Sei sulla pagina 1di 7

Downloaded 03/28/15 to 169.230.243.252. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.

org/

APPARENT VELOCITY FROM DIPPING INTERFACE REFLECTIONS?

F. K. LEVIN*

When a seismic reflector is a dipping plane and are not common, reflections stack even for steep
the subsurface can be approximated by a single dips. Stacking velocities for multiple reflections
bed, the velocity needed to stack CDP data is increase with the order of the multiple, but mul-
higher than the velocity with which the energy tiple reflections may also be stacked. To convert
travels in the subsurface. The ratio of the stacking from reflection time to depth rrcluires knowledge
velocity to the energy-travel velocity increases not given by the stacking velocity. In theory, a
from unity for strike lines to the secant of the dip split-spread procedure can furnish the needed
angle for dip lines. Although the reflection points information.

INTRODUCTION THE PRIMARY REFIX< TION

The effect of a dipping reflector on common The subsurface we shall conxider is shown in
depth point (CDP) processing has been consid- Figure 1. Because we have only one bed, the dis-
ered by several investigators. Cressman (1968) tinction between rms velocity, interval velocity,
pointed out that because of the way the normal and average velocity that must IX made when the
moveout (NM()) correction is applied, CDP subsurface is layered has no meaning here; we can
stacking works as well for dipping beds as it does assume there is a constant velocity 1’ betbveen the
for horizontal beds; although in the former case surface of the ground and the I-cflector. To keep
the reflections being summed do not have com- our deviation general, we write I IIC:equation of the
mon reflection points. Brown (1969) examined in reflecting plane as
detail the NM0 correction and velocity relations
for dipping beds. Both Cressman and Brown dealt s cos a + y cos /3 + 2 (OS + = d, (1)
with realistic subsurfaces, subsurfaces character-
ized by velocities varying with depth. In a recent where cr is the angle between the normal to the
paper, Taner and Koehler (1969) avoided the plane and the profile direction, taken here to be
difficulties associated with dipping bed calcula- the x axis, and 4 is the dip angle of the plane. LY,
tions by tracing rays with a digital computer. /3, and $I arc direction cosines ol’ (he normal to the
A physical understanding of the dependence of plane. We place our source at Ihc origin (0, 0, 0)
CDP-determined velocities on dip can be ob- and the geophone, at (X, 0, 0). d is the distance
tained if we restrict ourselves to an extremely from the source to the reflecting plane, measured
simple subsurface; namely, one with a single along the perpendicular to lh(, plane. n is the
reflecting plane. Although stemming from an corresponding distance from ;I point along the
unreasonably oversimplified picture, the expres- profile hallway between the source and geophone.
sions derived from such a subsurface are those Notice that the line along which D is measured
actually used in much CDP data processing. does not intersect the reflector at the reflection
Hence, there is some reason for presenting our point. The difference between silrgle-fold and CDI’
results in spite of their obvious limitations. geometries can be summarized as hollows: in the

t Manuscript receivedby the Editor September29, 1970;revisedmanuscriptreceivedJanuar!. 19, 1071.


* EssoProductionResearchCompany, Houston, Texas 7ioOl.
@ 1971by the Societyof Exploration Geophysicists.All rights rcscrvcd.
510
Velocity from Dipping Interface 511
Downloaded 03/28/15 to 169.230.243.252. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

FIG. I, Drawing of a layer bounded by the surface and a dipping, plane reflector. The source is at (0, 0, 0); the
geophone.at (X, 0, 0).

CUP method, 11 is constant; for single-fold ex- (4). First, to the degree that a constant velocity
ploration, d is constant. to a reflector is a good approximation, the use of
The mathematics of computing reflection Fi,~fo lets us stack CDP data lor steep or for
traveltime involves nothing more complicated small dips. The steepness of the dip doesn’t affect
than setting up an image geophone (or an image our ability to stack. Second, the time intercept for
source). As shown by Slotnick (1959), the travel- zero spread length is the two-way- distance at the
time to and from the reflector is given by halfway point divided by the constant velocity V.
As long as the source and geophone stations are
W” = Id” + X2 - 4dX cos a. (2) arranged symmetrically around a surface point,
Rewriting equation (2) in terms of D, we have the data may be stacked; although clearly we do
not stack data from the same subsurface points.
Third, the stacking velocity will always be the
same as or higher than I’.
Let us look more closely at the way I’~bro
depends on the direction of the profile line relative
to the direction of the dip of die reflector. The
angle (Y is not one that is available to an explora-
tion geophysicist. We normally think in terms of
the dip angle 4 and the angle between the profile
line and a dip line. We’ll call this latter angle 0
(Figure 2). In terms of 0 and 4, ccluation (3) is

