United States
Environmental Protection
Agency
Office of Solid Waste and
Emergency Response
Office of Emergency and
Remedial Response
Washington DC 20460
Office of Research and
Development
Hazardous Waste Engineering
Research Laboratory
Cincinnati OH 45268
Superfund
EPA/540/2-85/001 April 1985
vvEPA
Modeling Remedial
Actions at
Uncontrolled
Hazardous Waste
Sites
-------
MODELING REMEDIAL ACTIONS
AT UNCONTROLLED HAZARDOUS WASTE SITES
by
S.H. Boutwell, S.M. Brown, B.R. Roberts,
Anderson-Nichols & Co., Inc.
2666 East Bayshore Boulevard
Palo Alto, CA 94303
and Atwood
68-03-3116
Project Officers
Douglas C. Ammon
Thomas 0. Barnwell, Jr.
OFFICE OF EMERGENCY AND REMEDIAL RESPONSE
OFFICE OF SOLID WASTE AND EMERGENCY RESPONSE
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, DC 204060
HAZARDOUS WASTE ENGINEERING RESEARCH LABORATORY
OFFICE OF ENVIRONMENTAL ENGINEERING TECHNOLOGY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF ENVRIONMENTAL PROCESSES AND EFFECTS RESEARCH
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ATHENS, GA 30613
;,,-.,on Agency,
230 Soutii Dodi-bo.-n Street ,,
Chicago, Illinois 60604
-------
NOTICE
The information in this document has been funded, wholly or in
part, by the United States Environmental Protection Agency
under Contract No. 68-03-3116 to Anderson-Nichols & Co., Inc.
It has been subject to the Agency's peer and administrative
review and has been approved for publication as an EPA
document.
This report is intended to present information on the
selection and application of predictive tools for the control
of specific problems caused by uncontrolled waste sites. It
is not intended to address every conceivable waste site
problem or all possible applications. Mention of trade names
or commercial products does not constitute endorsement or
recommendation for use.
U,S. Environmental Protection Agency
ii
-------
FOREWORD
This is one of a series of reports being published to
implement CERCLA, otherwise known as Superfund legislation.
The documents in this series explain the hazardous response
program and, in particular, the technical requirements for
compliance with the National Contingency Plan (NCP), the
analytical and engineering methods and procedures to be used
for compliance, and the background and documenting data
related to these methods and procedures. The series may
include feasibility studies, research reports, manuals,
handbooks, and other reference documents pertinent to
Superfund.
This document provides guidance on the selection and use of
models for the purpose of evaluating the effectiveness of
remedial actions at uncontrolled hazardous waste sites. It
consists of four volumes, each covering a specific facet on
modeling remedial actions, including selection of models, and
the use of simplified, analytical, and numerical models for
the evaluation of subsurface, waste control, and surface water
remedial actions. This document provides a comprehensive set
of guidelines to regulatory officials for the incorporation of
models into the remedial action planning process at Federal
and State superfund sites.
111
-------
ABSTRACT
Assessment of remedial action performance at uncontrolled
hazardous waste sites in the past has largely been
accomplished using best engineering judgement. While this may
be appropriate for many sites, there are a number of sites
where site conditions are not understood, and selection and
design of appropriate remedial actions is not apparent.
Mathematical models can be used to supplement best engineering
judgement by providing a quantitative assessment of site
conditions and remedial action performance. This may allow
more accuracy and confidence in decisions concerning technical
and cost-effectiveness of remedial actions.
Given the number and complexity of models available, the
selection and application of appropriate models for remedial
action assessment can be confusing if one is not completely
familiar with the important site criteria and available
models. This document presents model selection and use
guidelines for such a purpose.
Four volumes comprise the document. Volume 1 presents a model
selection methodology based on flow charts and matrices.
Three basic decisions form the framework: 1) is modeling
required?; 2) if so, what level of modeling is required?; and
3) what are the required model capabilities? Volumes 2, 3, and
4 describe remedial action modeling requirements and model
application guidance for simplified methods for evaluation of
subsurface and waste control actions; numerical models for the
same actions; and analytical and numerical model use for
evaluation of remedial actions in surface water, respectively.
Remedial action modeling requirements guidance includes: 1)
the type or level of model(s) needed to evaluate an action or
group of actions; 2) the model dimensionality and grid
configuration needed to represent an action; 3) model
parameter adjustments required to simulate the effects of
implementing an action, and 4) design objectives and remedial
action configurations. Techniques and literature data useful
in estimating parameter values are provided.
Model application guidance is presented in terms of: 1) the
general capabilities of different types of models, including
sources of information on specific models; 2) factors to
consider when linking different numerical models to assess
iv
-------
complicated site and remedial action conditions; 3) the steps
to follow in applying models for remedial action assessment;
4) user expertise and resource requirements; 5) alternative
ways of analyzing remedial action performance; and 6) key
assumptions and limitations affecting the use of specific
models.
-------
CONTENTS
NOTICE ii
FOREWORD iii
ABSTRACT iv
FIGURES X
TABLES xvi
ACKNOWLEDGMENTS XX
INTRODUCTION xxii
VOLUME 1: SELECTION OF MODELS FOR REMEDIAL ASSESSMENT
1. Executive Summary 1-1
2. Introduction 1-3
2.1 Purpose of this Report 1-3
2 . 2 Definition of Models 1-4
2.3 The Role of Remedial Action Assessment
in the Remedial Response Process 1-5
2.4 Framework and Organization of Report 1-7
2.5 Caveats of Use 1-9
3. The First Decision: Is Modeling Necessary? 1-11
3.1 Overview 1-11
3.2 The Decision to Model: Flow Chart 1-11
4. Methodology for Model Selection For Soil and
Ground-Water Contamination Problems 1-15
4.1 Overview 1-15
4.2 What Level of Modeling is Required?..... 1-16
4.3 Required Model Capabilities for Each Level... 1-20
4.4 Resource and Data Availability 1-29
4.5 Model Selection Criteria for Soil and
Ground-Water Contamination Problems 1-33
5. Methodology For Model Selection For Surface
Water Contamination Problems 1-35
5.1 Overview 1-35
5.2 What Level of Modeling is Required? 1-37
5.3 Required Model Capabilities for Each Level... 1-41
5.4 Resource and Data Availability 1-49
5.5 Model Selection Criteria for Surface Water
Remedial Action Assessment 1-49
References 1-53
VI
-------
VOLUME 2: SIMPLIFIED METHODS FOR SUBSURFACE AND WASTE
CONTROL ACTION
1. Introduction 2-1
1.1 Purpose of Report 2-1
1.2 Report Organization 2-2
2. Summary and Conclusions 2-3
3. Remedial Action Evaluation with Simplified
Methods 2-6
3 .1 Overview 2-6
3.2 Subsurface Control Measures 2-11
3. 3 Waste Control 2-16
4. Theory Underlying Available Simplified Methods... 2-20
4.1 Overview 2-20
4.2 Well Hydraulics 2-20
4.3 Drain Hydraulics 2-35
4.4 Ground-Water Mounding Estimation Methods... 2-42
4.5 Seepage/Infiltration Estimation Methods.... 2-47
4.6 Superposition 2-49
4. 7 Trans format ion Methods 2-60
4. 8 Conf ormal Mapping 2-65
4.9 Contaminant Transport 2-69
5. Available Hand-Held Calculator and Micro-
Computer Programs 2-76
5.1 Overview 2-76
5.2 Available Programmable, Hand-Held
Calculator Programs 2-77
5.3 Available Programs for Micro-Computers 2-87
6. Example Applications... 2-97
6.1 Overview 2-97
6.2 Example 1: Water Table Suppression with an
Interceptor Trench 2-97
6.3 Example 2: Plume Capture with a Pumping/
Injection Doublet 2-102
6.4 Example 3: Ground-Water Pumping with and
without an Impermeable Barrier 2-108
6.5 Example 4: Recirculation System for
Ground-Water Clean-Up 2-116
6.6 Example 5: Drain Recirculation System 2-119
References 2-124
VOLUME 3: NUMERICAL MODELING OF SURFACE, SUBSURFACE
AND WASTE CONTROL ACTIONS
1. Introduction 3-1
1.1 Purpose of Report 3-1
1. 2 Report Organization 3-2
vii
-------
2. Conclusions 3-4
3. Migration and Fate Processes 3-6
3. 1 Overview 3-6
3.2 Processes Controlling Movement within Zones 3-10
3.3 Processes Controlling Transfers between
Zones 3-12
3.4 Processes Controlling Transformation/
Degradation 3-13
4. Remedial Actions and Affected Processes 3-14
4.1 Overview 3-14
4.2 Surface Control 3-16
4. 3 Subsurface Control 3-21
4.4 Waste Control 3-26
5. Numerical Model Application Guidelines 3-34
5.1 Overview 3-34
5.2 Numerical Model Capabilities 3-35
5 . 3 Interactions Between Models 3-44
5.4 Model Application Process 3-48
5.5 User Expertise and Resource Requirements... 3-53
5.6 Analysis of Remedial Action Performance
Using Numerical Models 3-55
6. Remedial Action Modeling Requirements 3-66
6.1 Overview 3-66
6.2 Modeling Requirements 3-67
6. 3 Parameter Estimation Guidance 3-87
References 3-114
Appendix
A. Supporting Information on HSPF, FEMWATER/
FEMWASTE and FE3DGW/CFEST 3-124
VOLUME 4: ANALYTICAL AND NUMERICAL MODELS FOR THE
EVALUATION OF REMEDIAL ACTIONS IN SURFACE
WATER
Introduction 4-1
1.1 Background 4-1
1. 2 Purpose of Report 4-2
1. 3 Report Content 4-3
Migration and Fate 4-4
2.1 Overview 4-4
2. 2 Physical Processes 4-4
2. 3 Chemical Processes 4-10
Remedial Actions and Affected Critical
Processes 4-20
3 .1 Overview. 4-20
3.2 Dilution 4-23
3. 3 Containment Actions 4-23
3 . 4 Removal Measures 4-28
3. 5 Treatment Measures 4-29
viii
-------
4. Use of Remedial Actions and Modeling: Case
Histories 4-31
4.1 Overview 4-31
4.2 Case Histories 4-31
4. 3 Summary 4-36
5. Use of Analytical and Simplified Assessment
Techniques for Remedial Action Screening and
Assessment 4-38
5.1 Overview 4-38
5.2 Uses of Simplified Assessment Techniques 4-39
5.3 Classification of Simplified Assessment
Techniques 4-41
5.4 Analytical Models 4-46
6. Use of Numerical Models for Remedial Action
Assessment 4-50
6. 1 Overview 4-50
6.2 Capabilities of Available Codes 4-51
6.3 The Model Development and Application
Process 4-59
7. Model Requirements for Surface Water Remedial
Actions 4-62
7.1 Overview 4-62
7.2 Modeling Requirements 4-65
7. 3 Parameter Estimation Guidance 4-75
References 4-90
IX
-------
FIGURES
Page
VOLUME 1: SELECTION OF MODELS FOR REMEDIAL ASSESSMENT
2.1 Three basic decisions in model selection 1-8
3.1 Flow chart to determine if modeling is required. 1-12
4.1 Flow chart to determine the level of modeling
required for soil and ground-water systems 1-17
4.2 Flow chart for required model capabilities for
soil and ground-water systems 1-22
5.1 Flow chart to determine the level of modeling
required for surface water systems 1-38
5.2 Flow chart for required model capabilities for
surface water systems 1-42
VOLUME 2: SIMPLIFIED METHODS FOR SUBSURFACE AND
WASTE CONTROL ACTION
4.1 Drawdown around a pumping well in a confined
aquifer 2-23
4.2 Drawdown around a pumping well in a leaky
aquifer 2-25
4.3 Assumed and actual flow patterns for a fully
penetrating well in a leaky aquifer 2-27
4.4 Drawdown around a pumping well in a water table
aquifer 2-30
4.5 Ideal conditions for applying a drain hydraulics
method based on Dupuit-Forchheimer assumptions
D«d and h«L 2-38
-------
4.6 Relationship between equivalent depth and total
depth for different drain separations 2-40
4.7 Plan view of flow to a drain of finite length... 2-41
4.8 Superposition of drawdowns for two pumping wells
in a confined aquifer 2-50
4.9 Superposition of drawdowns to obtain drawdown
after a step change in discharge 2-52
4.10 Flow pattern around a pumping well in a uniform
regional flow 2-53
4.11 Flow patterns around a recharge/recovery doublet
in a uniform regional flow 2-55
4.12 Dimensionless plot of doublet width.... 2-56
4.13 Percent recharge being discharged in a doublet.. 2-57
4.14 Inner and outer recirculation cells created
by pairs of pumping and injection wells 2-59
4.15 Real and image well configurations for wedge-
shaped aquifers 2-61
4.16 Real and image well configurations for strip
and rectangular aquifers 2-62
4.17 Configuration of an impermeable barrier that
partially penetrates into a single-layered
aquifer 2-67
4.18 Configuration of an impermeable barrier that
partially penetrates into a two-layered aquifer. 2-70
4.19 Relationship between and the flow under a
partially penetrating barrier in a layered
aquifer 2-71
4.20 Relationship between depth of penetration and
flow under a partially penetrating barrier in a
layered aquifer 2-72
6.1 Vertical cross-section through landfill 2-98
6.2 Depth of drain as a function of downgradient
distance from the landfill 2-101
XI
-------
6.3 Steady-state water table profile for a partially
penetrating drain located 200 ft downgradient
from the landfill 2-103
6.4 Change in water table elevation below the
landfill following installation of the drain.... 2-104
6.5 Aquifer characteristics and current extent of
methylene chloride plume 2-105
6.6 Dimensions of ground-water divide for a pumping/
injection Rate of 27 gpm 2-107
6.7 Plume location and aquifer characteristics for
example 3 2-109
6.8 Plume position 0, 10, 20, 40, 80 and 120 days
after initiation of pumping 2-110
6.9 Impermeable barrier configuration 2-112
6.10 Image well configuration for impermeable barrier 2-113
6.11 Plume position for 0, 10, 20, 40, 80, 120, 160,
320 and 640 days after initiation of pumping.... 2-114
6.12 Percent recovery as a function of time for
alternative well and barrier configurations 2-115
6.13 Aquifer characteristics and remedial action
configuration for well point recirculation
system 2-118
6.14 Comparison of water table elevations for
mound and injection well 2-120
6.15 Particle movement from the perimeter of the
cooling water pond to each well point 2-121
6.16 Equipotential contours and flow lines produced
by the drain recirculation system 2-123
VOLUME 3: NUMERICAL MODELING OF SURFACE, SUBSURFACE
AND WASTE CONTROL ACTIONS
3.1 Local environment zones surrounding an
uncontrolled hazardous waste site 3-7
3.2 Schematic overview of a waste site and
xii
-------
selected intra- and inter-zone processes
affecting water and waste constituent migration. 3-9
4.1 Hypothetical hazardous waste site (plan view)... 3-17
4.2 Hypothetical hazardous waste site (cross-
section) 3-18
4.3 Surface actions: grading, revegetation and
surface water diversion and collection (plan
view) 3-19
4.4 Grading, revegetation and surface water
diversion and collection (cross-section) 3-20
4.5 Subsurface drains and bottom liner 3-23
4.6 Extraction wells or interceptor trenches used to
lower water table (cross-section) 3-24
4.7 Extraction wells or interceptor trench combined
with a seepage basin to capture a plume
(cross-section) 3-25
4.8 Circumferential impermeable barrier (plan view). 3-27
4.9 Circumferential impermeable barrier
(cross-section) 3-28
4.10 Permeable treatment bed or chemical/microbe
injection 3-30
4.11 Bioreclamation and chemical injection using an
injection/withdrawal doublet (plan view) 3-31
4.12 Solution mining using injection/withdrawal
wells (cross-section) 3-33
5.1 Typical soft linkage of surface, unsaturated and
saturated zone codes 3-46
5.2 Typical dimensionalities used to represent
surface, unsaturated and saturated zones 3-49
5. 3 Model application process 3-51
5.4 Predicted performance of different remedial
action alternatives in reducing arsenic concen-
trations in the Cedar River under low flow
conditions 3-57
Xlll
-------
5.5 Predicted effects of two values of hydraulic
conductivity on the shape of the water table
with installed French Drain 3-60
5.6 Discharge to drains at Lipari Landfill for
different drain depths 3-62
5.7 Facility leachate loading and loading to ground
water 3-64
5.8 Relative leachate concentration at monitoring
well and stream 3-64
6.1 Two typical cap designs showing layers in each
zone 3-72
6.2 Example x-y representation and grid used to
evaluate the French Drains at Love Canal 3-77
6.3 Two-dimensional (x-z) grid configuration used by
Cohen and Miller to evaluate a proposed cut-off
wall at Love Canal 3-79
6.4 Two-dimensional (x-y) grid configuration used by
Anderson et al. to evaluate a proposed slurry
wall at the Lipari Landfill 3-80
6.5 Example representation and grid for a drain
system used to evaluate Uranium mill tailings
seepage into the unsaturated zone 3-84
6.6 Pan factors 3-90
6.7 The effect of soil type on soil-water retention. 3-96
6.8 Dependance of conductivity on suction in soils
of different textures 3-101
6.9 Variation of dispersivity with distance 3-103
6.10 Mineral bulk density 3-108
VOLUME 4: ANALYTICAL AND NUMERICAL MODELS FOR THE
EVALUATION OF REMEDIAL ACTIONS IN SURFACE
WATER
2.1 Flow diagram of important physical processes.. 4-7
2.2 Diagram of chemical and biological processes.. 4-11
xiv
-------
3.1 Typical boom or silt curtain deployment
configurations 4-25
3.2 Isolation for sediment excavation using single
cofferdam 4-26
3.3 Streamflow diversion for sediment excavation
using two cofferdams and diversion channel.... 4-27
4.1 Illustration of hypothetical spill incident... 4-33
6.1 Model development and application process 4-60
7.1 Reductions in total Kepone concentrations
from different dredging scenarios 4-73
7.2 Particle size vs. settling velocity for
suspended sediment 4-80
7.3 ^ and Tc for DuBoys relationship as functions
of median size of bed sediment, where T =
critical shear stress and i/^ = coefficient
depending on grain size 4-81
xv
-------
TABLES
Page
VOLUME 1: SELECTION OF MODELS FOR REMEDIAL ASSESSMENT
4.1
4.2
4.3
4.4
4.5
4.6
5.1
5.2
5.3
5.4
5.5
Soil and Ground-Water Remedial Measures
Remedial Actions vs. Simplified and Analytical
Methods Matrix
Remedial Actions vs. Required Processes Matrix..
Remedial Actions vs. Required Model Dimension-
Data Needs for Level I (Analytical) Methods for
Data Needs for Level II (Numerical) Methods for
Surface Water Remedial Actions
Simplified and Analytical Surface Water Methods
vs . Remedial Actions Matrix
Remedial Actions vs. Processes Matrix
Remedial Actions vs. Water Body Matrix
Data Needs for Level I (Analytical) Methods for
1-21
1-23
1-26
1-30
1-32
1-36
1-43
1-46
1-48
1-50
1-50
5.6 Data Needs for Level II (Numerical) Methods for
Surface Water Problems 1-51
VOLUME 2: SIMPLIFIED METHODS FOR SUBSURFACE AND WASTE
CONTROL ACTION
3.1 Grouping of Remedial Measures 2-7
xvi
-------
3.2 Applicability of Simplified Methods to the
Evaluation of Subsurface Control Actions 2-8
3.3 Applicability of Simplified Methods to the
Evaluation of Waste Control Actions 2-10
4.1 Inventory of Selected Well Hydraulics Solutions.. 2-33
4.2 Inventory of Selected Steady-State Drain
Hydraulics Solutions 2-43
4.3 Inventory of Selected Transient Drain
Hydraulics Solutions 2-45
4.4 Complete Elliptic Integrals of the First Kind.... 2-68
5.1 Available Hand-Held Calculator Programs
for Well Hydraulics 2-79
5.2 Inventory of Selected Hand-Held Calculator
Programs for Drain Hydraulics 2-83
5.3 Inventory of Selected Hand-Held Calculator
Programs for Ground-Water Mounding Estimation.... 2-86
5.4 Inventory of Selected Hand-Held Calculator
Programs for Contaminant Transport 2-88
5.5 Inventory of Selected Micro-Computer Programs for
Well Hydraulics 2-91
5.6 Inventory of Selected Micro-Computer Programs
for Ground-Water Mounding Estimation 2-93
5.7 Inventory of Selected Micro-Computer Programs
for Contaminant Transport 2-94
6.1 Saturated Hydraulic Conductivities and
Thicknesses for Example 1 2-100
VOLUME 3: NUMERICAL MODELING OF SURFACE, SUBSURFACE
AND WASTE CONTROL ACTIONS
3.1 Processes Controlling the Migration and Fate of
Hazardous Waste Constituents 3-8
4.1 Processes Affected by Different Remedial
Measures 3-15
xvi i
-------
5.1 General Capabilities of Selected Saturated,
Surface and Unsaturated Zone Models 3-38
5.2 Detailed Capabilities of Selected Surface,
Unsaturated and Saturated Zone Models 3-40
5.3 Relative Ranking of Potential Alternative
La Bounty Landfill Remedial Actions at Different
Points in Time Using Level of Contamination
Reduction in the Cedar River as a Measure of
Performance 3-58
6.1 Remedial Measures 3-68
6.2 Summary of Modeling Requirements for Each
Remedial Measure 3-70
6.3 Channel and Land Surface Manning's 'n1 Values
Applicable to Remedial Action Modeling 3-88
6.4 Interception Storage for Different Vegetative
Densities 3-91
6.5 Values of 'a' for Equation (6.2) 3-93
6.6 Coefficients for Linear Regression Equations for
Prediction of Soil Water Contents at Specific
Matric Potentials 3-97
6.7 Ranges of Hydraulic Conductivities for Different
Materials 3-99
6.8 Small Scale and Regional Dispersivity Values.... 3-104
6.9 Ranges of Porosity and Effective Porosity Values
for Selected Materials 3-106
6.10 Range of Bulk Density for Different Materials... 3-107
6.11 Regression Equations for the Estimation of K .. 3-110
6.12 Bioreclamation Degradation Rates for Selected
Waste Constituents 3-112
VOLUME 4: ANALYTICAL AND NUMERICAL MODELS FOR THE
EVALUATION OF REMEDIAL ACTIONS IN SURFACE
WATER
2.1 Important Processes: Physical and Chemical 4-5
xviii
-------
2.2 Pollutant vs. Processes Matrix 4-12
3 .1 Outline of Remedial Actions 4-21
3.2 Remedial Actions vs. Processes Matrix 4-22
4.1 Types of Discharge Scenarios 4-32
5.1 Simplified Assessment Techniques vs. Use and
Required Data Matrix 4-40
5. 2 Simplified Assessment Techniques 4-42
5.3 Analytical Models vs. Model Capabilities and
Required Data Matrix 4-48
6.1 Processes vs. Models Matrix 4-53
6.2 Remedial Actions vs. Water Body Matrix.. 4-56
6.3 Remedial Actions vs. Models Matrix 4-57
7.1 Grouping of Remedial Actions According to
Model Requirements 4-63
7.2 Modeling Requirements for Remedial Actions 4-66
7. 3 Sedimentation Grade Scale 4-78
7.4 Specific Weights of Sediments Showing Extreme
Variation 4-79
7.5 Determination of a Continuous Suspended
Sediment Source Term by Schnoor et al 4-83
7.6 Mass of Contaminated Sediments and Equivalent
Water Depth as a Function of Depth
contamination 4-84
7.7 Reported Values for the Longitudinal Mixing
Coefficients in Different Channels 4-87
7.8 Experimental Measurements of Transverse
Mixing in Open Channels with Curves and
Irregular Sides 4-89
xxx
-------
ACKNOWLEDGMENTS
This document represents the combination of efforts under
three work assignments of contract No. 68-03-3116 to provide
technical support for pesticides and other toxic substances
for the U.S. Environmental Protection Agency, Office of
Research and Development (ORD), by Anderson-Nichols & Co.,
Inc. (ANCo) in Palo Alto, CA. Mr. Douglas Ammon of the
Hazardous Waste Engineering Laboratory (HWERL) coordinated the
development of this document. Mr. Lee A. Mulkey of the
Environmental Research Laboratory (ERL) in Athens, GA, was the
project officer for this contract.
The technical project monitors, authors, and reviewers are
listed below.
Volume 1; Selection of Models for Remedial Action Assessment
(Work Assignment No. 20)
Technical project monitor: Mr. Thomas O. Barnwell, Jr.,
ERL
Authors: Mr. Scott H. Boutwell, Mr. Stuart M. Brown,
and Dr. Benjamin R. Roberts
Volume 2; Simplified Methods for Subsurface and Waste Control
Actions (Work Assignment No. 5)
Technical project monitor: Mr. Douglas Ammon, HWERL
Author: Mr. Stuart M. Brown
Volume 3; Numerical Modeling of Surface, Subsurface and Waste
Control Actions (Work Assignment No. 5)
Technical project monitor: Mr. Douglas Ammon
Authors: Mr. Stuart M. Brown, Mr. Scott H. Boutwell,
Dr. Benjamin R. Roberts, and Ms. Dorothy Fisher
Atwood
XX
-------
Volume 4; Analytical and Numerical models for the Evaluation
of Remedial Actions in Surface Water (Work
Assignment No. 10)
Technical project monitor: Mr. Thomas O. Barnwell, Jr.
Authors: Mr. Scott H. Boutwell and Dr. Benjamin R.
Roberts
Technical review of this document was provided by Mr. Anthony
S. Donigian, Jr. who also served as the project director for
the contract. Other reviewers are listed below.
Charles R. Cole, Battelle Pacific Northwest Laboratories
(Volume 3, interim report)
William Fallen, Office of Research and Development
(Volumes 1 and 4)
Wayne C. Huber, Ph.D., University of Florida (Volume 4)
Yasuo Onishi, Ph.D., Battelle Northwest Laboratories
(Volume 4, interim report)
Richard Stanford, Clean Sites Inc. (peer review)
Paul K. M. Van der Heijde, International Ground Water
Modeling Center (peer review)
David T. Williams, U.S. Army Waterways Experiment Station
(Volume 4)
David B. Watson, Anderson-Nichols & Co., Inc. (Volume 3)
Dr. Richard T. Y. Lo participated in the development of model
application guidelines and remedial action modeling
requirements in Volume 3.
Ms. Susan Reutter-Harrah supervised report production. Word
processing was provided by Ms. Carol McCullough and Ms.
Dorothy Inahara. Graphics were developed by Ms. Marythomas
Hutchins and Ms. Virginia Rombach.
xxi
-------
INTRODUCTION
The National Contingency Plan (NCP) sets forth a process for
the evaluation and selection of remedial actions at
uncontrolled hazardous waste sites. One element in this
process is the Engineering Feasibility Study. This study is
itself a staged process that involves the screening of
remedial action technologies, the detailed analysis of
potentially feasible alternatives and the conceptual design of
the most cost-effective alternative.
Throughout the feasibility study process a number of factors
are considered when evaluating remedial actions. These
factors include technical feasibility, costs, institutional
constraints, and potential environmental and public health
impacts. The level of attention given to each of these
factors depends upon which stage in the process is being
performed. During the screening stage, the intent is to
reduce the large number of potential technologies to a
workable number by identifying those that are clearly
infeasible or inappropriate. Best engineering judgement
supplemented by order-of-magnitude estimates of remedial
action performance are usually sufficient during this stage.
Once a set of potentially feasible alternatives has
beenidentified, each one has to be evaluated in detail.
Again, best engineering judgement supplemented by more
quantitative estimates of performance largely provide the
basis for the identification of one or more cost-effective
actions.
The final step is to develop a conceptual design for one or
more alternatives. This involves identifying the performance
expectations for the alternative, design criteria, preliminary
layout and process diagrams, operation and maintenance
requirements, monitoring requirements, and costs. This step
requires an even more quantitative analysis of performance.
Modeling, whether it be through the use of relatively simple
analytical solutions or more sophisticated numerical codes, is
beginning to be used more and more throughout the feasibility
study process. This four volume series is intended to provide
guidance on both the selection and use of a range of modeling
techniques applicable to the evaluation of remedial actions
xxii
-------
for ground-water and surface water contamination problems.
Volume 1, Selection of Models for Remedial Action Assessment,
provides a methodology for the selection of models. The
methodology addresses three key decisions: 1) whether modeling
should be considered; 2) if so, what level of model
sophistication is appropriate; and 3) what capabilities should
the model or models have. The first decision is critical
because modeling is appropriate for only certain sites. The
second decision is important because the level of model
sophistication will determine the level of resources that must
be allocated. The final decision ensures that the selected
model will be appropriate for the site conditions and remedial
actions that need to be assessed. Once a selection has been
made, the remaining volumes can be consulted to obtain
guidance on model use.
Volume 2, Simplified Methods for Subsurface and Waste Control
Actions, provides a compilation of simplified methods, or
analytical and semi-analytical solutions, applicable to the
evaluation of subsurface and waste control actions. The
primary emphasis of this volume is on identifying the methods
that can be used to evaluate specific actions and the
assumptions and limitations affecting their use. A
compilation of available hand-held calculator and
micro-computer programs for different types of methods is also
provided. The simplified methods contained in Volume 2 are
useful for screening remedial actions and, in some cases,
detail analysis and conceptual design.
Volume 3, Numerical Modeling of Surface, Subsurface and Waste
Control Actions, provides guidance on the use of numerical
models for sites where more detailed analyses are required and
where sufficient resources are available. The volume focusses
on the use of surface, unsaturated and saturated zone models.
Important considerations related to the use of numerical
models are discussed, as are modeling requirements for
specific surface, subsurface and waste control actions.
Modeling requirements include: 1) the type of model required
to analyze an action; 2) the dimensionality and grid
configuration required to represent an action; and 3) model
parameter adjustments required to simulate the effects of
implementing an action. Guidance on the estimation of model
parameters is presented for situations where site
characterization data are unavailable.
Volume 4, Analytical and Numerical Models for the Evaluation
of Surface Water Remedial Actions, provides remedial action
modeling guidance for surface water contamination problems.
Simplified methods and analytical and numerical models
applicable to the analysis of specific actions are discussed.
xxiii
-------
Considerations related to the application of both types of
models are presented along with modeling requirements for
different actions.
xxiv
-------
VOLUME 1
Selection of Models
for
Remedial Action Assessment
-------
-------
VOLUME 1: SELECTION OF MODELS FOR REMEDIAL
ACTION ASSESSMENT
SECTION 1
EXECUTIVE SUMMARY
This volume provides general guidance in the selection of
models for remedial assessment at uncontrolled hazardous
waste sites. Guidance is provided in the form of a series
of flow charts and matrices leading to model selection. The
purpose of this format is to make the model selection
methodology as utilitarian and "user-friendly" as possible.
As the methodology is used at different sites, user
confidence and expertise will increase. With successful
application experience, the model selection guide may become
an integral tool in the Remedial Investigation/Feasibility
Study process.
The selection of models is a function of the objectives of
the modeling study, complexity of the site, and type of
remedial actions being considered. These areas are
represented in the methodology by: the required level of
modeling flow chart; matrices of remedial actions vs.
processes and required dimensionality; and discussions of
time frame criteria. Other criteria that are important
include model performance and data/resource availability.
The same model selection methodology can be applied to the
selection of models for site characterization and exposure
assessment. The user again must identify the purpose of the
modeling study in order to determine the level of modeling
required, and evaluate required model capabilities based on
the complexity of the site, including significant environ-
mental pathways of exposure, and potential receptors. In
short, many of the model selection criteria for remedial
action assessment are applicable to exposure assessment and
site characterization as well.
1-1
-------
This volume is designed to provide guidance for model
selection and use in the Remedial Investigation/Feasibility
Study process. The document as a whole should provide a
comprehensive set of guidelines to Federal and State
officals for the incorporation of models into the remedial
action planning process at state and federal superfund
sites.
1-2
-------
SECTION 2
INTRODUCTION
2.1 PURPOSE OF THIS REPORT
Existing state and federal funds for the clean-up of
uncontrolled hazardous waste sites are limited.
Consequently, proper selection of remedial actions is
critical to ensure that effective measures are implemented
at as many sites as possible, and that future costs as a
result of inadequate actions are minimal.
Mathematical models can be used to assess the performance of
remedial actions, and thus complement the analyst's
expertise and judgement for selection and design of these
actions. Although the concept for using models for remedial
action assessment is relatively new, successful
demonstrations are evident. The recent application of
models for remedial action evaluation at the La Bounty
landfill site in Charles City, Iowa, the Gilson Rd. site in
Nashua, New Hampshire, and the Love Canal site in Niagara,
New York has proven that they can provide information useful
in selecting and designing actions.
Effective use of models for this purpose depends on the
selection of models most suitable to the job. There are
many models available today, varying in terms of complexity
and purpose of use. Actual selection can be difficult,
especially if one is not completely familiar with the
important site and remedial action criteria needed to choose
the appropriate model. This volume was developed for use as
a model selection guide for assessing remedial action
performance at uncontrolled hazardous waste sites. It is
intended to assist state or regional staff in assessing the
need for analytical predictive tools at these sites, for use
by themselves or to evaluate site contractor proposals.
1-3
-------
It should be emphasized that the model selection methodology
was designed for remedial action assessment for surface
water and subsurface contamination problems. However, many
of the same models and model selection criteria are
appropriate for both site characterization and exposure
assessments. Reports by Adkins et al. (1983) and Freed et
al. (1983) also provide methodologies for assessing exposure
to chemical substances.
The model selection methodology is specifically for surface
and ground-water contamination. Although some of overland
and unsaturated zone models consider air contamination
problems such as volatile emissions and fugitive dust
release, model selection for air contamination control is
not covered in this report. The user should refer to other
reports such as Farino et al. (1983), Thibodeaux (1981) and
Dynamac (1982) for models that simulate air emissions,
fugitive dust, and their appropriate control technologies at
hazardous waste sites.
2.2 DEFINITION OF MODELS
Before the model selection methodology is presented,
definition of the terms "model" and "level" must be
clarified. Throughout this volume we will refer to two
general classes of models, based on level of complexity.
They are analytical and numerical, designated as Levels I
and II, respectively.
Analytical and numerical models incorporate equations to
quantitatively predict results, with varying levels of
accuracy. The major difference between the two types of
models is the level of simplification. Analytical models
(Level I) rely on simplifying assumptions such as isotropic
(Hydraulic conductivity is equal for all directions: x, y,
z) and homogeneous conditions, steady flow, and regular
geometry. Their range of accuracy is around 1 order of
magnitude (EPA, 1982). Numerical models (Level II) utilize
the same equations, but can simulate varying processes,
fluxes, and geometries, by nature of their solution
techniques. Their range of accuracy is closer to a factor
of 2-4 (EPA, 1982). However, for both types of models,
accuracy will be also dependent upon application. Because
of the number of calculations required, all numerical models
and some analytical models require a computer to solve
equations that calculate exposure concentrations.
1-4
-------
2.3 THE ROLE OF REMEDIAL ACTION ASSESSMENT IN THE REMEDIAL
RESPONSE PROCESS
2.3.1 The Role of Models in Remedial Action Assessment
In the past, selection and design of remedial actions has
largely been accomplished through field data collection and
best engineering judgement. These approaches may be
sufficient for sites where environmental pathways and
potential receptors are clearly defined, and where previous
installations of a given remedial action have proved
successful. This past experience of both identifying
pathways and applying specific remedial measures may also be
sufficient for sound selection at new sites with similar
characteristics. However, there are a number of relatively
complex sites where best engineering judgement may not
provide enough guidance to allow for the proper selection
and design of a remedial action. For those sites, the use
of analytical and numerical predictive tools may be
appropriate to obtain a quantitative assessment of remedial
action performance. Best engineering judgement, then, may
be supplemented with quantitative results. This will
potentially lead to more accurate and confident decisions.
Models have potential use in the screening of alternatives,
analysis of alternatives, and conceptual design tasks.
Descriptions of model use for each task are provided below.
Actual selection of the level of model required is dependent
upon site and remedial action criteria, and to a lesser
extent, the current phase of the remedial investigation/
feasibility study. Other criteria of importance can include
resource and data availability and previous model
performance. This information will be covered in detail in
Sections 4 and 5.
Screening of alternatives is performed to eliminate those
actions deemed unfeasible due to technical, public
health/institutional, and cost reasons. Models may be used
at this stage to determine the general technical
feasibility, and any potential environmental impacts arising
from implementation of different remedial actions. Since
the screening analysis is essentially the first iteration of
subsequent, more detailed analyses, a ballpark or
order-of-magnitude estimate of effectiveness is usually all
that is required. For this reason, Level I models are often
sufficient.
1-5
-------
Those remedial actions that pass the initial screening
effort will then be subject to analysis of alternatives.
Models may be used to obtain informationonthe
effectiveness, durability, and expected exposure levels as a
result of the implementation of different actions. The
effectiveness of an action is the extent to which it meets a
design objective or site clean-up goal. The durability of
an action is the length of time it is effective. Durability
can be assessed by incorporating design life and risk of
failure considerations into modeling of the selected
actions. Such considerations may allow the prediction of
exposure levels in the event of progressive or catastrophic
failure of an action. Exposure levels of contaminants
expected with implementation of a remedial action can be
determined so compliance with regulatory criteria can be
ascertained. Models may be used in the detailed analysis
phase for the purposes mentioned above. Level I
(Analytical) models are limited to well characterized sites
and selected remedial actions. Level II (Numerical) models
are more appropriate in cases where site conditions or
remedial actions of interest require that changes in
material properties and multiple dimensions be considered.
The stage subsequent to the feasibility study is to develop
a conceptual design for the most cost-effective action.
Again, models can assist in this process by simulating
different configurations of the selected action. For
example, a ground-water pumping action may be conceptually
designed by evaluating pumping rates, number and placement
of wells, and location of screened intervals. Again, while
analytical or Level I models may be sufficient for some
sites and actions, the use of numerical models may be more
appropriate for complex configurations.
Because of the unique characteristics of each site/remedial
action scenario, the determination, selection and use of
modeling have to be addressed on a site-specific basis.
However, with sufficient user expertise, field data, and
guidance, models may become common tools for the Remedial
Investigation/Feasibility Study process.
1-6
-------
2.4 FRAMEWORK AND ORGANIZATION OF REPORT
2.4.1 Basic Framework for Model Selection
There are three basic decision points in the methodology
discussed in this volume. Figure 2.1 is a flow chart that
illustrates the framework. The decisions are:
1. Is modeling necessary
2. If so, what level of modeling is required and
3. What are the required model capabilities of that
level
Flow charts and matrices are used to facilitate the model
selection process. The sections of the volume that
correlate to these decision points are identified below.
2.4.2 Organization of Report
Section 3 discusses the first major decision: "Is modeling
necessary?" A flow chart is used to illustrate the
hierarchy of questions that must be asked in making this
basic decision.
Sections 4 and 5 comprise the major portions of the
methodology; Section 4 deals with model selection for
remedial action assessment for soil or ground-water
contamination problems, and Section 5 covers the same issues
for surface water problems. Each section contains the two
decision points:
o What level of modeling is required and
o What are the required model capabilities
A summary of general model selection criteria is provided in
each section to help the user identify his/her specific
model requirements and to direct the user to the appropriate
model use volume(s) for more information. Data and resource
considerations when applying the methodology are also
discussed. Section 6 contains references.
1-7
-------
IS MODELING NECESSARY?
YES
NO
MODEL SELECTION
IS NOT REQUIRED
WHAT LEVEL OF MODELING
IS REQURED?
LEVEL I: SIMPLE/ANALYTICAL
1
LEVEL II: COMPLEX/NUMERICAL
WHAT ARE THE REQUIRED
LEVEL I MODEL CAPABILITIES?
WHAT ARE THE REQUIRED
LEVEL II MODEL CAPABILITIES?
Figure 2.1 Three basic decisions in model selection,
1-8
-------
2.5 CAVEATS OF USE
There are a number of assumptions made in this volume
regarding user expertise and knowledge of the site, and
limitations of the model selection methodology. The user is
advised to assess his/her expertise and anticipated support
in light of the caveats of use.
2.5.1 Assumed User Knowledge and Expertise
The primary group of users is expected to be Federal (EPA)
regional and state environmental officials and staff. These
people often must evaluate field inspection reports, direct
efforts for data acquisition, and evaluate site contractor
proposals for remedial action, including recommended models
for exposure and remedial action assessment. To make sound
decisions using this methodology, the user should have at
the minimum a general understanding of a mix of
disciplines,such as Hydrology, Civil Engineering, Soil
Science, and Environmental Chemistry and an understanding of
the basic concepts of chemical transport and fate modeling,
such as levels of model complexity and expected accuracy,
processes that can be simulated, and parameter estimation
techniques. Ideally, academic background in any of the
above disciplines supplemented with experience, job
training, and/or exposure (e.g., short course attendance) in
the other disciplines provides a profile of the recommended
background of a user.
2.5.2 Understanding of the Site and Remedial Actions
In addition to required expertise, the user should be able
to characterize site conditions. This means that major
environmental pathways and potential receptors must be
identified. Such pathways and receptors can include
contaminated runoff into surface waters, leachate migration
into ground water, direct contamination of a drinking water
aquifer, and release of toxic volatile emissions. This
qualitative assessment of the site is a function of the
users expertise and the available data, such as observed
concentrations in surface water and ground water, and
knowledge of the history of the site, including type of
contaminants, methods of disposal, and release rates, if
such information is available.
1-9
-------
The user should be able to categorize groups of remedial
actions and correlate them to the major pathways and
potential receptors. As models will be evaluated according
to their ability to simulate various remedial actions, it is
important that the user be familiar with the purpose and
basic engineering design of the available remedial action
technologies.
2.5.3 Limitations of the Model Selection Methodology
This methodology was designed to provide guidance on the
selection of models for remedial action assessment in
surface and ground waters. Given the myriad possibilities
of site conditions, the methodology is directed towards
model selection at a generic level. That is, there may be
decision points in the methodology where more than one
answer exists. It is at these points that sufficient user
expertise and knowledge of the site is most critical.
Therefore, the information derived from the methodology
should supplement existing knowledge and expertise for
application to site-specific conditions.
The user should also refer to the subsequent volumes for
additional information about specific models, such as user
manuals and test case applications, before actual model
selection is determined.
It should be emphasized that this manual alone will not be
sufficient for actual model selection; the references cited
above should be examined, and, if necessary, outside
guidance should be obtained in order to facilitate the
selection process.
1-10
-------
SECTION 3
THE FIRST DECISION: IS MODELING NECESSARY?
3 .1 OVERVIEW
The decision to use models for remedial action assessment is
the first and perhaps most critical in the model selection
methodology. This section will help the user to answer the
question "Will the resources expended and results obtained
be worth the modeling effort?". If so, models can
supplement best engineering judgement for remedial action
assessment. If not, the user should explore other methods of
remedial action assessment, such as collection and analysis
of more field data.
The decision to use models is a function of the nature and
complexity of the site being considered, as well as the
extent of contamination and range of potential remedial
actions. As mentioned in the introduction, site
characteristics and remedial action criteria form the basis
of model selection throughout this report. Examination of
these issues allow the user to make the decision to model,
and if models are deemed necessary, they set the stage for
selection of required model capabilities for subsurface and
surface water problems in Sections 4 and 5, respectively.
3.2 THE DECISION TO MODEL: FLOW CHART
Figure 3.1 is a flow chart that can be used to determine if
modeling is required. The major issues or decision points
are represented using a flow chart in order to facilitate
ease of use. Each decision point is discussed below.
The first step is to develop a conceptual understanding of
the site. To do this, the user should make assumptions as
to the type and degree of hazards at the site, based on
1-11
-------
DEVELOP CONCEPTUAL
UNDERSTANDING OF SITE
I
CAN ASSUMPTIONS BE
CONFIRMED WITH
EXISTING DATA?
WILL ADITIONAL
DATA IMPROVE
UNDERSTANDING?
NO
NO
YES
DO YOU NEED
QUANTATIVE ESTIMATES
OF FUTURE CONDITIONS?
NO
YES
MODELING
IS NOT REQUIRED
MODELING
IS REQUIRED
DETERMINE LEVEL OF
MODELING REQUIRED
(FIGURES 4.1 OR 5.1)
Figure 3.1
Flow chart to determine if modeling is
required.
1-12
-------
existing data and best engineering judgement. Such
assumptions may include: the location of potential sources;
the relative importance of different contaminant migration
and fate pathways (e.g., air, surface water and ground
water); the relative importance of different transport,
transformation, and inter-media transfer processes (e.g.,
contaminated runoff from land into surface water, and
volatilization of a pollutant from water to air); and the
general type(s) of remedial measures that may be applicable,
such as waste isolation, removal, or treatment, leachate or
runoff control, or water diversion.
Once an understanding is developed, the user should ask the
following question: "Can assumptions be confirmed with
existing data?" Such assumptions can include the extent or
plume of contaminants and rate of spreading. For example,
can the contamination and fate pathways be identified by
examining contamination levels in different media?
If the assumption concerning important pathways and
receptors cannot be confirmed with existing data, the
following question should be asked: "Will additional data
improve understanding?" Quite often additional data
gathered from sampling programs in the Remedial
Investigation/Feasibility Study will be sufficient to
confirm one's understanding of the site and help identify
appropriate remedial measures. However, in some instances
existing data may be confusing or contradict the user's
assumptions and understanding of the site. For example,
pump tests may reveal that a wide range of hydraulic
conductivities are present. Or, estimates and location of
the source mass cannot be verified. In these cases, models
may serve to interpret and interpolate site conditions to
provide a better understanding.
If site pathways are well characterized (i.e., contaminated
runoff into a simple water body) and conditions are fairly
homogeneous (i.e., one general soil type or single layer
aquifer), additional data will probably suffice to confirm
assumptions. The user should then specify sampling
requirements, obtain more data, and re-iterate the decision
process by developing a new understanding of the site based
on the new data, and proceed from there.
If additional data will not serve to improve the
understanding of the site, modeling is required. The
decision to model in these cases is usually dependent
primarily on site complexity, as opposed to remedial action
criteria.
1-13
-------
If the question "Can assumptions be confirmed with existing
data?" was answered in the affirmative, the site itself is
well enough characterized not to warrant additional data or
modeling. The next step is to examine remedial action
criteria.
The question "Do you need quantitative estimates of future
conditions?" addresses the need to predict potential
contamination levels and the effectiveness of remedial actions
in reducing those contamination levels. Qualitative estimates
of future conditions can be made based on the conceptual
understanding of the site and judgement. When contamination
pathways are well characterized and past experience indicates
that appropriate remedial actions will work, a qualitative
assessment of future conditions may be sufficient. However,
if multiple or complex pathways are present, selection of the
appropriate remedial actions and their configurations are not
apparent, and specific regulatory criteria for contamination
levels must be met, a quantitative assessment of future
conditions may be required. For these cases an affirmative
response to the question posed above means that modeling is
required. A negative response indicates that modeling is not
required. In this latter case, the remainder of the model
selection methodology presented herein is not applicable, and
data and experience are sufficient for remedial action
assessment.
As evidenced by Figure 3.1, the decision to model may be a
result of site complexity (i.e., additional data will not
improve understanding) and the need for a quantitative
assessment of remedial action performance over time.
As the user proceeds through the Remedial Investigation/
Feasibility Study, he/she may arrive at different answers for
the need for modeling, depending on the current task at that
time. In light of this, the user should consider any future
modeling decisions (e.g., in the detailed analysis or
conceptual design stages) to be made. This can effect future
resource allocation and data collection. As mentioned before,
the methodology to decide to model does involve iteration, and
the user should expect to reassess the need to model
throughout the Remedial Investigation/Feasibility Study.
If modeling is required, the next step in the model selection
methodology is to determine the level of modeling required.
Sections 4 and 5 include flow charts for this purpose, for
both ground water and surface water problems, respectively.
These sections will assist the user in the third step of the
decision framework by identifying the required model
capabilities.
1-14
-------
SECTION 4
METHODOLOGY FOR MODEL SELECTION FOR SOIL AND
GROUND-WATER CONTAMINATION PROBLEMS
4.1 OVERVIEW
At this point, the user has ascertained the need for
modeling in remedial action assessment, and has identified
the specific media (e.g., air, subsurface, and surface
water) that are affected and are subject to contamination
control. This section helps to answer the second two
questions for subsurface contamination problems in the
modeling decision framework: 1) What level of modeling is
required?, and 2) What are the required model capabilities
of that level? In conjunction with these decisions data and
resource availability issues for each level of model are
also discussed. Section 5 covers these same issues for
surface water contamination problems. The formats of both
sections are similar. Flow charts and matrices are used to
guide the user towards model selection. The matrices will
introduce th*» user to the interacting relationships of the
remedial actions, environmental processes, and flow fields
(for required model dimensionality). Information describing
the remedial actions and environmental processes of concern
in the soil and ground-water systems is provided in Volume
3: Numerical Modeling of Surface, Subsurface, and Waste
Control Actions. Other sources of information on remedial
actions include: JRB (1982), Ehrenfield and Bass (1983) and
SCS (1981). A list of soil and ground-water remedial
actions considered for simulation is provided in Table 4.1.
This methodology can also be used to select models for site
characterization and exposure assessment. Selection
criteria will be based on site complexity and modeling
objectives, and may be less stringent than criteria for
remedial action assessment. In many of these cases,
analytical models may suffice.
1-15
-------
4.2 WHAT LEVEL OF MODELING IS REQUIRED?
There are seven questions to be answered when determining
the required level of model. Figure 4.1 is a flow chart
that illustrates the hierarchy of decisions to be made.
Each question or decision must be answered in the
affirmative for analytical (Level I) modeling to be chosen.
A "no" answer at any decision point pushes the user towards
the use of a numerical (Level II) model, whereupon data and
resource availability should be examined. This hierarchy
was developed to define the strict and limited conditions of
analytical model use in remedial action assessment. These
decisions are described below.
The first decision or question is:
"Are order-of-magnitude predictions acceptable?"
This is primarily a function of the current task of the
remedial investigation/feasibility study. In the screening
of alternatives, remedial actions are ranked for potential
use according to their general technical feasibility.
Therefore, order-of-magnitude assessments are usually
acceptable. In the analysis of alternatives and conceptual
design tasks, the selected remedial actions are subjected to
a more rigorous analysis. Quantitative assessments of
remedial action performance are needed at these stages so
that the most effective action or combination of actions in
terms of reducing concentration levels is chosen.
Therefore, it is possible that order-of-magnitude estimates
may not be acceptable, and Level II modeling is required.
The next three questions to be asked concern the degree of
variability or heterogeneity in site conditions. The first
is:
"Is it reasonable to assume that media properties are
uniform, and do not vary spatially?"
In actuality, site conditions or media properties are never
truly homogeneous; different levels of heterogeneity exist,
depending on site complexity and the size of area being
considered. However, in terms of modeling requirements,
assumptions can often be made to simplify site conditions.
If a high degree of accuracy is not critical, and
properties are relatively uniform (i.e., one soil type or
similar soil characteristics, single layer aquifer), Level I
1-16
-------
REASSESS
GOALS AND
DATA NEEDS
ARE ORDER OF MAGNITUDE
PREDICTIONS ACCEPTABLE?
NO
YES
IS IT REASONABLE TO ASSUME THAT MEDIA PROPERTIES
ARE UNIFORM, AND DO NOT VARY SPATIALLY ?
NO
YES
IS IT REASONABLE TO ASSUME THAT THE FLOW
FIELD IS UNIFORM, STEADY, AND REGULAR?
NO
YES
IS IT REASONABLE TO ASSUME THAT
THE SITE GEOMETRY IS REGULAR?
NO
YES
ARE THE SELECTED REMEDIAL ACTIONS
RELATIVELY SIMPLE IN CONFIGURATION?
NO
I
YES
DOES THE POLLUTANT HAVE RELATIVELY
THE SAME DENSITY AS WATER?
NO
YES
DO YOU HAVE
SUFFICIENT RESOURCES
AND AVAILABLE DATA
FOR NUMERICAL MODELS?
USE LEVEL I: ANALYTICAL MODEL
NO YES
USE LEVEL II: NUMERICAL MODEL
Figure 4.1 Flow chart to determine the level of modeling
required for soil and ground-water systems.
1-17
-------
modeling may be appropriate. If media properties cannot be
simplified, a Level II model is required. This question is
posed because only numerical models can explicitly represent
variability in media properties, such as porosity and
hydraulic conductivity. The number of simplifying
assumptions made is dependent upon both the user's expertise
and knowledge of the site. Therefore it is critical that
the user be able to characterize media properties, and be
cognizant of the limitations associated with these
assumptions.
The second of these questions concerning site conditions is:
"Is it reasonable to assume that the flow field is
uniform and steady?"
Unsaturated zone flow is most often unsteady and irregular,
except in those cases where infiltration is relatively
constant, as occuring under seepage/recharge basins. In
many cases, flow and transport processes in the unsaturated
zone may be neglected so that simulation of saturated zone
processes may be sufficient. Flow conditions must be
reasonably uniform, steady, and regular for analytical
(level I) models to be applied. Uniform flow refers to flow
that is in one direction (e.g., radial flow), and does not
vary across the width of the flow field. Steady flow does
not change over time. This occurs where boundary conditions
such as pumping/injection and recharge from rainfall or a
stream are constant over time.
The third site condition question to be asked is:
"Is it reasonable to assume that the site geometry is
regular?"
Examples of regular site geometry include constant aquifer
thickness and rectangular, circular or square shaped site on
a plan view. As with media property variability, no site is
totally rectangular, square, or conical, nor are aquifers
equally thick everywhere. However, some hazardous waste
sites can be approximated in this manner. Those with
rectangular surface impoundments and single or double layer
shallow aquifers are an example. Some examples of where
these simplifications cannot be made include aquifers with
fractured bedrock or aquifers with highly irregular
boundaries. If this question can be answered in the
affirmative, go on to the next question.
The next question deals with remedial action requirements in
1-18
-------
terms of model selection:
"Are the selected remedial actions relatively simple in
configuration?"
As noted in Section 2, remedial action criteria are an
integral part of model selection, and thus a central issue
in the selection methodology in this manual. Some remedial
actions and specific configurations cannot be simulated by
analytical models because they must be represented using
variable media properties, or they cause perturbations in
the flow field so that flow is not uniform or steady.
Remedial actions and configurations that fit this category
include: permeable treatment beds and partially-penetrating
wells and drains. If these actions are not to be selected
for detailed analysis and conceptual design, then proceed to
the next question.
Pollutant characteristics also affect the level of model
required. The primary question to be asked here is:
"Does the pollutant have relatively the same density as
water?"
If the pollutant has relatively the same density as water it
will be advected by the water although it can also be
retarded, or slowed, if it sorbs to the media. Pollutants
that are much heavier or lighter than water will not mix or
be advected entirely by the water; the result being
two-phase flow. In these cases the pollutant mass, either
as a liquid or a gas, exhibits its own flow with that of
water. This phenomena is extremely difficult to represent
and only a select group of complex numerical models are
capable of adequately representing pollutant movement under
these conditions.
If all of the above questions were answered in the
affirmative, a Level It Analytical Model is appropriate for
use. However, the user may have to use this flow chart
iteratively when working through the Remedial
Investigation/Feasibility Study, as objectives according to
each stage vary.
If any of the above questions were answered in the negative,
the user must ask the following question:
"Do you have sufficient resources and available data for
numerical models?"
1-19
-------
If the user has on the order of four to eight man-months of
time and has the requisite data sets (see Section 4.4.2) for
calibration and validation, the answer is affirmative, and a
Level II; Numerical Model is appropriate for use. If,
however, resources(Tncluding computer facilities, expertise
and time) and data are not adequate, the user must reassess
the goals and data needs for the current stage, and
re-iterate the level of modeling decision process again.
4.3 REQUIRED MODEL CAPABILITIES FOR EACH LEVEL
Once the appropriate level of model is chosen, the user
should identify the required model capabilities for the
site, based on flow field, critical processes, and remedial
action criteria. Section 4.3.1 covers the required model
capabilities for Level I models, Section 4.3.2 covers the
same for Level II models.
Matricies are used at this stage to facilitate the
identification of required model capabilities based upon a
wide range of potential scenarios. Figure 4.2 is a flow
chart that illustrates the framework for identification of
model capabilities and model selection. After the required
model capabilities are identified, a discussion on general
model selection criteria, and data and resource availability
for both levels of models is provided.
4.3.1 Level I Analysis
A Level I analysis is appropriate for remedial action
assessment if the user has answered "yes" to the first six
questions in Figure 4.1: Flow Chart To Determine the Level
of Modeling Required.
The available Level I methods for subsurface remedial action
assessment are fairly specific in terms of the type of
remedial actions that can be evaluated, and have limitations
for use that should be observed. For example, the conformal
mapping method is appropriate for simulation of
fully-penetrating and hanging impermeable barriers. Table
4.2 is a matrix of remedial action configurations vs.
simplified and analytical methods. The remedial action
configurations listed on the "Y" axis are the same as those
listed in Table 4.1.
1-20
-------
TABLE 4.1 SOIL AND GROUND-WATER REMEDIAL ACTIONS
Grading
Revegetalion
Surface Water Diversion
Capping
Seepage Basins and Ditches
Interception Trenches:
o Fully-penetrating
o Partially-penetrating
Ground-Water Pumping:
o Fully-penetrating levels
o Partially-penetrating wells
Impermeable Barriers
o Fully-penetrating
o Hanging
Drains
Permeable Treatment Beds
Bioreclamation/Chemical Injection
Excavation/Hydraulic Dredging
Solution Mining/Extraction
1-21
-------
LEVEL OF MODELING REQUIRED
(FROM FIGURE 3.1)
LEVEL I: ANALYTICAL MODELS
LEVEL II: NUMERICAL MODELS
i
IDENTIFICATION OF REMEDIAL
ACTION-SPECIFIC MODELS
AL
3
1
MODEL SELECTION CRITERIA
PROCESSES
DIMENSIONALITY
TIME
FRAME
RESOURCES/
DATA
MODEL SELECTION
(REFER TO VOLUME II FOR
REPRESENTATIVE MODELS)
MODEL SELECTION
(REFER TO VOLUME III FOR
REPRESENTATIVE MODELS)
Figure 4.2
Flow chart for required model capabilities for
soil and ground-water systems.
1-22
-------
TABLE 4.2 REMEDIAL ACTIONS VS. SIMPLIFIED AND ANALYTICAL METHODS MATRIX
REMEDIAL ACTIONS
SIMPLIFIED AND ANALYTICAL METHODS
tn Q
H
-01
4J
-------
There are 11 simplified and analytical methods available for
Level I remedial action assessment. These methods include:
runoff estimation; sediment yield; well hydraulics; drain
hydraulics, ground-water mounding; superposition; pond
seepage estimation; infiltration estimation; transformation;
conformal mapping; and contaminant transport. Some of these
methods encompass the theory used to develop different types
of solutions (e.g., well hydraulics, mounding, conformal
mapping and contaminant transport), whereas others encompass
the theory behind the use of these solutions to evaluate
relatively complex geohydrological conditions (e.g,
superposition and transformation). Volume 2: Simplified
Methods for the Evaluation of Subsurface Remedial Actions,
provides a good summary of selected analytical methods. The
user should refer to this volume for descriptions of the
methods, examples of applications, and appropriate
references. The user should identify all the methods that
are applicable to the remedial actions being screened and
consult the appropriate references in order to choose the
specific techniques that are applicable to the specific
site.
4.3.2 Level II Analysis
If the user answered 'no1 to any of the first six questions
in Figure 4.1, Level II (Numerical) methods should be
considered for remedial action assessment at this site.
Resources and data availability must also be answered in the
affirmative in Figure 4.1 to allow numerical model use.
Identification of required model capabilities for Level II
analysis is more complex than for Level I analysis, due to
the large number of models available with variable
capabilities. It is accomplished by evaluating three
primary groups of remedial action criteria: environmental
processes that are affected by remedial actions and thus
should be represented in a model; the minimum required
dimensionality; and time frame. The first two areas are
represented by matrices. The time frame requirements are
discussed in the text. Time frame requirements are
extremely site-specific, thus general guidance is given.
These areas are described below.
A measure of remedial action effectiveness is how well the
action controls or affects specific environmental processes
that are responsible for off-site contamination problems.
For example, impermeable barriers control the movement of
1-24
-------
ground water and contaminants. Remedial actions are most
often simulated by adjusting the parameters of specific
environmental process equations in the model. Therefore,
affected processes for each remedial action should be
identified in order to ensure that the selected model
considers them.
Table 4.3 is a matrix of remedial actions vs. required
processes. The processes are divided into 3 zones: overland
(or surface), unsaturated, and saturated. These zones also
correspond to the general types of models available. The
processes include: runoff, evapotranspiration, erosion, and
infiltration in the overland zone; percolation/leaching,
dispersion, retardation, degradation, and drainage in the
unsaturated zone; and ground-water movement, dispersion,
retardation, and degradation in the saturated zone. The
processes of infiltration and drainage can be considered as
inter-zone, as they describe water movement between the
overland-unsaturated zones and unsaturated-saturated zones,
respectively. A '^ ' in the blocks indicates that the
selected model should simulate that specific process in
order to represent the effects of a given type of remedial
action.
Once the user has identified processes to be represented in
the selected model based upon the remedial actions subject
to analysis, the required model dimensionality should be
identified. Table 4.4 is a matrix of required
dimensionalities for remedial actions as a function of
zone. The zones specify the number of land segments
(parcels of the surface zone separated into areas of uniform
properties) required for the overland zone, and the number
of dimensions and direction(s) for both unsaturated and
saturated zones. By reviewing this matrix, the user can
identify the required spatial domain or dimensionality of
the selected model.
The third area of remedial action criteria is the time frame
requirements. Numerical models can simulate chemical
transport and fate in two modes: using a steady-state mode,
where fluxes such as water velocity and pollutant loading
are constant or time invariant; or using a transient mode,
where flow and/or contaminant transport may vary over time.
Most models may run in either mode, depending on the input
data and specifications by the user. The flow field
throughout the system is usually established or calculated
first, then the transport part of the model utilizes the
flow field velocities to calculate pollutant transport. In
this way, time frame may be specified separately for both
1-25
-------
9Z-T
£ 5 S 7
go a • • • TJ • •
t-*-1 7 it
o o •
"-IS
:3!
1 ° s
9
I 3 M H-
nt ft (•
I O rr n
EVAPOTRANS-
PIRATION
INFILTRATION
PERCOLATION/
LEACHING
DISPERSION
RETARDATION
DEGRADATION
DRAINAGE
GROUND-WATER
MOVEMENT
DISPERSION
RETARDATION
DEGRADATION
M
*>
W
s
w
D
f
>
n
O
z
en
M
O
G
M
D
TJ
50
O
O
H
Cfi
cn
M
CO
50
H
X
-------
TABLE 4.4 REMEDIAL ACTIONS VS. REQUIRED MODEL
DIMENSIONALITY MATRIX
REMEDIAL ACTION
Grading
Revegetation
Surface Water
Diversion
Capping
Seepage Basins
Drains
Fully-penetrating
wells
Partially-penetra-
ting wells
Ground-water Pumping
Fully penetrating
wells
Partially-penetra-
ting wells
Impermeable Barriers
Fully-penetrating
Hanging
Interceptor Trenches
Overland
No. of
Segments
S
S
M
Zone
Unsaturated
No. of
Dimen-
sions
1
1
Direc-
tion^)
za
z
Saturated
No. of
Dimen-
sions
b
2
2
3
2
3
2
3
2
Direc-
tion( s)
b
X,Y
X,Y
X,Y,Z
X,Y
X,Y,Z
X,Y
X,Y,Z
X,Y
1-27
(continued)
-------
TABLE 4.4 (continued)
Zone
REMEDIAL ACTION
Permeable Treatment
Beds
Bioreclamation/
Chemical Injection
Excavation/Hydraulic
Dredging
Solution Mining/
Extraction
Dverland
So. of
Segments
Unsaturated
No. of
Dimen-
sions
1
1
Direc-
tion(s)
Z
Z
Saturated
No. of
Dimen-
sions
c
2d
Direc-
tion(s)
c
X,^
Denotes vertical direction
Only if Ground-water mounding from the seepage basin is
significant
ft
Assumes treatment bed is constructed so as not to modify
the flow field
Assumes injection/extraction wells are fully penetrating
S = single
M = multiple
1-28
-------
the flow and transport 'modules'. The time frames required
to properly represent the effects of a remedial measure
depend on the hydrologic zone, the important processes, and
the remedial measure itself. For example, remedial actions
designed for control of erosion and runoff such as grading
and surface water diversion could require a transient (or
dynamic) simulation with short time steps because rainfall
and runoff processes fluctuate rapidly. In the saturated
zone, the flow field is usually more steady, and
fluctuations occur on a scale of months and years. Thus, a
steady-state simulation is usually applicable. However, a
transient simulation may be required if recharge and
discharge from pumping/injection or hydraulically-connected
streams fluctuate over the simulation period, causing the
flow to be unsteady. Contaminant transport usually requires
a transient mode when simulating remedial actions, as model
results of interest include the variation in concentrations
in pre-and post-restorative periods.
4.4 RESOURCE AND DATA AVAILABILITY
As mentioned in Section 2, resource and data needs must be
examined when the level of modeling is determined. This
question is particularly important when it is apparent that
a Level II numerical model is required for simulation. A
brief overview of these issues is provided below; a detailed
examination of procedures for model use including estimation
of parameters for subsurface remedial actions are provided
in Volume 3.
4.4.1 Level I Analysis
Resource and data availability is not as critical to
performing a Level I analysis as is the user's expertise and
understanding of the site. Analytical and simplified
methods require very little data, can often be solved by
hand or with personal computers, and do not require a large
amount of time for implementation. However, use of these
methods does require an understanding of the assumptions and
limitations behind their development. A summary of basic
Level I (Analytical) model data needs is presented in Table
4.5. Sampling programs in the technology screening of
alternatives stage of the Remedial Investigation/
Feasibility Study should attempt to satisfy the data needs
of at least a Level I analysis.
1-29
-------
TABLE 4.5 DATA NEEDS FOR LEVEL I (ANALYTICAL) METHODS FOR
SUBSURFACE PROBLEMS (after Javandel et al.,
1983)
I. Geometry of System
o Average thickness and depth of aquifer
o Positions of significant features:
o Source(s) of contamination
o Discharge and recharge areas
II. Fluid (Water) Velocity
o Direction and magnitude of average regional
velocity in vicinity of study area
III. Concentration of Pollutant
o Source release rate
o History of operation of facility
IV. Dispersivity of Media
o Representative value of longitudinal dispersivity
for one-dimensional problems
o Representative values for both longitudinal and
transverse dispersivities for two-dimensional
problems
V. Pollutant Characteristics
o Retardation factors or distribution coefficients
for solutes that sorb to media
o Decay rate for solutes that are non-conservative
1-30
-------
4.4.2 Level II Analysis
Resource and data availability issues are more critical to
performing a Level II analysis compared to a Level I
analysis. Data sets that represent the range of values over
time and space are required for calibration and testing
(performing simulations to ensure "agreement" between
observed and predicted data), and remedial action
simulation, where parameters and site configurations are
adjusted to represent the selected measures. The
flexibility provided by numerical models to represent
spatial variability can lead to expanded data needs from the
increased number of e.i ements or compartments discretized in
the model. Table 4.6 provides a list of required data for
Level II methods. Since numerical models are more
applicable to detailed analysis and conceptual design tasks
in the Remedial Investigation/ Feasibility Study,
corresponding sampling programs in this stage should be
directed at meeting the needs for model simulations,
provided the need to model for a specific site is warranted.
Similarily, resource needs are more intensive than for Level
I. Demands will vary according to the complexity of the
modeling study, but generally, four to eight man-months of
time for an experienced user should be allocated. However,
time requirements can vary greatly, and will depend on the
application, also. This manpower requirement should be used
as a rough estimate, and is dependent on a number of
factors, including: expertise and experience of the user
with the selected model; availability of computer facilities
and software support; sufficient time to conduct the study;
and sufficient money to train the user, pay for his/her
time, and pay for the model, if it is proprietary and can be
obtained.
Resource needs are also affected by model performance
related criteria. The utility of a model depends not only
on its ability to represent site and remedial action
conditions, but also on model design, implementation,
testing, and documentation. These factors influence the
accuracy of model predictions, ease of use, data
requirements, and computer run costs. An additional, less
tangible, attribute of a model is its perceived reliability,
which is dependent on the number of times it has been
successfully implemented, the verification or testing of
model results against field measurements, and the technical
1-31
-------
TABLE 4.6 DATA NEEDS FOR LEVEL II (NUMERICAL) METHODS FOR
SUBSURFACE PROBLEMS (after Javandel et al., 1983)
I. Geometry of System
o Real extent of aquifer
o Location of natural or mathematical boundaries
o Thickness of aquifer and its variation with the
space
o Location and rates of discharge and recharge
areas
II. Fluid (Water) Velocity
o Distribution throughout the system
III. Concentration of pollutant
o History of operation of facility
o Present and future source rates of pollutant and
chemical composition
o Positions of sources relative to aquifer
IV. Dispersivity of Media
o Representative value of longitudinal dispersivity
for one-dimensional problems
o Representative values for both longitudinal,
transverse, and vertical dispersivities for
two- and three-dimensional problems
V. Pollutant Characteristics
o Retardation factors or distribution coefficients
for solutes that sorb to media
o Decay rate for solutes that are non-conservative
1-32
-------
(theoretical) basis for model calculations. These criteria
can be used to distinguish between models which satisfy all
of the site and remedial action criteria discussed earlier.
The most desirable models have extensive documentation, have
been applied to a number of diverse situations, have been
tested against several comprehensive data sets, and are
relatively efficient in terms of data preparation,
requirements and computer time. Models that exhibit these
characteristics will not require as much time and effort
compared to ones that do not exhibit the same
characteristics.
4.5 MODEL SELECTION CRITERIA FOR SOIL AND GROUND-WATER
CONTAMINATION PROBLEMS
By evaluating the matrices for Level I and II models the
user will be able to correlate model capabilities with
his/her site characteristics and selected remedial actions.
Some general trends of model selection criteria are apparent
in the matrices. These trends are described below to clarify
any confusion on the appropriate model capabilities
required.
In a Level I analyses, the selection of a specific method or
group of methods is primarily a function of the selected
remedial actions. For example, drain hydraulic methods are
applicable to drains, well hydraulic methods are applicable
to ground-water pumping/injection. The configuration of the
remedial action measure will also affect the type of
techniques that are applicable. For example, partially-
penetrating interceptor trenches and wells create more
complex flow patterns than do fully-penetrating drains and
wells; thus they require different techniques to account for
the more complex flow field.
For Level II analyses, there are general trends in each
major group of model selection criteria: processes, required
dimensionality, and time frame. These criteria are
discussed below.
The processes required for remedial action simulation are
often a function of the specific zone that is affected and
the pollutant characteristics. For example, remedial
actions such as grading, revegetation, and surface water
diversion require representation of the runoff and
infiltration processes in the selected model. Similarly, if
the pollutant has a high affinity for sorption, the
1-33
-------
mechanisms of retardation and soil erosion should be
represented. As it is expected that most subsurface
contamination problems will concern ground water, the
selected saturated zone transport model should usually
consider dispersion, retardation and degradation.
The minimum required dimensionality will also vary according
to the complexity of the site and selected remedial actions.
Some general requirements are apparent for each zone,
however. In the overland zone, the actions require a single
segment model or one that allows only uniform properties for
slope, roughness, and soil type. The other remedial action
requires a multiple segment model, or one that allows the
segmentation of the drainage area. Some overland zone
models allow multiple segments; such models give the user
flexibility to simulate all of the above actions.
In the unsaturated zone, a one-dimensional simulation in the
"Z" or vertical direction is often sufficient for the
evaluation of the few remedial actions that affect that
zone, such as excavation, seepage basins, and hydraulic
dredging. However, if there are soil layers with varying
permeability, a two dimensional horizontal-vertical (x-z)
simulation is required to represent the percolation of water
(Z direction) and the lateral interflow (x direction).
In the saturated zone, a two-dimensional simulation is the
minimum required dimensionality for most actions. If
mounding is not a problem, actions such as fully-penetrating
wells and interceptor trenches, fully-penetrating cut-off
walls, and drains may require a x-y flow pattern simulation.
Serious mounding problems also require a minimum of
two-dimensional x-y simulation, if not a three-dimensional
(x,y,z). Other actions can be represented with a x-y
simulation if the flow at the site has a neglible vertical
component. If it doesn't, a three dimensional (x,y,z)
simulation is required.
In terms of time frame, overland and unsaturated zone
measures such as grading, revegetation, surface water
diversion, and capping will require a dynamic simulation.
Remedial measures used in the saturated zone can often be
simulated using a steady-state mode for flow simulation,
although boundary conditions and monitoring data should be
evaluated to determine if a transient simulation is
necessary. Contaminant transport will almost always require
a transient simulation to examine reductions in
concentrations due to implementation of specific remedial
actions.
1-34
-------
SECTION 5
METHODOLOGY FOR MODEL SELECTION FOR SURFACE WATER
CONTAMINATION PROBLEMS
5.1 OVERVIEW
This section will assist the user in answering the second
two questions of the modeling decision framework for surface
water:
1. What level of modeling is required and
2. What are the required model capabilities for that
level
Flow charts and matrices form the basis of the methodology.
The matrices will introduce the user to relationships of
remedial actions to model selection criteria as processes,
dimensionality, and time frame. This section parallels
Section 4 for soil and ground-water contamination problems.
General model selection criteria and resource and data
availability issues are discussed in the latter part of this
section. Descriptions of surface water remedial actions
including required dimensionality and affected processes,
are provided in Volume 4: Simplified Methods and Numerical
Models for the Evaluation of Surface Water Remedial Actions.
If additional information is required, the user should refer
to available remedial action technology guides, such as JRB
(1982), Ehrenfield and Bass(1983), and SCS (1981). A list
of surface water remedial actions considered for simulation
is provided in Table 5.1. Environmental process
descriptions, along with the significant parameters,
environmental conditions of concern, and relation to other
processes are also provided in Volume 4. Additional sources
of information for these processes include Mills et al.
