Report of Investigations No. 257

USING AIRBORNE GEOPHYSICS TOIDENTIFY SALINIZATION IN WEST TEXAS

byJeffrey G. Paine, Alan R. Dutton,

and Martina U. Blüm

Bureau of Economic GeologyW. L. Fisher, Director ad interimThe University of Texas at AustinAustin, Texas 78713-8924

1999

CONTENTSABSTRACT 1INTRODUCTION 2

Geology and Soils 2Hydrogeology 7Causes of Salinization 9Oil and Gas Activity 10

GEOPHYSICAL METHODS 12Airborne Geophysics 12Ground-Based Geophysics 13

Single- and Multiple-Frequency Conductivity Profiling 13Time-Domain EM Soundings 15

WATER AND SOIL SAMPLING 16RESULTS 19

Airborne Geophysical Survey 19Magnetic Field Data 19EM Data 20

56,000-Hz Vertical Dipole Data 207,200-Hz Vertical Dipole Data 20900-Hz Vertical Dipole Data 25

Regional Conductivity Patterns and Local Anomalies 25Site Selection 26Regional Conductivity Patterns 26

Northwest Low Conductivity Zone 28Central Conductive Zone 30East-Southeast Low-Conductivity Zone 31

Known Wells Having Conductivity and Magnetic Anomalies 35Site 76 35Site 34 38Site 43 39Site 51 40

Conductivity and Magnetic Anomalies 41Site 17 41Site 71 43

Known Wells with Conductivity Anomalies 45Site 12 45Site 16 47Site 73 48

Other Local Anomalies 49Sites 65 and 67 49

Chemical Composition of Water Samples 50Water Salinity 50Chemical Composition 51Comparison with Regional Trends 52Comparison of Soil and Water Salinity with Measured Conductivity53

DISCUSSION 54Effectiveness of Airborne Geophysics 54Geophysical Profile of a Leaking Well 55Utility of Airborne and Ground-Based Geophysics 55

CONCLUSIONS 57ACKNOWLEDGMENTS 58REFERENCES 59

iii

APPENDIX A. Hatchel area geophysical sites 61APPENDIX B. Chemical composition of ground-water, surface-water, and oil-field-water

samples 67APPENDIX C. Chloride content and electrical conductivity of soils collected in this study

in Runnels County 69

Figures

1 . Map of the Hatchel, Texas, quadrangle 32 . Generalized geologic map of the Hatchel quadrangle 43 . Soil map of the Hatchel quadrangle 64. Total dissolved solids in brine from Guadalupian Series oil fields75. Potentiometric surface of brine in Guadalupian Series based on equivalent

fresh-water hydraulic head 86. Conceptual model of West Texas salinity sources 97. Oil- and gas-well locations in the Hatchel quadrangle 118. Photograph of helicopter lifting DIGHEM magnetometer and EM birds in

preparation for airborne geophysical survey 129. Photograph of RRC worker using a metal detector to locate an abandoned well13

10 . Photograph of Bureau worker using Geonics EM34-3 ground-conductivity meterto perform reconnaissance EM survey 14

11 . Exploration depth of various coil separations and orientations of the EM34-31412 . Protem 47/S transmitter input and receiver response 1513 . Instrument configuration of Protem l LEM sounding 1514. Photograph of LCRA and CRMWD staff sampling a water well1615 . Location of water and soil samples collected by CRMWD, LCRA, and the Bureau1716 . Map of enhanced total magnetic field strength 2117. Map of shallow ground conductivity at 56,000 Hz, vertical dipole coil orientation2218 . Map of moderately deep ground conductivity at 7,200 Hz, vertical dipole

coil orientation 2319. Map of deep ground conductivity at 900 Hz, vertical dipole coil orientation2420. Changes in estimated exploration depth with ground conductivity for

900-, 7,200-, and 56,000-Hz airborne EM coils 2521 . Transient decay and resistivity models for TDEM soundings at site 762922. Transient decay and resistivity models for TDEM sounding at site 823023. Transient decay and resistivity models for TDEM sounding at site 813124. Transient decay and resistivity models for TDEM sounding at site 833125. Transient decay and resistivity models for TDEM sounding at site 843226 . Transient decay and resistivity models for TDEM sounding at site 853327. Transient decay and resistivity models for TDEM sounding at site 8633

iv

28. Transient decay and resistivity models at TDEM soundings at site 17A3429. Transient decay and resistivity models for TDEM sounding at site 873530 . Sketch map of site 76 3631 . Apparent ground conductivity at site 76, measured using multiple coil separations

and horizontal and vertical dipole orientations 3732 . Two-layer conductivity models that fit multiple-coil-separation data for east-west

line 76B and north-south line 76C at site 76 3833 . Sketch map of site 34 3934. Apparent ground conductivity at site 34, measured using 20-m coil separation

and horizontal and vertical dipole orientations 3935. Sketch map of site 43 4036. Apparent ground conductivity at site 43, measured using 20-m coil separation

and horizontal and vertical dipole orientations 4037. Sketch map of site 51 4138. Apparent ground conductivity at site 51, measured using 20-m coil separation and

horizontal and vertical dipole orientations 4139. Photograph of vegetation-kill area adjacent to abandoned brine-disposal pit, site 174240. Sketch map of site 17 4341 . Apparent ground conductivity at site 17, measured using multiple coil separations

and horizontal and vertical dipole orientations 4442 . Sketch map of site 71 4543 . Apparent ground conductivity at site 71, measured using 20-m coil separation and

horizontal and vertical dipole orientations 4544. Sketch map of site 12 4645 . Apparent ground conductivity at site 12, measured using 20-m coil separation and

horizontal and vertical dipole orientations 4646. Sketch map of site 16 4747. Apparent ground conductivity at site 16, measured using 20-m coil separation and

horizontal and vertical dipole orientations 4748. Sketch map of site 73 4849. Apparent ground conductivity at site 73, measured using 20-m coil separation and

horizontal and vertical dipole orientations 4850 . Photograph of abandoned Early A No . 2 well leaking saltwater at site 674951 . Sketch map of sites 65 and 67 4952 . Apparent ground conductivity at site 65, measured using 20-m coil separation and

horizontal and vertical dipole orientations 5053 . Histogram of TDS in ground-, surface-, and oil-field-water samples in the Hatchel area5054. Relation between electrical conductivity and TDS and Cl content of water

and soil samples 5155 . Chemical composition of hydrochemical facies in ground-, surface-, and oil-field-water

samples in the Hatchel area 5156 . Relationship between Br:Cl ratio and Cl concentration in water samples collected

during this study, compared with regional data presented by Richter and others (1990) 52

V

57 . Relationship between Cl:S04 ratios and Na:Ca ratios in water samples collectedduring this study, compared with regional data presented by Richter and others52

58. Comparison of TDS of ground-water samples and ground conductivity measuredby 7,200- and 56,000-Hz surveys 53

59 . Comparison of measurement of ground conductivity by airborne geophysical surveyand measurement of Cl content and electrical conductivity of soil samples53

Tables

1 . Generalized stratigraphy of the Colorado River watershed 52 . Summary statistics for airborne geophysical anomalies 27

vi

ABSTRACTSalinization of soil and water is a

chronic environmental and agriculturalproblem in arid regions. In this study of a91-km2 area in Runnels County, Texas, weintegrated high-resolution airborne andground-based geophysical surveys andchemical analyses of soil and water toidentify near-surface salinization anddetermine its origin. Possible causes ofsalinization are migration of brine alongnatural conduits (faults, fractures, joints,and permeable stratigraphic units), infil-tration from brine-disposal pits and leak-ing oil and gas wells, and evaporativeconcentration of shallow ground water.

An airborne geophysical survey of theHatchel area, where more than 700 oiland gas wells have been drilled since the1920’s, measured magnetic-field intensityand ground conductivity at three electro-magnetic frequencies to identify (1) con-ductivity anomalies caused by salinizationand (2) magnetic-field anomalies causedby well casings and other ferrous objects.Water samples were analyzed to verify air-borne data and distinguish salinity types.We combined airborne geophysical datawith oil- and gas-well locations to identify107 conductivity anomalies consistentwith oil-field salinization.

Ground-based geophysical measure-ments, aerial-photograph interpretations,and record inspections of 54 anomaloussites revealed that at least 42 had oil-fieldsalinization and that 22 might be wellsthat are leaking or have leaked in the past.

We created a geophysical “profile” thatcaptured 20 of the 22 potentially leakingwells identified during field investigations:a site that (1) has a magnetic anomaly ora known well location and (2) has anoma-lously high conductivity as measured bythe high- and intermediate-frequency(56,000- and 7,200-Hz) airborne coils.These results suggest that airborne geo-physics can be combined with well loca-tions for identifying most potentiallyleaking wells without requiring groundinvestigations at every anomaly.

Used alone, airborne methods distin-guish natural salinization from oil-fieldsalinization but have difficulty discrimi-nating among oil-field sources (pits, spills,and leaking wells). Used alone, ground-based surveys can map salinization extentand determine whether wells mightbe leaking, but unknown salinization ismissed. In small areas where well loca-tions are known, ground-based surveyscan determine which wells might be leak-ing, and they are an inexpensive al-ternative to airborne surveys. Airbornemethods are most effective in typical oil-field areas of tens to hundreds of squarekilometers, where well locations are un-certain or multiple salinity sources areexpected. Airborne data can be used todetermine the extent and intensity ofsalinization, locate source areas, focusground investigations, and estimate chlo-ride mass in the ground.

Keywords: airborne geophysics, electromagnetic induction, oil-fieldpollution, salinization, water quality

1

INTRODUCTIONInfiltration of saline water into the

shallow subsurface through natural geo-logical conduits and oil-field sources canimpact wildlife habitat, restrict or elimi-nate agricultural uses of land, and polluteaquifers and surface water (Dutton andothers, 1989; Richter and others, 1990).Agricultural practices can exacerbate theproblem by locally raising water tablesand increasing dissolved mineral concen-tration by evaporation. Public concernabout the environmental effects of salini-zation has led to an increasing interest inmethods of determining whether oil-fieldbrines have entered the subsurface, wherethey have migrated, and whether they arecausing specific problems on the land sur-face, in water wells, or in surface water.

The purpose of this study was toevaluate the use of airborne and ground-based geophysics to identify and discernthe sources of near-surface salinization. Ahigh-resolution electromagnetic (EM) andmagnetometer survey of a 91-km2 studyarea (the Hatchel area) near Ballinger inRunnels County, Texas (fig. 1), was flownto locate areas having a geophysicalsignature consistent with that of a wellleaking brine. The EM survey was em-ployed to locate ground that is electricallyconductive because of salinization. Themagnetometer survey was employed to lo-cate magnetic anomalies caused by wellcasings. Airborne geophysical signaturesthat might indicate a leaking well in-cluded a conductivity anomaly and eitheran associated magnetic anomaly or aknown well location.

Ground-based investigations focusedon representative sites identified by theairborne data. These investigations in-cluded detailed EM surveys and collectionand chemical analyses of soil and watersamples. Because ground conductivity is

affected not only by pore-fluid chemistrybut also by soil type, rock type, andmoisture content, these factors werealso considered. Once the data for agiven site were analyzed, we interpretedthe most likely cause of the conductiv-ity anomaly at that site.

Geology and SoilsRock and soil type can influence

ground conductivity, but the effect of thehost material on ground conductivity isgenerally much less than that caused bymoisture content and ionic concentrationin pore fluids. Clayey soils and rocks aregenerally more conductive than sandyunits (McNeill, 1980a; Rhoades, 1981).Geologic units (Kier and others, 1976) atthe surface or in the shallow subsurfaceinclude (1) Quaternary alluvium (Qal,fig. 2; table 1) that is generally sandy, rela-tively high in water content, and found intopographic lows along streams; (2) thinPleistocene terrace deposits (Qu) that arecomposed of clay to gravel, are relativelydry and are mapped at higher elevationsalong the central part of the study area;and (3) outcrops of Permian Clear ForkGroup strata (Pcf) and the Lueders Forma-tion (Pl) containing units of sandstone,limestone, shale, and marl. Permian strati-graphic units dip west-northwestward intothe Permian Basin. The Coleman JunctionFormation, a limestone and shale forma-tion of the Permian Wichita-Albany Group(Kier and others, 1976), is an artesianbrine-bearing unit at depths of 200 to300 m that is commonly cited as a princi-pal source of near-surface brine (Richterand others, 1990). The Coleman Junctionis beyond the exploration depth ofairborne and ground-based geophysicalinstruments used in this study.

2

FIGURE 1. Map of the Hatchel, Texas, quadrangle, Runnels County. The airborne geophysical survey area is within theshaded rectangle. Shaded areas are oil and gas fields adapted from Abilene Geological Society (1992).

Soil type (fig. 3) can affect ground con-ductivities measured by high-frequency,shallow-penetrating conductivity instru-ments. Soils have been mapped in greaterdetail than have the geologic units in thestudy area (Wiedenfeld and others, 1970).In the lowlands, deep, loamy, relatively

wet soils of the Spur-Colorado-Miles asso-ciation are formed on Quaternary al-luvium (figs. 2, 3). These soils generallycontain less clay than do the upland soils,but they may have higher conductivity be-cause they are thicker and wetter, and inmany places they contain conductive pore

3

FIGURE 2. Generalized geologic map of the Hatchel quadrangle. Numbered sites are locations of local geophysical anomaliesand ground-based geophysical surveys (app. A). Adapted from Kier and others (1976).