VNnro,/V = (1 - sin* 9 c.05~0)-1’z. (5)


Hence, there is an apparent velocity l’r;hro which
will flatten the reflection; but this apparent In Figure 3, we plot the ratio C’N ,,f,,/V as a func-
velocity is not the same as V. tion of 0 for several dip angles. For profiles in the
We can draw several conclusions from equation strike direction, VNMO and I;’ arc: the same. For
512 Levin

/PROFILE LINE
Downloaded 03/28/15 to 169.230.243.252. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

---_--------

profiles in any other direction, Iys3t() is greater velocities found for lines in different directions,
than 1’. The greatest difference occurs ior profiles the average will he greater than I’ but less than
in the dip direction. For dip lines, l~sm~,,,;
Conversion of traveltime to depth for a layered
l’?m,u,, I’ = set 4. (6) subsurface is a complicated, sequential process.
.4 plot of this expression (Figure 1) shows that for
dips of less than 10 degrees, the velocity difference 1.04
can be ignored. It is obvious that if n-e average

1.03

0” 1.02
z
>’

1.01

/’
90 60 30 0
B - DEGREES l.OO-’
I 1 0 4 8 12 16
STRIKE DIP
+ - DEGREES
FIG. 3. Ratio of lys~,~ Lo
..
I’ as a function of 0 for several
.
dip angles. FIG. 4. Ratio of I’Nwo~,,~
“.r to 1’ for small angles of din.
Velocity from Dipping Interface 513

For the simple section of Figure 1, the conversion


is easy. If we want depth measured along a line
Downloaded 03/28/15 to 169.230.243.252. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

perpendicular to the reflector, we use the velocity


1.. On the other hand, if we want depth measured If the dip is small, we may replace sin 24 by 24.
along a line perpendicular to the surface and if the To get some idea as to the magnitude oi A, we
profile is a dip line, we use L-NN.\I().Sote that the consider a 10,000 it source-geophone separation,
line must be a dip line. Xn incidental result of the a reflector at 10,000 ft, and a 30 degree dip.
difference between I’ and I’slro is the following: A= 1100 ft. It is doubtful if the idea of a reflection
for a dip line, if l.shto is used to compute the dip, point moving along the reflector should be taken
the apparent dip from a record time section is the too seriously, since \vavefronts and not rays are
actual dip. reflected. The relation between the t\vo is dis-
In practice, a geophysicist finds l’Nsf\Ioby apply- cussedby Trorey (1970).
ing a velocity determination program to CI>I’- Due to our overly simplified model, we’ve
gathered data. To COI -;rt to depth, he will want ignored many important effects. Most important,
1.. If he is lucky enough to have cross-spreads, 1? of course, are all effects due to refraction at inter-
and 4 may be found by well-known procedures faces between the surface and the reflector being
(Slotnick, 1959). If, as is almost always true in considered. However, CDP data stack even when
marine seismology, he has only single-ended dips are appreciable. Hence, calculations of the
spreatls, 1’ may still be found if the field data type described by Taner and Koehler (1969) are
acquisition procedures involved split-spread shoot- likely to modify but not invalidate the conclusions
ing or if the data can be reassembled in split- drawn here.
spread form. The method is discussedat length by
MULTIPLE REFLECTIONS
Slotnick (1959). We designate as 1, and 1_ the
reflection times measured by geophones equally The mathematics developed for the primary
distant from but on opposite sides of the source. reflection carries over with very little change when
\Ve define an average time 7 by multiples are being considered. .\gain image
sources are used to get the time-distance relation
1, + t? 2d ? X’ 20 sin (PZ+ l)+ 2
r?=-___ _ + -kq (‘i)
2
t,, =
2 0V V sin C$ 1
1. is found from a plot of r2 against .V?. The r- .Y 1 - cos* 0 sin’ (~2+ 1)4
hyperbola is symmetrical about the source but the
1- .Y hyperbola has its head displaced in the up-
+ ----~~~----
1 .Y’. (l(l)

