/
United Stem
Environmental Protection
Agency
OH and Special Materia*
Control OMrion
WH-6M
MaeNnfton. O.C 20490
Wtm anj Wait* Management
Valley of the Drums
KfMOVCtOIT
NATIONAL TECHNICAL
INFORMATION SERVICE
«.«. hmimmi m eemmuu

-------
Efr-430/9-80-014
PB 80 22619 i
M TI T t ANW	• i
"Valley of the Drums" - Bullitt County, Kentucky
j	OA't
August 1980
Pi U*-OMV \( v	. I^A I	cOOl
52041001

U «'l i(-Or'Mi..w JhuA\	HLHOMT NO
N ' i « ^ A
li HiHrOHV.'.w wHv. ,
I Oil and Special Materials Control Division (Wll-548)
i 'Office of Water Program Operations
¦ Washington, D. C. 20450
t 3	* % « V t .» >«Cw
10 H,)w	LLlMLM NO
! B51B2D
• 1 C^NTmACT"" t NO
I Office of '..ater Program Operations
. U. S. Environmental Protection Agency
Washington, D. C. 20460
13	IVfJi lii.rjf u hu:
This document is a report outlininq the Vey actions of
the pe.i-?ral Dn-^renp Coordinator (OSC) who directed Federal
pollution control efforts at the cactioned event. The OSC
report is reaoirei hy ^c.ion 1510.56 of the? National Oil and
"lazardous ^uhstanc^s Polljtion Continqency Plan (Federal
Register, Vol. 45, No. 55 - Wednesday, March 19, 1930).
Of-Scene Coordinator report
L
Id	J ' ** J W * rfS » -» « ¦ i •
Release to the puolic
IJ i.e." • V C	t*t \ A., "	, J* w w
unclassified	' _
.
£p v r® • 3-; *-??
L • 5*. » C	* c

-------

»1
1
«
&
$5
%

-------
EPA-430/9-80-014
August 1980
VALLEY OF THE DRUMS
BULLITT COUNTY, KENTUCKY
U.S. ENVIRONMENTAL PROTECTION ACENCY
OIL AND SPECIAL MATERIALS
CONTROL DIVISION
WASHINGTON, D.C. 20460

-------
'.-DEFACE
This report reflects the activities of the Errvirormental
Dnergency Branch (EEB) Region IV, EPA, as it was concerned with
the captioned event. The report outlines the key actions of the
Federal Qt-Scene Coordinator, the Regional Response Team (RPT),
the state, local, ind Federal mmbers of the RRT. It reflects
the complexity of responding to environmental enargencies and the
need to ooordirvte and plan in advance for a majcr incident such
as this event.
This report is intended to satisfy the requirements of the
National Oil and Hazardous Substances ftjllution Contingency Plan
and to help others learn from our experiences.
Shnith, Chief
Environmental Qnergency Branch
1

-------
TABLE OF CONTENTS
SECTION	PAGE
1.	SUMMARY	1-1
2.	THE SITUATION	2-1
?. PHOTOGRAPHS	3-1
4. CONTAINMENT AND	TREATMENT 4-1
11

-------
LIST OF ILLUSTRATIONS
FlGUKE	PAGE
4-1	TREATMENT SYSTEM IN PLAN	4-6
4-2	TREATMENT SYSTEM PROFILE	4-7
4-3	WATER and SEDIMENT SAMPLING LOCATIONS	4-9
4-4	WATER QUALITY STATIONS	4-10
TAgLE
4-1	TEST RESULTS OF HATER and SEDIMENT SAMPLING	4-12
in

-------
1. SUMMARY
At 2:00 p.m., on March 2, 1979, Robert Sholar of Kentucky's Divi-
sion of Hazardous Materials and Waste Management contacted the Envi-
ronmental Emergency Branch of EPA Region IV. Mr. Sholar reported that
oil and other hazardous materials were spilling into Wilson Creek as a
result of rain runoff and leaching.
Under authority of Section 311 of the Clean Water Act (Federal
Water Pollution Control Act, amended 1972), Jack Stonebraker responded
as the federal On-Scene Coordinator and departed for Shepherdsvi1le,
Kentucky. Assistance was requested from the Coast Guard Gulf Strike
Team, the EPA Emergency Response Team (ERT), and the Coast Guard
Public Information Assistance lean. OH Materials of Findlay, Ohio,
was contracted for work required to control pollution occurring at the
dump site.
A survey of the dump site was conducted on March 3, 1979, and
construction began on an underflow dam approximately one-half mile
downstream from the source of the pollution. Lawyers for the property
owner, Mrs. A. L. Taylor, were contacted and permission was obtained
to continue work on the site in an effort to control the runoff and
eliminate the pollution of Wilson Creek. Men and equipment began
moving and relocating druns in order to clear work areas a.id provide
access to land at the base of the hill on which the dump is situated.
Wate*- samples were taken and a news release was issued to the media
late in the evening.
1-1

