United States Environmental Monitoring
Environmental Systems Laboratory February T987
Protection P.O. Box 15027
Agency Las Vegas, NV 89114
Research and Development
SEfifc GEOPHYSICAL AND
SOIL GAS INVESTIGATIONS
Phelps Collins Air National
Guard Base
Alpena, Michigan
Volume 1
prepared for
Air Force Engineering and Services Center
Tyndall Air Force Base, FL 32403-6001
-------
EJBD
ARCHIVE
EPA
-------
CXl TS-AMD-85580
^¦1 February 1987
TS~
VOLUME I
GEOPHYSICAL AND SOIL GAS INVESTIGATIONS
Phelps Collins Air National Guard Base
Alpena, Michigan
by
A11i son T. Baker
Lockheed Engineering and Management
Services Co., Inc.
Las Vegas, Nevada 89109
Douglas J. LaBrecque
Salt Lake City, Utah
C~
and
¦Jb
vV
LO Donn L. Marrin
Tracer Research Corporation, Inc.
Tucson, Arizona
^ and
Aldo T. Mazzell a
Advanced Monitoring Systems Division
Environmental Monitoring Systems Laboratory
Las Vegas, Nevada 89114
Contract No. 68-03-3245
Prepared for
Air Force Engineering and Services Center
Tyndall Air Force Base, Florida 32403-6001
Technical Monitor
Ann M. Pitchford
Advanced Monitoring Systems Division
Environmental Monitoring Systems Laboratory
Las Vegas, Nevada 89114
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
US EPA U-S- ENVIRONMENTAL PROTECTION AGENCY
Headquarters and Chemical Libraries?;!; B2127
NEVADA 89114
Mailcode 3404T
Repository Material
13wLrCtonD0c2oooN4W Permanent Collection
202-566-0556
-------
NOTICE
Although the information in this document has been funded by the Air Force
Engineering and Services Center and the United States Environmental Protection
Agency it does not necessarily reflect the views of the Center or the Agency
and no official endorsement should be inferred. The mention of trade names or
commercial products does not constitute either endorsement or recommendation
for use.
i i
-------
ABSTRACT
Geophysical and soil-gas surveys were conducted at the Phelps Collins Air
National Guard Base (ANGB) to help establish both the extent of contaminants at
that facility and the utility of the survey techniques to the Air Force Installa-
tion Restoration Program. Places examined include the drim disposal area, the
fire training pit, the water supply well area, a possible landfill by the
runway, an existing petrol eun, oil and lubricants (POL) storage area and two
planned POL areas. Techniques used include surface magnetics, electrical
resistivity soundings, pole-dipole resistivity profiles, electrical self-
potential, electromagnetic induction, seismic refraction, and the sampling and
analysis of soil gas.
A magnetic survey, an electromagnetic-induction survey (Geonics EM-31),
and a soil-gas survey were conducted in the drum disposal area. The geophysical
surveys established the extent of the disposal area; several broad positive
magnetic anomalies presumably identify concentrations of metallic debris. The
soil gas survey detected carbon tetrachloride between the drum disposal area
and the sewage treatment plant but no contamination was found in the remainder
of this area.
At the fire training area, a soil-gas survey identified a small volatile
organic contaminant plume. Contaminant concentrations were higher at shallow
i i i
-------
depths consistent with a surface source. An electrical self-potential survey
suggested ground-water flow direction to the north. A self-potential anomaly
was observed extending northeast from the fire training area. This pattern was
similar to that observed in the soil gas survey for toluene. One of two pairs
of resistivity soundings and seismic refraction soundings were inconclusive.
The other identified probable bedrock at 18 feet (5.5 meters). An
electromagnetic-induction survey (Geonics EM-34) did not detect evidence of
hydrocarbons in the fire training area.
The area surrounding the water plant and water supply wells no. 1 and no.
2 exhibits anomalously high electrical conductivity and the well water has
unusually high salinity. Pole-dipole resistivity profiling showed the conduc-
tive area to be centered southeast of the water plant and extending to an aban-
doned road-salt storage area. Resistivity and seismic refraction soundings
suggest the contaminant plume follows a depression in the bedrock surface.
Electromagnetic induction (Geonics EM-34) was inconclusive in this area probably
because of interference from cultural features. Water from well no. 3 is known
to be contaminated by low concentrations of trichloroethylene and chloroform.
A soil-gas survey around this well did not detect these compounds.
In the runway area, electromagnetic induction (Geonics EM-31) and magnetic
surveys did not detect signals indicative of a landfill or buried drums. This
area is assumed to be free of metallic contaminants. In the POL fuel storage
area, two soil-gas sampling locations had no detectable contamination and
remnants of a small (<100 gallons) JP-4 spill which had occurred two years
previously were not evident.
i v
-------
Of the two planned POL sites, the first had detectable amounts of methylene
chloride in the ground water and in some of the soil-gas samples. Soil-gas
sampling detected methylene chloride and tetrachloroethane in an area to the
northeast of this planned POL site and carbon tetrachloride in another separate
area. The second planned POL site appears to have minor, if any, volatile
organic contamination. However, contaminants were detected near the vehicle
maintenance building just to the west of this site.
v
-------
CONTENTS
Page
Abstract iii
Figures i x
Tables xi
Abbreviations and Symbols xii
Acknowledgement xii i
Introduction 1
Purpose of Investigation 1
Nature of the Problem 1
Scope of Work 4
Historical Background 6
Report Organization 8
Background Geology 9
Methods and Rationale 11
Geophysical Methods 11
Strategy 17
Drum Disposal Area 26
Site Description 26
Survey Results 27
Discussion of Results 33
Fire Training Area 34
Site Description 34
Site Objectives 35
Survey Results 35
Discussion of Results 46
Conductive Area 48
Site Description ' 48
Site Objectives 49
Survey Results 50
Discussion of Results 62
Miscellaneous Areas 64
Runway Area 64
(conti nued)
vi i
-------
CONTENTS (Continued)
Page
POL Fuel Storage Area 70
Planned POL Area No. 1 71
Planned POL Area No. 2 79
Study Area Geology 82
Objectives 82
Survey Results 82
Discussion of Results 88
Conclusions and Recommendations 89
Conclusions 89
Recommendations 92
References 93
Appendices
A. Daily Log of Field Work
B. Resistivity Profiling Data
C. Resistivity Sounding Data and Curves
D. Description of and Check on the Resistivity Algorithm ....
E. Electromagnetic Induction Data
F. Magnetics Data
G. Seismic "Time vs. Distance" Graphs
H. Soil Gas and Water Sampling Data and Contoured Maps of Results
(as per Tracer Research Corporation)
vi i i
-------
FIGURES
Number Page
1 Phelps Collins Air National Guard Base, Alpena, Michigan 2
2 Phelps Collins Air National Guard Base and Phelps Collins
Ai rport 3
3 Study areas for the Phelps Collins Air National Guard Base,
Al pena, Michigan. . « . . « . . * « . « . ¦ . . . 5
4 Arrays used for resistivity surveys at Phelps Collins ANGB. ... 12
5 Time-distance graph for seismic refraction data 15
6 Resistivity sounding lines at Phelps Collins ANGB 20
7 Resistivity profiling lines at Phelps Collins ANGB 21
8 EM-34 lines and points at Phelps Collins ANGB 22
9 Seismic refraction lines at Phelps Collins ANGB 23
10 Self-potential lines at Phelps Collins ANGB 24
11 Soil gas and water sampling locations at Phelps Collins ANGB. . . 25
12 EM-31-V survey lines and results at the drum disposal area. ... 28
13 Magnetics survey lines and results at the drum disposal area. . . 31
14 Self-potential lines and station locations in the fire
trainingarea. * . . . . . . . . . . * . . . . . . . . . . . . 37
15 Self-potential contours for fire training area 38
16 Approximate apparent resisitivies (in ohm-meters) derived from
EM-34-H data along line 4 41
17 Apparent resistivity of EM-34, horizontal dipole IOmeter data
in fire training area 43
18 Contours of toluene in soil gas at the fire training area .... 44
19 Computer contoured map of the base 10 log of apparent
resistivities pole-dipole (in ohm-meters) with 10-meter
dipoles and n - 3 . . . . . . . . . » . . . . . . . . . . . . . 51
20 Computer contoured map of the base 10 log of apparent
resistivities pole-dipole (in ohm-meters) with 10-meter
dipoles and n - 6 . . . . . . . . . . . . . . . . . . . . . . . 52
21 Computer contoured map of the base 10 log of apparent
resistivities pole-dipole (in ohm-meters) with 10-meter
dipoles and n - 1 . . . . . . . . . . . . . . . . . . . . . . . 53
22 Geoelectric section derived from two-dimensional modeling of
pole-dipole data from line 1 5 5
23 Geoelectric section derived from two-dimensional modeling of
pole-dipole data from line 22 . 56
24 Limestone-to-shale transition and survey line locations 57
25 Geoelectric section derived from two-dimensional modeling of
pole-dipole data from line 23 58
26 EM-34-H data from line 1 60
(conti nued)
ix
-------
FIGURES (Continued)
Number Page
27 EM-34-V data from line 1 61
28 EM-31 and magnetometer survey lines and contoured EM-31-H
results at the runway 66
29 Profile along line 14 of magnetic data 68
30 Methylene chloride distribution near planned POL no. 1 74
31 Carbon tetrachloride distribution near planned POL no. 1 75
32 TCA distribution near planned POL no. 1 76
33 PCE, TCE, and toluene concentrations in soil gas at planned POL
area no. 2.. «*»«.. «».«« 61
34 Geoelectric section derived from two-dimensional modeling of
pole-dipole data from line 4 84
x
-------
TABLES
Number Page
1 Phelps Collins Site Activity Schedule 6
2 Phelps Collins Site Survey Strategy 18
3 Vertical Soil Gas Profile at the Fire Training Area 45
4 Summary of Seismic Refraction Results 87
xi
-------
LIST OF ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
ANGB Air National Guard Base
EM electromagnetic induction
EMSCO Engineering and Management Services Co., Inc.
EMSL-LV Environmental Monitoring Systems Laboratory, Las Vegas
EPA Environmental Protection Agency
Hz hertz (cycles per second)
JP-4 jet fuel
I current (amps)
IRP Installation Restoration Program
K geometric factor
1 liter
m meter
mg mi 11i grams
mmho millimho
PCE perchl oroethyl ene
POL petroleum, oil, lubricants
S Siemen (= 1 mho)
SGA soil gas analysis
SP self potential
TCE trichl oroethyl ene
TDS total dissolved solids
ug micrograms
V voltage (volts)
SYMBOLS
p resistivity (ohm-meters)
pa apparent resistivity (ohm meters)
o" conductivity (mho/meter)
xi i
-------
ACKNOWLEDGMENTS
This study was accomplished as part of a research effort which was jointly
sponsored and funded by the Air Force Engineering and Services Center and the
U.S. Environmental Protection Agency's Environmental Monitoring Systems Labora-
tory-Las Vegas under interagency agreement number RW79931282-01-0. We thank
the Air Force Engineering and the Services Center Air National Guard for their
assistance and support. We extend special thanks to the Phelps Collins Air
National Guard Base civil engineers who provided background information and
continuous support throughout this study.
xii i
-------
CHAPTER 1
INTRODUCTION
PURPOSE OF INVESTIGATION
In 1984, the Environmental Protection Agency's Environmental Monitoring
Systems Laboratory (EMSL-LV) in Las Vegas, Nevada, and the Air Force Engineering
and Services Center entered into an interagency agreement concerning subsurface
contamination investigations for the Air Force Installation Restoration Program
(IRP). The agreement initiated research to demonstrate geophysical and soil
gas methods used in detecting and mapping ground-water and soil contamination.
