/ 33 j

CD,M. ._ F E D E R X'L C O R P O R A T I O N

January 27, 1992—'- *.--'-''"-'' "rlT" . •"" -L ..=-. ™~" . . '.-'-- •-. •- . - - . - . - . :

Ms. Lisa NicholsU.S. Environmental Protection"Agency841 Chestnut BuildingPhiladelphia, PA 1.9.107 -' -'- - - -

PROJECT:, . EPA'Contract Number: 68-W9-OOQ4 - -'

DOCUMENT: -. TES7-C03112-£P"-"CRFH

SUBJECT: - := Work Assignment C03112River Road Landfill . _Applicability of Geophysical Methods

Dear "Ms". Nichols: - - "-- - - - - - -

The purpose of this letter is to summarize ..our re.c.eixt -discussions regardingthe applicability of geophysical methods to locate .drums which have allegedlybeen disposed at the subject site. The information presented b.elow is basedon personal experience, discussions with individuals at CDM FEDERAL PROGRAMSCORPORATION (~FFC) experienced in the use of geophysical techniques, andconversations with several geophysiclsts employed with other firms. -

Under ideal circumstances, various geophysical methods can be utilized tolocate buried drums__with _a high degree of success. -::;However, the probabilityof success in using geophysics- at_the subject s,it_e is limited by severalfactors: •' - _ . _ . . . . - _ . . = - ------ • - • - ——- -•••—-

the high level of" interference that can be expected from othermetallic wastes disposed (such as refrigerators, stoves, etc.) in thelandfill;

- the method of drum disposal (random disposal vs. common location) isunknown; • "

the proximity of power" lines to the..disposal area;

- -the nature of the cover material;- '"--- —

/ • - - -the-potential ..depth of disposal of the drums which is unkrft g B 'S 7 9

992 Old Eagle School Rd., Suite 919 Wayrw, PA 19Q87 215 293-0450

COM FEDERAL PROGRAMS CORPORATION

Ms. NicholsPage 2

the various limitations of each candidate geophysical method (suchas the depth limitations of EMs and magnetometers (15 and 30 feet ;."„.respectively) , and the limited spatial coverage of GPR and itsinability to "see" below the first highly .conductive ..layer ).

In summary, geophysics can assist In locating drums disposed in a municipallandfill but the degree of success . is variable and difficult... , predict. Thegreater the probability that the drums are homogenous (electr.lcally andmagnetically) with the other burled materials, the lower the probability ofsuccess of locating the drums. It should be noted that positive confirmationof drum disposal will necessitate the excavation of test

Several experienced Individuals Indicated that, a preliminary "geophysicalsurvey would be valuable given the situation at the site. Such a survey wouldindicate the viability of a more detailed investigation and could be completedat a relatively low cost ($6,000 to $8,000). -* .-

For your Information, I have enclosed two papers that discuss geophysicalsurveys that have been used to locate burled drums . If you have any questionsregarding this submittal , please contact me at (215) 293-0450.

Sincerely,

CDM FEDERAL PROGRAMS CORPORATION

Bruce R. PlucaWork Assignment Manager

Enclosures

cc: Donna McGowan, TES VII Regional Project Officer, CERCIA Region IIIJean Wrlght, TES VII. Zone Project Officer (letter only}Constance V. Braun, FPC Program Manager

AR302880

Magnetic Surveying for Buried MetallicObjects

. • by Larry Barrows and Judith E. Rocdilo

Abstract .__.. ._.-... —..— ——- • ...field tests were conducted to determine representative total-intensity magnetic anomalies due lo the presence

of underground storage tanks and 55-gallon steel drums. Three different drums were suspended from a non-magnetictripod and the underlying field surveyed with each drum in an upright and a flipped plus rotated orientation. Atdrum-to-sensor separations of II feet, the anomalies had peak values of around 50 gammas and half-widths aboutequal to the drum-to-sensor, separation. Remanent and induced magnetizations were comparable; crushing one ofthe drums significantly reduced both. A profile over a single underground storage tank had a 1000-gamma anomaly,which was similar tothe modeled anomaly due to an infinitely long cylinder horizontally magnetized perpendicularto its axis. A profile over two adjacent tanks had a smooth 350-gamma single-peak anomaly even though modelsof two tanks produced dual-peaked anomalies. Demagnetization could explain why crushing a drum reduced itsinduced magnetization and why two adjacent tanks produced a single-peak anomaly.

A 40-acre abandoned landfill was surveyed on a 50- by 100-foot rectangular grid and alorig~scvcral detailedprofiles. The observed field had broad positive and negative anomalies that were similar to modeled anomalies dueto thickness variations in a layer of uniformly magnetized material. It was not comparable to the anomalies due toinduced magnetization in multiple, randomly located, randomly sized, independent spheres, suggesting tha; demagne-tization may have limited the effective susceptibility of the landfill material. A different 6-acre site survey conductedon a 10- by 10-foot grid was analyzed to determine the maximum station spacing and line separation that cou!d•have been used. Essentially, all of the anomalies at this site would have been resolved by a survey conducted on a20- by 20-foot grid and the larger anomalies would have been detected by a 50- by 50-foot grid.

Introduction __ - = _-=i--- of geophysical instruments including a total intensity- ;- 7" * 7- : rrm .... — ma ffetomeier. They found the magnetic detection limitMagnetic surveys have traditionally been used by for a single drum was 6 to II feet below the surface and

geologists to locate changes in rock type such, as might ' that the boundaries of a dump site containing steelbe associated with ore bodies, fault contacts, or igneous drums can be easily determined. Gilkesori and otherstatrustves. Another common application is determining (1986) describe a magnetic survey of a series of landfillthe probable depth to basement beneath sedimentary trenches that had been used to dispose of steel drums.rocfcs. For these applications, the principal geologic van- They found a distinctive pattern of magnetic highs overable is the distribution of ferromagnetic minerals, the trenches and lows over the inlertrench oorridors.mainly magnetite, within the earth. The theory and sur- Tfcey noted that these signals were similar to the calcu-vey procedures are described in a variety of references latcd magnetic anomalies due to infinitely long rectan-(e.g.. Grant and West 1965, S.E.G. 1966, Parasins 1975, gular bodies having dimensions comparable to theNettleton 1976,TeIford and others 1977, Robinson and. trenches and a magnetic susceptibility of k = 0.1. AsConih 1988). A particularly concise review of surveying explained later, a susceptibility of (his magnitude iswithportabIemagnetometcrsisgivcnbyBreiiier(1973). expected for a mass containing many disseminated

Magnetic surveys are used in hazardous waste site ferrous metal objects. Friscfaknecht and others (1985)investigations to locate 55-gallon drums, underground and Jachens and others (1986) describe fieffl $s§ gstorage tanks, buried pipes, and the edges of covered models that demonstrate the use of magnetic surveyslandfills. These applications usually involve shallow iron to locate covered abandoned well casings. This appZica-or steel objects, which influence the way the surveys lion is important because abandoned wells are potentialshould be conducted and interpreted. Tyagi and others pathways for the vertical migration of contaminated(1983) describe controlled field tests in which single ground water. They found strong positive anomaliesdrums and dusters of drums were, buried at various over the wells that closely resemble models of simpledepths. The test site was then surveyed with a variety magnetic dipolcs having the positive pole at the lop of204 Summer 1990 CWMR "~"

the casing and the negative pole at Its base. by its intensity, inclination (or angle 16 the horizontal),In this paper, the theory behind magnetic surveying and declination (or angle to geographic north). Breiner

