vEPA
          United States
           'vironmental Protection
           jency
           Office of
           Research and Development
           Washington, DC 20460
EPA/600/2-91/053
September 1991
Assessing LIST Corrective
Action Technology

A Scientific Evaluation of the
Mobility and Degradability of
Organic Contaminants in
Subsurface Environments

-------
                                      EPA/600/2-91/053
                                      September 1991
    ASSESSING UST CORRECTIVE ACTION TECHNOLOGY

          A Scientific Evaluation of  the
Mobility and Degradability of Organic Contaminants
            in Subsurface Environments
                        By
                  Varren J. Lyman
                 Patrick J. Reidy
                   Benjamin Levy
             Camp Dresser & McKee  Inc.
          Cambridge, Massachusetts  02142
            EPA Contract No. 68-03-3409
                  Project Officer

                   Chi-Yuan Fan
    Superfund Technology Demonstration  Division
       Risk Reduction Engineering  Laboratory
             Edison, Nev Jersey  08837
       RISK REDUCTION ENGINEERING LABORATORY
        OFFICE OF RESEARCH AND DEVELOPMENT
       U.S. ENVIRONMENTAL PROTECTION AGENCY
              CINCINNATI, OHIO  45286
                                          Printed on Recycled Paper

-------
                                DISCLAIMER
The information in this document has been funded by the U.S. Environmental
Protection Agency under Contract 68-03-3409 to COM Federal Programs
Corporation.  It has been subjected to the Agency's peer and administrative
review, and has been approved for publication.  Mention of trade names or
commercial products does not constitute endorsement or recommendation for
use.
                                     11

-------
                                 FOREWORD
Today's rapidly developing and changing technologies and industrial
products and practices frequently carry Vith them the increased generation
of materials that, if improperly dealt with, can threaten both public
health and the environment.  The U.S. Environmental Protection Agency is
charged by Congress with protecting the Nation's land, air, and water
resources.  Under a mandate of national environmental laws, the agency
strives to formulate and implement actions leading to a compatible balance
between human activities and the ability of natural resources to support
and nurture life.  These laws direct the EPA to perform research to define
our environmental problems, measure the impacts, and search for solutions.

The Risk Reduction Engineering Laboratory is responsible for planning,
implementing and managing research, development, and demonstration programs
to provide an authoritative, defensible engineering basis in support of the
policies, programs and regulations of the EPA with respect to drinking
water, wastewater, pesticides, toxic substances, solid and hazardous
wastes, and Superfund-related activities.  This publication is one of the
products of that research and provides a vital communication link between
the researcher and the user community.

An area of major concern is the health impacts associated with uncontrolled
releases of petroleum hydrocarbons from underground storage tanks.  This
document presents a structured data base of scientific information, drawn
primarily from the published literature, on the fate and transport of motor
fuel constituents in the soil/groundwater system.  It provides a strong
foundation for understanding the mobility and persistence of these
contaminants and, thus, an understanding of the potential effectiveness of
alternative remedial technologies.
                                      E. Timothy Qppelt, Director
                                      Risk Reduction Engineering Laboratory
                                    111

-------
                                 ABSTRACT
The problems associated with leakage of motor fuels and organic chemicals
from underground storage tanks (USTs) are compounded by a general lack of
understanding of the partitioning, retention, transformation, and transport
of these contaminants in the subsurface environment.  The research material
presented herein is the result of an intensive data collection and
evaluation effort.  The objective of this research was to compile the most
up to date and comprehensive knowledge of contaminant behavior in the
subsurface into a single document.  It provides an understanding of
micro-scale fate and transport processes as a means to understanding the
larger scale movement of contaminants.  This, in turn, leads to a more
thorough understanding of the application of corrective measures and
remediation techniques.  The micro-scale analysis focuses on 13 loci, each
of which represents a location and conditions in the subsurface environment
where and how contaminants may exist after an UST release.

The extensive research culminating in this report identifies the most
important "rules" governing transport, partitioning, retention, and
transformation of leaked motor fuels in the underground environment.  The
report is a data base of comprehensive scientific knowledge  that can be
drawn upon to define the key parameters and to increase the  level of
sophistication in the approach to the problem of motor fuel  leaking from
USTs.

Environmental scientists, engineers and managers of varying  levels of
technical expertise may find this technical handbook useful.  For the less
technical user, it serves to strengthen an understanding of  the fate and
transport processes vital to effective remedial response.  For the more
experienced user, it serves as a source book of information, data, and
equations to support more quantitative assessments of pollutant fate and
transport.

This work was submitted in partial fulfillment of Contract Number
68-03-3409 (Work Assignment 7) by Camp Dresser & McKee Inc.  under the
sponsorship of the U.S. Environmental Protection Agency.  This report
covers a period from 1987 to June 1991, and work was completed as of June
1991.
                                    IV

-------
                                 CONTENTS
Foreword 	   ill
Abstract	     iv
Acknowledgment 	     vi
Section
Section
Section
Section
Section
Section
Section
Section
Section
Section
Section
Section
Section
Section
1
?
•}
/,
•s
6
7
8
q
in
11
1?
13
14
— Locus
- Locus
- Locus
— Locus
- Locus
— Locus
— Locus
- Locus
- Locus
— Locus
- Locus
- Locus
— Locus
- Glossc
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
irv. ,
1 	
2 	
3 	
4 	
5 	
6 	
7 	
8 	
9 	
10 	
11 	
12 	
13 	

                                                                          10
                                                                          45
                                                                          70
                                                                         103
                                                                         126
                                                                         148
                                                                         173
                                                                         198
                                                                         229
                                                                         247
                                                                         270
                                                                         298
                                                                         324
                                                                         345
Note:  A detailed table of contents for the Introduction and Sections 1-13
       is provided at the beginning of each Section.

-------
                              ACKNOWLEDGMENT
This document was prepared for EPA's Office of Research and Development,
Risk Reduction Engineering Laboratory (RREL),  as part of Contract No.
68-03-3409.  Mr. Anthony Tafuri served as the Project Officer and Mr.
Chi-Yuan Fan served as the Technical Project Monitor.  Technical assistance
was supplied by Mr. John Farlow, Dr. John Brugger, Mr. Richard Griffith,
and Dr. Chien T. Chen of RREL.  Early work on this task was directed by Ms.
Claudia Furman of PEI Associates; the latter stages were directed by Dr.
Warren Lyman of Camp Dresser and McKee Inc. (COM).  Dr. Warren Lyman, Dr.
Myron Rosenberg and Patrick Reidy of COM, and Mr. William Thompson of PEI
are acknowledged for their technical and editorial assistance.  We also
acknowledge many helpful comments from Prof. John L. Wilson (New Mexico
Tech.) and Prof. Daniel D. Reible (Louisiana State University) who reviewed
an early version of this report.  Great appreciation is also extended to
the following PEI, COM, and Planning Resources Corp. (PRC) staff members
who conducted the necessary research and provided the technical writing and
support for the preparation of this document:

    o  Mr. Doug Bailey (PEI)
    o  Dr. Pinaki Banerjee (PRC)
    o  Mr. Jack Campion (COM)
    o  Mr. Roy Chaudet (PEI)
    o  Mr. Jim Curtis (COM)
    o  Mr. Paul Dean (PRC)
    o  Ms. Michel Kimpel (COM)
    o  Mr. Ben Levy (COM)
    o  Mr. Tom Pedersen (COM)
    o  Mr. Steve Pour (PEI)
                                    vi

-------
                               INTRODUCTION

                                 CONTENTS
                                                                      Page
Background and Purpose 	    2
Objectives of Document 	    5
Research Approach 	    5
Organization of Document	    8
                                  TABLES

1   Brief Descriptions of the thirteen physicochemical - Phase Loci     3
2   Matrix of Fate and Transport Process Considered for Each Locus      7
3   Generic Outline for Locus Sections                                  9
                                 FIGURES
    Schematic Representation of the 13 Loci in Terms of
    Unsaturated and Saturated Zones

-------
                               INTRODUCTION
BACKGROUND AND PURPOSE

The U.S. Environmental Protection Agency (EPA)  has developed a
comprehensive program for regulating certain underground storage tanks
(UST) that contain regulated substances.   With  this program comes the need
for development of guidance to assist those parties involved in complying
with regulatory requirements.   One significant  portion of the legislation
that has been developed pertains to corrective  actions for releases of
petroleum products such as gasoline and other motor fuels.  EPA estimates
that over 95 percent of the estimated 1.4 million UST systems are used to
store petroleum products.

Current regulations require that owners and operators take corrective
measures to mitigate releases  from underground  storage tanks.  To date,
guidance documents developed by the EPA and other organizations present
various technologies applicable to remediation  of the subsurface
environment.  However, only minimal amounts of  information have been
generated with regard to evaluating the movement and disposition of motor
fuel contaminants in the subsurface environment, and only recently has
significant research been devoted to this effort.  Currently, investigators
have no guidance available to determine corrective actions at UST sites
that is based upon the scientific principles governing the behavior and
degradation of motor fuel constituents in the subsurface environment.
Recognizing this, EPA sponsored the comprehensive, scientific literature
research effort resulting in this document.

As part of its philosophy, EPA developed the concept that a substance
leaking from an UST will be present in and transient between one or more
locations or settings in the subsurface environment.  A total of 13 of
these locations, referred to as physicochemical-phase loci, were identified
as the focal points for this research report.  Each of the 13 loci
represents a point in space (or location) and the physical state of the
leaked substance that together describe where and how these contaminants
may exist in the subsurface environment after an UST release.  For example,
contaminants may be dispersed as a component of soil gas, or they may be
dissolved in the water film surrounding a wet soil particle in the
unsaturated zone.  These are just two examples  of the 13  loci.
Collectively, the 13 loci represent all locations/states where and how
leaked material may be present in the subsurface environment.  The
distribution of contaminants among these loci is constantly changing, as
contaminants tend to move between different loci at varying rates over
time.  Table 1 presents a brief description of  each locus.  Figure 1
presents a schematic cross-section of the subsurface environment and
identifies where each locus may exist in terms  of the unsaturated and
saturated zones.

-------
                                  TABLE 1

                    BRIEF DESCRIPTIONS OF THE THIRTEEN
                        PHYSICOCHEMICAL-PHASE LOCI
Locus
Number                                Description
   1               Contaminant vapors as a component of soil gas in the
                   unsaturated zone.

   2               Liquid contaminants adhering to "water-dry" soil
                   particles in the unsaturated zone.

   3               Contaminants dissolved in the water film surrounding
                   soil particles in the unsaturated zone.

   4               Contaminants sorbed to "water-wet" soil particles or
                   rock surface (after migrating through the water) in
                   either the unsaturated or saturated zone.

   5               Liquid contaminants in the pore spaces between soil
                   particles in the saturated zone.

   6               Liquid contaminants in the pore spaces between soil
                   particles in the unsaturated zone.

   7               Liquid contaminants floating on the groundwater table.

   8               Contaminants dissolved in groundwater (i.e., water in
                   the saturated zone).

   9               Contaminants sorbed onto colloidal particles in water in
                   either the unsaturated or saturated zone.

  10               Contaminants that have diffused into mineral grains or
                   rocks in either  the unsaturated or saturated zone.

  11               Contaminants sorbed onto or into soil microbiota in
                   either the unsaturated or saturated zone.

  12               Contaminants dissolved in the mobile pore water of the
                   unsaturated zone.

  13               Liquid contaminants in rock fractures in either the
                   unsaturated or saturated zone.

-------
SOIL PARTICLES OR ROCK


LIQUID CONTAMINANT
(Organic Phase)

WATER WITH DISSOLVED
CONTAMINANT

SOIL AIR WITH
CONTAMINANT VAPORS
                                       CONTAMINANTS SORBED ON SOIL OR
                                       DIFFUSED INTO MINERAL GRAINS

                                       MOBILE COLLOIDAL PARTICLES
                                       WITH SORBED CONTAMINANT

                                       SOIL MICROBIOTA WITH
                                       SORBED CONTAMINANT


                                       LOCI NUMBER (SEE TABLE 1)
Figure 1.  Schematic Representation of the 13 Loci in
          Terms of Unsaturated and Saturated Zones

-------
The primary focus of this work assignment was to formulate a data base of
comprehensive scientific knowledge that can be drawn upon to define the key
parameters and to increase the level of sophistication in the approach to
the problem of motor fuel leaking from USTs.  The focus on organic
contaminants is based on the reality that over 95 percent of materials
stored in underground tanks is some type of petroleum product.  As a
starting point, EPA sponsored a two-day seminar where recognized experts
and specialists in eight specific disciplines presented information on the
physical, chemical, and biological properties and fundamentals of motor
fuel and organic chemical behavior in the subsurface environment.  Based on
the information presented at this seminar, an intensive evaluation of
available scientific data was conducted to provide a comprehensive,
up-to-date understanding of the physical and chemical "rules" governing
each locus for each of the 13 loci.

OBJECTIVES OF DOCUMENT

The EPA's Office of Underground Storage Tanks (OUST) is responsible for
establishing the Agency's regulatory program for managing underground
storage tanks.  The Risk Reduction Engineering Laboratory (RREL) of EPA's
Office of Research and Development is responsible for providing engineering
and scientific support to OUST.  One means of providing this technical
support is through the preparation of guidance materials such as handbooks,
manuals, and technical reports.  To date, there have been a number of such
documents developed by the EPA and other organizations on various aspects
of the UST corrective action process.  A review of these and other
currently available documents indicates that there was no pre-existing
guidance which directly relates fate and transport processes to corrective
actions, a shortcoming deemed serious by EPA.  Therefore, the primary
objective of this document was to provide a comprehensive, in-depth
research report with detailed coverage of fate and transport mechanisms
that also includes timely discussion of these mechanisms as they relate to
currently used or innovative approaches to UST releases.  This report, in
turn, provided the technical basis for the preparation of the desired
guidance documents recently released by RREL:

    o  ASSESSING UST CORRECTIVE ACTION TECHNOLOGIES:  Site Assessment and
       Selection of Unsaturated Zone Treatment Technologies.  Report No.
       EPA/600/2-90/011, Risk Reduction Engineering Laboratory, Cincinnati,
       OH, March 1990.

    o  ASSESSING UST CORRECTIVE ACTION TECHNOLOGIES:  Early Screening of
       Clean-up Technologies for the Saturated Zone Report.  No.
       EPA/600/2-90/027, Risk Reduction Engineering Laboratory, Cincinnati,
       OH, June 1990.

RESEARCH APPROACH

The approach for this research effort is based on the concept that a
chemical leaking from an UST will be present in and transient between one
or more of the 13 physicochemical-phase loci (locations or settings in the
subsurface environment) that were identified earlier.  These loci, as
described and presented in Table 1 and Figure 1, can be used in reference
to any type of contaminant that enters the subsurface environment.

-------
However, for the purpose of this research effort, the contaminants of
interest are petroleum product (e.g., gasoline) constituents, i.e.,
hydrocarbons and other chemicals, including common gasoline additives.  As
pointed out previously, the reason for focusing on petroleum product
constituents is that these products make up over 95 percent of the
materials managed in underground storage tanks.  All references to
contaminants in this document are, therefore, with regard to hydrocarbons
and associated organics.

Each locus, after being initially defined, was researched and evaluated in
terms of the mobilization, immobilization, transformation, and bulk
transport processes that pertain to it and in terms of how these various
processes affect the locus.  The objective was to evaluate and understand
micro-scale partitioning and transformation as a means to understanding the
larger scale movement of contaminants which will lead, in turn, to a
greater understanding of the application of corrective measures and
remediation techniques.  Table 2 is a matrix that identifies the different
processes that affect each locus.  For example, if the reader refers  to the
column "a" heading it identifies two phases (liquid contaminant water) and
a partitioning process in which they may be involved (dissolution).   The
dissolution process, when it involves air and water, appears as the heading
for column "d".  Bulk transport processes refer to advection and
dispersion.  These heading names identify only one direction of each
partitioning process, while the reverse process is inferred.  The
checkmarks that appear in each column identify those loci that may be
affected by that particular process.  Table 2 also identifies where
detailed discussions are presented on each process in terms of the loci
sections contained in this report.  Because there are many processes  that
are important to more than one locus, it was necessary to minimize
redundancy in the material presented.  This was accomplished by
identifying, up front, individual loci sections in which the detailed
process discussions would be included.  The section(s) containing detailed
discussions of a given topic would then be referenced in other sections
rather than repeat the material.  For example, if the reader refers to
Table 2, locus no. 7, column "a", a circle is shown around the checkmark.
This circled checkmark indicates that the detailed discussion of the
process involving partitioning from the liquid contaminant phase to water
and vice versa (i.e., dissolution/phase' separation) is included in the
section on locus no. 7 (Section 7).  This process is also discussed in the
sections on locus nos. 2, 3, 5, 6, 8, 12, and 13, denoted solely by a
checkmark, but in less detail than locus no. 7.

The one locus that is treated differently from others in the body of  this
report is locus no. 11 (contaminants sorbed into/onto biota).  It is
treated differently in terms of section organization and contents because
locus no. 11 is considered to be a transformation process in itself,  i.e.,
it is considered to be both a locus and a process.

Following discussion of partitioning, transformation, and transport
processes, each section provides guidance on calculating maximum and
average values for the contaminant storage capacity of the locus, including
estimation of parameter values and example calculations.  In addition,
example calculations are provided where appropriate for mass transport
processes (e.g., advection, dispersion, and diffusion) and partitioning

-------
                                  TABLE  2

                      MATRIX OF FATE AND TRANSPORT
                 PROCESSES CONSIDERED  FOR EACH  LOCUS












LOCUS
NO.


1
2
3
4
5
6
7
8
9
10
11
12
13
MOBILIZATION, IMMOBILIZATION, TRANSFORMATION,
AND BULK TRANSPORT PROCESSES
Partitioning and
Transformation Processes
obcdefgh



"e.
o
c 1
CO O
C (0
£-
O
o o
3 A
S v

(V)

V
4
V



V
V


•£•
£
5
c 5
CO CO
£ o
E >
CO *"""
0<
— *
•\j
•y

•\j





•\]



0
- £•
c o
CO (/)
i -
| o
c .
ob
It

3

(V)
fV)

(V)



(V)

^^
c
3
O
v>

Q
b.
o>
CO
A
V
-\]

3

(V)



V



—
5
§
o
52-
=
o
CO
A
V
^
i


V
3
V
7^
^^^

(V)



I
*-
c
0)
c
o
O
^^
0
en
A
V
3









c
0
lir~
•o
CO
o
O)
•o
o
5.
CO
0
ffi
A
*
V
-y
v
V
V
V
V
V
V-
V
^
V
V


c
o
a
u
o

a
u
E
«
0
A
V
•\J
A*
^
v
V
V
V
V
/
V
V
V
V

Bulk Transport Processes
i j k 1 m n o p q





c
CO
I _

C 0>
3!
P



3
V
t ^y
<^



/-—N
>T
sjx





c
CO
c
I
CO
0 5_
is

T

(V)
(V)
T
^v_X



X-T\
Vj
O'






o
c

t
3 at
CO ~
52
15
a1 c
J 3


(V)





3








= c
7 o
"Z N
S 'o
*S
TS 2
32
cr n
j CO




3














,_
"5
to
3









CO
•o
o
5
0 •-
s
2 H
I5
». £
°l
Is
coo
n
K- a.



•

oC

0>









5
£-
C
o
*V)
s
E




in
2
"5
a

^c
_o
in
•^
D
0









V
3
V
-\i
>
V


(V)
v
V





in
2
CO

£
O
To
O



(V)

(V)
T



V  " Important Process
   = Relatively Less Important
     Process
      = Detailed Discussion Provided In
        Corresponding Locus Section

( )   = All Media/Phases

-------
equilibrium processes (e.g.,  calculations for volatilization, dissolution,
and adsorption).  Lastly, the relative importance of each locus with regard
to fate and transport of contaminants is discussed in terms of remediation,
interaction with other loci,  identification of critical information gaps,
and recommendations for future research.

The ultimate objective in conducting research on the fate and transport of
hydrocarbons in the manner presented, i.e., locus by locus, and process by
process from a micro-scale perspective, is to gain a better understanding
of the basis for larger scale contaminant movement.  The ways in which the
behavior of hydrocarbons may be altered are directly related to remediation
techniques.  Therefore, a comprehensive understanding of the fundamental
"rules" of contaminant behavior engenders a better understanding of how
this behavior may be induced or prohibited to optimize a given remedial
strategy.

Even though remediation is not the subject of this report, indirectly the
ultimate purpose of conducting the loci research is to increase the
sophistication of current approaches to site assessment and corrective
action selection.  Brief discussions of remediation and corrective measures
are presented in each locus section in three places; in each introductory
subsection, in the subsections discussing effective partitioning processes,
and in the overview of each locus' relative importance presented in the
last subsection.

ORGANIZATION OF DOCUMENT

The main body of this document is organized into 13 major  locus sections.
Each locus section, except for Section 11 (locus no. 11),  consists of the
same five main subsections, which, in turn, contain discussions that are
presented similarly.  Table 3 presents a generic outline of  the locus
sections.  As previously explained, locus no. 11 is dealt  with differently
in terms of the structure of its section because of its nature, i.e., it  is
both a locus (biota) and a process (biodegradation).  Because of its
nature, the evaluation and discussion of locus no.  11 was  organized in a
different way as compared  to the other loci.  Although many  sections
deviate slightly from the standard outline, the same types of information
and evaluations have been  included in all sections.  The specific contents
of each locus section is presented at the beginning of each  section as
individual tables of contents.  The main table of  contents identifies only
the main section headings.

-------
                                  TABLE 3

                    GENERIC OUTLINE FOR LOCUS SECTIONS
                          SECTION X - LOCUS NO. X

X.I  Locus Description

     X.I.I  Short Definition
     X.I.2  Expanded Definition and Comments
            (Includes two figures; Figure X-l is schematic cross-sectional
            diagram of locus, and Figure X-2 is schematic diagram showing
            fate and transport processes and interaction with other loci)

X.2  Evaluation of Criteria for Remediation

     X.2.1  Introduction
     X.2.2  Mobilization/Remobilization

            X.2.2.1  Partitioning onto/into Mobile Phase
            X.2.2.2  Transport of/with Mobile Phase

     X.2.3  Fixation

            X.2.3.1  Partitioning onto Immobile (Stationary) Phase
            X.2.3.2  Other Fixation Approaches

     X.2.4  Transformation

            X.2.4.1  Biodegradation
            X.2.4.2  Chemical Oxidation

X.3  Storage Capacity in Locus

     X.3.1  Introduction and Basic Equations
     X.3.2  Guidance on Inputs for, and Calculation of, Maximum Value
     X.3.3  Guidance on Inputs for, and Calculation of, Average Value

X.4  Example Calculations

     X.4.1  Storage Capacity Calculations
     X.4.2  Transport Rate Calculations
            (Includes calculations of applicable mass transport rates,
            e.g., advection, diffusion, dispersion, and example
            calculations of important partitioning equilibrium process
            equations)

X.5  Summary of Relative Importance of Locus

     o  Remediation
     o  Loci Interactions
     o  Information Gaps

-------
                          SECTION 1.  LOCUS NO. 1

                                 CONTENTS
                                                                   Page No.

List of Tables	       11
List of Figures	       12
1.1  Locus Description	       13

     1.1.1  Short Definition	       13
     1.1.2  Expanded Definition and Comments	       13

1.2  Evaluation of Criteria for Remediation	       13

     1.2.1  Introduction	       13
     1.2.2  Mobilization/Remobilization	       13

            1.2.1.1  Partitioning into Mobile Phase	       13

                     Pure Chemical Vapor Pressure	       13
                     Partial Pressure	       16
                     Volatilization Rate	       19
                     Soil and Moisture Conditions	       19
                     Changes in Physicochemical Properties	       19

            1.2.2.2  Transport of the Mobile Phases	       21

                     Diffusion	       21
                     Impulse Input	       23
                     Step Input	       24
                     Tortuosity	       25
                     Advection	       26
                     Gravity-Driven Transport	       27
                     Dispersion	       28
                     Enhancement of Mobilization	       30

     1.2.3  Fixation	       30

            1.2.3.1  Partitioning onto Immobile  Phase	       30

                     Absorption  in Water	       31
                     Adsorption  on Dry Soils	       31

            1.2.3.2  Other  Fixation Approaches	       37
                                          10

-------
                                (Continued)


                                                                   Page No.

     1.2.4  Transformation	      37

            1.2.4.1  Biodegradation	      37
            1.2.4.2  Chemical Oxidation	      37

1.3  Storage Capacity in Locus	      37

     1.3.1  Introduction and Basic Equations	      37
     1.3.2  Guidance on Inputs for, and Calculation of,
              Maximum Value	      38
     1.3.3  Guidance on Inputs for, and Calculation of,
              Average Values	      39

1.4  Example Calculations	      40

     1.4.1  Storage Capacity Calculations	      40

            1.4.1.1  Maximum Quantity	      40
            1.4.1.2  Average Quant i ty	      40

     1.4.2  Transport Rate Calculations	      41

            Vapor Sorption of Benzene	      41

1.5  Summary of Relative Importance of Locus	      41

     1.5.1  Remediation	      41
     1.5.2  Loci Interaction	      42
     1.5.3  Information Gaps	      42

1.6  Literature Cited	      43
                                  TABLES
1-1A .Estimated Properties for a Synthetic Gasoline and its
      Constituents at 10°C                                            17

1-1B  Estimated Properties for a Synthetic Gasoline and
      its Constituents at 20°C                                        18

1-2   BET Monolayer Adsorption Capacities (X,,,) and
      Associated Molar Adsorption Heats (AHm) of Adsorbates
      on Dry Woodburn Soil at 20°C                                    36
                                         11

-------
                                  FIGURES

                                                                   Page No.

1-1  Schematic Cross-Sectional Diagram of Locus No.  1 -
     Contaminant Vapors as a Component of Soil Gas in the
     Unsaturated Zone                                                 14

1-2  Schematic Representation of Important Transformation
     and Transport Processes Affecting Other Loci                     15

1-3  Illustration of VOC Adsorption with three Moisture
     Regimes                                                          20

1-4  Boiling Point Distribution Curves for Samples of "Fresh"
     and "Weathered" Gasolines                                        22

1-5  Decay through Time of the Impulse Input                          24

1-6  Description of the Time-History of Diffusion Away from
     a Constant Source of Contamination                               25

1-7  Relation Between Peclet Number and Ratio of the Longitudinal
     Dispersion Coefficient and the Coefficient of Molecular
     Diffusion in a Sand of Uniform-sized Grains                      29

1-8  Schematic Diagram Shoving the Contribution of Molecular
     Diffusion and Mechanical Dispersion to the Spread of a
     Concentration Front in a Column with a Step-Function Input       30

1-9  Effects of Relative, Humidity on Soil Sorption of Organic
     Chemical Vapors                                                  32

1-10 Adsorption of Various Gases as a Function of their Normal
     Boiling Points                                                   33

1-11 Uptake of Organic Vapors and Moisture by Dry Voodburn Soil
     at 20°C vs Relative Vapor concentration                          35

1-12 Vapor Uptake of m-Dichlorobenzene and 1,2,4-Trichlorobenzene
     by Dry Woodburn Soil at 20 and 30°C                              35
                                    12

-------
                          SECTION 1 - LOCUS NO. 1


1.1  Locus Description

1.1.1  Short Definition

    Contaminant vapors as a component of soil gas in the unsaturated zone.

1.1.2  Expanded Definition and Comments

    Locus no. 1 includes soil gas with contaminant vapors anywhere in the
unsaturated zone, including the fissures of fractured bedrock.  The
contaminants in this locus are quite mobile, moving via both diffusion
through air pores (a concentration-gradient-driven process) and advection,
which may be driven by density gradients, barometric pressure pumping or
sweep flow from in situ gas generation.  Figures 1-1 and 1-2 present a
schematic cross-sectional diagram of locus no. 1 and a schematic
representation of the transformation and transport processes affecting
other loci, respectively.

1.2  Evaluation of Criteria for Remediation

1.2.1  Introduction

    The remediation of vapor-phase contaminant from locus no. 1, the soil
gas in the unsaturated zone, should be based on an understanding of the
physical and chemical processes that control the volatilization and
movement of vapor-phase contaminants.  These physical and chemical
parameters (e.g., soil air permeability, water content, vapor pressure) are
discussed in Sections 1.2.2 and 1.2.3, which discuss the effects that these
parameters have on the mobilization and fixation of volatilized liquid
contaminants in locus no. 1.

1.2.2  Mobilization/Remobilization

1.2.2.1  Partitioning into Mobile Phase

    Contaminant chemicals enter the soil gas of locus no. 1 by
volatilization of pure contaminant liquid (i.e., from loci 2, 6, 7, and
10).  The process of volatilization is discussed in detail in locus no. 7.
Contaminants may also enter locus no. 1 by volatilization from contaminated
water (locus no. 3).  This process is discussed in detail in locus no. 3.

Pure Chemical Vapor Pressure

    An important factor that determines the degree to which a chemical
volatilizes is its pure chemical vapor pressure, P°.  Tables 1-1A and 1-1B
show values of vapor pressure for several hydrocarbons comprising a
hypothetical, 23-component gasoline.  The C,-Cfi alkanes (e.g., isobutane,
n-butane, and isopentane) have the highest vapor pressure of any of the
constituents of gasoline.  The BTEX compounds (benzene, toluene,
ethylbenzene, and xylenes), normally regarded as volatile, have quite low
vapor pressures relative to other gasoline constituents.
                                    13

-------
 NOTE: NOT ALL PHASE BOUNDARIES ARE SHOWN.
                                 LEGEND
              CONTAMINANT VAPORS
            J AIR (PLAIN WHfTE AREAS)
              WATER (MOBILE PORE WATER
              IN UNSATURATED ZONE)
       AAAA  WATER FILM
X  BIOTA (MICROBKDTA INCLUDED)

   SOIL PARTICLE

 I COLLOIDAL PARTICLE
Figure 1-1. Schematic Cross-Sectional Diagram of Locus No. 1 - Contaminant
           Vapors as a Component of Soil Gas in the Unsaturated Zone.
                                 14

-------
                     TRANSPORT
                     RETARDATION
                     DISPERSION
DIFFUSION
IN AIR
      (BIOTA)
    LOCUS NO
11 - SORBED TO BIOTA
     VOLATILIZATION
                UPTAKE
        (ROCK)
       LOCUS NO
    10-DIFFUSED INTO
       MINERAL GRAINS
       OR ROCKS
                               LOCUS NO. 1
                           (CONTAMINANT VAPORS AS A
                           COMPONENT OF SOIL GAS IN
                             UNSATURATED ZONE)
        (LIQUID CONTAMINANTS)
             LOCUS NO.
        2- ADHERING TO •WATER-
          DRY' PARTICLES IN UN-
          SATURATED ZONE
        6-IN PORE SPACES IN UN-
          SATURATED ZONE
        7-FLOATING ON WATER
          TABLE
       13- IN ROCK FRACTURES
                                                     CONDENSATION
                                                    DISSOLUTION
               DIFFUSION
  PHASE
  SEPARATION
         (WATER)
        LOCUS NO.
 3 - DISSOLVED IN WATER FILM
 8 - DISSOLVED IN GROUNDWATER
12- DISSOLVED IN MOBILE
   PORE WATER
Figure 1-2. Schematic Representation of Important Transformation
            and Transport Processes Affecting Other Loci.
                                  15

-------
    Tables 1-1A and 1-1B show a range of physical and chemical properties
for a hypothetical gasoline developed by COM (1987).  This gasoline mixture
is based on data from several sources in the literature, and is an attempt
to define the constituents in gasoline to allow for estimation of various
properties of the gasoline mixture.  The chemicals comprising the gasoline
blend were chosen to represent classes of chemicals typically found in
gasoline.  Below each table are listed calculated properties for the bulk
liquid.  Table 1-1A is calculated at 10°C and Table 1-1B is calculated at
20°C; comparing the two tables shows how temperature influences the
properties.

    Vapor pressure is strongly temperature-dependent.  Tables 1-1A and 1-1B
show that the vapor pressure of isobutane, for example, rises from 1648 mm
Hg to 2253 mm Hg as the temperature changes from 10°C to 20°C.  Chemical
engineering literature (e.g., Perry and Chilton, 1973) contains several
pressure-temperature correlations.  The Antoine equation is usually used,
but correlations by Riedel, Kay, Reid, Miller, and Sherwood are also
discussed by Perry and Chilton (1973).

Partial Pressure

    For any mixture of compounds, the pressure exerted on the air by any
one chemical component equals the product of the pure chemical vapor
pressure and its liquid mole fraction:

                                P. = x..P.0                           (1.1)


    where   P.  =  partial pressure of contaminant i (mm Hg)


            x.  =  mole fraction of a contaminant i (dimensionless)

                   0 < X   < 1
            P.° =  vapor pressure of pure contaminant i (mm Hg)


    The total vapor pressure of a mixture is the sum of the partial
pressures of the chemicals comprising the mixture:


                                     N

                                P«. = Z  P<                             (1.2)
    where   P   =  Total pressure exerted on air by contaminant liquid

                   (mm Hg)
                                    16

-------
 II
o2fi-
*ii*
 ill
  -'-
 *1»!^

 ill
 *?s
 ill
 IP

 ii
^> CM Op **• 9 •

SS3RK
     r^*Mor**»»r»ir»r*«i*>»o^f*-o^ — »->i
     ^ •* C^- OB ^ •*  F^ ^ ^" "M P^ >M —*
     iisssissilsssssissssssl
     .w s .


     llll
                  .U.
                       il  i
                             «f
  ^ CD

  « "E

  8|
                                  Q. «

                                  •o £
                                 s <* »
                                 AS".

                                 alt

                                 I S «
                               ss •
                               Isg
                                          fe i-
                                    £ w-SS &£ .
                                    w -•* o ^ ••*• ^ *•

                                    I * "i"5 ""
                           -   S
                           ±* s
                           5 s r
                           m o *  ^-  —•  —
                           •ts r  s  s  s
                           a   s  1  s  a

                           ...a&^ls**;
                                    *i


                           |2|  »-fs-»l:
                         ^  II"  =i^E^=^
                         °  11=  i^iifi^£
.  3IP1F^R
!? _ •? fe - >B -3
                   17

-------
fi
S
             * 2?    * ooo»r*> — r«w»'>aDo*o^in*0P4W<«o»nt»»A<>*'~
                     * f-r-.-«-«-*r--r^«'>««r'»«««t«»o-«»«5'^-o5»-w
            ,r
          Q •• IE *^ * »^»o^-^»c<*^^>^i«r-c*«^o-<»—'-or^.^-ir»^-^-— ai*-»
                          §*r c^ &~ o •• — • -™ n ^ ^* <*«i «•< o «*" *^ ^ »• *•» ^ ••* ***
                          3»»-A**r«««.3c>*7 — o»S»^>*si5--«"5r^-*
                       »S5»s         aR"R2 = g:   ss      r
             8 =
P?

III
§11
s~
ill
i'P
            I
               §

              ii
                          R*
                                  *rO-^'O4C>a>«OlOO'V7O*tf>O
                                  — e.r-*oo-o^--*i^->-or^-n  o — ao  K> r-> ao
                                  — -o •«• m .n .0       r*>
                                     -
• ^ *^ w> ** ^ m •— IB r**> ** *v ^ n --^ -o o*1 ^ i

•04-0   »o^»    m ^ r^ —- **4 — —    o
• r* -*                                 CM
                          *^c*r— oot^.r^^«*»^- — r^-of— -
                       SSSSS5S3SSSSS335S3SSS33
                                       *-—>-'--
                                       trt-OO-CIJli^ — »-O-« — f>4 — -^>O — »
                       r*ir<«w^m»^ — 9-f^r^>^--a»-r>«-4^^


                       ODa>rMf>4^o(»r4<«^«-e^«i>^.«c>a>o4w-Voc>

                       (0
                          (B  .
                                   *l         <•  K      —.»»!*    —

                                   S      S 2 ^      ^ S £    f
                                   •*rf      * ^ ^- •»   •*-> ^ .O    *rf .
                                      S*    MO>»C    «» -fcJ —    * ^
                                      e    «—'£•*    v •* >~   •
                              •    &w*».«ks.MM •§•«»— j;    — .
                              —  *»*e-^>-«tcc*- i  **  » w
                                                                                        3
                                                                                     II

                                                                                      it
                                                                                      m Q-
                                                                                        .

                                                                                      Q. 00



                                                                                     II
                                                                                                   S         S    S   S
                                                                           o    "5  5



                                                                                 s-™3
                                                                                                                           S
                                                                                             •srs    ::•.-•.
                                                                                        ,    11J    5^  s-^ g
                                                          18

-------
    At 20°C, the hypothetical gasoline exerts a pressure of 274 mm Hg on
the air above the contaminant liquid, or 0.36 atmospheres (1 atm « 760 mm
Hg).  This may be expressed as 360,000 ppm on a volume per volume basis.

Volatilization Rate

    The rate at which contaminants volatilize from the liquid phase is
controlled by several physical and chemical parameters.  Perhaps most
important is the pure chemical vapor pressure.  As described previously,
this parameter is strongly temperature-dependent, and the rate of
volatilization is much greater at higher temperatures.  Soil temperature is
fairly constant at any location throughout the year, but it varies from
site to site (See Figure 8-5).  Other important factors that affect
volatilization in soil are described belov.

Soil Moisture Content

    The soil moisture content has two important effects on the
volatilization of liquid product into the soil pores.  Higher soil moisture
contents reduce the air-filled pore space, which reduces volatilization
(Farmer et al., 1980; Aurelius and Brown, 1987).  The volatilization rate
is also related to the water content through sorptive behavior.  Lighty et
al. (1988) and Houston et al. (1989) show that soil electrostatic force is
greater for drier soils; as water content increases, water molecules
displace contaminant molecules, and sorption decreases, increasing
volatilization.  Thus a plot of sorption vs. soil moisture content should
show a maximum in sorption at some point.  Figure 1-3 shows schematically
the effect of soil moisture content on volatilization.

Changes in Physicochemical Properties ("Weathering")

    Any product released into the subsurface environment is continually
changing in nature.  These changes, collectively referred to as
"weathering", cause the remaining liquid contaminant to have different
physical and chemical properties from fresh, or non-weathered,  product.
For example, when gasoline (a mixture of 200 or more compounds, each with
varying vapor pressures) is released into the soil, the most volatile
constituents are most likely to volatilize into the soil vapor and move
away from the residual liquid.  The constituents with the greatest
solubilities are the most likely to dissolve into infiltrating
precipitation and move toward the water table.  The most degradable
constituents are the most likely to be degraded.  The effect of these
changes is that the remaining liquid contaminant will have significantly
different properties than the unweathered product.  Johnson et al. (1989)
present the following properties for "fresh" and "weathered" gasoline:

                                       Fresh     Weathered

         Molecular Weight (g/mole)      95        111

         Vapor Pressure (atm)            0.34       0.049

         Vapor Density (mg/1)         1300        220
                                    19

-------
 VAPOR

 PHASE
0
             0
                    NON POLAR ORGANIC
                    H,0
ADSORBED f]  fI

UYER  LJnLJ
 0     o

nnnnrrn
             iLID
  a) DRY
     0
       .Q
    0   0
     0
                     o

  LJnLJnnooLJLJo
  b) DAMP
    n  °  °  ° rfn-
    U00°0o uuo
      iUD""~o	rro	'ou'
  c) WET




 Source: After Reible, 1989



 Figure 1-3. Illustration of VOC Adsorption with Three Moisture Regimes.
              20

-------
    Figure 1-4 shows boiling point distribution curves for these two
gasolines.  From the above property data and Figure 1-4, it is apparent
that the degree of weathering will be an important factor in the volatility
of a released product.

1.2.2.2  Transport of the Mobile Phase

    The mobilization/remobilization of the soil gas and vapor phase
contaminant is controlled by physical properties and conditions of the
soil, volatilized contaminant, and other immiscible phases present in locus
no. 1.  Three immiscible phases are considered — soil air, liquid
contaminant, and soil water — which are in direct physical contact with
one another.  The sections below discuss the transport processes that move
volatilized material through the soil air.

Diffusion

    Diffusion is the movement of molecules from an area of high
concentration to one of low concentration.  The transport of mass in a
direction of decreasing concentration is termed flux.  Flux is proportional
to the concentration gradient (the concentration difference divided by the
distance); the higher the concentration gradient, the stronger the flux.
This is stated in Pick's First Law (here shown for one-dimensional flux):
                               J = D    o-Cn                         (1.3)
                                    0   ~~
                                                             2
    where   J   =  mass flux of volatilized contaminant (g/cm  sec)
                                                2
            D   =  air diffusion coefficient (cm /sec)
            C   =  concentration of volatilized contaminant at X  (g/cm )
            C.,  =  concentration of volatilized contaminant at X- (g/cm )
            Ax  =  distance between X  and X1 (cm)


    Tables 1-1A and 1-1B list values for the air diffusion coefficients of
the constituents of the synthetic gasoline described previously.  Values
for D  for other compounds are available in the literature.

    Examination of Pick's First Law shows that a flux will continue to
occur until the concentration equilibrates (i.e., until the gradient is
zero).  Thus volatilization will cease when the vapor phase concentration
reaches the saturation value of the air in contact with the liquid
contaminant.  This law also shows that the rate of volatilization will be
at a maximum when the vapor concentration at X.. is zero.  Remediation by
soil venting operates by removing contaminant-saturated vapors and
replacing them by clean air, which helps to maximize the rate of
volatilization.
                                    21

-------
Cumulative
  Weight    08.
  Fraction
                                               "Weathered" Gasoline
               0.0
                 -40     0     40     80     120    160    200    240

                                    Tb(°C)
    Source: Johnson et al., 1990. (Copyright Water Well Journal Publishing Co., 1990.
           Reprinted with permission.)
   Figure 1-4. Boiling Point Distribution Curves for Samples of "Fresh"
                     and "Weathered" Gasolines.
                                 22

-------
    Pick's Second Law describes the change in concentration of volatilized
contaminant in the soil air with time at a given location.  Expressed in
one-dimensional form:
                               dC      . dC                          (1.4)
                               dt  ~  o  .2
                                         dx


    where   x  = distance (cm)
                                                               3
            C  = concentration of volatilized contaminant (g/cm )
                                              o
            D  = air diffusion coefficient (cm /sec)

            t  = time (sec)


    Analytical, closed-form solutions exist for these equations.  The form
of these solutions depends on the initial and boundary conditions on the
system which the equation describes.  An example of an initial condition is
the specification of zero concentration in the system prior to initiation
of diffusion.  A boundary condition might require that a certain location
in the system be always at some fixed concentration.  For example, the
ground surface could be set to equal zero concentration.  Two simple
analytical solutions to the one-dimensional form of Pick's Second Law are
presented below.

Impulse Input

    Under this solution, a slug of contaminant enters the system
instantaneously at one point and begins to diffuse.  No further mass is
added after the initial slug.   It is assumed that there is no background
contamination present in the system when the slug is added.  The change in
concentration with time and location is described by an exponential decay
function:
                          C(x,t) = CQ exp ( -x^                      (1.5)
UD  tj
                                                               3
    where   C(x,t) = concentration at point x and time t (mg/cm )
                                                      3
            C      = concentration at time zero (mg/cm )

            x      = distance from the source (cm)


Figure 1-5 shows graphically the effects of diffusion upon a step input.
                                    23

-------
                                   n
          -X
         Source: Fischer etal.. 1979. (Copyright Academic Press, Inc. 1979. Reprinted with permission.)
             Figure 1-5. Decay Through Time of the Impulse Input
Step Input

    A second  solution for Pick's Second  Law is for a continuous  source of
constant concentration (a "step input").   For this case, the  initial
condition is  zero concentration everywhere at time t=0.  The  boundary
conditions  are C(°°,  t)=0 for x=°° and  all t, and C(0,t)=C  for x=0 and all
t>0.  For these conditions, the solution to Pick's Second Law is:
                        C(x,t) = C   erfc
  x2  1
l4D~tj
(1.6)
    where    erfc ( ) = complementary  error function on the variable in ( )
                         erf c (x) =  1 -
                                          .V
                                           n=0
                                                n! (2n+l)
    Figure  1-6 shows the time history of diffusion away from  a  constant
source  of contamination.  With  time,  the concentration C will be found
throughout  the immediate area of  the  source,  an area that  increases with

-------
time.  The  air diffusion coefficient, D  ,  is  a key variable that determines
how quickly the concentration diffuses away from the source. A large
diffusion coefficient would lead to a greater rate of diffusion away  from
the source.
     o
    o
    o
      1-
                t increasing
   Source: Fischer£LaL 1979. (Copyright Academic Press, Inc. 1979. Reprinted with permission.)
             Figure 1-6. Description of the Time-History of Diffusion Away
                      From a Constant Source of Contamination
    Tortuosity
    The air  diffusion coefficient, D  , should  be adjusted to reflect
subsurface conditions.   For flow through  a  porous medium such as soil,  the
actual path  length that the molecules diffuse  across is dependent on  the
tortuosity of  the medium.   For highly tortuous paths,  the actual length
traveled by  the  molecules  is greater.  The  more tortuous the path,  the
longer it takes  for a molecule to Biffuse through the  soil, effectively
reducing the air diffusion coefficient.

    The air  diffusion coefficient may be  adjusted by a tortuosity factor.
Various expressions formulated for this factor have been compiled by  Weeks
et al.t 1982.  A commonly-used expression is presented by Millington  and
Quirk (1961):
                                     25

-------
            and
                                                                      (1.7)
                                et = ea + ew                          (1.8)
                                                        o
    where   D  = effective air diffusion coefficient (cm /sec)
                                                      2
            D  = diffusion coefficient in pure air (cm /sec)

            6  = air-filled porosity of soil
             cl
            6  = total porosity of soil

            9  = Water-filled porosity


    This expression shows that the effective diffusion coefficient is
substantially less than the air-diffusion coefficient even for dry soils.
For example, if the following conditions were present for a dry, sandy
soil:

            et = 0.35

            6  = 0.30
                        2
            D  = 0.07 cm /sec
             0          2
    then    D  = 0.01 cm /sec

    For the same soil with a higher water content ( e.g., 6  = 0.2) and a
lower air-filled porosity (9  = 0.15), the effective diffusion coefficient
is only about 0.015 D .  Obviously, then, diffusion through soils with
significant water content will be greatly retarded.

Advection

    Advection is the movement of the soil gas in response to a pressure
gradient exerted on the soil gas.  This pressure gradient may be caused by
barometric pressure fluctuations, artificially-induced gradients (for
example, by vacuum extraction), gravity, water table fluctuations, or by
wind-induced circulation.  Gravity is discussed in the follow sub-section.

    Advection refers to the average speed of the bulk soil gas as it moves
through the soil pores.  The velocity of the soil gas is described by
Darcy's Law:
                                    * (Pa + Pagh)/6                    (1.9)
                      9a      u   dx   a    a     a
                       a       a
                                    26

-------
    where   V    = gas velocity (cm/sec)
             X
            q    = volume flow rate per unit area (cm/sec)
                                            2
            k    = soil air permeability  (cm )
             cl
            y    = viscosity of air (g/cm sec)
             3.
            d/dx = differential along x direction
                                            2
            P    = pressure of air (g/cm  sec )
             a                         3
            p    = density of air (g/cm )
             a                                    2
            g    = acceleration of gravity (cm/sec )

            h    = elevation (cm)

            Q    = air-filled porosity of soil
             3.

    This form of Darcy's law accounts for both pressure-driven advection
(e.g., due to barometric "pumping" or changes in the barometric pressure or
from applied pressure gradients) and from gravity (discussed below).  An
examination of equation 1.9 shows that the air permeability is of great
importance in determining the vapor flow.  The air permeability depends on
several factors including the grain size  distribution, type and structure,
porosity, and water content of the soil.

    Advection and diffusion can be combined in the advection-diffusion
equation, expressed here for three-dimensional flow:

                          dC _ v^DC - WC                            (1.10)
                          dt ~

where V is the average linear pore velocity and C is the solute
concentration and 7 is the three dimensional gradient operator.


    This equation must be solved numerically.  Analytical solutions  to the
one-dimensional advection-diffusion equation exist that are similar  in form
to those presented in the diffusion subsection:
impulse-input: C(x,t) = C  exp  fx - V _t
                         O   r  	X—
(1.11)
step-input: C(x,t) = C  erfc  fx - V..t]                                (1.12)
                      \j       "~^~^^^^^~ ^~
Gravity-Driven Transport

    As stated above, gravity can cause  the downward migration of  soil gas
if the gas is denser than air.  The vapor density of pure, dry air  is about
1200 g/m  at 20°C.  The volatilization  of the chemicals  from the  synthetic
gasoline increases the vapor density of the soil gas, making it heavier
than pure air, as demonstrated by  the calculation of the vapor density of
the gasoline vapor-soil air mixture:

                                    27

-------
Pv -
                                     P M
                                     _§_
                                     RT
                                                                       3
    where   p  = vapor density of gasoline vapor-soil air mixture (gm/m )

            P  = atmospheric pressure (mm Hg)
             ct
            M  = average molecular weight of gasoline vapor-soil air

                 mixture (gm/mol)
                                                            3
            R  = universal gas constant, R = 0.06236 mm Hg m /mol K)

            T  = temperature (K)



    and

            M - M(P/P) + M(P  - P)/P                            (l.U)
    where   M  = average molecular weight of air, M  =28.96 g/mol
             d.                                     SL
            M  = average molecular weight of gasoline vapors, M  « 69.48

                 g/mol

            P  = equilibrium vapor pressure of liquid gasoline exerted on

                 soil air, Pt = 274 mm Hg (@ 20°C.  See Table 1-1B.)


    Therefore, (69.48 g/mol . 274 mm Hg/760 mm Hg) + (28.96 g/mol).

               (760 mm Hg - 274 mm Hg)/760 mm Hg)
               = 43.58 g/mol

            p  = 43.58 g/mol .  760 mm Hg/(293.15 K 0.06236 mm

                 Hg m3/mol .K)
                         3
            py = 1800 g/m



    This sample calculation (with values drawn from Table 1-1A) indicates
that the gasoline vapor-soil air mixture is half again as heavy as air.
Gravity will thus cause the downward migration of the contaminated soil
air.  The importance of gravity relative to advection and diffusion in the
transport of vapors in the unsaturated zone is uncertain and repie'jcut.s a
current area of much research.


Dispersion


    As the contaminated soil gas moves through the soil pores it mixes with
pure soil air, causing a spreading and reduction of the contaminant soil
gas concentration.  This mixing, known as hydrodynamic dispersion, is
composed of molecular diffusion and mechanical dispersion:
                                    28

-------
        D, =
                                       v + D*
(1.15)
                                                               2
    where   D, = coefficient of hydrodynamic dispersion (cm /sec)

             *  = characteristic mixing length (cm)


             v  = average flow velocity (cm/sec)
                                                            2
             D* = effective air diffusion coefficient  (cm /sec)
    The relative contribution of each  process depends  on the flow
conditions of the soil gas.  Figure  1-7 depicts the  ratio of mechanical
dispersion to molecular diffusion with the Peclet number.  The Peclet
number is a dimensionless number and is defined as follows:
                                  P =  v  .  d/D*                           (1.16)

    where   P  = Peclet number (dimensionless)

             v  = average flow velocity (cm/sec)
             d  = average pore diameter (cm)                    „
             D* = effective molecular diffusion coefficient (cm /sec)

    At  low velocities  (i.e., low Peclet  numbers) molecular diffusion
dominates over mechanical dispersion;  higher velocities (i.e., higher
Peclet  numbers) indicate that mechanical dispersion  dominates over
diffusion.
               100
                10
           *
           Q
               0.1
                   D* = Coefficient of diffusion
                   D; = Coefficient of dispersion
                    v = Average linear velocity
.y o
3-E   I
                              Transition
                           '   conditions
                           I
                                             Mechanical
                                              dispersion
                                              dominates
                                  vd/D*
Source: Freeze and Cherry, 1979. (Copyright Prentice-Hall Inc., 1979. Reproduced with permission.)

Figure 1-7.  Relation Between Peclet Number and Ratio of the Longitudinal Dispersion Coefficient
          and the Coefficient of Molecular Diffusion In a Sand of Uniform-sized Grains.
                                       29

-------
The hydrodynamic  dispersion causes some of the contaminant  concentration to
arrive before  and after the average concentration  traveling with the
average flow velocity.   This spreading about  the average velocity is shown
in Figure  1-8.
             o
> i: o
               o
                                                 f
v position of input
water at timet
                   0.5
                    0
                              Tracer front if
                              diffusion only
Dispersed
tracer front
                              Distance x
 Source: Freeze and Cherry, 1979. (Copyright Prentice-Hill Inc., 1979. Reproduced whh permission.)
 Figure 1-8. Schematic Diagram Showing the Contribution of Molecular Diffusion and Mechanical
          Dispersion to the Spread of  a Concentration Front in a Column with a Slap-Function Input


    Dispersion properties of a porous medium are  usually measured in field
or laboratory studies.  Much current research  focuses on the determination
of the  characteristic mixing length, 
-------
Absorption in Water (Details in Section 3)

    The process of absorption into water (or dissolution) is the reverse of
volatilization from water, which process is described in detail in Section
3.  The importance or degree of absorption in water is proportional to the
water solubility of the vapors and the amount of water present.  Although
volatilization and absorption at the air/water interface is rapid, the
overall rate of the processes may be controlled by the rate of diffusion of
solutes in water, a much slower process.

Adsorption on Dry Soils (Details in Section 2)

    Adsorption onto dry soils is also considered to be a rapid process
(Brunauer, 1945).  Thus, any immobilization or fixation due to sorption, is
likely temporary and reversible unless diffusion into mineral grains or
rocks (locus no. 10) or biodegradation (locus no. 11) is a significant
process.

    Water vapor and contaminant vapors compete for the same adsorption
sites on soils.  In general, water has the ability to dramatically reduce a
soils' sorption capacity for vapors of organic chemicals.  Water typically
will preferentially occupy sorption sites or displace previously adsorbed
organic molecules.  Figure 1-9 shows the effect that increasing the
relative water vapor humidity has on the soil sorption of benzene,
dichlorobenzene and 1,2,4-trichlorobenzene.  In the regions of low
humidity, the adsorption isotherms are typically non-linear and reflect
sorption - many molecular layers thick - on the soil surface.  At high
humidity, however, the sorption isotherms tend to be linear, as the
contaminant vapors dissolve into water rather than sorb to soil.

    Adsorption of Gases.  A significant distinction exists between the
adsorption of gases (temperature above critical point) and vapors
(temperature below critical point).  For gases, it is generally believed
that only monomolecular layers can be sorbed onto the solid surface (Grain,
1987).  Above the critical point of a gas, no further condensation from the
gas to the liquid phase is possible.

    A theoretical treatment of gas adsorption shows the amount adsorbed to
be proportional to the vapor concentration (Grain, 1987):

                                       K (P/PQ)                      (1.17)
                                x __ = 	
                                x    1 + K (P/P )
                                 m             o

    where   x  = mass of gas adsorbed per weight of solid adsorbent (mg/kg)

            x  = mass of gas adsorbed with a monolayer of coverage on

                 adsorbent (mg/kg)

            K  = a temperature-dependent empirical constant (dimensionless)
                                    31

-------
       40r-
             Bensene
             0.2    0.4   0.6   0.8    1.0
          Relative Vapor Concentration. P p*
•) Vapor Uptake of Benzene by Dry Woodburn Soil
   at 20 °C aa a Function of Relative Humidity.
                  0.2   0.4    0.6   0.8   1.0
               Relative Vapor Concentration, P/P«
b) Vapor Uptake of m-Dlchlorobenzene by Dry Woodburn Soli
  at 20? C as a Function of Relative Humidity.
                               "0     0.2   0.4   0.6   0.8   1.0
                                 Relative Vapor Concentration, P/P°


                  c) Vapor Uptake of 1,2,4-Trichlorobenzene by Dry Woodburn Soil
                     at 20°C as a Function of Relative Humidity.
Source: Chiou and Shoup, 1985 (Copyright American Chemical Society. Reprinted with permission)

Figure 1-9.    Effect of Relative Humidity on Soil Sorption of Organic Chemical Vapors
                                                32

-------
             P  = vapor concentration of gas  (atm)

             P  = saturation vapor pressure of gas (atm)


    There are many  empirical  correlations for gas adsorption on solids
(Brunauer,  1945).   One set of data showing the expected  correlation between
boiling  point (a vapor pressure  datum) and the amount  adsorbed on  charcoal
is shown in Figure  1-10.
              1000
              100
           !
           X
               10
                                                          so,
                                                       ci,
                                                         NH,
                                        Temperature: 15°C
                                        Adsorbent: Charcoal
                 0              100              200
                                  Boiling Point. Tb (K)


  Source: Grain, 1987. (Copyright Pergamon Press Inc. 1987. Reprinted with permission.)

  Figure 1-10. Adsorption of Various Gases as a Function of Their Normal Boiling Points.
                                       33

-------
    Many substances stored in USTs, such as gasoline, contain compounds
that, in their pure state, are gases at ambient temperatures.  Examples
include n-butane and isobutane (boiling points of -0.5°C and -11.9°C,
respectively).  These highly volatile constituents sorb to soils much less
than do compounds that have higher boiling points (which are vapors at
ambient temperatures), due to the limitations discussed above regarding
monomolecular layers.  The lover sorption of the gases would result in
higher mobility of these compounds.  For gases, the extent of adsorption is
likely to be directly controlled by the available surface area on the
soils.
    Adsorption of Vapors.  The vapors of compounds that are liquids at
ambient temperature can sorb onto soils in a process similar to
condensation.  The mass of vapors sorbed is not directly related to the
surface area of the soil because vapors can sorb many layers thick.  The
sorption of vapors is limited by the accessible pore volume of the soil.

    Sorption of vapors on soils is commonly analyzed via the
Bruanuer-Emmett-Teller (BET) isotherm equation for multilayer adsorption:
                               c (P/PQ)
(1.18)
                                         (P/PQ)]
    where   x, x , P and P  are as defined in equation 1.17
            c = temperature constant
    The constant c is related to the net molar heat of adsorption at x < x
by:
                 -In c = (AH  + AH )/RT
                            m     v'
(1.19)
    where   AH  = molar heat of adsorption (kcal/mol)

            AH  = molar heat of vaporization (kcal/mol)
            R   = gas constant, 0.001987 kcal/deg mol
            T   = temperature (K)

    Examples of sorption isotherms for several chemicals are shown in
Figure 1-11.  The temperature effect on vapor sorption for two chemicals is
shown in Figure 1-12.  This negligible temperature effect is evidence that
the heats of mineral adsorption (AH ) and vapor condensation (AH ) are
essentially the same.  Values of x , AHy and (AHm + AHy) derivedvfrom these
data are shown in Table 1-2.

    Relatively few data are available showing the extent of soil sorption
of organic vapors.  Publications that contain data include Chiou and Shoup
(1985), Isaacson and Sawheney (1984), Jurinak (1957) and Call (1957).
Sorption of organic vapors onto solid particles in the atmosphere, a
                                    34

-------
                               • Benzene
                               • Chlorobenzene
                                 p-Dichlorobenzene
                               o m-Oichlorobenzene
                               A 1.2.4-TrichloroMfuene
                          = 30ho Water (23.8 °C>
                          o
                             W0   0.2   0.4    0.6   0.8   1.0
                               Relative Vapor Concentration, P/P°
   Source:  Chiou and Shoup, 1985 (Copyright American Chemical Society. Reprinted with permission)

   Figure  1-11.   Uptake of Organic Vapors and Moisture by Dry Woodburn Soil at 20° C
                  vs.  Relative Vapor Concentration.
                             40r
                                m-Oichlorobenzene
                                 020°C  »30°C
                                1,2,4-Tnchlorobenzene
                                 B20°C  «30°C
                           ~ 30t-
                              0    0.2   0.4   0.6   0.8    1.0
                                Relative Vapor Concentration, P/P°
Source: Chiou and Shoup, 1985 (Copyright American Chemical Society. Reprinted with permission)

Figure 1-12.   Vapor Uptake of m-Dichlorobenzene and 1,2,4-Trichlorobenzene by Dry
               Woodburn Soil at 20 and 30° C.
                                            35

-------
                                 TABLE 1-2
                 BET MONOLAYER ADSORPTION CAPACITIES (x )
                                                       ffl

              AND ASSOCIATED MOLAR ADSORPTION HEATS (AH ) OF


                 ADSORBATES ON DRY WOODBURN SOIL AT 20°Ca
                                xm
Compound                  (mg/g of soil)    (kcal/mol)      (kcal/mol)
Benzene
Chlorobenzene
m-Di chlorobenzene
p-Di chlorobenzene
1,2, 4-Trichlorobenzene
Water (23.8°C)
5.57
7.53
7.42
5.54
9.53
11.7
8.09
9.43
10.6
15.0
11.7
10.3
1.52
1.82
1.86
2.54
1.93
2.14
a.  See equations 1.18 and 1.19 for definition of terms.


b.  Molar heats of vaporization determined from vapor pressure data.


Source:  Chiou and Shoup, 1985.
                                    36

-------
closely related process, is described in publications by Bidleman et al.
(1986) and Cantrell et al. (1986).

    The above discussion has indicated that modest soil sorption of organic
vapors is possible, but that the presence or introduction of water vapor
can lead to displacement of the sorbed vapors.  Thus, the temporary
fixation of organic vapors by soil sorption is enhanced by keeping the
soils dry.  Lowering the temperature would also increase the amount sorbed
to soils.

1.2.3.2  Other Fixation Approaches

    Other means of at least partially immobilizing vapors in soils include
the blocking of soil pores (e.g., with water or other liquid agents) and
the use of impermeable covers to prevent transport to the atmosphere.

1.2.4  Transformation

1.2.4.1  Biodegradation (Details in Section 11)

    Biodegradation in the subsurface occurs as soil microorganisms (mainly
bacteria) metabolize contaminants.  Oxygen, mineral nitrogen, and
micronutrients are required in addition to the contaminants themselves.
The optimum environment for this process is aqueous.

    The role of soil gas as a transport medium of contaminants to microbes
in liquid or solid phase loci may be more important than its role as a
locus for microbial biodegradation.

1.2.4.2  Chemical Oxidation (Details in Section 3.2.4.2)

    Gaseous components of petroleum hydrocarbons are subject to
photo-oxidation when exposed to sunlight or within environments where free
radicals exist.  Within the soil environment,  the oxidation of gaseous
components is not expected to be significant.

1.3  Storage Capacity in Locus

1.3.1  Introduction and Basic Equations

    The mass per unit volume of contaminant liquid volatilized into the
vadose zone air may be calculated from:

                          my = Co9a .  py                             (1.20)

                                                          3
    where  m   = mass of contaminant per unit  volume (kg/m )

            C  = equilibrium vapor concentration of contaminant
                               33
                 in soil air (m /m )

            0  = air-filled porosity (fraction)
             3                                  3
            p  = contaminant vapor density (kg/m )
                                    37

-------
    The air-filled porosity, 6 ,  is simply the difference between the total
                              a
and water-filled porosity:
                          ea . et - ev                               <


    where   9  = total soil porosity, (fraction)
            9  = water-filled porosity, _(fraction)

1.3.2  Guidance on Inputs for, and Calculations of, Maximum Value

    The quantity of contaminant vapor is at a maximum when the air-filled
porosity is at a maximum, or in other words, the water content is at a
minimum.  For this calculation, it is assumed that the water content in the
soil is at the wilting point of the soil.  The moisture content at the
wilting point varies with the soil type, as shown in Figure 12-4.

Guidance on parameter values is as follows:

    Total Porosity, 9 .

    o  Theoretical maximum porosity for1 sandy soil is 47.6% (thus, 9  =
       0.476).

    o  Table 12-2 lists representative total porosity values for a variety
       of soils.

    o  Figure 12-4 shows total porosity in graphical format.

    Water-Filled Porosity, 9y.

    o  Water-filled porosity can range from 0.05 (sand) to 9 .

    o  Figure 12-4 shows wilting point moisture content values for various
       soils.

    Equilibrium Vapor Concentration C .

    o  User should supply equilibrium vapor concentration of contaminant.

    o  For synthetic gasoline in locus no. 1, recommend:
                                                  3  3
              Temperature (°C)         C  range (m /m )

                      0                  0.11 - 0.28

                     10                  0.17 - 0.42

                     20                  0.24 - 0.61
                                    38

-------
    o  User should have information on contaminant composition, choosing
       high C  value for mixtures with high fraction of C,-C, alkane
             o                                           4  o
       compounds.

    Vapor Density, p .

    o  User should supply specific information on vapor density for
       contaminant of interest.

    o  Vapor density of gasoline vapors in Table 1-1A at 20°C is 1.042
       kg/m3.

1.3.3  Guidance on Inputs for, and Calculations, of Average Value
    The average quantity of contaminant vapor held in locus no. 1 is
assumed to occur when the water-filled porosity is at field capacity and
the water is at a static condition.
    Total Porosity, 6

    o  See guidance given in section 1.3.2.

    Water-Filled Porosity, 9

    o  Water-filled porosity should be chosen at field capacity, which may
       be obtained from Table 12-2.

    o  Recommend 9  = 0.09 at field capacity for a sandy soil.
                  W

    Vapor Density

    o  See guidance given in section 1.3.2.

    o  suggest using half vapor density value given in section 1.3.2 for
       average value calculation.
                                    39

-------
    Equilibrium Vapor Concentration, C


    o  See guidance given in section 1.3.2.


    o  Suggest using half equilibrium vapor concentration given in section

       1.3.2 for average value calculations.


1.4  Example Calculations

1.4.1  Storage Capacity Calculations

1.4.1.1  Maximum Quantity

    For this sample calculation, a sandy soil will be assumed.  Since we
wish to maximize the amount of volatilized contaminant liquid, we assume
the water content in the soil is at the wilting point.  For sandy soils,
the water content at the wilting point is about 5% (Dunne and Leopold,
1978).  The total porosity of the sandy soil is assumed to be 39.5% (Clapp
and Hornberger, 1978).  The air-filled porosity equals 34.5% (39.5% - 5%).

    For gasoline, the vapor density (from Table 1-2) is 1.04 kg/m .  The
maximum equilibrium concentration (from Table 1-1A) equals 0.61.  Thus  the
maximum mass in locus no. 1 can be calculated:


            mV = Co ' pv ' 9a     .
            my = (0.61).(1.04 kg/nT).(0.345)
            my = 0.22 kg/m3


    The maximum mass per unit volume in locus no. 1 for the above
conditions is 0.22 kg/m .

1.4.1.2  Average Quantity

    A sandy soil will also be assumed for  the calculation of the average
quantity of contaminant held in locus no.  1.  It is assumed that the
water-filled porosity at field capacity is about 9%.  As stated in section
1.4.1.1, the total porosity is 39.5%.  Therefore, the air-filled porosity
available at field capacity is 30.5%.

    The equilibrium vapor concentration used to calculate the average mass
of contaminant is 0.30 which is in  the middle of the range of C  reported
in Section 1.3.2:                                              °

            m  = C .p .9
             v    o Kv  a         _
            my = (0.30).(1.04 kg/nT).(0.305)

            my = 0.095 kg/m3
                                    40

-------
    The average mass per-unit volume in locus no. 1 for the above
conditions is 0.095 kg/m  .

1.4.2  Transport Rate Calculations

    Several examples of calculations for selected equations were provided
in Section 1.2 following  the discussion of the equation.  One example
calculation for vapor sorption is given below.

Vapor Sorption of Benzene

    Estimate the amount of benzene vapors that have sorbed to a soil given
xm = 5.57 mg/g of soil (Table 1-2), -(AHm + AHy) - 1.52 kcal/mole (Table

1-2), T= 20°C (293.15 K), PQ = 75.2 mm Hg (Table 1-1A), and P = 2.74 mm Hg

(Table 1-1A).

    From equation 1.19:

        - In c = (AH  + AH )/RT
                    m     v'
               = (-1.52)/[(1.987)(293.15)/(10 )]
               = -2.61


    Thus:   c  = 0.0736 (dimensionless)
    Also, P/P  = 2.74/75.2 = 0.0364


    Substituting in equation 1.18:


            x  = (5.57) 	(0.0736H0.0364)	

                        (1-0.0364)[1+(0.0736-1)(0.0364)]

               = 0.016 mg/g of soil

1.5  Summary of Relative Importance of Locus

1.5.1  Remediation

    The presence of contaminant vapors in the soil gas of the unsaturated
zone provides a source of contamination of unsaturated and saturated zone
water.  The contaminant vapors are very mobile and are capable of migrating
significant distances.  The vapors may pose an explosion hazard, especially
if they gather in sewer lines, basements, or other enclosed area.
Contaminant can sorb to soil particles or dissolve into subsurface water,
resulting in further groundwater contamination.  This migration pathway may
be significant in view of the mobility of vapors and the mass of
contaminant that may be held in locus no. 1 (Section 1-3).
                                    41

-------
    The amount of contaminant vapors that reach the water table depends on
the partitioning behavior of the contaminant.   Partitioning between the air
and water phases is governed by Henry's Law, while partitioning between the
contaminant liquid and soil air depends principally on the pure chemical
vapor pressure of the contaminant.  The equilibrium vapor pressure of the
liquid and the volume of soil air determine the amount of mass held in
locus no. 1.

    The importance of locus no. 1 relative to the other loci depends then
on the vapor pressure of the contaminant liquid, the air-filled porosity,
and the Henry's law constants of the contaminant liquid chemicals.  If
locus no. 1 holds a maximum quantity of contaminant mass, as described in
section I.A.I, it probably holds more mass per unit volume than loci nos.
3, A, 8, 9, 10, 11 and 12.

    Locus no. 1 provides an excellent medium by which to remediate
unsaturated zone contamination for volatile compounds.  Vacuum extraction
or soil vapor extraction has been used extensively to remove volatile
components from the unsaturated zone soils, while minimizing additional
groundwater contamination by removing a source of recharge.

    The success of vacuum extraction depends most heavily on the vapor
pressure of the contaminants and the air permeability of the soil.  Because
only relatively volatile compounds will exist in the soil gas, the vapor
pressure is the key contaminant variable with respect to the success of
vacuum extraction.  Bennedsen et al. (1985) state that compounds with a
vapor pressure above 0.5 mm Hg have sufficient volatility for successful
removal via vacuum extraction.  The soil air permeability is the most
important soil parameter.  Because the removal of contaminant vapors
depends on the ability to induce transport of the vapors through the
unsaturated zone, the air permeability of the soil is important in
determining whether this method can induce sufficient removal to make this
method feasible.

1.5.2  Loci Interactions

    Volatilization from liquid contaminant in loci 5, 6, and 7 into the
soil air represents the primary mechanism for partitioning into this locus.
Concentrations in the soil gas are affected by advection and dispersion,
gravity transport, and chemical retardation and sorption.  Volatilized
contaminant mass in the soil gas may partition to dry soil (locus no. 2) or
wet soil (locus no. A).  Contaminant mass may also dissolve into the pore
water of the unsaturated zone (loci nos. 3 and 12) or into the water of the
saturated zone (locus no. 8).  Contaminated soil gas is considered to be
relatively mobile and venting to the atmosphere is the chief loss
mechanism.

1.5.3  Information Gaps
                                       4
    Research into the following areas is important in better defining the
transport of soil gas:
                                    A2

-------
    1.  Importance of gravity  transport as a transport mechanism relative
        to other transport mechanisms such as advection and diffusion.

    2.  Conditions under which volatilized contaminants partition  to soil
        particles, including the influence of water content, soil  type and
        organic carbon content.

1.6  Literature Cited

Aurelius, M.W. and K.W. Brown.  1987.  Fate of Spilled Xylene as Influenced
    by Soil Moisture Content.  Water, Air and Soil Pollution.  36:23-31

Bennedsen, M.B., J.P. Scott, and J.D. Hartley.  1985.  Use of Vapor
    Extraction Systems for In-situ Removal of Volatile Organic Compounds
    from Soil.  Proceedings of the 5th National Conference on Hazardous
    Wastes and Hazardous Materials, HMCRI.  pp. 92-95.
                                 *•-
Bidleman, T.F., W.N. Billings and W.T. Foreman. 1986.  Vapor-Particle
    Partitioning of Semi-volatile Organic Compounds:  Estimates from Field
    Collections.  Environ. Sci. Technol., 20(10):1038-1043.

Brunauer, S.  1945.  The Adsorption of Gases and Vapors, Volume I:
    Physical Adsorption.  Princeton University Press, reprinted 1967 by
    University Microfilms, Ann Arbor, MI.

Call, F.  1957.  The Mechanism of Sorption of Ethylene Dibromide of Moist
    Soils.  J. Sci. Food Agric., 8:630-636.

Camp, Dresser & McKee Inc.  1987.  Modeling Vapor Phase Movement in
    Relation to UST Leak Detection.  Interim Report.  U.S. EPA, OUST.
    Washington, DC.

Cantrell, B.K., L.J. Salas, W.B. Johnson and J.C. Harper.  1986.  Phase
    Distributions of Low Velocity Organics in Ambient Air.
    EPA/600/3-86/064, U.S. EPA, Research Triangle Park, NC.

Chiou, C.T. and T.D. Shoup.  1985.  Soil Sorption of Organic Vapors and
    Effects of Humidity on Sorptive Mechanisms and Capacity.  Environ. Sci.
    Technol., 19 (12):1196-1200.

Clapp, R.B. and G.M. Hornberger.  1978.  Empirical Equations for Some soil
    Hydraulic Properties.  Water Resources Research.  Vol. 14, No. 4.

Dunne, T. and L.B.  Leopold.  1978.  Water in Environmental Planning.  W.H.
    Freeman & Co.

Farmer, W.J., M.S.  Yang, J. Letey, and W.F. Spencer.  1980.  Land Disposal
    of Hexachlorobenzene Wastes:  Controlling Vapor Movement in Soil.
    EPA/600/2-80/119.

Fischer, H.B.,  J.  Imberger, R.C.Y. Koh, and N.H.  Brooks.  1979.  Mixing in
    Inland and Coastal Waters.   Academic Press.

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


                                    43

-------
Grain, C.F..  1988.  Gas Adsorption on Solids.  In:  Environmental
    Inorganic Chemistry.  Pergamon Press, New York.

Houston, S.L., O.K. Kreamer, and R. Marwig.  1989.  A Batch-Type Testing
    Method for Determination of Adsorption of Gaseous Compounds on
    Partially Saturated Soils.  Journal of Testing and Evaluation, ASTM.
Hunt, J.R., N. Sitar, and K.S. Udell.  1986.  Organic Solvents and
    Petroleum Hydrocarbons in the Subsurface:  Transport and Cleanup.
    Univ. of California at Berkeley.  Sanitary Engineering and
    Environmental Health Research Laboratory.  Report 86-11.

Isaacson, P.J. and B.L. Savheney.  1984.  Measurement of Vapor Sorption:
    Influence of Organic Matter on Clay Sorption of 2,6-Dimethylphenol.
    Soil Science 138(6) :436-39.

Johnson, P.C., M.W. Kemblowski, J.D. Colthart, D.L. Byers, and C.C.
    Stanley.  1989.  A Practical Approach to the Design, Operation, and
    Monitoring of In-situ Soil Venting Systems.  Presented at the Soil
    Vapor Extraction Technology Workshop, Office of Research and
    Development, Edison, NJ.  June 28-89.  [Also published (1990): Ground
    Water Monitoring Review, 10(2) : 150-178] .

Jurinak, J.J.  1957.  The Effect of Clay Minerals and Exchangeable Cations
    on the Adsorption of Ethylene Dibromide Vapor.  Soil Sci. Soc. Amer.
    Proc., 21:599-602.

Lighty, J.S., G.D. Silcox, D.W. Pershing, and V.A. Cundy.  1988.  On the
    Fundamentals of Thermal Treatment for the Cleanup of Contaminated
    Soils.  APCA 81st Annual Meeting, Dallas, TX.  June 19-24.

Lord, A.E., Jr., R.M. Koerner, V.P. Murphy, and J.E. Brugger.  1987.
    In-situ Vacuum Assisted Steam Stripping of Contaminants from Soil.
    Proceedings of the 7th National Conference of Management of
    Uncontrolled Hazardous Wastes (Superfund '87).  HMCRI.  pp.  390-395.

