Bulletin of the

Association of Engineering Geologists

Vol. X XV II. N o. I. 1 990
pp. 37-50

Magnetometer and Gradiometer Surveys for Detection

of Underground Storage Tanks
CHARLES M. SCHLINGER
Department of Geology and Geophysics, University of Utah,
Salt Lake City, UT 84112

ABSTRACT

In recent years there has been a surge of interest in methods for rapid and reliable detection and location of underground storage
tanks and other cultural features related to hazardous substances in the subsurface. In the United States much of the motivation
comes from recent environmental protection legislation that regulates underground storage tanks, including both existing and
new installations. U.S. regulatory matters aside, ground-water contamination is a problem that knows no national borders;
remediation of sites where hazardous substances can invade or have invaded ground-water supplies is a global concern.
Detection and location of underground steel storage tanks can be readily accomplished using magnetometer and magnetic
gradiometer surveys, which arc a passive variety of remote sensing. This paper presents investigations at two sites at Hill Air
Force Base, northern Utah. In each case, magnetometer and gradiometer data have proven to be valuable for assessing the
possibilities of existence and location of buried underground storage tanks. Relevant magnetic-field principles are reviewed and
methods of data acquisition, reduction, analysis and interpretation are described.

IN TRO DUCT ION the purposes of ground water protection it is imperative

for both the public and private sector to locate
existing tanks, evaluate their condition, and if
There are numerous sites, under both private and necessary, remove or replace them. New regula-
public jurisdiction, throughout the world where haz- tions from the United States Government (Code of
ardous chemical materials arc thought to exist at Federal Regulations, 1988) stipulate conditions for
depth in the soil, however, the existence and specific construction and condition of underground storage
locations of these materials arc in fact not at all well tanks used for substances regulated by the U.S. Fed-
known. Of pressing concern in the United States is eral government.
the integrity of the subsurface containers of such
material; more specifically, the location and eval- -
An easily-used and interpretable method for rapid
uation of underground storage tanks (UST) and re- location of underground storage tanks is needed.
medi ati on o f l eaki ng t anks . So me o f t he und er - Magnetic surveys fill that need. Magnetic methods
ground tanks arc in use; others arc not. Both old have a history of application in mineral, geothermal,
and newer tanks present in the subsurface pose an and hydrocarbon exploration, archeology, and a va-
environmental problem in the form of hazardous riety of other areas. Here I review the applicati on
substances leaking into ground-water supplies. For of magnetometer and magnetic gradiometer surveys

