GB2138135A - Interpretation of seismic records - Google Patents

Abstract

In a method of interpreting seismic records, progressive changes in amplitude as a function of offset of common gathers can be more easily identified by emphasizing the degree of amplitude variation between "near" and "far" amplitude vs. time traces of each gather along a seismic line, and displaying the resulting near and far offset sections side-by-side. A significant - and progressive - change in P-wave reflection coefficient as a function of the angle of incidence (within sections) indicates valuable characteristics, say the fluid hydrocarbon-bearing potential and/or the lithology of the reflecting horizon.

Description

SPECIFICATION Improved method for the interpretation of seismic records to yield valuable characteristics, such as gas-bearing potential and lithology of strata The present invention pertains to the art of seismic prospecting for petroleum reservoirs by multiplepoint surveying techniques, and more particularly to the art of converting high-intensity reflection amplitude anomalies associated with one or more common centerpoints observed on seismic record traces into diagnostic indicators, say of both hydrocarbonbearing potential and lithology ofthe underlying subsurface strata.

Seismic prospecting for petroleum involves the creation of acoustic disturbances above, upon, orjust belowthe surface ofthe earth, using explosives, air guns, or large mechanical vibrators. Resulting acoustic waves propagate downwardly in the earth, and partially reflect back toward the surface when acoustic impedance changes within the earth are encoun tered.Achangefrom one rocktypeto another, for example, may be accompanied by an acoustic impedance change, so that the reflectivity of a particular layer depends on the velocity and density content between that layer and the layerwhich overlies it.

In early years, signal traces ofthe reflected acoustic waves were recorded immediately in the field as visible, side-by-side, dark, wiggly lines on white paper ("seismograms"). At present,the initial repro ductions -- i n a digital format- are on magnetic tape, and finally are reduced to visible side-by-side traces on paper orfilm in large central computing facilities.

At such centers, sophisticated processing makes possible the distinction ofsignalsfrom noise in cases thatwould have seemed hopeless in the early days of seismic prospecting. Until 1965, almost all seismic surveys conducted used an automatic gain control which continuously adjusted the gain of amplifiers in the field to account for decreasing amounts of energy from late reflection arrivals. As a result, reflection coefficients could not be accurately determined.

However, with the advent ofthe expander circuit and binary gain amplifiers, gain of the amplifiers can now be controlled and amplitudes recorded precisely; this makes it possible to conserve not only the special chai-acteristics of the reflections, but also their absolute amplitudes.

Today, more powerful computers with array processors and economical floating point capabilities also now enable modern geophysicists to maintian control ofthe amplitude of all recorded signals. The "floating point" capability is especially effective in expanding computerwork size by a large factor and in eliminating the need for computer automatic gain control. As a result ofthe above advances, reflections from many thousands of feet belowthe earth's surface can now be confidently detected and fol lowed through sometimes hundreds of side-by-side traces, the shortening or lengthening oftheircorres- ponding times of arrival being indicative of the shallowing or deepening of actual sedimentary strata of interest.

Apropos ofthe above has been use of ultra-high amplitude anomalies in seismic traces to inferthe presence of natural gas in situ. Seismic interpreters have used so-called "bright-spot" analysisto indicate several large gas reservoirs in the world, especially in the Gulf Coast of the United States. Such analysis is now rather common in the oil industry, but it is not without its critics. Not only cannot the persistence of such increased amplitude anomalies betaken as confirmation of the lateral extent of the gas reservoir, but also the anomaly itself (in some cases) maynot represent reflections of a discontinuity of a gasbearing medium and its over- or underlying associated rock strata.

However, the problem as to the degree an interpretercan rely on high-intensity anomalies, in these regards has recently been broughtto manageable proportions. In the above-identified copending ap plications, it istaughtthatgas-bearing potential and the lithology or one host and cap rock strata can be accurately determined by: (1) obtaining field data in which the data of common centerpoints are associated with more than onesource-detectorpair, (2) indexing the data whereby all recorded traces are indicated as being a product of respective sourcedetector pairs of known horizontal offset and centerpoint location, (3) thereafter, associating high-intensity amplitude anomalies in the traces in a manner that allows determination of both gas-bearing potential area the lithology ofthe host and cap strata to a surprisingly accurate degree.

The present invention further improves the ability ofthe seismologist to correctly identify the lithology and presence of hydrocarbonaceous fluids using certain occurrences in amplitude with offset of such high-intensity anomalies of the seismic records to differentiate the former from similarly patterned reflections of other types of configurations containing no accumulations.

In accordance with the present invention, prog ressive changes in amplitude as a function of offset of common gathers can be more easily identified by emphasizing the degree of amplitude variation between near offset and far offset traces (relative to a common sourcepoint) of each gather along a seismic line, and displaying resulting near and far offset sections. Result: the interpreter can easily follow progressive amplitude change in a mannerthat allows determination, say of both fluid hydrocarbon bearing potential, and the lithology of the host and cap rock strata to a surprisingly accurate degree.

In accordance with the present invention, systema tic generation of near and far offset traces and sections, is provided in a surprisingly efficient manner using pattern recognition methods that (although conventional in the data communications field) have not been used in the context here employed.

Assume, for example, that a T1th time sample exists for a gather of traces and the near and far offset traces for such time sample, are to be determined. In orderto carry out such a process, first, the traces amplitudes associated with the Tilth time sample are generated, i.e.,theAT,'sforexampleT,. Next, there is a fit ofthe formulated ATLAS to a series of linear and quadraticfunctionsoftheform: (;)A(X) = Co C1 x; (ii) A(x) = C0 + C2x2; and (iii) A(x) = C0 + C1x + C2x2.

Object : to obtain a least squares best fit of the generated AT1 amplitudes to either equation (i), (ii), or (iii), supra. Then amplitudes at preselected near and far values are calculated. That isto say, the best fitting above-identified linear and quadratic equation is solved for a preselected near and far offset value. The process then can be repeated, and then re-peated for sampletimesT2 . . .Tj, to generate a series of pairs of near and fartraces projected to preselected near and far locations offsetfrom the source point locations associated with the original gather oftraces.

