Unitecf States
         Environmental Protection
         Agency
         Technology Transfer
EPA/625/4-89/019
[>EPA    Seminar Publication

         Transport and Fate of
         Contaminants in the
         Subsurface

-------

-------
Technology Transfer	   EPA/625/4-89/019 - Sept. 1989
Seminar Publication

Transport and
Fate of Contaminants  in the
Subsurface
Center for Environmental Research Information
Cincinnati, OH 45268

and

Robert S. Kerr Environmental Research Laboratory
Ada, OK 74820

-------
                                     NOTICE
The information in this document has been funded wholly or in part by the United States
Environmental Protection Agency under Contract 68-C8-0014 to Eastern Research Group, Inc. It
has been subject to the Agency's peer and administrative review, and it has been approved for
publication as an EPA document. Mention of trade names or commercial products  does not
constitute endorsement or recommendation for use.
                                         11

-------
                                   CONTENTS
                                                                              Page
Chapter 1.  Introduction  [[[       1


Chapter 2.  Physical Processes Controlling the Transport of Contaminants
            in the Aqueous Phase  .............. . . . ........... ; . . ..........           5

Chapter 3.  Physical Processes Controlling the Transport of Non-Aqueous
           Phase Liquids in the Subsurface  .................. ........... .            23

Chapter 4.  Determination of Physical Transport Parameters  .........................  29

Chapter 5.  Subsurface Chemical Processes  ..........................................  41

Chapter 6.  Subsurface Chemical Processes: Field Examples ........ .............   57

Chapter 7.  Microbial Ecology and Pollutant Biodegradation in Subsurface
           Ecosystems  [[[        gy

Chapter 8.  Microbiological Principles Influencing the Biorestoration of Aquifers ........  85

Chapter 9.  Modeling Subsurface Contaminant Transport and Fate  ............ . .....  101


-------
                          ACKNOWLEDGEMENTS
This publication is based on a series of technology transfer seminars that were conducted in all
ten EPA Regions between October 1987 and February 1988. The seminars provided regulators
and technical specialists with a brief but intensive overview of the physical, chemical, arid
biological processes that govern the transport and fate of contaminants in the subsurface. A
secondary purpose of the seminar was to provide a summary of modeling  approaches used to
make predictions about the transport and fate of contaminants in the subsurface and to describe
the current and potential regulatory uses of such models.

The EPA Center for Environmental Research Information (CERI) and Robert S. Kerr
Environmental Research Laboratory (RSKERL) developed this project to assist the  technical
support and technology transfer efforts of the  EPA Office of Solid Waste and  Emergency
Response (OSWER) in the area of subsurface remediation. Seminar development and
coordination were aided by numerous personnel, representing the EPA Office of Research and
Development (ORD), OSWER, and the EPA Regions. Principal contributors to  the  project
include:

Authors:
    Richard L. Johnson, Oregon Graduate Center, Beaverton, Oregon
    Joseph F. Keely, Ground-water Quality Consultant, Portland, Oregon
    Carl D.  Palmer, Oregon Graduate Center, Beaverton, Oregon
    Joseph M. Suflita, University of Oklahoma, Norman, Oklahoma
    William Fish, Oregon Graduate Center, Beaverton, Oregon

Seminar Speakers:

    Johnson, Keely, Palmer, and Suflita
    Dermont Bouchard, ORD/RSKERL, Ada, Oklahoma
    Michael Henson, ORD/RSKERL, Ada, Oklahoma

Technical  Reviewers:

    Jack Keeley, Consultant, Ada, Oklahoma
    Ronald C. Sims, Utah State University, Logan, Utah
    David Ostendorf, University of Massachusetts, Amherst, Massachusetts

 Project Managers:

    Carol Grove, ORD/CERI, Cincinnati, Ohio
    Chuck Marshall, JACA Corp., Ft. Washington, Pennsylvania
    Marion R. Scalf, ORD/RSKERL, Ada, Oklahoma
    Mark Lennon and Sylvie Stanish, Eastern Research Group, Inc., Arlington, Massachusetts

 Seminar Development/Coordination:

    ORD/RSKERL: Don Draper, John Matthews
    ORD/OEPER: Will LaVeille
    OSWER: Janette  Hansen, Victor Hays, Amy Mills, Charles  Perry, Peter Tong, Burnell
             Vincent
    OGWP: Carey Carpenter
    Region I: David Lang, Stephen Mangion, John Zannos
    Region II: Ken Wenz
                                          IV

-------
Region III: Eileen Burrows, Dom DiGivlio, Mindi Snoparshy, Gerallyn Vails
Region IV: Elmer Akin
Region V: Ken Chin, Dale Helmers, Ken Westlake
Region VI: Debbie Wright
Region VII: Jack Coakley
Region VIII: Charles Brinkman
Region IX: John Duff

-------

-------
                                        CHAPTER 1


                                       INTRODUCTION

                                         Joseph F. Keely
Congress requires that the U.S. Environmental
Protection Agency (EPA), as well as other regulatory
entities and the  regulated  community, meet four
interrelated objectives for the protection of ground-
water quality. These objectives are:

1.  Assessment of the probable impact of existing
   pollution on ground water at  points of
   withdrawal or discharge (Safe Drinking  Water
   Act of 1974 & 1986 (SDWA)).
2.  Establishment of criteria for location, design,
   and operation of waste disposal  activities to
   prevent contamination of ground water, or
   movement  of contaminants to  points of
   withdrawal or discharge (Resource Conservation
   and Recovery Act of 1976 (RCRA),  and the
   Hazardous and Solid Waste Amendments of 1984
   (HSWA)).
3.  Regulation of the production, use, and disposal of
   specific chemicals possessing an unacceptably
   high potential for contaminating ground water
   when released  to  the environment (Toxic
   Substances Control Act (TSCA), and the Federal
   Insecticide,  Fungicide, and  Rodenticide  Act
   (FIFRA)).
4.  Development of remediation technologies that
   are effective in protecting and restoring ground
   water  quality without being  unnecessarily
   complex or costly and without unduly restricting
   other land  use activities  (Comprehensive
   Environmental Response, Compensation,  and
   Liability Act of 1980 (CERCLA or Superfund),
   and the Superfund Amendments and Reauthori-
   zation Act of 1986 (SARA)).

To achieve these  objectives, definitive  knowledge of
the transport and fate of contaminants in  the
subsurface environment is essential.  Without this
knowledge, regulatory agencies (such as EPA) run
the twin risks of under-control and  over-control.
Regulatory under-control would result  in inadequate
prevention and cleanup of ground-water contami-
nation. Regulatory over-control would result  in
costly preventative actions and  remedial responses
to contamination. However, gaining and using
knowledge about contaminant transport and fate can
be difficult because of the complexity of the  sub-
surface environment. The activities of site character-
ization and remediation can be used to illustrate this
complexity.
Site Characterization
Transport  and fate assessments require inter-
disciplinary analyses and interpretations because
the processes involved in  these activities are
naturally intertwined. Examining each process  in
isolation is much like taking photographs of an
object from different perspectives and then trying to
piece them together to describe the object. Each
transport process must be viewed from the broadest
of interdisciplinary viewpoints, and the interactions
between them identified and understood. In addition
to a sound conceptual basis, integrating information
on geologic, hydrologic, chemical, and biological
processes into an effective contaminant transport
evaluation requires data that are accurate, precise,
and appropriate at the intended problem scale.

The issues of contaminant transport and fate in the
subsurface  are particularly difficult to address  at
Superfund sites because of the complex array of the
chemical wastes involved.  The hydrogeologic
settings of these sites are usually measured  in

-------
hundreds of feet and, at this scale, are extremely
complicated when characterized for a  remediation
plan. The methods and tools used for large-scale
characterizations are generally applicable to the
specialized needs at hazardous waste sites; however,
the transition to smaller  scale is fraught  with
scientific and economic problems. Some stem  from
the highly variable nature  of contaminant
distributions at hazardous  waste sites, and other
problems result from the limitations  of available
methods, tools, and theories.
Even though a given parameter, such  as hydraulic
conductivity, can be  measured correctly and  with
great reproducibility, it is difficult to know how
closely  an observation actually represents the
vertical and horizontal distribution of conductivities
found at a site. Therefore, when using a conceptual
model to interpret contaminant transport processes,
it is crucial that special attention be  given to the
spatial and temporal variations of the collected data.

To circumvent the large numbers of measurements
and samples needed to reduce uncertainties in
dealing  with subsurface parameters, more compre-
hensive theories are constantly under development.
The use of these newly developed theories, however,
is often frustrating because many call for data which
are not yet practically obtainable, such as chemical
interaction coefficients or relative permeabilities of
immiscible solvents and water.  Therefore, modern
contaminant transport and  fate studies necessarily
involve  a compromise between sophisticated
theories, current limitations for acquiring data, and
economics.
Remediation
A major issue in  cleaning up ground-water
contamination is determining when remediation is
complete. In remedial  actions, the  level of
contamination measured at monitoring wells may be
dramatically reduced after a moderate  period of
time, but low levels of contamination usually persist.
In parallel, the  contaminant load removed by
extraction wells, for instance, declines over time and
gradually approaches a residual level in  the latter
stages (Figure 1). A decision must be made whether
to continue or end  remediation. By  continuing
remediation, efforts will be made to clean up small
amounts  of residual contamination. However, if
remediation is ended prematurely, an increase in the
level of ground-water contamination  may follow
(Figure 2).

There  are several contaminant transport processes
that may be responsible for the persistence of
residual  contamination  and the kind  of effect
depicted in Figure 2. In order to generate such an
     ON


    OFF

    MAX
 o
 I
 ui
 o
 o
RESIDUAL
CONTAMINATION
      \
                     —  TIME   -^~
 Figure 1. Apparent cleanup by pump-and-treat remediation.
    OFF

    MAX
 O
 i=
 m
 O
 o
 o
          — TIME
                          1     2
Figure 2. Contaminant rebound after  remediation
        ceases.
effect, releases of contaminant residuals must be
slow relative to water movement through  the
subsurface caused by pumping. Transport processes
that generate this kind of behavior include: (1)
diffusion of contaminants within  spatially variable
sediments; (2) hydrodynamic isolation; (3) sorption-
desorption; and (4) liquid-liquid partitioning.

Flow through the  zones of highest  hydraulic
conductivity results in rapid cleansing of these zones
by extraction wellfields, but cleanup of contaminants
in low permeability zones  can only occur after the
slow process of diffusion takes place (Figure 3). The
situation is similar, though reversed,  for  in  situ
remediations that require the injection and delivery
of nutrients or reactants  to the  zone  of intended
action. Because the surface area of low-permeability
sediments is greater than that for high-permeability
sediments, greater amounts of contaminants accum-
ulate on them. Hence, the  majority of contaminant
reserves may be available only  under diffusion-
controlled  conditions in many heterogeneous
settings.

-------
 LOW PERMEABILITY
 STRATA   I
                                                                                    ADVECTION
Figure 3. Remediation effectiveness limited by variations
        in permeability.
The operation of any remediation wellfield results in
the formation of stagnation zones downgradient of
extraction wells and upgradient of injection wells.
These zones are hydrodynamically isolated so that
mass transport within the isolated water will take
place only by diffusion. Because of this, the portion of
the contaminant plurne lying within a  wellfield's
associated stagnation zones will not  be  effectively
remediated.

For remediation efforts  involving compounds that
readily sorb to aquifer materials, the number of pore
volumes to be  removed depends not only on  the
sorptive tendencies of the contaminant, but also on
whether flow rates during remediation are too rapid
to allow contaminant  levels to  approach local
equilibrium (Figure 4). If insufficient contact time is
allowed, the affected water is advected away from
sorbed contaminants prior to reaching equilibrium
and is replaced by upgradient fresh water. This
method of removal generates large volumes of mildly
contaminated water where small volumes of highly
contaminated water would otherwise result.

When non-aqueous phase liquid (NAPL) residuals,
such as gasoline, are trapped in pores  by surface
tension, diffusive liquid-liquid partitioning controls
dissolution of the toxic compounds within the NAPLs
into the ground water. As with sorbing compounds,
flow rates during remediation may be too rapid to
allow saturation levels  of the partitioned contam-
inants to be reached (Figure 5) and large volumes of
mildly contaminated water will be generated.

The practical use of remediation wellfields and other
ground-water  cleanup technologies are highly
dependent on site-specific knowledge of the influence
of transport processes on contaminant levels. There
is still much to be learned about highly specific  and
cost-effective remediations; however,  far more could
be accomplished if the  processes  that govern  the
behavior and treatability of contaminants were
actively investigated at each site.  In general, con-
                                                                  ^ORGANIC CARBON OR
                                                                   MINERAL OXIDE SURFACE
    o
    UJ
    O
    o
    o
                                                            EQUIL, CONG.
                      SLOW
                      DESORPTION
INITIAL RAPID
DESORPTION
                    TIME

Figure 4. Sorption limitations on remediation effectiveness.

ventional field characterization efforts have not led
to satisfactory remediations. Recent transport-
process-oriented approaches of characterization are
resulting  in more permanent and cost-effective
remediations.

There are many misconceptions regarding  the
processes  affecting the transport  and fate of
contaminants in the  subsurface. Some of these are
relatively  easy to address  by educational efforts
while others can be addressed only by applied
research. This document will describe some of the
information  known  about transport and fate of
contaminants in the  subsurface. By understanding
and  using  this information, as well as information
derived from future research, the ground-water
protection  objectives required  by Congress may be
met.
EPA's Transport and Fate Research
Program
The U.S. EPA Office of Research and Development
(ORD) operates 12 national laboratories, several of
which address various aspects of  ground-water
contamination. The  Environmental Monitoring
Systems Laboratory in Las Vegas, Nevada develops

-------
          ADVECTION
                                UQUID:LIQUID
                                PARTITIONING
               SOLUBILITY
               LIMITED
     GROUND-WATER VELOCITY  	»~
Figure 5.  Liquid partitioning limitations on remediation
         effectiveness.

monitoring systems,  with emphasis on hazardous
waste site characterization, underground storage
tank leak detection, and soil vapor monitoring.
Engineering  aspects  of the prevention  and
remediation of ground-water  contamination are
addressed  by the Risk Reduction Engineering
Research Laboratory in Cincinnati, Ohio which
develops source control  measures (e.g., landfill
covers and liners, and waste caps) and waste disposal
techniques (e.g., solidification, pump-and-treat, and
incineration). Other ORD laboratories also support
 research that is germane to ground-water contam-
 ination issues (notably, those laboratories in Ada,
 Oklahoma; Athens, Georgia;  Edison, New Jersey;
 and Corvallis, Oregon).

 The  Robert S. Kerr  Environmental Research
 Laboratory (RSKERL) in Ada, Oklahoma focuses on
 these issues as its primary charge. The Kerr
 Laboratory serves as EPA's center for studies of the
 transport and fate of contaminants in the subsurface.
 The research program includes development of
 methodologies for  protection and  restoration of
 ground-water quality, and evaluation of the applica-
• bility and limitations of using natural soil  and
 subsurface processes for the treatment of hazardous
 wastes.  RSKERL's efforts  in the 1980s have
 increasingly focused on improvement of site charac-
 terization and remediation methods, with special
 emphasis on identification and quantification of the
 mechanisms by which natural processes govern the
 transport and fate of contaminants in the subsurface.

 Transport processes research at RSKERL is divided
 into three major areas:

    1.  Hydrologic processes, which act  to influence
       the movement of water (the primary vehicle
       for subsurface contaminant movement).
    2.  Abiotic processes, which are the physical and
       chemical interactions  that  cause  contami-
       nants to move at different rates than those of
       the ground water.
    3.  Biotic processes, which are the  microbially
       mediated transformations of contaminants
       in the subsurface to other compounds.

 In the subsurface,  however,  these  processes  are
 inseparable, and RSKERL's comprehensive research
 goal is to ultimately  have the knowledge to integrate
 the influences of these processes into  a singular
 understanding of contaminant behavior  in the sub-
 surface.

-------
                                       CHAPTER 2


        PHYSICAL PROCESSES  CONTROLLING THE TRANSPORT OF
                   CONTAMINANTS IN THE  AQUEOUS PHASE

                              Carl D. Palmer and Richard L. Johnson
introduction
Interest in the transport and fate of contaminants in
terrestrial subsurface environments is based on
concern for the protection and remediation of both
ground- and surface-water resources. To achieve this
protection, it is necessary to: (1) predict.the time of
arrival and concentration  of contaminants at  a
receptor such as a monitoring well, a water-supply
well, or a surface water body; (2) design safe, cost-
effective waste facilities; (3) install effective monitor-
ing systems; and (4) develop efficient and cost-
effective strategies for remediation of contaminated
aquifers. To attain these  goals, the processes
involved in the transport and  transformation of
contaminants in both porous and fractured media,
and under  either  saturated or  undersaturated
conditions must be understood. This chapter  will
discuss some issues associated with the  physical
processes controlling contaminant transport in the
subsurface. The advection-dispersion theory in
saturated, porous media will be described as well as
the issues pertaining to contaminant transport in the
vadose zone and in fractured rock. This information
can assist contractors, consulting engineers, and
scientists in designing more efficient and cost-
effective monitoring networks and remediation
strategies, and safer waste facilities.
 Contaminant Transport in Saturated,
 Porous Media

 Advection-Dispersion Theory
 If the concentration of a contaminant were measured
 in a monitoring well located between a contaminant
source and a receptor such as a water supply well
(Figure 6), a concentration breakthrough curve could
be obtained  (Figure  7).  In the graph, the
concentrations do not immediately increase in a step
function but, instead, increase more gently in an S-
shaped curve. In a one-dimensional, homogeneous
system, the arrival of the center of mass is the result
of advection  while the spread of the breakthrough
curve is the result of dispersion.

Advection is the  transport of a  non-reactive,
conservative tracer at an average ground-water
velocity. The average linear  velocity, v, at which
ground water flows through a porous aquifer is:
            v = - (K/6f) (dh/dx)
                      t>
(1)
where K  is the hydraulic conductivity  of the
formation  in the direction of ground-water flow, 6t is
the porosity of the formation, and (dh/dx) is the
hydraulic  gradient in the direction of ground-water
flow (Freeze and Cherry, 1979). The velocity given by
this equation  can be substantially different for
solutes that react through precipitation/dissolution,
adsorption, and/or partitioning within the geologic
media (see Chapter 5).

The study of dispersion phenomena is important for
predicting the  time when  an action limit,  a
concentration limit used in regulations such as
drinking water standards,  will be reached and for
determining optimal, cost-effective strategies for
aquifer remediation.  The  classical mathematical
approach  used to determine  solute transport in
porous media  is the advection-dispersion equation.
This equation is written in its one-dimensional form
as:

-------
            WASTE
   AQUIFER
                           VADOSE
                           ,  ZONE  \
MONITORING
   WELL
WATER
SUPPLY
WELL
 Figure 6. Site containing a monitoring well, contaminant source, and receptor.
                      BREAKTHROUGH CURVE
        1.0

        0.5
   o
   o
        0.0
                                        TIME
Figure 7.  Concentration of a contaminant in a monitoring well.

-------
     3 /   acy    ac   ac
     — D —  - v — = —-
     ax\   ax /   .ax   at
                    (2)
where D is the dispersion coefficient, v is the ground-
water velocity, C is the concentration of the solute, x
is the spatial coordinate, and t is time. This equation
is an expression of the mass balance of a contami-
nant within the aquifer as a result of dispersion,
advection, and change in storage. These processes
are represented respectively by the first, second, and
third terms of Equation 2.

Early studies of the dispersion coefficient revealed
that it  varies  with the ground-water  velocity
(Perkins and Johnston, 1963). A plot of the D/D0
versus vd/D0 (Figure 8) where D0 is the molecular
diffusion coefficient of the contaminant, v is the
ground-water velocity, and d is the grain diameter of
the porous medium,  shows that the  dispersion
coefficient is relatively constant at low velocities, but
increases linearly with velocity as the ground-water
velocities increase. These experimental results led
investigators  to propose that  the  dispersion
coefficient can be written as the sum of two terms: an
effective molecular diffusion coefficient, D^, and
mechanical dispersion coefficient, Dm:
                                        D = D, + D
                                              d     m
                                                    (3)
                                                  The mechanical dispersion coefficient is proportional
                                                  to the velocity:
                                            D  = av
                                             m
                                                    (4)
                           where a is the constant of proportionality and is
                           known as the dispersivity parameter. This param-
                           eter continues to be the focus of a great deal of
                           research and controversy. At a small scale of meas-
                           urement, mechanical dispersion is the result of: (1)
                           velocity variations within a pore; (2) different pore
                           geometries; and (3)  the divergence of flow  lines
                           around the sand grains present in a porous medium
                           (Gillham and Cherry, 1982).

                           The effective molecular diffusion coefficient is a
                           value for diffusion within the porous medium. It  can
                           be obtained from  the solution diffusion  coefficient,
                           Do, by:
                                               D, =
                                                d
                                                    (5)
                           where T is the tortuosity of the medium. Tortuosity is
                           a factor that accounts for the increased distance a
          10
           10
           10
             -1
          10
                    I   I  I I  Mill     I   I  I  I I Illl    I   I  I 1  I 11II
                   DL = Longitudinal Dispersion Coefficient
                   Do = Molecular Diffusion Coefficient
                   v  = Solute Velocity
                   d  = Average Grain Diameter
                                             I   I  I  I I Illl
                                           I   I  I  I/I I
                                          TRANSITION ZONE
                    DIFFUSION
                    DOMINATED
                     L=  D0 T
i  i  i i mi
I  I I Illl
i  i  i 1111     i   i  i  i mi i    i   i  i  i i in
                                                                               ADVECTION
                                                                               DOMINATED
               10
                 -3
       10
                                -2
   10
                                               -1
     10
                                                                                1
10
10
                                                 vd/D,
  Figure 8.
  Dispersion coefficient as a function of ground-water velocity, v, and grain diameter, d (after Perkins and Johnston.1963).

-------
diffusing ion must travel to get around the sand
grains. For granular, porous media, x is typically in
the range of 0.6 to 0.7 (Perkins and Johnston, 1963;
Gillham, et al., 1984). A simple theoretical calcu-
lation indicates that, rather than travelling through
a^grain of diameter d, the solute  must travel the
distance around the grain or nd/2. Thus, the
tortuosity should have a value of approximately 0.64.

The advection-dispersion equation  (Equation 2) can
be solved with appropriate  boundary and  initial
conditions to yield the concentration breakthrough
curves or concentration distribution curves. For a
one-dimensional system with a constant concen-
tration of solute at the boundary, the concentration
at a given distance from  the source results in an  S-
shaped curve (Figure 9). The concentration break-
through curves arrive at points further from the
source at later times and are more spread out. The
concentration distribution curves look like a mirror
image of the breakthrough  curves with greater
concentrations toward the source (Figure 10).

The discussion of the advection-dispersion equation
has been limited thus far to the one-dimensional
case. The same general principles can be applied  to
two- and three-dimensional  problems.  Figure 11
shows the transport of a contaminant slug through a
porous aquifer in two dimensions. As the center of
mass of the slug moves further from its initial
location within the aquifer by advection with the
ground-water flow, the slug spreads. The spreading
in the direction of ground-water flow is the longi-
tudinal dispersion, while the spread in the direction
perpendicular to the ground-water flow is known as
the transverse dispersion.

If the slug is viewed in three dimensions, there are
three dispersion coefficients,  one  longitudinal and'
two transverse. In its general mathematical-,
formulation, the dispersion coefficient is a second
rank tensor. More mathematically detailed descrip-
tions  of the advection-dispersion equation can' be ,'•
found in the works of Bear (1979 and 1969).         •
Application of Advection-Dispersion
Theory
One-, two-, and three-dimensional advection dis-
persion equations have been used  to simulate the
transport of contaminants. However, discrepancies
between theory and laboratory experiments  were
observed. Investigators attribute these discrepancies
to a variety of mechanisms including immobile zones
of water within experimental columns, solution-solid
                              BREAKTHROUGH CURVE
              1
           o.s
 LU
           0.4
 ceo
     O   0.2
     O
                               DISPERSION
                                                    TIME
 Figure  9.  Concentration breakthrough curves derived from the advection-dispersion equation.

-------
                        CONCENTRATION DISTRIBUTION
LU
LU
DC
                                                DISTANCE
Figure 10.  Concentration distribution curves derived from the advection-dispersion equation.
interface processes, anion exclusion, and diffusion in
and out of aggregates. An apparent scale dependency
of the dispersion coefficient also has been observed
(Table 1).  Laboratory scale experiments generally
Table 1.    Longitudinal Dispersivity Values Obtained from
          Different Scale Experiments (After Gillham and
          Cherry, 1982)
         Type of Test            Longitudinal Dispersivity
Laboratory Tests
Natural Gradient Tracer Tests
Single Well Tests
Radial and Two Well Tests
Model Calibration of Contaminant
Plumes '
0.0001 to 0.01
0.01 to 2 m
0.03 to 0.3 m
0.5 to 15 m
3 to 1 00 m
m
yield small values of longitudinal  dispersivity
(0.0001 to 0.01 m) while field tracer tests and model
calibration of contaminant plumes yield longitudinal
dispersivity values  in the 10 to 100 m range. Even
within a single  tracer experiment, longitudinal
dispersivity values  are observed to increase with
increasing  transport distance (Figure 12). This
apparent increase in the longitudinal dispersivity
parameter with an increasing spatial or time scale
indicates that the assumptions often used when
applying classic advection-dispersion theory in
natural geologic materials are not applicable.

Furthermore, longitudinal dispersivity values gen-
erally were considered to be  only 10 to 30 times
larger than transverse dispersivity values. If trans-
verse dispersivity  values are large, contaminant
plumes will spread over the entire thickness of an
aquifer (Figure 13A). This is contrary  to the  long,
thin plumes (Figure 13B) observed in the field where
detailed three-dimensional monitoring  was per-
formed (MacFarland, et al., 1983; Kimmel  and
Braids,  1980). These observations  indicate that
transverse dispersivity must necessarily be very
small and in many situations may even be close to
zero (Sudicky,  1986; Frind and Hokkanen, 1987). If
so, the transverse dispersion coefficient  would be
equal to the effective diffusion coefficient within the
medium.

There is a growing consensus among contaminant
hydrologists that the large longitudinal dispersion

-------
                       ADVECTION AND DISPERSION
                         OF A CONTAMINANT SLUG
cc
LLJ
                                                                                 o
    LL
O
CC
o
Figure 11.  Transport pf a contaminant slug through a porous aquifer.
                                                                   X2
                                                      A. HYPOTHETICAL CONTAMINANT PLUME



                                                      WITH A LARGE TRANSVERSE DISPERSIVITY
                                                                 •~



                                                         WASTE
to
   0.5






   "

   0.3



   0.2



   0.1



    0
                     LONGITUDINAL DISPERSIVITY


                     DATA FROM FHEYBERQ (1986)
 B. HYPOTHETICAL CONTAMINANT PLUME



WITH A SMALL TRANSVERSE DISPERSIVITY





   WASTE
            20     40     60     80

             DISTANCE FROM SOURCE (m)
                                       100
Figure 12.  Increase in longitudinal dispersivity

         with transport distance.
                                              le
                                              figure 13. Hypothetical contaminant plumes for

                                                      large (A) and small (B) dispersivities.
                                          10

-------
coefficients observed in field experiments and ob-
tained by model calibration of contaminant plumes
are the result of aquifer heterogeneity. The effects of
heterogeneity on the spread of contaminants is easily
illustrated in an ideally stratified aquifer with layers
of sediment of different hydraulic conductivities
(Mercado,  1967).  Contaminants move rapidly along
the layers with higher permeability and more slowly
along the lower permeability layers (Figure 14). If
water is sampled from monitoring wells that are
screened through the various layers, the sample  is an
integration of the concentrations in each layer.
Plotting concentration versus distance  from the
source reveals a curve with a large amount of spread
even  though only advection is considered. This
macroscopic dispersion is the result of aquifer hetero-
geneity and not the pore-scale processes  described
above.
  C0 ..*.




t



'




• f
\
~t



• 1

• •






ft


Ki
Ka
Ks
"Ki|
Ks
—
—

     o
     o
                     DISTANCE
Figure 14.
Contaminant  distributions and concentrations in an
ideally stratified aquifer (after Gillham and Cherry,1982).

Investigators can calculate contaminant transport
rates through stratified aquifers by assuming  that
hydraulic conductivities follow a Gaussian distribu-
tion. When completed, these  analyses indicate that
the spread in  the contaminant  distribution is
proportional  to  the  mean distance travelled, L,
rather than L1/2 which  is obtained for the  classic
advection-dispersion  theory (Mercado, 1967). Thus,
the spatial spread of the contaminant concentrations
is greater  than that predicted by the  advection-
dispersion theory.  If the stratified medium is
assumed to be homogeneous and if the  advection-
dispersion theory is  applied,  the apparent disper-
sivity parameters increase with distance from the
source.

All geologic formations are heterogeneous and none
are perfectly stratified. Bedforms are often lenticular
and may contain cross-stratification or graded bed-
ding. Detailing the distribution of hydraulic conduc-
tivity in the  subsurface is a  formidable task  that
cannot be done at waste sites on a routine basis.
Heterogeneity can, nonetheless, be quantified by
considering hydraulic conductivity to be a random
process, and determining its important statistics
such as the mean, variance, and autocorrelation
function.

Such  statistical analyses were performed on the
Borden aquifer (Ontario, Canada) where perme-
ameter measurements on more than 3,000 samples
were analyzed (Sudicky, 1986). In an aquifer that is
ostensibly described as homogeneous,  hydraulic
conductivity was found  to vary  by  more than two
orders of magnitude. Statistical analysis of this data
further revealed that the 'hydraulic conductivity is
log-normally distributed. Autocorrelation length-
scales are 1.6 m in the horizontal direction and 0,10
m in  the vertical direction.  Using this information
and applying the stochastic theory of Gelhar and
Axness (1983), Sudicky (1986)  calculated the
asymptotic longitudinal dispersivity for the Borden
aquifer to be 0.61 m, a value that is close to the 0.43
m reported by Freyberg (1986)  for  the  large-scale
tracer test conducted at the site  The  asymptotic
transverse dispersivity values calculated by Sudicky
(1986) are very  close to zero, indicating that the
transverse dispersion coefficient is on the order of the
effective molecular diffusion coefficient.

Gillham, et al. (1984) proposed  an advection-
diffusion  model for the transport of solutes in
heterogeneous, porous media. This model recognizes
that aquifers are heterogeneous  and that advection
is the key process controlling the rate of movement of
solutes through the layers of higher permeability. As
the front moves through those layers, some of the
solute is lost to the lower permeability layers via
molecular diffusion. This advection-diffusion model
was tested in the laboratory (Sudicky, et al., 1985)
using a sandbox model composed of a layer of sand
sandwiched between layers of  silt.  If molecular
diffusion in the transverse direction is ignored, the
one-dimensional advection-dispersion equation can
be applied (the dashed line in Figure  15). The
experimental data, however, revealed retardation in
the solute front and a  much greater spread in the
breakthrough curve. A  mathematical solution that
includes diffusion into the lower-permeability layers
was able to simulate the experimental data. If the
one-dimensional equation is used  to analyze the
experimental curves, a  large apparent dispersivity
would be obtained for the higher permeability layer.

While aquifer  heterogeneity  is a major factor
contributing to the spread of contaminants, other
processes  contribute  to contaminant  transport
characteristics as well.  Diverging flow lines spread
contaminants by advection over a larger cross-
section of the  aquifer. Temporal variations in the
water table can change the direction of ground-water
                                                 11

-------
 o
 <
 DC

 ill
 O
 O
 o
                                BREAKTHROUGH CURVES
                SHOWING EFFECT OF TRANSVERSE  DIFFUSION
0.5
 i
 i
 i
 i
 i
 i
i
i
i
                          10
                              i	1	,	,	

                              ADVECTION-DISPERS10N


                                ADVECTION-DIFFUSION
                                                  TRAVEL DISTANCE * 1.0 m
                                                  SAND THICKNESS =» 0.03 m
                                                  GROUNDWATER VELOCITY - 0.10 m/day
                                                  DISPERSIVITY - 0.001 m
                                                  DIFFUSION COEFFICIENT « 1.2 X 10'*ma/»
                                    20             30
                                    TIME (DAYS)
                                               40
50
 Figure 15.  Breakthrough curves showing the effect of transverse diffusion.
flow and contribute to the lateral spread of contam-
inants. Also, variations in the contaminant concen-
trations at the source can cause apparent dispersion
in the longitudinal direction (Frind and Hokkanen,
1987).

Apparent spreading in contaminant plumes also may
be the  result of ground-water sampling methods.
Insufficient  well purging  may result in under-
estimation of contaminant concentrations at select
locations within the aquifer. Monitoring wells with
different screen lengths integrate ground water from
disparate portions  of the aquifer and may  yield
dissimilar contaminant concentrations.

Dispersion is a phenomenon that often is used as a
mathematical convenience to correct for ignorance
about the aquifer's heterogenous nature and a poor
understanding of the processes occurring within the
aquifer. As the fundamental processes controlling
the distribution of contaminants in the subsurface
are understood, better and more cost-effective tech-
niques for aquifer remediation will be designed.
Diffusive Transport Through Low
Permeability Materials
Unfractured clays and rocks often have hydraulic
conductivities less than 1O9 m/s. A review of ground-
water flow through  such low-permeability forma-
                                           tions is provided by Neuzil (1986). However, for such
                                           materials, the diffusive transport of contaminants is
                                           large compared to advective transport.

                                           The diffusion of solutes through porous media has,
                                           been studied in the laboratory (Gillham, et al., 1984;
                                           Robin, et al., 1987). Experiments confirm the appli-
                                           cability of Fickian diffusion models and Equation 5
                                           for calculating the effective diffusion coefficient.
                                           Tortuosity factors in sand-clay mixtures were in the
                                           range of 0.59 to 0.84 for 36C1 and 0.33 to 0.70 for
                                           tritium (Gillham, 1984).

                                           A study of fine-grained Quaternary deposits in
                                           southern Ontario by Desaulniers, et al. (1981, 1986)
                                           established that  the movement  of Cl" and !8O
                                           through the deposits occurred through molecular
                                           diffusion. The transport of inorganic ions through a
                                           clay beneath a municipal landfill also was shown to
                                           be predominantly by molecular diffusion (Goodall
                                           and Quigley, 1977; Crooks and Quigley,  1984).
                                           Similarly, a study of the transport of organic contam-
                                           inants through a saturated clay beneath a hazardous
                                           waste site shows that the distribution of contam-
                                           inants is controlled by Fickian diffusion (Johnson, et
                                           al., 1989).

                                           Contaminants can contaminate aquifers by diffusing
                                           across natural aquitards or clay liners. The extent of
                                           ground-water contamination will depend on the
                                           diffusive  flux,  rate of ground-water  flow  in the
                                             12

-------
aquifer, and length of the source area in the direction
of ground-water flow.  Diffusion of organic contam-
inants is discussed in more detail in Chapter 6.  .
Effects of Density
The discussion so far has been limited to cases where
the contaminant plume does  not have a density
significantly different from the  native ground water.
In some  cases, dissolved concentrations are  large
enoiigh -that the density of the contaminant plume
may contribute to the direction of solute transport.
The contribution of density to the vertical component
of ground-water velocity, Vg, can be calculated using
the concept of equivalent freshrwater head (Frind,
1982) by:
      V  =
            •V
(6)
where KZ2 is  the  hydraulic conductivity in the
vertical direction, 0t is porosity, p is the density of
the contaminated water, and p0 is the density of the
native ground water.

As an example, assume that the density of ground
water within an  .aquifer  is l.OQ, the  natural
horizontal gradient is 0.005, and the natural vertical
gradient is 0.000. If the density of the contaminated
water is equal to the density of the native ground
water, the  contaminant plume moves horizontally
under-the naturally existing hydraulic gradient. If
the density of the contaminated water is 1.005, then
the driving force in the vertical direction is the same
as the driving force in the horizontal direction. If the
aquifer is isotropic, then the resultant vector of these
twd forces  plunges at 45° into the aquifer. Under
these conditions, a contaminant plume moves deeply
into the aquifer and may not be detected with shal-
low monitoring systems installed under the assump-
tion of horizontal  flow. The density of seawater,
which contains. 36,000 mg/L total dissolved solids, is
1.025 and the density of pure water is close to 1.000.
Therefore, the density  of  1.005  corresponds  to
approximately 7,000 mg/L.
Retardation of Contaminants
Not-all:, solutes are transported through- geologic
material at :the same rate. If solutes undergo chem-
ical reactions while being transported, their rate of
movement may be substantially less than the aver-
age' 'rate.of ground-water flow.  Such chemical
reactions include precipitation, adsorption, ion ex-
change,, and partitioning into soil organic matter or
       organic solvents. While these topics are discussed in
       more detail in Chapter 5, the  effect  of  this
       retardation on the breakthrough curves  is intro-
       duced here.

       One  simple  form of the differential  equation for
       contaminant transport with retardation is:
                a / D ac \   v  ac   ac
                              _
                dx \ R ax /   R
                                                    (7)
where  R is  a constant known as the retardation
factor and the other parameters are as defined above.
If the retardation factor is equal to 1.0, the solute is
nonreactive  and Equation 2  is obtained. If R is
greater than 1.0, the average velocity of the solute,
v/R, is less than the velocity of the ground water and
the dispersion of the solute, D/R, is likewise reduced.
If a monitoring well is located at such a distance from
a contaminant source  that it takes time ti  for a
nonreactive  solute to travel from the source  to the
well,  it will take 2ti  for a  contaminant with a
retardation factor of 2  to reach that same well and
4ti for a contaminant with a retardation factor of 4
(Figure 16).
       <
       CC
       o
       u
           1.0
           0.5
           0.0
                  R= 1
                               TIME
       Figure  16. Time required for movement of contaminants
                at different retardation factors.

       Contaminants  with lower retardation factors are
       transported greater distances  over a given time
       period than contaminants with larger retardation
       factors (Figure 17). A monitoring well network has a
       greater chance of encountering contaminants with
       low retardation factors simply because they occupy a
       greater volume of the aquifer. Thus, estimates of the
       total mass of a contaminant with a retardation factor
       of 1.0 may be more accurate than those for contam-
       inants with greater amounts of retardation. Also,
       estimates of the time to remove nonreactive contam-
       inants may, therefore, be more accurate than those
       time estimates for retarded contaminants.  This is
       particularly important  because  the slow movement
       of retarded contaminants may control .the time and
       cost necessary to completely remediate a site.
                                                 13

-------
                       RETARDATION AND MONITORING
                1,2,3      1 & 2
WASTE      DETECTED DETECTED
                                                                1 ONLY
                                                              DETECTED
Figure 17. Transport of contaminants with lower retardation factors at a waste site.
Flow and Transport in the Vadose Zone
The vadose zone is that portion of the soil between
the ground surface and the water table and includes
the capillary fringe. More generally,  the  zone is
defined as that region in which the pressure head is
less than zero. Because it  is part of the overall flow
path, this zone can be very important to the trans-
port of contaminants. Its length and the velocity of
the contaminants  passing through it, therefore,
should be included in estimates of transport times.

The vadose zone often contains greater amounts of
organic matter and metal  oxides than the saturated
zone. Contaminants can adsorb onto these materials,
making  their rate  of movement substantially less
than in the saturated  zone.  Further, materials
adhering to these adsorbents can  act as a source of
contaminants to the saturated zone even after reme-
diation.  In addition, the activity of microorganisms
in the vadose zone generally is considered to be much
greater  than below the water table. Finally,  the
unsaturated portion of the vadose zone can be a
pathway for  the transport of gases and volatile
organics. These characteristics of the vadose zone
can be important when predicting the  transport of
contaminants from a waste  site and designing
systems for remediation.
                                  Transport of Water and Solutes
                                  The flow of water through the vadose zone can be
                                  described by a differential equation that is analogous
                                  to the ground-water flow equation. The  equation's
                                  one-dimensional form is:


                                       3 /     dip \    d              dip
                                       -  K(ip) —   + - [K(ip)] = T (ip) —      (8)
                                       dz \     dz /   dz              dt

                                  where K(ip) is the hydraulic conductivity, F(ip) is the
                                  specific water capacity  (d9w/dip), 9W is the volumetric
                                  water content, and ip is the soil water pressure head.
                                  The key difference between this equation and the
                                  equation for one-dimensional ground-water flow is
                                  the dependence of the  hydraulic conductivity on ip
                                  and hence on 9W. This  makes Equation 8 nonlinear
                                  and more  difficult to solve than the  ground-water
                                  flow equation. If the hydraulic conductivity is
                                  constant and if the total head (h) is given as the  sum
                                  of the pressure head and the elevation head (z), then
                                  Equation 8 simplifies to the one-dimensional ground-
                                  water flow equation.

                                  The pressure head and the volumetric water content
                                  are related and often  are plotted as a characteristic
                                               14

-------
curve (Figure 18).  If a saturated soil with  zero
pressure head is drained, ip decreases while the
volumetric water content remains  constant and
equal to the porosity of the soil. The volumetric
water content is maintained at this level until
sufficient negative pressure (the air entry value) is
achieved to allow air to begin to enter the soil. At
this point, the volumetric water content decreases in
response to the decreasing pressure head. At low
pressure head, the curve begins to level off and
asymptotically  approaches the residual water
content of the soil.
               CHARACTERISTIC CURVE
                      Air Entry Valua
                                          ,,0.5
         MAIN
         DRAINAGE
         CURVE
      -300
                  -200          -100
            PRESSURE HEAD (CM OF WATER)
 Figure 18  Volumetric water content versus pressure head.

 If water is added to  the  drained soil, the plot  of
 volumetric water content versus pressure head does
 not follow the main drainage curve described above
 but instead follows another path known as the main
 wetting curve. Thus,  the relationship7between
 pressure head and volumetric water content is not
 unique. If a soil  is completely drained  and then
 saturated again, the volumetric water content would
 initially plot along the main drainage curve and then
 along the main wetting curve. If the  draining  or
 wetting process is interrupted before the cycle  is
 complete, the  data  plots between the two main
 curves along the primary scanning lines (Figure 19).
 If the wetting or drainage of the soil is reversed while
 the data is on the primary scanning lines, the data
 plots along yet a different path (secondary scanning
 lines). The dependence of volumetric water content
 (or any other soil property) at a given pressure head
 on the wetting and drying history of the soil is known
 as hysteresis.

 If the hydraulic  conductivity  is a function of the
 pressure head, it also must be a function of the volu-
 metric water content. For instance, the  hydraulic
 conductivity of a soil may decrease by more than two
 orders of magnitude as the volumetric water content
 decreases from saturation to residual water content.
             CHARACTERISTIC CURVE

          SCAN LINES
                                                         -300
                                                                     -200
            PRESSURE HEAD (CM OF WATER)
Figure 19. Volumetric water content versus pressure head
         showing primary scanning lines.

Solute transport in the vadose zone can be described
by an advection-dispersion equation with  an one-
dimensional form of:
                                                          -60 —
                                                               w  3z
                      a(gO
                       3z
at
            (9)
where C is the solute concentration, D is the vadose
zone dispersion coefficient, Ow is  the volumetric
water content, and q  is the volumetric water flux.
The dispersion coefficient has been assumed to be
analogous  to the dispersion term in the  saturated
zone:
         D = D t + av (0  )
               o       w
             (10)
where v(6w) is the solute velocity and is equal to
q/9w. Recent experiments by  Bond (1986) demon-
strate that Equation  10  is the correct form of the
dispersion coefficient for transport in the vadose
zone.

The application of Equations 9  and  10  to  field
situations is plagued by at least as many problems as
those discussed earlier for Equation 2. Heterogeneity
in the vadose zone may be the result of soil structure
(aggregates) or macroscopic pores such as earthworm
holes, decayed root channels,  animal burrows, and
fractures, all  of which can substantially alter the
flow of water and the transport of solutes through the
vadose zone (White, 1985). Van Genuchten and Jury
(1987) reviewed several  modeling  approaches  being
developed to investigate these situations.

Understanding the  processes that control the
movement of  water and solutes in the vadose zone
can provide  insight into  field  observations of
contaminant distributions  and can be used to design
                                                  15

-------
 storage facilities for wastes. Gillham (1984) used the
 general concepts of flow in the vadose zone to explain
 the role of the  capillary fringe in the apparently
 disproportionate rise in shallow water tables with
 small amounts of recharge. Such water table changes
 help to explain stream-flow generation (Abdul and
 Gillham, 1984). Further, such a theory suggests that
 the characteristics of the capillary fringe and  the
 vadose zone may contribute to  the spreading of
 contaminants below the water table. Finally, Frind,
 et  al. (1976) suggest that the differences in  the
 unsaturated hydraulic conductivity between coarse-
 and fine-grained materials can be utilized in waste
 storage facilities.
 Transport in the Gas Phase
 Transport of gases and contaminants through the
 unsaturated zone can be an important consideration
 in certain field situations. Some organic  contam-
 inants are volatile and can partition from the liquid
 phase into the  vapor phase. These vapors are
 transported  through the unsaturated zone  and
 eventually may diffuse into the atmosphere. The key
 physical processes that affect the  transport of gases
 in the vadose zone are diffusion and advection, with
 diffusion playing the largest role. This is the result of
 the large diffusion coefficient for gases (10-5 m2/s)
 compared to solutes (10-9 m2/s). Many  volatile
 organic chemicals have equilibrium concentrations
 that are high enough to increase  the density of the
 vapor phase to 1.5 g/cm3. This  high density, in
 principal, should  cause these vapors to sink to the
 capillary fringe. Cultural features, such as parking
 lots, streets, and foundations, can limit the exchange
 of gases with  the  atmosphere. The  transport of
 volatile organics through the soil-gas phase also will
 be affected by the partitioning of the gas phase into
 the soil water, adsorption, and biodegradation
 (Johnson and Pankow, 1987).

 There is an interest in using vapor monitoring wells
 to locate contaminant plumes in the saturated zone
 because  these  wells are much less expensive to
 install than standard ground-water monitoring
 wells. This concept  is based on  the premise that
 volatile organics within  a  contaminant  plume
 located just below the water table  will partition into
 the vapor phase in the overlying  unsaturated zone
 where they then  can be detected. In reality,  the
 transport of volatile organics  through the  un-
saturated zone  is  complicated by  the chemical and
physical processes discussed above and the general
heterogeneous nature of soils.  Vapors may be
transported along high permeability layers to points
distant from the ground-water source. Partitioning,
adsorption, and biodegradation may reduce concen-
trations of contaminants in the vapor phase to levels
 that cannot be detected. Therefore, the use of vapor
 sampling to  establish the existence of contaminant
 plumes in the subsurface is not always a reliable
 technology.

 The volatility of certain organic contaminants can be
 exploited for the purpose of remediation. If there are
 residual solvents or petroleum products in the un-.
 saturated zone, it is possible to  remove the volatile
 fraction through vapor pumping. Wells are installed
 in the unsaturated  zone  and  soil air extracted
 through a vacuum system. As the pressure drops and
 clean air passes through the soil, the organic contam-
 inants partition into the vapor phase where they are
 then extracted by the vapor pumping well. The use of
 such systems is becoming more commonplace, and
 more  efficient methods for their application are
 being developed.

 In addition to volatile organic  contaminants,  the
 transport of  the permanent gases,  such as carbon
 dioxide and oxygen, also are of great interest. Oxy-
 gen can oxidize sulfide minerals (such as  pyrite) to
 sulfuric acid, which can result in the degradation of
 ground-water quality. These problems  can  be partic-
 ularly acute in disturbed lands where mining or
 construction  has occurred. The presence of carbon
 dioxide also can alter ground-water quality by affect-
 ing mineral dissolution and  the adsorption of metal
 ions. Both  oxygen and carbon dioxide affect micro-
 bial activity  and  the rate of biodegradation in the
 vadose zone.
Contaminant Transport Through
Fractured Media
The models for solute transport discussed up to this
point only address porous media. While such models
are applicable  at sites  located on  recent alluvial
deposits and glacial sediments, they are not neces-
sarily appropriate when designing monitoring sys-
tems  or planning remedial activities at waste sites
on fractured rock.

Fractured  rock has both primary  and secondary
porosity. Primary porosity is the pore space formed
at the time of deposition and diagenesis of the rock
mass. Secondary porosity is the pore  space formed as
a result of fracture of the rock.

As in porous media, the transport mechanisms in
fractured media are advection and dispersion. In
fractured rock, however, contaminants are advected
only along the fractures.  Dispersion  phenomena
within fractured rock  is  the result of: (1) mixing at
fracture  intersections;  (2) variations  in aperture
across the width of the  fracture;  (3) variations in
aperture width along stream lines; (4) molecular dif-
fusion into microfractures penetrating the inter-
                                                16

-------
fracture  blocks; and  (5) molecular diffusion into
interfracture porous matrix blocks. Transport
through fractured media can be described by one of
four general types of models: continuum, discrete
fracture, hybrid, and channel models.

In continuum models, the individual fractures are
ignored and the entire medium is considered  to
behave  as  an equivalent porous medium.  These
models  may be either  single porosity or double
porosity models. Single porosity  models are
applicable  to fractured crystalline rocks such  as
granite and basalt where  the only porosity of the
rock mass is the fracture  porosity. Double porosity
models assume there  is both  primary porosity and
secondary porosity. These models are  applicable to
media such as sandstones and shales.

Discrete  fracture  models attempt to describe flow
and transport among individual fractures. These
models  require information  about each  fracture
within  the rock mass. The  great difficulty  of
obtaining this information led to  the development of
stochastic models, which utilize  information  about
the statistical distribution of fracture properties such
as orientation and aperture widths.

Hybrid models are combinations  of discrete fracture
and continuum models. An example of this model is
the "Multiple Interacting  Continua" (MING)  model
(Pruess and Narasimhan,  1985)  in which transport
takes place through  a three-dimensional fracture
network  (the discrete portion) while diffusion into
the  interfracture rock matrix (the continuum)
occurs.

Channel models  (Tsang  and Tsang,  1987) were
developed from both laboratory (Witherspoon, et al.,
1980) and field (Neretnieks, 1985) observations that
the transport  of solutes along fractures  does not
occur as a uniform  front along the width of a
fracture, but in many small fingers or channels. Such
models are only now being investigated.

Studies of fractured rock often make use of the cubic
law for the description  of fluid flow and hydraulic
conductivity.  The  hydraulic conductivity of a
fracture, Kf, is:
          K_ = (2brPg/(12u)
(ID
where 2b is the fracture aperture, p is the density of
the fluid, g is the acceleration of gravity, and u is the
dynamic viscosity of the fluid. The above equation
can be  modified for application in a continuum
model:
             = (2brpgN/(12p.B)
(12)
where B is the thickness of the medium and N is the
number of fractures  through that thickness. An
important consideration in the study of fractured
rock is knowing when a fracture  network can be
considered to behave like an equivalent porous
media. Numerical studies of single-porosity fracture
networks suggest that such networks behave more
like continua when: (1) fracture density is increased;
(2) apertures are constant rather than statistically
distributed; (3) the orientations of the fractures are
statistically distributed rather than constant; and (4)
larger sample sizes are tested (Long, et al., 1982).

Novakowski,  et al. (1985) conducted a tracer test in a
single fracture at about 100 m depth in gneiss at the
Chalk River Nuclear  Laboratories in  Ontario,
Canada. The  fracture aperture of 510 um obtained
from their two-well tracer test does not agree well
with the 60 um aperture obtained from interference
pumping tests. A survey of dispersivities in fissured
rock obtained at various sites in  Europe and in the
United States reveals a  range  of four orders of
magnitude  (Neretnieks, 1985). Novakowski, et al.
(1985) obtained a dispersivity of 1.4 m. Tracer tests
at the Oracle, Arizona site are within the range from
several tenths of a  meter to a  few meters (Cullen, et
al., 1985).  As in porous  media,  the dispersivity
values obtained  for fissured  media become larger
with increasing length of the flow path (Neretnieks,
1985).

Research into transport phenomena in fractured rock
has been supported primarily  by agencies interested
in the potential for transport of radionuclides from
high-level  radioactive waste repositories.  As a
consequence,  the bulk of the research performed has
been of fractured crystalline  rock such as granite.
However, many hazardous waste sites are located on
fractured sedimentary rocks such as sandstone  and
shale which  have primary porosities of 5 to 25
percent in the rock matrix. Fractures increase the
total porosity of these rocks and substantially
increase their hydraulic conductivity, making them
attractive for water  supply  aquifers. While  con-
taminants  are advected  through  these  fracture
systems, transport into and out of the porous matrix
is primarily by molecular diffusion (Figure 20). This
latter phenomena is much more important in
fractured, porous rock than in fractured, crystalline
rock.

Fractured, porous aquifers can be found throughout
the conterminous United States. The 20 states with
the greatest potential for ground-water  contamina-
tion in fractured,  porous  media (FPM) are found
primarily in  the northeastern United  States  and
around the Great Lakes (Figure 21). More than 1,500
Comprehensive Environmental Response  and Lia-
bility Act (CERCLA) or Superfund sites located on
                                                 17

-------
             FRACTURED POROUS ROCK
                              Diffusion
                              intp Rock
                              Matrix
   Diffusion
   Into Rock
   Matrix
the rate of movement of the  water through the
fractures (vf) by:
                = vf/Rf
(13)
                                                    where Rf is a retardation factor that accounts for the
                                                    loss of contaminant mass from the fracture to the
                                                    matrix:
            Rf = (b + nB'Vb
(14)
 Figure 20.  Transport in fractured porous rock.

FPM exist within  these  20 states (Johnson  and
Pankow, 1987).

The use of porous media models can be appropriate in
evaluating fractured, porous  formations  if the
concentration of the contaminant can quickly reach
equilibrium with  the concentration  found in the
fracture (Pankow,  et al., 1986).  The  rate of
movement of a conservative tracer through such a
fractured, porous aquifer (vs) can be calculated from
where b is the half-width of the fracture and B' is the
half-width of the matrix block.

These concepts have been applied to a site at Alkali
Lake, Oregon, and at Bayview  Park, Ontario
(Pankow, et al., 1986). At Alkali Lake, the sediments
are highly fractured, the matrix blocks are small (0.3
cm), and  the matrix diffusion coefficient is 0.1
cm2/day. In contrast, at Bayview Park, the matrix
blocks are 5 to 35 cm  and the matrix diffusion
coefficients are only 0.0032 cm2/day. At Alkali Lake,
the concentrations in  the matrix approach equi-
librium with the concentrations in the fractures in
                   o
 Figure 21.  States with a high potential for ground-water contamination in fractured, porous rock (bold line outlines EPA Re-
           gions).
                                                 18

-------
about 5 days, while more  than 6,000 days  are
required at Bayview Park. As a result, the equiva-
lent porous media model works well at Alkali Lake
but not at Bayview Park. Thus, simple calculations
of the time needed for the matrix block to reach
equilibrium within the fracture  can be used  as  a
guide for the applicability of the equivalent porous
media model for fractured, porous media.
Particle Transport Through Porous
Media
So far, this chapter has only considered the transport
of solutes through porous and  fractured media.
However, particles  also  may be of interest ,to
contaminant hydrologists. The term "particle" can be
used very broadly to include  bacteria, viruses,
inorganic precipitates, natural  organic  matter,
asbestos fibers, or clay.

Particles can be removed  from  solution by  three
major processes: (l) surface filtration; (2) straining;
and (3) physical-chemical processes (McDowell-
Boyer,  et al., 1986).  Determining which of these
processes is the most effective depends on the size of
the particles (Figure 22). If the particles are larger
than the  largest pore diameters,  they cannot
penetrate into the porous medium and will be
filtered at the surface of the medium. If the particles
are smaller than the  largest pores but  larger than
the smallest,  the particles  are transported into the
porous medium along the larger pore channels.
Eventually, the particles encounter a pore channel
with a smaller diameter and are removed by
straining. If  the particles are smaller than the
smallest pore  openings, the particles can be
transported great distances through  the porous
medium.

The rate at which the particles move through the
porous medium depends on a variety  of physico-
chemical processes. For example, the particles may
undergo random collisions with sand grains. A
certain percentage  of such collisions result in
particles adhering to the solid matrix (interception).
Chemical conditions also may affect particle trans-
port. For example, if the  pH changes, aggregates
may result due to changes in the particles' surface
properties. These larger aggregates then can be
strained or filtered from the water.

