MOCDENSE - two-constituent solute transport model for ground water having variable density
MOCDENSE Categories: flow models - saturated zone, saltwater intrusion models, solute transport models - saturated zone, solute transport models - unsaturated zone
MOCDENSE is the USGS model "A Two-Constituent Solute Transport Model for Ground Water Having
Variable Density" by L.F. Konikow and W.E. Sanford.
- Applicable to saltwater intrusion problems.
- Simulates solute transport and dispersion of one or two constituents in ground water with density dependent flow.
- Solves for fluid pressure rather than hydraulic head because of variable density.
- Flow is in a cross-sectional plane rather than a horizontal plane.
- Handles varying recharge, aquifer heterogeneities, variable aquifer thickness and complex boundary conditions.
- Method of Characteristics solves transport equation; finite-difference method solves flow equation.
- Density is considered a function of the concentration of one of the constituents.
- Grids to 150 X 150; saves time-history data for up to 25 observation points.
- Model results compare favorably with those of other numerical models.
- Executes in an easy-to-use, menu-driven batch file system.
- Includes source and executable codes: MOCDSIP (SIP Solver) and MOCD4
(Direct Solver); Input Processor, MOCDINP, Output Processor, MOCDOUT, MOCREC and TECVIEW.
Various numerical models have recently been developed that simulate variable density fluids in which the concentration of the solute of interest affects the
density of the fluid. These models typically have been applied to problems of sea water intrusion in coastal aquifers. However, there are many problems in which contaminants are introduced into an aquifer near the interface or
transition zone between freshwater and saltwater. Examples include the injection of waste water into coastal aquifers. In such cases, the injection will affect the fluid pressure and flow of both the freshwater and saltwater,
but the contaminants being injected are generally in such low concentrations that changes in concentration of the contaminants will not affect the fluid density. Simulation of such problems thus requires the ability to simulate
the simultaneous flow of variable-density ground water and the transport and dispersion of at least two solutes or soluble constituents. The fluid density needs to be related to the concentration of one of the constituents,
which in practice can be either salinity, dissolved-solids concentration, specific conductance, or chloride concentration. The objective of this report is to document a numerical simulation model that is applicable to these
types of problems.
is a modified version of the ground-water flow and solute-transport model of Konikow and Bredehoeft, which was designed to simulate the transport and dispersion of a single solute that does not affect the fluid density. This modified version simulates the flow in a cross-sectional plane rather than in an areal plane. Because the problem of interest involves variable density, the modified model solves for fluid pressure rather than hydraulic head in the flow equation; the solution to the flow equation is still obtained using a finite-difference method. Solute transport is simulated with the method of characteristics as in the original model. Density is considered to be a function of the concentration of one of the constituents. Use of this model depends on assumptions that:
- Flow is two-dimensional, with one of the principal axes being parallel to gravity,
- Constituents are conservative (nonreactive), and
- Density and viscosity are a function of concentration and not of other factors such as pressure and temperature.
These assumptions are often valid approximations where an aquifer system contains both freshwater and saltwater. This model is applicable in such situations
where, ln addition to that of the density-controlling species, the movement and concentration of another chemical species, such as a dissolved pollutant, needs to be predicted. MOCDENSE is also applicable to a two-constituent
system with no density-dependence, given that the other assumptions are valid, and to a single-constituent system with variable density.
The maximum dimensions of the flow and transport grids have been increased to 150x150. The limit on the number of particles has been increased from 3,200 to
80,000. The number of observation points which may be specified has been increased from 5 to 25. Observation point data (Density controlling and trace concentrations) may now be saved in a file for plotting with MOCDTIME (not
MOCDENSE utilizes extended memory.
This package contains two versions of MOCDENSE: MOCDSIP and MOCDD4. These two versions of MOCDENSE are identical except for the matrix solver. MOCDSIP uses a
Strongly Implicit Procedure (SIP) to numerically solve the flow equations while MOCDD4 uses a direct solver with D4 ordering. Also included in the package are the following programs: MOCDINP, MOCDOUT, MOCREC and TECVIEW.
is a program which uses Virtual Memory and page swapping to view and print portions of very large ASCII files. An amount of disk space slightly larger than the size of the file being viewed must be available.
is the input processor for the MOCDENSE model. MOCDINP. MOCDINP is used to Create or Modify the Data File required by MOCDENSE (MOCDD4 or MOCDSIP).
The USGS MOCDENSE report and associated MOCDENSE Computer Update Notes describe all inputs. MOCDINP provides a convenient means of producing the Input Data
Files required by MOCDENSE. See below for a description of changes to the structure of the Input Data Files required by MOCDENSE.
The number of nodes in the X and Z directions, NX and NZ, are restricted to 150 and 150, respectively. These values of 100 and 100 for the maximum number of
nodes compare to 24 and 20 in the original MOCDENSE source code.