-------
The following day, March 4, a sufficient number of drums and
debris had been relocated, allowing equipment to begin grading a shal-
low ditch along the base of the hill for control and collection of
runoff. Heavy rain the night before caused stream waters to rise and
it was necessary to enlarge the underflow dam.
Operations resulted on March 5. Crews continued to relocate drums
and segregate full and empty druns. Empty drums not suitable for pos-
sible reuse were crushed. Heavy equipment was used to dig a trench at
the base of the dump site. A catch pond was constructed at the lowest
point on the property to collect runoff and seepage from the diversion
ditches and trench. A limited Regional Response Team meeting was held
with representatives from the state. Alternatives for dealing with
the problems presented by the dunp site were discussed along with the
state's long-term plans for the "Valley of the Drums." Preliminary
sample analysis was performed by Environmental Consultants, Inc., in
Clarksville, Indiana. Results showed the presence of oil, benzenes,
toluene, and ketones in Wilson Creek. Based on this information, it
was decided to begin aeration of stream water with compressed air to
strip volatile organic chemicals. Aeration was begun in the stream
near the dunp and farther downstream, irrriediately behind the earthen
dam constructed on March 3.
On March 6, a third aeration system was established in the field
adjacent to the dam. Water was pimped from the pool behind (upstream)
the dam through spray nozzles on irrigation pipes and into the air.
The natural drainage of the field carried this aerated water back into
Wilson Creek, upstream of the dam.
The next day, March 7, it was decided to establish a filtration
system for treating runoff and seepage collected in the catch pond.
The system would pump this runoff through bins or cells. The first
cell would contain limestone, the second, peat moss, and the third,
activated carbon. It was felt that this treatment, in addition to
aeration, would be necessary before allowing the collected runoff to
enter Wilson Creek. Procurement of the materials required for the
treatment system began and drin relocation and segregation continued.
1-2

-------
The filtration system was not considered operational until March
12 because of the materials and logistics involved with its establish-
ment. The second cell containing peat moss did not work. The peat
moss had the same consistency as soil and rose to the toD. T.-v: peat
moss was removed and replaced with activated carbon.
Sampling for a biological and chemical survey began on March 8.
ERT members and state personnel collected samples that were sent for
analysis to EPA's laboratory in Athens, Georgia. A1 Smith, Chief of
EPA Region IV Environmental Emergency Branch, met with Eugene Mooney,
Secretary of the Kentucky Department of Natural Resources and Environ-
mental Protection (KDNREP), and 20 other representatives of the state
to discuss EPA plans and to establish a program of mutual effort and
cooperation.
On March 9-11, operations consisted primarily of filtration sys-
tem construction and drim relocation and segregation. Samples were
collected from the catchment basin and from effluent of the modified
treatment system. Samples were analyzed by Environmental Consultants,
Inc., on March i"?, 1979. The preliminary results indicated a 99%
reduction of organic chemicals through the treatment system. Only a
slight trace of toluene (1-5 ppb) was found in the effluent sample.
EPA and KDNREP tie Id a joint press conference on-site. The media
toured the work site and observed the treatment system in operation.
Approximately 20,000 drims of hazardous substances were uprighted
and placed in rows. Each drim was marked to indicate whether it was
empty or contained solids, sludge, or liquids. Personnel from KDNREP
conducted an inventory of the remaining druns. Tne empty drims that
could be reused were stacked while the damaged ones were crushed.
1-3

-------
2. THE SITUATION
During a routine inspection of the Taylor site, coimionly referred
to as the "Valley of the Druns," Robert Sholar, Kentucky Department of
Natural Resources and Enviromental Protection (KDNREP), Division of
Hazardous Materials and Waste Management, noticed a discharge of oil
and hazardous substances into Wilson Creek. The recent rains had
flushed the material into the creek and it was flowing downstream.
EPA was already aware of the "Valley of the Druns" and the poten-
tial for a discharge of oil into the creek via surface runoff. In
January 1979, several on-site visits revealed a large quantity of
druns (between 20,000 and 100,000) on the surface of the 23-acre site.
An unknown quantity of druns ir.ay also have been buried at this site.
At the time of the visits, however, the ground was frozen and covered
with snow. There wasn't any apparent discharge of oil or chemicals
into the surface waters.
RESPONSE AND OBSERVATIONS: CHRONOLOGICAL EVENTS
March 2, 1979
Mr. Sholar reported his findings to the Environmental Emergency
Branch (EEB) of EPA Region IV, and indicated that the material enter-
ing Wilson Creek was multicolored. The spilled material already
present on the ground would continue to enter the creek unless some
action wos taken.
2-1

-------
The OSC asked Mr. Sholar if oil was being discharged into Wilson
Creek. His answer was affirmative and he indicated that a laige
nunber of druns containing waste oil had been transported from Ford
Motor Company to this site.
EPA Region IV EEB Chief, A1 Smith, and the OSC notified John
White, Regional Administrator for EPA Region IV, of the incident. Mr.
White said he would inform Eugene Mooney, Secretary of KDNREP, that
EPA was responding to the spill under Section 311 of the Federal Water
Pollution Control Act.
The following teams were activated to assist the Qn-Scene Coordi-
nator (OSC) Jack Stonebraker:
EPA's Emergency Response Team (ERT)
National Strike Force - USCG bulf Strike Team (GST)
USCG Public Information Assistance Team (PIAT)
OH Materials, Inc.
The OSC arrived on-scene with personnel from OH Materials, Inc.,
and met with Nr. Sholar. There were three definite areas where oil
and chemical discharges were entering Wilson Creek.
A sheen of oil on the water surface was observed for about one-
fourth mile. The banks were covered with a gray residue that had been
there for some time. The sediment in the stream was covered with the
same type of material. When the sediment was disturbed, a strong odor
of solvent-type material was emitted.
It was clear that Section 311 of the Federal Water Pollution
Control Act was being violated. The problem was to contain the sur-
face runoff from this site. P'ans were to start work on March 3,
1979, when the equipment would arrive on-scene.
March 3, 1979
Bob Sholar notified Mrs. A.L. Taylor, owner of the property, that
oil and other chemical compounds were being discharged into Wilson
Creek. EPA had responded to this incident and needed permission to
get on her property and contain the discharge. It was believed that
Mr. A. L. Taylor's will had not been probated. Also, KDNREP had
2-2