Microorganisms are particles that can be transported
through geologic media. The movement of bacteria
and  viruses  in the  subsurface is  a  significant
problem. More than 50,000 individuals in the United
States  suffered from waterborne disease between
1971 and 1979 (Craun, 1984) and about 45 percent of
all reported cases involved ground-water sources. In
addition, increasing interest  in the use of
  SURFACE

  FILTRATION
 0O°O0O°
 o^o^o^o
   0O°O°O0
   o^o^o^o
 0°0°
 STRAINING
cgogogog
 O^~\;- ^^O°^~^OQ^~\
                 oP
                        o
                        o
 PHYSICAL-

 CHEMICAL
O
Figure 22.  Filtration mechanisms.
microorganisms for in situ remediation of aquifers
contaminated with organic chemicals necessitates a
greater understanding of the transport and  fate of
microorganisms within the subsurface.

There are many processes that limit the movement of
microorganisms through geologic  materials. Some
bacteria are large enough to be strained from the
water. In comparison, the  smaller viruses can pass
through the pores, but their surfaces  are charged
and, like charged ions, may undergo adsorption
under the proper chemical conditions. Like mole-
cules, microorganisms are transported by diffusion.
                                               19

-------
Some  microorganisms are motile  and move in
response to changes in chemical concentrations. Like
other living organisms, microbes grow and die, and
the rates of these processes must be included in the
description of the transport of microbes in the
subsurface.  All of these processes are reviewed in
detail by Yates, et al.  (1987)  and Matthess and
Pekdeger (1981).

Field tracer tests using baker's yeast (Saccharomyces
cerevisiae) were conducted at a field site in Stanton,
Texas  (Wood and Ehrlich, 1978). The baker's yeast
was injected into a sand and gravel  aquifer
containing clay lenses. In two separate tests where
bromine and iodine (Br" and D ions  were used as
chemical tracers, the  baker's yeast arrived at the
monitoring well  before the chemical  tracers. Wood
and  Ehrlich (1978) explain these observations  by
suggesting  that: (1)  the chemical  tracers were
retarded because of their adsorption onto the aquifer
material; and (2) the  yeast traveled  only through
solution channels in  caliche deposits within the
aquifer while  the chemical tracers flowed through
both the solution channels and the intergranular
pore structure. As  the chemical tracer  moves
through the larger solution pores, it loses mass to the
smaller pores by molecular diffusion, thereby
retarding the rate of movement of the chemical front.
The  larger yeast particles have smaller diffusion
coefficients and are excluded from the smaller pores.
Without the mass loss to the adjacent pores, the
yeast arrives more quickly at the observation well.

Champ and Schroeter (1988) used field tracer tests to
show that non-reactive particle tracers and bacteria
(Escherichia coli) can be rapidly transported through
fractured, crystalline rock. Similar  to the results of
Wood  and Ehrlich (1978), Champ and  Schroeter
(1988)  observed that the E. coli arrived before the
bromine tracer. Tracer experiments  using native
bacteria and different sizes and types  of  micro-
spheres were conducted  at a  Cape  Cod site  in
Massachusetts (Harvey,  et al.,  1987). The results
indicate that both size  and surface characteristics of
particles affect their movement through the aquifer.
Field evidence of the transport of inorganic particles
in the subsurface was obtained at the Otis Air Force
Base site on Cape Cod in Massachusetts (Gschwend
and Reynolds, 1987).  Secondarily  treated sewage
containing phosphates was recharged  to a sand and
gravel  aquifer through rapid infiltration beds over a
30-year period. Downgradient from the rapid
infiltration beds, ground-water samples are found to
contain 100 nm-diameter particles. These particles,
composed of phosphate and iron, may be the mineral
vivianite. The phosphate comes from the recharged
water while  the iron is derived from the dissolution
of naturally existing iron within the aquifer.
Another example of particle transport exists at the
Nevada Test  Site where particle  transport was
identified as a mechanism for  the movement of
lanthanide and transition element isotopes from
subsurface nuclear explosion cavities (Buddemeier,
1986).
References
Abdul and Gillham, 1984. "Laboratory Studies of the
  Effects of the Capillary Fringe on Streamflow
  Generation." Water Resources Research, Vol. 20,
  No. 6, pp. 691-698.
Bear, J., 1969. "Hydrodynamic Dispersion." In: Flow
  Through  Porous Media,  R.J.M. De  Wiest, Editor.
  Academic Press, New York, NY, pp.  109-199.

Bear, J., 1979. Hydraulics ofGroundwater. McGraw-
  Hill, New York, NY.

Bond, W. J., 1986. "Velocity-dependent  Hydrody-
  namic Dispersion During Unsteady, Unsaturated
  Soil Water Flow: Experiments." Water Resources
  Research, Vol. 22, No. 13, pp. 1881-1889.
Buddemeier, R. W., 1986. "Ground-water Transport
  of Radionuclides in Colloidal Form." EOS, Trans-
  actions of the American Geophysical Union, Vol. 67,
  No. 44, p. 955.
Champ, D. R. and J.  Schroeter,  1988. "Bacterial
  Transport in Fractured Rock: A Field Scale Tracer
  Test at the Chalk River Nuclear Laboratories."
  Proceedings of the International Conference on
  Waters  and Waste Water Microbiology,  Inter-
  national Association of Water Pollution Research
  and Control, Irvine, CA, Feb. 8-11, 1988, pp. 14-1 -
  14-7.
Craun, G. F., 1984. "Health Aspects of Ground-water
  Pollution." In: Ground-water Pollution Micro-
  biology, G.  Bitton and C.P.  Gerba,  Editors. John
  Wiley and Sons, New York, NY, pp. 135-179.
Crooks, V. E. and R. M. Quigley, 1984. "Saline
  Leachate  Migration Through Clay: A Comparative
  Laboratory and Field Investigation." Canadian
  GeotechnicalJournal, Vol. 21, pp. 349-362.

Cullen, J. J., K. J. Stetzenbach, and E. S. Simpson,
  1985. "Field Studies of Solute Transport in Frac-
  tured Crystalline Rocks Near Oracle, Arizona." In:
  Proceedings, Memoirs of the  17th International
  Congress of the International  Association  of
  Hydrologists, Vol. 17, International  Association  of
  Hydrologists, Tucson, AZ, pp. 332-344.

Desaulniers, D. E., J. A. Cherry, and P. Fritz, 1981.
  "Origin,  Age, and Movement of Pore Water  in
  Argillaceous Quaternary Deposits at Four Sites  in
  Southwestern Ontario." Journal of Hydrology, Vol.
  50, pp. 231-257.
                                               20

-------
Desaulniers, D. E., etal., 1986. "37C1-35C1 Variations
 in a Diffusion Controlled Ground Water System."
 Geochimica et Cosmochimica Acta, Vol.  50, pp.
 1757-1764.
Freeze, R. A. and J. A. Cherry, 1979. Ground-water.
 Prentice-Hall, Inc., Englewood Cliffs, NJ.
Freyberg, D. L., 1986. "A Natural Gradient Experi-
 ment on Solute Transport in  a  Sand Aquifer.  2.
, Spatial Moments  and the Advection and Dis-
 persion of Nonreactive Tracers." Water Resources
 Research, Vol. 22, No. 13, pp. 2031- 2046.
Frind, E. O., R. W. Gillham, and J. F. Pickens, 1976.
 "Application of Unsaturated Flow Properties in the
 Design of Geologic  Environments for Radioactive
 Waste Storage Facilities."  In: Proceedings, First
 International Conference on  Finite Elements  in
 Water Resources, Princeton University, July 1976.
Frind, E. O., 1982. "Simulation of Long-term Tran-
 sient Density-dependent Transport in Ground-
 water." Advances in Water Resources, Vol. 5, pp.
 73-88.
Frind, E. O. and G. E. Hokkanen, 1987. "Simulation
 of the  Borden Plume Using the  Alternating
 Direction Galerkin Technique." Water Resources
 Research, Vol. 23, No. 5, pp. 918-930.
Gelhar, L. W. and  C. L.  Axness, 1983. "Three-
 Dimensional Stochastic  Analysis of Macrodis-
 persion in  a Stratified Aquifer."  Water Resources
 Research, Vol. 15, pp. 1387-1397.
Gillham, R.  W. and J. A. Cherry, 1982. "Contam-
 inant  Migration in Saturated  Unconsolidated
 Geologic  Deposits." In: Recent Trends  in
 Hydrogeology, T. N. Narasimhan, Editor.  Geolog-
 ical Society of America Special Paper 189, pp. 31-
 62.
Gillham, R. A., 1984. "The Capillary Fringe and Its
 Effect  on  Water-table  Response." Journal  of
 Hydrology, Vol. 67, pp. 307-324.
Gillham, R. W., et al., 1984. "An Advection-diffusion
 Concept for Solute Transport in Heterogeneous
 Unconsolidated  Geologic  Deposits."  Water
 Resources Research, Vol. 20 (3), pp. 369-378.
Gillham, R.  W.,  M.  J.  L. Robin, and D.  J.
 Dytynyshyn, 1984.  "Diffusion of Nonreactive and
 Reactive Solutes Through  Fine-grained  Barrier
 Materials." Canadian Geotechnical Journal, Vol.
 21, pp. 541-550.
Goodall, D. C. and R. M. Quigley, 1977. "Pollutant
 Migration  for Two Sanitary Landfill Sites Near
 Sarnia, Ontario." Canadian Geotechnical Journal,
 Vol. 14, pp. 223-236.
Gschwend, P. M. and M. D. Reynolds, 1987. "Mono-
 disperse Ferrous Phosphate Colloids in an Anoxic
  Groundwater Plume." Journal of Contaminant
  Hydrology, Vol. 1, No. 3, pp. 309-327.
Harvey, R. W., et al.,  1987. "Transport of Bacteria
  Through A Contaminated Freshwater Aquifer." In:
  U.S. Geological Survey Program on Toxic Waste-
  Ground-Water Contamination: Proceedings of the
  Third Technical Meeting, Pensacola, FL, March
  23-27, 1987. U.S. Geological Survey Open-File
  Report 87-109, pp. B29-B31.
Johnson, R. L.,  et al.,  1987. "Mass Transfer of
  Organics Between Soil, Water, and Vapor Phases:
  Implications  for Monitoring, Biodegradation, and
  Remediation." Proceedings of the Petroleum-
  Hydrocarbons and Organic Chemicals  in Ground
  Water, Prevention, Detection,  and Restoration
  Conference  and Exposition, Houston,  TX.
  November 4-6, 1987.
Johnson,  R.  L.  and  J. F.  Pankow,  1987.
  "Contaminant  Transport in Fractured Porous
  Rock; A  Survey of Existing Ground-water
  Contamination Sites and a Summary of Existing
  Research  Tools." Final Report, Cooperative
  Research Agreement CR812553-010,  R.S.  Kerr
  Environmental Research Laboratory,  and U.S.
  Environmental Protection Agency.
Kimmel, G.  E. and O. C. Braids, 1980.  "Leachate
  Plumes in Groundwater  from  Babylon and Islip
  Landfills, Long Island, New York." U.S. Geological
  Survey, Professional Paper 1085.
Long, J. C. S.,  et  al.,  1982.  "Porous Media
  Equivalents of Discontinuous  Fractures."  Water
  Resources Research, Vol. 18, No. 3, pp. 645-658.
MacFarland, D. S., et al., 1983. "Migration of Con-
  taminants in Groundwater at a Landfill: A Case
  Study;  1. Groundwater  Flow  and Plume
  Delineation." Journal of Hydrology, Vol. 63, pp. 1-
  29.
Matthess, G. and A. Pekdeger, 1981. "Concepts of a
  Survival  and Transport Model of Pathogenic
  Bacteria and Viruses in Groundwater." Science of
  the Total Environment, Vol. 21, pp. 149-159.
McDowell-Boyer, L.M.,  J.R. Hunt,  and N.  Sitar,
  1986.  "Particle Transport Through Porous Media."
  Water Resources Research, Vol. 22, No. 13, pp.
  1901-1921.
Mercado,  A.,  1967.  "The Spreading Pattern of
  Injected  Water  in  a Permeability  Stratified
  Aquifer." Proceedings  of the  Symposium on
  Artificial Recharge and Management of Aquifers,
  International Association of Scientific  Hydrology,
  Vol. 72, pp. 23-36.
Molz, F. J., et al., 1982. "Performance, Analysis, and
  Simulation of a Two-well Tracer Test at the Mobile
  Site." Water Resources Research, Vol. 22, pp. 1031-
  1037.
                                               21

-------
Neretnieks, I., 1985. "Transport in Fractured Rocks."
  In: Proceedings, Memoirs of the 17th International
  Congress of the  International  Association of
  Hydrotogists, Vol. 17, International Association of
  Hydrologists, Tucson, AZ, 1985, pp. 301-308.
Neuman, S. P., et al., 1985. "Statistical Analysis of
  Hydraulic Test Data From Fractured Crystalline
  Rock  Near Oracle, Arizona." In: Proceedings,
  Memoirs of the 17th International Congress of the
  International Association of Hydrologists, Vol.  17,
  International Association of Hydrologists, Tucson,
  AZ, pp. 289-300.
Neuzil,  C.  E.,  1986. "Groundwater Flow in  Low-
  permeability Environments." Water Resources
  Research, Vol. 22, pp. 1163-1195.
Novakowski, et al., 1985. "A  Field Example of
  Measuring Hydrodynamic Dispersion in a Single
  Fracture." Water Resources Research, Vol. 21,  pp.
  1165-1174.
Pankow, J. F.,  et al., 1986. "An  Evaluation of
  Contaminant Migration Patterns at Two Waste
  Disposal Sites on Fractured Porous Media in Terms
  of the Equivalent Porous Medium (EPM) Model."
  Journal of Contaminant Hydrology, Vol.  1, pp. 65-
  76.
Perkins, T. K.  and O. C. Johnston, 1963. "A Review
  of Diffusion  and Dispersion in Porous Media."
  Journal of the Society of Petroleum Engineers, Vol.
  17, pp. 70-83.
Pickens, J. F.  and G. E. Grisak, 1981. "Scale-
  dependent Dispersion in a Stratified Aquifer."
  Water Resources Research, Vol. 17, pp. 1191-1212.
Pruess, K. and T. N. Narasimhan, 1985. "A Practical
  Method for Modeling Fluid and  Heat Flow in
  Fractured Porous Media." Society of Petroleum
  Engineers Journal, Vol. 25, No. 1, pp. 14-26.
Robin, M. J. L., R. W. Gillham, and D. W. Oscarson,
  1987. "Diffusion of Strontium and Chloride in
  Compacted Clay-based Materials." Soil Science
  Society of America Journal, Vol. 51, pp.  1102-1108.
Sudicky, E. A.,  1986. "A  Natural Gradient
  Experiment in a Sand Aquifer: Spatial Variability
  of  Hydraulic  Conductivity  and  Its  Role  in
  the Dispersion Process." Water Resources Research,
  Vol. 22, No. 13, pp. 2069-2082.
Sudicky, E. A., R. W. Gillham, and E. O. Frind, 1985.
  "Experimental Investigation of Solute Transport in
  Stratified Porous Media: 1. The Nonreactive Case."
  Water Resources Research, Vol. 21, No. 7, pp. 1035-
  1041.
Tsang, Y. W. and C. F. Tsang, 1987. "Channel Model
  of Flow  Through  Fractured  Media."  Water
  Resources Research, Vol. 23, No. 3, pp. 467-479.
Van Genuchten, M.  Th. and W.  A. Jury,  1987.
  "Progress  in  Unsaturated Flow and  Transport
  Modeling." Reviews of Geophysics, Vol. 25, No. 2,
  pp. 135-140.
White, R. E.,  1985.  "The Influence of Macropores on
  the  Transport of Dissolved Matter Through Soil."
  Advances in Soil Science, Vol. 3, pp. 95-120.
Witherspoon, P. A., et al.,  1980. "Validity of Cubic
  Law for Fluid  Flow in a Deformable  Rock
  Fracture." Water Resources Research, Vol. 16, No.
  6, pp. 1016-1024.
Wood, W. W. and G. G. Ehrlich, 1978. "Use of Baker's
  Yeast to Trace Microbial Movement  in Ground
  Water." Groundwater, Vol. 16, No. 6, pp. 398-403.
Yates, M. V., et al., 1987. "Modeling Virus Survival
  and  Transport in the Subsurface." Journal of
  Contaminant Hydrology, Vol. 1, pp. 329-345.
                                               22

-------
                                       CHAPTER 3

        PHYSICAL PROCESSES CONTROLLING THE TRANSPORT OF
            NON-AQUEOUS  PHASE LIQUIDS IN THE SUBSURFACE

                              Carl D. Palmer and Richard L Johnson
Introduction
Liquids that do not readily dissolve in water and can
exist as a separate fluid phase are known as  non-
aqueous phase liquids (NAPLs). Generally, NAPLs
are subdivided into two classes: those that are lighter
than water (LNAPLs); and those with a density
greater than water (DNAPLs). Most  LNAPLs are
hydrocarbon fuels  such as gasoline, heating oil,
kerosene, jet fuel, and aviation gas. Most DNAPLs
are chlorinated hydrocarbons such  as 1,1,1-
trichloroethane, carbon tetrachloride, chlorophenols,
chlorobenzenes, tetrachloroethylene, and PCBs.

Concern about NAPLs exists  because of  their
persistence in the subsurface and their ability  to
contaminate large volumes of water. For example, 7
L (10 kg) of trichloroethylene (TCE) can contaminate
108 L of ground  water  at 100 ppb (Feenstra and
Cherry, 1987). NAPLs are ubiquitous throughout
North America and have been identified at 4 out of 5
hazardous waste  sites in the United States (Plumb
and Pitchford, 1985). Greater understanding of the
transport and dissolution of NAPLs is necessary to
implement cost-effective techniques for the cleanup
of these contaminants.
Transport and Dissolution of NAPLs
As NAPLs  move through geologic media, they can
displace water and air. Because water is the wetting
phase relative to both air and NAPLs, it tends to line
the edges of the pores and cover the sand grains. The
NAPL is the non-wetting phase and tends to move
through the central portions of the pores. Neither the
water nor the NAPL phase occupies the entire pore.
Because of this, the permeability of the medium with
respect to these fluids is different than when the pore
space is entirely occupied by a given phase. This
reduction in permeability depends on the medium
and  often  is  described in terms of relative
permeability, krj, for phase i, which is defined as:
            k . = k.(S.)/k .
             ri    i  i  si
(D
where Si is the fraction of pore space occupied by
phase i, kj{Sj) is the permeability of the medium to
phase i at saturation S;, and ksi is the permeability of
the medium at complete saturation with phase i.
Thus, the relative permeability varies from 1.0  at
100 percent saturation to 0.0 at 0 percent saturation.

A  plot of relative permeability versus water
saturation for a hypothetical medium (Figure 23)
reveals some important features about multiphase
flow. At 100 percent water saturation, the relative
permeability of the water and the NAPL are 1.0 and
0.0, respectively. As the  fraction of the pore space
occupied by the NAPL  (Sn) increases, a corre-
sponding decrease occurs in the fraction of water
within the pore space (Sw). As Sw decreases, the
relative permeability with respect to the water phase
decreases to zero. Zero relative permeability is not
obtained at zero  Sw, but at the irreducible water
saturation (Srw). At this water saturation, the water
phase  is effectively immobile and  there is no
significant flow of water. These concepts are similar
to those discussed for unsaturated flow in  Chapter 2.
The relative permeability of the NAPL behaves in a
similar manner.  At 100 percent NAPL saturation,
the relative permeability for the NAPL is equal to
1.0, but as the NAPL saturation decreases, so does
the relative permeability. At  the  residual NAPL
saturation (Srn),  the relative permeability for the
                                              23

-------
       100%
    NAPL SATURATION
            Irreducible
         -  Water
            Saturation
                       Srw
              WATER SATURATION
                                100%
Figure 23.
Relative permeability as a function of satura-
tion.
NAPL is effectively zero and the NAPL is considered
to be immobile.

These immobile fractions of  NAPL cause great
concern because they cannot be easily removed from
the pores except by simple dissolution by flowing
ground water. An example of this could be  a cubic
meter of soil with  a 35 percent porosity and
containing TCE at a 20 percent residual saturation.
This situation implies that there is 0.07 m3 or 103 kg
of TCE within the soil. If the solubility of the TCE is
1,100 mg/L and if ground water  flows through the
soil at a rate of 1.7 cm/day, it would take 15.4 years
for the TCE to  be removed by dissolution. If the
contaminated aquifer is twice  as long (2 m), 30.8
years are necessary. If a NAPL  with a lower
solubility than TCE  is spilled,  the rate of mass
removal is lower,  requiring  even more  time for
dissolution. Thus, NAPLs that  enter the subsurface
can remain for decades and can  contaminate large
volumes of ground water. Some of the key aspects of
NAPL dissolution are considered in more detail in
Chapter 6.

Understanding how an NAPL moves within a porous
aquifer  can be useful. The movement of petroleum
products in the  subsurface is described in detail by
Schwille (1967), and  the movement  of DNAPLs is
examined closely by Schwille  (1988), Feenstra and
Cherry (1987), Kueper and McWorter (1988), and
Anderson (1988).
Light Non-Aqueous Phase Liquids
(LNAPLs)
As a spilled LNAPL enters the unsaturated zone, it
flows through the central portion of the unsaturated
pores. If the amount of product released is small, the
product flows until residual  saturation is reached
(Figure 24A). Therefore, a  three-phase system
consisting of water, product, and air is created within
the vadose zone.  Infiltrating water dissolves the
components  within the LNAPL (e.g.',' benzene,
toluene, and xylene) and Carries them  to the water
table. These  dissolved constituents  then form  a
contaminant plume emanating from the area of the
residual product. Many of the components commonly
found in LNAPLs are volatile and, as a consequence,
can partition  into  the soil air and be transported by
molecular diffusion to other portions of the aquifer.
As these vapors diffuse into adjacent soil areas, they
partition back into the water phase  and spread
contamination over a wider area. If the surface is not
covered with an impermeable material, these vapors
diffuse, across the surface boundary and into the
atmosphere. However, if'a relatively impermeable
boundary covers the area, np,,mass transfer occurs
with the atmosphere and the'Concentrations of
contaminants in the soil atmosphere may build up to
equilibrium concentrations;

If large volumes of product are spilled (Figure 24B),
the product flows through the pore space to the top of
the capillary fringe. The dissolved components of the
infiltrating product precede the product and may
change the wetting properties of the water, causing a
reduction in the residual water, Content and collapse
of the capillary fringe.         ;

The LNAPL product is lighter  than water and tends
to float on  top of  the capillary fringe^ AS  the he,ad
created by  the infiltrating /prodiict increases, the
water table is depressed and the product begins to
accumulate in the depression. If the source of the
spilled product is then turned off, the LNAPL within
the vadose zone continues to flow under the influence
of gravity until reaching residual saturation. As this
drained product continues to recharge the  product
pool, it spreads laterally on the top of the  capillary
fringe (Figure 24C).  The draining, of the upper
portions of the vadose zone also reduces the total
head at the interface between the product  and the
ground water, causing the water table to rebound
slightly. The  rebounding water can only displace  a
portion of the product because the latter remains at
residual saturation. Ground water passing  through
this area of residual saturation dissolves the com-
ponents within the residual  product,  creating  a
contaminant  plume. Water infiltrating from  tne
surface also can dissolve  the residual  product and
vapors within the  vadose zone, thereby contributing
to the overall contaminant load to the aquifer.

If the water table drops because , of 'seasonal
variations or pumping, the pool of product also drops.
If the water table  rises again,  part of the product is
pushed upward, but a portion remains at residual
saturation below the new water table,. Thus,
variations in the water table can spread'the product
                                                24

-------
                    PRODUCT SOURCE

                     ttttt
 GROUNDWATER
                            GROUNDWATER     WATERTABLE
                    PRODUCT SOURCE

                     Httt
OROUNDWATER
                             GROUNDWATER
                                        WATERTABLE
Figure 24.   Movement of LNAPLS into the subsurface:
           (a) distribution of LNAPL after small volume
           has been spilled;  (b) depression of the
           capillary fringe  and water  table; (c)
           rebounding of the  water-table as LNAPL
           drains from overlying pore space.

over a greater thickness of the aquifer, causing
increased volumes of soil to be contaminated. Clean-
up methods for LNAPLs in  the subsurface  should
take this principle into account and avoid moving the
product into uncontaminated  areas where more
product can be held at residual saturation.
Dense Non-Aqueous Phase Liquids
(DNAPLs)
DNAPLs can have great mobility in the subsurface
as a result of their relatively low solubility,  high
density, and low viscosity. The sparingly soluble
DNAPLs do not readily mix with water and therefore
remain as separate phases. The relatively  high
density of these liquids provides a driving force that
can carry product deep  into aquifers.  The
combination of this high density and low viscosity is
particularly important with regard to the transport
of DNAPLs in the subsurface. When a high density,
low viscosity fluid  (DNAPL) displaces a lower
density, higher viscosity fluid (water), the flow is
"unstable" and viscous fingering occurs (Saffman
and Taylor, 1958; Homsy, 1987; Chouke, et al., 1959;
Kueper and Frind,  1988).

During a spill (Figure 25A),  DNAPL flows  through
the unsaturated zone  under the influence of gravity
toward the water table. If only a small amount of
DNAPL is spilled, it  flows until reaching residual
saturation in the  vadose  zone. If there is water
within the unsaturated zone, the DNAPL exhibits
viscous fingering  during infiltration. No  viscous
fingering is observed  if the vadose zone  is dry. The
DNAPL can partition  into the vapor phase and these
dense  vapors may sink to the capillary fringe.
Infiltrating water can dissolve the residual  DNAPL
or the  vapors and  transport these contaminants to
the water table, creating a dissolved chemical plume
within the aquifer.

If a greater amount of DNAPL is spilled  (Figure
25B), the DNAPL flows until it reaches the capillary
fringe  and, once there, begins to penetrate  into the
aquifer. However, to do  this, the DNAPL must
displace the water by overcoming the capillary forces
between the  water and the medium. The  critical
height of  DNAPL required to  overcome these
capillary forces (zc) can be calculated from:
                                                            = 2y cos (9) (l/rt - 1/r )/ (Ap g)
                                            (2)
where y is the interfacial tension between the water
and the DNAPL, 9 is the contact angle between the
fluid boundary and the solid surface, rt is the radius
of the pore throat, rp is the radius of the pore, Ap is
the difference in the density between the water and
the DNAPL, and g is the acceleration of gravity
(Villaume, et al., 1983). As an example, calculated
critical  heights' required for perchloroethylene to
penetrate saturated porous media of different grain
size range from a few centimeters for coarse grains to
tens of meters for clays (Table 2). Thus, unfractured,
saturated clays and silts can be effective barriers to
the migration  of DNAPLs, provided the  critical
heights are not exceeded-

After penetrating the aquifer, the DNAPL continues
to move through the saturated zone until it reaches
residual saturation. The DNAPL then is dissolved by
ground  water passing through the contaminated
area, resulting in  a contaminant plume  that can
extend over a great thickness of the aquifer^ If finer-
grained strata are contained within the aquifer, the
                                                25

-------
 Table 2.   Critical Height for Perchloroethylene  to
          Penetrate Water Saturated  Media  (Anderson,
          1988).
                      DNAPL SOURCE

                      tttttt
Material
Coarse Sand
Fine Sand
Silt
Clay
Diameter
1.0
0.1
0.01
0.001
Critical
Height (cm)1
13
130
1,300
13,000
 'Calculated tor Ap » 0.62 g/cm^.
 Y ™ 47.5 dynes/cm, cos 9 = 1.

infiltrating DNAPL accumulates  on top of the
material, creating a pool. At the interface between
the ground water and the DNAPL pool, the solvent
dissolves into the  water and spreads vertically by
molecular diffusion. As water flows by the DNAPL
pool, the concentration of the contaminants with the
ground water increases until saturation is achieved
or the downgradient edge of the pool is reached. The
relative  density of pools and fingers  of DNAPL
within the aquifer  is important when controlling the
measured concentrations of dissolved contaminants
derived from DNAPLs. The existence of fingers and
pools of the DNAPL, rather than  relatively
continuous distributions, in the subsurface accounts
for the observation that the concentration of many of
the DNAPL compounds in ground water are far
below their saturation limit (Anderson, 1988).

If even larger amounts of DNAPL are spilled (Figure
25C), the DNAPL can, in principle, penetrate to the
bottom of the aquifer, forming pools in depressions. If
the impermeable boundary is sloping, the DNAPL
flows down the dip of the boundary. This direction
can be upgradient  from the original spill area if the
impermeable boundary slopes in that direction. The
DNAPL also can flow along bedrock troughs, which
may be  oriented differently from the  general
direction of ground-water flow. This flow along low
permeability boundaries can spread contamination
in directions that would not be predicted on the basis
ofhydraulics.

The transport of  DNAPLs in physical models  of
fractures illustrates the importance of fracture
aperture and roughness. Schwille (1988) found  that
if fracture apertures are greater than 0.2 to 0.5 mm,
the DNAPL moves directly  to the capillary fringe
where it spreads out. Eventually the  DNAPL
penetrates the capillary fringe and is transported  to
the bottom of the aquifer. There is very little residual
DNAPL for fractures  larger than 0.2  mm.  If the
fracture aperture is less than 0.2 mm and is smooth,
the DNAPL spreads out near the surface and a few
fingers migrate down to the capillary fringe. At the
capillary fringe, the DNAPL spreads out further and
a few relatively wide fingers  penetrate below the
B
DNAPL SOURCE

tttttt
Figure 25.   Movement of DNAPLs into the subsurface:
           (a) distribution of DNAPL after small volume
           has been spilled; (b) distribution of DNAPL
           after moderate volume has been spilled; (c)
           distribution of DNAPL after large volume has
           been spilled (After Feenstra and  Cherry,
           1988).
capillary fringe. If the  fracture is rough,  there is a
great  amount  of fingering  and  the  DNAPL
penetrates below  the  capillary  fringe  in small,
scattered fingers. Thus, for fractures with apertures
less  than  0.2 mm, there can be a large volume of
DNAPL that remains at residual saturation in the
fractures both above and below the capillary fringe.
While  similar behavior is expected  to occur in
fractured rock, the statistical distribution of fracture
                                                 26

-------
aperture and fracture roughness may preclude the
use of such a simple categorization based on the
mean apertures.


References
Anderson, M. R., 1988. "The Dissolution and
  Transport of Dense Non-aqueous Phase Liquids in
  Saturated Porous  Media." Ph.D Dissertation.
  Department of Environmental Science and
  Engineering, Oregon Graduate Center, Beaverton,
  OR.
Chouke, R. L., P. van Meurs, and C. van der Poel,
  1959. "The Instability of Slow, Immiscible, Viscous
  Liquid-liquid Displacements in Permeable Media."
  Transactions AIME, Vol. 216, pp. 188-194.
Feenstra, S. and J. A. Cherry, 1987.  "Dense Organic
  Solvents in Ground Water: An Introduction." In:
  Dense Chlorinated Solvents in  Ground Water,
  Institute for Ground Water Research, University of
  Waterloo,  Waterloo, Ontario, Progress Report
  0863985.
Homsy, G. M., 1987.  "Viscous Fingering in Porous
  Media."  Annual Review of Fluid Mechanics, Vol.
  19, pp. 271-311.
Kueper, B. H. and E. O. Frind,  1988. "An Overview
  of Immiscible Fingering in Porous Media." Journal
  of Contaminant Hydrology, Vol. 2, pp. 95-110.
Kueper, B. H. and D. B. McWorter, 1988. "Mechanics
  and Mathematics of Immiscible Fluids in Porous
  Media." In: Dense Chlorinated Solvents  in Ground
  Water, Institute for Ground Water Research, Uni-
  versity of Waterloo, Waterloo, Ontario, Progress
  Report 0863985.
Plumb, R. H., Jr. and A. M. Pitchford, 1985. "Volatile
  Organic Scans: Implication for Ground-water
  Monitoring." Proceedings Petroleum Hydrocarbons
  and Organic Chemicals in Ground Water, National
  Water Well Association, November 13-15,  1985,
  Houston, TX, pp. 207-222.
Saffman, P. G. and G. Taylor, 1958. "The Penetration
  of a Fluid into a Porous Medium or Hele-Shaw Cell
  Containing a More Viscous Liquid." Proceedings of
  the Royal Society London, A(245), pp. 312-329.
Schwille, F., 1967. "Petroleum Contamination of the
  Subsoil - A Hydrological Problem." In: The Joint
  Problems of the Oil and Water Industries,  Peter
  Hepple, Editor. Elsevier, Amsterdam, pp. 23-53.
Schwille, F., 1988. Dense Chlorinated Solvents in
  Porous  and Fractured Media. Translated by J.F.
  Pankow. Lewis Publishers, Inc., Chelsea, MI.
Villaume, J. F.,  P. C. Lowe, and D. F. Unites, 1983.
  "Recovery of Coal Gasification Wastes: An Innova-
  tive Approach." Proceedings of the Third National
  Symposium on Aquifer Restoration and Ground-
  Water Monitoring. National Water Well Associ-
  ation, Worthington, OH, pp. 434-445.
                                               27

-------

-------
                                       CHAPTER 4
         DETERMINATION OF PHYSICAL TRANSPORT PARAMETERS
                              Carl D. Palmer and Richard L Johnson
Introduction
The models used for the simulation and prediction of
contaminant transport in the subsurface are only as
good as the data input used in those models. The
range  of important variables for contaminant
transport are tabulated (Mercer, et al., 1982), but
ultimately site-specific values must be obtained.
Hydrogeological parameters such as hydraulic
conductivity, porosity, bulk density, ground-water
flux, and dispersion are important for modeling;
methods for determining these  parameters are
discussed below.

Both laboratory and field methods are useful for
determining hydrogeological parameters. Although
parameters  measured in  the laboratory are
applicable in small-scale situations, they may not be
representative  of the bulk properties of the
formation. Still, if many such measurements are
taken, not only can the average,  larger  scale
properties of the formation be estimated, but other
important statistical properties also can be calcu-
lated. For example, Sudicky  (1986)  applied such
techniques to determine the mean and  variance of
the saturated hydraulic conductivity of the Borden
aquifer in  order to calculate the macroscopic scale
dispersivity.

Field methods represent an increase in  the scale of
measurement relative to most laboratory methods.
This increase does not mean that such methods are
inherently better than laboratory methods, but
simply that  the field-measured variables represent
average properties over a larger volume. The key
advantage  of field methods is the  potential for
measuring less disturbed materials, thereby giving
more accurate representations of  the  relevant
parameters. The disadvantage to using such methods
is that during the data analysis, ideal models are
applied  to  non-ideal media. To reduce this
discrepancy, field and laboratory methods should be
developed or modified to complement one another.
Hydraulic Conductivity
Hydraulic conductivity can be measured both in the
laboratory and the field. Laboratory methods include
estimation from grain-size analysis, permeameter
tests, and soils-engineering tests. Slug tests, aquifer
tests, and flow net analyses are field methods that
provide increasing scales of measurement.


Laboratory Methods
Hydraulic conductivity, K, can be estimated from
grain-size distribution curves using either the Hazen
method (Freeze and Cherry, 1979) or the Masch and
Denny method (1966). Sieve, pipet, hydrometer, and
settling tube methods can be used to analyze grain
size, and light  scattering techniques have been
developed to look at  micron- and submicron-sized
particles. Sample preparation and methods for
particle-size analysis  are described in detail by Gee
and Bauder  (1986). In addition  to its use for
estimating hydraulic conductivity, grain-size
analysis can be used to properly size filter packs and
screens for monitoring, extraction, and injection
wells (Driscoll, 1986).

Saturated hydraulic conductivity  is measured with
either  constant-head  or falling-head permeameter
tests. Constant-head  tests are useful for measuring
hydraulic conductivity in the  range of 100 to 10-5
                                              29

-------
cm/s. Falling-head tests work best over a range of
10-3 to 10'7 cm/s. Undisturbed samples used in
permeameter tests offer the best results. If the soil
structure is disturbed, errors can be minimized by
repacking the soil to its original bulk density.

Soils-engineering tests in consolidometers and
triaxial cells provide coefficients of compressibility
and consolidation of soils. These  properties are
related to the hydrogeologic parameters of hydraulic
conductivity, K, and specific storage, Ss (Jorgensen,
1980). Parameters K and Ss for fine-grained
materials have been successfully calculated from the
coefficients of compressibility and consolidation
(Desaulniers, et al., 1981; Paul, 1987). In fractured
tills, these calculated values of K and Ss represent
values for the matrix.
Unsaturated hydraulic conductivity can be meas-
ured using steady-state head control  methods,
steady-state flux control  methods,  non-steady  state
methods, or sorptivity  methods. Details of methods
for both  saturated  and unsaturated media are
provided by Klute and Dirksen (1986). Empirical
models for predicting unsaturated hydraulic
conductivity as a function of water content or
pressure head are described by Mualem (1986).
Field Methods
Slug tests are the most common method for obtaining
hydraulic conductivity in the field. These  tests are
conducted  by instantaneously changing  the
hydraulic head within a well and measuring its
return to the static level. Different types of slug tests
include: (1) falling head; (2) rising head; (3) bail; and
(4) pressure/packer.

In a falling-head test, the water level is instan-
taneously increased by adding a slug of water, or,
preferably, some displacing volume of material such
as a metal rod. The head is then measured as it falls
back to its static level. In a rising head test, a volume
(e.g., a metal rod) is removed from the well, causing
the water level to instantaneously drop. The rising
hydraulic  head is then measured over a time
interval.

A bail test is a type of rising head test where the
water is removed using a bailer. This technique is
suitable in  low-permeability material where the
time needed to remove  the water is short relative to
the overall time needed for recovery in the well. In a
pressure/packer test,  an  interval of the well is
isolated by packers and a pressure pulse is applied to
that area. The  decay of the pressure pulse can be
measured with  a pressure transducer and the data
interpreted in the  same manner as the other test
methods. In principle, these data are  used  to
calculate the hydraulic conductivity and the specific
storage of the geologic material.

Several techniques used for analyzing slug test data
are described by Hvorslev (1951), Bouwer and Rice
(1976), Cooper, et al. (1967), and Nguyen and Finder
(1984). Both the Hvorslev (1951) and the Bouwer and
Rice (1976)  methods are based on steady-state flow
equations. If the  specific storage of the medium is
small, these techniques can provide a good estimate
of the hydraulic conductivity. In theory, the relative
recovery in the well should plot as a linear function
of time for both of these methods. The slope of this
recovery-time curve is used to calculate hydraulic
conductivity. If the formation has a non-zero specific
storage, a curve  rather  than a straight  line is
obtained and application of steady-state equations
results in over-estimation of the hydraulic  conduc-
tivity.

The Cooper, et al. (1967) and the Nguyen and Pinder
(1984) methods are based on transient ground-water
flow equations and, therefore, better represent  the
physical conditions within the soil. The Cooper, et al.
(1967) method assumes that the well fully penetrates
the aquifer  while the Nguyen and Pinder (1984)
method accounts for partial penetration of the well.
The Cooper, et al. (1967)  method is a curve-fitting
technique and is not very sensitive to the value of the
specific storage.  The  Nguyen and Pinder (1984)
model is the most general and, in principle, should be
the best  technique for interpreting slug test  data.
Instead of being a curve-fitting technique,  the model
uses the slopes of two different plots to calculate  the
specific storage and the hydraulic conductivity.

Many factors contribute to errors in the calculation
of hydraulic conductivity from slug test data (Table
3). Most of  these  factors  affect the estimation of
hydraulic conductivity by a factor of 2  or 3.  An
important exception is when a low-permeability skin
forms at the well-bore  interface.  Under these
circumstances, order of magnitude errors can result
(Palmer and Paul, 1987; Faust and Mercer, 1984). In
addition, simulated recovery data from wells that
have low permeability skins of finite thickness at the
well-bore interface can produce straighter  lines on a
Hvorslev plot than simulated data for ideal wells
without a skin. Because of this, a  straight-line
Hvorslev plot is not necessarily a valid criterion for a
"good" slug test.

Aquifer tests can provide larger scale information
about hydraulic properties than laboratory methods
or slug tests, and can be used to determine  hydraulic
conductivity,  specific  storage, leakage,  aquitard
diffusivity, anisotropy, and the general location of
boundaries. Also,  aquifer tests can be constant rate,
variable  rate, or constant head. Many different types
                                                30

-------
Table 3.    Potential Sources of Error in Slug Tests (Palmer
	and Paul, 1987)	

     Bridging of Seals
     Leaky Joints
     Formation of Low Permeability Skins
     Entrapped Air
     Presence of Fractures
     Stress Release Around Borehole
     Partial Penetration of Well
     Anisotropy of Formation
     Varying Regional Potentiometnc Surface
     Boundary Conditions
     Sand Pack Effects
     Uncertainty in Initial Head
     Radius of Influence of Test
     Thermal Expansion
of methods are used to analyze aquifer test data and
the choice of method depends on the conditions under
which the data were collected (steady-state flow or
non^steady-state  flow)  and the  type of aquifer
(confined, unconfined, or semi-confined). Methods for
analyzing aquifer  test data  can be found in
Kruseman and de Ridder (1970), Walton (1962, 1970,
and 1984), Lohman (1972), and Hantush (1964).

The classic test method of Theis (1935) can be used
for confined aquifers. However, this method is
limited to fully penetrating wells in isotropic media.
If the wells partially penetrate the aquifer, a variety
of methods by Hantush (1964) can be  used. These
methods also can be applied to leaky aquifers. If
monitoring wells are installed in the aquitard, then
the ratio method (Neuman and Witherspoon, 1972) is
useful for obtaining the hydraulic diffusivity of the
aquitard.

Investigators have  long debated the nature of
"delayed yield" observed in aquifer tests conducted
in unconfined aquifers (Neuman, 1972 and 1979).
The methodology presented by Neuman (1975) has a
stronger physical  basis  than other methods and
accounts for anisotropy and partial penetration.
Neuman's (1975) method is highly recommended for
use in unconfined aquifers.

Although the methods described above for analyzing
aquifer test data were derived for porous media, they
also can be applied to aquifer tests conducted in
fractured rock if the aquifer  behaves like a single
porosity medium. However, such fractured media are
often  highly anisotropic, and the methods by Weeks
(1969) or by Way and McKee (1982) may be more
appropriate in these situations.

If the test is conducted in a fractured, porous aquifer,
then the double porosity methods by  Barenblatt, et
 al. (1960) or by Boulton and Streltsova (1977) may be
 applied. Determining whether an aquifer is a single
 or double porosity system can be accomplished using
 a simple method (Gringarten, 1984), where a plot of
 ds/3 [ln(t)l versus t is made on log-log paper (Figure
 26).  In a single porosity aquifer, the plot should
 increase to  a maxima, then decrease and become
 constant. In a double porosity aquifer, the plot should
 go through  a maxima, and then decrease  and pass
 through a minima before finally becoming constant.
                             SINGLE POROSITY
           DOUBLE POROSITY
                       LOG (t)

Figure 26.    Differentiating double porosity media from
            single  porosity media  (after Gringarten,
            1984).

Flow-net analysis can provide a larger scale estimate
of hydraulic conductivity than aquifer tests. While
regional scale values may not be directly applicable
to a specific site, they may prove useful in regional
scale models to define  local boundary conditions.
Methods of flow-net analysis are described in Freeze
and Cherry (1979) and Cedergren (1967).

Unsaturated  hydraulic  conductivity can  be
measured in the field using either  steady-state or
non-steady-state flux methods. Such techniques are
described in detail by Green, et al.  (1986).
Bulk  Density,  Porosity,  and  Volumetric
Water Content
Bulk density, porosity, and volumetric water content
also are required for transport models. Porosity is
necessary to estimate the ground-water velocity from
the Darcy flux while both the porosity and the bulk
density are needed to calculate a retardation factor
from a  partition coefficient (see Chapter 5).  The
volumetric water content is required to estimate
water  and contaminant movement through the
vadose zone. Laboratory  techniques exist for
measuring these  parameters (Danielson  and
Sutherland, 1986;  Blake and Hartage,  1986;
Gardner, 1986). Field methods used to estimate these
parameters include neutron logs for porosity  and
water content, gamma-gamma logs for bulk density,
and electrical capacitance  for water content. In
                                                31

-------
addition, the porosity  can be calculated from the
average linear velocity, v, and the ground-water
flux, q, by q/v.
Ground-Water  Flux  and  Average  Linear
Velocity

Potentiometric Surface Data
The most common method for  estimating  the
magnitude and direction of ground-water flux is with
potentiometric surface data and hydraulic conduc-
tivity. The direction of ground-water flow in isotropic
media  is downgradient orthogonal to the equipo-
tential lines.  The flux rate, q, can be calculated
directly from  Darcy's equation and the product of
hydraulic conductivity and gradient. If the porosity,
0t,  is known, the average linear velocity, v, is
calculated as the ratio, q/0t. If the aquifer is aniso-
tropic, the flow lines and the equipotentials are not
orthogonal and the direction of flow  must  be
estimated from flow nets constructed  in  the
transformed section or through use of the inverse
hydraulic-conductivity ellipse (e.g.,  Freeze and
Cherry, 1979).
Borehole Dilution
Borehole dilution is a relatively simple technique to
determine the magnitude and, in principle,  the
direction of ground-water flow. In a borehole dilution
test, an interval within a well is  isolated with
packers and a tracer is injected and continuously
mixed. During the test, ground water enters the well
bore and dilutes the tracer. The rate at which the
tracer is diluted within the well bore is a measure of
the rate of ground-water flow. The relationship
between the concentration and ground-water flux is:
        In
(C - C')
CC0- C')
A0q
 w  " - V
(1)
where C' is the background concentration, C0 is the
concentration in the injected slug,  A is  the  cross-
sectional area of the borehole, W is the volume in the
isolated section of the borehole, and q is the ground-
water flux. The P parameter accounts for the con-
vergence of flow  lines on the open  borehole and is
often called the borehole factor. This convergence of
flow  lines results in a  greater  flux through the
borehole than through the aquifer. For an ideally
installed well with no sandpack and with the screen
permeability much greater than the formation, p has
                                        a value of 2.0. Methods for estimating 0 are discussed
                                        by Drost, et al. (1968) and by Halvely, et al. (1966).

                                        According  to Equation 1, a plot of the  logarithm of
                                        relative concentration versus time should produce a
                                        straight line with a slope of - A{5q/W. Thus, if A, 0,
                                        and W are known, the ground-water flux, q, can be
                                        calculated. Experiments conducted in a sandbox by
                                        McLinn and Palmer (1988) show that the logarithm
                                        of relative concentration versus time is  linear  and
                                        that the flux rates calculated from the borehole
                                        dilution test agree with the  measured flux values
                                        (Figure 27).
                                                                  q(calc.) = 5.9 cm/s

                                                                  q(meas.) = 5.7 cm/s
                                                        100
                                                                 200
                                                                          300
                                                                                   400
                                                          TIME (MINUTES)
                                        Figure 27.
           Results of borehole dilution test illustrating
           linear relationship between the logarithm of
           relative concentration and time.
There are several different types of borehole dilution
devices.  For example,  early investigators  used
radioisotopes with scintillation counters (Halvely, et
al., 1966), as well as specific ion electrodes (Grisak,
et al., 1977) and specific conductance  electrodes
(Bellanger, 1985).  Currently, a thermal device  is
commercially available (Kerfoot and Skinner, 1980;
and Kerfoot, 1982), and an electrical resistivity
device has been developed  (McLinn and Palmer,
1988). One key advantage of borehole dilution is its
potential to profile the distribution of velocities
along the length of the  well screen. Under appro-
priate conditions, therefore, the device can be used to
discern velocity variations within the aquifer and to
describe aquifer heterogeneity.
                                        Seepage Meters
                                        Seepage meters can be an extremely useful tool for
                                        measuring fluid flux at  the ground-wate'r/surface-
                                        water interface (Lee, 1977; Lee and Cherry, 1978).
                                                 32

-------
The seepage meter consists of an inverted section of a
55-gallon drum with a hole in the top that is covered
with a plastic bag (Figure  28). After a measured
period of time (At), the bag is removed and the
volume of water collected, V, is measured. The flux of
ground water through the sediments and into the
surface water is simply V/(AAt) where A is the area
of the  drum.  If a mini-piezometer is used in con-
junction  with the seepage meter,  the hydraulic
gradient between the  sediments  and  the surface
water can be calculated.  If the hydraulic gradient
and the flux rate are known, a simple calculation
using Darcy's equation yields the hydraulic conduc-
tivity of the sediments.  The seepage meter works
best in materials that are not easily compressed; if
the sediments are compressed during installation,
the hydraulic conductivity will be underestimated.
                    Mini-piezometer
                               Plastic Bag
Figure 28.   Seepage meter.
Dispersion Coefficient and Aquifer
Heterogeneity

Tracer Tests
Tracer tests have been  used to study physical,
chemical, and biological processes in the subsurface.
Physical parameters such as dispersion, average
linear velocity, porosity, and variation in hydraulic
conductivity have been obtained from tracer tests.
The tracer tests by Sudicky, et al. (1983) and Button
and Barker (1985) show that ground-water velocity
can vary substantially over small scales. In their
experiments, small slugs of contaminants broke into
discrete parts that moved at  different velocities.
Also, Sudicky, et al. (1983) and  Freyberg (1986) used
tracer tests to describe how the apparent dispersivity
value increases with the scale  of the  problem.
Pickens and Grisak (1981) were able to demonstrate
that when the concentration breakthrough curves
obtained with point samplers are analyzed, the
dispersivity  parameter is small  (0.007  m). In
addition, Palmer and Nadon (1986) and Taylor, et al.
(1988)  suggest that electrical resistivity  can be
coupled with  the execution of  single-well injection
tracer tests to obtain information about variation in
hydraulic conductivity.

Tracer tests  can be conducted in several ways.
Natural-gradient tracer tests are commonly used for
research purposes. In these tests, a slug of tracer is
injected into an aquifer and traced as it moves along
with the natural  hydraulic gradient. Such tests were
used by Sudicky, et al. (1983), Sutton and  Barker
(1985), and MacKay, et al. (1986).

Forced-gradient tracer-, experiments involve the
injection of a  tracer over a prolonged period of time
and measurement of the arrival of the tracer at one
or more monitoring wells during the injection period.
Forced-gradient  wells may be  either single-well or
double-well tests. In  single-well  tests,  only one
"active" well  is used for injecting  and withdrawing
the tracer. The fluid is injected into the aquifer and
monitored at  one or more sampling wells at some
radius  from the  active well  (Figure 29).  Two-well
tests (Figure  30) use two active wells, one to inject
the tracer and the other for withdrawal. These types
of forced gradient tests were used by Pickens and
Grisak (1981), Molz, et al. (1986), Palmer and Nadon
(1986), and Taylor, et al. (1988).

Experiments by Molz, et al. (1986) demonstrate that
tracer tests can be applied as an engineering tool and
numerical models can  be  used  to  predict the
transport of solutes in the subsurface.  Molz,  et al.
(1986) initially conducted a single-well tracer test in
a confined aquifer to measure  the variation in
hydraulic  conductivity over the  thickness of the
aquifer. These tests were on a scale of approximately
5.5 m. A two-well tracer test with travel distances of
38 to 90 m was also conducted. Using the hydraulic
conductivity distribution for the single-well test, the
concentration-versus-time curve of a pumping well
in a two-well test could  be  reasonably simulated
without recourse to model calibration (Figure 31).
Two models  were used:  (1) a three-dimensional
advection-dispersion model using small dispersivity
values; and (2) an advection model. The results of the
two models are indistinguishable. If the aquifer is
assumed to be homogeneous and the concentrations
                                                33

-------
INJECTION

       C|NJ(t)


     UPPER
                           WITHDRAWAL
                           Q=00UT
                          e.
      •K(z)
                      CONFINING LAYER
                          INJECTION-
                         -WITHDRAWAL
                          WELL
                   OBSERVATION
                   WELL   —2	.
                   WITH
                   MULTILEVEL
                   SAMPLERS
                                            -•-r
              LOWER CONFINING LAYER
Figure 29.   Vertical cross-section showing single-well
           test geometry (Molz, et al., 1986).
penetrating radar, induced electrical polarization,
resistivity, magnetometer, reflection seismics, and
electromagnetic surveys. Borehole methods include
geothermetry, electrical methods, acoustic methods,
and nuclear logging techniques. Surface geophysical
techniques such as resistivity, conductivity, seismic
refraction, and VLF can be  useful in identifying
lithology changes in the subsurface, depth to water
table, and depth to bedrock. Ground-penetrating
radar, electromagnetic methods (EM), and resistivity
are  effective in locating  buried  drums  and
containers. Borehole geophysical  techniques are
used to estimate hydraulic conductivity (resistivity),
lithology (natural gamma), bulk density (gamma-
gamma), and porosity (neutron). Discussions of these
topics can be found in Keys  and MacCary (1971),
Benson, et al. (1983), and Rehm, et al.  (1985). As
described in the discussion of tracer tests, borehole
resistivity logs can be  coupled with tracer tests to
obtain information about subsurface heterogeneity
(Taylor, et al., 1988; Palmer and Nadon, 1986).
      Injection well
        (source)
                    Withdrawal well
                         (sink)
                  Plan view
                                 Multi-Level
                                 Observation
                                 Well
xxx^xxxyxl Uxxx^
If

h -!*
i
i
•
/•yxyyxy

\
~KMJ
J
/
sssssssA \s
I

r*
1
I r*
it
SA \SSSSSS////S

-jj-*
"{h*
j i
"ir*"
           Vertical section  in x-z plane

Figure 30.   Two-well test geometry in a stratified aquifer
           (Molz, et al., 1986).
in the pumping well are calculated using average
hydraulic  properties,  poor replication of the
experimental data is obtained (Figure 32). These
results emphasize the importance of heterogeneity
and advection to the transport of solute through
porous media.
Geophysical Techniques
While geophysical techniques are recognized as
useful tools in characterizing waste sites, a detailed
discussion of geophysical techniques is beyond the
scope of this chapter. Surface geophysical techniques
used  include gravity, infrared imagery, ground-
Plume Detection

Ground-Water Monitoring
The most common method for detecting contaminant
plumes in the subsurface is by direct sampling of the
subsurface  fluid. However,  the common practice of
"plume  chasing" may not be the most efficient
method for determining the extent of ground-water
contamination (Dowden and Johnson, 1988). Greater
use of hydrogeologic data can greatly improve the
design of ground-water monitoring systems and the
quality of information provided.

Although ground-water monitoring networks can
provide the greatest certainty about  the  extent of
contamination, there are important factors that
control the  quality of this information, including the
amount of well  purging done prior to sampling
(Barcelona and Helfrich,  1986), the method of
sampling (Stolzenburg and  Nichols, 1985), and the
method of well construction and installation (Keely
and Boateng, 1987). Methods for ground-water
sampling are discussed in detail by Scalf, et al.
(1981), Ford, et al. (1984), and Barcelona, et al.
(1985).
Geophysical Techniques
Geophysical  methods can be  used  to  locate
contaminant plumes with high dissolved solids. For
example, surface resistivity methods have  been
successful at some sites (e.g., Stellar and Roux,  1975;
Kelly, 1976; Rogers and Kean, 1980), and electro-
magnetic methods are useful for obtaining  some
estimate of water quality changes in the subsurface
                                               34

-------

^
\
0
—
Z
O
h-
o:
i-
z
UJ
u
z
o
u



25.0
22.5
20.0
1 7.5

1 5.0
1 2.5

1 0.0

7.5

5.0
2.5
n
_
-
-
-

-
.

