US 4,241,429 AGrant

Velocity Determination and Stacking Process From Seismic Exploration of Three Dimensional Reflection Geometry

Issue Date:1980-12-23

•8 Claims

•6 Drawing Sheets

Abstract

In seismic exploration, linear, multiple fold, common depth point sets of seismograms with three dimensional reflection geometry are used to determine the dip and strike of the subsurface reflecting interfaces and the average velocity of the path of the seismic energy to the reflecting interface. The reflections in each set appear with time differences on a hyperbola with trace spacings determined by the source receiver coordinate distance along the lines of exploration. The offset of the apex of this hyperbola is determined from a normal move-out velocity search of the type performed on two dimensional CDP sets. This search identifies the correct stacking velocity and hyperbola offset which are used to determine dip, strike and average velocity.

Metadata

Assignee

Mobil Oil Corporation

Inventors

Marvin G. Bloomquist
Yung-liang Wang

Application Information

Application Number:US 8830878

Filing Date:1978-03-03

Priority Date:1978-03-03

Art Unit:222

Classifications

IPC:

G01V 136

Field of Search:

3403643315.5 TC;15.5 TD;15.5 DP4211 E;1 HH

Patent Drawings (6 sheets)

Description

Background of the Invention

This invention relates to seismic exploration and more particularly to the determination of velocity, dip and strike from common depth point sets of seismograms with three dimensional reflection geometry.

In seismic exploration, the measurement of velocity, dip and strike is important. From these parameters, the seismograms can be normal move-out corrected and common-depth-point traces can be stacked to closely represent the nature of the subsurface of the earth.

Normally, seismic exploration is carried out along a line. Pulses of seismic energy are generated by sources at spaced points along the line of exploration, and seismic energy reflected from subsurface interfaces is detected at spaced points along this line. The resulting seismograms can be easily corrected for normal move-out and velocity can be determined. Recently, a type of exploration in which seismograms are gathered into sets, each representing reflections from common depth points, has become prevalent. U.S. Pat. No. 3,697,939, Musgrave, describes common depth point (CDP) seismic exploration and the method of determining velocity from common depth point sets of seismograms.

U.S. Pat. No. 3,417,370, Brey, describes a method of determining velocity from CDP sets by searching the sets with various normal move-out velocities. The signal power across these seismic sets is detected and the maximum signal power identifies the correct normal move-out velocity.

U.S. Pat. No. 3,696,331, Guinzy et al, describes a velocity search procedure which also identifies the dip of the subsurface formations along the line of exploration.

Obtaining velocity from common depth point sets with two dimensional reflection geometry as described in the aforementioned patents is extensively used. However, there are certain situations in which seismic exploration with three dimensional reflection geometry is highly desirable. Three dimensional exploration has a high density of traces which provides greater accuracy and more detailed subsurface information that is especially important in geologically complicated areas. The processing of seismograms from three dimensional exploration is more complicated. However, the unit cost of three dimensional exploration is much less for the same information if a large number of receivers for each source are used. Moreover, in areas of unfavorable surface conditions, such as swamps, no permit areas, rivers or undershooting in salt-dome and shale-mass areas, the three dimensional CDP exploration becomes a necessity.

Summary of the Invention

In accordance with this invention, common depth point sets of seismograms with three dimensional reflection geometry are used to obtain the dip and strike of the subsurface reflecting interfaces and the average velocity over the paths of the seismic energy to the interfaces.

In accordance with one aspect of the invention, the CDP sets are normal move-out corrected with trace spacings determined by the source receiver coordinate distance along the lines of exploration instead of the true separation distances. With this arrangement, the reflection times on the traces before normal moveout correction of the set fall on a hyperbolic arc whose apex is offset from the point along the line closest to the shot point. By performing a linear step-out time search on the normal move-out corrected traces the stacking velocity and step-out time are measured. From these, the offset of the hyperbola and the correct hyperbola velocity can be determined.