(1982) and Callahan et al. (1979).
1-35
-------
TABLE 5.1 SURFACE WATER REMEDIAL ACTIONS
I. No Action
II. Physical/Mechanical Measures
o Mechanical dredging
o Hydraulic dredging
o Excavation
o Dilution
o Barriers/diversions
o Skimming
o Cofferdams
o Booms
o Silt curtains
o Capping
III. Treatment
o In-situ
o On-site
1-36
-------
The state-of-the-art for remedial action assessment modeling
in surface water is not as advanced as modeling for ground
water. The available Level I and Level II methods can
adequately represent the wide range of waterbody conditions,
but have not been as extensively tested for remedial action
assessment as have ground-water methods. A key problem is
estimating parameter values for specific remedial actions.
Quite often, best engineering judgement will have to suffice
for making the appropriate parameter adjustments for
remedial action simulation. Therefore, the user should
consider the uncertainty inherent in the representation of
remedial actions when selecting a model.
5.2 WHAT LEVEL OF MODELING IS REQUIRED?
There are six basic questions to be answered when
determining the level of model required for surface water
contamination problems. Figure 5.1 is a flow chart that
illustrates the hierarchy of decisions to be made. The
first five questions should be answered in the affirmative
for analytical or simplified methods (Level I) to be chosen.
A "no" answer for any of these questions forces the user to
consider the use of a numerical model (Level II), whereupon
a resource and data availability decision must be made. If
both resources and data are available, the numerical model
should be chosen. If either resources or data are not
available, the user will be directed to reassess project
goals and/or data needs, and re-iterate the decision process
flow chart. This hierarchy was developed as such in order
to define the strict and limited conditions of analytical
and simplified methods used in remedial action assessment.
Environmental conditions such as un-steady flow regimes,
non-uniform geometry, and complex sediment-water
interactions, cannot be accurately represented by Level I
methods. However, the use of Level I methods for remedial
action assessment in surface water, while more limited than
analogous use for soil and ground-water problems, may be
more appropriate given the limited testing and parameter
estimation available.
As with subsurface problems, the flow chart questions may
also be posed when selecting models for exposure assessment.
However, model requirements for conducting exposure
assessments will most likely be less stringent, due to the
fact that only the complexity of site conditions and
modeling objectives determine model selection. The
1-37
-------
REASSESS
GOALS AND
DATA NEEDS
ARE ORDER OF MAGNITUDE
PREDICTIONS ACCEPTABLE?
NO
YES
IS THE FLOW UNIFORM, STEADY.
AND VERTICALLY WELL-MIXED?
NO
I
YES
CAN THE WATER BODY GEOMETRY BE SIMPLIFIED
INTO REACH(ES) WITH UNIFORM PROPORTIONS?
NO
YES
ARE THE SELECTED REMEDIAL
ACTIONS RELATIVELY SIMPLE?
NO
YES
DOES THE POLLUTANT HAVE RELATIVELY THE SAME
DENSITY AS WATER, AND IS IT SOLUBLE IN WATER ?
NO
DO YOU HAVE
SUFFICIENT RESOURCES
AND AVAILABLE DATA
FOR NUMERICAL MODELS?
NO
YES
USE LEVEL I: ANALYTICAL MODEL
YES
USE LEVEL Ih NUMERICAL MODEL
Figure 5.1 Flow chart to determine the level of modeling
required for surface water systems.
1-38
-------
questions to determine the level of modeling required for
remedial action assessment are described below.
The first question to be asked is universal to both surface
water and subsurface contamination problems:
"Are order-of-magnitude predictions acceptable?"
This decision is a function of the current task at hand of
the Remedial Investigation/Feasibility Study, and the
relative complexity of the site. For screening of
alternatives, order-of-magnitude estimates are often
acceptable. When detailed analysis and conceptual design is
initiated, there is a need for higher accuracy in
quantitative assessments of remedial action performance.
Numerical models are much more accurate than .analytical
models when site conditions cannot be well-characterized,
and are usually more appropriate for use. There are many
cases however, where accuracy needs are independent of the
current stage of the Remedial Investigation/Feasibility
Study. These cases are dependent on the complexity and
number of remedial actions being evaluated. If
order-of-magnitude estimates are acceptable, proceed to the
next question.
The next two questions concern the degree of heterogeneity
in site conditions. The first question is:
"Is the flow uniform, steady, and vertically
well-mixed?"
As with soil and ground-water problems, the flow field must
be relatively simple if analytical models are to be used.
Again uniform flow refers to flow that does not vary
spatially i.e., along the length of the waterbody. Steady
flow refers to flow that does not vary over time. The
degree of vertical mixing is important in model level
selection because un-mixed, stratified waterbodies, such as
impoundments and estuaries, have complex and sometimes
bi-directional flow and thus require numerical methods. A
vertically well-mixed waterbody, then, has essentially
uniform flow in the "Z" or vertical dimension.
The second of the two questions concerning site
heterogeneity is:
"Can the waterbody geometry be simplified into reach(es)
with uniform proportions?"
1-39
-------
One of the limitations of analytical models is that the
system being represented must be simple or regularly shaped.
In surface waters, this means the waterbody has to be
segmented into reaches or lengths of uniform size.
Obviously, some waterbodies, such as estuaries and
delta-like rivers, do not lend themselves to such
simplification. The user must examine the shape and
configuration to determine if an average width, length, and
depth can be utilized.
The next question is:
"Are the selected remedial actions relatively simple?"
The use of analytical or simplified methods for remedial
action assessment in surface waters is limited to very few
measures. Some surface water measures usually cause
perturbations in the flow, or involve complex sediment
processes such as sediment transport and exchange from bed
sediments, both of which often cannot be accurately
represented by analytical or simplified methods. Such
methods include: hydraulic dredging, barriers, skimming, and
booms. Other measures such as mechanical dredging can be
represented with Level I methods, but only with gross
simplifications as to their effect in the waterbody.
The most important question to ask concerning pollutant
criteria is:
"Does the pollutant have relatively the same density as
water, and is it soluble in water?"
If the pollutant is soluble and has the same density as
water, it will be transported along with the water, and may
be accurately represented by a Level I method. If the
pollutant floats on the surface or sinks to the bottom, or
is insoluble, two different types of transport processes
must be coupled (or interact with each other): one for water
movement or advection, the other for the pollutant movement.
This phenomenon can be simulated only by a select group of
numerical models, and is similar to the two-phase flow
phenomenon mentioned in Section 4.2.
If all the above questions can be answered in the
affirmative, a Level I Method is appropriate for use. If
any of the above questions is answered in the negative, the
user must consider numerical models for use. Before
selection of such methods however, the following question
should be asked:
1-40
-------
"Do you have sufficient resources and data for numerical
models?"
Numerical models are both cost and data intensive.
Therefore a minimum of four to eight man-months of time, and
requisite data (see Section 5.3.2) for testing and
prediction purposes should be available. If the answer is
affirmative, Level II (Numerical) methods should be
selected. If, however, the resources and/or data are
inadequate, the user must reassess the goals and data, and
re-iterate the level of modeling decision process again.
5.3 REQUIRED MODEL CAPABILITIES FOR EACH LEVEL
Matrices are used at this stage to correlate remedial
actions with their required model capabilities for
simulation. Figure 5.2 is a flow chart that illustrates the
sequence of events that lead to model selection. After the
level of modeling is determined, the user should refer to
the specific matrices or text to identify required model
capabilities. In a Level II analysis three main groups of
model selection criteria are examined: environmental
processes, waterbody/dimensionality, and time frame. After
model capabilities are identified, the user is referred to
Volume 4, which contain model profiles and sources of
information on models.
5.3.1 Level I Analysis
A Level I analysis is appropriate for remedial action
assessment if the user has answered "yes" to the first five
questions in Figure 5.1. It is likely that Level I methods
will be sufficient in the screening of alternatives task of
the Remedial Investigation/ Feasibility Study; order of
magnitude estimates of technical performance are desired.
However, Level I methods are also appropriate for the
analysis of alternatives task if site conditions and
remedial actions can be simplified (i.e., steady, uniform
flow and selection of such actions as excavation, dilution,
or on-site treatment). Unlike Level I methods available for
soil and ground-water, surface water methods are not
specific in terms of remedial action simulation. Like
numerical methods, they have various capabilities and hence,
various uses. Table 5.2 is a matrix of simplified
1-41
-------
LEVEL OF MODELING REQUIRED
(FROM FIGURE 4.1)
LEVEL I: SIMPLIFIED AND
ANALYTICAL MODELS
LEVEL II: NUMERICAL MODELS
i
IDENTIFICATION OF
REMEDIAL ACTIONS THAT ARE
EASILY SIMULATED
I
MODEL SELECTION CRITERIA
PROCESSES
WATERBODY/
DIMENSIONALITY
TIME
FRAME
RESOURCES/
DATA
MODEL SELECTION
(REFER TO VOLUME IV
FOR REPRESENTATIVE MODELS)
MODEL SELECTION
(REFER TO VOLUME IV
FOR REPRESENTATIVE MODELS)
Figure 5.2
Flow chart for required model capabilities for
surface water systems.
1-42
-------
TABLE 5.2 SIMPLIFIED AND ANALYTICAL SURFACE WATER
MODELS VS. REMEDIAL ACTIONS MATRIX
Simplified Analytical Methods
Remedial
Actions
Contaminant
Transport
Sediment
Processes
Estimation
of Loading
Transformation
Processes
Mechanical1
dredging
Hydraulic1
dredging
Excavation
Barriers1
Skimming1
Dilution
Cofferdams
Boomsl
Silt
curtains1
Capping
On-site1
treatment
In-situJ
treatment
T,P
D,A
D,A
B,V
M
T,P
B,V
D,A
JThis action cannot be
represented accurately
using simplified or
analytical methods.
2 The transformation process
required will be a function
of the specific pollutant
characteristics.
Advection-Dispersion
Equation
B = Bed Exchange Analysis
D = Dilution Analysis
M = Mixing Zone Analyses
P = Partitioning (Sorption)
T = Sediment Transport Analysis
V = Vertical Distribution
of Sediments Analysis
1-43
-------
techniques and analytical methods vs. remedial actions.
Four major groups of model capabilities are listed across
the 'X' axis: contaminant transport, sediment processes,
estimations of loading, and transformation processes. In
the boxes of the matrix the specific methods of these major
groups are identified for each remedial action. Level I
methods include simplified assessment techniques and
analytical models.
There are a number of simplified assessment techniques that
can be used to represent contaminant transport, sediment
processes, estimation of loading and transformation
processes. Volume 4: Simplified Methods and Numerical
Models for the Evaluation of Surface Water Remedial Actions
provides descriptions of these techniques and appropriate
references.
Like the simplified assessment techniques, analytical models
require steady-state flow conditions and uniform geometry.
Their applicability can be limited, given the unsteady flow
regimes, non-uniform geometry, and complex sediment-water
interactions that characterize the environmental conditions
when remedial actions are implemented. Within this general
group of models, differences can include: complexity of
geometry allowed, mode of pollutant loading (instantaneous
or continuous), degree of mixing and dispersion (if any),
ability to calculate transfer of mass between the sediment
bed and the water column, methods of estimating sediment
transport (user input suspended sediment concentrations, or
concentrations calculated for each reach separately), lumped
or specific first-order decay reactions, and the range of
default values available for model parameters.
As was mentioned earlier in this section, Level I methods
have very limited roles for remedial action assessment in
surface waters. Of the thirteen measures listed in Table
5.2, five cannot be accurately represented using simplified
or analytical methods. Some require simulation of complex
geometry configurations (i.e., barriers), complex sediment
processes (dredging), and low/high density pollutant
transport and fate (skimming, booms). However, many of
these complex remedial actions, such as skimming and booms,
are usually employed under emergency response conditions,
and this would not allow the time or resources for numerical
simulation. Volume 4 contains a matrix of representative
methods vs. model capabilities, and descriptions of
representative models. Some other good sources for
compilations of available methods include Mills et al.
1-44
-------
(1982) for simplified methods, and Codell et al. (1982) and
Onishi et al. (1982) for analytical models.
5.3.2 Level II Analysis
If the user answered 'no' to any of the first five questions
in Figure 5.1, and determined that sufficient data and
resources were available, Level II methods should be
considered for remedial action assessment. This subsection
will help identify the required model capabilities according
to site and remedial action factors. The selection of
numerical models is appropriate when site/remedial action
conditions cannot be reasonably simplified, or when specific
configurations of a remedial action, such as a barrier, must
be evaluated for the best conceptual design.
Identification of required model capabilities for Level II
analysis is more complex than for Level I, due to the large
number of models available with variable capabilities.
Similar to identification of required model capabilities for
soil and ground-water problems, it is accomplished by
evaluating three groups of remedial action criteria:
environmenta.1 processes that are affected by remedial
actions and thus should be represented in a model;
dimensionality according to the remedial action and
waterbody: and the time frame. These groups are described
below.
A measure of remedial action effectiveness is how well the
action controls or affects specific environmental processes
that are responsible for off-site contamination problems.
For example, barriers control the advection and dispersion
of surface water and contaminants. Remedial actions are
most often simulated by adjusting the parameters of specific
environmental process equations in the model. Therefore,
affected processes for each remedial action must be
identified in order to ensure that the selected model
simulates them.
Table 5.3 is a matrix of remedial actions vs. environmental
processes. The processes are divided into two major groups
across the 'x' axis: transformation and physical.
Transformation processes include: hydrolysis, oxidation,
photolysis, volatilization, bio-degradation, and bio-
accumulation. Physical processes include adsorption,
sediment (transport and bed-exchange of contaminated
sediments), advection, and dispersion. Short descriptions
1-45
-------
TABLE 5.3 REMEDIAL ACTIONS VS. PROCESSES MATRIX
PROCESSES
ACTIONS
DEGRADATION
PHYSICAL
Illltilllll
DILUTION
REMOVAL
MECHANICAL
DREDGING
EXCAVATION
HYDRAULIC
DREDGING
BARRIERS/
DIVERSIONS
SKIMMING
CONTAINMENT
COFFERDAMS
BOOMS
SILT CURTAINS
CAPPING
TREATMENT
IN-SITU
ON-SITE
0
0
0
0
0
0
0
0
0
0
0
0
+
0
0
+
LEGEND:
+ = ENHANCES THE PROCESS IN RELATION TO NO ACTION
- = MITIGATES THE PROCESS IN RELATION TO NO ACTION
0 = DOES NOT AFFECT THE PROCESS
1-46
-------
of these processes are provided in Volume 4. A ' + ' in the
matrix boxes indicates that the remedial action enhances the
process in relation to no action. A '-' indicates that the
remedial action mitigates that process in relation to no
action. If either of these effects for remedial action are
identified, the selected model should be able to simulate
those processes. A '0' indicates that the remedial action
has no effect on that process; hence representation of that
process is not critical in the selected model.
The second major criteria group is the dimensionality
requirements. Table 5.4 is a matrix of remedial actions vs.
waterbody. The minimum model dimensionality is a function
primarily of the waterbody type. The waterbodies across the
'X1 axis are grouped as estuary, lake, or river, with
subgrouping within each according to system geometry and
degree of mixing. Numbers and letters in the matrix text
denote the type of simulation needed for that remedial
action in the specific waterbody. For example, "2L" denotes
that a two-dimensional (lateral/longitudinal) simulation is
required for that remedial action/waterbody scenario. A "0"
indicates that the remedial action is not suited for use
under the specific waterbody conditions. Some remedial
actions, such as dilution and barriers or diversions, often
may be simulated by adjusting the model boundary conditions
and system geometry. Most of the remedial actions require a
two-Dimensional (longitudinal/lateral) simulation. However,
as the mixing becomes more turbulent or complex (as in
estuaries and large lakes), a "pseudo" two-Dimensional
simulation (longitudinal/ vertical) with coefficients for
the horizontal or lateral dimension, or even three-
dimensional simulation, may be required.
The third area of remedial action criteria is the time frame
requirements. Like ground-water models, numerical models
for surface water can simulate chemical transport and fate
in two modes: using a steady-state mode, where fluxes such
as watar velocity and pollutant loading are constant or time
invariant; or using a transient mode, where flow and/or
contaminant transport may vary over time. Most models may
run in either mode, depending on the input data and
specifications by the user. The flow field throughout the
system is usually established or calculated first, then the
transport part of the model utilizes the flow field
velocities to move contaminant particles. In this way, time
frame may be specified for both the flow and transport
'modules'. The time frames required to properly represent
the affects of a remedial measure depends on the flow regime
or type of waterbody, the important processes, and the
1-47
-------
TABLE 5.4 REMEDIAL ACTIONS VS. WATER BODY MATRIX
ESTUARIES
REMEDIAL
ACTIONS
LAKES
RIVERS
NO ACTION
REMOVAL
MECHANICAL
DREDGING
EXCAVATION
HYDRAULIC
DREDGING
BARRIERS/
DIVERSIONS
SKIMMING
DILUTION
CONTAINMENT
COFFERDAMS
BOOMS
SILT
CURTAINS
CAPPING
TREATMENT
IN-S1TU
ON-SITE
1
1
2L
0
1
2L
3
2V
2V
2V
2V
0
2V
3
3
2V
21
0.
0
2L
2L
0
2L
2L
0
0
2L
3
2V
3
3
0
0
2V
0
3
3
2V
0
EPENDAfcT ON R
0
MOVAL
2L
2L
2L
2L
2V
0
2L
2V
0
0
:TION
2P
2L
2L
2V
2V
3
3
2V
2L
0
SED IN
0
2L
2L
0
IB
2L
0
0
2L
CONJUNt TION
IB
IB
2L
0
1
2L
0
2V
2L
2L
2L
21
21
21
0
'IB
2L
0
0
2L
LEGEND:
1 = 1-DIMENSIONAL
2 = 2-DIMENSIONAL
3 = 3-DIMENSIONAL
L = LATERALLY AVERAGED
V = VERTICALLY AVERAGED
0 = ACTION IS NOT APPLICABLE
TO THIS WATERBODY
B = BRANCHING OR NETWORK
1-48
-------
measure itself. For example, remedial actions that alter
the flow regime, such as hydraulic dredging and barriers may
require a transient or dynamic simulation of the flow
system. Similarly, some waterbodies, such as branching
estuaries and snowmelt-fed rivers, require a dynamic
simulation when the flows fluctuate within the simulation
period, which could be a number of days (for an estuary), or
over the course of a year (for a river). As with
ground-water models, the contaminant transport simulation
must always be in a transient or dynamic mode. When
simulating remedial actions, the interest is in predicting
variations in concentrations from baseline conditions to
different configurations of selected remedial actions.
5.4 RESOURCE AND DATA AVAILABILITY
The same issues of resource and data availability for
subsurface methods apply to surface water methods. Tables
5.5 and 5.6 list the data needs for surface water assessment
for simplified and analytical, and numerical methods,
respectively. The user should refer to Section 4.4 for a
generic overview of resource and data needs.
5.5 MODEL SELECTION CRITERIA FOR SURFACE WATER REMEDIAL
ACTION ASSESSMENT
At this point, the reader has become familiar with remedial
actions in terms of affected processes, minimum
dimensionality required, and time frame requirements.
In light of this information derived from the matrices, some
trends become apparent:
o Most removal measures affect sediment-water inter-
actions, particularly adsorption and sediment
deposition, erosion, and transport
o Physical processes such as longitudinal dispersion
and advection are more greatly affected by remedial
actions than are chemical/biological processes
o Remedial actions are specific to the type of
waterbody as well as to the type of discharge or
chemical
1-49
-------
TABLE 5.5 DATA NEEDS FOR LEVEL I (ANALYTICAL) METHODS FOR
SURFACE WATER PROBLEMS
I. Geometry of System
o Uniform reach or waterbody Size: length, width,
depth
II. Flow
o Average representative flow or velocity
III. Source of Pollutant
o Representative continuous rate, or specific
pulse, or
o Initial concentration from near field analysis
IV. Pollutant Characteristics
o Lumped decay or specific transformation rates if
pollutant is non-conservative
o Sediment concentrations and sorption coefficient,
or
o Sediment size, diameter, sorption coefficient,
and channel slope (if pollutant is hydrophobic)
1-50
-------
TABLE 5.6 DATA NEEDS FOR LEVEL II (NUMERICAL) METHODS FOR
SURFACE WATER PROBLEMS
I. Geometry of System
o Size of specific reaches: length, width, depth
II. Flow
o Distribution of flow or velocity (or depth and
width) throughout the system
III. Source of Pollutant
o Present and future source rates
o Locations of sources
IV. Dispersion
o Average representative longitudinal dispersion
coefficient for one-dimensional problems, both
longitudinal and transverse dispersion
coefficients for two-dimensional problems
o Time-varying coefficients for estuarial
simulation
V. Pollutant Characteristics
o Lumped decay rate or specific transformation
rates if pollutant is non-conservative
o Sorption coefficients for each sediment type;
sediment density and diameter, channel slopes,
and bed exchange rate if pollutant is hydrophobia
1-51
-------
Model selection criteria for a Level I analysis is a
function of the pollutant characteristics (i.e., whether it
sorbs and can be transported by sediments) and whether the
remedial action can be accurately represented by a
simplified or analytical method. Quite often sediment
processes will have to be represented on a gross level to
represent many of the remedial actions. However, in many
cases, parameter estimation for most remedial actions will
be very difficult. Numerical simulation with default or
uncertain parameter values will lead to potentially spurious
results. In these cases it is advisable to utilize
simplified or analytical techniques. The user should refer
to Volume 4 for backup information and references for Level
I methods.
Model selection criteria for Level II analysis are more
specific. Some general guidelines include:
o The simulation should usually be dynamic (time
varying) in order to simulate uneven flows (as in an
estuary) and pulse (spill) inputs of pollutants
o The spatial domain (dimensionality) will vary
according to the remedial action, but most actions
require a two-dimensional, vertical/longitudinal
simulation
o Many pollutants are hydrophobic; thus, the ability
to simulate sorption and sediment transport is
critical. Suspended sediments are a heterogeneous
mixture, requiring a model that can simulate the
various types, including organic matter which is
very important for sorbing organic pollutants. The
simulation of deposition and resuspension is
important also for the above mentioned reasons
o The ability to simulate degradation processes is
very important for those pollutants that readily
dissolve or are susceptible to volatilization and
photolysis. This is apparent when performing a
baseline assessment to determine the persistence of
the pollutant in the system. In addition, toxic
daughter products (degraded forms of the pollutant)
may be subject to specific sorption and degradation
effects. Therefore, degradation kinetics should
not be discounted in long-range fate analyses.
1-52
-------
REFERENCES
Adkins, L.M., J.J. Doria and M.T. Christopher. 1983. Methods
for Assessing Exposure to Chemical Substances - Vol. 3:
Methods for Assessing Exposure From Disposal Of Chemical
Substances, EPA 560/5-83-016, U.S. Environmental Protection
Agency, Office of Pesticides and Toxic Substances,
Washington, B.C.
Callahan, M.C., M.W. Slimak, N.H. Gabel, J.P. Map, C.F. Fowler,
J.R. Freed, P., Jennings, R.L. Durfee, F.C. Whitmore, B.
Maestri, W.R. Mabey, B.R. Holt and C. Gould. 1979.
Water-Related Environmental Fate of 129 Priority
Pollutants, EPA 440/4-79-029, Vol. 1,2, U.S. Environmental
Protection Agency, Washington, D.C.
Codell, R. B., K. T. Key and G. Whelan. 1982. A Collection of
Mathematical Models for Dispersion in Surface Water and
Groundwater, NUREG-0868, U.S. Nuclear Regulatory
Commission, Washington, D.C.
Dynamac Corporation. 1982. Methods for Assessing Exposure to
Windblown Particulates, U.S. Environmental Protection
Agency, Office of Health and Environmental Assessment,
Washington, D.C.
Ehrenfield, J.R. and J.M. Bass. 1983. Handbook for Evaluating
Remedial Action Technology Plans, EPA 600/2-83-76, U.S.
Environmental Protection Agency, Municipal Environmental
Research Laboratory, Cincinnati, OH.
Environmental Protection Agency. 1982. Workshop Summary, Level
II Predictive Exposure Assessment. April 27-29, 1982,
Atlanta, Georgia. U.S. Environmental Protection Agency,
Athens, GA.
Farino, W., P. Spawn, M. Jasinski and B. Murphy. 1983.
Evaluation and Selection of Models for Estimating Air
Emissions from Hazardous Waste Treatment, Storage and
Disposal Facilities. Revised Draft Final Report for the
U.S. Environmental Protection Agency, Office of Solid
Waste, Washington, D.C.
1-53
-------
Freed, J.R., T. Chambers, W.N. Christie and C.E. Carpenter.
1983. Methods for Assessing Exposure to Chemical
Substances - Vol. 2: Methods for Assessing Exposure to
Chemical Substances in the Ambient Environment, EPA
560/15-83-015, U.S. Environmental Protection Agency, Office
of Pesticides and Other Toxic Substances, Washington, D.C.
Javandel, I., C. Doughty and C.F. Tsang. 1984. Groundwater
Transport: Handbook of Mathematical Models. American
Geophysical Union Water Resources Monograph, Washington,
D.C.
JRB Associates. 1982.
Disposal Sites,
Protection Agency,
Handbook - Remedial Actions at Waste
EPA-625/6-82-006, U.S. Environmental
Cincinnati, OH.
Mills, W., J. Dean, D. Porcella, S. Gherini, R. Hudson,
W. Frick, G. Rupp and G. Bowie. 1982. Water Quality
Assessment: A Screening Procedure for Toxic and
Conventional Pollutants, EPA 600/6-82-004abc, Vol. 1,2,
U.S. Environmental Protection Agency, Athens, GA.
Onishi, Y., G. Whelan and R.L. Skaggs. 1982. Development of a
Multimedia Radionuclide Exposure Assessment Methodology for
Low-Level Waste Management, PNL-3370, Pacific Northwest
Laboratory, Richland, WA.
SCS Engineers. 1982. Costs of Remedial Response Actions at
Uncontrolled Hazardous Waste Sites, EPA 600/2-82-035, U.S.
Environmental Protection Agency, Environmental Research
Laboratory, Cincinnati, OH.
Thibodeaux, L. 1981.
From Hazardous
Materials.
Estimating the Air Emissions of Chemical
Waste Landfills. Journal of Hazardous
1-54
-------
VOLUME 2
Simplified Methods
for Subsurface and
Waste Control Actions
-------
-------
VOLUME 2: SIMPLIFIED METHODS FOR SUBSURFACE
AND WASTE CONTROL ACTIONS
SECTION 1
INTRODUCTION
1.1 PURPOSE OF REPORT
During the 1950's, the development of analytical and
semi-analytical solutions for flow in ground-water systems
dominated the literature. Even though attention shifted to
numerical modeling in the 60's, progress continued in the
development of analytical methods (Walton, 1979). In
particular, a number of solutions for contaminant transport
were developed during this time.
Many of these solutions, and analysis methods involving the
use of these solutions, are applicable to the evaluation of
subsurface control (e.g., ground-water pumping and impermeable
barriers) and waste control (i.e., in-situ treatment) remedial
action technologies. The purpose of this volume is to provide
general guidance on the use of these "simplified methods."
More specifically, the volume seeks to:
1. Identify the specific simplified methods applicable
to the evaluation of each subsurface and waste
control action;
2. Identify key assumptions and limitations affecting
the use of specific methods;
3. Provide a compilation of methods that have been
programmed for use on hand-held calculators and
micro-computers; and
4. Demonstrate the use of selected methods through
example evaluations of different remedial actions.
It is commonly assumed that the methods discussed herein are
"easy to use" because they require limited data, manpower,
2-1
-------
time and computer resources. This JLs a dangerous assumption.
The proper application of these methods requires considerable
judgement and experience, at times as much as would be
required to use a more sophisticated numerical model.
1.2 REPORT ORGANIZATION
A summary and brief set of conclusions are provided in the
next section.
Section 3 identifies the specific simplified methods that are
applicable to the evaluation of different subsurface and waste
control remedial action technologies. This section also
discusses, in general, how each method can be used.
Section 4 discusses the basic theory underlying different
groups of available analytical and semi-analytical methods.
This section does not attempt to provide complete derivations
for different methods. A number of excellent textbooks and
other publications cover their derivation; where appropriate,
these publications are identified for those readers interested
in more background. The focus of Section 4 is on the
assumptions and limitations associated with different groups
of methods, and how they affect the usefulness of these
methods for remedial action evaluation.
Section 5 is a compilation of the methods that have been
programmed for use with either hand-held calculators or
micro-computers. Tables showing many of the available
programs and sources for the programs are provided.
Section 6 provides a series of example applications that serve
to demonstrate how different methods can be used to evaluate
selected remedial action alternatives. The example
applications are largely for hypothetical sites, some of which
have been patterned after existing uncontrolled hazardous
waste sites.
2-2
-------
SECTION 2
SUMMARY AND CONCLUSIONS
A large number of the existing analytical and semi-analytical
solutions for ground-water flow and transport, and associated
simplified methods, are applicable to the evaluation of
subsurface and waste control remedial actions. The limited
data and resource requirements (i.e., time, manpower and
computer facilities) associated with the use .of these methods
make them ideally suited to the screening of remedial action
performance and, in some cases, to the detailed analysis and
conceptual design of remedial actions.
A number of the more commonly used methods have been compiled
in several publications that would be of use to state and
Federal Superfund staff and site contractors. A relatively
complete set of well hydraulics solutions, including tables
and graphs of well functions, can be found in a handbook by
Walton (1984a). A large number of drain hydraulic solutions
have been compiled by Cohen and Miller (1983). Finally, van
Genuchten and Alves (1982), Javandel et al. (1984) and Walton
(1984a) have compiled a number of contaminant transport
solutions.
Hand-held calculator and micro-computer programs have been
written for a subset of the more commonly used methods. These
programs greatly reduce the amount of work involved in making
numerous repetitive calculations when using these methods.
They also eliminate the need for tables and graphs of well
functions, and expand the capabilities of some methods by
incorporating simple numerical techniques that would be
difficult to solve by hand. Some of the programs have been
published in the open literature, while others can be obtained
directly from their developers. The International Center for
Ground Water Modeling at Holcomb Research Institute, Butler
University, provides a clearinghouse service for available
hand-held calculator and micro-computer programs.
Despite benefits associated with these simplified methods,
there are a number of important limitations and key
assumptions that must be considered when using them in a
practical evaluation or remedial action performance. Many of
2-3
-------
the analytical and semi-analytical solutions for flow and
transport were derived for specific types of aquifers (e.g.,
confined, leaky or water table) with highly idealized
characteristics. Typically, the aquifers are assumed to be
horizontal, infinite in extent, constant in thickness, and
composed of homogeneous and isotropic properties. Since few,
if any, aquifers can fully satisfy these assumptions, even on
a local scale, some degree of simplification or correction is
often required. Transformation methods like equivalent
sections and incremental methods and corrections for
anisotropy are commonly used to construct aquifers with
hydraulically equivalent characteristics. The method of
images is commonly used to construct aquifers that are
bounded. In using the method of images, however, it is only
possible to construct aquifers with highly idealized
geometries like wedges, strips and rectangles.
Many of the solutions were also derived for highly idealized
ground-water flow patterns. Typically, solutions are
available for radial or uniform, one-dimensional (horizontal)
flow patterns. Fortunately, through the use of
superposition, these idealized flow patterns can be combined
so that more complex flow patterns can be evaluated. The
superposition of solutions has its limitations, however,
particularly for water table aquifers. Superposition in water
table aquifers is only appropriate when changes in water table
elevations are small compared to the saturated thickness.
The other major limitation is that many of the solutions were
derived for specific well or drain configurations. Typically,
wells or drains are assumed to be fully penetrating. This
assumption makes it possible to neglect vertical flow
components. When evaluating wells or drains that are not
fully penetrating, solutions derived specifically for the
configuration of interest or appropriate corrections should be
used. This also holds for wells with finite diameters and for
flowing wells.
These and other limitations preclude the complete, detailed
analysis of all remedial action design objectives and
configurations with the simplifed methods discussed herein.
Changes in water table elevations or piezometric heads
associated with the implementation of most subsurface and
waste control remedial actions can generally be evaluated.
The major exception includes certain drain and impermeable
barrier configurations, particularly near the ends of
partially penetrating drains or barriers of finite length.
The other exception is one side of a fully penetrating,
impermeable barrier when the method of images is used.
Changes in ground-water flow patterns can also be evaluated
for most remedial actions, especially those that involve wells
2-4
-------
or drains. The one major exception is for remedial actions
implemented in a water table aquifer. If the remedial action
produces large changes in head relative to the saturated
thickness, it may not be possible to evaluate changes in flow
patterns with these methods. Ground-water flow around the
ends of impermeable barriers of finite length is another major
exception. All of the available simplified methods require
that impermeable barriers are assumed to be infinite in
length, keyed-in at the ends, or completely surrounding.
Changes in contaminant transport cannot be fully evaluated for
many remedial actions. Most of the solutions were derived for
radial or one-dimensional flow patterns. Thus, their use is
largely limited to remedial actions that can be treated as
point sources or sinks (e.g., recovery wells and injection
wells). They were also derived based on the assumption that
the properties affecting contaminant retardation and
degradation are homogeneous and isotropic. Therefore, the
spatial changes in these properties produced by many of the
waste control actions (e.g., bioreclamation and chemical
injection) cannot be evaluated.
Analytical solutions for contaminant transport typically
consider all of the key processes of importance (i.e.,
advection, dispersion, retardation and degradation), whereas
semi-analytical solutions typically only consider advection
and, in some cases, retardation. Semi-analytical methods,
however, offer great flexibility in terms of the complexity of
flow patterns that can be analyzed.
Despite their apparent "ease of use," considerable judgement
and experience are required to evaluate remedial action
performance with simplifed methods. In applying these methods
it is important to recognize the tradeoffs that are being made
between the ease of application and the accuracy with which
these methods can simulate the effects of implementing
different remedial actions. The reader is referred to Volume
1 of this series for guidance on how to determine whether to
select simplified methods or more detailed, numerical models.
Volume 3 provides those who chose numerical models with
guidance on their use in remedial action evaluation.
2-5
-------
SECTION 3
REMEDIAL ACTION EVALUATION WITH SIMPLIFIED METHODS
3.1 OVERVIEW
There are a large number of remedial action technologies that
can be implemented at uncontrolled hazardous waste sites.
These actions can be classified as either surface, subsurface
or waste control technologies; control can either be by waste
removal, containment or treatment. Many of the available
technolgies are described in remedial action handbooks like
those by JRB Associates (1982) and SCS Engineers (1982). In
Volume 3, the large number of available technologies are
condensed into fourteen "remedial measures." Essentially,
technologies with similar design objectives were grouped
together as remedial measures. Table 3.1 shows the measures
that were classified as either surface, subsurface or waste
control measures.
The analytical and semi-analytical methods discussed in the
next section can be used to evaluate many of the remedial
measures shown in Table 3.1. Since these methods are
applicable only to flow and contaminant transport in ground-
water systems, only subsurface and waste control measures can
be evaluated. This section will discuss which of these
measures can be evaluated and which specific simplified
methods can be used.
In reading this section it is important to recognize that each
of the subsurface and waste control measures listed in Table
3.1 can have different configurations and design objectives.
Impermeable barriers, for instance, can be installed
upgradient, downgradient and completely around a site. They
can be partially penetrating (i.e., hanging) or fully
penetrating (i.e., keyed-in). They can be used to lower the
water table, divert uncontaminated ground water around a site,
or preclude further migration of contaminated ground water. A
complete detailed analysis of every configuration and design
objective is not possible, however, because of the assumptions
and limitations inherent in most simplified methods. Tables
3.2 and 3.3 list typical configurations and design objectives
2-6
-------
TABLE 3.1 GROUPING OF REMEDIAL MEASURES
I. SURFACE CONTROL
o Grading
o Revegetation
o Surface Water Diversion
II. SUBSURFACE CONTROL
o Capping and Top Liners
o Seepage Basins and Drains
o Subsurface Drains, Ditches and
Bottom Liners
o Impermeable Barriers
o Ground-Water Pumping
o Interceptor Trenches
III. WASTE CONTROL
o Permeable Treatment Beds
o Bioreclamation
o Chemical Injection
o Solution Mining (Extraction)
o Excavation/Hydraulic Dredging
2-7
-------
TABLE 3.2 APPLICABILITY OF SIMPLIFIED METHODS TO THE
EVALUATION OF SUBSURFACE CONTROL ACTIONS
Remedial Action
Design Objective
Applicable Simplified Method(s)
Comments
I
00
Capping and Top
Liners
Seepage Basins and
Ditches
Subsurface Drains,
Ditches and Bottom
Liners
Impermeable Barriers
Reduce Infiltration
Reduce ground-water
contamination
Recharge water and
modify flow patterns
Capture leachate
Reduce ground-water
contamination
Divert ground water
Capture contaminated
ground water
SI, GM, TM
CT, TM
SI, GM, TM (Basins)
SI, DH, S, TM (Ditches)
SI (Drains and Bottom Liners)
NA (Ditches)
CT, TM (Drains and Bottom Liners)
NA (Ditches)
S (Fully-penetrating barrier)
CM, TM (Partially-penetrating
barrier)
CT, S, TM (Fully-penetrating
barrier)
NA (Partially-penetrating
barrier)
Solutions for injection wells may
have to be used if mounding is
significant
Note limitations on superposition
in water table aquifers
Solutions for injection wells may have
to be used if mounding is significant
Barrier created with method of images
is assumed to be infinitely long or
keyed-in at ends; flow around ends of
barrier cannot be considered.
Solutions available only for several
idealized aquifer geometries and
barrier is assumed to be infinitely
long or keyed-in at the ends.
Barrier created with method of images
is assumed to be infinitely long or
keyed-in at ends; contaminant migration
around ends of barrier cannot be
considered.
(continued)
-------
TABLE 3.2 (continued)
Remedial Action
Ground-water Pumping
Interceptor Trenches
NJ
vo
Design Objective
Divert ground water
Capture contaminated
ground water
Divert ground water
Capture contaminated
ground water
Applicable Simplified Method(s)
WH, S, TM
CT, S, TM
DH, S, TM
CT, S, TM
Comments
Corrections may be necessary for
partially- penetrating wells; note
limitations on superposition in water
table aquifers
-
Note limitations on superposition in
water table aquifers
Drain will have to be represented as a
line of closely spaced wells; note
limitations on superposition in water
table aquifers
LEGEND: WH - Well Hydraulics
DH - Drain Hydraulics
GM - Groundwater Mounding
SI - Seepage/Infiltration
S - Superposition
TM - Transformation Methods
CM - Conformal Mapping
CT - Contaminant Transport
NA - No method applicable
-------
TABLE 3.3 APPLICABILITY OF SIMPLIFIED METHODS TO THE EVALUATION OF
WASTE CONTROL ACTIONS
Remedial Action
Design Objective
Applicable Simplified Method(s)
Comments
Permeable Treatment
Beds
Bioreclamation
Chemical Injection
Solution Mining
to
i_i Excavation/
o Hydraulic Dredging
In-situ treatment of
ground water
In-situ treatment of
ground water
In-situ treatment of
Mobilize contaminants
Improve leachate quality
CT, TM
WH, S, CT, TM
WH, S, CT, TM
WH, S, GM, CT, TM
SI, GM, CT, TM
Contaminant transport solution must be
applied in a step-wise fashion to
treatment bed and up gradient and down
gradient portions of aquifer.
Reductions in contaminant concentrations
cannot be analyzed since applicable
solutions typically neglect degradation.
Reductions in contaminant concentrations
cannot be analyzed since applicable
solutions typically neglect degradation.
Selected solution must consider retardation
Contaminant transport solutions for
injection wells may have to be used if
mounding is significant.
LEGEND: GM - Ground-Water Mounding
SI - Seepage/Infiltration
TM - Transformation Methods
CT - Contaminant Transport
WH - Well Hydraulics
S - Superposition
-------
for each measure, as well as the applicable simplified
method(s). Important limitations and considerations
associated with the use of different simplified methods are
also listed.
3.2 SUBSURFACE CONTROL MEASURES
The primary goals of subsurface control measures are to
prevent leachate migration and reduce ground-water
contamination by diversion, containment or plume capture.
Subsurface control measures include capping and top liners;
seepage basins and ditches; subsurface drains, ditches and
bottom liners; impermeable barriers; ground-water pumping;
and interceptor trenches.
3.2.1 Capping and Top Liners
As Table 3.2 shows, caps and top liners are generally
implemented to reduce infiltration into a waste site, thereby
reducing the quantity of leachate that is generated. In
evaluating the performance of capping and top liner systems,
two design objectives are of concern: 1) the reduction in the
quantity of leachate that is generated and 2) the associated
reduction in ground-water contamination. Different methods
are required to evaluate each objective.
Methods applicable to the estimation of seepage rates for
landfills (see Subsection 4.5) can be used to evaluate the
effect of a cap or top liner on leachate generation.
Infiltration rates can be determined for both pre- and
post-restoration conditions. The ground-water mounding
estimation techniques discussed in Subsection 4.4 can also be
used to determine whether the reduction in leachate quantity
will have any effect on the degree of mounding, if any, below
the site.
Associated reductions' in ground-water contamination can be
evaluated with simplified methods for contaminant transport
(see Subsection 4.9). The choice of which type of contaminant
transport method to use will depend upon whether or not the
quantity of leachate generated by the site is sufficient to
cause mounding. If mounding is not significant either before
or after capping, almost any of the analytical or
semi-analytical solutions can be used.
If mounding is significant even after capping, only those
analyical or semi-analytical methods that consider radial flow
can be used. Analytical methods of this type are for
2-11
-------
injection wells; none exist for ground-water mounds.
Therefore, analytical methods can only be used if a flow
pattern equivalent to that of a mound can be simulated with
one or more injection wells. The same limitation holds for
those semi-analytical methods based on the complex velocity
potential concept (see Subsection 4.9). Except, these methods
often also provide a way of representing a circular source of
finite radius. In some cases, a flow pattern equivalent to
that around a mound can be represented with such a source.
The semi-analytical solution based on a simple numerical
technique discussed in Subsection 4.9 can also be used.
Again, an injection well or group of wells must be used to
create a flow pattern equivalent to that created by the mound.
Example Application 4 in Section 6 demonstrates the use of the
simple numerical technique to evaluate contaminant transport
from a site where mounding is significant.
As Table 3.2 shows, transformation methods are also useful in
evaluating capping and top liner actions. Transformation
methods are used to transform real world aquifers with
heterogeneous and isotropic conditions into equivalent,
idealized aquifers with homogeneous and isotropic conditions
(see Subsection 4.7). These methods are used in conjunction
with almost all analytical and semi-analytical solutions for
flow and contaminant transport. Thus, transformation methods
are useful in the evaluation of virtually all subsurface and
waste control actions.
3.2.2 Seepage Basins and Ditches
The primary objective for using seepage basins and ditches is
to recharge site runoff or water withdrawn by wells or drains.
A second objective is to improve the efficiency of plume
capture by modifying ground-water flow patterns. Thus, both
seepage (i.e., recharge) rates and the extent of ground-water
mounding are important when evaluating seepage basins and
ditches.
Subsections 4.5 and 4.4 discuss methods for estimating seepage
rates for ponded facilities and changes in water table
elevations (i.e., mounding), respectively. If mounding
occurs, its effect on the drawdowns at nearby wells or drains
can be evaluated by using the principle of superposition. The
use of superposition makes it possible to evaluate a number
of alternative locations for a seepage basin. It is important
to remember, however, the limitations associated with using
superposition in water table aquifers (see Subsection 4.3).
The effect of seepage from ditches can be evaluated in the
2-12
-------
same manner. Instead of using ground-water mounding
estimation methods, however, the drain hydraulics methods
discussed in Subsection 4.3 can be used. In using these
methods, the ditch is treated as a line source of finite
length.
If changes in contaminant migration patterns, as a result of
recharge from a basin or ditch, are also of interest, the
choice of which contaminant transport method to use will again
depend on the extent of mounding and whether one or more
injection wells or a circular source of finite radius can be
used to represent the mound created by the seepage basin or
ditch. Example Application 4 in Section 6 shows one approach
for analyzing a recharge basin.
3.2.3 Subsurface Drains, Ditches and Bottom Liners
Subsurface drains, ditches and bottom liners are usually
installed in the unsaturated zone to capture leachate before
it reaches the saturated zone. The infiltration estimation
techniques discussed in Subsection 4.5, in particular the
HELP model (Schroeder et al., 1984a and 1984b), could be used
to estimate reductions in leachate quantity associated with
subsurface drains or bottom lining. Given this change in
leachate quantity, changes in ground-water contamination
levels could be assessed with the contaminant transport
methods in Subsection 4.9. The same considerations as for
capping and top lining (see Subsection 3.2.1) would apply to
the selection of what type of method to use. None of the
simplified methods are applicable to the evaluation of
ditches.
3.2.4 Impermeable Barriers
Impermeable barriers are grout curtains, slurry walls and
sheet piling installed in the saturated zone to divert
uncontaminated ground-water around a site or limit the
migration of contaminated ground-water. Barriers can be
placed in a number of locations relative to a disposal site:
upgradient, downgradient or completely around. Barriers
designed to divert ground-water by lowering water levels can
either partially or fully penetrate the saturated zone. The
former must be keyed into impermeable strata to preclude water
movement around the ends. Barriers designed to contain ground
water are normally fully penetrating.
The analytical methods discussed in Section 4 are only
applicable to a few of the many possible barrier
2-13
-------
configurations. The amount of seepage likely to occur under a
partially penetrating barrier can be analyzed using the
conformal mapping methods described in Subsection 4.8. These
methods are for two-dimensional flow in the horizontal and
vertical directions only. Thus, it is assumed that the
barrier is infinitely long, keyed-in at the ends, or
completely surrounds the site. If the barrier does not have
one of these configurations, the conformal mapping methods
will only apply over those portions of the barrier where
horizontal flow components are essentially perpendicular to
the barrier.
The conformal mapping methods are limited to either
single-layered or two-layered saturated systems; in the latter
case, the layers must be of equal thickness. Therefore, it is
important to carefully consider site conditions before using
these methods.
The method for two-layered systems can also be used to
evaluate barriers that fully penetrate the saturated zone, but
are keyed into a leaky bedrock layer. Again the same
restrictions apply in terms of the barrier configurations and
site conditions that can be considered.
In cases where the barrier can be keyed into an impermeable
bedrock layer, the principle of superposition can be used.
Specifically, the method of images can be used to obtain an
impermeable boundary of infinite length by using real and
imaginary pumping wells (see Subsection 4.6). Different
barrier configurations, including upgradient, downgradient and
completely surrounding, can be analyzed through the proper use
of real and image wells. An impermeable barrier surrounding a
site is analyzed with the method of images in Example
Application 3 in Section 6.
Despite its flexibility, the method of images has two distinct
disadvantages. First, it requires that the barrier be assumed
to be infinite in length, keyed-in at the ends, or completely
surrounding the site. Flow conditions and heads for barriers
with other configurations cannot be considered except near the
center of relatively long barriers where flow directions are
essentially perpendicular to the barrier.
The second disadvantage is that only ground-water flow
patterns and heads on the side of the barrier with the real
wells (i.e., the real region) can be analyzed. The other side
(i.e., the image region) is of no value. Thus, the real
well(s) must be located on the same side of the barrier as the
disposal site if flow patterns and heads around the site are
of concern.
2-14
-------
3.2.5 Ground-Water Pumping
Ground-water pumping actions can have a number of
configurations and design objectives. Single pumping wells or
a line of well points can be used to capture a plume. Single
or multiple wells can be installed to divert ground water by
lowering the water table. They can also be used to prevent
unconfined aquifers from contaminating lower aquifers
separated by leaky formations. The water withdrawn by pumping
may be treated and subsequently reinjected through one or more
wells. The reinjection wells may be used to flush
contaminants toward the pumping wells or to create a hydraulic
barrier to preclude further plume migration.
The well hydraulics, superposition, transformation and
contaminant transport methods discussed in Subsections 4.2,
4.6, 4.7 and 4.9, respectively offer a relatively complete set
of methods for evaluating virtually all possible
configurations for ground-water pumping remedial actions.
They can be used to evaluate changes in ground-water flow
patterns, heads and contaminant movement. All of these are
important factors when evaluating ground-water pumping
schemes. In using the available analytical methods it is
important to recognize the key underlying assumptions and
limitations (see Table 3.2). Since pumping is often used
conjunctively with impermeable barriers it is also important
to recognize the limitations associated with the method of
images. Example Applications 2,3 and 4 in Section 6
demonstrate approaches for evaluating ground-water pumping
actions.
3.2.6 Interceptor Trenches
Interceptor trenches are drain systems that are installed in
the saturated zone. They can be used to: 1) divert ground
water by lowering the water table or 2) capture a plume.
The first design objective can be evaluated using the drain
hydraulics, superposition, and transformation methods
discussed in Subsections 4.3, 4.6 and 4.7, respectively. A
wide range of site conditions and drain configurations can be
considered with these methods. Example Applications 1 and 5
in Section 6 demonstrate an approach for evaluating how water
table elevations will change following the installation of a
drain.
The second objective can be evaluated using the contaminant
transport methods discussed in Subsection 4.9. Since none of
these methods explicitly consider drain systems, a drain must
2-15
-------
be represented by a line of closely spaced wells.
3.3 WASTE CONTROL
Waste control measures involve the removal or treatment of
hazardous wastes or contaminated water and sediments. Removal
can be accomplished through excavation or hydraulic dredging.
Treatment methods include permeable treatment beds,
bioreclamation, chemical injection, and solution mining
(extraction). Those treatment methods in the waste control
category are in-situ methods. That is, treatment is
accomplished in-place. On-site treatment methods like carbon
adsorption, precipitation, sedimentation, and activated sludge
are considered under the subsurface control category since
they are typically used in conjunction with ground-water
pumping systems, subsurface drains or interceptor drains.
3.3.1 Permeable Treatment Beds
Permeable treatment beds are trenches backfilled with
limestone activated carbon or another media that can
physically or chemically remove contaminants from ground
water. They are installed so as to penetrate into the
saturated zone, and are normally used in areas where the water
table is near the ground surface. Treatment occurs as
contaminated ground water passes through the bed. Permeable
treatment beds are typically designed to have the same
hydraulic conductivity as the surrounding materials. As a
result, their installation generally has little or no affect
on ground-water movement.
In evaluating the effectiveness of permeable treatment beds in
terms of reducing ground-water contamination it is important
to recognize that all of the simplified methods for
contaminant transport assume that aquifer properties are
homogeneous and isotropic. Thus, it is difficult to represent
the discontinuity produced by a treatment bed because it has
sorption properties different from those of the surrounding
aquifer materials. To analyze a treatment bed, the simplified
methods must be applied in a step-wise fashion, first to the
upgradient portion of the aquifer, then the treament bed
itself and then the downgradient portion of the aquifer.
3.3.2 Bioreclamation
Bioreclamation is an in-situ treatment method involving the
2-16
-------
injection of microbes, nutrients and oxygen into a plume to
initiate or accelerate contaminant degradation. It is
commonly used for hydrocarbons and other easily biodegradable
pollutants. Injection is accomplished through the use of one
or more wells. Pumping is also used to obtain water for the
injection system and to enhance treatment. Recirculation
between the injection and withdrawal wells is often an
important design consideration.
Since bioreclamation is essentially a form of a ground-water
pumping technique, there are a number of simplified methods
available to examine ground-water flow patterns, changes in
hydraulic heads, and pollutant movement between injection and
recovery wells. Despite the availability of a large number of
methods, they can only be used to evaluate a few of the design
objectives affecting the performance of bioreclamation
systems. The well hydraulics, superposition, and
transformation methods can be used to evaluate changes in flow
patterns and heads induced by the wells. Contaminant
transport methods can be used to estimate the size of a region
that will be treated by the injected mixture, the amount of
recirculation that might occur, and the time it will take for
the injected mixture to arrive at a recovery well. Reductions
in contaminant concentrations cannot be directly estimated,
however, since the applicable solutions typically neglect
degradation. Example Application 4 in Section 6 demonstrates
one approach for evaluating a bioreclamation action.
3.3.3 Chemical Injection
Chemical injection is used to treat the waste in a
landfill/lagoon or in a contaminated saturated zone. It is
usually applied to sites with well defined wastes, with
shallow landfill or lagoon depths, and where the vertical and
horizontal extent of the contamination is small (JRB
Associates, 1982). The objective of chemical injection is to
immobilize or destroy a pollutant. Numerous injection wells
may be employed depending on the size of the disposal site. A
water supply well is usually required for chemical dilution.
The effect of this measure is to substantially increase
retardation and degradation processes in either the
unsaturated or saturated zones.
As with bioreclamation, only a few of the design objectives
for chemical injection can be evaluated with available
simplified methods. Changes in flow patterns and hydraulic
heads can be evaluated with the well hydraulics,
superposition, and transformation methods. The size of the
zone treated by the injected fluid can be evaluated with
contaminant transport methods. The extent to which chemical
2-17
-------
injection will reduce ground-water contamination levels at
some point downgradient from a site cannot be assessed,
however. Again, this is due to the the fact that degradation
is often neglected in applicable transport solutions.
3.3.4 Solution Mining (Extraction)
Solution mining is similar to chemical injection in that both
methods chemically alter the pollutant in the waste itself.
However, solution mining involves the flooding of a landfill
with a chemical solvent, which may desorb or free the
pollutant so that it may be mobilized in a larger leachate
flow (JRB Associates, 1982). The leachate can then be
collected by drains and/or well points. The objective is to
increase the mobility of the contaminant.
The evaluation of the performance of a solution mining action
can be approached with the same types of methods used for
bioreclamation and chemical injection. Since an important
design objective for solution mining is the efficiency of
recovery, several of the analytical transport methods can be
used to identify which well configuration will provide for the
most efficient recovery. The effect of the solvent on
increasing contaminant mobility can be considered by simply
adjusting the retardation factor used in these methods.
Semi-analytical transport methods can be used in a similar
way.
3.3.5 Excavation/Hydraulic Dredging
Excavation/hydraulic dredging involves the removal of the
waste source itself, thus improving leachate quality.
Excavation is used on solids, sediments, or sludge materials.
Hydraulic dredging may be used to remove liquids and/or
sludges from lagoons or surface impoundments. After the waste
area has been excavated or dredged, it may be backfilled to
limit infiltration.
The effectiveness of excavation/hydraulic dredging can be
evaluated with a number of simplified methods. Seepage/
infiltration estimation methods can be used to evaluate
changes in leachate quantity. In the case of waste
excavation, those methods applicable to the estimation of
infiltration rates for landfills (e.g., HELP model) can be
used to determine whether the amount of water passing through
the site will change. The amount of change will, in part,
depend upon the properties of the materials used for backfill
relative to those of the excavated waste.
2-18
-------
Ground-water mounding estimation methods can be used to
determine whether the shape of the water table will change as
a result of changes in infiltration/seepage rates.
The type of contaminant transport method used to evaluate
changes in ground-water contamination levels will again be
determined by the extent of mounding. If mounding is
significant, a transport method that can consider radial flow
must be used. One or more injection wells or a circular
source of finite radius will need to be used to simulate the
effects of the mound. If mounding is not important, any of
the transport methods can be used.
2-19
-------
SECTION 4
THEORY UNDERLYING AVAILABLE SIMPLIFED METHODS
4.1 OVERVIEW
The basic theory underlying the simplified methods applicable
to remedial action evaluation can be divided into the
following areas: 1) well hydraulics, 2) drain hydraulics,
3) ground-water mounding, 4) seepage/infiltration,
5) superposition, 6) transformation methods, 7) conformal
mapping, and 8) contaminant transport. Some of these areas
encompass the theory used to develop different types of
solutions (e.g., well hydraulics, drain hydraulics,
ground-water mounding, conformal mapping and contaminant
transport), whereas others encompass the theory behind the use
of these solutions to evaluate relatively complex
geohydrological conditions (e.g., superposition and
transformation methods).
The applicable theory underlying each area will be summarized
in this section. Comprehensive discussions of the applicable
theories and derivations of analytical expressions will not be
provided, since this material is presented in a number of
standard references (e.g., Freeze and Cherry, 1979; Bear,
1979; Walton, 1970 and Harr, 1962) and handbooks (e.g. Walton,
1984a and Javandel et al., 1983). Rather, this section will
focus on the types of methods available in each area, and the
key assumptions and limitations governing their use.
4.2 WELL HYDRAULICS
Wells are used in many different types of remedial action
technologies. They can be used alone to control plume
movement, divert uncontaminated ground water or capture
contaminated ground water. They can also be used in
conjunction with other technologies for the same purposes, or
as part of in-situ treatment technologies where both injection
and extraction are required. As a result, well hydraulics
analyses are likely to be conducted at many sites.
2-20
-------
Fortunately, numerous solutions have been developed to
calculate the change in piezometric head or water table
elevation resulting from the introduction of a well. Some of
the earliest and perhaps most fundamental work in the area of
well hydraulics was conducted by Theis (1935). As Freeze and
Cherry (1979) note, Theis utilized an analogy to heat-transfer
theory to derive an analytical solution for flow to a well in
a highly simplified aquifer. This aquifer has the following
characteristics:
1. horizontal
2. confined between impermeable layers on the
top and bottom
3. infinite in horizontal extent
4. constant thickness
5. homogeneous and isotropic
Transient flow in this type of aquifer system with no sources
or sinks can be described by the following partial
differential equation:
5v2
s ah
T at
(4.1)
where
h = piezometric head, L
x,y = horizontal directions, L
S = storativity, dimensionless
T = transmissivity, L^/T
t = time, T
Recognizing that changes in head around a well are
radially-symmetric, Equation 4.1 can be rewritten in the
following form
a2h
ar2
§.
T at
(4.2)
where
r = radial distance from the well, L
4.2.1 Confined Aquifers
The work by Theis (1935) produced a solution to Equation 4.2
for the condition of a single, fully penetrating well with a
constant withdrawal (pumping) rate, an infinitesimally small
well diameter, and a uniform piezometric head prior to the
2-21
-------
initiation of pumping. Figure 4.1 shows the drawdown around a
well with this type of configuration in a horizontal confined
aquifer. Under these conditions, flow is strictly horizontal
and unidirectional toward the well. There are no vertical
flow components. Using a uniform piezometric head as an
initial condition, the assumption of no drawdown at infinity
and a constant pumping rate as boundary conditions, Theis
obtained the following solution for transient flow to a well:
u
where
u = r2S/4Tt, L
ho = initial piezometric head, L
Q = pumping rate, L.3/T
s = drawdown, L
As Freeze and Cherry (1979) note, the integral in Equation 4.3
is known as the exponential integral. Given the above
definition for u, the integral is also known as the well
function, W(u) . This gives the familiar Theis equation
s = -£- W(u) (4.4)
47TT
Values for W(u) can be evaluated using a series expansion as
noted by Bear (1979). Tabulated values for W(u) and a
graphical relationship between W(u) and 1/u are provided in
most ground-water textbooks .
In using the Theis equation it is important to recognize that
steady-state conditions can never be reached in an aquifer of
infinite extent. In the absence of any sources of recharge,
water must be continuously withdrawn from storage to meet the
demands of the pumping well. This requires that the cone of
depression must continually expand radially outward from the
well. From a practical point of view, however, peizometric
heads do reach a quasi-steady-state as the rate of propagation
of the cone of depression decreases. This is particularly
true for the region near the well. Thus, the Theis equation
can be used to obtain an estimate of steady-state conditions
when the time of pumping is assumed to be long.
Many of the hand calculator and micro-computer programs
discussed in Section 5 are for what is known as "Theis
condition aquifers." Theis condition aquifers are essentially
2-22
-------
o
E
OJ
C
C
O
u
it!
C
•H
(1)
C
•H
c
3
O
O Q)
13 IH
? -H
(0 3
>-l D1
Q fd
•H
fc.
2-23
-------
those that have the aquifer characteristics and well
conditions discussed above. Clearly few aquifers in the real
word, even with reasonably acceptable simplifications, can be
considered as Theis condition aquifers. Many are not
confined. Some have semi-impermeable confining layers on the
top and/or bottom. These are called "leaky" aquifers. Others
have a free surface on top and either an impermeable or leaky
layer on the bottom. These aquifers are known as "water
table" aquifers.
4.2.2 Leaky Aquifers
An analytical solution for leaky aquifer conditions was
initially developed by Hantush and Jacob (1955) and Hantush
(1956, 1960), and was later expanded by Neuman and Witherspoon
(1969a, 1969b, 1972). The expression derived by Hantush and
Jacob has a form similar to the Theis equation
s = __ W(u,r/B) '
47TT
where W(u,r/B) is called the leaky well function and r/B is a
dimensionless parameter given by
^ . [YS .!~YS (4.6)
r/B = rV - = rV -
Kbb1 Tb1
where K = hydraulic conductivity of aquifer, L/T
b = aquifer thickness, L
K1 = hydraulic conductivity of leaky layer, L/T
b' = thickness of leaky layer, L
The assumptions and limitations underlying Equation 4.5 are
essentially the same as those for the Theis equation. The
aquifer is horizontal, infinite in extent, has a constant
thickness, is homogeneous and isotropic, and has a uniform
piezometric head prior to pumping. The well is fully
penetrating with a constant pumping rate and infinitesimal
diameter. Figure 4.2 shows the drawdown around such a well in
a horizontal leaky aquifer. The unpumped aquifer above the
leaky aquifer is often called the "supplying aquifer." The
two aquifers are separated by the leaky layer.
In addition, Hantush and Jacob had to assume that: 1) the
hydraulic head in the supplying aquifer remains constant;
2) the rate of leakage across the leaky layer is proportional
to the difference in hydraulic heads between the pumped and
2-24
-------
t =
PQTENTIQMETRIC
SURFACE
NJ
I
SJ
(Jl
>r
Figure 4.2 Drawdown around a pumping well in a leaky
aquifer.
-------
unpumped aquifers; and 3) flow in the pumped aquifer is
strictly horizontal and unidirectional towards the well, while
flow in the leaky layer is vertical. As Freeze and Cherry
(1979) point out, the first assumption implies that the
unpumped aquifer can provide an unlimited supply of water to
the pumped aquifer. The second assumption neglects the effect
of storage in the leaky layer on delaying the delivery of
water. As a result, the rate of actual drawdown may be
over-predicted. The third assumption neglects the potential
for vertical flow components in the pumped aquifer and
horizontal components in the leaky layer. Figure 4.3 shows
how the assumed flow distribution (on the left side of the
well) is different from the actual distribution (on the right
side of the well). Huisman (1972) notes that less than 3%
error will be induced if vertical flow components are
neglected when A>3H, where X=1/B. Neuman and Witherspoon
(1969a) note that when the hydraulic conductivity of the
aquifer is at least two orders of magnitude greater than the
leaky layer, neglecting vertical flow components introduces
errors of no more than 5%.
Neuman and Witherspoon (1969a, 1969b) overcame the limitations
imposed by the first two assumptions by generating a more
rigorous leaky well function. Their expression takes the form
of:
47TT
W(u, r/B,., r/B .,
(4.7)
where
(4.8)
K,b and Ss are the hydraulic conductivity, thickness and
specific storage, respectively. The subscripts 1 and 2 are
for the pumped and unpumped aquifers, respectively. The prime
(') is for the leaky layer.
2-26
-------
t= 0
r
'/;////.///////./////./////////////////////////////
0RAWDOWN7-
UEAKY ;;;; AQUIFER
POTENfiO METRIC
SURFACE
Figure 4.3
Assumed (left side of well) and actual (right
side of well) flow patterns for a fully
penetrating well in a leaky aquifer (adapted
from Huisman, 1972).
-------
Values for the well function in Equations 4.6 and 4.7 have
been tabulated in many ground-water textbooks and the
publications referenced above.
Unlike confined aquifers, water levels in leaky aquifers of
infinite extent can achieve a steady-state condition. This
occurs once the entire discharge of the well is derived from
leakage.
Freeze and Cherry (1979) note that the simpler solution
(Equation 4.5) is widely used despite its limitations. Once
steady-state conditions are reached, the limitation imposed by
neglecting the storage effect in the leaky layer is removed.
The limitation imposed by assuming an unlimited supply of
water is not removed, however. Therefore, it is important to
examine the system of interest to determine whether it is
reasonable to simply rely on Equation 4.6.
Freeze and Cherry (1979) also note that the Theis equation can
be used in place of one of the leaky aquifer solutions,
because it provides a more conservative estimate of drawdowns;
drawdowns under leaky conditions will be less than those for
confined conditions because of the leakage. While this may be
appropriate for the analysis of wells for water supply
purposes, this line of reasoning is not appropriate for
remedial action evaluation. Since one of the intents of
ground-water pumping as a remedial action is to lower the
water table, a conservative estimate of drawdown may lead to
the design of an ineffective remedial action.
4^.2.3 Water Table Aquifers
In both confined and leaky aquifers with a fully penetrating
well it is generally reasonable to assume that flow is
strictly horizontal and unidirectional towards the well. In
water table aquifers this assumption may or may not be
reasonable. When a water table aquifer is pumped, vertical
flow components are created as the water table itself changes
shape around the well. Whereas in a confined aquifer, water
is produced by both the compaction of the aquifer and the
expansion of water, in a water table aquifer it is also
produced by gravity drainage. As Walton (1970) notes, the
release of water due to compaction of the aquifer and
expansion of water is instantaneous. The release due to
gravity drainage is not. As a result, .changes in hydraulic
head are initially rapid. The rate of change slows, however,
as the water released by gravity drainage reaches the cone of
depression. Once this occurs, the rate of change increases
and the cone of depression continues to expand as gravity
2-28
-------
drainage keeps pace with declining water levels. As with
confined aquifers, in the absence of any source of recharge,
the water table will continue to change as long as pumping
continues. Again, however, a quasi-steady condition will be
reached after a reasonably long time.
A number of analytical solutions have been developed for fully
penetrating wells in water table aquifers; Figure 4.4 shows
drawdown around a well in a horizontal water table aquifer.
The most general solution is the solution developed by Boulton
(1954, 1955, 1963) and later advanced by Neuman (1972, 1973,
1975). Their work produced an solution that takes the form of
S =—2- W(u_, u^Tj) (4.9)
47TT A ti
where W (UA,UB,T}) is known as the "unconfined well function."
Just following initiation of pumping, the unconfined well
function is W (u^,^), where
UA
4Tt
2 (4.10)
and S is now the elastic storativity of the aquifer. This
elastic storativity is responsible for the instantaneous
release of water. At some later time, the unconfined well
function is W (u „,*]), where,
JO
(4.11)
Sy is the specific yield responsible for delaying the release
or water. Details regarding the use of Equation 4.9 and
graphs and tables providing values of the unconfined well
function are given in Neuman (1975) as well as many
ground-water textbooks.
The expression developed by Neuman considers the effects of
both the delay caused by gravity drainage and vertical flow
components. The delay effect has its greatest impact during
the early stages of pumping. Bear (1979) states that the
2-29
-------
i
U)
o
SURFACE
WATER TABLE
::;v::::r;;:::;iAQUiFER;" "
///////////////v//////////////////////////////////////////////////
Figure 4.4 Drawdown around a pumping well in a water
table aquifer.
-------
specific yield increases at a diminishing rate with time.
Therefore, if water levels following long-term pumping are of
interest, the impact of the delay effect can be neglected.
Bear (1979) states further, however, that lenses of silt and
clay can produce significant delays.
Vertical flow components can affect water levels under certain
water table conditions. Bear (1979) points out that Boulton
(1954) and Hantush (1964) both suggested that vertical flow
components are of importance in the region of 0 ° e (4.12)
K
where H = hydraulic head at the well, L
HQ = initial hydraulic head, L
ne = effective porosity, dimensionless
Stallman (1965) found vertical flow components to be important
in the region
(4.13)
The importance of vertical flow components and the delay
caused by drainage should be examined given specific site
conditions, because the other general solutions for water
table aquifers assume both can be neglected.
In cases where they can be neglected, an expression developed
by Boulton (1954) in one of his earlier works can be used
(4.14)
where Cf = correction factor
V(p,T) = gravity well function for water table
aquifers
P = r/H0
T = Kt/ne H0
2,
T = KH0 = transmissivity, L /T
2-31
(1 + cf) V(p,T )
-------
Values of V( p , 1 ) and Cf are provided in a number of
publications, including Boulton (1954), Schoeller (1959) and
Hantush (1964). Bear (1979) presents several ways of
approximating V(P,T) for different ranges of T, and notes that
an error of less than 6% is obtained if Cf is assumed to be
zero for 0.055, V(/>,T) is approximately
equal to 1/2 W(u), where u = n^ /4Tt (Bear 1979). Combining
this result with Equation 4.14 shows that the analytical
solution for confined aquifers can, in some cases, be used to
estimate water levels in a water table aquifer, particularly
for long time frames where conditions approach a quasi-steady
state. Bear" (1979) states that for a thick aquifer with small
drawdowns that satisfy the following condition, a water table
aquifer can be treated as a confined aquifer:
(H0-h)«HQ or H0+h =s= 2HO (4.15)
4.2.4 Available Well Hydraulics Solutions
The previous discussions in this section have overviewed the
theory underlying some of the general analytical solutions
available for flow in confined, leaky and water table
aquifers. The basic assumptions upon which these solutions
are derived limits their use in a number of situations.
Fortunately, a number of other analytical solutions with less
restrictive assumptions have been developed. Table 4.1 was
adapted from tables in Walton (1984a). It provides a
reasonably complete inventory of available analytical
solutions for confined, leaky and water table aquifer systems,
respectively. Each solution in Table 4.1 is characterized in
terms of: 1) the aquifer characteristics and well
configuration that can be analyzed, 2) whether the solution is
for time-varying or steady-state conditions, and 3) the type
of output that can be obtained.
As this table shows, solutions are available for isotropic and
anisotropic conditions. Corrections for anisotropy, in
addition to those provided in these expressions, will be
discussed in Subsection 4.6, Transformation Methods.
A range of possible well configurations can also be
considered, including wells with a finite diameter, wells with
storage capacity, flowing wells, and partially penetrating
wells. During the early periods of pumping, drawdowns for
wells of finite diameter and/or with storage capacity will
deviate from those predicted using an expression for a well of
infinitesimal diameter and no storage. Thus, if piezometric
2-32
-------
TABLE 4.1 INVENTORY OF SELECTED WELL HYDRAULICS SOLUTIONS
(adapted from Walton, 1984a)
Type
C
C
C
KJ C
1
U) r
UJ U
C
C
C
C
C
I
L
L
L
L
L
L
Aquifer Cha
Properties
H.I
H,I
H,I
H.I
H,A
H.I
H.A
H.A
H,I
H,A
H,I
H.I
H.I
H.I
H.I
H,A
H.I
racteris
Extent
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
tics
Special
Cases
P
P
P
P.VE
P
P
F
r
P
P
P
P
P
P. AS
P
P
P
Well
Penetration
FP
FP
FP
FP
PP
FP
FP
PC
FP
FP
FP
FP
FP
PP
FP
PP
FP
Configura
Storage
NS
NS
S
NS
NS
NS
NS
NS
NS
NS
NS
NS
S
NS
NS
NS
NS
tion
Diameter
ID
FD
FO
ID
ID
ID
ID
ID
ID
ID
ID
ID
FD
ID
ID
ID
ID
Time
Frame
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
Output
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
0
References
Theis (1935)
Hantush (1964)
Papadopulos (1967)
Brutsaert and
Corapcioglu (1976)
Hantush (1964)
Jacob and
Lohmann (1952)
Boulton and
Streltsova (19776)
Boulton and
Streltsova (1977a)
Moench and
Prickett (1972)
Papadopulos (1965)
Hantush and
Jacob (1955)
Hantush (1959)
Lai and Chen Uu Su
(1979)
Witherspoon et al .
(1967)
Corapcioglu (1976)
Hantush (1964)
Hantush (1967b)
(continued)
-------
TABLE 4.1 (continued)
to
I
U)
Type
WT
WT
WT
WT
WT
Aquifer Cha
Properties
H,I
H,A
H,A
H,A
H,A
racteris
Extent
IN
IN
IN
IN
IN
tics
Special
Cases
P
P
P
F
P,LB
Well (
Penetration
FP
PP
PP
PC
PP
^onfigura
Storage
NS
NS
S
NS
NS
tion
Diameter
ID
ID
FD
ID
ID
Time
Frame
TV
TV
TV
TV
TV
Output
D
D
D
D
D
References
Neuman (1975)
Streltsova (1974)
Boulton and
Streltsova (1978)
Boulton and
Streltsova (1978)
Streltsova (1976)
(1) Additional selected solutions not listed in this table include:
- Boulton (1954a) solution for non steady-state water table drawdown.
- Cooper, H. H., Jr., and C. E. Jacob (1946) give a straight line graphical solution to the Theis equation.
- Thiem (1906) gives a steady state solution for flow to a well in a confined aquifer.
LEGEND: C - Confined FP -
WT - Water Table PP -
L - Leaky PC -
H - Homogeneous NS -
I - Isotropic S -
A - Anlsotropic ID -
IN - Infinite FD -
AS - Aqultard Storage
IB - Leaky Base
F - Fractured Media
P " Uniformly Porous
VE - Visco-Elastic Properties
Fully Penetrating
Partially Penetrating
Partially Cased
No Storage
Storage
Infinitesimal Diameter
Finite Diameter
TV - Time Varying
D - Drawdown
-------
heads just after initiation of pumping are of concern, one of
the appropriate solutions in Table 4.1 should be used.
Flowing wells are wells where the head in the well is held
constant and flow rates are allowed to vary with time. If
flowing wells are being considered, appropriate relationships
must be used since the ones discussed earlier are for wells
with constant flow rates and varying heads.
Partially penetrating wells are those that are screened only
over a portion of the aquifer. Partial penetration creates
vertical flow components that may preclude the use of
expressions based on the assumption of complete penetration.
As Bear (1979) states, the drawdown produced by a partially
penetrating well is greater than that for a fully penetrating
well. This difference is only significant for a distance 1.5
to 2.0 times the saturated thickness away from the well. As
Table 4.1 shows, a number of analytical solutions have been
developed for partially penetrating wells. Corrections to the
drawdowns predicted with solutions for fully penetrating wells
are also available in Bear (1979) for several different well
configurations.
