&EPA
United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA/600/R-98/010
January 1998
BIOPLUME
Natural Attenuation Decision
Support System
User's Manual
Version 1.0
The Plume of
Contaminated
Ground Water
The Source of
Contamination
-------
EPA/600/R-98/010
January 1998
BIOPLUME III
Natural Attenuation Decision Support System
User's Manual
Version 1.0
by
Hanadi S. Rafai
University of Houston
Houston, Texas
Charles J. Newell
Groundwater Services, Inc.
Houston, Texas
James R. Gonzales
Technology Transfer Division
Air Force Center for Environmental Excellence
Brooks AFB, San Antonio, Texas
Stergios Dendrou
Basil Dendrou
ZEi / MicroEngineering, Inc.
Annandale, Virginia
Lonnie Kennedy
Deerinwater Environmental Management Services
Norman, Oklahoma
John T. Wilson
Subsurface Protection and Remediation Division
National Risk Management Research Laboratory
Ada, Oklahoma
IAG#RW57936164
Project Officer
John T. Wilson
Subsurface Protection and Remediation Division
National Risk Management Research Laboratory
Ada, Oklahoma 74820
NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
NOTICE
BIOPLUME III and the field data were developed through a collaboration between the U.S.
EPA (Subsurface Protection and Remediation Division, National Risk Management Research
Laboratory, Robert S. Kerr Environmental Research Center, Ada, Oklahoma (RSKERC) and the
U.S. Air Force (U.S. Air Force Center for Environmental Excellence, Brooks Air Force Base, Texas).
EPA staff contributed conceptual guidance in the development of the BIOPLUME III mathematical
model, and contributed field data generated by EPA staff supported by ManTech Environmental
Research Services, Corp., the in-house analytical support contractor at the RSKERC. the computer
code for BIOPLUME III was developed by Groundwater Services, Inc. through a contract with the
U.S. Air Force. The graphical user interface (GUI) was developed by Deerinwater Environmental
Management Services, Inc. through a subcontract to ZEi Engineering Inc. Development of the GUI
was supported through a contract with the U.S. Air Force.
All data generated by EPA staff or by ManTech Environmental Research Services Corp. were
collected following procedures described in the field sampling Quality Assurance Plan for an in-
house research project on natural attenuation, and the analytical Quality Assurance Plan for ManTech
Environmental Research Services Corp.
BIOPLUME III and the User's Manual have been subjected to the Agency's peer and
administrative review and have been approved for publication as an EPA document. However,
BIOPLUME III is made available on an as-is basis without guarantee or warranty of any kind,
express of implied. Neither the United States Government (U.S. EPA or U.S. Air Force), Groundwater
Services Inc., Deerinwater Environmental Management Services Inc., or ZEi Engineering Inc., nor
any of the authors or reviewers accept any liability resulting from the use of BIOPLUME III and
interpretation of the predictions of the model are the sole responsibility of the user. Mention of
trade names or commercial products does not constitute endorsement or recommendation for use.
-------
FOREWORD
The U. S. Environmental Protection Agency is charged by Congress with protecting the Nation's
land, air, and water resources. Under a mandate of national environmental laws, the Agency strives
to formulate and implement actions leading to a compatible balance between human activities and
the ability of natural systems to support and nurture life. To meet these mandates, EPA's research
program is providing data and technical support for solving environmental problems today and
building a science knowledge base necessary to manage our ecological resources wisely, understand
how pollutants affect our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for investigation
of technological and management approaches for reducing risks from threats to human health and
the environment. The focus of the Laboratory's research program is on methods for the prevention
and control of pollution to air, land, water, and subsurface resources; protection of water quality in
public water systems; remediation of contaminated sites and ground water; and prevention and
control of indoor air pollution. The goal of this research effort is to catalyze development and
implementation of innovative, cost-effective environmental technologies; develop scientific and
engineering information needed by EPA to support regulatory and policy decisions; and provide
technical support and information transfer to ensure effective implementation of environmental
regulations and strategies.
An extensive investment in site characterization and mathematical modeling is often necessary
to establish the contribution of natural attenuation at a particular site. This document contains a
mathematical model (BIOPLUME III) intended to describe natural attenuation of organic
contaminants dissolved in ground water. The User's Manual provides instruction on the use of
BIOPLUME III, and contains field data from representative sites to illustrate its appropriate
application. This screening tool will allow ground water remediation managers to identify sites
where natural attenuation is most likely to be protective of human health and the environment. It
will also allow regulators to carry out an independent assessment of treatability studies and remedial
investigations that propose the use of natural attenuation.
Clinton W. Hall, Director
Subsurface Protection and Remediation Division
National Risk Management Research Laboratory
-------
ACKNOWLEDGMENTS
The authors would like to acknowledge the U. S. Air Force Center for Environmental Excellence
(AFCEE) for supporting the development of BIOPLUME III. We would like to specifically
acknowledge Lt. Col. Ross Miller and Marty Faile.
We also wish to acknowledge the following person and organization for providing valuable
input and comments on the development of the model: Dr. Michael Kavanaugh, Environ
The BIOPLUME III software was reviewed by a distingished review team. We wish to
acknowledge members of the team for their comments and suggestions:
Gilberto Alvarez, U.S. EPA Region V, Chicago, IL
Mike Barden, Wisconsin Department of Natural Resources
Curt Black, U.S. EPA Region X, Seattle, WA
Kathy Grindstaff, Indiana Department of Environmental Management (IDEM)
Bradley M. Hill, ManTech Environmental Technology, Inc.
Dr. Rashid Islam, ManTech Environmental Technology, Inc.
Robin Jenkins, Utah DEQ, LUST Program
Tim R. Larson, Florida Department of Environmental Protection
Dr. Ying Ouyang, ManTech Environmental Technology, Inc.
Luanne Vanderpool, U.S. EPA Region V, Chicago, IL
Dr. Jim Weaver, U.S. EPA National Risk Management Research Laboratory
Todd H., Wiedemeier, Parsons Engineering Science, Inc.
Joe R. Williams, U.S. EPA National Risk Management Research Laboratory
Kay Wischkaemper, U.S. EPA Region IV, Atlanta, GA
IV
-------
TABLE OF CONTENTS
Page
DISCLAIMER ii
ACKNOWLEDGMENTS iv
TABLE OF CONTENTS v
APPENDICES viii
LIST OF TABLES x
LIST OF FIGURES xi
1.0 INTRODUCTION 1
1.1 BIOPLUME III - An Extension of BIOPLUME I and II 2
1.2 Graphical User Interface 3
1.3 What is in this Manual? 3
1.4 Fundamentals of Intrinsic Remediation 4
1.4.1 Aerobic and Anaerobic Electron Acceptors 5
1.4.2 Kinetics of the Biodegradation Reactions 10
2.0 GETTING STARTED 19
2.1 Installing the Graphical User Interface 19
2.1.1 Microsoft Windows Fundamentals 19
2.1.2 What You Need to Get Started 24
2.1.3 How to Install the Graphical User Interface Platform 24
2.2 Description of the Platform Controls 25
2.2.1 Operating the Graphical User Interface Platform Commands and
Controls 25
2.2.2 Description of the Graphical User Interface Platform Menus 26
2.2.3 Navigating Through a Simulation 28
2.3 Checking Platform Installation 29
2.3.1 Checking the Platform Executables 30
2.3.2 Checking the Graphics and the Background Image 34
2.3.3 Checking the Animation Executables and Files 35
2.4 Checking the Installed Case Studies 37
2.5 Concluding Remarks 39
v
-------
3.0 Tutorial 40
3.1 Tutorial Overview 42
3.2 Session 1: Basic Model Development 43
3.3 Session 2: Basic Flow Modeling 45
3.3.1 Domain and Boundary Conditions 45
3.3.2 Hydraulic Head Conditions 50
3.3.3 Aquifer Thickness 53
3.3.4 Steady-State Simulation 55
3.4 Session 3: Non-Attenuated Hydrocarbon Mass Transport 58
3.4.1 Observed Contaminant Plume Addition 58
3.4.2 Transport Execution and Results 63
3.4.3 Constant Source Addition 65
3.5 Session 4: Simulated Microbial Attenuation 67
3.5.1 Addition of Electron Acceptors 67
3.5.2 Model Execution and Results 71
3.6 Session 5: Special Features 74
3.6.1 Adding Wells 74
3.6.2 Rivers, Drains and Lakes 77
3.7 Session 6: Video Animation 79
4.0 BIOPLUME III THEORETICAL DEVELOPMENT 82
4.1 Overview 82
4.1.2 Conceptual Model for Biodegradation 82
4.1.3 BIOPLUME III Applicability and Limitations 83
4.1.4 Comparison of BIOPLUME III to Analytical Models 84
4.2 Mathematical Model 85
4.2.1 Numerical Simulation of Oxygen Limited Biodegradation in
BIOPLUME II 85
4.2.1.1 Equation Formulation 85
4.2.1.2 Development of the BIOPLUME II Model 88
4.2.2 BIOPLUME III Equation Formulation 89
4.2.3 Biodegradation Kinetic Models in BIOPLUME III 93
4.2.3.1 First-Order Decay Model 93
VI
-------
4.2.3.2 Instantaneous Reaction Model 94
4.2.3.3 Monod Kinetic Model 94
4.3 Application of BIOPLUME III to Sites 96
4.3.1 Calibration, Verification and Prediction 97
4.3.2 Sensitivity Analysis 98
4.3.3 Impact of Non-BTEX Constituents on BIOPLUME III Modeling 102
4.3.4 Mass Balance Assessments 106
5.0 Platform User's Guide 108
5.1 User's Guide Overview 108
5.2 Modeling Steps Using the Platform 110
5.3 Reference on Menus and Toolboxes 116
5.3.1 Description of Menus and Menu Options 116
5.3.2 Secondary Menus 127
5.3.3 Available Menu Options in the Icon Bar 130
5.3.4 Toolbox Features 131
5.4 Reference on Dialog Boxes and Input Parameters 135
5.4.1 DialogBoxes Associated with Menu File 135
5.4.2 Dialog Boxes Associated with Menu Domain 137
5.4.3 Dialog Boxes Associated with Menu Loading 144
5.4.4 Dialog Boxes Associated with Menu Edit 148
5.4.5 Dialog Boxes Associated with Menu Grid 153
5.4.6 DialogBoxes Associated with Menu Initial Conditions 158
5.4.7 Dialog Boxes Associated with Menu Simulator 159
5.4.8 Dialog Boxes Associated with Menu Results 162
5.4.9 Dialog Boxes Associated with Menu View 168
5.4.10 Dialog Boxes Associated with Menu Annotation 169
5.5 Advanced Topics 170
5.5.1 Platform Software Architecture 170
5.5.2 Platform Input of Natural Attenuation Parameters 171
5.5.3 Sensitivity of Input Parameters 174
5.5.4 Concluding Remarks 175
6.0 References 176
vn
-------
APPENDICES
I. Input Data 179
I.I Discretization of Space 179
1.2 Discretization of Time 179
1.3 Hydrogeologic Characteristics of the Aquifer 191
1.4 Boundary Conditions 193
1.5 Initial Conditions 205
1.6 Sources and Sinks 205
1.7 Sorption, Source Decay, Radioactive Decay and Ion Exchange Variables 206
1.8 Biodegradation Variables 210
1.9 Numerical Parameters 211
1.10 Output Control Parameters 211
1.11 References 212
II. Interpretation of Output 214
III Standard Output File (SOF) 214
II.2 Graphical Output File (GOF) 214
II. 3 Resulting Heads 214
II.4 Resulting Concentrations 215
II. 5 Mass Balance Results 216
III. Questions Most Commonly Asked 218
III.l Can I use the model for anunconfined aquifer? 218
III.2 I need to model a larger grid 218
III.3 Should I assume steady-state or transient hydraulics? 218
III.4 I have large mass balance errors 218
III.5 My model runs forever 219
III.6 My model is generating particles. Is there something wrong? 219
III.7 My plume is running off the page. Is this OK? 219
III.8 I'm setting up all my cells as constant-head nodes to fix the ground water
elevations at the cells. Will it work? 219
III.9 What happens to particles that migrate off the grid? 219
APPENDIX A. Background Information on the USGS MOC Model 220
A.I Introduction 220
A.2 Theoretical Background 220
A.3 Stability Criteria 224
A.4 Boundary and Initial Conditions 225
A.5 Mass Balance 225
A.6 Evaluation of MOC - Comparison with Analytical Solutions 227
A.7 Mass Balance Tests 229
Vlll
-------
A.8 References 239
APPENDIX B. Implementing the Air Force Intrinsic Remediation Protocol Using the
Graphical Platform 240
B. 1 Context of the Remedial Investigation Using the Platform 240
B.I.I Review Existing Site Data 241
B.I.2 Develop Preliminary Conceptual Model 242
B.2 Site Characterization in Support of Intrinsic Remediation 245
B.2.1 Soil Characterization 246
B.2.2 Ground Water Characterization 249
B.2.3 Aquifer Parameter Estimation 256
B.2.4 Optional Confirmation of Biological Activities 258
B.3 Refining the Conceptual Model 259
B.3.1 Hydrogeologic Sections 260
B.3.2 Potentiometric Surface or Water Table Maps 260
B.3.3 Contaminant Concentration Contour Maps 261
B.3.4 Electron Acceptors and Metabolic Byproduct Contour Maps 261
B.4 Calculations and Sorting of Raw Data 266
B.4.1 Analysis of Contaminant, Electron Acceptor and Byproduct Data 266
B.4.2 Sorption and Retardation Calculations 270
B.4.3 Fuel/Water Partition Calculations 270
B.4.4 Ground Water Flow Velocity Calculations 270
B.4.5 Anaerobic Biodegradation Rate Constant Calculations 270
B.5 Simulate Intrinsic Remediation Using the Platform 271
B.5.1 Requirements for a Contaminant Biodegradation Simulation 272
B.5.2 Context of the Conceptual Model 272
B.5.3 Steps Specific to Biodegradation Modeling 274
B.5.4 Calibration of the Bioremediation Model 275
B.6 Conduct an Exposure Assessment 277
B.7 Prepare Long-Term Monitoring Plan 277
B.8 Additional Reading 279
IX
-------
LIST OF TABLES
Table 1.1 Redox Reactions for Benzene 6
Table 1.2 Redox Reactions for Toluene 7
Table 1.3 Redox Reactions for Ethylbenzene and Xylene 8
Table 1.4 Electron Acceptor/By-Product Data From Air Force Sites 10
Table 1.5 Biodegradation Capacity/Expressed Assimilative Capacity at AFCEE Intrinsic
Remediation Sites 11
Table 1.6a Utilization Factor Calculation for Benzene 15
Table 1.6b Utilization Factor Calculation for Toluene 16
Table 1.6c Utilization Factor Calculation for Ethylbenzene and Xylene 17
Table 1.7 Utilization Factors for BTEX 18
Table 2.1 Sub-directories of the BIOPLUME III Graphical User Interface Platform 30
Table 2.2 List of Installed Real Case Studies 37
Table 2.3 List of Installed Test Cases 38
Table 4.1 Sensitivity of Model Results to Changes in Hydrogeologic Parameters 100
Table 4.2 Sensitivity of Model Results to Linear Sorption and Radioactive Decay 101
Table 4.3 Sensitivity of Model Results to First Order Decay and Instantaneous Reaction
Biodegradation Kinetics 105
Table 5.1 Description of the Platform Menus 116
Table 5.2 Input Data Given Per Strata 172
Table I.I Input Data for BIOPLUME III 180
Table 1.2 Effective Porosity Estimates 192
Table 1.3 Dispersivity Estimates from Field Experiments 195
Table 1.4 Typical Bulk Densities and foc Values 207
Table 1.5 Typical Distribution Coefficients 209
Table A.I Model Parameters for the Tracer Slug Mass Balance Test Problem 232
Table A.2 Model Parameters for the Effects of Wells Mass Balance Test Problem 235
x
-------
LIST OF FIGURES
Figure 1.1 Distribution of BTEX, Electron Acceptors, and Metabolic By-Products vs.
Distance Along Centerline of Plume 13
Figure 2.1 Components of an MS-WINDOWS Application 21
Figure 2.2 Typical Controls in a Dialog Box 23
Figure 2.3 Graphical User Interface Platform Menu and Toolbox 26
Figure 2.4 Screen View of Case Study "TESTP31" 31
Figure 2.5 Defining the Simulation Period 32
Figure 2.6 Computed Hydraulic Heads at Time 2.5 Years 33
Figure 2.7 Computed Concentrations of Hydrocarbons at Time 2.5 Years 33
Figure 2.8 Screen View of Case Study "HILLAFB1" 34
Figure 2.9 Plume Migration for BTEX and Oxygen after a 1 Year Simulation 35
Figure 2.10 Playback Screen of AVI Files 36
Figure 3.1 Boundary Conditions in Test Simulation 50
Figure 3.2 Entering Piezometric Head Data Using the Line Tool 51
Figure 3.3 Kriged Piezometric Head Contour Map 52
Figure 3.4 Example of Log Point Distribution to Define Top and Bottom of Aquifer 54
Figure 3.5 Spreadsheet Representation of Simulated Hydraulic Heads 57
Figure 3.6 Contaminant Distribution Image 59
Figure 3.7 Example of Establishing Hydrocarbon Distribution Over the Grid 61
Figure 3.8 Example Kriged Hydrocarbon Distribution 62
Figure 3.9 Simulated Hydrocarbon Plume in 10 Years 64
Figure 3.10 Simulated Hydrocarbon Plume as a Constant Source 66
Figure 3.11 Addition of Electron Acceptor Source Areas 70
Figure 3.12 Simulated Hydrocarbon Plume at 10 Years Assuming Microbial Attenuation 73
Figure 3.13 Oxygen Distribution Showing Reaction Sag 73
Figure 3.14 Sulfate Distribution Showing Reaction Sag 73
Figure 3.15 Head Distribution Under Pumping Conditions 77
Figure 4.1 Principle of Superposition for Combining the Hydrocarbon and Oxygen Plumes in
BIOPLUMEII 90
Figure 5.1 Conceptual Model 110
Figure 5.2 Required Steps for a Groundwater Contaminant Migration Simulation 112
Figure 5.3 Layered Structure of the Graphical Platform Software Architecture 171
Figure 5.4 Estimates of Sensitivity Analysis 175
Figure I.I Grid Discretization in BIOPLUME III 189
Figure 1.2 Discretization of Time in BIOPLUME III 190
Figure 1.3 Longitudinal Dispersivity Chart 194
Figure 1.4 Illustration of Plume Length for Estimating Longitudinal Dispersivity 196
Figure 1.5 Hydraulic Conductivity for Different Type of Soils 197
Figure 1.6 Hydrogeologic Conditions for Site A 200
XI
-------
Figure 1.7 Hydrogeologic Conditions for Site B 201
Figure A.I Relation of Flow Field to Movement of Particles 223
Figure A.2 Comparison Between Analytical Model and MOC for Dispersion in One-
Dimensional Steady-State Flow 228
Figure A.3 Comparison Between Analytical Model and MOC for Dispersion in Plane Radial
Steady-State Flow 230
Figure A.4 Grid, Boundary Conditions and Flow Field for the Tracer Slug Mass Balance Test
Problem 231
Figure A.5 Mass Balance Errors for the Tracer Slug Mass Balance Problem 233
Figure A.6 Grid, Boundary Conditions and Flow Field for Effects of Wells Mass Balance
Test Problem 234
Figure A.7 Mass Balance Errors for the Effects of Wells Mass Balance Problem 236
Figure A.8 Effect of Number of Particles on Mass Balance Error 237
Figure A.9 Effect of Maximum Cell Distance (CELDIS) on Mass Balance Errors 23 8
Figure B. 1 Logical Connection Between Technical Protocol and the Platform 241
Figure B.2 General Configuration of the Platform 243
FigureB.3 Soil Sampling Using CPT Technology 247
Figure B.4 Typical CPTU Boring Log 248
Figure B.5 Platform Controls to Input Water Quality Data 250
Figure B.6 Entering Well Data in the Platform 251
Figure B.7 Platform Input of In Situ Measured Oxygen 251
FigureB.8 Contour of Hydraulic Heads 261
FigureB.9 Typical BTEX Contour Map 263
Figure B. 10 Measured Oxygen Plume 263
Figure B. 11 Measured Nitrate Plume 264
Figure B. 12 Measured Sulfate Plume 264
FigureB.13 Measured Methane Plume 265
Figure B. 14 Measured Ferrous Iron Plume 265
Figure B.I5 Oxygen Depletion 267
Figure B. 16 Nitrate Depletion 268
FigureB.17 Sulfate Depletion 268
Figure B. 18 Methane Creation 269
FigureB.19 Ferrous Iron Creation 269
Figure B.20 A Conceptual Intrinsic Remediation Model 274
FigureB.21 Typical Long-Term Monitoring Strategy 279
xn
-------
1.0 INTRODUCTION
The BIOPLUME III program is a two-dimensional, finite difference model for simulating the
natural attenuation of organic contaminants in ground water due to the processes of advection,
dispersion, sorption, and biodegradation. The model simulates the biodegradation of organic
contaminants using a number of aerobic and anaerobic electron acceptors: oxygen, nitrate, iron
(III), sulfate, and carbon dioxide.
Over the past several years, the high cost and poor performance of many pump and treat
remediation systems have led many researchers to consider natural attenuation as an alternative
technology for ground water remediation (Newell et al., 1996). Researchers associated with the
U.S. EPA's National Risk Management Research Laboratory in Ada, Oklahoma, have suggested
that anaerobic pathways could be significant, or even the dominant degradation mechanism at
many petroleum fuel sites (Wilson, 1994). As a result, The Air Force Center for Environmental
Excellence (AFCEE), Technology Transfer Division, launched a three-point technology
development effort in 1993, consisting of the following elements:
1) Field data collected at over 30 sites around the country (Wiedemeier et al., 1995a)
analyzing aerobic and anaerobic processes;
2) A technical Protocol, outlining the approach, data collection techniques, and data analysis
methods required for conducting an Air Force Intrinsic Remediation Study (Wiedemeier et
al., 1995b); and
3) Two intrinsic remediation modeling tools: the BIOSCREEN model developed by Dr.
Charles J. Newell of Groundwater Services, Inc. (GSI), and the BIOPLUME III model
developed by Dr. Hanadi Rifai at Rice University.
In addition, the Air Force also oversaw development of a modified version of a sophisticated
ground water modeling platform known as Environmental Information System (EIS) developed
by Dr. Stergios Dendrou and Dr. Basil Dendrou of ZEi/MicroEngineering, Inc., of Annandale,
Virginia.
This Windows"-based graphical platform model has been integrated with BIOPLUME III. The
integration effort of the platform and the BIOPLUME III model was managed by the prime
contractors, GSI and Deerinwater Environmental Management (DEM), and their subcontractors,
Rice University and ZEi/MicroEngineering, respectively. The "team" was formed after review of
the EIS system by AFCEE and EPA researchers who determined the inherent benefits of each
modeling system would result in a more advanced and user-friendly natural attenuation model.
Such a model was identified as a key requirement for broadening the use and acceptance of natural
attenuation during the 1994 EPA/Air Force Natural Attenuation Symposium in Denver,
Colorado.
Collectively, these software tools, the technical protocol, and the knowledge gained from
numerous natural attenuation studies by the Air Force Center for Environmental Excellence and
-------
the EPA's Risk Reduction Laboratory will provide users with the necessary assets to perform a
complete natural attenuation study.
1.1 BIOPLUME III - An Extension of BIOPLUME I and II
BIOPLUME III is a two-dimensional, finite difference model for simulating the biodegradation of
hydrocarbons in ground water. The model simulates both aerobic and anaerobic biodegradation
processes in addition to advection, dispersion, sorption and ion exchange. BIOPLUME III is
based on the U. S. Geologic Survey (USGS) Method of Characteristics Model dated July 1989
(Konikow and Bredehoeft, 1989; see Appendix A).
The BIOPLUME III code was developed primarily to model the natural attenuation of organic
contaminants in ground water due to the processes of advection, dispersion, sorption and
biodegradation. BIOPLUME III simulates the biodegradation of organic contaminants using a
number of aerobic and anaerobic electron acceptors: oxygen, nitrate, iron (III), sulfate, and carbon
dioxide. The model solves the transport equation six times to determine the fate and transport of
the hydrocarbons and the electron acceptors/reaction by-products. For the case where iron (III)
is used as an electron acceptor, the model simulates the production and transport of iron (II) or
ferrous iron.
Three different kinetic expressions can be used to simulate the aerobic and anaerobic
biodegradation reactions. These include: first-order decay, instantaneous reaction and Monod
kinetics. The principle of superposition is used to combine the hydrocarbon plume with the
electron acceptor plume(s).
Borden and Bedient (1986) developed the BIOPLUME I model based on their work at the
United Creosoting Company, Inc. Superfund site in Conroe, Texas. BIOPLUME I is based on
the assumption that aerobic biodegradation of hydrocarbons is often limited by the availability of
dissolved oxygen in ground water aquifers. Borden and Bedient (1986) simulated the aerobic
biodegradation of hydrocarbons as an instantaneous reaction between the hydrocarbon and
oxygen.
Rifai et al. (1988) developed the BIOPLUME II model by incorporating the concepts developed
by Borden and Bedient (1986) into the USGS two-dimensional solute transport model (Konikow
and Bredehoeft, 1978). The BIOPLUME II model tracks two plumes: oxygen and the
hydrocarbon. The two plumes are superimposed to determine the resulting concentrations of
oxygen and hydrocarbon at each time step. Anaerobic biodegradation in BIOPLUME II was
simulated as a first-order decay in hydrocarbon concentrations.
Other major differences between BIOPLUME II and BIOPLUME III include:
• BIOPLUME III runs in a Windows95 environment whereas BIOPLUME II was mainly
developed in a DOS environment.
• BIOPLUME III has been integrated with a modified version of a sophisticated ground
water modeling platform known as EIS developed by ZEi/MicroEngineering, Inc.
-------
1.2 Graphical User Interface
Intrinsic Remediation studies are data intensive and require the applicant to make the case that
natural attenuation is occurring at a site and that it will persist over time. To help the
environmental professional with the data management, visualization, and decision-making tasks
involved, the Air Force adopted the EIS Graphical User Interface Platform. EIS (Environmental
Information System) is the latest integrated software platform under Windows 95 in which to
register, sort, and evaluate the site-specific data of the physical processes influencing the ground
water migration of organic contaminants.
EIS is developed around the following integrating technologies:
Object-based simulation environment
Control tools for the creation of a spatial and temporal data base (4 dimensions)
A patented Macroengineering framework for managing different algorithmic solutions
Graphics that are embedded in a kriging scheme automatically adjusting to the required
spatial resolution
Open software architecture allowing a cost-efficient customization (other algorithmic
solutions, link to other GIS systems) and expansion of the platform (support of different
peripheral and field monitoring devices)
Integration and quantification of the simulation and data processing error to the risk of
health hazard
For these reasons, the EIS platform is at the forefront of the arsenal of tools that AFCEE is
making available to the engineering community in support of natural attenuation (intrinsic
remediation) studies.
1.3 What is in this Manual?
This user's guide is a stand-alone document for the BIOPLUME III model and the Graphical
User Interface Platform. Following this brief introduction, Section 2.0 provides instructions for
installing the software and getting started. Section 3.0 is a step-by-step tutorial that
demonstrates the main features of the platform. Section 4.0 is devoted to a detailed discussion on
the theoretical development of the BIOPLUME III model and Section 5.0 is a thorough user's
guide for the Graphical User Interface Platform.
-------
1.4 Fundamentals of Intrinsic Remediation1
Naturally occurring biological processes can significantly enhance the rate of organic mass
removal from contaminated aquifers. Biodegradation research performed by Rice University,
government agencies, and other research groups has identified several main themes that are crucial
for future studies of natural attenuation:
1. The relative importance of groundwater transport vs. microbial kinetics is
a key consideration for developing workable biodegradation expressions in
models. Results from the United Creosote site (Texas) and the Traverse
City Fuel Spill site (Michigan) indicate that biodegradation is better
represented as a macro-scale wastewater treatment-type process than as a
micro-scale study of microbial reactions.
2. The distribution and availability of electron acceptors control the rate ofin-
situ biodegradation for most petroleum release site plumes. Other factors
(e.g., population of microbes, pH, temperature, etc.) rarely limit the amount
of biodegradation occurring at these sites.
As mentioned previously, Borden and Bedient (1986) developed the BIOPLUME model, which
simulates aerobic biodegradation as an "instantaneous" microbial reaction that is limited by the
amount of electron acceptor, oxygen, that is available. In other words, the microbial reaction is
assumed to occur at a much faster rate than the time required for the aquifer to replenish the
amount of oxygen in the plume. Although the time required for the biomass to aerobically
degrade the dissolved hydrocarbons is on the order of days, the overall time to flush a plume with
fresh groundwater is on the order of years or tens of years.
Rifai et al. (1988) extended this approach and developed the BIOPLUME II model, which
simulates the transport of two plumes: an oxygen plume and a contaminant plume. The two
plumes are allowed to react, and the ratio of oxygen to contaminant consumed by the reaction is
determined from an appropriate stoichiometric model. The BIOPLUME II model is documented
with a detailed user's manual (Rifai et al., 1987) and is currently being used by EPA regional
offices, U.S. Air Force facilities, and by consulting firms. Borden et al. (1986) applied the
BIOPLUME concepts to the Conroe Superfund site; Rifai et al. (1988) applied the BIOPLUME
II model to a jet fuel spill at a Coast Guard facility in Michigan. Many other studies using the
BIOPLUME II model have been presented in recent literature.
The BIOPLUME II model has increased the understanding of biodegradation and natural
attenuation by simulating the effects of adsorption, dispersion, and aerobic biodegradation
processes in one model. It incorporates a simplified mechanism (first-order decay) for handling
other degradation processes, but does not address specific anaerobic decay reactions. Early
conceptual models of natural attenuation were based on the assumption that the anaerobic
1 Some of the information presented in this section is taken with permission from the BIOSCREEN
Manual developed by Groundwater Services, Inc. for the Air Force Center for Environmental Excellence
(Newell et al., 1996).
-------
degradation pathways were too slow to have any meaningful effect on the overall natural
attenuation rate at most sites. Accordingly, most field programs focused only on the distribution
of oxygen and contaminants, and did not measure the indicators of anaerobic activity such as
depletion of anaerobic electron acceptors or accumulation of anaerobic metabolic by-products.
1.4.1 Aerobic and Anaerobic Electron Acceptors
Naturally occurring biological processes can significantly enhance the rate of organic mass
removal from contaminated ground water aquifers. Biologically mediated degradation reactions
are oxidation/reduction (redox) reactions, involving the transfer of electrons from the organic
contaminant compound to an electron acceptor. Oxygen is the electron acceptor for aerobic
metabolism whereas nitrate, ferric iron, sulfate and carbon dioxide serve as electron acceptors for
alternative anaerobic pathways. Tables 1.1 through 1.3 list the redox reactions for benzene,
toluene, ethyl benzene, and xylene (BTEX).
In the presence of organic substrate and dissolved oxygen, microorganisms capable of aerobic
metabolism will predominate over anaerobic forms. However, dissolved oxygen is rapidly
consumed in the interior of contaminant plumes, converting these areas into anoxic (low oxygen)
zones. Under these conditions, anaerobic bacteria begin to utilize other electron acceptors to
metabolize dissolved hydrocarbons. The principle factors influencing the utilization of the
various electron acceptors include: 1) the relative biochemical energy provided by the reaction; 2)
the availability of individual or specific electron acceptors at a particular site; and 3) the kinetics
(rate) of the microbial reaction associated with the different electron acceptors.
The transfer of electrons during the redox reaction releases energy which is utilized for cell
maintenance and growth. The biochemical energy associated with alternative degradation
pathways can be represented by the redox potential of the alternative electron acceptors: the
more positive the redox potential, the more energetically favorable is the reaction utilizing that
electron acceptor. With everything else being equal, organisms with more efficient modes of
metabolism grow faster and therefore dominate over less efficient forms.
Electron
Acceptor
Oxygen
Nitrate
Ferric Iron
(solid)
Sulfate
Carbon Dioxide
Type of
Reaction
Aerobic
Anaerobic
Anaerobic
Anaerobic
Anaerobic
Metabolic
By-Pro duct
CO2
N2, C02
Ferrous Iron
(dissolved)3
H2S
Methane
Redox Potential
(pH = 7, in volts)
+ 820
+ 740
-50
-220
-240
Reaction
Preference
Most Preferred
fl
fl
fl
Least Preferred
Based solely on thermodynamic considerations, the most energetically preferred reaction should
proceed in the plume until all of the required electron acceptor is depleted. At that point, the
next most-preferred reaction should begin and continue until that electron acceptor is gone,
-------
Table 1.1. Redox Reactions for Benzene
Oxidation C6H6 + 12H2O
6CO2 + 30H+ + 30e'
Reduction 7.5O2 + 30H+ + 30e~ -
6NO3' + 36H+ + 30e'
15Mn4+ + 30e-
30Fe3+ + 30e'
3.75SO42 + 37.5H+ + 30e-
3.75CO2 + 30H+ + 30e-
> 15H2O
> 3N2+18H2O
> 15Mn2+
> 30Fe2+
> 3.75H2S + 15H2O
> 3.75CH4 + 7.5H2O
Oxygen
Nitrate
Manganese
Iron
Sulfate
Methanog.
Overall
C6H6 + 7.502
6CO2 + 3H2O
Oxygen
C6H6 + 6H+ +6NO3
6CO2 +3N2 +6H2O
Nitrate
C6H6 + 15Mn4+ + 12H2O -> 6CO2 + 30H+ + 15Mn2+ Mang
anese
CfiH,
66
12H2O
6CO2 + 30H+ + 30Fe
2+
Iron
6CO2 + 3.75H2S + 3H2O Sulfate
C6H6 + 4.5H20
2.25CO2 + 3.75CH4
Methanog.
-------
Table 1.2. Redox Reactions for Toluene
Oxidation C7H8 + 14H2O
6CO2 + 36H+ + 36e'
Reduction 9O2 + 36H+ + 36e' -» 18H2O
7.2NO3' + 43.2H+ + 36e' -» 3.6N2 + 21.6H2O
18Mn4+ + 36e' -» 18Mn2+
36Fe3+ + 36e' -» 36Fe2+
4.5SO42 + 45H+ + 36e- -» 4.5H2S + 18H2O
4.5CO2 + 36H+ + 36e' -^ 4.5CH4 + 9H2O
Oxygen
Nitrate
Manganese
Iron
Sulfate
Methanog.
Overall
C7H8 + 902
7CO2 + 4H2O
Oxygen
C7H8 + 7.2H++ 7.2NO3- -> 7CO2+3.6N2+7.6H2O Nitrate
C7H8+ 18Mn4++ 14H2O
7CO2 + 36H+ + 18Mn/+ Manganese
C7H8
14H2O
7CO2 + 36H+ + 36Fe
2+
Iron
2- _|_ OTU~'~
C7H8 + 4.5SO/- + 9H
7CO2 + 4.5H2S + 4H2O Sulfate
C7H8 + 5H2O
2.5CO2 + 4.5CH4
Methanog.
-------
Table 1.3. Redox Reactions for Ethylbenzene and Xylene
Oxidation C8H10 + 16H2O
8CO2 + 42H+ + 42e'
Reduction 10.5O2 + 42H+ + 42e'
8.4NO3' + 50.4H+ + 42e'
21Mn4+ + 42e'
42Fe3+ + 42e'
5.25SO42 + 52.5H+ + 42e-
5.25CO2 + 42H+ + 42e-
-» 21H2O
-^ 4.2N2 + 25.2H2O
-^ 21Mn2+
-^ 42Fe2+
-» 5.25H2S + 21H2O
-^ 5.25CH4+ 10.5H2O
Oxygen
Nitrate
Manganese
Iron
Sulfate
Methanog.
Overall
C8H10 + 10.502
8CO2 + 5H2O
Oxygen
C8H10 + 8.4H+ + 8.4NO3- -> 8CO2 + 4.2N2 +9.2H2O Nitrate
C8H10 + 21Mn4+ + 16H2O -> 8CO2 + 42H+ + 21Mn2+ Manganese
C8H
10
16H2O
8CO2 + 42H+ + 42Fe
2+
Iron
CgHio + 5.25SO/'+ 10.5H
8CO2 + 5.25H2S + 5H2O Sulfate
C8H10 + 5.5H20
2. 75CO2 + 5.25CH4
Methanog.
-------
leading to a pattern where preferred electron acceptors are consumed one at a time, in sequence.
Based on these principles, one would expect to observe monitoring well data with "no-detect"
results for the more energetic electron acceptors, such as oxygen and nitrate, in locations where
evidence of less energetic reactions is observed (e.g., monitoring well data indicating the presence
of ferrous iron).
In practice, however, it is unusual to collect samples from natural attenuation monitoring wells
that are completely depleted in one or more electron acceptors. Two processes are probably
responsible for this observation:
1. Alternative biochemical mechanisms having very similar energy potentials (such as aerobic
oxidation and nitrate reduction) may occur concurrently when the preferred electron
acceptor is reduced in concentration, rather than fully depleted. Facultative aerobes, for
example, can shift from aerobic metabolism to nitrate reduction when oxygen is still present
but in low concentrations (i.e. 1 mg/L oxygen; Snoeyink and Jenkins, 1980). Similarly,
noting the nearly equivalent redox potentials for sulfate and carbon dioxide (-220 volts and -
240 volts, respectively) one might expect that sulfate reduction and methanogenic reactions
may also occur together.
2. Standard monitoring wells, having 5 to 10 foot screened intervals, will mix waters from
different vertical zones. If different biodegradation reactions are occurring at different
depths, then one would expect to find geochemical evidence of alternative degradation
mechanisms occurring in the same well. If the dissolved hydrocarbon plume is thinner than
the screened interval of a monitoring well, then the geochemical evidence of electron
acceptor depletion or metabolite accumulation will be diluted by mixing with clean water
from zones where no degradation is occurring.
Therefore, most natural attenuation field sampling programs yield data that indicate a general
pattern of electron acceptor depletion, but not complete depletion, and an overlapping of electron
acceptor/metabolite isopleths into zones not predicted by thermodynamic principles. For
example, a zone of methane accumulation may be larger than the apparent anoxic zone.
Nevertheless, these general patterns of geochemical changes within the plume area provide strong
evidence that multiple mechanisms of biodegradation are occurring at many sites.
The data collected by Weidemeier et al. (1995a) and Newell et al. (1996) provides interesting
observations on intrinsic bioremediation. For example, while the energy of each reaction is based
on thermodynamics, the distribution of electron acceptors is dependent on site specific
hydrogeochemical processes and can vary significantly between sites as seen in Table 1.4.
At Hill AFB, the sulfate reactions are extremely important because of the large amount of
available sulfate for reduction. Note that different sites in close proximity can have quite
different electron acceptor concentrations, as shown by the two sites at Elmendorf AFB. For
data on more sites, see Table 1.5. Calculated biodegradation capacities at different U.S. Air Force
Natural Attenuation research sites have ranged from 7 to 70 mg/L (Table 1.5). The median value
for 28 AFCEE sites is 28.5 mg/L.
-------
Table 1.4. Electron Acceptor/By-Product Data From Air Force Sites
Measured Background Electron Acceptor/By-Product Concentration (mg/L)
Base Facility
POL Site,
Hill AFB, Utah*
Hangar 10 Site,
Elmendorf AFB, Alaska*
Site ST-41,
Elmendorf AFB,Alaska*
Bldg. 735,
Grissom AFB, Indiana
SWMU 66 Site,
Keesler AFB,
Mississippi
POL B Site,
Tyndall AFB, Florida
Background
Oxygen
6.0
0.8
12.7
9.1
1.7
1.4
Background
Nitrate
36.2
64.7
60.3
1.0
0.7
0.1
Maximum
Ferrous Iron
55.6
8.9
40.5
2.2
36.2
1.3
Background
Sulfate
96.6
25.1
57.0
59.8
22.4
5.9
Maximum
Methane
2.0
9.0
1.5
1.0
7.4
4.6
*Data from Wiedemeier et al. (1995a); all other data from Newell et al. (1996)
1.4.2 Kinetics of the Biodegradation Reactions
Aerobic biodegradation can be simulated as an "instantaneous" reaction that is limited by the
amount of electron acceptor (oxygen) that is available. The microbial reaction is assumed to
occur at a much faster rate than the time required for the aquifer to replenish the amount of
oxygen in the plume. Although the time required for the biomass to aerobically degrade the
dissolved hydrocarbons is on the order of days, the overall rate that groundwater is replenished in
most plumes is on the order of years or tens of years.
For example, microcosm data presented by Davis et al. (1994) show that microbes that have an
excess of electron acceptors can degrade concentrations of dissolved benzene (~1 mg/L) very
rapidly. In the presence of a surplus of oxygen, aerobic bacteria can degrade dissolved benzene in
about 8 days, which can be considered "instantaneous" compared to years required for flowing
ground water to replenish the plume area with oxygen.
Recent results from the Air Force Natural Attenuation Initiative indicate that the anaerobic
reactions, which were originally thought to be too slow to be of significance in ground water, can
also be simulated as instantaneous reactions (Newell et al., 1995). For example, Davis et al.
(1994) also ran microcosms with sulfate reducers and methanogens that indicated that benzene
could be degraded within a couple of weeks time frame (after acclimation). When compared to
the time required to replenish electron acceptors, the anaerobic reactions can also be considered to
be instantaneous at many sites.
This conclusion is supported by observing the pattern of anaerobic electron acceptors and by-
products along the plume at natural attenuation research sites:
10
-------
Table 1.5. Biodegradation Capacity/Expressed Assimilative Capacity at AFCEE Intrinsic Remediation Sites
Site
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Maximum
Total BTEX
Concentration
Base
Hill AFB
Battle Creek ANOB
Madison ANOB
ElmendorfAFB
ElmendorfAFB
King Salmon AFB
King Salmon AFB
Plattsburgh AFB
Eglin AFB
Patrick AFB
MacDill AFB
MacDill AFB
MacDill AFB
Offutt AFB
Offutt AFB
WestoverAFRES
WestoverAFRES
Myrtle Beach
Langley AFB
Oriffis AFB
Rickenbacker ANOB
Wurtsmith AFB
Travis AFB
Pope AFB
Seymour Johnson AFB
Grissom AFB
Tyndall AFB
Keesler AFB
State
Utah
Michigan
Wisconsin
Alaska
Alaska
Alaska
Alaska
New York
Florida
Florida
Florida
Florida
Florida
Nebraska
Nebraska
Massachusetts
Massachusetts
South Carolina
Virginia
New York
Ohio
Michigan
Califonia
North Carolina
North Carolina
Indiana
Florida
Mississippi
Site Name
Hangar 10
ST-41
FT-001
Naknek
Site 56
Site 57
Site OT-24
FPT-A3
FT-03
FT-08
SS-42
Bldg. 735
POLE
SWMU 66
Average
Median
Maximum
Minimum
(mg/L)
21.5
3.6
28.0
22.2
30.6
10.1
5.3
6.0
3.7
7.3
29.6
0.7
2.8
3.2
103.0
1.7
32.6
18.3
0.1
12.8
1.0
3.1
-
8.2
13.8
0.3
1.0
14.1
14.2
7.3
103.0
0.1
Biodegradation Capacity/
Observed Change in Concentration (mg/L)
O2
6.0
5.7
7.2
0.8
12.7
9.0
11.7
10.0
1.2
3.8
2.4
2.1
1.3
0.6
8.4
10.0
9.9
0.4
6.4
4.4
1.5
8.5
3.8
7.5
18.3
9.1
1.4
1.7
5.9
5.8
18.3
0.4
Nitrate
36.2
5.6
45.3
64.7
60.3
12.5
0
3.7
0
0
5.6
0.5
0
0
69.7
8.6
17.2
0
23.5
52.5
35.9
25.4
15.8
6.9
4.3
1.0
0.1
0.7
17.7
6.3
69.7
0
Iron
55.6
12.0
15.3
8.9
40.5
2.5
44.0
10.7
8.9
2.0
5.0
20.9
13.1
19.0
0
599.5
279.0
34.9
10.9
24.7
17.9
19.9
8.5
56.2
31.6
2.2
1.3
36.2
49.3
16.6
599.5
0
Sulfate
96.6
12.9
24.2
25.1
57.0
6.8
0
18.9
4.9
0
101.2
62.4
3.7
32.0
82.9
33.5
11.7
20.7
81.3
82.2
93.2
10.6
109.2
9.7
38.6
59.8
5.9
22.4
39.5
24.6
109.2
0
Methane
2.0
8.4
11.7
9.0
1.5
0.2
5.6
0.3
11.8
13.6
13.6
15.4
9.8
22.4
0
0.2
4.3
17.2
8.0
7.1
7.7
1.4
0.2
48.4
2.7
1.0
4.6
7.4
8.4
7.2
48.4
0
Aerobic
Respiration
1.9
1.8
2.3
0.3
4.0
2.9
3.7
3.2
0.4
1.2
0.8
0.7
0.4
0.2
2.7
3.2
3.1
0.1
2.0
1.4
0.5
2.7
1.2
2.4
5.8
2.9
0.5
0.5
1.9
1.9
5.8
0.1
Denitrification
7.4
1.1
9.2
13.2
12.3
2.6
0
0.7
0
0
1.1
0.1
0
0
14.2
1.8
3.5
0
4.8
10.7
7.3
5.2
3.2
1.4
0.9
0.2
0
0.1
3.6
1.3
14.2
0
'Expressed Assimilative Capacity (mg/L)
Iron
Reduction
2.6
0.6
0.7
0.4
1.9
0.1
2.0
0.5
0.4
0.1
0.2
1.0
0.6
0.9
0
27.5
12.8
1.6
0.5
1.1
0.8
0.9
0.4
2.6
1.5
0.1
0.1
1.7
2.3
0.8
27.5
0
Sulfate
Reduction
21.0
2.8
5.3
5.5
12.4
1.5
0
4.1
1.1
0
22.0
13.6
0.8
7.0
18.0
7.3
2.6
4.5
17.7
17.9
20.3
2.3
23.7
2.1
8.4
13.0
1.3
4.9
8.6
5.4
23.7
0
Methanogenesis
2.6
10.8
15.0
11.6
1.9
0.2
7.2
0.4
15.2
17.4
17.4
19.7
12.6
28.8
0
0.2
5.5
22.0
10.2
9.1
9.8
1.8
0.3
62.0
3.5
1.2
5.9
9.5
10.8
9.3
62.0
0
Total
Biodegradation
Capacity (mg/L)
35.4
17.1
32.5
30.9
32.5
7.2
12.9
8.9
17.0
18.7
41.5
35.0
14.4
36.8
34.9
40.0
27.5
28.2
35.3
40.2
38.7
12.9
28.9
70.5
20.0
17.4
7.7
16.7
27.1
28.5
70.5
7.2
Type of
Data/
Source of
Data
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
EAC/PES
BC/OSI
BC/OSI
BC/OSI
= Data not available.
EAC = Expressed Assimilative Capacity; BC = Biodegradation Capacity
PES = Parsons Engineering Science (Wiedemeier et. al., 1995a); GSI = Groundwater Services, Inc. (Newell et. al., 1996)
-------
If microbial kinetics were limiting the
rate of biodegradation:
If microbial kinetics were relatively
fast (instantaneous):
Anaerobic electron acceptors (nitrate and
sulfate) would be constantly decreasing
in concentration as one moved
downgradient from the source zone, and
Anaerobic electron acceptors (nitrate and
sulfate) would be mostly or totally
consumed in the source zone, and
Anaerobic by-products (ferrous iron and
methane) would be constantly
increasing in concentration as one
moved downgradient from the source
zone.
Anaerobic by-products (ferrous iron and
methane) would be found in the
highest concentrations in the source
zone.
Observed
Cone.
Cone.
Cone.
O2, NO3, SO4
FE2+, CH4
Observed
Cone.
Cone.
Cone.
O2, NO3, SO4
FE2+,CH4
The second pattern is observed at natural attenuation field sites (see Figure 1.1), supporting the
hypothesis that anaerobic reactions can be considered to be relatively instantaneous at most
petroleum release sites. The only cases where this might not be true is sites with very low
hydraulic residence times (very high groundwater velocities and short source zone lengths).
Biodegradation Capacity. To apply an electron-acceptor limited kinetic model, such as the
instantaneous reaction model, the amount of biodegradation that the groundwater that moves
through the source zone can support must be calculated. The conceptual model is that:
1. Ground water upgradient of the source contains electron acceptors;
2. As the upgradient ground water moves through the source zone, hydrocarbons in NAPLs
and contaminated soil release dissolved hydrocarbons (in the case of petroleum sites,
BTEX);
3. The biological reactions continue until the available electron acceptors are consumed (Two
exceptions to this conceptual model are the iron reactions, where the electron acceptor
ferric iron dissolves from the aquifer matrix; and the methane reactions, where the electron
acceptor CO2 is also produced as an end-product of the reactions. A simplifying
assumption can be made that the by-products ferrous iron and methane can be used as
12
-------
O)
o
15
14-1
C
u
g
o
1.0 T
0.5 --
0.0 —
40
20 --
4 --
2 --
0
0
» »-* *» »
Tyndall
-A—*-
-«t-
-A—A-A
/ —
200
400
600
800
* BTEX
• Sulfate
A Nitrate
* D. Oxygen
* Methane
• Iron
6 T
3
Hill
**±
*—•-
1*^
H
500
1000
1500 2000
O)
O
7=
ra
v
u
o
O
Patrick
200
400
600 800
BTEX
• Sulfate
A Nitrate
* D. Oxygen
• Methane
• Iron
10.0
5.0 --
0.0
-«-•
Elmendorf
ST-41
« »
20
10--
25 -r
/
200
400
*«—M
600
800
O
7=
2
I
o
o
25 -r
Keesler
100
200
300
Distance along plume centerline
* BTEX
• Sulfate
A Nitrate
* D. Oxygen
* Methane
• Iron
0.2
0.1 ...
0.0
Elmendorf
HG-10
6 -r
4 --
2 --
0 -P
-+-
1000
2000
3000 4000
Distance along plume centerline
Figure 1.1 Distribution of BTEX, Electron Acceptors, and Metabolic By-Products vs. Distance Along
Centerline of Plune.