Preferably, the near and far traces are grouped together to form a series of sections, best displayed on a side-by-side basis. Result: the interpreter can easily follow change in amplitude as a function of offset from section to section along the entire seismic survey line.

Furtherfeatures ofthe invention will become more apparent upon consideration of the following detailed description ofthe invention when taken in connection with the accompanying drawings, wherein: FIGS. 1,2 and 3 are geometrical plan and transformed views of a grid of centerpoints produced from (i) an array of seismic sources and detectors and (ii) seismic processing whereby a series of locational traces associated with individual centerpoints between respective source-detector pairs can be associated together in a meaningful way; FIG. 4 is a model oftypical reflecting horizons within an earth formation that can be associated with the characteristics of the locational traces of FIGS. 1 and3; FIGS.Sand 6are plots of reflection coefficientasa function of angle of incidence of seismic waves associated with the reflecting horizons of FIG. 4; FIGS. 7 and 8 are flow diagrams of processes akin to those shown in FIGS. 2 and 3 for carrying outthe method ofthe present invention, using a programmed digital computing system; FIGS. 9A and 9B are schematic diagrams illustrating certain steps oftheflow diagrams of FIGS. 7 and 8 in more detail; FIG. 10 is a schematic diagram of elements within a typical digital computing system; and FIGS. 11-16 are true record sections and portions of sections, illustrating the diagnostic capability of the method of the present invention.

Before discussion of an embodiment ofthe invention within an actual field environment, it may be of interest to indicate lithology limitations associated with the persent invention. For example, anomalies associated with gassands overshale cap rock are one example in which the method of the present invention offers surprising results; another relates to gas-saturated limestone over shale. Also of import ance is the relationship between Poisson's ratio and resulting high-intensity amplitude anomalies provided on seismic traces.

While Poisson's ratio (a) has the general formula

where Vp is compressional velocity andV5 is shear velocity of the medium, this concept is not without physical significance. For example, considering a slender cylindrical rod of an elastic material and applying a compressional forceto the ends, as the rod changes shape (the length of the rod decreasing: by AL, whilethe radius increasing byAR), Poisson's ratio is defined as the ratio ofthe relative change in radius (ARIR) to the relative change in length (#L/L).

Hence compressible materials have low Poisson's ratios, while incompressible materials (as a liquid) have high Poisson's ratios.

Equation (A) above indicates the relationship ofthe compressional and shearwave velocities of the material, Vp and Vs respectively; i.e., that Poisson's ratio may be determined dynamically by measuring the P-wave and S-wave velocities. Only two ofthe three variables are independent, however.

Recent studies on reflection and transmission seismic waves useful in geophysical applications include: (1) Koefoed, 0., 1955,for "On the Effect of Poisson's Ratios of Rock Strata in the Reflection Coefficients of Plane Waves", Geophysical Prospecting, Vol.3, No.4.

(2) Koefoed, O., 1962, for"Reflection and Transmission CoefficientsforPlane Longitudinal Incident Waves", Geophysical Prospecting, Vol.10, No.3.

(3) Muskat, M. and Meres, M.W., 1940, for "Reflection and Transmission Coefficients for Plane Waves in Elastic Media", Geophysics,Vol. 5,No.2.

(4) Tooley, R. D., Spencer, T. W. and Sagoci H. F., for"Reflection and Transmission of Plane Compressional Waves", Geophysics, Vol.30, No.4(1965).

(5) Costain, J. K., Cook, K. L. and Algermisshi, S.

T.,for "Amplitude,- Energy and Phase Angles of Plane SP Waves and TheirApplication to Earth Crustal Studies", Bull. Seis. Soc. Am.,Vol. 53p. 1639 et seq, All of the above havefocused on the complex modeling of variation in reflection and transmission coefflcientsasafunction of angle of incidence.

The problem is complicated, however. E.g., isotropic media with layer index ofthe strata, i=1 forthe incident mediurn and i=2 forthe underlying medium, have been modeled using equationsfdr P-wave reflection coefficient Apr and for P-wave transmission displacement amplitude coefficientApt. For each of the media, i.e., the incidentorunderlying medium, three independent variables exist: P-wave velocity, Q and bulk density, or a total of sixvariablesfor both media.

Butto provideforthe many combinations of possible variations, the above-listed studies have either: (a) generated many (literallythousands) plots of a mathematical natureforvarious parameters, values in which there was little relationship with true geophysical applications, since the latter were hopelessly obscured and unappreciated; or (b) made simplistic assumptions that, although using actual calculations, nevertheless did notex- press there nature of transmission and reflection coefficients, in particular lithological situations associatedwith the accumulation of gaseous hydro carbons within an actual earth formation.

While reference (2) concludes that change in Poisson's ratio in the two bounding media can cause change in the reflection coefficient as a function of angle of incidence, reference (2) does not relate that occurrence to lithology associated with the accumulation of gaseous hydrocarbons in the surprising manner ofthe present invention.

In the above-identified patent applications, it is taughtthat gas-containing strata have low Poisson's ratios and that the contrast with the overburden rock as a function of horizontal offset produces a surprising result: such contrast provides for a significant- and progressive-change in P-wave reflection coefficient at the interface of interest as a function of angle of incidence ofthe incidentwave. Thus, determining both the gas-bearing potential and lithology of host media is simplified by relating progressive change in amplitude intensity as a function of offset between source-detector pairs, i.e., angle of incidence being directly related to offset.

However, there is still a need in some cases to further emphasize the degree of amplitude change as a function of offset especially with respective sourcedetector pairs associated with near and far offset locations of a gather of traces.

By the terms "near" and "far" offset locations, it is meant that such are measured with respect to the source locations associated with the original source points where the seismic waves were generated.

Hence, they represent the degree of horizontal offset distance that exists between such field locations, the source and receiver as the data was collected in the field.

Now in more detail, attention should be directed to the Figures, particularly FIG. 1. Note that, inter alia, FIG. 1 illustrates in some detail how the terms of interest in this application are derived: e.g., the term "centerpoint" is a geographical location located midway between a series of sources S1,S2... Sn of a geophysical field system 9 and a set of detectors D1,D2.. . . .Dm ata datum horizon near the earth's surface. The centerpoints are designated C1 C"C2. . Cp in the Figure, and are associated with a trace derived by placement of a source atthatcenterpoint location followed immediately by relocating a detector thereat.