All of the analytical solutions shown in Table 4.1 are
included in a handbook by Walton (1984a). It contains the
actual expressions and supporting tables and graphs useful in
estimating values for different well functions. It also
discusses a number of other useful analytical solutions and
some of the available hand-calculator programs. The handbook
is a useful source for anyone planning to use analytical
solutions for the evaluation of remedial action performance.
It can be obtained from the International Ground Water
Modeling Center (IGWMC), Holcomb Research Institute, Butler
University in Indianapolis, Indiana (317-283-9458).
4.3 DRAIN HYDRAULICS
Drains are collection systems of finite length that can be
used, like wells, to control plume movement, divert
ground-water flow and depress water table levels. They can
have a number of configurations ranging from fully
penetrating, vertical trenches to partially penetrating
ditches, to perforated pipes. Unlike wells, drains are almost
always installed in water table aquifers. Rarely are they
used in confined or leaky aquifers. For this reason, most of
the analytical solutions that have been derived for drains are
for water table conditions, although several solutions have
been derived for confined and leaky systems. Due to the
limited usefulness of solutions for confined or leaky systems
for remedial action evaluation, the remainder of this
2-35
-------
subsection will focus on the theory, assumptions and
limitations for solutions applicable to water table
conditions.
The complete mathematical description of time-varying flow to
drains in water table aquifers is nonlinear and intractable
largely because of the effect of the moving water table
boundary. As a result, several simplifications must be made
before analytical expressions can be derived. The
simplifications upon which most of the available expressions
are based are the Dupuit-Forchheimer assumptions and
linearization.
The Dupuit-Forchheimer assumptions are based on the
observation that the slope of the water table in most aquifers
is very small. In addition, under steady-flow conditions
without accretion (i.e., recharge), the water table is a
streamline. These observations lead to the following
assumptions:
1) for small slopes on the water table, flow lines are
horizontal and equipotentials are vertical and
2) the hydraulic gradient is equal to the slope of the
water table and is invariant with depth
In effect, the Dupuit-Forchheimer assumptions make it possible
to neglect vertical flow components. As a result, the
mathematical description for steady flow in a horizontal,
homogeneous, isotropic, water table aquifer without sources or
sinks simplifies to
+ _5_
dy
^ff ^V
Boussinesq (1904) extended the Dupuit-Forchheimer assumptions
to include sources and sinks and time-varying conditions. The
Boussinesq equation is
dx dy 6y K K Ot
where N is the rate of accretion(L/T).
However, the Boussinesq equation is nonlinear. To simplify it
further, the concept of linearization must be invoked. The
most common linearization is to use a constant saturated
thickness when the change in water table elevation is small
compared to the saturated thickness. This is reasonable in
many cases. Cohen and Miller (1983) present a relationship
2-36
-------
for estimating h, the constant saturated thickness
h = d +
(4.18)
where d and D are defined in Figure 4.5. Thi£ thickness
be used to estimate an average transmissivity, T,
can
T = Kh
(4.19)
which can be substituted into the Boussinesq equation to
obtain
s ah
T at
(4.20)
This equation is linear in h, and can be solved for different
boundary conditions to obtain a number of useful analytical
solutions.
Before presenting these solutions, it is important to first
identify those limitations affecting their use. The first set
of limitations relate to the geometry of the water table
aquifer. The Dupuit-Forchheimer assumptions are only valid in
situations where D«d and d«L (see Figure 4.5). In addition,
Bear (1979) notes that errors in predicted heads will be small
when the square of the water table slope is much less than 1.0
(i.e., (Ah/Ax)2 «1.0). It is important to note that if
accuracy in the rate of discharge is more important than
water table elevations, the above geometry limitations can be
neglected in many cases (Bear, 1979).
The second set of limitations relate to specific conditions
where vertical flow components are significant. The
conditions include: 1) the seepage face near a drain, 2) a
ground-water divide, 3) near an impermeable barrier,
4) regions of significant accretion and 5) partially
penetrating drains. Bear (1979) states that at distances
greater than two times the saturated thickness away from these
conditions, the assumptions are valid.
Few of the expressions that will be presented later
incorporate corrections for the effect of partial penetration.
Huisman (1972) presents a series of formulas for partially
penetrating drains of different geometries. Cohen and Miller
(1983) note that Hooghoudt (1940) provided a means of
2-37
-------
I
OJ
00
h(x)
Figure 4.5
Ideal conditions for applying a drain hydraulics method based
on Dupuit - Forchheimer assumptions - D«d and h«L (taken
from Cohen and Miller, 1983).
-------
correcting predicted heads through the use of an equivalent
depth, d . Hooghoudt produced a table of equivalent depths
that can be substituted for d in many of the available
expressions. According to work by Wesseling (1964), the
equivalent depth approach is accurate to within 5% of results
obtained using a more rigorous mathematical approach.
Figure 4.6 shows the graphical relationship between de and d
for different drain spacings, L. Moody (1966) provides a
direct relationship:
(4.21)
de _ .. _
where 2
a = 3.55 - 1.6 - +2 (I)
L V-L'
For _ > 0.3
L = 8 [ lntt/r) - 1.15] (4.22)
de
When only a single drain is being considered, L goes to
infinity and de =d according to Cohen and Miller (1983).
Huismann and Olsthoorn (1983) note that the additional drawdown
due to partial penetration is negative when
(4.23)
where Q is the wetted circumference of the drain.
One other limitation noted by Cohen and Miller (1983) is that
the solutions for drain hydraulics do not consider flow in the
unsaturated zone that may be induced by drains.
Finally, most of the available solutions for drains are
one-dimensional. They are applicable only to those portions
of drains where flow is horizontal and perpendicular to the
axis of the drain. Near the ends of long drains and over most
the length of short drains there are flow components that are
perpendicular and parallel to the axis of the drain. This is
depicted in Figure 4.7. As a result, there is a variation in
flow and drawdown along the length of most drains.
Relationships that can be used to evaluate how both change
along the length of a drain are presented in Huisman (1972)
and Huisman and Olsthoorn (1983).
2-39
-------
QJ
t/1
tO
CO
-4-)
c
o
M
•r—
s_
o
cu
o
-p
c
cu
to
0 4 8 12 16 20 24 28 32 36 40 44 43 52 56 60 64 68 72 76 80 84 88 92 96 100
Measured Depth to Horizontal Base
Figure 4.6
Relationship between equivalent depth
and total depth for different drain
separations (taken from Cohen and
Miller, 1983).
2-40
-------
Figure 4.7
Plan view of flow to a drain of finite
length (taken from Cohen and Miller,
1983).
2-41
-------
Cohen and Miller (1983) recently compiled a large number of
available analytical expressions for flow to drains. Most of
them are for water table aquifers. Some were derived
specifically for confined and leaky conditions. Tables 4.2
and 4.3 include a large number of steady state and transient
drain hydraulic solutions, respectively. Each solution is
again characterized according to: 1) the aquifer
characteristics and drain configurations that can be analyzed,
2) whether the solution is for time-varying or steady-state
conditions, and 3) the type of output that can be obtained.
Another useful compilation of analytical solutions for flow to
drains is contained in Moore (1983). This technical resource
document provides procedures for evaluating the effectiveness
of sand or gravel drainage layers and drain pipes, as well as
compacted clay liners.
4.4 GROUND-WATER MOUNDING ESTIMATION METHODS
Large quantities of leachate can produce a mound in the water
table below certain types of waste disposal facilities and
remedial action technologies. The ponding of waste in lagoons
or impoundments can create large quantities of leachate,
particularly if no liner system is used or if the liner fails.
Certain landfill designs can also produce sufficient
quantities of leachate to cause mounding. Remedial action
technologies like seepage basins are generally used to dispose
of water from pumping wells or drains following treatment.
Since the objective of these technologies is to recharge
water, mounding can also occur under these facilities.
Mounding of a water table aquifer can have a major impact on
local ground-water flow patterns and the resultant movement of
contaminants. This impact often needs to be considered when
evaluating the effectiveness of different remedial action
alternatives.
As was discussed in Subsection 4.2 and 4.3, well hydraulics
and drain hydraulics methods can be used to represent point
sources of recharge (e.g., injection wells) and line sources
of recharge (e.g., seepage ditches or recharge ditches),
respectively. They cannot be used directly, however, to
represent areal sources of recharge (e.g, ponds, seepage
basins and landfills). Multiple point or line sources have to
be used to represent an areal source.
Several analytical methods have been developed for use in
evaluating changes in water table elevations as a result of
recharge from an areal source (Baumann, 1952; Glover, 1960;
2-42
-------
TABLE 4.2 INVENTORY OF SELECTED STEADY-STATE DRAIN HYDRAULICS SOLUTIONS
(adapted from Cohen and Miller, 1983)
Type
C
WT
WT
WT
to
•
*. WT
U)
WT
WT
WT
WT
WT
WT
WT
L
L
WT
WT
UT
H i ,
Aquifer Ch
roperties
H,I
H,I
H,I
H.I
H,I
H,I
H.I
H,I
H,I
IH.I
H,I
H.I
H.I
H.I
H.I
H.I
H.I
aracter
Extent
B
B
B
B
B
B
B
B
B
B
B
SI
IN
IN
IN
B
B
istics
Dimensional 1ty
X
X
X
R
R
R
R
X
X
X
X
R
X
X
X
X
X
Special
Cases
VT
RC
RC
RC
RC
RC.SB
RC.SB
RC
RC
RC
RC
RC
Lfl.RC
LB.RC
RC
Oral
Drain
Number
o
2
2
2
2
2
2
2
2
2
2
1
1
1
1
2
2
i Configurat
Penetration
FP
FP
PP
PP
JTP
FP
FP
FP
PP
PP
PP
PP
FP
FP
FP
PP
FP
ion
Length
IN
IN
in
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
Time
Frame
SS
SS
SS
SS
SS
SS
SS
SS
SS
SS
SS
SS
SS
SS
SS
SS
SS
Output
O.IF
D
D.IF
D
D
0
0
D
D
D
D
D
D.IF
D.IF
D.IF
D
D
References
Huisman (1972)
Jacob (1943)
Ferris et al . (1962)
Bear (1979)
Klrkham (1958)
Luthln (1973)
Harr (1962)
Moore (1983)
Moore (1983)
McBean et al .
(1982)
Klrkham (1967)
Klrkham et al . (1974)
Hooghoudt (1940) Luthln (1973)
van Schllfgaarde (1970)
Luthln (1973)
Youngs (1964", 1966a. 1966b)
Klrkham et al . (1974)
Bouwer (1974)
Huisman (1972)
Huisman (1972)
Huisman (1972)
Bear (1979)
Huisman (1972)
(continued)
-------
TABLE 4. 2 (continued)
LEGEND: C - Confined
WT - Water Table
L - Leaky
H - Homogeneous
IH - Inhomogeneous
I - Isotropic
B - Bounded
SI - Semi-Infinite
IN - Infinite
R - Radial
X - One-Dimensional
VT - Varying Aquifer Thickness
SB - Sloping Base
LB - Leaky Base
RC - Recharge
FP - Fully Penetrating
PP - Partially Penetrating
IN - Infinite
SS - Steady State
TV - Time Varying
IF - Inflow
D - Drawdown
-------
TABLE 4.3 INVENTORY OF SELECTED TRANSIENT DRAIN HYDRAULICS SOLUTIONS
(adapted from Cohen and Miller, 1983)
Type
WT
WT
WT
C
C
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
Aquifer Ch<
Properties
H.I
H.I
H I
H.I
H I
H.I
H.I
H.I
H.I
H.I
H.I
H.I
H.I
H.I
H.I
H.I
iracter
Extent
SI
SI
B
SI
SI
SI
B
B
B
B
B
B
B
B
B
IN
sties
Dimensional ity
X
X
X
X
X
X
X
X
X
X
R
R
R
R
X
X
Special
Cases
LB
RC
RC
SB
SB.RC
Drain
Drain
Number
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
Configurat
Penetration
FP
FP
FP
FP
FP
FP
PP
PP
PP
PP
PP
PP
PP
FP
FP
FP
on
Length
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
Time
Frame
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
Output
D
D.IF
D
D.IF
D
D.IF
D.IF
D
D
D.IF
D
D
D
D
D
D
References
Venetis (1968)
Moody and Ribbens (1965)
U.S. D.I. (1981)
Cooper and
Rorabaugh (1963)
Ferris et al .
(1962)
Ferris et al .
(1962)
Huisman (1972)
Glover (1974) Luthln (1973)
van Schllfgaarde (1974)
Brooks (1961)
Glover (1966, 1974)
Glover (1966, 1974)
Glover (1974)
van Schllfgaarde (1974)
Terzidis (1968)
van Schilfgaarde (1974)
van Schllfgaarde (1974)
van Schllfgaarde (1974)
Chauhan et al . (1968)
Singh and Jacob (1977)
(continued)
-------
TABLE 4.3 (continued)
LEGEND: C - Confined RC
WT - Water Table FP
L - Leaky PP
H - Homogeneous IN
Ih - Inhomogeneous SS
I - Isotropic TV
B - Bounded IF
SI - Semi-Infinite D
IN - Infinite
R - Radial
X - One-Dimensional
VT - Varying Aquifer Thickness
SB - Sloping Base
LB - Leaky Base
i
£*.
CTi
Recharge
Fully Penetrating
Partially Penetrating
Infinite
Steady State
Time Varying
Inflow
Drawdown
-------
Hantush, 1967a; Hunt, 1971 and Rao and Sarma, 1981). Most of
these methods were derived for an areal source, rectangular or
circular in configuration. They can be used to estimate
changes in water table elevations at different radial
distances away from the center of the source area. They can
be applied to sources with different areal configurations by
first converting the actual source area into an equivalent
rectangular or circular area.
In deriving solutions for mounding estimation, it is commonly
assumed that the aquifer is homogeneous, isotropic, infinite
in areal extent, and resting on a horizontal impermeable base.
Further, it is assumed that the seepage rate is uniform and
the water table remains below the base of the facility.
Estimates of mounding using the method by Hantush (1967a) have
been found to be reasonable if the rise of the water table is
not more that 50 percent of the original saturated depth.
In addition to typical aquifer properties like hydraulic
conductivity, specific yield and saturated thickness, the
seepage rate for the areal source must be known. This rate is
difficult to quantify without the use of relatively
sophisticated models or field methods; McWhorter and Nelson
(1980), for instance, present one approach to modeling seepage
from lagoons. At best, only estimates can be obtained with
relatively simple methods. Available methods are discussed
below.
4.5 SEEPAGE/INFILTRATION ESTIMATION METHODS
Many of the methods for estimating seepage rates for ponded
facilities are based on Darcy's Law. One example is a simple
graphical method presented by Knight et al. (1980). It only
requires an estimate of the permeability of the liner or
sludge materials in the bottom of the pond. This method
assumes a unit gradient (i.e., a gradient of one). It is
applicable to situations where the soil beneath the ponded
source is much more permeable than the liner or sludge
materials and where the depth of the liquid is small compared
to thickness of the liner or sludge.
In cases where this method is not applicable, a method
developed by Witherspoon and Narasimhan (1973) can be used.
It is a graphical technique based on results obtained from a
finite element computer model. It requires estimates of the
depth to the water table, pond depth, depth to an impervious
layer, length of the flow domain and drop in hydraulic head.
The first three parameters are generally easy to obtain. The
last two can be estimated using the graphical method by Knight
et al., (1980). Moore (1983) discusses the use of Darcy's Law
2-47
-------
for the purpose of estimating seepage rates and some of its
limitations. Sandberg et al., (1981) note that Darcy's Law
produces rates 2-5 times those calculated with numerical
models by McWhorter and Nelson (1980).
Bicknell (1984) recently developed a computer code that can be
used to estimate chemical emissions from ponded facilities.
Both volatile emissions and leachate quality can be
calculated. Seepage rates are estimated with Darcy's Law or
can be input if they have been measured or calculated with
another model.
Seepage rates for landfills or other areal sources without
ponded surfaces can be estimated with several methods. Fenn
et al., (1975) discuss the "water balance method." Given
monthly values for precipitation and potential
evapotranspiration, estimates of monthly evapo- transpiration,
runoff and infiltration can be obtained for different types of
soils. Seepage rates through multi-layered soil columns can
be estimated through successive applications of the method.
Thus, the water balance method is applicable to a landfill
with or without a cap. Dass et al. (1977) reported on a
similar method.
A somewhat more sophisticated method is incorporated in the
Hydrologic Evaluation of Landfill Performance (HELP) model
(Schroeder et al. 1984a and 1984b). HELP is a quasi-two-
dimensional model that computes a daily water budget for a
landfill represented as a series of horizontal layers. Each
layer corresponds to a given element of a landfill design
(e.g., cap, waste cell, leachate collection system, and
liner). HELP considers a broad range of hydrologic processes
including surface storage, runoff, infiltration, percolation,
evapotranspiration, lateral drainage and soil moisture
storage. The HELP model requires climatologic data, soil
characteristics, and design specifications as inputs.
Climatologic data consist of daily precipitation, mean monthly
temperatures, mean monthly solar radiation, leaf area indices,
root zone or evaporative zone depths, and winter cover
factors. Soil characteristics include porosity, field
capacity, wilting point, hydraulic conductivity, water
transmissivity, evaporation coefficient and Soil Conversation
Service (SCS) runoff curve numbers. Design specifications
consist of the number of layers and their type, thickness,
slope, and maximum lateral distance to a drain, if applicable,
and whether synthetic membranes are to be used in the cover
and/or liner.
While the water balance method described above can be solved
by hand, the large number of calculations performed by HELP
are most efficiently done on a computer. The program is
operational on EPA's National Computer Center in Research
2-48
-------
Triangle Park, North Carolina.
Bicknell (1984) recently modified HELP to include techniques
for estimating chemical emissions. Volatile emissions and
leachate quality can now be computed with HELP.
4.6 SUPERPOSITION
Many of the available analytical solutions were derived for
single wells or drains with constant flow rates or heads in
aquifers of infinite extent. Since few aquifers satisfy these
conditions, it is often necessary to consider the hydraulics
associated with and interactions between multiple wells and
drains and nearby boundaries. This is particularly true for
the evaluation of remedial action performance where wells,
drains and impermeable barriers are often used conjunctively.
It is the principle of superposition that makes it possible to
combine the solutions for single wells and/or drains to obtain
solutions for multiple well and drain systems with variable
flow rates and head conditions. One special type of
superposition, the method of images, makes it possible to add
the effects of boundaries, like streams, ground-water divides
and impermeable zones, to solutions for aquifers of infinite
extent.
Superposition, as defined by McWhorter and Sunada (1977), is
the method in which linear combinations of elementary
solutions are formed to provide additional solutions. The
method is valid for linear, homogeneous, partial differential
equations. Since many of the solutions for wells and drains
are linear, superposition can be used in most instances. The
major exception is flow in water table aquifers. As was noted
in Subsection 4.2, the governing equation is non-linear.
However, if simplification through linearization is
reasonable, superposition can even be used in water table
aquifers. Bear (1979) provides a theoretical description of
the principle of superposition and a procedure for determining
when it can be used.
The most common use of superposition is in the analysis of
multiple well systems. A multiple well problem can be
decomposed into a series of individual well problems. The
resultant draw-down at any point in the aquifer can be
obtained by summing the drawdown produced at that point by
each well. Figure 4.8 shows the drawdowns induced by pumping
each well individually and the resultant drawdown for both
wells together. The same procedure can be used to examine the
effect of varying drawdown along the length of a drain. This
problem can be decomposed onto a series of drains of different
length with different drawdowns.
2-49
-------
t=o -
••>•* • -..
N)
I
01
O
I g_s.'Sy
i i :>DRAWDOWN
UNWELL
r.
WELL
;£fJTIOMETRIC
CONFINED
AQUIFER
Figure 4.8 Superposition of drawdowns for two pumping
wells in a confined aquifer.
-------
Clearly, in some cases, a large number of tedious calculations
could be required to evaluate a remedial action involving
numerous wells and/or drains. Many of the hand-held
calculator and micro-computer programs discussed in Section 5
do this automatically, thus reducing the amount of work
required to evaluate a remedial action.
Superposition can also be used to evaluate time variable
pumping rates. Again, if the equations are linear, the time
variable solution for one pumping rate at a well can be added
to that for another. Figure 4.9 shows this use of
superposition for a well pumping at a rate of Q^ from t = 0 to
t = t2 and then at a rate of Q2 after t = t2«
Another use of superposition is to include the effects of
regional ground-water flow on the drawdowns induced by a well
or drain system. As Huisman (1972) notes, this problem can be
decomposed into two parts: 1) flow of ground water prior to
pumping and 2) flow of ground water due to pumping. The use
of superposition in this manner produces several analytical
expressions useful for remedial action evaluation. The first
of these is for steady flow to a single pumping well in a
uniform steady regional flow. The flow system for this case
is shown in Figure 4.10. The main features of importance in
this flow system are a stagnation point and ground-water
divide. Water within the envelope created by the ground-water
divide will eventually be captured by the well. Water on the
outside of the envelope will be affected only by the regional
ground-water flow. The following relationships can be used to
locate the stagnation point and to estimate the maximum width
of the envelope (Bear 1979):
(4.24)
w= -Q.
(4.25)
where xs = distance to stagnation point, L
Q = well pumping rate, L-^/T
q0 = specific discharge rate for aquifer, L /T
b = aquifer thickness,L
W = maximum width of envelope created by the
ground-water divide, L
Both xs and W are useful in determining what pumping rate
would be required to capture a ground-water plume. Equations
2-51
-------
PUMPING SCHEDULE DRAWDOWN RESPONSE
s
1 ^
Q2-Q1
Figure 4.9
Superposition of drawdowns
to obtain drawdown after a
step change in discharge
(taken from McWhorter and
Sunada, 1977) .
2-52
-------
NJ
I
Ul
P is a
Stagnation
Point
Ls
EQUIPOTENTIAL
WELL LINES
GROUNDWATER
DIVIDE
Figure 4.10 Flow pattern around a pumping well in a uniform
regional flow (after Powers et al., 1981).
-------
4.24 and 4.25 can also be used to examine the envelope that
would eventually be occupied by water or chemicals discharged
from an injection well in a uniform flow. In this case, qQ
has to be replaced by -qo and Q by -Q. This type of analysis
would be useful in examining the portion of an aquifer that
would be affected by a bioreclamation or chemical injection
scheme.
A related set of expressions that can be obtained from the
principal superposition is for a pair (doublet) of pumping and
injection wells in uniform flow. Depending upon the
orientation of the wells and their pumping rates relative to
the regional flow, recirculation of water can be avoided or
maximized. Figure 4.11 shows the envelope created by a
doublet oriented along a line parallel to the direction of
regional ground-water flow. Recirculation is maximized as a
result of placing the injection well directly upgradient from
the pumping well. This type of configuration may be desirable
for chemical injection or bioreclamation actions where the
intent is to perform in-situ treatment and, possibly,
recapture the injected fluid. Powers et al. (1981) provide a
relationship relating the maximum width of the envelope and
the distance between the two wells to the pumping/injection
rate and the properties of the aquifer :
tan"1 £ = *_
2
nvc b
(4.26)
where
c = half-width of the envelope, L
a = half the distance between the wells, L
V = pore water velocity of the regional flow
component, L/T .,
Q = pumping/injection rate, L /T
Figure 4.12 is a dimensionless plot of Equation 4.26.
In using this type of doublet configuration as a remedial
action the time required for partial or complete recovery of
the injected fluid is often of concern. Grove et al. (1970)
provide a solution for estimating this time. The solution is
provided graphically in Figure 4.13.
Recirculation can be minimized by reversing the position of
the wells. This may be important for a ground-water pumping
remedial action where the pumping well is used to capture the
plume and the injection well is used to dispose of treated
water. Bear (1979) shows that for wells with equal pumping
rates recirculation can be avoided when
2-54
-------
GROUND-WATER
DIVIDE
NJ
I
Ul
Ul
RECOVERY
WELL
RECHARGE
WELL
Figure 4.11
Flow patterns around a recharge/recovery doublet in a
uniform regional flow (after Powers et al., 1981).
-------
3 4 5678910
34 56789100
a
in
9
8.
7
6.
K
4
3
2
1.0
g
7
4.
3.
2.
.1
7.
-
"
-
"
01
/
^)
f\
f
1
1
-------
q
QV
I05 1
U6
4
2
I04 1
8
6
4
2
IC)38I
6
4
2
10 2 1
6
4
2
101 <
4
2
10° 1
10
2
c t-~^
V
S i
XX
X
\
\
X
4 6
JS
S, _
»
'v
^>
81 2
X,
x
'»
J\
tJK
"V . V
7>?s_
NJ\
>
2C
U
r
\
k
\
x
X
%
= l<
4
V
V
X
X
r
6
X
s
.
"i
\
^
s N
*!•
31 2
- . \_
"V
"V
^
* ,
$0%
V • >-
' ' ' "\_
' Js. ^
x\
1^
\
v
>
\
«fi
\
\
^^
4 6
X
V
X
FM
0%
™ L
s
h's
3
81 2
N
LL \
..75*/.
. . f\
X
^
\
, \
\
1 ; , \
i V \
s
\
^
1
\
\
v
V
s
4 6
v
^ c
\
\
5r
^
^
^,S
— 1 -
1
1
8 1 2
K>%
i
(
\
\
\
\
4 6
I-
4---
_.
-JTlT
1 1 If
\ f\ \ If
' \ 1 1 I
1
4 io"3 io~2 10"' o°
tv
•ZTFT
81
.0'
q = DISCHARGE/RECHARGE RATE PER UNIT PENETRATDN.
a - HALF THE DISTANCE BETWEEN THE WELLS.
v =DARCY VELOCITY = TRANS PORT VELOCITY TIMES
POROSITY.
n = POROSITY.
t =T1ME REQUIRED TO OBTAIN PERCENTAGE P.
p = PERCENTAGE OF PUMPED FLOW DERIVED FROM
RECHARGE.
Figure 4.13 Percent recharge being discharged
in a doublet (after Grove et al.,
1970).
2-57
-------
Qw < TTd b qQ (4.27)
where d is one-half the distance between the wells. If, for
some reason, this condition cannot be met, the proportion of
recirculation, Qwr, can be estimated by
VrL 2- i , • j_ ^ f w •« / •. ^_i\_a u ^. t \j i
; 7T f ^« K rv *4*-r X
n ( da b Qw ' dcfob ) (4.28)
Wilson (1984) presents analytical solutions for an extension
to the pumping/injection doublet. This extension involves the
use of two pumping wells and two injection wells oriented so
as to create inner and outer recirculation cells that
effectively capture a plume. The pumping and injection wells
are oriented as shown in Figure 4.14. As Wilson notes, the
outer cell reduces the time required to capture the plume and
the amount of water that must be treated. The outer cell also
provides a back-up should the chemical escape the inner cell.
Analytical solutions and type curves are provided by Wilson to
determine cell discussions and plume flushing times.
As was mentioned earlier, most of the available analytical
expressions are based on the assumption of infinite areal
extent. Real world aquifer systems are normally bounded,
however, by streams, lakes and geologic formations. Aquifers
also contain natural ground-water divides. Finally, the use
of impermeable barriers and drains as remedial actions can
also act as boundaries. A special type of superposition, known
as the method of images, can be used to examine the effects of
different boundaries.
The method of images is discussed in some detail by Ferris et
al., (1962), and is generally discussed in most ground-water
text books. The method involves the use of "imaginary" wells
placed in strategic locations to duplicate hydraulically the
effects of physical boundaries. To hydraulically duplicate
the effect of a no-flow boundary, an imaginary pumping well is
used. As Figure 4.8 shows, at the intersection of the cones
of depressions for two pumping wells, a ground-water divide
equivalent to a no-flow boundary is created. A stream can be
hydraulically duplicated by using an injection well as the
image well. In this case, the intersection of the cones of
depression act like a source of recharge. Image wells
normally have the same pumping rates as the real wells, and
are situated on a common line perpendicular to the boundary.
Under these conditions the boundary is located at a distance
2-58
-------
INJECTION
WELLS
PUMPING
WELLS
OUTER CELL
INNER CELL
Figure 4.14
Inner and outer recirculation cells created by
pairs of pumping and injection wells (after
Wilson, 1984). Copyrighted by National Water
Well Association.
2-59
-------
halfway between the image and real wells.
In using the method of images, it is important to recognize
several assumptions and limitations. First, when the method
is used to represent a stream, it is assumed that the real
pumping well does not lower the head in the stream. Second,
when it used to represent a no-flow boundary, the barrier that
is created is assumed to be vertical and fully penetrating.
Thus, situations like shallow streams, partially penetrating
drains or hanging slurry walls cannot be considered. Finally,
the barrier that is created is assumed to be infinite in
length. As a result, the types of aquifer geometries that can
be considered are limited to infinite-strips, semi-infinite
strips, wedge-shapes and rectangles. Figures 4.15 and 4.16
show the placement of image wells for these aquifer
geometries.
As Figure 4.15 shows, in addition to image wells on common
lines perpendicular to each boundary, an additional image well
is needed to balance the effect of the other two image wells
when evaluating "wedge-shaped" geometries. The former are
known as primary image wells, the latter is a secondary image
well (Ferris et al., 1962).
As Figure 4.16 shows, a long line of image wells is required
to evaluate a infinite-strip or semi-infinite strip
geometries. The wells on one side of the strip are required
to balance the effects of the wells on the other. While the
line of wells should be carried to infinity, in practice it is
only necessary to add wells until the next pair has a
negligible influence on the sum of the image well effects at a
point (Walton, 1984a). The number of image wells required to
evaluate a rectangular geometry can be large also (see
Figure 4.16). Such a geometry can lead to a large number of
tedious calculations. Again, many of the available hand-held
calculator and micro-computer programs will automatically sum
the drawdowns for a large number of image wells, thus reducing
the work required to use the method of images.
4.7 TRANSFORMATION METHODS
Most of the available analytical solutions are based on the
assumption that flow occurs in isotropic and homogeneous
media. This assumption is often limiting because all real
systems exhibit some degree of heterogeneity and anisotropy.
Fortunately, there are practical ways to circumvent this
limitation through the use of different "transformation"
methods: 1) equivalent sections, 2) incremental methods, and
3) corrections for anisotropy.
2-60
-------
•
I>
1,
O
Ij
o
I,
u>-
oo:
10 '
I z
WO
B •
i
o
12
y BARRIER
BOUNDARY
-------
-•—lUAGC P4TTf*tr
o
I.
KK
£0
•S TO iHfimTT "S
• o
»K»«»G««
•CtL WtLL
IUACC WILLS. !. t*t *u*tc*co M rue
StOUCHCt in WHICH Tntr r(»f COMSiOfffO
KJtf/ ftTTC»M tCff'Tf TO IftflHir
• • O O
rC wfu,
o o o o o o
-------
The use of equivalent sections basically involves converting
the irregular geometry of a real world aquifer system into an
equivalent system with a regular geometry. The geometries
typically used are those that can be obtained from the use of
the method of images (i.e., strips, rectangles and wedges).
This conversion is required because most analytical solutions
are derived for regular geometries.
In making the conversion to an equivalent system it is often
necessary to account for layered heterogeneities. Layered
heterogeneities are vertical changes in media properites. A
hydraulically equivalent vertical conductivity for a layered
system can be obtained by
5i (4.29)
where
Kz = equivalent vertical hydraulic conductivity, L/T
Ki = vertical hydraulic conductivity of each layer,
L/T
di = thickness of each layer, L
d = total thickness, L
A hydraulically equivalent horizontal conductivity can be
obtained by
K = V i i (4.30)
X f-' j
where K x = equivalent horizontal hydraulic conductivity,
L/T
Horizontal changes in media properties, or trending
heterogeneities, also have to be considered when converting to
equivalent systems. Hydraulically equivalent horizontal and
vertical hydraulic conductivities can be obtained by
K _ d_
*x~ H—:
i=l
2-63
-------
n K. d.
K = y -^-i- (4.32)
Z *- A
1=1 d
respectively, where d now is a horizontal distance rather than
the total depth. Walton (1984a) and Ferris et al., (1962)
recommend that when transient well or drain analyses are being
conducted in systems with trending heterogeneities, the
hydraulic conductivity be adjusted as the cone of depression
moves outward. The initial value of the hydraulic
conductivity would be equal to that for the media adjacent to
the well or drain. When the cone of depression encounters
another media, the hydraulic conductivity should be adjusted.
This procedure continues until drawdowns stabilize. Walton
(1984a) refers to this approach as the incremental method.
Another type of incremental method is to divide the aquifer
into regions with relatively uniform properties and then apply
the analytical solutions in a step-wise fashion to each
region. Bear (1979) recommends this approach for water table
aquifers with appreciable variations in head.
In many systems there may be distinct differences between
horizontal and vertical hydraulic conductivities. In these
cases, corrections need to be made before solutions based on
isotropic conditions can be used. Huisman and Olsthoorn
(1983) present a series of formulas for making corrections
K' = VKKz (4.33)
x' = x (4.34)
y1 = -- y (4.35)
z' =>/— z (4.36)
K,
2-64
-------
The resistance factor for leaky systems becomes:
c' =
K1
(4.37)
where c is the ratio of the leaky layer thickness to its
hydraulic conductivity.
The specific yield becomes:
S • = — S,, or S ' =
Huisman (1972) suggests the following correction for the
radius, r:
r1 =
(4.39)
He notes also that when the vertical hydraulic conductivity is
similar to the horizontal hydraulic conductivity and the
influence of the well is large compared to the saturated
thickness, there is no need to • consider the effect of
differences in the vertical and horizontal hydraulic
conductivities.
4.8 CONFORMAL MAPPING
Conformal mapping is a method for deriving analytical
solutions by transforming a problem from one geometrical
domain for which a solution is needed to one for which a
solution can be obtained. This method has been used to derive
expressions for selected two-dimensional ground-water flow
problems involving relatively complicated geometries (e.g.,
seepage under cut off walls and through earthen dams). The
theory behind the method is discussed by Harr (1962).
A major disadvantage of the method is that it is
mathematically involved, and often produces fairly complex
analytical solutions. The major advantage is that it provides
solutions for one class of flow problems that cannot be
considered with the methods discussed so far: flow under
partially penetrating impermeable barriers or barriers that
2-65
-------
are keyed into leaky formations. As was noted in Subsection
4.6, the method of images requires that an impermeable barrier
be fully penetrating and that no seepage occurs beneath the
barrier. In addition to the solutions presented below, Knox
(1984) recently developed an analytical technique for
estimating the flow under a partially-penetrating barrier. A
technique for developing breakthrough curves for contaminanted
ground-water is also presented.
Figure 4.17 shows a barrier (i.e., grout curtain, slurry wall
or sheet piling) that partially penetrates a horizontal water
table aquifer. The quantity of seepage under this barrier can
be estimated with the following relationship:
q =
khK1
2K
(4.40)
where q = flow rate per unit width, L /T
k = hydraulic conductivity, L/T
h = head difference, L
K1, K = values of complete elliptic integral of
the first kind
Values of K'/K have been tabulated for a range of values of
m^ , the modulus (see Table 4.4). The modulus for this case is
m
sin
7TS >
2T'
(4.41)
where s and T are defined in Figure 4.17.
The above expression is for a single layered, homogeneous,
isotropic aquifer. In many situations, aquifers with two
layers of differing permeabilities are encountered. Harr
(1962) provides a method for estimating the quantity of flow
under a barrier for a two-layered system. It involves the
calculation of a dimensionless parameter using the following
relationship:
tan
(4.42)
where
ki= hydraulic conductivity of the upper
2-66
-------
Y///////////////////////////////////////////////'
Figure 4.17
Configuration of an impermeable
barrier that partially penetrates
into a single-layered aquifer
(taken from Harr, 1962). Copy-
righted by McGraw-Hill.
2-67
-------
TABLE 4.4 COMPLETE ELLIPTIC INTEGRALS OF THE FIRST
KIND* (taken from Harr, 1962) Copyrighted
by McGraw-Hill
m»
0.000
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0.009
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
0.13
0.14
0.15
0.16
0.17
0.18
0.19
0.20
m'«
K
1.571
1.571
1.572
1.572
1.572
1.573
1.573
1.574
1.574
1.574
1.575
1.579
1.583
1.587
1.591
1.595
1.599
1.604
1.608
1.612
1.617
1.621
1.626
1.631
1.635
1.640
1.645
1.650
1.655
1.660
K'
K'
to
4.841
4.495
4.293
4.150
4.039
3.949
3.872
3.806
3.748
3.696
3.354
3.156
3.016
2.908
2.821
2.747
2.684
2.628
2.578
2.533
2.493
2.455
2.421
2.389
2.359
2.331
2.305
2.281
2.257
K
K
Jc
0.000
0.325
0.349
0.366
0.379
0.389
0.398
0.406
0.413
0.420
0.426
0.471
0.502
0.526
0.547
0.565
0.582
0.598
0.612
0.625
0.638
0.650
0.662
0.674
0.684
0.695
0.706
0.716
0.726
0.735
K'
~K
K'
~K
to
3.08
2.86
2.73
2.64
2.57
2.51
2.46
2.42
2.38
2.35
2.12
1.99
1.90
1.83
1.77
1.72
1.67
1.63
1.60
1.57
1.54
1.51
1.48
1.46
1.44
1.42
1.40
1.38
1.36
K
Jc
m'1
1.000
0.999
0.998
0.997
0.996
0.995
0.994
0.993
0.992
0.991
0.99
0.98
0.97
0.96
0.95
0.94
0.93
0.92
0.91
0.90
0.89
0.88
0.87
0.86
0.85
0.84
0.83
0.82
0.81
0.80
ml
m1
0.21
0.22
0.23
0.24
0.25
0.26
0.27
0.28
0.29
0.30
0.31
0.32
0.33
0.34
0.35
0.36
0.37
0.38
0.39
0.40
0.41
0.42
0.43
0.44
0.45
0.46
0.47
0.48
0.49
0.50
m'J
K
1.665
1.670
1.675
1.680
1.686
1.691
1.697
1.702
1.708
1.714
1.720
1.726
1.732
1.738
1 744
1.751
1.757
1.764
1.771
1.778
1.785
1.792
1.799
1.806
1.814
1.822
1.829
1.837
1 846
1.854
K'
K'
2.235
2.214
2.194
2.175
2.157
2.139
2.122
2.106
2.090
2.075
2.061
2.047
2.033
2.020
2.008
1.995
1.983
1.972
1.961
1.950
1.939
1.929
1.918
1.909
1.899
1.890
1.880
1.871
1.863
1.854
K
K
~K'
0.745
0.754
0.763
0.773
0.782
0.791
0.800
0.808
0.817
0.826
0.834
0.843
0.852
0.860
0.869
0.877
0.886
0.895
0.903
0.911
0.920
0.929
0.938
0.946
0.955
0.964
0.973
0.982
0.991
1.000
K'
/T
K'
~K
1.34
1.33
1.31
1.29
1.28
1.26
1.25
1.24
1.22
1.21
1.20
1.19
1.17
1.16
1.15
1.14
1.13
1.12
1.11
1.10
1.09
1.08
1.07
1.06
1.05
1.04
1.03
1.02
1.01
1.00
K
~K'
m'*
0.79
0.78
0.77
0.76
0.75
0.74
0.73
0.72
0.71
0.70
0.69
0.68
0.67
0.66
0.65
0.64
0.63
0.62
0.61
0.60
0.59
0.58
0.57
0.56
0.55
0.54
0.53
0.52
0.51
0.50
TO*
* From V. I. Aravin, and S. Numerov, "Seepage Computations for Hydraulic
Structures," Stpoitel'stvu i Arkhitekture, Moscow, 1955.
2-68
-------
layer, L/T
k2 = hydraulic conductivity of the lower
layer, L/T
The ratio of s/T is also calculated where s and T are defined
in Figure 4.18. It is important to note that the thickness of
each layer is assumed to be equal. Given values for e and
s/T, the seepage rate can be obtained for using Figures 4.19
and 4.20.
In cases where k^^k-i, the seepage rate can be calculated
directly by
(4.43)
One key assumption behind both of the above methods is that
the flow is occuring in two-dimensions only, the horizontal
and vertical dimensions. This is equivalent to assuming the
impermeable barrier is infinitely long and, therefore, no flow
occurs around the ends. Since impermeable barriers used as
remedial actions will always be of finite length, care must be
exercized in using the above methods.
4.9 CONTAMINANT TRANSPORT
All of the analytical methods discussed in the previous
subsections are useful for evaluating the changes only in
ground-water flow patterns and hydraulic heads associated with
wells, drains, mounds and impermeable barriers. Another area
of interest is the effect of these actions on contaminant
movement.
A number of analytical solutions for contaminant transport
have been developed. Most of them are based on the classical
convection-dispersion equation. In addition to these
analytical solutions, several semi-analytical methods have
also been developed. The theory behind the available
analytical and semi-analytical methods is discussed below.
One form of the classical partial differential equation for
contaminant transport in two-dimensions is
ac _ i pc _ R ac
s\*l&~ — I\ ~~~~~
ay 0X at
DV^+DT-^-V^-AFC = R^ (4.44)
•*-< a 2 •!• .a 2 Q »».
2-69
-------
QL-Z
Aq paq.q6TJAdoo '(2961 'JJ*?H uiojj
aajTnbe
UB jo
81
J
;;;;;;;z£;;;;;;;;
^
;;;;;;;;;;;;;;;;*;
xxxxxxx/xx/xx/xx/x
XXXXXXXXXXXXXXXXXX
XXXXXXX////XXXXXXX
/xxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxx
XXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXX
xxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxx
X X X X X X X ^ ^^ X X X X X X X
xxxxxxxxx^lxxxxxxx
xxxxxxxxxxxxxxxxx/
XXXXXXXXXXXXXXXXXX
xxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxx^
xxxxxxxxxxxxxxxxxx1
£
L
xx/xxxxxxxx/xxx<
•xxxxxxxxxxxxxxx*
•xxxxxxxxxxx/xxx/
'XXXXXXXXXXXXXXXy
'XXXXXXXXXXXXXXX
'XXXX/XXXXXXXXXX
'XXXXXXXXXXXXXXX
'XXXXXXX//////XX
'XXXXXXXXXXXXXXX
'XXXXXXXXXXXXXXX
'XXXXXXXXXXXXXXX
'////XXXXXXXXXXX
'XXXXXXXXXXXXXXX
'XXXXXXXXXXXXXXX
'XXXXXXXXXXXXXXX
^XXXXXXXXXXXXXX
XXXXXXXXXXXXXXX
k Jk
M
1 r
///
'/ /s
' / ' /
'/ / /
I
I
Jw
v9x
i
88
88
$
1
1
XXX
Rxx
I
^x/xxxxx
xxxxxxxx
xxxxxxx x
/XXXXXXX
XX /XXXJ
XXXXXX-
xxxxxx
xxxxxx
xxxxxx
xxxxxx
xxxxxx
XXX XXX
xxxxxx
xxxxxx
xxxxxx
XXXXX X
xxxxxx
xxxxxx.
XXXX XX 3
x/xxxxxx/x^
XXXXX XXXXXX
XXXXX X XX/XX
XX XXXXXX XXX
^X X XXXX XXXX
x xxxxxxxxxx
X XXXXXXXX/X
/XXXXXX XXXX
xxxxxxxxx^ x
xxxxxxxxO/
/xxxxx xx rxx
XXjt/XX XXXXX
x/Sxxxxxxxx
x xxxxxxx/xx
XXXXX XXXXX/
XXXXXXXXX XX
XXXXX XXXXXX
/XXXXXXXXXJ
XXXXX XXXXX^
A
...J
p
• • -1
1^
/ / / f
f / f /
/ / S f
"/ f / /
//"
f / f t
////
/ / / /
/ / "
/ / //
/ ' ' '
'/ s s /
/'"
/ s / s
/ // /
f'Sf
i' "
V S//
r
v;;'x'xV;;;;;
X XXX XXXX XXX
XXXXXXXXXXX
xxxxxxxxxx x
XXXXXXXXXX X
XXXXXXXXXXX
XXXX XXXXX X X
xxxxxxxxx/x
X//////XXXX
XXX X XXXXXXX
XXXXXXXXXXX
XXXXXXXXXXX
XXXXXXXXXXX
XXX X XXXXXXX
XXXXXXXXXXX
XXXXXXXXXX X
XX//XXXXXXX
xxx/xxxxxxx
XXXXXXXXXXX
xxxxxx/xxxx
-------
0.05 0.10 0.15
0.20 0.25
£
0.30 0.35 040 0.45
Figure 4.19 Relationship between e and
the flow under a partially
penetrating barrier in a
layered aquifer (taken from
Harr, 1962; after
Polubarinova-Kochina, 1952).
Copyrighted by McGraw-Hill.
2-71
-------
V4 Ve 1
Figure 4.20 Relationship between
depth of penetration
and flow under a
partially penetrating
barrier in a layered
aquifer (taken from
Harr, 1962; after
Polubarinova-Kochina,
1952). Copyrighted
by McGraw-Hill.
2-72
-------
where
DT =
C =
V =
A =
R =
P
ne
Kd
f-\
longitudinal dispersion coefficient, L /T
transverse dispersion coefficient, L^/T
contaminant concentration, M/L-3
average pore water velocity, L/T
contaminant degradation rate, /T
l+pKd/ne = retardation coefficient,
dimensionless 3
bulk density, M/L
effective porosity, dimensionless 3
equilibrium partitioning coefficient, L /M
In deriving this equation it is assumed that flow is steady
and uniform in the x direction, and that the aquifer is
composed of homogeneous and isotropic media; the contaminant
is assumed to have a constant density and viscosity. Almost
all of the available analytical solutions based on the
convection-dispersion equation are for steady and uniform flow
conditions, except for a few that were derived for radial flow
problems (i.e., flow to wells). Further, it is generally
assumed that contaminant adsorption/desorption can be
described as a linear and completely reversible process, and
that contaminant degradation can be described as a first-order
process. Finally, it is assumed that the dispersion and
diffusion can be grouped together and described as a Fickian
process (i.e., obeys Pick's first law). The coefficient for
the dispersion component is assumed to be directly
proportional to the pore water velocity and does not vary in
time or space.
The available analytical solutions based on the
convection-dispersion equation are consolidated in several key
publications. Van Genuchten and Alves (1982) provide
derivations for a relatively complete set of one-dimensional
analytical solutions. Walton (1984a) presents several
one-dimensional solutions, and a number of radial flow
solutions involving single injection and withdrawal wells with
and without regional flow. Other good sources of available
solutions include Bear (1979) and Javandel et al. (1984), and
Cleary and Ungs (1978).
Donigian et al., (1983) developed a methodology for evaluating
the potential for ground-water contamination under emergency
response conditions. The methodology uses a nomograph-based
solution to the one-dimensional, convection-dispersion
equation. Detailed guidance on parameter estimation is
provided as part of the methodology.
While a large number of analytical solutions are available,
their use in the evaluation of remedial action performance is
limited to two types of analysis. First, the solutions can be
2-73
-------
used to estimate the rate and direction of plume migration
away from an uncontrolled disposal site. This type of
information is useful when determining how the extent of
ground-water contamination may change with time and the type
of remedial action that may be needed sometime in the future.
If a remedial action needs to be implemented immediately, this
type of information will be of limited value since monitoring
and site characterization will have already determined the
extent of contamination. The second type of analysis applies
to those remedial action technologies that involve injection
and recovery (e.g., bioreclamation and chemical injection).
Analytical solutions for radial flow can be used to examine
the portion of an aquifer that will be affected by an injected
fluid and the time required for the injected fluid to reach a
recovery well. Both are needed in the evaluation and design
of in-situ treatment technologies. Some of the available
expressions also include the effect of a regional flow
component.
The first major type of semi-analytical method for contaminant
transport is based on the complex velocity potential concept.
Like superposition, this concept involves separating a complex
flow field that itself is intractable into a series of simple
flow fields for which tractable solutions are available. The
velocity potentials and stream functions for each simple flow
field are combined to obtain a complex velocity potential.
Javandel et al., (1984) provide a procedure for constructing
complex velocity potentials. Bear (1979) discusses the theory
behind the concept; he refers to it as the sharp front
approximation.
The complex velocity potential concept has several advantages
and disadvantages. Its major advantage is that it is
generally more powerful than analytical methods, largely
because more complex flow systems can be considered. Its
major limitation is that the concept generally applies only to
the transport of water-coincident contaminants. That is,
contaminants that move at the same velocity as the ground
water. As a result, the effects of dispersion are not
considered. In a few cases, retardation and decay can be
considered (see Nelson and Schur, 1980). Another limitation
is that some of the solutions developed using this concept are
mathematically complex. While they can be solved by hand with
the aid of tables or graphs of appropriate mathematical
functions, they are generally programmed for use on computers
or hand-held calculators to reduce the work involved in their
application. Those expressions that have been programmed for
use on hand-held calculators or micro-computers are discussed
in Section 4.
Javandel et al. (1984) and Bear (1979) overview some of the
available solutions that have been derived using the complex
2-74
-------
velocity potential concept. These expressions are applicable
to homogeneous, isotropic, saturated aquifers with uniform
flow. They can be used to predict contaminant transport in
two-dimensions (lateral and longitudinal). A large number of
injection wells, withdrawal wells and circular sources of
finite radius (e.g., ponds and lagoons) can be evaluated.
Thus, these expressions are applicable to a wide range of
subsurface and waste control remedial action technologies.
The other type of semi-analytical methods for contaminant
transport is based on a simple numerical technique discussed
by Bear (1979). This technique involves tracking the movement
of one or more particles of water with time. The rate and
direction of particle movement at any location in the
ground-water flow field is estimated by calculating the
component pore water velocity towards each pumping well and
away from each injection well. These component velocities are
summed together along with the regional component of pore
water velocity to obtain a resultant velocity vector. The
particle of water is then moved in the direction of the
resultant velocity. The distance it is moved is determined by
the magnitude of the resultant velocity and the time interval
selected by the analyst. The accuracy of the method improves
as the length of the time interval decreases, particularly as
the particle approaches a well.
The simple numerical method or particle mover method can be
used to analyze contaminant transport in two ways. The first
way involves tracking the movement of individual particles
released from the perimeter of a waste site or plume to
determine whether all particles will be recovered and how long
it will take. The second way is to determine the location of
a number of particles at the end of selected time increments.
If a sufficient number of particles are tracked, the position
of the perimeter of a plume can be estimated at the end of
each time increment.
In using the simple numerical technique it is assumed that the
contaminant is water-coincident. For contaminants that are
not water-coincident, the travel times obtained from the
technique can be adjusted by multiplying by the retardation
factor for the contaminant of interest.
Due to the numerous repetitive calculations that must be
performed when using this method, programs have been written
for hand-held calculators to reduce the effort involved in
their use. These programs allow the user to adjust the time
interval and/or distance a particle is allowed to move. They
also allow the user to consider a large number of injection
and recovery wells. As a result, a range of remedial action
technolgies can be considered.
2-75
-------
SECTION 5
AVAILABLE HAND-HELD CALCULATOR AND MICRO-COMPUTER PROGRAMS
5.1 OVERVIEW
The use of many of the analytical methods discussed in Section
4 to solve ground-water problems of practical interest will
generally require numerous, repetitive calculations. An
evaluation of the cone of depression for a single pumping
well, for instance, will involve solving one of the well
hydraulics equations for a number of radial distances away
from the well at different points in time. If the cone of
depression for a number of wells is of interest, the drawdown
for each well will have to be calculated and then summed to
obtain the total drawdown. This type of problem can lead to a
large number of calculations.
With the recent development of relatively powerful,
programmable hand-held calculators and micro-computers, the
work involved in using many of the analytical methods is
greatly reduced. Calculators and micro-computers can rapidly
perform a large number of repetitive calculations in minutes
that would otherwise require hours or days. As a result, the
level of manpower required to solve a given problem is greatly
reduced.
Calculators and microcomputers also have several other
advantages. First, they can reduce the need for tables and
graphs that are commonly required to solve analytical
expressions. Values for the "well functions" in most well
hydraulics equations generally have to be obtained from tables
or graphs. While the required tables and graphs can be found
in many ground-water textbooks and related publications, some
of the "well functions" can be approximated by series
expansions or mathematical functions that are easily solved on
a calculator. Many of the functions contained in other types
of analytical methods can also be approximated or solved
directly with calculators. In addition, simple integration
schemes that would require numerous, tedious calculations to
solve by hand can also be used to quickly solve certain
analytical equations.
2-76
-------
A second advantage is that calculators and micro-computers
offer peripherals that aid in the analysis of remedial
actions. The programs required to solve different analytical
expressions can be stored on magnetic cards, magnetic tape or
disks. When an analysis is required the programs can be
loaded rapidly. This reduces the level of effort required to
key in a program or to make repetitive key strokes on a
non-programmable calculator. Results of different analyses
can be stored for use later or printed immediately. This
reduces the level of effort involved in transcribing results.
The final advantage is that programmable calculators and
micro-computers are readily available. Most site contractors
and many state and Federal Superfund staff have access to
them. In addition, software availability is increasing
rapidly, particularly for micro-computers.
This section will identify what programs are currently
available for solving the analytical expressions discussed in
Section 4. It is important to note that there are a large
number of programs currently available, particularly for
hand-held calculators, and more are being written all the
time. These programs have been written to meet a number of
different needs, ranging from the solution of simple numerical
ground-water flow problems, to the solution of well hydraulics
equations, to the analysis of pump test data. This section
will focus only on those written for analytical or
semi-analytical methods that are of value in the evaluation of
remedial action performance. Those readers interested in
programs available for other types of analyses should consult
the International Ground Water Modeling Center (IGWMC),
Holcomb Research Institute, Butler University, Indianapolis,
Indiana. IGWMC provides a clearinghouse of available
programs.
It is also important to note that while an attempt was made to
be comprehensive in the identification of available programs,
resources were not available to consult every possible source.
Therefore, the programs identified herein should be considered
as representative of those that are available.
5.2 AVAILABLE PROGRAMMABLE, HAND-HELD CALCULATOR PROGRAMS
The large number of programs currently available for
programmable, hand-held calculator programs is an indicator of
their wide-spread use in solving practical ground-water
problems. Despite the large number of programs that are
available, they have only been written for a relatively small
number of analytical methods. This is, in part, due to the
2-77
-------
fact that hydro-legists have found that most problems can be
solved with just a few methods. It is also due to the fact
that programs are difficult, if not impossible, to write for
certain methods. These methods are generally ones where
reasonable approximations are not possible or graphical
solutions are required.
The largest proportion of available programs are for well
hydraulics, mainly because hydrologists are commonly faced
with problems involving wells and because many of the well
hydraulics equations are easily programmable. Table 5.1
provides a summary of a selected group of available programs
for well hydraulics. It shows some of the basic assumptions
and limitations for each program in terms of aquifer
characteristics and well configurations. It also shows
whether steady-state or time-varying analyses can be performed
and the output provided by the program. Finally, the table
lists the calculator for which the program is written and a
reference or source for the program. Again, these programs
were selected from the many that are currently available
because of their usefulness for remedial action evaluation.
As Table 5.1 shows, most of the programs were developed for
confined, homogeneous, isotropic aquifers of infinite extent.
These programs can also be applied to water table aquifers as
long as the assumptions discussed in Subsection 4.2.3 are not
violated. Some of the programs were written explicitly for
leaky aquifer systems. Corrections for heterogeneities and
anisotropy have to be made using the methods discussed in
Subsection 4.7 since only one of the programs considers other
than homogenous and isotropic conditions. Aquifers that are
bounded or remedial actions that include impermeable barriers
cannot be analyzed explicitly with these programs unless the
method of images is used. Since the method of images requires
at least one real well and one image well, those programs that
consider more than one well are particularly well suited to
the analysis of bounded aquifers. These programs
automatically sum the drawdowns attributable to each well.
Most of the well hydraulics programs are for fully penetrating
wells. Corrections for partial penetration need to be made,
if they are not explicitly considered. Walton (1984a) and
others discuss methods for making the needed corrections.
Time-varying estimates of drawdown at different locations are
typically provided by the available programs. Only a few
provide steady-state results or inflows to a well.
Finally, well hydraulics programs are available for both Texas
Instruments (mainly model 59) and Hewlett Packard (mainly
model 41C or 41CV) calculators. Listings for some of these
programs have been published in the open literature (e.g.,
2-78
-------
TABLE 5.1 AVAILABLE HAND-HELD CALCULATOR PROGRAMS FOR WELL HYDRAULICS
Program Title
General Aquifer Analysis
for Nonsteady Thels Condi-
tions
Multiple Well, Variable
Pumping Rate Problems
tvj
1 Constant or Variable
*-J Pumping (Injection) Rate,
^ Single or Multiple Fully
Penetrating Wells
Constant or Variable
Pumping (Injection) Rate,
Single or Multiple Fully
Penetrating Wells
Oewatering Well Design
Thels Condition Well Field
Point Sink Aquifer Model
Nonsteady State Nonleaky
Artesian-Single Produc-
tion Well
AQMODL (4)
Nonsteady State Nonleaky
Arteslan-Partially Pene-
trating Wells
Aquife
Type
C
C
C
C
C
C
C
C
C
C
r Character
roperties
H.I
H,I
H,I
H.I
H.I
H,I
H.I
H.I
H.I
H.I
istics
Extent
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
Well Co
Number
24
1
_
_
24
57
50
1
60
1
nf iguration
'enetration
FP
FP
FP
FP
FP
FP
FP
FP
FP
PP
Time-
frame
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
Vogram
Output
D
D
D
D
D
D
D
D
D
D
Calculator
Type
TI-59
HP-29C
TI-59
HP-97
TI-59
HP-41
HP-41
TI-59
HP-41
TI-59
Reference
Sandberg et al . 1981
Prickett and Vorhees 1981
Picking 1979
Warner and Yow 1979
Rayner 1981
Koch and Associates (1)
IGWMC (2)
Ulrick (3)
Walton 1983
Rayner 1983
Walton 1983
(continued)
-------
TABLE 5.1 (continued)
Program T1 tie
Constant Pumping
(Injection) Rate, Fully
Confined Aquifer, Parti-
ally Penetrating Well
Radial Flow to a Constant
Drawdown Hemisphere
Analysis fo Source or Sink
Flow Rates with Drawdown
as a Given
to Nonsteady Discharge of a
1 Flowing Well
CO
0 Anlsotroplc Confined
Aquifers
Jacob Leaky Artesian
Steady-State
Steady State Leaky Arte-
sian - Single Production
Hell
Nonsteady State Leaky
Artesian - Single
Production Well
Leaky Aquifer Drawdown
Constant Pumping
(Injection) Rate, Single
Aquif
Type
C
C
C
C
C
L
L
L
L
L
er Characte
Properties
H,I
H.I
H.I
H.I
H.A
H.I
H.I
H.I
H.I
H.I
ristics
Extent
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
Well C(
Number
1
1
7
1
1
25
1
1
1
1
jnfiguration
Penetration
PP
PP
FP
FP
PP
FP
FP
FP
FP
FP
Time-
frame
TV
TV
TV
TV
TV
SS
SS
TV
TV
TV
Program
Output
D
IF
IF
IF
D
D
D
D
D
D
Calculate
Type
TI-59
TI-59
TI-59
TI-59
TI-59
HP-41
TI-59
TI-59
TI-59
HP-41
TI-59
Reference
Warner and Yow 1980b
Koch and Associates (1)
Sandberg et al . 1981
PMckett and Vorhees 1981
Koch and Assocaites (1)
Parr et al . 1983
T.A. PMckett and
Associates (5)
Walton 1983
Walton 1983
Ulrlck (3)
Warner & Yow 1980a
Fully Penetrating Well,
Semlconflned Aquifer
(continued)
-------
TABLE 5.1 (continued)
Program Title
Hantush "Well Function"
Nonsteady State Two
Mutually Leaky Artesian
Aquifers - Single Pro-
duction Well
Steady Radial Ground-Water
Flow in a Finite Leaky
Aquifer
Successive Steady States -
Constant Head Points -
Unconfined Aquifer
Aquife
Type
L
L
L
WT
r Character
Properties
H.I
H.I
H.I
H.I
istics
Extent
IN
IN
B
IN
Well Cc
Number
1
1
1
7
nf iguration
Penetration
FP
FP
FP
FP
Time-
frame
TV
TV
SS
TV
Program
Output
0
D
D
IF.D
Calculator
Type
HP-41
TI-59
HP-41
TI-59
Reference
IGWMC(2)
Walton 1983
IGWMC(2)
Koch and Associates (1)
1
00
(1) Programs available as of October 1983 from Koch and Associates, 1660 S. Fillmore Street, Denver, Colorado, 80210
(2) Programs available as of May 1984 from the International Ground Mater Modeling Center, Holcomb Research Institute,
Butler University, 4600 Sunset Avenue, Indianapolis, Indiana, 46208
(3) Programs available as of August 1983 from James S. Ulrick and Associates, 2100 Los Angeles Avenue, Berkeley, California. 94707
(4) Programs can also consider regional water level changes with time and the effects of a regional gradient
(5) Programs available as of July 1983 from Thomas A. Prickett and Associates, Inc., 8 Montclalr Road, Urbana, Illinois, 61801
LEGEND: C - Confined
L - Leaky
WT - Hater Table
H - Homogeneous
I - Isotropic
A - Anisotropic
IN - Infinite
B - Bounded
FP - Fully Penetrating
PP - Partially Penetrating
TV - Time Varying
SS - Steady State
D - Drawdown
IF - Inflow
-------
Warner and Yow, 1979, 1980a, 1980b; and Rayner, 1981, 1983).
The rest are available for purchase from different sources.
Both documentation and pre-programmed magnetic cards are
available when they are purchased.
The available programs for drain hydraulics are summarized in
Table 5.2. It essentially follows the same format as Table
5.1, except this table shows the drain configurations that can
be considered.
Again, most of the available programs are for confined,
homogeneous, isotropic aquifers of infinite extent. While
only a few were explicitly written for water table conditions,
most of the others can be used as long as the assumptions
discussed in Subsection 4.3 are valid. The dimensionality
column in Table 5.2 refers to whether or not the drain is
assumed to be finite or infinite in length. If it is assumed
to be infinite, differences in drawdown and inflow along the
length of the drain cannot be considered.
As with the well hydraulics programs, the method of images is
required if a bounded aquifer is being analyzed. Programs
which can consider the drawdowns for multiple drains will
facilitate such analyses. They will also facilitate analyses
of multiple drains or drains with irregular boundaries.
Sandberg et al., (1981) and Prickett and Voorhees (1981)
provide a number of useful examples of how programs for
multiple drains can be used to represent a number of
ground-water conditions of practical interest (e.g.,
meandering river, lake shoreline or canal system).
All of the drain programs assume full penetration. The
equivalent depth correction discussed in Subsection 4.3 or a
different analytical expression (see Tables 4.4 and 4.5) will
need to be used for partially penetrating drains.
The available programs can be used to estimate both
steady-state and time-varying drawdown and inflow.
Table 5.3 lists the available programs for evaluating
ground-water mounding. They are all based on the theory by
Hantush (1967a). As was mentioned in Subsection 4.4, a
seepage rate estimate is needed to evaluate the potential for
mounding. For landfills this rate can be obtained from a
computer code like the HELP model (Schroeder et al., 1984 a
and 1984b) or the other simple techniques discussed in
Subsection 4.5. There are also several simple techniques for
ponds and impoundments (see Subsection 4.5). None of these
simple techniques have been programmed for hand-held
calculators. The only exception is a Hewlett-Packard 41
program written by James S. Ulrick and Associates in Berkeley,
California. It performs monthly water balance calculations
2-82
-------
TABLE 5.2 INVENTORY OF SELECTED HAND-HELD CALCULATOR PROGRAMS
FOR DRAIN HYDRAULICS
Program Title
Steady-State Draw-
down Around Fi-
nite Line Sinks
Successive Steady
States - Constant
Head Finite Line
Sinks - Compute
Drawdowns
to Finite Line Sinks
1 for Nonsteady
00 Conditions
Line Sink Aquifer
Model
Study of Steady-
State Flow to
Finite Line
Sources or Sinks
with Drawdown as
the Given
Successive Steady
States - One
Dimensional In-
flow to a Line
Successive Steady
States - Constant
Head Finite Line
Sinks - Compute
Inflows
Type
C
C
C
C
C
C
C
Aquifer (
Properties
H.I
H.I
H.I
H.I
H.I
H.I
H.I
"haract<
Extent
IN
IN
IN
IN
IN
IN
IN
;ristics
Dimensional ity
X-Y
X-Y
X-Y
X-Y
X-Y
X
X-Y
Ora
Number
10
10
15
15
6
1
6
in Configura
Penetration
FP
FP
FP
FP
FP
FP
FP
tion
Length
F
F
F
F
F
IN
F
Time-
frame
SS
SS
TV
TV
SS
SS
SS
Program
Output
D
D
D
D
IF
IN
IF
Calcu-
lator
Type
TI-59
TI-59
TI-59
HP-41
TI-59
TI-59
TI-59
Reference
Sandberg et al . 1981
Prickett and Vorhees
1981
Koch and Associates (1 )
Sandberg et al . 1981
Prickett and Vorhees
1981
Ulrick (2)
Sandberg et al . 1981
Prickett and Vorhees
1981
Koch and Associates (1)
Koch and Associates (1)
(continued)
-------
TABLE 5.2 (continued)
Program Title
One Dimensional ,
Nonsteady Flow
to a Constant
Drawdown, Infi-
nite Line Sink
or Source
One Dimensional ,
Nonsteady Flow
to an Increasing
Drawdown, Infi-
nite Line Sink
or Source
Boussinesq Solution
One Dimensional ,
Nonsteady Flow to
a Constant Draw-
down, Infinite
Line Sink or
Source with
Recharge
One Dimensional
Non-Steady Ground
Water Flow (3)
Type
C
C
WT
WT
WT
Aquifer
Properties
H.I
H.I
H.I
H,I
H.I
Charact
Extent
IN
IN
B
IN
IN
eristics
Dimensional It
X
X
X
X
X
Ora
Number
1
1
1
1
1
In Conflguri
Penetratlo
FP
FP
FP
FP
FP
itlon
Lengt
IN
IN
IN
IN
IN
Time-
frame
TV
TV
TV
TV
TV
Program
Output
IF.D
IF.O
IF.O
IF.O
IN.D
Calcu-
lator
Type
TI-59
TI-59
TI-59
TI-59
HP-41
Reference
Koch and Associates (1)
Koch and Associates (1)
Koch and Associates (1)
Koch and Associates (1)
Olsthoom (4)
(1) Programs available as of October 1983 from Koch and Associates, 1660 S. FUliwre Street, Denver, Colorado. 80210
(2) Programs available as of August 1983 from James S. Ulrlck and Associates, 2100 Los Angeles Avenue. Berkeley. California, 94707
(3) Program can consider four boundary conditions for drain: constant head, constant flux, linearly varying head and
linearly varying flux(Edelmn cases).
(4) Programs available «s of Hay 1984 from the International Ground Hater Modeling Center, Hoi comb Research Institute,
Butler University, 4600 Sunset Avenue, Indianapolis. Indiana, 46208
(continued)
-------
TABLE 5.2 (continued)
to
I
oo
LEGEND: C - Confined
WT - Water Table
H - Homogeneous
I - Isotropic
IN - Infinite
B - Bounded
X - Longitudinal
Y - Lateral
FP - Fully Penetrating
F - Finite
IN - Infinite
TV - Time Varying
SS - Steady State
0 - Drawdown
IF - Inflow
-------
TABLE 5.3 INVENTORY OF SELECTED HAND-HELD CALCULATOR PROGRAMS FOR
GROUND-WATER MOUNDING ESTIMATION
Program Title
Analysis of Ground Water
Mounding Beneath
Tailings Ponds
Circular Recharge Area
M Circular Basin Recharge
1 Mound
CO
Type
WT
WT
WT
Aquifer (
Properties
H.I
H.I
H.I
:haract<
Extent
IN
IN
IN
Bristles
Dimensional ity
R
R
R
Pond
Configuration
CI
CI
CI
Time-
frame
TV
TV
TV
Program
Output
HH
HH
HH
Calculator
Type
TI-59
TI-59
HP-41
Reference
Sandberg et al . 1981
Prickett and Vorhees
1981
Walton 1983
Ulrick (1)
(1) Programs available as of August 1983 from James S. Ulrick and Associates, 2100 Los Angeles Avenue, Berkeley, California, 94707
LEGEND: WT - Water Table
H - Homogeneous
I - Isotropic
IN - Infinite
R - Radial
CI - Circular
TV - Time Varying
HH - Hydraulic Head
-------
using the Thornthwaite method, but does not separate
infiltration and runoff. Another method would have to be used
to obtain an estimate of infiltration for landfills.
Available programs for contaminant transport are listed in
Table 5.4. The basic assumptions and limitations regarding
aquifer characteristics are similar to those for the other
types of programs. The available programs fall into two
groups. The first group includes those programs based on the
simple numerical technique discussed in Subsection 4.9. This
technique involves the tracking of particle movement over time
in response to injection/pumping wells and regional
ground-water flow. The only transport process considered in
these programs is advection. The output of programs in this
group is particle location with time. The second group of
programs includes those based on analytical solutions. These
programs consider advection and dispersion, and in some cases,
retardation and degradation. The output of these programs is
contaminant concentration at selected locations and points in
time. In situations where a programmable calculator is not
available, a nomograph-based solution to the
convection-dispersion equation can be used (see Donigian
et al., 1983).
All of the available programs are for point sources and sinks
(i.e., wells). None of them consider line sources and sinks
(i.e., drains) or area sources (i.e., ponds and landfills)
explicitly. A line of wells is often used to aproximate a
drain. A cluster of wells if often used to represent an area
source. Some of the programs consider regional ground-water
flow.
5.3 AVAILABLE PROGRAMS FOR MICRO-COMPUTERS
Access to micro-computers is increasing rapidly within
consulting firms and governmental agencies involved in the
evaluation of remedial action performance. These tools
provide capabilities that go far beyond those available on
programmable, hand-held calculators and which previously were
available only on mini-computers or large mainframes. Many
micro-computers are capable of running reduced versions of
some of the more sophisticated numerical models used to study
ground-water flow. For instance, there are a number of
versions of the Prickett-Lonnquist model (Prickett and
Lonnquist, 1971) that can be run on micro-computers.
Recognizing the benefits of micro-computers, many of the
programs written for hand-held calculators have been expanded
and programmed for use on a number of micro-computers. Many
new programs that take advantage of the computational
2-87
-------
TABLE 5.4 INVENTORY OF SELECTED HAND-HELD CALCULATOR PROGRAMS FOR
CONTAMINANT TRANSPORT
Program Title
Advective Mass Transport -
Theis Particle Hover
Streamlines and Travel
Times for Regional
Ground -Water Flow
affected by Sources
and Sinks
Advective Transport Model
co Advectlon and Dispersion •
co Regional Flow
Ground Water Dispersion
Plume Management Model
Calculator Code for Evalu-
ating Landfill Leachate
Plumes
Dissipation of a Concen-
trated Slug of Contami-
nant
Advectlon and Dispersion
from a Stream
Advection and Dispersion
from a Single Pumping
Well
Type
C
C
C
C
C
C
C
C
C
C
Aquifer
Properties
H.I
H.I
H.I
H.I
H.I
H.I
H.I
H.I
H.I
H.I
Charact
Extent
IN
IN
IN
IN
IN
IN
IN
IN
IN
IN
eristics
Dimensionality
X-Y
X-Y
X-Y
X-Y
X-Y
X-Y
X-Y
X-Y
X
R
Transport
Processes
AD
AD
AD
AD. DS.
RD, DG
AD. DS,
RD, DG
AD, DS,
RD, DG
AD, DS,
RD. DG
AD. DS,
RD. DG
AD. DS
AD, DS
Tlmeframe
TV
TV
TV
TV
TV
TV
TV
TV
TV
TV
Program
Output
PL
PL
PL
CN
CN
CN
CN
CN
CN
CN
Calculator
Type
TI-59
HP-41
HP-41
TI-59
TI- 58/59
TI-59
TI-59
TI-59
TI-59
TI-59
Reference
Sandberg et al . 1981
Prlckett and Vorhees
1981
Olsthoorn (4)
Ulrick (5)
Walton 1983
Kelly 1982
Sandberg et al . 1981
Prlckett and Vorhees
1981
Pettyjohn et al . 1982
T.A. Prlckett and
Associates (6)
Walton 1983
Walton 1983
(continued)
-------
TABLE 5.4 (continued)
Program Title
Advectlon and Dispersion
from a Single Solute
Injection Well
S-Paths
Aaulfer Characteristics
Type
C
C
Proper Ites
H,I
H.I
Extent
IN
IN
Dimensionality
R
X-Y
Transport
Processes
AD. DS
AO. RD.
DG
Tlraefrarae
TV
TV
Program
Output
CN
ML
Calculator
Type
HP-41
HP-41
Reference
Van der Heljde (4)
Oberlander
and Nelson 1984
i
CO
VO
(1) Considers 23 Injection or pumping wells
(2) Considers 63 Injection or pumping wells
(3) Considers 45 Injection or pumping wells
(4) Programs available as of May 1984 fro* the International Ground Water Modeling Center, Holcomb Research Institute.
Butler University. 4600 Sunset Avenue. Indianapolis, Indiana. 46208
(5) Programs available as of August 1983 from James S. UlMck and Associates, 2100 Los Angeles Avenue, Berkeley. California. 94707
(6) Programs available as of July 1983 from Thomas A. PMckett and Associates. Inc., 8 Montclalr Road, Urbana, Illinois, 61801
LEGEND: C - Confined
H - Homogeneous
I - Isotroplc
IN - Infinite
X - longitudinal
Y - Lateral
R - Radial
AD - Advectlon
DS - Dispersion
RD - Retardation
DG - Degradation
TV - Time Varying
PL - Particle Location with Time
CN - Concentration
ML - Mass Loading
-------
capabilities of micro-computers have been written as well.
In expanding existing programs or writing new ones, many
developers have also taken advantage of the interactive
features inherent in micro-computers. The programs have been
designed to query the user for input data and to generate
different types of graphical outputs. These added features
greatly enhance the usefulness of the programs and further
reduce the level of effort required to perform an analysis.