Sampling Date and Source of Data: Tyndall 3/95, Keesler 4/95, (Newell et al., 1996), Patrick 3/94 (note:
one NO3 outlier removed, sulfate not plotted), Hill 7/93, ElemdorfSite ST41 6/94, ElemdorfStie HG 10
6/94, (Wiedemeier et al, 1995a).
13
-------
proxies for the potential amount of biodegradation that could occur from the iron-
reduction and methanogenesis reactions.);
4. The total amount of available electron acceptors available for biological reactions can be
estimated by: a) calculating the difference between the upgradient wells and source zone
wells for oxygen, nitrate, and sulfate; and b) measuring the production of by-products
ferrous iron and methane in the source zone;
5. Using stoichiometry, a utilization factor can be developed to convert the mass of oxygen,
nitrate, and sulfate consumed to the mass of dissolved hydrocarbon that are used in the
biodegradation reactions. Similarly, utilization factors can be developed to convert the
mass of metabolic by-products that are consumed to the mass of dissolved hydrocarbon
that are used in the biodegradation reactions. Tables 1.6 a through c illustrate the method
for calculating utilization factors for benzene, toluene, ethyl benzene, and xylene and
Table 1.7 lists the overall utilization factors for BTEX;
6. For a given background concentration of an individual electron acceptor, the potential
contaminant mass removal or "biodegradation capacity" depends on the "utilization
factor" for that electron acceptor. Biodegradation capacity is also referred to as
"Expressed Assimilative Capacity" or EAC. Dividing the background concentration of an
electron acceptor by its utilization factor provides an estimate (in concentration units) of
the assimilative capacity of the aquifer by that mode of biodegradation.
When the available electron acceptor/by-product concentrations (Step 4) are divided by the
appropriate utilization factor (Step 5), an estimate of the "biodegradation capacity" of the
groundwater flowing through the source zone and plume can be developed.
14
-------
Table 1.6a. Utilization Factor Calculation for Benzene
Aerobic
or
—»
C6H6 + 7.502
—»
6CO2 + 3H2O
1 mole benzene reacts with 7.5 moles oxygen
(6x12 + 6) gms benzene react with (7.5x32) gms of oxygen
78 gms benzene react with 240 gms of oxygen
Utilization Factor
240/78
3.08
Nitrate
C6H6 + 6H+ +6NO3-
—»
6CO2 +3N2 +6H2O
1 mole benzene reacts with 6 moles nitrate
Utilization Factor = 372.06/78
4.77
Manganese
15Mn4+ + 12H9O
6CO9 + 30H+ + 15Mn2+
1 mole benzene reacts with 15 moles manganese
Utilization Factor = 824.1/78 = 10.57
Iron
C6H6 + 30FeJ+ + 12H2O
—»
6CO2 + 30H+ + 30Fe
2+
1 mole benzene reacts with 30 moles ferric iron
Utilization Factor = 1675.5/78 = 21.48
Sulfate
6CO2 + 3.75H2S + 3H2O
1 mole benzene reacts with 3.75 moles sulfate
Utilization Factor = 360.26/78
4.62
Carbon Dioxide
C6H6 + 4.5H20
—»
2.25CO2 + 3.75CH4
1 mole benzene reacts with 3.75 moles CO2 (see Table 1.2)
Utilization Factor = 165/78 = 2.12
15
-------
Table 1.6b. Utilization Factor Calculation for Toluene
Aerobic
or
—»
C7H8 + 9O2
—»
7CO2 + 4H2O
1 mole toluene reacts with 9 moles oxygen
(7x12 + 8) gms benzene react with (9x32) gms of oxygen
92 gms benzene react with 288 gms of oxygen
Utilization Factor
288/92
3.13
Nitrate
C7H8 + 7.2H+ +7.2NO3
7CO2 +3.6N2 +7.6H2O
1 mole toluene reacts with 7.2 moles nitrate
Utilization Factor = 446.5/92
4.85
Manganese
C7H8 + 18Mn4+ + 14H2O
7CO9 + 36H+ + 18Mn2+
1 mole toluene reacts with 18 moles manganese
Utilization Factor = 988.9/92 = 10.75
Iron
C7H8 + 36FeJ+ + 14H2O
7CO2 + 36H+ + 36Fe
2+
1 mole toluene reacts with 36 moles ferric iron
Utilization Factor = 2011/92
21.85
Sulfate
C7H8 + 4.5SCV' + 9H
7CO2 + 4.5H2S + 4H2O
1 mole toluene reacts with 4.5 moles sulfate
Utilization Factor = 432.3/92
4.70
Carbon Dioxide
C7H8 + 5H2O
2.5CO2 + 4.5CH4
1 mole toluene reacts with 4.5 moles CO2 (see Table 3.3)
Utilization Factor = 198/92 = 2.15
16
-------
Table 1.6c. Utilization Factor Calculation for Ethylbenzene and Xylene
Aerobic
C8H10+10.502
—»
8CO2 + 5H2O
1 mole ethylbenzene/xylene reacts with 10. 5 moles oxygen
or (8x12 +10) gms benzene react with (10.5x32) gms of O2
—» 106 gms benzene react with 336 gms of oxygen
Utilization Factor
336/106
3.17
Nitrate
C8H10 + 8.4H+ +8.4NO3-
8CO2 +4.2N2 +9.2H2O
1 mole ethylbenzene/xylene reacts with 8.4 moles nitrate
Utilization Factor = 520.9/106 = 4.91
Manganese
21Mn4+ + 16H9O
8CO9 + 42H+ + 21Mn2+
1 mole ethylbenzene /xylene reacts with 21 moles Mn
Utilization Factor = 1154/106 = 10.89
Iron
C8H10 + 42FeJ+ + 16H2O
8CO2 + 42H+ + 42Fe
2+
1 mole ethylbenzene /xylene reacts with 42 moles ferric iron
Utilization Factor = 2346/106 = 22.13
Sulfate
5.25SCV'+ 10.5FT
8CO2 + 5.25H2S + 5H2O
1 mole ethylbenzene /xylene reacts with 5.25 moles sulfate
Utilization Factor = 504.4/106 = 4.76
Carbon Dioxide
C8H10 + 5.5H20
—»
2.75CO2 + 5.25CH4
1 mole ethylbenzene /xylene reacts with 5.25 moles CO2
(see Table 3.4)
Utilization Factor = 231/106 = 2.18
17
-------
Table 1.7. Utilization Factors for BTEX
Aerobic
Nitrate
Manganese
Iron
Sulfate
Carbon Dioxide
Notes:
Average 1 =
Average! =
B T E
3.08 3.13 3.17
4.77 4.85 4.91
10.57 10.75 10.89
21.48 21.85 22.13
4.62 4.70 4.76
2.12 2.15 2.18
Arithmetic Average
Mass Weighted Average
X
3.17
4.91
10.89
22.13
4.76
2.18
Average 1
3.14
4.86
10.78
21.90
4.71
2.17
Average2
3.15
4.88
10.82
22.00
4.73
2.17
18
-------
2.0 GETTING STARTED
2.1 Installing the Graphical User Interface
This section tells you how to install the Graphical User Interface Platform using the automatic
"Install" program and what you need to know before you start running the program. To begin with,
your computer should be set up and running MS Windows 3.x or MS Windows 95. The
BIOPLUME m Graphical User Interface Platform consists of several executables (.EXE), and
Dynamic Link Libraries (.DLL). In addition, the software package includes the Bioplume3.EXE file.
All Graphical Platform software runs readily on both Windows 3.x and Windows 95. The
Bioplume3.exe executable runs on Windows 95. It also runs on Windows 3.x if the Win32s library
is installed. The Win32s library is available from Microsoft (can be downloaded from their Web
Site). Even so, some network Windows 3.11 versions exhibit problems in running Bioplume3.exe
(fatal Windows exception, returning to DOS). Clearly, the preferred medium for running
BIOPLUME m is Windows 95.
To proceed with the installation, the user should also know how to use the Basic Windows/DOS
commands for creating and changing directories, copying files and disks, and listing directory
information.
For more information about these commands consult the documentation provided with your
equipment. In this section you can briefly overview the following topics:
• Microsoft Windows Fundamentals (optional for those who want to refresh their memory
on the basic commands of MS-Windows),
• What you need to get started,
• How to install the Graphical User Interface Platform for BIOPLUME m,
• Description of the Graphical User Interface Platform Menus,
If you are familiar with the MS-Windows operations, you can go directly to Section 2.1.2 and proceed
with the installation of the program.
2.1.1 Microsoft Windows Fundamentals
The BIOPLUME m Graphical User Interface Platform runs under Microsoft Windows-3.x and
Windows 95, and makes extensive use of many Microsoft Windows features. This means that you
can adjust settings in your system without having to adjust the Platform.
In order to be able to activate all Windows features, you must install Windows separately. For
instructions on how to install Windows and run applications see the Microsoft Windows User's Guide.
19
-------
Before you begin using the Platform you need to be familiar with a few basic Windows features. This
will be done concisely with the instructions of this section. Those familiar with Windows can move
directly to the next section which provides important information on how to install the software.
At this stage we will focus on the following topics:
• What are the components of a Windows Application,
• How to use the mouse,
• How to use the menus,
• What is a dialog box.
WHAT ARE THE COMPONENTS OF A WINDOWS APPLICATION ?
Many different applications can be accessed from a basic Windows environment. Each open
application is displayed in a new window on your screen. Windows applications are made up of
several common components. These components, as shown in Figure 2.1 are:
MENU BAR:
Appears beneath the Windows title bar and contains the names of all principal menus used in
the "BIOPLUME III" program.
PULL-DOWN MENU:
A list of menu items that is "pulled-down" from the menu bar by clicking on a main menu item.
Windows menus are also called pull-down menus.
MAXIMIZE BOX:
A small box with an up-bar icon in the window's upper right comer. Allows the user with a
mouse to enlarge a window to its maximum size.
MINIMIZE BOX:
A small box with a down-bar icon in the window's upper right comer. Allows the user with a
mouse to shrink a window to an icon.
CLIENT AREA:
The "work area" of the screen over which the application has complete control.
20
-------
SMARTICONS:
Smartlcons are mouse shortcuts for "BIOPLUME III" features, functions, and commands.
"BIOPLUMEIII" displays a palette ofsmartlcons on the right-hand-side of the screen in a child
window that can be moved around the screen to fit the needs of the user by clicking on the
"Select" display area.
• • 'v.,'fv;-=.'.^;,;:;
File Domain Leading Edit find Initial Conditions Simulator Results View Annotation
[ D|
100 :
200 :
300 :
:
400 -
500 =
600 :
700 :
800 :
900 =
1000:
Q*|^
100
m
£Jfe
200
H
300
H?|^
1
400
1^|l^|?^|tfr|tdr|tatBigiBite
500
600
700
i i i n in [
800
1 1 1 Ml 1
1000
•Tnon r^^p Fi I^^^^^^^^^^^^^H
•
^1
BIOPTST1
^^^^^^^H
BIOPTST2
EGLIN
EGLIN1
EGLIN2
HILLAFB
HILLAFB1
LKTEST
PATRICK
TEST01
TESTP30
TESTP31
TESTP311
TESTP32
TESTP33
. ' , • i
i ..
0, ]
i Cancel ij
i
0
, 0.
"", '•; . '^^^s-
[%
d*
*
LI
M
^
«k^
•——
w, ^
(CT
%v*
H
^
^
M
'^^^"
A -,
I*S*I
^&
^f
*%
^»2)
y
' f '
i
Figure 2.1. Components of an MS-WINDOWS Application.
USING A MOUSE
A mouse is a hand-held pointing device. As you move the mouse across your desk, a pointer moves
on the screen. You can pick up the mouse and reposition it without moving the pointer on the screen.
All BIOPLUME m Platform actions require the main mouse button. These actions are the following:
POINT: Move the tip of the mouse pointer on top of an object on the screen.
CLICK : To quickly press and release a mouse button.
CLICK and DRAG: To press and hold a mouse button while dragging the mouse to highlight an
area.
21
-------
DOUBLE CLICK: To quickly press and release a mouse button twice in
succession.
RELEASE: To quit holding down a mouse button.
SELECT : To point on a menu or to highlight text or graphics so they will be affected by the next
action you take within the Platform.
The pointer assumes different shapes to denote different functions as you proceed with different tasks
of the Platform.
USING PULL-DOWN MENUS
Menus are lists of commands. When you select a menu in the Graphical Platform it drops down on
your screen showing all the items you can activate from that menu.
To select a MENU:
1. Point on the name of the menu you want.
2. Hold down the main mouse button. (The menu drops down on your screen.)
To activate an ITEM:
1. Select the menu that contains the item you want
2. Select the item you want. (The selected item is highlighted.)
When you click the main mouse button, the Platform carries out the action specified by the
highlighted item.
WHAT IS A DIALOG BOX ?
In many cases the Platform needs additional information from you before it can carry out a specific
command. In that case, the Platform displays a dialog box for you to fill in the information. Once a
dialog box appears, you must fill it in before continuing on. Sometimes you will type in text. Other
times you will simply select an option within the dialog box. Each dialog box has different kinds of
"Controls" that the user can select. They are as follows:
22
-------
Type of Control
Scroll Bar
List Box
Check Box
Text Box
Option Button
Command Button
How it Appears on the
Screen
Two arrows, two white
boxes and a white area
A set of items that you can
select from.
One option
Text or empty space
A set of options
A command name
What is your Action
Click on the arrows or the gray
area, or drag the small white box to
move the viewing area.
Drag the pointer down the list to
highlight the option you want and
click to activate it.
Click once to turn the option on,
click again to turn it off.
Review the text and if necessary
type in appropriate text.
Click the option you want.
Click once to carry out the
command shown inside the button.
These are typical controls needed to effectively use the Platform as illustrated in Figure 2.2. In
particular you enter the required input parameters using the corresponding "Text Editing" box. Then
you use the "Command Button" "OK" to accept these input values and move on to other modeling
activities.
Concentration Domain
Concentration (rng/L or |ig/L)
Default:
Minimum:
Maximum:
0.
100-
Cursor Increments
2.
Ok
•Ruler Tic Increments
Major:
Minor:
10.
Cancel
Figure 2.2. Typical Controls in a Dialog Box.
23
-------
2.1.2 What You Need to Get Started
First, make sure you have the correct equipment. While setting up MS-Windows 3.x or Windows 95,
you will be prompted to identify your pointing device, keyboard, printer(s), graphics adapter, and
monitor. Windows will copy the necessary driver files to your hard disk. We recommend that you use
equipment with the following specifications to run the BIOPLUME El Graphical User Interface
Platform:
IBM- Compatible Personal Computer, with Intel 486, or (preferably) Pentium processor.
Your system should include an 800 MB or larger (IGigaB) hard (fixed) disk.
At least 8 MB of RAM memory- possibly 16 MB, especially if you are using memory-resident
programs on a network, or Windows 95.
A Microsoft Windows -compatible graphics adapter and a compatible color graphics monitor.
Microsoft Disk Operating system (DOS) version 3.0 and above.
A Windows-compatible pointing device (mouse).
A Windows-compatible printer.
In general, if your equipment can run and print from MS-Windows, you can run and print
from the BIOPLUME m Graphical User Interface Platform.
2.1.3 How to Install the Graphical User Interface Platform
The easiest way to install the Graphical User Interface Platform is to rely on the "Install" program.
Once you start, the Install program prompts you on the screen for all the operations that need to take
place.
To Install the Platform proceed as follows:
1. Turn on your PC,
2. Insert the Platform Install disk (disk 1) into drive A or B,
3. To activate the MS-Windows-3.x environment, at the prompt > type WIN and press
the Enter key. For Windows 95, or if you are already in Windows proceed with step
4. From the Program Manager or File Manager (Start, Run in Windows 95) select file -
RUN, type A:\INSTALL or B:\INSTALL, and press the Enter key. The install
program begins.
5. Follow the instructions on your screen.
24
-------
Note: MS-Windows-3.x or Windows 95 must be previously set up on your system before installing
the BIOPLUME El Graphical User Interface Platform. You may find answers to any questions by
consulting the section "WHAT YOU NEED TO GET STARTED."
As you insert disks, the Install program copies the necessary files from the Platform disks onto your
hard disk. You can change the default drive and directory names of the install program to any drive
and names of your liking, when prompted.
The install procedure will create a sub-directory C:\EISBIOP on drive C: or any other drive of your
choice and will decompress all the executable files needed to run the program. Then, it will copy in
this new directory all the files required to run the different tutorials. Once the installation procedure is
completed, you will be automatically placed in the new sub-directory. Note that to run properly the
Platform you need at least 3MB of memory available on your hard disk. Consult your "Microsoft
Windows User's Guide" for complete information about Microsoft Windows-3.x or Windows 95.
BIOPLUME III
To check if everything is running properly test run the program: a
Group Application " Platform" has been created automatically by the
"Install" program; double click on the icon representing the
BIOPLUME m program.
This will activate the Platform to run under MS-Windows-3.x or Windows 95. Now you are ready to
consult Sections 2.2 and 2.3 to quickly navigate through the program and check that everything is
properly installed.
2.2 Description of the Platform Controls
2.2.1 Operating the Graphical User Interface Platform
Commands and Controls
Learning and using the BIOPLUME m Graphical User Interface Platform is easy and natural. The
system arranges windows in a hierarchy of parent, child, and sibling, starting with the desktop
(background) window. Each window is an Instance of a window class and each class has a window
procedure. All the user needs to know is the controls that allow him or her to activate these window
classes and associated procedures.
The Platform offers a very powerful set of controls that allow the user to build a Case Study "on the
fly." This set includes:
• A Menu of Program operations and,
• a Tool Box.
The Menu provides access to all operations of the Graphical environment: from file management, to
graphical editing, to the activation of a particular simulation and visualization of the results; the menu
control gives instantaneous access to all the tasks necessary for the simulation of a groundwater
contamination episode.
25
-------
The Toolbox provides the user all the necessary tools to build a model. The basic modeling features
(objects, lakes, wells, logpoints, etc.) are contained in the toolbox for easy access. All the user has to
do is point, click, and drag. Et voila! The selected modeling feature is created on the spot.
Figure 2.3 illustrates the general configuration of these various controls as they are displayed on the
screen.
File [Jornain Load ng Edit Grid Initial Conditions Simulator Results View Annotation
Figure 2.3. Graphical User Interface Platform Menu and Toolbox.
2.2.2 Description of the Graphical User Interface Platform
Menus
All Platform options fall under ten basic Menus. Each of the main menus is associated with secondary
pull-down menus which give access to the various options, allowing the user to generate pertinent
input data, and activate different tasks of the program. There is a logical sequence to activating these
menus. A particular case study necessitates several iterations, starting from a simple model and
26
-------
adding more refinements until we reach the desired accuracy. The Menus in the program are designed
to assist the user in the difficult tasks of model calibration and validation of the results.
Each Menu is in fact an editor with its own particular functions. These Menus are as follows (refer
also to Figure 2.3):
Menu Name
File
Domain
(Editor of Global Parameters)
Loading
(Editor of Heads and Concentrations)
Edit
(Editor of Modeling Features, Wells,
Sources, Lakes...)
Grid
(Editor of Boundary Conditions and all
Distributed parameters inside the
Modeling Grid Area)
Initial Conditions
(Editor of Simulation Period and Initial
Conditions)
Simulator
Results
(Graphical Editor of Simulation Results)
View
(Editor of Viewing Configurations)
Annotation
Menu Function
Performs all file management operations,
open, save, restore, delete, close, view file
content.
Control parameters defining the geometry of
the groundwater problem and the time
domain. Appropriate selection of the cursor
resolution.
Defining all existing loading (Hydraulic
heads, and concentrations) as a function of
time.
All editing capabilities for the modeling
features given in the toolbox for the
groundwater contaminant migration problem.
Definition and generation of the grid
geometry used for different resolution
processes. Editing of cell properties,
constant/variable flow, inactive cells.
Selecting initial conditions for the
simulation.
Selecting appropriate Simulation module to
run.
Visualization of all data related to the results
of various analysis options.
Select/Remove features appearing on the
screen of the Platform.
Activating/deactivating Annotations in all
graphical screens.
The BIOPLUME m Graphical User Interface Platform is a WYSIWYG ("What You See Is What
You Get") application. The Platform shows you on screen exactly how a document will appear when
it prints. It also adheres to Microsoft Windows conventions for using menus, menu commands,
dialog boxes, command buttons, option buttons, list boxes, check boxes, and a mouse. The mouse
27
-------
pointer becomes an essential tool in the Platform and is the most efficient vehicle to build a simulation
model. In that respect, the prompt assumes different shapes according to each editing mode, as shown
in the next section.
2.2.3 Navigating Through a Simulation
Typically, the user starts by opening a new file (case) and proceeds to the menu "Domain" to
characterize the geometric boundaries of the problem. Modeling tools available to the user include:
• mouse pointer for selecting modeling features for editing
• creation of wells
• creation of pollution source, and recharge areas
• impose boundary and initial conditions
Using the Smartlcons of the toolbox the user then selects the basic features of the model and proceeds
to the menu "Edit" to input their properties. The feature edit option is also accessible by double-
clicking on the feature (e.g. well) in the geographic domain. Similar tools are also available for editing
the numerical grid and for specifying boundary conditions (constant head, concentrations, general
head boundary).
The next step is to determine the loading conditions in the simulation through the options of menu
"Loading." Loading features, such as hydraulic heads and concentrations, pumping schedules, and
other boundary conditions, require the specification of time series. This operation is automated in the
Platform, where entries are limited to the times of change in loading attribute. All simulation time-
stepping is done automatically, with values interpolated at the simulation required time steps.
After defining the initial conditions, the program is now ready for activating the simulation (Menu
"Simulator"), and for viewing the results (Menu "Results"). If the results are not satisfactory,
several options are offered; the user can change the simulation domain, alter the loading parameters,
readjust the simulation grid, or redefine the initial conditions. The beauty of the program is that these
changes and alterations are built on-the-fly, without the need to reenter any of the fundamental data.
The program cleverly assists the user on each step, and keeps track of all the new parameters that enter
the simulation.
It is very simple to navigate through the Platform using the mouse pointer which helps you activate the
different menu options and select the appropriate commands and modeling features. What
distinguishes this program from other software is the fact that it provides the user with a completely
integrated computer environment for all modeling tasks: input data preparation, execution, and
analysis with interactive graphics, geostatistical (kriging) routines for input error control and optimal
use of existing geological information; and expert assistance in all phases of the simulation. The
Platform supervises the generation of all data needed to run the flow and migration problem as shown
below. The user operates in an "Object-Based" environment which offers remarkable flexibility in
28
-------
making the appropriate adjustments needed in the simulation of the groundwater contaminant
migration problem.
2.3 Checking Platform Installation
You can activate the Platform from either a DOS prompt or from the Program Manager window if
Microsoft Windows is already running (always the case for Windows 95).
To start the "Platform" from DOS:
• Display a DOS prompt for the drive that contains Windows. For example, C:\
• Type: WIN BIOPLUME. ( This assumes the existence of the appropriate set path
command in the autoexec.bat file)
• Press Enter.
BIOPLUME III
To start the "Platform" from within Windows:
• Display the Program Manager window.
• If necessary, open the group window that contains the "BIOPLUME" icon.
• Double click on the "BIOPLUME" icon.
For the implementation of this software architecture, the Platform sets up several sub-directories
(folders in Windows 95) to manage the flow of different software operations. Table 2.1 describes the
sub-directories that are automatically constructed during the installation procedure of the program.
File CONFIG.INI in sub-directory '..\CONFIG' initializes the version of the Platform activating the
appropriate modules. All the executables of the program reside in sub-directory '..\RELEASE'which
houses the engine of the program. Sub-directory '..\DATA' contains all the files pertinent to a
particular application. Sub-directory '..MMAGE' stores all the bitmaps (raster images) that are
necessary to build a remediation study. Finally, sub-directories '..MMPORT' and '..\EXPORT" contain
all pertinent peripheral data that need to be imported or exported from the platform for a case study.
But the heart of the platform resides in the user interface with its process scale operator. It controls all
simulation activities through the "Configuration" file, "Menus" and menu options.
29
-------
Table 2.1. Sub-directories of the BIOPLUME m Graphical User Interface Platform
Sub-directories
Config
Release
Data
Image
Import
Export
Report
Description
Contains the file with list of active Configuration options
Contains all executables and Dynamic Link Libraries (DLL) of
different modules of the platform
Contains all sub-directories related to different applications. The
name of these sub-directories corresponds to the name of the
different current applications.
Contains all the raster background images (*.BMP files).
Contains all pertinent Import Files
Contains all pertinent Export Files
Contains all pertinent Report files and result bitmaps
These sub-directories are all created automatically by the installation program. This program also
installs and checks the content of each sub-directory. However, you still need to ascertain that these
'Execution', 'BMP' (Graphics) and 'AW (Animation) files work properly on your system by running
the following cases:
1. Check that the BIOPLUME IE executables work properly
2. Check the Graphics executables that handle the background image
3. Check the Animation executables.
Note that the entire procedure requires only a few minutes of your time. All the input files and data
needed to run these cases are already installed by the installation program and the only thing that you
are asked to do is to activate the appropriate modules of the program following the instructions given
below.
2.3.1 Checking the Platform Executables
Test that all files are properly installed by following the steps below:
Step 1. Use the mouse to go to menu "File" and select option "Open". Among the different
input files that exist in your directory select case "TESTP31". Double click on TESTP31.
30
-------
Automatically all appropriate files are loaded into the Platform and the name of the opened
case is displayed at the top of the screen as shown in Figure 2.4.
Step 2: At this stage all input parameters needed to run this simple case are available to the
system, and all that is required from you is to initialize your particular run. This is done by
moving to Menu "Initial Conditions" and selecting sequentially (in the same order) options
"Simulation Period", "Starting Heads" and "Starting Concentrations". These options activate
appropriate dialog boxes in which you define the simulation period, the initial heads and
concentrations for your particular run. Figure 2.5 shows the dialog box allowing you to enter
the simulation period. You do not need to enter any value, just click on the "OK" button. You
do the same for the menu options 'starting heads' and 'starting concentrations'.
Step 3: You are now ready to execute the simulation. Move to Menu "Simulator," select
(click on) Bioplume HI and click on the "Save Data and Run Simulation" button. It should
take only a few minutes to run the BIOPLUME m algorithms, (sequentially, close window
Wbiop3.exe which creates the input stream; then close window Biopl3.exe when the cursor
starts blinking, after BIOPLUME m has finished executing; finally, close window Pbiop3.exe
after the graphics files have been executed, as explained in the next Step. "Exit Code zero"
signifies a successful run).
g^BIOPLUME III TESTP31 [Main Menu]
File Domain Loading §dit Grid Initial Conditions Simulator Results View Annotation
D
mi
100
2QOC
1000
2000 =
3000 =
4000 =
m
5000 :
GOOD:
7000 =
8000 =
9000 =
30001 40001 50001 60001 70001 800C
JMjJjj^yjMjjJjjjjMjjjJjjjMjjj^ jjjnjjjjJjjMjjjj^
•j--
m
m
m
m
m
K-K+HS&
.. 2290.
:CE
11
Figure 2.4. Screen View of Case Study "TESTP31".
31
-------
Simulation Period
Select Starting Time
Time
(Years)
Combined
Pumping Rale
Observed
Heads
Observed
Concentrations
•Starting Dale —
Feb 1, 1990
Ending Time (Years)'
2.5
Figure 2.5. Defining the Simulation Period.
Step 4: Three different modules are now activated sequentially by the program. All you have
to do is to click on the "Yes" button when the first module (Data file module) is finished, click
on the "No" button when the second module (BIOPLUME m module) is finished and finally
click on the "Yes" button when the third module (Graphics file module) is finished (Note that
"Exit Code zero" signifies a successful run). At this point if there are no messages displayed
on the screen, you have an indication that the execution and graphics files worked properly.
You can now visually inspect the output results in Menu "Results" under option "Hydraulic
Heads". Figure 2.6 shows the contours of the computed hydraulic heads at time 2.5 years that
you should obtain on the screen.
Note the cone of depression due to the pumping well. Also on the contour levels of the
hydraulic heads, you should read a maximum head of 100 ft. and a minimum head of 99.51 ft.
A final visual check requires activation of the concentration results.
Step 5: In Menu "Results" activate option "Concentrations/Hydrocarbons" to obtain the
computed distributions of the Hydrocarbon concentrations at time 2.5 years as shown in
Figure 2.7. As it can be seen this is a uniform field across the aquifer. Hydrocarbons
decreased from 100 ppb to 96 ppb after 2.5 years due to the biodegrading action of Iron
reduction. If these are the values that you read on your screen (Contour levels) then you have
successfully completed the installation of the test case execution files.
32
-------
< ii A <
••fffi
Cttl JiMl 30CQ Ji'Ml WM tWQ TOCO 6COQ
I f TTr MSI
a
Figure 2.6. Computed Hydraulic Heads at Time 2.5 Years.
IBIOPLUMEIII IIMf31 IHidiKlltMfl U«K
"- LI
LLJi
IBS.
ga>
UMf
• '.| •:•! ;.ij» i-m
WM HOT
Figure 2.7 Computed Concentrations of Hydrocarbons at Time 2.5 Years.
33
-------
2.3.2 Checking the Graphics and the Background Image
To check the graphics drivers you need to open a new file case ("HILLAFB1") and proceed with the
following steps:
Step 1. Use the mouse to go to menu "File" and select option "Open". Select case "HILLAFB1".
Double click on FflLLAFBl. Again automatically all appropriate files are loaded into the Platform.
This case study simulates the Hydrocarbon migration at the Hill AFB UST 870 site.
Site Characteristics
The UST 870 site at Hill AFB covers an area 2600 x 2000ft. It is seated on a plateau-like bench
formed by river deposits of the ancient Weber river. There are three aquifers present in the area.
However, the hydrocarbon contamination is believed to be limited to the shallow (unconfined)
aquifer, which is the subject of this simulation. Groundwater flow in this aquifer is in the SW
direction. Total dissolved BTEX within a contour level of 70 ppb is considered as the source of
contamination, as shown in Figure 2.8. More details of this case study can be found in the referenced
tutorial documents.
[Jornain Loading £dit (3rd mitial Conditions Simulator Flesults View Annotsbor
^d Myy M
>-\ 2000 2400 ;
V BTEX Source '
Figure 2.8 Screen View of Case Study "FflLLAFB 1".
34
-------
Now on your screen you have an image similar to Figure 2.8. Note that the background of the
working area is covered by the raster image (Bitmap) available for this case. The bitmap can be
found as file FnLLAFEVI.BMP in sub-directory "..Mmage". The graphics drivers automatically load
the image when you open the case. If for some reason you do not get the background image, check
whether the HILLAFIM.BMP file is corrupted using any available graphics program like "Paintbrush"
or "Paint".
Step 2: In this step you will only view the results of the simulation and check if the color palette is
loaded properly. Move to Menu "Results" and sequentially activate the following options:
"Concentrations\ Hydrocarbon" and "Concentrations\ Oxygen". On the screen you will obtain the
illustrations shown in Figure 2.9. On the left you have the hydrocarbon contaminant plume as it is
displayed on the screen at the end of a 1 year simulation and on the right you have the oxygen
depletion at the end of same time period. (Note that we removed the graphical representation of the
recharge zone at the boundaries by deactivating option " View\ Features"
Figure 2.9 Plume Migration for BTEX and Oxygen After a 1 Year Simulation.
As it can be seen, the computed hydrocarbon and electron acceptor concentrations are superimposed
on the raster image of the site. This greatly facilitates the location of points of compliance and long
term monitoring wells.
2.3.3 Checking the Animation Executables and Files
Finally, to complete the installation check-list you need to verify if the video animation (AVI) files
work properly. The video animation option runs on Windows 95 only. Note that the standard format
for Windows digitized video is the Audio-Video Interleaved (AVI) format. An AVI file can be played
in Windows with no additional hardware (of course it will be smoother and faster with a video
accelerator). The Platform supports Microsoft Video for Windows 95 AVT-format (*.AVT) video
files. To verify the video drivers you do not need to open another study, just continue with the
FflLLAFBl case, using the following step.
-------
Step 1: Move to Menu "Results" and activate option "AVT Animation". This will invoke the
animation module. As it can be seen a new menu bar appears at the top of the screen. Move to Menu
"File" and click on the option "Open AVT'. A dialog box appears on the screen with the list of all
available video clips (.AVT) files Select the file "HILLAFB.AVI" to obtain the screen shown in
Figure 2.10. To playback the video clip showing the simulated migration of hydrocarbons, just click
on the "Forward" play button that appears at the bottom left corner of the AVT window.
; EIS/GWM AVI Video Animation
M^vfe Is^do^j Help
-J HILLAFB.AVI JO - ttopped]
Figure 2.10 Playback Screen of AVT Files.
Et voila! The screen comes to life and the video clip stops after a few seconds. The detailed procedure
on how to create this AVT file is given in the User's Guide manual. All you need to know at this point
is that the "FflLLAFB.AVI" file was generated from only 4 Bitmaps (snapshots) depicting the
simulated plume at times 0.25, 0.50, 0.75 and 1 year. These bitmaps were selected and created using
the grasping tool activated from the available Platform tool box.
At this stage, if no error messages are encountered, the installation is successful and you may proceed
with the implementation of your own case studies. You can also consult the other sections of this
Manual, in particular: the Platform User's Guide, and the Tutorial.
36
-------
2.4 Checking the Installed Case Studies
Each case study is identified by its name and all pertinent input and output files reside in a sub-
directory (folder) that bears the same name as the case study and is located in sub-directory (folder)
C:\EISBIOP\DATA\. There are two categories of case studies that are installed with the program,
namely:
1. Simple academic cases that show the fundamental features and operations of the
program. The detailed description of their operations is given in the tutorial manual;
and,
2. Real case studies from different Air Force Bases across the U.S.
Table 2.2 lists all the installed real case studies while Table 2.3 shows the complete list of all the
installed test cases.
Table 2.2 List of Installed Real Case Studies.
Configuration
: i'1 . • i - "-
, , .- i. > , ,. , . ..,",, . L". ! •
Name
FflLLAFB
Description
Hill Air Force
Base
Features
Recharge +
Source
Size
25x20
37
-------
Table 2.3 List of Installed Test Cases.
Configuration
, ,._.,.__.._,.
, "*
.
Name
TESTP30
TESTP31
TESTP32
TESTP33
TESTP34
TESTP35
Description
Base case
Base case with
optimal input setup
Testing Drains
Testing Sources
Testing Lakes
Testing Rivers
Features
Recharge zone +
Well
+ Iron Reduction
Recharge zone +
Well
+ Iron Reduction
Recharge zone +
Well
+ Iron Reduction
+Drain
Recharge zone +
Well
+ Iron Reduction
+ Source
Recharge zone +
Well
+ Iron Reduction
+Lake
Recharge zone +
Well
+ Iron Reduction
+River
Size
9x10
9x10
9x10
9x10
9x10
9x10
38
-------
2.5 Concluding Remarks
This concludes the installation guide. The main objective of this Section is to guide the user through
all the installation procedures. A brief description of the basic features of the program allows the
user to quickly navigate through the platform. However, to get a better insight about the proper use of
the program we suggest to also consult the following Sections:
• Platform User's Guide
• Tutorial
• BIOPLUME m Theoretical Development, and
• Implementing the Air Force Protocol for Intrinsic Remediation.
39
-------
3.0
TUTORIAL
This section gives a series of tutorial examples which allow the user to learn by example how to
operate the Graphical User Interface Platform using the Platform menus and tools. Before
presenting the tutorial examples, we start by summarizing the fundamental steps required to
perform a complete simulation.
If the Platform software is not already running, start the Platform by double clicking on the
"BIOPLUME III" icon on your Windows Desktop:
BIOPLUME III
The main page of the Bioplume HI application looks as in the picture below. After closing (O.K.)
the "About Box," the user will have access to the main Menu which lists the following entries:
File,' Domain,' Loading,' Edit,' 'Grid,' Initial Conditions,' 'Simulator,' Results,' View,' and
'Annotation.'
File Domain Loading £dit ]3rid Initial Conditions Simulator Results View Annotation
E
100 =
200 =
300- =
400 =
500 =
600 =
700 i
800 =
900 =
1000 =
100
200
300
40C
500
600
700
80CI 900
1000
m
BIOPLUME III
Version 1.1
Developed for AFCEE
Air Force Center for Environmental Excellence
Brooks Air Force Base, San Antonio, Texas
I 370., 290.
£.;:;.tj(.;'; ^x_
c*
-------
• Specify any/all time series data of state variables -hydraulic heads and concentrations
(menu 'Loading'),
• Specify the numerical grid for simulation and perform parameter interpolation -kriging
(menu 'Grid'),
• Select initial conditions for simulation among previously entered Loading' data (menu
'Initial Conditions'); then,
• Perform Bioplume m simulation (menu 'Simulator') after selecting run-time options;
and finally,
• Graphically view the results of the simulation (menu 'Results').
Menus 'View' and 'Annotation' are service menus allowing to control the layout of the computer
screen. Also, menu Edit' is operated in conjunction with the 'Tool Box' displayed at the top right
corner of the screen. With the pointing device (mouse) the user selects features (e.g. wells,
pollution source, recharge areas), and places them on the screen within the modeled Domain and
specifies their properties.
Using the Smartlcons of the toolbox the user selects the basic features of the model and proceeds to
the menu 'Edit' to input their properties. The feature edit option is also accessible by double-
clicking on the feature (e.g. well) in the geographic domain. Similar tools are also available for
editing the numerical grid and for specifying boundary conditions (constant head, concentrations,
general head boundary).
The next step is to enter any time series data for hydraulic heads and concentrations using menu
'Loading.' All above data are real (not interpolated) data at log-points, and over appropriate time
intervals.
As in any software application (e.g. Word Processing), the user is encouraged to frequently "save"
his/her work, especially after entering new data, modeling features or Loading' data. This is done
by activating the 'Save' option in menu File.' (The 'Save As' option also allows the user to create a
new case name from a previous case, useful when creating a suite of "scenarios"). However, the
save operation in the Platform also automatically triggers a data synthesis operation, i.e. mapping of
log-point data to the numerical grid via Kriging. Therefore, the save operation is the last operation
to do prior to performing a simulation. The program is now ready to initiate a simulation.
The first step in preparing for an actual simulation is the selection of initial conditions (menu
'Initial Conditions.' The Platform is now ready for activating the simulation (menu 'Simulator') -
without performing a "save" operation again!
Then, for viewing the results graphically activate menu 'Results'. If the results are not satisfactory,
several options are offered; the user can change the simulation domain, alter the loading parameters,
readjust the simulation grid, or redefine the initial conditions. The beauty of the program is that
41
-------
these changes and alterations are built on-the-fly, without the need to reenter any of the fundamental
data. The program cleverly assists the user on each step, and keeps track of all the new parameters
that enter the simulation.
The best way to learn is by example: you are now ready to apply these general operating principles
to the tutorial examples of this section.
3.1 Tutorial Overview
The tutorial examples are grouped in sessions. The user is provided with instructions on how to
develop and run a new simulation. An "academic" example of a study area is provided as an
example. The attributes of the study area are highly simplified so that operational concepts can be
developed and tested without undue complexity.
The goal in this tutorial is to systematically develop site characteristics and test the model
incrementally. To that end, instructions are given in the form of six sessions:
• Session 1: Basic model development;
• Session 2: Ground water flow modeling;
• Session 3: Non-attenuated hydrocarbon mass transport;
• Session 4: Addition of electron acceptors and biological interaction;
• Session 5: Introduction of special features (wells, drains, lakes, etc.)
• Session 6: Animation and graphical presentation (Windows 95 version)
The Graphical User Interface Platform uses a hierarchical menu system using parent menus which,
in turn, activate a series of child or sub-menus. For many applications, input is required in several
sibling menus. To simplify the tutorial the following convention is used:
• Each menu or submenu is written in bold and italic typeface and preceded by a ^
symbol which can be interpreted as the command "go to".
• Each submenu will be sequentially indented under the parent menu. For example to go
to the parent menu "Domain" then the submenu "Surface Domain" would be written:
-> Domain
-> Surface Domain
• Entry fields listed on a submenu are written in bold and italic type face and are also
indented according to hierarchy.
• Data to be input will be placed in brackets. For example the instruction to enter 30%
porosity is given as [.30]. In the program you'd only enter 0.30 not the brackets!
• Commands to be executed with the mouse or keyboard are enclosed in <> symbols. For
example, after data is entered you may need to use the mouse to click on the OK button.
Here this step is written as .
• Where the enter button is to be pressed the following is used J.
• Grid coordinates are entered as (x-number, y-number)
42
-------
• All instructional comments within a command string are written in italic text. For
example, to describe the "Domain" level:
-> Domain - Used for defining the range of minimum and maximum values
to be used in the model and appearing on graphs that follow
Before beginning, copy the base images of the test site to the default EIS image directory. The base
images are provided on the accompanying floppy disk as Test01.bmp, Test02.bmp, Test03.bmp,
and Test04.bmp. Using DOS or the windows File Manager copy those files to
C:\EISBIOP\EVIAGE\ assuming C:\ is the root directory you installed the Platform on (check first,
these files may have already been copied from your self-installation package).
3.2 Session 1: Basic Model Development
Turn on your computer and click on the Bioplume icon to start running the program. The program
will appear on your screen along with a default uniform 10 by 10 model grid with an x and y range
of 0 to 1,000 feet.
Main
-* File
•} Save As [TestOl] - Saves the session as "TestOl "file name
•} Domain - Used for defining the range of minimum and maximum values to be used in
the model and appearing on graphs that follow. In some cases this directory also
establishes default aquifer parameters.
•^ Surface Domain
Surface Bounds
Left [0] - Sets minimum x grid value range for work area to 0
Right [2000] - Sets maximum x grid distance for work area to 2000
feet
Top [0] - Sets minimum y grid value range for work area to 0
Bottom [2000] - Sets maximum y grid value to 2000
Ruler Tic Increments
Horizontal Major [100] - Sets horizontal major ruler tic increment
to 100 for work area ruler
Horizontal Minor [10] - Sets horizontal minor ruler tic increment to
10 for work area ruler
Vertical Major [100] - Sets vertical major ruler tick increment to
100 for work area ruler
Horizontal Minor [10] - Sets vertical minor ruler tick increment to
10 for work area ruler
The 10 by 10 grid appears to have shrunk by half and now resides in the upper left corner of the
work space. Do not panic! This is normal and we'll fix it now.
43
-------
Main
-> Grid - Menu used for defining the model grid and major site aspects represented on that
grid. In addition to defining grid geometry, boundary conditions for hydraulic heads and
the concentrations of the contaminant and electron acceptors can be established across the
grid in this menu.
-> Generate Grid
Computational Bounds
Left [0] - Sets minimum x-direction grid value to Ofeet
Right [2000] - Sets maximum x-direction grid values to 2, 000 feet
Top [0] Sets minimum y-direction grid value to Ofeet
Bottom [2000] - Sets minimum y-direction grid value to Ofeet
Grid Size
Number of Columns [20] - Sets the number of columns in the grid
to 20 and automatically establishes that each cell will be 100 feet in
the x-direction,
Number of Rows [20] - Sets the number of rows
in the grid to 20 and automatically establishes that each cell will be
100 feet in the y-direction.
Note that under -> Grid -> Generate Grid, the "Grid Increment' entry items Column Increment
and Row Increment both changed to 100 feet when the number of rows and columns was set under
Grid Size to 20. It would have also been possible to directly set the Column Increment and Row
Increment to 100 then the values under Grid Size would have automatically been set to 20. It's
time to save your work:
Main
You will now notice that the model grid extends all the way across the work area. A useful aspect
of the Platform is that the grid geometry (number of rows or columns or length of the grid) can be
changed at any time during a simulation without affecting any previously entered model data (wells,
log-points, strata elevations and other features).
Next we want to incorporate the base image for the test site onto the working grid. The Platform
accepts windows meta (*.bmp) files. These files can be generated from many different graphics
programs including Paintbrush, CorelDraw!, AutoCad Rel. 12, and others. To import the
TestOl .bmp image for this session:
Main
-> Domain
-> Base Image
The site background image should now be pasted on the work area; however, you will notice that
the image does not conform to the ruler. The test site has dimensions of 2,000 by 2,000 feet and
44
-------
needs to be "registered" to the same dimensions as the grid. To adapt the image, click on the
Register button on the tool box which looks like this:
Note that the normal pointer icon changes
to a star-burst pattern. Move the pointer to the
upper, left-hand corner of the site image and click and hold down on the mouse button. While
holding down on the button drag over to the lower, right-hand corner of the site image. You'll note
that as you drag, a black box will form. When you position the cursor on the lower, right corner of
the image release the mouse button. The entire picture of the site should be covered with a dark
gray shading and a "Register" menu will appear. Enter values as follow:
^Register
First Point
X: [0] - 0 foot x-position for the first register point we entered in the upper
left hand corner
Y: [0] - 0 foot y-position for the first register point we entered in the upper
left hand corner
Second Point
X: [2000] - 2000 foot x-position for the second register point we entered in
the lower right hand corner
Y: [2000] - 2000 foot y-position for the second register
point we entered in the lower right hand corner.
The base map should now fit nicely on the defined model grid. If the image appears correct then
save the session:
Main
If it does not look right you'll need to redo this last step by exiting without saving and reopen the
TestOl model file:
Main
->File
-> Open
3.3 Session 2: Basic Flow Modeling
3.3.1 Domain and Boundary Conditions
Now we want to continue defining model range and basic aquifer values in preparation for flow
modeling under static (steady-state) conditions.
Main
•^Domain
45
-------
•^Elevation Domain - Defines default range for the top and bottom of the aquifer.
Note, that this does not set the actual top and bottom of the aquifer for the model; it
merely defines the range over which we will work. Set the top and bottom of aquifer
at some nice even increment slightly above and below the highest actualpiezometric
surface for your system. For this example, the piezometric surface will range from
an elevations of 2 to 8 feet; however, the base of the aquifer ranges in elevation
from about -3 to -7 feet in elevation. So, we 'II set the top elevation to 10 and the
bottom to -10.
Elevation
Maximum [10] - Sets maximum elevation range for the aquifer to 10
feet
Minimum [-10] - Sets minimum elevation range for the aquifer to 0
feet
Ruler Tic Increments
Major [10] - Sets major tic increments for every 10 feet on any
graphs rulers where aquifer elevations are shown. Here, as defined
we 'II only have one major tic of ten feet.
Minor [1] - Sets minor increments on any following rulers or graphs
to 1 foot. As defined here, our system will show ten minor tics.
Cursor Resolution [1] - Allows graphical entry of elevation data on
subsequent log points to be adjusted in increments of one foot.
Let's check to see that our latest set of commands works correctly by adding a log point. A "Log
Point" is a graphical method of assigning aquifer top and bottom data to the grid. You can think of
it as an actual data point derived from a soil boring or test hole, but log points can also be used to
assign grid values between actual test holes and "force" the grid to an interpretational value. The
log point icon is found on the button bar which is usually in the upper right hand corner of the work
space. It looks like:
Using the mouse, click on the log point icon. The mouse pointer will change to a kind of star-like
design. Move that pointer to any cell on the graph and click on the left mouse button once. The
star burst design will be transferred to the grid and, you will notice, that the mouse pointer has
returned to it's original arrow design. Using the mouse, place the cursor on the new log point and
double click the left mouse button(click that mouse button quickly twice). The "Log Point I"
menu will appear. Then click on "Cross Section". A graph should appear that shows the default
aquifer top and bottom (0 to 10) with a minor tick-increment every 1 foot (If you don't get this
repeat the steps above). Note the blue and red dots beside the elevation graph. By clicking and
dragging either the red or blue dot one can graphically adjust aquifer elevation top and bottom. Log
points can also be used to modify hydraulic conductivity, specific yield, effective porosity, and
longitudinal dispersivity as we'll discuss later.
Note that at least one log point must be entered to run the BIOPLUME III model. To delete
Log Point #1 you would simply click on it with the mouse then press the button on the
46
-------
key board. Maintain this Log Point for the time being. We will come back to log points later. For
now though let's save the model and continue establishing Loading Domain parameters.
Main
+ Domain
•> Loading Domain
•> Time
Maximum Time [10] - Sets maximum model run to ten years and for
all time domain graphs that follow
Ruler Tic Increments
Major [10] - Sets major tic increments on all time domain
graphs to every 10 years
Minor [1] - Sets minor tic increments on all time
domain graphs to one every year.
Now, as established above, any time domain graphs shown in the model will range from 0 to 10
years with minor tic increments set for one for each year. The pumping rate domain is the next
submenu under domain. This submenu item only has relevance if you are planning on
incorporating a pumping or injection well into your model and it can be skipped if not needed. We
will examine pumping effects in Session 5 but the topic will be skipped for now.
Main
•> Domain
•> Loading Domain
•> Pumping Rates - Skip this function for now. Keep all default settings.
Next, we continue with quickly establishing the balance (remainder) of the Loading Domain
values:
Main
-> Domain
-> Loading Domain
-> Concentrations - Skip this function for now and accept all default values
Main
•> Domain
•> Loading Domain
•> Infiltration - Sets infiltration or recharge range parameters. This is not
used in the current simulation just click on
Main
-> Domain
-> Loading Domain
-> Hydraulic Heads
Default [10] - Sets default upper limit hydraulic head to 10. This
value should be an even increment slightly higher than the highest
47
-------
piezometric head to be incorporated in the model. For the current
simulation the maximum piezometric head is 8 so we set this value to
10ft.
Maximum [10] - Sets maximum default piezometric head to 10ft
Ruler Tic Increments
Major [10] - Establishes that there will be only one major
tick increment of 10 feet for this simulation
Minor [1] - Divides the major tic increment into 10 equal
parts of one foot each.
Cursor Increment [1] - Permits graphical adjustment of
piezometric head graphs in one foot increments.
For the current session, we are not modeling chemical transport or biological reactions. Therefore,
for the time being, ignore the Chemical Species submenu of Domain but continue defining strata
properties:
Main
^Domain
•^Define Strata
Horiz. Hydraulic Conductivity (ft/sec) [3e-4] - Defines the default hydraulic
conductivity across the entire grid. If hydraulic conductivity is not modified
via log points then this is the value used for the whole grid.
Anisotropy [1] - Establishes that the aquifer is isotropic with respect to
vertical and horizontal hydraulic conductivity
Angle of Conductivity with x-axis (in degrees) [90] - For cases where the
principal axes of conductivity are at an angle with the x-y coordinate system
of the site.
