PP. 927-9.38,






Techniques are described whereby multiple coverage of the subsurface is obtained. Detector spreads and shotpoints are arranged so that the channels representing common depth points are recorded with appreciably different horizontal distances between the shotpoints and detector stations. The channels which have a common reflection point are combined, or stacked, after appropriate corrections for angularity and travel time to datum have been applied. Reflections which follow the assumed travel paths are greatly enhanced, and other events are reduced. Methods for attenuating multiple reflections with respect to primaries are discussedin considerable detail. Typical field comparisons between conventional and stacked traverses are shown to illustrate the degree of improvement which can be obtained in the signal-to-noise ratio. General considerations applicable to field usage, and the geographic range of field experience are summarized.

Improvements geophysicists have (1936), (1954), noise arrays. Applications either singly, to solutions multiplicity however, inherent and larger, tion yielded Poulter Parr have encountered.

in the




Multiple different suggested advocated velocity This recorded geophone

coverage of the same subsurface shot and for multiple detector positions several purposes. Green about

with (1938)

seismic signals has been a continuing to solution, From (1950), been (1955) new


with been

has been a com-

for many years. As certain problems ones have (1936), (1953), (1960), Kilipsch Woods to Graebner described of properties Rieber Reynolds various the

paths centered

mon depth point writer with

to eliminate proposed

the effect of dip on that the informapoint, but and Thus,

determinations. (1956) with a given reflection a multiplicity appropriate be combined


to utilize

tion associated


multi-element techniques, As the

of shotpoint



of these and related or in combination, many difficult of shotpoints and

after applying if the reflected

time corrections.

have produced detectors is inwith an

signals received

along the several their resultant of signals.


paths are adjusted Perturbations degraded incidence

for coincidence,

sum will be proportional following

to the number

creased to cope with still more difficult we are eventually limitation. the subsurface confronted As the arrays


other than the postulated and hence will be For random enhancement to pattern per-

ray paths will not be coincident, relative the average

become larger the summafrom a

to the reflections. theoretical

area which is averaged In practice arrivals

increases correspondingly. or integration subsurface area theoretically may result from pattern

will be proportional ber of signals. This formance. point The

to the square root of the numis analogous the source and receivthe limitations source and of recon-

of reflection

as great as ten acres of less than

Since, however,


ing points have been selected so that the reflection is common horizontal to all paths, techniques between by only spacing conventional ceiver pattern no longer apply.

700 ft. This, of course, tends to obscure the very detail which is being sought. The multiple-coverage, common-reflection vised to provide multiplicity without point technique was dea practical means of increasing

is restricted

the following

this limitation.

siderations: received by the Editor

* Presented at the 31st Annual SEG Meeting, Denver, November 8, 1961. Manuscript July 23, 1962. t Petty Geophysical Engineering Company, San Antonio, Texas. 927

928 1. in the spread. processes greatest distance over with which adequate the reflected signals persist simi- This is. Common depth points are recorded on the same channel throughout each sequence of twelve recordings. The both the spread and shotso that a syndepth points by The It paths which have a common the number between that the progressive two intervals is 12 or one-half distance setup is maintained. readily Loper and available. a total of 12 shotpoints the dots underneath is seen that must be for the end stations SHOTPOINT LOCATIONS GEOPHONE LOCATDNS REFLECTION POINTS V 0 ’ (SEOUENCE 2) FIG. The larity. 1. Detector spread advances one station. point to develop the desired spreads multiple will be is in a variety 24detector of ways. In this example are advanced metrical point on the same channel throughout of available point. in the arrangement of Figure 2 can be sequence of 12 shots. many be permissible extremely different mation tion. shotpoint backs up one station for a total of twelve recordings. and of a error in the postudistance of more than than n-ould mile if typical must be kept small with respect to the reflection period. The greatest distance adjustments The probable increases which of with will the W. preferred. station split-spread In the following configuration 24detector station unique property reflection number reflection stations variation shotpoint of recording used. ALTERNATE times longer with conventional attenuation apparent techniques. program. of the array summation Gth a total of the number length one coincidence accuracy. Magnetic Pittman convenient Figure This for obtaining tape recording equipment. Each thus provides effective great moveout the shots recorded of events with mill require transposi- wavelengths. 2. has become and has made the necessary summation and economical. Common-channel spread sequence giving 12.fold coverage of subsurface (multiplicity). of course. i.e. lated stepout 12 shots is equivalent station intervals to a 12 are used. channel in this manner correction but sum- can be made without (1954). Kate horizontal and detector corresponding channels from to each shot have been indicated the setup. the of the is 22 It is possible coverage examples. 1 illustrates multiple ARRANGEMENTS one simple field procedure coverage of the subsurface. arrangement has the a data from a specific multiplicity. . channels element one-half Thus. Harry permit requisite and Mayne intervals. If a standard used.