Representative programs for well hydraulics are summarized in
Table 5.5. The basic characteristics of these programs are
similar to those in Table 5.1. The main difference is in the
number of wells that can be considered. As Table 5.5 shows,
programs are available for a number of different types of
micro-computers. Walton (1984 b and c) has recently developed
a series of programs for several pocket and desk top
micro-computers; many of these programs are the same as those
programs listed in Table 5.5. Documentation on the programs
can be obtained from the International Groundwater Modeling
Center.
Only one program was found to be available for drain
hydraulics. It is called Mine Hydrology (FINITE). It is for
a confined, homogeneous, isotropic aquifer of infinite extent.
Up to 20 fully penetrating, finite length 'line sinks can be
considered. The program predicts time-varying drawdown and
inflow. It is available from Koch and Associates in Denver,
Colorado for the TRS-80, Apple, IBM-PC and Osborne computers.
Table 5.6 shows the available programs for mounding. These
programs are equivalent to those in Table 5.3. They are based
on the theory of either Hantush (1967a) or Glover (I960).
Similar programs for other micro-computers can be found in
Walton (1984 b and 1984 c).
The available programs for transport are summarized in Table
5.7. Of particular interest are the programs titled "Plume
Cross Section" and "Random Walk." The former provides a
method for evaluating the vertical mixing of a contaminant
plume. The latter is a micro-computer version of the
transport model developed by Prickett et al. (1981). Both
programs offer capabilities that are not available in existing
hand-held calculator programs. Again, similar programs can be
found in Walton, (1984 b and 1984 c).
All of the programs listed in Table 5.7 are for
micro-computers. It should be noted, however, there are also
a number of available programs based on analytical or
semi-analytical transport methods that can be run on
mini-computers and large mainframes. Since they are
analytical or semi-analytical, they generally require limited
2-90
-------
TABLE 5.5 INVENTORY OF SELECTED MICRO-COMPUTER PROGRAMS FOR WELL HYDRAULICS
Program Title
General Aquifer
Analysis (THEIS)
^ THWELLS
1
VD GWFLOW (3)
M
Nonsteady State
Nonleaky Arte-
sian - Single
Production Well
Nonsteady State
Nonleaky Arte-
sian - Partially
Penetrating Wells
Leaky Aquifer
Analysis (LEAKY)
Steady State Leaky
Artesian - Single
Production Well
Nonsteady State
Leaky Artesian -
Single Production
Well
Type
C
C
C,L
C
C
L
L
L
Aquifer C
Properties
H,I
H.I
H,I
H,I
H,I
H.I
H,I
H,I
haractt
Extent
IN
IN
IN
IN
IN
IN
IN
IN
'ristics
Dimensional ity
X-Y
X-Y
X-Y
R
R
X-Y
R
R
Well Cc
Number
100
-
1
1
1
100
1
1
nfiguration
Penetration
FP
FP
FP.PP
FP
PP
FP
FP
FP
Timeframe
TV
TV
TV
TV
TV
TV
SS
TV
Program
Output
0
D
0
0
D
D
D
D
Computer
Type
TRS-80
Apple
IBM-PC
Osborne
(2)
(2)
TRS-80
(5)
TRS-80
(5)
TRS-80
TRS-80
(5)
TRS-80
(5)
Reference
Koch and Associates (1)
IGWMC (4)
IGWMC (4)
Walton 1983
Walton 1983
Koch and Associates (1)
Walton 1983
Walton 1983
(continued)
-------
TABLE 5.5 (continued)
Program Title
Nonsteady State Two
Mutually Leaky
Artesian Aquifers -
Single Production
Well
Type
L
Aquifer (
'roper ties
H.I
:haracte
Extent
IN
rlstics
Dimensional Ity
R
Well Cc
Number
1
>nf iguration
Penetration
FP
Timeframe
TV
Program
Output
D
Computer
Type
TRS-80
(5)
Reference
Walton 1983
I
vo
to
(1) Programs available as of October 1983 from Koch and Associates. 1660 S. Flllmore Street. Denver. Colorado. 80210
(2) All programs from IGWMC available for Os borne, Kaypro, Superbraln and IBM
(3) GWFLOW Is a series of eight flow solutions, Including one for mounding estimation
(4) Programs available as of May 1984 from the International Ground Water Modeling Center, Holcomb Research Institute,
Butler University, 4600 Sunset Avenue, Indianapolis, Indiana, 46208
(S) Osborne, Kaypro, Superbraln, IBM, Radio Shack PC-1 and PC-2, and Sharp PC 1250 and 1500 programs available from
IGUMC; many of the program can handle multiple wells
LEGEND: C - Confined
L - Leaky
H - Homogeneous
I - Isotroplc
IN - Infinite
X • Longitudinal
Y - Lateral
R - Radial
FP - Fully Penetrating
PP - Partially Penetrating
TV - Time Varying
SS - Steady State
D - Drawdown
-------
TABLE 5.6 INVENTORY OF SELECTED MICRO-COMPUTER PROGRAMS FOR
GROUND-WATER MOUNDING ESTIMATION
Program Title
Circular Recharge
Area
Mounding
M
1
>£>
U)
Glover's Solution
Type
HT
MT
WT
Aquifer (
Properties
H.I
H.I
H,I
:haracte
Extent
IN
IN
IN.B
ristics
Dimensional 1ty
R
R
R
Pond
Configuration
CI
CI
RC
Tlmeframe
TV
TV
TV
Program
Output
HH
HH
HH.DS
Computer
Type
TRS-80
(1)
Appl e
Kaypro II
Victor
Vector
Apple
Reference
Walton 1983
NCGWR (2)
'to 1 den, Sunada and
Warner (1984)
(1) Osborne, Kaypro, Superbraln, IBM, Radio Shack PC-1 and PC-2, and Sharp PC 1250 and 1500 programs available from IGHMC.
(2) Programs available as of October 1983 from the National Center for Ground Water Research, Oklahoma State University,
Stlllwater. Oklahoma
LEGEND: WT - Water Table CI - Circular
H - Homogeneous RC - Rectangular
I - Isotropic TV - Time Varying
IN - Infinite HH - Hydraulic Head
B - Bounded DS - Discharge
-------
TABLE 5.7 INVENTORY OF SELECTED MICRO-COMPUTER PROGRAMS FOR
CONTAMINANT TRANSPORT
Program Title
Advectlon and Dis-
persion - Region-
al Flow
MAP PLUME
to
1
vo
*> PLUME
PLUME
PLOSBMB
PLUME CROSS-
SECTION
RANDOM WALK
RANDOM WALK
Type
C
C
C
C
C
C
C.L.
WT
C.L,
WT
Aquifer
Properties
H,I
H.I
H.I
H.I
H.I
H.I
H.I
H.I
'haract
Extent
IN
IN
IN
IN
IN
IN
IN
IN
eristics
Dimensional i ty
X-Y
X-Y
X-Y
X-Y
X-Y
X-Z
X-Y
X-Y
Transport
Processes
AD. OS,
RD, DG
AD, DS.
RD, DG
AD, DS,
RD, DG
AD. DS,
RD, DG
AD, DS,
RD, DG
AD, DS,
RD, DG
AD, DS,
RD, DG
AD, DS,
RD, DG
Tlmeframe
TV
TV
TV
TV
TV
TV
TV
TV
Program
Output
CN
CN
CN
CN
CN
CN
CN
CN
Computer
Type
TRS-80
(1)
Apple
Kaypro II
Victor
Vector
(3)
Sharp -
PCI 500
Osborne
Apple
Kaypro II
Victor
Vector
\pple
Kaypro II
Victor
Vector
TRS-80
Sharp -
PC 1500
Super-
jraln
)s borne
Sharp -
PCI 500
Reference
Walton 1983
NCGWR (2)
IGWMC (4)
NCGHR (2)
Voorhees (5)
NCGWR (2)
NCGWR (2)
IGWMC (4)
(continued)
-------
TABLE 5.7 (continued)
Program Title
RWOSBMB
RWMY
Advection and Dis-
persion from a
Stream
Advection and Dis-
persion from a
Single Pumping
Well
1
Type
C
C.L,
WT
C
C
Aquifer (
Properties
H,I
H,I
H,I
H.I
:haract<
Extent
IN
IN
IN
IN
eristics
Dimensional ity
X-Y
X-Y
X
R
Transport
Processes
AD, DS,
RD, DG
AD, DS,
RD, DG
AD. DS
AD, DS
Timeframe
TV
TV
TV
TV
Program
Output
CN
CN
CN
CN
Computer
Type
Osborne
Osborne
TRS-80
(1)
TRS-80
(1)
Reference
Voorhees (5)
Voorheos (5)
Walton 1983
Walton 1983
(1) Osborne, Kaypro, Superbraln, IBM, Radio Shack PC-1 and PC-2, and Sharp PC 1250 and 1500 programs available
<-" from IGWMC
(2) Programs available as of October 1983 from the National Center for Ground Water Research, Oklahoma State
University, Stillwater, Oklahoma
(3) All programs from IGWMC available for Osborne, Kaypro, Superbraln and IBM
(4) Programs available as of May 1984 from the International Ground Water Modeling Center, Holcomb Research Institute,
Butler University, 4600 Sunset Avenue, Indianapolis, Indiana, 46208
(5) Programs available as of November 1983 for Dr. Michael L. Voorhees of Warzyn Engineering, Inc., Madison, Wisconsin
LEGEND: C - Confined
H - Homogeneous
I - Isotropic
IN - Infinite
X - Longitudinal
Y - Lateral
Z - Vertical
R - Radial
AD - Advection
DS - Dispersion
RD - Retardation
DG - Degradation
TV - Time Varying
PL - Particle Location with time
CN - Concentration
-------
input data and computation time. Examples of available
programs include: 1) AT123D, a model by Yeh (1981); 2) the
GROUND and GRDFLX programs by Codell et al. (1982); 3) a
series of programs provided in Javandel et al. (1984); 4) the
PATHS Groundwater Hydrologic Model by Nelson and Schur (1980);
and 5) a computer code for evaluating landfill leachate plumes
by Pettyjohn et al. (1983). The latter is equivalent to the
calculator program by Pettyjohn et al. shown in Table 5.4 and
the PLUME, MAP PLUME and PLUME CROSS SECTION programs by the
National Center for Ground Water Research (NCGWR) shown in
Table 5.7.
2-96
-------
SECTION 6
EXAMPLE APPLICATIONS
6.1 OVERVIEW
The intent of this section is to demonstrate how some of the
analytical methods discussed in Section 4 can be used to
evaluate several typical remedial action alternatives. These
evaluations are for hypothetical sites, some of which have
been patterned after actual uncontrolled disposal sites.
6.2 EXAMPLE 1: WATER TABLE SUPPRESSION WITH AN
INTERCEPTOR TRENCH
The first example application is for an uncontrolled hazardous
waste landfill, portions of which are periodically inundated
by ground water. While inundated, considerable quantities of
waste materials leach directly into the saturated zone and
migrate downgradient to a nearby river. Data collected during
site characterization suggest that the quantity of leachate
generated by the landfill can be greatly reduced if ground
water is not allowed to come into contact with the waste
materials. An initial screening of remedial action
technologies during the Engineering Feasibility Study shows
that the ground-water table can be suppressed by pumping or
through the installation of an interceptor trench. If these
technologies are located downgradient from the site, they have
the added advantage of providing a means of actively removing
those contaminants that are already in the saturated zone.
This example application will focus on the analysis of the
trench alternative.
In evaluating the feasibility of an interceptor trench it is
important to first examine the characteristics of the local
ground-water system. Figure 6.1 is a vertical cross-section
taken through the landfill along the major direction of
ground-water flow. It shows that approximately the bottom 15
feet of the landfill are in contact with the water table.
This aquifer is composed of a relatively thin alluvial layer
2-97
-------
to
l
CD
100
120
140
160
180
i i i i i i i i i i~ i ~i ,r i ' i
TT i
I , I , I , I , I , I 7L
i.i.i
ill l r
I I I I Q
III
I I I I I I
i T T
TTTTT
i. i
III
i.i.i
'iVj.1
I I IT
i i i r
200
400
600
800
1000
1200
14OO
1600
DISTANCE (ft)
Figure 6.1 Vertical cross-section through landfill
-------
overlying a thicker layer of dolostone. At the base of the
water table aquifer is a leaky shale layer which confines an
artesian aquifer of regional extent. Peizometric heads in the
artesian aquifer are typically 10 feet higher than those in
the water table aquifer. Figure 6.1 also shows that there is
a regional recharge rate of 5 in./yr. Saturated hydraulic
conductivities and thicknesses for each layer are listed in
Table 6.1.
The major factors of importance in evaluating an interceptor
trench are: 1) what drawdown is required to suppress the
water table below the base of the landfill and 2) what
withdrawal rate is required to maintain the desired drawdown.
The first factor is of importance because it may determine
whether or not an interceptor trench can be used. If the
required drawdown is large, it may not be possible to install
a trench due to construction limitations. The second factor
is of importance because the quantity of ground-water
withdrawn by the trench needs to be known to design an on-site
treatment system and to determine whether or not to reinject
the treated water.
Given the above factors and the characteristics of the
ground-water system, there are several drain hydraulics
methods that can be used. Tables 4.2 and 4.3 show that
analytical solutions for flow to drains in leaky, water table
aquifers with recharge have been derived by Bear (1979) and
Huisman (1972); Bear provides a steady-state solution for a
partially penetrating drain and Huisman provides a transient
solution for a fully penetrating drain.
Both solutions assume homogeneous and isotropic conditions.
Since the water table aquifer is composed of two different
materials, an aquifer with an equivalent hydraulic
conductivity has to be generated. This is accomplished by
using the equivalent section methods discussed in Subsection
4.7. The hydraulic conductivity for an equivalent homogeneous
aquifer can be estimated with Equation 4.30. Using the
hydraulic conductivities for the alluvium and upper dolostone
layers given in Table 6.1, an equivalent conductivity of 12
ft/day is obtained.
Using this equivalent hydraulic conductivity and the rate of
leakage through the shale layer, the solution by Bear (1979)
can be used to estimate the elevation of the drain required to
suppress the water table below the waste; the rate of leakage
is estimated from the thickness and hydraulic conductivity of
the shale and the head difference between the artesian and
water table aquifers. Figure 6.2 shows how the required depth
of the drain varies with distance downgradient from the
landfill. The closer it is located to the landfill, the
shallower the required depth.
2-99
-------
TABLE 6.1 SATURATED HYDRAULIC CONDUCTIVITIES AND THICKNESSES
FOR EXAMPLE I
Saturated
Hydraulic Conductivity Thickness
Material Type (ft/day) (ft)
Alluvium 33.0 20
Dolostone 6.6 80
Shale 0.016 10
2-100
-------
STEADY STATE ANALYSIS
40
e
•*•«•
z
cc
o
30
Q
UJ
cc
UJ
cc
o
Q
CC
O
20
10
I
100 200 300 400 500
DISTANCE FROM FILL TO DRAIN (ft)
Figure 6.2
Depth of drain as a function of downgradient
distance from the landfill.
2-101
-------
The solution by Bear (1979) can also be used to estimate the
steady-state water table profile over the entire
cross-section, including between the drain and the river. The
procedure first involves evaluating the elevation of the water
table between a point located far upgradient from the landfill
and the drain. Once the elevation of the drain is known and
given the elevation of the river, the remainder of the profile
can be evaluated. Figure 6.3. shows the calculated profile
for a drain located 200 feet downgradient at a depth of 25
feet. This profile shows that the installation of the drain
will reverse the direction of the ground-water movement
between the drain and the river. This reversal will have a
beneficial affect in that any contaminants in this region will
be captured by the drain. It could also produce a negative
effect, however, since water will now be withdrawn from the
river.
Darcy's Law can be used to estimate the flow into the drain on
a per unit length basis. Given the equivalent hydraulic
conductivity calculated above, the depth of the water table
aquifer and the water table gradient, the flow to the drain
was estimated to be 0.47 gpm/ft. Approximately 62 percent of
the flow comes from the river.
The time required for the water table to be fully suppressed
can be estimated using the solution by Huisman (1972). While
this solution is for a fully penetrating drain, it should
provide reasonable results as long as the trench is more than
two times the saturated thickness away from the landfill. In
Subsection 4.3, it was noted that the effects of partial
penetration can be neglected beyond this distance. Figure 6.4
shows the change of elevation of the water table with time for
a drain located 200 feet downgradient from the facility and 25
feet below the water table. After approximately 65 days the
water table is suppressed below the required depth of 15 ft.
6.3 EXAMPLE 2: PLUME CAPTURE WITH A PUMPING/INJECTION
DOUBLET
This example is for the release of methylene chloride from an
abandoned waste storage lagoon. One year following the
release high concentrations are detected at a drinking water
well. The release occurred into a relatively permeable water
table aquifer composed of silty-sand. The lagoon is
essentially square with sides of 330 ft in length. Figure 6.5
shows characteristics of the aquifer and the current extent of
the methylene chloride plume as determined through a detailed
sampling program.
2-102
-------
lllllllllllll.il!
i i i i i i i i i i i i i i i i i i i i
NJ
I
M
O
U)
80
100
160
180
ill
I I I I I I
I , I , I
I I I I I"
I , \, FT I 71 , I
I .1,1
I . I . I . I . I . i
I . I . I . I . 1 . I . I . I
riiiiiiiiiiiiiii.ii.il
111
i i i i
1.1.1.1.1,1.1.171"1,1,1,1,1,1,1,1,1,1,1,i ,i,1,1,1,1. i.i ,i ,i
140- 4- I I I I I I
i i i i- i i i i i i i i rr
i i i i • i i i i i i i i i i i
E±r
i i 1 1 i
i 1 1 i 1 1 i
i 1 1 i
"FT
200
400
600
8OO
1000
1200
1400
1600
DISTANCE (ft)
RIVER
Figure 6.3 Steady-state water table profile for a partially pene-
trating drain located 200 ft downgradient from the
landfill.
-------
20 -
TRANSIENT ANALYSIS
REQUIRED DRAWDOWN
OBTAINED AFTER 65 DAYS
65 100
200
TIME
300
(days)
400
Figure 6.4
Change in water table elevation below the
landfill following installation of the
drain.
2-104
-------
O
(n
v=0.1 ft/day
K= 10 ft/day
b=40ft
METHYLENE
CHLORIDE PLUME
Figure 6.5 Aquifer characteristics and current extent of methylene
chloride plume.
-------
In evaluating available remedial action technologies, a
pumping/injection doublet was identified as a potentially
feasible alternative. The intent of the doublet would be to
create a ground-water divide that would completely encompass
the plume. This can be accomplished by locating the pumping
well downgradient from the lagoon and the injection well just
upgradient of the lagoon. If a line connecting the two wells
is parallel to the major direction of flow an envelope similar
to that shown in Figure 4.11 will be created. All water and
contamination within the envelope will be directed towards the
pumping well. The exact configuration of the envelope depends
upon the distance between the wells, the regional ground-water
velocity, and the pumping/injection rates.
The injection/withdrawal rate and location of the doublet
wells can be estimated using Equation 4.26 or the graphical
solution in Figure 4.12; both were discussed in Subsection
4.6. As Figure 6.5 shows, the maximum width of the plume is
currently 560 ft. To ensure complete capture, an envelope of
approximately 660 ft in width is assumed to be required. This
translates to a value of 660 -j- 2 or 330 ft for c in Equation
4.26 or in Figure 4.12.
The other dimension that is required is the overall length of
the envelope. This length is determined by the distance
between the injection and pumping wells and the
pumping/injection rate. The most appropriate location for the
injection well is just upgradient from the lagoon. The
pumping well can be located in the plume, but has to be near
enough to the leading edge of the plume to ensure its capture.
The exact location involves an iterative procedure wherein the
distance between the wells is selected and a pumping/injection
rate is calculated using Equation 4.26 or Figure 4.12; a is
the parameter corresponding to one-half the distance between
the wells. This rate is then used to estimate the distance,
x , from the pumping well to the stagnation point or edge of
the ground-water divide using the relationship for a single
pumping well in uniform flow (Equation 4.24). If this
distance is not long enough to ensure plume capture, the
pumping well needs to be moved closer to the edge of the
plume. Using this procedure a distance of 330 ft between
wells and a pumping/injection rate of about 27 gpm was found
to be adequate. Figure 6.6 shows the dimensions of the
ground-water divide using this distance and pumping rate.
2-106
-------
GROUND-WATER DIVIDE _ —A
I
M
O
LIMITS OF
CONTAMINATED
WATER
\
\
\
\
Figure 6.6 Dimensions of ground-water divide for a pump-
ing/injection rate of 27 gpm.
-------
6.4 EXAMPLE 3: GROUND-WATER PUMPING WITH AND WITHOUT AN
IMPERMEABLE BARRIER
In this example a number of abandoned underground storage
tanks were found to have lost their contents over a period of
years. A detailed field sampling program found that the
ground-water system was extensively contaminated. Figure 6.7
shows the location of the plume and the characteristics of the
ground-water system.
The screening of remedial actions during the Engineering
Feasibility Study suggested that the plume could be captured
with a line of pumping wells located near the leading edge of
the plume. It also suggested that an impermeable barrier
completely surrounding the plume might act to expedite the
clean-up action of the pumping wells. Thus, it was decided to
analyze the time required for plume extraction with and
without an impermeable barrier.
The technique selected for conducting the analysis was the
simple numerical technique (i.e., particle mover method)
discussed in Subsection 4.9. This technique involves tracking
the movement of a particle of water with time. The rate and
direction of particle movement are controlled by the
pumping/injection action of wells and the regional
ground-water flow.
The initial well configuration selected for analysis was a
line of three pumping wells located 100 ft upgradient from the
leading edge of the plume. Each well was assumed to be pumped
at a rate of 20 gpm.
Particles were released from a number of locations along the
perimeter of the plume. Their movement was tracked over time
until each particle arrived at one of the wells. The location
of each particle at the end of selected time increments was
noted. These locations were then used to estimate the
approximate location of the perimeter of the plume at the end
of each time increment.
Figure 6.8 shows the position of the plume 0, 10, 20, 40, 80
and 120 days after the initiation of pumping. The results
show that it takes approximately the same length of time for
contaminants to travel from the storage tanks to the wells as
it does for contaminants to travel from the leading edge of
the plume to the wells. This is because the net ground-water
velocity upgradient of the wells is the sum of the regional
velocity and the velocity induced by the pumping action of the
wells. Downgradient the net velocity is smaller because it is
2-108
-------
PERIMETER OF PLUME
I
M
O
REGIONAL
PORE WATER
VELOCITY*
0.33 ft/day
HYDRAULIC CONDUCTIVITY =10 ft/day
SATURATED THICKNESS = 40 ft
GRADIENT = 0.01ft /ft
UNDERGROUND
STORAGE TANKS
100 feet
Figure 6.7 Plume location and aquifer characteristics for Example 3
-------
Pumping Well
Figure 6.8 Plume position 0, 10, 20, 40, 80 and 120 days after initiation of pumping.
-------
the difference between the two velocities.
The results also show that most of the plume can be captured
in about 120 days. It is important to recognize, however,
that this timeframe is based on the assumption that the
contaminants are not retarded in their movement relative to
the movement of the ground water. The timeframes in this
example would have to be extended for contaminants that are
retarded by using an appropriate retardation factor. It is
also important to recognize that this analysis neglects the
effects of dispersion. The perimeter of the plume is assumed
to behave like a "sharp front."
The impact of installing an impermeable barrier around the
plume was examined with the same technique. Figure 6.9 shows
the resulting configuration of the remedial action
alternative.
The method of images was used to simulate the impact of
installing the barrier. The analysis was simplified somewhat
by only considering the effects of the upgradient and
downgradient portions of the barrier. The sides were
neglected. Figure 6.10 shows the image well configuration
used to create these barriers. Since they are assumed to be
infinite in extent, the "real aquifer" (i.e., the portion of
the aquifer inside the barrier) has the configuration of a
semi-infinite strip. A complete representation could be
obtained by using a more complex image well configuration
similar to that shown in the lower portion of Figure 4.15.
Particles were again released from the perimeter of the plume
and their movement towards the recovery wells was tracked with
time. Since the barrier eliminates the regional ground-water
flow component, it was assumed that the pumping rate of the
wells could be reduced. The wells were left in the same
location. Figure 6.11 shows the estimated position of the
plume after 0, 10, 20, 40, 80, 120, 160, 320, 480 and 640
days. These results show that the barrier wall reduces the
time required to capture contaminants downgradient from the
wells, but increases the time to capture contaminants between
the wells and the storage tanks. In part, this is due to the
use of a reduced pumping rate. However, it is also due to the
fact that there is no regional component of velocity within
the impermeable barrier. The velocity due to the pumping
wells is all that is affecting contaminant movement. This
velocity is very small near the facility.
The efficiency of several other well and impermeable barrier
configurations were also evaluated using this approach.
Figure 6.12 summarizes the results for four different
alternatives. Alternative 1 is the initial configuration
(i.e., no barrier and 3 wells pumping 20 gpm). Alternative 2
2-111
-------
FULLY PENETRATING
IMPERMEABLE BARRIER
K>«—100 ft->H—100 ft
100 feet
Figure 6.9 Impermeable barrier configuration.
-------
NJ
i
4-1
<**•
0|
o
CM
UPGRADIENT
1 IMPERMEABLE
1 RARRIFn k.
1 Dnnnicn w
\
\
\
-------
to
I
Figure 6.11 Plume position 0, 10, 20, 40, 80, 120, 160, 320 and 640 days after
initiation of pumping.
-------
01
100
80
OC
Ul
§ 60
Ul
oc
UJ 40
oc
UJ
flu
20
Alt.1
Alt. 2
Alt.3
Alt. 4
I
Pumping Well
Barrier Rate (gpm) Location
No
Yes
Yes
Yes
20 End
10 End
10 Center
20 Center
I L
100
200 300 400
TIME (days)
500
600
700
Figure 6.12 Percent recovery as a function of time for alternative well and
barrier configurations.
-------
is for the configuration shown in Figure 6.9. In Alternatives
3 and 4, the pumping wells were moved to the center of the
plume along a line parallel to the main axis of the plume.
Figure 6.12 shows the effectiveness of each alternative in
terms of percent removal. Here, percent removel refers to the
reduction in areal extent of contamination (i.e., plume size)
relative to the initial areal extent of the contamination.
Centering the wells reduces the recovery time by a factor of 3
for Alternative 3 and a factor of 6 for Alternative 4. In
addition, Alternative 1 and Alternative 4 have approximately
the same recovery time.
A large number of calculations were required to generate the
results in Figures 6.8, 6.11 and 6.12. The Advective Particle
Mover program by Ulrick (see Table 5.4) was used to reduce the
level of effort involved in this example application.
6.5 EXAMPLE 4: RECIRCULATION SYSTEM FOR GROUND-WATER
CLEAN-UP
This example is for a cooling water pond that fails and
releases several thousand gallons of ethylene glycol solution.
The affected ground-water system is a shallow, water table
aquifer composed of silty-sand type materials.
During the site characterization effort, a biological
feasibility study conducted on water samples indicated that
the ethylene glycol could easily be degraded. Based on an
analysis of the local hydrology and geology, a remedial action
alternative was proposed. The alternative consists of a
number of well points that would withdraw contaminated ground
water. This water would then be aerated, inoculated with
bacteria and then discharged back into the pond. As the
treated water seeps downward to the saturated zone it would
flush the remaining ethylene glycol towards the well points.
The bacteria would act to degrade the residual ethylene glycol
in place. The overall configuration of the remedial action is
shown in Figure 6.13 along with the characteristics of the
aquifer.
The analysis of the remedial action involved the use of
several analytical methods. Given the proposed pumping rate
for the well points of 2 gpm, the first step was to determine
the amount of drawdown produced at each well point and halfway
between well points. The drawdown at each well point needs to
be estimated to determine whether the water table aquifer can
be treated as a confined aquifer for purposes of the analysis.
As was noted in Subsection 4.2, a method developed for
confined conditions can only be used for water table
conditions if drawdowns are small relative to the total
2-116
-------
saturated thickness. The drawdowns halfway between well
points is of importance because without sufficient drawdown,
the plume may migrate past the line of well points,
particularly if there is a significant regional ground-water
flow component.
Drawdown around a well in a water table aquifer can be
estimated using a method by Neuman (1975). This method
involves the evaluation of the unconfined well function.
Values have been tabulated in a number of ground-water
textbooks. For the conditions listed in Figure 6.13 and for a
2 gpm pumping rate, the drawdown after 30 days is 2.4 ft at
the well point. This drawdown is relatively small compared to
the total saturated thickness. In addition, for these
conditions the unconfined well function is equivalent to the
confined well function. Therefore, it is reasonable to use
methods for confined aquifers in this analysis.
Using the same method the total drawdown halfway between well
points is 1.2 ft considering the effects of superposition.
This drawdown should be sufficient to direct contamination
towards one of the well points.
The next step in the analysis is to determine whether the
treated water discharged back into the pond will create a
mound and, if so, whether it would impact the effectiveness of
the well point system. The approximate height and extent of
the mound can be estimated using the method developed by
Hantush (1967a), as described in Subsection 4.6. Since the
withdrawal system is composed of four well points, the total
flow to the pond will be 8 gpm. Using Darcy's law it can be
shown that the seepage rate through the base of the pond is
approximately 2.4 gpd/ft^ . This rate produces a fairly
extensive mound that is 4.3 ft in height just below the pond.
Mounding with this seepage rate is so extensive that no
drawdown occurs at the well points. As a result, the well
points would be totally ineffective. Additional calculations
with different pumping rates and pond seepage rates showed
that the only way to ensure that a cone of depression would
occur around the well points was to discharge only a fraction
of the total flow back in to the pond. Using the original
pumping rate of 2 gpin, it was found that if 25 percent of the
total was discharged to the pond a cone of depression would
occur. At this discharge rate the maximum height of the mound
was found to be only 1.1 ft.
Residual ethylene glycol flushed into the saturated zone by
the treated water will tend to move radially away from the
pond as a result of this mound. Thus, it is important to
determine whether the well point system can effectively
capture all of the ethylene glycol.
2-117
-------
NJ
I
oo
REGIONAL
PORE WATER
VELOCITY=
0.04 ft/day
POND
ETHYLENE
GLYCOL
PLUME
HYDRAULIC CONDUCTIVITY =8.6 ft / day
Kv/Kh = 0.1
SATURATED THICKNESS = 30ft
DEPTH TO GROUNDWATER = 10 ft
SPECIFIC YIELD = 0.2
50ft
O
WELL POINT
100 ft
Figure 6.13 Aquifer characteristics and remedial action
configuration for well point recirculation
system.
-------
The effectiveness of the well points was evaluated using the
simple numerical technique discussed in Subsection 2.8.
Particles were released from different locations along the
perimeter of the mound to determine whether they would be
captured by the well point system or whether they would be
entrained in the regional flow field. Since this technique
can only consider point sources or sinks (i.e., injection or
withdrawal wells), an area source like the pond cannot be
considered directly. Instead, the effect of the mound has to
be simulated using one or more point sources (i.e., injection
wells). Through an iterative procedure it was found that an
injection well pumping at a rate of 3 gpm could produce a head
distribution roughly equivalent to the level of mounding that
would be expected, particularly beyond a distance of 40 ft,
which is the radius of the pond and the location where the
particles would be released. Figure 6.14 shows a comparison
between the estimated water table elevations for the mound and
for the injection well.
Using this injection rate, a well point pumping rate of 2 gpm
and the aquifer characteristics given in Figure 6.13,
particles were released from different positions along the
perimeter of the mound. Figure 6.15 shows the pathway
followed by each particle and the time in days for each
particle to arrive at one of the well points. These results
show how the glycol that is flushed into the saturated zone
will initially move radially away from the pond until it is
entrained in the regional flow and then directed towards one
of the well points. These results also show that even the
ethylene glycol that is initially on the far upgradient side
of the pond will be captured by one of the outside well
points. This ethylene glycol will take about seven times
longer than that which is initially on the downgradient side
of the pond. In fact, the results show that most of the
contamination will initially be captured by the inside wells.
As a result, it may be possible to initially just treat the
water from the two inside wells and then shift to the outer
wells when the remaining ethylene glycol ultimately arrives.
Due to the large number of calculations involved in tracking
the movement of different particles, this analysis was
conducted using a programmable hand-held calculator. The
Advective Transport program by Ulrick (see Table 5.4) was
used.
6.6 EXAMPLE 5: DRAIN RECIRCULATION SYSTEM
This final example involves the release of a solvent from a
large number of drums in a waste storage yard. The release
2-119
-------
~ 1.5
z
O
H
<
>
ui
LU
UJ
_J
CQ
CC
Ul
h-
ui
O
z
1.0
.5
MOUND
INJECTION WELL
50 100 150 200
DISTANCE FROM CENTER OF COOLING WATER POND (ft)
Figure 6.14 Comparison of water table elevations for mound and
injection well.
-------
1180 DAYS
590 DAYS
PERIMETER OF
COOLING WATER
POND
WELL
Figure 6.15
Particle movement from the perimeter of
the cooling water pond to each well point.
-------
occured into an area where the ground water is near the land
surface. The regional flow is negligible in the local area.
Access to areas beyond the perimeter of the storage yard is
limited on all sides by roads and buildings. Therefore, the
selected remedial action has to be implemented within the
perimeter of the storage yard.
In evaluating different remedial action technologies it was
decided that a recirculating drain system could be used. This
system would be composed of a fully penetrating interceptor
trench installed along one side of the storage yard. Water
withdrawn from the drain would be treated on-site and then
reinjected through another drain located along the other side.
This drain would create a mound that would direct the solvent
towards the recovery drain.
The flow system created by the installation of such a drain
system was evaluated using a hand-held calculator program that
estimates the drawdown around line sources and sinks of finite
length. The Line Sink program developed by Ulrick (see Table
5.2) was used.
The principle of superposition is used in this program to
obtain the total drawdown due to multiple line sources and
sinks. Using this program, the elevation of the water table
can be evaluated rapidly at a large number of locations.
These elevations can then be used to generate equipotential
contours (i.e., contours of equal elevation). These contours
can be used to generate flow lines (i.e., the direction ground
water will move).
Figure 6.16 shows a plan view of the site. It also shows the
location of the plume and the drains. Finally, equipotential
contours and flow lines for a recovery/reinjection rate of
60 gpm are also shown.
These results indicate that most of the plume outside the
perimeter of the yard is contained within the equipotential
contour corresponding to 1 ft of drawdown. Given the
negligible regional flow in the area, this drawdown should be
sufficient to ensure that the entire plume is captured by the
recovery drain. The remainder of the plume will be directed
towards the recovery drain by the mounding action of the
reinjection drain.
2-122
-------
U)
WASTE STORAGE
YARD PERIMETER /
REINJECT1ON DRAIN
Figure 6.16 Equipotential contours (in feet) and flow
lines produced by the drain recirculation
system.
-------
REFERENCES
Baumann, P. 1952. "Groundwater Movement Controlled through
Spreading," Transactions, ASCE, Vol. 117.
Bear, J. 1979. Hydraulics of Groundwater, McGraw-Hill Inc.,
New York, NY.
Bentall, R. 1963. Shortcuts and Special Problems in Aquifer
Tests, USGS Water-Supply Paper 1545-C.
Bicknell, B.R. 1984. Modeling Chemical Emissions from
Lagoons and Landfills, Final Report, U.S. Environmental
Protection Agency, Environmental Research Laboratory,
Athens, GA.
Boulton, N. S. 1954. "The Drawdown of the Water Table under
Non- Steady Conditions near a Pumped Well in an Unconfined
Formation," Proc. Inst. Civil Engrs., 3.
Boulton, N. S. 1955. "Unsteady Radial Flow to a Pumped Well
Allowing for Delayed Yield from Storage," Proc. Gen.
Assembly Rome, Intern. Ass. Sci. Hydrol. Publ. 37.
Boulton, N. S. 1963. "Analysis of Data from Nonequilibrium
Pumping Tests Allowing for Delayed Yield from Storage,"
Proc. Inst. Civil Engrs., 26.
Boulton, N. S., and T. D. Streltsova. 1977a. "Unsteady Flow
to a Pumped Well in a Fissured Water-Bearing Formation," J.
Hydrol., 35.
Boulton, N. S., and T. D. Streltsova. 1977b. "Unsteady Flow
to a Pumped Well in a Two-Layered Water-Bearing Formation,"
J. Hydrol., 35.
Boulton, N. S., and T. D. Streltsova. 1978. "Unsteady Flow
to a Pumped Well in a Fissured Aquifer with a Free Surface
Level Maintained Constant, "Water Resources Research,
14(3).
2-124
-------
Boussinesq, J. 1904. "Recherches Theoretiques sur 1'
Ecoulement des Nappes d'Eau Infiltrees dans le Sol et sur
le debit des Sources," J. Math. Pure Appl., 10.
Bouwer, H., 1974. "Design and Operation of Land Treatment
Systems for Minimum Contamination of Groundwater," Ground
Water, 12(3).
Brooks, R. H. 1961. "Unsteady Flow of Groundwater into Drain
Tile," ASCE. Proc. 87.
Brutsaert, W., and M. Y. Corapcioglu. 1976. "Pumping of
Aquifer with Viscoelastic Properties," Proceedings American
Society of Civil Engineering, Journal of the Hydraulics
Division, 102 (HY11).
Chauhan, H. S., G. 0. Schwab and M. Y. Hamdy. 1968.
"Analytical and Computer Solutions of Transient Water
Tables for Drainage of Sloping Land," Water Resources
Research, 4(3).
Cleary, R. W. and M. J. Ungs. 1978. Groundwater Pollution
and Hydrology, Mathematical Models and Computer Programs,
Water Resources Program, Report No. 78-WR-15, Princeton
University, Princeton, N.J.
Codell, R. B., K. T. Key and G. Whelan. 1982. A Collection
of Mathematical Models for Dispersion in Surface Water and
Groundwater, NUREG-0868, U.S. Nuclear Regulatory
Commission, Washington, DC.
Cohen, R. M. and W. J. Miller, III. 1983. "Use of Analytical
Models for Evaluating Corrective Actions at Hazardous Waste
Sites," Proceedings of the 3rd National Symposium on
Aquifer Restoration and Ground-Water Monitoring, National
Water Well Association, May 25-27.
Cooper, H. H., Jr., and M. I. Rorabaugh, 1963. Ground Water
Movements and Bank Storage Due to Flood Stages in Surface
Streams, USGS Water Supply Paper 1536-J.
Corapcioglu, M. Y. 1976. "Mathematical Modeling of Leaky
Aquifers with Rheological Properties," Proceedings of
Anaheim Symposium on Subsidence, Int. Assoc. Hydrol. Sci.
Publ. No. 121.
2-125
-------
Dass, P. et al. 1977. "Leachate Production at Sanitary Land-
fills," Proceedings of the American Society of Civil
Engineers, Journal of the Environmental Engineering
Division, 103 (EE6).
Donigian, A. S., Jr., T. Y. R. Lo, and E. W. Shanahan. 1983.
Rapid Assessment of Potential Ground-Water Contamination
Under Emergency Response Conditions, EPA-600/8-83-030, U.S.
Environmental Protection Agency, Office of Research and
Development, Washington, D.C.
Fenn, D. G., K. J. Hanley and T. V. De Geare. 1975. "Use of
the Water Balance Method for Predicting Leachate Generation
from Solid Waste Disposal Sites," Report SW-168, U.S.
Environmental Protection Agency, Cincinnati, OH.
Ferris, J. G., D. B. Knowles, R. H. Brown and R. W. Stallman.
1962. Theory of Aquifer Tests, USGS Water-Supply Paper
1536E.
Freeze, R. A. and J. A. Cherry. 1979. Groundwater, Prentice-
Hall, Inc., Englewood Cliffs, N.J.
Glover, R. E. 1960. Mathematical Derivations as Pertain to
Ground-water Recharge, Agricultural Research Service, USDA,
Ft. Collins, CO.
Glover, R. E. 1966. Theory of Ground Water Movement, U.S.
Bureau of Reclamation Engineering Monograph No. 13.
Glover, R. E. 1974. Transient Ground Water Hydraulics, Water
Resources Publications, Fort Collins, CO.
Grove, D. B., W. A. Beetem, and F. B. Sower. 1970. "Fluid
Travel Time between a Recharging and Discharging Well Pair
in an Aquifer Having a Uniform Regional Flow Field," Water
Resources Research, 6.
Hantush, M. S. 1956. "Analysis of Data from Pumping Tests in
Leaky Aquifers," Transactions American Geophysical Union,
37(6).
Hantush, M. S. 1959. "Non-Steady Flow to Flowing Wells in
Leaky Aquifers," Journal of Geophysical Research, 64(8).
Hantush, M. S. 1960. "Modification of the Theory of Leaky
Aquifers," Journal Geophysical Research, 65(11).
2-126
-------
Hantush, M. S. 1964. "Hydraulics of Wells," In Advances in
Hydroscience, V. T. Chow (Editor). Academic Press, New
York, N.Y.
Hantush, M. S. 1967a. "Growth and Decay of Ground water
Mounds in Response to Uniform Percolation," Water Resources
Research, 3(1).
Hantush, M. S. 1967b. "Flow to Wells in Aquifers Separated by
a Semi-Pervious Layer," Journal of Geophysical Research,
72(6).
Hantush, M. S., and C. E. Jacob. 1955. "Non-Steady Radial
Flow in an Infinite Leaky Aquifer and Nonsteady Green's
Function for an Infinite Strip of Leaky Aquifer,"
Transactions of the American Geophysical Union, 36(1).
Harr, M. E., 1962. Ground Water and Seepage, McGraw-Hill
Inc., New York, N.Y.
Hooghoudt, S. B. 1940. Bijdragen tot de Kennis van Eenige
Natuurkundige Grootheden van den Grond, 7, Algemeene
Beschouwing van het Probleem van de Detail ontwatering en
de Infiltratie Door Middel van Parallel Loopende Drains,
Greppels, Slooten, en Kanalen, Versl. Lanbouwk, Ond., 46.
Huisman, L. 1972. Groundwater Recovery, Winchester Press,
New York, N.Y.
Huisman, L. and T. N. Olsthoorn. 1983. Artificial Ground-
water Recharge, Pitman Publishing Inc., Marshfield, MA.
Hunt, B. W. 1971. "Vertical Recharge of Unconfined
Aquifers," Journal of Hydraulics Division, ASCE, 97(HY7).
Jacob, C. E. 1943. "Correlation of Ground Water Levels with
Precipitation on Long Island, N.Y.," Transactions of the
American Geophysical Union, pt. 2.
Jacob, C. E. and S. W. Lohman. 1952. "Nonsteady Flow to a
Well of a Constant Drawdown in an Extensive Aquifer,"
Transactions of the American Geophysical Union, 33(40).
Javandel, I., C. Doughty and C. F. Tsang. 1984. Ground-
Water Transport: Handbook of Mathematical Models,
Water Resources Monograph 10, American Geophysical
Union, Washington, DC.
2-127
-------
JRB Associates. 1982. Handbook of Remedial Actions at
Hazardous Waste Disposal Sites, EPA 625/6-82-006, U.S.
Environmental Protection Agency, Municipal Environmental
Research Laboratory, Cincinnati, OH.
Kelly, W. E. 1982. "Field Reports - Ground-Water Dispersion
Calculations with a Programmable calculator," Ground Water,
20(6) .
Kirkham, D. 1958. "Seepage of Steady Rainfall through Soil
into Drains," Transactions of the American Geophysical
Union, 39.
Kirkham, D. 1967. "Explanation of Paradoxes in the Dupuit-
Forchheimer Seepage Theory," Water Resources Research, 3.
Kirkham, D., S. Toksoz, and R. R. van der Ploeg. 1974.
"Steady Flow to Drains and Wells," In Drainage for
Agriculture, van Schilfgaarde, ed., American Society of
Agronomy, Inc., Madison, WI.
Knight, R. G., E. H. Rothfuss and K. D. Yard. 1980. FGD
Sludge Disposal Manual, Second Edition, CS-1515, Electric
Power Research Institute, Palo Alto, CA.
Knox, R.C. 1984. "Assessment of the Effectiveness of
Barriers for the Retardation of Pollutant Migration,"
Ground Water , 22(3).
Lai, R. Y. S., and Cheh Wu Su. 1974. "Nonsteady Flow to a
Large Well in a Leaky Aquifer," J. Hydrol., 22.
Luthin, J. N. 1973. Drainage Engineering, R. E. Krieger
Publishing Co., Huntington, NY.
McBean, E. A., R. Poland, F. A. Rovers and A. J. Crutcher.
1982. "Leachate Collection Design for Containment
Landfills," Proceeding of the ASCE, Journal of the
Environmental Engineering Division, 108(1).
McWhorter, D. B. and D. K. Sunada. 1977. Ground-Water
Hydrology and Hydraulics, Water Resource Publications,
Fort Collins, CO.
McWhorter, D. B. and J. D. Nelson. 1980. "Seepage in a
Partially Saturated Zone Beneath Tailings Impoundments,"
Mining Engineering, April.
2-128
-------
Moench, A. F. and T. A. Prickett. 1972. "Radial Flow in
an Infinite Aquifer Undergoing Conversion from Artesian to
Water Table Conditions," Water Resources Research, 9(2).
Molden, D., D. K. Sunada and J. W. Warner. 1984.
"Microcomputer Model of Artificial Recharge Using
Glover's Solution, " Ground Water, 22(1).
Moody, W. T. 1966. "Nonlinear Differential Equation for
Drain Spacing," Proceedings of the ASCE, 92.
Moody, W. T. and R. W. Ribbens. 1965. Ground Water-
Tetrama-Colusa Canal Reach No. 3, Sacramento Canal
Units, Central Valley Project, Memorandum to Chief,
Canals Branch, Bureau of Reclamation Office of Chief
Engineer, Denver, CO.
Moore, C. A. 1983. Landfill and Surface Impoundment
Performance Evaluation, SW-869, U.S. Environmental
Protection Agency, Office of Solid Waste and Emergency
Response, Washington, DC.
Nelson, R. W. and J. A. Schur. 1980. PATHS Groundwater
Hydrologic Model, PNL-3162, Prepared by Battelle, Pacific
Northwest Laboratories, Richland, WA for the Office of
Nuclear Waste Isolation, U.S. Department of Energy.
Neuman, S. P. 1972. "Theory of Flow in Unconfined Aquifers
Considering Delayed Response of the Water Table,"Water
Resources Research, 8.
Neuman, S. P. 1973. "Calibration of Distributed Parameter
Groundwater Flow Models Viewed as a Multiple-Objective
Decision Process under Uncertainty," Water Resources
Research, 9.
Neuman, S. P. 1975. "Analysis of Pumping Test Data from
Anisotropic Unconfined Aquifers Considering Delayed
Gravity Response," Water Resources Research, 11(2).
Neuman, S. P. and P. A. Witherspoon. 1969a. "Theory of
Flow in a Confined Two-Aquifer System," Water Resources
Research, 5.
Neuman, S. P. and P. A. Witherspoon. 1969b. "Applicability
of Current Theories of Flow in Leaky Aquifers," Water
Resources Research, 5.
2-129
-------
Neuman, S. P. and P. A. Witherspoon. 1972. "Field
Determination of the Hydraulic Properties of Leaky
Multiple-Aquifer Systems," Water Resources Research, 8.
Oberlander, P.L. and R.W. Nelson. 1984. "An Idealized
Ground-Water Flow and Chemical Transport Model
(S-PATHS)," Ground Water, 22(4).
Ogata, A. 1970. Theory of Dispersion in a Granular
Medium, U.S.G.S. Professional Paper 411-1.
Papadopulos, I. S. 1965. "Nonsteady Flow to a Well in an
Infinite Anisotropic Aquifer," Symp. Int. Assoc. Sci.
Hydrol., Dubrovnik.
Papadopulos, I. S. 1967. "Drawdown Distribution around a
Large-Diameter Well," Proceedings of a Symposium on Ground
Water, American Water Resources Association.
Parr, A. D., J. G. Melville and F. J. Molz. 1983. "HP41C
and TI59 Programs for Anisotropic Confined Aquifers,"
Ground Water , 21(2).
Pettyjohn, W. A., D. C. Kent, T. A. Prickett, H. E. Le Grand
and F. E. Witz. 1982. Methods for the Prediction of
Leachate Plume Migration and Mixing, Draft Report, U.S.
Environmental Protection Agency, Municipal Environmental
Research Laboratory, Cincinnati, OH.
Picking, L. W. 1979. "Field Reports - Programming a Pocket
Calculator for Solving Multiple Well, Variable Pumping
Rates Problems," Ground Water, 18(2).
Polubarinova-Kochina, P. Ya. 1952. "Theory of the Motion of
Ground Water," Gostekhizdat, Moscow.
Powers, M. A., P. P. Virgadamo and W. E. Kelly. 1981.
Groundwater Control at Hazardous Waste Sites, American
Society of Civil Engineers Meeting, St. Louis, MO.
Prickett, T. A. and C. G. Lonnquist. 1971. Selected Digital
Computer Techniques for Groundwater Resource Evaluation,
Bulletin No. 55, Illinois State Water Survey, Urbana, IL.
Prickett, T. A., T. G. Naymik and C. G. Lonnquist. 1981. A
"Random-Walk," Solute Transport Model for Selected
Groundwater Quality Evaluations, Bulletin No. 65,
Illinois State Water Survey, Urbana, IL.
2-130
-------
Prickett, T. A. and M. L. Vorhees. 1981. Selected Hand-Held
Calculator Codes for the Evaluation of Cumulative
Strip-Mining Impacts on Groundwater Resources, Prepared for
the Office of Surface Mining, Region V, Denver, CO.
Rao, N. H. and P. B. S. Sarma. 1981. "Ground-Water Recharge
from Rectangular Areas," Ground Water, 19(3).
Rayner, F. A. 1981. "Discussion of Programmable Hand
Calculator Programs for Pumping and Injection Wells: I -
Constant or Variable Pumping (Injection) Rate, Single or
Multiple Fully Penetrating Wells," Ground Water, 19(1).
Rayner, F. A. 1983. "Aquifer Modeling with a Handheld
Calculator - AQMODL," Ground Water, 21(1).
Sandberg, R., R. B. Scheibach, D. Koch and T. A. Prickett.
1981. Selected Hand-Held Calculator Codes for the
Evaluation of the Probable Cumulative Hydrologic
Impacts of Mining, Report H-D3004/030-81-1029F,
Prepared for the Office of Surface Mining, Region V,
Denver, CO.
Schoeller, H. 1959. Arid Zone Hydrology, Recent
Developments, UNESCO, Paris.
Schroeder, P.R., J.M. Morgan, T.M. Walski and A.C. Gibson.
1984a. Hydrologic Evaluation of Landfill Performance
(HELP) Model, Volume I. User's Guide for Version 1,
EPA/530-SW-84-009, U.S. Environmental Protection Agency,
Municipal Environmental Research Laboratory, Cincinnati,
OH.
Schroeder, P. R., A. C. Gibson and M. D. Smolen. 1984b.
Hydrologic Evaluation of Landfill Performance (HELP)
Model, Version II. Documentation for Version 1,
EPA/530-SW-84-010, U.S. Environmental Protection Agency,
Municipal Environmental Research Laboratory, Cincinnati,
OH.
SCS Engineers. 1982. Costs of Remedial Response Actions
at Uncontrolled Hazardous Waste Sites, EPA 600/2-82-035,
Prepared for the U.S. Environmental Protection Agency,
Municipal Environmental Research Laboratory, Cincinnati,
OH.
Singh, S. R. and C. M. Jacob. 1977. "Transient Analysis of
Phreatic Aquifers Lying between Two Open Channels,"
Water Resources Research, 13(2).
2-131
-------
Stallman, R. w. 1965. "Effects of Water Table Conditions on
Water Level Changes near Pumping Wells," Water Resources
Research, 1(2).
Streltsova, T. D. 1974. "Drawdown in Compressible Unconfined
Aquifers," Proceedings of the American Society of Civil
Engineers, Journal of the Hydraulics Division, 100(HY11).
Streltsova, T. D. 1976. "Analysis of Aquifer-Aquitard Flow,"
Water Resources Research, 12(3).
Terzidis, G. 1968. "Discussion, Falling Water Table Between
Tile Drains," ASCE Proceedings, 94.
Theis, C. V. 1935. "The Relation Between the Lowering of
Piezometrc Surface and the Rate and Duration of
Discharge of a Well Using Ground-Water Storage,"
Transactions American Geophysical Union, 16th Annual
Meeting, Part 2.
U.S. Department of the Interior. 1981. Ground Water Manual,
U.S. Department of the Interior.
van Genuchten, M. Th. and W. J. Alves. 1982. Analytical
Solutions of the One-Dimensional Convective-Dispersive
Solute Transport Equation, Technical Bulletin 1661,
USDA.
van Schilfgaarde, J. 1970. "Theory of Flow to Drains," In
Advances in Hydroscience, Vol. 6, Academic Press, New York,
NY.
van Schilfgaarde, J. 1974. "Nonsteady Flow to Drains," In
Drainage for Agriculture, van Schilfgaarde, ed.,
American Society of Agronomy, Inc., Madison, WI.
Venetis, C. 1968. "On the Impulse Response of an Aquifer,"
Bulletin of Int. Assn. of Scientific Hydrology, Vol.
13.
Walton, W. C. 1970. Groundwater Resource Evaluation.
McGraw-Hill, Inc. New York, NY.
Walton, W. C. 1979. "Progress in Analytical Groundwater
Modeling," In: W. Back and D. A. Stephenson
(Guest-Editors), Contemporary Hydrogeology - The George
Burke Maxey Memorial Volume, J. Hydrol; 43.
2-132
-------
Walton, W. C. 1983. Handbook of Analytical Ground Water
Model Codes for Radio Shack TRS-80 Pocket Computer and
Texas Instruments TI-59 Hand-Held Programmable
Calculator, GWMI 83-02/2, International Ground Water
Modeling Center, Holcomb Research Institute, Butler
University, Indianapolis, IN.
Walton, W. C. 1984a. Handbook of Analytical Ground Water
Models, GWMI 84-06/1, International Ground Water Modeling
Center, Holcomb Research Institute, Butler University,
Indianpolis, IN.
Walton, W. C. 1984b. 26 Basic Ground Water Model Programs
for Pocket Microcomputers, GWMI 84-06/3, International
Ground Water Modeling Center, Holcomb Research
Institute, Butler University, Indianapolis, IN.
Walton, W. C. 1984c. 35 Basic Ground Water Model Programs
for Desk Top Microcomputers, GWMI 84-06/4, International
Ground Water Modeling Center, Holcomb Research
Institute, Butler University, Indianapolis, IN.
Warner, D. L. and M. G. Yow. 1979. "Programmable Hand
Calculator Programs for Pumping and Injection Wells: I -
Constant or Variably Pumping (Injection) Rate, Single or
Multiple Fully Penetrating Wells," Ground Water, 17(6).
Warner, D. L. and M. G. Yow. 1980a. "Programmable Hand
Calculator Programs for Pumping and Injection Wells: II
Constant Pumping (Injection) Rate, Single Fully
Penetrating Well, Semiconfined Aquifer," Ground Water,
18(2).
Warner, D. L. and M. G. Yow. 1980b. "Programmable Hand
Calculator Programs for Pumping and Injection Wells:
III - Constant Pumping (Injection) Rate, Fully Confined
Aquifer, Partially Penetrating Well," Ground Water, 18(5).
Wesseling, J. 1964. "A Comparison of the Steady State Drain
Spacing Formulas of Hooghoudt and Kirkham in Connection
with Design Practice," Tech. Bull., Vol. 34, Inst. for Land
and Water Management Research, Wageningen.
Wilson, J.L. 1984. "Double Cell Hydraulic Containment of
Pollutant Plumes," Proceedings of the Fourth National
Symposium and Exposition on Aquifer Restoration and
Ground Water Monitoring, National Water Well
Association, Columbus, OH.
2-133
-------
Witherspoon, P. A., I. Javandel, S. P. Neuman and R. A. Freeze
1967. Interpretation of Aquifer Gas Storage Conditions
from Water Pumping Tests, American Gas Association, New
York, N.Y.
Witherspoon, P. A. and T. N. Narasimhan. 1973. "Seepage from
Recharge and Landfill Ditches," Specialty Conference on
Agricultural and Urban Considerations in Irrigation and
Drainage, Colorado Section of ASCE.
Yeh, G. T. 1981. AT123D: Analytical Transient One-, Two-,
and Three-Dimensional Simulation of Waste Transport in the
Aquifer System, ORNL-5602, Oak Ridge National Laboratory,
Oak Ridge, TN.
Youngs, E. G. 1964. "Horizontal Seepage through Unconfined
Aquifers with Hydraulic Conductivity Varying with
Depth," J. Hydrology, 3.
Youngs, E. G. 1966a. "Horizontal Seepage through Unconfined
Aquifers with Nonuniform Hydraulic Conductivity," J.
Hydrology, 4.
Youngs, E. G. 1966b. "Exact Analysis of Certain Problems of
Ground Water Flow with Free Surface Conditions,"J.
Hydrology, 4.
2-134
-------
VOLUME 3
Numerical Modeling
of Surface, Subsurface
and Waste Control Actions
-------
VOLUME 3: NUMERICAL MODELING OF SURFACE, SUBSURFACE
AND WASTE CONTROL ACTIONS
SECTION 1
INTRODUCTION
1.1 PURPOSE OF REPORT
Recent studies at several uncontrolled hazardous waste sites
have demonstrated the benefits of using numerical models to
evaluate remedial action performance. Models have been used
in the detailed analysis of alternative actions to identify
those that would be ineffective or would fail to meet site
clean-up goals. The quantitative measures of performance
derived from simulation results provide a useful basis for
comparison with other factors like remedial action costs.
Models have also been used to refine and, in some cases,
optimize conceptual designs prior to their implementation;
post-implementation modeling studies have also been conducted
to improve remedial action operation. Another beneficial use
has been in the prediction of future contamination levels for
purposes of exposure and risk assessment. Finally, the
increased level of understanding gained regarding important
processes/pathways and levels of uncertainty associated with
parameters that require additional characterization have been
a benefit to many model users.
The purpose of this volume is to provide guidance on the use
of surface, unsaturated and saturated zone models to evaluate
the performance of remedial actions. The guidance provided
herein focuses on: 1) important considerations related to the
application or use of numerical models and 2) modeling
requirements for specific remedial actions or groups of
actions. The guidance applies only to those actions commonly
implemented at hazardous waste sites, namely surface,
subsurface and waste control actions.
This volume will be of most value to two major groups: 1) EPA
and state Superfund staff and 2) certain site contractors.
EPA and state staff should gain an improved understanding of
3-1
-------
how numerical models can be used to assess remedial actions.
This information should be of particular benefit when
reviewing proposed site contractor plans to use numerical
models. This volume will aid site contractors that have
limited experience in using numerical modeling techniques.
1.2 REPORT ORGANIZATION
Brief conclusions regarding the use of numerical models for
remedial action evaluation are presented in Section 2.
Section 3 discusses the processes that act to transport,
transform and transfer hazardous waste constituents in the
local environment surrounding a hazardous waste site. Section
4 discusses specific surface, subsurface, and waste control
remedial action technologies and how their implementation
affects these processes. Both sections are meant to provide
the reader with a brief overview. They also set the stage for
Sections 5 and 6.
Section 5 discusses a number of important considerations
associated with the application or use of numerical models.
The section starts by overviewing the general capabilities of
surface, unsaturated and saturated zone models; brief reviews
of several representative models are also provided along with
sources of information on a number of other models.
Considerations related to the linkage of models for different
zones are presented for those situations involving the
analysis of relatively complicated site and remedial action
conditions. The process of "applying" a numerical model is
also presented, followed by a discussion of user expertise and
resources commonly required when using numerical models. The
section concludes by describing alternative ways of using
models to evaluate actions; a number of examples from
published modeling studies are included.
Section 6 provides modeling requirements for different
remedial actions or groups of actions. The requirements
include: 1) the type of model(s) (i.e., surface, unsaturated
or saturted zone), 2) dimensionality and grid configuration
(i.e., two-dimensional, x-y), and 3) parameter adjustments.
Where possible, parameter estimation guidance specific to a
given action is provided. Where this is not possible general
guidance is provided. The parameter guidance is meant to be
used only when site-specific data are not available.
To provide EPA and its site contractors with a modeling
capability that can be used to assess a broad range of site
and remedial action conditions, three models were selected
from the large number of available surface, unsaturated and
3-2
-------
saturated zone models. The selected models include: 1) the
Hydrologic Simulation Program-FORTRAN (HSPF) model for the
surface zone; 2) FEMWATER/FEMWASTE models for the unsaturated
zone; and 3) the Finite Element, Three-Dimensional Ground
Water (FE3DGW) and Combined Fluid, Energy and Solute Transport
(CFEST) models for the saturated zone. Each model is being
made available for use on the EPA National Computer Center
(NCC) in Research Triangle Park, N.C.
Appendix A to this report describes the rationale for
selecting each model, their capabilities, linkage
considerations, their implementation on NCC, available
documentation and user support, and specific parameters that
must be adjusted to represent selected actions.
3-3
-------
SECTION 2
CONCLUSIONS
Numerical models are finding increased use in the analysis of
remedial action performance. To date, most model applications
have been for the purpose of evaluating alternative remedial
action designs and the impacts associated with uncertain
estimates of key model parameters and assumptions. Both types
of uses have generally led to an improved understanding of
site conditions and an ability to quantitatively evaluate the
feasibility of different remedial actions. Numerical models
also have been used to a lesser extent to estimate reductions
in exposure levels associated with the implementation of
remedial actions. Such estimates have been used directly in
exposure assessment or as input to more comprehensive risk
assessments. Future uses of models include the analysis of
remedial action design life, the impacts associated with
remedial action failure and optimal remedial action design and
operation.
Limited field and laboratory data exist on the performance of
certain remedial actions. As a result, only limited guidance
can be provided on the model parameter adjustments required to
properly simulate the effects of implementing these actions.
In particular, data are lacking on: 1) in-place hydraulic
conductivities for different impermeable barrier materials; 2)
changes in chemical mobility resulting from chemical injection
and solution mining; 3) hydraulic properties and sorption
characteristics of permeable treatment bed materials; and 4)
changes in chemical susceptibility to degradation resulting
from bioreclamation. As these technologies are implemented at
different sites, laboratory and field experimental work should
be conducted to obtain data useful for future modeling
studies.
The modeling requirements for remedial actions ara highly
variable. If site contractors decide to use numerical
modeling for remedial action evaluation, the modeling
requirements for all potentially feasible actions should be
considered as early as possible in the Feasibility
Study/Remedial Investigation. Early consideration will allow
for the selection of a numerical model with the appropriate
3-4
-------
capabilities and level of sophistication. Early consideration
will also lead to more efficiency in terms of data collection
to support model application.
3-5
-------
SECTION 3
MIGRATION AND FATE PROCESSES
3.1 OVERVIEW
The local environment surrounding an uncontrolled hazardous
waste site can be subdivided into four zones, as defined below
and in Figure 3.1.
1. Atmospheric Zone: Segment of the total atmosphere
just above the land surface extending to areas
adjacent to the disposal site.
2. Surface Zone: Parcel of soil from the land surface
down to the root zone covering the waste site and the
surrounding drainage area.
3. Unsaturated Zone: Parcel of soil with boundaries
at the surface zone and the ground-water table; soil
pores may contain varying amounts of air and water.
4. Saturated Zone: Soil and rock below the ground-water
table, where all pores are filled with water and
extending down to impermeable basement rock.
There are a number of processes that act to control the
movement of contaminants within and between zones. These
processes can be grouped as follows: 1) processes
controlling movement within a zone (intra-zone), 2) processes
controlling transfers between zones (inter-zone) and 3)
processes controlling the transformation of chemicals. Table
3.1 lists the specific environmental processes that fall into
each group and defines the affected zones and key parameters
that influence each process. Figure 3.2 provides a schematic
overview of a waste site and the role that selected intra- and
inter-zone processes plySy in controlling water and waste
migration.
The following subsections provide brief descriptions of each
of the key processes listed in Table 3.1. These descriptions
are not meant to be comprehensive. Rather, they are meant to
3-6
-------
ATMOSPHERIC ZONE
U)
I
UNSATURATED ZONE
SATURATED ZONE
i^H
^*^- -^v v
Figure 3.1 Local environment zones surrounding an uncontrolled hazardous
waste site (adapted from JRB Associates, 1982) .
-------
TABLE 3.1 PROCESSES CONTROLLING THE MIGRATION AND FATE OF HAZARDOUS
WASTE CONSTITUENTS
Group
Processes
Affected Zones
Key Parameters
Intra-Zone
Advection: Runoff
Percolation
Ground-water
flow
Dispersion
Surface
Unsaturated
Saturated
Unsaturated
Saturated
Topography, vegetation, precipitation
soil moisture
Porosity, moisture content, infiltra-
tion rate
Porosity, hydraulic conductivity,
gradient
Soil/rock heterogeneity, pore size
distribution
U)
I
GO
Inter-Zone
Erosion
Sorption/Retardation
Evapotranspiration
Infiltration
Drainage
Volatilization
Transformation Photolysis
Hydrolysis, Oxidation,
Chemical Reaction
Bio-degradation
Surface
Surface
Unsaturated
Saturated
Surface to air
Surface to unsaturated
Unsaturated to saturated
Surface to air
Surface, air
All
All except air
Topography, vegetation, soil type,
precipitation
Organic content, sediment concen-
tration
Organic content, porosity, chemical
properties
Soil moisture, meteorologic condi-
tions, vegetation
Soil moisture, precipitation, soil
type, topography
Percolation rate, hydraulic conduc-
tivity, location of water table
Meteorologic conditions, chemical
properties
Meteorologic conditions.
Chemical properties, soil properties
Chemical properties, bacterial
activity
-------
I
VD
EVAPOTRANSPIRATION
GROUND-WATER FLOW
Figure 3.2 Schematic overview of a waste site and selected intra- and
inter-zone processes affecting water and waste constituent
migration (adapted from JRB Associates, 1982).
-------
provide elementary definitions of processes to set the stage
for subsequent discussions of remedial actions and the use of
numerical models to evaluate them.
It is important to note that the overall focus of this report
is on the use of models to analyze the effectiveness of
surface, subsurface and waste control remedial action
technologies, not surface water remedial action technologies.
Volume 4 discusses the use of both simplified and numerical
models to evaluate surface water technologies. Equivalent
guidance on modeling of gas migration technologies does not
currently exist.
3.2 PROCESSES CONTROLLING MOVEMENT WITHIN ZONES
3.2.1 Advection
Advection is the movement of a waste constituent as a result
of bulk water movement. Water movement in the surface zone
occurs in the form of overland flow or runoff, which can
entrain chemicals and transport them to stream channels.
Water movement in the unsaturated zone is primarily due to
percolation or vertical movement through the soil profile.
Passage of this water through waste materials can result in
the leaching of waste constituents. Lateral movement occurs
if the water reaches an impermeable strata or when the
vertical flux exceeds the saturated permeability of a given
strata. Water and associated contaminant movement in the
saturated zone are largely in response to natural and man
induced stresses (e.g., drainage from unsaturated zone and
pumping).
3.2.2 Dispersion
Dispersion is a dilution process that occurs as a result of
the spreading of a contaminant plume. In the surface zone,
dispersion in overland flow is normally not considered due to
the high velocities normally associated with runoff, and the
relatively short distances runoff travels before entering some
type of channel. In the unsaturated and saturated zones,
dispersion can be of importance and occurs as a result of:
o Molecular diffusion (in response to concentration
gradients).
o Mechanical dispersion: mechanical mixing on a micro-
scopic scale due to tortuosity (erratic pattern of
3-10
-------
flow through pores), branching, and changes in pore
size.
o Heterogenous properties of the media: layering and
differences in permeabilities and porosities on a
megascopic scale.
Dispersion in the unsaturated and saturated zones is primarily
a function of media properties and the scale at which the
heterogeneities of an aquifer system are considered.
3.2.3 Erosion
Erosion is the detachment of soil particles by rain droplets
and subsequent transport by overland flow originating upslope
from or on a waste site. This process occurs only in the
surface zone.
3.2.4 Sorption/Retardation
Sorption is the transfer of a portion of the soluble phase of
a waste constituent to the surface of soil, rock or organic
materials. In the surface zone, sorption is considered as a
separate process which determines the amount of a waste
constituent that will move with runoff, as opposed to with
eroded soil materials. Thus/ it is simply a partitioning
process.
In the unsaturated and saturated zones, sorption is usually
combined with a number of other processes to describe the
delayed movement of certain waste constituents relative to
that of water. The other processes include:
o Filtration
o Molecular diffusion into dead end pore spaces or
fractures
o Ion exchange
o Reversible chemical reactions with other contaminants
or the media
o Precipitation/dissolution
o Flocculation
Retardation is the general term used to describe the delay
constituents will experience due to all of these processes.
3-11
-------
3.3 PROCESSES CONTROLLING TRANSFERS BETWEEN ZONES
3.3.1 Evapotranspiration
Evapotranspiration collectively describes all processes which
act to transfer water from the surface zone and unsaturated
zone to the atmospheric zone. This includes evaporation from
water, soil, snow, ice, and vegetation, as well as
transpiration by plants.
3.3.2 Infiltration
Infiltration transfers water from the surface zone to the
unsaturated zone through progressive wetting of underlying
soils and movement due to hydrostatic pressure. The
infiltration rate is usually high just after the onset of
rainfall and decreases with time as soil pores become filled
with water.
3.3.3 Drainage
Drainage is the transfer of water between the unsaturated and
saturated zones. The hydraulics of drainage are complicated
by the fact that the soil pores in the unsaturated zone
contain both water and air. When the pores are almost
completely filled with water (i.e., near saturation), water
will tend to drain relatively freely in response to
gravitational forces. As the water content decreases,
capillary pressures increase and the amount of drainage that
can occur decreases sharply. This inter-zone process also
acts to transfer waste consitituents into the saturated zone.
3.3.4 Volatilization
The dominant mechanisms for vapor-phase transport of
constituents from the surface zone to the atmospheric zone are
gas phase molecular diffusion and convection by biogas venting
and barometric pressure pumping. Emissions from ponded wastes
are controlled primarily by volatilization at the air-water
interface.
3-12
-------
3.4 PROCESSES CONTROLLING TRANSFORMATION/DEGRADATION
Transformation refers to a number of chemical and biological
processes that act to change or degrade a specific waste
constituent. Quite often, the rate of transformation is
controlled by one or two processes. Key transformation
processes include photolysis, hydrolysis, oxidation, chemical
reaction, and biological (microbial) degradation.
Bio-degradation can also act to transfer contaminants into the
atmospheric zone through respiration of the degrading
organisms or changes from liquid to gas phase during chemical
reactions.
3-13
-------
SECTION 4
REMEDIAL ACTIONS AND AFFECTED PROCESSES
4.1 OVERVIEW
Remedial action technologies may be classified as surface
control, subsurface control, and waste control. Surface
control actions are directed at containing the waste.
Subsurface control actions prevent contamination of the
subsurface by directly containing the waste or by removal of
contamination. Waste control actions are directed at reducing
the source by direct removal or in-situ treatment. The
remedial action technologies that are described herein were
compiled from existing remedial action handbooks (JRB
Associates, 1982 and SCS Engineers, 1982). Sample
applications of many of these technologies to a hypothetical
waste site, including costs, is provided by Tolman et al.,
(1978).
For the purpose of this report, the large number of available
remedial action technologies have been condensed into fourteen
"remedial measures" within the three control groups mentioned
above. These measures are listed along the left axis of Table
4.1. This organization was based upon the similarity of
design objectives of the individual technologies. For
example, subsurface drains and ditches, as well as bottom
liners, were grouped into one measure because they all are
designed to control leachate migration (JRB Associates, 1982).
Remedial actions designed to reduce airborne emissions, such
as gas migration control and fugitive dust control, are not
considered.
The purpose of this section is to: 1) briefly overview the
design objectives of each of the measures listed in Table 4.1
and 2) identify which zones and processes are affected by
these measures and how they are affected. This type of
information is needed to support the guidance given in
Sections 5 and 6.
Table 4.1 summarizes the discussion provided in this section.
It lists each of the measures that will be discussed along
one axis and the zones and processes discussed in the previous
3-14
-------
TABLE 4.1 PROCESSES AFFECTED BY DIFFERENT REMEDIAL MEASURES
U)
i
/ SURFACE ZONE / / UNSATURATEO ZONE / / SATURATED ZONE JJJ.
REMEDIAL ACTIONS
SURFACE CONTROL
Grading
Revegetatlon
Surface Uater Diversion
and Collection
SUBSURFACE CONTROL
Capping and Top Liners
Seepage Basins and Ditches
Subsurface Drains / Ditches /
Bottom Liners
Impermeable Ba-Hers
Ground Uater Pumping
Interceptor Trenches
HASTE CONTROL
Permeable Treatment Beds
BioreclamatJon
Chemical Injection
Solution Mining Extraction
Excavation / Hydraul 1c
Dredging
//
+
-
_
+
-t-
-
-
__
-
±
-
_
-
+
+
+
-
+
_
+
+
-
_
+
+
_
+
-
+
_
+
+
-
+
+
+
-
+
+
+
+
+
-
+
+
+
Legend:
+ • Enhances Process In
-------
section along the other axis. The extent of impact on the
processes affected by each measure is denoted by either a (+)
or (-). The former indicates that the measure tends to
increase the effects of the process in terms of water and/or
contaminant movement. The latter indicates that the measure
tends to decrease the effects of the process. Figures 4.1 and
4.2 show, in plan view and cross-section, a hypothetical waste
site prior to the installation of any remedial actions.
Diagrams of each technology, including key process-related
changes, are provided in Figures 4.3 - 4.11.
4.2 SURFACE CONTROL
Surface control measures such as grading, revegetation, and
water diversion are designed to contain wastes reducing
infiltration and limiting runoff from waste disposal sites.
They can also reduce erosion, stabilize the surface of covered
landfills, and protect receiving water quality. This is
accomplished primarily by directing runoff away from a
hazardous waste site or by containing contaminated runoff.