[37]
38 BULLETIN OF THE ASSOCIATION OF ENGINEERING GEOLOGISTS

to the location of lost or imprecisely located un - such as a steel underground storage tank, and in the
derground steel storage tanks. The gradiometer, an same instance, from the presence of a nonmagnetic
instrument that is an adaptation of the conventional void in the subsurface, such as fiberglass storage
magnetometer, gives the gradient o f the magnetic tanks, if the surrounding soil materials are magnetic.
field. The gradient is especially useful for detecting The presence of a lateral variation in magnetic
objects buried at shallow depth (the gradient is the properties due to an object or void in the
quantity measured by magnetic locators used in land subsurface gives rise to a lateral variation in the
surveying). In addition to the application discussed magnetic field at the surface of the earth, above
in this paper, magnetic surveys have broad general the object. The variations in the magnetic field
application in passive surface searches for buried arise either because the object has a large magnetic
cultural objects, or searches for areas of prior human field of its own that adds to the background
disturbance. magnetic field, or, alternatively, in the case of a
The study presented in this paper was developed void at depth, because the absence of alluvial
for several reasons. In the first place, the Environ- material in the void gives rise to a local reduction in
mental Management Directorate at Hill Air Force the magnetic field. Note that the fluid in an
Base, northern Utah, responsible for the sites dis - underground storage tank does not directly give
cussed here, wished to see whether or not magnetic rise to a magnetic signal; it is the absence of
methods have an y demonstrabl e utilit y for envi - alluvium or the presence of highly magnetic ma-
ronmental and engineering site investigations. Sec- terial that gives rise to the signal that we seek to
ondly, the study was conducted to . investigate ad- measure at the surface. Values of the magnetic field
vantages and disadvantages of magnetometers over above or below expected background values are
magnetic gradiometers for these sorts of investiga- kn own as anomalous valu es. Collectivel y th ese
tions, and to highlight problems that might be ad- anomalous values define the spatial variations in
dressed by future research involving magnetic sur- the field that we measure, and constitute magnetic
veys of high-resolution and high -data -density. It is anomalies. Mapping magnetic anomalies on the sur-
worth noting that the application of magnetic sur- face allows us to infer the presence or absence of
veys described herein was not to definitively identify magnetic material in the subsurface.
underground storage tanks using magnetic methods.
The successful completion of magnetic surveys for
The purpose was to narrow down the range of sites
any site investigation requires that a number of dis-
for excavation, rather than pursue a course of ran-
tinct operations be carried out at each site. The first
dom excavation, or worse yet, a course of no action,
entails magnetic field measurements along profiles
due to a lack of information.
(or a grid) in the field area. Both accurate and precise
The following results are from a recent investi - measurements of the magnetic field strength are re-
gation of underground storage tanks at two sites at quired. Field strength is the magnitude of the geo -
Hill Air Force Base. Hill Field, as it is known, lies magnetic field, and therefore is a scalar quantity; it
on the Quaternary Weber Canyon Delta Formation, is commonly referred to as the 'total field.' The total
which consists of interbedded silt, sand, gravel and field is the most commonly and easily measured
clay lenses. quantity in surface magnetic surveys. Proton preces-
sion instrumentation is commonly used for such
measurements. In addition, instruments arc avail-
PRINCIPLES able that can simultaneously provide measurements
both of the field strength and of the vertical com -
For the purposes of an engineering or environ - ponent of its spatial gradient (the 'gradient').
mental surface site investigation, the objective of a
magnetic survey is to detect, by means of surface The earth's magnetic field varies not only with
measurements, variations in the magnetic proper- spatial position but with time also; consequently,
ties, or magnetization, of the subsurface. These the second operation of importance is measurement
variations in subsurface magnetization commonly of the time (diurnal) variation of the geomagnetic
arise due to geological structure, such as a pegmatite field at a fixed point at the site. The time variation,
vein or basaltic dike i n granite, a fault in once quantitatively documented, can be factored
bedrock, or hydrothermal alteration. They can into the data reduct ion procedure. The locat ion
also arise from prior human disturbance of where diurnal variation is established is often re-
alluvium or soil, from the presence of magnetic ferred to as a base station. Measurements at the base
objects in the subsurface. station made at time intervals from 5 to 30 minutes
 SCHLINGER— DETECTION OF UNDERGROUND STORAGE TANKS 39