isbricfiyrcvtewcdandthcficldtcstsconductedtodcicr- (1973) includes large-scale maps of these parametersmine the total-intensity magnetic anomalies due to 55- for the continental United Slates. Fabtno and otherstalloa steel drums and underground storage tanks arc (1979) give more detailed maps.t»rielK'de5cribcd.ThcanomaHcsduetothcthrcednims The geomagnetic field is not constant both in thetested had peak amplitudes of around 50 gammas and sense of diurnal variations of several tens of gammashalf-widths approximately equal to the II-foot separa- and occasional periods of rapid, irregular, transienttion between the drum and the magnetometer sensor. variations (magnetic storms). The diurnal variations canThe half-width is the distance between the two sides of be removed from survey data by drift corrections basedan aixxrtaly at intensities of one-half of its peak value. on cither regular base station tics or the record from aThese anomalies were similar to those due to isolated fixed base-station magnetometer. Magnetic storms varydipoles but in addition to the magnetization induced by in intensity and can make surveying impractical. Thethe earth's ambient Held, both remanent magnetization Space Environment Services Division of the Nationaland "demagnetization seemed to affect the signals. Oceanic and Atmospheric Administration providesRcmanem magnetization is a permanent magnetization recorded information on the current level of these flue-that B independent of the ambient Held. Detnagnetiza- tualions (telephone number (303) 497-3235) and a fore-[Ion B a limit on the strength of induced magnetization cast of the projected level for the next five days (tele-withtn art object imposed by the internal field due to phone number (303) 497-3171). It is normally desirablethe object itself. The surveyed anomaly due to one to know the condition of the earths field during eachunderground storage tank was similar to that due to an day of a field survey.infinixcfv long cylinder magnetized perpendicular to its . * , - , . , -. - * *. r j . 'V. j- ..' i Magnetic Anomaliesaxjs. However, the anomaly due to two adjacent tanks fealso resembled lh*t of a single body. Again, rcmanent FoUowmgTe!fordandothers(1976,P.llI,&iuationmagnetization and demagnetization are thought to 3*n>-lhe magnetic field at an external point <?0 due toaffect ihese signals. If the results described herein are a magnetized body can be expressed as:representative, then the strengths of anomalies due to — _ _ * ^ - » _ _» 1 —drums and tanks may depend more orftheir volume F(r° > =v J J J •v*r > * v (T-r*! dv(r>than OQ their metal content. Also, it may be difficult to volumeinfer from the shape of a magnetic anomaly the exact ^locadoa and nature of the causative body. Demagnetiza- where: £ is a position vector within the body. •cfon may also influence the magnetic signals from land- . is the gradient operator,fills coataming many metal objects. One site survey is" I?*- f*0 1 is the distance between the externaldescribed in which the total-intensity field resembled point and position within the body,that due to a uniformly magnetized layer of varying indicates the vector dot product, andhickness but did. not resembre that- due to an assem- M(r) isthe net magnetization per unit volume.ilcdge of magnetically independent objects. One impli- Th* net magnetization is the vector sum of inducedaiioa is that magnetic surveys may not be able to locate plus remanent magnetizations. Remanent magnetiza-MDceouations of metal objects, such as drums, within tion is a permanent magnetic moment per unit volumei UndfflL Another is that successful landfill surveys may and induced magnetization is temporary magnetization« conducted on a relatively coarse station .grid. The tnat disappears if the material is not in a magnetic field.lata from a second survey .conducted on a 10- by 10- Generally, the induced magnetization is parallel with .XK"square grid were analyzed to determine the max!- and proportional to the inducing field. Algorithms foryam station spacing that could have been used. For calculating the magnetic fields due to uniformly magne-ils particular ate, essentially an of the signal, would tized, simple geometric shapes are given in several texts*ve been resolved with stations on a 20- by 20-foot (e-g- Grant and West 1965, Nettletoo 1976.1elford andrid and the stronger anomalies would have been others 1976, Robinson and Coruh 1988).etcded on a 50-by 50-foot grid. . Magnetic fields are vectors and magnetometers mea-

sure some attribute of this vector field. Proton preces-aco«y sion magnetometers measure the mairimiim intensitybe Geomagnetic Environment (or total length of the vector) and flux gate magnetomc-Tbc signals in a magnetic survey are partially the tcrs measure the intensity in a particular direction (or

suit oC and strongly influenced by. the ambient mag- vector component). When used with dual sensors, mag-•Xx field of the earth. As a first approximation this netometers also measure the gradient of the attribute;omagnetfc Held resembles that due to a single axial * usually in the vertical direction.pole whose negative or south magnetic pole is toward The net magnetic field to which thec geographic north pole. The strength of this field responds is the vector sum of the field due *o localrice from 60,000 gammas near the poles, where it magnetized materials and the ambient field of the earth.inges vertically into the ground, to 25,000 gammas Figure 1 shows the total intensity field due to a simpleatr the equator where it parallels the earths surface, magnetic dipole, the ambient field of the earth, and theany particular region the ambient field is described total-intensity anomaly that would be detected during

a survey. In ihis case, the magnetization oj liejjbjeci ^ TOT«. IS—NSITY UACWETK: ANOMALYts parallel to ihe ambient field (induced magnetization).There is a magnetic low to the north of the center ofthe body and a larger high to the south.Effective Susceptibility A A T, „«-. - - x

Magnetic susceptibility, k, is the dimcnslomcsshpro- • _ > «*-=v*5-A v>portionality constant relating induced magnetization _ " rwithin a body to the inducing field. In general, the induc-ing field is the vector sum of both the earths ambientfield and the field due to the object itself. This feedbackis referred to as demagnetization and is expressed as areduction in the effective susceptibility of the object:

t**ma* .. _ _...,_ —--=*- • *

-• km4, is the material susceptibility, and X is thedemagnetization factor. Grant and West (1965) describeifoe physical basis for demagnetization and the deriva-tion of this relation.

Demagnetization factors are dependent on both theshape of the object and its orientation to the ambient5efd. For a sphere X = J/i IT, normal to ihe axis of the s i »cylinder X = 2 ir; and normal to a flat sheet X = 4 IT .-_-.- / / f(Strangway 1967).Figure2showstheresuItihgreIations" . . ' .berween effective and material susceptibility for these (hTI rectioi of he earth's°ambient field. The measured totalsample shapes and orientations. Note that for material magnetic intensity is ihe vector sum of the ambient field plussusceptibility less than about k = 0.05. the_effective and . the Held ijue to the body. ______________material susceptibilities are approximately equal. Most , ~~~

units have susceptibilities less than this, therefore,gnetization does not usually affect the interpreta-

tion of geologic surveys. However, ferrous metals havesusceptibilities of tens or hundreds, therefore, the effec-tive susceptibility of ferrous metal objects, like steeldrums, is limited by demagnetization to a few tenths. Ifthe ferrous metal content of a landfill is several percentof the volume, then the effective susceptibility of landfillmaterial would also be limited. In this case, local concen-trations of metal within the landfill would not beexpected to significantly increase the local effective sus- Uve «,««„!«* :o , ftvr teouis nuess of

. Of the natcrial (from Straayray 1967. p. 455).Demagnetization abo limits the applicability of the — ————————— " ———— —— •

algorithms used to calculate the magnetic effects of sim- field and a nearby base station was selected Surveypie models. These algorithms usually assume that mag- stations were at 3-foot intervals along north-south pnoetization is uniform throughout the material, a condi- spaced 6 feet apart (231 stations). In the center of thetioa not realized if the field due to the body itself is site a non-magnetic (PVC pipe) tripod was constructedirregular. For hazardous waste site investigations Acre from which the drums were suspended. The drums wereis a need to develop magnetic modeling techniques that 19 feet above ground level; therefore, with the 8-footaccommodate demagnetization phenomena. Unto this sensor height of the magnetometer, the signals wereis accomplished, magnetic models of ferrous metal similar to those from drums buried at 3 feet (Figure 3).objects (including the models in_this report) should be .- Running the surveys beneath, instead of over, theinterpreted cautiously. — objects , re versed the signals through an cast-west line.

- ThisresuHedinareversalofthepositionsoftheposiu'veField Tests peajc an(j rciated trough.55-Gallon Drums .- The site was first surveyed with an empty tripod to Q Q A The objectives of these tests were to establish the establish a baseline, which was removed from alpt s3-0 2 0 0 Jttgneticsignalofa55-gallonst«eldrumand!ocoinparc qucnt surveys. For each survey, the ends of the north-tins result with analytical models. A secondary objective south lines were first read and linearly drift-correctedm-as to determine the extent to which demagnetization 10 the base station and then the individual stations werelimits the effective susceptibility of a steel drum. read and linearly drift-corrected to the line ends. AU

A 60- by 60-foot test site was laid out in a flat empty data were relative to the first reading at the base station206 Summer 1990 GWMR

55 • Gafcxi Ooiminduced Total IntensityMaximum 51.3 gammasMinimum -7.2 gammas

io is So3

Figure 3. Test apparatus used to simulate the magaetkresponse of a buried 55-£*Ho« steel drum. The anomaly due totfce suspended drum Is similar to tkaf of * buried drum except * - * * . - , . . .chat the podfltw of the high aod low are reversed. f*uf? 4" T« n "" " ^ 'li0* -—————:.——————————2—————__—..———————————— m a single 55-gaUon drum. Tbe configured surface ts one-halfand all were gathered with a total intensity proton-proces- the sum of the anomaly due to a drum in its upright orieuta-sion magnetometer (Geometries Mode! 856). Figure 4 is <»on plus *hat due to the same drum in a nipped plus rotateda perspective diagram of a representative anomaly. The orientation. Survey stations *e« located at each of the grid

, . , ,._ . --. , . intersections.anomaly has a peak amplitude near 50 gammas, and has ———————————————.—————————————.s half-width about equal to the drum-to-sensor separation< U feet). It is less than*5 gammas at twice the sensor-to-drum separation.

Tests were run with three different drums. For eachdrum, the field beneath the drum was surveyed, thedrum was Hipped and rotated to reverse the directionof the rcmancnt magnetization, and the new field wassurveyed. Along the north-south center line, the averageof the two fields is attributable to induced magnetizationand one-half their difference is attributable to remanentmagnetization..For two of the three drums tested, theanomaly due to remanent magnetization was compara-ble to that due to induced magnetization. For the thirddrum, it was 40 percent as large.