Surface control measures mainly affect processes in the
surface zone (i.e., runoff, evapotranspiration and erosion)
and the transfer of water and waste constituents between the
surface and unsaturated zones via infiltration. Figures 4.3
and 4.4 show how surface control actions affect different
processes.
4.2.1 Grading
Grading is used to reshape the topography of landfills,
affecting surface zone processes in one of two ways. Usually,
the slope is increased and roughness is decreased to
facilitate runoff and decrease infiltration. The higher
velocities that result from these changes may cause increased
erosion and entrainment of contaminated soil unless other
measures are taken. A reduced slope and increased roughness
may be desired in some arid environments where clay capping
has been installed, to enhance infiltration and keep the cap
pliable. Grading is often used in conjunction with surface
sealing practices and revegetation.
4.2.2 Revegetation
This measure is used to stabilize the topsoil of a covered
landfill. Revegetation decreases erosion by reducing the
detachment of soil particles and reducing overland flow
3-16
-------
REGIONAL
GROUND-WATER
FLOW
u>
1
STREAM
Figure
Hyp
othetical
waste
site
n v
iew)
-------
-------
U)
i
DIVERSION
DITCH
RUNOFF
COLLECTION
DITCH
REDUCED .
PLUME SIZE
GRADED/REVEGETATED
SOIL COVER
Figure 4.3 Grading, revegetation and surface water diversion
and collection (plan view).
-------
.<&
•o
,v
-a-
^
tr
.£
?,
-------
velocities. The introduction of vegetation also increases
evapotranspiration and temporary water storage on ttys surface.
The net effect on infiltration is site dependent, but will
often be a decrease, particularly in less humid areas.
Transpiration capacities, rooting depth characteristics,
durability, preparation and planting characteristics affect
the impacts of revegetation at a site.
4.2.3 Surface Water Diversion and Collection
This type of measure is designed primarily to route runoff
away from a site. The techniques used to accomplish this
include: dikes and berms, ditches, diversions, and waterways;
terraces and benches; and chutes and downpipes. By removing
surface water from the site, these measures reduce the depth
of standing water on the surface, thereby limiting infil-
tration. Because overland flow is confined to collection
channels, erosion can be controlled and transport of
contaminated sediments can be eliminated.
4.3 SUBSURFACE CONTROL
Remedial measures that are included in this group primarily
affect processes in the unsaturated and saturated zones, as
well as processes acting to control the transfer of water and
contaminants between the two zones. Two exceptions are the
capping and top liner measure and the seepage basin and ditch
measure, which also affect processes in the surface zone. The
primary goals of subsurface control measures are to prevent
leachate migration and ground-water contamination through
diversion, containment or collection.
4.3.1 Capping and Top Liners
The placement of impermeable caps and top liners on waste
disposal sites reduces infiltration, increases runoff, reduces
erosion, and isolates the waste hydrologically. Cover soils,
such as clay, that have low permeabilities and are erosion
resistant are spread over the waste and then topsoil and
vegetation are added to stabilize the cap as shown in Figures
4.3 and 4.4. Capping, because it reduces infiltration of
water into the waste, minimizes the possibility that the waste
might reach field capacity and subsequently begin to leach.
Drainage to the saturated zone is also reduced. The water
table beneath the site may be lowered and contaminant plume
size may be reduced as a result of the decreased movement of
3-21
-------
leachate into the saturated zone.
4.3.2 Subsurface Drains and Bottom Liners
Subsurface drains are highly permeable trenches designed to
collect leachate or infiltrating water in the unsaturated
zone, thus reducing contamination of the aquifer. Drains may
also be used to collect leachate trapped by bottom liners
placed underneath the waste site. Bottom liners are low
permeability barriers, usually composed of injected slurries
or grout, that are installed underneath the waste site to
retard the percolation of contaminants. Bottom liners may
also be used to isolate the waste from a high ground-water
table. As shown in Figure 4.5, leachate generation is
minimized by these actions, resulting in a reduced plume area
and lower concentrations.
4.3.3 Ground-Water Pumping
Figures 4.6 and 4.7 show several ways that ground-water
pumping can be used alone or in combination with other
measures. Pumping of ground water is designed to lower the
ground-water table around the waste site or to contain a
ground-water plume. The ground-water table may be lowered to:
1) prevent contaminated ground water from discharging to a
receiving stream, 2) prevent direct contact between the waste
and the aquifer (as shown in Figure 4.6), and 3) prevent
leaky aquifers from contaminating other aquifers.
Ground-water pumping typically involves three steps: 1)
pumping to remove contaminated water and/or depress the water
table, 2) treatment of removed water to extract contaminants,
and 3) recharge of treated water through either injection
wells or seepage basins. A locally elevated ground-water
table is often created as a result of recharging treated
ground water. By depressing and elevating the ground-water
table in the right locations, a plume of contaminated ground
water can be isolated, as shown in Figure 4.7.
4.3.4 Interceptor Trenches
Interceptor trenches are used for the same purposes as
ground-water pumping: to lower the water table around the
site and to capture a plume by controlling the direction of
ground-water flow (see Figures 4.6 and 4.7). They are
characterized by high permeability material like gravel or a
slotted drain pipe in a trench which intersects the saturated
3-22
-------
,. -. • • V''J"*V" •'.",••';•*-',
,••"•'•••'. - -.. - ,- •.",••"•"."* '/•."«•• •'.''-'.v-'t- '•*.'•
-------
****
-------
3"
-------
zone. Water drains passively by gravitational forces into the
trench, thereby lowering the water table. They act in a
fashion similar to subsurface drains, but are used to capture
contaminated water in the saturated zone. Interceptor
trenches cause changes in processes similar to those caused by
extraction wells.
4.3.5 Seepage Basins and Ditches
Seepage basins and ditches are designed to recharge water from
surface collectors or extraction wells, drains and interceptor
trenches. They are sometimes used in conjunction with a
pumping system to change the ground-water table profile (see
Figure 4.7). The bottom of the basin itself is generally
constructed of highly pervious materials, allowing for
increased infiltration into the unsaturated zone. This
increase in infiltration leads to an increase in percolation
and drainage to the saturated zone. A localized rise in the
water table (i.e., mound) results. As a result, local changes
in ground-water flow directions can often be achieved.
4.3.6 Impermeable Barriers
Impermeable barriers are vertical walls of low permeability
material, such as bentonite slurry, cement, chemical grout, or
sheet piling, that are installed through the unsaturated zone
into the saturated zone. They are designed to either prevent
the migration of contaminated ground water away from a site or
to divert uncontaminated ground water away from a site.
Figures 4.8 and 4.9 show plan and cross-sectional views,
respectively, of a barrier completely surrounding the site.
Plume movement inside the barrier is reduced considerably.
However, the potential exists for the plume to escape if the
barrier is not keyed into an impermeable strata. Under
certain situations, partially penetrating or hanging barriers
can be used to reduce leachate generation by lowering the
water table.
4.4 WASTE CONTROL
Waste control measures are used to remove or treat wastes or
contaminated water and sediments. Removal may be accomplished
by excavation or hydraulic dredging. Treatment methods
include permeable treatment beds, bioreclamation, chemical
injection, and solution mining (extraction). These methods
are considered in-situ because treatment is accomplished
3-26
-------
1
-rl
V-4
W*
ri
Q)
I
W
0)
Q)
0)
I
O
"o
oo
•
<•
tr
•rl
3-21
-------
(0
*>
o
H-
rt
I
I
1
cr
&>
rt
«
H-
(0
o
H
O
w
01
\
(0
o
ft
-------
within the landfill/lagoon or plume itself. On-site, as
opposed to in-situ, treatment methods involve the extraction
of contaminated water and above ground treatment, followed by
disposal. Waste control measures have an effect on adsorption
and degradation processes, as well as advection and dispersion
processes.
4.4.1 Permeable Treatment Beds
Permeable treatment beds consist of limestone and/or activated
carbon, and are placed vertically in the saturated zone,
downgradient from a site, as shown in Figure 4.10. The
objective is to remove contaminants from the ground water as
it flows through the bed. Removal effectiveness may diminish
with time, however, as the adsorptive capacity of the bed
decreases or the bed becomes plugged. Permeable treatment
beds mainly increase retardation and degradation processes in
the bed itself.
4.4.2 Bioreclamation
In cases where the ground water has become contaminated with
biodegradable pollutants such as hydrocarbons, bioreclamation
may be considered as a remedial measure. It is an in-situ
ground-water treatment method, involving the injection of
microbial organisms, nutrients, and oxygen into a plume. The
objective is to greatly accelerate the degradation of a
pollutant. Bioreclamation acts to increase degradation
processes. It can also locally affect ground-water movement
and plume dispersion if injection and withdrawal rates are
high enough to substantially modify the ground-water flows.
Figure 4.11 shows the extent of in-situ treatment for a
bio-reclamation system that includes an injection/withdrawal
doublet.
4.4.3 Chemical Injection
Chemical injection is used to treat the waste in a landfill or
lagoon, or in a contaminated saturated zone. It is usually
applied to sites with wastes well defined in both location and
chemical composition with shallow landfill or lagoon depths,
and where the vertical and horizontal extent of the
contamination is small (JRB Associates, 1982). The objective
of the method is to immobilize or destroy a pollutant. The
effect of this measure is to substantially increase local
retardation and degradation processes. Figure 4.11 shows the
3-29
-------
^
.eve"1
-------
OJ
\
OJ
REGIONAL
GROUND-WATER
PLOW DIVERTED
PLUME .
TREATED AND
CAPTURED
L WW6CTIOH WELL
4 U
Figure 4.11 v
-------
extent of treatment for a chemical injection system composed
of an injection/withdrawal doublet.
4.4.4 Solution Mining (Extraction)
Solution mining is similar to chemical injection in that both
methods chemically alter the pollutant. However, solution
mining involves the injection into a landfill of a chemical
solvent, which desorbs or frees the pollutant so that it may
be mobilized in a larger leachate flow. The leachate is then
collected by interceptors and/or well points (see Figure
4.12). The objective is to increase the mobility of the
contaminant. Adequate confinement and collection of the
resultant leachate is necessary to prevent increased aquifer
contamination. Contaminant movement is also increased by
solution injection and collection.
4.4.5 Excavation and Hydraulic Dredging
Excavation and hydraulic dredging involve the removal of the
waste source itself. Hydraulic dredging may be used to remove
liquids and/or sludges from lagoons or surface impoundments.
After the waste area has been excavated or dredged, it may be
backfilled and capped to control infiltration.
Depending on the permeability of the backfill material and
other site restoration actions, infiltration to the
unsaturated zone may increase or decrease. This will in turn
lead to a decrease or increase in water percolation in the
unsaturated zone and drainage to the saturated zone. Since
the measure leads to the removal of waste materials, there
should be a major decrease in leachate generation and plume
size.
3-32
-------
I
Ul
U)
Figure 4.12
'NJECTION WELL
WITHDRAWAL WELL
-------
SECTION 5
NUMERICAL MODEL APPLICATION GUIDELINES
5.1 OVERVIEW
Numerical models of the subsurface and surface environments
provide capabilities which both complement and exceed those of
field data analysis and simplified methods. In contrast to
simplified methods, numerical models approximate process
equations using finite difference or finite element solution
techniques that make it possible to represent important
spatial and temporal variations in site conditions. Along
with this benefit, however, comes the cost of gathering the
field data required to describe key variations. Consequently,
a trade-off must be made between the ease of solution,
computational accuracy, limited resolution and limited
applicability of simplified methods and greater resolution,
more general applicability, increased complexity and increased
costs for numerical models. Key attributes of numerical
models can be summarized as follows:
1. Fewer simplifying assumptions are required, although
the simplicity and computational efficiency of the
solution algorithm depend, in part, on assumptions
made.
2. Values of key quantities (e.g., velocity and chemical
concentration) are computed at discrete space and
time intervals selected by the user. These intervals
can be adjusted to achieve the accuracy and
specificity required by site conditions and the
problem being addressed.
3. Multiple independent variables (e.g., velocity, temp-
erature, and chemical concentration) can be simulated
simultaneously, including interactions between these
variables.
4. Numerical solutions to the governing equations are
approximate and subject to computational errors due
to truncation, roundoff and numerical dispersion.
3-34
-------
Choice of solution scheme can have a substantial
effect on these errors.
5. Resources required to implement numerical models de-
pend on the dimensionality, resolution, number of
independent variables being predicted, and solution
scheme. Required resources include: user expertise
in modeling, field data, personnel time, and computer
facilities. It is reasonable to expect that needed
resources will be two to ten times those required for
simplified model applications.
A number of authors provide overviews of numerical models and
their use in the analysis of surface and subsurface problems,
including Mercer and Faust (1981), Javandel et al., (1984),
Bachmat et al., (1978), Orlob (1971) and Donigian (1981).
The reader is refered to these sources for more information on
model theory, structure, implementation and use.
Numerical models, as a result of the above attributes, are
most appropriately used for the analysis of physical and
chemical processes and site conditions which cannot be
adequately represented with simplified methods. Situations
which may justify numerical models include:
1. Local and/or off-site media properties which vary sig-
nificantly with location or direction causing complex
flow and transport conditions;
2. Highly variable, discontinuous or geometrically
complex boundary conditions (e.g., mixed flow and
no-flow boundaries) which require detailed
representation;
3. Time varying sources, sinks, or boundary conditions
(e.g., seasonal fluctuations in river water levels or
infiltration rates) which strongly influence flow and
transport; and
4. Remedial actions which, when implemented, result in
one or more of the conditions listed previously
(e.g., impermeable barriers).
Volume 1 of this report presents a methodology for determining
when numerical models should be used for analysis of surface
and subsurface remedial actions.
5.2 NUMERICAL MODEL CAPABILITIES
A broad spectrum of numerical models, potentially applicable
3-35
-------
to remedial action assessment, have been developed. For ease
of discussion, models, are often classified by the types of
problems they can solve and the solution techniques used.
Categories include: solution domain or zone, independent
variables considered and numerical solution technique.
As was noted in Section 3, the environment in the immediate
vicinity of a waste site can be divided into four zones: 1)
atmospheric, 2) surface, 3) unsaturated, and 4) saturated;
surface water bodies (e.g., lakes and rivers) are considered
to be a separate zone. Water and contaminant movement in each
of these zones is controlled by different processes; the
governing equations are enough different so that separate
solution schemes are usually required. The capabilities of
models for the atmospheric zone are outside the scope of this
volume and models for surface water bodies are discussed in
Volume 4. Models for the remaining zones or solution domains
are discussed below.
Independent variables can be grouped as flow-related and
transport-related. Flow models solve the applicable momentum,
continuity and pressure equations to yield estimates of fluid
movement and storage. Transport models use estimates of fluid
movement to predict chemical migration and fate. Flow and
transport calculations are often performed separately and in
sequence. Some code designers have chosen to have entirely
separate codes for the two computations so that each code is
as simple and efficient as possible. This latter approach
assumes that chemicals and other constituents being
transported will not affect fluid flow. Situations involving
chemicals that are denser or less dense than water, usually
cannot be simulated in this way.
Numerical solution procedures fall into two general
categories: finite difference methods (FDM) and finite element
methods (FEM). Other methods, such as integrated finite
difference and method of characteristics combine attributes of
FDM and FEM, but are generally referred to as finite
difference methods because of the way the solution domain is
represented. Mercer and Faust (1981) provide a brief
discussion of all of these techniques, including references to
in-depth treatments of each technique.
5.2.1 Surface Zone Models
Section 3 of this volume provides a description of key
processes controlling water and chemical movement in the
surface zone. In essence, the surface domain extends from the
surface of a hazardous waste site to the root zone and
downslope to a receiving water body (see Figure 3.1).
3-36
-------
Most readily available surface zone models represent the
surface zone with single or multiple "land segments," each
having uniform properties (e.g., slope, surface roughness, and
vegetative cover). The most comprehensive surface zone models
also simulate soil moisture in the unsaturated and saturated
zones to obtain improved estimates of infiltration rates.
Less comprehensive models use empirical relationships for
antecedent soil moisture. Numerical solution schemes for
these models use simple finite difference techniques. The
majority of surface zone models are event-based. That is,
they simulate one hydrologic event at a time. Models which
provide continuous simulation of runoff and soil moisture, as
well as water quality, require efficient data manipulation and
storage routines due to the large number of parameters and the
frequent time steps needed to represent runoff and
infiltration processes. Several models provide some data
management capability but only one, the Hydrologic Simulation
Program - FORTRAN (HSPF) (Johanson et al., 1981), provides
comprehensive data manipulation and storage capabilities.
Donigian (1981) discusses the evolution of surface zone
models and existing model capabilities and limitations in more
detail.
Volume 1 presents one approach to the selection of models for
remedial action evaluation. Examples of representative
surface runoff models which are potentially suitable for
remedial action evaluation are discussed below as a starting
point for those interested in applying such models. More
detailed listings of models and a discussion of model
attributes can be found in Onishi et al., (1983), and Donigian
(1981). Table 5.1 summarizes the capabilities of five surface
zone models, fifteen saturated and seven unsaturated zone
models. The characteristics of three particularly versatile
codes are shown in Table 5.2. These three models are
described briefly below.
HELP (Schroeder et al., 1984a and 1984b) estimates daily
water movement on the surface and through a landfill by
partitioning precipitation (and runoff entering the site) into
runoff, evapotranspiration, infiltration, and lateral
drainage. The SCS Curve Number method is used to estimate
runoff on a continuous (daily) basis, using a soil moisture
accounting procedure to determine infiltration. The landfill
is divided into discrete layers and moisture is routed
vertically from one layer to the next using Darcy's Law.
Although the original version of HELP does not simulate
leachate quality, Bicknell (1984) has modified HELP to
simulate chemical losses from a landfill. Both leaching and
volatilization losses can be estimated. HELP has been used in
the analysis of existing landfills and the design of new
sites.
3-37
-------
TABLE 5.1 GENERAL CAPABILITIES OF SELECTED SATURATED, SURFACE AND
UNSATURATED ZONE MODELS
MODEL NAME (References)
SURFACE ZONE MODEL
HSPF (EPA)
SHMN (EPA)
CREAMS (USDA / Corps of Engrs.)
SEASOIL (A.D. Little, Inc.)
HELP (EPA / Corps of Engrs.)
UNSATURATED ZONE MODEL
FEMUATER / FEMHASTE (ORNL)
TRUST / MILTRAN (LBL / Battelle)
COLUMN TRANSPORT WITH SORPTION
(Ktpp. Kenneth L.; England)
ODMOD (Argonne National Lab)
NMOOEL (Univ. of Florida)
PERCOL (Battelle)
PRZM (EPA /Athens)
(")
Unknown
(continued)
-------
TABLE 5.1 (continued)
/ SURFACE ZONE / UNSATURATEO ZONE / SATURATED ZONE / /
/t-31/ / / «. / / * / •& / MAJOR CODE LIMITATIONS
//? / . A A< j /AV/ ./ ////
SATURATED ZONE MOOEL
FEWA / FCMA (ORNL)
SWIFT (Intera)
HCTM (Intera)
FE3DGU / CFEST (Battelle)
AT123D (ORNL)
PLASM (PrUkett » Lonnqulst)
WASTE (Analytical Science Corp. )
GWSIM-II (Tenas 0>pt. of Water
Resources)
MOC (Konlko« > Bredehoeft. USGS)
GROUNDWATER COMPUTER PACKAGE
(Marlon-Lambert, J.; Canada)
PATHS (Battelle)
TRANSCOL / FRACSOl
(Prkkens. J.F.; Canada)
GETOUT (Burkholder. et al)
NEUSAM (Ledoux. E.; France)
VTT (Battelle)
fflf / //&// / /'
'///// / /
2
3
3
3
1
1
2
2
Z
2
1
1
1
2
X
I
X
I
X
I
X
I
X
X
I
X
X
X
X
X
X
X
I
I
X
I
X
X
X
X
«
X
X
'///
c
c
I
c
c
c
c
c
• Being analytical the model has limited
spatial >esolut1on
• 1 dimensional unsteady state or
2 dimensional steady state
• No adsorption and degradation
• No adsorption and degradation
• No documentation
• Being analytical the model his limited
spatial resolution
• 1 dimensional, no dispersion and degradation
• 1 dimensional
• 1 dimensional, no dispersion and degradation
• 2 dimensional , no pollutant transport sub-
module
U)
I
UJ
Legend:
(H) • Multiple Land Segments
(S) " Single Land Segment
X * Considered
C • Complete Documentation
I • Incomplete Documentation
or User's Guide
-------
TABLE 5.2 DETAILED CAPABILITIES OF SELECTED SURFACE, UNSATURATED AND
SATURATED ZONE MODELS
U)
I
/ SURFACE ZONE / / UNSATURATED ZONE / / SATURATED ZONE / tOVSieiMllQKS
CODE NAME (Reference)
SURFACE CODE
HSPF (EPA)
CBEUKS (USBA / Corps of Emirs.)
HELP (CPA) / Corps of Engrs.)
UNSATURATED ZONE CODE
FEMWATER / FEHHASTE '(ORNl)
TRUST / KLTR4.N (Ul / Battelle)
SATURATED ZONE CODE
FEUA / FEHA (ORNL)
SKI FT (Intera)
KCTM (Intera)
FE3DGH / CFEST (Bjttellt)
PLASH ( Pricket t 1 Lonnqulst)
(N)
IS)
(S)
X
I
I
1
I
I/S
I/S
L
I
2
2
X
X
I
X
X
X
X
X
2
2
X
X
X
X
X
X
0
X
X
X
X
0
X
X
X
X
X
0
X
X
•
Footnote:
1. FIox Model / Transport Model
lejend:
(N) • Multiple Land Sequent
(S) • Single Und Segment
I • Inflltritton
S • Seepage (handling of
seepage pond)
X • Considered
0 • Case studies - unpublished
Documentation - only for flow model
not for transport code
User's Guide - only for flow mode!
not for transport code
-------
CREAMS (Knisel, 1980) simulates surface hydrologic processes,
either continuously using the Green and Ampt formulation or
for discrete events using the SCS Curve Number approach. Like
HELP, it provides for only a single land segment and cannot
represent spatial variations in hydrologic conditions. It
simulates most of the important processes, including sediment
production and transport.
HSPF (Johanson et al. 1981) is the most recent version of a
family of watershed hydrology and quality models which have
the Stanford Watershed model as a base. HSPF simulates
surface and subsurface processes for multiple land segments
and is capable of representing complex hydrologic and chemical
transport conditions. Additional modules simulate transport
in surface water bodies and the interactions between surface
water and subsurface water and chemical movement. A
sophisticated data base management system is included as part
of this model.
5.2.2 Unsaturated Zone Models
For our purposes, the unsaturated zone begins at the base of
the root zone and extends to the water table (or capillary
fringe, if considered). Because moisture content is less than
porosity, the properties influencing water movement in this
zone (moisture content and hydraulic conductivity) depend upon
pressure head. Water movement is predominantly vertical.
Soil heterogeneities can result in lateral migration of water
and contaminants around clay layers and other discontinuities.
Available unsaturated zone models vary widely in their
capabilities and characterisitcs. While two-dimensional,
finite element codes appear to be the most common, finite
difference codes are also readily available. Separate codes
for flow and transport calculations are common, due to the
complexity of water movement. A number of codes can simulate
both unsaturated and saturated conditions and may be
potentially useful where fluctuating water table elevations or
perched saturated conditions are important. The most complex
models also simulate multi-phase flow and/or heat transfer and
may be appropriate if detailed modeling of multi-phase
transport is required.
The number and diversity of unsaturated zone codes often makes
selection difficult. Brief discriptions of several codes are
given below as a starting point. Surveys and critiques of
available codes can be found in Kincaid et al. 1984, Nelson et
al. (1982), and Oster (1982). The International Ground Water
Modeling Center (IGWMC) operated by Holcomb Research
3-41
-------
Institute, Butler University in Indianapolis, Indiana provides
a clearinghouse for information on the capabilities of a
number of different unsaturated zone models. Table 5.1
summarizes the capabilities of seven codes while the
characteristics of two particularly versatile codes are
presented in Table 5.2.
TRUST and MLTRAN (Narasimhan and Witherspoon, 1976, Reisenauer
et al. 1981 and 1982) are companion flow transport codes for
variably saturated media. TRUST is a two-dimensional
integrated finite difference code. TRUST considers soil
processes such as deformation, as well as the wetting-drying
front problem. TRUST output is formatted for direct input to
MLTRAN. MLTRAN was originally developed for the evaluation of
low level radioactive wastes and computes travel path and
travel time for water and a chemical. Retardation and
degradation of constituents are simulated, but dispersion is
not included.
FEMWATER and FEMWASTE (Yeh and Ward, 1979 and 1981) are
companion flow and transport codes that simulate two-
dimensional unsaturated/saturated ground-water systems.
Boundary condition options allow representation of seepage
from ponds, as well as surface infiltration. The simulated
plane can be vertical (x-z) or horizontal (x-y), allowing
simulation of lateral drainage. FEMWASTE represents all
important contaminant transport processes, including
dispersion. Heterogeneous soil properties, including the
effects of remedial actions, can be represented using a
variable finite element grid.
5.2.3 Saturated Zone Codes
The saturated zone extends from the water table downward to
underlying basement rock. Fluid and contaminant flow are
controlled by pressure head and hydraulic conductivity, and
are fundamentally three-dimensional.
A broad spectrum of saturated zone models are currently
available, varying from one-dimensional finite difference flow
codes to three-dimensional finite element codes that include
multiple phases, temperature effects, transport in fractured
media and geochemical and biological reactions. Numerous
surveys and critiques of saturated zone models are available,
including, Kincaid et al. (1984), Thomas et al., (1982),
Javandel et al. (1984), Gelhar (1977), Bachmat et al. (1978),
van Genuchten (1978a), Anderson (1979), Grove and Kipp (1980),
Knox and Canter (1980), Lappalla (1980), Moiser et al. (1980),
SAI (1981) and Koines (1982). The first two references
provide detailed reviews of a limited number of models, while
3-42
-------
the others are more comprehensive inventories. Again, IGWMC
provides information on a number of saturated zone models.
Brief descriptions of five saturated zone models (see Table
5.2) are provided below as examples of codes potentially
suitable for remedial action evaluation. Table 5.1 gives
general characteristics of 15 potentially applicable codes.
PLASM, or the "Random-Walk" Solute Transport Model, developed
by Prickett and Lonnquist (1981) is a two-dimensional (x-y),
transient model. It considers all important saturated zone
processes and inter-zone transfer processes. Judgement is
needed to arrive at an acceptable solution, since improper
discretization may cause the predicted concentrations to be
greater than the initial concentration. The "lumpy" character
of output (expressing concentration in terms of number of
particles) requires computer plotting and smoothing routines
to draw meaningful results. Such subroutines had not been
incorporated into the computer code at the time of this
review.
FE3DGW/CFEST (Gupta et al., 1979 and 1982) are two finite
element models which can simulate two-dimensional or
three-dimensional systems which are complex and multi-layered.
Flexible boundary conditions, an easily defined and modified
finite element structure, and the capability to model point
sources and sinks make this model both powerful and adaptable.
CFEST now simulates both retardation and degradation.
HCTM, or the Hydrologic Contaminant Transport Model, developed
by Intera, Inc., considers all the required saturated zone
processes such as adsorption, degradation and dispersion, as
well as inter-zone transfer processes. It handles
heterogeneous soil properties and provides variable spatial
resolution. It is a proprietary model and is not available to
the public, except by purchase.
SWIFT is generically related to HCTM. It was developed by
Intera and Sandia National Laboratories for the Nuclear
Regulatory Commission from the earlier USGS model SWIP
(predecessor of the DWDM - Deep Well Disposal Model). The
model, is more complex and costs more to run than the HCTM
code (Lantz, R., personal communication) as it couples a heat
transport sub-module to the original fluid and contaminant
transport codes. Unlike HCTM, SWIFT is not proprietary and
iswell documented with a user's manual and self teaching
guide (Dillion, et. al., 1978; Finley and Reeves, 1968;
Reeves and Cranwell, 1981).
FEWA/FEMA has been developed by Oak Ridge National Laboratory
(Yeh, G., unpublished draft). It is designed to be compatible
with FEMWATER/FEMWASTE. Like HCTM and SWIFT, it considers all
important saturated zone processes and inter-zone transfer
3-43
-------
processes. It also handles heterogeneous soil properties and
provides variable spatial resolution. Unlike HCTM and SWIFT,
FEWA/FEMA is not three-dimensional but two-dimensional, and is
less complex. It has a user's guide but complete
documentation was not available at the time of the review.
5.3 INTERACTIONS BETWEEN MODELS
Most hazardous waste sites hydrologically and chemically,
influence more than one zone. While most remedial actions
typically focus on a specific zone, they almost always change
water and chemical movement in other zones. Consequently,
more than one model will often be required to adequately
represent certain remedial actions. Table 4.1 lists the
inter-zone transfer processes (i.e., infiltration,
percolation, drainage, and pumping) affected by different
actions.
All numerical models have a limited solution domain because
differences in physics and, to a lesser extent, chemistry
between zones require substantially different governing
equations and solution techniques. Furthermore, ease of use
dictates that numerical codes be limited in size and
complexity, often yielding separate codes for flow and
contaminant transport calculations. When a complex hazardous
waste site must be modeled or the effects of certain remedial
actions predicted, several codes may be required. Because of
the interactions between zones, the codes must communicate
with one another.
Inter-code communication or linkage can be provided in one of
three ways:
1. Transfer of data between models by hand,
2. Integration of governing equations and solution tech-
niques into a separate computer code (hard linkage),
and
3. The use of external data management programs to indi-
rectly link the programs (soft linkages).
Hand transfer of data between codes is the most common, least
efficient and least reliable method for linking models. It
requires little advance preparation and no new software, but
can be very labor intensive if the number and extent of model
interactions are large. Hard linkage integrates the separate
computer codes so that all equations are solved simultaneously
and information is passed between the models during each
computation cycle (i.e., time step). Hard linkage requires,
3-44
-------
in a practical sense, that the codes be merged into a single
code. This type of undertaking is ambitious and can result in
a comprehensive, but complex, code. Soft linkage allows the
codes to remain separate and retain their original data
structures, computational sequences, and input/output
structures. Linkage is implemented via an external data
management program, often referred to as a "bridge program,"
which accepts output from one code, makes necessary
modifications, and inputs data to another code. Data
transfers typically occur only in one direction; consequently,
"feedback" from the second code to the first cannot occur.
The codes are run sequentially, with computations in the first
code proceeding independently from any conditions or results
in the second code. Of these three linkage techniques the
soft linkage or "bridge program" is most commonly used. The
next two sections discuss some of the considerations and
design procedures typically needed to use a soft linkage.
5.3.1 Soft Linkage of Codes
A typical soft linkage of surface, unsaturated and saturated
zone codes for the assessment of remedial action performance
is shown schematically in Figure 5.1. Site processes are
represented by the unidirectional movement of water and waste
constituents between zones. This requires that the surface
zone code includes the plant root zone where transpiration can
be removed from infiltration, leaving "net infiltration" for
input to the unsaturated zone code. It also requires that the
position of the water table remain fairly constant since there
is no feedback between the saturated and unsaturated zone
codes.
Remedial actions, such as subsurface drains and ground-water
pumpings can potentially cause feedback problems if water and
contaminants are withdrawn from one zone are re-introduced to
another zone through land application or seepage basins. To
account for this feedback, flow quantities and chemical
concentrations, including the affects of treatment, must
initially be estimated, checked, and possibly adjusted through
an iterative procedure.
For the type of linkage shown in Figure 5.1, the following
simulation steps would need to be performed to represent an
entire site, including remedial actions:
1. Input chemical/biological conditions and meteorologic,
hydrologic, and hydraulic conditions to the surface
zone code and run the code over a selected simulation
period. If remedial actions include land
application, estimate an application rate and waste
3-45
-------
SITE PROCESSES
TRANSLATION PRECIPITATION
* J,
OVERLAND FLOW_
ONTO SITE
REGIONAL FLOW
ONTO SITE *"
1 f ^
SURFACE ZONE OVERLAND FLOW
CODE Off SITE
1
NET INFILTRATION
*
I BRIDGE PROGRAM 1
1 +
f *+
UNSATURATED ZONE
CODE
I
DRAINAGE
*
1 BRIDGE PROGRAM 1 __X
1 ^
SATURATED ZONE REGIONAL FLOW
CODE OFF SITE
EFFECT OF
REflEDIAL ACTIONS
___ L*"" —
APPLICATION ~"
l L
TREATMENT
1 L
SEEPAGE
BASINS
^ 1 FAfUlTC
^ COLLECTION
^ '
TREATMENT
f
^ PUMPINS '
INTERCEPTOR DRAIN
M1THDRAWLS
Figure 5.1 Typical soft linkage of surface, unsaturated and
saturated zone codes.
3-46
-------
constituent concentration.
2. Transfer the predicted net infiltration rates for the
simulation period to the bridge program for
processing and then to the unsaturated zone code.
3. Run the unsaturated zone code over the simulation per-
iod using net infiltration inputs, as well as any
leachate collection rates and estimated seepage basin
water surface elevations or seepage rates.
4. Transfer the drainage rates for the simulation period
to the bridge program and then the saturated flow
code, making any necessary conversions.
5. Run the saturated zone code over the simulation period
using drainage rate inputs and any pumping/injection
rates and interceptor drain withdrawals.
6. Compare estimated land application rates and basin
flows and associated contaminant concentrations
assumed in Steps 1 and 3 with the model results in
steps 3 and 5. If the estimates are inappropriate,
adjust and rerun the models.
Since the soft linkage does not allow feedback between the
codes, an iterative procedure will often be necessary to
properly simulate transfers of water and waste constituents
associated with certain remedial actions.
5.3.2 Generic Bridge Program Design
The design of bridge programs to link codes basically involves
identifying the specific model results that need to be
transferred between codes. In general, these results will be
in the form of time series (i.e., a chronologically ordered
series of values). The design process also involves
determining whether any unit conversions are required.
Finally, the need to aggregate or disaggregate time series to
account for differences in model time step requirements and
the need to combine or separate time series to account for
differences in spatial discretization have to be considered.
The time series that must be transferred between the surface
and unsaturated zone codes are net infiltration and any
associated contaminant concentrations. The time series that
must be transferred between the unsaturated and saturated zone
codes are drainage of water and associated contaminant
concentrations.
3-47
-------
Due to the difference in time scale for processes in the three
zones, time stepping will differ between codes. Typical time
steps are minutes to hours, hours to days, and days to months
for the surface, unsaturated and saturated zone codes,
respectively.
Spatial discretization or computational element size will also
typically be different between models. Most surface zone
codes use relatively large single or multiple land segments.
Unsaturated zone codes may need to represent vertical and
horizontal variations in soil properties due to waste site
conditions and remedial actions. As a result, relatively
small, variable element sizes are often used. Saturated zone
code element sizes will vary with aquifer geometry and type of
remedial action, but will often be larger than the unsaturated
zone elements. In addition to differences in land segment and
element sizes between codes, different dimensionalities are
typically used. The surface zone is always represented in
one-dimension, while the unsaturated zone is usually
represented in either one (vertical or z).or two dimensions
(longitudinal-vertical or x-z). If a two-dimensional
representation is used, consideration must be given to how the
surface zone code results will be "mapped" onto the
two-dimensional unsaturated zone grid. A similar situation
arises when unsaturated zone code results in two-dimensions
(x-z) have to be mapped onto the y dimension of a two- or
three-dimensional saturated zone grid (see Figure 5.2). The
combination and separation of time series to account for
differences in element sizes and dimensionalities will be
specific to the codes selected and site being assessed.
The operation of multiple codes as a single system requires
that certain consistency checks be made to ensure accurate
results. The most important of these is conservation of mass.
Linkage procedures need to be checked to ensure that the total
mass of water and contaminant output from one code is input
exactly into the next code. This is often complicated by the
spatial and temporal differences between codes, as discussed
above. An input vs output mass balance should be computed
within each bridge program.
5.4 MODEL APPLICATION PROCESS
The process of "setting up" a computer code to simulate the
key processes controlling water and waste constituent movement
at a specific site is called the "model application" process.
It involves combining one's understanding of how a code
represents individual processes with one's understanding of
their actual occurrence in the field to obtain a model of the
site. Here, a code refers to the computer program that solves
3-48
-------
SURFACE ZONE
(SINGLE LAND SEGMENT REPRESENTATION)
UNSATURATED ZONE
(X - Z REPRESENTATION )
/ 7 / / /
SATURATED ZONE
(X - Y - Z REPRESENTATION)
Figure 5.2 Typical dimensionalities used to represent surface,
unsaturated and saturated zones.
3-49
-------
a set of equations. A model can either be generic or
site-specific. A generic model is the representation of a
generic physical system by a set of equations, conditions and
assumptions. A site-specific model is obtained by applying a
generic model to a particular site. The latter is based on
available site data and past experience. Application involves
using the model to analyze target situations, in this case the
performance of potential remedial actions. Mercer and Faust
(1981) discuss ground-water model development and application,
including data requirements, sources of error and
possibilities for misuse. Figure 5.3, taken from their
article, shows the steps in the model development and
application process. Once the need for numerical modeling has
been determined and appropriate models selected, the following
steps may be taken:
1. The conceptual understanding of site conditions is
further defined and quantified through the collection
and analysis of site data. This "conceptual model"
may also include approximate effects of potentially
feasible remedial actions.
2. The conceptual model is then used to define the nu-
meric model structure required for each zone, the
types of outputs needed, and the required spatial
(i.e., dimensionality) and temporal resolution.
3. Linkages between codes for each zone can also be spe-
cified by the conceptual model. The design of these
linkages will depend on the structure of each code
and the required interactions between zones.
4. Individual codes are installed on an appropriate
computer and the site model implemented by creating
an appropriate structure (i.e., grid configuration,
boundary conditions, and sink and source node
locations).
5. Values for individual model parameters are estimated
from field data and then verified by comparing model
predictions with available site data (i.e.,
calibration or history matching).
6. Appropriate linkages between zone models may be imple-
mented to form a complete model which represents all
important aspects of the site. In this way, the
inter-zone movement of water and contaminants can be
simulated.
7. Adjustments to model parameters and model structure
can then be made to represent the effect of
alternative remedial actions on water and constituent
3-50
-------
DETERMINE NECESSITY
OF NUMERICAL MODEL
COMPILE & INTERPRET
AVAILABLE DATA
±
COLLECT DATA AND
OBSERVE SYSTEM
Conceptualization
History Matching
(Field Problem)
PREPARE DATA
FOR MODEL
USING ESTIMATED
PARAMETERS
i
PREPARE DATA
FOR MODEL
USING ESTIMATED
PARAMETERS
Improve |
Conceptual
Model
INTERPRET
RESULTS
1
COMPARE RESULTS
WITH OBSERVED
DATA
Results
Satisfactory
Good
Comparison
Poor
Comparison
SENSITIVITY RUNS
ARE MORE DATA
NEEDED?
Yes
No
PREDICTIVE
SIMULATION RUNS
Figure 5.3
Model application process (from Mercer and
Faust, 1981). Copyrighted by National Water
Well Association.
3-51
-------
movement. Model parameter adjustments required to
represent specific remedial action alternatives are
discussed in detail in Section 6. The simulation of
certain actions (e.g., bioreclamation) may require
the adjustment of selected parameters with time.
Codes with a "restart" capability are particularly
well suited to this type of analysis. The restart
capability simply allows the user to stop a
simulation run, adjust one or more parameters, and
then start the simulation again.
8. The models (either individual or linked) can now be
run to predict future conditions with and without
remedial actions. Various combinations of actions
can be explored. Where data uncertainties exist,
sensitivity analyses can be used to estimate the
range of outcomes.
The development of a conceptual model for a site and the
collection of key site data to be used in models is discussed
by several authors, including Mercer and Faust (1981) and
Javandel et al. (1984). Model verification and parameter
adjustment is discussed in the user's guides for most codes,
and in numerous reports and papers—see the discussions and
references in Bedient et al. (1981) and Knox and Canter
(1980).
A number of important issues can be addressed when applying
numerical models, including:
1. Existing exposure routes and levels of exposure for
specific chemicals
2. Future exposures if no action is taken
3. Effects of alternative remedial actions on conditions
at and near the site
4. Future exposures during and after the implementation
of alternative remedial actions
Most of these questions will need to be answered during the
screening and analysis of alternatives. While screening may
require simplified methods, numerical models will find use in
the analysis of alternatives where complex site conditions
exist or complex remedial actions are anticipated.
During the remedial investigation, site characterization data
are collected. Site characterization could also include the
use of numerical models to specify chemical sources, chemical
migration pathways, and potential receptors.
3-52
-------
5.5 USER EXPERTISE AND RESOURCE REQUIREMENTS
The application of numerical models requires a level of
expertise that goes beyond that needed for the simplified
methods discussed in Volume 2. This is largely because both
computers and numerical methods are required to efficiently
solve practical problems, whereas simplified methods can often
be solved by hand or through the use of programmable
calculators and micro-computers. The following four basic
areas of expertise are required:
1. Hydrology/hydrogeology - Model users should have the
ability to conceptualize hydrologic systems and
identify key processes controlling water movement at
a site. Since both are largely derived from
available site characterization data, an
understanding of the limitations associated with
different field sampling methods is required.
2. Environmental Chemistry - Model users should be able
to identify important chemical migration and fate
processes, including the estimation of physical-
chemical properties, transfer coefficients, and rate
constants. The need to consider multi-phase
transport, density driven transport and interactions
that occur in complex mixtures is also required.
Again, since site characterization and literature
data provide much of the basis for parameter
estimation the user should have an understanding of
sampling procedures, analytical methods and chemical
property estimation methods.
3. Numerical Analysis - Numerical methods are used in all
numerical models, and even some simplified methods,
to solve basic driving equations. Errors can be
introduced in simulation results, unless the user
clearly understands the limitations associated with
different methods. These limitations can include
grid spacing (i.e., spatial discretization) and size
of the time step needed to obtain a stable, accurate
solution. A related area of expertise is in the
linkage of models. Users must understand how to
correctly transfer model results and map them onto
grids with different dimensionalities.
4. Computer Operations - At a minimum, numerical models
should be solved on a micro-computer. The
application of multi-dimensional models to large
problems will generally require a mini-computer or a
main-frame system. The efficient use of numerical
3-53
-------
models requires expertise in code implementation,
programming, data management, data processing and
computer graphics. Computer operations expertise is
especially important if models require linkage
through bridge programs.
Clearly, few individuals have all of the above expertise. For
this reason, it is common that a team approach will be
followed. It is important that the team members not only have
training in the above areas, but also considerable experience.
In many cases, the level of previous experience with similar
site conditions and a similar, or the same, model will
determine the success and quality of a modeling effort.
As with user expertise, more resources are generally required
to apply numerical models. Here resources refers to:
1. Computer facilities - As was stated above, access to
at least a micro-computer is required. Generally,
the user must have access to a mini-computer or main-
frame system.
2. Data - One of the major benefits of numerical models
is that spatial and temporal variations in site
conditions can be considered. To take advantage of
this benefit, data must be available to describe
variations in key parameters. Considerable data are
also required for model testing (i.e., calibration/
verification or history matching).
3. Time/manpower - The collection/reduction of site char-
acterization data, the development of a conceptual
understanding of important processes, and model
calibration/verification are the three most time
consuming steps in applying a numerical model;
relatively little time is required to analyze
remedial action performance once these steps are
completed. While it is difficult to specify the
exact time required for each step, a complete
numerical modeling study can easily require between 3
and 6 months of calender time and at least twice this
amount in manpower.
It is important to recognize the need to be able to commit
these levels of resources prior to initiating a numerical
modeling study.
3-54
-------
5.6 ANALYSIS OF REMEDIAL ACTION PERFORMANCE USING NUMERICAL
MODELS
The evaluation of remedial action performance initially
involves screening out those actions that clearly will not
meet site clean-up goals. Best engineering judgement
supported by the use of simplified methods, like those
discussed in Volume 2, are usually sufficient to determine
which general technologies are likely to work. This screening
effort is followed by a more detailed analysis of the
remaining actions in terms of technical feasibility and
environmental, public health, institutional and cost concerns.
The technical feasibility of an action relates to the degree
to which design objectives are achieved (i.e., effectiveness)
and the length of time that effectiveness is maintained (i.e.,
durability). It also relates to the ease of implementation of
a remedial action and possible concerns associated with risk
of failure. Environmental concerns are the incremental
impacts and benefits associated with the implementation of a
remedial action, while public health concerns relate to
reductions in human and environmental exposure levels.
Institutional concerns are related to relevant local, state
and Federal regulation. Cost concerns are the capital and
operating and maintenance costs for a given action.
Based on the results of the detailed analysis, one or more
actions are then selected for conceptual design. This effort
involves determining the optimal location, size and
configuration of a remedial action alternative.
Recent applications of numerical models (e.g., Silka and
Mercer, 1982; Cole et al., 1983; Mercer et al., 1983; Cohen
and Mercer 1984; and Anderson et al., 1984;) have shown how
they can be used to support the analysis of: 1) reductions in
exposure levels, 2) uncertainty regarding remedial action
performance, 3) optimization of remedial action designs, and
4) design life and impacts of failure.
5.6.1 Assessment of Reductions in Future Exposure Levels
Narrowing the large number of possible remedial actions down
to a set of technically feasible actions may be difficult if
one depends only on best engineering judgement. While it may
be relatively easy to determine that a subsurface control
measure is needed to clean-up a contaminated aquifer, site
conditions may make it difficult to determine whether a
pumping/injection system, up-gradient cut off wall,
downgradient cutoff wall, interceptor drains or combination of
these actions will be most effective in terms of providing the
3-55
-------
greatest reduction in exposure levels or which actions can
meet established site clean-up goals. Similarly, it is
difficult to determine what level of reductions are achievable
where specific clean-up goals are subjective or not
established. One of the benefits associated with using
numerical models is that environmental concentrations useful
for exposure and risk assessment can be estimated for a number
of locations of interest, including drinking water wells, site
boundaries or nearby surface water bodies.
Figure 5.4 shows the results of a model-based evaluation of
remedial action performance for the La Bounty Landfill in
Charles City, Iowa (Cole et al., 1984). This figure shows
predicted concentrations of arsenic levels in a river (the
Cedar River) adjacent to the landfill under low flow
conditions. The pre-restoration curve shows the predicted
build-up of arsenic concentrations from 1967 to 1983. The
curve labelled clay cap shows how concentrations were
predicted to change after the installation of a clay cap; this
curve represents the base case. All of the other curves are
for potential remedial actions proposed for implementation in
conjunction with the clay cap. Given a primary drinking water
standard of 50 ppb for arsenic as an example of a site
clean-up goal, the model results in Figure 5.4 can be used to
identify those actions that will lead to the greatest overall
reduction in exposure levels.
Figure 5.4 shows the importance of considering time when
comparing the performance of remedial actions. The pump and
treat and downgradient cut off wall alternatives reduce
arsenic concentrations within a few years of their
implementation, whereas the limited bottom lining,
stabilization and limited excavation alternatives take a
number of years to achieve the same reduction. However, these
three alternatives ultimately lead to the greatest reductions.
This point is more clearly demonstrated in Table 5.3. It
shows how the relative ranking of alternatives, in terms of
reductions in Cedar River contamination levels, changes
depending upon which point in time is chosen to evaluate
performance.
5.6.2 Uncertainty Regarding Remedial Action Performance
Field data are frequently insufficient to accurately
characterize site conditions. This is especially true for the
unsaturated and saturated zones. Additionally, the actual
performance of a remedial action may not be known until it has
been implemented and tested. Numerical models can be
particularly efficient and insightful tools for studying
potential uncertainties. Sensitivity analyses can be
3-56
-------
3
u.
Limited Bottom
Lining/Stabilization
25 .
o HYPOTHETICAL TINE (YEWS)
Figure 5.4 Predicted performance of different remedial action
alternatives in reducing arsenic concentrations in
the Cedar River under low flow conditions (after Cole
et al., 1984) .
3-57
-------
TABLE 5.3 RELATIVE RANKING OF POTENTIAL ALTERNATIVE LA BOUNTY
LANDFILL REMEDIAL ACTIONS AT DIFFERENT POINTS IN
TIME USING LEVEL OF CONTAMINATION REDUCTION IN THE
CEDAR RIVER AS A MEASURE OF PERFORMANCE (Taken from
the results by Cole et al., 1984)
Years After Implementation
Remedial Action 2 6 12
Downgradient cutoff 1 11
wall
Pump and Treat 2 24
Upgradient cutoff 3 55
wall
Limited excavation 5 43
Limited bottom lining 4 32
stabilization
1 = Largest reduction in contamination levels
5 = Smallest reduction in contamination levels
3-58
-------
performed by varying uncertain parameters, making runs and
observing the changes in model outputs. Such changes include
water levels, flow directions and rates, and chemical
concentrations.
Silka and Mercer (1982) used sensitivity analyses to
investigate the probable effects of installing a subsurface
drain at Love Canal, NY. Parameters evaluated included
hydraulic conductivity, effective porosity and recharge.
Figure 5.5 shows the effect of changes in shallow aquifer
hydraulic conductivity on water table elevations, as simulated
by a two-dimensional saturated zone model. Substantial
differences in elevation near the drain are predicted,
indicating that accurate specification of hydraulic
conductivity is important. Through comparison of model
predictions of drain flux and water table elevation with field
measurements, the authors were able to estimate average or
bulk hydraulic conductivities for the shallow aquifer.
Mercer et al. (1983) conducted both sensitivity analyses and
more rigorous uncertainty analysis in a later remedial action
evaluation at Love Canal. Their sensitivity analyses
considered conditions along two of the model boundaries,
aquifer transmissivity, confining bed hydraulic conductivity
and shallow system water levels. Their uncertainty analysis
considered the effects of uncertainties in hydraulic
conductivity and porosity on contaminant travel times. A
Monte Carlo technique was used to select conductivity and
porosity values from estimated frequency distributions.
5.6.3 Optimization of Remedial Action Design
Once remedial action alternatives have been identified, their
design may be refined as part of the development of a
conceptual design. Optimizing a design involves evaluating
alternative locations, pumping rates and remedial action
configurations to identify which specific combination will be
most effective. Modeling is ideally suited to this type of
analysis because a number of alternative designs can be
evaluated rapidly and quantitatively.
Cole et al. (1984) evaluated several alternative designs for a
proposed upgradient cut off wall at the La Bounty Landfill.
They showed that by changing the location of the cut off wall
and by lowering the head in the subsurface drain located on
the upgradient side of the wall, arsenic loadings to the Cedar
River could be reduced by about 30 percent. This reduction
was sufficient to make the upgradient cut off wall a feasible
alternative.
Anderson et al. (1984) analyzed alternative remedial action
3-59
-------
UJ
_l
UJ
1 73
•» MODELED K : 1(T4m/s
— MODELED K r 10'7m/s
17J
CAYUQA CREEK
FRENCH DRAIN
17 1
aoo 400 too
DISTANCE FROM DRAIN (m)
• 00
Figure 5.5
Predicted effects of two values of hydraulic
conductivity on the shape of the water table
with installed French Drain (from Silka and
Mercer, 1982}.
3-60
-------
designs for the Lipari landfill in New Jersey using a
two-dimensional (x-y) ^inite difference model of the saturated
zone. Actions simulated included slurry walls, drains and
clay caps, alone and in combination. The effect of drain
depths, a partial or full clay cap and a slurry barrier wall
on discharge of ground water to seeps and to drains over time
were estimated. Figure 5.6 shows predicted variations in
drain discharge with time for different drain depths.
Cohen and Mercer (1984) used a two-dimensional (x-y) model to
evaluate proposed additional remedial measures for Love Canal.
They analyzed the effectiveness of different designs that
included a concrete cut off wall and a synthetic cover. The
simulation results showed that the cut off wall would provide
only a minor reduction in drainage to the French Drain
compared to the cover. As Cohen and Mercer note, based in
part on the model results, the State of New York decided not
to construct the proposed cut off wall.
Optimization of remedial action design can and has been taken
one step further in sophistication. Mathematical programming
such as linear or quadratic programming can be used in
conjunction with numerical modeling to directly optimize
pumping rates and well locations, eliminating a tedious trial
and error search. This technique involves defining an
objective function (such as minimizing the costs of pumping)
and a set of constraints that might require that certain
hydraulic gradient or head conditions are met. The
mathematical algorithm then finds the optimal solution for the
given problem. Atwood (1984) used linear programming to
optimize well selection and pumping rates in a hydraulic
containment and extraction design. The optimal design called
for an outer set of wells to initially stabilize contaminated
ground water. As the plume diminishes in size, an inner set
of wells is determined to be more efficient. An optimal
schedule of extraction and injection rates for the sixteen
year clean-up period was also determined by the program.
Another example of combining optimization techniques with
numerical modeling to design extraction well systems is
discussed by Shafer (1984). Gorelick (1983) reviews the
state-of-the-art research on these management techniques for
water quality and water allocation problems. A more recent
article by Gorelick et al. (1984) introduces the use of non
linear programming techniques.
5.6.4. Assessment of Design Life and Impacts of Failure
All remedial actions have a finite life that needs to be
considered when evaluating their performance. Covers erode,
drainage systems clog and wells collapse. Remedial actions
3-61
-------
350
300
250
o
o
o
(0
z
5
K
O
ui
O
X
O
in
200
ISO
a 100
- \
SO -
• RUN 7 FULL 30 FT DRAIN
© RUN 8 FULL 120 FT DRAIN
• RUN 14 PARTIAL 30 FT DRAIN
(NO SLURRY WALL. FULL CLAY CAP)
ARITHMETIC MEAN
HARMONIC MEAN
3 4
TIME
-------
can also simply fail, either catastrophically or
progressively. Synthetic liners for example, should be
effective for a number of years. However, if they are not
properly selected, installed and used, they can fail in
relatively short periods of time. Recent research has shown
that the permeability of clay liner materials can change over
time if exposed to hydrophobic pollutants (Green et al.,
1983). The permeability of bentonite slurry materials used in
impermeable barriers has also been found to increase in the
presence of certain organic and inorganic compounds (Spooner
et al., 1983).
Numerical modeling has not been used to any large extent to
evaluate the impact of these types of failures when assessing
remedial action performance. In at least one situation,
however, design life and failure mode considerations were
incorporated into a model-based analysis of new waste disposal
facilities. A multi-disiplinary team was assembled by the EPA
Office of Solid Waste to investigate the influence of site
conditions, disposal facility design, and failure mode on
leachate migration. As described by Brown et al., (1984),
three numerical models were used. HELP (Schroeder et al.,
1984a and 1984b) was used to estimate leachate generation,
vertical movement through the facility and release through the
liner into the unsaturated zone. The Pesticide Root Zone
Model (Carsel et al., 1984) was used to predict the transport
of leachate vertically through the unsaturated zone. The
Combined Fluid, Energy, Solute Transport (CFEST) model (Gupta
et al., 1979) was used to estimate chemical movement in the
saturated zone.
Examples of typical model results for one of the facility
design/failure mode scenarios are given in Figures 5.7 and
5.8. Figure 5.7 shows the calculated leachate loading from
the base of a landfill with a single clay cover and a leachate
collection system. The progressive increase in leachate
loading over the first 20 years shows the impact of increasing
facility size by opening new waste cells. The slight decrease
in loading is due to the installation of a cap after the
facility is closed. The rapid increase in loading after about
50 years is due to the failure of the leachate collection
system. In this scenario, the facility is sited in a humid
location with high intensity rainfall. Figure 5.7 also shows
the predicted mass loading of leachate to the saturated zone
over a 200 year time frame.
Figure 5.8 shows predicted time histories of leachate
concentration at a monitoring well 100 m downgradient from the
facility and a ground-water discharge point (i.e., stream)
300 m downgradient. The concentrations shown in this figure
are "relative concentrations." That is, Figure 5.8 shows
predicted ground-water concentrations relative to the initial
3-63
-------
£ o.e
«
M
e
z
S '
o
g 0.2
0.0
FACILITY LOADING
LOADING TO GROUND WATER
100
TIME (yr>
ISO
200
Figure 5.7 Facility leachate loading and loading to
ground water.
0.06
o
O
5 0.04
O 0.02
O
100 i
I: STREAM
so
100
TIME (yr)
ISO
200
Figure 5.8 Relative leachate concentration at
monitoring well (100 m) and fetream.
3-64
-------
leachate concentration (Co). Thus, for this scenario the
maximum relative concentration (C/Co) in the saturated zone is
0.04 or 4 percent of the original concentration.
Results such as those shown in Figures 5.7 and 5.8 were used
to evaluate the impacts of a number of facility designs and
modes of failure on ground-water quality.
3-65
-------
SECTION 6
REMEDIAL ACTION MODELING REQUIREMENTS
6.1 OVERVIEW
In using numerical models to evaluate remedial action
performance it is important to recognize that different
remedial actions can have substantially different modeling
requirements. The type of model(s) (i.e., surface,
unsaturated or saturated zone) required to simulate the
effects of an action can vary, as can the dimensionality and
grid configuration. In addition, the model parameters that
must be adjusted to represent the effects of implementing
different actions can vary. As Volume 1 notes, all of these
modeling requirements must be considered, hopefully early
enough in the Feasibility Study/Remedial Investigation process
to have an impact on the specific model(s) selected for use in
remedial action evaluation.
Section 6 seeks to define numerical modeling requirements for
specific remedial actions and groups of actions. Here,
modeling requirements refer to: 1) the type of model(s) that
are required, 2) dimensionality and grid configuration
considerations, and 3) model parameter adjustments. Guidance
is provided on sources of information and available techniques
for parameter estimation for situations where field data are
not available. The modeling requirements defined herein were,
in large part, taken from previous remedial action modeling
studies (e.g., Cole et al., 1984; Mercer and Silka, 1981;
Mercer et al., 1983; Anderson et al., 1984; and Cohen and
Mercer, 1984).
As was noted in Section 4, similarities in design objectives
and remedial action configuration made it possible to condense
the large number of available technologies into fourteen
remedial measures under the general categories of surface
control, subsurface control, and waste control. These
fourteen measures can be condensed further due to similarities
in modeling requirements. An example would be the grouping
together of bio-reclamation and chemical injection. Both of
these measures can be modeled in a similar fashion: injection
3-66
-------
and extraction wells are used and the degradation rates
assigned to different elements or blocks in the model grid are
adjusted to represent the enhanced degradation of the chemical
in the treated zone. The fourteen remedial measures were also
re-grouped because they tend to be used conjunctively. For
example, the remedial actions of capping, grading, and
revegetation were grouped together because they are often
implemented as a group to control infiltration and runoff.
Given the above, the fourteen remedial measures discussed in
Section 4, were reduced to the nine remedial action groups
shown in Table 6.1.
Prior to presenting modeling requirements for each group of
remedial measures, several key points need to be addressed.
1. Only those modeling requirements associated with a
given group of remedial measures are discussed.
Requirements associated with the use of numerical
models for site characterization and assessment are
not presented. Thus, the guidance presented herein
is in addition to that needed to develop a model of
the site.
2. Certain model parameter adjustments are highly site-
specific. Thus, it is difficult to provide guidance
on their estimation.
3. Data on certain model parameters are, on the whole,
quite sparse due to a lack of field data on the
performance of some remedial measures. In many
cases, only laboratory or pilot scale data or
parameter values from previous modeling studies are
available.
6.2 MODELING REQUIREMENTS
The modeling requirements for each group of measures are
presented in terms of the following:
1. Model Type - Model type refers to whether a surface,
unsaturated or saturated zone model, or some
combination of the three, is required.
2. Dimensionality and Grid Configuration - Dimensionality
refers to the directions (i.e., x, y, and z) of water
and chemical movement that can be simulated; grid
configuration refers to the spatial discretization
needed to represent a site and the remedial action.
3. Parameter Adjustments - Parameter adjustments refer to
3-67
-------
TABLE 6.1 REMEDIAL MEASURES
Capping, Grading and Revegetation
Surface Water Diversion and Collection
Ground-Water Pumping and Interceptor Trenches
Impermeable Barriers
Subsurface Drains and Solution Mining
Excavation
Hydraulic Dredging and Seepage Basins
Bioreclamation and Chemical Injection
Permeable Treatment Beds
3-68
-------
the model inputs that must be modified to represent a
remedial measure.
Table 6.2 summarizes the modeling requirements for each
measure. The following discussion provides more detailed
guidance.
6.2.1 Capping, Grading, and Revegetation
Capping, grading, and revegetation are often used to reduce
infiltration and control erosive runoff. Since these three
remedial actions are commonly implemented together, they were
grouped into one remedial measure. The purpose of modeling is
to: 1) estimate reductions in chemical loadings to adjacent
surface water bodies and 2) estimate reductions in
infiltration into the waste site and associated leachate
generation.
Model Type
Figure 6.1 shows two typical cap designs composed of
vegetative, barrier, gas channel, filter, and buffer layers
overlying waste materials.
Two types of models may be required to evaluate this measure:
a surface zone model and an unsaturated zone model.
Typically, the surface zone model is applied to only the upper
portion of the cap. The vegetative layer would constitute the
surface zone for the designs shown in Figure 6.1. Time series
of rainfall, potential evapotranspiration and possibly other
meterological conditions are input to the surface zone model
to generate time series of net infiltration into the layer
below the vegetative layer and time series of runoff, erosion,
and contaminant loadings from the site.
The remainder of the cap and the waste itself would be
analyzed with the unsaturated zone model. The net
infiltration time series generated by the surface zone model
can be used as a flux boundary condition in the unsaturated
zone model. This boundary condition is applied to the first
compartment representing the interface between the surface and
unsaturated zones.
Dimensionality and Grid Configuration
Either a single or a multiple land segment configuration can
be used to represent the disposal site with the surface zone
model. If runoff, erosion and chemical loadings from the site
itself are of concern, a single land segment with uniform
properties (e.g., slope, roughness and infiltration capacity)
3-69
-------
TABLE 6.2 SUMMARY OF MODELING REQUIREMENTS FOR EACH REMEDIAL MEASURE
Remedial
Measure
Dimensionality
Grid Configuration
Parameter
Adjustments
Comments
I
•~j
o
Capping, Grading SF
and Revegetation UZ
Surface Water Diversion SF
and Collection
Ground-Water Pumping SZ
and Interceptor
Trenches
Impermeable Barriers SZ
Subsurface Drains and UZ
Solution Mining SZ
Excavation UZ
SZ
Hydraulic Dredging UZ
and Seepage Basins SZ
Bioreclamation and SZ
Chemical Injection
Permeable Treatment SZ
Beds
S, M
ID/z
M
2D/x-y
2D/x-y, 2D/X-2
3D
2D/x-z
2D/x-z
ID/z
2D/x-y
ID/z
2D/x-y
2D/x-y
20/x-z
SR, ER, ET, IN,
MC, HC, PO, DS, BD
SR, ER
NW, NC
HC
NW, NC, AD
AD
MC, HC, NC, PO, BD
HC, NC
NW, NC
NW, NC, DG
AD, PO, HC, BD
Number of land segments
depends on =ite conditions
Channel segments will also
be required
3D model may be needed for
partially-penetrating
wells/drains
Dimensionality dependant
upon barrier design
Model type dependant on
site conditions
Model type dependant on
site conditions
Saturated zone model not
required if mounding not
of concern
Estimation of extent of
treated zone must be esti-
mated prior to degradation
rate
Hydraulic conductivity
adjustment dependant on
materials in treatment bed
(continued)
-------
TABLE 6.2 (continued)
U)
i
LEGEND: SF Surface zone model
UZ Unsaturated zone model
SZ Saturated zone model
S Single land segment
M Multiple land segment
ID One-dimensional
2D Two-dimensional
3D Three-dimensional
x Longitudinal direction
y Lateral direction
z Vertical direction
SR Surface roughness
ER Soil erodibility
ET Evapotranspiration
IN Interception
IF Infiltration
MC Moisture content
HC Hydraulic conductivity
DG Degradation
DS Dispersivity
PO Porosity
BD Bulk density
AD Sorption
NW Nodal water flux or
held head
NC Nodal chemical flux or
held concentration
-------
LOAM (FOR VEGETATION
...-.•TivvV •:«•. •.•'....•••.'.'•'.'.'*••'.';. •."•' '•...;•'. *.
a: •• '•'.;••• -.'«• .'•• .'•• •• ".• •'•••' • •'.••-.'•: •• v.- •-.-.-;•. :"-J-». •':• ?
•V--V•':..-':•:; :.;.Gf?AVEL (GAS CHANNEL).';'.-;;•.->•:
SURFACE
UNSATURATED
BARRIER
GAS CHANNEL
WASTE
U)
I
'///////////JdLKI (BARRIER
IWASTE
fiinnnnrimnnniimiiij]
ACTUAL CAP DESIGNS
SURFACE
UNSATURATED
ZONES
GRID CONFIGURATION
BARRIER
FILTER
BUFFER
WASTE
Figure 6.1 Two typical cap designs showing layers in each zone
(after JRB Associates, 1982) .
-------
can be used. In cases, however, where runoff from areas
surrounding the site or loadings to a nearby surface water are
of concern, multiple land segments may be required. As the
areal extent of the surface zone increases, care must be
exercised in selecting the number of land segments and their
characteristics. In addition, users should recognize the
possible need to represent channel processes should the
drainage area encompass well-defined surface drainage
features.
The minimum dimension of the unsaturated zone model should be
one, in the vertical or z-direction. As Figure 6.1 shows, the
cap and disposal site can be represented as a series of
compartments of equal thickness corresponding to the layers
below the vegetated layer and the waste materials. Each layer
can be assigned varying properties (e.g., hydraulic
conductivities and porosities), depending upon the site
conditions and choice of materials in the cap design. The
thickness of the cap and drainage layers can range from 0.5 to
1.0 meters (JRB Associates, 1982; Mercer and Silka, 1982).
Parameter Adjustments
The general surface zone model parameters that need to be
adjusted to represent the effects of capping, grading and
revegetation are:
o interception storage
o surface roughness
o infiltration capacity
o evapotranspiration rate
o soil erodibility
The first four parameters largely affect runoff and
infiltration, while the remainder affect soil erosion.
Subsection 6.3 provides guidance on the estimation of these
parameters. The simulation of this measure will not require
the adjustment of those parameters affecting chemical
migration and fate. If a prior modeling study was not
conducted during site characterization, sorption coefficients
and degradation rates will have to be estimated.
Unsaturated zone model parameters that need to be adjusted
include:
o moisture content characteristics
o hydraulic conductivity
o porosity
They include those parameters related to the hydraulic
properties of the individual layers used in the cap. In their
analysis of the clay cap at Love Canal/ Silka and Mercer
3-73
-------
(1982) used a hydraulic conductivity of 10~^ m/sec. Cole et
al., (1984) used a conductivity of 3.5x10 m/sec for the
Charles City clay cap.
In situations where a synthetic material is used as a cover, a
common assumption in modeling the unsaturated zone is to use a
zero infiltration rate; this is the assumption Cohen and
Mercer (1984) made in their analysis of a synthetic cover
extension for Love Canal. A similar asssumption can also be
made for more regional analyses of clay or synthetic covers;
Anderson et al., (1984) made this assumption in their analysis
of the Lipari Landfill.
Again, if an unsaturated zone model was not used for site
characterization, parameters related to the hydraulic
properties of the waste materials will need to be estimated,
as will those related to chemical transport and fate. The
latter include sorption coefficients, degradation rates,
dispersivities and bulk densities.
6.2.2 Surface Water Diversion and Collection
Surface water diversion and collection actions are designed
primarily to route runoff away from a hazardous waste site.
Reductions in runoff, erosion, infiltration and off-site
transport of waste constituents are the primary changes that
need to be analyzed when using models to evaluate this
measure.
Model Type
The evaluation of this remedial measure can be accomplished
with a surface zone model. As with the capping, grading, and
revegetation measure, time series of meteorological conditions
are input to the model to generate time series of runoff
losses, erosion losses and chemical loadings.
Dimensionality and Grid Configuration
A model that is capable of considering multiple land and
channel segments is required. At least one land segment is
required for the waste site, the others are required to
represent areas adjacent to the site and channels collecting
diverted runoff. Runoff from land segments upgradient from
the site can be used as input to channel segments to represent
the diversion of runoff around the site.
3-74
-------
Model Parameter Adjustments
The parameters that must be adjusted to represent the effect
of surface water diversion and collection are those related to
changes in the topography of the land surface and those
related to the addition of drainage structures. These
parameters include:
o surface/channel roughness
o soil erodibility
Parameters related to surface hydrology (e.g., infiltration
rate and interception storage) and waste constituent transport
(e.g., sorption coefficients and degradation rates) will need
to be estimated, if they are not available from an earlier
modeling study.
6.2.3 Ground-Water Pumping and Interceptor Trenches
Numerical models can be used to evaluate a number of different
changes induced by the implementation of a ground-water
pumping system or interceptor trenches. Changes in heads,
directions of water and contaminant migration, and rates of
plume withdrawal can all be evaluated.
Model Type
A saturated zone model is required to evaluate this measure.
Pumping wells are represented by assigning heads or fluxes to
nodes in the grid; injection wells are represented in a
similar manner except contaminant concentrations also have to
be assigned if any residual contamination will be reinjected
following treatment. Trenches are normally represented by
assigning heads to a line of nodes. The performance of such a
measure can be assessed by modifying the number, placement,
and withdrawal rates of the wells or trenches.
Dimensionality and Grid Configuration
At a minimum, a two-dimensional (x-y) simulation is required
to represent mounding and depression of the water table. It
is important to note, however, that a two-dimensional (x-y)
representation inherently assumes that the wells or trenches
fully penetrate the saturated zone. If field conditions
dictate that withdrawal/injection occur over specified depth
intervals or the trenches be partially penetrating, a more
rigorous three-dimensional representation may need to be used.
A two-dimensional (x-z) representation is rarely used for
wells because they must be represented as a trench or line of
closely spaced wells in the y dimension. This can create a
3-75
-------
problem when specifying pumping rates, since water is not
withdrawn from the y-direction. In addition, this
representation makes it impossible to examine the potential
for plume excursion around or between pumping wells. An x-z
representation can be used for an interceptor trench, however.
Such a representation is only reasonable however, near the
middle of the trench where flows are mainly in the x and z
directions.
In designing a grid for a ground-water pumping system or
interceptor trench both regular and irregular grid spacings
can be used. The key constraint is that a node be positioned
near the proposed location of each well/trench. If several
wells are close together, their discharges may be combined and
assigned to a single node. The grid should be designed to
accommodate a number of different well/trench locations to
avoid having to restructure the grid for each alternative.
Depending on the level of analysis, the size of the grid
blocks or elements may be reduced near the wells/trenches to
obtain greater spatial resolution of predicted heads and
ground-water flow directions. Figure 6.2 shows the grid
configuration used by Silka and Mercer (1982) to represent the
French Drain at Love Canal. Note the change in grid spacing
near the french drain.
Parameter Adjustments
The parameter adjustments for this measure are relatively
straight forward. Heads or fluxes for the nodes representing
the wells or trenches need to be specified. The heads or
fluxes can be constant or time varying. The only other
required parameter adjustment is to assign contaminant
concentrations to those nodes representing injection wells.
These concentrations will have to be estimated based on the
concentration of waste constituents in the aquifer and the
efficiency of the on-site treatment system.
6.2.4 Impermeable Barriers
As is noted in Volume 2, simplified methods can be used to
analyze only a few of the many impermeable barrier design
objectives and possible barrier configurations. For this
reason, numerical models will often be used. Of primary
interest are the extent to which a barrier will prevent
contaminated ground water from migrating away from the site or
divert uncontaminated ground water around a site.
Model Type
A saturated zone model is needed to evaluate impermeable
3-76
-------
FRENCH
DRAIN
Figure 6.7 Example x-y representation and grid used to evaluate the
French Drains at Love Canal (taken from Silka and Mercer,
1982).
3-77
-------
barriers.
Dimensionality and Grid Configuration
The required dimensionality of the saturated zone model is
highy dependent upon the design of the barrier. Barrier
designs can include: 1) an upgradient barrier keyed into an
impermeable layer, with an optional drainage system upgradient
of the wall to reduce mounding; 2) a partially penetrating,
upgradient barrier that is keyed into natural impervious
boundaries at each end; 3) a fully-penetrating downgradient
barrier with or without a drainage system; 4) both upgradient
and downgradient barriers; or 5) a fully-penetrating barrier
that completely surrounds a site.
A two-dimensional (x-y) model can be used to evaluate all of
designs where the barrier is keyed in at the bottom; as with
ground-water pumping, the use of an x-y representation
inherently assumes that the barrier fully penetrates the
aquifer and that there is no flow under the barrier. Designs
involving partially penetrating barriers or barriers keyed
into leaky formations require at least a two-dimensional (x-z)
representation. Figure 6.3 shows how an x-z representation
was used by Cohen and Mercer (1984) to evaluate the benefits
of a new cut-off wall at Love Canal. They modeled a
cross-section of the site and assumed symmetry along the
centerline of Love Canal. Thus, the grid in Figure 6.3 is
only for one-half of the total cross-section. Using an x-z
representation, however, assumes that the barrier is
infinitely long in the y-direction. Thus, flow conditions
near the end of a barrier cannot be analyzed. In relatively
complex situations a three-dimensional representation can be
used to evaluate the potential for contaminant movement both
under and around the ends of a barrier.
Simulation of the effects of an impermeable barrier involves
designing a grid with elements or blocks in the approximate
location of the barrier. Element or block widths are a
function of the barrier design; the usual thickness is around
1 meter (JRB Associates, 1982). Generally an irregular grid
spacing is used particularly in the immediate vicinity of the
barrier where directions of water movement change rapidly.
Figure 6.4 shows the grid configuration used by Anderson et
al. (1984) for the Lipari Landfill. This grid is for a
two-dimensional (x-y) representation. The grid blocks
representing the barrier are shown. Figure 6.3 shows the x-z
grid configuration used by Cohen and Mercer (1984). Notice
the variable grid spacing around the partially-penetrating
barrier and drain.