Storage Coefficient [.20] - Sets storativity if model is to be run under
transient conditions. For this simulation, steady state is assumed, but it is
acceptable to put a realistic value here (the selection of steady-state or
transient conditions is specified later at the run-time options irrespective of
the S value entered here).
Effective Porosity [0.20] - Sets the porosity for the model. For unconfined
aquifers Storage Coefficient = Effective Porosity. For confined systems
Effective Porosity » Storage Coefficient.
•^Transport Properties - Sets mass transport properties and is not used in
the current session. Just click to return to the work space.
Boundary conditions are required so that the numerical model can approximate flow across the grid.
Each general cell in the grid is fundamentally defined in terms of Darcy's Law wherein water flow
across a cell boundary is directly proportional to head flux and hydraulic conductivity. The goal is
to allow head elevations for most cells in the grid to be variable based on the prevailing hydraulic
conditions and imposed stress (such as pumping). This is accomplished by simultaneously solving
a series of mathematical equations describing each cell. To provide mathematical stability;
however, a certain number of known values must be defined yielding a system of equations
48
-------
consisting of ^-equations for w-unknowns. Known values are supplied through the use of boundary
conditions.
BIOPLUME HI supports three types of boundary conditions, inactive, constant head, and constant
flow (flux). Constant flow cells produce the effects of pumping wells and may or may not be used
in a simulation. Constant head and inactive cells; however, are required for BIOPLUME m
simulations. Inactive cells are technically excluded from the active portion of the model domain.
All the cells around the perimeter of the grid must be defined as inactive for the model MOC and
BIOPLUME m. These perimeter cells are automatically set to inactive in the Platform; you can also
define those cells manually as described below.
The constant head condition fixes the water table elevation at a constant value in certain cells
throughout the simulation. For example, often the second row from top or bottom of the grid, or
the second column from left and right, are defined as constant heads^ Other cells, however, can also
be defined as constant head cells.
It is important to note that BIOPLUME m, as was the case with the predecessor program MOC,
calculates head values across the body of the grid based on the values of the constant head cells
defined within the grid. The Platform allows you to essentially draw the complete hydraulic head
surface map across the grid (initial conditions). By Kriging, interpolated head values are
established across the entire grid. The Platform then assigns the Kriged value to the defined
constant head cell automatically. So at this stage we establish boundary condition types and not the
actual head values.
To establish boundary condition types for the current session:
Main
-*Grid
-> Edit Grid
Using the mouse click on the inactive cell icon in the tool box which looks like this:
Note that the normal pointer icon changes to an x-shape. By clicking and dragging you can paint
the outermost perimeter of the grid (one row and one column at a time) with inactive cells. Next,
we need to establish constant head conditions. Click on the constant head cell icon that looks like
this:
Now the pointer icon will be shaped in a triangular pattern. Paint around the next-to-the-outside
cells. Note that there is a normal cell icon in the tool box that looks like this:
49
-------
X
•*
x
*
X
V
*
v
«
:•:
X
*
JC-
X
*
X
V
+
+
t
+
+
+
*
+
*
+
+
+
+
1
+
+
.<
••
*i*
*i+
* *
SF
*• t
* +
+ +
x X
•
+
*
t
>
-
•
*
+
+
+
+
+
+
+
X
+
+
+
+
*
+
*
:•
*
*
t
*
•
+
'
i
+
+
+
+
I
+
+
.!
'
* 1-
* +
* t-
* +
* >
X X
+
+
+
+
t
+
+
-•
*
*
+
»
•
-
•
r
+
•l-
+
*
»
+
*
•-
+
+
-f
*
•
-f
•
i
+
+
+
+
(
+
+
r
fe
4^
ix
;•*
" u
x :
.«.
X
x
x
X
X
|X
'*
i.
•- A
If you make a mistake simply click on the
proper tool box icon and correct it. When
you're finished the grid pattern should
appear like that shown on Figure 3.1. If it
doesn't keep working with it till it does then
click on to return to the main
menu. Save the file:
Main
Figure 3.1 Boundary Conditions in Test
Simulation.
3.3.2 Hydraulic Head Conditions
Now that the boundary conditions have been established head values need to be assigned to
constant head cells. The Platform ascribes actual site head elevations to the defined constant head
cells automatically. The best method to enter the piezometric surface for the model is to actually
trace the water level isopleths (head contour lines) shown on the site map image. Go to:
Main
^Loading
•> Observed Heads
On the tool box click on the Head Line tool that looks like this:
Notice that the mouse pointer is changed to a "+" sign indicating the Head Line tool is active. This
tool will allow you to draw a line on the grid and give it the proper elevation attribute. Start at one
end of the 7.0 isopleth and click once then continue to click along the trace of the contour. When
you reach the end, double click to complete the line. Every point you clicked to define the line will
be represented by a small triangle. Next, put the mouse pointer anywhere on the line and double
click to activate the "Head Contour T' menu, then:
50
-------
Main
->Head Contour 1
Piezometric Heads
Head [7] - Enters the head value of seven for the contour line just drawn
Repeat the process by drawing on top of each
contour (3 through 7 ft.) shown on the base
map. Remember, after drawing a line you
need to append the elevation attribute to the
line otherwise a default value of "0" will be
entered for that isopleth. If desired, you can
enter values for the 2 and 8 foot elevations by
interpolating the approximate location of
those contours. Note that regions of constant
(uniform) elevation can be introduced by
using the Area Head tool that looks like this:
Figure 3.2 Entering Piezometric Head Data
Using the Line Tool When finished tracing the contours your
screen should appear something like that
shown in Figure 3.2. Next, click on
to return to the Main Menu. A File (Case) Save automatically Krigs all entered data using
the default Kriging procedure. Custom (selected) Kriging can also be performed as follows:
Main
•> Select Kriging
-> Observed Heads - Selects the Kriging method
used to interpolate entered head values across the grid.
The next time the file is saved the observed head values just entered will be automatically
interpolated across the grid using the selected Kriging technique. Save now:
Main
-> File
Now, let's check the hydraulic head system we just defined:
51
-------
Main
->Grid
Observed Heads <2-D Contours>
A color shaded map will appear on
the screen showing the hydraulic
heads across the site in 2-
dimensions as shown in Figure 3.3.
Note that the value of any cell can
be viewed by clicking on that cell or
group of cells (click and drag over a
window). Contours can also be
viewed in a three dimensional
perspective by:
Main
or by:
Main
•^ Observed Heads
<3-D Contours>
->Grid
Figure 3.3 Kriged Piezometric Head Contour Map.
-> Observed Heads
<3-D Distribution
Return to the main menu by clicking on the button. Next, let's check to see if certain
distributed aquifer properties have been entered correctly:
Main
->Grid
-> Distributed Properties
•> Horiz. Hydraulic Conductivity <2-D Contours> - Click on any cell to
read a value of3e-4
Continue going through each element of the Main +Grid ^Distributive Properties submenu to
check for each of the following values:
• Storage Coefficient = 0.20
52
-------
• Effective Porosity = 0.20
• Longitudinal Dispersivity = 0.0
• Horizontal Hydraulic Conductivity = 0.0003
If there are any deviations from the values listed above return to -^Domain and track the problem
down. For the current model, these distributed properties are constant across the site. By
introducing log points, variations in these parameters can be introduced to the grid. To view these
options click on the existing log point:
-> Log Point 1
•^Cross Section
-> Stratum Properties
+Edit at Selected Elevation
Note that horizontal hydraulic conductivity, storage coefficient, effective porosity, and longitudinal
dispersivity can all be activated by clicking on the box. When only one log point is
entered these distributed properties are automatically defined as being constant across the grid.
When more than one log point is used the log point values are interpolated across the grid by
Kriging when the model is saved. Return to the Main Menu by clicking .
3.3.3 Aquifer Thickness
For this simulation we want the base of the aquifer to be approximately 10 feet below the water
table surface yielding ten feet of saturated thickness. To do this we'll need to enter a series of log
points and define the base of the aquifer. Log points can represent actual well or core hole points;
however, it is often convenient to add a sufficient number of additional "log points" to better
define the condition of interest.
First, delete the existing log point. Using the mouse pointer, click on the log point symbol then
press the key. Click on the log point tool. Again, this button looks like this:
•••••••••••-•••
Click anywhere near the end of the 7.0 ft. contour line. Double click on the log point symbol you
just inserted.
+ Log Point 1
Section
Click on the upper dot next to the footage scale and drag it down to 7.0 ft. Click on the lower dot
next to the footage scale and drag it up to the -3.0 ft level. The aquifer is now defined as being 10
foot thick at that point. Add five to ten additional log points along the 7.0 foot contour line.
Continue by adding about the same number of log points to each of the other contour lines. For
each log point drag the upper dot to the elevation of it's associated contour line and set the bottom
53
-------
dot 10 feet below the top. When you're finished the log point distribution should look something
like Figure 3.4.
Figure 3.4 Example of Log Point Distribution to Define
Top and Bottom of Aquifer.
Next, the thickness values we just entered need to be geostatistically distributed across the grid.
Let's use Select Kriging as follows:
Main
->Grid
-> Select Kriging
-> Cross-Sectional Parameters
Parameter
[Elevation] - Selects elevation as the current Kriging
parameter
Kriging Options
Main
M
J
Now check aquifer thickness by observing graphical output:
Main
Layer Thicknesses
•> 2D Contours
54
-------
After viewing aquifer thickness return to Main. Experiment a little by slightly changing one or
more log point elevations, but ultimately return to about a 10 foot thickness across the grid.
3.3.4 Steady-State Simulation
We are now finally ready for the first model run. From the Main Menu click on Simulator. Notice
that Bioplume m is "grayed out" indicating that the program is not yet available (operational). A
number of steps must be visited prior to the simulation, and the Platform assists you by not
allowing you to skip a step. Before any simulation can be run the following set of initial conditions
must be defined:
Main
^Initial Conditions
•> Simulation Period
Ending Time [10] - Sets simulation period to 10 years
Main
-> Initial Conditions
-> Starting Heads
Use Observed Values - Should be checked
Main
-> Initial Conditions
-> Starting Concentrations - Use Observed Values should be checked
Next, we move on to the simulator:
+ Simulator
•> Bioplume III
Data Set Heading
Runtime Options
•> Time Parameters
Maximum No. of Time Steps [10] - Divides model run
observation periods into ten one year increments which will
can be viewed sequentially after the simulation is concluded.
Time Increment Multiplier [0] - Not used
Initial Time Step in Seconds [0] - Not used for steady state
runs but should be set to 3,600 for transient conditions
Steady State Run ® - Should be checked indicating steady
state conditions
-> Execution Parameters
No. of Iteration Parameters [7]
55
-------
Convergence Criteria [.001]
Maximum No. of Iterations [100]
Maximum Cell Distance per Move of Particles [0.50]
Maximum No. of Particles [3000]
No. Particles per Node [16]
^Program Options
Time Step Interval for Complete Printout [1] - All other
input items in this submenu should be equal to zero.
-> Transport Subgrid - Not used. Default values only
•> Biodegradation - Not used in this simulation. Oxygen, nitrate,
ferric iron, sulfate, and carbon dioxide all should be checked
-> Save Data and Run Simulation
The translator to generate the BIOPLUME m input stream from the graphics files has now been
executed. If successful the following message will appear "Program Terminated with exit code 0".
Click .
A second window will open stating that the Input File name is BIOP3in.dat; the ASCII Output File
name is: BIOP3out.dat; and that the Graphical Output File name is BIOP3g.dat. BIOPLUME m is
now running. Notice a black cursor square in the Input/Output window. While BIOPLUME m is
running this square is solid. When the simulation is complete the cursor square will begin to blink.
When the simulation is completed close the BIOPL3.exe window:
BIOPL3.exe
- Exits the window
- Does not save the Input/Output window
The window "PBIOP3" will pop up on the screen and begin scrolling a series of variable values.
Upon completion the following message will appear "Program Terminated with exit code 0. Exit
Window?". Enter .
Model results can be graphically viewed immediately upon completion of the simulation. Recall,
that in this first test only a steady state hydraulic head surface was modeled. To see the results:
Main
-> Results
-> Hydraulic Heads <2-D Contours>
The simulated static water level should look a lot like the piezometric surface shown on the site
base map. Because we specified steady-state conditions the piezometric surface does not vary with
respect to time; however, one can view different time steps by using these icons (the opening snap
shot is always the last time step):
56
-------
For those familiar with the program output file, the ASCII output file generated by BIOPLUME m
can also be accessed and viewed within the Platform as follows:
•^ File - After viewing
Finally, graphics are good but graphics together with numerical values are even better: with the
cursor in the "spy glass" mode (data tool in the tool box, top right corner), over a
window of grid cells; , and a spread sheet table with the corresponding hydraulic head
values appears, graphics and numbers all at once as shown in Figure 3.5.
Main" fontour Range lime Layer View Annotation
100.
99.946
99.893
99.839
99.786
99.732
99.679
99.6Z5
99.572
99.518
Figure 3.5 Spread-Sheet Representation of Simulated Hydraulic Heads.
After viewing, go back to .
57
-------
3.4 Session 3: Non-Attenuated Hydrocarbon Mass Transport
3.4.1 Observed Contaminant Plume Addition
In this session we will build on the existing model by introducing a hydrocarbon source to the
system. First, the source will be modeled as a single pulse of hydrocarbon. Later, a constant
source of contamination will be examined. At this time the solute will be modeled with
retardation effects but without biological attenuation. If you do not already have case TestOl
active then open the file now (or go to the previous section and create it):
Main
*File
•> Open
First, save this simulation to a new file name:
Main
*File
- Saves current session as Test02.
Next, change the base image to show the current distribution of hydrocarbon contamination at the
site:
-> Domain
-> Base Image
* Select
-> From File
A new site map should appear on the work space showing the distribution of the hydrocarbon
plume as shown in Figure 3.6. As in Session 1, register the image to adjust it to the model
coordinate system.
First, we need to revisit the Domain menu to establish the range and default values for
hydrocarbon.
58
-------
Figure 3.6 Contaminant Distribution Image.
Main
Domain
-> Loading Domain
-> Concentrations
Default [0] - Sets default concentrations for all contaminants and
electron acceptors to 0 mg/L or ug/L. Note that the units can be
either mg/L or ug/L but that those units must be consistent for the
organic contaminant as well as for all electron acceptors. For this
example we use units of mg/L.
Minimum [0] - Sets minimum concentration to 0 mg/L
Maximum [100] - Sets maximum concentration to 100 mg/L
Ruler Tic Increments
Major [10] - Sets major tic increments on concentration
graphs in 10 mg/L units
Minor [2] - Sets minor tic in increments of 10 mg/L each
between each major tic increment
Cursor Increment [2] - Permits graphical changes in
concentrations in 2 mg/L increments
59
-------
-> Domain
-> Chemical Species
-> Contaminant
-> Reaction
- Establishes that contaminant will
interact with soil according to linear sorption.
•> Sorption Parameters
Distribution Coefficient [0.093] - Sets
approximate retardation factor (Rf) of 2 by solving
the equation Rf = (l+(bulk density * distribution
coefficient)/porosity).
-> Bulk Density
Bulk Density [2.14] - Here
we assume the aquifer is comprised of quartz sand
(2.68 g/cm3) with 20%porosity. Therefore, the bulk
density is 2.68 x (1-0.20) = 2.14 g/cm .
-> Domain
-> Define Strata
-> Transport Properties
Dispersivity [10]
Dispersivity Ratio [0.10] - Ratio of transverse to longitudinal
dispersivity is 1:10 or 0.10
Vertical Dispersivity Ratio [1] - Not used
Molecular Diffusion [0] - Not used
Bulk Density (g/cm3) - [2.14]
-> File
(For more information on these parameters see Section 4, Theoretical Development, and
Appendix B, Air Force Intrinsic Remediation Protocol Implementation).
Now, we will graphically introduce hydrocarbon concentrations to the grid with a slight
modification of the procedure used in developing the piezometric surface in Session 2. Recall,
that earlier we traced over the piezometric surface using the line tool. Here we use that technique
again; but for concentrations it is also necessary to specifically define the area outside the plume
as having zero concentration.
To begin this process:
•^Loading
-> Observed Concentrations
-> Hydrocarbon
Select the Concentration Area Tool from the tool box. It looks like this:
60
-------
Begin by clicking just outside of the 0.10 contour line. Continue tracing around that isopleth by
clicking a point about every 1A - inch. Then extend the field around the perimeter of the grid.
When you have traced nearly all the way back to your point of origin double click to exit the
Concentration Area Tool. The traced area should appear shaded with small triangles representing
line points. Next, double click anywhere on the shaded area to activate the Concentration
Contour 1 dialog box:
^Concentration Zone 1
Concentrations
Concentration [0] - Enters 0 mg/L concentration for the area
defined just outside of the plume.
Next, use the Concentration Line Tool to specify the concentrations of the plume. Recall that
the Concentration Line Tool looks like this:
Ui&Affc*
After clicking on the Concentration Line
Tool, trace over the top of the 0.1 mg/L
contour line and double click when
finished. Next, double click on the line
you've just drawn and enter a
concentration of 0.10. Repeat this
process with the 1 and 10 mg/L contour
lines. Finally, click on the concentration
line tool one more time and add a single
point in the center of the plume giving it
a value of 15 mg/L. The finished product
should look something like Figure 3.7.
Figure 3.7 Example of Establishing Hydrocarbon
Distribution Over the Grid.
Finally, you'll need to distribute the input hydrocarbon concentrations just entered to the grid:
61
-------
•* Grid
-> Select Kriging
-> Observed Concentrations
•> Hydrocarbon - Selects Kriging
Method.
-> File - Saves session, automatically Krigs data and extrapolates concentration
values to grid.
Next, we want to view and check the concentrations just entered:
•* Grid
•> Observed Concentrations
•> Hydrocarbon
•> 2-D Concentrations
A 2-dimensional shaded contour map should appear on the work space showing hydrocarbon
concentrations that looks like Figure 3.8. You can click on any cell to view the concentration at
that point. Note that in some areas several contour lines passed through a single cell; however,
the Platform has assigned a single concentration value to that cell from the latest contour line. If
the concentrations are not what you want, you can always modify or delete the old contours.
Figure 3.8 Example Kriged Hydrocarbon
Distribution.
It is a good idea to check on the
condition of the piezometric head
distribution under Main -> Grid ->
Observed Head •> 2D Contours. If the
head distribution looks correct then
return to the Main Menu. Otherwise go
through the Kriging process again for
the piezometric surface under Main •>
Grid -> Select Kriging -> Observed
Heads. When finished return to the
Main Menu.
62
-------
3.4.2 Transport Execution and Results
You are now ready to run the mass transport simulation. As above, step through the Initial
Conditions menus:
Main
^Initial Conditions
•> Simulation Period
Ending Time [10] - Sets simulation period to 10 years
Main
-> Initial Conditions
-> Starting Heads
Use Observed Values - Should be checked
Main
-> Initial Conditions
-> Starting Concentrations - Use Observed Values should be checked
Finally, we need to check the simulation run time parameters and run the model:
•> Simulator
•> Bioplume III
Data Set Heading [Static Piezometric Surface]
Runtime Options
•> Time Parameters
Maximum No. of Time Steps [10]
Time Increment Multiplier [0]
Initial Time Step in Seconds [0]
Steady State Run ®^]
•> Execution Parameters
No. of Iteration Parameters [7]
Convergence Criteria [.001]
Maximum No. of Iterations [100]
Maximum Cell Distance per Move of Particles [0.50]
Maximum No. of Particles [3000]
No. Particles per Node [15]
-> Program Options
Time Step Interval for Complete Printout [1] - All other
input items in this submenu should be equal to zero.
-> Transport Subgrid - Not used. Default values only
•> Biodegradation - Not used in this simulation. Oxygen, nitrate,
ferric iron, sulfate, and carbon dioxide all should be checked
63
-------
-> Save Data and Run Simulation
The WBIOP3 window will pop up to translate the graphics files into BIOPLUME in input
stream. When the translation is complete the following message will appear "Program
Terminated with exit code 0". Click . The BIOPL3.EXE window will pop up indicating
that BIOPLUME HI is running. When the simulation is complete the black cursor square will
begin to blink. When the simulation is completed close the BIOPL3.exe window (and the
input/output window):
BIOPL3.exe
- Exits the window
- Does not save the Input/Output window
The window "PBIOP3" will pop up on the screen and begin scrolling a series of variable values.
Upon completion the following message will appear "Program Terminated with exit code 0. Exit
Window?." Enter . The simulation is complete and model results can now be viewed
graphically:
Main
-> Results
-> Concentrations
Hydrocarbon <2-D Contours>
The initial view will be the predicted
hydrocarbon concentration at 10 years
which should look something like Figure
3.9. Various time steps can be viewed by
clicking on the following buttons:
After viewing .
Figure 3.9 Simulated Hydrocarbon Plume in 10
Years.
64
-------
3.4.3 Constant Source Addition
Now we will set up the model to simulate a constant contaminant source. If you do not already
have Test02 active then open that file now:
Main
->File
-> Open
First, save this simulation to a new file name, as TestOS.
Main
->File
->Save As - Saves current session as Test03.
Next, we need to erase (delete) all existing hydrocarbon concentration values:
Main
-> Loading
-> Observed Concentrations
-> Hydrocarbon
Click on the shaded area concentration then press . Click on each of the line
concentration values and also press the delete button. This will remove all observed
concentration values from the grid. Return to the Main menu. Click on the source tool that
looks like this:
The cursor will change to a "+" symbol. Trace around the 10 mg/L concentration contour. When
you've finished double click to complete the polygon. Next, double click again on the shaded
area to bring up the Source 1 menu. Then:
Sourcel
+ Loading
•> Concentration
[Hydrocarbon] - Defines the current source as being for hydrocarbon.
Sources can be applied for each electron acceptor independently under
this option.
•> Set Value - A graph will appear. Click on the red dot and drag it
vertically up to the 10 mg/L mark. This sets the constant source of
hydrocarbon at 10 mg/L.
65
-------
Main
File
You're ready again to run the
simulation. Repeat the steps listed in
Section 3.4.2 above. This time the
results for a 10 year migration period
look like Figure 3.10.
Figure 3.10 Simulated Hydrocarbon Plume as
Constant Source.
66
-------
3.5 Session 4: Simulated Microbial Attenuation
3.5.1 Addition of Electron Acceptors
We have previously developed a model simulating flow and non-attenuated hydrocarbon mass
transport. Now we will take the process one step further by introducing biological degradation
that will be stoichiometrically balanced against Oxygen and Sulfate reduction. First, create a
new file:
Main
->File
-> Open
Save this simulation to a new file name:
Main
->File
->SaveAs
For this simulation we assume that the original concentration of dissolved oxygen and sulfate in
the ground water is 8 and 80 mg/L respectively. Additionally, here we assume that the general
chemical equation for the hydrocarbon solute is CyHg (toluene). Therefore, the stoichiometric
expression for complete mineralization of toluene with oxygen is (for additional information see
Section 4, "Theoretical Development," and Appendix B, "Air Force Intrinsic Remediation
Protocol"):
C7H8 + 902 v 7C02 + 4H20
Here, 9 molar volumes of molecular oxygen are required to convert 1 mole of organic to carbon
dioxide and water. BIOPLUME HI reactions are assumed to be on a mass per mass basis. By
conversion we see that:
C7H8 = 92 g/mole so 92 g/mole * 1 mole = 92 g
02 = 32 g/mole so 32 g/mole * 9 moles = 288 g
Therefore, the mass/mass ratio of oxygen to organic is 288 g/ 92 g = 3.1. In this case, 3.1 mg/L
oxygen is needed to oxidize 1 mg/L hydrocarbon.
The stoichiometric expression for sulfate reduction, assuming hydrogen sulfide (FL:S)
production, is:
67
-------
C7H8 + 3.67S04 v 7C02 + 3.67H2S + 0.33H20
Here, 3.67 moles SC>4 oxidize 1 mole CyHg. Converting to a mass/mass basis yields 3.8 mg/L
sulfate required to oxidize 1 mg/L toluene. Other assumptions used here are that:
• Oxygen is used before sulfate;
• Oxygen and sulfate concentrations are initially constant across the entire grid;
• Oxygen reacts instantaneously with hydrocarbon according to the stoichiometry
expressed above;
• Sulfate reacts with hydrocarbon according to a first order rate constant of 0.001.
As with the sessions above, to begin the model we need to make some changes under the
Domain directory related to electron acceptor reaction parameters:
-> Domain
-> Chemical Species
•> Electron Acceptors
•> Oxygen
-> Reaction
•> Interaction
Stoichiometric Ratio of Electron Acceptor to
Hydrocarbon [3.1]
Electron Acceptor Threshold [.50] - Sets minimum
level that aerobic bacteria can remove oxygen to
0.50 mg/L
First Order Decay Rate [0]
Maximum Hydrocarbon Utilization [0]
Hydrocarbon Half-Saturation [0]
Electron Acceptor Half Saturation [0]
Microbial Concentration [0]
Retardation Factor for Microorganisms [0]
* Sulfate
•> Reaction
-> Interaction
Stoichiometric Ratio of Electron Acceptor to
Hydrocarbon [3.82]
Electron Acceptor Threshold [0]
First Order Decay Rate [.001]
Maximum Hydrocarbon Utilization [0]
68
-------
Hydrocarbon Half-Saturation [0]
Electron Acceptor Half Saturation [0]
Microbial Concentration [0]
Retardation Factor for Microorganisms [0]
Next, we need to define the concentrations of oxygen and sulfate across the grid. This is done in
two phases. First, the current concentration of oxygen and sulfate must be entered:
-> Loading
-> Observed Concentrations
-> Oxygen
Select the Area Concentration Tool from the Tool Box:
Beginning in one corner of the grid draw a square encompassing the entire grid area. When
complete, double click to end. Place the cursor on the shaded square and double click again.
Then:
Concentration Zone 1
Concentration
Concentration [8.0] - Enters a value of 8.0 mg/Lfor the concentration of oxygen
across the entire grid.
Repeat the process for sulfate:
-> Loading
-> Observed Concentrations
* Sulfate
As with oxygen use the Area Concentration Tool in the Tool Box to enter the concentration
attribute of 80.0 mg/L for sulfate then return to the Main menu.
The concentrations of oxygen and sulfate we just entered are mobile solutes. Without the
addition of a source, oxygen and sulfate would "migrate" off the grid so that some areas near the
grid boundary would have little or no electron acceptor concentrations over time. To eliminate
that problem a source of electron acceptors must be added. In the Main menu click on the
Recharge tool in the Tool Box:
69
-------
The cursor will change to a "+" symbol. Draw a rectangular shape covering the top row of
constant head cells. When finished, double click to end. Place the cursor on the shaded area and
double click again to activate Recharge Region 1. From that point:
Recharge Region 1
Loading
Concentration [Oxygen] - Selects oxygen as the source parameter
->Set Values
Concentration [8.0] - Using the cursor you can click on the red
dot and move it vertically up to the 8.0 mg/L level on the graph or
you can directly enter 8.0 mg/L .
Concentration [Sulfate] - Selects sulfate as the source parameter.
•*Set Values
Concentration [80.0] - Using the cursor you can click on the red
dot and move it vertically up to the 80.0 mg/L level on the graph
also.
Repeat this process by introducing recharge areas to the constant head cells on the bottom, left,
and right grid cells. Define the concentration for each area as 8.0 mg/L oxygen and 80 mg/L
sulfate. When finished the grid should look like Figure 3.11.
Figure 3.11 Addition of Electron Acceptor Source Areas.
70
-------
Next, the Kriging Method needs to be selected for each electron acceptor:
•* Grid
•> Select Kriging
•> Observed Concentrations
-> Oxygen
•* Grid
-> Select Kriging
-> Observed Concentrations
->Sulfate
•> File - Saves and appends values entered for electron acceptor concentration
to the grid.
As above, we want to check the concentrations just entered to make sure there are correct:
•* Grid
-> Observed Concentrations
-> Oxygen
-> 2-D Concentrations
Click anywhere on the grid to see a cell concentration of 8 mg/L. Repeat the process to
check sulfate:
•* Grid
•> Observed Concentrations
-> Sulfate
•> 2-D Concentrations
Click anywhere on the grid to see a cell concentration of 80 mg/L. Also, it's a good
idea to check the hydrocarbon concentration and piezometric head distribution at this time under
Main •> Grid •> Observed Concentrations. If these appear correct you're ready for the next
step. If not, then re-Krig, save, and check the distribution again until correct.
3.5.2 Model Execution and Results
We are now ready to execute the simulation. As above, step through the Initial Conditions
submenus in preparation for executing the simulation:
•^Initial Conditions
-> Simulation Period
-> Initial Conditions
-> Starting Heads
•> Initial Conditions
71
-------
•> Starting Concentrations
-------
As discussed above, step through the
WBIOP3, BIOPL3.EXE, and BIOPL3.exe
windows. Model results can be viewed
graphically under the Results menu. You'll
note that the plume concentration for
hydrocarbon is greatly diminished over the
results obtained in Test02 (Figure 3.12).
Additionally, you should see electron acceptor
concentration sags for both oxygen and sulfate
similar to those shown on Figures 3.13 and
3.14.
Figure 3.12 Simulated Hydrocarbon Plume
at 10 Years Assuming Microbial
Attenuation.
Figure 3.13 Oxygen Distribution Showing
Reaction Sag.
Figure 3.14 Sulfate Distribution Showing
Reaction Sag.
73
-------
3.6 Session 5: Special Features
A number of special features are available with the Graphical Platform that can be used to adapt
a model to more closely simulate field conditions. In the Main menu, the Tool Box contains
buttons that simplify the addition of lakes, rivers, and drains. The addition of these attributes to a
simulation are discussed below. First, create a new file using the simulation developed in
Session 2:
Main
*File
-> Open
Save this simulation to a new file name:
Main
*File
->SaveAs
3.6.1 Adding Wells
In this section we will examine the addition of a well to the simulation. It is necessary to define
the pumping domain:
Main
-> Domain
-> Loading Domain
-> Pumping Rates
Default [0] - Sets default pumping rate to 0 ft3/sec. Note that all
positive pumping rates are for pumping wells and negative rate
numbers are for injection wells
Minimum [-0.1] - Sets default lower pumping rate on all well
graphs to -0.1 ft3/sec.
Maximum [0.1] - Sets default maximum pumping rate on all well
graphs to 0.1 ft3/sec (44.8 gpm).
Ruler Tick Increments
Major [. 1] - Sets one major tic increment at. 1 ft3/sec.
Minor [.02] - Divides the one major tick increment into 10
minor units for all pumping graphs
Cursor Increments [.02] - Permits graphical adjustment of
pumping rates for all wells in increments of .02 ftVsec.
-> File
On the button bar find and click on the well icon button that looks like this:
74
-------
Note that the normal mouse pointer changes to a circle shape. Point the cell on the grid
containing the water well at coordinates (1,510, 1,010) and click. The circle shape will be
transferred to the grid and now represents a new injection or withdraw well. Note that the mouse
cursor returns back to its normal configuration. Next, place the cursor on the well and double
click which will open the Well 1 menu.
For this example we'll set up two separate stress periods of five years each. For the first five
year period the well will not be pumped. For years six through ten the well will be produced at
0.02 ft3/sec. First we need to define the stress periods.
Wettl
-> Time Steps
•^ Add Timestep [5] - Defines a new stress period beginning after
the fifth year.
Note: Although many stress periods can be added all must be of equal duration. For example,
in the current simulation we could define five stress periods; however, each period would need to
be set at two years. Considerable pumping versatility can be achieved by defining many stress
periods and setting the same discharge rate for several sequential periods.
Next, the discharge rate for each time step must be specified. Continuing in the Well 1 menu,
click on Pumping Rates. It is possible to manually enter the pumping rates by clicking on the
Time period desired then entering the discharge value under Rate These values can also be
easily entered graphically. On the displayed graph click on the small dot at time zero and hold
down on the mouse button. While holding the button drag the dot to the zero pumping rate
value. Next, click on the dot at time five and drag it up to 0.02. Click on to return
to the Main menu.
Though not needed in the current example, note that in the Well 1 menu concentrations of
contaminants and electron acceptors could also be introduced to the simulation by direct
injection. Injection, however, requires a negative pumping rate.
We are ready to run the simulation again. First, update the current simulation by saving the file:
Main
As always, it's a good idea to check the condition of the head and hydrocarbon distribution under
Main -> Grid -> Observed Heads or Observed Concentrations. If all is correct proceed with
the simulation:
75
-------
Main
•^Initial Conditions
-> Simulation Period
Main
•> Initial Conditions
•* Starting Heads
Main
-> Initial Conditions
•> Starting Concentrations
Main
•} Simulator
•> Bioplume III
Data Set Heading
^Runtime Options
•> Time Parameters
Maximum No. of Time Steps [5] - Sets 5 time steps for
each of the two stress periods or one time step per year for
the 10 year duration of the simulation.
Pumping Period in Years [0]
Time Increment Multiplier [0]
Initial Time Step in Seconds [0]
-> Execution Parameters
No. of Iteration Parameters [7]
Convergence Criteria [.001]
Maximum No. of Iterations [200]
Maximum Cell Distance per Move of Particles [0.50]
Maximum No. of Particles [3000]
No. Particles per Node [15]
•> Program Options
Time Step Interval for Complete Printout [1]
-> Transport Subgrid
•> Biodegradation
Oxygen
Nitrate
Ferric Iron
Sulfate
Carbon Dioxide
-> Save Data and Run Simulation
Follow the steps outlined above to run through the model. After processing is complete, view the
dynamic piezometric head under -^Results as described above. The piezometric surface for the
first five years should look like Figure 3.3 in Session 1. A cone of depression like that shown in
Figure 3.15 should be evident for the last five year time period. Plume distribution should also
manifest pumping effects:
76
-------
Main
-> Results
Concentrations ->
-> Hydrocarbon
-> 2 D Contours
•> Contour Range - Sets the
range of hydrocarbon concentration contours to span the
current time step.
After viewing plume distribution return to the Main menu.
Figure 3.15 Head Distribution Under Pumping
Conditions.
3.6.2 Rivers, Drains, and Lakes
Rivers, drains, and lakes can easily be added to a simulation using special buttons located in the
Tool Box These features are simulated by specifying designated cells to have constant head
values and high hydraulic conductivity. Using special tools makes this application easier. For
this example a river and lake will be added to the model. As in the examples above, start a new
simulation based on Test02:
Main
Open
77
-------
Main
->File
->SaveAs
Next, change the base image to Test04.bmp following the steps listed in Session 1. The new
image should show the addition of a stream and lake. Click on the River Tool in the Tool Box
that looks like this:
The cursor should change to a "+" shape. Click on one end of Little Creek and continue clicking
along the trace of the stream. When finished double-click exit the drawing mode. Place the
cursor on the line you just drew and double click to activate the River 1 menu:
River 1
-> River Bed
Bed Level [0] - Sets base of river at 0 but also diminishes the thickness of the
aquifer concurrently. Therefore, value must be above the base of the aquifer.
Thickness [1]
Hydraulic Conductivity [.1]
-> Loading
-> Levels
Surf ace Levels [3] - Sets hydraulic elevation to 3. This item could also be
set graphically.
A lake can be added in a simmilar manner. Click on the Lake Tool in the Tool Box that looks
like this:
Again, the cursor should change to a "+" shape. Using the cursor, draw around Tiny Lake.
When finished double click to end the drawing mode. Place the cursor on the lake area and
double click again to activate the Lake 1 menu. Then:
Lake 1
* Lake Bed
Bed Level [4] -Must be less than the thickness of the aquifer.
Thickness [10]
Hydraulic Conductivity [.1]
+ Loading
•> Levels
78
-------
Surface Level [6.5] - Sets the hydraulic head to 6.5 feet defining the water
surface of the lake.
We are now ready to run another simulation. Step through the Initial Conditions and Simulator
menus as described in sections above. When the simulation is completed view the hydraulic
heads under Results. A constant head value of 6.5 will be found in the lake area. Head contours
will converge on the stream. An examination of the plume distribution with respect to time will
show hydrocarbon intersecting and following the path of the creek.
3.7 Session 6: Video Animation
With the Windows 95 version of the Platform video animation of plume migration is possible.
Through AVI animation, sequential time steps can be viewed in rapid succession. To
demonstrate this feature reopen Test04:
Main
-*File
•> Open [Test04]
Go to Main ^Results. If the Results submenu is inactive (grayed out) then:
Main
-*File
- Restores prior simulation results.
If the Results submenu is still inactive you'll need to rerun the simulation. Now, assuming the
Results submenu is now active view the hydrocarbon concentration plume:
Main
-> Results
-> Concentrations
-> Hydrocarbon
-> 2 D Contours
-> Contour Range
Next, using the Timestep tool click back to view Time 0. Recall that the Timestep tool looks
like this:
79
-------
Next, capture this image by clicking on the Save Image tool in the Tool Box. The save image
tool looks like this:
The cursor will change to a large square shape. Identify a rectangular area on the grid of which
you want to create an image. This area need not be the entire grid; however, the same area must
be identified on successive time steps. So, whatever area you select to work with note the
coordinates. For this session place the cursor over the cell in position (600, 300). Click and hold
the mouse button. Drag down and to the right to the cell located at (1,800, 1,400) then release
the mouse button. A shaded rectangle will appear over the blocked area and the Save Bitmap
File As menu will pop up:
Save Bitmap File As
-> File Name [A.bmp] - Saves the designated grid area as a windows bit map
under the file name "userOO.bmp".
Next, using the Timestep tool go to Time 1. Click on the Save Image tool and block out the
same grid range as before. This time save the image as "B.bmp". Repeat this process for each
successive time step saving the images as "A.bmp" through "J.bmp". When finished return to
the Main menu.
We are now ready to tie the sequential time steps together in a single video clip:
Main
-> Results
-> A vi A nimation
EIS Video Animation
-*File
-> Select Bitmap Files to Compile
Select All Bitmap Files to Compile
•> Folders - Go to
the designated subdirectory under the root
of C: assuming \EISBIOP\Bioplume III was
installed on the C: directory.
While holding down on the "Ctrl" button on your keyboard use the mouse and click on the file
names "A.bmp" through "J.bmp". When finished click on . The Save AVI Movie As
menu will popup:
Save A VI Movie As
File Name [Testavi]
80
-------
The following message will appear "Done building AVI!" . To view the animation clip:
EIS\GWM Video Animation
-> Open A VI [Testavi]
The image of Timestep 0 will appear. To activate the animation sequence use the mouse to click
on the "scissors" symbol in the lower left hand corner of the image.
81
-------
4.0 BIOPLUME III THEORETICAL DEVELOPMENT
4.1 Overview
The BIOPLUME III model simulates aerobic and anaerobic biodegradation processes using
oxygen, nitrate, iron (III), sulfate, and carbon dioxide as electron acceptors. In addition, the
model simulates advection, dispersion, sorption, and ion exchange. The model solves the
transport equation six times to determine the fate and transport of the hydrocarbons and the
electron acceptors/reaction by-products. For example, in the case of iron (III), the model
simulates the production and transport of iron (II) or ferrous iron. The following two sections
describe in more detail the conceptual model for biodegradation used in BIOPLUME III and
provides a summary of the limitations of the program. For more information on the processes of
advection, dispersion, and sorption, the user is encouraged to consult Appendix A and Konikow
and Bredehoeft (1978 & 1989).
4.1.2 Conceptual Model for Biodegradation
Recent research suggests that hydrocarbons are degraded both aerobically and anaerobically in
subsurface environments. The main electron acceptors include oxygen for aerobic biodegradation
and nitrate, iron (III), sulfate, and carbon dioxide for anaerobic biodegradation. Manganese has
also been identified as an anaerobic electron acceptor; however, manganese has not been
incorporated into the current version of BIOPLUME III.
The conceptual model used in BIOPLUME III to simulate these biodegradation processes tracks
six plumes simultaneously: hydrocarbon, oxygen, nitrate, iron (II) , sulfate, and carbon dioxide.
Iron (III) is input as a concentration matrix of ferric iron in the formation. Once ferric iron is used
for biodegradation, BIOPLUME III simulates the production and transport of ferrous iron.
Biodegradation occurs sequentially in the following order:
Oxygen -» Nitrate -» Iron (III) -» Sulfate -» Carbon Dioxide
The biodegradation of hydrocarbon in a given location using nitrate, for example, can only occur if
oxygen has been depleted to its threshold concentration at that location.
Three different kinetic expressions can be utilized for the biodegradation reaction for each of the
electron acceptors:
1. First-order decay
2. Instantaneous reaction
3. Monod kinetics
These kinetic expressions are discussed in more detail in Section 4.2.3. The first-order decay
model implemented in BIOPLUME III for any of the electron acceptors is limited by the
82
-------
availability of the electron acceptor in question. In other words, the model allows the first-order
reaction to take place up to the point that the electron acceptor concentration available in the
aquifer has been depleted.
The Monod kinetic model used in BIOPLUME III assumes a constant microbial population for
each of the aerobic and anaerobic reactions and does not simulate the growth, transport and decay
of the microbial population in the subsurface.
4.1.3 BIOPLUME III Applicability and Limitations
The BIOPLUME III model has been mainly developed to simulate the natural attenuation of
hydrocarbons using oxygen, nitrate, iron (III), sulfate, and carbon dioxide as electron acceptors
for biodegradation. BIOPLUME III is generally used to answer a number of questions regarding
natural attenuation:
1. How long will the plume extend if no engineered/source controls are implemented?
2. How long will the plume persist until natural attenuation processes completely dissipate
the contaminants?
3. How long will the plume extend or persist if some engineered controls or source reduction
measures are undertaken (for example, free phase removal or residual soil contamination
removal)?
The model can also be used to simulate bioremediation of hydrocarbons in ground water by
injecting electron acceptors (except for iron(III)) and can also be used to simulate air sparging for
low injection air flow rates. Finally, the model can be used to simulate advection, dispersion, and
sorption without including biodegradation.
As with any model, there are limitations to the use of BIOPLUME III. The assumptions used in
the USGS MOC code include:
1. Darcy's law is valid and hydraulic-head gradients are the only driving mechanism for flow.
2. The porosity and hydraulic conductivity of the aquifer are constant with time, and
porosity is uniform in space.
3. Gradients of fluid density, viscosity, and temperature do not affect the velocity
distribution.
4. No chemical reactions occur that affect the fluid properties, or the aquifer properties.
5. Ionic and molecular diffusion are negligible contributors to the total dispersive flux.
6. Vertical variations in head and concentration are negligible.
83
-------
7. The aquifer is homogeneous and isotropic with respect to the coefficients of longitudinal
and transverse dispersivity.
The limitations imposed by the biodegradation expressions incorporated in BIOPLUME III
include:
1. The model does not account for selective or competitive biodegradation of the
hydrocarbons. This means that hydrocarbons are generally simulated as a lumped organic
which represents the sum of benzene, toluene, ethyl benzene or xylene. If a single
component is to be simulated, the user would have to determine how much electron
acceptor would be available for the component in question.
2. The conceptual model for biodegradation used in BIOPLUME III is a simplification of
the complex biologically mediated redox reactions that occur in the subsurface.
4.1.4 Comparison of BIOPLUME III to Analytical Models
The testing program for BIOPLUME III was based on: i) previous modeling projects performed
by several different researchers, ii) articles in the related literature, and iii) testing performed
directly by the project team.
The BIOPLUME III model is based on the Method of Characteristics (MOC) code (Konikow
and Bredehoeft, 1978 and 1989), which was first modified by Borden to develop the
BIOPLUME model (Borden and Bedient, 1986a and 1986b). Rifai then modified the
BIOPLUME model to develop the BIOPLUME II model (Rifai et al., 1987). BIOPLUME II
formed the basis for the BIOPLUME III model, in part using source code from a research model
developed by Rifai (Rifai and Bedient, 1990).
The following comparison/checking operations have been conducted throughout the development
of the BIOPLUME models:
1. The MOC model was compared to an analytical solution (Konikow and Bredehoeft,
1978). This work is summarized in Appendix A.6 of the BIOPLUME III User's
Manual.
2. BIOPLUME was successfully calibrated to a field site (Borden and Bedient, 1986b).
3. BIOPLUME II was successfully calibrated to a field site (Rifai et al., 1988).
4. An analytical solution supplemented by a superposition technique for the instantaneous
biodegradation reaction was compared against BIOPLUME II by Connor et al. (1994),
who concluded that "incorporation of the simple oxygen-superposition function into the
Domenico model provides a steady-state plume prediction in close agreement with the
BIOPLUME II model." In addition, Ollila (1996) performed a similar comparison and
determined that the analytical solution was in "close agreement with BIOPLUME II."
84
-------
6.
4.2
BIOPLUME III was tested by comparing i) the results from BIOPLUME III simulations
using a single electron acceptor against ii) BIOPLUME II runs with the oxygen
concentrations adjusted to reflect the different utilization factors for each non-oxygen
electron acceptor. These comparisons were performed for oxygen, nitrate, sulfate, and
methane (because the iron reaction was based on dissolution from a solid, it could not be
compared in the same way). The BIOPLUME III Monod biodegradation modules,
originally developed by Rifai for a research code (Rifai and Bedient, 1990), were tested by
reducing the half-saturation constants and increasing the maximum utilization rates until
the results approached the instantaneous reaction solutions. There are no analytical
solutions which include Monod kinetics that can be used for model testing.
As part of this project, the model was successfully calibrated to 8 field sites by the
project team. After calibration, the simulated BTEX and electron acceptor/by-product
concentrations matched observed conditions in the field.
Mathematical Model
4.2.1 Numerical Simulation of Oxygen Limited Biodegradation in
BIOPLUME II
4.2.1.1 Equation Formulation. Borden and Bedient (1986) simulated the growth of
microorganisms and removal of hydrocarbon and oxygen using a modification of the Monod
function where:
dH
dt
dO
dt
dMt
~dT
where
H
O
Mt
k
Y
(4.1)
(4.2)
(4.3)
hydrocarbon concentration
oxygen concentration
total microbial concentration
maximum hydrocarbon utilization rate per unit mass microorganisms
microbial yield coefficient (g cells/g hydrocarbon)
hydrocarbon half-saturation constant
oxygen half-saturation constant
85
-------
kc = first-order decay rate of natural organic carbon
C = natural organic carbon concentration
b = microbial decay rate
F = ratio of oxygen to hydrocarbon consumed
Equations 4.1 through 4.3 were combined with the advection/dispersion equation for a solute
undergoing linear instantaneous adsorption to result in:
V • (DVH - vH)/Rh - Mt • k • br-TT hr—n (4.4)
V • (DVO - vO) - Mt • k • F • y ^TT v ^o (4.5)
(Kh + Hj^K0 + Oj
where
D = dispersion tensor
v = ground water velocity vector
R^ = retardation factor for hydrocarbon
The movement of naturally occurring microorganisms will be limited by the tendency of the
organisms to grow as microcolonies attached to the formation. Borden and Bedient (1986)
assumed that the transfer of microorganisms between the solid surface and ground water will be
rapid and will follow a linear relationship with total concentration, thus allowing them to simulate
the transport of microorganisms using a simple retardation factor approach:
V • (DVMs-vMs)/R
s-sm
H V 0 "1 kcYC .
M -k-Y'l — — — II — - — 1+ — -- s (46)
s IC+H K +0 v '
h A o
where
Ms = concentration of microbes in solution
Ma = concentration of microbes attached to aquifer
Km = ratio of microbes attached to microbes in solution
Rm = microbial retardation factor
Ma = Km «MS
Mt = Ms +Ma =(l+Km)*Ms =Rm -Ms
86
-------
Borden and Bedient (1986) conducted one-dimensional simulations with equations (4.4) - (4.6)
and determined that there are three general regions where different processes control the rate and
extent of degradation: near the contaminant source, in the heart of the plume and at the leading
edge of the plume. Biodegradation rates will be very high near the source and will result in
depleted oxygen levels. Biodegradation in the heart of the plume will be limited by the
availability of oxygen. The primary mass transfer processes include horizontal mixing with
oxygenated formation water, advective fluxes of oxygen and vertical exchange with the
unsaturated zone. The limited oxygen supply to the heart of the plume will result in a region of
reduced oxygen and hydrocarbon concentrations. At the leading edge of the plume, oxygen is
present in excess and hydrocarbons will be absent or present in trace quantities.
Sensitivity analyses with the one-dimensional model indicated that the microbial parameters had
little or no effect on the hydrocarbon concentration in the body of the plume and on the time to
hydrocarbon breakthrough. This led Borden and Bedient (1986) to assume that the consumption
of hydrocarbon and oxygen might be approximated as an instantaneous reaction between oxygen
and hydrocarbon:
H(t+l) = H(t)-O(t)/F O(t+l) = 0 where H(t)>O(t)/F (4.7)
O(t+l) = O(t)-H(t)»F H(t+l) = 0 where O(t)>H(t)»F (4.8)
where H(t), H(t+l), O(t), O(t+l) are the hydrocarbon and oxygen concentrations at time t and
t+1.
Borden and Bedient (1986) concluded that the instantaneous reaction assumption is a close
approximation to equations (4.4) through (4.6). Their simulations indicate that the most
significant errors using this assumption occur in the region near the source area especially when
ground water velocities are very high or for poorly degradable hydrocarbons.
Borden and Bedient (1986) used the instantaneous reaction assumption to simplify the system of
equations (4.4) - (4.6) to:
dH 1 / d2H d2H dH\
~dT = R~ Dl—2~+Dt—2~~v~dx~ ~b (4'9)
\ J I
dO d2O d2O dO
3T = Dl—9~+Dt—9~-yl~-6F (4-10)
C/t gXZ QyZ. C/X
where
6 = min (H, O/F)
D{ = transverse dispersion coefficient
«t = transverse dispersivity
D^ = «{ • v
y = coordinate orthogonal to the flow
87
-------
Two-dimensional simulations of equations (4.9) and (4.10) indicate that biodegrading plumes are
generally less laterally spread than their non-degrading counterparts. Simulations also indicated
that transverse mixing is the major source of oxygen to the plume.