are commonly acceptable, depending on the rate of fields in free space can be referred to in terms of
the diurnal variation. Importantly, one must also the induction, B, or the field int ensit y H . Th a t is ,
record the times at which data arc acquired by the i n free sp ace = B and H arc vector fields, with
other "roving" magnetometer(s) used to establish magnitude B and H. In the cgs cmu system B has
the spatial variation of the magnetic field. units of Gauss (G) and H has units of Oersted (Oe).
Keeping tally of the geomagnetic diurnal varia - Dimensionally, the units of these two quantities
tions brings up an important consideration. Electric ar e equivalent. For example, the geomagnetic field
power lines can create problems in magnetic surveys has an average magnitude at the earth's surface of
because the current flow gives rise to an alternating about 0.5 Oe, or 0.5 G. For practical reasons,
magnetic field, which interferes with sensor operation— workers in geophysics use a unit known as the
especially problematic for proton precession gamma (γ). 1 γ = 10 - 50e = 10 - 5 G . The
magnetometers. Because of this it is difficult and magnetometers commonly used in applied geo-
often impossible to carry out a magnetic survey in physics have sensitivities of 0.1 to 1.0 γ. Anomalies
the vicinity of electrical power lines. The problem in the geomagnetic field commonly range from 10's
is not limited to high-tension AC and DC lines; to 1,000's of γ, depending on the depth and size of
innocuous-look i ng rural li nes can be a source of the source and its intensity of magnetization.
grief. While it is not reasonable to offer a safe
Turning to the SI (System Internationale), applied
distance that can be us ed in a gen eral situation,
geophysicists commonly use B, rather than H , to
repeatability of the measurements is usuall y a
describe magnetic fields in free space. Contrary to
sure sign that power lines arc not a problem. In
the cgs cmu system, in the SI, B is not the same as
areas where power lines arc present, fluxgate
H in free space; neither do they have the same units.
magnetometers (discussed below) offer a decided
The unit of B in the SI is the Tesla; units of 10 -9
advantage in that the sensor is not overwhelmed
Tesla are used in practice. These units arc called
by 60 Hz alternating current
nanoTesla, or simply nT. Fortunately, a 1 nT field
The third important st ep is assigning all obser - is equivalent to a 1 γfield. Since the geophysical
vations unique locations in space. Accurate and community is moving towards exclusive use of SI
precise measurements of the magnetic field and magnetic units, their use is increasingly common.
gradient must be spatially located. From an All field strength values arc reported in nT; all spa-
operational point of view this is essential for tial gradients of the field arc reported in nT/m.
producing reliable contour maps, or for
comparison of individual profiles that cross an Instrumentation
area of interest. From an interpre tational and
applications point of view, if we wish to actually As mentio ned above, proton precession magne-
locate and recover a tank or some other source tometers arc commonly used in geophysical appli-
buried in the subsurface, the accuracy of our cations. The principles of operation of these devices
measurement locations becomes important. In arc discussed in some detail by Telford and others
terms of practice, the effort required to locate (1976), Griffiths and King (1981), Dobrin and Savit
observations with an error of 0.2-0.3 m is minimal. (1988). and Robinson and Coruh (1988). The proton
Achieving this end entails some l and surveyin g, precession magnetometer is based on a transducer
which can be accomplished by means of a variety that converts the earth's field strength into an al -
of procedures, varying in complexity from ternating voltage, which has a frequency propor -
measuring with a tape, to using a total-station tional to the field strength. From a classical physics
electronic theodolite with electronic distance meter point of view the working of a proton precession
(EDM). magnetometer can be understood as follows. Within
Magnetic Quantities and Units the sensor, a relatively large magnetic field produced
by electric current in a coil aligns the nuclear mag-
Before discussing the magnetic surveys acquired netic moments of hydrogen nuclei (protons).present
at our field sites, a brief review of magnetic quan - in a hydrocarbon -rich fluid (e.g., white gas). The
tities and units will be of use. When working with current is turned off and an induced emf (electro -
magnetic fields in free space, i.e., above the ground motive force) is generated within the same coil due
surface, we need to distinguish between cgs (centimeter- to Larmor precession by the magnetic moments of
gram-second) and mks (meter-kilogram-second) or SI protons. The frequency of precession and conse -
units. In the cgs system of electromagnetic units (cmu), quently the frequency of the induced emf is pro-
used in applied geophysics for many years, magnetic
 40 BULLETIN OF THE ASSOCIATION OF ENGINEERING GEOLOGISTS