I.e., if the sources S1. . .Sn are excited in sequence at the source locations indicated, traces received at the different detector locations shown can be related to common centerpoints therebetween, and a gather or group oftraces is formed. I.e., if the reflecting interface isaflat horizon, the depth pointwhere a reflection occurs will define a vertical line which passes through the centerpoint of interest. Applying static and dynamic corrections to the field traces is equivalent (underthe above facts) to placing the individual sources S1,S2...Sn at the centerpoint in sequence followed by replacement with the detectors D1... Dm of interest at the same locations.If the traces associated with a common centerpointaresummed, a series of enhanced traces, sometimes called CDPS (Common Depth Point Stack) traces, is provided. But before such traces are summed, such display cap be enhanced to surprisingly indicate the presence of fluid hydrocarbons in a host strata as well as the lithology ofthe latter.

FIG. 2 illustrates reflection phenomena of a threelayer model typical of a young, shallow geologic section 10, such as found in the Gulf Coast, illustrating how reflection phenomena associated with the traces associated with the field system 9 of FIG. 1 can be related to the presence of gas.

Section 10 includes a gas sand 11 embedded in a shale stratum 12. Assume a Poisson's ratio of 0.1 for the gas sand and of 0.4 forthe shale, a 20% velocity reduction at interface 13, say from 1 0,000'/sec to 8000'/sec, and a 10% density reduction from 2.40 g/cc to 2.16 g/cc.

The actual P-wave reflection coefficientAprcan be related to section 10 by Equation (1) below; also, P-wave transmission displacement amplitude coefficient Apt can similarly be related in accordance with Equation (2) below.

E . k21k22 (a1c2 + a2C1) (3) t = b2n2 # a1c1v2 (4) x.

= a2c2 (@2 # 4a1c1@2b2 (5) x n . 6c2-rl (6) C = 6=1 (7) V = 6C2 + 2b2 (8) X = C1 + 26b2 (9) @i = = k2i - 2b2 (10) 5 = 2/ 1 (11) Pi = Pivsi (12) b = h1sin# (13) ai2 = hi2 - b2 (14) C2j = kl - b2 (15) hi = 1/vpi (16) ki = 1/V5i (17) Vp p-wave velocity Vsj s-wave velocity P1 density layer index e angle of incidence Equations (1) and (2) are, of course, the two basic equations of wave travel in an earth formation and are for isotropic media with the layer index being i=1 forthe incident medium and i=2forthe underlying medium. Equations (3) through (17) simply define intermediate variables.

As an example of calculations associated therewith, if @=0 (normal incidence), the P-wave reflection coefficient Apr is equal to about -0.16 and +0.16, respectively.

FIG. 3 illustrates change in reflection coefficient as a function of angle of incidence # for the three-layer model of FIG. 2.

Note that solid lines 20, 21 illustrate the effects of reflection (and transmission, by omission) on the top and base ofthe gas sand. In line 20, at #=0 , note that the Apr equals -0.16; while at 9=400, the Apr is about -0.28. That is, rather a surprisingly large change in the reflection coefficient as a function of angle of incidence occurs, with the greatest change occuring between 9=200 and û=40 .

Forthe bottom layer, line 21 changes at about the same rate, but in opposite sign. I.e., at 0=0 , Apr is about+0.1 6 and at #=40 , Apr is about +0.26. Again, the greatest change in Apr occurs between 0=20 and #=40 . As a result, the amplitude ofthe seismic wave reflected from this model would increase about 70% overthe angle of incidence range shown, i.e., overthe incremental 40 degrees shown.

While angles of incidence equal to 40 may seem a little large for reflection profiling (heretofore, most data arriving beyond 30 being thought useless and muted out), experience has now nevertheless shown that reflection data can and do arrive at reflection angles greaterthan 30 . Hence, the angles of incidence must be determineå, ana the straight-ray approach to estimate such angles of incidence (using depth-to-reflector and shot-to-aetector and shot-to group offset), is useful.

FIG. 4 illustrates another plotassociated with a three-layer model akin to that shown in FIG. 4, but in which the sandstone contains gas but is buried deep below the earth's surface. The values for the three layer model of FIG. 2 are again used except that the velocity change from shale to sand is only 10%, or from 10,000'/sec to 9000'/sec. As shown, curves 25,26 are even more significant: both curves are seen to increase in magnitudeform overthe 40" of change in the angle of incidence. However, field results have not verified these results, prior to the inventions described in the above applications, since Poisson's ratio in such gas sands may be strongly affected by depth.

FIG. 5 is a diagram which illustrates a data "addressing" technique as practiced bythe present invention; in the Figure, the traces are generated using an end-shooting array of 48 detectors with source and detectors advancing one detector interval per shot point. Resuit: a 24-fold CDP-stacked record section. Note further each centerpoint is associated with 24 separate traces to varying offset.

Tofurther understand the nature of FIG. S, assume that the sources S1,S2...Sn are sequentially located at shotpoints SP1,SP2 . . . SPn at the top of the Figure.

Assume also that the detectors are placed in line with the sources, i.e., along the same line of survey atthe detector locations D1'D2...Dm.After each source is activated, reflections are received at the detectors, at the locations shown. Then by the "rollalong" techni que,the source and detectorspreads can be moved in the direction B of survey line A and the process repeated to provide a series of traces. The latter are associated with centerpoints midway between the respective detector-source pairs. In the Figure, assume source S1 has been located atshotpoint SP1 and excited. Midway between SP1 and each ofthe detectors, a D1,D2...Dm is a series of centerpoints C1,C2... , , , Cn. The latter ace each associted with a trace.

In this regard and for a further description of such techniques, see U.S. Patent 3,597,727 for "Method of Attenuating Multiple Seismic Signals in the Deter mination of Inline and Cross-Dips Employing Cross Steered Seismic Data", Judson etal, issuedAugust 3, 1971, and assigned to the assignee ofthe present application. With appropriate static and dynamic corrections, the data can be related to the common centerpoints midway between individual source points and detectors, as discussed in the above-noted reference.