. v . . . . Figure 13 illustrates a sequence which also yieltls six-fold multiplicity using three shots per spread.Jll . .?24 ...Horizontal Data Stacking Techniques 929 I 0 0 0 0 0 0 0 0 0 SHOTPOINT LOCATIONS V GEOPHONE LOCATIONS o .P20 .oooooooooo I 00000000000 I I 0000000 lq3d~o. . .. . . . . l . . . . . . . . ..p 5 Sj7 spa 10 S. . .l? I ! l “20s. . . . .q.. . ...%. I Detector sprc~dan11 shotpoint are advancedtogether by two stations for successive recordings..p5.e2* S.?24 . . . .:. .*24S. .. REFLECTION POINTS 0 0 0 0 0 3 0 00000 0 0 0 0 . . -A 0 0 0 0 0 0 0 0 0 0 0 0000000 0 0 0 0 . . . .p7. The preceding examples point the way to the .*&Spg . .. ... but allows the spread to be moved six stations ahead each time Table 2 shows the corresponding channel composition schedule. tJ0000000000 00000000000 I 000000000000 . 9.. . .. .p3. .ps.p . . . . . . . . .e. .. .ys S. . . . . . #l S. . .*g l l l l . .. XI SpS. . . . . . . . . . .oooooooo 0 0 0 00000000 I “24 #l6 S. used to obtain the full multiplicity of six. . In each of the previous examples only one shot is made into each instrument spread. . . . . v. . . . . 000000 -_-id --. .f?tj . .V. Table 1 shows the channels recorded from each shotpoint which are used to compose the 24 depth points under Shotpoint 7. . . . 2. .::. . . Sf4V. . . . S. ..p l . . ..po 0 0 0 0 0 &oq3000000000 \~o. .. Symmetrical straddle spread sequence giving six-fold multiplicity. . .P12 I DEPTH POINTS I 24 I’G. . zc5Se6. . *gSs. .. . . .