portional to the earth's field (about which the mag- The deeper the source (e.g., a storage tank) the more
netic moments arc "precessing") strength. closely does the calculated gradient approximate the
In addition to proton precession instruments there arc a actual gra dient.
number of other instruments that can be used for
magnetic surveys. In efforts to accurately record the Characteristic Signals
spatial variations of the field strength or gradient, From magnetic field theory (Grant and West,
continuous-reading vertical component fluxgate 1965) the magnetic field due either to a point (di -
magnetometers and gradiometers (Clark, 1986), offer pole) source, or a three -dimensional (3D) finite vol-
an alternative to the discrete sampling inherent in the ume of magnetized material, decays in proportion to
proton-precession magnetometer (and its more sensitive -3
r as we move away from the source; r is the
and more expensive cousin, the optically-pumped separation between the source and the magnetometer.
magnetometer). As mentioned above, the fluxgate sensor The gradient of the field, on the other hand, decays
is insensitive to 60 Hz "noise" associated with power -4
in proportion to r . By means of Fourier transform
lines. Overhauser effect magnetometers (Dobrin and Savit, it is possible to show that a signal proportional to r -4
1988), based on the principle of nuclear magnetic (the gradient of the field) has more power at higher
resonance (NMR), are a v a i l a b l e f o r h i gh -p r e c i s i o n spatial frequencies, relative to a signal proportional
( ~0 . 0 0 1 n T = 1 picoTesla) high-sampling (~ 10 to r -3 (the field itself). Consequently, the magnetic
samples per second) applications, however, at this time gradient signal due to a given 3D source is more
such instruments are built to customer specification and limited in spatial extent, compared to the field itself.
are used primarily for military applications. Given its This will be evident in the magnetometer and
precision and sampling rate, the Overhauser -effect gradiometer survey data dis cussed below.
magne tometer may have great future potential in The field strength and gradient of an ideal source at
geophysical applications. While the continuous reading middle magnetic latitudes, near the magnetic
nature of fluxgate magnetometers gives them an ad- equator, and at high -latitudes (near the magnetic
vantage over proton precession and opticall y- pol e) are given in Figure 1. An additional consid-
pumped instruments, the mechanical and electronic eration from the r -dependence of each quantity is
calibration of f luxgate magnetometers and gradi - that the gradient decays much faster than the field as
ometers is much more critical, because each fluxgate we move away from the source. Therefore, the
sensor does not give an absolute reading, but has a deeper a given 3D source, the less manifestation it
continuously adjustable baseline—a problem for will have in gradient measurements as compared to
gradient measurements, in which the readings of two measurements of the magnetic field. Both the gra -
carefully aligned sensors must be differenced. In spite of dient measurements and the field measurements
these difficulties, fluxgate gradiometers, while not in have their merits, depending on the source depth and
wide use have proven advantage over other magnetometers extent at depth, and the variety of sources pres ent at a
in some circumstances.
site. F inally, whereas it is possible in principle, using
The spatial gradient of the magnetic field is ob- Fourier analysis, to obtain the field from the gradient and
tained by using two magnetometers in tandem. The most vice versa, for the purposes of our application it is easier to
common configuration has one magnetometer vertically measure and record both simultaneously.
above the other, with a separation ranging from 0.5 to 1
m. The vertical component of the spatial gradient, or METHODS
simply the gradient, is obtained by differencing two Magnetic surveys at the two sites at Hill Air Force Base
simultaneous measurements of B and dividing by the were conducted on the 10th and 11th of October,
sensor separation. Clearly, this is an approximation of 1988. In addition to measuring the magnetic field on
the gradient, due to the finite separation of the sensors. the surface at these sites, we measured the rate of
For example, given a point dipole source (which is change of the field with elevation—the vertical
equivalent to a uniformly magnetized spherical magnetic gradient.
distribution of a magnetic medium) at mid-latitude,
The magnetic and land survey data were acquired in
buried at 2 m depth, this approximation, obtained with about 12 hr, spread over two days. This was in spite
two sensors 1.75 and 2.25 m above the surface, is of the fact that the crew members were not familiar
within 2 percent of the actual vertical gradient at 2.0 m. with the equipment, which did not influ-
 SCHLINGER— DETECTION OF UNDERGROUND STORAGE TANKS 41

ence our results. A single experienced person and an precession magnetometer sensors and electronics package,
inexperienced assistant could easily conduct the magnetic and yields both total magnetic field strength and vertical
and land survey operations. A data logger on the magnetic field gradient measurements. The sensors arc
theodolite/EDM and an automated bas e station mounted vertically on a light-weight non-magnetic pole,
magnetometer (to record the time variation of the field) 2 to 3 m above the ground. The instrument used for our
would have eliminated the need for any manual data surveys was configured with a sensor spacing of 0.5 m. The
entry into the computer . D ata reduction, checking, measurements of the field strength and its vertical
analysis and plotting took a day's time, and could be gradient, along with the time of measurement are recorded
done in half the time by automating and in instrument memory.
concatenating the separate steps into one. On the other
A Geometries model 816 proton precession mag-
hand, stepping through the process and checking the data
netometer was used for tracking the diurnal variation, with
at each stage has its benefits.
measurements made manually about every five minutes.
Acquisition of Magnetic Data The magnetometers used for the project arc factory
calibrated although we did check them against one another
There were no problematic power lines in the vicinity of
to make sure that their readings of the field at a specific
our sites, and consequently we used proton precession total but arbitrary point in space were in agreement. At the end
field magnetometers. An EDA Omni Plus of each day's survey, the magnetometer/gradiometer was
magnetometer/gradiometer was used as the “roving” connected to a PC-type computer, into which its data
magnetometer. This instrument combines two proton- were transferred.
42 BUL LETIN OF THE AS SOCIATIO N OF EN GIN EERING G EO LOGISTS

