&EPA
                            United States
                            Environmental Protection
                            Agency
                         Risk Reduction
                         Engineering Laboratory
                         Cincinnati, OH 45268
Research and Development    EPA/600/M-89/033   July 1990
 ENVIRONMENTAL
 RESEARCH   BRIEF
       INVESTIGATION OF FAILURE  MECHANISMS AND MIGRATION OF
               INDUSTRIAL CHEMICALS AT WILSONVILLE,  ILLINOIS
                             Beverly L. Herzog and Robert A. Griffin*
Abstract
In late 1981, ground water contamination was discovered in
a monitoring well at  the  Earthline disposal  facility  near
Wilsonville, Illinois.  This meant that organic chemicals had
migrated at a rate of  100 to 1000 times the rate predicted
when the site was given its  permit in 1976. The Illinois  State
Geological Survey  (ISGS) then conducted  a  3-yr,
multidisciplinary study to determine why the site failed  to
perform  as predicted. Geology, hydrogeology,  geo-
chemistry, engineering geology, and x-ray mineralogy  were
disciplines used in the study of this site.

Postulated failure mechanisms included migration through
previously undetermined (unmapped)  permeable zones,
subsidence of an underground mine, organic chemical/clay
interactions, acid mine drainage/clay interactions, and trench
cover settlement and erosion. The study concluded that the
primary reason for the rapid migration was the presence  of
previously undetermined permeable zones that' included
fractured and jointed glacial till formations. The inaccurate
predictions were  caused by  the use of laboratory-
determined hydraulic  conductivity values, which did not
adequately measure the effects of fractures and joints in the
transit  time calculations. Field-measured hydraulic
conductivity values were generally 10 to 1000 times greater
than their laboratory measured  counterparts,  thus largely
accounting  for the  discrepancy  between predicted  and
actual migration rates in the transit time calculations.  This
however was compounded by the burial of liquid wastes and
by trench covers that allowed excess infiltration to enter the
trenches. Organic chemical/clay interactions may also  have
 The authors are with Illinois State Geological Survey, Champaign, Illinois
 61820.
                      exacerbated the  problem in areas  where  liquid organic
                      wastes were buried.

                      This Research Brief was developed  by  the  principal
                      investigators and EPA's Risk Reduction Engineering
                      Laboratory, Cincinnati, OH, to announce key findings of the
                      research project that is fully documented in the reports and
                      publications listed at the end.

                      introduction

                      Earthline Corporation,  a subsidiary of SCA  Services,  Inc.,
                      began operating a 130-acre landfill near Wilsonville, Illinois,
                      on November 15,  1976. The operation was a trench-and-fill
                      procedure that relied on a clayey glacial till deposit native to
                      the site for natural attenuation of any leachate. A compact-
                      ed clay liner was reportedly used to supplement native till in
                      at least one of the trenches. The company had applied for
                      and  received a  permit  from  the Illinois  Environmental
                      Protection Agency (IEPA) to  dispose  of  industrial  and
                      hazardous wastes at the site.
                      Several months after  the landfill was  in operation, the
                      citizens of Wilsonville became  alarmed  at the disposal of
                      hazardous wastes near their community. They filed suit to
                      stop the disposal of wastes and  to have them removed from
                      the site. A lengthy court battle ensued,  and  Earthline
                      continued to bury wastes. In  1981, the Illinois  Supreme
                      Court affirmed the  1978 trial court's  ruling  that the
                      hazardous wastes buried in the  26 trenches (each 10 to 20
                      ft deep, 50 to 100 ft wide, and 175  to 400 ft long) at
                      Wilsonville must be exhumed and removed  from the  site.
                      SCA Services, Inc., announced in March 1982 that they
                      were dropping  further appeals and would comply with the
                      court order. The preparation began in the summer of 1982,
                      and the actual exhumation and  removal process, begun on
                      September 7, 1982, was expected to take approximately 4
                      yr. Earlier that  year, in January 1982, the IEPA confirmed