In accordance with this invention, the step-out time offset and the hyperbola velocity are easily converted into average velocity, dip and strike. From these parameters the normal-move-out time is derived and applied for stacking. SHORT DESCRIPTION OF THE DRAWINGS

FIG. 1A depicts a prior art seismic exploration technique for obtaining common depth point sets of seismograms;

FIG. 1B depicts the equivalent field procedure to FIG. 1A but with a non-dipping reflecting interface;

FIG. 1C depicts the CDP set for the four detected traces;

FIG. 2 depicts a field procedure with three dimensional reflection geometry;

FIG. 2A depicts the CDP sets with reflections which occur on an offset hyperbola;

FIG. 3 is a flow sheet depicting the process of the invention;

FIG. 4 depicts plots which aid in understanding the process;

FIG. 4A depicts a CDP set having three reflections;

FIG. 5 depicts the outputs of the entire normal moveout and stacking procedure;

FIG. 6 is an example of another form of three dimensional seismic exploration;

FIG. 6A is another example of three dimensional seismic exploration; and

FIG. 7 is a flow chart depicting a sub routine in more detail.

Description of the Preferred Embodiment

FIGS. 1A-1C depict the prior art of seismic exploration to obtain common depth point sets of seismograms with two dimensional reflection geometry.

In FIG. 1A, a pulse of acoustic energy is generated at 10. It is reflected from the common reflecting point 14 and detected by a detector at 10 to produce the reflection 10a in the set of CDP traces shown in FIG. 1C. Similarly, a pulse is generated at 11 and detected at 11a, a pulse is generated at 12 and detected at 12a, and a pulse of seismic energy is generated at 13 and detected at 13a.

FIG. 1B shows the equivalent field procedure to that of FIG. 1A as far as the reflection times are concerned but with a non-dipping reflecting interface and the median velocity is the same as the stacking velocity.

The four detected traces are the CDP set shown in FIG. 1C. All contain reflections from a short straight line segment through the common reflecting point 14 in the up-dip direction. The reflections across the CDP set in FIG. 1C fall on the hyperbolic arc 15 which has its apex at the time T.sub.O which is the time of the reflection on the trace with zero source-receiver spacing. The prior art teaches how to normal move-out correct such a CDP set to the zero source-receiver time T.sub.O. Then the traces can be stacked to produce a signal trace having a reflection with enhanced signal to noise power as shown in FIG. 1D.

In CDP sets with two dimensional reflection geometry as shown in FIGS. 1A-1C, the reflections fall on a hyperbola which has its apex at the zero source receiver position. Prior art techniques, such as described in the aforementioned Brey and Guinzy et al patents, search along the different hyperbolas associated with different searching velocities. The traces are cross-correlated or summed along these various hyperbolas to determine which hyperbola produces the maximum signal power. The hyperbola which produces the maximum signal power is associated with the correct stacking velocity which is thereby determined. The aforementioned Guinzy et al patent also describes, column 5, line 30, a searching technique for use on field, or monitor, records which are not CDP sets. In such field records, the reflections also occur on a hyperbola but the apex is offset when there is a dipping formation. The Guinzy et al technique searches through all velocities and dips to find the correct hyperbola velocity and the correct hyperbola apex offset.

Three dimensional reflection geometry, as shown in FIG. 2, produces CDP sets with reflections which occur on an offset hyperbola as shown in FIG. 2A. In accordance with this invention, the offset h and hyperbola velocity V.sub.a are easily determined. From this offset h and the determined hyperbola velocity, the average velocity, dip and strike of the subsurface interface are easily determined. These are used to correct the traces of the set to the time T.sub.M, which is the zero source-receiver time at the CDP, so that the traces can be stacked for signal enhancement.

