTERDAM
                                               WELL FIELD
                                                    SCHENECTADY
                                                    WELL FIELD


                                                            A'
0   500  1000 FEET
Figure  C-15.  Areal view of the Mohawk River Valley showing location of the
             Schenectady and Rotterdam well fields (Winslow 1962).
     MOHAWK
      RIVER
     0
  g '5
    50
 u
 u
 LL
   150
      0
                                                                    A1
         MOHAWK
          RIVER
                                      FLOOD PLAIN DEPOSITS
               SANDY GRAVEL AND SAND
               GLACIAL  TILL
     SHALES AND SILTSTONE
    	1	L_
          IOOO
                                   2000
                                  FEET
                                       3000
4000
Figure C-16.
  Cross-sectional view of the Mohawk River alluvial aquifer along
  section A-A1  of Figure C-15.
                                -121-

-------
     400

     600
                           MOHAWK  RIVER
  SI400
    2200
2









197
3

/







198
4

V
;






199
5


J^»
X >





200
6



t
i





7





S A




8





^




9





1




10





VEL




I
WE




_ Fl



206
12
_L




ELD



207
13









208
14









209
15









210
         0
1600           3200
       FEET
4800
Figure C-17.  Grid used to discretize the Mohawk River problem.  (The
            node numbers in the first and last rows are listed.)
                              -122-

-------
   MOHAWK RIVER PROBLEM
   UNITS  ARE GALLONS DAYS FEET
 182   210 17
 30001  1E16
 (lx,15F6.2)
 GROUP II
 14  15
 400  450  500  600  800  1000  1200  1400  1600  1800  2000   2200  2400  2600
 5600  5200  4800  4480  4000  3600  3200  2800  2400  2000  1600  1200 800  400
 GROUP III
 111  3785.68
 10 0 1
 0-11
  100 100 100 100  100 100 100 100 100 100 100 100 100 100 100
                              000000000000000
000000000
oooo-ooooo
000000000
000000000
000000000
000000000
999
001
GROUP IV
4
111111111
0-11
111111111
222222222
222222222
422222222
444222222
444422222
444444222
444444222
444444422
444444422
444444443
444444443
444444443
1 100000 1000000
2 100000 1000000
0
0
0
0
0
0
1
2
2
2
2
2
2
2
2
2
3
3
3
0
0
000
000
000
000
000
000
111
222
222
222
222
222
222
222
222
222
333
333
333
13000
13000
0000000000000
0000000000000
0000000000000
0000000000000
0000000000000
0000000000000
1
2
2
2
2
2
2
2
2
2
3
3
3 999
13000 21000 10 1
13000 21000 10 1
0000
0000
0000
0000
0000
0000
  3 500 500 0 13000 13000 21000  10 1
  4 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001
  GROUP VI
  0 0
  5 5
  130 -4.5E5 131 -4.5E5 145 -4.5E5 146 -4.5E5 70 -2E5
  130 1 131 1 145 1 146 1 70 1
                                                                (continued)
Figure C-18.   The  data deck used  to model heat  flow in  the Mohawk
                River alluvial aquifer.
                                      -123-

-------
       0 0
       GROUP VI  1
       1
       1 3
       0-11
       1 1 1 1 1 10 20 30 40 50  50 50  50  50 50
       1 1 1 1 1 10 20 30 40 50  50 50  50  50 50
       1 1 1 1 1 10 20 30 40 50  50 50  50  50 50
       1 1 1 1 1 10 20 30 40 50  50 50  50  50 50
       1 1 1 1 1 10 20 30 40 50  50 50  50  50 50
       99 99 99 20 20 25 35  45 50 55 55 55 55 55 55
       99 99 99 99 40 45 50  55 60 60 60 60 60 60 60
       99 99 99 99 99 99 65  65 65 60 60 60 60 60 60
       99 99 99 99 99  99 70 70  70 70  70  70 70  70  70
       75 75 75 75 75 75 75  75 75 75 75 75 75 75 75
       808080808080808080808080808080
       80 80 80 80 80 80 80  80 80 80 80 80 80 80' 80
       808080808080808080808080808080
       80 80 80 80 80 80 80  80 80 80 80 80 80 80 80 999
       100 0 1
       GROUP VII
       14
       1 3
       8 3
3000 2 3
3000 9 3
             3000 3 3 3000 4 3  3000 5 3 3000 6 3 3000 7 3 3000
             3000 10 3 3000 11 3  3000  12 3 3000 13 3 3000 14 3 3000
000000000000000
14
132333435363738393 10 3 11 3123 13 3 14 3
000000000000000
GROUP VIII
14 0
                              3 9 3 10 3 11 3 12 3 13 3 14 3
132333
GROUP IX
110000
0 220 220 1
277 860
1.0000
2.0000
3.0000
4.0000
5.0000
6.0000
7.0000
8.0000
9.0000
10.0000
10 10 10 10
435363738


220 220 1 220 220

29.5000
28.5000
27.5000
27.4600
27.4100
27.3700
27.3300
27.2900
27.2400
27.2000
10 1 1 1
                                     111
                                         41.0000
                                         35.9000
                                         30.8000
                                         30.5000
                                         30.2000
                                         29.9000
                                         29.6000
                                         29.3000
                                         29.0000
                                         28.7000
Figure C-18.
     The data  deck used to model heat  flow  in the  Mohawk
     River alluvial  aquifier.
                                      -124-

-------
S3
Ui
I
                                 MOHAWK RIVER PROBLEM
                                 UNITS ARE GALLONS DAYS FEET
                                                                           04/07/78
                                                                           17:01:49
        # OF NODES 210  # OF ELEMENTS  182  BANDWIDTH 17

        VELOCITIES ARE BEING CALCULATED AT CURRENT TIME STEP 1

        THE X-SPACING IS
          400.0000      450.0000      500.0000     600.0000      800.0000     1000.0000     1200.0000     1400.0000
         1600.0000     1800.0000     2000.0000     2200.0000    2400.0000     2600.0000

        THE Y-SPACING IS:
         5600.0000     5200.0000     4800.0000     4480.0000    4000.0000     3600.0000 .    3200.0000     2800.0000
         2400.0000     2000.0000     1600.0000     1200.0000     800.0000     400.0000         0.0

        EQUATION 2 TIME STEP 1.00000              HEAT OR MASS CAPACITY  COEFFICIENT  3785.68

        PARAMETER FACTORS- IN ORDER     1.00000   1.00000   1.00000   1.00000    1.00000   1.00000    1.00000    1.00000

              MATERIAL   HOR PERM    VER PERM    STORAGE    HOR THERM COND    VERT THERM COND

                         1   .1000+06   .1000+06   .0000        .1300+05           .1300+05
SPECIFIC DISPERSION COEFS
HEAT
 .2100+05  10.00   1.000


THE VALUES IN
1.00
2.00
2.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
1.00
2.00
2.00
2.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
2
3
4
. 1000+06
500.0
.1000-02
. 1000+06
500.0
. 1000-02
0000
0000
1000-02
.1300+05
.1300+05
. 1000-02
. 1300+05
. 1300+05
. 1000-02
THE TYPE OF MATERIAL ARE
1.00
2.00
2.00
2.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
1.00 1.00
2.00 2.00
2.00 2.00
2.00 2.00
2.00 2.00
4.00 2.00
4.00 4.00
4.00 4.00
4.00 4.00
4.00 4.00
4.00 4.00
4.00 4.00
4.00 4.00
1.00
2.00
2.00
2.00
2.00
2.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
1.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
4.00
4.00
4.00
4.00
4.00
1.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
4.00
4.00
4.00
1.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.0C
2.00
3.00
3.00
3.00
Figure C-19. Program output for Mohawk
1.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
3.00
3.00
3.00
River
1.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
3.00
3.00
3.00
1.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
3.00
3.00
3.00
1.00 1.00
2.00 2.00
2.00 2.00
2.00 2.00
2.00 2.00
2.00 2.00
2.00 2.00
2.00 2.00
2.00 2.00
2.00 2.00
3.00 3.00
3.00 3.- 00
3.00 3.00
problem. (<
                                                                                                  .2100+05   10.00    1.000

                                                                                                  .2100+05   10.00    1.000

                                                                                                  .1000-02   .1000-02.1000-02
                                                                                          (continued)

-------
Ni
     POINT SOURCES  OF WATER—NODE  # AND AMOUNT
        130   -450000.       131    -450000.
     POINT SOURCES OF HEAT
        130   1.00000
                       131    1.00000
                                           145   -450000.      146    -450000.      70    -200000.
145   1.00000       146    1.00000       70    1.00000
     ELEVATION OF THE AQUIFER BOTTOM
1.00 1.00
1.00 1.00
1.00 1.00
1.00 1.00
1.00 1.00
99.00 99.00
99.00 99.00
99.00 99.00
99.00 99.00
75.00 75.00
80.00 80.00
80.00 80.00
80.00 80.00
80.00 80.00
1.00
1.00
1.00
1.00
1.00
99.00
99.00
99.00
99.00
75.00
80.00
80.00
80.00
80.00
1.00
1.00
1.00
1.00
1.00
20.00
99.00
99.00
99.00
75.00
80.00
80.00
80.00
80.00
1.00
. 1.00
1.00
1.00
1.00
20.00
40.00
99.00
99.00
75.00
80.00
80.00
80.00
80.00
10.00
10.00
10.00
10.00
10.00
25.00
45.00
99.00
99.00
75.00
80.00
80.00
80.00
80.00
20.00
20.00
20.00
20.00
20.00
35.00
50.00
65.00
70.00
75.00
80.00
80.00
80.00
80.00
30.00
30.00
30.00
30.00
30.00
45.00
55.00
65.00
70.00
75.00
80.00
80.00
80.00
80.00
40.00
40.00
40.00
40.00
40.00
50.00
60.00
65.00
70.00
75.00
80.00
80.00
80.00
80.00
50.00
50.00
50.00
50.00
50.00
55.00
60.00
65.00
70.00
75.00
80.00
80.00
80.00
80.00
50.00
50.00
50.00
50.00
50.00
55.00
60.00
60.00
70.00
75.00
80.00
80.00
80.00
80.00
50.00
50.00
50.00
50.00
50.00
55.00
60.00
60.00
70.00
75.00
80.00
80.00
80.00
80.00
50.00
50.00
50.00
50.00
50.00
55.00
60.00
60.00
70.00
75.00
80.00
80.00
80.00
80.00
50.00
50.00
50.00
50.00
50.00
55.00
60.00
60.00
70.00
75.00
80.00
80.00
80.00
80.00
50.00
50.00
50.00
50.00
50.00
55.00
60.00
60.00
70.00
75.00
80.00
80.00
80.00
80.00
INFORMATION FOR MASS OR HEAT TRANSFER ACROSS  A SPECIFIED  HEAD  BOUNDARY
ELEMENT NUMBER—BOUNDARY CODE—TEMPERATURE OR CONCENTRATION OF INCOMING FLUID
  1  3   .00000       2  3     .00000       3  3     .00000       4   3     .00000
  6  3   .00000       7  3     .00000       8  3     .00000       9   3     .00000
 11  3   .00000      12  3     .00000      13  3     .00000       14   3     .00000
                                                                                         5   3     .00000
                                                                                        10   3     .00000
     CONVENTIVE OUT AND CONDUCTIVE BOUNDARY INFORMATION
       1  3   3000.00      2  3     3000.00      3  3
       6  3   3000.00      7  3     3000.00      8  3
      11  3   3000.00     12  3     3000.00     13  3

     WATER FLOW BOUNDARIES--ELEM 0 AND BOUNDARY CODE
       13      23     3343     5363
      10  3     11  3    12  3   13  3    14  3
                                                          ELEM  //--BOUNDARY CODE—TRANSFER COEFFICIENT
                                                    3000.00      4   3     3000.00     5   3    3000.00
                                                    3000.00      9   3     3000.00    10   3    3000.00
                                                    3000.00      14   3     3000.00
                                                       7   3
                     8  3
9  3
                                                                                  10  3
                                Figure  C-19.   (continued)

-------
     ITERATIONS  NEEDED FOR CONVERGENCE
No
vj
I
POTENTIAL DISTRIBUTION AT TIME STEP .000000
100.00 100.00 100.00
100.00 100.00 100.00
100.00 100.00 100.00
100.00 100.00 100.00
100.00 100.00 100.00
100.00 100.00 100.00
99.97 99.97 99.98
99.94 99.94 99.94
99.91 99.91 99.91
99.90 99.90 99.90
99.90 99.90 99.90
99.90 99.90 99.90
99.90 99.90 99.90
99.90 99.90 99.90
100.00
100.00
100.00
100.00
100.00
100.00
99.99
99.96
99.91
99.90
99.90
99.90
99.89
99.89
100.00 100.
100.00 100.
100.00 100.
100.00 100.
100.00 99.
99.99 99.
99.99 99.
99.98 99.
99.91 99.
99.89 99.
99.89 99.
99.89 99.
99.89 99.
99.89 99.
00
00
00
00
99
99
99
97
91
89
88
86
87
87
100.00
100.00
100.00
99.99
99.98
99.97
99.96
99.94
99.93
99.89
99.87
99.85
99.84
99.84
100.00 100.00 100.00
100.00
99.99
99.98
99.97
99.95
99.93
99.91
99.88
99.84
99.83
99.81
99.79
99.79
CUMULATIVE MASS BLANCE
B. FLUX RECHARGE
B. FLUX DISCHARGE
CHANGE IN STORAGE
QUANTITY PUMPED
DIFFERENCE



-

.200989+07
.000000
.000000
.200000+07
9886.42















99.99 99.99
99.99 99.98
99.97 99.96
99.95 99.92
99.92 99.89
99.89 99.84
99.85 99.78
99.82 99.68
99.79 99.63
99.77 99.65
99.72 99.69
99.70 99.70
99.70 99.70
RATES FOR THIS
.200989+07
.000000
.000000
-.200000+07
9886.42
100.00 100.00
99.99
99.98
99.97
99.93
99.89
99.84
99.78
99.68
99.63
99.65
99.70
99.72
99.72
TIME STEP





99.99
99.99
99.97
99.94
99.91
99.88
99.84
99.82
99.80
99.78
99.77
99.77
99.77






100.00 100.00
99.99 100.00
99.99
99.98
99.96
99.93
99.91
99.89
99.88
99.87
99.86
99.84
99.83
99.83






99.99
99.98
99.97
99.95
99.93
99.92
99.91
99.90
99.90
99.88
99.87
99.87






100.00
100.00
99.99
99.98
99.97
99.95
99.94
99.93
99.92
99.91
99.91
99.90
99.89
99.88






THE DISPERSION ROUTINE IS BEING USED
NODE # AND TEMP AT INF.
9 15.04 10 15.04
NODE # AND TEMP AT INF.
9 15.52 10 15.52
NODE // AND TEMP AT INF.
9 16.00 10 16.00
NODE // AND TEMP AT INF.
9 16.48 10 16.48
NODE // AND TEMP AT INF.
9 16.96 10 16.96
NODE # AND TEMP AT INF.
9 17.44 10 17.44
NODE # AND TEMP AT INF.
9 17.92 10 17.92
1 15.04
11 15.04
1 15.52
11 15.52
1 16.00
11 16.00
1 16.48
11 16.48
1 16.96
11 16.96
1 17.44
11 17.44
1 17.92
11 17.92
2 15.04
12 15.04
2 15.52
12 15.52
2 16.00
12 16.00
2 16.48
12 16.48
2 16.96
12 16.96
2 17.44
12 17.44
2 17.92
12 17.92
3
13
3
13
3
13
3
13
3
13
3
13
3
13
15.04
15.04
15.52
15.52
16.00
16.00
16.48
16.48
16.96
16.96
17.44
17.44
17.92
17.92
4 15.04
14 15.04
4 15.52
14 15.52
4 16.00
14 16.00
4 16.48
14 16.48
4 16.96
14 16.96
4 17.44
14 17.44
4 17.92
14 17.92
5 15.04 6

5 15.52 6

5 16.00 6

5 16.48 6

5 16.96 6

5 17.44 6

5 17.92 6

15.04 7

15.52 7

16.00 7

16.48 7

16.96 7

17.44 7

17.92 7

15.04

15.52

16.00

16.48

16.96

17.44

17.92

6 15.04

8 15.52

8 16.00

8 16.48

8 16.96

8 17.44

8 17.92





























                                             Figure  C-19.   (continued)

-------
NODE I AND TEMP Af  INF.  1  18.40   2  18.40   3  18.40   4   18.40  5  18.40  6  18.40  7  18.40  8  18.40
9
NODE t
9
NODE 1
9
18.40 10 18.40 11
AND TEMP AT INF. 1
18.88 10 18.88 11
AND TEMP AT INF. 1
19.36 10 19.36 11
TEMPERATURE DISTRIBUTION AT
10.69
9.82
10.04
9.98
10.01
10.00
10.01
10.01
10.00
10.00
10.00
10.00
10.00
^ 10.00
N>
oo
10.70
9.82
10.04
9.99
10.00
10.00
10.01
10.01
10.00
10.00
10.00
10.00
10.00
10.00


10.76
9.81
10.04
9.99
10.00
10.00
10.01
10.02
10.00
10.00
10.00
10.00
10.00
10.00


10.91
9.77
10.04
9.99
10.00
10.00
10.00
10.02
10.00
10.00
10.00
10.00
10.00
10.00


18.40
18.88
18.88
19.36
19.36
12 18.40 13 18.40
2 18.88 3 18.88
12 18.88 13 18.88
2 19.36 3 19.36
12 19.36 13 19.36
TIME STEP
11.38
9.67
10.06
9.99
10.00
10.00
10.00
10.01
10.00
10.00
10.00
10.00
10.00
10.00


12.20
9.52
10.07
9.99
10.00
10.00
10.00
10.01
10.00
9.99
10.00
10.00
10.00
10.00


10.000
13.50
9.39
10.07
9.99
10.00
10,00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00



14.99
9.46
10.02
10.00
10.00
10.00
10.00
10.00
10.00
9.99
10.00
10.00
10.00
10.00


14 18.40
4 18.88 5
14 18.88
4 19.36 5
14 19.36

16.49
9.93
9.91
10.02
10.00
10.00
10.00
10.00
10.02
10.04
10.00
10.00
10.00
10.00



17.48
10.67
9.84
10.03
9.99
10.00
10.00
10.02
9.98
9.92
10.03
9.99
10.00
10.00


18.88 6 18
19.36 6 19

17.25
10.49
9.86
10.02
10.00
10.00
10.00
10.02
9.98
9.92
10.03
9.99
10.00
10.00



16.63
10.03
9.90
10.02
10.00
10.00
10.00
10.00
10.01
10.03
10.00
10.00
10.00
10.00


.88 7
.36 7

15.86
9.69
9.96
10.01
10.00
10.00
10.00
10.00
10.00
9.99
10.00
10.00
10.00
10.00


18.88
19.36

15.15
9.52
10.01
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00


8 18.88
8 19.36

14.87
9.48
10.02
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00


B. FLUX DISCHARGE
CHANGE IN STORAGE
QUANTITY PUMPED
DIFFERENCE

CONDUCTIVE TRANSFER
CONVECTIVE TRANSFER—OUT
CONVECTIVE TRANSFER—IN
B. FLUX RECHARGE
  .000000
  .620931+12
 -.755490+12
 -.471091+07
CUMULATIVE MASS  BLANCE
  .676988+11
  .000000
  .130872+13
  .000000
 .000000
 .794988+11
-.754061+11
-502208.
RATES FOR THIS TIME STEP
 .759752+10
 .000000
 .147307+12
 .000000
                          Figure C-19.   Program output for Mohawk River problem.

-------
The Columbia Generating Station Site

    This example is  representative of the application described  in  section 5
of this report  in which temperatures and water flows in the alluvial aquifer
along the Wisconsin  River were simulated after construction of a cooling lake
on the river flood plain.  Temperatures and water flows were simulated  in an
aquifer cross section showi in the schematic diagram in figure C-20, and a
finite element  grid  was used to discretize the cross section (figure C-21).
The parameters, boundary conditions, and initial conditions used in the
simulation are  listed in Table C-5.

    The input data and the program output for a 9-day simulation are listed in
Figure C-22 and figure C-25.  ENTfli BOUND was called at each time step  to
change the temperature of the water in the cooling lake and the  air
temperature.
                                     -129-

-------
     TABLE C-5.  PARAMETERS,  INITIAL CONDITIONS,  AND BOUNDAR1 CONDITIONS
               USED IN THE COLUMBIA  GENERATING STATION  PROBLEM
hater flow

     Boundary conditions

          River

          Marsh

          Lake

          All other boundaries

     Parameters

          K11» K22
Heat flow

     Initial conditions

     Boundary conditions

          Lake

          Marsh, River


          All other boundaries

     Parameters

          K11» K22> Pcs
          Time step
Specified head at 778 fta

Specified head at 780 ft

Specified head at 789.2 ft

No flow
Ten sets of values specified
(refer to program output)
   = 10°C
Convective flux in

Convective flux out and
conductive flux

No flow
Ten sets of values specified
(refer to program output)

0.33 ft

0.08 ft

3 days
afhe units of feet, gallons, calories, and days were used in this simulation.
                                     -130-

-------
       790
       780

       770
           COOLING
             LAKE
        750
      O
      m
        3801
          -565
                MARSH
    PEATS  AND CLAYS
  WISCONSIN RIVER
  AND DUCK CREEK
 (water, marsh and swamp)
-\n	,  ,^—-^—  	1778
             .SILTS_AND^FINE_SANDS	
              MEDIUM TO COARSE SANDS
    FINE SANDS WITH SILT AND CLAY LENSES
                            SANDSTONE
                                                                 J
                                                                2400
                                    FEET
Figure.  C-20.  Schematic cross section of the Columbia Generating Station site
             along an east-west line.
UJ °
0
2 2
cr
ID 5
CO D
£ 10
l-
5 "5
O
di 20
CD
t 4°
UJ
u- 4.on
COOLING
• 	 LAKE 	 	




























-DIKE-














	 	 MARSH SURFACE










































»















RIVER
^LOWLANDS








           -565   -180    -60 -30  0  20  40 60 105 160    420   2396
              -308   -110                            282   1100
                           FEET FROM THE DIKE
Figure  C-21.
The grid used to discretize the cross section simulated at the
Columbia Generating Station site.
                                  -131-

-------
         CGS  130'S
            UNITS DSHD ARE   GALLONS FOR FLOW, FT FOR DISTANCE,  CAL, CENTIGRADE
         105 128 10
         30001 1E14
         (1X.8G12.6)

         (1X.7G12.6)

         GROUP 22
         16 8
         -565 -308 -180 -110 -60 -30 0 20 40 60  105 160 260 420 1100 2396
         0 -2 -5 -10 -15 -20 -40 -400
         GROUP III
         311 3785.68
         10 0 1
         0-11
         89.30000000 89.30000000 89.30000000
         89.30000000 89.30000000 84.50000000
         80 0000000800000000 80 0000000 80 0000000
         800000000 80 0000000 80 0000000 79 0000000
         78.5 0000000 78.5 0 0 0 0 0 0 0 999
         001
         GROUP IV CGS 130'S CARDS USED FOR SAVANNAH PAPER
         10
1.6 1.6 1 .7 .7 1 .2 .2
1-11
1234465
1234465
1234465
1234465
7234465
7 2 10 8 8 6 5
1 9 10 8 8 6 5
1 9 10 8 8 6 5
1 9 10 8 8 6 5
1 9 10 8 8 6 5
1 9 10 8 8 6 5
1 9 10 8 8 6 5
1 9 10 8 8 6 5
1 1 10 8 8 6 5
1 1 10 8 8 6 5
1
















999

100 20 0 20000 20000 26000 .33
2
110 8000 8000 26000
3

.33 .08

                                       .08
                                                                  (continued)

Figure C-22.   The data deck used to  model heat flow at the  Columbia
                Generating Station site.
                                        -132-

-------
      100 20 0 17000 17000 21000 .33 .08
      4
      500 50 0 17000 17000 21000 .33 .08
      5
      10 10 0 8000 8000 26000 .33 .08
      6
      63 1 72 1 81 1 144 1 153 1 162 1
      63 1 72 1 81 1 144 1 153 1 162 1
      0 0
      GROUP VI
      0
      GROUP VII
      000000000000000000000000000000 00 0000 00 0
      0 0
      GROUP VIII
      0 9
      132333435363738393 10  3  11  3
      GROUP IX
      1 1 40 1 0 0
      2 9 16 24 26 33 40 48 50 57 64 72
                                 74 81 88 96 98 105 112 120 122 129 136  144  146
      153 160 168 170 177 184 192 194 201 208 216  218  225 232 240
      -1.0001 11100
      11111111
      1
                11 1  15             5
  860
             857   10.2222       24.6111
             858   12.5000       26.0555
             859   12.7222       25.1666
             860   11.8333       24.6111
11111110
-81 1
-142 -284 -142 142 284 142
000 3.15E6 6.36E6 3.15E6
12000 12000 12000 12000  12000 12000 12000  12000
12000 12000 12000 12000  12000 12000 12000  12000
0 0
000000
0000000
12000 12000 12000 12000  12000 12000 12000  12000
12000 12000 12000 12000  12000 12000 12000  12000
81 1
561.5 1123 561.5 -561.5  -1123 -561.5
-9.08E6 -1.82E7 -9.08E6  000
12000 12000 12000 12000 12000 12000 12000  12000
12000 12000 12000 12000 12000 12000 12000  12000
0 0
0000000
0000000
12000 12000 12000 12000 12000 12000 12000  12000
12000 12000 12000 12000 12000 12000 12000  12000
Figure  C-22.
          The data deck used  to model heat flow at the Columbia
          Generating  Station  site.
                                           -133-

-------
   CGS 130'S  UNITS USED ARE   GALLONS FOR FLOW,FT  FOR DISTANCE,CAL,CENTIGRADE
ti OF NODES 128  9 OF ELEMENTS 105  PANDWIDTH 10

VELOCITIES ARE  BEING CALCULATED AT CURRENT TIME STEP   1
THE X-SPACING IS
-565.0000 -508.0000
40.0000 60.0000
THE Y-SPACING IS
0.0 -4.0000
EDUCATION 2 TIME STEP 3
-160.0000
105.0000

-5.0000
. 00000
PARAMETER FACTORS— IN ORDER 1.60000
1
u» MATERIAL HOR PERM
1 160.0
2 1,800
3 160.0
4 800.0
5 16.00
6 200.0
7 1.800
8 1600.
9 3.200
10 280.0
VER PERM
32.00
1.600
32.00
80.00
16.00
19.20
1.600
160.0
3.200
40.00
-110.0000
-160.0000

-10.0000
HEAT OR MASS
1.60000
-60.0000 -30
260.0000 420

-15.0000 -20
CAPACITY COEFFICIENT
1.00000 .700000
.0000 0
.0000 1100

.0000 -40
3785.68
. 700000 1
STORAGE HOR THERM COND VERT THERM COND
.0000 .1400+05 .1400+. 05
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
5600.
.1190+05
.1190+05
5600.
.1190+05
5600.
.1190+05
5600.
.1190+05
5600.
.1190+05
.1190+05
5600.
.1190+05
5600.
.1190+05
5600.
.1190+05
.0 20.0000
.0000 2396.0000

