United States        R.S. Kerr            Office of Solid Waste    Publication: 9355.4-16FS
                       Environmental       Environmental        and Emergency        EPA/540/F-94/049
                       Protection Agency    Research Laboratory   Response            PB94-963317
                                                                                September 1994
SEPA             DNAPL SITE CHARACTERIZATION
                                   Response
 Office of Emergency and Remedial Response
 Hazardous Site Control Division (5203G)                                                     Quick Reference Fact Sheet


 INTRODUCTION

 Dense nonaqueous phase liquids (DNAPLs), such as some chlorinated solvents, coal tar wastes, creosote based wood-
 treating oils, and some pesticides, are immiscible fluids with a density greater than water.  As a result of widespread
 production, transportation,  use, and disposal  of hazardous DNAPLs,  particularly since 1940,  there are numerous
 DNAPL contamination sites in North America. The potential for significant long-term groundwater contamination by
 DNAPL chemicals at many sites is high due to their toxicity, limited solubility (but much higher than drinking water
 limits),  and  significant migration  potential in soil gas, groundwater, and/or as a separate phase liquid.  DNAPL
 chemicals, particularly chlorinated  solvents, are among the most prevalent groundwater contaminants identified in
 groundwater supplies and at waste disposal sites. For these and other reasons, it is important to investigate for DNAPLs
 (Table  1).
                      Table  1.  Why Investigate DNAPL?
      To facilitate early removal action (e.g., if the DNAPL zone is still expanding)
      To adequately account for it in performing a risk assessment
      To evaluate possible cosolvency effects (e.g., NAPL mobilizing immobile chemicals)
      Because DNAPL transport principles and properties differ from those associated with solutes
      To adequately characterize the source area
      To characterize mass loadings and mass-in-place (e.g., most mass will be in the DNAPL)
      To adequately account for the dissolution or DNAPL (e.g., can persist for decades; effluent changes over
      time leaving behind a residual that may be more difficult to remediate)
    •  To avoid spreading contamination via characterization and/or remediation
 INVESTIGATION STRATEGY
L

 Site characterization, a process following the scientific method, is performed in phases (Figure 1). During the initial
 phase, a hypothesis or conceptual model of chemical presence, transport and fate is formulated using available site
 information and an understanding of the processes that control chemical distribution. Based on the initial hypothesis,
 a data collection program is designed in the second phase to test and improve the site conceptual model and thereby
 facilitate risk and remedy  assessment.  After analyzing the newly acquired data within the context of  the initial
 conceptual model, an iterative step of refining the hypothesis is performed using the results of the analysis, and
 additional data may be collected. As knowledge of the site increases and becomes more complex, the working hypothesis
 may take the form of either a numerical or analytical model. Data collection continues until the hypothesis is proven
 sufficiently. During implementation of a remedy, the subsurface system is stressed. This provides an opportunity to
 monitor and not only learn about the effectiveness of the remediation, but to learn more about the subsurface. Therefore,
 remediation (especially pilot studies) should be considered part of site characterization, yielding data that may allow
 improvements to be made in the conduct of remediation.

-------
             NOTE:

Characlenzahon should t>e conducted in a
phased,  evolutionary manner starting  with
review of avatla6l« data.   Early field work
should focus on areas fceyond1 the DHAPL
zone and use nontnvastve methods in Uu
DHAPL zone   Each characterization
activity should be designed to test the
conceptual model in a manner Out will
increase the capacity to perform risk
and remedy  analysis

To limit  tht  potential for promoting
contaminant  migration, avoid:  (I) conducting
unecessary field wort' (2) drilling through
capillary barriers beneath DNAPL- (3)  pumping
from beneath DNAPL zones:  and. (4) using
invasive characterization  or remediation
methods without due consideration for the
potential  consequences.
                                          REVIEW EXISTING DATA
                                                (CHAPTER 7)
                                          Indusliy type and processes used
                                         Documented use or disposal ol DNM>L
                                           Avoitabte site, tocol. or  regional
                                               investigation report!
                                             Corporate/client record!
                                               Government records
                                       Uruveriily. library. historical society records
                                                   nth key personnel
                                                Aerial photographs
                                                                UNDERSTANDING  OF THE  DNAPL PROBLEM,
                                                                   DNAPL AND  MEDIA  PROPERTIES, AND
                                                                          TRANSPORT  PROCESSES
                                                                            (CHAPTERS 3.  4. & 5)
                                   [DEVELOP  INITIAL  CONCEPTUAL MODEL [«-
                                       SITE CHARACTERIZATION  ACTIVITIES
                           NONINVASIVE METHODS
                               (CHAPTER 8)

                              Surface geophysics
                               Soil 901 analysis
                             INVASIVE METHODS
                               (CHAPTER 9)

                                 lest pill
                                 Borings
                                  Wens
                               Hydraulic tests
                                 LABORATORY METHODS
                                     (CHAPTER  10)
   ESTIMATE
   QUANTITIES
    OF  DNAPL
   RELEASED
     AND IN
   SUBSURFACE