Millington, R.J. and J.M. Quirk.  1961.  Permeability of Porous Solids.
    Trans. Faraday Soc., 57:1200-1207.

Perry, R.H. and C.H. Chilton.  1973.  Chemical Engineer's Handbook, 5th ed.
    McGraw-Hill, New York.

Reible, D.D. 1989.  Introduction to Physi co-Chemical Processes Influencing
    Enhanced Volatilization.  Presented at the workshop on Soil Vacuum
    Extraction.  R.S. Kerr Environmental Research Laboratory, Ada, OK.
    April 27-28.

Weeks, E.P.  1982.  Unsaturated Zone Diffusion Parameters.  Water Resources
    Research.  18(5).
                                    44

-------
                          SBCTION 2.  LOCUS NO. 2

                                 CONTENTS
                                                                   Page No.

List of Tables	      46
List of Figures	      47
2.1  Locus Description	      48

     2.1.1  Short Definition	      48
     2.1.2  Expanded Definition and Comments	      48

2.2  Evaluation of. Criteria for Remediation	      48

     2.2.1  Introduction	      48
     2.2.2  Mobilization/Remobilization	      52

            2.2.2.1  Partitioning into Mobile Phase	      52

                     Displacement of NAPL by Water	      53

            2.2.2.2  Transport of Mobile Phase	      53

                     Soil Permeability	      55
                     Viscosity	      57
                     Pressure Drops	      58

     2.2.3  Fixation	      63

            2.2.3.1  Partitioning onto Immobile (Stationary)
                       Phase	      63

     2.2.4  Transformation	      63

            2.2.4.1  Biodegradation	      63
            2.2.4.2  Chemical Oxidation	      64

2.3  Storage Capacity in Locus	      64

     2.3.1  Introduction and Basic Equations,	      64
     2.3.2  Guidance on Inputs for, and Calculation of,
              Maximum Value	      64
     2.3.3  Guidance on Inputs for, and Calculation of,
              Average Values	      66
                                    45

-------
                                (Continued)



                                                                   Page No.

2.4  Example Calculations	      66

     2.4.1  Storage Capacity Calculations	      66

            2.4.1.1  Maximum Value	      66
            2.4.1.2  Average Value	      66

     2.4.2  Transport Rate Calculations	      67

2.5  Summary of Relative Importance of Locus	      67

     2.5.1  Remediation	      67
     2.5.2  Loci Interaction	      68
     2.5.3  Information Gaps	      68

2.6  Literature Cited	      68
                                  TABLES


2-1  Relationship Between Diameter of Particles and Surface Area      56

2-2  Fluid Viscosities at 20°C                                        57

2-3  Surface Tension Values, Liquid-Vapor Interfaces                  61

2-4  Retention of Gasoline in Soil                                    65
                                    46

-------
                                  FIGURES


                                                                   Page No.

2-1  Schematic Cross-Section Diagram of Locus No. 2 - Liquid
     Contaminants Adhering to "Water-dry" Soil Particles or Rock      49
     Surfaces in the Unsaturated Zone

2-2  Schematic Representation of Important Transformation and
     Transport Processes Affecting Other Loci                         50

2-3  Schematic Cross-section of Locus No. 2 Contaminants Before
     and After Draining by Gravity                                    51

2-4  Capillary Pressure Effects on Variation of Water Content
     and Elevation                                                    55

2-5  Schematic Cross-section of Soil Particles Shoving Liquid
     Held by Capillary Forces                                         60

2-6  Capillary Rise                                                   61

2-7  Interfacial Surface Tensions and Contact Angle                   62
                                    47

-------
                          SECTION 2 - LOCUS NO. 2
2.1  Locus Description

2.1.1  Short Definition

    Liquid contaminants adhering to "water-dry" soil particles in the
unsaturated zone.

2.1.2  Expanded Definition and Comments

    The principal feature of this locus is the presence of liquid
contaminants, either as a continuous phase (e.g., in a spill front) or as a
discontinuous phase (e.g., separate droplets), adhering directly to soil
surfaces.  In either case, the liquid contaminants are somewhat mobile, the
movement being due to gravity, the pressure (from above) of infiltrating
water, and capillary tension.  Such liquid contaminants, whether adhering
directly to soil surfaces or not, are often referred to as "non-aqueous
phase liquids", or NAPL.

    This locus is intended to include the so-called "oil-wet" case where
the liquid contaminants preferentially wet the soil surface, displacing
water.  It also obviously includes the situation where the soils are dry
(thus there is no water to displace), so that the liquid contaminants come
into direct contact with the soil.  This locus does not include either
liquid contaminants in  the pores where water is covering the soil surfaces
(this "water-wet" scenario is covered by locus no. 6) nor the presence of
the "oil-wet" case in the saturated zone which is included in the locus no.
5 definition.  Contaminants  that sorb out of the vapor phase onto dry soils
are considered a part of this locus.  The sorption mechanism in this case
is analogous to condensation and the formation of an ultra-thin liquid film
on the surface of the particles.  Figures 2-1 and 2-2 present a schematic
cross-sectional diagram of locus no. 2 and a schematic representation of
the transformation and  transport processes affecting other loci,
respectively.

2.2  Evaluation of Criteria  for Remediation

2.2.1  Introduction

    Liquid contaminants in pore spaces between soil particles in the
unsaturated zone may exist either as a continuous bulk liquid phase, or as
discontinuous films and rings.  Initially, after a relatively large
release, the entire pore volume may be filled with organic liquid and any
residual water not displaced by  the descending organic plume.  Over  time,
much of  the liquid contaminant will drain from the pores by gravity,
leaving  the residual "attached"  to the soil particles.  Figure 2-3 shows
the saturated and drained states of locus 2.

    Contaminants may also enter  this locus through condensation of vapors.
English  and Loehr (1989)  found  that hydrocarbon  vapors partition onto  the
solid phase in the unsaturated  zone, and  that sorption is greatest at  low
moisture content where  organic  vapors  find less  competition with water
                                     48

-------
       NOTE: NOT ALL PHASE BOUNDARIES ARE SHOWN.
                                       LEGEND
                    CONTAMINANT VAPORS

                    AIR (PLAIN WHITE AREAS)
   SOIL PARTICLE

f  COLLOIDAL PARTICLE
                    WATER (MOBILE PORE WATER   ff*f LIQUID CONTAMINANTS
                    IN UNSATURATED ZONE)       *

                    BIOTA (MICROBIOTA INCLUDED)
Figure 2-1. Schematic Cross-Sectional Diagram Of Locus No. 2 - Liquid
           Contaminants Adhering To "Water-dry" Soil Particles Or Rock
           Surfaces In The Unsaturated Zone.
                                     49

-------
                                  (SOIL GAS)
                                  LOCUS NO
                            1 - CONTAMINANT VAPORS
11
     (BIOTA)
    LOCUS NO
  •SORBED TO BIOTA
                      VOLATILIZATION
DEPURATION
                      CONDENSATION
               UPTAKE
               DIFFUSION
        (ROCK)
       LOCUS NO
    10 - DIFFUSED INTO
       MINERAL GRAINS
       OR ROCKS
                              LOCUS NO. 2
                         (LIQUID CONTAMINANTS ADHERING
                         TO -WATER-DRY- SOIL PARTICLES
                             OR ROCK SURFACES IN
                             UNSATURATED ZONE)
                                   DISSOLUTION
                                   PHASE
                                   SEPARATION
        (WATER)
       LOCUS NO
 3 - DISSOLVED IN WATER FILM
12- DISSOLVED IN MOBILE
   PORE WATER
                      PERCOLATION / INFILTRATION
                      CAPILLARY SUCTION
                      RETARDATION OF CONSTITUENTS
                      DIFFUSION IN LIQUID
                              (LIQUID CONTAMINANTS)
                                   LOCUS NO.
                          6-
          UQUID CONTAMINANTS IN PORE
          SPACES IN UNSATURATED ZONE
        7-FLOATING ON WATER TABLE
       13 - IN ROCK FRACTURES
Figure 2-2. Schematic Representation of Important Transformation
            and Transport Processes Affecting Other Loci.
                                  50

-------
                  (a)
                                    (b)
Organic Liquid Saturation Of Soil Pores.
    Displaced Water Is Also Shown.
                      Oil-wet residual organic liquid after
                     draining. Some residual water remains
                           trapped after drainage.
Entrapped Water
                                            Organic Liquid
                  Soil Particle
                    o
Blank = Air Space
 Figure 2-3. Schematic Cross-section of Locus No. 2 Contaminants
             Before and After Draining By Gravity
                                      51

-------
vapors for sorption sites.  Detailed discussion of contaminant vapors is
presented in Section 1.

    The locus preferentially wets the soil particles, displacing residual
water from the surface into the pore space and creating an "oil-wet"
scenario.  For the purposes of this discussion, it is assumed that residual
water can be trapped in the pore space and not transported out of the
locus, as shown in Figure 2-3.

    Mobilization of the contaminants in locus no. 2 can be accomplished by
volatilization, dissolution, or pressure gradients arising from gravity and
liquid surface tension.  Transport of contaminants through the locus will
be a function of the liquid properties of the contaminant, such as
viscosity, and the bulk properties of the soil, such as porosity and
surface area.  Transport can be increased by affecting soil properties,
liquid properties or a combination of both.

    Immobilization of residual contaminants can be enhanced by minimizing
the parameters that increase contaminant mobilization and transport.  In
addition, partitioning within the contaminant phase itself will increase
the potential for immobilization.

2.2.2  Mobilization/Immobilization

2.2.2.1  Partitioning into Mobile Phase

    Contaminants enter locus no. 2 as a mobile phase.  The phase may be a
liquid such as continuous liquid contaminant, or a vapor such as the
volatile components ahead of a liquid front.  In the former case, the
liquid contaminant wets the soil particles, displacing any residual water
from the surface.  In the case of contaminated infiltrating water entering
the locus, the oil-wet nature of locus no. 2 assumes a preferential
sorption or partitioning of the contaminant to the soil surface.
(Water-wet partitioning of contaminants is examined in locus no. 5).  In
the case of contaminants entering locus no. 2 as vapors, the vapors may
sorb and condense on the "dry" soil.

    Because contaminants enter locus no. 2 as a mobile phase, the
contaminants in the locus may generally be viewed as being somewhat mobile,
with the exception of the first molecular layer(s) of contaminant.
Contaminants sorbed to (or direclty contacting)  the soil particles will be
somewhat less mobile because  the proximity of  the liquid and solid enhance
the attractive forces between them.  This phenomenon has implications
regarding fixation as described in Section 2.2.3 below.

    Mobilization of contaminants in locus no. 2  is accomplished through
three basic mechanisms for either the continuous liquid phase or residual
saturation:  pressure gradients, volatilization, and dissolution.  Pressure
gradients arise from gravity, including the weight of infiltrating liquid
as well as liquid surface tension.  They also affect the transport of  the
mobile phase (Sec. 2.2.2.2).  Volatilization involves a contaminant phase
change from liquid to vapor and is a function  of the vapor pressure of  the
contaminant, the temperature, and vapor space  saturation of the soil gas.
                                    52

-------
Conversely, contaminant vapors can condense, partitioning into locus no. 2
from the soil gas.  Locus no. 1 describes in detail the factors affecting
volatilization of contaminants.  Dissolution involves dissolving
contaminants into water solution and is treated in detail in locus no. 5.

Displacement of NAPL by Vater

    Whenever a porous medium is saturated with more than one fluid (e.g.,
NAPL and water), it will demonstrate a capacity to retain both fluids (Sale
and Piontek, 1989).  Locus No. 2 contaminants can be displaced by
infiltrating rain water or through a remedial flushing process.  The degree
to which displacement occurs depends on the fluid viscosities and the
effective permeabilities of the soil for the fluids.  The mobility ratio is
a measure of the ability of one fluid (e.g., water) to displace another
(e.g., NAPL), and is defined by:

                 M =  (k  .y Wfk  .y 1                                (2.1)
                      I ew  ol  I eo  wl


where M   = mobility ratio (dimensionless)
                                                                           2
      k   = effective permeability of water at residual NAPL saturation  (cm )
       ew                                                                  „
      k   = effective permeability of NAPL at residual water saturation  (cm )

      y   = viscosity of water (g/cm-sec)
       w
      y   = viscosity of NAPL (g/cm-sec)


    Effective permeability is a measure of the soil's ability to conduct a
particular fluid.  In a two fluid system, the effective permeability of  one
fluid will increase as the fraction of pore volume it occupies increases,
while at the same time the effective permeability of the other fluid will
decrease.  As water is passed through the system, NAPL is displaced until
the effective permeability of NAPL approaches zero and the remaining NAPL
is bound to the soil too tightly for further displacement.

    The addition of surfactants (e.g., alkalis or polymers) to water can
enhance displacement of NAPL.  Surfactants reduce the interfacial tension
(ITF) between NAPL and the aqueous phase, reducing the forces binding the
contaminant to the soil surface.  Alkalis can also reduce ITFs between NAPL
and the aqueous phase by forming surfactants.  In addition, alkaline agents
can neutralize soil surfaces, reducing the soil/NAPL affinity.  Addition of
a polymer to water increases its viscosity.  This will reduce the
preferential flow of water through the larger pores and result in more
thorough contact between the two fluids.

2.2.2.2  Transport of Mobile Phase

    Transport of contaminants within locus no. 2 is primarily restricted to
the continuous liquid phase (as shown in Figure 2-3a).  Transport is a
function of liquid pressure gradients due to gravity, viscous forces, and
liquid surface tension.  Transport in or out of the locus is described by
volatilization (locus no. 1) and dissolution (loci no. 3 and 12).  In as
                                    53

-------
much as condensation from the vapor phase (or sorption from solution)
represents transport of contaminant to locus no. 2, this section will be
limited to transport of the condensed or liquid phase.  In addition,
although the liquid contaminant phase may be in contact with air or water,
this section will treat the phases as essentially immiscible.

    Contaminant flow as a continuous phase through the porous medium of
locus no. 2 is inversely proportional to the viscosity of the liquid
contaminant, and proportional to the permeability of the soil and the force
(pressure) acting on (and by) the liquid.  Darcy's law states that the
volume flow rate per unit area through a porous medium treated as a bundle
of capillaries is:
              0 -
              Q -
                     8u
                               3
where  Q =  volume flowrate (cm /sec)
       r = radius of capillaries (cm)
                                2
       k = soil permeability (cm )

       u = dynamic viscosity (g/cm.sec)
                                               2
      AP = pressure drop of the fluid (g/cm.sec )
                                                    2
       A = cross sectional area of porous medium (cm )

       1 = length of the porous medium (cm)
    Equation 2.2 applies-only to conditions where soil pores are  fully
saturated with liquid.  An example calculation is provided in Section
2.4.2.  The permeability of various soils can be estimated from Table 12-2.
Table 2-2 (in the subsection on Viscosity) lists viscosities for  some
common fluids.  Equation 2.5 can be used to obtain  the fluid pressure drop.
For a unit area or length, A or 1 can be specified  accordingly.

    The capillaries consist of the interconnecting  pores between  the
particles, where the  radii vary with the grain-size distribution  in the
unsaturated zone.  For example, Figure 2-4 illustrates two representations
of the capillaries just  above the saturated zone termed the "capillary
fringe."  (Water remains/flows above the saturated  zone in the capillary
fringe due to interfacial surface tension.  Contaminant capillary flow is
discussed further below.)

    To increase the volume flow rate of liquid contaminant in locus no. 2,
the pressure on the contaminant, the capillary radii, or the soil
permeability must be  increased, or contaminant viscosity must be  decreased.
Conversely, decreasing r, P, or k, or increasing u  will reduce contaminant
flow, effectively immobilizing contaminants in the  locus.  Changing the
size of the capillary radii is not considered feasible.
                                     54

-------
  z
  o
  K
  >
  LU
  _l
  UJ
CAPILLARIES OF  Z
 LIKE  DIAMETER  p
          '      '
     WATER
     TABLE
   0   50   100
     WATER
    CONTENT  (%)
 CAPILLARIES .OF
UNLIKE. DIAMETER
                                                          ./•
                                                 I'
                    0   5'0  100
                      WATER
                    CONTENT (%)
   Source: Schwilte, 1967

   Figure 2-4.  Capillary Pressure Effects on Variation of Water Content with Elevation.




Soil Permeability

    As illustrated in equation 2.2, soil permeability  (k) is  a  fundamental
parameter governing contaminant flow in a porous medium.  Essentially,  soil
permeability is a measure of resistance to  liquid  flow, and is  independent
of the physical properties of the liquid.   (It  is  therefore an  important
parameter for any liquid flow in the subsurface, including air  and  vapors.)
Soil permeability is generally related to soil  grain-size distribution, and
is directly proportional to soil porosity and inversely proportional  to
soil surface area.  Relatively large-grained soils such as sands  are
generally more permeable than small-grained materials, such-as  clays.

    The surface area of the soil particles  plays an  important role  in
intrinsic soil permeability.  This is because surface  features  of the  soil
particles, such as depressions or micropores, produce  resistance  to liquid
flow in the unsaturated zone in addition to that produced by  interparticle
capillaries.  Altering soil surface area may have  implications  for
mobilization or immobilization of contaminants  in  the  unsaturated zone, but
its practical application in remediation may be less feasible than  other
techniques.

    Table 2-1 shows typical surface areas and particle diameters  for
various soils.  As indicated, heat activated (sintered) clays have  a very
large surface area compared to other materials.  Increasing the surface
area of the soil in locus 2 will markedly decrease permeability.  A
decrease in permeability will have a corresponding decrease on  contaminant
flow rate in equation 2.2.
                                    55

-------
                                 TABLE 2-1

        RELATIONSHIP BETWEEN DIAMETER OF PARTICLES AND SURFACE AREA
  Diameter of                                    Approximate Surface
Particles (mm)          Description                 Area (m /g)
 1.0 - 2.0              Very Coarse Sand            0.001 - 0.003

 0.5 - 1.0              Coarse Sand                 0.003 - 0.005
 0.25 - 0.5             Medium Sand                 0.005 - 0.01

 0.1 - 0.25             Fine Sand                   0.01 - 0.03

 0.05 - 0.1             Very Fine Sand              0.03 - 0.1

 0.002 - 0.05           Silt                        0.1 - 1

 <0.002                 Clay                        >1

 <0.002                 Activated Clay              1.50 - 2.25



 Source:  Adapted from Hillel, 1980
    If lowering the mobility of a contaminant in locus no. 2 is desired,
heat activating the soil (if sinterable) will greatly increase the area
available for sorption or adhesion, and reduce liquid transport out of the
locus.  Liquid transport will occur on a microscale, however, as  the
contaminant spreads over the new available surface area.  (As a practical
matter, supplying heat to locus no. 2 will have competing effects, even for
activated soils.  Heat will increase volatilization as described  in locus
no. 1, and will lower viscosity; see below).

    In addition, heat treating soils that can be activated beneath UST
installations may serve as a useful in-situ preventative measure, providing
a sub-bedding of low liquid permeability and high adsorbent capacity  for
soil vapors.  Alternatively, if native soils can not be heat activated, a
sub-bedding of activated adsorbent may be desirable.

    To increase permeability and contaminant flow rate, soil porosity may
be increased by physically increasing the void space per unit volume  of
soil.  This can be accomplished by plowing or mechanically breaking the
soil, increasing permeability in equation 2.2.  Increasing porosity
effectively increases capillary radii in equation 2.2 as well.  Conversely,
soil compaction reduces the interparticle radii and soil porosity, reducing
contaminant flowrate.  Again, the practical application of these  techniques
at UST sites, where the tank may be several meters below ground surface,
may be limited.
                                     56

-------
    Soil permeability may be determined indirectly from hydraulic
conductivity measurements, since hydraulic conductivity is a function of
soil permeability, liquid density, and liquid viscosity.  (See Section
6.2.2.2).  Estimates of soil permeability can also be found for various
soil types in Table 12-2.  Other liquid phases produce resistance to liquid
flow in addition to resistance caused by the soil particles, and also
compete for available flow area.  Section 6 discusses relative permeability
factors that account for the presence of additional immiscible liquid
phases (such as water) and can be used to adjust soil permeability.
However increasing porosity or interparticle dimensions still increases
transport, regardless of the presence of other phases.

Viscosity

    Whereas permeability is a measure of external resistance to fluid flow,
viscosity is a measure of the internal resistance to flow exhibited by a
fluid.  (Specifically, it is the ratio of shearing stress to rate of
shear.)  Table 2-2 compares viscosities of some common fluids.
                                 TABLE 2-2

                         FLUID VISCOSITIES AT 20°C
              Fluid                         Viscosity (cP)


              Air                                0.018

              Gasoline                         0.4-0.6

              Benzene                             0.64

              Water                                1.0

              Kerosene                             2.0

              Crude oil                            7.2


              Source:  Nash, 1987


    The viscosities listed are dynamic viscosities and may be used directly
in equation 2.2.  However, fluid flow rate is not linearly dependent on
dynamic viscosities because different liquids exhibit different pressure
drops for a given porous bed.  Kinematic viscosity, which is the dynamic
viscosity divided by fluid density, produces a direct comparison of fluid
flow rate.  For example, "thinner" water (less viscous by a factor of 2)
does not flow twice as fast as kerosene in a porous medium, because
kerosene has a density of 0.8 g/cm ; water flows 2.5 times faster (by
volume) than kerosene.
                                    57

-------
    Whether comparing dynamic or kinematic fluid viscosities, decreasing
viscosity increases contaminant flow rate.  One way to change liquid
viscosity is to change the liquid temperature.  Thus,  one way to increase
contaminant transport is to raise the temperature.  Because soil
temperature remains fairly constant on a daily and seasonal basis below a
depth of 2 meters, heat will typically be applied by artificial means.  Two
common methods used for tertiary recovery in oil fields are steam and fire.
As mentioned above with respect to soil permeability,  high heat may cause
localized activation of the soil, potentially immobilizing or decreasing
the transport of liquid contaminants.  (Steam is typically used to activate
carbon, clays, alumina and other adsorbents).

    Another method of changing liquid viscosity is by chemical addition.
For example, carbon dioxide is often employed by oil companies for tertiary
recovery.  Chemical additives such as viscosity index improvers or
viscosity extenders are added to motor oil to lower the viscosity at low
temperatures and to raise it at high temperatures.  Agents used for this
purpose are polymers of alkyl esters of methacrylic acid, polyisobutylenes,
and others.  Two drawbacks to the use of viscosity extenders for UST leak
remediation are in situ mixing, and the addition of hydrocarbons to the
subsurface environment.

    The mobility of contaminants in locus no. 6 can also be reduced by
emulsification, which involves the addition of emulsifiers (surfactants) to
the locus to increase viscosity.  There are drawbacks to emulsification
because it requires the presence of considerable amounts of water to convey
the emulsifier to the locus.  Dissolution of contaminants into the aqueous
phase may occur.  Care in emulsifier selection is also required to avoid
unintended consequences.  For example, while emulsifiers increase
viscosity, slowing transport, they may not completely immobilize
contaminants.  Furthermore, emulsions can break (separate), particularly if
the dispersed (i.e., dilute) phase was not completely emulsified.  An
unemulsified portion of the dispersed phase will serve as a "seed" to break
the emulsion.

Pressure Drops

    Pressure gradients in locus no. 2 arise from gravity (including
hydrostatic pressure) and surface tension.  Gravity produces a downward
force on the contaminant, pulling the liquid  through locus no. 2.  An
additional force is the weight (due to gravity) of liquid at a higher
elevation acting on liquid beneath it.  This hydrostatic pressure (or
"head") can be described by

                        AP  *    Ap gh                                 (2.3)

where  Ap = difference in density between the liquid and the gas phase
            (g/cm3)
                                                   2
        g = acceleration due to gravity (980  cm/sec )

        h = height of liquid column (cm)
                                     58

-------
    As can be seen by substituting AP of equation 2.3 into equation 2.2,
increasing the density of a liquid contaminant or increasing the height of
the liquid column by addition of liquid will increase the flow rate of a
contaminant in locus no. 2.  However, the former is impractical, while the
latter is generally unacceptable (except for water flushing) and in an
oil-wet regime possibly ineffective.  Because gravity varies so little with
elevation (0.05 percent per kilometer, ignoring tidal effects), and
hydrostatic pressure is normally localized and of small magnitude,
hydrostatic pressure is normally ignored in the subsurface environment.
For a release from an UST, there is an instantaneous pressure from the
height of the liquid in the tank.  However, the pressure quickly dissipates
laterally.  Even for no lateral dissipation, as in the case of rainwater
infiltration, the static pressure is minimal.  For example, an
instantaneous rainfall of one inch would result in an increased pressure of
only 0.035 psi in the locus.  Also, density is a stronger function of
temperature, resulting in a net reduction of flow rate with reduced
temperature.

    Thus, for the continuous phase (Figure 2-3a), gravity is the principal
pressure gradient in locus no. 2.  (This holds true for flushing with a
continuous water phase as well).  If the height of liquid is comparable to
the pore dimensions, and discontinuous films, rings, and blobs form
residual saturation (Figure 2-3b), then static pressure approaches zero and
surface tension and capillary forces dominate.  Under these conditions,
contaminants are effectively immobilized.

    A curved interface between two liquids or between a liquid and a gas
indicates a difference in pressure exists between the two (Hoag and Marley,
1986).  Capillarity concerns interfaces that are sufficiently mobile to
assume an equilibrium shape.  The most common examples are meniscuses and
drops formed by liquid in air.  The fundamental equation of capillarity was
given in 1805 by Young and Laplace (Adamson, 1981):

                   AP =  Y  1     1                                   (2.4)
                             "
                                                        2
where        AP = capillary pressure or tension (dyne/cm )

              Y = liquid surface tension (dyne/cm)

      R1 and R~ = radii of curvatue for the solid surface (cm)


    Figure 2-5 shows a cross-section of locus 2 contaminants held by
capillary forces.  One of the two radii of curvature is identified.  In the
case of a cylinder, R., = R2 and equation 2.4 reduces to:

                        AP =  2jr
                              r                                        (2.5)

where r is the radius of the cylinder.  Assuming the pores in locus no. 2
are cylindrical, and the liquid wets the soil, equation 2.5 can be used to
show the pressure exerted by the liquid contaminant on the pores within the
locus.

-------
                              "Adsorbed" Phase
                                                         Soil particle
                                 Radius of curvature of
                                 liquid /gas interface
         Source: Adapted from Hillel, 1980
            Figure 2-5. Schematic Cross-section of Soil Particles
                       Showing Liquid Held By Capillary Forces.
If the radius  is  not  too large,  the surface tension of  the liquid will
cause the liquid  to spread across the particle surface  in  all directions,
thereby "vetting" the soil particles.  The vertical rise (up or down) from
a flat liquid  surface as shown in Figure 2-6, (for which AP must be zero)
can be found by combining equations 2.3 and 2.5
                                                                        (2.6)
                                    60

-------
      Figure 2-6. Capillary Rise (capillary much magnified in relation to dish)
    As can be seen from equation  2.5,  capillary tension will be the
greatest for the smallest  pores,  regardless  of  saturation.   Indeed, for
water in loam the capillary  pressure  is  in excess  of 4 atmospheres (Brooks
and Corey, 1964).  Table 2-3 gives  surface tensions  for some liquids.

                                  TABLE 2-3

              SURFACE TENSION VALUES,  LIQUID-VAPOR INTERFACES
Liquid
Water
Benzene
Toluene
Octane
Temp.(°C)
20
20
20
20
Surface Tension;
72.88
28.88
28.52
21.62
(dyne/cm)




    Source: Adamson, 1981

    Thus, it is inherently exceedingly  difficult  to  transport  the entire
residual saturation in locus no.  2.   Equation  2.5 indicates that increasing
pore radius reduces capillary  tension,  resulting  in  increased  hydrostatic
pressure differential in the locus.   As shown  in  equation 2.2,  an increase
in pressure differential will  cause  an  increase in contaminant  flow rate.
This corresponds to increasing porosity by  mechanical  plowing.
                                     61

-------
Alternatively, liquid surface tension is a weak function of liquid
temperature, decreasing slightly with increased liquid temperature.
Increasing temperature will reduce capillary pressure (and reduce
viscosity), increasing contaminant transport.  Conversely, compaction will
reduce pore radii, increasing capillary pressure and decreasing contaminant
flow rate.

    Discontinuous rings and films adhering to soil particles can also be
mobilized by increasing the contact angle between the liquid and the soil.
The contact angle is defined by
                        6  =  cos
                                 -1
                                         - Y
(2.7)
                                       cw
Where Y  » Y   and Y   are the interfacial surface tensions for the
particli and water (pw), particle and contaminant (pc), and contaminant and
water (cw).  Figure 2-7 illustrates the relation between interfacial
tension forces and contact angle for a globule of organic liquid in a
water-saturated pore space.  The magnitude and direction of the ITFs at the
soil/water/organic liquid interface are shown.  Decreasing the surface
tension between the particle and water with a surfactant will permit the
water to wet the soil, mobilizing the contaminant.
                                                           Water
        Figure 2-7.  Interfacial Surface Tensions and Contact Angle
                                     62

-------
    However, choice of surfactant is crucial.  In an oil-wet regime where
the soil substrate has a surface electronegativity from hydrated silicates,
an anionic surfactant will help to immobilize the contaminant.  To
transport the contaminant out of the locus, a cation surfactant should be
used.

2.2.3  Fixation

2.2.3.1  Partitioning onto Immobile (Stationary) Phase

    As mentioned in Section 2.2.1, contaminant enters locus no. 2 as a
mobile liquid or condensing vapor phase. The contaminant adheres to and
spreads on the soil particles by surface tension and moves through the
locus by gravity and capillary pressure gradients.  Equation 2.2 described
the parameters governing mobility in the locus.

    Just as contaminant transport can be increased by increasing (or
decreasing) relevant parameters of flow rate, fixing contaminants (or
greatly reducing flow rate) can be achieved by decreasing the parameters
that favor mobility or increasing those that inhibit mobility.  Flow rate,
or mobility, varies with soil permeability, liquid viscosity, and pressure
differentials.

    Soil permeability can be reduced by heat activation of the soil,
effectively increasing soil surface area and reducing mobility.  Viscosity
of the liquid can be altered with polymers and surfactants.  Pressure
differentials are influenced by soil/liquid interfacial tension forces and
changes in liquid column height.  The specific physical and chemical
principles that influence these parameters as they relate to fixation of
liquid contaminant were discussed in detail earlier in this section.

    A likely first step in a remedial fixation approach would be to
minimize pressure differentials caused by infiltrating rain water.  Other
techniques are more complex and require careful preparation and execution.
To increase liquid viscosity, water may be required to convey the viscosity
enhancers to the locus.  This may result in displacement and/or dissolution
of contaminant.  Also, the chemical enhancers must be evenly distributed
through the contaminated zone at the proper concentration or results will
be less than optimal.

2.2.4  Transformation

2.2.4.1  Biodegradation (Details in Section 11)

    Biodegradation is most likely to take place in an aerated aqueous
environment in the unsaturated zone.  Contaminant liquids in locus no. 2
exist in a predominantly water-free environment and while some microbes can
live suspended in pure hydrocarbon, these are rare and metabolize oil
slowly.  Therefore, locus no. 2 contaminants would likely diffuse into
other loci before they could be biodegraded.
                                    63

-------
2.2.4.2  Chemical Oxidation (Details in Section 3.2.4.2)

    Liquid contaminants in locus no. 2 are in contact with mineral surfaces
and may undergo catalytic transformations.  Chemical oxidation in this
locus may also be enhanced by introducing an oxidant such as ozone or
peroxide.

2.3  Storage Capacity in Locus

2.3.1  Introduction and Basic Equations

    The bulk or gross storage capacity of contaminant in locus no. 2 is a
function of soil porosity.  Soil with high void space per unit volume can
store more contaminant than soil with low porosity.  The soil void space
available for storage is

                    ea  = et - ev                                     (2.8)

where 0  is the total porosity of the soil, 6  is the porosity filled by
residual water saturation, and 0  is the resultant porosity of the air
space (i.e., not soil or water).  As stated in the Locus Description,
although water may be displaced by  the contaminant, it is assumed that it
does not leave the pore space.

    To calculate the maximum mass of contaminant per unit volume of soil,
the air space porosity is multiplied by the contaminant density to obtain

                  ra   =   p 8                                         (2.9)
                             SL

where m = maximum storage capacity  (g/cm )
                                    3
      p = contaminant density (g/cm )
    After the contaminant has flowed or drained  through  the locus,  residual
films and rings remain.  The mass of residual saturation per unit volume  is
also a function of contaminant density and porosity.

                 m   =   p0  R                                         (2.10)
                  IT       3

    Here P is retention  capacity, the fraction of  total  soil porosity
storing residual contaminants.  Table 2-4 lists  experimentally derived
retention capacity values for gasoline.  "Mixed" sand  type contains equal
parts of fine, medium and coarse sand.

2.3.2  Guidance on Inputs for, and Calculation of, Maximum Value

    Maximum storage  capacity in locus no. 2  is a function of porosity and
liquid density, and  may  be calculated from equations 2.8 and 2.9.

    o   Values of porosity for various soils can be estimated  from  Figure
        12-4.  If dry soils are assumed, the total porosity is used.  If
        some water is assumed to be  retained as  residual saturation,
        air-filled porosity (6 ) can be estimated  from equation 2.8 by

-------
                                  TABLE 2-4




                        RETENTION OF GASOLINE IN SOIL
Run
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Sand
Type
Medium
Medium
Medium
Coarse
Fine
Medium
Mixed
Medium
Medium
Coarse
Fine
Medium
Mixed
Medium
Medium
Coarse
Fine
Medium
Mixed
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Fine
Fine
Fine
Dry density,
g/cm
1.44
1.44
1.44
1.47
1.47
1.55
1.74
1.55
1.55
1.57
1.62
1.71
1.85
1.71
1.68
1.72
1.72
1.48
2.00
1.48
1.47
1.57
1.56
1.71
1.55
1.72
1.71
1.47
1.62
1.72
Moisture
State
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Field
Dry
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Mass of
Gasoline
Retained, ga
123.0
93.0
112.0
90.5
330.0
117.0
218.0
118.0
120.0
100.0
290.0
128.0
218.0
132.0
117.0
98.0
290.0
100.0
200.0
77.0
84.0
81.0
69.0
88.0
85.0
83.0
85.0
121.4
100.0
124.3
(R)
Degree of Saturation
of gasoline, percentage
19.6
14.8
17.8
14.9
54.0
20.5
46.3
20.7
21.0
17.8
54.3
26.3
52.6
27.1
24.0
19.3
60.2
16.4
59.5
12.7
13.8
14.5
12.3
18.0
15.2
16.9
17.4
19.9
18.7
25.8
a.  Volume of soil contacted = 1,831




Source:  Hoag and Mar ley, 1986.
                                      65

-------
        substituting field capacity or wilting point values from Figure
        12-4 for 0 .

    o   The density of liquid contaminants can be found in various,,
        literature sources.  Liquid density of gasoline (0.74 g/cm ) is
        shown in Table 1-1A.

2.3.3  Guidance on Inputs for, and Calculations of, Average Values

    For residual contaminant saturation, equation 2.10 may be used to
calculate average storage values in locus no. 2.  Porosity and liquid
density values are as discussed in Section 2.3.2.  Retention capacity (R)
for gasoline can be estimated from Table 2-4.

2.4  Example Calculations

2.4.1  Storage Capacity Calculations

    Assume a release of gasoline in a medium sand with moisture content at
field capacity.

2.4.1.1  Maximum Value
                                               3
    o   Liquid density of gasoline is 0.74 g/cm  (Table 1-1A).

    o   Soil porosity for medium sand is 0.42 and field capacity is 0.10
        (Figure 12-4).

Using equation 2.8, air-filled porosity is:

    0  = 0.42 - 0.10 = 0'.32
     3.

Using equation 2.9, maximum storage capacity is:

    m = 0.74 g/cm3  (0.32) = 0.24 g/cm3

2.4.1.2  Average Value

    o   Liquid density of gasoline is 0.74 g/cm  .

    o   Air-filled  soil porosity is 0.32.

    o   The range of retention capacity values from Table 2-4 for medium
        sand at field capacity moisture content  is 0.123 to 0.18.  The
        average R is 0.15.

Using equation 2.10, average storage capacity is:

    mr = 0.74 g/cm3  (0.32)(0.15) = 0.036 g/cm3
                                     66

-------
2.4.2  Transport Rate Calculation

    As mentioned in Section 2.2.2.1, contaminant transport out of or
through the locus can occur through volatilization, dissolution, and/or
pressure gradients (neglecting biological and chemical oxidation
processes).  Rate of volatilization, and dissolution are examined in loci
nos. 1 and 5 respectively.  Rate of liquid contaminant transport in the
locus can be calculated from equation 2.2.

    If gasoline is flowing through locus no. 2, the flow rate can be
calculated by equation 2.2.

    o   Average pore diameter is assumed to be 0.01 cm (0.005 cm in radius).

    o   Liquid viscosity is 0.005 poise (Table 2-2).

    o   Surface tension of gasoline (taken as octane) is 21.62 dyne/cm
        (Table 2-3).

    o   AP is determined by equation 2.5.  „
        AP  = 2(21.62)70.005 = 8648 dyne/cm
                                                           2
    o   Unit area (A) and length (1) are assumed to be 1 cm  and 1 cm
        respectively.

Then, using equation 2.2, flow rate is:

    Q = 0.0052 (8648)/(8)<0.005) = 5.4 cm3/sec

The~flow rate is a linear velocity of fluid volume per unit time per area
(cm /sec/cm ).  This flow rate (5.4 cm/sec) is very high because flow
length (1) was assumed to be very short.  Increasing the length of the
medium increases surface area~and resistance to flow.  Thus, for the same
release, the flow rate per cm  for a porous bed 1 m long would be:

    Q = 5.4/100 = 0.054 cm3/sec

2.5  Summary of Relative Importance of Locus

2.5.1  Remediation

    Residual contaminant in the oil-wet unsaturated zone can provide a
continual source of dissolved contamination via infiltrating water as well
as vapor phase contamination through volatilization.  Active venting of
contaminated soils to enhance removal by volatilization is an increasing
popular remedial approach.  Decreasing viscosity through heating or chemical
additives can also be used to remove residual contaminants, although small
capillaries may remain unaffected.  (Heating the locus will also facilitate
volatilization).  Alternatively, the residual saturation can be immobilized
by emulsification, although mixing is a problem.  Contaminants can also be
immobilized by reducing soil porosity, either by compaction or filling the
unsaturated porosity with an immiscible, immobile phase.
                                    67

-------
2.5.2  Loci Interactions

    Because residual saturation of liquid contaminant can be significant
(greater than 50 percent of porosity),  locus no.  2 can be a source of
significant contamination for other loci.  For residual saturation, the
principal partitioning processes are volatilization into locus no. 1, and
dissolution into residual or infiltrating water (loci nos.  3 and 12,
respectively).

2.5.3  Information Gaps

    The principal information gaps in understanding mobilization or
immobilization of oil-wet contaminants in the unsaturated zone include:

    (1)  Fundamental data on multi-component contaminant surface tension

    (2)  Distribution of pore geometries and radii

    (3)  Data on emulsifier selection and in situ influence on contaminant
         viscosity

    (4)  Data on surfactant selection and in situ influence on contaminant
         contact angle and mobilization

    (5)  Understanding of extent of in situ heat activation of soils

    (6)  Biodegradation and chemical oxidation in the locus.

2.6  Literature Cited

Adamson, A.W.  1981.  Physical Chemistry of Surfaces.  John Wiley & Sons,
    New York.

Brooks, R.H. and Cory, A.T.  1964.  Hydraulic Properties of Porous Media,
    Hydrology Papers, Colorado State University, #3, Ft. Collins.

English, C.W. and R.C. Loehr. 1989.  Sorption of Volatile Organic Compounds
    in Soil.  Proceedings of Petroleum Hydrocarbons and Organic Chemicals
    in Groundwater.  NWWA. Houston. 383-391.

Hillel, D.   1980.  Fundamentals of Soil  Physics. Harcourt, Brace, and
    Joranovich. New York.

Hoag, G.E. and M.C. Marley.  1986.  Gasoline Residual Saturation in
    Unsaturated Uniform Aquifer Materials.  Journal of Environmental
    Engineering.  Vol.  112, No. 3. 586-604.

Nash, J.H. 1987. Chemical and Physical Properties of Gasoline.  Proceedings
    of Underground Environment of an UST Motor Fuel Release.  EPA Office of
    Research and Development.  Edison, NJ.

Sale, T. and K. Piontek.  1989.  Chemically Enhanced  In Situ Soil Washing.
    Procedures of Petroleum Hydrocarbons and Organic Chemicals in
    Groundwater.  NWWA. Houston, 487-504.
                                         68

-------
Schwille, F. 1967.  Petroleum Contamination of the Subsoil - A Hydrological
    Problem.  In:  Joint Problems of Oil and Water Industries Institute of
    Petroleum, pp. 23-54.  London, England.
                                    69

-------
                          SECTION 3.  LOCUS NO. 3

                                 CONTENTS
                                                                   Page No.

List of Tables	      72
List of Figures	      72
3.1  Locus Description	      73

     3.1.1  Short Definition	      73
     3.1.2  Expanded Definition and Comments..	      73

3.2  Evaluation of Criteria for Remediation	      73

     3.2.1  Introduction	      73
     3.2.2  Mobilization/Remobilization	      76

            3.2.2.1  Transfer to Mobile Pore Water	      76

                     Advection (Mixing)	      76
                     Diffusion in Water	      77

            3.2.2.2  Transfer to Soil Gas (Volatilization
                      from Solution)	      79

                     Equilibrium Partitioning	      79
                     Rate of Volatilization	      81

            3.2.2.3  Transport of/with Mobile Phase	      83

     3.2.3  Fixation	      83

            3.2.3.1  Partitioning onto Immobile (Stationary)
                      Phase	      83

                     Sorption	      83

            3.2.3.2  Other Fixation Approaches	      83

     3.2.4  Transformation	      83

            3.2.4.1  Biodegradation	      83
            3.2.4.2  Abiotic Transformation of UST Contaminants.      84

                     Complexation	      84
                     Photo-oxidation	      84
                     Hydrolysis	      85
                                    70

-------
                                (Continued)
                                                                   Page No.

                     Elimination	      85
                     Dehydrogenation	      85
                     Redox	      86
                     Polymerization	      89
                     Summary	      93
                     Applicability to Locus 3	      93

3.3  Storage Capacity in Locus......	      93

     3.3.1  Introduction and Basic Equations	      93
     3.3.2  Guidance on Inputs for, and Calculation of,
              Maximum Values	      95

            3.3.2.1  Concentration of Contaminants in Water	      95
            3.3.2.2  Soil Moisture Content:  Water-Filled
                       Porosi ty	      95
            3.3.2.3  Soil Bulk Density	      95

     3.3.3  Guidance on Inputs for, and Calculation of,
              Average Values	      95

            3.3.3.1  Concentration of Contaminants in Water	      95
            3.3.3.2  Soil Moisture Content:  Water-Filled
                       Porosity	      95
            3.3.3.3  Soil Bulk Density	      95

3.4  Example Calculations	      95

     3.4.1  Storage Capacity Calculations	      95

            3.4.1.1  Maximum Value	      95
            3.4.1.2  Average Value	      96

     3.4.2  Transport Rate Calculations	      96

            3.4.2.1  Diffusion Flux	      96
            3.4.2.2  Air/Water Concentrations	      97

3.5  Summary of Relative Importance of Locus..	»..      97

     3.5.1  Remediation	      97
     3.5.2  Loci Interaction	      98
     3.5.3  Information Gaps	      98

3.6  Literature Cited	      98
                                    71

-------
                                  TABLES
                                                                   Page No.

3-1  Diffusion Coefficients of Selected Organic Chemicals in
     Water                                                            78

3-2  Estimated Concentrations of Gasoline contaminants in Locus
     After Losses to Air and Soil                                     82

3-3  Oxidation Processes                                              87

3-4  Selected Values for Soil Bulk Density                            94
                                  FIGURES
3-1  Schematic Cross-Sectional Diagram of Locus No. 3 -
     Contaminants Dissolved in the Water Film Surrounding
     Soil Particles or on Rock Surfaces in the Unsaturated Zone        74

3-2  Schematic Representation of Important Transformation and
     Transport Processes Affecting Other Loci                          75

3-3  Henry's Lav Constants for Gasoline Components                     80

3-4  Humic Substances Component Structures                             90

3-5  Abiotic/Biotic Humic Substance Formation Pathways                 91
                                    72

-------
                          SECTION 3 - LOCUS NO. 3
3.1  Locus Description

3.1.1  Short Definition

    Contaminants dissolved in the water film surrounding soil particles in
the unsaturated zone.

3.1.2.  Expanded Definition and Comments

    This locus includes only the water (with its dissolved contaminants)
that exists as a thin film around, or in narrow pores between, soil
particles.  As such, it represents the relatively immobile water retained
in a well-drained soil.

    This locus does not include pore water in the unsaturated zone that is
not in close proximity to soil surfaces.  This water, which is covered in
the definition of locus no. 12, is quite mobile and may, at times,
completely fill the available soil pores.

    The locus no. 3 water will exist in times and places of dryness and/or
low infiltration.  It could thus include well drained soils under an
impervious cover (e.g., paved road), or slightly moist soils in unpaved
areas where infiltration is minimal.  The thickness of the film is not
precisely defined, and there is clearly some gradual change between water
(and its dissolved contaminants) that will be considered in locus no. 3 vs.
that which is considered to be in locus no. 12.  It will comprise only a
fraction (e.g., 1 to 10 percent) of the total soil porosity.  Figures 3-1
and 3-2 present a schematic cross-sectional diagram of locus no. 3 and a
schematic representation of the transformation and transport processes
affecting other loci, respectively.

3.2  Evaluation of Criteria for Remediation

3.2.1  Introduction

    As stated above, locus no. 3 consists of contaminants dissolved in a
thin, immobile film of water surrounding soil particles in the unsaturated
zone.  During times of rainfall infiltration, percolating water (locus no.
12) is in intimate contact with, arfd partially mixes with, some of the
water of locus no. 3.  However, a significant fraction of the locus 3
water may be held in small pores and fractures that play little or no role
in infiltration and are, thus, bypassed by the percolating water.

    The distinction between loci 3 and 12 is discussed in more detail in
Section 12.2.2.2.  The point is made there that locus 12 (mobile) water
might be considered to certainly include water held in excess of a soil's
field capacity, and, possibly, also the less mobile water that is in excess
of the wilting point but less than the field capacity.  Figure 12-4 (in
Section 12) provides data on the moisture content of various soils at field
capacity and at the wilting point.  Wilting points occur at about 5 percent
moisture content in sandy soils and about 20 percent in clayey soils.  It
                                    73

-------
  NOTE: NOT ALL PHASE BOUNDARIES ARE SHOWN.
                                  LEGEND


               CONTAMINANT VAPORS         X  BIOTA (MICROBIOTA INCLUDED)
AIR (PLAIN WHITE AREAS)
                                           SOIL PARTICLE
           O   WATER (MOBILE PORE WATER     I COLLOIDAL PARTICLE
               IN UNSATURATED ZONE)
                                        ... CONTAMINANTS
         AAAA  WATER FILM
Figure 3-1.  Schematic Cross-Sectional Diagram of Locus No. 3 -
            Contaminants Dissolved in the Water Film Surrounding
            Soil Particles or on Rock Surfaces in the Unsaturated Zone.
                                74

-------
         (BIOTA)
        LOCUS NO
    11 - SORBED TO BIOTA
10
  (ROCK)
 LOCUS NO.
• DIFFUSED INTO
 MINERAL GRAINS
 OR ROCKS
                                                         DISSOLUTION
                                                          SORPTION
                DIFFUSION
                                      (SOIL GAS)
                                      LOCUS NO
                                1 - CONTAMINANT VAPORS
                           VOLATILIZATION
                                           CONDENSATION
                                                  PHASE
                                                  SEPARATION
    LOCUS NO. 3
  (DISSOLVED IN WATER FILM
SURROUNDING SOIL PARTICLES
  OR ON ROCK SURFACES IN
    UNSATURATED ZONE)
                                    (LIQUID CONTAMINANTS)
                                        LOCUS NO.
                                 2 - ADHERING TO 'WATER-DRr
                                   SOIL PARTICLES
                                 6-IN PORE SPACES IN
                                   UNSATURATED ZONE
                                 7 - FLOATING ON WATER
                                   TABLE
                                |I3 - IN ROCK FRACTURES
                                                          DISSOLUTION
 (SORBED CONTAMINANTS)
      LOCUS NO.
4 - SORBED TO "WATER-WET
  SOIL PARTICLES
9 - SORBED TO COLLODIAL
  PARTICLES
                                       (WATER)
                                       LOCUS NO.
                                 12 - DISSOLVED IN MOBILE
                                    PORE WATER
      Figure 3-2. Schematic Representation of Important Transformation
                   and Transport Processes Affecting Other Loci.
                                           75

-------
should be understood, however, that soil moisture content cannot be used to
classify all of the moisture in a soil as being either locus no. 3 or 12.
In many (perhaps most) soils, both will be present.

    The properties of the water held in locus no. 3 are assumed here to be
the same as for bulk water.  It is known, however, that water within a few
molecular diameters of a solid phase, such as a clay particle, may have a
more rigid structure and be more immobile than is water at greater
distances.  This structure is induced by electronic interactions between
surface charges on the soil and the dipole of the water molecule.  Some
hydrogen bonding may also be involved.  Such interactions are the molecular
basis for interfacial tension forces which, in turn, are the basis for
capillary suction.

3.2.2  Hobilization/Remobilization

    Contaminants held in locus no. 3 are, by definition, in an  immobile
phase.  The only two approaches available to mobilize these contaminants
are by transfer to mobile pore water (locus no. 12) or by volatilization
into the mobile soil gas (locus no. 1).  Each process is discussed below.

3.2.2.1  Transfer to Mobile Pore Water

    Two mechanisms are available to transfer contaminants into  mobile pore
water (locus no. 12):  advection (or mixing) and diffusion.

Advection (Mixing)

    Although locus no. 3 is defined as being an immobile water  film, there
is clearly a continuum of mobility between the strongly held water (a few
molecules thick) essentially  touching  the soil surfaces and the free
mobility of locus no. 12 waters.  When both loci are present  (i.e., the
moisture content is above  the wilting  point),  then gravitational flow -  and
other pressure gradients - will cause  some intermingling of the two.

    Because the distinction between loci 3 and 12  is an invention of this
report,  there are - presumably - no published  reports which directly
address  the extent of mixing during periods of infiltration.  It would
appear obvious, however, that mixing would be  facilitated:

    o    During periods of rapid infiltration  associated with saturated
         soils and high hydraulic heads;

    o    In high porosity  soils;

    o    In homogeneous soils; and

    o    With more soluble contaminants  (since sorption  is  less important
          for them).

    It cannot be assumed that advection  and mixing would  increase in direct
proportion  to  the amount of  infiltrating water.   As  is  discussed in Section
12.2.2.2,  the presence of  macropores  often leads  to  a  situation where a
small fraction of a  soil's porosity  is responsible for  a  large  fraction  of
                                     76

-------
the total water flow during periods of rapid infiltration.  For example, 10
percent of the porosity - in the form of macropores - might be responsible
for 80 percent of the flow.  The water held in the remaining 90 percent of
the porosity moves at a much slower rate and is mostly bypassed.  In such
situations, one might anticipate having to flush a soil with the equivalent
of several pore volumes to achieve a degree of mixing and dilution that a
well mixed system (e.g., a high porosity soil) would give with a single
pore volume.

    When advection and mixing do occur, solute retardation due to soil
sorption will occur.  Detailed discussions of sorption and retardation are
given in Sections A and 8, respectively.

Diffusion in Water (Additional details in Section 7.2.2)

    When both loci 3 and 12 are present in a soil, diffusion of
contaminants through the aqueous phase can lead to some mobilization even
if the water is stagnant.  Diffusion is driven by concentration gradients.
The constant of proportionality between the concentration gradient (AC) and
the induced flux (J) is called the diffusion coefficient (D), i.e.:

                                 J = D AC/Ax                          (3.1)

  where  J  = net molal flux of contaminant across a hypothetical plane
              (mol/cm .s)
                                                               2
         D  = diffusion coefficient of contaminant in water (cm /s)
         AC = concentration gradient of contaminant at the hypothetical
                           3
              plane (mol/cm )

         Ax = distance over which concentration gradient is measured (cm)
    Selected values of D are provided in Table 3-1.  As can be seen from
this table, most organic chemicals - especially neutral ones of low
molecular weight - diffuse at about the same rate in water.  Also,
diffusion rates decrease with decreasing temperature.

    For diffusion on a macroscale, the blocking action of the soil
particles must be accounted for; this is usually done by including a
dimensionless tortuosity factor (0 < T < 1) and the calculation of an
effective diffusion coefficient:

                             Deff = T D                               (3.2)

    Soils with a high degree of tortuosity in the water-filled pores have
lower values of T and, thus, lower effective diffusion rates.  Values of T
are typically in the range of 0.01 to 0.5 (Tucker and Nelken, 1981).
Although expressions have been given for the estimation of T for air-filled
pores - and diffusion in air (see Section 1.2.2) - it is not clear if the
same expressions can be used for diffusion in soil water.
                                    77

-------
                                 TABLE 3-1

                         DIFFUSION COEFFICIENTS OF
                    SELECTED ORGANIC CHEMICALS IN WATER
Chemical
HYDROCARBONS
n-Butane
n-Pentane
Cyclopentane
Cyclohexane
Methylcyclopentane
Benzene
Toluene
Ethylbenzene
OXYGEN CONTAINING
Methanol
Ethanol
Formic acid
Acetic acid
Valeric acid
Acetone
Ethyl acetate
Benzyl alcohol
5 2
Diffusion Coefficient x 10 (cm /s)
25°C 20°C 2°C Source

0.89 0.50 <4°C) a
0.84 0.46 (4°C) a
0.93 0.56 a
0.84 0.46 a
0.85 0.48 a
1.02 0.58 a
0.85 0.45 a
0.81 0.44 a

1.66 — — b
1.24 — — b
1.52 — — b
1.19 ~ — b
0.82 — — b
1.28 — — b
1.12 — — b
0.93 — — b
a.  Witherspoon and Bonoli, 1969

b.  Hayduk and Laudie, 1974
                                    /8

-------
    From the above it is seen that mobilization via diffusion is enhanced
by increasing temperature, increasing porosity and increasing concentration
gradient.  While remedial actions can be taken to provide some increases in
temperature and concentration gradient (the latter, for example, by vacuum
extraction of gases), little can be done to change the porosity.  Tucker
and Nelken (1982) suggest that diffusion can probably be ignored if pore
water velocities exceed 0.002 cm/s.

3.2.2.2  Transfer to Soil Gas (Volatilization from Solution)

Equilibrium Partitioning

    Contaminants held in locus no. 3 can be mobilized by transfer into the
soil gas phase.  This air-water transfer is a natural process.  When
contaminants such as petroleum hydrocarbons are present in a system
containing both water and air, they will partition themselves between the
two phases in proportion to their*-value of Henry's law constant:

                             Ca = H Cw                                (3.3)

                                         3
  where  H  = Henry's law constant (atm m /mol)
         C  = concentration in air (atm)
          3                                3
         C  = concentration in water (mol/m )
          w


    Values of H for selected gasoline constituents and additives are
provided in Figure 3-3.  (Numeric tabulations were provided in Tables 1-1A
and IB.)  As can be seen from these data, values of H range over eight
orders of magnitude.  Hydrocarbons range over three orders of magnitude,
with the saturated aliphatics being the most volatile (H ~ 1-10 atm m /mol
@ 25°C), unsaturated and cyclo-aliphatics slightly lower in volatility (H ~
0.1-1 atm m /mol @ 25°C), and light aromatics being the least volatile (H -
0.007 - 0.02 atm m /mol @ 25°C).  Some additives (e.g., ethanol) are of
such low volatility that mobilization into the soil gas will be extremely
difficult unless the soil is taken to dryness.

    Values of H for hydrophobic solutes may be estimated from the ratio of
vapor pressure to water solubility (i.e., H = P/S).  The influence of
environmental variables on H can, thus, often be estimated from the
corresponding influence on vapor pressure (P) and solubility (S).  For
hydrocarbons, we can expect H to: "increase significantly with increasing
temperature; increase slightly with increasing water salinity; and decrease
slightly with increasing concentration of dissolved organic carbon (DOC) in
water.  Some data that are specific to hydrocarbons are provided by Mackay
and Shiu (1981), Leighton and Calo (1981), Gossett and Lincoff (1981),
Mackay et al. (1979), and Lion and Garbarini (1983).  Additional
information on the variability of vapor pressure and solubility with
various environmental parameters is provided in Sections 1.2.2 and 7.2.2,
respectively.

    It is interesting to note that the combined effects of volatilization
and sorption lead, for a typical gasoline, to a very strong (relative)
                                    79

-------
                                     10
                                     10

                          2-METHYLHEXANE -
                               N-OCTANE '
                         2-METHYLPENTANE •
                              N-PENTANE '
                               N-3UTANE -

                      ME7HYLCYCLOHEXANE ,

                           CYCLOHEXANE

                                     10'
    NOTE:
    Values of H are
    in units of arm x rrr /mol.
    T = 25°Cfor
    hydrocarbons and 20 °C
    for additives

    BOLD TYPE « ADDITIVES
              i
ETHYLBENZENE-
    TOLUENE•
    BENZENE-
             i-2
                                     10 -3
                                     1CT
                                     10
           10
                                     10
                                      -7
                        2.2. s. s - TETRAMETHYLHEXANE
                        DODECANE
                       ^ 2.2.4 - TRIMETHYLHEXANE
                       ^— 2. 4 - DIMETHYLHEXANE
                       ^-N-HEPTANE
                          N-HEXANE
                          ISOPENTANE
                          'ISO8UTANE
                          • TETRAETHYL  LEAD
                      \-1-HEXENE
                        ^I-PENTENE
                                                 1,4-DIETHYLBENZENe

                                                 M-XYLENE
                                                 1.3.5 - TRIMETHYLBENZENE
                                                    1. 2 - DICHLOROETHANE
                                                    ETHYLENEDIBROMIOE
                                                    METHYL-T- BUTYLETHER
                           METHYLCYCLOPENTAOIENYL
                           MANGANESE TRICARBONYL
METHANOL

ETHANOL




TRI-0-CRESYL PHOSPHATE
Source: Most hydrocarbon data are from Mackay and Shiu, 1981.  Remainder are estimates by
       Camp, Dresser & McKee, Inc., 1987.

     Figure 3-3.  Henry's Law Constants for Gasoline Components
                                     80

-------
enrichment in the concentration of light aromatics:  benzene, toluene,
ethylbenzene and xylene (BTEX).  In one model calculation (W. Lyman,
unpublished data), a hypothetical gasoline with 23 constituents was
considered to have contacted pore water in the unsaturated zone until
disolution had reached equilibrium.  The dissolved constituents were then
allowed to partition into the soil gas and, via sorption, to the soil.  The
model results are shown in Table 3-2.

Rate of Volatilization

    In instances when the concentration in water exceeds the equilibrium
concentration, volatilization will occur.  The rate of volatilization can
be approximated from an approximation to Picks' diffusion law (c.f. eqn.
3.1) (Thomas, 1982):

                         Jyl = k AC                                   (3.4)

                                                                   2
  where  J , = flux of contaminant across air-water interface (g/cm s)

           k = a first-order exchange constant (cm/s)
                                                              3
          AC = concentration difference across interface (g/cm )


    In practice, k is usually expressed in terms of a gas-phase (k ) and a
liquid-phase (k,) mass transfer coefficient, expressions which always
reflect the movement (i.e., degree of mixing as reflected in a velocity
term) of the fluid phase.  Several expressions for k  and k-, have been
derived for surface waters and impoundments (Thomas, 1982), and they show
the value of k going to zero as the fluid velocity (and  thus mixing) goes
to zero.  Thus, in quiescent environments - such as the  stagnant water of
locus no. 3 - the rate of volatilization may reach very  low levels and be
controlled by the rate of diffusion of contaminants in the air and water
phases on either side of the interface.

    In cases where vacuum extraction is implemented, the air phase may be
relatively well mixed* but the locus 3 water may not be.  In any case,
where H is high (>10   atm m /mol), as it is for most petroleum
hydrocarbons, the resistance of the water film dominates by a factor of ten
at least, i.e., k, is at least ten times less than k .   The volatilization
transfer is liquia-phase-controlled and equation 3.4gcan be written as:

                          Jvl - Kl 
-------
a
                1-1
                 4-> C*<
3 4J OH HH (tf
•on) < c/5 co
•rn M 01 i cu
co 4-1 co IH cu co
cu e o cu a Q
fv* (U 'O 4-* »H 1-4
C/ (u {o rH
a > > o IH
5 J|
c
•H
c
T5 0
Q> -rt
4-1 4-1
§fd
M
•H 4->
« s
U 0

6


•H CO
§53
Cu
•H X
l-l
tfl rH
•o 'c8
4-> a,
« c
M Tl
w










^J
Kff
rH 5^
•H s^
W M
TJ 'H
ra o
•H
U 4->
•rt -rt
^t:
2,2
U
4->
2 rH*
3 "v
1 "^
CU ^4
C 4)
•H 4->
rH Cd
O t*
i
rH
•H
£
•H
>^

0)
5

c
•H
X
l-l
•H
<3
•H
4-1
C
(U
c
8,
e
cS
(U
c
•H
rH
O
                                                  t-l «H \O CM
                                                                 00
               CM
               r>. oo f^
               O rH CO
                                                                                                               CM
                                                                                                               ID
                                                                                                   00
                                                           CO CO
                                                                                          CM
                                                     8 CM CM O
                                                     m vo &\ o
•*-*incMOc*icM-*
CMOOCMOCOOrHr-t
in
in
8SS
   rH VO O  - .
»H o co o  in
                                                                                                                      '
                                                  rHrHrHCOCOrHOrHCM
                                                                                          CM
                                                     X


                                                     CO
                                                  rHrHr^-St    rHCOCM        
                                                                                             -* m        -«•    m
                                                                                                          U U



                                                                                                    •^ ^^."i
                                                                                              .   .  .&&CMCO
                                                                                              I   « 4->  I   ^  •>
                                                                                              ~ CM  W XI CM rH
                                                  £
                                                  4-1
                                                   CU
                                                  •H
                                                  Q

                                                  4
                          CU
                                                                                                                            IT)
                                                                                                                            SO
                                                                           CO

                                                                           rH


                                                                           ft?
                                                                           4-f

                                                                           •o



                                                                           1
                                                                           co
                                                                           •H
                                                                                                                                            1-3


                                                                                                                                             •

                                                                                                                                            >
                                                                        82

-------
    Again, however, K, is directly related to water velocity (and air
velocity, too); and wnen either of these go to zero, so does K,.  Thus, it
would appear that, in vacuum extraction remediation, a state would soon be
reached where the rate of movement of (and diffusion in) water would be the
rate-limiting step for mass recovery.  Any attempts to increase mass
recovery by changing only the air velocity (and air concentration) will be
futile until the soils have dried out.

    Once the contaminants have entered the soil gas (locus no. 1) their
movement in, and transport with, soil gas is controlled by a variety of
factors.  Such transport is described in Section 1.

3.2.2.3  Transport of/vith Mobile Phase

    Transport of contaminants in soil gas is described in Section 1.
Transport of contaminants in mobile pore water within the unsaturated zone
is described in Section 12.

3.2.3  Fixation

3.2.3.1  Partitioning onto Immobile (Stationary) Phase

Sorption (Details in Section 4.2.2)

    The contaminants in locus no. 3 are already in a relatively immobile
phase.  They may be made more immobile by sorption to soil.  Sorption will
be most important for the constituents of low water solubility and low
vapor pressure.  For petroleum hydrocarbons, this generally means saturated
(and some unsaturated) compounds having roughly nine or more carbons.  Soil
sorption is not very effective at immobilizing the light aromatics (BTEX
fraction).

    The sorption equilibrium is not easily changed by parameters that can
be adjusted during common remediation activities.  It is, in fact, easier
to enhance the desorption process via addition of water, detergents,
solvents or - in some cases - heat.  Only if the soils are taken to dryness
will the higher molecular weight fraction of contaminants, especially those
that are solids at the system temperature, become fairly well immobilized.

3.2.3.2  Other Fixation Approaches

    Isolation from infiltration (e.g., via the use of impermeable covers)
will prevent transfer to locus no. 12 (mobile pore water).  An impermeable
cover will also reduce volatilization to the air.

3.2.4  Transformation

3.2.4.1  Biodegradation (Details in Section 11)

    Biodegradation by microbes occurs most readily when microbes, organic
contaminants and other nutrient substrates (nitrogen, oxygen, phosphorus,
potassium) come in close physical contact.  This may well be the case for
locus no. 3.  However, the discontinuous nature of the water film precludes
significant communication with other sources of substrate.  Therefore,
                                    83

-------
biodegradation may be high initially but drop off significantly when any of
the above named substrates, including the organic contaminant becomes
limiting.  A full treatment of biodegradation kinetics is found in the
locus no. 11 section.

3.2.4.2  Abiotic Transformation of UST Contaminants

    Contaminants released to the environment from an UST will be subject to
biotic and abiotic "weathering" reactions in the soil/groundwater media.
The rate and degree of weathering will be related to the characteristics of
the contaminant, whether it be gasoline or no. 2 fuel oil, for example,
mediated by environmental factors including temperature, moisture content
and soil chemical, physical and microbial properties.

    Abiotic and biotic properties operate in concert to weather
hydrocarbons introduced to the soil environment.  Although mineralization
may be envisioned as the ideal end point for remediation of hydrocarbon
contaminated media, certain degradation products may remain as stable
components in soils for extended periods of time.  These degradation
products include aliphatic acids and aromatics that are integral components
of soil humus (Alexander, 1977).

    Humic substances are brown to black colored organic acids which have
high molecular weight and are structurally complex.  Humic substances play
a significant role in both the biotic and abiotic transformation of
introduced hydrocarbons as well as being formed from the degradation of the
hydrocarbons.  Degradation of petroleum hydrocarbons to humic substances
rather than to the mineralization products of CCL and H~0 could therefore
be expected to beneficially impact soil physical and chemical properties.

    Abiotic processes which are involved in the transformation of UST
contaminants include:

    o   complexation
    o   photo-oxidation
    o   hydrolysis
    o   elimination
    o   dehydrogenation
    o   redox and
    o   polymerization

Complexation

    Clays have been shown  to be of significance in  the stabilization of
organic carbon by  complex  formation with active aluminum groups or
sesquioxides.  The silica-alumina humus complex may also increase the
resistance of bound humic  substances  to degradation  (Huang,  1987).

Photo-oxidation

    Photo-oxidation  is an  important abiotic transformation process when
petroleum is spilled on soil svu faces or in aquatic  systems.  The excited
state brought about  by sunlight that  leads  to photo-oxidation does not
occur in  the subsurface soi] system and is, therefore, not considered  a

-------
significant factor in abiotic transformations for USTs.  However,
photo-oxidation could potentially be incorporated in certain pump and treat
schemes.

Hydrolysis

    Hydrolysis is the reaction of a compound with water, hydronium or
hydroxide ions resulting in bond cleavage.  Equation 3.6 is an example of
an hydrolysis reaction transforming an epoxide to a diol (Valentine, 1987):

         R-CH - CHR0  +  H00  ->  R.,CHCR0                            (3.6)
          *\/   *      2      \l\  2
              0                    OH  OH

    At a constant pH, hydrolysis appears to occur as a first order or
"pseudo-first order" reaction (Amdurer et al., 1986).  The rate of
disappearance of the substrate (R£) follows:
                     -d(Rx)/dt  =  k(Rx)                              (3.7)

    Hydrocarbons resistant to hydrolysis include saturated and unsaturated
aliphatics, aromatics, phenols, alcohols, ketones and polycyclic aromatic
hydrocarbons (Harris, 1982).

Elimination

    Abiotic mechanisms have been shown to be operable in the degradation of
chlorinated hydrocarbons.  The solvent 1 ,1,1-trichloroethane (TCA) is
transformed abiotically to yield 1, 1-dichloroethylene (1,1-DCE) by
elimination and acetic acid by hydrolysis (Cline et al., 1986).  Vogel &
McCarty (1987a and 1987b) confirmed the process of elimination of H  and
Cl~ from TCA producing 1,1-DCE in neutral dilute aqueous solutions at 20°C.

    Abiotic homogeneous elimination of HCl from 1,1,2,2-tetrachloroethane
(TeCA) produced 1,1,2-trichloroethene (TCE) in a 0.1 M phosphate-buffered
distilled water solution within a pH range of 5 to 9, and temperature range
of 30 to 95°C (Cooper, et al., 1987).  The half life of TeCA at 25°C is
reportedly on the order of 102 days at pH 7 and 1.02 days at pH 9.

Dehyd rogena t i on
    Dehydrogenation can be described as the transfer of H  ions and
electrons from a reduced substrate to a terminal H  and electron acceptor,
A, in the absence of 02 (Tabatabai, 1982):

                     XH2 + A -» X + AH2                                (3.8)

    Dehydrogenation is most often mediated in soils by microbial enzymes.
Enzymes are proteins produced by living cells that can catalyze chemical
reactions in soils.  The enzyme causes reactions to proceed at accelerated
rates but do not themselves undergo permanent alteration.  Extra-cellular
enzymes are important in the degradation of large molecules which can not
enter the microbial cell (Alexander, 1967).  Enzymatic coupling of aromatic
compounds has been demonstrated (Maloney et al., 1986) and it is known that
                                    85

-------
toluene can strongly inhibit dehydrogenase activity in soil at elevated
levels (Tabatabi, 1982).  Likewise, the oxidation inhibitors added to
petroleum products to retard microbial growth during storage also limit
microbial degradation in soils.  Dehydrogenase activities of soils exposed
to leaded gasoline, diesel fuel, kerosene or motor oil are reduced compared
to those of soils exposed to crude oils which do not contain oxidation
inhibitors (Frankernberger & Johanason, 1982).  Additional discussion of
enzyme mediated transformation processes are contained in Section 11.

Redox

    Redox reactions may be the most significant of the abiotic processes
that lead to transformation of contaminants in soils.  Redox refers to the
reduction/oxidation process whereby reduction of one compound is
accompanied by the oxidation of another compound.  Oxidation occurs vhen a
reactive electron-deficient substrate (oxidant) removes electrons from more
electron-rich substrate.  The latter substrate is oxidized (loses
electrons) while the oxidant is reduced (gains electrons).  The redox half
reactions for the oxidation of glucose by H.O is written (Valentine, 1987):

         C6H12°6  +  6H2°  *  6C02 + 24H+ * 24e~                     (3>9)
         602  + 24H+  +  24e~  •*  12H20                              (3.10)

         C6H12°6  + 602  "*  6C02 * 6H2°                              (3-11)

    Organic oxidation frequently involves the gain of oxygen and loss of
hydrogen.  Table 3-3 presents a general schematic of oxidation processes.

    The majority of naturally occuring redox reactions in soils are
expected to be mediated by soil humus and metallic clay complexes which
serve as catalysts for oxidation reactions.  Oxidation of n-alkanes
proceeds via terminal oxidation with the formation of a saturated primary
alcohol at the expense of oxygen.  Aromatic ring fission requires
dehydroxylation followed by introduction of oxygen.  Electron acceptors
other than oxygen can mediate oxidation in soil environments.  Denitrifying
bacteria have been shown to be capable of degrading alkanes in the absence
of Oy if NO- is present to serve as an electron acceptor (Thomas and Ward,
19897.

    Oxidants present in the environment include peroxy radicals (RO-),
hydroxyl radicals (OH), oxygen (02) and ozone (03) (Mill, 1980).  Free
radicals, metal ions or other catalysts combine with triplet oxygen to form
peroxy radicals which remove hydrogen atoms from the substrate to form
hydroperoxide (Larson et al., 1979).

    Incorporation of oxygen into hydrocarbons increases water solubility;
for instance, 2-naphthol solubility is approximately 1000 mg/1, whereas
naphthalene's solubility is approximately 35 mg/1 (Larson et al., 1979).
Oxygenated hydrocarbon compounds were also found to be toxic or inhibitory
to heterotrophic microorganisms by Larson et al. (1979).
                                    86

-------
                                 TABLE  3-3


                            OXIDATION PROCESSES
H-Atom transfer
                       RO  + H-C-->ROH-»-   R09 - C - C
                          /      \                |    |


HO'  addition to aromatics

                          HO' +   f|  ^   *
                                  \^s
                                           HO



R00"   transfer of 0-atoms to nucleophilic species
                           R02*  + NO  -> RO*  +  N02
R = alkyl or H
SOURCE:  Mill et al.  (1980)
                                     87

-------
    Ozonation - Ozone is used extensively for the oxidation of organic
constituents in wastewaters and drinking water supplies.  Ozone acts to
oxidize compounds directly by the formation of hydroxyl free radicals from
decomposed ozone or by interaction of ozone with the solute to induce
oxidation.  Ozone occurs in the atmosphere but is not expected to be a
major factor in in-situ degradation unless introduced to the soil.

    Ozone attack on carbon-hydrogen bonds results in transformation of
aldehydes to carboxyl acids, alcohols to carbonxylic acids and/or aldehydes
or ketones, ethers to alcohols and esters and hydrocarbons to alcohols and
ketones (Bailey, 1972).

    Ozone reacts readily with benzene, toluene, xylene, and benzopyrene,
resulting in formation of products which are usually more susceptible to
biodegradation and less toxic (Rice et al., 1981).  Maloney et al. (1985)
have determined that ozonation of dissolved organic carbon, humic
substances, increases their biodegradability.  However, oxidation of
certain organics such as pesticides by ozone, for example, has been shown
to lead to production of degradation products which are more toxic than the
parent compounds (Rice et al., 1981).

    Ozone decomposes very rapidly in water, yielding hydroxyl radicals  that
can,.react with aromatic compounds such as benzene (rate constant k=6.7  x
10~  1/mol.s) and cyclohexene (k=8.8 x 10~  1/mol.s) (Hoigne and Bader,
1976).  However, only 0.5 mole of OH radicals per mole of ozone decomposed
was shown to be effective in reacting with the organics.

    The following conclusions are among those drawn by Rice et al. (1981)
with respect to ozonation:

    o   Saturated aliphatic hydrocarbons are unreactive with ozone.

    o   Alcohols are slowly oxidized to aldehydes and ketones, which  then
        slowly oxidize to acids.

    o   Benzene is slowly oxidized, and other aromatics are easily
        oxidized except when electron withdrawing constituents are
        present.

    o   Reaction products are more polar and more readily biodegradable.

    o   Complete reaction  to CO,, and water rarely occurs.

    Utilization of ozone for in-situ oxidation of petroleum contaminated
soil may  not be practical  for several reasons:   its reactive nature and
rapid decomposition necessitates on-site production; relatively high  cost;
and the availability of other oxidants, such as  hydrogen peroxide.

    Hydrogen Peroxide  - Hydrogen peroxide  (HpOO  is fully miscible in water
and commercially available  in aqueous solutions  over a  wide range of
concentrations.  The oxidation of aliphatic and  aromatic hydrocarbons by
use of hydrogen peroxide is  similar  to  that which occurs with ozone.
However,  peroxide  can  be more easily  introduced  into the subsurface

-------
environment by injection wells and is commonly used for accelerating
in-situ biogedegradation by increasing the available oxygen.

Polymerization

    Polymerization is the joining of small molecules together to form large
molecules.  In soils, clay minerals are responsible for the polymerization
of unsaturated organics.  Benzene, toluene, phenol and quinone have been
shown to polymerize spontaneously on smectite surfaces if Cu(II) or Fe(III)
are present to serve as a catalyst (Bohn et al., 1979).

    Humic substances are naturally occurring heterogeneous refractory
organics of high molecular weight that can be categorized based on their
solubility (Aiken et al., 1985):

    o   Humin - insoluble in water

    o   Humic acid - insoluble in water under acidic conditions (pH<2)
        but becomes soluble at higher pH

    o   Fulvic acid - soluble in water

    The characteristics of the humic substances within a soil are
relatively stable as biotic inputs are balanced by mineralization.  With
time, humic substances are ultimately converted to kerogen, a coal-like
material, which is insoluble in non-oxidizing acids, bases and most organic
solvents.