day's work (data for Site 1 and Site 2 were acquired on

separate days) are shown in Figure 2. The maximum
variations are generally less than about 30 nT. This
magnitude of variation, seen by both the base station and
the roving magnetometers, is small compared to the
observed spatial variations, which are on the order of
100's to 1,000's of nT, nonetheless, each day's magnetic
observations were corrected by removing these diurnal
variations, both positive and negative. Linear interpolation
was used to estimate the variation at times intermediate
to the observation times (every 5 minutes). Cubic spline
interpolation can also be used, however, an unconstrained
application of splines can cause interpolation problems
due to oscillations of the interpolating cubic
polynomial(s). In applications where the anomalies
in field strength are on the order of a few hundred or few
tens of nT, sampling of the diurnal variations needs to be
done more frequently, e.g., every minute. Diurnal
Location of Measurements
corrections are not applied to the gradiometer data
Each measurement was located by means of land- because each of the two magnetometer sensors used for
surveying methods. For the sites described here, our calculating the gradient see essentially the same diurnal
magnetic readings were obtained along parallel or nearly- variation. This underscores an obvious advantage of
parallel lines laid out on the ground. The estimated gradient measurements—the diurnal variation need not be
precision for locations is ~0.1 m. The endpoints of each established.
line were located by surveying with a total station Once diurnal corrections were applied to the magnetic
theodolite with EDM, and the locations of equally-spaced field observations, anomalies were calculated by
intermediate points of measurement along each line were
subtracting a value appropriate for the background
located by interpolation. Locations of cultural features field 2t each site. The background value at each site was
such as lamp posts, boundary or cadestral monuments,
determined in a purely qualitative fashion by visual
building corners, etc., were also determined.
inspection of contour maps of the field strength, and was
Additionally, the locations of known magnetic objects,
taken as the average value of the field intensity in areas
such as road signs, parked trucks, trailers, and other cultural
of the site where the field showed minimal spatial
objects located on the site or adjacent to it were surveyed. variability. These values are: 54,000 nT for Site 1, area A;
If simple square or rectangular areas are selected for
54,450 nT for Site 1, area B; 54,400 nT for Site 2.
investigation, a surveying scheme that locates only two
Removing the background value from observations at
corners of the grid, and takes advantage of a grid laid out
each site yields magnetic anomalies, which must be
with cord would considerably simplify the land surveying
interpreted. A similar procedure could be applied in the
operation and subsequent survey data reduction.
case of the gradiometer data, but in areas with large
After entry into a computer the survey data were gradients (Sites 1 and 2) this is unnecessary. Finally, data
reduced using software previously developed for files of x-y-field strength or x-y-gradient were prepared
other projects. All x-y locations arc cast in arbitrary local for each site. These data were then gridded and contoured.
x-y coordinate systems for each field site. No nearby It is worthwhile to keep in mind that gridding and
control points were available for easy merger of these contouring arc themselves filtering operations that can
local coordinate systems with Utah State Plane or either degrade or enhance the signals present in the raw
Universal Transverse Mercator coordinate systems. numerical data.