Lockheed Engineering and Management Services Company, Inc. (EMSCO) performed
geophysical and soil gas surveys on several Air Force bases under contract to
EPA. The surveys and results are described in individual site reports such as
this one.
NATURE OF THE PROBLEM
The Phelps Collins Air National Guard Base (ANGB) is one site which was
studied under the agreement. The Base is located approximately 5 miles west of
the town of Alpena in Alpena County, Michigan (see Figures 1 and 2). It is
adjacent to the Alpena County airport, and both the Air National Guard, and
Alpena use the airstrip. The Base is used primarily as a training area during
1
-------
Figure 1. Phelps Collins Air National Guard Base, Alpena County, Michigan.
2
-------
-------
the summer when up to 2,000 persons can be accommodated. Approximately 50
people work on the Base year round.
Low levels of trichloroethylene (TCE), chloroform, and other contaminants
have been identified in the drinking water of the Base supply wells. Contami-
nant concentrations range from trace amounts up to approximately 0.010 mg/1.
The water wells are currently the only water sources for the Base.
Additional areas of primary interest to the Air National Guard regarding
ground-water contamination potential are the fire training area and the drum
disposal site. Other sites of lesser importance include a 1983 fuel spill site
at the petroleum, oil and lubricants (POL) fuel storage area, the old county
road salt storage site, an old county maintenance garage site, and a current
vehicle maintenance facility.
SCOPE OF WORK
Field surveys were performed at the Base during July and September 1985.
(See Table 1 for a schedule of site activities.) Electromagnetic induction
(EM), resistivity sounding, pole-dipole profiling, self potential (SP), magnetics,
and seismic refraction were the geophysical methods used. Soil gas analysis
(SGA) and water sampling were also used.
Sites with more intensive survey coverage are shown in Figure 3. Electro-
magnetic induction using the Geonics EM-31 terrain conductivity meter and
magnetics surveys were used at the drum disposal area and at the runway area to
4
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
TABLE 1. PHELPS COLLINS SITE ACTIVITY SCHEDULE (1985)
April 17
Preliminary site visit
May 1 to June 30
Preliminary site research
July 7 and 8
Visit local (Michigan) information sources
July 9 to 21
Geophysical survey: EM, SP, magnetics, and resistivity
sounding, and profiling
August 12 to 14
Site visit to meet with ANGB personnel and to observe
drilling operations
September 3 to 13
SGA, water sampling, and geophysical surveys including
seismic refraction and magnetics
detect and to locate buried metallic objects. The SP survey was experimental
in nature and was used around the fire training area to determine the utility
and limits of the method for detecting the ground-water gradient. Seismic
refraction and resistivity soundings were used to determine depth to water
and to gain a better understanding of the local geology. Electromagnetic
induction using the Geonics EM-34 terrain conductivity meter and resistivity
profiling were used across the Base to detect ground-water contamination and to
determine site geology. SGA sampling locations were chosen for the entire base
to locate areas of contamination and their sources. Ground water was sampled
at two locations to obtain an analysis of the ground water and also to compare
the results obtained with two independent analyses at the same locations.
HISTORICAL BACKGROUND
The Phelps Collins ANGB was first used just after World War I for a summer
training facility. In the late 1930's and early 1940's, the First Pursuit
6
-------
Group of the Army Air Corps used the Base for flight training and gunnery
exercises. In 1942, because of World War II and the expanding Army Air Corps
and its need for traininy facilities, the War Assets Administration acquired
the site and built housing, operations buildings, and runways. The installation
was turned over to Alpena County in 1947, and it was again used as a civilian
ai rport.
The site remained civilian, in the possession of Alpena County, until
1952, at which time negotiations were completed for the Air National Guard to
reassume possession of the training facilities with joint use of the airstrip.
The Base has been maintained as a field training area since that time with a
number of improvements having been made to the training facilities. The County
Road Commission occupied several buildings on Base from 1952 to 1958 to provide
airport and Base maintenance. The Base is now maintained by its own personnel,
has its own water supply and treatment facilities, and is fully equipped as a
field training site with equipment maintenance and training facilities for
approximately 2,000 persons (Hazardous Materials Technical Center, 1985).
Within the past decade, TCt was first detected in the water supply wells
of the Base, particularly in well no. 3 (Figure 3). Communication w.ith the
Department of Natural Resources of the State of Michigan prompted action to be
taken on the site regarding water supply (well no. 3 is no longer being used as
potable water). Soil and water samples were taken (August 1985) to detect
contamination at a planned POL fuel storage facility (planned POL storage area
no. 1). Wells were drilled in October and November 1985 to locate uncontami-
nated water sources.
7
-------
REPORT ORGANIZATION
The remainder of this report is divided into ten chapters and several
appendices. Chapter 2 describes the geology of the site. Chapter 3 describes
the methods used for this study and the rationale for method selection for this
site. Chapters 4 through 7 describe the work performed and the results from
particular areas on the Base where work was concentrated. In these chapters,
each method used, the results, and the interpretation are presented. Chapter 8
describes the site geology as determined through work performed on the Base
since the beginning of this project. Conclusions and recommendations are
presented in Chapter 9. Appendices to this report contain data collected at
the Base.
8
-------
CHAPTER 2
BACKGROUND GEOLOGY
The Phelps Collins ANGB is located in the Upper Devonian Traverse Group of
the Upper Michigan Basin. At the Base, the surface consists of glacial till
and soil an estimated 40 to 50 feet thick, according to well and borehole logs.
The bedrock which underlies the unconsolidated sediments is the Potter Farm
Formation, which is predominantly limestone and calcareous shale. The forma-
tions below the Potter Farm Formation are also predominantly limestone, and the
Alpena area and adjacent counties contain numerous sinkholes and other express-
ions of karst topography. The sinkholes of northern lower Michigan are most
common within exposures of the Traverse Group and tend to form linear trends
with southeast-northwest orientations along possible faults (Black, 1983).
A sinkhole is adjacent to the fire training area in the northwest portion
of the Base. It is approximately 350 feet (107 meters) wide and is of unknown
depth. Water stands about 40 feet (12 meters) below the land surface which is
about 30 feet (9 meters) below the level of surface water in the area. The
lower water level of the sinkhole may be indicative of a deep regional flow
system with a lower hydraulic head than the shallow unconfined aquifer (Black,
1983).
Wildermuth et al. (1924) classified the soil types on the Base into three
9
-------
general categories: (1) well drained soils, consisting of the Rubicon sand and
the Grayling sand; (2) mineral soils developed under conditions of deficient
drainage and consisting of the Saugatuck sand and the Granby sand; and (3)
poorly drained organic soils consisting of the Haughton muck, the Rifle Peat,
and the Lupton muck.
Existing data poorly defined the geological conditions of the site. The
interaction of the sinkhole with the upper and lower aquifers is unknown as is
the interaction of the South Branch Thunder Bay River with the geology of the
western edge of the site. Several of the geophysical surveys were performed
during this study to better determine the site geology. These surveys and
their results are presented in Chapter 8, Study Area Geology.
10
-------
CHAPTER 3
METHODS AND RATIONALE
Eight geophysical methods and soil gas and water sample analysis were used
over the course of this study. The geophysical methods consisted of electro-
magnetic induction, resistivity sounding, pole-dipole profiling, magnetics,
seismic refraction and self potential measurements. This chapter discusses the
theory involved for each of these methods. The second part of this chapter
describes where and why each method was used on the Base.
GEOPHYSICAL METHODS
Resistivity
The Wenner array for electrical resistivity was used at the Phelps Collins
ANGB (Figure 4). All of the soundings were made with an ABEM SAS300 resistivity
meter and a SAS2000 transmitter booster. The Bison offset sounding cable
system and steel stake electrodes were used for all soundings. The cable
system was designed to perform Wenner soundings quickly.
Some form of numerical modeling is generally required to relate the
resistivity of the earth to the instrument response. One-dimensional computer
11
-------
WENNER
ARRAY
-a +
a-
POLE-
DIPOLE
ARRAY
Electrode
at Infinity
H na
"vD-
- a- -*
Figure 4. Arrays used for resistivity surveys at Phelps Collins ANGB.
12
-------
inverse modeling (LaBrecque et al., 1984) was used to obtain layered earth
model resistivities from the resistivity sounding data.
Profiling determines horizontal changes in resistivity by taking a series
of measurements along a single line with a single a-spacing. Resistivity
profiling was performed at Phelps Collins using the pole-dipole array which is
depicted in Figure 4.
Profiling was performed by using an ABEM SAS300 resistivity meter and a
SAS2000 transmitter booster. The electrodes were steel stakes with attached
clips. Also used was a cable and switchbox system designed by Lockheed-EMSCO
to make pole-dipole measurements very rapidly.
The data were interpreted in two ways. One-dimensional modeling was used
for all data. In the conductive plume area, data from several lines were
combined and contoured. Two-dimensional modeling was used in areas of more
complex geoelectrical structures.
The electromagnetic induction instruments used in the Phelps Collins study
were the Geonics EM-31 and EM-34 terrain conductivity meters. Throughout this
report, EM-31 or EM-34 will be used to refer to those instruments or to the
surveys performed using them.
The EM instruments are used in two orientations. For horizontal dipoles,
the directions of the transmitted and received magnetic fields are parallel to
the surface of the ground at both the transmitter and receiver. For vertical
13
-------
dipoles, these magnetic fields are perpendicular to the ground surface.
Throughout this report, data collected with the EM-31 and EM-34 using the
horizontal dipole configuration will be designated EM-31-H and EM-34-H data,
respectively. Similarly, data collected with vertical dipoles will be desig-
nated EM-31-V and EM-34-V data.
The EM-31 has a fixed coil separation of 15.6 feet (3.7 meters) and an
operating frequency of 9.8 KHz. The EM-34 uses dipole separations of 33, 66,
and 131 feet (10, 20, and 40 meters), and operating frequencies of 6,400;
1,600; and 400 Hz, respectively. Both instruments are designed to read approxi-
mate apparent conductivity (McNeill, 1980).
Magnetics
Magnetic measurements were taken with a GISC0 G-856 proton precession
magnetometer.
Seismic Refraction
Equipment used for the seismic method included a seismic source, geophones,
and a seismograph. The refraction survey at Phelps Collins was performed using
the Bison Geopro 8012A signal processing seismograph. Twelve 10-hertz geophones
were planted into the ground at 3.05 m (10 ft.) intervals, and were connected
through a cable to the seismograph. The source was a 16-pound sledgehammer
striking a steel plate, usually at a 3.05 m (10 ft.) offset from each end of
the geophone line. Copies of the seismic records were made by printing the
14
-------
waveforms onto paper and by storing them on tape via the Bruel and Kjaer model
7400 digital cassette recorder.