4

5

TABLE 1. Generalized stratigraphy of the Colorado River watershed. Modified from Brown and others (1972),American Association of Petroleum Geologists (1973), Eifler (1975), Lee (1986), and Richter and others (1990).

6

FIGURE 3. Soil map of the Hatchel quadrangle. Numbered sites are locations of local geophysical anomalies and ground-basedgeophysical surveys (app. A). Adapted from Wiedenfeld and others (1970).

FIGURE 4. Total dissolved solidsin brine from Guadalupian Seriesoil fields. Modified from McNeal(1965).

water. Soils on the uplands are generallythe shallow, clayey to loamy, and relativelydry residual soils of the Portales-Potter-Mereta association that form on the Pleisto-cene terrace deposits and Permian ClearFork Group strata (figs. 2, 3). Despite theirrelatively high clay content, these soilshave low measured conductivity becauseof their thinness and low moisture content.

HydrogeologyThe Hatchel area lies on the east flank

of the Southern Great Plains, which slopesgently southeastward from eastern NewMexico to Central Texas and at the southend of the Rolling Plains. The informallynamed Southern Great Plains regionalground-water flow system includes diversehydrostratigraphic units in Ordovician toNeogene rocks (Jorgensen and others,1988). Physiographic features of theHatchel area include gently rolling, dis-

sected uplands and valleys formed by thesouthward-draining Elm and CoyoteCreeks and their tributaries.

The Upper Permian rocks that con-tain fresh ground water at shallow depthsin the Hatchel area contain brine and hy-drocarbons only tens of kilometers to thewest (McNeal, 1965; Core Laboratories,1972). Oil and gas fields occur at depthsas shallow as 300 m in Permian forma-tions to more than 1,800 m in Pennsylva-nian and Ordovician rocks (table 1). Brineis prevalent throughout the Paleozoic sec-tion except where it is displaced by locallyrecharged meteoric water. For example,salinity of water in the GuadalupianSeries, which overlies the Clear Fork Group(table 1), increases from 50,000 mg/Ljust west of the outcrop to more than200,000 mg/L westward in the PermianBasin (fig. 4).

Regional and local topographic reliefhas large effects on ground-water flow

7

FIGURE 5. Potentiometric surfaceof brine in Guadalupian Seriesbased on equivalent fresh-waterhydraulic head. Modified fromMcNeal (1965).

(Tóth, 1962). Potentiometric surfaces inthe Southern Great Plains regional flowsystem (Jorgensen and others, 1988) areinclined toward the east, indicating thepotential for eastward flow of brine inPaleozoic rocks toward outcrops in theHatchel area (fig. 5). Eastward fluid flowprobably influenced migration of hydro-carbons into reservoirs across the eastshelf of the Permian Basin. Richter andKreitler (1986) showed that brine at shallow(~30 m) depths in the south part of theRolling Plains is derived from deep parts ofthe regional flow system. Bein and Dutton(1993) mapped connate and meteoricwater throughout the stratigraphic sectionacross the Southern Great Plains andshowed eastward movement of salinewater into the Rolling Plains. Comparison

of potentiometric surfaces of hydrostra-tigraphic units in Paleozoic rocks mappedby McNeal (1965) indicates that there ispotential for upward movement of brineacross confining layers if pathways such asfractures and unplugged boreholes exist. Po-tentiometric surfaces of brines beneath thevalleys in the study area could be close to,and in some Wolfcampian formations (forexample, Coleman Junction) higher than,ground surface elevations of 490 to 550 m.

Potentiometric surfaces of shallowground water in different aquifer unitsmost likely are inclined toward Elm andCoyote Creeks and their tributaries (fig. 1),reflecting topographic influence on re-charge and discharge locations in the localground-water flow system (Tóth, 1962).The upland areas are probably local re-

8

FIGURE 6. Conceptual model of West Texas salinity sources. Potential sources are (1) natural brine discharge through perme-able stratigraphic units, fractures, and joints; (2) infiltration of saline water beneath brine-disposal pits; (3) upward flow ofbrine through inadequately plugged and leaky boreholes; and (4) evaporative concentration of shallow ground water. Adaptedfrom Richter and others (1990).

charge zones, and ground-water dischargefrom local and regional flow systems ismost likely focused in the valleys of Elmand Coyote Creeks.

Causes of SalinizationSalinization of soil and water is a

common problem in the central andsouthwestern United States (Dutton andothers, 1989; Richter and others, 1990).Potential causes of soil and water saliniza-tion in the Hatchel area include (1) natu-ral discharge of subsurface brines throughpermeable units, fractures, or joints; (2) in-filtration of produced brine beneath dis-posal pits; (3) upward movement of brineacross confining beds through unpluggedoil, gas, and deep water wells; and (4) evap-orative concentration of ground waterfrom shallow water tables that have risenin response to agricultural landscapingand consequently increased ground-waterrecharge (fig. 6).

Oil and gas wells provide potentialpaths for brine to move to the near-surface

environment. If brine is associated withthe produced oil or gas, it moves to thesurface along with the hydrocarbons. InTexas, before the practice was banned in1969 (Richter and others, 1990), pro-duced brine was commonly dischargedinto unlined pits and left to evaporate orinfiltrate the ground. Brine within non-producing geologic units, such as theColeman Junction Formation, can alsoreach the near-surface environmentthrough deep water wells and improperlyplugged oil and gas wells reaching deeperstrata.

Although salinization from any of thenatural, oil-field, or agricultural sourcescould cause similar increases in groundconductivity, each type may have a uniquegeophysical or chemical signature thatwould allow it to be distinguished. Forexample, ground conductivity that de-creases with depth may indicate surfacesalinization associated with a brine-disposal pit or evaporative concentration.Conductivities that increase downwardsuggest deeper salinity sources, such as a

9

leaking well or natural flow path. Arealconductivity patterns and magnetic datafrom airborne surveys help further distin-guish potentially leaking wells (pointsources having magnetic anomalies) fromnatural conduits (curvilinear features havingno magnetic anomalies).

Oil and Gas ActivityThere are 18 named oil and gas fields

that cover a substantial part of theHatchel quadrangle (fig. 1). RailroadCommission of Texas (RRC) records indi-cate that there are at least 963 wells in thequadrangle and 718 in the Hatchel area(fig. 7) and that most of the explorationand production occurred before the no-pitorder. Although these wells are distributedunevenly, few areas are unexplored. Andbecause drilling, production, and plug-ging practices have improved over the de-cades that have passed since the discovery

of oil and gas, areas of early activity are ofgreater environmental concern than arefields discovered recently.

Oil and gas production in the Hatchelarea is largely from the PennsylvanianCisco, Canyon, and Strawn Groups(Fitzgerald, 1952; Gardner and Phifer,1953; Abilene Geological Society, 1992).The oldest wells are found in McMillanand Beddo fields, which, discovered inthe late 1920’s, are located in the north-east and east parts of the study area (fig. 1).Most fields were discovered between thelate 1940’s and the early 1960’s, includingBallinger (1947), Elm Creek (1950), Patsy(1951), Hays and Watts (1951), Spill(1952), Hall (1953), Richardson (1957),Byers (1957), Andergram (1958), and D&Gfields (1963). Hatchel-area discoveries inthe 1980’s include QV field, Bear Cat field(1981), North Big Ed field (1982), Quickoil and gas field (1984), and GHG field(1985).

FIGURE 7. Oil- and gas-well locations inthe Hatchel quadrangle. Well locationsfrom Railroad Commission of Texas. Fieldlocations shown in figure 1.

10

GEOPHYSICAL METHODSWe used electromagnetic induction

(EM) methods (Parasnis, 1973; Frisch-knecht and others, 1991; West andMacnae, 1991) to measure apparent elec-trical conductivity of the ground with air-borne and ground-based instruments. EMmethods employ a changing primary mag-netic field that is created around a trans-mitter coil to induce a current to flowwithin the ground, which in turn creates asecondary magnetic field that is sensed bythe receiver coil. In general, the strengthof the secondary field is proportional tothe conductivity of the ground. We as-sume that the near-surface environmentconsists of horizontal layers of infinite lat-eral extent. Although not strictly true any-where, near-surface layers have sufficientlateral extent in the Hatchel area to vali-date the assumption.

Airborne GeophysicsGeoterrex–Dighem, a geophysical ser-

vice company based in Canada, surveyedthe Hatchel study area using helicopter-based geophysical instruments in January1996 (Garrie, 1996). Principal instrumentsmounted in the helicopter were the DIGHEMV

system that measures ground conductivityby using five EM coil pairs and that mea-sures magnetic-field strength by using aCesium vapor magnetometer. The EM coilsand the magnetometer were slung beneaththe helicopter (fig. 8) at nominal heights of30 and 40 m, respectively. The helicoptermaintained a height of 60 m and flew at anaverage speed of 108 km/hr (Garrie, 1996).Supporting instruments included a differen-tial GPS navigation system having a loca-tional accuracy to 5 m and a radar altimeter.

FIGURE 8. Photograph of helicopter lifting DIGHEM magnetometer and EM birds in preparation for airbornegeophysical survey.

12

Flight lines for this high-resolutionsurvey, oriented east-west and spaced100 m apart, covered a total length of 940 kmwithin the 91-km2 Hatchel area (fig. 1). EMand magnetometer data were sampled at0.1-s intervals, corresponding to a samplespacing of about 3 m along each flightline. EM coil separations of 6 to 8 m anda bird height of 30 m suggest a footprint afew tens of meters wide. EM and magne-tometer data were processed by Geoterrex-Dighem (Garrie, 1996). Images of groundconductivity at three coil frequencies(900, 7,200, and 56,000 Hz in a verticaldipole orientation) and total and en-hanced magnetic-field strength were im-ported into a geographic informationsystem for spatial analysis that includedcomparisons with maps of soil type, geo-logic unit, and well location.

Ground-Based GeophysicsBureau of Economic Geology (Bureau)

staff investigated representative sites usingground-based EM instruments. Frequency-domain EM methods were used in a re-connaissance mode to confirm salinizationand in multifrequency mode to delineatelateral and vertical extent of salinization.Time-domain EM (TDEM) soundings wereused to estimate plume thickness and toexamine changes in conductivity deeperthan depths reached by other airborneand ground-based methods. RRC staffused a portable metal detector to locatewell casings and guide conductivity sur-vey placement at the selected sites (fig. 9).

Single- and Multiple-FrequencyConductivity Profiling

Reconnaissance conductivity profileswere completed at 28 sites (app. A) to lo-cate highly conductive ground identifiedfrom the airborne data that may containsites of salinization. In these surveys, a

FIGURE 9. Photograph of RRC worker using a metal detectorto locate an abandoned well.

Geonics EM34-3 ground-conductivitymeter (fig. 10) was used to measure ap-parent conductivity (McNeill, 1980b)along lines crossing the sites. The EM34-3supports a 10-, 20-, or 40-m transmitter-and receiver-coil separation (fig. 11) andtwo principal coil orientations (horizontaland vertical dipole). A 10- or 20-m coilseparation was used for reconnaissancesurveys, having exploration depths of 6 or12 m in the horizontal dipole orientationand 12 or 25 m in the vertical dipole ori-entation (fig. 11). The conductivity valuesrepresent a weighted average conductivityof the sensed soil volume beneath thetransmitter and receiver coils. Althoughreconnaissance profiling can delineatesalinized areas rapidly, it reveals little in-

13

FIGURE 10. Photograph of Bureau worker using Geonics EM34-3 ground-conductivity meter to perform reconnais-sance EM survey across abandoned Gardner No. 1 well at site 33. Transmitter coil shown.

formation on conductivity changes withdepth.

Multiple-frequency profiling, one oftwo ground-based methods used to exam-ine conductivity variations with depth,was completed at 32 sites by using theEM34-3 (app. A). The exploration depthof the EM34-3 increases with increasingcoil separation (decreasing frequency) fora given coil orientation (fig. 11). Conse-quently, conductivity measured at differ-ent coil separations and orientations canbe used to infer conductivity changes withdepth (McNeill, 1980b, c) to a maximumdepth of about 30 m. We measured appar-ent conductivity (in mS/m) first at 10-mcoil separation using horizontal and verti-cal dipole orientations. After we com-pleted a line using the 10-m separation,the 20-m separation was selected, the in-strument was recalibrated, and horizontaland vertical dipole data were collectedalong the same line. Finally, the 40-m sep-aration was selected, the instrument was

recalibrated, and apparent conductivity forboth dipole orientations was measured.

EMIX34 Plus, a computer programpublished by Interpex, was used to pro-cess the data and produce simple modelsof lateral and vertical conductivity

FIGURE 11. Exploration depth of various coil separations andorientations of the EM34-3.

14

changes. Horizontal and vertical dipoleconductivities for each coil separation wereentered in the program, and a starting con-ductivity model was entered, consisting oflayer thicknesses and conductivities thatqualitatively fit the observed data. Thecomputer then displayed both the ob-served conductivities and those calculatedfrom the model. The user then adjustedthe model to better fit the observed data.After reasonable agreement was obtained,the program adjusted layer thicknessesand conductivities to obtain the best fit.The program then determined the range ofmodel thicknesses and conductivities,producing an equivalent fit to the ob-served data.