-------
issued a complaint against Mrs. Taylor to remove the druns of indus-
trial waste. Mrs. Taylor informed Bob Sholar that she would not give
permission until her attorney had been advised.
At 1:00 p.m., Mrs. Taylor met the response group at the site.
Credentials were presented to her and she unlocked tl.e gate to the
property. She stipulated that no one from the media b*» allowed in,
anJ that the gate be locked every night. Mrs. Taylor was told that
her request would be honored.
The first action was to construct an underflow dam about one-half
mile downstream from the site. This would contain the floatable
materials, but not the soluble chemical compounds. Samples were
needed to determine what type of chemicals were present. It was evi-
dent that volatile organics were in the water colunn. These organics
could be removed by aeration.
Plans were made to put in ring aerators at two locations, one
about one-fourth mile downstream from the site and the other at the
dam.
Work began on constructing a catchment basin at a low point near
Wilson Creek where the discharge of oil and hazardous substances was
entering the stream. Trenches would also be constructed so the sur-
face runoff could be collected in the catchment basin.
Druns of industrial waste were scattered all over the property,
impeding the flow of the surface runoff. Some of these druis were
leaking, others were empty. In order to get to the leaking druns, the
empty druns had to be moved and arranged in rows.
Data was needed on what was being discharged into the stream.
Bob Sholar said that Envlromental Consultants, Inc., in Clarksville,
Indiana, had done some work for the state. He contacted Robert Fuchs
of that firm. Mr. Fuchs agreed to assist in collecting samples and
promised a quick turn-around time for qualitative analyses.
March 4, 1979
The rain that fe'l during the night caused the waters of Wilson
Creek to rise considerably. The water was flowing around the end of
the underflow dam. The design of the dam was changed to allow water
2-3

-------
to flow through pipes near the top. This drop in elevation would
provide additional aeration and increase the stripping effect of the
volatile o-gamc compounds.
Work continued on the construction of trencher and catchment
basins. The druns were arranged in rows and marked as solid, liquid,
sludge, or empty. The empty aruns that could not be reused were
crushed and piled 1.. one area.
The qualitative* analyses of the samples collected by Bob Sholar
on March 2, 1979, and analyzed by Environmental Consultants, Inc.,
were received. Ten compounds were identified; xylene, toluene, ben-
zene, and ketones were the primary compcunos. These could be removed
by aeration and sorbent materials. Additional samples were collected
this date.
Members of the Regional Response Team (RRT) were notified that a
meeting would be held at the Conroand Post in Shepherosv11le, Kentucky,
at 10:00 a.m. on March 5, 1979. It would be a limited activation of
the RRT. Representatives from KUNREP, the Department of the Interior
(D01], and the U.S. Coast Guard were asked to attend.
March 5, 1979
At 10:00 a.m., the RRT met to discuss the situation created by
the discharge of oil and hazardous substances into Wilson Creek. The
following members attended:
Jack Wilson, KUNREP
Clyde Baldwin, KONREP
Captain Ralph Barte11, USCG
Waynon Johnson, DOl
Jack Stonebraker, EPA OSC
John Gilber* EPA ERT
Mike Donohoe, USCG Gulf Strike Team
The RRT suggestca that a chemical and biological study on the
affected drainage basin be conducted. Personnel from EPA, KDNREP, and
DOI would plan and conduct the survey.
Z-4

-------
Jack Wilson advised the RRT that rearranging the druns at the
site nay have jeopardized the state's legal action against the prop-
erty owner and its effort to encourage the generators to remove the
druns. When questioned about the state's case against the gener-
ators, M*. Wilson indicated that, legally, rt was questionable. How-
ever, tne state felt that the generators of the industrial waste had a
"moral" obligation to remove the druns from the site.
To control the source of pollution, it was necessary to move full
and empty druns in order to get to the leaking druns. All that that
involved was setting the druns upright in the same general area. It
was difficult to understand how this action would jeopardize the
state's complaint against Hrs. Taylor or the generators.
It was finally agreed to continue chis phase of the operation.
The RRT would convene at a later date.
The in-stream treatment devices would be utilized until the
catchment basin was in place. In addition to the ring aerator placed
at the dam site, water would be pumped through a header system whereby
the volatile organics could be removed by air stripping.
Gene Coker and Fred Wolfe of EPA arrived on-scene to assist in
the collection of core sampling. This operation had already been
planned. The hydrogeological survey would begin on the next day.
March 6, 1979
Work continued on construction of trenches and the catchmer.t
basin. In consultation with Gene Coker of EPA, it was decided to cut
the lower trench to a depth of five feet. This would intercept the
lateral underground migration of oi> and chemical compounds.
An area about two acres in size was discovered where the trees
and other vegetation had been destroyed by surface runoff from the
Taylor site. The contaminated soil in this area was 2-3 feet deep.
Another trench was constructed to intercept the runoff. The trench
was designed to direct the runoff into the catchment basin.
During the process of clearing an area just above the catchment
basin, a partially exposed drun was noticed.1 A small area around this
drun was uncovered and a pit or trench containing several buried druns
was discovered.
2-5

