You are on page 1of 8

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/241209541

Crosshole seismic testing---Procedures and pitfalls

Article  in  Geophysics · January 1981


DOI: 10.1190/1.1441134

CITATIONS READS

18 1,190

1 author:

Dwain K. Butler
applied geophysics consultancy lld
88 PUBLICATIONS   577 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Dwain K. Butler on 09 July 2018.

The user has requested enhancement of the downloaded file.


GEOPHYSICS, VOL. 46, NO I (JANUARY 1981). P. 23 2’). 15 FIGS

Crosshole seismic testing-Procedures and pitfalls

Dwain K. Butler* and Joseph R. Curro, Jr.*

Crosshole seismic testing is a valuable technique for determining seismic velocity profiles for critical structure siting investigations. Im-
proper field procedures, however, can result in the acceptanceof apparent velocities as true velocities. leading to unconservative design
values. Common pitfalls in crossholeseismic testing are ( 1) the use of source-receiverborehole spacingstoo large to enable the determination
of true formation velocities, (2) the use of in-hole station spacingstoo large to allow resolution of thin velocity layers of interest, (3) the use
of incremental traveltimes between widely spaced receiver boreholes, and (4) the naive assumptionthat boreholes are vertical and hence
parallel. A procedure is presented whereby field programs can be more rationally planned.

INTRODUCTION be referred to as a crosshole set. A-B and A-C are crossholepairs,


An appropriatephilosophy for site investigations is to strive for where A is the sourceborehole and B and C are receiver boreholes
“adequate” subsurfacedefinition, at least, adequateto the extent in each case. The perpendicular arrangement (Figure la) is useful
that there will be no surprises during subsequent site use. The for areal coverage, resulting in a set of perpendicularprofiles when
results of the geophysical aspects of a coordinated site investiga- conducted on a grid pattern. In the absence of significant aniso-
tion should be presented in forms immediately useful to those in- tropy, the results of the two crosshole pairs in Figure la can be
volved in site planning and development, i.e., engineers, man- averagedto give a single representativedetermination. The in-line
agers, etc. arrangement(Figure 1b) is useful for a structure-controlledsurvey.
Crossholeseismic testing is a valuable technique for site investi- suchas along the center line of a proposeddam, along the base of
an existing dam, or along the proposedroute of a tunnel, highway,
gations, since it can be used to produce detailed compression (P)-
and shear @)-wave velocity profiles. These velocity profiles can canal, etc. Also, the in-line arrangement is invaluable for model
then be used to compute elastic moduli profiles for dynamic verification, where pair A-B is used to develop the model and pairs
analyses. Crosshole seismic testing, however, should never be A-C and B-C are used to verify the model.
used alone, but should always be utilized in a coordinated site- In-hole procedure
investigation program. The crosshole survey should tie in with
The opposed source-receiver in-hole geometry (Figure 2) is pre-
existing boreholes, and all available geologic information should
ferred, in which the source and receiver are synchronouslyraised
be used in the interpretation process. In fact, one of the primary
from the bottom to the top of the boreholes. With this procedure,
disadvantagesof the crosshole method is the requirement of bore-
holes for conducting the tests; thus, the use of any existing bore-
holes at a site assumesdual importance. The primary advantage,
however, is subsurfaceresolution which is not possible with sur-
face methods. This alone justifies use of the crosshole seismic
method in many cases.
The object here is not to examine the mechanics or instrumen-
“t-4
tation requirements of crosshole seismic testing. Rather, the pri- 0 SOURCE BOREHOLE
mary concern is with proper procedures in planning the field a. I 0 RECEIVER BOREHOLE
survey and pitfalls to avoid. Through an example and an analytical
study, procedures are suggested for more rational planning of
crosshole surveys. Finally, two field examples illustrate the use
of crossholesurveysin coordinatedgeophysical site investigations.

PROCEDURES
b. A~--&-----~
Crosshole set borehole geometry
FIG. I. Typical geometries for crosshole seismic tests: (a) areal
One of the first considerations in planning a crosshole seismic coverage and for detection of horizontal anisotropy, and (b) model
investigation is the borehole geometry. Figure 1 illustrates two verification, structure-controlled surveys, and detection of
commonly used borehole geometries. Boreholes A, B, and C will inhomogeneity.