But by such afield technique, data provided generate 24 separate traces associated with the same centerpoint C1... Cn. To correctly index ("address") these traces as a function of several factors including horizontal offset and centerpoint location, involves the use ofstacking chart 44.

FIG. 6 illustrates stacking chart44 in detail. As shown, Chart 44 is a diagram in which thetraces are associated with either a plurality of oblique common profile lines PL1,PL2. . . or a series oofcommon offset and centerpoint locations at 90 degrees to each other.

For best illustration, focus on a single shotpoint, say SPa, and on a single detector spread having detectors D1,D2.. . . .Dm of FIG. 6 along survey line A. Assume a source is located at shotpoint SP1 and activated thereafter. The detector spread and souce are "rol- led"forward along survey lineAin the direction B, being advanced one station per activation. Then after detection has occurred, and if the resulting center point pattern is rotated 450 about angle 46 to profile line PL1 and projected below the spread as in FIG. 6 as a function of common offsetvalues and centerpoint positions, the chart44 of FIG. 6 results.Of course, each centerpoint has an amplitude vs. time trace associated therewith, and for didactic purposes that trace can be said to project along a line normal to the plane ofthe Figure.

It should be emphasized that the centerpoints provided in FIGS. SandS are geographically located along the line ofsurveyA in line with the source points SP1,SP2...As the locational traces are generated, the chart 44 aids in keeping a "tag" on each resulting trace. As the detector spread an sources are rolled forward one station and the technique repeated, another series of traces is generated associated with centerpoints on new profile line PL2.That is, although the centerpoints are geographically still associated within positions along the survey line A of FIG. 5, by rotation along the angle 46, the new centerpoint pattern C1',C2'.. .Cn' can be horizontally and vertically aligned with centerpoints previously generated. I.e., at common offset values (in horizontal alignment) certain centerpoints are aligned, viz, centerpoint C1 aligns with C1' as shown; further C2 is aligned with C2,, etc. Also, there are traces that have common centerpoints. I.e., at common centerpoints (in vertical alignment) centerpoint C2 aligns with centerpoint C1', and centerpoints C3,C2' and C1" are similarly aligned. Thus, via chart 44, each trace associated with a centerpoint can be easily "addressed" as to: (i) its actual geographical location (i.e., along phantom lines normal to diagonal profile lines PL1,PL2. . - along common location lines LL1,LL2. .

so that its actual field location is likewise easily known; (ii) its association with other traces along common horizontal offset lines COL1,COL2. . .COLx; and (iii) its association with still other traces along common vertical centerpoint location lines CPL1,CPL2 Also, "addressing" the traces allows such traces to be easily enhanced as by using amplitude projection (ofthe trace gather) to new "near" and "far" offset locations, as in the manner of FIGS. 7 and 8. Briefly, as shown in FIGS. 7 and 8, by using onlythe amplitudes variation between the near and far offsets, the interpreter can more readily determine valuable characteristics, such as the fluid hydrocarbon-bearing potential andlithologyofthehostmedia,while simplifying the number of parameters required to render a viable display of such data.

Now, in more detail, FIGS.7 and 8 are flow diagrams illustrative of a computer-dominated process in which the functions required by the method of the present invention can be easily ascertained.

Preliminaryto the steps shown in FIG. 7, assumethat a section of seismic data has been analyzed for "bright spots"; such events are known by geographical location and/or a time/depth basis; and the traces have been dynamically and staticallycorrected, as hereinbefore described.

The steps of FIG. 7 include generating addresses for the data that include a common offset address in the manner of FIG. 2, common centerpointaddress and an actual geographical location address also in the manner of FIG. 2. Next, near and fartraces (and ultimately sections thereof) are generated based on an analytic relationship thatfirst best approximates the actual variation in amplitude versus offset within each gather oftraces for a series of time samples and then determine the projected amplitudes of the near and far traces based on the functional form ofthe best fitting curve. Finally, the generated near and far sections are displayed side-by-sidewherebythe character ofthe amplitude event of interest is indicated as a function of changing centerpoint values.If the event character abruptly changes from the neartofarsections (normalized to common centerpointvalues) then there is a high likelihood that the event is indicative of strata containing hydrocarbons. Also the lithology ofthe host strata is easily determined based on the assumption and operations described in detail in the above-identified applications as well as in less detail below.

In more detail, after the addresses have been generated, amplitudes of side-by-side traces of each gathercan be re-indexed asafunction oftimeand offset. That is to say, for a time sample, say time sample T1, ofthetrace gather G1, all amplitudesfor that sample are first re-indexed as a function of offset (if not already so ordered, see FIG. 9A). Next, the generated amplitude vs. offset data are compared to a series of analytic functions and the "best fit" determined based on a least-square analysis. That is to say, the linear or quadratic equation that bestfits the data isthe one in which the sum ofthe squares of the distances (associated with the amplitudes of the trace gather) is minimum.

In orderto simplifythe step of bestfittingtheform of the actual data to a specified linear or quadratic equation, usuallythree mathematical functions are specified: a linear equation on the form A (x) = CO+ C1x; and quadratic equations oftheform A (x) = CO + C2x2andA(x) = CO + Crx= C2x2whereA(x) isthe variation in amplitude of the data as a function of offset and C,, C1 and C2 are constants determined by standard pseudo inverse matrix methods conventionally available in the seismic processing industry.

After a bestfit of the data associated with the time sampleT1 has been obtained,the near and far offset amplitude values are next generated based on the functional form ofthe bestfitting equation, see Fig.

9B.The data is then stored. Then after the near and far offset amplitudes associated with the remaining time samples T2.. .Tj ofthe gather G1 have been generated, the process is repeated for neighboring gathers G2. . . Gj.

For generating near and far offset traces (and ultimately sections thereof), new offset values to which the field data is projected (which of course are outside ofthe set of offsets associated with the traces of each original gather) must be chosen. In the case of the near offset location, the choice is a conventional one-zero (FIG. 9B). That isto say,forthe linear or quadratic equations setforth above is set equal to zero and the amplitude solutions as afunction of differenttimesamplesTi . . .Tj, determined.While in the process of generating near trace data, the choice can be said to be somewhat obvious [i.e., setting Xto zero in either equation (i), (ii), or (iii)], not so in the case offar offset trace and section determination.