Two new flags, NHYDRO and NOUTPR, have been added to Card 2 of the MOCDENSE data files by TECSOFT. If NHYDRO is non-zero and NUMOBS is non-zero, the Density
Controlling and Trace Concentrations for all defined observation points will be saved in an ASCII file. The TECSOFT program, MOCDTIME, may then be used to obtain plots of data automatically. NOUTPR is a flag used to reduce the
size of the hardcopy output file. The flag may have the following values:
0: Normal hardcopy output file
1: Eliminates printing of observation point data from the hardcopy output file
2: Eliminates printing of particle tracking data from the hardcopy output file
3: Eliminates both
is the Output Processor for the MOCDENSE model. MOCDOUT takes, as input, the concentration/velocity vector unformatted file from MOCDENSE (MOCDSIP or MOCDD4) and permits the user to save selected sets of data in an ASCII file. The ASCII file may be later processed by MOCDGRAF to produce graphics.
The hardcopy output file from a MOCDENSE (MOCDSIP or MOCDD4) run contains FORTRAN Carriage Control characters. The FORTRAN Carriage Control characters
are ignored by most dot matrix printers. The hardcopy output, therefore, may be difficult to read. The MOCREC (Remove or Execute Control Characters) program removes the FORTRAN Carriage Control characters and substitutes ASCII
(not included) can automatically process any or all concentration (trace and density controlling solutes) or velocity-vector grids contained in the file produced by MOCDOUT and stack the contour/velocity vector plots in a metacode graphics file for subsequent processing with TRANSLATE. The graphics file produced by MOCDGRAF could contain up to 500 plots.
MOCDENSE PROGRAM USAGE
executes in a unique batch processing environment designed by TECSOFT INC. The batch processing environment is initiated by typing MOCDENSE. The four (4) major functions within the batch shell are: FILES, SELECT, EXECUTE and QUIT. FILES permits the creation (CREATE) or modification (MODIFY) of batch files necessary for execution. Any number of MOCDENSE batch files may exist concurrently. Prior to execution, a batch file must be selected. After entering SELECT, a batch file is selected using a file pick list and its contents are displayed. Next, the EXECUTE function is selected. A choice of either solver (MOCDSIP or MOCDD4) may then be made. Following execution, QUIT (or <ESC> key) exits the batch shell. A mouse may be used.
Help is available for most functions within the shell by pressing the <F1> function key while highlighting that function. Exit help by pressing the
The following sequence applies to a MOCDENSE simulation:
1. Create the MOCDENSE Data File.
2. Run MOCDENSE. Create or Modify a Batch File. Select a Batch File. Choose MOCDSIP or MOCDD4. For most problems MOCDD4 should be used. However,
for steady-state cases when grid nodes have zero transmissivity and for cases in which the transmissivity is highly anisotropic, MOCDSIP should be used. The output files are: a Hardcopy Output File and an Unformatted File (if
NPNCHV in the data file is non-zero). During Batch File Creation or Modification, if NPNCHV is non-zero, the program prompts for the files (Data File, Hardcopy Output and Unformatted) and the types of data (concentration,
velocity vector) which will be saved in the Unformatted File. Any or all may be saved.
3. Process the hardcopy output file with MOCREC, if desired, to remove carriage control characters and replace with appropriate ASCII control
4. Run MOCDOUT
if desired. MOCDOUT takes as input the unformatted file from MOCDSIP (or MOCDD4) and outputs an ASCII file. Unformatted data may have been saved for all simulation time steps (NPNCHV=1). Rather than process all of this data with MOCDGRAF, when only a subset is required, MOCDOUT gives the user an opportunity to thin out the data while converting to ASCII.
5. Run MOCDGRAF
to contour data and draw velocity vectors. MOCDGRAF is used to contour concentrations (both the trace and density-controlling solutes) and to generate velocity vectors which may be superimposed on the contours. MOCDGRAF reads and processes the ASCII file from MOCDOUT.
6. Run MOCDTIME
to plot time vs. Density Controlling concentration or Trace Concentration at specified observation points. MOCDTIME reads an ASCII file produced by MOCDENSE.
MOCDGRAF and MOCDTIME are separate programs specifically developed for use with MOCDENSE. These programs are not included with the MOCDENSE package.
MOCDENSE Code Modifications
Certain necessary changes were made to the MOCDENSE version to accommodate the larger arrays.
- Arrays sizes were modified throughout the code to reflect the increase in the maximum number of X,Z nodes from 24x20 to 150x150.
- The limit for NPMAX has been increased from 3,200 to 80,000 because of the larger permissible grids. To accommodate this increase, the format in Card 2 for
NPMAX was changed from I4 to I5.
- The maximum number of observation points has been increased from 5 to 25.