Referring now to FIG. 2 in more detail, the basic three dimensional seismic exploration geometry includes generating pulses at spaced points along at least one line of exploration 16 and detecting the seismic energy at spaced points along at least one other parallel line 17. The lines of exploration 16 and 17 are on the horizontal datum plane ABCD. Seismic energy is reflected from the subsurface interface plane EFGH which has a dip .theta. and a strike .phi.. The origin 18 is referred to as the common depth point or common surface mid-point. The source line 16 and the geophone line 17 are parallel to the X axis and the distance between them is .eta.. Such exploration produces linear (parallel to exploration line), multiple fold, common depth point sets of seismograms with three dimensional reflection geometry. In accordance with this invention, the dip, strike and velocity are easily determined from such CDP sets.

FIG. 3 depicts the process of this invention and FIGS. 4 and 4A show plots which aid in understanding the process. FIG. 4A depicts a CDP set having three reflections.

The first step of the process, indicated at 19 FIG. 3 is to correct the set for various normal move-out (NMO), or searching velocities V.sub.NMO. After time shifting these traces for an assumed NMO velocity of 5,000 feet per second, the reflections appear along the heavy lines shown at the upper left hand corner of FIG. 4. Each trace has been normal move-out corrected by an amount ##EQU1## where x is the distance along the x-axis between shotpoint and receiver, T.sub.x is the uncorrected time for the trace at x, and V.sub.NMO is 5,000 feet per second. After time shifting by this amount, the first reflections all fall along the straight line 20. The second and third reflections occur along the curved lines 21 and 22.

In accordance with this invention, after normal move-out for each searching velocity, a signal power search is performed along straight lines associated with various step-out times (.DELTA.t), the procedure being denoted 20 in FIG. 3. One signal power detection technique is to detect or sum the amplitude of the traces along the straight line. For example, the samples at 23-28 are all summed to produce a signal power which is plotted at 29 in FIG. 4. Another signal power detection technique is cross-correlation. By cross-correlating the windows of the traces at the times 23-28, a similar high signal power will be detected. Because the first reflections occur on the straight line 20, after normal move-out correction with V.sub.NMO =5,000 feet per second, the high signal power indicated at 29 is produced. However, a high signal power is not produced for the second and third reflections.

The same search is performed for other searching velocities. The set 30 represents the traces after being NMO corrected for a searching velocity of 6,000 feet per second. None of the three reflections occurs along a straight line and there is no high signal power output. The set 31 has been NMO corrected with a searching velocity of 7,000 feet per second. The second reflection occurs along the straight line 32. Accordingly, detection of the signal power along the straight line 32 produces a high signal power which is recorded at 33. The set 34 has been NMO corrected with a searching velocity of 17,000 feet per second. The third reflections occur along the straight line 35. Detection of the signal power along this straight line produces a high indicated signal power plotted at 36.

The upper right-hand corner of FIG. 4 shows the signal powers 29, 33 and 36 plotted as a function of velocity. This identifies the correct stacking velocity V.sub.S (that is the value of V.sub.NMO which gives maximum power) for the first reflection to be 5,000 feet per second, the correct stacking velocity V.sub.S for the second reflection to be 7,000 feet per second, and the correct stacking velocity V.sub.S for the third reflection to be 17,000 feet per second.

The step of selecting the straight line which produces the maximum detected signal power is indicated at 37 in the flow chart of FIG. 3. In accordance with this invention, the slope .DELTA.t of the straight line which produced maximum detected signal power is also determined. The slope of the line 25 is determined to be +0.040 milliseconds per foot. The slope of the straight line 32 is determined to be -0.015 milliseconds per foot. The slope of the straight line 35 is determined to be -0.045 milliseconds per foot. The plot at the lower right-hand corner of FIG. 4 shows the three signal powers 29, 33 and 36 plotted on a dip scale which identifies the aforementioned slopes of the straight lines 20, 32 and 35.