Time-Domain EM SoundingsWe acquired time-domain, or tran-

sient, EM (TDEM) soundings (Kaufmanand Keller, 1983; Spies and Frischknecht,1991) using the Geonics PROTEM 47/S toproduce vertical conductivity profiles to amaximum depth of 75 to 100 rn. Ratherthan changing coil frequencies and sepa-

FIGURE 12. Protem 47/S transmitter input and receiver re-sponse. Adapted from Geonics Limited (1992).

FIGURE 13. Instrument configuration of Protem TDEMsounding.

rations to vary exploration depth, TDEMinstruments measure the decay ofa secondary EM field (the “transient”)produced by the collapse of a primaryEM field (fig. 12). The secondary fieldstrength is measured by the receiving coilat moments in time (or “gates”) followingtransmitter-current termination. Second-ary field strength at early time gives infor-mation about conductivity in the shallowsubsurface; field strength at late time isgoverned by conductivity at depth. Thecomputer program TEMIX, by Interpex,was used to construct vertical conduc-tivity profiles that best fit the observedtransient decay for each sounding.

Thirteen TDEM soundings were con-ducted at nine sites in April 1996 (app. A).We collected all time-domain soundingsusing a 40 x 40-m transmitter loop, withthe receiver coil outside the transmitterloop (fig. 13). Whereas results of multiple-coil-separation soundings are given in theconductivity unit mS/m, results of TDEMsoundings are customarily given in theresistivity unit ohm-m. These units are theinverse of each other and are converted bythe equationConductivity (mS/m) = 1,000/Resistivity (ohm-m).

15

WATER AND SOIL SAMPLINGGround-water, surface-water, and soil

samples were collected to (1) evaluatehow well the airborne geophysical datacharacterize ground-water and soil salinityand (2) identify sources of salinity in theHatchel area. Data collected by theBureau were merged with data collectedby the Lower Colorado River Authority(LCRA) and the Colorado River MunicipalWater District (CRMWD). Ground waterwas sampled from most of the accessiblewater wells, including 6 sampled by theBureau and 19 by LCRA and CRMWD(figs. 14, 15). Sampled wells were in areasof high and low conductivity mapped bythe airborne EM survey. Soil samples werepaired with ground-water samples andchosen in areas of high and low conduc-tivity shown on the 56,000-Hz ground-

conductivity map. Surface water wassampled in January 1996 by LCRA andCRMWD and in April 1996 by the Bureau(fig. 15). In addition, water samples werecollected at two abandoned oil wells beingplugged by RRC and at an unpluggedcore hole.

Ground-water samples were taken byBureau staff from abandoned and activewells. Where a well had no pump, we in-stalled a temporary pump to purge thewell bore and obtain a sample. Groundwater was sampled by the LCRA andCRMWD from wells equipped with apump or windmill. Three of the wellssampled by Bureau staff were hand-dugwells holding 475 to 2,480 L in storage.The others were drilled wells holding lessthan 40 L in storage.

FIGURE 14. Photograph of LCRA and CRMWD staff sampling a water well during the airborne geophysical survey.

16

We collected water samples after thetemperature, pH, and Eh had stabilized,which took less than 1 hr. During thattime the water columns in the wells weredrawn down as much as 80 percent at our15- to 19-L/min pumping rate. Watersampling followed standard techniques(Brown and others, 1970; Wood, 1976), andsamples were passed through a 0.45-µ mfilter and collected in bottles. We mea-sured alkalinity by titration of unfilteredsamples using H2SO4 within several hoursof collection. Waters to be analyzed forcations were acidified in the field bymeans of 6N HNO3; waters to be analyzedfor anions were collected without treat-ment. Sample bottles were sealed andstored at 4°C before analysis. We mea-sured cation concentrations using induc-tively coupled plasma-optical-emissionspectroscopy. Fluoride, chloride, bromide,nitrate, and sulfate concentrations weremeasured by ion chromatography. Bicar-bonate was determined by titration. TDS(total dissolved solids) of the CRMWDsamples was determined by gravimetricanalysis of a filtered sample. TDS of theBureau samples was calculated as the sumof reported dissolved constituents (cat-ions, anions, and dissolved silica).

Water from the abandoned oil wellswas sampled at the wellhead. Fluid pres-

sure allowed the wells to dischargethrough the valve used to control flowduring plugging. The collected liquid con-tained black suspended solids (pipe scale)and some oil. The samples were pre-filtered through glass wool and thendrawn through an in-line 0.45-µ m filter.Water temperature was measured in thestream flowing from the wellhead; pH wasmeasured in the filtered sample. The sam-ple from the core hole was drawn into thefilter chamber through a 0.45-µ m filter.

The four surface-water samples col-lected by the Bureau were drawn througha 0.45-µ m filter and decanted into polyeth-ylene bottles. Temperature was measured inthe stream, and pH was measured on site.

Soil samples were collected with a handauger, air dried, and milled. Electrical-conductivity and chloride measurementswere performed on 1:1 extractions.Samples were shaken for 1 hr then centri-fuged to obtain a particle-free liquid. Elec-trical conductivity, which indicates theconcentration of soluble salts in an aque-ous sample, was determined on the super-natant by means of a microconductivitycell and corrected to 25° C. Chlorideconcentrations were determined by argen-tometric titration or ion chromatography.

18

RESULTSWe analyzed airborne geophysical

data to determine whether airborne meth-ods can reliably delineate salinized areas,distinguish oil-field salinization fromother causes, locate oil and gas wells andassociated salinization, and guide groundinvestigations of salinization.

Airborne GeophysicalSurvey

One of the goals of the airborne geo-physical survey was to identify areaswhere conductive ground coincided witha magnetic anomaly. This geophysical sig-nature, insofar as a well casing produces amagnetic anomaly and salinization pro-duces a conductivity anomaly, might indi-cate areas where saltwater has entered theground through a brine-disposal pit or aleaking oil or gas well. Airborne-surveydata, including maps of magnetic-fieldstrength and ground conductivity, wereanalyzed to identify regional and localanomalies for field investigations.

Magnetic-Field DataTotal magnetic-field data from the

Hatchel area show abundant local anoma-lies superimposed on the regional gradi-ent of Earth’s magnetic field (fig. 16).Magnetic-field strength increases alongthe regional gradient from the lowestvalue of 50,186 nanoTeslas (nT) in thesouthwest corner of the Hatchel area tothe highest value of 50,382 nT in thenortheast corner (Garrie, 1996). These re-sults are consistent with regional maps ofmagnetic-field strength (Bennett andothers, 1985).

Linear and oval magnetic anomaliesare superimposed on the regional gradient(fig. 16). The anomalies are weak relative

to the total field strength (as much as 30 nT,or about 0.06 percent) but are well abovethe 0.01-nT sensitivity of the magnetom-eter. The linear features, some of whichextend completely across the study area,are generally 150 to 200 m wide. Their loca-tions correspond to those of known pipelines.Two of the anomalies, the northwest-south-east-trending line in the southwest cornerand the north-south-trending line on theeast side that bends to the northeast, areboth known pipelines that are mislocatedon maps (fig. 16). Mapped pipelines arealso in the study area that are not visibleon the magnetic field map, perhaps becausethey are constructed of nonferrous mate-rial or oriented parallel to the flight lines.

Most of the oval anomalies are about160 m across, although they range from80 to 200 m. These are anomaly dimen-sions in an east-west direction (along theflight lines), where magnetometer mea-surements were acquired at 3-m intervals,rather than north-south (across the flightlines), where measurements were 100 mapart. Most of the oval magnetic anoma-lies coincide with known wells (fig. 16)and are interpreted to be magnetic-fieldperturbations caused by ferrous elementsof the well (casing and pump jack). Othermagnetic anomalies coincide with struc-tures containing significant magneticallysusceptible material, such as some homes,metal barns, and windmills.

Some oil and gas wells (known fromRRC records and field investigations) arelocated where no magnetic anomalies aremapped. These wells were undetected be-cause of either small magnetic anomalieslocated between flight lines or lack of wellcasing. Typical oval anomaly diameters of80 to 200 m suggest that the 100-mflight-line spacing was a good compromisebetween anomaly detection and survey

19

cost. Line spacings of 200 m that weresimulated during processing reduced thenumber of detected magnetic anomalies.

EM DataData for constructing ground-

conductivity maps of the Hatchel areawere obtained from EM coils operatingat 56,000 (fig. 17), 7,200 (fig. 18), and900 Hz (fig. 19) in a vertical dipole orienta-tion. Because exploration depth depends onfrequency and ground conductivity, deeperexploration depths were attained at lowercoil frequencies and, for a given coil fre-quency, less conductive ground (fig. 20).

56,000-Hz vertical dipole dataGround conductivities measured by

the 56,000-Hz coils, the shallowest-exploration-depth frequency, showedgood contrast between the lowest ob-served values of about 60 to the highestvalues of 730 mS/m (fig. 17). Maximumexploration depths range from 2.5 m overhighly conductive ground to as much as9 m over ground having the lowest ob-served conductivities (fig. 20). This shal-low depth is most affected by soil type(clay soils are more conductive than sandysoils), moisture content (wet soils aremore conductive than dry soils), andpore-water chemistry (saline water is moreconductive than fresh water).

Highly conductive areas visible on the56,000-Hz map include numerous smallovals that are generally 80 to 250 macross, curvilinear features that are a fewhundred meters wide and hundreds ofmeters long, and large, irregular featurescovering many square kilometers (fig. 17).

Significant curvilinear features in-clude (1) a zone about 200 m wide andmore than 2 km long located along a bluffwest of Bluff Creek in the north part ofthe study area, (2) a segment about 400 m

wide along Coyote Creek that extendsfrom the north edge to near the south edgeof the Hatchel area, (3) a 200-m-wide zonealong a bluff west of Elm Creek, and (4) ashort segment of Elm Creek just down-stream from its confluence with EagleBranch (fig. 17). The curvilinear conduc-tivity highs located at similar elevationsalong a bluff or hill slope probably markground-water seeps. Curvilinear featuresalong the creeks most likely representareas where higher salinity water flows inthe creek or associated alluvium.

Ground investigations focused on thesmaller, oval-shaped anomalies in the56,000-Hz and lower frequency data.These anomalies are the ones most likelyto be caused by oil-field salinization be-cause of the limited lateral distance thatsaltwater plumes generally move frombrine-disposal pits and leaking wells.Some anomalies in the 56,000-Hz datamay also include small patches of conduc-tive soil, springs, and stock tanks.

The most significant of the large, irregu-lar areas of high conductivity are (1) thewest third of the study area, (2) the areasouth of Hatchel and east of CoyoteCreek, and (3) the area just east of ElmCreek and south of its confluence withMud Creek (fig. 17). These features aretoo extensive to be caused by individualleaking wells, but some may indicate nu-merous and closely spaced saltwatersources. Most of these large features,probably not oil-field related, are insteadareas where soils or shallow geologic unitshave more clay, have become wetter, orcontain water that has a higher dissolvedmineral content.

7,200-Hz vertical dipole dataGround conductivity sensed by the

7,200-Hz airborne coils ranges from 60 tomore than 700 mS/m, only slightly lower

20

FIGURE 20. Changes in estimated exploration depth(skin depth) with ground conductivity for 900-, 7,200-, and56,000-Hz airborne EM coils. Shaded area indicatesconductivity range observed in the Hatchel area.

than the range observed for the higher fre-quency coils (fig. 18). The 7,200-Hz coilsexplore deeper than do the 56,000-Hzcoils, ranging from 24 m for the least con-ductive ground to 7 m for the most con-ductive ground (fig. 20).

Highly conductive areas are morecommon on the 56,600-Hz map than onthe 7,200-Hz map, suggesting that manyof the conductivity anomalies visible onthe 56,000-Hz map represent very shal-low features. Local, oval-shaped anomaliesthat are ubiquitous on the high-frequencymap are less common on the 7,200-Hzmap. They are most common along theupper reaches of Mud Creek, east of ElmCreek and south of its confluence withMud Creek, and east and northeast ofHatchel, and a few more are scatteredacross the study area (fig. 18). Large areasof anomalously high conductivity arefound along and east of Coyote Creek,along and west of Bluff Creek upstream ofits confluence with Elm Creek, and overthe entire west-northwest section of thestudy area. The strongest large anomaliesare along Coyote Creek and Bluff Creek.Linear and curvilinear features such asthose evident on the 56,000-Hz map areabsent at 7,200 Hz except along someroads. These are caused by power lines.

900-Hz vertical dipole dataHighly conductive ground measured

by the 900-Hz airborne coils covers lessarea than that detected by the 7,200-Hzcoils (fig. 19). Total range in conductivityis also lower for the 900-Hz coils, rangingbetween 60 and 400 mS/m. Maximumexploration depths are 68 m for the leastconductive ground and 27 m for the mostconductive ground (fig. 20).

Few local conductivity anomalies arevisible in the 900-Hz data. The principalfeatures are three areas of similar groundconductivity (fig. 19). The smallest is anarea of low to moderate conductivity (100to 150 mS/m) in the northwest part of theHatchel area. This low-conductivity zoneis bordered to the southeast by a 4- to5-km-wide band of highly conductive(150 to 400 mS/m) ground trendingnorth-northeast. Southeast of this zoneis an extensive area of low conductivity(60 to 120 mS/m).

The north-northeasterly trend of theconductive band matches the strike ofPermian geologic units (Kier and others,1976). These units dip gently to the west-northwest into the Permian Basin. The900-Hz coils, which are sensing the upper30 to 40 m in this zone, are probably de-tecting a unit bearing saltwater. This unit,more conductive than units above andbelow it, shallows to the east-southeast.Probably too deep to be detected by theairborne instruments in the northwestcorner of the study area, it has been re-moved by erosion in the southeast part ofthe area. We acquired more deeply explor-ing TDEM soundings to further investigateconductivity patterns evident at 900 Hz.