    The presence of naturally occurring soil humic substances will
influence the rate and abiotic transformation mechanisms of hydrocarbons
introduced to the environment from leaking underground storage tanks.  The
quantity of naturally occurring organic matter present in soils varies
significantly, with soils that have developed in anaerobic environments,
such as swamps, having significant accumulations of organics, while those
that developed under aerobic conditions having lower organic matter
contents in general.  The amount of rainfall and the temperature under
which the soil formed will also influence its organic matter content, with
increasing organic matter generally being accumulated to a greater extent
under conditions which favor biological activities.  The major
accumulations of organic matter are associated with the most biologically
active zone of soils.  This zone is generally found from the ground surface
to the depth to which plant roots and soil organisms (earthworms) are
active.  Exceptions to these generalities are those cases where buried soil
horizons or petrogenetic organic matter exists.

    Organic matter inputs to soil include plant residues and microbial
cells but can also include introduced hydrocarbon contaminants.  Plant
organic matter is comprised of cellulose, hemicellulose, lignin and various
simple sugars, amino acids, aliphatic acids ethers, alcohols and proteins
(Alexander, 1964).  Humic substance are formed in soils as organic matter
decomposes in soils.  Although degradative processes lead to humic
formation, polymerization also has a role in its formation.
                                    89

-------
    Underground storage tanks ate generally installed at a depth of 3  to 12
feet beneath the soil surface.  At these depths the organic matter content
of most mineral soils is less than 1 percent.  The backfill and bedding
materials used for UST installations is also expected to be low in organic
matter content.  The naturally occurring humic substances required for
initiation of redox reactions with soil clay minerals may, therefore, not
be present to any great extent within UST soil materials.  However, the
release of hydrocarbons from an UST can introduce the reducing agents
required for initiation of abiotic polymerization type reactions.

    Fulvic acids also play a role in the transport of insoluble organic
substances, such as alkanes, in water by acting as carriers (Stevenson,
1985).  Petroleum can become associated with humic materials forming
non-ionic polar organic macromolecules.

    Figure 3-4 presents some postulated structures for components of humic
substances.  Recent research using carbon magnetic resonance (CMR) and wet
chemistry oxidative degradation studies have provided insight into humic
structure.  Although no generally accepted structure for humic substances
is found in the literature, Ghosal and Chian (1985) have determined that
the presence of 3 or more condensed aromatic ring systems in humic
substances is almost absent, but heterocylic compounds derived from
carbohydrates are present.  Chen and Schnitzel (1976), in evaluating the
viscosity of fulvic humic acids, determined that the materials behaved as
linear polyelectrolytes, not like structures exclusively composed of
condensed rings.  The molecular weights of the materials they studied
ranged from 1081 to 6171 with diameters of 25 to 107 A.  The
characteristics of the humic materials present in soil will affect the
degree to which abiotic reactions take place.
                        COOH
                        COOH
                                              O     OH
                                              M      i
                                             -C-CH = C-
                                                         —* n
               Source: Stevenson, 1985
               Figure 3-4. Humic Substance Component Structures

-------
     Evaluation of organic matter  from an Aquoll indicates  that  the humic
acids extracted from clays had a higher content of aliphatics  and lover
content of  aromatics than did the  humic acids extracted from  the larger
sized soil  separates, silt and sand  (Catroux and Schnitzer, 1987).

    Abiotic factors have been shown  to play a vital role in the
polymerization of phenolic compounds leading to the formation  of humic
substances  (Vang et al., 1986).  These findings are significant  with
respect to  the discharge of hydrocarbons from an UST because mineralization
of the materials may not be required to achieve remediation.
Transformation of the hydrocarbons into stable humic substances  would not
only provide for elimination of the  contaminant but impart  the beneficial
properties  associated with organic matter to the soil.

    Figure  3-5 presents the biodegradation pathway for hydrocarbons and the
postulated  abiotic polymerization  reactions leading to formation of humic
substances.   Evidently, the interaction of  biotic and abiotic pathways may
play a major role in the transformation of hydrocarbons introduced  into the
soil systems.
 TOLUENE   BENZYL ALCOHOL  KNZM.DCHYE   KN2OCWD
                                               MINERALIZATION
                                              	»•  CXJ^.H.0
              PHCNCL

                I
                |  F«M CLAY CATALYST
                                               KXYMENZATON
                                                           HUMC
                                                         SUBSTANCES
                                                                     MICROBIAL
                                                                    DEGRADATION
                                                                     ABIOTIC
                                                                  TRANSFORMATION
                                                                    (POSTULATED)
          Source: Adapted from Alexander, 1964 and Voudrais & Reinhard, 1986
          Figure 3-5. Abiotic/Biotic Humic Substance Formation Pathways
                                     91

-------
    The abiotic reaction leading to humic substance formation requires that
a catalyst be present in the soil system.  Catalysts that have been shown
to be effective in oxidizing organic compounds include Cu (Voudrius and
Reinhard, 1986) Mn, Fe and layer silicates (McBride, 1986).   These
catalysts are present in most soils as oxide coatings or mineral
components.  The presence of cations and anions on oxide surfaces can
reduce the reaction rate by limiting release of Mn (II) to solution (Stone
& Morgan, 1984a).

    Primary minerals such as olivines, pyroxenes,  and amphiboles have been
shown to accelerate polymerization rates to a greater extent than micas and
feldspars, whereas microclene and quartz were ineffective in catalyzing
polymerization.  McBride (1987) investigated the adsorption and oxidation
of catechol and hydroquinone by Fe and Mn oxides.   Oxidation of the organic
compounds was explained by the formation of a Fe (III)-hydroquinone complex
at the Fe-oxide surfaces for a sufficient period of time to allow for the
electron transfer.  The oxidized molecule subsequently is released into
solution by dissociation of the complex followed by rapid oxidation of Fe
(II).

    Reduction of Mn(IIl) and Mn(lV) by hydroquinone occurs on the oxide
surface (Stone and Morgan, 1984a).  Hydroquinone forms a surface complex
with the oxide prior to electron transfer, and the rate of the surface
chemical reaction is rate limiting, not the rate of surface complex
formation.

    In the oxidation of hydroquinone, Fe oxide serves as the catalyst
resulting in the reduction of Fe(lII).  McBride (1987) postulates that the
oxidized organic desorbs from the oxide or that electron transfer from the
reduced Fe occurs prior to 02 adsorption.  Significant quantities of Mn(ll)
were released  into solution upon phenol oxidation in McBride's (1987)
study. He concludes that Mn oxide was the primary oxidizing agent since
Mn(ll) oxidation by dissolved 02 is slow except at elevated pHs.

    In a study on the dissolution of Mn oxides by aromatic and aliphatic
substances, Stone & Morgan (1984b)  found that saturated alcohols, aldehydes,
ketones and carboxylic acid were non-reactive except for pyruvic and oxalic
acid.  Catechols, hydroquinones, methoxyphenols, resorcinols and ascorbate
all reduced and dissolved Mn oxides.  These molecules, except ascorbate,
form core  structure for humic substances and  the researchers, therefore,
conclude that  humic substances  should dissolve Mn oxides under natural
conditions.  Complexation of organics onto Fe and Mn oxides facilitates
electron transfer required for  abiotic transformation  (Stone & Morgan,
1984a).

    Longmire (1986) studied  the impact of leakage of approximately 39,000
gallons of gasoline on the reduction  of  iron  in soils.  The strongly
reducing conditions resulting at  this UST site resulted in dissolution of
iron, which was detected as uncomplexed  and complexed  Fe in the reduced
groundwater.   The dissolution of Fe in this case may be related to the
depression of  the  soil Eh due to  consumption  of 02 by microorganisms.
However, dissolution may also be due  to  abiotic reactions with catechol or
phenolics  in the  soil.   In either  case,  the ferrous  iron generated can serve
to  complex introduced organics  to  further the transformation processes.
                                         92

-------
    Voudrais and Reinhard (1986) indicate that obtaining mechanistic
information for reactions catalyzed by mineral surfaces is difficult
because:

    o  active surface sites are difficult to characterize and quantify
    o  competition for reaction sites by substrates occurs
    o  several products may be formed from a single substrate
    o  recovery of products from mineral matrix is difficult
    o  of the presence of undefined active phase
    o  reactions may be mass transfer limited

Summary

    Abiotic transformation of hydrocarbons in soils proceeds by a number of
reaction types including elimination, dehydrogenation, redox and
polymerization.  These abiotic reactions are closely related to the
microbial degradation pathways from which precursors to redox-
polymerization reactions are derived.  Determination of the rates of
reactions in soils for a specific loci may be quantifiable provided
estimates on the characteristics of the soil and contaminant can be
defined.  When mixtures of contaminants are involved, such as petroleum
hydrocarbons, however, the utility of such an exercise seems questionable.

    A systematic approach that allows a user to estimate the degree of
abiotic transformations in soil may be a useful tool for those involved in
remediation of petroleum-contaminated soil.  Soil factors that play a role
in abiotic transformation processes include:

    o  humic content and characterization
    o  particle size distribution
    o  moisture retention characteristics
    o  mineralogy including sesquoxide content
    o  redox potential, and
    o  pH

    Values for these parameters are easily obtained for soils, and general
ranges can be established for use in a guidance document, allowing an
estimate of the potential (low, medium, high) for abiotic transformations.

Applicability to Locus 3

    Moist soil conditions provide optimum conditions for migration of
contaminants to soil sites where abiotic oxidation can be initiated.

3.3  Storage Capacity in Locus

3.3.1  Introduction and Basic Equations

    The mass of contaminants per unit volume of soil is given by:

-------
                                                             o
where m  = mass of contaminants per cubic meter of soil (mg/m )

      C  = concentration of contaminants in water film of locus no. 3
       w
           (mg/L)

      0  = water-filled porosity (dimensionless, 0 < 9  <1)
       w                                              w


    A calculation of mass per unit weight of soil is given by:


                             mg = Cw 0w ' pb

where m  = mass of contaminants per kilograms of dry soil (mg/kg)
       O
      p,  = dry soil bulk density (kg/L)
    Selected values of 9  representative of soils at their field capacity
and wilting point are given in Figure 12-4 (in Section 12).  Selected
values of soil bulk density are given in Table 3-4.
                                 TABLE 3-4

                   SELECTED VALUES FOR SOIL BULK DENSITY
                                                                 Bulk    3
                   Soil Type                                Density  (g/cm )
    SURFACE SOILS FOUND IN AGRICULTURAL AREAS:

    A)  Well-Decomposed Organic Soil                          0.2 - 0.3
    B)  Cultivated Surface Mineral Soils                      1.25 - 1.40
    C)  Clay, Clay Loam, Silt Loam                            1.00 - 1.60
    D)  Sands and Sandy Loams                                 1.20 - 1.80

    MATERIALS USED IN ROAD & AIRFIELD CONSTRUCTION:

    A)  Silts and Clays                                       1.3 - 2.0
    B)  Sand and Sandy Soils                                  1.6 - 2.2
    C)  Gravel and Gravelly Soils                             1.8-2.3

    VERY COMPACT SUBSOILS                                Up  to 2.3 or  2.5
SOURCE:  Donigian  et  al.  (1984)
                                     94

-------
3.3.2  Guidance on Inputs for, and calculations of Maximum Values

3.3.2.1  Concentration of Contaminants in Water (C )

    Use the solubility for the liquid contaminant in water.  For the
hydrocarbon components of a typical gasoline (i.e., excluding consideration
of additives), a reasonable value for C  at 20°C is. 200 mg/1.  For gasoline
with oxygenated additives (e.g., ethanol or MTBE), a value of 2000 mg/1 is
recommended.

3.3.2.2  Soil Moisture Content:  Water-Filled Porosity (9 )

    Use the data shown in Figure 12-4.  Specifically, the line representing
the lower limit of the field capacity for different soils probably
represents a maximum for the amount of immobile pore water in locus no. 3.
This line shows 6  ranging from about 0.10 for sandy soils to about 0.30
for clayey soils.

3.3.2.3  Soil Bulk Density (p )
                             o
                                          3
    A recommended value for p,  is 2.0 g/cm .  This should be used for
compacted subsoils unless site-specific information indicates a higher or
lower value should be used.  See Table 3-7 for selected values of p, .

3.3.3  Guidance on Inputs for, and Calculation of Average Value

3.3.3.1  Concentration of Contaminants in Water (C )

    Use a value that is about 10 percent of the solubility limit.  This is
an arbitrary value, but the dilution by a factor of 10 from the maximum
value represents a reasonable amount of dilution with clean pore water.
For a gasoline with no additives, this value would be about 20 mg/1;  with
an oxygenated additive it would be about 200 mg/1.

3.3.3.2  Soil Moisture Content:  Water-Filled Porosity (6w)

    It is recommended that the line in Figure 12-4 that represents the
wilting point be used.  This line shows 0  ranging from about 0.05 for
sandy soils to 0.25 for clayey soils.

3.3.3.3  Soil Bulk, Density (p )
                              s
                                                        3
    Use same value as for "maximum" case (i.e., 2.0 g/cm  unless special
conditions apply.)

3.4  Example Calculations

3.4.1  Storage  Capacity Calculations

3.4.1.1  Maximum Value

    Estimate the maximum storage capacity for an additive-free gasoline in
a loam soil.  We use equation 3.12:

-------
                         m  = C   0  103
                          V    W   W

                 with   C  = 200 mg/L (recommended in Section 3.3.2)
                         w

                        6  =0.20 (lower limit of field capacity according
                         W   to Figure 12-4)

                 Thus,  my = (200)(0.20)(103) = 40,000 mg/m3

                           = 40 g/m3

                           = 20 mg/kg (with equation 3.13 and soil bulk
                             density of 2 kg/L)

3.4.1.2  Average Value

    Estimate the average (typical) storage capacity for an additive-free
gasoline in a sandy soil.  We use equation 3.12:

                        m  = C   0  103
                         V    W   W

                 with   C  = 20 mg/L (recommended in Section 3.3.J.,
                         w

                        0  = 0.05 (wilting point as shown in Figure 12-4)
                         w

                 Thus,  my = (20)(0.05)(103) = 1,000 mg/m3

                           = 1 g/m

                           = 0.5 mg/kg (with equation 3.13 and soil bulk
                             density of 2 kg/L)

3.4.2  Transport Rate Calculations

3.4.2.1  Diffusion Flux

    Estimate the diffusive flux of benzene through a 1-cm long (straight)
tube of pore water given a water temperature of 2°C, and concentrations of
benzene of 0 and 5 mg/L at either end of the tube.

    Since a tortuosity factor is not involved with a straight tube,
equation 3.1 is used:

                        J = D AC

                 with   D = 0.58 x 10"5 cm2/s @ 2°C (Table 3-2)

                       AC - 5 mg/L = 6.4 x 10~5 mol/L = 6.4 x 10~8  mol/cm3

                 Thus,  J = (0.58 x 10~5)(6.4 x 10~8)

                          = 3.7 x 10~13 mol/cm2s  = 2.9 x 10"11 g/cm2  s

                                    -4       2
                          = 2.5 x 10   mg/cm  day
                                    96

-------
    Note that the total surface area of all the pores in the soil exposed
to flowing soil gas would have to be known to calculate a flux in units of
mass per unit time.

3.A.2.2  Air/Water Concentrations

    Estimate the vapor phase concentration of toluene for a soil gas that
is in equilibrium with pore water containing 100 mg/L of toluene (mol. wt.
= 92).

    Equation 3.3 is used:

                        C  = H C
                         a      w

                 with   H  = 8 x 10~3 atm m3/mol @ 25°C (Figure 3-3)

                        C  = 100 mg/L = 1.09 x 10~3 mol/L =1.09 mol/m3
                         w

                 Thus,  C  = (8 x 10~3)(1.09) = 8.7 x 10~3 atm = 6.6 mm Hg
                         3

3.5  Summary of Relative Importance of Locus

3.5.1  Remediation

    Contaminants held in solution in the immobile pore water of the
unsaturated zone usually comprise a very small fraction of the total mass
of leaked material.  In spite of this low storage capacity, the locus may
play an important role, first in the long term retention of some
contaminants in the unsaturated zone, and second as an impedance to
remediation technologies such as vacuum extraction.

    The long term retention is associated with the fact that the water is
primarily held in the narrow pores away from the larger pores that may see
the cleansing action of infiltrating water or flowing soil gas.
Hydrocarbon contaminants of relatively low volatility and high solubility
(specifically the BTEX fraction) will have enhanced concentrations and long
residence times in this locus.

    The phenomenon of impedance to vacuum extraction is related to the
phenomenon of long term residence.  In this case, in addition to having to
yield up its initial contaminant load (acquired prior to vacuum
extraction), the waters in this locus will act as an absorbent to all
contaminant vapors, especially those of high water solubility.  This
ability to trap vapors mobilized by the vacuum extraction system will slow
down the remediation process.  As noted above, diffusion of dissolved
contaminants in the water of this locus probably controls the rate of mass
recovery in cases where there is no liquid contaminant (free product)
present.  Only when the soil is taken to dryness - not an impossible feat -
will this effect be eliminated.
                                    97

-------
3.5.2  Loci Interactions

    The major loci interactions are intermixing with (or advecting to)
mobile pore water (locus no. 12), volatilization into soil gas (locus no.
1), and sorption to soils (locus no. 4).  In all cases, diffusion of
contaminants through the aqueous phase may be a rate-limiting step.
Interactions with soil bacteria may be important in some cases, however,
the low water content associated with this locus may be detrimental to
bacterial activity and population growth.

3.5.3  Information Gaps

    Perhaps the most significant data gap for this locus is in the
understanding of the mobility of the aqueous phase, both in an absolute
sense (e.g., micropore water velocities) and a relative sense (i.e.,
relative to locus no. 12).  Without a knowledge of just how immobile the
locus no. 3 waters are, it is impossible to predict quantitatively the rate
of contaminant transport between them and other loci.

    The importance of biodegradation in this locus is also unclear due - as
mentioned above - to the possible hindrance of low water content on
bacterial activity and growth.

3.6  Literature Cited

Aiken, G.R., D.M. McKnight, R.L. Wershaw and P. MacCarthy (Eds.). 1985.  An
    Introduction to Humic Substances in Soil, Sediment and Vater.  In:
    Humic Substances in Soil Sediment and Water.  John Wiley and Sons, New
    York.

Alexander, M. 1961. Introduction to Soil/Microbiology.  John Wiley and
    Sons, New York.

Amdurer, M., R.T. Fillman, J. Roetzer, and C. Russ. 1986.  Systems to
    Accelerate In-Situ Stabilization of Waste Deposits.  EPA/540/2-86-002.

Bailey, P.S. 1972.  Organic Groupings Reactive Toward Ozone:  Mechanisms in
    Aqueous Media.  In;  Ozone in Water and .Wastewater Treatment.  F.L.
    Evans (Ed).  Ann Arbor Scientific Publishers, Ann Arbor, MI.

Bonn, H.L., B.L. McNeal and G.A. O'Connor. 1979.  Soil Chemistry.
    Wiley-Interscience, New York.

Catroux, G. and M. Schnitzer.  1987.  Chemical, Spectroscopic and
    Biological Characteristics of the Organic Matter in Particle Size
    Fractions Separated from an Aquisol.  Soil Sci. Soc. America J.,
    51(5):1200-1206.

Chen, Y. and M. Schnitzer.  1976.  Viscosity Measurements on Soil Humic
    Substances.  Soil Sci. Soc. America J., 40: 866-872.

Cline, P.V., J.J. Delfino and W.J. Cooper. 1986.  Hydrolysis of
    1,1,1-Trichloroethane; Formation of  1,1-Dichloroethene.  In;
    Proceedings of the NWWA/API Conference on Petroleum Hydrocarbons and
    Organic Chemicals in Groundwater.  November 12-14.  Houston, TX.

                                    98

-------
Cooper, W.J., M. Mehran, O.J. Riusech, and J.A. Joens.  1987.  Abiotic
    Transformations of Halogenated Organics. 1. Elimination Reaction of
    1,1,2,2-Tetrachloroethane and Formation of 1,1,2-Trichloroethene
    Enviro. Sci. Technol.  21(11):1112-1114.

Donigian, A.S., Jr., T.Y.R. Lo and E.W. Shanahan.  1984.  Groundwater
    Contamination and Emergency Response Guide, Part III, Noyes
    Publications, Park Ridge, NJ.

Frankenberger, W.T. Jr. and J.B Johanson. 1982.  Influence of Crude Oil and
    Refined Petroleum Products on Soil Dehydrogenase Activity.  J. Environ.
    Qual, 11(4):602-607.

Ghosal, M. and E.S.K. Chain. 1985.  An Evaluation of Aromatic Fraction in
    Humic Substances.  J. Soil Sci. Soc. Am, 49:616-618.

Gossett, J.M. and A.H. Lincoff. 1981.  Solute-Gas Equilibria in
    Multi-Organic Systems.  Report No. AFOSR-TR-81-0858 prepared for the
    Air Force Office of Scientific Research by Cornell University, Ithaca,
    NY.  (Available from NTIS as AD A109 082.)

Harris, J.C. 1982.  Rate of Hydrolysis.  In:  Handbook of Chemical Property
    Estimation Methods, V. Lyman, V. Reehl and D. Rosenblatt (Eds.) McGraw
    Hill, New York.

Hayduk, W. and H. Laudie. 1974.  Prediction of Diffusion Coefficients for
    Nonelectrolytes in Dilute Aqueous Solutions.  AICLE Journal,
    20(3):611-15.

Hoigne, J. and H. Bader. 1976.  The Role of Hydroxyl Radical Reactions in
    Ozonation Processes in Aqueous Solutions.  Vater Resources, 10:377-386.

Huang, P.M. 1987.  Impact of Mineralogical Research on Soil and
    Environmental Sciences.  In;  Future Developments in Soil Science
    Research.  SSSA, Madison, VI.

Larson, R.A., T.L. Batt, L.L. Hunt and K. Rogenmuser. 1979. Photo Oxidation
    Products of Fuel Oil and Their Antimicrobial Activity.  Env. Sci.
    Technol., 13(8):965-969.

Leighton, D.T., Jr. and J.M. Calo. 1981.  Distribution Coefficients of
    Chlorinated Hydrocarbons in Dilute Air-Water Systems for Groundwater
    Contamination Applications.  J. Chem. Eng. Data, 26:382-85.

Lion, L.W. and D. Garbarini. 1983.  Partitioning Equilibria of Volatile
    Pollutants in Three-Phase Systems.  Report No. ESL-TR-83-51,
    Engineering and Services Laboratory, Air Force Engineering and Services
    Center, Tyndall AFB, Florida.  (Available from NTIS as AD A137 207.)

Longmere, P. 1986.  Iron Dissolution Resulting from Petroleum - Product
    Contamination in Soil and Groundwater.  1.  Thermodynamic
    Considerations.  Proceedings of NWA/API Conference on Petroleum
    Hydrocarbons and Organic Chemicals in Groundwater.  November 12-14.
    Houston, TX.
                                    99

-------
Mackay, D. and W.Y. Shiu.  1981.  A Critical Review of Henry's Law
    Constants for Chemicals of Environmental Interest.
    Data, 10(4):1175-99.
       J.  Phys. Chem. Ref.
Mackay, D.,  W.Y. Shiu, and R.P. Sutherland. 1979.  Determination of
    Air-Water Henry's Law Constants for Hydrophobic Pollutants.  Env. Sci.
    Technol., 13(3):333-337.

Maloney, S.W., J. Marem, J. Mallerialle, and F. Fiessinger. 1986.
    Transformation of Trace Organic Compounds in Drinking Water by
    Enzymatic Oxidative Coupling.  Env. Sci. Technol., 20(3):249-253.

Maloney, S.W., I.H. Suffet, K. Bancroft, and H.M. Neukrug. 1985.  Ozone-GAC
    Following Conventional U.S. Drinking Water Treatment.  JAWWA, 8:66-73.

McBride, M.B.  1987.  Adsorption and Oxidation of Phenolic compounds by
    Iron and Manganese Oxides.  Soil Sci. Soc. Am. J. Vol. 51.
Mill, T., D.G. Hendry and H. Richardson.  1980.
    Natural Waters.  Science, 207:886-887.
Free-Radical Oxidants in
Rice, R.G., C.M. Robson, G.W. Miller and A.G. Hill.  1981.  Uses of Ozone
    in Drinking Water Treatment.  J. Amer. Water Works Assn., 73(1): 44-57.

Stevenson, F.J.  1985.  Geochemistry of Soil Humic Substances.  In:  Humic
    Substances in Soil Sediment and Water.  John Wiley and Sons, New York.

Stone, A.T. and J.J. Morgan.  1984a.  Reduction and Dissolution of
    Manganese (II) and Manganese (IV) Oxides by Organics, 1:  Reaction with
    Hydroquinone.  Environ. Sci. Technol. 18:450-456.

Stone, A.T. and J.J. Morgan.  1984b.  Reduction and Dissolution of
    Manganese (II) and Manganese (IV) Oxides by Organics.  2:  Survey of
    the Reactivity of Organics.  Environ. Sci. Technol. 18:617-624.

Tabatabi, M.A.  1982.  Soil Enzymes.  In: Methods of Soil Analysis.
    Agronomy Monograph No. 9, Part 2, 2nd edition, A.L. Page, R.H. Miller
    and D.R. Keeny (eds.), American Society of Agronomy, Madison, VI.

Thomas, R.G.  1982.  Volatilization from Water.  In:  Handbook of Chemical"
    Property Estimation Methods, W. Lyman, W. Reehl and D. Rosenblatt
    (eds.) McGraw-Hill, New York.

Thomas, J.M. and C.H. Ward.  1989.  In Situ Biorestoration of Organic
    Contaminants in  the Subsurface.  Environ. Sci. Technol., 27(7):
    760-766.

Tucker, W.A. and L.H. Nelken. 1982.  Diffusion Coefficients  in Air and
    Water.  In;  Handbook of Chemical Property Estimation Methods, W.
    Lyman, W. Reehl  and D. Rosenblatt (eds.), McGraw-Hill, New York.

Valentine, R.L. 1987. Nonbiological Transformation.  In;  Vadose Zone
    Modeling of Organic Pollutants.  Lewis Publishers Inc.,  Chelsea, MI.
                                     100

-------
Vogel, T.M. and P.L. McCarty.  1987a.  Abiotic and Biotic Transformations
    of 1,1,1-Trichloroethane under Methanogenic Conditions.  Env. Sci.
    Technol., 21(12):1208-1213.

Vogel, T.M. and P.L. McCarty.  1987b.  Rate of Abiotic Formation of
    1,1-Dichloroethylene from 1,1,1-Trichloroethane in Groundwater.  J.
    Contam. Hydro., 1:299-308.

Voudrias, E.A. and M.R. Reinhard. 1985.  Abiotic Organic Reactions at
    Mineral Surfaces.  In;  Geochemical Process at Mineral Surfaces.  Davis
    and Hayes (Eds.), ACS Symposium Series No. 323.

Wang, T.S.C., P.M. Huany, C.H. Chou, and J.H. Chem. 1986.  The Role of Soil
    Minerals in the Abiotic Polymerization of Phenolic Compounds and
    Formation of Humic Substances.  In;  Interactions of Soil Minerals with
    Natural Organics and Microbes.  Huang and Schutzer (eds.),  SSSA Spec.
    Pub. No. 17, Madison, VI.

Witherspoon, P.A. and L. Bonoli. 1969.  Correlation of Diffusion
    Coefficients for Paraffin, Aromatic, and Cydoparaffin Hydrocarbons in
    Water.  Ind. Eng. Chem. Fundam., 8(3):589-91.
                                    101

-------
                          SECTION 4.  LOCUS NO. A

                                 CONTENTS
                                                                   Page No.

List of Tables	     103
List of Figures	     103
4.1  Locus Description	     104

     4.1.1  Short Definition	     104
     4.1.2  Expanded Definition and Comments	     104

4.2  Evaluation of Criteria for Remediation	     104

     4.2.1  Introduction	     104
     4.2.2  Mobilization/Remobilization	     107

            4.2.2.1  Partitioning into Mobile Phase	     107
            4.2.2.2  Transport with Mobile Phase	     116

     4.2.3  Fixation	     116

            4.2.3.1  Partitioning onto Immobile
                      (Stationary) Phase	     116
            4.2.3.2  Other Fixation Approaches	     116

     4.2.4  Transformation..	     117

            4.2.4.1  Biodegradation	     117
            4.2.4.2  Chemical Oxidation	     117

4.3  Storage Capacity in Locus	     117

     4.3.1  Introduction and Basic Equations	     117
     4.3.2  Guidance on Inputs for, and Calculation of,
              Maximum Value	     117
     4.3.3  Guidance on Inputs for, and Calculation of,
              Average Values	     118

            Comments	     118

4.4  Example Calculations	     118

     4.4.1  Storage Capacity Calculations	     118

            4.4.1.1  Maximum Value Calculations	     118
            4.4.1.2  Average Value Calculations	     119
                                    102

-------
                                (Continued)



                                                                   Page No.

     4.4.2  Transport Rate Calculations	     119

            4.4.2.1  Time Required to Desorb 50, 90, and 95
                       Percent of Contaminant	     120
            4.4.2.2  Retardation of Bulk Transport	     121

4.5  Summary of Relative Importance of Locus	     122

     4.5.1  Remediation	     122
     4.5.2  Loci Interaction	     122
     4.5.3  Information Gaps	?•	     123

4.6  Literature Cited	     123
                                  TABLES


4-1  Ranges of Values of Porosity                                     110

4-2  Key Parameters Affecting Sorptive Behavior                       111

4-3  Estimated Soil Sorption Constants for Selected Gasoline
     Constituents                                                     113

4-4  Selected Soil Characteristics and Sorption Coefficient
     Data                                                             115

4-5  Loci Interactions With Hydrocarbons Sorbed Onto Soil
     Surfaces                                                         123
                                  FIGURES
4-1  Schematic Cross-Sectional Diagram of Locus No. 4 -
     Contaminants Sorbed to "Water-vet" Soil Particles or Rock
     Surfaces (After Migrating Through Water) in Either the
     Unsaturated or Saturated Zone                                    105

4-2  Schematic Representation of Important Transformation and
     Transport Processes Affecting Other Loci                         106
                                    103

-------
                          SECTION 4 - LOCUS NO. 4
4.1  Locus Description

4.1.1  Short Definition

    Contaminants sorbed to "water-wet" soil particles or rock surface
(after migrating through the water) in either the unsaturated or saturated
zone.

4.1.2  Expanded Definition and Comments

    In this locus definition, contaminants must first be dissolved in water
and then sorbed to a soil or rock surface.  The concentration of sorbed
contaminants is, therefore, limited by the initial dissolution process.
Although a partitioning process involving fairly rapid sorption and
desorption is presumed to exist, the sorbed contaminants are considered
immobile.  The sorbed contaminants are either absorbed into a thin layer of
naturally-occurring organic matter surrounding the soil particles, or
adsorbed in a thin layer to exposed mineral (e.g., clay) surfaces.  They
are not able to form a separate liquid phase on the particle's surface.
Sorbed contaminants that subsequently diffuse into the solid rock or
mineral particles are covered by locus no. 10.  Figures 4-1 and 4-2 present
a schematic cross-sectional diagram of locus no. 4 and a schematic
representation of the transformation and  transport processes affecting
other loci, respectively.

4.2  Evaluation of Criteria for Remediation

4.2.1  Introduction

    Sorption of contaminants onto soil surfaces is generally treated by
assuming equilibrium conditions between the sorbed and the solution phase.
At equilibrium, the relationship between  the sorbed and solution phase
concentration can be described by the Freundlich isotherm (described
below).

    There  is considerable ongoing debate  among researchers as to the exact
mechanisms involved in governing sorptive behaviors.  Thermodynamically,
equilibrium sorption is defined as the state in which the fugacities of
contaminants in the sorbed and solution phases are equal.  The general
approach in defining sorptive behavior has been to identify key parameters
that affect sorption and desorption.  An  understanding of these key
parameters then facilitates  the assessment of remedial actions.

    Mobilization of sorbed contaminants can be accomplished by desorption
into the aqueous phase in  both saturated  and unsaturated zones.
Immobilization  can be accomplished by enhancing sorption.  Biodegradation
and other  transformation processes can, for many compounds, act as a
removal mechanism.  However,  the rates for such processes are somewhat slow
and biodegradation is not  often attempted as  the sole corrective action.
                                     104

-------
     UNSATURATED
         ZONE
WATER
TABLE
          I
          1
      SATURATED
         ZONE
     NOTE: NOT ALL PHASE BOUNDARIES ARE SHOWN.
                                      LEGEND
                  CONTAMINANT VAPORS

                  AIR (PLAIN WHITE AREAS IN
                  SATURATED ZONE)

                  WATER (MOBILE PORE WATER
                  IN UNSATURATED ZONE)
            AAAA  WATER FILM
-Zi WATER IN SATURATED ZONE

X   BIOTA (MICROBIOTA INCLUDED)

    SOIL PARTICLE

 "•  CONTAMINANTS

 f   COLLOIDAL PARTICLE
  Figure 4-1. Schematic Cross-Sectional Diagram of Locus No. 4 -
            Contaminants sorbed to "Water-Wet" Soil Particles or
            Rock Surfaces (After Migrating Through Water) in Either
            the Unsaturated or Saturated Zone.
                                 105

-------
     (BIOTA)
    LOCUS NO
11-SORBEDTOBIOTA
                   DEPURATION
                UPTAKE
                               LOCUS NO. 4
                           (CONTAMINANTS SORBED TO
                          "WATER-WET" SOIL PARTICLES
                             OR ROCK SURFACES IN
                          UNSATURATED OR SATURATED
                                   ZONE)
                 SORPTION
     (ROCK)
    LOCUS NO
 10- DIFFUSED INTO
    MINERAL GRAINS
    OR ROCKS
                DIFFUSION
                             SORPTION
                            DESORPTION
        (WATER)
       LOCUS NO. •
 3- DISSOLVED IN WATER FILM
 8- DISSOLVED IN GROUND-
   WATER
12- DISSOLVED IN MOBILE
   PORE WATER
DESORPTION
       Figure 4-2. Schematic Representation of Important Transformation
                   and Transport Processes Affecting Other Loci.
                                         106

-------
4.2.2  Mobilization/Remobilization

4.2.2.1  Partitioning into Mobile Phase

    The only mobile phase in contact with this locus is water (loci nos.  3,
8, or 12).  The contaminants initially dissolved in water are considered  to
have sorbed into a thin layer of naturally occurring organic matter
surrounding the soil particles or adsorbed in a very thin layer to exposed
clay surfaces.  A partitioning process involving rapid sorption and
desorption is presumed to be controlling the movement of contaminants into
and out of the immobile surface region of the soil particles.

    Desorption of sorbed contaminants from the soil surfaces into the
mobile aqueous phase may be expressed by the Freundlich equation, which
states:

              CS = Y(CV)N                                           (4.D


    where     C  = concentration of contaminant in soil (mg/kg)
               s
              C  = concentration of contaminant in water (mg/L)
               w
              K, = sorption coefficient (L/kg)

              N  = measure of deviation from linearity

    Rao and Davidson (1980) compiled a list of mean values of N for several
pesticides and pesticide related compounds.  The mean values were derived
from sorption experiments conducted with a number of soils.  The reported
mean values range from 0.70 to 1.18 and the overall mean value is 0.87.  N
values from experiments conducted with gasoline constituents are not
readily available in the literature.

    The sorption coefficient K, is determined by plotting the sorbed phase
concentration versus the solution phase concentration.  For dilute
solutions such as those normally encountered in natural environments,
equation 4.1 may be written as the linear equation (i.e., with N = 1) shown
below:

                       Cs  =  K.Cw                                    (4.2)

    where    K - sorption coefficient for linear isotherm (L/kg)

    K, or K values for the same compound can range over several orders of
magnitude depending on soil characteristics.  The estimation technique for
K is discussed later in this section.

    Karickhoff et al. (1979) observe that for hydrophobic organics (such  as
most gasoline constituents), if the concentration of contaminants in the
aqueous phase is below 10~  mole/liter, or is less than one-half of the
water solubility (whichever is lower), then the sorption isotherms are
linear.
                                    107

-------
    Rao and Davidson (1980) indicate that for a solution phase
concentration of 10 mg/1 and a N value of 0.7, the amount adsorbed could be
over-predicted by a factor of two by assuming linearity.  For large
solution concentrations, the amount adsorbed could be overestimated by an
order of magnitude or more.  Concentration of contaminants in water and
soil is expected to be high in the region impacted by underground releases.
For such systems, equation 4.1 is more appropriate for defining
sorption-desorption.

    Kinetically, sorption has been viewed as a two step process involving
rapid initial sorption followed by a slower sorption process.  Karickhoff
(1980) attributed the first step to fast adsorption onto the external soil
surfaces and the second step to slow diffusion of the solute into the
interstices of the soil particle (e.g., Locus 10).  It is apparent that the
contact time between the solid and the solution phase should be long enough
for reaching equilibrium conditions.

    Several models exist that describe solute transfer between the immobile
and mobile soil-water regions in terms of the two stage process.  While the
first stage is usually treated as instantaneous, the second stage is either
a kinetically-controlled or a diffusion-controlled process.

    Wu and Gschwend (1986) describe sorption kinetics in terms of
intraparticle diffusion.  They describe an effective intraparticle
diffusivity as:

              Deff =      D/
                        (1-0). K.p                                     (4<3)

                                                            2
    where     f> ff - effective Intraparticle Diffusivity  (cm /s)

              D    - pore Fluid Diffusivity of the Sorbate  (cm  /s)

              6    - porosity (dimensionless)
                                             3
              K    = sorption coefficient (cm /g)
                                                          3
              p,   = bulk Density of  the Dry Sorbent  (g/cm  )


    D ff can be calculated  from this  equation if values for other
                                                                  -5    2
parameters are available.   D  for neutral organics is  typically 10    cm /s

at 25°C.  A value of 2.5 g/cm  may  be used as a rough  estimate  for  p
                                                                    S
(Arthur D. Little, 1987).   Wu and Gschwend (1986) chose a value of  0.13 for

the porosity as typical for silts.  K may be estimated using equation 4.9.
    Wu and Gschwend  (1986) showed  that  sorption  rates were  lower  for  larger
soil aggregates and  more hydrophobic  compounds.  The results  also indicated
that desorption rates  can be described  in  terms  of  reversible exchange
mechanisms.  They  further stated  that if the  particle size  distribution  is
sufficiently narrow  to allow estimation of  an average particle size,  then
                                     10H

-------
the analytical solution to equation 4.3 may be used to estimate the time
scale of sorption and desorption.  An estimate of time for 50 percent
desorption or sorption can be estimated from (Arthur D. Little, 1987),

              t0i5 = 0.031 r2/Deff                                    (4.4)


    t~ c = time for 50 percent of the chemical to desorb or time for 50

           percent of the ultimate amount to sorb (sec)


    r    = average particle radius (cm)

                                        2
    D r  = intraparticle diffusivity (cm /s)
                                                 _
    Wu and Gschvend (1986) gave a value of 2 x 10   cm for the particle
radius of river sediments.  However, the range of particle sizes for
natural soils and sediments can span an order of magnitude or more.  For
natural soils, a weighted average radius should be used.  The weighted
average radius can be calculated by using the equation (Arthur D. Little,
1987),

              r  =  (Zr^.m./m )1/2                                   (4.5)

    where     r. = mean radius of the i-th size fraction (cm)

              m. = mass of particles in the i-th fraction (gm)

              m  = total mass of particles (g)


Values of r., m. ,  and m  can be obtained by conducting a standard particle

size analysis.


    Equations 4.3 through 4.5 thus can provide valuable information on the
time required for immobilizing contaminants sorbed onto soil surfaces,
provided equilibrium conditions exist.  Arthur D. Little (1987) further
presented equations for estimating time required to desorb 90 (t~ g) and 95
(tn QC) percent of the sorbed contaminant as follows,

    t0>9 (sec)          =       0.18 r2/Deff                          (4.6)

    t0>95 (sec)         =       0.25 r2/Deff                          (4.7)


    Bulk transport of hydrocarbons dissolved in groundwater is discussed in
detail in section 8.  A term which is relevant for discussions under this
loci is the retardation factor, R.  Retardation of a contaminant's movement
relative to the bulk mass of groundwater can be stated as follows (Huyakorn
and Finder, 1983):
                                    1.09

-------
              R   =  vw/vc = l + Kpb/G
    where     v  = average velocity of water (cm/sec)
               w
              v  = average velocity of the contaminant (cm/sec)
               C                           3
              K  = sorption coefficient (cm /g)
                                          3
              p,  = soil Bulk Density (g/cm )

              6  = soil Porosity (dimensionless)


The range of porosity values for natural soils is listed in Table 4-1.


    In general, sorption and desorption are often assumed as having the
same partition coefficient.  Several researchers have stated that such
assumptions are not valid.  Irreversibility and hysteresis have been
observed indicating that the adsorption and desorption mechanisms might be
different.  Jaffe and Ferrara (1983) stated that for hydrophobic substances
with high carbon content, if the product of sorption coefficient and soil
density is larger than one, then desorption rate is slower than the rate of
sorption.  Several other researchers, however, have attributed this
phenomenon to experimental conditions.

                                 TABLE 4-1

                        RANGE OF VALUES OF POROSITY
              Unconsolidated Deposits         Porosity, %

                   Gravel                        25-40
                   Sand                          25-50
                   Silt                          35-50
                   Clay                          40-70

              Rocks
Fractured basalt
Karstlimestone-
Sandstone
Limestone, dolomite
Shale
Fractured crystalline rock
Dense crystalline rock
5-50
5-50
5-30
0-20
0-10
0-10
0-5
              Source:  Freeze and Cherry (1979)
                                    110

-------
     The key parameters that influence sorptive behavior are listed in Table
 4-2.   Many studies have established relationships between a sorption
 property,  K   (sorption coefficient normalized for organic carbon) with a
 solution property (solubility or octanol/water partition coefficient, K  ).
 For a complete list of such relationships,  the reader is referred to A.D.
 Little,  Inc.  (1987).

     The K   values so obtained,  can then be used to estimate the sorption
 coefficient by using the relationship.

                      K   =   K   x  f                                 (4.9)
                               oc     oc

     Where f   is the weight fraction of organic carbon in the soil.  Soil
 carbon content can span over two orders of  magnitude.   For example, in
 surface soils of Oklahoma and Pennsylvania, organic carbon ranged from 0.5
 to over 6 percent (Banerjee et al., 1985a;  USDA,  1960).  The range is
 probably much higher spanning greater areas.   In the saturated zone, most
 soils are expected to have a carbon content less than 0.1 percent (Newsom,
.1985).   Banerjee et al. (1985a)  and Banerjee et al. (1985b) measured
 organic carbon contents of two subsurface core samples collected from
 surface down to same depth in the subsurface.   In one set of 20 contiguous
 sections of soil, organic carbon content ranged from 0.08 to 0.89 percent.
 In another set of six samples collected from surface down to the saturated
 zone at 90 cm intervals, the organic carbon content of the surface soil was
 reported at 1.33 percent and the other five samples, including two from the
 saturated zone, had carbon content ranging  from 0.03 to 0.05 percent.
 Although organic carbon is generally associated with the finer sized
 particles, the organic carbon content can not  be estimated from other soil
 properties.

                                  TABLE 4-2

                 KEY PARAMETERS AFFECTING SORPTIVE BEHAVIOR
 Soil  Properties           Solute Properties        Solution Properties

 Organic matter content    Water solubility        Dissolved organic carbon
                                                  (both true solutions and
                                                  colloids)

 Clay  content              Octanol/water           Temperature
                          partion coefficient

 Cation exchange           Polarity                Salinity
 capacity

 Surface area              Chemical class          pH

                                                  Presence of co-solvents

                                                  Presence of other solutes
                                     111

-------
Organic carbon content can be measured in the laboratory, and it is
strongly recommended that the traction organic carbon content to be used in
equation 4.9 be a measured value.

    For hydrocarbons such as gasoline dissolved in water, the contaminants
should be viewed as the individual constituents of gasoline. For a few
constituents, measured K   values are available in the literature.  One
                        oc
convenient source of estimated values is the Superfund Public Health
Evaluation Manual (U.S. EPA, 1986).

    K   values can also be estimated from equations presented in Arthur D.
Little? Inc. (1987) as follows:

              log K     =  0.779 log K    +  0.46                    (4.10)
                   O C,                 \J W
              log KQC   = -0.602 log S    +  0.656                   (4.11)


    where     S   = water solubility (moles/L)

              K   = octanol/water partition coefficient  (dimensionless)
               ow


    Experimentally measured K   values foi gasoline constituents are not
readily available in the literature.

    It is generally believed that for hydrophobic compounds, the
relationships based on K   are superior  to those based on water solubility.
However, for gasoline constituents with  low K   values,  solubility-based
relationships are probably superior to those Based on K  (W. Lyman, pers.
comm.)  The sources for K   values should be the Superfund  Public Health
Evaluation Manual, equation 4.10, or equation 4.11.

    The chemical species that are most likely to desorb  from the soil
surfaces are also the ones  that are least likely to sorb.   These include
the compounds with high water solubility and low K   values.  Water
solubility, log K   and estimated K   values, for gasoline  constituents are
listed in Table 4™.               °C

    The estimation techniques described  in the preceding paragraphs are
applicable for most soils in the unsaturated zone where  the carbon content
is expected to be greater than 0.1 percent.  Several researchers
(Schwarzenbach and Uestall, 1981; Karickhoff, 1984) have shown  that the K
relationships may not be valid for soils with carbon content less  than O.I
percent, which is the type  of soil expected  to be present in the saturated
zone.  McCarty et al.  (1981) state  that  sorption studies conducted with
pure minerals indicate that hydrophobic  solutes are sorbed  by inorganic
surfaces.  They defined a critical level of  organic carbon  (f   *)  in
inorganic matrices below which the  inorganic phase contributes  to  the  total
sorption.   f  * may be expressed as,

              f  *  =   As       1
               OC
                         200  K   0.84                                  (4.12)
                              ow
                                     112

-------
             TABLE 4-3

 ESTIMATED SOIL SORPTION CONSTANTS
FOR SELECTED GASOLINE CONSTITUENTS
GASOLINE COMPONENT
N-Butane
Isobutane
N-Pentane
Isopentane
1-Pentene
N-Hexene
1-Hexene
2-Methylpentane
Cyclohexane
Benzene
N-Heptane
2-Methylhexane
Methylcyclohexane
Toluene
N-Octane
2 , 4-Dimethylhexane
Ethylbenzene
M-Xylene
2,2, 4-Tr ime thylhexane
1,3, 5-Trimethlyhexane
2,2,5, 5-Tetrame thylhexane
1,4-Diethylbenzene
Dodecane
Log
2.89
2.76
3.39
3.37
2.84
4.00
3.39
3.80
3.44
2.13
4.66
4.41
3.97
2.69
5.18
4.82
3.15
3.20
5.23
3.42
5.64
4.35
7.12
Water5
Solubility
at 25°C
K a
°W (mg/L)
61.4
48.9
41.2
48.5
148
12.5
50
14.2
59.7
1780
2.68
2.54
15
537
0.66
1.5C
167
162
0.8°
72.6
0.13C
15C
0.005
No Additives Considered
a. From Leo (1983)
b. From Mackay and shiu (1981), except as noted
c. Estimated by Lyman (personal communication, 1987)
d. Based upon limited experimental data, the values
solubility are considered more reliable.
Soil Sorption .
Constant, Koc (L/kg)
Estimated From:
Kow S
490
420
910
880
460
1,900
910
1,500
960
190
4,300
3,200
1,800
380
8,200
5,200
680
720
8,700
940
14,000
2,900
88,000
estimated
240
270
320
300
180
600
320
560
290
62
1,300
1,300
580
110
2,600
1,800
200
210
2,500
320
5,900
670
28,000
from
                113

-------
                                                    2
    where     A   =  silica specific surface area (m /g) (McCarty et al.,
                                               2
                     1981, used a value of 13 m /g)

              K   =  octanol/water partition coefficient
               ow


    For benzene, f  * = 0.001,  which translates to an organic carbon
content of 0.1 percent.  For compounds with higher K   values, f  * would
be lower indicating that carbon-based sorption is more important tor those
compounds.

    In such low carbon soils, the sorption coefficient may be expressed as
(McCarty, et al., 1981):

              K  =  K   xf   + K .  x f .                             (4.13)
                     OC    OC    Ol     01


    where     K   = sorption coefficient (L/kg)

              f  . = inorganic fraction

              K  . = partition coefficient  in the inorganic phase (L/kg)


    The inorganic fraction or clay content of natural soils may range from
less than 1 to greater than 99 percent.

    At present, relationships are not available for estimating K .  based on
solute properties.  However, qualitative and semi-quantitative estimates of
the contributions to sorption due to the inorganic fraction of soils is
possible.  Karickhoff  (1984) states that if the ratio of the  total clay
content to the carbon  content is greater than 60, then the contribution due
to mineral components  are significant to the total sorption of neutral
organics.  Another study (Banerjee et al., 1985b) was conducted to determine
the magnitude of the non-carbon based sorption of benzene, a gasoline
constituent, on low carbon soils.  The extent of carbon-based sorption and
K   can be established using the soil sorption coefficient determined using
tne soil with 0.26 percent carbon content.  The organic carbon content, clay
content, and sorption  coefficients for benzene on these soils are listed in
Table 4-4.  The KQ  value so determined was 46 (i.e., K   = 0.12/0.0026 = 46
since sorption to inorganic fraction not significant).  This K   value is
preferred over those listed in Table 4-4,  since all other sorption
coefficient values were determined under the same experimental conditions.
The partition coefficient, K ., can thus be calculated by substituting
appropriate values into equation 4.13.  For example, for soil B-l,

    0.038     =   46 x  0.031  +  K .  x  11
                                   01
                         100              100
Therefore, KQ. = 0.22
                                    114

-------
                                 TABLE 4-4

        SELECTED SOIL CHARACTERISTICS AND SORPTION COEFFICIENT DATA
Soil
J-10
B-l
C-l
N-6
foc
(X)
0.26
0.031
0.031
0.028
Clay

-------
    where     X     = mole-fraction solubility of the contaminant in
                      mixed solvent

              d     = constant related to the contaminants melting point


    Nkedi-Kizza et al. (1985) used a value of 0.83 for a.   An example
calculation, showing the effect of organic co-solvent on contaminant
solubility is presented in Section 8.

    The effect of temperature on sorption coefficients of  neutral organics
for soils in the subsurface does not appear to be significant.  Ambient
temperature in the subsurface environment is not expected  to vary beyond a
factor of two.  Most researchers (Wu and Gschwend, 1986; Wauchope et al.,
1983) found little or no effect for temperature variations over a factor of
two.

    For solutions containing significant amount of inorganic salts, such as
groundwater affected by salt water intrusion, sorption potential is
expected to increase.

    Several researchers have indicated that competitive sorption may occur
for organic cations, chlorinated phenols and inorganic ions.  Sorptive
competition between immobile adsorbates and mobile colloidal adsorbates may
also occur.  However, results from several studies indicate that these
phenomena may not be important for neutral organics.  Chiou et al. (1983)
showed that nonionic organics sorb on soil independently from mixtures.
However, Abdul and Gibson (1986) noted competitive sorption in experiments
conducted with fluorene and naphthalene.  They observed that K values for
naphthalene and fluorene were greater by a factor of 1.1 and 1.3,
respectively, in single compound experiments compared to experiments having
a mixture.  Considering the imprecision associated with determining
sorption coefficients, event under similar experimental conditions, the
extent of decreased sorption coefficients observed for multiple solutes
does not appear to be significant.

4.2.2.2  Transport with Mobile Phase

Transport of contaminants dissolved in water is discussed in Sections 8 and
12.

4.2.3  Fixation

4.2.3.1  Partitioning onto Immobile (Stationary) Phase

    Contaminants adhering to  the soil surface are already considered to be
immobile.   Sorption, however, is not an effective mechanism for fixation
since desorption into the mobile phase could continue for an extended
period of  time.

4.2.3.2  Other Fixation Approaches

    None of  the available processes can significantly affect  fixation.
                                     116

-------
4.2.4  Transformation

4.2.4.1  Biodegradation (Details in Section 11)

    Since sorbed contaminants are in close contact with a dissolved aqueous
phase, contaminants in locus no. 4 are very likely subjected to
biodegradation.  The concentrations in the dissolved and sorbed phases are
related by a partition coefficient, as explained earlier.  Therefore, as
microbes remove contaminants from the aqueous phase, more contaminant
molecules desorb into aqueous phase to allow continued biodegradation.

    It is important to realize, however, that transport of necessary
substrates to the microbes, including nitrogen, phosphorous, and
particularly oxygen, may limit  the rate of biodegradation.  This is
especially true for oxygen transport below the water table.

4.2.4.2  Chemical Oxidation (Details in Section 3.2.4.2)

    Under wet soil conditions, anaerobic conditions can develop resulting
in a shift in pH.  Such conditions can increase the occurrence of reduced
clay-metal complexes which can catalyze abiotic transformations and lead to
polymerization.

4.3  Storage capacity in Locus

4.3.1  Introduction and Basic Equations

    The maximum amount oi gasoline dissolved in water that can be held at
the soil surfaces can be stated as,

              C   =  K  C
               s         w


    where     C   =  concentration of contaminants in soil (mg/kg)

              K   =  sorption coefficient (L/kg)

              C   =  concentration in aqueous phase (mg/L)


4.3.2  Guidance on Input for, and Calculations of Maximum Value

    o    The sorption coefficient can be estimated from equation 4.9, which
         states

              K  =   K   x f
                      oc    oc

    o    It is recommended that fraction organic carbon be a measured
         value.  For soils with carbon content of less than 0.1 percent,
         assume 0.1 percent carbon to maximize sorption.

    o    For retention of gasoline, assume C  = water solubility of
         gasoline.  Also, estimate K   from equation 4.11.
                                    117

-------
    o    K   values for gasoline constituents are listed in Table 4-3.  For
         constituents not listed in Table 4-3,  obtain literature value or
         use equations 4.10 or 4.11.

    o    For maximum value calculation,  assume water solubility of gasoline
         as 200 mg/1 (source of this value is discussed in Section 8.3).
         This value applies to non-oxygenated gasolines.

    o    For calculating storage capacity for individual gasoline
         constituents, estimate solubility in water by using

              C.w=X.g  S.                                          (4.17)


    where     S.  = solubility of Constituent i in water (mg/L)

              X.  = mole Fraction of Constituent i in gasoline

              C.  = concentration of Constituent i in water (mg/L)
               iw


4.3.3  Guidance on Inputs for, and Calculations of Average Value

    Use average value of 20 mg for concentration of gasoline.  The source
of this value is discussed in detail in Section 8.4.

Comments

    o    It is assumed that concentration of contaminants in the aqueous
         phase is linearly related to concentration in soil.

    o    Effects of salinity, co-solutes, co-solvents, and non-carbon-based
         sorptions are ignored.

4.4  Example Calculation

    Problem:  Calculate mass of gasoline and benzene for release of non-
oxygenated gasoline at a site where the soil organic carbon content is  1
percent.

4.4.1  Storage Capacity Calculations

4.4.1.1  Maximum Value Calculations

    o    Storage capacity  for gasoline:


              S    =  200 mg/L = 0.002 molar  (assuming molecular weight of

                                               gasoline as 100)

              f    =  1/100    =   .01
               oc
                                    118

-------
    Use equation 4.11 to estimate K  ,
    log K    =  -0.602 x log(0.002) + 0.656
             =  2.28
        K    =  191
         oc
    Then  K  =  (191)(0.01)   -  1.91 L/kg
    and   C  =   1.91L/kg x 200 mg/L = 3.82 x 102 mg/kg
    o    Storage capacity for benzene:

    Estimated concentration of benzene in gasoline saturated water
                C   =  64.8 mg/L   (from Table 8-3)

                K   =  62
                 oc
    Therefore,  K   =  0.62

                    =  0.62

                    =  40.2 mg/kg
             w
            K   =  62          (from Table 4-3)
             oc                               '
and,        C   =  0.62 x 64.8 mg/kg
             o
4.4.1.2  Average Value Calculations

    For calculating average retention of gasoline, the concentration of
gasoline in water is assumed to be 20 mg/L.

              K    =  191 (from above)
               oc
    Then,     K    =  1.91 L/kg

    and       C    =1.91 L/kg x 20 mg/L  =  38 mg/kg
               o


    All other parameter values remain the same.


4.4.2  Transport Rate Calculations

    In this section, we evaluate the transport and fate of gasoline
constituents initially dissolved in water which has sorbed onto soil
surfaces.  Two questions are specifically addressed here:  1) What would be
the time scale for desorbing the contaminants and 2) what would be the
effect of sorption on bulk transport of gasoline dissolved in water.

-------
4.4.2.1  Time Required to Desorb 50, 90, and 95 percent of Contaminant
    The guiding equations are:

              D ff  =  	                                       (4.3)
                       (1-9) PgK
               eff
              lo.5  -

              '0.9  =

              lo.95 =

    Input values for benzene:

                       -5   2
              D   =10   cm /s(assumed)
              0   =  0.13 (from Uu and Gschwend, 1986)
              p   =  2.5 g/cm
               s
              K   =  KQc x fo{i -r  62 (from Table A-3) x 0.005  (assuming an
                                                               f    content
                                                               oc
                                                               of 0.5
                                                               percent)
                             -5         2
    Then      Dgff =       10   x (0.13) _
                        (1 - 0.13) x 2.5 x 0.31
                                 -7   2
                   f-     2.5 x 10   cm /sec
                                              _3
    Then, assuming a particle radius of 2 x  10   cm,  the  time  required  to
desorb 50 percent of the sorbed benzene,
                       0.031 x  (2 x  10"3)2
                                          sec
                             2.5  x  10
                    =  0.5  second
                                     120

-------
The time required to desorb 90 and 95 percent of the sorbed benzene are:


                        0.18 x (2 x 10~3)2
              t~ g  =   	          sec = 2.9 sec
    and
                             2.5 x 10~7
                        0.25 x (2 x 10~3)2

              '0.95 '   	=	S6C = 4 S6C
                            2.5 x 10
    For a compound such as dodecane, which has a very high K   value, the
time taken to desorb a certain fraction of the sorbed contaminate would be
much higher.   The time required to desorb 90 percent of dodecane is
calculated below:

    K   of dodecane  =  88,000
     oc


    Then,   K   =  88,000 x 0.005

                =  440


    Then,  D ff = 1.76 x 10~10 cm2/s
    and    tn 9 = ^091 sec = 1.14 hours


4.4.2.2  Retardation of Bulk Transport

    The governing equation for calculating the retardation of a contaminant
relative to the bulk movement of water is,


                          K  P
              R   =   1 +     b
                             e
    Input values for benzene:
              K,p, :    same as above
              6:      0.4 for silt (from Table 4-1)
                                    121

-------
                  =  1 + 0.31 x 2.5
                            0.4


                  =  2.9
    Due to the uncertainties associated with predicting K   from K   and
                                                         oc       ov
solubility, a retardation factor of 2 or below is not significant.  In the
saturated zone, where f   can be below 0.001, sorption coefficient will be
much lower than 0.31 ana consequently, R will be less than 2.  For such envi-
ronments, benzene should be considered as moving with the bulk flow of water.

    The retardation factor for dodecane in an environment similar to that
presented for benzene will be

    R        1  + 88,000 x 0.005 x 2.5
                         0.4


         =   2750


4.5  Summary of Relative Importance of Locus

4.5.1  Remediation

    Sorbed contaminants on soil surfaces in  the unsaturated and the
saturated zone can act as a source of contamination for extended period of
time.  Natural flushing would continue to "wash" contaminants away from the
soil surfaces and into the groundwater, acting as a ready source of
contamination.  Therefore, sorption onto soil surfaces is not a desirable
process for long term remediation.  Flushing of the system and removal
combined with treatment, surface capping and subsurface barriers are
probably the most effective remediation techniques.  Biodegradation,
although a slow process, can potentially be  an effective remedial
alternative.

4.5.2  Loci Interactions

    Table 4-5 lists all the loci that interact with hydrocarbons sorbed
onto soil surfaces from aqueous phase.  Bulk, transport is influenced by
sorption-desorption reactions which influences exchange of contaminants
(Loci nos. 3, 8, 9, 11, 12).  Diffusion in and out of the soil surfaces
depends on particle size distribution and geometry of soil particles (Loci
nos. 3, 8, 9, 11, 12).  Volatilization into  the soil gas from these loci is
probably not significant.
                                    122

-------
                                  TABLE 4-5

                 LOCI INTERACTIONS WITH HYDROCARBONS SORBED
                             ONTO SOIL SURFACES
                           Phases       Interacting       Relative
 Process	  in Contact	Loci	Importance
Mobility
Diffusion
Desorption
Wet soil 3, 8, 9,
11, 12
Wet soil 3, 8, 9,
11, 12
Modest
High
.Immobility

 Sorption                Rock            10,  13          Low



 4.5.3  Information Gaps

     In order to improve predictive capabilities for transport and fate of
 liquid hydrocarbons sorbed onto soil  surfaces, the following information is
 needed:

    1.     Sorption isotherm data on gasoline  constituents that include low
          carbon soils.

    2.     Influence of additives on sorption  of gasoline constituents.

    3.     Data showing errors associated with assuming linear isotherms.

    4.     Development of models relating soil characteristics to sorption in
          low carbon soil.

 4.6  Literature Cited

 Abdul, A.S.  and T.L. Gibson. 1986.   Equilibrium Batch Experiments with Six
     Polycyclic Aromatic Hydiocarbons  and Two Aquifer Materials.  Hazardous
     Waste &  Hazardous Materials,  3(2):125-137.

 Arthur D.  Little,  Inc.  1987.  Prediction of  Soil and Sediment Sorption for
     Organic  Compounds.   Office of Water Regulations and Standards, U.S.
     EPA,  Washington, D.C.

 Banerjee,  P.,  M.D.  Piwoni, and K. Ebeid.  1985a.   Sorption of Organic
     Contaminants to a Low  Carbon  Subsurface  Core.   Chemosphere,
     14(8):1057-1067.

-------
Banerjee, P., M.D. Piwoni, and K. Ebeid. 1985b.  Sorption of Organic
    Solvents to Subsurface Soil.  Presented at the annual SETAC Symposium.
    St. Louis, MO.

Chiou, C.T., P.E. Porter, and D.W. Schmedding. 1983.  Partition Equilibria
    of Nonionic Organic Compounds Between Soil Organic Matter and Water.
    Environ. Sci. Technol., 17:227-231.

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

Huyakorn, P.S. and G.F. Pinder. 1983.  Computation Methods in Subsurface
    Flow, Academic Press.

Jaffe, P.R. and R.A. Ferrara. 1983.  Desorption Kinetics in Modeling of
    Toxic Chemicals.  Journal of Environmental Engineering,
    109(40):859-867.

Karickhoff, S.W., D.S. Brown, and T.A. Scott. 1979.  Sorption of
    Hydrophobic Pollutants on Natural Sediments.  Water Research,
    13:241-248.

Karickhoff, S.W. 1980.  Sorption Kinetics of Hydrophobic Pollutants in
    Natural Sediments.  In:  Contaminants and Sediments, Volume 2 (R.A.
    Baker, ed.).  Ann Arbor Science Publishers, Inc., Ann Arbor, MI.

Karickhoff, S.W. 1984.  Organic Pollutant Sorption  in Aquatic Systems.
    Journal of Hydraulic Engineering, 110:707-735.

Leo, A.L. 1983.  Log P and Parameter Database.  Obtained from Pomona
    College, Claremont, CA.

McCarty, P.L., M. Reinhard, and B.E. Rittman 1981.  Trace Organics in
    Groundwater.  Environ. Sci Technol., 15(1):40-51.

MacKay, D. and W. Shiu. 1981.  A Critical Review of Henry's Law Constants
    for Chemicals of Environmental Interests.  J. Phys. Chem. Ref. Data,
    10:1175-1199.

Newsom, J. 1985.  Transport of Organic Compounds Dissolved in Groundwater.
    Groundwater Monitoring Review, Spring.

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., 19(10):975-979.

Rao,  P.S.C.. and J.M.  Davidson.  1980.   Estimation of Pesticide Retention
    and Transformation Parameters Required  in Nonpoint  Source Pollution
    Models.  In:  Environmental  Impact  of Nonpoint  Source Pollution (M.R.
    Overcash and J.M.  Davidson,  eds.).   Ann Arbor Science Publishers, Inc.,
    Ann Arbor, MI.
                                     124

-------
Schwarzenbach, R.P. and J. Westall. 1981.  Transport of Nonpolar Organic
    Compounds from Surface Water to Groundwater:  Laboratory Sorption
    Studies.  Environ. Sci Technol., 15:1360-1367.

U.S. Department of Agriculture (USDA). 1960.  Soil Survey:  Erie County,
    PA.

U.S. 1986.  EPA Superfund Public Health Evaluation Manual, Office of
    Emergency and Remedial Response, Washington, D.C.

Wauchope, R.D., K.E. Savage, and W.C. Koskinen. 1983.  Adsorption-
    Desorption Equilibriums of Herbicides in Soil:  Naphthalene as a Model
    Compound for Entropy-Enthalpy Effects.  Weed Science, 31:744-751.

Uu, S. and P.M. Gschwend. 1986.  Sorption Kinetics of Hydrophobic Organic
    Compounds to Natural Sediments and Soils.  Environ. Sci. Technol.,
    20(7):717-725.
                                    125

-------
                          SECTION 5.  LOCUS NO. 5

                                 CONTENTS
List of Tables	
List of Figures	
5.1  Locus Description.
     5.1.1  Short Definition
     5.1.2  Expanded Definit

5.2  Evaluation of Criteria
     5.2.1  Introduction...
     5.2.2  Mobilization/Rem
            5.2.2.1  Partiti
            5.2.2.2  Transpo

                     Physica
                     Hydraul
                     Mobiliz
     5.2.3  Fixation.
            5.2.3.1  Partiti
                      (Stati
            5.2.3.2  Other F
     5.2.4  Transformation.
            5.2.4.1  Biodegr
            5.2.4.2  Chemica

5.3  Storage Capacity in Loc

     5.3.1  Introduction and
     5.3.2  Guidance on Inpu
              Maximum Value.
     5.3.3  Guidance on Inpu
              Average Values

            Comments	
on and Comments.

or Remediation..
bilization.
ning into Mobile Phase.
t with Mobile Phase....
 Dislodgement/Displacement,
c Removal	
tion Schemes	
ning onto Immobile
nary) Phase	
xation Approaches..
dation	
 Oxidation.
Basic Equations	
s for, and Calculation of,
5 for, and Calculation of,
                                    126
Page No.

  127
  128
  129

  129
  129

  129

  129
  134

  134
  134

  135
  136
  139

  140
  140
  140

  140

  140
  140

  140

  140

  141

  141

  141

-------
                                (Continued)



                                                                   Page No.

5.4  Example Calculations	     142

     5.4.1  Storage Capacity Calculations	     142

            5.4.1.1  Maximum Storage Capacity	     142
            5.4.1.2  Average Storage Capacity	     142
            5.4.1.3  Maximum Mass	     143
            5.4.1.4  Average Mass	     143

     5.4.2  Transport Rate Calculations	     143

            5.4.2.1  Dissolution Rates	     143
            5.4.2.2  Transport Rate of Residual Liquid
                       Contaminant	     143

5.5  Summary of Relative Importance of Locus	     144

     5.5.1  Remediation	     144
     5.5.2  Loci Interaction	     144
     5.5.3  Information Gaps	     145

5.6  Literature Cited	     146
                                  TABLES


5-1  Typicical Fluid Properties of Hydrocarbons                      138

5-2  Water-Hydrocarbon Experimental Results                          141

5-3  Loci Interactions with Residual Liquid Contaminant in
     Groundwater                                                     145
                                    127

-------
5-1  Schematic Cross-Sectii
     Contaminants in the Pi
     or Rock Surfaces in tl

5-2  Schematic Representat:
     Transport Processes A

5-3  Sketches of Low Capil
     Using the Pore Double
     passing Below

5-4  Effect of Aspect Rati
     Non-uniform Table

5-5  Residual Hydrocarbon
     Residual Saturation (
     (S*or) as a Function

5-6  Hydraulic Gradient Ne
     (at N* ) in Soils of
     of Various Interfacia

5-7  Hydraulic Gradient Ne
     for Various Hydrocarb
     Recovery of Residual
     = 10 dyne/cm

5-8  Percentage of Total V
     Residual Hydrocarbon
     the Curves
lal Diagram of Locus No. 5 - Liquid
re Spaces Between Soil Particles
e Saturated Zone

 n of Important Transformation and
 lecting other Loci

 ry Number Trapping Mechanisms,
 Model:  Snap-off at Top, By-
 on Hydrocarbon Trapping in a
                                  FIGURES
                                                                   Page No.
130
131
                                          132
aturation Ratio, Relating Final
  ) to Initial Residual Saturation
? Capillary Number (NC>

essary to Initiate Blob Mobilization
arious Permeabilities for Hydrocarbons
 Tension

essary to (a) Initiate Blob Mobilization
ns and (b) Obtain Reported Percentage
ydrocarbon for Interfacial Tension
lume of a Heterogeneous Sand in which
as been Reduced by the Amount Shown on
                                          132
136
137
                                          137
                                          139
                                    128

-------
                          SECTION 5 - LOCUS NO. 5
5.1  Locus Description

5.1.2  Short Definition

    Liquid contaminants in the pore spaces between soil particles in the
saturated zone.

5.1.2  Expanded Definition and Comments

    The liquid contaminant in this locus may be in either the "water-wet"
configuration (i.e. particle surfaces wet by water) or the "oil-wet"
configuration (i.e. particle surfaces wet by oil); no air is present.  The
liquid contaminants are likely to be present as a discontinuous phase,
derived from fluctuating water table level.  A continuous phase is more
likely for liquid contaminant such as tetrachloroethylene or PCBs which are
denser than water.  The liquid contaminants in locus no.  5 are subject to
relatively weak buoyancy forces due to density differences and entrainment
forces due to moving groundwater.  Thus the liquid material is considered
relatively immobile and dissolution is the principal loss mechanism.

    This locus differs from locus no. 13 which considers liquid
contaminants in fractured bedrock, not in porous media.  Figures 5-1 and
5-2 present a schematic cross-sectional diagram of locus no. 5 and a
schematic representation of the transformation and transport processes
affecting other loci, respectively.

5.2  Evaluation of Criteria for Remediation

5.2.1  Introduction

    Liquid contaminant in pore spaces between particles in the saturated
zone is most likely to be present as a discontinuous phase in pore volume
otherwise filled with water, and is derived from displacement of the main
body of liquid contaminant by a fluctuating groundwater table (described in
locus no. 7).  Once the main liquid contaminant body is displaced, a
portion is retained in the porous media because of capillary forces,
becoming immobile at lower residual saturations.  Residual saturation is
the volume of discontinuous immobile liquid contaminant per unit void
volume.

    Residual liquid contaminant is retained in pore spaces in the saturated
zone by two mechanisms:  by-passing and snap-off (Mohanty, et al., 1980;
Chatzis, et al., 1983) illustrated in Figure 5-3.  The doublet model of
pore structure (Chatzis, et al., 1983) consists of a tube split into two
pores, one generally narrower than the other, which rejoin.  Liquid
contaminant is trapped in the upper tube by snap-off, which strongly
depends on pore shape and wettability (Figure 5-3).  For pores with a high
aspect ratio (i.e. where the pore throats are much smaller than the pore
bodies), snap-off is common.  For pores with a low aspect ratio, liquid
contaminant is easily displaced (Figure 5-4a).  The remaining liquid
contaminant in the upper pore is displaced downstream.  If the downstream
                                    129

-------
    NOTE: NOT ALL PHASE
         AIR IS NOT PRESENT.
                                    LEGEND


                               WATER FILM

                               WATER IN SATURATED ZONE

                               BIOTA (MICROBIOTA INCLUDED)

                               SOIL PARTICLE

                          |    COLLOIDAL PARTICLE

                                               PHASE
Figure 5-1. Schematic Cross-Sectional
           Liquid Contaminants in
           Particles or Rock Surfaces
    Diagram of Locus No. 5 -
the Pore Spaces Between Soil
   in the Saturated Zone.
                                  130

-------
                                            (WATER)
                                           LOCUS NO.
                                      8 - DISSOLVED IN GROUND-
                                        WATER   	
                                  DISSOLUTION
           PHASE SEPARATION
(LIQUID CONTAMINANTS)
     LOCUS NO
   7-FLOATING ON
     WATER TABLE
  13- IN ROCK
     FRACTURES
  LOCUS NO. 5
(LIQUID CONTAMINANTS
  IN PORE SPACES IN
  SATURATED ZONE)
DIFFUSION
    (ROCK)
   LOCUS NO,
10-DIFFUSED INTO
   MINERAL GRAINS
   OR ROCKS
                                   DEPURATION
                                             (BIOTA)
                                           LOCUS NO.
                                       11-SORBEDTOBIOTA
              Figure 5-2. Schematic Representation of Important Transformation
                          and Transport Processes Affecting Other Loci.
                                                 131

-------
       STAGE 1


Source: Chatzis et. al., I983.  (Copyright Society of Pe

     Figure 5-3   Sketches Of Low
                  The Pore Doublet
                              (a) LO
                                                         SNAP-OFF
                    BY-PASSING


     STAGE 2


oleum Engineers. 1983. Reprinted with Permission.)
STAGE 3
apillary Number Trapping Mechanisms, Using
i/lodel: Snap Off at Top, By-passing Below
 ASPECT RATIO
                                                                                         OIL
       OIL TRAPPED BY SNAP-OFF
                                                         COLLAR OF WATER
                              (b) HIGH ASPECT RATIO


  Source: Chatzisetal. 1983. (Copyright Society of Petroleum Engineers. 1983. Reprinted with permission.)
             Figure 5-4    Effect of Aspect Ratio on Hydrocarbon Trapping in a
                          Non-Uniform Tube
                                              132

-------
junction is unable to form a stable meniscus (boundary) in the downstream
tube, two menisci between water and liquid contaminant develop:  one that
is displaced downstream from the upper pore and another in the lower pore.
Liquid contaminant in the lower pore becomes trapped by by-passing (Figure
5-3).  As the aspect ratio increases, the proportion of liquid contaminant
trapped by snap-off increases.  Heterogeneity in soil or rock promotes
trapping through by-passing.

    Naturally occurring hydraulic gradients in groundwater are too small  to
displace the highly-curved menisci between liquid contaminant and water in
the pore space, and cannot "squeeze" the liquid contaminant through the
pore throats.  Thus, the menisci block pores and interfere with water flow.
For relatively low liquid contaminant saturations, these capillary forces
cause disconnections of the liquid contaminant phase, resulting in oil
blobs.  This residual liquid contaminant in the saturated zone can occupy
from 15 to 40 percent or more of the pore volume (Felsenthal, 1979; Morrow
and Chatzis, 1982; Chatzis, et al., 1984).  Porous media structure and
particle size have much less influence on residual saturation in the
saturated zone than in the unsaturated zone.

    When considering the problem of remediation of residual liquid
contaminant in the saturated zone, some fundamental questions are:

    (1)  What is the amount of residual liquid contaminant in the porous
         media interstices?

    (2)  Which constituents can be remobilized from, or immobilized in, the
         interstices?

    The influence of capillary forces in this locus is an important factor
for remediation.  In order for corrective action mechanisms to mitigate
residual liquid contaminant, these capillary forces must be either
surmounted (or enhanced) for mobilization (or immobilization) to occur.

    Mobilization of residual liquid contaminant can be accomplished either:
(1) by dissolving the liquid contaminant into groundwater; or (2) by
physically dislodging the droplets or blobs from soil pore volume,
entraining the droplets in the groundwater flow.  Dissolution of residual
hydrocarbon is not as effective overall as physical removal, and acts as a
continual source of contamination.  Physical removal of residual liquid
contaminant can be relatively effective and is accomplished either by
hydraulic sweeping (i.e. increasing* hydraulic gradient) and/or by lowering
of interfacial tensions between the residual liquid and water with
surfactants.

    Immobilization of residual liquid contaminant can be accomplished by
enhancing capillary trapping via lowering the hydraulic gradient.