The third drum was then crushed to a 1.1 cubic foot,drum-shaped mass. This crushed drum was surveyed in Figure SL North-south profiles beneath a suspended 5S-gaJloaboth its upright aad reversed orientations. Figure 5 dmm in its upright orientation and its nipped plus rotatedShows the observed data along the north-south central orientation. The IoHer-ampBt.de profiles are for the same

... . t -t ,,_ » j ^ j j ••*..(. drum after it bad been crusbed to a LI cubic foot dnun-pcofile for both the whole and crushed drums in botlj shaped mass. ^ ^of their orientations. In Its crushed configuration, the - - •drum showed very little rexnanant magnetization, pos-sibly because ttie magnetized sheet metal had beenfolded over on itself. The anomaly due to induced mag-netization was only 30 percent of that of the uncrusheddrum even though both configurations contained thesame steel.

Modeled profiles of the total intensity anomaly ofuniformly magnetized spheres were matched to theinduced-magnetization anoraaSes of both the whole and ___________crushed drums. The sphere volumes were identical to « —10 » a a / i ^ V w * S « n~their respective drums. For the whole drum, the mod- ' o-t-w 0*4 v_ydcd sphere had an effective susceptifofliry of k « 6.10, *"*"«**• J?1""'* "~!*5dae to"«od«rgro«»j*«-r t, t j j i, n 10 age tank. The modeled aao» Jj g that d«c to «a lofiahelyfor tne crUSHed drum fc 0.1S. long cyfioder wfth a boriziwtal aogaedntiM of 2750gmmm«Underground Storage Tanks ' per oiMctoot___________ &&-%€*&

Magnetic surveys are tied to locate underground ston magnetometer,storage tanks either for their removal or as an aid In The first example is a single tank on a narrow landpositioning boreholes in which fcafc detectors are to be spit extending from the south shore of Lake Mead, Ari-mstallcd. The following field tests demonstrate the zona. The tank had been part of a marine fuel dockcharacter of the associated total-intensity signals. Again, until the facility was destroyed by high water and aban-ihe data were fathered with a total-field proton-proces- doncd. There were no remaining buildings, power lines.

—— r-ir-r T*VVt

pipes or other soufccjs.;pf cultural noise. Mooeteo TOTAL INTENSITY — /*~\ ~*-EASTFigure 6 is a profile normal to the long axis of this * \

tank along with a matching model based on the actualtank diameter. The model has a uniform horizontal mag-netization of 2750 gammas per unit volume. This netmagnetization is the vccTor sum of induced plusremancm magnelizaiiqn and cannot be resolved withoutreorienting; Tfic~ tank. One simple possibility is an ~ £induced magnetization of 5500 gammas per unit volume «(k = O.I) plus an upward remancnt magnetization of f o-

oeseRvgoTOTAL INTENSITY -.,

gammas per unit volume. a— -100-1oThe second example is a profile over two adjacent *~ •«»-

tanks located 22 feet to one side of a large vehicle main-tenance garage. Figure 7 shows the observed data, an .assumed Hnear regional, which may be due to the garage.and the residual anomaly along with the profile due todie indicated model. In this case, the entire anomalv . _ - - y , 1~^...... . , , .... . . . ,*, figure 7. Toliil-mtenMiv anonuu due (o IHW underground stne-might be due to induced magnenzalion in a single small ag* tenkfc ^ ^ ^ rcso(vc (hc p ncc <)fCHXly that is considerably deeper than the actual tanks. <wo object* even ihuugh calculated modcfe ofiwo near-surfaceNote that the data did not resolve two tanks even though magneticaliy independent tank> produced dual-peakedgeometrically correct models of two magneticallyindependent tanks had dual-peaked anomalies. A possi-ble explanation is that the inducing field within eachrank is the sum of the earths ambient field, the fieldcue to the tank, andjhe field due to the adjacent tank.The tanks would then not be magnetically independentand the two-tank model would not apply.

Field Surveys.The Landfill :..__ = -•--_-

The first example is a survey of a 70-acre coveredlandfill in south-central Indiana. A 20-acre lake occupiesthe center of the site and a river flows along the northernand northeastern sides. The landfill had been used todispose of approximately 40,000 drums of chemicalwastes along with a vai |e y;o| domestic Jiiid| industrialrefuse. The survey was i etea to better define theUteral extent of the land ff'ahd. if possible, to locateclusters of drums.- 33:."

Two survey methods were used. One was a recon-naissance survey, with stations at 50-foot intervals alongGoes spaced 100 feet apart. The other was a series ofmore detailed north-south profiles with stations at 10-or 20-foot intervals. Both were conducted with GeoMet-rics Model 856 total-intensity magnetometers.

Figure 8 is a contour map of the total intensity datafrom the reconnaissance surveK&he dots are measure- _ _ _, _..*_,-_ =^ __Fifmc A. Mjujnrtjc loOu tuttuun *rera «*.•*.» •• i nm •*•meat stations and the contour interval is 1000 gammas. ««wfc.«nirtl lodUnm. The co**Mr fetcmrf fa 1000 gmoosAx this location, the ambient Held is 56 00 gammas and and Ihc dots *re stetioos ** «fei* 4mt* wen grtbered. Profile

measured values range from 52,600 to 64,500 gam- A-A' fa shown tn f vre «.mas. so the anomalous field ranges from 3600 to +8300 The continuity of the contoured field may be duegammas. The reconnaissance survey clearly showed to demagnetization limiting and homogenizing the effec-areas in the southeast corner and west-central side of tiye susceptibility of the landfill material. Figure 9 is a n cfbe site that are magnetically smooth and arc not north-south profile along line A-A. Stations are § Op3 Q 2 8 8 bdieved to contain buried debris. The data contoured foot intervals. Also shown is a ample model andfeod-Pqprisingly well, considering that drums and metallic cicd field configured to match the larger features in thedebris are exposed on the surface. At the relatively observed data. The model is an east-west trending in-cnarse 50- by 100-foot station spacing (necessitaled by finitejy jong polygon with a uniform susceptibility ofthe size of the site and limited field time) many "single k = 0.24 (the Talwanii algorithm, e.g.. Grant and Westpoint anomalies" and ambiguities in the contours were ___ 1955). The horizontal scale « as shown, but there is a"ticipalcd. " '_ __ .;__ _ _,_ ...-.- IOX vertical exaggeration in ihe model and iis o>«- "-—X!* Summer t<W» CWMtt

NORTH

sa.ooo

s 30Oci O£* -10O-

*--<oo-

j[00

so *e la *a w too

Figure 10. A total-intensity magnetic model of in asscmfalcdgcof randomly located, randomly sired spheres magnetized in the

__ _ direction of the earth1* field. The modeled field is dominatedby a few narrow peaks due to the shallowest objects,

nately positive and dominated by a few high-intensity1OX V«riicai e**g9*c»t*>" •* r f °m J

100 aoo 300 4io soo «oo r±o «oo n*mw pcafcs due to the shallowest objects. TheKH.AC* w*«i> observed field at the landfill had broad anomalies with

Figure 9. Observed and modeled magnetic total intensity over both positive and negative parts.a covered tandfilL The modeled anoma&es are doe to varia- ' •dons in (he wnOpmrtion of the landfill material. llie sludge Ponds———————.——-————————————————————— Hie next example is a survey of some abandonedthickness is only 5 feet. The important point is that the sludge ponds on a 6-acre site south of Houston, Texas.anomalies can be attributed to modest thickness, varia- The ponds had been used 10 dispose of broken-«!abs ofcions in the layer. Comparable results were obtained reinforced concrete and had then been covered withwith models having an irregular upper surface and a fiat earth. The survey objective was to locate areas where jbase. In contrast. Figure 10 shows the effect of induced the slabs had been dumped so they could be avoidedmagnetization in an assembledge of randomly Ideated, when drilling ground water monitoring wells. It wasrandomly sized spheres. This modeled Held is predomi- thought that the steel reinforcing bars in the concrete

~1

I!. Modeled magnetic t«Cal Intensity over some abandoned sludge ponds near Houston, Texas. The contours are at±20, ±40, ±SO, ±160, ±320, ±€40 gammas and stations were on a 10- by 10- foot «*««« erid.

would produce a detectable magnetic anomaly. ......This survey was conducted with an OMNI-IV tic-

line magnetometer system. Stations wcrcTestablished at<0-fooi intervals along lines 10 feet apart and both thestal intensity and its vertical gradient were recorded,Tbe OMNI-IV monitors the quality of each reading andthe data are reliable except in an 80-foot strip along theeastern side of the site where there is an overheadpower-line. In this area about 20 percent of the readings•were unreliable and were edited from the data.