For generating such far offset traces and sections, the offset values chosen must not only be constant and outside the set of usual offsetvaluesofthe common gather (as in neartrace processing), but also they must be values customarily acceptable to those skilled in the art. In this regard, mute offsets as used in conventional seismic processing, have been found to be adequate. Also offset locations associated with common emergence angles of adequatefrequent content, are likewise useful.In thatway, the offset values chosen forfar projection purposes are those values either (i) that are conventionally associated with the process of excluding from the early parts of the offset traces, signals dominated by refraction energy, or (ii) that are associated with emergence angles such that a long offset traces associated therewith has a frequency content that is not appreci- ably lowerthan those of neighboring tracers.

However, the space coordinates ofthe final traces is not an offset coordinate but in fact is a centerpoint location that is common to the common gatherfrom which the nearand fartraceswere produced. Hence the resulting plots easily correlate with actual field addresses.

In carrying outthe above processes on a highspeed basis, a fully programmed digital computer can be useful, and moreover is the best modefor carrying out the present invention. But electromecha nical systems well known in the art can also be used.

In either case, the field traces must first undergo static and dynamic correction before the traces can be displayed as a function of offset to determine their potential as a gas reservoir. Such correction techni ques arewell known in the art-see,e.g.,U.S. Patent 2,838,743, of O.A. Fredriksson etal,for"Normal Moveout Correction with Common Drive for Record ing Medium and Recorder and/or Reproducing Means", in which a mechanical device and method are depicted.Modern processing today uses properly programmed digital computersforthattask in which the data words are indexed as a function of, inter alia, amplitude, time, datum height, geographical loca tion, group offset, velocity, and are manipulated to correct for the angular and horizontal offset; in this latter environment, see U.S. 3,731,269, Judson et al, issued May 1,1973, for "Static Corrections for Seismic Traces by Cross-Correlation Method", a computer-implemented program of the above type.

Electromechanical sorting and stacking equipment is also well known in the art and is of the oldest ways of canceling noise. See,forexample,thefollowing patents which contain sorting and stacking techni ques, including beam steering techniques: Patent Issued Inventor Title 3,597,727 12/30/68 Judson et al Method of Attenuating Multiple Seismic Signals in the Determination of Inline and SDips Employing Cross-Steered Seismic Data 3,806,863 4/23/74 Tilley et al Metbod of Collecting Seismic Data of Strata Underlying Bodies of Water 3,638,178 1/25/72 Stephenson Method of Processing Three Dimensional Seismic Data to Select and plot Said Data on a Two-Dimensional Display Surface 3,346,840 10/10/67 Lara Double Sonogramming for Seismic Record Improvement 3,766,519 10/16/73 Stephenson Metbod for Processing Surface Detected Seismic Data to Plotted Representations of Subsurface Directional Seinnin Data 3,784,967 1/8/74 Graul Seismic Record Processing Method 3,149,302 9/15/74 Klein et al Information Selection Programmer Employing Relative Amplitude, Absolute Amplitude and Time Coherence 3,149,303 9/15/64 glein et al Seismic Cross-Section Plotter FIG. 10 illustrates particular elements of a comput ing system for carrying outthe steps of FIGS. 7,8, 9A and 9B. While many computing systems are available to carry outthe process ofthe invention, perhaps to best illustrate operations at the lowest cost per instruction,a microcomputing system 50 is didacti cally best and is presented in detail below.The system 50 of FIG. 10 can be implemented on hardware provided by many different manufacturers, and for this purpose, elements provided by Intel Corporation, Santa Clara, California, may be prefer red. However, where a central center for seismic data processing is available, a large main-frame comput ing system (such as an IBM 370/65) is usually already in place; and thus for most applications involving the present invention, such a system becomes the best mode for carrying it out.

System 50 can include a CPU 51 controlled by a control unit52. Two memory units 53 and 54connect to the CPU 51 through BUS 55. Program memory unit 53 stores instructions for directing the activities of the CPU 51 while data memory unit 54 contains data (as data words) related to the seismic data provided by the field acquisition system. Since the seismic traces contain large amounts of bit data, an auxiliary memory unit55 can be prnvided.TheCPU 51 can rapidly access data stored through addressing the particular input port, say at 56 in the Figure.

Additional input ports can also be providedto receive additional information as required from usuatexter- nal equipment well known in the art,e.g.,floppy disks, paper-tape readers, etc., including such equip- ment interfaced through input interface port 57 tied to a keyboard unit 58 for such devices. Using clock inputs, control circuitry 52 maintains the proper sequence of events required for any processing task.

After an instruction is fetched and decoded, the control circuitry issues the appropriate signals (to units both internal and external) for initiating the proper processing action as setforth above.

In addition to providing both mathematical projec tionsofthetracedata of each original gatherand displays of such projections on a side-by-side basis, the system 50 can also allowforthetesting of the contents ofthe projections against certain known trends in the original data to better pinpointthe, say fluid hydrocarbon-bearing potential and/or lithology ofthe surveyed earth formation. Such decisions relate to certain relationships inherently involved in the data.

Note that prior related patent applications, op. cit., teach that zones of gaseous hydrocarbon accumulation can be accurately identified by determining iffirst high-intensity event exist in the trace gathers of interest and then next if the events can be associated with the presence of gaseous hydrocarbons, viz., answering the question, "Does the amplitude of such events change progressively as a function of horizontal offset?", in the affirmative. Such a conclusion involves a precursorstep in which the events of interest (from one gather with same event in another gather) are contrasted with each other.And if there appears to be a detectable change in the amplitude characterofthe event of interest, say a reversal in UPor DOWN-scaletrend, then the conclusion that such change was brought about bythe presence of gaseous hydrocarbons has a high probability of being true. And after interrogation via a look-up table, the lithologic character ofthe underlying strata is also capable ofdetermination.

Such decisions and the results of those decisions are automatically controlled by the system 50. After picking and codifying the amplitudes ofthe event(s) of interest, i.e., projecting near and far amplitudes via a least-squares fit, the system 50 also can automatically determine their UP-or DOWN-scale trend; determine if the trend is a reversal of priorcalculted data, and depending on whether or nota reversal is found (assumethat it has been), highlightthe reversal; and then compare its direction with singlevariable lithologytable so asto indicate both the gas-bearing potential and the lithology of the strata.

