GEOPHYSICS, VOL. 60, NO.6 (MAY-JUNE 1995); P. 629--630.

Special Issue

Crosswell methods: Where are we, where are we going?

James W. Rector III, Editor*

FOREWORD

This special issue on crosswell methods was constructedin response to the large number of crosswell papers submit­ted during 1993 and 1994. The editorial staff felt that ratherthan distributing these papers throughout the year the soci­ety would be better served by having an archival volumecomposed completely of crosswell-related papers. This wasnot a publicly solicited issue, and we undoubtedly did notreceive some excellent papers that may have been submittedwith a public solicitation. Nevertheless, we feel that thisvolume is an excellent snapshot of crosswell technology as itstood in 1993-1994.

This issue is primarily composed of papers related tocrosswell seismic methods, although there are also a numberof papers on crosswell electromagnetic imaging. The issuecontains a wide variety of topics, from case histories tomodeling and inversion algorithms. A common theme in theseismic methods section is the implementation of full-wave­form analysis and imaging algorithms on real data. Mostearlier papers on crosswell seismic imaging of real data haveused only the direct arrival traveltime (e.g., Lines andLaFehr, 1989) and most published studies of full-waveformimaging have been used on model data (e.g., Pratt andGoulty, 1991). This issue contains the first high-resolutioncrosswell reflection images from oil fields. The issue alsocontains some of the first elastic inversions of real crosswelldata. These real data results demonstrate the fundamentaladvantages and potential of crosswell seismic imaging indetecting and imaging features below the resolution of sur­face seismic data.

The interest in crosswell methods has also extended toelectromagnetic methods for imaging interwell electricalconductivity. The sensitivity of conductivity to porosity,pore fluid type, saturation, and temperature has led to thedevelopment of crosswell electromagnetic systems and im-

aging algorithms. Prototype systems described in this issueoffer useful images for well spacings up to several hundredmeters, and field examples and numerical simulations showremarkably good resolution of interwell features as com­pared to surface electromagnetic techniques.

As recently as five years ago, seismic crosswell tomogra­phy was a hot technology at the SEG annual meetings. Mostof the major oil companies had active research programs andsubstantial budgets. Since 1990, research budgets have beenreduced, and crosswell research has suffered a dispropor­tionate share of the reduction. The initial excitement incrosswell seismic imaging can probably be attributed toseveral factors. Laboratory results from the late 1970s andearly 1980s (Ito et aI., 1979; Wang and Nur, 1988) showedthat P-wave velocities were strongly affected by steamphases and temperature. At the same time, steam floodingand thermal stimulation were being used to enhance produc­tion.

Led by an extensive project at Mobil, monitoring of steamfronts by imaging the P-wave velocity distribution was oneof the first applications of crosswell seismic imaging. Cross­well seismic imaging of steam floods is represented byseveral papers in this issue (Mathisen et aI.; Lee et aI.).Crosswell was considered to be the geometry of choice forsteam flood imaging because velocities derived from thesurface were thought to have insufficient resolution. In theSan Joaquin Valley of California, where much of the initialcrosswell seismic fieldwork was performed, there weremany closely spaced wells, and surface seismic data wasoften difficult to obtain and of poor quality. In addition,computerized tomographic imaging, initially developed inthe medical field (Kak and Slaney, 1985), was receivingsubstantial attention in geophysics as a means of estimatingsubsurface velocity distributions. Of all the seismic geome­tries for data acquisition, crosswell was most like the geom­etry used in medical imaging. Crosswell direct-arrival trav­eltime tomography produced robust images of velocity thatwere interpreted in terms of steam distribution (e.g., Justice

'Department of Material Science and Mineral Engineering, University of California at Berkeley, 557 Evans Hall, Berkeley, CA 94720.© 1995 Society of Exploration Geophysicists. All rights reserved.

629

Dow

nloa

ded

11/2

5/14

to 1

34.7

1.13

5.19

1. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

630 Rector

et aI., 1989). However, some of the initial interest in direct­arrival traveltime tomography as a high-resolution imagingtool waned as investigators discovered that without externalconstraints the resolution of the technique was fundamen­tally limited by recording aperture and the Fresnel zone ofthe direct wave (Williamson and Worthington, 1993; Rectorand Washbourne, 1994). In particular, the lack of aperture atthe base of the wells reduced the potential applicationbreadth of crosswell direct-arrival traveltime imaging. Someof the more recent tomographic inversion algorithms incor­porate the effect of aperture and Fresnel zone, and producetomograms that are quite smooth (Schuster and Quintus­Bosz, 1993; Vasco and Majer, 1993). It is interesting to notethat the resolution of crosswell electromagnetic methods isof the same order as the smoothed seismic tomographyresults.