-------
The druns were in poor condition and leaked a black oily sub-
stance. The pit remained open for several days and filled with the
black substance.
An additional pit was discovered on the north side of the road
near two power line poles. Apparently, while Mr. Taylor was operating
at this site, he constructed several pits, dunped the contents of the
drums into them, and covered the pits with soil. There were probably
several other underground pits on this site. Where they were located
was unknown.
Mr. Fuchs of Environmental Consultants, Inc., reported some
preliminary data on the samples collected from the leach<».te that was
entering Wilson Creek. The results were as follows:
SUBSTANCE	CONCENTRATION
Oil
252.0
ppm
Methyl Ethyl Ketone
5.0
ppm
2 - butanol
21.0
pp
-------
ai-ed, it was decided to extend the trench from the catchment basin up
to the Mock house, work began on the trench this date. With the
completion of toe trench, it was possible to collect the surface
runoff in the catchment basin.
Now that the collecticn system was about to be completed, a
decision had to be made on how to treat the wastewater in the catch-
ment basin. A1 Smith arrived at the site to assist the OSC in the
design of a temporary wastewater treatment facility. Considering the
type of chemicals being discharged into Wilson Creek, Mr. Smith sug-
gested a three-cell treatment facility. The first cell would contain
crushed limestone, the second, peat moss, and the third, activated
carboi.. To remove the volatile organic compounos, the first cell
could be aerated.
Biologists from EPA, DOI, and KDNREP met to discuss plans for the
collection of biological and chemical samples. They decided to col-
lect six water samples and five sediment samples from Wilson Creek,
the catchment basin, and rn open pit located on the site. In addi-
tion, a static bioassay would be conducted on water collected from
certain stations.
A1 *j»iith was instructed to attend a meeting of Secretary Mooney
and staff in Frankfort, Kentucky. The purpose of the meeting was to
discuss plans for a solution to the problems of disposal and contain-
ment of surface runoff.
Crews continued to segregate and organize the druns.
harch 8, 1979
It was decided to use twq dempsey dtmpsters as containers for the
limestone and activated carbon. They would arrive at the site on
March 9, 1979. The contractor began ordering PVC pipe and other
materials to be wid in the treatment system. Samples were collected
for the biological/chemical survey analyses to be conducted in the EPA
laboratory in Athens, Georgia.
Mr. Smith returned from the meeting in Frankfort, Kentucky, with
the following information:
2-7
-------
Jack Wilson would be the coordinator for state actions with the
drums.
?tate personnel agreed with the drun organization and segregation
and wanted to assist in operations.
The state planned to initiate an inventory of the drums on March
1?, 1979.
Mr. Mooney requested a joint press conference between EPA and
KONREP to discuss treatment systems and future plans for this
site.
M^rch 9-11, 1979
The dunpsters were set on a gravel base. Piping was arranged to
provide upflow through the first cell of crushed limestone and second
cell of activated carbon. (The use of pe-.it moss in the second cell
had proved unsuccessful because it kept flowing over the top. Acti-
vated carbon was substituted to alleviate the problem.) Compressed
air would be pumped through the ring in the bottom of the first cell
for aeration. A weir was constructed in the final cell for distribu-
tion over the activated carbon (downflow).
An electrical submergible pump was installed to pump water from
the ped to the first cell. The pump would be activated when water
in the catchment basin reached a certain level. The water for the
aeration system would also be activated when the pimp was operating.
A bridge or catwalk was constructed from the edge of the catch-
ment basin to the center. This would enable the operator to raise or
lower the pump and to provide the necessary maintenance.
Ben Eason of the P.AT made arrangements with Ed Shelley of KDrfREP
for a joint news conference on March 13, 1979. The conference would
be held at the work site.
As of this date, 8,641 druns had been segregated and organized in
rows. This amounted to one-third of the work area, or two acres.
2-8
-------
March 12, 1979
Work continued on the construction of the filtration plant. The
primary problem encountered was the leaking of the containers. It was
necessary to weld all of the seams in the containers.
Tt-e druns were organized at the rate of about 900 per day.
KDNREP said that the inventory of druns would be delayed until the
druns could all be arranged.
Plans were made for a final RRT meeting to be held in Shepherds-
vilie, Kentucky, on March 14, 1979.
Operation of the filtration plant began at 5:00 o.m. The flow
rate through the system was varied from 60 gallons per minute to 300
gallons per minute The best flow rate was 60 gallons per minute, or
tt,64G gallons in a 24-hour period. Samples were collected from the
catchment basin ar.d the effluent. The samples were analyzed by
Environmental Consultants, Inc., to determine the efficiency of the
treatment system.
March 13. 1979
KDNREP obtained permission from Mrs. Taylor and her attorney for
representatives from the news media to tour the work site.
The preliminary analysis of the samples of influent and effluent
of the filtration p'ant indicated a 99% reduction of orgaric (onpounds
through the system. Only tcluene, at 3 ppb, was found in the effluent
sample.
The press conference was held on-scene. Jack Wilson, Deputy Com-
missioner for KDNREP and Jack Stonebraker, EPA 0SC, were the spokesmen
for their respective agencies. Afterward, the news media toured the
work site, took photographs, and interviewed the 0SC and Mr. Wilson.
The media repr»sentatlves were impressed with the accomplishments of
the work crews in controlling the runoff and installing the treatment
system.
A1 Smith notified the 0SC that polychlorinated biphenyls (PCBs),
Aroclor 1254 and Aroclor 1260, had been found in samples collected by
EPA on February 14, 1979. These were preliminary results and not for
publication. It was believed that PCBs were in the sediment samples
and not in the ambient water samples.
2-9
-------
An estimated 1,000 drums were organized in rows and marked as
solid, liquid, sludge, or empty on this day.
March 14, 1979
The RRT meeting was held at 10:00 a.m. The following topics were
discussed:
Mr. Wilson stated that Bob Sholar would monitor the site and
operate the treatment system. However, the state did not have
funds to replace any equipment or to provide any maintenance to
the treatment system.
EPA agreed to maintain the treatment system for an undetermined
period of time (at least through spring, because of the heavy
rains that generally fall djring this time).
Mr. Sholar would collect sanples of influent and effluent after a
heavy rain. It had not railed since the collection system had
been constructed.
State personnel would continue the inventory of drins. EPA would
continue to upright druns "'or source control.
About 5190,000 from the contingency fund had been spent. The OSC
requested that another 5100,000 be authorized under the project
nimber.
March 15-21, 1979
Work continued on segregating and organizing drinrs. This work
was completed on March 21, 1979. A total of 20,000 drums had	been
marked solid, liquid, sludge, or empty. The empty druns that	were not
usable were crushed.
It was necessary to replace the wood barrier between the	two
activated carbon cells with a metal barrier in the filtration	plant.
Otherwise, the treatment system was working well.
Approximately 5290,000 was expended from the contingency	fund to
control and provide treatment to the surface runoff from this	site.
2-10
-------
3. PHOTOGRAPHS
3-1
-------
AERIAL VIEW OF VALLEY OF THE DRUMS NEAR
LOUISVILLE, KENTUCKY
3-2
-------
DRUMS CONTAINING SOLVENTS. PAINTS, AND ACID
-------
DRUMS PRIOR TO CLEANUP ACTIVITIES
-------
ARRANGEMENT OF DRUMS IN ROWS
-------
ANOTHER VIEW OF DRUMS PRIOR TO CLEANUP
-------
»>V, -/
" «£».r
1 . ¦
"w," * ££'
r '	1
DRAINAGE FROM VALLEY AFTER RAIN
-------
RUNOFF FROM VALLEY INTO WILSON CREEK
-------