-

-

~
,

         0
                                                          —•—Experiment

                                                          	Geotrans  Calculation (3-D)
                                                            0   TWAM Calculation
                                                                (Advection)
                     •J	1	1	1	1	1	1	\	III'
Figure 31.
  40  80  I 20 160 200 240 280 320 360 400 440 480 520 560 600 640 680 720 760

                                 TIME  (MRS)
Comparison  of  predicted  and  modeled concentrations in the pumping well of a two- well tracer test (Molz,
etal., 1986).
(Stewart, 1982; Slaine and Greenhouse, 1982;
Glaccum, et al., 1982; Ludwig, 1983). The borehole
counterparts to  these surface geophysical methods
should work equally as well in determining vari-
ations in water quality over the thickness of the
aquifer. Detailed discussions of these methods and
their limitations can be found in Keys and MacCary
(1971), Benson, et al. (1983) and Rehm, et al. (1985).
Chemical Time-Series Sampling Tests
Chemical time-series sampling tests are conducted
by repeatedly sampling wells that are continuously
pumped (Keely, 1982). The concentration data can
then be plotted as a function of time, volume
removed, or an equivalent radius from the well bore.
The particular shape of the curve obtained from such
a test  varies according to the distribution  of
contaminants in the subsurface. Some suggested
interpretations are provided in Figure 33. Data from
the time-series test have been used to determine the
proper amount of well  purging that should  be
performed before collecting  water samples at the
site,  and to identify the  source  of a contaminant
entering a water-supply well (Keely and Wolf, 1983).
While  there  is  no  unique  interpretation  of the
                                      resultant curve,  its general shape can provide
                                      information that must be reconciled with hypotheses
                                      concerning the distribution of contaminants near the
                                      well bore.
                                      References

                                      Barcelona, M. J., et al.,  1985. "Practical Guide for
                                        Ground-water Sampling." U.S. EPA Report,
                                        EPA/600/2-85/104. Robert S. Kerr Environmental
                                        Research Laboratory, Office of Research and
                                        Development, U.S.  Environmental Protection
                                        Agency, Ada, OK.
                                      Barcelona, M. J. and J.  A. Helfrich, 1986. "Well
                                        Construction and Purging Effects on Ground-water
                                        Samples." Environmental Science and Technology,
                                        Vol. 20, pp. 1179-1184.
                                      Barenblatt, G. I., I. P. Zheltor, and I. N. Kochina,
                                        1960. "Basic Concepts in the Theory of Seepage of
                                        Homogeneous Liquids in Fissured Rocks."
                                        Prikladnaya Matem, Mekh (USSR), Vol. 24, No. 5,
                                        pp. 852-864.
                                      Bellanger, D., 1985. "Ground-water Velocity
                                        Measurements Using An Electrical Conductance-
                                        Borehole Dilution Device."  Unpublished Master's
                                              35

-------
<
r-
•z.
UJ
o
z
o
o
              45
              40
              35
              3O
             25
              I 5
              I O
                              8
                              o  o
                             o

                             0°   °
                            o
                O  4O  8O  ISO I6O ZOO 240 2BO 32O 360 400 44O 480 320 560 600 640 680 720 760

                                           TIME (hrs.)

Figure 32,   Comparison of predicted and  modeled concentrations in the pumping well of  two-well tracer test
           assuming a homogeneous aquifer (Molz, et al., 1986).
 Thesis, Department of Earth Sciences, University
 of Waterloo, Waterloo, Ontario.

Benson, R. C., R. A. Glaccum, and M. R. Noel, 1983.
 "Geophysical Techniques for Sensing Buried
 Wastes and Waste Migration." Contract Report No.
 68-03-3050, Environmental Monitoring Systems
 Laboratory, Office of Research and Development,
 U.S. Environmental Protection Agency, Las Vegas,
 NV.

Blake, G. R. and K. H.  Hartage, 1986. "Bulk
 Density." In:  Method of Soil Analysis, Part I,
 Physical and  Mineralogical Methods,  A. Klute,
 Editor. Agronomy, No. 9, Part 1, American Society
 of Agronomy, Inc., Soil Science Society of America,
 Inc., Madison, WI, pp. 363-375.
                                          Boulton, N.S. and T.D. Streltsova, 1977. "Unsteady
                                           Flow to  a Pumped Well in a Fissured Water
                                           Bearing  Formation." Journal of Hydrology, Vol.
                                           35, pp. 257- 270.
                                          Bouwer, H. and R.C. Rice, 1976. "A Slug Test for
                                           Hydraulic Conductivity of Unconfined  Aquifers
                                           with Completely or Partially Penetrating Wells."
                                           Water Resources Research, Vol. 12, No. 3, pp. 423-
                                           428.
                                          Cedergren, H., 1967. Seepage, Drainage, and Flow
                                           Nets. John Wiley & Sons, New York.
                                          Cooper,  H.  H.,  J. D.  Bredehoeft  and  I.  S.
                                           Papadopulos, 1967. "Response of a Finite-diameter
                                           Well to an Instantaneous Charge of Water." Water
                                           Resources Research, Vol. 3, No. 3, pp. 263-269.
                                               36

-------
         o
         <
         cc

         UJ
         o
         o
         o
Figure 33A.
         O
         I
         O
         o
Figure 33B.
TIME OR VOLUME PUMPED
Hypothetical curve obtained from chemical
times-series sampling test. (The results may
be interpreted as data from a  monitoring
well placed at the edge of a plume where
the concentrations  are reduced due  to
increasing  amounts of uncontaminated
water entering the borehole.) (After Keely,
1982).
                                  B
         CC
         I
         o
         I
TIME OR VOLUME PUMPED
Hypothetical curve obtained from chemical
times-series sampling test.  (The  results
may be interpreted as  data  from a
monitoring well  that  may  have been
contaminated during drilling or installation.
The concentrations rapidly decrease  as
uncontaminated  water  from  the  aquifer
enters the well bore.) (After Keely, 1982).
             TIME OR VOLUME PUMPED
 Figure 33C.   Hypothetical curve obtained from chemical
             times-series sampling test. (The results may
             be interpreted  as data from  a  monitoring
             well in which  chemical  reactions have
             reduced concentrations  in the stagnant
             water column.  When pumping  is started,
             contaminated ground water from  the aquifer
             enters  the  well  and  contaminant
             "concentrations rapidly increase.) (After
             Keely, 1982).
Danie'lson, R.E.  and P.L. Sutherland,  1986.
  "Porosity." In: Method of Soil Analysis, Part I,
  Physical and Mineralogical Methods, A. Klute,
  Editor. Agronomy, No. 9, Part 1, American Society
  of Agronomy, Inc., Soil Science Society of America,
  Inc., Madison, WI, pp. 443-461.
Desaulniers, D. E., J. A. Cherry, and P. Fritz, 1981.
  "Origin,  Age, and  Movement of  Pore Water in
  Argillaceous Quaternary Deposits at Four Sites in
  Southwestern Ontario." Journal of Hydrology, Vol.
  50, pp. 231-257.

Dowden, J. and  L.  Johnson,  1988.  "Cost-benefit
  Analysis  of Alternative Remedial Investigation
  Methodologies:  A  Case Study." Proceedings of the
  Second National Outdoor Conference  on Aquifer
  Restoration,  Ground-water Monitoring, and Geo-
  physical  Methods,  Las  Vegas,  NV, May 23-26,
  1988, pp.  1495-1507.

Driscoll, F.G.,  1986. Groandwater and  Wells,  2nd
  Edition. Johnson Division, St. Paul, MN.

Drost, W. D., et al.,  1968. "Point Dilution Methods of
  Investigating Ground  Water Flow by Means of
  Radioisotopes." Water Resources Research, Vol. 4,
  No. I, pp. 125-146.

Faust, C. R. and J.  W. Mercer, 1984. "Evaluation of
  Slug Tests in Wells Containing a Finite Thickness
  Skin."  Water Resources Research, Vol. 20,  No. 4,
  pp. 504-506.

Ford,  P. J., P. J. Turina, and D. E. Seely, 1984.
  "Characterization of Hazardous Waste Sites  — A
  Methods  Manual, Volume  II, Available  Sampling
  Methods." U.S. EPA Report, EPA/600/4-84/076.
  Robert S. Kerr Environmental Research Labora-
  tory, Office  of  Research and Development, U.S.
  Environmental Protection Agency, Ada, OK.

Freeze, R.  A. and J. A. Cherry, 1979. Groundwater.
  Prentice-Hall, Inc., Englewood Cliffs, NJ.

Freyberg, D. L.,  1986. "A Natural Gradient
  Experiment on Solute Transport in a Sand Aquifer.
  2. Spatial Moments and the Advection and  Dis-
  persion of Nonreactive Tracers." Water Resources
  Research, Vol. 22, pp. 2031-2046.

Gardner, W. H., 1986. "Water Content."  In: Methods
  of Soil Analysis, Part /, Physical and Mineralogical
  Methods, A. Klute, Editor. Agronomy, No. 9,  Part
  1, American Society of Agronomy, Inc.,  Soil Science
  Society of America, Inc., Madison, WI, pp. 493-544.

Gee, G.  W. and J.  W. Bauder, 1986. "Particle-size
  Analysis." In: Methods of Soil  Analysis, Part I,
  Physical and Mineralogical Methods, A.  Klute,
  Editor. Agronomy, N'o. 9, Part 1, American Society
  of Agronomy, Inc., Soil Science Society of America,
  Inc., Madison, WI, pp. 383-411.

Glaccum, R. A., R. C. Benson, and M.R. Noel, 198?
  "Improving  Accuracy  and  Cost-effectiveness  of
                                                 37

-------
   Hazardous Waste Site Investigations." Ground-
   water Monitor ing Review, Vol. 2, pp. 35-40.
 Green, R. E., L.  R. Ahuja, and S. K. Chong,  1986.
   "Hydraulic Conductivity, Diffusivity,  and
   Sorbivity of Unsaturated Soils: Field Methods." In:
   Methods of Soils Analysis, Part I, Physical and
   Mineralogical Methods, A. Klute, Editor. Agron-
   omy, No. 9, Part 1, American Society of Agronomy,
   Inc., Soil Science Society of America, Inc., Madison,
   WI, pp. 799-823.
 Gringarten, A.C., 1984. "Interpretation of Tests in
   Fissured and Multilayered Reservoirs with Double
   Porosity Behavior: Theory and Practice." Journal
   of Petroleum Technology, April, 1984, pp. 549-564.
 Grisak, G., W. Merritt, and D. Williams, 1977. "A
   Fluoride Borehole Dilution Apparatus for Ground-
   water Velocity Measurements." Canadian  Geo-
   tecknicalJournal, Vol. 14, No. 4, pp. 554-561.
 Halvely, E.,  et  al., 1966. "Borehole  Dilution
   Techniques - A  Critical Review." In: International
   Atomic Energy Agency, Isotopes in Hydrology,
   Vienna.
 Hantush, M. S., 1960. "Modification of the Theory of
   Leaky Aquifers." Journal of Geophysical Research,
   Vol. 65, No. 11, pp. 3713-3725.

 Hantush, M.  S.,  1964. "Hydraulics of Wells." In:
   Advances in Hydroscience, Vol. 1, Academic Press,
   Inc., New York,  NY, pp. 281-432.
 Hvorslev, M.  J.,  1951. "Time  Lag and  Soil
   Permeability in  Ground-water Observations." U.S.
   Army Corps of Engineers, Waterways Experiment
   Station Bulletin 36, Vicksburg, MS.
 Jorgensen, D.  G., 1980. "Relationships  between
   Basic Soils-engineering Equations and Basic
   Ground-water Flow Equations." U.S. Geological
  Survey Water-Supply Paper 2064, 40 p.
 Keely, J. F., 1982.  "Chemical Time-series Sampling."
  Ground-water Monitoring Review, Fall  1982  pp
  29-38.
 Keely, J .F. and K. Boateng, 1987. "Monitoring Well
  Installation, Purging, and Sampling Techniques.
  Part II: Case Histories." Groundwater, Vol. 25, pp.
  427-439.
 Keely, J. F. and F. Wolf, 1983. "Field Applications of
  Chemical Time-series Sampling." Ground-water
  Monitoring Review, Fall, 1983, pp. 26-33.
 Kelly, W.  E.,  1976. "Geoelectric  Sounding  for
  Delineating  Ground-water Contamination."
  Groundwater, Vol. 14, pp. 6-10.

Kerfoot, W. B., 1982. "Comparison of 2-D  and 3-D
  Ground-water  Flowmeter  Probes in  Fully
  Penetrating Monitoring Wells." In: Proceedings of
  the  Second National Symposium on  Aquifer
  Restoration and  Ground-water Monitoring, D. M.
   Nielson, Editor. Worthington, OH, Water Well
   Journal Publishing Co., pp. 264-269.
 Kerfoot, W.  B.  and S.  M. Skinner, 1980. "Direct
   Ground-water Flow Measurement." In:  Proceed-
   ings of 5th National  Ground Water Quality
   Symposium, Las Vegas, October 1980.
 Keys, S. W. and L. M. MacCary, 1971. "Application
   of Borehole Geophysics to Water Resources
   Investigations." Techniques of Water Resources
   Investigations  of the  United  States Geological
   Survey, Chapter El, Book 2.
 Klute, A. and C. Dirksen, 1986. "Hydraulic Con-
   ductivity and Diffusivity: Laboratory Methods." In:
   Methods of Soil Analysis, Part I, Physical and
   Mineralogical  Methods,  A. Klute, Editor. Agron-
   omy, No. 9, Part 1, American Society of Agronomy,
   Inc., Soil Science Society of America, Inc., Madison,
   WI, pp. 687-734.
 Kruseman, G.  P. and N.  A. de Ridder, 1970.
   "Analysis and Evaluation of Pumping Test Data."
   Bulletin 11, International Institute for Land
   Reclamation and Improvement, Wageningen, The
   Netherlands.
 Lee, D. R., 1977. " A Device for Measuring Seepage
   Flux in Lakes and Estuaries." Limnology and
   Oceanography, Vol. 22, No. 1, pp. 140-147.
 Lee, D. R. and J. A. Cherry, 1978. "A Field  Exercise
   on Groundwater Flow  Using Seepage Meters and
   Mini-piezometers." Journal of Geological Educa-
   tion, Vol. 27, pp. 6-10.

 Lohman, S. W.,  1972. "Ground-water Hydraulics."
   U.S. Geological Survey, Professional Paper 708.

 Ludwig, K. J., 1983. "Electromagnetic Induction
   Methods for Monitoring Acid  Mine  Drainage."
  Ground-water Monitoring Review, Vol. 4  pp  46-
  57.

 MacKay, D. M.,  et al.,  1986. "A  Natural Gradient
   Experiment on Solute Transport  in  a Sand
   Aquifer. 1.  Approach and  Overview of Plume
   Movement." Water Resources Research,  Vol  22
   No. 13, pp. 2017-2030.

 Masch, F. D.  and K. J.  Denny,  1966. "Grain-size
   Distribution and Its Effect on the Permeability of
  Unconsolidated Sands." Water Resources Research
  Vol. 2, pp. 665-677.

 McLinn, E. L.  and C.  D. Palmer, 1988.  "An
  Electrical-resistivity Borehole-dilution Device for
  the Determination  of Ground-water Flux." Pro-
  ceedings of the Second National Outdoor Confer-
  ence on Aquifer Restoration, Ground-water Moni-
  toring, and Geophysical Methods, Las Vegas  NV
  May 23-26,1988, pp. 851-874.

Mercer, J. W., S. D.  Thomas,  and B.  Ross, 1982.
  "Parameters  and Variables  Appearing in
                                               38

-------
  Repository Siting Models." U.S. Nuclear Regula-
  tory Commission, Report NUREG/CR-3066.
Molz, F., et al., 1986. "Performance, Analysis and
  Simulation of a Two-well Tracer Test at the Mobil
  Site," Vol. 22, pp. 1031-1037.
Molz, F. J., et al., 1986. "Performance and Analysis
  of Aquifer Tracer Tests  with  Implications  for
  Contaminant Transport Modeling." U.S. Envi-
  ronmental  Protection Agency,  Research  and
  Development Report 600/2-86/062.
Morrison, R. D., 1983.  Ground-water Monitoring.
  Technology, Procedures, Equipment, and Applica-
  tions. TIMCO Mfg., Inc., Prairie du Sac, WI.
Mualem, Y., 1986. "Hydraulic Conductivity of Un-
  saturated Soils: Prediction and  Formulas." In:
  Methods of Soil Analysis, Part I, Physical and
  Mineralogical Methods, A. Klute, Editor. Agron-
  omy, No. 9, Part 1, American Society of Agronomy,
  Inc., Soil Science Society  of America,  Inc.,
  Madison, WI, pp. 799-823
Neuman, S. P., 1972. "Theory of Flow in Unconfined
  Aquifers Considering Delayed Response of the
  Water  Table." Water  Resources Research, Vol. 8,
  pp. 1031-1045.
Neuman, S. P. and P. A. Witherspoon, 1972. "Field
  Determination of the Hydraulic Properties of
  Leaky Multiple-aquifer Systems." Water Resources
  Research, Vol. 8, pp. 1284-1298.
Neuman, S. P., 1975.  "Analysis  of Pumping  Test
  Data from  Anisotropic  Unconfined Aquifers
  Considering Delayed Gravity Response." Water
  Resources Research, Vol. 11, No. 2, pp. 329-342.
Neuman, S. P.,  1979.  "Perspective on Delayed
  Yield." Water Resources Research, Vol. 15, No. 4,
  pp; 899-908.
Nguyen, V. and G. F. Finder, 1984. "Direct Calcula-
  tion of Aquifer Parameters in Slug Test Analysis."
  In: Ground Water Hydraulics, J. S. Rosenshein and
  G. D. Bennett, Editors. Water Resources Mono-
  graph Series 9, American Geophysical Union,
  Washington, DC, pp. 222-239.
Palmer,  C. D. and R. L. Nadon,  1986,  "A Radial
  Injection Tracer Experiment  in a Confined
  Aquifer, Scarborough, Ontario." Groundwater,
  Vol. 24, pp. 322-331.
Palmer, C. D. and D. G. Paul, 1987. "Problems in the
  Interpretation of Slug  Test  Data from  Fine-
  grained Glacial Tills." Proceedings of the Focus
  Conference on Northwestern Ground Water Issues,
  Portland, OR, May 5-7,1987.
Paul, D. G.,  1987. "The Effect of Construction,
  Installation, and Development Techniques on the
  Performance of Monitoring Wells in Fine-grained
  Glacial Tills." M.S. Thesis, Department of Geo-
  logical and Geophysical  Sciences, University of
  Wisconsin-Milwaukee.
Pickens, J.  F. and G. E. Grisak, 1981. "Scale
  Dependent  Dispersion in a Stratified Aquifer."
  Water Resources Research, Vol. 17, pp. 1191-1212.
Rehm, B. W., et al.,  1985.  "Field Measurement
  Methods for Hydrogeologic Investigations: A
  Critical Review of  the  Literature." Electrical
  Power Research Institute Report EA-4301, Project
  2485-7, Electrical Power Research Institute, Palo
  Alto, CA.
Rogers, R. B. and W.  F. Kean, 1980. "Monitoring
  Groundwater Contamination at a Fly-ash Disposal
  Site  Using Surface Electrical Resistivity
  Methods." Groundwater, Vol. 18, pp. 472-478.
Scalf, M. R., et al., 1981. "Manual of Ground-water
  Sampling Procedures." Robert  S. Kerr Environ-
  mental Research Laboratory, U.S. Environmental
  Protection Agency, Ada, OK, pp. 43-71.
Slaine,  D. D. and J. P. Greenhouse,  1982.  "Case
  Studies of Geophysical Contaminant Mapping at
  Several Waste Disposal Sites."  In: Proceedings of
  the Second National Symposium  on Aquifer
  Restoration and Ground Water  Monitoring, D. M.
  Nielsen, Editor. National Water Well Association,
  Columbus, OH, pp. 299-315.
Stewart, M. T., 1982. "Evaluation of Electromagnetic
  Methods  for  Rapid Mapping of  Salt-water
  Interfaces in Coastal Aquifers." Groundwater, Vol.
  20, pp. 538- 545.
Stollar, R. L. and P. Roux, 1975.  "Earth  Resistivity
  Surveys -  A Method for Defining Ground-water
  Contamination." Groundwater, Vol.  13, pp. 145-
  150.
Stolzenburg, T. R. and D. G. Nichols, 1985. "Prelim-
  inary Results  on Chemical Changes in Ground-
  water Samples Due to Sampling Devices." Electric
  Power Research Institute Report, EPRI EA-4118.
Sudicky, E. A., J. A. Cherry and E. O. Frind. 1983.
  "Migration of Contaminants in Ground Water at a
  Landfill: A Case Study. 4. A Natural-gradient
  Dispersion Test." Journal  of Hydrology, Vol. 63,
  No. 1/2, pp. 81-108.
Sudicky, E.  A.,  1986.  "A Natural  Gradient
  Experiment in a Sand Aquifer: Spatial  Variability
  of Hydraulic  Conductivity and Its  Role  in the
  Dispersion Process." Water Resources Research,
  Vol. 22, No. 13, pp. 2069-2082.
Sutton, P. A. and J. F. Barker, 1985. "Migration and
  Attenuation of Selected  Organics in a  Sandy
  Aquifer — A Natural  Gradient Experiment."
  Groundwater, Vol. 23, No. 1, pp. 10-16.
Taylor, K., F. Molz, and J.  S. Hayworth, 1988. "A
  Single Well  Electrical  Tracer Test for the
                                                39

-------
  Determination of Hydraulic Conductivity and
  Porosity as a Function of Depth." Proceedings of
  the Second National Outdoor Action Conference on
  Aquifer Restoration, Ground-water Monitoring
  and Geophysical Methods, pp. 925-938.
Theis, C. V. 1935. "The Relation Between the Lower-
  ing of the Piezometric Surface and the Rate and
  Duration of Discharge of a Well  Using Ground-
  water Storage." Transactions of the American
  Geophysical Union, Vol. 2, pp. 519- 524.
Walton, W. C.,  1962. "Selected Analytical Methods
  for Well and  Aquifer Evaluation." Illinois  State
  Water Survey Bulletin 49.
Walton, W. C., 1970. Groundwater Resource Evalua-
  tion. McGraw-Hill, New York, NY.
Walton, W. C., 1984. Practical Aspects of Ground
  Water Modeling, National Water  Well Associ-
  ation, Dublin, OH.

Way, S. C. and C.  R. McKee, 1982. "In-situ Deter-
  mination of Three-dimensional Aquifer Perme-
  abilities." Groundwater, Vol. 20, pp. 594-603.
Weeks, E. P., 1969. "Determining  the  Ratio of
  Horizontal  to Vertical Permeability by Aquifer-
  test Analysis." Water Resources Research, Vol. 5
  No. 1, pp. 196-214.
                                              40

-------
                                       CHAPTER 5
                      SUBSURFACE CHEMICAL PROCESSES
                        Richard L. Johnson, Carl D. Palmer, and William Fish
Introduction
Risk assessment and remediation of ground-water
contamination require an  understanding of how
chemicals move through and interact with the
subsurface environment.  However,  subsurface
transport of contaminants is often controlled by
complex interactions between chemical, physical and
biological processes. This means that mathematical
models often must be used to predict chemical
movement. Models for ground-water systems are
frequently based on the  advection-dispersion
equation described in Chapter 2 and shown here in
one-dimensional form:
      D
                aC
        ax
      - v —
    2     ax
                     at
                                     (D
This equation requires quantitative information to
simulate all subsurface processes. In the context of
the advection-dispersion  equation,  chemical
processes  important in controlling contaminant
transport and fate usually take the form of reaction
terms (RXN) added to the basic equation:
D
a2c
          ac
      _ v _ =
        ax
                     ac
                     _
                     at
                                  (2)
This chapter will discuss the chemical processes that
affect the  movement of organic and inorganic
contaminants in ground-water systems and how
those processes can be represented in the advection-
dispersion  equation.  Methods and  limitations of
experimental data for modeling chemical processes
also will be examined.
Reactions of Organic Compounds
Hundreds of thousands of organic chemicals are
currently used  in industrial  and  domestic
applications in the United States. These chemicals
represent an extremely broad range of physical and
chemical properties and are subject to different
physical, chemical, and biological processes in the
subsurface. For example,  because  chlorinated
solvents are only slightly soluble and are more dense
than water, they can penetrate deep into aquifers
and remain as an immiscible phase for prolonged
periods of time. In contrast, a spill of acetone will not
penetrate into the ground water because of its low
density. Also, because of its  miscibility with water,
the acetone will dissolve  quickly and become
available for  further chemical  and biological
reactions.

Organic reactions may transform one compound into
another, change the state of a compound, or cause a
compound to combine with other organic or inorganic
chemicals. In the context of the advection-dispersion
equation, these reactions represent changes in the
distribution of mass within the elementary volume
through which the movement of the  chemicals is
modeled. Although, many of these reactions  have
been studied in the laboratory and observed in the
field, there is a lack of good quantitative information
about the processes under complex, real-world
conditions.
Chemical reactions in the subsurface are frequently
characterized on a kinetic basis as equilibrium or
zero- or first-order, depending upon how the rate is
affected by the concentrations of the reactants. For
example, a zero-order reaction is one that proceeds at
a rate independent of the  concentration of the
                                              41

-------
reactant(s). In a first-order process, the rate of the
reaction is directly dependent on the concentration of
one of the reactants. In many cases, grouping a
chemical  process into one of these classes over-
simplifies the actual process. However, more
realistic, higher-order processes often are difficult to
measure and/or  model in complex environmental
systems.  The following  example of hydrolysis
illustrates some of the difficulties in obtaining
quantitative kinetic information and applying it to
the advection- dispersion equation.
Hydrolysis
Hydrolysis is  the  direct reaction of dissolved
compounds with water molecules, and can be an
important abiotic degradation process in ground-
water systems (Mabey and Mill, 1978). For example,
hydrolysis of chlorinated hydrocarbons is significant
because many chlorinated compounds are not readily
degraded by reactions  such  as biodegradation
(Siegrist and McCarty, 1987). This hydrolysis of
chlorinated compounds often yields an alcohol or an
alkene (Figure 34).
  RX + HOH
  HX
                 ROH +  HX
   I    I
  C-C
 Figure 34.
              OC  + HX
Schematic hydrolysis reactions for halo-
carbons.
Hydrolysis rates are typically difficult to measure in
the field due to other competing reactions and slow
degradation rates, and as a result, most hydrolysis
data come from laboratory studies. Even in the
laboratory, conditions must be optimized to minimize
other reactions and provide good quantitative data.
For example, data in Figure 35 (Ellington, et al.,
1986) for the hydrolysis of 1,2,4-trichlorobenzene
were collected at 70°C to speed the reaction and
eliminate competing microbial degradation. The
data then were fitted as a first-order reaction, which
assumes a straight  line on a semi-log plot, even
though the data suggest some higher-order effects.
From these data, the hydrolysis rate constant (K) at
the elevated temperatures was  estimated (Figure
36). The extrapolation of K at ambient temperatures
                                               100 Q
                                                50
                                       PERCENT
                                      REMAINING
                                                           10
               70°C
                                                          pH"7.11
                                                                    HALF-LIFE-160 HOURS
                                                                    50     100     150
                                                                    TIME (HOURS)
                                      .Figure 35.   Hydrolysis data  for 1,2,4-trichlorobenzene
                                                 (adapted from Ellington, et al., 1986).
                                                        dC
                                                                 In
           CCO)
                                                                 -Kt
                                                    at the half-life:
                                                                          C(0)
                                                                       0.5
                                                                               t s 160 hours
thus,  K = 0.69/160
       K =4.3x 103 hr"1
                                       Figure 36.   Calculation of the first-order hydrolysis rate
                                                 constant  for the degradation of 1,2,4-
                                                 trichlorobenzene.
                                      from this data could potentially result in additional
                                      error.

                                      After estimating K, the behavior  of a specific
                                      compound  can be modeled using a form  of the
                                      advection-dispersion equation that includes a first-
                                      order degradation term:
„
D
  a2c
                                                   _v
                                                            sc
                                                          = _  +KC
(3)
                                               42

-------
Although first-order reactions are not always the
most realistic, they are easy to incorporate into
transport models. Hydrolysis  and other chemical
process data on the EPA priority pollutants can be
found in Mabey, et al. (1982).
Sorption
Of all the  chemical processes discussed  in  this
chapter, sorption is probably the most important and
the most studied. In ground-water applications,
sorption of non-polar organics often can be treated as
an equilibrium-partitioning process between the
aqueous phase and the porous medium (Chiou, et al.,
1979). When solute  concentrations are low (e.g.,
either ^ 10-5 Molar, or less than half the solubility,
whichever is lower), partitioning often is described
using a linear Freundlich isotherm (Karickhoff, et
al. 1979; Karickhoff, 1984):
                                               where
                 S = K C
                       P
                                         (4)
where S is the sorbed concentration (mg/g), C is the
aqueous  concentration (mg/mL), and Kp  is  the
partition coefficient. Kp typically is measured in the
laboratory using batch equilibrium tests, and the
data are plotted as the concentration in the aqueous
phase versus the amount sorbed onto the solid phase
(Figure 37) (Chiou, et al., 1979). Under conditions of
o
V)
   1200
 o>
O
<
oc


I
o
o
S
m
oc
o
W
    800
400
               1,1,1 -TRICHLOROETH ANE

                1,1,2,2-TETR ACHLOROETH ANE

                        1,2-DICHLOROETHANE
   0   400  800  1200 1600 2000  2400

       AQUEOUS CONCENTRATION (ug/U
Figure 37.   Batch equilibrium data for 1,1,1-TCA, 1,1,2,2-
           TeCA and 1,2-DCA (adapted from Chiou, et
           al., 1979).
linear equilibrium partitioning, the sorption process
can be represented in the advection-dispersion equa-
tion as a "retardation factor," R:
           32c     ac   ^ac
         D—  -v — =R —
           ax2     *<     at
                                        (5a)
                                                                     ,K
            R= 1 +
                                                                                           (5b)
                                               and pb = soil bulk density (g/cm3) and 0t = porosity.

                                               The dominant mechanism of organic sorption is the
                                               hydrophobic  bond between a chemical and natural
                                               organic matter associated with aquifers (Karickhoff,
                                               et al., 1979; 1984;  Tanford, 1973;  MacKay and
                                               Powers,  1987;  Chiou, et al., 1985). The extent of
                                               sorption can  be reasonably estimated if the organic
                                               carbon content of the  soil is known (Figure 38)
                                               (Karickhoff, et al., 1979; Karickhoff, 1984) by using
                                               the expression:
                                                              K  =K  f
                                                               p    oc oc
                                             (6)
where foc is the fraction organic carbon content of the
soil and Koc is a proportionality constant character-
istic of the specific chemical. This approach works
reasonably well for a wide range of soils, providing
the soil organic content is sufficiently high (e.g., foc
> 0.001). For lower carbon-content soils, sorption of
the neutral organics onto the mineral phase  can
cause important errors in the estimate of Kp (Chiou,
et al, 1985). Still, the Koc relationship allows estima-
tions of sorption-based retardation developed from
measured foc values, rather than the more expensive;
batch equilibrium tests.
       1800

       1500
                                                          1200
                                                PYRENE
                                                       900
                                                       600
                                                       300
                                                                 SLOPE • K
                       OO

                     'PHENANTHRENE
600


500


400

      Kp
300 PHENAN-
    THRENE

200


100
           0.0  .005  .010  .015  .020  .025
             FRACTION ORGANIC CARBON

Figure 38.   Sorption of pyrene and phenanthrene  on
           various soils as  a function of soil foc (after
           Karickhoff, 1981).
Koc values for many compounds  are unknown.
Because  of  this, numerous  researchers have
developed correlation equations to relate Koc to more
commonly available chemical properties, such as
solubility  or octanol-water  partition  coefficient
(Chiou, etal., 1982 and 1983; Schwarzenbach  and
Westall, 1981; Kariekhoff,  1981; Kenaga and Goring,
                                                43

-------
 1980) (Figures 39 and 40). Some of these correlations
 cover a broad range of compounds, in which case the
 errors associated with the Koc estimates can be large.
 However, within a compound class, Koc values
 derived  using these expressions often provide
 reasonable estimates of sorption.
      7

      6

      5

      4

      3

      2

      1

      0

    -1
  log Koc * -0.55 log S + 3.64
               (S in mg/L )
    Kenaga and Goring, 1980
D a
       -4  -3-2-101   2

                   log S (mg/L)
                  4  5
 Figure 39.   Correlation of Koc and  solubility data
            (adapted from Kenaga and Goring, 1980).

                 Regression Equations

           tog KOO - -0.55 log S + 3.64
           log Koc = 0.544 log Kow + 1.377
           tog Koc = 0.681 log BCF + 1.963

 Figure 40.   Regression equations for Koc  versus solu-
            bility (S in  mg/L), octanol/water partition
            coefficient  (Kow), and bioconcentration
            factor (BCF) (after Kenaga and Goring, 1980).
Unfortunately, the linear equilibrium approach to
sorption  is  not adequate for some  real-world
situations. For  example, trichloroethylene (TCE)
sorption onto a glacial till shows a change in Kp of
more than 50-fold over the  parts per billion range
(Figure 41) (Johnson, et al., 1989; Myrand, et al.,
1989; McKay and Trudell,  1987). In addition, for
many very hydrophobia organics, adsorption and
desorption occur over time scales of many  months
(Figure  42) (Karickhoff and Morris,  1985;
Witkowski, et al., 1988; Coates and Elzerman, 1986;
Wu and Gschwend, 1986). The importance  of non-
ideal sorption will be discussed again in Chapter 6.

Field data or field-scale experiments also can provide
good measures of sorption-based retardation.  Field
data have the advantages  of operating on  more
realistic scales,  time frames,  and conditions than
those typically reproduced  in the laboratory. For
                                                  r McKay and Trudell
                                             LOG OF THE FINAL AQUEOUS CONCENTRATION (PPB)
                                   Figure 41.   TCE sorption  on a glacial  till  from  near
                                              Sarnia, Ontario (after  McKay and Trudell,
                                              1987;  Myrand,  et al. 1989; Johnson, et al.,
                                              1989).
                                                DESORPTION OF HEXACHLOROBENZENE
                                   Figure 42.   Desorption of  hexachlorobenzene from two
                                              sediments (after Karickhoff and Morris,
                                              1985).

                                  example, data on chlorophenol transport collected
                                  downgradient from a hazardous chemical disposal
                                  site at Alkali Lake, Oregon (Figure 43) (Johnson, et
                                  al.,  1985), demonstrate compound-specific retarda-
                                  tion that can  be explained by  equilibrium-parti-
                                  tioning sorption. Also, numerous large-scale tracer
                                  experiments were conducted in  recent years  to
                                  demonstrate and quantify certain physical  and
                                  chemical processes (Figure 44) (LeBlanc, et al., 1987;
                                  MacKay, et al., 1986; Barker,  et al., 1987; Molz,  et
                                  al., 1986). However, field data are often expensive, as
                                  well as difficult and time-consuming to gather.


                                  Cosolvation and lonization
                                  Cosolvation and  ionization are processes that can
                                  decrease sorption and, therefore, increase transport
                                  velocity. Cosolvents decrease the entropic forces that
                                  favor sorption of hydrophobic organics by increasing
                                  the interactions between the solute and the solvent
                                  (Nkedi-Kizza, et al., 1985; Zachara, et al., 1988). The
                                  thermodynamic basis for the Cosolvation effect was
                                  described by Rao, et al. (1985), and Woodburn, et al.
                                  (1986). For many of the more  hydrophobic priority
                                                 44

-------
pollutants, the presence of biologically derived or
anthropogenic compounds in the range of 20 percent
by volume or greater can increase the solubility of
those pollutants by an order of magnitude or more.
As seen in Figure 45, this results in a log-linear
decrease in sorption that inversely parallels changes
in the solubility of the solute in the mixed solvent
system (Nkedi-Kizza, et al., 1985). The data for three
soils in Figure 45 clearly show that  the  cosolvent
concentration must be large in order for the  solute
velocity to be  substantially increased. For this
reason, cosolvation is important primarily near
sources of ground-water contamination.

                    ALKALI LAKE
                   2,3,4,5-TETRACHLOROPHENOL
                      DICHLOROPHENOXYPHENOL
                         200
                                400
Figure 43.
        DISTANCE (M)
Chlorophenol distributions dow.ngradient of
the chemical disposal site at Alkali Lake,
Oregon (adapted from Johnson, et al., 1985).
                       STANFORD/WATERLOO
                  CARBON TETRACHLORIDE
                         TETRACHLOROETHYLENE
                      200
                                     400
 Figure 44.
         TIME (DAYS)
Tracer distribution data from the Stanford/
Waterloo tracer  experiment (adapted from
Mackay, et al., 1986).
                                       Acidic compounds, such as phenols or organic acids,
                                       can lose a proton in solution to form anions (Figure
                                       46) that, due to their charge, tend to be very water
                                       soluble (Zachara, et al., 1986). Thus, the  Koc of a
                                       compound like 2,4,5-trichlorophenol can decrease
                                       from 2,330 for the phenol, to near  zero for the
                                       phenolate (Figure 47). Acidic compounds tend to
                                       ionize more as the pH increases. However, for many
                                       compounds like the chlorophenols, substantial ion-
                                       ization can occur at neutral pHs.  Positively charged
                                       organic ions also can be present in ground-water
                                       systems. Zachara, et al. (1986) and Ainsworth, et al.
                                       (1987) demonstrated that in low-carbon soils and
                                              1000
                                                         100
                                                 Kp     10
                                                          0.1
                                                                         ANTHRACENE
                                                                          .2    .3    .4   .5
                                                  Figure 45.
             FRACTION CO-SOLVENT
                  (METHANOL)
         Effect of methanol as a cosolvent on anthra-
         cene sorption for three soils (Adapted from
         Nkedi-Kizza, et al., 1985).
                                                       Koc-2330
                                                                           Koc~°
Figure 46.   Koc values for 2,4,5-trichlorophenol and 2,4,5-
           trichlorophenolate.
                                               45

-------
        2500


        2000


 KOC   1500

        1000


        500
                      2,4,5-
                 TRICHLOROPHENOL
                                  (Roberts, et al., 1986) show a reasonably good fit to a
                                  first-order degradation process.
                 6.0   6.5  7.0   7.5   8.0  8.5

                           P"
 Figure 47.   Koc versus pH for 2,4,5-trichlorophenol.

 soils containing clays, quinoline is sorbed primarily
 by ion exchange.

 Zachara, et al. (1988) and Fu and Luthy (1986a,
 1986b) examined the combined processes of ioniza-
 tion and cosolvation using quinoline as a  model.
 They found that sorption decreased in a log-linear
 fashion when methanol and acetone were used  as
 cosolvents. Zachara, et al. were able to relate the
 increased cosolvent effect of acetone,  relative  to
 methanol, to the greater solvating properties  of
 acetone.
Biodegradation
Although biodegradation is described in greater
detail in Chapters 7 and 8, two chemical-reaction-
related aspects of biodegradation  are discussed in
this section. The first aspect is that biodegradation in
natural systems often can be modeled as a first-order
chemical reaction. Both laboratory and field data
suggest that this is true when none of the reactants
are in limited supply. For example, data from  the
Stanford/Waterloo tracer experiment (Figure  48)
   Cl
   Cl
oc
Cl
Cl
Cl
Cl
c-c
                                                                        TETRACHLOROETHYLENE
                                                                           •  X
                                                                        CARBON TETRACHLORIDE
                                                                             A
                                                                                BROMOFORM
                                                                        DIOHLOROBENZENE
                                                         200   400    600
                                                          TIME (DAYS)
                                           Figure 48.   Degradation of chlorinated organics in  the
                                                      Stanford-Waterloo natural-gradient tracer
                                                      experiment (adapted  from  Roberts, et  al.,
                                                      1986).

                                           The second chemically related aspect of biodegra-
                                           dation is that, like most abiotic reactions, not all
                                           biodegradation reactions result in complete mineral-
                                           ization of the reactants. In many cases, the micro-
                                           organisms that degrade the contaminants produce
                                           an  intermediate chemical which they do not  or
                                           cannot degrade. Other organisms or other conditions
                                           may allow degradation to continue, but the lifetime
                                           of the intermediate chemical  may be significant.
                                           Perhaps the best  known example of this is the
                                           sequential degradation of tetrachloroethylene to the
                                           more toxic vinyl chloride (Figure 49) which is not
                                           readily degraded and tends to accumulate. In the
                                           process, the original moderately sorbing contam-
                                           inant is transformed to a weakly sorbing compound
                                           with a partition coefficient lower than the original
                                           chemical by a factor of approximately eighty.


                                           Volatilization and Dissolution
                                           Spills of volatile organic compounds (solvents  or
                                           petroleum-based liquids) are  very common in all
                                           parts of the United  States. Two important transport
        364                    126                      59

 Figure 49.   Schematic of the sequential degradation of tetrachloroethylene to vinyl chloride.
                                                                               8.2
                                                46

-------
pathways for these materials are volatilization into
the unsaturated zone and dissolution into the ground
water. In addition to being important transport
pathways, volatile organic compounds also carry the
contaminants from the free-product phase into  the
aqueous and vapor phases where  they are  more
amenable to degradation.

The importance of volatilization is determined by the
area of contact between the free product and  the
unsaturated zone, the vapor pressures of the spilled
compounds, and the rate at which the compound
diffuses in the subsurface.  The contact between a
compound and the unsaturated zone is determined
by the nature of the medium (e.g., grain size, depth to
water, water content, etc.) as well as of the compound
(e.g., surface tension  and liquid density).   As
discussed  in Chapter 3, when immiscible liquids
move downward through unsaturated porous media,
portions of the liquid are left behind as  "trapped
residual." This residual provides a very large surface
area  for volatilization.  Laboratory experiments
suggest that vapor concentrations in the vicinity of
the  residual  are  maintained  at saturation
concentrations. Movement of the vapor away from
the residual is typically controlled by molecular
diffusion, as described by Fick's Second Law:
eax2
                        at
                                             (7)
Mass transfer is controlled by the effective diffusion
coefficient, De, of the compound in the porous
medium. De is defined as:
               D  =D T
                e    a
(8)
where Da =  free-air diffusion coefficient, and T; =
tortuosity factor. Tortuosity factors for moist porous
media  can be  determined  experimentally  or
calculated theoretically. Millington (1959) developed
the most widely used theoretical expression for T;:
                       0
                        2.33
                   T =
                        e
                                             (9)
where 9a is the air-filled porosity of the medium.
Thus, diffusion is significantly reduced in high
water-content (low 9a) soils (Figure 50).

In addition to the effect of tortuosity, diffusion may
be further reduced by the partitioning of the vapors
out of the gas phase and into the solid or aqueous
phases. This process can be described in a manner
that is very analogous to retardation due to sorption
      in the saturated zone, with the addition of a term to
      describe the partitioning between the  vapor and
      aqueous phases (Baehr, 1987):
                            PbKHKP
       Figure 50.
                                               0.3
                                              Tortuosity as a function of air-filled porosity
                                              (total porosity  = 0.35)  (adapted  from
                                              Millington, 1959).
       where KH  = dimensionless water-air partition
       coefficient and 9W = volumetric pore-water content.
       Equation 7 can be modified to the form:
                      R
                                                            ac
                                                            at
                                                  (ii)
       As with the saturated zone  R (Equation 5b),  this
       formulation of the retardation factor assumes  that
       partitioning is at equilibrium and the factors Kn and
       KH are not functions of solute concentration. These
       assumptions are probably valid for most unsaturated
       zone conditions.

       When immiscible fluids  reach the capillary fringe,
       their behavior is dictated by the fluids' density
       relative to water (Schwille, 1988; Scheigg, 1984). As
       discussed in Chapter 3, fluids less dense than water
       (LNAPLs) pool up on the water table while dense
       fluids (DNAPLs) penetrate into the ground water.
       Floating pools of LNAPL  also  can provide  sub-
       stantial surface  area  for  volatilization.  Again,
       diffusion frequently controls the mass transfer of
       organics into the vapor phase.

       The transport and fate of DNAPLs that penetrate
       into  the  ground-water  zone  is  controlled by
       dissolution. Anderson (1988) and Anderson, et al.
       (1987) conducted  a series of dissolution experiments
                                                 47

-------
in a three-dimensional physical model (Figure 51). A
cylindrical volume of sand laden with tetrachloro-
ethylene (TeCE) was surrounded by identical sand
without the TeCE. Ground water flowing through
the tank dissolved the TeCE, and concentrations
downgradient of the source quickly rose to saturation
values. Experiments at higher ground-water
velocities  showed that saturation values were

       CYLINDRICAL SPILL OF TeCE
QROUNDWATER,

  FLOW  £f
        EXPERIMENTAL SAND TANK
Figure 51.
     1m x 1m x 1m
Schematic of the experimental apparatus for
measuring dissolution from DNAPL residual in
saturated porous media used by Anderson
(1988).
                                    maintained, even when the flow rate was 1 m/day
                                    (Figure 52). These data suggest that ground water is
                                    flowing relatively unimpaired through the zone of
                                    residual, and that the dissolution process should be
                                    effective,  even at the high ground-water velocities
                                    present during remediation.
Chemical Reactions of Inorganic
Components

Spec/at/on
For organic contaminant studies, researchers are
primarily interested in the total concentration of the
compound in a given phase (e.g., in water vs. in the
aquifer  matrix).  Study of inorganic  compound
behavior, on the other hand, is greatly complicated
by the lack of sufficient knowledge of the total
concentration of material. Inorganic materials can
occur in  many chemical forms, and knowing these
forms or "species" is critical to predicting the
behavior of inorganic compounds (Morel, 1983;
Sposito, 1986).

In ground water, an element may occur in any of the
following six categories of species:
                    10 cm/day
                               30 cm/day
                 100 cm/day
    o
    ,§
    U!
    o
    0)
 200
 180
 160
 140
 120
 100
   80
   60
   40
   20
 Figure 52.
     -20-16-12-8   -40    4    8   12   16   20

           DISTANCE FROM PLUME CENTER  (cm)

 Tetrachloroethylene dissolution data for ground-water velocities of 10 cm/day, 30 cm/day, and 100
 cm/day (adapted from Anderson, et al., 1987).
                                          48

-------
1.  "Free" ions (i.e.,  surrounded only by  water
   molecules)
2.  Insoluble species (e.g.,Ag2S,BaSO4)
3.  Metal/ligand complexes (e.g., A1(OH)2 + ,  Cu-
   humate)
4.  Adsorbed species (e.g., lead sorbed onto a ferric
   hydroxide surface)
5.  Species held on a surface by ion exchange (e.g.,
   calcium ions on clay)
6.  Species that differ  by oxidation state (e.g.,
   manganese (II) and (IV); iron (II) and (III); and
   chromium (III) and (VI)

The mobility, reactivity, biological availability, and
toxicity of metals and  other inorganics depend upon
the speciation; knowing only the total concentration
of an inorganic compound is frequently of little use.
The primary reactions governing these six categories
of inorganic chemicals are discussed below.
Solubility and Dissolution
The dissolution and weathering of minerals deter-
mines the natural composition of ground water. A
useful distinction can be drawn between these two
related phenomena. "Dissolution" refers to the
dissolving of all components within a mineral, for
example:

            halite (NaCl) -> Na + , Cl-

      gypsum (CaSO4 • 2H2O) ->Ca2+, SO4=

 calcite/aragonite (CaCOailimestone) -»Ca2 + , COa",
                     HCO3-

             quartz (SiO2) -» H4SiO4

Dissolution of such minerals  is the source of most
inorganic ions in ground water. The extent of the
dissolution can be estimated from calculations using
thermo'dynamic constants known  as the solubility
products, Ksp. In principal, a mineral can dissolve up
to the limits of its solubility; however, in many cases,
the reactions occur at such a slow rate that  true
equilibrium is never attained (Morgan, 1967).

Natural systems are further complicated by the
simultaneous presence of many minerals containing
common ions. The contribution  of ions from one
mineral affects the solubility of other minerals
containing  the same ion. This  is  the so-called
"common-ion effect." Computer calculation schemes,
such as MINTEQ  (Felmy, et al., 1984), MINEQL
(Westall,  et al., 1976),  or WATEQ2 (Ball, et al.,
1980), yield the equilibrium distribution of chemical
species in the ground water and indicate if the water
is undersaturated, supersaturated, or at equilibrium
with various mineral phases. Some of these  pro-
grams also can be  used to  predict the ionic
composition of ground water in equilibrium with
assumed mineral phases (Jennings, et al., 1982).

"Weathering" is  a  partial dissolution process in
which certain elements  leach out of a mineral,
leaving others behind. The weathering of alumino-
silicates (such  as feldspars) contributes cations,
primarily Ca2 + ,  Mg2 + ,  K + ,  Na + , and silica, to
water and forms secondary weathering products such
as kaolinite and montmorillonite. This weathering of
silicate  minerals increases the alkalinity of the
water; hence, natural ground water often is  more
alkaline than its rainwater origins. Weathering and
dissolution also can  be sources  of contaminants.
Leachates from mine tailings (Hem, 1970) can yield
arsenate, heavy metals, and strong mineral acids.
Also,  leachates from fly-ash piles yield selenium,
arsenate, lithium ions (Li + ),  and heavy metals
(Honeyman, et al.,  1982; Murarka and Macintosh,
1987; Stumm and Morgan, 1981).

The converse of dissolution reactions is the pre-
cipitation  of minerals or contaminants out of
aqueous solution. During precipitation, the least-
soluble mineral is removed from solution as shown
for iron in Figure 53 (Stumm and Morgan, 1981;
Williams,  1985). A  thermodynamic restriction
known as the Gibbs  Phase Rule limits the number of
solid phases that can form from a given solution
(Sillen, 1967). An element is removed by precipita-
tion when its solution concentration saturates the
solubility of one of its solid compounds. If the
solution concentration later  drops below the
solubility limit, the  solid will begin to dissolve until
the solubility level  is attained again. Thus,
contaminants may initially precipitate, then slowly
dissolve later  after  "remediation"  reduced the
                                                   Figure  53.
                    2    46    8    10   12  14

                         PH
           Log C-pH diagram for iron (adapted from
           Stumm and Morgan, 1981).
                                                49

-------
solution concentration. Remediation may take years
to complete under such conditions.

In another scenario, a contaminant initially may be
soluble, but later precipitates after  mixing with
other waters or after contact with other minerals
(Williams, 1985; Drever, 1982;  Palmer,  1989). For
example,  pumping water from an aquifer during
remediation might cause dissolved lead to be
mobilized until it converges and mixes with high
carbonate waters from a different formation. At that
point, much of the lead would precipitate as PbCOs
solid. Changes in the oxidation  state of an element
also can  cause contaminants  to precipitate or
dissolve; this topic is addressed later in this chapter
in the discussion on redox chemistry.
Complexation Reactions
In a complexation reaction, a metal ion reacts with
an anion that functions as a so-called ligand. The
metal and the ligand bind together to form a new,
soluble species called a complex. Transition metals
are the most important metals involved in complexa-
tion (Stumm and Morgan,  1981); alkaline earth
metals only form weak complexes  while alkali
metals essentially do  not form complexes at all
(Dempsey, et al., 1983). The approximate order of
complexing strength of metals is:

Fe(III) > Hg > Cu > Pb > Ni > Zn > Cd > Fe(II)
                > Mn > Ca > Mg

Important  inorganic ligands include most of the
common anions (Hanzlik, 1976), and their strength
depends highly on the metal ion with which they are
complexing. Common ligands are OH-, C1-, SC>4 =,
CO3=, S=, F-, NH3, PO43-, CN-, and polyphosphates.
Inorganic ligands frequently are in great excess
compared to the "trace" metals they bind  and,
therefore, affect metal  chemistry, not vice versa
(Morel, 1983).

Organic ligands generally form much  stronger
complexes than inorganic ligands. Important organic
ligands include synthetic compounds from  wastes
such as amines, pyridines, phenols, and other
organic bases and weak  acids. Natural organic
ligands are mostly humic materials (Stevenson, 1982
and 1985; Hayes and Swift, 1978; Schnitzer, 1969),
and  the complexation  behavior of these  diverse
substances  is difficult to predict (Dzombak, et al.,
1986; Fish, et al., 1986;  Perdue, 1985; Sposito, 1984;
Perdue and Lytle, 1983). Humic materials  are
generally found in significant concentrations only in
shallow aquifers, but in such systems they  may
dominate the metal chemistry of the ground water
(Thurman, 1985).
Equilibrium among reactants and complexes for a
given reaction is predicted by an equilibrium (or
"stability") constant (K) which defines a mass-law
relationship among the species. For example:

Reaction:        Hg2+ + Cl' = HgCl +

Described by:     [HgCl + ]/[Hg2 + ][Cr] = KHgci+
                 = 107.2

For given total  ion concentrations (measured  by
analysis), stability constants can be used to predict
the concentration of all possible species (Figure 54)
(Smith and Martell, 1976).
                         pH
 >-       7.19     7.2  7.5  8.0          8.29
    18
 \Ji
i  20

    22
 *
 u>
 o
                0.1
                          1.0     10.0
                    SALINITY (0/00)
Figure 54.   Speciation of mercury  as  a  function of
           salinity (adapted from Smith  and Martell,
           1976).
Because complexes decrease the amount of free ions
in solution, less metal may  adsorb onto aquifer
matrix or precipitate. That is, the metal is more
soluble because it is mostly bound up in the soluble
complex. Rueter, et al. (1979) found that a metal
undergoing complexation may be less toxic to aquifer
microbes.
Adsorption and Surface Chemistry
A vast amount of surface area exists in an aquifer
and, in many cases, surface adsorption is the most
important process governing toxic metal transport in
the subsurface. Changes in metals concentration, as
well as pH, can have a significant effect on the extent
of adsorption (Figure 55). Unfortunately, a general
model of ionic adsorption on natural surfaces still
has not been developed (Dzombak,  1986; Dzombak
and Morel, 1986). Numerous theories and models of
adsorption exist, but no truly general principles have
been  defined.  Current  models  still  are   highly
                                               50

-------
    100
 s
 CO
    80
     60
 <   40
     20
                 Fe(lll)
                            Pb
                                          Cd
                        4
                       PH
                                6
8
Figure 55.
           Adsorption of metal ions on amorphous silica
           as a function of pH (adapted from Schindler,
           et al., 1976).
empirical and therefore  are calibrated to specific
data (Westall, 1980). Because of this, it is difficult to
base generalized regulations or remediation plans on
the computer-predicted behavior of metal ions.

Despite these shortcomings, some useful approaches
to the problem have been developed. Adsorption data
usually are presented graphically as "isotherms"
(called this because they represent data collected at a
fixed  temperature). Isotherms can  be  fitted to
mathematical representations; the two most common
forms are  the Freundlich  and  the Langmuir
isotherms (Figure 56). The Freundlich isotherm:
                       iN
                                           (12)
is purely empirical, and sorbed (S) and aqueous (C)
concentration data  are  fitted by  adjusting the
parameters K and N. The Langmuir formulation, in
contrast, is based  on  the theory of surface
complexation (Morel, 1983):
            S = S
                      KG
                "maxl + KC
  (13)
where Smax is the maximum amount which can be
sorbed, and K is the partition coefficient.

Surface complexation models represent adsorption as
ions binding to specific chemical functional groups
on a reactive surface. All surface sites may be identi-
cal or may be grouped into different classes of sites
(Benjamin and Leckie, 1981), and each type- of
surface site has a set of specific adsorption constants,
one for each adsorbing compound. Electrostatic
forces at the  surface also contribute to the overall
adsorption constant  (Davis, et al.,  1978) and
sometimes are explicitly included in the adsorption
constant as  the Coulombic term (Stumm and
Morgan, 1981). The binding of ions to the surface is
                                                   logS
                                                                           logC
                                                  Figure 56.
                     Schematic  drawing of  Freundlich and
                     Langmuir  isotherm  shapes  for  batch
                     equilibrium tests.
computed from constants with mass-law equations
identical to those used to calculate solution-complex
formation (Stumm, et al., 1976; Schindler, et al.,
1976; Dzombak and Morel, 1986).

Although this approach is conceptually simple, there
are practical problems in using the models. Model
parameters are effectively data-fitting parameters.
The parameters can fit a specified set of data to a
particular model, but have no true thermodynamic
meaning and,  therefore, no generality beyond the
calibrating data set (Westall, et al., 1980). Some
models contain many "knobs" that can be adjusted to
get the best fit of the experimental data. However,
once a model is calibrated for one set of conditions,
lack of generality may preclude dynamic predictions
for changing  aquifer  conditions. Parameters
calculated using one model should not be used in
another model unless modified according to each
model's assumptions.
                                                  Ion-Exchange Reactions
                                                  Ion-exchange reactions are similar in effect  to
                                                  adsorption, but  have  some  key  distinctions.
                                                  Adsorption is viewed as the coordination bonding of
                                                  metals (or anions) to specific surface sites considered
                                                  to be two-dimensional. In contrast, an ion-exchanger
                                                  is visualized as a three-dimensional, porous matrix
                                                  containing  fixed charges.  Ions  are held by
                                                  electrostatic forces rather than by coordination
                                                51

-------
bonding (Helfferich,  1962). Surface complexation
uses  fixed  stability constants for mass law
calculations whereas ion-exchange "selectivity
coefficients" are strictly empirical (Reichenberg,
1966) and vary with the amount of ion present. Ion
exchange best describes the binding of alkali metals,
alkaline  earths, and some anions to clays and
condensed humic matter (Sposito, 1984; Helfferich,
1962).

