USGS Groundwater Information: Branch of Geophysics
by J.W. Lane Jr., F.P. Haeni, Gary
U.S. Geological Survey, 11 Sherman Place, U-5015, Storrs, CT 06103
and David L. Wright
U.S. Geological Survey, MS 964, Box 25046, Federal Center, Denver, CO 80225
Cross-hole tomography and single-hole reflection borehole-radar surveys were conducted at the U.S. Geological Survey's Fractured Rock Research Site at Mirror Lake, Grafton County, New Hampshire, USA, using electric dipole omni-directional and magnetic dipole directional antennas to detect the presence of a saline tracer in fractured crystalline bedrock. The tracer was used to enhance the radar's ability to determine the location and orientation of permeable fractures.
The well field used in this study consisted of four boreholes that approximate a square 9 m (meters) on a side. Injection and recovery boreholes were located along a diagonal of the square, approximately 13m apart. Two observation boreholes were located along the other diagonal of the square. A hydraulically conductive zone, known to connect the injection and recovery boreholes, was isolated with straddle packers. A sodium chloride (NaCl) solution was continuously injected into the isolated zone to achieve and maintain a steady-state concentration along the tracer transport path while borehole-radar data were collected.
Cross-hole tomography surveys were conducted with two borehole-radar systems. Data collected before and during tracer injection were subtracted to produce difference tomograms that delineate high-attenuation zones caused by the spatial distribution of the saline tracer in permeable fracture zones. Single-hole reflection surveys showed significant changes in the reflection amplitude, frequency content, and reflector continuity between the background surveys and those collected during tracer injection. A qualitative ranking of reflectors with the most significant changes in reflection amplitude and continuity was used to identify those reflectors that represent permeable fractures between the injection and recovery boreholes.
Fluid movement and chemical transport in crystalline rock is largely controlled by the hydraulic properties of the fractures within the rock. Characterization of the hydrologic properties of a fracture system and the effect of these properties on ground-water flow and transport in a bedrock aquifer is an important component of water supply and ground-water contamination studies.
Borehole-radar methods can provide information about the extent and orientation of fractures and fracture zones and can be used to detect fractures and other structures not intersected by the borehole. Cross-hole tomography methods and single-hole radar reflection methods (using omni- and directional-receivers) have been used to detect and map fractures and fracture zones for ground-water investigations (Gaylor and others, 1994; Lane and others, 1994; Hansen and Lane, 1995). Borehole-radar methods can maximize the information obtained from a single borehole and reduce the number of boreholes needed to characterize a fractured-bedrock aquifer.
Cross-hole tomography methods can delineate velocity or attenuation anomalies caused by fractures and fracture zones, and single-hole reflection methods can detect reflections from fractures and fracture zones. Geologic changes such as lithologic contrasts, foliation, or alteration can also be detected with borehole-radar methods. Integrating the interpretation of borehole-radar data with the results of borehole-geophysical logs, such as acoustic televiewer, borehole video, and heat-pulse flowmeter, can help distinguish between velocity and attenuation anomalies and reflections induced by lithologic changes and those caused by fractures.
Hydrologic properties of a fracture or fracture zone are determined using hydraulic and tracer testing. Because fracture systems are complex, unique models of ground-water flow and solute transport are difficult to obtain from hydraulic or tracer tests alone. Other information is needed to accurately describe the flow geometry, boundaries, and heterogeneity (Hsieh and others, 1993). Integrating a tracer test, using an electrically conductive tracer, with borehole-radar methods, which could monitor the movement of the tracer through the fracture system, could constrain the flow and transport models by providing the geometry of the permeable fracture system.
Use of borehole-radar methods to detect a saline tracer capitalizes on the sensitivity of electromagnetic waves to the electric properties of the medium through which it propagates. Introducing a saline, electrically conductive solution into the fracture system changes the electric properties of the water in the fractures. Changes in the electrical properties will affect both cross-hole tomography and reflection measurements. For cross-hole radar surveys, transmission of an electromagnetic wave through fractured bedrock filled with tracer will cause additional wave attenuation. Amplitude data collected before and during tracer injection can be subtracted. Tomographic inversion of these residual data will produce a difference attenuation tomogram that outlines permeable fractures filled with tracer. Saline tracer tomography has been successfully used to identify permeable zones and interpret transport paths in several studies (Ramirez and Lytle, 1986; Niva and others, 1988; Olsson and others, 1992; Kong and others, 1994).