If a drainage system is used in conjunction with a barrier,
it may be represented by a set of nodes with fixed heads
3-78
-------
CATUdA CREEK
CUT-OFF WALL
LOCATION
FRENCH DRAIN
LOCATION
CENTER OF LOVE CANAL
35 ft
Figure 6.3 Two-dimensional (x-z) grid configuration used by Cohen
and Mercer to evaluate a proposed cut-off wall at Love
Canal. Copyrighted by National Water Well Association.
3-79
-------
SOO FtCT
SCALE
CONSTANT HEAD NODE
PARTIAL SLURRY WALL
SLURRY WALL EXTENSIONS
Figure 6.4
Two-dimensional (x-y) grid configuration
used by Anderson et al. (1984) to evaluate
a proposed slurry wall at the Lipari
Landfill.
3-80
-------
corresponding to the estimated water elevation in the drain.
Parameter Adjustments
The only parameter adjustment required to represent an
impermeable barrier is the hydraulic conductivity assigned to
those grid blocks or elements representing the barrier. The
actual conductivity values will depend on the material used
for the barrier and the construction method. As will be
discussed later in this section, there are some data available
for the hydraulic conductivity of soil-bentonite and
cement-bentonite slurry materials. No data were found for
other grout materials or sheet pile barriers.
6.2.5 Subsurface Drains and Solution Mining
Subsurface drains and solution mining are grouped together
because of similarities in required model type, dimensionality
and parameter adjustments. In analyzing these actions, the
primary use of a model is to determine the extent to which
leachate generation can be controlled.
Model Type
The type of model required to evaluate subsurface drains or
solution mining will depend upon site conditions. In
situations where the wastes are disposed above the water
table, an unsaturated zone model should be used. When the
wastes are inundated by ground water both unsaturated and
saturated zone models may be required. To properly represent
the effects of these actions on water movement and contaminant
migration, the unsaturated and saturated zone models may have
to be linked or, if possible, a combined unsaturated/saturated
zone model can be used.
Dimensionality and Grid Configuration
In modeling the effectiveness of these actions the primary
focus will usually be on changes in leachate generation rather
than reductions in ground water concentrations some distance
from the site. For this reason, the site can be represented
with a typical cross-section. Thus, a two-dimensional (x-z)
model can be used for both the unsaturated and saturated
zones.
A single node or group of nodes can be used to represent a
subsurface drain. Fluxes or heads must be assigned to each
node to obtain proper withdrawal rates. A similar approach
can be used to represent drains or well points used to extract
leachate generated as a result of solution mining. The
3-81
-------
injection of the chemical solvent used to mobilize
contaminants can be simulated by assigning fluxes or held
concentrations to nodes in the waste layer.
The grid spacing for either action can be regular or
irregular. Often the size of the grid will be reduced near
the nodes simulating the drains or well points. Figure 6.5
shows an example x-z grid configuration for the analysis of a
subsurface drain (Nelson et al., 1983). This grid was used in
an evaluation of leachate migration from a uranium mill
tailings disposal site.
Parameter Adjustments
No additional parameter adjustments other than assigning heads
or fluxes to selected nodes are required to evaluate
subsurface drains.
For solution mining, however, the sorption coefficient or
retardation factor must be adjusted for those elements or grid
blocks receiving the injected chemical solvent. Either
parameter needs to be reduced to reflect the effect of
increased mobility. The amount of reduction is waste and
solvent specific. No data are available on possible parameter
ranges, largely because this technology has received limited
use in the field.
6.2.6 Excavation
In the evaluation of excavation actions, models can be used to
estimate reductions in leachate quality associated with the
removal of waste materials.
Model Type
As with the previous measure, the required model type depends
on site conditions. If the wastes are disposed of above the
water table, an unsaturated zone model can be used. If the
wastes extend into the saturated zone, the use of a saturated
zone model is required.
Dimensionality and Grid Configurations
The minimum dimensionality for the unsaturated zone is a
one-dimensional (z) representation. If there are lateral
heterogenities in the waste materials or subsoils, a
two-dimensional (x-z) representation should be used. For the
saturated zone, a two-dimensional (x-z) representation is
appropriate for near field analyses. A two-dimensional (x-y)
representation is more appropriate for a more regional
3-82
-------
analysis.
Parameter Adjustments
Since limited excavation involves the replacement of waste
materials with other clean soils, those parameters related to
material properties need to be adjusted. These parameters
include moisture content characteristics, hydraulic
conductivity , bulk density, and porosity. The back fill
materials will probably be taken from a locally available
source. Values for a range of different material types are
presented in Subsection 6.3.
6.2.7 Hydraulic Dredging and Seepage Basins
These remedial actions are represented in a group because of
similar simulation requirements. Hydraulic dredging is used
to remove liquids and/or sludge from lagoons or surface
impoundments. Seepage basins are used to recharge treated
water back into the ground. Such water may originate from
pumping wells or surface water diversion structures.
Model Type
An unsaturated zone model can be used for both actions. If
the extent of mounding caused by a seepage basin is of
concern, a saturated zone model can be linked to the
unsaturated zone model.
Dimensionality and Grid Configuration
A one-dimensional (z), representation would be the minimum for
the unsaturated zone. The grid configuration would be a
series of compartments or elements representing the soils
below the base of the pond or basin. Vertical variations in
soil characteristics can be represented by assigning different
properties to the compartments/elements. In cases where
lateral variations are important or seepage from the sides of
the pond/basin need to be considered, a two- dimensional (x-z)
representation is needed.
The dimensionality for the saturated zone model would be x-y.
Such a representation would make it possible to predict
changes in water table elevations (i.e., mounding) produced by
recharge from a seepage basin.
Parameter Adjustments
The required parameter adjustments are limited to changing
3-83
-------
OJ
I
00
Figure 6.5
Example representation and grid for a drain system used to
evaluate Uranium mill tailings seepage into the unsaturated
zone {after Nelson et al., 1983).
-------
heads or fluxes to represent the removal of water and waste
materials as in the case of hydraulic dredging or to represent
the ponding of water as in the case of seepage basins.
6.2.8 Bioreclamation and Chemical Injection
The simulation requirements for these actions are similar to
those for ground-water pumping. The major difference is in
the parameter adjustments required to simulate the in-situ
treatment of waste constituents.
Dimensionality and Grid Configuration
The dimensionality and grid configuration requirements are
basically the same as for a ground-water pumping action:
two-dimensional (x-y) with either regular or irregular grid
spacing. The analysis of partially-penetrating injection/
withdrawal wells may require a three-dimensional model.
Parameter Adjustment
Heads or fluxes and held concentrations must be specified for
those nodes representing injection wells. The held
concentrations will depend upon the efficiency of the on-site
treatment system. Withdrawal wells are represented by
assigning nodes or fluxes to appropriate nodes.
The effects of chemical injection and bioreclamation require
that the degradation rate assigned to some of the grid blocks
or elements be adjusted. Such an adjustment is complicated by
the fact that the elements requiring adjustment will change
with time as the chemical or bacteria migrate away from the
injection wells. Few,if any, saturated zone models can handle
such changes in parameter values unless they offer a restart
capability. Therefore, the only way to evaluate the
performance of these technologies is through a steady-state
analysis. Such an analysis would initially involve using a
flow model in a steady-state mode, to identify the region of
the flow field influenced by the injection and extraction
wells (i.e., the treated zone). All water within this region
is influenced by the wells, while all water outside the region
is influenced by the regional ground-water flow system. Once
the region has been identified, the steady-state flow field
can be input to a transport model. The degradation rate for
those elements encompassing the region can be assigned values
typical of those for the action. The degradation rates for
the other elements would remain unchanged.
Limited data are currently available on ranges of degradation
rates for bioreclamation and chemical injection.
3-85
-------
6.2.9 Permeable Treatment Beds
The purpose of modeling a permeable treatment bed is to
determine the extent to which plume concentrations are reduced
as a result of in-situ treatment.
Model Type
A saturated zone model is appropriate for most analyses.
Dimensionality and Grid Configuration
In most situations, a two-dimensional x-z representation can
be used. In using a x-z representation it is impossible to
evaluate the possibility for plume excursion around the ends
of the treatment bed. An x-y representation would be required
if plume excursion is of concern. Such a representation,
however, presumes the treatment bed is fully-penetrating.
The only major consideration in designing the grid is to
ensure that grid blocks or elements are established to
represent the treatment bed. Since the treatment beds are
designed to have permeabilities similar to the surrounding
media, they generally will not alter ground-water flow
patterns significantly (JRB Associates, 1982). For this
reason, there is no need to modify grid spacing near the
treatment bed.
Parameter Adjustments
Assuming the permeability of the material selected for the
treatment bed is similar to that for the surrounding media
only the retardation factor assigned to those grid blocks or
elements representing the treatment bed need to be adjusted.
In some models, the retardation factor is actually adjusted.
In others, the sorption coefficient and/or porosity are
adjusted so that the internally calculated retardation factor
is correct. Only limited data are available on ranges of
parameter values for treatment bed materials. Roberts (1982)
reported that equilibrium partition coefficients for activated
carbon can range between 0.005 to 0.1 ml/gin.
If the permeability of the bed materials is different, the
material properties assigned to the elements representing the
bed also need to be adjusted. No data were found on the
in-situ permeability of activated carbon or crushed limestone
materials. Subsection 6.3 provides parameter estimation
guidance for natural aquifer materials.
3-86
-------
6.3 PARAMETER ESTIMATION GUIDANCE
The parameters requiring adjustment to simulate the remedial
measures discussed in the previous section can be grouped as
follows: 1) surface zone modeling parameters, and 2)
unsaturated and saturated zone modeling parameters. This
section seeks to provide sources of data and techniques for
the estimation of selected model parameters.
The guidance presented herein is only meant to be used in
support of, rather than in place of on-site field
measurements, sampling and laboratory studies. To the extent
possible, values for model parameters should be determined as
part of the site characterization phase of the Remedial
Investigation process. This process is meant to fill
limitations in the existing data base for a site and provide
the data required to evaluate remedial action alternatives
(EPA, '1984). Hopefully, this section can be used to more
fully understand those data required for remedial action
modeling and, in the absence of site specific data, aid in
parameter estimation.
Where available, data sources and estimation techniques
pertinent to remedial action specific parameters are provided.
Both are extemely limited, however. For this reason, more
general data sources and estimation techniques are discussed
to provide a basis for at least the initial determination of
appropriate parameter values.
6.3.1 Surface Zone Model Parameter Guidance
The key surface zone model parameters requiring adjustment are
those related to: 1) channel/surface roughness, 2)
evapotranspiration, 3) interception, 4) infiltration, and 5)
soil erodibility. Available guidance on the estimation of
remedial action performance is provided below.
6.3.1.1 Channel/Surface Roughness—
In most surface zone models, the roughness of land or channel
segments is defined in terms of a parameter known as the
roughness coefficent, the most common being Manning's "n."
Donigian et al., (1983) note that most of the published values
for Manning's "n" are for channel rather than overland flow.
Most standard open channel flow references provide ranges of
values for different channel types. Table 6.3 lists values
for lined and unlined channels typical of those that might be
3-87
-------
TABLE 6.3 CHANNEL AND LAND SURFACE MANNING'S 'n1
VALUES APPLICABLE TO REMEDIAL ACTION
MODELING
Channel Type* Manning's 'n' Value
Smooth concrete 0.012
Ordinary concrete 0.013
Shot concrete, untroweled and
earth channels in good condition 0.017
Straight unlined earth channels in
good condition 0.020
Grass covered waterways 0.02-0.4
Land Surface Condition**
Smooth fallow 0.15-0.20
Rough fallow, cultivated 0.20-0.30
Light turf 0.25-0.35
Heavy turf 0.30-0.40
* Values taken from Chow (1964)
** Values recommended by Donigian et al. (1983)
3-88
-------
constructed at a hazardous waste site. As this table shows,
Manning's "n" values for grass covered waterways can be
highly variable. The actual value depends upon the vegetal
retardance, flow velocity and hydraulic radius of the channel.
Chow (1964) provides guidance on the estimation of appropriate
values given the design of the waterway.
Table 6.3 also lists ranges of values recommended by Donigian
et al., (1983) for different land surface conditions.
Specific published values include those by Ree et al. (1977).
They calculated values of 0.25 to 0.62 for grass cover
conditions. Ross et al. (1977) used values of 0.35 to 0.40
for agricultural areas and 0.30 for forested areas.
6.3.1.2 Evapotranspiration—
Evapotranspiration is the process by which water is carried
from the soil by either direct vaporization from the soil or
by transpiration of plants. The maximum rate of
evapotranspiration (potential ET) depends on the demand from
the atmosphere and the nature of the evaporating surface be it
soil or plant. The actual rate depends on the moisture
available to evaporate from the surface and the soil. Linsley
et al. (1982) discuss methods for calculating ET based on
water and energy budget methods, meteorological data and pan
evaporation data.
One common method is a two-phased approach. First, the
potential ET is calculated using pan evaporation data. Most
U.S. weather stations provide pan evaporation data along with
other standard meteorological data. The pan factor is then
used to convert the daily pan evaporation data into daily
potential ET. The second phase involves calculating the
actual ET from the surface and soil, based on the available
water and potential ET. Models using this method only require
the appropriate pan evaporation factor along with pan
evaporation data from a nearby weather station. Figure 6.6
provides pan factors for the entire contiguous United States.
6.3.1.3 Interception—
The interception parameter in surface zone models represents a
storage depth or volume for precipitation that is trapped on
the surface of vegetation. Precipitation in excess of the
interception storage is assumed to reach the soil surface.
Interception storage is related directly to the density of the
vegetation cover. Several publications provide ranges of
values for different agricultural crops (Woolhiser, 1976;
Donigian et al., 1983; Knisel, 1980; and Carsel et al., 1984).
Typical values range from 0.0 to 0.25 cm. Table 6.4 lists
general ranges of values for different vegetation densities.
3-89
-------
79
JS1*
u>
I
79
75 SB
70
Figure 6.6 Pan factors (source: Carsel et al., 1984)
-------
TABLE 6.4 INTERCEPTION STORAGE FOR DIFFERENT VEGETATIVE
DENSITIES
Density Interception Storage (cm)
Light 0.0 - 0.15
Moderate 0.20 - 0.30
Heavy 0.30 - 0.45
3-91
-------
6.3.1.4 Infiltration —
Infiltration is a complex process that depends on many
physical factors: 1) the soil type, 2) antecedent moisture, 3)
organic matter, 4) vegetative cover, and 5) rate of water
supply to surface. The sophistication with which surface zone
models handle infiltration varies. Often the infiltration
rate is calculated within the model and does not require any
special parameters. Two examples of infiltration estimations
are presented.
One of the simplest approaches is the calculation of the
average infiltration rate, the W index (see Linsley et al . ,
1982):
W = £ (P - Qs - S) (6.1)
where: W = average infiltration rate, L/T
t = duration of precipitation, T
P = total precipitation during time, L
Qs = surface runoff, L
S = effective surface retention, L.
Another approach developed by Holtan et al . , (1975)
incorporates the effects of vegetative cover in the
calculation of the maximum infiltration rate, f:
(6.2,
where f = infiltration rate, L/T
a = vegetative parameter, L/T (see Table 6.5)
S = soil water capacity exceeding wilting point,
S = soil water in excess of wilting point, L /L
fc = minimum infiltration rate after prolonged
wetting, L/T
The maximum infiltration capacity as suggested by the equation
above depends on the antecedent moisture content. The minimum
infiltration rate, f , is the saturated hydraulic
conductivity. Once the°profile is saturated, the infiltration
rate is limited by the speed at which water can move in the
soil represented by the saturated hydraulic conductivity.
6.3.1.5 Soil Erodibility —
A range of algorithms are used in surface zone models to
simulate the process of soil erosion. Some are based on more
3-92
-------
TABLE 6.5 VALUES OF 'a' FOR EQUATION (6.2)
(From Holtan et al., 1975)
(in./hr per in.**1.4 of
available storage)
Cover
Fallow
Row crops
Small grains
^ Legumes
Sod
Pasture
Bunchgrass
Temporary (sod)
Permanent (sod)
Woods and forests
Poor Condition
0.10
0.10
0.20
0.20
0.40
0.20
0.40
0.80
0.80
Good Condition
0.30
0.20
0.30
0.40
0.60
0.40
0.60
1.00
1.00
3-93
-------
mechanistic descriptions while others are strictly empirical.
Thus, it is difficult to provide guidance on the adjustment of
specific parameters because they are often model-dependent.
Some general guidance can be provided, however, since most of
the commonly used algorithms were derived to take advantage of
the wealth of information generated by the Soil Conservation
Service in their development of the Universal Soil Loss
Equation (USLE):
Y(s)
A (R • K • LS • C • P) S
(6.3)
where Y(s) = sediment loading from surface erosion,
tons/year
A = drainage area, acres
R = rainfall factor, expressing the erosion
potential of average annual rainfall
K = soil erodibility factor, expressed in tons
per acre per R unit
LS = topographic factor, a combination of the
slope-length and slope-steepness, dimen-
sionless
C = Cover management factor, representing the
degree of soil disturbance and vegetative
cover density, dimensionless
P = erosion control practice factor, accounting
for practices that act to reduce erosion,
dimensionless
S-, = sediment delivery ratio, dimensionless
Parameter values for most of the factors are well documented
for agricultural areas. However, values have also been
derived for construction and mining conditions that would be
applicable to hazardous waste sites. Detailed guidance and
estimation methods are available for each factor. Rather than
repeat it herein, it is suggested that the following sources
be consulted: Wischmeier and Smith, 1978; EPA, 1975; and
Mills et al., 1982.
6.3.2 Subsurface Modeling Parameters
The subsurface modeling parameters are divided into two
categories: 1) flow-related parameters and 2)
transport-related parameters. Flow parameters affect the
hydraulic flow field and the general velocity of the ground
water. The transport-related parameters affect the migration
and fate of the contaminant.
3-94
-------
6.3.2.1 Flow-related Parameters—
Three key parameters that affect the flow of water are
moisture content characteristics, hydraulic conductivity and,
for transient flow problems, the storage coefficient. All
three of these parameters vary over a large range and are
highly specific to the materials being considered.
6.3.2.1.1. Moisture Content Characteristics—In the
unsaturated zone, where all the pores are not filled with
water, the soil moisture content is an important physical
property which affects plasticity, strength, microbial
activity and the chemical state of the soil. The negative
pressure head (suction) determines the degree with which water
is held in the soil matrix. This is the amount of suction
that is required to remove the water. At zero pressure all
the pores are filled with water and the soil is saturated. As
the pressure decreases (suction increases) the water content
decreases.
The relation between moisture content and pressure head is
described by the soil moisture characteristic curve. The
shape and intercepts of this curve depends on the physical
properties of the soil matrix: the pore size distribution,
grain size, and mean pore diameter. Figure 6.7 illustrates
the influence of soil type on the characteristic curve.
Empirical measurements in the lab or field must be made for
each site to accurately determine the characteristic curve.
Hillel (1982), discusses methods for measuring water content
and pressure head.
In the absence of any laboratory or field data several methods
have been developed to determine the moisture content at given
pressure heads. One method, developed by Rawls and Brakensiek
(1982), uses bulk density and organic matter content, as well
as soil texture. They developed the following regression
equation to estimate the water content at given negative
pressure heads (suction):
0= a + (b x sand%) + (c x clay%) + (d x organic matter)
+ (e x bulk density (gr/cm^ ))
(6.4)
where 0 = water content, L3/L3
a-e = regression coefficients
Table 6.6 shows the values of the regression coefficients to
to be used at selected pressure heads.
3-95
-------
c
o
u
3
cn
Clayey soil
Sandy soil
Water content
Figure 6.7 The effect of soil type on soil-water retention
(source: Hillel, 1982). Copyrighted by Academic
Press.
3-96
-------
TABLE 6.6 COEFFICIENTS FOR LINEAR REGRESSION EQUATIONS FOR
PREDICTION OF SOIL WATER CONTENTS AT SPECIFIC
MATRIC POTENTIALS (source: Carsel et al. , 1984)
Ma trie
Sand
Intercept (%)
Coefficient a
-0.20
-0.33
-0.60
-1.0
-2.0
-4.0
-7.0
-10.0
-15.0
0.4180
0.3486
0.2819
0.2352
0.1837
0.1426
0.1155
0.1005
0.0854
b
-0.0021
-0.0018
-0.0014
-0.0012
-0.0009
-0.0007
-0.0005
-0.0004
-0.0004
Clay
(%)
c
0.0035
0.0039
0.0042
0.0043
0.0044
0.0045
0.0045
0.0044
0.0044
Organic
Matter
(%)
d
0.0232
0.0228
0.0216
0.0202
0.0181
0.0160
0.0143
0.0133
0.0122
Bulk
Density
(g cm~3)
e
-0.0859
-0.0738
-0.0612
-0.0517
-0.0407
-0.0315
-0.0253
-0.0218
-0.0182
R2
0.75
0.78
0.78
0.76
0.74
0.71
0.69
0.67
0.66
3-97
-------
6.3.2.1.2 Hydraulic Conductivity — The hydraulic conductivity
is a measure of the ease with which a fluid is transmitted
through a porous medium. It is one of the most important and
most variable physical properties governing flow in both the
saturated and unsaturated zones. For remedial action
modeling, the correct assessment of hydraulic conductivities
for example in a slurry trench is critical for the accuracy of
a modeling effort. Both the fluid properties and the media
properties contribute to the hydraulic conductivity:
K
where K = hydraulic conductivity, L/T 2
k = intrinsic permeability of the porous medium, L
*? = viscosity of fluid, M/TL 2
g = gravitational constant, L/T
P = fluid density, M/L3
For most studies, the fluid of concern is water. However when
a dissolved contaminant affects the density and viscosity of
the ground water, these factors should be considered.
Viscosity accounts for the fluid's internal resistance to
flow. Density compensates for the effects of gravity. The
tables (and equations) that follow assume water is the fluid.
The effects of the media on hydraulic conductivity are
encompassed in the intrinsic permeability. Attempts have been
made to quantify the media effects based on porosity, pore
size distribution, and surface area without much success. The
Kozeny-Carman theory estimates the hydraulic conductivity of
the well graded sands based on a pore shape factor, porosity,
specific surface area and tortuosity. Its application however
is limited by the difficulty to measure the dependent
variables.
The most reliable method is to directly measure the hydraulic
conductivity in the field or laboratory. Roberts (1984)
describes a variety of. measurement techniques for hydraulic
conductivity. In the absence of any field or laboratory
measurement, Table 6.7 lists order of magnitude estimates of
hydraulic conductivity for selected materials that could be
encountered in remedial action modeling. For more detailed
guidance on hydraulic conductivities relating to slurry walls
and grout curtains, see JRB Associates (1984) and Shafer et
al. (1984).
In a .previous modeling study Cohen and Mercer (1984) used 3.5
x 10 m/sec as a conductivity for a proposed concrete cut-off
wall at Love Canal. Anderson et al. (1984) used conducivity
of 10~6-times lower than that of the surrounding aquifer
materials in their analysis of a slurry wall at the Lipari
3-98
-------
TABLE 6.7 RANGES OF HYDRAULIC CONDUCTIVITIES FOR
DIFFERENT MATERIALS (adapted from Spooner
et al., 1983, Freeze and Cherry, 1979, and
Morris and Johnson, 1967)
Hydraulic Conductivity
Material (m/s)
Clay 10-10-10-8
Soil Bentonite 5xlO~10 - 10~7
Cement Bentonite 10~8
Silt/Loess 10~9 - 10~5
Sand
fine 10~5 - 10""
medium 10"1* - 10~3
course 10~3 - 10~2
Gravel 10~3 - 10"1
3-99
-------
Landfill. si^-ka and Mercer (1982) used conductivities of 10~5
m/sec and 10 m/sec for barrier drain materials and a clay cap
in a analysis of Love Canal.
In the unsaturated zone the hydraulic conductivity depends on
soil moisture and the pressure head, as well as fluid and
media properties. The hydraulic conductivity in the soil can
vary over several orders of magnitude simply depending on the
soil moisture. The physical properties of the soil influence
the relation between hydraulic conductivity and suction as
shown by the curve in Figure 6.8.
As for the soil-moisture characteristic curve, it is best to
determine the relationship between hydraulic conductivity and
pressure head in the field or laboratory. Alternately, the
unsaturated hydraulic conductivity can be determined by
knowing the saturated hydraulic conductivity and the
soil-moisture curve. van Genuchten (1978b) has developed a
closed form analytical solution for unsaturated hydraulic
conductivity based on the soil-moisture characteristic curve.
6.3.2.1.3 Storage Coefficient—The storage coefficient is the
volume of water released from storage per unit surface area of
aquifer for a unit decline in the water table elevation or
piezometric surface in an unconfined or confined aquifer.
This parameter is necessary when simulating transient flow
conditions.
In unconfined aquifers the storage coefficient more commonly
known as the specific yield, is much higher in magnitude than
the storativity of a confined aquifer. Specific yield
generally ranges from 0.01 to 0.30 (Freeze and Cherry, 1979).
Table 6.9 in the section on porosity shows typical values for
selected materials (porosity and effective porosity).
In confined aquifers storativity ranges from .005 to .00005
(Freeze and Cherry, 1979). The higher values correspond to
aquifers with more easily compressed materials. The water
released from storage comes from the compaction of the aquifer
material and the expansion of water under lower-pressure. In
an unconfined aquifer the water released from storage actually
drains from the aquifer. These two processes account for the
difference in magnitude.
6.3.2.2 Transport-Related Parameters—
The important transport parameters that require adjustment for
remedial action modeling are: 1) dispersivity, 2) porosity,
3) bulk density, 4) sorption coefficient, and 5)
degradation rate. All five of these parameters affect the
movement of hazardous wastes constituents and must be
correctly adjusted to represent the effects of different
3-100
-------
Sandy soil
Clayey soi
Figure 6.8
Suction
Dependence of conductivity on suction in
soils of different texture—log-log scale
(source: Hillel, 1982). Copyrighted by
Academic Press.
3-101
-------
remedial actions.
6.3.2.2.1 Dispersivity—Dispersion is the spreading of a
contaminant due to two processes: 1) velocity- and
flow-related mechanical dispersion and 2) molecular
diffusion. Mathematically, the dispersion coefficent is
estimated by:
D = a V + D
m
(6.6)
where
D =
a =
V =
D =
m
dispersion coefficient, L /T
dispersivity, L
mean ground-water velocity, L/T
molecular diffusion, L^/T
In most modeling studies the molecular diffusion is considered
negligible. Mechanical dispersion is the product of
dispersivity and ground-water velocity. It is the.result of
velocity variations in an individual pore space and in pores
of differing sizes and because of the tortuous flow path that
water must take around the grains in the porous medium.
Experimental and theoretical work has shown that dispersivity
is scale dependent. Lab experiments report small
dispersivities on the order of centimeters. Field scale
experiments have dispersivities ranging from a few meters up
to hundreds of meters (see Figure 6.9). However, there has
been little success in developing techniques for accurate
estimation. For the most part dispersivity has been used as a
model calibration parameter, not necessarily reflecting a
physically meaningful number (see Anderson, 1984 for more
detailed discussion). Table 6.8 shows dispersivity values
that have been measured and used in other modeling studies.
6.3.2.2.2
Porosity—Porosity is defined as the percent void
In
In
volume in a representative volume of the porous medium.
the saturated zone this entire volume is filled by water.
the unsaturated zone the pore space is filled by both water
and air. In modeling contaminant transport, porosity is
necessary to determine the average ground-water velocities and
associated contaminant velocities. For a given average flux
rate (specific discharge), a porous medium with high porosity
will have a slower pore water velocity than material with a
low porosity. Low porosity material has fewer voids for water
to flow through so a higher velocity is necessary for the
specific discharge to equal that of a high porosity material.
Porosity depends on particle size, particle size distribution
and degree of lithification. For a single particle size
class, in general, the larger the particle size the higher the
porosity. However, if the aquifer material can fill in the
3-102
-------
100
c/)
o:
LU
Q_
O
ID
t I
i n
_ SAND.GRAVEL,
* SANDSTONE
A LIMESTONE, BASALT,
GRANITE a SCHIST
I 10 100 1000
DISTANCE (m)
Figure 6.9 Variation of dispersivity with distance
(source: Anderson 1984"; adapted
from Lallemand-Barres and Peaudecerf,
1978) .
3-103
-------
TABLE 6.8
SMALL SCALE AND REGIONAL DISPERSIVITY VALUES
(adapted from Anderson, 1984)
Localized Scale
Distance
u>
I
Aquifer Type
Alluvial
Fractured Dolomite
Fractured Chalk
Chalk
Location Between Wells
Chalk River, Ontario
Lyons , France
Barstow, CA 6.4
Tucson, AZ 80
Carlsbad, NM 38.1-54.9
Dorset, England 8
Dorset, England 8
Ufc ^V /
ax /ay
0.34 - 1
5.0
12.0 .9 - 3.9
8.0 8 - 530
5.0 .34 - 34.5
7.0 7.0 - 780
12.0 3.0
15.2
15.2
38.1
3.1
1
Aquifer Type
Alluvial
Limestone
Glacial Deposits
Glacial Till
Location
Rocky Mt. Arsenal, CO
Colorado
California
Sutter Basin, CA
Alsace, France
Cutler, FL
Brunswick, GA
Long Island, NY
Alberta, CA
Regional Scale
Distance
Between Wells
30.5
30.5
30.5
80 - 2000
15
22
61
21.3
3-6
1
3.3
3.3
10
15
10.0
3.3
5
5
-------
space between large particles, the porosity will decrease.
Lithification decreases porosity by compacting the sediments
and eliminating the pore space.
Porosity and effective porosity should be distinguished. Most
transport equations use effective porosity which does not
include dead-end and unconnected pores. Effective porosity
approximately equals the specific yield, which is the amount
of water that will drain from a given saturated soil sample.
Table 6.9 lists ranges of porosity and effective porosity for
selected materials that may be used in different remedial
actions.
6.3.2.2.3 Bulk Density—Bulk density is the dry particle mass
per unit volume of soil. It is a basic property in the
estimation of the retardation factor. Table 6.10 and Figure
6.10 can be used to estimate values for different materials.
6.3.2.2.4 Sorption Coefficients—The most common sorption
parameter in unsaturated and saturated zone models is the
partition coefficient or Kd. In using this parameter to
simulate sorption, it is assumed that the process is linear,
completely reversible and rapid relative to the time step in
the model. Where these assumptions are not valid, other
descriptions must be used (e.g., Freundlich and Langmuir
isotherms). Cherry et al. (1984) discuss the assumptions and
important limitations inherent in the use of different
isotherms to describe the sorption process, particularly with
respect to inorganic pollutants whose mobility is controlled
by precipitation/dissolution, oxidation/reduction and chemical
speciation reactions. Both Cherry et al. (1984) and Rai et
al. (1984a) discuss the role these reactions play in
controlling the chemical mobility.
There are three basic approaches for estimating sorption
parameter values. The most preferred is to conduct in-situ
tracer experiments or laboratory batch or column experiments
using soil and ground-water samples from the site. In many
cases, however, time and resource limitations preclude the use
of such procedures.
A second approach is to use literature data derived from field
or laboratory experiments. There are a growing number of
useful compilations of sorption data for both organic and
inorganic pollutants,. Dawson et al. (1980) reviewed
literature data on 250 chemicals commonly found in hazardous
waste streams. Available determinations of K
-------
TABLE 6.9 RANGES OF POROSITY AND EFFECTIVE POROSITY VALUES
FOR SELECTED MATERIALS (sources: Morris and
Johnson, 1967 and Davis and DeWeist, 1966)
Material
Porosity
Range Average
Effective Porosity
Specific Yield
Range
Average
Clay
Silt
Sand
Gravel
fine
medium
coarse
fine
medium
coarse
33-65
33-61
25-53
27-49
30-46
25-40
24-44
24-35
4/
46
43
39
39
34
32
28
0-18
1-39
1-46
16-46
18-42
12-40
17-43
12-26
b
15
30
32
29
28
24
22
3-106
-------
TABLE 6.10 RANGE OF BULK DENSITY (gm/cm3) FOR DIFFERENT
MATERIALS (source: Morris and Johnson, 1967
and Baes and Sharp, 1983)
Material
Sandstone fine
medium
Siltstone
Claystone
Shale
Sand fine
medium
coarse
Gravel fine
medium
coarse
Silt
Clay
Silt Loams
Clay and Clay
loams
Sandy Loams
Silt Loams
Loams
All Soils
Range
1.34 -
1.50 -
2.52 -
2.50 -
2.47 -
1.13 -
1.27 -
1.42 -
1.60 -
1.47 -
1.69 -
1.01 -
1.18 -
0.86 -
0.94 -
1.25 -
1.02 -
1.16 -
0.86 -
2.32
1.86
2.89
2.76
2.83
1.99
1.93
1.94
1.99
2.09
2.08
1.79
1.72
1.67
1.54
1.76
1.58
1.58
1.76
Mean
1.76
1.68
2.65
2.66
2.69
1.55
1.69
1.73
1.76
1.85
1.93
1.38
1.49
1.32
1.30
1.49
1.22
1.42
1.35
3-107
-------
o
00
_g
o
10 20
30 40 50 60 70
Sand (%)
90 100
Figure 6.10 Mineral bulk density (g/ciri)
(source: Carsel et al.f 1984)
-------
on sorption parameters, including details on experimental
conditions (i.e., soil properties, water chemistry and
laboratory methods). Published data on the sorption of
inorganic constituents commonly found in utility solid wastes
can be found in a recent publication by Rai et al. (1984a and
1984b). The first volume provides summaries of sorption
parameter values for each constituent, while the latter is an
annotated bibliography of over 350 publications. In using any
of the above databases, it is critical to remember that
sorption parameters are only applicable to the experimental
conditions under which they were measured. The application of
parameter values to other conditions should be approached with
great care.
The final approach is to use one of the many available
empirical estimation techniques that provide relationships
between the sorption coefficient and other basic
physical-chemical properties (e.g., solubility and octanol
water partition coefficient). One useful summary of available
relationships can be found in Lyman et al., (1982); this
summary is given in Table 6.11. Most of the available
relationships give
K
oc
or organic carbon partition
coefficient. A Koc can be converted to a Kd by:
Kd ~ Koc 'foc
(6.7)
where
oc
= weight percent of the solid phase composed of
organic carbon.
Typical values of foc range from 0.4 to 10.0 percent (Brady,
1974). Values of solubility or octonal-water partition
coefficients can be found in a number of compilations of
chemical properties (Dawson et al., 1980; Mabey et al., 1982;
Leo et al., 1971; Hansch and Leo, 1979; Verschueren, 1977; and
Sax, 1979). Lyman et al., (1982) provide techniques for
estimating solubility and octonal-water partition coefficients
when literature data are not available.
In using the relationships in Table 6.11, or other similar
relationships, it is important to recognize that sorption is
assumed to be "keyed" solely to the organic carbon content of
the soil/sediment (Cherry et al., 1984). McCarty et al.,
(1981) note that at low organic contents (say <1%) typical of
those found in deep aquifer materials, the inorganic
composition of a soil can have a larger affect on sorption.
McCarty et al., (1981) provide a relationship determining the
critical organic fraction:
3-109
-------
TABLE 6.11 REGRESSION EQUATIONS FOR THE ESTIMATION OF K
(source: Lyman et al., 1982.)
by McGraw-Hill
Copyrighted
oc
Regression Equations for the Estimation of K
oc
ES No.
41
44
4.7"
48
49
4-10
4-11
4-12
4-13"
4-I4"-'
4 16
4-16
Equitlon'
log KM » -0.68 log S » 3.64 |S In mg/LI
log K.e- -0.64 logS + 0.44
(S ln.mol< Irictkinl
tog K(c- -0.667 log 5*4.377
|Sln«imolii/LI
log K>( - 0.644 log KO%¥ + 1 J77
log K.ft • 0.837 log KOW - 0.008
tog KM- 1.00 log K>f< - 0.21
log K.( • 0.94 log *.„ + 0.02
log K( » 0.681 log BCHD « 1-B63
log K(( - 0.681 log BCFIt) t 1J88
N..b
108
10
16
46
19
10
9
13
30
29
13
22
,'*
0.71
0.94
0.99
0.74
0.96
1.00
t
0.91
0.64
0.69
0.76
0.83
Chtmlc*! OUMI RtprwHitcd
Wid« nrltty, moitly ptillcUti
Moitly iromttlcor polynucku tromttlct; two chlotlniltd
Chlorlnittd hydroctrbont
Wid« vwltty. moitly p«itlcld»i
Aromitici. polynudtw •romttlct. trliilnti ind dinluo-
tnlllni rwrblcld*!
Moitly aromitlc or potynuclair tromitlct; two chlorlnnid
1-TrltilMi «nd dinllrotnllln* hxblcldtl
Vtrltty ol muciicldii. rwrtalcidti »nd (unglcld*!
Subitltuud pninylur*M tnd ilkyl-N-phtnylorbimitii
Atomtttc tompoundi: uttu. 1,3,6-trliimn. uitxmi|n.
•nd uracill
Wid< virlity, moitly p«illcld«i
Wldt vwUty, moitly piitlcld*!
Hit.
126]
1261
111)
1201
191
128)
IT)
1381
(61
118)
126)
|28|
t. K( • toll (or wdimtnt) (dioiption cotflicitnt; S • wiur tolubillty: Kc<> • ocund-w»uf pirtitlon eotllidint; BCF(I) • bioeonuntriiion Itctor
trom llowtnfwiiir mil; BCF It) * bioeonctnuitlon tutor from modtl Koiyiitmi; t • ptricnor; N • numbtr of liui In molicult which cin p»r-
Help*If In thi lormilion ol t hydiogin bond.
b. No. • numbw ol dwmietlt uud lo ebtiin r«jf«uk>n tquilion.
c. r1 • cori*lallon co«liiei«ni IM rcgranion tquviion.
d. Equation of igmtlly gi*tn In ncmi ol K>m. Th« nlniomhip K>m - K0C/1.724
-------
200 K -
ow
2
where S = silica - specific surface area, m /gm
6.3.2.2.5 Degradation Rate—The degradation of organic
pollutants can occur as a result of chemical and biological
reactions. The most important chemical reactions are
oxidation, hydrolysis and reduction. Oxidation requires the
presence of an oxidant (e.g., gaseous oxygen or free radicals
like OH or peroxy RO2). Hydrolysis involves the introduction
of a hydroxyl group into a compound. Organics susceptible to
hydrolysis can be found in Callahan et al. (1979). Reduction
involves the removal of a halogen atom through
oxidation-reduction. This process is generally only of
importance in low-redox state ground waters (Cherry et al.,
1984).
Biological reactions are generally enzymatic reactions induced
by bacteria. Until recently, biodegradation in ground water
was neglected because microbial activity was assumed to be
limited. Research by Wilson et al. (1982) and others has
shown that significant microbial activity can occur in the
saturated zone and that indigenous species are capable of
degrading selected organics. Kobayashi and Rittman (1982)
note that only broad guidelines can be given regarding the
susceptibility of compounds to biodegradation.
It is important to note that only limited data exist on
changes in biodegradation rates associated with the
implementation of in-situ treatment remedial actions. This is
in part due to the proprietary nature of certain treatment
schemes and a general lack of data on the performance of the
limited number of treatment systems that have been
implemented.
As with sorption parameters, three basic approaches exist for
estimating degradation rates for remedial action assessment.
Field and/or laboratory determinations are clearly preferred
whenever possible.
Literature data provide a second source. Table 6.12 lists the
limited data that were found for bioreclamation actions.
There are a number of useful compilations of measured rate
constants for both chemical and biological degradation
reactions (Callahan et al., 1979; Mills et al., 1982; Lyman et
al., 1982; and Dawson et al., 1980). The FOCIS database being
developed by Battelle, Pacific Northwest Laboratories for
3-111
-------
TABLE 6.12 BIORECLAMATION DEGRADATION RATES FOR SELECTED
WASTE CONSTITUENTS (Source: Personal communi-
tion with Mr. John Zikopoulos, Polybac, Inc.,
Allentown, Pennsylvania, April, 1983)
Waste Constituent Degradation Rate (/day)
Polyvinyl alcohol 0.63-2.5
Benzoic acid 0.076-1.0
Chloropropham 0.01-0.03
3-112
-------
EPA-Ada will also contain literature data on degradation
rates, including field or laboratory experimental conditions.
In using any of these sources it is again important to
recognize the difficulties inherent in extrapolating rate
constants to different site-specific conditions, particularly
since much of the published data are for surface water rather
than ground-water systems.
The third approach is to use estimation procedures such as
those given in Lyman et al. (1982). Mills (1980) notes that
oxidation and hydrolysis rates can be estimated within a
factor of 3-5 and 2-3, respectively. The predictions of
biodegradation rates is all but impossible according to Cherry
et al. (1984).
3-113
-------
REFERENCES
Anderson, M.P. 1979. Using Models to Simulate the Movement
of Contaminants Through Ground-Water Flow Systems, In: CRC
Critical Reviews in Environmental Control, Vol 9, No. 2.
Anderson, M. 1984. "Movement of Contaminants in Ground Water
Transport - Advection and Dispersion," In: Ground Water
Contamination, Studies in Geophysics, National Research
Council, National Academy Press, Washington, D.C.
Anderson, P.F., C.R. Faust and J.W. Mercer. 1984. "Analysis
of Conceptual Designs for Remedial Measures at Lipari
Landfill, New Jersey," Ground Water, Vol. 22., No. 2.
Atwood, D.F. 1984. Management of Contaminated Ground Water
with Aquifer Simulation and Linear Programming: The
Development of a Hydraulic Control Procedure, MS Thesis,
Stanford University, Stanford, CA.
Bachmat, Y., B. Andrews, D. Holtz, and S. Sebastian. 1978.
Utilization of Numerical Ground Water Models for Water
Resource Management, EPA 600/8-78-012, U.S. Environmental
Protection Agency, Environmental Research Laboratory, Ada,
OK.
Baes , C.F., III and R.D. Sharp. 1983. "A Proposal for Esti-
mation of Soil Leaching Constants for Use in Assessment
Models," J. Environmental Quality, Vol. 12, No. 1.
Barnwell, T.O., Jr. and R.C. Johanson. 1981. HSPF: A Com-
prehensive Package For Simulation of Watershed Hydrology
and Water Quality. In: Nonpoint Pollution Control: Tools
and Techniques for the Future. Interstate Commission on
the Potomac River Basin Rockville, MD.
Bedient, P.B., N.K. Springer, C. J. Cook and M. B. Thompson
1982. "Modeling Chemical Reactions and Transport in
Groundwater Systems: A Review," In: Modeling the Fate of
Chemicals in the Aquatic Environment, K. L. Dickson, A. W.
Maki and J. Cairns, Jr. eds., Ann Arbor Science Publishers,
Ann Arbor, MI.
3-114
-------
Bicknell, B.R. 1984. Modeling Chemical Emissions from
Lagoons and Landfills, Final Report, U.S. Environmental
Protection Agency, Environmental Research Laboratory,
Athens, GA.
Brady, N.C. 1974. The Nature and Properties of Soils. 8th E.
McMillan Publishing Co., New York, N.Y.
Brown, S.M., A.S. Donigian, Jr., S.B. Yabusaki and J.T.
Bachmaier. 1984. "Locational Factors Affecting Leachate
Migration," Proceedings of the 1984 Specialty Conference on
Environmental Engineering, ASCE, New York, N.Y.
Callahan, M.A., M.W. Slimak, N.W. Gable, I.P. May, C.F.
Fowler, J.R. Freed, P. Jennings, R.L. Durfee, F.C Whitmore,
B. Maestri, W.R. Mabey, B.R. Holt, and C. Gould. 1979.
Water Related Environmental Fate of 129 Priority
Pollutants, Volumes 1 and 2, EPA 600/3-82-023ab, U.S.
Environmental Protection Agency, Office of Water Planning
and Standards, Washington, D.C.
Carsel, R.F., C.N. Smith, L.A. Mulkey, J.D. Dean, and P.
Jowise. 1984. User's Manual for the Pesticide Root Zone
Model (PRZM), U.S. Environmental Protection Agency,
Environmental Research Laboratory, Athens, GA.
Cherry, J.R., R.W. Gillham and J. F. Barker. 1984. "Contami-
nants in Ground Water: Chemical Processes," In: Ground
Water Contamination, Studies in Geophysics, National
Research Council, National Academy Press, Washington, DC.
Chow, V.T. 1964. Handbook of Applied Hydrology, McGraw-Hill
Company, New York, N.Y.
Cohen, R.M. and J.W. Mercer. 1984. "Estimation of a Proposed
Synthetic Cap and Concrete Cut-Off Wall at Love Canal Using
a Cross-Sectional Model," Proceediungs of the Fourth
National Symposium and Exposition on Aquifer Restoration
and Ground Water Monitoring, National Water Well
Association, Washington, D.C.
Cole, C.R., F.W. Bond, S.M. Brown and G.W. Dawson. 1984.
Demonstration/Application of Ground-Water Modeling
Technology for Evaluation of Remedial Action Alternatives,
Draft Final, U.S. Environmental Protection Agency,
Municipal Environmental Research Laboratory, Cincinnati,
OH.
Davis, S.N. and R.J.M. DeWiest. 1966. Hydrogeology, John
Wiley and Sons, New York, N.Y. 463 p.
3-115
-------
Dawson, G.W., C.J. English and S.E. Petty. 1980. Physical
Chemical Properties of Hazardous Waste Constituents, U.S.
Environmental Protection Agency, Environmental Research
Laboratory, Athens, GA.
Dillion, R.T., R.B. Lantz and S.B. Pahwa. 1978. Risk
Methodology for Geologic Disposal of Radioactive Waste:
The Sandia Waste Isolation Flow and Transport (SWIFT)
Model, SAND78-1267.
Donigian, Jr., A.S., J.C. Imhoff, B.R. Bicknell and J.L.
Kittle, Jr. 1984. Guide to the Application of the
Hydrological Simulation Program - FORTRAN (HSPF). EPA
600/3-84-065, U.S. Environmental Protection Agency, Athens,
GA.
Donigian, Jr., A.S., T.Y.R. Lo and E.W. Shanahan. 1983.
Rapid Assessment of Potential Ground-Water Contamination
under Emergency Response Conditions, EPA 600/8-83-030, U.S.
Environmental Protection Agency, Washington, D.C.
Donigian, Jr., A.S. 1981. "Water Quality Modeling in Rela-
tion to Watershed Hydrology," In: Modeling Components of
Hydrologic Cycle, ed. V.P. Singh, Water Resource
Publications, Littleton, CO.
Donigian, Jr., A.S., J.L. Baker, D.A. Haith and M.F. Walter.
1983. HSPF Parameter Adjustments to Evaluate the Effects
of Agricultural Best Management Practices, EPA
PB83-247-171, U.S. Environmental Protection Agency,
Environmental Research Laboratory, Athens, GA.
Duguid, J.O. and M. Reeves. 1976. Material Transport in
Porous Media: A Finite-Element Galerkin Model, ORNL-4928,
Oak Ridge National Library, Oak Ridge, TN.
Edwards, A.L. 1972. TRUMP: A Computer Program for Transient
and Steady-State Temperature Distributions in
Multidimensional Systems, Report UCRL-14754, NTIS,
Springfield, VA.
Environmental Protection Agency. 1975. Control of Water
Pollution from Cropland, Vol. I, A Manual for Guideline
Development, EPA-600/2-75-026a, U.S. Environmental
Protection Agency, Environmental Research Laboratory,
Athens, GA.
Finley, N.C. and M. Reeves. 1968. SWIFT - Self Teaching Cur-
riculum, NUREG/CR -1968, SAND81-0410.
Freeze, R. and J. Cherry. 1979. Groundwater, Prentice-Hall
Press, N.Y.
j-116
-------
Gelhar, L.W. 1977. A Comparison of Ground Water Quality
Modeling Techniques, Paper presented at Water Resource
Center Conference, September 6-8, 1976, Berkshire, U.I.C.
Gorelick, S.M., C.I. Voss, P.E. Gill, W. Murray, M.A. Saunder
and M.H. Wright. 1984. Aquifer Reclamation Design: The
Use of Contaminant Transport Simulation Combined with
Nonlinear Programming", Water Resources Res., Vol. 20, No.
4.
Gorelick S.M. 1983. "A Review of Distributed Parameter
Groundwater Management Modeling Mehtods," Water Resources
Research, Vol. 19, No. 2.
Green, W.J., G.F. Lee, R.A. Jones and T. Palit. 1983. "Inter-
action of Clay Soils with Water and Organic Solvents:
Implications for the Disposal of Hazardous Wastes,"
Environmental Science and Technology, Vol. 17, No. 5.
Grove, D.B. and K.L. Kipp. 1980. Modeling Contaminant
Transport in Porous Media in Relation to Nuclear Waste
Disposal, A Review, In: Modeling and Low-Level Waste
Management, An Interagency Workshop, ORD-821, Oak Ridge
National Laboratory, Oak Ridge, TN.
Gupta, S. K., C. R. Cole and F. W. Bond. 1979. Finite-Ele-
ment Three-Dimensional Ground-Water (FE3DGW) Flow Model
Formulation, Program Listings and User's Manual,
PNL-2939, Battelle, Pacific Northwest Laboratory, Richland,
WA.
Gupta, S.K., C.T. Kincaid, P.R. Meyer, C.A. Newbill and
C.R. Cole. 1982. A Multi-Dimensional Finite Element Code
for the Analysis of Coupled Fluid, Energy and Solute
Transport (CFEST), PNL 4260, Battelle, Pacific Northwest
Laboratory, Richland, WA.
Hansch, C. and A. Leo. 1979. Substituent Constants for Cor-
relation Analysis in Chemistry and Biology, Wiley, New
York, NY.
Hillel, D. 1982. Introduction to Soil Physics, Academic Press,
New York, N.Y.
Holton, H.N., G.J. Stiltner, W.H. Menson and N.C. Lopez. 1975.
USDAHL-74, Revised Model of Watershed Hydrology, Tech.
Bull. 1518, USDA, Washington, D.C.
3-117
-------
Javandel, I., C. Doughty and C.F. Tsang. 1984. Groundwater
Transport: Handbook of Mathematical Models, American
Geophysical Union, Water Resources Monograph 10,
Washington, D.C.
Johanson, R., H. Davis, J. Imhoff, J.L. Kittle, and A.S.
Donigian, Jr. 1981. User's Manual for the Hydrological
Simulation Program - FORTRAN (HSPF), Release 7.0, U.S.
Environmental Protection Agency, Environmental Research
Laboratory, Athens, GA.
JRB Associates. 1982. Handbook for Remedial Action at Waste
Disposal Sites, U.S. Environmental Protection Agency,
Cincinnati, OH. EPA 625/6-82-006
JRB Associates, 1984. Slurry Trench Construction for Pollu-
tion Migration Control, EPA 540/2-84-001, U.S.
Environmental Protection Agency, Office of Emergency and
Remedial Response, Cincinnati, OH.
Kincaid, C.T., J.R. Morrey and J.E. Rogers. 1984. Geohydro-
chemical Models for Solute Migration, Volume 1: Process
Description and Computer Code Selection, EA-3417, Electric
Power Research Institute, Palo Alto, CA.
Knisel, W.G. 1980. CREAMS: A Field Scale Model for Chemical
Runoff, and Erosion from Agricultural Management Systems,
United States Department of Agriculture, Conservation
Research Report Number 26.
Knox, R.C. and L.W. Canter. 1980. Summary of Ground Water
Modeling in the United States, Report No. NCGWR 80-20,
National Center for Ground Water Research, Norman, OK.
Kobayashi, H. and B.E. Rittman. 1982. "Microbial Removal of
Hazardous Organic Compounds," Environmental Science and
Technology, Vol. 16, No. 3.
Koines A. 1982. Technical Review of Ground-Water Models,
Contract No. 68-01-6404, Draft Report, U.S. Environmental
Protection Agency, Office of Solid Waste, Washington, D.C.
Lappala, E.G. 1980. Modeling of Water and Solute Transport
Under Variable Saturated Conditions: State-of-the-Art, In:
Modeling and Low-Level Waste Management, An Interagency
Workshop, ORD-821, Oak Ridge National Laboratory, Oak
Ridge, TN.
Leo, A., C. Hansch and D. Elkins. 1971. "Partition Coeffi-
ceints and Their Use," Chem. Rev., Vol. 71.
3-118
-------
Linsley, R.K., M.A. Kohler and J.H.L. Paulhus. 1982. Hydrology
for Engineers, McGraw-Hill, New York, N.Y.
Lyman, W.J., W.F. Reehl and D.H. Rosenblatt. 1982. Handbook
of Chemical Property Estimation Methods. McGraw-Hill, New
York, N.Y.
Mabey, W.R., J.H. Smith, R.T. Podoll. H.L. Johnson, T. Mills,
T.W. Chou, J. Gates, I. Waight Partridge and D. Vandenburg.
1982. Aquatic Fate Process Data for Organic Priority
Pollutants, U.S. Environmental Protection Agency, Office of
Water Regulation and Standards, Washington, D.C.
McCarty, P.L., M. Reinhard and B.E. Rittman. 1981. "Trace
Organics in Ground Water," Environmental Science and
Technology, Vol. 15, No. 1.
Mercer, J.W. and C.R. Faust. 1981. Ground-Water Modeling,
National Water Well Association, Worthington, OH.
Mercer, J.W., S.D. Thomas and B. Ross. 1982. Parameters
and Variables Appearing in Repository Siting Models,
NUREG/CR-3066, U.S. Nuclear Regulatory Commission,
Washington, D.C.
Mills, W., J. Dean, D. Porcella, S. Gherini, R. Hudson, W.
Frick, G. Rupp and G. Bowie. 1982. Water Quality
Assessment: A Screening Procedure for Toxic and
Conventional Pollutants, Vol. 1 and 2, EPA 600/6082-004ab,
U.S. Environmental Protection Agency, Athens GA.
Moiser, J.E., J.R. Fowler, C.J. Barton, W.W. Tolbert, S.C.
Meyers, J.E. Vancil, H.A. Price, M.J.R. Vasco, E.R. Rutz,
T.X. Wendel and L.D. Rickertson. 1980. Low-Level Waste
Management: A Complilation of Models and Monitoring
Techniques, ORNL/SMB-79/13617/2, SAI/OR-565-2, Oak Ridge
National Laboratory, Oak Ridge, TN.
Morris, D.A. and A.I. Johnson. 1967. Summary of Hydrologic
and Physical Properties of Rock and Soil Materials as
Analyzed by the Hydrologic Laboratory of the U.S.
Geologiocal Survey 1948-60, USGS Water Supply Paper 1839-D.
Narasimhan, T.N. and P.A. Witherspoon. 1976. An Inte-
grated Finite Difference Method for Analyzing Fluid Flow in
Porous Media, Water Resources Research Vol. 12, No. 1.
Nelson, R.W., G.W. Gee and E.M. Arnold, Eds. 1982. "Pro-
ceedings of the Symposium of Unsaturated Flow and Transport
Modeling," NUREG/CP-0030, PNL-SA-10325, Battelle, Pacific
Northwest Laboratory, Richland, WA.
3-119
-------
Nelson, R.W., P.R. Meyer, P.L. Oberlander, S.C. Sneider, D.W.
Mayer and A.E. Reisenauer. 1983. Model Evaluation of
Seepage from Uranium Tailings Above and Below the Water
Table, NUREG/CR-3078, PNL-4461, Battelle, Pacific Northwest
Laboratory, Richland, WA.
Onishi, Y., R.J. Serne, E.M. Arnold, C.E. Cowan and F.L.
Thompson. 1981. Critical Review: Radionuclide Transport,
Sediment Transport, and Water Quality Modeling; and
Radionuclide Adsorption/Desorption Mechanisms, U.S. Nuclear
Regulatory Commission, NUREG/CR-1322, Washington, D.C.
Orlob, G. 1971. "Mathematical Modeling of Estuarial Systems,"
International Symposium of Mathematical Modeling
Techniques.
Oster, C.A. 1982. Review of Ground-water Flow and Trans-
port Models in the Unsaturated Zone, NUREG/CR-2917,
PNL-4427,Battelle, Pacific Northwest Laboratory, Richland,
WA.
Prickett, T.A. and C.G. Lonnquist. 1971. Selected Digital
Computer Techniques for Groundwater Resource Evaluation,
Bulletin 55, Illinois State Water Survey, Champaign, IL.
Prickett, T.A., T.G. Naymik and C.G. Lonnquist. 1981. A
"Random-Walk" Solute Transport Model for Selected Ground
Water Quality Evaluations, Bulletin 65, Illinois State
Water Survey, Champaign, IL.
Rai, D., J.M. Zachara, R.A. Schmidt and A.P. Schwab. 1984a.
Chemical Attentuation Rates, Coefficients and Constants in
Leachate Migration, Volume 1: A Critical Review, EA-3356,
Volume 1, Electric Power Research Institute, Palo Alto, CA.
Rai, D., J.M. Zachara, R.A. Schmidt and A.P. Schwab. 1984b.
Chemical Attenuation Rates, Coefficients and Constants in
Leachate Migration, Volume 2: An Annotated Bibliography,
EA-3356, Volume 2, Electric Power Research Institute, Palo
Alto, CA.
Rawls, W.J. and D.L. Brakensiek. 1982. "Estimating Soil Water
Retention from Soil Properties," Proc. ASCE, Vol 108, No.
IR2.
Ree, W.U., F.L. Wimerley and R.W. Crow. 1977. "Manning's n
and the Overland Flow Equation," ASAE Trans., Vol. 22,
No. 1
3-120
-------
Reeves, M. and R.M. Cranwell. 1981. User's Manual for the
Sandia Waste Isolation Flow Transport Model (SWIFT),
Release 4.81, NUREG/CR-2324, SAND81-2516, Washington, D.C.
Reeves, M. and J.O. Duguid. 1976. Water Movement Through
Saturated Unsaturated Porous Media: A Finite-Element
Galerkin Model, ORNL-4297, Oak Ridge National Laboratory
Report, Oak Ridge, TN.
Reisenauer, A.E., S.K. Gupta, R.W. Nelson and C.A. Newbill.
1981. Advective Radionuclide Transport with Soil
Interaction Under Variably Saturated Flow Conditions,
PNL-3994, Battelle, Pacific Northwest Laboratory, Richland,
WA.
Reisenauer, A.E., K.T. Key, T.N. Narasimhan and R.W. Nelson.
1982. TRUST: A Computer Program for Variably Saturated
Flow in Multidimensional Media, NUREG/CR-2360, PNL-3975,
Battelle, Pacific Northwest Laboratory, Richland, WA.
Roberts, P.V. 1982. "Performance of GAC for FOG Removal,"
Journal of American Water Works Association, Feb.
Roberts, D.W. 1984. Soil Properties, Classification, and
Hydraulic Conductivity Testing, SW-925. Draft Technical
Resource Document for Public Comments, U.S. Environmental
Protection Agency, Municipal Environmental Research
Laboratory, Cincinnati, OH.
Ross, B.B., D.N. Contractor, and V.O. Shanholtz. 1977. "Finite
Element Simulation of Overland and Channel Flow," Trans.
ASAE.
SAI. 1981. Tabulation of Waste Isolation Computer Models,
ONWI-78, Prepared by Science Applications, Inc. for the
Office of Nuclear Waste, Battelle, Memorial Institute,
Columbus, OH.
Sax, N.I. 1979. Dangerous Properties of Industrial Materials
4th ed., Van Nostrand Reinhold, New York, N.Y.
Schroeder, P.R., J.M. Morgan, T.M. Walbki, and A.C. Gibson.
1984a. The Hydrologic Evaluation of Landfill Performance
(HELP) Model, Volume I., User's Guide for Version 1,
EPA/530-SW-84-009, U.S. Environmental Protection Agency,
Office of Solid Waste and Emergency Response, Washington,
D.C.
3-121
-------
Schroeder, P.R., A.C. Gibson, and M.P. Smolen. 1984b. The
Hydrolgic Ealuation of Landfill Performance (HELP) Model,
Volume II., Documentaiton for Version I, EPA/530-SW-84-010,
U.S. Environmental Protection Agency, Office of Solid Waste
and Emergency Response, Washington, D.C.
SCS Engineers. 1982. Costs of Remedial Response Actions at
Uncontrolled Hazardous Waste Sites, EPA 600/2-82-035, U.S.
Environmental Protection Agency, Municipal Environmental
Research Laboratory, Cincinnati, OH.,
Segol, G. 1976. A 3-D Galerkin Finite-Element Model for the
Analysis of Contaminant Transport in Variably Saturated
Porous Media, Proceeding of the 1st International
Conference on Finite Elements in Water Resources, Princeton
University, Pentech Press, London, England.
Shafer, J.M., P.L. Oberlander, and R.L. Skaggs. 1984. Miti-
gative Techniques and Analysis of Generic Site Conditions
for Ground-Water Contamination Associated with Sewer
Accidents, NUREG/CR-3681 PNL-5072. Prepared for U.S.
Nuclear Regulatory Commission, Washington, D.C.
Shafer, J.M. 1984. "Determining Optimum Pumping Rates for
Creation of Hydraulic Barriers to Ground Water Pollutant
Migration," Fourth National Symposium and Exposition on
Aquifer Restoration and Ground Water Monitoring, National
Water Well Association, Washington, D.C.
Silka, L.R. and J.W. Mercer. 1982. "Evaluation of Remedial
Actions for Groundwater Contamination at Love Canal, New
York," Proceedings of National Conference on Management of
Hazardous Materials Control Research Institute, Silver
Spring, MD.
Thomas, S.D., B. Ross and J.W. Mercer. 1982. A Summary of
Repository Siting Models, NUREG/CR-2782, U.S. Nuclear
Regulatory Commission, Washington, D.C.
van Genuchten, R. 1978a. Simulation Models and Their Appli-
cation to Landfill Disposal Siting: A Review of Current
Technology, In: Land Disposal of Hazardous Waste,
Proceedings of the Fourth Annual Research Symposium.
EPA-600/9-78-016, San Antonio, TX.
van Genuchten, R. 1978b. Calculating the Unsaturated
Hydraulic Conductivity with a new Closed-Form Analytical
Model. Research Report 78-WR-08 Water Resources Program,
Department of Civil Engineering, Princeton University, N.J.
Verschueren, K. 1977. Handbook of Environmental Data on
Organic Chemicals, Van Nostrand Reinhold Co., N.Y.
3-122
-------
Wilson, J.T., J.F. McNabb, D.L. Balkwill and W.C. Ghiorse.
1983. "Enumeration and Characterization of Bacteria
Indigenous to a Shallow Water-Table Aquifer," Ground
Water, Vol. 21, No. 2.
Wischmeier, W.H. and D.D. Smith. 1978. Predicting Rainfall
Erosion Losses—A Guide to Conservation Planning,
Agriculture Handbook No. 537, USDA, Agricultural Research
Service.
Woolhiser, D.A. 1976. Hydrologic Aspects of Nonpoint Source
Pollution, In: B.A. Stewart et al., Control of Water
Pollution from Cropland, Vol. II, EPA-600/2-75-026b, U.S.
Environmental Protection Agency, Washington, D.C.
Yeh, G.T. and D.S. Ward. 1979. FEMWATER: A Finite-Element
Model of Water Flow Through Saturated-Unsaturated Porous
Media, ORNL-5567, Oak Ridge National Laboratory, Oak Ridge,
TN.
Yeh, G.T. and D.S. Ward. 1981. FEMWASTE: A Finite-Element
Model of Waste Transport Through Saturated-Unsaturated
Porous Media, ORNL-5601, Oak Ridge National Laboratory, Oak
Ridge, TN.
Yeh, G.T. 1982a. Training Course No. 1: The Implementation
of FEMWATER Computer Program—Final Report, NUREG/CR-2705,
ORNL-5567, Oak Ridge National Laboratory, Oak Ridge, TN.
Yeh, G.T. 1982b. Training Course No. 2: The Implementation
of FEMWATER Computer Program—Final Report, NUREG/CR-2706,
ORNL-5601, Oak Ridge National Laboratory, Oak Ridge, TN.
3-123
-------
APPENDIX A
SUPPORTING INFORMATION ON HSPF, FEMWATER/
FEMWASTE AND FE3DGW/CFEST
3-124
-------
APPENDIX A
SUPPORTING INFORMATION ON HSPF, FEMWATER/
FEMWASTE AND FE3DGW/CFEST
At the request of EPA, surface, unsaturated and saturated zone
models applicable to the assessment of a broad range of
remedial actions were selected for implementation on the
EPA-NCC computer system in Research Triangle Park, N.C. The
intent was to provide a general capability for remedial action
modeling. This appendix discusses the selection of the three
codes; linkage considerations; their implementation on the NCC
system; sources of code documentation and user assistance; and
specific parameters requiring adjustment for each remedial
measure discussed in Section 6.
A.I CODE SELECTION
The process of reviewing and selecting models for remedial
action assessment is described in detail in Volume 1. Section
5 of this volume discusses the model development and
application process and evaluates a number of candidate codes.
Starting with the candidate codes listed in Section 5.2, the
following criteria were used to select one code for each zone.
1. Dimensionality - multiple land segment surface zone
model, two - dimensional (x-z) unsaturated zone
model, and three - dimensional saturated zone model.
2. Time Frame - continuous simulation with variable time
step.
3. Flow Processes - advection, infiltration to the
unsaturated zone, drainage to the saturated zone, and
evapotranspiration.
4. Transport Processes - advection, dispersion, sorption,
retardation, and degradation.
5. Data Structure - flexible input and output sequences,
data management and storage capabilities.
3-125
-------
6. Ease of Use - clear and complete documentation,
program maintenance and user assistance available.
7. Code Testing - model tests under a variety of
conditions, validation against field data.
Table A.I summarizes these criteria for each zone.
The following codes were selected based on the above criteria:
HSPF for the surface zone, FEMWATER/FEMWASTE for the
unsaturated zone and FE3DGW/CFEST for the saturated zone. The
principle reasons for selecting the codes are:
1. HSPF provides simulation of multiple land segments,
which allows the representation of both the site and
the drainage area surrounding the site. This is
particularly important when modeling surface water
diversion and collection type actions, which cause
local changes in runoff.
2. The datg base management software associated with HSPF
is both flexible and powerful, and can provide a
structure for easy linkage with the codes selected
for the other zones.
3. FEMWATER/FEMWASTE provides all of the needed processes
and also has been tested and applied to case studies.
Documentation a/id ease of use are generally good.
4. FE3DGW/CFEST supports two-dimensional or three-dimen-
sional simulation. Code tests have been conducted.
Documentation of FE3DGW is reasonably complete;
similar documentation is under prepration for CFEST.
Both models were used for remedial action evaluation
(Cole et al., 1983).
5. User support and code implementability can be major
factors when selecting codes. HSPF is actively
supported by the Water Quality Modeling Center at
EPA-Athens (Contact: Mr. Thomas Barnwell) and
through a maintenance contract that provides for
training workshops and code updates. Individual
support contracts can also be arranged. HSPF has
been implemented on numerous computer systems,
including the IBM at the National Computer Center.
User support for FEMWATER/FEMWASTE is limited to that
offered by ORNL. Both codes have been implemented on
a number of computer systems. User support for
FE3DGW and CFEST is offered by Battelle, Pacific
Northwest Laboratory. The Department of Energy's,
Office of Nucltar Waste Isolation has recently
3-126
-------
TABLE A.I CHARACTERISTICS OF CODES (BY ZONE) THAT WOULD SATISFY EPA'S
MEED FOR A COMPREHENSIVE, MODEL-BASED REMEDIAL ACTION
ASSESSMENT C7iPABILITY
U>
I
to
DIMENSIONALITY AND SPATIAL
RESOLUTION
MIGRATION AND FATE PROCESSES
Legend:
• - Process Should
Be Considered
X • Longitudinal
Y = Lateral
Z = Vertical
S - Single Segment R = Retardation
M • Multiple Segment D « Degradation
C » Hydraulic SS • Steady State
Conductivity
P « Porosity
T •= Time Varying
-------
expanded the level of user support. CFEST has been
implemented on several computers, including EPA's IBM
system at the National Computer Center.
These models, as a group, provide most of the capabilities
required for analysis of complex site conditions, including
the following:
1. Representation of the surface hydrologic system,
including precipitation, snow melt,
evapotranspiration, runoff, and infiltration (HSPF).
2. Representation of sediment and sediment-related
contaminant transport, including soil detachment,
scour and deposition (HSPF).
3. Representation of percolation through the unsat-
urated zone, including soil wetting front movement,
seepage from ponds, and lateral drainage
(FEMWATER/FEMWASTE).
4. Representation of flow through heterogeneous aquifers
and multi-aquifer systems with variable water table
elevations (FE3DGW/CFEST).
5. Representation of all key chemical transport processes
(advection, dispersion, retardation, and degradation)
in the three zones.
6. Representation of complex boundary conditions caused
by ponds, streams, aquicludes and basement rock, as
well as by different remedial actions.
7. Representation of changes in most of the key processes
affected by remedial actions.
A.2 LINKAGE OF HSPF, FEMWATER/FEMWASTE AND CFEST
In cases where the use of two or more of these models is
required to evaluate remedial action performance, linkage of
the models may be required (see Section 5.3 of this volume).
"Soft linkage" of the three codes is likely to be the most
viable approach. No such linkage currently exists. The
following discussion, however, presents a possible approach
for linking the models.
Figure A. 1 provides a schematic diagram for the proposed
linkage. The linkage of HSPF and FEMWATER/FEMWASTE would take
advantage of the data management utility routines already
included in HSPF. TSMS, the Time Series Management System, is
3-128
-------
HSPF
i TSMS |
i j
NET INFILTRATION 4 TIME SERIES
BRIDGE PROGRAM
FEMWATER
DRAINAGE
TIME
SERIES
DARCY'S
VELOCITIES^ f
FEMWASTE
CONTAMINANT
CONCENTRATION
TIME SERIES
BRIDGE PROGRAM
CFEST
Figure A.I Schematic diagram showing soft linkage of HSPF,
FEMWATER/FEMWASTE and CFEST with bridge programs,
3-129
-------
a series of routines that provide time series storage,
transfer, conversion and mass balance checking capabilities.
TSMS could be used aggregate or disaggregate and combine or
separate net infiltration time series (from HSPF) for
individual land segments to • obtain net infiltration time
series for input to individual nodes in the FEMWATER/FEMWASTE
grid. Once these time series have been prepared with the TSMS
routines, they could be transferred to a bridge program that
simply reformats the time series for direct input to FEMWATER
and FEMWASTE. Thus, the bridge program between these two
codes would be relatively simple.
The bridge program between FEMWATER/FEMWASTE and FE3DGW/CFEST
would be more complicated. It should provide for the
aggregation and disaggregation of drainage and associated
contaminant concentration time series, as well as the
combination and separation of these time series. The former
would make it possible to account for differences in time
stepping, the latter would make it possible to account for
differences in computational grids. The program also should
provide for unit conversions, mass balance checks, and the
reformatting of time series for input to CFEST.
A.3 MODEL CODE IMPLEMENTATION
The codes as currently implemented are available on the NCC
IBM computer system; a valid account is required for access to
the codes. Existing program load modules are executed via
input files containing the program input and required Job
Control Language (JCL). JCL is used to execute load modules,
create scratch files and output files, and pass control to
subsequent modules.
The following procedure should be used for implementing the
codes:
1. Contact the appropriate EPA official and request that
all necessary files (i.e., program load modules,
sample input/JCL files, and other necessary files) be
copied to your account. Program documentation should
also be obtained.
2. Using the sample input files and code documentation,
develop input for your application.
3. Modify the JCL to reflect your account; and modify all
file names.
4. Run the code.
3-130
-------
Interested parties should contact Mr. Douglas Ammon for
current information or program status and implementation of
FE3DGW, CFEST, FEMWATER and FEMWASTE. Mr. Thomas 0. Barnwell,
Jr. should be contacted for information on HSPF.
A.4 SOURCES OF CODE DOCUMENTATION AND USER ASSISTANCE
A major reason for the selection of the specified codes was
the availability of user guidance. Key sources of information
and help are summarized below.
HSPF was created and is supported by the EPA and is currently
available on magnetic tape, as a source code for mainframe
computers from:
The Water Quality Modeling Center
U.S. EPA, Environmental Research Lab
Athens, GA 30613
Contact: Mr. Thomas O. Barnwell
The model is currently maintained, under a contract with EPA,
by:
Anderson-Nichols & Co., Inc.
2666 East Bayshore Road
Palo Alto, CA 94303
Contact: Mr. Jack Kittle or Mr. Brian Bicknell
A current release of the user's manual for HSPF can be
obtained from the Water Quality Modeling Center. Key
references for HSPF design, structure and application include
Johanson et al., (1981) and Donigian et al., (1984).
FEMWATER/FEMWASTE were developed at Oak Ridge National
Laboratory and are currently maintained by ORNL staff. The
source code and documentation are available from:
Oak Ridge National Laboratory
Environmental Science Division
P.O. Box X
Oak Ridge, TN 37830
Contact: Dr. George T. Yeh
Key references on the design, structure, implementation and
use of these codes include: Reeves et al., 1975; Duguid et
al., 1976; Yeh and Ward, 1980 and 1981; and Yeh 1982a and
1982b. The last two sources are self-contained training
3-131
-------
courses.
FE3DGW/CFEST were developed by Battelle, Pacific Northwest
Laboratory (PNL) and are currently maintained by PNL staff.
The source code and documentation can be obtained from:
Battelle, Pacific Northwest Laboratory
P.O. Box 999
Richland, WA 99352
Contact: Mr. Charles R. Cole
Key references on the design, implementation and use of these
codes include: Gupta et al., (1984), Gupta et al., (1982) and
Cole et al., (1984).
A.5 PARAMETER ADJUSTMENTS REQUIRED FOR EACH REMEDIAL MEASURE
This section presents the specific parameters and input
boundary conditions that must be adjusted in HSPF,
FEMWATER/FEMWASTE and CFEST to represent each remedial measure
discussed in Section 6. The recommended adjustments are
presented in a series of tables for each measure and for each
code needed for that measure. The following is a list of the
tables.