-------
that organic pollutants had migrated as far as 50 ft from the
trenches in a 3-yr  period. The levels of volatile, organic
priority  pollutants found were reported  by Johnson jet al.
(1983). This discovery was a separate issue from the  court
proceedings  and exhumation order. The migrationi  rates
were 100 to 1000 times faster than  predicted. Two opvious
questions  were  posed:  (1) why were  these  organic
compounds migrating faster than predicted and (2) what are
the implications  to land disposal of similar wastes at  other
sites?  This research  project  was designed  to provide
answers to these and many  other questions regarding the
efficacy of land  disposal of hazardous wastes - particularly
organic liquids.                                   '
The  study of the site  was  a cooperative  effort between
several agencies  and  the  site  owner.  Thej  U.S.
Environmental Protection Agency supplied a major part  of
the funding through a cooperative agreement with  the ISGS,
whose personnel performed  a majority  of the  work!   The
IEPA provided a drill rig and crew for the field  studies and
performed  organic analyses on soil and water samples from
the site. The site  owner, SCA  Services,  Inc.,  pr0vided
access to the site, safety training, and a substantial amount
of the  materials for construction of the  monitoring wells.
Waste Management, Inc., which later took over ownership of
the  site,  continued to cooperate  with  the  study  and
additionally provided some analytical support for monitoring
well samples.                                    '

Objectives
To answer the two major research  questions, the scjope of
work needed to include studies of several aspects [of site
behavior.                                         :
     1.  Site characterization-detailed  descriptions  of
        geologic materials,  geomorphology,  and; hydro-
        geology; comparison  of field  and  laboratory
        measurements of hydraulic  conductivity; bignifi-
        cance of fracture flow.                     j
    2.  Groundwater  quality-distribution  of organic
        chemicals  across the  site, sampling methodology
        for volatile organic compounds  in fine-grained sed-
        iments.
    3.  Organic  chemical/clay interactions-laboratory
        studies to determine effects of organic  chemicals
        on permeability and  pore structure of clay sojils.
     4.  Gob pile  effects-effects of  acidity  and high
        inorganic salt  content  of  leachate from adjacent
        coal refuse pile (gob pile) on the trench materials.
     5.  Trench cover studies-observations of  condition of
        trench  covers, differential settlement,  and  mine
        subsidence.                              I
     6.  Condition of  drums  and  wastes-photographic
        documentation of effects of leachate on drums.
A multidisciplinary/multiagency approach was adopted early
in the  planning  of this project.  Cooperation with  the  site
operators  improved access  to the site and reduced legal
problems. Consolidation of efforts also reduced the number
of borings, samples, and analyses required.
Site Characterization

Geology
An extensive geologic  investigation and description of the
site was carried out to  place this site in the proper regional
geologic  framework and to collect sufficient  baseline data
for extrapolating the results from this  investigation to other
sites. The geologic characterization was carried out by four
principal  means: (1) examination of all previously gathered
data  and  information,  (2) investigation  of  outcrops and
exposures at and in the vicinity of the site, (3) study  of the
trench walls  themselves  and  backhoe  pits at selected
locations on and around the  site, and  (4)  study of drill hole
samples  collected on  and around the site.  The  location of
the drill hole samples coincides with well  nests and profiles
shown on the site map  (Figure  1).
                            Mine Cleaning Refuse
       — Water table (m), April 1984
 Figure 1. Map showing water table elevation (m) in April, 1984;
         locations of trenches, wells and cross section.

 A detailed geologic description of the site was presented by
 Follmer (1984); a cross section is given in Figure 2 following
 the orientation shown in Figure 1.
 The first unit encountered at the land surface is the  modern
 soil  developed in the Peoria  Loess. It is orange-brown  in
 color and is variable in thickness across the site. The loess
 itself is a windblown, clayey silt. The combined  thickness of
 the Peoria Loess and modern soil ranges from  1 ft to 4 ft.
 Beneath the  Peoria  Loess is  the  Roxana Silt with the
 Farmdale Soil developed  at  the top of the silt. This silt,
 which has a higher sand content than the Peoria Loess, is 1
 ft to  3-ft thick. Underlying the loess  is the Vandalia Till,
 which   has   three  phases:  an   upper,   brown, stiff
 weathered zone, a brown, brittle  weathered zone, and a gray

-------
 unweathered zone. An ancient soil, the Sangamon paleosoi,
 is developed in the upper portion of the till and varies from 8
 ft  to 12  ft in  thickness at the site. Discontinuous sand
 lenses are found widely scattered throughout the Vandalia
 Till, but are more common at the  base of the unit. The
 Vandalia Till, below the  Sangamon Soil, typically varies in
 thickness from  5 ft  to 20 ft. This till overlies an older till of
 the Banner Formation (Follmer,. 1984). The thickness of this
 formation is unknown at most  locations, since few  borings
 penetrated  it  completely.  One  boring,  at Nest  1,
 encountered nearly 50 ft of the Banner Formation.

