CANADIAN JOURNAL OF EXPLORATION GEOPHYSlCS YOL 32. N”~ 1 IJUNE 1996), P. 6~23

CROSS-HOLE SEISMIC AND SINGLE-HOLE GEOPHYSICAL SURVEYS TO CHARACTERIZE AN AREA OF MODERATELY FRACTURED GRANITE’

J.G. HAYLE.S~. M.H. SEW’, G.S. LODHA~, R.A. EVERIT? AND D.K. TOMSONS*

ABSTRACT

Ceophy\ical surveyr were done in hurehukr MF~, and MF-2 drilled “ndrrgro”nd ill AECL‘\ “ndergr”“nd Keieilrch Lab”rar”ry ,“RI.~ in,,1 il region of l”demlely fractured grimia. Rorehnle MF~I is 220 m in length and was dnllcd from ,hr 130 111 shaft station in the URL ill a” uimutil of 191. and plunging at about 1Y. B”reh”k bw2 is 210 m in length and was drilled fmn thr 240 Level ilt the lame arilwth a\ MF-, but WiCh a shil,,owrr plunge of atn,ur Y” The bore holes ilre mughly parallel IO he large-scale c”mp”siti”nal taping in dle granite. The *iswnce between he ,w<, hmhdc Cdlilp. is 108 m and the hulrs cunvrrgr 10 wide 7x m of each other at heir mds. mc nrughly recrang”lar region between there boreh”lrs “lw\“rcs about 2 1” m by WI 111. The p,ilnr ddi”.?d by ,k arcil between ,he hulrr *trike* N ItrE and dips at ah,“, 82w Fu,,~wwrf),r,” smic. smic uelocity. S”“iC ampli,“dc. “il,“Gll gamma and caliper logs were rccordcd in h”lh holes. fM0wed by a cross-hale seismic rurvry. txrween the holes using AECL‘S cross-tiolr Audio-frequency RlEk Testing systmn KHARTS~. CHARTS was designed specifically fir cross~hole seismic romogn@lic swvrys in gmitc i,“d has ken in use fc,r ,m yrilrr. The CHARTS tranrlniner generates a signal centre* ilf irhout 44X) Hr. which pmduces 8~ seismic signal wavrlmgth of ahut I.1 m for P-waves in the granite. About 1780 gmd-a,-fair quality P- wile arnvalr were obtained tr<,m the cr‘xGx,,e rurvq Tk loqest raypaths are -220 m. S-wave mivillr ilrr poor to nmcxistent. Stmng P-waves were ah recorded late in romr of rhe crusr~hole seismic records: these were caused by mude wnvrrsionr of ruk waves in hurehule MP- I at a” open fracture.

There is an anirotmpy in the P-wave velocity &ta that is mm or less constant throughout the area uf rrudy. Vertically travelling wil~es are ahout 4%’ faster than those travelling horizontally. The minimum P~wave velocity ii probably caused hy vrnicill (0 nears Yeltical fracture sets “Iat EL,, acruss tile local lifhology, If the miiotmpy is remuvcd. Ihe r&dual velociry image correlates well 10 the geology and fracturing intersected by the horcholcr. Low veloci- ties along the boreholrr correlate with zones of increased fracturing and owocia~ed alteration. The alterhun mnes also coincide with areas of lower sonic vetwily and sonii. amplitude and an increase in nimral gamma count rilter in rhr singbhole surveyr.

IN1.KODlICTION

The Canadian Nuclear Fuel Waste Management Program is researching the concept of disposal of nuclear fuel waste in a vault excavated 500 to 1000 m below surface at a suit- able site in plutonic rock of the Canadian Shield (Dormuth and Nuttall, 1987). One aspect of the program has been to develop methods to assess the in situ physical properties of the rock. Methods of locating fractures and fracture zones are of particular interest. This paper describes the use of cross-hole seismic surveys and single-hole acoustic velocity surveys to help to locate and define fractures in a granite.

The surveys were done from the underground workings at AECL’s Underground Research Laboratory CURL) in June, 1993. The URL is located about 150 km northeast of Winnipeg, Manitoba and is situated on the grenitic Lzc du Bonnet batholith. Figure I shows the location of the URL, the underground workings and boreholes MF-I and MF-2 used in this study. The geophysical surveys were done to help charac- terize the fracturing in an area of moderately fractured granite at the 240 m level of the URL as part of a broader experiment to determine the solute transport properties of the moderately fmctured granite. The overall experiment is referred to as the Moderately Frectured Rock (MFR) Experiment.

The region of moderately fractured rock at the 240 m level of the URL occurs about 200 m south of the URL shaft (see Figure I) and has been defined by a series of boreholes drilled from the underground workings of the URL and from surf&. The volume of interest for the MFR solute transport experiment ~nensures roughly SO m by SO m by 50 m (see Figure I) and the area between boreholes MF-I and MF-2 cuts through the middle of the volume of interest (see Figures I and 2 1.