Presentation of Data
Magnetometer Data Reduction
The locations of measurements of magnetic field and
Diurnal variations of the geomagnetic field as recorded gradient for Areas A and B of Site I are given in Figure
manually with the base station magnetometer for each 3, along with the locations of several ref-
 SCHLINGER — DETECTION OF UNDERGROUND STORAGE TANKS 43

gradient maps (Figures 4B, 5B, and 7B). For example, in

erence points. Figures 4A and 5A are contour maps of Figure 4B, the contour lines show more curvature near the
anomalies in the magnetic field strength for these Areas A lines along which data (locations marked by dots) were
and B. Figures 4B and 5B are contour maps of the vertical acquired. From a logistical point of view, it is
gradient of the field strength in these same areas. Figure 6 unreasonable to acquire a high-density two-dimensional
shows the locations of magnetic field and gradient data set, because of time considerations and because it is
observations for Site 2, along with locations of superfluous. Instead, we acquire data in profiles, with a
reference points. Figure 7A is a contour map of small sample spacing along the profiles and a larger sample
anomalies in the magnetic field strength for Site 2 spacing between the profiles (Figures 3 and 6). If one
and Figure 7B is a contour map of the gradient at this has an idea of the strike direction of the object(s) of interest
site. then the profiles can be aligned at right angles to the
strike. It is worthwhile to remember that a number of
DISCUSSION AND INTERPRETATION minor features in the contour maps of the data are a
manifestation of 1) discrete rather than continuous
A detailed discussion of anomalies in the field strength
and the vertical gradient of the field strength for each sampling, or 2) anisotropy in the spatial density of data.
survey will be illustrative of qualitative interpretational Spatial aliasing (Figure 8) along the profiles is not a
procedures for anomalies caused by underground problem because the sample spacing (spacing between
storage tanks and other cultural features. points of measurement) was small compared to the
One feature of most magnetic surveys, including the expected spatial wavelength of the magnetic field strength
ones discussed here, is that we sample at discrete points, and gradient signals due to a storage tank. Conceivably
rather than continuously. Furthermore, magnetic there is a potential for some aliasing as far as sampling
data arc commonly acquired in profile form and this can perpendicular to the profile lines is concerned, but
have an effect on the contour maps. The effect is experience tells us that this is not a serious problem.
readily apparent for data sets that exhibit large When looking for the relatively high-frequency (short
variations over short distances on the ground, e.g., the spatial wavelength)
44 BULLETIN OF THE ASSOCIATION OF ENGINEERING GEOLOGISTS
46 BULLETIN OF THE ASSOCIATION OF ENGINEERING GEOLOGISTS
46 B U LLETI N OF TH E A S S OC I ATI ON OF EN GI N EER I N G G EO L OGI ST S

variations in the gradient, one should generally use a tank. In fact, results from magnetometer profiles across
smaller sampling interval than would be used for steel storage tanks of known location (not illustrated)
measurements of the magnetic field strength alone. indicate that anomalies of 3,000-5,000 nT can be expected
Overall then, the contour maps (Figures 4, 5, and 7) from tanks that hold 1,000-10,000 gallons (4,000- 40,000
offer quite good representations of the magnetic field and liters), buried a meter or so beneath the ground
its gradient at the earth's surface at each site. A site by site surface. The large anomaly in Area A is broad,
interpretation follows. which indicates relatively deep burial of a large tank
(alternatively, this could indicate numerous smaller
sources clustered together). The gradient data (Figure 4B)
Site 1 show two more-localized areas of positive gradient
The area of Site 1 is a paved parking lot with a (maximum of about 1,200 nT/m) that can be used to
maintained grassy area adjacent to it. With reference to estimate the location of what may be the ends of the
Figure 3, Area A covers the parking lot and Area B tank.
covers a portion of the grassy area. There is thought to be a Magnetometer and gradiometer data for Area B of
steel fuel oil storage tank of unknown size and location Site I (Figure 5) show much smaller anomalies in the
at the site. magnetic field and its gradient, compared to Area A. A
The large positive anomaly in magnetic field linear trend of small anomalies in the gradient, located at
strength at Area A of Site I (Figure 4A) is consistent with the top of the contour map in Figure 5B, are thought to be
a three-dimensional magnetic source in the subsurface. related to a buried 9-cm diameter welded-steel pipeline
The positive nature of the anomaly indicates that the (abandoned steam line), unrelated to the manhole
earth's magnetic field is stronger in this part of the area found in Area B (Figure 3). The manhole is for access to
than elsewhere. The amplitude of the anomaly, about a 15-cm diameter vitreous clay: pipeline, which is
3,500 nT, is very high, and is consistent with a magnetic presumably nonmagnetic. The manhole and its cast iron
iron or steel source—presumably an underground storage cover do not yield much of an anomaly at the ele-
SCH LINGER- DETECTION OF UNDERGROUND STORAGE TANKS 47
48 B ULL ETI N OF T HE AS SOC I AT I ON OF ENGI NE ER I NG GEO LOG I STS