 An  investigation  designed  to  detect  chemical  and
 mineralogical  effects  in  Pleistocene  deposits  by  the
 overlying gob  pile  led  to the observation of mixed-layer
[ kaolinite/smectite (K/S) in the Sangamon  paleosoi. This clay
 mineral is an indicator of weathering in well-drained soils. Its
 irregular distribution  suggests  a complex pattern  of well-
[ drained  and poorly drained soils  during the  Sangamon
 interval. It  is probable  that permeability and  degree  of
 cracking in the paleosoi also vary with the paleodrainage
 and/or paleotopography. No alteration of feldspar to kaolinite
 was observed.

 Hydrology

 Hydrogeologic  and geochemical investigations  of  the  site
 were  carried  out using  borings  for  installation  of 11
 piezometer nests,  nine monitoring well nests and two well
 profiles (Figure.  1). Piezometer nests included three to six
 piezometers, whereas well nests contained two to four wells.
 Eight to ten wells were  included in well profiles. Details of
 piezometer construction  are presented in Herzog and Morse
 (1986)  and Griffin et  al. (1984). These piezometers were
 initially installed for in-situ hydraulic conductivity tests at the
 various depths of the piezometers. Later these piezometers
 were used to  establish  the long-term piezometric surface
 and, in turn, the hydraulic gradient and flow across  the site.
 Core samples  from these borings were used for chemical
 analysis. The  chemistry of the water was not analyzed
 because of interference caused by the water added during
 the hydraulic  conductivity  slug  tests.  Slug tests were
 analyzed  by three methods:  Cooper et  al. (1967), with
 additional type curves by Papadopulos et al. (1973); Nguyen
 and  Pinder (1984); and Hvorslev (1951). These tests are
 discussed in  detail in  Herzog and  Morse  (1986).  Results
 from these slug tests are given in Table 1.

 Vertical holes were used to measure hydraulic conductivity
 in the  horizontal direction,  and angle holes  at separate
 locations to measure  hydraulic conductivity in  the  vertical
 direction.  Three nests,  containing three  to four holes per
 nest were drilled  on a 45° angle to intersect possible
 vertical   fractures.  In-situ  hydraulic  conductivity
 measurements  were carried out; these results are also
 presented in Table 1.

 Differences between  vertical  and  horizontal  hydraulic
 conductivity ranged from near agreement for a soft ablation
 zone to more  than a factor of 10  in  the  unaltered basal
 Vandalia.  The  higher ratios in the  basal Vandalia till are
 believed due to joints in  the till  that were observed to have a
 predominately vertical orientation. Vertical  joints were
 observed  in nearly all  exposures of the till, in backhoe
 trenches on the site, and in the continuous cores collected
from angle-hole drilling. These joints were  common  in the
Sangamon Soil and decreased in. number with depth.

The  field-determined, hydraulic  conductivity values were
consistent between methods (Herzog and Morse, in press).
Although not all piezometers could be reasonably analyzed
using all three slug test methods, all could  be  analyzed by
at least one method. These  values are 10 to  1000 times
greater than the laboratory-determined values (Griffin  et. al.,
1984; Herzog, and  Morse,  1986).  These  differences  are
believed due to the discontinuous joints present in the tills
and to scale effects.

A separate set of monitoring wells was constructed for water
quality samples. Hydraulic conductivity was measured on
the monitoring wells during development with the use of the
recovery test as described by Todd (1980). These results
are also given in Table  1.  Recovery  test results were
generally higher than were slug test  results. This is believed
to be the result of  an attempt to locate the  monitoring wells
in the more permeable zones.

ISGS monitoring well piezometers and SCA  monitoring wells
were  used  to  measure  the potentiometric  surface in each
zone. The water table is shown in Figure 1. Potentiometric
surfaces of deeper zones showed the same general pattern
with  lower  water  levels,  indicating  that  the zones  are
connected and the vertical gradient is downward. Figure 1
also shows the influence of the gob pile on groundwater flow
at the site; groundwater flows radially outward from the gob
pile.