The seismic rurvry ccveds that the southern half of the survey Borehole MF-I is 220 m in length and was drilled from a

panel is likely mure frilcturrd dum the nonhern half as suggested by shaft station at I30 m depth in the URL. The borehole was

the frivzture h,gs 01 dx hoceholes. Drilling subsequent to these sw drilled with an azimuth of 191” and plunges at about l9*. veyr inferrectrd 8 cataclastic horizon within Frilcture Zone 2.5 in the Borehole MF-2 is 210 m in length and was drilled from the 240 southern half of rhe panrl which coincides with a region of very Iow Level at the sxne azimuth as MF-I but with a slightly shal- apparent velocities. lower plunge of about 9~. The distance between the borehole

‘Prrsrnwd a( the CSEGKSPG Joinr Niltiond Cmven~ion. Calgary. Alberta. Mily 12. 1994. Mnnuacript received hy the Editor Novrmher 22. 1995: rcvivxl mirnurcript received January 10. 1906. ‘AFCL. Whiteshell Laborilturier. Pinawa. Manitoh ROE IL0 Reviews by C.C. Davison, K.M. Stevens. E.T. Kouk and L.H. Frost were very whmhlc. ?k Operations stcdffffat the URL helprd in scheduling, reducing noise levels and pmuiding the xcunite hordwle survey information. The graphics office at URL and Dvdc Wodclcock provided several figures and iusislimce ~11s received hn D. 00x and R. McCrcgor in graphics preparation. A majwily of this work was fundrd juimly hy AEC:L and Ontario Hydn, under the ;mrpiccr d chc Cimdu Ownen Cimup.

(‘EC 6 1”“C I”0

rKoSS~llol.,i AN,) SIN~;I.I:~Ho,.,~~;,~OPHYS,t‘AI. SIIKVI:YS

Shaft

I

Planned Location of MFR Experiment

Fig. 1. Location of the Underground Research Laboratory and location of boreholes MF-1 and MF-2,

collars is I08 m and the holes converge to within 78 m at their ends. The roughly rectangular region between these boreholes measures about 210 m by 90 m. The plane defined by the area between the holes strikes NIO”E and dips at about 82”W. The single-hole and crass-hole surveys were performed in boreholes MF-I and MF-2 to assist in physical property characterization along and between the boreholes.

The surveys were done to determine the variations in rela- tive fracture frequency between the low-dipping boreholes MF- 1 and MF-2.

GEOLOGY

The URL is within the granitic Lx du Bonnet Batholith, a large Archean-age complex, with dimensions of about 100 km

east-west by 40 km north-south by perhaps IS to 20 km in depth extent. It is believed this batholith complex formed at a depth of about I5 to 20 km about 2.6 billion years ago (Stone et al., 1989). The present erosion surface is believed to be very close to the upper contact of the complex with the country rocks (Brown et al., 1995; Everitt and Brown, in press). Gneissic (compositionally layered) granite is the main rock type at the URL, but several xenolithic and leucwxatic zones also exist and the batholith is also crosscut by granodiorite and pegmatite dykes and sills. Red to pink alteration is pervasive in the upper 200 to 300 m but below these depths the granite is grey except for red alteration adjacent to fracture zones (Everitt and Brown, 1986; Whitaker, 1987: Everitt et al., 1991).

Figure 2 shows the structural geology of the URL in a vert- CBI plane that roughly contains the two boreholes. Three

North

0

100

m Moderately Fractured m Fracture Zones

n Sparsely Fractured

Fig. 2. Geology of the Underground Research Laboratory site. The locations of boreholes MF-1 and MF-2 are shown

low-dipping fracture zones cut through the rock mass at the URL and the intervening rocks range from moderately frac- tured to sparsely fractured. These fracture zones and the vat- al fractures and joints associated with them are the dominant pathways for groundwater flow through the rock. The location

of boreholes MF-I and MF-2 are shown in Figures I and 2.

INSTRUMENTS AN” SURVEY DESCRIPTION

Figure 3a shows the lithologic phases of grey granite between boreholes MF-I and MF-2 and the location of two low-dipping fracture zones (Fracture Zones 2 and 3) above and below these two holes. Both holes were drilled in the plane of a coarse scale compositional layering (Figure 3a). Fracture Zone 2.5 is related to Fracture Zone 2 but where it intersects the URL shaft it comprises a zone of shallow-dip- ping fractures each less than 1 mm in width infilled by chlo- rite and hematite. In addition, Fracture Zone 2.5, at the shaft,

does not have a cataclastic horizon. or dilation gaps produced by offset as exhibited by Fracture Zone 2 or 3. Figure 3b

shows the alteration and fracture sets that overprint the granitic rock lithology. Figures 2, 3a and 3b define the extent of geologic knowledge of the area prior to the geophysical surveys described in the following sections.