CONSTRUCTION OF INTERCEPTOR TRENCH
FOR COLLC CTING RUNOFF
-------
.*¦1
HOLDING POND FOR CONTAMINATED RUNOFF
-------
FLOATABLE MATERIALS ON HOLDING POND
-------
CONSTRUCTION OF WASTE TREATMENT FACILITY
-------
ANOTHER VIEW OF WASTE TREATMENT FACILITY
-------

; ' ." ' \
:Wrv-v;/ *•
'• r 4 •	,
CONTAMINATED WATER BEING PUMPED FROM HOLDING
POND TO FIRST CELL CONTAINING CRUSHED
LIMESTONE FOR AERATION
-------
LEACHATE AND RUNOFF KILLED VEGETATION IN
THIS AREA
3-15
-------
IN-STREAM AERATION OF WILSON CREEK
-------
AERATION OF WILSON CREEK TWO MILES
DOWNSTREAM FROM VALLEY
-------
AEP'AL VIEW Of VALLEY OF TNE DRUMS NEAR
COMPLETION OF CLEANUP ACTIVITIES
-------
4. CONTAINMENT AND TREATMENT
Introduction
Several industries located in the metropolitan area of Louis-
ville, Kentucky, are involved in the production and/or manufacturing
of automobiles, plastics, chemicals, paints, varnish, and organic
solvent materials. They generate large quantities of industrial waste
that must, or should be, disposed of in an approved manner. In past
years, some of these industries transferred their waste products into
drims and paid other companies or individuals to transport and dispose
of the drums. During that period, the generators did not question the
procedures utilized by the waste haulers in the disposal of these
potentially hazardous substances.
On a 23-acre site in Bullitt County, Kentucky, nestled between a
private golf course and a county park, more than 27,000 drians of m-
dust-ial waste «ere found in 1978 by state and EPA officials. The
collection and transport of the drums to this particular site began in
1967 and continued until the property owner's death in 1977. Due to
limited space, the owner had constructeo pits and trenches, filled
them witK industria1 waste, and sold or crushed the empty drums.
Later, the trenches were covered with soil from a nearby hill. Some
estimates indicate that more than 100,000 druns were stored, buried,
or dumped on this site. Eventually, weather conditions deteriorated
the stored drims and the contents were spilled onto the ground, con-
taminating Wilson Creek via surface runoff and leachaLe from subsur-
face drainage. The adjacent property owners reported that, at times,
4-1
-------
the waste materials on the water surface were multicolored. They also
complained that the smell was unbearable during warm weather. There
were also occasions when the spilled material ignited and created a
fire on the stream and adjacent shorelines. The property owners also
expressed concern that their w;ter wells may have become contaminated
by the spilled industrial waste.
On March 2, 1979, when the snow began to melt, the surface runoff
from the dunp site, now commonly referred to as the "Valley of"the
Drums," once again contaminated adjacent Wilson Creek. Personnel for
the Kentucky Department of Natural Resources and Environmental Protec-
tion j(KDNREP) advised the U.S. Environmsntal Protection Agency (EPA)
of the incident and requested assistance. Since oil and mixed chem-
icals were discha-ged into surface waters, the federal Contingency
Fundi PL: 92-500 Section 311(K), could be used to relieve the environ-
mental emergency. Therefore, the Environmental Emergency Branch (EEB)
of EPA Region IV was authorized to use the fund to control the runoff
and eliminate further pollution of the stream.
;Extensive sampling for both chemical and biological data was con-
ducted in the drun storage area and the Wilson Creek drainage basin.
The sample analyses indicated the presence of several organic com-
I
pounds such as benzene, toluc le, ketones, xylene, styrene, phthalates,
t
and polycli lor mated biphenyls (PCBs) in runoff and sediment samples.
I
Low concentrations were also found in the benthic organisms. Alto-
gether, there were 142 chemical compounds identified in the various
samples collected in the "Valley of the Drims."
In order to relieve the emergency situation, the following tem-
porary devices were constructed on-site:
Trenches to intercept the surface runoff and the underground
lateral migration of pollutants through the soil.
A catchment basin to receive the runoff via the trenches.
A field-constructed activated carbon unit to treat contaminated
water in the catchment basin.
In addition, druns were segregated and marked according to their
contents, and organized in rows for future handling.
4-2
-------
Although the problem of disposal still remains at this site for a
large quantity of hazardous substances, a technically sound and econ-
omically feasible methodology for treating and controlling contami-
nated runoff was instituted.
In-Stream Treatment
Wilson Creek flows in an easterly direction to its confluence
with Southern Ditch Creek. Subsequently, these waters flow into Pond
Creek and the Ohio River near West Point, Kentucky. The "Valley of
the Druns" lies in the headwaters of Wilson Creek where the normal
flow is low and subject to fluctuations from seasonal water contribu-
tions. The stream bottom in these upper reaches consists mainly of
sand and clay deposits. Years of contaminated runoff entering the
stream covered the sediment with a gray residue. When the sediment
was disturbed, a strong odor of solvent-type material was emitted.
Floatable materials such as oil, xylene, and toluene were present on
the water surface at the time. This condition persisted downstream