Presentedat the48thAnnualInternational
SEGMeetingOctober31, 1978, in SanFrancisco.Manuscriptreceivedby theEditorJune28, 1979;revisedmanu-
scriptreceivedNovember26, 1979.
‘U.S. Army Corpsof EngineersWaterwaysExperimentStatmn,P 0. Box 631, Vicksburg,MS 39180.
0016-8033181/0101-0023$03.00. 0 1981Societyof ExplorationGeophysicists.
All rightsreserved.

23
24 Butler and Curro

it is possible to obtain an optimum data set for the determination


I of true formation velocities and interface depths. In order to
delineate a velocity-stratigraphic unit’, utilizing first arrival time
data, it is necessaryto have a true velocity and at least one apparent
velocity’ determination from source receiver stations in the unit.
For a section with velocity increasing with depth, one true and one
SOURCE <ECEIVER
apparent velocity determination in each unit is the optimum (and
minimum) data set. This minimum data set requirement places
restrictions on source-receiver borehole spacing and in-hole sta-
tion spacing.

PITFALLS
Realizing that some of the pitfalls described here could be
avoided by conducting other type tests such as upholeidownhole
surveys, this paper is designed to addressprimarily the crosshole
technique. The pitfalls encountered in crosshole seismic field pro-
cedures and data interpretation techniques that have contributed
to erroneousresults are discussedbelow. One of the troublesome
productsof these pitfalls is the acceptanceby many investigators
of velocities as typical of various materials, which, in fact, are
too large. This acceptancecan lead to the rejection of true velocities
even when determined and to unconaervative designs.
A common pitfall is the use of source-receiver borehole spac-
ings which are too large to enable true velocity determination. It
is common to find references to crosshole surveys with borehole
FIG. 2. Vertical section view of crossholeseismictesting geometry spacingsof 100 ft or more. In cases where the interpretation pro-
showing hypothetical raypaths for layered subsurface. cedure consists solely of dividing direct-path distances by first

‘A velocity-stratigraphicunitis a soil/rock unit which is velocity mappable,


i.e., a unit which can be mapped or delineated by seismic techniques.

‘An apparent velocity is defined simply as the direct-path distancedivided


by the first arrival time In this sense, prior to interpretation, the true
velocity (if defined by the data) is an apparent velocity.

time MSEC
GEOLbzG’C
LAYER 1 FILL
----------
BAY
MUD

UPPER
SANDS

STIFF
CLAYS
&
SANDS

FIRM
CLAY

STIFF
CLAY

SAND

/ P-ARRIVALS

ROCK
S-ARRIVALS

FIG. 3. Velocity profiles for site in San Francisco Bay area (from Schwarz, 1970)
Crosshole Seismic Testing 25

LAYER 1-Q
b LAYER 2 - LAYER 2 -% LAYER 3 -
k k%

IlOO-

6
0 I I I I 1
0 100 200 300 400 500
BOREHOLE SPACING D, FT

FIG. 4. Variation of apparent S-wave velocity with borehole spac-


ing for source-receiver station centered in the Bay Mud layer.

FIG. 5. Apparent S-wave velocity variation with source-receiver


position in the Bay Mud layer (layer 2) for a fixed borehole spacing
of 20 ft.

arrival times, the result is an apparent velocity profile. Another


pitfall is the use of large in-hole station spacings. This contributes
D
to the difficulty of determining true velocities and also does not
allow resolution of thin layers, which could be of considerable
importance. The source-receiver borehole spacing and in-hole
station spacingare the critical parametersto consider in planning a
II VI
LAYER 1
INTERFACE

crosshole survey.
A third pitfall is the use of incremental traveltimes between
two receiver boreholes to determine the velocity profile. This is
often done to avoid the need for knowing zero time However, II\ IIT 2

II
considering the receiver borehole spacings commonly used and
the thinly stratified nature of the near-surface deposits at many
/
II
sites, the first arrivals at the two receivers probably do not repre-
sent direct-path arrivals through the same unit or even refracted _1’ --_II

arrivals along the same interface. A fourth pitfall is the assumption


3
that any two boreholes are parallel.