The effect of the injection of a saline tracer into a fracture on electromagnetic reflection measurements can be approximated using GPRMODV2 (Powers and Olhoeft, 1995) or a simplified plane-wave, normal-incidence, parallel-planar reflection model such as presented by Barber and Morey (1994). Reflections from tracer-filled fractures will have a higher amplitude than reflections from fractures where tracer is absent.
This paper presents the results of experiments using borehole-radar methods, including cross-hole tomography and single-hole reflection surveys, during a tracer-injection test to delineate permeable fractures along the tracer flowpath.
The field experiment to test the applicability of the borehole-radar methods for saline tracer detection was conducted in September 1995 and was funded by the U.S. Geological Survey's (USGS) Toxic Substances Hydrology Program. The site is located in the U.S. Forest Service Hubbard Brook Experimental Forest in the Mirror Lake area near West Thornton, New Hampshire.
The Mirror Lake site is located at the lower end of the Hubbard Brook valley in the southern part of the White Mountains of New Hampshire (fig. 1). Fracture orientation is highly variable, with a concentration of observed strike azimuths at 30° E and fracture dips at 7° NW, 50° SE, and 82° SE (Barton, 1993). The fracture system is poorly connected, which indicates fluid flow in the fracture system is heterogeneous, tortuous, and follows highly channeled flowpaths (Barton, 1993). The average horizontal hydraulic conductivity of the upper 100 m of bedrock is 3 x 10S-7 m/s (meters per second) but ranges over several orders of magnitude (Hsieh and others, 1993). Fracture mineralization includes iron-oxide coatings on the fracture surfaces that have penetrated as much as 1 m into the bedrock. These iron oxide precipitates may induce a halo of anomalously high electrical conductivity surrounding permeable zones.
Figure 1: Location of study area and FSE well field, U.S. Geological Survey's Fractured Rock Research site, Mirror Lake, Grafton County, N.H., USA.
The experiment was conducted at the FSE 1-4 cluster of the FSE wellfield at the USGS's Fractured Rock Hydrology Test Site. Approximately 20m of glacial drift overlie the bedrock at the well cluster. The bedrock consists mostly of granite and smaller amounts of schist, intruded by pegmatite and aplite dikes (C.D. Johnson, U.S. Geological Survey, written commun., 1994). The boreholes are completed with steel casing into rock, with open hole to the drilled depths. The borehole locations approximate a square, with a side length of 9 m (fig. 2).
Figure 2: FSE-1 to FSE-4 borehole arrangement and experiment design.
The boreholes have been studied using borehole-geophysical methods, hydraulic tests, and tracer tests. These studies have determined that two approximately horizontal, hydraulically conductive zones connect the boreholes. The upper zone is near the bottom of casing and has hydraulic conductivities from 10-7 to 10-6 m/s. The lower zone is approximately 40 m below the top of casing (TOC) and has hydraulic conductivities from 10-6 to 10-5 m/s (P.A. Hsieh, U.S. Geological Survey, written commun., 1994) (fig. 2). Although the hydraulically conductive zones are horizontal, the fractures in the boreholes are not. Results from acoustic televiewer (F.L. Paillet, U.S. Geological Survey, written commun., 1994), borehole video (C.D. Johnson, U.S. Geological Survey, written commun., 1994) and previous single-hole directional radar surveys indicate that the fractures in the boreholes are steeply dipping. The main concentration of fractures strike to the northeast, with dips greater than 45°, which is consistent with the observations of fractures exposed on the surface (Barton, 1993). Therefore, the hydraulic connection between the boreholes is probably controlled by interconnections between the steeply dipping fractures. This implies that the saline tracer must travel through multiple fractures of differing orientations to move from the injection to the extraction borehole, offering multiple opportunities to image the tracer using borehole-radar methods.