Knowledge of ion exchange in soils and aquifers is
very important for understanding the behavior of
major (natural) ions. In principle, ion exchange
theory also is useful for understanding  low-level
contaminant ions. In practice, this theory is very
empirical when applied to these types of ions, and
surface complexation models  probably are better
choices for trace metals.  Ion exchange models need
further development, but may be the most useful
representation of competition among metals for
surface binding (Sposito, 1984).
Redox Chemistry
Reduction-oxidation (redox) reactions involve  a
change in the oxidation state of elements. The level
of that change is determined  by the number of
electrons on  the element transferred during  the
reaction (Stumm and  Morgan, 1981.)  Redox
reactions can greatly affect contaminant transport.
For  example,  in  slightly acidic  to alkaline
environments, Fe(III) precipitates as a  highly
adsorptive solid phase (ferric hydroxide), whereas
Fe(II) is  very soluble and does not retain other
metals. The reduction of Fe(III) to Fe(II), therefore,
releases not only Fe2+ to the water, but also any
contaminants that  were adsorbed to the  ferric
hydroxide surfaces (Evans, at al., 1983; Sholkovitz,
1985). The behavior of chromium and selenium also
illustrates the importance  of redox chemistry to
contaminant movement.  Cr(VI) (hexavalent) is  a
toxic, relatively mobile  anion whereas trivalent
Cr(III) is inert, relatively insoluble, and strongly
adsorbs to surfaces. Selenate (Se(VD) is mobile,  but
less toxic; however, selenite (Se(IV)) is more toxic,
but less mobile.

The  redox state of an aquifer usually is closely
related to the microbial activity and the type of
substrates available to  the organisms. Organic
contaminants provide the reducing equivalents for
the microbes. Because  of the inherently "enclosed"
nature of aquifers, oxygen is readily depleted and
chemically reducing (anaerobic) conditions form. The
redox reactions that occur depend entirely  on  the
dominant electron potential, which is defined by the
primary redox-active chemical species. For example,
the combination of Fe(II)/Fe(III) defines a particular,
narrow  range of electron  potentials, whereas
  S( + IV)/S(-II) defines a more broad range. These
  pairs of chemical species are called redox couples.

  After oxygen is depleted from the water, the most
  easily reduced materials begin to react  and, along
  with  the  reduced product, dictate the dominant
  potential. After that material is  more or  less
  completely reduced, the next most easily reduced
  material  begins to react, and so  on. Microbes
  typically catalyze this series of reactions, and an
  aquifer is described as "mildly reducing" or "strongly
  reducing," depending on where it is in the chemical
  series (Stumm and Morgan, 1981).

  The electron potential of a water body can be given in
  volts  (as the EH), or expressed by the "pe," which is
  the negative logarithm of the electron "activity" in
  the water. The pe sounds complex but is exactly
  analogous to pH (negative log of hydrogen ion
  activity). A convenient way to summarize a set of
  reactions is on a pH-pe (or pH-En) diagram, some-
  times called a Pourbaix diagram. Inspection of the
  diagram yields the predominant redox species at any
  specified pH and EH (Figure 57).

     + 1.0

     +0.8

     +0.6

     +0.4


Eh  +0-2
     +0.0

     -0.2

     -0.4

     -0.6

     -0.8

     -1.0
       %* IX
       *<  xX,
       Svftv. *->%
                      ^
 Figure 57.
02    468    10   12  141

                 pH
 pH-Eh diagram showing the ranges of various
 aquatic environments.
 In this theoretical approach, only one redox couple
 should define the redox potential of the system at
 equilibrium. In practice, this series of events is not
                                              52

-------
clearly defined and many redox couples  not in
equilibrium can be observed simultaneously
(Lindberg and Runnels, 1984). Hence,  it can be
nearly impossible  to predict redox behavior of
chemicals in aquifers from equilibrium predictions.
However, the redox status of an aquifer cannot be
ignored because of the effects discussed above  and
the potential effects on  biodegradation of organic
contaminants. Anaerobic (reducing) conditions  are
inefficient for the degradation of hydrocarbons,  but
reducing conditions favor dehalogenation of chlorin-
ated and other halogenated compounds.

Much more research is needed in the area of redox
chemistry in aquifers because researchers only now
are realizing the importance of these phenomena in
controlling contaminant  behavior. Among the areas
for future research  are the development  of better
measuring techniques for redox potentials,  micro-
biological studies, and new approaches for  computa-
tional models.
References
Ainsworth, C. C., J. M. Zachara, and R. L. Schmidt,
  1987. "Quinoline Sorption on Na-Montmorillonite:
  Contributions of the  Protonated and Neutral
  Species." Clays Clay Miner., Vol. 35, pp. 121-128.
Anderson, M.  A., 1988.  "Dissolution  of Tetra-
  chloroethylene into  Ground Water." Ph.D. Disser-
  tation, Oregon Graduate Center, Beaverton, OR.
Anderson, M. A., J. F. Pankow,  and R. L. Johnson,
  1987. "The Dissolution of  Residual Dense Non-
  aqueous Phase Liquid (DNAPL) from a Saturated
  Porous  Medium."  In: Proceedings, Petroleum
  Hydrocarbons and Organic  Chemicals in Ground-
  water, National Water Well Association and  the
  American Petroleum Institute, Houston, TX,
  November, 1987, pp. 409-428.
Baehr,  A. L., 1987. "Selective Transport  of
  Hydrocarbons in the Unsaturated Zone Due to
  Aqueous and  Vapor Phase Partitioning." Water
  Resources Research, Vol. 23, No. 10, pp. 1926-1938.
Bahr, J. M. and J. Rubin, 1987. "Direct Comparison
  of Kinetic and Local Equilibrium Formulation for
  Solute Transport Affected by Surface Reactions."
  Water Resources Research, Vol.  23, pp. 438-452.
Ball, J. W., D. K. Nordstrom,  and E. A- Jenne, 1980.
  "Additional and  Revised Thermochemical Data
  and Computer Code for WATEQ2: A Computerized
  Chemical Model for Trace and Major Element
  Speciation and Mineral  Equilibria  of Natural
  Waters." U.S. Geological  Survey Water  Resources
  Investigations Nos. 78-116.
Barker, J. F., G. C. Patrick, and D. Major,  1987.
  "Natural Attenuation of  Aromatic Hydrocarbons
  in a Shallow Sand Aquifer."  Ground-water
  Monitoring Review, Winter 1987, pp. 64-71.
Benjamin, M. M. and J. O. Leckie, 1981. "Multiple-
  site  Adsorption of Cd, Cu, Zn,  and  Pb on
  Amorphous Iron Oxyhydroxides." J. Coll. Interface
  ScL, Vol. 79, No. 2, pp. 209-221.

Chiou, C. T., T. D. Shoup, and P. E. Porter, 1985.
  "Mechanistic Roles of Soil Humus and Minerals in
  the Sorption of Nonionic Organic Compounds from
  Aqueous  and Organic  Solutions." Organic
  Geochemistry, Vol. 8, pp. 9-14.

Chiou, C. T., D. W.  Schmedding, and M. Manes,
  1982. "Partitioning of Organic Compounds on
  Octanol-water Systems." Env. Sci.  Tech.,  Vol. 16,
  pp. 4-10.
Chiou C. T., L. J. Peters, and V. H. Freed, 1979. "A
  Physical  Concept of Soil-water  Equilibria for
  Nonionic  Organic Compounds." Science, Vol. 206,
  pp. 831- 832.
Chiou, C. T., P. E. Porter, and D.  W. Schmedding,
  1983. "Partition Equilibria of Nonionic Organic
  Compounds Between Soil Organic  Matter and
  Water." Env. Sci. Tech., Vol. 17, pp. 227-231.
Coates, John T. and Alan W. Elzerman,  1986.
  "Desorption Kinetics for Selected PCB Congeners
.  from  River Sediment." J. Contaminant Hydrology,
  Vol. 1, pp. 191- 210.

Davis,  J. A., R. O. James, and J. O. Leckie, 1978.
  "Surface  lonization and Complexation  at the
  Oxide/Water Interface: I. Computation of Elec-
  trical  Double  Layer  Properties  in Simple
  Electrolytes." J. Coll. Interface Sci., Vol. 63, No. 3,
  480-499.
Dempsey, B. A. and C. R. O'Melia, 1983. "Proton and
  Calcium  Complexation  of Four Fulvic Acid
  Fractions." In: Aquatic and Terrestrial Humic
  Materials, R. F. Christman and  E. T. Gjessing,
  Editors. Ann Arbor Science, Ann Arbor, MI.
Drever, J. I., 1982. The Geochemistry of Natural
  Waters, Prentice-Hall, Englewood Cliffs, NJ.
Dzombak, D. A., W. Fish, and F. M. M. Morel, 1986.
  "Metal-humate Interaction. 1. Discrete Ligand and
  Continuous Distribution Models." Environ. Sci.
  Technol., Vol. 20, pp. 669-675.
Dzombak, D. A. and F. M. M. Morel,  1986. "Sorption
  of Cadmium on Hydrous Ferric Oxide at High
  Sorbate/Sorbent Ratios: Equilibrium, Kinetics, and
  Modelling." J. Colloid Interface Sci., Vol. 112, No.
  2, pp. 588-598.
Dzombak, D. M., 1986. "Towards a Uniform Model
  for Sorption of Inorganic Ions on Hydrous Oxides."
  Ph.D. Dissertation, Department of Civil Engineer-
  ing, Massachusetts Institute of Technology.
                                               53

-------
Ellington, J. J., F. E. Stancil, and W. D. Payne, 1986.
  "Measurement of Hydrolysis  Rate Constants for
  Evaluation of Hazardous Waste  Land Disposal:
  Volume 1. Data on 32 Chemicals." EPA/600/3-
  86/043, Office of Research and Development, U.S.
  Environmental Protection Agency, Athens, GA.
Evans, D. W., J. J. Alberts, and R.  A. Clark, 1983.
  "Reversible Ion-exchange of Cesium-137 Leading
  to Mobilization  from  Reservoir Sediments."
  Geochem, Cosmochim. Acta.,  Vol. 47, No. 11, pp.
  1041-1049.
Felmy, A. R., D. C.  Girvin, and E.  A. Jenne, 1984.
  "MINTEQ: A  Computer Program for Calculating
  Aqueous Geochemical Equilibria." EPA/600/3-84-
  032, Environmental Research Laboratory,  U.S.
  Environmental Protection Agency, Athens, GA.
Fish, W., D. A. Dzombak, and F. M.  M. Morel, 1986.
  "Metal-humate Interactions.  2. Application and
  Comparison of Models." Environ. Sci. Technol.,
  Vol. 20, pp. 676-683.
Fu, J. K.  and R. G. Luthy,  1986a. "Aromatic
  Compound Solubility in Solvent/Water Mixtures."
  J. Environ. Eng., Vol. 112, pp. 328-345.
Fu, J. K. and R. G. Luthy, 1986b. "Effect of Organic
  Solvent on Sorption of Aromatic Solutes  onto
  Soils." J. Environ. Eng., Vol. 112, pp. 346-366.
Hanzlik, R. P., 1976. "Inorganic  Aspects of Biological
  and Organic Chemistry." Academic Press,  New
  York.
Hayes, M. H. B. and R. S. Swift, 1978. "The Chem-
  istry of Soil Organic Colloids." In: The Chemistry of
  Soil Constituents,  D. J. Greenland and M. H. B.
  Hayes, Editors. Wiley Interscience, New York.
Helfferich, F., 1962. Ion Exchange. McGraw-Hill,
  New York.
Hem,  J. D., 1970. "Study and Interpretation of the
  Chemical Characteristics of Natural Water." U.S.
  Geological Survey Water-supply Paper 1473.
Honeyman, B.  D., 1984. "Cation and Anion
  Adsorption at the Oxide/Solution Interface in Sys-
  tems Containing Binary Mixtures of Adsorbents:
  An  Investigation  of the Concept of Adsorptive
  Additivity." Ph.D. Thesis, Stanford University,
  Stanford, CA.
Honeyman, B. D., K. F. Hayes, and J.  O. Leckie,
  1982. "Aqueous Chemistry of As, B, Cr, Se, and V
  with Particular Reference  to Fly-ash Transport
  Water." Project Report EPRI-910-1, Electric Power
  Research Institute, Palo Alto, CA.
Jennings, A. A., D. J. Kirkner, and T. L. Theis, 1982.
  "Multicomponent Equilibrium  Chemistry in
  Ground Water Quality Models." Water Resources
  Research, Vol. 18, pp. 1089-1096.
Johnson, R. L., et al.,  1985.  "Migration of
  Chlorophenolic Compounds at the Chemical Waste
  Disposal Site at Alkali Lake, Oregon - 2. Contam-
  inant Distributions, Transport, and Retardation."
  Groundwater, Vol. 23, No. 5., pp. 652-666.
Johnson, R. L., J. A. Cherry, and J. F. Pankow, 1989.
  "Diffusive Contaminant Transport  in Natural
  Clay: A Field Example and Implications for Clay-
  lined  Waste  Disposal  Sites."  Environ. Sci.
  Technol., Vol. 23, pp. 340-349.
Karickhoff, S. W. and K. R. Morris, 1985. "Sorption
  Dynamics of Hydrophobic Pollutants  in Sediment
  Suspensions." Environ. Toxicol. Chem., Vol. 4, pp.
  469-479.
Karickhoff, S. W., D. S. Brown, and T. A. Scott, 1979.
  "Sorption of Hydrophobic  Pollutants  on Natural
  Sediments." Water Research, Vol.  13, pp. 241-248.
Karickhoff, S. W., 1984. "Organic Pollutant Sorption
  in Aquatic Systems." J. Hydraulic Engineer., Vol.
  110, pp. 707-735.
Karickhoff, S. W., 1981. "Semi-Empirical Estimation
  of Sorption of Hydrophobic Pollutants on Natural
  Sediments and Soils." Chemosphere,  Vol. 10,  pp.
  833-846.
Kenaga,  E. E. and C. A. I.  Goring, 1980. ASTM
  Special Technical Publication 707. ASTM,
  Washington, DC.
LeBlanc, D. R., et al., 1987. "Fate and Transport of
  Contaminants  in Septic-contaminated Ground
  Water on Cape Cod, Massachusetts." U.S.G.S Open
  File Report 87-109, Chapter B.
Lindberg, R. D. and D. D.  Runnels, 1984. "Ground
  Water  Redox Reactions: An Analysis of Equi-
  librium State Applied to Eh Measurements and
  Geochemical Modeling." Science, Vol. 225, pp. 925-
  927.
Mabey, W. R. and T. Mill, 1978. "Critical Review of
  Hydrolysis of Organic Compounds in Water Under
  Environmental Conditions." J. Phys. Chem. Ref.
  Data, Vol. 7, pp. 383-415.
Mabey, W. R., et al., 1982. "Aquatic Fate Process
  Data for Organic Priority Pollutants." Chapter 4,
  EPA/440/4-81-014, Office of Water  Regulations
  and Standards, U.S. Environmental Protection
  Agency, Washington, DC.
MacKay, D. and B.  Powers, 1987. "Sorption of
  Hydrophobic Chemicals from Water: A Hypothesis
  for the Mechanism of the Particle Concentration
  Effect." Chemosphere, Vol. 16, pp. 745-757.
MacKay, D. M.,  et al., 1986. "A Natural Gradient
  Experiment on  Solute  Transport  in a  Sand
  Aquifer.  1. Approach and Overview  of Plume
  Movement." Water Resources Research, Vol.  22,
  No. 13, pp. 2017-2030.
                                               54

-------
McKay, L. D. and M. R. Trudell, 1987: "Sorption of
  Trichloroethylene in Clayey Soils  at the Tricil
  Waste Disposal  Site  near Sarnia,  Ontario.
  Unpublished report,  University  of Waterloo
  Institute for Ground Water Research.
Millington, R.  J.,  1959. "Gas Diffusion in Porous
  Media." Science, Vol. 130, pp. 100-102.
Molz, F. J., et al.,  1986. "Performance and Analysis
  of Aquifer Tracer Tests with  Implications for
  Contaminant Transport Modeling." EPA/600/2-
  86/062, Office of Research and Development, R.S.
  Kerr Environmental Research Lab, U.S. Environ-
  mental Protection Agency, Ada, OK.
Morel, F. M.  M., 1983. Principles of Aquatic
  Chemistry. Wiley Interscience, New York.
Morgan, J. J., 1967. "Application and Limitations of
  Chemical Thermodynamics in Water Systems." In:
  Equilibrium Concepts in Natural Water Systems,
  Advances  in Chemistry Series No. 67, American
  Chemical Society, Washington, DC.
Murarka, I.  P. and  D. A. Mclntosh,  1987. "Solid-
  waste Environmental Studies (SWES):  Descrip-
  tion, Status, and Available Results." EPRI EA-
  5322-SR, Electric Power Research Institute, Palo
  Alto, CA.
Myrand, D., et al., 1989. "Diffusion of Volatile
  Organic Compounds in Natural Clay Deposits."
  Submitted to J. Contam. Hydrology.
Nkedi-Kizza, P., P.S.C. Rao, and A.G. Hornsby, 1985.
  "Influence of Organic Cosolvents on Sorption of
  Hydrophobic Organic Chemicals by Soils."
  Environ. Sci. Technol., Vol. 19, pp. 975-979.
Palmer,  C.  D., 1989. "Hydrogeochemistry of the
  Subsurface."  In:  Chemistry of Ground Water, Carl
  Palmer, Editor. Lewis Publishers, Chelsea, MI. (In
  Review).
Perdue, E. M. and C. R. Lytle, 1983. "A Distribution
  Model for  Binding of Protons and Metal Ions by
  Humic  Substances." Environ. Sci. Technol., Vol.
  17, pp. 654-660.
Perdue, E. M., 1985. "Acidic Functional Groups of
  Humic  Substances." In: Humic Substances in Soil,
  Sediment and Water, G.  R. Aiken, D. M.
  MacKnight,  R.  L.  Wershaw, and P. McCarthy,
  Editors. Wiley Interscience, New York.
Pye, V. I., R. Patrick, and J. Quarles, 1983. Ground-
  water  Contamination in the  United States,
  University of Pennsylvania Press, Philadelphia,
  PA.
Rai, D., et al., 1987. "Inorganic and  Organic
  Constituents in Fossil Fuel Combustion Residues.
  Volume 1: A Critical Review." EPRI-EA-5176,
  Electrical Power Research Institute, Palo Alto, CA.
Rao, P. S. C., et al., 1985. "Sorption and Transport of
 Toxic Organic Substances in Aqueous and Mixed
 Solvent Systems." J. Environ. Quality, Vol. 14, pp.
 376- 383.
Reichenberg, D. 1966. "Ion Exchange Selectivity."
 In: Ion Exchange, Vol.  1., J. A. Marinsky, Editor.
 Marcel Dekker, New York.
Roberts, P. V., M. N. Goltz, and D. M. Mackay, 1986.
 "A Natural  Gradient Experiment  on Solute
 Transport in a Sand Aquifer: 3. Retardation
 Estimates and Mass Balances for Organic Solutes."
 Water Resources  Research, Vol. 22, No. 13,  pp.
 2047-2058.
Rueter, J. G., J. J. McCarthy, and E. J. Carpenter,
 1979. "The Toxic Effect of Copper on Occillatoria
 (Trichodesmium)  theibautii." Limnol.  Oceanogr.,
 Vol. 24, No. 3, pp.  558-561.
Scheigg, H. O., 1984. "Considerations on Water, Oil
 and Air in Porous Media." Water Sci. Technol., Vol.
 17, pp. 467-476.
Schellenberg,  K. C., C. Leuenberger, and  R.  P.
 Schwarzenbach, 1984. "Sorption of Chlorinated
 Phenols by Natural  Sediments and  Aquifer
 Materials." Environ. Sci. Technol., Vol. 18,  pp.
 1360-1367.
Schindler, P. W. and W. Stumm, 1987. "The Surface
 Chemistry of Oxides, Hydroxides, and Oxide
 Minerals." In: Aquatic Surface Chemistry,  W.
 Stumm, Editor. Wiley and Sons, New York, pp. 83-
 110.
Schindler, P. W., et al., 1976. "Ligand Properties of
 Surface Silanol Groups. I. Surface Complex Forma-
 tion with Fe3 + , Cu2 + ,  Cd2 + , and Pb2+."  J. Coll.
 Interface Sci., Vol. 55, No. 2, pp. 469-475.
Schnitzer, M., 1969. "Reactions Between Fulvic Acid,
 A Soil  Humic Compound, and Inorganic Soil
 Constituents." Soil Sci. Soc. Am. Proc., Vol. 33, pp.
 75-81.
Schwarzenbach, R.  and J. Westall, 1981. "Transport
 of Nonpolar  Organic Compounds  from Surface
 Water to  Ground Water: Laboratory Sorption
 Studies." Env. Sci. Tech., Vol. 15, pp. 1360-1367.
Schwille, F., 1988. Dense Chlorinated Solvents  in
 Porous and Fractured Media: Model Experiments.
 Translated by J.F. Pankow. Lewis Publishers,
 Chelsea, MI.
Sholkovitz,  E.  R., 1985.  "Redox-related Geochem-
 istry in Lakes: Alkali Metals,  Alkaline Earth
 Metals, and Cesium-137." In: Chemical Processes in
 Lakes, W.  Stumm, Editor. Wiley-Interscience, New
 York.
Siegrist,  H. and P, L. McCarty, 1987. "Column
 Methodologies  for Determining  Sorption  and
                                               55

-------
  Biotransformation Potential for Chlorinated Ali-
  phatic Compounds in Aquifers." J. Contaminant
  Hydrology, Vol. 2, pp. 31-50.
 Sillen,  L. G., 1967. "Gibbs Phase Rule and Marine
  Sediments." In: "Equilibrium Concepts in Natural
  Water Systems." Advances in Chemistry Series No.
  67, American Chemical Society, Washington, DC.
 Smith, R.M.  and A.E.  Martell, 1976. Critical
  Stability Constants. Plenum, New York.
 Sposito, G., 1986. "Sorption of Trace Metals  by
  Humic Materials in Soils and Natural  Waters."
  CRG Crit. Revs. Environ. Control, Vol. 16, pp. 193-
  229.
 Sposito, G., 1984. The Surface Chemistry of Soils.
  Oxford University Press, New York.
 Stevenson, F. J., 1985. "Geochemistry of Soil Humic
  Substances." In: Humic  Substances  in Soil,
  Sediment and Water, G. R. Aiken, D. M. McKnight,
  R. L. Wershaw, and P. MacCarthy, Editors. Wiley
  Interscience, New York.
 Stevenson, F. J., 1982 Humus Chemistry: Genesis,
  Compositions, Reactions. Wiley Interscience, New
  York.
 Stumm, W.,  H. Hohl,  and  F.  Dalang, 1976.
  "Interaction  of Metal Ions with Hydrous Oxide
  Surfaces." Croat.  Chem. Acta., Vol. 48, No. 4, pp.
  491-504.
 Stumm, W. and J.  J.  Morgan, 1981.  Aquatic
  Chemistry, 2nd edition.  Wiley Interscience, New
  York.
Tanford, C.,  1973.  The Hydrophobic  Effect:
  Formation of Micelles and Biological Membranes,
  Wiley and Sons, Inc., New York.
Thurman, E.  M.,  1985. "Humic Substances  in
  Ground Water." In: Humic Substances in Soil,
  Sediment and Water, G. R. Aiken, D. M. McKnight,
  R L. Wershaw,  and P. MacCarthy, Editors. Wiley
  Interscience, New York.
 Westall, J. C., J. L. Zachary, and F. M. M. Morel,
   1976.  "MINEQL: A Computer Program for the
   Calculation of Chemical Equilibrium Composition
   of Aqueous Systems." Technical Note No. 18,
   Massachusetts Institute of Technology.
 Westall, J. C. and H. Hohl,  1980. "A Comparison of
   Electrostatic Models for  the  Oxide/Solution
   Interface." Advances Coll. Interface Sci., Vol. 12,
   No. 2, pp. 265-294.
 Williams, P. A., 1985. "Secondary Minerals: Natural
   Metal Ion Buffers." In: Environmental Inorganic
   Chemistry, K. J. Irgolic and A. E. Martell, Editors.
   VCH Publishers, Deerfield Beach, FL.
 Witkowski, P. J., P. R. Jaffe,  and R. A. Ferrara, 1988.
   "Sorption arid Desorption Dynamics of Aroclor
   1242 to Natural Sediment." J. Contaminant
   Hydrology, Vol. 2, pp. 249-269.
 Woodburn, K. B., et al., 1986. "Solvaphobic Approach
   for Predicting Sorption of Hydrophobic Organic
   Chemicals  on  Synthetic Sorbents and Soils." J.
   Contaminant Hydrology, Vol. 1, pp. 227-241.
 Wu, S.-c. and P. M. Gschwend, 1986. "Sorption
   Kinetics of Hydrophobic Organic Compounds to
   Natural Sediments and Soils." Env. Sci. Tech.
   Vol. 20, pp. 717- 725.
 Zachara, J. M., et al., 1987.  "Sorption of Binary
  Mixtures of Aromatic Nitrogen Heterocyclic
  Compounds on Subsurface Materials." Environ.
  Sci. TechnoL, Vol. 21, pp. 397-402.
Zachara, J. M., et al., 1986.  "Quinoline Sorption to
  Subsurface Materials: Role of pH and Retention of
  the Organic Cation." Environ. Sci. TechnoL, Vol.
  20, pp. 620-627.
Zachara, J.M., et al., 1988. "Influence of Cosolvents
  on Quinoline Sorption by Subsurface Materials
  and Clays." J. Contaminant Hydrology, Vol. 2, pp
  343- 364.
                                               56

-------
                                        CHAPTERS
          SUBSURFACE CHEMICAL PROCESSES: FIELD  EXAMPLES
                         Richard L. Johnson, Carl D. Palmer, and William Fish
Introduction
Understanding chemical processes in the subsurface
is essential for accurate site characterization and for
the design and implementation of efficient remedi-
ation systems. This chapter will focus on examples of
how chemical processes can control transport and
fate in the subsurface.
Petroleum Hydrocarbons in the
Unsatu rated Zone
Gasoline in the subsurface has been examined from
both experimental and theoretical points of view. For
example, theoretical analyses of spills  by Baehr
(1987), Corpacioglu and Baehr ;(1987), and Baehr and
Corpacioglu (1987) demonstrate that volatilization is
a dominant mechanism of transport for many
hydrocarbon spills (Figure 58). Gasoline is a complex
mixture of hundreds  of compounds, some of which
are more amenable to volatilization than  to
dissolution, and vice versa. Also,  other components
within gasoline are not particularly prone to either
and will tend to persist in the subsurface.

As  discussed in Chapter  5, the  importance  of
volatilization is determined by the composition of the
free-product, the vapor pressure  of the specific
compound, and the rate at which the  chemical
diffuses in the vapor  phase. For example, aromatic
hydrocarbons in  gasoline are moderately volatile,
but also dissolve  readily  into  the  pore water.
Therefore, they diffuse more slowly and may persist
in the unsaturated zone longer than the less-soluble
alkanes and alkenes (Baehr, 1987).

Biodegradation also plays an important role in the
fate of gasoline.  Many components  of gasoline are
                                                           REMAINING
                                                           IN VADOSE ZONE
    50
    40
    30
    20
    10
                                            2000
                        1000
                                            2000
            C5 ALKENES
C5-C6 ALKANES

    C6 NAPHTHENES
                        1000
                        DAYS
                                            2000
 Figure 58.   (a) Transport pathways for subsurface gaso-
            line;  (b)  masses  of  aromatics  in the
            unsaturated zone after a hypothetical spill;
            and (c) masses of hydrocarbons in the
            unsaturated zone after a  hypothetical spill
            (after Baehr, 1987).
readily degraded by subsurface microorganisms. In
the saturated zone, biodegradation frequently makes
the aquifer anaerobic, resulting in much slower rates
of degradation. This trend towards anaerobic condi-
                                               57

-------
 tions is demonstrated in data collected by Wilson, et
 al., (1986) and others at a Traverse City, Michigan
 site (Figures 59 and 60).
   Heart Of
   the Plume
Anaerobic
 Zone
Aerobic
 Zone
"Renovated"
   Zone
Figure 59.   Ground-water quality parameters down-
           gradient from a hydrocarbon spill at Traverse
           City, Michigan (after Wilson, et al., 1986).
In the unsaturated zone, vapor-phase molecular
diffusion can maintain an oxygen supply even at
distances of tens of feet below the ground surface
(Figure 61) (Hult and Grabbe, 1985). The data of
Hult, et al. also show elevated carbon dioxide and
methane concentrations (Figures 61 and 62) which
are the result of biodegradation.  In field experi-
ments, Allen, et al.  (1987) infiltrated  wa'ter
containing simple aromatics into unsaturated porous
media.

Naturally present microorganisms quickly degraded
all of the components. These data,  when contrasted
with the fact that the aromatics are widely observed
in the unsaturated zone, suggest that the  inter-
actions between gasoline and the subsurface envi-
ronment are complex.
Indicator Compounds
RCRA regulations identify a number of indicator
parameters to be used as an early warning system for
ground-water contamination, including specific
conductance, pH,  chloride,  bromide, total organic
carbon (TOG), and total organic halides (TOX). In
general, these parameters were selected because
they are conservative and non-reactive, more or less
                                                    I
                                          10
                                           8
                                           6
                                           4
                                           2
                                           0
                                                             Fe
                                                    I
                                                       200-
                                                   1
                                                      -200
                                                                                     eh
                                                          Heart of
                                                          the Plume
                                                     Anaerobic
                                                      Zone
                             Aerobic      "Renovated"
                             Zone         zone
  Figure 60.   Additional ground-water quality parameters
            downgradient from  a hydrocarbon spill  at
            Traverse City, Michigan (after Wilson et al.,
            1986).
unique to and commonly found  in contaminated
ground water, and easily analyzed. Also, compounds
detectable  at  the leading edge of a contaminant
plume were emphasized.

Most of the RCRA-specified indicator parameters,
however, only partially satisfy the criteria listed
above. Changes in specific conductance, pH, chloride,
and  bromide are 'typically  not  unique to
contaminated groundfwater. Similarly, some natural
total organic carbon .(TOC),  chloride, and  bromide
are present  in virtually all  ground water. TOC  is
particularly problematic in that natural TOC is often
found in ground water in the range of 1 to  10 mg/L,
whereas it  is often desirable  to measure organic
contaminants at the ug/L level.  Total organic halides
(TOX) determinations have a much lower detection
limit than  TOC, but  are  prone to  analytical
difficulties.

Plumb (1985),  Plumb and Fitzsimmons (1984),  and
Plumb and Pitchford (1985) examined leachate data
from a large number of RCRA and Superfund sites to
determine if other components of leachate  would
better serve as indicator parameters. These studies
show that purgeable priority pollutants (PPP) were
found at a high percentage of these sites. The PPP
have  low detection limits and are not normally
present in ground water; therefore, they seem an
excellent alternative to the bulk TOC and TOX
                                                58

-------
                     OXYGIEN (ATM)
            CARBON DIOXIDE (ATM)
    1420 '
Figure 61.  Oxygen and carbon dioxide concentrations in
          the unsaturated zone; above an oil spill at
          Bemidji, Minnesota (after Hult, et al., 1985).
analyses. However, the sum of l\ll priority pollutants
in landfill leachates often represent only 1 to 5
percent of the total dissolved  organic material
present in the ground water (e.g., Reinhard, et al.,
1984; Johnson, et al., 1989'i  The bulk-of .the.
unidentified organic contaminants are made up of a
wide variety of primarily hyclrophilic compounds
(Figure 63). An examination of major  chemicals
produced in the United  Stages (Webber, 1986)
indicates that 7 of the top 35 ape priority pollutants
and only 5 are purgeable priority pollutants (PPP).
Furthermore, many of the tqp  35  chemicals are
significantly more hydrophilic j;han the PPP (Figure
64).  This means that if sorption plays a  role  in
retarding chemical  movement,  there are many
common compounds  that will move more  rapidly
than the PPP. Many of those non-PPP compounds
cannot be detected by standard EPA methods.
Nevertheless, they may have (Significant impacts on
ground-water quality due  to/«their mobility,  solu-
bilityj degradability, toxicity, etc. A better under-
standing of the complexity itff landfill leachate is
necessary before an accurate assessment of the
transport and fate of the wide range of compounds
found there can be made.
              TOTAL VOLATILE
         HYDROCARBONS  (g/m3)
                  METHANE (PPM)
 Figure  62.  Hydrocarbon and methane concentrations in
           the unsaturated zone above  an oil spill in
           Bemidji, Minnesota (after Hult, et al., 1985).
                    Alcohols
                    Analines
                    Acetates
                    Amines
                    Thiols
                    Furans
                    Nitriles
                    Phenols
                    Aldehydes
                    Ketones
                    Acids
 Figure 63.  Polar and ionizable compound  classes
           commonly present in leachates.
Organic Transport on Microparticles
A,growing number of field data suggest that very
hydrophobic compounds can be transported in the
subsurface at rates which are significantly greater
                                              59

-------
Q 15-
U. 13-
§1

§ 9:
Q 7:
g.l
Ul 3
CC .
          444-
                                                              2,3,4,5,6,2'>5'-
                                                      HEPTACHLOROBIPHENYL
         NOT        PRIORITY  OTHER PURGEABLE
       PRIORITY    POLLUTANTS    PRIORITY
      POLLUTANTS                POLLUTANTS
Figure 64.   Estimated retardation factors for common
           organics and priority pollutants (foc = 0.01).

than would be predicted by equilibrium  sorption
(Means and Wijayaratne, 1982). In addition, nearly
all ground water contains some levels of "dissolved"
organic carbon (DOC) as microparticles. (DOC levels
are  commonly 0.1  to  10 mg/L, but  may be
significantly higher in the  vicinity of landfills or
other waste disposal facilities.)

The  importance of microparticle transport of
organics in ground water depends upon the extent of
partitioning to the microparticles. For example,
when the carbon associated with non-settling
particles (NSP) is 1 mg/L, a compound with a Koc of
106 will only experience a decrease in retardation
factor of approximately two. Given the  very large
retardation values common for many of these
compounds,  a factor  of two increase in  velocity
generally will  not be important. However,  if
desorption occurs more slowly than adsorption onto
the microparticles, the relatively rapid movement of
the microparticles could result in significantly
enhanced transport velocities for the hydrophobic
organics.  Laboratory data suggest that  desorption
times can be quite long, however, the processes
controlling the sorption kinetics of hydrophobic
organics on microparticles currently are not well-
understood.

In a related observation,  several  investigators
reported that the partition coefficients of hydro-
phobic organics appear  to  increase  at  low
sediment/water ratios (Figure 65) (Voice,  et al., 1983;
Voice and Weber, 1985; Gschwend and  Wu, 1985).
The  most  satisfactory explanation for this
phenomena is that  the process  of separating
microparticle-bound  organics from the aqueous
phase is incomplete (Gschwend and Wu,  1985). This
effect increases with increasing soil/water ratio, thus
(somewhat counter-intuitively), the  Kp  values
observed at low soil/water ratios may be more
representative of partitioning in the undisturbed
medium.  This observation is important  because
                                                            107.
                                                  (ML/G)
                                                            10V
                                                                   102     1Q3     104

                                                   SEDIMENT CONCENTRATION

                                                                 i(MG/L)
                                                 Figure 65.   Apparertf: decrease in Kp with  increasing
                                                           solids concentration in  batch equilibrium
                                                           tests (adapted from  Gschwend and  Wu,
                                                           1985).
                                                 batch tests  are  frequently  used to estimate
                                                 retardation factors for many very hydrophobic
                                                 compounds at sedilment/water ratios at or above
                                                 those in Figure 65. In those cases, Kp values could be
                                                 underestimated by ah  order of magnitude or more,
                                                 leading to a corresponding overestimation of
                                                 transport velocity arid an underestimation of the
                                                 time required for rertiediation.
                                                Clay Liner Failure
                                                Although clay liners have been a standard method of
                                                containment for land-filled wastes, questions persist
                                                as to their suitability and durability. Three possible
                                                mechanisms of clay-liner failure  are advective
                                                transport, waste-induced changes in permeability,
                                                and molecular diffusion. Advective  transport
                                                through the  liner"can occur if the  liner has
                                                mechanically failedy or if a  very large hydraulic
                                                gradient exists acro.'ss the liner. An example of the
                                                latter is a landfill filled with water, located in the
                                                unsaturated zone. Under these conditions, gradients
                                                of 10 or higher may make  transport velocities
                                                significant, even for at "tight" liner. In such cases,
                                                installation of a gravel sump between two liners can
                                                minimize the hydrautlic gradient across the lower
                                                liner. Advective velocities for a landfill sited in a
                                                saturated medium are expected to be  of relatively
                                              60

-------
minor  importance, ^except where breaks in the
barrier exist.

There is some experimental evidence that solvents
present  in landfills can  degrade  otherwise
impermeable clay  liners (Anderson, et al., 1985;
Brown and Thomas, 1984; Brown and Anderson,
1983; Green, et al.,  1983). The  main question is
whether or not organic solvents can cause the clay to
shrink and crack. Experiments indicate that if pure
solvents penetrate into the clay, they may be able to
displace water within the clay structure, leading to
shrinkage, cracking, and an increase in perme-
ability. In these experiments, the  solvents were
forced into the clay at high pressures equivalent to
many tens of meters  of fluid head. This behavior is
consistent with the critical pressure required for
entry of the solvents into the clay (estimated from
the grain diameter of the clay) (Chapter 3; Villaume,
et al.,  1983), but is probably not realistic in  most
landfill situations.

In the  absence of sufficient pressure, entry of the
solvents occurs by  simple Fickian diffusion in the
aqueous  phase. However, the extent of entry is
limited by the solubilities of the solvent in the water
and vice  versa. For many solvents, solubilities are
thousands of milligrams per  liter or less, in which
case, the impact  of the  solvents  6n the clay is
expected to be minimal.

The  presence of mechanical failures or  other
imperfections in the clay may provide focal points for
degradation of the clay liners. Miscible liquids, such
as alcohols, acids, bases, and ketones, also have the
potential to affect clay structure. Because they are
not limited by solubility, large concentrations can
diffuse into the clay.  Miscible compounds generally
are polar  molecules which can interfere  with the
electrostatic forces within the clay  (Fernandez and
Quigley,  1985). This interference can result in  a
breakdown of the  clay structure and subsequent
failure of the liner.

Even though chemical concentrations of thousands of
milligrams per liter within clay  liners  are not
expected  to affect the hydraulic  properties of the
clay, such concentrations could have a major impact
on ground-water quality because of the  diffusive
movement of chemicals through the liner. In the
absence of advection, mass transport is controlled by
Pick's Second Law:
                                           \ a2c
                                          R
                                 dt
                                                                      (2)
                          A  number  of  field studies  demonstrated the
                          importance of simple Fickian diffusion (Johnson, et
                          al., 1989; Goodall and Quigley, 1977; Crooks and
                          Quigley, 1984; Desaulniers, et al., 1981). Johnson, et
                          al. collected clay cores from beneath a five-year-old
                          waste disposal cell and  determined concentration
                          profiles for chloride  and several  organic contam-
                          inants. The chloride diffused nearly one meter while
                          the organics moved  much shorter distances.  The
                          principal reason for the  slower  diffusion of the
                          organics was their sorption onto the clay (foc~0.01).
                          For non-sorbed contaminants  (including organics
                          when the foc is low), breakthrough of a one-meter
                          liner could occur in less than 10 years. Within a few
                          decades, steady-state concentrations could develop
                          across the clay, possibly resulting in significant mass
                          transfer into the underlying aquifer.

                          Under steady-state conditions, mass flux follows
                          Pick's First Law:
                                                  ac
                                             D9  —
                                               e
                                                    (3)
                          As the calculation in Figure 66 suggests, the mass
                          flux through a liner via diffusion can be on the order
                          of 1 gram per square meter per year. For a large
                          liner, a mass transfer of thousands of grams of
                          individual contaminants per year to the aquifer
                                      1000 ppm
                                                                     1m
                                         0 ppm

                                           dC
                                                 t
                            = -0.37 x 1.5x10"6 cm2 /sec x  1g/L-m
               D
                e ,2
                  dX
ac
dt
(1)
As seen in Chapter 5, for sorbing compounds, Fick's
Law can be modified to  handle  equilibrium
partitioning:
                                                    = 5.55x10   g/cm /sec
                           •» 1.75 g/m /year

                          Figure 66.   Example calculations of steady-state diffu-
                                     sive mass flux through a one-meter thick clay
                                     liner.
                                                61

-------
 could occur. Similar conditions  can exist  across
 natural aquitards as well, and the  large contact area
 between aquifers could result in  significant mass
 transfer into an uncontaminated aquifer.
 "Plume Sniffing"
 Because soil vapor samples are relatively easy to
 collect from the subsurface, interest has grown in
 using soil-gas sampling or "plume sniffing" to search
 for ground-water contamination  (Marrin and
 Kerfoot, 1988). Shallow soil-gas sampling detects
 vapors emitted into the unsaturated zone by ground-
 water contaminant plumes that contain volatile
 organics. Many factors affect vapor concentrations in
 the unsaturated zone. For example, vertical trans-
 verse dispersion in the saturated zone may control
 mass flux into the unsaturated zone. Because air-
 diffusion coefficients are much greater than the
 values typically reported for vertical dispersion (10-2
 cm2/s vs. 10'6 cm2/s), unsaturated zone vapor concen-
 trations might be expected to be low. Nevertheless,
 substantial unsaturated concentrations  have been
 reported at several sites. Swallow and Gschwend
 (1984) examined the steady-state distribution  of
 volatile organic vapors in the ground  water and
 unsaturated zones (Figure 67).  For vertical trans-
 verse dispersivities (atv) in the ground water of 1 cm
 or less, concentrations at steady-state were found to
 be near  zero through the  unsaturated zone.  A
 limitation of the steady-state approach is that data
 are not provided during the time required for the
 system to reach  steady-state.  For the  10-meter
 aquifer in the Swallow and Gschwend model system,
 with atv = 1 cm, hundreds of years would be required
 before the system reached the steady-state.

 As previously mentioned, vapor concentrations  in
 the unsaturated zone may deviate from idealized
ui
ui
                                      UNSATURATED
                                      CAPILLARY
SATURATED
        CONCENTRATION
Figure 67.   Schematic drawing  of Steady-state concen-
           tration profiles for several vertical dispersion
           values  (adapted  from  Swallow and
           Gschwend, 1984).
             values due to environmental conditions or biodegra-
             dation. Data from Hult and Grabbe (1985)  (Figure
             68) for the Bemidji, Minnesota site show that vapor
             concentrations at the depths typically used for plume
             sniffing  (3  to  6  feet) are very low because of
             biodegradation. This is the case even though a large
             organics  source is present as free-product at the
             capillary fringe. Vapors from ground-water  plumes
             also can be masked by clean water which forms a cap
             over the plume. This occurs if local recharge
             displaces the contaminants from the capillary fringe,
             or if the  plume moves downward into  the  ground
             water.

             Numerous case histories demonstrate the usefulness
             of plume sniffing (Marrin and  Thompson, 1987;
             Marrin and Kerfoot, 1988), especially for chlorinated
             solvents.  As a result,  plume sniffing continues to
             receive a  great deal of attention, despite the fact that
             its general applicability has  not been  thoroughly
             demonstrated.  As  Marrin and Kerfoot  (1988)
             correctly  point out:

               Conventional technologies available  for sub-
               surface investigation (e.g.,  monitoring .wells
               and soil borings) always will  be required to
               confirm and monitor subsurface contamina-
               tion; however, quicker and less expensive tech-
               niques are useful for preliminary site evalu-
               ations.

             They  also suggest that, while  positive results
             generally suggest detection of ground-water plumes,
             failure to detect plumes does not assure that ground-
             water contamination is not present.
Ground-Water Contamination by
Chromium
Chromium (Cr)  in the environment causes great
concern because of its wide use in industry and its
potentially high toxieity. Although chromium can
exist in oxidation states ranging from -2 to + 6, the
trivalent (Cr(III)) and hexavalent  (Cr(VT))  species
are the common stable forms found in the environ-
ment. Three recent surveys of the environmental
significance of chromium (U.S. EPA, 1983;  Radian
Corp., 1983; the American Petroleum Institute,
1981) emphasize that the chemical form of this metal
determines its environmental behayior and toxieity.
Cr(III) is relatively insoluble and exhibits little or no
toxieity (van Weerelt,  et al.,  1984), while Cr(VI)
usually occurs as the  highly soluble CrO42-, HCrO^,
and Cr2O72- anions.

Cr(III) is the most common form of chromium in the
earth's crust; the predominant source of hexavalent
chromium  in the environment is anthropogenic
activities.  For example, industry has used Cr(VI)
                                                62

-------
                         TOTAL VOLATILE HYDROCARBONS
 UJ
 O
 <
 1L
 DC
 O

 Hi
 UJ
 LU
 IL
         30
                         RELATIVE CONCENTRATION
Figure 68.
Vertical hydrocarbon profiles in the unsaturated zone above an oil spill near Bemidji, Minnesota site (adapted
from Hult, et al., 1985) .
primarily in metal plating and leather tanning
applications for over 100 years.  As  a result,
numerous waste lagoons, dumps, and landfills are
contaminated with chromate wastes (Black and Heil,
1982; Cook and DiNitto, 1982; Owen, 1982; Massa-
chusetts  Department of Environmental Quality
Engineering, 1981; Keely and Boateng,  1987). At
many sites, the chromium coexists with a variety of
other inorganic and organic  wastes, and under these
conditions, a wide range of chemical interactions are
possible.                   t

Chromium transport in aqueous systems strongly
depends on sorption, chelation, and redox reactions.
The redox reactions are only poorly understood, yet,
they are of key importance because the redox state of
chromium dictates  its  sorptive and chelation
behavior. Despite the strong tendency for chromium
to partition to  many  mineral surfaces,  there are
reports of extensive chromium migration (Keely and
Boateng,  1987).

Surface  sorption is an important element of
chromium behavior.  Electrostatic forces near the
                                      solid-liquid interface form a transition range near
                                      neutral pH in which both positively and negatively
                                      charged surface sites coexist (Swallow, et al., 1980).
                                      Several published studies discuss adsorption of
                                      chromium species at oxide interfaces (Bartlett and
                                      Kimble, 1976; Bartlett and James, 1979; Huang and
                                      Wu, 1977). If the mineral composition of the aquifer
                                      is known, appropriate sorption parameters for Cr(III)
                                      or Cr(VI) can be roughly estimated. Unfortunately,
                                      the redox state of the chromium also must be known
                                      and this is difficult to predict because natural redox
                                      chemistry of Cr(HI)/Cr(VI) is not well understood.

                                      Although  Cr(III) and Cr(VT) are each quite stable
                                      and tend to be kinetically inhibited from undergoing
                                      redox transformations,  there are systems which
                                      catalyze oxidation-reduction reactions of chromium.
                                      In strong acid solutions,  Cr(VI) will oxidize organic
                                      compounds and be reduced to  Cr(III) (Bartlett and
                                      Kimble, 1976). This reaction eliminates toxic Cr(VI)
                                      and generates the  relatively insoluble trivalent
                                      species. The reverse reaction can be driven by MnC>2
                                      in a process in which the MnOg appears to act as both
                                      an oxidizing agent and a catalytic surface (Bartlett
                                              63

-------
 and James, 1979). James and Bartlett (1983)
 reported the oxidation of Cr(III)  in a mixture of
 tannery sludge and moist soil  containing MnC>2.
 Thus, Cr(VI) may be generated at waste sites in
 which only Cr(III) is deposited. It is possible that at
 an actual waste disposal site, the organic oxidation
 and Mn02 reduction reactions may form a cycle.

 Another chromium/organic association that is
 largely unexplored yet potentially important to
 chromium behavior at waste sites is the solvation of
 chromate by organic phases. Essentially all studies
 of chromium in the environment focus on the  solid-
 or aqueous-phase chromium. However, chromate is
 typical of a class of relatively large  inorganic anions
 that dissolve to an appreciable extent  in certain
 organic solvents. This  phenomenon is well  docu-
 mented for highly acidic conditions in  analytical
 solvent extraction studies (Zolotov, et al., 1967).
 Recent studies show that organic solvation  of
 chromate occurs even near neutral pH; the phe-
 nomenon,  therefore,  has wider  environmental
 significance than would be deduced  from the solvent
 extraction  literature.  Reactions also may  be
 enhanced by the close  association of the chromate
 with the solvent.

 Partitioning of chromate into non-aqueous  phases
 may actually increase the mobility of chromium in
 the environment. Both Cr(VI) and dense chlorinated
 solvents are frequently associated with plating
 operations. This suggests that the  chromate could
 become stabilized in  the dense solvents and  be
 transported well below the water table.

 In summary, reactions between solvated Cr(VI) and
 organic compounds may affect the composition and
 behavior of ground-water contaminants by:  (1)
 removing more hexavalent chromium from the
 aqueous phase than would occur by simple solvation;
 (2)  stabilizing  Cr(VI)  in organic phases; and  (3)
 mediating the oxidation of organic waste components
 into new compounds with greater or lesser toxicity
 and mobility.
Ground-Water Contamination by Leaded
Gasoline
The subsurface movement of lead is expected to be
minimal in most aquifers because of its low solubility
and strong tendency to sorb to aquifer materials.
However, studies have found that gasoline-derived
lead can be transported over hundreds of meters. The
mechanism(s) of this transport remain  unclear,
although  transport as  the  alkyl lead,  on
microparticles, and/or in  the free-product gasoline
may be involved. An example case history of lead
movement was conducted at the site of a large
gasoline spill in Yakima, Washington (Fish, 1987).
 Up to 20,000 gallons of leaded and unleaded gasoline
 were released over several years. At the  time of
 sampling, the lead had  been transported over 100
 meters in a heterogeneous alluvial aquifer.

 At least three hypotheses have been made to explain
 these data. The first is that some of the lead is
 associated with fine-colloidal particles transported
 through the aquifer. The presence of large quantities
 of lead in the solid phase would support this possi-
 bility. However, the relatively  low concentrations
 observed in ground water filtered  through a 0.1 um
 filter suggests that if particle transport is the mecha-
 nism, the bulk of the particles must be greater than
 0.1p.m.

 A second hypothesis is that the  bulk of the lead
 transport occurred when the lead existed  in less
 strongly sorbed alkylated forms. However, no alkyl-
 ated lead could be found in the aquifer at the time of
 sampling.

 A third hypothesis is that the bulk of the transport of
 the lead, as well as the organics,  took place while
 both were associated with the free-product gasoline.
 The volume of the spill is such that the product could
 have  extended over significant distances. However,
 as discussed in Chapter 3,  the erratic behavior of
 immiscible phases in the subsurface is poorly under-
 stood. At the time of sampling, free product could be
 observed in wells at distances of up  to 100 m from the
 source. The highly heterogenous, interbedded nature
 of the aquifer also might allow relatively narrow,
 preferential pathways for free-product movement
 that could enhance traveling distances.
References
Allen, R. M., R. W. Gillham, and J. F. Barker, 1987.
  "Remediation of Gasoline-contaminated Ground
  Water  by Infiltration Through Soil." National
  Water Well Association, Fourth Annual Eastern
  Regional Ground Water Conference, July 14-16,
  1987, Burlington, VT.

American Petroleum Institute (API), 1981. Sources,
  Chemistry,  Fate, and Effects  of Chromium  in
  Aquatic Environments, American Petroleum
  Institute.     ,

Anderson, D.  C., K. W. Brown, and J. C. Thomas,
  1985. "Conductivity  of Compacted Clay Soils  to
  Water and Organic Liquids." Waste Manage. Res.,
  Vol. 3, No. 4, pp. 339-349.

Baedecker, M. J.,  et al., 1985.  "Geochemistry of a
  Shallow Aquifer Contaminated with Creosote
  Products." In: Proceedings, Second U.S. Geological
  Survey  Toxic Waste Technical Meeting, Cape Cod,
  MA, October, 1985.
                                               64

-------
Baehr, A. L. and  M. Y. Corpacioglu, 1987. "A
  Compositional Multiphase Model for Ground-water
  Contamination  by Petroleum  Products.  2:
  Numerical Solution." Water Resources Research,
  Vol. 23, No. 1, pp. 201-213.
Baehr, A. L., 1987. "Selective Transport of Hydro-
  carbons in the Unsaturated Zone Due to Aqueous
  and Vapor Phase Partitioning.1'  Water Resources
  Research, Vol. 23, No. 10, pp. 1926-1938.
Bartlett, R. and B.  James,  1975). "Behavior of
  Chromium in Soils: III. Oxida.tion." J. Env.
  Quality, Vol. 8, pp. 31-35.
Bartlett, R. and J.  M. Kimble, 197:6. "Behavior of
  Chromium in Soils: II. Hexavalerit Forms." J. Env.
  Quality, Vol. 5, pp. 383-388.
Black, J. A. and J. H. Heil,  1982.  "Municipal Solid
  Waste Leachate and Scavenger Waste: Problems
  arid Prospects  in Brookhaven Town."  In:
  Proceedings, N.E. Conference  on the Impact of
  Waste Storage and Disposal on  Ground-water
  Resources, R. P. Novitsky and G. Levine, Editors.
  U.S. Geological Survey/Cornell University.
Brown, K. W. and D.  C. Anderson,JL983. "Effects of
  Organic Solvents on the PermeaTbility  of Clay
  Soils." EPA/600/2-83-016,  Office of Research and
  Development, U.S. Environmental Protection
  Agency.                       .T,.
Brown, K. W. and J. C. Thomas, 1984. "Conductivity
  of Three Commercially Available Clays to Petrol-
  eum Products and Organic Solvents." Hazard.
  Waste, Vol. 1, No. 4, pp. 545-553.
Cook, D. K. and R. G. DiNitto, 1982. '"Evaluation of
  Ground Water Quality in East and'North Woburn,
  MA."  In: Proceedings, N.E. Conference on the
  Impact of Waste Storage and Disposal on  Ground-
  Water Resources, R. P.  Novitsky and G. Levine,
  Editors.  U.S.  Geological Survey/Cornell  Uni-
  versity.                   ;
Corpacioglu, M. Y.  and A.  L.  Bae.hr, 1987. "A
  Compositional Multiphase Model for Ground-water
  Contamination by Petroleum Products.  1: Theo-
  retical Considerations." Water Resources Research,
  Vol. 23, No. 1, pp. 191-200.
Crooks,  V.  E. and R. M. Quigley,  1984.  "Saline
  Leachate Migration through Clay:  A  Comparative
  Laboratory and Field Investigation." Can. Geotech.
  J., Vol. 21, pp. 349-362.
Desaulniers, D. E., J. A. Cherry, and P. Fritz, 1981.
  "Origin, Age, and Movement  of  Pore Water in
  Argillaceous Quaternary Deposits at-JFour Sites in
  Southwestern Ontario." J. Hydrology, Vol. 50, pp.
  231-257.
Fernandez, F.  and R. M.  Quigley, 1985. "Hydraulic
  Conductivity of  Natural Clays Permeated with
  Simple  Liquid Hydrocarbons." Can. Geotech. J.,
  Vol. 22, No. 2, pp. 205-214.
Fish, W., 1987. "Subsurface Transport of Gasoline-
  derived Lead at a Filling  Station Contamination
  Site in Yakima,  Washington." In:  Proceedings,
  NWWA FOCUS:  Northwestern  Ground  Water
  Issues Conference, National Water  Well Associ-
  ation, Portland, OR, May 1987.
Godsy E. M. and D..F. Goerlitz, 1984. "Anaerobic
  Microbial Transformations of Phenolic and Other
  Selected Compounds in Contaminated Ground
  Water at a Creosote Works, Pensacola, FL." In:
  Movement and Fate of Creosote Waste in Ground
  Water, Pensacola, FL, U.S.  Geological Survey Open
  File Report 84-466.
Goodall, D. C. and  R. M. Quigley,  1977. "Pollutant
  Migration from Two Sanitary Landfill Sites near
  Sarnia, Ontario." Can. Geotech. J., Vol. 14, pp. 223-
  236.
Green, W. J., et al., 1983. "Interaction of Clay Soils
  with Water and Organic Solvents: Implications for
  the  Disposal of Hazardous Wastes."  Environ. Sci.
  TechnoL, Vol. 17, No. 5, pp. 278-282.
Gschwend, P.  M. and S.-c. Wu.,  1985. "On the
  Constancy of Sediment-water Partition  Coeffi-
  cients of Hydrophobic Organic Pollutants." Env.
  Sci. Tech., Vol. 19, pp. 90-96.
Huang, C. P. and M. H. Wu, 1977.  "The Removal of
  Chromium (VI) from Dilute Aqueous Solution by
  Activated Carbon." Water Res., Vol. 11, pp. 673-
  678.
Hult, M. F. and R. R. Grabbe, 1985. "Permanent
  Gases and Hydrocarbon Vapors in the Unsaturated
  Zone." In: Proceedings, U.S.  Geological Survey
  Second Toxic-Waste Technical Meeting, Cape Cod,
  MA, October, 1985.
James, B. R. and R. J. Bartlett, 1983. "Behavior of
  Chromium in  Soils: IV.  Interactions Between
  Oxidation-reduction and Organic Complexation."
  J. Env. Quality, Vol. 12, pp. 173-176.
Johnson, R. L., C. D. Palmer, and J. F. Keely, 1987.
  "Mass Transfer of Organics Between Soil,  Water,
  and Vapor Phases: Implications for Monitoring,
  Biodegradation,  and  Remediation." In:
  Proceedings, Petroleum Hydrocarbons and Organic
  Chemicals in Ground Water, National Water Well
  Association and the American Petroleum Institute,
  Houston, TX, November, 1987.
Johnson, R. L., J. A. Cherry, and J. F. Pankow, 1989.
  "Diffusive Contaminant  Transport in Natural
  Clay: A Field Example and Implications for Clay-
  lined Waste  Disposal  Sites." Accepted for
  publication in Environ. Sci. TechnoL
                                                65

-------
 Keely, J. F. and K. Boateng, 1987. "Monitoring Well
   Installation, Purging, and Sampling Techniques.
   Part 2: Case Histories." Groundwater, Vol. 25, No.
   4, pp. 427-439.
 Marrin, D. L. and G. M. Thompson, 1987. "Gaseous
   Behavior  of TCE  Overlying a Contaminated
   Aquifer." Groundwater, Vol. 25, pp. 21-27.
 Marrin, D. L. and H. B. Kerfoot, 1988. "Soil-gas
   Surveying Techniques." Environ. Sci.  Technol.,
   Vol. 22, No. 7, pp. 740-745.
 Massachusetts Department of Environmental
   Quality  Engineering (DEQE), 1981.  "Interim
   Report on Chemical Contamination in Massa-
   chusetts."
 Mattraw, H. C. and B. J. Franks, 1984. "Description
   of Hazardous Waste Research at a Creosote Works,
   Pensacola, FL." In: Movement and Fate of Creosote
   Waste in Ground  Water,  Pensacola FL, U.S.
   Geological Survey Open-File Report, pp. 84-466.
 Means, J. C. and R. Wijayaratne, 1982. "Role of
   Natural Colloids in the Transport of Hydrophobic
   Pollutants." Science, Vol. 215, pp. 968-970.
 Owen, K. C. 1982. "Ground-water Resources  in
   Niagara Falls, NY and the Potential Impacts of
   Hazardous Waste Contamination." In: Proceed-
   ings, N.E. Conference on the Impact of Waste
   Storage and Disposal on Ground-water Resources,
   R. P. Novitsky and G.  Levine, Editors. U.S.
   Geological Survey/Cornell University.
 Plumb, R. H., 1985. "Disposal Site Monitoring Data:
  Observations and Strategy Implications."  In:
  Proceedings, Second Annual Canadian/American
  Conference on Hydrogeology: Hazardous Wastes in
  Ground Water — A Soluble Dilemma, National
  Water Well Association, Dublin, OH, June, 1985.
 Plumb R.  H. and C. K.  Fitzsimmons,  1984.
  "Performance Evaluations of RCRA  Indicator
  Parameters." In: Proceedings, Second Annual
  Canadian/American Conference on Hydrogeology:
  Practical Applications of Geochemistry, National
  Water Well Association, Dublin, OH, June, 1985,
  pp. 125-137.
 Plumb, R. H. and A. M. Pitchford, 1985. "Volatile
  Organic Scans: Implications for Ground-water
  Monitoring." In: Proceedings, Petroleum Hydro-
  carbons and Organic Chemicals in Ground Water,
  National Water Well Association and the Amer-
  ican Petroleum Institute, Houston, TX, November,
  1985, pp. 13-15.
Radian Corp., 1983.  Estimates of Population Expo-
  sure  to Ambient Chromium Emissions, Radian
  Corp., Durham, NC.
 Reinhard, M., N. L. Goodman, and J. F. Barker,
   1984. "Occurrence and Distribution of Organic
   Chemicals in  Two Landfill Leachate  Plumes."
   Environ. Sci. Technol., Vol. 18, No. 12, pp. 953-961.