Borden (1986) examined the vertical exchange of oxygen with the unsaturated zone. His
simulations indicated that the effects of gas exchange with the unsaturated zone may be
approximated as a first-order decay in space and time of hydrocarbon concentrations. Borden
(1986) developed a regression function at the United Creosoting Company site to determine the
reaeration first-order decay coefficient:
r\ rrn / -10.5 B \
K' = 2611 Dv 0-79 exp (-^-^ (4.11)
where
B = saturated thickness
Dv = vertical dispersion coefficient
4.2.1.2 Development of the BIOPLUME II Model. Rifai et al (1988) incorporated the
conclusions developed by Borden and Bedient (1986) into the USGS two-dimensional solute
transport model more commonly referred as the Method of Characteristics (MOC) model. The
MOC model was modified from a single particle mover to a dual particle mover model to simulate
the transport of hydrocarbon and oxygen. The system of transport equations solved in
BIOPLUME II is given by:
dHb 1 / d / 5H\ d \ H'W
— = R a bDiJ a - a (bHVi) ' ~ (4' 12)
dOb / d I dO\ d \ O'W
— — = — bD«T— -T— (bOVO - - (4.13)
dt Idx I l) dx I dxv iyl n v '
where
H = concentration of hydrocarbon
O = concentration of oxygen
H' = concentration of hydrocarbon in source or sink fluid
O' = concentration of oxygen in source or sink fluid
n = effective porosity
b = saturated thickness
W = volume flux per unit area
Vj = seepage velocity in the direction of xj
Rh = retardation factor for hydrocarbon
DJ = coefficient of hydrodynamic dispersion
-------
xj /xj = cartesian coordinates
t = time
The hydrocarbon and oxygen plumes are combined using the principle of superposition and
equations (4.7) and (4.8). The principle of superposition is best portrayed in Figure 4.1. It is
noted from Figure 4.1 that wherever the hydrocarbon is present in relatively high concentrations,
oxygen is absent. The oxygen plume forms an envelope for the hydrocarbon plume with oxygen
concentrations gradually increasing to initial background levels as one moves away from the
centerline of the hydrocarbon plume. In profile view, the hydrocarbon plume is less spread out,
and has lower concentrations than the nonbiodegraded plume.
4.2.2 BIOPLUME III Equation Formulation
Much like the approach used in developing BIOPLUME II, the 1989 version of the MOC model
was modified to become a six-component particle mover model to simulate the transport of
hydrocarbon, oxygen, nitrate, iron(II), sulfate and carbon dioxide. Since the biodegradation of
hydrocarbon uses iron (III) as an electron acceptor, iron (III) concentrations are simulated as an
initial concentration of ferric iron that is available in each cell. Once the iron (III) is consumed,
hydrocarbon concentrations are reduced and ferrous iron is produced. The resulting ferrous iron
is then transported in the aquifer. The BIOPLUME III equations include:
aHb 1 / a / 3H\ a \ H'W
(4.14)
aob
at
aNb
at
aFb
acb
at ~ 13xi ["^Jdxil axiw^"iy| n
where
N = concentration of nitrate
N' = concentration of nitrate in source or sink fluid
89
-------
Zone of reduced
hydrocarbon con-
centrations
Zone of treatment
H
A
Without oxygen
xygen
A1
B
Zone of oxygen
depletion
B1
[Background DO|
DO
Zone of reduced
oxygen concentration
B
Background DO
tfai
Depleted
oxygen
B1
Source: Bedient, Rifai, and Newell (1994)
Figure 4.1. Principle of Superposition for Combining the Hydrocarbon
and Oxygen Plumes in BIOPLUME II
90
-------
F = concentration of iron (II)
F = concentration of iron (II) in source or sink fluid
S = concentration of sulfate
S' = concentration of sulfate in source or sink fluid
C = concentration of carbon dioxide
C = concentration of carbon dioxide in source or sink fluid
All other parameters as defined previously.
The biodegradation of hydrocarbon using the aerobic and anaerobic electron acceptors is
simulated using the principle of superposition and the following equations:
H(t+l) = H(t)-R -R -R -R -R (4.20)
HO HN HFe HS HC '
O(t+l) = O(t)-R (4.21)
OH
N(t+l) = N(t)-RNH (4.22)
Fe(t+l) = Fe(t)-RpeH (4.23)
= RFeH 6 (4.24)
RsH (4.25)
C(t+l) = C(t)-R (4.26)
CH
where RTT^ , RTT^T , RTTT, , R^ , R^ are the hydrocarbon concentration losses due to
rlO rlN rlre rlS rlC
biodegradation using oxygen, nitrate, ferric iron, sulfate and carbon dioxide as electron
acceptors, respectively. The terms RQH , RNH , RpeH , RSH , RCH are the corresponding
concentration losses in the electron acceptors. These reaction terms are computed using
one of the three biodegradation expressions: first-order, instantaneous or Monod. For
example, and for the instantaneous model, the reaction terms are computed as follows:
RHO = 0(t)/F0
RHN = N(t)/FN
RHFe = Fe(t)/FFe
RHS = S(t)/Fs
RHC = C(t)/Fc (4.27)
ROH = H(t)-F0
RNH = Hfl+l)1^
RFeH = H(t+l)2-FFe
RSH = H(t+l)3-Fs
RCH = H(t+l)4-Fc (4.28)
91
-------
where F0, FN, FFe, Fs, and Fc are the stoichiometric ratios for each of the electron
acceptors, respectively and H^+l)1, H(t+l)2, H(t+l)3, and H(t+l)4 are the hydrocarbon
concentrations modified by loss due to the reaction with oxygen; oxygen and nitrate;
oxygen, nitrate and iron; and oxygen, nitrate, iron and sulfate; respectively in the given
time step.
For each of the electron acceptors, the following constraints are applied:
Htt+l)1 = 0 where O(t)>H(t)»F0
O(t+l) = 0 where H(t)>O(t)/F0 (4.29)
H(t+l)2 = 0 where N(t)>H(t+l)1»FN
N(t+l) = 0 where H(t+l)1>N(t)/FN (4.30)
H(t+l)3 = 0 where Fe(t) > H(t+l)2»FFe
Fe(t+l) = 0 where H(t+l)2 > Fe(t)/F (4.31)
Fe
H(t+l)4 = 0 where S(t) > H(t+l)3»Fs
S(t+l) = 0 where H(t+l)3 > S(t)/F (4.32)
H(t+l) = 0 where C(t) > H(t+l)4»Fc
C(t+l) = 0 where H(t+l)4>C(t)/Fc (4.33)
Furthermore, these reaction terms are subject to additional constraints. For first-order decay,
instantaneous and Monod kinetic models:
RHN = 0 if O(t+l)>O (4.34)
0 if O(t+l)>O
min
(4.35)
mm
RHS
or N(t+l) > N
min
(4.36)
mm
RHC
(4.37)
nun
where Omin, Nmin, Femin, Smin, Cmin are the threshold concentrations for the corresponding
electron acceptor below which no biodegradation will take place.
92
0
0
0
0
if
if
or
if
or
or
if
or
or
or
O(t+l) > O
O(t+l) > O
min
N(t+l) > N
min
0(t+l) > Omin
N(t+l) > N
min
Fe(t+l) > Fe
min
O(t+l) > O
min
N(t+l) > N
min
Fe(t+l) > Fe
min
S(t+l)>S
-------
For the first-order decay and Monod kinetic models the reaction terms are compared to the
concentration of the electron acceptor:
RHO < 0(t)/FQ (4.38)
RHN < N(t)/FN (4.39)
RHFe < Fe(t)/Fpe (4.40)
RHS * S(t)/Fg 6 (4.41)
RHr < C(t)/F (4.42)
JrlL_. (""
4.2.3 Biodegradation Kinetic Models in BIOPLUME III
The BIOPLUME III model simulates three types of kinetic reactions to represent the aerobic and
anaerobic biodegradation of the hydrocarbon: first-order decay, instantaneous reaction and
Monod kinetics. These expressions are described in the following sections.
4.2.3.1 First-Order Decay Model. One of the most commonly used expressions for
representing the biodegradation of an organic compound involves the use of an exponential decay
relationship:
C = C0»ekt (4.43)
where C is the biodegraded concentration of the chemical, C0 is the starting concentration, and k
is the rate of decrease of the chemical. First-order rate constants can be expressed in terms of a
half-life for the chemical:
0.693
ti/2 = IT" (4.44)
The general literature contains a large number of studies that have determined the half-lives of
many organics detected in ground water. For example, literature values for the half-life for
benzene range from 10 to 730 days while those for TCE range from 10.7 months to 4.5 years
(Howard et al., 1991).
The first-order decay model does not account for site-specific information such as the availability
of electron acceptors. This explains, in part, why the reported half-lives for a given chemical
vary over a broad range of values. Another consideration is the fact that the reported values for
first-order decay rates may have been derived from laboratory experiments conducted under a
specific set of conditions. There has been very little work done to correlate first-order decay
rates developed from laboratory experiments to an equivalent rate that would apply at the field
scale. A number of investigators have alternatively developed methods for estimating first-order
decay rates from natural attenuation field data (Wiedemeier et al., 1995b and Buscheck et al.,
1993).
93
-------
4.2.3.2 Instantaneous Reaction Model. The instantaneous reaction expression, an
expression first proposed by Borden and Bedient (1986), assumes that microbial biodegradation
kinetics are fast in comparison with the transport of oxygen, and that the growth of
microorganisms and utilization of oxygen and organics in the subsurface can be simulated as an
instantaneous reaction between the organic contaminant and oxygen.
From a practical standpoint, the instantaneous reaction model assumes that the rate of utilization
of the contaminant and oxygen by the microorganisms is very high, and that the time required to
mineralize the contaminant is very small, or almost instantaneous. Using oxygen as an electron
acceptor, for example, biodegradation is calculated using the expression:
ACR = - (4.45)
where ACR is the change in contaminant concentration due to biodegradation, O is the
concentration of oxygen, and F is the ratio of oxygen to contaminant consumed. The
instantaneous reaction model has the advantage of not requiring kinetic data. The model,
however, is limited to situations where the rate of biodegradation is fast relative to the rate of
ground water flow.
4.2.3.3 Monod Kinetic Model. One of the most common expressions for simulating
biodegradation is the hyperbolic saturation function presented by Monod (1942) and referred to
as Monod or Michaelis-Menten kinetics:
C
M- =
where ^ is the growth rate (time'l), Mmax is the maximum growth rate (time'l), and C is the
concentration of the growth-limiting substrate (mg/L). The term Kc is known as the half-
saturation constant or the growth-limiting substrate concentration which allows the
microorganism to grow at half the maximum specific growth rate.
The rate equation describing ^ as a function of C contains first-order, mixed-order, and zero-order
regions. When C » Kc , Kc + C is almost equal to C, and the reaction approaches zero-order
with:
V = ^max (4.47)
and ^imax becomes the limiting maximum reaction rate. When C « Kc , Equation (4.46) reduces
to:
94
-------
M^max
M- = -JT~ ' C (4-48)
Kc
Mmax
and the reaction approaches first-order with— JT — equal to the first-order rate constant.
KC
In ground water, the Monod growth function is related to the rate of decrease of an organic
compound. This is done by utilizing a yield coefficient, Y, where Y is a measure of the organisms
formed per substrate utilized. The change in substrate concentration can then be expressed as
follows:
dC ^max M c
dt = Y(KC + C)
where M is the microbial mass in mg/L. Because of the relationship between substrate utilization
and the growth of microbial mass, Equation (4.49) is accompanied by an expression of the change
in microbial mass as a function of time:
-df = ^ax M Y (Kc + Q -b-M (4.50)
where & is a first-order decay coefficient that accounts for cell death.
The advantage of using Monod kinetics is that the constants Kc and ^ax uniquely define the
rate equation for mineralization of a specific compound. The ratio — — also represents the
Kc
first-order rate constant for degradation when C « Kc . This rate constant incorporates both the
activity of the degrading population and the substrate dependency of the reaction. It therefore
takes into account both population and substrate levels, and provides a theoretical basis for
extrapolating laboratory rate data to the environment.
The reduction of contaminant concentrations using Monod kinetics can be expressed as:
AC = Mt Mmax At <
where C = contaminant concentration, Mt is the total microbial concentration, Mmax = maximum
contaminant utilization rate per unit mass microorganisms, Kc = contaminant half saturation
constant, and At is the time interval being considered.
Incorporating Equation (4.51) into the one-dimensional transport equation, for example, results
in:
95
-------
dC d2C dC I C
— - Mt MmaxlFTrl (4-52)
3x2 3x
where v is the seepage velocity, and Dx is the dispersion coefficient.
One of the main difficulties with the Monod kinetic model is estimating the necessary
biodegradation parameters for using the model (the maximum growth rate and the half-saturation
constants).
4.3 Application of BIOPLUME III to Sites
Sufficient field data are essential when using the BIOPLUME III model for simulating existing
flow and/or contaminant conditions at a site or when using the model for predictive purposes.
However, it may be desirable to model a site even when little data exist. The modeling in this
case may serve as a method for identifying those areas where detailed field information needs to
be collected.
Applying an appropriate modeling methodology will increase the confidence in modeling results
with BIOPLUME III. Anderson and Woessner (1992) propose a general modeling protocol that
can be applied to any site. Specific steps which apply to BIOPLUME III include:
1. Establish the purpose of the model;
2. Develop a conceptual model of the system;
3. Calibrate the site model;
4. Determine the effects of uncertainty on model results;
5. Verify the calibrated model;
6. Predict results based on the calibrated model;
7. Determine the effects of uncertainty on model predictions;
8. Present modeling results;
9. Postaudit and update model as necessary.
Stating the purpose of the modeling effort with BIOPLUME III helps focus the study and
determine the expectations from the analysis. Typical objectives include:
1. Determining the effectiveness of natural attenuation for remediating a given site; and
2. Designing a ground water pump-and-treat and/or bioremediation system.
Formulating a conceptual model of the site is essential to the success of a BIOPLUME III effort.
A conceptual model is a pictorial representation of the ground water flow and transport system,
frequently in the form of a block diagram or a cross-section. The nature of the conceptual model
will determine the dimensions of the BIOPLUME III model and the design of the grid.
96
-------
Formulating a conceptual model for the BIOPLUME III model includes: (1) defining the
hydrogeologic features of interest, i.e., the aquifers that will be modeled; (2) defining the flow
system (including boundary and initial conditions) and sources and sinks of water in the system
such as recharge from infiltration, and pumping; and (3) defining the transport system (velocity,
dispersion, sorption and biodegradation) and sources and sinks of chemicals in the system
(including boundary and initial conditions).
4.3.1 Calibration, Verification and Prediction
Calibrating the BIOPLUME III model is the process of demonstrating that the model is capable
of producing field-measured values of the head and concentrations at the site. For the case of
ground water flow, for example, calibration is accomplished by finding a set of parameters,
boundary and initial conditions, and stresses that produces simulated values of heads that match
measured values within a specified range of error.
The procedure for calibrating the BIOPLUME III model is by manual trial-and-error selection of
parameters. The main parameters that are used for calibrating the flow at a site include:
transmissivity, thickness, recharge and boundary conditions. The main parameters that are used
to calibrate the transport and fate of chemicals at a site include: source definition, dispersion,
sorption, and biodegradation parameters. In addition, the transmissivity, thickness and recharge
data used in calibrating the flow solution determine the transport velocity and should be checked
for accuracy against observed field velocities.
Obtaining the information necessary for the BIOPLUME III model is a process that involves
interpreting field data to estimate the values for the model parameters. This process, while not
straight forward in some cases, is crucial to the modeling effort. In general, the site hydrogeologic
and water quality data are analyzed with the objective of predicting trends and estimating the
parameter values for BIOPLUME III. The subsurface geologic data are usually interpreted to
yield values for transmissivity, thickness, and porosity. The water level or potentiometric
surface data are analyzed to determine the direction of ground water flow and the water level
contours. Water quality data are analyzed to determine the spatial and temporal trends in
contaminant distributions at the site.
An emerging tool in spatial data analysis that should be mentioned here is geostatistics.
Geostatistics can be viewed as a set of statistical procedures for describing the correlation of
spatially distributed random variables and for performing interpolation and aerial estimation of
these variables (Cooper and Istok, 1988). Kriging, for example, is one of the most widely used
geostatistical methods to determine spatial distributions of the hydraulic conductivity (or
transmissivity and thickness) at a site. Contouring data using other statistical methods can also
be used as an alternative to kriging.
A quantitative evaluation of the calibration process involves an assessment of the calibration
error. The calibration error is determined by comparing model predicted values to observed
values of the heads and concentrations. Two equations are commonly used for this purpose:
97
-------
n
Mean Error = —\(xm-xs)i (4.53)
7 = 1
Root Mean Squared (RMS) Error
.,0.5
n
— (Xm -
(4.54)
7 = 1
where xm and xs are the measured and simulated values, respectively.
It should be noted that the calibration error is very different and distinct from the computational
error which is a result of the numerical approximation procedures used in the BIOPLUME II
model. Computational errors are discussed in more detail in Section II.5.
Verifying the calibrated site model is the process of using the calibrated model to predict a second
set of measured data from the site. The purpose of this step is to ensure that the calibrated
model is indeed capable of simulating observed site conditions. If the modeling results for the
verification step do not match within reasonable error the observed field data, the model might
require fine-tuning and "re-calibration".
Prediction is the process of using the calibrated/verified model to determine site conditions in
response to an anticipated set of future events. The prediction process is often associated within
a sensitivity analysis similar to that completed with the model after calibration. This is
necessary to determine which parameters specifically impact the predicted results.
4.3.2 Sensitivity Analysis
The purpose of a sensitivity analysis is to quantify the effects of uncertainty in the estimates of
model parameters on model results. During a sensitivity analysis, calibrated values for
transmissivity, thickness, recharge, dispersivity, etc. are systematically changed within a
prescribed range of applicable values. The magnitude of change in heads and concentrations from
the calibrated model is a measure of the sensitivity of the model results to the particular
parameter. The results of this analysis are expressed as the effects of the parameter change on
the spatial distribution of heads and concentrations.
The sensitivity of BIOPLUME III model results to the input parameters is a key analysis that
the user should perform for each site application. This section will present in general the relative
sensitivity of the model to various input parameters using a hypothetical case study scenario.
The user is encouraged to repeat some of these analyses for their specific sites.
98
-------
The hypothetical site "base case" scenario used in the sensitivity simulations was set-up as
follows:
Grid Size 9x10
Cell Size 900' x 900'
Aquifer Thickness 20 ft
Transmissivity 0.1 ft2/s
Porosity 0.3
CELDIS 0.5
Longitudinal Disp. 100 ft
Transverse Disp. 30 ft
Simulation Time 2.5 yrs
Source of Contamination 1 injection well @ 0.1 cfs and 100 mg/L source
cone.
Recharge 0 cfs
Boundary Conditions Constant head, upgradient and downgradient
Chemical Reactions None
Biodegradation Reactions None
Three categories of parameters were analyzed: hydrogeologic, chemical and biodegradation model
parameters. In each category and for each parameter analyzed, the value of the parameter was
changed by a factor of up to one order of magnitude from the "base case" scenario. The
associated model results were then analyzed to determine the impact of the changed parameter
values on the contaminant plume shape, size and concentrations.
Hydrogeologic Parameters. Five hydrogeologic parameters were evaluated: porosity, aquifer
thickness, transmissivity, longitudinal and transverse dispersivity. Overall, model results were
most sensitive to changes in porosity, thickness and transmissivity. This is to be expected since
the three parameters affect the seepage velocity for the aquifer. The data in Table 4.1 indicate
that model results are most sensitive to changes in the transmissivity and aquifer thickness.
Chemical Parameters. Two variables, linear sorption and radioactive decay, were used in this
analysis to illustrate the sensitivity of the model to selected chemical parameters. The user is
encouraged to determine the sensitivity of model results to the remaining chemical parameters
(Langmuir and Freundlich sorption parameters and ion exchange) if they apply to their site.
Both linear sorption and radioactive decay have a substantial impact on the model results as can
be seen in Table 4.2. A retardation factor of 2 caused plume concentrations to decline by 27%
from the "base case" scenario and a half-life of 2 x 107 seconds or 231 days caused plume
concentrations to decline by over 50%.
Biodegradation Parameters. The BIOPLUME III model simulates biodegradation using two basic
methods. The first method involves specifying an overall first-order decay rate to simulate both
aerobic and anaerobic processes. The second method involves specifying the background electron
99
-------
Table 4.1. Sensitivity of Model Results to Changes in Hydrogeologic Parameters
Variable
Porosity
Thickness
(ft)
Transmissivity
(sq. ft. / sec)
Longitudinal
Dispersivity
(ft)
Transverse
Dispersivity
(ft)
0.15
0.3 *
0.45
10
20*
40
0.01
0.1 *
0.2
10
50
100*
10
30*
60
Max. Plume
Concentration
(mg/1)
75
67
80
75
67
47
90
67
57
70
69
67
68
67
66
Plume
Length
(# cells)
6
4
4
6
4
2
3
4
5
3
4
4
4
4
4
Plume
Width
(# cells)
5
3
3
5
3
2
3
3
3
3
3
3
O
3
3
* Base Case
100
-------
Table 4.2. Sensitivity of Model Results to Linear Sorption and Radioactive Decay
Variable Max. Plume
Concentration
(mg/1)
1 * 67
R 2 49
5 28
0* 67
THALF 107 20
(sec) 2 x 107 33
Plume
Length
(# cells)
4
3
2
4
2
2
Plume
Width
(# cells)
O
2
1
3
2
3
* Base Case
101
-------
acceptor concentrations in the aquifer and the selecting an associated kinetic model for the
analysis.
The sensitivity analyses conducted for the biodegradation parameters involved simulating the
impact of using an overall first-order decay parameter as well as the impact of specifying electron
acceptor concentrations with instantaneous kinetics. The results from the analyses are shown in
Table 4.3.
The data in Table 4.3 illustrates that model results are very sensitive to biodegradation
parameters. Regardless of the modeling methodology and biodegradation kinetics, the simulated
concentrations using biodegradation are likely to differ substantially from their counterparts
without biodegradation.
4.3.3 Impact of Non-BTEX Constituents on BIOPLUME III
Modeling
BTEX constituents only comprise a small percentage of the total organic mass in gasoline and JP-
4 mixtures. However, the best available information suggests that most JP-4 and gasoline plumes
will be dominated by BTEX components, and that only a small fraction of the plumes contain
dissolved non-BTEX compounds. This is due to the BTEX compounds having very high
solubilities relative to the remaining fraction of organic mass in these fuel mixtures. In other
words, most of the non-BTEX constituents of gasoline and JP-4 are relatively insoluble, creating
dissolved-phased plumes that are dominated by the BTEX compounds. The following
calculations support this conceptual model of BTEX-dominated plumes from JP-4 and gasoline
releases.
Gasoline composition data presented by Johnson et al. (1990a and 1990b), and JP-4 composition
data presented by Stelljes and Watkin (Stelljes and Watkin, 1993; data adapted from Oak Ridge
National Laboratory, 1989) were used to determine the effective solubility of these hydrocarbon
mixtures in equilibrium with water (effective solubility = mole fraction x pure phase solubility;
see Bedient et al., 1994). The total effective solubility of all the constituents was then compared
to the effective solubility of the BTEX constituents. The following tables show this calculation
for fresh gasoline, weathered gasoline, and virgin JP-4:
102
-------
FRESH GASOLINE
(data from Johnson et al., 1990)
Constituent
Benzene
Toluene
Ethylbenzene
Xylenes
TOTAL BTEX
58 Compounds
TOTAL
Mass Mole
Fraction Fraction
0.0076
0.055
0.0
0.0957
0.16
0.84
1.00
0.0093
0.0568
0.0
0.0858
0.15
0.85
1.00
Pure-Phase Solubility
(mg/L)
1780
515
152
198
152 - 1780 (range)
0.004 - 1230 (range)
-
% BTEX = (63 m.,/1
Effective Solubility
(mg/L)
17
29
0
17
63
30
93
.) (93 mg/L)
= 68 %
WEATHERED GASOLINE # 1
(data from Johnson et al., 1990a)
Constituent
Benzene
Toluene
Ethylbenzene
Xylenes
TOTAL BTEX
58 Compounds
TOTAL
Mass Mole
Fraction Fraction
0.01
0.1048
0.0
0.1239
0.24
0.76
1.00
0.0137
0.1216
0.0
0.1247
0.26
0.74
1.00
Pure-Phase Solubility
(mg/L)
1780
515
152
198
152 - 1780 (range)
0.004 - 1230 (range)
-
% BTEX = (772 mg/L)
Effective Solubility
(mg/L)
24
63
0
25
112
14
126
(726 mg/L)
= 89%
WEATHERED GASOLINE #2
(data from Johnson et al., 1990b)
Constituent
Benzene
Toluene
Ethylbenzene
Xylenes
TOTAL BTEX
64 Compounds
TOTAL
Mass Mole
Fraction Fraction
0.0021
0.0359
0.013
0.080
0.13
0.87
1.00
0.003
0.043
0.014
0.084
0.14
0.86
1.00
Pure-Phase Solubility
(mg/L)
1780
515
152
198
152 - 1780 (range)
0.004 - 1230 (range)
-
Effective Solubility
(mg/L)
5
22
2
15
44
21
65
BTEX = (44 mg/L) (65 mg/L) = 68%
103
-------
VIRGIN JP-4
(data from Stelljes and Watkin, 1993; Oak Ridge N. Lab, 1989)
Constituent
Benzene
Toluene
Ethylbenzene
Xylenes
TOTAL BTEX
1 3 Compounds
TOTAL
Mass
Fraction
0.005
0.0133
0.0037
0.0232
0.045
(4.5%)
0.27
(27%)
0.315
(31.5)%
Mole
Fraction
0.023
0.053
0.013
0.080
0.168
0.832
1.000
Pure-Phase Solubility
(mg/L)
1780
515
152
198
152 - 1780 (range)
0.004 - 1230 (range)
_
Effective Solubility
(mg/L)
42
27
2
16
87
4
91
BTEX = (87 mg/L) (91 mg/L) = 95
In each of these four fuel samples, BTEX compounds comprise the majority of the dissolved
organic mass in equilibrium with water. The non-BTEX components represent a much smaller
portion of the dissolved mass. As expected, the theoretical dissolved-phase concentrations from
these samples are much higher than what is typically observed in groundwater samples due to
factors such as dilution, the heterogeneous distribution of non-aqueous phase liquids, and the low
level of mixing occurring in aquifers (see Bedient et al., 1994 for a more complete discussion).
Note that the total effective solubility of weathered gasoline #1 (125.4 mg/L) is greater than the
total effective solubility of the fresh gasoline (92.8 mg/L). A comparison of the two samples
indicates that the fresh gasoline includes a significant mass of light, volatile compounds that have
pure-phase solubilities that are much lower than that of the BTEX compounds (e.g., isopentane
with a vapor pressure of 0.78 atm and a solubility of 48 mg/L, compared to solubilities of 152 -
1780 mg/L for the BTEX compounds). When these light compounds are weathered (probably
volatilized), the mole fractions of the BTEX components (the only remaining components with
any significant solubility) increase, thereby increasing the total effective solubility of the
weathered gasoline. On the other hand, weathered gasoline #2 has a total effective solubility that
is significantly lower than fresh gasoline (65.0 mg/L vs. 92.8 mg/L), suggesting that this gasoline
has weathered to the point where there has been significant removal of both volatile and soluble
components from the gasoline.
In their analysis, Stelljes and Watkin (1993) identified only 17 compounds representing 31% by
mass of a complete JP-4 mixture. However, a comparison of the relative make-up of the
quantified mixture to the reported make-up of JP-4 (also from Stelljes and Watkin, 1993) shows
the various classes of organic compounds to be equivalently represented in both mixtures. The
quantified mixture can therefore be assumed to be generally representative of the complete JP-4
mixture.
104
-------
Table 4.3. Sensitivity of Model Results to First Order Decay and
Instantaneous Reaction Biodegradation Kinetics
Variable
DEC1
(I/sec)
02
(mg/1)
O2, NO3
Fe, S04,
CO2
Max. Plume
Concentration
(mg/1)
0* 67
0.116xlO~7 58
0.116xlO~6 43
0* 67
3 67
12 66
0, 0, 0, 0, 0 * 67
3,3,3,3,3** 62
Plume
Length
(# cells)
4
4
2
4
4
2
4
2
Plume
Width
(# cells)
O
3
1
3
3
3
3
2
* Base Case
** Threshold Cone. = 0.5 mg/L for all; Stoichiometric Coefficients
in order shown = 3,5, 22, 5, 2
105
-------
% benzenes, alkyIbenzenes in identified compounds: 14% (note: equals 4.5% of 31.5%)
% benzenes, alky Ibenzenes incomplete JP-4 mixture: 18% (from Stelljes and Watkin, 1993)
% branched alkanes in all identified compounds: 26%
% branched alkanes in complete JP-4 mixture: 31%
% cycloalkanes in all compounds identified: 7%
% cycloalkanes in complete JP-4 mixture: 16%
% naphthalenes in all compounds identified: 6%
% naphthalenes in complete JP-4 mixture: 3%
% normal alkanes in all compounds identified: 47%
% normal alkanes in complete JP-4 mixture: 32%
Finally, it is important to note that there is considerable variability among different fresh fuels,
and even more variation among weathered fuels. Therefore, these results should only be used as a
general indicator that the BTEX compounds comprise the majority of the soluble components in
plumes originating from JP-4 and gasoline releases. These results should not be used as absolute,
universal values for all sites.
With regards to biodegradation modeling, however, it is probably appropriate to assume that
BTEX compounds exert the majority (i.e. ~ 70% or greater) of the electron acceptor demand at
JP-4 and gasoline sites. To make modeling BTEX more accurate, however, the total
concentrations of available electron acceptors can be reduced by some fraction to account for the
electron acceptor demand posed by biodegradable non-BTEX organics in groundwater. Two
examples of how to account for the impact for non-BTEX components is to multiply all electron
acceptor/by-product concentrations used in the model by either i) the ratio of BTEX/TOC
concentrations, or ii) the ratio of BTEX/BOD concentrations (if TOC and BOD data are
available). If these data are not available, a conservative approach would be to reduce all available
electron acceptor/by-product concentrations used in the model by 30% to account for the
possible impacts of non-BTEX organics in groundwater.
4.3.4 Mass Balance Assessments
The output from the BIOPLUME III Model includes a hydraulic mass balance and a chemical
mass balance assessment (see Section A.5 in Appendix A). These mass balance assessments
inform the user of how well the numerical techniques are performing in terms of simulating the
specific site conditions. In general, water balances of less than 1% and chemical mass balances of
less than 15% are desirable.
arise
It is the authors' experience that high mass balance errors (for the contaminant) generally
from:
1. Unreasonably high pumping or injection rates for the particular site conditions; or
2. A relatively high seepage velocity in the system; or
3. An inadequate grid that does not accommodate the plume being modeled.
Chemical mass balance errors can be possibly lowered by adjusting the parameters associated
with the above listed reasons. The user should be cautioned, however, that changing these
106
-------
parameters in some instances does not produce the desired effect on mass balance results. The
authors attribute this to the nature of the method-of-characteristics and to the specific algorithm
used in estimating chemical mass balance errors in the model.
107
-------
5.0 PLATFORM USER'S GUIDE
5.1 User's Guide Overview
This is the Platform User's Guide which continues on the themes introduced in Section 2, "Getting
Started." It provides further details on the Platform operations in conjunction with the modeling
requirements of Intrinsic Remediation (Natural Attenuation) studies and the most commonly used
menu options and controls. This document serves to:
• Highlight the basic modeling requirement of Intrinsic Remediation making a smooth
transition from theory to practice.
• Highlight the unique editing tools of the Platform that greatly facilitate the mandatory
creation and validation activities, which are an integral part of the modeling process.
The Platform provides the scientist and engineer the means to work in an interactive computer
graphics environment where the remediation model under consideration (study) is continuously
displayed on the screen. The user navigates through the various parts of the Platform by means of
menus which are always displayed next to the model abstraction. Menu choices are selected and
interaction with the model is performed by pointing with a mouse. By pointing to the screen, rather
than typing commands, a natural dialogue is developed between the user and the platform. In this
perspective, the different Menus become sophisticated graphical editors that speed up the various
steps (chores) of the modeling process.
It is therefore natural to start with a quick review of the modeling steps given in Section 5.2, followed
by the detailed description of the available menus, controls and input parameters given in Sections 5.3
and 5.4. Finally, for the demanding user who wants to have a deeper insight of the Platform
operations, Section 5.5 offers a brief description of the software architecture, a file description and
details of auxiliary video technologies supported by the platform.
Framework
The Platform uses a hierarchical menu system using a main menu (parent menu) which, in turn,
activates a series of secondary menus (child menus). For many applications, input is required in
several secondary menus (child menus). To simplify the description of the different input
requirements and quickly navigate through the User's Guide we summarize the various procedures in
a general framework starting with the basic requirements of the modeling process, gradually
introducing the Platform tools needed to implement the model, and finishing with the detailed input
data necessary to run the problem. This approach is illustrated below.
108
-------
Site
Characterization
i
Section 5.2
Modeling
Ctpn 1
Step 2
Qt^»r\ 7 .....
Step 4
Step 5
otep o
Qt^»r\ 1 ....
1
Intrinsic
Remediation
Protocol
Section 5.3 Menus
and Tools
Main Menus
. . Secondary .Menus
Section 5.4 Input
Parameters
Dialog Boxes
Instructions on
Input Parameters
Activation of
Graphics
As it can be observed, "Raw Data" from the site characterization (site investigation) should be
processed and sorted according to the Intrinsic Remediation Protocol before using the Platform.
These data must be categorized in different modeling steps as indicated in Section 5.2 to speed up
computer implementation. In each modeling step a detailed reference is given to the entry points of
the Platform through the main and secondary menus introduced in Section 5.3 and the corresponding
dialog boxes given in Section 5.4.
109
-------
5.2 Modeling Steps Using the Platform
A model is a word description of the components of a contaminated aquifer system, the "loads" or
"forcing" to the system, and the processes operative on the system. This description is made on the
basis of preexisting data, regional aquifer atlases, or previous site studies. Pictures complement word
descriptions (proverbially "worth a thousand words"). A graphical representation of the contaminated
aquifer is part of the conceptual model. Figure 5.1 illustrates a typical representation of a
contaminated aquifer system with hydrocarbons. Present in the conceptual model shown in Figure 5.1
are a source of contamination, a fuel tank farm leaking at the surface; the vadose or unsaturated zone
through which the Tree product' seeps; the mass of free product that "floats" atop the water table, i.e.
that portion of the aquifer is saturated with fuel; a vapor zone, i.e. unsaturated zone filled with fuel
vapors; and a zone of contact between free product and water table, where the fuel is dissolved into
the saturated aquifer. The dissolved contaminant creates a plume which is advected and dispersed by
the flow of the aquifer. In most instances, the immediate concern is about the quality of the aquifer and
therefore how to control the level of concentration of the dissolved contaminant. The rest of the
phases, leaking source, free product, characterize the release mechanism.
In this typical case study the Platform allows a quick simulation of the dissolved plume, its origin, its
evolution, its migration and biodegradation. The program deals only with light hydrocarbons
(LNAPLs -light non-aqueous phase liquids).
Figure 5.1 Conceptual Model.
Most of the steps that go into preparing for a model using the Platform are grouped in their logical
sequence shown in Figure 5.2. They are as follows:
110
-------
Stepl:
The first thing that needs to be done is the determination of the modeling domain, that is the
geographic extent of the simulation area. Typically this domain will start from the area of interest (for
example a waste site or a well field) and extend to where secure boundary conditions may exist (that is
conditions that are unaltered by forcing that may be imposed within the simulation domain), or
beyond the radius of influence of anticipated forcing mechanisms. The element to consider in
defining the simulation domain is a bitmap of the site showing as many features as available,
including topographic contour lines, surface features, lakes, rivers, drains, observed piezometric heads
and plume delineation. This bitmap is imported in the platform and "registered" to the scales of the
simulation domain defined earlier. It provides the canvas on which to build the ground water
biodegradation model using the platform tools.
Platform Implementation:
Menu "Domain" and its various options is the 'Domain Editor' allowing to enter all of the
above mentioned parameters (see Figure 5.2 for the parameters that need to be defined in this
step).
Recommended Approach:
• Recognize the impact of the advective process in the migration of the Hydrocarbon plume
for the given flow regime. Select a grid domain that covers the anticipated migration of
the plume.
• Combine all possible bioremediation processes to reproduce the observed contaminant
plume (e.g. within a period of '365 days).
Step 2:
Define an optimum grid for the simulation. Grid definition is automated in the Platform and offers
absolutely no inconvenience to the user on two counts:
1. It is drawn by specifying the number of cells in the x (top) and y (left) axes spanning the
simulation domain. Note that the simulation domain can be any rectangular area inside
the topographic domain.
2. The aquifer properties (conductivities, porosity, dispersion) are interpolated to the grid
centers from observed data points by Kriging. A complete assortment of advanced kriging
options are available for the user to control the geostatistical interpolation error. In fact
this is one of the strong points of the Platform because once the raw data are entered the
user does not ever have to revisit them although he/she may test a wide variety of different
grid configurations.
Ill
-------
Modeling Steps
Modeling Activity
Bioplume III Editors
Menu/Option
Stepl: Register the site map with
contours of observed heads and
concentrations. Define surface
bounds (left and right), max and
min elevations. Register raster
background image if available.
Domain/Surface Domain
Domain/Elevation
Domain/Base Image
Step 2: Define grid to better capture
contaminant plume migration.
(Define computational bounds and
grid size number of columns and
number of rows)
Grid/Generate
Step 3: Define Properties of
simulation domain delineated by
the grid extent. Define constitutive
properties of modeled strata.
(Hydraulic conductivities, transport
properties, and others). Fora better
refinement use Log-points and
Kriged zones.
Domain/Define Strata
Edit/Features (log-points, kriged
zones)
Step 4: Define Simulation Loadings
Hydraulic Heads, Concentrations,
and models features (wells, rivers,
lakes, ponds)
Loading/Heads
Loading/Concentrations
Tool-Box -Edit/Features (log-points,
kriged zones)
Step 5: Define Boundary
Conditions and Recharge at the
boundaries.
Grid/Edit
Step 6: Define simulation period
and initial conditions
Initial Conditions/Simulation Period
Step 7: Select simulation
parameters and activate run
Simulator/Bioplume III
Step 8: View graphics of output
results
Results/Heads
Results/Concentrations
Figure 5.2 Required Steps for a Groundwater Contaminant Migration Simulation.
112
-------
Platform Implementation of Step 2:
Menu "Grid" and its different options is the 'Grid Editor' allowing to enter the number of
required cells to create the grid. In this step you define the grid to better capture the contaminant
plume migration. You need to define the computational bounds and the grid size number of columns
and number of rows.
Step 3:
In this step you need to define the properties of the simulation domain delineated by the grid extent.
Then you need to define the constitutive properties of the simulation strata. Such parameters are the
hydraulic conductivities and the transport properties. For a better characterization of the strata in-
homogeneities, if they exist at a particular site, use the Log-points and the Kriged zones available in
the tool box. Then edit these features entering the appropriate constitutive parameters.
Platform Implementation of Step 3:
The menu "Domain" and the toolbox are the program controls to enter the values of these parameters.
Step 4:
Now it is time to define the time dependent "loads" for the simulation. These are the site
measurements of the hydraulic heads, hydrocarbon contours, well pumping schedules, source
mechanisms and other modeling features. The best way to enter these parameters in the program is to
trace actual water and contour level isopleths shown on the site map image. Different modeling
features such as wells, sources, rivers and lakes are external effects that are considered also as loads to
the simulation. These features are created with the help of the toolbox.
Platform Implementation of Step 4:
The menu "Loading" and the toolbox are the program controls to enter the values of these loading
parameters.
Step 5:
Boundary conditions are required so that the numerical model can approximate the flow and
contaminant migration across the grid. BIOPLUME El supports three types of boundary conditions,
inactive, constant head, and constant flow. Constant flow cells produce the effects of pumping wells
and should be used with caution (in many cases they are not justified). All the cells around the
perimeter of the grid must be defined as inactive for the models MOC and BIOPLUME EL They are
automatically set to inactive in the Platform. The constant head condition prescribes water table
elevations at a constant value in certain cells over the entire simulation. Boundary conditions must
also be specified for concentrations, specifically for electron acceptors when recharging conditions
prevail at the boundary of the grid.
113
-------
Platform Implementation of Step 5:
The menu "Grid" and the toolbox are the program controls to enter the values of these boundary
conditions.
Step 6:
At this stage we need to select the initial conditions to run the simulation. In that respect we need to
enter the simulation period (usually several years) and the starting heads and concentrations.
Platform Implementation of Step 6:
Menu "Initial Conditions" allows you to select the above parameters.
Step 7:
We are now ready to run the model. However, you need first to select the time, and execution
parameters as well as the program run time options and the biodegradation parameters. Depending on
the size of the model several minutes are needed to complete the run.
Platform Implementation of Step 7:
Menu "Simulator" includes all the controls needed to run the program.
Step 8:
This is the most enjoyable step of the simulation. The output results come to live in a variety of
graphics. The computed hydraulic heads and the concentrations of the hydrocarbon and the various
electron acceptors can be visually inspected in two-dimensional maps or three-dimensional oblique
views of the two-dimensional data. The calibration errors can also viewed as well as the results at the
observation wells.
Platform Implementation of Step 8:
Menu "Results" allows to view the output graphics.
The conceptual model development is arguably the most important phase for a simulation study where
experience counts the most. The automated/integrated Platform breaks rank with this tradition in two
important ways:
1. Because data entry and model setup are performed by the Platform very efficiently, the
user can concentrate on the physical, chemical, and biological aspects of the problem and
gain experience very quickly.
2. The user does not need to switch from simple to more complicated models: all entered
data are immediately accessible for use in testing new model setups, adjusted numerical
grids, boundary conditions etc.
114
-------
The Menus of the program are in fact sophisticated text and graphical editors that simplify the model
creation. These editors include a variety of objects (e.g. a modeling feature) that take into account
their event-driven nature. These objects have a context that determines their relationship to other
objects, a set of properties that determine their characteristics, and built-in methods that determine
their behavior in response to events.
Objects in the Platform are self-contained. You can change the behavior of one object in the model
without changing the behavior of the remaining objects in the model. This object-oriented design
offers great advantages. Features like recharge and pumping wells, interaction with rivers, drains,
ponds, lakes and aquifers can now be included in the model graphically on the spot with a click of the
mouse. No more hassle trying to input and track the simulation parameters. No need to blindly edit
ASCII files to readjust input parameters. Instead the user can now focus on the model
conceptualization without the need to micromanage input and output files. However, complete
reports on the output results can be found in several output files that reside in the sub-directory of the
case study.
All the details of the input modeling procedures are given in Section 5.3, following the framework of
the above presented modeling steps. Each Menu and its options is considered as a graphical editor
that allows the user to implement particular parameters and features in the program.
Along with the information provided in the next section the user should also consult the Tutorial
(Section 3) which gives several examples, starting with simple cases and ending with the computer
implementation of a real case study.
For the implementation of real case studies pertaining to Air Force Base Sites across the U.S. also
consult Appendix B, "Implementing the Air Force I. R. Protocol Using the Graphical Platform,"
which explains how to analyze 3D field data and extract 2D information for the purpose of running
BIOPLUME m.
115
-------
5.3
Reference on Menus and Toolboxes
This is the Reference Section to which you will refer every time you have a question or need a detailed
description of a particular feature of the Platform. You have a working knowledge of the various
program graphical capabilities if you have read Section 2, "Getting Started". This Reference Section
lists complete and definitive information on all program features.
In particular, you will find a detailed description of:
• The Menus (Editors) and menu options appearing in different Windows.
• The dialog boxes and their input parameters appearing on the screen when a particular
menu option is invoked at various stages of the modeling process.
5.3.1 Description of Menus and Menu Options
As described in the Getting Started section, the Platform has ten basic Menus. Each of the main
menus is associated with secondary pull-down menus which give access to various Platform options,
allowing the user to generate pertinent input data, and activate different tasks of the program. There is
a logical sequence to activating these menus. A particular case study necessitates several iterations,
starting from a simple model and adding more refinements until we reach the desired accuracy. The
Menus in the program are designed to facilitate the user in the difficult tasks of calibrating the model
and validating the results. The description of the Menus is given in Table 5.1.
Table 5.1 Description of the Platform Menus.
Menu Name
File
Domain (Editor of Global Parameters)
Loading
(Editor of Heads and Concentrations)
Edit
(Editor of Modeling Features, Wells,
Sources, Lakes...)
Grid
(Editor of Boundary Conditions and all
Distributed parameters inside the Modeling
Grid Area)
Initial Conditions (Editor of Simulation
Period and initial Conditions)
Simulator
Menu Function
Performs all file management operations, open,
save, restore, delete, close, view file content.
Control parameters defining the geometry of the
groundwater problem and the time domain.
Appropriate selection of the cursor resolution.
Defining all existing loading (Hydraulic heads, &
concentrations) as a function of time.
All editing capabilities for the modeling features
given in the toolbox for the groundwater
contaminant migration problem.
Definition and generation of the grid geometry
used for different resolution processes. Editing of
cell properties, constant/variable flow, inactive
cells.
Selecting initial conditions for the simulation.
Selecting appropriate Simulation module to run.
116
-------
Results
(Graphical Editor of Simulation Results)
View (Editor of Viewing Configurations)
Annotation
Visualization of all relevant data, input as well as
results of various analysis options.
Select/Remove features appearing on the screen
of the "BIOPLUME III" program.
Activating/deactivating Annotations in all
graphical options.
Each of the above main menus is associated with secondary pull-down menus which give access to
the various Platform options, allowing the user to generate pertinent input data, and activate different
tasks of the program. The complete Description of these options (secondary pull down menus)
follows for each Menu Individually.
Menu "Fffle"
On screen you get the following display:
Bippyj.ME.iii HILLAFBO [Main Menu]
J fjomain Loading £dit Grid Initial Conditions Simulator Ftesults View Annotation
New
fiestore
Update Control
Menu Option
New
Open
Restore
Save
Save_As
Delete
Description
Generate a new File which can be saved under option Save_As
Open an existing file from the existing Data Base
Restore output files of a previous simulation (run)
Save the active Input file
Save the active case under a different name
Delete existing files
117
-------
Report
Edit Memo
Transfer
Exit
Browse or Print existing Input or Output ASCII files
An "Editor" to report on the simulation
Transfer via "Modem" selected files to another installation
"BIOPLUME III"
of
Exit the "BIOPLUME III" program
Menu "Domain"
On screen you get the following display:
IBIOPLUMEIII HILLAFBO [Main Menu]
File
""•"""""•••I
D
Loading £dit jjrid initial Conditions Simulator flesults View Annotation
Surface Domain..
jElevation Domain...
Loading Domain
Chemical Species
Define Strata...
Base]mage->
^••sgMMmft
2000
i i i
ilx?
Gth'Street
i i i
V
/
Menu Option
Surface Domain
Elevation Domain
Loading Domain
Chemical Species
Define Strata
Show Layers
Base Image
Description
Global control data defining surface features of the problem
Global control data defining vertical scale features of the problem
Bounds of the "Loading" parameters
Define parameters of Electron Acceptors
Input data defining properties of layered geologic medium
Organization of strata into Computational layers
Import of raster image of surface domain (BMP format)
118
-------
Menu "Loading"
On screen you get the following display:
BIOPLUMEIII HILLAFBO [Main Menu]
Domain BJBiiJl Edit £rid Initial Conditions Simulator flesults View Annotation
Hydrocarbon->
Qxygen->
N[itrate->
Ferric lron->
Sulfate->
Carbon Dioxide->
Menu Option
Background Recharge
Observed Heads
Observed Concentrations
Description
Defining the characteristics of the background
recharge
Locating the Observed Heads in the background of
the simulation domain
Locating the Observed Concentrations in the
background of the simulation domain
119
-------
Menu "Edit*
On the screen you get the following display:
llBiopLUMEiii HILLAFBO [Main Menu]
.Domain Loading 1J!M (3rid initial Conditions Simulator Ftesults View Annotation
Delete All Features
Preferences..
Menu Option
Properties
Cross Section
Delete Feature
Delete All Features
Preference
Description
Input characteristics of modeling features selected with
smartlcons
Stratum properties of selected cross-section
Delete selected modeling feature
Delete all selected modeling features
Parameters for copying and pasting
120
-------
Menu "Grid"
On the screen you get the following display:
BIOPLUMEIII HILLAFBO [Main Menu]
File Domain Loading £dit
D
I I I
EPA-JJ2
-JJ2/E,
» Initial Conditions Simulator Besults View Annotation
generate Grid...
EditGrid->
S elected Kriging
Computational Gnd
Layer £levations
Layer Jhicknesses
Distributed Properties
Kriging £rror (Dist'd Props)
Observed jHeads
Observed Concentrations
EPI
.-82-?
Menu Option
Generate Grid
Edit Grid
Selected Kriging
Computational Grid
Layer Elevations
Layer Thickness
Distributed Properties
Kriging Error (Dist'd Props)
Observed Heads
Observed Concentrations
Description
Automatic generation of the grid according to specified increments
Refined editing of the grid, its boundary and initial conditions
Full "Kriging" procedure of selected properties of geologic medium
View 3D display of computational grid
Show contours of layer elevations
Show contours of layer thickness
Complete "Kriging" procedure for all properties of geologic
medium
Show contours of kriging error distribution
Contours of observed Heads
Contours of observed concentrations
121
-------
Menu "Initial Conditions"
On the screen you get the following display:
JBIOPLUMEIII HILLAFBO [Main Menu]
Initial Conditions
File Domain Loading £dit £rid
D
\Sj— n-
iMll
Simulator Flesults View Annotation
Simulation Period...
i i i i i i
Starting Concentrations..
Gth Street
Menu Option
Simulation Period
Starting Heads
Starting Concentrations
Description
Define starting and end time for the simulation
Select starting head conditions for the simulation
Select starting concentrations for the simulation
122
-------
Menu "Simulator"
On the screen you get the following display:
I BIOPLUMEIII HILLAFBO [Main Menu]
File .Domain Loading £dit find initial Conditions
Results View Annotation
D
Biofilume III
10001
Menu Option
Bioplume III
Description
Activation of Bioplume III model
123
-------
Menu "Results
On the screen you get the following display:
BIOPLUMEIII HILLAFBO [Main Menu]
File .Domain Loading £dit (3rid initial Conditions Simulator
View Annotation
T;i; Hydraulic JHeads
Water Table
Head Prediction Deviations > Pert
Cone Prediction Deviations
Carl
Engineering 6raphs->
AVI Animation
Menu Options
Hydraulic Heads
Concentrations
Water Table
Head Prediction Deviations
Concentration Prediction Deviations
Velocities
Observation Wells
Engineering Graphs
AVI Animation
Description
Contours of computed Heads
Contours of computed Concentrations
Contours of water table
Contours of the computational error on the Heads
Contours of the computational error on the Concentrations
Graphical vector representation of computed "Velocities"
Graphical representation of Observed variables
Graphs of Concentrations
Animation of plume migration
124
-------
Menu "View"
On the screen you get the following display:
BIOPLUMEIII HILLAFBQ [Main Menu]
File D_omain Loading Edit firid initial Conditions Simulator Flesults ^^ffl Annotation
i In
Zoom flut
V1 Zoom to Overview
Snap Cursor
Show grid
Gth'Sti ^ Show Interaction Jools
Show 3D View
Show Base image
V" Make Base Image Gray-Scale
•K— Print...