    Historical data
     F«M data
                         T
   DELINEATE

     DNAPL
    SOURCE

     AREAS

     EXAMINE
  SUSPECT AREAS

now drains ant sumps
    Catch basins
 Pits, ponds,  lagoons
 Other disposal oreos
    Septic lanki
    leach heMs
    French drains
      Sews
    Process tanks
   •ostewter tones
      USIs
  Mxweo/aund tanki
Cnemcal staraae areas
Chemical transfer areas
     Pipelines
 Vasle staraae areas
  Loadnt dock areas
     Wor1< anas
   Discolored sain
  Stressed waetatnn
   DaturtMd earth
Oislurtevt tow-rfr^ areas
                        T
DETERMINE

  DNAPL
   ZONE

   Delineate
   mobile
   DNAPl
   Cilnote
  saturations
                                            DJMPl
                     T
  DETERMINE
STRATIGRAPHY
  Suoljgiopnic traps
  Capillary bgrriera
 (presence and slope)
 Preferential pothMys
                          T
  DETERMINE
     FLUID
  PROPERTIES

      Density
     Vacosity
  InUrfociol tenwon
     Solubily
  SorplM prapgrties
 por transport properties
 Clwrrucal composition
CoKlvtncy and rnctnrity
                        T
 DETERMINE
FLUID-MEDIA
 PROPERTIES

    WettobiMy
  Pc (s,) relalnns
Relative pernteobiUies
                      T
DETERMINE
   MEDIA
PROPERTIES
  permeabilities
    Porosities
  Organic cortwn
    content
  Heterogeneities
                        UOR£ DATA IS M£DtD
                                                            REFINE  THE  CONCEPTUAL MODEL
                                                                   RISK  ASSESSMENT)
       -\ REMEDY  ASSESSMENT
                                                                                                    TREATABIUTY AND
                                                                                                     PILOT STUDIES
                                                SITE REMEDIATION
                                                 AND  LONG-TERM
                                                   MONITORING
 DETERMINE THE
NATURE, EXTENT.
MIGRATION  RATE,
  AND FATE OF
 CONTAMINANTS

NATURE AND EXTENT
       MM.
                                                                                              oraundvaler contamination
                                                                                                Mscrbtd soJ or*
                                                                                                rack contamination
                                                                                                   Soil an
                                                                                                  cantaminalian
                                                                                                 Solace Mler
                                                                                                  aid sediment
                                                                                                  contamination

                                                                                               MIGRATION RATE

                                                                                                    own.
                                                                                              Aqueous phase cNsmcoU
                                                                                                 do. directions
                                                                                                  end «locitiee

                                                                                                    FATE
                                                                                                 DWIPL dOJoUwn
                                                                                                ONlin. «eloUiie«ion
                                                                                                OHlft mrobiiation
                                                                                                 Adsorption oft
                                                                                                 dearadalion of
                                                                                                aqueous and «apor
                                                                                                phose canlamiiunts
                                Figure 1.   DNAPL  site  characterization flowchart
                                (modified  from Cohen and  Mercer,  1993).

-------
SITE CONCEPTUALIZATION

The subsurface movement of DNAPL is controlled substantially by the nature of the release, the  DNAPL density,
interfacial tension, and viscosity, porous media and capillary properties, and usually to a lesser extent, hydraulic forces.
Below the water table, non-wetting DNAPL migrates preferentially through permeable pathways such as soil and rock
fractures,  root holes, and sand layers that provide relatively little capillary resistance to  flow. There  are several
conceptualizations that can be made for DNAPL sites; two are illustrated in Figure 2. As shown, DNAPL chemicals can
migrate  in the subsurface as volatiles in soil gas, dissolved in groundwater, and  as a mobile, separate phase liquid.
Depending on how they are released, DNAPLs may be fairly widespread throughout the subsurface (Figure 2a) or they
may have a limited distribution (Figure 2b), making detection difficult. Several relatively simple quantitative methods
for examining DNAPL presence, migration, containment, dissolution, and mobilization within the context of various
conceptual models are described by Cohen and Mercer (1993).
DNAPL SITE  CHARACTERIZATION OBJECTIVES

The objectives of DNAPL site characterization include: (1) to determine DNAPL properties, (2) to identify DNAPL
release/source areas, (3) to define stratigraphy,  (4) to delineate DNAPL distribution, and (5) to minimize investigation
risk. DNAPL and media properties control DNAPL migration in the subsurface; Table 2 lists these properties. It is also
important to identify stratigraphic barriers and traps, and to  determine migration pathways, including man-made
features. As part of the DNAPL distribution, a distinction needs to be made between mobile versus DNAPL at residual.
Visual detection of DNAPL in soil and groundwater samples may be difficult where the DNAPL is colorless, present in
low saturation, or distributed  heterogeneously.  These factors confound  characterization of the  movement and
distribution of DNAPL even at sites with relatively homogeneous soils and a known, uniform DNAPL source.  The
difficulty of characterization is further compounded by fractured bedrock, heterogeneous strata, multiple DNAPL
mixtures and releases, and other complicating factors.
NONINVASIVE METHODS

Noninvasive methods can often be used during the early phases of field work to optimize the cost-effectiveness of
a DNAPL site characterization  program.   Specifically, surface geophysical surveys, soil  gas analysis, and
photointerpretation can facilitate characterization of contaminant source areas, geologic controls on contaminant
movement (i.e., stratigraphy and utilities), and the extent  of  subsurface contamination.  Conceptual model
refinements derived using these methods reduce the risk of spreading contaminants during subsequent invasive
field work.