Along with developments in tomography and rock phys­ics, downhole equipment development, and in particulardownhole seismic source development projects producedequipment that could be used to collect field data. Downholeseismic sources such as the air gun, the hydraulic vibrator(Paulsson, 1989), piezoelectric transducers, and controlledexplosives were all deployed initially in the late 1980s. Thecost of performing crosswell surveys was also reducedthrough the development of multilevel geophone and hydro­phone strings. Multichannel recording systems capable ofsampling intervals of 0.2 ms or less were also fielded initiallyduring this period.

Perhaps the most important result from the early period ofcrosswell research was the discovery that very high frequen­cies could be propagated over distances of several hundredmeters when both source and receiver were located in deepboreholes (e.g., Harris, 1988). As shown in one ofthe papersin this issue (Lee et al.), P-waves with frequencies of up to2000 Hz have been propagated over path lengths of morethan 2000 ft (600 m). The discovery of very high bandwidthwith downhole sources and receivers confirmed that most ofthe frequency loss in surface seismic profiling was a result ofpropagation through the near surface. Although VSP re­moves one leg of this near-surface propagation, removingboth legs is necessary for very high bandwidth in mostenvironments. The increased bandwidth available fromcrosswell seismic profiling translates into over an order ofmagnitude improvement in subsurface resolution. Layersthat are at or below the tuning thickness of surface seismicdata can be resolved with crosswell seismic techniques.However, to exploit the bandwidth of crosswell seismicdata, components of the "scattered" wavefield had to beextracted and imaged. As shown in this issue, many cross­well investigators are now using secondary arrivals such asreflections to obtain very high-resolution subsurface images.

The progression from first-arrival imaging to full-waveformimaging is similar to the progression in surface seismic,where refraction imaging was gradually replaced by reflec­tion imaging. In VSP, the checkshot survey led to VSPreflection images.

The future of crosswell seismic imaging is bright. There isno other technology on the horizon that promises the poten­tial resolution of crosswell seismic imaging. Although 3-Dseismic has revolutionized geophysical exploration, the fun­damental band-width and resolution limitations of the tech­nique can be overcome only by going downhole. The 2-Dnature of crosswell is overshadowed by the improvement inresolution obtained. As crosswell survey costs drop andimage reliability improves, we will undoubtedly see a grow­ing use of crosswell seismic data. The electromagneticmethods are in an early stage of development. At present thesystems are designed for open-hole or plastic casing, andonly preliminary measurements have been done in steel­cased wells. The attenuation of high frequencies by thecasing will require low frequencies of operation, highertransmitter moments, and casing corrections. All these seemwell along in development, and several groups are movingquickly to realize such a system.

REFERENCES

Harris, J. H., 1988, Crosswell seismic measurements in sedimentaryrocks: 58th Ann. Internat. Mtg., Soc. Expl. Geophys., ExpandedAbstracts, 147-150.

Ito, H., DeVilbiss, J., and Nur, A., 1979, Compressional and shearwaves in saturated rock during water-steam transition, J. ofGeophys. Res., 84, 4731-4735.

Justice, J. H., Vasilliou, A. A., Singh, S., Logel, J. D., Hansen,P. A., Hall, B. R., Hutt, P. R., and Solanki, J. 1., 1989, Acoustictomography for monitoring enhanced oil recovery: The LeadingEdge, 7, 6, 12-19.

Kak, A. C., and Slaney, M., 1985, Principles of computerizedtomographic imaging, IEEE Press.

Lines, L. R., and LaFehr, E. D., 1989, Tomographic modeling of across borehole data set: Geophysics, 54, 1249-1257.

Paulsson, B. N. P., 1989, Three-component downhole seismicvibrator: 58th Ann. Internat. Mtg. Soc. Expl. Geophys., Ex­panded Abstracts, 139-142.

Pratt, R. G., and Goulty, N. R., 1991, Combining wave-equationimaging with traveltime tomography to form high resolutionimages from crosshole data: Geophysics, 56, 2, 208-225.

Rector, J. W., and Washbourne, J., 1994, Characterization ofresolution and uniqueness in cross-well direct arrival traveltimetomography using the Fourier projection slice theorem: Geophys­ics, 59, 11, 1056--1064.

Schuster, G. T., and Quintus-Bosz, A., 1993, Wavepath eikonaltraveltime inversion: Theory, Geophysics, 58, 1314-1323.

Vasco, D. W., and Majer, E. L., 1993, Wavepath traveltimetomography, Geophys. J. Int., 115, 1055-1069.

Wang, Z., and Nur, A., 1988, Effect of temperature on wavevelocities in sands and sandstones with heavy hydrocarbons, SPEReservoir Engineering, 3, I, 158-164.

Williamson, P. R., and Worthington, M. H., 1993, Resolution limitsin ray tomography due to wave behavior: Numerical experiments:Geophysics, 58, 727-736.

Dow

nloa

ded

11/2

5/14

to 1

34.7

1.13

5.19

1. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/