During the tracer test, the lower hydraulically conductive zone between FSE-1 and FSE-4 was isolated using straddle packers. The packed-off interval in FSE-4 was pumped at about 9L/min (liter per minute) while a NaCl solution with concentrations of 20 to 28 g/L (grams per liter) was injected into the packed-off interval in FSE-1 at about L/min. The experiment involved the repeated, near-continuous injection of NaCl tracer during the collection of borehole-radar data. Radar data were collected before and during tracer injection to establish the `background' conditions and to measure any changes resulting from the presence of the tracer.
Radar data were collected using the RAMAC1 system and a USGS system. The RAMAC system used 60-MHz (megahertz) transmitting and receiving electric dipole antennas and a 60-MHz directional magnetic-dipole antenna (Falk, 1992). The USGS system used 60-MHz electric dipole antennas. Cross-hole surveys were conducted between FSE-3 and FSE-2 with the RAMAC and USGS systems before and during steady-state tracer injection. Cross-hole surveys between FSE-1 and FSE-4 were conducted only with the USGS system. Single-hole reflection surveys were undertaken before and after tracer injection in FSE-1, FSE-2, and FSE-3.
CROSS-HOLE TOMOGRAPHY SURVEYS
Cross-hole data were collected from about 20m to 65 m below TOC on a 2 m-by-2 m grid using the RAMAC system. The total grid resulted in about 700 separate raypaths. Data-collection procedures for the USGS system are described in Wright and others (1996). Although about 6,000 rays were collected for each data set with the USGS system, only about 3,000 rays were used for the tomographic inversion.
The cross-hole tomography data collected with the different systems were interpreted separately. Data collected with the RAMAC system were interpreted with TOMOCG (Ivansson, 1984). Difference tomograms were produced by subtracting residual amplitude data collected before and during tracer injection (Niva and others, 1988) (fig. 3). The data collected with the USGS system were interpreted with a conjugate-gradient method developed by the USGS (Ellefsen, 1995), and difference tomograms were produced using a dual frequency band-amplitude-ratio method (Wright and others, 1996) (fig. 4).
Figure 3: 60 MHz borehole-radar attenuation tomography between FSE-3 and FSE-2, showing the sequence of tomograms produced from data sets collected with the RAMAC system before and during tracer injection, and the resultant difference tomogram.
Figure 4: 60 MHz difference attenuation tomogram produced from data collected with the U.S. Geological Survey's borehole-radar system between FSE-1 and FSE-4.
The most attenuative anomaly in the background tomogram between FSE-3 and FSE-2 is near the top of the tomogram, which correlates with the upper hydraulically conductive zone. The magnitude of this anomaly indicates that conductive minerals or iron compounds are present in the upper zone. Although isolated anomalies are present near both boreholes and may be related to zones of increased fracture porosity, a continuous, horizontal, high-attenuation anomaly linking FSE-2 and FSE-3 at the depth of the lower hydraulically conductive zone is not present.
The difference tomograms show attenuation anomalies that are interpreted as saline tracer in permeable fractures within the plane of the tomograms. A linear attenuation feature extends from the upper part of the tomogram near FSE-3 to a high-attenuation anomaly near FSE-2 between 42 and 49m below TOC. This anomaly is interpreted as a fracture zone connecting the lower and upper hydraulically conductive zones. Other attenuation anomalies observed near FSE-3 are smaller and of a lower magnitude than the attenuation anomalies near FSE-2, indicating FSE-3 is not well connected to the lower hydraulically conductive zone. This is supported by water-level data, which showed that the response of FSE-3 lagged that of the other boreholes when FSE-4 was pumped (A.M. Shapiro, U.S. Geological Survey, written commun., 1995).
The FSE-1 to FSE-4 difference tomogram (fig. 4) shows that high attenuation anomalies are not continuous from FSE-1 to FSE-4, indicating the tracer path moves out of the plane of the tomogram toward FSE-2.