A Mount Sopris Instruments Ltd. Series III logging system was used for the single-hole geophysical surveys in MF-I and MF-2. Three-arm caliper, n&.turirl gamma, sonic velocity and sonic amplitude measurements were performed. A set of interlocking aluminum rods were used to push the probes to the end of each borehole and then readings were taken as the probes were pulled back toward the borehole collar. Analog and digital measurements were recorded every 0. I m in the holes. Data were logged directly to a portable computer attached to the acquisition system. The digital dttta was then edited and plotted using a process developed at AECL’s Whiteshell Laboratories.

The AECL Cross-Hole Audio-frequency Rock Testing System (CHARTS) was used for the cross-hole seismic surveys. This instrument was developed by G.F. West and associates at the University of Toronto specifically for sur- veys in a high-strength rock. Results with an earlier ver- sion of this instrument are described by Wong et al. (1983, 1984, 1985). Several surveys have been completed by AECL with a newer version of the instrument constructed in 1986.

Cl& 8 l”“C IW6

a. L-~ ,f---‘ UAL Main Shaft

-

Red/Pink Granite

enollthlc Gnelsslc .-.. ____I,_

r ‘-& . , , . . . . . ; ,’ ... ., ,

y. : Y-w

. . . . . . .

ranite

I Fracture , sets

Fracture Sets

Fig. 3. (a) Lithology of the plane containing boreholes MF-I and MF-2. (b) Alteration overprinting the geology is shown with the fracture logs for ea hole. The fractures intersecting the walls of the shaft are also shown at the left.

ch

CIEG s June IWb

Flg. 4. Basic elements 01 the AECL Cross-Hole Audio-frequency Rock Testing System (CHARTS).

The new system operates at about 4400 Hz which yields a P- wake with a wavelength of -1.3 m in the URL granite. Both the transmitter and receiver use piezoelectric elements as the active transducer. The system uses a Pseudo Random Binary Sequence (PRBS) to increase the signal-to-noise ratio. Figure 4 shows the basic elements of CHARTS.

CRoss-How Sutwev METHOII

The accuracy of a cross-hole velocity image is very depen- dent on the 3-D space position of the boreholes. The average uncertainty in the borehole collar coordinates is ?2 mm and the azimuth and dip angles are considered good to about

The cross-hole data was acquired with the following parameters:

transmitter borehole: MF-I water filled and probe unclamped receiver borehole: MF-2 water tilled and probe unclamped source spacing: 5 m recel”er spacmg: 5m AID sample rate: 20 1(” record length: 163 ms transmttter: piezc&ctric in hybrid design, PRBS

drive, 62 mm diameter PRBS parameters: 1023 element PRBS EXl”W x- , y- and z-oriented detectors in hybrid

design, 62 mm diameter Rxffx centre frequency:4400 Hz

+O.OS’. A Pajari survey was performed to measure the receiver channels: 3 components (x = radial, y = azimuthal azimuth and dip of the borehole at roughly 20-m intervals and z = axial) along the length of the borehole by using a magnetic compass number of rays shot: 1930 and a dip sensor mounted in a gravity-oriented gymbal Since the holes are approximately horizontal, a set of arrangement to an average uncertainty of about kO.2”. interlocking aluminum push rods was used to push the

CJFU 10 ,mr \Wh

rKoSS Ho,.,: ANI) \IN<il.l: ,,Ol.l~ ~i~iONlYSl~‘hl. SIIKVI:YS

1 Fiel;fpy )

,+,

1 Deconv,olution ]

Fig. 5. Flow chart for data processing and analysis of cross-hole data

probes along the boreholes. The receiver was pushed to the

bottom of MF-I and the survey was started by moving the

transmitter at S-m intervals in MF-I between 5 to 210 m

using the push rods. The error in transducer depth is esti-

mated to be +2.5 cm for depths O-100 m and ?5 cm for

depths >I00 m.

PKOCESSING AND ANALYSIS

Figure 5 shows the steps used in processing the cross-hole

data. Binary data, recorded on 9 track tape, is cross-correlated

with the reference PRBS and converted to ASCII format and

is then plotted, timepicked, analyzed, edited and finally imaged using the Simultaneous Iterative Reconstruction

Tomography (SIRT) algorithm. The most common error that

occurs in the measurement of signal transit time is caused by

low signal-to-noise ratios. If low quality or questionable data

is included with good quality data the resultant image may

have some irregular features that detract from, or bias, the

image. The decision to remove questionable data is normally

based on the signal-to-noise ratio. Normally, at the URL, the

transit-time measurement is accurate to I to 2 parts in 1000

CEO 1, he lU’,O

J.G. HAYtx5.M.H. SERZ”, 0.x LODHA, R.A. E”EKn”rand D.K.T”MS”NS

when signal-to-noise ratios are high (greater than 2.5 to 3) with CHARTS. The data is corrected for velocity anisotropy and in the final step the residual velocity data is imaged using six iterations of a straight-ray SIRT algorithm. A tw- metre pixel size was used for the imaging and this image was then smoothed to a 0.6-m pixel size. As mentioned above, a good measurement of seismic wave velocity requires an accurate measurement of both the signal transit time and the separation between the signal source and detector.