With regard to the operation of the latter table, it comprises a LOOK-UPfunction in which the UP- or DOWN-scale trend ofthe amplitude direction (with offset) of the individual gathertriggersthe printing of an appropriate lithologictag.

For outputting information, the system 50 can include a printer unit 59 by which say, the results of the lithology determination steps (of the interrogation ofthe lithology LOOK-UP table) are printable.

Of more use as an output unit, however, is disk unit 60, which can temporarily store the data. Thereafter, and eff-line digital plotter capable of generating a series of displays is used in conjunction with the data on the disk unit 60. Such plotters are available in the art, and one proprietary model that lam familiar with uses a computer-control led CRTfor optically merging onto photographic paper, as a display mechanism, the seismic data. Briefly, in such a plotterthe data are converted to CRT deflection signals; the resulting beam is drawn on the face ofthe CRT and the optically merged record of the event indicated, say via photographic film.After a predetermined number of side-by-side lines have been drawn, the film is processed in a photography laboratory and hard copies returned to the interpreters of their review.

EXAMPLES Diagnostic capability provided by the method of the present invention is better illustrated in the Examplessetforth below Example I Seismic data were obtained in the Sacramento Valley, California. These data, in CDP-stacked form are shown in FIG.11. Three discovery wells (located at about CP-109, CP-98, and CP-85) penetrates a 100-foot sand which is almost fully gas-saturated.

The developed portion of the field extends from about CP-75 to CP-130. Gas occurs at a depth of about 7,000 feet which corresponds to a time of about 1.7 seconds on the plots.

The near and far projected trace sections 80,81 are shown in FIGS. 12 and 13. Note by comparing the sections that the amplitudes over the regions and depths of interest increase with offset within the plots, and moreover correlate well with the gas find of interest.

Example II Seismic data were also obtained from Alaska, and are depicted in CDP-gathered format in FIG. 14.

Discovery wells a re Iocated at a bout CP-140 and CP-110 and penetrate a series of stringers containing gaseous hydrocarbons.

The field extends from about CP-1 OS to CP-150. Gas occurs at a depth of about 3,500 feet which corresponds to a time of about 0.9 seconds on the plot Near and far projected trace sections 84 and 85 are shown in FIGS. 15 and 16. A comparison of the sections shows thatthe amplitudes overthe regions and depths of interest increase with offset within the plots, and moreover correlates well with the gasfind of interest.

It should thus be understood that the invention is not limited to any specific embodiments setforth herein, as variations are readily apparent.

For example, envelopes of the amplitudes of each ofthe near and fartraces can be generated using conventional averaging processing (say, using the root mean square of many amplitude values associated with the zero and 90 wavelets of the stored data, i.e., using several amplitude samples over several time samples first at zero phase and then shifted in phase 90 ), followed bythe subtraction of the generated envelopes (one from the other) as a function of common centerpoint location. The resulting difference envelope is more reliable since any one segment is an average of several amplitude values over manytime samples. Hence, noisewithinthe original section tends to be suppressed.

In addition, note that the prior-mentioned prediction techniques for determining near and far trace projections, can also utilize matrix equations ofthe following form.

PSEUDO INVERSE MATRIX EQUATIONS Assu me that these N traces per gather with offsets Xj and amplitudes Aj available for processing and that the analytic functions or of the form described above, viz: A(j) = C0 + C1x (B) A(j) = C0 + C2x2 (C) A(j) = C0 + C1x + C2x2 (D) For equation (C) least squares solution in matrix form for constants C0 and C2 is: C0 N #Xi2 -1 #Ai = C2 #Xi2 #Xi4 #AiXi2 Letting: DET=N.#Xi4-(#Xi2)2 Then the evaluation ofthe constants C0 and C2 are further simplified.

#X4 #X2 C0 = # #Ai - # #AiXi2 DET DET #X2 N C2 = # #Ai - # #AiXi2 DET DET Asthe amplitudes ofthe near and fartraces are projected, such processing involves substitution of a finite offset value, i.e., offset value Xp, in the above equations, as follows: Projection at offset Xp: A(Xp) = C0 + C2Xp2 = zx4-xp#x . zAi OET = NXp-#x . #Ai xi = ZWiAi where: #X4 - Xp2 #X2 NXp2 - #X2 Wi = + # Xi2 DET DET and: Wi=K0+K2.Xi2 Moreover, since the constants Ko and K2 are easily evaluated, equation (C) can be rapidly solved. In like manner, equations (B) and (D) are also solvable.For equation (B), e.g., application ofthe above matrix form form provides for a projection at Xp of the form: A (Xp) = SWj Aj where: Letting DET = N zX2- (#X)2 #X2 - Xp #X NXp - #X Wi = + # Xi DET DET or Wj= Ko+ K1 Xi Note that in the above form, that if the projected offset location is about equal to 2/3 ofthe maximum offset value, then Xp = # X2/EX = 213 xmas; Ko = 0; and Wj= KlXj Likewise for equation (D), application of the above matrixform providesfora projection of Xp of the form A(Xp) = #WiA where: Wi = K0 + K1Xi + K2Xi2 and K's = #(#Xi,#Xi2,#Xi3,#Xi4,Xp,N) Thus, the invention is to be given the broadest possible interpretation within the terms of the following claims.