.0000 -400.0000





.00000 .200000 .200000
SPECIFIC DISPERSION
HEAT
.2600+05 .6600-01
.2600+05 .6600-01
.2100+05 .6600-01
.2100+05 .6600-01
.2600+05 .6600-01
.2100+05 .6600-01
.2600+05 .6600-01
.2100+05 .6600-01
.2600+05 .6600-01
.2100+05 .6600-01
COEFS
. 1600-01
. 1600-01
. 1600-01
.1600-01
. 1600-01
.1600-01
. 1600-01
.1600-01
.1600r.01
.1600-01
           Figure C-23.   Program output  for Columbia Generating Station problem,   (continued)

-------
          IN THE TYPE OF MATERIAL MATRIX ARE
LU T*»*JW»p»fc* *
1.00000
1.00000
1.00000
1.00000
7.00000
7 . 00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
2.00000
2.00000
2.00000
2.00000
2.00000
2.00000
9.00000
2.00000
9.00000
9.00000
9.00000
9.00000
9.00000
1.00000
1.00000
3.00000
3.00000
3.00000
3.00000
3.00000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
4.00000
4.00000
4.00000
4.00000
4.00000
8.00000
8.00000
8.00000
8.00000
8.00000
8.00000
8.00000
8.00000
8.00000
8.00000
4.00000
4 . 00000
4.00000
4.00000
4.00000
8.00000
8.00000
8.00000
8.00000
8.00000
8.00000
8.00000
8.00000
8.00000
8.00000
6.00000
6.00000
6.00000
6.00000
6.00000
6.00000
6.00000
6.00000
6.00000
6.00000
6.00000
6.00000
6.00000
6.00000
6.00000
5.00000
5.00000
5.00000
5.00000
5.00000
5.00000
5.00000
5.00000
5.00000
5.00000
5.00000
5 . 00000
5.00000
5.00000
5.00000
I
I—"
Ul
INFORMATION FOR MASS OR HEAT TRANSFER ACROSS A SPECIFIED HEAD BOUNDARY
ELEMENT NUMBER—BOUNDARY CODE—TEMPERATURE OR CONCENTRATION OF INCOMING FLUID
     1  i    .00000        8  1   .00000      15   1    .00000       22  1     .00000
   36  1
CONVECTIVE OUT AND CONDUCTIVE BOUNDARY INFORMATION
                                  29  1
                                                                                                         .00000
1 1 12000.0
50 1 .000000
85 1 .000000
64 1 8000.00
99 1 8000.00
8 1 12000.00
57 1 .000000
92 1 .000000
71 1 8000.00
15 1 12000.0
64 1 .000000
99 1 .000000
75 1 8000.00
         ELEM //—BOUNDARY  CODE—TRANSFER COEFFICIENT
          22  1  12000.0         43  1   .000000
         71 1  .000000         78 1  .000000
         50 1  8000.00         57 1  8000.00
         85 1  8000.00         92 1  8000.00
WATER FLOW BOUNDARIES--ELEM # AND BOUNDARY CODE
   11         81       15  1     22  1     29  1
  71  1        78  1       85  1     92  1     99  1
POTENTIAL DISTRIBUTION AT TIME STEP
89.3000
89.3000
89.3000
89.3000
89.3000
84.5000
80.0000
89.3805
89.2615
89.2224
89.2390
89.0895
89.8503
80.2692
88.70000
88.3858
86.9984
85.9284
84.7128
83.3893
82.5679
.00000
88.6525
87.9897
85.8184
85.6591
84.3900
84.3395
82,7641
36  1
                                                   88.6355
                                                   87.9564
                                                   86.7581
                                                   85.5716
                                                   84.3048
                                                   83.3358
                                                   82.8003
                               Figure C-23.   (continued)
                                                                       43  1
        88.6218
        87.9318
        86.7171
        85.5204
        84.2692
        83.3626
        82.8165
                    50  1
                                                                                 88.4519
                                                                                 87.6714
                                                                                 86.3573
                                                                                 85.2024
                                                                                 84.2345
                                                                                 83.6783
                                                                                 83.1117
                                                                                           57  1
                                        64   1
86.8035
85.9030
84.9464
84.4145
84.0487
83.8094
83.5782

-------
-51.5399 -39.3041 -31.2270 -22.3388
-31.6491 -34.3912 -33.3403 -24.6573
-30.5509 -29.3358 -28.9887 -22.8616
-53.6430 -54.0320 -52.3883 -41.7264
-48.1144 -45.1488 -44.7073 -35.9204
-34.1526 -38.6777 -33.9313 -20.7011
-35.987 -15.4392 -35.8504 -31.1358
-134.856 -156.379 -129.908 -105.243
12.9762 18.4182
THE DISPERSION ROUTINE IS
MODE t AND TEMP AT INF.
71 9.00 78 9.00
71 9.00 78 9.00
TEMPERATURE DISTRIBUTION
23.0526 10.6176
24.0723 10.8383
23.2023 11.3082
28.3750 12.0244
JL 28.4317 11.7903
w 9.36313 9.22293
I 10.6118 10.1791
9.68002 9.95359
9.38139 10.0437
9.48225 10.0148
9.42817 10.0091
9.44609 10.3090
9.35584 9.97909
9.43319 9.99364
9.38966 9.98055
9.40101 9.98236
12.3757 9.10431
BEING USED
1 25.76 8 25.76
85 9.00 92 9.00
35 9.00 92 9.00
AT TIME STEP .00000
9.98973 10.0160
9.88251 10.0196
9.35211 10.0222
9.32307 10.0226
9.80178 10.0184
10.8940 10.0030
9.97268 10.9030
10-0947 9.99947
9.99198 10.0016
9.99649 10.0011
9.99830 10.0004
10.6900 10.0791
10.8035 9.99838
10.0908 9.99995
10.0016 9.99975
10.0016 9.99369
-10.7826 -4.78335
-12.7486 -4.72700
-12.2211 -4.62473
-23.2787 -9.53972
-20.6327 -9.53262
-19.2395 -11.5253
-22.1957 -14.5765
-64.2157 -34.1659
3.91385

15 25.76
99 9.00
99 9.00

9.99720
9.99675
9.99657
9.99680
9.99758
9.98476
9.99928
10.0001
9.99978
9.99965
9.99997
10.0001
10.0001
10.00000
10.0001
10.0001
CUMULATIVE MASS BLANCE RATES
CONDUCTIVE TRANSFER
CONVECTIVE TRANSFER—OUT
CONVECTIVE TRANSFER—IN
B. FLUX RECHARGE
B. FLUX DISCHARGE
CHANGE IN STORAGE
QUANTITY PUMPED
DIFFERECE
QXQT
Figure
.114876409
.446673+08
.118618409
.000000
.000000
.188595409
.000000
32501.0










.408686

22 25.76
50 9.00


10.0002
10.0002
10.0002
10.0002
10.0001
10.0030
10.0001
10.0000
10.0000
10.0001
9.99996
9.99995
9.99998
9.99996
9.99997
9.99998
-.416066
-.530258
-.619473
-1.59924
-2.14461
-3.85953
-5.71147
-12.2147
-.678204

43 25.76
57 9.00


9.99999
9.99999
9.99999
9.99999
10.00000
10.0000
10.00000
10.00000
10.0000
10.00000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000




50 25.76
64 9.00


10.0000
10.0000
10.0000
10.0000
10.0000
10.00000
10.0000
10.00000
10.0000
10.0000
10.00000
10.00000
10.00000
10.00000
10.00000
10.00000
FOR THIS TIME STEP
.114876409
.448673408
.112618409
.000000
.000000
. 188595409
.000000
32501.0



















C-23. Program output for Columbia Generating Station probl
57   9.00   64   9.00

-------
80.0000 80.1100
80.0000 80.1228
80.0000 80.1019
80.0000 80.0735
80.0000 80.0499
80.0000 80.0004
79.0000 79.0329
78.5000 78.4972
78.5000 78.5814

B. FLUX RECHARGE
B. FLUX DISCHARGE
CHANGE IN STORAGE
QUANTITY PUMPED
DIFFERECE
82.1493 82.4126
81.6793 82.1055
81.6384 81.8778
81.1874 81.3707
80.7483 80.8439
80.1561 80.1600
79.0889 79.1506
79.4910 78.4856
79.0047 78.5070
CUMULATIVE MASS BLANCE
417.916
417.716
.000000
.000000
.200280
82.4586
82.1513
81.8771
81.3596
80.5631
80.1619
79.1630
78.4846
78.5075






82.4821
82.1748
81.3987
81.3761
80.8744
80.1647
79.1703
78.4842
78.5077
RATES FOR THIS
417.916
417.916
.000000
.000000
.200280
82.7735
82.4630
82.1775
81.6172
81.0584
80.2634
79.2952
78.4326
78.5085
TIME STEP





83.4091
83.2320
83.0495
82.8280
82.1134
81.2531
80.1966
78.5328
78.4860






TIME STEP .000000
FLOWS IN THE X DIPRECTION


1
U)
T4










-.
_,
— .
-.
-.
rm ^
— 1
.
™ •
-.
-.
-.
-1
-.
,
996091-02
414756-01
281315-01
378952
453898
366091
.08199
372479-01
141957
359611-01
532992-01
673201-01
.67195
208530
443338-03
-.502178-02
-.179398-01
-.315544-01
-.552483-01
-.444803
-.499161
-.117809
-.524130-01
-.533644-01
-.434412-01
-.343201-01
-.251925-01
-.515699-01
-.339984
.275349-02
-1.
-5.
-10
-16
-26
-27
-22
-17
66834
99535
.8207
.8848
.8867
.6845
.8950
.1250
-1511524
-12
-10
-7.
-7.
-1.
.6539
.0149
56603
71698
10422
.159093-01
-8.87056
-31.4491
-56.9305
-86.1615
-114.767
-105.815
-117.353
-104.200
-92.2431
-77.3605
-60.7450
-47.0623
-42.0485
-6.71201
.115948
-9.05049
-32.
-57.
-85.
-106
-122
-114
-104
-93.
-78.
-61.
-47.
-42.
0244
8404
5518
.215
.489
.864
.412
4833
5292
6645
9377
3296
-6.81793
.121541
-9.
-33
-57
-75
-82
-63
-57
-52
-47
-40
-32
-25
-20
-3.
72060
.5617
.0730
.3980
.8756
.0097
.1372
.4734
.7045
.8782
.7559
.5635
.8387
74431
.647594-01
-16.0007
-43.3976
-58.9506
-65.2530
-64.8656
-65.0647
-62.0496
-59.6368
-57.2477
-53.3715
-47.7426
-40.4953
-30.9582
-8.90867
-.394883-01
FLOWS  IN THE Y DIRECTION
101.654
100.988
80.2117
109.236
8.97147
3.87355
102.223
98.5690
87.2790
86.4973
32.7361
-12.4882
100.305
96.0592
85.4840
80.4588
30.5889
-14.9164
87.9404
81.4697
70.3572
58.7120
18.1331
-13.2480
66.8651
57.0099
43.8252
29.4592
1.79816
-17.5263
45.0807
32.3655
19.3443
7.18937
-3.43804
-7.47321
18.5753
7.68152
2.90531
.918946
.399501-01
-.338393
                                Figure C-23.   (continued)

-------
The Heat Pump Problem

    This problem was coded to study the impact of the injection of cooled
waters from a heat pump into a shallow ground-water aquifer.  The problem is
discussed in more detail by Andrews (1978).

    The finite element grid used to discretize the problem is shown in Figure
C-24.  Since the problem is symmetric, only half the system was modeled.   A
novel feature of the finite element grid is that the thickness of each element
is proportional to distance along the horizontal axis from the wells—a
quasi-radial formulation.  The program input data and a sample program output
are listed in Figure C-25 and Figure C-26.

    A few unusual features are incorporated into this problem.  The heat  pump
problem was programmed to recompute flows at odd intervals, corresponding to
the end of each of the seasons.   The intervals at which flows were recomputed
are listed on cards 3 and 4 of data group IX.   ENTRY CHAN was called each time
flows were recomputed to read in the pumping schedule for the current
interval.   Air temperatures for the entire  simulation period were read in by a
call to ENTRY LAKE at the beginning of program execution.

    The major subroutines and the tasks performed by each are listed in a
program flow chart (Figure C-27).
o
SCREEN FOR
SHALLOW WELL
100
1-
LJ
LJ
U.
200-
SCREEN FOR
DEEP WELL
300-
>



























^^







=





i i
— •
=





	 «

z





1 1

oc.
•61.0 H
LJ
-91.5
 60     120           240
(18.3)                (73.2)
               FEET
           (METERS)
                                                          800    2000
                                                                (609.8)
 Figure C-24.   Grid used to discretize the aquifer simulated  in the  heat
               pump problem.
                                  -138-

-------
   HOME HEATING AND COOLING—SYMMETRICAL SET-UP
  UNITS  FEET GALLONS DAYS CALORIES
168  198  11
30001 1E17
(lx,9(G12.6)        )

(lx,8(G12.6)        )

GROUP II
22 9
0 2 10 25 40 48 50 75 100 102 110 135 165  180 195 200 225 250 255 270 285 350
2000 800 250 160  120 80 60 30 0 -30  -300
GROUP III
15 1 0 3785.68
10 0 1
0-11
10 10 10 10 10  10 10 10 10
    00000000000000000000000000000000000000
0000000000000000000000000000000000000000
0000000000000000000000000000000000000000
0000000000000000000000000000000000000000
                 0000000000000000000000000000000
999
0-11
111111111 10 10000000 10 00000000
10 00000000 10 00000000 10 00000000
10 00000000 10 00000000 10 00000000
10 00000000 10 00000000 10 00000000
10 00000000 10 00000000 10 00000000
10 00000000 10 00000000 10 00000000
10 00000000 10 00000000 10 00000000
10 0  0  0 0 0 0 0 0 999
GROUP IV
4
111.7.7111
0-11
11111111222222222222222233333333
4444444444444444444444444444444444444444
4444444444444444444444444444444444444444
4444444444444444444444444444444444444444
4444444444444444   999
1
 .1 .1 0 8000 8000 17000 10 1
2
10 100 0 17000 17000 21000 10 1
3
25 250 0 17000 17000 21000 10 1
4
 10 100 0 17000 17000 21000 10 1
GROUP V
 0 0
 6 6
 125 12 0 17000 17000 21000 .33 .08
 7
 110 8000 8000  26000 .33 .08
 8                                                          (continued)


  Figure  C-25.  The  data  deck used to model the heat pump  simulation
                 with no regional groundwater flow.


                                   -139-

-------
       1000 100 0 17000 21000 .33 .08
       9220 8000 8000 26000 .33 .08
       10
       175 25 0 17000 17000 21000 .33 .08
       GROUP V
       0 0
       0 0
       0 0
       GROUP VI
       0
       GROUP VII
       21
       1 1 12000 8 1 12000 15 1 12000 22 1 12000
       43 1 0 50 1 0 57 1 0 64 1 0 71 1 0 78  1 0 85 1 0 92 1 0 99 1 0
       50 1 8000 57 1 8000 64 1 8000 71 1 8000 78 1 8000 85 1 8000 92 1 8000 99 1 8000
       000000000000000000000
       6
       1 1 8 1 15 1 22 1 29 1 36 1
       0000000
       GROUP VIII
       15 0
       1 1 8 1 15 1 22 1 29 1 36 1 43 1 50 1  57 1 64 1 71 1 78 1 85 1 92 1 99 1
       GROUP IX
       100008
       53 69 93 101 52 68 92 100
       1 140 140 1 140 140 1 140 140 1 3
       5
       10 860
            1.0000     29.5000    41.0000
            2.0000     28.5000    35.9000
            3.0000     27.5000    30.8000
       11111110
Figure C-25.   The data deck used to model the heat pump simulation
                with no regional groundwater flow.
                                        -140-

-------
              HOME HEATING AND COOLING—SYMMETRICAL SET-UP
              UNITS FEET GALLONS  DAYS CALORIES
                                                                                04/07/78
                                                                                17:02:33
// OF NODES 198   # OF ELEMENTS  168   BANDWIDTH 11

VELOCITIES ARE BEING CALCULATED AT CURRENT TIME STEP
THE X-SPACING IS
    0,0          2.0000       10.0000
  100.0000     102.0000      110.0000
  225.0000     250.0000      255.0000

THE Y-SPACING IS
 2000.000      800.0000      250.0000
    0.0

EQUATION 2 TIME STEP  15.0000
   25.0000
  135.0000
  270.0000
  160.0000
 40.0000
165.0000
285.0000
120.0000
 48.0000
180.0000
350.0000
 80.0000
 50.0000
195.0000
 60.0000
 75.0000
200.0000
 30.0000
HEAT OR MASS CAPACITY COEFFICIENT   3785.68
PARAMETER FACTORS—IN ORDER   1.00000   1.00000   1.00000   .700000   .700000    1.00000   1.00000   1.00000

  MATERIAL   HOR PERM     VER  PERM     STORAGE   HOR THERM COND   VERT THERM COND    SPECIFIC   DISPERSION   COEFS
                                                                                   HEAT

1 . 10004-00
2 10.00
3 25.00
4 10.00
.1000+00 .0000
100.0 .0000
250.00 .0000
100.0 .0000
5600. 5600.
.1190+05 .1190+05
. 1190+05 . 1190+05
.1190+05 .1190+05
.1700+05 10.00 1.000
.2100+05 10.00 1.000
.2100+05 10.00 1.000
.2100+05 10.00 1.000
THE VALUES IN THE TYPE OF MATERIAL MATRIX ARE
1.00000
2.00000
2.00000
3.00000
4.00000
4.00000
4.00000
4.00000
4.00000
1.00000
2.00000
2.00000
3.00000
4. 00000
4.00000
4.00000
4.00000
4.00000
1.00000
2.00000
2.00000
3.00000
4. 00000
4.00000
4.00000
4.00000
4.00000
1.00000
2.00000
2.00000
3.00000
4.00000
4.00000
4.00000
4.00000
4.00000
1.00000
2.00000
2.00000
3.00000
4.00000
4.00000
4.00000
4.00000
4.00000
1.00000
2.00000
2.00000
3.00000
4.00000
4.00000
4.00000
4.00000
4.00000
1.00000
2.00000
2.00000
3.00000
4.00000
4.00000
4.00000
4.00000
4.00000
1.00000
2.00000
2.00000
3.00000
4.00000
4.00000
4.00000
4.00000
4.00000
             Figure C-26.   Program output for  the heat pump problem.
                                                 (continued)

-------
4.00000
A. 00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4,00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
4.00000
POINT SOURCES OF WATER—NODE //  AND AMOUNT
  63   1.00000          72   1.00000
162   1.00000
81
                                                    1.00000
144
1.00000
153   1.00000
POINT SOURCES OF HEAT
  63   1.00000          72   1.00000          81
162   1.00000
CONVECTIVE OUT AND CONDUCTIVE BOUNDARY  INFORMATION
    0 0  .00000
      1.00000         144    1.00000        153    1.00000

               ELEM //--BOUNDARY  CODE—TRANSFER   COEFFICIENT
HEAT FLOW BOUNDARIES—NODE // AND BOUNDARY  CODE
  13        23        33        43         53
                  6  3
       7  3
            8  3
                                                                                                 9  3
TEMPERATURE DISTRIBUTION AT TIME STEP
10.0000
10. 0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10,0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
.250000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000

10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000

10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000

10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000

10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
10.0000
                                 Figure C-26.   (continued)

-------
10.0000
10.0000
10.0000
10.0000
10.0000
POTENTIAL
10.00000
10.0013
10.0014
10.0015
10.0015
10.0015
10.0015
10.0016
10.0017
10.0017
10.0017
10.0018
10.0019
10.0019
10.0020
10.0020
10.0020
10.0019
10.0019
10.0019
10.0019
10.0018
10.0000
10.0000
10.0000
10.0000
10.0000
DISTRIBUTION
10.00000
10.0011
10.0012
10.0013
10.0014
10.0015
10.0015
10.0018
10.0022
10.0022
10.0023
10.0026
10.0029
10.0031
10.0033
10.0034
10.0037
10.0041
10.0042
10.0044
10.0046
10.0051
10.0000
10.0000
10.0000
10.0000
10.0000
AT TIME
10.00000
9.98874
9.98832
9.98773
9.98772
9.98799
9.98807
9.98989
9.99331
9.99365
9.99511
10.0005
10.0075
10.0109
10.0139
10.0148
10.0183
10.0201
10.0203
10.0206
10.0206
10.0198
10.0000
10. 0000
10.0000
10.0000
10.0000
STEP
10.00000
9.97939
9.97842
9.97589
9.97457
9.97279
9.97238
9.97052
9.97712
9.97797
9.98172
9.99548
10.0129
10.0215
10.0295
10.0319
10.0393
10.0383
10.0373
10.0336
10.0298
10.0222
10.0000
10.0000
10.0000
10.0000
10. 0000
000000
10.00000
9.97459
9.97322
9.96883
9.96588
9.96126
9.96016
9.95306
9.96360
9.96504
9.97125
9.99222
10.0164
10.0291
10.0419
10.0458
10.0557
10.0527
10.0502
10.0421
10.0346
10.0227
10.0000
10.0000
10.0000
10.0000
10.0000

10.00000
9.96998
9.96813
9.96082
9.95443
9.94321
9.94052
9.92129
9.94163
9.94436
9.95580
9.98861
10.0203
10.0391
10.0611
10.0683
10.0903
10.0755
10.0698
10.0527
10.0396
10.0229
10.0000
10.0000
10.0000
10.0000
10.0000

10. 00000
9.96806
9.96599
9.95687
9.94715
9.92931
9.92456
9.89284
9.92412
9.92863
9.94514
9.98716
10.0221
10.0449
10.0753
10.0860
10.1192
10.0933
10.0842
10.0588
10.0417
10.0228
10.0000
10.0000
10.0000
10.0000
10.0000

10.00000
9.96583
9.96346
9.95193
9.93387
9.89177
9.88408
9.81196
9.88125
9.88812
9.92506
9.98618
10.0242
10.0528
10.1086
10.1293
10.2003
10.1367
10.1176
10.0669
10.0442
10.0227
10.0000
10.0000
10.0000
10.0000
10.0000

10.00000
9.96433
9.96173
9.94858
9.91940
9.78885
9.72716
9.52379
9.72221
9.78247
9.90286
9.98653
10.0256
10.0565
10.1655
10.2876
10.4890
10.2951
10.1746
10.0708
10.0460
10.0225
B. FLUX RECHARGE
B. FLUX DISCHARGE
CHANGE IN STORAGE
QUANTITY PUMPED
DIFFERENCE
CUMULATIVE MASS BLANCE
    .000000
    .000000
    .000000
    .000000
    .000000
RATES FOR THIS TIME STEP
        .000000
        .000000
        .000000
        .000000
        .000000
             Figure  C-26.  Program output for the heat pump  problem.

-------
I.  BRIEF DESCRIPTION OF THE  PROGRAM ROUTINES

    A.  Main Routine

        This routine controls the flow of the  program and the program
        dimensions.  The parameter variables L,  N,  M control  the maximum
        number of elements, nodes, and bandwidth respectively.

        1.  Internal Subroutine ITERAT
            This subroutine controls the solution of a steady-state areal
            problem by an iterative method.
        2.  Internal Subroutine FLOADD
            This subroutine adds in a constant specified flow rate to the
            values calculated by the water-flow equation.

    B.  Subroutine DATAIN

        This subroutine reads in the program data.

        1.  ENTRY DI
            This entry point  passes arguments  from the main routine to
            subroutine DATAIN.
        2.  Internal Subroutine ELREAD
            This subroutine reads in the element node numbers.
        3.  Internal Subroutine P
            This subroutine creates a rectangular grid and numbers the nodes
            and elements in the grid.

    C.  Subroutine STRUCT

        This subroutine assembles the global structure matrix and the global
        capacitance matrix.

    D.  Subroutine GAUSS

        This subroutine forms the element structure and capacitance matrices
        by Gaussian Quadrature.

        1.  ENTRY DERIVE
            The main entry point for subroutine GAUSS.
        2.  Internal Subroutine CONVEC
            This routine handles the convective boundaries.

    E.  Subroutine SHAPE1

        This subroutine computes the shape functions for a linear element.

    F.  Subroutine SHAPE

        This subroutine computes the shape functions for elements with one or
        more nonlinear sides.
                                    -144-

-------
G.  Subroutine LOAD

    This subroutine computes the global recharge matrix by adding in the
    various sources and sinks.

    1.  ENTRY HEATE
        The main entry point for subroutine LOAD.
    2.  Internal Subroutine CB
        This subroutine computes mass or heat flow across a specified
        potential boundary.

H.  Subroutine SOLVE

    This subroutine is a symmetric banded matrix equation solver.

    1.  ENTRY BACK
        Entry point for the back substitution of the recharge matrix.
    2.  ENTRY MULTI
        Entry point for the multiplication in a transient problem.

I.  Subroutine ASOLVE

    This subroutine is an  assymmetric banded matrix equation solver.

    1.  ENTRY ABACK
        Entry point for the back substitution.
    2.  ENTRY AMULTI
        Entry point for matrix multiplication in a transient problem.

J.  Subroutine BALAN

    This subroutine computes  a mass  balance.

    1.  ENTRY MASBAL
        The main  entry point  for subroutine BALAN.
    2.  ENTRY BPRINT
        The entry point for the printing  of the mass balance.
    3.  ENTRY WATER
        Entry point for computing  water flows in each  element.
    4.  ENTRY VELO
        Computes  water velocities  at the  Gauss points.
    5.  ENTRY VCENT
        Computes  water velocities  at the  center  of each element.
    6.  Internal  Subroutine FLOWW
        Subroutine for computing element  flows.

K.  Subroutine FLOWS

    This subroutine prints out element  flows and nodal values.

     1.  ENTRY FFLOW
        Entry point  for the printing of nodal potentials on  file  14.

                                -145-

-------
    2.  ENTRY FFFLOW
        Entry point for the printing of nodal temperatures or
        concentrations on file 14.
    3.  ENTRY WFLOW
        Entry point for printing water flows for each element.
    4.  ENTRY HPRINT
        Entry point for printing nodal temperature or concentration
        values.
    5.  ENTRY WPRINT
        Entry point for printing nodal potential values.

L.  Subroutine PARAM

    This subroutine computes element parameters that are a function of
    velocity or temperature and localized coordinates.

    1.  ENTflY MECD
        Entry point for computing dispersivity.
    2.  ENTRY CORD
        Entry point for the computation of localized coordinates.
    3.  ENTRY PE
        Entry point for computing hydraulic conductivity as a function of
        temperature.
    4.  ENTRY PEE
        Entry point for the computation of aquifer thickness in an areal
        problem.

M.  Subroutine BOUNDA

    This routine is used to change parameter values or boundary conditions
    at each time step.  Five entry points, ENTRY LAKE, ENTRY BVAL, ENTRY
    BOUND, ENTRY CHANG, ENTRY CHAN, are provided for user programming.

N.  Subroutine EIGEN
    This subroutine is used to compute the maximum stable time step for a
    transient problem.

0.  Subroutine ADJUST

    This subroutine moves the upper boundary in a cross-section problem.