The first arrival of the acoustic wave was picked on the seismic record
displayed on the seismograph. This was done in the field by moving a time line
on the seismograph and noting the time.
Time versus distance (T-x) plots (see Figure 5) were constructed from the
field notes. Arrival times (in milliseconds) for each geophone were plotted
against distance (in feet) from the source and best fit lines were drawn through
the points. Both forward and reverse shots were plotted on the same graph,
dots and x's respectively. From the apparent velocities and the intercept
times of both forward and reverse shots, the actual layer velocity, layer
depth, and slope of the interface were computed.
"O
c
. o
0)
(/>
100
75
50
25
vs
V3
(Xj
.Y2>
V
25 | 50
XCi
(X^Y,)
SOURCE LOCATION
75J
Xc2
100 125 150
DISTANCE (feet)
Figure 5. Time-distance graph for seismic
refraction data (modified from Mooney, 1984).
15
-------
Self Potential
The self potential method measures the naturally occurring electrical
potential of the earth without applying electrical current. Such natural
potentials are caused by oxidation-reduction reactions, temperature gradients,
and the flow of ground water. The latter source was the intended target at
Phelps Col 1ins.
Equifjment used in this survey consisted of three nonpolarizing copper-
copper sulfate electrodes, several reels of wire, and a high-impedance Fluke
77 multimeter.
Soil Gas Analysis
Geological and hydrological conditions were optimal for soil gas analysis
over most of the Phelps Collins site with sandy soils overlying a relatively
shallow aquifer. Soil gas sampling therefore should have detected any ground-
water contamination with the exception of methylene chloride, which has a
greater aqueous solubility and is less amenable to detection than many other
halocarbon compounds.
Equipment for this survey was supplied by Tracer Research Corporation. It
consisted of an analytical field van capable of hydraul ically driving and
pulling soil gas probes and transporting all equipment. Analysis equipment
used at Phelps Collins consisted of a Varian 6000 gas chromatograph and two
Spectra-Physics SP4270 computing integrators. Soil gas samples were collected
16
-------
at depths from 1 to 12 feet (0.3 to 3.8 meters). Detection limits for the
halogenated hydrocarbons ranged from 0.000001 to 0.08 ug/1 when an electron
capture detector was used while limits for hydrocarbons ranged from 0.03 to 20
yg/1 when a flame-ionization detector was used.
Water Sampling
Two water samples were collected and analyzed for contaminants. The same
equipment as that for SGA was used to collect and to analyze the water samples
with the addition of a peristaltic pump to draw the water samples. This process
is not recognized as a standard EPA analytical procedure.
STRATEGY
A series of measurements was developed to identify contaminants at specific
t.
sites across the Base and also to gain a better understanding of the local
geology. The locations, methods used at those locations, and the rationale for
using the particular method are shown in Table 2. Refer to Figure 3 for a map
of the area locations.
Figures 6 through 11 show where the measurements were made. These maps
show line or point locations for resistivity soundings, resistivity profiling,
EM-34, seismic refraction, self potential, and soil gas analysis and water
sampling. Maps of line locations and coverage for EM-31 and magnetics are not
included here, but can be found on pages 28 and 31 of the drun disposal area
(Chapter 4) and on page 66 of the runway area (Chapter 7).
-------
TABLE 2. PHELPS COLLINS SITE SURVEY STRATEGY
Locati on
Method
Rati onale
Drum Disposal
area
Fire Training
area and the
sinkhole
EM-31
Magnetics
SGA
EM-34
SP
SGA
Conductive
area
EM-34
This instrument was used to locate the buried
drums and to detect any saline contaminant plume
originating from the druns.
This method was used to determine the boundaries
of the buried druns and of other ferrous
objects.
SGA was used to determine if any drims were
leaking and, if so, to determine the contaminant
identity and the extent of any plume.
EM-34 was used over a large area in and around
the fire training area to detect any resistive
layers caused by hydrocarbons from by fire
training exercises.
SP was used to determine the ground-water
gradient around the sinkhole.
This method was used to map any hydrocarbon
contamination, especially in relation to
the sinkhole, and to determine the vertical
contamination distribution (a depth profile).
EM-34 was used in an attempt to define the edge
of the conductive area.
Runway area
Resistivity- This method was used to determine if a suspected
sounding and clay layer exists, to define its boundary, and
profiling to define the boundary of the conductive area.
SGA
EM-31
Magnetics
SGA was used near well no. 3, on the edges of
the area, to detect and map contaminants known
to exist in well no. 3.
EM-31 was used to determine the extent and the
characteristics of an anomalous area near the
runway.
This technique was used to detect and map any
ferrous objects that may have been buried near
the runway.
(conti nued)
18
-------
TABLE 2. (Continued)
Locati on
Method
Rationale
POL fuel
storage
facilities,
planned and
exi sti ng
Entire study
area
SGA
Water
sampling and
analysis
EM-34,
Resistivity
profi1i ng,
and sound-
ing
SGA
SGA was intended to determine if any contamina-
tion exists in these areas and, if so, to
determine the contaminants and their extent in
the soil gas.
This was intended to determine the constituents
and the concentration of contaminants and also
to compare results with previous results.
These surveys were used to gain a better under-
standing of the geology of the site. In so
doing, possible contaminant transport paths and
other anomalous or contaminated areas which were
previously unknown may be indicated.
Samples were collected at many locations on
Base to locate areas of contamination.
the
19
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
CHAPTER 4
DRUM DISPOSAL AREA
SITE DESCRIPTION
The drum disposal area encompasses approximately 5 acres in the north-
western part of the study area. It is adjacent to a swampy portion of the
south branch of the Thunder Bay River and is immediately north of the sewage
lagoon for the water treatment plant. It is covered with grass and demolition
fill, and the surrounding area is heavily wooded with dense underbrush. The
land surface is approximately 10 feet above the exposed water surface in the
ri ver.
Large numbers of drums containing solid and liquid waste that originated
with operations at the Base were discarded at this site. The number of drums
and their content are not known. The drums were allegedly discarded where the
bank drops off to the river and were then covered with fill. Demolition fill
is still dumped at the edge of the river (as observed by a field crew on site).
The change in the river bank can be noted by comparing old maps of the base with
the current topography of the area.
The sewage lagoon is also a suspected source of contamination. Wastes
26
-------
from chemical spills on the ANGB, in addition to normal sewage, were deposited
in the lagoon for many years. The exact chemicals are not known. A feasibility
study was begun in September 1985 by Phelps Collins ANGB to determine the
extent of seepage from the lagoon.
SURVEY RESULTS
EM-31
Measurements were made at 5-meter spacings along 4 lines spaced 33 feet
(10 meters) apart (see Figure 12). Four measurements were made at each station.
These consisted of north-south and east-west measurements with horizontal
dipoles, and north-south and east-west measurements with vertical dipoles. The
vertical and horizontal measurements are two individual data sets since these
represent different penetration depths.
The extent of the buried drums could not be determined from the EM-31
results since the survey did not cover the entire drum area; however, effects
of the buried drums and other conductive debris were readily apparent in the
data. The north-south measurement was often radically different from the east-
west measurement; whereas, in an undisturbed area, the two measurements should
be the same. Anomalously high conductivities were noted throughout the area.
The most practical way of displaying results of the EM-31 survey with such
dramatic cultural influences was to produce two maps of the data. The first
map is a difference map which is the absolute value of the difference between
27
-------
PHELPS COLLINS
/ DRUM DISPOSAL AREA
APPROXIMATE SCALE
FEET
50
-T"
20
100
I
"r~
40
2 0 0
f
60
METERS
Legend
X
Deer Fence
X X X X
Barbed Wire Fence
tt
Swamp
Fill
*» ^
Road
Sewage Lagoon
Buried Drums/
Ferrous Objects
Alleged Buried
Drums
EM31V
High readings, conductivity
in mlliimhos/meter
Contour Interval =
1 0 millimhos /meter
EM3 I V
Absolute difference
between readings
Contour Interval =
1 millimho/meter
-I EM31 Data
Location Lines
Figure 12. EM-31-V survey lines and results from the drum disposal area.
28
-------
the north-south and the east-west values. The second map is a highest value
map which is a map of the largest of the two measurements at that location. A
conductor is indicated when there is a large difference between the two values
or when the highest value is large. A representation of the highly conductive
areas can be obtained by comparing the two maps.
EM-31-V data are- presented in Figure 12. On the highest value map, anomalies
parallel the suspected edge of buried objects in the northwestern part of the
surveyed area. An anomaly in the southwest shows the large conductivity ampli-
tude expected for buried drums. The difference map targets a northern suspected
drum area and also detects larger, lower amplitude, electromagnetic anomalies.
Armored personnel vehicles on the site caused high anomalies in the mid- to
upper-right portion on both maps.
Only the EM-31-V data are presented here. The EM-31-H was influenced by
surface trash as much if not more than by the buried conductors.
Magnetics
The magnetics survey was performed after the armored personnel vehicles
had been removed. Since the EM-31 survey showed the disposal site to be larger
than had been previously anticipated, the area of the magnetometer survey was
extended. Measurements were made at 6.4-feet (2-meter)spacings on 8 lines spaced
33 feet (10 meters) apart (see Figure 13). Lines 3 through 6 are in the same
locations as the EM-31 lines.
29
-------
Two sets of magnetometer measurements were made at each station: total
field and gradient. Total field anomalies were as great as 1,600 gammas over
the center of the area. Gradient anomalies were as high as 240 gammas per
meter. A deer fence at the south end of the area produced gradient anomalies
on the order of 200 gammas per meter when the sensor head was within 2 feet
(0.6 meters) of the fence. Other ferrous objects on the surface also influenced
measurements, and these were noted in the field and accounted for when determin-
ing buried object boundaries.
Suspected boundaries of buried drums and other ferrous objects (or both)
are shown in Figures 12 and 13. These boundaries were determined from the
magnetometer results which are shown on the overlays in Figure 13. Two types
of areas have been delineated: (1) those with lower amplitude and higher
spatial frequency, indicating low iron content; and (2) those with high ampli-
tude and lower frequency indicating a higher concentration of metallic objects.
The second type area, as depicted by darker shading, represents where the
drums are suspected to have been buried. The total field is very high in those
areas. The gradient indicates that there may be two areas of buried drums
adjacent to each other in the southwestern portion of the area; whereas, the
total field does not make this distinction. Models of total field and gradient
component anomalies indicate the latter to be more focused on the boundaries of
areas containing considerable iron content (Gilkeson et al., 1986).
The lightly shaded area is more representative of demolition fill which
covers most of the site. One such anomaly is near the fill edge just north of
30
-------
/ PHELPS COLLINS
I DRUM DISPOSAL AREA
APPROXIMATE SCALE
FEET
50
L.
T~
20
100
I
I
40
200
{
60
METERS
Legend
Xx
Deer Fence
X X X X
Barbed Wire Fence
a*" «»
»', ^ *s
Swamp
.-V
Fill
Road
Sewage Lagoon
Buried Drums/
Ferrous Objects
Alleged Buried
Drums
Total Magnetic Field
(difference between
background and
point location readings)
Contour Interval = 100-gammas
Magnetic Field Gradient
(difference between
background and
point location readings)
Contour Interval = 30 gammas
Magnetometer Data
Location Lines
Figure 13. Magnetics survey lines and results from the drum disposal area.