CRITICAL PARAMETER CONSIDERATIONS FIG. 6. Three-layer case geometry.


An example
A familiar example illustrates the requirements for borehole
spacing and in-hole station spacing. Figure 3 shows the P- and
S-wave velocity profiles at a constructionsite in the San Francisco
Bay area (Schwarz, 1970). The profiles were obtained by down-
hole seismic methods. Assuming these profiles, consider a hypo-
thetical crosshole survey at this site to determine the S-wave
velocity profile.
The primary feature of interest and concern at this site is the
Bay Mud low-velocity layer. For a crosshole station centered in
the Bay Mud layer, the effect of various borehole spacingson the
computedapparentS-wave velocity basedon first arrival time data
is shown in Figure 4. Note that the apparent velocity equals the
true velocity only for borehole spacingsless than about 25 ft. For
a borehole spacing of 100 ft (typical of many surveys conducted
in the past), the apparent velocity is too large by a factor of two.
For a properly chosen borehole spacing of, say, 20 ft, Figure 5
illustrates the variation of apparent S-wave velocity as a function
of depth below the upper layer (layer 1). There is a zone approxi-
mately 6.5 ft thick within which a source-receiver station records
I I I I I
a direct-path first arrival through layer 2, and a zone only 9 ft thick 0 0.2 0.4 0.6 0.8 1.0
in which the computed apparent velocity is within IO percent of +$)

the true velocity. Thus, it is apparent that if the Bay Mud were
an unknown stratumin a crossholesurvey, with a borehole spacing FIG. 7. Plot of dimensionlesstraveltime equation for four values of
of 20 ft and an in-hole station spacing of IO ft or greater, the true velocity ratio va/va.
26 Butler and Cum

velocity might not be determined. However. with an in-hole sta-


0.6 r-
tion spacingof 6 ft or less, the true velocity would be determined.

Generalization
These concepts regarding the borehole spacing and in-bole
station spacing can be generalized. Assume a three-layer case
with velocity increasingdownward (18~< v2 < vs), so that only a
direct path or a refraction along interface II can give a first arrival
for a crossholestation at depth z below interface I (Figure 6). The
traveltime equation for the refracted path can be nondimen-
sionalized as follows:
T = [I - 2Ztan(sinP’V)]V + 2Z/cos(sin-‘V) (1)
= ma/D,
where
T-2

O- where t is the traveltime of the first arrival, T is the thickness of


0 0.5 I .o layer 2, and D is the borehole separation. Figure 7 presentsplots
of equation (1) for four values of the velocity ratio V.
v (= vz/vs)
The value 7 = 1.O representsthe case where the direct and re-
fracted path traveltimes are equal. For 1/T < 1.O, the direct ray
FIG. 8. Variation of dimensionlessdistance with velocity ratio for
case where direct and refracted path traveltimes are equal. will be the first arrival and the apparent velocity (VA) equals the
true velocity (vz). The values of Z for which 7 = 1.0, Z,, are
significant since for all 2 2 Z,, for a given V. vA = v2.
Solving for Z, gives
Z, = l/2 cos (sin-‘V)/(l + V), (2)
which is plotted in Figure 8. This simple relation can be examined
for various limiting cases of interest. As one example, consider
the case for z = 0, i.e., where the source-receiver station is at the
top of layer 2 and T becomes the lninimum layer thickness for
which VA = v2. Figure 9 illustrates the simple relation between

I-

k-x
S.P.
v

I f
Z
LAYERI VI=1150 FT/SEC AZ,= 20 FT

II
INTERFACE1

2 V,=710 FTiSEC AZ,=20 FT

Tr

3 I',=1500 FT!SEC AZ,= 40 FT

5 IO 15
T (ft or m)

FIG. 9. Variation of minimum layer thickness (T) with borehole 4 V,=5650 FT/SEC
spacing(0) for which the first arrival (with source-receiverstation
at top of layer) will yield the true layer velocity. FIG. 10. Idealized P-wave profile deduced from crosshole survey.
Crosshole Seismic Testing 27

0.20

5650 FT/SEC)

0. I5

v
z
w^ 0.10
I
I=

0.05

0 I
50 100 150 200 250 300
DISTANCE X, FT

FIG. 11. Time-distance plot for hypothetical seismic refraction survey over crosshole profile of Figure 10