Figure II is a contour map of the edited total inten-sity data. Values range from 49,150 to 50350 gammas.The ambient field at this location is 49,«00 gammas sotfae anomalies range from -650 to +550 gammas. The

•.-exponentially spaced contour intervals were used so thatboth subtle features in the relatively smooth areas anddie shapes of the larger anomalies are displayed. Thism?p along with a map of the vertical gradient sucessfullyidentified undisturbed areas in which the monitoringwells could be placed. ,# r : -

On Figure 11 there is a tendency for the magnetichighs to be flanked to the north by lows of comparable~ ... . A - ,.„. i_,jf:ii -„-,,«, ,t,« Arr« < ~.,..\A u,. Figure 12. Magnetic total in(en>n* aoonwlv due to m 80- liv 80-anrolitude. As in the landfill survey, the effect could be . * _ _ •* . . . . , - . • . . __ -^ . . . . . . bv 4-fool Ihicfc sfab of material **itfa a horizontal northerly

modeled as thickness variations in a continuous layer. magnetization of 1000 gammas per cubic foot. TUc conioiirHowever, at this site, enough of the slab dumps (and interval is 50 gammaxanomalies) are sufficiently isolated to suggest a differentinterpretation. Figure 12 shows the total intensity ano- o.os~ ~———" —:———"———rr —•——<~ v*-iomaly due to a horizontally magnetized slab at a depthof 4 feet. For horizontal magnetization at the latitude«ihe site, the magnetic highs and lows arc of compara-

: amplitude but for.steeply plunging magnetizationparallel'to the ambient field, the highs are significantlylarger than the lows. Demagnetization limits the effec-tive susceptibility perpendicular to the surface of a slabor to the axis of a bar. However, it docs not limit thesusceptibility parallel to a thin slab or along the axis ofa bar. Therefore, for flat-lying reinforced slabs, the hori-zontal component of induced magnetization is expectedto be larger than the vertical component, which is consis-tent with our model.

Selecting station spacing and line separation involvesa tradeoff between survey resolution and the amountof field work. If the distance between measurements istoo large, the data will be uncertain by an amount com- R TO tt Tm>-dim«t«wal torier transform of the totalparable to the amplitude of the narrower anomalies; intensity data In Fi&m 1L Tfce cMowed values we the mod-e*cflifthcmeasurementsarcrjrecise.(ThisspatJal-alias- oH of rt»e uanstonn after dKj *H WHXHJKM! by a oloc-poiniing phenomena is simflar to the temporal aliasing that ••* <n**rf)U "" «""»' inttml o to percent of (be maxtmamoccurs when a continuous time signal is digitized.) On i—— ..die other hand, if the distance between measurements and orientations. The graph axes are wave number or-is too small, the time and cost of the survey may be one-half the reciprocal wavelengths in the ENE-WSWprohibitive. For the survey over the sludge ponds, the and NNW-SSE directions of the survey grid. The coit-rdati vely short 10- by 10-foot grid was selected because tours are the moduli of the transform (an array of com-tfae nature of the signal was not known beforehand and plex numbers) after they had been filtered or averagedtfae surveyor wanted to detect all significant anomalies. with a nine-point unit matrix. The contoufl ij ejjvQ £887

i determine the maximum distance between sta- - . 10 percent of the peak filtered value.that would have adequately resolved the field Note that almost all of the amplitude spectra are at

variations, the Fourier transformation was used on the wavcnumbers less than 0.025 ft' (0.5/20 ft). This impliestotal intensity data.and then the resulting amplitude that most features would be adequately resolved by aspectra (Figure 13} was smoothed and contoured. This survey conducted on a 20-foot grid, assuming the datamap shows the relative amplitudes of the variations in were reliable. Some data redundancy is desirable andmagnetic total intensity as a function of their widths it is more efficient to make closely spaced readings along210 Summer 1990 GWAT R~

more widely separated lines than it is to make the same tometer surveys on hazardous waste disposal sites —number of readings on a square grid. An optimum sur- A case history. Ground Water Monitoring Review,ccy grid at this site might have stations at 10-foot inter- v. 6. n. 1, pp. 54-61.vals along lines no more than 30 feel apart. Contour Grant. ES. and G.F. West. 1965. Interpretationmaps constructed from alternate stations and lines (a in Applied Geophysics. McGraw-HHI. New0- by 20-foot grid) and from even- third line (a 10- by pp. 306-381.30-foot grid) resolved all of the anomalies on the total- Jachens. R.C.. M.W. Webring. and F.C Frischknecht.intensity map, 1986. Abandoned-well study in the Santa Clara Val-

Thc highest peaks on the amplitude spectra occur Icy. California. U-S- Geological Survey, Open-Fileat wave numbers near0.005 ftl (0.5/100 feet) and ampli- Report 86-350. 13p.tudcs arc generally less that 50 percent of the peak value Netllelon. L.L. 1976. Gravity and Magnetics in Oil Pros-at wave numbers greater than about 0.01 ft'1 (0-5/50 ft). peering. McGraxv-Hill. New York, pp. 305-426.This implies that the larger amplitude anomalies would Parasnis. D.S. 1986. Mining Geophysics. Elsevier Scien-have been detected by a survey on a 50-foot grid. A tific Publishing Co- Amsterdam, pp. 3-60.map constructed from every fifth line and station Robinson. E.S. and C- Coruh. 1988. Basic Explorationdetected all the major anomalous areas but did not Geophysics. JAViley & Sons. New York. pp. 333-444,resolve the shapes of the anomalies. Soc.of Exp.Geophysicist C*ds). 1967. Mining Geophys'_ /cs.v.II.Soc. of Exp.Geophvsicists.TuIsa, Oklahoma.Summary pp. 423-620.

Magnettcsurveyscanbeanirnportantpartofhazard- Strangway. DAV. 1967. Magnetic Characteristics ofous waste site investigations but the physical principles Rocks'. 1967. Soc. of Exp.Geophysicists, pp. 454-473.must be understood before the data are interpreted. In Telford. W.M.. L.P. Geldart. R.E. Shriff, and D.A. Keys.particular, these surveys often involve ferrous metals 1976. Applied Geophysics. Cambridge Universityandeffectivesusceptibilitycanbelimitedbydemagneti- Press. Cambridge, pp. 105-217.zation to a few tenths. The magnetic field variations Tyagi. S.. A,E. Lord Jr_ and R.M. Koerncr. 1983. Usewill then be due to the configuration of the magnetized " of a proton precession magnetometer to detect buriedmaterial rather than to local concentrations of metal. drums in sandy soil. Jour, of Hazardous Materials,

The, detectable anomaly due to an isolated steel,. Vt gt pp.'ll-23. ""drum has a width of about twice the distance betweenthe drum and the magnetometer sensor. A survey "to Biographical Sketchesreliably detect single drums would have to be conducted Larry Barrows received his M3. and Ph.D. degreeswith a station spacing and line separation less thaa this *« geophysics from the Colorado School of Mines inwidth. However, if a hazardous waste site contains suffi- 1973 and 1978, respectively. He has worked as a systemsdent disseminated metal for demagnetization to occur, engineer on Sky tab remote-sensing experiments and asthe stronger anomalies can have dimensions comparable an exploration gcophysitist in frontier areas of Alaska.10 the landfill cells. In this case, relatively coarse station Since 1979 he has specialized in geophysical surveyingspacings and line separations may be adequate. for ground water and hazardous waste site investiga-Acknowledgments . tions, first as the project geophysidst on a radioactive

. . waste disposal program and then as a research scientistThe landfill survey in south-central Indiana was con- r«^^t-rfc J?CM /« toss t.- s~:~~j c-. c. c-.j---*-~j. . . _. j~,~,. . t, ' . f ft f < /*" *nc (/.j. dTf\~ Mil j7oo nejvinea tzann csenceanaducted by Mike Gibbons and Aldo Mazzelta for the ~ - —- r * r r j n L x*r^.rfo e- • . i n . — - A T*. j .j Engineering Inc. and LaCoste and Romoere GravityU.S. Environmental Protection Agency.The drum study ..* — , ,*ont c- *— j * .* DJ ou *. .. .. _. T -.L - t_1 Meters Inc. (4807 Spuxtcood Springs Rd~ Slag. 2,was conduced by the authors, also with support from . rf -™. -I-™., !T __ , *_* j - V.». no V=OA T « j ~£ * j!T!_/ . t» Austin TX 7S759). C&rrau projects include using mt-thcU .EPA- The study of underground storage tanks _._„ */* »—a j *-* i._-J__*_ f TV *•*_ . _ -— • —u. »xt. o —™ crogravifr to detect solution conduits In karst terrainswaspartofaBoyScoutservjceprpiectbyMikeBarrows j * * • j«j * _r,*[. ij j J _ * j i _ » oxid using seismic provMdrou ta determine near-surf aceand the sludge pond survey was conducted .by Jerry , Jda/*