Geoc/iem/sfry

Sampling Protocol
As part of  the groundwater  sampling program, time-series
samples were analyzed to  help establish  a  protocol  for
sampling wells finished in fine-grained sediments.  These
tests, described  more fully in Griffin et al. (1985), showed
dramatic changes  in concentration with time after  purging.
Most volatile organic compounds were found in their  lowest
concentration  before  purging  and  their  greatest
concentration  after 2 hr  to  8  hr of recharge.  Additional
research is  needed however to verify the trends observed
during this series of experiments.

Contaminant  Distribution'
Organic compounds  were analyzed  both from  soil  samples
and from monitoring  wells. The  soil  samples were collected
as the bore holes for placement of the monitoring wells were
drilled.  As  reported  by Griffin et al.  (1984),  the highest
concentrations of  organic chemicals  were found  in the
Vandalia Till ablation and weathered basal zones over most
of the site. At nest V, the zone of highest contamination was
the weathered basal  Vandalia Till. These zones have  higher
hydraulic conductivities  than  both the  overlying  and
underlying zones, as can be seen in  Table 1.

Monitoring  well  results  indicated that organic  compounds
were also found in the unweathered Vandalia Till. A  typical
contaminant distribution pattern with trichlorethylene used
as an example, is  shown in  Figure 2. The highest  levels of
contamination  on  the  site were  found on the southwest
corner  of French  Area A,  at  nests W1  and  B,  in the
unaltered basal  Vandalia, where  endrin and dieldrin were

-------
                Table 1. Geometric Means of Field Hydraulic Conductivity Data
                                                            Slug tests (cm/s)
Material
Peoria Loess/gob
contact
Vandalia Till ablation
aone-soft,clayey
Vandalia Till ablation
zone-soft, mushy
Base of ablation
zone/top of altered
Vandalia Till
Basal Vandalia Till-
altered, jointed
Basal Vandalia Till-
unaltered
Vandalia Till/Banner
Formation Contact
Banner
Formation/bedrock
contact
Orientation
Vertical

Vertical 45°
angle
Vertical
45 "angle
Cooper et al.
method


1.3 x 10~7
3.8 x 10-5
7.5 x 10-5
Vertical j

Vertical 45°
angle
Vertical
45°angle
Vertical

1.2 x 10-6
8.5 x 10-6
8.4 x 10-7
5.7 x 10'7


Vertical

i
Nguyen &
Finder


1.2 x 20-7
6.4 x 10-7
1.9 x 10-5
2.4 x 10-s


2.3 x ID'6
1.2 X10-S
6.5 x 10-8
9.4 x 10-7


1.1 x 10-7


Hvorslev
method


1.2 x 10-7
4.3 x 10-7
1.2 x 10-7
1.2 x 10-


8.4 x 10-7
3.0 x 10-6
3.9 x 10-8
8.0 x 10-7


2.8 x 10'6


Recovery test
4.8 x 10-3

2.0 x 10-6
1.6 x 10-5
1.3 x 10-6

4.0 x 10-6
6.8 xio-7

9.9 X10-7

1.7 x 10-7


present in concentrations  greater than 1%.  High levjels of
contamination  in  the unaltered Vandalia  suggest that
contaminants are moving  downward through  sand lenses
and interconnected joints.                           '
                                                  i
Organic Chemical/Clay Interactions

X-Ray Diffraction Studies                        :

To establish  a procedure for rapidly  screening chemical
effects on clay  liners and covers,  a  study  was  done of
changes in  interplanar spacing  of  smectite  in  a sodium
bentonite gel caused by different chemicals. The  purpose of
this  research  was to study  the  mechanisms involved in
organic chemical/clay interactions. It was  not  meant]to be
site-specific to Wilsonville.

X-ray diffraction (XRD) was used  to  rapidly  determine
changes in basal interplaner spacing (001). A smectite [gel in
demineralized water has an (001) spacing of 100 to 150 + A.
Many chemicals induce a collapse of the  (001)  spaqing to
12 to 20 A. This collapse is associated  with flocculatibn and
crack formation in  the gel, which, if occurring in the| field,
could result  in large increases  in  permeability  of a clay
barrier.  Past studies had suggested  that the  effects of
chemicals on a smectite gel would be similar to  the effects
on nonswelling colloidal particles, although  the magnitude of
permeability change should be much less.            }
                                                  I
The  XRD study confirmed estimates of the type of changes
that  occur. It was found  that salts, alcohols, and  several
other chemicals that  were miscible in water caused^ rapid
collapse, flocculation, and crack formation in the bentonite
gel.  High concentrations of sodium collapsed smectit^ to 17
A as compared with  19A when saturated with calciurn ions.
Calcium replacement usually causes collapse. Expansion of
smectite  by organic chemicals  is  known  to  be  roughly
proportional to their dielectric constants. Unfortunately, only
compounds miscible in water could  be studied  by this
technique,  and  many  of the  problem  chemicals  at
Wilsonville and other hazardous waste  sites  are  mostly
immiscible in water.