Surface Geophysics

At contamination sites, geophysical surveys are usually conducted to: (1) assess stratigraphic and hydrogeologic
conditions; (2) detect and map electrically  conductive contaminants; (3) locate and delineate buried wastes and
utilities; (4) optimize the location and spacings of test pits, borings, and wells; and (5) facilitate interpolation of
subsurface conditions among boring locations.  Surface geophysical methods that potentially may be used at
DNAPL sites are identified in Table 3. Subsurface DNAPL is generally a poor target for conventional geophysical
methods.  Although ground-penetrating radar, EM conductivity, and complex resistivity have been used to infer
NAPL presence at a very limited number of sites, direct detection and mapping of non-conductive subsurface
DNAPL using surface geophysical techniques is an unclear, and apparently limited, emerging technology.  The
value of surface geophysics at  most  DNAPL sites will be  to aid characterization of waste disposal areas,
stratigraphic conditions, and potential routes of DNAPL migration.

Soil Gas  Analysis

Many DNAPLs, including most halogenated solvents, have high vapor pressures and will volatilize in the vadose
zone to form a vapor plume  around  a DNAPL source.  Volatile organic compounds (VOCs) dissolved in
groundwater can also volatilize at the capillary fringe into soil gas. Thus, soil gas surveys may be used to help
identify contaminated zones for further study.  Table 4 presents characteristics of various contaminants and their

-------
(a)  DNAPL chemcials are distributed in several phases: dissolved in
groundwater, adsorbed to soils, volatilized in soil gas, and as residual and
mobile immiscible fluids (modified from Huling and Weaver, 1991; WCGR,
1991).
                    Pool
                                 Finger
(b)  DNAPL chemicals in fingers and pools (modified from Anderson et al.,
1992).
              Figure 2.  DNAPL site conceptualizations.

-------
Table 2. Properties of DNAPL and Media that Influence Contaminant
Migration (from Mercer and Waddell, 1992).

PROPERTY
Saturation
Residual saturation
Interfacial tension
Wettability
Capillary pressure
Relative permeability
Solubility
Volatilization
Density
Viscosity
DEFINITION
The volume traction of the total void volume occupied by that
fluid at a point.
Saturation at which NAPL becomes discontinuous and is
immobilized by capillary forces.
The free surface energy at the interface formed between two
immiscible or nearly immiscible liquids. Surface tension is the
inlerfacial tension between a liquid and its own vapor.
Describes the preferential spreading of one fluid over solid
surfaces in a two-fluid system; it depends on imerfacial tension.
The difference between the non-wetting fluid pressure and the
wetting fluid pressure. It is a measure of the tendency of a
porous medium to attract the wetting fluid and repel the
non-wetting fluid.
The ratio of the effective permeability of a fluid at a fixed
saturation to the intrinsic permeability.
The maximum concentration of the chemical that will dissolve
under a set of chemical and physical conditions.
The transfer of matter from liquid and soil to the gaseous phase.
Mass per unit volume of a substance. Specific gravity is the
ratio of a substance's density to that of some standard substance,
usually water.


RANGE OF VALUES
0-1
Approximately 0. 1 -0.5.
Values of interfacial and surface tensions
for NAPL-forming chemicals generally
range between 0.01 5-0.5 N/m.
Determined through contact-angle
studies. Commonly, NAPL is the wetting
fluid in the vadose zone and the
non-wetting fluid in the saturated zone.
Depends on the interfaeial tension,
contact angle, and pore size.
0.0- 1 .0, depending on the fluid
saturation.
Varies widely depending on chemical
and aquifer conditions.
Depends on organic partitioning between
water and air, and NAPL and air.
Specific gravities of petroleum products
may be as low as 0.7, whereas
chlorinated aliphatic compounds can be
as high as 1.2 to 1.5.
Varies depending on fluid and
temperature.



relation to soil gas surveys. Several chemical characteristics indicate whethera measurable vapor concentration can be
detected. Ideally, compounds such as VOCs monitored using soil gas analysis will: (1) be subject to little retardation
in groundwater; (2) partition significantly from water to soil gas (Henry's Law constant >0.0005 atrn-m3/mole); (3) have
sufficient vapor pressure to diffuse significantly upward in the vadose zone (>0.00 1 3 atm @ 20o C); (4) be persistent:
and (5) be susceptible to detection and quantitation by affordable analytical techniques.
Studies during the 1980s generally indicated the utility of soil gas survey ing for delineating VOC source areas and VOC-
contaminated groundwater (e.g., Marrin and Thompson, 1987; Thompson and Marrin, 1987). More recently, well-
documented field experiments were conducted at the Borden, Ontario DNAPL research site to investigate the behavior
and distributions of TCE in soil gas caused by (1) vapor transport from a DNAPL source in the vadose zone and (2)
dissolved transport with groundwater from a DNAPL source below the water table (Hughes et al., 1990a; Rivett and
Cherry, 1991). The extent and magnitude of soil gas contamination derived from the vadose zone source was much
greater than that derived from the groundwater source. Rivett and Cherry ( 1 991 ) attribute the limited upward diffusion
of groundwater contaminants to the low vertical transverse dispersivities (mm range) which are observed in tracer
studies. Soil gas contamination, therefore, is not a reliable indicator of the distribution of DNAPL or groundwater
contamination at depth below the water table. Although the vadose zone source produced TCE concentrations in soil
gas and groundwater over a much wider area than the groundwater source site, the TCE in groundwater derived from the
vadose zone source was less concentrated than at the groundwater source site and was restricted to the upper 5 ft of the
saturated zone. Overall, these studies suggest that soil gas contamination will usually be dominated by volatilization
and vapor transport from contaminant sources in the vadose zone rather than from groundwater, and that the upward
transport of VOCs to the vadose zone from groundwater is probably limited to dissolved contaminants that are very near
to the water table.