Robinson for the ENSR Corp. The contributions of all » j . » « « » . » ^, - j^ -j » j -- _ ^ « _i j Judith E. Rocchio ts the future air resource special-tnese indtvtauais ana agencies is grateluliy acxnowi- , »— *. . » , , — „.«»..— .. . iston the Stanislaus National Forest (1977 Greenleycugcd. ~ *^ Rd., Sonora, CA 9S37O). She will obtain the positionReferences upon completion of her J&A. degree In air resourceBreiner, S. 1973. Applications Manual for Portable Mag- management at Colorado State University Department

nctometcrs: GeoMetrics. Sunnyvale. California. of Natural Resources (1990). She war tflth LockheedFabiano, BB WJ. Jones, and M.W. Peddie. 1979. The Engineering Management Services Co., Las Vegas,

Magnetic Charts of the United States for Epoch 1975: Nevada, prior to attemdlngCSU where she was InvolvedUS. Geological Survey-Grcular 810,15 p. *n several environmental monitoring programs

Frischfcnecfat,EC.andP.V,Raab.l985Xocationofaban- authored the paper with DC. Barrows. From JftlVdoncd wdls with geophysical methods: KTIS Report 1986 Rocchlo was a project geologist for Cower OilPB 85-122638, National Technical Information Ser- Co. and Consolidation Coal Co Denver, Colorado.vice. Sorincfxeld. Vireinia, 48 p. She received herB.S. in geology (1981)from the Univer-

Gilkcson, R.H., P.C Heigold and D.E. Laymon. 1986. «"0' of Nevada, Las Vcgas.Practical application of theoretical models to magne-

Sumnirt- 1990 <7U'MK 21!

DETECTION OF BURIED DRUMS AND TANKSAT WASTE DISPOSAL SITES AND INDUSTRIAL PLANTS

By•* * John W. Fowler

Sarah M. HetzneckerAmin Ayubcha

ABSTRACT

In any comprehensive waste site evaluation, the delineation of poten-tial contaminant source areas such as buried underground storagetanks and buried.drums is very important. Magnetic surveying using aproton precession magnetometer and/or terrain conductivity surveyingusing, electromagnetic induction techniques represent rapid, effectivemethods of delineating anomalies which may result from buried ferro-magnetic materials or conductive wastes.

These techniques allow for rapid data collection over large areasand, coupled with the use of computers for data reduction, are verycost effective.

This paper compares magnetic surveying and terrain conductivity astools for locating areas of buried drums and tanks at waste disposalsites and industrial plants. Included in this discussion are: sur-vey design and field procedures, data reduction and processing, datainterpretation, and an illustrative case history.

This study attempts to circumvent some of the shortcomings due toinadequate survey layout and field procedures.

A discussion of the identification, of anomalous zones by examinationof contour maps and profiles is presented. In addition, a techniquefor estimating burial depths of magnetic bodies is also presented.

AR302889

INTRODUCTION

The accurate delineation of regions of buried steel drums or tanks hasbecome a common problem In the investigation of waste disposal and indus-trial plant sites. Both drums and tanks may serve as source areas forgroundwater contamination, and decisions regarding investigative andremedi al activities may be dependent upon the 1ocation of former drumdisposal areas or abandoned tanks.

The survey methodologies presented herein are particularly applicable toindustrial plant sites and industrial waste landfills. These sites usu-ally present particular challenges to the geophysicist due to potentialinterferences to data collection caused by cultural features or the char-acter of wastes contained within landfilled areas. Site characteristicssuch as topography, depth to groundwater, groundwater conductivity, sitegeology* and characteristics of potential contaminants must also be con-sidered during initial survey design.

The authors have found that an initial interview with current or formeremployees or personnel familiar with past plant or landfill operationsmay prove useful in survey planning. Information regarding former tanklocations may be available on engineering drawings of plant sites. How-ever, drawings and employee recollections may often be inaccurate due tochanges in land surface configuration, building demolition, or lack ofaccurate knowledge.

When exact location of buried objects is unknown, both magnetic surveyingand terrain conductivity surveying represent rapid, cost-effective meansof locating buried features. Both nondestructive techniques are accom-plished with lightweight, portable instrumentation and require a minimumof site preparation. A typical survey day may cover 3 to 5 acres of areadepending upon data point spacing and survey objectives. Generally, twodays of data processing and interpretation are required for each fieldday.

The objectives of the survey and type of subsurface body sought willdetermine the most applicable techniques. In some situations it may beadvisable to conduct both magnetic surveying and terrain conductivitysurveying over the same area in order to produce data sets that are com-plementary and discover unique information. Magnetic surveying is thepreferred technique if the target objects are ferromagnetic (iron orsteel) bodies. The magnetic survey interpretation method presentedherein also allows a depth estimate of magnetic sources.

Terrain conductivity will not discriminate between ferromagnetic materialand other conductive bodies; however, unlike magnetics it will locatenonferromagnetic conductors. Current instrumentation for electromagneticinduction has fixed exploratory depths, and depth estimates for bodieslocated are not easily made.

Following is a brief discussion of the theory of magnetics and electro-magnetics and a description of instrumentation with which the author^ are Q Qfamiliar. In later sections of this paper, survey design and datftftojQiZO Jsition techniques and an illustrative case study are presented.

MAGNETIC SURVEYING THEORY

The Earth's magnetic field resembles that of a uniformly polarizedsphere. The two poles of the sphere, are located near the geographicalnorth and south poles. The unit commonly used in magnetic field measure-ments is the gamma (1 gamma - 10~9 webers/m2). The intensity of themagnetic field varies, being twice as large at the poles than at theequator (60,000 and 30,000 gammas, respectively). The intensity of themagnetic field, in the vicinity of the northeastern United States isapproximately 55,000 gammas (Figure 1).

The objective of a magnetic survey is to identify spatial changes (anoma-lies) in the Earth's magnetic field. In a typical waste site or indus-trial plant survey, the changes range from a few gammas to severalthousand gammas^ These variations are normally caused by manmade objectscontaining ferromagnetic materials, and electrical power lines in proxim-ity to the magnetometer.

A magnetic survey entails conducting a series of measurements of the mag-netic field intensity. Measurements are taken at intervals along succes-sive parallel traverse lines, which together form a grid. Spatialchanges in the field are identified by two methods: examination oftwo-dimensional graphs of the magnetic field along traverse lines, andexamination of contour maps of the magnetic field intensity.

Two-dimensional graphs of total magnetic field intensity disturbances(anomalies) are highly variable in shape and amplitude. They are almostalways asymetrical due to the dipole nature of the field (i.e., they haveboth positive and negative components). A significant characteristic ofthe profile of a magnetic anomaly is its variation with depth; the deeperthe source, the larger the period (or the broader the anomaly) (Figure2). It is this property that allows the determination from profiles ofapproximate depth to the magnetic source without prior informationregarding the source. Anomaly shape and amplitude also may be affectedby the shape of the source, and by the orientation of the source in theEarth's magnetic field. As a result, anomalies sometimes appear complex,even from simple dipole sources.

MAGNETIC SURVEYING INSTRUMENTATION

Magnetic surveying for the study described herein was accomplishedthrough the use of a proton precession magnetometer. The proton preces-sion magnetometer uses the precession of spinning protons in a sample ofhydrocarbon fluid to measure the total magnetic field intensity. Brieflydescribed, the protons behave as small, spinning dipole magnets. Thespin axes of the protons are temporarily aligned by application of a

AR302891

Figure 1THE TOTAL INTENSITY OF THE EARTH'S MAGNETIC FIELD

-DEPTH

I*HLv —— ^^- " • • " "- • -1- •

^ DEPTH

— DEPTH

Figure 2 EFFECT OF DEPTH OF ON ANOMALY WIDTH

current in a coil of wire. When the current is removed, the spin axes ofthe protons precess about the direction of the Earth's magnetic field,similar to the way a spinning top will precess about the gravity field.The processing protons generate a small signal in the same coil used topolarize them. The frequency of this signal is precisely proportional tothe total magnetic field intensity at the point of measurement.

The authors are most familiar with a portable Model G856 magnetometer,manufactured by EG&G Geometries of Sunnyvale, California (Figure 3). Inthe mode in which it was utilized, the magnetometer has a sensitivitycapable of measuring the absolute value of the Earth's magnetic fieldwith an accuracy of 0.2 gammas. Therefore, with a total field intensityof 55,000 gammas, the sensitivity of the instrument makes it possible torecord variations in the Earth's magnetic field to an accuracy of twoparts in 550,000.

The instrument, which is portable and battery operated, has a digital LEDdisplay and a memory capable of storing 1,000 readings. The memory fea-ture allows the data to be transferred electronically to a computer fordata processing. ..... ... - - -

TERRAIN CONDUCTIVITY THEORY

Terrain conductivity surveying utilizes the application of inductiveelectromagnetic techniques for the direct measurement of apparent terrainconductivity. The term "apparent conductivity" is used because themeasured value is a surface and/or depth-weighted average of conductivitybeneath the measurement point. For convenience apparent terrain conduc-tivity is often referred to simply as terrain conductivity.