Table A.2
Table A.3
Table A.4
Table A.5
Table A.6
Table A.7
Table A.8
Table A.9
Capping, Grading and Revegetation Parameter
Adjustments for HSPF
Capping, Grading and Revegetation Parameter
Adjustments for FEMWATER/FEMWASTE
Surface Water Diversion and Collection
Parameter Adjustments for HSPF
Ground-Water Pumping and Interceptor Trench
Parameter Adjustments for CFEST
Impermeable Barrier Parameter Adjustments
for CFEST
Subsurface Drains and Solution Mining Parameter
Adjustments for FEMWATER/FEMWASTE
Subsurface Drains and Solution Mining Parameter
Adjustments for CFEST
Excavation Parameter Adjustments for FEMWATER/
FEMWASTE
3-132
-------
Table A.10 Excavation Parameter Adjustments for CFEST
Table A.11 Hydraulic Dredging and Seepage Basin Parameter
Adjustments for FEMWATER/FEMWASTE
Table A.12 Bioreclamation and Chemical Injection Parameter
Adjustments for CFEST
Table A.13 Permeable Treatment Bed Parameter Adjustments
for CFEST
3-133
-------
TABLE A.2 CAPPING, GRADING AND REVEGETATION PARAMETER
ADJUSTMENTS FOR HSPF
Parameter
NSUR
LSUR
SLSUR
KRER
KGER
INFILT
Purpose
Surface roughness
(Manning's n)
Slope length
Slope
Coefficient for
soil detachment
Coefficient of
soil scour
Infiltration
capacity
Rang e/Units
0.25-0.4
(S) L
5-18%
0.08-0.28
COVER Canopy development
LZETP Lower zone
evapotranspiration
parameter
CEPSC Maximum
interception
AFFIX Soil compaction
factor
0.0-1.0
0.05-1.0
in/hr
0-1.0
0.2-0.9
0.06-0.25
in.
0.1-.001
Reference
Donigian
et al., 1983
JRB Associates,
1982
Donigian
et al., 1983
Johanson
et al., 1981
Donigian
et al., 1983
Donigian and
Davis, 1978
Donigian
et al., 1983
Donigian
et al., 1983
Donigian
et al., 1983
Donigian
et al., 1983
(S) Site-specific
L Length
3-134
-------
TABLE A.3 CAPPING, GRADING AND REVEGETATION PARAMETER
ADJUSTMENTS FOR FEMWATER/FEMWASTE
FEMWATER
Parameter
NMAT
PROP(3,1)
HPROP(J,K)
THPROP(J,K)
AKPROP(J,K)
PROP(4,I)
PROP(5,1)
Purpose
Range/Units Reference
Number of materials (S)
(cap, drainage and
filter layers)
Porosity of each
material 'I'
Pressure head of
Jth point for
material 'K1
Moisture-content
of Jth point for
material 'K1
Relative conduc-
tivity of Jth point
for material 'K1
(S) %
(S) L
(S) L3/L3
(S)
xx-component of (S) L/T
saturated hydraulic
conductivity for
material 'I'
FEMWASTE
Parameter
PROP(3,1)
PROP(4,1)
PROP(6,1)
PROP(9,1)
zz-component
of saturated
hydraulic conduc-
tivity for
material 'K1
Purpose
Longitudinal
dispersivity for
material 'I'
Lateral
dispersivity for
material 'I'
Porosity of
material 'I'
Tortuosity of
material 'I'
Kz/Kx=0.1
(initial
estimate)
Range/Units
(S) L
(S) L
(S) %
0.0-5.0
Section 6.3.2.2.2
Section 6.3.2.1.1
Section 6.3.2.1.1
Section 6.3.2.1.1
Section 6.3.2.1.2
Freeze and
Cherry, 1979
Reference
Section 6.3.2.2.1
Section 6.3.2.2.1
Section 6.3.2.2.2
Yeh, 1982
3-135
-------
TABLE A.4 SURFACE WATER DIVERSION AND COLLECTION PARAMETER
ADJUSTMENTS FOR HSPF
Pervious Land Segments
Parameter Purpose
Range/Units Reference
NSUR
LSUR
SLSUR
Surface roughness 0.15-0.4
(Manning's n)
Slope length
Slope
(S) L
12-18%
Donigian
et al., 1983
JRB Associates,
1982
Channel Segments
Parameter Purpose
NSUR
SLSUR
Channel roughness
(Manning's n)
Channel Slope
Range/Units Reference
- Section 6.3.1.1
6-12%
JRB Associates,
1982
IS) Site-specific
L Length
3-136
-------
TABLE A.5 GROUND WATER PUMPING AND IMPERMEABLE BARRIER
PARAMETER ADJUSTMENTS FOR CFEST
Parameter
NODE
BIV
NODALQ
NQNDOE
BIVF
BIVFC
Purpose
Held head node
number
Value of held head
Time-constant
nodal flux
Node number
having nodal flux
Integrated flow
volume
Concentration of
injection fluid
Range/Units
(S)
(S) L
2
(S)
(S) L3/T
(S) M/T
Reference
(S) Site-specific
L Length
M Mass
3-137
-------
TABLE A.6 IMPERMEABLE BARRIER PARAMETER ADJUSTMENTS FOR CFEST
Parameter
MAT
XK
YK
ZK
THETAO
HTHETA
Purpose
Number of
materials (barrier,
surrounding media)
Hydraulic conduc-
tivity (for 'x'
direction)
Hydraulic conduc-
tivity (for 'y1
direction)
Hydraulic conduc-
tivity (for 'z1
direction)
Porosity at refer-
ence pressure head
Pressure head at
which THETAO is
defined
Range/Units
(S)
Reference
ALPHAL(I) Longitudinal
dispersivity for
each material 'I
ALPHAT(I) Lateral disper-
sivity for each
material 'I'
(S) L/T Section 6.3.2.1.2
(S) L/T Section 6.3.2.1.2
(S) L/T Section 6.3.2.1.2
(S) %
(S) L
(S) L
(S) L
Section 6.3.2.2.2
Section 6.3.2.2.1
Section 6.3.2.2.1
S) Site-specific
L Length
3-138
-------
TABLE A.7 SUBSURFACE DRAINS AND SOLUTION MINING PARAMETER
ADJUSTMENTS FOR FEMWATER/FEMWASTE
FEMWATER
Parameter
NMAT
HPROP(J,K)
THPROP(J,K)
AKPROP(J,K)
PROP(3,1)
PROP(4,I)
PROP(5,1)
THDBF(J,I)
HDBF(J,I)
NPDB(I)
Purpose
Number of materials
(waste, gravel and
surrounding area)
Pressure head of (S) L
Jth data point
for material 'K'
Moisture-content
of Jth data point
for material 'K1
Relative conduc- (S)
tivity of Jth data
point for
material 'K1
Porosity of (S) %
medium 'I'
xx-component of (S) L/T
saturated hydrualic
conductivity for
material 'I'
zz-component of (S) L/T
saturated hydraulic
conductivity for
material 'I'
Time of Jth data (S) T
point on Ith held
head profile
Total head of Jth (S) L
data point in Ith
profile
Global node number (S)
of Ith node
Range/Units Reference
(S)
Section 6.3.2.1.1
(S) L3/L3 Section 6.3.2.1.1
Section 6.3.2.1.1
Section 6.3.2.1.2
Section 6.3.2.1.2
Section 6.3.2.1.2
(continued)
3-139
-------
TABLE A.7 (continued)
FEMWASTE
Parameter Purpose Range/Units Reference
NMAT Number of materials (S)
(waste, gravel and
surrounding area)
PROP(1,1) Distribution (W) L3/M Section 6.3.2.2.4
coefficient for
materials 'I'
TSOSF(J,I) Time of Jth (S) T
data point on Ith
Cauchy flux profile
SOSF(J,I) Source/sink value (S) L 3/T/L
of Jth data point
in Ith profile
TCRSF(J,I) Time of Jth data (S) T
point on Ith
incoming-concentra-
tion vs. time-
profile
CRSF (J,I) Concentration (S) M/L
(S) Site-specific
(W) Waste-specific
L Length
T Time
M Mass
3-140
-------
TABLE A.8 SUBSURFACE DRAINS AND SOLUTION MINING PARAMETER
ADJUSTMENTS FOR CFEST
Parameter
MAT
XK.YK.ZK
THETAO
HTHETA
NODE
BIV
NODALQ
NQNDOE
BIVF
BIVCF
Purpose
Number of
materials (waste,
surrounding media)
Hydraulic con-
ductivity for
'x1, 'y1 and 'z1
directions
Porosity at
reference pressure
head
Pressure head at
which THETAO is
defined
Held head node
number
Value of held
head
Time-constant
nodal flux
Node number
having nodal flux
Integrated flow
volume
Concentration of
injected fluid
Range/Units
(S)
(S) L/T
(S) %
(S) L
(S)
(S) L
2
(S)
(S) L3/T
(S) M/L3
Reference
Section 6.3.1.2.2
Section 6.3.2.2.2
(S) Site-specific
L Length
T Time
M Mass
3-141
-------
TABLE A. 9 EXCAVATION PARAMETER ADJUSTMENTS FOR FEMWATER/
FEMWASTE
FEMWATER
Parameter
NMAT
PROP (3,1)
HPROP(J,K)
THPROP(J,K)
AKPROP(J,K)
Purpose
Number of
materials (waste
and surrounding
media)
Porosity of
material 'I1
Pressure head of
Jth data point
for material 'K1
Range/Units
(S)
(S) %
(S) L
(S) L/L
PROP(4,I)
PROP (5,1)
Moisture-content
of Jth data point
material 'K1
Relative conduc- (S)
tivity of Jth data
point for
material 'K1
xx-component of (S) L/T
saturated hydrau-
lic conductivity
for material 'I'
zz-component of (S) L/T
saturated hydraulic
conductivity for
material 'I'
Reference
Section 6.3.2.2.2
Section 6.3.2.1.1
Section 6.3.2.1.1
Section 6.3.2.1.1
Section 6.3.2.1.2
Section 6.3.2.1.2
(continued)
3-142
-------
TABLE A.9 (continued)
FEMWASTE
Parameter
NMAT
PROP(1,1)
PROP(2,I)
PROP(3,1)
PROP(4,I)
PROP(6,1)
PROP(9,1)
TCDBF(J,I)
CDBF (J,I)
Purpose
Range/Units
Reference
Number of materials (S)
materials (waste and
surrounding media)
Distribution (W) L/M
coefficient for
material 'I'
Bulk density (S) M/L'
Longitudinal (S) L
dispersivity for
material 'I'
Lateral dispersi- (S) L
vity for material 'I'
Porosity of (S) %
material 'I'
Tortuosity of 0.0-0.5
material 'I'
Time of Jth data (S) T
point in Ith held
concentration profile
Held concentration (S) M/L'
of Jth data point
in Ith profile
Section 6.3.2.2.4
Section 6.3.2.2.3
Section 6.3.2.2.1
Section 6.3.2.2.1
Section 6.3.2.2.2
Yeh, 1982
(S) Site-specific
(W) Waste-specific
L Length
T Time
M Mass
3-143
-------
TABLE A.10 EXCAVATION PARAMETER ADJUSTMENTS FOR CFEST
Parameter
MAT
XK
YK
THETAO
Purpose
Number of mater-
ials (waste and
surrounding media)
Hydraulic conduc-
tivity in 'x1
direction
Range/Units Reference
(S)
(S) L/T Section 6.3.2.1.2
Hydraulic conduc- (S) L/T Section 6.3.2.1.2
tivity in 'y1
direction
Porosity at (S) %
reference pressure
head
Section 6.3.2.2.2
HTHETA Pressure head at (S) L
which THETAO is
defined
ALPHAL(I) Longitudinal (S) L
dispersivity for
material 'I'
ALPHAT(I) Lateral dispersi- (S) L
vity for
material 'I'
NODBC Held concentration (S)
node number
BIVC Value of held (S) M/L
concentration
Section 6.3.2.2.1
Section 6.3.2.2.1
(S) Site-specific
L Length
T Time
M Mass
3-144
-------
TABLE A.11 HYDRAULIC DREDGING AND SEEPAGE BASIN PARAMETER
ADJUSTMENTS FOR FEMWATER/FEMWASTE
FEMWATER
Parameter
THDBF(J,I)
HDBF(J,I)
Purpose
Time of Jth
data point in Ith
held head profile
Total head of
Jth data point
in Ith profile
Range/Units
(S) T
(S) L
Reference
FEMWASTE
Parameter
TCDBF(J,I)
CDBF (J,I)
Purpose
Time of Jth
data point in Ith
held concen-
tration profile
Held concentration
of Jth data point
in Ith profile
Range/Units
(S) T
(S) M/L
Reference
(S) Site-specific
L Length
T Time
M Mass
3-145
-------
TABLE A.12 BIORECLAMATION AND CHEMICAL INJECTION PARAMETER
ADJUSTMENTS FOR CFEST
Parameter
NODE
BIV
NODBC
BIVC
DECAY
Purpose
Held head node
number
Value of held
head
Held concen-
tration node
number
Value of held
concentration
Degradation
rate
Range/Units
(S)
(S) L
(S)
(S) M/L3
(W) /T
Reference
Section 6.3.2.2.5
(S) Site-specific
L Length
M Mass
3-146
-------
TABLE A.13 PERMEABLE TREATMENT BED PARAMETER ADJUSTMENTS FOR
CFEST
Parameter Purpose Range/Units Reference
RETARD Retardation (W) L/M Section 6.3.2.2.4
factor
3-147
-------
VOLUME 4
Analytical and Numerical Models
for the
Evaluation of Remedial Actions
in Surface Water
-------
VOLUME 4: ANALYTICAL AND NUMERICAL MODELS FOR THE EVALUATION
OF REMEDIAL ACTIONS IN SURFACE WATER
SECTION 1
INTRODUCTION
1.1 BACKGROUND
Releases of hazardous substances into rivers, lakes, and
estuaries have been a major concern of the Environmental
Protection Agency (EPA) for many years. EPA, the U.S. Coast
Guard, and state and local agencies have responded to numerous
release episodes which derived from a variety of sources.
Agency actions involved problem identification and
quantification, assessment of hazards to health and the
environment, selection and implementation of appropiate
responses, and follow-up monitoring of contaminant levels in
the water body. Analytical and numerical predictive tools are
available and have been used to varying degrees to (1)
identify current chemical locations and concentrations, (2)
predict future movement of chemical plumes, and (3) evaluate
the responses of the receiving water body and the chemical
plume to alternative actions.
Models may be used in the selection and design of removal and
other long-term remedial actions. Emphasis is placed on the
representation of complex processes in the water bodies and
the physical and chemical effects of specific actions. Both
analytic and numerical models are considered; however,
emphasis is placed on numerical models which are capable of
representing a broad variety of complex conditions. Guidance
on model selection is intended to assist EPA and other
agencies in performing in-house studies and in working with
and evaluating the results of studies by other organizations.
This report constitutes one volume of a four volume set of
reports designed for the selection and use of models for
remedial action assessment at hazardous waste sites. Volume
1: Selection of Models for Remedial Action Assessment provides
a model selection methodology for ground-water and surface
4-1
-------
water contamination problems. This report (Volume 4) is
designed to complement Volume 1 by providing guidance on the
evaluation of available surface water models for remedial
action assessment. Volumes 2 and 3 of the set provide similar
guidance for the evaluation of simplified and numerical models
for subsurface problems, respectively.
1.2 PURPOSE OF REPORT
The primary goal of this report was the development and
documentation of a model evaluation and application procedure.
This procedure enables the user to identify models which are
most appropriate for his or her site-specific needs and apply
specific models to specific applications. Development of the
model evaluation procedure involved the following steps.
1. Identify in-stream processes which may be impacted by
discharges and/or remedial actions.
2. Identify potentially viable remedial actions and
relate such actions to specific discharge scenarios
and water body types.
3. Relate remedial actions to in-stream processes, in-
cluding a determination of whether the action will
enhance or retard each process.
4. Relate specific hazardous chemicals to in-stream pro-
cesses, including the importance of each process in
determining chemical migration and fate.
5. Identify available analytic and numeric models for
chemical transport and fate and evaluate their
potential applicability to each scenario (i.e., water
body, discharge, and chemical type).
6. Evaluate representative models using matrices relating
key model capabilities to in-stream processes, water
body characteristics, and remedial action modeling
requirements.
7. Identify representative types of models which are
suitable for various scenarios.
8. Develop modeling requirements in the form of model
type, required dimensionality and grid configuration,
and parameter adjustment, for each remedial action
group.
4-2
-------
1.3 REPORT CONTENT
This report presents the model evaluation procedure and
supporting information needed to (1) identify potential
remedial actions for a given discharge scenario, (2) identify
key processes which should be simulated, (3) evaluate
alternative models, and (4) evaluate specific models for
application. Sections 2 and 3 summarize migration and fate
processes and remedial actions, respectively. These summaries
are basic and assume little prior reader familiarity with
chemical transport and fate. The experienced reader may wish
to use these sections only for reference while reading the
remainder of the report.
Section 4 is a summary of eight case histories which show
typical discharge scenarios and remedial responses. Where
appropriate, use of models in each response process is
described. These cases form a background for the model
selection process. Analytic and simplified assessment
techniques are described in Section 5, while numerical models
are discussed in Section 6. Descriptions of each level of
technique summarize capabilities, data needs, and general
computation method. Section 7 provides modeling requirements
for surface water remedial actions. These requirements
include specific adjustments required for each remedial
measure, as well as parameter estimation guidance.
References, including those for models mentioned in the
report, follow Section 7.
4-3
-------
SECTION 2
MIGRATION AND FATE
2.1 OVERVIEW
The migration and fate of chemicals in surface waterbodies
results from both physical and chemical processes. Physical
processes cause movement of chemicals, while chemical
processes cause degradation and transformation of pollutants.
Chemical processes affect migration when changes of state or
physical properties occur (e.g., precipitation of a chemical
or sorption onto sediments). Mills, et al., (1982) provide
survey level discussions of all important processes, while
Callahan, et al., (1979) discuss chemical processes affecting
the fate of priority pollutants. Table 2.1 lists important
instream migration and fate processes in the order they are
discussed in this chapter. The following subsections provide
brief descriptions of each of the key processes and are not
intended to be comprehensive. Rather, they offer a basis for
later discussions of remedial actions and model applications.
2.2 PHYSICAL PROCESSES
2.2.1 Overview
Physical processes may be lumped into 3 groups.
1. Advection: Transport of the pollutant at the same
velocity as surrounding water molecules, in vertical
or horizontal directions.
2. Dispersion: Spreading of the pollutant plume in the
water column as a result of molecular diffusion,
turbulent diffusion and shear-flow dispersion.
Molecular diffusion represents the scattering of
molecules from random motions, and is dependant on the
viscosity of the fluid and the size of the particles.
Turbulent diffusion operates on a larger scale. It
4-4
-------
TABLE 2.1 IMPORTANT PROCESSES: PHYSICAL AND CHEMICAL
I. Physical
Advection of flow
Vertical
Horizontal
Dispersion
Longitudinal
Transverse and vertical
Sedimentation
Advection of sediment
Erosion: scour of native material and
resuspencsion of contaminated
sediments
Deposition: settling
II. Chemical
o Hydrolysis
o Oxidation
o Photolysis
o Bio-degradation
o Bio-accumulation
o Precipitation/dissolution
o Volatilization
o Adsorption
4-5
-------
represents mixing and spreading of particle clouds
from un-steady flow conditions and random velocity
fluctuations. Shear-flow dispersion is caused by
spatially-averaged gross velocity differences in the
flow which create shearing and spreading movement
during advection. It applies in 2-D flow when a
spatial average is taken. These three mechanisms will
often be lumped into a term called "effective"
dispersion for each dimension (x, y, z) (Orlob, 1971).
3. Sedimentation: This may be considered a form of
advective transport for particulate matter.
Pollutants sorbed onto suspended or bed sediments
differ in their transport rates from that of the water
because of the varying densities and radii of the
contaminated sediments. Sediments may be transported,
deposited, and resuspended in the water column,
depending on the hydrodynamics of the system.
Surface water bodies differ substantially in the way physical
processes operate due to the overall size, geometry and
boundary conditions of each water body. The full range of
surface water bodies can be classified into one of three
types: impoundments, rivers and estuaries. The energy needed
to drive these processes in water bodies may result from the
following sources:
o Density differences in water due to temperature,
distribution of dissolved solids (including salinity),
and sediment concentration (rare cases)
o Energy gradient due to wind shear hydraulic forces,
gravity, or boat/ship traffic
o Coriolis force which is due to the earth's rotation
and imposes lateral forces on a flow (Important only
when large waterbodies are analyzed)
o Water momentum at boundaries including tidal and fresh
water inflow and outflow
o Mechanical energy transfer due to wind shear on the
water surface and waves induced by wind
Figure 2.1 is a diagram of the important physical processes in
lakes, rivers, and estuaries. The range of important
processes for each water body are represented by the
horizontal lines along the bottom of the figure.
4-6
-------
HEAT TRANSFER
£>.
I
--J
TIDAL (SALINE)
ADVECTION
r—>• ADVECTION + DISPERSION
IMPOUNDMENTS
RIVERS
ESTUARIES
Figure 2.1 Flow diagram of important physical processes,
-------
2.2.2 Rivers/Streams
Rivers are characterized by uni-directional flow which is
often well-mixed. When a discharge into a river occurs,
three stages of mixing may follow (Neely et al., 1976):
(1) initial buoyancy and momentum of the spilled material, (2)
lateral dispersion across the channel width, and (3)
longitudinal dispersion downstream.
Dominant transport mechanism: Longitudinal advection and
longitudinal ("effective") dispersion. Assumes river is
relatively shallow, and well-mixed across cross-section and
depth.
Significant parameters:
o Longitudinal dispersion coefficient
o Mean velocity
o Cross-sectional area
o Depth
o Bottom roughness and slope
o Sediment size
o Vertical and lateral locations of inlets/outlets
Environmental conditions of concern:
o Precipitation and surface runoff of watershed
o Evaporation
o Scouring of the channel bed
o Deposition of sediments
o Water temperature (affects sediment transport)
2.2.3 Impoundments
Lakes, impoundments, and reservoirs are characterized by
relatively low velocities (except at inflow/outflow areas),
high retention time, and large surface areas. All of these
properties enhance heat transfer with the atmosphere. Of
primary concern are impoundments in the North American
temperate region: lakes that are monomictic or dimictic (one
or two turnovers of lake water per year, respectively). This
annual cycle of vertical mixing followed by stratification is
caused by wind stress on the surface, density differences in
the water caused by solar insolation, and changes in air
temperature (Fischer et al., 1979).
Dominant transport mechanism: Vertical and longitudinal
advection and effective dispersion in the "x" and "z"
directions. Representation of transport often requires a
4-8
-------
2-D (vertical) simulation. In large, shallow lakes, a 2-D
x-y simulation may be more appropriate.
Significant parameters (in addition to those for rivers):
o Wind stress
o Boundary conditions: Shoreline shape, mixed depth,
boundary roughness
o Vertical and horizontal dispersion coefficients
o Inflow/outflow rates
o Rate of heat transfer at surface
o Detention time
Environmental conditions of concern (in addition to those
for rivers):
o Air temperature
o Specific humidity
o Solar insolation
o Dominant wind direction and speed
2.2.4 Estuaries
Estuaries may be the most complex waterbodies in hydrodynamic
terms. Primary forces affecting transport are tidal
variations in water surface elevation, wind stress, fresh
water inflow, and internal density differences. Some
estuaries with large surface areas may also be influenced by
Coriolis forces.
Dominant transport mechanism: Longitudinal and lateral
advection and dispersion. Although a 3-D representation is
desirable in many stratified estuaries, most numerical
simulations use a 2-D vertical/longitudinal, or lateral/
longitudinal modeo., which is tidally varying. However,
pseudo 2-D (network) simulations have been applied to
shallow, well-mix<»rt systems.
Significant parameters (in addition to those for
impoundments):
o Tidal exchange at seaward boundary
o Tidal circulation within estuary
o Tidal height variation
o Tidal period
o Freshwater inflow (inland boundary condition)
4-9
-------
Environmental conditions of concern (in addition to those
for impoundments):
o Types and concentrations of suspended material
2.3 CHEMICAL PROCESSES
The instream processes listed in Table 2.1 constitute the
various means of degradation and transformation of a pollutant
in the water body. Figure 2.2 is a flow chart of these
processes and their interactions. Descriptions of each
process follow in standardized form, including significant
parameters, environmental conditions of concern, and relation
to other processes. Speciation processes such as acid-base
equilibria are different for each pollutant, and are
considered implicitly in the descriptions of the
chemical/biological processes. Table 2.2, a matrix adapted
from Callahan et al. (1979), summarizes the relative
importance of degradation processes affecting the aquatic fate
of priority pollutants. The matrix originally incorporated
only organics; heavy metals and inorganics have been added.
As the matrix is reviewed, trends between pollutant group and
important processes will be noticed. For example, dominant
fate processes for pesticides include sorption, volatilization
and bio-degradation; for aliphatic hydrocarbons (compounds
with carbon atoms formed in open chains, such as chloroform
and vinyl chloride), volatilization; and for metals and
inorganics, sorption and bio-accummulation.
2.3.1 Hydrolysis
Hydrolysis may be defined as any reaction (without the aid of
light or micro-organisms) in which a chemical combines with
water molecules to form a new compound. Many hydrolysis
reactions are pH-dependent. Significant parameters include
hydrolysis rate coefficients which are dependent on the
chemical structure of the compound, pH, and temperature. Rate
constants for particular compounds can be obtained in
literature or be determined by standard laboratory tests
(Mills et al., 1982). Environmental conditions of concern
include pH and water temperature.
Hydrolysis affects other chemical processes by either creating
new, more active compounds or replacing active compounds with
relative inert ones. Biodegradation, volatilization and
bio-accumulation may be affected in this way.
4-10
-------
VOLATILIZATION
ATMOSPHERE
SENSITIZED
PHOTOLYSIS
DIRECT
PHOTOLYSIS
ADSORPTION
PARTICULATE
POLLUTANT
DESORPTION
HYDROLYSIS-}
HYDROLYSIS
^-BIOACCUMULATION
OXIDATION-*
BIOACCUMULATION
ADSORPTION
DESORPTION
SEDIMENTATION
\ ' v ' \ \ \ \
SEDIMENTS
\ \ \ \ \
\ \ \ \ \ \
BIODEGRADATION
DEPURATION
BIODEGRADATION
PARTICULATE
* DAUGHTER PRODUCTS
ALSO SUSCEPTABLE TO
CHEMICAL PROCESSES
Figure 2.2 Diagram of chemical and biological processes (after Schnoor
and McAvoy, 1981).
-------
TABLE 2.2 POLLUTANT VS. PROCESSES MATRIX
(after Callahan et al., 1979)
Compound
Process
PESTICIDES
Aerolein
Aldrin
Chlordane
ODD
DOE
DOT
Dleldrin
Endosulfan and Endosulfan Sulfate
Endrln and Endrin Aldehyde
Heptachlor
Heptachlor Epoxlde
Hexachlorocyclohexane (o,Bi5 isomers)
-Hexachlorocyclohexane (Lindane)
Isophorone
TCDD
Toxaphene
PCBs and RELATED COMPOUNDS
Polychlorinated Biphenyls
2-Chloronaphthalene
HALOGENATED ALIPHATIC HYDROCARBONS
Chloromethane (methyl chloride)
Dichlorotnethane (methylene chloride)
Trichloromethane (chloroform)
Tetrachloromethane (carbon tetrachlorlde)
Chloroethane (ethyl chloride)
1,1-Oichtoroethane (ethylidtne chloride)
1,2-Dichloroethane (ethylene dichloride)
1,1,1-Trichloroethane (methyl chloroform)
1,1,2-Jrichloroethane
1,1,2,2-Tetrach1oroethane
II 5
d -
I 2
+
+
+
•f
Key to Syntols:
+ Predominant fate determining process - Not likely to be an important process
+ Could be an important fate process 7 Importance of process uncertain or not
known
(continued)
4-12
-------
Compound
TABLE 2.2 (continued)
Process
I
Hexachloroethane
Chloroethene (vinyl chloride)
1,1-Oichloroethene (vinylidene chloride)
1,2 - trans_-01 ch 1 oroe thene
Trichloroethene
Tetrach1oroethene (perch1oroethy1ene)
l,2-01chloropropane
1,3-01ch1oropropene
Hexachlorobutadiene
Hex ach1orocyc1opentad1ene
Bromomethane (methyl bromide)
Bromodichloromethane
Dlbromochloromethane
Trlbromomethane (bromoform)
Dlchlorodifluoromethane
Trichlorofluoromethane
HAL06ENATED ETHERS
Bls(choromethyl) ether
Bis(2-chloroethy1) ether
Bis(2-chloroisopropyl) ether
2-Chloroethyl vinyl ether
4-Chlorophenyl phenyl ether
4-Bromophenyl- phenyl ether
Bis(2-chloroethoxy) methane
MONOCYCLIC AROMATICS
Benzene
Chlorobenzene
l,2-01chlorobenzene (£-d1Chlorobenzene)
1,3-Dichlorobenzene (m-d1Chlorobenzene)
1,4-01Chlorobenzene (p_-d1chlorobenzene)
1,2,4-Trichlorobenzene
Hexachlorobenzene
Key to Symbols:
++ Predominant fate determining process - Not likely to be an Important process
+ Could be an important fate process 7 Importance of process uncertain or not
known
(continued)
4-13
-------
Compound
TABLE 2.2 (continued)
Process
Ethylbenzene
Nitrobenzene
Toluene
2,4-Dinitrotoluene
2,6-Dinitrotoluene
Phenol
2-Chlorophenol
2,4-Dichlorophenol
2,4,6-Trichlorophenol
Pentachlorophenol
2-Nitrophenol
4-Nitrophenol
2,4-Dim'trophenol
2,4-Dimethyl phenol (2,4-xylenol)
£-chloro-m-cresol
4,6-Dinitro-£-cresol
PHTHALATE ESTERS
Dimethyl phthalate
Diethyl phthalate
Di-n-butyl phthalate
Di-n-octyl phthalate
Bis(2-ethylhexyl) phthalate
Butyl benzyl phthalate
POLYCYCLIC AROMATIC HYDROCARBONS
Acenaphthene
Acenaphthylene
Fluorene
Naphthalene
Anthracene
Fluoranthene^
Phenanthrene
Benzo(a)anthracene
Benzo(b)fluoranthene
Benzo (.k) f 1 uor anthene
Chrysene
*»
tb
4-
•f
+
Key to Symbols:
^ Predominant fate determining process - Not likely to be an important process
+ Could be an important fate process ? Importance of process uncertain or not
known
(continued)
4-14
-------
TABLE 2.2 (continued)
Compound
Process
Pyrenec
Benzo(ghi )perylene
Benzo(a)pyrene
Dibenzo(a,h)anthracene
Indeno(l,2,3-cd)pyrene
NITROSAMINES AND MISC. COMPOUNDS
Dimethylnitrosamine
Diphenylnitrosamine
Di-n-porpyl nitrosamine
Benzidine
3,3'-Dichlorobenzidine
1,2-Diphenylhydrazine (Hydrazobenzene)
Acrilonitrile
METALS AND INORGANICS
Asbestos
Antimony
Arsenic
Berylumm
Cadmium
Copper
Chromium
Cyanides
Lead
Mercury
Nickel
Selenium
Silver
Tha11i urn
Zinc
Key to Symbols:
++ Predominate fate determining process - Not likely to be an important process
+ Could be an important fate process ? Importance of process uncertain or not
known
Notes
aBiodegradation is the only process knoen to transform polychlorinated biphenyls
under environmental conditions, and only the lighter compounds are measurably
biodegraded. There is experimental evidence that the heavier polychlorinated
biphenyls (five chlorine atoms or more per molecule) can be photolyzed by ultra-
violet light, but there are no data to indicate that this process is operative
in the environment.
Based on information for 4-m'trophenol
cBased on information for PAH's as a group. Little or no information for these
compounds exists.
4-15
-------
2.3.2 Oxidation
Two general
environment:
types of chemical oxidation occur in the aquatic
o photo-oxidation, in which photolysis, either direct or
by interaction with a photosensitizer, serves as the
driving force
o thermal or auto-oxidation, known simply as oxidation
(occurs when the pollutant reacts with oxidants in
solution)
The term oxidation in this report will refer to all oxidizing
processes except photo-oxidation. Significant parameters
include the base oxidant rate coefficient for a pollutant and
the free radical oxygen concentration. Environmental
conditions of concern include water temperature and reaeration
rate which affect oxygen concentration. Oxidation can affect
other processes in three ways: by producing reducing
conditions (inhibits bio-degradation), by altering solubility
(affects precipitation), and by lowering reactivity (affects
volatilization and photolysis).
2.3.3 Photolysis
Photo-chemical transformation may occur directly or
indirectly. Direct photolysis involves the absorption of
light by the pollutant, placing electrons in an excited state
from which reactions can transpire. Indirect photolysis
occurs when another chemical absorbs light, and in its excited
state, undergoes reaction with the pollutant (Mills et al.,
1982). Significant parameters include the molar absorption
coefficient (specific to each chemical) and the incident light
intensity at a specific wavelength, which is a function of the
mixed depth of water and attenuation of light by natural
waters. Environmental conditions of concern include vertical
mixing of the water column, turbidity caused by suspended
sediments, water temperature, and incident light at the water
surface. The oxidation of material may result in reducing
conditions, inhibiting bio-degradation.
2.3.4 Volatilization
Volatilization is actually a physical process in which the
dissolved pollutant changes state and is transported from the
water to the atmosphere. Current evidence indicates that it
4-16
-------
is the dominant aquatic fate process for low molecular weight,
non-polar compounds that don't easily degrade biologically or
chemically (Callahan et al., 1979). Significant parameters
include Henry's Law constant for compounds (essentially a
partitioning coefficient between the gas and liquid phases),
and reaeration rate, which is a function of wind speed and the
mixed depth of water. Environmental processes of concern
include water temperature, dissolved oxygen concentration, and
vertical mixing. Increased turbulence increases the
reaeration rate, enhancing volatilization (Smith et al.,
1981).
2.3.5 Adsorption
The adsorption process involves the exchange of a pollutant
between the dissolved and adsorbed states. Usually this
includes chemi-adsorption, or chelation with the sorbent, as
well as physical adsorption, in which the sorbate is loosely
held by ionic attraction. Consequently, the type and amount
of suspended sediments strongly influence the type of
adsorption isotherm (graph of sorbed material vs. material
dissolved at a specific temperature) that describes the
sorption process. A linear isotherm is often assumed at low
pollutant concentrations (Karickhoff, 1979). Because
contaminated particulates may settle out of the water column,
the bed sediment may serve as a repository or sink allowing
release and/or resuspension over a long time period. For most
organic and non-polar compounds the amount of organic carbon
in the sediment determines the extent of sorption (Mulkey et
al., 1982).
Significant parameters include partition coefficient at
equilibrium (for hydrophobic or low solubility pollutants) or
expressed on an organic carbon basis, and dissolved
concentration of pollutant. Environmental conditions of
concern include pH (particularly important when particulates
are clay or organic material), water temperature, and sediment
concentrations and organic content. Adsorption rates may be
increased by vertical mixing and turbulence which causes
suspension of sediments or may be decreased by deposition of
sediments. Sorbed chemicals are not generally subject to
photolysis or volatilization, but may be more or less
available to bio-degradation.
2.3.6 Bio-degradation
Microbial breakdown is significant because of the high species
diversity and metabolic rates of microbes in the natural
4-17
-------
environment. Pollutants are most susceptible to breakdown
when they sorb with suspended sediments or settle out of the
water column onto the bed. This resulting increased surface
area can cause an increase in biodegradation (Mills et al.,
1982). Significant parameters include pollutant
concentration, standing microbial biomass, specific growth
rate constant for the bacterial group, and metabolic pathways.
Environmental conditions of concern include pH, water
temperature, reaeration and resultant dissolved oxygen
concentrations, trace nutrient concentrations, and specific
toxicity to bacteria.
2.3.7 Bio-accumulation or Bio-magnification
Bio-accumulation or magnification is an important process for
the partitioning of hydrophobic pollutants. Such pollutants
are usually lipid-soluble; hence, uptake via absorption or
ingestion results in the accumulation of the pollutant in the
fatty tissue of an organism. An octanol-water partition
coefficient is used to describe the uptake as octanol
resembles body fat (Neely et al., 1974).
Significant parameters include an octanol-water partition
coefficient (determined from laboratory test or structure-
activity relationship) and solubility of pollutant in water.
Environmental conditions of concern include fish and other
biomass standing crops, water temperature (can affect rates of
uptake and metabolism of organisms), and food chain order.
2.3.8 Precipitation/Dissolution
The solubility of a contaminant in water is defined as the
maximum amount of that chemical that will dissolve in pure
water at a specified temperature (Lyman et al., 1982). Above
this amount, two phases may exist: the saturated aqueous
solution and the precipitated solid. Most organic pollutants
have low solubilities (Lyman et al., 1982). It is probable
that their maximum solubility would not be reached in the
aquatic environment except where high, localized,
concentrations exist (as in a spill). However, fluctuating
environmental conditions, such as pH or temperature, may
alternately cause a pollutant to dissolve or precipitate and,
as a result, will affect the mode of transport and importance
of some chemical or biological processes.
Significant parameters include octanol-water coefficient
(Kow )/ solubility product (KSp), and distribution (partition)
coefficient (K^). Environmental conditions of concern include
4-18
-------
pH, temperature, total dissolved solids, dissolved organic
matter, degree of mixing in water column, and pressure (rare
cases).
The form of the contaminant (dissolved or solid) will control
the transport mechanisms in the aquatic system. Soluble
pollutants can be easily distributed, as they move with water
molecules. These pollutants usually exhibit low sorption and
bio-concentration characteristics (Lyman et al., 1982).
Insoluble pollutants may behave similarly to suspended
sediments; they may be deposited, resuspended, and partitioned
between the sediment and biotic compartments. The solubility
will affect other processes: photolysis, hydrolysis and
oxidation are enhanced by high solubility, while sorption and
bioconcentration are often enhanced by precipitation.
4-19
-------
SECTION 3
REMEDIAL ACTIONS AND AFFECTED CRITICAL PROCESSES
3.1 OVERVIEW
Remedial actions may be classified into four groups:
dilution, containment, removal and treatment. Individual
remedial actions (such as mechanical dredging) are divided
into these groups and described herein, with attention given
to the dimensionality required for simulating that action, as
well as the affected critical processes. Table 3.1 provides
an outline of the remedial actions considered.
The purpose of this section is to: 1) briefly overview the
design objectives of each of the measures listed in Table 3.1
and 2), identify which water bodies and processes are affected
by these measures and how they are affected. This type of
information is needed to support the development of guidance
on the use of models to evaluate remedial action performance.
Detailed information regarding design of these remedial
actions, potential applications and their effect on surface
water bodies can be found in JRB (1982), Raj and O1Parrel
(1977), Thibodeaux (1979), and other sources.
Table 3.2 is a matrix of environmental processes vs. remedial
actions. Environmental processes are grouped as either
chemical/biological or physical processes, similar to the
format for descriptions of processes presented earlier.
Remedial actions are grouped in a fashion similar to their
descriptions earlier. This matrix will allow the reader to
identify specific remedial actions with affected processes.
This information should be kept in mind while the following
matrices are reviewed.
As an example of matrix interpretation, consider the following
example. Mechanical dredging is a common form of waste source
removal for contaminated sediments in a shallow, low flow
waterbody. The use of this action may enhance the following
in-stream processes, as denoted by a "+" on the matrix:
photolysis, volatilization, sedimentation, and dispersion.
These processes, then, are more important in assessing
4-20
-------
TABLE 3.1 OUTLINE OF REMEDIAL ACTIONS
I. Dilution
II. Containment
o Booms
o Silt Curtains
o Cofferdams
o Barriers/diversions
o Capping
III. Removal
o Skimming
o Hydraulic dredging
o Mechanical dredging
o Excavation
IV. Treatment
o In-situ
o On-site
4-21
-------
TABLE 3.2 REMEDIAL ACTION VS. PROCESSES MATRIX
PROCESSES
ACTIONS
TRANSFCRriATION
/
PHYSICAL
NO ACTION
REMOVAL
MECHANICAL
DREDGING
EXCAVATION
HYDRAULIC
DREDGING
BARRIERS/
DIVERSIONS
SKIMMING
DILUTION
CONTAINMENT
COFFERDAMS
BOOMS
SILT CURTAINS
CAPPING
TREATMENT
IN-SITU
ON-SITE
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
•f
t
LEGEND:
+ = ENHANCES THE PROCESS IN RELATION TO NO ACTION
- = MITIGATES THE PROCESS IN RELATION TO NO ACTION
0 = DOES NOT AFFECT THE PROCESS
4-22
-------
transport and fate of a pollutant using this remedial action.
Also of importance, the processes of bio-degradation and
adsorption may be decreased (denoted by "-"). The rest of the
in-stream processes should not be affected (denoted by "0").
Model dimensionality required to adequately represent the
effects of these actions is not typically dependent on the
action, but relates to the water body shape, size and type.
Exceptions to this are noted.
3.2 DILUTION
This action can reduce in-stream concentrations by increasing
flow and reducing hydraulic retention time. This may be
accomplished in water systems with controllable flows, as in
rivers with dams upstream. Dilution will not appreciably
affect the geometry or dimensionality of the flow but may
increase dispersion due to mixing and higher velocities.
Affected processes are limited to dispersion and advection,
which are both increased.
3.3 CONTAINMENT ACTIONS
Containment actions separate chemicals from the rest of the
waterbody. Consequently, they often alter the geometry of the
body and flow direction. Both advection and dispersion change
as a result. These changes can usually be represented by a
2-D (horizontal plane) model. Changes in chemical processes
depend on the action, as discussed below.
3.3.1 Booms
Booms can be used to intercept or contain light, miscible
pollutants (Specific Gravity <1) in a surface slick. This
limits their use to a period immediately after the spill
before the plume disperses or to small impoundments or
dead-end branches in estuaries where surface wave action and
wind shear are at a minimum (Raj and O'Farrel, 1977).
Skimming may be used in conjunction for removal of pollutants.
Figure 3.1 shows possible deployments of booms. Advection and
dispersion will typically be decreased in the surface layer
due to the blockage of wave action and surface currents by the
booms. Because the chemical slick is contained, chemical
concentrations will remain high and processes which depend on
concentration (e.g., volatilization) may increase.
4-23
-------
Conversely, self shading due to slick capacity may decrease
photolysis.
3.3.2 Silt Curtains
Silt curtains and nets serve a function similar to booms, but
may also trap suspended material, such as plumes downstream
from a dredging operation. Figure 3.1 shows possible
deployments of silt curtains. These actions have much the
same effects on processes as booms, except that they can
affect the entire water column and contain chemicals and
sediments from surface to bottom. Use of silt curtains should
be limited to situations with low velocities and minimal wave
action to avoid failure.
3.3.3 Cofferdams
Cofferdams are single wall barriers usually made out of earth
or steel, and constructed for shallow streams or rivers, or
for those waterbodies with low flow. The dams divert or
contain streamflow so that an area can be dewatered or
isolated in preparation for excavation or dredging actions.
Two possible configurations for cofferdams are shown in Figure
3.2. These structures typically confine the flow and,
especially in rivers, will cause increased velocities and
concomitant increases in dispersion, scour, and sediment
deposition. If a cofferdam isolates contaminants from the
water body all contaminant release and transport processes are
minimized.
3.3.4 Barriers/Diversions
This group includes all physical structures that impede flow
and divert water away from contaminated area by using a
separate diversion channel. An example of this is a complete
stream flow diversion around a contaminated area described by
Zaccor (1981) or as shown in Figure 3.3. Complete diversions
are usually required when an entire stream cross-section is
heavily contaminated and removal of the contaminants is
required. The waterbody boundaries are changed, the flow is
entirely removed from the vicinity of the contaminants, and
chemical processes are stopped.
4-24
-------
..^» Maze (Not Recommended)
Legend:
O Mooring Buoy
X Anchor
J* Single Anchor
or Piling
U-Shaped
In-Stream
Curtain Movement Due \
to Reversing Currents ~
"C"
U-Shaped
Anchored On-Shore
Estuary
Circular or Elliptical
Figure 3.1 Typical boom or silt curtain deployment
configurations (from Barnard, 1978).
4-25
-------
Area of
sediment
dewatering
and excav-
ation
OR:
Temporary
sheet-pile
bank
reinforcement
Sediment
excavation
Temporary
sheet-pile bank
reinforcement
Figure 3.2 Isolation for sediment excavation using single cofferdam
(from JRB, 1982).
-------
Temporary sheet-pile;
remove after pipeline construction
Diversion
channel;
excavate, place
corrugated metal
pipe or similar
conduit
^±i±±±±±i2 Flow
Uostream cofferdam
cccc^ upstream coneraam
Sediment
dewatering
and excavation
Downstream
Temporary x£~~
sheet-pile
Riprap for
outlet protection
Figure 3.3 Streamflow diversion for sediment excavation
using two cofferdams and diversion channel
(source: JRB, 1982).
4-27
-------
3 . 3.5 Capping
Capping with impervious materials may be applied to localized
"hotspots" on the sediment bed, particularly where
indentations occur. Problems can occur if stream velocity
causes scour or depth is great, making verification of
effectiveness difficult. Other problems include locating and
treating waste deposits. Once installation is complete,
movement of contaminants into the water column by scour,
desorption or other processes will effectively cease. No
other processes will be affected. During installation, scour
and mixing may temporarily increase contaminant mobilization.
3.4 REMOVAL MEASURES
Removal measures are designed to eliminate the source of
contamination from the water body. All contaminant-related
processes will, consequently, be minimized. Four types of
methods are available (see Table 3.1) and can be chosen to fit
particular water body and contaminant conditions. These
measures are often used in conjunction with containment
measures to ensure that any chemicals mobilized by the removal
process are retained at the site.
3.4.1 Skimming
Skimming is used when the pollutant has a specific gravity
less than 1 and is contained within an impoundment or by means
of surface barriers (booms) (Raj and O'Farrel, 1977). It is
not as efficient when there is significant turbulence near the
surface as in an estuary or fast moving stream or when strong
winds are present. During skimming, increased turbulence and
higher local velocities will tend to disperse chemicals unless
effectively contained.
3.4.2 Hydraulic Dredging
This type of dredging includes the use of centrifugal pumping
systems and portable hydraulic pipeline dredges. Centrifugal
pumping systems can cut and chop heavy, viscous material (JRB,
1982). It may be applicable to spills of immiscible, high
specific gravity material that settles in pools (Thibodeaux,
1979). Both types of hydraulic dredges may be used in
impoundments or streams. Advantages over mechanical dredging
include: minimal turbidity is created, dewatering of spoils
4-28
-------
isn't necessary, and it is suitable for removal of material in
a wide range of consistencies, from floating liquid to
sediment/sludges. However, spoil management actions are more
important due to the large volume of material removed and must
be included along with the use of diversions or barriers in
any dredging plan. During dredging, turbulence and locally
high velocities may resuspend and/or disperse chemicals unless
effectively confined. These effects are less severe than
those caused by mechanical dredging. Disposal of spoils
involves large quantities of water, which may contaminate the
same or other water bodies unless proper storage or treatment
is implemented.
3.4.3 Mechanical Dredging
This measure may be used under conditions of slow, shallow
flow. It should be used conjunctively with either streamflow
diversion or silt curtains to prevent uncontrolled transport
of resuspended contaminated sediments. Applicable waterbodies
include streams, small rivers, lake shorelines, and small and
then dewatered. However, supernatant from the dredge spoil
poses an additional problem. Mechanical dredging will disturb
bottom sediments and distribute them over the water column,
resulting in increases in all migration and fate processes.
3.4.4 Excavation
This action may be used in conjunction with barriers and
diversions, or may be applied to marshes or soil where
contaminants are entering the surface water via leaching or
runoff. Since excavation implies the removal of dry soil, the
dewatering action (containment, diversion) is always
considered as a conjunctive measure. Because the excavation
site is isolated from the water body, removal of materials
causes no changes in processes.
3.5 TREATMENT MEASURES
Remedial actions relating to the treatment of discharged
hazardous materials in waterways are minor in importance.
Similar actions are much more important at uncontrolled
hazardous waste sites. Quite often treatment actions will be
used in conjunction with a removal action, such as dredging.
In these cases, the removal action has the greatest impact on
in-stream processes.
4-29
-------
Treatment methods may be physical, chemical, or biological.
They may be applied in-situ, or on-site. In-situ treatment
applications are rare, and limited to physical or activated
carbon filtration systems. On-site applications are more
common. Because on-site actions are outside of the waterway,
their effect on in-stream processes is rarely felt, especially
if the contaminated sediments and/or water are hauled offsite.
However, if the material is treated and then released back
into the waterbody, some impacts may be felt. Advection and
dispersion may be increased locally by the discharge. All
chemical migration and fate processes will operate on the
discharge plume as they would on any point source of
contamination.
4-30
-------
SECTION 4
USE OF REMEDIAL ACTIONS AND MODELING: CASE HISTORIES
4.1 OVERVIEW
In order to illustrate the need for remedial action assessment
tools, case histories of discharge incidents and EPA responses
are described below. These represent "typical" or common
discharge scenarios that may occur in rivers, lakes, and
estuaries. Descriptions of discharge scenario types are
provided in Table 4.1. A hypothetical release incident that
illustrates some of the release mechanisms is provided in
Figure 4.1.
4.2 CASE HISTORIES
Eight case histories are briefly described below. Contained
within each is an identification of the critical environmental
processes, remedial actions considered, and modeling efforts.
4.2.1 Hudson River PCS Spill
Approximately 500,000 Ibs. of PCBs were discharged into the
Hudson River near Troy, New York over a period of time. It
was estimated through an Environmental Impact Statement that
$40 million would be needed for remedial actions to get the
PCB concentration down to 50 ppm. Critical processes were
identified as sorption and sedimentation; at high flows, the
PCBs would desorb from scouring action on the sediment bed.
Remedial actions chosen were: mechanically dredge (using a
clamshell dredge) 40 "hotspots" and discharge off-site.
Another remedial action of "capping" was considered infeasible
due to costs and the fact that the river is a navigable
waterway. The models were used to estimate PCB transport.
The numerical sediment model HEC-6 (Hydrologic Engineering
Center, 1977) was used with the WASP model (Water Quality
4-31
-------
TABLE 4.1 TYPES OF DISCHARGE SCENARIOS (after Mills et al., 1982)
I
U)
to
DIRECT
• MAY EMANATE FROM BARGE/SHIP DUMPING,
OR PIPELINE RUPTURE
• SPECIFIC GRAVITY >1,0, HYDROPHOBIC, OR
HAVING HIGH SORPTIONj POLLUTANT SETTLES
ON BED
- ADVECTED ALONG BOTTOM
- RE-ENTRAINED BY RESUSPENSION
OF SEDIMENTS
- DIFFUSION FROM SEDIMENT BED
- MAY UNDERGO REDUCTION OXIDATION
VIA MICROBIAL ACTIVITY IN THE BED
• SPECIFIC GRAVITY<1,0, HYDROPHILIC, OR
HAVING LITTLE SORPTION; POLLUTANT IS
ENTRAINED IN WATER COLUMN
- VOLATILIZATION AND PHOTOLYSIS MAY
BE IMPORTANT
- OTHER REACTIONS (l.E,, HYDROLYSIS)
MAY AFFECT SOLUBILITY OR ABILITY TO
SORB
- ADVECTED AND DISPERSED ACCORDING TO
BUOYANCY, MOMENTUM (NEAR FIELD),
AND DOMINANT MIXING PROCESSES
INDIRECT
• MAY RESULT FROM TRUCK/RAIL OR WASTE
SITE ACCIDENT, OR FROM STORM EVENT.
SIMILAR BEHAVIOR AS "DIRECT" DISCHARGES
• SURFACE RUNOFF FROM SPILL ON LAND
- TRANSPORT VIA FIRST STORM EVENT
• CONTAMINATED TRIBUTARY INFLOW
- SMALL ENOUGH TO BE CONSIDERED
- A POINT SOURCE, OR,
- OUTSIDE SYSTEM BOUNDARIES
• GROUND WATER RECHARGING SURFACE
WATER, OR DIRECT LEACHING
- RECHARGE DEPENDANT ON WATER
TABLE LEVEL AND STREAM FLOW
- VIEWED AS CONTINUOUS INPUT
• WET/DRY DEPOSITION FROM AIR TO
SURFACE WATER (l,E., ACID RAIN ON
LARGE LAKES)
-------
-------
Analysis Simulation Program by DiToro et al.,1982) to
determine PCB distribution in the biotic and abiotic
compartments.
4.2.2 Waukegan Harbor PCB Spill
A corporation was discovered in 1975 to be discharging wastes
containing PCBs into Waukegan Harbor, on Lake Michigan. Total
amount of PCBs dumped was estimated to be 1.3 to 1.7 million
pounds. Levels of contamination exceeded the F.D.A. fish
tissue and sediment criteria level of 50 ppm (EPA, 1982).
Simulations were performed using the WASP program by R. V.
Thomann at Hydroqual, Inc. The objective of the study was to
quantify loading into Lake Michigan from Waukegan Harbor and
the drainage ditch where most of the waste had been dumped.
Again the critical process was identified as sorption to bed
sediments in the harbor, with sediment and advective transport
from natural flushing and dredging operations affecting
desorption from the bed. Bio-accumulation was important,
also, in light of fish tissue concentrations, but was
simulated as a source/sink; depuration (excretion and death)
and uptake ratios were simplified. Remedial actions chosen
were to mechanically dredge the harbor with turbidity control
(barriers) and to excavate the ditch.
4.2.3 Iron Mountain Mine Site
Iron Mountain Mine is defunct, and drains to the Sacramento
River via a tributary creek near Redding, California. Tailing
ponds, portals, and a pit on top of the mountain contribute a
variety of heavy metals such as zinc, copper, and cadmium in
point-source and non-point source pollution. Problems occur
in the spring when snow-melt and rains lead to a large
contaminated runoff flow. Two treatment plants can remove 75%
of the copper from controlled flows (little runoff) and as
much zinc and cadmium required with the control technology;
the problem then is exacerbated when high runoffs can by-pass
the plants. The only remedial action taken to date is the
construction of a dam on Cheswick Lake, leading into the
Sacramento River, to control or dilute flows downstream.
Critical processes are identified as advective transport
(dilution of acid drainage), and hydrolysis (metal mobility).
No modeling efforts have been done, although funding may be
provided via a feasibility study for clean-up through
Superfund.
4-34
-------
4.2.4 Kepone Contamination in the James River
In the 1970's high concentrations of the pesticide Kepone were
discovered in fish tissue and sediments in the
tidally-influenced James River, VA. The material had been
discharged over a period of time into Bailey Creek (river mile
120). Kepone is hydrophobic (low solubility/high sorption)
and is predominately transported by contaminated sediments.
Distribution is also complicated by the facts that Kepone may
bio-accumulate and that the James River is essentially an
estuarial system with complex hydrodynamics. The most
critical process identified was sorption onto specific
sediments. Estuary systems have a number of sediment types,
each with specific sorption capacities. Modeling efforts
were conducted by Onishi of Battelle using FETRA (Finite
Element Transport Model, (Onishi et al., 1979), and by
O'Conner of Hydroqual (O'Connor and Farley, 1981). The FETRA
model was used to simulate the transport of Kepone and
sediments utilizing simulated velocities and flow depth from
the EXPLORE-I Model. Organic sediments are considered to be
important carriers of organic pollutants. Hence, the FETRA
code simulated dissolved Kepone and particulate Kepone with
their sorption and transport mechanisms for noncohesive
(sand), cohesive (clay and silt), and organic sediment
separately. O'Conner also simulated transport mechanisms but
examined bio-accumulation also. No simulation of remedial
actions was conducted. Drinking water and fish harvesting
bans were temporarily enacted as remedial measures.
4.2.5 Formalin Spill on the Russian River
A one-time, finite duration spill of formalin occurred on the
Russian River in Cloverdale, CA, in May 1982. The pollutant
entered the river via surface flow and leaching into a
tributary. Critical processes were identified as
volatilization and sorption. Because of the number of
drinking water intakes along the river, a drinking water ban
was enacted in conjunction with controlling the flow by
closing the upstream dams, allowing the discharge of formalin
over time into the Pacific Ocean. However, afterwards the
in-stream concentrations were found still to be high, so the
Army Corps of Engineers decided to use dilution as a remedial
action, opening the upstream dams. No modeling efforts were
initiated.
4-35
-------
4.2.6 Triana DDT Site
This site has been releasing DDT over a long period into a
stream which leads into wetlands in Wheeler Reservoir near
Decatur, Alabama. The most critical process identified was
bio-accumulation as fish tissue concentrations were high (50
ppm) and the fish were a staple food item to the indigenous
population. The Army Corps of Engineers evaluated remedial
actions including dredging, coating of the sediments with an
impervious layer (capping) and the creation of a sediment dam
with channel diversion around the wetlands area. Modeling the
systems with EXAMS (Exposure Analysis Modeling System, Burns
et al., 1982) was suggested by TVA but has not been done to
this date.
4.2.7 Marathon Oil
In July 1982, between 150,000 - 250,000 gallons of heavy crude
oil spilled into an irrigation ditch that leads to the
Shoshone River in northern Wyoming. The spill occurred as a
result of a pipeline accident with the Marathon Oil Company.
The critical process was considered to be advective transport.
Deflection booms and siphon trucks were utilized as part of a
clean-up program. Response was more of an emergency nature
than a long-term remedial one.
4.2.8 Chlorine Barge Spill
A barge containing chlorine gas ruptured in San Francisco Bay
in 1981. Hazard from toxic fumes was considered imminent, so
the emergency response team at EPA took charge. Data from
CHRIS (Chemical Hazard Research Information System) were
utilized, as well as a gas dispersion model. No attempt was
made to examine in-stream processes as the immediate need was
to assess the toxic cloud formation.
4.3 SUMMARY
Clean-up programs have traditionally been used in response to
emergency conditions where limited time and data require
rapid, simple screening techniques. However, most of the
earlier spill incidents of note were petroleum products with
known or simple chemical properties. Modeling efforts
concerned the simulation of circulation processes in open
waters. The influx of more complex and toxic materials that
4-36
-------
degrade slowly, however, now presents additional problems over
a longer time period.
Critical processes identified in the case histories most often
were sorption, sediment migration (transport, scouring and
deposition), advective transport, volatilization and some
degradation processes such as bio-accumulation. Indeed, for
the 103 organic priority pollutants, sorption processes are
important for 60, and volatilization is important for 52
(Mills et al., 1982). Many of these pollutants are
hydrophobic and thus sorb readily and can be transported with
sediment. Advection and dispersion are also quite important,
as evidenced by the James River and Russian River cases, and
are specific for each waterbody.
Modeling efforts are not commonplace in remedial action
programs, as seen in the case histories. Simulations were
applied where it was apparent that long-term hazards could
arise from fluctuating environmental conditions and the slow
degradation of the pollutant (i.e. Kepone and PCB). In the
future, models may provide guidance for implementation of
remedial actions, including design considerations, such as
placement and size of barriers. Simplified assessment
techniques and analytical models may also be used for
screening purposes and to characterize the site conditions.
4-37
-------
SECTION 5
USE OF ANALYTICAL AND SIMPLIFIED ASSESSMENT TECHNIQUES FOR
REMEDIAL ACTION SCREENING AND ASSESSMENT
5.1 OVERVIEW
Simplified assessment techniques and analytical models play an
important part in the screening of hazards and assessment of
exposure from contaminant discharges. These simpler models
are presented to allow the user a choice between levels of
model complexity, depending upon the nature of the problem.
Their relatively simple nature allows application with little
data and resources. They can, consequently, be used on a site
for initial evaluation of site conditions and testing of
hypotheses regarding gross contaminant transport processes.
However, they can be considerably less accurate than numerical
models and are not able to adequately simulate complex
environmental conditions or the detailed effects of remedial
actions. Despite their simplicity they may require
substantial user experience and judgment to estimate
appropriate parameter values and to apply the procedures
effectively.
Simplified techniques and analytical models are similar in
that they use analytical solutions for the flow and transport
equations. Such solutions require that numerous assumptions
be made, including steady-state conditions, homogeneous
physical and chemical properties and simple flow geometries.
The simplified techniques usually produce one value, because
they are essentially comprised of one equation. These
techniques are most useful for predicting steady-state
contaminant concentrations under fixed environmental
conditions. While they may use the same equations, analytical
models can calculate concentrations over extended time periods
with variations in parameters such as flow rate. A computer
program is used to solve the analytic equation(s) repeatedly
as time steps are taken. This allows the use of analytic
models for time-dependent problems and for sensitivity
analyses where the effects of parameter uncertainty are
evaluated.
4-38
-------
5.2 USES OF SIMPLIFIED ASSESSMENT TECHNIQUES
Simple methods are useful for screening and preliminary
exposure assessments where the primary goal is to determine
compliance with instream water quality standards. They can
also be used to better define objectives, estimate the level
of study required to Attain the objectives, and to determine
the nature of analysis required (i.e., numerical, analytical
or physical modeling). Specific uses include: mixing zone
water quality criteria development and determination of peak
concentration, travel time, and concentration as a function of
distance. These uses are referred to in the simplified
assessment techniques vs. use and data matrix (Table 5.1).
The mixing zone or near-field is that area where initial
dilution of the contaminant takes place. The degree of
dilution and mixing determines the initial concentration (Co).
Because of the limited dispersion that occurs near the
discharge site, concentrations tend to be high and chronic
toxicity to biota is often a problem if the discharge is
continuous. For this reason, mixing zone criteria are
established. Simple techniques can be used, based upon the
buoyancy and momentum of the discharge, water depth, and
stream velocity, to determine whether and how compliance can
be attained. Determination of peak concentration is important
when a worst case scenario is assumed. The contaminant is
considered to be conservative (no degradation and minimal
mixing is assumed), so that a maximum concentration is
predicted.
Determination of travel time in regions of the waterbody away
from the discharge is probably the most common use of
simplified techniques. Velocity and distance are used to
determine the time it takes for a slug input to reach a given
point downstream. This point could be a drinking water
intake, or other area where health effects may be felt. Flow
is assumed to be steady and non-dispersive (plug flow).
Dispersion is considered for time of travel of a slug input.
Degradation of the contaminant is represented as a function of
time. Far field techniques are designed for this use and are
particularly applicable to rivers where advection dominates.
Variations in far field concentrations with distance from the
source can be readily determined through solution of the
analytic equation(s) at different locations. Such profiles
provide a one-, two-, or three-dimensional picture of the
effluent plume for either continuous or short duration
discharges.
4-39
-------
TABLE 5.1 SIMPLIFIED ASSESSMENT TECHNIQUES VS. USE AND REQUIRED DATA
I
*>
O
-------
5.3 CLASSIFICATION OF SIMPLIFIED ASSESSMENT TECHNIQUES
These techniques include computations that require few
parameters and may be performed on a hand calculator. They
may be used for site screening purposes to provide an initial
assessment of the extent of the hazard and to determine what,
if any, subsequent analyses should be employed. Table 5.2 is
a list of these methods with references and uses, and includes
the general groups of mixing zone and far field approaches,
transformation equations, and sediment-water interactions.
These groups are described below.
5.3.1 Near-Field Analyses
Several techniques can be used to determine the discharge
concentration after initial mixing: degree of initial mixing,
initial dilution, and mixing across width. Critical
parameters are usually stream velocity and the buoyancy,
momentum, and flow rate of the discharge. This group uses
initial dilution processes to determine the maximum
concentration after near field mixing. The calulated
concentration allows determination of mixing zone water
quality criteria, and is used as an initial concentration (Co)
in far field analyses.
The degree of initial mixing analysis can be used on rivers
to determine the distance downstream below a point source
where complete mixing occurs, or to define the boundaries of
the mixing zone. Pollutant loading is assumed to be
instantaneous. River width is a sensitive parameter in the
analysis. The simple equation computes downstream distance as
a function of lateral dispersion, river width, and stream
velocity.
Initial (near-field) dilution analysis is designed for
estuaries or coastal waters where the pollutant is discharged
through submerged diffusers (Frick, 1981). The dominant
mixing process is different from that of a river, where width
and velocity govern mixing. Mixing occurs as the buoyant
effluent plume rises from the diffuser and entrains the
ambient fluid (Mills et al., 1982). Critical to the
calculation is the degree of density stratification, port
spacing, effluent velocity to current velocity ratio, and
depth. Initial dilution values as a function of depth and
Froude number have been developed by Frick (1981) using a
plume model under various physical conditions.
The mixing across width analysis is designed to determine the
mixing zone size for lakes and wide rivers with irregular
4-41
-------
TABLE 5.2
SIMPLIFIED ASSESSMENT TECHNIQUES
FOR SURFACE WATER
Technique
I. Near-Field Analysis
o Degree of initial mixing
o Initial dilution
o Mixing across width
(lateral dispersion)
II. Far-Field Approaches
o Estuaries
Fraction of freshwater
Modified total prism
o Rivers/Lakes
Point source-
continuous
Non point source-
continuous
Spills of pollutants
III. Transformation Equations
IV. Sediment-Water Interactions
o Vertical distribution of
sorbate
o Desorption from sediment
bed
o Transport of high density/
sorbed pollutants
Reference
Codell et al., 1982
Mills et al., 1982
Fischer et al., 1979
Mills et al., 1982
Fischer et al., 1979
Tracor , 1971
Mills et al., 1982
Codell et al., 1982
Neely et al., 1976
Raj and O'Farrell, 1977
Krenkel and Novotny, 1980
Thomann , 1972
Csanady , 1973
Mills et al., 1982
Mills et al., 1982
4-42
-------
geometries, especially where it's not apparent that the far
shore affects mixing. This method is similar to the degree of
initial mixing approach, except that the discharge velocity
and geometry control near field pollutant dispersion, because
of the relatively low ambient velocities present. Critical
parameters also include depth and width of the waterbody.
5.3.2 Far-Field Analyses
Far field approches are used to determine downstream transport
of pollutants, including time of travel of a pulse input, peak
concentrations, and concentration profiles or extent of
plumes. Most often, results from mixing zone analysis (such
as Co) are used as input because far field methods do not
consider such parameters as buoyancy or momentum of the
discharge. Geometry is usually simplified, and complete
mixing across a stream width is assumed. An analytical
solution is derived from the one-dimensional transport or mass
balance equations using steady flow parameters (velocity,
depth, and cross-sectional area).
The fraction of freshwater method is a simple calculation for
pollutant transport in estuaries. Transport is determined
using the flushing time, which is the time of travel required
to move a pollutant to the mouth of the estuary. The
calculation assumes that the salinity is uniform throughout
the estuary and that net seaward flow of saline water is
proportional to the river discharge for that tidal cycle.
Mixing is assumed to be instantaneous within each estuary
segment. Plume movement is calculated based on net seaward
velocity during a tidal cycles.
The modified tidal prism approach is used to calculate
flushing time in estuaries also. Flushing time is calculated
by dividing the estuary into segments with lengths determined
by the maximum flow path of water during a tidal cycle (Mills
et al., 1982). The tidal prism is compared to the total
volume for each segment, as a measure of flushing potential.
Salinity distribution is not required. A disadvantage is that
in order to predict the flushing time of a pollutant midway in
the estuary, the method has to be applied to the whole system.
Parameter requirements include the river discharge over each
tidal cycle and segment dimensions.
The point source analysis is applicable to both a continuous
source effluent and a finite duration release of a pollutant.
Uses include prediction of: steady-state and transient
concentrations as a function of distance, advection rate past
a specified location, and transformation to other species over
a specified reach. Plug flow (no dispersion) is sometimes
4-43
-------
assumed. Concentrations are calculated by the transformation
of a given initial concentration over time. This
transformation rate is represented by an exponential term
containing transformation coefficients and a distance/velocity
ratio (which denotes the time of travel). Thus the amount of
data required is not extensive. Transformation of the
dissolved fraction can be calculated provided that the
partition coefficient for the pollutant is known.
The nonpoint source analysis is designed to calculate
steady-state or transient concentration profiles and time of
travel. The far field analysis for downstream transport is
similar to that of the point source assessment; however, the
initial concentration is calculated by estimating loading into
a specified volume of water from an adjacent land segment.
Mixing is assumed to be complete and instantaneous for each
event. Besides the data mentioned for the point source
method, river and runoff flows as well as segment length are
needed. Runoff flow may be estimated using SCS
runoff-infiltration curves. If the pollutant is not highly
soluble, a partition coefficient is needed and runoff of
contaminated sediments must also be estimated. The user is
referred to Donigian (1981), O'Connor (1967) or Mills et al.,
(1982) for more information.
A number of specific methods for one-time discharges of highly
soluble contaminants are available for determining time of
travel, concentration profile, and peak concentration. The
analyses are designed for calculating initial concentration
and downstream transport. Because the contaminant is released
as a "slug" input and not a continuous release, a different
solution technique from the continuous point source analysis
is required. The dissolved phase concentration is calculated
by using an expression containing the dissolved mass fraction,
cross-sectional area (assumed to be constant) and time. This
expression replaces initial concentration as used in
continuous effluent analyses. The transformation exponential
expression is also more complex, utilizing a simplified form
of the advection/dispersion equation (containing a steady
velocity, distance, and longitudinal dispersion coefficient),
transformation coefficients, and elapsed time. An
instantaneous mixing analysis can be performed first, in order
to find the volume of water needed to dilute the pollutant to
its solubility limit. Assuming concentrations near the
solubility limit are rapidly attained, the far field analysis
can be performed. The user is referred to Mills et al. (1982)
for further detail.
4-44
-------
5_._3.3 Transformation Equations
Transformation equations primarily serve as screening tools
based on chemical characteristics. They may also be used in
lakes where advection is not a dominant means of transport and
fate. Point source and nonpoint source loading data, as well
as other hydrological data, must be compiled if the
application is for a specific waterbody.
These equations describe the fate of pollutants over time
rather than over space. For simplicity, these removal
processes are based on equilibrium rates, and are first-order
reactions (e.g., dependent only on the concentration of the
pollutant and a fixed coefficient). Because waterbody
parameters such as advection and mixing are not part of the
analyses, the equations are usually not suitable for assessing
water quality criteria. However, once travel time is known,
transformation equations can be used to obtain concentration.
In addition, they can be used as screening tools for
persistence of pollutants. Lyman et al., (1982), Callahan et
al. (1979), and Mills et al. (1982) can provide additional
details.
5.3.4 Sediment-Water Interactions
Hydrophobic (low solubility) pollutants are subjected to
different transport and fate mechanisms than are hydrophilic
(highly soluble) ones. They may be more dense and/or sorb
strongly to sediments. A dissolved phase may exist and can
present an environmeatal hazard, although it is usually small
compared to the sorbed phase. A series of specific analyses
may be performed to determine peak concentrations,
concentration profiles, and plume extent of these pollutants.
Before downstream transport can be calculated, a vertical
distribution of suspended material analysis must be
determined. It is particularly useful when the pollutant's
partition coefficient is high (the dissolved phase is small or
neglible). Required data and parameters include: settling
velocity of the sediments or particulate phase, pollutant
density, hydraulic radius of the reach, slope, and shear
velocity (related to flow velocity and bottom roughness).
The desorption from sediment bed analysis is used to calculate
contaminant concentrations in the water column. A high
percentage of the dense or sorbed pollutant can be deposited
on the sediment bed. If the pollutant is not very susceptible
to degradation, it may slowly desorb back into the water
column over a long time period. This desorption process can
4-45
-------
be calculated bas^ on an initial concentration in the
sediment bed, dissolved concentration in the water column, and
desorption rate coefficients. The water column concentration
is derived using the stream velocity, mass of the pollutant
per unit area of bed, equivalent depth of water in the
sediment, and a partition coefficient (Mills et al., 1982).
The analysis for a spill of low solubility/high density
pollutants provides a means to calculate the water column
concentration (dissolved and particulate) that is subject to
downstream advection. Primary to the calculation is the
diffusion coefficient and thickness of the diffusive sublayer
over the bed. Depth and stream velocity will affect this
thickness. Before this analysis is used, however, the
dimensions of the contaminated zone must be known or
calculated (using a mixing zone analysis), as well as the
solubility limit of the pollutant in water. Refer to Raj and
O'Farrel (1977) or Mills et al. (1982) for more information.
5.4 ANALYTICAL MODELS
5.4.1 Overview
Analytical models are presented as an intermediate technique
in terms of level or complexity, between simplified assessment
techniques and numerical models. The difference between
simplified assessment techniques and analytical models is
often small: analytical models often employ the same
equations as simplified techniques but require computers
because of the number of calculations to be solved, as for a
multi-reach stream. They can be applied to more complex
problems where some variation in properties occur.
These models require steady-state flow conditions and uniform
geometry. They have limited applicability to remedial
actions, given the unsteady flow regimes, non-uniform
geometry, and complex sediment-water interactions that
characterize environmental conditions when remedial actions
are implemented. They are used for time of travel, peak
concentration, and concentration profile determinations.
Within this group of models, differences can include:
complexity of geometry allowed, mode of pollutant loading
(instantaneous or continuous), degree of mixing and dispersion
(if any), ability to calculate transfer of mass between the
sediment bed and the water column, method of estimating
sediment transport (user input suspended sediment
concentrations, or concentrations calculated for each reach
separately), lumped or specific first order decay reactions,
and the range of default values available for model
4-46
-------
parameters.
Table 5.3 is a matrix comparing selected analytical models
with respect to model capabilities and required data. The
model group is not meant to be comprehensive; rather, it
represents a cross-section of available analytical models and
is designed for comparison purposes only. Descriptions of
models follow.
5.4.2 Selected Analytical Models
STTUBE and TUBE (Codell et al., 1982) are steady-state,
conservative river models which are used in conjunction. Both
use simple geometries (representing the river as a rectangular
channel) and constant coefficients to analytically solve a
standard dispersion equation. TUBE generates dispersion
coefficients and the velocity field for STTUBE, which then
simulates dilution and travel times. Computations are
performed for stream-tube coordinates, in which the
cross-sectional areas are mapped onto a new river discharge
based coordinate system, thus simplifying the mathematical
representation. STTUBE simulates a steady release of
pollutants, and is restricted to portions of the river removed
from the influences of discharge (far field). These models do
not simulate sorption or transformation processes.