31
-------
the two lower anomalies. This anomaly is found on both sets of magnetics
results. Demolition fill consisting of rebar-reinforced concrete and 6-inch
piping interspersed with dirt and rock fill can be seen where the fill area
drops off toward the swamp. The iron content known to exist in this area is
less dense than that in the three targeted anomalies.
SGA
Seven soil gas samples were collected in the vicinity of the drum disposal
area and the sewage lagoon. Those taken to the north and east of the drum area
were to check for any contaminants leaking from the drums. The samples between
the drum area and the sewage lagoon were collected to determine if contaminants
were present there although it could not be determined whether the source was
the drums or the lagoon. One sample was collected near the lagoon and away from
the drum area to check for contamination caused solely by the lagoon.
Caution was necessary when driving probes near the suspected drum area
since the drun locations were not known. The operators did not want to puncture
any drums for safety reasons since the drum contents were unknown and might
be dangerous. In three instances when the probes abutted hard objects, the
probe was withdrawn and relocated.
Carbon tetrachloride was detected in samples collected between the drums and
the sewage lagoon (points 4A, 93, and 95, Figure 11); however, it is impossible
to determine which is the source because of the close proximity to both possible
sources. Any attempt to isolate the source by collecting samples close to one
32
-------
area and away from the other was inconclusive since nothing was detected in any
other samples. No contamination was found in any of the remaining samples
north or east of the drum area (points 1, 2, and 3) or east of the sewage
lagoon (point 94).
Carbon tetrachloride was evident in the area between the sewage lagoon and
the drum disposal area. The lack of contamination north and east of this area
suggests (1) that the ground-water flow is in a westerly direction, toward the
river, (2) either or both areas could be sources, and (3) contamination, if it
is migrating from the source, flows toward the river.
DISCUSSION OF RESULTS
Results from EM-31, magnetics, and SGA surveys were used to determine the
extent of ferrous and metallic object burial, the types of contaminants in the
area, and contaminant migration directions.
The EM-31 and magnetics surveys suggest three areas containing a high
density of ferrous objects. These probably represent the drums which were
allegedly buried at this site. Considerable demolition fill with lesser iron
content is indicated over a greater area.
The SGA survey detected carbon tetrachloride between the buried drum area
and the sewage lagoon. No contaminants were detected east or north of either
area. This indicates that carbon tetrachloride is coming from either or both
areas and if it is migrating, flows into the swamp portion of the river.
33
-------
CHAPTER 5
FIRE TRAINING AREA
SITE DESCRIPTION
The fire training area is approximately 100 meters in diameter and is
adjacent to the sinkhole in the northern part of the study area. The area
northeast, and immediately surrounding the fire training area is covered with
grass. It changes into heavily wooded areas both to the west and as it approaches
the sinkhole to the south. The hydrogeology of the fire training area is
considered to be linked to that of the sinkhole because of their proximity.
A circular concrete pad covers the fire training area. Old airplane bodies
and vehicles have been placed on the pad, saturated with fuel (JP-4), and
ignited for fire training exercises. The concrete is jointed and cracked and
therefore not impervious to fuels or chemicals used to extinguish fires. A
hydrocarbon odor is prevalent in the center of the area. A small channel has
eroded through the west side of the fire training area embankment, and runoff
from that part of the area drains into the north side of the sinkhole.
The sinkhole is approximately 330 feet (100 meters) across and has water
standing about 40 feet (12.5 meters) below the land surface. The water surface
34
-------
of the sinkhole is about 30 feet (9 meters) lower than that of the South Branch
Thunder Bay River (west of the site) and of Lake Winyah (north of the site).
SITE OBJECTIVES
The objectives at this site were to determine (1) the extent of the hydro-
carbon contamination (2) possible modes of transportation or migration of con-
tamination from the fire training area, (3) the role of the sinkhole with
respect to hydrocarbon migration, and (4) the role of the sinkhole with respect
to ground-water flow in the area. Methods used on this site were resistivity
soundings, seismic refraction, EM-34, SP, and SGA. EM-34 was used to detect
and to determine the extent of any resistive area caused by hydrocarbons migrat-
ing from the fire training area. SP was used to determine the ground-water
flow in the fire training area. SGA was used to determine the extent of con-
tamination laterally and with depth. Resistivity measurements were used in
conjunction with seismic refraction measurements to determine the geology of
the study area.
SURVEY RESULTS
Self Potential
The self potential survey was performed along line 4 and along lines 7
through 11. Measurements were made at 160-feet (50-meter) intervals with a
base station electrode located at station 0 on line 4. Measurements were made
at the base station before and after each line survey. The data were then
35
-------
corrected for the variations observed in the base station values. These
corrections ranged from 1.6 to 5.2 millivolts and were at most 10 percent of
the self potential variations observed over the survey area (on the order of 50
to 95 millivolts).
The self potential results are shown in Figure 14. The lines, station
locations, and corrected self potential values are shown in Figure 14 and the
contoured map is shown in Figure 15. Interpretation of self potential data is
a difficult task and can depend on a wide range of variables. For example, the
polarity of the electrokinetic (streaming) self potential can be either positive
or negative with respect to the ground-water flow depending upon the mineral
composition of the subsurface and the chemical composition and pH of the elec-
trolyte. If one assumes that the upper aquifer is dominated by the presence
of clay and the electrolyte is basic (Miller, 1978), then an excess of negative
ions would exist in solution and the electrokinetic self potential would decrease
in the direction of flow (Dakhnov, 1962). The self potential data could also
be reflecting flows in the deeper fractured limestones. In this case, assuming
the formation is calcite and the electrolyte pH is less than 9, the electrokinetic
self potential would again decrease in the direction of flow. With these
assumptions, it would appear that the self potential data reflect a.northward
ground-water flow as indicated by the arrow between lines 10 and 11 in Figure
15. This is in agreement with the regional ground-water flow towards the lake
(Hazardous Materials Technical Center, 1985).
A self potential anomaly is observed in the fire training area, line 9,
station 0. The anomaly extends about 330 feet (100 meters) towards the northwest
36
-------
A'
0 50 100
1 I I
SCALE (meters)
SELF POTENTIAL (millivolts)
Figure 14. Self potential lines and station locations in
fire training area. The approximate location of an underground
water line is indicated (Scott, 1970).
37
-------
Approximate location of_
underground water line"
L/'ne 4
S.P. GROUNDWATER
FLOW DIRECTION
GROUNDWATER FLOW DIRECTION
SELF POTENTIAL (S.P.) (millivolts)
CONTOUR INTERVAL 10 millivolts
Figure 15. Self potential contours for fire training area. The datum
at line 8, station zero (0) was questionable and so was excluded
in the drawing of the self potential contours.
38
-------
and then trends northward. The initial portion of this pattern is similar to
the results, from the soil gas survey for tolune shown in Appendix H, Figure 4.
The soil gas survey did not cover the area to the north and so no comparison
can be made there. The approximate location of an underground water pipe line
is indicated in Figure 14 (Scott, 1970). It lies in the area of the self
potential anomaly. No other information is known about this pipe line, such as
its composition or whether it still exists, and so its effect on the self
potential cannot be estimated. In any case, the existence of a disturbed soil
trench could provide a higher permeability path for the contamination migration.
An exact interpretation of this self potential anomaly is not possible with the
existing information. It may be reflecting the contamination zone, changes in
the geology, cultural features or a combination of all three.
The sinkhole does not appear to dominate the ground-water flow of the Base,
but it may affect the flow in its immediate vicinity. A positive self potential
anomaly could be expected in the limestone formation for ground-water flow into
the sinkhole. The sinkhole does lie in an area of higher self potential,
however, there is insufficient data coverage around the sinkhole to clearly
evaluate the streaming potential flow in that area. High density coverage of
stations to the east, west, and south of the sinkhole are needed to adequately
interpret any results.
Resistivity
Two resistivity soundings, numbers 4 and 9, were performed to delineate
structures near the sinkhole. Depths to water and bedrock, as derived from
39
-------
seismic refraction conducted in the same locations, were used to constrain the
layer thicknesses in resistivity sounding inversions.
Data from sounding 9 could not be fit with a model having the same layer
thickness as required by the refraction data. Lateral changes in the subsurface
geology, which are probable as they are evidenced by the sinkhole, may have
affected the resistivity data, the refraction data, or both. The results of
both methods at that line location must be considered questionable. The sound-
ing does indicate the depth of bedrock to be approximately 66 feet (20 meters).
At sounding 4, located about 800 feet (250 meters) northwest of the sink-
hole, a 1337 ohm-meter layer at 17.6 feet (5.5 meters) probable depth (refer to
sounding 4 data in Appendix C) appears to be bedrock. The resistivity of the
layer above the bedrock layer is somewhat higher than that of saturated sediments
on other parts of the Base, and it appears that the water table occurs at or
below the top of the bedrock in this location.
EM-34
Measurements were made at 164-feet (50-meter) intervals along .lines 1, 4.
Measurements were made using 33-, 66-, and 131-feet (10-, 20-, and 40 meter)
coil separations along lines 7 through 10. Figure 16 plots the measurements
for horizontal dipoles along line 4. Values have been converted from approxi-
mate apparent conductivities to approximate apparent resistivities for compari-
son with resistivity data. Results of the survey along this line are rather
40
-------
LINE 4 EM34 DATA
600 r
500
400
300
200 -
Coil Separation
O 10 Meters
<3 20 Meters
~ 40 Meters
100
_L
400
WEST
300
200
100
00
EAST
METERS
Figure 16. Approximate apparent resistivities (in ohm-meters) derived
from EM-34-H data along line 4.
41
-------
disappointing. Although there is a visible increase in resistivity at the east
end of the line, closer to the sinkhole, there are no distinct anomalies.
Results from lines 7 through 10 proved equally disappointing. Conductivity
was very low throughout the entire area and showed no indication of the hydro-
carbons' in the fire training area. The latter is probably a direct result of
the former observation. The hydrocarbons did not lower the subsurface conduc-
tivity since it is already low. Therefore, because of the inherent geological
conditions, it was impossible with EM to detect the hydrocarbon extent.
The data for the EM-34, horizontal dipole, 33-feet (10-meter) spacing were
contoured and are shown in Figure 17. The most significant apparent resistivity
variations are seen along line 8. Low apparent resistivity values are observed
along this line, becoming less resistive to the north. These data may be
reflecting the underground water pipe line, a disturbed soil trench, a change
in the geology, or a combination of all three.
SGA
Soil gas samples were taken from depths of 5 feet (1.5 meters), at eleven
points in the fire training area. Elevated levels of toluene (up to 1000 mg/m^),
benzene (700 mg/m^), methylene chloride (up to 2 mg/m^), trichloroethylene (0.3
mg/m3), and perchloroethylene (PCE) (0.003 mg/m^) were detected in the soil. A
plume which appears to be very limited in size appears to have migrated northeast
from the center of the fire training area. Figure 18 shows toluene concentra-
tions in the area. (For complete results, refer to Appendix H.) The highest
42
-------
APPARENT RESISTIVITY (ohm-meters)
CONTOUR INTERVAL
25,50,100,200,400 ohm-meters
Figure 17. Apparent resistivity of EM-34, horizontal dipole 10 meter data
in fire training area. A conductive channel is suggested along line 8.