Menu Options
Zoom In
Zoom Out
Zoom to Overview
Snap Cursor
Show Grid
Show Interaction Tools
Show 3D View
Show Base Image
Make Base Image Gray
Print
Description
Reduce viewing scale
Enlarge viewing scale
Zoom to fit the image in the Window of working area
Activate resolution of the cursor movement
Show generated grid
Show box with smart-Icons (toolbox)
Show 3D View window of simulated domain
Show selected raster image in the background
Show selected raster image in gray scale
Printing the Screen/Window
125
-------
Menu "Annotation"
On the screen you get the following display:
BIOPLUMEIII HILLAFBO [Main Menu]
File fJomain Loading £dit Carid initial Conditions Simulator flesults View
View/Edit
Menu Options
View/Edit
Delete
Delete All
Show Annotation
Description
View or Edit Annotation Window in working area
Delete particular Annotation (Highlighted)
Delete all Annotations in current simulation case
Display Annotation Marks in working area
126
-------
5.3.2 Secondary Menus
In the previous section were described the full list of primary menu options. Several menus such as:
"Loading", "Grid", and "Results" lead to submenus that are specifically designed to support
graphical interactive procedures: a small arrow next to the menu option indicates the existence of
such secondary menu as show below.
QBIOPLUMEIM HILLAFBO [Main Menu]
Edit Grid Initial Conditions Simulator Results View Annotation
Carbon Dioxide->
For example, in menu "Loading" option "Observed Concentrations/Hydrocarbons" has an
arrow to the right indicating that it accesses a secondary menu. Click with the mouse at this type of
menu option and automatically you move to a Secondary Menu, or "Sub-Menu" domain as shown in
the figure below.
Main Menu
Layer
Secondary
Menu Layer
~7
To move from the "Secondary" menu back to the main menu check on the submenu "Main". The
complete list of the secondary menus follows.
Note that the calling option appears in the title bar of the window display for your reference. A total
of 120 graphical options are offered in the Platform allowing you to view the results in 2D and 3D
127
-------
(perspective) configurations. As you may have noticed most of these secondary menus (Sub-menus)
belong to one of five categories as shown below:
Category 1(Base Image Operations)
Category 2 (Loading Conditions)
Category 3 (Grid Editing)
Category 4(lnput Graphics)
Category 5(Output graphics)
"Main"
"Main"
"Main"
"Main"
" Select
"Edit"
"Edit"
"Contour
"Edit"
"Time"
"Layer"
range"
"Main" "Contour range
"Annotation"
"View1
"View"
'Layer"
"View"
"Annotation"
"Annotation"
"View" "Annotation"
"Time" "Layer" "View"
The submenus are activated from all menus that are followed by a small arrow (•*). The list of the
basic items of these "Submenus" is given in the following tables. (The result of their activation is self
descriptive).
Category 1 (Base Image Operations)
Main
Select
From File
From Clipboard
Deselect
Edit
ImageCenter
Copy to Clipboard
Register Image
View
Zoom In
Zoom Out
Zoom to Overview
Snap Cursor
Show Features
Show InteractionTools
Print
128
-------
Category 2 (Loading Conditions)
Main
Edit
Properties
Delete Load Object
Delete All Load
Objects
Time
Previous Time-step
Next Time-step
Select Time-step
Edit Time-step
View
Zoom In
Zoom Out
Zoom to Overview
Annotation
View/Edit
Delete
Delete All
Show Annotations
Category 3 (Grid Editing)
Main
Edit
Edit Head
Boundaries
Edit Concentration
Boundaries
Delete grid-line
Layer
Previous layer
Next layer
Select layer
View
Zoom In
Zoom Out
Zoom to Overview
Annotation
View/Edit
Delete
Delete All
Show Annotations
Category 4 (Input Graphics)
Main
Contour Range
Span current layer
Span all layers
Set number of
levels
Layer
Previous layer
Next layer
Select layer
View
Zoom In
Zoom Out
Zoom to Overview
Annotation
View/Edit
Delete
Delete All
Show Annotations
129
-------
Category 5 (Output time dependent graphics)
Main
Contour
Range
Span current
layer
Span all layers
Set number of
levels
Time
Previous Time-
step
Next Time-step
Select Time-
step
Edit Time-step
Layer
Previous layer
Next layer
Select layer
View
Zoom In
Zoom Out
Zoom to
Overview
Annotation
View/Edit
Delete
Delete All
Show
Annotations
5.3.3 Available Menu Options in the Icon Bar
The 'Icon Bar' which is located below the 'Menu Bar' contains different sets of "Smartlcons" that
can enhance the interactive operations. These '"Smartlcons" are mouse shortcuts of the most
commonly used menu options. A brief description of their function is provided in the following
graphs.
For the Primary Menu System
130
-------
For the Secondary Menu System
5.3.4 Tool-Box Features
The tool-box provides to the user all necessary tools to build a model or to select a particular graphical
mode (function). These tools cover the basic modeling features such as the log-points, wells,
contaminant sources, rivers, ponds and lakes. They are all contained in the tool-box for easy
access. All the user has to do is point, click and drag. The selected modeling feature is created on the
spot. Once a feature is created in the working area of the screen (highlighted), the user proceeds to the
menu item "Edit" to input its properties. The feature edit option is also accessible by double-clicking
on the feature (well, lake, river) in the working domain. Similar Tools are also available for editing
the computational grid and specifying boundary conditions (constant head, concentrations).
The tool-box also features tools that allow the evaluation of linear and polygonal distances and areas
graphically on the screen using point-and-click procedures. These tools are useful to quickly evaluate
contaminant extent and migration rates. The tables below offer the description of all features in the
various tool-boxes.
131
-------
Tool-Box in Main Menu Layer
Contamination
Source
For more details on how to use these tools you can refer to Section 2, 'Tutorial' and the next section.
Tool-Box in Secondary Menu (Activated by "Loading/Observed Heads, or
Concentrations)
Observed Heads
Observed concentrations
Zone of Constant
Concentrations
132
-------
Tool-Box in Secondary Menu (Activated by "Grid Editor for Hydraulic Heads")
Pointer .
v
\EEfm^^Mx
Constant
Head/
Concentr.
Variable
Head/
Concentr.
/
w
^
f+\
K!
VI
iAi
2
>
®
-
Inactive
/ Head/
Concentratr.
Tool-Box in Secondary Menu For 2D Contour Maps
Data Capturing
Tool
K
Bitmap
Capturing Tool
133
-------
Tool-Box in Secondary Menu For 3D Elevations
(BIOPLUME m supports only 2D modeling)
Cross-section
Selection
Rotating the 3D
View
With the presentation of the tool-box features we conclude the description of the basic operations of
the program that allow an entry point to the implementation of the modeling steps introduced in
Section 5.2. In the next section we continue with the definition of the input parameters that are
required to run the BIOPLUME m simulation.
134
-------
5.4 Reference on Dialog Boxes and Input Parameters
This is the Reference Section to which you will refer every time you have a question or need a detailed
description on a dialog box and its corresponding input parameters appearing on the screen when a
particular menu option is invoked at various stages of the modeling process. All necessary data
needed to build and validate a simulation are easily handled using only a few dialog boxes that are
properly managed from the main menu as shown below:
Operation
File Manipulation
Simulation Domain
Loading Parameters
(Heads,
Concentrations)
General Input Data
Initial Conditions
Numerical Simulation
Graphics of Results
Detailed Modeling Activity
New, Open, Save
Define area, limiting values of basic
parameters, & electron acceptors
Define Heads and Concentrations
at different times.
Define numerical grid and
associated distributed properties of
the soil media
Define simulation period, select
initial Heads and Concentrations
Select running options and run the
case study
View different graphics of output
results
Corresponding Menu
"File"
"Domain"
"Loading"
"Edit", "Grid"
" Initial Conditions"
"Simulator"
"Results"
All input data required to run BIOPLUME El is prompted from the user by the Platform in a
completely interactive manner. The error-prone formatting chores of editing input streams are totally
eliminated. Furthermore, the user is guided by the appropriate dialog boxes as to what type of input is
expected, the default value, and the range of values that are expected. The details on how to use these
dialog boxes are discussed in the following pages. To facilitate the presentation, we show the input
parameters of the dialog boxes exactly as they appear on the screen. Most of these dialog boxes are
self explanatory. However, where needed you will find a brief explanation and description on the
nature of these parameters. You can also consult Section 2, 'Tutorial' which has examples on how to
use these tools.
5.4.1 Dialog Boxes Associated with Menu 'File'
Menu Item: "New"
This item initializes the Platform for a new case study.
135
-------
Menu Item: "Open"
It allows the user to open an existing case study. It will retrieve all the files found in sub-directory
..\EISBioplume\Data\"Case Name" bearing the name of the case study. These files are automatically
loaded into the system.
Use the mouse to open an existing case by double-clicking on the selected application name. Now
you can access these files, to upgrade the case or view existing graphics and output results.
Menu Item: "Restore"
If for some reason the changes that you have made into an application case are not satisfactory you can
restore the files to the original case (prior to the last change) by activating this menu option.
Menu Item: "Save"
It allows the user to save an existing case including the changes affected (edited) on that file.
However, be aware that in this case all existing (simulation) output files will be deleted. A warning
screen appears and gives you the option to "Cancel". As a rule, it is advisable to maintain simulation
results in the original case name and to give the new version a new name using the "Save_As" menu
option.
Menu Item: "Save As"
It allows the user to change the name of an existing case by typing the new name in the appropriate
box.
Menu Item: "Delete"
This item deletes all data and output files of the specified case. A warning screen appears on the
screen that allows you to backtrack by "Canceling". BE AWARE, this is a strong command: all files
136
-------
from the case, input and output will be deleted from the system for good! (at least consider making a
back up tape).
Menu Item: "Report"
This option allows the user to view the ASCII (output) file from a BIOPLUME m simulation. See
Section 4, 'Theoretical Development' and Appendices for more information on the contents of this
file.
IH)r:.H H» ,1 '.!'.
'"MHSFQ8T WITH HIIT-PI1 n.HTTIS* ft:".>71flR piaiOTMMTI
i M p • i DAT*
MtCRIFTOK
m cMjHimi or
jw (Minnni or mm
1DU . ar'nim-.tK, n.hiDDst
MINI (.PUIIIIKi fSHJUl IN ViAlt>>
LIHK i. lilt I If/liHIjUII II!. LI I If lil W
. \Ml I'lHlTlRI. I I HI: sriH Ih »!•'..
I.HHM
.HH
w.
-Li, J
Menu Item: "Edit Memo"
A small editor to report on the simulation. Simple editing tools are supported.
Menu Item: "Transfer"
A simple module allowing you to transfer selected files via modem to another installation of the
Platform.
Menu Item: "Exit"
Exit command to close out the Platform Windows application.
5.4.2 Dialog Boxes Associated with Menu 'Domain'
Menu Item: "Surface Domain"
This item allows the user to define the domain of interest for the simulation. It also allows the user to
determine the scaling of the vertical and horizontal rulers and cursor resolution. The cursor resolution
controls the "snap" cursor action. Use a consistent set of units for length. For your reference the
values displayed in the dialog box are taken from the case study "Hillafbl". In this case the
coordinates of the origin at the top left corner of the domain are (0., 0.), while the coordinates at the
137
-------
bottom right corner are (2500.,2000) in feet. Note, however that any pair of coordinates can be
selected at the origin.
I Surface Domain
- Surface Bounds [Viewable Area of Interest, flj : - Cursor Resolution
i Left: [II
i Top: |0.
i - Horizontal
i Major: 400.
i Minor: |100.
Flight:
Bottom:
rVertica
Major:
Minor:
2500. 10-
2000-
I 1
400, OK
100.
Ldricel
Menu Item: "Elevation Domain"
This item allows the determination of vertical (elevation) domain, Ruler increments, and cursor
resolution. Top and bottom elevations of the aquifer layer vary between 4560 ft and 4690 ft. above
sea level in the example.
I Elevation Domain
Elevation (ft)
•Ruler Tic Increments
Minimum:
Major:
Minor:
Cursor Resolution
Menu Item: "Loading Domain"
This menu option allows the user to select the bounds (Upper and lower limits) of the following
simulation parameters.
Time
Pumping Rates
Concentrations
Infiltration
Hydraulic Heads
138
-------
Each item activates a dialog box to enter pertinent input information. These dialog boxes are shown
below.
•_..r_tl |
II.,. I, .......
"V ,l*
Mm* *'
The ruler tic increments in all of these boxes determine the appearance of the rulers in all
corresponding graphical displays. Note that if the increment is too small, the numerical characters will
be difficult to read.
Menu Item: "Chemical Species"
This is an option that allows the user to enter the properties of the hydrocarbon contaminant and
potential "Electron Acceptors" used in the simulation. The dialog box for the hydrocarbon is shown
below.
_S=lJ
139
-------
To enter the reaction properties for the hydrocarbon click on the "Reaction" button as shown below.
I Chemical Reaction Parameters for Hydrocarbon
Sorption Options
f* ;No Sorption Isotherm;
("* Linear Isotherm
f~' Freundlich Isotherm
C" Langmuir Isotherm
Sorption Parameters I
Decay Options
(• No Decay
'"~" Radioactive Decay
l'~" Biodegration Decay
Decaf
Ion-exchange Options
'•*" No ion exchange
("" Monovalent exchange
f~' Divalent exchange
f~' Monovalent-divalent exch.
f~" Divalent-monovalent exch.
Bulk Density
OK
This is where you input the "Chemical Reaction Parameters"; for selection of chemical reaction type
among Sorption, Decay, and Ion-exchange; Parameters buttons activate the next dialog box. Note that
all relevant parameters are entered once. Selection of chemical reaction type, if any to include
in the simulation is done at "Run Time Option" dialog box discussed in menu "Simulator". The
next three dialog boxes show the editing boxes of the "Sorption", "Ion Exchange", and "Decay"
parameters.
| Decay Parameters
Biodegradation
First-older rate for |[jj
dissolved phase *-—
First-order rate for
sorbed phase [1/T]
Set Defaults
- Layer
Ok
The same dialog boxes also appear for the reaction parameters of the "Electron Acceptors".
140
-------
To enter the reaction and interaction properties for the "Electron Acceptors" click on
"Domain\Chemical SpeciesVElectron Acceptors to obtain the dialog box shown below.
I Bradegratilaliofi fe, lection AceepMifs ma I
r Se
t E lectron Acceptor Parameters
[Jl^Pfil" I Heaetion InteraeUcm
Nitrate I Reacticm Interaction
Ferrb Iron j Reaction Interaction
Sulfate fteaction Interactiofi
Carbon Dioxide fleaclion InteractJotii
OK | Cancel |
This dialog box includes all the electron acceptors supported by BIOPLUME El. Since the reaction
parameters are the same as for the hydrocarbon, we illustrate here only the "Interaction" parameters
shown in the dialog box below.
IBiodeqiaoatian Parameters tor Oxygen
Stoichiometric Ratio of Election
Acceptor to Hydrocarbon (-):
Electron Acceptor Threshold
Concentration:
First Order Decay Rate;
Maximum Hydrocarbon Utilization
Rate:
Hydrocarbon Half -Saturation
Constant:
Electron Acceptor Half-Saturation
Constant:
Total Mierobial Concentration:
Retardation Factor for
Microorganisms:
i
0.
0.
0.
a
0-
0.
0.
OK
Cancel
This dialog box shows the parameters for oxygen only; the list of Interaction parameters is the same
for all "Electron Acceptors".
141
-------
For further information on the nature and impact of these parameters on the simulation of "Intrinsic
Remediation" consult Section 2, "Tutorial," and Appendix B, "Intrinsic Remediation Protocol".
Menu Item: "Define Strata"
The "Strata Definition" Box is provided for specification of background (constant, default) properties
of the computational layer. The "Transport Properties" activates the screen for the dispersion
coefficient.
Transport Properties for Stratum 1
Dispersivity (ft):
Dispersivity Ratio (-):
Veitical Dispersivily Ratio:
Molecular Diffusion Coefficient (-):
Bulk Density (g/cm3):
10.
D
Ok
Menu Item "Base Image"
This option allows the selection and identification of the base image on top of which the modeling
features will be defined using the tool-box. Upon activation of this option a secondary menu appears
on the screen as shown below.
142
-------
The big "Up Arrow" in the upper left coiner of the screen indicates that you are in a Secondary Menu
(Child Menu). The displayed image is obtained in two steps. In the first step we load an existing
bitmap from the bitmap files stored in Sub-directory "..\EIS\Bioplume\Image\". This is done by
activating menu option "SelectMmage File" and enter in the editing box the name of the required
bitmap file (files with extension .BMP). In a second step register the image that is displayed on the
screen.
Select Image File
File
afhillim.bmp
eglin0.bmp
eglinl _bmp
gwig1.bmp
mbtutl.bmp
mftut~l.bmp
patrick.bmp
test01.bmp
List of type:
.BMP
Polders;
c:\eisbio~1 \image
OK
c:\
eisbio™1
•$ image
_
c: hp_pavilion
To register the image select the appropriate icon from the toolbox (the button adjacent to the zooming
tool) and click on the working area to define the top left coiner of the simulation domain. Then drag
the mouse to the bottom right coiner to register the area for the simulation. Releasing the mouse
causes a dialog box to appear. Now all you need to do is enter the coordinates of these two points and
the registration is completed.
The background image can be used to fulfill different objectives within the framework of the same
application. For example, the background can be used to:
• locate the modeling features for a particular simulation
• graphically enter the initial and observed hydraulic heads, and
• graphically enter the initial and observed concentrations
The Platform allows interchanging of the background image. However, the user should check that the
image registration points are the same in all background images for compatibility.
143
-------
5.4.3 Dialog Boxes Associated with Menu 'Loading'
This menu allows the user to enter all known observed (measured) data from a monitoring activity.
In that respect all the parameters that are edited with this menu are time dependent.
Menu Item "Background Recharge"
This option allows to impose a specified recharge rate (flux) and concentrations (hydrocarbon and
electron acceptors) throughout the simulation domain at all known times as illustrated below.
[ Background Recharge
r Lc
• *
lading
Time Steps
; Infiltration 1
Concentration
Hydrocarbon ^||
Set Values
Ok | Cancel
Desired time-steps in which site recharges are known can be entered by pressing on the "Timesteps"
button. This activates the following screen.
(Loading Timesteps
-Edit Timesteps —
Time (Days)
Ok
Add Timestep
Cancel
Add
Copy Timestep*
Note that you can add, copy, paste and delete different times.
144
-------
To enter the observed fluxes click on the "Infiltration" button and specify the recharge rate (fluxes) at
all specified times. Copy and Paste buttons allow to duplicate previous record. The entry is numerical
and/or graphical as shown below.
Time
Flux
BOO
1000 =
900
air
500 -
400 =
300 :
100 =
20.
Time (Days)
You can either enter the values at different times in the editing box or with the mouse by clicking and
dragging the nodes for each specified time at the desired level. Follow the same procedure to specify
"Recharge Concentrations" as shown below.
| HydfocarDor* Concentrations
Time
Concentration
400
0.
.200
.EDO
IB-
IB.
46.
52.
0.
50 E
Time (Days)
145
-------
The previous three dialog boxes are typical of all time-dependent parameters. The same dialog boxes
are encountered in many other parts of the program. Note that you can also specify "Recharge
Concentrations" for all "Electron Acceptors".
Menu Item "Observed Heads"
This item allows specification of known hydraulic heads at all specified times in a secondary menu
environment shown below. You can input the hydraulic head contours (e.g. as shown on the
background bitmap) using the appropriate tool from the tool box. Just point and click.
Once a contour is placed on the map, double click or edit this feature through menu edit to enter the
contour level using the following dialog box.
Head Contour 40
Contour Path
Point X-Coocd Y-Coord
Layer
Head
1320.
940.
Ok
Cancel
You then proceed with the next contour of hydraulic heads. This procedure takes only a few minutes
and the rest is taken care of by the Platform. The hydraulic heads will be automatically distributed at
each grid cell where it is needed. Note also that the contours for the starting time can automatically be
considered as initial conditions for your simulation. Subsequent times are considered as target values
146
-------
for your calibration. The platform automatically tracks down the error between the observed and
simulated values, (see option "Head Prediction Deviations" in Menu "Results").
Menu Item "Observed Concentrations"
This item allows specification of known concentration of the hydrocarbon and the various electron
acceptors at all specified times in the secondary menu environment shown below. You can input the
concentration levels (e.g. as shown on the background bitmap) using the appropriate tool from the tool
box. Just point and click.
Once a Concentration zone is placed on the map, double click or edit this feature through menu edit to
enter the concentration level using the following dialog box.
Concentration Zone 1
Concentrations —
Layer Concentration
Zone Perimeter —
Point X-Coord
RI 11880.
Y-Coord
840.
Note that with the concentration zone you need also to specify a logpoint with zero or background
concentration to trigger the kriging module. The concentrations will be automatically distributed at
each grid cell where needed. Note also that the contours for the starting time can automatically be
considered as initial conditions for your simulation. Subsequent times are considered as target values
147
-------
for your calibration. The platform automatically tracks down the error between the observed and
simulated values (see option "Concentration Prediction Deviations" in Menu "Results").
5.4.4 Dialog Boxes Associated with Menu 'Edit'
This menu allows the user to input the appropriate input data (properties) associated with the
modeling features that are available in the toolbox. In the Platform these modeling features are:
Recharge Zone,
Contaminant Source,
Wells,
Rivers,
Drains,
Evapotranspiration zone,
Subkriging Domain,
Logpoints,
Transects
Note that these features must be activated in the graphics working area. You activate a feature by
setting the mouse in the pointer mode and by clicking on the feature with the mouse. Bioplume m
responds by highlighting this feature. Now you are ready to activate the appropriate dialog box to enter
the input parameters of this feature by either double-clicking on the highlighted feature or pressing
option "Properties in menu "Edit".
Log-points, wells and recharges are the most commonly used modeling features and are described in
details herein.
Log-points
Several successive dialog boxes are needed to edit its parameters as shown below.
148
-------
Log Point 13
Surface Coordinates
X: 750.
Y: 440.
Cross Section
Cancel
I Point 1 3 Cf oss S ection Q
-
-
4650^
4600 _
'
Elevations
Level Elevation
|D | 4625.
2 4602.
Stratum Properties ,
Edit at Selected Elevation
Copy All Paste All
| Ok J Cancel |
The cross section properties show the thickness of the computational layer at this particular location.
To enter the parameters for the aquifer click on the button "Edit at Selected Elevation". You will get
the dialog box shown below to enter the Horizontal Conductivities, Storage Coefficient, Effective
Porosity and Longitudinal Dispersivity.
Horiz. Hydraulic Conduct.
l~" Activate I
Storage Coefficient
I~ Activate I
Effective Porosity-
["" Activate I
Longitudinal Dispersivity
P" Activate I
Wells
Ok
The dialog box for editing a well is shown below.
Loading
Time Steps
Pumping Rates
Concentration
Hydrocarbon
Set Values
Screen
r Surf
X:
Y:
ed Stratum: |l |
ice Coordinates i
1830.
460
OK
Cancel
149
-------
You start again by specifying the time steps in which the pumping rates changes and then enter the
Pumping rates and various concentrations as shown in the next three dialog boxes.
1 Loading Timesteps
Edil Timesteps
Time (Days)
Add Timestep
Add
Delete Tiroes! ep
0,
Cancel
Recharge
For the recharge zone the procedure is the same: double-click on the recharge zone to obtain the
following dialog box.
150
-------
Irtecriarge Region 1
Active Stratum: P
Loading
Time Steps
Huxes
- Concentration -
Hydrocarbon
Set Values
Perimeter
Point X-Coord
Y-Coord
130.
130.
130.
2250.
2240,
2BO.
250.
140.
To enter the recharge parameters the procedure is the same as before. Enter the Time steps followed
by the input recharges and concentrations.
The other modeling features have similar requirements that the user will find easy use.
Menu Item "Cross Section"
Cross Sections are defined by two points in the working area. The "Cross Section" button brings up
the 2-D graph of the cross section (transect). New log-points are added mid-way to the right of an
active log-line. Move it to its exact location by clicking on the vertical line and holding until properly
placed. Adjust the thickness of the computational layer and input layer properties. The dialog box
below shows a typical cross section along with the buttons to enter its properties.
Transect 1 trass Section
4650
4600
r Transect 4
i Distance from |
| Transect Endooint I
| Elevations
i Level Elevation
Stratum Properties
Edit at Selected Elevation 1
151
-------
Clicking on the button "Strata Properties" activates a new dialog box to enter aquifer's properties as
shown below.
1 Straitum Pioperli^s £il Sefeetetl Etevailidri
— HorizL Hydr-au
|T~ Activate
F~ Activate 1
ic Conduct. —
1 Activate
Longitudinal D
1 Activate 1
ispersivitji
Ok I
i Cancel |
Menu Item "Delete"
Activating this option will delete the feature that is highlighted in the working area.
Menu Item "Delete All Features"
Activating this option will delete all the existing features of the simulation model. Beware, this is a
very drastic command.
Menu Item "Preference"
This option allows the selection of different copying and pasting parameters.
Input Pfeferenees
New Well
C**|Use default properties;
'"""Use previous well properties
OK
Cancel
- New Log Point
" Use default properties
" Use previous log properties
-Surface Object Creation
f* Switch to Object Select mode
'""'" Stay in Object Create mode
There are three sets of parameters. The first set allows the user to change the properties of the new
well (pasted) from the default values to the properties of the previously created well. The same effects
have the other two sets that are dealing with logpoints and selected contour levels.
152
-------
5.4.5 Dialog Boxes Associated with Menu 'Grid'
This menu allows the definition and generation of the grid which is used to discretize the aquifer. It
allows to access and use an enhanced "Kriging" procedure to infer aquifer properties at grid points.
Menu Item "Generate Grid"
The user defines a computational grid, either by increments, or by number of columns and rows. Note
that the computational domain can also be narrower than the base map domain. These grid
parameters are shown in the following dialog box.
- Computational Bounds -
Right: [2500. |
Bottom: [2000. |
-Grid Size-
Number of Columns: 20
Number of Rows: 15
Generate Grid
Cancel
Grid Increments
Column Increments:
Row Increments:
Once the "Generate Grid" button is clicked the new grid appears in the working area of the screen.
Menu Item "Edit Grid"
This option is for editing grid features, specifying constant head/concentration cells, inactive cells and
observation points. Clicking on the button "Edit Grid" spawns a secondary menu as shown below.
-------
iTTsim^T
To select an observation
well point just double
click on a location in the
working area as shown
above and activate this
option using the
prompted dialog box. To
select an inactive or
constant head grid cell,
use the tool box and the
mouse to graphically edit
the grid.
1 "*?w™" 0
tej W.UI *
.•Jlj1' yf Y" '-.vsi ^""ntf fife^i ^ ^ i"v w^JtrT" *"itVs*«VsrJr
* . ' •* -,( ; r+ '•
• -- : * • T . •.. i
Menu Item "Selected Kriging"
All input parameters such as material properties, hydraulic heads and concentrations of hydrocarbons
and electron acceptors are usually given at particular locations (logpoints). Therefore they need to be
distributed throughout the computational grid. This is done automatically in the Platform whenever a
"Save" procedure is activated. A default simple kriging procedure is used. However, for better
accuracy you need to explicitly activate the "Quick Kriging" algorithm. This is the case when this
particular menu option is activated. First select the type of parameter you need to re-krig and you will
obtain the following screen (in this example the observed Hydraulic Heads).
Layei
Kriging Options
Show Statistics
Delete Krig Data
To re-adjust the parameter distribution obtained by the default kriging click on the "Delete Krig
Data" button and activate option "Quick Kriging". After a few seconds the operation is completed
and you can proceed with the re-adjustement of another parameter. After "Saving" the case, the new
distribution will be available to inspect graphically. The demanding user may also want to inspect the
154
-------
statistics of the examined parameter. All you need to do is to click on "Show Statistics" to obtain the
dialog box below.
This allows the user to quickly validate the selection and orientation of the computational grid. The
Platform computes the variogram of the parameter (e.g. observed hydraulic heads) in all pertinent
directions. In the example the Northing variogram (y-direction) is flat showing little variation, while
the Easting (x-direction) variogram is smooth with exponential growth. "Show Variogram" displays
the graphics below. The variogram in the "Easting" direction confirms that the predominant flow
regime is in the "x" direction.
Menu Item "Computational Grid"
This is an option that allows the user to display a perspective view of the computational grid and the
node wells.
155
-------
Menu Item "Layer Elevations"
This is also an option that displays the computational layers in a perspective configuration. Cross
Sections are obtained on the fly using the appropriate tool from the toolbox. Just point and click along
the red or blue line to get the cross section below.
Menu Item "Layer Thickness"
The distribution of the thickness of the computational layer across the simulation domain can also be
obtained using this option. Contours of the layer thickness are shown below.
E9 T SESS
156
-------
Menu Item " Distributed Properties"
Essentially, this option displays on the screen the results of the kriging procedures on the various
distributed parameters (input parameters). The distribution of the hydraulic conductivity is shown in
the example below.
Menu Item "Observed Heads"
The results of the kriged observed heads at different times are shown in this option as shown below.
Menu Item "Observed Concentrations"
The results of the kriged observed concentrations at different times are also displayed in this option.
157
-------
5.4.6 Dialog Boxes Associated with Menu 'Initial Conditions'
Menu Item "Simulation Period"
This option allows specification of starting and ending time of the simulation. To aid in this selection,
a summary table is given of all the times when pumping or recharge are specified.
i Simulation Hanod :
s*
jlect Starling Time
Time Combined Observed Observed
(Years! Pumping Rate Heads Concentrations
r Starting! BatB~""""i pEnding Time (Years)*™] I jp'j^ 1
Febl.1990 : 1
J
Menu Item "Starting Heads"
The following dialog box allows the selection of starting conditions among the following options:
'Observed values' (earlier entered in menu "Loading"), 'Previously Generated Heads' (from a
previous run), 'Constant Heads' and 'Field Filled to Capacity". More information about these options
can be found in Section 3, Tutorial.
(Initial Conditions foi Hydraulic Heads
Initialization Options
f"° : ::•-:
f~ Use Constant Head Values
f* Fill Field to Capacity
Ok
Cancel
Menu Item "Starting Concentrations"
The following dialog box allows the selection of starting conditions for hydrocarbon and electron
acceptors among the following options: 'Observed Values' (entered in menu "Loading"), 'Previously
158
-------
Generated Concentrations' (from a previous run), and 'Constant Concentrations'. More information
about these options can be found in Section 3, Tutorial.
[in
tial Conditions for Concentiati
Dxygen
titrate
-erne lion
Sulfale
Zaibon Dioxide
ons
^^^^^^^^^^^^^^^^^^^^^fx\
(• Use Observed Values !
f* Use Pie? Geneialed Concerilralions j
C Use Constant Concentrations j
[ OK i| Cancel
5.4.7 Dialog Boxes Associated with Menu 'Simulator'
This is the menu that allows activation of the BIOPLUME El run after selection of appropriate
runtime options shown below.
Note that the "Bioplume III" button is originally grayed out (not accessible). It becomes
accessible only if the starting "Heads" and "Concentrations" are specified in the previous menu.
':-•.•• ''!.-!.. a.I RL.. IjiaitUtHI
The run time options for BIOPLUME m are the following: Time parameters, Execution parameters,
Program options, Transport Subgrid, Biodegradation. These options are discussed below.
159
-------
Time Parameters
I Time Parameters
Maximum No. of Time Steps: Q
Time Increment Multiplier: 0,
Inital Time Step in Seconds: 0-
F?" Steady State Run
The time parameters allow the selection of the number of time steps for a particular run. Note that for
BIOPLUME HI the initial time step should be given in seconds.
Execution Parameters
i Execution Parameters
No. of Iteration Parameters:
Convergence Criteria:
Maximum No. of Iterations:
Maximum Cell Distance per
Move of Particles (MOC):
Maximum No. of Particles:
No. of Particles per Node:
i
.00999
100
.5
3000
9
OK
These are Standard MOC program runtime parameters. For details, see Section 4, BIOPLUME HI
Theoretical Development.
160
-------
Program Options
Program Options
Time Step Interval far Complete
Printout:
Move Interval for Chemical
Concentration Printout:
Time Step Interval for Velocity
Printout:
[0 = No)
(-1 = First time step)
(-2 = All time steps)
Dispersion Coetf. Print Option:
(0 = No)
(1 = First time step)
[2 = All time steps)
Time Step Interval for Velocity
Printout on File Unit 7:
(0 = No)
(-1 - First time step)
1-2 = All time steps)
Concentration Change Print Option
10 - 20):
OK
These are again Standard MOC program parameters. For details, see Section 4, BIOPLUME
Theoretical Development.
Transport Subgrid
[Transport Subgrra
OK
This is also sub-gridding option of the MOC program. Specify the top left cell (column and row
number) and the bottom right cell which defines a sub-region over which the transport simulation will
be performed. The flow simulation is performed over the entire grid. This option is less useful with
faster and more powerful PC computers.
161
-------
Biodegradation
This is the Platform
biodegradation option
selection table. It
allows to
activate/deactivate
electron acceptors, and
to select interaction
mechanisms
(Instantaneous Reaction,
Zero Order Reaction
and Monod Kinetics).
IBiodepadalion Q pliers
I Election Acceptor
| Byproduct
| Ogjigen
E Nitrate
| Ferric Iron
E Sulfate
| Carbon Dioxide
Inactive First Order Instantaneous j Zero Order Monod
Decay Reaction i Reaction Kinetics
ft- r r \ r r
(f r r j r r
(f r r j r r
(f r r j r r
(j- r r j r r
OK | Cancel
XI
Now you are finally ready to run the simulation. Press the button "Save Data and Run Simulation"
to activate sequentially three executables as follows: an executable to generate BlOPLUMElll input
stream from the graphics files, an executable that runs the B1OPLUME 111 algorithm, and an
executable to create output graphics. For more details on the proper sequence with which these
executables are run see also Section 3, Tutorial.
5.4.8 Dialog Boxes Associated with Menu 'Results'
This menu offers a variety of graphics to display the simulation results. Graphics include: computed
Hydraulic Heads, computed Concentrations and Computed Velocities.
Menu Item "Hydraulic Heads"
This option produces the following secondary screen displaying the end of simulation period results.
162
-------
Menu Item "Concentrations
For the Hydrocarbon the following graphics is displayed.
For the Oxygen the following results are displayed at the end of a one year simulation.
163
-------
For the Nitrate the following results are displayed at the end of a one year simulation.
Menu Item "Head Prediction Deviations"
This option produces the distribution of the errors between observations and model predictions of the
Hydraulic Heads. Error contours are only displayed at observation times that match the computed
times.
164
-------
Menu Item "Concentration Prediction Deviations"
This option produces the distribution of the errors between observations and model predictions of the
Concentrations of various constituents.
•<- .«^'-J Jlffp^^jftte1'^pK
M Til ~41 ' V :l I | 'I
Menu Item "Velocities"
This menu option displays the velocities as shown below.
•&
165
-------
Menu Item "Observation Wells"
This option displays the computed time series of selected concentrations at specified observation
points.
Menu Item "Engineering Graphs"
This option allows the user to display in a secondary screen the results using an engineering format.
The depth (z-direction) shows the intensity of the computed concentrations.
166
-------
The direct comparison of the computed concentrations of different constituents is also possible by
activating option "New" in the secondary menu "Edit".
Menu Item "AVI Animation"
Finally, the final option of this menu concerns the video animation (AVI) files. Note that the standard
format for Windows digitized video is the Audio-Video Interleaved (AVT) format. An AVT file can be
played in Windows with no additional hardware (of course it will be smoother and faster with a video
accelerator). Now activate option "AVTanimation". This will invoke the animation module. As it
can be seen, a new menu bar appears at the top of the screen. Move to Menu "File" and click on the
option "Open AVT. A dialog box appears on the screen with the list of all available video clips
(.AVI) files. Select the file "HILLAFB.AVI" to obtain the screen shown below. To playback the
video clip showing the simulated migration of hydrocarbons, you only have to click on the "Forward"
play button that appears at bottom left corner of the AVI window.
167
-------
Et voila! The screen comes to life and the video clip stops after a few seconds. The detailed procedure
on how to create this AVI file is given in Section 3, Tutorial. All you need to know at this point is that
the "HTLTAFR.AVT" file was generated from only 4 Bitmap snapshots depicting the simulated plume
at 0.25, 0.5 o.75 and 1 year. These bitmaps were selected and created using the grasping tool
activated from the available tool box in the secondary menu "Results/Concentrations".
5.4.9 Dialog Boxes Associated with Menu 'View'
This menu provides all the options to change the appearance of the screen. In particular it allows the
user to do the following:
Description
Reduce viewing scale
Enlarge viewing scale
Zoom to fit the image in the Window of working
area
Activate resolution of the cursor movement
Menu Option to
Activate
Zoom In
Zoom Out
Zoom to Overview
Snap Cursor
168
-------
Show generated grid
Show box with smart-Icons (toolbox)
Show 3D View window of simulated domain
Show selected raster image in the background
Show selected raster image in gray scale
Printing the Screen/Window
Show Grid
Show Interaction Tools
Show 3D View
Show Base Image
Make Base Image Gray
Print
5.4.10 Dialog Boxes Associated with Menu 'Annotation'
One of the nice features of the Platform is the possibility to graphically create annotations. In fact the
"Annotation" data network allows the user to write his/her notes (Impressions) at a particular location
of the input and output graphics of a particular run. To activate an "Annotation" use the appropriate
tool from the toolbox, and with the mouse click at the desired location for the annotation. Then write
your remarks in the editing dialog box. Automatically your annotation will be linked to the displayed
graphical representation. The options in menu "Annotation" allows you to delete and edit existing
Annotations.
169
-------
5.5 Advanced Topics
5.5.1 Platform Software Architecture
The main objective of the development of the Graphical Platform is to provide the scientist and
engineer the means to work in an interactive computer graphics environment where the remediation
model under consideration (study) is constantly displayed on the screen. The user navigates through
the various modules of the program by means of menus which are always displayed next to the model
abstraction. Menu choices are picked and interaction with the model is performed by pointing with a
mouse. By pointing to the screen, rather than typing commands, a natural dialogue is developed
between the user and the platform.
The key aspect of the Platform protocol is the integration of data management, graphics and
algorithmic routines, into a coherent platform which is flexible and simple to use. Integration is
achieved by layering the various parts of the program. This layering insulates the high level functional
routines from the low level details of data storage and management. This layered approach also
promotes program modularity.
The core of all procedures under the Platform is the file data base which stores all data, files, and
information pertinent to a particular application. This is the repository of all information used in
various parts of the platform (see Figure 5.3). The file data base is only accessible through the data
base access routines. The layer above the data base is a collection of routines which implement the
computational functionality of the program. These are loosely grouped into several categories. These
categories include: Grid editor, Geologic features editor, editor of initial conditions, 2D and 3D
graphics routines, Kriging routines, Scientific Engines.
Encircling the functional routines is the user interface with its process scale operator. This is a
collection of menu drivers and display routines which allow the analyst to interact with the platform.
An important aspect of this part of the platform is that the user need only deal with one interface.
There is a reassuring continuity of display and type of interaction as the analyst moves from one part
of the program to another.
All parts of the program are coordinated by the process scale operator which automates a great portion
of the management chores, and shadows the user's modeling and simulation activities. For the
implementation of this software architecture the Platform uses several sub-directories to manage the
flow of different software operations.
170
-------
Modeling
Features
(wels,
»rivers
Figure 5.3 Layered Structure of the Graphical Platform Software Architecture.
5.5.2 Platform Input of Natural Attenuation Parameters
The specific data that drive a groundwater contaminant migration simulation model are listed in Table
5.2. They address each and everyone of the mechanisms that the model simulates, namely flow
through the porous medium, interaction with surface waters, evapotranspiration losses, drains, other
forcing mechanisms such as wells and recharge, multiple dissolved species plumes, chemical
reactions, and boundary conditions.
171
-------
Table 5.2 Input Data Given Per Strata.
Physical
Process
Flow
Dispersion
Sorption
SurfaceWater
Physical
Parameter
Horizontal
Conductivity
Anisotropy
Vertical
Conductivity
Storage
Coefficient
Water table
Storage factor
Effective
Porosity
Longitudinal
Dispersivity
Transverse
Dispers. ratio
Vertical Dispers.
ratio
Effective Mol.
Diffusion
Bulk Density of
Medium
Linear
Isotherm
Distribution
Coef.
Freundlich
Isotherm
Equilibrium
Constant
Freundlich
Exponent
Langmuir
Isotherm
Equilibrium
Constant
Total
Concentration
Bed Elevation
Bed
Conductivity
Surface
Elevation
Function
Conveyance
Conveyance
Transient
Computations
Intersticial
Velocity
Gradients with
Aquifer
Link with
Aquifer
Gradients with
Aquifer
Link
Link with
Aquifer
DefaultValue
1000.
1.
1000.
0.100
0.009
0.200
0.
1.
1.
0.
0.
0.
0.
1.
0.
0.
1,000
Menu
Domain/Define
Strata
Domain/Define
Strata
Domain/
Chemical Species
172
-------
Physical
Process
Decay
lonExchange
Wells/Drain
Slurry Walls
Discontinuity
Fault
Physical
Parameter
Radioactive
Decay
Half Life (T)
Rate Constant
(1/T)
Biodegradation
Decay
First order rate
for dissolved
phase
Fist order rate for
sorbed phase
Monovalent
Exchange
Ion-exchange
selectivity
coefficient
Ion-exchange
capacity
Divalent
exchange
Monovalent-
Divalent
Divalent-
Monovalent
Drain Elevation
Bed
Conductivity
Wall
Elevation/Thickn
ess
Wall
Conductivity
Fault
Elevation/Thickn
ess
Fault
Conductivity
Function
Gradients with
Aquifer
Link with
Aquifer
Gradients with
Aquifer
Link with
Aquifer
Gradients with
Aquifer
Link with
Aquifer
Link
Link with
Aquifer
Link with
Aquifer
Link with
Aquifer
Default
Value
100
.006
0.
0.
0.
0
1,000
1,000
1,000
Menu
Domain/
Chemical Species
Domain/
Chemical Species
173
-------
Physical
Process
Recharge/E. T.
Evapotranspirati
on
Boundary
Conditions
Initial Conditions
Chemical
Reactions
Physical
Parameter
Time Schedule
of rate
Aquifer Layer
Time Schedule
of rate
Aquifer Layer
Constant Head
No Flux
Boundary
General Head
Boundary
Constant
Concentrations
Initial Head
Initial
Concentrations
Soil Bulk
Density
Function
Aquifer Source
Aquifer Drain
Fix Head
Control Flux
Impose Flux via
conductivity
Fix
Concentrations
Fix Head
Fix
Concentrations
Link
Default
Value
Menu
5.5.3 Sensitivity of Input Parameters
The model input parameters should be subjected to sensitivity analyses to test model response to the
potential range of key parameters. These analyses permit evaluation of the effects on model output
(Concentrations) of varying: hydraulic, hydrologic, hydrogeologic properties, dispersivities, source
loading rates and other parameters within conceivable ranges quantified by the available "Kriging"
procedure.
Each remediation site has its own idiosyncrasies. However, to properly perform a calibration analysis,
there is a need to know the relative effects of these input parameters. Figure 5.4 below provides a
rough estimate of the importance of these parameters in evaluating the contaminant migration
concentration. These estimates allow a quick determination of the parameters needing readjustment
174
-------
during the calibration process. Clearly the parameters with a high influence on the estimated
concentrations must be calibrated first.
Hydraulic p5W)
Conductor.
Figure 5.4 Estimates of Sensitivity Analysis.
5.5.4 Concluding Remarks
This concludes the formal presentation of the Platform and the input parameters that are needed to set
up a groundwater remediation investigation. However, for a more thorough understanding of the
Platform and its use you must also consult the following Sections:
1. Sections, Tutorial
2. Section 4, Theoretical Development, and
3. Appendix B, Intrinsic Remediation Implementation Protocol
175
-------
6.0 References
Anderson, M. P. and W. W. Woessner, 1992. Applied Groundwater Modeling, Academic Press,
San Diego, CA.
Bedient, P. B., H. S. Rifai, and C. J. Newell, 1994. Ground Water Contamination, Transport and
Remediation., PTR Prentice-Hall, Inc., Englewood Cliffs, NJ.
Borden, R. C. and P. B. Bedient, 1986. "Transport of Dissolved Hydrocarbons Influenced By
Oxygen-Limited Biodegradation: 1. Theoretical Development," Water Resources Research,
13:1973-1982.
Borden, R. C., P. B. Bedient, M. D. Lee, C. H. Ward, and J. T. Wilson, 1986. "Transport of
Dissolved Hydrocarbons Influenced by Oxygen-Limited Biodegradation: 2. Field Application,"
Water Resources Research, 13:1983-1990.
Borden, R. C., 1986. "Influence of Adsorption and Oxygen Limited Biodegradation on the
Transport and Fate of a Creosote Plume: Field Methods and Simulation Techniques," Houston,
TX.
Buscheck, T. E., K. T. O'Reilly, and S. N. Nelson, 1993. "Evaluation of Intrinsic Bioremediation
at Field Sites," Chevron Research and Technology Company, Proceedings of the 1993 Petroleum
Hydrocarbons and Organic Chemicals in Ground Water: Prevention, Detection, and Restoration,
Houston, Texas.
Connor, J. A., C. J. Newell, J. P. Nevin, and H. S. Rifai, 1994. "Guidelines for Use of
Groundwater Spreadsheet Models in Risk-Based Corrective Action Design," Proceedings of
NGWA Pet. Hydro. Conf, Houston, TX, November 1994.
Cooper, R. M., and J. D. Istok, 1988. "Geostatistics Applied to Groundwater Contamination.
I: Methodology," Journal of Environmental Engineering, 114(2):270-286.
Davis, J. W., N. J. Kliker, and C. L. Carpenter, 1994. "Natural Biological Attenuation of
Benzene in Ground Water Beneath a Manufacturing Facility," Ground Water, Vol. 32, No. 2, pp.
215-226.
Howard, P. H., R. S. Boethling, W. F. Jarvis, W. M. Meylan, and E. M. Michalenko, 1991.
Handbook of Environmental Degradation Rates, Lewis Publishers, Inc., Chelsea, MI.
Johnson, P. C., M. W. Kemblowski, and J. D. Colthart, 1990a. "Quantitative Analysis of
Cleanup of Hydrocarbon-Contaminated Soils by In-Situ Soil Venting," Ground Water, Vol. 28,
No. 3, May - June, 1990, pp. 413-429.
Johnson, P. C., C. C. Stanley, M. W. Kemblowski, D. L. Byers, and J. D. Colthart, 1990b. "A
Practical Approach to the Design, Operation, and Monitoring of In Site Soil-Venting Systems,"
Ground Water Monitoring and Remediation, Spring 1990, pp. 159-178.
176
-------
Konikow, L. F. and J. D. Bredehoeft, 1978. "Computer Model of Two-Dimensional Solute
Transport and Dispersion in Ground Water," Techniques of Water Resources Investigation of the
United States Geological Survey, Book 7, Reston, VA,
Konikow, L. F. and J. D. Bredehoeft, 1989. "Computer Model of Two-Dimensional Solute
Transport and Dispersion in Ground Water," Techniques of Water Resources Investigation of the
United States Geological Survey, Book 7, Reston, VA.
Monod, J., 1942. Recherches sur la croissance des cultures bacteriennes, Herman & Cie, Paris,
1942.
Newell, C. J., J. W. Winters, H. S. Rifai, R. N. Miller, J. Gonzales, and T. H. Wiedemeier, 1995.
"Modeling Intrinsic Remediation With Multiple Electron Acceptors: Results from Seven Sites,"
National Ground Water Association, Proceedings of the Petroleum Hydrocarbons and Organic
Chemicals In Ground Water Conference, Houston, TX, November 1995, pp. 33-48.
Newell, C. J., R. K. McLeod, and J. R. Gonzales, 1996. BIOSCREEN Natural Attenuation
Decision Support System User's Manual, Version 1.3, EPA/600/R-96/087, August 1996. Robert
S. Kerr Environmental Research Center, Ada, OK.
Oak Ridge National Laboratory, 1989. The Installation Restoration Program Toxicology Guide,
DOE Interagency Agreement No. 1891-A076-A1, Volumes III and IV, July, 1989.
Ollila, P. W., 1996. "Evaluating Natural Attenuation With Spreadsheet Analytical Fate and
Transport Models," Ground Water Monitoring and Remediation, Vol. 16, No. 24, pp. 69-75.
Rifai, H. S. and P. B Bedient, 1990. "Comparison of Biodegradation Kinetics With an
Instantaneous Reaction Model for Groundwater," Water Resources Research, Vol. 26, No. 4, pp.
637-645.
Rifai, H. S., P. B. Bedient, R. C. Borden, and J. F. Haasbeek, 1987. BIOPLUME II Computer
Model of Two-Dimensional Contaminant Transport Under the Influence of Oxygen Limited
Biodegradation In Ground Water, User's Manual, Version 1.0, National Center for Ground
Water Research, Rice University, Houston, TX.
Rifai, H. S., P. B. Bedient, J. T. Wilson, K. M. Miller, and J. M. Armstrong, 1988.
"Biodegradation Modeling at Aviation Fuel Spill Site," Journal of Environmental Engineering,
114(5):1007-1029.
Snoeyink, V. L. and D. Jenkins, 1980. Water Chemistry. John Wiley & Sons, New York.
Stelljes, M. E. and G. E. Watkin, 1993. "Comparison of Environmental Impacts Posed by
Different Hydrocarbon Mixtures: A Need for Site Specific Composition Analysis," in
Hydrocarbon Contaminated Soils and Groundwater, Vol. 3, P. T. Kostecki and E. J. Calabrese,
Eds., Lewis Publishers, Boca Rotan, p. 554.