    1.  ENTRY ADJUST
        The main entry point for subroutine ADJUST.
                                -146-

-------
              MAIN PROGRAM
                 Q)—
DATAIN
Reads the  data in and
prints the data
If transient
problem go to
3


                             PE
                             Adjusts hydraulic
                             conductivity for
                             temperature
                                     STRUCT
                                     Assembles the
                                     structure matrix
                                     LOAD
                                     Assembles the  load
                                     vector—the right hand
                                     side of tha equation
                              ITERATE
                              Called in areal
                              steady state problems
MASBAL
Computes
balance
a mass
                             ADJUST
                             Called in a moving
                             boundary problem—
                             adjusts the boundary
                             and directs program
                             flow to 2 or 3 depending
                             upon how much the boundary
                             is moved
                              WPRINT, WTLOW,  BPRINT
                              Prints potentials' flows
                              and mass balance
                                       GAUSS
                                       Performs integration
                                       for an element using
                                       Gaussian Quadrature
                                                                     SOLVE
                                                                     Performs forward
                                                                     reduction of the
                                                                     structure matrix
                                       BOUNDA
                                       Reads in  new
                                       boundary
                                       conditions  for
                                       the current
                                       time step
                                       MULTI
                                       In a transient
                                       problem multiplies
                                       load vector by
                                       potential values
                                       at last time step
                                       BACK
                                       Solves  for the
                                       unknown potentials
Figure  C-27.    Program flow chart.
                                                  -147-

-------
                         If not a linked
                         problea STOP
                         STRUCTURE
                         Assemble the structure
                         natrix for equation number
                         two
                          LOAD
                          Assemble the load vector
                          for equation number two
                         IttSBAL
                         Computes a m»»» balance
                         for equation number two
                         HRPICT. BPRI8T
                         Prints out «ass or
                         energy concentrations at
                         nodes and the BBSS
                         balance
GAUSS
Performs integration
for an element using
Gaussian Quadrature
                                                                  ASOLVE
                                                                  Performs forward
                                                                  reduction of the
                                                                  structure maxtrix
                                                                  (matrix is not
                                                                  symmetrical)	
BOUNDA
Reads in new
boundary
conditions for
the current tine
step	
AMULTI
In a transient
problem multiplies
load vector by heat,
nass values at last
tine step
 ABACK
 Solves  for  the
 unknown potentials
PEE
Computes aquifer
thickness
                                 CORD
                                 Computes
                                 localized
                                 coordinates
                         CONTROL TRAHSFER
                         Three possible types of
                         transfers—one chosen
                         depends upon type of
                         problea    	
                                 SHAPE
                                 Computes the
                                 shape
                                 functions
VELCO
Computes
velocity
at the Gauss
points
MECD
Computes the
dispersion
coefficients
at the Gauss
points	
 CONVEC
 Computes
 coefficients for
 convective and
 flow
 boundaries
Figure  C-27  (continued)
                                                      -148-

-------
 THE FINITE ELEMENT PROGRAM LISTING
C
C
C
C
C
C1
C
C
C
C
C
C
C
    THE  MAIN  ROUTINE FOR THE FINITE ELEMENT PROGRAM
           BY CHARLES ANDREWS
MAI
MAI
MAI
MAI
MAI
MAI
MAI
MAI
MAI
         THE PARAMETER STATEMENTS
       L—NUMBER OF ELEMENTS, N —NUMBER OF NODES.
       M— BANDWIDTH, Z—SIZE OF ARRAYS CONTAINING SOURCE
        SINK, AND BOUNDARY INFORMATION

            PARAMETER H=17,N=160,L=130,Z=50


            **t*****************************************************
            PARAMETER P=M»2-1,BBB=N
    INTEGER AY(4),AZ(4)
    DATA AY/4,1,4,3/
    DATA AZ/3,2,1,2/
   DOUBLE PRECISION S,T
   REAL MAT  INFLOW
    COMMON/ACC/KK(12)-K1 ,K2,K3,K4
     COMMON/AM/NLA,PCT/CON/MC1,MC2,HCONV
  «/CONT/LA,LB,LC,LD.LE,LF,LG LH
   COMMON/HI/TITLE(25),V(26),VV(26)
   COMMON/HM/PXX  PYY ,PXY,KAD/HH/LFLOW,LON/ME/MEQ
    COMMON/ATHICK/ASIZE.NTHICK.THICK  ERROR
MODEL  PARAMETERS--'.
   DIMENSION WXARIES
         DIMENSION NWATER(Z) .AWATER(Z) ,NHEAT(Z) ,AHEAT(Z)
 C— ELEVEATION AT THE BOTTOM OF THE AQUIFER
         DIMENSION BOT(BBB)
 C— INFORMATION ON CONVECTIVE  BOUNDARIES                      CBALCZ 2)
      •  ,NCON(Z,3) .TINF(Z),CONV(Z),ALOC(Z),NEL(Z,2).AEL(Z,2),CBALU,^
 C— STRUCTURE MATRICIES
         DIMENSION GG(N), S( N,P) ,T( N ,P)
     100
     200
     300
     100
     500
     600
     700
     800
     900
MAI 1000
MAI 1100
MAI 1200
MAI 1300
MAI 1400
MAI 1500
MAI 1600
MAI 1700
MAI 1800
MAI 1900
MAI 2000
MAI 2100
MAI 2200
MAI 2300
MAI 2400
MAI 2500
MAI 2600
MAI 2700
MAI 2800
MAI 2900
MAI 3000
MAI 3100
MAI 3200
MAI 3300
MAI 3400
MAI  3500
MAI  3600
MAI  3700
MAI  3800
MAI  3900
MAI  4000
MAI  4100
 MAI  4200
 MAI  4300
 MAI  4400
 MAI  4500
 MAI  4600
 MAI  4700
 MAI  4800
 MAI  4900
 MAI  5000
 MAI  5100
 MAI 5200
 MAI  5300
 MAI 5400
 MAI 5500
 MAI 5600
 MAI 5700
                                       -149-

-------
       CALL URDATE(IDATE.IYEAR)
       CALL URTIMD(ITIME,ISEC)
       FORMAKIX,'MAXIMUM  NUMBER OF NODES PERMITTED EXCEEDED')
3      FORMATC1X,'MAXIMUM  NUMBER OF ELEMENTS EXCEEDED')
4       FORMAK IX, 1H1)
       READ 7,(TITLE(J),J=1,24)
       PRINT 300 (TITLE(J),J=1,24),IDATE,IYEAR,ITIME,ISEC
7      FORMAT(12A6)
      READ,LM,LN,MBAND
       PRINT 8,LN,LM,MBAND
8      FORMAT(1X,'*OF NODES',14  '  * OF  ELEMENTS', 1*4  '  BANDWIDTH1
       IF(LM.GT.L)  PRINT 3
       IF(LN.GT.N)  PRINT 1
      MPP=MBAND»2-1
       NAA=0
C*»****«**»*«*«*»***«*«***»«*****«****««*»*»«*«*««»*«*»**»*
c
c
     INPUT THE DATA

      CALL DATAINUM.LN MBAND,MAT,R1 ,R .HEA,HEAD,WX,WY,STO,INFLOW,PX,PY,
     •PCX,HEAT,XLOC.YLOC,NOD,TPX TPY,KRANA,CFACT,CFACT1,
     •ALPHA.KSTEP.KBOUND,KFT,PCW,KRAN.ALPH,KSTE,KBOUN,KF,PCH,DIFF,
     •Z,NCON,TINF,CONV,ALOC,LWATER HWATER,AWATER,LHEAT,NHEAT,AHEAT.
     •LFLUXW,NFLUXW,AFLUXW,LFLUXH,NFLUXH,AFLUXH,BBB,BOT,KTYPE
     «,NODE,NEL,AEL)
       CALL DI(KAREAL,ALPHM,ALPHAM,ER,ITER.LEL
     •-LINEW,LINEH,NLINEW,NLINEH,ALINEW,ALINEH)
C
CC««**fti**«***»**«***«t****ft»»«»»«****«ttft*»t»*«ft***«*ftff**ft
C  INITIALIZE THE STARTING ADDRESSES  AND STARTING VALUES
C
       CALL FLOWS(LM,LN,R.R1,FLOWX,FLOWY)
C
       CALL BOUNDA(LM,LH.Z,R,R1.HEAD,HEA,FLOWX,FLOWY,HEAT,INFLOW
     • NWATER AWATER,NHEAT.AHEAT.NCON,TINF
     • CONV,ALOC,NEL,AEL,CBAL,LEL,LHEAT,LWATER,PCW,AY,AZ
     •,LINEW,LINEH,NLINEW,NLINEH,ALINEW,ALINEH)
       CALL PARAMUM.LN.BBB.XLOC,YLOC.NOD.PCX.DIFF,R,R1 ,WX,WY,TPX
     «,TPY,BOT,KAREAL>
       CALL VELOC(LN,LM,R1,WX,WY)
C
       MU=0

C  PROGRAM CONTROL  IS  ESTABLISHED
       READ,LA,LD,LE,LF,LG,LH
       IF(LE.GT.O)  READ,(NTIME(I),1=1.LE)
       IF(LF.EQ.I)  READ,ASIZE,NTHICK,MTHICK ERROR,XADD,YADD
       IFCLH GT.O)  READ,(HSPACE(I),1=1,LH)
      IF(LA.EQ.I) CALL LAKE
       IF(LG.GT.O)  CALL ADJUSTUG LM.LN Z.HEA.XLOC,YLOC.NOD
     •  ,FLOWX,FLOWY.R1)
        GO T0(10.20,30,30,30), KTYPE
10      CONTINUE
        MAr-1
        MB=1
        MC=1
        MD=1
        ME=1
       READ.MW
        GO TO 120
20      READ,MA,MB,MW,MC,MD
               IF(MD.LE.O) MU=1
        ME=1
        GO TO 120
MAI 5800
MAI 5900
MAI 6000
MAI 6100
MAI 6200
MAI 6300
MAI 6400
MAI 6500
MAI 6600
MAI 6700
MAI 6800
MAI 6900
MAI 7000
MAI 7100
MAI 7200
MAI 7300
MAI 7400
MAI 7500
MAI 7600
MAI 7700
MAI 7800
MAI 7900
MAI 8000
MAI 8100
MAI 8200
MAI 8300
MAI 8400
MAI 8500
MAI 8600
MAI 8700
MAI 8800
MAI 8900
MAI 9000
MAI 9100
MAI 9200
MAI 9300
MAI 9*400
MAI 9500
MAI 9600
MAI 9700
MAI 9800
MAI 9900
MAI10000
MAI10100
MAI10200
MAI10300
MAI10400
HAI10500
MAI10600
MAI10700
MAI10800
MAI10900
MAI11000
MAI11100
MAI11200
MAI11300
MAI11400
MAI11500
MA111600
MAI11700
MAI11800
MAI11900
MAI12000
                                     -150-

-------
30      READ,MO,MP.MV,MQ,MR.MS,MT MU.ME
        MA = -1
        MB=1
        MC=1
        MD=1
        ME = 0
       MW=9999999999
        ALPH=0.0
120      CONTINUE
C   NA IS A COUNTER
       NA=0
C THESE LINES ESTABLISH NO CONVECTIVE BOUNDARIES IN EQN 1
       NCC=NCONV
       NCONV=0
C NO DISPERSION IN EQN 1
        KAD = 0
C   CONTROL FOR SYM OR ASSYM SOLUTION TECHNIQUE
        NLB=NLA
        NLA=LC
       IFUTYPE.EQ.5) GO TO  160
       CONTINUE
       CONTINUE
124
125
C
 C
 C
 C
      EQUATION NUMBER ONE
 C  ADJUST  PERMEABILITIES  FOR  CHANGING  TEMPERATURES
        IF(MS.EQ.I)   CALL PE(MS)
 C   MULTIPLICATION  FACTOR FOR TIME STEP
        IF(KS.GT.O.OR.NA.GT.O) ALPH=ALPH*ALPHM
 C  INITALIZE  STARTING ADDRESSES IN HEATER GAS AND BALANCE FOR EQUATION
        MEQ=1
       CALL LOAD(LM,LN,MBAND,NOD,NODE,R1,RI,G
      ».GG,INFLOW,HEA.ALPH,KBOUN,KF,PCH,KRANA,CFACT1 KS,
      *Z NCON,TINF,CONV,ALOC,LWATER,NWATER,AWATER,CBAL,AY,AZ
      » ,AEL,NEL,MEQ,LEL,CBALA,LINEW,NLINEW,ALINEW)
         CALL  BALAN(LM,LN,R1,WX,WY,STO HEA,INFLOW,Z,LWATER.AWATER
      *   LFLUXW,NFLUXW,WA,WB,WC,WD,WE,MPRINT:NTYPE,RI,NODE,NOD
      *FLOWX FLOWY.PCW,MEQ,CBAL,CBALA,AY,AZ,NCON,KRANA)
 C
 C
 C

 130
  C
  C
       CALL GAUSSULL-LM,LN,XLOC,YLOC,NOD,WY.WX,INFLOW,STO,ALPH,KBOUN,
     * KRANA,Z,NCON,CONV,ALOC,NODE,CBAL,AY,AZ)
      CALL STRUCT(LN,LM,MBAND,MPP,R1.S  T,G .GG ,RI ,HEA,XLOC ,
     *YLOC,NOD,NODE,CFACT1,KRANA,ALPH,KBOUN)

       NAA=NAA+1
       NA=NA+1
         BKS=BKS-«-ALPH
       IFULPH.LT.0.0001)  BKS=AKS

       IFCKAREAL.EQ.1)  CALL ITERATCER,ITER,KSA,$12M)
       KSA~0
         IF(MC.GT.O)  CALL MASBAL(ALPH)
         M1=NA/MD
         MM' — Ml^M 0
        IF(MMI.EQ.NA.AND.MU.EQ.O)  CALL WATER


        IF(LG.GT.O) CALL ADJUS($125,$130)
MAI12100
MAI12200
MAI12300
MAI12MOO
MAI12500
MAI12600
MAI12700
MAI12800
MAI12900
MAI13000
MAI13100
MAI13200
MAI13300
MAI13400
MAI13500
MAI13600
MAI13700
MAI13800
MAI13900
MAI14000
MAI14100
MAI14200
MAI11300
MAI14400
MAI14500
MAI14600
MAI14700
MAI14800
MAI14900
MAI15000
 MAI15100
 MAI15200
 MAI15300
 MAI15400
1 MA115500
 MAI15600
 MAI15700
 MAI15800
 MAI15900
 MAI16000
 MAI16100
 MAI16200
 MAI16300
 MAI16400
 MAI16500
 MAI16600
 MAI16700
 MAI16800
 MAI16900
 MAI17000
 MAI17100
 MAI17200
 MAI17300
 MAI17400
 MAI17500
 MAI17600
 MAI17700
 MAI17800
 MAI17900
 MAI18000
 MAI18100
 MAI18200
 MAI18300
 MAI18400
                                      -151-

-------
c
c
c
c
c
      MUNA/MB
      M1=M1«MB
      IF(MI.EQ.NA)  CALL WPRINT(BKS)
      M1=HA/MC
      M1=M1»MC
      IF(Ml.EQ.NA)  CALL BPRINT

ADDS IK A CONSTANT  BACKGROUND FLOW RATE

     IF(XADD.EQ.O.AND.YADD.EQ.O) GO TO 145
C
145
C
C
       CALL FLOADD
     CONTINUE
        IF(MMI.EQ.NA.AND.MU.EQ.O)  CALL  WFLOW(BKS)
       M1=NA/MW
       H1=M1»MW
       IF(MI.EQ.NA) CALL FFLOW
       IF(MA.LE.BKS) GO TO 155
       IF(ALPHM.GT.I) GO TO 125
       GO TO 130
155    CONTINUE
        IF(ME.EQ.I) STOP
160    CONTINUE
£*••«»••••*»•*•*••••«•**•*••«••«•••*••*••**••*••••••ft*«««t»*
C
C    EQUATION NUMBER TWO
C
£••***•••••••••••••••••t«»ll*«*«««**«»««»«»*«*t««•••*••*•**«
       NTHICK=MTHICK
       NCONVrNCC
        KAD=MT
        NUUNLB
(;•*«•*•••••••*«•«*•••*•••*•**«*t•*«*••«•»•»«•••••»••««••••••!•••
C MULTIPLICATION FACTOR FOR THE TIME  STEP
       IF(KS.GT.O) ALPHA=ALPHA«ALPHAM
C
C
C
 INITIALIZE  STARTING  ADDRESSES
     MEQ=2
IN HEATER AND BALANCE FOR EQUATION 2
      CALL LOADUM.LN.HBAND.NOD,NODE,R,RI,G
     «GG,HEAT ,HEAD ,ALPHA,KBOUND,KFT,PCW,KRAN.CFACT,KS,
     »Z,NCON.TINF,CONV,ALOC,LHEAT,NHEAT  AHEAT.CBAL,AY,AZ
     * ,AEL,NEL,HEQ,LEL,CBALA,LINEH,NLINEH,ALINEH)
        CALL BALANCLM,LN,R,PX,PY,PCX,HEAD,HEAT,Z,LHEAT,AHEAT.LFLUXH,
MAI18500
MAI18600
MAI18700
MAI18800
MAI18900
MAI19000
MAI19100
MAI19200
MAI19300
MAI19400
MAI19500
MAI19600
MAI19700
MAI19800
MAI19900
MAI20000
MAI20100
MAI20200
MAI20300
MAI20400
MAI20500
MAI20600
MAI20700
MAI20800
MAI20900
MAI21000
MAI21100
MAI21200
MAI21300
MAI21400
MAI21500
MAI21600
MAI21700
MAI21800
MAI21900
MAI22000
MAI22100
MAI22200
MAI22300
MAI22400
MAI22500
MAI22600
MAI22700
MAI22800
MAI22900
MAI23000
MAI23100
MAI23200
MAI23300
MAI23400
C
C
c
c
c
   •   NFLUXHtHA,HB,HCfHD,HE,MPRINT.NTYPE,RI,NODEfNODsFLOWX,FLOWY,PCWf MAI23500
   *   MEQ,CBAL,CBALA,AY.AZ,NCON,KRAN)                                 MAI23600
                                                                     MAI23700
                                                                     MAI23800
                                                                     MAI23900
     CALL GAUSS(LLL,LM,LN.XLOC,YLOC,NOD,PY.PX.HEAT.PCX,ALPHA,KBOUND,KRMAI24000
   •AN,Z,NCON,CONVfALOC,NODE,CBAL,AY,AZ)                              MAI24100
                                                                     MAI24200
   CALL STRUCT(LN,LM,MBAND,MPP R,S T-G,GG,RI,HEAD,XLOC,YLOC,         MAI24300
   «  NOD.NODE,CFACT,KRAN,ALPHA.KBOUND)                                MAI24400
                                                                     MAI24500
                                                                     MAI24600
                                     -152-

-------
        CONTINUE
      KS=KS+1
      AKS=AKS+ ALPHA
        CALL HEATE(KS)
       IF(LH.EQ.O)  GO TO 209
       DO 208 1=1.LH
       J=NSPACE(I)
       YSPACEU) = R(J)
       WRITEC15.V)  (YSPACE(I),I=1,LH)
       CONTINUE
       IF(MQ.GT.O)  CALL MASBAL(ALPHA)
       IFUSIZE.NE.-1 .0)        GO TO 212
      DO 211 1=2,LN 2
200
C
C
C
C
 208

209
011    RfT — l^ — RTT^
212   CONTINUE
        M1=KS/MP
        M1=M1«MP
        IF(M1.EQ.KS) CALL HPRINT(AKS)
       M1=KS/MV
       M1=M1«MV
       IF(M1.EQ.KS) CALL FFFLOW
        M1=KS/MQ
        M1=M1«MQ
        IF(M1.EQ.KS) CALL BPRINT
        IF(MO.LE.AKS) STOP
C NTIME CONTAINS INFORMATION ON WHEN FLOWS ARE TO BE RECOMPUTED
C  NAA IS A COUNTER INDICATING THE NUMBER OF TIMES FLOWS HAVE BEEN
       IF(LE.EQ.1.AND.NTIME(NAA).EQ.KS) GO TO 120
        M1=KS/MR
        M1=MR*M1
       IF(MLNE.KS) GO TO 200
       GO TO 120
C
300
  FORMAT(1H1/1X,130( t*1)/1X,3('
* 2('*' ,35X,12A6,T130,'*'/1X), '
                                      ,T130, '*
                                      ,T130,'*'/1X,
            ,T100,A6,A2,T130.'*'/1X) ,130('*M)
C»«»»«(«»»«*««*»*««««»«*»«»«»««*»«********»*******************
C
       SUBROUTINE ITERATCER.ITER.KS.*)
       IF(KRANA.NE.O) RETURN
        A=0
       DO 1J=1,LN
       B=RI(J)-RHJ)
       B=ABS(B)
1      A=AMAX1(B,A)
       KS=KS+1
       IF(A.LT.ER.OR.KS.GT.ITER) PRINT 5.KS
       IF(A.LT.ER.OR.KS.GT.ITER) RETURN
       RFTII RN ll
5      FORMAT(1X//1X,'ITERATIONS NEEDED  FOR  CONVERGENCE   ,15)
C
     MAI24700
     MAI2U800
     MAI24900
     MAI25000
     MAI25100
     MAI25200
     MAI25300
     MAI25UOO
     MAI25500
     MAI25600
     MAI25700
     MAI25800
     MAI25900
     MAI26000
     MAI26100
     MAI26200
     MAI26300
     MAI26UOO
     MAI26500
     MAI26600
     MAI26700
     MAI26800
     MAI26900
     MAI27000
     MAI27100
     MAI27200
     MAI27300
     MAI27400
     MAI27500
     MAI27600
     MAI27700
COMPUMAI27800
     MAI27900
     MAI28000
     MAI28100
     MAI28200
     MAI28300
     MAI28MOO
     MAI28500
     MAI28600
     MAI28700
     MAI28800
     MAI28900
     MAI29000
     MAI29100
     MAI29200
     MAI29300
     MAI29MOO
     MAI29500
     MAI29600
     MAI29700
     MAI29800
     MAI29900
     MAI30000
     MAI30100
     MAI30200
     MAI30300
                                     -153-

-------
c««»*«»«*«««*•»««•»«««*»•««««««»*«««**»«»»»««»»«**««««««*»»*            MAI 30500
C                                                                      MAI30600
       SUBROUTINE FLOADD                                               MAI30700
C                                                                      MAI30800
        DO TJO KSS=1.LM                                                MAI30900
       CALL CORD(KSS)                                                   MAI31000
       Y=YLOC(Kfl)-YLOC(K1)+YLOC(K3)-YLOC(K2)                            MAI31100
       Y=ABS(Y)/2                                                      MAI31200
       X=XLOC(K1)+XLOC(K1)-XLOC(K3)-XLOC(K2)                            MAI 31300
       X=ABS(X)/2                                                      MAI3140G
       XX=1.0                                                          MAI31500
       YY=1.0                                                          MAI31600
       IF(NTHICK.EQ.O)  GO TO 135                                        MAI31700
       YY=(YLOC(K1)+YLOC(K4))/2*ASIZE                                   MAI31800
       XX=(XLOC(K3)+XLOC(KU))/2*ASIZE                                   MAI31900
       XX=ABS(XX)                                                      MAI32000
       YY=ABS(YY)                                                      MAI32100
135    CONTINUE                                                        MAI32200
       FLOWY(KSS)=FLOWY(KSS)+YADD*X/YY                                 MAI32300
       FLOWXCKS3)=FLOWX(KSS)-»-XADD«Y/XX                                 MAI32400
140    CONTINUE                                                        MAI32500
       END                                                             MAI32600
                                     -154-

-------
C*««««**«««««***»***«**»»«*«»»»****««*t«*«***««*«««»**««*«»««»«
c
C    THIS  ROUTINE  IS  USED TO READ IN  THE DATA
C
      SUBROUTINE DATAINCLM,LN,MBAND,MAT,R1,RtHEA,HEAD,WX,WY,STO
     «,INFLOW,PX.PY.PCX,HEAT,XLOC.YLOC,NOD,TPX,TPY,KRANA,CFACT.CFACT1,
     •ALPHA,KSTEP,BOUND,KFT,PCW.KRAN,ALPH,KSTE,BOUN,KF,PCH,DIFF,
     *Z,NCON,TINF,CONV,ALOC,LWATERfNWATER,AWATER,LHEAT,NHEAT,AHEAT,
     ,LFLUXW,NFLUXW,AFLUXW,LFLUXH,NFLUXH,AFLUXH,BBB,EOT.KTYPE
     *,NODE,NEL,AEL)
       INTEGER BOUND,BOUN,BBB,Z
      REAL INFLOW,MAT
      COMMOM/HI/TITLE(25),V(26),VV(26)/HH/LFLOW,LON
      COMMON/A1/M,MM,NUMNP,NSIZE/AM/NLA,PCT
       COMMON/CON/MC1,MC2,NCONV
       DIMENSION NCON(Z,3),TINF(Z),CONV(Z),ALOC(Z),NWATER(Z),
     *  AWATER(Z),NHEAT(Z),AHEAT(Z),NFLUXW(Z,2),AFLUXW(Z),NFLUXH(Z, 2) ,
     *AFLUXH(Z),BOT(BBB),AEL(Z,2),NEL(Z,2)
      DIMENSION MAT(LM),R1(LN),R(LN),HEA(LN),HEAD(LN),WX(LM),WY(LM)
      DIMENSION STO(LM).INFLOW(LM),PX(LM),PY(LM),PCX(LM),HEAT(LM)
      DIMENSION XLOC(LN),YLOC(LN),NOD(12,LM), TPX(LM),TPY(LM)
       DIMENSION FF(8),Q(12), A(8 ,50) , DIFFUM ,2) , NODEC LM)
       RETURN
C
C
       ENTRY DI(KAREAL,ALPHM,ALPHAM,ER,ITER,LEL,
     «LINEW,LINEH,NLINEW,NLINEH,ALINEW,ALINEH)
       DIMENSION NLINEW(Z,2),NLINEH(Z,2),ALINEW(Z),ALINEH(Z)
£f ••*«»»««*«««««««»««*«««**«*»«*»«*«*»•*«**«««*«»««**»»«»*»«
C
      DO 2 J=1,LM
      HEAD(J)=0.0
 2    HEA(J)=0.0
C
C
£«*«**ft**ft«*«»ft**««***ftX««»***tt«*«X«»»»****«*»****«««*««*»*»
C
    7 FORMATC12A6)
        READ,KTYPE,KPRINT,LON,MQ,NLA,CFACT
       CFACT1=CFACT
        NUMNP=LN
      NSIZE=NUMNP
        M=LM
      MM=M
      READ UO,(V(I),1=1,26)
      READ 40,(VV(I),1=1,26)
       IF(NLA.GT.O)  PRINT 155,NLA
C
       READ 7,(QU),J=1,12)
CXX»XXXXXXXftXXXX*XX*XXXXX«XXXX»*XXX*«XXX***X«*«**X**«XXX*»«XXX
C
c
c
c
     DATA GROUP II
  READ IN DATA ON THE SPATIAL STRUCTURE