In summary, water, acetone, and various  dispersants cause
smectite  to expand and  seal clay  liners.  Most  organic
chemicals, metal cations with a charge above +1, and high
levels of  univalent ions cause collapse of the smectite with
flocculation and cracking.  These results  indicate that non-
expanding  clay minerals might provide a better barrier to
waste  migration,  especially  if  these  liner  materials  are
locally available and could  be obtained at low enough cost
to produce a significantly thicker barrier.
Viscosity and Test Tube Studies

Because the XRD experiments were limited by the types of
chemicals that could be evaluated, a series of experiments
were  designed to see  if viscosity measurements  and/or
change in gel structure in a sealed test tube could be used
to rapidly screen new chemicals and mixtures of chemicals
for their effect on liner integrity. The viscosity tests were run
on  sodium smectite gels,  and  XRD determinations  were
made before the samples were sealed in test tubes for still
further observation.

Results matched very well with earlier XRD studies. It was
again impossible to perform meaningful tests  on immiscible
chemicals.  An attempt was also made to use acetone as an
immiscibility "bridge," but the approach was not successful.

-------
                        ft   m
                      640-4-195
                                      SW
                      630-
                       620-
                       610
                       600-
                       590-
                       580
                           -190
                           -185
                           -180
^afer Table JA_priM2,J984)|_


       Trench Area B
                                                           Va'ndalia ablation tiHT
                                                       / /Basal Vandalia Till, altered, jointed
                 Basal Vandalia Till,
                    unaltered
                                  - 625 - Equipotential line (ft)
                                    h    Well screen (well 1D)
 Figure 2. Cross from Profile V through Trench Area B to the gob pile showing distribution and trichlorethylene (ppb) in
 groundwater, September 1984.
A viscosity decrease, flocculation,  crack formation,  and
(001) collapse were all closely correlated.

Future work should include  more  attention to partitioning
coefficients for immiscible chemicals. After the degree  of
chemical  adsorption  on a water-wetted clay surface  is
determined, other tests should be  made to determine the
effect on  liner integrity.  More effort should be devoted  to
evaluate locally available clays for use as clay barriers and
to compare those materials with bentonite. A study of rate of
loss  of water-immiscible  wastes from  organophillic
adsorbent clays is also needed to estimate waste migration.
Finally, both hydrophilic and  organophilic systems need  to
be studied with  complex mixtures of chemicals  under
realistic field conditions.

Gob Pile  Effects

A study of the gob pile (coal cleaning refuse) was performed
to determine  if acid  mine drainage increased the rate  of
contaminant migration. Measurements of soil pH were taken
at increments of approximately 1 ft at Nest I. Values of pH in
the gob ranged from 1.85 to 2.40, with the  lower  values
occurring  at  the bottom  of  the gob pile.  The  pH
measurement immediately below the gob  pile was 3.25.
Within 8 ft of the bottom of  the  gob pile, pH values were
greater than 7 and between 7 and 9 for the remainder of the
stratigraphic column.

Measurements of pH  were also made at monitoring  wells in
Nest A and  Profiles  V  and W. All water samples  had pH
values between 6 and 7. It was concluded that the effect of
.acid mine  drainage was highly localized and did  not affect
chemical migration in the trenches at the site.
    Trench Cover Studies

    Cover Condition

    The condition of the  covers was examined as the covers
    were removed to determine the effect of surface  and  inter-
    nal  erosion. Investigative  techniques  included  photo-
    interpretation, remote  sensing, and soil  erosivity  testing
    (Stohretal., 1985).