    Transformation of residual liquid contaminant by biodegradation (locus
no. 11) is seldom attempted as the only corrective action for removal of
residual liquid contaminant; the quantity of liquid contaminant at many of
these sites is simply too large.  Biodegradation, in conjunction with
mobilization mechanisms, can result in effective removal of residual liquid
contaminant, leading to ultimate restoration of the aquifer.

-------
    In locus no. 5 two fluid
contaminant.  In such a two-fluid
usually considered to be the n|on
wets the particle surfaces
where the liquid contaminant i
in loci nos. 2 and 7.  Only th
considered in this discussion.
5.2.2  Mobilization/Remobilization
5.2.2.1  Partitioning into Mobile Phase
                    Biases are present:   water and liquid
                         phase system,  the liquid contaminant is
                       -wetting fluid (i.e.  water preferentially
                         than the liquid contaminant).   The case
                       considered the wetting fluid is  discussed
                       "water-wet" configuration will be
rather
    Partitioning of residual
water depends on:  (1) the
contact with groundwater (locus
constituents of a liquid contajminant
that blend (see locus no. 7
Equilibrium dissolution is approximated
                    l|iquid contaminant (e.g., gasoline) in pore
                          area of the liquid contaminant in
                       no. 8); and,  (2) the solubilities of the
                            in water and their concentrations in
                       a more detailed discussion of dissolution).
                               by:
surface
where    C. = concentration of

         X. = mole fraction of
Calculations are more difficul
additives are present.  The
for these additives (e.g., amolunt
properties, etc.).  The gasoline
are the light aromatics:  ben
Many gasoline additives, such
higher solubilities.  Solubili
gasolines are presented in Tat:
5.2.2.2  Transport with Mobile
    Once residual liquid conta
mobilized either by dissolution
groundwater,  transport occurs
constituents,  the advective-
transport and  is presented in
become dislodged and entrained
artificially  increased hydraulic
contaminant exists as a discoqt
not describe  adequately the
                                 C. = X.S.
                                i = 1, 2, 3... n constituents
                                                             (5.1)
                      constituent- i in the water phase (mg/1)

                      constituent i in liquid contaminant

S.  = solubility of constituent i in pure water (mg/1)
                     t if significant amounts of oxygenated
                   difficulty is compounded by the lack of data
                         in gasoline blend, solubility, fluid
                        constituents with the highest solubilities
                     ene, toluene, xylenes, and ethylbenzene.
                     as MTBE, ethanol, and methanol, have much
                     ties of chemicals found typically in
                     les 4-3 and 7-1.
                      Phase
                     minant in the saturated phase has become
                       or dislodgment and entrainment in the
                     with groundwater flow.  For dissolved
                   dispersive equation (8.1) describes bulk
                     locus no. 8.  Residual liquid contaminant can
                      in the groundwater as a result of an
                        gradient.  Because residual liquid
                      inuous phase, continuum flow equations may
                   transport of entrained liquid contaminant.

-------
Physical Dislodgement/Displacement

    The capillary forces which resist mobilization can be overcome by
viscous forces associated with hydraulic gradient or by gravity buoyancy
forces.  In this locus, residual liquid contaminant (e.g., gasoline) is
subjected to relatively weak buoyancy forces due to density differences.
Capillary and viscous forces are more significant than buoyancy forces with
regard to mobilization of residual liquid contaminant in pore spaces in the
saturated zone.

    The dimensionless ratio of capillary forces to viscous forces is known
as the capillary number (N ):

                           Nc=k.pv.g.J                               5>2)
                                                         2
where    k  = intrinsic permeability of porous medium (cm )
                                        3
         p  = the density of water (g/cm )
          w
         3  = interfacial tension (dyne/cm)
         g  = gravitational acceleration (cr

         J  = hydraulic gradient (cm/cm)
    Experimental studies by Wilson and Conrad (1984) established a strong
correlation between displacement of residual liquid contaminant and
capillary number, when the hydraulic gradient was greater than the critical
value needed to initiate motion of some of the residual liquid contaminant
blobs or droplets.  This critical value of the capillary number depends on
the residual saturation.

    Liquid contaminant saturation is the volume of liquid contaminant per
unit void volume:

                             SQ = Vo/Vv                               (5.3)

                                                          3   3
where    S  = organic or liquid contaminant saturation (cm /cm )
         V  = volume of hydrocarbons (cm )
         V  = volume of voids (or pore space)(cm )


    As the liquid contaminant is displaced, its saturation is reduced to a
residual:

                             S   = V . /V                              (5.4)
                              or    do  v                                 '

                                                         3   3
where    S   = residual liquid contaminant saturation (cm /cm )

         V,  = volume of discontinuous liquid contaminant (cm )

          V  = volume of voids (cm )


                                    13'j

-------
     The relationship betweer
 is  shown in Figure 5-5;  resi
 valye (i-e^ ratio of final
 S    ).   N   reo.r.esents  the  c
 initiated.   N    denotes  the
 the.residual hydrocarbon.   lor
 10   .   Experiments on carborate
 the sandstone curve in  Figur
 the capillary  number and residual  saturation
dual saturation is  normalized by  its  initial
 sidual saturation,  S   to initial  saturation,
 itical capillary number at which motion  is
 capillary number, necessary to dispj^ce all of
 •" sandstone, N  = 2 x 10   and N     = 2 x
    rocks produce curves lying to the left of
  5-5 (Morrow,  1984).
                            • w
                      NC.CAP
Source: Wilson and Conrad, 1984. (Copyright
Figure 5-5.  Residual Hydrocarbon Sat
          Initial Residual Saturation
Hydraulic Removal

     Some  of the liquid conta
it will be removed when N,
hydraulic gradient of thecwa
porous medium (k)^ using  the
critical  value N   and for
tensions, Table §-1.  These
initiate  blob mobilization.

     For example, n-octane wi
will tend not to dissolve in
of«50 dyne/cm.  As evident
cm  ,  a gradient of J = 0.1
a steep gradient, but not un
5-6  represents the gradient
hydrocarbon for 9 = 10 dyne/
                                                           Nc represents initiation of
                                                           blob motion
                                                           NC represents complete
                                                           removal of blobs
          10-'
LLARY NUMBER
                                                    10-2
Water Well Journal Publishing Co., 1984. Reprinted with permission.)
 ration Ratio, Relating Final Residual Saturation (Sor) to
S*or) as a Function of Capillary Number (Nc).
ningnt wil] be  removed when N  > N
 N   .  Figures  5-6  and 5-7a are p5<
                                      and all  of
        Figures 5-6 and  5-7a are pfots of
 er (J) versus intrinsic permeability of the
 capillary number of  equation 5.2, for the
 rious liquid contaminant water interfacial
jlots describe hydraulic gradients necessary  to
 h a low solubility  (S = 0.66 mg/L) at  20°C
 the water phase;  it has an interfacial tension
 om Figure 5-6,  in a coarse sand, with  k =  10
  necessary  to  initiate blob mobility.   This is
 easonable.  The upper right curve of Figure
lecessary for complete removal of all
•.m.

-------
                           0.00 1
                                                                '   io"
                                - CLEAN SAND— H
                                S«.TYSANO-H
                                              R (en.2)

Source: Wilson and Conrad, 1984. (Copyright Water Well Journal Publishing Co. , 1984. Reprinted with permission.)


Figure 5-6.  Hydraulic Gradient Necessary to Initiate Blob Mobilization (at N^ ) In Soils of Various
            Permeabilities for Hydrocarbons of Various Interfaclal Tensions.
  (a)
         o.o
(b)
                      k«-»iX10*
                                10  12  T4
 Source: Wilson and Conrad, 1984. (Copyright Water Well Journal Publishing Co., 1984. Reprinted with permission.)

Figure 5-7.  Hydraulic Gradient Necessary to (a) Initiate Blob Mobilization for Various
            Hydrocarbons and (b) Obtain Reported Percentage Recovery of Residual Hydrocarbon
            for Interfaclal Tension = 10 dyne/cm.
                                              137

-------
                 TYPICAL FL1
                        TABLE 5-1

                   ID PROPERTIES OF HYDROCARBONS
  Fluid
     Interfacial
       Tension

      (dyne/cm)
   Density
   (g/cm3)
   Viscosity
    _2
(xlO   g/cm.sec)
                                                                Temperature
Water
n-Butanol
n-Pentanol
n-Hexanol
Benzene
n-Heptane
n-Octane
n-Hexane
          0.0
          1.6
          4.4
          6.8
         35.0
         50.2
         50.8
         51.0
   0.9982
   0.8098
   0.81443
   0.8136a
   0.87865
   0.68376
   0.7025
   0.6603
     1.002
     2.948
     0.652
     0.409
     0.542
     0.326
20
20
25
25
20
20
20
20
a. 20°C
Source:
Wilson and Conrad,
Publishing Co., 19
    Figure 5-7b is a plot o
recovered as a function of
interfacial tension, a = 10
5-6 is the 100 percent reco
other liquid contaminants c
a/10.  For the coarse sand
mobilize n-octane blobs, a
a liquid contaminant with a
gradient is (50/10)
large gradient.
             5 tim
    The residual saturation
for homogeneous media.  Het
when considering the  total
spatially variable medium-t
Smith  (1981), and used by W
in Figure 5-8.  The curves
swept  from each portion of
c - 10 dyne/cm, only  7 perc
percent or more of the resi
pore space has been swept o
because it neglects spatial
heterogeneity significantly
liquid contaminant.
1984 (Copyright Water Well Journal
4.  Reprinted with permission.)
 the percentage of residual hydrocarbon
radient and soil permeability, for^a. specific
dyne/cm.  The top curve labeled N    in Figure
ery curve for the same hydrocarbon.  Removal of
n be examined by multiplying these results by
xample, which required a gradient of 0.1 to
radient of 1.0 is required to completely remove
  10 dyne/cm.  For n-octane, the necessary
s as large, of J = 5; this is an impossibly
                   curves in Figures 5-5 through 5-7 were derived
                   rogeneity, however, plays a significant role
                   olume of residual hydrocarbon recovered.  A
                   -coarse sand was statistically evaluated by
                   Ison and Conrad (1984).  The results are shown
                   epict the percentage of residual hydrocarbon
                   he soil volume.  Thus, for a gradient of 10 and
                   nt of the pore space has been swept of 75
                   ual hydrocarbon, while over 80 percent of the
                    50 percent or more.  This plot is crude
                   correlation, but it demonstrates that
                   affects the removal/mobilization of residual
                                    138

-------
                iu
                   100
                                                 O=10
                UJ
                O
                u
                ff
                UJ
                Q.
                                                          100
.00002
                   E(logk)--6.4;ff|ogk -4.17; and r^.*  -0

 Source: Wilson and Conrad, 1984. (Copyright Water Well Journal Publishing Co. , 1984. Reprinted with permission.)
 Figure 5-8. Percentage of Total Volume of a Heterogeneous Sand In which Residual
           Hydrocarbon has been Reduced by the Amount Shown on the Curves.
    In summary, residual liquid contaminant  is  easier to mobilize for
contaminants with lower interfacial  tensions with water, for more permeable
porous media, and in the laboratory  where  extremely high hydraulic
gradients can be produced.  Liquid contaminant  is more difficult to
mobilize in the field where there are  severe practical constraints on the
hydraulic gradient; it may be possible  to  mobilize some of the residual
liquid contaminants, but it is difficult to  mobilize it all.

Mobilization Schemes

    Mobilization of residual liquid  contaminants  is possible by a number of
different schemes involving an increased hydraulic gradient or lowering of
interfacial tensions.  These schemes are in  different stages of development
and understanding, and usually do not  result in mobilization of all the
residual liquid contaminants.  Hydraulic gradient can be maximized in the
field by employing closely-spaced vertical sweeps and/or hydraulic
fracturing of aquifer (Wilson and Conrad,  1984).   A scheme for mobilization
which can be used in conjunction with  increased gradients, is the reduction
of interfacial tensions (Melrose and Brander, 1974;  Taber, 1981) by
surfactants and alkaline flooding.   There  does  not appear, however, to be a
safe means of implementation without further contamination of the
groundwater.  Thermal or steam floods may  be effective in lowering
interfacial tensions, but also increases the dissolution of certain
constituents.   These schemes have been  utilized to some degree by the oil
industry for enhanced oil recovery (EOR).  EOR  methods have not been
applied to groundwater pollution problems, but  may prove to be feasible.
                                           139

-------
5.2.3  Fixation
5.2.3.1  Partitioning onto
lobile  (Stationary)  Phase
    The residual liquid contaninant liquid phase is already considered to
be relatively immobile.  In the "water-wet" configuration, liquid
contaminant constituents could partition onto the porous medium particles
after migrating through a water film surrounding a water-wetted porous
medium particle or rock surface (see locus no. 4 for discussion).  Residual
liquid contaminant could be made more immobile if the wetting conditions
changed to an "oil-wet" configuration, perhaps brought about by lowering
the water table (see locus no. 7 for discussion).

5.2.3.2  Other Fixation Approaches
    Residual liquid contaminant
capillary trapping and remain
enhanced by decreasing the hydraulic
and the capillary number (see
flow, however, dissolution oc
isolate liquid contaminant in
porous medium in the saturate
5.2.4  Transformation
   in  pore  spaces  has  already  undergone
5  relatively immobile.   Immobilization can be
        gradient,  thus lowering dissolution
Figure 5-5).   Even with static groundwater
:urs;  it is not possible to completely fix or
pore  spaces between the particles of the
I  zone.
5.2.4.1  Biodegradation (Deta
rate of aerobic degradation.
increase the dissolved oxygen
 Is in Section 1)
    Below the water table, the supply of oxygen is limited, reducing  the
 The rise and fall of the water table might
 level if the layer of liquid contaminant upon
                                             t"
                               is very thin or absent.


                               tails in Section 3.2.4.2)

                               ered in the saturated zone can increase  the
likelihood of contaminant cat*
transformations.  Polymerizat:
conditions.

5.3  Storage Capacity in Locu;

5.3.1  Introduction and Basic
in the pore space between  the
saturated zone is:
 lyst interaction and thereby lead to abiotic
 on reactions can also proceed under anaerobic
 Equations
    The maximum amount of residual liquid contaminants  that can  be  stored
 particles of the porous medium in the
                        R = S
                             cr
where    R =  retention  (liters
                         3
              of soil, L/m  )
     0 .  10-
(5.5)
  of residual liquid contaminant/cubic meters
                                     140

-------
       S   = residual saturation (0 < S   < 1)
        or                             oir ~~
         9 = soil porosity (0 < 0 < 1)
    o  R and S   values have largely been determined experimentally on
       natural sandstone cores or on glass beads (see Table 5-2).

    o  The mass of residual liquid contaminant can be calculated by
       multiplying (R)(p ), where p  is the density of the liquid
       contaminant.

                                 TABLE 5-2

                  WATER-HYDROCARBON EXPERIMENTAL RESULTS
Media
Glass Beads
Sandstones
Sor <%>
14
27-40
R (L/m3)
57
54-80
0
0.38
0.20
         Source:  Wilson and Conrad (1984)


5.3.2  Guidance on Inputs for, and Calculations of, Maximum Value

    Residual hydrocarbon saturation (S  ) as determined by experiments run
at low capillary number ranged between 27 percent and 43 percent (Chatzis
and Morrow, 1981).

    o  The upper value of 43 percent is recommended for calculating the
       maximum value.

    o  For appropriate porosity values, see Table 12-2 or Figure 12-4.

5.3.3  Guidance on Inputs for, and Calculations of, Average Value

    o  Experimentally determined residual hydrocarbon saturation (S  ) run
       at various capillary numbers were found to have a mean of 28 percent
       (Felsenthal, 1979).  S   =28 percent is recommended for calculating
       the average value.

Comments

    o  Retention values neglect subtraction or addition of liquid
       contaminant by partitioning processes.
                                    141

-------
    o  The retention values vill likely decrease with time in response to
       loss of constituents to biodegradation and dissolution into
       groundwater.

    o  Retention is not signi
       temperature.
ficantly affected by modest changes in
    o  Soil or rock heterogereity and porosity are the most significant
       factors affecting retention.

    o  Retention in the saturated zone is significantly larger than in the
       unsaturated zone and is essentially independent of particle size.
    o  Most hydrocarbon reten
tion values reported in the literature refer
       to vadose zone experiments.

5.4  Example Calculations

5.4.1  Storage Capacity Calculation

5.4.1.1  Maximum Storage Capacity

  „ If the total volume of voids for a sandstone is 4 cubic meters (V,
4m ). and the volume of residual hydrocarbon is 1.7 cubic meters (V
1.7m ), then the residual sat
              S   = V  /V
               or    or  v
              S   = 43%
               or
uration can be determined using:
                                                                   or
    Using the maximum residual saturation value (S   = 43%), and a porosity
value for a very fine-grained sandstone (0 = 0.19)°rthe maximum amount of
retention (R) from equation 5

    R = (0.43.)(0.19.)103

    R = 81.7 liters of residual hydrocarbon/nr' of sandstone

5.4.1.2  Average Storage Capacity
                3
    Using an average residual
porosity (0 = 0.19), the aver

    R = (0.28.)(0.19.)103
    R = 53.2 liters of residue
    The mass of residual gaso
determined if the density of
 saturation value (S = 28%), for the same
 ge amount of retention is:
 1 hydrocarbon/m  of sandstone

 ine per unit volume of porous medium can be
 he liquid contaminant (p ) is known:
                                                      g
                                = p  .(R/1(T)
                                   6
                                    I/.?

-------
    Where:    m  = residual mas.s of gasoline/unit volume of porous medium
               r   (kg/mj)

              p  = 710 kg/m3 (from Table 7-1)
               O
5.4.1.3  Maximum Mass

    mr = (710) (82/103)

       =58.2 kg/m3

5.4.1.4  Average Mass

    mr = (710) (53/103)

       =37.6 kg/m3

5.4.2  Transport Rate Calculations

5.4.2.1  Dissolution Rates (Details in Section 7.4.2)

    The rate of dissolution from residual liquid contaminant in the pore
space in the saturated zone depends strongly on the partition coefficient and
mixing within each phase (e.g., greater mixing with an increased groundwater
flow).  Dissolution rates are discussed in detail in locus no. 7.

5.4.2.2  Transport Rate of Residual Liquid Contaminant

    Since residual liquid contaminant in porous medium interstices in the
saturated zone exists as a discontinuous phase, continuum flow models do
not describe adequately the transport of entrained liquid contaminant.  At
present, there is not adequate understanding of how discontinuous blobs or
droplets are transported in groundwater.  These droplets may move as single
blobs and may or may not coalesce.  Bigger blobs may breakup into single
blobs which have a greater propensity for retrapping.  If surfactants are
present, the likelihood of coalescence is diminished.  Overall, the
heterogeneity of the porous media will undoubtedly have a significant
influence on blob distribution and transport.

    A useful calculation might be to determine the hydraulic gradient
necessary to initiate blob mobilization.  For example, in a typical
sandstone the critical capillary number required to initiate blob
mobilization, N * = 2 x 10"  (Chatzis and Morrow, 1981), can be used to
determine the hydraulic gradient.

    Using equation 5.2:

              N  = k.p.g.J
               c     a

    and rearranging to solve for J:
                   *
                  N    'd
              J = __L_
                  k.p.g

                                    HI

-------
    N
      8

      k

      g
          2  x 10   (given in example)
          50 dyne/cm (from Table 5-1)
        = 10"
               cm
(for a coar
          980.7  cm/sec'
                   3
se sandstone)
     p  = 1.0 gm/cm"
      w
      J =
            (2x 10 5)(50)	

         (10~5) (1.0) (980.7)
      J = 0.1

    For a coarse sandstone, n-
to initiate blob motion.  A hydraulic
completely remove all residual
5.5  Summary of Relative Impor

5.5.1  Remediation

    The largest amount of liqu
                              )ctane would require a steep gradient of 0.1
                                      gradient of 10 would be required to
                               n-octane from the coarse sandstone.
residual liquid contaminant cai
of magnitude greater than that
                               ance of Locus
                               d contaminant stored in the saturated zone
is most likely the liquid contaminant in porous media interstices.  This
                              i ho present in amounts of one or two orders
                               dissolved in the groundwater or sorbed onto
particulate matter of the porouis media.  Residual liquid contaminant in
pore space in the saturated zome provides a continual source of groundwater
contamination by dissolution or the more soluble constituents.  Because of
dissolution, it is not particularly feasible to promote retention of this
residual liquid contaminant.
pragmatic and could possibly result in more effective aquifer restoration.
This removal could be accompli;
transformation mechanisms.  Di<
contaminant by increased hydrai
tensions could reduce the quan
appropriate for biodegradation.
                               emoval of the residual appears to be more
                               her) by a number of mobilization and
                               lodgment of some of the residual liquid
                               lie gradient and/or decreased interfacial
                               ity of liquid contaminant to a level
                                 The efficacy of these corrective action
mechanisms (or combination of mechanisms) is not currently known.  Specific
studies of each mechanism howe\ei,  indicates a substantial potential for
the remediation of residual licuid contaminant in groundwater.

5.5.2  Loci Interactions

    The loci interacting with residual liquid contaminant in groundwater
during partitioning and mass tiansport process are summarized in Figure 5-2
and Table 5-3.  For residual liquid contaminant, the principal partitioning
process is dissolution  into  the
advection and diffusion.  Bulk
7).  Bulk  transport out of  the
                                groundwater (locus no. 8) by mixing due to
                               transport into the water-filled pore space
in the saturated zone  is  the result of a fluctuating water  table  (locus no.
                               locus is influenced by the hydraulic
gradient of the groundwater  (Iccus no. 8) and  the permeability of  the  media
                                     144

-------
(loci nos. 5, 10, and 13).  The degree to which the residual liquid
contaminant is retained is related to the sorptive properties of the porous
media (loci nos 4, 10, and 13) and wetting conditions (loci 7, 8).
Residual liquid contaminant is considered to be relatively immobile and
dissolution is the chief loss mechanism.

                                 TABLE 5-3

                  LOCI INTERACTIONS WITH RESIDUAL LIQUID
                        CONTAMINANT IN GROUNDWATER
    Process
              Phases in
            Direct Contact
                                   a
Interacting
    Loci
 Relative
Importance
Mobility

Dissolution
 (Phase separation)

Bulk Transport
 (Displacement,
  Entrainment)
Immobili ty

Sorption


Wetting Conditions
                water
                water
              wet soil
                rock
                rock
            liquid hydrocarbon
              wet soil
                rock

                water
     8
     4
    10
    13
     7
     4
    13

  7, 8
                  high
  high
moderate
very low
  high
  high
moderate
  low

moderate-high
a. Biota (locus no. 11) are potentially in direct contact with all phases.
5.5.3  Information Gaps

    For a better understanding and definition of the behavior of residual
liquid contaminant in pore spaces in the saturated zone, and the influence
and utility of partitioning process on this residual liquid contaminant,
the following information is needed:
   1.
   2.
   3.
fundamental data on solubility and dissolution rates as well as
the composition and physicochemical properties of liquid
contaminant;

specific concentration of additives in different liquid
contaminants and their behavior in an aquifer;

sorption data (k,,k  ) that represent natural soil systems of
interest as well as nonequilibrium effects (DOC, OC, time,
temperature, etc.);

-------
   4.     data on the influeni
         liquid contaminant;

   5.     fundamental underst;
         fluid-phase in a po
   6.     further understand!;
         heterogeneity; and,

   7.     aerobic and anaerob
         date have focused 01

5.6  Literature Cited

Chatzis, I. and N. Morrow.
    Relationships for Sandst
    Exhibit, San Antonio, TX

Chatzis, I., N.R. Morrow, an
    Structure of Residual Oi
    Journal, Vol. 23, no. 2.

Chatzis, I., M.S. Kuntamukkl
    13213.  1984 S.P.E. Annu
    TX.

Felsenthal, M. 1979.  A Stat
    Journal of Petroleum Tec

Melrose, J. C. and C. F. Bra
    Determining Displacement
    Journal of Canadian Petr

Mohanty, K.K., H.T. Davis, a
    Entrapment in Water-Wet
    Technical Conference and

Morrow, N.R. (ed).  1984.  M
    Entrapment and Mobilizat
    2-70-3304, New Mexico En
    Fe, NM

Morrow, N.R. and I. Chatzis.
    Conditions for Entrapmen
    DOE/BC/10310-20, Departm

Smith,  L.  1981.  Spatial Var
    Sand.   Math. Geol. Vol.

Taber,  J.J. 1981.  Research
    Future.  In:  Surface Ph
    Shah  (Ed.) Plenum Publis
  of surfactants on mobility of residual
iding of transport of a discontinuous
jus media;
g of trapping mechanisms in relation to media
c biodegradation in groundwater.  Studio *
 surface and ocean water.
981.  Correlation of Capillary Number
nes.  S.P.E. Annual Technical Conference and
 H.T. Lim.  1983.  Magnitude and Detailed
 Saturation.  Soc. Petroleum Engineering
a, and N.R. Morrow.  1984.  Paper S.P.E.
  Technical Conference and Exhibit, Houston,
 tical Study of Core Waterflood Parameters.
nology, 31:1303-1304.

dpi.  1974.  Role of Capillary Forces in
Efficiency for Oil Recovery by Waterflooding.
leum Technology, Vol. 13.

d L.D. Scriven.  1980.  Physics of Oil
ock.  Paper S.P.E. 9406, 1980 S.P.E. Annual
Exhibit, Dallas, TX.

asurement and Correlation of Conditions for
on of Residual Oil.  Report No. NMERDI
rgy Research and Development Institute, Santa
 198?.  Measurement and Correlation of
 and Mobilization of Residual Oil.
nt of Energy, Washington, D.C.

ability of Flow Parameters  in a Stratified
3.

n Enhanced Oil Recovery, Past, Present,  and
 lomcna in Enhanced Oil Recovery 1981, D.O.
                                     146

-------
Wilson, J.L. and S. H. Conrad.  1984.  Is Physical Displacement of Residual
    Hydrocarbons a Realistic Possibility in Aquifer Restoration?  In;
    Proceedings of Petroleum Hydrocarbons and Organic Chemicals in Ground
    Water Prevention, Detection, and Restoration, Houston, TX, pp. 274-298.
                                     147

-------
                          SECTION 6.  LOCUS NO. 6

                                 CONTENTS
                                                                   Page No.

     List of Figures	     149
6.1  Locus Description	     150

     6.1.1  Short Definition	     150
     6.1.2  Expanded Definition and Comments	     150

6.2  Evaluation of Criteria for Remediation	     150

     6.2.1  Introduction	     150
     6.2.2  Mobilization/Remobilization	     150

            6.2.2.1  Partitioning Into/Onto Mobile Phases	     150

                     Partitioning into Air	     150
                     Partitioning into Water	     153

            6.2.2.2  Transport of/with Mobile Phase	     153

                     Introduct Ion	     153
                     Intrinsic Permeability and
                       Hydraulic Conductivity	     154
                     Physical Property Temperature-
                       Depende: ice	     155
                     Hydraulic and Fluid Conductivity	     155
                     Relative Permeability	     156
                     Enhancement of Transport	     157
                     Darcy's Lav	     158
                     Capillary Tension	     160
                     Mob ill ty Enhancement	     162

     6.2.3  Fixation	,	     162

     6.2.4  Transformation....	     163

            6.2.4.1  Biodegradc tion	     163
            6.2.4.2  Chemical (xidation	     163

6.3  Storage Capacity in Locus	     163

     6.3.1  Introduction and Basic Equations	     163
                                    148

-------
                                (Continued)
                                                                   Page No.

     6.3.2  Guidance on Inputs for, and Calculation of,
              Maximum Value	.	     163

            6.3.1.1  Basic Equations	     164
            6.3.1.2  Guidance on Parameter Values	     164

     6.3.3  Guidance on Inputs for, and Calculation of,
              Average Value	     165

            6.3.3.1  Basic Equations	     165
            6.3.3.2  Guidance on Parameter Values	     166

6.4  Example Calculations	     168

     6.4.1  Storage Capacity Calculations	     168

            6.4.1.1  Maximum Quantity	     168
            6.4.1.2  Average Quantity	     168

     6.4.2  Transport Rate Calculations	     170

6.5  Summary of Relative Importance of Locus	     170

     6.5.1  Remediation	     170
     6.5.2  Loci Interaction	     170
     6.5.3  Information Gaps	     170

6.6  Literature Cited	     171
6.7  Additional Reading	     171
                                  FIGURES
6-1  Schematic Cross-Sectional Diagram of Locus No. 6 - Liquid
     Contaminants in the Pore Spaces Between Solid Particles
     or Rock Surfaces in the Unsaturated Zone                        151

6-2  Schematic Representation of Important Transformation and
     Transport Properties Affecting Other Loci                       152

6-3  Flow Response of Soils to Hydraulic Gradient                    158

6-4  Hydraulic Conductivity Versus Capillary Suction for
     Different Soils                                                 161
                                    149

-------
                          SFCT10N 6 - LOCUS NO. 6
6.1  Locus Definition

6.1.1  Short Definition

    Liquid contaminants in the pore spaces between soil particles in the
unsaturated zone.

6.1.2  Expanded Definition and Comments
    The liquid contaminants
configuration, (i.e, no subs
in locus 6 are in the "water-wet"
tantial contact between liquid contaminant and
the surfaces of the soil particles).  Soil air is present and liquid
contaminant-air interfaces exist.  When a spill front from a large liquid
release passes through the unsatuiated zone, the liquid contaminants may
form a continuous phase, filling a large fraction of the void volume.
After the spill front has passed, a discontinuous phase is more likely.
The mobility of liquid contaminants in locus 6 depends on the volume of the
spill, its physicochemical p
roper ties, and the hydraulic properties of the
diagram of locus no. 6 and a
and transport processes affe

6.2  Evaluation of Criteria
6.2.1  Introduction
porous medium.  The liquid contaminants move in response to gravity and
capillarity.  Figui.es 6-1 ani 6-2 present a schematic cross-sectional
 schematic representation of the transformation
cting other loci, respectively.
tor Remediation
    The evaluation of criteria foi remediation of liquid contaminant in
locus no. 6 is based on the  physical and chemical processes which control
the movement of the liquid contaminant through the unsaturated zone.  These
physical and chemical factor.3 (e.g., hydraulic conductivity, kinematic
viscosity, and capillary tension) are discussed in sections 6.2.2 and
6.2.3.

6.2.2  Mobilization/Remobil i ation
6.2.2.1  Partitioning Into/Ofrto

Partitioning into Air (Detai
    Since the liquid contami ant
discussion of partitioning o
the liquid contaminant leave
the movement of  the liquid c<
section 6.2.2.2  and applies
well.
    Mobile Phase

 s in Section 1.2.2)
     enters locus no. 6 as a mobile phase,
  the liquid contaminant should focus on how
  the mobile phase.  A detailed discussion of
 ntaminant through locus no. 6 is presented in
 o movement of liquid water in locus no. 12 as
    Depending on whether air and water are present and  in  contact with  the
liquid contaminant in locus no. 6,  the contaminant may  volatilize into  the
air or dissolve into the wateir.  One of  the factors controlling  the
volatilization of a liquid contaminant is its pure chemical vapor pressure.
                                     150

-------
  NOTE: NOT ALL PHASE BOUNDARIES ARE SHOWN.
                                  LEGEND
              CONTAMINANT VAPORS
                              PHASE
        AAAA WATER FILM
 X  BIOTA (MICROBKDTA INCLUDED)

:~D SOIL PARTICLE

 I  COLLOIDAL PARTICLE
            J AIR (PLAIN WHITE AREAS)
Figure 6-1. Schematic Cross-Sectional Diagram of Locus No. 6 -
          Liquid Contaminants in the Pore Spaces Between Soil
          Particles or Rock Surfaces in the Unsaturated Zone.
                                  151

-------
              VOLATILIZATION
(SC
LO(
1 - CONTAk
ILGAS)
:USNO
INANT VAPORS
CONDENSATION

(BIOTA)
LOCUS NO
1 1 - SOBBED TO BIOTA

DEPURATION
«4—
UPTAKE
/ LOCUi
SPACES BETWEE
— OR ROCK J
\ UNSATUR
s NO. 6 \
N SOIL PARTICLES ^
>URFACES IN "^"
ATED ZONE) /
DISSOLUTION
—+'
PHASE
SEPARATION
(WATER)
LOCUS NO
3- DISSOLVED IN WATER FILM
12- DISSOLVED IN MOBILE
PORE WATER
                                  PERCOLATION
                                  CAPILLARY SUCTION
                                  RETARDATION OF CONSTITUENTS
                                  ADVECTION
                                  DIFFUSION
(LIQUID C
1£
2- ADHERING 7
PARTICLES
7- FLOATING O
13 - IN ROCK FR>
DNTAMINANTS)
CUS NO.
O -WATER-DRY* SOL
M WATER TABLE
CTURES
Figure 6-2. Schematic Representation
           and Transport Processes
      of Important Transformation
      Affecting Other Loci.
                                152

-------
A liquid contaminant with a high puie chemical vapor pressure (e.g.,
isobutane) may readily volatilize into the air.  If the liquid contaminant
has a relatively low vapor pressure (e.g., benzene), it may not evaporate
as readily.  The pure chemical vapor pressure of a liquid increases
exponentially with temperature.  The variation of temperature in a natural
environment is discussed in some detail in section 6.2.2.2.

    Another factor affecting the volatilization of the liquid contaminant
is the volume of pure air into which the liquid contaminant volatilizes.
Once the saturation vapor pressure in the air above the liquid contaminant
is reached, the air and liquid contaminant phases are in equilibrium; no
further net exchange of mass between the two phases occurs.  If an ambient
or artificial pressure gradient exists, air containing the volatilized
contaminant is replaced with fresh air into which more liquid contaminant
volatilizes.

    Diffusion may limit volatilization if it is the predominant transport
mechanism for moving the volatilized contaminants away from the liquid-
vapor interface.  The factors which influence the extent of volatilization
and subsequent vapor phase transport are discussed in detail in locus 1.

Partitioning into Water (Details in Section 7.2.2)

    Although most organic contaminants are sparingly soluble in water, some
liquid contaminant may dissolve.  Aromatic hydrocarbons (e.g., benzene,
toluene, xylene) dissolve more readily into water than chemicals such as
isobutane.  The factors influencing dissolution of liquid contaminant into
water are discussed in detail in locus no. 7.  Solubilities of compounds
typically found in gasoline are presented in Table 7-1.

6.2.2.2  Transport of/with Mobile Phase

Introduction

    The mobilization/remobilization of a separate immiscible contaminant
phase within the pore spaces of the unsaturated zone is controlled by
physical properties and conditions of the porous medium, contaminant phase,
and other immiscible phases present in locus no. 6.  Three immiscible
fluids — liquid contaminant, liquid water, and air — may be in direct
physical contact with one another.  In the sections below, the influence of
the properties of the porous medium and various phases on the mobilization
and subsequent transport of the liquid contaminant through the unsaturated
zone are discussed.

    The physical properties of the porous medium and of the liquid
contaminant (e.g., intrinsic permeability, liquid density, and liquid
viscosity) responsible for transport of the liquid contaminant through
locus no. 6 are also responsible for initiating the movement of the  liquid
contaminant in locus no. 6.  These same properties which allow movement may
change spatially and temporally within a natural environment, arresting the
movement of the contaminant liquid within locus no. 6.

-------
Intrinsic Permeability and Hy

    One of the key parameters
liquid contaminant is the int
Intrinsic permeability is sol
of the porous medium and gene
to clay soils.  Liquid contam
easily, with little resistanc
intrinsic permeability.

    Intrinsic permeability is
unsaturated zone to the flow
properties of the liquid.  A
fluid conductivity, K    , wh
permeability of the porous me
kinematic viscosity of the li
conductivity is adopted; flui
porous medium transmits any f
                    K
                     sat ,w
    where
K
               sat ,w
              v
               w
               w
               w
          = sat

          = int

          = kin

          =- liq

          = gi'a

          = dyn
    The kinematic viscosity o
viscosity to its density.  A
viscosity has a low kinematic
porous medium than a liquid w
dynamic viscosity, density, a
selected organic contaminant
al., 1982):

      Compound

liquid density, (g/cm  )

liquid dynamic viscosity,
x 10  , (g/cm'sec)

liquid kinematic viscosity,
x 10" , (cin /sec)

    The lower kinematic  visco
and isopentane will flow fast
A more extensive table of  kin
7-1.
                 aulic Conductivity

                 overning the initiation of movement of
                 nsic permeability, k, of the porous medium.
                ly a function of the grain-size distribution
                ally decreases from sands to loams to silts
                nant may move through the unsaturated zone
                 to flow, through a porous medium with a high
                a measure of the resistance of the
                 : a liquid and is independent of the physical
                ore useful flow parameter, however, is the
                ch is proportional to the intrinsic
                lum and inversely proportional to the
                uid,  v,.   For water, the term hydraulic
                 conductivity describes the ease with which a
                uid.
                 g/v  = k p  g/u
                 6  w     Kw 6  w
                                        (6.1)
rated hydraulic conductivity (cm/sec)
                      2
insic permeability (cm )
                            2
matic viscosity of water (cm "/sec)
                         3
id density of water (g/cm )
                              2
ity constant (g = 980.7 cm/sec )

mic viscosity of water (g/cm sec)
                 a liquid is the ratio of its dynamic
                igh density liquid with a low dynamic
                viscosity; it moves more quickly through a
                th a higher kinematic viscosity.  Values for
                d kinematic viscosity of water and several
                iquids at 20°C are tabulated below (Lyman et
                Water

                 1.0


                 1.0


                 1.0
                Benzene
Isopentane
                 0.89
                 0.64
                 0.72
  0.62
  0.52
  0.84
                ity values indicate  that both liquid benzene
                  than water through  the same porous medium.
                matic viscosity values  is presented in Table
                                     15/4

-------
Physical Property Temperature- Dependence

    Both liquid density and dynamic viscosity are temperature-dependent,
increasing as temperature decreases.  Dynamic viscosity is, however, more
strongly dependent on temperature than is density.  As temperature
decreases, the kinematic viscosity of the liquid increases; colder liquids
have a higher kinematic viscosity and move more slowly through the same
porous medium than a warmer liquid.

    Because temperature varies in the subsurface on a daily and seasonal
basis, the kinematic viscosity of any liquids (e.g., water or liquid
contaminant) may vary similarly.  Temperature varies in the subsurface with
a diurnal (i.e., daily) cycle, warming to a peak by mid-day and cooling
during the night (Rosenberg et al., 1983).  Soil temperature also
fluctuates seasonally, freezing the top several meters of the soil in some
latitudes.  Beyond two meters, however, temperature of the soil remains
fairly constant daily and seasonally.

    Soil type and moisture content also influence the response of a soil to
temperature changes.  Generally, sandy soils transmit heat more efficiently
than loamy and clayey soils.  The same increase in surface temperature
produces a larger temperature change in the sandy soil than in the clayey
soil for the same depth.  The presence of water in the soil ameliorates
temperature extremes because of the high heat of vaporization of water.
Soil type, moisture content, latitude, and season are important factors
governing the fluctuations of temperature in the subsurface, which in turn
affects the kinematic viscosity of the water or liquid contaminant in locus
no. 6 and its mobility within the locus.

Hydraulic and Fluid Conductivity

    Hydraulic conductivity is measured in the field or in more accurate
laboratory experiments on extracted soil samples.  The intrinsic
permeability may be calculated from the experimentally-determined hydraulic
conductivity of the soil materials to a given liquid.  The calculated
intrinsic permeability may then bp used to estimate the fluid conductivity
of the same soil material for another liquid with a different kinematic
viscosity.  A sample calculation is given below based on properties of a
synthetic gasoline blend in Table 7-1.
                                                    _2
   i.    measured K      for soil with water, K = 10   cm/sec
         density for witer, p  =1.0 g/cm     „
         dynamic viscosity for water, y  = 10   g/cm sec

                                      -7   2
  ii.    calculated k for soil, k = 10   cm
                                            3
 iii.    density for gasoline, p, = 0.7 g/cm            „
         dynamic viscosity for gasoline,  u, = 4.92 x 10"  g/cm sec
                                                               _2
  iv.    estimated K for soil with gasoline, K      = 1.44 x 10   cm/sec
                                              sat,g
                                    15.-i

-------
    The sample calculation indicates that the same porous medium is able to
conduct the synthetic gasoline about forty four percent faster than water
because of the difference in

Relative Permeability
                 liquid  kinematic viscosity.
    Variation in saturation of a fluid in locus no. 6 affects the fluid
conductivity of the partially
there are several immiscible
contaminant) the increase in
fluids.  When the porous med
                 -saturated porous medium to that fluid.   If
                 fluids present  (i.e.,  air,  water,  and liquid
                 saturation of one immiscible liquid reduces
the fluid conductivity of th<> porous medium relative the other immiscible
                 urn is relatively unsaturated with the liquid
contaminant, the fluid conductivity of the porous medium for the liquid
contaminant is at a minimum
immobile.
    For an immiscible fluid in a porous medium, the variation in fluid
conductivity with the degree
permeability for two or more
Dept., Colorado State Univer.<
relationship estimates relati
the saturation of the porou.r
                 nd the liquid contaminant may be relatively
                 of saturation is known as relative
permeability.  A commonly-used expression for estimating the relative
                 immiscible fluids is the Brooks-Corey
relationship (Personal Commuricat ions with O.K. Sunada, Civil Engineering
                 ity,  Ft.  Col-lins,  1987).   The Brooks-Corey
                 ve poumeabi lity for an immiscible liquid from
                 medium relative to that fluid and the
empirical grain-size distribution parameter:
wetting liquid
    k   =  l(S-S
     rw        r
non-wetting liquid
                              b)/b
    k    =  [1-(S-S )/(] -S  )]'
     nrw    l  v   r      r/J
    where  k

           kr
           S

           S
rw
nrw
     = ratio of i

     = grain-size
                                                          (6.2)
                                                          (6.3)
= wetting liquid relative permeability (dim.)
= non-wettin; liquid relative permeability (dim.)
= ratio of vjlume of water to pore volume (cm3/cm3)
            rreducible water volume to pore volume (cm3/cm3)
             distribution parameter (dim.)
and  the  range  in values  of  re
                   non-wettin
                   0  <  k    <
                      -  rw  -
                  ative permeability for the wetting and
                   liquids are:
                   0  <  k
                         nrw -
    o   Sample  calculation  of
        permeability  at  b  =  2.H  and  S   =  0.05.
                  etting and non-wetting liquid relative

-------
                   Wetting Liquid             ,          ,
                   Saturation (S)              rw        nrv

                        0.0                  0.0        1.0
                        0.25                 0.003      0.580
                        0.50                 0.062      0.200
                        0.75                 0.322      0.028
                        1.0                  1.0        0.0

    This example shows the importance of saturation on the relative
permeability of a fluid in the presence of another immiscible fluid; the
saturation must be high for the fluid to flow readily.

    The increase in fluid conductivity with saturation is due to the
decrease in frictive resistance to flow per unit cross-sectional area which
decreases with increasing pore size.  Small pores exert a higher resistance
to flow per unit cross-sectional area than larger pores and have,
therefore, lower fluid conductivity than larger pores for the same liquid.

    At low saturation levels, the smallest pores of the medium are filled
with a liquid because of high capillary tension exerted by the smaller
pores.  Because the liquid exists within the smaller pores at higher
capillary tension, the fluid conductivity of the medium to that liquid is
small.  As saturation increases, larger pores begin to fill with liquid,
and the fluid conductivity of the porous medium to the liquid increases to
a maximum at complete saturation.

Enhancement of Transport

    The preceding discussion described how fluid conductivity affects
mobilization and transport of the liquid contaminant within locus no. 6.
If mobilization and transport of the liquid contaminant is desired, some
enhancement of the fluid conductivity of the porous medium to the liquid
contaminant is necessary.  Such an enhancement of the fluid conductivity
might be accomplished by changing one or more of the parameters (i.e.,
intrinsic permeability, liquid density, dynamic viscosity, and saturation)
upon which the fluid conductivity depends.

    As stated above, the fluid conductivity is directly proportional to the
intrinsic permeability of the porous medium.  It is possible, albeit
perhaps impractical, to enhance the intrinsic permeability of the porous
medium by removing some of the fine-grained material of the porous medium.
Such a removal of fine-grained material often occurs when a newly-installed
water well in unconsolidated porous materials first pumps water.

    The removal of fine-grained material from the porous medium by the
newly-installed well usually is limited to the immediate volume of porous
medium surrounding the well and usually ceases shortly after initiation of
pumping, if the water well has been properly designed.  The removal of
fine-grained material from the porous medium may enhance locally the
intrinsic permeability, but would be expensive to implement on a large
scale.  Furthermore, substantial removal of porous medium materials may
result in subsidence of the Jand surface with an attendant decrease in the
fluid conductivity because of compaction.  In summary, the intrinsic
                                    157

-------
permeability probably  cannot
mobilization and  transport.

    It is difficult  to imagii
contaminant could  be enhancec
porous medium  to  the liquid
come with a lowering of the
liquid contaminant possibly  c
emulsifying compound,  alterir
state.  The reduced dynamic
would allow the liquid contan
unsaturated porous medium of

    The relative  permeability
liquid contaminant could be
liquid contaminant in  locus  r
increasing the potential for
saturation, the other  princi
reduction of water content ir
permeability of the  porous mf
contaminant; the  liquid contc

Darcy's Law

    The role of hydraulic co
visualized as  the slope of t
area against the  hydraulic g
depicts the flow  response of
hydraulic gradient.  The ste
reflects its higher  hydrauli
                  Sandy Soil
            Source: Hillel, 1980. (Copyrig

              Figure 6-3. Flow Resp
   significantly enhanced  to induce
  how the liquid density  of  the liquid
 to increase the fluid  conductivity of the
>ntaminant; small  increases  in density would
 mperature.  The dynamic  viscosity of the
)uld be altered with  the  addition of some
g the liquid contaminant  to  a more fluid
 scosity of the emulsified  liquid contaminant
 nant to flow more readily  through the
 ocus no. 6.

 of the porous medium of  locus no. 6 to the
ncreased by increasing  the  saturation of
   6.  But a healthier  approach toward
mobilization would be to  reduce the water
al immiscible phase in  locus no. 6.  The
 locus no. 6 would reduce the relative
 lium to water and  increase  it for the liquid
minanl would then  be  more mobile.
 lurtivity or  fluid  conductivity can be
  line produced  by  plotting the flow per unit
adient which  produces  that flow.  Figure 6-3
a sandy soil  and a  clayey soil to changes in
 ei  slope associated with the sandy soil
 conductivity relative to the clayey soil.
iraulic Gradient (AH/AX)

 Academic Press, Inc. 1980. Reprinted with permission.)

 nse of Soils to Hydraulic Gradient
                                     158

-------
    This simple linear relationship is known as Darcy's Law.  Darcy's Lav
states that the rate of flow of a volume of the liquid contaminant through
the porous medium of locus no. 6 is proportional to the stress applied to
the liquid; the higher the stress, the more liquid flows.  The stress is
applied by ambient and artificial pressure gradients and from the downward
force of gravity.  A one-dimensional, vertical expression for Darcy's Law
is:

                   q = (k k /y) (dP/dz + p g dh/dz)                   (6.4)


    where     y      = dynamic liquid viscosity (g/cm sec)
                                                 2
              k      = intrinsic permeability (cm )

              q      = discharge of liquid per unit cross-sectional flow
                            /  3/      2^
                       area (cm /sec cm )

              k      = relative permeability (dim.)

              dP/dz  = vertical ambient or artificial pressure
                                         -2   2
                       gradient (g cm sec  /cm  cm)

              dh/dz  = gravity gradient (cm/cm)

              z      = depth (cm)
                                           3
              p      = liquid density (g/cm )
                                                         2
              g      = gravity constant, g = 980.7 cm/sec
    Generally, in most natural settings of a liquid contaminant in the
unsaturated zone, it is difficult to apply an external artificial pressure
stress to the unsaturated zone having anything more than a limited areal
effect (Personal Communication with D.L. Marrin, Consultant, LaJolla, CA,
1987).  By the same token, ambient pressure gradients, while certainly
affecting the air phase of locus no. 6, generally are of small magnitude
and have a small effect on the movement of liquid contaminant.  By
neglecting the effects of artificial and ambient pressure gradients on the
liquid contaminant in locus no. 6, the remaining force which transports the
liquid contaminant is gravity.  Assuming a unit vertical gravity gradient
and no external pressure gradients:

              dh/dz  = 1

and

              dP/dz  = 0

Darcy's Law becomes:

              q = k kr p g/y                                          (6.5)

-------
Capillary Tension

    Gravity produces a down
6.  Acting against the down
force of capillary tension.
partially saturated porous
result of the surface tensi
solid grains of the porous
exerted on the liquid by t\
also varies with saturation
lowest saturation levels.
tension and saturation was
    where
              e
              e
capilla
capilla
(cm H20
moistur
moistur
empiric
    Representative values :
Table 12-2 (Section 12).

    Generally,  the smallei
range in capillary tension
(i.e., 1 atm) in  the large
suction at -15  atm) in  the
locus no. 6  is  open to  the
gradient between  a liquid
less-than-atmospheric press
calculations with paramete
               6  (cm  /cm  )

               
-------
               Same  materiaIs, different  moisture content.
                       ^e ' ~                  vp
                    /        x
                    (cm /cm )
                    (cm H20)
Sand.
Sand,
0.20
0.35
190
 20
    The sample  calculations indicate  that  the loam holds the  same  amount of
water as  the  sand  at a much higher  capillary tension.  The calculations
also show  that,  for the same sand,  a  smaller water volume is  held  at a
higher capillary tension than a larger  water volume.

    The capillary  pressure gradient acts on a liquid entering the
unsaturated medium of locus no. 6,  causing the movement of liquid  in the
direction  of  decreasing capillary pressure (i.e., from high  to low
pressure).  The movement ot liquid  in locus no. 6 from high  to low pressure
areas typically causes the liquid contaminant to spread laterally  as it
moves downward.

    The dependence of hydraulic conductivity and capillary tension on
moisture content implies that hydraulic conductivity is also  dependent on
capillary  tension.  As the soil is  dewatered, the capillary  tension
increases, and  the hydraulic conductivity  decreases.  The effect of
increasing capillary tension on hydraulic  conductivity varies with soil
texture and can be seen in Figure 6-4.
                                Sandy Soil


                                      Clayey Soil
                                Suction
            Source: Hillel, 1980. (Copyright Academic Press, Inc. 1980. Reprinted with permission.)

                Figure 6-4. Hydraulic Conductivity Versus Capillary
                        Suction for Different Soils (log-log scale)
                                      16

-------
    Figure 6-4 indicates tha
clayey soils decreases with
important point to be drawn
tension reaches a certain po
soil actually exceeds that o
the fact that the sandy soil
moisture is held more strong
resulting in a larger hydrau
capillary tension.

Mobility Enhancement

    The mobilization of the
enhanced by reducing the eff
obvious but unfeasible metho
contaminant, increasing the
appropriate alternative migh
contaminant which would lowe
emulsified liquid contaminan

6.2.3  Fixation

    As stated above, many of
that allow the initiation of
contaminant in locus no. 6 a
liquid contaminant.  The fix
thus arises from natural, or
parameters which may be botl

    The intrinsic permeabili
the result of compaction'of
by groundwater removal.  The
reducing its ability to trar
the porous medium could be e
giving the site added protec
contaminant.

    Variability in sediment
natural environment and inti
spatially variable.  For exa
have an intrinsic permeabili
liquid contaminant with tran
that the sediments of locus
permeability of the porous m
movement of  the liquid conta
permeability.  A new UST pla
vertical decrease in grain
natural protection against w
contaminant migration.

    The relative permeabili
contaminant liquid could be
This reduction could be acco
liquid contaminant by pumpii
 hydiaulic conductivity for the sandy and
ncreasing capillary tension.  Another
rom Figure 6-4 is that when the capillary
nt, the hydraulic conductivity of the clayey
 the sandy soil.  This means that, in spite of
may conduct fluids more readily at saturation,
y by the finer pores of the clayey soil
ic conductivity of the clayey soil at high
ontaminant liquid in locus no. 6 could be
ct of capillary tension on the liquid.  An
 for doing this would be to add more liquid
aturation - not a desirable situation.  A more
 be the addition of emulsifiers to the liquid
 the surface tension of the liquid, making the
 more mobile.
the hitherto described transport parameters
movement and translocation of a liquid
so can retard and arrest movement of  the
tion of the liquid contaminant in locus no. 6
rtificial changes induced in the transport
spatial and temporal.

y of the porous medium may reduce with  time as
he giound by application of external  weight or
result would compress the porous medium,
mit any liquid or fluid.  Such a compaction of
fected prior to emplacement of a new  UST,
ion against migration of the liquid
izes and distributions  is often  large  in  the
nsic permeability of a  porous medium is
iple, the upper portion  of locus  no. 6  might
y high enough to initiate movement of  the
location.  Moving downward,  it is conceivable
o. 6 become  finer,  reducing  the  intrinsic
dium.  At some depth, it is  possible that the
linant would  become  arrested  by a low intrinsic
ed in such a natural environment with  a
ze and intrinsic permeability would have  a
despread vertical and lateral liquid
 of  the  porous  medium  of  locus  no.  6  to  the
 educed sufficiently  to arrest  the  liquid.
 plashed  by  lowering  the saturation of the
   This  may be  a  costly and  inefficient
                                     16?

-------
remedial alternative if the liquid contaminant exists in locus no. 6 as a
discontinuous phase.  Alternatively, the relative permeability of
the liquid contaminant may be reduced by the increase in saturation of
water.  The problem with this alternative, apart from the increase in
contaminated water, is that the added water might force the liquid
contaminant ahead of it, displacing it instead of immobilizing it.

    The density and dynamic viscosity of the contaminant liquid changes
with time because of changes in the composition of the liquid contaminant.
If the air phase is in contact with the liquid contaminant, volatile
components of the liquid contaminant may evaporate.  For example, in the
case of gasoline, the lighter-molecular weight alkane chemicals (e.g.,
isobutane, isopentane, n-butane) volatilize readily.  As the contaminant
liquid contacts pure water in the unsaturated zone, the more soluble
components of the contaminant liquid (e.g., simple aromatic
hydrocarbons) dissolve into the water.  The aged contaminant liquid left as
residuum by the processes of volatilization and dissolution will be a more
viscous immobile liquid compared to fresh contaminant liquid.

6.2.4  Transformation

6.2.4.1  Biodegradation (Details in Section 11)

    Biodegradation is most likely to take place in an aerated aqueous
environment in the unsaturated zone.  Therefore, liquid contaminant in
locus no. 6 is likely to be degraded only after passing into the water film
which wets the soil particles aiound which they flow.  Some microbes can
live while suspended in pure hydrocarbon but these are rare and metabolize
oil slowly.  Therefore, locus no. 6 contaminants would likely diffuse into
other loci before they could be biodegraded in place.

6.2.4.2  Chemical Oxidation (Details in Section 3.2.4.2)

    Liquid contaminants in soil which do not come in contact with mineral
surfaces may not undergo catalytic transformations.  In order for chemical
oxidation to occur in this locus, an oxidant such as ozone or peroxide must
be introduced.

6.3 Storage Capacity in Locus

6.3.1  Introduction and Basic Equations

    Determination of the amount of liquid contaminant held in locus no. 6
is based on equations and guidance on selection of appropriate parameter
values given below.

6.3.2  Guidance on Inputs for, and Calculations of, Maximum Value

    The maximum quantity of liquid contaminant held in locus no. 6 occurs
in the presence of a liquid contaminant spill front.  Under this condition,
it is assumed that:

    o   the migrating contaminant liquid fills the air-filled porosity;

-------
    o   the liquid contaminant and the soil water are immiscible; and

    o   the moisture-filled porosity is assumed to be minimal so that it
        does not impede the
           vertical migration of the contaminant front nor
        occupies a significant fraction of the total porosity.
    It is also assumed that
larger than any quantity of
           the quantity of liquid contaminant is very much
           contaminant mass lost to volatilization into
and uptake, and attenuation
6.
the soil air, dissolution into the vadose zone water, microbial degradation
            o soil  particles.
6.3.1.1  Basic Equations

    Mass of liquid contamination per unit volume of soil held in locus no.
    where
              g
= mass of 1
                   contaminant (kg/m )
                 = density o
    Air-filled porosity in 1
                                                                      (6.7)
quid contaminant per unit volume of
             9   = air-filled void space filled with contaminant liquid
              3      33
                   (m /m )
    where    0   = total porosity (m /m )
             liquid contaminant (kg/m )
           )cus no. 6.
                                    3, 3,
                                                                       (6.8)
             0   - watei-f il
              w
                          3  3
            ed porosity (m"/m )
6.3.1.2  Guidance on Parameter Values

    Total Porosity, 0t.

    o   Theoretical maximum  orosity for a sandy soil  is  47.6%.
    o   Refer to Table 12-1
        variety of consolida

    Water-Filled Porosity, 9

    o   Water-filled porosit.v

    o   The air-filled poros:
        at wilting point, uhi
        12-4.
            or representative total porosity values  for a
            ed and unconsolidated sediments.
             ranges from 0 to 6  .

            ty  is maximized by choosing a moisture  content
            cli  is available for  various soils  from  Figure
                                     164

-------
6.3.3  Guidance on Inputs for, and Calculation of, Average Value

6.3.3.1  Basic Equations

    The average quantity of liquid contaminant in locus no. 6 is defined as
the quantity of liquid remaining in the locus after the passage of the
contaminant spill front.  The available porosity within which the volume of
residual liquid contaminant may bp held in locus no. 6 under these
conditions is affected by the water-filled porosity:

                          6    =9-9                               (6.9)
                           ac     t    w                              v   '
                                                               3  3
    where   6   = air-filled and contaminant-filled porosity (m /m )
             clC

    The air-filled porosity, 6 , is calculated from:
                              3

                            9=9-9                           (6.10)
                             a      ac    c                          v
                                                3  3
    where:  9,  = contaminant-filled porosity (m /m )

    With the passage of the contaminant front, some quantity of liquid
contaminant remains trapped by capillary forces or by other trapping
mechanisms.  At this point, the gravity forces causing drainage are
balanced by capillary forces and the liquid contaminant volume is at the
field capacity of the soil.  Clapp and Hornberger (1978) estimated the
contaminant-filled porosity from:

                      9c = 9^Ufr/>s)~1/b                           (6.11)

                                                              3
    where   9   =  saturation content of liquid contaminant (m  of liquid
             S                        3
                   contaminant per 1 m  of soil volume)
              ;  =  capillary tension at field capacity (cm H«0)
                =  capillary tension at saturation (cm H«0)
                =  grain-size distribution parameter (dim.)
    The contaminant-filled porosity can be converted to a mass of residual
liquid contaminant per unit volume of soil:

                           m     .-  9 .p,                             (6.12)
                            i es      c  1                            ^

    where   m    = mass or residual liquid contaminant per unit soil volume
             I- fc.O        n
                                                      3
            p-,   = density of liquid contaminant (kg/m )


    Since not all the porosity is tilled by a continuous supply of liquid
contaminant as described in section 6.3.2, liquid contaminant can be lost
to dissolution to vadose zone water, volatilization into vadose zone air,
microbial degradation, and chemical attenuation to soils.  When these

-------
losses are subtrnrted from th
mass of liquid contaminant pe
                  m  = m
                   g    res
    where   m    = mass of li
             g
                   no. 6 afte

            m    = mass of li
             res
                   residual q

            m    = mass of li
             w
                   dissolutio

            m    = mass of li
             v
                   volatiliza

            m,    = mass of li
             b
                   microbial

            m    = mass of li
             s
                   attenuatio
    Losses to microbial degra
negligible for the present an
contaminant lost to dissoluti
from:
    where
S = solubility o
    The mass of liquid contam
zone air is calculated from:
    where    p   = vapor densi
             sg
                  mixture (kg
6.3.3.2  Guidance on Paramete

    Grain-Size Distribution P

    o   Empirical factor rela
        increasing with deer*
        representative value?
                  residual  mass  of  contaminant  liquid,  the
                  unit  volume  of soil  held in locus no.  6 is:
                 m  -  m  -  m,  -  m
                  w     v    b     s
                                       (6.13)
                 uid  contaminant  per unit  volume held in locus
                                                    3
                  contaminant  front  has  passed (kg/m )
                 uid  contaminant  per unit  volume held as
                                            3
                 antity in  locus  no. 6 (kg/m )
                 uid  contaminant  per unit  volume lost to
                                              3
                  into vadose  zone water (kg/m )
                 uid  contaminant  per unit  volume lost to
                                               3
                 ion  into vadose  zone air  (kg/m )
                 uid  contaminant  per unit  volume lost to
                                 3
                 egradation (kg/m )
                 uid  contaminant  per unit  volume lost to
                  to  soils  (kg/m  )
                 at ion and attenuation will be assumed to be
                 lysis,  m,  = m  =0.   The mass of liquid
                 11 into the vaaose zone water is calculated
                               -  s'e /icr
                                     w
                                       (6.14)

 liquid contaminant in water (mg/L)

nanl  lost to volatilization into the vadose
                 y of volatilized liquid contaminant-air



                  Values

                 rameter, b.

                 ed to soil grain size and distribution,
                 sing grain size.  Refer to Table 12-2 for
                 of b foi a variety of soils.
                                     106

-------
    Capillary Tension at Field Capacity, (J/, .

    o   Use whatever capillary tension desired for field capacity.
        Recommend using 0.3 bai tension (i.e., 306 cm H~0).

    Capillary Tension at Saturation, \|/ .

    o   Use values reported in Table 12-2 for various soils.

    Saturation Content, Q .

    o   Use values reported in Table 12-2 for various soils.

    Solubility,  S.

    o   User should supply solubility of liquid contaminant of interest.

    o   For gasoline containing no oxygenated additives, use 200 mg/L for
        solubility.  See locus 3 for detailed information of solubilities
        of organic chemicals commonly found in gasolines.

    Vapor Density, p
                                                                     3
    o   For gasoline vapors, Table 1-2 indicates a range of 0.51 kg/m  at
        0°C to 1.04 kg/m  at 20°C.

    o   Vapor density may be calculated for any temperature and gasoline
        from:

                           p    =  M   P /T R                        (6.16)
                            sg      ca  a                            ^

        where   M     = average molecular weight of volatilized liquid
                 C3.
                        contaminant-air mixture (g/mol)
                P     = atmospheric pressure, P  = 760 mm Hg
                 3.                             3.
                T     = temperature (K)

                R     = universal gas constant,
                        R = 0.06236 mm Hg m3/mol K
    The average molecular weight ol the volatilized liquid contaminant air
mixture is calculated from:

                   M    -  M  (P../P )  +  M  (1 - P./P )             (6.17)
                    ca      c v t  a       a       t  a'             ^    '

    where:  M  = molecular weight of volatilized liquid contaminant (g/mol)

            M  = molecular weight of air = 28.97 g/mol
             3
            P  = equilibrium vapoi pressure of liquid contaminant (mm Hg)
                                    K./

-------
6.4  Example Calculations

6.4.1  Storage Capacity Calci

6.4.1.1  Maximum Quantity

    For this sample calculat
quantity of contaminant liqu
in the soil.  The wilting po
minimum water content.  For
point is about 5 percent (Se
sandy soil is assumed to be
air-filled porosity in which
    e   =  e. - e
     a      t    w
    et  =  0.395

    e   =  0.05
     w
    9   =  0.345
     a
    Assuming a liquid contam
in Table 7-2:
    p  = 710 kg/m
     o

    The maximum mass of liqu
locus no. 6 is:
     g      g  a

    m   =  (710 kg/m3) (0.34
     o
    m   =  245 kg liquid ron
     o

    The maximum of liquid co
no. 6 is 245 kg/m  or 345 L/

6.4.1.2  Average Quantity

    A sandy soil will also b
quantity of contaminant liqu
water-filled porosity is 5%
section 6.4.1.1, the maximum
34.5%.  Using equation 6.11,
field capacity is:
    6c  '  V<*fc'V
                      -1/b
    Values for thf> grain .six
at field capacity are taken
                             I ations
                             on,  a sandy soil is assumed.   The available
                             d is maximized by minimizing the water content
                             nt is a convenient measuring point for this
                             andy soils, the water content at the wilting
                              Figure 12-4).  The total porosity of the
                             9.5% (Clapp and Hornberger, 1978).  The
                             contaminant liquid may be stored is:
                             nant density similar to the gasoline reported
                             d contaminant per unit volume of soil held  in
                             ami nant ppr 1 m  of soil volume

                             tarninant mass per unit volume held in locus
                              assumed for the calculation of the average
                             d held in locus no. 6.  It is assumed  that the
                             nd the total porosity is 39.5%.  As stated in
                             available porosity for contaminant liquid is
                             the porosity occupied by contaminant liquid at
                              distribution parameter  and  capillary  tension
                              rom Clapp  and Hornbeiger (1978):
                                    1 f.H

-------
    \J/fc = 306 cm H?0



     <710 kg/'"3)



    m    = 110 kg/m3 or 155 L/m3
     res        6


    Mass lost to Volatilization:



    m    = 9  p
     v      a Msg


    m    = (0.19) (1.04 kg/m3)



    mv   = 0.198 kg/m3



    Mass lost to Dissolution:



    my   = S 6/103



    my   = (130 mg/L) (0.05)/103



    m    = 6.5 x 10~3 kg/m3  or  9.2 x 10"3 L/m3
     W
                                  *


    Mass held in Locus no. 6 uncloi average conditions:



    m    = m    - m  - m
     g      res    v    w


    m    = 110 kg/m3 - 0.198 kg/m3 -  6.5 x 10~3 kg/m3
     O


    m    = 109.8 kg/m3  or  154.6 L/m3
     O


    The mass of liquid contaminant, lost to biodegradation  and  alteration to

soil is assumed to be negligible; m,  = in  =0.

-------
6.4.2  Transport Rate Calcu

    Sample calculations aie

6.5  Summary or Relative Imp

6.5.1  Remediation

    The presence of liquid
for contamination of the so
contaminates unsaturatecl zon
The extent to which the liqu
locus no. 1 or loci no.'s 3
solubility, and the kinemat

    The amount of mass per i
density of the liquid and tl
volume of a denser liquid ma
calculations in section 6.4
fraction of liquid contamin
per unit volume as loci no.
1-4 and 8-13.

    If the liquid contaminar
fraction of the unsaturated
kinematic viscosity, and if
the contaminant liquid may
liquid contaminant may be a
contaminant is removed by p
in the pores of the unsatur
the saturation level of the
progressively harder to wit
volatility of the liquid co
much of the remaining Jiiquic

6.5.2  Loci Interactions

    Volatilization from loc
partitioning to locus no. I
unsaturated zone water (loc
particles  (loci no. 2 and 4
quantity or is mobile (i.e.
loci no.'s 7 and 8.

6.5.3  Information Gaps

    Information in the foil
understanding of immiscible
zone:

    1.  Chemical compositio
        contaminant;
tions

rovided in Section 6.2 for selected equations,

rtance of Locus
ntaminant in  the unsaturated zone  is a source
 air  (locus no. 1) by volatilization.  It  also
 water  (loci  no.'s 3 and  12) by dissolution.
d contaminant  in locus no. 6 contaminates
nd 12 depends  on the vapor pressure,
 viscosity of  the liquid  contaminant.

it volume held  in locus no. 6 depends on  the
 volume of pore space.  More mass  per unit
 be stored in  the same volume.  Sample
 indicate that  a dry sandy soil holds a large
t.  In  fact,  locus no. 6-can hold  as much  mass
 5 and  7, and  certainly more than  loci no.'s
  in  locus  no.  6  occupies  a  significant
one  porosity,  if the  liquid has  a  low
he intrinsic  permeability is sufficient,  then
  mobile.   If  this  is  the  case,  the pumping of
iable  remedial option.  As  the  liquid
iping,  however, some  fraction remains  trapped
ed zone.   As  described  in section  6.2.2.2,  as
iquid  contaminant  drops,  it will become
raw  more liquid  contaminant.  Depending  on the
aminant, however,  vacuum  extraction may  remove
contaminant.
  no.  6  represents  the  primary  mechanism for
  Liquid contaminant  may  dissolve into the
 no.'s 3 and  12)  and  attenuate  to soil
   If  the liquid  contaminant  is of sufficient
 low kinematic  viscosity),  it may transfer to
 ing areas  is  necessary for improved
 iquid  contaminant  transport in the unsaturated
 and  phy.sirochemical  properties of the liquid
                                     170

-------
    2.  Information on effects of trapping mechanism and capillary tension
        on liquid contaminant.

    3.  Understanding the behavior of three-phase flow in the unsaturated
        zone.

6.6  Literature Cited

Clapp, R.B. and G.M. Hornbergcr.  1978.  Empirical Equations for Some Soil
    Hydraulic Properties.  Water Resources Research, 14(4):601-604.

Hillel, D.  1980.  Fundamentals of Soil Physics, Academic Press.

Lyman, U.J., W.F. Reehl, and D.H. Rosenblatt.  1982.  Handbook of Chemical
    Property Estimation Methods, McGraw-Hill Book Co., New York.

Rosenberg, N.J., B.L. Blad, and S.B. Verma. 1983.  Microclimates:  The
    Biological Environment, 2nd ed., Wiley-Interscience.

6.7  Additional Reading

Aleman, M.A. and J.C. Slattery. 1988.  Estimation of Three-phase Relative
    Permeabilities. Transport in Porous Media, 3:111-131.

Baehr, A.L., G.E. Hoag, and M.C. Marley. 1989.  Removing Volatile
    Contaminants from the Unsaturated Zone by Inducing Advective Air-phase
    Transport. Journal of Contaminant Hydrology, 4:1-26.

Conrad, S.H., J.L.  Wilson, W. Mason, and W. Peplinski. 1988.  Observing the
    Transport and Fate of Pen oleum Hydrocarbons in Soil and in Groundwater
    Using Flow Visualization Techniques. Proc. of American Association of
    Petroleum Geologist? Symposium, May 10.

Cyr, T.J., V. de la Cruz, arid T.J.T. Spanos.  1988.  An Analysis of the
    Viability of Polymer Flooding as an Enhanced Oil Recovery Technology.
    Transport in Porous Media, 3:591-618.

Delshad, M. and G.A. Pope.  No date.  Comparison of the Three-phase Oil
    Relative Permeability Models, submitted to Transport in Porous Media.

Darcos, T.  1978.  Theoretical Considerations and Practical Implications on
    the Infiltration of Hydrocarbons in Aquifers, Int. Symp. on Ground
    Water Pollution by Oil Hydrocarbons, Prague, pp. 127-137.

Faust, C.R., J.H. Guswa, and J.W. Mercer. 1988.  Simulation of
    Three-dimensional Flow of Immiscible Fluids within and below the
    Unsaturated Zone, submitted to Water Resources Research.

Kuppusamy, T., J. Sheng. J.C. Parker, and R.J. Lenhard. 1987.
    Finite-element  Analysis of Multiphase Immiscible Flow Through Soils.
    Water Resources Research, ?3(4):6?5 631.
                                    1/1

-------
Reible, D.D., T.H. IHangasc
    Infiltration of Immisci
    submitted to Ground Watelr
Schramm, M., A.W. Warrick,
    to Four Organic Liquids,
    Materials, 3(1):21.
 kcir<", D.V. Doshi, and M.E. Malhiet. 1988,
  c Contaminants in the Unsaturated Zone,
a|nd W.H. Fuller.  1986.  Permeability of Soils
  and Water.  Hazardous Waste and Hazardous
Touma, J. and M. Vauclin.  1986.  Experimental and Numerical Analysis  of
    Two-phase Infiltration i|n a Partially Saturated Soil.  Transport  in
    Porous Media, 1:27-55.

Tyler, S.W., M.R. Whitbeck, IM.W. Kirk, J.W. Hess, L.G. Everett,  O.K.
    Kreamer, and B.H. Wilson.   1987.  Processes Affecting  Subsurface
    Transport of Leaking Und
 eiground Storage Tank Fluids, EPA/600/6-87/005,
    EPA Environmental Monitoring Systems Laboratory,  Las Vegas,  NV.
Van der Waarden, M., A.1..A.M
    Transport of Mineral Oil
    Experiments on  the Transfer
    Trickling Water. Water R
Van der Waarden, M., A.L.A.M
    of Mineral Oil  Component
    and Organic Soil Compone
    11:359-365.
   Mrdie, and W.M. Groenewoud.   1971.
  Components to Groundwater, I.   Model
     of Hydrocarbons from a Residual Oil  Zone  to
 esearch, 5:213-226.
   Birdie, and W.M. Groenewoud. 1977.  Transport
 s to Groundwater, II.  Influence of Lime,  Clay,
  its on the Rate of Transport. Water Research,
                                     I/'.-'

-------
                          SECTION 7.  LOCUS NO. 7

                                 CONTENTS
                                                                   Page No.

List of Tables	     172
List of Figures	     173
7.1  Locus Description	,	     176

     7.1.1  Short Definition	     176
     7.1.2  Expanded Definition and Comments	     176

7.2  Evaluation of Criteria for Remediation	     176

     7.2.1  Introduction	     176
     7.2.2  Mobilization/Remobilization	     179

            7.2.2.1  Partitioning onto Mobile Phases	     179

                     Volatilization	     179
                     Dissolution	     180

            7.2.2.2  Transport of/with Mobile Phase	     180

     7.2.3  Fixation	     188

            7.2.3.1  Partitioning onto Immobile
                       (Stationary) Phase	     188

                     Abandonment	     188
                     Adsorption	     188
                     Dissolution	     189
                     Diffusion	     189

            7.2.3.2  Other Fixation Approaches	     189

     7.2.4  Transformation	     189

            7.2.4.1  Biodegradation	     189
            7.2.4.2  Chemical Oxidation	     189

7.3  Storage Capacity in Locus	     189

     7.3.1  Introduction and Basic Equations	     189
     7.3.2  Guidance on Inputs for, and Calculation of,
              Maximum Value	     191
     7.3.3  Guidance on Inputs for, and Calculation of,
              Average Value	     191

                                    173

-------
7.4  Example Calculations....

     7.4.1  Storage Capacity

            7.4.1.1  Maximum
            7.4.1.2  Average

     7.4.2  Transport Rate Ca

7.5  Summary of Relative Impo
     7.5.1  Remediation....
     7.5.2  Loci Interaction.
     7.5.3  Information Gaps.
7.6  Literature Cited.
7.7  Other References.
7-1  Physicochemical Properti
     of a Synthetic Gasoline

7-2  Physicochemical Properti
     of an "Aged" Synthetic G

7-3  Average Grain Size and F
     Geologic Materials

7-4  Examples of Product Reco

7-5  Total Porosities for Com
     Sediments

7-6  Loci Interactions with L
     Groundwater
                                (Continued)
alculations.
torage Capacity.
torage Capacity.
culations	

tance of Locus.
                                  TABLES
s of Selected Chemical Constituents
t 20°C

s of Selected Chemical Constituents
soline at 20°C

nicular Zone Thickness for Various
ery after Tank Leaks

on Consolidated and Unconsolidated


quid Contaminant Floating on
Page No.

  192

  192

  192
  192

  192

  193

  193
  193
  194

  194
  195
  181


  183


  185

  186


  190


  193
                                    174

-------
                                  FIGURES
                                                                   Page No.