-------
Air Photo Interpretation

Historic aerial photographs can be analyzed to document waste disposal practices and locations, geologic conditions,
drainage patterns, pooled fluids, site development including excavations for pipelines and underground storage tanks,
signs of vegetative stress, soil staining, and other factors relevant to assessing subsurface chemical migration. The Earth
Science Information Center  (ESIC) at the U.S. Geological Survey  in Reston, Virginia catalogs and disseminates
information about aerial photographs and satellite images available from public and private sources.  ESIC will provide
a listing of available aerial photographs for any location in the United States and order forms for their purchase.

Another use of air photos is to perform  fracture trace analysis. Fracture trace analysis involves stereoscopic study of
aerial photographs to identify surface expressions of vertical or nearly-vertical subsurface zones of fracture concentration.
In fractured rock terrain, particularly in karst areas, groundwater flow and chemical transport are usually concentrated
in fractures.
 Table  3.   Surface Geophysical  Methods for Evaluating DNAPL  Site
 Contamination (from Cohen and Mercer, 1993).
                   OBJECTIVE
                         METHODS
     Delineate limits of waste disposal areas.
Following review of historical documentation, interviews, aerial
photographs, and available site data, consider use of ground-penetrating radar
(GPR), electromagnetic (EM) conductivity, magnetometer, and metal
detection surveys.
     Delineate buried utility corridors.
Following review of historical documentation, utility records, interviews,
aerial photographs, and available site data, consider use of GPR, EM
conductivity, magnetometer, and metal detection surveys.
     Map stratigraphy, particularly permeable pathways
     and the surfaces of fine-grained capillary barrier
     layers and bedrock, to determine potential routes of
     DNAPL migration and stratigraphic traps.
Following review of aerial photographs and available site, local, and regional
hydrogeologic data, consider use of GPR, EM conductivity, electrical
resistivity, and high resolution seismic surveys.
     Delineate conductive inorganic contaminant
     plumes that may be associated with DNAPL
     contamination.
Following review of available data on waste disposal practices, chemical
migration, and hydrogeologic conditions, consider use of EM conductivity,
electrical resistivity surveys.
     Delineate geophysical anomalies that may result
     from an accumulation of DNAPL.
Consider using GPR to infer DNAPL accumulations at shallow depths in low
conductivity soils because DNAPL presence will probably alter the dielectric
properties of the subsurface (Olhoeft, 1986; WCGR, 1991). Also consider
using electrical methods to infer DNAPL accumulation based on the presence
of low conductivity anomalies (Davis, 1991). Note that the use of surface
geophysical methods for direct detection of DNAPL presence is an emerging,
but limited, technology that may not be cost-effective. Very few
geophysicists are experienced in the application of surface survey methods
for DNAPL detection.  These applications, therefore, should be treated with
caution.
     Note:  Survey results depend on site-specific conditions. At some sites, an alternate method may provide better results.

-------
Table 4. Characteristics of Contaminants in Relation to Soil Gas Surveying
(modified from Marrin, 1987).

GROUP/CONTAMINANTS
APPLICABILITY OF SOIL-GAS SURVEY
TECHNIQUES
Group A: Halogenated methanes, ethanes, and ethenes
Chloroform, vinyl chloride, carbon tetrachloride,
trichlorofluoromethane, TCA, EDB, TCE
Detectable in soil gas over a wide range of environmental
conditions. Dense nonaqueous phase liquid (DNAPL) will sink in
aquifer if present as pure liquid.
Group B: Halogenated propanes, propenes, and benzenes
Chlorobenzene, trichlorobenzene, 1 ,2-dichloropropane
Limited value; detectable by soil-gas techniques only where probes
can sample near contaminated soil or groundwater. DNAPL
Group C: Halogenated polycyclic aromatics
Aldrin, DDT, chlordane, heptachlor, PCBs
Do not partition into the gas phase adequately to be detected in soil
gas under normal circumstances. DNAPL
Group D: C, - Cg petroleum hydrocarbons
Benzene, toluene, xylene isomers, methane, ethane,
cyclohexane, gasoline, JP-4
Most predictably detected in shallow aquifers or leaking
underground storage tanks where probes can be driven near the
source of contamination. Light nonaqueous phase liquids
(LNAPLs) float as thin film on the water table. Can act as a
solvent for DNAPLs, keeping them nearer the ground surface.
Group E: C9 - C,2 petroleum hydrocarbons
Trimethylbenzene, naphthalene, decane, diesel and jet A
fuels
Limited value; detectable by soil gas techniques only where probes
can sample near contaminted soil or groundwater. DNAPL
Group F: Polycyclic aromatic hydrocarbons
Anthracene, benzopyrene, fluoranthene, chrysene, motor
oils, coal tars
Do not partition adequately into the gas phase to be detected in soil
gas under normal circumstances. DNAPL