177
-------
Wiedemeier, T. H., R. N. Miller, J. T. Wilson, and D. H. Kampbell, 1995a. "Significance of
Anaerobic Processes for the Intrinsic Bioremediation of Fuel Hydrocarbons," National Ground
Water Association, Proceedings of the Petroleum Hydrocarbons and Organic Chemicals in
Ground Water Conference, Houston, TX, November 1995.
Wiedemeier, T. H., J. T. Wilson, D. H. Kampbell, R. N. Miller, and J. E. Hansen, 1995b.
"Technical Protocol for Implementing Intrinsic Remediation With Long-Term Monitoring for
Natural Attenuation of Fuel Contamination Dissolved in Groundwater ", Vol. 1, Air Force Center
for Environmental Excellence, Technology Transfer Division, Brooks AFB, San Antonio, TX.
Wilson, J. T., 1994. Presentation at Symposium on Intrinsic Bioremediation of Ground Water,
Denver, CO, August 1-September 1, EPA/600/R-94-162.
178
-------
APPENDIX I. INPUT DATA
The BIOPLUME III input data are listed in detail in Table I.I. The key variables and concepts
used in the model will be described more thoroughly in this section. A number of examples will
be given throughout the section to better illustrate some of the variable definitions.
I.I Discretization of Space
The first step in applying the BIOPLUME III model to a field site involves selecting the size of
the model grid and the number of cells contained within the grid. Four variables are used to define
the selected grid: NX, NY, XDEL and YDEL (Figure 1.1). The number of grid cells in the x- and
y- directions are defined in NX and NY, respectively and the size of the individual cells are
defined in XDEL and YDEL, respectively (see Figure 1.1).
Since the model requires that no-flow boundaries be specified around the site, "extra" cells need
to be incorporated into the grid design. In other words, if an "active" domain of 12 x 12 cells is
needed, a 14 x 14 grid is specified in order to allow for the outer rows and columns to serve as no-
flow boundaries.
There are a number of conventions used in the model which are useful to note. First, flow is
generally along the y-direction. The origin is designated at the upper left-hand corner of the grid
(this means that flow is essentially "down the page"). The x-cells are then counted starting with
1 at the origin and through NX moving to the right of the origin. Similarly, the y-cells are counted
starting with 1 at the origin and through NY moving downwards from the origin (see Figure I.I).
These conventions may be changed but caution needs to be exercised in entering the input data
and analyzing the resulting output to avoid confusion.
1.2 Discretization of Time
BIOPLUME III uses three variables to define simulation time in the model: NTIM, PINT and
NPMP. The relationship between these variables is illustrated in Figure 1.2.
NTIM - is defined as the number of times in a given simulation period that the user may receive
model results.
PINT - is defined as the length of time in a given simulation period.
NPMP - defines the number of pumping periods to be simulated. A pumping period is defined
as a specified length of time during which the hydrologic conditions at the site remain unchanged.
179
-------
Table 1.1. Input Data for BIOPLUME III
Line # Parameter
1 TITLE
2 NTIM
NPMP*
NX
NY
NPMAX
NPNT
NITP
NUMOBS
ITMAX
NREC
NPTPND
NCODES
NPNTMV
NPNTVL
NPNTD
NPDELC
NPNCHV
IREACT
3 PINT
TOL
POROS
BETA
S
TIMX
TINIT
XDEL
YDEL
DLTRAT
CELDIS
ANFCTR
Type
Alphanumeric
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Integer
Real
Real
Real
Real
Real
Real
Real
Real
Real
Real
Real
Real
Definition
Short description of dataset
Maximum number of time steps in a pumping
period
Number of pumping periods
Number of nodes in x direction
Number of nodes in y direction
Maximum number of particles
Time step interval for printing
Number of iteration parameters
Number of observation points
Maximum allowable number of iterations
Number of pumping or injection wells
Initial number of particles per node
Number of node identification code
Particle move interval for printing chemical
output data (Specify 0 to print at end of time step)
Option for printing computed velocities (0 = do
not print; 1 = print for first time step; 2 = print for
all time steps)
Option for printing computed dispersion equation
coefficients (0 = do not print; 1 = print for first
time step; 2 = print for all time steps)
Option for printing computed changes in
concentration 0 = do not print; 1 = print)
Not used
Reaction type specifier for contaminant
Pumping period in years
Convergence criteria for flow equation
Effective porosity
Longitudinal dispersivity in ft
Storage coefficient (S=0 for steady-state flow)
Time increment multiplier for transient flow
(disregarded if S = 0)
Size of initial time step in seconds for transient
flow (disregarded if S = 0)
Width of cell in x direction in ft
Width of cell in y direction in ft
Ratio of transverse to longitudinal dispersivity
Maximum cell distance per particle move
Ratio of Tyy to Txx
Range
1 ... 100
3... 35
3 ...35
NX*NY*NPTPND
4... 7
0... 5
100... 200
0... 50
4, 5, 8, 9 or 16
1 ...9
0, Ior2
0, Ior2
Oorl
-lorO... 7
£0.01
0.01 ... 1
0... 1
180
-------
Table 1.1. Input Data for BIOPLUME III
Line#
Parameter
Type
Definition
Range
One of the following
possibilities depending
on the value oflREACT
All data in this series are for contaminant
IREACT = -1
IREACT = 0
IREACT = 1
IREACT = 2
IREACT = 3
IREACT = 4
IREACT = 5
IREACT = 6
IREACT = 7
5
6
THALF
no line 4
DK, RHOB, THALF
RHOB, EKF, XNF,
THALF
RHOB, EKL, CEC,
THALF
RHOB, EK, CEC,
CTOT, THALF
RHOB, EK, CEC,
CTOT, THALF
RHOB, EK, CEC,
CTOT, THALF
RHOB, EK, CEC,
CTOT, THALF
THALFS
DEC1
Real
Real
Real
Real
Real
Real
Real
Real
Real
Real
Decay half-life for a radioactive compound in
seconds
DK - linear sorption distribution coefficient
(volume/mass), RHOB - aquifer bulk density
(mass/volume). These two parameters need to
have consistent units, for example DK in cc/g and
RHOB in g/cc
EKF - Freundlich sorption coefficient, XNF -
Freundlich sorption exponent (dimensionless)
EKL - Langmuir sorption coefficient
(volume/mass, for example, ml/g), CEC -
Maximum sorption capacity or ion-exhange
capacity (mass/mass, for example, ng/g)
EK - Ion exchange selectivity coefficient (units
depend on stoichiometry), CTOT - total
solution concentration of two exchanging ions
(equivalents/volume)
Variables defined previously
Variables defined previously
Variables defined previously
Source decay half-life in seconds
Lumped decay coefficient for aerobic and
181
-------
Table 1.1. Input Data for BIOPLUME III
Line#
Parameter
Type Definition
Range
IRECO
Integer Biodegradation type specifier for oxygen
One of the following
possibilities depending
on the value of IRECO
IRECO = 0
no line 8
IRECO = 1
DCO, FO, DOMIN
Real
First-order decay rate for aerobic biodegradation
(I/day), stoichiometric ratio of oxygen required
to degrade contaminant, threshold concentration
of oxygen below which biodegradation does not
occur (concentration units, mass/volume, for
example, mg/1)
IRECO = 2
FO, DOMIN
Real
Variables defined previously
IRECO = 3
FO, DOMIN,
CMSO, RMO,
RKHO, RKMAXO,
RKO
Real
One of the following
possibilities depending
onthevalueoflRECN
Start w. CMSO - cone, of microorganisms
(mass/volume), retardation factor for
microorganisms, half-saturation constant for
contaminant (mass/volume), maximum
utilization factor for contaminant (I/days), half-
saturation constant for oxygen (mass/volume)
9
IRECN
Integer
Biodegradation type specifier for nitrate
10
IRECN = 0
no line 10
IRECN = 1
DCN, FN, NMIN
Real
Variables similar to those defined for oxygen
IRECN = 2
FN, NMIN
Real
Variables similar to those defined for oxygen
IRECN = 3
FN, NMIN, CMSN,
RMN, RKHN,
RKMAXN, RKN
Real
Variables similar to those defined for oxygen
11
IRECF
Integer Biodegradation type specifier for iron
182
-------
Table 1.1. Input Data for BIOPLUME III
Line#
Parameter
Type
Definition
Range
12
One of the following
possibilities depending
onthevalueoflRECF
IRECF = 0
no line 12
IRECF = 1
DCF, FF,FMIN
Real
Variables similar to those defined for oxygen
IRECF = 2
FF,FMIN
Real
Variables similar to those defined for oxygen
IRECF = 3
FF, FMIN, CMSF,
RMF,RKHF,
RKMAXF,RKF
Real
Variables similar to those defined for oxygen
13
IRECS
Integer Biodegradation type specifier for sulfate
14
One of the following
possibilities depending
onthevalueoflRECS
IRECS = 0
IRECS = 1
IRECS = 2
no line 14
DCS, FS, SMIN
FS, SMIN
Real Variables similar to those defined for oxygen
Real Variables similar to those defined for oxygen
IRECS = 3
FS, SMIN, CMSS,
RMS, RKHS,
RKMAXS, RKS
Real
Variables similar to those defined for oxygen
15
IRECC
Integer Biodegradation type specifier for carbon dioxide
16
One of the following
possibilities depending
on the value of IRECC
IRECC = 0
no line 16
IRECC = 1
DCC, FC, CMIN
Real
Variables similar to those defined for oxygen
IRECC = 2
FC, CMIN
Real
Variables similar to those defined for oxygen
IRECC = 3
FC, CMIN, CMSC,
RMC, RKHC,
RKMAXC, RKC
Real
Variables similar to those defined for oxygen
Data Set 1 IXOBS
# of lines = NUMOB S I YOB S
Integer x coordinate for observation points
Integer y coordinate for observation points
183
-------
Table 1.1. Input Data for BIOPLUME III
Line # Parameter
Data Set 2 IX
# of lines = NREC IY
REC
CNREC
CNREC1
CNREC2
CNREC3
CNREC4
CNREC5
Data Set 3 INPUT
# of lines = 1 or NY FCTR
VPRM
Type
Integer
Integer
Real
Real
Real
Real
Real
Real
Real
Integer
Real
Real
Definition Range
x coordinate of pumping or injection wells
y coordinate of pumping or injection wells
Pumping (+) or injection (-) rate in cfs
Concentration of injected contaminated water
(mass/volume)
Concentration of injected oxygenated water
(mass/volume)
Concentration of injected nitrate water
(mass/volume)
Concentration of injection iron water
(mass/volume)
Concentration of injected sulfate water
(mass/volume)
Concentration of injected carbon dioxide water
(mass/volume)
Parameter card for transmissivity . If 0 then a
constant transmissivity is specified in FCTR. If 1 0 or 1
then transmissivity is read from an array
Constant transmissivity value in sq ft/sec OR
factor to multiply transmissivity array
Array of transmissivity data in sq ft per sec
Data Set 4
INPUT
Parameter card for thickness. If 0 then a constant
Integer thickness is specified in FCTR. If 1 then
thickness is read from an array
Oorl
# of lines = lor NY
FCTR
THCK
Real
Constant thickness value in ft OR factor to
multiply thickness array
Real
Array of thickness data in ft
184
-------
Table 1.1. Input Data for BIOPLUME III
Line#
Parameter
Type
Definition
Range
Data Set 5
INPUT
Parameter card for recharge. If 0 then a constant
Integer recharge is specified in FCTR. If 1 then recharge
is read from an array
Oorl
# of lines = lor NY
FCTR
RECH
Real
Constant recharge (-) or discharge (+) value in
ft/sec OR factor to multiply recharge array
Real
Array of recharge (-) or discharge (+) data in ft
per sec
Data Set 6
INPUT
Parameter card for node identification. If 0 all
Integer nodes are identified by FCTR. If 1 then node
identification is read from an array
Oorl
# of lines = lor NY FCTR
NODEID
Data Set 7 ICODE
# of lines = NCODES FCTR1
FCTR2
FCTR2O
FCTR2N
FCTR2F
FCTR2S
FCTR2C
FCTR3
OVERRD
Real
Integer
Integer
Real
Real
Real
Real
Real
Real
Real
Real
Integer
Node identification OR factor to multiply node
identification array
Array of node identification data
Instructions for using the NODEID array. When
NODEID = ICODE, then the following factors
are set
Leakance
Concentration of contaminated water
Concentration of oxygenated water
Concentration of nitrate water
Concentration of iron water
Concentration of sulfate water
Concentration of carbon dioxide water
Recharge (-) or discharge (+)
If OVERRD=Om then the value of RECH is not
changed. If OVERRD is nonzero, then the value
ofRECHissettoFCTRS
Data Set 8
# of lines = 1 or NY
INPUT
FCTR
WT
Parameter card for water table. If 0 then a
constant water table is specified in FCTR. If 1
then water table is read from an array
Constant water table value in ft OR factor to
multiply water table array
Array of water table data in ft
Oorl
185
-------
Table 1.1. Input Data for BIOPLUME III
Line#
Parameter
Type
Definition
Range
Data Set 9
INPUT
Parameter card for initial contaminant
concentration. If 0 then a constant initial
concentration is specified in FCTR. If 1 then
initial contaminant concentration is read from an
array
Oorl
# of lines = 1 or NY
FCTR
Constant initial contaminant concentration value
(mass/volume) OR factor to multiply initial
contaminant concentration array
CONC
Array of initial contaminant concentration data
(mass/volume)
Data Set 10
INPUT
Parameter card for initial oxygen concentration.
If 0 then a constant initial concentration is
specified in FCTR. If 1 then initial oxygen
concentration is read from an array
Oorl
# of lines = lor NY
FCTR
CONC1
Constant initial oxygen concentration value
(mass/volume) OR factor to multiply initial
oxygen concentration array
Array of initial oxygen concentration data
(mass/volume)
Data Set 11
INPUT
Parameter card for initial nitrate concentration. If
0 then a constant initial concentration is specified
in FCTR. If 1 then initial nitrate concentration is
read from an array
Oorl
# of lines = 1 or NY
FCTR
CONC2
Constant initial nitrate concentration value
(mass/volume) OR factor to multiply initial
nitrate concentration array
Array of initial nitrate concentration data
(mass/volume)
186
-------
Table 1.1. Input Data for BIOPLUME III
Line#
Parameter
Type
Definition
Range
Data Set 12
INPUT
Parameter card for initial ferrous iron
concentration. If 0 then a constant initial iron
concentration is specified in FCTR. If 1 then
initial ferrous iron concentration is read from an
array
Oorl
# of lines = 1 or NY
FCTR
CONC3
Constant initial ferrous iron concentration value
(mass/volume) OR factor to multiply initial
ferrous iron concentration array
Array of initial ferrous iron concentration data
(mass/volume)
Data Set 13
# of lines = lor NY
INPUT
FCTR
CONC3A
Parameter card for initial ferric iron
concentration. If 0 then a constant initial ferric
iron concentration is specified in FCTR. If 1 then
initial ferric iron concentration is read from an
array
Oorl
Constant initial ferric iron concentration value
(mass/volume) OR factor to multiply initial
ferric iron concentration array
Array of initial ferric iron concentration data
(mass/volume)
Data Set 14
INPUT
Parameter card for initial sulfate concentration.
If 0 then a constant initial concentration is
specified in FCTR. If 1 then initial sulfate
concentration is read from an array
Oorl
# of lines = 1 or NY
FCTR
Constant initial sulfate concentration value
(mass/volume) OR factor to multiply initial
sulfate concentration array
CONC4
Array of initial sulfate concentration data
(mass/volume)
Data Set 15
INPUT
Parameter card for initial carbon dioxide
concentration. If 0 then a constant initial
concentration is specified in FCTR. If 1 then
initial carbon dioxide concentration is read from
an array
# of lines = lor NY
FCTR
CONC5
Constant initial carbon dioxide concentration
value (mass/volume) OR factor to multiply
initial carbon dioxide concentration array
Array of initial carbon dioxide concentration data
(mass/volume)
187
-------
Table 1.1. Input Data for BIOPLUME III
Line#
Parameter
Type
Definition
Range
Data Set 16
# of lines = 1
ICHK
Parameter to check whether data will be revised
for subsequent pumping periods (1 means data
will be revised, and remainder of data are
specified; 0 means no revisions will be made and
data from previous pumping period are used)
# of lines = 1
NTIM
NPNT
NITP
ITMAX
NREC
NPNTMV
NPNTVL
NPNTD
NPDELC
NPNCHV
PINT
TIMX
TINIT
Previously defined variables (see lines 1 and 3)
# of lines = NREC
IX
IY
REC
CNREC
CNREC1
CNREC2
CNREC3
CNREC4
CNREC5
Previously defined variables (see dataset 2)
Notes:
* - If NPMP > 1, then data set 16 must be completed
188
-------
' Origin
NX = 12
10
11
12
13
10 11 12
Direction of
-Groundwater
Flow
•No Flow Cells
Figure 1.1. Grid Discretization in Bioplume III
189
-------
Real
Time
1970
2000
Tank
installed
1985 1987 1990
Contamination Tank Pump-and-
identified removed treat system
installed
Model
Definitions
1st pumping period
2nd pumping
period
1—I—I—\-
3rd pumping
period
H—I—I—I—I—I—I—I—h
PINT =1-17 years
NTIM = 1
PINT = 3 years
NTIM = 3
PINT= 10 years
NTIM = 10
[NPMP = 3]
Figure 1.2. Discretization of Time in BIOPLUME III
190
-------
Example I.I:
An underground storage tank was installed at Site A in 1970. A ground water contamination problem
was subsequently discovered in 1985. The underground storage tank was removed in 1987 and a pump-
and-treat remediation system was installed at the site in 1990 (see Figure I.I). Ground water
monitoring was undertaken quarterly throughout the period from 1985 to present day. The modeling
objective for this site is to determine the status of the ground water plume in the year 2000.
In order to simulate the period from 1970 to 2000 (total of 30 years), three distinct pumping periods
(NPMP = 3) should be used:
1st Pumping Period: 1970 - 1987. This period is characterized by the leaking event which may have
taken place any time after the tank was installed (in general, tank failure occurs within a
period of 7 - 10 years after installation). The simulation time for this period (PINT) may be
anywhere from 1 to 17 years depending on when it is assumed that the tank began leaking. The
number of time steps (NTIM) will be one because data were only collected in the last three years
of the period (1985 to 1987).
2nd Pumping Period: 1987 through 1990. This period is characterized by the removal of the leaking
tank and therefore "cutting off" the source of ground water contamination. The simulation time
(PINT) is equal to 3 years and NTIM can be anywhere from 1 to 12 (since ground water data are
collected every three months in the three year period). In general, however, it is not efficient
to run the model on a quarterly basis because of the possibly long run times, large amounts of
data for analysis, and the lack of necessity for that much resolution in model results. An NTIM
of 1 or 3 is suggested in this case.
3rd Pumping Period: 1990 through 2000. This period is characterized by the installation of the pump-
and-treat system. The simulation time (PINT) is equal to 10 years and an NTIM of 1 or 10 is
suggested (this allows viewing of the model results in the year 2000 or annually, respectively).
Note: The BIOPLUME III model internally calculates a "computational time step" to minimize
the transport mass balance errors. The "computational time step" can be manipulated by the
user to improve the mass balance error or to shorten run times (see Sections 1.9, A.3 and A.7).
1.3 Hydrogeologic Characteristics of the Aquifer
A number of variables are used in BIOPLUME III to identify the hydrogeologic characteristics of
the aquifer. These include: porosity (POROS), longitudinal dispersivity (BETA), storativity
(S), ratio of transverse to longitudinal dispersivity (DLTRAT), ratio of longitudinal
transmissivity to transverse transmissivity (ANFCTR), transmissivity (VPRM), recharge
(RECH) and thickness of the aquifer (THCK).
POROS - (effective porosity) is the dimensionless ratio of the volume of interconnected voids to
the bulk volume of the aquifer solids. The porosity is obtained from site specific measurements
or from literature values (see Table 1.2).
BETA - (longitudinal dispersivity) defines the longitudinal spreading of a plume in the direction
of flow. Selection of dispersivity values is difficult because of the impracticability of measuring
dispersion in the field. Values for BETA may be obtained using:
191
-------
Table 1.2. Effective Porosity Estimates
Media Porosity
Gravel, fine .25 - .38
Sand, coarse .31 - .46
Sand, fine .26 - .53
Silt .34 - .61
Clay .34 - .60
Sandstone .005-.10
Limestone .001 - .05
Source: Domenico and Schwartz, 1990.
192
-------
• Data compiled from 50 sites by Gelhar et al. (1985) shown in Figure 1.3;
Data from recent field studies as shown in Table 1.3, or;
• Using the relationship suggested by Pickens and Grisak (1981):
BETA = 0.1 Lp, where Lp is the plume length (see Figure 1.4) or distance to
measurement point in ft.
S - (storativity) is the product of the specific storage and the thickness of the aquifer, where the
specific storage is defined as the volume of water that a unit volume of aquifer releases from
storage when the pressure head in the unit volume changes a unit amount. Storativity is only
used for transient flow analyses and is estimated from pump tests conducted at the site.
DLTRAT - is the ratio of transverse to longitudinal dispersivity. Much like the longitudinal
dispersivity, this variable is difficult to estimate. The data in Table 1.3 or one of the following
relationships may be used:
DLTRAT = 0.33 (Gelhar et al., 1992)
DLTRAT = 0.1 (U. S. Environmental Protect!on Agency, 1986)
ANFCTR - (ratio of longitundinal transmissivity to transverse transmissivity) is rarely
characterized at field sites and is mostly set to 1.
VPRM - (transmissivity) is the product of the hydraulic conductivity and the thickness of the
aquifer. VPRM can be obtained from slug tests or pump tests at the site or from published
literature values for the hydraulic conductivity (Figure 1.5).
RECH - (recharge) is generally obtained from rainfall data and soil infiltration characteristics.
This variable is rarely, if ever, measured at field sites. It is usually estimated from local or
regional data published by the USGS and the Soil Conservation Service or calibrated.
THCK - (aquifer thickness) is generally obtained from well logs, soil borings and other geologic
characterization efforts at the site.
Note: The last three parameters: VPRM, RECH and THCK, may be specified as a constant for
the whole site or as a spatially variable parameter such that a different value is entered for each
cell in the model grid.
1.4 Boundary Conditions
In order to simulate a field site with the BIOPLUME III model, it is necessary to identify the
hydrogeologic conditions that prevail around the site. These hydrogeologic conditions are
referred to as boundary conditions. The two types of boundary conditions that can be simulated
with BIOPLUME III include: constant head and constant flux.
193
-------
10
103
102
10
10°
-1
10
10-2
10-3
Longitudinal Dispersivity
= 10% of scale
Q longitudinal Dispersivity
= 0.83 [LoglO (scale)]
2.414
RELIABILITY
o
o° o
o Low
O Intermediate
O High
I I I Mill II III Mil I I III Mil I I III Mil I I III Mil I I III Mil I Mill
1Q'1 10° IO1 IO2 IO3 11
Scale (m)
io5 ID'
Note: Data includes Gelhar's reanalysis of several dispersivity studies
Size of circle represents general reliability of dispersivity estimates.
Location of 10% of scale linear relationship plotted as dashed line
(PickensandGrisak, 1981)
Reference: Gelharetal., 1985
Figure 1.3. Longitudinal Dispersivity Chart
194
-------
Table 1.3. Dispersivity Estimates from Field Experiments
Site
Longitudinal Transverse Vertical
dispersivity (m) dispersivity (m) dispersivity (m)
Canada Base,
Borden,
Ontario
.49
.039
.023
MADE Site,
Columbus,
Mississippi
9.5
2.2
NE
Cape Cod,
Massachusetts
.96
.018
.0015
NE = Not Estimated
Sources: Boggs, et. al., 1992
LeBlanc, Garabedian, et al., 1991
Garabedian, Gelhar, et al., 1991
Freyberg, 1986
Mackay, et. al., 1986
195
-------
Source
Center of mass of plume
Contour depicting MCL
or detection limit cone.
of contaminants
Lp - Plume length
Figure 1.4. Illustration of Plume Length for Estimating
Longitudinal Dispersivity
196
-------
Unconsolidated
Rocks
--
c
;
i
i
-f
L> 1
U
3
|
.§ 1
12 TO
~ a
c^ ^
I
C
fc,
4
\i
H
3
H •
W
n ^
deposits
»-
^ <-*
0 P
U 1H
r
T
|
«2
§1
3 e
ij $
*%
M
H ^H
•a
C3
u
U
a
c
4.
0
•>
rt
O
a
'C
5s
e
T:
p
a
|
t7
n
n
>
5
H
4
5
'c
k k K K K
(darcy) (cm2) (cm/s) (m/s) (gal/d/ft2)
. - « . . i
1
o
"F
c
?
1
c_
5
»%
5
j
-10J
-104
-103
-io2
-10
-1
-10-1
-10-2
-10-3
-10-4
-10-5
-IO-6
-10-7
L10-8
-10'J
-lO-4
-10-5
_ 10'6
-lO-7
- 10'8
-10-9
- 10-10
- 10-11
- 10-12
- 10-13
- 10-14
- 10-15
- 10-16
- ioz
- 10
- 1
_io-i
-10-2
- lO-3
-10-4
-10-5
- 10'6
-10-7
- io-8
-lO-9
- 10-1°
L lo-n
- 1
r io6
-10-1
-10-2
- lO-3
- lO-4
- lO-5
- io-6
-10-7
_ 10'8
_10-9
- 10-1°
-lo-ii
- 10-12
- 10-13
-105
-io4
-103
-io2
- 10
- 1
-10-1
-lO-2
-10-3
-lO-4
-10-5
- 10'6
-10-7
Source: Freeze and Cherry, 1979
Figure 1.5. Hydraulic Conductivity for Different Types of Soils
197
-------
Constant head boundaries refer to cells where the water level is constant throughout the
simulation. The head or water level value is specified by the user at the constant head
boundaries.
Constant flux boundaries, on the other hand, refer to cells that allow water (and possibly
contaminants and electron acceptors) to flow through. In this case, the user specifies the rate of
water flow (or flux of water) through the cell and specifies whether the cell is also a source of
contamination or electron acceptor(s).
A number of variables in BIOPLUME III allow the user to specify the boundary conditions for
the site. These include: NCODES, the NODEID matrix, and the parameters ICODE, FCTR1,
FCTR2, FCTR2O, FCTR2N, FCTR2F, FCTR2S, FCTR2C, FCTR3, and OVERRD.
The NCODES variable is used to define the number of boundary condition types to be used in
the model. For example, if constant head boundary conditions without chemical concentration
inflow are to be used for the site, then the NCODES variable is set to one. If, on the other hand,
constant head boundaries without chemical concentration inflow are to be used in one portion of
the site and constant head boundaries with chemical concentrations inflow are to be used in
another portion of the site, then the NCODES variable is set to two.
The NODEID matrix is used to specify the cells at which the boundary conditions will be
designated. The NODEID matrix can be thought of as an ON/OFF switch designator. If
BIOPLUME III encounters a number between 1 to 9 at any of the cells, the model interprets that
as "an ON switch" for additional boundary condition information. BIOPLUME III anticipates
that more data would be provided for those cells. The data include the variables ICODE,
FCTR1, FCTR2, FCTR2O, FCTR2N, FCTR2F, FCTR2S, FCTR2C, FCTR3, and OVERRD.
Additionally, if constant head boundaries are used at any of the cells, the water table or WT
variable needs to be specified for those cells.
198
-------
Example 1.2:
Figures 1.6 and 1.7 illustrate two different hydrogeologic scenarios at given field sites. What types of
boundary conditions can be used to simulate these conditions?
Site A - The water level conditions at this site are best represented using constant head nodes. The
background electron acceptors are specified as input through the boundaries and the contaminant
concentrations are specified as input through some of the cells in the upgradient boundary. The
measured water levels at the boundaries are used as input water table elevations (WT). The resulting
parameter set-up is as follows:
NCODE = 2
NODEID Matrix:
00000000000000
01111222211110
00000000000000
00000000000000
00000000000000
00000000000000
00000000000000
00000000000000
00000000000000
00000000000000
00000000000000
00000000000000
01111111111110
00000000000000
ICODE, FCTR1, FCTR2, FCTR2O, FCTR2N, FCTR2F, FCTR2S, FCTR2C, FCTR3, and OVERRD
1, 1,0,8, 10,0,20,18,0,0
2, 1,100,8,10,0,20, 18,0,0
WT Matrix
11
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
199
-------
10 11 12
2
3
4
10
11
12
13
14
Background electron acceptor concentrations:
Oxygen = 8 mg/L
Nitrate = 10 mg/L
Ferric Iron = 0 mg/L
Sulfate = 20 mg/L
CO2 =18 mg/L
100 -
v
\
94'-
\
\
G
rouncK
vater 1
evel
•
K
\
Cor
Sourc
(e.g. c
stant 1
*
3 Of CC
ontam
lead IK
ntamii
inated
3des
lation
soils)
Figure 1.6. Hydrogeologic Conditions for Site A
200
-------
10 11 12
123456789
Source of contamination
(e.g. contaminated soils)
roundwater level
Leaking Pond
Constant 1
ead nodes
14
Figure 1.7. Hydrogeologic Conditions for Site B
201
-------
Site B - The water levels at this site are also represented with constant head boundaries. The pond in
the middle of the site can be represented using three options: (1) constant head nodes with the
corresponding water level in the lake being specified for the appropriate cells; (2) recharge nodes in
the NODEID matrix; or leakance cells in the NODEID matrix.
Option 1 - Constant Head Nodes for Pond
NCODE = 2
NODEID Matrix:
00000000000000
01111111111110
00000000000000
00000000000000
00000000000000
00000222200000
00000222200000
00000222200000
00000222200000
00000000000000
00000000000000
00000000000000
01111111111110
00000000000000
ICODE, FCTR1, FCTR2, FCTR2O, FCTR2N, FCTR2F
1, 1,0,8, 10,0,20,18,0,0
2, 1,0,0,0,0,0,0,0,0
WT Matrix
11
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
101
101
101
101
0
0
0
94
0
0
100
0
0
0
101
101
101
101
0
0
0
94
0
0
100
0
0
0
101
101
101
101
0
0
0
94
0
0
100
0
0
0
101
101
101
101
0
0
0
94
0
FCTR2S, FCTR2C, FCTR3, and OVERRD
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
202
-------
Option 2 - Recharge Cells for Pond
NCODE = 2
NODEID Matrix:
00000000000000
01111111111110
00000000000000
00000000000000
00000000000000
00000222200000
00000222200000
00000222200000
00000222200000
00000000000000
00000000000000
00000000000000
01111111111110
00000000000000
ICODE, FCTR1, FCTR2, FCTR2O, FCTR2N, FCTR2F
1, 1,0,8, 10,0,20,18,0,0
2, 1,0, 0, 0, 0, 0, 0, -l.Oe-7, 1
WT Matrix
11
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
FCTR2S, FCTR2C, FCTR3, and OVERRD
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
203
-------
Option 3 - Leakance Cells for Pond
NCODE = 2
NODEID Matrix:
00000000000000
01111111111110
00000000000000
00000000000000
00000000000000
00000222200000
00000222200000
00000222200000
00000222200000
00000000000000
00000000000000
00000000000000
01111111111110
00000000000000
ICODE, FCTR1, FCTR2, FCTR2O, FCTR2N, FCTR2F,
1, 1,0,8, 10,0,20,18,0,0
2,-1.0e-9, 0,0, 0,0, 0,0,0, 0
WT Matrix
11
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
Note that if the
injection
wells.
0
100
0
0
0
0
0
0
0
0
0
0
94
0
lake
0
100
0
0
0
0
0
0
0
0
0
0
94
0
was
0
100
0
0
0
0
0
0
0
0
0
0
94
0
leaking
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
0
100
0
0
0
0
0
0
0
0
0
0
94
0
contaminants, the
0
100
0
0
0
0
0
0
0
0
0
0
94
0
lake
FCTR2S, FCTR2C, FCTR3, and OVERRD
0
100
0
0
0
0
0
0
0
0
0
0
94
0
can
0
100
0
0
0
0
0
0
0
0
0
0
94
0
000
100 100 0
000
000
000
000
000
000
000
000
000
000
94 94 0
000
additionally be represented using
Note:
A related variable to boundary conditions is the hydraulic gradient measured at the site. The
BIOPLUME III model generates water level information throughout the site that should "mimic"
the measured water levels. Therefore, hydraulic gradients observed at the site should be similar
to those generated by the model.
204
-------
1.5 Initial Conditions
The head and concentrations at the start of the simulation period need to be specified in the
BIOPLUME III input. The specific variables include: initial water table (WT), initial
concentration of contaminants (CONC), oxygen (CONC1), nitrate (CONC2), ferrous iron
(CONC3), ferric iron (CONC3 A), sulfate (CONC4), and carbon dioxide (CONC5).
The initial water table (WT) may be developed by contouring water level data measurements and
entering the resulting values into each cell in the model grid. This is however, time consuming and
not entirely necessary since the model will recompute the water table anyway. The user can
enter "0" for the initial water table elevation everywhere except where constant head nodes have
been specified (the actual water level is entered for those).
The initial concentration of contaminants (CONC) and the initial concentrations for all the
electron acceptors (CONC1 through CONC5) are determined from monitoring well data.
Note: The BIOPLUME III model does not require specific units for concentration. The user
may select between mg/L and mg/L. The model doesrequire that the user use a consistent set of
units for all the concentration input. Therefore, if mg/Lfor example are to be used, then all the
concentration data need to be entered in that systems of units. The output concentrations
generated by the model will also reflect the same units as the input.
1.6 Sources and Sinks
The introduction of water or release of water from the aquifer (including contaminants and
electron acceptors) is referred to as sources and sinks. These can be simulated using
injection/pumping wells, recharge/discharge cells or constant head cells. The use of recharge and
constant head nodes to represent sources and sinks has been illustrated in the previous section.
This section will focus on the use of injection wells to represent sources and/or pumping and
injection scenarios. The model parameters for injection/pumping wells include: NREC, REC,
CNRECH, CNRECO, CNRECN, CNRECF, CNRECS, and CNRECC.
NREC - defines the number of injection or pumping wells that will be used in the model input.
REC - specifies the injection (-ve) or pumping rate (+ve) for each of the wells.
CNRECH - specifies the concentration of contaminant in injected water (parameter set to 0 for
pumping wells).
CNRECO - specifies the concentration of oxygen in injected water (parameter set to 0 for
pumping wells).
CNRECN - specifies the concentration of nitrate in injected water (parameter set to 0 for
pumping wells).
205
-------
CNRECF - specifies the concentration of ferrous iron in injected water (parameter set to 0 for
pumping wells).
CNRECS - specifies the concentration of sulfate in injected water (parameter set to 0 for
pumping wells).
CNRECC - specifies the concentration of carbon dioxide in injected water (parameter set to 0 for
pumping wells).
1.7 Sorption, Source Decay, Radioactive Decay and Ion-
Exchange Variables
A number of variables are used in the model to represent source decay, radioactive decay,
sorption and ion-exchange reactions. The parameter IREACT is used to designate which of these
reactions are to be used in the current simulation:
IREACT REACTION PARAMETERS TO BE SPECIFIED
-1 decay only THALF
0 no reaction None
1 linear sorption DK, RHOB, THALF
2 Freundlich sorption RHOB, EKF, XNF, THALF
3 Langmuir sorption RHOB, EKL, CEC, THALF
4 monovalent exchange RHOB, EK, CEC, CTOT, THALF
5 divalent exchange RHOB, EK, CEC, CTOT, THALF
6 mono-divalent exchange RHOB, EK, CEC, CTOT, THALF
7 di-monovalent exchange RHOB, EK, CEC, CTOT, THALF
THALF is the decay half-life used for radioactive compounds. This half-life is applied both to
the specified source concentration and to the dissolved concentrations in the model.
RHOB - is the aquifer bulk density. Typical values are included in Table 1.4.
DK - is the linear sorption distribution coefficient more typically referred to as Kd. The
distribution coefficient is generally computed using the following relationship:
Kd = Koc • foc where Koc is the normalized distribution coefficient (see Table 1.5 for
typical values) and foc is the fraction of organic carbon found in uncontaminated soils at
the site. The variable foc can be determined from laboratory analyses of the soils at the
site or using the typical values listed in Table 1.4.
EKF - is the Freundlich sorption coefficient.
XNF - is the Freundlich sorption exponent.
EKL - is the Langmuir sorption coefficient.
206
-------
Table 1.4. Typical Bulk Densities and foc Values (part 1 of 2)
Representative Values of Dry Bulk Density for Common Aquifer Matrix Materials
Aquifer Matrix Dry Bulk Density (gm/cm3)
Clay 1.00-2.40
Glacial Sediments 1.15-2.10
Loess 0.75-1.60
Fine Sand 1.37-1.81
Medium Sand 1.37-1.81
Coarse Sand 1.37-1.81
Gravely Sand 1.37-1.81
Fine Gravel 1.36-2.19
Medium Gravel 1.36-2.19
Coarse Gravel 1.36-2.19
Sandstone 1.60-2.68
Shale 1.54-3.17
Limestone 1.74-2.79
Granite 2.24 - 2.46
Basalt 2.00 - 2.70
Sources: Walton, 1988
Domenico and Schwartz, 1990
207
-------
Texture
Table 1.4. Typical Bulk Densities and foc Values (part 2 of 2)
Representative Values of Total Organic Carbon for Common Sediments
Depositional Environment Fraction Organic Carbon Site Name
medium sand
fine sand
fine to coarse sand
organic silt and peat
silty sand
silt with sand, gravel and
(glacial till)
medium sand to gravel
loess (silt)
fine - medium sand
fine to medium sand
fine to coarse sand
sand
coarse silt
medium silt
fine silt
silt
fine sand
medium sand to gravel
fluvial-deltaic
back-barrier (marine)
glacial (lacustrine)
glaciofluvial
c&iycial moraine
glaciofluvial
eolian
glaciofluvial or
glaciolacustrine
glaciofluvial
glaciofluvial
fluvial
fluvial
fluvial
fluvial
lacustrine
glaciofluvial
glaciofluvial
0.00053
0.0006 -
0.00026
0.10-0.
0.0007 -
0.0017 -
0.00125
0.00058
- 0.0012
0.0015
- 0.007
25
0.008
0.0019
- 0.0016
< 0.0006 - 0.0061
0.00021
0.00029
0.0057
0.029
0.020
0.0226
0.0011
0.00023
0.00017
-0.019
- 0.073
- 0.0012
- 0.00065
Hill AFB, Utah c/
Boiling AFB, D.C. c/
Patrick AFB, Florida c/
Elmendorf AFB, Alaska c/
Elmendorf AFB, Alaska c
Elmendorf AFB, Alaska c
Elmendorf AFB, Alaska c
Offutt AFB, Nebraska c/
Truax Field, Madison,
Wisconsin c
King Salmon AFB, Fire
Training Area, Alaska c
Dover AFB, Delaware c
Battle Creek ANGB,
Michigan c
Oconee River, Georgia a
Oconee River, Georgia a
Oconee River, Georgia a
Oconee River, Georgia a
Wildwood, Ontario b/
Various sites in Ontario
Various sites in Ontario
a/Karickhoff, 1981
b/ Domenico and Schwartz, 1990
c/ Wiedemeier et al., 1995b
208
-------
Table 1.5. Typical Distribution Coefficients
Compound KOC
(cm3/g)
Benzene 83
Toluene 300
Ethylbenzene 1100
Xylene 240
Source: Texas Natural Resource Conservation
Commission, 1994.
209
-------
CEC - is the maximum sorption capacity or ion exchange capacity.
EK - is the Ion-exchange selectivity coefficient.
CTOT - is the total solution concentration of two exchanging ions.
THALFS - represents the source decay rate.
1.8 Biodegradation Variables
A number of variables are used in BIOPLUME III to simulate the aerobic and anaerobic
biodegradation of contaminants. An overall first-order decay biodegradation rate (DEC1) can be
designated to simulate the lumped effect of aerobic and anaerobic processes. Alternatively,
detailed information about the electron acceptors may be provided to simulate their individual
impacts. A biodegradation type specifier for each of the electron acceptors (IRECO, IRECN,
IRECF, IRECS, IRECC) is used to select between the first-order, instantaneous, and Monod
kinetic models. The electron acceptor data for oxygen, for example, depends on the selected
kinetic model:
First-Order Decay Simulations:
DCO - is the first-order decay rate for oxygen.
FO - is the concentration of available oxygen in the ground water. This value is needed
to allow the model to decay hydrocarbons as long as oxygen is present in the
aquifer. If oxygen concentrations reach their specified threshold concentrations
(DOMIN), the biodegradation reaction is terminated. This ensures that the first-
order decay model does not overestimate the amount of biodegradation that is
likely to occur in the aquifer.
DOMIN - is the threshold oxygen concentration.
Instantaneous Reaction Simulations:
The variables FO and DOMIN (defined previously) are required.
Monod Kinetic Simulations:
In addition to FO and DOMIN (as in the case of the first-order model, these parameters
are provided to ensure this model does not overestimate the amount of biodegradation
that is likely to occur in the aquifer), the following parameters are required:
CMSO - is the concentration of microorganisms in the aquifer.
RMO - is the microbial retardation coefficient.
RKHO - is the half-saturation constant for the contaminant.
RKMAXO - is the maximum growth rate for the contaminant.
RKO - is the half-saturation constant for oxygen.
Similar variables are defined for all the electron acceptors.
210
-------
1.9 Numerical Parameters
A number of variables used in BIOPLUME III define "numerical parameters" needed in the
solution method used by the model. These include: NPTPND, NPMAX, CELDIS, NITP,
ITMAX, TOL, TIMX, and TINIT
NPTPND is the number of particles to be used in each cell in the model. The number of particles
used in each cell will impact the runtime required for the model. A smaller number of particles
will allow the model to run in a shorter period of time but may increase the mass balance errors in
the simulation. In general, 9 particles provide adequate model accuracy without causing
excessively long runtimes.
NPMAX is the maximum number of particles for the whole grid. In general, NPMAX should be
set to a number greater than NX»NY»NPTPND.
CELDIS defines the maximum allowable distance within the cell that a particle is allowed to
move in a time step. A CELDIS of 0.5 implies that a particle is not allowed to move more than
half a cell length during the time step. This variable is needed in order to control the movement of
the particles and the mass balance errors in the model (see Section A.3 in Appendix A). A
smaller CELDIS causes lower numerical mass balance errors but may increase runtimes. In
general, a CELDIS of 0.5 is recommended.
NITP is the number of iteration parameters used in solving the flow equation. A value of 7 is
recommended for this variable.
ITMAX is the maximum number of iterations to be used in solving the flow equation. A value of
200 (the maximum) is recommended for this variable. If the model is unable to arrive at a solution
of the flow equation using this value, an error message will be generated and the model run will be
terminated. In this case it is recommended that the user review the input data for possible errors.
TOL is the convergence criterion that is used to iteratively solve the flow equation. A value £
0.001 is recommended.
TIMX and TINIT define the time steps for transient simulations. TINIT is the size of the initial
time step, and TIMX is the multiplier that will be used to generate subsequent time steps from
TINIT. For example, if TINIT is set to 1000 seconds and TIMX is set to 10, the second time
step will be 10 x 1000 = 10,000 seconds, the third time step will be 10,000 x 10 = 100,000
seconds, and so on.
1.10 Output Control Parameters
A number of variables in BIOPLUME III can be used to control the amount of output that is
generated by the model. These include: NPNT, NPNTMV, NPNTVL, NPNTD, NPDELC, and
NPNCHV. The majority of these parameters, except for NPNT, are typically set to "0. NPNT
is usually set to "1" to allow viewing of the output at the end of the time step.
211
-------
Other variables are used to control the type of output that can be generated by the model. These
include variables that designate the number and location of observation points: NUMOBS,
IXOBS and IYOBS
NUMOBS is the number of observation or monitoring wells to be specified. A maximum of 5 is
allowed.
IXOBS and IYOBS define the locations of the specified number of monitoring wells.
1.11 References
Boggs, J. M., S. C. Young, L. M. Beard, L. W. Gelhar, K. R. Rehfeldt, and E. E. Adams, 1992.
"Field Study of Dispersion in a Heterogeneous Aquifer. 1. Overview and Site Description,"
Water Resources Research, Vol. 28, No. 12, pp. 3281-3291.
Domenico, P.A. and F. W. Schwartz, 1990. Physical and Chemical Hydrogeology, John Wiley
and Sons, New York, NY, 824 p.
Freeze, R. A. and J. A. Cherry, 1979. Ground-water, Prentice Hall, Englewood Cliffs, NJ.
Freyberg, D. L., 1986. "A Natural Gradient Experiment on Solute Transport in a Sand Aquifer.
2. Spatial Moments and the Advection and Dispersion of Nonreactive Tracers," Water Resources
Research, Vol. 22, No. 13, pp. 2031-2046.
Garabedian, S. P., D. R. LeBlanc, L. W. Gelhar, and M. A. Celia, 1991. "Large-Scale Natural
Gradient Tracer Test in Sand and Gravel, Cape Cod, Massachusetts. 2. Analysis of Spatial
Moments for a Nonreactive Tracer," Water Resources Research, Vol. 27, No. 5, pp. 911-924.
Gelhar, L. W., A. Montoglou, C. Welty, and K. R. Rehfeldt, 1985. "A Review of Field Scale
Physical Solute Transport Processes in Saturated and Unsaturated Porous Media," Final Proj.
Report., EPRI EA-4190, Electric Power Research Institute, Palo Alto, CA.
Gelhar, L. W., C. Welty and K. R. Rehfeldt, 1992. "A Critical Review of Data on Field-Scale
Dispersion in Aquifers," Water Resources Research, Vol. 28, No. 7, pp. 1955-1974.
Karickhoff, S. W., 1981. "Semi-Empirical Estimation of Sorption of Hydrophobic Pollutants on
Natural Sediments and Soils," Chemosphere, Vol. 10, pp. 833-846.
LeBlanc, D. R., S. P. Garabedian, K. M. Hess, L. W. Gelhar, R. D. Quadri, K. G. Stollenwerk,
and W. W. Wood, 1991. "Large-Scale Natural Gradient Tracer Test in Sand and Gravel, Cape
Cod, Massachusetts. 1. Experimental Design and Observed Tracer Movement," Water
Resources Research, Vol. 27, No. 5, pp. 895-910.
212
-------
Mackay, D. M., D. L. Freyberg, P. V. Roberts, and J. A. Cherry, 1986. "A Natural Gradient
Experiment on Solute Transport in a Sand Aquifer. 1. Approach and Overview of Plume
Movement," Water Resources Research, Vol. 22, No. 13, pp. 2017-2029.
Pickens, J. F. and G. E. Grisak, 1981. "Scale-Dependent Dispersion in a Stratified Granular
Aquifer,"/. Water Resources Research, Vol. 17, No. 4, pp. 1191-1211.
Texas Natural Resource Conservation Commission, 1994. "Risk-Based Corrective Action for
Leaking Storage Tank Sites," Austin, TX.
U. S. Environmental Protection Agency, 1986. Background Document for the Ground-Water
Screening Procedure to Support 40 CFR Part 269 - Land Disposal, EPA/530-SW-86-047,
January 1986.
Walton, W. C., 1988. Practical Aspects of Groundwater Modeling, National Water Well
Association, Worthington, OH, 587 p.
213
-------
APPENDIX II. INTERPRETATION OF OUTPUT
II.l Standard Output File (SOF)
The BIOPLUME III model generates a Standard Output File (SOF) that lists the results from a
specific model run. It is often very useful to review this file for two purposes: (1) to ensure the
accuracy of the input data since the SOF contains a summary of this data; and (2) to verify that
the model run was completed without significant errors or warnings during execution. The SOF
file first lists the input data for the run followed by computed head and concentration maps.
Additionally, the SOF contains mass balance data for the hydraulic and transport calculations
from the model run. Finally, the SOF file contains data at observation wells if they had been
specified by the user.
II.2 Graphical Output File (GOF)
A companion output file to the standard file discussed above is the Graphical Output File
(GOF). The GOF contains all the significant output data from a model run that can be used to
generate graphical output such as contour maps for heads and concentrations. The data in the
GOF can be extracted and used in conjunction with a graphics generation software program to
generate mapped results from the model.