 Swallow, J.  A. and P. M. Gschwend, 1984. "Volatil-
   ization of  Organic Compounds from Unconfined
   Aquifers." In: Proceedings, Petroleum Hydro-
   carbons and Organic  Chemicals in Ground Water,
   National  Water Well  Association  and  the
  American  Petroleum  Institute, Houston, TX,
  November 1984.,

 Swallow, K. C.,  D. N.  Hume, and F. M. M. Morel,
   1980. "Sorption, of Copper and Lead by Hydrous
  Ferric Oxide." Environ. Sci.  Technol., Vol. 14, No.
  8, pp. 1326-13311..

 U.S.  Environmental  Protection  Agency, 1983.
   "Health Assessment Document for Chrome." EPA-
   600/83-104A.

 van Weerelt, M., W. C. Pfeiffer, and M.  Fiszman,
  1984. "Uptake and Release of Chromium (VI) and
  Chromium (HI) by Barnacles." Mar. Environ. Res.,
  Vol. 11, No. 3, pp. 201-211.

 Villaume, J. F., IP: C. Lowe, and D. F. Unites, 1983.
   "Recovery of' Coal  Gasification Wastes:  An
   Innovative Approach." In:  Proceedings, Third
  National Symposium on Aquifer Restoration and
  Ground-water ^Monitoring, National Water Well
  Association, V/orthington, OH, pp. 434-445.

 Voice, T. C., C.rJp; Rice and W. J. Weber, Jr., 1983.
  "Effect of Solids Concentration  on the  Sorptive
  Partitioning of Hydrophobic  Pollutants in Aquatic
  Systems." Em*-. Sci. Technol., Vol 17, pp. 513-518.

 Voice, T. C.  anjd W. J.  Weber, Jr., 1985.  "Sorbent
  Concentration!  Effects in Liquid/Solid Partition-
  ing." Env. Sci, Technol., Vol 19, pp. 789-796.

 Webber, D.,  19;86. "Top 50 Chemicals Production
   Dropped Modlerately  in 1985." Chem. Eng. New,
   April 21,198B,-pp. 13-15.
Wilson, B.  H./ et al.,  1986.  "Biological Fate of
   Hydrocarboris at an Aviation Gasoline Spill Site."
   In:  Proceedings, Petroleum Hydrocarbons and
   Organic Chemicals in Ground Water, National
  Water Well Association and  the American Petrol-
  eum Institute, Houston, TX, November, 1986.
Zolotov, Y. A.,[V. V. Bagreev, and O. M. Petrukhin,
  1967. "Solvent Extraction Chemistry."  In: Pro-
  ceedings, Inil.  Conference on Solvent Extraction
  Chemistry, Gothenburg, North-Holland,  Amster-
  dam.      '•       '•
                                              66

-------
                                       CHAPTER?

       MICROBIAL ECOLOGY AND  POLLUTANT BIODEGRADATION IN
                            SUBSURFACE ECOSYSTEMS

                                       Joseph M. Suflita
Introduction
Interest in ground water-related sciences is growing,
and individuals involved with all facets of ground^
water use and protection are developing a greater
awareness of the  chemistry, physics, hydrology,
geology, and biology of the terrestrial subsurface.
Information gathered from ground-water contam-
ination sites helps to focus questions regarding the
possibility that aquifers possess natural attenuation
mechanisms for pollutant abatement. Consideration
of this possibility has led investigators to realize that
natural physico-chemical processes may,  at best,
only partially transform aquifer contaminants.

Parallel to the general interest in ground water came
the beginnings of serious study of ground-water
microbiology. Because of the biochemical versatility
of natural microbial communities, investigators
thought that the microflora indigenous to aquifers
might function to remove problem contaminants. In
fact, microbial metabolism of ground-water pollu-
tants is the only technology that has the potential to
completely degrade pollutants in situ and convert
them to more environmentally acceptable forms.

This chapter will introduce the historic and current
scientific perspectives regarding microbial ecology of
the terrestrial subsurface and will focus on how these
perceptions evolved. Examples will be given of the
diverse types of subsurface microorganisms, micro-
bial communities, and associated metabolic activ-
ities. Also, the metabolic principles governing pollu-
tant biodegradation  in  other habitats  will .be
extrapolated to ground water. Limits  to pollutant
biodegradation will be considered in the context of
existing environmental conditions, physiology of the
indigenous microflora, and chemical structure of the
contaminants. Finally, this chapter will discuss how
these principles can be applied to either the in situ or
above-ground bioremediation of contaminated
aquifers.


Historical Perceptions
Several historical misconceptions about ground-
water  microbiology hindered  the use of bio-
reclamation and probably contributed to the abuse of
aquifers. First, ground water was perceived as a safe
water source, protected by a metabolically diverse
"living filter" of microorganisms in the soil root zone
which functioned  to convert organic pollutants to
innocuous endproducts. However, recent evidence
indicates that, in many instances, ground water may
be at least as contaminated  as surface waters
(Council on Environmental Quality, 1981; Page,
1981).  In retrospect, the  historical perspective is
somewhat understandable since ground-water flow
is generally slow (10 to 100 m per year) and transport
processes are  complex (McKay,  et al, 1985). A
significant time lag for the movement of chemicals
from their subsurface source to even nearby wells is
common, and many years can pass before environ-
mental or health  impacts of ground-water contam-
ination become evident.

Secondly, aquifers were considered abiotic environ-
ments  since early studies indicated that microbial
numbers decreased  with soil depth (Waksman,
1916), and  later studies showed that the vast
majority of cells were attached to soil particulates
(Balkwill, et al., 1977). In addition, by estimating the
time required for  surface water  to  vertically
penetrate subsurface formations,  investigators
reasoned that  microbes traveling with the water
                                              67

-------
would quickly exhaust  available  nutrients and
rapidly die off. These beliefs led to the perception
that aquifers were sterile  and essentially irre-
trievably lost when contaminated by polluting chem-
icals. Now, however, microscopic, cultivation, meta-
bolic, and biochemical investigations,  often using
aseptically obtained aquifer material, reveal that the
terrestrial subsurface harbors a surprisingly rich
assemblage of procaryotic and some eucaryotic life
forms. The evidence supporting this perception  is
described in the following sections.
Ground-Water Contamination
The sources of ground-water contamination have
important implications for the design of biorestora-
tion techniques. Contaminants can originate from
non-point sources, such as  agricultural chemicals
and road salts, and from point sources including
residential septic systems, leaking underground
storage tanks, surface impoundments, landfills, and
transportation losses. Comprehensive reviews of the
sources and types of ground-water pollution are
presented by Keswick (1984); Craun (1984); and
Zoetman (1985).

It is important to point out that ground-water
pollution is not just a water pollution problem. As
shown in Figure 69, a spill of gasoline hydrocarbons
tends to exist in multiple phases. The gasoline free
product represents the most severe form of contam-
ination, but also is the most limited in terms of area
affected. As gasoline moves from a spill site,  it
contaminates  soil in  the vadose zone. Since its
components are largely water-insoluble and less
dense than water, hydrocarbon free product tends to
reside and spread along the water table boundary.
Hydrocarbon free product can easily pollute wells
within the zone of contamination, and also can sorb
to and contaminate those soil areas influenced by
water table fluctuations. This sorbed material tends
to be another  more subtle  source of secondary
contamination.  In addition, free product can contam-
inate surface waters, and gasoline vapors can collect
in basements of buildings and create  inhalation or
explosion risks.

As discussed in earlier chapters, the severity of soil
contamination tends to be a function of the proper-
ties  of the soil. Water, too, affects the amount of
contamination because it is an excellent solvent and
can even dissolve, to some extent, those substances
deemed insoluble. The concentrations of hydrocar-
bons in the water phase are generally low, but the
amount of area affected by such contamination can
be very high. Consequently, this form of contamina-
tion  has the highest potential for human exposure.
All remediation technologies  must consider where
the contaminant resides in the subsurface; treatment
of only  the water phase may provide  a temporary
solution to an  immediate problem, but also  may
effectively ignore the more subtle and, many times,
more important problem areas.
Evidence for Subsurface Microbiota
Microbiological  investigations of the terrestrial
subsurface have revealed that all aquifers examined
thus far possess a sometimes surprisingly rich micro-
flora.  Such studies include direct microscopic, culti-
vation, metabolic, and biochemical evidence for
microorganisms in aseptically obtained aquifer
material. Table  4  is only a partial  listing of the
typical numbers of bacteria detected in  various
geological settings. This table illustrates that rela-
tively high numbers of microorganisms can be
detected in both contaminated and pristine aquifers
of varying depth and geological composition. There is
                                           Contaminated
                                           Water
                                           •Supply
I           Fume/ Explosion
           Hazard
                                                                                site
       Surface
       Contamination
                           Contaminated
                            Soil
                                  Water
                                                                                     Table
                                                                  Free Product
                                              Contaminated Water
Figure 69.   Multiple problems associated with hydrocarbon release to the terrestrial subsurface.
                                                68

-------
Table 4. Microbial Numbers in Various Geological Settings and Depths
Aquifer Site
Sand/Gravel
Lula, OK
Sand
Pickett, OK
Loamy Clay
Fort Polk, LA
Sand/Clay
Conroe, TX ,
Burn Cavern
Tenn. Colony, TX
Carrizo Aquifer
TX
Mogothy Aquifer
Long Island, NY
Sand
Ontario, Canada
Gravel
Dayton, OH
Glacial Till
St Louis Pk., MN
Marmot Basin
Alberta, Canada
Bacatunna, Clay
Pensacola, FL
Sand/Gravel
Gape Cod, MA
Sand
Norman, OK
Sample
Type3
S
S
S
S
S
w
w
w
w
w
w
w
S
S
S
Depth
(m)
5
5.5
5
5.5
7.5
91
76
146
10
10-12
25-35
1.5
410
12-32
1.8
Aquifer
Contaminant
None
None
None
None
Creosote
Phenolics
None
Treated sewage
Septage
None
Creosote
None
Acid waste
Sewage
Landfill leachate
Total Countb
x 106
Cells/gdw
or Cells/ml
3.8-9.3
5.2
9.8
1-1 0<*
5-49
—
™
-
0.14
0.036-0.06
0.07-10
0.05-2.5
10<*
11-34
11-17
Viable Count<=
x 104
CFU/gdw
or Cells
20-110
250
-
—
4.2
0.04
0.01-5
0.031
0.052-0.06
0.09-4006
*"
-
~
~
References
Wilson, etal., 1983;
Balkwill and Ghiorse, 1985
Balkwill and Ghiorse,
1 985; Ghiorse and
Balkwill, 1985
Ghiorse and Balkwill,
1983; White, etal., 1985
Webster, et al., 1 985
Humenick, etal., 1982
Humenick, et al., 1 982
Godsey and Ehrlich, 1978
Ventullo and Larson, 1 985
Ventullo and Larson, 1 985
Ehrlich, etal., 1983
Ladd, etal., 1982
White, etal., 1983
Harvey, et al., 1 984
Beeman and Suflita, 1 987
aAquifer Solids
"Ground Water
*>Acridine Orange Direct Count (Unless Otherwise Noted)
cPlate Count Assay (Unless Otherwise Noted)
dSignature Lipid Analysis
eMost Probable Number Assay
good evidence that even deep geological formations
can be suitable habitats for bacteria (Kuznetsov, et
al., 1963; Updegraff, 1982).
The  picture  that emerges  from  microbiological
studies is that subsurface microorganisms tend to be
small, capable of response to the influx of nutrients,
and primarily attached to solid surfaces. Although
detectable, eucaryotic  microorganisms are few in
number relative to surface soil  microorganisms and
presumably of minor importance in the terrestrial
subsurface. Microeucaryotes  (such as fungi and
protozoa)  probably exist  in most aquifers as
metabolically inert resting structures. These data
reinforce information  published in a recent book
(Bitton and Gerba, 1984) which reiterates that there
is little to preclude microbial growth in the terres-
trial subsurface (McNabb and Dunlap, 1975). This
conclusion is important because it shows that those
aquifers that are most susceptible to contamination
by chemicals (those less than 100 meters deep), do
possess diverse microbial communities.
Sample Procurement
The development of suitable sampling technology
has helped to show the existence of a sizable  sub-
surface microbiota. Any microbiological probe of the
terrestrial subsurface is critically dependent on the
quality,  integrity, and representativeness of the
samples  obtained. To make valid interpretations of
                                                 69

-------
 the data resulting from subsurface investigations,
 the samples received must not be contaminated with
 nonindigenous microorganisms. Potential sources of
 contamination include organisms present on the
 drilling machinery, in surface soil layers, in drilling
 muds, and in water used to  make up the drilling
 muds. Since most subsurface biology is associated
 with the solid matrix, the discussion of sampling will
 be limited to aquifer sediments. However,  McNabb
 and Mallard (1984) have  published an excellent
 review of all the requirements for microbiological
 sampling of aquifers.

 No sampling effort can be  performed without some
 microbiological contamination, regardless  of
 whether drilling tools are employed to reach deeper
 regions of the subsurface or if excavation pits are dug
 in relatively shallow areas. With  this realization,
 most current sampling efforts rely on the recovery
 and subsequent dissection of cores to  remove
 microbiologically compromised portions. This dissec-
 tion can be performed in the field or, if the sampling
 sites are nearby, when the cores are returned to the
 laboratory. In all cases, the outer few centimeters
 and the top and bottom portions of aquifer cores are
 removed because of possible contamination by nonin-
 digenous bacteria, and only the center portions of an
 aquifer core are used for subsequent analysis.

 Ideally, this dissection process occurs as  soon  as
 possible after the core is removed from the ground so
 that nonindigenous microorganisms do not have a
 chance to penetrate to the inner portions of the core.
 In the field, an alcohol-sterilized paring device is
 used in the dissection process (McNabb and Mallard,
 1984). The paring device has an inner diameter that
 is smaller than the diameter  of the core itself. The
 aquifer material is extruded  out of the core barrel
 used for sampling and over  the sterile paring device
 to strip away the potentially contaminated material.
 For anaerobic aquifers, this field paring procedure is
 performed inside plastic anaerobic glovebags while
 the latter is purged with nitrogen to minimize
 exposure of  the microflora  to oxygen (Beeman and
 Suflita, 1987). Samples received in this manner are
 termed "aseptically obtained" and  are suitable for
 microbiological analysis.
Metabolic Activity in Aquifers
Many of the studies cited in Table 4 employed
aseptically obtained aquifer material and, thus,
provide direct and conclusive  evidence for the
existence of subsurface bacteria. Several questions
immediately become apparent: (1) Are the  indig-
enous ground-water organisms metabolically active;
(2) How diverse is their metabolism; (3) What factors
serve to  limit and/or  stimulate  the growth  and
 metabolism of these organisms; and (4) Can one take
 advantage of the inherent metabolic versatility of
 aquifer communities  to remediate contaminated
 areas?

 The answers to the first two questions can be gleaned
 from a review of the scientific literature (Table 5).
 Even though many of  the studies cited in the table
 were conducted prior to the development of aseptic
 sampling procedures,  the list does  illustrate that
 subsurface microbial activity can be detected, and
 major and minor elements  may potentially be
 recycled in subsurface ecosystems.  The metabolic
 processes referred to in the table include a variety of
 aerobic and anaerobic carbon transformations, many
 of which are pertinent  to  the  biodegradation of
 aquifer contaminants. In this respect, the recent
 reviews by Ghiorse and Wilson (1988) and Lee, et al.
 (1988) are very illuminating. Other  metabolic
 processes noted are those required for the cycling of
 nitrogen, sulfur, iron, and manganese.

 The various metabolic processes listed in Table 5 are
 not mutually exclusive. As shown  schematically in
 Figure 70, as labile  organic  matter enters an
 oxygenated aquifer, microbial metabolism will likely
 degrade the contaminating substrate. That is, the
 indigenous microorganisms utilize the pollutant as
 an electron donor to support heterotrophic microbial
 respiration. The aquifer microbiota use oxygen as a
 co-substrate and as an electron acceptor to support
 their respiratory activities. This demand on oxygen
 often results in its depletion and the establishment of
 anaerobic conditions. When oxygen becomes  limit-
 ing, aerobic respiration slows, but other groups of
 microorganisms may then come  into play and
 continue  to degrade the contaminating  organic
 matter. Under conditions  of anoxia, anaerobic
 bacteria can  use organic chemicals or several  inor-
 ganic anions as alternate electron acceptors.

 Nitrate present in the ground water,  as indicated in
 Figure 70, generally is not  rapidly  depleted until
 oxygen is  utilized. Organic matter can  still be
 metabolized, but, instead of oxygen, nitrate serves as
 the terminal electron acceptor during the process of
 denitrification. Similarly,  sulfate can  serve  as a
 terminal electron acceptor when nitrate  is limiting.
 This information supports  microbial metabolism
 linked to sulfate-reduction. When this occurs, hydro-
 gen sulfide often can be detected in the ground water
 as a metabolic endproduct.  When very highly
 reducing conditions prevail in aquifers, carbon
dioxide also can serve  as an electron acceptor and
 metabolism is linked to methane formation. Some-
times a spatial separation of dominant metabolic
processes can occur in  aquifers, depending on the
availability of electron acceptors, the presence of
suitable microorganisms, and the energy benefit of
                                                70

-------
Table  5.     Selected Microbial Metabolic Processes and Oxygen Requirements

              Metabolic Process                       Oxygen Requirement
                        in Subsurface Environments

                                           Reference
   I.  Biodegradation of Organic Pollutants
      a.   Petroleum
          Hydrocarbons
      b.   Alkylpyridines

      c.   Creosote
          Chemicals
      d.   Coal Gasification
          Products

      e.   Sewage Effluent
      f.   Halogenated
          Organic Compounds
      g.   Nitrilotriacetate (NTA)


      h.   Pesticides




   II.   Nitrification



  III.   Denitrification



  IV.   Sulfur Oxidation

   V.  Sulfur Reduction
  VI.   Iron Oxidation


  VII.  Iron Reduction


  VIII. Manganese Oxidation

   IX. Methanogenesis
 Aerobic
 Aerobic/Anaerobic

 Aerobic/Anaerobic



 Aerobic


'Aerobic



 Aerobic/Anaerobic
 Aerobic/Anaerobic


 Aerobic/Anaerobic




 Aerobic



 Anaerobic



 Aerobic

 Anaerobic  ,
  Aerobic


  Anaerobic


  Aerobic

  Anaerobic
Jamison, etal., 1975
Lee and Ward, 1985
McCarty, etal., 1984
Raymond, et al., 1976
Roberts, etal., 1980
Wilson, et al., 1983
Wilson, et al., 1985

Rogers, et al., 1985

Ehrlich, etal., 1983
Smolensk! and Suflita, 1987
Wilson, etal., 1985

Humenick, etal., 1982
Aulenbach, etal., 1975
Godsey and Ehrlich, 1978
Harvey, etal., 1984'

Gibson and Suflita, 1986
McCarty, etal., 1984
Suflita and Gibson, 1985
Suflita and Miller, 1985
Ward, 1985
Wilson,  etal., 1983
Wood, etal., 1985

Ventullo and Larson, 1985
Ward, 1985

Gibson  and Suflita, 1986
Suflita and Gibson, 1985
Ventullo and Larson, 1985
Ward, 1985

Barcelona and Naymic, 1984
Idelovitch and Michail, 1980
Preul, 1966

Ehrlich, etal., 1983
Lind, 1975
Ward, 1985

Olson, etal., 1981

Beeman and Suflita, 1987
Bastin,  1926
Hvid-Hansen, 1951
Jacks, 1977
Olson, etal., 1981
van Beek, etal., 1962

Olson, etal., 1981
Hallburg and Martinell, 1976

 Godsey and Ehrlich, 1978
 Ehrlich, et al., 1983

 Hallburg and Martinell, 1976

 Beeman and Suflita, 1987
 Belyaev and Ivanov, 1983
 Davis, 1967
 Gibson and Suflita, 1986 ,,
 Godsey and Ehrlich, 1978
 Suflita and Miller, 1985
 van Beek, et al., 1962
                                                               71

-------
                            GROUNDWATER FLOW
                CO
                »-
                2
                LU
               ID
               o
               LU
             + 10
             •  0
             UJ
              a
             -10
                                                   CHEMICAL SPECIES
                                                 ELECTRON ACCEPTORS
           ACETATE	-CO,
                                             SO'
                     CO,
                                  BIOLOGICAL CONDITIONS
                  AEROBIC
                  HETEROTROPHIC
                  RESPIRATION
                        DENTTRFICATICN
SULFATE
RESPIRATION
                                                      METHANOGEhCSIS
Figure 70.
Mfcroblally mediated changes in chemical species, redox conditions, and spatial regions favoring different types
of metabolic processes along the flow path of a contaminant plume (adapted from Bouwer and McCarty, 1984):
the metabolic process to the catalyzing microbial
communities. As organic matter is transported in a
plume, a series of redox zones can be established
which range from highly oxidized to highly reduced
conditions. The biodegradation potential available
under these various conditions and the expected
                                       rates of metabolism will be very different in each
                                       instance.

                                       The biodegradation  of a known ground-water
                                       contaminant, p-cresol, can be used as an example
                                       (Figure 71). This compound is reported  to be
                                               72

-------
                             OH
                                                         COO"
   Redox  Conditions      Biodegradability   Lag Time      Relative  Rate     Ref.
   Aerobic

   Denitrifying

   Sulfate  Reducing

  (.Methanogenic
                                     1

                                     2

                                    3,4

                                  4,5,6
    (1) Hopper,  1976,  1978^;  (2) Bossert.& Young,  1986;  (3)  Bak  &'Widdel, 1986;
    (4) Smolenski & Suflita,  1-987;  (5)  Godsy  et al.,  1983;  (6)  Senior  & Balba,
    1984.          :.;,,•
Figure 71.   Proposed degradation pathway for p-cresol under various redox conditions.
degradable under aerobic, dinitrifying, sulfate-
reducing, and methanogenic conditions. Moreover,
the initial stages of the metabolism of p-cresol are
similar under the various redox conditions.  Under
aerobic conditions, the initial conversion  of the
substrate to aromatic alcohol requires molecular
oxygen. However,  oxygen inhibits the anaerobic
metabolism of p-cresol, so, presumably, a water
molecule  is employed to effect this transformation
during the catalysis of this substrate under the other
conditions. Comparing the length of the lag  time
before the onset of biodegradation and the rate of
metabolism once it begins, aerobic decomposition is
relatively fast, especially compared to methanogenic
incubation conditions. Thus, if rapid  biotrans-
formation  is the goal,  then aerobic incubation
conditions should be maintained in an aquifer in any
sort of a bioremediation strategy. However, consider-
ing the often slow rates of ground-water movement,
even anaerobic removal  may effectively compete
with the faster aerobic metabolism of this substrate.

This does not imply that aerobic transformation
mechanisms are always faster'than the  anaerobic
counterparts. Some compounds may biodegrade at
faster rates when anaerobic conditions prevail. For
example, several halogenated aliphatic substrates
prove amenable to microbial metabolism under
reducing  conditions, but persist in comparable
aerobic incubations (Bouwer  and McCarty,  1984;
Bouwer, et al., 1981). Similarly,  the haloaromatic
pesticide 2,4,5-T  often is considered recalcitrant
under aerobic conditions. Although specialized,
genetically manipulated microorganisms now exist
that aerobically degrade this substance (Kilbane, et
al., 1982), 2,4,5-T can be completely mineralized
without genetic  manipulations when incubated
under methanogenic conditions (Gibson and Suflita,
1986).

When assessing microbial metabolism in aquifers,
the existing scientific literature must be used as a
guidepost for determining the types of biotrans-
formations to expect. For some substances, existing
information will be available  and instructive; how-
ever,  for  more exotic substances,  the available
literature will prove disappointingly small. In either
case, laboratory experimentation is required to eval-
uate a pollutant's susceptibility to biotransforma-
tion.
Microcosms
Even when the bulk of scientific literature indicates
that a pollutant chemical is likely to biodegrade
easily in most environments, an assay should be used
to confirm that contention.  This is necessary
because, while a specific ground-water contaminant
                                                73

-------
 may be the same as that encountered in other
 environments, the particular physical, chemical, and
 hydrological  site characteristics can be quite
 different and can influence biodegradation processes.
 A reliable tool or method is required for evaluating
 the susceptibility of a ground-water contaminant to
 biodegradation. In addition, the experimental  sys-
 tem should be scientifically rigorous, yet flexible
 enough to facilitate a critical evaluation of various
 bioremediation strategies. In this respect, the use of
 microcosms seems  to hold great  promise. Some
 examples of microcosm technology are discussed by
 Pritchard (1981) and Pritchard and Bourquin (1984).
 Also, Wilson and Noonan (1984) have analyzed the
 specific application of microcosms to aquifer systems.

 The term microcosm means different things to
 different  individuals. In the strictest sense, a
 microcosm is:

   ...a calibrated laboratory simulation of a portion
   of a natural environment in which  environ-
   mental components, in as undisturbed a condi-
   tion as possible, are enclosed within definable
   physical and chemical boundaries and studied
   under a standard set of laboratory conditions
   (Pritchard, 1981).

 Implicit in this definition is that microcosms may be
 used as surrogates for ground-water field studies
 without  the associated  logistical, financial,
 administrative, and regulatory problems. In addi-
 tion, when properly employed, microcosms incor-
 porate the site-specific variables of the area under
 investigation and allow for the assessment of risk
 and prediction of the transport  and fate charac-
 teristics of a contaminant. Further, such systems
 help to evaluate the waste assimilative capacity of
 an environment (Pritchard and Bourquin, 1984).
 This evaluation can be achieved without a detailed
 knowledge of  those endogenous environmental
 factors that may  stimulate or retard microbial
 metabolism.

 The design of microcosms depends a great deal on the
 nature of the question being posed (Wilson and
 Noonan, 1984;  and  Pritchard and Bourquin, 1984).
 Microcosms range  from simple batch incubation
 systems to large and complex flow-through devices,
 and all can be  used in a variety of studies on  the
 transport and fate of subterranean contaminants
 influenced by microorganisms. Microcosms are used
 to help identify biodegradable pollutants  and those
 pollutants that  tend to persist. Often, by following
 the fate of a pollutant in a microcosm, the predom-
 inant pathways of biotic or abiotic transformation
can be described. Lastly, the decay  of a particular
contaminant in a microcosm may relate to the rate of
biotransformation in situ without relying on indirect
 measures of microbial biodegradation for rate predic-
 tions such as the enumeration of catalytic micro-
 organisms.

 This last point can be illustrated with an example.
 An assay for the anaerobic biodegradation of the
 pesticide 2,4,5-T was conducted with microcosms
 made of sediments and ground water sampled from
 two distinct sites within the same aquifer (Gibson
 and Suflita, 1986). Microorganisms from one site
 were able to completely mineralize this substrate
 while parallel experiments  revealed that the
 pesticide persisted in microcosms made from the
 other site. Further experimentation showed that the
 requisite microorganisms were present at both  sites,
 but their activity at the second site was at  least
 partially inhibited by locally high levels of sulfate in
 the ground water. Predictions of the rate of substrate
 utilization in the sulfate-rich ground water would
 have been overly optimistic if based primarily on
 assays for the number of degrading microorganisms.
 Conversely, predictions of transport and  fate
 behavior based on microcosm experimentation would
 likely prove more realistic,  even without the
 appreciation for the role of sulfate in a portion of the
 aquifer. This example also illustrates how  site-
 specific aquifer characteristics can drastically influ-
 ence the evaluation of biodegradation results.

 Microcosms possess several experimental  advan-
 tages. For example, microcosms are replicable and
 allow appropriate controls to  be employed.  This
 advantage cannot be  overstated since it allows the
 resulting data to be statistically evaluated. Further,
 such experiments can meet the usual requirements
 for scientific rigor. While environmental simulation
 is a goal of microcosms, such systems may be subject
 to controlled perturbations  of their chemical or
 physical parameters. In that manner, the influence
 of such perturbations can be quantitatively  evalu-
 ated for effects on biotransformation  processes.
 Similarly, the trophic structure of microcosms can be
 varied in order to study complex biological inter-
 actions. However, this advantage may not exist since
 microeucaryotes tend to be of minor importance in
 most aquifers examined thus far.  Microcosms are
 accessible and  containable laboratory  tools with
 which  the experimenter controls inputs (water,
 substrate, nutrients, etc.) as well as exports (cells,
 metabolites, etc.). When properly employed,  micro-
 cosms avoid field pollution, yet provide a time-
 efficient way of evaluating the likely fate of environ-
 mental contaminants.

 Conversely, microcosms also suffer from several
 limitations that should not be ignored at the onset of
experimentation and during an evaluation  of the
resulting data. One limitation is that the start-up
and operating costs for complex microcosms can be
                                                74

-------
high. Also, microcosms can disturb the normal struc-
tural and functional features of an ecosystem and
generally possess abnormally high surface to volume
ratios. The simple sampling and containerization of
microbial  communities within altered physical
limits causes the initiation of rapid population shifts
that may influence biodegradation estimates. How-
ever, the advantages of microcosm  techniques out-
weigh the inherent limitations  (Pritchard and
Bourquin,  1984), and such studies can provide useful
and environmentally reliable information.
Extrapolation of Metabolic Information
Although microcosms provide a tool that can be used
to address some of the questions regarding metabolic
proficiency of ground-water microorganisms, other
questions remain. For example,  is the information
collected from microcosm studies reliable, and can
the metabolic information obtained from microcosm
studies be reliably extrapolated to terrestrial
subsurface environments?  These questions stem
from a consideration of the basic principles that
generally govern biodegradation. Although subsur-
face microbes may or may not prove to be unusual,
metabolism tends to be a unifying feature of life. As
Dagley (1984) has pointed out,  diverse life forms
often exhibit similar  metabolic pathways, and
information pertinent to even human physiology
often can be obtained from  microbiological studies
(Dagley, 1984). Therefore,  metabolic principles
gleaned from the study  of xenobiotic compound
metabolism in surface  habitats are useful  for
subsurface site evaluations and the design of aquifer
remediation strategies.  The existing literature
provides an excellent scientific foundation that can
be used to make cautious extrapolations of metabolic
principles observed in surface environments to those
that regulate microbial activities in the subsurface.

Microorganisms play prominent  roles in the
transformation and degradation of contaminating
organic chemicals in virtually every major habitat
except the atmosphere. Microbial communities in
nature are able to metabolize many types and
quantities of synthetic organic  compounds (Alex-
ander, 1981; Kobayashi and Rittman, 1982). At the
extremes, some xenobiotic compounds can supply the
carbon and energy requirements associated with
microbial growth, while  other substrates tend to
resist significant biotransformation. Still other
chemicals are "cometabolized" to form degradation
products that are usually structurally similar to the
parent molecule (Alexander, 1981). By definition,
the latter process does  not result in growth of the
catalyzing microorganism,  but  does  result in the
transformation of a substrate such that the molecule
can be available for subsequent utilization by other
microorganisms. In an ecological sense, it is common
that several microbes act in a combination to effect
some overall degradation of a substrate (Slater and
Lovatt, 1984). In  anaerobic  ecosystems, this con-
certed metabolic activity is essential for the normal
cycling of carbon  and energy on earth (Mclnerney
and Bryant, 1981).

Knowledge of the mechanisms of  energy  and
materials cycling in the terrestrial subsurface,
however, is limited and reflects  the difficulties in
sampling and studying a complex and unfamiliar
environment  (McNabb  and Mallard,  1984).  As
pointed out earlier, the likely metabolic fate of
contaminating chemicals sometimes can be predicted
if the governing microbial ecology and the limits of
extrapolation are understood.
Recalcitrance
Researchers have learned that there are limits to the
metabolic versatility of microorganisms. Many xeno-
biotic substrates are transformed so slowly under
most conditions that some degradation of environ-
mental quality occurs. These chemicals are referred
to as recalcitrant. However, recalcitrance or resis-
tance to biodegradation is  not a feature strictly
associated  with anthropogenic chemicals'. Many
naturally occurring materials also can persist for
long periods of time (Alexander, 1965 and 1973). For
example, entire portions of human corpses  (i.e., hair)
were recovered from various environments (peat
bogs, desert cemeteries, etc.) and estimated to be
hundreds to thousands of years old. Similar ages
have been ascribed to proteolytic enzymes  recovered
from permafrost soils. Microbial spores can persist
for tens  of thousands of years. These spores are
useful to help "date"  geological deposits. Finally,
most of the world economies are based on petroleum
deposits that have  managed to persist  for many
millennia within subterranean environments.

In aquifers, as in other environments, organic matter
decomposition should be considered in the context of:
(1) existing environmental conditions; (2) physiology
of the requisite microorganisms; and (3)  chemical
structure of the particular substance(s) under consid-
eration. The general principles of these interrelated
topics are considered below; detailed reviews can be
found in Alexander's publications (1965 and 1973).
Environmental Barriers to
Biodegradation
In order to grow, microorganisms need a suitable
physical and chemical environment.  Extremes  of
temperature, pH, salinity, osmotic  or hydrostatic
pressures, radiation, free water limitations, contami-
                                               75

-------
nant concentration, and/or the presence of a heavy
metal  or other toxicant materials can adversely
influence and even limit the rate of microbial growth
and/or substrate utilization. Often, two or more
environmental factors interact to  limit microbial
decomposition processes, and, in fact, environmental
barriers can act to render normally labile substances
persistent. This explanation was used by Woods Hole
researchers (Jannasch, et al., 1971) who found food
substances from  an  accidentally sunken and
subsequently recovered submarine to  be almost
preserved after 10 months  exposure to deep-sea
conditions. However, when the food was incubated in
sterile seawater at in situ temperatures (3°C), the
materials putrefied after only a few weeks. The
authors suggested that the hydrostatic pressures of
the sea (150 ATM) effectively raised the minimal
temperature necessary for microbial proliferation.
Once  this  increase  exceeded  incubation
temperatures, microbial activity slowed 10 to  100
fold.

Recognition of the  nature of the limiting environ-
mental factor(s) and a consideration of its practical
application in subsurface environments will help
dictate which type of bioreclamation procedures to
use. For example, the temperature of aquifers prob-
ably could not be significantly altered to stimulate in
situ microbial growth and metabolism. However, the
same is not true for a surface biological treatment
process.
Physiological Barriers to Biodegradation
In addition to the immediate environment, various
microbial physiological factors can influence the
biotransformation of pollutant chemicals. Like all
forms of life, the requisite microorganisms present in
an aquifer are primarily composed of carbon,
hydrogen, oxygen, nitrogen, phosphorus, sulfur, and
a variety of trace elements. These  substances are
required  in varying degrees  for aquifer micro-
organisms to proliferate. Aquifer microorganisms
can utilize such substances to the point where one or
more of  the  requirements are exhausted and
effectively limit further microbial growth  or
metabolic activity. For aquifer remediation efforts to
succeed, these materials must already be present in
the aquifer or be supplied in the proper form.

Ideally, the organic pollutants in the aquifer repre-
sent  an appropriate supply of carbon  and energy
necessary for heterotrophic  microbial growth.
However, that supply can be too high or too low. Too
high a substrate concentration can  limit microbial
metabolism due to the toxicity of the substrate to the
requisite  microflora. In  contrast, ground-water con-
taminants also can be present at concentrations that
are too low to allow microbial  response and/or may
 not be suitable  growth substrates. Growth  and
 energy sources need not come from the same carbon
 substrate. The growth and metabolism of degrading
 microbes sometimes can be stimulated by providing
 them with a non-harmful primary carbon substrate
 so that the rate  and extent of pollutant decompo-
 sition can be proportionally increased (McCarty,
 1985; McCarty, et al., 1981; McCarty, et al., 1984).

 A chemical also  will be metabolized poorly if it is
 unable to enter microbial cells. This may occur with
 either natural or  anthropogenic polymers. While the
 monomeric units may be  inherently amenable to
 microbial destruction, the  larger molecular weight
 polymers persist because they often fail  to  gain
 access to intracellular catabolic enzymes. A sub-
 strate also  will persist if it fails to derepress the
 enzymes necessary for its destruction. It  may be
 possible to induce the appropriate enzymes with an
 alternate chemical  compound. Occasionally, initial
 biochemical reactions result in metabolites that tend
 to  inhibit  the degradation of even the  parent
 molecule and can  adversely affect the biodegradation
 of some pollutants.

 Lastly, the absence  of other necessary micro-
 organisms can limit the destruction of a contami-
 nant. Often, several microbial groups are needed for
 the destruction of a pollutant. In anaerobic environ-
 ments, this type of relationship  is a prerequisite for
 completion of the carbon cycle  (Mclnerney and
 Bryant,  1981). The anaerobic mineralization of
 organic matter is critically dependent on obligate
 microbial consortia; if any of the individual members
 of a consortium are absent, the biodegradation of the
 parent material effectively ceases.
Chemical Barriers to Biodegradation
One of the most important factors that influence the
degradation of a contaminant in aquifers  is the
contaminant's structure, which can dictate the
pollutant's  physical state (i.e., soluble, adsorbed,
conjugated, etc.) and thereby alter its tendency  to
biodegrade. However, it is important to be specific
when referring to biodegradation. When a compound
undergoes primary attack, initial metabolic  events
often result in reaction products with their  own
environmental impact and persistence  character-
istics. An example of this is the fate of the pesticide
carbaryl (Sevin). Carbaryl is widely  used as a
substitute for DDT and is touted as being readily
degradable. However, this pesticide is known to have
a myriad of environmental  fates (Figure 72). An
examination of Figure  72 reveals  that  somte
metabolic routes do lead to the ultimate conversion
of carbaryl  to carbon dioxide, while others result in
the production of complex aromatic metabolites, the
fate of which is still unknown. Therefore,  when
                                               76

-------
                                 OCONHCH3
                                                      OH
Figure 72.   Proposed pathway of carbaryl degradation in soils and microbial cultures (adapted from Rajagopal, et al., 1984).
evaluating the scientific literature, the distinction
must be  made  between "biodegradation" and
"mineralization." The mere loss of a chemical from
an environment may or may  not  be a desirable
consequence of biotransformation processes. If bio-
degradation results in the production of undesirable
metabolites, it may be best to choose a non-biological
strategy for aquifer cleanup. The  example  above
illustrates the need to understand the overall fate of
a contaminating chemical in an aquifer and how
informed decision-making must be based on solid
metabolic information.

Aquifer pollutants  may contain various chemical
linkages that tend to  favor or hinder microbial
attack. However, broad generalizations on the biode-
gradability of various linkages tend to be of limited
value since  substitute effects drastically alter the
                                                 77

-------
    the molecule proved available to only a few organ-
    isms  and permitted minimal  growth at best. Very
    often, when a quaternary carbon atom occurs at the
    terminal end of an alkane chain, the resulting
    molecule is quite  resistant to aerobic microbial
    attack.

    Structural effects also can be  observed with
    anaerobic microorganisms. A comparison of the rate
    of aryl-reductive dehalogenation  by an anaerobic
    consortium of bacteria revealed that the type and
    position of the halogen substituent influenced  the
    degradation of the resulting molecules.  Table 7
    shows that all  brominated and iodinated benzoates
    were  transformed by these  bacteria, but no fluori-
    nated benzoates and only one chlorinated  benzoate
susceptibility of even simple organic molecules to
biotransformation. The number, type, and position of
substituents must be considered  when  evaluating
the metabolic fate of particular contaminants in
aquifers.

The effect of branching is illustrated  in Table 6
which is adapted from the work  of McKenna and
Kallio (1964).  Approximately 15  bacterial strains
from several different genera were  evaluated for
their ability to use a series of structurally related
substrates. Cultures exhibiting  profuse growth,
slight growth, or no growth on a particular substrate
were assigned a qualitative index of 2, 1, or 0,
respectively. All cultures grew well with 1-phenyl-
decane. The organisms most likely initiated their
attack on this molecule through the oxidation of the
terminal methyl group  in order to convert  the
molecule to a fatty acid derivative. Continued metab-
olism of the molecule would likely be by 0-oxidation
of the side chain. However, a single methyl substitu-
ent on  the side chain drastically influenced  the
susceptibility of  the resulting structure  to
metabolism by the test organisms. All the myco-
bacteria and nocardia were able to grow on 1-phenyl-
4-methyldecane. However,  the pseudomonads and
micrococci were no longer able to do so. When  the
substrate contained an "internal" quaternary carbon
atom, as is the case for l-phenyl-4,4-dimethyldecane,


Table 6.    Microbial Growth Response to Phenyldecanes: 0 = No Growth; 1 = Questionable or Slight Growth; 2 = Moderate
           or Abundant Growth (Adapted from McKenna and Kallio, 1964)
   Table 7.
The  Rate  of  Anerobic  Monohalobenzoate
Metabolism  Exhibited  by  an  Enrichment of
Dehalogenating Bacteria (DeWeerd, et al., 1986)

         Dehalogenation Rate (nmoles/L/hr)
        Position
                Cl
                                      Bf
Ortho
Meta
Para
n.d.
0
0
0
4.63
0
1.20
3.70
0.05
0.50
0.89
0.66
Organism
M/crococcus cerificans




Pseudomonas aeruginosa


Mycobactenum phlei
M, fortuitum
M, rhodochrous
M. smegmatis
Nocardia opaca
N, rubra
N, erythropolis
N, polychromogenes
N, corallma
Strain
H.0.1-N
H.0.3
H.0.4
S-18.2
S-14.1
119JWF
191 JWF
Sol 20 JS
No. 451
No. 389
No. 382
No. 422





1 -Phenyldecane
2
2
2
2
2
2
2
2
2
?.
2
2
2
2
2
2
2
1 -Phenyl-4-methyldecane
' 0
0
• • o
,0
" °
0
0
0
2
2
2
: 2
2
2
2
2
2
i -rnenyi-q-,4-
dimethyldecane
0
0
0
0
0
0
0
0
0
0
0
1-
1
1 .
1
1
1
78

-------
were transformed.  For transformed compounds,
slower rates of dehalogenation were correlated with
increasing halogen size. However, the rate of tirtKo,
meta, and para dehalogenation differed less with
increasing halogen size. Most importantly, the meta-
substituted halobenzoates always exhibited the
fastest rates of biodegradation when compared  to
their isomeric counterparts. Thus, when halobenzo-
ates are considered, the meto-substituted molecules
are the most susceptible to anaerobic biotf ansforma-
tion.    —                                   :J%
These few  examples illustrate that the chemical
structure of a contaminant also influences its sus-
ceptibility to biodegradation. Despite the difficulty
in generalization, the scientific literature shows that
the closer a contaminant  structurally resembles a
naturally  occurring compound,  the better the
possibility  that the former will be able to enter a
microbial cell, derepress the synthesis of catabolic
enzymes, and be converted by those  enzymes  to
central metabolic intermediates. In addition, biode-
gradation is less  likely, but not entirely precluded,
for those molecules having unusual structural
features  only infrequently encountered  in nature.
Ease of biodegradation can be viewed as a continuum
ranging from very labile compounds to those that are
recalcitrant (Figure 73).
  Labile
                              Recalcitrant
  Structural
  Analogs of
  Natural Materials
Chemicals with
No Natural
Counterpart
 Figure 73.  The continuum of biodegradation ease.

 When viewed in this context,  it should not be
 surprising that xenobiotic compounds tend to persist
 in nature because microorganisms have not evolved
 the necessary  metabolic machinery to attack those
 compounds. However, as a  group, microorganisms
 are nutritionally versatile, have the potential  to
 grow rapidly, and possess only a single copy of DNA.
 Consequently, any genetic mutation or recombina-
 tion is immediately expressed. If the alteration is of
 adaptive significance, new species of microorganisms
 can arise and proliferate. The polluted environment
 supplies selection pressure for the  evolution'of
 organisms with novel metabolic potential. Ulti-
 mately, the organisms may not only survive in the
 polluted environment, but also may be capable of
 growing at the  expense  of the contaminating
 substance.                                 .......
Bioremediation of Aquifers
Once an aquifer contaminant is recognized as being
susceptible to biodegradation, the goal of bioremedi-
ation/bioreclamation efforts  is  to utilize the meta-
bolic capabilities of the indigenous microflora to
destroy and eliminate that contaminant. Currently,
enhanced bioreclamation is the use of common
aquifer bacteria to degrade  organic contaminants.
This practice generally does  not require the innocu-
latidn of the terrestrial subsurface  with foreign
bacteria.

Bioremediation  attempts  to impose particular
conditions in an aquifer that encourage microbial
proliferation and the development of desirable micro-
organisms. With knowledge of the chemical and
physical needs of microorganisms and the  pre-
dominant metabolic pathways,  a strategy can be
developed to stimulate biodegradation. Most often,
microbial activity is stimulated by supplying the
nutrients necessary for microbial growth. These
efforts can take place either above ground or in situ.

An example of above-ground  bioreclamation is
schematically  shown in Figure 74.  In  the illus-
tration, ground water and the associated contam-
inant are pumped to the surface through a series of
recovery wells. On the surface, the water is treated
in a series of  steps, which include the supply of
appropriate nutrients.  The  treated (i.e., decontam-
inated) water is  then reinjected into the  aquifer.
Injection and recovery wells are positioned so that
they  intersect  a zone of contamination within the
aquifer along natural ground-water flow paths. If
excess nutrients are in the treated water,  it is
conceivable that this system also will stimulate the
in situ biotransformation of the contaminants.

Moving the contaminant to the surface for treatment
is often impossible or impractical since even mildly
adsorbing chemical species may require many
decades  of pumping before being reduced to suf-
ficiently low levels. In this case, an in situ approach
 may be more feasible.  Figure 75 shows two  such
systems.
 Figure 75A illustrates a system in which the
 microbial  nutrients are mixed  with  ground water
 and circulated through the  contaminated portion of
 the aquifer via a series of injection and recovery
 wells. In  this case, air compressors are used to
 impose aerobic conditions on the indigenous micro-
 flora. Nutrients also can be circulated via an infiltra-
 tion gallery as depicted in Figure 75B. This method
 provides a potential mechanism for the microorgan-
 isms to  attack contaminants trapped in the pore
 spaces of the unsaturated zone. Each of the examples
                                                 79

-------
                                         DIRECTION OF GROUND WATER FLOW
               outside water source
                           A
                          INJECTION  SYSTEM
•* —
t
' treated water / /
A"\ y ^
r clean \ 	 -, ^ ^
. pool ) S^\
/ t
M a »
^
• M M • a
V
                                              ZONE  OF CONTAMINATION
          V      I
                  ^
    contaminated
      water    T    T
                                                           I       I        J      1
                                                           f       w        w      ir
                                                T    T    T    T   T    T   T    t
r —
c
pump
1
1
1
1
1
1
^ RECOVERY
/


A


1
SYSTEM
4
1


1


J


J
i
1
 Figure 74.
                          •

treatment system design for the biorestoration of a contaminated aquifer (adapted from Lee, et
 discussed above presume an adequate understanding
 and control of the hydrogeology of the area under
 treatment. This topic  will be  dealt with more
 specifically in Chapter 8.

 Biorestoration approaches have several advantages
 over other potential cleanup strategies. First, these
 approaches can be  used to treat some common
 aquifer pollutants, as described in Chapter  8. Bio-
 remediation is environmentally sound in  that it
 results in the complete destruction of a contaminant.
 This is not to imply that the technology can remove
 99.999 percent of the material, but rather, it results
 m the complete change in the chemical properties of
 the  parent contaminant. Aquifer  biorestoration
 approaches utilize the indigenous microflora  so that
 mass inoculation efforts are often unnecessary.
 There is insufficient scientific evidence to suggest
 that the intentional addition of desirable  micro-
 organisms actually aided an aquifer biorestoration
 effort (Lee, et al.,  1988; Chapter 8).  Finally,
 compared to other technologies, biorestoration also is
 economical.

 However, biorestoration also suffers from several
drawbacks (Lee, et al., 1988). At some sites, many
types of wastes (acids,  bases, and organic  and
inorganic materials) are co-disposed and any of these
                            wastes alone or in combination  could inhibit
                            microbial  growth and metabolism. Incomplete
                            degradation of some  substances may lead to taste
                            and odor problems in the ground  water. Also, the
                            operation  of biorestoration programs can  be
                            maintenance- and analytically intensive. Further, to
                            date, the technology  has generally been limited' to
                            aquifers possessing  high permeabilities.  Despite
                            these limitations, the cost of competing technologies
                            will likely insure that bioreclamation efforts will
                            continue and assume an increasingly larger role as a
                            viable aquifer cleanup strategy.


                            References

                            Alexander, M., 1965.  "Biodegradation: Problems of
                             Molecular Recalcitrance and Microbial Fallibility."
                             Adv. Appl. Microbiol., Vol. 7, pp. 35-80.
                            Alexander, M., 1973. "Nonbiodegradable and Other
                             Recalcitrant Molecules." Biotech.  Bioeng Vol  15
                             pp. 611-647.                               '   '
                            Alexander, M., 1981. "Biodegradation of Chemicals
                             of Environmental Concern." Science, Vol 211 nn
                             132-138.                                 • ' PP'
                           Aulenbach, D. B., N. L. Clesceri, and T. J. Tofflemire,
                             1975.  "Water Renovation by  Discharge into Deep
                                                80

-------
               To Sewer or
               Recirculate
                               Air
                               Compressor
      Nutrient
      Addition
      Tank
                         Production
                         Well
                                    Coarse Sand
                                     Water Table-
                            ^_ —  Spilled Materials   __
                                                                         Water Supply
                       Injection
                       Well
                                                                   — Sparger
                                •. •  .  Clay  .'  • '.  •   '
                      Air Compressor or
                     Hydrogen  Peroxide Tank
Nutrient Addition
                                   -Infiltration Gallery

                                     ^Trapped Hydrocarbons
                       Recirculatec
                       Water  8 Nutrients/


                      Monitoring Well
              Water/"-.^
              Table

            Recovery Well
Figure 75.   Two designs for the in situ remediation of contaminated aquifers (adapted from Lee, et al., 1988).
  Natural Sand Filters." Proceedings of AIChE
  Conference, May 4-8,1975, Chicago, IL.

Bak, F. and F. Widdel,  1986. "Anaerobic Degrada-
  tion of Phenol and Phenol Derivatives by Desulfo-
  bacterium Phenolicum sp. nov." Arch. Microbiol.,
  Vol. 146, pp. 177-180.