Regional ConductivityPatterns and Local Anomalies

Several natural and cultural effectscan cause highly conductive ground.

25

Shapes of conductivity anomalies in mapview and changes in conductivity withdepth should allow us to distinguishnatural conductivity anomalies (salinesprings, clayey or wet soil, and salinegeologic formations) from oil-field anoma-lies, but we might not be certain from air-borne data alone whether a given oil-fieldanomaly represents a leaking well, a brine-disposal pit, or a spill. In an attemptto discover geophysical signatures forsalinization and its various sources, wecompleted field investigations at sitesrepresenting the diversity that might beencountered in geophysical studies ofoil-field salinization. These investigationsincluded ground-based geophysical sur-veying and soil and water sampling. Wealso noted the soil type, geologic unit,and topographic setting for examining thepotential effects of those properties on thegeophysical signature.

Site SelectionWe made ground-based geophysical

measurements at 63 field sites chosen onthe basis of airborne geophysical signatures(app. A). We examined regional conductiv-ity patterns evident in the deep-exploring,900-Hz conductivity images using TDEMsoundings, and we examined local anoma-lies evident in the 56,000- and 7,200-Hzconductivity images using primarily single-and multiple-coil-separation conductivityprofiling.

We identified 107 local geophysicalanomalies (table 2; app. A) by visuallyinspecting the airborne magnetic and EMdata. These anomalies were classified intofive common airborne signature types:CMW, CM, CW, C, and MW. C denotes aconductivity anomaly detected by the56,000-Hz coils, M denotes a magneticanomaly, and W denotes a known welllocation. For example, type-CMW sites are

those that have high-conductivity anoma-lies on the 56,000-Hz map, have magneticanomalies, and have known wells nearby.Of the 107 local anomalies, 69 are typeCMW and indicate possible salinizationnear known and detected wells. There are15 CM sites, which might indicate salin-ization near tank batteries or near wellsthat are either mislocated or not invento-ried. The 12 type-CW sites might indicatesalinization at known wells undetected bythe airborne magnetometer or at wellshaving no casing. The 11 type-C sitesmight be caused by salinization near anunknown well that went undetected bythe airborne magnetometer. In addition toinvestigations at the conductivity anom-alies, we visited two type-MW sites. Thesesites, which number in the hundreds,represent known wells having detectedmagnetic anomalies. Lack of an associatedconductivity anomaly indicates no evi-dence of salinization.

Ground-based geophysical investiga-tions consisted of (1) 43 reconnaissanceconductivity profiles across 29 sites, (2) 60conductivity profiles at multiple explora-tion depths at 32 sites, (3) 6 TDEMsoundings at two local anomalies, and (4) 7TDEM soundings at locations chosen onthe basis of regional conductivity patterns.

Regional Conductivity PatternsConductivity patterns on the 900-Hz

map differ fundamentally from those onmaps showing conductivity at shallowerdepths (compare figs. 17 through 19),having a northeasterly trend resemblingthat on the regional geologic map (Kierand others, 1976). TDEM soundingsdetect vertical conductivity changes tomaximum depths of 80 to 100 m, whichare comparable to the 30- to 60-m explo-ration depth of the 900-Hz airborne coils(fig. 20) and greater than depths reached

26

TABLE 2. Summary statistics for airborne geophysical anomalies (figs. 17 through 19; app. A). Type-CMWsites are those that have a conductivity anomaly, a magnetic anomaly, and a known well. Type-CM sites havea conductivity anomaly and a magnetic anomaly, but no known well. Type-CW sites have a conductivityanomaly and a known well, but no magnetic anomaly. Type-C sites have a conductivity anomaly, but neithera known well nor a magnetic anomaly.

TABLE 2 (cont.)

1Sites that fit a profile of a potentially leaking well are those that show conductivity anomalies measured by both the56,000-Hz and 7,200-Hz airborne coils and have a known well location, a magnetic anomaly, or both.

by the EM34-3 (fig. 11). TDEM soundingsites (app. A; fig. 19) represented large ar-eas having similar conductivities on the900-Hz map. For sounding purposes, theHatchel area was divided into a low-con-ductivity zone in the northwest part of thestudy area, a high-conductivity band trend-ing north-northeast across the centralpart of the study area, and another low-conductivity zone in the east-southeastpart of the study area.

Northwest low-conductivity zoneA triangular zone of relatively low

conductivity on the 900-Hz map is lo-cated in the northwest part of the Hatchelarea (fig. 19). Typical conductivities inthis zone are 130 to 180 mS/m. FourTDEM soundings were acquired at twosites within this area: three at site 76 andone at site 82 near the upstream end ofTurkey Creek.

Conductivity measured by the 900-Hzairborne coils is about 140 mS/m at site 76(fig. 19). The deeper parts of the con-ductivity models for the site-76 TDEMsoundings are similar (fig. 21), showing a7.8-ohm-m (128 mS/m) layer extendingfrom 10 m to 41 m below the surface thatis underlain by a conductive (2.3-ohm-mor 430-mS/m) layer below 41-m depth.

The 900-Hz coils, which sense to depthsof 40 to 50 m in this area, measure con-ductivity that is close to that of the 10- to41-m layer in the sounding models andare affected by only the upper part of theconductive layer beginning at 41 m andextending to an unknown depth. Ground-surface elevation at this site is 538 mabove sea level; the top of the conductivelayer is 497 m above sea level.

Mapped 900-Hz conductivity forthe TDEM sounding at site 82 is about150 mS/m (fig. 19). The conductivitymodel for sounding S82A (fig. 22) hasdeep layering similar to that at site 76: a rel-atively resistive (8.5-ohm-m or 118-mS/m)layer at depths of 15 to 37 m overlies aconductive (2.6-ohm-m or 384-mS/m)layer below 37-m depth. The conductivelayer is modeled at a depth shallower thanthat of site 76, which accounts for theslight increase in conductivity sensed bythe 900-Hz coils. With a ground-surfaceelevation of 530 m above sea level, thetop of the conductive layer at site 82 is493 m above sea level.

Relatively low conductivity sensed bythe 900-Hz airborne coils largely reflectsconductivities characteristic of a strati-graphic interval that extends to a depth ofabout 40 m in this area. The stratigraphic

28

FIGURE 21. (left) Transient decay and (right) resistivity models at TDEM soundings (a) S76A, (b) S76B, and (c) S76C at site 76(figs. 17 through 19; app. A). Symbols in the decay curve at left represent field measurements; the solid line through the datapoints at left represents the calculated decay of the resistivity model at right. The resistivity model at right that has the solid linerepresents the model fitting the field data best; the dashed lines at right represent models fitting the field data to within10 percent of the fit provided by the best model.

29

FIGURE 22. (left) Tran-sient decay and (right)resistivity models atIDEM sounding S82Aat site 82 (figs. 17through 19; app. A).

interval does not contain large volumes ofsaltwater. TDEM soundings detected a moreconductive stratigraphic interval deeperthan about 40 m that is likely to containsaltwater. This zone follows the regionalgeologic trend of west-northwestward deep-ening into the Permian Basin.

Central conductive zoneThe central conductive zone, as

depicted on the 900-Hz airborne map(fig. 19), trends north-northeastwardfrom the southwest corner of the Hatchelarea. Conductivities are high, rangingfrom 175 to 350 mS/m. At these conduc-tivities, exploration depth at 900 Hz is 30to 40 m (fig. 20). Two TDEM soundingswere located within the most conductiveparts of this trend: site 81, located in a cul-tivated field west of Hatchel, and site 83,near Bluff Creek at the north edge of thestudy area (fig. 19).

At site 81, conductivity measured bythe 900-Hz airborne coils is about 340 mS/m.TDEM sounding S81A indicated the presenceof a conductive (2.7-ohm-m or 365-mS/m)layer at a depth of 11 m that extends to 61 m(fig. 23). This thick layer encompasses

most of the effective exploration depth ofthe 900-Hz airborne coils and matchesthe measured conductivity well. A moreresistive (8.2-ohm-m or 123-mS/m) layerlies below 61-m depth. The top of the model’s50-m-thick conductive layer is 509 mabove sea level.

The conductivity profile that best fitsthe TDEM data at site 83 also indicates aconductive layer close to the land surface(fig. 24). Conductivity measured by the900-Hz airborne coils is about 300 mS/m,slightly lower than at site 81 (fig. 19). Theconductive layer detected by the soundingis also slightly less conductive (3.2 ohm-mor 313 mS/m) than the conductive layer atsite 81 and again matches the airbornemeasurements. The conductive layer ex-tends from 10 to 71 m below the surfaceand is underlain by a resistive (13.5-ohm-mor 74-mS/m) layer. The top of the conduc-tive layer is 514 m above sea level.

The main differences between con-ductivity models constructed for TDEMsoundings in this zone and those in thenorthwest low-conductivity zone are that(1) the conductive layer is detected atshallower depths and (2) the base of the

30

FIGURE 23. (left) Tran-sient decay and (right)resistivity models atIDEM sounding S81Aat site 81 (figs. 17through 19; app. A).

FIGURE 24. (left) Tran-sient decay and (right)resistivity models atTDEM sounding S83Aat site 83 (figs. 17through 19; app. A).

conductive layer is detected. Soundings top approaches the land surface on the south-in the central conductive zone are con-sistent with the interpretation that theconductive layer is a saltwater-bearinggeologic interval that is shallowing east-southeastward away from the PermianBasin. This unit is 50 to 60 m thick, and its

east margin of the central conductive zone.

East-southeastlow-conductivity zone

This zone is the largest of the threeregional zones, covering the east-southeast

31

FIGURE 25. (left) Tran-sient decay and (right)resistivity models atTDEM sounding S84Aat site 84 (figs. 17through 19; app. A).

part of the Hatchel area. Groundconductivities measured by the 900-Hzairborne coils are lowest here, rangingfrom 60 to 150 mS/m (fig. 19). Conse-quently, maximum exploration depths aregreatest (45 m in 150-mS/m terrain and70 m in 60-mS/m terrain, fig. 20).

Seven TDEM soundings were ac-quired at five sites in this zone. At site 84in the northeast corner of the study area(fig. 19), conductivity measured by the900-Hz airborne coils is 120 mS/m.A conductivity model constructed forsounding S84A (fig. 25) includes a layerhaving similar conductivity (7.4 ohm-m or136 mS/m) between 19- and 52-m depth,an interval that covers most of the explo-ration depth range of the 900-Hz airbornecoils. No highly conductive layer was de-tected in the sounding.

Soundings at sites 85 and 86 (figs. 26,27) are located at the base (S85A) and thetop (S86A) of a hill in Beddo field over-looking Elm Creek (fig. 19). Becausesounding S85A was acquired at lowerelevation, it explores to elevations lowerthan that of S86A. Conductivity measured

using the 900-Hz airborne coils is 100 mS/m,which falls between those of two relativelyresistive layers modeled for the sounding(fig. 26). The more resistive (86.8-ohm-mor 12-mS/m) layer lies between the depthsof 5 and 20 m and is underlain bya 7.5-ohm-m (133-mS/m) layer below20 m. The 900-Hz airborne coils senseboth of these layers. No highly conduc-tive layer was detected in either TDEMsounding.

Of the five sounding sites in this low-conductivity zone, conductivities mea-sured by the 900-Hz airborne coils are thelowest (about 80 mS/m) at site 17 (fig. 19).The deeper parts of three soundings ac-quired as part of a local anomaly investi-gation are similar (fig. 28). Representativesounding S17B (fig. 28b) shows the pres-ence of a relatively resistive (12.4-ohm-mor 81-mS/m) layer between the depths of11 and 39 m that corresponds well withobserved 900-Hz airborne coil measure-ments. No highly conductive layer otherthan the surface layer was detected. Theconductive layer at the surface is relatedto salinization at a pit.

32

FIGURE 26. (left)Transient decay and(right) resistivity mod-els at TDEM soundingS85A at site 85 (figs. 17through 19; app. A).

FIGURE 27. (left)Transient decay and(right) resistivity mod-els at TDEM soundingS86A at site 86 (figs. 17through 19; app. A).

Site 87 is located in the southeastcorner of the Hatchel area, where conduc-tivity measured by the 900-Hz airbornecoils is about 100 mS/m (fig. 19). Theconductivity model constructed forsounding S87A (fig. 29) indicates three rel-atively conductive layers within the depth

range sensed by the airborne coils. Themost conductive (6.2 ohm-m or 161 mS/m)of the three layers extends from 14- to 24-mdepth and is underlain by a more resistive(17.5-ohm-m or 57-mS/m) layer between24- and 42-m depth. Below 42 m is a layerof intermediate resistivity (9.4 ohm-m or

33

FIGURE 29. (left)Transient decay and(right) resistivity mod-els at TDEM soundingS87A at site 87 (figs. 17through 19; app. A).

107 mS/m). These three layers have acombined resistivity similar to that mea-sured by the airborne coils. No highlyconductive interval was detected.