  IF(MQ.LT.0.0001) GO TO 12
  DO 8 1=1,NUMNP
  READ,J,XLOC(J),YLOC(J)
8 CONTINUE
  DO 9 11=1,LM
   CALL ELREAD
9 CONTINUE
  GO TO 15
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
DAT
 100
 200
 300
 400
 500
 600
 700
 800
 900
1000
1100
1200
1300
moo
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
2500
2600
2700
2800
2900
3000
3100
3200
3300
3400
3500
3600
3700
3800
3900
4000
4100
4200
4300
4400
4500
4600
4700
4800
4900
5000
5100
5200
5300
5400
5500
5600
5700
5800
5900
6000
6100
6200
                                     -155-

-------
12      CONTINUE
      READ.KSPACX.KSPACY
       CALL P(KSPACX.KSPACY)
       MQ=ABS(MQ)
       DO 14 J=1,LM
14     NODE
-------
31

32
33
35
C
  DO 32 JJ=1,NMATA
    IFCKTYPE.LT.3) READ.J ,(A(I,J>,1=1,3)
   IF(KTYPE.EQ.5)  READ,J,(A(I,J),1=4,8)
  IFCKTYPE.GT.2)  READ,J , (A(I,J),1=1,8)
   DO 31 1=1 ,8
   A(I,J)=A(I,J)»FF(I)
  PRINT 104,J,(A(I,J),I=1 ,8)
  CONTINUE
  DO 34 K=1,LM
  DO 33 J=1 ,50
   KZX=MAT(K)
  IF(KZX.NE.J) GO TO 33
  WX(K)=A(1,J)
  WY(K)=A(2,J)
  STO(K)=A(3,J)
  PX(K)=A(4,J)
  PY(K)=A(5,J)
  PCX(K)=A(6,J)
  DIFF(K,1)=A(7,J)
  DIFF(K,2)=A(8,J)
  GO TO 34
  CONTINUE
   CONTINUE
  DO 35 J=1,LM
  TPX(J)=WX(J)
  TPY(J)=WY(J)

    PRINT 120
  PRINT VV,(MAT(J),J=1,LM)
       READ 7,(Q(J),J=1,12)
£•»«««*»»«»*»»««*««*«««»»«»*»««»*»««««»»«««»»«««»»««»««*«««««•««*
C
C
C
INPUT DISTRIBUTED SOURCES
        READ.KGEN.KGENH
       IF(KGEN.EQ.I) CALL READER(LM.INFLOW,I)
       IFU.NE.999) STOP INFL
       IF(KGENH.EQ.I) CALL READERUM ,HEAT ,1)
       IFCI.NE.999) STOP HEAT
C«»««**»«*•**»*«**•**»»*«**««*««»«*******«**ftftx«*ft*«««ft*«*ff*»««
C
C  INPUT POINT SOURCES OF WATER AND  HEAT
      READ.LWATER.LHEAT
      IF(LWATER.EQ.O) GO TO 61
      READ,(NWATER(I),AWATER(I),1=1,LWATER)
      PRINT 131,(NWATER(I),AWATER(I),1=1.LWATER)
61    CONTINUE
      IF(LHEAT.EQ.O) GO TO 63
      READ,(NHEAT(I),AHEAT(I),1=1,LHEAT)
      PRINT 132,(NHEAT(I),AHEAT(I),1=1,LHEAT)
63    CONTINUE
C
C
C INPUT LINE SOURCES OF WATER AND  HEAT
       READ.LINEW.LINEH
       IF(LINEW.EQ.O) GO TO 66
       READ,(NLINEW(I,1),NLINEW(I,2),I=1,LINEW)
       READ,(ALINEW(I) ,1=1 .LINEW)
       PRINT 278,(NLINEW(I,1),NLINEW(I,2),ALINEW(I),I=1,LINEW)
66     IF(LINEH.EQ.O) GO TO 68
       READ,(NLINEH(I,1),NLINEH(I,2),I=1,LINEH)
       READ,(ALINEH(I),1=1.LINEH)
       PRINT 279,(NLINEH(I,1),NLINEH(I,2),ALINEH(I),I=1,LINEH)
DAT12600
DAT12700
DAT12800
DAT12900
DAT13000
DAT13100
DAT13200
DAT13300
DAT13400
DAT13500
DAT13600
DAT13700
DAT13800
DAT13900
DAT14000
DAT14100
DAT14200
DAT14300
DAT14400
DAT14500
DAT14600
DAT14700
DAT14800
DAT14900
DAT15000
DAT15100
DAT15200
DAT15300
DAT15400
DAT15500
DAT15600
DAT15700
DAT15800
DAT15900
DAT16000
DAT16100
DAT16200
DAT16300
DAT16400
DAT16500
DAT16600
DAT16700
DAT16800
DAT16900
DAT17000
DAT17100
DAT17200
DAT 17300
DAT17400
DAT17500
DAT17600
DAT17700
DAT17800
DAT17900
DAT18000
DAT18100
DAT18200
DAT18300
DAT18400
DAT18500
DAT18600
DAT18700
DAT18800
DAT18900
                                     -157-

-------
68     CONTINUE
C
C
       READ 7,(QU),J = 1,12)
etc***!****************************************************
C
C       DATA GROUP VI
C
C INPUT INFORMATION NEEDED FOR  A  2-D AREAL VIEW OF THE SYSTEM
       READ.KAREAL
       IF(KAREAL.EQ.O) GO TO 41
       READ.ER.ITER
       CALL READER(LN,BOT,I)
       IFU.NE.999)  STOP BOT
       CALL READER(LN,R1,I)
       IF(I.NE.999) STOP R1  2
10     FORMAT(2(13A6,A2/))
       PRINT 276
       PRINT V,(BOT(I),I=1,LN)
C
41    CONTINUE
       READ 7,(Q(J),J=1,12)
£«*««>**«•*«»«***•«•*•««**««««»«»*»»*»«««»«•***««**«»**«»*»•
C INPUT CONVECTIVE BOUNDARY  INFORMATION
C
C        DATA GROUP VII
       IFUTYPE.LT.3) GO TO 59
        READ, NCONV
      IF(NCONV.EQ.O) GO TO 55
       READ,(NCON(I,1),NCON(I,2),CONV(I),I=1,NCONV)
 55
      READ,(TINF(I),1=1,NCONV)
      CONTINUE
C
C
C INFORMATION NEEDED FOR HEAT OR MASS TRANSFER  ACROSS  A  BOUNDARY
C  WITH A SPECIFIED HEAD
C
       READ.LEL
       IF(LEL.EQ.O) GO TO 56
       READ,(NEL(J,1),NEL(J,2),J=1,LEL)
       READ,(AEL(J,2),J=1,LEL)
       PRINT 157,(NEL(J,1),NEL(J,2),AEL(J,2),J=1,LEL)
       CONTINUE
        PRINT 110,{NCON(I,1),NCON(I,2),CONV(I),1=1,NCONV)
       CONTINUE
       READ 7,(Q(I),1=1,12)
56

59

C
C«
C
C
C
C
71
73
C
   DATA GROUP VIII
INPUT LOCATION OF FLOW BOUNDARIES
   INPUT NODE I AND THEN  THE  BOUNDARY  CODE
  READ,LFLUXW,LFLUXH
  IF(LFLUXW.EQ.O) GO TO 71
  READ,(NFLUXW(I,1),NFLUXW(I,2),I=1,LFLUXW)
  PRINT 133,(NFLUXW(I,1),NFLUXW(I,2),I=1.LFLUXW)
  CONTINUE
  IF(LFLUXH.EQ.O) GO TO 73
  READ,       (NFLUXH(I,1),NFLUXH(I,2),I=1,LFLUXH)
  PRINT 13* i   (NFLUXH(I,1),NFLUXH(I,2),I=1,LFLUXH)
   CONTINUE

   READ 7,(Q(I),J=1,12)
DAT19000
DAT19100
DAT19200
DAT19300
DAT19400
DAT19500
DAT19600
DAT19700
DAT19800
DAT19900
DAT20000
DAT20100
DAT20200
DAT20300
DAT20400
DAT20500
DAT20600
DAT20700
DAT20800
DAT20900
DAT21000
DAT21100
DAT21200
DAT21300
DAT2T400
DAT21500
DAT21600
DAT21700
DAT21800
DAT21900
DAT22000
DAT22100
DAT22200
DAT22300
DAT22400
DAT22500
DAT22600,
DAT22700
DAT22800
DAT22900
DAT23000
DAT23100
DAT23200
DAT23300
DAT23400
DAT23500
DAT23600
DAT23700
DAT23800
DAT23900
DAT24000
DAT24100
DAT24200
DAT24300
DAT2H400
DAT21500
DAT24600
DAT24700
DAT2H800
DAT24900
DAT25000
DAT25100
DAT25200
DAT25300
                                    -158-

-------
c
C FORMAT INFORMATION—INFORMATION PRINTED ON EVERY RUN
102    FORMATC1X//1X.T10,'MATERIAL',T20,'HOR PERM',T32,' VER PERM'.TSO,
     "'STORAGE',T65,'HOR  THERM COND1,T85,'VERT THERM COND',T103,
     "'SPECIFIC  DISPERSION COEFS'/1X,T103,'HEAT')
104   FORMATC1X,T12,G9.4,T22,G9.4,T34,G9.4,T50,G9.4,T68,G9.4,T88,
     «G9.4,T105,G9.4,T115,2G9.4/)
105     FORMATC1X//1X,'PARAMETER  FACTORS—IN ORDER      ',8CG10.6,1X))
110    FORMATC1X,'CONVECTIVE OUT  AND CONDUCTIVE  BOUNDARY INFORMATION'
     *,'         ELEM # —BOUNDARY CODE—TRANSFER  COEFFICIENT'
     * /IX,20(IX,5(14,IX,II,1X,G11.6,2X)/))
120   FORMATC1X//1X,'THE  VALUES IN  THE  TYPE OF MATERIAL  MATRIX  ARE')
131   FORMATC1X//1X,'POINT  SOURCES  OF WATER—NODE # AND  AMOUNT'/IX,
     «5(I4,1X,G12.6,5X))
132   FORMATC1X//1X,'POINT  SOURCES  OF HEAT1/1X,5CI4,1X,G12.6,5X))
133   FORMATC1X//1X,'WATER FLOW BOUNDARIES—ELEM # AND BOUNDARY CODE'
     «10CI4,2X,I1,5X))
134   FORMATC1X//1X,'HEAT FLOW BOUNDARIES—NODE # AND BOUNDARY  CODE'/
     «10CI4,2X,I1,5X))
150   FORMATC1X//1X,1EQUATION 1 TIME STEP',G12.6)
151   FORMATC1X//1X,1EQUATION 2 TIME STEP',G12.6,10X,
     *ITY COEFFICIENT1,G12.6)
155    FORMATC1X//1X,'VELOCITIES ARE BEING CALCULATED
     *EP     ',11)
156    FORMATC1X//1X,'VELOCITIES ARE BEING CALCULATED
     •ME  STEP')
157    FORMATC1X//1X,'INFORMATION FOR MASS OR HEAT TRANSFER ACROSS'     DAT28000
     •,1X,'A SPECIFIED HEAD BOUNDARYV1X,'ELEMENT NUMBER--BOUNDARY CODE'DAT28100
     «,IX,'--TEMPERATURE OR CONCENTRATION OF INCOMING FLUID'            DAT28200
     «/1X,5(2X,I4,1X,I2,1X,2X,G10.5)/)                                   DAT28300
C»»»««»»»»*»»»»«»««»«»«»«*»«»»»*»««*«*»«»»*»**«***»«*«*««»»»«««»«»«*»»*»DAT28400
                    DAT25400
                    DAT25500
                    DAT25600
                    DAT25700
                    DAT25800
                    DAT25900
                    DAT26000
                    DAT26100
                 .1XDAT26200
                    DAT26300
                    DAT26400
                    DAT26500
                    DAT26600
                    DAT26700
                    DAT26800
                 '   DAT26900
                    DAT27000
                    DAT27100
                    DAT27200
                    DAT27300
               CAPACDAT27400
                    DAT27500
             TIME STDAT27600
                    DAT27700
  AT THE PREVIOUS TIDAT27800
                    DAT27900
                                                     'HEAT OR MASS

                                                      AT CURRENT
C
C
C
            THE ROUTINE THAT PRINTS OUT THE INITIAL
             SPECIFIED BY THE INPUT DATA
CONDITIONS
      IF(KPRINT.EQ.O) GO TO 300
      PRINT 199.CTITLECJ),J=1,24)
  199 FORMATC1H1,1X//,20('«',12A6,20('*')/1X,1 OX,'THE UNITS USED ARE',
     *12A6))
      PRINT 210,NUMNP,M
  210  FORMATC1X,'NUMBER OF NODES ',13,'   NUMBER OF ELEMENTS  ',I3//)
      PRINT 211,ALPHA,PCW
  211 FORMATC1X,10X,'TIME STEP',F5.2,' UNITS',10X,'PC=',G10.5)
      PRINT 220
  220 FORMATCIX,'ELEMENT DATA'/1X,'NODE NUMBER',T20,'X-LOCATION',T40,
     •'Y-LOCATION',T60,'SPEC. HEAD',T75,'SPEC.  T',T90,'INITIAL T',
     « T110,'INITIAL HEAD')
      DO 240 J=1,NUMNP
      PRINT 225,J,XLOCCJ),YLOCCJ),HEA(J),HEAD(J),R(J),R1(J)
225   FORMATC1X,T5,I3,T20,G10.5,T40,G10.5,T60,G10.5,T75,G10.5,T90,G10.5
     «  T110.G10.5)
  240 CONTINUE
      PRINT 250
250    FORMATC1X.2X///1X,'ELEMENT PROPERTIES'/1X,
     «T22,'HCY',T32,'STO',T42,'TCX',T52,'TCY',T63,'PC',T72,
     •  '  THE ELEMENT NODES    ',T102,'HEATIN',T112,'WATER  IN')
      DO 270 J=1,M
      PRINT 275,J,WXCJ),WYCJ),STO(J),PX(J),PY(J),PCXCJ),CNOD(LZ,J),LZ=1,DAT31000
     *   4),HEAT(J),INFLOWCJ)                                           DAT31100
                                                   'ELEMENT1,T12,'HCX'
DAT28500
DAT28600
DAT28700
DAT28800
DAT28900
DAT29000
DAT29100
DAT29200
DAT29300
DAT29400
DAT29500
DAT29600
DAT29700
DAT29800
DAT29900
DAT30000
DAT30100
.DAT30200
DAT30300
DAT30400
DAT30500
DAT30600
DAT30700
DAT30800
DAT30900
275

276
278
      FORMATC1X,T4,I3,T10,G8.3,T20,G8.3,T30,G8.3,T40,G8.3,,T50IG8.3,    DAT31200
     «  T60,G8.3,T70,4I3,T100,G8.3,T110,G8.3)                           DAT31300
       FORMATC1X//1X,'ELEVATION OF THE AQUIFER BOTTOM',1X/)             DAT31400
       FORMATCIX,'LINE SOURCES OF WATER--ELEMENT NUMBER.BOUNDARY CODE.RADAT31500
     «TE'/1X,6(I4,1X,I2,1X,G10.5,2X))                                   HAT?,A™
                                     -159-

-------
279    FORMAT(1X,'LINE SOURCES OF  HEAT—ELEMENT NUMBER .BOUNDARY  CODE,  RADAT31700
     *TE'/1X,6(I4,1X,I2,1X,G10.5,2X))                                   DAT31800
  270 CONTINUE                                                         DATS 1900
300    CONTINUE                                                        DAT32000
C ROUTINE TO ORIENTATE THE  MATRIX—INSURES THAT ALL  PROBLEMS ARE ORIENTADAT32100
C   IN THE SAME MANNER                                                 DAT32200
282
284

288

290
294
C
       IF(MQ.EQ.O) GO TO 294
       IF(MQ.EQ.2) GO TO 294
       AC=0
       AB=0
       DO 282 J=1,LN
       AB=AMAX1(XLOC(J),AB)
       AC=AMAX1(YLOC(J),AC)
       IF(MQ.EQ.4) GO TO 288
       DO 284 J=1,LN
       YLOC(J)=AC-YLOC(J)
       IF(MQ.EQ.I) GO TO 294
       CONTINUE
       DO 290 J=1,LN
       XLOC(J)=AB-XLOC(J)
       CONTINUE
                                                                 DAT32300
                                                                 DAT32400
                                                                 DAT32500
                                                                 DAT32600
                                                                 DAT32700
                                                                 DAT32800
                                                                 DAT32900
                                                                 DAT33000
                                                                 DAT33100
                                                                 DAT33200
                                                                 DAT33300
                                                                 DAT33400
                                                                 DAT33500
                                                                 DAT33600
                                                                 DAT33700
                                                                 DAT33800
                                                                 DAT33900
C*« *««««§»« t«»»«f*ci**>«««*»t*«««i*«t««*cft«*««*«t•«•••!•«t«ff«*« ««***«*** DAT34000
54
55

1
8
9

10
11
12
13
RETURN
 SUBROUTINE  ELREAD
 DIMENSION DIGIT(IO),  CHAFU80)

 FORMAT(1X,'PROBLEMS—BAD CHARACTER',IX JA1,3X, •ELEMENT',16)
 FORMATdX,1 PROBLEM—NOT ENOUGH NODAL DATA FOR ELEMENT    ',16)
 READ  1,(CHAR(J),J=1,80)
 FORMAT(80A1)
 KC=0
 KEr4
 KS=2
 J=0
 KL=0
 NUH=0
 J=J+1
 IF(CHAR(J).EQ.'  ') GO TO 9
 1=0
 1=1+1
 IF(I.LE.IO) GO TO 12
 PRINT 54,CHAR(J),L
 STOP
 IF(CHAR(J).NE.DIGIT(D) GO TO 11
 NUM=10*NUM+I-1
 J=J+1
 IF(CHAR(J).EQ.'  ') GO TO
 IF(CHAR(J).EQ.'«») GO TO
 GO TO 10
 KL=KL+1
 IFUL.EQ
                                13
                                14
                                                                        DAT34100
                                                                        DAT34200
                                                                        DAT34300
                                                                        DAT34400
                                                                        DAT34500
                                                                        DAT34600
                                                                        DAT34700
                                                                        DAT34800
                                                                        DAT34900
                                                                        DAT35000
                                                                        DAT35100
                                                                        DAT35200
                                                                        DAT35300
                                                                        DAT354CO
                                                                        DAT35500
                                                                        DAT35600
                                                                        DAT35700
                                                                        DAT35800
                                                                        DAT35900
                                                                        DAT36000
                                                                        DAT36100
                                                                        DAT36200
                                                                        DAT36300
                                                                        DAT36UOO
                                                                        DAT36500
                                                                        DAT36600
                                                                        DAT36700
 IF(KL.EQ
 KL=KL+1
                1)
                I)
                   L=NUM
                   GO TO 8
14
       IF(KC.EQ.S)  GO TO 15
       NOD(KC,L)=NUM
       IF(KS.EQ.O)  KErKE-t-2
       IF(KS.EQ.I)  KE=KE+1
       KS=0
       GO TO 8
       K£=KE+1
       KS=KS+1
DAT36900
DAT37000
DAT37100
DAT37200
DAT37300
DAT37400
DAT37500
DAT37600
DAT37700
DAT37800
DAT37900
DAT38000
                                    -160-

-------
15



16


C
C'
C
 1
 2
  3
  4
 20

 10
 11
NOD(KE,L)=NUM
IFU.LT.65) GO  TO  8
PRIMT 55,L
STOP
MTM=M
DO  16 1=5,12
IF(NOD(I,L).EQ.O)  GO TO 16
MTM=MTM+1
CONTINUE
NODE(L)=MTM
RETURN
       SUBROUTINE P(KX,KY)
       DIMENSION D(50),  B(50),C(50)
      J = 0
       MN = 0
       IF(KX.LT.O) MN=1
       KX=ABS(KX)
      KXA=KX-1
      KYA=KY-1
      DO 2 L=1 ,KXA
      DO 1 K=1 ,KYA
NOD(3,J>=L+J+KYA
NODC1 ,J)=L+J
NOD(2,J)=L+J+KY
CONTINUE
CONTINUE
 READ,(B(J),J=1,KX)
  PRINT 10
 PRINT ,(B(J),J=1,KX)
READ,(C(J),J=1,KY)
 PRINT 11
 PRINT, (C(J),J=1TKY)
J=0
DO 4 L=1 ,KX
DO 3 K=1 ,KY
YLOC(J)=C(K)
XLOC(J)=B(L)
CONTINUE
CONTINUE
 IF(MN.EQ.O) RETURN
 READ.MLD
 MLB=1
 IF(MLD.EO.O) MLB=KY
 READ, (D(J),J=1,KX)
 KXX=LN-KY+1
 DO 20 K=MLB,KY
 L=0
 DO 20 JJ=1,KXX,KY
 J=JJ+K-1
 L=L+1
 YLOC(J)=YLOC(J)+D(L)
RETURN
 FORMAT(1X/1X,'THE  X-SPACING
 FORMAT(1X/1X,'THE  Y-SPACING
 RETURN
 SUBROUTINE  READERUZ ,Q ,1)
                              IS')
                              IS')
DAT38100
DAT38200
DAT38300
DAT38MOO
DAT38500
DAT38600
DAT38700
DAT38800
DAT38900
DAT39000
DAT39100
DAT39200
DAT39300
DAT39400
DAT39500
DAT39600
DAT39700
DAT39800
DAT39900
DAT40000
DAT40100
DAT40200
DAT40300
DAT40400
DAT40500
DAT40600
DAT40700
DAT40800
DAT40900
DAT41000
DAT41100
DAT41200
DAT41300
DAT41400
DAT41500
DAT41600
DAT41700
DAT41800
DAT41900
DAT42000
DAT42100
DAT42200
DAT42300
DAT42400
DAT42500
DAT42600
DAT42700
DAT42800
DAT42900
DAT43000
DAT43100
DAT43200
DAT43300
DAT43400
 DAT43500
 DAT43600
 DAT43700
DAT43800
 DAT43900
 DAT44000
 DAT44100
 DAT44200
                                      -161-

-------
DIMENSION  Q(LZ)                                                  DAT1U300
READ,QI.IZ,FACT                                                  DAT44400
DO 1 IPs'! ,LZ                                                    DAT4M500
Q(IP)=QI                                                        DATJ44600
IF(IZ.GE.O)  GO TO 2                                             DAT44700
READ,(Q(IP),IP=1,LZ),I                                           DAT44800
GO TO 3                                                         DAT4U900
1=999                                                           DATM5000
IF(IZ.EQ.O)  GO TO 3                                             DATU5100
READ,(J,Q(J),IP=1,IZ)                                            DATU5200
CONTINUE                                                        DAT45300
DO 5 IP=1,LZ                                                    DAT45400
Q(IP)=Q(IP)«FACT                                                DATM5500
RETURN                                                          DAT45600
END                                                             DAT45700
                             -162-

-------
 *«*»»«»«»»«*»«»««
C
C
C
                                                                        STR
50
                                                                        STR
      THIS  ROUTINE  ASSEMBLES THE STRUCTURE MATRICIES AND PREFORMS       STR
         A  FORWARD  REDUCTION OF THE STIFNESS MATRIX BY GAUSS ELIMINATIONSTR
                                                                        STR
      SUBROUTINE  STRUCK LN ,LM ,MBAND ,MP ,R , S,T ,G ,GG ,RI .HEAD ,XLOC,          STR
     *YLOC, NOD, NODE, CFACT.TT, ALPHA, KBOUND)                              STR
       INTEGER  TT                                                       STR
       REAL  JAC                                                         STR
      DOUBLE PRECISION  AA,SE,TE,S,T                                     STR
      COMMON/AAA/XL(4),YL(4) , DETJAC, JAC( 4 ,4 )/A1/M ,MM,NUMNP .NSIZE        STR
       COMMON/AB/AAC6) , W( 4)/ABB/SE( 12 , 1 2) ,TE(12,12)                     STR
       COMMON/AC/RE(12) , GE( 12)/ACC/KK( 12) ,K1 ,K2,K3,K4                   STR
       COMMON/HH/LFLOW .LON/HM/PXX , PYY , PXY ,KAD/AM/KSYM , PCT               STR
      DIMENSION XLOC(LN) ,YLOC(LN) , NOD( 1 2 , LM) , R(LN)  .S(LN,MP)              STR
      DIMENSION T(LN,MP) ,G(LN) ,GG(LN) ,RI(LN) ,NODE(LM) ,HEAD(LN)          STR
      LFLOW=1                                                           STR
      AA(1)=-(1/3)«*-5                                                   STR
                                                                        STR
                                                                        STR
                                                                        STR
                                                                        STR
                                                                        STR
                                                                        STR
                                                                        STR
      AA(2)=-AA(1)
DO 50 J=1,LN
GG( J)=0.0
 DO 50 1=1 ,MP
S(J,I)=0.0
T(J,I)=0.0
C PRINT MESSAGE IF DISPERSION ROUTINE IS USED                           STR
       IF(KAD.EQ.I)  PRINT 10                                            STR
C                                                                       STR
Q*ff««««»«»*«>«»*«*«»«*»««*»«»*«**«
-------
  1«»0 CONTINUE                                                         SIR  6500
  150  CONTINUE                                                        STR  6600
C                                                                      STR  6700
C                                                                      STR  6800
C*»t»*««»f»t«»»««•*»»•*»«»•»»»«»«»«»»»««««««»•t•«»»»«»**«»«»«»»*«««««»»#STR  5900
C                                                                      STR  7000
C—ROUTINE FOR SPECIFIED HEAD  BOUNDARY  CONDITIONS                       STR  7100
      DO 200 N=1,NUHNP                                                 STR  7200
      IF(HEADCN).EQ.O) GO TO 200                                        STR  7300
      IF(KSYM.EQ.O)  S(N,1)=S(N,1)+CFACT                                STR  7400
      IF(KSYM.GT.O)  S(N,MBAND)=S(N,MBAND)+CFACT                         STR  7500
200     CONTINUE                                                       STR  7600
C                                                                      STR  7700
£*••«»•»*•**••*****•«•**«»•*•**•**•*•••»«•*«***««*«»**«•»**»***«**«*«**«sjR  7800
C                                                                      STR  7900
C   CALL THE ROUTINE THAT CHECKS  FOR  THE  STABILITY  OF THE  CHOSEN  TIME STSTR  8000
       KSMrKSYM                                                        STR  8100
      IFUON.EQ.99.AND.TT.EQ.1)  CALL  EIGEN(LN ,MBAND,MP,KSM ,CFACT ,S,T ,RI ,STR  8200
     •HEAD)
C GAUSS ELIMINATION—FORWARD REDUCTION  OF STIFNESS  MATRIX               STR  8400
C                                                '                      STR  8500
        IF(KSYM.EQ.O) CALL SOLVEUN MBAND ,S,T,R,RI)                     STR  8600
        IF(KSYH.NE.O) CALL ASOLVE(S,R,RI,LN,MBAND-1,LN,MBAND»2-1,T)     STR  8700
C                                                                      STR  8800
C                                                                      STR  8900
10     FORHAT(1X/1X,'THE DISPERSION ROUTINE IS BEING USED  )             STR  9000
20     FORHAT(1X/1X,'MAX. BANDWIDTH EXCEEDED IN  ELEMENT',II,            STR  9100
     »  'WHERE BANDWIDTH IS  •,!«)                                      STR  9200
      RETURN                                                           STR  9300
C                                                                      STR  9400
C                                                                      STR  9500
       END                                                             STR  9600
                                     -164-