II.3 Resulting Heads
Typically, the SOF and GOF contain the computed heads for the site based on the input data
provided by the user. The head data are listed in three different but associated formats as
follows:
1st Format - computed head matrix in decimal format
N= 1
NUMBER OF ITERATIONS = 20
1HEAD DISTRIBUTION - ROW
NUMBER OF TIME STEPS = 1
TIME(SECONDS) = 0.78894E+08
TIME(DAYS) = 0.91312E+03
TIME(YEARS) = 0.25000E+01
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
100.0000000
99.9510147
99.9024628
99.8567306
99.8223890
99.8219954
99.8955020
100.0000000
.0000000
.0000000
100.0000000
99.9506046
99.8996127
99.8453541
99.7884550
99.7480648
99.8645340
100.0000000
.0000000
.000000
100.0000000
99.9518232
99.8999880
99.8366370
99.7380312
99.5172336
99.8146013
100.0000000
.0000000
.0000000
100.0000000
99.9567243
99.9118469
99.8631898
99.8098146
99.7682048
99.8766615
100.0000000
.0000000
.0000000
100.0000000
99.9632506
99.9274539
99.8944758
99.8698476
99.8690780
99.9238635
100.0000000
.0000000
.0000000
100.0000000
99.9688484
99.9402101
99.9174271
99.9060372
99.9143640
99.9497385
100.0000000
.0000000
.0000000
100.0000000
99.9719650
99.9470698
99.9290041
99.9225293
99.9325607
99.9607587
100.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
.0000000
214
-------
2nd Format - computed head matrix in integer format
1HEAD DISTRIBUTION - ROW
NUMBER OF TIME STEPS = 1
TIME(SECONDS) = 0.78894E+08
TIME(DAYS) = 0.91312E+03
TIME(YEARS) = 0.25000E+01
0000000000
0 0100100100100100100100 0
0 0100100100100100100100 0
0 0100100100100100100100 0
0 0100100100100100100100 0
0 0100100100100100100100 0
0 0100100100100100100100 0
0 0100100100100100100100 0
0 0100100100100100100100 0
0000000000
3rd format - computed drawdown in integer format
1DRAWDOWN
000000000
000000000
0-100-100-100-100-100-100-100 0
0-100-100-100-100-100-100-100 0
0-100-100-100-100-100-100-100 0
0-100-100-100-100-100-100-100 0
0-100-100-100-100-100-100-100 0
0-100-100-100-100-100-100-100 0
000000000
000000000
II.4 Resulting Concentrations
Typical output from BIOPLUME III lists the resulting concentrations for the contaminant and
the electron acceptors at the site. The concentration matrices are listed in the SOF in integer
format, while they are listed in decimal format in the GOF. For example, the concentration
matrix for hydrocarbon in the SOF might be:
00 000 00 0 00
0 0 98 98 98 98 98 98 98 0
0 0 98 98 98 98 98 98 98 0
0 0 98 98 98 98 98 98 98 0
0 0 98 98 98 98 98 98 98 0
0 0 98 98 98 98 98 98 98 0
0 0 98 98 98 98 98 98 98 0
0 0 98 98 98 98 98 98 98 0
0 0 98 98 98 98 98 98 98 0
00 0000 0 000
215
-------
The equivalent matrix in the GOF would be:
.000000
.000000
.000000
.000000
.000000
.000000
.000000
.000000
.000000
.000000
.000000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
.000000
.000000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
.000000
.000000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
.000000
.000000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
.000000
.000000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
.000000
.000000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
.000000
.000000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
97.5000
.000000
.000000
.000000
.000000
.000000
.000000
.000000
.000000
.000000
.000000
.000000
II.5 Mass Balance Results
Two types of mass balances are reported in the BIOPLUME III output: (1) the hydraulic mass
balance; and (2) the transport mass balance. The hydraulic mass balance is reported in general
following the water table calculations. Typically, the information includes the following:
CUMULATIVE MASS BALANCE - (IN FT**3)
RECHARGE AND INJECTION
PUMP AGE AND E-T WITHDRAWAL
CUMULATIVE NET PUMPAGE
WATER RELEASE FROM STORAGE
LEAKAGE INTO AQUIFER
LEAKAGE OUT OF AQUIFER
CUMULATIVE NET LEAKAGE
MASS BALANCE RESIDUAL
ERROR (AS PERCENT)
RATE MASS BALANCE - (IN C.F.S.)
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
0
O.OOOOOE+00
LEAKAGE INTO AQUIFER
LEAKAGE OUT OF AQUIFER
NET LEAKAGE (QNET)
RECHARGE AND INJECTION
PUMPAGE AND E-T WITHDRAWAL
NET WITHDRAWAL (TPUM)
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
The majority of the output is self-explanatory. The first part lists the total volumes of water
into and out of the aquifer in ft3 and the second part lists the data in terms of rate in cu. ft/sec.
The last variable in the first part of the data is the hydraulic mass balance error for the flow. The
hydraulic mass balance error should be relatively low (less than 1%).
The transport mass balance is provided for the contaminant and for the electron acceptors. The
data provided for the contaminant, for example, includes:
216
-------
CHEMICAL MASS BALANCE
MASS IN BOUNDARIES = O.OOOOOE+00
MASS OUT BOUNDARIES = O.OOOOOE+00
MASS PUMPED IN = -O.OOOOOE+00
MASS PUMPED OUT = -O.OOOOOE+00
MASS LOST W. O2 BIODEG. = O.OOOOOE+00
MASS LOST W. NO3 BIODEG. = O.OOOOOE+00
MASS LOST W. Fe BIODEG. = O.OOOOOE+00
MASS LOST W. SO4 BIODEG. = O.OOOOOE+00
MASS LOST W. CO2 BIODEG. = O.OOOOOE+00
MASS LOST BY DECAY = O.OOOOOE+00
MASS ADSORBED ON SOLIDS= O.OOOOOE+00
INITIAL MASS ADSORBED = O.OOOOOE+00
INFLOW MINUS OUTFLOW = -O.OOOOOE+00
INITIAL MASS DISSOLVED = O.OOOOOE+00
PRESENT MASS DISSOLVED = O.OOOOOE+00
CHANGE MASS DISSOLVED = -O.OOOOOE+00
CHANGE TOTL.MASS STORED= 0.68039E+09
COMPARE RESIDUAL WITH NET FLUX AND MASS ACCUMULATION:
MASS BALANCE RESIDUAL = -O.OOOOOE+00
ERROR (AS PERCENT) =-O.OOOOOE+00
COMPARE INITIAL MASS STORED WITH CHANGE IN MASS STORED:
ERROR (AS PERCENT) = O.OOOOOE+00
Mass in/out Boundaries estimates the amount of mass that enters or leaves the boundaries of the
specified grid. Mass pumped in/out estimates the mass entering or leaving the model grid through
injection and pumping wells. Losses due to biodegradation and first-order decay mechanisms are
listed individually in the matrix. Similarly adsorbed mass (initial and current) is also listed.
Present Mass Dissolved represents the mass currently remaining in the aquifer. The remainder of
the mass balance data illustrate the various steps in the mass balance calculations to estimate the
resulting error. The user is referred to Section A.5 for further details.
Note: The units used in the mass balance calculations depend on the concentration units specified
by the user. For example, if the user specified all the input concentration data using mg/L, then
the mass balance information would have units of mg/L»ft3. In order to obtain the mass in mg, the
user needs to multiply the numbers in the mass balance matrix by 28.03 to convert the L to ft3.
217
-------
APPENDIX III. QUESTIONS MOST ASKED
III.l Can I use the model for an unconfined aquifer?
The model, while designed for a confined aquifer, may certainly be used for simulating an
unconfined aquifer. The only condition placed on the model would involve injection and
pumping activities. As a general rule of thumb, any change in head due to injection and pumping
should not exceed 10 - 15% of the specified saturated thickness of the aquifer. If this condition is
violated and heads are allowed to vary outside this range (due to pumping/injection), the accuracy
of the hydraulic solution would decline thus potentially causing errors in the transport solution.
This is mainly because the specific difference between a confined and unconfined aquifer has to
do with the saturated thickness being constant in one and variable in the other as water is
removed from or added to the aquifer.
III.2 I need to model a larger grid.
The model grid, in principle, can be as large as needed. In practice, however, there is a limitation
based on the amount of memory available in the particular computer platform being used.
Consult your particular platform implementation version to determine the maximum grid size
that can be accommodated. If a larger grid size is still desired, you can modify and recompile the
code.
III.3 Should I assume steady-state or transient hydraulics?
The majority of hydraulic conditions at sites, unless a pump test or a specific transient scenario
is being simulated, can be adequately modeled assuming steady-state hydraulics. This is
particularly true if a long period of time is being simulated (on the order of years). While most
sites experience seasonal variations, it is not practical to simulate these events individually over a
long period of time. A common approach involves establishing "an average" hydraulic condition
for the site and using it for the model simulations.
III.4 I have large mass balance errors ...
Mass balance errors are influenced by a number of model parameters. Pumping or injecting
significant amounts of water into or out of the aquifer generates relatively high velocities around
the pumping/injection zones and might cause large mass balance errors. Two model parameters:
NPTPND and CELDIS, directly affect mass balance errors. A larger number of particles per cell
generates less error but requires longer runtimes. The relationship between CELDIS and mass
balance errors is not as direct. Decreasing CELDIS might improve the mass balance error or cause
it to oscillate in different time periods. Again, a smaller CELDIS causes longer runtimes.
218
-------
III.5 My model runs forever ...
Model runtime is determined by the number of particle moves that the program has to complete.
The number of particle moves required is determined internally by the model based on one of
four criteria (see Section A.3 in Appendix A).
In order to reduce runtimes, you need to determine which criterion is being used for the run in
question and change the parameter that influences the internal calculation of the number of
particle moves required. To determine the criterion being used, you need to terminate the
computer run in question (by using an "escape" sequence of keys and not by rebooting
computer) and examine the output file generated from the run by searching for the words
STABILITY CRITERIA.
III.6 My model is generating particles. Is there something wrong?
In some situations, a cell may become void of particles. In order to ensure numerical accuracy,
the model limits the number of cells that can be void of particles to a small percentage of the total
number of cells that represent the aquifer. If the limit is exceeded, the numerical solution of the
transport equation is terminated at the end of the time increment and the concentrations are
saved. Then the model regenerates the initial particle distribution throughout the grid and assigns
the "final" concentrations at the time of termination as the new "initial" particle concentrations.
The solution is then continued in time.
III.7 My plume is running off the page. Is this OK?
This is one of the most common mistakes made in using BIOPLUME III. The grid used is
basically too small and the plume migrates beyond the edge of the model grid. The resulting
model may not accurately represent site conditions. You need to re-examine your grid design and
lengthen your grid as necessary to allow the plume to be contained within the model grid.
III.8 I'm setting up all my cells as constant-head nodes to fix the
ground water elevations at the cells. Will it work?
This is not a particularly useful approach. A significant step of modeling the site involved
calibrating the hydraulics. If you were to fix the water table, you would not be able to judge the
effectiveness of the model in predicting the site conditions.
III.9 What happens to particles that migrate off the grid?
Neither water or dissolved chemicals are allowed to cross a no-flow boundary in the model.
Under certain conditions it might be possible for a particle to be advected across the boundary
during a time increment. The model responds to this situation by relocating the particle within
the aquifer by reflection across the boundary.
219
-------
APPENDIX A. BACKGROUND INFORMATION ON THE
USGS MOC MODEL
A.I Introduction
The USGS Method of Characteristics (MOC) Model was developed by L. F. Konikow and J. D.
Bredehoeft in 1978 (Konikow and Bredehoeft, 1978). The model has been revised numerous
times since its development. The latest modification completed in 1989 incorporated into the
model decay and equilibrium-controlled sorption or ion exchange (Goode and Konikow, 1989).
The 1989 version of MOC was used to develop BIOPLUME III.
The MOC model simulates solute transport in flowing ground water. The model is flexible in
that it can be applied to a wide range of problem types. It is applicable to one- or two-
dimensional problems involving steady-state or transient flow. The model computes changes in
concentration over time caused by the process of advection, hydrodynamic dispersion, and
mixing or dilution from fluid sources. In its 1978 version, the model assumed that the solute is
non-reactive and that gradients of fluid density, viscosity, and temperature do not affect the
velocity distribution. The 1989 version of the model simulates exponential decay such as
radioactive decay; reversible equilibrium-controlled sorption with Linear, Freundlich, or Langmuir
isotherms; and reversible equilibrium-controlled ion exchange for monovalent and divalent ions.
The aquifer may be heterogeneous and/or anisotropic.
The MOC model couples the ground water flow equation with the solute transport equation.
The computer code uses an alternating-direction implicit procedure to solve a finite-difference
approximation to the ground water flow equation, and it uses the method of characteristics to
solve the solute transport equation. The latter uses a particle tracking procedure to represent
advective transport and a two-step explicit procedure to simulate hydrodynamic dispersion, fluid
sources and sinks, and divergence of velocity. The explicit procedure used in the MOC model
has several stability criteria which are used internally to address time-step limitations.
A.2 Theoretical Background
A.2.1 Flow Equation
The equation solved in MOC describing the transient two-dimensional flow of homogeneous
compressible fluid through a nonhomogeneous anistropic aquifer is given by:
s|j- + W i/j = 1,2 (A.I)
where
Tj; is the transmissivity tensor, L^ /T;
220
-------
h is the hydraulic head, L;
S is the storage coefficient, (dimensionless);
t is the time, T;
W = W(x,y,t) is the volume flux per unit area (positive sign for outflow and
negative for inflow), L/T; and
xj and xj are the Cartesian coordinates, L.
The fluxes considered in MOC include direct withdrawal or recharge, such as well pumpage,
injection, or evapotranspiration, and steady leakage into or out of the aquifer through a confining
layer, streambed, or lakebed:
Kz
W(x,y,t) = Q(x,y,t)- — (He - h) (A.2)
where
Q is the rate of withdrawal (positive sign) or recharge (negative sign), L/T;
Kz is the vertical hydraulic conductivity of the confining layer, streambed or, lakebed,
L/T;
m is the thickness of the confining layer, streambed or, lakebed, L; and
He is the hydraulic head in the source bed, stream, or lake, L.
The average seepage velocity of ground water is derived from Darcy's law:
Kii ah
where
V{ is the seepage velocity in the direction of xj , L/T;
Kj; is the hydraulic conductivity tensor, L/T; and
n is the effective porosity of the aquifer, (dimensionless).
A.2.2 Transport Equation
The transport equation solved in MOC is given by:
a(Cb) a / ac\ a C'W
——- = — bD» - — /bCV;\ ii=12 (A 4)
at dxj I l) 3xi 3xj (. !; n '> ' \ • t
where
is the concentration of the dissolved chemical species,
is the coefficient of hydrodynamic dispersion (a second-order tensor), L^ /T;
221
-------
b is the saturated thickness of the aquifer, L; and
C' is the concentration of the dissolved chemical in a source or sink fluid,
The first term on the right side of the equation represents the change in concentration due to
hydrodynamic dispersion. The second term describes advection while the third term describes a
fluid source or sink.
A.2.3 MOC Assumptions
The main assumptions in MOC include:
1. Darcy's law is valid and hydraulic-head gradients are the only driving mechanism for flow.
2. The porosity and hydraulic conductivity of the aquifer are constant with time, and
porosity is uniform in space.
3. Gradients of fluid density, viscosity, and temperature do not affect the velocity
distribution.
4. No chemical reactions occur that affect the fluid properties, or the aquifer properties.
5. Ionic and molecular diffusion are negligible contributors to the total dispersive flux.
6. Vertical variations in head and concentration are negligible.
7. The aquifer is homogeneous and isotropic with respect to the coefficients of longitudinal
and transverse dispersivity.
A.2.4. Numerical Methods
A.2.4.1 Flow Equation. The flow equation (Equation A.I) is approximated with an
implicit finite difference equation. The resulting finite difference equation is solved using an
iterative alternating-direction implicit (ADI) procedure. After the head distribution has been
computed for a given time step, the velocity of ground water flow is computed at each node using
an explicit finite-difference form of Equation (A.3).
A.2.4.2 Transport Equation - The Method of Characteristics. The Method of
Characteristics is used to solve the transport equation in this model. The approach taken by
MOC is not to solve Equation (A.4) directly, but rather to solve an equivalent system of
ordinary differential equations. Representative fluid particles are advected with the flowing
ground water and changes in their chemical concentrations are observed as they move (see Figure
A.1).
The first step in MOC involves placing a number of traceable particles in each cell of the finite-
difference grid. These particles are distributed in a uniform geometric pattern throughout the grid
222
-------
EXPLANATION
Nod« of finite-difference grid
Location gl particle p
X or Y cemporienl of velocity
Arts of influence for interpolating velocity
in X direction at particle ft
Ana of ir»f|u«nc« for interpolating velocity
In Y direction at particle j>
Source: Konikow and Bredehoeft, 1978
Figure A. 1 Relation of Flow Field to Movement of Particles
223
-------
(four, five, eight, nine or sixteen particles are allowed per cell in MOC). The location of each
particle is specified by its x- and y- coordinates. The initial concentration assigned to each point
is the initial concentration associated with the node of the cell containing the particle.
For each time step every particle is moved a distance proportional to the length of the time
increment and the velocity at the location of the particle. After all particles have been moved, the
concentration at each node is temporarily assigned the average of the concentrations of all
particles then located within the area of that cell. The moving particles thus simulate advective
transport because the concentrations at each node of the grid will change with each time step as
different particles having different concentrations enter and leave the area of that cell.
The changes in concentration caused by hydrodynamic dispersion and fluid sources are then
computed at each node of the grid rather than directly at the location of each particle because of
the difficulty in computing concentration gradients at a large number of moving points.
A.3 Stability Criteria
The explicit numerical solution of the solute-transport equation has a number of stability criteria
associated with it. These may require that the flow time step be subdivided into a number of
smaller time increments to accurately solve the transport equation. There are four stability
criteria that drive the transport time-step determination: dispersion, mixing and advection in the
x- and y- directions.
Dispersion Citerion:
At < Min over grid
0.5
Dxx Dyy
Ax2 Ay2
(A. 5)
Mixing Criterion:
At < Min over grid
/
k
(A.6)
Velocity Criteria:
At
yAx
Vx
max
(A.7)
At
(A. 8)
where y is the fraction of the grid dimension that particles will be allowed to move (CELDIS in
the input stream).
224
-------
If the time step used in the flow computations exceeds the smallest of the time steps computed
using equations (A.5 through A.8), then the time step will be subdivided into the appropriate
number of smaller time increments required for solving the solute transport equation.
These criteria determine to a large degree the length of runtimes for a model simulation. Therefore
to decrease runtimes (with the possible outcome of increasing numerical errors), one has to
determine which of the stability criteria influences the time step calculation and modify the
parameters involved accordingly.
A.4 Boundary and Initial Conditions
To obtain a solution to the equations that describe ground water flow and solute transport, it is
necessary to specify boundary and initial conditions for the domain of the problem. Two types
of boundary conditions are incorporated into MOC: constant-flux and constant-head boundary
conditions. These can be used to represent artificial boundaries for the model as well as to
represent the real boundaries of the aquifer.
A constant-flux boundary can be used to represent aquifer underflow, well withdrawals, or well
injection. A finite flux is designated by specifying the flux rate as a well discharge or injection
rate for the appropriate nodes. A no-flow boundary is necessary around the area of interest for
MOC. No-flow boundaries can also be located elsewhere in the grid to simulate natural limits or
barriers to ground water flow. No flow boundaries are designated by setting the transmissivity
equal to zero at appropriate nodes, thereby precluding the flow of water or dissolved chemicals
across the boundaries of the cell containing that node.
A constant-head boundary is used to represent parts of the aquifer where the head will not
change with time, such as recharge boundaries or areas beyond the influence of hydraulic stresses.
In MOC, constant head boundaries are simulated by adjusting the leakage term at the appropriate
nodes. This is accomplished by setting the leakage to a sufficiently high value (such as 1.0 s'1)
to allow the head in the aquifer at a node to be implicitly computed as a value that is essentially
equal to the value of He , which would be specified as the described constant-head altitude.
A.5 Mass Balance
Mass balance calculations are performed after specified time intervals to help check the numerical
accuracy and precision of the solution. The principle of conservation of mass requires that the
total mass inflows and outflows (or net flux) must equal the accumulation of mass (or change in
mass stored):
ZInflows - ^Outflows = AMass Stored (A.9)
The difference between the net flux and the mass accumulation is the mass residual (Rm ):
Rm = AMS -Mf (A. 10)
225
-------
where
AMS is the change in mass stored in aquifer, M; and
Mf is the net mass flux, M.
The change in mass stored is evaluated using the equation:
AMS = 22 bij nAxAy (Cijk - Cijo) (A. 11)
where Cijo is the initial concentration at node (i,j), M/L^ ; and
Mf = 2 2 2 AxAyAtk Wijk C'ijk (A. 12)
t j k } }
The percent error (E) in the mass balance is computed in two different ways. First, the residual
is compared with the average of the net flux and the net accumulation:
100.0 (Mf - AMS)
El = 0.5 (Mf + AMS) (A'13)
Equation (A. 13) is a good measure of the accuracy of the numerical solution when the flux and
the change in mass stored are relatively large. Equation (A. 13) does not account for initial mass
stored in the aquifer. A second type of error computed by the model accounts for this situation
by comparing the residual with the initial mass of solute (Mo ) in the aquifer:
100.0 (Mf - AMS)
E2 = M^ (A. 14)
Equation (A. 14) becomes meaningless, however, when Mo is zero or small relative
to AMS . In these cases, the model will compute a third type of error measure:
100.0 (Mf - AMS)
E3 = M0-Mf (
In general, either one or both of Ej or £3 is computed by the model.
226
-------
A.6 Evaluation of MOC - Comparison with Analytical Solutions
The accuracy of the MOC model was evaluated by comparing one-dimensional model
simulations to an analytical solution of contaminant transport (steady-state flow through a
homogeneous isotropic medium; Konikow and Bredehoeft, 1978). The analytical solution
consisted of the following system of equations:
C(x,t)-C0
Initial Conditions C = Co for t < 0 and -oo < x < 0
C = CL for 0
•}/~~t
Boundary Conditions -j— = 0 for t > 0 and x = ±
C = C forx = + oo
C = C0
where
erfc is the complimentary error function, and
q = nV is the specific discharge, LT~1
Figure A.2 presents the results from the comparisons for two different values of dispersivity.
As can be seen from Figure A.2, there was exact agreement between modeled results and the
analytical solution for higher dispersion. There is a small difference in the modeled results at
some nodes for the case of low dispersion. The authors of MOC attribute these differences to
the error in computing the concentration at a node as the arithmetic average of the concentrations
of all particles located in that cell.
The MOC model was also evaluated by comparing it to the analytical solution for the problem of
plane radial flow in which a well continuously injects tracer at a constant rate qw, and a constant
concentration Cr>:
C 1
C0 - 2
TT erfc
r2/2-Gt
4/Scur3 ,
1 /
(A. 17)
227
-------
Analytic*!
DtSlANCE, IN KIT
Source: Konikow and Bredehoeft, 1978
too
Source: Konikow and Bredehoeft, 1978
Figure A.2 Comparison Between Analytical Model and MOC
for Dispersion in 1-D Steady-State Flow
228
-------
where
G =qw/2itnb=Vr
r is the radial distance from the center of the well, L; and
r = V(2Gt) is the average radius of the body of inj ected water, L.
Figure A.3 presents the results from the comparison. Here again, there is good agreement
between the modeled data and the analytical solution. Some numerical dispersion can be seen in
the modeled results. The authors attribute this problem to the regeneration of particles in the
model.
A.7 Mass Balance Tests
The accuracy and precision of the numerical methods used in MOC were evaluated by computing
the mass balance error for three problems: (1) the spread of a tracer slug; (2) the effects of wells;
and (3) the effects of user options.
Spreading of a Tracer Slug. This problem illustrates the mass balance errors resulting from
advection and dispersion modeling in MOC. A slug of tracer was placed in four cells of a grid
whose boundary conditions generated a steady-state flow field that was moderately divergent in
some places and moderately convergent in other places (see Figure A.4). The aquifer was
assumed to be homogeneous and isotropic. The parameters used in the model run are listed in
Table A. 1. The slug of known tracer was allowed to move downgradient for 2 years. The model
was run for two cases: (1) no dispersion; and (2) a longitudinal dispersivity of 100 ft. Figure A.5
illustrates the resulting mass balance errors for the two cases. As can be seen from Figure A.5,
the mass balance errors ranged from -8 to +8%.
Effects of Wells. This problem evaluates the mass balance errors for scenarios where the flow
field is influenced by wells. The grid and boundary conditions used for this problem are shown
in Figure A.6. The problem simulates one injection well and one pumping well. The parameters
for this problem are listed in Table A.2. The aquifer was assumed to be isotropic and
homogeneous. The problem was simulated for 2.4 years and assumed steady-state flow. This
problem was also simulated for two cases: (1) no dispersion; and (2) a longitudinal dispersivity
of 100 ft. Figure A.7 illustrates the resulting mass balance errors for the two cases. As can be
seen from Figure A.7, the mass balance errors ranged from -8 to +8%.
Effects of User Options. There are two parameters that are specified by the user that impact
the accuracy, precision and efficiency of the model results. These include the initial number of
particles per node (NPTPND) and the maximum fraction of the grid dimensions that particles are
allowed to move (CELDIS). The set-up for the effects of wells problem was used to evaluate the
effect of these two parameters on mass balance results. Figure A.8 illustrates the relationship
between the number of particles (NPTPND) and the mass balance error. The errors are smaller
for higher numbers of particles. However, it is noted that longer runtimes are also required.
Figure A.9 illustrates the relationship between CELDIS and mass balance errors. It can be seen
from Figure A.9 that the impact of CELDIS on the error is more complicated. A decrease in
229
-------
TECHNIQUES OF WATER RESOURCES INVESTIGATIONS
1,0 1—
0.0
100
200
RADIAL DiSTAMCE, IN FEET
Source: 1978
Source: Konikow and Bredehoeft, 19
Figure A.3 Comparison Between Analytical Model and MOC for
Dispersion Plane Radial Steady-State Flow
230
-------
MODEL OF SOLUTE TRANSPORT IN GROUND WATER
and 1978
EXPLANATION
(CoJ
90- Computed poleritiomtuic •Ilitudt
Cf>^looi ir»1e»v«I 2,0 (»«t
(0.61 meiei)
6^= S0§ (c*
£»/= 800 letl (274
Source: Konikow and Bredehoeft, 1978
Figure A.4 Grid Boundary Conditions and Flow Field
for the Tracer Slug Mass Balance
Test Problem
231
-------
Table A. 1. Model Parameters for the Tracer Slug Mass Balance Test Problem
Aquifer Properties Numerical Parameters
K = 0.005 ft/s. AX = 900 ft
= (1.5xlO-3m/s) (274m)
b =20.0 ft AY = 900ft
= 6.1 m (274m)
S =0.0 CELDIS = 0.49
n =0.30 NPTPND= 9
ocT/aL = .30
232
-------
TECHNIQUES OF WATER RESOURCES INVESTIGATIONS
10.0
z
LU
u
EC
UJ
cc
O
cc
£ °*°
UJ
U
< -5.0
to
4/5
-10.0
Q.O
-J-
a,= 0.0 feel /S ' \
*• V r t
•SA^
0.5
1.0
TIME, IN YEARS
1.5
2.0
Source: Konikow and Bredehoeft, 1978
Figure A.5 Mass Balance Errors for the
Tracer Slug Mass Balance Problem
233
-------
EXPLANATION
No-flow boundary
Constanl»head boundary
Injection
Withdrawal
!l
Conipwled poltsni torn* trie •
Conloyr interval 2,0 feat
{0,61
itity,
s WO I»«l |2M
= 900 feet 1274 «wil*r«J
Source: Konikow and Bredehoeft, 1978
Figure A.6 Grid, Boundary Conditions and Flow Field for
Effects of Wells Mass Balance Test Problem
234
-------
Table A.2. Model Parameters for the Effects of Wells Mass Balance Test Problem
Aquifer Properties Numerical Parameters
K = 0.005 ft/s. DX = 900 ft
= (1.5xlOJm/s) (274m)
b =20.0 ft DY = 900ft
= (6.1m) (274m)
S =0.0 CELDIS = 0.49
n =0.30 NPTPND= 9
aT/aL = .30
C' = 100.0
C0 = 0.0
qw =1.0ft3/s
= (0.028 m3 / s)
235
-------
10,0
2.5
TIME, IN YEARS
Source: Konikow and Bredehoeft, 1978
Figure A.7 Mass Balance Errors for the Effects of
Wells Mass Balance Problem
236
-------
MODEL OF SOLUTE TRANSPORT IN GROUNDWATER
20. Or
t NPIPND* 4
NPlPNDs 5
£> £ NP1PND* t
o.s
1.0 1.5
TIME. IN YEARS
Source: Konikow and Bredehoeft, 1978
Figure A.8 Effect of Number of Particles
on Mass Balance Error
237
-------
TECHNIQUES OF WATER RESOURCES INVESTIGATIONS
II..G
CiLDlS«0.2S
O O CUDtSsO.SO
£> OCILOISs 0.75
0 O CELD1S«1.00
TIME, IN YEARS
Source: Konikow and Bredehoeft, 1978
Figure A.9 Effect of Maximum Cell Distance
(CELDIS) on Mass Balance Errors
238
-------
CELDIS from 0.5 to 0.25 caused the mass balance errors to oscillate for the first 1.5 years before
the solution converged to a small error. Again, a smaller CELDIS caused longer runtimes.
The effects of NPTPND and CELDIS on the mass balance error are problem dependent. In
problems where CELDIS is not the influencing stability criterion, varying CELDIS will not have
an effect on mass balance errors. In general, it is recommended that the user specify 9 particles
per cell (NPTPND = 9) and a CELDIS of 0.5 for model runs (initial model set-up or calibration
runs can be developed using a smaller number of particles (4 or 5) and a higher number for
CELDIS (0.75 or 1) as a first-cut).
A.8 References
Goode, D. J. and L. F. Konikow, Modification of a Method of Characteristics Solute Transport
Model to Incorporate Decay and Equilibrium-Controlled Sorption or Ion Exchange, USGS, Water
Resources Investigation Report 89-4030, Reston, Virginia, 1989.
Konikow, L. F. and J. D. Bredehoeft, Computer Model of Two-Dimensional Solute Transport
and Dispersion in Ground Water, Techniques of Water Resources Investigation of the United
States Geological Survey, Book 7, Reston, Virginia, 1989.
239
-------
APPENDIX B. IMPLEMENTING THE AIR FORCE
INTRINSIC REMEDIATION PROTOCOL
USING THE GRAPHICAL PLATFORM
B.I Context of the Remedial Investigation Using the Platform
Remediation and containment are the enabling technologies for the immediate control of the
spread of contamination, and for the long-term management strategy, especially where natural
degradation is allowed to play a constructive role. The mechanism for determining the overall
best remediation strategy is by implementing comprehensive Remedial Investigation/ Feasibility
studies. These two complementary investigations represent respectively, the diagnostic and the
prognostic aspects of the remediation process. The Remedial Investigation (diagnosis) is
conducted concurrently with the Feasibility Study (prognosis), and emphasizes data collection
and site characterization. The data collected during the diagnostic phase of the study are used to
evaluate the existing state of the site, and to support the analysis and decision-making activities
of the feasibility study, including the formulation of remedial alternatives. Comprehensive
geomedia rehabilitation must include both aspects of the solution.
The primary objective of an Intrinsic Remediation (IR) or Natural Attenuation investigation is to
show that natural processes cause contaminant degradation and can reduce contaminant
concentrations in groundwater to below regulatory standards before exposure pathways reach
susceptible populations. This requires that the potential, extent, and concentration of the
contaminant plume must be projected in time and space. This projection should be based on
historic variations and the current extent of the contaminant plume as well as the measured rates
of contaminant attenuation.
It is the responsibility of the proposer to provide sufficient evidence to demonstrate that a
selected remediation mechanism will reduce contaminant concentrations to acceptable levels
before potential receptors are reached. This requires the use of a model for both the diagnostic
and prognostic phase of the remediation study, so that consideration be given to all possible
contaminant migration scenarios. In what follows we give more details on how to implement
Remedial Investigations and Feasibility Studies using the Graphical User Interface Platform in
conjunction with BIOPLUME III.
Quantification of contaminant migration and attenuation rates and successful implementation of
a remediation scheme require the following steps:
In the Diagnostic phase
1. Review existing data
2. Develop preliminary conceptual model for the site and assess potential impact of
selected remediation [Enter hydrogeologic data in the Platform]
240
-------
3. Perform site characterization in support of selected remediation [Use kriging results
to enhance site boring location]
4. Refine conceptual model based on site characterization data [Calibrate flow and
contaminant migration models in the Platform]
In the Prognostic phase
5. Model intinsic bioremediation scenarios using different features of the Platform.
6. Prepare long-term monitoring plan, and
7. Present findings to regulatory agencies and obtain approval for the selected
remediation plan with long-term monitoring options.
These activities follow the recommendations of the "Technical Protocol for Implementing
Intrinsic Remediation with Long-Term Monitoring for Natural Attenuation of Fuel
Contamination Dissolved in Ground Water" by Todd H. Wiedemeier, John T. Wilson,
Donald H. Kampbell, Ross N. Miller, and Jerry E. Hansen, AFCEE, 1995.
In the present document we elaborate on the implementation of the tasks described in the
above referenced Protocol using the Platform. For ease of reference we adopt the same
order as the Protocol. In fact, as shown in Figure B.I, the present document provides all
the logical connections between the technical protocol on intrinsic remediation and the
Graphical Platform.
1 ecnmcal
Protocol
forlR
Software
Protocol
,7
\\
EIS Platform
BIOPLUME III
Figure B.I Logical Connection Between Technical Protocol and the Platform.
B.I.I Review Existing Site Data
The first step in the remediation investigation is to review existing site-specific data to determine
if a selected procedure is a viable remedial option. Critical review of existing data also allows
development of a preliminary conceptual model. The preliminary conceptual model is an
essential tool for identifying any shortcomings in the data, and in developing a scientifically
advantageous and cost-effective plan for additional data collection.
241
-------
The information that must be obtained during data review includes the following categories:
Soil and ground water quality data:
• Firstly, the three-dimensional distribution of mobile and residual NAPL and
dissolved-phase contaminants. The distribution of mobile and residual NAPL will be
used to define the dissolved-phase plume source area.
• Ground water and soil geochemical data.
• Historic water quality data showing variations in contaminant concentrations through
time.
• Chemical and physical characteristics of the contaminants.
• Potential for biodegradation of the contaminants.
Geologic and hydrogeologic data:
• Lithology and stratigraphy of the geologic medium.
• Grain-size distribution (percent sand, silt, and clay).
• Aquifer hydraulic conductivity.
• Ground water flow gradients and potentiometric or water table surface maps (over
several seasons, if possible).
• Preferential flow paths.
• Interactions between ground water and surface water and rates of infiltration/recharge.
Location of potential receptors:
• Groundwater well locations.
• Ground water discharge points downgradient of site.
If little or no site-specific data are available, then all future site characterization activities should
also include collecting the data necessary to support the intrinsic remediation. Use of the
Platform can greatly streamline this additional data collection activity which is otherwise
necessary for a convincing and successful implementation of the remedial action.
B.1.2 Develop Preliminary Conceptual Model
242
-------
The existing site characterization data are used to develop a conceptual geohydrologic model of
the site and a preliminary assessment of the potential for alternative remediation schemes. The
conceptual model is a word description of the three-dimensional representation of the ground
water flow and contaminant migration system based on available geological, hydrological,
climatological, and analytical data for the site. This type of conceptual model is more detailed
than generic descriptions commonly used for risk analysis that consider the location of
contaminant sources, transport pathways, exposure points, and receptors only qualitatively. The
conceptual model is the most important step in properly developing a site contamination
simulation model (diagnostic phase) which will be used to determine optimal placement of
additional data collection points as necessary to aid in the remediation investigation.
The Platform offers an ideal framework to build this model in a graphical interactive
environment that allows the user to visually inspect all his modeling choices. Figure B.2 shows a
typical Platform screen with a background image which is used as a canvas on which to build the
appropriate model data at the appropriate locations.
Practically all the site data described in Section B.I can now be input into the model through the
use of logpoints, wells and several dialog boxes. The Platform shadows the user's activity
through all steps of the modeling process. Errors, when they occur, will be identified and
corrected on the spot considerably reducing the time to validate the model.
'C"!
F*.
!«-•
.
j
1
IS-'-'i rir;- "S'SJ '"i1""".' ",!.'"/ '"
Q,**.. t«w*»j e« 'i«1 W-'!«.wlfc*,. £««*»!« .=*.»*• w~ ihrww
ftffafel: f> *». ^ lvp,»M^li5|fnfefe
1 .1
PI ? *i»ii«ii ^,
,' - 1" '•'•-,
' *>' ; • ..lii^s
BMri^tiftg Tf "* ^ | ^ " * *" '
'/ -," ,".'-/. ;. / *"''; ./»
' /* / / / v
, ' / iJ it / ,»• - y
• - . • / / / •
/ / •' / (•**!• 1
"1 / / ,'/ / •• .'
/ /
. ' c ,> , ••
?' „.-* * y • j ^ » . • »
IM /7 / ' «• J I
• - 'X- •/ ' '
/" '"' '" ' /
Kl J f if t
: ' * ' f' '" ^
3fi. \
ItrrU f.,,*", ".">
*" ---l| '
/
. /
o " / ':'IM|1I
V ,
isi / Woi"*1}
i . ^t
i_ .**..*' / v
,' 1 '" I •' '
> • ,ttUJ* 'f f -J
' ',', '• ' / »
,-, -J »,v i a j|>
tOwifi--. if I^1™
I . ^ •%^-^_,,
<, , \r'*"" r'^s.^t™*
\ ^ \.^ ' +-
j-*"1* * -1»
^X' l • d'V (
'^iirr Hi*l,
119.^1. m '
SPra
* V
1-
w
r,,t
cs
•E-
*r;
h *
Hffia^
K
.,,
1
W
tb
<
El
%«
Figure B.2 General Configuration of the Platform.
Successful model development involves the following steps:
• Definition of the problem to be solved (usually the unknown nature and extent of
existing and future contamination),
243
-------
• Integration of all available data in the Platform including:
• Local geologic and topographic maps,
• Hydraulic data,
• Site stratigraphy, and
• Contaminant concentration and distribution data.
• Conceptual model development, including extent of the site, boundary
conditions, loading conditions, driving mechanisms, assimilative capacity.
• Determination of additional data requirements including:
• Borehole locations and monitoring well spacing,
• An approved sampling and analysis plan, and
• Any data requirements that have not been adequately addressed.
The purpose for developing the conceptual model is to assess the potential for remediation.
Existing data can be useful in determining if intrinsic remediation will be sufficient to prevent a
dissolved-phase contaminant plume from completing exposure pathways, or from reaching a
predetermined point of compliance, in concentrations above applicable regulatory standards. The
goal of the remedial investigation is to determine the likelihood of pathway completion by
estimating the migration and future extent of the plume based on geologic and contaminant
properties, biodegradability, aquifer properties, head gradients, and the location of the plume and
contaminant source relative to the potential receptor (i.e., the distance between the leading edge
of the plume and the potential receptor).
If the contaminant plume is in its migration phase, then the remediation scheme should be based
on containment and on contaminant reduction, and site characterization activities should be
designed in support of this remedial option. On the other hand, if exposure pathways have
already been completed and contaminants pose an unacceptable risk, then other remedial
measures should be considered. However, even in the latter case attention should be given to use
to the fullest extent the natural attenuation capacity of the site. In this case, containment of the
plume may prevent the further migration of the plume along the exposure pathway while giving it
time to self-destruct over the natural or enhanced biodegradation cycle. This combination of
containment in favor of biodegradation and treatment of the receptor areas can help reduce the
overall cost and duration of the remedial action.
The backbone to all these activities is again the Platform main screen (see Figure B.2), which
provides all the necessary tools to build the conceptual model with the data collected during the
site investigation.
244
-------
B.2 Site Characterization in Support of Intrinsic Remediation
Detailed site characterization is necessary to document the potential for remediation at a site. As
discussed in Section B.I, review of existing site characterization data is particularly useful before
initiating site characterization activities. Such review allows identification of data gaps and in
conjunction with the diagnostic model guide the most effective placement of additional data
collection points, using the kriging (generalized covariance) scheme supported by the Platform.
The site characterization phase of the remediation investigation provides two important pieces of
information about the site: whether natural mechanisms of contaminant attenuation are occurring
at rates sufficient to protect human health and the environment; and sufficient site-specific data
for diagnostic phase model development to allow prediction of the future extent and
concentration of the contaminant plume. Site characterization in support of remediation should
include at least the following set of parameters:
• Extent and type of soil and ground water contamination.
• Location and extent of contaminant source areas) (i.e., areas containing mobile or
residual NAPL).
• The potential for a continuing source due to leaking tanks or pipelines.
• Ground water geochemical parameters.
• Regional hydrogeology, including:
• Drinking water aquifers, and
• Regional confining units.
• Local and site-specific hydrogeology, including:
• Local drinking water aquifers.
• Location of industrial, agricultural, and domestic water wells.
• Patterns of aquifer use.
• Lithology.
• Site stratigraphy, including identification of transmissive and non-transmissive
units.
• Grain-size distribution (percent sand, silt, and clay).
• Aquifer hydraulic conductivity.
• Ground water hydraulic information.
245
-------
• Preferential flow paths.
• Location and type of surface water bodies.
• Areas of local ground water recharge and discharge.
• Definition of potential exposure pathways and receptors.
The following sections describe technologies that can be used in collecting site characterization
data.
B.2.1 Soil Characterization
In order to adequately define the subsurface hydrogeologic system and to determine the amount
and three-dimensional distribution of dissolved hydrocarbons, mobile and residual NAPL and
other dissolved organic contaminants that can act as a continuing source of ground water
contamination, extensive soil characterization must be completed. Soil characterization is
important in determining the source mechanism of a dissolved pollutant plume, but also for
direct remedial action if the soil contamination is above soils standards.
Soil Sampling
The purpose of soil sampling is to determine the subsurface distribution of hydrostratigraphic
units and the distribution of mobile contaminants. These objectives can be achieved through the
use of conventional soil borings or cone penetrometer testing. All soil samples should be
collected, described, and analyzed in accordance with EPA standard procedures. Figure B.3
illustrates a typical site investigation using a CPT rig. The data collected at the borehole of each
CPT location can be sorted and directly input into the program using the logpoint control from
the available toolbox.
246
-------
Contours of Benzene in (ma .'I)
CPTUTeite
Contaminant Plume
Modeling Layer for EIS.'Biopldme
Figure B.3 Soil Sampling Using CPT Technology.
Soil Analytical Protocol
This analytical protocol includes all of the parameters necessary to document remediation of fuel
hydrocarbons, including the effects of sorption and biodegradation (aerobic and anaerobic) of
fuel hydrocarbons. Some analytes are given as a reference below. These data are usually
collected at a logpoint and their distribution given at different depths as a profile. Figure B.4
shows a typical CPTU boring log. These raw data must be interpreted and sorted to determine
the proper value that will be entered into the program. A usual approach is to identify the
geologic strata where the contamination takes place and average the soil properties though the
thickness of this strata as illustrated in Figure B.4.
247
-------
Typical CPTU Bortog-Log
Borehole
Ttp Res. ffffctton R*t1o POTS Contain A
Computational
Layer
Figure B.4 Typical CPTU Boring Log.
Total Volatile and Extractable Hydrocarbons
Knowledge of the location, distribution, concentration, and total mass of TPH sorbed to soils or
present as mobile NAPL is required to calculate contaminant partitioning from these phases into
the dissolved phase. The presence or absence of TPH is also used to define the edge of the
nonaqueous phase liquid (NAPL) plume. Knowledge of the location of the leading edge of the
NAPL plume is important in properly setting up the BIOPLUME III model.
Aromatic Hydrocarbons
Knowledge of the location, distribution, concentration, and total mass of fuel-derived
hydrocarbons of regulatory concern (especially BTEX) sorbed to soils or present as mobile
NAPL is required to calculate contaminant partitioning from mobile and residual NAPL into the
dissolved phase.
Total Organic Carbon
Knowledge of the total organic carbon (TOC) content of the aquifer matrix is important in
sorption and solute-retardation calculations. TOC samples should be collected from a
background location in the zones where most contaminant migration is expected to occur.
248
-------
Dehydrogenase Activity
The dehydrogenase test is a qualitative method used to determine if aerobic bacteria are present
in an aquifer in quantities capable of biodegrading fuel hydrocarbons. If the test gives a positive
result, a sufficient number of microorganisms capable of aerobic metabolism and/or
denitrification are present in the aquifer. If the test is negative, the number of aerobic
microorganisms capable of aerobic metabolism is insufficient in the aquifer. However, the
dehydrogenase test gives no indication of the relative abundance of anaerobic microorganisms
capable of utilizing sulfate, iron III, or carbon dioxide for anaerobic biodegradation.
Grain Size Distribution
The grain size distribution of the aquifer matrix is an important indicator of hydraulic
conductivity. In addition, clay minerals can be important sites for contaminant adsorption,
especially when organic carbon comprises less than about 0.1 percent of the aquifer matrix.
Because of this, knowledge of the relative abundance of clay minerals is important in sorption
and solute retardation calculations.
Soil Gas Analysis
The concentrations of soil gas oxygen, carbon dioxide, and total combustible hydrocarbons are
important in defining the extent of NAPL contamination. This information can be used to define
the edge of the free-phase plume and to estimate the potential for natural biodegradation of
vadose zone fuel residuals. Depleted oxygen levels and elevated carbon dioxide levels in soil gas
are indicative of aerobic biodegradation of fuel hydrocarbons in the unsaturated zone, which may
be enhanced if additional oxygen is provided through bioventing.
B.2.2 Ground Water Characterization
Sufficient information must be collected about the ground water system to adequately determine
the amount and three-dimensional distribution of dissolved-phase contamination and to
document its biochemical evolution. Ground water samples must be collected to show
measurable changes in the chemistry of ground water in the affected area which is brought about
by biodegradation. By measuring these changes, a case can be made for the presence of intrinsic
remediation taking place at the site.
Ground Water Sampling
Ground water samples may be obtained from monitoring wells or point-source sampling devices
such as the 'Geoprobe', Hydropunch', or the cone penetrometer. All ground water samples
should be collected in accordance with EPA standards.
249
-------
Ground Water Analytical Protocol
The analytical protocol must include all parameters necessary to document the remediation
process. For intrinsic remediation of fuel hydrocarbons the analytical protocol should include the
effects of sorption and aerobic and anaerobic biodegradation. Data obtained from the analysis of
ground water for these analytes will be used to scientifically document intrinsic remediation of
fuel hydrocarbons and to model the past behavior of the plume (diagnostic phase) and the long
term prediction of its evolution (prognostic phase). The following sections describe the most
prevailing ground water analytical parameters used in the Platform. Most of these data are
usually entered as distributed or uniform parameters throughout the 2D modeling domain.
Appropriate dialog boxes allow the user to input their values as shown in Figure B.5.
Etiodegiadation Election Acceptors
• Select Electron Acceptors and Set Parameters
R? Oxygen
Reaction
Interaction
Nitrate
f~ Ferrous Iron
Manganese
Sulfate
}~ Methane
Set Default Names
OK
Cancel
Figure B.5 Platform controls to input water quality data.
To enter the values of the measured contaminant, electron acceptors or byproducts, use the well
control tool from the toolbox. This operation is presented in Figure B.6, where after properly
locating the well in space you enter and input the physical parameters in an editing mode. To
enter this editing mode you double click on the well and obtain the dialog box shown in Figure
B.6. At this stage you enter or edit the time sequence of the in situ information and enter the
values of the variable in the dialog box shown in Figure B.7. This figure shows a typical input
stream of the measured oxygen at the site. The Oxygen value at time 0 constitutes the initial
condition while the subsequent entries can be used as targets for the model calibration. Note that
the Platform automatically distributes the logpoint information to the area covered by the
simulation.
250
-------
File
(To
[Domain Loading £dit £rid initial Conditions Simulator Fjesults View Annotation
EEH
I.-2
R-1
-FuaM
R2
_«__.
Fence ^
-f-
Loading
Time Steps
Pumping Rates
Concentration
Screened Stratum: |J _ |
Surface Coordinates <
X: 111.
Y: 122.
! SeiValues |
OK
*' A
Ert-1
Pt^^
Stc
;oc|
ECA83.5
rm D
£?
•"reek
11
o
Figure B.6 Entering Well Data in the Platform.
Oxygen Concentrations
Time Concentration
400.
10.
4.
2.
Copi? Values
Pas! e V a I a e s
Ok
Cancel
14-
1IJ;
\
ioq 200) soq 400( 500) eoq 7oq soq aoq IQOC
Time (Days)
Figure B.7 Platform Input of in Situ Measured Oxygen.
251
-------
The same input procedure is used for most of the components influencing intrinsic remediation
described below.
Dissolved Oxygen
Dissolved oxygen is the most thermodynamically favored electron acceptor used in the
biodegradation of fuel hydrocarbons. Dissolved oxygen concentrations are used to estimate the
mass of contaminant that can be biodegraded by aerobic processes. As a rule, the stoichiometric
ratio of dissolved oxygen consumed by microbes to destroyed BTEX compound is 1.0 mg/L of
dissolved oxygen consumed to approximately 0.32 mg/L of BTEX compounds destroyed.
During aerobic biodegradation, dissolved oxygen levels are reduced as aerobic respiration occurs.
Also, anaerobic bacteria (obligate anaerobes) generally cannot function at dissolved oxygen
levels greater than about 0.5 mg/L. Therefore, higher values of dissolved oxygen indicate that
aerobic biodegradation is likely at work.
Dissolved oxygen measurements should be taken during well purging and immediately before
and after sample acquisition using a direct-reading meter. Because most well purging techniques
can allow aeration of collected ground water samples, it is important to minimize potential
aeration.
Oxidation/Reduction Potential (Ejj)
The oxidation/reduction (redox) potential of ground water (EH) is a measure of electron activity
and is an indicator of the relative tendency of a solution to accept or transfer electrons. Redox
reactions in ground water are usually biologically mediated and therefore, the redox potential of a
ground water system depends upon and influences rates of biodegradation. Knowledge of the
redox potential of ground water is also important because some biological processes only operate
within a prescribed range of redox conditions. Knowledge of the redox potential of ground water
can be used as an indicator of certain geochemical activities such as sulfate reduction. The redox
potential of ground water generally ranges from -400 millivolts (mV) to 800 mV.
Redox potential can be used to provide real time data on the location of the contaminant plume,
especially in areas undergoing anaerobic biodegradation. Mapping the redox potential of the
ground water while in the field allows the field scientist to determine the approximate location of
the contaminant plume. To map the redox potential of the ground water while in the field it is
important to have at least one redox measurement (preferably more) from a well located
upgradient of the plume. The redox potential of a ground water sample taken inside the
contaminant plume should be somewhat lower than that taken in the upgradient location. Redox
potential measurements should be taken during well purging and immediately before and after
sample acquisition using a direct-reading meter. Because most well purging techniques can allow
aeration of collected ground water samples (which can affect redox potential measurements), it is
important to minimize potential aeration.
252
-------
pH, Temperature, and Conductivity
Because the pH, temperature, and electric conductivity of a ground water sample can change
significantly within a short time following sample acquisition, these parameters must be
measured in the field in unfiltered, unpreserved, "fresh" water collected by the same technique as
the samples taken for laboratory analyses. The measurements should be made in a clean glass
container separate from those intended for laboratory analysis and the measured values should be
recorded in the ground water sampling record.
The pH of ground water has an effect on the presence and activity of microbial populations in
ground water. This is especially true for methanogens which may be active after all aerobic,
sulfate reduction, and nitrate reduction degradation has taken place. Microbes capable of
degrading petroleum hydrocarbon compounds generally prefer pH values varying from 6 to 8
standard units.
Electric conductivity is a measure of the ability of a solution to conduct electricity. For ground
water, conductivity is directly related to the concentration of ions in solution, increasing as ion
concentration increases. Like chloride, conductivity is used to ensure that ground water samples
collected at a site are representative of the water comprising the saturated zone in which the
dissolved-phase contamination is present. If the conductivities of samples taken from different
sampling points are radically different, then the waters may be from different hydrogeologic
zones.