An accurate measure of the separation distance between the signal source and the detector is often more troublesome than the measurement of signal transit time. The measure- ment of distance between transducers in two boreholes is a 3-D function of I2 variables (i.e., 6 variables for each bore- hole). The variables are: the borehole collar position (x,y.r), the azimuth and dip of the borehole and, finally, the trans- ducer depth within the borehole. A well designed cross-hole velocity survey attempts to match the uncertainty in the mea- surement of distance to the uncertainty in the travel time.

SINC;I.E-HOIX GEOPHYSICAI. LOCCING RESULTS

Figure 6a shows the results of the single-hole geophysical surveys in MF-I, along with the geologic units and fracture frequency as observed in the core from the borehole. The acoustic velocity in MF-I averages about 5700 m/s from 0 to 80 m and then the average velocity decreases tu about 5650 m/s from 80 m to the bottom of the hole at 210 m. The acoustic velocity decreases about 200.300 m/s between the depths: O-12 m, 40.43 m, 139.141 m and 156.157 m and these lows correlate with areas of increased fracture fre- quency in the core. The acoustic amplitude also decreases in areas of increased fracturing in MF-I but the amplitudes of the lows do not correlate exactly with the amplitude of the decrease in acuusitic velocity.

No fractures were intersected between 57 m to IO1 m in MF-I and this is also a region of uniform response on the acoustic logs.

The caliper log shows decreases in borehole diameter of about 0.5 mm and 0.15 mm below 140 m and I58 m, respec- tively. These changes appear to be at areas of increased frac- turing and may indicate a problem with the measurement. The caliper arms may not generate enough force to counter the heevy weight of the probe, especially in this nearly hori- zontal borehole, and so the measurements may be in error. A 0.4 mm widening of the borehole at 140 m indicates a cavity associated with the fracturing. The fractures and cavity at 140 m also cause large decreases in the acoustic velocity and acoustic amplitude. The natural gamma log is uniform at about 500 counts per second (cps) and there is no significant correlation in this log to the fracturing or the other logs.

Figure 6b shows the results of the single-hole geophysical surveys in MF-2 with the fracture frequency as observed in the core and the geologic units intersected in this hole. MF-2 intersects far fewer fractures than MF-I and the acoustic velocity and amplitude logs show a much more uniform

CEO 12

response than that observed in MF- I. There are no anome- lous velocity lows in the acoustic logs associated with the first three fractures logged in the core and even towards the end of the borehole (from I68 m to 210 m) the increase in fracture frequency does not affect the acoustic logs very much. The acoustic velocity in MF-2 averages about 5750 mls from 0 to I I2 m where the average drops tu about 5650 mls for the remainder of the hole.

The caliper log of MF-2 shows the unusual situation of an increase in diameter of about 0.4 mm between about 20 m and 60 m. No fractures have been logged in this zone. This apparent increase in diameter is probably due to B problem in the culiper instrument itself, as mentioned earlier, rather than a true change in hole diameter. There are no fractures at these depths and there are no correlations to the other geo- physical logs to explain these apparent changes.

The natural gamma log in MF-2 averages about 600 to 650 cps from 0 to about I50 m and then the level gradually increases to about 1500 cps. We can not explain this gradual increase in the natural gamma count rate. The xenolithic granite, with its coarsely crystalline matrix, appears to give slightly higher natural gamma count rates than the homoge- neous and medium- to coarse-grained grey granite but the correlation is weak. Geological mapping of this unit at the 240 Level indicates that this unit varies considerably in com- positional texture.

CROSS-HOLE SEISMIC REWI.TS

Figure 7 shows the raypath and ray density plots for the cross-hole seismic panel between MF-I and MF-2. The best information in such a tomographic survey is usually in the centre where ray concentration is greatest and where the greatest angular coverage exists. Figure 8a shows an exam- ple of a complete fan of waveforms for the receiver fixed at 150-m depth in borehole MF-2 while the transmitter was moved at 5-m intervals from 5 m to about 220 m in MF-I. Figure 8a shows good P-arrivals but only irregular and poor quality S-arrivals. This figure also illustrates strong tube wave events caused by an 0.4 mm change in diameter of borehole MF-I at about 140 m that coincides with an open permeable fracture which causes a secondary P-wave to be generated. Figure 8b shows how these waves are created and detected. Tube wave information has not been studied in detail.