RIVLAK (Codell et al., 1982) computes concentrations in a
river or near shore region of a large lake from a non-steady
source. RIVLAK requires uniform geometry, steady flow, pulse
input, constant dispersion coefficients, and release of the
contaminant from a vertical line source. This type of
analysis provided is more applicable to near field
concentrations, and is useful for mixing zone criteria as well
as for determining peak concentration. STTUBE, TUBE, and
RIVLAK program listings, as well as user manuals, are provided
in Codell et al. (1982).
HAGS, or Hazard Assessment Computer System (Raj and O'Farrel,
1981) contains eight analytical models designed for water
quality assessment as well as explosion and flammability
hazards and toxic cloud formation assessment. Four of the
models are suitable for distribution of water borne
pollutants. These models all: assume instantaneous release
of pollutants; are unable to simulate dispersion, degradation,
or sorption processes; and provide peak concentration results.
The primary differences between the models are based on
chemical characteristics such as density and solubility. The
four water quality assessment models are subsequently
described.
4-47
-------
TABLE 5.3 SELECTED ANALYTICAL MODELS VS. MODEL CAPABILITIES AND REQUIRED
DATA/FACTORS
I
4^
CD
7
SELECTED AHALmCIL MODELS
3TTUBE
TUBE
RIYLAt
HAZARD ASSESSMENT
COMPUTER 3>3-i"J1S;
Mixing A Dilution or
Soluble Pollutant*
Spreading of Light
Pollutants
Dispersion A Dlasolutlon
of Pollutants Vlth
rinlta Solubility
Spreading A Sinking of
Insoluble Pollutants
HITCHARD BOX MODEL
Codell et. al., I9B2
Codell et. al., 19(2
Codell et. al., 1982
Horrow et. al., 1981
Raj and O'Parrell, 1981
Raj and O'Farrell, 1981
Raj and O'Farrell, 1981
Hills et. al., 1982
-------
The mixing and dilution of the soluble pollutants model
(Morrow et al., 1981) simulates instantaneous and continuous
releases of hazardous chemicals into navigable non-tidal
rivers. Very near field, near field, and far field
computations are performed. The near field analyses are based
on buoyancy, momentum of discharge, and turbulence; far field
analyses predict steady-state concentrations as a function of
distance downstream. Volatilization is the only
transformation process simulated.
The spreading of light pollutants model (Raj and O1Parrel,
1977) examines the dispersion of low density (specific gravity
less than one), low Solubility pollutants on the surface of a
waterbody. The pollutant is mixed based on eddy diffusivity
so river turbulence parameters are required for simulation.
The dissolution and dispersion of pollutants with finite
solubility model (Raj and O'Farrel, 1977) was designed to
simulate pollutants that are soluble in low concentrations.
The dissolution rate is predicted using solubility and
dilution parameters, then dispersion is predicted for
uni-directional flow.
The spreading and sinking of insoluble pollutants model (Raj
and O'Farrel, 1981) is based on two stages: gravity-inertia
and hydrodynamics. The model predicts the shape of the pool
and duration of pool spreading. Its use is limited to
turbulent rivers. No slope effects, complex geometry, or
long-term bed/water interactions are included. HAGS is
operational on the Cybernet System of Control Data
Corporation. Authorization and access procedures for the
system are provided by the National Response Center of the
U.S. Coast Guard in Washington, DC.
The Pritchard Box Model (Pritchard, 1969) is a steady-state,
conservative, 2D (x-z) estuary model. It is designed for
stratified estuaries and is sensitive to longitudinal salinity
profile inputs. If the estuary is uniform and has little
variation in salinity along its axis, it may be divided into
two segments, whereupon a hand calculator can suffice for
performing the analyses. The model should be implemented on a
computer if more than five segments are defined. The model
accepts only continuous pollutant release, preferably from the
head of the estuary.
4-49
-------
SECTION 6
USE OF NUMERICAL MODELS FOR REMEDIAL ACTION ASSESSMENT
6.1 OVERVIEW
Numerical models provide the investigator with the ability to
represent chemical transport in complex water bodies where
multi-dimensional flow, stratification, tidal variations
and/or complex boundary conditions are important. Although
such models involve substantially greater resources, their use
may be justified where the effects of candidate remedial
actions cannot be adequately represented by simplified
methods. This section introduces a number of potentially
useful models, discusses their capabilities, and provides a
framework for their application.
Numerical models, in contrast to conceptual models
(physically-based equations representing key processes) and
analytic models (simplified process equations solved exactly
using direct mathematical manipulation), approximate the
process equations using finite difference or finite element
techniques and separate the site into discrete segments. In
this way, the full process equations can be solved with a
minimum of restrictive assumptions. The solution, however,
will not be exact. Consequently, a trade-off must be made
between 1) ease of solution, computational accuracy,
simplicity and limited applicability for analytical models and
2) greater resolution, more general applicability, increased
complexity and increased solution costs for numerical models.
Key attributes of numerical models can be summarized as
follows.
1. Few simplifying assumptions are required, although the
simplicity and computational efficiency of the
solution algorithm depend, in part, on assumptions
made.
2. Values of key quantities (e.g., velocity and chemical
concentration) are computed at discrete space and time
intervals selected by the user. These intervals
(i.e., model resolution) can be adjusted to achieve
4-50
-------
the accuracy and specificity required by the site and
problem being addressed.
3. Numerical solutions to the governing equations are
approximate and subject to computational errors due to
truncation, roundoff and numerical dispersion. Choice
of solution scheme can have a substantial effect on
these errors.
4. Resources required to implement numerical models
depend on the dimensionality, resolution, number of
independent variables being predicted, and solution
scheme. Required resources include: user expertise
in developing and applying such models, field data,
data on chemical sources, sinks and reaction rates,
personnel time, and (typically) substantial computer
facilities. It is reasonable to expect that needed
resources will be two to ten times those required for
analytic model applications.
5. Multiple independent variables (e.g., velocity,
temperature, chemical concentration, etc.), can be
simulated simultaneously, including interactions
between these variables.
A number of authors provide overviews of numerical models and
their use in problems related to surface water bodies.
Donigian (1981) reviews runoff and instream contaminant
transport and fate models, Onishi, et al., (1981) review
sediment transport and water quality models, and Orlob (1971)
discusses estuary models. Other current model reviews include
Basta and Bower (1982) and EPA (1983). Additional information
on surface water models can be obtained from the Center for
Water Quality Modeling, EPA Environmental Research Laboratory,
Athens, GA.
6.2 CAPABILITIES OF AVAILABLE CODES
The development of numerical models for surface water
hydrodynamics and chemical transport has been ongoing since
the early 1960's. Consequently, a large number of codes
providing various degrees of sophistication are available.
Some 35 codes were screened for possible use in remedial
action assessment. Eleven codes were selected for further
evaluation and inclusion in this discussion based on recent
applications to toxic pollutant transport and fate studies, or
ability to represent complex flow and mixing processes. These
models serve as examples of codes which are potentially useful
in remedial action assessment and a starting point for
evaluations of suitable codes. Other codes (both existing and
4-51
-------
under development) may be of similar use.
Numerical codes can be differentiated by several aspects of
their capabilities: type of water body that can be simulated,
spatial domain (dimensionality), temporal domain (steady state
versus dynamic time frame), and ability to represent chemical
fate. A code is typically written for a certain type of water
body (river, lake, or estuary); this target water body often
defines the dimensionality and time frame of the code. In
many cases a code written for estuaries can also be used for
lakes or rivers or a code written for lakes can be used for
rivers because all of the required elements of the less
complex water body may be contained in the code. A
disadvantage of using a complex code on a simple water body is
the need to input parameters and data which may be extraneous
to the problem and the added computer costs associated with a
more sophisticated model.
Unlike ground-water models, which tend to use separate codes
for flow modeling and chemical transport modeling, surface
water models typically solve both flow and transport equations
at the same time. There are two primary reasons for this: 1)
there is usually limited interest in water movement without
transport of heat, salinity, or chemicals and 2) the movement
of heat, salinity and some chemicals affects hydrodynamics and
so cannot be separated from the computation of flows. Most of
the models discussed here are combined flow and transport
models.
The majority of surface water codes provide dynamic (time
varying) simulation of flow and transport. Dynamic
simulations allow variations in chemical loadings due to
changes in meteorology and discharge rates and in water flow
rates due to the effects of tides, reservoir operation and
streamflow.
Certain estuary models use tidally-averaged flow conditions to
eliminate the effect of tides and reduce model complexity and
run costs. Such an approach can produce meaningful results
when the effects of flow reversals, movement of salinity
gradients, and tidally-induced mixing can be ignored or
approximated by steady-state parameters. Similarly,
conditions in rivers and lakes which are steady over time
(i.e., no significant variations in flows, temperature, or
chemical inputs) can be simulated by steady-state models.
Such models provide results similar to those obtained from
analytic models.
Table 6.1 is a matrix of selected codes vs. environmental
processes and waterbody conditions. Models chosen for
detailed evaluation include: DEM: Dynamic Estuary Model
(Ambrose and Roesch, 1982); FETRA: Finite Element Transport
4-52
-------
TABLE 6.1 PROCESSES VS. MODELS MATRIX
I
Ul
OJ
PROCESSES
SELECTED
NUMERICAL
MODELS
TOXIWASP
HSPF
TODAM
EXAMS
SERATRA
DEM
FLESCOT
FETRA
LEENDERTSE 2D
LEENDERTSE 3D
LARM
WATFRnnnv SPATIAL nt«;ppp<;TnN TEMPORAL
WATERBODY DQMAIN DISPERSION DOMAJfJ
SEDIMENTATION
TRANSFORMATION
PROCESSES
c c c
B
FOOTNOTES:
1. FETRA DOES EMPLOY
FIRST-ORDER DECAY FOR A
POLLUTANT, HOWEVER, THIS
LUMPED PARAMETER MAY BE
TOO SIMPLE FOR POLLUTANTS
WITH MULTIPLE DEGRADATION
PROCESSES.
LEGEND
S = SINGLE ITEM
B = BRANCHING OR NETWORK
L = LATERALLY INTEGRATED
V = VERTICALLY INTEGRATED
C = COMPARTMENTS
-------
Model (Onishi et al., 1979); TODAM: Transient One Dimensional
Degradation and Migration Model (Onishi et al., 1982);
SERATRA: Sediraent-Radionuclide Transport Model (Onishi and
Wise, 1979); FLESCOT: Flow Energy Salinity Sediment Transport
Model (Onishi and Trent, 1982); HSPF Hydrologic Simulation
Program - Fortran (Johanson et al., 1981); TOXIWASP: Water
Quality Analysis Program (Ambrose et al., 1983 and Ditoro et
al., 1982); EXAMS: Exposure Analysis Modeling System (Burns
et al., 1982); Leenderste two-dimensional and
three-dimensional circulation models (Liu and Leenderste,
1972); and LARM: Laterally-Averaged Reservoir Model (Edinger
and BuchaTc, 1982) .
The codes represented in Table 6.1 are divided into three
groups based on the types of processes represented. The first
group (TOXIWASP, HSPF, etc.) model water flow, chemical
advection, sedimentation processes, and chemical trans-
formation. The second group (DEM, FLESCOT, and FETRA)
represent all other processes. The third group provides only
hydrodynamic modeling, with some capability to advect and
degrade single pollutants. These three groups also differ in
the sophistication of their hydrodynamic computations: the
first group uses compartmental or simple branching 1-D models
(except for SERATRA), the second group provides a wide range
of hydrodynamic solution techniques, and the third group
provides relatively sophisticated two-dimensional and
three-dimensional hydrodynamic codes. The model user must, in
most cases, make a trade-off between detailed representation
of chemical transport and transformation and representation of
complex flows.
The parameters on the top axis of Table 6.1 may require
further explanation. Spatial domain refers to the number of
dimensions (1, 2, or 3) that the model may simulate. The
two-dimensional models are further described as either lateral
("y" direction) or vertical ("z" direction) along with the
normal longitudinal ("x") direction. Dispersion may be
simulated by turbulence calculated within the program
(generated by velocity differences or shear within the flow
field); or it may be simulated empirically via user input
dispersion coefficients. Temporal domain refers to the
model's ability to simulate steady, continuous events or
unsteady, pulse events. Steady-state refers to continuous
waste input and flows over the duration of the simulated time
period. Tidally averaged is also steady state but refers to
simulating steady estuary hydrodynamics for each tidal period.
Quasi-dynamic refers to the model's ability to simulate some
variables in a steady-state and others dynamically in the same
simulation. A dynamic simulation means that flows and waste
loading may vary for each time step within a simulation.
4-54
-------
Sedimentation refers to the whole range of sediment-water
interactions (sediment transport, deposition, and erosion)
that may occur. Sediment transport and suspended sediment
sorption were described in Section 2. Direct bed exchange
encompasses diffusion, scouring, deposition, and resuspension
of contaminated material between the sediment bed and the
water column. Armoring refers to the sorting of bed sediments
during flows such that the bed surface is more resistant to
scour than the underlying material. This situation may affect
contaminant concentrations in the water and in the bed through
modification of exchange rates.
Transformation processes have been described in Section 2.
The lumped decay refers to a simple (usually first-order)
reaction that accounts for the pollutants' aquatic fate. For
some complex pollutants however, this degradation model
formulation is an over-simplification and may not provide an
accurate picture. "Daughter Products" refers to the model's
ability to track the pollutant after it has degraded to
another form. This "new" pollutant may be susceptible to the
same physical, biological, and chemical processes as its
parent. An example of this process is the degradation of DDT.
The metabolites (or products) of chemical/biological
degradation are ODD or DDE. Both of these compounds are more
toxic than DDT, and warrant examination of transport and fate.
Table 6.2 is a matrix of the type of simulations needed for
remedial actions and specific waterbodies. The waterbodies
are grouped as estuary, lake, or river, with subgrouping
within each according to system geometry and degree of mixing.
Numbers and letters in the matrix denote the type of
simulation needed for that remedial action in the specific
waterbody. For example, "2L" denotes that a two-dimensional
(lateral-longitudinal) simulation is required for that
remedial action/waterbody scenario. A "0" indicates that the
remedial action is not suited for use under the specific
waterbody conditions. The simpler remedial actions such as
dilution and the use of barriers or diversions, often may be
simulated by adjusting the boundary conditions and system
geometry. Most of the remedial actions require a
two-dimensional (longitudinal-lateral) simulation. However,
as the mixing becomes more turbulent or complex (as in
estuaries and large lakes), a two-dimensional
(longitudinal-vertical) simulation with coefficients for the
horizontal or lateral (third) dimension, or a full
three-dimensional simulation may be required.
The remedial actions vs. models matrix (Table 6.3) is a
culmination of the previous two matrices. The critical
processes of transport and fate of each remedial action are
matched against model capabilities. As the matrix is reviewed
the reader should refer to the previous matrices and the
4-55
-------
TABLE 6.2 REMEDIAL ACTIONS VS. WATERBODY MATRIX
REMEDIAL
ACTIONS
RIVERS
NO ACTION
REMOVAL
MECHANICAL
DREDGING
EXCAVATION
HYDRAULIC
DREDGING
BARRIERS/
DIVERSIONS
SKIMMING
DILUTION
CONTAINMENT
COFFERDAMS
BOOMS
SILT
CURTAINS
CAPPING
TREATMENT
IN-SITU
ON-SITE
21
0
1
2L
3
2V
2V
2V
2V
0
2V
3
3
2V
2L
0
* I
0.
0
2L
2L
0
2L
2L
0
0
0
EPENDA
2V
3
3
0
0
2L
3
2V
0
3
3
2V
0
^T ON R
0
MOVAL
2L
21
21
21
2V
0
2L
2V
0
0
:TION
2P
2L
2L
2V
2V
3
3
2V
2L
0
JSED IN
0
2L
2L
0
IB
2L
0
0
2L
CONJUN
IB
IB
2L
0
1
2L
0
2V
2L
2L
TION
2L
2L
2L
2L
0
IB
2L
0
0
2L
LEGEND:
1 = 1-DIMENSIONAL
2 = 2-DIMENSIONAL
3 = 3-DIMENSIONAL
L = LATERALLY AVERAGED
V = VERTICALLY AVERAGED
0 = ACTION IS NOT APPLICABLE
TO THIS WATERBODY
B = BRANCHING OR NETWORK
4-56
-------
TABLE 6.3 REMEDIAL ACTIONS VS. MODEL MATRIX
FOOTNOTES:
1. FETRA DOES EMPLOY FIRST-ORDER DECAY FOR A POLLUTANT,
HOWEVER, THIS LUMPED PARAMETER MAY BE TOO SIMPLE FOR
POLLUTANTS WITH MULTIPLE DEGRADATION PROCESSES.
LEGEND:
A = REPRESENTS ALL IMPORTANT PROCESSES (P,S,
AND C).
P - REPRESENTS THE THE CRITICAL PHYSICAL PROCESSES.
C - REPRESENTS THE CRITICAL CHEMICAL/BIO-DEGRADATION
PROCESSES.
S - REPRESENTS THE CRITICAL SEDIMENT-WATER INTERACTION.
B c BRANCHING OR NETWORK
L - LATERAL
V « VERTICAL
ID, 2D, 3D - REFERS TO DIMENSIONS REQUIRED
4-57
-------
remedial action description (Section 3) for reference. Model
evaluation criteria are based upon the environmental processes
affected by the remedial action and the dimensionality needed
to represent these processes. For simplicity and ease of
use, these criteria have been stated as questions and
organized into the following groups.
Physical processes (denoted by "P"):
o Can the model simulate inflow and outflows?
o Is the dimensionality sufficient to represent a change
in the system boundaries and geometry due to the
remedial action?
o Can dispersion be adequately simulated using empirical
coefficients, or should it be calculated within the
model equations to account for the effect of new
barriers, or inflow/outflow?
o Can the removal of the waste source such as a
contaminated sediment bed be simulated (i.e.: Are
there adequate source/sink terms?)
Sediment/Water Interactions (denoted by "S"):
o Can partitioning between sorbed/desorbed phases of the
pollutant be simulated?
o If the pollutant is sorbed or in particulate form, can
sediment transport be simulated?
o Can the model simulate bed-water transfers, such as
scouring, deposition, and diffusion over time?
Chemical/Biological Degradation (denoted by "C"):
o Does the model simulate the important degradation
processes?
The above groups will be represented by the letters P, S, and
C, respectively. Appearance of a letter under a model
corresponding to a specific remedial action indicates that the
model can represent the critical environmental processes
affected by the remedial action (or answering "yes" to the
hypothetical questions posed within the specific group above).
Important factors or groups are listed beside each remedial
action for easy reference.
4-58
-------
6.3 THE MODEL DEVELOPMENT AND APPLICATION PROCESS
The process of setting up a computer code so that it will
simulate water and waste constituent movement at a specific
site is called the "model development" process. It involves
combining one's understanding of how a code represents
individual processes with one's understanding of their actual
occurrence in the field. The latter is based on available
site data, information and past experience. Model application
is the use of a developed and tested model to analyze target
situations, in this case the performance of a potential
remedial action. While a numerical code is often quite
general, a developed model is specific to a particular site
and, when applied, to a particular condition at that site.
Figure 6.1, taken from Mercer and Faust (1981), represents one
process for model development and application. Once the need
for numerical modeling has been determined and appropriate
models selected for each affected zone, the following steps
may be taken.
1. The conceptual model of the site used to select model
codes is further defined and quantified through the
collection and analysis of site data. This conceptual
model may also include approximate effects of
potentially feasible remedial actions.
2. The conceptual model is then used to define the model
structure required for the water body of concern, the
types of outputs needed, and the required spatial and
temporal resolution of model simulations.
3. The individual code is installed on an appropriate
computer and the site model implemented by creating an
appropriate model structure (i.e., element or grid
size and orientations, boundary conditions, and sink
and source node loctions).
4. Values for individual model parameters are estimated
from field data and then verified by comparing model
predictions with available site data (i.e., history
matching). The model can then be calibrated on a
different set of site data to identify the ranges of
values for critical parameters. This process is based
on the assumption that the ability to obtain complete
sets of data is not limited by time or money
constraints. If the data is incomplete, best
engineering judgement of information from field sites
with similar characteristics should be applied.
5. Adjustments to model parameters and localized model
structure can then be made to represent the effects of
4-59
-------
DETERMINE NECESSITY
OF NUMERICAL MODEL
COMPILE & INTERPRET
AVAILABLE DATA
COLLECT DATA AND
OBSERVE SYSTEM
Conceptualization
History Matching
(Field Problem)
PREPARE DATA
FOR MODEL
USING ESTIMATED
PARAMETERS
i
PREPARE DATA
FOR MODEL
USING ESTIMATED
PARAMETERS
Improve
Conceptual
Model
INTERPRET
RESULTS
COMPARE RESULTS
WITH OBSERVED
DATA
Results
Satisfactory
Good
Comparison
Poor
Comparison
SENSITIVITY RUNS
ARE MORE DATA
NEEDED?
Yes
No
PREDICTIVE
SIMULATION RUNS
Fiqure 6.1 Model development and application process
(from Mercer and Faust, 1981). Copyrighted
by National Water Well Association.
4-60
-------
alternative remedial actions on water and constituent
movement. Procedures or adjusting model parameters to
represent specific remedial action alternatives are
discussed in detail in Section 5.
6. The verified and adjusted model can now be run to
predict future conditions with and without remedial
actions. Various combinations of actions can be
explored. Where data uncertainties exist, sensitivity
analyses can be used to estimate the range of
outcomes.
Numerical models are potential tools for answering several
important questions raised by the feasibility study process
for evaluating remedial action alternatives, including:
1. existing exposure routes and levels of exposure for
specific chemicals
2. future exposures if no action is taken
3. effects of alternative remedial actions on conditions
at and near the site
4. future exposures during and after the implementation
of alternative remedial actions
Most of these questions will need to be answered during the
screening and analysis of alternatives. While screening will
require, at most, analytical models, numerical models may find
use in the alternatives analysis where complex site conditions
exist or complex remedial actions are anticipated.
Site characterization involves data collection and evaluation
efforts, (including the potential use of numerical models)
required to specify chemical sources, chemical migration
pathways, chemical fate, potential receptors, and human health
and environmental effects. These efforts will be accomplished
during the site investigation and analysis steps of the
Remedial Investigation/Feasibility Study process.
4-61
-------
SECTION 7
MODEL REQUIREMENTS FOR SURFACE WATER REMEDIAL ACTIONS
7.1 OVERVIEW
This section provides modeling requirements for surface water
remedial actions. Such requirements may apply to either
analytical models (Level I) or numerical models (Level II).
Model requirements refer to the type of model required
(analytic or numerical) and any unique capabilities such as
sediment transport; the model dimensionality and grid
configuration? and parameter adjustments. For each remedial
action, guidance is provided for the model adjustments
required to simulate the environmental effects of that action.
Most of these model adjustments involve parameters. As such,
model parameter estimation guidance is also provided here to
assist the user in deriving appropriate values for critical
parameters. The model parameters that must be adjusted to
simulate the effects of implementing different actions can
vary. As Volume 1 notes, modeling requirements for all
potential remedial actions must be considered early enough in
the Feasibility Study/Remedial Investigation process to have
an impact on the specific model(s) selected for use in
remedial action evaluation. The remedial actions described in
Section 3 were condensed into eight groups, according to their
design objectives and conjunctive use with other actions in
the same group. The actions are listed in Table 7.1.
Each remedial action scenario produces unique effects in the
waterbody. Modeling requirements will be dictated by the
spill/discharge mode, the degree of initial mixing or dilution
and, to a lesser extent, the migration of contaminants through
the waterbody to an exposure site. The processes governing
contaminant transport and fate are different between the spill
site and the exposure site. Spill site processes of
importance include rate, duration, and type (i.e., point,
nonpoint, pulse, continuous) of contaminant discharge,
momentum and buoyancy of the contaminant flow, in-stream
velocity distribution, and turbulent mixing. These processes
are commonly incorporated in what is termed near-field models.
Exposure or far-field models incorporate advection as the
4-62
-------
TABLE 7.1 GROUPING OF REMEDIAL ACTIONS ACCORDING
TO MODEL REQUIREMENTS
Dilution
Containment Measures
Booms and partial barriers
Cofferdams and full barriers
Silt curtains
Capping
Removal Measures
Hydraulic and mechanical dredging
Excavation
Treatment
4-63
-------
primary process of contaminant transport, with degradation and
sediment transport also being important. Therefore, model
selection and application will be different for analysis of
spill site processes and remedial actions than for evaluation
of chemical transport and fate away from the source.
Unlike soil and ground-water remedial actions, surface water
remedial actions have limited influence on the contaminant
migration path, except in cases where flow is disrupted, as
with barrier implementation. Consequently, remedial action
modeling can often be confined to the immediate vicinity of
the source. For some of the removal-type remedial actions a
change in the source term for a far-field or exposure model is
sufficient for representation. In these cases, selection of
an appropriate model will be based on the complexity of the
receiving waterbody; the remedial action should not affect
model selection criteria.
The key questions for remedial action simulation are, then:
must the source (remedial action site) be modeled, and once it
is adequately represented, should remedial action impacts be
input to a separate far-field model that will predict pre- and
post-restorative concentration levels at an exposure site? As
mentioned above, the source term can be empirically derived
for some removal-type remedial actions, and can be input into
an exposure model. If the remedial action's effects in the
near-field cannot be simplified, the source and spill area
must be modeled. The focus of this section, then, will be
source modeling or representation, and will include near-field
modeling requirements for the remedial actions that affect
initial mixing and dilution processes. Far-field or migration
modeling will not be discussed except where the remedial
action affects advection. Modeling needs dictated by the
waterbody characteristics will be addressed only as they
affect remedial action modeling.
The following subsections detail modeling requirements for
specific measures and provide parameter estimation guidance
for those parameters that may be adjusted in order to
represent the environmental effects of a remedial action.
Prior to presenting modeling requirements for each group of
measures, several key points need to be addressed.
1. Only those modeling requirements associated with a
given group of remedial measures are discussed.
Requirements associated with the use of numerical
models for site characterization and assessment are
not presented. Thus, the guidance presented herein is
in addition to that needed to develop a model of the
site.
2. Certain model parameter adjustments are highly site-
4-64
-------
specific. Thus, it is difficult to provide guidance
on parameter estimation.
3. Data on certain model parameters are, on the whole,
quite sparse due to a lack of field data on the
performance of most remedial measures. In many cases,
only laboratory or pilot scale data or parameter
values from previous modeling studies are available.
7.2 MODELING REQUIREMENTS
The modeling requirement for each group of measures are
presented in terms of the following:
1. Model Type - Model type refers to the level of
sophistication required in the selected model.
Choices include numerical (or Level II, as referred to
in Volume 1), analytical (or Level I), or empirical
computation.
2. Dimensionality and Grid Configuration - Dimensionality
refers to the directions (i.e., x, y, and z) of water
and chemical movement that can be simulated, while
grid configuration refers to the spatial
discretization used to represent a site and the
remedial action.
3. Parameter Adjustments - Parameter adjustments refer to
the model inputs that must be modified to represent a
remedial measure.
Table 7.2 summarizes the modeling requirements for each
measure. The following discussion provides more detailed
guidance.
7.2.1 Dilution
Dilution is the most simple remedial action to simulate, as
well as being relatively simple to implement in the field.
This type of measure was used to reduce aqueous concentrations
of formalin in the Russian River (Ca.) following a spill in
1982.
Model Type
A source model is not required because changes in source
concentrations and flow rates can be estimated directly. An
exposure or far-field model can be used by applying new source
4-65
-------
TABLE 7.2 MODELING REQUIREMENTS FOR REMEDIAL ACTIONS
Dimensionality/ Parameter
Remedial Actions Model Type Grid Configuration Adjustment
Dilution WB WB S/S
Containment
Measures:
Booms and partial 2P 2D(x-z) BC
barriers
Cofferdams and WB WB S/S
full barriers
Silt curtains SD 2D(x-z) BC
Capping WB 2D(x) S/S, SB
Removal Measures:
Hydraulic and SD 2D(x-z) S/S, SB
mechanical
dredging
Excavation WB WB S/S
Treatment WB WB S/S
LEGEND: BC = Boundary conditions
SB = Sediment bed parameters
SD = Sediment transport
S/S = Source on sink terms
WB = Requirements are a function of the waterbody
characteristics
ID(x) = One-dimensional, longitudinal direction
2D(x-z)= Two-dimensional, longitudinal and vertical
directions
2P = Two-phase flow
NOTE: Grid configurations are generic in nature and are
described in text.
4-66
-------
concentration and flow rates.
Dimensionality and Grid Configuration
The required model dimensionality and grid configuration will
be a function of the velocity distributions and geometry of
the waterbody. However, the exposure model must be able to
represent a new input distribution for the source.
Parameter Adjustment
The parameters that can be adjusted to represent dilution are
the input concentration (or mass) or the source flow rate.
The concentration can be reduced and be input for a longer
period of time, thus assuring that no change in input mass is
realized. If the contaminant is released into the receiving
water via a waste stream, as with a point source, the source
flow rate can be increased in order to dilute the incoming
plume.
7.2.2 Containment: Booms and Partial Barriers
This group of measures is directed at controlling the
spreading of light, immiscible contaminants on top of the
water column. The use of skimmers is also included in this
group, as skimming and booms are usually used in conjunction.
A number of references provide information on deployment and
configurations, such as Department of Transportation (1978);
Petty et al., (1982); Fussell et al., (1981); and Huibregste
et al., (1977).
Model Type
The primary model requirement is the ability to simulate 2
phase flow (e.g., water and a floating, immiscible liquid).
Only a very few numerical models have this capability. The
selected model should incorporate turbulent mixing and shear
between the two liquid layers, in order to adequately
represent dissolution and mixing of soluble components of the
pool into the water column. However, the control of pool
spreading along with a specified dissolution rate may also be
modeled on a gross level using empirical entrainment and
containment calculations. One such model for pool spreading
is in HACS (Hazard Assessment Computer System), a set of
analytical models intended for use in rapid response
situations of chemical spills. Raj and O'Farrel (1981)
provide details on this model as well as the other water
quality models used in HACS.
4-67
-------
Dimensionality and Grid Configuration
The recommended model dimensionality for most waterbodies is
two-dimensional in the longitudinal and vertical planes. The
two-dimensional (x-z) configuration allows resolution of the
water column, which is important in controlling the spread of
a selected layer or depth of water. A one-dimensional (x)
model is sufficient for narrow rivers, where the pool spreads
across the water surface laterally to both side boundaries.
However, this dimensionality and configuration (well-mixed
reaches) do not allow for the selective containment of the
surface slick. Herbes et al., (1982) present such a model for
transport of coal liquefaction product spills in rivers. A
two-dimensional (x-z) configuration will allow variable grid
spacing along the vertical plane. Most grid points should be
specified around the boom or barrier in order to allow better
resolution of the containment, and to reduce any numerical
instability caused by a no flow boundary and high shear stress
between the two liquid layers.
Parameter Adjustment
The physical barrier or booms are represented by a no flow
boundary within the grid. The removal of material via
skimming must be approximated by reducing the solubility of
the contaminant in water. This will serve to reduce or
eliminate aqueous concentrations downstream, which is the
purpose of containment and skimming.
7.2.3 Containment: Cofferdams and Full Barriers
This group of measures has only minor modeling requirements
because the source is assumed to be completely isolated from
the waterbody, including the period of implementation of the
remedial action.
Model Type
The model required, if any, will be dependent on the waterbody
characteristics. If an exposure model is required because of
waterbody complexity, the actions are represented by a change
in the source term and possibly the boundary shape of the new
shoreline.
Dimensionality and Grid Configuration
Dimensionality and grid configuration will be a function of
the waterbody characteristics.
4-68
-------
Parameter Adjustment
The only parameter that needs to be adjusted is the source
term, which is reduced according to the assumed efficiency of
the barrier in isolating the waste, and the degree of
dewatering of the spill site area.
7.2.4 Containment: Silt Curtains
Silt curtains are designed to reduce suspended sediments in
the near-field water column resulting from dredging,
excavation, and non-point sources. They usually force the
turbid water to a lower elevation with minimal deposition and
the suspended sediments resurface further down-stream. Since
silt curtains are often used in conjunction with mechanical
dredging, the user should refer to those modeling requirements
as well when evaluating these actions.
Model Type
A numerical model with sediment and contaminant transport
capabilities is required for simulation. The model should
incorporate turbulent mixing and shear, and sediment scour and
deposition processes.
Dimensionality and Grid Configuration
A minimum of 2 dimensions (longitudinal-vertical) is required
to simulate the vertical distribution of sediments and allow
better resolution of the trapping effect of the silt curtain.
If the area to be contained is irregularly shaped, a 3D
simulation may be required. It is important to compute the
velocity distribution in the water column in order to simulate
sediment scour and deposition accurately. The grid
configuration along the vertical plane should reflect more
points around the curtain and between the curtain and the
bottom. This is done to represent the turbulent mixing and
shear, and associated sediment deposition and transport in
these locations in the water column. If a three-dimensional
model is used, the boundaries should be set away from the
curtains in order to mitigate any influences the artificial
boundaries may have on the flow field.
Parameter Adjustment
No parameters need to be adjusted, as the curtain is simulated
by no-flow grid points in the model. In this sense, water is
not allowed to flow through the curtain as it would in the
waterbody. Because the curtain impedes the flow of water and
causes more turbulence and increased velocities around curtain
4-69
-------
edges, mixing-related parameters (dispersion coefficients) may
also need adjustment.
7.2.5 Containment; Capping
The purpose of capping is to prevent desorption of
contaminants and erosion of contaminanted sediments from the
sediment bed. This type of action is limited in use because
of the difficulty in locating and covering the total
contaminated sediment beds.
Model Type
The model required will be dependent upon the waterbody
geometry and flow field complexity. The action is represented
by reducing the source term in an exposure model.
Dimensionality and Grid Configuration
A minimum of one-dimension (longitudinal) with a boundary
layer profile calculation for sediment entrainment is
required. Two-dimensional (x-z) models with sediment
transport may provide better resolution of desorbed
contaminant concentrations and sediment entrainment in the
water column immediately above the cap.
Parameter Adjustment
The simplest method of simulating the cap is to reduce the
contaminant mass per unit area of bed or the concentration in
the sediment bed. The degree of reduction will depend on the
percentage of contaminated bed that is assumed to be isolated
in each segment or reach. This method was employed by Onishi
(1979) when he simulated the effects of dredging (or removing)
the Kepone-contaminated bed along a portion of the James River
(see Section 7.2.6). The caps can also be simulated with more
detail if the user wishes to examine potential erosion of the
cap, exposure of the contaminated bed, or diffusion of the
contaminant through the cap into the water column. Parameter
adjustment could include: assign the contaminant
concentration to deep burial in the lower portion of the bed;
modify the sediment bed characteristics, such as bed shear
strength, particle size, diameter, and density, to reflect the
cap material (probably clay); and decrease the resuspension
velocity and/or increase the settling velocity of the sediment
particles. If a two-dimensional simulation is used, the
bottom profile can also be adjusted to represent potentially
increased velocities around the raised, capped areas. In this
case, the depth of the cap should also be specified.
4-70
-------
7.2.6 Removal; Hydraulic and Mechanical Dredging
Hydraulic and mechanical dredging constitute the most commonly
employed remedial actions for restoration of contaminated
surface waters. For some dredging scenarios, two sets of
modeling requirements must be applied: one for the dredging
period in order to examine potential adverse effects; and one
for post-restoration, in order to examine concentration levels
from residual contamination. Both sets of requirements are
described seperately below. The first set of modeling
requirements (i.e., during dredging) can be omitted for those
cases where the dredging effects are considered to be
completely isolated in the spill area. Examples of this would
include scenarios when silt curtains which are 100% effective
are used in conjunction, or when the spill site is isolated
using a full barrier.
7.2.6.1 During Dredging Operations
Model Type
To model the effects of the dredging operation, a numerical
model with sediment transport capabilities and a vertical line
source is required. Johnson (1981) evaluates a number of such
models designed to simulate dredging and barge dumping
activities. The selected model should incorporate sediment
scour and deposition also. Schnoor et al., (1982) utilized
such a model (Wechsler and Cogley, 1977) to simulate the
suspended sediment concentrations resulting from open water
disposal of dredged material on the Mississippi River.
Dimensionality and Grid Configuration
A minimum dimensionality of 2D (x-z) is recommended. However,
most of the dredging models reviewed by Johnson (1981) are
three-dimensional. The vertical dimension allows better
resolution for the resuspension and deposition areas.
Parameter Adjustment
The parameters adjusted should reflect the effects of the
increased turbulence induced by hydraulic dredging and
increased supended sediment concentrations from both types of
dredging. The flow source and sink terms associated with
hydraulic dredging shoud be negligible compared to the
in-stream flow, especially on a large river. Turbulent
diffusion or mixing coefficients in the lateral and vertical
directions should be increased. Gradation of the source
sediment should be specified because it affects the transport
4-71
-------
of the material. A vertical sediment concentration
distribution must be set for the line source. For the side
bank disposal of dredged material, Schnoor et al., (1982)
utilized the dredging rate of the barge and the channel depth,
width, and velocity to determine input sediment concentration.
More guidance on estimating suspended sediment loading is
provided in Section 7.3.
7.2.6.2 Post-Restoration
Model Type
A numerical far-field or exposure model with sediment
transport capabilities can be used to evaluate post-
restoration conditions. The model can utilize an empirical
source term or predicted suspended and deposited sediment
concentrations for initial conditions. The ability to
simulate scour and deposition of dredged material is required.
Dimensionality and Grid Configuration
A minimum two-dimensional (x-z) sediment transport simulation
is recommended, if vertical distribution of suspended
sediments is important. The grid spacing should be closer
along the bottom to represent large suspended sediment
concentration fluctuations.
Parameter Adjustment
The chemical concentration in the bed must be adjusted in
order to reflect the presence of deposited contaminated
sediments. Bottom topography, in the form of sediment bed
thickness, and water column depth may have to be adjusted for
those areas of dredge related heavy scour or deposition.
Suspended sediment concentrations predicted from the dredging
model can be used for initial levels. If no dredging modeling
is performed, the removal of contaminated bed by dredging,
capping, or any isolation and complete removal methods may be
simulated in a fashion similar to Onishi (1979). He used the
two-dimensional (x-z) sediment transport model FETRA (Onishi
et al., 1979) to locate areas along the river where
contaminated sediment was being deposited. Ten locations for
clean-up were simulated by removing the contaminated bed along
selected reaches. Figure 7.1 illustrates the changes in
Kepone concentrations from different clean-up areas. As
evidenced in Figure 7.1, a 34.5 km length clean-up region
reduced concentrations the most (55%), although a 22 km
clean-up region was quite close in level of reduction (48%).
This study did not evaluate the transport of contaminated
sediments over a period of time due to dredging itself.
However, given the size of the tidal river and location of
deposition areas, such effects would have been local, and were
4-72
-------
at
0.020
0.018
a 0.016
a.
z 0.014
aoi2
0.010
0.008
0.006
0.004
0.002
0
0.
UJ
_
o
:ASES
fcs^X v s
^ \t>
AN UP REGIONS
30 40 50 60 70 80 90
RIVER KILOMETERS
100
110
120 130
Figure 7.1 Reductions in total Kepone concentrations from
different dredging scenarios (Onishi, 1979).
4-73
-------
not important in terras of viewing Kepone loading into
Chesapeake Bay.
7.2.7 Removal: Excavation
This action is usually used in conjunction with full barriers,
as it can only be used on dry dewatered solids. Because the
spill area is assumed to be completely isolated during
implementation, no source area modeling is required.
Model Type
The model type for post-restoration will be dependent upon the
waterbody.
Dimensionality and Grid Configuration
There will be no change in dimensionality or grid
configuration; they are used dependent on waterbody
characteristics.
Parameter Adjustment
The only parameter adjustment will be the change in source
term. Before restoration, it is assumed that the source term
will be from a contaminated sediment bed, in either
concentration or mass per unit area form. Upon restoration,
any contaminant leaving the spill area will be from dewatering
operations. This source term may be represented for a finite
period of time with a empirically derived aqueous
concentration and flow rate.
7.2.8 Treatment
Modeling requirements for treatment actions will often be
represented as reductions in the source term for the
re-introduced waste water. Remedial action modeling is not
required if the treatment action does not affect in-stream
processes.
Model Type
Source area modeling is not required for either in-situ or
on-site actions.
Dimensionality and Grid Configuration
Dimensionality and grid configuration requirements will be
4-74
-------
dictated by the waterbody characteristics.
Parameter Adjustment
The only parameter adjustment is the reduction of a source
term for the in-situ action, and the addition of sink and
source terms for the on-site action. The new introduced
source will be empirically reduced from in-stream
concentrations.
7.3 PARAMETER ESTIMATION GUIDANCE
The parameters requiring adjustment to simulate the remedial
measures described previously can be characterized by four
groups: 1) source term parameters for contaminants and flow;
2) sediment-related parameters for bed and suspended
sediments; 3) boundary conditions, including channel geometry;
and 4) dispersion parameters. This section provides sources
of data and techniques for estimation of model parameters.
The guidance presented herein is only meant to be used in
support of, rather than in place of on-site field
measurements, sampling and laboratory studies. To the extent
possible, values for model parameters should be determined as
part of the remedial investigation process. This process is
meant to fill the data required to evaluate conceptual,
remedial action alternatives (JRB Associates, 1983).
Hopefully, this section can be used to more fully understand
those data required for remedial action modeling and in the
worst case aid in their estimation in absence of site-specific
data.
Where available, data sources and estimation techniques
pertinent to remedial action specific parameters are provided.
Both are extremely limited, however. For this reason, more
general data sources and estimation techniques are discussed
to provide a basis for at least the initial determination of
appropriate parameter values. Zison et al. (1978) is a good
general source for transport and fate parameters and
formulations.
7.3.1 Source Term Parameters
As evidenced throughout Section 7.2 on modeling requirements:,
the most common parameter adjustment for surface water
remedial actions is the modification of source terms. The
primary source discussed here is the introduction of a
dissolved contaminant and flow into receiving water. Sources
4-75
-------
of contaminated sediments from a discharge or sediment bed are
discussed in Section 7.3.2.
The simplest description for a concentration of a discharged
contaminant into a waterbody involves mass, flow, and time in
the following equation:
r _ M _ M
Ci~V~Qt (7.1)
where c i = initial concentration
M = mass of contaminant
V = volume of released water
Q = flow of discharged water
t = time
This initial concentration is subsequently mixed or diluted in
a mixing zone proximal to the discharge point. Complete and
instantaneous mixing across the channel width is often
assumed, except for wide rivers, estuaries, and lakes. Mills
et al. (1982) provide a number of expressions to describe
mixing zone geometry and concentrations. Reductions in
sources terms from dilution will often be accomplished by
increasing the source flow rate (Q). Reductions in the source
term according to contaminant mass removal or isolation
actions (i.e., excavation, installation of cofferdams) is
reflected by lower mass or concentration inputs, whichever the
exposure model requires. Actual reduction is determined
empirically? the user may want to evaluate different source
'term reductions to reflect various clean-up efficiencies
(i.e., 90%, 70%, 50%, 10%). Source terms may also be adjusted
according to the period of release. The timing of the release
during a spill can range from instantaneous to continuous. If
we assume that clean-up is never 100% efficient, residual
contamination may enter the water body for a finite period of
time during and after restoration. The new source term is
represented by a series of instantaneous releases at finite
time intervals. Estimates of the removal mass and time over
which discharge continues can be used to determine a new
source concentration for an exposure model and the number of
time steps it will be active.
7.3.2 Sediment Parameters
Many of the hazardous wastes that are discharged to receiving
waters are transported as particulates or via contaminated
sediments. Their ultimate sink can be the sediment bed, where
the contaminant can desorb back into the water column or be
resuspended with sediment particles as the result of scour.
4-76
-------
Most of the remedial actions described previously are designed
to isolate or remove contaminated sediments and solids. Two
areas of parameter adjustments are important: 1) those
affecting sediment transport, including scour and deposition;
and 2) those affecting the contaminated sediment as a source,
including contaminant mass, area of bed, and desorption and
diffusion coefficients. Each group is described below.
7.3.2.1 Sediment Transport Parameters
The important parameters of sediment transport are the
sediment diameter, specific gravity, settling (or fall)
velocity, the critical shear stress associated with
deposition of sediment, and the suspended sediment loading
term. Sediment diameter and specific gravity along with shear
velocity are the primary parameters used to describe sediment
transport. These parameters are not often adjusted, except to
represent new sediment material. Table 7.3 is a sediment
grade scale with sizes of different materials. Table 7.4
lists specific weights (gravities) for sediments in
representative waterbodies. If contaminant transport is
associated with certain particle sizes, the model should have
the ability to transport by grain size, not by total sediment
load only.
The settling velocity strongly affects the rate at which
sediments will be deposited on the bed. Net settling velocity
may be calculated in the model using bed shear stresses or be
a parameter input. In the latter case, it can be adjusted to
represent different sediment types, such as clay, from a
capping remedial action. Barnard (1978) developed an
empirical relationship between settling velocity and particle
diameter, presented in Figure 7.2. This relationship is based
on the assumption that the particle has a certain specific
gravity and shape factor at a specific water temperature. In
estuary and sometimes in reservoir analysis, parameters that
affect cohesive (silts and clays) sediment transport are
chemical conditions (e.g., salinity, pH, valence of cations),
concentration of suspended material, and mineral properties of
the particles such as sodium adsorption ratio (SAR) and cation
exchange capacity. This is paticularly important because the
sorption/desorption activity of contaminants is usually
associated with clays and silts.
The critical shear strength of the sediment bed will determine
the degree of erosion that can occur. It is a function of the
sediment type, in particular, the median sediment diameter in
the bed. Figure 7.3 is a diagram that can be used for such a
purpose.
Suspended sediment source terms, used to represent the effects
4-77
-------
TABLE 7.3 SEDIMENTATION GRADE SCALE (from Vanoni, 1975)
Copyrighted by the American Society of Civil
Engineers
Class Name
Very large boulders
Large boulders
Medium boulders
Small boulders
Large cobbles
Small cobbles
Very coarse gravel
Coarse gravel
Medium gravel
Fine gravel
Very fine gravel
Very coarse sand
Coarse sand
Medium sand
Find sand
Very fine sand
Coarse silt
Medium silt
Fine silt
Very fine Silt
Coarse clay
Medium clay
Fine city
Very fine clay
Size
Millimeters
2-1
1-1/2
1/2-1/4
1/4-1/8
1/8-1/16
1/16-1/32
1/32-1/64
1/64-1/128
1/128-1/256
1/256-1/512
1/512-1/024
1/1024-1/2048
1/2048-1/4096
4096-2048
2048-1024
1024-512
512-256
246-128
128-64
64-32
32-16
16-8
8-4
4-2
2.000-1.000
1.000-0.500
0.500-0.250
0.250-0.125
0.125-0.062
0.062-0.031
0.031-0.016
O.OV6-0.008
0.008-0.004
0.004-0.0020
0.0020-0.0010
0.0010-0.0005
0.0005-0.00024
Range
Microns Inches
160-80
80-40
40-20
20-10
10-5
5-2.5
2.5 -1.3
1.3 -0.6
0.6 -0.3
0.3 -0.16
0.16-0.08
2000-1000
1000-500
500-250
250-125
125-62
62-31
31-16
16-8
8-4
4-2
2-1
1-0.5
0.5-0.24
Approximate Sieve Mesh
Openings Per Inch
United States
Tyler Standard
2-1/2
5 5
9 10
16 18
32 35
60 60
115 120
250 230
4-78
-------
TABLE 7.4 SPECIFIC WEIGHTS OF SEDIMENTS SHOWING EXTREME
VARIATION (Vanoni, 1975) Copyrighted by the
American Society of Civil Engineers
Location
(1)
Lake Niedersonthofen, Bavaria, upper layer
Lake Niedersonthofen, 20 m depth
Lake Arthur, South Africa
Iowa River at Iowa City, Iowa
Missouri River near Kansas City, Mo.
Lake Claremore, Oklahoma
Lake McBride,Iowa
Powder River, Wyoming
Castle wood Reservoir, Colorado
Cedar River near Cedar Valley, Iowa
Lake Arthur, South Africa
Predominant
class of
sediment
<2)
marl*
marl*
clay
silt
silt
silt
silt
silt
sand
sand
sand
Specific weight,
in pounds per
cubic foot
(3)
21.6
89.6
38
52
74
54
60
81
92
109
100
a As used herein, marl is a mixture of calcium carbonate or dolomite and clay.
4-79
-------
SETTLING VELOCITY [cm/s]
I
CD
O
c:
l-f
tt>
— < ^
W CD CD
P> I-1 i-(
H O rt
3 O H-
(l> H- O
H rt M
CD
HI 01
M O H-
<£> M N
^J 0)
oo 0)
— C <
• VI 01
X! •
fl>
3 01
QJ n>
fl> rt
DJ rt
I-1
01 H-
(B 3
3
rt
-------
200
100
40 g
tn
20 £
3
O
10
.2 .4 .6 .8 I
MEDIAN SIZE OF BED SEDIMENT, d50
(MM)
Figure 7.3
\j/ and r for DuBoys relationship
as functions of median size of
bed sediment, where T = critical
shear stress and = coefficient
depending on grain size (Task
Committee on Preparation of Sedi-
mentation Manual, 1971). Copyrighted
by the American Society of Civil
Engineers.
4-81
-------
of a dredging operation, are often developed empirically based
on the mode of dredging, type of sediment, and location of
disposal. Schnoor et al. (1982) developed a suspended
sediment source term for a sediment transport/dredging model
(Wechsler and Cogley 1977) on the Mississippi River. Table
7.5 lists the factors they considered in order to obtain a
suspended sediment concentration of 120 mg/1.
7.3.2.2 Contaminated Sediment Bed Parameters—
These parameters are useful in deriving a source term to
represent desorption and diffusion from a sediment bed, as
well as to represent the residual sorbed contaminants that are
subject to scour and transport with sediment. Desorption and
diffusion parameters can be adjusted in a sediment transport
model to represent a reduced contaminated bed size or
concentration.
During the period of desorption (that is, after the
contaminant has been spilled and some has advected through the
waterbody with the rest settling in the bed in a sorbed phase)
the average aqueous concentration can be described by the
following equation:
C (7.2)
where Xo = concentration of pollutant in bed at time t=0
a = equivalent depth of water in sediment Ms, cm
S = specific gravity
Kp = partition coefficient
Ms = mass of contaminated sediment per unit area of
river bed, g/cm^
If data are not available, Ms and a can be estimated based on
the depth of contaminated sediments and percent solids by
weight values in Table 7.6.
In a sediment and contaminant transport model, the aqueous and
sorbed contaminant concentration will be computed using the
parameters above and flow parameters. The concentration (Xo )
at t=0 can be reduced empirically to represent some removal
action as can the mass of contaminated sediment per unit area
of bed (M§). The partition coefficient (Kp) does not need to
be changed unless some treatment action is applied to the bed
itself.
The contaminant can also diffuse back into the water column if
there is a sufficient gradient. Ditoro et al. (1981) define
an overall diffusive exchange coeficient (KL ) with the
4-82
-------
TABLE 7.5 DETERMINATION OF A CONTINUOUS SUSPENDED SEDIMENT
SOURCE TERM BY SCHNOOR ET AL.(1982)
Parameter
Mean velocity
Depth
Source width
Dredge capacity
Percent of Spoil
that is solids
Percent of solids
that are actually
entrained
Value
0.5 m/s
4.0 m
20m
1072 yd/hr
85%
Source
Field measurement
Field measurement
Observation
Communication with
Army Corps of
Engineers
Best Engineering
Judgement
Shallow depth near
shore, most of solids
were sand
4-83
-------
TABLE 7.6 MASS OF CONTAMINATED SEDIMENTS AND
EQUIVALENT WATER DEPTH AS A FUNCTION
OF DEPTH CONTAMINATION (Mills et al.,
1982)
Depth (mm) Percent Solids by Weight
1 20
50
80
5 20
50
80
10 20
50
80
20 20
50
80
50 20
50
80
100 20
50
80
Ms (g/cm2)
0.02
0.06
0.11
0.11
0.30
0.55
0.23
0.60
1.1
0.45
1.2
2.2
1.1
3.0
5.5
2.3
6.0
11.0
6(mm)
0.9
0.6
0.3
4.5
3.0
1.4
9.1
6.0
2.7
18.
12.
5.5
45.
30.
14.
91.
60.
27.
4-84
-------
following equation:
D
2
(7'3)
where KL = overall diffusive exchange coefficient, cm/day
D2 = interstitial water diffusion coefficient
<|> = porosity of sediments in bed
<7Z = length or depth or the gradient in the bed, cm
D2 can be estimated using the following equation from Manheim
(1970):
(7.4)
where D2Q = molecular diffusion coefficient of the
chemical
Lyman et al . (1982) provide a method to determine D2Q based on
molecular weight. The diffusion of the chemical into the
water column can then be adjusted in the model in order to
represent the addition of new material such as a clay cap or
another sediment type deposited on top of the bed, affecting
the sediment porosity and gradient length, or depth of
contamination .
7.3.3 Boundary Condition Parameters
Parameter adjustment for boundary conditions is very site
specific, and will vary from case to case according to the
channel geometry, the model dimensionality, and remedial
action configuration.
Channel geometry changes as a result of the addition of a
barrier can be simulated in two different ways, according to
the model dimensionality. For example, the parameter
adjustment in a one-dimensional (x) model to represent the
barrier involves reducing the reach width for the spill site
area. In a two-dimensional (x-y) simulation, the same barrier
can be represented by applying no-flow conditions for the grid
points or nodes along the barrier length.
Some remedial actions, such as silt curtains, require no-flow
boundary conditions for certain grid points in the water
column. In a two-dimensional (x-z) simulation, the no-flow
points should be defined to a specific depth in the water
column at some distance x which represents the curtain in the
4-85
-------
waterbody
7.3.4 Dispersion Parameters
As described in Section 2, dispersion is the aggregate result
of molecular diffusion, turbulent diffusion, and shear flow
dispersion for each dimension (x, y, and z) . Remedial actions
such as barriers, cofferdams, silt curtains and dredging can
cause an increase in dispersion, particularly turbulent
diffusion, and cause the contaminant to spread more rapidly.
Dispersion coefficents can be specified in numerical models to
represent the change in mixing from remedial action
implementation. Parameter adjustments for the longitudinal
dispersion coefficient, transverse mixing coefficient, and
vertical mixing coefficient are provided below. Fischer et
al. (1979) is a good source for dispersion parameters.
7.3.4.1 Longitudinal Dispersion Coefficient (Kx)—
Several simple methods for evaluating Kx are available in the
literature. Compilations of available methods include Fischer
et al. (1979) and Benedict (1978). The method Liu (1977) used
(based on the work of Fischer, 1967) is presented here because
it is relatively easy to calculate. Table 7.7 provides
reported values K of representative channels.
The longitudinal dispersion coefficient Kx is determined as
follows:
u 2 B3
Q
B
(7.5)
where (Liu, 1978):
where
D =
B =
u* =
ux "
A =
mean depth
mean width
bed shear velocity
mean stream velocity
cross sectional area
= river discharge
This parameter may be adjusted locally when barriers cause a
change in the depth, width, and cross-sectional area of the
river.
4-86
-------
TABLE 7.7 REPORTED VALUES FOR THE LONGITUDINAL MIXING
COEFFICIENTS FOR DIFFERENT CHANNELS
(Benedict, 1978)
Channel Depth (cm) Kv (M2/sec)
Chicago Ship Canal 807 3.0
Sacramento River 400 15.
River Derwent 25 4.6
South Platte River 46 16.2
Yuma Mesa Canal 345 0.76
Trapezoidal Lab Channel 2.1-4.7 0.123-0.22
Green-Duwamish River 110 6.5-8.5
Missouri River 270 1500.
Clinch River 58-210 14-47
Copper Creek, VA 48-85 9.5-21
Powell River, TN 85 9.5
Sinuous Laboratory 2.7-7.0 .51-3.1
Channel
4-87
-------
7.3.4.2 Transverse Mixing Coefficient—
The transverse or lateral coefficient Ky has been described by
Elder (1959) as:
K = a Du* (7.6)
where a = coefficient
D = depth
u^ = bed shear velocity
Transverse mixing becomes very important to contaminant
dispersion when barriers are placed in rivers. The
coefficient a can be adjusted to represent changes in channel
geometry from such a cause. Yotsukura and Cobb (1972)
reported values of a from 0.1 - 0.2 for straight channels, and
0.6 - 10 in the Missouri River. Fisher et al. (1979)
recommend a value of 0.6. Table 7.8 lists Ky values for
represenative channels. In a two-dimensional (x-y) simulation
with barriers that cause sharp bends in channel geometry,
higher values of a should be used.
7.3.4.3 Vertical Mixing Coefficient—
Fischer et al. (1979) provide the following _equation to
determine the average vertical mixing coefficient 8 v:
8 - 0.067 du
(7.7)
where 8V = vertical mixing coefficient
d = depth of an open channel flow
u* = shear velocity at a wall boundary or channel
bottom
This coefficient can be adjusted in two dimensional (x-z) or
three dimensional models in order to change the value for Kz
for remedial actions that cause an increase in vertical
mixing, such as hydraulic dredging and silt curtains.
4-88
-------
TABLE 7.8
EXPERIMENTAL MEASUREMENTS OF TRANSVERSE MIXING IN OPEN CHANNELS WITH
CURVES AND IRREGULAR SIDES (Fischer et al.f 1979) Copyrighted by
Academic Press
i
oo
VD
Channel
Missouri River near
Blair, Nebraska
Laboratory
Chnnnel
geometry
Meandering river
Smooth sides find
Channel
width,
W
(m)
200
2.2
Menn depth
of flow,
d
(m)
2.7
0.097
Menn
velocity,
u
(m/s)
1.75
0.11
Shear
velocity,
u*
-------
REFERENCES
Ambrose, R. 1982. Simulation Models for Assessing
Eutrophication Problems in River Basins, U.S. EPA,
Environmental Research Lab, Athens, GA. Presented at the
International Workshop on the Comparison of Application of
Mathematical Models for the Assessment of Changes in Water
Quality in River Basins, both surface and ground waters.
La Coruna, Spain.
Ambrose, R.B., S.I. Hill and L.A. Mulkey. 1983. Users Manual
for the Chemical Transport and Fate Model TOXIWASP, Version
1, EPA 600/3-83-005, U.S. Environmental Protection Agency,
Athens, GA.
Ambrose, R., and S. Roesch 1982. Dynamic Estuary Model
Performance. ASCE, Vol. 108, No. EE1.
Barnard, W. 1978. Prediction and Control of Dredged Material
Dispersion around Dredging and Open-water Pipeline Disposal
Operations. U.S. Army Engineer Waterways Experiment
Station. Vicksburg, MS.
Basta, D.J. and B.T. Bower, eds. 1982. Analyzing Natural
Systems. Resources for the Future, Johns Hopkins
University Press.
Benedict, B. 1978. Review of Toxic Spill Modeling,
AD/A-073-22, U.S. Department of Commerce.
Burns, L., R. Lassiter and D. Cline. 1982. Documentation for
the Exposure Assessment Modeling System (EXAMS), EPA
600/3-82-023, U.S. Environmental Protection Agency,
Environmental Research Laboratory, Athens, GA.
Callahan, M.C., M.W. Slimak, N.H. Gabel, J.P. Map, C.F.
Fowler, J.R. Freed, P. Jennings, R. L. Durfee, F.C.
Whitmore, B. Maestri, W.R. Mabey, B.R. Holt and C. Gould.
1979. Water-Related Environmental Fate of 129 Priority
Pollutants, EPA - 440/4-79-029ab, Volumes 1 and 2,
U.S. Environmental Protection Agency, Washington, D.C.
4-90
-------
Codell, R.B., K.T. Key and G. Whelan. 1982. A Collection of
Mathematical Models for Dispersion in Surface Water and
Groundwater. NUREG-0868, Division of Engineering, Office
of Nuclear Realtor Regulation, U.S. Nuclear Regulatory
Commission.
Colonna, G. 1982. Hazard Assessment Computer System (HAGS) -
Models Review and Validation. From Proceedings of the
Control of Hazardous Material Spills Conference, April
19-22, Milwaukee, WI.
Csanady, G.T. 1973. Turbulent Diffusion in the Environment.
D. Reidel Publ., Boston, MA.
Department of Transportation. 1978. CHRIS Response Methods
Handbook. COMDTINST MI6465.14. U.S. Coast Guard,
Washington, D.C.
Department of Transportation, U.S. Coast Guard. 1981. CHRIS
Hazardous Assessment Handbook C6-446-3.
Di Toro et al., 1981. Analysis of Fate of Chemicals in
Receiving Waters. CMA Project Env-7-W, Chemical
Manufacturers Association, Washington, D.C.
Di Toro, D., J. Fitzpatrick and R. Thomann. 1982.
Documentation for Water Quality Analysis Simulation Program
(WASP) and Model Verification Program (MVP). Environmental
Research Laboratory, U.S. Environmental Protection Agency,
Duluth, MN. EPA Contract Nl-68-01-3872.
Donigian, A.S. 1981. Water Quality Modeling in Relation to
Watershed Hydrology, In: Modeling Components of Hydrologic
Cycle, ed. V.P. Singh, Water Resources Publications,
Litteton, CO.
Edinger, J. and E. Buchak. 1982. Description and Appli-
cations of the LARM Code (Laterally-Averaged Reservoir
Model). By J. E. Edinger and Associates, Wayne, PA.
Elder, J.W. 1959. The Dispersion of Marked Fluid in
Turbulent Shear Flow. Journal of Fluid Mechanics, Vol. 5,
No. 4, pp. 544-560.
Environmental Protection Agency. 1982. The PCB
Contamination Problem in Waukegan, Illinois. U.S.
Environmental Protection Agency, Region V.
4-91
-------
Environmental Protection Agency. 1983. Environmental Data
Base and Model Directory. Two Volumes, EPA Information
Clearinghouse (PM 211A), U.S. Environmental Protection
Agency, Washington, D.C.
Fischer, H.B. 1967. The Mechanics of Dispersion in Natural
Streams. Journal of the Hydraulics Division, American
Society of Civil Engineers, Vol. 93, HY6, pp. 187-216.
Fischer, H.B., E.J. List, R.L.Y. Koh, J. Imberger and N.H.
Brooks. 1979. Mixing in Inland and Coastal Waters.
Academic Press, New York, NY.
Frick, W. E. 1981. A Theory and Users Guide for the Plume
Model MERGE. Tetra Tech, Inc. Corvallis, OR.
Fussell, D.R., H. Godjen, P. Hayward, R.H. Lillie, A. Manlo
and C. Panisi. 1981. Revised - Inland Oil Spill Clean-up
Manual. Conservation of Clean Air and Water - Europe
(Concawe), Report No. 7/81, Den Haag, Netherlands.
George, J.F. 1978. An Analysis of the Functional
Capabilities and Performance of Silt Curtains. Technical
Report No. D-78-39. Army Engineer Waterways Experiment
Station, Vicksburg, MI.
Herbes, S.E., G.T. Yeh and G.R. Southworth. 1982.
Transport Model for Coal-Liquefaction Product Spills in
Rivers. Presented at the Third Annual Meeting of the
Society of Environmental Toxicology and Chemistry,
Arlington, VA.
Huibregste, K.R., R.C. Scholz, R.E. Wullschleger, J. H. Moser,
E.R. Bellinger and C.A. Hansen. 1977. Manual for the
Control of Hazardous Material Spills, Vol. 1: Spill
Assessment and Water Treatment Techniques. EPA
600/2-77-227.
Hydrologic Engineering Center. 1977. HEC-6-Scour and
Deposition in Rivers and Reservoirs, Users Manual. U.S.
Army Corps of Engineers, Davis, CA. 723-62-62470.
JRB Associates. 1982. Handbook - Remedial Actions at Waste
Disposal Sites, EPA 625/6-82-006, Environmental Protection
Agency, Cincinnati, OH.
Johanson, R., J. Imhoff, H. Davis, J. Kittle and A. Donigian.
1981. User's Manual for the Hydrological Simulation
Program - Fortran (HSPF): Release 7.0, EPA 600/6-82-0046,
U.S. Environmental Protection Agency, Athens, GA.
4-92
-------
Johnson, B. 1981. Discussion of Mathematical Models for
Computing the Physical Fate of Solids Released in Open
Water During Dredging or Dredged Material Disposal. In:
Contaminants and Sediments, R. Baker, Ed. Chap. 15. Ann
Arbor Science, Ann Arbor, MI.
Karickhoff, S., D. Brown and T. Scott. 1979. Sorption of
Hydrophobic Pollutants on Natural Sediments. Wat. Res.
Vol. 13, pp. 241-248.
Karickhoff, S. 1979. Semi-Empirical Estimation of Sorption
on Hydrophobic Pollutants on Natural Sediments and Soils.
Chemosphere. Vol. 10, No. 8, pp. 833-846.
Krenkel, P.A. and V. Novotny. 1980. Water Quality
Management, Academic Press, New York, NY.
Liu, H. 1977. Predicting Dispersion Coefficient of Streams.
Journal of the Environmental Division, American Society of
Civil Engineers, Vol. 103, No. EE1, pp. 59-69.
Liu, H. 1978. Discussion of, Predicting Dispersion
Coefficient of Streams. Journal of the Environmental
Division, American Society of Civil Engineers, Vol. 104,
No. EE4, pp. 825-828.
Liu, S. and J. Leenderste. 1972. A Three - Dimensional
Model for Estuaries and Coastal Seas: Volume VI, Bristol
Bay Simulations. R-2405-NOVA. Rand Corp., Santa Monica,
CA.
Lyman, W.J., W.F. Reehl and D.H. Rosenblatt. 1982. Handbook
of Chemical Property Estimation Methods—Environmental
Behavior of Organic Compounds. McGraw-Hill Book Co., NY.
Maki, A., J. Dickson and J. Cairns, Eds. 1979.
Biotransformation and Fate of Chemicals in the Aquatic
Environment. Am. Soc. of Microbiology Press, Washington,
D.C.
Manheim, F.T. 1970. The Diffusion of Ions in
Unconsolidated Sediments. Earth Planet Sci. Letter, Vol.
9, pp. 307-309.
Mercer, J.W. and C.R. Faust. 1981. Ground-Water Modeling.
National Water Well Association, Worthington, OH.
Mills, W., J. Dean, D. Porcella, S. Gherini, R. Hudson, W.
Frick, G. Rupp and G. Bowie. 1982. Water Quality
Assessment: A Screening Procedure for Toxic and
Conventional Pollutants, EPA 600/6-82-004abc, Volumes 1,2.
4-93
-------
Morrow, T., F. Dodge, E. Bowles and D.W. Astleford. 1981.
Mixing and Dilution of Water Soluble Hazardous Chemical
Released in Navigable Rivers. Am. Soc. Mech. Eng.
81-FE-17.
Mulkey, L., R. Ambrose and T. Barnwell. 1982. Aquatic Fate
and Transport Modeling Techniques for Predicting
Environmental Exposure to Organic and Other Toxicants — A
Comparative Study. U.S. Environmental Protection Agency,
Environmental Res. Lab. Athens, GA. Presented at the
International Workshop on the Comparison of Application of
Mathematical Models for the Assessment of Changes in Water
Quality in River Basins, both Surface and Ground Waters.
La Coruna, Spain.
Neely, W., G. Blau and T. Alfrey. 1976. Mathematical Models
Predict Concentration - Time Profiles Resulting From
Chemical Spill in a River. Env. Sci. Tech., Vol. 10, No.
1.
Neely, W., D. Branson and G. Blau. 1974. Partition
Coefficient to Measure Bioconcentration Potential of
Organic Chemicals in Fish. Env. Sci. Tech.
O'Conner, D.J. 1967. The Temporal and Spatial Distribution
of Dissolved Oxygen in Streams, Water Resources Research,
Vol. 3, No. 1.
O'Connor, D. and K. Farley. 1981. Preliminary Analysis of
Kepone Distribution in the James River. Environmental
Engr. and Sci. Program, Manhattan College, Bronx, N.Y.
Onishi, Y. 1979. Mathematical Transport Modeling for Deter-
mination of Effectiveness of Kepone Clean-up Activities in
the James River Estuary. Presented at the ASCE Hydraulics
Division Conference.
Onishi, Y. 1981. Sediment - Contaminant Transport Model.
ASCE Vol 107, No. HY9, J. Hydraulics Division.
Onishi, Y., E.M. Arnold and D.W. Mayer. 1979. Modified
Finite Element Transport Model, FETRA, for Sediment and
Radionuclide Migration in Open Coastal Waters,
NUREG/CR-1026, PNL-3114, Pacific Northwest Laboratory,
Richland, WA.
Onishi, Y. and S. Wise. 1979. User's Manual for the
In-stream Sediment Contaminant Transport Model, SERATRA.
U.S. Environmental Protection Agency, Athens, GA.
4-94
-------
Onishi, Y., R.J. Serve, E.M. Arnold, C.E. Cowan and P.L.
Thompson. 1981. Critical Review: Radionuclide Transport,
Sediment Transport, and Water Quality Modeling; and
Radionuclide Adsorption/Desorption Mechanisms.
NUREG/CR-1322, Nuclear Regulatory Commission, Washington,
D.C.
Onishi, Y., G. Whelan and R.L. Skaggs. 1982. Development of
a Multimedia Radionuclide Exposure Assessment Methodology
for Low-Level Waste Management. PNL-3370, Pacific
Northwest Laboratory, Richland, WA.
Onishi, Y. and D.S. Trent. 1982. Mathematical Simulation of
Sediment and Radionuclide Transport in Estuaries - Testing
of Three-Dimensional Radionuclide Transport Modeling for
the Hudson River Estuary, N.Y. NUREG/CR-2423, PNL-4109,
Pacific Northwest Laboratory.
Orlob, G. 1971. Mathematical Modeling of Estuarial Systems.
International Symposium on Mathematical Modeling
Techniques.
Petty, J.E., W. Wakamiya, C.J. English, J.A. Strand and D.D.
Mahlum. 1982. Assessment of Synfuel Spill Clean-up
Options. PNL-4244, Prepared for the U.S. Department of
Energy by the Pacific Northwest Laboratory, Battelle
Memorial Institute.
Pritchard, D.W. 1969. Dispersion and Flushing of Pollutants
in Estuaries. Journal of Hydraulics Division, ASCE, HY1,
pp. 115-124.
Raj, P.K. and P.M. O'Farrel. 1977. Development of
Additional Hazard Assessment Models. CG-D-36-77, U.S.
Coast Guard, Dept. of Transportation, Washington, D.C.
Rausch, A. H., R.M. Kumar and L.J. Lynch. 1977. A Critical
Technical Review of Six Additional Hazard Assessment
Models. CG-D-54-77. U.S. Coast Guard, Dept. of
Transportation, Washington, D.C.
Schnoor, J. 1981. Fate and Transport Modeling for Toxic
Substances. Presented at the Pellston Conference on
Modeling the Fate of Chemicals in the Aquatic Environment,
Pellston, MI.
4-95
-------
Schnoor, J.L., A.R. Giaquinta, C. Sato, C.P. Robinson and
D.B. McDonald. 1982. Refinement and Verification of
Predicted models of Suspended Sediment Dispersion and
Desorption of Toxics from Dredged Sediments. IIHR Report
No. 249, Iowa Institute of Hydraulic Research, The
University of Iowa, Iowa City, IA.
Schnoor, J. and D.C. McAvoy. A Pesticide Transport and
Bioconcentration Model, Submitted to Journal of American
Society of Civil Engineers, Engineering Division. (In
press).
Smith, R.C. and K.S. Baker. 1981. Optical Properties of
the Clearest Natural Waters (200-800 NM). Applied Optics,
Volume 20, No. 2, pp. 177-184.
Task Committee on Preparation of Sedimentation Manual, 1971.
Sediment Discharge Formulas. Journal of the Hydraulics
Division, ASCE. Vol. 97, No. HYU, Proc. Paper No. 7786.
Thibodeaux, L. 1979. Chemodynamics; Environmental Movement
of Chemicals in Air, Water, and Soil. Wiley-Interscience,
NY.
Thomann, R.V. 1972. Systems Analysis and Water Quality
Management. McGraw-Hill, New York, N.Y.
Torguimsen, G. 1982. A Comprehensive Model for Oil Spill
Simulation. NOAA, Modeling and Simulation Studies,
Seattle, WA.
Tracer Inc., ed. 1971. Estuarine Modeling: An Assessment-
Water Pollution Control Research Series, No. 16070 DZV
02/71, U.S. Environmental Protection Agency, Water Quality
Office, Washington, D.C.
Vanoni, V.A., editor. 1975. Sedimentation Engineering.
ASCE-Manuals and Reports on Engineering Practice - No. 54,
New York, N.Y.
Wechsler, B.A. and D.R. Cogley. 1977. A Laboratory Study of
the Turbidity Generation Potential of Sediments to be
Dredged. Technical Report D-77-14, U.S. Army Engineer
Waterway, Experiment Station, CE, Vicksburg, MS.
Wolfe, N.L. 1980. Determining the Role of Hydrolysis in the
Fate of Organics in Natural Waters. In: Dynamics,
Exposure, and Hazard Assessment of Toxic Chemicals (R.
Haque,Ed.),Ann Arbor Science,Ann Arbor,MI.
4-96
-------
Yotsukura, N. and E.D. Cobb. 1972. Transverse Diffusion of
Solutes in Natural Streams. U.S. Geological Survey
Professional Paper 582-C.
Zaccor, J. 1981. A Mobile Stream Diversion System for
Hazardous Materials Spills Isolation. EPA - 600/52-81-219.
Zison, S.W., W.B. Mills, D. Deimer and C.W. Chen. 1978.
Rates, Constants, and Kinetics Formulations in Surface
Water Quality Modeling. EPA 600/3-78-105, U.S.
Environmental Protection Agency, Environmental Research
Laboratory, Athens, GA.
4-97
• US GOVERNMENT PRINTING OFFICE 1985 - 559-111/10815
------- |