43
-------
Figure 18. Contours of toluene in soil gas at the fire training area.
44
-------
concentration of toluene was at point SG10 which is close to a drainage channel
from the center fire ring. The high concentration of TCE at point SG8 may
represent either a secondary source area or a slug of TCE which has migrated
downgradient from the ring center.
In an attempt to confirm whether the high concentrations of hydrocarbons
were a result of soil or ground-water contamination, a soil gas measurements
were made at several depths. Samples were collected from 3, 8, and 12 feet
(0.9, 2.4, and 3.7 meters) at a point adjacent to sample point SG10. Benzene
and toluene results are shown in Table 3. The general trend is decreasing
concentrations with depth, which is generally indicative of soil contamination.
Although the ground water beneath the fire training pit may contain some benzene
and toluene, the shallow soil is apparently the major source of these compounds
in soil gas. Shallow soil contamination is not unusual at sites where volatile
organic chemicals are applied directly to the ground surface.
TABLE 3. VERTICAL SOIL GAS PROFILE ADJACENT TO POINT SG10
AT THE FIRE TRAINING AREA
Soil Gas Concentration (mg/m3)
Depth (feet) Benzene Toluene
3 700 1000
8 <10 1000
12 <10 500
45
-------
DISCUSSION OF RESULTS
Self potential, resistivity soundings, EM, and SGA were performed in the
vicinity of the fire training area to determine contaminant concentrations in
and around the area, to determine the role of the sinkhole in contaminant
migration, and to gain a better understanding of the geology of the area,
particularly concerning the sinkhole and the ground-water flow in its vicinity.
The geophysical methods were useful in determining geology and ground water
conditions surrounding the sinkhole, but the soil gas survey (SGA) was the
instrumental method for detecting contamination levels.
Resistivity soundings indicate that depth to bedrock is about 66 feet (20
meters) adjacent to the sinkhole and about 18 feet (5.5 meters) approximately
820 feet (250 meters) northwest of the sinkhole. The water table is probably
at or below the top of the bedrock at the latter location.
Self potential results suggest a ground-water flow direction to the north
and northwest in agreement with the regional ground-water flow toward the
lake and river. A self potential anomaly is observed extending northeast from
the fire training area and then trending northward. The initial portion of
this pattern is similar to that observed in the results of the soil gas survey
for toluene. The soil gas survey did not cover the area to the north and so no
comparison can be made there. The self potential anomaly may be reflecting an
underground water line or disturbed soil trench. Additional soil gas measure-
ments across this north-trending self potential anomaly would help confirm
whether this could be a conduit for contamination migration.
46
-------
There is insufficient self potential data around the sinkhole to clearly
evaluate the streaming potential flow in that area.
EM results proved to be inconclusive in this area for detecting hydro-
carbons. A north-trending conductive zone is suggested possibly reflecting the
underground water pipe line, a disturbed soil trend, a change in the geology,
or a combination of all three.
SGA proved most valuable for detecting the contaminant conditions. High
concentrations of toluene and benzene with lesser concentrations of methylene
chloride, trichloroethylene, and PCE were measured in and around the fire train-
ing area. A plume of contaminants appears to have migrated to the northeast.
This would suggest fluid flow and migration away from the sinkhole. The highest
concentrations were detected close to the center of the fire ring. Vertical
concentrations decrease with depth which indicates that the soil surface is the
major source of the soil contamination.
A potential problem concerning contamination characterization at this site
revolves around the sinkhole and possible transport paths and other structures
which may be associated with it. High ground water flow velocities through
fissures and solution cavities can transport contaminants great distances in a
relatively short time. Additional complexities are created with the possibility
of the sinkhole being connected to the lower, semi confined aquifer which appears
to be drawing ground water and thus any contaminants in the ground water from
the sinkhole into the lower aquifer system. Hence, the lower aquifer may be
the contamination victim with relatively uncontaminated water above it. However,
the SGA data suggest that this possibility did not occur because the contamina-
tion plume apparently moved away from the sinkhole.
47
-------
CHAPTER 6
CONDUCTIVE AREA
SITE DESCRIPTION
The conductive area surrounds the water plant in the center of the study
area. It encompasses the Base water supply wells no. 1 and no. 2. The majority
of the area is covered by heavy woods and dense underbrush. Those areas which
are clear tend to be close to overhead power lines, roads, or other cultural
objects or obstructions.
Well no. 3 is not within the conductive area but will be included in this
discussion since it is close to the conductive area and is one of the three
water supply wells, the other two of which are within the conductive area.
Although it produces less than the other wells, well no. 3 had been the best
water supply of the Base with much lower total dissolved solids (TDS) than
either well no. 1 (greater than 500 ppm) or well no. 2 (greater than 1000 ppm).
However, well no. 3 was closed when low levels of TCE, chloroform, and other
contaminants were detected in the water. These contaminants have not been
found in wells 1 or 2. Well no. 3 is also in a heavily wooded area. The roads
which lead to the well are fenced off from vehicle traffic and are becoming
overgrown.
48
-------
Possible explanations of the conductive area include (1) high TDS in the
water and (2) a clay layer. Well logs identify an existing clay layer near the
river. There are also two possible salt sources.
In the past, Alpena County stored road salt in an area southwest of planned
POL storage area no. 1 (refer to Figure 3, p. 5). The salt was not confined
in any manner to prevent percolation into the aquifer system. Also, a well
drilled prior to 1926 encountered rock salt at 200 feet. The exact location of
this well is unknown; however, Base personnel suspect it is located in the
vicinity of ttie water plant. It is not known whether or not the well was cased
or plugged, or if it was plugged, whether or not it was plugged properly. With
improper abandonment procedures, the current water pumping in the area could
draw brine up the abandoned well and into the existing aquifer.
SITE OBJECTIVES
Objectives at the site were to determine the edge of the conductive area,
the source of the high conductivity, whether a clay layer exists, and its
boundaries. Objectives near well no. 3 were to detect and to map contaminants
known to exist in the well. EM-34 and pole-dipole profiling and resistivity
sounding were the methods used in the conductive area, and SGA was used near
well no. 3. EM-34 and resistivity were used to determine the edge of the con-
ductive area. Resistivity measurements were also used to determine the geology
of the area and in particular to check for the presence of clay. SGA was
intended to map contaminants from well no. 3 and to determine their source.
Some seismic refraction was also performed to aid in the interpretation of
resistivity results in addition to determining the geology of the study area.
49
-------
SURVEY RESULTS
Resistivity
Resistivity profiling in this area was performed with 33-feet (10-flieter)
dipoles along lines 1, la, 2, 3, 5, 22, and 23 (refer to Figure 7, p. 21). The
results showed a large low apparent resistivity area in the vicinity of the
Base water supply wells. The low resistivity area may be due to high salinity/
high TDS ground water; however, it may also be reflecting changes in the geology
which is also more complex than had been originally anticipated.
Resistivity profiling results can be shown as cross sections, or if data
from multiple lines are available, as contour plots at a selected depth. The
depths selected are represented by values for n. Larger values of n indicate
increased depths. Figure 19 is a computer contoured map of pole-dipole data
with n = 3. The base 10 log of the apparent resistivity values was contoured
because of the broad range of values across the study area. The log = 2 contour
interval represents the pattern of the low resistivity plume. The anomaly is
elongated along a northwest-southeast trend with the lowest apparent resistivity
values centered south-southeast of the water plant. The lateral boundaries of
this low resistivity area requires two-dimensional modeling and is discussed
below.
Figures 20 and 21 are computer-contoured maps of the pole-dipole data with
n = 6 and n = 1, respectively. The n = 6 data contour is approximately the same
50
-------
(FEET)
0 200 400 800
APPROXIMATE | ¦ ^ i | 1
SCALE
50 100 150 200
(METERS)
Figure 19. Computer contoured map of the base 10 log of apparent resistivity
pole-dipole data (in ohm-meters) with 10-meter dipoles and n = 3.
51
-------
APPROXIMATE
SCALE
i 1 r
100 150 200
(METERS)
Figure 20. Computer contoured map of the base 10 log of apparent resistivity
pole-diploe data (in ohm-meters) with 10-meter dipoles and n = 6.
52
-------
(FEET)
0 200 400 800
APPROXIMATE i | i ¦
SCALE 1'II1
0 50 100 150 200
[METERS)
Figure 21. Computer contoured map of the base 10 log of apparent resistivity
pole-dipole data (in ohm-meters) with 10-meter dipoles and n = 1.
53
-------
as the n = 3 data. The n = 1 data, however, produced erratic results since the
depth of penetration is less than the plume depth.
The plume is evident in pseudosections of the data and in the two-
dimensionally modeled results. (All profiling data and pseudosections can be
found in Appendix B.) Figure 22 depicts the results of two-dimensional modeling
of the southern portion of line 1. The fluctuating resistivities in the segments
of the shallow layer probably reflect fluctuations in depth to the water table
and not actual resistivity changes. Of these, the more resistive segments
indicate greater depth to water. The plume appears as the 55 ohm-meter segment
beginning at station 340S and extends to the south. The layer extends from 16
to 66 feet (5 to 20 meters) in depth.. This boundary corresponds to about the
2.2 apparent resistivity contour in Figure 18.
A change in bedrock type was also found to exist near the northern edge of
the conductive plume. The bedrock appears to change from limestone to shale
beginning near station 390 on line 1 (Figure 22). The two-dimensional modeling
results of line 22 (Figure 23) indicate that the conductive plume and the
shale/limestone boundaries nearly overlie each other at station 90. The plume
is now represented by the 75 and 45 ohm-meter zones and extends from about 10
to 67 feet (3 to 20.5 meters) in depth. The transition from shale to limestone
is distinct and can be traced by an essentially straight line represented in
Figure 24. Lines la, 2, and 3 appear to be completely underlain by shale, as
indicated by their low apparent resistivities for n = 6 data. Although line 23
does not overlie the shale, two-dimensional modeling (Figure 25) suggests that
it crosses the northeast edge of the plume. The conductive plume is represented
54
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
V)
t-
3
(/)
Ui
cc
o
Ui
O
o
D
i
CM
CM
CM
UI
z
(SJSjeui) Hld30
Figure 23. Geoeloctric section derived from two-dimensional modeling of
pole-dipole data from line 22. Values are in ohm-meters.
56
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
o
IT)
o
o
m
o
CM
o
fO
(SJ0J9LU) Hld30
Figure 25. Geoelectric section derived froin two-dimensional modeling of
pole-dipole data from line 23. Values are in ohm-meters.
58
-------
by a 50 ohm-meter zone that extends from about 36 to 85 feet (11 to 26 meters)
in depth.
Resistivity soundings, in the vicinity of the conductive area, were made
at sounding locations 2, 3, 6, and 7 (see Figure 6, p. 20). Depth to water, as
derived from seismic refraction conducted in the same locations, was used to
constrain the layer thickness in resistivity sounding inversions.
Inversion results show the depth to bedrock at sounding 6 to be substan-
tially less than that at soundings 3 and 7, closer to the conductive plume.