7-1  Schematic Cross-Sectional Diagram of Locus No. 7 - Liquid
     Contaminants Floating on the Groundwater Table                  177

7-2  Schematic Representation of Important Transformation and
     Transport Processes Affecting Other Loci                        178

7-3  Schematic of Immiscible Liquid Contaminant Accumulating
     Upon a Deflected Water Table                                    179

7-4  Schematic of the Capillary Zone With a Typical Water
     Saturation Curve                                                185

7-5  Mobilization of Contaminants in Locus 7                         187
                                    175

-------
                          SECTION 7 - LOCUS NO. 7
7.1  Locus Definition
7.1.1  Short Definition

    Liquid contaminants floating upon the water tables.
7.1.2  Expanded Definition am
 Comments
    The leaked liquid contaminant in this locus floats upon the water table
and is generally less dense than water; in some circumstances, liquids
which are denser than water may also float upon the water table.  If the
liquid contaminant has been released in sufficient quantity, it infiltrates
downward through the vadose zone (unsaturated zone), forming a continuous
pool in the vicinity of the leak.  The weight of the accumulating liquid
contaminant deflects the water table downward.  At the fringes of the pool,
the liquid contaminant may forn viscous stringers as it migrates into zones
of low interfacial tension 01 high hydraulic conductivity.  The liquid
contaminant spreads laterally under its own weight away from the leak
source, or it may move downgralient atop a sloping water  table.
    While aspects of this locu
6, the main feature is the exi
immiscible liquid upon the wat
contaminant floating upon peic
of locus 7 and Figure 7-2 repr
s  are  obviously  close to those of loci 5 and
;tence of  a floating layer of relatively
^i  table.   This  locus also includes liquid
led  water  tables.   Figure 7-1 is a schematic
=sents the transformation and transport
processes affecting other loci

7.2  Evaluation of Criteria for Remediation
7.2.1  Introduction
    Liquid contaminant is pres
occurred in the subsurface 01
liquid contaminant loses mass
dissolution into soil water.
in the voids of the vadose zon
quantity of liquid contaminant
;nt  in locus 7 because of a leak which
ipon the land surface.  The infiltrating
   volatilization into soil air and by
Jquid contaminant is also lost to retention
?  by capillarity.  If the leak has released a
 larger than losses to volatilization,
dissolution, and capillary retention,  the liquid contaminant accumulates
upon the water  table.  For a CM
and unconfined porous medium, a mound of liquid contaminant may  form, which
deflects  the water  table under
    The remainder of  this  Sect
affecting  the mobilization,  fij
contaminants in  locus  7.   This
contaminant storage capacity o
sample calculations.
 ntinuous leak in a homogeneous,  isotropic,
 its weight (see Figure 7-3).

 on describes in some detail the factors
nation,  and transformation of  liquid
 is followed by an assessment  of the liquid
  locus  7,  guidance for the assessment, and
                                     176

-------
    UNSATURATED
        ZONE
WATER
TABLE
      SATURATED
        ZONE
     NOTE: NOT ALL PHASE BOUNDARIES ARE SHOWN.
                  CONTAMINANT VAPORS

                  AIR (PLAIN WHITE AREAS IN
                  UNSATURATED ZONE)

                  WATER (MOBILE PORE WATER
                  IN UNSATURATED ZONE)
           AAAA  WATER FILM
_^3 WATER IN SATURATED ZONE

X   BIOTA (MICROBIOTA INCLUDED)

    SOIL PARTICLE

|   COLLOIDAL PARTICLE

    LIQUID CONTAMINANTS
  Figure 7-1. Schematic Cross-Sectional Diagram of Locus No. 7 -
             Liquid Contaminants Floating on the Groundwater Table
                                 177

-------
                                    L
                               1 -CONT
                         VOLATILIZATION
     (ROCK)
    LOCUS NO
10 - DIFFUSED IN
   MINERAL GRAINS
   OR ROCKS
DIFFUSION
               DEPURATION
         (BIOTA)
        LOCUS NO
    11 - SOR3ED TO BIOTA
       -#
     UPTAKE     \^^^
      PARTITIONING MAY NOT BE REQUIRED FOR
      LIQUID CONTAMINANTS TO MOVE TO LOCUS
      NO 2. RATHER THE UP AND DOWN MOVEMEW
      OF WATER TABLE WILL LEAVE LIQUID CONTA
      BEHIND IN LOCUS NO. 2
           Figure 7-2. Schematic Re
                       and Transport
                                      3IL GAS)
                                       US NO
                      /1INANT VAPORS
                          CONDENSATION
                                      a NO. 7
                                      flNANTS FLOATING
                                      WATER TABLE)
                                                       DISSOLUTION
                                                        PHASE
                                                        SEPARATION
                                                           (WATER)
                                                          LOCUS NO.
                                                    3 - DISSOLVED IN WATER FILM
                                                    8- DISSOLVED IN GROUND-
                                                      WATER
                                                   12 - DISSOLVED IN MOBILE
                                                      PORE WATER
(UOUIOC
LO
2- ADHE
DRY-
5- IN POP
SATU
6- INPOI
UNSA
13- IN OCX
3NTAMINANTS)
:us NO.
IING TO "WATER-
OIL PARTICLES-
IE SPACES IN
ATED ZONE
*E SPACES IN
VIRATED ZONE
:K FRACTURES
                      NANTS
                      resentation of Important Transformation
                      Processes Affecting Other Loci.
                                         178

-------
                                       \ \%%    %  •.   ••.       V. <•  S
                                       « V-      v\  -v  ••  <*• ' v vC- •.
I
s
.2
a
•§
B
LU
                               Infiltrating Hydrocarbon Liquid
                      Flgura 7-3. Schematic of Immlaclbla Liquid Contaminant
                               Accumulating upon a Deflacted Watar Table.
7.2.2  Mobilization/Remobilization

7.2.2.1  Partitioning into Mobile Phase

    Locus  7  consists  of free (mobile) liquid contaminants  floating upon the
water  table.   Contaminants in the free phase of  this  locus are present
because of  their  relative immiscibility with water and  high mobility in the
subsurface.   Historically, remedial efforts (e.g., free-product  pumping)
exploited  these properties of the contaminants;  alternative corrective
action may  involve removal of contaminants via other  loci.  Figure 7-2
presents these loci which may be affected by the liquid contaminants in
locus  7 and  the pathways or partitioning processes whereby inter-loci
transfer occurs.   A discussion of the factors affecting the movement and
transfer of  liquid contaminants in locus 7 follows.

Volatilization

    Volatilization is the primary direct mechanism by which partitioning of
liquid contaminant from locus 7 to locus 1 occurs.  Once contaminants have
moved  into  the soil vapoi phase of locus 1, the  volatilized liquid
contaminants move through the vado.se zone by diffusion  and advection (see
Section 1.2.2) where  they may be treated or removed from the subsurface by
various corrective actions such as vacuum extraction.
                                     17'-)

-------
Dissolution

    Dissolution is the primary
liquid contaminant from locu
which contaminant transfer-
is limited, in part, by the
    Solubility of additive-fr
mg/L in water at 20°C.  Solubil
a hypothetical, 23-component
mole-fraction-weighted averag
about 130 mg/L.
                      direct mechanism by which partitioning of
                     7 to loci  3,  8,  and 12 occurs.   The extent to
                  occurs by the dissolution of liquid contaminant
                   aqueous solubility of the liquid  contaminant.
                    ee gasoline ranges between about 130 and 250
                      ity values for the chemical constituents of
                    gasoline are reported in Table 7-1;  the
                      solubility of the synthetic gasoline is
    The addition of hydrophilic
methyl-tertiary-butylether in
result in gasoline solubiliti
Site-specific factors such as
salinity, particulate matter
in the aqueous phase also aff
                       compounds such as methanol,  ethanol, and
                     quantities of as much as 10% by volume may
                     s of greater than 2000 mg/L (Lyman, 1987).
                     ambient temperature and pressure, pH,
                     ontent, and concentrations of other compounds
                    2ct dissolution.
    Dissolution reduces the mkss of liquid contaminant in locus 7.  The
rate of dissolution is affected by the concentration gradient across the
liquid contaminant-water phas ? interface:  the higher the gradient, the
greater the diffusive flux ac
the process and mathematics o
and diffusion/dispersion of tie dissolved contaminants in the aqueous phase
of locus
8.2.2.2.
7 affects this conce
    Dissolution of the liquid
infiltrating rainwater and th
surface.  The rise ot the wat
                     oss the interface.   Section 1.2.2.2 describes
                      diffusion in greater detail.   The advection
itiation  gradient  and  is  discussed  in Section
                     contaminant is facilitated by contact with
                      attendant fluctuations in the water table
                     r table covered with liquid contaminants will
probably increase the dissolution of liquid contaminant into the water of
the vadose zone, the fate of fhich is described in Sections 3.2.2.2 and
12.2.2.2.

7.2.2.2  Transport of/with Mobile Phase

    Contaminants present in locus 7 are, by definition, mobile.  The
transport of liquid contaminar
Law (see Section 6.2.2.2) wit!
appropriately scaled  to reflec
contaminant.

    The saturated hydraulic
be scaled by the ratio of kin
liquid contaminant to obtain i
the liquid  contaminant:
                     ts within this locus is governed by Darcy's
                      the fluid conductivity of the porous medium
                     t the kinematic viscosity of the liquid
                   conductivity of a porous medium for water may
                    efmatic viscosity of the water to that of the
                      fluid conductivity of the porous medium for
                                  K   (u /u  )
                                   W   W  0
                                                             (7.1)
                                     180

-------
ri
                   oooooooooooooooooooo   oo   o
                   oooooocoocc>oocococ>
                   88888888S8888888SR88S88

•S-a

2
a-
                                                                                     y

                                                                                      >«
                                                                                     ac tJ
                                                                                     S-S
                                                                          H n  n

                                                                          rt J3  o
                                                                                "  "
                                             181

-------
where:

    K
     o
    K  =
V  =
 o

 w
         fluid  conductivity

         (cm/sec)

         saturated  hydraulic

         (cm/sec)

         kinematic  viscosity

         kinematic  viscosity
                        of a porous medium to a liquid contaminant
                         conductivity of a porous medium to water
                             of liquid contaminant (cm /sec)
                                         2
                             of water (cm /sec)
    Values of saturated hydraul-i
12-2, and for aquifer material
    Kinematic viscosity, the
liquid density, affects the
through a porous medium.  Th
the more resistance to flow,
definition, low-density fluijds
kinematic viscosity and flow
discussion of the effects of
contaminant flow and Darcy's
                               .i c conductivity for soils reported in Table
                              s in Table 8-1,  may be thus scaled.
                             ratio of dynamic viscosity of a liquid to its
                            tate at which the liquid contaminant may flow
                            =• higher the kinematic viscosity of a liquid,
                             resulting in slower flow.   Because of its
                               of high dynamic viscosity have a high
                             less readily through the porous medium.  A
                             dynamic viscosity and liquid density on
                             Law appears in Section 6.2.2.2.
    Kinematic viscosity valu
gasoline are reported in Tab
viscosity of water at 20°C i
of 0.00721 cm /sec.  Pure li
through the same porous medi

    In fact, most of the che
viscosities less than that o
in a pure liquid state than
kinematic viscosity of the s
                             s for selected constituents of a synthetic
                             e 7-1.   As a basis for comparison, the kinematic
                              0.01 cm /sec; benzene has a kinematic viscosity
                            jiiid benzene flows about 40% faster than water
                            im,  all  othei  factors remaining equal.
                            deals reported in Table 7-1 possess kinematic
                              watei,  indicating that they flow more readily
                            rater.  The mole-fraction-weighted average
                            mthetic gasoline is 6.36 x 10   cm /sec.
the vadose and  the saturated
    With time, the liquid contaminant loses its more volatile constituents
to the air of the vadose zone and more soluble constituents to the water of
                             zone.  The remaining liquid contaminant
becomes more dense and viscov
     In  the  vadose  zone  above-
 liquid  contaminant  to migiate
                             s.  This "aging" of the liquid contaminant by
losses  to volatilization and dissolution  increases  the kinematic  viscosity
of  the  liquid  contaminant.  Ls a  result,  the  "aged"  liquid  contaminant
flows less  readily  through  tie porous medium.  To estimate  the  effects  of
volatilization and  dissolution upon  the kinematic viscosity of  the
synthetic gasoline,  the more volatile (vapor  pressure >  20  mm Hg) and
soluble (solubility >  10 mg/L) chemical constituents of  the synthetic
gasoline in Table 7-1  were  removed  (see Table 7-2);  the  resultant kinematic
viscosity is  1.34 x 10  cm /.sec,  indicating  that the "aged" gasoline would
flow about  half  as  fast as  the unweathered  synthetic gasoline.
 the  liquid  contaminant  reaches  the  water  table  and  accumulates,  it  begins
 to  spread  out  laterally  until
                             the watei table, gravity and capillarity cause
                              downward and away from the leak source.  As
                              gravitational and capillary forces balance.
                                     18?

-------
s
i-i 3 S
Jo w a?

> £§
0 £>/-Js
'XJ 4-J -rH QU
O -C^* Cy
3 § O E
a£|s
^O £j x^
•rj -H O
d w o
• H S "Sb
rS
£?
M rH 3
•^ ^9 ~|p
£t r-t ^s
t/J
S
rH* '£
S C<3
d-.
^If
ci 2 ^
O *j v

•M rH
-G. Q
cu ob t
rH -r-l <


tative Chemical
S

^J


in on r-- m
o^^f^
i— 1 i-H v£> O
O
en en oj
pb ti S W
f"*-» ^?N r^-
§§§!
o") O O g;
in in so < — i
O cxl SO
CD O O


so c^ rn ^H
SO CO t — 1 O
o o o o

§rsl so r-i
' — I O so
o o o o

c^ ct^ ~*^ r^-
cn i — . rn ON
so rn o co
rH rH SO


CO SO CN CN
CN CN CN CN
^ CO CN ^5
rH CN •>* CM
i — It — 1 rH rH

91
imethylhexane
Tet rame thylhexa
phatic)
CU >-i |.,H
C H iTl r- j
3 4 in !_

O CN CN •-
C CN c-T H



CN
I'
rH
CN
2
rH
^ 1
o


co
^ 1
6 '

Q
1 O
rH

8
!8
t— i


, B

cu
c§
ct ion-weigh ted ,
ns:
8 2
U— ( |TT'
QJ E
I~H E
§ cTi





















cV5
cQ
4->
nJ
^d
o so
£ 4^
ubility values
sity reported a
Available.
Calculable.
rH C *J 4J
&&'%'&.
II II II II

<* -° £%
                                          i
                                           s
                                          O *-'
                                         11
                                           01
                                           M
                                           ™
                                          r- rt
                               183

-------
    Hantush (1967) presents a| i
and decay of the groundwater
technique may also describe
upon the water table for a cont
                      analytical solution that predicts the rise
                    table in response to deep percolation.  This
                       growth of a mound of liquid contaminant
                      inuous leak of liquid contaminant:
                   the
    h(r) = [(Q/(2-n-K))-W(r

for distances larger than the
    h(r)
    Q
    n
    K


    r
    S
    t
    b
                    S/(4-K-b-t))]

                     radius of the leak, where:


= liquid contaminant thickness upon the water table (m)
                                                3
  leak rate of liquip contaminant from source (m /day)

  pi (3.14159...)
                                                             (7.2)
= fluid conductivity
  contaminant (m/day

  radial distance from leak (m)
= specific yield of
= leak duration (day:;)
= average thickness
         = exponential integia] function
                     of porous medium to a liquid
                                    3, 2
                     orous medium (m /m /m)
                     >f liquid contaminant (m)
Once a guess for b is made, a
thickness, h(r), can be calcu
Values of h(r) are re calcula
    The use of equation 7.2 inquires an initial guess at  the value of  the
average thickness of liquid contaminant, b, which is not  known a  priori.
                     range of values of liquid contaminant
                     ated, from which b may be re-estimated.
                     ed in an iterative procedure until the
difference in average  thickness between any  two successive  iterations  is
negligible.  This  technique may be used to estimate  the areal extent and
distribution of liquid contaminant upon the  water  table.

    The  thickness  of liquid contaminant accumulating upon  the water  table
depends  on the rate at which  the liquid contaminant  leaks  from  the source,
the intrinsic permeability of  the porous medium, and the kinematic
viscosity of the liquid contarr inant.  Capillary forces may  cause  liquid
contaminant to form a  layer upon the water table which varies in  thickness
depending on the geologic material.  Fine-grained  materials exhibit  more
pronounced capillarity than coarse-grained materials.  A zone of  capillary
tension  and partial saturation  forms above the capillary fringe which  has
been  termed the funicular zone  (see Figure 7-4).

    Within the funicular zone,  liquid contaminant  may be trapped  by
capillarity;  the  stronger  the
                     capillary force,  the  thicker  the entrapped
layer.  Dietz  (1971)  states  that  oil  spreads  over  a  water  table  in  a  layer
roughly equivalent  to the  thickness of  the  funicular zone.

Table  7-3  lists average  funirujlar zone  thicknesses tor  various geologic
materials.
                                     1H4

-------
                      UNSATURATED ZONE
                                      Capillary rise above phreatic level
                                      in slimmest continuous pores
                          V
                               FUNICULAR ZONE
                                       Capillary rise above phreatic level
                                       in widest continuous pores
                                          CAPILLARY FRINGE
Phreatic level, as found
in observation wells
                                          SATURATED ZONE
                                100%
                                Water Saturation
               Source: Adapted from Deitz, 1971

Figure 7-4. Schematic of the Capillary Zone with a Typical Water Saturation Curve
                                   TABLE 7-3

               AVERAGE  GRAIN SIZE AND FUNICULAR ZONE THICKNESS
                     FOR  VARIOUS GEOLOGIC MATERIALS
         Material
 Very coarse sand
 Moderately coarse sand
 Moderately fine sand
 Very fine sand
 Silt
 Clay
;rage Grain Size (mm)
0.5
0 . 2
0.05 -
0.015 -
0.004 -
<0.004*
2.0
0.5
0.2
0.05
0.015*

Funicular Zone
Thickness (cm)
1.8
9.0
22.4
28.1
45
120
- 9.0
- 22.4
- 28.1
- 45
- 90
- 240
 Source:  Driscoll,  1986.
                                      1H1

-------
    Figure 7-5 depicts the ef
liquid contaminant upon the w<
Initially, liquid contaminant
in the liquid contaminant poo
7-5b).  Pumping of the oil cav
at the extraction well; the ur
up-coning at the extraction w
sufficient rate, more and mor«
oil.  Eventually, the cone of
may intersect; because of the
presence of water, no further
treating contaminated water ar
removing as much oil as possi

    A similar situation occurs
withdrawn and becomes gradual
caused by pumping watei cause
the extraction well.  Suffici<
the screened interval, reduci
Figure 7-5d of two extraction
causes a cone of depression t<
thickness which is then more

    Numerous remediation atte
success.  Remediation of many
been performed with free prod
a few gallons to tens of thou
Larger leaks have occurred at
presents recovery histories o
               EXAMPLES OF PR

               	 	Type of

                   Abandoned
    Location	

West Coast, USA

South Korea
East Coast, USA
North Coast,
Venezuela

Eastern USA

Southwest USA

New Zealand


Midwestern USA
                   Fuel loadi
                   UST/load  r
                   Pipeline/b


                   Bulk stora

                   Refinery/b

                   Refinery/b


                   Retail  fue
Source:  Yaniga,  1984a and  19
                               cts of various remedial pumping schemes on
                               er table in locus no. 7 (see Figure 7-5a).
                               s withdrawn from an extraction well screened
                                which is assumed to be oil (see Figure
                               es a cone of depression to form in the soil
                               erlying water also responds to the pumping,
                               1.  If the extraction well is pumped at a
                               water is withdrawn from the well instead of
                               epression in the oil and the up-coning water
                               ov/ relative permeability of oil in the
                               >il is removed.  Apart from the difficulty of
                                oil simultaneously, the objective of
                               e from locus 7 by pumping is not met.

                               in Figure 7-5c where water is initially
                                replaced with oil.  The cone of depression
                               the oil to accumulate in a greater thickness at
                               it pumping of water may cause oil to intrude in
                              K water withdrawal.  The scenario depicted in
                               ells simultaneously pumping water and oil
                               form in the water, collecting oil in greater
                              fficiently extracted by the second well.

                                s have been conducted, some with considerable
                               .eaks at service stations with gasoline USTs has
                               t recovery.  These leaks are usually limited to
                              ands of gallons of leaked liquid contaminant.
                               ndustrial facilities and airports.  Table 7-4
                               some larger remediation efforts.

                                 TABLE 7-4

                              DUCT RECOVERY AFTER TANK LEAKS
Facility    _0ngoing Recoveries	Product Type
ulk <;toiage  500-1,000 gal/wk

  rark       1,000-1,500 gal/wk

:k           1,000-1,500 gal/wk

Ik storage   28,000 gal/day
 .k storage

Ik storage
500-700 gal/day

3,000-5,000 gal/day

1,500-2,000 gal/day
 facility    700 gal/month
Mixed fuels
Jet fuel/Avgas
Jet fuel
Gasoline

Fuel oil
Mixed fuels
Gasoline/
gas oil
Gasoline
                              4b

-------
     Impermeable Stratum
                                                Impermeable Stratum
A. Before Remediation
                                            B. With Oil Pumped Only
                         Water
                                                                    Water
Impermeable Stratum
                                                     Impermeable Stratum
C. With Water Pumped Only
                                            0. With Oil and Water Pumped
           Figure 7-5. Mobilization of Contaminants in Locus 7.
                                     187

-------
7.2.3  Fixation

    As noted in Section 7.2.2
high mobile liquid state.  Co
rendered less threatening if
considerations addressed in
1) enhancing phase exchange a
one of the loci that, presuma
state; and, 2) fixing the loc
or as a permanent corrective

7.2.3.1  Partitioning onto Im

    The processes of abandonm
contaminants to move from loc

Abandonment

    The ultimate direction of
site-dependent.  A falling w,
water-dry soil particles at r
coating a water-dry soil part
conditions, however, the cont
vapor phase (locus 1) in the
mass transport mechanisms, as
infiltrating water may disp]a
until it flows downward undei
Locus 2 is then displaced dow
front, or is dissolved in the
locus 8 or remains relatively
from locus 7 into locus 3, by
mobile because of their assoc
held in place by capillary te
locus 3 can be, however:, easi
saturated by either  infiltrat
volatilization.  Contaminants
a rising water table, which
main body of liquid contamina
immobile and are held in the

    These contaminants can be
Section 7.2.2.1) or mechanica
effect is an increase in pote
in groundwater).  Should the
reabsorbed in locus 7.  Conta
locus 6 by the rising and low
partially fills  the void spac
liquid contaminant is termed
water-dry and water-wet  soil

Adsorption

    Contaminants  in  locus 7 i
particles or biological  matt
 contaminants in locus 7 are in a relatively
taminants present in locus 7 could be
ixed into a less mobile phase.  The
e following subsections are directed toward:
  bulk transport out of the locus and into
ly,  hold the contaminants in a less mobile
  as a whole, for possible future remediation
ction.

obile (Stationary) Phase

nt,  dissolution, and adsorption may cause
s 7  to comparatively less mobile phases.
partitioning between loci 2 and 7 is
er table may leave a contaminant film over
sidual saturation.  The contaminant film
 le is relatively immobile.  Under these
minant film may transform to the more mobile
nsaturated zone and migrate by diffusion or
detailed in Section 1.  At an unpaved site,
e the contaminant film which may accumulate
greater-than-residual saturation conditions.
ward to locus 7 ahead of the advancing water
advancing water front and becomes mobile in
immobile in locus 3.  Contaminants moving
the mechanism described above, will be less
ation with the unsaturated pore space water
si on.  The contaminants transferred into
y remobilized when the soil pore spaces are
rig water, a rising groundwater table, or
may be transferred from locus 7 to locus 5 by
olates blobs of liquid contaminant from the
t.  Once in locus 5, the contaminants are
nterstitial spaces by capillary tension.

remobilized by dissolution (as described in
 dispersion of minute droplets.  The overall
tial partitioning to locus 8 (mobile solute
ater table again fall, locus 5 could be
inants can be transferred from locus 7 to
ring water table, so that liquid contaminant
s of the unsaturated zone.  This remaining
esidual saturation and may be different for
articles.
y come  into direct  contact with water-dry  soil
   As a  result of contact, contaminants  may
                                     urn

-------
either adsorb (mono-molecuiai layer film) or adhere (multi-molecular layer
film) to particles in these loci and be less mobile than when in locus 7.
For detailed discussions of these processes, see Sections 2, 4, and 9.

Dissolution (Detailed in Section 7.2.2)

    Dissolution of contaminants from locus 7 in the relatively immobile
water film surrounding soil particles in the unsaturated zone (locus 3)
could take place.  As discussed above, however, there is great potential
for contaminants to be remohilized from this temporary sink.

Diffusion (Details in Section 10.2.3)

    Contaminant species may diffuse from locus 7 into the water-dry surface
of a mineral particle (locus 10).  Although contaminants in locus 10 are
believed to be immobile, the total impact of this process on reducing the
mobility of contaminants in locus 7 and its impact on remediation of the
locus is exceeding small.

7.2.3.2  Other Fixation Approaches

    Hydrodynamic control (groundwater pumping or draining) and grout
curtains are control options in some geologic settings that may effectively
halt mass migration of locus 7  to a new physical location.  There are,
however, no other fixation approaches that could conceivably prevent
migration of contaminants i iom  thoii  nearly pure state in locus 7 to some
other locus.

7.2.4  Transformation

7.2.4.1  Biodegradation (Details in Section 11)

    Bulk liquid hydrocarbons are much less subject to biodegradation than
organic contaminants dissolved  in water.  This is because of the limitation
of oxygen in organic liquids which is requisite for aerobic biodegradation.
Microorganisms which can survive in,  and cause "spoilage" of, organic
liquids are known.  The rate of reaction is low, however, due to
limitations of oxygen 01 othn  nutrients.  It is likely that contaminants
in locus 7 would disperse and diffuse to other loci, where they may degrade
more quickly, before they degraded as a bulk liquid in locus 7.

7.2.4.2  Chemical Oxidations (Details in Section 3.2.4.2)

    Contaminants floating on the water surface can undergo abiotic
transformation as the constituent comes into contact with oxidized metal
complexes.

7.3  Storage Capacity in Locus

7.3.1  Introduction and Basic Equations

    The storage capacity of locus 7 can be described by the total porosity
(Driscoll, 1986), which is defined as:

-------
where the difference between
solid is filled by geologic
total porosity for a variety
                         Volume of total pore space
                           Volume of bulk solid
                                          (7.3)
 the pore space volume and that of the bulk
material.  Table 7-5 presents ranges of percent
 of unconsolidated and consolidated sediments.
                                 TABLE 7-5
                 TOTAL POROSITIES FOR COMMON CONSOLIDATED
                       AND UNCONSOLIDATED SEDIMENTS
Unconsolidated
Sediments
Clay
Silt
Sand
Gravel
Sand & gravel
Glacial till

Po
(P
4
3
2
2
11
1(

rotal
-osity
ncent)
3 55
3-50
3-40
5-40
)-35
)-?!>

Consolidated
Sediments
Sandstone
Limestone
Shale
Fractured
crystalline rock
Vesicular basalt
Dense, solid rock
Total
Porosity
(percent)
5 30
1-20
0-10

0-10
10-50
<1
         Source: Drisroll, 1(
    Those porosity values me;
reported in Table 7-5 becaus
reducing the volume availabl
space-occupying factors are:

    1)  Entrapped water;
    2)  Biomass;
    3)  Entrapped .soil gas.

    The volumes of the factoi
 86.
 sured in nature may be lower than those values
  additional phases may occupy void volume,
 > for contaminant storage.  These other
decrease in response  to nutri
   listed above are site specific and may vary
with time.  Entrapped wnter nay b<-' displaced, biomass may  increase  or
 ent levels, and entrapped soil gas may be
displaced or absorbed  into  the contaminant or aqueous  phases.   Several  of
these factors are briefly dis
poros i ty-reduc t ion.
    The quantity of entrappec
and uniformity of  the grains.
by capillarity, the quantity
to the wilting point of  the
Section 12.2.2.2.  Figure  12-
capacity and wilting point of
entrapped water may be estima
for contaminant  liquid.  Air
 cussed below as they relate to
  water in a porous medium depends on the size
   Presuming that the entrapped water is held
 of water held ranges from the field capacity
pjorous medium.  These terms are discussed in
 4 gives ranges of values for the field
  various soils with which the quantity of
 ted.
    Entrapped soil gas may  reducf  the  storage  capacity  of  the  porous  medium
 dissolution in a liquid contaminant, however,

-------
may be relatively rapid, making this consideration relatively unimportant
in calculating the storage capacity of locus 7.

    As a final note on factors which may offset the storage capacity,
changes in the crystalline structure of clays due to incorporation of
organic compounds have been reported.  This may be a significant factor in
calculating storage and transmission of liquid contaminants in some
geologic settings.  Certain clays also adsorb water, swelling and reducing
void volume.  The increased percent volume in clay due to structural
changes may be as high as 400%.

    Under these concepts, the storage capacity of locus 7 is:


                      mc = ecpc =  f6t * 9a ' 6w - 9^ p-              (7<3)
where:

      m   = storage capacity for contaminant (kg/L)

      6   = total porosity (dim.)

      6   = contaminant storage capacity (dim.)

      6   = entrapped soil gas fraction (dim.)
       3.
      6   = entrapped water fraction (dim.)
       W                       33
      9,   = biomass fraction (m /m  dim.)

      p   = density of contaminants (g/cm  or kg/L)


7.3.2  Guidance on Impacts for, and Calculations of, Maximum Value

    For the calculation of a maximum value for the contaminant storage with
equation 7.3, the following assumptions were made:

    1)  no entrapped soil gas, 0=0;
                                3.

    2)  entrapped water is at wilting point;

    3)  no biomass, 9,  = 0.

    As stated above, Figure 12-4 gives the fraction of water held in a soil
at the wilting point.  Under these assumptions, equation 7.3 reduces to:
7.3.3  Guidance on Impacts for, and Calculations of, Average Values

    To calculate an average value for contaminant storage capacity, some of
the assumptions in the previous section will be altered.  For lack of
better information, it is assumed that the volume of biomass is negligible
(9,  = 0), and that the volume of entrapped air is also small (9  =0).
Instead of relying upon the wilting point of a soil for an estimation of
entrapped water volume, the field capacity of the soil will be used.

                                    191

-------
7.4  Example Calculations

7.4.1  Storage Capacity Ca

    Assume gasoline in a s<
maximum total porosity for
porosity for sand from Tab
g/cin  (Table 1-1A).

7.4.1.1  Maximum storage c
                           culation

                            d formation for these calculations.  The
                            and from Table 7-5 is 0.40.  The average
                           e 7-5 is 0.33.  The density of gasoline is 0.74
    Assuming 0  = 6, = 0 a
12-4), and using equation
                            >acity:
                             6  = wilting point, which is 0.05 (Figure
                            3:
    m
         (0.4 - 0.05) 0.74
     *• = 0.26 g/cni

7.4.1.2  Average storage c

    Assuming 6  = 6,  = 0 a
              cl    D

    m  = (0.33 - 0.09) 0.7
       = 0.18 g/cni

7.4.2  Transport Rate Calc

    Based on flow equation
simple calculation of liqu
Returning to the sandy soi
for contaminant storage ra
kinematic viscosity of the
cm
       ^. For andean sand
from 10   to 10  cm/sec (F
cm/sec is chosen for the h
saturated hydraulic conduc
fluid conductivity of the
K  =
               V
                     K
                      w
where
         \)  = 0.01 cm /sec
         v  = 0.00636 cm2/
          o
The scaled fluid conductiv

    Assuming a hydraulic g
liquid contaminant (synthe
                           pacity:

                           d GW = field capacity of 0.09 (Figure 12-4):
                            .at ions
                        presented in Sections 2, 5, 6, 8 and 12, a
                       d contaminant flow velocity is presented.
                        of the previous section, the porosity available
                       ges from 0.25 to 0.40.  From Table 7-1, the
                       unweathered liquid contaminant is 0.00636
                        the saturated hydraulic conductivity may range
                       eeze and Cherry, 1979); a medium value of 10~
                       draulic conductivity.  With equation 7.1, the
                        vity of the sand for water may be scaled to a
                       and for the synthetic gasoline:
                           for water
                           ec for synthetic gasoline

                                          _2
                           ty is 1.57 x 10   cm/sec.

                           idient of 0.001 cm/cm, the flow velocity of  the
                           Lc gasoline) is:
                                                                       (7.5)
                                    192

-------
where
         dh/dl = hydraulic gradient (cm/cm)
    From equation 7.5, the estimated flow velocity of ,-synthetic gasoline in
the sand ranges from 3.93 x 10   cm/sec to 6.33 x 10   cm/sec.

7.5  Summary of Relative Importance of Locus

7.5.1  Remediation

    The remediation of contaminants in locus 7 is essential because
contaminants in this locus serve as a continual source of contamination for
the vadose zone and the saturated zone.  Remediation of other loci may be
necessary as well to remove the health and/or environmental threat posed by
the contaminants.                *"

    Remediation oi contaminants in this locus has been by the direct
removal of liquid contaminants by active pumping.  In the relatively recent
past, other corrective actions have been utilized in conjunction with
active removal.  While there is no known evidence of any other corrective
action being as effective or efficient as direct liquid removal, other
alternatives, such as enhanced biodegradation, active soil venting (vacuum
extraction, steam stripping, etc.), and in situ treatment should be
evaluated before a corrective action is selected for a given site.

7.5.2  Loci Interaction

    The fact that locus 7 consist^ of hulk liquid contaminant and is mobile
in its bulk state in thiee dimension.', moans that the contaminants in this
locus can, ultimately, affect nil other loci.  Table 7-6 summarizes the
partitioning and iclative importance of that partitioning between locus 7
and the other loci.

                                 TABLE 76

                 LOCI INTERACTIONS WITH LIQUID CONTAMINANT
                          FLOATING ON GROUNDWATER
Interacting
Locus
1
2
3
5
6
8
10
11
12
13
Phase
Contacted
air
dry solid
water
watei k solid
solid
water
so] id
biota
wa t e i
voids
Transfer
Process
volatilization
adhering
dissolution
adhering
adhering
dissolution
adsorption
sorption
dissolution
bulk transfer-
Relative
Importance
high
high
moderate
moderate
high
high
low
moderate
moderate
high

-------
    As is evident from the t
no's. 1, 2, 6, 8, and 13.  T
contaminant to loci 1, 2, 6
Of these, the largest mass o
loci 1 and 13, with lesser p
potential for partitioning t
the large potential surface
the duration of contact.

    Loci 3, 5, 11, and 12 ar
partitioning from locus 7, b
those mentioned above becaus
transferred there.  Locus 10
from locus 7, but appears to

7.5.3  Information Gaps

    There is a considerable
data generated that is relev
information is, however, sit
information gaps  that, if fi
behavior of contaminants in

    (1)  Solubility data of
         groundwater  conditi

    (2)  Data on  the  effects
         parameters (e.g., s

    (3)  Data on  physical an
         mixtures;

    (4)  Data on  hydraulic p

    (5)  Areal distribution

7.6  Literature Cited

Dietz, D.N.   1971.  Pollutio
    Water Pollution by Oil.

Driscoll, F.G.   1986.  Gi oui
    Paul, MN.

Freeze,  R.A.  and  J.A. Cherry
    Cliffs, NJ.

Lyman, W.J.   1987.  Environn
     Soil/Groundwater  Compart
    Underground  Environment
    N.J.
•>le',  the loci most affected by locus 7 are
e ability of locus 7 to transfer pure
id 13 make it a critical distribution locus.
 contaminants will probably be transferred to
tentinl partitioning to loci 2 and 6.  The
 locus 8 is also relatively high because of
rea of contact at the bottom of locus 7 and
 also potential receptors of contaminants
t are of considerable less importance than
 of the mass of contaminants that could be
has high potential to receive contaminants
be a relatively small-volume sink.
mount of research that has been completed and
nt to the study of locus 7.  Most of this
 specific.  The following is a brief list of
Ird, would improve the understanding of the
ocus 7:

ndividual contaminants under various
ns;

of gasoline additives on fundamental
lubility, activity, volatility);

 chemical characteristics of "weathered"
operties of  the site; and

f the liquid contaminant in  the subsurface.
 of Permeable Strata by Oil Components.   In;
Institute of Petroleum, London,  England.

water and Wells, Johnson Well  Division,  St.
  1979.  Groundwater.   Prentice-Hall,  Inglewood
ntal  Partitioning  of  Gasoline  in
ents.   In:   Final  Seminar  Proceedings  of  the
f an  UST~Motor  Fuel Release, U.S.  EPA,  Edison,

-------
Yaniga, P.M.  1984a.  "Hydrocarbon Contamination of Groundwater:
    Assessment and Abatement," testimony presented to U.S. Senate Committee
    on Environmental and Public Works and Subcommittee on Toxic Substances
    and Environmental Oversale.

Yaniga, P.M.  1984b.  Hydrocarbon Retrieval and Apparent Hydrocarbon
    Thickness:  Interrelationships to Recharging/Discharging Aquifer
    Conditions.  Proceedings of the NWWA/API Conference on Petroleum
    Hydrocarbons and Organic Chemicals in Groundwater, Houston, TX, pp.
    299-325.

7.7  Other References

Case Studies

Testa, S.M., D.M. Baker, and P.L. Avery.  1989.  Field Studies and
    Occurrence, Recoverability, and Mitigation Strategy of Free Product
    Liquid Hydrocarbon.  In:  Environmenta] Concerns in the Petroleum
    Industry.  American Association of Petroleum Geologists, pp. 57-81.

Lapham, S.M.  1986.  Advanced Stages of a Hydrocarbon Recovery/Closed Loop
    Hydrocarbon Flushing System Successful in a High Permeable Glacial
    Outwash in Apple Valley, MN.  Proceedings of the NWWA/API Conference on
    Petroleum Hydrocarbons and Oiganic Chemicals in Groundwater, Houston,
    TEX, pp. 755-769.

General Ground Water Theory

Driscoll, F.G.  1986.  Groundwater and Wells.  Johnson Well Division, St.
    Paul, MN.

Fetter, C.W., Jr.  1980.  Applied Hydrogeology.  C.E. Merrill Publishing
    Co., Columbus, OH.

Immiscible-phase Experimentation

Schiegg, H.O.  1988.  Physical Simulation of Three-phase Immiscible Fluid
    Displacement on Porous Media.  VAW No. 40(5), U.S. Department of
    Energy.

Schwille, F.  1988.  Dense Chlorinated Solvents in Porous And Fractured
    Media, Lewis Publishers, Chelsea, MI.

Immiscible-phase Simulation

Abriola, L.M. and G.F. Pindei.  1985.  A Multiphase Approach to Modeling
    Porous Media Contamination by Organic Compounds, 1.  Equation
    Development.  Water Resources Rpsearch, 21(1):11-18.

Abriola, L.M. and G.F. Pindei.  198C>.  A Multiphase Approach to Modeling
    Porous Media Contamination by Organic Compounds, 2. Numerical
    Simulation.  Water Resources Research, 21(1):19-26.
                                     4 j

-------
Aziz, K. and A. Settari.  1
    Science Publishers, Lon

Faust, C.R.  1985.  Transpo
    Unsaturated Zone - A Nu
    21:587-596.

Falta, R.W. and I. Javandel
    Multicomponent Contamin
    American Geophysical Un

Hochmuth, D.P. and D.K. Sun
    Immiscible Flow in Coar

Hunt, J.R. and N. Sitar.
    Cleanup, 1. Analysis of
    24(8):1247-1258.

Hunt, J.R., N. Si tar, and K
    Transport and Cleanup,
    Research, ?4(8):1247-12

Huyakorn, P.S. and G.F. Pin
    the Solution  of Two plia
    Resources, 1.

Kaluarachchi, J.J., J.C. Pa
    for Water and Light Hyd
    Vertical Equilibrium.

Kuppusamy, T., J, Sheng, J.
    Element Analysis of Mul
    Resources Research, 23:

Osbourne, M. and  J. Sykcs.
    Transport at  the Hyde P
    22(l):25-33.

Parker, J.C., J.J. Kaluarac
    of  Free  Product Recover
    Proceedings of the  NWWA
    Organic  Chemicals  in GL

Parker, J.C., R.J. Lenhard,
    for Constitutive Proper
    Water Resources Researc

Immiscible-phase  Flow  Theor

Cary, J.W.,  J.F.  McBridc, c
    Residuals  in  Capillary
    Environ. Qual.,  JH(1):7
 9.   Pen oleum Reservoir Simulation, Applied
on.

  of [mmiscible Fluids Within and Below the
erica!  Model.  Water Resources Research,
  1987.  A Numerical Method for Multiphase
it Transport in Groundwater Systems.  Trans.
on,  68(44).

da.   1985.  Ground-water Model of Two Phase
e Material.  Ground Water, 23(5):617-626.

88.   Nonaqueous Phase Liquid Transport and
Mechanisms.  Water Resources Research,
cl. Udell.  1988.  Nonaqueous Phase Liquid
.  Experimental Studies.  Water Resources
ei .   lc)78.  New Finite Element Technique for
  Flow through Porous Media.  Advances in Water
ker, and A.K. Katyal.  1988.  A Numerical Model
:>carbon Migration in Unconfined Aquifers under
dvanres in Water Resources, in review.

  Parker, and R.J. Lenhard.  1987.  Finite
iphase Immiscible Flow through Soils.  Water
25-631.

1986.  Numerical Modeling of Immiscible Organic
rk Landfill.  Water Resources Research,
chi, and A.K. Katyal.  1988.  Areal Simulation
 from a Gasoline Storage Tank Leak Site.
API Conference on Petroleum Hydrocarbons and
undv.it CM . Houston, TX, pp. 315-332.

and T. Kuppusamy.  1987.  A Parametric Model
ies Governing Multiphase Flow in Porous Media.
  23:618-624.
d C.r,. Simmons.  1989.   "Trichloroethylene
 inge as Affected by Air-entry  Pressures," J.
 77.
                                     196

-------
Parker, J.C. R.J. Lenhard, and T. Kuppu.samy.  1987.  Physics of Immiscible
    Flow in Porous Media.  EPA/600/2-87/101.

Stallman, R.W.  1964.  "Multiphase Fluid Flow in Porous Media - A Review of
    Theories Pertinent to Hydrological Studies," USGS Prof. Paper 411-E,
    Department of Energy.

Monitoring and Measurement Studies

Abdul, A.S., S.F. Kia, and T.L. Gibson.  1989.  "Limitations of Monitoring
    Wells for the Detection and Quantification of Petroleum Products in
    Soils and Aquifers,"  Ground Water Monitoring Rev., 9(2), 90-99.

Wilson, J.L., S.H. Conrad, E. Hagan, W.R. Mason, and W. Peplinski.  1988.
    "The Pore Level Spatial Distribution and Saturation of Organic Liquids
    in Porous Media, "Proceedings of Petroleum Hydrocarbons and Organic
    Chemicals in Groundwater, 107-134, National Water Well Association.

Remediation

Camp Dresser & Mc.Kee Inc.  1988.  Cleanup of Releases from Petroleum USTs:
    Selected Technologies.  EPA/530/UST-88/001.

Hunt, J.R., N. Sitar, and K.S. Udr-11.  1986.  "Organic Solvents and
    Petroleum Products in the Subsurface:  Transport and Cleanup," U.
    Cal-Berkely, UCB-SEERHL Report No. 86-11, San. Eng. and Env. Health.

Somers, J.A.  1973.  The Fate of Spilled Oil in the Soil.  Hydrological
    Sciences, 19(4):501 521.

Testa, S.M. and M.T. Paczkowski.  1989.  Volume Determination and
    Recoverability of Free Hydror.irbon.  Ground Water Monitoring Review,
    9(1):120-12R."

Yaniga, P.M.  1984.  Hydrocarbon Retrieval and Apparent Hydrocarbon
    Thickness:  Interrelationship.1-- to Recharging/Discharging Aquifer
    Conditions.  Proceedings of the NWWA/AP1 Conference on Petroleum
    Hydrocarbons and Organic Chemicals in Groundwater, Houston, TX, pp.
    299-325, National Water Well Association.

-------
                          SECTION 8.  LOCUS NO. 8

                                 CONTENTS
List of Tables	
List of Figures	
8.1  Locus Description.
     8.1.1  Short Definition.
     8.1.2  Expanded Definiti

8.2  Evaluation of Criteria f
     8.2.1  Introduction...,
     8.2.2  Mobilization/Remo

            8.2.2.1  Partitio
            8.2.2.2  Transpor

                     Transpor
                     Dispersi
                     Diffusio
                     Advectio
                     Retardat
     8.2.3  Fixation.
            8.2.3.1  Partitio
                       Phase.
            8.2.3.2  Other Fi
     8.2.4  Transformation.
            8.2.4.1  Biodegra
            8.2.4.2  Chemical

8.3  Storage Capacity in Locu

     8.3.1  Introduction and
     8.3.2  Guidance on Input
              Maximum Value
     8.3.3  Guidance on Input
              Average Values.
n and Comments.

r Remediation..
ilization.
ing onto Mobile Phase.
 with Mobile Phase....
 Rate.
n.
                              on.
ing onto Immobile (Stationary)
• ••••••••••••••••••••••••••we**
ation Approaches	
ation	
Oxidation.
asic Equations	
 for, and Calculation of,
 for, and Calculation of,
                                    198
Page No.

  199
  199
  201

  201
  201

  201

  201
  204

  204
  205

  208
  209
  209
  210
  213

  217
  217
  218

  218

  218
  218

  219

  219

  219

  221

-------
                                (Continued)


                                                                   Page No.

8.4  Example Calculations	     221

     8.4.1  Storage Capacity Calculations	     221

            8.4.1.1  Maximum Value	     221
            8.4.1.2  Average Value	     221
            8.4.1.3  Example of Solubility Enhancement	     221

     8.4.2  Transport Rate Calculations	     221

8.5  Summary of Relative Importance of Locus	     224
                                 *-,
     8.5.1  Remediation	     224
     8.5.2  Loci Interactions	     225
     8.5.3  Information Gaps	     225

8.6  Literature Cited	     226
                                  TABLES
8-1  Estimated Retardation Factors for Hydrocarbon Gasoline
     Constituents                                                    216

8-2  Water Soluble Fractions for Three Gasolines                     220

8-3  Estimated Solubility of Gasoline in Water                       222

8-4  Loci Interactions with Hydrocarbon Dissolved in Groundvater     226



                                  FIGURES
8-1  Schematic Cross-Sectional Diagram of Locus No. 8 -
     Contaminants Dissolved in Groundvater (i.e., Water in the
     Saturated Zone)                                                 202

8-2  Schematic Representation of Important Transformation and
     Transport Processes Affecting Other Loci                        203

8-3  Processes of Dispersion on a Microscopic Scale                  206

8-4  Concentration Distribution of Dissolved Components as a
     Result of Hydrodynamic Dispersion                               206
                                    199

-------
8-5  Approximate Temperature
     United States at Depths
8-6  Relations Among Hydraulic Conductivity, Uniformity
     Coefficient, and Median

8-7  Advance of Reactive and
     a Column
                                (Continued)
                        of Groundvater in the Conterminous
                        of 10 to 25 Meters
                        Grain Diameter
                        Nonreactive Contaminants through
8-8  Effect of the Distribution Coefficient on Contaminant
     Retardation During Transport in a Shallow Groundvater
     Flow System

8-9  Relationship of Retarda
     Partition Coefficient b;
     Equations

8-10 Normalized Breakthrough
                       :ion Factors to Octanol-Vater
                         Conventional and Revised
                        Curves for Displacement of
Benzene of 15-cm LincolJi Fine Sand Soil Column as a
Function of Water Velocity
Page No.


  210


  212


  213



  214



  217



  218
                                    200

-------
                          SECTION 8 - LOCUS NO. 8
8.1  Locus Description

8.1.1  Short Definition

    Contaminants dissolved in groundwater (i.e., water in the saturated
zone).

8.1.2  Expanded Definition and Comments

    Locus no. 8 consists of groundwater contaminated with dissolved
pollutants.  The dissolution process may have occurred in the unsaturated
zone (from locus nos. 2 or 6), at the groundwater table (from locus no. 7),
or from the saturated zone (locus no. 5).  The locus forms a continuous
phase in the saturated zone and is mobile.  This saturated flow may be in
the vertical or horizontal direction; the flow is more even than in the
unsaturated zone because of the phase continuity and a more consistent
water supply.  The flowing groundwater may transport the dissolved
contaminants (up to several kilometers) from the leak site.  No air is in
contact with locus no. 8 except at the surface of the groundwater table,
which, in combination with the fact that groundwater is far from being a
well-mixed system, essentially eliminates volatilization as a significant
loss mechanism.  Figures 8-1 and 8-2 present a schematic cross-sectional
diagram of locus no. 8 and a schematic representation of the transformation
and transport processes affecting other loci, respectively.

8.2  Evaluation of Criteria for Remediation

8.2.1  Introduction

    Contaminants dissolved in groundwater become part of a continuous,
mobile phase.  The contaminant's behavior is affected by the partitioning
characteristics, particularly the soil sorption partitioning, and by
transformation processes, mainly biodegradation.

    Remedial measures used for contaminants in this locus tend to fall into
three categories:  mobilization, immobilization, and transformation.
Mobilization techniques, such as pump and treat schemes and artificial
recharge,  seek to move the contamination to an extraction point where it is
captured and treated.  During these processes, the contaminants in the
groundwater become diluted in concentration.  Immobilization techniques
(e.g., slurry walls) attempt to limit the migration of contaminants by
reducing the mobility of a certain part of a groundwater flow regime.  This
technique alone does not treat the contaminants to dilute the
concentrations, but may be combined with other techniques.  Transformation
processes generally refer to biodegradation, or the enhancement of the
natural biological transformation processes, usually through the addition
of nutrients, particularly oxygen.

    The dissolved oxygen concentration is typically the rate-limiting
factor for aerobic biodegradation.  Groundwater typically contains very low
concentrations of dissolved oxygen,  although some systems do contain
                                    201

-------
NOTE: NOT ALL PHASE BOUNDA
     AIR IS NOT PRESENT.
        :><; WATER INS

          X   BIOTA (MICR

         CD SOIL PARTIC
Figure 8-1. Schematic Cr
           Contaminant!
           (i.e., Water in
 ES ARE SHOWN.
TURATED ZONE

 'BIOTA INCLUDED)

LE
I COLLOIDAL PARTICLE

" CONTAMINANTS
 ss-Sectional Diagram of Locus No. 8
 Dissolved in Groundwater
 he Saturated Zone)
                              202

-------
                                          (SOIL GAS)
                                          LOCUS NO.
                                    1 - CONTAMINANT VAPORS
                              VOLATILIZATION
              CONDENSATION
                                                       PHASE
                                                       SEPARATION
                   DEPURATION
        (BIOTA)
       LOCUS NO
   11-SORBEDTOBIOTA
NOTE: DASHED ARROW REPRESENTS RELATIVELY
     INSIGNIFICANT PROCESS MECHANISMS.
    LOCUS  NO. 8
(DISSOLVED IN GROUNDWATER
.e. WATER IN SATURATED ZONE)
     (LIQUID CONTAMINANTS)
         LOCUS NO.
 5 - IN PORE SPACES IN SATURATED
   ZONE
 7 - FLOATING ON WATER TABLE
13- IN ROCK FRACTURES
                                                            DISSOLUTION
                      UPTAKE
                                                            SORPTION
                                                        DESORPTION
                                                ADVECTION
                                                DISPERSION
                                                DIFFUSION
                                    (SORBED CONTAMINANTS)
                                          LOCUS NO.
                                  4 - SORBED TO "WATER-WET
                                    SOIL PARTICLES
                                  9 - SORBED TO COLLODIAL
                                    PARTICLES
           Figure 8-2. Schematic Representation of Important Transformation
                        and Transport Processes Affecting Other Loci.
                                                203

-------
sufficient oxygen to allow 1:
of oxygen to groundwater,
the addition of oxygenated
used to stimulate degradatior
phosphorus, may also be addec
    Anaerobic biodegradation
degradation.
degradation.
               mi ted aerobic degradation.   The introduction
            through air sparging,  pure oxygen injection, or
                ipounds such as hydrogen peroxide, is often
                  Other nutrients,  including nitrogen and
com
               seems to be far less important for hydrocarbon
Far less is kncwn about anaerobic degradation than aerobic
Because the rates of anaerobic degradation are far slower,
only aerobic degradation is considered for hydrocarbon remediation.
    Remedial measures used tc
contaminated by dissolved hyc
following questions:
    (1)  What is the extent
         vertically?
                ameliorate groundwater that has been
               rocarbon constituents must address the
              of the contaminant plume, both areally and
    (2)  What contaminants are present in the plume, and what are the
         physical and chemical characteristics of those compounds?
    (3)  What is the rate of
         feasible would it be
         immbolization or by

    (4)  What is the partitio
         that affect the pote
         contaminant sorbed t
         very difficult to re
               transport of the contaminant plume?  How
                to change the rate of transport, either by
               nobilization?

               ling behavior of the contaminants and how does
               ntial removal (e.g., is much of the
               ightly in dead pore space, where it will be
               move)?
    The following sections adiress the answers to these questions and
especially discuss how these mswers affect remediation decisions.

8.2.2  Mobilization/Remobilizition

8.2.2.1  Partitioning into Mo>ile Phase
    Locus no. 8 consists of contaminants dissolved in the groundwater.
this locus, the contaminant already exists in a mobile phase.  The
dissolution process may have occurred in the unsaturated zone (from locus
no. 2 or no. 6), at the groundwater table (from locus no. 7), from the
                                                          For
saturated zone (locus no. 5),
rocks (locus no. 4) or colloids (locus no. 9).  The dissolution of liquid
contaminant into the water ph<
    Partitioning of dissolved
result from volatilization in
no. 1) or phase separation frd>m
liquid hydrocarbon (locus no.
locus no. 8 at the water tab!
well-mixed system, volatiliza
loss mechanism.  Phase separa
would be insignificant in comparison
                or from desorption from water-wetted soil or
                se is described in detail in Section 7.
                hydrocarbon into another mobile phase can
                o the soil gas in the unsaturated zone (locus
                  the dissolved hydrocarbon back into the
                5 and 7).  Since air is only in contact with
                 and groundwater is not necessarily a
                ion is not considered to be a significant
                ion, defined as the reverse of dissolution,
                       to dissolution.
                                    204

-------
8.2.2.2  Transport with Mobile Phase

    Transport of solutes in the groundwater is typically described by a
conservation-of-mass statement, which states that the net change in an
elemental volume over some time period equals the inflow minus the outflow,
adjusted to account for any reactions that occur within the elemental
volume during that time period.  The flow into and out of the volume may be
described by advection and hydrodynamic dispersion.  (See Figures 8-3 and
8-4.)  Reactions that occur within that volume (e.g., biodegradation) can
be difficult to describe mathematically, and are often assumed negligible.
Transport may then be described by the advection-dispersion equation, the
basic groundwater flow equation.  If we consider nonreactive (conservative)
solutes in saturated, homogeneous, isotropic materials under steady state,
uniform flow conditions, the one-dimensional form of the equation is
(Freeze and Cherry, 1979):
    at  ~ R

where

    C. = concentration of contaminant i dissolved in groundwater (mg/L)
    D, = hydrodynamic dispersion coefficient (cmVsec)
    V, = average linear groundwater velocity (cm/sec)
     1 = distance along flowline (cm)
     R = retardation factor (dimensionless)
     t = time (sec)


    The first term on the right-hand side of equation 8.1 represents the
diffusive flux, while the second term represents the advective flux.  The
coefficient of hydrodynamic dispersion is composed of two components:
mechanical dispersion and molecular diffusion.  The diffusion coefficient
becomes significant in relatively quiescent systems.  Dispecsion can also
be important when considering dilution or retardation of the contaminant.
The dispersion coefficient can be calculated as follows:

    D = o^ V + D*                                                     (8.2)

where

     D = dispersion coefficient (cmVsec)
    a. = longitudinal dispersivity of the media (cm)

     V = average linear groundwater velocity (cm/sec)

    D* = effective diffusion coefficient (cm2/sec)
    The average linear groundwater velocity, V,, in equation 8.1, is equal
to the specific discharge divided by the effective porosity (q/9).  The
influence of this transport mechanism dominates the effect of diffusion for
                                    205

-------
    Mixing in individual pores
                            M
Source: Freeze and Cherry, 1979. (C


            Figure 8-3. Processed
                                                    lii^r^
                                                    Mixing by molecular

                                                         diffusion
xing of pore channels
pyright Prentice-Hall Inc., 1979. Reprinted with permission.)



 of Dispersion on a Microscopic Scale.
     1  100-
      a
      tS
      Q
        200-
        Source: After Schwille, 1981


        Figure 8-4. Concentration C

                 Result of Hydro*
                                                      1%

                                                  After 12 months
 istribution of Dissolved Components as a
 ynamic Dispersion (flow velocity: 0.5 m/day).
                                       206

-------
most cases.  Remedial measures  that work by mobilizing the groundwater
(e.g., extraction) depend on advective flow.  Mathematically, this velocity
may be represented:


    Vl = q/6 =  (1/6) *_£J[  g                                        (8.3)


where:

    v, = average linear groundwater velocity (cm/sec)

     q = volume flow per unit area (cmVsec cm2)

     k = intrinsic permeability of the media (cm2)

     p = density of the fluid (g/cm3)

     g = gravitational constant (cm/sec2)

     u = dynamic viscosity (g/cm-sec)

     6 = effective porosity of  the media (dimensionless)

    -TJ- = hydraulic gradient (cm/cm)


    Retardation can have a significant influence on the spread of the
dissolved contaminant plume with time.  Retardation is simply the ratio of
the velocity of a dissolved contaminant plume in relation to the bulk
velocity of the groundwater.  It can be described by the conventional
retardation equation:

    R = Vw/Vc = 1 + Kdpb/6                                            (8.4)

where:

     R = retardation factor (dimensionless)

    V  = average velocity of water (cm/sec)

    V  = average velocity of chemical pollutant (cm/sec)

    K, = sorption coefficient (cm3/g) (see Section 4.2.2)

    p,  = soil bulk density (g/cm3) (see Section 3.3.2)

     6 = soil porosity (dimensionless) (see Section 3.3.2)


    Equation 8.4 shows that the retardation factor moves inversely to the
porosity.  That is, at low porosity values the retardation increases, while
retardation decreases as the porosity increases.  The retardation factor is
related to the sorption behavior of solutes in groundwater.  Sorption will
be discussed more fully in the following section.

    Equation 8.1 considered only one-dimensional flow.  In the field,
however,  dispersion occurs in three dimensions.  For the three dimensions,
the advective-diffusion equation reads:
                                    207

-------
    at
' - E-)  [
Dx
             3x:
                      i + Dy
                             
-------
    Vc - V/R                                                          (8.8)

where:

    V  = average velocity of the contaminant (cm/sec)

    V  = average linear groundwater velocity (cm/sec)

    R  - retardation factor (dimensionless)


Examples of transport calculations using equations 8.6 and 8.7 will be
shown in Section 8.4.2.

Dispersion

    In Equation 8.1, the dispersion term
    i_  (D ^
    ax  r ax
is defined in accordance with Pick's Laws.  Near the source of
contamination, however, dispersion is a non-Fickian phenomenon.  In a
porous media, dispersion from the source becomes Fickian at distances on
the order of tens to hundreds of meters.  This is due to the variation in
dispersivity (a,) in equation 8.2 from the source (usually increasing with
distance).  The change in dispersivity is considered to be a second-order
effect relative to the heterogeneity of the media.

    Dispersivities (a,), which are dependent on grain size distribution but
independent of grain shape, are easily measured in a laboratory.  They are
more difficult to measure in the field, however; here values range from
0.01 to 2 cm, higher than laboratory values due to heterogeneities in the
macroscopic flow field.

Diffusion (Details in Section 7.2.2)

    Diffusion in solutions is the process whereby molecular constituents
move under the influence of their kinetic activity in the direction of
their concentration gradient (from high concentration to low).  Diffusion
occurs in the absence (or presence) of advection and ceases only when
concentration gradients become nonexistent.  In equation 8.2, the effective
diffusion coefficient, D*, takes into account the tortuous path for
nonadsorbed species in porous media.

                             D* = TD                                  (8.9)

where:

    D* = effective diffusion coefficient (cmVsec)

    D  = diffusion coefficient (cmVsec) (see Section 3.2.2)

    t  = tortuosity factor (0 < w < 1) (dimensionless)
                                    209

-------
    The tortuosity factor may
the media (Millington and Qui
      be expressed as a function of porosity (9) of
      •k, 1961):
    For neutral organic
range of 10"   to 10   cm2/sec
is temperature dependent; for
as it is at 25°C.  Groundwatei
stable, although it does vary
                                 6
                                  1.33
                                             (8.10)
chemicals in water, typical values for D are in the
      (see Table 3-1).  The diffusion coefficient
      example, at 5°C D is only about half as large
       temperature within any region is relatively
      across larger areas (Figure 8-5).
Source: Heath, 1983

Figure 8-5.  Approximate Temperature
           States at Depths of 10
          of Groundwater in the Conterminus United
       o 25 Meters (degrees Celsius).
Advection
    The average  linear groundwe
is the main component of  the
conductivity  (K) and hydraulic
average linear groundwater  veldcity
       ter velocity (V), defined in Equation 8.3,
       vective term in equation 8.1.  The hydraulc
       gradient (dh/dl) strongly influence  the
                                     210

-------
    The saturated hydraulic  conductivity  is often measured  rather  than  the
intrinsic permeability of a  media.
    K «                                                   (see equation 8.3)


where

    K = saturated hydraulic conductivity  (cm/sec)

    k = intrinsic permeability  (cm2)
    p = density of fluid  (g/cm3)
    g = gravitational constant  (cm/sec2)

    y = dynamic viscosity  (g/cm-sec)

                                 *»-
    Hydraulic conductivity is related to  the grain size of a porous media
(see Figure 8-6).  A simple empirical relationship to estimate the
saturated hydraulic conductivity is based on the grain size distribution.

    Hydraulic fracturing may also be useful for increasing the conductivity
of an area surrounding a well (Murdoch et al., 1990).  This method involves
pumping fluid into a well until pressures exceed a critical value, beyond
which fracturing begins.  Sand  is then pumped into the fractures  to hold
them open and provide a high permeability channel.  This method,  developed
for enhanced oil recovery many years ago, is especially useful in regions
of low permeability, and may help to improve remediation methods  that
depend on mobilizing contaminants (groundwater pumping, soil venting,
etc.).

    The main component of advection is the average linear groundwater
velocity, which depends on the hydraulic conductivity and gradient of the
media.  The average linear groundwater velocity is greatest in homogeneous
soils with high porosity, and with more soluble constituents.  The
conductivity can be increased for the purpose of remediation (e.g.,
flushing, extraction, etc.) by  increasing the gradient.

    The grain size distribution, easily measured in a laboratory  from a
soil sample, should be used to  find the following parameters:

    D,Q = grain size diameter larger than 60% (by weight) of the  soil's
          grains

    Dcn = median grain size

    D.,. = grain size diameter larger than 10% (by weight) of the  soil's
          grains

    The uniformity coefficient is equal to D,n/D1Q.  Once these values are
known, they can be used with Figure 8-6 (Robson,  1978) to find the
hydraulic conductivity.  The range of values for hydraulic conductivity is
very large (Table 8-1), from 100 to 10    cm/s.  Hydraulic conductivity
values can be measured in the field through a variety of aquifer  test
                                    211

-------
   3QOOOL. M.I.
    3000
w
P-

H
W
W
>
M
H



Q

O



a
M
J
     300
           EXPLANATION


Cu - Uniformity Coeff. =

--- Extrapolated Curve
                                              - 60

                                               10
           MEDIAN GRAIN DIAMETER (D  ) IN nun
 Source: Robson, 1978.
 Figure 8-6. Relations
            Coefficient,
                                                   0.01
 Among Hydraulic Conductivity, Uniformity
   and Median Grain Diameter.
                              212

-------
 methods (e.g.,  slug  test or pump test).  For neutral organic compounds in
 water, typical  groundwater velocities may range  from tens to hundreds of
 meters/year.  Naturally-occurring hydraulic gradients are typically on the
 order of 0.01 (dh/dl).

 Retardation

     Contaminants  dissolved in groundwater can become retarded by sorption
 onto mineral grains  in  soils or rocks in the saturated zone.  Sorption
 retards the movement  of a contaminant compared to  the groundwater movement
 (see locus no.  4  and  no.  9).  When partitioning  of a contaminant between
 the liquid and  solid  phases is completely reversible,  the retardation of
 the contaminant relative to the bulk mass of the groundwater can be
 described by equation 8.4.

     Species in  a  mixture of reactive contaminants  that are dissolved in
 groundwater will  travel at  different rates depending on their retardation
 factors (see Eq.  8.4).   Thus,  the contaminant plume  will  exhibit a
 chromatographic effect  over time.  As a contaminant  plume advances along
 flow paths, the front is  retarded as a portion of  the contaminant mass
 sorbs to the solid phase  (Figure 8-7).  If the input  of contaminant mass to
 the groundwater is discontinued, and the groundwater  concentrations drop
 (for example, as  a groundwater pump and treat scheme  progresses),
 contaminants may  desorb from the soil back into  the  groundwater.  Thus, it
 is doubtful that  all  of the contaminant can be permanently immobilized,
 even though the retardation of the concentration front may be strong.
 However,  a portion of the contaminants may be irreversibly fixed relative
 to the time scale of  interest  (see Figure 8-8).
                   Continuous tracer supply at time t>0
                           Distance	*•

      (a) Dispersed front of non-retarded solute
      (b) Front of solute that undergoes equilibrium partitioning between liquid and solids
      (c) Front of solute that undergoes slower rate of transfer to solids

Source: Freeze and Cherry, 1979. (Copyright Prentice-Hall Inc., 1979. Reprinted with permission.)

Figure 8-7. Advance of Reactive and Nonreactive Contaminants Through a Column.
     Retardation is strongly affected  by subsurface carbon content.   The
effect  on retardation differs, however,  based on the phase of  the  carbon.
Dissolved organic carbon (DOC) (e.g.,  highly soluble compounds such as MTBE
and  tertiary butyl alcohol and colloidal particles of humic and  fulvic
                                      213

-------
                 Continuous source
                 of contominotioi
                   u in
Steady - state flow
                                                  K,j= Distribution coefficient
                                                  Transport time =60 years
               Porosity - 0.3
               Hydraulic conductivity • 0.5 m/day
               a
               Transport time —„ ,	
               Concentration contours at C/Co = 0.9,0.7, 0.5, 0.3, and 0.1

Source: Freeze and Cherry, 1979. (Colyright  Prentice-Hall Inc., 1979. Reproduced with permission.)

Figure 8-8.   Effect of the Dlstrll >ution Coefficient on Contaminant Retardation
              During Transport in a Shallow Groundwater Flow System.
                                         214

-------
acids) may significantly lower retardation.  Organic matter in the solid
phase increases sorption and hence retardation.

    Garrett et al.  (1986) report on how the presence of methyl tertiary
butyl ether (MTBE)  in gasoline spills affects the movement of the plume and
the detection of a  release.  MTBE has multiple effects related to
retardation and movement.  First, MTBE, which may comprise up to 11% by
volume of gasoline, is soluble in water to 43,000 mg/L (Garrett et al.,
1986), about 25 times more soluble than benzene, otherwise the most soluble
gasoline constituent at 1,780 mg/L.  Generally, sorption of organic
compounds is inversely proportional to their solubility.  Thus, MTBE is
likely to be minimally retarded, allowing it to travel faster than other
gasoline constituents.  Indeed, Garrett et al. (1986) report that MTBE
often forms a "halo" around a dissolved gasoline plume, where MTBE can be
detected but other  gasoline compounds are not detected.

    MTBE also decreases the retardation of other gasoline constituents by
increasing their dissolved concentrations in the groundwater.  Benzene,
toluene, ethylbenzene, and xylenes (BTEX) are more soluble in ether than in
water.  When MTBE is present in the gasoline plume, the BTEX concentrations
in the groundwater  are higher.  Because the BTEX compounds are more
soluble, they are likely to be more mobile (retarded less) in the
groundwater.

    For groundwaters that contain significant dissolved DOC, the
conventional retardation equation (Eq. 8.4) can be modified to reflect the
effect that DOC has:

                  K,p,/e
    R = 1 + 	—	-?                                      (8.11)
            1 + KDQC (DOC) 10-b

where:

       R = retardation coefficient (dimensionless)
      K, = sorption coefficient (L/kg)

      p,  = soil bulk density (kg/L)
       6 = porosity (dimensionless)

    KQf)r = sorption constant for compound in DOC (L/kg) (~Koc)

     DOC = dissolved organic carbon concentration (mg/L)


    This equation presumes that the solid organic matter is less than 0.1%,
so Kd-type constants may replace Koc constants.  Higher levels of organic
carbon in the solid matter can increase sorption and elevate retardation
effects.

    Figure 8-9 shows that retardation factors are significantly lower at
DOC concentrations of 100 ppm, particularly for high-molecular weight
organic compounds.  The variability in the retardation factors due to
dissolved carbon adds to the inherent variance; Table 8-1 shows that
                                    215

-------
                                 TABLE 8-1

                       ESTIMATED RETARDATION FACTORS
                   FOR HYDROCARBON GASOLINE CONSTITUENTS

Estimated
Soil
Sorption Constants


n-Butane
Isobutane
n-Pentane
Isopentane
1-Pentene
n-Hexane
1-Hexene
2-Methylpentane
Cyclohexane
Benzene
n-Heptane
2-Methylhexane
Methylcyclohexane
Toluene
n-Octane
2, 4-Dime thylhexane
Ethylbenzene
m-Xylene
2,2, 4-Trime thylhexane
1,3, 5-Trime thylbenzent.
2,2,5, 5-Te trame thylhexane
1 , 4-Diethylbenzene
Dodecane
Koc
(L/kg) (f
490
420
910
880
460
1,900
910
1,500
960
190
4,300
3,200
1,800
380
8,200
5,200
680
720
8,700
940
14,000
2,900
88,000
Kd
oc
0.49
0.42
0.91
0.88^
0.46^
1.90
0.91
1.50
0.96
0.19
4.30
3.20
1.80
0.38
8.20
5.20
0.68
0.72'
8.70
0.94
14.0
2.9
88.0
Estimated
Retardation
Factors
R
4.5
4.0
7.6
7.3
4.3
14.7
7.6
11.8
7.9
2.4
32.0
24.0
14.0
2.7
60.0
38.4
5.9
6.2
63.6
7.8
101.8
21.9
634.0
a. Kd  = KQ f
   f   = Fraction of organic carbon

b. Using equation 8.4, a bulk
   0.20

c. Nonlinearity is important f
   concentrations are close to
density p  =1.8 L/kg and soil porosity 0
Or these constituents when their
 their solubility limit.
                                    216

-------
retardation factors  range  over two orders of magnitude.  The higher
molecular weight aliphatic compounds are most strongly retarded.
                           OC  -  0.001
      log  R
5

4
3
2




r\
0




CD
CD
3
N
CD
3
CD
-^"
^***^
0 1 2

H
CD
S
O
3-
0
O
CD-
^^^
/•&
CD
3
CD>
s*S--"

3
log


•o
cF
3
CD "'
^



4
Ko>
3-
CD
3
&
3
•-*
3-
CD
3
CD,
/
r
^ ^




W
O
en
1
•y.
O
CD
Xo


100

1000

5 (

0
o
H
D
o'
X
5X
r
ppm


pprr

pprr

3









i

7

 Source: Kan and Thomson, 1986. (Copyright Water Well Journal
       Publishing Co., 1986. Reprinted with permission.)

 Figure 8-9. Relationship of Retardation Factors to Octanoi-Water Partition Coefficient
          by Conventional (	) and Revised (—  —  -,	) Equations.
8.2.3  Fixation  (Details  in Section 4.2.3)

8.2.3.1  Partitioning  onto Immobile (Stationary) Phase

    Sorption onto soil particles (locus no.  4) is a possible means of
fixation.  Soil  sorption  is an important factor in solute retardation,
discussed in the preceding section.   Sorption is not reliable for
preventing transport with the  ambient groundwater flow for all compounds.
For other compounds, however,  sorption to soil does virtually prevent
significant transport  (especially high molecular-weight compounds).  Thus,
sorption would probably not be an effective  mechanism for permanent
fixation of all  of  the contaminant,  although a portion of the contaminant
may be irreversibly fixed relative to a specific time scale.

    Sorption could  possibly be an effective  mechanism for fixation of most
of the contaminant  in  a relatively quiescent groundwater system.  A slower
                                     217

-------
  groundwater velocity  results i
  lower concentrations  in  the tr
  benzene in Figure 8-10.
               0 8




               0 6


             C/Co


               0 4




               0.2




               0 0
Source: Clark et al., 1986. (Copyright Water W

 Figure 8-10.  Normalized Breakthrough
             Linclon Fine Sand Soil Co
 Journal Publishing Co. ,1986. Reprinted with permission.)

Curves for Displacement of Benzene of 15-cm
umn as a Function of Water Velocity
   8.2.3.2  Other Fixation Approaches
       None are immediately evide

   8.2.4  Transformation

   8.2.4.1  Biodegradation (Detai

       Groundwater  is  largely dev
   lack of surficial contact with
   microorganisms primarily respo
   large numbers in this locus.
   groundwater, therefore,  is usu
   However, if contaminants are t
   (due to fluctuations  in the gr
   oxygenated, biodegradation can

   8.2.4.2  Chemical Oxidation (D

       Contaminants in groundwate
   (provided precursors  such as p
   of aerobic microbial  process.
 a  greater amount of sorption and, thus,
nsmitted water.  An example  is given for
              a - Banz.n. - 10.3 ll/d«y

              b - Benzen. - 2.1 ft/day

              c - Bcnzeni - 3.3 M/d»y
        C = chemical concentration in column effluent
        Co » chemical concentration in column influent
        V = volume of column effluent
        Vo = pore volume of column
                  I
I
 t.
 s  in Section 11)

 id of oxygen due to  its  slow movement and
 air.   As a result, the aerobic
 sible for biodegradation cannot survive  in
 iodegradation of contaminants in the
 lly an insignificant  means of removal.
 ansported to the water  table or vadose zone
 undwater elevation)  which are much more  well
 occur much more rapidly.

 tails in Section 3.2.4.2)

  can undergo abiotic transformation
 enol and catechol are present) as a result
                                        218

-------
8.3  Storage Capacity  in Locus

8.3.1  Introduction and Basic Equations

    Some of the information below considers the case for gasoline, which is
the product most commmonly held in an UST.  The maximum concentration of an
unadulterated gasoline that can be dissolved in groundwater is directly
related to the solubilities of each constituent in a gasoline blend (for
further discussion of dissolution, see Section 7).

    For unoxidized hydrocarbon in contact with pure water the concentration
of the solute in aqueous solution is proportional to the concentration of
the constituent in gasoline and the solubility of the solute.

                        Ciw =Xig Siw                                 <8

    i = 1, 2, 3 ... n constituents

where
    C.  = concentration of constituent i in the water phase (mg/1)
    X.  = mole fraction of constituent i in gasoline (dimensionless
    S.  = solubility of constituent i in pure water (mg/1)
    o  Calculations are more difficult if significant amounts of oxygenated
       additives are present.

8.3.2  Guidance on Inputs for, and Calculation of, Maximum Value

    An approximate maximum for non-oxygenated gasoline blends or for
hydrocarbon constituents of oxygenated fuels is on the order of 200 mg/1.
The value for oxygenated fuels (i.e., those containing MTBE, TEA, ethanol,
methanol, etc.) is poorly constrained at present, but is at least an order
of magnitude greater than that for non-oxygenated fuels.  The general
effect of co-solvents on solubility can be stated as:
                        In Sm = In SW + fff                           (8.13)
where
    S  = mole fraction solubility in water (dimensionless)

    S  = mole fraction solubility in mixed solvents (dimensionless)
     a = parameter related to solutes' surface area and interfacial free

         energy

     f = volume fraction of co-solvent (0 < f < 1) (dimensionless)
    The value of 200 mg/1 stated above for the maximum concentration of
gasoline without additives dissolved in groundwater is an approximation
based on a laboratory study (API, 1985) and solubility calculations from
                                    219

-------
literature - reported values
found in the field are typics
value.  The maximum values ii
advection and dispersion ser
concentrations.

    Hydrocarbon solubility ir
changes in temperature, ioni
pressure are nearly constant
somewhat with increasing ion
may be affected, perhaps qui
Garrett et al. (1986) note tt
compounds are more soluble
BTEX are almost completely sc
equation 8.13) are empirical
gasoline products.  In gener
dissolved organic carbon.
   Lyman,  1987).   The actual concentrations
  lly much lower  than the maximum estimated
   the field typically diminish with time,  as
  e to spread the plume and dilute the
   water is not significantly affected by modest
   strength,  pH,  or pressure.  Temperature and
  for most groundwaters.   Solubility is reduced
    strength in aqueous solution.   Solubility
  e significantly,  by co-solvent effects.
  at in gasolines that contain MTBE, the BTEX
than in gasolines not containing MTBE, because
   .uble in ether.  Co-solvent values for a (see
   -derived;  their availability is limited for
   ., solubility increases with increasing
    The composition of the d
composition of the bulk prodx
gasolines, the most soluble
toluene, xylene, and ethylber
studies of the water soluble
products often stored in und
BTX compounds comprise 23-5S/
benzene concentrations are e
product (Guard et al., 1983)
solubilization experiments ar
contained 49% aromatic compoi
hour of mixing.