INVASIVE METHODS
Following development of the site conceptual model based on available information and noninvasive field methods,
invasive techniques will generally be required to advance site characterization and enable the conduct of risk and remedy
assessments. Generally, these invasive activities include: (a) drilling and test pit excavation to characterize subsurface
solids and liquids; and (b) monitor well installation to sample fluids, and to conduct fluid level surveys, hydraulic tests,
and borehole geophysical surveys. Invasive field methods should be used in a phased manner to test and improve the
site conceptual model based on careful consideration of site-specific conditions and DNAPL transport processes.
Invasive Method Risks
The risk of enlarging the zone of chemical contamination by use of invasive methods is an important consideration that
must be evaluated during a DNAPL site investigation. Drilling, well installation, and pumping activities typically
present the greatest risk of promoting DNAPL migration during site investigation. Drilling and well installations may
create vertical pathways for DNAPL movement. In the absence of adequate sampling and monitoring as drilling
progresses, it is possible to drill through a DNAPL zone without detecting the presence of DNAPL (USEPA, 1992).
Increased hydraulic gradients caused by pumping may mobilize stagnant DNAPL. Pumping from beneath or adjacent
to the DNAPL zone can induce downward or lateral movement of DNAPL, particularly in fractured media due to the
development of relatively high fluid velocities. In general, groundwater should not be pumped from an uncontaminated
aquifer directly beneath a capillary barrier and overlying DNAPL zone. If the risks cannot be adequately minimized,

-------
alternate methods should be used, if possible, to achieve the characterization objective; or the objective should be
waived.

Drilling  Methods

Conventional drilling methods have  a high potential for promoting downward DNAPL migration  (USEPA, 1992).
Specific conditions that may result in downward DNAPL migration include: an open borehole during drilling and prior
to well installation; an unsealed or inadequately sealed borehole; a well screen that spans a barrier layer and connects
an overlying zone with perched DNAPL to a lower transmissive zone; an inadequately sealed well annulus that allows
DNAPL to migrate through the well-grout interface, the grout, the grout-formation interface, or vertically-connected
fractures in  the disturbed zone adjacent to the well; and, structural degradation of bentonite or grout sealant, or well
casing,  due to chemical effects of DNAPL or the groundwater environment.

To minimize the risk of inducing DNAPL migration as a result  of drilling, site  investigators should:  (1) avoid
unnecessary drilling within the DNAPL zone; (2) minimize the time during which a boring is open;  (3) minimize the
length of hole which is open at any time; (4) use telescoped casing drilling techniques to isolate shallow contaminated
zones from deeper zones; (5) use high-quality continuous sampling procedures (e.g., coring or split-spoon sampling) in
a potential DNAPL zone and carefully examine subsurface materials brought to the surface as drilling progresses to
avoid drilling through a barrier layer beneath DNAPL (i.e., stop drilling at the top of the barrier layer); (6) consider using
a dense drilling mud (i.e., with barium sulfate additives, also known as barite) or maintaining a high hydrostatic head
with water to prevent DNAPL from  sinking down the  borehole during drilling;  (7)  use  less-invasive direct-push
sampling methods where appropriate (e.g., cone penetrometer, Geoprobe®, and HydroPunch techniques); (8) select
optimum well materials and grouting  methods based on consideration of site-specific chemical compatibility; and (9)
if the long-term integrity of a particular grout sealant  is questionable, consider placing alternating layers of different
grout types and sealing the  entire distance between the well screen and surface to minimize the potential for vertical
migration of DNAPL.

The risk of  causing DNAPL  migration generally increases where there are fractured media, heterogeneous strata,
multiple release locations, large DNAPL release volumes, and barrier layers that are subtle (e.g., a thin silt layer beneath
sand) rather than obvious (e.g., a thick soft clay layer beneath sand). At many sites, the DNAPL zone can be adequately
characterized by limiting drilling to shallow depths. Characterization of deeper units can be accomplished by deeper
borings and  wells beyond the edge of the DNAPL zone. The "outside-in" approach whereby invasive activities are
conducted beyond suspected DNAPL areas to improve the conceptual model before drilling in the DNAPL zone also is
used to  reduce risks associated with DNAPL site characterization (USEPA, 1992).

Well  Construction

The design and construction of wells at DNAPL sites require special consideration of (1) the effect of well design and
location on immiscible fluid movement and distribution in the well and near-well environment; (2) the compatibility of
well materials with NAPLs and dissolved chemicals; and (3) well development options. Based on experiments, field
experience,  and the principles of DNAPL movement, it is apparent that  the following factors may cause the elevation
and thickness of DNAPL in a  well to differ from that in formation and/or lead to vertical DNAPL migration.

(1) If the well screen or casing extends below the top of a DNAPL barrier layer, a measured DNAPL pool thickness may
    exceed that in the  formation by the length of the well below the barrier layer surface.

(2) If the well bottom  is set above the top of a DNAPL barrier layer, the DNAPL thickness in the well may be less than
    the  formation thickness.