Balkwill, D. L. and W. C. Ghiorse, 1985. "Character-
  ization of Subsurface Bacteria Associated with Two
  Shallow Aquifers in Oklahoma." Appl. Environ.
  Microbiol., Vol. 50, pp. 560-588.
      Balkwill, D. L., T. E. Ruzinski, and L. E. Casida,
       1977. "Release of Microorganisms from Soil with
       Respect  to Transmission  Electron Microscopy
       Viewing and Plate Counts." Antonie Van Leeu-
       wenhoekJ. Microbiol. Serol., Vol. 43, pp. 73-87.
      Barcelona, M. J. and T. G. Naymik, 1984. "Dynamics
       of a Fertilizer Plume in Ground Water." Environ.
       Sci. Technol., Vol. 18, pp. 257-261.
      Bastin, E. S., 1926. "The Problem of the Natural
       Reduction of Sulfates." Bull. Amer. Assoc. Petrol.
       Geol., Vol. 10, pp. 1270-1299.
                                                81

-------
Beeman, R. E. and J. M. Suflita, 1987. "Microbial
  Ecology of a Shallow Unconfined Ground-water
  Aquifer Polluted by Municipal Landfill Leachate."
  Microb. Ecol., Vol. 14, pp. 39-54.
Belyaev, S. S. and M. V. Ivanov, 1983. "Bacterial
  Methanogenesis in  Underground Waters." Ecol.
  Bull, Vol. 35, pp. 273-280.
Bitton, G. and C. P.  Gerba, 1984.  "Ground-water
  Pollution Microbiology: The Emerging Issue." In:
  Groundwater Pollution Microbiology, Bitton, G.,
  and C. P. Gerba, Editors. John Wiley & Sons, New
  York, pp. 1-7.
Bossert, I. D. and L. Y. Young, 1986. "Anaerobic
  Oxidation of p-cresol by a Denitrifying Bacterium."
  Appl. Environ. Microbiol, Vol. 52, pp. 1117-1122.
Bouwer, E. J.  and P. L. McCarty, 1984. "Modeling of
  Trace  Organics Biotransformation in the Sub-
  surface." Groundwater, Vol. 22, pp. 433-440.
Bouwer,  E. J., B. E.  Ritman, and P. L.  McCarty,
  1981. "Anaerobic Degradation of Halogenated 1-
  and 2-Carbon Organic Compounds." Environ. Sci.
  Tecknol, Vol. 15, pp. 596-599.
Craun, G. P., 1984. "Health Aspects of Ground-water
  Pollution." In:  Ground-water Pollution Microbi-
  ology,  Bitton, G., and C. P. Gerba, Editors. John
  Wiley & Sons, New York, pp. 135-195.
Council  on Environmental Quality,  1981. "Contam-
  ination of  Ground Water by  Toxic Organic
  Chemicals." U.S. Government Printing Office,
  Washington, DC.
Dagley, S., 1984. "Introduction."  In: Microbial
  Degradation of Organic Compounds, Gibson, D. T.,
  Editor. Marcel Dekker, Inc., New York, pp. 1-10.
Davis, J.B., 1967. Petroleum Microbiology. Elsevier
  Publishing Co., Amsterdam, p. 604.
DeWeerd, K.,  et al.,  1986. "The  Relationship
  Between Reductive Dehalogenation and Other
  Aryl Substituent Removal Reactions Catalyzed by
  Anaerobes." FEMS Microb. Ecol., Vol. 38, pp. 331-
  339.
Ehrlich, G. G., et al., 1983. "Microbial Ecology of a
  Cresote-contaminated Aquifer at St. Louis  Park,
  Minnesota." Dev. Ind. Microbiol., Vol. 24, pp. 235-
  245.
Ghiorse, W. C. and D. L. Balkwill, 1983.  "Enumer-
  ation  and  Morphological Characterization  of
  Bacteria Indigenous to Subsurface Environments."
  Dev. Ind. Microbiol., Vol. 24, pp. 213-224.
Ghiorse, W. C. and D.   L. Balkwill,  1985.
  "Microbiological Characterization of Subsurface
  Environments." In:  Ground-water Quality,  C. H.
  Ward,  W. Giger and P. L. McCarty, Editors. John
  Wiley & Sons, Inc., New York, pp. 536-556.
Ghiorse, W. C.  and J. T. Wilson, 1988. "Microbial
  Ecology of the Terrestrial Subsurface." Adv. Appl.
  Microbiol., Vol. 33, pp. 107-172.
Gibson, S. A. and J. M. Suflita, 1986. "Extrapolation
  of Biodegradation Results  to  Groundwater
  Aquifers:  Reductive Dehalogenation of Aromatic
  Compounds." Appl. Environ. Microbiol., Vol.  52,
  pp. 681-688.
Godsey, E. M.  and  G.  G.  Ehrlich,  1978.
  "Reconnaissance for  Microbial Activity  in the
  Magothy  Aquifer, Bay Park, New  York, Four
  Years After Artificial Recharge." J. Res. U.S. Geol.
  Survey, Vol. 6, pp. 829-836.
Godsey, E.  M.,  D. F. Goerlitz, and G. G. Ehrlich,
  1983. "Methanogenesis of Phenolic Compounds by
  a Bacterial Consortium from a  Contaminated
  Aquifer in St. Louis  Park,  Minnesota." Bull.
  Environ. Contam. Toxicol., Vol.  30, pp. 261-268.
Hallburg, R. O. and R. Martinell,  1976. "Vyredox - In
  Situ Purification of Ground-water." Ground-water,
  Vol. 14, pp. 88-93.
Harvey, R.  W.,  D.  L. Smith, and L. George, 1984.
  "Effect of Organic Contamination Upon Microbial
  Distribution and Heterotrophic Uptake in a Cape
  Cod,  Massachusetts  Aquifer."  Appl. Environ.
  Microbiol., Vol. 48, pp. 1197-1202.
Hopper, D. J., 1976. "The Hydroxylation of p-cresol
  and Its Conversion to  p-hydroxybenzaldehyde in
  Pseudomonas putida." Biochem. Biophys. Res.
  Commun., Vol. 69, pp. 462-468.
Hopper, D. J., 1978. "Incorporation of [18Q] Water in
  the Formation of p-hydroxybenzyl Alcohol  by the
  p-cresol Methylhydroxylase from Pseudomonas
  putida." J. Biochem., Vol. 175, pp. 345-347.
Humenick,  M. J., L. N.  Bitton,  and C. F. Mattox,
  1982. "Natural Restoration of Ground Water in
  UCG."InSitu, Vol. 6, pp. 107-125.
Hvid-Hansen, N.,  1951.  "Sulphate-reducing and
  Hydrocarbon-producing Bacteria in Ground
  Water." Act. Pathol. Microbiol.  Scand., Vol.  29, pp.
  314-334.
Idelovitch,  E. and M. Michail,  1980. "Treatment
  Effects and Pollution Dangers of Secondary
  Effluent Percolation to Groundwater." Prog. Wat.
  Tech., Vol. 12, pp. 949-966.

Jacks, G., 1977.  "The 'Amber River.' An Example of
  Sulfate Reduction." Proc. Second Int. Symp.  Water-
  Rock Interaction. Strasburg, Aug. 17-25, Vol. I, pp.
  259-266.
Jamison, V.  W., R. L. Raymond, and J. O. Hudson,
  1975. "Biodegradation of High-octane Gasoline in
  Groundwater." Dev. Ind. Microbiol., Vol. 16,  pp.
  305-312.
                                               82

-------
Jannasch, H., et al., 1971. "Microbial Degradation of
  Organic Matter in the Deep Sea." Science. Vol. 171,
  pp. 672-675.
Keswick B .H.} 1984.  "Sources of  Ground-water
  Pollution." In:  Groundwater Pollution Micro-
  biology, Bitton, G., and C. P. Gerba, Editors. John
  Wiley & Sons, New York, pp. 39-64.
Kilbane, J. J., et al., 1982. "Biodegradation of 2,4,5-
  trichlorophenoxyacetic Acid by a Pure Culture of
  Pseudomonas cepacia." Appl. Environ. Microbiol.,
  Vol. 44, pp. 72-78.
Kobayashi, H. and B. E. Rittmann, 1982. "Microbial
  Removal of Hazardous Organic Compounds."
  Environ. Sci. TechnoL, Vol. 16, pp. 170a-183a.
Kuznetsov, S. I., N. V. Ivanov, and N. N. Lyalikova,
  1963.  "The Distribution of Bacteria  in Ground-
  waters and Sedimentary Rocks." In: Introduction to
  Geological Microbiology, C. Oppenheimer, Editor.
  McGraw-Hill Book Co., New York.
Ladd, T. I., et al.,  1982. "Heterotrophic Activity and
  Biodegradation  of  Labile and Refractory  Com-
  pounds by  Ground Water and Stream Microbial
  Populations." Appl. Environ.  Microbiol., Vol. 44,
  pp. 321-329.
Lee, M.'D., et al.,  1988. "Biorestoration of Aquifers
  Contaminated with Organic Compounds." CRC
  Crit. Rev. Environ. Control, Vol. 18, pp.  29-89.
Lee, M.  D.  and C.  H. Ward, 1985. "Biological
  Methods for the Restoration  of Contaminated
  Aquifers." Env. Toxicol. Chem., Vol. 4, pp. 721-726.
Lind, A.  M.,  1975.  "Nitrate Reduction in the
  Subsoil." Proc.  Int. Assoc.  Water Pollut. Res.
  Copenhagen, Aug. 18-20, Vol. 1,-p. 14.
MacKay, D. M., P.  V. Roberts, and I. A. Cherry, 1985.
"Transport of Organic Contaminants  in Ground
Water." Environ. Sci. TechnoL, Vol. 19, pp. 384-392.
McCarty, P. L., 1985. "Application of Biological
  Transformations in Groundwater." Proc. Second
  Int. Conf. Ground-water Quality Res., Durham,  N.
  N. and A. E. Redelfs,  Editors. Natl. Cntr. Ground-
  water Res., Stillwater, OK, pp. 6-11.
McCarty, P.  L., M. Reinhard, and B. E. Rittmann,
  1981.  "Trace Organics in Groundwater." Environ.
  Sci. TechnoL, Vol. 15, pp. 47-51.
McCarty P. L., B. E. Rittmann, and E. J. Bouwer,
  1984.  "Microbiological Processes Affecting Chemi-
  cal Transformations in Groundwater." In: Ground-
  water Pollution Microbiology, Bitton, G., and C. P.
  Gerba, Editors. John Wiley & Sons,  New York.
Mclnerney, M. J.  and M. P. Bryant, 1981. "Basic
  Principles  of Bioconversions  in Anaerobic Diges-
  tion and Methanogenesis." Biomass  Conversion
  Processes for Energy and Fuels, Soferr, S. S. and O.
  R. Zaborsky, Editors. Plenum Publ. Corp., New
  York, pp. 277-296.

McKenna, E. J. and R. E. Kallio, 1964. "Hydrocarbon
  Structure: Its Effect on Bacterial Utilization of
  Alkanes." In: Principles and Applications in
  Aquatic Microbiology, H.  Heukelekian and N. C.
  Dondero, Editors. John Wiley & Sons, New York,
  pp. 1-14.

McNabb, J. F. and W. J. Dunlap, 1975. "Subsurface
  Biological Activity in Relation to Groundwater
  Pollution." Groundwater, Vol. 13, pp. 33-34.
McNabb, J. F. and G.  E.  Mallard, 1984. "Micro-
  biological  Sampling in the Assessment of
  Groundwater  Pollution." Groundwater Pollution
  Microbiology,  Bitton, G. and C. P. Gerba, Editors.
  John Wiley & Sons, New York, pp.  235-260.
Olson, G. S., et al., 1981. "Sulfate-reducing Bacteria
  from Deep Aquifers in Montana." Geomicrobiol J.,
  Vol. 2, pp. 327-340.
Olson, G. J., G. A. McFeters, and K. L. Temple, 1981.
  "Occurrence and  Activity of Iron and Sulfur-
  oxidizing Microorganisms in Alkaline Coal Strip
  Mine Spoils." Microb. EcoL, Vol. 7, pp. 40-50.

Page, G. W., 1981. "Comparison of Groundwater and
  Surface  Water for  Patterns  and  Levels of
  Contamination  by Toxic Substances."  Environ.
  Sci. TechnoL, Vol. 15, pp. 1475-1481.

Preul,  H. C., 1966.  "Underground Movement of
  Nitrogen." Adv. Water Pollut. Res.: Proc. Third
  Int. Conf., pp. 309-328.
Pritchard,  P. H., 1981.  "Model Ecosystems." In:
  Environmental Risk Analysis for Chemicals, R. A.
  Conway, Editor. Van Nostrand  Reinhold, New
  York, pp. 257- 353.
Pritchard, P. H. and A. W. Bourquin, 1984.  "The Use
  of Microcosms for Evaluation of Interactions
  Between Pollutants and Microorganisms." In: Adv.
  Microbial.  Ecology,  K.  C. Marchall,  Editor.
  Plenum Publishing Corporation, Vol. 7, pp. 133-
  215.
Pye, V. I. and R. Patrick, 1983. "Groundwater Con-
  tamination in the United States." Science, Vol.
  221, p. 713.
Rajagopal, B. S., et al., 1984. "Effect and Persistence
  of Selected Carbamate Pesticides in Soil." Res.
  Rev., Vol. 93, pp.  1-199.
Raymond, R. L., V. W. Jamison, and J. O. Hudson,
  1976. "Beneficial Stimulation of Bacterial Activity
  in Ground Water Containing Petroleum Products."
  Aice Symposium Series, Vol. 73, pp. 390-404.
Roberts, P. V., et  al., 1980. "Organic Contaminant
  Behavior During Groundwater Recharge." J. Water
  Pollut. Control Fed., Vol. 52, pp. 161-172.
                                               83

-------
Rogers, J. E., et al., 1985. "Microbial Transformation
 of Alkylpyridines in Ground-water." Wat. Air Soil
 Poll., Vol. 24, pp. 443-454.
Senior, E. and M. T. M. Balba, 1984. "The  Use of
 Single-stage and Multi-stage Fermenters to Study
 the Metabolism of Xenobiotic  and Naturally
 Occurring Molecules by Interacting Microbial
 Associations." In: Microbiological Methods for
 Environmental Biotechnology, Grainger, J.  M. and
 J.  M.  Lynch, Editors.  Society for  Applied
 Bacteriology, Academic Press, Inc., Orlando, FL.
 pp. 275-293.
Slater, J. H. and D. Lovatt,  1984.  "Biodegradation
 and the Significance of Microbial Communities."
 Microbial. Degradation of Organic Compounds, D.
 T. Gibson, Editor. Marcel Dekker, Inc., New York,
 pp. 439-485.
Smolenski, W. J. and  J. M.  Suflita, 1987. "Biode-
 gradation of Cresol Isomers in Anoxic Aquifers."
 Appl. Environ. Microbiol., Vol. 53, pp. 710-716.
Suflita, J. M. and S. A. Gibson, 1985. "Biodegra-
 dation of Haloaromatic Substrates in a Shallow
 Anoxic  Groundwater Aquifer." Proc.  Second Intl.
 Conf. Groundwater Quality Res., N. N. Durham
 and A.  E. Redelfs, Editors. Natl. Cntr. Ground-
 water Res., Stillwater, OK, pp. 30-32.
Suflita, J. M. and G. D. Miller, 1985. "The Microbial
 Metabolism of Chlorophenolic  Compounds  in
 Groundwater Aquifers." Env.  Toxicol. Chem., Vol.
 4, pp. 751-758.
Updegraff, D. M., 1982. "Plugging and Penetration of
  Petroleum Reservoir Rock by Microorganisms."
 Proceedings of 1982 International Conference  on
  Microbial Enhancement of Oil Recovery, May 16-
  21, Shangri-La, Afton, OK.
van Beek, C. G. E. M. and D. van  der Kooij, 1962.
 "Sulfate-Reducing Bacteria in Ground Water from
 Clogging and Nonclogging Shallow Wells in the
 Netherlands River Region." Groundwater, Vol. 20,
 pp. 298-302.
Ventullo, R. M. and R. J. Larson,  1985. "Metabolic
 Diversity and Activity of Heterotrophic Bacteria in
 Ground Water." Env. Toxicol. Chem., Vol. 4, pp.
 321-329.
Waksman,  S. A., 1916. "Bacterial Numbers  in Soil,
  at Different Depths, and in Different Seasons of
  the Year." Soil Science, Vol. 1, pp. 363-380.
Ward, T. E., 1985. "Characterizing the Aerobic and
  Anaerobic  Microbial Activities in Surface and
  Subsurface Soils." Environ. Tox. Chem., Vol. 4, pp.
  727-737.
Webster, J. J., et al., 1985.  "Determination of
  Microbial Numbers in Subsurface Environments."
  Groundwater, Vol. 23, pp. 17-25.
White, D. C.,  et al., 1985. "Biochemical Measures of
  the Biomass, Community  Structure, and Meta-
  bolic Activity of the Ground Water Microbiota."
  In: Ground Water Quality,  C. H. Ward, W. Giger,
  and P. L. McCarty, Editors. John Wiley & Sons,
  Inc., New York, pp. 307-329.
White, D. C., et al., 1983. "The Groundwater Aquifer
  Microbiota: Biomass Community Structure and
  Nutritional Status." Dev. Ind. Microbiol., Vol. 24,
  pp. 201-211.
Wilson, J. T., et al., 1983. "Enumeration and Charac-
  terization of Bacteria Indigenous to a Shallow
  Water-table Aquifer."  Groundwater, Vol. 21, pp.
  134-142.                         ;
Wilson, J. T., et al,, 1985, "Influence of Microbial
 . Adaption on the Fate of Organic Pollutants in
  Ground Water." Env. Toxicol. Chem., Vol. 4, pp.
  743-750.              ..-',.,
Wilson, J.  and M. J. Noonan, 1984. "Microbial
  Activity  in Model Aquifer Systems." In: Ground-
  water Pollution Microbiology, Bitton, G., and C. P.
  Gerba, Editors. John Wiley & Sons, New York, pp.
  117-133.
Wilson, J. T.,  M. J. Noonan, and J. F.! McNabb, 1985.
  "Biodegradation of Contaminants in the Sub-
  surface." In: Ground Water Quality, C. H. Ward,
  W. Giger, and P. L. McCarty, Editors. John Wiley
  & Sons, Inc., New York, pp. 483- 498.
Wood,  P. R., R. F. Lang,  and I. L. Payan,  1985.
  "Anaerobic Transformation,  Transport and
  Removal of Volatile Chlorinated Orgariics in
  Ground Water." In: Ground Water Quality, C. H.
  Ward, W. Giger, and P. L. McCarty; Editors. John
  Wiley & Sons, Inc., New York, pp. 493-511.
Zoetman, B. C. J., 1985. "Overview of Contaminants
  in Ground Water." In: Ground Water Quality, C. H.
  Ward, W. Giger, and P. L. McCarty, Editors. John
  Wiley & Sons, Inc., New York, pp. 27-38.
                                                84

-------
                                       CHAPTER 8

             MICROBIOLOGICAL PRINCIPLES INFLUENCING  THE
                          BIORESTORATION OF AQUIFERS

                                       Joseph M. Suflita
Introduction
Ground-water pollution problems can be enormously
complex. Symptoms of aquifer contamination include
odor and taste problems and the occasional appear-
ance of a free product phase in wells. Besides the
health concerns associated with the latter type of
contamination, volatile and flammable fumes can
accumulate in buildings and result in explosions
and/or inhalation risks. Ground-water contamina-
tion also can conceivably impact surface water
supplies and lead to more obvious signs of pollution
Including color and odor problems, fish kills, and if
an immiscible chemical is involved, the formation of
seeps or slicks.

The purpose of this chapter is  to briefly consider
various treatment options for cleaning up contam-
inated aquifers and to illustrate how biorestoration
techniques fit into the myriad of pollution mitigation
tools. This chapter also describes the types of consid-
erations that must be made prior to implementing a
field aquifer biorestoration program. An example of
spilled gasoline in an aquifer is chosen to illustrate
how basic microbiological and biochemical principles
meld into an overall aquifer treatment strategy. In
addition, guidelines are provided for the critical
evaluation of claims for aquifer restoration,  with
suggestions for the types of information that might
be  collected  to support such claims. Particular
attention is  paid to in situ biorestoration attempts
that rely on the inoculation of desirable micro-
organisms. Lastly, a perspective on bioreclamation
techniques is provided through a consideration of the
practical limitations of the technology.
Environmental Fate of
Contaminants and Treatment Options

At first glance, it may seem unusual to consider the
environmental fate of a contaminant together with
various options for the abatement of that contam-
inant. However, these  topics are governed by the
same two phenomena - the transport characteristics
of a pollutant and the  reaction of the contaminant
with the environment.

Table 8 generally depicts the fate of a contaminant in
a specific environment in terms of movement, reten-
tion, and reaction  processes as a function  of the
properties of both the environment and the contam-
inant. The fate of a contaminant is largely a function
of the chemistry of the pollutant in its environment.
Similarly, the reaction  mechanisms that a pollutant
may undergo (i.e., hydrolysis, precipitation,  oxida-
tion/reduction, etc.) are a function of that chemical
and the existing environmental conditions.  These
various processes are  interrelated  (Table 8).  For
instance, a chemical may be inherently susceptible
to microbial  attack, but is deposited in an environ-
ment where the production of low molecular weight
acids from the microbial metabolism of other forms of
organic matter results  in a decrease in overall pH.
The immediate pH conditions may  prevent the
continued metabolism of that chemical. The inter-
relationships between environmental conditions,
chemical structure, and the physiology of micro-
organisms are described in Chapter 7.

Treatment options for contaminated aquifer remedi-
ation are governed by the same basic characteristics
                                               85

-------
 (Table 9). Transport and reactivity properties of a
 contaminant can be envisioned as a continuum from
 low to high for each characteristic, and can help
 decision-makers who are considering an aquifer
 treatment option. For example, if a contaminant does
 not transport well and tends  to be a recalcitrant
 molecule, containment may  be the most  likely
 treatment option. However, if the pollutant is subject
 to desirable biotransformation reactions and reactiv-
 ity  is high, then a bioremediation approach  to
 contaminated aquifers should at least be considered.
 The simple analysis in Table 9 suggests that restora-
 tion efforts centered on biodegradation strategies are
 not applicable in all circumstances.
 Table 8.    Summary of  the Mechanisms Influencing the
           Fate of Contaminants in Environments
Process
Movement
Retention
Reaction
Environmental
Conditions
Water Flow Rate
Formation
Permeability
Water Motion
Gravity
Surface Tension
Soil/Sediment
Organic Matter
Content
Sorptive Capacity
pH
Redox Status
Microbial
Communities
Contaminant
Amount of Material
Physical State
Solubility
Viscosity
Type Solubility
Ionic Character
Chemical
Transformation
Biodegradability
Table  9.
 Reactivity
          Treatment Options as a Function of the Trans-
          port and Reactivity Characteristics of the
          Contaminant
                            Transport
      High
    Low
 High

 Low
Extraction and/or
 Biodegradation

Extraction
Biodegradation
                                  Containment
In fact, all remediation technologies have inherent
limitations that must be kept in mind when evalu-
ating a variety of treatment options. For instance,
containment techniques generally rely on physical or
hydrological barriers to keep a  contaminant from
spreading underground. Typical containment meth-
ods involve slurry walls,  clay  caps, interceptor
trenches, or hydraulic barriers. These techniques can
be costly, plus there is no guarantee of their long-
term effectiveness and, most importantly, they do
not really address the pollution problem.
 Similarly, extraction-based remediation efforts also
 have disadvantages. As noted in Table 9, the technol-
 ogy tends to be used on substances that are easily
 transported in the subsurface. The simplest extrac-
 tion approach is to excavate the contaminated soils
 and sediments in the problem area. However, this
 often is not feasible because of the  depth of the
 problem area or because excavation activities could
 undermine the  foundation of roads, buildings, or
 other structures.

 The pumping of ground water is another extraction
 technique. Ground-water  pumping technologies
 generally are combined with some sort of surface
 treatment like air stripping, carbon  adsorption,
 biological or chemical reaction, or simple discharge
 to other locations. Often such extraction techniques
 tend  to  treat only a  limited  amount  of the total
 contaminant problem;  that is, only contaminants
 dissolved in the water phase are removed from the
 aquifer.  Many  decades often are needed  to extract
 even mildly adsorbing chemical pollutants.

 One  of  the  major  disadvantages of extraction
 technologies is that the contaminants  often are
 merely transferred from one location to another. The
 problems associated with trans-media  pollution of
 different environmental compartments continue to
 gain  increasing attention  among the regulatory
 community. Other methods, such as soil venting,
 tend to address volatile contaminants present in the
 soil atmosphere, yet again, only address a portion of
 the problem.

 Extraction often tends to be the treatment of choice
 because it is straightforward, understandable, and
 predictable and generally has small yearly invest-
 ment costs. However, this approach ignores the long-
 term  commitment to the extraction process, partic-
 ularly for sorbed contaminants and the maintenance
 involved in the treatment operation.

 Reaction technologies include both chemical and
 biological methods of converting contaminants  to
 more environmentally acceptable forms. Generally,
 chemical  transformations  convert  pollutants
 through hydrolysis  or initial oxidations. Such
 reactions also can be catalyzed by microorganisms,
 but they often are the rate-limiting step in biological
 treatment scenarios. On the other hand,  biological
 treatment processes have the potential to completely
 mineralize contaminants  and convert them to
 innocuous substances like carbon dioxide, water, and
cell material. Reaction processes, when successful,
tend to be relatively short-term solutions to pollution
problems and are  associated  with high operation
costs. The treatment chemicals  and/or microbial
nutrients must  be transported  to the contaminated
                                                86

-------
 portion of the aquifer so that both the dissolved and
 adsorbed phases of the contaminant are attacked.


 Aquifer Remediation Considerations
 Many considerations must be made before seriously
 considering an aquifer bioremediation  approach
 (Lee, et al., 1988). The first  consideration  is
 determining the contaminant  type, the phases  in
 which it exists, the solubility of the material, and its
 inherent  susceptibility to biodegradation.  The
 microbial ecology scientific literature is an excellent
 source for the latter consideration. Biodegradation
 information resulting from the study of a  variety of
 environments indicates whether  the contaminant
 can be degraded and, if so, under what ecological
 conditions.  This information does  not insure  that a
 contaminant will be degraded in all instances since a
 variety of site-specific characteristics can interplay
 to influence the success of bioremediation efforts.

 The scientific literature also can provide  some as-
 sessment of the likely  pathways of degradation,
 which  in  turn allows  for an evaluation of the
 consequences of biorestoration.  Disappearance  is
 only one aspect of biodegradation since a contam-
 inant can be converted to a series of metabolites  that
 also are of environmental concern. For instance, the
 reductive removal of tetrachloroethylene (TeCE)
 under anaerobic conditions leads to a series of
 dehalogenated intermediates. TeCE's halogens  are
 removed and replaced  by protons in a series of
 sequential  steps. However,  the rate of reductive
 dehalogenation tends to  decrease as fewer and fewer
 halogens remain on the ethylene moiety. Conse-
 quently, vinyl chloride accumulates and, from  a
 regulatory viewpoint, causes greater concern than
 the parent contaminant. Thus, biodegradation path-
 way information can be critical for the design of
 bioremediation efforts.

 Successful  implementation of biorestoration efforts
 also depends on determining site characteristics, in
 particular, the site-specific hydrogeological vari-
 ables. Factors such as thickness of the vadose zone,
 its permeability,  geologic complexity, and organic
 matter content can impact biorestoration programs.
 Saturated zone characteristics such as the  type  and
 composition of an aquifer, its permeability, thick-
 ness, interconnection to other aquifers, location to
 discharge areas, magnitude of  water table fluctu-
 ations, and  ground-water flow  rates all influence
 aquifer bioremediation planning and decision-
 making. To date, bioremediation has been attempted
 in aquifers possessing  a variety  of flow charac-
 teristics including pumping rates ranging from 25 to
 380 L per minute, and flow rates from 0.6 to 800 m
per year or hydraulic conductivities of 10-5 to 10-3 cm
 per second. Generally, bioremediation efforts center
 around more permeable aquifer systems where the
 movement of ground water can be more successfully
 controlled.

 Another consideration  in  biorestoration is the
 removal of free product. This is extremely important
 since many substances, while suitable nutrients for
 microbial growth when  present at low  concen-
 trations, are inhibitory at high concentrations.  Such
 concentration effects can  easily be observed  with
 substances ranging from simple sugars to gasoline.
 Consequently, in a biorestoration effort, it is impor-
 tant to remove as much free product as possible by
 physical or  chemical means and use biological
 methods to treat the remainder.

 The next biorestoration consideration is whether the
 system design will be above or below the  ground.
 Again, this decision is influenced by the ease of
 transporting the contaminant from its location to the
 area where  the  treatment will take place. Since
 many contaminants are in multiple phases in the
 subsurface, in situ treatment strategies often are the
 best option. In in situ treatment efforts, nutrients are
 transported to the  requisite microorganisms in an
 effort to create the correct environmental conditions
 for microbial proliferation.

 A critical component in biorestoration is a laboratory
 evaluation of the site pollutants' susceptibility to
 biodegradation. The appropriate use of microcosms
 in rigorously defined and controlled  experiments
 allows the investigator to  evaluate biodegradation
 under conditions that are more environmentally
 realistic. The microcosm approach allows the investi-
 gator to incorporate numerous site-specific variables
 without a detailed knowledge of what these variables
 actually are.  For  instance,  scientific literature  may
 indicate that a particular pollutant is subject to bio-
 degradation,  but this may not  be confirmed by
 microcosm studies due to the absence of an essential
 microbial nutrient or the presence of an inhibitory
 substance at the site.  Some of these limitations  may
 be relatively easy to  overcome, while others may
 prove more difficult.

 If initial microcosm evaluations are unsuccessful,
 subsequent studies should attempt to stimulate
 biodegradation of the problem contaminants using
 nutrients or adjusting other variables. If biodegra-
 dation in microcosms cannot be successfully stimu-
 lated, the chances of success in the field are minimal
 but if initial microcosm evaluations prove successful,
 the chances for success in the field are significantly
 improved. These preliminary laboratory investiga-
 tions not only allow an investigator  to design a
biorestoration  program, but  also to optimize  the
treatment strategy.
                                                87

-------
As a final consideration, some plan should be devised
to monitor the progress of a biostimulation activity.
A more detailed discussion of the types of informa-
tion needed for monitoring are presented later in this
chapter.

From a bioremediation viewpoint, the ideal site has:
(1) a homogeneous and permeable aquifer; (2) a con-
taminant originating from a single point source; (3) a
low ground-water gradient; (4) no free product; (5) no
soil contamination; and (6) an easily degraded, ex-
tracted, or immobilized contaminant. Obviously, the
above combination of characteristics describes very
few sites. The following sections will attempt to show
how information on the chemistry, microbiology, and
biochemistry of a contaminant can form a  bio-
restoration strategy.
Pathways of Hydrocarbon Metabolism
One of the most frequently encountered aquifer con-
taminants is gasoline, which is largely a mixture of
hydrocarbons including three major chemical
classes: the alkanes; the cycloalkanes; and aromatic
chemicals (Table  10). Table  10 illustrates that
greater than 73 percent of several  gasolines are
composed of these three hydrocarbon classes. Gaso-
line hydrocarbons within each class possess a variety
of substituents and numerous substitution patterns
that ultimately influence the susceptibility of the
resulting structures to biodegradation.

Table 10.  Major Chemical Components (Percent of Total)
          of the Gasoline Fraction of Several Petroleums
          (Adapted from Perry, 1984)
Hydrocarbon
Class <
Alkanes
Cycloalkanes
Aroma tics

Texas
16.8
47.1
19.5
Gasoline from
California
18.0
55.5
10.2

Louisiana
24.5
38.4
15.6
Alkanes exist in gasoline  and many gasoline
components contain an alkane moiety in their carbon
skeleton. An enormous amount of scientific litera-
ture exists on the types of alkanes that are subject to
microbial attack and the diversity of microorganisms
involved in alkane metabolism as detailed by Singer
and Finnerty (1984).  An extremely wide  range of
alkanes are susceptible to biodegradation, but some
microorganisms preferentially  utilize only short
chain molecules while others metabolize only longer
chain structures.
Microorganisms can initiate the aerobic degradation
of alkanes through mechanisms that involve mono-,
di-, or subterminal oxidation of the molecule. Prob-
ably the most frequently reported degradative route
involves the oxidation of a terminal methyl group to
convert the alkane to an alcohol of equivalent chain
length (Figure 76). This alcohol probably undergoes
a series of dehydrogenation steps  to form the
corresponding aldehyde and eventually a fatty acid
(Singer and Finnerty, 1984).

Once formed, the fatty acid is further metabolized by
{5-oxidation. This process  liberates  a two-carbon
fragment as acetylcoenzyme A, a central metabolic
intermediate. Two carbon fragments are then con-
tinuously cleaved from the resulting shorter fatty
acid until the entire molecule is  degraded. Alkanes
with odd numbers of carbon atoms ultimately yield a
molecule of another central metabolite - propionyl
coenzyme A. The significant aspect of Figure 76 is
that the initial oxidation  of an alkane molecule
involves the incorporation of molecular oxygen. This
basic theme also is pertinent to the other major
classes of gasoline hydrocarbons.
CH3 - (CH2)n
 + 02.2Hl
          P^-
 CH3 - (CH2)n - CH3 Alkane

                                  / Methyl
                H20              ! Group

 CH3 - (CH2)n - CH2OH Alcohol

          < ~2H
 CH3 — (CH2)n — CHO Aldehyde
  H20  -v  -2H
         Ti r                    • J
 CH3 — (CH2)n — COOH  Fatty Acid
                                                            I
                                                    0 -Oxidation
                                                        CO2
Figure 76.
                                                            t
           The initial steps  in the  aerobic  microbial
           metabolism of alkane hydrocarbons.
It is more difficult to enrich for microorganisms that
are able to use cycloalkanes as a sole source of carbon
and energy. This appears to be widespread in nature
and is discussed by Trudgill (1984) and Perry (1984).
It may be that microorganisms interact frequently in
commensalistic relationships based on the cometab-
olism of these substrates (cometabolism will be dis-
cussed later in this chapter). However, it is clear that
individual or mixtures of microorganisms can metab-
olize cycloalkanes in the manner shown for  cyclo-
hexane (Figure 77).
                                                88

-------
O-*
                             2H
           Cyclohexane
                            H2O
                                          -2H
               O2 + 2H
                          Cyclohexanol
Cyclohexanone
                                                                    H20
                                    COOH

                                    CH2
       0-Caprolactone
                         CH2

                         CH2
                                              |3-Oxidation
                                                                •+• CO2 + H2O
Figure 77.
                         CH2
                          I
                         COOH

                     Adipic Acid
The Initial steps in the aerobic microbial metabolism of alicyclic hydrocarbons.
As shown in Figure 77, cyclohexane hydroxylation
by a microbially produced monooxygenase leads to
the formation of an alicylic alcohol. Subsequent
dehydrogenation of the alcohol forms a ketone and
further oxidation of the  ketone results in the
formation of a lactone ring structure. The lactone is a
suitable substrate for ring opening and is eventually
converted to a dicarboxylic acid  which, in turn, is
subject to oxidation. Molecular oxygen participates
in the biodegradation pathway, but, in this case, is
involved in two separate steps.

Aromatic compounds of gasoline  also are subject to
microbial attack by many different types of bacteria
and fungi. The aromatic hydrocarbons are aero-
bically metabolized by bacteria  to dihydroxylated
compounds through  cis-dihydrodiol intermediates.
Figure 78 shows how, in the case of benzene,  a
bacterial dioxygenase incorporates both  atoms of
molecular oxygen to form cis-benzene dihydrodiol
which is subsequently dehydrogenated to  result in
the formation of catechol. Catechol is then a suitable
substrate for ring cleavage. Other dioxygenases open
the ring most often via an 'ortho' or 'meta' cleavage
route. The subsequent intermediates produced by
these pathways eventually enter central metabolic
reaction sequences of the bacterial cell. The aerobic
catabolism of homocyclic aromatic compounds is
reviewed by Gibson and Subramanian (1984) and by
Bayly and Barbour (1984).

Oxygen plays a critical role in the metabolism of
aromatic hydrocarbons. All of the metabolic path-
                                       ways discussed above require oxygen as a coreactant.
                                       In addition, the organisms catalyzing these biocon-
                                       versions use oxygen as a terminal electron acceptor
                                       (described in Chapter 7). Hydrocarbon metabolism
                                       puts a large demand on oxygen resources; therefore,
                                       plans for biorestoration  activities should consider
                                       how this oxygen demand will be supplied.

                                       In addition to oxygen,  other potential  limiting
                                       nutrients also must be supplied in suitable form for
                                       the microorganisms to proliferate at the expense of
                                       the hydrocarbons. Along with a suitable environ-
                                       ment, microorganisms need nitrogen, phosphorus,
                                       sulfur,  and trace elements and without  the two
                                       former nutrients, hydrocarbon metabolism may be
                                       limited  even  when oxygen supplies are adequate.
                                       With the proper nutrients, the microorganisms can
                                       convert gasoline hydrocarbons  to the  environ-
                                       mentally innocuous  products of  carbon dioxide,
                                       water, and additional cell material.

                                       Once the chemical and physical requirements for
                                       microbial growth and the  predominant metabolic
                                       pathways are known, attempts can be  made to
                                       stimulate  the biodegradation  of gasoline by
                                       superimposing the correct nutrients in situ. Figure
                                       79A represents an aquifer contaminated with gaso-
                                       line hydrocarbons. Free  product already has been
                                       removed from this schematic site, but both water and
                                       soil contamination still exist.  Preliminary testing
                                       shows that hydrocarbon-degrading microorganisms
                                       are present and active at the  site and that  these
                                       organisms can be stimulated to increase hydrocarbon
                                              89

-------
                Benzene
                     cis-Benzene Oihydrodiol
                                           Muconic Semialdehyde
                                                                            COOH
                                                                            COOH
                                                                   cis, cis-Muconic Aci<
_ Figure 78.
                                              Mineralization
                                               CO2 + H2O

The initial steps in the aerobic microbial metabolism of aromatic hydrocarbons.
  biodegradation. Figure 79B depicts the attempt to
  gain hydrogeologic control over the contaminated
  area through the insertion of a series of injection and
  recovery wells. The goal is to recirculate the ground
  water in an effort to create an underground reaction
  chamber. The figure shows that the pumping and
  recirculation also change the water table. Ideally,
  the 'reaction chamber1 will  encompass the contam-
  inated area; however, this often is not possible due to
  the magnitude of the contamination problem and the
  permeability limits of the aquifer. In the latter case,
  the contamination  problem can be approached in
  stages that are governed largely by the efficiency of
  water recirculation at the site.
  Nutrients are then added to the recirculation system
  and transported with the injection water into the
  aquifer (Figure 79C) where the indigenous micro-
  organisms will start to proliferate (the  bioactive
  area). The efficiency of the nutrient transport will
  then dominate the success of the remediation effort.
  A variety of mechanisms for supplying oxygen to the
  requisite microorganisms is considered in the review
  byLee,etal.(1988).
  Initially, most of the bioactive area probably will be
  centered in areas adjacent to the injection well. As
  the electron donors in that area are  depleted, the
  major bioactivity shifts to other areas of the contam-
  inant plume where additional hydrocarbon and,
  thus, electron donors exist, and the requisite nutri-
                                        ents penetrate the plume (Figure 79D). This process
                                        continues until the site is considered remediated
                                        (Figure  79E). At this point, the nutrients  are
                                        removed from the recirculation stream. The aquifer
                                        tends to exhibit homeostatic controls (described
                                        below) and the microflora that grew in response to
                                        the hydrocarbon and nutrient input return to the
                                        levels originally present before the contamination
                                        incident.

                                        Bioreclamation of gasoline  in  aquifers has an
                                        excellent chance of success because of the wealth of
                                        scientific information that forms the basis of this
                                        applied technology. Researchers have studied hydro-
                                        carbon  metabolism by microorganisms for  many
                                        decades and learned that gasoline hydrocarbons are
                                        xenobiotic substrates only in the sense that they
                                        occur in various environmental media at higher than
                                        acceptable concentrations.  Most hydrocarbons in
                                        gasoline are natural substrates and microbial com-
                                        munities have evolved mechanisms for their degra-
                                        dation. These hydrocarbon-degrading microorgan-
                                        isms are not unusual and tend to be ubiquitously
                                        distributed in many and probably most natural
                                        environments.  Scientific  investigations also have
                                        helped elucidate the nutrient requirements neces-
                                        sary for microbial growth and proliferation at the
                                        expense of hydrocarbon substrates so  that remedia-
                                        tion activities are relatively easy to construct, test in
                                        microcosms, and extrapolate to the field.
                                                   90

-------
                   ^-Contominoted oreo
                                   "-.Original
                                     Water Table
       Recycled
       Ground
       Woter-r
                                    New
                                    Water Table

                                  "-Original
                              /     Water Table
       Recycled
       Ground
       Water
      Mutrient Flow
] Bioactive Area
                                     New
                                     Water Table

                                     Original
                                     Water Table
                                     New
                                    Water Table
                                    Original
                                    Water Table
                                       New
                                     Water Table
                                     Original
                                     Water Table
                                       \ Contaminant
Figure 79.   Schematic illustration of the various steps
           (A-E)  in an aquifer biorestoration program
           (see text for description).
A similar information base does not exist for more
exotic substrates; therefore, microcosm research may
be necessary in order to design effective remediation
processes. Principles gleaned from the  study of
hydrocarbon degradation may or may not be applica-
ble to other contaminants.
Critical Evaluation of Biorestoration
Claims have been made that a variety of hydro-
carbons, solvents, and other contaminants were
treated by in situ bioremediation (Table 11) (Lee, et
al., 1988). However, these claims generally lack firm
scientific evidence that biorestoration actually was
                                responsible for the removal of the substances. Many
                                claims for in situ biorestoration tend to appear in
                                publications that were not peer reviewed, therefore,
                                legitimate questions persist as to the effectiveness of
                                biological treatment programs for various contami-
                                nant types. Consequently, clients or regulators have
                                difficulties evaluating such claims. Since biorestora-
                                tion is still an emerging technology, practitioners
                                must gather the necessary evidence for prospective
                                clients and regulators to critically interpret  the
                                chances of success for such programs.

                                Table 11.   Contaminants Treated by In Situ Bioremedi-
                                          ation (Adapted from Lee, et al., 1988)
                                                        Contaminant Type
                                                             Representatives
                                                      Hydrocarbons
                                 Solvents
                                Other Compounds
                              gasoline
                              mineral oil
                              aliphatic plasticizers

                              methyl chloride
                              n-butanol
                              acetone
                              ethylene glycol
                              isopropanol
                              tetrahydrofuran
                              chloroform

                              dimethyl aniline
In remedial activities, practitioners should carefully
document the reduction in substrate concentration(s)
as a result of their efforts. Implicit in this suggestion
is reliable information on mass balances. While it is
often impossible to estimate the quantity of material
lost in an accident, sometimes such estimates are
feasible. For instance, logs are kept of inputs and
removals of chemical substances from storage facil-
ities. This type of information may prove valuable in
estimating the maximum amount  of contaminant
involved in a pollution incident. Alternately, the
remedial efforts can be gauged relative to the initial
amount of contaminant present in the aquifer. This
approach requires good definition of the plume,  an
appreciation for where the contaminant exists, and
reliable and specific measurement methods. If biore-
mediation is successful, contaminant levels should
begin to fall in areas that receive the treatment and
remain relatively unchanged in areas that do not.

Bioremediation efforts often seek  to increase the
activity of microorganisms  in an aquifer. The
confidence in bioremediation technologies  would be
greater if the increase in microbial  numbers and/or
activities were quantitated relative to: (1)  plume
areas prior to any treatment;  (2) areas within the
plume that did not receive the treatment; or  (3)
control areas outside the plume. The latter will give
some indication of background levels of micro-
organisms and allow comparisons to be made before,
during, and after a remediation program.
                                                  91

-------
Often, the production of microbial catabolites can be
measured in areas that receive treatment but not in
contaminated areas left untreated. For instance, the
production of a variety of lesser halogenated PCB
congeners in river  sediments contaminated with
Aroclors was previously noted (Brown, et al., 1987;
Brown, et al., 1984). Many of the congeners produced
did not comprise a significant portion of the original
contaminant Aroclors. Their  production  was sug-
gested as field evidence  that the original  PCB
materials were metabolized  by anaerobic micro-
organisms and reductively dehalogenated (Brown, et
al., 1987; Brown, et  al.,  1984). Laboratory evidence
in microcosm experiments designed to  test this
hypothesis confirmed that suspicion (Quensen, et al.,
1988). Similarly, the production of metabolites like
hydrogen sulfide or methane often are testament to
the metabolic activities  of microorganisms. If the
degradation  pathways of particular contaminants
are known prior to a remedial  effort, it is sometimes
possible to specifically assay  for the production of
likely catabolites.

Many times, biorestoration programs rely on the
contaminant itself to supply the  electron donors
necessary for microbial  growth and proliferation.
One of the objectives of the remediation is to supply
the necessary electron acceptor. The consumption of
added terminal electron  acceptors could possibly be
measured. For instance, if oxygen is added to help
stimulate hydrocarbon metabolism in aquifers, its
depletion in the treatment area should be relatively
easy to measure.

Finally, rapid microbial biodegradation typically
seems to be  preceded by a variable length of time
where little or no activity is measured. This length of
time is referred to as the adaptation or acclimation
period (Linkfield, Suflita, and Tiedje, 1989). So far,
adaptation seems to be a unique biological response.
Observation  of this type  of phenomenon in response
to a biorestoration effort can be taken as presump-
tive evidence that microbial action is operative.

It is  important to  emphasize that  all  of the
techniques and measurements  suggested above must
be made relative to  appropriate controls. For some
techniques, that may mean the assay of untreated
areas or areas outside the plume. Many clients may
not be willing to invest the  necessary time and
financial resources to gather  this level of informa-
tion.  Even with controls, the  evidence garnered in
the above fashion tends to be largely circumstantial.
However, the stronger the evidence, the greater the
degree of confidence in the technology.
Lag, Adaptation, or Acclimation Periods
In biodegradation studies, a period of time often is
observed where very little substrate is turned over
and correspondingly little product is formed.  This
phase of 'metabolism' is potentially more environ-
mentally significant than other phases because of its
possible effects on the level of the exposure of
humans and ecosystems to specific pollutants. The
lag, adaptation, or acclimation phase can be one of
the most frustrating portions of bioremediation
programs, since, despite all efforts, virtually nothing
seems to be happening to the problem contaminant.

For example, Figure 80A«'shows the reductive
dehalogenation of 2-bromobenzoate in anoxic sedi-
ment microcosms.  The substrate lags for several
weeks after which  it is rapidly metabolized. The
production of an intermediate  catabolite, benzoic
acid, occurs before its rapid degradation. Ultimately,
the substrate and intermediate are converted to the
gaseous endproducts,  methane and carbon dioxide.
At this  point, the microorganisms are considered
acclimated to the degradation of the parent substrate
and subsequent substrate additions will be degraded
without a  lag. This phenomenon is graphically
illustrated for the  reductive dehalogenation of a
related substrate 3-iodobenzoate (Figure  SOB).  In
this case, the second addition of 3-iodobenzoate is
presented as time  zero. Note the  immediate  con-
sumption of the substrate, the much lesser accumu-
lation of benzoic acid, and the more rapid conversion
of the parent molecule to gaseous endproducts.

In the examples  given in Figure 80A and B,
acclimation might reasonably be expected since the
microorganisms were able to mineralize the parent
substrates. Some of the carbon and energy gained
from the metabolism of those substrates  was
presumably used for the proliferation of the  cata-
lyzing microbial communities. If the concentration of
the requisite microorganisms was initially very low,
the lag period could conceivably be a function of the
time required for these organisms to  grow  to
sufficient density to effect some significant amount
of substrate depletion relative  to the  large back-
ground added as the amendment. However, acclima-
tion also is observed for substrates that are not
mineralized. As seen in Figure 80C, the substrate 4-
amino-3,5-dichlorobenzoate  can be reductively de-
halogenated to form the monohalogenated product
following a three-week lag period. Even though the
parent substrate is not mineralized and does not
supply carbon for microbial growth, additions of the
parent substrate following the removal of the initial
                                                92

-------
                                                                         ,4-Amino-3,5
                                                                           dichloro-
                                                                           banzoate
                                                          4-Amino-3-chloro
                                                               btnjoalt
                                                                                             100
                                                                                     oc.
                                                                                     o
                                                                                             50°
                                                                                     CE
                                                                                     O
                                                                                                a
                                                                                                o
                                                                                                cc
                                                                                                o_
                                                                       21
                                                                 24   27   30   33
                                                                                           J  0
                                     INCUBATION  TIME  (DAYS)
Figure 80.
Patterns of anaerobic reductive dehalogenation of halobenzoates by sediment microorganisms: (a) degradation
of 2-bromobenzoate in fresh  sediment microcosms; (b) degradation of 3-iodobenzoate by sediment micro-
organism previously acclimated to 3-iodobenzoate degradation; (c) degradation of 4- amino-3,5-dichlorobenzoate
by fresh sediment and the accumulation of the monodehalogenated endproduct (Horowitz, et al., 1983).
amendment are metabolized without an additional
lag period.

Examining and  understanding the factors which
influence the lag period can be difficult, but may
ultimately lead  to  biorestoration scenarios with
controlled,  reduced, or even eliminated adaptation
times. The requirement for the growth of the
requisite microorganism most often is touted as the
operational reason for lag periods. While undoubt-
edly true, the example above illustrates that there
may be other reasons for the delay in biodegradation.

The  structure of the  chemical itself is known to
influence the rate  of biodegradation. Figure 81
compares the adaptation period for  a variety of
halobenzoates in methanogenic sediment micro-
cosms. Note that all the substrates were added at the
same starting concentration and that they all
possessed halogen substitutions at one or both meta
(3 or 5) positions. When degraded, all of the
substrates  were  metabolized via reductive  dehalo-
genation reactions. The various substrates were
degraded in a specific order; that is, 3-bromobenzoate
(3-Br-Bz) degraded before  3-iodobenzoate  (3-I-Bz)
which in turn degraded faster than 3,5-dichloro-
benzoate (3,5-diCl-Bz), etc. Therefore,  chemical
structure very definitely influences the length of the
lag period, and, perhaps more  significantly,  the
length of the lag periods are relatively reproducible.
The insert table in Figure 81 illustrates that a repeat
experiment not only gives the same relative order of
degradation, but also approximately the same length
                                        of the lag period. The lag period does not appear to
                                        correlate with whether the substrate is mineralized.
                                        This  result implies that a specific physiological or
                                        chemical basis exists  for the  characteristic lag
                                        periods.

                                        Other experiments show that  concentration  can
                                        markedly influence the length of the lag period.
                                        Figures  82  and 83 compare  the reductive dehalo-
                                        genation of several halobenzoates at various concen-
                                        trations  and illustrate  a number of characteristic
                                        patterns. In the case of the anaerobic biodegradation
                                        of both 3,5-dichloro-  and 3-chlorobenzoate at sub-
                                        strate concentrations ranging from  20-800  uM
                                        (Figure  82), the dichloro-substrate  exhibits  a
                                        characteristic lag time prior to rapid biodegradation
                                        regardless  of the substrate concentration  range.
                                        However, the length of the lag period associated with
                                        the lower concentrations of monochlorobenzoate is
                                        much shorter than  those observed with higher
                                        substrate concentrations. This perhaps is not sur-
                                        prising since the benzoates are known bacteriostatic
                                        agents and higher concentrations tend to inhibit
                                        microbial activity.

                                        A similar situation also can be observed in the case of
                                        low vs.  high concentrations of 3-fluorobenzoate
                                        (Figure 83). However, the opposite result is observed
                                        when the substrate is changed to 4-amino-3,5-
                                        dichlorobenzoate. With this substrate, concentra-
                                        tions 3:  40  uM were  degraded with a characteristic
                                        lag period.  However, concentrations of 20 uM ex-
                                        hibited lag periods in excess of one year (data not
                                                 93

-------
                               4-NH2-3,5-diCI-BZ

                               3,5-diCI-BZ
                                           ADAPTATION  PERIOD (DAYS)"
                             EXPT    3-Br   3-1  3,5-diCI   4-NH2- 3,5 di Cl  3'CI   3~F
                                                              150-170   >I70

                                                              125-148   >365
                                               28  29-35
                                                        60
                                        TIME  (DAYS)
                                                       200
240
Figure 81.
Length and reproducibility of the adaptation period prior to  the reductive  dehalogenation of  several
halobenzoates at initial concentration of 800 urn (Linkfield, et al., 1989).
completely shown). If the concentrations of this
substrate were only doubled, biodegradation would
proceed in typical fashion and reach levels far below
20 uM. From these examples, one can see that when
substrate concentrations are either too high or too
low, biodegradation activity can be  adversely
affected and increased lag periods could result.
Potential Biostimulation Approaches
Many other factors besides  the necessity for
microbial growth, chemical structure, and substrate
concentration are known to influence the length  of
the lag period. These factors include the need  to
deplete competing substrates and exchange genetic
information and the lack of required nutrients. An
important issue is whether biodegradation can be
stimulated so that the lag or adaptation period can
be overcome as quickly as possible. This can happen,
provided that the factors controlling the lag and
adaptation periods are understood. Although some
biostimulation approaches  were discussed earlier,
three additional approaches will be considered: cross
adaptation; analog enrichment; and biomass enrich-
ment.
                                       The principle of cross  adaptation involves the
                                       addition of a readily degradable substrate to help
                                       bring about the rapid biotransformation of more
                                       recalcitrant molecules. An example of this phenom-
                                       enon is illustrated in  Table 12 and the study
                                       described below.

                                       Anoxic sediment microcosms were exposed to one of
                                       two haloaromatic substrates, the relatively easily
                                       degradable 3-bromobenzoate and the more recalci-
                                       trant 4-amino-3,5-dichlorobenzoate. As  indicated  in
                                       the table,  the former substrate would  start to de-
                                       grade in as little as a few days, whereas the latter
                                       substrate normally took about eight weeks. In both
                                       cases, subsequent additions of the same substrates to
                                       the microcosms were degraded without an observ-
                                       able lag period. Complete degradation of  4-amino-
                                       3,5-dichlorobenzoate reamendments took two  to
                                       three weeks while 3-bromobenzoate additions took
                                       less than one week.

                                       If other halobenzoates were added  to  adapted
                                       sediment instead of additional parent substrate, a
                                       variety of responses  were observed.  When the
                                       sediment microflora was adapted to the  degradation
                                       of a relatively labile substrate (3-bromobenzoate), it
                                               94

-------
                                  800 pM 3-GI-BZ
                                     3,5-diCl-BZ

                           400uM 3,5-diCl-BZ
                                   80       120       160       200

                                          TIME (DAYS)
                                                                 240
Figure 82.
Effect of substrate concentration on the length of the adaptation period prior to the reductive dehalogenation of
3,5-dichloro- and 3-chlorobenzoate (Linkfield, et al., 1989)
also was cross-adapted and capable of rapid
degradation of more recalcitrant materials like 4-
amino-3,5-dichlorobenzoate and 3,5-dichloroben-
zoate. However,  this technique did not result in a
significant improvement in the degradation of other
substrates like 3-iodobenzoate or 3-chlorobenzoate.
Sediment microflora capable of degrading either of
the two  starting substrates exhibited the  same
apparent substrate specificity, while similar experi-
ments with 3-iodobenzoate-adapted organisms could
only degrade subsequent additions  of the parent
substrate.

A technique that may be related to cross adaptation
is referred to as analog enrichment and is illustrated
by the experiments of You and Bartha (1982). Their
objective was to stimulate the mineralization of 14C-
labelled  dichloroaniline by soil microorganisms.
These investigators added aniline as a structural
analog of the halogenated pollutant to soil experi-
ments. They found that the amount of mineralization
of dichloroaniline (determined as the  amount of
14COa) was proportional to the amount of aniline
added as a substrate analog to soil (Figure 84). Pre-
sumably, the dichloroaniline was not a particularly
good inducer of  the requisite enzymatic machinery
among the microbial communities in soil. The addi-
tion of the analog may have derepressed the enzymes
                                        Table 12.   Cross-adaptation of Anaerobic Microorganisms
                                                  to the Reductive Dehalogenation of a Variety of
                                                  Halobenzoates (Adapted from Horowitz, et at.,
                                                  1983)
                                                                    Time (wk) for Complete
                                                                    Degradation in Sediment
Substrate Tested
for Cross-
Adaptation
rT)

O
A
OL
Adaptation
Time
(wk)
3-8
0.5-4
2-3
2-3
32-40
Adapted to:
d j&
2-3 2-3
<1 <1
2-3 . 2-3
<1 <1
Lag Lag
                                        responsible for aniline biotransformation and these
                                        enzymes also might have recognized the halogenated
                                        aniline as a suitable substrate. This  study also
                                        showed an increase in the microbial  mineralization
                                        of humus-bound dichloroaniline as a function of
                                        analog enrichment  with aniline;  humus-bound
                                        residues generally are considered to  be much more
                                        recalcitrant than unbound residues.

                                        Finally, the results of Wilson and Wilson (1985) may
                                        illustrate  an enrichment of desirable microorgan-
                                                 95

-------
                     40OpM 4-NH2-3,5-diCI-BZ
                               160
                                     20O   240
                       TIME (DAYS)
 Figure 83.   Differences in the adaptation period as a
            function of substrate concentration for two
            halobenzoates at 20  and  400  pm initial
            concentrations (Linkfield, et al., 1989).
                                        75
                        Days
Figure 84.   Effect of aniline on the mineralization of 3,4-
           dichloroaniline in  soil: (a) 1.8 mg of aniline
           added per gram; (b) 0.4 mg of aniline added
           per gram; (c) no aniline added; (d) HgCI2-
           poisoned control  (adapted  from You  and
           Bartha, 1982).
isms in a complex microcosm. These authors found
that trichloroethylene (TCE) could be removed  from
soil microcosms that were treated with a combina-
tion of methane and air. However, the TCE was not
removed (other than due to abiotic loss) in soil
microcosms that were not so treated.  When the
methane- and air-amended microcosm was treated
with a bactericidal substance, the removal of  TCE
stopped. The authors speculated that the micro-
 organisms responsible for the biotransformation of
 TCE were a unique group collectively called the
 methane-oxidizing bacteria. These organisms pro-
 duce a powerful oxygen-requiring enzyme called a
 methane monooxygenase that is responsible for the
 initial bioconversion of methane. However,  the
 enzyme has a broad substrate specificity and also can
 oxidize other substrates including TCE. Other types
 of bacteria not involved in methane metabolism can
 act similarly (Nelson, et al., 1987;  Nelson, et al
 1986).