The surveying technique utilizes a small transmitter coil, which is situ-ated at the Earth's surface. An alternating current at an audio fre-quency is applied to the terminals of this coil. This current generatesan alternating magnetic field, which causes electrical currents to beinduced in the Earth. The induced currents in the Earth generate asecondary magnetic field, which is detected by a receiver coil locatednear the transmitter coll. Currents are not induced in the air, which iseffectively, infinitely resistive. When the energizing frequency is suf-ficiently low, the magnetic field generated is linearly proportional toground conductivity. The more conductive the ground, the larger thecurrent flow, and the larger the resultant secondary magnetic fieldmeasured. Terrain conductivity meters are instruments which measureresultant magnetic fields and are calibrated to read ground conductivitydirectly. The_ MKS^jnits of _conductivity are the mho (Siemen) per meteror millimho per meter.

AR302893

Figure 3PROTON PRECESSIONMAGNETOMETER

Figure 4 TERRAIN CONDUCTIVITY

Terrain conductivity is dependent upon the nature of the soil, subsurfaceporosity, hydraulic permeability, moisture content, depth to and type ofbedrock, and the concentration of dissolved electrolytes. The presenceof such objects as electric lines, wire, pipes, buildings, tanks, buriedmetal, fences, and foundations may also influence conductivity.

Most physical constituents of a soil are electrical insulators, which donot allow appreciable current flow through them. The medium which doesallow current flow is the relatively conductive interstitial moisture.Water conductivity is determined mainly by ionic 'content; therefore,variations in soil conductivity can be used to define areas where ground-w ater may contain a higher concentration of dissolved electrolytes.

In general, the conductivity of the Earth varies with depth. In a typi-cal vertical profile, the conductivity will initially increase with depthdue to increasing soil moisture becoming constant at the water table dueto saturation. If the bedrock underlying soils has a low porosity, theconductivity within this bedrock medium will decrease. In addition, anincreased clay content can affect conductivity because of a surface con-duction phenomenon which occurs in clay.

Since cultural, geological, and hydrogeological .factors can affect con-ductivity surveys, surveys intended to delineate a contaminant area orburied metallic objects must include a sufficient density of measurementsboth over the suspect region and beyond into the surrounding area, sothat the possible influence of any of the above-mentioned factors can bedetermined. Furthermore, these factors must be considered when the dataare interpreted.

The actual magnitude of conductivity values measured is less importantthan the trends and anomalies in the measurements. To be meaningful,survey results must be correlated with known subsurface conditions fromtest borings and wells.

A terrain conductivity survey consists of a series of measurements takenwithin a grid system or along profile lines. Spatial changes in terrainconductivity , are identified by two means: examination of conductivitycontour maps; or examination of two-dimensional graphs of conductivityvariations by distance, known as conductivity profiles.

TERRAIN CONDUCTIVITY INSTRUMENTATION

The terrain conductivity instruments used most frequently by the authorswere manufactured by Geonics, Limited of Mississauga, Ontario, Canada.Geonics manufactures several instrument types. The authors are familiarwith the EM 31 and EM 34-3 (Figure 4).

ftR302895

The EM 31 is a one-person portable unit- which has two coils separated bya fiberglass pole. Intercoil spacing is fixed at 3.7 meters (12 feet).The exploration depth commonly used with this instrument is 6 meters {20feet) below land surface.

The EH 34-3 is a two-person portable unit which has two coils connectedby flexible cables. Exploration depth for the EM 34-3 is controlled byintercoil spacing. The instrument is calibrated to allow several inter-coil spacings and several consequent exploration depths. Intercoilspacings of 10 meters (33 feet), 20 meters (66 feet)", and 40 meters (131feet) are available.

The EM 34-3 coils may be operated in a vertical configuration or may belaid flat on the ground surface during measurements. Coil configurationaffects exploration depth. When the coils are held vertically, thesecondary magnetic field is perpendicular to the plane of the receivercoll. This configuration is known as the horizontal dipole mode. Whenthe coils are held horizontally, the configuration is known as the verti-cal dipole mode. Both horizontal and vertical coil configurations maycommonly be used.

At most industrial plant sites the coils are usually carried with theirplanes vertical (horizontal dipole mode), because in this configurationthe measurement is relatively insensitive to misalignment of the coils.Table 1 summarizes EM 34-3 exploration depths and intercoil spacings.

SURVEY DESIGN AND DATA COLLECTION TECHNIQUES

One of the most important elements of a geophysical survey 1s a topo-graphic map of the study area. The map should include sufficient detailto allow accurate location of field activities. An accurate survey gridmust be established over the study area. The grid should be marked inthe field with flagged stakes of nonmagnetic or nonconductive materialand plotted exactly on the site topographic map.

Survey lines should be oriented north-south for magnetic surveys, as thisresults in greater peak-to-trough magnetic anomalies in the northeasternUnited States. Grid spacing and measurement density will, of course,depend upon the objectives of the survey, types of bodies sought, andfield conditions. It is important to select an interval that results insufficient coverage without redundancy and that reduces field time. Ifgreater definition is needed in a particular region of a survey area,additional lines may be added.

A constant 10-foot data interval along each line was found to be theoptimum spacing for most magnetic surveys. A data interval of less than10 feet may result in redundancy, large volumes of data to be processed,and greater field time. Although data spacing of greater than 10 feetmay yield satisfactory results, smaller shallow features may be missed.In regions of suspected shallow burial of drums or tanks (less than 20feet), a 10-foot spacing ensures adequate coverage.

AR302896S

TABLE 1

EM 34-3 EXPLORATION DEPTHS

Intercoil Spacing • Exploration DepthMeters (mjFeet (ft) Horizontal D i p o l e s V e r t i c a l Oipoles

10 m (33 ft) 7.5 (25 ft) 15.0 m (49 ft)20 m (66 ft) 15.0 (49 ft) 30.0 m (100 ft40 m (131 ft) 30.0 (100 ft) 60.0 m (195 ft

Source: Geonics Limited

flR302897

A constant 20-foot data Interval is generally the optimum spacing forelectromagnetic surveys using the EM 31. Surveys using the EM 34-3, con-ducted by the authors, have used data spacings of 50 feet. Other dataspacings may yield satisfactory results; however, the spacing interval isdependent upon the particular survey.

Field notes are very important during the acquisition of geophysicaldata. As the survey lines are traversed, the location of any surfaceferromagnetic material, construction debris (which may contain steelreinforcing material), or cultural features (fences," power lines, build-Ings) should be noted on the data sheet along with the exact position ofthe feature. In addition, topographic and geologic features should benoted. The field notes are then correlated with the data during theinterpretation process. Interferences due to surface features may thenbe identified.

FIELD PROCEDURES - MAGNETIC SURVEY

Magnetic field surveys are most efficient when carried out by two fieldtechnicians. One person carries the instrument and records the measure-ments, while the other person records field notes and helps to maintainstrict quality control (i.e., location and instrument positions).

Prior to each day's activities, a diurnal station should be established.Measurements at this station should be obtained periodically throughoutthe day to check for magnetic storm interferences and natural diurnalvariations of the magnetic field.

Each measurement station must be recoverable within 5 feet of the actualmeasurement station. To maintain a consistent data interval, a 300-foottape measure may be marked off with paint at 10-foot intervals. Thisallows the survey to proceed rapidly and accurately from station tostation.

The authors have also found that at most waste disposal sites and indus-trial plant sites the elevation of the magnetic sensor some distanceabove the land surface serves to attenuate the effects of surfacedebris. A more easily interpreted data set may be obtained using thismethod, because a better signal-to-noise ratio is generally observed.

FIELD PROCEDURES - TERRAIN CONDUCTIVITY

The field procedures for terrain conductivity surveying are similar tothose for magnetic surveying. Terrain conductivity field surveys aremost efficient when carried out by two field technicians. One personcarries the instrument, and calls out terrain conductivity measurements,while the other person records data and field notes (surface features,conductivity measurements). It is also important for both field tech-nicians to maintain strict quality control (i.e., location, instrumentposition).

BR302898

Prior to each day's data collection, a control station should be estab-lished. If possible, it is ideal to have the cont.rol station representnatural "background" conditions. The geophysicist should return to thecontrol station several times daily to verify the repeatability of datameasurements.

The authors recommend that for EH 31 surveys, a minimum of two measure-ments must be obtained at each measurement station. Ideally, the EM 31should be rotated through a full 360 degrees, and a maximum and minimumconductivity value should be recorded. In homogeneous strata the con-ductivity measurement will remain constant through the full rotation. Innonhomogeneous strata or near buried conductors there will be a definitemaximum and minimum value of conductivity. In general field practice,time constraints do not allow a full 360 degrees rotation at eachmeasurement location. As a consequence, the authors have found that twomeasurements may be made at 90 degrees to each other. These measurementsare averaged to determine the station conductivity. In addition, a highdegree of, directionality (i.e., difference) between the two measurementsmay indicate close proximity of a conductive body.