-------
£»**«*»»«***«««««**«****«««*« ft* ****•««*»«««»««* *»««!« ««»**«««•««»«««*«

C*                                                                      GAU
C   THIS ROUTINE FORMS ELEMENT STIFFNESS MATRIX BY GAUSS QUADRATURE    'GAU
C                                                                       GAU
       SU BROUTINE GAUSSC LLL , LM , LN ,XLOC , YLOC , NOD , PY , PX , HEAT , PCX , ALPHA , KBOGAU
     *UND ,TT, Z , NCON , CONV ,ALOC ,NODE ,CBAL ,AY , A2)                          GAU
      REAL JAC
      INTEGER VE, TT ,Z , AY( 4) ,AZ(4)
       DOUBLE PRECISION AA ,NOT,NXI ,NET,SE ,TE ,BB ,DUM1 ,DUM2 , DUM3
       COMMON/AA/NXI(12),NET(12),NOT(12)
       COMMON/AAA/XLC4) ,YL(4) , DETJAC , JAC( 4 ,4)
       COMMON/AB/AA(6) , W( 4) /ABB/SE( 12 , 12) , TE( 12, 12)
       COMMON/AC/RE(12),GE(12)/ACC/KK(12),K1,K2,K3,K4
      COMMON/HH/LFLOW,LON/HM/PXX,PYY,PXY.KAD/ME/MEQ
      COMMON/CON/MC1,MC2,NCONV/AS/BBC2,12)
                DIMENSION  NCON(Z,3),CONV(Z),ALOC(Z)
      COMMON/CON1/HE(4),HEE(4,4)/AM/NLA,PCH/AT/XIH4) ,ETI(U)
       COMMON/ATHICK/ASIZE.NTHICK,THICK,ERROR
       DIMENSION VE(4),AVY(4),AVX<4),VEL(4)
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
      DIMENSION           XLOC(LN),YLOC(LN),PY(LM),PX(LM),HEAT(LM),PCXCLGAU
     *M),CBAL(Z,2),AG(6),NODE(LM),NOD(12.LM)                            GAU
       DATA XII/-1.,1.,1.,-1./                                          GAU
       DATA VE/1 ,4,2,3/                                                 GAU
       DATA ETI/-1.,-1.,1 .,1 ./                                          GAU
       DATA W/.3478549,.6521451,.6521451,-347S549/                      GAU
       DATA AG/-.5773503,.5773503,-.8611363,-.3399810,.3399810,.8611363/GAU
      RETURN                                                            GAU
£«»««»«»•«»»*»«»»««*»««««««»**»««»*«»««**«**«««««««»»»«»«««»«««»*»*«««««GA[j
C                                                                       GAU
       ENTRY DERIVE(LLL,M4)                                             GAU
C                                                                       GAU
C                                                                       GAU
£*«>*«««*«»«*«»*»«**««»*«ft»«*ft««»*«***«*««**««»*««*ft«ft*****ft**»ft*       GAU
C                                                                       GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
                                                                        GAU
 20
 30
     CALL CORD(LLL)
    T- 1  0
    CALL PEE(T,LLL)
   DO 10 K=1 ,12
   GE(K)=0.0
   DO 10 L=1 ,12
   TE(K,L)=0.0
10 SE(K,L)=0
  START THE QUADRATURE LOOP
    NP=2
    IFCM4.GT.M) NP=H
    KVE=0
     DO 200 II=1,NP
    DO 200 JJ=1,NP
    KVE=KVE+1
    IFCNP-EQ.**) GO TO 20
    XI=AG(JJ)
    YI=AG(II)
    GO TO 30
    XI=AG(JJ+2)
    YI=AG(II+2)
    IF(LFLOW.EQ.O) XI=0.0
    IF(LFLOW.EQ.O) YI=0.0
    CONTINUE
    IF(M14.GT.M) CALL SHAPECLLL ,MU ,XI ,YI ,XLOC, YLOC.LN ,LM,NOD)
     IFCM4.EQ.4) CALL SHAPE1(II,JJ)
c»»«»»«*»«»»«»««»«*»««»**««»««»»»»«•»«»»»«**»««
 100
 200
 300
 400
 500
 600
 700
 800
 900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
2500
2600
2700
2800
2900
3000
3100
3200
3300
3400
3500
3600
3700
3800
3900
4000
4100
4200
4300
4400
4500
4600
4700
4800
4900
5000
5100
5200
5300
5400
5500
5600
5700
5800
5900
6000
6100
6200
                                     -165-

-------
175    CONTINUE
       THICK=1.0
       IF(NTHICK,EQ.O)  GO  TO  176
       XX=(XLOC(K2)+XLOC(K1))/2
       XX=ABS(XX)
       THICK=XX*ASIZE
      YY=(YLOC(K3)+YLOC(K1))/2
      YY=ABS(YY)
      IF(ASIZE.LT.O) THICK=-YY«ASIZE
176      DUM1=W(II)«W(JJ)«DETJAC»THICK«T
       IF(M4.EQ.4)  DUM1=DETJAC«THICK«T
C»«*»»»«»«*»»»»«»«««««»*»»»»»«*»»»»«»«»*»«»««««
C
       IF(LFLOW.EQ.O)  DETJAC=DETJAC«T«THICK
       IF(LFLOW.EQ.O)  RETURN
C VELEOCITY AND DISPERSION CALCULATIONS
C
       VX=0.0
       VY=0.0
       IF(NLA.EQ.O) GO TO  178
        IFCNLA.EQ.1) CALL  VELO(LLL,VX,VY)
        IFCNLA.EQ.2) CALL  VCENTCLLL,VX,VY)
       IF(KAD.EQ.I) CALL HECD(LLL,VX,VY)
C
       IFCKVE.GT.4) GO TO  178
       KEE=VE(KVE)
       AVY(KEE)=VY
       AVXCKEE)=VX
178    CONTINUE
C
£ft««f«t«»lt«*«tf*l«lf»«*f*««»f***l««*««>l*t««»ft
C
c
180
      DO 200 NROW=1,M4
      GE(NROW)=GE(NROW)+NOT(NROW)*HEAT(LLL)«DUM1
      DO 200 NCOL=1,M4
        IF(NLA.EQ.O) GO TO 180
        DUM2=BB(1,NROW)»NOT(NCOL)»PCH»VX+BB(2,NROW)*NOT(NCOL)«PCH«VY
        SE{NROW,NCOL)=SE(NROW,NCOL)+DUN2*DUM1
        CONTINUE
      DUM2=BB(1,NROW)*BB{1,NCOL)«(PX(LLL)+PXX)
      S£(NROW,NCOL)=SE(NROW,NCOL)+DUM1*DUM2
      DUM2=BB(2,NROW)«BB(2,NCOL)*(PYULL)
      SE(NROH,NCOL)=SE(NROW,NCOL)+DUH1«DUM2
                                                                       GAU
                                                                       GAU
                                                                            8100
                                                                            8200
                                                                            8300
                                                                            8400
GAU 6300
GAU 6400
GAU 6500
GAU 6600
GAU 6700
GAU 6800
GAU 6900
GAU 7000
    7100
    7200
GAU 7300
GAU 7400
GAU 7500
GAU 7600
GAU 7700
GAU 7800
GAU 7900
GAU 8000
GAU
GAU
GAU
GAU
GAU 8500
GAU 8600
GAU 8700
GAU 8800
GAU 8900
GAU 9000
GAU 9100
GAU 9200
GAU 9300
GAU 9400
GAU 9500
GAU 9600
GAU 9700
GAU 9800
GAU 9900
GAU10000
GAU10100
GAU10200
GAU10300
GAU10400
GAU10500
GAU10600
GAU10700
GAU10800
      THESE NEST TWO LINES ADD IN  TRANSVERSE  TERMS OF CONDUCTIVITY TENSOGAU10900
       DUM2=BB(1,NROW)«BB(2,NCOL)»PXY+BB(2,NROW)*BB(1,NCOL)«PXY
      SE(NROW,NCOL)=SE(NRO¥,NCOL)+DUM1*DUM2
      IF(TT.EQ.O) GO TO 200
      DUM3=NOT(NROW)«NOT(NCOL)«PCXULL)
       IF(NLA.EQ.O) DUM3=DUM3/T
       TE(NROW,NCOL) = DUM3»DUMUTE(NROW,NCOL)
200     CONTINUE
C
£•**««•*»*•»Ift***»*ft««*«»ftfft*fftt«•••*••«*•*•*««*»***»»«****
C
c
     THE CONVECTIVE  BOUNDARIES ARE HANDLED

      IF(NCONV.EQ.O) GO TO 400
       IF(MEQ.EQ.I)  GO  TO 400
       DO 301  Isl.NCONV
      11=1
       IF(LLL.EQ.NCON(I,1))  CALL CONVEC
 300   CONTINUE
GAU11000
GAU11100
GAU11200
GAU11300
GAU11400
GAU11500
GAU11600
GAU11700
GAU11800
GAU11900
GAU12000
GAU12100
GAU12200
GAU12300
GAU12400
GAU12500
GAU12600
                                    -166-

-------
301
400
C
C
C
 450
800
C
C
 505
 C
 C
  550
 •«

 :«
 c
 600

 601

 602

 603

 604
 605
 C
 CONTINUE
 CONTINUE
 VYY=(AVY(1) + AVY(2)-t-AVY(
 IF(KBOUND.EQ.99)  RETURN
 IF(TT.EQ.O)  RETURN
  t>«*«*««»1
   CRANK-NICOLSON  METHOD FOR TREATING TRANSIENT CONDITIONS
 DO 450  NROW=1,M4
 GE(NROW)=GE(NROW)*ALPHA
 DO 450  NCOL=1,M4
 DUM4=SE(NROW,NCOL>
 SECNROW,NCOL)=ALPHA/2*SE(NROW,NCOU+TECNROW,NCOL)
 TECNROW,NCOL)=TE(NROW,NCOL)-ALPHA/2*DUM4
 CONTINUE
  CONTINUE
**ft«*«******«***»*********t*«»*«*«*«»*****«**«****
  RETURN
  ROUTINE THAT ADDS IN CONVECTIVE BOUNDARY TERMS

  SUBROUTINE CONVEC
  LA=NCON(I,2)
  MC1=AY(LA)
  MC2=AZ(LA)
  A=XL(MC1)-XL(MC2)
  MC3=MC1+MC2
  IFCMC3.EQ.5) A=YL(MC1)-YL(MC2)
  A=ABS(A)
  DO 505 J=1,4
  HE(J)=0.0
  HE(MC1)=.5
  HE(MC2)=.5
  ALOC(II)=A«THICK*T
  CO=CONV(II)
  IFCCO.LT.0.00001) GO TO 600
  CBAL(II,1)=ALPHA«CO«THICK*T*A
 DO 550 K=1,4
 DO 550   J=1,4
  SE(K,J)=HE(J)*HE(K)»CO«THICK*T«A+SE(K,J)

  RETURN
   CONTINUE
   GO  TO  (603,601,601,603),LA
   DO  602 J=1,4
   IF(MC3.EQ.5)  VEL(J)=AVX(J)
   IFCMC3-NE.5)  VEL(J)=AVY(J)
   GO  TO  605
   DO  604 J=1,4
   IF(MC3-EQ.5)  VEL(J)=-AVX(J)
   IF(MC3.NE.5)  VEL(J)=-AVY(J)
   CONTINUE

   C0=-C0
   NCO=CO
   VT=VEL(MC1)/2+VEL(MC2)/2
   IF(NCO.LT.995.0R.NCO.GT.1000) GO TO 700
   IF(NCO.EQ.999) CO=VT«PCH*ERROR
   IF(NCO.EQ.996) C0=(-VT+VYY)«PCH*ERROR
GAU12700
GAU12800
GAU12900
GAU13000
GAU13100
GAU13200
GAU13300
GAU13400
GAU13500
GAU13600
GAU13700
GAU13800
GAU13900
GAU14000
GAU14100
GAU14200
GAU14300
GAU14400
GAU14500
GAU14600
GAU14700
GAU14800
GAU14900
GAU15000
GAU15100
GAU15200
GAU15300
GAU15400
GAU15500
GAU15600
GAU15700
GAU15800
GAU15900
GAU16000
GAU16100
GAU16200
GAU16300
GAU16400
GAU16500
GAU16600
 GAU16700
 GAU16800
 GAU16900
 GAU17000
 GAU17100
 GAU17200
 GAU17300
 GAU17400
 GAU17500
 GAU17600
 GAU17700
 GAU17800
 GAU17900
 GAU18000
 GAU18100
 GAU18200
 GAU18300
 GAU18400
 GAU18500
 GAU18600
 GAU18700
 GAU18800
 GAU18900
                                      -167-

-------
 650
IF(NCO.EQ.997)  CO=ERROR
 CBAL(II,1)=ALPHA«CO«THICK»T«A
DO 650  !(=•!,«
DO 650   J=1,4
 SE(K,J)=HE(J)«HE(K)»CO»THICK»T«A+SE(K,J)
  RETURN
£»•***»»»*•**•*•«*••««*»*««»****«««•**•«««•«««*«*•*•«»*•
c
700    CONTINUE
C CALCULATE HEAT TRANSFER OUT WITH FLOWING WATER
C
       CBALCII,1)=ALPHA«VT«PCH«THICK*T«A
C
       DUM1=1-AA(1)
       DUM2=1-AA(2)
       HE(MC1) = DUM1M
       HE(MC2)=DUM2/4
       VELL=VEL(MC1)
       DO 720 K=1,4
       DO 720 J=1,4
720    SE
-------
c«*«««*«*«««*»*«t*«««»»«**«««»*******««»****»*«**»
       SUBROUTINE SHAPE1(II,JJ)
      REAL JAC

       DOUBLE PRECISION AA.NOT ,NXI,NET,SE,TE,BB ,DUM1,DUM2
       COMMON/AA/NXI(12),NET(12),NOT(12)
       COMMON/AAA/XL(4),YL(4),DETJAC,JAC(4,4)
       COMMON/AB/AA(6),W(4)/ABB/SE(12,12),TE(12,12)
       COMMON/AC/RE(12),GE(12)/ACC/KK(12),K1,K2,K3,K4
      COMMON/HH/LFLOW,LON/HM/PXX,PYY,PXY,KAD
      COMMON/AT/XII(4),ETI(4)/CON/MC1,MC2,NCONV/AS/BB(2,12)
    COMPUTE THE SHAPE FUNCTIONS AND THEIR DERIVATIVES
      DO 100 1=1 ,4

      DUM2=(l!+ETI(I)»AA(JJ))*.25
         NXI(I)=XII(I)«DUM2
      NET(I)=ETI(I)*DUM1
       NOT(I)=4.«DUM1«DUM2
   100  CONTINUE
 C   COMPUTE  JACOBIAN,  AND  ITS DETERMINANT
       DO  150 1=1 ,2
       DO  150 J=1 ,2
   150  JAC(I,J)=0.0
       DO  160 1=1,4
       JAC(1,1) = JAC(1,1)-»-NXI(I)*XL(I)
        JAC(1,2)=JAC(1,2)+NXI(I)*YL(I)
       JAC(2,1)=JAC(2,1)*NET(I)*XL(I)
       JAC(2,2)=JAC(2,2)+NET(I)*YL(I)
   160   CONTINUE
       DETJAC=JAC(1,1)*JAC(2,2)-JAC(2,1)*JAC(1,2)
       DUM1=JAC(1,1)/DETJAC
       JAC(1,1)=JAC(2,2)/DETJAC
       JAC(1,2)=-JAC(1,2)/DETJAC
       JAC(2,1)=-JAC(2,1)/DETJAC
       JAC(2,2)=DUM1
         DO 170  L=1,4
       BB(1,L)=JAC(1,1)«NXI(L)+JAC(1,2)»NET(L)
       BB(2,L)=JAC(2,1) »NXI(L)+JAC(2,2)*NET(L)
 170     CONTINUE
        RETURN
        END
SHI
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SH1
SHI
SH1
SH1
SH1
SH1
SH1
SH1
 SH1
 SH1
 SH1
 SH1
 SH1
 SH1
 SHI
 SHI
 SH1
 SH1
 SHI
 SHI
 SHI
 SHI
 SHI
 100
 200
 300
 400
 500
 600
 700
 800
 900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
2500
2600
2700
2800
2900
 3000
 3100
 3200
 3300
 3400
 3500
 3600
 3700
 3800
 3900
 4000
 4100
 4200
 4300
 4400
                                       -169-

-------

c
c
c
c
c
c
c
c
c
c
c
c
c
c
c




c



c
c
c




c
c
c




c








c
c











c


SUBROUTINE SHAPEU ,M,XI ,YI ,X ,Y ,NN ,NE , IN)
«*««»»«*»«***»*«*••««««««*****«««««*«««<»*»«

XI/YI THE GAUSS POINTS
X/Y THE X AND Y LOCATIONS OF THE ELEMENT NODES
NN NUMBER OF NODES
NE NUMBER OF ELEMENTS
IN ARRAY REFERED TO IN MAIN PROGRAM AS NOD
M REFERED TO IN MAIN PROGRAM AS Ml , NUMBER OF. NODES IN THE ELEMENT[
DET IS THE DETERMINENT, ONE QUARTER OF THE AREA OF THE ELEMENT


CALLED FROM MNPGM
PURPOSE: TO COMPUTE BASIS FUNCTIONS
ORIGINALLY PROGRAMED BY EMIL 0. FRIND

DOUBLE PRECISION DFX.DFY ,FF, BB
DIMENSION X(NN), Y(NN) , IN(12,NE)
DIMENSION ALF(4), DAX(4), DAY(4), BTX(4), BTYU) , DBX(4),
1 DBYCO
f
COMMON/AA/DFX( 12) , DFY( 12) , FF( 12)
«/AAA/ZZZCO,ZZZZZ(4),DET,DJAC(4,4)
COMMON/AS/BB(2,12)


XI/YI - LOCAL COORDINATES OF THE INTEGRATION POINTS
XIU1.-XI
XI2=1.+XI
YIIsl.-YI
YI2=1.+YI

CORNER NODE SHAPE FUNCTIONS, BASIC PART
ALF - ALPHA PART OF SHAPE FUNCTION
ALF(1)=.25«XI1«YI1
ALF(2)=.25*XI2«YI1
ALF(3)=.25*XI2«YI2
ALF(1)=.25«XI1«YI2
DAX/DAY - X- AND Y-DERIVATIVE OF ALPHA PART
DAX(1)=-.25«YI1
DAX(2)=.25»YI1
DAX(3)=.25«YI2
DAX(i|) = -.25«YI2
DAY(1)=-.25«XI1
DAY(2)=-.25«XI2
DAY(3)=.25*XI2
DAY(M)=.25«XI1

CORNER NODE SHAPE FUNCTIONS, SIDE-DEPENDENT PART
XQ1=XI-.5
XQ2=-XI-.5
YQ1=YI-.5
YQ2=-YI-.5
XC1=1.125*XI»XI-.625
XC2=2.25«XI
YC1=1.125«YI»YI-.625
YC2=2.25»YI
J1 = 1
J2=2
J3=5
FOR BETA X PART (BTX) OF SHAPE FUNCTION
DO 50 J=1,2
IF (IN(J3,D.EQ.O) GO TO 10
SHA 100
SHA 200
SHA 300
SHA 100
SHA 500
SHA 600
SHA 700
SHA 800
SHA 900
SHA 1000
SHA 1 100
SHA 1200
SHA 1300
SHA 1400
SHA 1500

SHA 1700
SHA 1800
SHA 1900
SHA 2000
SHA 2100
SHA 2200
SHA 2300
SHA 2400

SHA 2600
SHA 2700
SHA 2800
SHA 2900
SHA 3000
SHA 3100
SHA 3200
SHA 3300
SHA 3400
SHA 3500
SHA 3600
SHA 3700
SHA 3800
SHA 3900
SHA 4000
SHA 4100
SHA 4200
SHA 4300
SHA 4400
SHA 4500
SHA 4600
SHA 4700
SHA 4800
SHA 4900
SHA 5000
SHA 5100
SHA 5200
SHA 5300
SHA 5400
SHA 5500
SHA 5600
SHA 5700
SHA 5800
SHA 5900
SHA 6000
SHA 6100
SHA 6200
SHA 6^00
-170-

-------
   IF (INCJ3+1,L).EQ.O) GO TO 20
   GO TO 30
10 CONTINUE
   LINEAR . .  .
   BTX(J1)=.5
   BTX(J2)=.5
   DBX -.BETA  X  DERIVATIVE
   DBX(J1)=0.
   DBX(J2)=0.
   GO TO 40
20 CONTINUE
   QUADRATIC .  .  .
   BTX(J1)=XQ2
   BTX(J2)=XQ1
   DBX(J1)=-1.
   DBX(J2)=1.
   GO TO 40
30 CONTINUE
   CUBIC .  . .
   BTX(J1)=XC1
   BTX(J2)=XC1
   DBX(J1)=XC2
   DBX(J2)=XC2
40 CONTINUE
   J1 = 4
   J2=3
   J3=9
50 CONTINUE
   J1 = 2
   J2=3
   J3=7
   FOR BETA Y PART  (BTY)
   DO  100  J=1,2
   IF  (IN(J3,L).EQ.O)  GO  TO  60
   IF  (IN(J3+1 ,D .EQ.O) GO TO 70
   GO TO 80
60 CONTINUE
   LINEAR  . . .
   BTY(J1)=.5
   BTY(J2)=.5
   DBY - BETA Y DERIVATIVE
   DBY(J1)=0.
   DBY(J2)=0.
   GO  TO 90
70 CONTINUE
   QUADRATIC  .  .  .
   BTY(J1)=YQ2
   BTY(J2)=YQ1
   DBY(J1) = -1 .
   DBY(J2) = 1 .
   GO  TO  90
80 CONTINUE
   CUBIC  .  .  .
   BTY(J1)=YC1
   BTY(J2)=YC1
   DBY(J1)=YC2
   DBY(J2)=YC2
 90 CONTINUE
   J1=1
   J2=4
   J3=11
100 CONTINUE
OF SHAPE FUNCTION
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
-SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
SHA
   6400
   6500
   6600
   6700
   6800
   6900
   7000
   7100
   7200
   7300
   7400
   7500
   7600
   7700
   7800
   7900
   8000
   8100
   8200
   8300
   8400
   8500
   8600
   8700
   8800
   8900
   9000
   9100
   9200
   9300
   9400
   9500
   9600
   9700
   9800
   9900
SHA10000
SHA10100
SHA10200
SHA10300
SHA10400
SHA10500
SHA10600
SHA10700
SHA10800
SHA10900
SHA11000
SHA11100
SHA11200
SHA11300
SHA11400
SHA11500
SHA11600
SHA11700
SHA11800
SHA11900
SHA12000
SHA12100
SHA12200
SHA12300
SHA12400
SHA12500
SHA12600
                                   -171-

-------
c
c
c
    SHAPE FUNCTION  DERIVATIVE MATRIX  -  CORNER  NODES
    DO 110 J = 1,4
    DFX(J) = DAX(J)«(BTX(J) + BTY(J»+DBX(J)»ALF(J)
    DFY(J)=DAY(J)«(BTX(J)+BTY
-------
  200
  210
  220
  230
  2UO
c
c
FF( J)= .28125*XEQ*XE1*YI2
DFY(J)=.28125»XEQ*XE1
IF (IN(11 ,L) .EQ.O) GO TO  230
IF (IN(12,L).EQ.O) GO TO  210
GO TO 220
J = J+1
DFX( J)=-.5*YEQ
DFY(J)=-YI*XI1
FF(J)=.5*XI1*YEQ
GO TO 230
J = J+1
DFX(J)=-.28125*YEQ»YE2
DFY( J)= .28125*XI1*(3.*YEQ-2.*YI*YE2)
FF(J)=.28125*XI1*YEQ*YE2
J = J + 1
DFX( J)=-.28125*YEQ*YE1
DFY(J)=-.28125*XI1*(3.*YEQ+2.«YI«YE1)
FF( J)= .28125*YEQ*YE1*XI1
CONTINUE
CONTINUE

JACOBIAN
SUM1=0.
SUM2=0.
SUM3=0.
SUMM=0.
      DO 260 1=1 ,M
  250 K=K+1
      IF (IN(K.L).EQ.O) GO  TO  250
      KI=IN(K,L)
      SUM1 = SUMUDFX(I)«X(KI)
      SUM2=SUM2-t-DFX(I)*Y(KI)
      SUM3=SUM3>DFY(I)«X(KI)
      SUM1=SUM4+DFY(I)»Y(KI)
  260 CONTINUE
      DET=SUM1*SUM4-SUM2*SUM3
      DET1=1./DET
      C11=DET1*SUM4
      C12=-DET1*SUM2
      C21=-DET1»SUM3
      C22=DET1*SUM1

                DJAC(1,1)=C11
                DJAC(1,2)=C12
                DJAC(2,1)=C21
                DJAC(2,2)=C22
      SHAPE  FUNCTION  DERIVATIVES - GLOBAL
      DO  270 J=1,M
      BB(1,J) = C11»DFX(J) + C12«DFYU)
      BB(2,J)=C21*DFX(J)+C22«DFY(J)
   270 CONTINUE
      RETURN
      END
 SHA19000
 SHA19100
.SHA19200
 SHA19300
 SHA19100
 SHA19500
 SHA19600
 SHA19700
 SHA19800
 SHA19900
 SHA20000
 SHA20100
 SHA20200
 SHA20300
 SHA20100
 SHA20500
 SHA20600
 SHA20700
 SHA20800
 SHA20900
 SHA21000
 SHA21100
 SHA21200
 SHA21300
 SHA21400
 SHA21500
 SHA21600
 SHA21700
 SHA21800
 SHA21900
 SHA22000
 SHA22100
 SHA22200
 SHA22300
 SHA22UOO
 SHA22500
 SHA22600
 SHA22700
  SHA22800
  SHA22900
  SHA23000
  SHA23100
  SHA23200
  SHA23300
  SHA23'»00
  SHA23500
  SHA23600
  SHA23700
  SHA23800
  SHA23900
                                                                    SHA24100
                                                                    SHA24200
                                                                    SHA2H300
                                      -173-