 Presumably, the addition of methane and air to the
 soil columns  stimulated the proliferation of one or
 several populations of methane-oxidizing  micro-
 organisms. While normally present in soil,  the
 relative numbers of these organisms represented a
 greater proportion of the total microbial community
 receiving the methane and air treatment. While
 oxidizing methane, the organisms also cometabolized
 TCE. That is, the organisms growing on a particular
 substrate  gratuitously oxidized  a second substrate
 which they were unable  to use as a source of carbon
 and energy for microbial growth. The cometabolized
 substrate is not usually assimilated by the first
 organism, but the oxidation products are then avail-
 able for other organisms. This type of microbial
 interaction forms the basis of many different types of
 commensalistic relationships between microbial
 populations. In fact, many different substrates  are
 known to  be cometabolized, ranging from simple
 short chain aliphatic hydrocarbons to complex halo-
 genated pesticides (Horvath, 1972).
 Bioremediation with Microbial Inoculants
 The use  of specialized microbial  cultures for
 bioremediation of contaminated aquifers is contro-
 versial (these cultures may or may not involve the
 use of genetically engineered organisms). Conclusive
 proof that the use of such  inoculants constitutes a
 wise management practice has not been shown. Lee,
 et al. (1988) wrote that the role of inoculants in
 ground water cleanup efforts generally can not be
 conclusively determined because adequate controls
 many times are not employed in most experimental
 designs. Fundamentally, questions concerning the
 addition of microorganisms to  aquifers  are  not
 different from those surrounding the use of microbial
 inoculants in soil to elicit some desirable response. In
 the late 1970s, manufacturers captured the imag-
 ination of the popular press with claims for such
 products. These ranged from  an assurance for
 increased crop yields to the ability of such products to
 make "depleted soils come  alive." Generally, these
inoculants or microbial fertilizers have failed to live
up to their claims, which should not be surprising to
microbial  ecologists. An excellent consideration of
                                                96

-------
the relationships between microbial inoculants and
microbial ecological principles is  given by Miller
(1979).
                                             ";' 4#v
Scientific literature shows examples of microbial
inoculation attempts in soil. The early work of
Katznelson (1940a;  1940b)  can be considered
pioneering. Table 13  summarizes  some of his
experiments on the survival of several bacteria,
actinomycetes, and fungi upon reinoculation of these
organisms into manured or manured and limed soil.
In general, these organisms (as well as others) died
back very rapidly upon inoculation.

A spore-forming Bacillus  species was apparently
able to maintain its numbers under some conditions.
Eventually, none of the  organisms could be detected
in numbers significantly above their baseline levels
in  soil.  Similarly,  the review of Miller (1979)
summarizes the findings of Van Donsel, et al. (1967)
on the  survivability of fecal coliforms and fecal
streptococci introduced to soil at various times of the
year (Figure 85).  In both winter and summer, the
fecal bacterial numbers were reduced  by  several
orders  of magnitude.  However, the  decreased
temperatures of winter  allowed the fecal organisms
to  survive for a longer  time before  they were
eventually eliminated.

Implicit in Figure  85 is the role of abiotic factors like
temperature on the mechanisms for elimination of
non-indigenous or foreign microorganisms inocu-
lated in soil. In fact, a variety of abiotic factors may
act alone or collectively to inhibit the survival of
inoculant microorganisms. These  factors are those
known  to influence microorganisms in general and
include  pH, temperature,  salinity, water content,
and osmotic or hydrostatic pressure among others.

In addition to abiotic factors, biotic mechanisms also
may be responsible for the demise of inoculants in
                                                        100
     10-
 en

 "c
 CD
 U
 ^
 ID
 0_
0.1-
    0.01-
  0.001-
Figure 85.
             T
             10
              T	1	T
              2O   30   40
                Time (days)
       Survival of fecal  coliforms  (FC) and fecal
       streptococci (FS)  in soil during summer and
       winter (adapted from Miller, 1979)
70
soil. This is illustrated by the failure of Rhizobium
japonicum cells inoculated in soil (Miller,  1979).
Figure 86A  shows that R. japonicum strain 123
survives when inoculated in sterile soil but declines
rapidly in non-sterile soil. The die-back of strain 123
in non-sterile soil was accompanied by the  simul-
taneous  population increase of a  lytic agent,
presumably a bacteriophage (Miller, 1979) (Figure
86B). Other biotic factors such as the  production of
microbially produced toxins or antibiotics, predatory
eucaryotes, parasitic procaryotes, lytic enzymes, etc.,
can influence the success of inoculation efforts
(Miller, 1979).

It is likely that the microbial communities existing
in aquifers represent a climax ecological community.
The organisms  are found there  because they sur-
vived  an extensive period of natural selection and
are best able to occupy the available niche. That is,
  Table 13.   The Survival of Microorganisms Inoculated into Soil
                                       Manured Soil
                    Manured and Limed Soil
                                           45
                                                         Incubation (Days)

                                                      100          0
                            45
  *reinoculated
                                         100
Organism
Penicillium sp.
Actinomycetes cellulosae
Bacillus cereus
Pseudomonas fluorescen
Azotobacter chroococcum
Numbers per Gram Dry Soil x 105
24.7
8.4
23.2
142.8
200.
7.7
0.1
57.4
0.
0, 300"
7.1
0.
49.3
0.
0.
33.9
7.6 ''
86.9
.175.
360. •
2.7
0.04
8.6
1.1
120. ,
2.2
0.
12.3
0.
0.
                                                  97

-------
                 10   2O    30   40
                     Time (days)
                               Phage
50
      _   0   2  4   6   8   10   12  14
                    Incubation (days)


 Figure 86.   (a) The survival of Rhizobium japonicum
            strain 123 in sterile and nonsterile soil;  (b)
            The relationship between the dieback of
            strain 123 and the increase in the population
            of  a  bacteriophage (adapted from Miller,
            1979).
 the indigenous organisms are best able to assume the
 functions or "occupations" (Miller,  1979) of organ-
 isms  in that habitat. Gause's ecological  principle
 states that only one species can occupy a specific
 niche in a habitat (Gause,  1934). Once the available
 niche^is filled, climax ecological communities tend to
 exhibit the property of homeostasis. The community
 tends to be both  quantitatively and  qualitatively
 stable when subjected to moderate levels of biotic or
 abiotic stresses.

 Little is known about  the mechanisms of homeo-
 stasis in aquifers. However, it seems certain that
 multiple mechanisms, both biotic and abiotic, will
 serve to maintain homeostasis. The inoculation of
 foreign  microorganisms can be viewed as biotic
 stress and the mechanisms responsible for  the
elimination of these organisms can be perceived as
part of the homeostatic controls. When viewed in this
context, it is unreasonable to expect competitive
success from an inoculant  that was likely grown to
 high numbers in the laboratory and subsequently
 forced to compete in situ with indigenous organisms
 naturally selected for their ability  to survive the
 adverse conditions of nature.

 This view on the potential for success of microbial
 inoculants in the terrestrial subsurface is pessimistic
 and  largely  based on  ecological  principles.
 Supporting this view though are the problems (not
 dealt with here) associated with the  transport of
 microorganisms in aquifers and the acceptability of
 this practice to  the regulatory community. Still,
 inoculants may occupy a significant role in pollution
 mitigation scenarios. For example, homeostatic
 control mechanisms of complex environments can be
 overwhelmed  and frequent inoculation can take
 place in order to achieve some desired result. Thus, it
 may be possible  to superimpose alternate environ-
 mental conditions on polluted aquifers with the aim
 of selectively favoring a desirable inoculant.  How-
 ever, inoculants  will  prove most useful  when the
 contaminant of interest is exotic and difficult to
degrade by the indigenous microflora. Further, given
 the difficulties in transporting bacteria in the
subsurface, inoculants might  be most  useful in
above-ground and contained treatment processes.
             Conclusion
             The astonishing metabolic versatility of micro-
             organisms has fueled a great deal of excitement
             among regulators, researchers, and business people
             regarding economical methods for the restoration of
             contaminated environments  such as aquifers.
             However, as with any technology,  bioremediation
             has both promises and limitations. It is critical to the
             future development of this  technology that  the
             practitioners, clients, and regulators recognize  the
             problems and promises of bioremediation; only with
             this recognition can biorestoration be properly con-
             sidered  as another part of the pollution mitigation
             arsenal.
             References
             Bayly, R.  C.  and  M. G.  Barbour, 1984.  "The
               Degradation of Aromatic Compounds by the Meta
               and Gentisate Pathways." In: Microbial Degrada-
               tion of Organic Compounds, D. T. Gibson, Editor.
               Marcel Dekker, Inc., New York, pp. 253-294.
             Brown, J. F. Jr., et al., 1987.  "Polychlorinated
               Biphenyl  Dechlorination in Aquatic Sediments."
               Science, Vol. 236, pp. 709-712.
             Brown, J. R. Jr., et al., 1984. "PCB Transformations
               in Upper Hudson Sediments." Northeast Environ
               Sci.,Vol. 3, pp. 167-169.
                                                98

-------
Gause, G. F., 1934. The Struggle for Existence.
  Williams and Wilkins, Baltimore, MD.
Gibson, D. T. and V. Subramanian, 1984. "Microbial
  Degradation  of  Aromatic Hydrocarbons."  In:
  Microbial Degradation of Organic Compounds, D.
  T. Gibson, Editor. Marcel Dekker, Inc., New York,
  pp. 181-252.
Horowitz, A., J. M. Suflita, and J. M. Tiedje, 1983.
  "Reductive Dehalogenations of Halobenzoates by
  Anaerobic Lake Sediment Microorganisms." Appl.
  Environ. Microbiol., Vol. 45, pp. 1459-1465.
Horvath, R. S., 1972. "Microbial Co-metabolism  and
  the Degradation of Organic Compounds in
  Nature."Bacterial. Rev., Vol. 36, pp. 146-155.
Katznelson, H., 1940a. "Survival of Azotobacter in
  Soil." Soil Sci., Vol. 49, pp. 21-35.
Katznelson, H., 1940b. "Survival of Microorganisms
  Introduced into Soil."  Soil Sci., Vol. 49, pp. 283-
  293.
Lee, M. D.,  et al.,  1988.  "Biorestoration of Aquifers
  Contaminated with Organic Compounds." CRC
  Crit. Rev. Environ. Control., Vol. 18, pp. 29-89.
Linkfield, T. G., Suflita, J. M., and J. M. Tiedje, 1989.
  "Characterization of the Acclimation Period Prior
  to the Anaerobic Biodegradation of Haloaromatic
  Compounds." Appl. Environ. Microbiol., (in press).
Miller, R. H., 1979. "Ecological Factors Which Influ-
  ence the Success  of Microbial  Fertilizers or
  Activators." Dev. Ind.  Microbiol., Vol. 20, pp. 335-
  342.
Nelson, M.  J. K.,  et al., 1987. "Biodegradation of
  Trichloroethylene and Involvement of an Aromatic
  Biodegradative Pathway." Appl. Environ.  Micro-
  biol., Vol. 53, pp. 949-954.
Nelson, M. J. K., et al., 1986. "Aerobic Metabolism of
  Trichloroethylene by a Bacterial  Isolate." Appl.
  Environ. Microbiol., Vol. 52, pp. 383-384.
Perry, J. J., 1984. "Microbial Metabolism of Cyclic
  Alkanes." In: Petroleum Microbiology, R. M. Atlas,
  Editor, Macmillan Publishing Co., New York, pp.
  61-97.
Quensen, J. F. Ill, J. M. Tiedje, and S. A. Boyd, 1988.
  "Reductive Dechlorination of Polychlorinated
  Biphenyls by Anaerobic Microorganisms from
  Sediments." Science, Vol. 242, pp. 752-754.
Singer, M. E. and W. R. Finnerty, 1984. "Microbial
  Metabolism of Straight-chain and Branched
  Alkanes." In: Petroleum Microbiology, R. M. Atlas,
  Editor. Macmillan Publishing Co., New York, pp.
  1-59.
Trudgill, P. W., 1984. "Microbial Degradation of the
  Alicyclic Ring: Structural Relationships and Meta-
  bolic Pathways." In:  Microbiol. Degradation of
  Organic Compounds, D. T. Gibson, Editor. Marcel
  Dekker, Inc., New York, pp.131-180.
Van Donsel, F. J., E.  E. Geldreich, and N. A. Clarke,
  1967. "Seasonal Variations in Survival of Indicator
  Bacteria in Soil and Their Contribution to Storm
  Water  Pollution." Appl. Microbiol.,  Vol. 15, pp.
  1362-1370.
Wilson, J. T. and B. H. Wilson, 1985. "Biotrans-
   formation of Trichloroethylene  in  Soil."  Appl.
   Environ. Microbiol., Vol. 49, pp. 242-243.
You, I.-S. and R.  Bartha, 1982. "Stimulation of 3,4-
  dichloroaniline  Mineralization by Aniline." Appl.
  Environ. Microbiol., Vol. 44, pp. 678-681.
                                                 99

-------

-------
                                         CHAPTER 9
     MODELING SUBSURFACE CONTAMINANT TRANSPORT AND FATE
                                          Joseph F. Keely
Introduction
When scientists and engineers attempt to simulate
the effects of natural phenomena, they are engaging
in modeling. Models are simplified representations of
real-world processes and events, and their creation
and use require many observation-based judgments.
The key  theme that  drives and focuses the
development and application of models is the faithful
simulation of the specific  natural processes. Such
simulation must be demonstrated under a variety of
defined conditions  and must incorporate known
scientific facts before a model can be considered
reliable. Many forms  of models exist, each having
specific advantages and disadvantages.

Physical models, such as sand-filled tanks  used to
simulate aquifers (Figures 87 and 88) and laboratory
columns used to study the relative motion of various
contaminants flowing through aquifer materials
(Figure 89), provide an element of reality that is
enlightening  and  satisfying from an  intuitive
viewpoint. The main disadvantages of physical
models are the extreme effort and time required to
generate  a meaningful amount  of data.  Other
difficulties  relate to  the  care  required  to obtain
samples of subsurface material for  the construction
of these models, without significantly disturbing the
natural condition of the samples.

Analog models also are physically  based, but their
operating principle is  one of similarity, not true-life
representation. A typical example is the  electric
analog model (Figures 90 and 91), where  capacitors
and resistors are able to closely replicate the effects
of the rate of water release from storage in aquifers.
As  is the case with other physically based models,
Figure 87.   Large "sand tank" physical aquifer model.
           (The model is constructed of glass walls and
           external metal braces. A thick layer of silty
           loam overlies a layer of fine sand, which, in
           turn, overlies a layer of  clay. Scores of
           stainless-steel  piezometers penetrate the
           three layers.)
 Figure  88.   Close-up view of piezometers in a  large
            "sand tank" physical aquifer model. (Shallow,
            intermediate, and deep  piezometers are
            wired together in bundles. Wax sheets are
            molded  over the piezometer bundles to
            protect them between samplings.)
                                                101

-------
 data generation is slow and little flexibility exists for
 experimental design changes.
 Figure 89.   Laboratory  column  housed  in  constant-
             temperature environmental chamber.  (Con-
             taminated solutions are injected into column
             through  inlet tubing in top,  by  action  of
             hydraulic  press in foreground. Samples  of
             the advancing front  are withdrawn through
             ports visible on right-hand side and bottom
             of column.)

Figure  90.   Electric analog aquifer model constructed by
            Illinois State Water Survey. (The regular array
            of resistors and the visible electric  "pump"
            are hard-wired into a board papered with the
            appropriate geologic map.)
Mathematical models are non-physical and rely on
the quantification of relationships between specific
parameters and variables to simulate the effects of
natural processes (Figure 92). Because of this,
mathematical  models are abstract and  typically
provide little in the way of an intuitive link to real-
world situations.  Despite this, mathematical models
can generate powerful insights into the functional
dependencies between causes and effects in the real
world.  Large amounts of data can be generated
quickly, and experimental modifications made with
minimal effort,  making it possible for  many
situations to  be studied in great detail for a given
problem.
     1. NEAR - SOURCE ZONE
     2. INTERMEDIATE ZONE
     3, NEAR • W6UF16I.O ZONE
                                                              1. NEAR-SOURCE ZONE
                                                              Z-tNTEHMEOIATEZOfJE
                                                              3. NEAR • WEU-FIELO ZONE
Figure 91.   Control panel for electric analog model
            shown in Figure 90.
Figure 92.   Typical ground-water contamination scenario
            and a possible contaminant transport model
            grid  design for its simulation. (Values  for
            natural  process  parameters  would  be
            specified  at  each node of the grid  in
            performing simulations. The grid density is
            greatest  at the  source  and at potential
            location.)
                                                    102

-------
Categories of Models
Appropriate models  do not exist yet for many
subsurface  contamination problems because a
number of natural processes have yet to be fully
understood. This is especially true for transport and
fate evaluations, where chemical and biological
processes  are very important  but are still poorly
defined. Although great advances are being made in
understanding the behavior of individual contam-
inants, studies of the interactions between contam-
inants are still in their infancy. Also, the current
understanding of physical processes, such as multi-
phase transport, transport through fractured rock,
and transport through karst aquifers,  lags behind
needed knowledge. Moreover, certain well-under-
stood phenomena, such as  the effects of partially
penetrating wells in unconfined (water-table)
aquifers under  varying pumping  rates, pose  unre-
solved difficulties for mathematical formulations.

A variety of schemes exist for  categorizing the
technical  underpinnings and capabilities of mathe-
matical models, but the following categorization can
be useful  (Bachmat, et al., 1978; van der Heijde, et
al., 1985):    -.

•   Parameter identification models
•   Prediction models
•   Resource management models
•   Data manipulation codes

Parameter identification models most often are used
to estimate  the aquifer coefficients that determine
fluid flow and contaminant transport characteristics,
for example, annual recharge (Puri, 1984), coef-
ficients of permeability and storage (Shelton, 1982;
Khan, 1986a and b), and dispersivity (Guven, et al.,
1984; Strecker and Chu, 1986).  Prediction models are
the most  numerous kind of model because they are
the primary tools used for  testing hypotheses
(Andersen, et al., 1984; Mercer  and Faust, 1981;
Krabbenhoft and Anderson, 1986).

Resource management models are combinations of
predictive models, constraining functions (e.g., total
pumpage allowed), and optimization routines for
objective  functions (e.g., scheduling wellfield opera-
tions for minimum cost or minimum drawdown/
pumping lift). Very few of these  models are  devel-
oped and supported enough to be considered practi-
cally useful and  there does  not appear  to be  a
significant drive to improve this  situation (van der
Heijde, 1984a and b; van der Heijde, et al., 1985).

Data  manipulation codes also received  little
attention until only recently.  These codes are now
becoming increasingly popular because they simplify
data entry (e.g., preprocessors) to other kinds of
models and facilitate the production of graphic dis-
plays (e.g., postprocessors) of model outputs (van der
Heijde and Srinivasan, 1983;  Srinivasan, 1984;
Moses and Herman, 1986). Other software packages
are available for routine and advanced statistics,
specialized  graphics, and database  management
needs (Brown, 1986).


Quality Control
Quality control measures are  greatly needed for
modeling the transport and fate of subsurface con-
taminants,  particularly in the use  of numerical
models.  Huyakorn, et al. (1984) suggested  three
levels of quality control:

1.  Validation of the mathematical basis of a model
    by  comparing its output with known analytical
    solutions to specific problems.
2.  Verification of the applicability  of a model to
    various  problem categories  by successful  simu-
    lation of observed field data.
3.  Benchmarking the efficiency of a  model in
    solving  problems by comparison  with the per-
    formance of other models.

These levels of quality control address the soundness
and utility of the model  alone but do not  treat
questions of its application to  a  specific problem.
Hence, at least three additional levels of quality
control appear justified:

4.  Critical review of the problem conceptualization
    to  ensure that the modeling effort considers all
    physical, chemical, and biological processes that
    may affect the problem.
5.  Evaluation of the specifics of the model's applica-
    tion, e.g., appropriateness of the boundary condi-
    tions, grid design, time steps, etc.
6.  Appraisal of the match between the  mathe-
    matical sophistication of the model and the
    temporal and spatial resolution of the data.
Validation of the mathematical framework of  a
numerical model  is deceptively simple. The  usual
approach for ground-water flow models involves  a
comparison of drawdowns predicted by the  Theis
analytical solution to those obtained by  using the
model. The deceptive part is that the Theis solution
can treat only simplified situations as compared with
the scope of situations addressable by the numerical
model. In other words,  analytical solutions cannot
test most of the capabilities of numerical models in a
meaningful way; this is particularly true in simula-
ting complex aquifer boundaries  and  irregular
chemical distributions.
                                                103

-------
 Field verification of a numerical model consists of
 two steps, calibrating the model  using one set of
 historical records (e.g.,  pumping rates and water
 levels from a certain year), and then, attempting to
 predict a subsequent set of observations. In the
 calibration phase, the aquifer coefficients and other
 model parameters are adjusted to achieve the best
 match between model outputs  and known  data. In
 the predictive  phase, no adjustments are made
 except for actual changes in pumping rates, etc.
 Presuming that the aquifer coefficients and other
 parameters are known with  sufficient accuracy,  a
 mismatch means  that  the model either is not
 correctly formulated or does not treat all of the
 important phenomena affecting the actual field
 situation, such as leakage between two aquifers.

 Field verification usually leads to additional data-
 gathering efforts because existing data for the
 calibration procedure often are insufficient to pro-
 vide unique estimates of key parameters. This means
 that a black box solution may be obtained,  which is
 valid only for the observation period  used in the
 calibration. For this reason, the blind prediction
 phase is an essential check on the uniqueness of the
 parameter values used in the model. In this regard,
 field verification of models  using datasets from
 controlled research experiments may be  more
 practical to achieve  than with  the data generated
 during a Superfund site investigation.

 Benchmarking routines are available that compare
 the efficiency of different  models in solving the same
 problem (Ross, et al., 1982; Huyakorn, et al., 1984);
 however, more  must be done in this area. For
 example, common observations indicated that finite
 element models (FEMs) have an inherent advantage
 over finite difference models (FDMs) in terms of
 ability to incorporate irregular boundaries  (Mercer
 and Faust, 1981) (the number of points (nodes) used
 by FEMs is considerably less due to the flexible nodal
 spacings allowed). Benchmarking routines, however,
 show that the  large amount of computer time
 required to evaluate FEM nodes reduces  the cost
 advantage for simulations of comparable accuracy.


 A Field Sample
 Field experience using special geotechnical  methods
 and state-of-the-art research findings was gained at
 the 20-acre Chem-Dyne solvent reprocessing site in
 Hamilton, Ohio (Figure 93), where over 250 chemical
 waste  generators disposed of drummed  or  bulk
 wastes during its operational lifetime  (1974-1980).
 Poor waste handling practices, such as purposeful
 pn-site spillage of a wide variety of industrial chem-
 icals and solvents, direct  discharge of liquid wastes
 to a stormwater drain beneath the site, and mixing of
incompatible wastes, occurred routinely at Chem-
 Dyne. These practices caused extensive soil  and
 ground-water contamination, massive fish kills in
 the Great Miami River, and major on-site fires and
 explosions.

 The stockpiling of liquid and solid wastes resulted in
 a long-term threat to the environment. More than
 50,000 drums of hazardous waste  were stored at the
 site at its peak of operations (CH2M-Hill, 1984a).
 The drums were stacked improperly, in tiers five and
 six drums high, causing the drums at the bottom to
 buckle and corrode. After the remedial investigation
 began in the spring of 1982, more than 20,000 drums
 still remained; at least 8,500 of these  were so badly
 corroded that they could not be identified. A number
 of bulk chemical storage tanks also were abandoned
 on site. Visual observations indicated that  raw
 chemical  salts and oils had been poured out on the
 sand-and-gravel ground surface.
 The FIT Investigation
 The seriousness of the ground-water contamination
 problem at Chem-Dyne became evident during the
 initial site survey (1980-1981), which included the
 construction and sampling of over twenty shallow
 monitoring wells (Ecology and Environment, 1982).
 The initial  survey indicated that the contaminant
 problem was much  more limited than was later
 shown to be the case  (Roy F. Weston Inc.,  1983;
 CH2M-Hill, 1984a). A  good portion of the improve-
 ment in delineating the plume was brought about by
 a better understanding of the natural processes
 controlling transport of contaminants at the site.

 The initial site survey indicated that ground water
 flowed to the west of the site (toward the Great
 Miami River), but that a shallow trough paralleled
 the river as a result of weak and temporary stream
 influences. The study concluded that contaminants
 already in the aquifer would be discharged into the
 river and would not need to be removed (Ecology and
 Environment, 1982). That study also concluded that
 the source  was limited to highly contaminated
 surface soils, and that removal  of the uppermost
 three feet of the soil would essentially eliminate the
 source of contaminants.

 That conclusion, however, was based on faulty soil
 sampling procedures. The soil samples taken were
 not preserved in air-tight containers, so most of the
 volatile organic chemicals leaked out prior to
 analysis.  The uppermost soil samples probably
 showed high volatile organic levels because of the co-
occurrence of viscous oils and other organic chem-
icals that  may have served to entrap the volatiles.
The more viscous and highly retarded chemicals did
not migrate far enough into the vertical profile to
                                               104

-------

                                                           CHEM-DYNE SUPERFUND SITE

                                                                       Hamilton, Ohio

                                                                        Location Map
                                       LEGEND
                                  • monitoring well locations
                                MW1 monitoring well Identity
                                • • • site boundary
Figure 93.   Monitoring well location map for the Chem-Dyne site.
exert a similar influence on samples collected at
depths greater than a few feet.
Additional Site Investigations
Subsequent studies of the  site corrected these
misinterpretations by producing data from proper
soil samplings and incorporating much more detailed
characterizations of the fluvral sediments and the
natural flow system. In those studies, vertical profile
characterizations were obtained  from each  new
borehole drilled by continuous split-spoon samples of
subsurface solids. The split-spoon samples helped to
confirm the general locations of interfingered  clay
lenses and clearly  showed the  high degree of
heterogeneity of the sediments (Figures 94 and 95).
For example, a dense clay lens was found at similar
elevations (at 570 to 580  ft MSL)  along the valley
axis (Figure 94), but was  found only intermittently
perpendicular to the valley axis (Figure 95). This
should be expected by hydrogeologic intuition since,
during flooding, rivers dissect deposits that were laid
down in low energy  periods and rivers undergo
natural channel changes (e.g., meandering) as they
mature.  These phenomena  create  lenticular clay
bodies of very limited extent and  structurally
anisotropic conditions. The major axis of anisotropy
parallels the average downstream direction, which
itself is generally co-linear with the valley axis.

Clusters of vertically separated monitoring wells
were constructed  during the remedial  investiga-
tions/feasibility studies at Chem-Dyne.  While an
extensive network of shallow wells confirmed earlier
indications of general  ground-water flow toward
Great Miami River (Figure 96), the clusters of
vertically separated wells revealed that downward
gradients existed adjacent to the river. Figures 97
and  98 show that  these vertical gradients, which
                                                105

-------
                    CHEM-DYNE GEOLOGIC  CROSS-SECTION
             NNW
    V)
    o
     600


     590


     580


     570


     560


     550


     540


     530
                                                                                    SSE
-J  520
111

    510


    500
m
                                            Water Level Elevation"   ]
                                          Approximately 563 ft. MSL j

 -October 30, 1983
                                                           DATA SOURCE:  CH2M-HIII, 1984a
                  Fill (sandy gravel)
                  Clayey silt, silty clay
                                        Sand
                                        Sandy gravel, gr. sand
                                                      Silty sand
                                                  f-Xi Clayey gravel, glacial till
Figure 94.   NNW-SSE geologic cross-section at Chem-Dyne site.
ranged from a 1- to 3-foot drop over the 20-foot
vertical separation between the bottom of the shal-
low wells and the top of the screens in the deep wells
(or about 0.100), are quite dramatic relative to the
horizontal gradient across the contaminant plume
(which averages about 0.001). This finding indicated
that the migrating plume would not be discharged to
the river, but would flow under the river.

The presence of major industrial wells on the west
bank of Great Miami River provided an explanation
for the observed downward  vertical gradients
(normally, one would have expected the river to be
gaining water from the aquifer at this point in the
basin), and supported the conclusion  that
contaminants could not be discharged to the  river
from the aquifer. The plume would be drawn to
greater depths in the aquifer by the locally severe
downward gradient, but it could not be determined if
the industrial wells would actually  capture the
plume. That determination would require careful
evaluation of the hydrogeologic features beneath the
river (an activity not attempted  because of the
associated costs)  and expectations that planned
                                              remedial actions would stop the plume before
                                              substantial encroachment could occur.
                                              Hydrologic Complications
                                              Unfortunately, the hydraulic interplay of the river
                                              with the aquifer was not well appreciated by the field
                                              crews taking routine measurements of water levels.
                                              Observations late in the final study found  that,
                                              during preparations for  a pump test, river stage
                                              variations cause as much as three feet of water level
                                              change during a single day at wells close to the river.
                                              This effect was virtually  negligible at wells much
                                              closer to the site. This sort of situation  makes it
                                              crucial to obtain water levels at all wells within only
                                              a few hours, otherwise, the sort of confusing water
                                              level maps shown in Figures 99 to 101 may result.
                                              These figures  were  prepared with water-level
                                              elevation data that were  measured over periods of
                                              several days (CH2M-Hill, 1984a).

                                              Investigators decided to use a  major pump test to
                                              estimate  the hydrogeologic characteristics of the
                                               106

-------
    600


    590

O
_J  580


21  570


    560
                    CHEM-DYNE GEOLOGIC  CROSS-SECTION

             WSW                                                                      ENE
    A
    O
    550


    540


    53D
&
>: Clayey gravel, glacial till
 Figure 95.   WSW-ENE geologic cross-section at Chem-Dyne site.
heavily contaminated portion of the aquifer. The
pump test was technically difficult because the
pumping well had to be drilled  on site  due to
potential liability and lack of access elsewhere. The
'drillers were substantially slowed by the need to
wear air-tanks when encountering particularly con-
taminated subsoils that emitted volatile fumes and
presented unacceptable  health risks. Since the
pumped water was expected to.be contaminated, ten
large contemporary holding tanks (100,000  gallons
each) were constructed on site to impound the waters
for testing and possible treatment  before being
discharged to the local sewer system (CHgM-Hill,
1984a).

Although there  had been some resistance  to
conducting the pump test because of its cost, the test
results were very valuable. The water levels in
thirty-six monitoring wells were observed during the
test, providing a very detailed picture of area! trans-
missivity variations (Figures 102). This information
helped explain the unusual configuration of the
plume shown in Figures 103, 104, and 106 (Figure
105 is an updated location map for 1985 data
                                              presented in Figure 106 and some later figures). The
                                              information also was used to guide the design of a
                                              pump-and-treat system. Storage coefficients  also
                                              were estimated and, though the short duration of the
                                              test (14 hours)  did not provide many definitive
                                              estimates, qualitative confirmation of the generally
                                              non-artesian (water-table) nature of the aquifer
                                              beneath the site was clearly confirmed, as were the
                                              increasingly artesian conditions from the west edge
                                              of the site towards the river.
                                              Anisotropic Flow Biases
                                              On-site transmissivity estimates from a trio of wells
                                              (MW-23, MW-26, and MW-29) indicated a 2:1 aniso-
                                              tropic bias  toward the  river as  opposed to
                                              downvalley, whereas nearer the river a second trio of
                                              wells (MW-28, MW-33, and MW-35) yielded esti-
                                              mates for  which the bias appeared to be 10:1
                                              downvalley (CH2M-Hill, 1984a). These trends coin-
                                              cide with the nature of the system; that is, there are
                                              few clay occurrences on site and east of the site from
                                              whence recharge waters flow toward the river,  and
                                               107

-------
                                                                HYDRAULIC  CANAL
                                                                                                   I
                                                           CHEM-DYNE  SUPERFUND SITE

                                                                       Hamilton, Ohio

                                                             April, 1983 Water Level Elevation Contours
                                                                      for Shallow Wells
                                                                       (values in feet)
 Figure 96.   April 1983 water level elevation contours at Chem-Dyne site using shallow well data only, in feet MSL.
significant occurrences of clays at the west fenceline
of the site and adjacent to the river. The latter clays
had lenticular shapes that paralleled the axis of the
river, causing  a strong downvalley bias. These
indications mean that flow would not proceed at an
average velocity perpendicular to the water level
contours that would be west from the northern half of
the site, and south to southwest from the southern
half of the site (Figure 95). Instead, flow would move
westerly first and then southerly as it neared  the
river. By the time areas adjacent to the  river were
reached, the net position would be roughly the same,
but the  path taken to get there would be  strikingly
different. The contaminant masses leaving both the
northern and southern halves of the site would be
concentrated in west-trending tongues.  Moreover,
the future travel paths  with the anisotropic bias
interpretation  would not be the same  as with a
presumption of isotropicity.
The distributions of contaminants observed at Chem-
Dyne seem to support this anisotropic bias  inter-
pretation. The highest concentrations lie  along an
axis that does not appear to be influenced by the
southerly components of flow expressed by the water-
level contours offsite (Figures 103 and 104). While it
is true that the pumpage of major production wells
on  the other side of the river offers a  nominal
explanation for this,  the actual water level contours
contradict  the notion that  the industrial  wells
dominate the entire flow field.  This has  to be
appreciated in terms of capture zones; the industrial
wells will indeed affect all of the ground-water flow-
lines locally,  but may not capture all of them. It is
likely that  the plume is bifurcating near the river,
with one member travelling on flowlines captured by
the industrial wells, and a second member tangen-
tially affected, but eventually released to continue on
down the valley with the rest of the aquifer waters.
                                                108

-------
     570
     560
     550
  <  540
  111
     530
     520
                                  MB-20
MIU-19
 D
         DEO 82  APR 83  JUN 83  OCT 83
                                   HELL SCREENS
         COMPARISON OF WATER LEVEL ELEVATIONS IN
          MONITORING WELLS MW-19 AND MW-2D

 figure 97.  Water levels in a cluster of two  vertically
           separated monitoring wells adjacent to Great
           Miami River and due west of Chem-Dyne site.
    570
     560
     550
  o -.. . ,

 '<  540
  ILI  '
     530
    520
                                  rw-21
                                       MU-22
                                        D
        DEC 82  APR 83  JUN 83  OCT 83   HELL SCREENS

    .  (  COMPARISON OF WATER LEVEL ELEVATIONS IN
          MONITORING WELLS MW-21 AND MW-22

Figure  98.   Water levels in a cluster  of two vertically
 • ..- .    •   separated monitoring wells at the confluence
    •'.•'    of Great Miami River and Ford  Hydraulic
           Canal, due west of north boundary of Chem-
           Dyne site.
Field Evidence for Biotransformations
Finally, the distribution patterns of contaminant
species  that emerged from the investigations at
Chem-Dyne were understood by  considering re-
search results and theories regarding chemical and
microbiological influences. Contaminant distribu-
tion maps derived from samples taken at the end of
the field investigation (October,  1983, only  months
after the last drums of solvents were removed from
the site)  suggested  that  the transformation of
tetrachloroethene (Figure 107) to less halogenated
daughter products such as trichloroethene (Figure
108), dichloroethene (Figure 109),  and vinyl
chloride/monochloroethene  (Figure 110) was occur-
ring.

In such circumstances, one would expect to  see the
progressive  disappearance of tetrachloroethene and
successive increases in the concentrations and extent
of potential  daughter products. This seems to be the
case at  Chem-Dyne, according to the October 1983
data. One might argue that too little vinyl chloride is
observed  (Figure 110) to  show the full series of
degradation expected,  but there are  plausible
reasons why the  distributions might  be  as  shown.
For example, with  a  continuous source  input, the.
concentrations  of tetrachloroethene might be high
enough that there would  be no need for further
biotransformation of daughter products because an
ample food  supply is  available in the parent mate-
rial. Alternatively, the concentrations of tetrachloro-
ethene with continuous source inputs might indeed
be  so high as to  limit biotransformation by  toxic
effects.  Since the relative kinetics of the various
transformations in this sequence still require further
definition, it is impossible to make rigorous conclu-
sions with regard to these possibilities.

But consider the data obtained .during  a chemical
sampling conducted two  years later in preparation
for activation of the pump-and-treat system used to
remediate the plume  (Figures 110 to 113). At least
two years of freedom from surface inputs of solvents
had occurred as well  as  two years of healthy rain-
water flushing the  unsaturated zone of stored
residues. Investigators  found that the daughter
products (Figures 111 to 113) contain much greater
mass than that in  the tetrachloroethene contours
(Figure 110),  and  are spread over  significantly
greater areas.  The increase in  the  vinyl chloride
component  of  the  plume is staggering (compare
Figure 113 to Figure 109). These data highly suggest
active  degradation  of  tetrachloroethene to  its
possible daughter products. Knowledge of this kind
of possible  transformation should be valuable to
those attempting to design and estimate costs for
treatment systems  since  treatment efficiencies vary
with  the  contaminant and its contribution to the
                                                109

-------
                                                                HYDRAULIC  CANAL
                                                                                                   I
                                        LEGEND
                                   • monitoring well location
                                MW1 monitoring well Identity
                                • • • site boundary
                                   x water level elevation contour
          CHEM-DYNE SUPERFUND SITE

                      Hamilton, Ohio

          December, 1982 Water Level Elevation Contours
                    for Shallow Wells
                     (values in feet)
 Figura  99.   December 1982 water level elevation contours at
 overall loading. Vinyl chloride is much more volatile
 and easier to remove than tetrachloroethene.

 The relative rates of movement of these and other
 common solvents like benzene and chloroform at
 Chem-Dyne generally conformed to  predictions
 based  on sorption principles. This is useful in a
 practical sense. The remediation efforts made use of
 these contaminant transport theories in estimating
 the capacity of the treatment system and the length
 of time needed to remove residuals from the aquifer
 solids (CH2M-Hill, 1984b).


 The Role of Mathematical Models
 During the latter stages of negotiations with  the
 Potentially Responsible Parties (PRPs), the State of
Ohio government  contractors prepared mathe-
matical models of the flow system and contaminant
Chem-Dyne site using shallow well data only in feet MSL.
  transport at Chem-Dyne  (GeoTrans, 1984).  These
  models were used to estimate the possible direction
  and rate of migration of the plume in the absence of
  remediation, the  mass of contaminants removed
  during various remedial options, and the effects of
  sorption and dispersion on those estimates. Because
  of the wide range of sorption properties associated
  with the variety of contaminants found in significant
  concentrations, it was necessary to select values of
  retardation constants  that  represented the  likely
  upper and lower limits of sorptive effects. It also was
  necessary to estimate or assume the values of other
  parameters such as dispersion coefficients known to
  affect transport processes.

  Ward and  his co-workers recently published  this
  application of models to the Chem-Dyne site in an
  article  highlighting the development and recom-
  mendation of the Telescopic Mesh Refinement (TMR)
                                                110

-------
                                        LEGEND
                                  • monitoring well location
                                MW1 monitoring well identity
                                • • • site boundary
                                «»••»->• water level elevation contour
        CHEM-DYNE  SUPERFUND  SITE

                    Hamilton, Ohio

         June, 1983 Water Level Elevation Contours
                  for Shallow Wells
                   (values in feet)
Figure 100.    June 1983 water level elevation contours at ChenvOyne site using shallow well data only, in feet MSL.
modeling approach (Ward, et al., 1987). The TMR
approach involves a staged evaluation of the prob-
lem, proceeding from the regional level to the local
level to the site-specific  level (Figure 114). This
approach  makes it possible to assure the appro-
priateness and consistency of boundary conditions
such as recharge rates to the aquifer and interactions
with streams. GeoTrans chose to develop this set of
models using the SWIFT  model, which utilizes the
finite difference method (FDM), that, in turn, utilizes
rectangular grids  (Figure 115). FDM is  the most
mathematically straightforward and easily de-
bugged of the numerical analysis techniques used for
mathematical  modeling (Mercer and  Faust, 1981).
Recognizing the need to account for the influences of
the vertical gradients previously noted to  be severe
adjacent to Great Miami River, GeoTrans created a
quasi-three-dimensional (layered) model at the local
scale (Figure 116).

The results of their modeling efforts included maps of
the potentiometric surfaces represented  by water
level elevations in wells tapping the different model
"layers" (Figures 117 and 118). These maps indicated
that local industrial wells control flow in the deeper
portions  of the aquifer. This site-scale model was
used to investigate the probable effectiveness of the
pump-and-treat remediation proposed by the
Potentially Responsible  Parties for  Chem-Dyne
(Figures  119 and 120). Geo-Trans concluded that the
proposed pumping scheme appeared to  be quite
effective  in the  interior zone of the contaminant
plume, but could not be completely  effective  in
developing an inward  hydraulic gradient at the
                                                 111

-------
                                                                HYDRAULIC  CANAL
                                        LEGEND
                                   • monitoring well location
                                MW1 monitoring well identity
                                • • • site boundary
                                s*-^s water level elevation contour
        CHEM-DYNE  SUPERFUND  SITE

                    Hamilton, Ohio

        October, 1983 Water Level Elevation Contours
                   for Shallow Wells
                    (values in feet)
Figure 101.    October 1983 water level elevation contours at Chem-Dyne site using shallow well, data only, in feet MSL.
periphery of the plume. Their predictions turned out
to be correct; during the first year of operation, the
net withdrawal rate of the remediation wellfield was
increased substantially over the originally proposed
values, and inward hydraulic gradient control was
still not  fully established (Figure 121). GeoTrans
modeling efforts also included predictions of the
extraction well concentrations  versus time (Figure
122), but comparisons with actual performance are
difficult because of the many remediation shut-down
periods (including lost time due to clogging of the air-
stripper by precipitated iron and manganese).

Large uncertainties  were associated with those
modeling efforts due to lack of information about the
actual history of chemical inputs and other impor-
tant data. However, there was agreement  between
the government and PRP technical experts about the
helpfulness of modeling efforts  in assessing the
magnitude of the problem and in determining mini-
mal requirements  for remediation. Consequently,
modeling efforts continue at Chem-Dyne. Data gen-
erated during the remediation phase are being used
to refine models in an ongoing process so that the
effectiveness of the remedial action can be evaluated
properly.                             .
Summary
Models of the transport and fate of contaminants in
the subsurface environment are created by organ-
izing known information and relationships into a
functional representation (e.g., a sand-filled tank, a
circuit board, or an equation). Models may be used to
simulate the response of specific problems to a wide
variety of possible solutions.
                                                112

-------
                              LEGEND
                              • monitoring well locations
                            MW1 monitoring well identity
                            ... site boundary
                            ^ux Transmissivity Isopleth in thousands
                                of square feet per day
        CHEM-DYNE  SUPERFUND SITE

                    Hamilton, Ohio

           Transmissivity Estimate from October
                   1983 Pump Test
               (Deep Well Contours Only)
Figure 102!    Transmissivity estimates obtained  from October  1983 Pump test at Chem-Dyne site, in thousands of
             square feet/day (plotted values of estimates from all wells).

The application of mathematical models is subject to
considerable error in practical situations when
appropriate field determinations of natural process
parameters are lacking. Contrary to popular beliefs,
this source of error is not addressed adequately by
sensitivity analyses or by the  application  of
stochastic techniques for estimating  uncertainty.
Rather, the high degree of hydrogeological, chemical,
and microbiological complexity typically present in
field situations forces the use of site-specific charac-
terization of the influences of various natural
processes by detailed field and laboratory  investi-
gations.

Both the mathematics that describe models and the
parameter inputs to those models must be subjected
to rigorous  quality control procedures. Otherwise,
results from held applications of models are likely to
be qualitatively, as well as quantitatively, incorrect.
Quality control methodologies must focus on the
accuracy of the problem conceptualization and the
representativeness of parameter values, and recog-
nize that accuracy and precision determinations are
insufficient measures of quality.
References
Andersen, P. F., C. R. Faust, and J. W. Mercer, 1984.
  "Analysis of Conceptual  Designs for Remedial
  Measures at Lipari Landfill." Groundwater, Vol.
  22, No. 2.

Bachmat, Y., et al., 1978. "Utilization of Numerical
  Groundwater Models for Water  Resource Manage-
  ment." EPA-600/8-78-012, R.  S. Kerr Environ-
  mental Research Laboratory, U.S. Environmental
  Protection Agency, Ada, OK.
                                                 113

-------
                                                               HYDRAULIC  CANAL
                                                                                	if
                                                                                                1
            Scali
                 3OOfl

                  lOOro
  LEGEND
  • monjtorjng well locations
MW1 monitoring well Identity
. . . site boundary
    Isopleth In parts per billion
       (shallow wells only)
        CHEM-DYNE SUPERFUND SITE

                    Hamilton, Ohio

         TOTAL VOLATILE ORGANIC CHEMICALS
                APRIL 1983 SAMPLING
Figure 103.    April 1983 total VOC concentration contours (ppb) at Chem-Dyne site, using shallow well data only.
Brown, J., 1986. "Environmental Software Review."
  Pollution Engineering, Vol. 18, No. 1.

CH2M Hill, 1984a. "Remedial Investigation Report:
  Chem-Dyne Site, Hamilton, Ohio." Unpublished
  report, Contract No. 68-01-6692 to U.S.  EPA
  Region 5, Chicago, IL.

CH2M Hill, 1984b. "Feasibility Study Report: Chem-
  Dyne Site, Hamilton, Ohio." Unpublished report,
  Contract No.  68-01-6692 to U.S. EPA Region 5,
  Chicago, IL.

Ecology and Environment, 1982. "Field Investigation
  of Uncontrolled Waste Sites  — Ground Water
  Investigation  of Chem-Dyne Sites." Unpublished
  report, Contract No. 68-01-6506 to U.S.  EPA
  Region 5, Chicago, IL.
Faust, C. R., L. R. Silka, and J. W. Mercer, 1981.
  "Computer Modeling and Ground-water Protec-
  tion." Groundwater, Vol. 19, No. 4.

GeoTrans, 1984. "Evaluation  and Analysis of
  Groundwater and Soil Contamination at  the
  Chem-Dyne Site, Hamilton, Ohio." Report to the
  Ohio Environmental Protection Agency, October
  30,1984.

Guven, O., F. J. Molz, and J. G. Melville, 1984. "An
  Analysis of Dispersion  in a Stratified Aquifer."
  Water Resources Research, Vol. 20, pp. 1337-1354.

Huyakorn, P. S., et al., 1984. "Testing and Valida-
  tion of Models for Simulating Solute Transport in
  Ground Water: Development and Testing of Bench-
  mark Techniques." IGWMC Report No. GWMI84-
                                               114

-------
                                  LEGEND
                                  •  monitoring well locations
                                MW1  monitoring well identity
                                - - -  site boundary
                                    Isopleth in parts per billion
                                       (shallow welfs only)
        CHEM-DYNE SUPERFUND SITE

                    Hamilton, Ohio

         TOTAL VOLATILE ORGANIC CHEMICALS
                JUNE 1983 SAMPLING
Figure 104.   June 1983 total VOC concentration contours (ppb) at Chem-Dyne site, using shallow well data only.
  13, International Ground Water Modeling Center,
  Holcolm Research Institute.
Khan, I.  A., 1986a, "Inverse Problem in Ground
  Water:  Model Development." Groundwater, Vol.
  24, No.  1.
Khan, I.  A. 1986b, "Inverse Problem in Ground
  Water:  Model Application." Groundwater, Vol. 24,
  No. 1.
Krabbenhoft, D. P. and M. P. Anderson, 1986. "Use
  of a Numerical Ground-water Flow Model for
  Hypothesis Testing." Groundwater, Vol. 24, No.l.
Mercer, J. W. and C. R. Faust, 1981. Ground-Water
  Modeling. National Water Well  Association,
  Worthington, OH.
Moses, C. O. and J. S. Herman, 1986. "Computer
  Notes — WATIN — A Computer Program for Gen-
  erating Input Files for WATEQF." Groundwater,
  Vol. 24, No. 1.
Puri, S., 1984. "Aquifer Studies Using Flow Simula-
  tions." Groundwater, Vol. 22, No. 5.
Ross,  B., et al., 1982. "Benchmark Problems for
  Repository Siting Models." U.S. NRC Publication
  No.  NUREG/CP-3097, U:S. Nuclear Regulatory
  Commission, Washington, DC.
Roy F. Weston, Inc., 1983. "Preliminary Hydro-
  geologic Investigation and Preliminary Evaluation
  of Remedial  Action Alternatives Feasibility:
  Chem-Dyne Hazardous Materials Recycling Facil-
  ity,  Hamilton, Ohio." Unpublished  Report, Con-
  tract No.  68-03-1613 to  U.S. EPA Region 5,
  Chicago, IL.
Shelton, M. L., 1982. "Ground-water Management in
  Basalts." Groundwater, Vol. 20, No. 1.
                                               115

-------
                                        LEGEND
                                   • monitoring well location
                                 MW1 monitoring well identity
                                 • • - site boundary
       CHEM-DYNE SUPERFUND SITE

                   Hamilton, Ohio

               UP DATED LOCATION MAP
                 (G-WELLS ADDED)
Figure 105.    Updated location map for the Chem-Dyne site.

Srinivasan, P., 1984. "PIG - A Graphic Interactive
  Preprocessor for Ground-water Models." IGWMC
  Report No. GWMI 84-15. International Ground
  Water Modeling Center,  Holcolm  Research
  Institute.
Strecker,  E.  W. and W.  Chu, 1986. "Parameter
  Identification of a Ground-water  Contaminant
  Transport Model." Groundwater, Vol. 24, No. 1.
van der Heyde, P.  K. M., 1984a. "Availability and
  Applicability of Numerical  Models for Ground
  Water Resources Management."  IGWMC Report
  No. GWMI 84-19. International Ground Water
  Modeling Center, Holcolm Research Institute.
van der Heyde, P.K.M., 1984b. "Utilization of Models
  as Analytic Tools for Groundwater Management."
  IGWMC Report No. GWMI 84-18.  International
  Ground Water Modeling Center, Holcolm Research
  Institute.
van der Heijde, P.K.M., et al., 1985. "Groundwater
  Management: The Use of Numerical Models, 2nd
  Edition." AGU Water Resources Monograph No. 5.
  American Geophysical Union, Washington, DC.
van der Heijde, P.  K. M. and P. Srinivasan, 1983.
 "Aspects of the Use of Graphic Techniques in
 Ground-water Modeling."  IGWMC Report  No.
 GWMI 83-11. International  Ground Water Model-
 ing Center, Holcolm Research Institute.
Wagner, B. J.  and  S. M. Gorelick, 1987. "Optimal
  Groundwater Quality Management Under Param-
  eter Uncertainty." Water Resources Research, Vol.
  23, No. 7.
Ward, D. S., et al.,  1987. "Evaluation of a Ground-
  water Corrective Action at the Chem-Dyne Haz-
  ardous Waste Site Using a Telescopic Mesh
  Refinement Modeling Approach. "Water Resources
  Research, Vol. 23, No. 4.
                                              116

-------
                                                                      HYDRAULIC   CAN.AL
   LEGEND
  - • monitoring well locations
 MW1 monitoring well identity
 ... site boundary
O^s^ Isopleth in parts per billion
        (shallow wells only)
                                                                  CHEM-DYNE  SUPERFUND  SITE

                                                                               Hamilton, Ohio

                                                                   TOTAL VOLATILE ORGANIC CHEMICALS

                                                                        DECEMBER 1985 SAMPLING
Figure  106.    December 1985 total VOC concentration contours (ppb) at Chem-Dyne site, using shallow well data only.
                                                      117

-------
                                                                      HYDRAULIC  CANAL
                                                                 CHEM-DYNE  SUPERFUND  SITE

                                                                              Hamilton, Ohio
                                                                            TETRACHLOROETHENE

                                                                           OCTOBER 1983 SAMPLING
  LEGEND
   • monitoring well locations
MW1 monitoring well identity
    site boundary
    Isopleth in parts per billion
       (shallow wells only)
Figure  107.    October 1983 tetrachloroethane concentration contours (ppb) at Chem-Dyne site, using shallow well data
              only.
                                                    118

-------
                                                                     HYDRAULIC  CANAL
                                                                CHEM-DYNE  SUPERFUND SITE

                                                                             Hamilton, Ohio
  LEGEND
.... • monjtormg well locations
MW1 monitoring well identity
- - . site boundary
    Isopleth in parts per billion
       (shallow wells only)
                                                                          TRICHLOROETHENE
                                                                        OCTOBER 1883 SAMPLING
Figure 108.    October 1983 trichloroethane concentration contours (ppb) at Chem-Dyne site, using shallow well data
              only.
                                                   119

-------
                                      LEGEND
                                      • monitoring well locations
                                    MW1 monitoring well identity
                                    ... site boundary
                                      ^ Isopletn In parts per billion
                                           (shallow wells only)
CHEM-DYNE  SUPERFUND  SITE

              Hamilton, Ohio


         trans-DICHLOROETHENE

         OCTOBER 1983 SAMPLING'"
Figure  109.    October 1983 trans-dichloroethene concentration contours (ppb) at Chem-Dyne site, using shallow well
              data only.
                                                     120

-------
                                                                      HYDRAULIC  CANAL
  LEGEND
   • monitoring well locations
MW1 monitoring well identity
. . . site boundary
-^*» Isopleth in parts per billion
       (shallow wells only)
                                                                 CHEM-DYNE SUPERFUND  SITE

                                                                               Hamilton, Ohio

                                                                              VINYL CHLORIDE
                                                                          (MONOCHLOROETHENE)
                                                                          OCTOBER 1983 SAMPLING
Figure  110.    October 1983 vinyt chloride concentration contours (ppb) at Chem-Dyne site, using shallow well data only.
                                                     121

-------
                                   LEGEND
                                   • monitoring well locations
                                 MW1 monitoring well identity
                                 ... site boundary
                                     Isopleth in parts per billion
                                        (shallow wells only)
CHEM-DYNE  SUPERFUND  SITE

             Hamilton, Ohio

          TETRACH LOROETH EN E

        DECEMBER 1985 SAMPLING
Figure  111.    December 1985 tetrachloroethene concentration contours (ppb) at Chem-Dyne  site, using shallow well
              data only.
                                                    122

-------
   I
                                                                                              1
                                            LEGEND
                                          ..,.,* monitoring well locations
                                          MW1 monitoring well Identity
                                          ... site bounda-
                                          --N^ Isopletl '
                                                   CHEM-DYNE SUPERFUND  SITE

                                                                 Hamilton, Ohio

                                                              TRICHLOROETHENE

                                                           DECEMBER 1985 SAMPLING
                               i boundary
                               pleth In parts per billion
                                (shallow wells only)
w   Figure  112.
December 1985 trichloroethene concentration contours (ppb) at Chem-Dyne site, using shallow well data
only.
                                                        123

-------
                                                                      HYDRAULIC  CANAL
                                          LEGEND
                                           • monitoring well locations
                                        MW1 monitoring well identity
                                        • - . site boundary
                                            Isopleth in parts per billion
                                               (shallow wells only)
CHEM-DYNE  SUPERFUND  SITE

             Hamilton, Ohio


        trans-1-2-DICHLOROETHENE

        DECEMBER 1985 SAMPLING

Figure 113.    December 1985 trans-dichloroethene concentration contours (ppb) at Chem-Dyne site, using shallow well
              data only.                                               ,
                                                     124

-------
                                                                     HYDRAULIC   CANAL
                                                                CHEM-DYNE SUPERFUND  SITE

                                                                              Hamilton, Ohio
                                                                             VINYL CHLORIDE
                                                                          (MONOCHLOROETHENE)

                                                                        DECEMBER 1985 SAMPLING
  LEGEND
.... * monitoring well locations
MW1 monitoring well Identity
. . . site boundary
P -••*«"•  soofi.
                                      Isopleth in parts per billion
                                         (shallow well
Figure 114.  ,.December. 1985 yinyl chloride concentration contours (ppb) at Chem-Dyne site, using shallow well data
             only.                                                   •                    ..".-.'
                                                    125

-------
                                              REGIONAL
                                                                                          200 FEET




                                                                                      0  30 CO UETEOS
Figure  115.   Conceptual diagram of the telescopic mesh refinement modeling approach (Ward, et al., 1987).
                                                    126

-------
Figure 116.
Finite-difference grid used for the regional-
scale flow model of the Great Miami River
Valley-fill aquifer (Ward, et al., 1987).
                       STREAM
                       REACHES
     HAMILTON
 NORTH WELLFIELD
                        B CHAMPION
                             PAPER
LOCAL MODEL
                                                                 CONSTANT PRESSURE
                                                                 BOUNDARY
                                                              •  PUMPING CENTER
                               HAMILTON
                           SOUTH WELLFIELD
                                               127

-------
Figure 117.    Local-scale model conceptualization
            (not to scale) (Ward, et al., 1987).
                                            TWO-MILE
                                               0AM
                                     GREAT
                                     MIAMI
                                     RIVER
                                                           FORO
                                                        HYDRAULIC
                                                          CANAL
                           CHAMPION
                            PAPERS
     MODEL
     LAYER
J_


~3
                    4

                    5
                       •Kh^ooft/d:
                                       jijirty ;Sands';and 'firbvelV:-::;
SiUs'Sands',Clays''  '•  '-'•••.'  '' ',".
 •. .   •  . .   . •. c!ea.n Sand and Gravels.
                                             rNOTE
: Conversion factor
1ftd~':0.3049md'}
                                             128

-------
            i r-j ri-7 :;--:•'" :-•••••- •.
            _M  ! i i i '•$..;! | i i! j  <
                 till i iff hrr ri"1
                                        577
                     HYDROELECTRIC DAM—
                 561
 562
                                       564
                                       563
     /CALCULATED
    «V  POTENTIOMETRIC
  <§*  SURFACE CONTOUR
 /    IN  FEET

    •  PUMPING CENTER
    o    1000 FEET
    HV-H
    0    300 METERS

NOTE: Conversion  Factor

      Iff* 0.3049m
Figure 118.    Calculated potentiometric surface for the
            shallow interval (local-scale  flow model,
            layer 1) (Ward, et al., 1987).
                     129

-------
                                               563
                   561
562
/CALCULATED
A/ POTENTIOMETRIC
«T SURFACE CONTOUR
/ IN FEET
• PUMPING CENTER
0 1OOO FEET
I I ' I I
0 300 METERS
NOTE: Conversion Factor
1ft= 0.3049m
si
•



















































•X
4:





































|
i
1





• i
I
i
l



1 0
1






. REMEDIATION WELLS
. R
- s
E:
ii
. E
II
»TE IN 6PM 	 IS
HALLOW
(TRACTION 	 0
4TERMEDIATE
(TRACTION 	 A
SECTION 	 •
1
1








~
;



:



















\












j'





f








•


-





20






\m
\



to








!jl















u

i









o
It












49


-,14







a
it



















1 MOTS •' Ccnnnmn Fxur
l«w> • OMJILfi




|






*


bl

L
Bis


•^





:



•*







o2



















u


V




































V





















i






















































0 100 «0 FttT
o 30 (o Hrrcn
                                                            Figure  120.    Finite-difference grid used in the site-scale
                                                                          flow and transport model  for the Chem-
                                                                          Dyne site. Relationship to  local-scale
                                                                          model is shown in Figure 115 (Ward, et al.,
                                                                          1987).
Figure  119.    Calculated  potentiometric  surface for the
              deep interval (local-scale flow model,  layer
              4) (Ward, etal., 1987).
                                                      130

-------
                                                                        MASS UNREMEDIATED
                                                                           LEAVING GRID
  PROPOSED
     PLAN
                                                      \    \MASSREMOVED
                                                        ^»      X DV OCIiCniATI.
                                                                 BY REMEDIATION
                         MASS IN GROUNDWATER
            H0it'- Conversion Factor
                 lib - 0.4535
                                                        100

                                              TIME  (doys)

Figure 121.  Nviass of VOCs versus time (Ward, et al., 1987).
                             1000
      10,000
          0.1        1         10
                           TIME (days)
1000
 Figure 122.    Extraction well concentrations versus time
              (Ward, etal., 1987).
                                                 131

-------

-------
                                       CHAPTER  10
   MANAGEMENT CONSIDERATIONS IN TRANSPORT AND FATE  ISSUES
                                        Joseph F. Keely
Perspectives for Site Characterizations
and Remediations
The preceding chapters discussed many concepts
pertinent to investigating and predicting the trans-
port and fate of contaminants in the subsurface.
Recognition  of the fact that  these  concepts are
evolving is  important when  making  decisions
regarding detection and remediation  of subsurface
contamination. This importance lies not  only in an
awareness of existing uncertainties, but also in the
realization that conventional site characterization
approaches have fallen considerably behind the
state-of-the-art. From a practitioner's perspective,
most subsurface contamination assessments do not
adequately emphasize the need to obtain detailed
information about preferential pathways and the
natural  processes affecting the transport behavior
and ultimate fate of contaminants. When site
characterization efforts incorporate state-of-the-art
characterizations of natural process parameters,
rather than relying almost exclusively on conven-
tional collection of ground-water samples for chem-
ical analyses, the quality and  cost-effectiveness of
subsurface contamination remediations may be
improved significantly.
Site Characterization Approaches
Tables 14, 15, and  16 provide summaries of the
principal activities, benefits, and shortcomings of
three possible site characterization approaches: con-
ventional; state-of-the-art; and state-of-the-science.
Each activity of the conventional approach can be
accomplished with semi-skilled labor and off-the-
shelf technology. Together  with  moderate to  low
costs,  these readily available tools and techniques
are reason enough for perpetuation of the conven-
tional approach - until one notes the shortcomings.
Conventional approaches cannot thoroughly charac-
terize the extent and probable behavior of a sub-
surface contaminant plume; they are, by design, a
compromise between the desire to discover the key
problems at a site and the  equal  desire to keep
expenses to an absolute minimum.