Figure 9 shows a close-up of P- and S-wave arrivals with the receiver fixed at 55 m in MF-2 and the transmitter moved from 155-m to 175-m depth. This figure illustrates the qual- ity of the P-wave arrivals: often the arrival time uncertainty is I to 2 parts in 1000. The plot gain is constant for all seis- mograms in this figure and the gradual decrease in P-wave

amplitude is related to the greater travel path distance. The dominant frequency of the P-wave arrival was about 4400 Hz for this survey and this corresponds to a I .3-m wave- length in the granite.

There is a pronounced curvature in the plot of observed

,unr IWO

20

Fraslm 10

0

Fracture Frequency (core)

11h.t

! ocQo1’l. ! , , . . , . . . , 0000

2.0”“. ” ” ” “. 1.. . 1 1 Acoustic AmDlltude (Volts)

2.5

2.0 11 1

2.0 l.0 1.5 1.0

~~~~

‘cm j Natural Gamma (cps) IWO

10 - : Fracture Frequency (core)

Fraclm .

0 ‘. I t . ., 11. 9ooo . I .( I . . . I I . t 0000

s$ Acoustic Amplitude (Volts)

Diameter (mm)

0000

2.0

2.0

1.6

1.0 97

90

90

gnelsslc granite u/ /50/ 100 ““~redalt xeno’lthlc granite Downhole Length (m)

Fig. 5. Single-hole geophysical logs for MF-I and MF-2 compared to fracture frequency and geology.

J.C. HAYLES. M.H. SERZ”, G.S. LODHA. R.A. EVEKlrr and D.K. TOMSONS

240

z 48 ,~~~~~~~, x E 40 2 >I ix

32

24 >I -= 2 16

p” 4

2 a

MF-1

Distance (m)

Fig. 7. Raypath and ray density diagrams for the CHARTS survey between MF-I and MFG. The interior region with the greatest range of angular coverage (and greatest ccsncentration of rays) is the area where the image is most reliable. The ray densities are calculated for 2 by 2-m pixels that have been interpolated to 0.6 by 0.6-m pixels and then smoothed.

velocity verws ray angle for this data (see Figure IOa) and Velocity anisotropy within the rock mass can obscure the this usually indicates a velocity anisotropy in the rock more subtle P-wave velocity variations and their correlation to mass. Rays that travel at 90~ with respect to the axis of the the geology and Figures IOb and IOc show the procedure used transmitter hole tend to travel faster than rays with other to remove the velocity anisotropy from the data. A sinusoid orientations. was fitted to the central population of the data on the velocity-

c,m 14 ,“l,C IW4

(‘ROSS-Hot.,: AN,) SINGLE-HoLt Gb.OPHYSI~‘.~I. SI’I~“I-YS

0 0.033 Time (milliseconds)

0.065 0.098 0.131 0.16

,.,. IL xu-

ET

m I hi I m La. I”” 111

, ,

Indirect Secondary

P-Wave -5800 m/s

Flg. 8. Tube waves generated within the water column of the transmitter borehole. MF-1, travel at about 1500 m/s along the hole and create sec- _ onaary !+vaves at a con~trlctlon (impedance cnange) !n MI-1 near 140-m depth. This secondary source of P-wave energy is observed late in the ?.eismic record with a distinctive slope.

CIEC 15 ,vne ,996

J.C. HAYLES, M.H. SERZU. G.S. LODHA. R.A. E”ERlTT and D.K. T”MSONS

12 1

18 I

Time (milliseconds)

24 30 I I

36 42 48 I I I

20 Microsecond Sample Rate

160 55

Y File name

range 128.4 m VP - 5773 m/s

165 55

range 131.7 m I je153215 vp = 6771 m,s

range 136.1 m 170 55 vp - 6764 ml.9

‘7 P-wave

\ S-wave

range 138.6 m 175 55 vp = 6765 m/s

Fig. 9. Close-up of P- and Swave arrival quality for the x-component tensor in the receiver at 55-m depth in MF-2 and the transmitter moving from 155 to 175 m in MF-1. The vertical gain is constant for each of the five seismograms shown here.

6200

zf 6000 E %

x

z 5800

5600

Observed

9b 1 Ray Angle (degrees)

90

Ray Angle (degrees)

Anisotropy = Correction Residual

Ray Angle (degrees)

Flg. 10. P-wave velocity anisotropy was removed by fining a sinusoid to the observed data. The anisotropy is probably caused by veltical and near-ver- tical fracturing that occurs across the entire panel. Tomographic images based on the residual velocity distribution then should show patterns related to the geology, without the strong overprinted effect of vertical fractures.

ClEG 16 June 1496

vs-ray angle plot. The amplitude of the anisotropy function that was removed is about 250 m/s (peak-to-peak), or 4% of the average velocity, and is centred at an angle of about 90 from the dip of the transmitter hole. A model of weak elastic anisotropy caused by a set of parallel microcracks (Thornsen, 1986) is shown in Figure I I. The theory predicts a gentle curve on a velocity-vs.ray angle plot that is similar to a sinu- soidal function. The cosine function we have used to account for the velocity anisotropy is probably a reasonable approxi- mation based on the available information for the rock mass.