    Depressions  and sinkholes in the trench covers were found
    by field inspection, supplemented by interpretation of large
    scale aerial photography. Most severe depressions occurred
    over trenches  that extended to  nearby  contaminated
    monitoring wells (Stohr et al.,  1985). Not all contaminated
    monitoring  wells  were  associated with  depressions,
    however.  Freely drained  depressions were  distinguished
    from water-holding depressions with the use of a  post-
    sunset  thermal  infrared  ground survey.  Freely drained
    depressions were slightly  warmer than the adjacent ground,
    whereas water-retaining depressions were relatively cooler.

    Soil erosivity was  tested using the  pinhole techniques
    developed by  Sherard et al. (1976) to determine  the
    susceptibility of trench cover materials to  piping.  The
    qualitative results ranged between  "unsusceptible"  and
    "highly susceptible" to piping. A cover in French Area B,
    which  contained  sinkholes,  included both severely  erosive
    and very resistant soils. No  depressions were  found in
    French Area A  and its cover was  moderately resistant to
    piping  (Stohr et al., 1987).  This suggests that the  pinhole
    test would be useful for selecting cover material.

-------
Differential Settlement

Tilt plates and a precision (third-order) survey were us|ed to
monitor settlement, differential settlement, and tilt of jsome
of the trench  covers (Griffin et al., 1983). A  "Tiltrrjeter"
(manufactured  by Slope Indicator  Company*) recorded
tilting measurements at points on 11 trench covers.    I
                                                  i
Multiple  linear regression was  employed to  model  the
relationships between tilting  plates and  climatic  variables
(temperature and precipitation) for 3, 7, and 14 days before
surveying. The  model  accounts for only  34% of  the
movement of the tilt plates.
                                                  t
Both  temperature and  precipitation,  at  days  3  and  14,
however,  correlated  inversely with the tilt plate  elevation
data during the study period. This movement  is believed to
loosen the compacted soil thereby increasing void pace  and
pore  size. The  surface of the  cover  appears  ti)  be
particularly vulnerable to these effects.               |

Mine Subsidence
                                                  i
Subsidence  of an underground coal mine, located 300 ft
below the site, was postulated to cause  trench instability
and, therefore, to increase rates of chemical migration.

Stereoscopic examination of historical aerial photographs
showed no depressions  adjacent to  the  landfill site. Tree
canopies obscured the ground in some areas; however, no
pattern  of tilting  trees  was  observed.  A  ground
reconnaissance  survey  done  before  the  exhumjation
activities found no surface depression or other indications of
mine subsidence at the landfill site.

Monthly  precision (3rd  order)  vertical surveys  of deep
settlement probes set at or below the bottom of the burial
trenches were  done to  measure near-surface  movements
that indicate underground mine stability. Statistical analysis
by  Pearson correlation  coefficients showed that all,  except
one,  of the deep  probes  were  highly  correlated.; The
movement of the exception probe is  believed to be d|ue to
slope instability, because  of its proximity to a steep slope.
Therefore, because no mine subsidence was present at the
site, mine subsidence did not affect the  rate of  chemical
migration.

Conditions of Drums and Wastes               >

As  drums and other  wastes  were  removed from  the
trenches,  their condition was documented to  help interpret
the effects of the leachate on the drums and earth materials,
the relative leachate strength, and drum life expectancy^.
On three occasions, stereo photography was used to record
trench investigations. A 2-ft aluminum cube assembled on
site  was used  for interior  orientation of the photography
(Stohr, 1983). Photographic  observations of drums and other
waste containers were  made periodically  during  the landfill
exhumation. Observations and photographs were not made
systematically, however.                            ;
 "Mention of  trade names or commercial products does not constitute
 endorsement or recommendation for use..                   i
Results from one study of inclination of drums in one trench
indicated that they varied from 0° to 32° from the vertical as
measured  from a  photograph  wjth a  level  and  plum
aluminum cube atop the drums. Void space between  the
drums was calculated to range from 17% to 38% of  the
trench volume. The condition of the exhumed  drums and
wastes varied  greatly: the inclination of drums varied from
0° to 90° from vertical and the  condition of the waste was
from good to highly degraded. The good condition of some
drums (paint  intact, drums  undeformed, and sacks  of
herbicide  intact) suggests that the  longevity  of  a waste
container  in a landfill  depends  on  its original condition,
handling,  contents, and the composition of  fluids that
surround it. During the excavation  of trench  24,  drums
buried 3 yr earlier, sitting in  an  unknown  orange-brown
liquid,  had intact paint (although  some  showed  surface
corrosion).