(3) If the well connects a DNAPL pool above a barrier layer to a deeper permeable formation, the DNAPL elevation and
    thickness in the well are likely to be erroneous and the well will cause DNAPL to short-circuit the barrier layer and
    contaminate the lower permeable formation. The height of the DNAPL column at the well bottom will tend to equal
    or be less than the critical DNAPL height required to overcome the capillary resistance offered by the sandpack and/
    or formation.

(4) DNAPL  which enters a coarse sandpack may sink to the bottom of the sandpack rather than flow through the  well
    screen. Small quantities of DNAPL may elude detection by sinking down the sandpack and accumulating below the
    base of the well screen.

(5)  Similarly, if the bottom of the well screen is set above the bottom of the sandpack and there is no casing beneath the
    screen, small quantities of DNAPL may elude detection by sinking out the base of the screen and into the underlying
    sandpack.                   ^	

                                                     8

-------
(6)  Hydrophilic (e.g., quartz) sandpacks generally should be coarser than the surrounding media to ensure that mobile
    DNAPL can enter the well. Screen or sandpack openings that are too small may act as a capillary barrier to DNAPL
    flow. Laboratory experiments (Hampton et al., 1991; Hampton and Heuvelhorst, 1990) indicate that (a) NAPL flows
    more readily into wells with uniform coarse hydrophilic sandpacks than wells with finer-grained and/or nonuniform
    hydrophilic sandpacks; and (b) NAPL recovery can be optimized using a hydrophobic sandpack material (such as
    teflon chips or made by applying a water-repellent coating to sand) with angular grains and a nonuniform grain size
    distribution.

(7)  If the well screen is located entirely within a DNAPL pool and water is pumped from the well, DNAPL will upcone
    in the well to maintain hydrostatic equilibrium causing the DNAPL thickness in  the well to exceed that in the
    formation.

(8)  The elevation of DNAPL in a well may exceed that in the adjacent formation by a length equivalent to the DNAPL-
    water capillary fringe height where the top of the pool is under drainage conditions.

(9)  DNAPL will not flow into a well where it is present  at or below residual saturation or at negative pressure.

A well that is completed to the top of a capillary barrier and screened from the capillary barrier surface to above the
DNAPL-water  interface  is most likely to provide DNAPL thickness and  elevation data that are representative of
formation conditions. Well development should be limited in wells containing DNAPL to gentle pumping and removal
of fine particles to minimize DNAPL redistribution. Measurements should  be made of immiscible fluid stratification
in the well prior to and after development.

Fluid Thickness,  Elevation, and Sampling  Surveys

Fluid elevation and thickness measurements are made in wells to assist determination of fluid potentials, flow directions,
and immiscible fluid distributions.  With knowledge of DNAPL entry areas, the surface slope of capillary barriers,
hydraulic data, and other observations of DNAPL presence, well data can be evaluated to infer the directions of DNAPL
migration. Interpretation of fluid data from wells containing NAPL may be complicated  by several factors related to the
measurement method, fluid properties, well design, or well location. DNAPL in wells, therefore, should be evaluated
in conjunction  with evidence of geologic conditions and DNAPL presence obtained during drilling.

While conducting immiscible fluid level and thickness measurements, care should be taken to slowly lower and raise the
measuring device within the well to avoid disturbing the immiscible fluid equilibrium and creating emulsions. Similarly,
measurements should be  made prior to purging and sampling activities.  The cost of purchase and decontamination of
the  measuring  device should be  considered when selecting a measurement method, particularly given uncertainties
involved in interpreting NAPL thickness and elevation data. Measurements are typically made using interface probes,
bailers, hydrocarbon-detection paste, or other methods (e.g., with a weighted cotton string).

Fluid sampling surveys should be conducted at potential DNAPL sites to examine wells for the presence of LNAPLs and
DNAPLs. NAPL samples can be tested for physical properties and chemical composition. Various sampling devices
can be employed to acquire fluid samples from the top and bottom of the well fluid column.   Villaume  (1985)
recommends use of a  bottom-loading bailer or mechanical discrete-depth sampler for  collecting DNAPL samples.
Huling and Weaver (1991) suggest that the best DNAPL sampler is a double check valve bailer which should be slowly
lowered to the well bottom and then slowly raised to provide the most reliable results.  DNAPL can be sampled from wells
with a shallow water table (<25 ft deep) with a peristaltic pump and from depths to approximately 300 ft using a simple
inertial pump.  An advantage of the peristaltic and inertial pumps is that fluid contact is  confined to inexpensive tubing
(and a foot-valve with the  inertial pump).  The cost to decontaminate or replace DNAPL-contaminated equipment is
usually a major factor  in selecting a sampling method.

Determining  DNAPL Presence

DNAPL presence can be:  (1) determined directly by  visual examination of subsurface samples; (2) inferred by
interpretation of chemical analyses of subsurface samples; and/or (3) suspected based on interpretation of anomalous
chemical distribution and hydrogeologic data. Methods to visually detect DNAPL in subsurface samples are identified
in Table 5.  Indirect methods for assessing the presence of DNAPL in the subsurface rely on comparing measured
chemical concentrations to effective solubility limits for groundwater and to  calculated equilibrium partitioning
concentrations for soil and  groundwater. Chemical data  indicative and/or suggestive of DNAPL presence are given in
Table 5.