The highly conductive, saltwater-bearing geologic unit that was detected atprogressively shallower depths to the eastand southeast across the northwest low-conductivity zone and the central conduc-tive zone is absent in the southeast zone.Considering the flatness of the studyarea and the west-northwestward dip ofPermian strata, part of the unit probablyreaches the surface and has been eroded,has discharged saltwater, or has been infil-trated by fresher water. Explorationdepths of the airborne and ground-basedinstruments were insufficient to reach theColeman Junction Formation, a knownbrine-bearing interval beneath the Hatchelarea.

Known Wells Having Conductivityand Magnetic Anomalies

Type-CMW sites, where a conductiv-ity anomaly (C), a magnetic anomaly (M),

and a known well location (W) coincide,were the most common of the 107 localconductivity anomalies identified from theairborne geophysical data (table 2; app. A).We investigated 34 of these sites usingground-based geophysical methods, fourof which are detailed next. The remainingsites are summarized in table 2 andappendix A.

Site 76Site 76, located in the northwest part

of the Hatchel area, is an anomalouslyconductive area on the shallow (56,000-Hz)and moderately deep (7,200-Hz) airborneconductivity maps, but not on the 900-Hzmap (figs. 17 through 19). The conductiv-ity anomaly measures about 120 m east-west and 200 m north-south on the56,000-Hz map and coincides with a mag-netic anomaly that is similar in size (fig. 16).The site encompasses the abandoned VancilNo. 1 well (fig. 30), which was drilledand completed in 1966 and plugged in1975. Thin Quaternary surficial depositscover Permian Clear Fork strata (fig. 2);soil is mapped as Portales clay loam (fig. 3).

35

FIGURE 30. Sketch map of site 76 (figs. 17 through 19;app. A). One reconnaissance conductivity profile (76A), twomultiple-coil-separation profiles (76B and 76C), and threeTDEM soundings (S76A, S76B, and S76C) were acquirednear the abandoned Vancil No. 1 well.

Ground investigations consisted of areconnaissance conductivity profile, twomultiple-coil-separation conductivity pro-files, and three TDEM soundings. The re-connaissance profile revealed that elevatedground conductivity extends about 130 macross the site from east to west for the20-m coil separation. This distance issimilar to that measured from the airbornesurvey maps. High conductivities for thehorizontal dipole mode (>100 mS/m) arefound near the well, with highest valueslocated 20 m west of the well.

Multiple-coil-separation profiles (fig. 31)were acquired along an east-west lineand a north-south line across the well todetermine vertical and lateral ground-conductivity patterns. Horizontal dipolemeasurements show well-defined conduc-tivity peaks (fig. 31a, b); vertical dipolemeasurements show characteristic reduc-

tions in apparent conductivity over themost conductive ground (McNeill,1980b). Highest ground conductivitiesmeasured in the horizontal dipole con-figuration on line 76B, the east-west line,are found for the 10- and 20-m coil sepa-rations (fig. 31a), indicating that theupper 6 to 12 m is more conductive thanthe 12- to 24-m depth range.

The highest conductivity for the 10-mcoil separation on line 76B is 10 m westof the abandoned well. For the 20-m sepa-ration, the peak is 10 m farther west.Highest conductivities for the 40-m sepa-ration, although lower than those for theshorter separations, cover a broader areaand are also centered about 20 m west ofthe well (downslope). East and west of thepeak, conductivities increase with coilseparation from 50 to 60 mS/m at 10-mseparation, to 70 to 80 mS/m at 20 mS/m,to 80 to 90 mS/m at 40-m separation.This increase suggests that, away from thepeak, conductivity increases with depthwithin the upper 20 m.

Similar trends are visible on thenorth-south line 76C (fig. 31b). Peaks ateach of the three separations have similarconductivities (~100 mS/m) and are allcentered within 5 m of the well. Peaksdrop off more quickly north of the well(upslope) for the 10- and 20-m coil sepa-rations than they do south of the well.

Horizontal dipole data were used toconstruct two-layer conductivity modelsalong lines 76B and 76C (fig. 32). Online 76B, a low-conductivity layer (23 to90 mS/m) overlies a layer having higherconductivities (144 to 211 mS/m) onthe east and west flanks of the abandonedwell (fig. 32a). The surficial low-conductivity layer thins from about 4 m ateach end of the line to less than 1 m inthickness near the well. At the well, thelow-conductivity layer is replaced by alayer 2 to 4 m thick that is more conduc-

36

FIGURE 32. Two-layer conductivity models that fit multiple-coil-separation data for (a) east-west line 76B and (b) north-south line 76C at site 76. Vancil No. 1 well is located at100 m on line 76B and at 90 m on line 76C.

tive than the layer below it. Two-layermodels constructed along north-southline 76C (fig. 32b) are similar: a poorlyconductive surface layer overlies a moreconductive layer and is a few meters thickaway from the well, is thinner toward thewell, and is replaced at the well by a layerthat is 1 to 2 m thick and that is moreconductive than the underlying layer.

We located TDEM soundings at threesites along line 76B (fig. 30) to examinedeeper variations in ground conductivity.Transients show decreasing apparentresistivity with time (fig. 21), suggestingthat conductivity increases with depth.Four-layer conductivity models providegood fits to the observed transients. Thesemodels (fig. 21) confirm that resistivity

decreases (or conductivity increases) ingeneral with depth, and they differ signifi-cantly only in the upper 10 m.

A thin surface layer that is more con-ductive than are underlying units thickensfrom less than 5 m at the most upslopesounding (S76A, fig. 21a) to 7 or 8 m atthe soundings closest to the well (S76B,fig. 21b) and downslope from the well(S76C, fig. 21c). This conductive surfacelayer, correlating with the lower conduc-tive layer detected on the multiple-coil-separation profiles, probably representssalinization adjacent to the well.

Despite a lack of surface evidence ofsaltwater leakage (no barren zone or saltwater at the surface), this well is probablyleaking or has leaked recently. Evidencesuggesting a recent or ongoing leak in-cludes conductivity anomalies visible on56,000- and 7,200-Hz airborne conduc-tivity maps, conductivity peaks centerednear the abandoned well for the 10-, 20-,and 40-m coil separations, a modeledconductive layer that shallows near thewell, and the presence of a near-surfacehigh-conductivity zone downslope of thewell. The most likely cause of elevatedground conductivity near the well is thepresence of saltwater. Background con-ductivity levels are reached within about50 m of the well, suggesting that littlesaltwater has migrated farther than thatdistance.

Site 34Site 34 is located south-southeast of

Hatchel (fig. 17) on a cultivated and ter-raced field. Portales clay-loam soil gradeswestward into thinner Mereta clay-loamsoil across the site (fig. 3). Caliche frag-ments are present at the surface where theclay loam is thinnest. These soils formedon thin Quaternary surficial deposits thatoverlie Clear Fork Group rocks (fig. 2).The site includes the location of the

38

FIGURE 33. Sketch map of site 34 (figs. 17 through 19;app. A). Reconnaissance conductivity profile 34A crossesthe Waller No. 1 dry hole.

Waller No. 1 well, a dry hole that wasdrilled and plugged in 1977 (fig. 33).

This site has a weak magneticanomaly (fig. 16) and anomalously highconductivity on the 56,000-Hz map(fig. 17). Conductivity is high on the7,200- and 900-Hz maps, but no higherthan that of adjacent areas (figs. 18, 19).A reconnaissance east-west conductivityprofile (fig. 33) shows relatively lowground conductivity in both dipole orien-tations (fig. 34). There is no conductivitypeak at the well, but ground conductivityin the horizontal dipole mode graduallyincreases from near 55 mS/m at the eastend of line 34A to 75 mS/m at the westend of the line. This gradual westward in-crease is also indicated on the 7,200- and900-Hz conductivity maps (figs. 18, 19).

Absence of high ground conductivi-ties near the well suggests that it is notleaking. The gradual westward increase inground conductivity coincides with awestward decrease in elevation, an obser-vation that supports an interpretation of

FIGURE 34. Apparent ground conductivity at site 34, mea-sured using 20-m coil separation and horizontal and verticaldipole orientations. Line 34A crosses the abandoned WallerNo. 1 well from east to west.

increasing conductivity related to a reduc-tion in depth of saline ground water.

Site 43Site 43 is located in the northwest

part of the Hatchel area near a tributary toCoyote Creek. The airborne survey revealsa shallow conductivity anomaly (fig. 17)that coincides with a magnetic anomaly(fig. 16) near a known well. The conduc-tivity anomaly does not appear on the7,200- or 900-Hz maps (figs. 18, 19).

The site lies in rangeland thatstraddles alluvial deposits of the CoyoteCreek tributary and Clear Fork Groupstrata (fig. 2). Soil is classified as moder-ately deep Portales clay loam (fig. 3). A re-connaissance conductivity profile (line 43A,fig. 35) was acquired across an elongatebarren area. The landowner reports thatthis barren area occasionally flows salt-water that runs off into the tributary andthat there is an abandoned well locatednear the barren area. Aerial photographstaken in 1940 show road access to the siteand a disturbed area. A buried well wasnot located by RRC staff during a metal-detector survey of the barren area.

39

FIGURE 35. Sketch map of site 43 (figs. 17 through 19;app. A). Reconnaissance conductivity profile 43A crosses abarren area on rangeland.

Line 43A shows a distinct horizontaldipole conductivity peak centered on thebarren area (fig. 36). The peak, which isnearly 150 m wide, has conductivities ashigh as 135 mS/m. These values are signi-ficantly higher than background conduc-tivities of 70 to 80 mS/m. Vertical dipoledata show characteristic lowering of appar-ent conductivity beneath the barren area.

Two possible interpretations of thegeophysical data at this site are that (1) anabandoned well at or near the barren areais leaking or has leaked brine into theshallow subsurface or (2) a natural salinespring periodically discharges. The broadconductivity peak on the ground-baseddata, the presence of a conductivity anom-aly on the shallow airborne data, surfaceevidence of brine leakage, aerial photo-graphic evidence of oil-field activity, andlandowner reports of an abandoned wellsuggest that a potentially leaking well isat the site.

FIGURE 36. Apparent ground conductivity at site 43, mea-sured using 20-m coil separation and horizontal and verticaldipole orientations. Line 43A crosses the barren area fromnorthwest to southeast.

Site 51Site 51 is south of Hatchel (fig. 17)

on an alluvial terrace west of CoyoteCreek. The active Cederholm B No. 2 wellwas drilled and completed on the terracein 1994. Deep, clayey Rowena and Tobosasoils grade eastward into deep Spur loamon the floodplain west of Coyote Creek(fig. 3). Clear Fork Group strata underlie thethick alluvial and eolian deposits (fig. 2).

There is a magnetic anomaly at thewell site (fig. 16). Conductivities at thesite are high on the 56,000-, 7,200-, and900-Hz maps (figs. 17 through 19) butare higher than at adjacent areas only onthe 56,000-Hz map. Conductivities forthe 900-Hz data are uniform at the site,whereas the 7,200-Hz map shows con-ductivity increasing eastward.

Reconnaissance profile 51A extendsnortheastward from higher to lower flood-plain elevations (fig. 37). This profilecrosses a barren area adjacent to the wellthat was disturbed during drilling andcompletion activities. Horizontal dipoleground conductivities are relatively highat this site but show no peak at either thebarren area or the well (fig. 38). Conduc-

40

FIGURE 37. Sketch map of site 51 (figs. 17 through 19;app. A). Reconnaissance conductivity profile 51A crossesthe Cederholm B No. 2 well and an adjacent barren area.

tivities decrease slightly from the upslope,southwest end of the line to the down-slope, northeast end. This trend is similarto that observed in the 7,200-Hz data(fig. 18).

Ground-based geophysical measure-ments suggest that this well is not leaking.The slight and gradual decrease in con-ductivity along line 51A toward CoyoteCreek is best explained as an effect ofdecreasing clay content from the thickRowena and Tobosa clay and clay-loamsoils at the southwest end of the line tothe thick, loamy Spur soils lower on theCoyote Creek floodplain at the northeastend of the line.

Conductivity andMagnetic Anomalies

Type-CM sites are those where a con-ductivity anomaly (C) and a magneticanomaly (M) coincide, but there is nomapped well. These sites are importantbecause they might indicate unknown ormislocated oil or gas wells that are leakingbrine. Ground-based geophysical methods

FIGURE 38. Apparent ground conductivity at site 51, mea-sured using 20-m coil separation and horizontal and verticaldipole orientations. Line 51A crosses the well from south-west to northeast.

were employed at 6 of the 15 CM sitesidentified to determine the cause of theairborne signatures. Two sites are discussednext in detail; the remaining sites aresummarized in table 2 and appendix A.

Site 17This site, located on the east side of

the Hatchel area along a tributary of MudCreek, is defined by a large conductivityanomaly on the 56,000-Hz map anda weak anomaly on the 7,200-Hz map(figs. 17, 18). The conductivity anomaly iselongate northwest-southeast and mea-sures about 200 × 400 m. There is a coin-cident magnetic anomaly (fig. 16), butconductivity on the 900-Hz map is notanomalously high (fig. 19).

Aerial photographs and site visitsreveal that the conductivity anomalyextends northwestward from a formerbrine-disposal pit northwest of the activeWilliams No. 1 well. Northwest of the pit(downslope toward the creek) is a barrenarea (figs. 39, 40). Soils are classified asthe moderately deep Valera-Vernon siltyclay, the shallow Kavett silty clay, or themoderately deep Tobosa clay (fig. 3).

41

FIGURE 39. Photograph of vegetation-kill area adjacent to abandoned brine-disposal pit, site 17.