The plot 38 shows maximum signal power as a function of record time. It identifies the T.sub.O 's of the three reflections as occurring at 0.11 seconds, 0.20 seconds, and 0.285 seconds. From the determined values of stacking velocity V.sub.S, slope .DELTA.t and record time T.sub.O, the hyperbola-velocity and the apex offset of the hyperbola on which the reflections occur can be determined. The hyperbola-velocity V.sub.a and hyperbola offset h are related to the determined stacking velocity V.sub.S, slope .DELTA.t and time T.sub.O by ##EQU2## From two sets of (V.sub.a, .DELTA.t), the average velocity V, the dip .theta. and the strike .phi. of the reflecting interface can be determined as indicated by step 39 in FIG. 3. Specifically, for two CDP along the line of exploration as shown in FIG. 6, the velocity dip and strike are determined from ##EQU3##

.delta.x=distance between CDP points,

.DELTA.t is the slope of the straight line for a CDP set,

T.sub.O is the travel time at x=0 for a CDP set,

.eta. is the distance between the line of exploration on which pulses of seismic energy is generated and the line of exploration on which seismic energy is detected and

subscripts 1 and 2 denote two adjacent CDP sets.

From average velocity V, dip .theta. and strike .phi., the time shift .delta.t required to correct each trace to the zero source receiver distance can be determined. This is indicated at 40 in the flow sheet of FIG. 3 and is carried out in accordance with: ##EQU4## where

T.sub.x = uncorrected travel time at trace located at x,

sin .alpha..sub.c =cos .phi. sin .theta., and

sin .beta..sub.c =sin .phi. sin .theta..

This time .delta.t can be used to normal move-out correct the traces as indicated by step 41 in FIG. 3.

The steps 19, 20 and 37 are performed in a routine referred to as VIPSKEW. This procedure is repeated for adjacent sets of traces as indicated by step 42 in the flow chart. All adjacent sets are processed as indicated by step 43. This can be better understood from reference to FIG. 5 which depicts the outputs of the entire normal move-out and stacking procedure. The top of FIG. 5 depicts the two parallel lines 16 and 17 on the surface of the earth. Pulses of seismic energy are generated at spaced points along the line 16 and detected at the detectors at spaced points along the line 17. As indicated in FIG. 5, fifty-seven CDP sets are produced. Each set is comprised of twelve traces, with each set representing reflections from a common depth point. These include the CDP sets 44-48 and others. For each two CDP sets, values of average velocity V, dip .theta. and strike .phi. are determined for each reflection window on each CDP set. From these values of V, .theta. and .phi., the normal move-out time shift .delta.t is determined in accordance with step 40 of the flow sheet in FIG. 3. The normal move-out corrected sets, the result of step 41 in FIG. 3, are shown at 49-53 in FIG. 5. These are stacked to produce the stacked section 54, the stacking being indicated as step 55 in FIG. 3.

The present invention may be practised with other forms of three dimensional seismic exploration. Another example is shown in FIG. 6A. In this case, pulses of acoustic energy are produced at spaced points along two parallel lines of exploration 56 and 57. Seismic energy is detected at spaced points along the parallel lines 58 and 59. The searching procedure performed on the data is the same as that discussed above. However, in this case, the determination of hyperbola velocity and hyperbola offset h are given by:

where

T.sub.a1 =minimum (apex) travel time for set 1

T.sub.a2 =minimum (apex) travel time for set 2 ##EQU5## This is a quadratic in .alpha. which is solved for the real positive root.

FIG. 7 is a flow chart depicting procedure VIPSKEW in more detail. The inputs to the procedure are linear, multiple fold, common depth points sets of seismograms with three dimensional reflection geometry. The output is a plot of the type shown at the bottom of FIG. 4. The output is also digitized for further processing in accordance with steps 39, 40, 41 and 55. It will be understood that these steps are standard computational or normal move-out correction steps.

An exemplary program listing for carrying out the VIPSKEW procedure is given in the appendix. This program is written in Fortran language for performance on commercially available Control Data Corporation computers.