Ground water temperature directly affects the solubility of oxygen and other geochemical
species. The solubility of dissolved oxygen is temperature dependent, being more soluble in cold
water than in warm water. Ground water temperature also affects the metabolic activity of
bacteria. Rates of hydrocarbon biodegradation roughly double for every 10° C increase in
temperature ("Q"io rule) over the temperature range between 5° and 25° C. Ground water
temperatures less than about 5° C tend to inhibit biodegradation, and slow rates of biodegradation
are generally observed in such waters.
Alkalinity
The total alkalinity of a ground water system is indicative of a water's capacity to neutralize acid.
Alkalinity is defined as the net concentration of strong base in excess of strong acid with a pure
CC>2-water system as the point of reference (Domenico and Schwartz, 1990). Alkalinity results
from the presence of hydroxides, carbonates, and bicarbonates of elements such as calcium,
magnesium, sodium, potassium, or ammonia. These species result from the dissolution of rock
(especially carbonate rocks), the transfer of CC>2 from the atmosphere, and respiration of
microorganisms. Alkalinity is important in the maintenance of ground water pH because it
buffers the ground water system against acids generated through both aerobic and anaerobic
biodegradation processes.
253
-------
Nitrate
In the hierarchical order of processes occurring in the microbiological treatment zone, after
dissolved oxygen has been depleted, nitrate may be used as an electron acceptor for anaerobic
biodegradation. Nitrate concentrations are used to estimate the mass of contaminant that can be
biodegraded by denitrification processes. By knowing the volume of contaminated ground water,
the background nitrate concentration, and the concentration of nitrate measured in the
contaminated area, it is possible to estimate the mass of BTEX lost to biodegradation.
Stoichiometrically, each 1.0 mg/L of ionic nitrate consumed by microbes results in the
destruction of approximately 0.21 mg/L of BTEX compounds. Nitrate concentrations are a direct
input parameter to the Platform.
Sulfate andSulfide Sulfur
After dissolved oxygen and nitrate have been depleted in the microbiological treatment zone,
sulfate may be used as an electron acceptor for anaerobic biodegradation. This process is termed
sulfanogenesis and results in the production of sulfide. Sulfate concentrations are used as an
indicator of anaerobic degradation of fuel compounds. By knowing the volume of contaminated
ground water, the background sulfate concentration, and the concentration of sulfate measured in
the contaminated area, it is possible to estimate the mass of BTEX lost to biodegradation through
sulfate reduction. Stoichiometrically, each 1.0 mg/L of sulfate consumed by microbes results in
the destruction of approximately 0.21 mg/L of BTEX. Sulfate concentrations are a direct input
parameter for the Platform.
Ferrous Iron
Ferric iron is also used as an electron acceptor during anaerobic biodegradation of petroleum
hydrocarbons after nitrate or sulfate depletion, or some times in conjunction with them. During
this process, ferric iron is reduced to the ferrous form which may be soluble in water. Ferrous
iron concentrations are used as an indicator of anaerobic degradation of fuel compounds. By
knowing the volume of contaminated ground water, the background ferrous iron concentration,
and the concentration of ferrous iron measured in the contaminated area, it is possible to estimate
the mass of BTEX lost to biodegradation through ferric iron reduction. Stoichiometrically, the
degradation of 1 mg/L of BTEX results in the production of approximately 21.8 mg/L of ferrous
iron during ferric iron reduction. Iron concentrations are used as a direct input parameter to the
Platform. The equivalent amount of Ferric Iron is estimated from the measured Ferrous Iron,
which is used as model input. BIOPLUME HI simulates the hydrocarbon reduction and the
corresponding Ferric Iron depletion.
Carbon Dioxide
Metabolic processes operating during biodegradation of fuel hydrocarbons result in the
production of carbon dioxide (CO2). Accurate measurement of CC>2 produced during
biodegradation is difficult because carbonate in ground water (measured as alkalinity) serves as
both a source and sink for free CC>2. If the CC>2 produced during metabolism is not removed by
254
-------
the natural carbonate buffering system of the aquifer, CC>2 levels higher than background may be
observed. Comparison of empirical data to stoichiometric calculations can provide estimates of
the degree of microbiological activity and the occurrence of in situ mineralization of
contaminants.
Methane
During methanogenesis (an anaerobic biodegradation process), carbon dioxide (or acetate) is
used as an electron acceptor, and methane is produced. Methanogenesis generally occurs after
oxygen, nitrate, and sulfate have been depleted in the treatment zone. The presence of methane
in ground water is indicative of strongly reducing conditions. Because methane is not present in
fuel, the presence of methane in ground water above background concentrations in contact with
fuels is indicative of microbial degradation of fuel hydrocarbons. Methane concentrations can be
used to estimate the amount of BTEX destroyed in an aquifer. By knowing the volume of
contaminated ground water, the background methane concentration, and the concentration of
methane measured in the contaminated area, it is possible to estimate the mass of BTEX lost to
biodegradation through methanogenesis reduction. The degradation of 1 mg/L of BTEX results
in the production of approximately 0.78 mg/L of methane during methanogenesis. Methane
concentrations are used as an indirect input parameter to the Platform. The equivalent
amount of COi is estimated from the measured methane, which is used as model input.
BIOPLUME III simulates the hydrocarbon reduction and the corresponding COi
depletion.
Chloride
Chloride is used to ensure that ground water samples collected at a site are representative of the
water comprising the saturated zone in which the dissolved-phase contamination is present (i.e.
to ensure that all samples are from the same ground water flow system). If the chloride
concentrations of samples taken from different sampling points are radically different, then the
waters may be from different hydrogeologic zones.
Total Petroleum Hydrocarbons and Aromatic Hydrocarbons
These analytes are used to determine the type, concentration, and distribution of fuel hydrocarbon
in the aquifer. Of the compounds present in most gasolines and jet fuels, the BTEX compounds
generally represent the regulatory contaminants of concern. For this reason, these compounds are
generally of significant interest in the fate and transport analysis. At a minimum, the aromatic
hydrocarbon analysis (Method SW8020) must include the BTEX compounds and the
trimethylbenzene and tetramethylbenzene isomers. The combined dissolved-phase concentrations
of BTEX, trimethylbenzene, and tetramethylbenzene should not be greater than about 30 mg/L
for a JP4 spill. If these compounds are found in concentrations greater than 30 mg/L then
sampling errors such as emulsification of NAPL in the ground water sample have likely occurred
and should be investigated.
255
-------
B.2.3 Aquifer Parameter Estimation
Hydraulic Conductivity
Hydraulic conductivity is a measure of an aquifer's ability to transmit water and is perhaps the
most important aquifer parameter governing fluid flow in the subsurface. The velocity of ground
water and dissolved-phase contamination is directly related to the hydraulic conductivity of the
saturated zone. In addition, subsurface variations in hydraulic conductivity directly influence
contaminant fate and transport by providing preferential paths for contaminant migration.
Estimates of hydraulic conductivity are used to determine residence times for contaminants and
tracers and to determine the seepage velocity of ground water.
The most common methods used to measure hydraulic conductivity in the subsurface are aquifer
pumping tests and slug tests. One drawback to these methods is that they average hydraulic
properties over the screened interval of the well. To help alleviate this potential problem, the
screened interval of the well should be selected after consideration is given to subsurface
stratigraphy. Information about subsurface stratigraphy should come from geologic boring logs
completed on continuous cores. An alternate method to delineate zones with high hydraulic
conductivity is to use pressure dissipation data from CPT logs.
Pumping Tests
Pumping tests generally give the most reliable information on hydraulic conductivity but are
difficult to conduct in contaminated areas because the water produced during the test generally
must be contained and treated. In addition, a minimum 4-inch-diameter well is generally required
to complete pumping tests in highly transmissive aquifers because the 2-inch submersible pumps
available today are not capable of producing a flow rate large enough for meaningful pumping
tests. In areas with fairly uniform aquifer materials, pumping tests can be completed in
uncontaminated areas and the results used to estimate hydraulic conductivity in the contaminated
area. Pumping tests should be conducted in narrowly screened wells that are screened in the most
transmissive zones in the aquifer.
Slug Tests
Slug tests are a commonly used alternative to pumping tests. They are relatively easy to conduct
and, in general, produce reliable information. One commonly cited drawback to slug testing is
that this method generally gives hydraulic conductivity information only for the area immediately
surrounding the monitoring well. Slug tests do, however, have two distinct advantages over
pumping tests; they can be conducted in 2-inch monitoring wells, and they produce no water. If
slug tests are going to be relied upon to provide information on the three-dimensional distribution
of hydraulic conductivity in an aquifer, multiple slug tests must be performed. It is not advisable
to rely on data from one slug test in one monitoring well. Because of this, slug tests should be
256
-------
conducted at several monitoring wells at the site. Like pumping tests, slug tests should be
conducted in wells that are narrowly screened in the mosttransmissive zones in the aquifer.
Hydraulic Gradient
The hydraulic gradient is the change in hydraulic head between two points over the distance
between these points. Hydraulic gradients are most easily visualized if all known (measured)
heads are portrayed on a contoured map. This is done automatically with the Platform using one
of the kriging options. A static representation of the hydraulic head map is not sufficient because
seasonal variations in ground water flow direction can have a profound influence on contaminant
transport. Sites in upland areas are less likely to be affected by seasonal variations in ground
water flow direction than those sites situated near surface water bodies such as rivers and lakes.
In situ measured gradients are the most commonly used variables for the calibration of the
groundwater flow model (advective part of the contaminant migration problem).
To determine the effect of seasonal variations in ground water flow direction on contaminant
migration, quarterly ground water level measurements should be taken over a period of at least 1
year. For many sites, historic data to that effect already exist.
Processes Causing an Apparent Reduction in Total Contaminant Mass
Several processes cause a reduction in contaminant concentrations and an apparent reduction in
the total mass of contaminant in a system. Processes causing an apparent reduction in
contaminant mass include dilution, sorption, and hydrodynamic dispersion. In order to determine
the mass of contaminant removed from the system it is necessary to correct observed
concentrations for the effects of these processes. The following sections give a brief overview of
these processes and their evaluation with the Platform.
To estimate the degree of biodegradation, it is important to adjust measured BTEX
concentrations for those processes that cause a concentration reduction without reduction in
contaminant mass. This is accomplished by normalizing the measured concentration of each of
the BTEX compounds to the concentration of a tracer that is at least as sorptive as the BTEX
compounds, but which is biologically inactive. Two trace chemicals found in fuel hydrocarbon
plumes are trimethylbenzene and tetramethylbenzene (Cozzarelli et al, 1994). This aspect of the
data collection is very important for the calibration phase of the BIOPLUME HI model. These
parameters are considered as constant parameters throughout the duration of the simulation.
Dialog boxes allow to enter their appropriate values.
Dilution
Dilution results in a reduction in contaminant concentrations. It can be caused by the improper
vertical extent of the screened interval of the monitoring wells, or by infiltration which causes an
apparent reduction in contaminant mass by mixing with the contaminant plume, thereby causing
dilution. Monitoring wells screened over large vertical distances may dilute ground water
257
-------
samples by mixing water from clean aquifer zones with contaminated water during sampling.
This problem is especially relevant for dissolved-phase BTEX contamination which may remain
near the ground water table for some distance downgradient of the source. To avoid potential
dilution, monitoring wells should be screened over relatively short vertical intervals (less than 5
feet), and nested wells should be used to define the vertical extent of contamination in the
saturated zone.
Sorption (Retardation)
The retardation of organic solutes caused by sorption is an important consideration when
modeling intrinsic remediation. Sorption of a contaminant to the aquifer matrix results in an
apparent decrease in contaminant mass that must be accounted for. Dissolved oxygen and other
electron acceptors present in the ground water travel at the advective transport velocity of the
ground water. Any slowing of the solute relative to the advective transport velocity of the ground
water allows replenishment of electron acceptors into upgradient areas of the plume. Sorption
and retardation are explicitly taken into account in the Platform (menu "Domain/Chemical
Properties").
Hydrodynamic Dispersion
Dispersion is the expansion of a plume in the apparent absence of ground water flow (due to
subgrid scale movement, from subgrid scale to Brownian motion). For intrinsic bioremediation
the dispersion of organic solutes in an aquifer is an important consideration because the
dispersion of a contaminant into uncontaminated portions of the aquifer allows the solute plume
to mix with uncontaminated ground water containing higher concentrations of electron acceptors.
B.2.4 Optional Confirmation of Biologic Activity
Extensive evidence showing that biodegradation of fuel hydrocarbons frequently occurs under
natural conditions can be found in the literature. The following sections describe two techniques
that may be used to show that microorganisms capable of degrading fuel hydrocarbons are
present at a given site.
Field Dehydrogenase Test
The field dehydrogenase test is a qualitative method used to determine if aerobic bacteria are
present in an aquifer in quantities capable of biodegrading fuel hydrocarbons. Positive results
indicate that a sufficient number of microorganisms capable of aerobic metabolism and/or
denitrification are present in the aquifer. However, a negative result for the dehydrogenase test
gives no indication of the relative abundance of anaerobic microorganisms capable of utilizing
sulfate, iron III, or carbon dioxide during biodegradation.
258
-------
Microcosm Studies
Microcosm studies are necessary only when there is considerable skepticism about the
biodegradation of fuel hydrocarbons at a specific site. If more evidence of intrinsic remediation
of fuel hydrocarbons is required, then a microcosm study using site-specific aquifer materials and
contaminants can be undertaken. Microcosm studies conducted using aquifer materials collected
in a contaminated area at a site can be used to show that the microorganisms necessary for
biodegradation are present and can be used as a good line of evidence to support the intrinsic
remediation demonstration. Microcosm studies also provide site-specific estimates of
biodegradation rate constants that can be used to verify rates of biodegradation measured in the
field. It should be kept in mind, however, that the preferable method of fuel hydrocarbon
biodegradation rate constant determination is by in situ field measurement. The collection of
material for the microcosm study, the procedures used to set up and analyze the microcosm
study, and the interpretation of the results of the microcosm study, must follow EPA standards.
B.3 Refining the Conceptual Model
The additional site investigation data must first be used to refine the conceptual model. This
refinement is facilitated by estimation of the rate of ground water flow, and the rates of
dispersion, sorption, dilution, and biodegradation. The results of these calculations are then used
to scientifically document the occurrence and the rates of natural biodegradation and to help
model intrinsic remediation. No single piece of data is sufficient to successfully support the
intrinsic remediation option at a given site. Because the burden of proof is on the proponent, all
available data must be integrated in such a way that the evidence in support of intrinsic
remediation is sufficient and irrefutable.
Conceptual model refinement involves integrating newly gathered field data to the preliminary
conceptual model that was developed based on previously existing site-specific data. This
involves integrating into the Platform the newly obtained data to develop an accurate three-
dimensional representation of the hydrogeologic and contaminant migration system. Conceptual
model refinement consists of several steps including boring log preparation, hydrogeologic
section preparation, potentiometric surface map preparation, contaminant contour map
preparation, and preparation of electron acceptor and metabolic byproduct contour maps.
Geologic Boring Logs
Geologic boring logs are entered in the Platform for all subsurface materials encountered during
the soil boring or cone penetrometer testing (CPT) phase of the field work. Description of the
aquifer matrix includes relative density, color, major textural constituents, minor constituents,
porosity, relative moisture content, plasticity of fines, cohesiveness, grain size distribution,
structure or stratification, relative permeability, and any other significant observations such as
visible fuel or fuel odor. It is also important to correlate the results of volatiles screening using
headspace vapor analysis with depth intervals of geologic materials. The depth of lithologic
contacts and/or significant textural changes should be recorded to the nearest 0.1 foot. This
resolution is necessary because preferential flow and contaminant migration paths may be limited
to stratigraphic units as thin as a few inches.
259
-------
Cone Penetrometer Data
Cone penetrometer logs data come in the form of the ratio of sleeve friction to tip pressure.
Cone penetrometer logs may also contain fluid resistivity data and estimates of aquifer hydraulic
conductivity. To provide meaningful data, the cone penetrometer must be capable of providing
stratigraphic resolution on the order of 3 inches. To provide accurate stratigraphic information,
cone penetrometer logs must be correlated with continuous subsurface cores. Cone penetrometer
logs are a cost effective means of completing the hydrogeologic section information initially
based on cores.
B.3.1 Hydrogeologic Sections
Hydrogeologic sections are entered in the Platform based on boring logs or CPT data. A
minimum of two hydrogeologic sections are required, one parallel to the direction of ground
water flow and one perpendicular to the direction of ground water flow. Hydraulic head data
including potentiometric surface and/or water table elevation data are automatically generated by
the Platform. These sections are useful in locating potential preferential contaminant migration
paths and in modeling the site using the simulation models of the program.
B.3.2 Potentiometric Surface or Water Table Maps
A potentiometric surface or water table map is a two-dimensional graphic representation of
equipotential lines shown in plan view. These maps are generated automatically by the Platform
based on water level measurements and surveyor's data. Because ground water flows from areas
of high hydraulic head to areas of low hydraulic head, such maps are used to estimate the
probable direction of plume migration and to calculate hydraulic gradients. Care must be
exercised to use water levels measured in wells screened in the same relative position within the
same hydrogeologic unit. To document seasonal variations in ground water flow, separate
potentiometric surface or water table maps should be prepared for quarterly water level
measurements taken over a period of at least one year. In areas with mobile NAPL, a correction
must be made for the water table deflection caused by the NAPL. Typical contours of the
hydraulic heads are shown in Figure B.8 as they appear on the computer screen.
260
-------
•.i-?m EI | -fc. : : i_i. T.T*. :.
? i I U -^/i * i "'^Sk^-i^lt&sliilli
L^J^^L-JK^iiL. -L- li-L.J..iJ^.Mi'lS::m::;d:;;M:;l;:M^:ii!g
y ^ps=^p^^™i
'^
-------
biodegradation. The electron acceptor and metabolic byproduct contour maps provide evidence
of the occurrence of intrinsic remediation at a site.
Electron Acceptor Contour Maps
Contour maps are needed for the electron acceptors including dissolved oxygen, nitrate, and
sulfate. During aerobic biodegradation, dissolved oxygen concentrations will decrease to levels
below background. Similarly, during anaerobic degradation, the concentrations of nitrate and
sulfate will be seen to decrease to levels below background. The electron acceptor contour maps
allow interpretation of data on the relative migration and degradation rates of contaminants in the
subsurface. The Platform allows direct input of all these parameters. Thus, electron acceptor
contour maps provide visible evidence of biodegradation and a visual indication of the
relationship between the contaminant plume and the various electron acceptors.
Metabolic Byproduct Contour Maps
Contour maps should be prepared for the metabolic byproducts iron n, sulfide, and methane.
During anaerobic degradation, the concentrations of iron n, sulfide, and methane are seen to
increase to levels above background concentrations. These maps allow interpretation of data on
the microbial degradation of fuel hydrocarbons and the relative migration and degradation rates
of contaminants in the subsurface. Thus, metabolic byproduct contour maps provide visible
evidence of biodegradation and a visual indication of the relationship between the contaminant
plume and the various metabolic byproducts.
Typical contour maps of BTEX, electron acceptors and by-products as generated by the Platform
are shown below:
• Figure B.9 shows a typical BTEX contour map
• Figure B. 10 shows the measured Oxygen plume
• Figure B. 11 shows the measured Nitrate plume
• Figure B. 12 shows the measured Sulfate plume
• Figure B. 13 shows the measured Methane plume
• Figure B. 14 shows the measured Ferrous Iron plume.
262
-------
Nortn
POC
POC
Figure B.9 Typical BTEX Contour Map.
Figure B.10 Measured Oxygen Plume.
263
-------
.' \\'. ..-./A-^iiliplpiiali
; \ \ •
Figure B. 11 Measured Nitrate Plume.
.•norm
\\
Figure B. 12 Measured Sulfate Plume.
264
-------
/ '
I ( Harrington^
UL.QOP
N>
•'XX'IXN.
. Hortft
^1
POC
*--..s_,jj\ - K,\VLTM
POC
POC
Figure B. 13 Measured Methane Plume.
/ •
(( Harrington
HLoop
N
Hortfi
POC
POC
POC
Figure B. 14 Measured Ferrous Iron Plume.
265
-------
B.4 Calculations and Sorting of Raw Data
Several estimations must be made prior to the full implementation of the simulation model to
predict future trends of contaminant migration. These calculations include:
• An in depth comparison of the hydrocarbons, electron acceptors and by-products
plumes to estimate biodegradation rate constants,
• Sorption and retardation rates,
• fuel/water partitioning calculations,
• ground water flow velocity calculations.
Each of these calculations is discussed in the following sections.
B.4.1 Analysis of Contaminant, Electron Acceptor and Metabolic
Byproduct Data
The extent and distribution (vertical and horizontal) of contaminant and electron acceptor and
metabolic byproduct concentrations are of paramount importance in documenting the occurrence
of biodegradation of fuel hydrocarbons and in simulation model implementation.
Electron Acceptors and BTEXData
Dissolved oxygen concentrations below background in an area with fuel hydrocarbon
contamination are indicative of aerobic hydrocarbon biodegradation. Similarly, nitrate and
sulfate concentrations below background in an area with fuel hydrocarbon contamination are
indicative of anaerobic hydrocarbon biodegradation. These relationships can be established on
the basis of the Platform generated contour maps. Generally, dissolved oxygen and nitrate are
used in areas with dissolved-phase fuel- hydrocarbon contamination at rates which are
instantaneous relative to the average advective transport ground water velocity. This results in
the consumption of these compounds at a rate approximately equal to the rate at which they are
replenished by advective flow processes. For this reason, the use of these compounds as electron
acceptors in the biodegradation of dissolved-phase fuel-hydrocarbons is a mass-transport-limited
process (Borden and Bedient, 1986; Wilson et al., 1985). The use of dissolved oxygen and
nitrate in the biodegradation of dissolved-phase fuel-hydrocarbons can be modeled using the
Platform.
Microorganisms generally utilize sulfate, iron IE, and carbon dioxide in areas with dissolved-
phase fuel-hydrocarbon contamination at rates that are slow relative to the advective transport
velocity of ground water. This results in the consumption of these compounds at a rate that could
be slower than the rate at which they are replenished by advective flow processes. Therefore, the
use of these compounds as electron acceptors in the biodegradation of dissolved-phase fuel-
hydrocarbons may be a reaction-limited process that can be approximated by first-order kinetics.
266
-------
The BIOPLUME in model uses a first-order rate constant to model such biodegradation.
Determination of first-order decay rate constants is discussed in the next section.
Metabolic Byproduct and BTEX Data
Elevated concentrations of the metabolic byproducts Iron n and methane in areas with fuel
hydrocarbon contamination are indicative of hydrocarbon biodegradation. Contour maps can be
used to provide visible evidence of these relationships.
Indicative of the existence of these biodegradtion processes are the contour maps described in
the previous section. To further examine their delicate interaction, the Platform offers the
possibility to compare various distributions of the hydrocarbons versus the electron acceptors
and by-products. Figure B. 15 for example, clearly shows the Oxygen depletion in the area of
the corresponding hydrocarbon plume. Similar patterns are observed for nitrates and
Sulfates, (Figure B.I6 and B.17). Figures B.I8 and B.19 by comparison show the creation of
byproducts, namely ferrous iron and methane, in plumes similar in shape to the hydrocarbon
plume.
BTEX
Oxygen Demand
August 1993 DT=000 days
Figure B. 15 Oxygen Depletion.
267
-------
BTEX
Nitrate
August 1993 DT=000 days
Figure B. 16 Nitrate Depletion.
BTEX
Sulfate
August 1993 DT=000 days
Figure B. 17 Sulfate Depletion.
268
-------
BTEX
Methane
August 1993 DT=000 days
Figure B. 18 Methane Creation.
BTEX
Ferrous Iron
POC i
POC :
August 1993 DT=000 days
Figure B.I9 Ferrous Iron Creation.
269
-------
B.4.2 Sorption and Retardation Calculations
Contaminant sorption and retardation calculations should be made based on the total organic
carbon (TOC) content of the aquifer matrix and the organic carbon partition coefficient (Y) of
each contaminant. The average TOC concentration from the most transmissive zone in the
aquifer should be used for retardation calculations. At a minimum, these calculations should be
completed for each of the BTEX compounds and any tracers. Sorption and retardation
calculations are described in the next section.
B.4.3 Fuel/Water Partitioning Calculations
If NAPL remains at the site, fuel/water partitioning calculations should be made to account for
the partitioning from these phases into the dissolved phase in ground water. Several models for
fuel/water partitioning have been proposed in recent years, including those by Hunt et al. (1988),
Johnson and Pankow (1992), Cline et al. (1991) and Bruce et al (1991). The models presented
by Cline et al. (1991) and Bruce et al. (1991) represent equilibrium partitioning, i.e. they are the
most conservative models. Equilibrium partitioning is conservative because it predicts the
maximum dissolved-phase concentration when LNAPL in contact with water is allowed to reach
equilibrium. The results of these equilibrium partitioning calculations can be used in a
simulation model to simulate a continuous source of contamination.
B.4.4 Ground Water Flow Velocity Calculations
The average linear ground water flow velocity of the most transmissive aquifer zone containing
contamination should be calculated to check the accuracy of the ground water flow simulation
model and to allow calculation of first-order biodegradation rate constants.
B.4.5 Anaerobic Biodegradation Rate Constant Calculations
One of the advantages of BIOPLUME HI is that it does account for anaerobic degradation to the
same degree of accuracy as for aerobic degradation, i.e. one can also use a first-order anaerobic
decay constant.
In order to calculate anaerobic rate constants for each chemical, the apparent degradation rate
must be dissociated from the effects of dilution and volatilization.
This is accomplished by normalizing the concentration of each contaminant to the concentration
of a tracer that is at least as sorptive, but which is biologically inactive. Two chemicals that have
good potential as tracers that are found in fuel hydrocarbon plumes are trimethylbenzene and
tetramethylbenzene. Both have been shown to be recalcitrant under anaerobic conditions. It is
important to note however, that all refined fuel components will degrade in a ground water
system undergoing intrinsic remediation. Trimethylbenzene and tetraethylbenzene, while being
recalcitrant under anaerobic conditions, will degrade under aerobic conditions.
270
-------
When sulfate is being used as an electron acceptor and sulfate concentrations are greater than 10
milligrams per liter (mg/L), the first-order rate constant is appropriate. To adequately describe
biodegradation rates using a first-order rate constant during methanogenesis, the total alkalinity
for the system should be greater than about 50 mg/L. An example anaerobic biodegradation rate
constant calculation is given in the next section. The Platform allows direct input of anaerobic
electron acceptor data so that both aerobic and anaerobic degradation can be simulated.
B.5 Simulate Intrinsic Remediation Using the Platform
Modeling of the intrinsic bioremediation processes is necessary because we need to predict the
migration and attenuation of the contaminant plume over time. Unlike other technologies,
Intrinsic Remediation requires the ability to predict the future behavior of a contaminant plume.
Indeed the whole IR technology is based on the claim that the applicant has deep knowledge of
the behavior of the contaminant plume over time, and can guarantee with a high degree of
reliability that the plume will experience loss of contaminant mass and that in any event it will
not reach receiving points along established exposure pathways, and that concentrations will not
exceed regulatory standards at compliance points.
Simulation can be used to do site-specific predictions of the fate and migration of solutes under
governing physical, chemical, and biological processes, provided the model has been calibrated
and verified to site data. Therefore, the simulation model cannot prove or disprove that intrinsic
remediation is occurring at a given site alone. But in conjunction with site-specific data analysis
simulation models can be very powerful indeed. A calibrated and verified model can prove the
case that it reproduces historical data and can be used in a predictive mode.
Scenario analysis vs. Prediction
A calibrated and verified model will be used in one of two capacities: I/ to predict (exactly) what
will happen in the immediate or longer term future (like the weather forecast in the nightly
news); and, II to predict extreme conditions (scenario analysis). The distinction is subtle but
important: any difference between scenario predictions and the subsequent plume behavior
should not be used to discredit the model. Instead, these differences should be first explained
(e.g. wet year, or presence of new contaminant source) and then used to strengthen the validity of
the scenario predictions. The exact future prediction depends on imponderables such as the
hydrological cycle (wet/ dry year), exact behavior of the contamination source, which may be of
lesser importance in ascertaining that contamination will be contained. The scenario analysis
will have fulfilled its role if under the simulation conditions the plume does not exceed the safety
envelope around the site as monitored by the long term monitoring wells and the compliance
wells.
271
-------
B.5.1 Requirements for a Contaminant Biodegradation
Simulation
A ground water simulation model for intrinsic remediation is developed to formulate and
substantiate ground water restoration decisions. It is used for the following reasons:
1. To give insight into the physical bioremediation processes influencing the study area,
2. To provide predictions of system behavior under changing circumstances, and
3. To test hypotheses on system behavior by organizing the collection of additional data
to improve site characterization and increase the confidence level in the management
decisions.
B.5.2 Context of the Conceptual Model
A conceptual model is a word description of the components of a prototype contaminated aquifer
system, the "loads" or "forcing" to the system, and the processes operative on the system. This
description is made on the basis of preexisting data, regional aquifer atlases, or previous site
studies. Pictures complement word descriptions (proverbially "worth a thousand words"). A
graphical representation of the contaminated aquifer is part of the conceptual model. Figure B.20
illustrates a typical conceptual representation of a contaminated aquifer system.
Present in this conceptual model are a source of contamination, a fuel tank leaking at the surface
(a typical problem that can be handled with intrinsic remediation); the vadose or unsaturated
zone through which the Tree product' seeps; the mass of free product that "floats" atop the water
table, i.e. that portion of the aquifer is saturated with fuel; a vapor zone, i.e. unsaturated zone
filled with fuel vapors; and a zone of contact between free product and water table, where fuel is
dissolved into the saturated aquifer. The dissolved contaminant creates a plume which is
advected and dispersed by the flow of the aquifer. In most instances, the immediate concern is
about the quality of the aquifer and therefore how to control the level of concentration of the
dissolved contaminant. The rest of the phases, leaking source, free product, characterize the
release mechanism.
Of course, any long term remediation will have to start with the removal of the source and the
free product. But even that cannot be achieved one hundred percent, and therefore is controlled
by the desirable or achievable concentration levels of dissolved contaminants in the aquifer
which represents the major 'end point'. A modern approach to risk assessment then works
backwards from the accepted concentration level in the aquifer to target residual levels in the
soil.
This document deals primarily with the dissolved plume, its origin, its evolution, its simulation
and its remediation by biodegradation. And of course, a fuel leakage does not produce one
contaminant plume but many, as many as the constituents of the fuel. Not all are toxic, and
therefore the discussion focuses on the stable toxic constituents. For fuel hydrocarbons these are
the BTEX sequence (Benzene, Toluene, Ethylene, Xylene). All above discussion deals with light
272
-------
hydrocarbons (LNAPLs -light non-aqueous phase liquids). Dense organic liquids, DNAPLs,
form free product masses which tend to sink in the aquifer, that is they have a higher mobility.
They are mostly solvents which are significantly more difficult to degrade biologically.
This completes the component description of the conceptual model. But a conceptual model is
not complete without providing additional information about boundary conditions, or how the
vicinity of the aquifer near the site of interest relates to the surrounding aquifers at the regional
scale; the presence and interaction with other surface features such as rivers and ponds or drains;
and the "forcing" mechanisms or loads to the system, wells, recharge, evapotranspiration and
other losses.
The conceptual model development is arguably the most important phase for a modeling
exercise, where experience counts the most. The automated/integrated Platform breaks rank with
this tradition in two important ways:
1. Because data entry and model setup are performed by the program very efficiently, the
user can concentrate on the physical, chemical, and biological aspects of the problem
and gain experience very quickly.
2. The user does not need to switch from simple (analytic) to more complicated models:
all entered data are immediately accessible for use in testing new model setups,
adjusted numerical grids, boundary conditions etc.
These are extraordinary advantages that the Platform offers when dealing with a subject as
complicated as the modern multi-disciplinary theory of biodegradation.
273
-------
Figure B.20 A conceptual Intrinsic Remediation Model
B.5.3 Steps Specific to Biodegradation Modeling
Most of the elements that go into preparing for a biodegradation modeling exercise are already
mentioned in the conceptual model phase described above.
The first thing that needs to be done is the determination of the modeling domain, that is the
geographic extent of the simulation area. Typically this domain will start from the area of
interest (for example a waste site or a well field) and extend to where secure boundary conditions
may exist (that is conditions that are unaltered by 'forcing' that may be imposed within the
simulation domain), or beyond the radius of influence of anticipated forcing mechanisms. With
variable spacing grids available one should err on the side of safety and retain a larger rather than
a narrower domain. The element to consider in defining the simulation domain is a bitmap of the
site showing as many features as available, including topographic contour lines, surface features,
lakes, rivers, drains, observed hydraulic heads and plume delineation. This bitmap is imported in
the program and "registered" to the scales of the simulation domain defined earlier. It provides
the canvas on which to build the ground water intrinsic remediation model using the Platform
tools.
Grid definition is automated in the Platform and offers absolutely no inconvenience to the
modeler on two counts:
1. It is drawn by specifying the increments or the number of elements in the x (top) and
y (left) axes spanning the domain. The drawn grid can then be graphically edited by
moving horizontal or vertical grid lines at will; or inserting new lines (rows or
columns) as the need arises.
274
-------
2. The aquifer properties (conductivities, porosity, dispersion) are interpolated to the
grid centers from observed data points by Kriging. A complete assortment of
advanced kriging options are available for the modeler to control the geostatistical
interpolation error. In fact this is one of the strong points of the Platform because
once the raw data are entered the modeler does not ever have to revisit them although
he/ she may test a wide variety of different grid configurations.
Next is the depiction of the soil medium aquifer stratification or layering. In fact a distinction is
drawn between soil strata and aquifer layers: strata are physical units in the soil medium which
have different conductances and other properties and can be aquitards or even confining units.
The last piece of data necessary to perform a simulation pertains to the initial and boundary
conditions. Contaminant and oxygen or other terminal electron acceptor concentrations must be
known or assumed (constant for example) at some point in time. A simulation model always
starts from some known initial conditions and marches the solution in time. Often these
conditions are measured in the field from observation wells or piezocones at one or more points
in time, and are interpreted into concentration contours. These data also are interpolated to the
model grid via kriging as explained above. In fact, more than one set of data are needed so as to
calibrate the model against one set and validate it against additional sets of observed heads.
With all the necessary data entered, the remaining tasks for the modeler are to "create" his/ her
conceptual model interactively on screen with the program tools by selecting and specifying
features to include, for example rivers, ponds, drains, wells and their pumping schedule, as well
as man-made features such as liners, slurry walls, even geologic faults, and of course the
contaminant sources. With the interactive/ graphical selection of appropriate boundary
conditions, the modeler is then ready to fire up their first simulation. Nothing to be timid about:
if there is an error in the data or model setup, it will become immediately apparent in the
distorted simulated results. Or the Platform will guide you about any discrepancies that may not
allow you to activate the ground water biodegradation model. Corrections can then be made on
the spot so that one can proceed very quickly to a series of trial-and-errors.
The Platform in fact gives a good name to the old trial-and-error procedure; only it does it while
increasing efficiency and productivity!
In the following sections we look in more detail into the types of data that are needed for a
ground water biodegradation simulation; and into the crucial task of model calibration and
validation. Of course, the specific steps for biodegradation modeling presume a calibrated/
verified flow model.
B.5.4 Calibration of the Bioremediation Model
A ground water bioremediation model is an integrated suite of interacting simulators for the flow
through the porous medium, interaction with surface waters, evapotranspiration losses, drains,
other forcing mechanisms such as wells and recharge, and boundary conditions; and the
advection and dispersion of constituent plumes and their interaction. A calibrated model is one
where there is a balance between grid resolution and data accuracy, layering and the vertical
275
-------
structure of the hydraulic head and concentration distributions, the number of processes and their
importance or influence on the simulated results. Here is how to proceed to develop a balanced
model.
The calibration picture is at once complicated but also tractable because there is a hierarchy to
follow. First, we must begin with a calibrated flow model, that is the flow model can be
calibrated independently of the migration or biodegradation processes. Of course unresolved
discrepancies at the migration or biodegradation level can point to necessary adjustments in
aquifer layer thickness' or conductivities which are the primary flow model calibration
parameters. This is why the calibration process is an iterative one. But do not equate iterations
with confusion: iterations bring order to a complex process.
The calibrated flow model will then be used to calibrate the intrinsic remediation migration
model for a conservative constituent. This will allow to calibrate the advective and dispersive
properties of the aquifer. For example, starting from one set of observed concentrations as initial
conditions the 1-species simulation will be used to reproduce the concentrations as observed at a
later time.
With a calibrated advection/ dispersion model one then will attempt to model degradation
processes, chemical reactions or biodegradation. For the case of biodegradation, a second plume
will be simulated of dissolved oxygen for aerobic conditions, or a "compound" plume for the
case where additional degradation conditions exist, for example denitrifying conditions. The
calibration consists of comparing the simulated degraded hydrocarbon plume (dissolved BTEX)
and the corresponding depletion of dissolved oxygen, nitrates, other electron acceptors and by-
products against the observed (measured) concentrations. The calibration parameters are the
stoichiometric ratio of hydrocarbon consumption to oxygen or nitrate; alternatively, the
interaction between electron acceptor and hydrocarbon plumes may be modeled instantaneously
or using Monod kinetics theory. Additional tuning parameters such as the reaeration coefficient
can also be used to resolve any residual discrepancies.
These are the essential steps to follow for a systematic biodegradation simulation model
calibration. In the foregoing discussion little mention is made about source mechanisms. This
and other features are discussed next.
Identifying Loadings, Sources for the Simulation Model
First determine what drives the contaminant migration system: how does recharge from
precipitation, influx from neighboring regions of the aquifer or surface water impoundments
affect the migration process; and where are the withdrawals, pumping wells, drains, seepage into
rivers or lakes. Then determine how the cyclicity of the aquifer regime, either annual or drought-
wet year cycle can affect the long term behavior of contaminant plumes. Then consider the
contamination source mechanism. There are several possibilities depending on the data
available: it can be simulated as a recharge zone with a given flux and concentration; or it can be
simulated as a zone of constant concentration at the maximum dissolution rate below the free
product mass; or it can be emulated by a series of injection wells. Before proceeding with a
simulation however, there is a series of boundary and initial conditions to define.
276
-------
Identifying Boundary and Initial Conditions of the Simulation Model
Boundary conditions are usually the concentrations prevailing at the boundary of the modeled
domain. When they are known from observations and when they are not affected by processes
taking place in the simulation domain (e.g. cone of depression from pumping wells) they may as
well be considered as constant. Under these conditions the model can be calibrated for
conservative constituents as discussed above. If other processes are important, for example
recharge from precipitation or interaction with surface waters, then the fixed boundary condition
can be relaxed and the model can be calibrated explicitly for the source mechanism. Thus, the
individual processes can be isolated and calibrated separately in a hierarchical manner. The
Platform offers automation and integration which allow the user to perform all these formidable
looking calibration tasks very efficiently, accurately and effortlessly.
Identifying Soil Layering and Grid Resolution for the Simulation Model
Finally, a word must be said about grid resolution. It should be looked at as a calibration
parameter in the sense that the user should start out simple with a clearly distinctive coarse grid.
After initial calibration and if the accuracy of the contaminant plume field data warrant it then the
user can consider increasing the grid resolution for final calibration. Once the data have been
entered in the program data structure, switching grid resolution is easy and painless. This is how
with the Platform, the complicated steps for a professional model calibration do become tangible.
B.6 Conduct an Exposure Assessment
After the rates of natural attenuation have been documented and predictions of the future extent
and concentration of the contaminant plume have been made using the BIOPLUME HI fate and
migration model, the proponent of intrinsic remediation should prepare a permit application for
implementation of this remedial option. Supporting the intrinsic remediation option generally
will involve implementation of an exposure assessment. The results of numerical fate and
migration modeling are central to the exposure assessment. Conservative model input
parameters should give conservative estimates of contaminant plume migration. From this
information, the potential impacts to human health and the environment from contamination
present at the site can be estimated. The exposure assessment in support of the remediation
option is described in a separate Risk Assessment AFCEE document.
B.7 Prepare Long-Term Monitoring Plan
The long-term monitoring plan consists of locating ground water monitoring wells and
developing a ground water sampling and analysis strategy. This plan is used to monitor plume
migration over time and to verify that intrinsic remediation is occurring at rates sufficient to
protect potential downgradient receptors. The long-term monitoring plan should be developed
based on the results of the BIOPLUME III model simulations.
Point-of-compliance (POC) monitoring wells are wells that are installed at locations
downgradient of the contaminant plume and upgradient of potential receptors. POC monitoring
277
-------
wells are generally installed along a property boundary or at a location approximately 5 years
downgradient of the current plume at the seepage velocity of the ground water or 1 to 2 years
upgradient of the nearest downgradient receptor, whichever is more protective. The final number
and location of POC monitoring wells will depend on regulatory considerations. Long-term
monitoring wells are wells that are placed upgradient of, within, and immediately downgradient
of the contaminant plume. These wells are used to monitor the effectiveness of intrinsic
remediation in reducing the total mass of contaminant within the plume. Requirements are, one
well upgradient of the contaminant plume, one well within the anaerobic treatment zone, one
well in the aerobic treatment zone and one well immediately downgradient of the contaminant
plume. The final number and location of longterm monitoring wells will depend on regulatory
considerations.
Figure B.21 shows a hypothetical long-term monitoring scenario. The results of a numerical
model such as BIOPLUME HI can be used to help locate both the long-term and POC monitoring
wells. In order to provide a valid monitoring instrument, all monitoring wells must be screened
in the same hydrogeologic unit as the contaminant plume. This generally requires detailed
stratigraphic correlation. To facilitate accurate stratigraphic correlation, detailed visual
descriptions of all subsurface materials encountered during borehole drilling should be prepared
prior to monitoring well installation. The final placement of all monitoring wells should be
determined in collaboration with the appropriate regulators.
The ground water sampling and analysis plan should be prepared in conjunction with POC and
long-term monitoring well placement. Analyses should be limited to determining BTEX,
dissolved oxygen, nitrate, and sulfate concentrations. Water level and NAPL thickness
measurements must be made during each sampling event. Sampling frequency is dependent on
the final placement of the POC monitoring wells. For example, if the POC monitoring wells are
located 2 years upgradient of the nearest downgradient receptor, then an annual sampling
frequency should be sufficient. If the POC monitoring wells are located 1 year upgradient of the
potential receptor, then a semiannual sampling frequency should be sufficient. The final
sampling frequency should be determined in collaboration with regulators.
278
-------
Point-of- C oni p I lane *
Me niter ing Well
Lonj-Tcii
Well
Figure B.21 Typical Long-Term Monitoring Strategy
B.8 Additional Reading
Blake, S.B., and Hall., R.A., 1984, Monitoring petroleum spills with wells - some problems and
solutions: In, Proceedings of the Fourth National Symposium on Aquifer Restoration and
Ground Water Monitoring: May 23-25, 1984, p. 305-3 1 0.
Borden, R.C. and P.B. Bedient, 1986, Transport of dissolved hydrocarbons influenced by oxygen
limited biodegradation - theoretical development: Water Resources Research, v, 22, no. 13, p.
1973-1982.
Bouwer, E.J., 1992, Bioremediation of Subsurface Contaminants, In R. Mitchell, editor,
Environmental Microbiology: Wiley, New York, p. 287-318.
Bouwer, H., and Rice, R.C., 1976, A Slug Test for Determining Hydraulic Conductivity of
Unconfined Aquifers With Completely or Partially Penetrating Wells: Water Resources
Research, v. 12, no. 3, p. 423-428.
Bouwer, H., 1989, The Bouwer and Rice slug test - an update: Ground Water, v. 27, no. 3, p.
304-309.
279
-------
Bruce, L., Miller, T., and Hockman, B., 1991, Solubility versus equilibrium saturation of
gasoline compounds - a method to estimate fuel/water partition coefficient using solubility, In,
A. Stanley, editor, NWWA/API Conference on Petroleum Hydrocarbons in Ground water-
NWWA/API, p. 571-582.
Brown, D.S. and Flagg, E.W., 1981, Empirical prediction of organic pollutant sorption in natural
sediments: Journal of Environmental Quality, v. 10, no. 3, p. 382-386.
Briggs, G.G., 1981, Theoretical and experimental relationships between soil adsorption, octanol-
water partition coefficients, water solubilities, bioconcentration factors, and the parachlor:
Journal of Agriculture and Food Chemistry, v. 29, p. 1050-1059.
Chiou, CT., Porter, P.E., and Schmedding, D.W., 1983, Partition equilibria of nonionic organic
compounds between soil organic matter and water: Environmental Science and Technology: v.
17, no, 4, p. 227-23 1.
Cline, P.V., Delfino, J.J., and Rao, P.S.C., 1991, Partitioning of aromatic constituents into water
from gasoline and other complex solvent mixtures: Environmental Science and Technology, v.
25, p. 914-920.
Cozzarelli, I.M., Baedecker, M.J., Eganhouse, R.P., and Goerlitz, D.F., 1994, The geochemical
evolution of low-molecular-weight organic acids derived from the degradation of petroleum
contaminants in groundwater: Geochimica et Cosmochimica Acta, v. 58, no. 2, p. 863-877.
Dawson K.J. and Istok, J.D., 1991, Aquifer Testing - Design and analysis of pumping and slug
tests: Lewis Publishers, Chelsea, Michigan, 344 p.
de Pastrovich, T.L., Baradat, Y., Barthel, R., Chiarelli, A., and Fussell, D.R., 1979, Protection of
groundwater from oil pollution: CONCAWE, The Hague, 61 p.
Fetter, C.W., 1993, Contaminant Hydrogeology: MacMillan, New York, New York, 458 p.
Hampton, D.R., and Miller, P.D.G., 1988, Laboratory investigation of the relationship between
actual and apparent product thickness in sands.
Hassett, J.J., Means, J.C., Banwart, W.L., and Wood, S.G., 1980, Sorption Properties of
Sediments and Energy-Related Pollutants: EPA/600/3-80-041, U.S. Environmental Protection
Agency, Washington, D.C.
Hughes, J.P., Sullivan, C.R., and Zinner, R.E., 1988, Two techniques for determining the true
hydrocarbon thickness in an unconfmed sandy aquifer: In Proceedings of the Petroleum
Hydrocarbons and Organic Chemicals in Ground water: Prevention, Detection, and Restoration
Conference: NWW A/API, p. 291 -314.
Hunt, J.R., Sitar, N., and Udell, K.S., 1988, Nonaqueous phase liquid transport and cleanup, 1.
Analysis of mechanisms: Water Resources Research, v. 24, no. 8, p. 1247-1258.
280
-------
Hvorslev M.J., 195 1, Time lag and soil permeability in ground-water observations: United
States Corps of Engineers Waterways Experiment Station Bulletin 36 Vicksburg Mississippi 50
P-
Karickhoff, S.W., Brown, D.S., and Scott, T.A., 1979, Sorption of hydrophobic pollutants on
natural sediments: Water Resources Research, v. 13, p. 241-248.
Karickhoff, S.W., 1981, Semi-empirical estimation of sorption of hydrophobic pollutants on
natural sediments and soils: Chemosphere, v. 10, p. 833-846.
Kemblowski, M.W., and Chiang, C.Y., 1990, Hydrocarbon thickness fluctuations in monitoring
wells: Ground Water v. 28, no. 2, p. 244-252.
Kenaga, E.E., and Goring, C.A.I., 1980, ASTM Special Technical Publication 707- American
Society for Testing Materials, Washington, D.C.
Johnson, R.L., and Pankow, J.F., 1992, Dissolution of dense chlorinated solvents in ground
water, 2. Source functions for pools of solvents: Environmental Science and Technology, v. 26,
no. 5, p. 896-901.
Lenhard, R.J., and Parker, J.C., 1990, Estimation of free hydrocarbon volume from fluid levels in
monitoring wells: Ground Water, v. 28, no. 1, p. 57-67'.
Lyman, W.J., Reidy, P.J., and Levy, B., 1992, Mobility and Degradation of Organic
Contaminants in Sub surf ace Environments: C.K. Smoley, Inc., Chelsea, Michigan, 395 P.
McCall, P.J., Swann, R.L., and Laskowski, 1983, Partition models for equilibrium distribution of
chemicals in environmental compartments, In, R.L. Swann and A. Eschenroder, editors, Fate of
Chemicals in the Environment: American Chemical Society, p. 105123.
Rao, P.S.C., and Davidson, J.M., 1980, Estimation of pesticide retention and transformation
parameters required in nonpoint source pollution models, In, M.R. Overcash and J.M. Davidson,
editors, Environmental Impact of Nonpoint Source Pollution: Ann Arbor Science Publishers,
Inc., Ann Arbor, Michigan, p. 23-67.
Sellers, K.L., and Schreiber, R.P., 1992, Air sparging model for predicting ground water clean up
rate: Proceedings of the 1992 NGWA Petroleum Hydrocarbons and Organic Chemicals in
Ground Water, Prevention, Detection, and Restoration Conference, November, 1992.
Shwarzenbach, R.P., and Westall, J., 1985, Sorption of hydrophobic trace organic compounds in
ground water systems: Water Science Technology, v. 17, p. 39-55.
Vroblesky, D.A., and Chapelle, F.H., 1994, Temporal and spatial changes of terminal electron-
accepting processes in a petroleum hydrocarbon-contaminated aquifer and the significance for
contaminant biodegradation: Water Resources Research, -v. 30, no. 5, p. 1561-1570.
Walton, W.C., 1988, Practical Aspects of Ground Water Modeling: National Water Well
Association, Worthington, Ohio, 587 p.
281
-------
Wiedemeier, T.H., Guest, P.R., Henry, R.L., and Keith, C.B., 1993, The use of Bioplume to
support regulatory negotiations at a fuel spill site near Denver, Colorado, In Proceedings of the
Petroleum Hydrocarbons and Organic Chemicals in Ground water: Prevention, Detection, and
Restoration Conference: NWWA/API, p. 445 -459.
Wilson, J.T., McNabb, J.F., Ccichran, J.W., Wang, T.H., Tomson, M.B., and Bedient, P.B.,
1985, Influence of microbial adaptation on the fate of organic pollutants in ground water:
Environmental Toxicology and Chemistry, v. 4, p. 721-726.
282
------- |