Sounding 2 indicates a bedrock high north of the plume as well. Two-dimensional
modeling results of the profiling data taken on line 3 substantiate these
findings and indicate that this bedrock high probably continues to the southeast.
Thus, with the shale acting as an aquiclude below and with probable aquitards
to the north and south, the conditions are ideal for contaminant migration from
the old county salt storage area to the well field. Migration would be assisted
by drawdown caused by the water wells.
EM-34
A plot of EM-34-H data taken along lines 1 and la is shown in Figure 26.
The data show a subtle decrease in apparent resistivity corresponding to the
conductive plume (centered near station 500). EM 34-V data for the same lines
are shown in Figure 27. Vertical dipole results are very erratic particularly
for the 131 feet (40-meter) coil separation and are inconsistent with the
geology as determined by the pole-dipole data.
59
-------
fl c*
{SU313W-HHO) A1IAI1SIS3U !N3UVddV
Figure 26. EM-34-H data from line 1. Values have been converted to
approximate apparent resistivities (in ohm-meters).
60
-------
<
I-
<
o
*
«
I
5
UJ
UJ
z
o
O
tn
in
«s>
w
QJ
a>
**
a>
a>
a>
£
Z
S
O
o
o
ca
o
<
~
o
o
^7
(SU3JL3W-WHO) A1IAI1SIS3H !N3UVddV
Figure 27. EM-34-V data from line 1. Values have been converted to
approximate apparent resistivities (in ohm-meters).
61
-------
SGA
The soil gas survey was ineffective in determining the source of contamina-
tion in well no. 3. Results showed no contamination adjacent to the well.
DISCUSSION OF RESULTS
Results from resistivity soundings and profiling, EM-34, and SGA were used
to determine characteristics of the conductive plume and characteristics of the
contaminants in well no. 3. The resistivity methods were the methods most
successful in determining characteristics of the conductive plume and of the
geology of the area.
Pole-dipole results show the plume to be centered south-southeast of the
water plant and to be elongated in a northwest-southeast trend. The mapped
results show the approximate extent of the plume. Two-dimensional modeling
indicated a low resistivity area ranging from 45 to 70 ohm-meters and from about
10 to 85 feet (3 to 26 meters) in depth. A change in bedrock type was detected
from two-dimensional modeling of the data. A limestone-to-shale bedrock transi-
tion was located (Figure 24). Lines la, 2, and 3 appear to be completely under-
lain by shale, while line 23 is underlain by limestone. Line 23 also crosses
the northeast edge of the conductive plume.
Sounding results show the conductive area to be associated with a bedrock
low. These results and the results showing both location and elongation direc-
tion of the plume, combined with probable drawdown caused by well pumping from
62
-------
wells no. 1 and no. 2 indicate the old salt storage area to be the source of the
conductive plume.
EM-34 data did indicate higher conductivity in the vicinity of the water
plant; however interference from deer fences and power lines both prevented
many readings from being taken and caused occasional spurious readings. Survey
coverage was therefore inadequate to produce interpretable results. This was
not a serious problem since the resistivity methods produced yood results with
better resolution.
The source of well no. 3 contamination could not be determined. SGA
results showed no contamination adjacent to the well.
63
-------
CHAPTER 7
MISCELLANEOUS AREAS
In this chapter, four areas will be discussed. They are the runway area,
the POL fuel storage area, and the two planned POL fuel storage areas. These
areas are indicated by shading in Figure 3 on page 5.
RUNWAY AREA
Site Description
The runway area is adjacent to taxi way C in the southeastern part of the
study area. It is located at the end of the EM-34 and resistivity survey line
5. The ground is fairly level, appears undisturbed, and is covered with grass.
This area was not an intended target in the original site survey layout.
EM-34 and resistivity profiling surveys indicated anomalous conditions at the
end of line 5. This prompted a more intensive survey with EM-31 and magnetics
to determine the characteristics of the anomaly, its extent, and its origin.
Conflicting reports concerning this site were received from the Phelps
Collins ANGB personnel who indicated nothing should be in the area and from the
64
-------
Michigan Department of Natural Resources that indicated the possibility of
drums buried in the area.
Site Objectives
Objectives at this site were to determine the characteristics, the extent,
and the origin of an anomaly located near the runway. Methods used on this
site were EM-31 and magnetics. EM-31 was used to determine the extent and the
characteristics of the anomaly. Magnetics were used to detect and to map any
ferrous objects that may have been buried there. EM-34 and resistivity pro-
filing had been used at the southwestern portion of the area and were instru-
mental in detecting the anomalous conditions of the area.
Survey Results
EM-31 results from the horizontal dipoles mode are presented in Figure 28.
This figure shows the high values and the absolute difference between the two
values. The measurements were made perpendicular and parallel to the survey
lines. The anomaly of interest is the linear feature which extends from left
to right at the top of the figure.
Measurements were made at 16-feet (5-meter) spacings along 4 parallel
lines 49 feet (15 meters) apart. Four measurements were made at each station.
Results from the horizontal dipole measurements are presented. Results from
the vertical dipole mode did not show the anomaly as strongly when the high
values were plotted, and did not show the anomaly at all when the difference
65
-------
PHELPS COLLINS
RUNWAY AREA
X
APPROXIMATE SCALE
FEET
0 20 40 80
1 L
i
1 0
METERS
20
EM3 1 H
Absolute difference
between readings
Contour Interval =
5 millimho/meter
EM31 H
High readings,
conductivity in
millimhos/meter
Contour Interval =
5 millimhos/meter
-I EM31 and
Magnetometer
Data Location
Lines
Figure 28. EM-31 and magnetometer survey lines and contoured EM-31-H
results from the runway.
66
-------
values were plotted. This indicates a shallow source since the depth of
penetration of the EM-31-H is shallower than that of the EM-31-V.
Contoured maps of results and field observations both indicated a N20W
trend in the anomaly. Very high values were observed in the field, and the
instrument was rotated above those points to obtain minimum and maximum values.
Maximum values trended in the N20W direction.
One other feature on the plot of high readings is a series of apparent
anomalies with a north-south trend. The contours are 5 mmhos/m which is a
value that is very low and is not anomalous for the area. This apparent trend is
due to topographic effects caused by the small hills at that location.
Magnetics showed only one small anomaly which was concentrated at one
point on line 14. The total magnetic field and the magnetic field gradient
gave identical results: one high value on line 14 with background values at
all remaining points on line 14 and the other lines. The maximum total field
value was about 140 gammas, while the maximum field gradient reading was -80
gammas. Neither of these is very high nor as extreme as the values at the drum
disposal area. This indicates that a small ferrous object is buried in the
vicinity of the anomaly.
A profile along line 14 of the total magnetic field values with 58.900
ganmias background magnetic field subtracted is shown in Figure 29. All but one
value are in the -5 to -20 gammas ranye, with a general declining trend in
values probably indicating instrument drift. The single anomaly of 139 gammas
is toward the east end of the line.
-------
Figure 29. Profile along line 14 of magnetic data (total field with
58.900 garrenas background subtracted).
68
-------
One minor anomaly (-40 gammas) was observed in the total field at the west
end of line 15. It indicated where the magnetometer detected the survey vehicle,
which had been parked approximately 100 meters from the survey area.
Discussion of Results
This site was briefly revisited in August 1985. At that time, it was
found that the anomalous area is underlain with 26 high-intensity power lines
(4,800 to 7,200 volts). The lines had not been mapped, and their location was
not known until the field work had been completed and the data had been interpreted.
EM-31 results indicated a linear anomaly along a N20W trend. This aligned
exactly with two manhole covers which house end points of the power lines. The
EM-31 results also indicated a shallow source. The power cables were approxi-
mately 3 feet (1 meter) deep at each end point, and which corresponds to the
same depth at the anomaly since the land surface within the survey area and
over the power lines was fairly flat.
Magnetics results indicated a small magnetic anomaly on line 14 just west
of the EM-31 power line anomaly. Magnetics results tend to indicate the approxi-
mate location of buried objects, often with some offset from the actual location.
The anomaly shown by the magnetics survey is probably related to the power
lines which pass through the area. The amplitude of the anomaly suggests a
small ferrous object (e.g., a splice box) and does not indicate drum burial in
the area.
69
-------
POL FUEL STORAGE AREA
Site Description
This area is an existing fuel storage area in the southwestern part of the
study area. It contains above-ground fuel storage tanks and refueling units
for the jet fuel (JP-4) which is used at the Base. The area is partially paved
and partially covered with grass. There are storage tanks in the grassy areas
which have bemied soil around them to contain product in the event of a spill.
In May 1983, an estimated 500 to 800 gallons of JP-4 overflowed from one
of the tanks. Base personnel have estimated that nearly 90 percent of the fuel
was recovered within minutes of the spill. Soil sampling after the spill
cleanup showed no evidence of contaminant migration. Two ground-water monitoring
wells were installed downgradient to monitor any future contaminant migration.
Site Objectives
Objectives at the site were to determine if any contamination exists in
the spill area and, if so, to determine its extent. The method used to fulfill
these objectives was SGA. No geophysical methods were used because of the
numerous cultural structures in the area.
Survey Results
Two soil gas samples were collected in the vicinity of the fuel spill. One
70
-------
was within the berm into which the fuel spilled, and the other was outside and
downgradient of the spill area between the berm and the ground-water monitoring
wells. Soil gas samples were collected from a depth of about 1.5 feet (0.5 meters)
(instead of the 5-foot (1.5 meter) depth elsewhere) because the water table was
so shallow (at approximately 2 feet (0.6 meters) deep). This was discovered
when the soil gas probe encountered water at that depth.
Samples were analyzed for constituents normally found in JP-4. Neither
sample showed any evidence of contamination by any of the volatile organic
chemicals analyzed at this Base.
Discussion of Results
SGA was used at the POL fuel storage area to determine if any contamination
exists because of a fuel spill at the site. The results gave no evidence of
contami nation.
PLANNED POL AREA NO. 1
Site Description
Planned POL area no. 1 is located in the south-central part of the ANGB
It was used from 1947 through 1972 by the Alpena County Road Commission for
vehicle maintenance. It had been intended as the location of a new POL fuel
storage facility until the area was found to be contaminated. The area is
71
-------
currently covered with grass and gravel and has trees to the east and south and
a road to the north.
In August 1985, a contractor for the Phelps Collins ANGB performed chemical
analyses on soil and ground-water samples which they had collected during drill-
ing operations. They drilled seven wells to the water table and twenty-eight
auger borings 6 feet (1.8 meters) in depth. Results of analyses showed the
area to be contaminated with purgeable halocarbons (i.e., 1,1-dichloroethane,
methylene chloride, 1,1,2-trichloroethane, and trichloroethene), volatile
aromatic hydrocarbons (i.e., toluene, xylenes, ethylbenzene, and benzene), and
oil and grease. The areal extent of the contamination could not be determined
from samples taken during drilling since the contamination extends beyond the
perimeter of the area sampled.
Site Objectives
The objectives at the site were to determine the contaminant identities,
the extent of contamination, and the concentration of contamination in the ground
water. The results were also intended to be compared with those obtained by
the ANGB contractor. Methods used at this site were SGA and water sampling.