These soils all overlie Clear Fork Grouplimestone (fig. 2).

Ground-based investigations con-sisted of two reconnaissance profiles, twomultiple-coil-separation profiles, and threeTDEM soundings (fig. 40). The recon-naissance profiles confirm the presence ofa conductivity peak centered on the pitand adjacent barren zone. Horizontaldipole conductivities increase from back-ground levels of 50 to 60 mS/m at distancesof 100 m northwest of the pit to 150 mS/m.The total width of the conductive zone,greater than 140 m northwest-southeast,is about 200 m northeast-southwest.

Multiple-coil-separation profiles 17C(fig. 41a) and 17D (fig. 41b) show that, ascoil separation (and, thus, explorationdepth) increases, ground conductivitydecreases for the horizontal dipole mode.Conductivity reaches a peak of 220 mS/mat the 10-m coil separation, 160 mS/m at

the 20-m separation, and 110 mS/m atthe 40-m separation. Peak conductivitieson line 17D, which crosses the pit andthe barren area from southeast to north-west, are offset to the northwest (down-slope) of the pit. Ground conductivityaway from the peaks on both lines showslittle variation with increasing coil separa-tion.

Decreasing ground conductivitiesaccompany increasing coil separation,implying that conductivity decreases withdepth near the pit. Three TDEM sound-ings support this observation. Thesesoundings have transients that are gener-ally similar (fig. 28), differing the mostat early times (or shallow depths). Deeperthan 10 to 20 m below the surface, resis-tivities are about 10 ohm-m. Nearer thesurface, thin conductive layers overlie amore resistive layer. The conductive layersare thickest (7 m) and least resistive at

42

FIGURE 40. Sketch map of site 17 (figs. 17 through 19;app. A). Two reconnaissance conductivity profiles (17A and17B), two multiple-coil-separation profiles (17C and 17D),and three TDEM soundings (S17A, S17B, and S17C) wereacquired near the surface pit.

S17A (fig. 28a), which was acquired clos-est to the pit and barren area. Thicknessesof the surface conductive layer decreaseto about 5 m, and resistivities increaseat soundings S17B (fig. 28b) and S17C(fig. 28c), which are farther from the pit.

Airborne data, ground surveys, andvisual observations support the interpreta-tion that the shallow subsurface is salin-ized at site 17 and saltwater has migratednorthwestward (downslope). Infiltrationdepths are probably 5 to 7 m. Decreasing

conductivities with depth imply a surfacesource. The presence of a pit having anadjacent barren area that coincides with aground-conductivity peak suggests thatsaltwater was discharged to the pit andsubsequently infiltrated the shallow sub-surface.

Site 71Site 71 is located near an abandoned

residence southeast of Hatchel. An anomalyis visible on the shallow (56,000-Hz) con-ductivity map (fig. 17) and the magnetic-field map (fig. 16). Maps of groundconductivity at deeper levels show slightlyanomalous values at 7,200 Hz (fig. 18)and background values at 900 Hz (fig. 19).No wells are mapped nearby. Thin Meretaclay loam and thicker Portales clay loamcover the area (fig. 3), which is cultivatedin places. These soils are formed on thinQuaternary deposits that cover Clear ForkGroup strata.

Two reconnaissance conductivity pro-files cross the site east-west (line 71A) andnorth-south (line 71C) (fig. 42). Line 71Ashows low, background conductivities(<50 mS/m) on the grassland at the eastend of the line, a horizontal dipole conduc-tivity peak of about 125 mS/m in the middleof the cultivated field, and a return to back-ground conductivity values westward(fig. 43a). Northward from the county roadon line 71C (fig. 43b), horizontal dipole con-ductivities increase to a peak of 125 mS/mand decline northward to background valuesof 50 to 60 mS/m. Peak conductivity on bothlines was found in a barren area. A metal-detector search by RRC staff at the barren area(fig. 9) failed to locate a well but did locatenumerous scattered metallic objects.

On geophysical evidence alone, onemight conclude that there is a buried wellnear the barren area. Aerial photographsshow that oil-field tanks were installed atthis site between 1940 and 1970 and have

43

FIGURE 42. Sketch map of site 71 (figs. 17 through 19;app. A). Reconnaissance conductivity profiles 71A and 71Ccross a barren area in a cultivated field near an abandonedresidence.

been removed. No evidence of a well wasfound, either on aerial photographs orduring site visits. Salinization at the siteprobably arose from spills or leakage fromthe tanks while they were operating.

Known Wells HavingConductivity Anomalies

Type-CW sites are those where a con-ductivity anomaly (C) and a known well(W) coincide, but there is no magneticanomaly. These sites are important be-cause the airborne signature combinedwith the knowledge that there is a well atthe site could be interpreted as presenceof an uncased well that might leak.Ground-based conductivity profiles wereacquired at 7 of the 12 CW sites identi-fied to determine the cause of the air-borne signatures. Three of these sites arediscussed next.

Site 12Located in the northeast part of the

Hatchel area, site 12 has two known wellswith no associated magnetic anomaly(fig. 16) and high conductivity anomalies

FIGURE 43. Apparent ground conductivity at site 71, mea-sured using 20-m coil separation and horizontal and verticaldipole orientations. (a) Line 71A crosses an abandoned oil-field facility, residence, and windmill from east to west.(b) Line 71B crosses the abandoned oil-field facility fromsouth to north.

on the 56,000- and 7,200-Hz airbornemaps (figs. 17, 18). No 900-Hz anomalyis apparent (fig. 19). Soils are classified asPortales clay loam and Potter clay loam(fig. 3), consisting of thin clay loam over-lying pedogenic carbonate. The soils haveformed on Permian Clear Fork Grouprocks (fig. 2).

Two reconnaissance geophysical pro-files were acquired at site 12 (fig. 44).Line 12A extends east-west across the

45

FIGURE 44. Sketch map of site 12 (figs. 17 through 19;app. A). Two reconnaissance conductivity profiles (12A and12B) cross the abandoned McMillan No. 6 and No. 7 wellsand adjacent surface pits.

abandoned McMillan No. 7 well. Line 12Bextends north-south across the No. 7 welland an adjacent brine-disposal pit, thenturns southwestward to cross the aban-doned McMillan No. 6 well and its dis-posal pit. RRC records indicate that theNo. 6 well was drilled in 1934 andplugged in 1942 and that the No. 7 wellwas drilled in 1935.

No conductivity peaks were found cen-tered on either of the two wells at site 12(fig. 45). For the horizontal dipole data at20-m coil separation, the highest conduc-tivities were just above 100 mS/m in twoareas: 30 m east of well No. 7 on line 12A(fig. 45a) and 35 to 45 m south of theNo. 7 well on line 12B (fig. 45b). Thebroad conductive zone south of wellNo. 7 extends 90 m downslope (south) ofthe well. Highest conductivities in thiszone are adjacent to the No. 7 well pit.Horizontal dipole ground conductivitiesnear the McMillan No. 6 well are relativelylow (50 to 60 mS/m) both near the welland across the adjacent pit (fig. 45b).

FIGURE 45. Apparent ground conductivity at site 12, mea-sured using 20-m coil separation and horizontal and verticaldipole orientations. (a) line 12A crosses abandoned McMillanNo. 7 well from east to west. (b) Line 12B crosses McMillanNo. 7 and No. 6 wells from north to south.

Neither well has surface evidence ofleakage. Well No. 6, which RRC recordsshow to have been plugged, has no ground-based geophysical evidence of leakage andis probably not leaking. Elevated conduc-tivities near well No. 7, which are visible inboth airborne and ground-based geo-physical data, suggest that the ground issalinized near this well. The presence ofa conductivity peak adjacent to the pitrather than at the well implies that salt-

46

FIGURE 46. Sketch map of site 16 (figs. 17 through 19;app. A). Reconnaissance conductivity profiles 16A and 16Bpass near the abandoned McMillan No. 3 well in a cultivatedfield south of Mud Creek.

water discharged into the pit is the likelysource of elevated conductivities.

Site 16Site 16 contains an abandoned well

in a cultivated field south of Mud Creek.The site is associated with a shallowanomaly on the 56,000-Hz map (fig. 17)that weakens on the 7,200-Hz map(fig. 18) and is absent on the 900-Hz map(fig. 19). There is no magnetic anomaly.Ground investigations revealed the pres-ence of concrete blocks and well debrisassociated with the McMillan No. 3 well,which was drilled in 1928 and abandonedin 1930. Deep Olton clay-loam soil coversthe site and overlies Clear Fork Groupstrata.

Two conductivity profiles pass nearthe abandoned well (fig. 46). A horizontaldipole conductivity peak more than 100 mwide is present on north-south line 16A(fig. 47a). The peak, which reaches conduc-tivities of 100 mS/m, is centered within afew meters of well No. 3. The east-west line16B reveals a slightly broader peak of simi-lar magnitude, but the highest values are

FIGURE 47. Apparent ground conductivity at site 16, mea-sured using 20-m coil separation and horizontal and verticaldipole orientations. Lines (a) 16A and (b) 16B cross theabandoned McMillan No. 3 well.

centered 20 to 50 m west (downslope) ofthe well (fig. 47b). Background conduc-tivities of 40 to 50 mS/m are reachedwithin 160 m west of the well.

The airborne conductivity and mag-netic signature suggest an uncased wellthat might be leaking. Well casing wasdetected during the ground surveys de-spite the lack of a magnetic anomaly inthe airborne survey. Saltwater is in theshallow surface near the well that was de-tected by both the airborne and ground-based instruments. Conductivity profilessuggest that either (1) the well is leakingsaltwater that is migrating to the west

47

FIGURE 48. Sketch map of site 73 (figs. 17 through 19;app. A). Reconnaissance conductivity profiles 73A and73B pass near the abandoned Russell No. 1 well and anadjacent pit.

(downslope) or (2) the pit was locatedwest of the well and brine that was dis-charged into it has infiltrated the subsur-face. RRC reentered and plugged this wellin 1997 after the airborne and ground-based investigations. Saltwater flow wasencountered during reentry.

Site 73This site surrounds the Russell No. 1

well, which was drilled and abandoned in1928. The site, located west of Flag Creekin the southwest corner of the Hatchelarea, has thin Potter clay-loam soil overClear Fork Group strata. Conductivitiesare high and slightly anomalous on the56,000-Hz airborne map (fig. 17). Thedeeper maps neither indicate anomalouslyhigh conductivities (figs. 18, 19) nor evi-dence a magnetic anomaly (fig. 16).A metal-detector survey by RRC stafflocated the well.

Two conductivity profiles were ac-quired across the site (fig. 48). Line 73Areveals a conductivity peak in the hori-

FIGURE 49. Apparent ground conductivity at site 73,measured using 20-m coil separation and horizontal andvertical dipole orientations. (a) Line 73A crosses the aban-doned Russell No. 1 well and pit from northwest to south-east. (b) Line 73B crosses the well and pit from northeastto southwest.

zontal dipole mode that is 110 m wide(fig. 49a). Conductivities increase frombackground values of 50 mS/m to nearly100 mS/m. The peak is centered at a pit10 to 20 m southeast (downslope) of thewell. Line 73 B, which passes between thewell and the pit (fig. 48), measured con-ductivities as high as 80 mS/m near thewell and pit (fig. 49b).

The location of peak conductivities atthe pit rather than the well and the ab-sence of anomalous conductivity on the

48

FIGURE 50. Photograph of abandoned Early A No. 2 well leak-ing saltwater at site 67.

7,200- and 900-Hz maps support an in-terpretation that this well is not leaking.Despite the passage of more than 70 yrsince abandonment in 1928, the saliniza-tion appears to be related to saltwater dis-charge into the pit during drilling oroperation.

Other Local AnomaliesIn the course of field investigation, a

few other site types were examined. Forexample, there are numerous type-C siteswhere conductivity anomalies (C) visibleon the airborne maps do not coincidewith magnetic anomalies (M) or well lo-cations (W). There are also many type-MW sites, where magnetic anomalies

FIGURE 51. Sketch map of sites 65 and 67 (figs. 17 through19; app. A). Reconnaissance conductivity profiles 6567A and6567B pass near the leaking Early No. 1 and Early A No. 2wells in a cultivated field upslope of Bluff Creek.

associated with known wells have no coin-cident conductivity anomaly. Most of thewells in the study area are in this category,suggesting no evidence of extensive salin-ization at these wells.

Sites 65 and 67These are adjacent type-MW sites in

the northeast part of the Hatchel area(fig. 17). Site 65 consists of a magneticanomaly associated with the Early No. 1well, a recently abandoned oil well thatwas drilled in 1982. Site 67 is a separatemagnetic anomaly encircling the Early ANo. 2 well, another recently abandonedoil well also drilled in 1982. Both wellsrecently began leaking brine at the surface(fig. 50) and were plugged in 1996 duringthis study. These wells are in a cultivatedfield on deep Miles sandy-loam soil justupslope from Bluff Creek (fig. 51). Thesoil grades downward into Clear ForkGroup strata and Quaternary alluvium.

49

FIGURE 52. Apparent ground conductivity at site 65, mea-sured using 20-m coil separation and horizontal and verticaldipole orientations. (a) Line 6567A crosses the abandonedand leaking wells from east to west. (b) Line 6567B crossesEarly A No. 2 from south to north.

The airborne survey shows no con-ductivity anomaly at either site. On the56,000-Hz map, conductivities are gener-ally low and decrease eastward toward thecreek (fig. 17). On the deeper, 7,200- and900-Hz maps, conductivities are highand they increase toward the creek, al-though they show no anomalies at thewells (figs. 18, 19).