Conclusions

The  primary  reason that  the Wilsonville  industrial waste
disposal  site  failed  to  perform  as  predicted  was that  the
earth materials had higher  hydraulic conductivity  values
than the original tests indicated. The  original  predictions
were based on laboratory-determined values of hydraulic
conductivity, which  were confirmed  by this  study.  Field-
determined values  of  horizontal  hydraulic  conductivity,
however, were 10  to  1000 times greater than  laboratory-
determined values. In addition, the original investigation  did
not  recognize the importance  of vertical  joints  and
disconnected sand lenses. These joints  cause the vertical
hydraulic conductivity to be up to 10  times greater than  the
horizontal value. Joints and sand  lenses also presented
preferential pathways for downward movement of chemicals
at some locations in and around the site.

Rates of chemical migration may also have been enhanced
by differential settlement and reactions of organic chemicals
with  clay.  Highly erodible  earth materials allowed freely
draining depressions to develop in the trench covers, and
thus permitted surface  water to  enter the trenches, interact
with, the waste and increase groundwater gradients out of
the trenches.

The  highest levels of contamination  found at Nest B and
Profile W are  immediately downgradient of an area where
large quantities of liquid wastes were  buried. Interactions
between  these chemicals and  the  clay may  have opened
joints, and increased downward flow.  This mechanism does
not appear to be significant elsewhere on the site.

Although the  site  was not affected by the  acid mine
drainage or subsidence, the coal refuse  pile did create  a
groundwater mound that affected the local flow direction and
gradients in the shallow groundwater flow system.

This Research Brief and  the  articles by Follmer  (1984),
Griffin et al. (1983, 1984, 1985), Herzog et al. (1989), Herzog
and Morse (1986, and in press), Johnson et al. (1983), Stohr
(1983), and Stohr et al. (1985 and 1987) were submitted by
the Illinois  State Geological  Survey in partial fulfillment of
Cooperative Agreement CR810442 under the sponsorship of
CR810442 under the sponsorship of the U.S. Environmental
Protection  Agency. The  Principal  Investigator  was  Dr.

-------
Robert A. Griffin  of the  Illinois State Geological  Survey,
Champaign, IL 61820. Dr. Michael H. Roulier was the EPA
Project Officer

References
Cooper, H. H., J. S. Bredehoeft, and I.S. Papadopulos, 1967.
    Response of a Finite-Diameter Well to an Instantaneous
    Charge of Water. Water  Resources Research, 3:263-
    269.

Follmer, L. R. 1984. Soil-An  Uncertain Medium for Waste
    Disposal. In:  Proceedings of the  Seventh  Annual
    Madison  Waste Comference.  University of Wisconsin-
    Extension, Madison, Wl p. 296-311.

Griffin, R. A.,  K. Cartright, P. B. DuMontelle, L. R.  Follmer,
    C. J. Stohr, T. M. Johnson, M. M. Killey, R. E.  Hughes,
    B. L. Herzog, and W. J. Morse.  1983. Investigation of
    Clay  Soil  Behavior and Migration  of Industrial
    Chemicals at  Wilsonville, Illinois. In: Proceedings of the
    Ninth Annual  Research  Symposim  of the Solid  and
    Hazardous  Waste Research  Division,  U.S.
    Environmental Protection Agency, Cincinnati, OH. EPA-
    6009-83-018. p. 70-79.

Griffin, R. A.,  R. E. Hughes, L R. Follmer, C. J. Stohr, W. J.
    Morse, T.M.  Johnson, J.  K.  Bartz, J.  D. Steele,  K.
    Cartright, M.  M. Killey,  and  P.  B.  DuMontelle. 1984.
    Migration of  Industrial  Chemicals and Soil-Waste
    Interactions at Wilsonville, Illinois.  In: Proceedings of
    the Tenth Annual Research  Symposium  of the Solid
    and Hazardous Research  Division, U.S. Environmental
    Protection Agency,  Cincinnati, OH.  EPA-600/9-84-007.
    p. 61-77.

