GEOS 5311 Lecture Notes: Introduction to

Groundwater Modeling
Dr. T. Brikowski

Spring 2013

file:introduction.tex,v (1.10, January 23, 2013), printed January 23, 2013

Purpose of the Class

Learn to apply quantitative modeling techniques wisely to

obtain useful and valid results

An enormous number of quantitative tools exist in hydrology

(see Section 2)

The challenge is to learn basic rules and principles that are

useful no matter which tool is selected

We will do this in lecture, gaining hands-on experience using

one general computer interface (GMS1 , and the most common
flow (MODFLOW) and transport (MT3D) model programs.
Note GMS development was funded by the USACE.

http://www.aquaveo.com

What is modeling?

Model - any construction that approximates a field situation

I

conceptual model: a qualitative approximation of a system

used to understand its gross character (this is how we think it
works . . . )
mathematical model: mathematical construction, based on
governing equation(s) and boundary/initial conditions that are
thought to approximate a field setting
numerical model: computer approximation of mathematical
model

Mathematical Model Solutions

To make predictions or analyses, the system of governing

equations must be solved
Solutions take two basic forms:
I

analytic solution: an exact or algebraic solution of the

mathematical model, requires simple problem
numerical solution: an approximate solution, usually generated
on a computer, allows for great complexity

Uses of Models

Predicting response of system to specific, isolated changes

Studying effect of various processes on the system

Limited ability to predict future state of system

because of limitations in data (most often

incompletely-characterized heterogeneity)
because of limitations in theory (poorly understood
phenomena, e.g. hysteretic behavior in unsaturated flow)

Future Directions
I

historically groundwater modeling was motivated by (in

approximate order of development):
I

future large-scale projects or concepts:

I

introduction of EPA regulation of hazardous chemical spills

(CERCLA/Superfund), e.g. Love Canal
Federal waste storage projects, e.g. Yucca Mountain, WIPP
site, Rocky Flats, Pantex
climate-change impacts on water supply

future tools:
GSFLOW free USGS program for coupled
groundwater-surfacewater interaction (see USGS
website)
VIC main model currently used for runoff forecasting
in the Western U.S. See U. Washington website

Hydrologic Model Types

Model types can be sorted most generally by their governing

equation (van der Heijde et al. 1985; Mangold and Tsang 1991)
(also IGWMC software reviews2 ):
I

Flow model - solves water mass balance equation, dependent

variable is pressure or head, velocity is a derived quantity

Transport model - solves water and dissolved species mass

balance equation, dependent variable is concentration. More
difficult to solve numerically.

http://igwmc.mines.edu

Available Flow Models: USGS

See USGS WRD website3 for more details.
I Prickett-Lonnquist (aka PLASM, Prickett and Lonnquist
1971)
I

Modflow (Harbaugh 2005; Harbaugh et al. 2000; McDonald

and Harbaugh 1988)
I
I

The most popular flow modeling program from mid-1970s to

mid-1980s.
Finite difference (ADI), original versions required
customization for each setting (i.e. no provision for general
boundary conditions).

industry standard for 20 years

3-D saturated flow (confined or unconfined), finite difference
(block-centered grid).
Written to be a modular (easily-updatable) version of
Prickett-Lonnquist. Many add-on modules available (e.g. Hill
1990).

Sutra (Voss 1984)

I

2-D variable-density flow code (i.e. includes buoyancy effects)

Available Flow Models: USGS (cont.)

models fluid flow or transport of a single (density-affecting)

species, i.e. heat or Cl .
Developed for work on saline-water intrusion problems,
applicable to simple geothermal-hydrothermal settings.

http://water.usgs.gov

Other Flow Models

Other useful flow models:
I TOUGH+ (supercedes TOUGH, Pruess 1991, 2004; Pruess
et al. 1999)
I

WhAEM, ModAEM, GFLOW2000

I

uses Analytic Element Method (Haitjema 1995; Strack 1989),

really just superimposed analytic solutions

FEMwater
I

multi-phase (gas and liquid) flow code, developed for work in

nuclear waste disposal, derived from geothermal modeling
codes.
integrated finite difference (boundary conditions specified
unnaturally, unstructured grid). Difficult to use successfully
since models highly non-linear equations.
about the only useable code for volatile organic compounds
(e.g. underground gasoline vapors)

Finite Element flow code, allows great flexibility in grids (2D

and 3D)

HYDRUS2D/3D

Other Flow Models (cont.)