    Solubility data for gaso
presented by Kramer and Haye
experimental work are shown
    Benzene
    Toluene
    MTBE
    The concentration of ben
tripled in the super unleade
of 966 mg/1 is 27 times regu
state the concentrations of
increased levels of benzene
gasoline may be due to cosol
  ssolved gasoline is quite different from the
  ct (see Table 3-2).  For non-oxygenated
  omponents are the light aromatics:   benzene,
  zene (BTEX).   The literature includes several
  iraction (WSF) of gasoline and other petroleum
   ground storage tanks.   The results show that
   of the bulk product but 42-74% of the WSF;
  riched ten times in the WSF versus the bulk
    Coleman et al. (1984) performed
  d found that the unleaded gasoline product
  nds, but the WSF contained 95% after one-half
  me that contains oxygenated additives is
   (1987).  Partial results of their
  n Table 8-2.
                                TABLE 8-2
             WATER SOLUBLE FRACTIONS FOR THREE GASOLINES (PPH)
                             Regular
                             Leaded
                 Regular
                 Unleaded
               Super
               Unleaded
  30.5
  31.4
  43.7
28.1
31.1
35.1
67
107.4
966
  ene is roughly doubled and toluene, more than
   brand.  The super unleaded MTBE concentration
  ar unleaded.  While Kramer and Hayes do not
  snzene and toluene in the bulk product, the
  id toluene in the WSF of the super unleaded
  snt effects from the MTBE.
                                    220

-------
8.3.3  Guidance on Inputs for, and Calculations of, Average Values

    o  Typical value for non-oxygenated gasoline blends is on the order of
       10-20 mg/L.

    o  Oxygenated fuels are at least an order (if not several orders) of
       magnitude greater than for non-oxygenated fuels.

8.4  Example Calculations

8.4.1  Storage Capacity Calculations

8.4.1.1  Maximum Value

    o  Use maximum value of 200 mg/L given in Section 8.3.2 for
       unadulterated gasoline.

8.4.1.2  Average Value

    o  Use average value of 20 mg/L given in Section 8.3.3.

    These values are based, in part, on a weighted sum and average of
calculated concentrations using equation 8.12:

                       Ci = Xig Sig

For illustrative purposes, an example is provided in Table 8-3.

8.4.1.3  Example of Solubility Enhancement

    For oxygenated fuels, the effect of co-solvents on solubility can be
shown by the following example.  Consider the enhanced solubility of
benzene in water due to the presence of ethanol using equation 8.13:

    Sw = 1780 mg/L (25°C) in pure water (see Table 8-3)
       = 4.11 x 10~4 mole  fraction

    Assume ethanol present at 10% volume and assume o = 10, f = 0.1

    In Sm = In (4.11 x lO"4) + 10  (0.1) = -7.797 + 1
          = -6.797
       Sm = 1.118 x 10"3 mole  fraction
          = 4840 mg/L

    As can be seen from the example, ethanol increases the solubility of
benzene nearly three times.

8.4.2  Transport Rate Calculations

    For unadulterated gasoline, the transport of each constituent needs to
be considered in calculating the effect of transport on the concentration
of the blend.  The following example considers the transport of benzene
dissolved in groundwater and uses equation 8.12.
                                          221

-------

ESTIMATED S(

Gasoline, Component
n-Butane
Isobutane
n-Pentane
Isopentane
1-Pentene
n-Hexane
1-Hexene
2-Methylpentane
Cyclohexane
Benzene
n-Heptane
2-Methylhexane
Methylcyclohexane
Toluene
n-Octane
2 , 4-Dimethylhexane
Ethylbenzene
m-Xylene
2,2, 4-Trimethylhexane
1,3, 5-Trimethylbenzene
2,2,5, 5-Te t rame thylhexane
1 , 4-Diethylbenzene
Dodecane

Source: Lyman, 1987
Note: No additives considere

TABLE 8-3


LUBILITY OF GASOLINE IN WATER
X
Selected
Cone, in
Gasoline
(Mol. Fr.)
0.0163
0.0326
0.0394
0.184
0.0203
0.0990
0.0169
0.0880
0.0338
0.0364
0.0142
0.0473
0.00966
0.0515
0.00830
0.0664
0.0179
0.0625
0.0148
0.0394
0.0100
0.0353
0.0558
2 = 0.998


d.
222
S
Reported
Solubility in
Pure Water
25°C (mg/L)
61.4 (1 ATM)
48.9 (1 ATM)
41.2
48.5
148
12.5
50
14.2
59.7
1780
2.68
2.54
15
537
0.66
1.5
167
162
0.8
72.6
0.13
15
0.005
= 3240




C
Est. Cone.
in Gasoline-
Saturated
Water (mg/L)
1.00
1.59
1.62
8.92
3.00
1.24
0.84
1.25
2.02
64.80
0.38
0.12
0.14
27.6
0.0055
0.10
2.99
10.1
0.012
2.86
0.0013
0.53
3 x 10~4
= 131
219 (10°C)




-------
    o  What is the concentration after one year of an instantaneous
       benzene-saturated point source of 1 m3 in groundwater, with an
       average linear pore velocity in the x-direction of 100 m/yr and
       dispersivities of 1 m and 10 cm in the x and y directions,
       respectively?  Steady-state groundwater flow in a homogeneous,
       isotropic media are implicitly assumed.   Using equation 8.7:
                   m
                       D
                      x y        x

o  Contaminant Mass, m, in system
                                   ( -x2    v2 ^
                                   4D~T - 4DtJ
Where:
       m = (C.) (V)
       C. = benzene concentration--at saturation
        V = unit volume

        m = (1.78 x 106 mg/m3) (1 m3) (for benzene)
          = 1.78 x 106 mg
    o  Use equation 8.2 to determine dispersion coefficients, D ' & D '
                                                               x     y

        D* = 1 x 10"  cmVs (benzene diffusion in water)
    a  = 100 cm
     J\
       = 10 cm

     V = 100 m/yr
        a  = 10 cm determined by measured extent of the plume
       D '  = a  V + D*
        xx                                        ,.
           = (100 cm) (10,000 cm/31,536,000 sec) + (10   cmVs)

           = 0.03172 cm2/s
       D '  = (10 cm)(10,000 cm/31,536,000 sec) + (10 5 cm2/s)
           = 0.00318 cm2/s
    Taking into account retardation (R = 1.5 for benzene)
       D  = D' /R = 0.03172/1.5 = 0.02115 cmVs
        x     x
       D  = 0.00318/1.5 = .001414 cm2/s


    Since the mass, M, is defined as a point source that spreads with
transport,  the peak concentration will be at the center of the
                                    223

-------
concentration profile (Gaus
maximum concentration at th

    The concentration at an
contaminant has traveled th
occurs when the distance tr
velocity at a particular ti

    x -» 0 when x = Vt ,
                   r (see
    Therefore, to determine
set x = 0.

    o  Input Values

        t = 1 year = 31,536

        m = (C.) (V) = 1.78

       D* =1 x 10~5 cm2/s

       a  = 100 cm
        X
       a  = 10 cm

        V = 100 m/yr

        R = 1.5
       D
        >
       D
x
        y
        x = 0
        y = 0
    C (x,y,t) =
0.02115 cm'/s (

0.001414 cm2/s
                4 (n)(31,53

       = 10,150 mg/m3
       =10.2 mg/L

    Thus, for 1,780 grams o
the final concentration is

    The rate of transport
       = 100/1.5 = 66.6 m/j

8.5  Summary of Relative Im

8.5.1  Remediation

    Hydrocarbons that are
serious pollution problems
                    ian distribution).   Thus, in determining the
                     centerline,  we can let y = 0.

                     point at any time  is at a maximum when the
                     least distance,  that is, as x -» 0.  This
                    veled is equal to the average linear pore
                    e:
                    efinition of x for equation 8.6)

                    the maximum concentration at time, t, we can
                    000 seconds
                    x 106 mg
 x 10 4) = 2.1 x 10 6 m2/s

1 x 10~4) = 1.4 x 10~7 m2/s
                               (1.78 x 10 )
                    ,000)/(2.1 x
                                           v-6
                                              — exp (0)
                     benzene that traveled 66 meters in one year,
                    0.2 mg/L.

                    sing equation 8.8) is simply:
                             Vc = V/R
                    iortance of Locus
                    ssolved in groundwater pose one of  the more
                     The solubility of the lighter hydrocarbon
                                     224

-------
constituents greatly exceeds the concentration levels at which groundvater
is considered to be seriously polluted.  For example, the typical
solubility of a gasoline might be approximately 20 mg/1, but it can be
detected by taste and odor at concentrations of less than 0.005 mg/liter.

    This low taste and odor threshold and the toxicity of many hydrocarbons
(even at low levels) argues against fixation as a remedial measure.  Natual
flushing, dispersion and biodegradation serve to reduce concentrations of
hydrocarbons in the aquifer, but the time scale for this approach is
sufficiently long to preclude this means (the "no-action" approach) from
consideration for all but a few sites.

    Many contaminated sites will require removal and treatment to restore
the aquifer to a safer condition.  Historically, this has meant pumping of
the contaminated groundwater followed by above ground treatment, usually
via air stripping or activated carbon.  Biostimulation - the addition of
oxygen and other nutrients to improve the natural biodegradation processes
- is being used at an increasing number of sites but is still a far less
popular option than either air stripping or activated carbon.

    All three groundwater treatment methods have shown the ability to
reduce aquifer concentrations to below regulated levels.  Their application
is site-specific, however, and some sites prove difficult to restore to
pre-contaminated condition.

8.5.2  Loci Interactions

    The loci that interact with hydrocarbon constituents dissolved in
groundwater during partitioning and mass migration processes are summarized
in Figure 8-2 and Table 8-4.  In an open system, groundwater can act as a
sink for mobile constituents from other loci.  Dissolution of hydrocarbon
constituents may have occurred in the unsaturated zone (loci nos. 2 or 6),
at the groundwater table (Locus no. 7), from the saturated zone (loci no.
4, 5, or 9).  Bulk transport out of the locus is influenced by the media
through which the groundwater flows (loci 4, 10, or 13) and the overburden.
Aromatic constituents dissolved in groundwater can volatilize into the
overlying soil gas (locus no. 1), although the relative amounts are
probably not significant.  The distribution of hydrocarbon in or out of the
groundwater may depend on the nature of groundwater movement and proximity
to discharge points. Of these loci, those containing liquid hydrocarbon are
relatively more important.

8.5.3  Information Gaps

    For a better understanding and definition of factors affecting
hydrocarbon constituents dissolved in groundwater,  the following
information is needed:

    (1)  Fundamental data on solubility, especially for multi-component
         products like petroleum, in addition to co-solvent effects for
         additives.

    (2)  Composition and physicochemical properties of gasoline.
                                    225

-------
                                TABLE 8-4
                    LOCI INTERACTIONS WITH HYDROCARBON
                          DISSOLVED IN GROUNDWATER
    Process
   Phas
     Co
    Mobility
    Dissolution
    Bulk Transport
    (advection and
    dispersion/
    diffusion)

    Volatilization

    Immobility
    Sorption
   wet
   liqu
   hydr

   liqu
   hydr
   wet
   rock
   air
   wet
   rock
s in
itact
                                 a
Interacting
   Loci
 Relative
Importance
oil
d
carbon

d
carbon
oil
oil
   4
   5,7
   5,7

   4
   13
   1
   4
   10,13
low
high


high

moderate
moderate-high
low
low
low
    a.Biota (locus no. 11) a
       re potentially in contact with all phases
    (3)  Dissolution rates f
         from unsaturated zo

    (4)  Better definition o
         anaerobic systems,

    (5)  Concentration of ad
         behavior as co-solv
         data of additives.
        r product into groundwater,  including recharge
        e residual.

       E natural biodegradation rates in aerobic and
       md factors affecting these rates.

       litives in different gasoline blends, their
       snts and in groundwater, and basic property
    (6)  Sorption data (K,,
         include nonequilibr
          ) representing natural soil systems that
        urn effects.
    (7)  The influence of
         transport and
     me lia structure and heterogeneity on bulk
  retardation.
    (8)  Further definition
         heterogeneity and p

8.6  Literature Cited

API.  1985.  Subsurface Vent
    Aquifer.  Publication No
       >f dispersivities in relation to media
       roximity to the source.
       Lng of Hydrocarbon Vapors from an Underground
         4410.  Washington, D.C.
Baetsle, L.H. 1969.
    Progress in Nuclear
    (Ed.) Pergamon Press,
Migration of Radionuclides in Porous Media.
   Enerjry, Series XII, Health Physics.  A.M.F. Duhamel
     Eljisford, NY.  p. 707-730.
                                    226

-------
Clark, G.L., A.T. Kan, and M.B. Tomsom.  1986.  Kinetic Interaction of
    Neutral Trace Level Organic Compound with Soil Organic Material.  In:
    Proceedings of Petroleum Hydrocarbons in Ground Water:  Prevention,
    Detection, and Restoration.  NWWA/API, Houston, TX.

Coleman, W.E., J.W. Munch, R.P. Streicher, H.P. Ringhand, and F.C. Kopfler.
    1984.  The Identification and Measurement of Components in Gasoline,
    Kerosine, and No. 2 Fuel Oil that Partition into the Aqueous Phase
    After Mixing.  Arch. Environ. Contain. Toxicol. 13:  171-178.

Ebach, E.A. and R.R. White.  1958. Mixing of Fluids Through Beds of Packed
    Solids.  American Institute of Chemical Engineering Journal, Vol. 4,
    no. 2.

Freeze, R.A. and J.H. Cherry.  1979.  Groundwater.  Prentice-Hall Inc., New
    Jersey.

Garret, P., M. Moreau, and J.D. Lowry. 1986.  Methyl Tertiary Butyl Ether
    as a Ground Water Contaminant.  In:  Proceedings of Petroleum
    Hydrocarbons and Organic Chemicals in Ground Water Conference.
    NWWA-API.

Guard, H.E., J. Ng and R.B. Laughlin. 1983.  Characterization of Gasolines,
    Diesel Fuels and their Water Soluble Fractions.  Naval Biosciences
    Laboratory, Naval Supply Center, Oakland, CA.

Heath, R.C. 1983.  Basic Ground-water Hydrology.  USGS Water Supply Paper
    2220.  84 pp.

Kan, A.T. and M.B. Tomson.  1986.  Facilitated Transport of Naphthalene and
    Phenanthrene in a Sandy Soil column with dissolved Organic Matter -
    macromolecules and Micelles.  In:  Proceedings of Petroleum
    Hydrocarbons and Organic Chemicals in Groundwater.  NWWA/API, Houston,
    TX.

Kramer, W.H. and T.J. Hayes.  1987.  Water Soluble Phase of Gasoline:
    Results of a Laboratory Mixing Experiment.  New Jersey Geological
    Survey Technical Memorandum 87-5.  Trenton, NJ.

Lyman, W.L. 1987.  Environmental Partitioning of Gasoline in
    Soil/Groundwater Compartments.  Final Seminar Proceedings on
    Underground Environment of an UST Motor Fuel Release.  USEPA RREL,
    sponsors, Edison, NJ.

Millington, R.J. and J.M. Quirk. 1961.  Permeability of Porous Solids.
    Transactions Faraday Society, 57:1200-1207.

Murdoch, L.C., G. Losonsky, I. Klich, and P. Cluxton.  1990.  Field Tests
    of Hydraulic Fracturing to Increase Fluid Flow in Soils.  Poster
    Session, Sixteenth Annual EPA Hazardous Waste Research Symposium,
    Cincinnati, OH.  April 3-5.

Ogata, A. 1970.  Theory of Dispersion in a Granular Medium.  USGS
    Professional Paper 411-1.
                                    227

-------
Ogata, A. and R.B. Banks. 19
    Longitudinal Dispersion
    411-A.

Rifai, M.N.E., W.J. Kaufman
    Laminar Flow through For
    Sanitary Engineering Res
    Berkeley.
jl.   A Solution of the Differential Equation of
 n Porous Media.  USGS Professional Paper
and O.K. Todd.  1956.  Dispersion Phenomena in
 us Media.  Report No. 3IER, Series 90,
earch Laboratory, University of California,
Robson, S.  1978.  Applicati
    Groundwater Solute Transport
    Paper 2050.
Schville, F. 1981.  Groundwa
    Immiscible with Water.
    International Symposium,
 n of Digital Profile Modeling Techniques to
     at Barstow, CA. U.S.G.S. Water Supply
 er Pollution in Porous Media by Fluids
 n:  Quality of Groundwater, Proceedings of an
 Noordwijkehout, The Netherlands, p. 451-463.
                                    228

-------
                          SECTION 9.  LOCUS NO. 9

                                 CONTENTS
                                                                   Page No.

List of Tables	      230
List of Figures	      230
9.1  Locus Description	      231

     9.1.1  Short Definition	      231
     9.1.2  Expanded Definition and Comments	      231

9.2  Evaluation of Criteria for Remediation	      231

     9.2.1  Introduction	      231
     9.2.2  Mobilization/Remobilization	      235

            9.2.2.1  Partitioning onto Mobile Phase	      235
            9.2.2.2  Transport with Mobile Phase	      238

     9.2.3  Fixation	      241

            9.2.3.1  Partitioning onto Immobile (Stationary)
                       Phase	      241

     9.2.4  Transformation	      242

            9.2.4.1  Biodegradation	      242
            9.2.4.2  Chemical Oxidation		      242

9.3  Storage Capaci ty in Locus	      242

     9.3.1  Introduction and Basic Equations	      242
     9.3.2  Guidance on Inputs for, and Calculation of,
              Maximum Value	      243
     9.3.3  Guidance on Inputs for, and Calculation of,
              Average Values	      243

9.4  Example Calculations	      243

     9.4.1  Storage Capacity Calculations	      243

            9.4.1.1  Maximum Value Calculations	      243
            9.4.1.2  Average Value Calculations	      244
                                    229

-------
                                (Continued)


                                                                   Page No.

9.5  Summary of Relative Importance of Locus	      244

     9.5.1  Remediation	      244
     9.5.2  Loci Interaction	      244
     9.5.3  Information Gaps	      245

9.6  Literature Cited	      245


                                  TABLES


9-1  Relative Sizes of Particles                                      234

9-2  Factors Influencing Fixation of Colloidal Particles              241


                                  FIGURES
9-1  Schematic Cross-Sectional Diagram of Locus No. 9 -
     Contaminants Sorbed to Colloidal Particles in Water in
     Either the Unsaturated or Saturated Zone                         232

9-2  Schematic Representation of Important Transformation and
     Transport Processes Affecting Other Loci                         233

9-3  Continuum of Particulate and Dissolved Organic Carbon in
     Natural Waters                                                   234

9-4  Typical Distribution of Dissolved Organic Carbon in
     Interstitial Water of Soils                                      238

9-5  Increase in Mobility of Contaminants Versus Relative
     Concentration of Colloidal Particles                             239

9-6  Mobility of a Hydrophobic Compound Relative to the Mobility
     of the Same Compound without the Presence of a Macromolecule
     as a Function of Octanol-Water Partition Coefficient and
     Amount of Organic Carbon in the Mobile Phase                     240
                                    230

-------
                          SECTION 9 - LOCUS NO. 9
9.1  Locus Description

9.1.1  Short Definition

    Contaminants sorbed onto colloidal particles in water in either the
unsaturated or saturated zone.

9.1.2  Expanded Definition and Comments

    Hydrophobic compounds in solution, which may preferentially sorb to the
stationary soil matrix, may also sorb to "free" colloidal particles in
water.  This locus is recognized as being potentially important as it
allows for the mobilization of strongly-sorbed contaminants that would
otherwise remain immobile due to sorption on a stationary phase.  The
definition includes liquid contaminants "attached" to colloidal particles
rather than just "sorbed".  Water in both the saturated and unsaturated
zones is included with regard to transport on colloids.

    Research on this locus is somewhat limited.  However, recent studies
suggest that the process of deposition of dissolved contaminants onto the
mobile colloidal phase is similar to that for locus 4 (sorption onto the
stationary soil phase).  Figures 9-1 and 9-2 present a schematic
cross-sectional diagram of locus no. 9 and a schematic representation of
the transformation and transport processes affecting other loci,
respectively.

9.2  Evaluation of Criteria for Remediation

9.2.1  Introduction

    Colloidal particles form a homogeneous continuous phase dispersed in
water.  They are distinguished from truly dissolved materials of ionic or
molecular size and larger suspended materials which are influenced by
gravity.   Colloids are electrically charged particles (usually negatively
charged)  that may be comprised of small solid particles, macromolecules,
small droplets of liquids, or small gas bubbles.  Because colloidal
particles are very small (0.45 to 0.001 urn), gravity is an insignificant
force acting on them when compared with the electrical forces acting on the
surface of the particles.  Table 9-1 compares the size of colloidal
particles to those of other particles.  Figure 9-3 demonstrates the general
colloidal particles sizo range in groundwater systems.  Note that
colloidal-sized particles can be "free" or "bound", and the colloidal phase
discussed herein is associated only with free particles.  Soils such as
clays may be comprised to a large degree by immobile, or bound,
colloidal-sized particles, but these are not associated with locus 9.
                                    231

-------
      UNSATURATED
         ZONE
WATER
TABLE
       SATURATED
         ZONE
      NOTE: NOT ALL PHASE BOUNDARIES ARE SHOWN.
                                      LEGEND
                   CONTAMINANT VAPORS

                   AIR (PLAIN WHITE AREAS)
^; WATER IN SATURATED ZONE

 X  BIOTA (MICROBIOTA INCLUDED)
              O    WATER (MOBILE PORE WATER    d*> SOIL PARTICLE
                   IN UNSATURATED ZONE)
                                             |   COLLOIDAL PARTICLE
            AAAA  WATER FILM
                                            •—  CONTAMINANTS
  Figure 9-1.   Schematic Cross-Sectional Diagram of Locus No.9 -
               Contaminants Sorbed to Colloidal Particles in Water
               in either the Unsaturated or Saturated Zone.
                                  232

-------
TRANSPORT WITH GROUNDWATER OR
MOBILE PORE WATER EITHER BY THEMSELVES
OR ATTACHED TO SOIL PARTICLES
     ADHESION OF COLLODIAL PARTICLE
     TO STATIONARY SOIL PARTICLES
             ADVECTION
             DISPERSION
             DIFFUSION
                                 LOCUS NO.  9
                             (CONTAMINANTS SORBED TO
                              COLLODIAL PARTICLES IN
                              WATER IN SATURATED OR
                                UNSATURATED ZONE)
                 DEPURATION
              (BIOTA)
           • LOCUS NO
         11-SORBED TO BIOTA
         SESORPTION /
         ISSOLUTION
PHASE SEPARATION
SORPTION
      (WATER)
      LOCUS NO.
 3 - DISSOLVED IN WATER
  FILM
 8 - DISSOLVED IN GROUND-
  WATER
12-DISSOLVED IN MOBILE
  PORE WATER
         Figure 9-2. Schematic Representation of Important Transformation
                     and Transport Processes Affecting Other Loci.
                                               233

-------
                                  TABLE  9-1

                         RELATIVE SIZES OF PARTICLES
                    Particle
  Approximate Diameter
         Atoms
         Molecules
         Molecular groups
         Colloidal particles
         Microscopically visible particles
         Colloidal clay
         Clay
         Silt
         Very  fine sand
         Fine  sand
          0.1 to 0.6  my
          0.2 to 5.0  my
          0.5 to  10  my
        0.001 to  10  my
                >250  y
                <0.2  y
                  <2  y
      0.05 to 0.002  mm
        0.10 to 0.05  mm
        0.25 to 0.10  mm
                                           _Q
                                           "
                                _          _
         a.  1A (Angstrom)  7  10   mm  =  10"  cm; 1  my (millimicron)
           =  10   mm =  10"  cm; 1 y  (micron) = 10"   mm = 10"   cm

         Source:   Stevenson,  1982.


                 Organic Carbon  Continuum
                10"
                        10'6      10'6      10'7      10''
          10'9      10'10 meters
            Zooplankton
                                                            FA
              Phytoplankton
      Fulvic
       Acid
                           Bacteria
   Humic i
    Acid j
                                     Viruses
                                 Clay-humtc-matal complaxasjr • ;•
        Hydropflilic t
         Acidt i
                                   ,
                POC
DOC
                                  10*   10'   10*  10s   10*  103   10*   IO1
                                                             Molecular Weight

                                   —	General colloid range	»-i



         FA I Fatty acidi     'CHOI Carbohydraln   AA '  Ammo Acids     HC I Hydrocarbons
Source: Thurman, 1985. (Copyright Prentice Hall, Inc. Reprinted with permission.)

Figure 9-3. Continuum of Particulate and Dissolved Organic Carbon in Natural Wtaers.

                                       234

-------
    There are two general classes of colloidal particles:  1) hydrophobic
colloidal particles, which repel water (i.e., clay particles) and 2)
hydrophilic colloidal particles which have affinity for water (i.e., humic
and fulvic substances).  Organic colloidal particles like humic and fulvic
substances are present in soils, sediments, and aquifer materials.  Organic
matter plays a large role in adsorption of dissolved contaminants to a
solid phase.  Humic substances constitute 40 to 60 percent of natural
dissolved organic carbon (DOC) and are the largest fraction of organic
matter in natural waters (Thurman, 1985).  Inorganic colloidal particles
include clays, metal oxides, and inorganic precipitates in the submicron
size range.

    Colloidal particles of interest here are mobile with water.  They can
adsorb organic and inorganic contaminants which have been dissolved in
water and enhance their mobility.  According to Nightingale and Bianchi
(1977), colloidal clay particles, which were mobilized from the surface
soils, were the cause of turbidity in wells several hundred meters distant
from the recharge site.

    Corrective action mechanisms that may immobilize colloidal particles by
transformation from aqueous phase to solid phase are:

    o  Decrease in groundwater velocity which would enhance the attachment
       process of colloidal particles onto soil particles.

    o  Increase in ionic strength in groundwater.  As ionic strength
       increases, the precipitation of organic acids increases.

    o  Decrease in pH of groundwater, which causes precipitation of
       colloidal humic substances.

    Conversely, the preceding mechanisms may be reversed to facilitate
continued mobilization of contaminants in locus 9.  The following
discussions present information about the factors affecting the
mobilization, fixation and transformation of contaminants in locus 9.

9.2.2  Hobilization/Remobilization

9.2.2.1  Partionining onto Mobile Phase

    Contaminants are tound in locus 9 when dissovled contaminants sorb to
free colloidal particles.  The contaminants remain in a mobile phase with
the colloid.  Immobilization occurs through desorption back to the
dissolved phase (locus 8) and subsequent re-sorption to the immobile soil
matrix (locus 4) or through immobilization of the colloidal particle
itself.

    As discussed in Section 9.2.1, there are two types of colloidal
particles, organics and inorganics.  Clays are inorganic colloidal
particles which have layered structures of silicates and metal hydroxides,
a planar geometry, and very large surface area with negative charges.
Organic colloidal particles such as humic and fulvic substances are
hydrophobic, organophilic,  and have high cation exchange capacities because
of negative charges at their surfaces.
                                    235

-------
    In soils which contain more than 0.1 percent (by weight) of organic
matter, sorption of neutral organics is controlled by this organic
fraction; otherwise, sorption is slight and is controlled primarily by
surface area of the soil particles.  Because of some similarities between
soils and colloidal particles, the processes involved in water:soil
partitioning may be similar to those of water: colloid partitioning.

    Colloidal particles may adsorb contaminants as a result of one of the
following processes:

    o  Physical adsorption, which is related to Van-der Vaals forces and is
       caused by a weak electrostatic interaction between atoms and
       molecules of the contaminants and the colloidal particles.

    o  Chemisorption, which involves direct formation of a chemical bond
       between the adsorbed contaminant molecule and the colloidal
       particle.

    o  Exchange adsorption, which occurs when contaminant ions concentrate
       at the surface of the colloidal particles as a result of
       electrostatic attraction between the contaminant and the negative
       surface of the colloidal particle.

    Without considering water:colloid partitioning of contaminants,
water: soil partitioning models may overestimate the mass of immobilized
contaminants.  Water: soil partitioning is often approximated by the
following retardation equation:


                             R ' l + Kd pb/G                          (9.1)

where    R = retardation of dissolved contaminant plume

        K, = partition coefficient for soil and water (L/kg)

        p,  = bulk density of solid soil (g/cm3)

         6 = soil porosoity (dimensionless)


    Water: soil partitioning of hydrophobic contaminants is covered in
detail in Section 4.

    Hutchins et al. (1985) presents a modified retardation equation based
on water:soil:colloid partitioning of organic solutes:
                                                                      <9'2)
where

    K  = colloid:water partition coefficient (L/kg) ; and other parameter
     "   as defined in equation (1).
                                    236

-------
    The modified equation accounts for a decreased proportion of
immobilized contaminant due to sorption onto mobile colloidal particles.
Experimental data indicate that the modified retardation equation provides
a reasonable estimation of this facilitated transport (Kan and Mason,
1986).

    Typical values for soil density and porosity can be found in Section 4.
The soil:water partition coefficient can be estimated from:

                             K, = K   f                               (9.3)
                              d    oc  oc                             v    '

where   K   = organic carbon partition coefficient; and

        f   = fractional organic content of soil by weight.


The colloid:water partition coefficient (K ) can be estimated by:

                             K  = KOC (DOC) 10"6                      (9.4)

where   K   = organic carbon partition coefficient; and

        DOC = dissolved organic carbon content of water (mg/L).


    Equations 9.3 and 9.4 point out the importance of organic carbon
content in sorption of dissolved contaminants.  Other key parameters
affecting sorption were presented in Table 4-3.  Although sorption on to
colloidal particles is less well understood than sorption on to soil, many
of the same parameters are likely to be involved.

    The distribution of dissolved organic carbon (DOC) in the subsurface
environment is especially important for locus 9.  Figure 9-4 shows a
typical distribution of DOC in interstitial waters of soils.  DOC content
is usually much higher in the pore water of the unsaturated zone,
particularly near the surface.  This suggests that facilitated transport of
contaminants in locus 9 may be significantly greater vertically downward in
mobile pore water of the unsaturated zone than in groundwater.

    Among the factors that may be of importance in mobilizing/immobilizing
locus 9 contaminants are temperature, salinity, and pH of the solution.  As
temperature increases, contaminant solubility generally increases, and the
degree of partitioning on to colloidal matter is likely to decrease.
However, control of mobilization by controlling temperature may be of
limited practical value.

    In general, waters low in specific conductivity will have lower
concentrations of DOC, and therefore a smaller fraction of locus 9
contaminants than highly conductive waters.  Increased specific
conductivity, associated with increased salinity, has been shown to cause
precipitation of DOC.  Sholkovitz (1976) found a sudden removal of DOC at a
salinity level of 0.5 to 1.5 percent.  The maximum removal of DOC from
solution appears to occur between 1.5 and 2.0 percent salinity.   Humic
substances, the major fraction of colloidal organic matter in natural
waters, are particularly subject to precipitation when salinity increases
                                    237

-------
           tThen, DOC complexes Fe and Al
            'Clays  sorb 'I'M-O^ \>\
            '.;-. •••.(..,-.:iv-'i;(r  \:
            DOC-metal complexes j (
                                              0   5   10  15  20
                                                 DOC, mg/l
Source: Thurman, 1985. (Copyright Prentice Hall, Inc. Reprinted with permission.)

Figure 9-4.  Typical Distribution of Dissolved Organic Carbon in Interstitial Water of Soils.
 are encountered.  Artificially raising  or  lowering the salinity and/or
 conductivity of subsurface waters may help control mobility of locus 9
 contaminants.

     Sorption of organic acids, such as  aquatic  humic substances, is
 influenced by pH.  For organic acids to adsorb  completely,  the pH of the
 solution should be 2 pH units below the pKa,  the ionization constant for a
 weak acid (Thurman, 1985).  Adsorption  of  organic acids is  greatest at pH 2
 or less, and desorption is greatest at  pH  6 or  more.  Increasing the
 acidity of groundwater or pore water may result in immobilization of both
 colloidal humic substances and attached contaminants.   Conversely,
 increasing alkalinity may result in increased mobilization  of locus 9
 contaminants.

 9.2.2.2  Transport with Mobile Phase

     Partitioning of hydrophobic contaminants  from the dissovled phase to
 free colloidal particles as well as stationary  soil has recently been
 demonstrated.  The colloidal matter is  transported with groundwater,
 carrying with it the sorbed contaminants.   Enfield (1985) has noted that
 hydrophobic contaminants such as DDT have  been  moved farther in groundwater
                                      238

-------
 than laboratory models would suggest  due  to  increased mobility on colloidal
 particles.

     A fraction of the colloidal particles will  "filter out"  of solution
 during transport, as is discussed in  Section 9.2.3,  but the  remaining free
 colloid moves at roughly the average  velocity of groundwater.   In some
 cases, the average free colloid velocity may be greater than that of
 groundwater.  Enfield et al. (1989) estimated that  the velocity of
 colloidal matter was 10 to 16 percent greater than  groundwater during
 column studies using fine sand.  They concluded that  the  relatively
 large-sized colloidal particles were  excluded from  the smaller pore spaces
 in the soil, reducing the effective void volume that  can  be  occupied by the
 colloid.

     The degree to which contaminant mobility is enhanced  via locus 9 is
 influenced in large part by the sorptive properties of the contaminant,  the
 concentration of free colloidal matter, and  the organic carbon content of
 the colloidal matter.  The average velocity  of highly  sorptive compounds
 may be increased significantly in the presence of waters  with  high DOC
 content in colloidal form.   For less hydrophobic compounds,  transport of
 contaminants via locus 9 is generally only a  small  fraction  of the total
 mobile contaminant.   Figure 9-5 demonstrates  that the  presence of 1 and 5
 mg/1 of organic colloidal particles in water  increases  mobility  of
 contaminant by 10 and 50 times respectively;  a significant increase but
 still several orders of magnitude less mobile than conservative  compounds.
                  HOC
                                                                  Relative
                                                                  Concentration
                                                                  (C/Co)
           -6
                              Log Distance
      Source: McCarthy and Wobber, 1986.
      Figure 9-5.  Increase in Mobility of Contaminants Versus Relative
                Concentration of Colloidal Particles.
    Figure 9-6 shows  the  relative  change  in  mobility of hydrophobic
contaminants due to  transport with colloidal particles.   As the figure
shows, the importance of  colloidal particles on  mobility of contaminants
diminishes as the octanol-water partition coefficient (K  ) decreases.
                                    239

-------
                                                     o
                                                     u
                                         •
                                         u
















£
z
o
x

*
~
c














"T"
g
U
0
E
o
k»
~~
E
0
0

•
•
a
e

^
u
E
£
U
O
*
0
0
f
e

a
a






""""^ 1000

.
'a
u
0

0
w
u
E
3 100 -
O
£

_
•
U
i
•
u
0
£ 10
o
7
*
e
a

«
1 '
S
* 5 2 * K"

5l I I a «
ent- o < o x N
ii i i iii
macromol«cul«

concentration (mg/L)
	 10
	 100 " /


	 10000 /
t
//
r
1
/
/ ,•
/ /
f /
' /
/
/
/
t •' '
.' '
/ •' /'
/ / /
• / '
/ /' '
,' ,••' ,'
' .-•' f'

t 4 e 8
109 kow
Source: Enfield and Bengtsson, 1988. (Copyright Water Well Journal Publishing Co.
       Reproduced with permission.)

Figure 9-6.  Mobility of a Hydrophobic Compound Relative to the Mobility of the
            Same Compound without the Presence of a Macrornolecule as a
            Function of Octanol-Water Partition Coefficient and Amount of
            Organic Carbon in the Mobile Phase.
                                240

-------
Higher K   generally implies a greater preference for the sorbed rather
than dissolved phase.  Figure 9-6 shows that for colloidal concentrations
less than 100 mg/L, typical for most groundwaters, only compounds with K
greater than 10,000 are significantly affected by this mechanism.       OW

9.2.3  Fixation

9.2.3.1  Partitioning onto Immobile (Stationary) Phase

    Fixation of locus 9 contaminants can occur through desorption to the
mobile aqueous phase and subsequent re-sorption to the stationary soil
matrix.  There is currently little information available describing the
process of desorption of contaminants from colloidal paraticles.  Results
of experiments by Enfield et al. (1989) suggest that water:colloid
partitioning of organic solutes is a reversible reaction and therefore,
some degree of fixation probably occurs via this process.  However, the
implications of this process on remediation are not well understood.
Colloidal particles have similarities with soil particles; therefore, the
processes of adsorption/desorption of contaminants and soil particles
discussed in Sections 2 and 4 may be applicable to colloidal particles, and
the materials in these sections may provide insight into this fixation
mechanism.

    Locus 9 contaminants may also be fixated through immobilization of the
colloidal particles onto the stationary soil matrix.  The particles can be
"filtered out" of solution during transport as they come into contact with
the soil.  Brownian diffusion, gravitational sedimentation, and
interception cause the particles to contact the soil matrix.  The colloidal
particles may then become attached to the soil particles by electrostatic,
van der Waals, hydrodynamic, and specific chemical interaction forces
between the respective particle surfaces.  Electrostatic forces are the
dominant forces that control the attachment of contaminated colloidal
paraticles to soil particles.  These forces result from isomorphic
substitution within the crystal lattice, surface group ionization, and ion
adsorption.

    Table 9-2 shows some of the factors that influence fixation of
colloidal particles onto immobile soil.  A larger soil porosity and a
larger soil particle size provide a lesser surface area exposure to the
colloid, and would suggest a lower fraction of colloidal attachment to the
soil matrix.  A lower groundwater velocity would,  in general, allow a
greater frequency of contact between colloid and stationary soil, enhancing
immobilization of the colloid.

      TABLE 9-2.  FACTORS INFLUENCING FIXATION OF COLLOIDAL PARTICLES
             Site Parameters	Solution Parameters

           Soil Porosity                Colloidal Particle Size
           Soil Particle Size           Colloidal Concentration
           Groundwater Velocity         pH
                                        Ionic Strength
                                    241

-------
    Larger colloidal particle size and higher colloidal concentration
generally indicate increased contact with stationary soil and, therefore,
greater likelihood of attachment.  Lowering of pH causes precipitation of
the humic acid fraction of the colloid onto immobile soil.  Ionic strength,
which can be indirectly measured by specific conductivity or salinity, also
influences colloidal fixation.  A sudden increase in the ionic strength of
water can cause precipitation of colloidal matter.

    Of the parameters in Table 9-2, groundwater velocity, pH, and ionic
strength are most easily controlled artificially, and therefore, of
greatest interest in remediation.  Decreasing groundwater velocity can
result in lower mobility of contaminants in the free colloidal phase,
however, the relative effect of this mechanism on overall contaminant
mobility may be small.  On the other hand, increased groundwater velocity
could cause increased mobility of locus 9 contaminants under high hydraulic
gradient conditions (i.e., groundwater pumping).  The effect of pH and
ionic strength on colloidal fixation are discussed in greater detail in
Section 9.2.2.

9.2.4  Transformation

9.2.4.1  Biodegradation (Details in Section 11)

    By comparison, the conditions for biodegradation in locus no. 9 closely
resembled those of locus no. 4; the sorbed contaminants are in equilibrium
with dissolved aqueous phase contaminants.  Since nearly all microbial
transformation likely occurs within the cell, the contaminants must first
desorb to the liquid phase (loci nos. 8 and 12) prior to biodegradation.
In the vadose zone, oxygen is much more available than below the water
table, and biodegradation is therefore much more likely to occur there.

9.2.4.2  Chemical Oxidation (Details in Section 3.2.4.2)

    Colloids serve as the major mechanism by which abiotic transformations
occur.  Chemical oxidation of contaminants is mediated by reduced
clay-metal colloids and polymerization occurs as organic colloids are
joined.

9.3  Storage Capacity in Locus

9.3.1  Introduction and Basic Equations

    The basic equation for calculating the storage capacity in locus no. 9
is as follows:
                                                                      (9.5)
Where:  C  « concentration of contaminant on colloidal particles (ug/kg)
        K  = partition coefficient (L/kg)
        C  = concentration of contaminant in aqueous phase (ug/L)
                                    242

-------
9.3.2  Guidance on Inputs for, and Calculation of, Maximum Value

    o  The partition coefficient, K  , can be estimated by equation 9.4:

       Kp = Koc (DOC) 1CT6

    o  DOC concentrations for groundwater range from 0.2-17 Mg/L  (Thurman,
       1985).  Use maximum value of  17 yg/L.

    o  K   values for gasoline constituents can be found in Table 4-4.  For
       bulk gasoline, use K   = 191  L/kg, as discussed in Section 4.4.1.

    o  The pure water solubility of  a compound will generally be higher
       than found in situ.  Use solubilities of gasoline constituents in
       gasoline-saturated water found in Table 8-4.  Use bulk gasoline
       maximum solubility of 200 mg/L as discussed in Section 8.3.

9.3.3  Guidance on Inputs for, and Calculation of, Average Value

    o  Determine K  as discussed in  Section 9.3.2.

    o  Use average DOC concentration of 0.7 ug/L (Thurman, 1985).

    o  Use K   values as discussed in Section 9.3.2.
            oc

    o  Use average C  value of 20 mg/L for gasoline, as discussed in
       Section 8.4.

9.4  Example Caculations

9.4.1  Storage Capacity Calculations

9.4.1.1  Maximum Value Calculations

    o  Storage capacity (C ) for gasoline:

Using equation 9.4 to estimate K :

    K  = 191 (17) 10~6 = 0.0032 L/kg

    Cw = 200 mg/L

Then using equation 9..5:

    C  = (0.0032)(?00) = 0.64 mg/kg

    o  Maximum storage capacity for  benzene:

From Table 4.4, K   for benzene = 62
                 oc

Then for equation 9.4:

    K  = 62 (17) 10~6 = 0.001 L/kg
                                    243

-------
From Table 8-4, gasoline-saturated water solubility of benzene = 64.8 mg/L.

Then using equation 9.5:

    C  = (0.001) (64.8) = 0.065 mg/kg

9.4.1.2  Average Value Calculation

    o  Average storage capacity for gasoline:

    K  * 0.0032 L/kg

    Cy = 20 mg/L

Then using equation 9.5:

    C  = (0.0032) (20) = 0.064 mg/kg

9.5  Summary of Relative Importance of Locus

9.5.1  Remediation

    Under most natural conditions, the significance of contaminated
colloidal particles in groundwater contaminant is very low.  Remediation
efforts targeting locus 9 contaminants may be warrranted for highly
hydrophobic compounds (K   > 10,000) and/or for waters very high in DOC (>
10 ug/L).  Under these conditions, contaminant mobility may be greater than
normally expected.

    For the saturated zone, immobilization techniques would likely focus on
enhanced precipitation of free colloidal matter through adjustments in pH
or ionic strength.  Removal of locus 9 contaminants from groundwater by
traditional pumping methods should be effective.

    In the unsaturated zone, enhanced fixation of colloidal particles
through pH or ionic strength adjustments will introduce considerable
amounts of water to the subsurface.  This is likely to facilitate mobility
in this zone rather than retard it, and will generally not be feasible.
Traditional soil venting methods may serve to immobilize colloidal
particles in pore water as water vapors are extracted and the soil "dried  -
out".

9.5.2  Loci Interactions

    Loci 3, 5, 6, 7, 8 and 12 interact with attachment of contaminants onto
colloidal particles in groundwater.  For the contaminated colloidal
particles, the partitioning process into groundwater (aqueous phase) and
soil particles (solid phase) is not clearly defined and not enough
information is available.  Bulk transport into the locus occurs from locus
nos. 3, 5, 6, 7, 8 and 12.

    Currently little is known about the desorption process of contaminants
from the colloidal particles in groundwater; therefore, bulk transport out
                                    244

-------
of locus no. 9 is not well defined.  It may be related to locus nos. 2, 3,
4, 5, 8, 11 and 12.

9.5.3  Information Gaps

    There is limited information available on the subject of colloidal
particle behavior and the influence of these particles on the mobility of
contaminants in the subsurface environment.  Critical areas in which
additional research is needed to understand the role of colloidal particles
in the rate and transport of contaminants are as follows:

    1.  Factors affecting the process of adsorption/desorption of
        contaminants on and off colloidal particles.

    2.  Effects of colloidal particles on the solubility of contaminants in
        groundwater (i.e., do these particles increase solubility).
        Additional qualitative and quantitative data are needed.

    3.  Factors affecting the mass transport of contaminated colloidal
        particles with groundwater.

    4.  Effect of precipitation (filtration) of free colloidal particles on
        the permeability of the soil and the effect of sorbed contaminant
        on precipitation/filtration of free colloidal particles.

    5.  Concentrations, properties and constituents of inorganic colloidal
        particles in the groundwater.

    6.  Importance of organic and inorganic colloidal particles in
        contaminant transport.

    7.  Type of contaminants most influenced by locuS 9 transport.

9.6  Literture Cited

Enfield, C.G.  1985.  Chemical Transport Facilitated by Multiphase Flow
    Systems.  Water Sci. Technol., 17:1-12.

Enfield, C.G. and G. Bengtsson.  1988.  Macromolecular Transport of
    Hydrophobic Contaminants in Aqueous Environments.  Ground Water,
    26(1):64-70.

Enfield, C.G., G. Bengtsson, and R. Lindquist.  1989.  Influence of
    Macromolecules on Chemical Transport.  Environ. Sci. Technol.  Vol. ?1,
    no. 10.

Hutchins, S.R., M.B. Tomson, P.B. Bedient, and C.H. Ward.  1985.  Fate of
    Trace Organics During Land Application of Municipal Wastewater.  CRC
    Critical Reviews of Environmental Control.  Vol. 15. pp. 355-416.

Kan, A.T. and B.T. Mason.  1986.  Facilitated Transport of Naphthalene and
    Phenanthrene in a Sandy Soil Column with Dissolved Organic Matter.
    Proceedings of the NWWA/API Conference on Petroluem Hydrocarbons and
    Organic Chemicals in Groundwater.  Houston, TX.  pp. 93-106.
                                    245

-------
McCarthy, J.F. and F.J. Wobber.  1986.  Transport of Contaminants in the
    Subsurface:  The Role of Organic and Inorganic Colloidal Particles.
    U.S. Department of Energy.

Nightingale, H.I. and W.C. Bianchi.  1977.  Groundwater Turbidity Resulting
    from Artificial Recharge.  Groundwater, 15(2):  146-152.

Sholkovitz, E.R.  1976.  Flocculation of Dissolved Organic and Inorganic
    Matter During the Mixing of River Water and Seawater.  Geochemica et
    Cosmochemica Acta.  40:831-845.

Stevenson, F.J.  1982.  Humus Chemistry.  John Wiley & Sons, New York.

Thurman, E.M.  1985.  Organic Geochemistry of Natural Waters.  M. Nijhoff
    and W. Junk Publishers, Boston,  pp. 1-17.
                                    246

-------
                         SECTION 10.  LOCUS NO. 10

                                  CONTENTS
                                                                    Page No.
List of Tables	    248
List of Figures	    248
10.1  Locus Description	    250

      10.1.1  Short Definition. ..7	    250
      10.1.2  Expanded Definition and Comments	    250

10.2  Evaluation of Criteria for Remediation	    250

      10.2.1  Introduction	    250

              10.2.1.1  Description of Pore Spaces	    250
              10.2.1.2  Description of Contaminant Phases	    253
              10.2.1.3  Summary of Corrective Action Potential...    253

      10.2.2  Mobilization/Remobilization	    253

              10.2.2.1  Partitioning onto Mobile Phase	    253

                        Diffusion	    256
                        Diffusion Between Primary and Secondary
                          Porosity	    256
                        Diffusion Between Strata of Different
                          Hydraulic Conductivities	    260
                        Diffusion in the Unsaturated Zone..:	    262
                        Sorption	    263
                        Advection	    263

              10.2.2.2  Transport of/with Mobile Phase	    264

      10.2.3  Fixation	    264

              10.2.3.1  Partitioning onto Immobile (Stationary)
                          Phase	    264

      10.2.4  Transformation	    264

              10.2.4.1  Biodegradation	   264
              10.2.4.2  Chemical Oxidation	   265
                                     247

-------
                                (Continued)


                                                                   Page No.

10.3  Storage Capacity in Locus	    265

      10.3.1  Introduction and Basic Equations	    265
      10.3.2  Guidance on Inputs for, and Calculation of,
                Maximum Values	    265

              10.3.2.1  Concentration of Contaminants in Secondary
                          Pore Water	    265
              10.3.2.2  Total Primary Porosity	    265

      10.3.3  Guidance on Inputs for, and Calculation of,
                Average Values	    265

              10.3.3.1  Concentration of Contaminants in Secondary
                          Pore Water	    265
              10.3.3.2  Total Primary Porosity	    266

10.4  Example Calculations	    266

      10.4.1  Storage Capacity Calculations	    266

              10.4.1.1  Maximum Values	    266
              10.4.1.2  Average Values	    266

      10.4.2  Transport Rate Calculations	    266

10.5  Summary of Relative Importance of Locus	    267

      10.5.1  Remediation	    267
      10.5.2  Loci Interactions	    268
      10.5.3  Information Gaps	    268

10.6  Literature Cited	    268


                                   TABLES


10-1  Typical Primary Porosities                                     259

10-2  Typical Effective Diffusivities for Nonsorbing Species         259


                                  FIGURES
10-1  Schematic Cross-Sectional Diagram of Locus No. 10 -
      Contaminants  that have Diffused into Mineral Grains or
      Rocks  in either  the Unsaturated or Saturated Zone              251
                                     248

-------
                                (Continued)


                                                                   Page No.

10-2  Schematic Representation of Important Transformation and
      Transport Processes Affecting other Loci                       252

10-3  Schematic Diagram of Rock Matrix                               254

10-4  Schematic Diagram of Diffusion Across a Fissure-Matrix
      Boundary                                                       255

10-5  Physical Representation of Equation 10.1 for Mass Diffusing
      Through a Rock Slab                                            258

10-6  Comparison Between Measurements and Simulations for
      Low-Velocity Case                                              261

10-7  Comparison Between Measurements and Simulations for
      High-Velocity Case                                             261

10-8  Diagram of Diffusion Between Strata                            262
                                     249

-------
                         SECTION 10 - LOCUS NO. 10
10.1  Locus Description

10.1.1  Short Definition

    Contaminants that have diffused into mineral grains or rocks in either
the unsaturated or saturated zone.

10.1.2  Expanded Definition and Comments

    This locus focuses on contaminants that have gone beyond surface
sorption (e.g., locus no. 2 and 4) and migrated or diffused into the
inorganic mineral grains or rocks.  While somme of the contaminants might
penetrate intact mineral crystals or amorphous solids by pure diffusion, a
larger fraction is likely to have entered through microfractures,
micropores or the thin spaces between mineral (e.g., clay platlet) layers.
In the latter case, capillary tension would be the major driving force.  In
both cases, the contaminants are very immobile because they are trapped
within a stationary phase and would have to be extracted from the minerals
(a slow process) in many remediation schemes.  Figures 10-1 and 10-2
present a schematic cross-sectional diagram of locus no. 10 and a schematic
representation of the transformation and transport processes affecting
other loci, respectively.

10.2  Evaluation Of Criteria For Remediation

10.2.1 Introduction

    As indicated in the definition of this locus, there are essentially two
types of locations within a rock into which an organic contaminant might
move:  1) intergranular pore spaces, and 2) between intramineral layers.
Contaminants could enter these spaces in the gaseous, aqueous or pure
liquid phases, or as a combination of these.  The migration of these phases
could occur in the unsaturated or the saturated zone according to processes
of diffusion, sorption and advection.

10.2.1.1  Description of Pore Spaces

    The pore spaces addressed in this discussion encompass the primary
porosity of a rock, or those intergranular and intramineral spaces which
were created during the rock formation process.  Diameters of these pore
spaces may range from many angstroms (10~ m) to a few microns n<>~ m).
Secondary porosity, or spaces formed in fracturing or geochemical weather-
ing, is addressed in locus no. 13, yet there is some relationship between
primary and secondary porosity within the context of diffusion into the
rock matrix:  higher levels of secondary porosity increase the accessibi-
lity of primary pore spaces.  Thus, secondary porosity is discussed here
inasmuch as it relates to primary porosity, while transport and remediation
within the secondary porosity is  discussed in locus no. 13.

    The minerals of primary concern regarding intramineral contaminant
migration are the phyllosilicates, especially micas such as biotite and the
                                    250

-------
     UNSATURATED
        ZONE
WATER
TABLE
      SATURATED
        ZONE
     NOTE: NOT ALL PHASE BOUNDARIES ARE SHOWN.
                                      LEGEND
CONTAMINANT VAPORS

AIR (PLAIN WHITE AREAS IN
UNSATURATED ZONE)

WATER (MOBILE PORE WATER
IN UNSATURATED ZONE)

WATER FILM
                                          ;>^ WATER IN SATURATED ZONE

                                            X   BIOTA (MICROBIOTA INCLUDED)

                                           CD SOIL PARTICLE

                                            I   COLLOIDAL PARTICLE

                                                CONTAMINANTS
   Figure 10-1. Schematic Cross-Sectional Diagram of Locus No. 10 -
                Contaminants That Have Diffused Into Mineral Grains or
                Rocks in Either the Unsaturated or Saturated Zone.
                                  251

-------
    (WATER)
   LOCUS NO
 3 - DISSOLVED IN
   WATER FILM
 8 - DISSOLVED IN
   GROUNDWATER
12- DISSOLVED IN
   MOBILE PORE
   WATER
PHASE
SEPARATION

DIFFUSION^

DISSOLUTION
                                        (SOIL GAS)
                                        LOCUS NO
                                  1 - CONTAMINANT VAPORS
                            VOLATILIZATION
                              CONDENSATION
                              DIFFUSION
  LOCUS  NO. 10
(CONTAMINANTS .DIFFUSED
 INTO MINERAL GRAINS OR
 ROCKS IN UNSATURATED
  OR SATURATED ZONE)
                                DEPURATION
                                DIFFUSION
DIFFUSION
 (LIQUID CONTAMINANTS)
      LOCUS NO
2 - ADHERING TO -WATER-
  DRY" PARTICLES IN UN-
  SATUFIATED ZONE
5-IN PORE SPACES IN
  SATURATED ZONE
13 - IN ROCK FRACTURES
                                                           .SORPTION
                                                           DIFFUSION
                                                          DESCRIPTION
                                          (BIOTA)
                                        LOCUS NO.
                                    11-SORBEDTOBIOTA
                                                       (SORBED CONTAMINANTS)
                                                            LOCUS NO
                                                     4 - SORBED TO "WATER-WET
                                                       SOIL PARTICLES
            Figure 10-2. Schematic Representation of Important Transformation
                         and Transport Processes Affecting Other Loci.
                                                252

-------
clay minerals, including kaolinite, illite, vermiculites, and
montmorillonite.  Phyllosilicates have the special feature that they are
composed of tetrahedral and octahedral silicate layers, which allow ions,
water molecules, and, potentially, small organic molecules to move between
those layers.  Although the concentration of these minerals is usually low
in comparison to the one or two major constituents of a rock, their
presence is occasionally important in some rocks as they are the only
minerals which might sorb or otherwise take up enough of the diffused
contaminant to have ramifications for corrective action.  Sorption to the
surfaces of mineral grains within a rock occurs much like the sorption
processes discussed in locus no. 4.

10.2.1.2  Description of Contaminant Phases

    The contaminant phase that will be addressed in this section is the
aqueous phase.  Although multi-phase contaminant/groundvater systems are
likely to occur in the vicinity o€ this locus, such combinations are
extremely difficult to model and will not be considered here.  Furthermore,
information concerning gas phase and liquid phase diffusion into the rock
matrix is very sparse, with most studies, both theoretical and
experimental, focusing on solute diffusion.

10.2.1.3  Summary of Corrective Action Potential

    The three processes which will be discussed are diffusion, sorption and
advection.  As might be expected, each of these processes holds different
implications for remediation.  In a very general sense, the most important
remediation mechanisms involve controlling the concentration gradient and
affecting the sorption potential of the petroleum deriviatives.
Contaminants can be mobilized from the rock matrix by reversing the
concentration gradient, encouraging the chemicals to diffuse out of the
rock.  Such diffusion may be reduced if the minerals composing the rock
matrix have a high capacity to sorb organic chemicals.  On the other hand,
in the absence of advection, if the direction of diffusion can be
maintained, then the contaminant may be contained in a limited area.
Regardless of diffusive processes, minerals with a high sorption capacity
will tend to immoblize the contaminants unless those organiq molecules are
displaced by substances with higher affinities for the minerals, in which
case the contaminant may be remoblized.  The discussion that follows will
provide the important details of the processes (mainly diffusion and
sorption) affecting the migration of organic chemicals in the rock matrix
and investigate, where possible, remediation alternatives for this locus.

10.2.2  Mobilization/Remobilization

10.2.2.1  Partitioning onto Mobile Phase

    The most important partitioning mechanisms that could move contaminants
from the rock matrix to a mobile phase are diffusion and desorption.
(Contaminants being transported by advection are already in the mobile
phase).  Figure 10-3 shows an enlarged portion of Figure 10-1 in order to
clarify what the consolidated matrix looks like, and Figure 10-4 shows a
                                    253

-------
      KEY:
               matrix grain

               contaminant
               vapor phase
            X  bacteria
             ^ mobile water in unsaturated zone

               water film
                                                           Unsaturated
                                                               Zone

                                                            Saturated
                                                              Zone
                                                     See Figure 10-4
                                                     for detail
            *+*' mobile water in saturated zone

            I colloid  particles
Figure 10-3. Schematic Diagram of Rock Matrix
                                254

-------
                                                                    Unsaturated
                                                                       Zone
                                                                     Saturated
                                                                       Zone
      LEGEND
             matrix grain
             aqueous phase
             diffusing molecules
      • • •  sorbed contaminant
      xxx  bacteria
               id Phase contaminant
Figure 10-4.  Schematic Diagram of Diffusion Across a Fissure-Matrix Boundary
                                  (Detail from Figure 10-3)
                                       255

-------
schematic representation of the various phases through which a diffusing
chemical might pass as it moves into or out of the matrix.  Within the rock
matrix, the chemical could be in either the gaseous, aqueous or the liquid
contaminant phase.

Diffusion

    Molecular diffusion is one of a number of dispersive processes.  It
occurs as a result of Brownian motion, with net migration of contaminant
molecules from areas of higher to lower concentration.  In the case of a
completely inert contaminant, this process is theoretically reversible so
that a reversed concentration gradient will draw the contaminant back out
of the rock matrix.

    Diffusion differs from most other dispersive processes in that it does
not require advection to occur but rather is driven by concentration
gradients.  In fact, higher advective velocities increase the viability of
other dispersion mechanisms so that the relative importance of diffusion is
lessened.  One study has suggested that for average advective velocities of
less than 1 m/day, diffusion is the dominant transverse dispersion
mechanism (Grane and Gardner, 1961, cited by Sudicky et al., 1985).

    Little work has been published regarding liquid or gas phase diffusion,
so this section will address only the aqueous phase.  Liquid phase and
aqueous phase diffusion are likely to be quite similar.  As shown in Figure
10-4, an organic chemical in the aqueous phase that has diffused into the
rock may diffuse out by moving through several phases.  In the saturated
zone, the diffusing hydrocarbon may move between water-wet or
contaminant-wet matrix grains into the aqueous or the liquid phase, while
in the unsaturated zone, the diffusing hydrocarbon may move additionally
between dry grains and into the vapor phase.  In this figure, diffusion is
depicted as movement into a secondary macropore, such as a fracture,
because the bulk of diffusion into a rock will occur in zones of high
secondary porosity, which provides a relatively high level of available
surface area for diffusion.  At any rate, this discussion will be limited
to aqueous phase diffusion with no attention to the moisture
characteristics of the particles, as there is little theoretical
information and no experimental data available at this level of detail.

Diffusion Between Primary and Secondary Porosity

    The type of diffusion depicted in Figure 10-4 (matrix to secondary pore
channel) can be described by the following equation, derived from Pick's
first and second laws, which describes the mass of chemical expected to
diffuse into the rock matrix from a macropore per unit area (Skagius and
Neretnieks, 1986a and 1986b):

                             CH D          C1 1 a
                       Q = 	—— t -   —	                   (10.1)
                                    256

-------
    where   Q  = mass of diffused substance, at time t, per unit diffusive
                                   2
                 surface area (mg/m )
            1  = distance of diffusion (normal to macropore surface)(m)
                                                       3
            C. = concentration in macropore water (mg/m )

            t  = time (s)

            D  = effective diffusion coefficient, or effective diffusivity
                 (m2/s)

            a  = rock capacity factor (unitless)
    The rock capacity factor measures a rock's capacity to store, instead
of transmit, as a function of both porosity and sorption:

                            a  = 6t + Kdp                            (10.2)


    where   9  = total (primary) porosity (dimensionless)
                                       3
            K, = sorption coefficient(m /kg)

            p  = rock density (kg/m )
    The effective diffusion coefficient measures a rock's capacity to
transmit by diffusion, and is a combination of diffusivity in the primary
porosity and the portion of that porosity that is devoted to transport:

                             Dg = n+D                                (10.3)


    where   D   = effective diffusion coefficient

            0   = transport porosity (dimensionless)
                                               2
            D   = pore diffusion coefficient (m /s)


    It is worth noting that the total porosity may be differentiated into
the effective porosity, including those pores that are interconnected and
serve as conduits, and storage porosity, including those pores that have a
dead end.  With these definitions, the rock capacity factor might better be
defined in terms of the effective porosity instead of the total porosity.
It is also interesting to note that, in the derivation of equation 10.1,
the generic diffusion coefficient of Pick's second law, D, was detined as D
= D /a, with D  and a being independent of one another.  The importance of
this observation is discussed below in the portion of this section that
addresses sorption.

    Figure 10-5 shows a schematic picture of the physical setting where
this equation applies.  This equation applies for diffusion through a rock
slab that has an initial contaminant concentration of zero and an outlet
concentration that is effectively zero.
                                    257

-------
                                            Rock Matrix Slab
                                Distance (L)

                             Direction of advection
              C0- Inlet concentration of contaminant
               C - Concentration at distance, x, from inlet
            Figure 10-5. Physical Representation of Equation 10.1
                        for Mass Diffusing Through a Rock Slab
    Table 10-1 gives  some  examples  of primary porosity values for various
rock types, and Table 10-2 gives  examples of effective diffusion
coefficients.  For sorption  coefficients, discussed in locus no. A, few
data are available.   Rock  densities will vary considerably with mineral
composition, but  they are  relatively simple to determine.  Thus, for a
given macropore concentration  and diffusional length of interest, such as
the distance between  macropores,  the flux rate of an organic chemical into
the rock matrix may be calculated.   The time frame of interest, should be
chosen carefully  because the effective diffusion coefficient has been
assumed to be constant in  time, which may be inaccurate over short periods
of time (Sudicky  et al., 1985).

    Assuming that the petroleum derivatives considered here are nonreative,
then the reverse  process,  diffusion out of the rock matrix, ought to be
described by the  same equation in the saturated zone, where no significant
discontinuity would exist  between the matrix pore water phase and the
fracture aqueous  phase.  The implication for remediation is that
contaminants that have diffused into the rock matrix will re-enter the
mobile phase as natural flushing  or dilution changes the concentration
gradient between  primary and secondary porosity.  Examining the rock
capacity term reveals that the addition of a desorbing agent will increase
                                     258

-------
                                TABLE 10-1

                        TYPICAL PRIMARY POROSITIES
          Rock Type
     Typical Value or Range
        of Porosities (X)
                         Source
granite
gneiss
limestone
sandstone (semiconsolidated)
basalt
limestone
granite
clay-loam till
equal sized spheres
loosest packing
tightest packing
SOURCES: 1. Skagius and Neretnieks,
2. Heath, 1983
3. Garrels et al., 1949
4. Neretnieks, 1980
5. Grisak et al. , 1980
0.02-1.14
0.02-0.74
10
10
10
3.79-34
0.05-0.09
30-35

48
26
1986a




1
1
2
2
2
3
4
5

2
2





                                TABLE 10-2

          TYPICAL EFFECTIVE DIFFUSIVITIES FOR NONSORBING SPECIES
        Rock Type
Typical Value or Range    Diffusing
Diffusivities (cm2/s)	Substance*
                              Source
granite
granite
gneiss
gneiss
unconsolidated sand, silt
granite
gneiss
limestone
granite
clay-loam till

*I = ion
 0 = organic molecule
   4.1-66
 0.22-6.9
   1.8-13
 0.01-7.8
     1.21
   5-36.3
 3.3-28.7
        2
  0.25-10
  1.9-5.0
x 10
x 10
x 10
x 10
x 10
x 10
x 10
x 10
x 10
x 10
-10
-10
-10
-10
-10
-5
-8
-7
i
o
i
i
i
i
i
i
i
i
i
i
2
3
3
4
s
6
SOURCES:  1.  Skagius and Neretnieks, 1986a
          2.  Sudicky et al., 1985
          3.  Skagius and Neretnieks, 1986b
          4.  Garrels, et al., 1949
          5.  Neretnieks, 1980
          6.  Grisak et al., 1980
                                    259

-------
the effectiveness of a reversed concentration gradient, provided that the
desorbing agent is also able to move into the rock matrix.

    The information provided combined with some simple field measurements of
rock bulk densities allow the mass flux rate per unit area to be calculated
readily.  The usefulness of this flux rate is limited, however, by the
ability to determine accurately the fracture extent or advective velocities
within the fractures, because these matrix characteristics contribute to
fracture diffusion and thus, to the fracture mass flux of contaminants (see
Section 13 for a discussion of fracture characterization).  If an order of
magnitude estimate of this contribution is acceptable, then this mass flux
rate may be useful.  However, further difficulties arise in determining
whether, for example, diffusivities determined from this equation for
non-sorbing species may be applied to sorbing species or whether this
analysis may be applied in the unsaturated zone.

Diffusion Between Strata of Different Hydraulic Conductivities

    Diffusion occurs not only between primary and secondary porosity but
also in porous media, especially in stratified regions.  Sudicky et al.
(1985) describe dispersion of a solute from an advective sand layer into
the surrounding, lower conductivity silt layers.  For a seven-day,
continuous input of a nonsorbing tracer (NaCl), diffusion was found to be
the main component of transverse dispersion.  The experiments were not
allowed to run long enough for the effluent concentration to approach zero;
however, some differences were discovered between the high and low velocity
cases.  For a low advective velocity (0.1 m/day) in the sand layer,
approximately 73 percent of the input mass had been recovered in 36 days,
or 23 days after first appearance.  On the other hand, in the high velocity
case (0.5 m/day), approximately 63 percent of the input mass had been
recovered in 12 days, or 10 days after first appearance.  Concentration
profiles for the low and high velocity case are shown in Figures 10-6 and
10-7.  Although the leveling off of the concentration profile in Figure
10-7 suggests that some time might pass before the level of recovery
reached that depicted in Figure 10-6, the high level of recovery up to this
point supports the notion that higher velocities allow less diffusion into
the silt layer.  In addition, the fact that the curve in Figure 10-7 does
level off instead of continuing to go quickly doVnward shows that some of
the material that had diffused into the silt layer subsequently diffused
back out.  Sudicky et al. (1985) assume that this diffusion could be
described by Pick's first law as illustrated in Figure 10-8, (at y = b):

                                 dc'
                        J = 6 D* dy                                    (10.4)

                                                     2
   where   J  = flux across sand-silt interface (kg/m /s)

           0  = silt porosity (dimensionless)
                              2
           D* = diffusivity (m /s)

           c' = concentration of solute in silt (kg/m )

           y  = distance from center of sand layer (m)

           b  = distance from center of sand layer to sand-silt interface  (m)
                                    260

-------
    100
 o
 u
  . 080
 o

 | 0.60
 o
 c
 o
 O
   0.40
 I 020
 K
                            Measured

                            Thick-layer Solution, DfO

                            Thin-layer Solution, Dj*0

                            Thin-layer Solution, DfO

                            No Diffusive Loss
                   8      12
            16     20    24

                 Time (days)
36
             Q. » Longitudinal hydrodynamic dispersion coefficient.



Source: Sudicky et al., 1985.


Figure 10*6.  Comparison Between Measurements and Simulations

              for Low-Velocity Case.
            10
            08
         I  06
         o
         c  04
         o
         u
         •I

         -02
         o
•   Measured

    Thick-layer Solution, Dj«0

    Thin-layer Solutions,
                     2      4      6      8      10      12

                              Time  (days)



                   > Longitudinal hydrodynamic dispersion coefficient.
 Source: Sudicky et al., 1985.

 Figure 10-7.  Comparison Between Measurements and Simulations

               for High-Velocity Case.
                                   261

-------
                                      (t/2) C/C0   SAND
                                                        Direction of advectfon
            Co= Inlet concentration of contaminant
             C = Concentration at distance, y, from center of sand stratum
            Figure 10-8. Diagram of Diffusion Between Strata
    Had the experiments been allowed  to run until  the  relative
concentration reached (nearly) zero,  some simple mass  balance calculations
could have been performed to check  the assumed values  for  porosity (9) and
D*, which are reported here in Tables 10-1 and 10-2.   Values for these
parameters have been determined using unconsolidated media and  should not
be used in reference to consolidated media (rocks), however, the principles
governing their derivation are the  same, and  the values  are provided here
for comparison.

Diffusion in the Unsaturated Zone

    Little is known about diffusion in the unsaturated zone.  Wang and
Narasimhan (1985) have proposed a model in which flow  in an unsaturated,
fractured porous medium takes place mostly within  the  rock matrix, with
fractures being bridged at points where asperities allow the sides of the
fracture to touch and hold a small  amount of  water.  In  fact, it is
expected that any water residing in non-touching areas of  fractures will
have some (small) velocity into the matrix.   The Darcian advective
velocities predicted for various model configurations  are  on the order of
only 10   to 10"  m/day, so that diffusion is the  major  component of solute
movement in the unsaturated zone, according  to this model. As  the authors
themselves point out, however, more work is needed beyond  the study of
fluid flow to the study of solute movement.
                                     262

-------
Sorption (Details in Section 4.2)

    Sorption is  the other main process governing the movement of a
contaminant in the rock matrix.  The process of surface sorption is
discussed in locus no. 4.  The sorption of organic chemicals to mineral
surfaces, and especially to clay minerals, is controlled by van der Waal's
forces and depends upon both the surface area and the solubility of the
sorbing chemical in water.  The sorption process described in Section 4
deals with sorption in soils, yet some of the principles discussed may be
applicable for sorption to rock minerals.  The principles of soil sorption
may not apply to rock mineral sorption, however, because of the potentially
considerable differences between the two loci.

    While diffusion determines whether a solute may move into the matrix,
sorption is largely responsible for the degree to which that contaminant
may continue to move from one matrix block to the next and whether the
contaminant will be able to diffuse back out.  In fact, as indicated in
equations 10.1 and 10.2, the mass diffused depends on not only the
diffusivity but also the sorption coefficient which, together with primary
porosity, determines the rock capacity factor.  The process of sorption in
soils is discussed in detail in Section 4; suffice it to say here that the
principles are likely to be the same for minerals residing in a rock
matrix, and the amount of contaminant which is able to sorb to the minerals
will likely be limited by the amount of contaminant that is able to diffuse
into the mineral grains.  Of the two loci dealing with contaminants sorbed
to or in close proximity with soil particles (4 and 9), locus no. 4 is
probably the more similar physically to locus no. 10, although there are
differences such as the degree of organic humus coating on, and surface
area to volume ratio of, the particles of locus no. 4.

    Clearly, there is a relationship between the degree of sorption and the
diffusivity.  As mentioned above in the discussion of diffusion into the
rock matrix, this relationship is stated in the derivation of equation
10.1, where the generic diffusion coefficient of Pick's second law, D, is
defined as D = D /a (Skagius and Neretnieks, 1986a).  Thus, the generic
diffusion coefficient depends upon both the capacity to transmit, reflected
by D , and the capacity to store, reflected by a.  These two parameters (D
and a) have been considered independent of one another.  Although no
correlation was demonstrated by Skagius and Neretnieks (1986a) for the
nonsorbing species tested, it is conceivable that a relationship does exist
if sorption takes place in such a manner as to remove the sorbed molecules
from the concentration gradient which is "felt" by subsequently diffusing
molecules.   In the context of diffusion between strata, experiments to
examine a diffusivity retardation factor, which is a function of a
solute-solid partitioning coefficient, have been performed (Starr et al.,
1985).  However,  the models thus obtained did not fit observed data, with
discrepancies increasing with decreased velocity in the higher conductivity
stratum.

Advection

    Advection affects this locus both directly and indirectly, but with
less impact than either diffusion or sorption.  As indicated in the
discussion of diffusion, the degree of diffusive flux of a contaminant into
                                    263

-------
the rock matrix is largely dependent upon the degree of secondary porosity.
Thus, the advective transport of a contaminant into the secondary porosity
(fractures and dissolution channels) is indirectly important to this locus.
Advective transport in secondary porosity is discussed in Section 13.  On
the other hand, contaminant transport in an unfractured, consolidated,
porous media is governed by an interplay of two processes: longitudinal (in
the direction of flow) advective transport in higher conductivity strata,
and tranverse diffusion into the surrounding lower conductivity strata.

    In cases of consolidated and unfractured media, the modeling of
advective transport is essentially the same as modeling in unconsolidated
media, except that the actual values of the horizontal and vertical
conductivity parameters will be different.  Thus, in this unfractured case,
the discussion of advective transport given in Section 8 is applicable here
as well.  The difficulty lies in the fact that there is a continuum between
"fractured" and "unfractured" consolidated media.  The level of fracturing
for which unconsolidated media models are still accurate is not well
defined.  For the purpose of this locus, advection is only discussed
qualitatively (in Section 10.2.2.1) as it directly affects diffusion
processes.

10.2.2.2  Transport of/with Mobile Phase

    Once the contaminant moves into the mobile phase, which occurs mainly
in high conductivity strata or fractures in the saturated zone and is
restricted by sorption, it may be transported with that phase.  Transport
is discussed in a number of other potentially applicable loci including
loci 13 and 8.

10.2.3  Fixation

10.2.3.1  Partitioning onto Immobile (Stationary) Phase

    The most important mechanism that could immobilize contaminants in the
rock matrix is sorption.  Although diffusion may be very slow, it will
continue to move the contaminant for as long as a concentration gradient
exists.  Furthermore, as seen in the discussion of diffusion in stratified
porous media, diffusion may significantly retard but not halt the advective
movement of a non-sorbing contaminant.  Thus, diffusion followed by
sorption within the rock matrix will be the most significant retardation
mechanism.  Once again, the details of surface sorption are discussed  in
Section 4.2.3.

10.2.4  Transformation

10.2.4.1  Biodegradation (Details in Section 11)

    Bacteria are the primary agents of biodegradation in  the subsurface
environment.  Since they range in size from approximately 1 to 10 microns,
they are excluded from any pore or crack with a smaller diameter of width.
It is therefore unlikely that significant biodegradation occurs in the
interior of minerals or rocks.  Transport of necessary substrates such as
oxygen and mineral nitrogen  to the microbes may also be severely limiting.
                                    264

-------
10.2.4.2  Chemical Oxidation (Details in Section 3.2.4.2)

    Chemical oxidation is not expected to occur within crystalline mineral
lattices.  No data are, however, available.

10.3  Storage Capacity in Locus

10.3.1  Introduction and Basic Equations

    The mass of contaminant per unit volume of unfractured rock is simply
the following:
                                                                    (10.5)
    where   m  = mass of contaminant per unit volume of rock (mg/L)
            C  - concentration of ^aqueous phase contaminant in secondary

                 pore (fractures and solution channels) waters (mg/L)

            0. = total primary porosity of rock (dimensionless)


    Some values of 6  for both crystalline and porous media were given in
Table 10-2.  The total primary porosity includes both effective and storage
porosity.