for approximately 0.5 miles (0.9 kilometers). The initial response to
this situation was to impede the flow of the contaminated waters of
Wilson Creek. This was done by the construction of an underflow dam
across Wilson Creek, downstream from the "Valley of the Oruns.M The
underflow dam would contain the floatable material, but not the sol-
uble organic compounds that were present in the water colunn. To
achieve the removal of these organic compounds, ring aerators were
placed in the stream at two locations: one immediately downstream
from the "Valley" the other just upstream from the underflow dam. In
addition, water was pumped from behind the dam throi'qh a header pipe
and sprayed onto the adjacent land. Thus, aeration of the stream
water released the volatile organics. The stream water was polished
as it drained overland back into the creek above the dam.
Control and Collection of Surface Runoff
Drums of industrial wastes were scattered all over the 10-acre
"Valley" site. In some areas, the drums were stacked 8-10 feet high.
4-3
-------
They were badly weathered, some leaking, others empty. The soil was
contaminated tn some places to a depth of 10-12 inches. There were
I
two small impoundments on the property. One contained ethylene
glycol, the other about eight feet of sludge and waste material.
During periods of rainfall or thaw, the ground became saturated and
hazardous substances were flushed into Wilson Creak. On several occa-
sions, especially after a heavy rainfall, these impoundments over-
flowed and drained into Wilson Creek.
Another serious problem lay underground where, for years, the
"operator had dimped industrial waste into pits or trenches. Once the
pits were filled, they were covered with soil from a nearby hillside.
Three large underground trenches were discovered 011 the "Valley" site.
One trench, located about 50 yards upgrade from Wilson Creek, measured
10' x 150* x 8'. Other trenches, as yet undiscovered, may also
exist. Thus, hazardous substances were not only entering Wilson Creek
via surface runoff, but also by means of a lateral migration through
the soil.
To control and collect both surface and subsurface drainage of
contaminated waters, interceptor trenches were constructed at the base
of the, slope on the east side of the "Valley." In certain areas
adjacent to the creek, the trenches were excavated to a depth of six
feet. This was necesssary to ensure the interception of tne subsur-
face drainage,
A catchment basin was also excavated to collect the runoff from
the "Valley" site. The basin configuration was to accommodate a flow
of 21 cfs for a retention time of 15 minutes. This design was based
on a drainage area of 10 acres, a five-year storm, a runoff coeffi-
cient of 0.35, and the "Rational" form q * ACi. At the time of the
construction, it was estimated tnat the basin would last approximately
three months. After nearly one year of service, the basin was still
adequate and no overflow had been recorded.
Carbon Absorption Treatment System
In devising a system capable of treating the contaminated runoff
held in the catchment basin, certain criteria had to be considered.
4-4
-------
For example, the treatment system would have to be practical, inex-
pensive, easy to operate, and require a minimun of maintenance. In
addition, the system must be temporary in nature, not a permanent
facility, and capable of reducing the concentrations of the chemical
constituents to an acceptable level.
Activated carbon is specific for nonpolar molecules, which makes
it useful in the removal of organic compounds. On nunerous occasions,
carbon colunns have been utilized in treating spills of hazardous sub-
stances in the field. EPA was successful in removing PCBs and other
organic compounds with a modified activated carbon treatment system in
Whitehouse, Florida. Having examined the chemical data of the samples
collected in February 1979, it was decided to construct a carbon
absorption unit preceded by aeration and pH adjustment chamber.
Figures 4-1 and 4-2 depict the design and flow pattern of the
treatment system. Two trash dumpsters (20' x 10" x 5') were made
watertight by welding seams. They were placed on a bed of crushed
rock. The first dumpster contained 8.U00 pounds of crushed limestone.
The iimestone was used to elevate the pH to about 8. By doing this,
some of the inorganic compounds could be precipitated. Compressed air
was also punped through a ring aerator, located at the bottom of the
first unit, to remove the volatile organic compounds from the water
colnnn. The other dunpster was divided into two cells, both contain-
ing activated carbon. The piping was arranged so the contaminated
water would flow up through the carbon in the second cell and be dis-
tributed over the carbon through a "V-notch weir in t^e third and
final c®'1
The treatment system used a submersible pump to transfer the con-
taminated water from the catchment basin to the first dumpster. The
pimp was activated when the water in the basin reached a predetermined
level. From the first through the third cell, the flow was by grav-
ity. The effluent was discharged into Wilson Creek. The most effi-
cient flow rate through the system was 60 gallons per minute.
The chemical analyses of the samples collected from the catchment
basin and the effluent indicated a 99% reduction of organic compounds