Claims (27)

1. A method for determining valuable characteristics of strata in the earth using high-intensity amplitude events in seismic records, comprising the steps of: (a) generating seismic data, including a record of signals from acoustic discontinuities associated with said strata of interest by positioning and employing an array of sources and detectors such that centerpoints between selected pairs of sources and detec torsform a series of centerpoints along a line of survey, said recorded signals being the output of said detectors; (b) by means of automated processing means, statically and dynamically correcting said recorded signals to form corrected traces whereby each of said corrected traces is associated with a centerpoint horizontally midway between a source-detector pair from which said each corrected trace was originally derived;; (c) by means of automated processing means, indexing said corrected traces so that each of said corrected traces is identified in its relationship to neighboring traces on the basis of progressive changes in common centerpoint location; (d) determining among aseriesofanalyticfunc- tions of known mathematical character, a best fit to amplitude vs. horizontal offsetvariations of a gather of said corrected traces, said gather of traces being identified with a common centerpoint location and a set of progressively changing horizontal offset values;; (e) predicting near and far amplitude vs. time trace projections for said gather of corrected traces at new offset locations based on said best fitting analyticfunction, said predicted nearandfaroffset trace projections being identified with offset loca tionsfalling on opposite sides of said set of changing horizontal offsetvalues;; (f) displaying a first series ofsaidtrace projections of step (e) associatedwith near offset locations, side-by-side with a second series oftrace projections also of step (e) associated with far offset locations, said first and second series of displayed traces all being associated with at least the same general common groups of centerpoints so that progressive change in a high-intensity amplitude event in said displayed traces is identified as a function ofprogres- sivechange in centerpointvalues.

2. The method of Claim 1 in which said one of said series of analytic function of known mathematical character of step (e) is selected from a group of linear and quad ratic equations ofthe form: A(x) = C0 + C1 x; A(x) = C0 + C2x2; and A(x)=C0+C1 +C2x2 where A(x) is the amplitude of the projected trace as a function of offsetx; and C0, C1 and C2 are constants determined by conventional seismic processing steps.

3. The method of Claim 2 in which selection of said one analytic function of known mathematical characterofstep (d) is based on a least squares fit, of said best fitting function to said amplitude vs. offset variations of said gather of corrected traces.

4. The method of Claim 2 further characterized in that step (e) of predicting near and far amplitude vs.

offsettrace projections for each of said gathers of corrected traces is determined by solving said best fitting analyticfunction for preselected near and far offset values, using constants determined by conventional processing steps.

5. The method of Claim 4 in which each of said near amplitude vs. offsettrace projections of step (e) is determined by solving said one selected analytical function for a near offset location X = 0 and using constants determined by conventional processing steps.

6. The method of Claim 4 in which each of said far amplitudevs. offsettrace projections of step (e) is determined by solving said one selected analytical function forthe far offset location x = the mute offset location for CDP processing of said corrected traces.

7. - The method of Claim 4 in which each of said far amplitude vs. offsettrace projections of step (e) is determined by solving said bestfitting analytical function for the far offset location x = the offset location usedforCDP processing of said corrected traces so as to provide for an emergence angle that produces minimum acceptable trace distortion.

8. The method of Claim 7 in which said emerg- ence angle of minimum acceptabletrace distortion is between 40-50 degrees measured from a vertical normal to a horizontal rejecting horizon.

9. The method of Claim 1 with the additional step of: (g) determining the lithologic character ofthe strata based on the direction of the progressive change in the amplitude event between said first and second series of trace projections.

10. The method of Claim 9 in which step (g) is further characterized by the substeps of (a) observing thatthe amplitude event of interest increases from said first series of traces to said second series, and (b) concluding thatthe lithologic character of the strata is more likely than nota sandstone underlying an impervious shale.

11. The method of Claims 9 in which step (g) is further characterized by the substeps of (a) observing thatthe amplitude event of interest decreases from said first series oftraces to said second series, and (b) concludingthatthe lithologiccharacterofthe strata is more likely than nota limestone underlying an impervious shale.

12. A method for converting an original multitrace seismic record into an improved section having increased capability asto fluid hydrocarbon-bearing potential and/or lithologic nature of high-intensity amplitude events related to reflections from subsurface strata, said improved section being composed ol a plurality of amplitude -versus - centerpoint - and time traces, said original record consisting of a plurality of multitrace seismic traces of amplitude versus - horizontal coordinate - and -time, each of said traces constituting energy derived in association with a particular source-detector pair of known horizontal offset and of known centerpoint location, and representing, in part, event reflections from said subsurface strata, said conversion comprising the steps of:: (a) classifying said original traces on the basis of common but progressively changing horizontal offset values and common but progressively changing common centerpoint locations, whereby each resulting trace is identified by a centerpoint location common to at least anothertrace and a known horizontal offset value; (b) determining among a series of analyticfunctions of known mathematical character, a bestfitto amplitude vs. offset variations of said each resulting trace and said at leastanothertrace;; (c) predicting near and far amplitude vs. offset trace projectionsforsaid resulting trace and another trace at new offset locations based on said best fitting analytic function, said predicted near and far offset trace projections being identified with offset loca tions falling on opposite sides of said set of changing horizontal offset values.

(d) displaying a first series of said trace projections of step (c) associated with near offset locations, side-by-side with a second series of trace projections also of step (c) associated with far offset locations, to form at least a segment of said improved section; said first and second series of displayed traces all being associated with at least the same general common group of centrepoints so that progressive change in a high-intensity amplitude event in said displayed traces is identified as a function of progressivechange in centerpointvalues.

13. The method of Claim 12 in which said best fitting analyticfunction of known mathematical character of step (b) is selected from a group of linear and quadratic equations of the form: A(x) = CO + C1 x; A(x)= C0 + C2x2; and A(x)=CO+Cax+C2x2 whereA(x) is the amplitude ofthe projected trace as a functionofoffsetx; and CO, C1 andC2areconstants determined by conventional seismic processing steps.

14. The method of Claim 13 in which selection of said best fitting analytic function of known mathematical character of step (b) is based on a least squares fit, of said function to said amplitude vs. offset variations of said gather of corrected traces.

15. The method of Claim 13further characterized in the step (c) of predicting near and far amplitude vs.

offset trace projections for each of said resulting trace and said anothertrace is determined by solving said best fitting analyticfu nction for preselected near and far offset values, using constants determined by conventional processing steps.

16. The method of Claim 15 in which each of said near amplitude vs. offset trace projections of step (c) is determined by solving said bestfitting analytical function for a near offset location X = 0 and using constants determined by conventional processing steps.

17. The method of Claim 15 in which each of said far amplitude vs. offset trace projections of step (c) is determined by solving said bestfitting analytical function forthefar offset location x = the mute offset location for CDP processing of said corrected traces.

18. The method of Claim 15 in which each of said far amplitude vs. offset trace projections of step (c) is determined by solving said bestfitting analytical function for the far offset location x = the offset location used for CDP processing of said corrected traces so as to providefor an emergence anglethat produces minimun acceptable trace distortion.