SGA was intended to determine contaminant identities and their extent in the
soil gas. Water sampling and analysis was intended to determine contaminant
identities and concentrations. The water samples were analyzed using the SGA
gas chromatographs immediately after collection. This is not recognized as a
standard EPA analytical procedure.
72
-------
Survey Results
SGA detected methylene chloride, carbon tetrachloride, and 1,1,1-
trichloroethane (TCA) in and around this area. The only contaminant detected
within PUL area no. 1 was methylene chloride. All other chemicals for which
the soil gas was analyzed were below detection limits at all points within the
area. Carbon tetrachloride, TCA, and methylene chloride were detected northeast
of the area.
Figure 30 shows the methylene chloride distribution in this area and in
the adjacent area to the northeast. Concentrations of 5 and 0.3 mg/m3 methylene
chloride were detected at points 25 and 21, respectively. Tracer Research
Corporation also included point 43, with 0.5 mg/m3, in a methylene chloride
plume containing all three points. It is more probable that point 43 is
separate since (1) samples taken between point 25 and 43 were below detection
limits, (2) conditions for detecting contamination in ground water were optimal
with well-sorted sands down to the water table, and (3) another contaminant
from a different probable source was detected at point 43.
Carbon tetrachloride, TCA, and methylene chloride were detected northeast
of the POL area no. 1. Figures 31 and 32 show the carbon tetrachloride and the
TCA distributions in this area. Concentrations of 1 and 0.002 mg/m3 TCA were
detected at points 43 and 41, respectively. Concentrations of 0.004, 0.003,
0.001, and 0.001 mg/m3 carbon tetrachloride were detected at points 27, 26, 44,
and 28, respectively.
73
-------
FIRE STATION
42
(BD)
1
28
(BD)
IP
ill
44
(BD)
64
[BD)
\
~ 4 5
(BD)
»57
(BD)
J
63
(BD)
46
(BD)
30
(BD)
"56
(BD)
43
(.3)
-.l->
(BD)
LEGEND
SOIL GAS SAMPLING POINT #43
METHYLENE CHLORIDE CONCENTRATION
IN SOIL GAS OR GROUNDWATER lug/L)
METHYLENE CHLORIDE CONTOUR
IN SOIL GAS
BELOW DETECTION LIMITS
(approximately 0.02 /ig/L)
FEET
250
-i r-
50 100
METERS
500
J
r
150
~ W-5 GROUND-WATER SAMPLING POINT W-5
Figure 30. Methylene chloride distribution near planned POL area no. 1.
74
-------
43
LEGEND
SOIL GAS SAMPLING POINT #43
(.004) CARBON TETRACHLORIDE CONCENTRATION
IN SOIL GAS OR GROUNDWATER (Mg/L)
^01n» CARBON TETRACHLORIDE CONTOUR
IN SOIL GAS
(BD) BELOW DETECTION LIMITS
(approximately 0.00008 Mg/L)
FEET
250
i
-r
50 100
METERS
500
J
r
150
~ W-5 GROUND-WATER SAMPLING POINT W-5
Figure 31. Carbon tetrachloride distribution near planned POL area no. 1.
75
-------
-FIRE STATION
2 8
(BD)
Bp
i 45
(.0006)
46
(BD)
~ 57
(BD)
*30
(BD)
"56
(BD)
CONTROL
65 TOWER
(BD)
40
(BD)
«l
63
(BD)
39
(BD)
36
(BD)
LEGEND
43 SOIL GAS SAMPLING POINT #43
(.002) 1,1,1-TRICHLOROETHANE CONCENTRATION
IN SOIL GAS OR GROUNDWATER (pg/L)
1.1,1-TRICHLOROETHANE CONCENTRATION
CONTOUR IN SOIL GAS
(BD) BELOW DETECTION LIMITS
(approximately 0.00009 fig/L)
FEET
0 250 500
1 1 1 1 r1
0 50 100 150
METERS
~ W-5 GROUND-WATER SAMPLING POINT W-5
Figure 32. TLA distribution near planned POL area no. 1.
76
-------
The carbon tetrachloride plume appears to have originated at the site of a
former aircraft hangar. It is oriented in a plume and was detected at sample
point 27, where it was necessary to drill through a concrete pad to obtain the
sample. Past experience has shown that volatile chemicals tend to be trapped
and to concentrate below concrete pads.
The highest TCA concentration (in this area) was detected at point 43.
This is also the location where methylene chloride was detected outside of the
POL area no. 1. The relatively high concentration of TCA with such rapid
attenuation with distance from point 43 probably indicates a localized source.
A reasonable assumption is that the methylene chloride detected at this location
originated at the same localized source.
Water sampling detected 5000 and 4000 pg/1 methylene chloride in well
nos. 5 and 24, respectively, with concentrations of all other compounds at or
below detection limits. The ANGB contractor did not detect any methylene
chloride (with a detection limit of 0.001 mg/1) but did detect 0.003 mg/1
toluene and 0.033 mg/1 xylenes in well no. 5 and 0.069 mg/1 xylenes in well no.
24. The Tracer Research Corporation did not analyze any samples for xylenes or
the sample from well no. 5 for toluene; therefore those values could not be
compared.
SGA sample points 21 and 23 were within 15 feet of well nos. 5 and 24,
respectively. Soil gas results were at or below detection levels with the
exception of 0.3 mg/m3 methylene chloride at point 21. The large concentration
77
-------
difference between soil gas and water sampling results may be a result of the
relatively high aqueous solubility of methylene chloride.
Discussion of Results
Results from SGA and water sampling indicate three contaminated areas near
the planned POL area no. 1. Two areas to the northeast of the area are local-
ized and contaminant concentrations are low. One area is a north-south trending
carbon tetrachloride plume which appears to have originated at the site of a
former aircraft hangar. The other area is a TCA and methylene chloride plume
which appears to have originated at some localized source near sampling point
43.
The third area of contamination is within the planned POL area. The ANGB
contractor's water sampling results showed the entire area to be contaminated.
SGA detected only methylene chloride at only one location. Methylene chloride
is not as easily detected in soil gas as other volatile organic chemicals
because of its relatively high aqueous solubility. Methylene chloride was
detected in high concentrations in both water samples collected and analyzed
while SGA work was being performed. SGA should have detected the other con-
taminants. The absence of the other contaminants in the SGA results may be due
to the differing sample collection depths. Water sample results from the two
sampling events differed in both contaminant identities and concentrations.
78
-------
PLANNED PUL AREA NO. 2
Site Description
The planned POL area no. 2 is located in the east-central part of the ANGB.
It was the second planned location for the new POL fuel storage facility.
Most of the area is heavily forested. It is surrounded and crossed by roads,
and it contains a number of buildings.
This area was selected for study when POL area no. 1 was found to be
contaminated by the ANGB contractor. ANGB personnel wished to find any indica-
tion of contamination problems at the new facility location prior to financing
an expensive program to assure that it was not contaminated.
Site Objectives
The objectives at this site were to determine if and to what degree the site
was contaminated. SGA was used to fulfill these objectives.
Survey Results
Based on SGA results, the planned POL area no. 2 appears to have minor, if
any, contamination. All volatile organic chemicals are below detection limits
or are at concentrations resulting from atmospheric sources.
79
-------
Just west of the area, PCE, toluene, and TIE were detected near a vehicle
maintenance building currently in use. Figure 33 shows the PCE, TCE, and
toluene distributions. The first nirober indicates the PCE concentraiion. The
second and third numbers indicate TCE and toluene, respectively. The highest
PCE concentration is at the location nearest the vehicle maintenance building,
and it decreases with distance from that building. TCE and toluene are isolated
spots of contamination. The highest TCE concentration (0.2 yg/l) is at point
88, and a minimal level of TCE (0.0009 yg/1) was detected at point 90 which is
the only location where toluene (0.2 ug/1) was detected.
Discussion of Results
Planned POL area no. 2 appears to have minor, if any, contamination.
However, it should be noted that according to SGA, the planned POL area no. 1
had only minor contamination. In contrast, soil borings in the same areas by
a different contractor showed significiant contamination. Further investiga-
tion will be necessary to confirm the extent of the contamination. West of the
area, minor PCE, TCE, and toluene contamination exist near the vehicle main-
tenance building. Characteristics of these contaminant concentrations indicate
that the PCE originates near the vehicle maintenance building, and the TCE and
toluene are isolated spots of contamination possibly from localized spills.
80
-------
90*
(.0009/. 0009/. 2)
88 <
(.009/.2)
62
85
(.0006)
83 ,
(.02)'
84
(.0002)
VEHICLE
MAINTENANCE
AREA
LEGEND
43 SOIL GAS SAMPLING POINT #43 FEET
0 250 500
(.09/.01/.2) PCE/TCE/TOLUENE CONCENTRATION
IN SOIL GAS
I i 1i r1
0 50 100 150
METERS
Figure 33. PCE, TCE, and toluene concentrations in soil gas (above
minimum detection limit) at planned POL area no. 2.
81
-------
CHAPTER 8
STUDY AREA GEOLOGY
OBJECTIVES
Very little local geological or hydrological background information exists.
Resistivity profiling and sounding techniques, EM-34, and seismic refraction
were, therefore, performed across the study area to yain a better understanding
of the site geology. In so doing, possible contaminant transport paths and
other anomalous or contaminated areas might also be indicated.
SURVEY RESULTS
Resistivity
Resistivity profiling and soundings were performed across the study area
with closer measurement intervals near the fire training area and the conductive
area. Ten profiles were constructed from pole-dipole measurements made using
33 feet (10-meter) dipoles. The geology at the conductive area appears to have
a significant influence on conductive plume characteristics. Therefore, results
pertaining to the geology of that area are presented in Chapter 6.
Two-dimensionally modeled results of a portion of the line 4 pole-dipole
82
-------
data (near the fire training area) are shown in Figure 34. Unfortunately, it
is difficult to determine the uniqueness of solutions obtained through forward
modeling. Still, very little lateral change is observed from stations 0 to
200 in the geoelectric section. The unsaturated zone resistivities are extremely
high (probably in excess of 10,000 ohm-m), and the water table appears to be
about 16 feet (5 meters) deep; the bedrock appears to be about 66 feet (20
meters) deep. A rather abrupt change in bedrock in the saturated zone, and in
the unsaturated zone resistivities occurs near station 220. The 4000 and 2000
ohm-m blocks probably indicate a decrease in depth to water rather than distinct
changes in subsurface geology. A local bedrock high appears to coincide with
the change in bedrock resistivity.
Line 1 shows resistivity changes similar to those of line 4. (Pole-dipole
data and pseudosections can be found in Appendix B of this report.) A line
connecting the line 4 bedrock high with that of line 1 gives the presumed
strike (surface contact) of the beds, N40W, which agrees with those results
obtained near the conductive area and which is approximately that found in the
formation sequence near Alpena as shown in Ehlers and Kesling (1970). A transi-
tion in saturated zone resistivities from 350 to 170 ohm-m going south along
line 1 is most likely caused by the increase in ground-water salinity near the
conductive area. A bedrock high appears to limit the flow of ground water
perpendicular to the strike of the bedding; this would explain the change in
depth to ground water.