I
I

vadose zone modeling, very popular

most used as Modflow package

Available Transport Models: USGS

See USGS WRD website4 for more details.
I Method-of-Characteristics (or MOC, Konikow and
Bredehoeft 1978)
I

MT3D (Zheng 1992, 1995)

I
I

Solves flow equation on discrete time steps, then transport as

a steady-state process at each time step.
Robust approach (stable for large time and space steps), but
limited accuracy. The most commonly used model in the 70s
and early 80s.
Adds MOC to Modflow.
Probably the most popular transport model in use today. Also
robust and computationally efficient.

http://water.usgs.gov

Other Transport Codes

RT3D:
I

Reactive transport code, adds rock/matrix chemical reaction

to MT3D. From PNL.
same capabilities now in MT3D

TRACR3D (Los Alamos National Lab, (Travis and Birdsell

1991). Efficiently written finite-difference program, includes
two-phase flow, deformable and reactive porous media
modeling.

Reaction Models
I

These compute water-rock reactions, including evolution of

waters along a flowpath.

Essentially a multi-component mass and electrical charge

balance, with thermodynamic rules (reactions) used to
determine distribution of species
Main codes are:

PHREEQC (Parkhurst 1995, USGS)

I

Charge balance, speciation & saturation index calculations

(i.e. chemical significance of individual analyses);
reaction-path modeling (prediction of water-rock interaction,
or deduction of paths from analyses at differing locations).
Command-line interface, but quite robust and applicable for
many problems

Geochemists Workbench (Bethke 1996). Nice point-and-click

interface, similar functionality as PHREEQC, with more ability
to model reaction paths. Expensive.

Reaction Models (cont.)

EQ3/EQ6 (Wolery 1979): water-rock interactions over wide

range of T & P. Most useful for nuclear waste problems,
hydrothermal systems, etc. The most complicated and
thorough reaction model readily available.

Coupled-Process Models
At the leading edge of current modeling practice are models that
attempt to do it all:
I

usually simulate combined flow, transport and

deformation/deposition.

primarily involved in oilfield reservoir models (e.g. secondary

and tertiary production efforts)

also important for advanced waste disposal problems, e.g.

high level nuclear waste (TOSPEC, Yucca Mountain)

many appear as new modules for Modflow:

PHAST reactive transport combining Modflow with
PHREEQC. See PHAST homepage
HYDRUS vadose zone model for accurate recharge
estimation to saturated zone. See overview
article

Why model?

Hypothesis testing

System exploration (interpretation, study of process) e.g.

(Brikowski and Norton 1989)
Prediction

the basic goal of most consulting contracts (i.e. this is what

pays well)
VERY RISKY because (Konikow 1986; Konikow and
E. P. Patten 1985):
I

geologic systems will always be incompletely characterized

[surprise, ][]Bredehoeft-2005
conditions may change in the future

Use and Misuse of Models

General procedures in modeling (Mercer and Faust

1980, 1981)

Hypothesis testing: Aquifer-Lake interaction (Krabbenhoft

and Anderson 1986)

System exploration: Heat pipe phenomena (Pruess 1985)

Misuse: Inputs pre-determine results (Fehn and Cathles 1979)

Demonstration of Modeling Procedure

For example, development of a model of Hays, KS valley aquifer
I Develop conceptual model of site
I

Typically will have surface topography/culture (e.g. USGS

topo maps, Hays location figure), geologic maps/crosssections, and some borehole information (usually well lithology
logs, some hydrologic tests or observations, Hays boreholes
figure)
From this break system down into hydrostratigraphic units,
system boundaries and source/sinks
I

at Hays years of observation indicate three units, bedrock

Cretaceous shale of very low permeability, overlying Qal
sand-gravel aquifer of high permeability, generally overlain by
Qt silt-clay overbank deposits of moderate-low permeability
lateral aquifer boundaries are generally defined by the geologic
map (Kc-QT contact at the surface)
vertical boundaries (top and bottom) are defined from
borehole information (demo of shale-top picking in GMS)

other properties, such as hydraulic conductivity, are fit

parameters for model

Demonstration of Modeling Procedure (cont.)

Calibrate model
I

carry out flow model, calculate residual error with observed

head

Models

Two main models of this system have been made

I

DOE-Yucca Mountain (Eddebbarh et al. 2003; Zyvoloski et al.

2003). Limited to region around waste repository
USGS Model (Belcher et al. 2010)
I

includes entire flow system

USGS Study
To avoid adding lots of unreadable figures, start with list of PDF
page #s to Belcher et al. (2010), accessible at UTD here
I

Ultimate scale: 27 hydrgeologic units, 16 layers of defined

thickness, a finite-difference grid consisting of 194 rows and
160 columns, and uniform cells 1,500 meters (m) on each side.