-------
Table  5.   Determinant,  Inferential,  and  Suggestive  Indications of  DNAPL
Presence (modified from  Cohen and Mercer,  1993,  Cherry and Feenstra, 1991;
Newell  and  Ross, 1992;  and  Cohen  et  al.,  1992).
    DETERMINING DNAPL PRESENCE
      BY VISUAL EXAMINATION OF
          SUBSURFACE SAMPLES
INFERRING DNAPL PRESENCE
      BY INTERPRETING
    CHEMICAL ANALYSES
  SUSPECTING DNAPL PRESENCE
   BASED ON ANOMALOUS FIELD
             CONDITIONS
   Methods to detect DNAPL in wells:

   * NAPL/water interface probe detection of
   immiscible phase at base of fluid column

   * Pump from bottom of fluid column and
   inspect retrieved sample

   * Retrieve a transparent, bottom-loading bailer
   from the bottom of a well and inspect the fluid
   sample

   * Inspect fluid retrieved from the bottom of a
   well using a mechanical discrete-depth sampler

   * Inspect fluid retained ona weighted cotton
   string that was lowered down a well
   Methods to enhance inspection of fluid
   sample for DNAPL presence:

   * Centrifuge sample and look for phase
   separation

   * Add hydrophobic dye (such as Sudan IV or
   Red Oil) to sample, shake, and ltx>k for
   coloration of DNAPL fraction

   * Examine UV fluorescence of sample (many
   DNAPLs will fluoresce)

   * Assess density of NAPL relative to water
   (sinkers or floaters) by shaking solution or by
   using a syringe needle to inject NAPL globules
   into the water column
   Methods to detect DNAPL in soil and rock
   samples:

   * Examine UV fluorescence of sample (many
   DNAPLs will fluoresce)

   * Add hydrophobic dye and water to soil sample
   in polybag or jar, shake, and examine for
   coloration of the NAPL fraction

   * Conduct a soil-water shake test without
   hydrophobic dye (can be effective for NAPLs
   that are neither colorless nor the color of the soil

   * Centrifuge sample with water and look  for
   phase separtion

   * Perform a paint filter test, in which soil is
   placed in a filter funnel, water is added, and the
   filter is examined for separate phases
Chemical analysis results from which
DNAPL presence ca be inferred (with
more or less certainty depending on the
strenth of the overall data):

* Concentrations of DNAPL chemicals in
groundwater are greater than 1 % of the
pure phase solubility on effective
solubility

* Concentrations of DNAPL chemicals
on soils are greater than 10,000 mg/kg
(equal to 1% soil mass)

* Concentrations of DNAPL chemicals in
groundwater calculated from water/soil
partitioning relationships and soil samples
are greater than pure phase solubility or
effective solubility

* Organic vapor concentrations detected
in soil gas exceeds 100-1000 ppm
Field conditions that suggest DNAPL
presence:

* Concentrations of DNAPL chemicals increase
with depth in a pattern that cannot be explained
by advective transport

* Concentrations of DNAPL chemicals increase
up the hydraulic gradient from the contaminant
release area (apparently due to containated soil
gas migration and/or, DNAPL movement along
capillary and/or permeability interfaces that
slope counter to the hydraulic gradient)

* Erratic patterns of dissolved concentrations of
DNAPL chemicals in groundwater which are
typical of DNAPL sites due to heterogeneity of
(I) the DNAPL distribution, (2) the porous
media, (3) well construction details, and (4)
sampling protocols

* Erratic, localized, very high contaminant
concentrations  in soil gas, particularly located
just above the water table (where dense gas
derived from DNAPL in the vadose zone will
tend to accumulate)

* Dissolved DNAPL chemical concentrations in
recovered groundwater that decrease with time
during a pump-and-treat operation, but then
increase significantly after the pumps are turned
off (although complexities of contaminant
desorption, formation heterogeneity, and
temporal and spatial variations of the
contaminant source strength can produce similar
results)

* The presence of dissolved DNAPL chemicals
in groundwater that is older than potential
contamiant releases (using age dating) suggests
DNAPL migration

* Deterioration of wells and pumps (can be
caused by DNAPL; i.e., chlorinated solvents
degrade PVC)

* Patterns of dissolved chemicals that may be
indicative of pulsed releases associated with
recharge events through a DNAPL zone

* Plume behavior where most of the plume mass
is still  near the source area, even after adequate
time has passed to allow dissolved transport
away from the source area

* Mass removed during remediation far exceeds
original calculation of dissolved and adsorbed
mass-in-place
                                                             10

-------
Where present as a separate phase, DNAPL compounds are generally detected at <10% of their aqueous solubility limit
in groundwater. This is due to the effects of non-uniform groundwater flow, variable DNAPL distribution, the mixing
of groundwater in a well, and the reduced effective solubility of individual compounds in a multi-liquid NAPL mixture.
Typically, dissolved contaminant concentrations >1% of the aqueous solubility limit are highly suggestive of NAPL
presence.  Concentrations <1 %, however, do not preclude the presence of NAPL.  In soil, contaminant concentrations
in the percent range are generally indicative of NAPL presence.  However, NAPL may also be present at much lower soil
concentrations.  Feenstra et al. (1991) detail an equilibrium partitioning method for assessing the presence of NAPL in
soil samples based on determining total chemical concentrations, soil moisture content,  porosity, organic carbon
content, approximate composition of the possible NAPL, sorption parameters, and solubilities.
INTEGRATED DATA ANALYSIS