Ground-based measurements verifythe absence of significant conductivityanomalies. Line 6567A, which passes nearboth wells (fig. 51), shows a gradual in-crease in conductivity (horizontal dipole

FIGURE 53. Histogram of TDS in ground-, surface-, and oil-field-water samples in the Hatchel area (app. B). Mean valuesof surface and ground waters are similar.

mode) eastward and a modest increase inconductivity near the Early A No. 2 well(fig. 52a). There is no evidence of any in-crease in conductivity near the No. 1 well.The modest peak near the No. 2 well isabout 50 m across and reaches peak val-ues of 80 mS/m. Line 6567B, passing nearthe No. 2 well along a north-south line,shows only a small ground-conductivityincrease near the well (fig. 52b).

These wells were leaking, but air-borne and ground-based geophysical datashow that little salinization occurred be-fore the wells were plugged. Leakage be-gan recently and has occurred at low ratesthrough the top of the surface casing.

Chemical Composition ofWater Samples

Water SalinityTotal dissolved solids (TDS) in

ground-water samples (app. B) range from480 to 7,620 mg/L (fig. 53). TDS, orsalinity, of water generally follows a loga-rithmically normal distribution. The best

50

FIGURE 54. Relation between electrical conductivity and(a) TDS and Cl content of water samples and (b) Cl contentof soil samples having linear least-squares regression lines(apps. B, C). Electrical conductivity is a good predictor ofCl and TDS.

estimate of average TDS, 1,746 mg/L, isbased on the logarithm of TDS values.TDS does not significantly differ with welldepth. Surface-water samples have nearlythe same range and mean in TDS (fig. 53;app. B).

The three oil-field samples have mark-edly higher salinity (app. B) than surfaceon ground water. TDS of the two samplesfrom abandoned oil wells (sample sites 27and 28, fig. 15) was about 66,000 mg/L.Water collected from the seeping core holeon Elm Creek (sample site 26, fig. 15) had aTDS of more than 10,200 mg/L.

TDS of water samples (fig. 54a) andchloride content of soil (fig. 54b; app. C)

FIGURE 55. Chemical composition of hydrochemical facies inground-, surface-, and oil-field-water samples in the Hatchelarea (app. B).

correlate with electrical conductivity mea-sured on the samples.

Chemical CompositionGround water and surface water in the

Hatchel area have similar chemical composi-tions that include calcium-bicarbonate,calcium-sulfate, mixed-cation—mixed-anion,and sodium-chloride hydrochemical facies(Back, 1966) (fig. 55). In this area, we canassume that the calcium-bicarbonate hy-drochemical facies is derived from the reac-tion of ground water flowing throughlimestone and that the calcium-sulfatetype results from ground water encounter-ing gypsum. The Permian formations aredominated by limestone, gypsum, andshale. A change from calcium- to sodium-dominated cations can take place along aflow path because of ion exchange or be-cause of mixing with sodium-chloridewater. The water samples in which nosingle cation or anion dominates mostlikely reflect mixtures of water of different

51

FIGURE 56. Relationship between Br:Cl ratio and Cl concen-tration in water samples collected during this study (pluses,circles, and squares), compared with regional data presentedby Richter and others (1990). Ground- and surface-watersamples collected in the Hatchel area resemble ground waterelsewhere in the region. The oil-field water sampled in thisstudy resembles the Permian (Pr) more closely than it doesthe Pennsylvanian (Pn) end member of oil-field brine definedby Richter and others (1990).

FIGURE 57. Relationship between Cl:SO4 ratios and Na:Caratios in water samples collected during this study (pluses,circles, and squares), compared with regional data presentedby Richter and others (1990). Ground- and surface-watersamples collected in the Hatchel area are similar to groundwater elsewhere in the region. The oil-field water sampled inthis study resembles the Permian (Pr) more closely than itdoes the Pennsylvanian (Pn) end member of oil-field brinedefined by Richter and others (1990). Some surface-watersamples have a high Na and Cl content similar to that of thePermian brine end member.

hydrochemical facies, movement of waterfrom one lithology to another, or samplestaken at intermediate positions along aflow path.

Samples from the two abandoned oilwells (samples 27, 28; fig. 15; app. B)have a sodium-chloride hydrochemical fa-cies typical of oil-field waters. The dilutedwater from the unplugged core hole hashigher proportions of calcium and sulfate(fig. 55).

Comparison withRegional Trends

Dutton and others (1989) and Richterand others (1990) studied salinity sourcesaffecting ground-water quality in andnear the Hatchel area. The ground- andsurface-water samples collected for thisstudy do not differ from ground water de-scribed in the earlier regional study. For

example, the ratio of bromide:chloride(Br:Cl) of the new samples overlaps thatof the older ground- and surface-watersamples (fig. 56). The new samples alsooverlap the regional samples on a crossplot of the chloride:sulfate (Cl:SO4) ratioand the sodium:calcium (Na:Ca) ratio(fig. 57). Dutton and others (1989) andRichter and others (1990) identified thesegraphical techniques as particularly effec-tive in distinguishing among differentsources of water. Note that the surfaceand ground water lie along a well-definedtrend on the Cl:SO4 versus Na:Ca graph(fig. 57). The trend “points” toward thePermian formation end member of oil-fieldwater defined by Dutton and others(1989) and Richter and others (1990). Thenew oil-field samples plot at that Permianformation end member (figs. 56, 57).

52

FIGURE 58. Comparison of TDS of ground-water samplesand ground conductivity measured by (a) 7,200- and(b) 56,000-Hz surveys.

Comparison of Soil andWater Salinity withMeasured Conductivity

Although the sample size is small,ground conductivity measured by the56,000- and 7,200-Hz airborne coilsgenerally increases with increasingground-water TDS (fig. 58; app. B). The56,000-Hz ground-conductivity surveydoes not correspond well to chloride con-tent or electrical conductivity measured insoil samples (fig. 59; app. C). There are

FIGURE 59. Comparison of measurement of ground con-ductivity by airborne geophysical survey (56,000 Hz) andmeasurement of (a) Cl content and (b) electrical conductivityof soil samples.

several reasons for this poor correlation,including the difficulty in preparing soilsamples for electrical conductivity mea-surements, the difference between theelectrical conductivity of a large soil vol-ume sensed by the airborne coils and avery small volume of soil measured in thelaboratory, and the effect of in situ soilmoisture on measured conductivity (Paineand others, 1998b).

53

DISCUSSIONHigh-resolution airborne geophysical

measurements, when combined with de-tailed ground-based geophysical surveys,water soil sampling and chemical analyses,information on soil type, and rock typeand oil- and gas-well distribution, providean excellent opportunity for understandingground-conductivity changes and theircauses. Specifically these data from theHatchel area allow us to distinguish oil-fieldand non-oil-field causes of salinization, as-sess the merits of ground-based and air-borne geophysical methods in the study ofsalinization, determine the usefulness ofthe method in discriminating amongcauses of oil-field salinization, and developa geophysical profile of a potentially leakingwell.

Effectiveness ofAirborne Geophysics

As the ground surveys at anomaliesdetected by the airborne survey show, theeffectiveness of the airborne geophysicalmethod depends on the objective. If thegoal is to identify areas of salinization,then the airborne method is obviouslysuccessful. It measures ground conductiv-ity accurately, locates anomalies withadequate precision, determines the lateralextent and intensity of salinization, and il-lustrates lateral and vertical conductivitychanges adequately to determine whethersalinization is likely due to natural or oil-field causes. If the goal is to locate oil andgas wells with an airborne magnetometer,this study demonstrates that wells can bedetected this way, but that some magneticanomalies are small enough to be missedif the line spacing is too coarse. If the goalis to distinguish leaking wells from otheroil-field-salinity sources, then the results

are mixed. The airborne method increasesthe likelihood of locating leaking wells,but it cannot distinguish leaking wells frombrine pits in all cases and cannot distinguishleaking wells from those that were onceleaking and subsequently were plugged.

An understanding of the airborneconductivity data, combined with aknowledge of oil-field history, qualitativeplume-migration rates, and land use,allows us to distinguish conductivityanomalies related to agricultural practices,soil and rock types, and soil moisturefrom those that are oil-field related. Wecan infer that most oil-field salinizationwill be manifested as point sources show-ing limited lateral migration (typically tensto hundreds of meters). Conversely, long,curvilinear conductivity anomalies thatfollow topography are likely to be relatedto agricultural practices, natural hillsideseeps, or water-saturated alluvium nearflowing creeks. Large conductivity anoma-lies covering several square kilometersare most likely related to changes in soilor rock type. The airborne method is quitesuccessful in identifying these broadsource types.

Ground investigations are requiredto help researchers using the airbornemethod to distinguish among commonoil-field-salinity sources. Airborne signa-tures of leaking tank batteries, brine-disposal pits, and leaking wells are similar,requiring follow-up work to verify thelikely source type. In the Hatchel area,where more than 700 known wells havebeen drilled, 107 conductivity anomalieswere identified in the airborne data thatare consistent with oil-field salinization.Ground-based surveys suggest that most areindeed sites with oil-field-related saliniza-tion (42 of the 54 conductivity anomaliesvisited), but fewer are attributable to po-

54

tentially leaking wells (22 of 54 anomaliesvisited).

Most (99) of the conductivity anoma-lies were identified on the 56,000-Hz con-ductivity map, which is the map thatshows the shallowest exploration depth(fig. 17; table 2). Many of these anomaliesappear to be related to surface sourcessuch as brine pits or tank batteries. As ex-ploration depth increases, fewer sites haveanomalous conductivities. Only 56 of the107 sites show anomalous conductivitieson the 7,200-Hz map, only 48 showanomalous conductivities on both the7,200- and 56,000-Hz maps, and only 13show anomalous conductivities at allthree frequencies (table 2). Ground sur-veys indicate that anomalies visible onboth the 56,000- and 7,200-Hz mapswere more likely to be potentially leakingwells than anomalies visible only on the56,000-Hz map.

Geophysical Profileof a Leaking Well

Ground investigation of airborne geo-physical signatures allows development ofa profile of a well that may be leaking salt-water. Sites that have a suspicious air-borne signature are those that (1) haveeither a magnetic anomaly or a knownwell location and (2) have anomalouslyhigh ground conductivity on both theshallow-exploring 56,000-Hz airbornecoils and the deep-exploring 7,200-Hzcoils (table 2; app. A). The suspicious sitemight have a conductivity high on mapsproduced from the deeply penetrating900-Hz coils.

In the Hatchel area, 39 of the 107sites fit the airborne profile of a poten-tially leaking well (table 2). Ground sur-veys at 34 of these sites show that at least29 of them have significant oil-fieldsalinization and that 20 might be actively

or formerly leaking wells. The 20 profile-matching, potentially leaking wells areCMW (16), CM (1), or CW (3) sites. Ofthe four surveyed sites having conductiv-ity and magnetic anomalies but no knownwell (a potentially leaking unknown well),one was a potentially leaking well, onewas a feedlot, one was metal debris at aseep, and one was a tank or pit.

Ground investigations reduce thenumber of anomalous sites to a list ofwells that might be leaking or might havehad significant leaks in the past. Theseinvestigations should include precise lo-cation of the well and reconnaissanceground-conductivity profiles that crossthe well site and any other potential salt-water sources, such as pits or tank batter-ies. If a well has been leaking brine intothe shallow subsurface, measured groundconductivities should increase by a factorof two or three from background levels tomaximum levels recorded within a fewmeters of the well. Wells that have begunto leak recently may show only a smallanomaly at the shallowest investigationdepths; conversely, wells that have beenleaking longer will have the highest con-ductivity peaks and the broadest areas ofinfluence. Further information on the lat-eral extent and depth of salinization canbe obtained by using more sophisticatedgeophysical surveys, such as those em-ploying multiple coil separations andTDEM soundings.

Utility of Airborne andGround-Based Geophysics

Airborne geophysics measuredmagnetic-field strength and ground con-ductivity that allowed us to focus ourground investigations. The airborne mag-netometer located many, but not all, wellsin the study area; it also located somewells that were absent from, or mislocated

55

on, RRC maps. The magnetic data servebest as a reconnaissance tool in areaswhere well locations are unknown orinaccurate. Typical measured magneticanomaly widths of 80 to 200 m suggestthat, as flight-line spacing lengthensbeyond the 100 m used in this study,progressively more well-related magneticanomalies will be missed. The 100-mspacing was a compromise between sur-vey cost, which is roughly proportional toline length flown, and the percentage ofwells detected. Closer flight lines arerequired to detect all wells that have cas-ings. Wells having no casings cannotbe detected by using airborne or ground-based magnetometers.

Airborne EM instruments mapped thelateral and vertical conductivity distribu-tion to depths of a few tens of meters. Theconductivity data acquired at multiple fre-quencies allowed us to discriminate be-tween near-surface causes of elevatedconductivity, such as brine-disposal pits,and deeper causes, such as brine-bearinggeologic formations. Used alone, each coilfrequency had limitations. Data collectedby the 56,000-Hz airborne coils showedthe most detail, but some of the anoma-lies had non-oil-field causes. Conversely,the deep data collected by the 900-Hzcoils showed conductivity changes thatare probably related to the subsurface dis-tribution of geologic units. Distinct anom-alies were not, however, always visiblenear wells that ground surveys suggest areleaking (table 2; app. A). The single mostuseful coil frequency was 7,200 Hz, whichwas insensitive to most shallow conduc-tivity anomalies related to soils, soil mois-ture, brine pits, and surface-water bodies,as well as deeper natural patterns.