Specific-conductance measurements made in the boreholes during tracer injection support the interpretation that the pattern of attenuation anomalies represents the distribution of saline tracer and are not due to antenna loading caused by tracer invasion into the boreholes.
SINGLE-HOLE REFLECTION SURVEYS
Reflection data were collected every 0.5 m from about 20 to 70 m. The data were interpreted using the RAMAC software RADINTER. An adaptive subtraction filtering algorithm was developed to analyze the difference in reflection data collected before and during tracer injection. The results from FSE-1, FSE-2, and FSE-3 show that the injection of a saline tracer has a measurable effect on the amplitude, continuity, and frequency content of reflections from fractures (fig. 5). After tracer injection, the reflector amplitudes were generally higher, more laterally continuous, and contained more low-frequency energy than before tracer injection.
Figure 5: Azimuthal images of 60 MHz magnetic dipole directional data from FSE-1, FSE-2, and FSE-3 after the application of a moving average filter showing examples of the effects of the injection of a saline tracer on reflector amplitude, frequency content, and lateral continuity.
The orientation and location of reflectors that had the most substantial increases in reflection amplitude and continuity are interpreted as the most permeable fractures transporting the tracer. The most substantial changes were observed in the data from FSE-1 and FSE-2.; therefore, most tracer transport involves fractures located near or intersecting these two boreholes. The changes in or near FSE-3 are not as large as those observed in the data from FSE-1 and FSE-2, indicating that fractures near FSE-3 are either not highly permeable or are not well connected to the other permeable fractures.
Based on the single-hole reflection data, it is interpreted that 12 permeable fractures transported tracer from the injected zone in FSE-1 to the extracted zone in FSE-4 (table 1). This interpretation, which is consistent with the cross-hole tomography surveys and the acoustic televiewer and oriented video logs, involves the intersection and hydraulic connection of at least five fractures in and near the injection zone, allowing the tracer to partition into the upper and lower hydraulically conductive zones. The tracer moved through at least five fractures near FSE-2 and was extracted from fractures or fracture zones in FSE-4 that join the upper and lower hydraulically conductive zones.
Table 1: Radar reflectors interpreted to be the primary permeable fractures for transport of tracer from FSE-1 to FSE-4
[Interval hydraulic conductivity data from PA. Hsieh, written commun., 1994; acoustic televiewer data from F.L. Paillet, written commun., 1994; borehole video data from C.D. Johnson, written commun., 1994]
Borehole-radar methods, including cross-hole tomography and single-hole reflection surveys, were successfully used to detect a saline tracer and interpret permeable fractures in the FSE 1-4 well cluster at Mirror Lake, Grafton County, New Hampshire, USA.
The tomograms and reflection surveys successfully detected the presence of the saline tracer and showed that the saline tracer moved from FSE-1 to FSE-4 through at least 12 permeable fractures. The tracer was transported between two hydraulically conductive intervals known to connect the FSE-1 to FSE-4 boreholes.
This experiment demonstrates that borehole-radar methods can be integrated with saline-tracer experiments to provide insights into the nature of fluid flow and chemical transport in fractured crystalline bedrock and that a saline tracer has a measurable effect on the amplitude, continuity, and frequency content of reflections from fractures.
1Use of tradenames is for identification purposes only and does not constitute endorsement by the U.S. Geological Survey.
Barber, W. B., and Morey, R., 1994. Radar detection of thin layers of hydrocarbon contamination. Proceedings, GPR Ontario, Canada, pp. 1215-1228.
Barton, C.C., 1993. Characterizing bedrock fractures in outcrop for studies of ground-water hydrology: an example from Mirror Lake, Grafton County, New Hampshire. Morganwalp, D.W., ed., U.S. Geological Survey Toxic Substances Hydrology Program: Proceedings, Technical Meeting, Sept. 1993, Colorado Springs, Colorado, 16 pp.
Ellefsen, 1995, High resolution mapping of stratigraphy using seismic tomography--a feasibility test at the M-Area Basin, Savannah River site, South Carolina. U.S. Geological Survey Open-File Report 95-518. 46 pp.