Figure l2a shows the P-wave velocity image from the observed P-weve arrival times prior to anisotropy correction. The residual velocity image of Figure 12b represents the rock velocity structure after correction for anisotropy. The fracture occurrence along each borehole is plotted with the residual P-wave velocity image and the sonic velocity logs in

Figure 13. These sonic velocity logs in Figure I3 have also been corrected for the same anisotropy as observed in the cross-hole seismic survey. Lower-hemisphere equal-area stereographic projections of poles to fracture planes are also included in this figure. Both stereograms indicate two steeply dipping fracture sets that strike roughly NE and NW (the boreholes trend -19 IO). There is a good correlation between the areas of increased fracturing and regions of low apparent P-wave velocity in the panel. The first 40 m in MF-I is the most fractured and this area also has a lower then average P-

wave velocity. The average P-wave velocity tends to increase where there is no fracturing in MF-I (from about 40 m to 100 m) and then decreases again where fractures are observed (from about 100 m to the end of MF-I).

Higher than average velocities occur in MF-2 from 0 to I IO m than for the remainder of the borehole from I IO m to

Plane

h 90 degrees

Parallel Fractures

6000

-90 0 90

Incident Angle (degrees)

Y

I Fig. 11. Theory of weak elastic anisOtropy (Thornsen, 1986) that relates P-wave velocity to the angle of propagation within a medium with plane-par- allel fracturing developed along one orientation. The P-wave travels fastest parallel to the fracturing.

ClK 17 ,une ,996

I

I.G. “AMES, M.H. SERZU.0.S. L0DHA.R.A. E”ERn-rand O.K. TOMSONS

North

-Y, r Observed

6100

6050 si z 6000

; 5950

-5 5900

$ 5650

5800

5750

South

MF-2 -=-msy

\ MF-1

Fig. 12. P.wave velocity tomograms for the area between boreholes MF-I and MF-,Z. The upper tomogram (a) shows the Velocity image pflor lo anisotropy wrreclion and the lower tamgram (b) shows the residual P-wave velocity ~?~ge after mrre~tvm for a 4% (peak-to-peak) anisotropy.

CIEG 18 ,unc 19%

CIEG 19 ,uns IWO

210 m. The high between 164 m to 170 m in the MF-2 veloc- ity log also shows up faintly in the cross-hole seismic date. The velocity log of MF- I is much more variable than in MF- 2, perhaps due to the greater number of fractures intersected by MF-I. A higher than average acousitic velocity is observed between hl m to 7X m in MF- I and this correlates to a region of higher than average P-wave response in the tomographic image. A high P-wave velocity region about 20 m to 2.5 m in diameter exists between MF- I and MF-2 cen- tred about 95 m down MF-I nnd about 25 m toward MF-2 from MF-I. This high-velocity zone correlates roughly with a xenolithic horizon (see Figure 3a ).

Figure I4 shows a comparison between the residual seis- mic velocity image between MF-I and MF-2 with an image of the cross-hole radar velocity that was measured between the same boreholes (Stevens et al.. 1994). The radar survey was done using a RAMAC system at 22 MHz which pro- duced a wavelength of about 4 m in the rock. A 4-m interval was used for the source and detector measurement spacing and over 2500 rays were shot for the survey. Instrument problems during the radar survey reduced data quality but there are clear similarities between the radar and seismic images. Where the P-wave velocity image shows a high velocity along the northern half of MF-2, the radar survey shows a similar high. Similarly, low seismic velocity and low radar wave velocity occur all along the lower half of borehole MF-2. The isolated seismic velocity high, about 20 m to 25 m in diameter mentioned above, however, correlates with a low radar wave velocity anomaly.

INTERPRETATION

The apparent P-wave velocities in the northern half of the panel BE not reduced from the average value observed from other areas of sparsely fractured granite at this level. This suggests that the granite in this region is sparsely fractured and that Fracture Zone 2.5 is thin and lacks a cataclastic zone.

The apparent P-wave velocities are reduced in the south- ern half of the panel to a level that is below the average expected for sparsely fractured rock at this level. The inten- sity of fracturing (and/or alteration) must increase in the southern half of the panel. An increase in vertical and/or near-vertical fracturing could explain the observed velocities but horizontal and near-horizontal fracturing may also increase. Increased vertical fracturing and alteration is observed in MF-2 from I IO m to 220 m. Fracture Zone 2.5 may widen or become more intense in the southern half of the panel. (Borehole MF-I2 drilled subsequent to this survey intersected a narrow cataclastic zone within Fracture Zone 2.5 in the southern half of the cross-hole survey panel that is absent where Fracture Zone 2.5 intersects the URL shaft.)