dip direction. I-\s a practical method, the applica- In equation (lo), jz is the order oi the multiple;
tion of equation (7) is less attractive than it first for the primary reflection, 11=O. \Ve define a
appears, since the technique demands that a time- stacking velocity l’~lro, as
distance plot be available and, unlike modern
velocity determination computer programs, does V NII(),/V= [l-cos2,sin2 (?z+l)+]P”?. (11)
not operate directly on recorded data.
l’~r,lo,/V= 1 for a strike line and the maximum
We indicated earlier that D does not intersect
deviation from unity occurs for a dip line. For a
the reflector at the reflection point (Figure 1). The
dip line,
reflection point, which depends both on the spread
length S and on L), is up-dip relative to the inter- 1’NhIO, t V = sec(72 + 1)1$. (12)
<,I”
section point. If \ve call the distance between the
t\vo points A, we find (see appendix) that Plots of this function (Figure 5) show that the
ratio of velocities increases rapidly with multiple
X” .Y’ number, so that even for very small dips, the
A = 3. sin u cos 01 = ix sin 2a
multiple reflection stacking \relocity is appreci-
(8) ably different from 1’. The effect is even more
x pronounced when the data of Figure 5 are re-
= lz sin f#~cos 041 - sin* I#Icos *O.
plotted as Figure 6. (The curves in Figure 6 are
drawn to assist the eye; only the values at integer
For a strike line, A=O. For a dip line, 11’shave physical meaning.) Figure 7 differs from
Levin

432 1-N
Downloaded 03/28/15 to 169.230.243.252. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

9 - DEGREES #J- DEGREES

l:m. 5. Ratio 1’N41,hC,,p/l~’


as a function of dip
angle for n = 0 to a= 4.

Figure 5 in that the ratio ~~?;Mo~~~~/~;~NIN~~o~,,~~ is since to,, corresponds to S=O. As the multiple
plotted against #I, It is this ratio, rather than the number increases, the time selxiration of succes-
ratio of equation (12), that an interpreter is able sive multiples decreases (Figui-c 8). The reader,
to measure. distressed by a time intercept i ntlependent of dip
In contrast to their behavior for primary re- for primary reflections but dip tlependent for mul-
flections, time intercepts for multiple reflections tiple reflections, should realize that the apparent
depend on the dip angle. The intercept anomaly results from our defining to, in terms 0i
D, a distance that, for the 1)rimary reflection,
20 sin(n + 1)4 ‘Lpins” the reflector at a dip~in~l~,l)endelltdistance
to, = 7 -s,;r- (13)
from the spread midpoint.
Those who have had occasion to study light
from equation (I 0) is independent of the direction fringes in a wedge will recogni/.c> equation (1.3).
of the profile line relative to the direction of dip,

0 10 20 30
+ - DEGREES
I:Ic. 6. Ratio VNMO,~~_,/V
for small dip angles, I:m. 8. time interceptsas a function of dit>for primitry
n=O ton=9. reflection and first four multilJe reflections.
Velocity from Dipping Interface 515
Downloaded 03/28/15 to 169.230.243.252. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

\ RAY
2D/Sin+/POINT
/ \ /

1‘1~. 9. The geometry involved in the concel~t of image sources or image geophones for multillle reflections.

Physicists (Born and Wolf, 1YOA) have long the water layer is effectively deccqled from the
known that the image sources lie on a circle whose underlying section; the water we(lge can then be
center is the edge of the wedge (the outcrop oi the thought of as modellcd by Figure I. The velocity
reflector) and whose radius passes through the increase and time separation decrease with mul-
point being examined (Figure Y). This construc- tiple number predicted by equatiorrs (12) and (13)
tion and the fact that angles inscribed in a semi- are seen in the surface-to-bottom multiple reflec-
circle arc: right angles are sufficient to yield equa- tion velocities and times.
tion (I 3).
REFERENCES
Corresponding to equation (7) ior primary
reflections, vie can form for split spread data Born, M., and Wolf, E.. 1964, Princil)lcs of optics: New
York, Pergamon Press.
l3rown, R. J. S., 1969. Normal-mow~l~ul and velocity
relations for Oat and dipping lxds iu1~1for long offsets:
Geophysics, v. 34, p. 1X-195.
Cressman, K. S., 1968, How velocity Ix~zring and steep
dip affect (‘III’: Geophysics, v. 33, 11.399-411.
Slotnick. hl. n,f.,1959, Ixssons in wiimic computing:
Tulsa, Society of Exploration Geophysicists.
‘I’aner, M. ‘I‘.. and Roehler, I,‘., 1969, Velocity spectra-
digital computer derivation anti applications of
velocity functions: Geophysics. v. .U, p. 859-881.
The slope of the _Yz--7fLcurve gives I’ regardless Trorey, A. Lb’., 1970. 4 simple theor! for seismic dif-
of dip hut the time intercept is identical to that fractions: Geophysics, v. 3.5, p, 762 7x4.
found with CL)I’-ordered data.
APPENDIX
.A11the reservations expressed in our discussion
of primary reflections with regard to our over- All of the relations in the body of the report can
simplified picture of the subsurface are equally be derived by the use of geometry alone; however,
valid for multiple reflections. There is one situa- since some oi the readers may haw the author’s
tion for which our model is a good approximation difficulty with seeing rays in the-ee dimensions,
of reality. In marine exploration over a slope, the I’ll sketch the analytical geomet r! of the dcriva-
bottom in some casesis such a good reflector that tions.
516 Levin