Griffin, R. A., B. L. Herzog, T. M. Johnson, W. J. Morse, R.
    E.  Hughes,  S. F. J. Chore,  and L. R.  Follmer, 1985.
    Mechanisms  of Contaminant Migration Through a Clay
    Barrier-Case Study,  Wilsonville, Illinois. In: Proceedings
    of the Eleventh  Annual  Research  Symposium of the
    Solid  and  Hazardous   Research Division,  U.S.
    Environmental Protection Agency, Cincinnati, OH. EPA-
    600/9-85-013. p. 27-38.

Herzog, B.  L.,  and  W.  J. Morse.  1986. Hydraulic
    Conductivity  at  a Hazardous Waste  Disposal  Site:
    Comparison  of Laboratory and  Field-Determined
    Values. Waste Management and Research, 4:177-187.

Herzog, B. L, and W. J. Morse. In  press.  Comparison of
    Slug Test Methodologies for  Determination of Hydraulic
    Conductivities  in  Fine-Grained  Sediments.  In:
    Proceedings  of the  ASTM Symposium  on Standards
    Development for Ground Water  and  Vadose Zone
    Investigations, January 24-28, 1988. American Society
    for Testing and Materials, Philadelphia, PA 19103.

Hvorslev, M. J. 1951.  Time  Lag and Soil Permeability in
    Ground-Water Observation.  Bulletin  36. Waterways
    Experiment Station,  U.S.  Army  Corps of Engineers,
    Vicksburg, MS. 50 p.
Johnson, T. M., R. A. Griffin, K. Cartwright, L. R. Fallmer, B.
    L. Herzog, W.J. Morse, P. B. DuMontelle, M. M. Killey,
    C. J. Stohr,  and R.  E.  Hughes.  1983,  Hydrologic
    Investigations of Failure Mechanisms  and  Migration of
    Organic  Chemicals  at  Wilsonville,  Illinois.  In:
    Proceedings  of  the  Third National  Symposium on
    Aquifer  Restoration  and  Ground-Water  Monitoring,
    National Water Well Association, Dublin, OH.  p.  413-
    420.

Nguyen, V.,  and G. F. Pinder. 1984. Direct Calculation of
    Aquifer  Parameters in  Slug  Test  Analysis. In
    Groundwater  Hydraulics  (J.  Rosensheim  and G. D.
    Bennet,  eds.). Monograph 9, American Geophysical
    Union Water Resources, p. 222-299.

Papadopulos, S. S.,  J. D. Bredehoeft, and  H.  H.  Cooper,
    1973. On the Analysis  of  Slug Test  Data. Water
    Resources Research, 9:1087-1089.

Sherard, J. L., L.  P.  Dunnigan,  R.  S. Decker, and E. F.
    Steele. 1976.  Pinhole test for  identifying dispersive
    soils. Journal of the  Geotechnical Engineering  Division,
    Am. Soc. of Civil Engineers, 102  (No.  GT1): 69-81.

Stohr,  C. J.  1983. Applications of Close-Range  Photo-
    grammetry  for  Geologic Investigations  During the
    Exhumation  of a Hazardous Waste  Disposal  Site. In:
    Technical Papers of the  49th  Annual Meeting of the
    American Society of  Photogrammetry, Washington,
    D.C., American Society of Photogrammetry.p. 267-273.

Stohr, C. J., W.  J. Su, P. B. DuMontelle,  and R. A. Griffin.
    1985. Photointerpretation  and Remote  Sensing
    Investigation  of  a  Hazardous  Waste Landfill. In:
    Proceedings  of  the 1985  Annual   Meeting  of the
    American Society of Photo-grammetry,  Washing-ton,
    D. C. p. 262-272.

Stohr, C. J., W.  J. Su, P. B. DuMontelle,  and R. A. Griffin.
    1987. Remote Sensing  Investigation of a Hazardous
    Waste Landfill. Photogrammetric  Engineering  and
    Remote Sensing, 53 (11):1555-1563.

Todd,  D. K.  1980. Groundwater Hydrology (2nd  Edition).
    John Wiley and Sons New York,  N.Y., Chap. 4.
 Bibliography
 Herzog, B. L. , R. A. Griffin, C. J. Stohr, L. R. Follmer, W. J.
     Morse and  W. J. Su. 1989.  Investigation  of  Failure
     Mechanisms and Migration  of  Organic  Chemicals at
     Wilsonville,  Illinois. Ground Water Monitoring Review, 9
     (2):82-89

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
     BULK RATE
POSTAGE & FEES PAID
         EPA
   PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/M-89/033

-------