While a particular embodiment of the invention has been shown and described, various modifications are within the true spirit and scope of the invention. The appended claims are intended to cover all such modifications. ##SPC1##

Claims

What is claimed is:

1. The method of producing a stacked seismic section in seismic exploration comprising: generating linear, multiple fold, common depth point sets of seismograms with three dimensional reflection geometry, the reflections of seismic energy appearing across each set with time differences which are substantially on a hyperbola, normal move-out correcting said sets with trace spacings determined by the source receiver-coordinate distance along the lines of exploration and with given values of searching velocity and travel time, detecting the power of the seismic signal in the normal move-out corrected sets, determining from the maximum detected signal the hyperbola velocity and the offset of the apex of said hyperbola from the point of minimum source-receiver separation, from said hyperbola velocity and apex offset, determining the dip and strike of the subsurface interface producing reflections in said seismograms and the average velocity of the path of said seismic energy to said subsurface interface, and correcting said seismograms to zero source receiver time with time corrections determined from said average velocity, dip and strike, and stacking the corrected seismograms to produce a three-dimensional common depth point stacked seismic section.

2. The method recited in claim 1 further comprising: from said average velocity, dip and strike, determining the time correction for correcting each trace of a set to zero source-receiver time, and correcting said seismograms to zero source-receiver time.

3. The method recited in claim 2 further comprising: stacking the corrected seismograms to produce a three dimensional common depth point stacked seismic section.

4. The method recited in claim 1 wherein hyperbola velocity V.sub.a and hyperbola offset h are determined from stacking velocity V.sub.s and slope .DELTA.t by the following: ##EQU6##

5. The method recited in claim 1 wherein the step of normal move-out correcting includes time shifting each trace in a set by an amount: ##EQU7## where T.sub.x is the travel time on each trace in a set, X is the source receiver coordinate distance along said lines of exploration, and V.sub.nmo is the searching velocity.

6. The method recited in claim 1 wherein the step of generating sets of seismograms includes: generating pulses of seismic energy at spaced points along at least one line of exploration; detecting seismograms representing the reflection of seismic energy at spaced points along at least one other parallel line of exploration, and gathering detected seismograms into sets representing reflections from common depth points.

7. In seismic exploration wherein seismograms are obtained from three dimensional reflection geometry, the method of correcting said seismograms to zero source-receiver times and producing a stacked seismic section comprising: generating pulses of seismic energy at spaced points along at least one line of exploration; detecting seismograms representing the reflection of seismic energy at spaced points along at least one other parallel line of exploration; gathering detected seismograms into sets representing reflections from common depth points; normal move-out correcting said sets for given values of searching velocity V and travel time T.sub.O, detecting the power of the seismic signal in the normal move-out corrected sets, determining the stacking velocity associated with the maximum detected signal and the slope of the straight line through common reflections on each normal move-out corrected set which produced said maximum detected signal; from said stacking velocity and slope, determining the dip and strike of the subsurface interface producing reflections in said seismograms and the average velocity of the path of said seismic energy to said subsurface interface; from said average velocity, dip and strike, determining the time correction for correcting each trace to zero source-receiver time; and stacking the corrected seismograms to produce a three dimensional common depth point stacked seismic section.

8. The method recited in claim 7 wherein the step of determining dip, strike and average velocity is carried out using two adjacent CDP sets and wherein the step of determining the dip .theta., strike .phi., and average velocity V is performed in accordance with ##EQU8## where ##EQU9## .delta.x=distance between CDP points, V.sub.a is hyperbola velocity, .DELTA.t is the slope of the straight line for a CDP set, T.sub.O is travel time for trace at x=0, .eta. is the distance between the line of exploration on which pulses of seismic energy is generated and the line of exploration on which seismic energy is detected and, subscripts 1 and 2 denote the two adjacent CDP sets along the line of exploration.

Patent Citations (2)

Patent	Date	Inventor	Cited By
US3417370	1968-12-01	Brey
US3696331	1972-10-01	Guinzy et al.