through the treatment system. Only toluene (3.0 mg/1) was detected in
the effluent.
4-5
-------
CATCHMENT BASIN
15 Ml [JUTE
RETENTION
HUNOFF
SINGLE CELL UNIT
(8000 lbs. CRUSHED LIMESTONE)
WITH A£HATION.
TWO CELL UNIT ACTIVATED
CABBON
lrtEATMENT SYSTEM IN PLAN
tmiovio avi
AJS
06C
3/6/79
tATI
VALLEY 0} THE DRUMS
CMUWIM MUMJ
HEPARD5VILLEf KENTUCKY
-------
20'XIO'X5' STEEL
TRASH DUMPSTERS
PERFORATED PIPE
BOOO lb. CRUSHED
LIMESTONE
ACTIVATED
CARPON
catchme:jt
BASIN
TREATMENT SYSTEM PROFILE
VALLEY OF THE DRUMS
UWIN9 NUHW
Figure 4—2
SHEPARDSVILLE, KENTUCKY
-------
One of the problems encountered with this system was the silting
of carbon in the final two chambers. This *as especially a problem
after a heavy rain trfien the surface runoff contained a lot of soil.
However, this problem was solved by spraying the basin with polyelec-
trolyte, which settled the soil to the bottom.
Maintenance for the treatment facility has been minimal. How-
ever, after a period of ten months, it was necessary to replace the
motor for the aerator and the activated carbons in both units.
Chemical Analyses of Water and Sediment Samples
During t:ie entire episode, water and sediment samples were col-
lected from Wilson Creek and the treatment system. This was done by
EPA Reg*"on IV personnel and a private contactor hired by EPA. The
sampling at sites downstream from the "Valley" was part of a program
to determine the effect of various hazardous waste disposal sites in
and around the Louisville area. Ouring the latter part of February
1979, specific sampling was performed on the "Valley" site to
determine the types and quantities of hazardous substances present
(Figure 4-3). At this time, water samples were collected from stand-
ing water and impoundments that were discharged into Wilson Creek. To
determine the extent of contamination in the surface soils, specific
site samples were collected. None of the drums was sampled because it
would be cost-prohibitive and time-consuming for the emeryency phase
of the exercise.
Samples were transported to EPA's laboratory in Athens, Georgia,
and analyzed according to acceptable analytical procedures.
Analysis revealed the presence of a vast array of inorganic and
organic substances in the "Valley" proper and adjacent Wilson Creek.
In treating the waste, the primary concern was the organic conmounds.
Seven water and sediment samples were collected in the drum storage
area. Station AT5, a large pool of water in a small depression, had a
large assortment of chemical compounds. Station ATI and AT2 had the
least number of chemical compounds present (Figure 4-4).
A large number of solvents, ketones, phthalates, benzene, phen-
ols, and other organic compounds were found in lh? settling basin,
Station AT7.
4-8
-------
OQ
0° O
O Oo
o o
iC
DRUMS
o
oo
op;
CO
O O
AT-3
AT-5
AT-U
[AT-2
DRAINAGE
AT-6
WATER AMD SEDIMENT SAMPLING LOCATIONS
APPAOVCO »V
DRAWN BY
2/22/79
RDS
OtTI
VALLEY OF THE DRUMS
. AAWING Nuxaill
SIIEPARDSVILLE, KENTUCKY
Figure 4—3
-------
X-
o.
WILSON CREEK
o
	WC-U
WC-3
VALLEY
OF
DRUMS
NATIONAL
TURNPIKE
STITE'S STATION RD
SHEP11ER3»5VILLE. KENTUCKY
WATER QUALITY STATIONS
AfPROVlU OTl
OOAWN «V
O PDS
OA.
HAZARDOUS SITES INVESTIGATIONS
PAaWIHO NUMftM
LOUISVILLE, KENTUCKY
Figure 4—4
-------
Station AT4 had the largest nunber of compounds; analysis of sed-
iment samples identified 28. PCBs (Aroclor 1254 and 1260} were among
the compounds identified from this particular station.
In general, the water samples contained more organic compounds
than did the sediment samples. A total of 126 organic compounds were
identified in the surface runoff. The results of the sampling are
reported in Table 4-1.
Four stations along Wilson Creek were sampled (Figure 4-4) and
analyzed by EPA and a private laboratory, benerally, the same organ-
_ics detected in highest concentrations in water from the drun area
were also present in Wilson Creek samples. Compounds present in the
highest concentrations in these samples were: ethyl benzene, toluene,
ethoxypropanol, acetone, and methyl ethyl ketone. Station WC2,
located approximately 170 yards downstream from the site, had the
highest concentration of compounds detected.
A nunber of organic compounds were detected in the stream sedi-
ments at relatively low concentiations (less than 5 mg/kg). The
largest nunber was detected at Station WC4, located approximately 600
yards downstream from the "Valley" site. At this point, the stream
had a more defined streambed, which facilitated greater sedimentation.
Low concentrations (0.01-3.0 mg/kg) of PCBs (Aroclor 1254 and Aroclor
1^60) were detected in all of the Wilson Creek sediment samples. How-^
ever, no PCBs were detected in the water samples.
On-Site Housekeeping and Source Control
Prior to closing the emergency phase of the exercise, druns were
segregated and marked according to their contents, i.e., liquid,
solid, sludge, or empty. The empty drums that could not be reused
were crushed and piled in various locations on the property. The
druns were also arranged in rows to facilitate future handling and
disposal of the waste.
Cost
This five-week intensive effort cost the Government S300.000.
The treatment system, including one year maintenance, power, and car-
bon recharge, accounted for $45,000 of the amount.
4-11
-------
Table 4-1 TEST RESULTS OF WATER AND SEDIMENT SAMPLING
cohivund uk/i
W'.t, r
-nl.
AT-}
Wnii r
a r-