In general, the cross-hole and single-hole seismic velocity surveys show lower seismic velocities in the region proposed for the Moderately Fractured Rock experiment. In addition, there is a 4% (peak-to-peak) anisotropy in the seismic wave velocity that correlates well with the orientation of vertical

J.O. HAYLES. Mkt. SE.RZII.C.S. LODHA. R.A. EVERrrTaand O.K. TOMSONS

and near-vertical fracturing that has been mapped in the rock mass throughout the upper 220 m of the rock ar the URL shaft location (see Figure 2).

A comparison between cross-hole seismic and cross-hole radar images, shown in Figure 14, indicates that the areas of lower P-wave velocity are also areas of lower radar wave velocity. Despite the reduced data quality in the radar image there are clear similarities between the radar and seismic images. One of the main controls of the radar wave velocity within the granite is water content. Areas of higher water content (as either free water and/or hydrous mineral alter- ation) tend to decrease the radar wave velocity. This means that the areas of increased water content seen in the radar survey tend to be the same areas that the seismic survey pre- dicts to be more fractured. In contrast, Figure I4 also shows one location where a high in the residual P-wave velocity coincides with lower than average radar velocity. This fea- ture is about 20 m in diameter and is located about one third the way down MF-I and is offset from MF-I toward MF-2 by about 25 m. This anomaly may be caused by an increased concentration of pyrite, a metallic mineral, that is a common association within the xenolithic domain in this horizon. Radar transmission velocity decreases in the prexncr of good to excellent electrical conductors like pyrite. The increase in seismic velocity is also consistent with the higher bulk density of the minerals contained in the xenolith. Xenoliths with 2-m to 4-m dimensions have been observed at the 130 Level containing up to 20% pyrite by volume.

Figure I5 shows an interpretation of the results of the borehole logging and cross-hole seismic survey for the study area. In this figure, the predicted intensity of fracturing (of mixed orientation) is inferred from the P-wave velocity. Lowest velocities are assumed tu be the areas of greatest rel- ative fracturing and alteration.

There is a good correlation between the seismic survey results and the known geologic structure between the two boreholes. Unaltered regions of granite exhibit the highest acoustic velocities and areas of higher than average P-wave

velocity in the cross-hole tomography. The areas of greatest alteration and fracturing also coincide with lows in the acous- tic velocity logs and lows in the cross-hole velocity image.

The cross-hole seismic tomography survey has helped to define the geologic conditions between boreholes MF-I and MF-2 at the URL. There is a good agreement between the fracture frequency observed in the borehole core from these two boreholes with the single-hole acoustic velocity and acoustic amplitude logs and with the velocities observed in the residual cross-hole velocity image.

The main cause of the 4% anisotropy in seismic velocity observed in this study is probably due to pervasive vertical/steeply dipping fractures and microfractures in the rock between the two boreholes. A similar anisotropy has been observed on a smaller scale survey (Hayles et al., 1993)

Cl& 20 ,unu IW4

6100

F E

- 0000 ),

.=

8 9 5900

5800

123

z =t

z

* 113 I .=

.E

3

103

MF-1

MF-1

Fig. 14. Comparison of cross-hole seismic and cross-hole radar tomographic images for the area between boreholes MF-I and MF-2. A 4% anisotropy was removed from the P-wave velocity to produce the image *how here (see Figures 10 and 12).

c,eCi 2, ,unc 1Wh

J.G. HAYLES. M.H. SERL”. G.S. LODHA. R.A. E”ERtTTand D.K. T”MS”NS

Shaft Wall Profile

FZ3

FZ 2.5

North . b South

CreC, 200,~ m

MF-2 Open Fractures

Total Fractures

Increasing Fracture Intensity

Fig. 15. Relative fracture frequency between boreholes MF-1 and MF-2 interpreted from geology and geophysics. The fractures intersecting the walls of the shaft are shown at the left.

looking at blast damage around the perimeter of Room 214 at