D and T for this case (i.e., where z = 0), for the two limiting were 150 to 200 ft long. Only a single velocity of - 1110 ft/sec
values of the velocity ratio and an intermediate value. For a given was indicated, seemingly disproving the crosshole results. How-
value of D and v2/v3, the relation in Figure 9 gives the minimum ever, accepting the crosshole profile (Figure lo), it is easy to see
T for which the first arrival will define ~2. Also, for all values of that a surface refraction time-distance plot would appear as shown
D and T to the right of v2/va = 0, VA = ~2. in Figure 11, where the solid lines represent first arrivals. Note
Thus, if information is available for a given site, suchas a boring that to a distance of 205 ft, only a single velocity is indicated by
log, which includes typical thicknesses of geologic units and first arrivak. Not only is the !owue!ocity !ayer (!ayer 2) net in-
general descriptions, it is possible to plan the crosshole survey dicated, but layer 3 is not representedas a first arrival. Thus, the
(borehole spacing and in-hole station spacing) to optimize refraction survey results, based on 150- to 200.ft lines, are entirely
delineation. consistenf with the crosshole profile.

FIELD EXAMPLES Coordinated seismic investigation of +I building foundation

Detection of velocity inversions The second field example illustrates the proceduresfor arriving
at a moduli profile for the foundation of a building site (Curro and
The routine detection of velocity inversions or low-velocity
Marcuson, 1978). Figure 12 illustrates the seismic test layout for
zones is one of the great advantages of the crosshole method.
the foundation of a proposed structure. OFviously, the drawing is
Figure 10 illustratesan idealized P-wave velocity profile, exhibit-
ing a low-velocity zone (layer 2). This profile was deducedfrom a
crosshole investigation along the base of an existing dam. The
in-line crosshole set geometry was used with source-receiver
spacings of 20 and 120 ft. The shorter source-receiver spacing
data were used to deduce the model profile, and the longer spacing
data and the incremental data were used for model verification.
Data from boring logs and Standard Penetration Tests indicated
a “loose zone” corresponding to the low-velocity zone. Also,
the S-wave velocity profile showed a velocity inversion over the
same depth interval (v, - 360 ft/sec for the low-velocity zone).
However, the results of this investigation were reviewed by a con-
sultant, who maintained the velocities were too low compared to
velocities reported in the literature for similar materials. Un-
fortunately, the data for similar materials used by the consultant
were basedprimarily on crossholetest resultsobtained using bore-
hole spacingswhich were far too large to define true velocities.
Sukequently, a surface refraction survey was conducted over
the crosshole test locations by a third party, using lines which FIG. 12. Seismic test layout for building foundation (not to scale).
28 Butler and Curro
BORING 809 BORING 809 P-WAVE VELOCITY P-WAVE VELOCITY
WATER CONTENT, % BORING 809 VISUAL INSPECTION PROFILE PROFILE
DRILLER’S LOG LOG-JAR SAMPLES 809.810 809-811

CH 3573
I-l
Br CH ii/SEC
Laminated Br-Gr
Layered
10 3618

1
FT/SEC

ML
ML
Firm, Y-Br
Br
w/gravel
4367
20
Clayey
.---.
t Tr, R-Br CL
4422
z .---_
Clayey Br
Y-B,
i d
Gr
Gr-Br 3317 3275
30
c9 Br

ML
% b
Gr
2211 2068
ML
.I cl Firm, Gr,
CL-ML, Gr-Br
w/gravel
40
0
w/gravel
SM: Firm, Gr. ML-GM
0
w/gravel Gr
tong
at hot
SH 7935 8978
SH
H, dGr
Bk-dGr
50 F ?O

LEGEND

CH = INORGANIC, HIGH PLASTICITY CLAY


ML = INORGANIC SILTS AND VERY FINE SANDS
CL = INORGANIC, LOW PLASTICITY CLAY
SM = SILTY SANDS
GM = SILTY GRAVELS
SH = SHALE

FIG. 13. Physical property data for subsurfacematerials and P-wave true velocity profiles for crosshole set 8OY-810-811 (see Figure 12).