Decide boundary of system (Setting, p. 6)

potentially complicated hydrologic subdivisions (schematic

block diagram, p. 138)
Hydrostratigraphic units (internal heterogeneity):

I
I

I
I

what kind of rocks possible (Surficial geology, p. 24)

deep carbonates important (large 2ndary porosity) and very
thick (E-W deep fence diagram, p. 26)
surficial volcanics lower perm, have visible chemical impact
(surface map, p. 27)
structure (normal faults) important, (structure map, p. 31)
vertical zonation crucial in this area (deep high conductivity
layers, see x-secns., p. 34, maps p. 42, 48-50)

USGS Study (cont.)

I
I
I

Boundary fluxes (recharge-discharge)

I

I
I
I

again carbonates most important (maps p. 62-3)

some confining beds present (map. p 66)
together with structural juxtaposition stratigraphy quite
complicated (3D block diagram, p. 74)
Walker Lane-related strike-slip faults thought to be deep
conduits for groundwater flow (3D block digram, p. 81)
also faults as flow barriers (p. 83)
potentiometric surface map, with recharge-discharge areas (p.
100)
anthropogenic (pumping) effects (p. 106, see abandoned farms
in Google Maps)
pumping can be included, but with significant errors (p. 110)
precipitation-infiltration is main water source (p. 112-3)
inter-basin flow possible, even across flow system boundaries
(small compared to pump rate, p. 115)
also ungauged natural springs (p. 155), at least providing
water level estimate

Calibration data (history matching)

USGS Study (cont.)

I

Build model grid

I
I
I

I
I
I

use holes-of-opportunity, usually poorly distributed (p. 122)

and records questionable
grid map view, p. 166,
hydrostratigraphic unit list, p. 167
assign hydrolithology (and therefore rock properties) in 3D (p.
174, explan. p 171)
block & fence diagrams of model units, p. 182-184, 243-6,
263)
viewable as thickness maps also (e.g. volcanic-sed p. 203,
lower carbonate aquifer p. 231)
model layers vs. hydrostraigraphy (p. 261)
fence diagram of model units (p. 263)
simulated recharge & discharge (p. 266-7)

Calibration results
I

head errors (p. 331, large in Eleana Range, good in Amargosa

Valley)

References
Belcher, W.R., DAgnese, F.A., OBrien, G.M.: Death valley regional
groundwater flow system, nevada and california; hydrogeologic
framework and transient groundwater flow model; introduction. U.
S. Geological Survey Professional Paper pp. 3 17 (2010),
http://pubs.usgs.gov/pp/1711/
Bethke, C.M.: Geochemical Reaction Modeling: Concepts and
Applications. Oxford University Press, New York (1996)
Brikowski, T.H., Norton, D.: Influence of magma chamber geometry on
hydrothermal activity at mid-ocean ridges. Earth Planet. Sci. Lett.
93, 241255 (1989)
Eddebbarh, A.A., Zyvoloski, G.A., Robinson, B.A., Kwicklis, E.M.,
Reimus, P.W., Arnold, B.W., Corbet, T., Kuzio, S.P., Faunt, C.:
The saturated zone at Yucca Mountain: an overview of the
characterization and assessment of the saturated zone as a barrier to
potential radionuclide migration. J. Contam. Hydrology 62-63,
477493 (April-May 2003), http://www.sciencedirect.com/
science/article/pii/S0169772202001547

References (cont.)
Fehn, U., Cathles, L.M.: Hydrothermal convection at slow-spreading
mid-ocean ridges. Tectonophysics 55(1-2), 239260 (1979)
Haitjema, H.M.: Analytic Element Modeling of Groundwater Flow.
Academic Press, San Diego, CA (1995), iSBN 0-12-316550-4
Harbaugh, A.W.: Modflow-2005, the u.s. geological survey modular
ground-water modelthe ground-water flow process. Techniques and
Methods Book 6-A16, U. S. Geol. Survey, Denver, CO (2005),
http://pubs.usgs.gov/tm/2005/tm6A16/PDF/TM6A16.pdf
Harbaugh, A.W., Banta, E.R., Hill, M.C., McDonald, M.G.:
Modflow-2000, the u.s. geological survey modular ground-water
model user guide to modularization concepts and the ground-water
flow process. Open File Rept. OFR00-92, U. S. Geol. Survey,
Denver, CO (2000), http://water.usgs.gov/nrp/gwsoftware/
modflow2000/ofr00-92.pdf, 121 p
van der Heijde, P., Bachmat, Y., Bredehoeft, J., Andrews, B., Holtz, D.,
Sebastian, S.: Groundwater Management: The use of numerical
models, Water Resources Monograph, vol. 5. Amer. Geophys.
Union, Washington, D.C. (1985)