-------
 c
 c
 c
 c
 c
 c
 c
 c
 c
 c
 c
c
c
                                                                        LOA
      THIS ROUTINE COMPUTES  THE  COLUMN MATRIX R BY ADDING IN THE VARIOUS LOA
         SOURCES AND SINKS AND THEN  SOLVES FOR THE UNKNOWN COLUMN MATRIX LOA
                                                                        LOA
                                                                        LOA
       SUBROUTINE LOAD(LM,LN,MBAND,NOD,NODE,R,RI,G                       LOA
      »,GG,HEAT,HEAD,ALPHA,KBOUND,KFT,PCW,LT,CFACT,KS,                   LOA
      *Z,NCON,TINF,CONV,ALOC.LSOU RC,NSOU RC,ASOU RC,CBAL,AY,A Z             LOA
      « ,AEL,NEL,MEQ,LEL,CBALA,LINE,NLINE,ALINE)                         LOA
        REAL JAC                                                        LOA
        INTEGER Z,AY(4),AZ(4)                                            LOA
       DOUBLE PRECISION AA,NET,NXI,NOT                                   LOA
        COMMON/AAA/XL(4),YL(4),DETJAC,JAC(4,4)/AB/AA(6),W(4)             LOA
        COMMON/A1/M,MM,NUMNP,NSIZE/AA/NXI(12),NET(12),NOT(12>            LOA
        COMMON/ACC/KK(12),K1,K2,K3,K4/AM/KSYM,PCT/HH/LFLOW,LON           LOA
       COMMON/BAL/BLINE/CON/MC1,MC2,NCONV/CONT/LA,LB,LC,LD,LE,LF,LG,LH   LOA
        COMHON/ATHICK/ASIZE,NTHICK.THICK,ERROR,XADD,YADD                 LOA
       DIMENSION NCON(Z,3),TINF(Z),CONV(Z),ALOC(Z),NSOURC(Z)             LOA
       DIMENSION AEL(Z,2), CBAL(Z,2), ASOURC(Z),NODE(LM),NOD(12,LM)      LOA
         DIMENSION R(LN),G(LN),GG(LN),RI(LN),HEAD(LN),HEAT(LM)           LOA
      *,NLINE(Z,2),ALINE(Z),NEL(Z,2)                                     LOA
         RETURN                                                         LOA
                                                                        LOA
           ENTRY  HEATE(KS)
         DO 1  J=1,LN
         RI(J)=R(J)
         G(J)=0.0
         CONTINUE
        BLINEsO.O
        IFCLD.EQ.1.AND.MEQ.EQ.1) CALL CHAN
        IF(LT.EQ.O) GO TO 300
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                        LOA
                                                                       >LOA
                                                                        LOA
   TRANSIENT LOOP  GENERATION AND BOUNDARY CONDITIONS ARE CHANGED AT EACLOA
       IF(KBOUND.EQ.I) CALL BOUND(KS)                                   LOA
C THREE ENTRY POINTS ARE PROVIDED FOR CHANGING RECHARGE RATES           LOA
C AND BOUNDARY CONDITIONS AT EACH TIME STEP—ALL ENTRIES IN BOUNDA      LOA
C                                                                       LOA
       IF(KBOUND.EQ.2) CALL BVAL(KS.ALPHA)                              LOA
C                                                                       LOA
       IF(LD.EQ.1.AND.MEQ.EQ.2) CALL CHANG                              LOA
                                                                        LOA
C

C

C
      CONVECTIVE BOUNDARY ROUTINE
       IF(HEQ.EQ.I)  GO TO 50

      IF(NCONV.EQ.O) GO TO 40

       DO 30   I=1,NCONV
      LLL=NCON(I,1)
       LA=NCON(I,2)
LOA
LOA
LOA
LOA
LOA
LOA
LOA
LOA
  100
  200
  300
  400
  500
  600
  700
  800
  900
 1000
 1100
 1200
 1300
 1400
 1500
 1600
 1700
 1800
 1900
 2000
 2100
 2200
 2300
 2400
 2500
 2600
 2700
 2800
 2900
 3000
 3100
 3200
 3300
 3400
 3500
 3600
 3700
 3800
 3900
 4000
 4100
 4200
 4300
 4400
 4500
 4600
 4700
 4800
 4900
 5000
 5100
 5200
 5300
 5400
5500
5600
5700
 5800
 5900
 6000
6100
6200
                                     -174-

-------
       MCUAYUA)
       MC2=AZ(LA)
       MA=NOD(MC1,LLL)
       NA=NOD(MC2,LLL)
       K = NA
       ONV=ABS(CONV(I))
       CBAL(I,2)=ONV«ALOC(I)*TINF(I)*ALPHA
C   H*LENGTH PF BOUNDARY'TEMP AT INFINITY*TIME STEP
      G(K)=ONV«ALOC(I)«.5*TINF(I)«ALPHA+G(K)
       K = MA
      G(K)=ONV*ALOC(I)*.5*TINF(I)«ALPHA+G(K)
 30   CONTINUE
C
 40   CONTINUE
C
C ROUTINE  THAT CALCUALTES CONVECTIVE INPUTS IS CALLED
       IF(LEL.GT.O) CALL CB(ALPHA)
C
C
50     CONTINUE
                                                                       LOA 6300
                                                                       LOA 6400
                                                                       LOA 6500
                                                                       LOA 6600
                                                                       LOA 6700
                                                                       LOA 6800
                                                                       LOA 6900
                                                                       LOA 7000
                                                                       LOA 7100
                                                                       LOA 7200
                                                                       LOA 7300
                                                                       LOA 7400
                                                                       LOA 7500
                                                                       LOA 7600
                                                                       LOA 7700
                                                                       LOA 7800
                                                                       LOA 7900
                                                                       LOA 8000
                                                                       LOA 8100
                                                                       LOA 8200
£«**«******«*«*»«*»«»»»****«**»»****f«*«***»*»»*f»***»****»***«*«*»«*»»»LOA 8300
C—MATRIX MULTIPLICATION LOOP                                           LOA 8400
C                                                                       LOA 8500
         IF(KSYM.EQ.O) CALL MULTI                                       LOA 8600
        IF(KSYM.NE.O) CALL AMULTI                                       LOA 8700
C                                                                       LOA 8800
C»t««»*»*»*««*«•*»*»««»»«**«*«**»*»****»**»»*»**»**«*»««*»«*»*»«**«**«**LOA 8900
   GENERATION TERMS ARE ADDED INTO THE RIGHT SIDE OF THE EQUATION
C
C
C
C
C
C GENERATION TERMS FROM POINT SOURCES
C
300
350
360
C
C
C
        IF(LSOURC.EQ.O) GO TO 360
        DO 350 LLL=1,LSOURC
        K=NSOURC(LLL)
       IF(ALPHA.LE.O) AS=ASOURC(LLL)
       IF(ALPHA.GT.O) AS=ASOURC(LLL)»ALPHA
       IF(LT.EQ.O) GG(K)=GG(K)+AS
       IF(LT.EQ.I) G(K)=G(K)+AS
       CONTINUE
       CONTINUE

    GENERATION TERMS FROM LINE SOURCES ARE COMPUTED AND ADDED IN

       IF(LINE.EQ.O) GO TO 378
       DO 375 L=1,LINE
       LLL=NLINE(L,1)
       CALL CORD(LLL)
       LA=NLINE(L,2)
       LY=AY(LA)
       LZ=AZ(LA)
       MC3=LY+LZ
       IFCMC3.EQ.5)  A=ABS(YL(LY)-YL(LZ))
       IF(MC3-NE.5)  A=ABS(XL(LY)-XL(LZ))
       AS=ALINE(L)«A
       IF(ALPHA.GT.O) AS=AS*ALPHA
       BLINE=BLINE+AS
       LY=NOD(LY,LLL)
       LZ=NOD(LZ,LLL)
       IF(LT.EQ.O) GO TO  372
LOA 9000
LOA 9100
LOA 9200
LOA 9300
LOA 9400
LOA 9500
LOA 9600
LOA 9700
LOA 9800
LOA 9900
LOA10000
LOA10100
LOA10200
LOA10300
LOA10400
LOA10500
LOA10600
LOA10700
LOA10800
LOA10900
LOA11000
LOA11100
LOA11200
LOA11300
LOA11400
LOA11500
LOA11600
LOA11700
LOA11800
LOA11900
LOA12000
LOA12100
LOA12200
LOA12300
LOA12400
                                     -175-

-------
372

375
C
378
C

380
C
       G(LY)=G(LY)+AS72
       G(LZ)=G(LZ)+AS/2
       GO TO 375
       GG(LY)=GG(LY)+AS/2
       GG(LZ)=GG(LZ)+AS/2
       CONTINUE

       CONTINUE

      DO 380 N=1,NSIZE
       R(N)=R(N)+G(N)
                                                                   LOA12500
                                                                   LOA12600
                                                                   LOA12700
                                                                   LOA12800
                                                                   LOA12900
                                                                   LOA13000
                                                                   LOA13100
                                                                   LOA13200
                                                                   LOA13300
                                                                   LOA13400
                                                                   LOA13500
                                                                   LOA13600
£*««««*»»•«««««««**»««»*«»»«««»««••«*••**•*•*«*•»•••«*•««*•••««««««*»»«*^QA-i 3700
C
C
C
C
  GENERATION TERMS FROH SOURCES  ARE  ADDED  IN
 AND SPECIFIED HEADS ARE TREATED
                                                                       LOA13800
                                                                       LOA13900
                                                                       LOA1HOOO
                                                                       LOA1H100
                                                                       LOA14200
                                                                       LOA14300
                                                                       LOA14400
C                                                                      LOA14500
£•*••*»••••»•*****•»•**•»*»»»••••••*••*»••»«»•**»••*»»•»•« ••»««»••«»»*»«LOA 1^600
       DO mo N=1,NSIZE
      IF(LT.EQ.O) R(N)=CFACT»HEAD(N)+GG(N)
      IF(LT.EQ.I) R(N)=R(N)+CFACT«HEAD(N)+GG(N)
C
C  FORWARD REDUCTION OF THE LOAD MATRIX IS PREFORMED  AND BACK
C    SUBSTITUTION IS PREFORMED TO SOLVE FOR THE UNKNOWN  MATRIX
C
         IF(KSYM.EQ.O)  CALL BACK
         IF(KSYM.NE.O) CALL ABACK
550      CONTINUE
C
^•••••••••••••••••••••**tl«ftl»t**«••••••§•••§••••*•**••••*t»*••»»t*
      RETURN
C

C
C
        SUBROUTINE CB(ALPHA)
C
(;•**••••*»«••**«*»••••*»****••»«•*•••*•»•*•••*••«•••*•*»••»«•*«*««*
       DIMENSION JA(  A=XL(LY)-XL(LZ)
   A=ABS(A)»T*THICK
   LYsNODUY.LLL)
   LZ=NOD(LZ,LLL)

COMPUTE VELOCITIES AT THE GAUSS POINTS
                                                                        LOA14700
                                                                        LOA14800
                                                                        LOA15000
                                                                        LOA15100
                                                                        LOA15200
                                                                        LOA15300
                                                                        LOA15UOO
                                                                        LOA15500
                                                                        LOA15600
                                                                        LOA15700
                                                                        LOA15800
                                                                        LOA15900
                                                                        LOA16000
                                                                        LOA16100
                                                                        LOA16200
                                                                        LOA16300
                                                                        LOA16400
                                                                        LOA16500
                                                                        LOA16600
                                                                        LOA16700
                                                                        LOA16800
                                                                        LOA16900
                                                                        LOA17000
                                                                        LOA17100
                                                                        LOA17200
                                                                        LOA17300
                                                                        LOA17500
                                                                        LOA17600
                                                                        LOA17700
                                                                        LOA17800
                                                                        LOA17900
                                                                        LOA18000
                                                                        LOA18100
                                                                        LOA18200
                                                                        LOA18300
                                                                        LOA18UOO
                                                                        LOA18500
                                                                        LOA18600
                                                                        LOA18700
                                                                        LOA18800
                                     -176-

-------
c
c
 10
 20
VX IS ASSOCIATED WITH ARRAY AY
   JJ=JA(LA)
   II=JB(LA)
   CALL SHAPEKJJ.II)
   CALL VELO(LLL,VX,VY)
   JJ=JC(LA)
   II=JD(LA)
   CALL SHAPEKJJ,!!)
   CALL VELO(LLL,VXX,VYY)
   IFUA.GT.2) GO TO 6
   V1=(VY«A/2)
   V2=(VYY«A/2)
   GO TO 7
   V1=(VX*A/2)
   V2=(VXX»A/2)
   CONTINUE
   GO TO (9,8,8,9),LA
   V1=-V1
   V2=-V2
   CONTINUE
   IF(VI.GE.O) Q=V1*PCW«AEL(J,2)
   IFCV1.LT.O)  Q=V1*PCW«R(LY)
   IF(V2.GE.O) QQ=V2»PCW«AEL(J,2)
   IFCV2.LT.O) QQ=V2«PCW*R(LZ)
        i = G(LY)+Q»ALPHA*FACTUQQ»ALPHA«FACT2
        l = G(LZ)-t-QQ*ALPHA*FACT1+Q*ALPHA«FACT2
   CBALA=CBALA+Q+QQ
   CONTINUE
   FORMATdX,'BOUND TEMPS  • ,7( 14 ,G8 -3 ,3X))
   RETURN
   END
VXX IS ASSOCIATED WITH ARRAY
  LOA18900
AZLOA19000
  LOA19100
  LOA19200
  LOA19300
  LOA19400
  LOA19500
  LOA19600
  LOA19700
  LOA19800
  LOA19900
  LOA20000
  LOA20100
  LOA20200
  LOA20300
  LOA20400
  LOA20500
  LOA20600
  LOA20700
  LOA20800
  LOA20900
  LOA21000
  LOA21100
  LOA21200
  LOA21300
  LOA21100
  LOA21500
  LOA21600
  LOA21700
  LOA21800
  LOA21900
  LOA22000
  LOA22100
                                       -177-

-------
 c
 c
 c
      FORWARD REDUCTION OF THE S MATRIX BY GAUSS ELIMINATION

      SUBROUTINE SOLVECLN,MBAND,S,T,R,RI)
      DOUBLE PRECISION S,T,C
      COMMON/AT/ M, MM, NUMNP, NSIZE
      DIMENSION S(LN,MBAND),T(LN,MBAND) ,R(LN) ,RI(LN)
C GAUSS ELIMINATION—FORWARD REDUCTION OF STIFNESS MATRIX
      DO 550 N=1,NSIZE
      DO 530 L=2,MBAND
      IF(S(N,L).EQ.O) GO TO 530
      I=N+L-1
      C=S(N,L)/S(N,1)
       J=0
      DO 510 K=L,MBAND

 510

530
550
        S(I,J)=S(I,J)-C«S(N,K)
      S(N,L)=C    t
        CONTINUE
        CONTINUE
      RETURN
 C
 C

 C
 C
 C
 C
620
630
650
660
C
      BACK SOLUTION OF THE R VECTOR  BY GAUSS ELIMINATION

      ENTRY BACK
   FORWARD REDUCTION OF THE LOAD MATRIX  IS  PREFORMED  AND BACK
     SUBSTITUTION  IS PREFORMED  TO SOLVE  FOR THE  UNKNOWN  MATRIX
      DO 630 N=1.NSIZE
      DO 620 L=2,MBAND
      IF(S(N,L).EQ.O) GO TO 620
      I=N+L-1
      R(I)=R(I)-S(N,L)*R(N)
      CONTINUE
      R(N)=R(N)/S(N,1)
      DO 660 M=2,NSIZE
      N=NSIZE+1-M
      DO 650 L=2,MBAND
      IF(S(N,L).EQ.O) GO TO 650
      K=N+L-1
      R(N)=R(N)-S(N,L)»R(K)
         CONTINUE
      CONTINUE

      RETURN
C  MULTIPLICATION OF THE T MATRIX BY THE NODAL VALUES AT THE LAST TIME
C
C
      ENTRY MULTI
C
C    MATRIX MULTIPLICATION LOOP—CRANK-NICOLSON METHOD
\*
       DO 150 Hsl,NSIZE
       R(N)=0.0
      DO 110 M:1,MBAND
      NO=H+N-1
      IF(NO.GT.NSIZE)  GO TO 120
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 (SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 •SOL
SSOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
 SOL
  100
  200
  300
  400
  500
  600
  700
  800
  900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
2500
2600
2700
2800
2900
3000
3100
3200
3300
3400
3500
3600
3700
3800
3900
4000
4100
4200
4300
4400
4500
4600
4700
4800
4900
5000
5100
5200
5300
5400
5500
5600
5700
5800
5900
6000
6100
6200
6300
                                     -178-

-------
  110   R(N>=RI(NO)*TCN,M)+R\N>                                          SOL 6400
  120   CONTINUE                                                         SOL 6500
      MBA=MBAND-1                                                        SOL 6600
       DO 130 M=1,MBA                                                   SOL 6700
      NO=N-M                                                            SOL 6800
      IF(NO.LE.O)  GO TO 140                                             SOL 6900
  130   R(N)=RI(NO)*T(NO,M+1)+R(N)                                       SOL 7000
  140   CONTINUE                                                         SOL 7100
150   CONTINUE                                                          SOL 7200
C                                                                       SOL 7300
      RETURN                                                            SOL 7400
      END                                                               SOL 7500
                                      -179-

-------
£*««*«*****•*•*«*•»«*»••«••«•«*«•«*»«••«»»*•»*«»»«*«*«»*»***»•« »«««»» »««ASQ  -JQQ
C                                                                       ASO  200
      SUBROUTINE ASOLVE(B,R,RI,NEQ,IHALFB,NDIM,MDIM,T)                  ASO  300
C                                                                       ASO  400
C   ASSYMMETRIC BAND MATRIX EQUATION SOLVER                             ASO  500
C    ORIGINALLY PROGRAMED BY J.O. DUGUID                                ASO  600
C                                                                       ASO  700
      DOUBLE PRECISION B,T,PIVOT,C                                      ASO  800
      DIMENSION B(NDIM,MDIM),R(NDIM),T(HDIM,MDIM)                       ASO  900
      DIMENSION RI(NDIM)   .                                             ASO TOGO
      NRS=NEQ-1                                                         ASO 1100
      IHBP=IHALFB+1                                                     ASO 1200
C                 ,                                                      ASO 1300
C   TRIANGULARIZE MATRIX B USING DOOLITTLE METHOD                        ASO 1400
C                                                                       ASO 1500
      DO 20 K=1,NRS                                                     ASO 1600
        PIVOT=B(K,IHBP)                                                  ASO 1700
      KK=K+1                                                            ASO 1800
      KC=IHBP                                                           ASO 1900
      DO 10 I=KK,NEQ                                                    ASO 2000
      KC=KC-1                                                           ASO 2100
      IF(KC.LE.O) GO TO 20                                              ASO 2200
      C=-B(I,KC)/PIVOT                                                  ASO 2300
      B(I,KC)=C                                                         ASO 2400
      KI=KC-f1                                                           ASO 2500
      LIMsKC+IHALFB                                                     ASO 2600
      DO 10 J=KI,LIM                                                    ASO 2700
      JC=IHBP+J-KC                                                      ASO 2800
10    B(I,J) = BU,J) + C»B(K,JC)                                           ASO 2900
20    CONTINUE                                                          ASO 3000
        RETURN                                                           ASO 3100
C                                                                       ASO 3200
^•••••••••••••••••••••••••••••••«t»*»»«*****>«*»«»»*»«•«»«••«**««**•«**»ASO 3300
C                                                                       ASO 3400
      ENTRY ABACK                                                       ASO 3500
C                                                                       ASO 3600
      NN=NEQ+1                                                          ASO 3700
      IBANDs2«IHALFB+1                                                  ASO 3800
      DO 70 1=2,NEQ                                                     ASO 3900
      JC=IHBP-I+1                                                       ASO 4000
      JI=1                                                              ASO 4100
      IF(JC.LE.O) GO TO 40                                              ASO 4200
      GO TO 50                                                          ASO 4300
40    JC=1                                                              ASO 4400
      JI=I-IHBP+1                                                       ASO 4500
50    SUM=0.0                                                           ASO 4600
      DO 60 J=JC,IHALFB                                                 ASO 4700
      SUM=SUM+B(I,J)«R(JI)                                              ASO 4800
60    JI=JI+1                                                           ASO 4900
70    R(I)=R(I)+SUH                                                     ASO 5000
C                                                                       ASO 5100
C    BACK SUBSTITUTION                                                  ASO 5200
C                                                                       ASO 5300
      R(NEQ)=R
-------
80    SUM=SUM+B(I,J)*R(JP)                                              ASO 6300
90    R(I)=(R(I)-SUM)/B(I,IHBP)                                         ASO 6400
100   RETURN                                                            ASO 6500
C                                                                       ASO 6600
c»««*»«»»*»**********»»»****»***»»*»*»»*******»*»*****»****«»*»»»»»*«**«ASO 6700
C                                                                       ASO 6800
      ENTRY AMULTI                                                      ASO 6900
C                                                                       ASO 7000
      LN=NDIM                                                           ASO 7100
      MBAND=(MDIM+1)/2                                                  ASO 7200
      DO 150 N=l,LN                                                     ASO 7300
      R(N)=0.0                                                          ASO 7400
      DO 110 M=1,MDIM                                                   ASO 7500
      NO=M+N-MBAND                                                      ASO 7600
      IF(NO.GT.LN) GO TO  120                                            ASO 7700
      IF(NO.LT.I) GO TO 110                                             ASO 7800
      R(N)=RI(NO)*T(N,M)+R(N)                                            ASO 7900
 110     CONTINUE                                                         ASO 8000
 120   CONTINUE                                                           ASO 8100
 150     CONTINUE                                                          ASO 8200
      RETURN                                                             ASO 8300
      END                                                                ASO 8400
                                       -181-

-------
 £«»«»•*«»«•»«•*««t*«*««««»««*»•«•»««»«««««»  *»«*»»»»«**••«««ft«»»ft»»»««ft«BAL
 C    THIS SUBROUTINE COMPUTES A MASS  BALANCE                            BAL
 C                                                                       BAL
       SUBROUTINE BALAN(LM,LN,R,PX.PY.PCX.HEAD,HEAT,Z,LSOURC,ASOURC,      BAL
      • LFLUX,NFLUX,A,B,C,D,E,MPRINT,NTYPE,RI,NODE,NOD,FLOWX,FLOWY,PCW   BAL
      »  ,MEQ,CBAL,CBALA,AY.AZ,NCON,KRAN)                                 BAL
       REAL  JAC                                                         BAL
       INTEGER Z.AY.AZ                                                   BAL
        DOUBLE PRECISION AA.NXI,NET,NOT                                   BAL
        COMMON/AAA/XL(4),YL(4),DETJAC,JAC(4.4)/AB/AA(6)-W(4)              BAL
        COMMON/CON/M1,M2,NCONV/AA/NXI(12),NET(12),NOT(12)                 BAL
        COHMON/ACC/KK(12)>K1,K2,K3,K4/HH/LFLOW,LON/BAL/BLINE              BAL
       DIMENSION RI(LN),R(LN),PX(LM),PY(LM),NODE(LM)                      BAL
        DIMENSION FLOHX(LM),FLOWY(LM),AY(4),AZ(4)                         BAL
        DIMENSION NCON(Z,2),CBAL(Z,2)  ,  NOD(12,LM)                        BAL
       DIMENSION PCX(LM),HEAD(LN),HEAT(LM),ASOURC(Z),NFLUX(Z,2)           BAL
        DIMENSION LOA(4),LOB{4),LOC(4),LOD(4)                            BAL
        DATA LOA/1,2,3,4/LOB/2,3,4,1/LOC/5,7.9,11/LOD/6 8.10,12/          BAL
       RETURN                                                            BAL
 £«•»**•»*«»»«•««•»*•«*»»**••«**»*»***«*•**«
       ENTRY MASBAL(ALPHA)
 C
 C	COMPUTE BOUNDARY FLUXES
 C
        ST=0.0
        PUMP=0.0
        F=0.0
        LFLOW=0
        RECHsO.O
        DISCH=0.0
        IF(LFLUX.EO.O).GO TO 15
        AA(1)=0,0
        AA(2)=0.0
 C
       DO 10  LLA=1,LFLUX
       LLL =NFLUX(LLA,1)
       K=NFLUX(LLA,2)
 C
        CALL FLOWW(LLL)
 C
        LT=NFLUX(LLA,2)
        FLUXX=0.0
        FLUXY=0.0
       IFCLT.GT.2) FLUXX=DHX»ARX«PX(LLL)
       IF(LT.LT-3) FLUXY=DHY*ARY»PY(LLL)
        IFCMEQ.EQ.1}  GO TO 8
        LY=AY(LT)
        LZ=AZ(LT)
        LY=NOD(LY,LLL)
        LZ=NOD(LZ,LLL)
        RR=(R(LY)+R(LZ))/2
       IFCLT.GT.2) FLUXX=FLUXX«-FLOWX(LLL)»PCW»RR
       IF(LT.LT.3) FLUXY=FLUXY+FLOWY(LLL)«PCW»RR
8      CONTINUE
       IFCLT.EQ.1.AND.FLUXY.LT.O)
       IF(LT.EQ.1.AND.FLUXY.GT.O)
       IFCLT.EQ.2.AND.FLUXY.LT.O)
       IFUT.EQ.2.AND.FLUXY GT.O)
       IFCLT.EQ.3-AND.FLUXX.LT.O)
       IF(LT.EQ.3-AND.FLUXX.GT.O)
       IF(LT.EQ.4.AND.FLUXX.LT.O)
       IFCLT.EQ.4.AND.FLUXX.GT.O)
10    CONTINUE
          ***«»»*«**««*»««*«*««»««t»»*BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
                                      BAL
DISCH=DISCH-FLUXY
RECH=RECH+FLUXY
RECH=RECH-FLUXY
DISCH=DISCH+FLUXY
RECH=RECH-FLUXX
DISCH=DISCH+FLUXX
DISCH=DISCH-FLUXX
RECH=RECH+FLUXX
BAL
BAL
BAL
BAL
BAL
BAL
BAL
BAL
BAL
BAL
BAL
BAL
 100
 200
 300
 400
 500
 600
 700
 800
 900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
2500
2600
2700
2800
2900
3000
3100
3200
3300
3400
3500
3600
3700
3800
3900
4000
4100
4200
4300
4400
4500
4600
4700
4800
4900
5000
5100
5200
5300
5400
5500
5600
5700
5800
5900
6000
6100
6200
6300
                                     -182-

-------
J5    CONTINUE                                                          BAL

C*»******»*»***»*»*. **««»««*,,»«,,«««,„,», **««««„ ««,»««««««««*, »«»,,»BAL 6600
C   COMPUTE RATE OF CHANGE IN STORAGE                                   BAL 6700
         n                                                                  6900
        =0 .0
      DO 20 LLL=1,LM
       M4=H
       CALL DERIVECLLL.M4)                                              0AL 7300
               IF(MEQ.EQ.I) CALL PEECT.LLL)                             BAL 7400
               IF(MEQ-EQ.I) DETJAC=DETJAC/T                             BAL 7500
       j|B=0                                                             BAL 7600
       £A=0                                                             BAL 7700