As with magnetic surveying, each measurement station must be recoverableand must be plotted accurately on a topographic map.

DATA REDUCTION AND INTERPRETATION .

MAGNETICS

The G-856 magnetometer, used fay the authors, has a memory capabilitywhich allows on-board storage of data from 1,000 stations. A more recentmagnetometer, produced by the same manufacturer, allows on-board storageof data up to 3,000 stations. This storage feature is especially usefulsince it eliminates the need for manual data transcription.

A typical field day will include approximately 600 readings. Magneticsurvey data may be transferred electronically in the field from the mag-netometer to an HP 110 portable computer. Data from the HP 110 and datastored within the magnetometer memory may then be transferred to a mainframe computer system.

Data may be processed as follows:

1, The x and y coordinates of each measurement point are com-puter generated, and added to the raw data file. .Alsoincluded within the raw data file is information such asJulian day, station number, time, and intensity of totalmagnetic field.

AR302899

2. Profiles of traverse lines can be plotted using either theraw data file or the, modified file with x and y coordi-nates. The distance between measurement stations are plot-ted on the x-axis and intensity of the total magnetic fieldis plotted on the y-axis. The authors have used aTextronlcs 4662 interactive digital plotter utilizing HP3000 graphic routines to generate profiles.

3. A magnetic field contour map is produced from the modifieddata file. The contour interval should be" selected basedon a statistical analysis of the data and the expectedintensity of anomalies produced by target objects.Generally, a 500-gamma contour interval may be initiallyselected. Later, a 100-gamma contour interval may be usedfor refinement of anomaly locations.

INTERPRETATION

Anomalies in the Earth's magnetic field are the result of disturbancescaused by changes in the magnetization of material. A magnetic profileover a magnetically homogeneous region does not contain any anomalies.

A concentration of ferromagnetic material, such as steel drums, acts as amagnetic dipole which distorts the total magnetic field. A typicaldipole source and the resultant anomaly in profile is depicted in Figure5. It should be noted that the asymmetry of the anomaly is due primarilyto the orientation of the magnetic field lines of force in the vicinityof the dipole.

A proton precession magnetometer measures the total magnetic field of theEarth and its localized perturbations. This type of magnetometer doesnot discriminate between the vertical or horizontal components of thefield. The magnetic field is a vector quantity which is inclined atabout 45 degrees to the north at the latitudes of the United States. Theinclination of the field becomes increasingly vertical as one movestowards the magnetic pole, and more horizontal as the magnetic equator isapproached. For moderate magnetic field inclinations in the northernhemisphere, a magnetic anomaly is composed of high total field valueslocated just south of the magnetic source and corresponding low totalfield values directly north of the source. The source itself liesbeneath the region of maximum gradient of the field. A total magneticfield anomaly is depicted in map view in Figure 6.

flR302900

NORTHOTAL FIELD PROFILE

UANO SURFACE __ AREA OFMAXIMUMGRADIENT

MAGNETICSOURCE

DIPOLE SOURCE \ H— MAGNETICTOTAL FIELD' ! / HIGHCONTOURS

SOUTH

Figure 5 MAGNETIC FIELD ANOMALY Figure 6 TOTAL MAGNETICDIPOLE SOURCE FIELD ANOMALY(Field Inclination 45) MAP VIEW

MAGNETIC FIELD(GAMMAS)

STEEL DRUMANOMALY

STRAIQHTLINETANGENT

BUILDINGINTERFERENCE

ANOMALY

'MAGNETIC SOURCEDEPTHcilSFEET

4R30290DISTANCE (FEET)

Figure 7 MAGNETIC PROFILE OVER BURIED SOURCE

The amplitude of a magnetic anomaly decreases with increasing depth ofburial of the magnetic source. For a magnetic dipole, the decrease inintensity of the magnetic field is inversely proportional to the cube ofthe distance between the sensor and the source. The "signature" of amagnetic anomaly will delineate the location of the ferromagnetic mate-rial; it also is indicative of the approximate depth of burial of themagnetic source. A magnetic anomaly over an area of buried steel drumsis depicted in Figure 7.

Methods used for interpretation of the survey data included a qualitativedetermination of the regions of burial of the ferromagnetic material anda more quantitative determination of the approximate depth of burial ofthese materials. The following is a discussion of techniques used by theauthors.

The qualitative procedure is employed initially. This involves an exami-nation of the computer-generated magnetic field contour maps. Regions ofhigh gradient are noted on the map. Regions of probable burial on eachline are also identified. Relative .magnetic highs are indicated on themagnetic field contour maps as red "+" s 1 gns, and the negatives werenoted using blue w-" signs. The color coding proves useful in the inter-pretation of complex regions.

After the location of the buried material has been indicated on each sur-vey line, anomaly trends are noted and indicated on the contour map.

The final process in the qualitative interpretive procedure involvessegregating the map into various magnetic regions - "Highly MagneticZones," "Weakly Magnetic Zones," and "Non-Magnetic Zones." "Highly Mag-netic Zones" often represented regions of tank or drum burial,"Weakly-Magnetic Zones" represented areas of minor amounts of 'buried mag-netic debris, and "Non-Magnetic Zones" represented "clean" areas.

The second phase in the interpretation of the survey data is a quantita-tive procedure. This involves an estimation of the depth of burial ofthe ferromagnetic material. This is based- on the theory that, as thedistance between the magnetic source and the sensor increased, the ampli-tude of the anomaly decreased proportionally.

Several curve-matching methods and numerous deconvo1ut1on and mode1 ingtechniques have been developed to determine the depth to a magneticsource based on this premise. However, many of these techniques requirerigorous mathematical computations. Excellent results were obtained bythe authors using a graphic technique described by Vacquier, Steenland,and Henderson (1951) and Nettleton (1940). This method, commonly knownas the "slope estimate," is based on the fact that the distance from mag-netic source to sensor is proportionally related to the horizontal extentof a straight 1 ine drawn para! lei to the "straight" portion of the

AR302902

maximum gradient of the anomaly. The, "straight slope" distance, X, 1ssometimes multiplied by a factor varying from 0.5 to 1.5, but for practi-cal considerations is usually considered to be 1. As shown in Figure 8,X is considered to be equal to D, the distance from magnetic source tosensor.

The slope estimate technique is applied to each smoothed profile usingthe contour map as a reference. Line-to-line continuity should bechecked during interpretation. If depth estimates, vary significantlyfrom line to line, the profiles should be reexamined to ascertain whethera reinterpretation is required, or if the depth to the ferromagneticmaterial does indeed vary. Subsurface depths of burial can be estimatedby subtracting sensor height from estimated depth to magnetic source.

DIURNAL CORRECTION •

Diurnal variations in the Earth's magnetic field are small, but rapidchanges in the total field, intensity are generally cyclic in nature.Diurnal variations generally have an amplitude of 20 to 50 gammas over a24-hour time period. The effect of such a low-amplitude, long-wavelengthvariation is negligible on a survey to locate buried steel drums ortanks, where anomalies of hundreds to thousands of gammas may commonlyoccur. In addition to diurnal variations, short period micropulsationsin the magnetic field caused by magnetic storms may occur during magneticsurveys. These micropulsations generally have a maximum amplitude ofseveral tens of gammas.

During an aeromagnetic survey, these magnetic field variations are com-monly recorded with an additional magnetometer in a fixed location,referred to as a "diurnal base station." A constant record of magneticfield variation is obtained, and variations in the magnetic field of suf-ficient amplitude may then be subtracted from the magnetic survey data.Another common practice is to review the diurnal record and determinewhether the data collected during a given time period is acceptable orshould be voided based on the amplitude of the diurnal variations.

The use of an additional magnetometer for a diurnal base stationincreases survey costs. An acceptab1e method used by the authors con-sists of taking readings from a fixed point frequently during the mag-netic survey. Diurnal base station data were plotted to determine if theeffects of the diurnal variations were negligible (Figure 9). Generally,the magnetic field variation with time plotted as a smooth curve wil 1have an amplitude of approximately 10 to 40 gammas during the survey.

It is the authors' opinion that the effect of small diurnal variationsmay be considered negligible, on search techniques for concentrations ofburied drums and tanks.

AR302903

Shallow Body

X LATERAL DISTANCE

0 DEPTH TO BODY

Figure 8 METHOD OF ANOMALY DEPTH ESTIMATIONSS29O

ssaac58270

35260TOTAL

MAGNETIC 592 60FIELD

(GAMMAS) 53240

asaao

1 l:OO 12:00 13:00 l*:OO 15:00TlME(HOUflS)

Figure 9 DIURNAL VARIATIONS MARCH 26, 1984

290U

Figure 10 SURVEY GRID LOCATIONS

TERRAIN CONDUCTIVITY . .