When data from each of the coil fre-quencies are combined, they provide apowerful method of visualizing lateraland vertical conductivity changes. Pat-

terns evident at individual frequenciesallow us to distinguish source types andmap salinization extent and intensity,whereas differing exploration depthsamong the frequencies allow us to distin-guish surface salinity sources from sub-surface ones. By combining lateral andvertical conductivity information, we canestimate chloride mass within saltwaterplumes (Paine and others, 1998a).

Ground-based geophysical surveyscomplement the airborne data, forming acritical component in the understandingof ground-conductivity changes in generaland in the search for potentially leakingwells in particular. Metal-detector searchesof suspected well sites that were chosenfrom the airborne geophysical signatureallowed precise location of abandonedwells. Reconnaissance conductivity pro-files across these well sites allowed usto determine whether the conductivityanomaly was likely to be related to salt-water from a leaking well or to saltwaterthat entered the subsurface through anearby pit or surface spill. Multiple-coil-separation surveys and TDEM soundingsenabled us to delineate the lateral andvertical extent of salinization. The TDEMsoundings also extended the maximumexploration depth of the airborne survey,located a subsurface saltwater-bearingunit, and helped us better understandconductivity patterns visible on thedeeper, 900-Hz conductivity maps.

Without the ground-based geophysi-cal surveys, it is possible to determinesalinization extent and intensity and dis-tinguish oil-field from natural salinization,but it is difficult to distinguish amongthe various oil-field sources. Without theairborne data, ground-based surveys canhelp determine whether known wellsmight be leaking and may be used as arelatively inexpensive method of testingwhether a well is leaking saltwater into

56

the shallow subsurface. With ground-based geophysics alone, however, un-known wells and unsuspected salinizationarea would most likely be missed and site-specific conductivity, soil, and water datawould lack a broader context.

No matter how successful the air-borne method is judged to be, its applica-tion is hindered by cost. Mobilizationcosts preclude its use in small areas of afew tens of square kilometers or less,where ground-based methods are moreappropriate. Furthermore, short flight-linespacings required to locate small conduc-tivity and magnetic anomalies make themethod uneconomical in excessively largeareas. A realistic upper limit might be atypical oil-field size of a few tens to a fewhundreds of square kilometers. Areaslarger than a given oil field would involve

surveying non-oil-field land for oil-fieldsalinization. In addition to survey cost,the airborne data commonly require follow-up, ground-based investigations to distin-guish likely leaking wells from otheroil-field sources.

In summary, ground-based surveyingof known wells is a relatively inexpensivemeans of determining wells that shouldbe plugged in small areas (a few tens ofsquare kilometers or less). Over large areas(oil-field size) where well locations areuncertain or the extent of salinization isunknown, an airborne survey should beconsidered because it identifies salinizedground, determines the lateral and verticalextent and intensity of salinization, distin-guishes natural from oil-field salinitysources, and focuses ground investigations.

CONCLUSIONSAnalysis of airborne and ground-

based geophysical data and soil- and water-quality data in the Hatchel area of WestTexas shows that

airborne methods accurately por-tray location, size, and magnitudeof conductivity and magnetic-fieldanomalies;airborne instruments can gener-ally be used to distinguish oil-field from natural and agriculturalsources of salinization, particu-larly if well locations are known;salinization causes include up-ward migration of brine alongnatural pathways, seepage of salt-water into the subsurface at brine-disposal pits, brine discharge atabandoned wells and coreholesthat are either unplugged or not

plugged in compliance with pres-ent standards, and evaporativeconcentration of chloride in shal-low ground water;airborne geophysical data alonecan be used to identify wells thathave evidence of salinization, butground investigations are requiredto reliably distinguish surface spillsand pit discharge from wells thatare potentially leaking saltwater;persistence of salinization in arid,near-surface environments makesformerly leaking wells geophysi-cally indistinguishable from leak-ing wells;airborne instruments are bestused in relatively large areas whereit is likely that there are multipleunknown sources and it is impor-

57

tant to determine salinization ex-tent and intensity, source areas, orvolume of infiltrated saline water;ground-based instruments arebest used in surveying small areashaving few sources, determiningwhether specific known wellsmight be leaking, and verifyingairborne geophysical data;

chemical analyses of soil and wa-ter can help differentiate betweennatural, agricultural, and oil-fieldcauses of salinization; andground-water salinity correlatesbetter than soil salinity with mea-sured ground conductivity.

ACKNOWLEDGMENTSThis project was funded by the Railroad

Commission of Texas (RRC) under ContractsIAC 94-0443 and IAC 96-0034, Jeffrey G.Paine, Principal Investigator. The LowerColorado River Authority (LCRA) contributedfinancially to the project. Interpretations andconclusions presented in this report are notnecessarily those of the RRC, LCRA, or theColorado River Municipal Water District(CRMWD).

Airborne geophysical data were acquiredand processed by David Miles, Megan Sheffer,Mark Nicholas, and Douglas Game of Geoterrex-Dighem. Staff from LCRA and CRMWDsampled surface and ground water during theairborne survey and provided chemical analyses.

Jose Mayorga and Randall Ross of RRC,Geoff Saunders of LCRA, Okla Thornton ofCRMWD, and Brace Smith of the U.S. Geologi-

cal Survey made many suggestions that helpedus complete the project. Isack Ramirez andRobert Luxton of the San Angelo District RRCoffice provided invaluable assistance in search-ing well records, locating wells, contacting land-owners, and facilitating project progress.

Bureau staff Thomas A. Tremblay andErica R. Boghici conducted GIS analyses ofdata; Steven Tweedy collected and analyzedsoil and water samples; and Ianthe Nelson re-searched the history of oil and gas explorationin the Hatchel area. The report was reviewed byJames C. Gibeaut, Rebecca C. Smyth, James L.Simmons, and Jay A. Raney. Jamie H. Coggindrafted the figures, did the typesetting, anddesigned the cover and layout under the direc-tion of Joel L. Lardon, Graphics Manager.Susan Lloyd did the word processing, and LanaDieterich edited the report.

REFERENCESAbilene Geological Society, 1992, The stratigraphic distribu-

tion of hydrocarbon production from 19 counties in theAbilene area: Abilene, Texas, variously paginated.

American Association of Petroleum Geologists, 1973,Geological highway map of Texas: U.S. geological highwaymap of Texas: U.S. Geological Highway Map Series,Map No. 7, scale 1:500,000.

Back, W., 1966, Hydrochemical facies and ground-water flowpatterns in the northern part of the Atlantic coastal plain:U.S. Geological Survey Professional Paper 498-A, 42 p.

Bein, Amos, and Dutton, A. R., 1993, Origin, distribution,and movement of brine in the Permian Basin (U.S.A.):a model for displacement of connate brine: GeologicalSociety of America Bulletin, v. 105, no. 6, p. 695–707.

Bennett, J. E., Collins, C. L., and Colby, R. J., 1985, Aero-magnetic map of Texas: Bendix Field EngineeringCorporation, scale 1:750,000.

Brown, E., Skougstad, M. W., and Fishman, M. J., 1970,Methods for collection and analysis of water samples fordissolved minerals and gases: U.S. Geological Survey

58

Techniques of Water-Resources Investigations, book 5,chapter Al, 160 p.

Brown, L. E., Jr., Goodson, J. L, and Harwood, Peggy, 1972,Abilene sheet: The University of Texas at Austin, Bureau ofEconomic Geology, Geologic Atlas of Texas, scale1:250,000.

Core Laboratories, Inc., 1972, A survey of the subsurfacesaline water of Texas: Texas Water Development BoardReport 157, v. 1, 12 p. + illustrations.

Dutton, A. R., Richter, B. C, and Kreitler, C. W., 1989, Brinedischarge and salinization, Concho River watershed, WestTexas: Ground Water, v. 27, no. 3, p. 375–383.

Eifler, G. K., Jr., 1975, San Angelo sheet: The University ofTexas at Austin, Bureau of Economic Geology, GeologicAtlas of Texas: scale 1:250,000.

Fitzgerald, N. D., 1952, The Beddo field, Runnels County,Texas: Abilene Geological Society Contributions, p. 7–8.

Frischknecht, F. C., Labson, V. F., Spies, B. R., and Anderson,W. L., 1991, Profiling using small sources, in Nabighian,M. N., ed., Electromagnetic methods in applied geophysics—applications, part A and part B: Tulsa, Oklahoma, Societyof Exploration Geophysicists, p. 105–270.

Gardner, F. J., and Phifer, R. L., 1953, The oil and gas fieldsof west Texas, part 1: Railroad Commission District 7-C:Five Star Oil Report, p. 148–198.

Garrie, D. G., 1996, DIGHEMV survey for Bureau ofEconomic Geology, University of Texas, Hatchel study area,Runnels County, Texas: Mississauga, Ontario, Dighem,Report 625, variously paginated.

Geonics Limited, 1992, Protem 47 operating manual:Mississauga, Ontario, Geonics Limited, variously paginated.

Jorgensen, D. G., Downey, Joe, Dutton, A. R., and Maclay,R. W., 1988, Region 16, central nonglaciated plains, inBack, William, Rosenshein, J. S., and Seaber, P R., eds.,Hydrogeology: Geological Society of America, The Geol-ogy of North America, v. O-2, p. 141–156.

Kaufman, A. A., and Keller, G. V., 1983, Frequency and tran-sient soundings: Amsterdam, Elsevier, Methods inGeochemistry and Geophysics, no. 16, 685 p.

Kier, R. S., Brown, L. F., Jr., and Harwood, P., 1976,Brownwood sheet: The University of Texas at Austin,Bureau of Economic Geology, Geologic Atlas of Texas,scale 1:250,000.

Lee, J. N., 1986, Shallow ground-water conditions, Tom GreenCounty, Texas: U.S. Geological Survey, Water ResourcesInvestigations Report 86-4177, 88 p.

McNeal, R. P., 1965, Hydrodynamics of the Permian Basin:American Association of Petroleum Geologists, Memoir 4,p. 308–326.

McNeill, J. D., 1980a, Electrical conductivity of soils androcks: Mississauga, Ontario, Geonics Limited, TechnicalNote TN-5, 22 p.

_______ 1980b, Electromagnetic terrain conductivitymeasurement at low induction numbers: Mississauga,Ontario, Geonics Limited, Technical Note TN-6, 15 p.

_______ 1980c, EM34-3 survey interpretation tech-niques: Mississauga, Ontario, Geonics Limited, TechnicalNote TN-8, 16 p.

Paine, J. G., Dutton, A. R., Hovorka, S. D., Blüm, M. U.,Mahoney, M. P., and Sullivan, E. J., 1998a, Brine in thenear-surface environment: determining salinization extent,identifying sources, and estimating chloride mass usingsurface, borehole, and airborne EM, in Bell, R. S., Powers,M. H., and Larson, Timothy, eds., Proceedings, Sympo-sium on the Application of Geophysics to Environmentaland Engineering Problems: Environmental and Engineer-ing Geophysical Society, p. 215–219.

Paine, J. G., Goldsmith, R. S., and Scanlon, B. R., 1998b,Electrical conductivity and gamma-ray response to clay,water, and chloride content in fissured sediments, Trans-Pecos, Texas: Environmental and Engineering Geoscience,v. 4, no. 2, p. 225–239.

Parasnis, D. S., 1973, Mining geophysics: Amsterdam, Elsevier,395 p.

Rhoades, J. D., 1981, Predicting bulk soil electrical conduc-tivity versus saturation paste extract electrical conductivitycalibrations from soil properties: Soil Science Society ofAmerica Journal, v. 45, p. 42–44.

Richter, B. C., Dutton, A. R., and Kreitler, C. W., 1990,Identification of sources and mechanisms of salt-waterpollution affecting ground-water quality: a case study,West Texas: The University of Texas at Austin, Bureau ofEconomic Geology Report of Investigations No. 191, 43 p.

Richter, B. C., and Kreitler, C. W., 1986, Geochemistry of saltwater beneath the Rolling Plains, North-Central Texas:Ground Water, v. 24, no. 6, p. 735–742.

Spies, B. R., and Frischknecht, F. C., 1991, Electromagneticsounding, in Nabighian, M. N., ed., Electromagneticmethods in applied geophysics—applications, part Aand part B: Tulsa, Oklahoma, Society of ExplorationGeophysicists, p. 285–386.

Tóth, J., 1962, A theory of groundwater motion in smalldrainage basins in central Alberta: Journal of GeophysicalResearch, v. 67, no. 11, p. 4375–4387.

West, G. F., and Macnae, J. C., 1991, Physics of the electro-magnetic induction exploration method, in Nabighian,M. N., ed., Electromagnetic methods in applied geophysics—applications, part A and part B: Tulsa, Oklahoma, Societyof Exploration Geophysicists, p. 5–45.

Wiedenfeld, C. C., Bamhill, L. J., and Novosad, C. J., 1970,Soil survey of Runnels County, Texas: U.S. Department ofAgriculture, Soil Conservation Service, 60 p.

Wood, W.W., 1976, Guidelines for collection and fieldanalysis of ground-water samples for selected unstableconstituents: U.S. Geological Survey, Techniques ofWater-Resources Investigations, book 1, chapter D2, 24 p.

59