10.3.2  Guidance on Inputs for, and Calculation of, MaxiBUB Values

10.3.2.1  Concentration of Contaminants in Secondary Pore Water, C

    As indicated in Section 3.3.2, the maximum concentration may be taken
as the contaminant solubility in water.  The hydrocarbons of a typical
gasoline without additives have an approximate solubility of 200 mg/L at
20°C.  A gasoline containing oxygenated additives, such as ethanol or MTBE,
has an approximate solubility of 2,000 mg/L.

10.3.2.2  Total Primary Porosity, 6t

    For the calculation of a maximum storge capacity, the higher typical
values of porosity (Table 10-1) should be used.  Since porosities differ
for different rock types, a porosity appropriate to the rock type of
interest should be used.

10.3.3  Guidance on Inputs for, and Calculation of, Average Values

10.3.3.1  Concentration of Contaminants in Secondary Pore Water, C

    An arbitrary but reasonable estimate of the average concentration is 10
percent of the maximum solubility.  A factor of 10 allows for the dilution
of contaminated water with clean pore water as a result of transport, as
well as other factors which may reduce the contaminant level in time, such
as volatilization or biodegradation.  For a gasoline without additives, the
average concentration is 20 mg/L, and for a gasoline with oxygenated
additives, the average concentration is 200 mg/L.
                                    265

-------
10.3.3.2  Total Primary Porosity, 9

    A reasonable estimate of average primary porosity may be made by again
consulting Table 10-2 and choosing values in the middle of the range of
values shown.  As mentioned before, average values should be chosen
appropriately for the specific rock type of interest.

10.4  Example Calculations

10.4.1  Storage Capacity Calculations

10.4.1.1  Maximum Values

    To estimate the maximum storage capacity in a limestone for a gasoline
containing oxygenated additives, use equation 10.5.  As recommended in
Section 10.3.2, the concentration in secondary pore water (C) should be
2,000 mg/L.  From Table 10-1, the maximum limestone porosity given is 34
percent.  Thus, the maximum storage capacity is:

       MC = C6t = (2000 mg/L)(0.34)

                « 680 mg/L, or
                = 680 g/m3


10.4.1.2  Average Values

    To estimate the average storage capacity in a granite for a gasoline
containing no additives, use equation 10.5 again.  As recommended in
Section 10.3.3, the concentration in secondary pore waters should be 20
mg/L.  From Table 10-1, a reasonable value of average granite porosity is
0.3 percent.  Thus, the average storage capacity is:

       MC . C9t - (20 mg/L)(0.003)

                = 0.06 mg/L, or
                         3
                = 60 mg/m


10.4.2  Transport Rate Calculations

    As indicated in Section 10.2.2.1, the mass of a pollutant diffusing
into the matrix from the secondary porosity can be calculated using the
following equations:

                            C. D0      C. 1 a
                             1  e  t - —	                        (10.1)
                        a = 6t + Kdp                                 (10.2)

with parameters as defined above.  Maximum values for diffusive capacity
can be determined by making some generous estimates of the parameters
involved.  A generous estimate of the slab length (1) for example, would
                                    266

-------
assume a highly fractured system so that the matrix blocks would be very
small.  Such a system might yield a matrix block size of 0.25 m (see
Section 13).  A generous estimate of the rock capacity factor (a) would
assume a non-sorbing contaminant (K, = 0) and a low porosity (6 ).  From
Table 10-1, a low estimate of porosity is 0.02 percent, which corresponds
to a crystalline rock, or 4 percent corresponding to a porous media.  A
generous estimate of the effective_diffusivity (D ) would lie on the high
end, or, from Table 10-2, 6.6 x 10"  cm /s for crystalline and 2 x 10
cm /s for porous rock.  A high estimate of the concentration (C.) may be
obtained as recommended in Section 10.3.2.  Finally, an estimate of time
(t) could be chosen as 1 year, to make the calculation for a unit time
frame.  Then, for crystalline and porous rocks,„the maximum flux parameter
Q would be 7.5 x 10~  mg/m  and 2.5 x 10~  mg/m , respectively.

    From these numbers, it is clear that the combination of D  and t would
have to increase by four (4) orders of magnitude for crystalline rocks to
gain the same storage capacity as porous rocks.  Since an increase in the
diffusivity of more than two (2) orders of magnitude is highly unlikely,
this means that a time scale of at least 100 years would have to be of
interest in order for diffusion into crystalline rocks to be as significant
as the diffusion which would take place in porous media in just one year.

10.5  Summary of Relative Importance of Locus

10.5.1  Remediation

    This discussion has indicated that the major mechanisms governing
contaminant migration into or out of the rock matrix are diffusion and
sorption.  In fact, it has been indicated that the diffusive flux into even
a porous medium is quite low relative to the presence of contaminants in
other loci.  Although sorption is an important process, it was not dealt
with qualitatively in this section for two reasons:  first, sorption is
discussed in detail elsewhere; and second, and more importantly, sorption
cannot provide a truly significant degree of fixation in this locus for two
reasons.  Although the sorbing minerals are often listed as common
constituents of various rock types, they rarely comprise a significant
fraction of any rock.  In addition, even in cases where they do comprise
significant portions of a rock, the low diffusion rate of contaminants into
the rock limits the degree of sorption.

    The real significance of findings for this locus lies in the fact that
once contaminants are introduced to the locus they are not easily removed.
This demonstrates the difficulty of cleaning a release site to zero-level
concentrations.  Although this locus poses neither a significant help nor a
significant impedence to most methods of remediation, the fact that minor
amounts of contaminant will diffuse into the rock indicates that zero-level
remediation may not be possible.  In cases where such a zero-level
remediation effort is deemed desirable, the drastic measure of vitrifica-
tion as a means of driving off the contaminants into the air by literally
melting the rock is probably the only option which will truly work.
                                    267

-------
10.5.2  Loci Interactions

    The loci that interact with locus no. 10 are summarized in Figure 10.2.
The first, essential step in interaction between locus no. 10 and other
loci is aqueous phase diffusion out of the rock matrix.  Then, the diffused
contaminants in the unsaturated zone may: partition out of water and into
the vapor phase (locus no. 1), partition out of water and onto water-dry
soil particles or rock surfaces (locus no. 2), remain dissolved in the
relatively immobile water film surrounding well drained soil particles or
rock surfaces (locus no. 3), remain dissolved in relatively mobile pore
waters just above the capillary fringe or associated with heavy rainfall
(locus no. 12), or undergo biodegradation (locus no. 11).  In the saturated
zone, on the other hand, the diffused contaminants may: become sorbed to
water-wet soil particles or rock surfaces (locus no. 4); remain dissolved
in mobile pore waters (locus no. 8) or in rock fracture and dissolution
channels (locus no. 13), in which cases the contaminant may alternate
continuously between diffusive and advective transport; or undergo
biodegradation (locus no. 11).  As has been indicated throughout this
section, diffusion into the rock matrix from any other locus is so very
slow that the other loci are more likely to control contaminant transport,
with diffusion into and out of the rock matrix providing some small degree
of retardation.

10.5.3  Information Gaps

    In order to better understand the movement of contaminants into and
through the rock matrix, the following data are needed:

    1.  Additional measurements of diffusivity and primary porosity;

    2.  A better understanding of fracture densities and what level of
        fracturing may still be modeled as an unfractured porous media;

    3.  A better understanding of sorption processes within the rock
        matrix, and especially of the capacity, if any, for sorption to
        effectively remove molecules from the total concentration gradient
        which governs diffusion; and

    4.  A better characterization of advective transport in stratified
        porous media, with special attention given to the extent of
        diffusion as the mechanism controlling transverse dispersion and  to
        the relation between advective longitudinal velocity and transverse
        diffusion.

10.6  Literature Cited

Garrels, R.M., Dreyer, R.M. and Howland, A.L.  1949.  Diffusion of Ions
    Through Intergranular Spaces in Water-Saturated Rocks.  Bull. Geol.
    Soc. Am., No. 60.  pp. 1809-1828.

Grisak, G.E., J.F. Pickens, and J.A. Cherry.  1980.  Solute Transport
    Through Fractured Media 2.  Column Study of Fractured Till.  Water
    Resources Research, 16(4):731-739.
                                    268

-------
Heath, R.C.  1983.  Basic Groundwater Hydrology.  USGS Water-supply Paper
    No. 2220.  Alexandria, VA:  USGS.

Neretnieks, I.  1980.  Diffusion in the Rock Matrix:  An Important Factor
    in Radionuclide Retardation?  Journal of Geophysical Research,
    85(B8):4379-4397.

Skagius, K. and I. Neretnieks.  1986a.  Porosities and Diffusivities of
    Some Nonsorbing Specifies in Crystalline Rocks.  Water Resource
    Research, 22(3):389-398.

Skagius, K. and I. Neretnieks.  1986b.  Diffusivity Measurements and
    Electrical Resistivity Measurements in Rock Samples Under Mechanical
    Stress.  Water Resource Research, 22(4):570-580.

Starr, R.C., R.W. Gillham, and E.A. Sudicky.  1985.  Experimental
    Investigation of Solute Transport in Stratified Porous Media 2.  The
    Reactive Case.  Water Resources Research, 21(7):1035-1041.

Sudicky, E.A., R.W. Gillhan, and E.D. Frind.  1985.  Experimental
    Investigation of Solute Transport in Stratified Porous Media 1.  The
    Nonreactive Case.  Water Resources Res., 21(7):1035-1041.

Wang, J.S.Y. and T.N. Narasimhan.  1985.  Hydrologic Mechanisms Governing
    Fluid Flow in a Partially Saturated, Fractured, Porous Medium.  Water
    Resources Research, 21(12):1861-1874.
                                    269

-------
                         SECTION 11.  LOCUS NO. 11

                                 CONTENTS
                                                                   Page No.

List of Tables	     271
List of Figures	     271
11.1  Locus Description	     272

      11.1.1 Short Definition	     272
      11.1.2 Expanded Definition and Comments	     272

11.2  Evaluation of Criteria for Remediation	     272

      11.2.1 Introduction	     272
      11.2.2 Microbial Uptake of Hydrocarbons	     276
      11.2.3 Biodegradation	     280

             Mechanisms	      280
             Kinetics	      280
             Additional Factors Affecting Biodegradation	      289

      11.2.4 Mobility....	      290
      11.2.5 Viability	      291

11.3  Storage Capacity in Locus	     291

      11.3.1 Introduction and Basic Equations	     291
      11.3.2 Guidance on Inputs for, and Calculation of,
               Maximum Value	     293
      11.3.3 Guidance on Inputs for, and Calculation of,
               Average Values	     293

11.4 Example Calculations	      294

      11.4.1 Storage Capacity Calculations	     294

             11.4.1.1  Maximum Value	     294
             11.4.1.2  Average Value	     294

      11.4.2 Transformation Rate Calculations	     294

11.5  Summary of Relative Importance of Locus	     295

11.6  Literature Cited	     295
                                    270

-------
                                  TABLES

                                                                   Page No.

11-1  Factors Affecting Removal                                       276

11-2  Bioremedial Actions                                             277

11-3  Bacteria-Water and Octanol-Water Partition Coefficients
      for Several Polyaromatic Hydrocarbons                           279

11-4  Six Models for Mineralization Kinetics with only the
      Variable of Substrate Concentration and Cell Density            285

11-5  Parameters and Asymptotic Standard Deviations of Four
      Models of Substrate Disappearance Kinetics Fit to
      Data on Metabolism of Nine Concentrations of
      [U-ring-  C] Benzoate by Pesudomonas Sp.                        288

11-6  Relative Biodegradability of Hydrocarbons                       289


                                  FIGURES
11-1  Schematic Cross-Sectional Diagram of Locus No. 11 -
      Contaminants Sorbed Onto or Into Soil Microbiota in
      Either the Unsaturated or Saturated Zone                        273

11-2  Schematic Representation of Important Transformation
      and Transport Processes Affecting Other Loci                    274

11-3  Representation of Hydrocarbon Inclusion body,
      "H", in Acinebacter sp. Grown on Hexadecane                     275

11-4  Pathways of Biodegradation for Several Hydrocarbons             281

11-5  Kinetic Models as a Function of Initial Substrate
      Concentration and Bacterial Cell Density                        286

11-6  Kinetic Disappearance Curves for Six Kinetic Models
      of Biodegradation                                               287

11-7  Breakthrough Curves for Escherichia Coli and
      Coliphage F2                                                    292
                                    271

-------
                         SECTION 11 - LOCUS NO. 11
11.1  Locus Description

11.1.1  Short Definition

    Contaminants sorbed onto or into soil microbiota in either the
saturated or unsaturated zone.

11.1.2  Expanded Definition and Comments

    Organic chemicals can be taken up by soil microbiota by several
mechanisms: 1) sorption to the exterior cell membranes or to exuded
extracellular material; 2) molecular transport across the cell membranes
into the cells' cytoplasm; or 3) ingestion of particles or liquid droplets.
In the latter case, inclusions of nearly pure liquid contaminant may exist
inside the biota.  While uptake is possible from air, liquid contaminant,
or aqueous solution, uptake from aqueous solution or liquid contaminant are
by far the more important.  Soil biota include bacteria, fungi, and yeasts,
but bacteria are by far the most numerous and, therefore, the most
important.  The mobility of microbes is proportional to the degree of
moisture saturation in the soil, and the sorbed contaminants possess a
similar degree of mobility.  The potential importance of this locus as a
temporary reservoir for contaminants has never been assessed.

    Because only aerobic degradation is considered important for
hydrocarbons, and the concentration of biota in soil drop off sharply with
increasing depth, the unsaturated zone is probably much more important than
the saturated zone for biodegradation of hydrocarbons.  However, some
groundwater (in the saturated zone) does contain dissolved oxygen
sufficient to allow the aerobic biodegradation of a few milligrams/liter of
dissolved hydrocarbons.  Figures 11-1 and 11-2 present a schematic
cross-sectional diagram of locus no. 11 and a schematic representation of
the transformation and transport processes affecting other loci,
respectively.

11.2  Evaluation of Criteria for Remediation

11.2.1  Introduction

    Locus no. 11 is by nature unique among the thirteen loci described in
this report.  Since microbes may exist inside or adjacent to each of the
other loci, there is close interaction between locus no. 11 and each of the
others.  Also, being animate organisms, microbes are chemically complex and
dynamic.  They act as agents of physical and chemical change to leaked
hydrocarbons in the subsurface.  As a population, microbes can increase or
decrease in number, adapt, and be transported from locus to locus,
depending on environmental conditions.  For these reasons, this section is
structured differently than the others.  Emphasis is placed on the
thermodynamic and kinetic aspects of "bioremoval," which involves multiple
mechanisms.  The importance of locus no. 11 clearly lies in its potential
to remove leaked hydrocarbons from  the subsurface.
                                    272

-------
     UNSATURATED
         ZONE
WATER
TABLE
          t
          1
      SATURATED
        ZONE
     NOTE: NOT ALL PHASE BOUNDARIES ARE SHOWN.
          I      I
                    LEGEND


CONTAMINANT VAPORS        -^^i WATER IN SATURATED ZONE

AIR (PLAIN WHITE AREAS IN        X  BIOTA (MICROBIOTA INCLUDED)
UNSATURATED ZONE)
                                           CD SOIL PARTICLE
             Q    WATER (MOBILE PORE WATER
                  IN UNSATURATED ZONE)         f  COLLOIDAL PARTICLE
           AAAA  WATER RLM
                          • ••• CONTAMINANTS
   Figure 11-1. Schematic Cross-Sectional Diagram of Locus No.11 -
               Contaminants Sorbed onto or Into Soil Mlcroblota In
               either the Unsaturated or Saturated Zone.
                                   273

-------
                                (SOIL GAS)
                                LOCUS NO.
                          1 - CONTAMINANT VAPORS
10
       (ROCK)
     10CUS NO
     DIFFUSED INTO
     MINERAL GRAINS
     OR ROCKS
                       DEPURATION
DEPURATION
              UPTAKE
                                         UPTAKE
                                             DEPURATION
                                                   UPTAKE
                             LOCUS NO. 11
                           (CONTAMINANTS SORBED
                             ONTO OR INTO SOIL
                          MICROBIOTA IN SATURATED
                           OR UNSATURATED ZONE)
               UPTAKE
 (SORBED CONTAMINANTS)
      LOCUS NO
4 -SORBED TO •WATER-WET-
  SOIL PARTICLES
9-SORBED TO COLLODIAL
  PARTICLES
   (LIQUID CONTAMINANTS)
       LOCUS NO
 2- ADHERING TO 'WATER DRr
   SOIL PARTICLES
 5-IN PORE SPACES IN
   SATURATED ZONE
 6-IN PORE SPACES IN
   UNSATURATED ZONE
 7- FLOATING ON WATER
   TABLE
13 - IN ROCK FRACTURES
                                  (WATER)
                                 LOCUS NO.
                            3- DISSOLVED IN WATER
                              FILM
                            8 - DISSOLVED IN GROUND-
                              WATER
                           12- DISSOLVED IN MOBILE
                              PORE WATER
                                                       TRANSPORT OF BIOTA
                                                       WITH WATER
Figure 11-2. Schematic Representation of Important Transformation
            and Transport Processes Affecting Other Loci.
                                    274

-------
    Biological removal of contaminant compounds from soil consists of  two
distinct, yet somewhat interdependent, mechanisms:  microbial uptake and
biodegradation.  Microbial uptake is the physical-chemical removal of
compounds by microbes from the surroundings.  It may take the form of
either adsorption to the microbe's surface or transport  through  the
microbe's membrane covering into its interior.  Transport through the  cell
membrane can itself take the form of either diffusion-like flow  or bulk
transport.  Diffusion occurs at the molecular level and  is strongly
influenced by hydrophobic attraction between the contaminant and the cell.
Bulk transport, also called inclusion, resembles "swallowing" by the cell
of a volume of solution or pure product.  Figure 11-3 shows several
inclusion bodies of hexadecane inside a bacterium.
                                                        0.2 urn
  Source:  Scott and Finnerty, 1976. (Copyright American Society for Microbiology. 1979.
         Reprinted with permission.)

            Figure 11-3. Representation of Hydrocarbon Inclusion Body,
                       "H", in Aclnebacter sp. Grown on Hexadecane
    Biodegradation is the microbially mediated chemical transformation of
organic compounds to form new product compounds.  Biodegradation is largely
attributed to reactions occurring within the cell, although many microbes
excrete enzymes that catalyze transformations outside the cell.  In the
case of reactions within the cell, uptake is a necessary prerequisite to
biodegradation.

    Since microbes, consisting largely of bacteria, exist and thrive
abundantly in surface soils, biodegradation is a spontaneous and "natural"
process.  The bacteria adapt to, and interact with, organic contaminants as
they would other organic compounds.  The rates of uptake and transformation
                                    275

-------
can be artificially altered by manipulation of various factors.
Bioremediation, the term applied to manipulating conditions to facilitate
biodegradation, is a viable option that requires a clear understanding of
the mechanisms of transport and adsorption and the kinetics of microbial
metabolism and growth.  An examination of the factors that limit the rates
of these processes is thus in order.  The rate limiting factors, which may
be grouped according to their effects on uptake or metabolism, are listed
in Table 11-1 and will be addressed in the following sections.

                                TABLE 11-1

                       FACTORS AFFECTING BIOREMOVAL
    A.   Uptake

         -  bacterial concentration (population density)
         -  bacterial species
            contaminant concentration
         -  degree of microbial and contaminant dispersion in soil

    B.   Biodegradation

         -  molecular structure of contaminant compounds
         -  contaminant concentration
            bacterial concentration
         -  bacterial species
         -  nutrient (carbon, oxygen, nitrogen, phosphorus)
            concentrations
            temperature '
         -  pH
    Various modes of bioremediation are listed in Table 11-2.  All focus on
the transformation of contaminants and not removal via sorption.  This is
because sorption is reversible and, therefore, not permanent whereas
transformation is essentially irreversible.  Composting and groundwater
recirculation, with and without the addition of specially selected
bacteria, are the more common and are proven remedial techniques.  The
vapor phase bioreactor and slurry reactors are both in development, but
hold the promise of unique advantages for specific remedial requirements.

11.2.2  Microbial Uptake of Hydrocarbons

    As stated previously, the two major bacterial uptake mechanisms are
adsorption to the exterior membrane surface of the bacteria and absorption
through the membrane to the cell's interior (cytoplasm).  A review by
Baughman and Paris (1981) revealed that researchers have tended to ignore
the distinction between the two.  However, some have assumed adsorption (or
"uptake" in their terms) while only one suggested adsorption as the sole
mechanism.  One group proposed sequential adsorption and absorption.
                                    276

-------
                                TABLE 11-2

                            BIOREMEDIAL ACTIONS

1.
Mode
Composting
Soil
Handling
Excavated
Mixing
Addition
Direct Soil
Comments
- Simple direct method;
"low tech"
2.  Groundvater
    Recirculation
In situ
3.  Selected
    microbes
Excavated
or in situ
4.  Vapor Phase
    Bioreactor
In situ
5.  Slurry
    Reactor
Excavated
Delivered via
Recirculation
H2°
Either of
the above
Mixed directly
into bioreactor
Mixed directly
into bioreactor
  requires control of
  volatilized compounds

  similar to land
  treatment of refinery
  wastes

  dependent on site
  hydrogeology

  dependent on soil
  transmissivity

  may be used with any
  other mode

  most effective in new
  spill cleanup

  limited or no additional
  effect for old spills
  where indigenous
  bacteria are well
  acclimated.

  used downstream of vapor
  extraction systems

- bioreactor may be solid
  phase (soil) or
  submerged (aqueous)
  culture

- still experimental

- high energy input
  required

  high reaction rate

  capital intensive

  still experimental
                                    277

-------
    All experimental work reported in the literature has been performed in
mixed aqueous solution.  Transport limitations make it nearly impossible to
measure the bulk partitioning of organic compounds between bacterial cells
and soil.  Thus, the available equilibrium partitioning data probably are
only applicable in the immediate vicinity of soil microbes, especially in
the vadose zone.

    The relevant mathematical expressions applicable to uptake are the same
as those for other equilibrium partitioning phenomena.  They are the
partition coefficient and the Freundlich equation.  The partition
coefficient, K_, is defined as the ratio of the contaminant concentration
in the cell, [C]  ,,, to the concentration in the bulk medium, [C]   ,, at
equilibrium.    ceii                                              med


                   K_ =    cel1                                      (11.1)
where:  [C]  ,, = grams hydrocarbon on or in microbes per gram dry soil

        [C]  .  m grams hydrocarbon per gram dry soil

    The Freundlich equation, proposed in 1909, adds an exponential
parameter:


                   ^cell - Kf

    where:  Kf   = Freundlich coefficient (dimensionless)

            1/n  = linearity constant (dimensionless)


    Although both equations have been used, it is important to note that
1/n approaches unity for many systems at low concentration.  Therefore, K,
will equal KR in many instances.

    There is a severe lack of partitioning data for hydrocarbons.  Nearly
all available data are for pesticides and other chlorinated compounds
(Baughman and Paris, 1981; Bell and Tsezos, 1987).  Some values of Kfi for
polyaromatic hydrocarbons (PAHs) are given in Table 11-3.  Baughman and
Paris (1981) found no data for non-cyclic hydrocarbons and no Freundlich
parameters for any hydrocarbons.

    In the absence of experimentally-derived K_ values, an alternative
method of assessing potential partitioning behavior involves octanol-water
coefficients.  As previously mentioned, uptake by microbes is strongly
dependent upon the hydrophobicity of the contaminant.  Hydrophobicity is
strongly a function of chemical structure.  Specifically, for hydrocarbons,
the higher the molecular weight, the more hydrophobia is the compound.  For
example, n-octane, CgH.g, is more hydrophobic than n-hexane, CL-H. ,.  It
could be expected, tnen, that the partition coefficient, K_, is greater for
octane than for hexane.  This "hydrophobic effect" is confirmed when the
partitioning of a compound between octanol and water (expressed as the
octanol-water coefficient, K  ) is compared with Kg.
                                    278

-------
                                  TABLE 11-3

         BACTERIA-WATER AND OCTANOL-WATER PARTITION COEFFICIENTS FOR
                      SEVERAL POLYAROMATIC HYDROCARBONS
Compound
Benz [a]
anthracene
Benz [a]
pyrene
Benz [f]
quinoline
Pyrene
Phenanthrene
References:
Organisms
Viable bacteria
Killed bacteria
Killed bacteria
Viable bacteria
Heat killed
bacteria
Unspecified
bacteria
Unspecified
bacteria
(1) Smith et al., 1978
(2) Karickhoff et al. ,
(3) Leo, A.J. , 1975.
KB
(3.6 + 0.3)xl04
(9.9 + 0.7)xl04
(3.5)xl05
149
540
2.7xl04
6.3xl03
1979.
Reference K^ Reference
(1) 4.1xl05 (3)
(1) 2.2xl06 (3)
(1) 1.6xl03 (3)
(2) l.SxlO5 (2)
(2) 3.3xl03 (2)

    Table 11-3 includes the octanol-water partition coefficients, for
several PAHs.   The relationship between KB and  Kou is quite linear
considering that several orders of magnitude are spanned.   This is
consistent with previous correlations made between Kou and Kp values for
higher organisms (Baughman and Paris, 1981).  Thus,  K^ may be a  useful
surrogate for predicting KB.

    No detailed kinetic studies are available in the literature although
some workers have made measurements over time to verify equilibrium
(Baughman and Paris, 1981).  Generally, it appears that equilibrium is
reached in a matter of minutes to several hours.  There have been no
exceptions involving bacteria under experimental conditions.

    The four principal factors affecting uptake were listed in Table 11-1.
Two important factors are the bacterial population density and the
contaminant concentrations.  For uptake to occur, a hydrocarbon molecule
must come in close contact with a bacterium.  The higher the concentrations
of both bacteria and contaminants, the more frequent will  be the contacts
between the two.  Assuming uniform dispersion of bacteria  in the volume of
soil, a high degree of dispersion of the contaminant within the soil is
desirable to facilitate microbial/contaminant contact.   It should be noted,
                                           279

-------
however, that some minimum contaminant concentration is required for any
biodegradation to occur (Alexander, 1985).

    Baughman and Paris (1981) reviewed microbial uptake studies in which
various bacterial species were contacted with the same compound.  This
review showed that the partition coefficient, K_, was somewhat dependent
upon the species of bacteria in the test.  There are no data available,
however, addressing the dependence of Kg on bacterial species for
hydrocarbons.  The number of species of bacteria present in a soil is
dependent on the site history, especially with regard to previous
contamination.

11.2.3  Biodegradation

Mechanisms

    Examining the mechanisms of molecular transformation allows one to
judge the effectiveness of detoxification by biodegradation.  A series of
biodegradation reactions, called the biodegradation pathway, typically
occurs as more complex substances are transformed into simpler compounds.
For most hydrocarbons, these simpler compounds are less toxic than the more
complex hydrocarbons.  For most cases, degradation results in carbon
dioxide and water as final end products; this is termed mineralization.
Other times, however, the hydrocarbons do not reach a mineralized endpoint
but rather result in relatively stable aliphatic and aromatic compounds.
These compounds may become integral parts of the soil humus (Alexander,
1977) or may be flushed out of the immediate vicinity in the form of highly
soluble acids, ketones and alcohols (e.g., acetic acid, acetone).  In the
case of chlorinated hydrocarbons, mineralization is less often the
biodegradation pathway.  Chlorinated compounds may be transformed into
recalcitrant and sometimes more toxic compounds.

    The precise pathways of hydrocarbon oxidation are debated in the
literature.  However, transformations likely occur stepwise from end
carbons, producing alcohols, aldehydes and fatty acids in sequence.
Several examples are presented in Figure 11-4 (Singer and Finnerty, 1984).

Kinetics

    Although a large body of literature exists on the kinetics of microbial
growth, relatively little is available on biodegradation rates.  Simkins
and Alexander (1984) performed examination of degradation kinetics.
Several assumptions are made for this analysis:

    1.  There is no transport limitation of any substrate.

    2.  The only substrate which may be limiting is carbon; 0«, N, and P
        are always in excess.

    3.  The microbial population consists of a single species and may be
        rate-limiting.

    4.  All organic contaminants present are equally metabolizable.
                                    280

-------
                                PRISTANE
                                   \
                                   HYI
                                  1
     2,6,10,14-TETRAMETHYLPENTADECANOIC ACID

w-OXIDATION           f             B-OXIDATION
 2,6,10,14-TETRAMETHYLPENTADECANEDIOIC    4,9,12-TRIMETHYLTRIDECANOIC ACID


2,6,10,-TRIMETHYLTRIDECANEDIOIC ACID        2,6,10-TRIMETHYLUNDECANOIC ACID
2,6,10-TRIMETHYLUNDECANEDIOICACID
  2,6,-DIMETHYLNONANEDIOIC ACID
  2,6-DIAMETHYLHEPTANEDIOIC ACID
    2-METHYLPENTANEDIOIC ACID
                Pathway of pristane oxidation by Brevlbactorium erythrogenes.
        H3C-(CH2)U-CH3
        (n-HEXADECANE)
     H3C-(CH2)U-CH2OOH
 (n-HEXADECYLHYDROPEROXIDE)
      H3C-(CH2)U-CH2OH
       (n-HEXADECANOL)
      H3C - (CH2 )u - CHO                         (HEXADECYLHEXADECANOATE)
    (n-HEXADECYLALDEHYDE)
      H3C - (CH2 )u - COOH
    (n-HEXADECANOIC ACID)-
        Proposed pathway of hexadecane metabolism in Acinetobacter sp. H01-N.


     Source: Adapted from Singer and Rnnerty, 1984
    Figure 11-4. Pathways of Biodegredation of Several Hydrocarbons

                                                                     (Continued)

                                      281

-------
      H3C-(CH2)11-CH3
        (n-TRIDECANE)
                                                 H3C-(CH2)n -CH3
                                                    (n-ALKANE)
                                                           O,
                OH
    H3C-(CH2)10-CH-CH3
      (TRIDECAN-2-OL)
                                                H3C-(CH2)n -CH2OH
                                              (PRIMARY FATTY ALCOHOL)
             J
                                                        \
     H3C * -HYDROXY FATTY ACID)
H3C - (CH2 )9 - CH2 OH + CH3-COOH
 (UNDECAN-1 -OL)  (ACETIC ACID)
                                                OHC-(CH2)n -COOH
                                              (CO-ALDEHYDE FATTY ACID)
      H3C -(CH2)g -COOH
      (UNDECANOICACID)
                                                HOOC-(CH2)n -COOH
                                              (DICARBOXYLIC FATTY ACID)
Pathway of Subtermlnal Alkane
  Oxidation In Pseudomonas.
                                           Pathway of Dltermlnal Alkane Oxidation
                       Figure 11-4. (Continued)
                                      282

-------
    The implication of these assumptions when applying the kinetic models
in subsurface conditions vill be discussed later.

    Although several models of cell growth have been proposed that depend
upon only a single substrate, the classic model of kinetics was proposed by
Monod :

                     u   C
              u =     max                                            (11.3)
                    
-------
    If the mass balance equation and its time derivative are solved for X
and dX/dt, respectively, these expressions can be substituted into equation
11.5.  The result is:

                  -dC     umav C (C  + Xn - C)
                                   o    o
                   dt            K  + C
                                  s

    Equation 11.7 is a general expression of substrate disappearance for
systems in which only cell densities and substrate concentration determine
the rate of substrate removal.

    Simkins and Alexander (1984) identified six special cases which may be
expressed as simplifications of the general expression.  These are
presented in Table 11-4.  Figure 11-5 illustrates the approximate relative
values of X ,  C  and K  required for each of these cases.  Figure 11-6
shows the snapes of all six kinetic curves.  Table 11-5 gives parameter
values for disappearance of benzoate with Pseudomonas bacteria using four
of these models.

    The assumption that transport of CL, N, or P is not rate-limiting is
not always robust.  The assumption becomes more reasonable, however, as the
size of the unit volume decreases and is most reasonable at the microscale
level of locus no. 11 (i.e., a single cell which, if alive, is surrounded
by water film).  Also, transport of both microbes and substrates, including
contaminants,  increases with moisture content, especially under saturation
conditions.  Saturation exists below the water table at all times, as well
as above the water table, to some extent, after heavy rainfall.

    The carbon limitation assumption is also open to question.  Oxygen is
probably most often the limiting substrate, except perhaps after a heavy
rainfall.  However, soluble carbon sources such as liquid hydrocarbons can
be rate limiting when their concentration is sufficiently low.  In fact,
for some compounds, a minimum "threshold" concentration is necessary in
order to detect any mineralization at all (Alexander, 1985).  No threshold
values for hydrocarbons were found but  they are likely in the parts per
trillion (ppt) range.  The threshold for chlorinated species are generally
higher, in the parts per billion (ppb)  range, but vary from soil to aqueous
media.

    It was assumed that both a single bacterial species and a single carbon
source are operative in the development of the kinetic equations.  In fact,
multiple species and multiple contaminant compounds are likely.  This
presents a problem since u    is different for each species-compound
combination at a given temperature.  The precise kinetics of biodegradation
would be extremely complex if all of the many combinations were considered.
The approximation  that all degradation  is due to a single species is a good
one if the contamination is old, however, because the population of
hydrocarbons degraders will predominate by natural selection.
Physiologically,  the population will be sufficiently uniform that u    may
be used to approximate the growth for all active microbes.
                                    284

-------
                                          TABLE 11-4

                SIX MODELS FOR MINERALIZATION  KINETICS WITH ONLY THE
                VARIABLES  OF  SUBSTRATE CONCENTRATION AND CELL  DENSITY
     Model and charactcristio
                 Equation and incqualriie*
 I.   Zero order
       Differential form
       Integral form
       Derived parameter
       Necessary conditions

 II.   Monod. no growth
       Differential form
       Integral form
       Derived parameter
       Necessary condition

 111.  First order
       Differential form
       Integral form
       Derived parameter
       Necessary conditions

 IV.  Logistic
       Differential form
      Imecral form
 -d.V'dl = A,
 i' = 5,, - A,t
 "I =  M-m.iv*!>
 A',, » 5,i and 5,, » A',
 -d5'dt = A,5/(A\ -^  5)
 AMn(5/5,,) +  5 - 5,,  =
 A | = Hm:ixA(l
 A',, » 5,,
 -d5/dt = A,5
 5 = 5,, e\p(-A,U
 M = M-m,i\Ao/A,
 A',, » 5,i and 50 « A',
 -d5/dt = k4S(SH + A',, -  5)

     .V,, - A',,
5 =
                                                                         f*,, + A',,)t]
      Derived parameter
      Necessary condition

V.  Monod with growth
      Differential form
      Integral form
      Derived parameter
      Necessary condition

VI. Logarithmic
      Differential form
      Integral form
      Derived parameter
      Necessary condition
k* ~ M-m.i\''A,
5,, « A\
-d5/dt = |(xni.1N5(50 + A'(1 - 5)]/(A', + 5)
AMn(5/5,,) = <50 T A',, + A',)\n(XIX») - (5,, +
None
None
-d5/dt = (J.miiv(5,, -»- A',, - 5)
5 = 5,, -i- A()|l - exp(|xmiivl)]
None
t'  •* *^  1 *
Ju -s>  A,
   Source:   Simkins  and  Aldexander  (1984)
                                                 285

-------
               9-
               8-
I
s
.Q
3

"oj    6 -

1
'E


5    54
                                                                       VI
                    0.001     001      0.1        1        10       100      1000
                                 Initial substrate concentration (//g/mL)
           Ji II III IV V ana VI designate regions in wnicti zero-oroer Monoo iwitnout growini. first-order logistic
           Monoa iwun growth! and logamnmic kinetics are expected t3T I


Source:  Alexander, 1985.  (Copyright American Chemical Society. 1985. Reprinted with permission.)



 Figure 11-5.  Kinetic Models as a Function of Initial Substrate Concentration

                              and Bacterial Cell Density.
                                             236

-------
         100
          80-
                                  Time
Source: Alexander, 1985 (Copyright American Chemical Society.  Reprinted with permission)

Figure 11-6.  Kinetic Disappearance Curves for Six Kinetic Models of
                           Biodegradation
                                    287

-------









M
H
1
co •
M £^
fc Z C/3
O H
Z M
CO ^C
nJ fc SB
g O O
tc S 8
HH 5
S o to
£ 3 *
fc ^

o E w
u
C/3 Z IH
§ ° 0

H <3 N
"? 3 Sd H

•H HOW
iH >
M Q H tj
5§ § &J-1,
t«4 ^Q pt* t&p
Q H
Z C/3 -H
H H V
j_j S
H U CO
S 0 Z
N

e -o
CO O
H U



^n.
J
e
1*
V— '
o
X




j"
B
"bb
3
x>

CO





*•
n)
4-J
CO
c
o
U
2
(4
M












i-H
V
T)
0
as




**.
rH i— 1
n) B
4-> u bO
O 3
<; s-^
VD O 00 ON CO O
O 
odd

2< SZ SC eg j2 Sz
co oo m
rH CO «*
* • •
O O O


td (d (d
>^^ ^^ vO
§tH rHU O
o o CM m
O O O O OO T3 'V T>
• • • o o ^o ^ to co
+l+l+l°°°°°^
ogg *'*'*' *'*'*'
CMCMCMOrH . . . «
O O O O O II II II II

II II II II 11 X X X X
COCOCOrHrH S S B 8
i^ J^ ,*^ J^ K*{ 3. 3i 3, 3,




XI X!
t* r^
O 0 X!
^4 ^ 4-f
l-l ti M bO bO »^ O O O
0) 0) 01 O 'H -H «H
TJTJ'OOOKBBB
U U U C C bCXi Xi Xl
O O O *J 4J *J
« is « .H «H -H
4-1 4>J 4^ T3 ^O H3 ^J M M
MMMOOOrtrtrt
UbiiHCCCbObObO
•H -H -H O O O O O O




O CM
rH CO O C\l
O O t-H CO O CM

rH CO rH















4
0)
4J
9
C
•H
^
0)
Q,
l-l
V

•H
rH •
rH 1
rH C
•H «H
s a

tro

• O.O
rH rH
1 CO
CO B MH
(U ID O
•H U
3 bo co
G 0 4->
•H IH -H
a u c
•H P
m a
CO O «rl
4-> >
•H CO
§*J ><
•H to
X! § S8
X! 3 3.
•H bO CO
> C -H
•rt
> CO
CO «) rH
*t X!  OS
Stx
0)
*-• XCO
CO «3
u
a, "J
« a) jJ*4
U rH 4->
U «fl C OJ
cy o «j 4->
no «H 4-» CO
M rH CO C
o a c o
i a o o
4J «J O
CO (U
IH 4-< ey 4->
•H o xi rt
b Z E-i 04
• • • •

-------
    Table 11-1 noted that molecular structure is an important variable in
determining rates of biodegradation.  This contradicts the fourth
assumption, that all organic contaminants are equally degradable.  Table
11-6 lists qualitatively the relative order of hydrocarbon biodegradability
(Bartha and Atlas, 1977).  Unfortunately, a quantitative compilation of
u    values is not available.  Microbe-specific u    values for individual
compounds may be found in the literature.   A "characteristic" u    for the
mixture with the microbial population may be used for calculation!.

                                TABLE 11-6

                 RELATIVE BIODEGRADABILITY OF HYDROCARBONS
                            (decreasing order)


1.  N-ALKANES, C1()-C25

2.  iso-alkanes; biodegradability decreases as degree of branching
    increases

3.  olefins

A.  low-molecular-weight aromatic hydrocarbons; may be toxic to microbes at
    high concentrations

5.  polycyclic aromatic hydrocarbons

6.  cycloalkanes; degraded by very few organisms
Source:  Bartha and Atlas, 1977


    Nevertheless, the multiplicity of contaminant compounds decreases with
time as the more soluble and more volatile compounds are physically
removed.  This process is a part of "weathering" that all contaminants in
the subsurface undergo.  Also, for narrow "cuts" of petroleum products
(e.g., gasoline) the biodegradability is similar for the majority of the
volume of the mixture.

    These considerations must be aecounted for to practically apply
theoretical kinetics to locus no. 11.  This is difficult given the
variability in subsurface conditions, variability of contamination from
point to point, and ever-changing weather conditions.  Order-of-magnitude
rate estimates may be made by measuring cell density and hydrocarbon
concentration and determining u    of the system.

Additional Factors Affecting Biodegradation

    The effects of molecular structure, contaminant concentration,
bacterial number and species, and nutrient concentrations have been
discussed.  It is worth noting that temperature and pH also affect rates of
biodegradation.
                                    289

-------
    Optimum temperatures are definitely between 20°C and 40°C (Atlas,
1981).  The soil temperatures of most areas of the United States fall
within this range.  Therefore, soil temperatures in the continental U.S.
are not limiting except perhaps during winter months in northern latitudes.
Optimum pH for biodegradation is near the neutral range or slightly basic.

11.2.4  Mobility

    The mobility of organisms in the subsurface is highly dependent on the
hydrology of the site of interest.  Microbes move through entrainment in
aqueous streams either independently or while attached to colloidal soil
particles (see also locus no. 9).   Thus, in the vadose zone, bacterial
movement occurs almost exclusively by surface water percolation.  In the
saturated zone, transport occurs via groundwater movement.  In both zones,
microbial movement is attenuated due to the filtering effect of the soil.

    Experiments have been performed by Krone et al. (1958) to model
bacterial movement in the vadose zone.  By applying E. coli to sand
columns, the filtration or screening effect was measured.  At first, there
was a rise in bacterial content in the column effluent owing to detention
of bacteria by sand.  A maximum was reached followed by an asymptotic
decrease to a constant effluent bacterial count.  The decrease is due to
the enhanced filtering effect by trapped microbes.  As microbes are trapped
in and clog soil pores, the effective porosity of the soil decreases.
Fewer and fewer bacterial particles are able to pass until only larger
sized pores remain which cannot retain microbes.  If the rates of bacterial
and water loading remain constant, the rate of bacterial percolation will
thus also become constant.

    The foregoing observations were made after applying a large homogeneous
population of bacteria to the soil surface.  The situation of a
heterogeneous population of indigenous bacteria being mobilized by
percolating groundwater is somewhat different.  The bacterial concentration
is likely smaller.  Therefore, the enhanced filtering effect will be less
evident.  The size and shape of the bacteria also vary, which complicates
the overall percolation rate determination.  Also, there is not a constant
application of new microbes.  A somewhat different model is needed
reflecting the intermittent nature of applied surface water and a
relatively small population of microbes.  Nevertheless, the model does lend
some general insight into the resistance to movement of soil bacteria.

    The soil structure of the site of interest also has a strong impact on
microbial mobility.  Smith et al. (1983) compared the bacterial filtering
of intact and repacked soil columns.  Up to 96 percent of bacteria applied
to intact columns was recovered while only 40 to 60 percent was recovered
from repacked columns.  The authors concluded that flow through soil
"macropores," whereby the bacteria bypass most of the adsorptive and
retentive actions of the soil grains, is a common phenomenon.

    Harvey et al. (1989) suggest that microbial growth may play an
important role in transport through aquifers.  Growth of bacteria, as they
are transported through the aquifer, would partially compensate for the
removal of bacteria due to sorption, biological adhesion, filtration
("straining"), predation, and lysis.  An example presented by those
                                    290

-------
researchers showed that without growth, the attenuation measured in one
experiment would preclude bacteria from traveling more than 100 meters
through a Cape Cod aquifer.

    Movement below the water table may best be illustrated by the work of
Bitton et al. (1974).  Figure 11-7 is a "breakthrough" curve depicting the
arrival of a bacterium and a much smaller virus at an observation well 500
feet from the injection point.  Filtration and dilution reduced the
concentration by two orders of magnitude.  The largest bacterial cells
arrived at the observation point earlier than the smaller virus because
they are less subject to filtration.

11.2.5  Viability

    Soil microbes capable of degrading petroleum products include
Pseudomonas. Flavobacterium. Achromobacter. Arthrobacter. Microccus and
Acinetobacter. among others.  In fact, more than 200 soil microbial species
have been identified that can assimilate hydrocarbon substrates (Savage et
al., 1985).  Total microbial counts of fertile soils range from 107 to 109
per gram of dry soil, and hydrocarbon degrader counts range from 10s to 106
per gram in soils with no history of pollution (Bossert and Bartha, 1984).
Soils which have been exposed to petroleum products have bacteria counts on
the order of 106 to 108  per  gram.  As  previously mentioned,  this effect  is
due to the selective pressure placed on a microbial population after a
hydrocarbon release.  Bacteria that readily utilize hydrocarbons will grow
rapidly while those that do not will remain stable or decrease in numbers.
Growth may be described by the Monod equation found in Section 11.2.3.

    Survival times of bacterial cells have been studied for populations of
pathogenic (disease causing) bacteria applied to the soil surface such as
£. tvphimurium. E. coli and £. faecalis (Bitton et al., 1983).  Population
decay is logarithmic with time after surface application.  For indigenous
microbial geni such as Pseudomonas and Arthrobacter. populations rise and
fall with available substrate concentrations, temperature and moisture
content.  The individual lifespan of a bacterium probably ranges between
two weeks and three months (Corapcioglu, 1984).

11.3  Storage Capacity in Locus

11.3.1  Introduction and Basic Equations

    The storage capacity of hydrocarbons by microbes in the subsurface is
relatively small as compared with other loci.  Equation 11.8 is derived
intuitively from the foregoing discussions.

            CHC,  microbes = CHC,  soil  .  KB  . Bsoil .  "cell         (11.8)


    where  CHC,  microbes =  microbial hydrocarbon storage capacity; the
                            grams of hydrocarbons in or on microbes per
                            gram of dry soil (g/g)
                                          291

-------
      25 i-
  *?
   O
                                                                 2:00
             C0= Conentration of tracer in Injected water
             C = Concentration measured in observation well
Source: Bitton et al.. 1974. (Copyright Kluwar Academic Publishers. 1974. Reprinted with permission.)

Figure 11-7.  Breakthrough Curves for Escherichia QoJLand Coliphage f2
                                     292

-------
                          = the weight fraction of hydrocarbons in the dry
                            soil (g/g)

    KB                    = the equilibrium partition coefficient of
                            hydrocarbons between microbes and soil.  It is
                            assumed, for calculation purposes, that Kg
                            values in Table 11-3 for bacteria and water
                            apply equally well between bacteria and soil.
                            Dimensions:

                      (grams bacterial hydrocarbon)/(grams soil hydrocarbon)
                      (      gram bacteria        )/(       gram soil      )

    B  .^          =  the soil population count of hydrocarbon-degrading
                      bacteria.  It is assumed that only hydrocarbon
                      degraders can take up (absorb or adsorb)
                      hydrocarbons.  Typical range is 10 -10  cells per
                      gram of dry soil.

    M  ,,          =  cellular mass; the mass of one bacterium is assumed
                      to be a constant, estimated by Alexander (1977) to be
                      1.5 x 10    grams per cell.

11.3.2  Guidance on Inputs for, and Calculations of, Maximum Value

1.  Hydrocarbon soil concentration; CHC
    A maximum value of 10% (0.1 g/g dry soil) by weight of hydrocarbons in
    soil is assumed.

2.  Partition coefficient, Kg                            5
    The maximum K_ value found for hydrocarbons is 3.5x10  for benz[a]
    pyrene (Table 11-3).  K_ values for lighter hydrocarbons are
    unavailable.  This is a very high concentration factor that likely is
    applicable only in very dilute solutions.  Somewhat arbitrarily, it is
    reasonable to limit cell uptake of hydrocarbons to fifty percent of the
    cell mass.  In a soil with 10 percent mass fraction of hydrocarbons,
    this corresponds to Kfi equal to five.

3.  Bacterial soil population count, B  .,
    10  hydrocarbon degrading bacteria per gram of dry soil is the maximum
    value stated in Section 11.2.5.  It is assumed that non-hydrocarbon
    degraders do not uptake hydrocarbons significantly.

11.3.3  Guidance on Inputs for, and Calculation of Average Value

1.  Hydrocarbon soil concentration; CHC_,
    An average value of 100 ppm (1,0x10"  grams per gram of dry soil) is
    assumed.

2.  Partition coefficient, K_                         »
    The partition coefficient for phenanthrene, 6.3x10 , is arbitrarily
    chosen as an "average" value (Table 11-3).
                                    293

-------
3.  Bacterial soil population count, B  .,
    Assuming that the likelihood of a given cell population is a Gaussian
    function of IggiQ (B  }]) the "average" value of bacterial population
    count is 5x10  cells per gram of dry soil.

11.4  Example Calculations

11.4.1  Storage Capacity Calculations

11.4.1.1  Maximum Value

    Using equation 11.8 and the maximum values from Section 11.3.2, a
    maximum microbial hydrocarbon storage capacity, Curi, may be calculated:
                                                     a\s

    Cur, microbes = (0.1) . (5) . 108 . (1.5xlO~12)
     nt/

                  = 7.5x10"  grams hydrocarbon in or on microbes per gram
                    of dry soil

11.4.1.2  Average Value

    Using equation 11.8 and "average" values cited in section 11.3.3, an
    "average" microbial hydrocarbon storage capacity, CHC, may be
    calculated.

    Cnr,, microbes - (l.OxlO'4) . (6.3xl03) . (5xl06) . (1.5xlO~12)
     nu
                  « 4.7x10"  grams hydrocarbon in or on microbes per gram
                    of dry soil

11.4.2  Transformation Rate Calculations

    As developed in section 11.2.3, the form of the equation describing the
rate of biodegradation of any organic contaminant depends upon both the
cell population and the substrate concentration.  Figure 11-5 shows the
approximate ranges of initial substrate and cell number for each kinetic
model.

    1.  Sample calculation:  First order kinetics

    If the substrate concentration is known-to be 0.03 ug/mL and the
initial cell population is approximately 10  microbes per mL, first order
kinetics apply:
                   "     =  k, C
                     dt      J

    For benzoate, k~ = .0293 minutes"  according to Table 11-5.  Therefore,
a volumetric rate or degradation can be calculated:

                   - dC  =  (.0293) (.030) = 8.8 x 10"4 yg/min-mL

                     dt
                                    294

-------
    2.  Sample calculation:  Monod kinetics with growth

    If  the substrate concentration is known to be 1.5 ug/mL, and X - 2.5
ug/ml,  then Monod kinetics with growth apply.  Rearranging equation 11.5
yields.


                   - dC      umax C X
                     dt       Ks + c

                                 -3        -1
    For benzoate, u    = 7.2 x 10   minutes   and K  «= 0.45 ug/mL,
according  to Table 11-5.  Therefore, a volumetric biodegradation rate can
be calculated.

                   - dC       (7.2 x 10"3) (1.5) (2.5)
                     dt             (.45 + 1.5)

                         « 0.013 ug/min-mL

11.5  Summary of Relative Importance of Locus

    As stated in the introduction, the importance of locus no. 11 lies in
its functions as the site of biodegradation.  Biodegradation is the
principal means by which subsurface organic pollutants can be mitigated
naturally.  The rate of natural biodegradation in the subsurface is usually
severely limited by oxygen availability.  The biodegradation rate can be
increased by a large factor if bioremedial actions are taken.

    Enhancement of biodegradation principally attempts to increase the
availability of oxygen to soil and groundwater microbes.  Bioremediation is
a viable and advantageous option for abatement of leaked organic
contaminants.

    Attempts to predict rates of biodegradation by the use of mathematical
models typically involve assumptions that may not be true in the
underground environment.  A better way to predict natural and enhanced
biodegradation is actual laboratory or field microcosm studies that mimic
actual environmental conditions.  Decisions that affect the long term
environmental condition of a leak site should consider enhanced
biodegradation as a remedial alternative.

11.6  Literature Cited

Alexander, M.  1977.  Introduction to Soil Microbiology.  2nd Edition, John
    Wiley and Sons, New York.

Alexander, M.  1985.  Biodegradation of Organic Chemicals.  Environ. Sci.
    Technol., 19(2):106-111.

Alexander, M.  1986.  Introduction to Soil Microbiology, John Wiley & Sons
    Inc., New York.
                                    295

-------
Atlas, R.M.  1981.  Microbial Degradation of Petroleum Hydrocarbons:  An
    Environmental Perspective.  Microbiol. Rev., 45(1):180-209.

Bartha, R. and R.M. Atlas.  1977.  The Microbiology of Aquatic Oil Spills.
    Adv. Appl. Microbiol., 22:225-226.

Baughman, G.L. and D.F. Paris.  1981.  "Microbial Bioconcentration of
    Organic Pollutants from Aquatic Systems - A Critical Review."  In; CRC
    Critical Reviews in Microbiology, January 1981.  pp. 205-228.

Bell, J.P. and M. Tsezos.  1987.  Removal of Hazardous Organic Pollutants
    by Biomass Adsorption.  JWPCF, 59(4):191-198.

Bitton, G., S.R. Farrah, R.H. Ruskin, J. Butner and Y.J. Chou.  1983.
    Survival of Pathogenic and Indicator Organisms in Ground Water.  Ground
    ¥ater, 21(4):405-410.

Bitton, G., N. Lahav, and Y. Henis.  1974.  Movement and Retention of
    Klebsiella Aerogenes in Soil Columns.  Plant Soil, Vol 40, no. 373.

Bossert, I. and R. Bartha.  1984.  The Fate of Petroleum in Soil
    Ecosystems.  In;  Petroleum Microbiology, R.M. Atlas (Ed.), McMillan
    Publishing Co., New York.

Corapcioglu, M.Y.  1984.  Study and Prediction of Bacterial and Viral
    Contamination of Ground Water form Landfills and Septic Tank Systems.
    Project Completion Report, Grant No. 14-08-0001 - G833/02, Dept. of
    Civil Engineering, University of Delaware, Newark, Delaware.

Harvey, R.W., L.H. George, R.L. Smith, and D.R. LeBlanc.  1989. Transport
    of Microspheres and Indigenous Bacteria through a Sandy Aquifer.
    Results of Natural- and Forced-Gradient Tracer Experiments.  Environ.
    Sci. and Technol., 23:51-56.

Karickhoff, S.W., D.S. Brown and T.A. Scott.  1979.  Sorption of
    Hydrophobic Pollutants on Natural Sediments.  Water Research, Vol. 13,
    no. 241.

Krone, R.B., G.T. Orlob, and C. Hodgkinson.  1958.  Sew. Indust. Wastes,
    Vol. 30, no. 1.

Leo, A.J.  1975.  Calculation of Partition Coefficients Useful in  the
    Evaluation of the Relative Hazards of Various Chemicals in the
    Environment.  In;  Structure - Activity Correlation in Studies of
    Toxicity and Bioconcentration with Aquatic Organisms, Veith, G.D. and
    Konasewich, D.G. (Eds.)  International Joint Commission.

Savage, G.M., L.F. Dias and C.G. Golueke.  1985.  Biocycle, 26(1):31-34.

Scott, C.C.L. and W.R. Finnerty.  1976.  J. Bacteriology, 127(481-489).

Simkins, S. and M. Alexander.  1984.  Models for Mineralization Kinetics
    with the Variables of Substrate Concentration and Population Density.
    Appl. Environ. Microbiol., 47:1299-1306.
                                    296

-------
Singer, M.E. and V.R. Finnerty.  1984.  Microbial Metabolism of
    Straight-Chain and Branched Alkanes.  In;  Petroleum Microbiology, R.
    Atlas (Ed.)  p. 1-60.

Smith, J.H., V.R. Mabey, N. Bohonos, B.R. Holt, S.S. Lee, T.V. Chou, D.C.
    Booberger, and T. Mill.  1978.  Environmental Pathways of Selected
    Chemicals in Freshwater Systems.  Part II, Laboratory Studies.  U.S.
    EPA, Athens, GA.

Smith, M.S., G.V. Thomas, and R.E. White.  1983.  Movement of Bacteria
    through Macropores to Ground Water.  Kentucky Water Resource Institute,
    Lexington, KY.  NTIS PB83-246546.
                                    297

-------
                         SECTION 12.  LOCUS NO. 12

                                 CONTENTS
                                                                   Page No.
List of Tables	   300
List of Figures	   300
12.1  Locus Description	   301

      12.1.1  Short Definition	   301
      12.1.2  Expanded Definition and Comments	   301

12.2  Evaluation of Criteria for Remediation	   301

      12.2.1  Introduction	   301
      12.2.2  Mobilization/Remobilization	   301

              12.2.2.1  Partitioning onto Mobile Phases	   301
              12.2.2.2  Transport of /with Mobile Phase	   304

                        Introduction	   304
                        Definition of Mobile Water	   304
                        Permanent Wilting Point	   305
                        Porosity and Field Capacity	   305
                        Hysteresis	   308
                        Darcy' s Law	   308
                        Soil Moisture Profiles	   311
                        Macropore Flow	   313
                        Enhancement of Mobility	   313

      12.2.3  Fixation	   315

              12.2.3.1  Partitioning onto Immobile Phase	   315

                        Impermeable Cover	   315
                        Evaporation	   315
                        Vacuum Extraction/Wind	   315
                        Transpiration	   316

              12.2.3.2  Other Fixation Approaches......	   316

                        Freezing	   316
                        Sorption	   316
                                    298

-------
                                (Continued)


                                                                   Page No.

      12.2.4  Transformation	   316

              12.2.4.1  Biodegradation	   316
              12.2.4.2  Chemical Oxidation	   316
12.3  Storage Capaci ty in Locus	   317

      12.3.1  Introduction	   317
      12.3.2  Guidance on Inputs for, and Calculation of,
                Maximum Value	   317

              12.3.2.1  Total Porosity	   317
              12.3.2.2  Water-Filled Porosity	   317
              12.3.2.3  Solubility	   317

      12.3.3  Guidance on Inputs for, and Calculation of,
                Average Values	   318

              12.3.3.1  Total Porosity	   318
              12.3.3.2  Water-Filled Porosity	   318
              12.3.3.3  Solubility	   318

12.4  Example Calculations	   318

      12.4.1  Storage Capacity Calculations.	   318

              12.4.1.1  Maximum Quantity	   318
              12.4.1.2  Average Quantity	   319

      12.4.2  Transport Rate Calculations	   319

              Depth of Infiltration Water Front	   319

12.5  Summary of Relative Importance of Locus	   320

      12.5.1  Remediation	   320
      12.5.2  Loci Interactions	   321
      12.5.3  Information Gaps	   321

12.6  Literature Cited	   321

12.7  Other References	   322
                                    299

-------
                                  TABLES

                                                                   Page No.

12-1  Representative Porosity Values                                307

12-2  Representative Values for Hydraulic Parameters                311


                                 FIGURES

12-1  Schematic Cross-sectional Diagram of Locus No. 12 -
      Contaminants Dissolved in the Mobile Pore Vater of the
      Unsaturated Zone                                              302

12-2  Schematic Representative of Important Transformation and
      Transport Processes Affecting Other Loci                      303

12-3  Vater Film Surrounding a Vet Solid Particle                   304

12-4  Water-Holding Properties of Various Soils on the Basis of
      their Texture                                                 306

12-5  Moisture Content vs. Capillary Pressure Head                  309

12-6  Changes of Soil Moisture with Depth During Infiltration
      into Sand vith an Initial Moisture Content of 0.5 Percent     312

12-7  Soil Vater Profiles During Drainage of a Silt Loam for
      Various Durations                                             314

12-8  Depth of Infiltration vs. Moisture Content at One Day         320
                                    300

-------
                         SECTION 12 - LOCUS NO. 12
12.1  Locus Description

12.1.1  Short Definition

    Contaminants dissolved in the mobile pore vater of the unsaturated
zone.

12.1.2  Expanded Definition and Contents

    The definition of this locus includes vater occupying a large fraction
of the total porosity of the soil at certain times or places (e.g., after a
heavy rainfall or above the capillary fringe).  The pore vater moves
because of gravity and capillary tension, carrying dissolved contaminants
vith it.  Individual contaminants', however, vill be retarded by sorption to
soil surfaces.

    This locus excludes thin films of tightly-bound vater of very lov
mobility; this situation is discussed in locus no. 3.  Figures 12-1 and
12-2 present a schematic cross-sectional diagram of locus no. 12 and a
schematic representation of the transformation and transport processes
affecting other loci, respectively.

12.2  Evaluation of Criteria for Remediation

12.2.1  Introduction

    The evaluation of criteria for remediation of dissolved liquid contamina-
tion in locus no. 12 is based on the physical and chemical processes vhich
control the movement of the contaminated mobile vater through the unsaturated
zone.  These physical and chemical factors (e.g., hydraulic conductivity,
capillary tension, and solubility) are discussed in Sections 6.2.2 and 6.2.3.
The effects of these physical and chemical properties on the mobilization and
fixation of contaminated vater in locus no. 12 are discussed belov.

12.2.2  Mobilization/Remobilization

12.2.2.1  Partioning into/onto Mobile Phase

    Contaminants in the vater of locus no. 12 are, by definition, mobile.
Contaminants may enter the mobile vater of locus no. 12 by the dissolution
of pure liquid contaminant in the unsaturated zone (locus no. 6).  In the
case of gasoline, those compounds vhich dissolve more readily into vater are
oxygenated additives (e.g., ethanol, methanol, MTBE), phenols, and simple
aromatic hydrocarbons (e.g., benzene, toluene, and xylenes) (V. Lyman, pers.
coram., 1987).  Once dissolved into the mobile water, these contaminants move
vith the mobile vater as a single phase.  The process of dissolution of
liquid contaminant into vater is discussed in locus no. 7.

    Mobile vater of locus no. 12 may become contaminated through contact vith
polluted vater in locus no. 3 as contaminated and pure vaters mix.  Contact
vith soil air (locus no. 1) containing volatilized liquid contaminant may
                                    301

-------
NOTE: NOT ALL PHASE BOUNDARIES ARE SHOWN.
                                 LEGEND
             CONTAMINANT VAPORS

     I       I AIR (PLAIN WHITE AREAS)
             WATER (MOBILE PORE WATER
             IN UNSATURATED ZONE)
       AAAA  WATER FILM
X  BIOTA (MICROBIOTA INCLUDED)

   SOIL PARTICLE

 f COLLOIDAL PARTICLE
                                           CONTAMINANTS DISSOLVED
                                           IN PORE WATER
   Figure 12-1. Schematic Cross-Sectional Diagram of Locus No. 12 -
               Contaminants Dissolved in the Mobile Pore Water of
               the Unsaturated Zone
                                    302

-------
                                  (SOIL GAS)
                                  LOCUS NO.
                             1 - CONTAMINANT VAPORS
     (BIOTA)
    LOCUS NO
11 - SORBED TO BIOTA
                        VOLATILIZATION
              DISSOLUTION

                    DISSOLUTION
                    DEPURATION
                 UPTAKE
(ROCK)
LOCUS NO
10 - DIFFUSED INTO
M'NERAL GRAINS
OR ROCKS
DIFFUSION^

-**
\
                                      (UOUID CONTAMINANTS)
                                           LOCUS NO.
                                    2 - ADHERING TO "WATER DRY-
                                      SOIL PARTICLES
                                    6-JN PORE SPACES IN
                                      UNSATURATED ZONE
                                    7 - FLOATING ON WATER
                                      TABLE
                                    13 - IN ROCK FRACTURES
                                                       PHASE
                                                       SEPARATION
                               LOCUS NO. 12
                            (CONTAMINANTS DISSOLVED
                              IN MOBILE PORE WATER
                              IN UNSATURATED ZONE)
                 DESORPTION
     (SOBBED CONTAMINANTS)
           LOCUS NO.
    f. • SORBED TO -WATER-WET*
      SOIL PARTICLES
    0 - SORBED TO COLLOWAL
      PARTICLES
SORPTtON
              ADVECTTON
              DIFFUSION
              DISPERSION
              RETARDATION OF
              CONTAMINANTS
      (WATER)
      LOCUS NO.
3- DISSOLVED IN WATER
  FILM
                                   (WATER)
                                  LOCUS NO.
                                8- DISSOLVED IN
                                  OROUNDWATER
Figure 12-2. Schematic Representation of Important Transformation
            and Transport Processes Affecting Other Loci.
                                       303

-------
  also contaminate  the  mobile water of locus no. 12.  The compounds which most
  readily partition from  the air phase to the mobile water phase are those
  compounds  with  lower  Henry's Law Constants (e.g., benzene, toluene, xylenes);
  those compounds with  high Henry's Law constants (e.g., 2-methylhexane,
  n-pentane,  2,2,5,5-tetramethylhexane) present in the mobile water of locus
  no.  12 partition  more readily into the soil air of locus no. 1.  The process
  of air/water  partitioning is discussed in detail in locus no. 3.

      Partitioning  to the mobile water of locus no. 12 can result from contact
  with soil  particles onto which the liquid contaminant adheres (locus no. 2)
  or into which the liquid contaminant has diffused (locus no. 10).  Contact
  with soil  particles serves also as a mechanism for removal of contaminants
  from the mobile water.  Finally, the mobile water of locus no.  12 can become
  polluted by contact with contaminated water adhering to the soil particles
  (locus no.  3),  with water containing contaminated microbiota (locus no. 11),
  or with water containing contaminated colloids (locus no.  9).

  12.2.2.2 Transport of/with Mobile Phase

  Introduction

      Locus no. 12  concerns the movement of contaminants dissolved in the
  mobile water  of the unsaturated zone.  This discussion assumes  that there are
  two  immiscible  fluid  phases (i.e.,  air and water) present  in the unsaturated
  zone.   The  definition of mobile water in the unsaturated zone and the
  influences  on its movement in the unsaturated zone will be discussed below.

  Definition  of Mobile Water

      In  the  unsaturated zone,  a continuum in capillary tension exists between
  water  held  at zero capillary tension (i.e.,  gravity drainage) and water held
  by very  high  capillary tensions to  soil grain surfaces (Figure  12-3).
  THIN FILM OF WATER
    ADHERING VERY
      TIGHTLY TO
     SURFACE OF
      SOIL GRAIN        J ^JLJ_m^pin^,  —«_•
                        ™«^i™  —™           PARTLY DRAINED
                                                             SOIL PORE
    SOIL PARTICLE
                                                          WATER HELD IN
                                                         NARROW NECKS
                                                          OF SOIL PORES
Source: Adapted from Dunne & Leopold, 1978

           Figure 12-3. Water Film Surrounding a Wet Solid Particle
                                     304

-------
    Water, held at gravity moves freely downward through the unsaturated
zone; water held by high capillary tension is relatively immobile.  Because
of the range in capillary tension, a continuum in mobility also exists
between water draining by gravity and immobile water held by capillary
tension.

    As a consequence of this continuum in water mobility, it is difficult to
assign a natural boundary between mobile and immobile water.  Nevertheless,
an arbitrary lower limit to the volume of mobile water can be defined, below
which the volume of water can be considered relatively immobile in the
unsaturated zone.

Permanent Wilting Point

    A convenient point for this arbitary boundary between mobile and
immobile water is the permanent wilting point.  The permanent wilting point
is the soil moisture condition at which plants wilt.  This occurs when
plants can no longer remove water from the soil; the remaining water is too
tightly held by the soil particles.  Since most plants can exert up to 15
atm of capillary tension (155.1 m H^O) (Dunne and Leopold, 1978), the volume
of water remaining in the soil at capillary tensions equal to or greater
than 15 atm will be considered immobile (locus no. 3).  The immobile water
occurs as a thin film adhering to the surface of soil grains (Figure 12-3).

    Because of natural variation in size and distribution of grains of
undisturbed soils, the volume of water held at the wilting point varies
with soil type (Figure 12-4).  Figure 12-4 indicates that sandy soils hold
about five percent water by volume at the wilting point; at the fine end of
the grain-size spectrum, clayey soils hold as a much as twenty five percent
water by volume.  The increase in water content with decreasing grain-size
occurs because the smaller pores in finer soils hold water more firmly.
The dependence of capillary tension on saturation is discussed in section
6.2.2.1 of locus no. 6.

Porosity and Field Capacity

    Figure 12-4 also shows the variation in porosity and field capacity
with soil type; additional porosity data are given in Table 12-1.
Rainwater infiltrates into the soil and fills empty pore spaces.  The
maximum amount of water that the soil can hold equals the non-solid volume
of the soil, i.e., the total porosity; at this point the soil is saturated.
After the rainstorm, the moisture content declines from saturation.

    Drainage of the soil moisture under gravity continues until capillary
tension balances gravity.  At this point, drainage ceases and the soil is
at field capacity.  Porosity and field capacity define upper limits to the
volume of mobile water in locus no. 12.

    The lower limit on the volume of mobile water was arbitrarily set at 15
atm; the maximum volume of mobile water is defined by the porosity of the
soil.  These limits on the volume of mobile water range widely from 5 to
48.7 percent for sandy soils; the range of 20 to 55 percent is narrower for
clayey soils (Figure 12-4).
                                    305

-------
       .60
       .50
      .40
      .30
c
CD
a
CD
0.
.20
      .10
        01—
         Sand
                                                                Porosity
                                                                Field
                                                                Capacity
                                                                      , Wilting
                                                                      'Point
                                                        Clay
                                                    Heavy
                                                    clay loam
                                                    Clay loam
                            Fine
                            sandy loam
                                           Light clay loam
                                            Silt loam
                Sandy loam
                 Fine sand
 Source: Adapted from Dunne and Leopold, 1978

 Figure 12-4.  Water-Holding Properties of Various Soils on the Basis of Their Texture.
                                     306

-------
                                TABLE 12-1

                      REPRESENTATIVE POROSITY VALUES
  Material
Porosity,
 Percent
        Material
Porosity,
 Percent
Gravel, coarse

Gravel, medium

Gravel, fine

Sand, coarse

Sand, medium

Sand, fine

Silt

Clay

Sandstone, fine-grained

Sandstone, medium-grained

Limestone

Dolomite

Dune sand
  28a

  32a

  34a

  39-

  39

  43

  46

  42

  33

  37

  30

  26

  45
Loses                        49

Peat                         92

Schist                       38

Siltstone                    35

Claystone                    43

Shale                         6

Till, predominantly silt     34

Till, predominantly sand     31

Tuff                         41

Basalt                       17

Gabbro, weathered            43

Granite, weathered           45

Granite, weathered           45
a. These values are for repacked samples; all others are undisturbed.

Source:  Pettyjohn and Haunslow (1982).
                                    307

-------
    The condition of maximum mobile water content occurs while rainwater is
infiltrating into the soil and is transient.  After the rainstorm has
ceased, the soil water drains under gravity, until field capacity is reach-
ed.  The water remaining in the soil at field capacity may still move under
gravity and capillary tension.  The range in mobile water volume between
field capacity and at 15 atm limit for sandy soils is narrow, 5 to 9
percent; for clayey soils, the range is from 25 to 37 percent (Figure 12-4).

Hysteresis

    The relationship between saturation of a soil and the capillary tension
at which the water is held is not unique, but depends on the wetting
history of the soil.  More water is held in the soil during drainage than
in wetting at the same capillary tension (Figure 12-5).  This phenomenon is
known as hysteresis.

    Hysteresis is most pronounced in coarse-grained soils at low capillary
tensions and has been attributed to several possible causes (Hillel, 1980).
Although the discussion of the causes of hysteresis is beyond the scope of
this report, the phenomenon is important because it affects unsaturated
hydraulic conductivity as well as the moisture content of the soil.

Darcy's Lav

    The flow of water and transport of solutes in the unsaturated zone is
governed by Darcy's Law.  This empirical law relates the discharge of water
through a porous medium to the product of the hydraulic conductivity of the
soil and the hydraulic gradient acting on the water.  Hydraulic
conductivity measures the resistance of a porous medium to water flow and
depends on physical properties of the porous medium and of water; hydraulic
gradient depends on gravity and pressure.  A detailed discussion of Darcy's
Law and its variables is presented in section 6.2.2.2 of locus no. 6.

    Numerical and analytical models of water flow through the porous medium
of soils are available which predict the transport of an infiltrating front
or draining of water front through a soil.  The three-dimensional equation
for unsaturated water flow through a porous medium is (Hillel, 1980):

    3§     _ !_ (K 8*1 - i- fifi*l   1- fclSl + i£
    at  =    ax I axj   ay I ayJ "" azl azj    az                        (12. i)


                                     3   3
where        0 = moisture content (cm /cm )

             t = time (sec)

         x,y,z = space coordinates (cm)

             
-------
Q
IU
oc


2
IU
c
a.
IMBIBITION (OR WETTING)





. BOUNDARY DRYING CURVE
                BOUNDARY


                WETTING CURVE
                                                              DRAINAGE (OR DRYING)
                                                               STARTING WITH A


                                                               SATURATED SAMPLE
                       MOISTURE CONTENT
                                    ENTRAPPED AIR

                                    (MAY BE REMOVED WITH TIME)
       Source: Bear, 1979.  (Copyright McGraw-Hill Inc. 1979. Reprinted with permission.)



        Figure 12-5. Moisture Content vs. Capillary Pressure Head
                                  309

-------
vertically downward, since this water may recharge aquifers at depth.  A
relatively simple, analytical model can be derived to predict the vertical
movement of water with time.  This model is based on the one-dimensional
continuity equation (Hillel, 1980):

                             39      83
                             3t  " ~ 9z                              (12.2)

                                 3   3
where    9 = moisture content (cm /cm )

         q = discharge per unit area (cm/sec)
         z = depth (cm)

         t « time (sec)

    Using the definition of a partial differential equation (Svokowski,
1975) and re-arranging equation 12.2:

                             dz     3
                             dt  "  6                                (12.3)

    The discharge of water through a porous medium is described by Darcy's
Law:
               q . (k.kr/y).(dP/dz + p.g.dh/dz)                      (12.4)

                                             2
where         k  = intrinsic permeability (cm )
              k  = relative permeability (dimensionless)
               r                                 3
              p  = density of water, p * 1.0 g/cm
                                            2
              g  • gravity, g » 980.7 cm/sec

              y  = dynamic viscosity (g/cm.sec)
                                                       -2   3
          dP/dz  = vertical pressure gradient (g.cm.sec  /cm )

          dh/dz  = vertical head gradient (cm/cm)

    These parameters are described in detail in locus no. 6.  If no
pressure gradient exists, and a unit head gradient is assumed, equation
12.4 becomes:

                q  =  p.g.k.k /y  -  K .k                            (12.5)
                             i        Si

where    K  = saturated hydraulic conductivity (cm/sec)
          s

    An expression for the relative permeability is reported by Clapp and
Hornberger (1978):

                kr  «  (6/es)2b+3                                    (12.6)

                                                3   3
where    0  = moisture content at saturation (cm /cm )
          S
         b  = grain-size distribution parameter (dimensionless)

    Selected values of 9 ,b,V » and K  are given in Table 12-2.
                        s    s       s
                                    310

-------
                                TABLE 12-2

              REPRESENTATIVE VALUES FOR HYDRAULIC PARAMETERS
Soil Texture
Sand
Loamy sand
Sandy loam
Silt loam
Loam
Sandy clay loam
Silty clay loam
Clay loam
Sandy clay
Silty clay
Clay
b
4.05
4.38
4.90
5.30
5.39
7.12
7.75
8.52
10.4
10.4
11.4
Ys
(cm)
12.1
9.0
21.8
78.6
47.8
29.9
35.6
63.0
15.3
49.0
40.5
3®s 3
(on /on )
0.395
0.410
0.435
0.485
0.451
0.420
0.477
0.476
0.426
0.492
0.482
Ks
(cm/min)
1.056
0.938
0.208
0.0432
0.0417
0.0378
0.0102
0.0147
0.0130
0.0062
0.0077
Source:  Adapted from Clapp and Horberger (1978)

    Substituting equation 12.5 and 12.6 into equation 12.3 gives:

               dz/dt  =  (K /0).(0/e
                                                                     (12.7)
which may be integrated to give the depth, z(t), of the infiltrating water
front at time t:
               z(t)
                           .t/ej.o/ej
                                     0
                                       2b+3
(12.8)
    This simple, analytical expression assumes that capillary forces are
negligible compared to gravity transport in the downward vertical movement
of water.  This assumption will be most reasonable for a moderately
well-saturated, sandy soil.

Soil Moisture Profiles

    As described in detail in section 6.2.2.2, the hydraulic conductivity
of the soil increases with moisture content to a maximum at saturation.   As
rainwater infiltrates into a soil, the moisture content increases as pores
fill with water.  At complete saturation, the infiltration rate is limited
by the saturated hydraulic conductivity.