DATA REDUCTION

The terrain conductivity instruments used by the authors do not haveon-board memory capabilities. As a consequence, data must be manuallytranscribed in the field. Another possible data gathering methodology,reported by others, is the use of voice-activated tape recorders torecord station, conductivity, and other field notes. Geonics is also nowmanufacturing an instrument with on-board memory.

In a typical field day approximately 400 readings may be obtained.

Data are processed as follows:

1. Data collected in the field are keyed into an HP 110 micro-computer. A spread sheet software program is used to cal-culate average conductivity and directionality.

2* X and Y coordinates are assigned to the data using the samemethodology as described for magnetic survey data,

3. Conductivity data are contoured and profiled by computer.

INTERPRETATION

The primary application of terrain conductivity with which the authorsare familiar includes the use of the EM 31 instrument at an explorationdepth of approximately 20 feet, and the use of the EM 34-3. The EM 34-3was used with the 10- and 20-meter coil spacings. The authors have foundthat the 40-meter coil spacing is generally too susceptible to culturalinterferences at industrial plant sites. Generally, the EM 31 is theinstrument of choice for location of buried drums or tanks, because thesebodies are almost exclusively located within the upper 20 feet of thesubsurface.

Anomalies detected along a survey line' are^related to the electromagneticcontrast of the source object with the surrounding soils. .Drums, tanks,and other buried metallic objects cause local distortions -of the conduc-tivity readings. The amplitude of the anomalies varies with the amount,shape, and depth of the buried source objects.

Terrain conductivity profiles are examined for conductivity highs and/orlows which are not attributable to a known surface source body. Thesehigh and low conductivities represent buried conductive or resistivebodies which are located beneath the anomaly location. For the detectionof buried drums and/or tanks, only conductive regions are considered.

AR302905

CASE STUDY

Statement of the Problem

An Industrial concern initiated a comprehensive environmental study of aninactive industrial waste landfill at their plant site. Exact informa-tion regarding the location and condition of buried wastes was unknown.

As a preliminary screening tool for initial evaluation of the 120-acresite, geophysical techniques were selected to ' provide a rapid,cost-effective and efficient means of assessing subsurface conditions.

Results of the geophysical surveys were used to plan other exploratoryefforts, including location of a comprehensive monitoring well networkand exploratory tests pits designed to assess landfill contents.

Site Description

The site is located along a river within the Coastal Plain physiographicprovince. Surficial geologic materials consist of unconsolidated allu-vial sediments, primarily sand, gravel and cobbles which range from 20 to60 feet in thickness. Underlying the unconsolidated sediments is schistbedrock which exhibits an uneven surface. The top 10 to 20 feet of bed-rock consists of saprolite formed by weathering.

The landfill surface is fairly level with a subdued rolling to flat ter-rain, A small stream cuts through the area, roughly dividing it inhalf. Much of the area consists of dredged river sediments. Grain sizesrange from silt to cobbles.

The landfill is lightly vegetated throughout, with low growing grassesand some small trees and shrubs. Surface debris and surficial evidenceof buried materials is minimal. Chain link fences and electrical powerlines are located on the perimeter of landfill areas (Figure 10).

Approach

Two geophysical techniques were chosen to evaluate the site. These tech-niques provided complementary data sets, while also providing informationunique to each method of exploration.

A magnetic survey was chosen to locate buried drums and ferromagneticdebris.

A terrain conductivity survey was chosen to complement the magnetic sur-vey, and also to locate areas of contaminated groundwater.

AR302906

Grid Spacing = <

A 100-foot x 100-foot grid system was established over the entire studyarea.

The magnetic survey was conducted on traverse lines separated by 50 feetwith measurements taken at 10-foot intervals.

The terrain conductivity survey was conducted on traverse lines separatedby 50 feet with measurements taken at 10-foot and 20-foot intervals.

Data Reduction

Over 1,300 measurements of terrain conductivity were collected during thesurvey. In addition, calibration and background data were collected toensure proper instrument function and assist in interpretation.

Data were keyed into BCM's mainframe computer for processing as describedearlier. A computer program was applied to the data and automaticallycalculated average conductivity and directionality of conductivity. Theprogram also assigned x and y values for each station and generated afile ready for computer plotting.

Following computer data reduction, conductivity contour maps and profilesof selected traverses were generated. Profiles and contour maps wereannotated using field notes taken during the survey to identify anomalies,(areas of conductivity highs or lows) caused by surficial objects.

Interpretationi

A primary step in the interpretation of the terrain conductivity data wasthe establishment of background conductivities at the area of study.This was accomplished by taking data in areas close to, but uneffectedby, the area of study. Background conductivity at the subject siteranged from 4 mmhos/m to 7 mmhos/m.

Following the establishment of background conductivities, terrain conduc-tivity contour maps were examined for the exi stence of 1owered and/orelevated conductivities that could not be attributed to known, naturallyexisting subsurface or manmade features. Areas of elevated conductivi-ties that were not attributable to surficial or known buried featureswere identtfjjed as anomalies.

Magnetometry and.Terrain Conductivity

Within the suspected drum disposal area, magnetometry and terrain conduc-tivity were conducted on the same grid system. Variations in the mag-netic field ranged up to several thousand gammas. Depth estimates wereobtained from magnetic field profiles.

AR302907

Variations in terrain conductivity (exploration depth 6 meters) rangedfrom 4 millimhos/meter (mmhos/m) to 400 mmhos/m.

Similar trench outlines were detected using both techniques (Figures 11and 12).

Magnetometry provided a more quantitative data including burial depthestimates,

Confirmation

Information from the geophysical surveys was used to locate a comprehen-sive network of monitoring wells and exploratory test pits at the site.A total of 96 monitoring wells was installed during a 2-year period tomonitor groundwater quality and provide a long-term monitoring network onthe site perimeter. Information gathered during the geophysical surveysallowed the placement of wells within landfilled areas.

Test pits, constructed to confirm the geophysical surveys and provide anassessment of the condition of buried waste, provided a high degree ofcorrelation with geophysical results. Test pits dug on magnetic surveyanomalies revealed the presence of buried metallic objects (Figure 13).Test pits dug in areas which did not exhibit magnetic anomalies were freeof metallic debris (Figure 14). Test pits excavated on terrain conduc-tivity anomalies revealed waste material.

Summary

Magnetometry and terrain conductivity provided complementary data setswhich were used in the initial characterization of a waste disposal site.

- Magnetometry detected trenches of buried drums and ferromag-netic debris

- Terrain conductivity provided confirmation of trench burialareas, and located contaminated groundwater plumes and buriedwastes

Data from both surveys were used to plan a comprehensive network of moni-toring wells and a test excavation program.

AR3029089

Figure 13 CONFIRMATORY TEST PIT CLEAN AREA

02909Figure 14 CONFIRMATORY TEST PIT MAGNETIC ANOMOLY

CONCLUSIONS

The authors have evaluated the use of magnetic surveying and terrain con-ductivity for the delineation of buried drums and tanks. The followingconclusions may be drawn:

1. Magnetic surveying, when properly conducted and inter-preted, is the preferred method for the location of buriedtanks and drums.

2. Magnetic surveying allows the geophysicist to estimate thedepth of buried bodies.

3. Terrain conductivity surveying will locate buried drums andtanks; however, some difficulties may be encountered inclose proximity to strong conductors.

4. - Both techniques are rapid, cost-effective, and may be con-ducted with a minimum of site preparation.

5. If both techniques are conducted over a study area, thedata sets produced may be complimentary and they may eachprovide specific site information unique to that technique.

AR3029

REFERENCES

Brelner, S., 1973. "Applications Manual for Portable Magnetometers."EG&G Geometries of Sunnyvale, California, 58p.

Oobrin, H. B., 1976. Introduction to Geophysical Prospecting. McGraw-Hill, 630p.

Fowler, John W.s 1985 "Magnetic Survey Methods Used in the InitialAssessment of A Waste Disposal Site".Presented at The National WaterWell Association's Second National Conference and Exposition on Surfaceand Borehold Geophysical Methods in GroundWater Investigation Fort Worth,Texas, 1985

Griffiths, D. F. and R. F. King, 1981. Applied Geophysics for Geologistsand Engineers. Pergamon Press, London.

Nettleton, L. L., 1940. Geophysical Prospecting for Oil. McGraw-Hill,New York.

Parasnis, 0. S., 1972. Principles of Applied Geophysics. Chapman & HallLimited, London.

Reford, M. S., 1964. Airborne Magnetometer Surveys for PetroleumExploration. Aero Service Corporation, Houston, Texas, 39p.

Vacquier, V. and N. C. Steenland and R.G, Henderson, 1951. Interpreta-tion of Aerocnagnetic Maps, Geological Society of America, Memoir #47.Geological Society of America, New York.

iMcNeill, 0. 0., 1980. Electronic Terrain Conductivity Measurement at LowInduction Numbers". Geonics Limited, Ontario, Canada.

AR3029I2