of anomalies in the area, and, the occurrence of

small anomalies along a linear trend, suggest cultural
features other than a "generic" steel underground
storage tank in Area B. One of the difficulties of
interpreting anomalies due to cultural features at
many sites, including this one is that recordkeeping
has not always been given the priority that we would
like it to have had.

Site 2
The area of Site 2 is a gravel lot adjacent to a
utility building, which is indicated in outline on
Figure 6. There was thought to be a steel gasoline
storage tank of unknown size and location at the
vation of the magnetometer sensor(s)—about 2.5 m site.
above the ground. While I have not investigated this in Locations of data points and of reference points
much detail, from the available literature (Bo zorth, for the magnetic survey at Site 2 are given in Figure
1951; Brandes, 1983), it seems that cast iron is not 6. Anomalies in the magnetic field strength (Figure
that strongly magnetized, compared to heavy steel 7A) and the gradient (Figure 7B) need to be inter-
plate steel traditionally used for underground storage preted in the light of known cultural features (Figure
tank construction. It is not unlikely that the 6). When interpreting these data, one must keep in
combination of a large nonmagnetic void (the man- mind that selection of contour interval is a filtering
hole) and a magnetic disk of cast iron (the cover) process; e.g., an interval of 1,000 nT (Figure 7A)
yields not much of an anomaly 2.5 m above the will exclude isolated anomalies with amplitudes less
structure. There is no doubt that were one to repeat than about 1,000 nT. Looking first at the anomalies
the survey with the gradiometer near ground level, the in magnetic field strength (Figure 7A), one large cen-
cover would yield a substantial signal. This un- tral positive anomaly is clearly evident, with an am-
derscores an alternative method for doing these types of plitude of almost 6,000 nT —a likely signal from a
surveys: bring the magnetometer sensors down close large buried steel object, presumably an under-
to the ground when looking for weak signals (Clark, ground storage tank. To the left of this anomaly is a
1986), but be ready for extremely high magnetic field smaller negative anomaly. However, a check of
and gradient readings when crossing over objects such Figure 6 reveals that this anomaly is related to the
as manhole covers. northern edge of building 1141, and is therefore not
The main anomaly in magnetic field strength, lo- of interest for the purposes of this study. The contour
cated in the left-central region of Figure 5A, is a map of anomalies in the gradient (Figure 7B) shows a
negative anomaly (the field strength here is weaker large anomaly, with an amplitude of almost 4,000
than that in the surrounding area) of low magnitude, nT/m, at the same central location. This anomaly in
about 320 nT above background. The shape of this the gradient has more-limited area extent than the
anomaly is consistent with either a void in weakly anomaly in field strength and clearly marks the likely
magnetic soil (which describes the soil at the site location of the underground storage tank thought to
fairly well), or remnant magnetization. While the exist in the area. Interestingly enough, the gradient data
amplitude of the anomaly is rather low for the steel do not show a signal from the northern edge of
underground storage tank thought to possibly exist in building 1141, pointing out another advantage of the
the subsurface of this area, the anomaly is larger and gradiometer. The gradient data probably do not show
more localized than what one would expect for a this feature because the source within the building
fiberglass or unreinforced concrete tank. The negative was probably at the same vertical level as the
sign could indicate remnant magnetization of iron. The sensors, rather than in the ground. The line of small
spatial extent of the anomaly is limited, indicating anomalies on the left edge of the gradient map
either shallow depth of burial or small source (Figure 7B) arc thought to be related to a utility line
dimension. in the subsurface—possibly a steel water line
Collectively, the amplitude, negative sign of the connecting the fire hydrants marked in Figure 6. The
largest anomaly in the area, spatially-limited nature anomalies it the top of Figure 7A and
 SCHLINGER—DETECTION OF UNDERGROUND STORAGE TANKS 49