There is no practical cookbook approach to site investigation or data analysis. In addition to the site characterization
techniques described herein, many additional methods (i.e., using tracers, interpreting chemical distributions and ratios,
and conducting hydraulic tests) can be used to enhance site evaluation. Each site presents variations of contaminant
transport conditions and issues.  Site characterization, data analysis, and conceptual model refinement are iterative
activities which should satisfy the characterization objectives outlined in Figure 1 as needed to converge to a final
remedy.  During the process, acquired data should be utilized to guide ongoing investigations. For example, careful
examination of soil, rock, and fluid samples obtained as drilling progresses should be made to identify DNAPL presence
and potential barrier layers and thereby guide decisions regarding continued drilling, well construction, and/or borehole
abandonment.  Geologic, fluid elevation, and chemical distribution data should be organized (preferably using database,
CAD, and/or GIS programs) and displayed on maps that are updated periodically to help determine the worth of
additional data collection activities. With continued refinement of the site conceptual model, the benefit, cost, and risk
of additional work can and should be evaluated with improved accuracy.  This is the advantage of a flexible, phased
approach to site characterization.
REFERENCES

Anderson, M.R., R.L. Johnson, and J.F. Pankow, 1992. Dissolution of dense chlorinated solvents into groundwater: 3.
    Modeling contaminant plumes from fingers and pools of solvent,Environmental Science and Technology, 26(5):901 -
    908.
Cherry, J. A. and S. Feenstra, 1991. Identification of DNAPL sites: An eleven point approach, draft document in Dense
    Immiscible Phase Liquid Contaminants in Porous and Fractured Media, short course notes, Waterloo Centre for
    Ground Water Research, Kitchener, Ontario.
Cohen, R.M., A.P. Bryda, S.T. Shaw, and C.P. Spalding, 1992. Evaluation of visual methods to detect NAPL in soil and
    water, Ground Water Monitoring Review, 12(4): 132-141.
Cohen, R.M. and J.W. Mercer, 1993. DNAPL Site Evaluation, Lewis Publishers, Chelsea, MI.
Davis, J.O., 1991. Depth zoning and specializing processing methods for electromagnetic geophysical surveys to remote
    sense hydrocarbon type groundwater contaminants, Proceedings of the Fifth National Outdoor Action Conference
    on Aquifer Restoration, Ground Water Monitoring,  and Geophysical Methods, Las Vegas, NV, pp. 905-913.
Huling, S.G. and J.W. Weaver, 1991. Dense nonaqueous phase liquids, USEPA Groundwater Issue Paper, EPA/540/4-
    91, 21 pp.
Marrin, D.L.,  1987.  Soil gas  sampling strategies:  Deep vs. shallow aquifers, Proceedings of 1 st National Outdoor
    Action Conference on Aquifer Restoration, Ground Water Monitoring and Geophysical Methods, National Water
    Well Association, Dublin, OH, pp. 437-454.
Mercer, J.W. and R.K. Waddell, 1993. Contaminant transport in groundwater (Chapter 16) in Handbook of Hydrology,
    D.R. Maidment (Editor), McGraw-Hill, New York.
Newell, C.J. and R.R. Ross, 1992. Estimating potential for occurrence of DNAPL at Superfund sites, USEPA Quick
    Reference Fact Sheet, R.S. Kerr Environmental Research Laboratory, Ada, OK.
Olhoeft, G.R., 1986.  Direct  detection of hydrocarbons and organic  chemicals with ground penetrating radar and
    complex resistivity, Proceedings of Petroleum Hydrocarbons andOrganic Chemicals in GroundWater: Prevention,
    Detection, and Restoration, National Water Well Association, American Petroleum Institute, Houston, TX, pp. 284-
    305.
USEPA, 1992. Dense nonaqueous phase liquids - A workshop summary, Dallas, TX, April 17-18, 1991, EPA/600/R-
    92/030, Robert S. Kerr Environmental Research Laboratory, Ada, OK.
Villaume, J.F., 1985. Investigations at sites contaminated with DNAPLs, Ground Water Monitoring Review, 5(2):60-
    74.
WCGR,  1991.  DNAPL short  course notes, Dense Immiscible Phase Liquid Contaminants (DNAPLs) in Porous and
    Fractured Media, A short course, October 7-10,  Kitchner Ontario, Canada, Waterloo Centre for Ground Water
    Research, University of Waterloo.


                                                   11

-------
NOTICE: The policies and procedures set out in this document are intended solely as guidance.  They are not
intended, nor can they be relied upon, to create any rights enforceable by any party in litigation with the United
States. EPA officials may decide to follow the guidance provided in this memorandum, or to act at variance with
the guidance, based on an analysis of specific site circumstances.  The Agency also reserves the right to change
this guidance at any time without public notice.
       For more information, contact:
               Randall Breeden
               U.S. Environmental Protection Agency
               Hazardous Site Control Division (5203G)
               401 M Street, S.W.
               Washington, DC  20460

       Authors:
               Office of Emergency and Remedial Response/Office of Research and Development
               401 M Street, S.W.
               Washington, DC  20460
                                                  12

-------