19. The method of Claim 18 in which said emergence angle of minimum trace distortion is betweeen 40-50 degrees measured from a vertical normal to a horizontal reflecting horizon.

20. The method of Claim 12 with the additional step of (e) determining the lithologic character ofthe strata based on the direction of progressive change in the amplitude eventcommon to saidtraces.

21. Amethod for determining hydrocarbon-bearing potential and/or lithology of strata in the earth using high-intensity amplitude events in seismic records, comprising the steps of: (a) generating seismic data, including a record of signals from acoustic discontinuities associated with said strata of interest by positioning and employing an array of sources and detectors such that centerpoints between selected pairs of sources and detectorsform a series of centerpoints along a line of survey, said recorded signals beingtheoutputofsaid detectors; ; (b) by means of automated processing means, statically and dynamically correcting said recorded signals to form corrected traces whereby each of said corrected traces is associated with a centerpoint horizontally midway between a source-detector pair from which said each corrected trace was originally derived; (c) by means of automated processing means, indexing said corrected traces so that each of said corrected traces is identified in its relationship to neighboring traces on the basis of progressive changes in common centerpoint location; (d) determining from among a series of analytic functions of known mathematical character, a best fit to amplitude vs. horizontal offset variations of a gather of said corrected traces, said gather of traces being identified with a common centerpoint location and a set of progressively changing horizontal offset values;; (e) predicting near and far amplitude vs. time trace projections for said gather of corrected traces at new offset locations based on said best fitting analyticfunction, said predicted near and far offset trace projections being identified with offset locationsfalling on opposite sides of said setof changing horizontal offset values; (f) generating a first envelope of said trace amplitude projections of step (e) associated with near offset locations and a second amplitude envelope of said trace projections of step (e) associated with far offset locations and subtracting the two envelopes onefrom the otherto form a difference envelope.

(g) displaying said difference envelope of step (f) so asto depict amplitude vs. time change as a function of centerpoint coordinate so that progressive change in a high-intensity amplitude event in said displayed traces is identified as a function of progressive change in centerpointvalues.

22. Method of Claim 21 in which step (e) of predicting near and fartrace projections is in accordance with solution of the general equation A(x) = Alin matrixformat,where: Wj is a function thatvaries with the selected linear or quadratic equation that best fits with the said amplitude vs. time variation of the trace gathers, and Aj is the amplitude ofthe gathers for the time samples T1...Th

23. Method of Claim 22 where step (e) of predicting projection amplitudes provides that such projections be for an offset location at Xp and wherein Win the general equation A(x) = WiAi is equal to: X4 Xp2 X2 + NXO2 - X2 . Xi2 OET 1

24.A methof for converting an original multitrace seismic record into an improved section having increased capability as to hydrocarbon-bearing potential and/or lithologic nature of high-intensity amplitude events related to reflections from subsurface strata containing the hydrocarbons, said improved section being composed of a plurality of amplitude-versus-centerpoint-and-time traces, said original record consisting of a plurality of multitrace-seismic traces of amplitude - versus horizontal coordinate- and -time, each of said traces constituting energy derived in association with a particular source-detector pair of known horizontal offset and of known centerpoint location, and representing, in part, event reflections from said subsurface strata, said conversion comprising the steps of:: (a) classifying said original traces on the basis of common but progressively changing horizontal offset values and common but progressively changing common centerpoint locations, whereby each resulting trace is identified by a centerpoint location common to atleastanothertrnceand a known horizontal offset value; (b) determining with regard a series ofanalytic functions of known mathematical character,that one of said series is the best fit to amplitudvs. time variationsofsaid each resulting and said least anothertrace;; (c) predicting near and far arnplitudevs. time trace projections for said resulting and arrothertraces at new offset locations based on said one, best fitting analytic function, said predicted near and faroffset trace projections being identified with offsetlocationsfalling on opposite sides of said set of changing horizontal offsetvalues; (d) generating afirstseries of trace amplitude projections vs. time of step (c) associated with near offset locations, and a second series of trace amplitude projections vs. time associated with far offset locations and subtracting increments of said firstand second series normalized to the sametime sample, one from the other to form at least a segment of said improved section;; (e) displaying said improved section depicting amplitude change as a function of time and centerpoint coordinate so that progressive change in a high intensity amplitude event in said displayed traces is identified as a function of progressive change in centerpointvalues.

25. The method of Claims 24 in which said one of said series of analyticfunction of known mathematical character of step (b) is selected from a group of linear and quadratic equations ofthe form: A(x) = CO + C1.x; A(x) = CO + C2x2; and A(x)=CO+Clx+C2x2 whereA(x) is the amplitude ofthe projected trace as afunction of offsets and C0, C1 and C2 are constants determined by matrix algebra.

26. A method for determining valuablecharacter- istics of stratå inthe earth using high-intensity amplitude events in seismic records, comprisins the steps of: (a) generating seismic data, including a record of signals frnrn acoustic discontinuities associated with said strata of interest by positioning and employing an array of sources and detectors such thatcenterpoints between selected pairs of sources and detec torsforrna series of centerpoints Elong alive of survey, said recorded signals being the output of said detectors; ; (b) by means of automated processing means, stati.cally and dynaminally correctingsaid recorded signals to form correed traces whc-reby each of said corrected traces is associatedwith acenterpoint horizontally midway between-a scaurce-detector pair from which said each corrected trace was originally derived; (c) by means of automated processing means, indexing said correctedtraces intwo dimensìons whereby each of said correctedtraces is identified in it relationship to neighboring traces on the basis of progressive changes in horizontal offset value versus progressive changes in common centerpoint locazion; ; (d) weighting said series of traces of step (c) by semblance coefficients wherein after a series of normalized ratios of output-to-input energy is generated by stacking, events in the traces associated with a limited number of phases are better indicated; (e) displaying saidweightedtraces or representations of said weighted traces of (d), whereby progressive change in a high-intensity amplitude event of said displayed traces or representations of said traces isidentifiedasafunction of progressive change in said horizontal offset values.

27. A method of seismic prospecting, substantiallyas hereinbefore described with reference to the accompanying drawings.