7B can be related to trailers and trucks parked at the edge of an anomaly using surface measurements, and make
of the site. an interpretation in terms of what the source actually
A few final comments on interpretation arc warranted. is, using some a priori knowledge of what the possible
In magnetic interpretation modeling commonly is used to sources are (as in the interpretations made in this
determine geometric and physical characteristics of the study), determination of source geometry (depth of
source(s). Reasonable objectives of modeling might be to burial, shape) is a problem with an inherently
estimate the depth of burial or to determine the nonunique solution. Working to limit the possible
amplitudes and shapes of anomalies that can be expected solutions is a worthwhile endeavor and reaching this
from various sources with simple geometry, or, the effects goal invariably entails bringing a suitable number of
of latitude (Figure 1). Interpretation of data by means of 'constraints to bear on the interpretation problem. In the
simple modeling entails either a qualitative or a case of underground storage tanks, a knowledge of tank
quantitative comparison of the observed data with the materials, sizes, depths of burial commonly used, and
magnetic field or gradient due to ideal objects, such as a known cultural features in the area (e.g., buried utility
cylindrical shell with a specific radius, thickness, length, lines) may well constrain the interpretation to the point
depth of burial and magnetization. Sophisticated modeling, where the family of solutions is manageable.
either in the forward or inverse sense (Telford et al., 1976;
Griffiths and King, 1981; Dobrin and Savit, 1988; and EX C A V A T I O N
Robinson and Coruh, 1988) commonly requires greater
expenditure of time, and may be of limited value for site Based on the magnetometer and gradiometer results
investigations of the type described here. This is (Figure 7) an excavation at Site 2 began on the 20th of
especially true for steel and cast iron sources, because the November 1989. Excavation was started at the center of
magnetization of objects such as fabricated steel large anomaly in the vertical gradient (Figure 7B).
Buried 1 m beneath the surface was a steel underground
underground storage tanks is likely to be heterogeneous,
and it may depend greatly on the level of remnant storage tank that contained approximately 42,000 liters
magnetization; point to point variations in the (11,000 gallons) of a mixture that was predominately
water, with a layer of oil on top. As of November
direction of magnetization, which are not known a
priori, cannot be reasonably established from 22 plans were being made for drainage and transfer
observations of the field or its gradient. of this liquid to a treatment facility, to be followed by
removal attic tank.
In contrast to steel tanks, fiberglass or unreinforced
concrete tanks can be expected to produce small signals of
CONCLUSIONS
only 10 to 20 nT (40,000 liter tank buried 1 m deep)
provided that the surrounding alluvium is weakly 1. Magnetometers and magnetic gradiometers
magnetic (susceptibility of 10-3 dimensionless SI units). If offer excellent potential for location of underground
detected, these signals can be readily modeled, because for storage tanks and other buried cultural features of
such tanks, the source of the magnetic anomaly is the interest for site investigations that focus on hazardous
void itself, and for this class of tanks the void has materials. The methods could be useful in other types
ideal geometry and magnetization for modeling (e.g., a of investigations as well. These magnetic surveys can
uniform cylinder, the magnetic field or gradient of which be applied even in areas where known cultural features
can be calculated). However, the detection of signals due arc abundant.
to fiberglass or concrete tanks would certainly require 2. Measurements of the magnetic field strength
high-precision work with careful analysis (including and its vertical gradient arc easily obtained with com-
filtering of the data) and interpretation. Contamination of mercially available instrumentation. Precise horizontal
the signal by cultural features present at the site may be location of measurements is critical if interpretable
problematic. In any case, a sensor near ground level, results arc desired.
closer to the source, will enhance the signal. 3. Measurements of magnetic field strength often
A final precaution centers on the non-uniqueness complement gradient measurements. Since both can be
inherent in magnetic interpretation. While one certainly acquired at the same time, with no additional effort, the
can evaluate the intensity, polarity and size added constraints on interpretation warrant the use of
both for many site investigations.