AT-5
AT-/.
AT-7
Wtiii r
.• J.
Wdtfl frl.
Watrr S<*<1.
Water Zed.


1-*
..,0
10-S0
i0-100

1-25
1-10
10-50
1-10

l-2i
1-10
1-10
1-10


1-10
1-10

__3L-._. !-•?	 _
... __ 	
20 1
it
1 ?!•
11
3 JO
13
1

120

li/0
- . _
62
-
3bo
37
37
T3
170
10-100
220
2e
200
?1

7.2
5.7
A
C6
150
-
1
1
i
3
— 6d"

10-100


9
11
1-10
26
)
-
*
l.U
2ft
6b
7i3 —'
s
2.7
18
e.7

0.6	
"13
" 16
97
20

7.2
2.7
29
<<.9

29
17
26
U.7
1
1
!
N H
•<.9
120
3b
7.U
37
10-50
1-10
| heniolc Aciu
i Methyl benzoic Acid
C? Alkyl benzoic Acid
| CU ALkjrl Benzoic Acid
I Methyl Benicne SolrunaAide _	_
| Unidentified Compound* 1-iO ug/l
j Unidentified Compound* 10-100 Utf/1
; Xylene (? lflaaar*)
| Methyl Pentanol
Hethyl Cyclopenttnol		 	
r Dine I hyi Pentanone
Buta. yeth&ncl
I Methrl Hydroxy Prntajione
C? Altyl Beoieu* <3 Itoaert)
j CyclcaKxanone _	__ 	
' Oibutoxyaethajiol (2 U Alkyl Styrene _
Methyl Be niene Methanol
Isophorone
Phenaxy Methyl Cfelrene
f^ienyl Clh«r
CUlorophenoxy propanol	
Dloethyl nitn*iat«
ftienoiy Propanediol
Tetra oxadodecane
j j
1
6

| Diethyl Hithklate
I Wettyrl Propyl E«t«r of_Beniolc
pkcld
Pent*o»pentadec*n«
Bit (2-ethylhexyl) Ptithalat«
I Methyl Butajiolc Acid
9>
13
r.-o
17
3-0
•<5
10-ICO
10-lUO
"18
;i
?6o
?u
tu
<"'>
ttt
I.b
!•. 1
6.6
l<2
7.5
27
3.0
-------
Table 4-1 TEST RESULTS OF WATER AND SEDIMENT SAMPLING (Cont.)

AT-1
AT-?
AT-3
AT-U
*Izi
AT-6
AT-7
COMPOUND uk/1
Water Se.

l-?5

1-10
Ethyl Hexanolc Acid


I,?
10-100
To-100
10-50
Dimethyl Proponol



10-100

380
10-50
Phthallc Acid



1-10 8
1-25
100
y}°
tr&ns-l,2-dlchloroethylene
T < 5 T < .025
T<5
75 T < . 0?5
?6
36
T<5
6.U |
Toluene


H . 3U°_ ..
160
. 180
" "
160
Ethyl Beniene


69 0.6
10


31
Acetone


6.0

UOO

730
TrLchlorotrtfluoroeth&ne 1





5."»
Methyl Ethyl Ketone


330

1(10


Hexane






?.3
Methyl Iaobutyl Ketone



800
" Tioo

1600
2-Hexanol



78
9.6
r<5
37
2-methyl-l-prcpanal



T <5
T 5

5.2
Tetrachloroethane
6.U
18
12
5.7
0.2
6.2

1-butanol




T«5

T<5
1-hexanol




T< 5


1, 1-dlchloroethane


T < 5
T <5
T « 5

T<5
Vinyl Chloride







Dimethyl dlaulflde


6
T<5


T«5
Methylene Chloride
6.6
7. j	
5.3


T.k

Trl methyl Cyclohexanone


3?


15

Methyl Beniene Methanol





1.2

Carbonic Acid





3-5

Propoxr Butane


38


1.2

Dl-n-butyl Phth&late
39
7.5
o
!
vO
U"V
10
9.6
16

Fluoranthcne



CI9
7.2
T «1

Dimethyl Phenol





1-10

Hexanol Acid





12

Hydroxy Methoxy Benzene


**
1
t—
O

1-25
1-10

Pentachlorochenol





l». 1

Naphthalene


1.9 11
T < 1
M


1, 1, -trlchloroethane




9."»


Camphor




230


-------
Table 4-1 TEST RESULTS OF WATER AND SEDIMENT SAMPLINQ (Cont.)
AT-l	AT-?	AT-3	AT-U	AT-5	AT-6	AT-7
COMPOUND ufl/1	Water Sed. Wate. Sed. Mater Sed. Water Sed. Water Sed. Water tied. Water Sed.
Cj Alkyl Benzoic Acid
Methyleatera of Ethyl Hexanolc Acid
Benzene Propanlc Acid
Heptanol
HeDtanone


1-10
1-10
1-25
• -25
1-25
1U
220


Incene
Methyl (methylethanol) cyclohexanol
a-Terplneol
Terplnene
Iaophorona


l»9
20
2.U
62
l<00
110
19


Qulnollne
Hethyl Naphthalene
Acenaphthene
Dlpenzofuraa
Beta-Naphthmltrlle



1-10
9.6
9.6
12
U.8
U.8


J'luorene
Trlbutyl Eater of Riosphorlc Acid
Phenanthrene/Anthracene
Butyl Methyl Benzene Sulfonamide
[ Pyrene



1-10
72
7.2
5.7
17
2.2


Methyl Octanone
Butoxy Propanol
Cj Alkyl Styrene
Methyl (Propoxyloxy) Ethoxy Propanol
Butyoxy Ethoxy Ethanol


<<.7
2*4
3-5
6.2
110
12
2.2


Dimethyl Rienol
Tetra methyl Fentanone
Dlbutyl Ether
C5 Alkyl Benzene
Alkyl Phenol


5.6
l.U
11
11
1-10
0.2
1-10



Alkyl Styrene
Methyl Heptenona

>•.5
2.5
1-16



-------