       S. 9 1.1.4
       L3=LOC(I)                                                        BAL 3200
       W=LOD(I)                                                        BAL 8300
       L5=MOD(L1,LLL)                                                   BAL 8400
       L6=NOD(L2,LLL)                                                   BAL 8500
       L7=NOD(L3,LLL)                                                   BAL 8600
       L8=NOD(L4,LLL)                                                   BAL 8700
       IF(LT.GT.O) GO TO 23                                             BAL 8800
       RA=RA+.5«R(L5)+.5*R(L6)                                          BAL 8900
       RB=RB+.5*RI(L5)+.5*RI(L6)                                        BAL 9000
       GO TO 19                                                         BAL 9100
23     IF(LS.GT.O) GO TO 24                                             BAL 9200
       RA=RA+1/3»R(L5)+2/3*R(L7) + 1/3»RU6)                              BAL 9300
       RB=RB+1/3*RI(L5) + 1/3*RIU6)+2/3»RI(L7)                           BAL 9400
       GO TO 19                                                         BAL 9500
24     RA=RAt-1/8«R(L5) + 1/8*R(L6)+3/8»R(L7)+3/8*R(L8)                    BAL 9600
       RBsRB+1/8»RI(L5)+1/8»RI(L6)+3/8»RI(L7)+3/8»RI(L8)                BAL 9700
19     CONTINUE                                                         BAL 9800
       RA=RA/4                                                          BAL 9900
       RB=RB/4                                                          BAL 10000
      ST=ST+(RA-RB)»PCX(LLL)«U«DETJAC                                   BAL10100
20    CONTINUE                                                          BAL10200
21    CONTINUE                                                          BAL10300
C                                                                       BAL 10 400
£*««**«««*****»***«*«****««**«*»*»*»»«***»*  *»*»*»»»»*»*»*»*•»****«*«* »*BAL 10500
C      ADD IN PUMPING SOURCES                                           BAL10600
C                                                                       BAL10700
      PUMP=0.0                                                          BAL10800
       PUMP=PUMP+BLINE                                                  BAL10900
      IF(LSOURC.EQ.O) GO TO 31                                          BAL11000
      DO 30 LLL=1,LSOURC                                                BAL11100
      PUMP=PUMP+ASOURC(LLL)                                             BAL11200
30     CONTINUE                                                         BAL11300
31     CONTINUE                                                         BAL11400

C     DISTRIBUTED SOURCES                                               BAL11500
        DO 40 LLL=1,LM                                                  BAL11600
       IF(HEATULL).EQ.O) GO TO 40                                      BAL11700
        M4=NODE(LLL)                                                    BAL1 1800
       CALL DERIVE(LLL,H4)                                              BAL11900
       PUMP=PUMP+HEAT(LLL)«4«DETJAC                                     BAL12000
40     CONTINUE                                                         BAL12100
C                                                                       BAL12200
C *t ««««»»««•»**««««»*««»«»*«*•»»»*««»*»»««»««••«««»««»*«*«»»«*         BALI 2300
C                                                                       BAL12400
                                     -183-

-------
c
c
43
C
C
15
48
     ROUTINES  TO  COMPUTE FLUXES ACROSS CONVECTIVE BOUNDARIES

        IF(MEQ.EQ.I)  GO TO  48
        IF(NCONV.EQ.O) GO TO 45
        C0=0.0
        COA=0.0
        DO 43  I=1,NCONV
        LLL=NCON(I,1)
        LA=NCON(I,2)
        MC1=AY(LA)
        MC2=AZ(LA)
        MA=NOD(HC1,LLL)
        NA=NOD(MC2,LLL)
        NC=CBAL(I,2)«100
        IF(NC.EQ.O)  COA=COA+(R(MA)/4+R(NA)/4+RI(MA)/4+RI(NA)/4)«CBAL(I,
        IF(NC.EQ.O)  GO TO 43
        CO=CO+CBAL(I,2)-(R(MA)/4+R(NA)/4+RI(MA)/4+RI(NA)/4)*CBAL(I,l)
        CONTINUE
        CONTINUE
        CONTINUE
 C»»««»»»«»»*»»«*»»»»»*»»»»«»»*»•«»««»»»«*««««*»»«
 C
 c
        AL = ALPHA
        IFULPHA.LT.0.0001) AL=1
        RECH=RECH«AL
        DISCH=DISCH»AL
        PUMP=PUMP»AL
         CBALA=CBALA*AL
 £•»»*•««•t»tt*f*•»•»**•*•*t*If«f»«*«»*•tt*»t»«t*fft«**»!«»*»* t
       A=A+RECH
       B=B+DISCH
       C=C+ST
       D=D+PUMP
        0=0* CO
        00=00+COA
        P=P+CBALA
        IF(MEQ .EQ . 1) F=RECH-DISCH-«-PUMP-ST
        IF(MEQ.EQ.2) F=RECH-DISCH+PUMP-ST+CBALA+CO-COA
       E=E+F
         RETURN
         ENTRY BPRINT
       PRINT 50
         IFCMEQ.EQ.2) PRINT 51,0,CO,00,COA,P.CBALA
       PRINT 52,A,RECH,B,DISCH,C,ST,D,PUMP,E,F
50    FORMATCIX,T30,'CUMULATIVE MASS BLANCE1.
     •T60,'RATES FOR THIS TIME STEP')
51     FORMATdX,'CONDUCTIVE TRANSFER',
     «  ,T30,G12.6,T60,G12.6,
     •  1X/1X,'CONVECTIVE TRANSFER—OUT  '.T30.G12.6,T60,G12.6
     «,1X/1X,'CONVECTIVE TRANSFER—IN',T30,G12.6,T60,G12.6)
52     FORMAT(
     »1Xf'B. FLUX RECHARGE',T30.G12.6,T60, G12.6.
     «/1X,'B. FLUX DISCHARGE',T30,G12.6,T60.G12.6,
     •/«, 'CHANGE IN STORAGE',T30,G12.6,T60,G12.6,
     »/U,'QUANTITY PUMPED'.T30.G12 .6 .T60 ,G12.6 ,
     •/U.'DIFFERECE • ,T30,G12.6,T60,G12.6)
       RETURN
  BALI 2500
  BALI 2600
  BAL12700
  BAL12800
  BAL12900
  BALI 3000
  BALI 3100
  BALI 3200
  BALI 3300
  BAL13400
  BALI 3500
  BAL13600
  BALI 3700
  BALI 3800
DBAL13900
  BAL14000
  BAL14100
  BAL14200
  BAL14300
  BAL14400
  BAL14500
  BALI 4600
  BAL14700
  BAL14800
  BAL14900
  BAL15000
  BAL15100
  BAL15200
  BALI 5300
  BAL15400
  BAL15500
  BAL15600
  BAL15700
  BAL15800
  BAL15900
  BALI6000
  BAL16100
  BALI6200
  BAL16300
  BAL16400
  BAL16500
  BALI6600
  BAL16700
  BAL16800
  BAL16900
  BAL17000
  BAL17100
  BAL17200
  BAL17300
  BAL17400
  BAL17500
  BAL17600
  BAL17700
  BAL17800
  BAL17900
  BAL18000
  BAL18100
  BAL18200
  BAL18300
  BAL18400
  BAL18500
                                    -184-

-------
C«f*««*»««*****»»**«»«*««***«*«»«»*«»*«»»
c
       ENTRY WATER
       AA(1)=0.0
       AA(2)=0.0
       LFLOW=0
       DO 100 LLL=1,LM
       CALL FLOWW(LLL)
                                          **««*«*«*»*****»»«*»«*
C
C
C
      FLOW EQUALS THE SLOPE   *   AREA   *  PERMEABILITY
      FLOWX(LLL)=DHX*ARX*PX(LLL)
       FLOWY(LLL)=DHY*ARY*PY(LLL)
C
100    CONTINUE
C «*««»*****»*********«**«*****«****»»***«»*««»*«»»»**»«****»****»
       RETURN
C
C
C
c
       ENTRY VELOC(LN,LM,RW,WX WY)
       DIMENSION WX(LM),WY(LM),RW(LN)
       RETURN
100
500
C
C
       DO 500 1=1, Ml
         K=K+1
       IF(NOD(K,LLLQ) .EQ.O) GO TO 400
        KI=NOD(K,LLLQ)
       DHE=DHE+NXI(I)«RW(KI)
       DHN=DHN+NET(I)«RW(KI)
         CONTINUE
      DHX=JAC( 1 , 1)*DHE+JAC( 1 ,2)*DHN
      DHY=JAC(2,1)«DHE+JAC(2,2)«DHN
       VX=DHX»WX(LLLQ)
       VY=DHY«WY(LLLQ)
                RETURN
                ENTRY VCENTCLLLQ.VXQ,VYA)
    THE AREA TO BE USED IN THE FORMULA Q=KIA IS COMPUTED
      ARX=(YL(3) + YLC4)-YL(1)-YL(2))/2
      ARX=ABS(ARX)
      ARY=(XL(2)+XL(3)-XL(1)-XL(4))
      ARY=ABS(ARY)/2
                VXQ=FLOWX(LLLQ)/ARX
                VYA=FLOWY(LLLO)/ARY
                RETURN
                                                                       BAL18600
                                                                       BAL18700
                                                                       BAL18800
                                                                       BAL18900
                                                                       BAL19000
                                                                       BAL19100
                                                                       BAL19200
                                                                       BAL19300
                                                                       BAL19400
                                                                       BAL19500
                                                                       BAL19600
                                                                       BAL19700
                                                                       BALI 9800
                                                                       BAL19900
                                                                       BAL20000
                                                                       BAL20100
                                                                       BAL20200
                                                                       BAL20300
                                                                       BAL20400
                                                                       BAL20500
                                                                       BAL20600
                                                                       BAL20700
                                                                       BAL20800
C
C
              ENTRY VELO(LLLQ,VX,VY)                                  BAL21100
                                                                      BAL21200
                                                                      BAL21300
                                                                      BAL21400
                                                                      BAL21500
                                                                      BAL21600
                                                                      BAL21700
                                                                      BAL21800
                                                                      BAL21900
                                                                      BAL22000
                                                                      BAL22100
                                                                      BAL22200
                                                                      BAL22300
                                                                      BAL22400
                                                                      BAL22500
                                                                      BAL22600
                                                                      BAL22700
                                                                      BAL22800
«**«**«**«***»»»*»**«*««***«««**«««*««»***«*****»*»*****«*******»*****BAL22900
                                                                      BAL23000
                                                                      BAL23100
                                                                      BAL23200
                                                                      BAL23300
                                                                      BAL23400
                                                                      BAL23500
                                                                      BAL23600
                                                                      BAL23700
                                                                      BAL23800
                                                                      BAL23900
                                                                      BAL2MOOO
                                                                      BAL24100
                                                                      BAL21200
                                                                      BAL24300
                                                                      BAL24400
                                                                      BAL24500
                                                                      BAL21600
                                                                      BAL21700
                                                                      BAL24800
       SUBROUTINE FLOWW(LLL)

       DHE=0.0
       DHN=0.0
       M4=NODE(LLL)
                                     -185-

-------
        CALL  DERIVEULL.MH)
       K=0
       DO 7 1=1 ,MH
7


C
c
c
   IF(NOD(K,LLL).EQ.O) GO TO H
   KI=NOD(K,LLL)
   DHE=DHE+NXI(I)«R(KI)
   DHN=DHN+NET(I)«R(KI)
   CONTINUE
  DHX= JAC( 1,1) »DHE+ J AC( 1,2) »DHN
  DHY=JAC(2,1)«DHE+JAC(2,2)«DHN
THE AREA TO BE USED IN THE FORMULA Q=KIA IS COMPUTED
  ARX=(YL(3)+YL(4)-YL(1)-YL(2))/2
  ARX=ABS(ARX)
  ARY=(XL( 2)+XL{ 3)-XL( 1 )-XL( «) )
  ARY=ABS(ARY)/2
   ARXX=ARX
   ARX=DETJAC»i»/ARY
   ARY=DETJAC«J»/ARXX
         t

   RETURN
BAL24900
BAL2500C
BAL2510C
BAL2520C
BAL2530C
BAL25'40C
BAL2550C
BAL2560C
BAL2570C
BAL2580C
BAL2590C
BAL2600d
BAL26100
BAL26200
BAL26300
      END
BAL26500
BAL26600
BAL26700
BAL26800
BAL26900
BAL27000
BAL27100
BAL27200
BAL27300
                                     -186-

-------
C THIS ROUTINE IS USED TO PRINT FLOWS AND FLUXES                        FLO  100
f*

       SUBROUTINE FLOWS(LM,LN ,R,R1,FLOWX.FLOWY)                         FLO  300
       DIMENSION R(LN),RULN),FLOWX(LM),FLOWY(LM)                       FLO  400
       COMMON/HI/TITLE(25),V(26),VV(26)                                 FLO  500

C*«»*««**««ft*«»««**««ft**»««»««»»»t««««ft««*»ft«*                          pLQ  ^QQ
       RETURN                                                           FLO  800
       ENTRY FFLOW                                                      FLO  900
       VRITEdU.V) (R1(I),I=1,LN)                                       FLO 1000
       RETURN                                                           FLO HQO
       ENTRY FFFLOW                                                     FLO 1200
       HHITEdl.V) (R(I),I=1,LN)                                        FLO 1300
       RETURN                                                           FLO 1400
C                                                                       FLO 1500
      ENTRY WFLOW(S)                                                    FLO 1600
      PRINT 1,S                                                         FLO 1700
 1    FORMAT(1X//1X,'TIME  STEP   ',G12.6/1X,'FLOWS  IN THE X DIRECTION1)  FLO 1800
       PRINT VV,(FLOWX(I),I=1,LM)                                       FLO 1900
      PRINT 2                                                           FLO 2000
  2    FORMATCJX/1X,'FLOWS IN THE Y  DIRECTION')                         FLO 2100
      PRINT VV,(FLOWY(I),1=1,LM)                                        FLO 2200
      RETURN                                                            FLO 2300
C                                                                       FLO 2400
C***«*«»****«««**f**x*ft««***«x»**»*ftft«««»***x>                          FLO 2500
C                                                                       FLO 2600
      ENTRY HPRINT(S)                                                   FLO 2700
      PRINT 10,S                                                        FLO 2800
 10    FORMATdX//1X, 'TEMPERATURE DISTRIBUTION  AT  TIME  STEP   ',G12.6)   FLO 2900
      PRINT V,(R(I) ,I=1,LN)                                             FLO 3000
      RETURN                                                            FLO 3100
C                                                                       FLO 3200
£t»ft«*«ft«»»*««>«««*»*>»««*«c***«*««»»»ft«««»««««*«                       FLO 3300
C                                                                       FLO 3400
       ENTRY WPRINT(S)                                                  FLO 3500
      PRINT 100,S                                                       FLO 3600
100       FORMAT(1X//1X,'POTENTIAL DISTRIBUTION AT TIME STEP   '.G12.6)  FLO 3700
      PRINT V,(R1(I),1=1,LN)                                            FLO 3800
      RETURN                                                            FLO 3900
      END
                                      -187-

-------
£**•**•••******•*•«•«***•*****«*«***»••«*•*»•***••**•*•••»***•••*«»**•»• PAP
       SUBROUTINE PARAM(LM,LN,BBB,XLOC, YLOC.NOD.PCX ,DIFF ,R ,R1 ,WX,WY,     PAR
C                                                                       PAR
C    THIS SUBROUTINE IS USED TO COMPUTE THE DISPERSION  COEFFICEIENTS,    PAR
C     THE LOCALIZED COORDINATES, AND  CHANGES IN HYDRAULIC               PAR
C    CONDUCTIVITY WITH TEMPERATURE                                      PAR
C                                                                       PAR
^•••••••••••••••••«»»>*«»«««»i •>»>•* «•»**>«»«»»»*««*« i ««*««***»»«**«««•  PAR
     • A, B, EOT, K AHEAD     '                                            PAR
       INTEGER BBB                                                      PAR
       DIMENSION XLOC(LN),YLOC(LN),NOD(12,LM>,PCX(LM),DIFF(LM,2),R(LN)   PAR
     *,  RHLN),HX(LM),WY(LM),A(LM)tB(LM),BOT(BBB)                       PAR
       REAL  JAC                                                         PAR
       COMMON/AAA/XL(4),YL(4),DETJAC,JAC(4,4)/ACC/KK(12) ,K1 ,K2,K3,K4     PAR
     »/HM/PXX,PYY,PXY,KAD/ AM/NLA, PCH                                    PAR
C                                                                       PAR
       RETURN                                    -                       PAR
C                                                                       PAR
£««*••«••*«**«»»»*•«««««»*»««»*«•«*««*«*»•*«•*******»»«**
C  THE ROUTINE THAT CALCULATES THE DISPERSION COEFFICIENTS
       ENTRY MECD(LLL,VX,VY)
        V=VX»VX+VY«VY
        IF(V.LE.O) RETURN
        V=SQRT(V)
       DL=DIFF(LLL,1)»V
       DT=DIFF(LLL,2)«V
        PXX=(DL«VX*VX/V/V+DT«VY«VY/V/V)«PCW
       PYY=(DT»VX«VX/V/V+DL»VY*VY/V/V)»PCW
       PXY=«DL-DT)»VX»VY/V/V)*PCW
       PXY=ABS(PXY)
       RETURN
C
£••*»•••••**•*••••*••**»*•••*»*«•••»•*••»•«•**»****»*»*»*
C
        ENTRY CORD(LLLQ)
       COMMON/AF/II,JJ,L1(4)
C  THIS ROUTINE COMPUTES LOCALIZED COORDINATES
       K1=NOD(1,LLLQ)
       K2=NOD(2,LLLQ)
       K3=NOD(3,LLLQ)
       K4=NOD(4,LLLQ)
       AQrXLOC(KI)
       BB=XLOC(K2)
       C=XLOC(K3)
       D=XLOC(K4)
       E=AMIN1(AQ,BB,C,D)
      XL(1)=AQ-E
      XL(2)=BB-E
      XL(3)=C-E
      XL(4)=D-E
      AQ=YLOC(K1)
      BB=YLOC(K2)
      C=YLOC(K3>
      D=YLOC(K4)
      E=AHIN1(AO,BB,C,D)
      YL(1)=AQ-E
      YL(2)=BB-E
      YL(3)=C-E
      YL(H)=D-E
C
C
      RETURN
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
PAR
                                        -JQQ
                                        200
                                        300
                                        400
                                        500
                                        600
                                        700
                                        goo
                                        900
                                       1000
                                       1100
                                       1200
                                       1300
                                       1400
                                       1500
                                       1600
                                       1700
                                       1800
                                       1900
                                       2000
                                       2100
                                       2200
                                       2300
                                       2400
                                       2500
                                       2600
                                       2700
                                       2800
                                       2900
                                       3000
                                       3100
                                       3200
                                       3300
                                       3400
                                       3500
                                       3600
                                       3700
                                       3800
                                       3900
                                       4000
                                       4100
                                       4200
                                       4300
                                       4400
                                       4500
                                       4600
                                       4700
                                       4800
                                       4900
                                       5000
                                       5100
                                       5200
                                       5300
                                       5400
                                       5500
                                       5600
                                       5700
                                       5800
                                       5900
                                       6000
                                       6100
                                       6200
                                       6300
-188-

-------
C
C«»t****t*«**t*ft«**«*«***«****«*«««***xft***x«*«**««*«c«***ft«x»«*
C
       ENTRY PE(MS)
C
       IF(MS.EQ.O) RETURN
      DO 1 J=1 ,LM
      N1=NOD(1,J)
      N2=NOD(2,J)
      N3=NOD(3,J)
      N4=NOD(4,J)
      T=(R(N1) + R
c
C   THE VISCOSITY  RELATIONSHIP
c
      U=1.917-.05635*Ti-.00071*T*T
      WX(J)=1.21/U«A(J)
      WY(J)=1.2VU*B(J)
1     CONTINUE
      RETURN
       ENTRY  PEE(BBQQ.JQQ)
       IF(KAREAL.EQ.O)  RETURN
       N1=NOD(1,JQQ)
       N2=NOD(2,JQQ)
       N3=NOD(3,JQQ
       N4-NOD(U  JOO)
 C     GEOMETRIC MEANS  ARE CALCULATED FOR AQUIFER DEPTH
       BBQQ=(R1(Sl)-BOT(N1))«(R1(N2)-BOT(N2))»(RHN3)-BOT(N3))«(R1(NM)
      »OT(NU))
       BBQQ=ABS(BBQQ>
       BBQQ=BBQQ«*0.25
        RETURN
        tnu
                                 PAR 6400
                                 PAR 6500
                                 PAR 6600
                                 PAR 6700
                                 PAR 6800
                                 PAR 6900
                                 PAR 7000
                                 PAR 7100
                                 PAR 7200
                                 PAR 7300
                                 PAR 7400
                                 PAR 7500
                                 PAR 7600
                                 PAR 7700
                                 PAR 7800
                                 PAR 7900
                                 PAR 8000
                                 PAR 8100
                                 PAR 8200
                                 PAR 8300
                                 pAR 8400
                                 PAR 8500
                                 PAR 8600
                                  PAR 9100
                                 BPAR 9200

                                  pAR
-189-

-------
 c«««»*»««««»«*««»««»»«»*«t•«»§«»«**«§«»»««»»«»««*«*«»»f«»»»»»,»,«»»,
 C   THIS ROUTINE IS USED TO CHANGE BOUNDARY CONDITIONS OR  PARAMETERS
 C     AT EACH TIME STEP     FIVE ENTRY POINTS ARE PROVIDED
 C
        SUBROUTINE BOUNDA(LM,LN,Z,R,R1.HEAD.HEA.FLOWX,FLOWY,HEAT.INFLOW
      »,NWATER,AWATER,NHEAT,AHEAT,NCON,TINF,CONV,ALOC,NEL,AEL,CBAL
      »,LEL,LHEAT,LWATER,PCW,AY,AZ,LINEW,LINEH,NLINEW,NLINEH,
      « ALINEW.ALINEH)
        INTEGER Z,AY(4),AZ(4)
        REAL INFLOW,N11
        COMMON/ATHICK/ASIZE.NTHICK,THICK,ERROR/BM/RYC4800,3)
      */BOUND/DIST,N1,N2,N3,N4,N5,N6,N7,N8,N9,N10
      */CONT/LA,LB,LC,LD,LE,LF,LG,LH/CON/MC1,MC2,NCONV/ME/MEQ
 C—LINKING INFORMATION
         DIMENSION FLOWX(LM),FLOWY(LM)
 C--INFORMATION FOR POINT SOURCES
         DIMENSION NWATER(Z),AWATER(Z),NHEAT(Z),AHEAT(Z)
 C  RECHARGE RATE INFORMATION
      «, HEAT(LM),INFLOW(LM),NLINEW(Z>2),NLINEH(Z,2),ALINEW(Z),ALINEH(Z)
 C—INFORMATION ON CONVECTIVE BOUNDARIES
      • ,NCON(Z,3),TINF(Z),CONV(Z),ALOC(Z),NEL(Z,2),AEL(Z,2),CBAL(Z,2)
 C—BOUNDARY CONDITIONS                                         ^L,*I
       DIMENSION HEAD(LN),HEA(LN)
 C—INITIAL CONDITIONS AND ANSWERS
        DIMENSION R(LN).RKLN)
  110
  111
        RETURN
  ENTRY LAKE
 READ,  DIST,N1,N2,N3,N4,N5,N6,N7,N8,N9,N10,N11
 READ,M,NUMBER
 DO 110 K=1,M
  READ.A.B.C
 HM=NUMBER-M
 DO 111  K=1,MM
  READ,RY(K,3),RY(K,1),RY(K,2)
 RETURN
                                                           *»»
100

C
C'
C

C
59
 ENTRY BVALUSQ,ALPHA)
 K=KSQ*ALPHA
 T1=RY(K,1)-5.0
 DO 100 J=1,LEL
   AEL(J,2)=T1
 RETURN
 EMTRY BOUND(KS)

        K=KS«N10+8
TlsRY(K.I)
T2=RY(K,2)
AA=ALOG(T1)
AB=(ALOG(T2)-AA)/54
T3=AA+DIST«AB
T3=EXP(T3)
 DO 59 J=N1,N2,N3
 HEAD(J)=T3
DO  60 J=N1,N5,N6
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
 BOU
BOU
BOU
  100
  200
  300
  400
  500
  600
  700
  800
  900
 1000
 1100
 1200
 1300
 1400
 1500
 1600
 1700
 1800
 1900
 2000
 2100
 2200
 2300
 2400
 2500
 2600
 2700
 2800
 2900
 3000
 3100
 3200
 3300
 3400
 3500
 3600
 3700
 3800
 3900
 4000
 4100
 4200
 4300
 4400
 4500
 4600
 4700
 4800
 4900
 5000
 5100
 5200
 5300
 5400
 5500
 5600
 5700
 5800
 5900
 6000
 6100
6200
6300
6400
                                     -190-

-------
 60     HEAD(J)=T3-2.5                                                  BOU 6500
       DO 200  J=1,LEL                                                  BOU 6600
       TINF(J)=T3                                                       BOU 6700
200       AEL(J,2)=T3                                                   BOU 6800
      IF(NCONV.LT.I) GO TO 300                                          BOU 6900
       LEA=LEL+1                                                        BOU 7000
      DO 500 J=LEA,NCONV                                                BOU 7100
       TINF(J)=RY(K-2,1)-N11                                            BOU 7200
      IF(RY(K-2,3).LT.50) TINF(J)=RY(K-2,3)-N11                         BOU 7300
 500    CONTINUE                                                        BOU 7400
       PRINT 555, (NCON(J,1),TINF(J),J=1,NCONV)                         BOU 7500
555    FORMATCIX.'NODE # AND  TEMP  AT  INF.',8(2X,13,2X,F5.2))            BOU 7600
      GO TO 400                                                         BOU 7700
 300     CONTINUE                                                       BOU 7800
      DO 700  J=N7,N8,N9                                                BOU 7900
 700     HEAD(J)=T3-N11                                                 BOU 8000
 400     CONTINUE                                                       BOU 8100
      RETURN                                                            BOU 8200
C                                                                       BOU 8300
£**«***«****»****«**«*«»****««»**«*********»*««*«*«*                   BOU 8400
C                                                                       BOU 8500
C                                                                       BOU 8600
       ENTRY CHANG                                                      BOU 8700
C                                                                       BOU 8800
       RR=R(130) + R(13D + R( 145)+R(146)                                  BOU 8900
       RR=RR/4                                                          BOU 9000
         DO  10 J=1,LHEAT                                                 BOU 9100
       N=NHEAT(J)                                                       BOU 9200
10     AHEAT(J)=R(N)»PCW*AWATER(J)                                     BOU 9300
       RETURN                                                           BOU 9400
c                                                                       BOU 9500
       ENTRY CHAN                                                       BOU 9600
       RETURN                                                           BOU 9700
       END                                                              BOU 9800
                                      -191-