A comparison of Tables 14, 15, and 16 suggests that
state-of-the-art and state-of-the-science approaches
may be more costly to implement in site character-
izations,  but also that the increased value of the
information obtained is likely to save costs because
of dramatic improvements in the technical effective-
ness (e.g., all portions of the  zone of contamination
cleansed) and efficiency (e.g.,  treatment of the mini-
mum volume at the lowest cost) of the site cleanup.
Key management uncertainties regarding the
degree of health threat posed  by a site, the selection
of appropriate remedial action technologies, and the
duration and effectiveness  of the remediations
should decrease significantly  with the implementa-
tion of more sophisticated site characterization
approaches.

The economic  benefits of advanced site charac-
terization approaches are illustrated conceptually in
Figure 123. The  illustration implies that modest
increases in site characterization  expenses (pre-
sumably  for  more sophisticated data collection and
interpretation efforts) will generate  large decreases
in cleanup costs  by virtue of greater effectiveness
and efficiency of the remedial design and operation.
In kind, total costs would fall  dramatically since
cleanup costs normally comprise the majority of site
expenditures. Maximum return on increased invest-
ments is expected for the state-of-the-art approach
                                              133

-------
and will  dimmish as  the state-of-the-science ap-
proach is reached because highly specialized equip-
ment and personnel are not widely available.

Testing these conceptual relationships directly is not
possible because  an investigation and  remediation
cannot be carried out to fruition along each approach

Table 14.  Conventional Approach to Site Characterization
 Actions Typically Taken
   Install several dozen shallow monitoring wells
   Sample and analyze numerous times for 129 + priority
   pollutants
   Define geology primarily by driller's log and cuttings
   Evaluate hydrology with water level maps only
   Possibly obtain soil and core samples for chemical analyses
 Benefits
   Rapid screening of problem
   Moderate costs involved
   Field and lab techniques standardized
   Data analysis relatively straightforward
   Tentative identification of remedial options possible
 Shortcomings
   True extent of problem often misunderstood
   Selected remedial alternative may not be appropriate
   Optimization of remedial actions not possible
   Cleanup costs unpredictable and excessive
   Verification of compliance uncertain and difficult
Table 15.   State-of-the-Art  Approach to  Site  Charac-
           terization
 Recommended Actions
   Install depth-specific well clusters
   Sample and analyze for 129 +• priority pollutants initially
   Analyze selected contaminants in subsequent samplings
   Define geology by extensive coring/split-spoon samplings
   Evaluate hydrology with well clusters and geohydraulic tests
   Perform limited test on solids (grain size, clay content)
   Conduct limited geophysical surveys (resistivity soundings)
 Benefits

 • Conceptual understanding of problem more complete
 • Better prospect for optimization of remedial actions
 • Predictability of remediation effectiveness increased
 •. Cleanup costs lowered, estimates improved
 • Verification of compliance more soundly based

 Shortcomings
   Characterization costs somewhat higher
   Detailed understanding of problem still difficult
   Full optimization of remedial actions not likely
   Field tests may create secondary problems
   Demand for specialists increased
 Table  16.   State-of-the-Science Approach to Site Charac-
            terization
  Idealized Approach

  •  Assume "state-of-the-art-approach" as starting point
  •  Conduct tracer tests and borehole geophysical surveys
  •  Determine percent organic carbons, exchange capacity, etc., of
    soils and subsurface sediments
  •  Measure redox potential, pH, dissolved oxygen, etc., of soils
    and subsurface sediments
  •  Evaluate sorption-desorption behavior using select cores
  •  Assess potential for biotransformation using select cores

  Benefits
    Thorough conceptual understanding of problem obtained
    Full optimization of remedial actions possible
    Predictability of remediation effectiveness maximized
    Cleanup costs lowered significantly, estimates reliable
    Verification of compliance assured
                                                           Shortcomings
    Characterization cost significantly higher
    Few previous field applications of advanced theories
    Field and laboratory techniques not yet standardized
    Availability of specialized equipment low
    Demand for specialists dramatically increased
simultaneously. The best that can be done is to note
the magnitude of changes in  perceptions, decisions,
work plans,  etc., when advanced  techniques are
applied at a ground-water contamination site  that
has undergone a conventional  level of site charac-
terization. The latter situation is fairly common
because many first attempts  at a remedial investi-
gation turn up additional problems or complexities
not suspected when the investigation was budgeted
and implemented, and do not  generate consistent or
meaningful information.

Recognizing the need for  more   technically
sophisticated site characterization efforts is only the
first  step toward an improved  remedy.  Until
recently,  implementation of technically sophisti-
cated site characterization approaches was consid-
ered  difficult due to the scarcity of skilled labor and
professionals  knowledgeable in specialized tech-
niques. Now, however, there  is rapid  growth in the
number of skilled professionals, as witnessed  by the
hundreds  of training courses  offered  annually and
the technical assistance  and  information transfer
programs within EPA and other Federal  and State
agencies. Legally, the  passage of the Superfund
Amendments and Reauthorization Act (SARA) gave
EPA and  interested parties the opportunity to test
promising technologies at Superfund sites (e.g., the
Superfund Innovative Technology Evaluation (SITE)
program  at  EPA's  Office  of Research  and
Development).
                                                      134

-------
 HIGH
   CO
   O
   o
   111
 LOW
       CONVENTIONAL   STATE-OF-     STATE-OF-
         APPROACH      THE-ART    THE-SCIENCE
 Figure 123.   Conceptualization of trade-offs of costs in
             investigations and cleanups, as a function
             of approach used for site characterization.

Illustrative Scenarios
It is helpful to examine possible scenarios that might
result from these different  site investigation
approaches. Figure 124 depicts  a  hypothetical
ground-water and soil contamination site located in
a mixed residential and light industry section of a
town in the Northeast. As illustrated, there are three
major plumes: an acids plume (e.g., from electrolytic
plating operations); a phenols plume (e.g., from a
creosoting operation that used large amounts  of
pentachlorophenol); and a volatile organics plume
(e.g., from solvent storage leaks). In addition, on-site
soils are heavily contaminated in one area with
spilled pesticides and in another area with spilled
transformer oils that contain high concentrations of
PCBs.

The hydrogeologic setting for the hypothetical site is
a productive alluvial aquifer that is composed of an
assortment of sands and gravels interfingered with
clay and silt remnants of old  streambeds and
floodplains deposits that were continually dissected
and crosscut by a central river as the valley matured
over geologic  time.  The deeper portion of the
sediments is highly permeable and is the zone most
heavily used for municipal and industrial supply
wells, whereas the shallow portion of the sediments
is only moderately permeable since it contains many
more occurrences of clay and silt lenses. The pre-
dominant ground-water flow direction in the deeper
zone parallels the river (which also is parallel  to the
axis of the valley), except in localized areas around
municipal and industrial wellfields. The predom-
inant direction of flow in the  shallow zone  is
seasonally dependent, with a strong component of
flow toward the river during periods of low flow in
                                                         tributary
           river
                                                     s-^-v^ v
                                                         "*»     ,''       \
                                                          ••--.' acids plume
 Figure 124.   Hypothetical ground-water contamination problem.
                                                135

-------
the river, and a strong component of flow parallel to
the river during periods of high river flow. Strong
downward components of flow carry water from the
shallow zone to  the deeper zone throughout
municipal and industrial wellfields, as well as along
the river during  periods of high flow.  Slight
downward components of flow exist elsewhere due to
local recharge by infiltrating rainwater.

A conventional level of site characterization would
ostensibly define the  horizontal extent  of the  most
mobile/widespread plume, but would provide only a
superficial understanding of variations in the compo-
sition of the sediments. An average coefficient of
permeability (hydraulic conductivity) would  be
obtained from  review of previously published
geologic reports and assumed to be representative of
the entire aquifer for the purpose of estimating
flowrates.

The kind of cleanup likely to result from a conven-
tional site investigation is illustrated in Figure 125.
The volatile organics plume would be considered the
most important  to remediate since  it is the most
mobile, and an extraction system would be installed.
Extracted fluids would be air-stripped of volatiles
and then passed through a treatment plant for
removal of non-volatile residues, probably by rela-
tively expensive filtration through granular acti-
vated carbon.

Extraction wells would be  placed along the
downgradient boundary of the volatile organic
compound (VOC) plume to withdraw contaminated
ground water. A couple of injection wells would be
placed upgradient and used to return a portion of the
extracted and treated waters to the aquifer. The
remainder of the pumped and treated waters would
be discharged to the tributary under a National
Pollution Discharge Elimination System (NPDES)
permit. Information obtained from the drilling logs
and samples of the monitoring wells could do no more
than position all of the screened sections  of .the
remediation wells at the same depth (shallow). The
remediation wellfield would be scheduled to operate
for the amount of time needed to remove a volume of
water somewhat greater than that estimated to
reside within the bounds of the zone of contam-
ination. The latter would amount to perhaps three to
five times the nominal value of contaminated water
and would be based on average retardation values
(found in the  scientific literature) for contaminants
found at the  site. The PCB-laden soils would be
excavated and sent off to an incinerator or approved
waste  treatment  and disposal  facility.  The
decision-makers would have based their approval of
                                                      tributary
          river
                                                  S-A--XV
                                                     ~-^ •evy-'     A
    $®®&r-             vr.-' H W *&
    W&^    , *km* (. •&.. ?. .;;>VK:; "p
    :jfe •;  > ••  . ,'S «;fa-r*., *!*•  ,; -;^ ^,a^
    !1|::!:	^^|s».rftVv*^ ••>. -|^'::!^
    '*•,
Figure 125.   Typical conventional cleanup applied to ground-water contamination problem in Figure 124.
                                             136

-------
such a remedy on the presumption that the plume(s)
were adequately defined, or that the true magnitude
of the problem does not differ substantially from
these definitions, save for the possibility of a longer
period of pumpage.

Incorporation of some of the more common state-of-
the-art site investigation techniques, such as pump
tests, installation of vertically separated clusters of
monitoring wells (shallow,  intermediate, and deep)
and river stage monitors, and chemical analysis of
sediment and soil samples would likely result in the
kind of remediation illustrated in Figure 126. Since a
detailed understanding of the geology and hydrology
would be  obtained, optimal selection of well loca-
tions, wellscreen positions, and flowrates (gallons
per minute)  for  the remediation  wells could be
determined. A special program to recover the  acid
plume and neutralize it could be instituted as well as
a special program for  the pesticide plume. This
approach would probably lower treatment costs over-
all, despite the need for separate treatment trains for
the different plumes, because substantially lower
amounts of ground water would be treated by expen-
sive carbon filtration.

The extraction wellscreens' positions would  become
increasingly deeper as one gets closer to the river
because monitoring well clusters would have indi-
cated that the plume is migrating beneath shallow
accumulations of clays and silts to the deeper, more
permeable sediments.  Approximately two-thirds of
the extracted and treated ground water would be
reinjected through injection wellscreens positioned
deep to avoid diminishing the effectiveness of nearby
extraction wells. As in the conventionally based
remedy, the  remediation wellfield would be sched-
uled to operate for the amount of time  needed to
remove a volume of water based on average contam-
inant retardation values and the volume of ground
water residing in the  zone of  contamination. The
detailed  geologic and hydrologic information ac-
quired, however, would result in an expectation of a
more rapid cleansing of specific portions of the zone
of contamination. The decision-makers would have
based their approval of this remedy on the presump-
tion that  the remediation is optimized to the point of
providing the most effective cleanup, though the
efficiency of the remediation may be  less than
optimal.

If all state-of-the-art investigation tools were used at
the site, there would be an opportunity to evaluate
the desirability of using a subsurface barrier wall to
enhance remediation efforts (Figure 127). The wall
would not be expected to entomb the plumes, but
                                                          tributary
          river

                                 pesticide area
                              (pump & treat w/ GAG)
                       treatment
                       facility
                          EW6* %l     phenols plume
                                •
                            VOC plume
                            (pump & air strip)
                                                                  acids plume
                                                               (pump & neutralize)
                             •• .  ;•;:;:;:;; •'••:.-..' .:/screen settings of.-
                              '   ':-::':'-'.- .' ':•:] reclamation .wells  :
  Figure 126.    Moderate state-of-the-art cleanup applied to ground-water contamination problem in Figure 124.
                                                 137

-------
would limit pumping to contaminated fluids (rather
than having the extracted waters diluted with fresh
waters available to the extraction wells, as was true
of the two previous approaches). The volume pumped
would be lower because the  barrier wall  would
increase the drawdown at each well by hydraulic
interference effects, thereby maintaining the same
effective hydrodynamic control with lesser pumpage
(note  the lower values in the sets of parentheses at
each  well  in  Figure 127,  given in gallons per
minute). Treatment costs would decrease  too,
because the waters pumped would contain higher
concentrations of contaminants  (treatment  effi-
ciencies normally fall with  decreasing  concentra-
tions). Soil washing techniques would be used on the
pesticide-contaminated area  to minimize future
source releases to ground water.

The efficiency and effectiveness of the remediation
would appear to be optimal, but that is a perception
based on the presumption  that contaminants are
readily released. Given the potential limitations to
pump-and-treat remediations  discussed in earlier
sections  of this document, however, it  is doubtful
that this advanced state-of-the-art site investigation
precludes  further  improvement. Chemical  and
biological peculiarities must be  given as much
                     attention as the site geology and hydrology. The use
                     of average retardation values from the literature
                     infers that additional improvements in effectiveness
                     and efficiency can be garnered by detailed evaluation
                     of contaminant retardation  at this site. Likewise,
                     detailed examination of the potential for biotrans-
                     formation would be expected to provide additional
                     effectiveness and efficiency.

                     At the state-of-the-science level of site character-
                     ization, tracer tests could be undertaken that would
                     provide good information on the potential for
                     diffusive restrictions in low  permeability sediments
                     and  on anisotropic biases in the flow  regime.
                     Sorption behavior of the VOCs could be evaluated in
                     part by determination of the total  organic  carbon
                     contents of the  subsurface sediments. Similarly, the
                     cation  exchange capacities of subsurface sediment
                     samples could be  determined to obtain estimates of
                     release rates and mobilities of toxic metals. The
                     stabilities of various possible forms of elements and
                     compounds could be evaluated with measurements of
                     pH, redox potential, and dissolved oxygen. Finally, if
                     state-of-the-science findings regarding potential bio-
                     transformations were used,  it might be  possible to
                     effect in situ degradation of the phenols plume and
                     remove volatile residues (Figure 128).
                                                          tributary
                                                                          V
           river

'/0EW7
', (100' pesticide area —
\ (wash soils, pump
\ & treat w/ GAG)
"Sx; » * Y Yt Y,
"?' ,' PCB soils
^ 	 " ~ . (remove)
-. ."-N ^
X X.

                          EW6
                          (150)
        phenols plume
                   ^V /
-:   _Ews       •m  •'
X»(150i..- —--—f-7.!).  /
  '-—-*"      >^      \X
                                                                           treatment
                                                                           facility
IW2  ^-^/gf.
(iso) »s fh
    (150)/^''J
                       S
                   shallow subsurface wall
                   (to maximize pumpage of plume)
                                    acids plume
                                 (pump & neutralize)
                                                        :BWt
                                       4-'.
                KEW6-'
                      VST
                                  ? -wwf^«> -. -^;:
Ffgura 127.    Advanced state-of-the-art cleanup applied to ground-water contamination problem in Figure 124.
                                                138

-------
                                                         tributary
          river
                               pesticide area
                               (wash soils, pump
                               & treat w/GAG)
                                    treatment •
                                    facility
          .' PCBsoilsl
            (remove)
                                     phenols plume
                                     (bioreclaim)
                  shallow subsurface wall
                  (to maximize pumpage
                  of plume)
(pump & air strip
majority, bioreclaim
residuals)
               acids plume
            (pump & neutralize)
Figure 128.    State-of-the-science cleanup applied to ground-water contamination problem in Figure 124.
Performance Evaluations of
Remediations
Large expenditures are made each year to prepare
for and operate remediations of ground-water con-
tamination. Regulatory responsibilities require that
adequate oversight of these  remediations be made
possible by  structuring appropriate compliance
criteria for monitoring wells. The oversight efforts
are nominally directed at answering the question,
"What can  be done  to  show whether or  not a
remediation is generating the desired contamination
control?" Recently, other questions have developed
because of the realization that pump-and-treat
remediations do not function as well as has been
presumed. Such questions include: "What can be
done to determine whether the  remediation will
meet its timelines?"  and "What can  be done to
determine whether the remediation will stay in
budget?"

Conventional wisdom states that these questions can
be answered by the use of sophisticated data analysis
tools, such as computerized mathematical models of
ground-water flow and contaminant transport. Com-
puter models can indeed be used to make predictions
about future performance, but such predictions are
highly dependent on the quality and completeness of
 the field and laboratory data. The latter is just as
 true for models evaluating pump-and-treat remedi-
 ations, in contrast to the common  belief that an
 accurate performance evaluation can be made simply
 by comparing data obtained from monitoring wells
 during remediation to the data generated prior to the
.onset of remediation. Historical trends of contam-
 inant levels at local monitoring wells are rendered
 useless by the extraction and injection wells used in
 pump-and-treat remediations. This is a consequence
 of the fact that the extraction and injection wells
 produce complex flow patterns locally, where  pre-
 viously there were comparatively simple flow
 patterns.
 Complex ground-water  flow patterns present great
 technical challenges in terms of characterization and
 manipulation (management)  of the  associated
 contaminant transport pathways. In Figure 129, for
 example, waters moving along the flowline  that
 proceeds  directly into  a pumping  well  from up-
 gradient are moving the most rapidly, whereas those
 waters lying at the lateral limits of the capture  zone
 (indicated by the bold  curved line in Figure  129)
 move  much more slowly. One result is that certain
 parts of the aquifer are flushed quite well and others
 are  remediated relatively poorly. Another result is
 that those previously uncontaminated portion^ of the
                                                139

-------
 MODERATE
   FAST
Figure 129.
                    MODERATE
             Flowline pattern generated by an extraction
             well.
 aquifer that form the peripheral  bounds  of the
 contaminant plume may become contaminated by
 the operation of an extraction well that is located too
 close to the plume boundary. This occurs because the
 flowline pattern extends  downgradient of the well.
 The latter is not a trivial situation avoided without
 repercussions by simply locating the extraction well
 far enough inside the plume  boundary that its
 flowline pattern does  not extend beyond the down-
 gradient edge of the plume, because doing so results
 in very poor cleansing of the aquifer  between the
 location of the extraction well and the downgradient
 plume boundary.

 It is not possible to determine precisely where the
 various flowlines generated by a pump-and-treat
 remediation are located unless detailed field  evalu-
 ations are made during remediation. Neither con-
 taminant  nor velocity distributions are constant
 throughout the zone of action (that portion of an
 aquifer actively manipulated by the pumping wells).
 Consequently, more data must be generated during
 the remediation (especially inside the boundaries of
 the contamination plume) than were generated
 during the entire remedial investigation/feasibility
 study process at a site, and interpretations must be
 made of those data that require much more sophisti-
 cated tools. Indeed, it  might be successfully argued
 that in most settings, monitoring well data collected
 during remediation are useless unless a mathe-
 matical model is used to  organize and analyze the
 data.

 Decisions  regarding the frequency and density of
chemical samplings must consider the detailed flow-
paths generated by the remediation wellfield, and
include changes in contaminant concentrations
resulting from variations in the influences of trans-
port processes along those flowpaths. The need to
reposition  extraction wells occasionally to remediate
 portions of the contaminated zone previously subject
 to slow flowlines means that the chemical samplings
 may generate results that are not easily understood.
 It also means that the chemical compliance points
 may have to be moved during the course of a
 remediation.

 Nor are evaluations  of the  hydrodynamic per-
 formance  of remediation wellfields easily accom-
 plished. For example, an inward hydraulic gradient
 is usually required to be maintained at the periphery
 of a contaminant plume undergoing remediation by
 use of a pump-and-treat wellfield. This requirement
 is imposed to ensure that no portion of the  plume is
 free to migrate away from the zone of action.  To
 assess this performance adequately, the hydraulic
 gradient must be measured accurately in three
 dimensions between each pair of adjacent pumping
 or injection wells. The design of an array of
 piezometers (small diameter wells  with very short
 screened intervals, used to measure the pressure
 head of selected positions  in  an aquifer)  for this
purpose can be difficult. Two points define a line and
three points define a planar surface, but many more
are needed to define the convoluted  water-table
surface that develops between  adjacent pumping or
injection wells. Not only are there velocity divides in
the horizontal dimension near active wells, but in
the vertical dimension, too, because the pressure
influence of each well extends to only a limited depth
in practical terms.
                                                  Innovations in Pump-and-Treat
                                                  Remediations
                                                  One of the promising innovations in pump-and-treat
                                                  remediations  is intermittent operation or pulsed
                                                  pumping of a remediation wellfield. Pulsed operation
                                                  of hydraulic systems is the cycling of extraction or
                                                  injection wells on and off in  "active"  and "resting
                                                  phases" (Figure 130). The resting phase of a pulsed-,
                                                  pumping  operation can  allow sufficient time for
                                                  contaminants to diffuse out of low permeability, zones
                                                  and into  adjacent high permeability zones, until
                                                  maximum concentrations are achieved in the higher
                                                                1   'a   '3   '4  's  'e   '?   'a
                                                 Figure 130.    Reduction of residual contaminant mass by
                                                              pulsed pumping.
                                              140

-------
permeability zones. For sdrbed contaminants and
NAPL residuals, sufficient time can be allowed for
equilibrium concentrations to be reached in local
ground water. Subsequent to each resting phase, the
active phase of the pulsed-pumping cycle removes
the minimum volume of contaminated ground water,
at the maximum possible concentrations, for  the
most efficient treatment. By occasionally cycling
only select wells, these wells' stagnation zones may
be brought into active flowpaths and remediated.

Pulsed operation of remediation  wellfields incurs
certain additional costs and concerns that must be
compared with its advantages for site-specific appli-
cations^" During the rest phase of pulsed-pumping
cycles, peripheral gradient control may be needed to
ensure adequate hydrodynamic control of the plume;
in an ideal situation, peripheral gradient control
would be unnecessary. This might be the case where
there are no active wells, major streams, or other
significant hydraulic stresses nearby to influence the
contaminant plume while  the  remedial action
wellfield is in the resting phase.  The plume would
migrate only a few feet during the tens to hundreds
of hours that  the  system was at rest, and  that
movement would be rapidly recovered by the much
higher'flow velocities  back toward the extraction
wells during the active phase.

When significant hydraulic stresses are  nearby,
however, plume movement during the resting phase
may be unacceptable. Irrigation or water-supply
pumpage, for example,  might cause plume
movement on the order of several tens of feet per day.
It then might be impossible to  recover the lost
portion of the plume  when the active phase of the
pulsed-pumping cycle commences. In such cases,
peripheral gradient control during the resting phase
would be essential. If adequate storage capacity is
available, it may be  possible to provide gradient
control in the resting phase  by injection of treated
waters downgradient of the remediation wellfield.
Regardless of the mechanics of the compensating
actions, their capital and operating expenses must be
added to those of the primary remediation wellfield.

Pump-and-treat remediations currently are under-
way that incorporate some of the principles of pulsed
pumping. For instance, pumpage from contaminated
bedrock aquifers and other low permeability forma-
tions results in intermittent  wellfield operations by
default; the wells are pumped dry even at low flow
rates.  In such cases, the wells are operated on
demand with the help of fluid-level sensors that
trigger the onset and cessation of pumpage.  This
simultaneously accomplishes the goal of pumping
ground water  only after it has reached chemical
equilibrium, since equilibrium occurs on the same
time frame as the fluid recharge event (both are
diffusively restricted).  In settings of moderate to
high permeability, the onset and cessation of
pumpage  could  be  keyed  to  contaminant
concentration levels in the pumped water, independ-
ent of flow changes required to  maintain proper
hydrodynamic/gradient control. As indicated in the
discussion of pulsed pumping, this may be acceptable
(pose no unreasonable risk) in circumstances where
the contaminant plume would not be  subject to
substantial movement in the absence of pumpage.

Other strategies for improvement of the performance
of pump-and-treat remediations include:

1.  Flow scheduling of wellfield operations to satisfy
    simultaneously hydrodynamic/gradient control
    and contaminant concentration trends  or other
    performance criteria.
2.  Physical  repositioning of extraction  wells to
    effect major flowline/transport pathway altera-
    tions.
3.  Integration of wellfield operations with other
    subsurface technologies (e.g., barrier  walls to
    limit plume transport and minimize pumping of
    fresh water, or infiltration  ponds to maintain
    saturated flow conditions for flushing  contam-
    inants from (normally) unsaturated soils and
    sediments).

The first of these alone would allow for  flushing of
stagnant zones by occasionally turning off individual
pumps, but  the  flushing could not be  done as
efficiently as repositioning or adding pumping wells
(the second means of improvement). The first and
second approaches differ in effects, however, because
repositioning or adding wells requires access for
drilling and necessarily precludes capping of the site
until after completion of the pump-and-treat opera-
tions. The third improvement approach, combining
pump-and-treat  with subsurface  barrier  walls,
trenching, or in situ techniques (all of which may
occur at any time during remediation), also may
require postponement of capping until after comple-
tion of the remediation.

The latter  strategy raises latent fears of lack of
control of the contaminant source, which is almost
always mitigated by isolation of the contaminated
soils and subsoils that remain long after man-made
containers  are removed from the typical site.
Fortunately,  vacuum extraction of contaminanted
air/vapor from soils  and subsoils has  recently
emerged as a potentially effective means of removing
VOCs, steam flooding is being evaluated for removal
of the more retarded organics, and in situ  chemical
fixation techniques are being  tested for the  isolation
of metals wastes. Vacuum extraction is capable of
removing several pounds of VOGs per day (since the
VOCs  readily  volatilize into the soil gas/vapor),
                                               141

-------
 whereas air stripping of VOCs from comparable
 volumes of contaminated ground water typically
 results in the removal of only a few grams of VOCs
 per day (because VOCs are so poorly soluble  in
 water). Similarly, steam flooding can be an econom-
 ically attractive means of concentrating contami-
 nant residuals as a front leading the injected body of
 steam. Regardless of the efficacy of vacuum extrac-
 tion, steam flooding, or chemical fixation in terms of
 permanent and complete remediation of the contami-
 nation in the unsaturated zone each have excellent
 potential for control of fluid and  contaminant
 movement in the unsaturated zone and should be
 considered as potentially significant additions to the
 list of source control options.  In addition,  soils
 engineering and landscape maintenance techniques
 can minimize infiltration of rainwater in the absence
 of a multilayer RCRA-style cap.

 In terms of performance evaluation of a remediation,
 the presence of a multilayer RCRA-styled cap poses
 major limitations. The periodic removal of core
 samples of subsurface solids from the body of the
 plume and the source zone, with subsequent extrac-
 tion of the chemical residues on the solids, is the only
 direct means of evaluating the true magnitude of the
 residuals and their depletion rate. Since this must be
 done periodically, capping should be postponed until
 closure of the site.

 If capping can be postponed or forgone, there will be
 great flexibility for management of pump-and-treat
 remediations which  can improve effectiveness and
 lower costs.  Also, the soils and subsoils can be
 cleansed of contamination without waiting in isola-
 tion for  eventual breakdown of a cap. Given the
 parallel theme of SARA - true remediation, not just
 stabilized problems - innovative pump-and-treat re-
 mediations and  source removal techniques may be
 the most economical  and responsible choices for
 remediations.
Managerial Considerations in Using
Transport and Fate Models

Effective Communications
One of the principal problems underlying  the
continuing  difficulties in transferring  technical
information about transport and fate issues is the
poor level of communications between specialists and
decision-makers that often results in poorly focused
remediation activities. Part of this problem occurs
because specialists and decision-makers have  dif-
ferent perceptions of their roles  and the  situations
they face. The problem of effective communication is
not easy to solve. Questions presented in Tables 17,
 18, and 19 can be used to stimulate more effective
 dialogues and may provide  insights to presenting
 material for public consumption, both in terms of
 clarity  and honest appraisals  of the  costs and
 limitations that must be accepted by the public.


 Table 17.   Screening Level Questions to Help  Focus
            Ground-water Contamination Assessments

 General Problem Definition
 •   What are'the key issues: quality, quantity, or both?
 •   What are the controlling geologic, hydralogic, chemical, and
     biological features?
 •   Are there reliable data (proper field scale, quality controlled,
     etc.) for preliminary assessments?              ,„_„
 •   Do the model(s) needed for appropriate simulations exist?

 Initial Responses Needed
 •   What is the time-frame for action (imminent or long-term)?
 •   What actions, if  taken now, can  significantly delay or
     minimize the projected impacts?
 •   To what degree can mathematical  simulations  yield
     meaningful results for the action alternatives, given available
     data?
 •   What other techniques or information (generic models, past
     experience, etc.) would be useful for initial estimates?

 Strategies for Further Study
 •   Are  the critical data gaps identified; if  not, how well can
     specific data needs be determined?
 •   What are the trade-offs between  additional data and
     increased certainty of the assessments?
 •   How much additional manpower and resources are necessary
     to improve mathematical modeling efforts?
 •   How long will it take to produce useful simulations, including
     quality control and error-estimation efforts?
Technical Support
The return on investments made in using transport
and fate models rests principally with the  training
and experience of the technical support staff
applying the model to a problem and on the degree of
communication between those persons and manage-
ment. In discussing the potential uses of computer
modeling for ground-water protection efforts, Faust,
et al. (1981) noted that the final worth of modeling
applications depends on the people who apply the
models. Managers should be aware  that specialized
training and experience is necessary to develop and
apply mathematical models,  and  relatively  few
technical support staff can be expected to have such
skills (van der Heijde, et al., 1985). This is due in
part to the need for the modeler to have familiarity
with a number of scientific disciplines so that the
model is structured faithfully to simulate real-world
situations.
                                                 142

-------
Table 18.    Conceptualization  Question* to Help  Focus
            Ground-water Contamination Assessments

 Field Techniques and Data Production
 •   Are  the installation and sampling techniques to be used
     accepted? innovative? controversial?
 •   Where are the weak spots in the assessment, and can these
     be further minimized or eliminated?
 •   What are the limitations of field tests that estimate the natural
     processes parameters of this problem?

 Model Input Parameters and Boundary Conditions
 •   How reliable are the estimates of the input parameters; are
     they quantified within accepted statistical bounds?
 •   What are  the boundary conditions,  and why are  they
     appropriate to this problem?
 •   Have the initial conditions with which the model is calibrated
     been checked for accuracy and consistency?
 •   Are  the spatial grid design(s) and time-steps of the  model
     optimized for this problem?

 Model Quality Control and Error Estimation
 •   Have these models been mathematically validated against
     other solutions to this kind of problem?
 •   Has anyone field-verified these models before,  by  direct
     applications or simulation of controlled experiments?
 •   How  do these models compare with others in terms of
     computational efficiency, ease of use, or modification?
 •   What special  measures are  being taken to estimate  the
     overall errors of the simulation?
What levels of training and experience are necessary
to apply mathematical models properly? Are "Ren-
naissance  specialists"  needed or  can  interdisci-
plinary teams be effective? The  answers to these
questions are not clear-cut, but the more informed an
individual is, the more effective he or she can be. It is
doubtful that any individual can master  each
discipline with the same depth of understanding as
specialists  in those fields, but a working knowledge
of many disciplines is necessary so  that  appropriate
questions may be put to specialists, and some sense
of integration of the various disciplines can evolve.
In  practice, this means  that modelers should be
involved in continuing education efforts with  the
support of  management. The benefits to be gained
are tremendous, and the costs of not doing so may be
equally large.

Managers and technical support  staff alike should
appreciate  the difficulty in explaining the results of
complicated models to non-technical audiences such
as  in public  meetings and courts of  law. Many
scientists find it  difficult to discuss the details of
their labors without the convenience of their scien-
tific  jargon. Some  of the  more useful means of
overcoming this limitation involve the production of
highly simplified audio-visual aids.
Table 19.    Sociopolitical Questions to Help Focus Ground-
            water Contamination Assessments

 Demographic Considerations
 •   Is there a larger population endangered by the problem than
     we are able to provide sufficient responses to?
 •   Is it possible to present this assessment in both non-technical
     and technical formats to reach all audiences?
 •   What role can modeling play in  public information  efforts
     (e.g., effective graphics)?
 •   ;How prepared are we  to respond to criticism  of  this
     assessment (e.g., supportive materials)?

 Political Constraints
 •   Are there non-technical barriers to the techniques to be used
     to produce this assessment, such as "tainted by association"
     with a controversy elsewhere?
 •  *Do we have the cooperation of all  involved parties in
     obtaining the necessary data and implementing solutions?
 •   Are similar technical efforts for this problem being undertaken
     by friend or foe?
 •   Can the results of the assessment  be turned against us; are
     the results ambiguous or equivocal?


 Legal Concerns
 •   Will these activities meet all regulations?
 •   If we are dependent on others for key inputs, how  do we
     recoup losses stemming from possible non-performance?
 •   What liabilities are incurred for projections arising from poor
     data, misinterpretations, or models used?
 •   Do any of the issues to be addressed by this assessment
     require the advice of attorneys?
Potential Liabilities
Some of the liabilities in using transport and fate
models relate to the degree  to which predictive
models are used in permitting or banning  specific
practices  or products.  If a model is incapable  of
treating specific applications properly, substantially
incorrect decisions  may result. Depending on the
application, unacceptable environmental effects may
begin to accumulate long before the nature of the
problem is  recognized. Conversely, unjustified
restrictions may be imposed on  the regulated
community. Inappropriate or inadequate models also
may cause the re-opening clause of a negotiated
settlement agreement to  be  invoked when, for
instance, compliance requirements that were guided
by the predicted plume  behavior  generated by the
model are not met.

Certain liabilities relate to the use of proprietary
codes in legal settings, where the inner workings of a
model may be subject to disclosure in the interests of
justice.  The desire for confidentiality by the model
developer would likely be subordinate to the public
right to full disclosure of actions predicated on
modeling  results. The mechanisms for protection of
                                                      143

-------
 proprietary rights typically do not extend beyond
 extracted promises of confidentiality by reviewers or
 other interested parties.  Hence, a developer of
 proprietary codes still risks exposure of innovative
 techniques.

 Other liabilities may arise  as the result of mis-
 applications  of transport and  fate  models,  or
 applications of models later found to be faulty.
 Frequently, the choices of boundary  and initial
 conditions for a given application are hotly con-
 tested; misapplications of this kind are  undoubtedly
 responsible for many of the reservations expressed
 by would-be model  users. Also, many times in the
 past, a  well-used and highly regarded model code
 was found to contain errors or lack the  ability to
 faithfully simulate  certain situations for which it
 was widely presumed to be applicable  (El-Kadi,
 1988). The best way to minimize these liabilities is to
 adopt strict quality control procedures  for each
 application.
Economic Considerations
The nominal costs of the support staff, computing
facilities, and specialized graphics production equip-
ment associated with transport and fate modeling
efforts can be  high. In addition, quality  control
activities can result in substantial costs. The deter-
mining factor in controlling these costs is the degree
to which a manager must be certain of the charac-
teristics of the model and the validity of its output.

As a general rule, costs are greatest for personnel,
moderate for hardware, and minimal for software.
The exception to this ordering relates  to the com-
bination of software and hardware purchased. An
optimally outfitted business computer (e.g.,  VAX
11/785 or IBM 3031) costs higher than $100,000, but
can rapidly pay for itself in terms of dramatically
increased speed and computational power. A  well-
complemented  personal computer (e.g., Compaq
Deskpro 386 or Macintosh II) may cost $10,000 with
moderate accessories, but the significantly slower
speed and limited  computational power may  infer
hidden costs in terms of the inability to perform
specific tasks. For  example, highly desirable statis-
tical packages like SAS and SPSS are only available
with reduced capabilities or altogether  unavailable
for personal computers. Many of the most sophisti-
cated mathematical models are  only available in
their fully capable form on business computers.

Figure 131 gives a brief comparison of typical costs
for software at different levels of computing power.
Obviously, the software for less capable computers is
cheaper,  but the  programs  are not equivalent;
therefore,  managers need to consider the appropri-
ateness of a chosen level of computer power. If the
decisions to be made will be based on very little data,
it may not make sense to insist on the most elegant
software and hardware. If the intended use involves
substantial amounts of data  and sophisticated
analyses are desired, it would be unwise to opt for the
least expensive combination.
   300
       Mainframe Fortran    PC Fortran      PC BASIC
      Ground-Water Modeling Software Categories
              minimum cost
                           maximum cost
 Figure 131.
              Average  price  per category for ground-
              water models from  the  international
              ground-water modeling center.
There seems to be an increasing drive  away from
both ends of the spectrum of computing power and
toward  its middle; that is, the use of. powerful
personal computers is increasing rapidly, whereas
the use of small programmable calculators and large
business computers is declining.  In part, this stems
from the significant improvements in the computing
power and quality of printed outputs obtainable in
recent years from personal computers.; Also,  the
telecommunications capabilities  of personal com-
puters now commonly include  emulation of  the
interactive terminals of large business computers so;
that vast computational power can be accessed and
the results retrieved with no  more than a phone call.
Recently, many of the mathematical models and data
packages have been 'down-sized' from mainframe
computers to personal computers and more are now
being written directly for this market.
                                               144

-------
Figure  132 provides some  idea  of the  costs of
available software and hardware for personal  com-
puters.  Table 20 lists  recent  salary ranges and
desired background for the technical  support  staff
needed to operate such systems. Figure 133 attempts
to place all of  the nominal costs of  subsurface
contaminant transport modeling in perspective.
                           Table 20.    Desired Backgrounds and Salary Ranges Adver-
                                       tised  for  Positions' Requiring Ground-water
                                       Modeling
2.00O
g 1.600
CO
g 1,000
oZ
CO 500
1
ft


-


Jin


•-






Jl





•nil
-



1





•











Jinn
            2  3  4  5  6  7   8   9  10 11  12  13  14  15
             Vendors of Ground-Water Models
                           |   I
    1. International Ground

      Water Modeling Ctr.

    2. Computapipe Co.

    3. Data Services, Inc.

    4. GeoTrans, Inc.

    5. Hydrosoft, Inc.

    6. In Situ, Inc.

    7. IrriscoCo.
 8. Koch and Assoc.

 9. KRS Enterprises, Inc.

10. Michael P. SpinksCo.

11. RockWare, Inc.

12. Solutech Corp.

13. T.A. Prickett & Assoc.

14. James S. Ulrick Co.

15. Watershed Research, Inc.
Figure 132.     Price ranges for IBM-PC  ground-water
               models available from various sources.

The technical considerations discussed in previous
sections indicate that  the desired accuracy of the
modeling effort directly  affects  the total costs  of
mathematical simulations. Thus, managers will
want to determine the incremental benefits  gained
by increased expenditures, especially for more in-
volved mathematical modeling efforts. While many
economic theories exist for determining these bene-
fits, the most straightforward of these are the cost-
benefit approaches commonly used to evaluate the
economic desirability of water resource projects.
There are two generalized  approaches  commonly
used:  the Benefit/Cost Ratio method and the Net
Benefit method.

The  Benefit/Cost (B/C)  Ratio  method involves
tallying the economic value of  all  benefits and
dividing that sum by the total cost involved in gener-
ating those benefits. A ratio greater than  one  is
required for  the project to be considered viable,
                            Position Title:
                            Salary Offered:
                            Desired Background:
                                                          Position Title:

                                                          Salary Offered:
                                                          Desired Background:
                                                          Position Title:
                                                          Salary Offered:
                                                          Desired Background:
                            Position Title:

                            Salary Offered:
                            Desired Background:
Position Title:

Salary Offered:
Desired Background:
Position Title:

Salary Offered:
Desired Background:
                            Position Title:

                            Salary Offered:
                            Desired Background:
Hydrogeologist (Argonne Natl. Lab.)
$31,619-$48,876
Familiarity and experience in field testing
and monitoring of ground-water flow and
the use of numerical models

Hydrologic Modeler (University of Wis-
consin)
$23,000-$25,000
Primary strength in application of
numerical models to ground-water flow
and chemical transport; strong chemical
background

Soil Scientist (USEPA)
$31,619-$41,105
Knowledge of: (1) soil physics;  (2)
processes governing transport and fate
of chemical and biological species (3)
math, statistics, and geostatistics; and
the ability to develop computer  codes

Ground-water Hydrologist (Inyo County
Water Dept, CA)
Starting up to $32,000
At least three years experience including ,
field work, surface/ground-water resource
evaluations, environmental assessments,
flow modeling; FORTRAN

Hydrogeologist (S.W. Texas State
University)
$24,444-$30,096
Academic training in hydrogeology, min.
2 years experience, knowledge of
limestone aquifers and computer
operations

Geochemist {U.S. Nuclear Regulatory
Com.)
$21,170-$41,105
Knowledge of solute and radionuclide
transport, including speciation,
attenuation (sorption), numerical
modeling.

Hydrogeol./Civil Engr. (typical consulting
firm)
"commensurate with experience"
Strong background in applied ground-
water flow and contaminant transport
modeling, knowledge of Federal/State
regulations
                           though there may be sociopolitical reasons  for
                           proceeding with projects  that do not meet this
                           criterion.  An example would be a new  project that
                           has gained considerable social or political momen-
                           tum,  but  which begins  to  exceed initial cost
                           estimates. Not proceeding or substantially altering
                           the work may be economically wise; however, such a
                           decision may be viewed as a breach  of  faith by  the
                           public. Regardless of how this  kind of situation
                           evolves, it is not uncommon for certain costs to be
                                                      145

-------
                Compaq 386
                  & Accessories
                   w/Financing
                     ($15K)
Ground-Water
Models ($5K)
 Benefits -
25% Salary
                                Service,
                              Expendables,
                              and Optional
                             Software ($10K)
                                                            Salary: 1.0 FTE
                                                               - $40K/yr
                                                        Total Cost:  $280K
                                 Ground-Water
                                 Models ($15K)
                    Benefits -
                   25% Salary
                VAX 11/785
                  & Accessories
                  w/Financing
                    ($125K)
                    Service,
                 Expendables,
                 and Optional
                Software ($25K)
                 Salary:  1.0 FTE
                    - $40K/yr
                                                        Total Cost:  $415K
Figure 133.    Costs of sustaining ground-water modeling capabilities at two different computing levels for a five-year period.
forgiven or subsidized, muddying the picture for
incremental benefits or trade-off analyses.

The Net Benefit method involves determining the
arithmetic difference of the total benefits and total
costs. The obvious criterion in this method is that the
proposed work result in a  situation where total
benefits exceed total costs. This approach is most
often adopted by profit-making enterprises because
they seek to maximize the difference, which is their
source of income. The B/C Ratio method, by contrast,
has long been used  by government agencies and
                other non-profit organizations because they seek to
                show the simple viability of their efforts, irrespective
                of the costs involved.

                In a very real sense, these two general economic
                assessment methods stem from different  philoso-
                phies, yet they share many common difficulties and
                limitations. For example,  there is a need to predict
                the present worth of future costs and to amortize
                benefits over the life of a  project. The mechanics of
                such calculations are well  known, but they necessar-
                ily involve substantial  uncertainties.  The  present
                                                146

-------
worth of a series of equal payments for equipment or
software can be computed by:

        P = A * [((1 + i)n - l)/(i * (1 + i)n)]

where P is the present worth, A is the series payment
each interest period, i is the interest rate per period,
and n is the number of interest periods (White, et al.,
1984). Note, however, that the interest rate must be
estimated. A  small  difference in the interest  rate
results  in tremendous differences in the present
worth estimate  due to the exponential nature of the
equation.

Managers also  may compute the future worth of a
present investment, calculate  the percentage of
worth annually  acquired through single payments or
serial investments, and so on. One should be aware
that these methods of calculating costs belong to the
general family of single-objective, or mutually exclu-
sive alternative analyses which assume that the cost
of two actions is obtained by simple addition of their
singly computed costs. In other words,  the  efforts
evaluated are presumed to have no interactions; for
some aspects  of ground-water modeling efforts, this
assumption may not be valid. For instance, one may
not be able to specify software and hardware costs
independently.  In addition, these methods  rely on
the expected  value concept, wherein the expected
value of an alternative is viewed as  the  single
product of its effects and the probability of their
occurrence. This means that  high-risk,  low-proba-
bility alternatives and low-risk, high-probability
alternatives have the same expected value.

To overcome these difficulties, methods can be  used
which incorporate functional dependencies between
various alternatives and do not rely on the expected
value concept,  e.g., the multi-objective decision
theories (Asbeck and Haimes,  1984; Haimes and
Hall, 1974; Haimes, 1981).  A conceivable use would
be the estimation of lowered health risks associated
with various remedial action  alternatives at a haz-
ardous waste site. In  such a  case, the  output of a
contaminant transport model would be used to
provide certain inputs (i.e., contaminant concentra-
tions and transport velocities)  to a health  effects
model, and that model would produce  the inputs
(e.g., probability of additional cancers per level of
contaminant) for the multi-objective decision model.
The primary difficulty with multi-objective ap-
proaches is estimating the probabilities of each
alternative so that  the objectives which are to be
satisfied may be ranked in order of importance. A
related difficulty is the need to specify the functional
form of the inputs (e.g., the population distribution
function of pumpage rates or contaminant levels).
Historical  records  about the inputs may  be too
insufficient  to allow their functional forms to be
determined.
Another problem compounding the cost-benefit
analysis of mathematical modeling efforts relates to
the need to place an economic value on intangibles.
For example, the increased productivity a manager
might expect as a result of rapid machine calcu-
lations replacing hand calculations may not be as
definable in terms of the improved  quality  of
judgments made as it is in terms of time released for
other duties. Similarly, the estimation of improved
ground-water  quality  protection benefits  may
necessitate some valuation of human life and suffer-
ing. Hence, there is often room for considerable
adjustment of the values of costs and benefits. This
flexibility can be used inappropriately to improve
otherwise unsatisfactory  economic evaluations. For
instance, Lehr (1986) offers a scathing indictment of
the Tennessee Valley Authority for conducting
hydroelectric projects which have "incredibly large
costs" and "negative cost benefit ratios."

Finally, some costs and benefits may be incorrectly
evaluated because they are based on probabilistic
data, a fact which goes unrecognized. For instance,
the key parameters affecting ground-water computa-
tions (e.g., hydraulic conductivity) are only known
within an order-of-magnitude due to data collection
limitations.  In these situations, great caution must
be exercised.
Summary
This chapter described the present and future status
of ground-water contamination assessments and the
large difference between what  may be  known in a
theoretical context and what is put into practice.
This difference exists because field methods  used to
characterize important natural process  parameters
are still relatively crude  and there remains  the
unwarranted perception that mathematical  models
can estimate these important parameters accurately
with small amounts of data.

Historically, decision-makers have been reluctant to
fund state-of-the-art site characterization  ap-
proaches. Such approaches are  more  frequently
being recognized as the appropriate means by which
to design, implement, and complete the most effec-
tive and efficient remediations of ground-water con-
tamination. Mathematical models can  be used to
gain insights to potential behavior of a plume and to
test hypotheses about conceptualizations, so as to
generate better understandings of important phys-
ical, chemical, and biological processes which affect
specific ground-water  problems  — but only  if
adequate data are available.
                                                147

-------
  Pump-and-fcreat remediations are  far more
  complicated than previously thought. The variations
  in ground-water flow velocities and directions that
  are imposed on natural systems by remediation
  wellfields  tremendously complicate attempts to
  evaluate the progress of pump-and-treat remedi-
  ations, in  part because of the tortuosity  of the
  flowlines that are generated, and because historical
  trends of contaminant concentrations at  monitoring
  wells are rendered useless for further predictions and
  evaluations. Just as it is improbable that a  proper
  understanding of the true extent of the contam-
  ination problem can be obtained unless sophisticated
  site characterization approaches are utilized, it is not
  possible to optimize the effectiveness and efficiency
  of a pump-and-treat remediation unless the geologic,
  hydro logic, chemical, and biological complexities of a
  site are defined adequately. While it is evident that
  no guarantee can be given  that a remediation will
  indeed be completely effective and optimally efficient
  by virtue of the use of state-of-the-art and state-of-
  the-science techniques, it is  equally evident  that
  their use ensures that the best remediation practical
  can be achieved.

  The  use of mathematical models in the decision-
  making process means that the user will inevitably
  incur certain liabilities. Anticipation of problem
  areas and some sensitivity to the possible misuses of
  models will greatly minimize potential  liabilities.
  Rigorous quality control programs also will achieve
  the same goal. A number of direct and indirect costs
  attend the use of models,  not the least of which
  involve efforts to obtain and  retain specialized
  experts. There will always be a significant degree of
  professional judgment called for in  ground-water
  contamination assessments, and, judging from the
  frequency with which significant errors are intro-
  duced by poor field work, this is an area that needs
  much attention. More strict  licensing of engineers or
  scientists engaged in this kind of work will probably
    do little to improve the situation. Rather, much more
    effective  communication between specialists and
    decision-makers, and their communication with the
    public is needed. Bringing transport and fate issues
    out of the research community and into the political
    arena, and describing and addressing the problem of
    subsurface contamination is - with all its costs and
    technical limitations - the real solution.
    References
    Asbeck,  E. and Y.  Y.  Haimes,  1984.  "The
     Multiobjective Risk Method." Large Scale Systems,
     Vol. 13, No. 38.

    El-Kadi,  A.,  1988.  "Applying .the  USGS  Mass
     Transport Model (MOC) to Remedial Actions by
     Recovery Wells." Groundwater, Vol. 26, No. 3.
    Faust, C.  R., L. R. Silka, and J. W. Mercer,  1981.
     ."Computer  Modeling  and  Ground-water
     Protection." Groundwater, Vol. 19, No. 4.
    Haimes, Y. Y.,  1981. Risk/Benefit Analysis in  Water
     Resources Planning and  Management. Plenum
     Publishers, New York.

    Haimes, Y. Y. and W. A. Hall, 1974. "Multiobjectives
     in Water Resources  Systems  Analysis: the
     Surrogate Worth Trade-off Method." Water
     Resources Research, Vol. 10.
    Lehr, J. H., 1986. "The Myth of TVA." Groundwater,
     Vol. 24, No. 1.
    van der Heijdej P. K. M., et al., 1985.  Groundwater
     Management:  The Use of Numerical Models, 2nd
     Edition,  AGU Water Resources Monograph No. 5,
     American Geophysical Union, Washington, DC.
    White, J.  A., M.  H. Agee, and K.  E. Case,  1984.
     Principles of Engineering Economic Analysis, 2nd
     Edition. John Wiley and Sons, New York.
•C.S. GOVERNMENT PRINTING OFFICE: 1994-550-001/00175
148

-------

-------

-------

-------
                                                                                TJ O
                                                                                CD ^


                                                                                Q) O
                                                                                              > mi
                                                                               8
                                                                                              Q
                                                                                              0>
                                                                                              00
§
NJ
      «I*   §2?
      g Q) O   •  0) Q)
      2 n c   rr o «
      _ IT o   Q) J CD

      2.0°   i.°3
OO
O
M
vo
        Si
      1 S 3
              s ^ ^
                ? 2.
                                                                                       m

-------