the URL using signal frequencies of >60 kHz. Vertically ori- ented microcracks, uniformly distributed in the rock mass,

were thought to cause the velocity anisotropy at this smaller scale.

Cross-hole radar velocity results complement the cross-hole

seismic velocity results. Areas of low radar wave velocity are

also areas of low P-wave velocity. This suggests the frac-

tures responsible for the lower P-wave velocity are also water filled. An anomalous region measuring about 20 m by

20 m has been observed in both the seismic and radar SW

veys and is interpreted as a metallic sulphide-bearing xeno-

lith similar to that seen on the I30 Level.

Higher P-wave velocities correlate with areas of more competent granite while areas of low velocity correlate with

areas of known fracturing. The lowest P-wave velocities in the panel lie in the southern half of the panel and overlie the

position of Fracture Zone 2.5 and suggest that Fracture Zone

2.5 may include some dilational and cataclastic zones.

Drilling subsequent to our surveys tends to confirm this interpretation. If remains to be seen how the hydraulic tests

that follow will correlate with these results.

CIEG 22 ,une I*)4

(‘K0SS~H0LE ANI) SINCX.E~HOLE G,iOP,lYSK‘At. s,II<v,;Ys

REFERENCES Stevens, KM., Everit,. K.A.. s,rect. P.J. and Lodha, G.S.. ,994. Lirho-struck turd characIerira,i”n in granite rocks using single~hotr and cross-hole

Brown. A., Everi” R.A.. Martin, C.D. and Davis”“. C.C.. ,995, Past and future fracturing in AECL research areas in the Superior Province of the

radar techniques: Procerdings of the Fifth tntrmahwd Gmference on

Canadian Precambrian Shield: with emphasis on the Lx d” Sonnet Ground Penetrating Radar, Kitchcner. Ontario. Vol. 2.625-633

Batholith: Atomic Energy of Canada Limited. Repon I t214*. COG 94. Stone, D.. Kamineni. D.C., Brown. A. and Everirt. R.A.. 19X0, A compw- 528. SO” of fracture styles in two granite bodies uf ,hC superior Pruvince: can.

Dormuth. K.W. and Nuttall, K., ,987. The Canadian Nuclear Fuel wilste 1. Earth Science 26. 387-403. ivtanagemrnt Pmgram: Radioacdve Waste Management and the Nuclear Thamam. L.. 1986. Weak etmfic misomov: Croohvrics 51. 1954.1966. Fuel Cycle 8,931”4. .,

tv.nu. R.N. ano L)ruwn, A., bYa”, auoaurlacr ~C”,“E EvenII. R.A. and BrnW”, A.. ,986. Subsurface geology of rile ““dergmnd Whitaker. S.H., 1987. Geoscience research for the Canadian Nuclear Waste

Research Laboratory, an werview of recent develop Research Laboratory, an werview of recent developments, in Proceedings Manageme”, Program: Radiuactive Waste hlanilgemrnt and the Nuclem

of the Twentieth lnfomx~tion Meeting of the Canadi; of the Twentieth lnfomx~tion Meeting of the Canadian Nuclear Fuel Waste Fuel Cycle 8, 145-196.

Management Program: “0,. I, 146.181. *turni Management Program: “0,. I, 146-181. *tumic Energy of Canada Wang J., Hurtey. P. and West. G.F., 1983. Cross-hde srismotogy and ieis- Limited 1 Limited Technical Record, TR-375’. mic imaging in crystalline m&s: Geophyi. Res. Mt. 10, ht(h-h*9.

~ and ~. in press. Geological mapping of AECL’s Underground Research Labmtury. a cross-section of thrust faults and associated frac- ures in the roof zone of an Archcan batholith: Proceedings of Fractured and Jointed Rock Masses. a reeional confemm of the tnternationat

and ~. 1984. Cros*~hote audio-frequency reismotogy in granitic rocks using pieroetectric transducers as sources and detectors: Geuenpt. 22,x-279.

Society of Rock Mechanics. Graniibakken, California, Vol. I, t-t I.

-- Davison, C.C., Gaacoyne, M. and Martin, CD., 1991, Reeionat and IocaI settine of the Undereround Research Laboratorv. in Pr&edings of an tnter&ional Symp&um on Unique Undergr&nd Structures: “0,. 2, 64~1 to 64-12.

Haytes. J.C.. Stebner, D.T.. SW,“. M.H. and Lodha, G.S., ,993. Geophysical inspection for the engineered blast test using low-shock- energy explosives phase 111: Repan for *gence Nationale pour ta Gestion der Dechets Radioactifs WWRA). Fontmay-aur-Roses. France. Rep. st2RPAEL9MQ1.

and ~. 1985, tnvesrigation of subsurface geological structure at the Underground Research Laboratory with cross-hole seismic scanning, in The geoscience program - proceedings of the Sevenreenth Information Meeting of the Nuclear Furl Wtule Management Program: Vol. II, 593-608, Atomic Energy uf Canada Limited Technical Record. TR-299*.

*Available from Scientjfic Document Dktribution Office tSDDO1. Afmnic Energy of Canada Limited. Chalk River. Ontario. KOJ 110 Fan: (613) SW 1745 Tel: (613) 584-331 I ext. 4623

cm? 23 h’s IW6

t