the spread, (.1-, 2, 0, 01, to the plane along a line


perpendicular to the surface \VC nectl only
Downloaded 03/28/15 to 169.230.243.252. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

find
.Y cos Ly+ ?’ cos @ + 5 co.5 4 = (1, (1) the intersection 0i the line ha\ing the equation
a source at (0, 0, O), and a geophone at (S, 0, 0). s= 51’2, y=O with the plane. \\.e call the inter-
\Ve compute the position of an image geophone section point (.I-, 2, 0, Z’).
at the same distance below the plane along the s 2 cos a + %’ cos c#l = tl
perpendicular to the plane from (S, 0, 0) as the Or
geophone is above the plane. Since this type of %’ = (tl - s 2 cos a) (‘OS #B.
computation reoccurs repeatedly, 1’11 give it in
IVe let 11’ be the distance measured along the
detail. Live\zant the intersection point (.1-l, Y1, Z,)
which lies both on the plane and the normal to the perpendicular to the surface
plane: D’ = Z’ = D cos 4. (23)
If we define to=2D, TTand recall that TTx\l,,,,,,,
= I’,lcos 4, \ve find that

x1 cos cy + 1’1 cos p + %I cos I$ = d. 20’


[,, = - ~~- (h)

Hence, (S,, Y,, ZI) = [(d- rcos CY)cos


Ly+ s, I~S.\l,,,,,,,

(d-_Scos cu)cos8, (& Scos u)cos $1 and the 1‘ 0’ derive equation (8) for A, lye must find the
image geophone is at coordinates (SO, Y,, 20) of the point at which 11
(S,, I’,, %.‘) intersects the plane and the coordinates ( .Ys,J/3,Z:,)
of the reflection point. (_Yo,Y,, Z,) are i0untl 1~~
= [2(d-Xcos a) co5 o( + x, (d-X cos a) cosp,
the same procedure dkcussed in the second para-
2(d-x cos a) cos l#l]. graph of the appendix.
The distance travelccl by the reflection is the (.YO, 1’0, Z(l)
distance from the source at (0, 0, 0) to the image (A&)
geophone at (?ig, Y?, Z,) or = (D cos cy+ X 2, D cos /j, D cos 4).
1’” jS3, Yt, Z,) is the intersection point of the line
c’ t =x:+ ~~9+z~=ld?+?(?--~Xcosoc. (2) connecting the source at (0, 0, 0) and the image
D=(-_Scosa 2 or d=D+.Ycosa 2. (la) geophoiie at (S4, Y?, Zz).

(52)

Substituting equation (la) into equation (2), we A = [(S, - S,,)’ + (17:i - Y,,)
get
+ (Z:, - Z!(,)‘1’ ”

= ~- sin LYcos LY.


1D

To compute the travel paths oi multiple re-


flections, we extend the method used to find the
.I little manipulation shows that cos CY
= sin C#XOS
0. travel path oi the primary reflection. ‘The image
‘The expression for l_\‘h~~ becomes geophone for the first multiple is found by reflect-
ing (S,, YP, Z,) in the surface and then reflecting
V,v.~o = V(l - sin2 + cos* 0)-l ‘?. (3 this second image in the dipping plane. Higher-
D is measured along a line perpendicular to the order multiples require us to repeat the process of
plane. 7’0 find the distance from center point of reflecting first in the surface anti then in the plane.

Potrebbero piacerti anche