Falk, L., 1992. Directional borehole antenna--Theory. Stripa Project 92-16, SKB, Stockholm, Sweden, 138 pp.
Gaylor, R., Lieblich, D.A., Mariani, D., and Xia, J., 1994. Environmental logging with a borehole radar tool. Proceedings, Second Int. Conf. on Ultra-Wideband Short-Pulse Electromagnetics, Polytechnic Univ., New York, April 5-7, 20 pp.
Hansen, B.P., and Lane, J.W., 1995. Use of surface and borehole geophysical surveys to determine fracture orientation and other site characteristics in crystalline bedrock terrain, Millville and Uxbridge, Massachusetts. U.S. Geological Survey Water-Resources Investigations Report 95-4121 25 pp.
Hsieh, P.A., Shapiro, A.M., Barton, C.C., Haeni, F.P., Johnson, C.D., Martin, C.W., Paillet, F.L., Winter, T.C., and Wright, D.L., 1993. Methods of characterizing fluid movement and chemical transport in fractured rock. Cheney, J.T. and Hepburn, J.C., eds., Field Trip Guidebook for the Northeastern United States, Boston, Mass., Oct. 25-28, pp. R1-R29.
Ivansson, S., 1984, Crosshole investigations--tomography and its application to crosshole seismic measurements. Stockholm, Sweden, Stripa Project IR-84-08.
Kong, F., Westerdahl, H., By, T.L., and Kitterod, N., 1994. Radar tomography for environmental geotechnology: field and simulation tests. Proceedings, GPR `94, Fifth Int. Conf. on Ground Penetrating Radar, Kitchener, Ontario, Canada, June 12-16, pp. 1249-1260.
Lane, J.W., Haeni, F.P., and Williams, J.H., 1994. Detection of bedrock fractures and lithologic changes using borehole radar at selected sites. Proceedings, GPR `94, Fifth Int. Conf. on Ground Penetrating Radar, Kitchener, Ontario, Canada, June 12-16, pp. 577-592.
Niva, B., Olsson, O., and Blumling, P., 1988. Radar crosshole tomography at the Grimsel Rock Laboratory with application to migration of saline tracer through fracture zones. Nationale Genossenschaft fur die lagerung radioaktiver Abfalle, NTB 88-31.
Olsson, O., Andersson, P., Carlsten, S., Falk, L., Niva, B., and Sandberg, E., 1992. Fracture characterization in crystalline rocks by borehole radar. Pilon, J., ed., Ground Penetrating Radar, Geological Survey of Canada Paper 90-4, pp. 139-150.
Powers, M.H., and Olhoeft, G.R., 1995. GPRMODV2--One-dimensional full waveform forward modeling of dispersive ground-penetrating radar data. U.S. Geological Survey Open-File Report 95-58, 41 pp. Ramirez, A.L., and Lytle, R.J., 1986. Investigation of fracture flow paths using alterant geophysical tomography. Journal of Rock Mechanics and Mining Sciences and Geomechanics, Vol. 23, No. 2, pp. 165-169.
Wright, D.L., Grover, T.P., Ellefsen, K.J., Lane, J.W., Jr., and Kase, P.G., 1996. Radar tomograms at Mirror Lake, New Hampshire--3D visualization and a brine tracer experiment. Bell, R.S., and Cramer, M.H., eds., Symposium on the Application of Geophysics to Engineering and Environmental Problems, Keystone, Colo., April 28-May 2, 1996, Environmental and Engineering Geophysical Society, pp. 565-575.
Final copy as submitted to Sixth International Conference on Ground-Penetrating Radar (GPR'96) for publication as: Lane, J.W., Haeni, F.P., Placzek, G., and Wright, D.L., 1996, Use of borehole-radar methods to detect a saline tracer in fractured crystalline bedrock at Mirror Lake, Grafton County, New Hampshire, USA, in Sixth International Conference on Ground-Penetrating Radar (GPR'96), Sendai, Japan, September 30-October 3, 1996, Proceedings: Sendai, Japan, Tohoku University Department of Geoscience and Technology, p. 185-190.