    Figure 12-6 shows the change in soil moisture with depth during
infiltration of water into a sand (Dunne and Leopold,  1978).  The
isochrones (i.e., lines of equal time) indicate that the time for water to
reach a given depth decreases with increasing moisture content; a vetter
soil allows faster infiltration.

    The effect of rainfall intensity is seen also in Figure 12-6.  Figure
12-6b was produced at a higher rainfall intensity than Figure 12-6a.  The
result is that the soil in Figure 12-6b holds more water and percolation is
                                    311

-------
(a)
              2
              ¥
              o
       0
      10
      20
      30
      40
      50
      60
                                                         i    I    r
                                                         I	I
                      0  2   4   6   8   10  12  14  16  18 20  22  24
                              Soil Moisture Content (percent volume)
(b)
2
4>
O

1
i.
 o
10
20
30
40
50
60
                     0   2   4   6   8  10  12  14  16  18  20  22  24
                              Soil Moisture Content (percent volume)
   NOTE: Rainfall intensity in (b) is 3.8 times greater than in (a). As a result, the soil moisture
          content had to be greater (about 24 percent in (b) versus 16 percent in (a)) in order
          to transmit the water at the applied rate (saturated moisture content = 38 percent).
   Source: Rubin, 1966.
  Figure 12-6. Changes of Soil Moisture with Depth During Infiltration into Sand
                with an Initial Moisture Content of 0.5 Percent.
                                             312

-------
faster.  If the rainfall intensity exceeds the saturated hydraulic conduc-
tivity, however, ponding of water on the soil surface and runoff occurs.

    After infiltration ceases, percolation slows as the moisture is
re-distributed throughout the soil depth (Figure 12-5).  Gravity drainage
declines with time, until field capacity is reached (Figure 12-7).

Nacropore Flov

    It has been noted that natural soils commonly contain preferential
flowpaths, known as macropores, which are formed by a variety of processes
including activity of soil fauna, plant roots, natural soil piping, and
cracks and fissures (Seven and Germann, 1982).  Seven and Germann (1981)
suggested that a macropore would be any soil pore where the capillary
tension was larger than -0.1 kPa (0.1 cm H.O); the diameter of this
arbitrary macropore would be larger than 0.3 cm.  Other definitions of
macropores are reported by Seven and Germann (1982).

    The presence of macropores in soil allows the transmission of water
very quickly, by-passing the matrix of soil grains through which water
moves more slowly by percolation.  An example of the rapid flow of
macropores was observed during a ponded infiltration experiment on a forest
soil, in which the flow of water from an empty tree root located 2 meters
from the measurement site occurred within 3 minutes after initiation of
ponding (Levy, 1987).  A common field method of determining the macropore
and matrix flow of a natural soil is the measurement of infiltration of
water into the soil under conditions of ponded water and under capillary
tension (Watson and Luxmoore, 1986).

    Macropores can transmit large volumes of water through the soil, even
though they may comprise only a small fraction of the total soil volume.
For example, measurements of macropore flow in a forest soil by Watson and
Luxmoore (1986) indicate that 73 percent of the water ponded on the soil
surface was carried by macropores.  Calculations suggest that these
macropores have a diameter of > 0.1 cm and comprise less than 0.1 percent
of the soil volume (Watson and Luxmoore, 1986).

    The presence of macropores can reduce infiltration of water through the
porous medium of the soil from the surface; a remedial program of artifi-
cial recharge may fail to clean the contaminated water from the soil as a
result.  The phenomenon of macropore flow is likely to occur during high
intensity rainstorms, increasing in importance to flow as the rainfall
intensity increases.

Enhancement of Mobility

    The mobility of water in locus no. 12 cannot be enhanced easily by
changing the physical properties of the porous medium or of water.  The
reasons for this limitation is discussed in section 6.2.2.3.  The mobility
of water in locus no. 12 can be enhanced, however, by increasing the
moisture content of the unsaturated zone.
                                    313

-------
            0
       2
       fl>
       s
       I
            10
20
            30
            40
              0.0    0.1
                       OJ     0.4
                     Soil Water Content (g/cm)
Source: Biswas aLaL, 1966.


Figure 12-7.  Soil Water Profiles During Drainage of a Silt Loam for Various

                     Durations, hours (h) or days (d).
                                  314

-------
    An increase in saturation in locus no. 12 by the addition of clean
water increases the hydraulic conductivity.  The increase in hydraulic
conductivity allows contaminated water to percolate more readily.  Enough
water must be added as clean infiltrate to enhance the mobility of the
water present in locus no. 12.  Too much clean water added to the soil in a
short time period, however, will initiate macropore flow; much water may
by-pass the contaminated soil matrix with no resulting remediation.

12.2.3  Fixation

12.2.3.1  Partitioning onto Immobile Phase

    Several methods are possible alternatives by which the contaminants
dissolved in the mobile water of locus no. 12 may become immobilized.
These alternatives are described below in order of decreasing feasibility.

Impermeable Cover

    Of the various fixation alternatives, emplacement of an impermeable
cover over the land surface above a contaminated site is the most easily
effected.  The presence of the impermeable cover would prevent recharge of
infiltrating rainwater to the unsaturated zone.  Interdiction of recharge
to locus no. 12 would prevent an increase in hydraulic conductivity with
increased moisture content, preventing any further downward migration of
contaminated water.

Evaporation (Details in Section 3)

    The reduction of hydraulic conductivity, and subsequent immobilization
of contaminated water, can also be accomplished by the removal of water by
evaporation.  The removal of water from locus no. 12 by evaporation
increases the capillary tension exerted by the porous medium on the
remaining water.  The remaining water in locus no. 12 will be held more
firmly and will be consequently more immobile.

    Evaporation of water from locus no. 12 naturally occurs if the soil
surface is open to the atmosphere.  Evaporation is most pronounced in sandy
soils, and is largest during summer months when monthly average
temperatures are the highest.  Because of its dependence on temperature,
evaporation could be enhanced by increasing the water temperature of locus
no. 12; this enhancement may be accomplished by the pumping of dry, heated
air into the ground through wells.

Vacuum Extraction/Wind

    A more practical method of enhancing evaporative water loss from locus
no. 12 would be to apply vacuum extraction to the soil, removing
water-saturated air.  More water evaporates into the new, dry soil air
which replaces the evacuated, moist air.   Evaporation increases, and the
remaining water in locus no. 12 is immobilized.

    Forced evaporation may result in the deposition of calcium carbonate and
other naturally occurring salts in the upper sub-surface of the soil.
Precipitation of these salts from evaporating water may reduce the hydraulic
                                    315

-------
conductivity of the soil sub-surface with time.  This phenomenon occurs
naturally, producing a very hard, impermeable crust in salty, desert soils
known as a caliche.  The presence of caliche in a soil can prevent infiltra-
tion of water downward, acting in the same manner as an impermeable cover.

Transpiration

    Moisture is removed from locus no. 12 by the plant cover on the soil
surface, causing an increase in capillary tension exerted on the remaining
water.  The hydraulic conductivity of locus no. 12 could be reduced by
covering the site with plant species that transpire large quantities of
water (e.g., alfalfa, soybeans) (Rosenberg et al., 1983).

12.2.3.2  Other Fixation Approaches

    The following fixation approaches are suggested for immobilization of
contaminated water in locus no. 12, despite their apparent impracticality.

Freezing

    Very simply, if the mobile water of locus no. 12 could be frozen, it
would be effectively immobilized.  Large-scale freezing of the unsaturated
zone in areas other than in tundra must be highly unfeasible.

Sorption

    It is unlikely that an enhancement of the sorptive properties of the
porous medium of locus no. 12 could be achieved.

12.2.4  Transformation

12.2.4.1  Biodegradation (Details in Section 11)

    Of all loci, locus no. 12 probably holds the greatest potential for
biodegradation.  Due to its location in the vadose zone and contact with
infiltrated rainwater, the oxygen content is higher than in any other
locus.  Because the phase is aqueous and mobile, transport of all
substrates (oxygen, carbon, minerals) is maximized.  Nevertheless, oxygen
or nitrogen concentration may still be rate-limiting.  Since concentrations
are subject to rainfall, wide variations in oxygen or other substrates may'
be encountered.  Rates of natural biodegradation may be enhanced by the
application of oxygenated waters containing hydrogen peroxide and nutrients
through injection wells.

12.2.4.2  Chemical Oxidation (Details in Section 3.2.4.2)

    Soil water provides the optimum mechanism for chemical oxidation of
contaminants since this media is subject to microbial activities, is
oxygenated by precipitation entering the system and by diffusion, and comes
in contact with catalysts in the soil.
                                    316

-------
12.3  Storage Capacity in Locus

12.3.1  Introduction and Basic Equations

    The mass of liquid contaminant dissolved into the vadose zone water is
calculated from:

                             mw  -  S.0W/103                         (12.9)

where    m  = mass per unit volume of .-liquid contaminant dissolved in the
              vadose zone water (kg/m )

         S  - solubility of liquid contaminant in water (mg/L)
                                       3   3
         6  = water-filled porosity (cm /cm )

    The water-filled porosity in locus no. 12 is the difference between the
total and air-filled porosity:

                             ew  -  et - ea                         (12.io>
                                     3   3
where    8  = total soil porosity (cm /cm )
                                     33
         9  = air-filled porosity (cm /cm )
          cl

12.3.2  Guidance on Inputs for, and Calculations of, Maxima Value

12.3.2.1  Total Porosity, 0 .

    o   Theoretical maximum porosity for sandy soils is 47.6%.

    o   User should refer to Table 12-1 or 12-2 for representative total
        porosity values for a variety of soils.

12.3.2.2  Water-Filled Porosity, 6 .

    o   The maximum quantity of dissolved liquid contaminant held in locus
        no. 12 occurs when the water-filled porosity is at a maximum; the
        air-filled porosity is at a minimum.  It is assumed that the water
        content in the soil is the total porosity of the soil.

    o   Water-filled porosity ranges from 0 to 0t«

12.3.2.3  Solubility, S.

    o   User should supply the solubility of liquid contaminant.

    o   For gasolines with no oxygenated additives, a value of 200 mg/L for
        solubility is suggested.  Refer to locus 3 for detailed information
        on solubility.
                                    317

-------
12.3.3  Guidance on Inputs for, and Calculations of, Average Value

12.3.3.1  Total Porosity, 6 .

    o   User should refer to Table 12-1 or 12-2 for representative total
        porosity values for a variety of soils.

12.3.3.2  Water-Filled Porosity, 8y.

    o   The average quantity of dissolved liquid contaminant held in locus
        no. 12 occurs when the water-filled porosity is at field capacity.
        Suggest using 0.09 for water-filled porosity for a sandy soil at
        field capacity.

    o   Water-filled porosity ranges from 0 to 9 .

12.3.3.3  Solubility, S.

    o   User should supply the solubility of liquid contaminant.

    o   For gasolines with no oxygenated additives, a value of 100 mg/L for
        solubility is suggested as half the value reported in section
        12.3.2.  Refer to locus 3 for detailed information on solubility.

12.4  Example Calculations

12.4.1  Storage Capacity Calculations

12.4.1.1  Maximum Quantity

    For this sample calculation, a sandy soil is assumed.  Since we wish  to
maximize the amount of dissolved contaminant liquid, we maximize the water
content in the soil at the total porosity of the soil.  For sandy soils,
the water content at the total porosity is about 39.5% (Clapp and
Hornberger, 1978).  The air-filled porosity is zero.

                             ey = et - ea
                             6t = 0.395

                             e  = o.oo
                              £1
                             ew = 0.395

    In computing the maximum quantity of dissolved liquid in locus no. 12,
it is assumed that the solubility is at a maximum.   A value of 200 mg/L for
S as reported in section 12.3.2 for gasoline will be used to calculate the
maximum mass per unit volume of dissolved contaminant liquid held in locus
no. 12 from Equation 12.9.

         m  = S.6 /103
          w      w                 _
         mw = (200 mg/L).(0.395)/KT
         m  = 0.08 kg/m3
                                    318

-------
12.4.1.2  Average Quantity

    A sandy soil is also assumed for the calculation of the average
quantity of dissolved contaminant liquid held in locus no. 12.  It is
assumed that the water-filled porosity is at field capacity.  For a sandy
soil, the water-filled porosity at field capacity is about 9%.  As shown in
Table 12-2, the total porosity is 39.5%.


         9w - 6t - 6a
         et = 0.395

         ea = 0.09

         ey . 0.305

    In computing the average quantity of dissolved liquid in locus no. 12,
it is assumed that the solubility,. is half that reported in section 12.3.2.
A value of 100 mg/L for S will be used to calculate the average mass per
unit volume of dissolved contaminant liquid held in locus no. 12:

         mw = s.ew/io3
         mw . (100 mg/L).(0.305)/103
         my = 0.03 kg/m3

12.4.2  Transport Rate Calculation

Depth of Infiltrating Water Front

    The depth of an infiltrating water front can be estimated using
equation 12.8.



                                    3
    This equation was derived under the following assumptions:  no pressure
gradient; unit hydraulic head; negligible capillary forces; 'and, uniform
vertical moisture distribution.  If time is one day, and appropriate values
from Table 12-2 are used, values for the depth of an infiltrating water
front for different types of soil can be found.  Figure 12-8 shows
graphically the effect of moisture content on the depth of infiltration for
three soil types.  As can be expected, sand allows for deeper penetration
than the other soil types.  Also illustrated is the important role the
moisture content plays in the depth of infiltration.
                                    319

-------
                                                       -O-
                                                             Clay

                                                             Loam

                                                             Sand
            0.30    0.32    0.34     0.36    0.38    0.40
                        Moisture Content
   Figure 12-8. Depth of Infiltration Versus Moisture Content at One Day.
12.5  Summary of Relative Importance of Locus

12.5.1  Remediation

    Contaminated mobile water in che unsaturated zone can contaminate
groundwater (locus no. 8) by mixing.  Partitioning between the water and
air phases may contaminate the soil air (locus no. 1).  The extent of
contamination of the mobile water depends on the solubility of the liquid
contaminant in water.

    The relative importance of locus no.  12 depends on the solubility of
the liquid contaminant and the moisture content of the soil.   If locus no
12 holds a maximum mass per unit volume,  it probably holds more mass per
unit volume than loci no.'s 4,9,10 and 11.   Locus no.  12 should hold the
same mass per unit volume of water as loci  no. 3 and 8.   Because locus no
12 holds more water than locus no. 3, it  will also hold more mass;
similarly, locus no.  12 holds less mass than locus no. 8.   Under the right

                                        320

-------
conditions, (e.g., low volatility, low Henry's Lav constant, locus no. 12
may be as important as locus no. 1.  When the contaminant liquid is
sparingly soluble in water, locus no. 12 will hold less mass per unit
volume than loci no. 2,5,6,7 and 13.

    If the contaminated mobile water in locus no. 12 occupies a large
fraction of the unsaturated zone porosity, and if the soil permeability is
sufficiently high, then the contaminated water may be fairly mobile.
Pumping may remove much of the contaminated water in this case.

12.5.2  Loci Interactions

    Dissolution from liquid contaminant in locus no. 6 represents the
primary mechanism for partitioning into this locus.  Contaminants may
partition into the air phase (locus no. 1) if the Henry's Law constants are
high enough.  Soil particles may attenuate contamination by retardation and
sorption (loci no.'s 2 and 4).  If the contaminated mobile water is in
sufficient quantity, it may migrate vertically, contaminating the
groundvater (locus no. 8).  It may be aided in this regard by mobile
colloids (Locus no. 9).

12.5.3  Information Gaps

    Research into the following area is important to better define the fate
of contaminated unsaturated soil water:

    1.  Attenuation of hydrocarbons from water to soil particles.

    2.  Exchange rates between air and water phases.

12.6  Literature Cited

Bear, J.  1979.  Hydraulics of Groundwater.  McGraw-Hill, Inc., New York.

Beven, K. and P.P. Germann.  1981.  Water Flow in Soil Macropores, 2, A
    Combined Flow Model.  Journal of Soil Science, 32:15-29.

Beven, K. and P.P. Germann.  1982.  Macropores and Vater Flow in Soils.
    Water Resources Research, 18(5):1311-1325.

Biswas, T.D., D.R. Nielsen and J.W. Bigger.  1966.  Redistribution of Soil
    Water after Infiltration.  Water Resources Research, 2:513-524.

Clapp, R.B. and G.M. Hornberger.  1978.  Empirical Equations for Some Soil
    Hydraulic Properties.  Water Resources Research, 14(4):601-604.

Dunne, T. and L.B. Leopold.  1978.  Water in Environmental Planning.  W.H.
    Freman & Co.

Gilliam, R.W., A. Klute, and D.F. Heermann.  1979.  Measurement and
    Numerical Simulation of Hysteretic Flow in a Heterogeneous Porous
    Medium.  Soil Science Society of America Journal, 43(6):1061-1067.
                                    321

-------
Giesel, V., M. Renger, and 0. Strebel.  1972.  Numerical Treatment of the
    Unsaturated Water Flow Equation:  Comparison of Experimental and
    Computed Results.  Water Resources Research, 9(1):174-177.

Hillel, D.  1980.  Fundamentals of Soil Physics.  Academic Press.

Levy, B.S.  1987.  Kinematic Wave Approximation to Solute Transport in a
    Forest Soil.  M.S. Thesis, University of Virginia.

Pettyjohn, W.A. and A.W. Haunslov.  1982.  Organic Compounds and
    Groundvater Pollution.  In: Proceedings of 2— National Symposium on
    Aquifer Restoration and Ground Water Monitoring.  D.M. Nielsen
    (editor).  National Water Well Association.

Rosenberg, N.J., B.L. Verma, and S.B. Verma, 1983.  Microclimates:  the
    Biological Environment.  2nd ed.  Wiley - Interscience.

Rubin, J. 1966.  Theory of Rainfall Uptake by Soils Initially Drier than
    their Field Capacity and its Applications.  Water Resources Research,
    2:739-750.

Svokovski, E.W.  1975.  Calculus with Analytic Geometry.  Prindle, Weber
    and Schmidt.

Watson, K.W. and R.J. Luxmoore.  1986.  Estimating Macroporosity in a
    Forest Watershed by Use of a Tension Infiltrometer.  Soil Science
    Society of America Journal, 50:578-582.

12.7  Other References

Baehr, A.L., G.E. Hoag, and M.C. Marley.  1989.  Removing Volatile
    Contaminants from the Unsaturated Zone by Inducing Air-phase Transport.
    Journal of Contaminant Hydrology, 4, 1-26.

Baek, N.H., L.S. Clesceri, and N.L. Clesceri.  1989.  Modeling of Enhanced
    Biodegradation in Unsaturated Soil Zone.  Journal of Environmental
    Engineering, 115(1):150-172.

Bouyoucos, G.L. 1915.  Effect of Temperature on the Movement of Water Vapor
    and Capillary Moisture in Soils.  Journal of Agricultural Research,
    5:141-172.

Hutzler, N.J., J.S. Gierke, and L.C. Krauss.  1989.  Movement of Volatile
    Organic Chemicals in Soils.  In: Reactions and Movement of Organic
    Chemicals  in Soils, Soil Science Society of America Special Publication
    No. 22. pp. 373-403.

Jury, W.A., W.F. Spencer, and W.J. Farmer.  1983.  Behavior Assessment
    Model  for  Trace Organics in Soil, I. Model Description.  Journal of
    Environmental Quality, 12(4).
                                    322

-------
Krearner, O.K., G.M. Thompson, E.S. Simpson, K. Stetzenbach, and A. Long,
    1985.  Seasonal, Diurnal, and Storm-related Changes in Concentrations
    of Trichloroethylene (TCE) and Other Volatile Organic Compounds in the
    Unsaturated Zone.  Report for USDI, USGS - Arizona Vater Resources
    Research Center.

Marrin, D.L. 1984.  Investigation of Volatile Contaminants in the
    Unsaturated Zone above TCE-polluted Groundwater.  U.S. EPA,
    #CR811018-01-0.

Vagenet, R.L., J.L. Hutson, and J.V. Biggar.  1989.  Simulating the Fate of
    a Volatile Pesticide in Unsaturated Soil:  A Case Study vith DBCP.
    Journal of Environmental Quality, 18(l):78-84.

Wierenga, P.J. and M.T. Van Genuchten, 1989.  Solute Transport through
    Small and Large Unsaturated Soil Columns.  Ground Vater, 27(1):35-42.

Wilson, J.T., C.G. Enfield, V.J. Dunlap, R.L. Cosby, D.A. Foster, and L.B.
    Baskin, 1981.  Transport and Fate of Selected Organic Pollutants in a
    Sandy Soil.  Journal of Environmental Quality, 10(4):501-506.
                                    323

-------
                         SECTION 13.  LOCUS NO. 13

                                 CONTENTS
                                                                   Page No.
List of Figures	   325
13.1  Locus Description	   326

      13.1.1  Short Definition	   326
      13.1.2  Expanded Definition and Comments	   326

13.2  Evaluation of Criteria for Remediation	   326

      13.2.1  Introduction	   326

              13.2.1.1  Fractures and Dissolution Zones	   326

                        Distinction Between Primary and
                          Secondary Porosity	   326
                        Sources and Types of Fractures	   329
                        Importance of Locus	   329
                        Fracture Characteristics	   329
                        Fracture Density	   329
                        Jracture Width	   330
                        Fracture Orientation and Connectivity....   330
                        Fracture Porosity	   330
                        Level of Present Understanding	   332

      13.2.2  Mobilization/Remobilization	   333

              13*2.2.1  Partitioning into Mobile Phases	   333
              13.2.2.2  Transport of/with Mobile Phase	   333

                        Saturated Flow	   333
                        Unsaturated Flow	   335

      13.2.3  Fixation	   335

              13.2.3.1  Partitioning onto Immobile
                          (Stationary) Phase	   335
              13.2.3.2  Other Fixation Approaches	   337

      13.2.4  Transformation	   337

              13.2.4.1  Biodegradation	   337
              13.2.4.2  Chemical Oxidation	   337
                                    324

-------
                                (Continued)

                                                                   Page
13.3  Storage Capacity in Locus .......................... ........   337

      13.3.1  Introduction .......................................   337
      13.3.2  Guidance on Inputs for, and Calculation of,
                Maximum Value ....................................   338
      13.3.3  Guidance on Inputs for, and Calculation of,
                Average Values ...................................   338

              Unsaturated Zone ...................................   338

13.4  Example Calculations .......................................   339

      13.4.1  Storage Capacity Calculations ......................   339

              13.4.1.1  Maximum Quantity .........................   339
              13.4.1.2  Average Quantity at Unsaturated Conditions  340
              13.4.1.3  Average Quantity at Saturated Conditions.   340

      13.4.2  Transport Rate Calculations .......................    340

              Parallel-Plate Approximation ......................    341

13.5  Summary of Relative Importance of Locus ....................   341

      13.5.1  Remediation ........................................   341
      13.5.2  Loci Interaction ...................................   342
      13.5.3  Information Gaps ...................................   342

13.6  Literature Cited ...........................................   342

13.7  Additional Reading .........................................   343

                                  FIGURES

13-1  Schematic Cross-sectional Diagram of Locus No. 13 - Liquid
      Contaminants in Rock Fractures in Either the Unsaturated
      or Saturated Zone                                             327

13-2  Schematic Representation of Important Transformation and
      Transport Processes Affecting Other Loci                      328

13-3  Map of Fracture Spacings and Orientations from Borehole
      Records                                                       331

13-4  Depiction of Strike,  Dip Direction,  and Dip                   332

13-5  Saturation Versus Pressure Head for Fractures and Rock
      Matrix                                                        336

13-6  Effective Flow Areas  in Fractures as a Function of Pressure   336


                                    325

-------
                         SECTION 13 - LOCUS NO. 13
13.1  Locus Description

13.1.1  Short Definition

    Liquid contaminants in fractured rock or karstic limestone in either
the unsaturated or saturated zone.

13.1.2  Expanded Definition and Comments

    The presence of liquid contaminant in the fractures of rock or the
dissolution cavities of karstic limestone distinguishes this locus from
loci no's. 2, 5, 6 and 7, which discuss liquid contaminants present in
granular, consolidated and unconsolidated porous media.  Essentially, this
locus discusses the presence of liquid contaminants in the secondary
porosity of consolidated rocks; the presence of liquid contaminants in the
primary porosity of consolidated rocks is discussed in locus no. 10.

    Figure 13-1 presents a schematic cross-section of the locus; figure
13-2 is a schematic representation of the transformation and transport
processes affecting other loci.

13.2  Evaluation of Criteria for Remediation

13.2.1  Introduction

    The presence of fractured rock or karstic limestone at a leaking UST
site may mean greater difficulty in locating, modeling, and remediating the
contamination than if no fractions or dissolution zones were present.
Rocks with fracture zones or dissolution zones are typically very
permeable, and leaked product can travel very quickly from the spill site.
Many fracture or dissolution zones are highly anisotropic, and thus cannot
be simulated using models which assume an isotropic medium.

    Remediation is also made more difficult by the presence of fractures
and dissolution zones.  Recovery of free product is generally very low.
Garrett (1987) reports recoveries from six sites that range up to only 70
gallons; up to 6000 gallons were recovered from eleven sand and gravel
sites.  Among available remedial solutions, most often used are pumping
methods, which induce a gradient to remove product or water-product
mixtures; biodegradation is also used.

13.2.1.1  Fractures and Dissolution Zones

Distinction Between Primary and Secondary Porosity

    Total porosity is the volume of void space divided by the total unit
volume of a rock.  Total porosity may be divided into primary porosity and
secondary porosity.  Primary porosity is porosity due to void space in the
rock matrix.  Secondary porosity is void space due to fractures or
dissolution zones.
                                    326

-------
   UNSATURATED,
       ZONE
WATER
TABLE
     SATURATED
       ZONE
   NOTE: NOT ALL PHASE BOUNDARIES ARE SHOWN.
                                   LEGEND

                 WATER IN SATURATED ZONE

                 BDTA (MICROBIOTA INCLUDED)  I
COLLOIDAL PARTICLE
DISCONTINUOUS
LIOUiQ CONTAMINANT
                                         	 HHHit
                 WATER (MOBILE PORE WATER     jwr CONTAMINANT VAPORS
                 IN UNSATURATED ZONE)       **^
 Figure 13-1. Schematic Cross-Sectional Diagram of Locus No. 13
            Liquid Contaminants in Rock Fractures in Either the
            Unsaturated or Saturated Zone.
                                 327

-------
                                    (SOIL GAS)
                                    LOCUS NO
                               1 - CONTAMINANT VAPORS
                        VOLATILIZATION
          (BIOTA)
        LOCUS NO
    11 - SORBED TO BIOTA
                  UPTAKE
     (ROCK)
    LOCUS NO
10- DIFFUSED IN
   MINERAL GRAINS
   OR ROCKS
              CONDENSATION
    LOCUS  NO. 13
(LIQUID CONTAMINANTS IN ROCK
 FRACTURES IN UNSATURATED
   OR SATURATED ZONE)
                DIFFUSION
                                                        DISSOLUTION
PHASE
SEPARATION
       (WATER)
      LOCUS NO
3- DISSOLVED IN WATER FILM
8 - DISSOLVED IN GROUND-
  WATER
2 - DISSOLVED IN MOBILE
  PORE WATER
                                            ADVECTION
                                            DIFFUSION
                                            DISPERSION
                               (LIQUID CONTAMINANTS)
                                    LOCUS NO.
                           2- ADHERING TO "WATER-DRY-SOIL
                             PARTICLES
                           5 - IN PORE SPACES IN SATURATED
                             ZONE
                           6- IN PORE SPACES IN UNSATURATED
                             ZONE
       Figure 13-2. Schematic Representation of Important Transformation
                   and Transport Processes Affecting Other Loci.
                                         328

-------
    Locus no. 13 describes rock fractures and dissolution zones, focusing
on secondary porosity.  Locus no. 10, which discusses the rock matrix,
concerns primary porosity.  Crystalline rocks, such as granite, may have
primary porosities which range from 0-10 percent.  Typical secondary
porosities are 0.001-1 percent (Freeze and Cherry, 1979).  These very small
values for secondary porosities have important implications for transport
within fractures.  This is discussed in Section 13.2.2.2.

Sources and Types of Fractures

    Secondary porosity is formed after deposition and induration of the
rock (Duguid and Lee, 1977); primary porosity is formed during deposition
of a sedimentary rock or the solidification of an igneous rock.  Secondary
porosity generally results from dissolution and geochemical weathering, or
fracturing by tectonic action.  The range in size of fractures and
dissolution zones is very large; microcracks may be as small as a few
microns, whereas limestone caverns in karst topography may be tens of
meters deep and wide.  This variation in size contributes to the
anisotropic nature of transport in fractured rocks and karstic limestone.

Importance of Locus

    The importance of fracture and dissolution zones in the fields of
earthquake engineering, geothermal energy, and deep underground burial of
radioactive waste has been long acknowledged.  Fractures and dissolution
zones may be equally important with regard to groundwater contamination.
Where a rock contains both primary and secondary porosity, the fractures
and dissolution zones typically provide the greater permeability (sometimes
by several orders of magnitude) due to larger openings (Tang et al., 1981).
Results published by Garrett (1987) show that many contaminated well
incidents occur in fractured bedrock, after only a few gallons or tens of
gallons of gasoline have been spilled.  Garrett concludes that fracture
zones are more vulnerable than sand and gravel aquifers to groundwater
pollution and should be given more weight in hazardous waste site ranking
systems, such as DRASTIC (U.S. EPA, 1985).

Fracture Characteristics

    Although the following discussion focuses on the characterization of
fracture systems in rock, it may be extended to describe the dissolution
zones of karstic limestones.  Fracture systems in rocks may be described by
several properties, including fracture density and spacing, fracture width,
and fracture orientation and connectivity.  The  properties of the
fractures control, to a large extent, the flow and transport of water and
contaminants through the fractured rock.  Understanding mass transport
through rock fractures and modeling such systems must be based on a good
understanding of these properties.  The difficulty in gaining such an
understanding, however, has thus far impeded a complete understanding of
flow of liquid contaminants in locus no. 13.

Fracture Density

    Fracture density is the number of fractures per unit rock volume and
has a direct relationship to the permeability of a rock zone.  Highly
                                    329

-------
fractured rocks have high fracture densities and are more permeable than
less fractured rocks, all other factors being equal.  Fracture spacing is
typically determined by measuring the number of fractures intersecting a
rock core of a given length.  An example of a diagram showing fracture
intersections in a geologic cross-section is shown in Figure 13-3.

Fracture Width

    Fracture width refers to the distance between fracture walls measured
perpendicular to the walls.  The width varies along a fracture from zero
(where the fracture walls are in contact)  to some maximum value.  This
maximum value may be very large, as in the case of karst limestone
dissolution cavities or large joints; typical reported_values are usually
on the order of tens of microns to one millimeter (10~  to 10" m).
Fractures typically contain asperities (irregularities) that make the
effective width smaller than the maximum opening at some places.

Fracture Orientation and Connectivity

    The spatial orientation of a fracture or fracture plane is described by
its strike and dip (see Figure 13-4).  The strike of the fracture refers to
the surficial expression of the fracture upon a horizontal, flat land
surface.  The orientation of the strike of a fracture is measured as an
angle from north.  For example, a fracture with a strike of N30°E has a
surficial expression whose orientation is 30 degrees east of north.

    The dip of the fracture refers to the angle from horizontal of the
fracture plane and is measured perpendicular to the direction of strike.
Continuing from the example, since the strike of the fracture is
northeast-southwest.  The dip is either to the northwest or the southeast.
For example, if the dip is 60°SE, it dips 60° from horizontal to the
southeast.

    Characterizing fracture orientation is important because it affects the
flow direction and fracture connectivity; it may be delineated by surficial
mapping of fractures in bedrock outcrops or by interpretation of borehole
geophysical logs.

    Connectivity is extremely important in determining the hydraulic
conductivity of a fracture network, because fractures that do not connect
to other fractures or to a permeable rock matrix do not transmit fluid.
Connectivity is a property which must be described indirectly.  As such,
the information collected on connectivity tests (e.g., pump tests, natural
and artificial tracers) or by mapping is inferential and probably
incomplete.

Fracture Porosity

    The fracture porosity is the fraction of a unit volume that is occupied
by fracture openings.  A more important parameter is the effective
porosity, which is less than or equal to the total porosity, and which
includes only connected fractures.  Thus, void volume formed by
non-connected fractures is not included.
                                    330

-------
• S82W
   O
I    II   XI
Ground Surface
 •10
C
i
E
 •20
 -30
                            LEGEND

                            probable major flow path
                            probable minor flow path
                            area of low recovery and loss of orientation
                            expression of fracture(s) on ground surface
          NOTE:  Each fracture intersected by a borehole is shown at the angle of
                  Intersection.

                  Bearing and plunge of boreholes • 162? 745 SW
      Source: Rasmussen and Evans, 1987

      Figure 13-3.  Map of Fracture Spacings and Orientations from Borehole Records.
                                              331

-------
                                            Strike
                                            Direction
                                                               Dip
                                                              Direction
        a  Angle of Strike
        b  Angle of Dip
Fracture
 Plane
           Figure 13-4. Depiction of Strike, Dip Direction, and Dip.
    Freeze and Cherry (1979)  report effective porosities on the order of
0.001-1 percent.  Data from several other published reports (Garret,  1988;
Duguid and Lee, 1977;  Snow, 1968;  and Hendry et al., 1986) all fall within
this range, with most porosities on the order of 0.01-0.1 percent.
Torbarov  (1976) reports porosities in karst regions that range from one to
two percent.

Level of  Present Understanding

    Despite the potential importance of locus no. 13 in groundwater
contamination scenarios,  relatively little is known about the flow of
liquid contaminants  in fractures.   The fluid mechanics of multi-phase flow,
as opposed to single-phase flow, is especially in need of field experiments
to allow  researchers to develop and validate flow models applicable to
leaking UST sites.   Most  of the research that has been conducted  on flow of
contaminants in fractured rocks in recent years on this topic has been due
to the need for siting of high-level radioactive waste repositories and
                                     332

-------
understanding the transport of radionucleides through fractures.  This
research may be possibly applied to the problem of liquid hydrocarbon
transport.

    Flow models developed for sand or gravel aquifers generally assume
homogeneity and isotropy in a porous medium.  These assumptions do not hold
for fractured rock; the models are not applicable for fracture-flov
conditions unless the rock is so extensively fractured so as to represent
effectively a porous medium.

13.2.2  Mobilization/Remobilization

13.2.2.1  Partionining into Mobile Phase

    In this discussion, contaminants are assumed to be mobile in the
fractures (i.e., contaminants are already in the mobile phase in locus no.
13).  While some researchers (Wang and Narasimhan, 1985) have reported that
fractures can serve as barriers to flow under conditions of high capillary
tension (high negative pressure), fractures can increase mobility due to
characteristically high permeabilities.

    Liquid contaminant from locus no. 13 may also partition into water or
air, by dissolution or volatilization (see figure 13-2).  Only liquid
contaminant in the unsaturated zone may volatilize, whereas liquid
contaminant in either the saturated or unsaturated zones may dissolve into
water.  Volatilization is discussed in Section 1; dissolution is discussed
in Sections 8 and 12.

13.2.2.2  Transport of/vith Mobile Phase

    According to Wang and Narasimhan (1985), transport of contaminant
through the fractures of rocks with high primary porosity depends highly on
the pressure within the fractures.  At high capillary pressure
characteristic of unsaturated conditions, fractures are barriers to flow.
At saturation, however, the fractures carry most of the contaminant flow
through the rock.  Where fractures comprise all the porosity of a geologic
structure, all flow occurs in those fractures.  Transport of volatilized
liquid contaminant or of dissolved liquid contaminant (locus no. 1 and loci
no.'s 8 and 12, respectively) also occurs through fractures.  Transport is
treated separately for the saturated and unsaturated zones.

Saturated Flow

    A lack of understanding of the actual flow mechanics of fracture flow
impedes a better knowledge of transport phenomenon in locus no. 13.  Most
research on fracture flow relates to radionucleide-containing water; little
has been published regarding systems containing water and petroleum or
other contaminants, or multi-phase flow in general.

    A potential source of major error is the application of Darcy's Law to
multi-phase flow in fractures.  While conclusive evidence is not yet
available, this assumption is probably not valid (personal communication
from J.L.  Wilson, New Mexico Technology, 1988).  This uncertainty
                                    333

-------
notwithstanding, several researchers have developed theories regarding flow
through fractures.  The main results are summarized below.

    On a regional scale, researchers have assumed that Darcy's Law can be
used to describe flow through fractures:


                             V = Ksat, w .  dh                        (13.1)
                                    er     di

       _                                V
where: V = groundwater velocity (m/sec), sat,w = hydraulic conductivity
           (m/sec), dh/dl = hydraulic gradient (meter/meter) and 6  =
           transport porosity of rock (dimensionless).
                                                                     _3
   .Because the effective fracture porosity is typically very low (10   to
10~ ), groundwater velocity in fractures is much larger than in soils and
other porous media; velocities maybe on the order of meters to tens of
meters per day or more, making liquid contaminant very mobile in this
locus.  Section 13.4 gives a sample calculation for this equation.

    Many researchers (Snow, 1968; Wang and Narasimhan, 1985) have assumed
that individual fractures can be modeled as parallel-plate flow.  Under
this assumption, the flow is described by:


                        Q = g . w . £b3   dh                         (13.2)
                              12 . v    ' dl

where:
                                                     3
        Q = volumetric flow rate through fractures (m /sec)
        b = fracture width (meters)
                                             2
        g = gravitational constant (9.8 m/sec )
        w = lateral width of fracture (meters)
                              '   2
        v = kinematic viscosity (m /sec)

    dh/dl = hydraulic gradient (meter/meter)

    The summation includes all fractures, which are not of equal width.
The use of this equation makes several assumptions which may be erroneous:
all apertures remain constant in width throughout their length; fluid does
not influence the medium through which it flows; all the fluid channels
remain parallel to each other and to the fracture; and, flow is
non-turbulent, single-phase, non-compressible, and Newtonian.  Vang and
Narasimhan (1985) report that this cubic law overestimates flow in cases
where there is a high percentage of contact points, or points where stress
is transferred, within  the fractures.  Thus, the application of parallel
plate theory to liquid  contaminant transport seems to be limited.
                                    334

-------
    Some work has attempted to relate flow in fractures to viscous,
gravitational, and capillary forces in order to derive flow equations
(Wilson and Conrad, 1984; Hunt et al., 1986).  Thus far, a complete theory
based on this approach has not been published.

Unsaturated Flow

    Vang and Narasimhan (1985) discuss the interaction of the fracture
network and a porous rock matrix under conditions of capillary tension
(i.e., in the unsaturated zone).  The authors argue that, while the
fractures transport most of the flow under saturated conditions, the rock
matrix transports most of the flow when conditions of capillary tension
exist.  Under these conditions, Vang and Narasimhan (1985) postulate that
fractures act as barriers to flow.

    Rasmussen and Evans (1987) report that the unsaturated hydraulic
conductivity of a fractured rock is a function of saturation.  As capillary
tension increases, the degree of saturation decreases (see Figure 13-5).
At higher capillary tension, water or liquid contaminant is held in smaller
fractures which have a higher surface area to volume ratio.  As a result,
the smaller fractures have an increased resistance to flow and a lower
hydraulic or fluid conductivity.  Therefore, there is an inverse
relationship between capillary tension, and saturation and hydraulic
conductivity.

    The increase in capillary tension also decreases the effective flow
area because of a lowered fluid content in the fractures (see Figure 13-6).
The decrease in cross-sectional area of flow reduces the ability of the
fractured rock to transmit fluids.

    For rock with low or non-existent primary porosity in the unsaturated
zone, fractures provide the only means of transport for fluid flow.  Flow
under this condition is analogous to rain drops on a windshield; flow is a
function of interfacial surface tensions, capillary forces, wettability
patterns, and gravity (personal communication, J.L. Wilson, New Mexico
Technology, 1988).

13.2.3  Fixation

13.2.3.1  Partitioning onto Immobile (Stationary) Phase

    Liquid contaminants in locus no. 13, as a mobile phase, may become
stationary by either diffusing into the rock matrix (locus no. 10) or by
means of a high capillary tension (high negative pressure).  Liquid
contaminant held in locus no. 10, however, (as explained in Section 10.3)
is not entirely stationary.  Liquid contaminants my also adhere to the
walls of the fractures.  Unlike typical soils, however, the ratio of
surface area to volume is very small and relatively little of the liquid
contaminant will be trapped.  Also, inducing a suction in the fractures may
result in the removal of liquid contaminant from the fractures into the
matrix (as explained earlier) but will not immobilize liquid contaminant.
Thus, liquid contaminants in this locus generally do not partition to a
stationary phase.
                                    335

-------
                      0.6
                      0.4
                      0.2
                                                •lit tltt]  c l 1 rnir
                                                   \Motrix
                               .Vertical
t-
 \)  frocture

 \\
 '^Horizontal
  \\ frocture
                      ~\O~2   -IO'1    -10°    -10'    -10*     -10s
                                     Pressure (m)


   Source: Wang and Narasimhan, 1985.


   Figure 13-5. Saturation Versus Pressure Head for Fractures and Rock Matrix
                      O8
                    o06
                      0.2
                                \\  Vertical
                                \ ^fracture
                                 \\
                           frocture
                       -KT*
Source:  Wang and Narasimhan, 1985.
        -10°    -10'    -10*
         Pressurt(m)
 Figure 13-6. Effective Flow Areas in Fractures as a Function of Pressure.
                                         336

-------
13.2.3.2  Other Fixation Approaches

    The immobilization of liquid contaminants in rock fractures is not
generally attempted.  The very high mobility of liquids in rock fractures,
as discussed in section 13.2.2, makes it infeasible to trap and
indefinitely hold liquid contaminants in this locus.

    Several techniques to fix contaminants in a particular locus are known,
including stabilization, slurry vails, vitrification, and foam
stabilization.  Stabilization techniques generally use a siliceous material
to bind material into a solid mass.  For fractures, it is difficult to
transmit the cementing agent throughout the zone to be stabilized, and
therefore this is not usually done.  Slurry vails are often used to control
the flov of groundvater.  In a fractured rock zone, it is difficult to seal
off all possible routes of contaminant migration to ensure long-term
stabilization.  Vitrification uses electric fields to turn a soil or rock
mass into a glass-like material. '-This technique may prove technically
feasible, but at the present the technique is costly and leaves the site
unusable for months to years due to high temperatures.  Foam stabilization
is currently in the experimental stage (personal communication, S. Benson,
Lavrence Berkeley Laboratory 1988).  This technique involves injecting foam
into the soil pores and allowing solidification of the foam.  The
permeability is greatly decreased.  This technology is veil-advanced, but
may never be appropriate for aquifer situations due to the constituents of
the foam.

13.2.4  Transformation

13.2.4.1  Biodegradation (Details in Section 11)

    As with other bulk liquid loci, biodegradation in locus no. 13 is
probably insignificant.  Biodegradation is largely mediated by aerobic
bacteria which require oxygen as well as mineral nutrients.  For this
reason, liquid contaminants from locus no. 13 are most likely to be
biodegraded only after they have been dispersed and dissolved into one of
the aqueous phase loci.

13.2.4.2  Chemical Oxidation (Details in Section 3.2.4.2)

    Rocks may have secondary clay mineral coatings or contain primary
minerals which could catalyze chemical oxidation.

13.3  Storage Capacity in Locus

13.3.1  Introduction

    In general, the storage capacity in rock fractures is very small, and
is limited by the pore space constituted by the fractures.  For this locus
ve are assuming that the primary porosity is negligible; all the available
void volume is constituted by the fractures.  For calculation of the amount
stored in the fractures, the total porosity is appropriate; for transport
equations, the effective hydraulic porosity was used.  This quantity, which
equals the total porosity minus the porosity due to non-connected
fractures, describes the porosity available for liquid transport through a
                                    337

-------
rock zone.  Non-connected fractures ("dead zones"), however, function as
storage; the total porosity is thus used in this section.

13.3.2  Guidance on Inputs or, and Calculation of Maximun Value

    The porosity of the fractures determines the volume of liquid
contaminant stored in locus 13.  The fracture porosity forms an upper-bound
for the amount of liquid contaminant potentially held in this locus.  As
stated previously, the fracture porosity is typically less than one
percent.  In regions of karst limestone, where solutions channels and
caverns exist, porosities of one to three percent are possible (Duguid and
Lee, 1977) on a regional scale.  Locally, however, the fracture porosity
may be considerably higher.

    The maximum amount of liquid contaminant held in this locus is
determined by multiplying the fracture porosity by the density of the
liquid contaminant:

                        mr " 6rtpl                                   (13.3)

where
                                                                 3
    m  * mass of liquid contaminant per unit volume of rock (kg/m )

    9  « total porosity of rock fractures
     r                                   3
    p, = liquid contaminant density (kg/m )


    For the calculation of the maximum quantity of liquid contaminant held
in locus 13, it is assumed that there are no losses of liquid contaminant
from volatilization, dissolution, or microbial degradation, and that the
fracture porosity is occupied by only liquid contaminant.

13.3.3  Guidance on Inputs for, and Calculations of, Average Values

    While the fracture porosity places an upper limit on the volume of
liquid contaminant which may be stored in locus 13, the actual volume
stored in the fractures may be less than-this due to the presence of air
and water in the fractures.

Unsaturated Zone

    In this case both air and water phases are present in the fracture and
are presumed to be in direct contact with the liquid contaminant.  The air
allows liquid contaminant to volatilize; dissolution of liquid contaminant
into the water phase occurs.

    This case is characteristic of the condition where a slug of liquid
contaminant passes a fracture zone, leaving liquid contaminant held in
residium.  No guidance was found for typical values of residual saturation
in fractures, but it would depend on fracture width, interfacial tension of
the liquid contaminant, and losses due to microbial degradation,
dissolution, and volatilization, etc.
                                    338

-------
    The average mass of liquid contaminant in the fractured rock is:

  mr = mres ~ mw ' mv ' mb ' ml
where
                                                         3
    m    « contaminant mass per unit volume of rock (kg/m )
         » residual mass of liquid contaminant per unit volume of rock
     res   /»	y_j\
m
       (kg/mj)
                                                              3
m    » contaminant mass in vater per unit volume of rock (kg/m )
                                                                 3
m    = contaminant mass volatilized per unit volume of rock (kg/m )
m,    « contaminant mass lost to microbial degradation per unit volume
 D     of rock 
-------
    m  = (0.0005).(713 kg/m3)
                                            3
       = 0.36 kg of liquid contaminant per m  of fractured rock

13.4.1.2  Average Quantity at Unsaturated Conditions

    In this example, the values for fracture porosity and liquid density
are retained for comparison.  Additionally, it is assumed that the losses
of liquid contaminant to microbial degradation and chemical reaction are
negligible.  The residual, water-filled, and air-filled porosities are
0.0003, 0.0001, and 0.0001, respectively.  Values of solubility and vapor
density are assumed to be 130 mg/1 and 1600 g/m , respectively.  The mass
of liquid contaminant per unit volume of fractured rock is:

 m  - (713 kg/m3).(0.0003) - (130 mg/1).(10~4).(10~3) - (10~4).(1.6 kg/m3)
  r                                      3
 m  - 0.21 kg of liquid contaminant per m  of fractured rock

13.4.1.3  Average Quantity at Saturated Conditions

    A similar calculation may be made for saturated conditions, under the
same assumptions.  Here, the fraction of fracture porosity occupied by air is
zero.  The mass of liquid contaminant per unit volume of fractured rock iai

    in. . (0.0003).(713 kg/m3) - (130 mg/1).(10~4).(10~3)
     r                                      3
    m  = 0.21 kg of liquid contaminant per m  of fractured rock

13.4.2  Transport Rate Calculations

    As stated earlier, the flow velocity of fluids in fractures can be very
large relative to that in sand and gravel aquifers.  In the following
example, it is assumed that:

    K  .. , = fluid conductivity            = 10   m/sec
     sat,l
    dh/dl  = gradient                      a 0.1 m/m
    0      = fracture porosity             « 0.0005

    V,     = velocity of immiscible fluid, m/sec

    Then, using Equation 13.1,


    V,  = Ksat, 1 . dh
            6.      dl

      = (10~7 m/sec)'(0.1)/0.0005

      - 2x10   m/sec

      =1.73 m/day

      This is a high velocity compared to typical groundwater  flow, using
conservative values; in the case where the fracture porosity is only
0.0001, which is the low end of the range quoted in Freeze and Cherry
                                    340

-------
(1979), the velocity would be 8.64 m/day.  The high seepage velocities in
fracture zones are cited by Garrett (1988) as the main reason that
aquifers in fractured rock zones are so sensitive to contamination, even
after only a few gallons or tens of gallons leaked.  These very high flow
rates may also be responsible for the non-Darcian flow patterns (personal
commuciation, J.L. Wilson, New Mexico Technology, 1988).

Parallel-Plate Approximation

      Equation 13.2 may not be entirely appropriate, but it may estimate
the quantity of flow occurring through a single fracture.  Assume that
the following values characterize a fracture:
    dh/dl = hydraulic gradient           = 1 cm/cm

    b     = aperture opening             - 0.01 cm
     8i
    w     = lateral width                =0.1 cm
                                                    -3   2
    v,    = kinematic viscosity          = 6.36 x 10   cm /sec
             of liquid contaminant
                                                       2
    g     = gravitational constant       = 980.7 cm/sec
                  Q = (980.7 cm/sec2).(0.01cm)3 .(O.lcm)

                          (12).(6.36 x 10~3 cm2/sec)


        Q = 1.28 x 10~9 m3/sec

          = 0.11 liters/day
          = 10.7 gallon/year

    This represents the flow of the synthetic gasoline blend through a
single idealized fracture with aperture of 100 microns, a typical value for
fractured igneous rock (Snow, 1969) with a unit gradient.  Large joints and
conduits will transmit a far greater volume of fluid.

13.5  Summary of Relative Importance of Locus

13.5.1  Remediation

    Contamination in fractures is highly mobile and therefore relatively
difficult to remediate.  Where the primary porosity is negligible, the
fractures provide the principal pathway for flow.   The relatively large
size of fracture openings permit fluid to move quickly away from a leak
site, contaminating large areas in a short time.

    The first step to remove liquid contaminant from this locus is usually
to induce a local hydraulic gradient.  In this manner, free product may be
removed as well as pure liquid product blobs within the groundwater.
Because of the low porosity and high permeability of the fracture network,
relatively little free product is typically recovered.  Groundwater pumping
                                    341

-------
with above-ground treatment by air stripping or activated carbon is another
method that is often used for the saturated zone.

    Immobilization measures are generally not undertaken for this locus.
The high mobility of contaminants in the fractures and the lack of knowledge
regarding fracture connectivity lover the feasibility of these actions.

13.5.2  Loci Interactions

    Figure 13-2 shows the major interactions among locus no. 13 and the
other loci.  Liquid contaminant is highly mobile and can interact with
almost all of the other loci.  Bulk transport to fractures from soil, or
vice versa, is important; contaminant may diffuse into loci no. 2, 5, or 6,
or gather as a free product lens (locus no. 7); contaminant may adsorb onto
the rock face (locus no. 4); or contaminant may be adsorbed to micells
traveling through the fractures (locus no. 9).  Also, biodegradation and
inorganic transformations may occur in this locus.  Volatilization of
contaminant (locus no. 1) may occur; or contaminants may dissolve into
water (locus no. 3, 8, or 12).  Important interactions among locus no. 13
and others are discussed in the appropriate sections of this report.

13.5.3  Information Gaps

    In general, the behavior and mechanics of multi-phase flow through
fractures is poorly understood.  This topic has yet to attract widespread
attention among researchers.  In particular need of research are the
following areas:

    (1)  field and laboratory experiments with actual fracture networks to
         develop a comprehensive data base for fracture porosities,
         permeabilities,, aperture sizes, and other basic data;

    (2)  the formulation of appropriate flow models to use to represent
         flow through fractures;

    (3)  the influence of wettability patterns, interfacial tensions,
         capillary pressure, gravitational and viscous forces on flow
         through fracture networks;

    (4)  the interactions between fractures and the rock matrix in both the
         saturated and unsaturated zones; and,

    (5)  data and a better understanding of fractures, including methods to
         determine connectivity of fractures.

13.6  Literature Cited

Duguid, J.O. and P.C.Y. Lee.  1977.  Flow in Fractured Porous Media.  Water
    Resource Research, 13(3):558-566.

Freeze, R.A. and J.A. Cherry.  1979.  Groundwater.  Prentice-Hall,
    Inglewood Cliffs, N.J.
                                    342

-------
Garrett, Peter.  1987.  Comparative Vulnerability to Oil Spills of Bedrock
    versus Sand and Gravel Aquifers.  In;  Proceedings of Focus on Eastern
    Regional Ground Water Issues.  NWA.

Hendry, M.J., J.A. Cherry, and E.I. Wallick.  1986.  Origin and
    Distribution of Sulfate in a Fractured Till in Southern Alberta,
    Canada.  Water Resources Research, 22(1): 45-61.

Hunt, J.R., N. Sitar, and K.S. Udell.  1986.  Organic Solvents and
    Petroleum Products in the Subsurface:  Transport and Cleanup.  Univ.
    Calif. (Berkeley), UCB-SEERHL Report No. 86-11.  Sanitary Engineering
    and Environmental Health Research Laboratory.

Rasmussen, T.C. and D.D. Evans.  1987.  Unsaturated Flow and Transport
    Through Fractured Rock Related to High-Level Waste Repositories, Final
    Report.  Prepared for Nuclear Regulatory Commission, Washington, D.C.

Snow, D.T.  1968.  Rock Fracture Spacings, Openings, and Porosities.  Journal
    of the Soil Mechanics and Foundations Division, Proc. ASCE, 94:73-91.

Snow, D.T.  1969. Anisotropic Permeability of Fractured Media.  Water
    Resources Research, 5(6):1273-1289.

Tang, D.H., E.O. Frind, and E.A. Sudicky.  1981.  Contaminant Transport in
    Fractured Porous Media:  Analytical Solution for a Single Fracture.
    Water Resources Research, 17(3):  555-564.

Torbarov, K.  1976.  Estimation of Permeability and Effective Porosity in
    Karst on the Basis of Recession Curve Analysis.  Proc. of the
    U.S.-Yugoslavian Symposium.  Karst Hydrology and Water Resources,
    Volume I.  Water Resources Publications, Fort Collins, CO.

U.S. EPA. 1985.  DRASTIC:  A Standardized System for Evaluating Ground Water
    Pollution Potential Using Hydrogeologic Settings, EPA/600/2-85/018.

Wang, J.S.Y. and T.N. Narasimhan.  1985.  Hydrologic Mechanisms Governing
    Fluid Flow in a Partially Saturated, Fractured, Porous Medium.  Water
    Resources Research, 21(12):1861-1874.

Wilson, J.L. and S.H. Conrad.  1984.  Is Physical Displacement of Residual
    Hydrocarbons a Realistic Possibility in Aquifer Restoration?  In;
    Proceedings of Petroleum Hydrocarbons and Organic Chemicals in
    Groundwater.  NWWA/API.

13.7  Additional Reading

Flov Hydraulics in Fractured and Karstic Media

Long, J.C.S., J.S. Remer, C.R. Wilson, and P.A. Witherspoon.  1982.  Porous
    Media Equivalents for Networks of Discontinuous Fractures. Water
    Resources Research, 18(3):645-658.

Moore, G.K.  1973.  Hydraulics of Sheetlike Solution Cavities.  Ground
    Water, 11(4):4-11.
                                    343

-------
Tsang, Y.W. and C.F. Tsang.  1987.  Channel Model of Flow Through Fractured
    Media.  Water Resources Research, 23(3):467-479.

Characterization of Fractured and Karstic Media

Silliman, E. and R. Robinson.  1989.  Identifying Fracture Interconnections
    between Boreholes Using Natural Temperature Profiling:  The Conceptual
    Basis.  Ground Water, 27(3):393-402.

Smart, C.C.  1988.  Artificial Tracer Techniques for the Determination of
    the Structure of Conduit Aquifers.  Ground Water, 26(4):445-453.

Taylor, R.W. and A.H. Fleming.  1988.  Characterizing Jointed Systems by
    Azimuthal Resistivity Surveys.  Ground Water, 26(4):464-474.

Contaminant Transport in Fractured and Karstic Media

Andersson, J. and R. Thunvik.  1986.  Predicting Mass Transport in Discrete
    Fractures, Water Resources Research, 22(13):1941-1950.

Grisak, G.E. and J.F. Pickens.  1980.  Solute Transport Through Fractured
    Media, 1.  The Effect of Matrix Diffusion.  Water Resources Research,
    16(4): 719-730.

Rasmuson, A.  1985.  Analysis of Hydrodynamic Dispersion in Discrete
    Fracture Networks using the Method of Moments.  Water Resources
    Research, 21(11):1677-1683.

Ross, B. 1986.  Dispersion in Fractal Fracture Networks.  Water Resources
    Research, 22(5): 823-827.

Simmleit, N. and R. Herrmann.  1987.  The Behavior of Hydrophobic, Organic
    Micropollutants in Different Karst Water Systems.  Water, Air, and Soil
    Pollution, 34:97-109.

Sudicky, E.A. and E.O. Frind.  1982.  Contaminant Transport Porous Media:
    Analytical Solutions for a System of Parallel Fractures.  Water
    Resources Research, 18(6):1634-1642.

Models of Flow and Transport in Fractured and Karstic Media

Andersson, J., A.M. Shapiro, and J. Bear.  1984.  A Stochastic Model of a
    Fractured Rock Conditioned by Measured Information.  Water Resources
    Research, 201(1):79-88.

Endo, H.K., J.C.S. Long, C.R. Wilson, and P.A. Witherspoon.  1984.  A Model
    for Investigation Mechanical Transport in Fractured Networks.   Water
    Resources Research, 20(10):1390-1400.

Oda, M.  1986.  An Equivalent Continuum Model for Coupled Stress and Fluid
    Flow Analysis in Jointed Rock Masses.  Water Resources Research,
    22(13):1845-1856.
                                    344

-------
                                SECTION 14

                                 GLOSSARY
Anaerobe.  An organism that does not require air or free oxygen to maintain
    its life process.

Abiotic.  Referring to the absence of living organisms.

Advection.  The process of transfer of fluids (vapors or liquids) through a
    geologic formation in response to a pressure gradient which may be
    caused by changes in barometric pressure, water table levels, wind
    fluctuations, or rainfall percolation.  Advection can result from a
    thermal gradient caused by a heat source.  See Darcy's Law.

Aliphatic.  Of or pertaining to a broad category of carbon compounds
    distinguished only a straight, or branched, open chain arrangement of
    the constituent carbon atoms.  The carbon-carbon bonds may be saturated
    or unsaturated.

Anisotropy.  The dependence of property upon direction of measurement
    (e.g., hydraulic conductivity, porosity, compressibility, dispersion,
    etc.).

Aromatic.  Of or pertaining to organic compounds that resemble benzene in
    chemical behavior.

Bentonite.  A colloidal clay, largely made up of the mineral sodium
    montmorillonite, a hydrated aluminum silicate.

Brovnian Movement.  Random movement of molecules or colloids suspended in a
    fluid.

Biodegradation.  A process by which microbial organisms transform or alter
    through enzymatic action the structure of chemicals introduced into the
    environment.

Biotic.  Of or pertaining to life and living organisms.  Induced by the
    actions of living organisms.

Biomass.  The amount of living matter in a given area or volume.

Biota.  A term that encompasses the spectrum of living things within a
    given area.

Caliche.  A layer of calcium and magnesium carbonate deposited in a
    near-surface soil horizon by evaporating soil water or groundwater.  It
    may occur as a soft thin soil horizon, as a hard thick bed just below
    the land surface, or as a surface layer exposed by erosion.

Capillarity.  The pressure difference across the interface of two
    immiscible liquids or between a liquid and a solid.  In a porous
                                    345

-------
    medium, capillarity is zero in the saturated zone  and less than zero in
    the vadose zone.   See surface tension,  saturated zone,  vadose zone.

Capillary Fringe.   The zone of a porous medium above the  water table within
    which the porous  medium is saturated but  is at  less  than atmospheric
    pressure.  The capillary fringe is considered to be  part of the vadose
    zone, but not  of  the unsaturated zone.

Catalyst.  A substance that alters the rate of a chemical reaction and may
    be recovered,  essentially unaltered, in form and amount at the end of
    the reaction.

Chemisorption.  A process in which weak chemical bonds are formed between
    gas or liquid  molecules and a solid surface.

Clay.  A mineralogical term describing a family of  aluminosilicate minerals
    formed by the  decomposition of primary minerals (e.g., micas and
    feldspars) and re-composed into clay minerals.   A  textural term
    referring to particles  2 $m in diameter.

Coalescence.  The  bonding of welded materials into one body, or the uniting
    by growth in one body as particals, gas or liquid.

Colloid.  Particles having dimensions of 10-10,000 angstroms (1-1000
    nanometers) and which are dispersed in a different phase, such as a
    fluid or liquid.

Complex.  A combination of two or more atoms into a molecular species,
    usually charged,  and existing in water or some other fluid.

Compressibility.  The change in volume of a porous medium in response to an
    applied stress which is counterbalanced by the incompressibility of the
    saturating fluid and the granular skeleton of the  porous medium which
    it saturates.   Compressibility has units of inverse force (Pa-1).

Constituents.  An essential part or component of a system or group:
    examples are an ingredient of a chemical system,  or a component of an
    alloy.

Containment.  The prevention of  the spreading of oil or other hazardous
    materials by the placing of  booms or physical barriers and the use of
    the  absorbents, gelling, or  herding agents or to other materials to
    restrain, entrap, and collect a spill.

Corrective Action.  The  removal  of chemicals and/or contaminated soils,
    objects  and groundwater, from a site, or other clean-up activities
    designed  to restore  the local environment to acceptable conditions.
    Corrective actions may include, for example, vacuum extraction of the
    vadose zone, soil washing, and the  extraction and treatment of
    contaminated groundwater.

Cytoplasm.   The protoplasm of  an animal or plant cell external to  the
    nucleus.
                                         346

-------
Darcy's Lav.  An empirical relationship between hydraulic gradient and the
    viscous flow of water in  the saturated zone of a porous medium under
    conditions of laminar flow.  The flux of vapors through the voids of
    the vadose zone can be related to pressure gradient through the air
    permeability by Darcy's Law.  See hydraulic conductivity, air
    permeability, hydraulic gradient, pressure gradient, laminar flow,
    vadose zone, saturated zone.

Dehydrogenation.  Removal of  hydrogen from a compound.

Density.  The amount of mass  of a substance per unit volume of the
    substance (g/cm ).

Desorption.  The process of removing a sorbed substance by the reverse of
    adsorption or absorption.

Diffusion.  The process whereby the molecules of a compound in a single
    phase equilibrate  to a zero concentration gradient by random molecular
    motion.  The flux  of molecules is from regions of high concentration to
    low concentration  and is  governed by Pick's Second Law.  See Pick's
    Second Law, effective diffusion coefficient.

Dispersion.  The process by which a substance or chemical spreads and
    dilutes in flowing ground water or soil gas.  On a microscale,
    dispersion is due  to mixing within individual pores, mixing between
    pore channels, and mixing due to molecular diffusion.  At larger
    scales, geologic heterogeneity and anisotropy cause.dispersion.
    Dispersion has units of squared length per time (cm /sec).

Dissociation.  Separation of  a molecule into two or more fragments (atoms,
    ions, radicals) by collision with a second body or by the absorption of
    electromagnetic tadiation.

Dissolution.  Dissolving of a material in a liquid solvent (e.q., water).

Dynamic Viscosity.  The measure of internal friction of a fluid that
    resists shear within the  fluid; the constant of proportionality between
    a shear stress applied to liquid and the rate of angular deformation
    within the liquid, having units of mass per length per time (gm/cm
    sec).

Effective Diffusion Coefficient.  The constant of proportionality in Pick's
    Second Law which is dependent on tortuosity, porosity, and moisture
    content and properties of the diffusing compound, having units of
    squared length per time (cm /sec).   See tortuosity, porosity, moisture
    content.

Eoulisification.   The  process of dispersing one liquid in a second
    immiscible liquid.

Entrainment.  A process in which solid particles or liquid droplets are, by
    force of friction with a passing fluid (e.g., air or water),  lifted
    from a resting place and carried along with the flowing fluid.
                                    347

-------
Pick's Second Lav.  An equation relating the change of concentration with
    time due to diffusion to the change in concentration gradient with
    distance from the source of concentration.  See diffusion, effective
    diffusion coefficient.

Field Capacity.  The percentage of water remaining in the soil 2 or 3 days
    after gravity drainage has ceased from saturated conditions.

Fluid Conductivity.  The constant of proportionality in Darcy's Lav
    relating the rate of flov of a fluid through a cross-section of porous
    medium in response to a hydraulic gradient.  Fluid conductivity is a
    function of the intrinsic permeability of a porous medium and the
    kinematic viscosity of the fluid vhich flovs through it.  Fluid
    conductivity has units of length per time (cm/sec).

Flux.  The rate of movement of mass through a unit cross-sectional area per
    unit time in response to a concentration gradient or some advective
    force, having units of mass per area per time (g/cm *sec).

Freundlich Isotherm.  An expression relating the equilibrium betveen the
    sorbed chemical concentration and its aqueous phase concentration
    through a power lav.

Fugacity.  A function used as an analog of the partial pressure in applying
    thermodynamics to real systems; at a constant temperature, it is
    proportional to the exponential of the ratio of the chemical potential
    of constituent of a system divided by the product of the gas constant
    and the temperature, and it approaches the partial pressure as the
    total pressure of the gas approaches zero.

Fulvic Acid.  A term of varied usage but usually referring to the mixture
    of organic substances remaining in solution upon acidification of a
    dilute alkali extract from the soil.  Thus, fulvic acids are soluble
    under all pH conditions.

Funicular Zone.  A narrov band above the groundvater table bounded above by
    the capillary rise in the slimmest continuous pores, and bounded belov
    by the capillary rise in the widest, pores.  (See Figure 7-4).

Half-life.  The time required for half of a substance to decay or alter by
    a process, (e.g., radioactive decay, biodegradation, volatilization,
    photolysis, hydrolysis, oxidation, etc.).

Henry's Lav.  The relationship betveen the partial pressure of a compound
    and the equilibrium concentration in the liquid through a constant of
    proportionality known as Henry's Lav Constant.  See partial pressure.

Heterogeneity.  The dependence of property upon location of measurement
    (e.g., hydraulic conductivity, porosity, compressibility, dispersion,
    etc.).  Heterogeneity may be due to grain size trends, stratigraphic
    contacts, faults, and vertical bedding.
                                    348

-------
Homogeneity.  The independence of property with location of measurement
    (e.g., hydraulic conductivity, porosity, compressibility, dispersion,
    etc.).

Humin.  That fraction of humic substances that is not soluble in water at
    any pH value.

Humic acid.  That fraction of humic substances that is not soluble in water
    under acid conditions (below pH 2), but becomes soluble at greater pH.

Humic substances.  A general category of naturally occurring, biogenic
    heterogeneous organic substances that can generally be characterized as
    being yellow to black in color, of high molecular weight, and
    refractory.  Humic substances include humin, humic acids and fulvic
    acids.

Hydraulic Conductivity.  The constant of proportionality in Darcy's Law
    relating the rate of flow of water through a cross-section of porous
    medium in response to a hydraulic gradient.  Also known as the
    coefficient of permeability, hydraulic conductivity is a function of
    the intrinsic permeability of a porous medium and the kinematic
    viscosity of the water which flows through it.  Hydraulic conductivity
    has units of length per time (cm/sec).

Hydraulic Head.  A measure of mechanical energy per unit weight density of
    water as the sum of elevation head and pressure head, having units of
    length (cm).

Hydraulic Gradient.  The change in piezometric head between two points
    divided by the horizontal distance between the two points, having
    dimensions of length per length (cm/cm).  See piezometric head.

Hydrocarbon.  One of a very large group of chemical compounds composed only
    of carbon and hydrogen; the largest source of hydrocarbons is from
    petroleum crude oil.

Hydrolysis.  Decomposition or alteration of a chemical substance by
    reaction with water.

Hydrophobic.  Lacking an affinity for, repelling, or failure to absorb or
    dissolve in, water.

Hysteresis.  The dependence of the'state of a system on direction of the
    process leading to it; a non-unique response of a system to stress,
    responding differently when the stress is released.  Compressibility,
    moisture content, soil adsorption and unsaturated hydraulic
    conductivity exhibit hysteretic behavior.

Illite.  A group of 2:1, non-expanding, hydrous potassium-magnesium
    aluminosilicate clay minerals of intermediate properties between
    kaolinite and montmorillonite.

Immersion.  Placement into or within a fluid, usually water.
                                    349

-------
Infiltration.  The downward movement of water through a soil from rainfall
    or from the application of artificial recharge in response to gravity
    and capillarity.

Intrinsic Permeability.  A measure of the ease with which a porous medium
    transmits a fluid which is independent of»the properties of the fluid
    and which has units of squared length (cm ).

Isotropy.  The independence of a property with direction of measurement
    (e.g., hydraulic conductivity, porosity, compressibility, dispersion,
    etc.).

Kaolin!te.  A 1:1, non-expanding, hydrous aluminosilicate clay mineral
    formed by the alteration of micas and feldspars.

Kinematic Viscosity.  The ratio of the dynamic viscosity?of a fluid to its
    density, having units of squared length per time (cm /sec).

Locus (pi., Loci).  In this report, the term locus is used to refer to one
    of 13 generic contaminated environments in the subsoil region.  Each
    locus is defined by considering:  (1) its position relative to the
    groundwater table (above, on, below); (2) the phase of the contaminant
    (liquid, vapor, aqueous solution, sorbed to soil); and (3) the nature
    of the local natural matter (unconsolidated sediments, fractured rock,
    flowing water).

Meniscus.  The curved surface of a liquid between solid boundaries (e.g.,
    capillary tube, mineral grains) which is caused by the surface tension
    of the liquid.  The geometry of the meniscus is affected by the surface
    tension of the liquid, the difference in density between the liquid and
    overlying air, the distance between the solid boundaries, and the
    hydrophobic nature of the liquid.

Mica.  A family of platy, igneous and metamorphic, aluminosilicate minerals
    which weather and form clays.  Micas separate readily into thin sheets
    or flakes.

Microbe.  A microorganism, especially a bacterium.

Microorganisms.  Microscopic organisms including bacteria, protozoans,
    yeast, fungi, viruses and algae.

Macropore.  A large pore in a porous medium which may be formed by physical
    phenonema or biological activity, and through which water, or other
    fluids, flows solely under the influence of gravity, unaffected by
    capillarity.

Mobilization.  The process or processes by which a liquid contaminant in a
    locus is made more mobile in the locus or more transferable between
    loci.

Montmorillonite.  A 2:1, expanding, hydrous magnesium aluminosilicate clay
    mineral exhibiting pronounced swelling-shrinkage behavior and high
    plasticity and cohesion.
                                    350

-------
Moisture Content.  The amount of water lost from the soil upon drying to a
    constant weight, expressed as the weight per unit weight of dry soil or
    as the volume of water per unit bulk volume of the soil.  For a fully
    saturated medium, moisture content equals the porosity; in the vadose
    zone, moisture content ranges between zero and the porosity value for
    the medium.  See porosity, vadose zone, saturated zone.

Mole Fraction.  The ratio of the number of moles of a substance in a
    mixture or solution to the total number of moles of all the components
    in the mixture or solution.

Monovalent.  A radical or atom whose valency is 1.

Oxidation.  A chemical reaction that increases the oxygen content of a
    compound, or raises the oxidation state of an element.

Oxidation Potential.  The difference in potential between an atom or ion
    and the state in which an electron has been removed to an infinite
    distance from this atom or ion.

Partial Pressure.  The portion of total vapor pressure in a system due to
    one or more constituents in the vapor mixture.

Percolation, Soil Water.  The downward movement of water through soil.
    Especially, the downward flow of water in saturated or nearly saturated
    soil at hydraulic gradients of the order of 1.0 or less.

Phreatic Level.  The groundwater level that would be seen in an observation
    well projecting down into an aquifer. (See Figure 7-4.)

Polymerization.  The bonding of two or more monomers to produce a polymer.

Porosity.  The volume fraction of a rock or unconsolidated sediment not
    occupied by solid material but usually occupied by water and/or air.
    Porosity is a dimensionless quantity.

Redox.  A chemical reaction in which an atom or molecule loses/gains
    electrons to/from another atom or molecule.  Also called
    oxidation-reduction.  Oxidation is the loss of electrons; reduction is
    the gain in electrons.

Reduction.  A chemical reaction in which an atom of molecule gains
    electrons.  Sometimes results by reaction of the substance with
    hydrogen.

Refractory.  A nonspecific characteristic of some chemicals implying
    resistance to biodegradation or other degradation or treatment
    processes.

Residence Time.  The average time that water remains in a porous medium or
    a particle remains in a reservoir.  Residence time is calculated as the
    ratio of reservoir volume to total water inflow rate having units of
    time (sec).
                                          351

-------
Residual Saturation.  The amount of water or oil remaining in the voids of
    a porous medium and held in an immobile state by capillarity and
    dead-end pores.

Salinity.  The quantity of anions and cations in water, usually between 33
    and 37 parts per thousand in sea water.

Sinter.  A chemical sedimentary rock deposited by precipitation from
    mineral waters, especially siliceous sinter and calcareous sinter.

Slurry.  A thick mixture of liquid, especially water, and any of several
    finely divided substances, such as cement or clay particles.

Solubility.  The maximum amount of mass of a compound that will dissolve
    into a unit volume of solvent, usually water, having units of mass per
    volume (gm/cm ).

Sorption.  A general term used to encompass the process of absorption,
    adsorption, ion exchange, and chemisorption.

Specific Gravity.  The ratio of the weight of a given volume of the
    material at 4°C (or some stated temperature) to the weight of an equal
    volume of distilled water.  Materials with specific gravity of less
    than 1 will float on water; materials with specific gravity over 1 will
    sink in water.

Surface Tension.  A measure of the interfacial tension due to molecular
    attraction between two fluids in contact or between a liquid in contact
    with a solid, having units of mass per squared time (dyne/cm).

Surfactant.  Also Surface Active Agent.  A chemical material which provides
    a "linking action" between two materials, such as oil and water, which
    normally resist mixing or readily joining in solution.  Surfactants
    provide the emulsification forces which allow oil and water to mix and
    remain in either oil-in-water or water-in-oil solutions.

Tortuosity.  The ratio of path length through a porous medium to the
    straight-line flow path which describes the geometry of the porous
    medium.  Tortuosity is a dimensionless parameter which ranges in value
    from 1 to 2.

Transpiration.  The release of water withdrawn from the soil by plants
    during photosynthesis and other life processes.

Unsaturated Zone.  The portion of a porous medium, usually above the water
    table in an unconfined aquifer, within which the moisture content is
    less than saturation and the capillary pressure is less than
    atmospheric pressure.  The unsaturated zone does not include the
    capillary fringe.

Vadose Zone.  The portion of a porous medium above the water table within
    which the capillary pressure is less than atmospheric and the moisture
    content is usually less than saturation.  The vadose zone includes the
    capillary fringe.
                                     352

-------
Vapor Pressure.  The equilibrium pressure exerted on  the  atmosphere by a
    liquid or solid at a given  temperature.  Also a measure  of  a
    substance's propensity  to evaporate or give off flammable vapors.   The
    higher the vapor pressure,  the more volatile the  substance.

Vermiculite.  A 2:1, hydrous ferro-magnesium aluminosilicate clay mineral
    similar in structure and property  to montmorillonite.

Vitrification.  Formation of a  glassy  or noncrystalline material.

Volatilization.  The process of transfer of a chemical from  the water  or
    liquid phase to the air phase.  Solubility, molecular weight,  and  vapor
    pressure of the liquid and  the nature of the air-liquid/water interface
    affect the rate of volatilization.  See solubility, vapor pressure.

Water Table.  The water surface in an  unconfined aquifer  at  which the  fluid
    pressure in the voids is at atmospheric pressure.

Wilting Point.  The point at which a plant wilts, no  longer  able to
    withdraw water from a soil  or sediment to support transpiration
    processes and retain turgor pressure.
 GOVERNMENT HUNTING OmCE: 199:-s.,,.! a* ,0622   353

-------