BORINGS BORINGS
809, 810, 811 812, 813, 814

T
0 36001
_ 1700 16901 0
---
2500’ 0 3600’ --- 16901 0
T
0 4ifi-’ - 2775*
-- 3125
0 4175’ --
5000* --
0 63901 4800 ---
---
--
0 32951 ---_

3000* 0 21401
2650
0 21401
---_
I
0 8600’ ---_ -- --
8500 6000’

--I-

:420*’

FIG. 14. P-wave true velocity profile for building foundation. *-from downhole tests. **-from refraction tests. $-from crosshole
tests.
Crosshole Seismic Testing 29

Y Jooo’~

BORINGS BORINGS
809, 810, 811 812, 813, 814
-
G = 3700 psi E = 11,000 psi 1’ = 0.47
o----------__-_-__o

0 0

0 G = 11,200 psi E = 32,700 psi v = 0.46 0

-PO O P
-- ----
0 ---_---_ -0

0 0

0 G = 23,800 psi E = 69,300 psi v = 0.46 0


- 0 0
-_-_ ---_
0 ----__

G = 61,300 psi E = 182,200 psi ” = 0.49

619 L

FIG. 15. Moduli profile for building foundation. G = shear modulus; E = Young‘s modulus: I’ - Poisson’s ratio,

not to scale. The source-receiver borehole spacing for the cross- the existing knowledge of the geology of the site and the resolution
hole tests is 20 ft. A downhole test is indicated at each source required by subsequentuse of the data. Proper advance planning
borehole, and the in-hole station spacing was 5 ft for both the can ensure that at least the optimum (minimum) data set is ob-
crosshole and downhole tests. The surface refraction lines are tained at a site (for the known geology and required resolution).
240 ft long. The crosshole procedure in this paper avoids the collection of
Figure 13 illustrates the second phase of the interpretation large volumes of data such as generatedby methods using multiple
process, where the data for the two source-receiver pairs in one fixed-depth receivers. Data from the multiple-receiver surveys
of the crosshole sets have already been processed with an inter- are properly analyzed with a sophisticated iterative, ray-tracing,
pretive computer code (Butler et al, 1978) to yield true P-wave finite-difference computer code (Bois et al. 197 I), and the results
velocity profiles. The two profiles are compared with each other, from such an analysis are commonly not in a form readily usable
with boring logs, and with any available physical property data. by engineers. While the procedure described in this paper should
In the absence of any apparent, significant anisotropy, the two not be expected to work well in delineating localized anomalies
profiles are averaged to give a single representative profile for the between the boreholes, it has produced excellent results in most
crosshole set. applications to date.
The results of the two crosshole sets are then compared with
REFERENCES
the downhole and surface refraction results to deduce a P-wave
velocity profile for the building site as shown in Figure 14. Sim- Bois, P., LaPorte, M., Lavergne. M., andThomas. G., 1971, Com-
puterized determination of seismic velocities between well shafts: Geo-
ilarly, the S-wave velocity data are analyzed to deduce the S-wave phys. Prosp.,v. 19, p. 42-83.
velocity profile. Finally, the resultsof primary interestto engineers, Butler, D. K., Skoglund, G. R., and Landers, G. B., 1978. Crosshole:
the recommended moduli profile for the site (Figure 1.5) are de- An interpretive computer code for crosshole seismic test results.
documentation. and examples: Misc. paper S-78-8. US Army Engineer
duced from the velocity profiles and measured densities. Waterways Experiment Station, CE, Vicksburg. MS.
Curro. J. R.. and Marcuson, W. F.. 1978. In-situ and laboratory deter-
minations of shear and Young’s moduli for the Portsmouth. Ohio,
gaseousdiffusmn add-on site: MISC. paper S-78-12. US Army Engineer
CONCLUSIONS Waterways Experiment Station. CE. Vicksburg, MS.
Schwarz, S. D.. 1970, Site evaluation-geophysical exploration, in Obser-
Crosshole seismic surveys for determining velocity profiles for rational methods in soil and rock mechamcs: Lecture Series, University
critical structure-sitingstudies can be rationally planned based on of Illinois at Chicago Circle.

View publication stats

You might also like