Still on line 1, a highly resistive bedrock type is probably a relatively
dense, low porosity limestone with little shale content. The less resistive
83
-------
LINE 4 2-D MODELLING RESULTS
WEST
260 250 240 230 220 210 200 190 180
OH 1 1 1 1 1 1 1 1
4000
2000
14000
170
350
20-
1300
5000
30-1-
Figure 34. Geoelectric section derived from two-dimensional modeling of
pole-dipole data from line 4. Values are in ohm-meters.
84
-------
bedrock which lies to the southwest could be either a much higher porosity
limestone or a limestone containing a significant number of shale zones. A
shallow resistive anomaly which occurs between stations 340 and 420 is a terrain
effect from crossing over a hill and does not indicate a change in geology.
EM-34
The EM-34 survey did not produce data which aided in determining the
geology of the area. Although most of the study area was covered with dense
forest and underbrush, a considerable number of cultural obstructions either
prevented readings from being taken or interferred with the signal obtained.
Deer fences and power lines were nunerous on the Base, and the close grid of
data needed for useful interpretation could not be obtained. The apparent
resistivities obtained through the EM-34 survey were substantially lower than
those indicated by the resistivity method. The most likely explanation for
this is the difference between the measurement techniques.
Seismic Refraction
Seismic refraction measurements were performed on thirteen lines located
along sounding lines SI through S5 and along both S7 and S14. These were
performed primarily to aid in interpretation of resistivity data. The resistivity
measurements were unable to distinguish between a conductive pi lane in the
ground water and a possible clay lens near well no. 3.
Depth to the water table was determined for most of the lines. Depth to
85
-------
bedrock was determined for a few locations. The results and an interpretation
of those results are shown in Table 4. The time versus distance (T-x) graphs
from which these results were determined are in Appendix G.
86
-------
TABLE 4. SUMMARY OF SEISMIC REFRACTION RESULTS
Average Average
Line Layer Velocity* Depthj Elevationt Interpretation
1 1 1,070 0 690 Unsaturated
2 5,750 19 671 Saturated
2 1 1,140 0 682 Unsaturated
2 5,930 14 668 Saturated
3 1 1,250 681 Unsaturated
2 5,450 10 671 Saturated
4 UNRESOLVED
5 1 1,210 0 683 Unsaturated
2 5,430 9 674 Saturated
7 1 1,150 682 Unsaturated
2 5,150 13 669 Saturated
8 1 1,150 0 682 Unsaturated
2 5,350 15 667 Saturated
9 1 990 0 683 Unsaturated
2 4,820 22 661 Saturated or
CI ay
10 1 1,120 0 689 Unsaturated
2 5,420 17 672 Saturated
11 1 1,080 0 680 Unsaturated
2 1,690 9 671 Saturated
12 1 1,080 0 681 Unsaturated
2 5,130 12 669 Saturated
3 18,920 53 628 Bedrock
13 1 1,120 0 682 Unsaturated
2 10,610 21 661 Bedrock
14 1 940 0 684 Unsaturated
2 12,580 11 673 Bedrock
*In ft/sec (rounded to nearest 10 ft/sec)
fin feet
87
-------
DISCUSSION OF RESULTS
Resistivity profiling and sounding, EM-34, and seismic refraction were
performed across the study area to better determine the local geology of the
site. A tight grid of reliable data could not be obtained with the EM-34
because of cultural interferences and instrument limitations. Resistivity
profiling and sounding, used in conjunction with seismic refraction to constrain
layer thicknesses, gave a better indication of bedding on the study area.
Near the conductive area, a limestone-to-shale bedrock transition was found
running in a northwest-southeast direction (Figure 22). Lines la, 2, and 3
appear to be completely underlain by shale, and line 23 appears to be completely
underlain by limestone. A bedrock low appears to be associated with the con-
ductive area. Limestone bedrock continues northward toward the fire training
area.
A distinct change in bedrock lithology with an N40W strike occurs west of
the sinkhole. The change appears to be from a more highly resistive bedrock
type, probably a relatively dense, low porosity limestone with little shale
content, to a less resistive bedrock, probably a much higher porosity limestone
or one containing a significant nimber of shale zones. The goephysical data
alone are insufficient to determine whether the change is caused by an increase
in porosity, which implies an increase in permeabil ity, or by an increase in
shale content, which implies a decrease in permeability. An increase in porosity
better explains some of the ground-water level changes as observed in the
geophysical data, but this is speculative.
88
-------
CHAPTER 9
CONCLUSIONS AND RECOMMENDATIONS
Surveys consisting of resistivity profiling and sounding, self potential,
electromagnetic induction, magnetics, seismic refraction, soil-gas analysis,
and water sampling were performed at the Phelps Collins ANGB. Intensive surveys
were performed in areas of special interest. Those areas consisted of the drum
disposal area, the fire training area, the conductive area, the "runway" area,
the POL fuel storage area, and the planned POL fuel storage areas 1 and 2.
CONCLUSIONS
The general geology of the study area appears to be gently dipping beds
with an N40W strike (surface contact). The north part of the site is underlain
by what appears to be a relatively dense, low porosity limestone with little
shale content. West of the sinkhole, the lithology changes to a less resistive
bedrock, which is probably a much higher porosity limestone or one containing a
significant number of shale zones. A limestone-to-shale bedrock transition was
found north of the water plant near the conductive area.
Pole-dipole data depict an elongated plume of high salinity water originat-
ing southwest of well no. 2 in the southern end of the study area. Data show
89
-------
the plume area to be underlain by a shale layer with a probable bedrock high
southeast of the well field. If one assumes that the strike of the bedrock
lows and highs follows that of the bedding, N40W, then high salinity water
presumably channels from the old salt storage area into the well area.
The uppermost limestone which forms tne bedrock beneath the fire training
area appears to be of fairly low permeability. Therefore, the actual influx of
ground water from the upper aquifer is probably relatively low even though the
sinkhole provides a hydro!ogical connection between the surface aquifer and a
deeper aquifer of lower hydrostatic head. This should not be construed to
imply that the low water influx is insufficient to draw down any contamination.
The large gradient creates a rather localized, but large, self potential anomaly.
Depth to bedrock appears to be about 66 feet (20 meters) deep adjacent to
the sinkhole and about 18 feet (5.5 meters) deep approximately 820 feet (250
meters) northwest of the sinkhole. The water table is at or below the top of
bedrock at the latter location.
High levels of toluene and benzene, with lower concentrations of methylene
chloride, trichoroethylene, and PCE contaminate the soil below and near the
fire training area. A plume of contaminants appears to have migrated northeast
from the area and away from the sinkhole, with the highest concentrations being
close to the center of the fire ring. Vertical profile concentrations indicate
the soil surface to be the major source of the soil contamination in the fire
training area.
90
-------
In the drum disposal area, surveys suggest three areas of dense ferrous
objects, probably representative of the alleged buried drums. Considerable
demolition fill with less ferrous material content is indicated over a greater
area. The area between the drum disposal area and the adjacent sewage lagoon
has elevated levels of carbon tetrachloride; however, survey results do not
indicate which may be the contaminant source.
No evidence of contamination caused by a recent spill was found at the
existing POL fuel storage facility. Planned POL area no. 2 showed similar
results. However, south of the planned POL area no. 2, minor PCE, TCE, and
toluene contamination appears to originate from the vehicle maintenance build-
ings, and other isolated spots of TCE and toluene were detected.
Planned POL area no. 1 and areas northeast of it are contaminated with
methylene chloride, TCA, and carbon tetrachloride. The two areas to the north-
east consist of a carbon tetrachloride plume and a TCA and methylene chloride
"spot." Methylene chloride also contaminates the planned POL area no. 1.
The runway area is underlain by 26 high-intensity power cables at a depth
of approximately 1-meter and in an N20W trend.
The source of water well no. 3 contamination was not determined. No
contamination was found adjacent to the well. The geological conditions are
ideal for detection of contamination by the method used; therefore, one con-
clusion, which is not necessarily the only possibility, is that the contami-
nation is a result of a local spill at the well site.
91
-------
RECOMMENDATIONS
The following are recommendations based upon the results obtained through
the surveys performed at the Phelps Collins ANGB:
1. Water sampling could be done at the edge of the river near the drum
disposal area to determine if the chemicals are contaminating the
river, to determine the concentrations involved and, possibly, to
determine the location of the source.
2. The drum disposal area could be excavated to better determine drum
conditions and their content. This would provide true determination of
possible contaminants and of the hazards involved in addition to
determining the actual conditions at the site. It may also be more
cost effective than installing monitor wells to determine contaminant
extent. The drums are not very deep (less than 15 feet), and the area
is not very large.
3. Should more information be desired concerning sinkhole effects on sur-
rounding ground-water flow, an expanded self potential survey to
encompass the area to the south of the sinkhole could be used to
expand the data base currently in existence.
92
-------
REFERENCES
Black, T. J. 1983. Tectonics, Structure, arid Karst in Northern Lower Michigan.
Michigan Basin Geological Society 1980 Field Conference. 33 pp.
Corry, C. E. 1985. Spontaneous Polarization Associated with Porphyry Sulfide
Mineralization. Geophysics, 50, pp. 1020-1034.
Dakhnov, V. N. 1962. Geophysical Well Logging, Quarterly of the Colorado
School of Mines, Vol. 57, No. 2.
Ehlers, G. M., and R. V. Kesling. 1970. Devonian Strata of Alpena and Presque
Isle Counties, Michigan. Guide book for field trips: Michigan Basin
Geological Society and Geological Society of America (North-central Section
Meeting).
Gilkeson, R. H., P. C. Heigold, and D. E. Laymon. 1986. Practical Application
of Theoretical Models to Magnetometer Surveys on Hazardous Waste Disposal
Sites--A Case History. Ground Water Monitoring Review, winter, pp.
54-61.
Hazardous Materials Technical Center. 1985. Installation Restoration Program
Records Search for Phelps Collins Air National Guard Base, Alpena, Michigan.
National Technical Information Service, Virginia.
LaBrecque, D. J., D. D. Weber, and R. B. Evans. 1984. Numerical Modeling of
Geophysical Data Over Contaminant Plumes in Hazardous Wastes and Environ-
mental Emergencies. HMCRI, Washington, D.C. pp. 117-122.
Keller, G. V., and F. C. Frischknecht. 1966. Electrical Methods in Geophysical
Prospecting. Pergamon Press, New York. 517 pp.
McNeill, T. D. 1-980. Electromagnetic Terrain Conductivity Measurements at Low
Induction Numbers. Geonics Technical Note TN-6.
Miller, Lynn. 1978. Hydrogeological Evaluation for A1pena County Department of
Public Works, Ann Arbor, Michigan, Project C262781.
Mooney, H. M. 1984. Handbook of Engineering Geophysics, Volune 1: Seismic.
Bison Instruments, Inc. Minneapolis, Minnesota.
Scott, R. S. 1970. Engineering Company, Inc., Topographic Map, Alpena,
Michigan.
93
-------
Sill, W. R. 1983. Self-Potential Modeling from Primary Flows. Geophysics,
48, 76-86.
Wildenmuth, R. et al., 1924. Soil Survey of Alpena County, Michigan. Bureau
of Chemistry and Soils, U.S. Government Printing Office. 30 pp.
94
------- |