-------
       SUBROUTINE EIGEN(LN,MBAND.MP,KSYM,CFACT,S,T-RI,HEAD)             EIG  100
      DOUBLE PRECISION S,T                                              EIG  200
      DIMENSION HEAD(LN), RI(LN), SUN MBAND) ,T(LN,MBAND)               EIG  300
       COMMON/HI/TITLE(25),V(26),VV(26)                                 EIG  400
       PRINT 10                                                         EIG  500
      COMMON/HH/LFLOW.LON                                               EIG  600
C                                                                       EIG  700
      C=1000                                                            EIG  800
      DO 1 J=1,LN                                                       EIG  900
      A=0                                                               EIG 1000
       IF(KSYM.GT.O) GO TO 6                                            EIG 1100
       IF(HEAD(J).EQ 0) A=S(J,1)                                        EIG 1200
      DO 2 K=2,MBAND                                                   'EIG 1300
       IF(HEADU).NE.O) GO TO 1                                         EIG 1400
       E=S(J,K)                                                         EIG 1500
2      A=ABS(E)+A                                                       EIG 1600
      MBA=MBAND-1                                                       EIG 1700
      DO 4 K=1,MBA                                                      EIG 1800
      NO=J-K                                                            EIG 1900
      IF(NO.LE.O) GO TO 5                                               EIG 2000
       E=T(NO K+1)                                                      EIG 2100
4      A=ABS(E)+A                                                       EIG 2200
    5 CONTINUE                                                          EIG 2300
      B=T(J,1)».5/A                                                     EIG 2100
       GO TO 8                                                          EIG 2500
6      DO 7 K=1,MP                                                      EIG 2600
       IF(S(J,K).GT.CFACT) GO TO 1                                      EIG 2700
       E=S(J,K)                                                         EIG 2800
       E=ABS(E)                                                         EIG 2900
       IFU.EQ.MBAND) E=S(J,K)                                          EIG 3000
7      A=E+A                                                            EIG 3100
       B=T(J,MBAND)»0.5/A                                               EIG 3200
8      CONTINUE                                                         EIG 3300
       RI(J) = B                                                          EIG 3*»00
       IF(B.LT.C) LQ=J                                                  EIG 3500
      C=AMIN1(B,C)                                                      EIG 3600
    1 CONTINUE                                                          EIG 3700
       PRINT V,(RI(J),J=1,LN)                                           EIG 3800
      PRINT 11.C.LQ                                                     EIG 3900
       STOP                                                             EIG 4000
10     FORMAT(1X/1X,'STABILITY OF SOLUTION USING CRANK NICOLSON METHOD' EIG 4100
     «,4X/3X,'MAXIMUM TIME STEP FOR EACH NODE')/                        EIG 4200
11     FORMAT(1X/1X,'THE CRITICAL TIME STEP IS '.G12.6,'                EIG 4300
     •CONSTRAINED AT NODE',15)                                          EIG 4400
      END                                                               EIG 4500
                                     -192-

-------
c
c
c
c
SUBROUTINE THAT ALLOWS FOR  A MOVING  BOUNDARY  IN  A
       SUBROUTINE ADJUST(NJUST.LM.LN,Z,HEAD,XLOC,YLOC,
       I
       DIMENSION NMOV(40,2),  FLOWX(LM),FLOWY(LM),A(2),
       DIMENSION BMOV(40,6),HEAD(LN),XLOC(LN),YLOC(LN)
       READ,NSTEP.NPRINT,NPR,ERROR,FACTOR
        READ , (NMOV ( J , 1) , NMOV ( J , 2), BMOV ( J , 3) , BMOV ( J , 6 ) ,
       KLT=0
       KSS=1
       RETURN
        ENTRY  ADJUS(«,«)
        IFCKSS.EQ.2)  GO TO 5
        DO  4 J=1 .NJUST
        LL=NMOV(J,1)
        LL=ABS(LL)
        IF(LL.NE.O)  GO TO 2
        LL=NMOV(J,2)
        LL=ABS(LL)
        L=NODE(4,LL)
        GO  TO  3
 2      CONTINUE
        L=NODE(3,LL)
 3      CONTINUE
        BMOV(J,5)=YLOC(L)
 4      CONTINUE
 5      CONTINUE
        KLT=KLT+1
        READ,BKS,TIME,FACT
        DO 6 J=1,NJUST
 6      BMOV(J,4)=BMOV(J,6)*FACT
        DO 8 J=1,NJUST
 8      BMOV(J,4)=BMOV(J,4)/FACTOR
        SLENG=0.00001
        FLOW=0
        DO 40 J=1,NJUST
        DO 10 K=1,2
        NS=1
        LLL=NMOV(J,K)
        IF(LLL.LT.O)  NS=2
        LLL=ABS(LLL)
        IF(LLL.EQ.O)  GO  TO 10
        CALL  CORD(LLL)
        CALL  VCENTULL.VX.VY)
        LA=NODE(3,LLL)
        LB=NODE(4,LLL)
        AA=(XLOC(LA)-XLOC(LB))
        A(K)=ABS(AA)
        B(K)=VY
  10     CONTINUE
        LC=NMOV(J,1)
        LC=ABS(LC)
        IF(LC.EQ.O)  A(1)=A(2)
        IFUC.EQ.O)  B(1) = B(2)
        IF(LLL.EQ.O)  A(2)=A(1)
        IF(LLL.EQ.O)  B(2)=B(1)
         IFULL.EQ.O)  LB=NODE(3,LC)
VERTICAL CROSS SECADJ   100
                  ADJ   200
                  ADJ   300
                  ADJ   400
NODE,FLOWX.FLOWY.RADJ   500
                  ADJ   600
B(2),R(LN)        ADJ   700
,NODE(12,LM)      ADJ   800
                  ADJ   900
J=1,NJUST)        ADJ  1000
                  ADJ  1100
                  ADJ  1200
                  ADJ  1300
                  ADJ  1400
                  ADJ  1500
                  ADJ  1600
                  ADJ  1700
                  ADJ  1800
                  ADJ  1900
                  ADJ 2000
                  ADJ 2100
                  ADJ 2200
                  ADJ 2300
                  ADJ 2400
                  ADJ 2500
                  ADJ 2600
                  ADJ 2700
                  ADJ 2800
                  ADJ 2900
                  ADJ 3000
                  ADJ  3100
                  ADJ  3200
                  ADJ  3300
                  ADJ  3400
                  ADJ  3500
                  ADJ  3600
                  ADJ  3700
                  ADJ  3800
                   ADJ  3900
                   ADJ  4000
                   ADJ  4100
                   ADJ  4200
                   ADJ  4300
                   ADJ  4400
                   ADJ  4500
                   ADJ  4600
                   ADJ  4700
                   ADJ  4800
                   ADJ 4900
                   ADJ 5000
                   ADJ 5100
                   ADJ 5200
                   ADJ 5300
                   ADJ 5400
                   ADJ 5500
                   ADJ 5600
                   ADJ 5700
                   ADJ 5800
                   ADJ 5900
                   ADJ 6000
                   ADJ 6100
                   ADJ 6200
                                       -193-

-------
         IF(NS.EQ.2) GO TO 20
         C=(B(1)*A(1)+B(2)«A(2))/(A(1)/2+A(2)/2)
        GO TO 10                                                         ADJ 660°
 20     CONTINUE                                                         ADJ 67°°
        FLOW=B(1)»A(1)/2+B(2)«A(2)/2+FLOW                                ^ %*™
        SLENG=SLENG+A(1)/2+A(2)/2                                        A°, °?°9
 40     CONTINUE                                                         ™ 7000

 c                                                                       ADJ 7200
 C LAKE LEVEL LEVELLER

        HE=FLOW/SLENG
        DO 50 Jrl.NJUST
        IF(NMOV(J, 2). GT. -0.0001) GO TO 50
        LLL=NHOV(J,2)
        LLL=ABS(LLL)
        IF(LLL.EQ.O) GO TO 15                                              1
        MD=NODEC«,LLL)
        GO TO 1?
 15     LLL=NMOV(J,1)
        LLL=ABS(LLL)
        HD=NODE(3,LLL)
 17     CONTINUE
 50         lME                                                         ADJ  890°
 5U     CONTINUE                                                         ADJ

 C   PRINTING ROUTINE

        LP=KLT/NPRINT                                                        9300
        LQ=KLT/NPR
        LP=LP»NPRINT
        IF(KLT.EQ.LP)  CALL  WPRINT(BKS)                                    AD,
        IF(KLT.EQ.LQ)  CALL  WFLOW(BKS)                                     Jn]  qqOO
        IF(KLT.EQ.LQ)  CALL  BPRINT                                         ADJIOObo
        IF(KLT.GT.NSTEP)  STOP
        J|P n                                                              ADJ 10300
        S'TOJ.I.MJUST
       IF(LL.HE.O) GO TO 60                                             ADJ '0800
         -         )                                                         .UBUO
                                                                        ADJ 10900
                LL)                                                     ADJ11000
       rn  n  c                                                         ADJ 11 100
       GO TO 65                                                         Ani
60     CONTINUE                                                         ADJ
       L=NODE(3,LL)                                                     ADJ
65     CONTINUE                                                         ADJlisOO
        AD=BMOV(J,5)-YLOC(L)                                            ADJ^SOO
       AD=ABS(AD)                                                       ADJ 11700
       AC=AHAX1(AD,AC)                                                  ADI^Rnn
70     CONTINUE                                                         JSJii'SSS
       IFCAC.GT. ERROR) KSS=1                                            ADJ12000
       IF(KSS.EQ.2) RETURN 2                                            ADJ12'00
       PRINT nO.BKS.KLT                                                ADJ12200
110    FORHAT(1X/1X,'STUCTURE MATRIX RECOMPUTED AT TIME STEP   '.G8.3,   ADJ12300
     "   "SR*TION  ''I5)                                              ADJ 12400
       ^IURN 1                                                         ADJ 12500
       EMD                                                              ADJ 12600


                                     -194-

-------
                                  APPENDIX D

                  ESTIMATION OF AQUIFER PARAMETERS BY USING
                         SUBSURFACE TEMPERATURE DATA


    Flowing ground water* distorts the normal distribution of subsurface
temperatures.  Researchers interested in determining the geothermal heat flux
have long been aware of the distorting influence moving ground water can have
on the measured geothermal gradient (Kirge 1939).  Only a few ground-water
researchers, however, have attempted to use subsurface temperature information
to calculate the rate and direction of ground-water flow.

    Stallman (1960) aroused interest in the use of subsurface temperature as
an indirect manifestation of ground-water velocity with a presentation of a
differential equation describing heat transport in the subsurface and with the
suggestion that temperature measurements might be a useful means for
indirectly determining aquifer characteristics.  An analytical solution was
developed by Stallman (1965) for determining ground-water velocities in a
homogeneous medium when the boundary conditions of heat and water movement
are, respectively, (1) a sinusoidal temperature fluctuation of constant
amplitude at the land surface, and (2) a constant and uniform percolation rate
normal to the land surface.  Bredehoeft and Papadopulos (1965) developed an
analytical solution for determining the ground-water velocity from subsurface
temperature data in an isotropic, homogeneous, and fully saturated
semiconfining layer in which all flow is vertical.

    The analytical solution developed by Stallraan (1965) has not been used
extensively.  Its main drawbacks are the assumptions that the water table is
at the surface and that all flow is vertical.  Taking a different approach,
Nightingale  (1975) estimated vertical recharge rates from an infiltration pond
with a sinusoidal temperature distribution, but he assumed that conductive
transfer processes were negligible and that velocity could be calculated by
using only the lag between surface temperatures and temperatures at a point in
the subsurface.

    Several  researchers have used Bredehoeft and Papadopulos's analytical
solution.  Cartwright (1970) successfully used temperature anomalies to
estimate the amount of vertical movement of ground water in  the Illinois
Basin.  He matched the average temperature profile to the 3=-1 curve of
Bredehoeft and Papadopulos  (1965) to obtain a discharge rate of 1.52 cm/yr
from the deep aquifer, a value that agrees with estimates of ground-water
discharge into streams in the basin.  Sorey (1971) used the  Bredehoeft and
Papadopulos  technique to estimate the rate of upward movement through
semiconfining beds in the San Luis valley of Colorado and the Roswell basin of
New Mexico.  His conclusions suggest that pumping tests and water-budget

                                    -195-

-------
methods are often preferable because  of limitations imposed  by  instabilities
in the borehole fluids and the measurement detail  required.   Boyle and Saleem
(1978) used the technique to estimate vertical  flow rates through a clay-rich
glacial drift semiconfining layer in  the Chicago area.   They obtained good
agreement with values calculated by using the water-budget method.

    The available analytical techniques only provide solutions  for a very
small group of problems in which flow is in the vertical direction and
boundary conditions are specialized.   In most aquifers  the dominant direction
of water movement is horizontal.  Noticeable temperature variations have been
observed in the horizontal direction  in ground-water systems (Winslow  1962,
Schneider 1962 and 1964, Mink 1964, Parsons 1970,  Suptow 1971,  Cartwright
1973).  The researchers all implicitly assumed  that the variations in
temperature were caused by ground-water flow, but  none  were  able to quantify
flows by using this information.

    In the course of the research reported here an attempt was  made to use  the
measured horizontal and vertical distribution of subsurface  temperatures to
estimate the rate of flow from a cooling lake situated  on an alluvial  aquifer.
A numerical technique was used by which ground-water velocities and hydraulic
conductivities in a two-dimensional system with nonuniform boundary conditions
can be determined by a trial-and-errror procedure  from  subsurface temperature
information.  The procedure is not recommended  for use  in routine ground-water
flow system analysis since traditional methods  for defining  flow systems are
simpler to use and more reliable.

MATHEMATICAL MODtL

    Given the following assumptions:   (1) thermal  equilibrium between  the
liquid and soil particles is achieved simultaneously, (2) the density  of the
soil particles is constant, (3) the heat capacity  is constant,  and  (4) the
system is chemically inert, then the  general differential equation  for
simultaneous heat and fluid flow in a two-dimensional aquifer is

                              8(q,T>
where T = temperature, T; DJJ = coefficient of dispersion, H/tTL;  pCw = heat
capacity of water, ti/L^T; pCs = heat capacity of the saturated media, H/L^T;
Qi = specific discharge or ground-water velocity, L/t; R = rate of heat
injection or discharge, H/L^t; and xj,S2 = cartesian coordinates L.

    The goal is to solve Eq. (D-1) for the velocity distribution;  then
hydraulic conductivities can be determined from
where KJJ = hydraulic conductivity tensor, L/t; and  4 = head, L.

    Velocities in the aquifer can be determined from Eq. (D-1) either directly
or indirectly according to whether velocities are obtained directly by using

                                    -196-

-------
the temperature distribution as a known in the differential equation, or
whether velocities are obtained as a solution to a nonlinear optimization
problem in which a set of calculated temperatures is matched to an  observed
set.  Direct methods require that the temperature distribution be known
completely and that it closely represent the true solution to the differential
equations (Neuman 1973).  These types of solutions to the inverse problem are
currently a subject of active research and were not used in the present study.

    The equation describing the two-dimensional flow of water through a
nonhomogeneous aquifer in steady state may be written as
                         i        xj                                     (0-3)

where h = water recharge rate per unit area, L/t; and b = thickness, L.

    The ground-water velocity or specific discharge can then be determined
from
    Acquisition of some information on subsurface permeability is desirable
for initial estimates of velocities by using the general differential  equation
describing water flow in an aquifer.  Hydraulic conductivities can  then  be
adjusted until the computed velocity distribution, when input  into  Eq. (D-1),
produces a temperature distribution that is reasonably close to the observed
temperature data.

    The parameter estimation problem could be viewed as a classical nonlinear
regression problem in which a solution to the linked differential equations
[Eq. (D-1), Eq. (D-3), and Eq. (D-4)] forms the regression equation and  in
which all unknown quantities are parameters (Cooley 1977).  The problem  was
not viewed in this manner because (1) sufficient temperature data will
generally not be available to insure a well-conditioned solution, and  (2) Eq.
(D-1 ) behaves as a hyperbolic paraboloid when convective transport  dominates,
which means that small errors in the input parameters can cause large  errors
in the output .

    Instead, a trial-and-error procedure was used to estimate  hydraulic
conductivities.  This technique is conceptually simple, but has several
drawbacks:  (1) It can be expensive, (2) it is very time consuming, (3)  an
answer cannot always be obtained, and (4) if an answer is obtinaed, it is
probably not the best estimate, and its relation to the best estimate  is
unknown .

    Several difficulties are encountered when Eq. (D-1) and Eq. (D-3)  are
linked and solved in a trial-and-error procedure to estimate hydraulic
conductivities.  The best documented difficulty is that, when hydraulic
conductivities are known, researchers have not been very successful in solving
for a known conservative contaminant distribution, even though the
conservative mass transfer problem is mathematically simpler than the  heat

                                    -197-

-------
 transport problem.  The differential equations are of the same form,  but  the
 conservative mass transport problem has two fewer parameters because  of the
 common assumption that molecular diffusion, which is equivalent to thermal
 conductivity, is negligible.

     Attempts to simulate a known conservative contaminant distribution in an
 aquifer with a deterministic model are not numerous.  Finder (1973) simulated
 the  observed chromium contamination on Long Island;  Bredehoeft and Finder
 (1973) simulated the known chloride distribution in the Brunswick aquifer;
 Konikow (1976) simulated the chloride distribution in the vicinity of the
 Rocky Mountain Arsenal; Robertson (1974) simulated chlorides and other
 contaminants in the vicinity of the Idaho Test Site; and fiobson (1978)
 simulated total dissolved solids in a shallow alluvial aquifer near Barstow,
 Calif.  In all these simulations velocities within the aquifer were determined
 by using equations similar to Eq. (D-3) and Eq. (D-4).  In these equations
 hydraulic conductivities were adjusted by a trial-and-error procedure until
 predicted aquifer heads closely matched observed aquifer heads.  An equation
 similar to Eq. (0-1) was used to simulate transport of contaminants,  and  the
 porosity and dispersivity parameters were then adjusted until the observed
 chemical concentration patterns were matched.  Even when hydraulic
 conductivities are known, the chemical concentration distribution could not be
 simulated without a trial-and-error adjustment procedure.  Also, even after
 the  researchers had obtained what they considered to be a best fit between  the
 simulated and the observed data, the fit in all cases was less than perfect.
 Anderson (1979) discusses some of the problems involved in applying
 contaminant transport models.

     Since the temperature data will generally be sparse, the trial-and-error
 procedure of parameter estimation for the linked Eq. (D1) and (0-3) is
 generally non-tractable unless several of the parameters are constrained.  In
 this analysis, in which hydraulic conductivities in a steady-state water-flow
 problem were estimated, thermal conductivities, heat capacities, and
 dispersivities were fixed.  The rationale fbr fixing these parameters was that
 the  thermal conductivity and the heat capacity of most saturated glacial
materials fall within a small range, and they can be measured accurately  in
 the  laboratory.  Dispersivities were assumed to be small because element  sizes
were chosen so that intra-element inhomogeneities were minimized.

    These procedures left only hydraulic conductivities to be adjusted.
Generally, even with hydraulic conductivities as the only unknowns, problems
were ill conditioned .unless the hydraulic conductivities were constrained to
be within a small range.  Therefore, this technique is only useful for
refining estimates of hydraulic conductivity that are determined with other
procedures.

    The following procedure was used to refine estimates of the hydraulic
conductivity distribution at the Columbia Generating Station site by  using
subsurface temperature data:

 1) Hydraulic conductivity distributions were proposed on the basis of
   stratigraphic information and field tests, and these distributions were


                                     -198-

-------
   then tested using Eq. (D-3) to determine if the known potential
   distribution could be predicted within a set error criterion.

2) Once the potential distribution could be predicted, the relative magnitude
   of the hydraulic conductivities was adjusted in a model linking Eq.  (D-1)
   and Eq.  (D-3) to determine which factor gave the best fit to the observed
   temperature data.

3) If the best fit of the temperature data obtained in step 2 was not good,
   steps 1  and 2 were repeated.  If the fit was judged to be acceptable,
   thermal  conductivities and dispersivities were changed to determine  the
   sensitivity of the estimate to changes in these parameters.

SOLUTION PROCEDURE

    The finite element method was used to solve Eq. (D-1) and Eq. (D-3) for
temperatures in an aquifer subject to the boundary conditions described in
section 5 and appendix B.  Each cross section modeled was divided into  100-150
quadrilateral elements.  The procedure used to link the equations was:  (1)
Eq. (D-3) was solved for head at each node and Eq. (D-4) was solved for
velocity; (2) £q. (D-1) was solved for temperature at each node;  (3) the
solution to Eq. (D-1) was stepped forward in time by using the Crank-Nicolson
approximation for the tine derivative for a specified number of time steps;
(4) the hydraulic conductivities, which are a function of temperature, were
adjusted for the new temperature distribution; and (5) steps (1)  to (4) were
repeated.

FIELD DATA

    Temperatures were recorded weekly at 40-110 points in the subsurface in
the vicinity of the Columbia Generating Station (figure A-1) from August 1976
to January  1978.  The monitoring techniques and the locations of the data
points are  described in appendix A.  The data collected on 7 October 1977 are
presented in a fence diagram in Figure (D-1).  The data collected in
cross-section A-A' and B-B'  of Figure 6 are presented in section 5.
Temperatures recorded in an array of wells on the east side of the cooling
lake are shown in Figure (D-2).

RESULTS

    The temperature data from cross-sections A-A' and B-B1 were used to refine
estimates of hydraulic conductivity and to refine previous estimates of flow
in these cross sections from the cooling lake to the wetland west of the lake.
Andrews (1976) modeled flow in these cross sections and obtained average
annual flow rates of 4.5 n^/day (m^/day per meter width of the dike and 3.5
rn^/day, respectively.  By using the trial-and-error procedure described in
this appendix, values of 5.2 m^/day and 4.3 n^/day were obtained for now in
these two cross sections, respectively.  An error criterion of 0.05 m was used
in step 1.   Thermal conductivities were set equal to values obtained in the
laboratory.  Longitudinal dispersivity was set to 20 cm, and transverse
dispersivity was set to 5 cm on the basis of intro-element homogeneity
considerations.

                                     -199-

-------
                  TwnperaturM in the Marsh on October 7, 1977
Figure D-l.  Temperatures (°C) in a section of the marsh adjacent  to  the
             cooling lake at the Columbia Generating Station  site  on
             7 October 1977.  The three-dimensional diagram depicts
             temperatures in a 2,500-m section along the dike which
             extends outward from the dike for a distance  of  110 m and
             to a depth of 10 m.
                                   -200-

-------
   LAKE
   2


   4


   6
                                               JUNE9
                                           7.0"
    10°
12.4°
                                          8.8°
              10
  30
50     METERS
                                           OCTOBER 24
                                  30
                      50     METERS
Figure D-2.   Temperatures in a 60-m portion of cross-section B-B' of
             Figure 6 adjacent to the drainage ditch east of the cooling
             lake on 9 June 1977 and 24 October 1977.
                                 -201-

-------
    The hydraulic conductivity distribution that  best  reproduced  the observed
temperature data was not sensitive  to  changes in  the  thermal  conductivities or
dispersivities within reasonable limits.   Adjusting the  thermal conductivities
and the dispersivities did,  however, reduce the residuals  between the  observed
and the simulated temperatures.   The changes in simulated  temperature  that
resulted when dispersivities were changed  are described  in section 5.

    This trial-and-error procedure  does not lend  itself  to estimation  of the
standard error of the parameters in the best fit  mode.   It is only an
intuitive observation that the standard error of  the  estimated hydraulic
conductivities are reduced by using this tehcnique rather  than stopping  at
step  1 in the parameter estimation.

    The temperature data were also  used to attempt to determine flow rates  in
five other cross sections (Figure 15)  at the Columbia Generating  Station site.
In all these cross sections only limited potential data  were  available,  and
several hydraulic conductivity distributions could be found that  would
reproduce the known potential distribution equally well.   Since the
temperature data were also limited, several hydraulic  conductivity
distributions which represented widely varying flow rates  could reproduce the
observed data equally well.   Only in the simulated cross section  located near
the intake of the cooling lake were the results satisfactory. (Of the five
cross sections, the best temperature and potential data  were  available for
this  cross section.) In this cross  section the flow rates  into the wetland
were  estimated to average 7.4 nrVday.

CONCLUSIONS

    Unfortunately, the information  in  Figure (0-1) was not sufficient  to
estimate flow rates from the cooling lake  into the wetland along  the dike.
Figure (0-1) does illustrate vividly the complexity of ground-water flow
patterns and hydraulic conductivity distributions near the dike.   Although
flow rates from the cooling lake to the wetland in the area shown in Figure
(D-1) are most likely all within the range of 3.5-7.5 rn^/day, the temperature
contrasts are dramatic.  Small changes in  peat thickness or the presence or
absence of a clay lens can cause very  different temperature patterns even
though flows are approximately the  same.   Keys and Brown (1978) reached  a
similar conclusion.  The temperature data, although indicating that flows are
much greater in some areas than others, are not sufficient for determining
hydraulic conductivity distributions.
                                    -202-

-------
                                  TECHNICAL REPORT DATA
                          (Please read Instructions on the reverse before completing)
 REPORT NO.
                            2.
                                                          3. RECIPIENT'S ACCESSION NO.
TITLE AND SUBTITLE
  Impacts of Coal-Fired Power Plants  on  Local
  Ground-Water  Systems
Wisconsin Power Plant Impact  Study
                                                          5. REPORT DATP
                                                               August 1980 Issuing Date.
                                                           i. PERFORMING ORGANIZATION CODE
 AUTHOR(S)
  Charles B. Andrews
  Mary P. Anderson
                                                          8. PERFORMING ORGANIZATION REPORT NO.
 PERFORMING ORGANIZATION NAME AND ADDRESS
  Institute for Environmental Studies
  Environmental Monitoring  and Data Acquisition  Group
  University of Wisconsin-Madison
  Madison, Wisconsin   53706	
                                                          10. PROGRAM ELEMENT NO.

                                                            1BA820
                                                          11. CONTRACT/GRANT NO.

                                                            R803971
2. SPONSORING AGENCY NAME ANOADDRESS
   Environmental Research Laboratory
   Office of Research and Development
   U.S.  Environmental Protection Agency
   Duluth, Minnesota  55804	
                                                          13. TYPE OF REPORT AND PERIOD COVERED
                                                           Final  7/75 - 6/78
                                                          14. SPONSORING AGENCY CODE
                                                               EPA/600/03
5. SUPPLEMENTARY NOTES
6. ABSTRACT Quantitative  techniques for simulating  the impacts of a coal-fired power
plant on the ground-water system of a river  flood-plain wetland were developed and
tested.  Effects related to the construction and  operation of the cooling lake and ash-
pit had the greatest  impact.   Ground-water flow system models were used to simulate
ground-water flows  before and after the cooling lake and ashpit were filled.  The sim-
ulations and field  data indicate that the cooling lake and ashpit altered local flow
systems and increased ground-water discharge.   Chemical changes in the ground-water
system were minor.  Contaminated ground water was confined to a small area near the
ashpit.  Thermal changes in the ground water are  a major impact of the cooling lake.
Changes in water temperature and levels have altered the vegetation of the wetland, a
major ground-water  discharge area.  Ground-water  temperatures near the cooling lake
were monitored.  A  model was used to simulate the response of subsurface temperatures
to seasonal changes in a lake and air temperatures.  Long-term substrate temperature
changes expected in the wetland were predicted.  Using ground-water temperatures to
estimate flow rates was investigated.  Simulated  temperature patterns agreed with
field data, but were  sensitive to the distribution of subsurface lithologies.  It is
predicted that by 1987 ground-water temperatures  will be increased, resulting in an
increase in ground-water flow.
                               KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS C.  COSATI Field/Group
Ground, water models, cooling lakes,  tempera
ture monitoring, river flood-plain wetlands
coal fired power plants, ash pit, flow rate:
vegetation
                                               Wisconsin Power Plant
                                               Study, Impacts on Biota
                                               Flow measurement, terraii
                                               models, artesian water,
                                               cooling water
                           06-F,  T
                           08- D,  F,  H,  M
                           10-C
                           13-B,  M
18. DISTRIBUTION STATEMENT
       Release to the public
19. SECURITY CLASS (ThisR
  unclassified
                                                                           215
                                              20. SECURITY CLASS (Thispage)
                                                unclassified
                                                                         22. PRICE
EPA Form 2220-1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE
                                                                      PRINTING "FFICF: 1 980--657-1 5?/OOB4
                                            203

-------