References (cont.)
Hill, M.C.: Preconditioned conjugate-gradient 2 (pcg2), a computer
program for solving ground-water flow equations. Water-resour.
investig. rept. 90-4048, U.S. Geol. Survey, Denver, CO (1990)
Konikow, L.F.: Predictive accuracy of a groundwater model - lessons
from a post-audit. Ground Water 24, 173184 (1986)
Konikow, L.F., Bredehoeft, J.D.: Computer model of two-dimensional
solute transport and dispersion in ground water, vol. 7, chap. Chp.
C2. U. S. Geol. Survey (1978)
Konikow, L.F., E. P. Patten, J.: Groundwater forecasting. In: Anderson,
M., Burt, T. (eds.) Hydrological Forecasting. John Wiley (1985)
Krabbenhoft, D.P., Anderson, M.P.: Use of a numerical ground-water
flow model for hypothesis testing. Ground Water 24, 4955 (1986)
Mangold, D.C., Tsang, C.F.: A summary of subsurface hydrological and
hydrochemical models. Rev. Geophys. 29, 5179 (1991)
McDonald, M.G., Harbaugh, A.W.: A modular three-dimensional
finite-difference ground-water flow model. Techniques of Water
Resour. Investig. A1, Book 6, 200 (1988)

References (cont.)
Mercer, J.W., Faust, C.R.: Ground-water modeling: An overview.
Ground Water 18, 108115 (1980)
Mercer, J.W., Faust, C.R.: Ground-Water Modeling. Nat. Water Well
Assn. (1981)
Parkhurst, D.L.: Users guide to PHREEQC-A computer program for
speciation, reaction-path, advective-transport, and inverse
geochemical calculations. Water-Resources Investig. Rept. 95-4227
(1995), http:
//wwwbrr.cr.usgs.gov/projects/GWC_coupled/phreeqc/
Prickett, T.A., Lonnquist, C.G.: Selected digital computer techniques for
groundwater resource evaluation. Bulletin, Illinois State Water
Survey, Urbana, IL (1971)
Pruess, K.: A quantitative model of vapor dominated geothermal
reservoirs as heat pipes in fractured porous rocks. Geotherm.
Resour. Council Transact. 9, 353361 (1985)
Pruess, K.: Tough2 - a general-purpose numerical simulator of
multiphase fluid and heat flow. Lbl-29400/uc-251, Lawrence
Berkeley Laboratory, Berkeley, CA (1991)

References (cont.)
Pruess, K.: The TOUGH codes; a family of simulation tools for
multiphase flow and transport process in permeable media. Vadose
Zone J. 3(3), 738746 (Aug 2004)
Pruess, K., Oldenburg, C., Moridis, G.: Tough2 users guide, version 2.0.
Report LBNL-43134, Lawrence Berkeley Nat. Lab, Berkeley, CA
(NOV 1999), http://esd.lbl.gov/tough2/LBNL_43134.pdf
Strack, O.D.L.: Groundwater Mechanics. Prentice Hall, Englewood Cliffs,
NJ (1989)
Travis, B.J., Birdsell, K.H.: Tracr3d: A model of flow and transport in
porous media model description and users manual. La-11798-m,
Los Alamos Nat. Lab., Los Alamos, NM (1991)
Voss, C.I.: A finite-element simulation model for saturated-unsaturated,
fluid-density-dependent groundwater flow with energy transport or
chemically-reactive single-species solute transport. Water Resour.
Investig Rept. 84-4369, U.S. Geol. Surv. (1984)
Wolery, T.J.: Calculation of chemical equilibrium between aqueous
solution and minerals: the eq3/6 software package. Ucrl-52658,
Lawrence Livermore Nat. Lab., Livermore, CA (1979)

References (cont.)
Zheng, C.: MT3D: A modular three-dimensional transport model for
simulation of advection, dispersion and chemical reactions of
contaminants in groundwater systems. S.S. Papdopulos &
Associates, Inc., Bethesda, MD (1992)
Zheng, C.: Applied Contaminant Transport Modeling. van Nostrand
(1995), iSBN 0-442-01348-5
Zhou, Y., Li, W.: A review of regional groundwater flow modeling.
Geoscience Frontiers 2(2), 205 214 (2011),
http://www.sciencedirect.com/science/article/pii/
S167498711100020X
Zyvoloski, G., Kwicklis, E., Eddebbarh, A.A., Arnold, B., Faunt, C.,
Robinson, B.A.: The site-scale saturated zone flow model for Yucca
Mountain: calibration of different conceptual models and their
impact on flow paths. J. Contam. Hydrology 62-63, 731750
(April-May 2003), http://www.sciencedirect.com/science/
article/pii/S0169772202001900