where Z=depth below the shot. 1 2 __________~_____~. Table Powerful evidence confirming the existence of multiple reflections was presented by Ellsworth et al (1948). Table 3 shows the channel composidevelops multiplicity. Harry Mayne Table 1. Figure 5 illustrates a rather typical example. of shotpoint 7. MULTIPLE REFLECTIONS of detector stations in the spread. Stacked depth point 13 is under shotpoint 7. and stacked depth points 14 thru 24 correspond to those recorded on channels 13 thru 2 3 of shotpoint 7 respectively. so.22. P. This type of arrangement is particularly efficient in attenuating multiple reflections..000+0. The common-reflection point technique can be an excellent tool in reducing multiple reflections even though their source in many cases cannot be precisely determined. and only a single shot per spread is taken. D No. Thus if: (as previously by which M= path multiplicity N-number ~=number tion schedule. has been as- 2. as recorded from the indicated shotpoints. the shot is located at the end of the spread. Table illustrating the channel combination schedule for data recorded using the spread SCquence of Figure 2. -4 hypothetical velocity function of 6. the spread progression.1. 21 1 22 23 1 24 c 4. the following ity expression will define the multiplic- for any sequence: M Figure 6 illustrates six-fold = ss/212.: 24 22 23 24 \C 123456789101~2 L_-A~-----__-r_1/~3 . assuming for simplicity that the shot and the near-surface velocity contrast generating the multiples are at the same level. Depth 3 19 17 15 4 20 18 16 5 21 19 li 6 22 20 18 7 23 21 1Y 8 9 10 11 12 point Ko. Later experience and theoretical studies on synthetic seismograms have indicates! that occurrence of this insidious form of “noise” is dangerously common. respectively. arrangement which a third This is similar to the sequence of Figure 2 since the shot location and spread are both advanced two stations. s. Depth points of the stacked resultant are composed of the channels included in each vertical column. Depth points represented by the stacked data are the same as those recorded from shotpoint 2 C.930 W. T- 3 ~ 4 I 23 19 15 12 8 4 I 5 / 24 20 16 ~ 9 5 I2 6 7 10 11 12 13 14 15 16 17 18 19 20 I -1 21 - 22 24 : 3 4 5 6 ~ 16 i20 24 21 17 13 22 18 14 11 7 3 21 17 13 10 6 22 18 14 11 7 3 23 19 15 12 8 4 24 20 16 9 3 1 i I 21 17 13 10 6 2 22 18 14 11 : I 23 19 15 12 i ~ 24 20 16 ~ 9 5 1 21 17 13 10 6 2 22 18 14 11 3 23 19 I5 12 8 4 24 20 lb 9 3 1 21 Ii 13 10 6 2 22 18 14 11 7 3 23 1Y 15 12 8 4 24 20 16 9 3 1 21 17 13 10 6 2 23 I’ ) IS 12 8 4 general particular expression for the multiplicity of an> defined).~___~ 17 15 13 18 16 14 C 24 23 22 20_-21 124 22 23 24 17 15 13 4 1 2 12 10 8 13 11 9 1 14 12 10 2 15 13 11 3 1 16 14 12 4 2 18 16 14 19 17 15 20 18 16 21 19 17 24 2’ : . S= number of shot positions for each spread. Depth points of the stacked resultant correspond to those recorded on channels one thru 12. However. of stations is advanced. 1 2 Dmth mint No . which changes the geometry of the common-reflectionpoint paths. Table illustrating the channel combination schedule for data recorded using the spread sequence of Figure 3.

........... .... . .... ..... 3.. B................... .. ............. .. Shotpoints spaced two intervals apart.... ......... . _ 11 12 13 14 21 17 13 9 5 1 22 18 14 10 6 2 23 19 24 20 21 22 18 14 10 6 2 23 fi 11 i 3 20 12 i 13 5 9 1 14 10 6 2 15 11 i 3 21 22 23 _ I i _ ’ I‘ SHOTPOINT LOCATIONS GEOPHONE LOCATIONS REFLECTION POINTS V 0 0 s....... Depth points represented by the stacked data are the same as those recorded from shotpoint 6.........F! 3 I c A ...... . The detector spread is advanced six stations forward after each sequence of three recordings......... I 9 DEPTH POINTS I “24 FIG... B .... s. ......... 2 22 18 14 10 2” 3 23 10 15 11 7 3 4 24 20 16 12 8 4 6 _ 22 18 ti 6 2 7 9 I 10 23 19 15 11 7 3 Depth point No. c. .. s......c I c A.. c ..... Nonsymmetrical spread sequence which gives six-fold multiplicity using three shotpoint locations for each detector spread.. .... .... ................. Table illustrating the channel combination schedule for data recorded using the spread sequence of Figure 4...“............ . .l? 2 c ...... ....Horizontal Data Stacking Techniques 931 Table 3... .. A c ... B .....

019 . the stepout will. depend ~II the Since the travel shown by the will follow is confined to a lower velocity 5. . of course. the multiples will be out of phase as indicated by the lower curve.035 Actual Stepout Stepout for Optimum Attenuation 0 : 1: 17 21 2785 1925 3640 4500 _ . Station NO. Shotpoint and spread advanced together bY FIG. exhibit coefficients involved. Single-ended spread sequence giving’ -fold two stations between successive recordings. since there is so little difference in stepout between channels one and five. it will geometry. Hence if a number of channels with different shotpoint-to-detector distances arc combined so that the primaries are in coincidence. 4. For example. latcd at a depth of 8.898 ft would also have a total travel time of about 2. and the multiple has a period of 0.960 ft which corresponds to a time of 2.028 .007 . Station interval is 2 I5 ft.028 Actual attenuation of 5 channelsummation>4. suppose we combine five paths. Table comparing the actual stepout dikferences obtained with the desired theoretical optimum values when the spread arrangement of Figure 4 is adapted to the example shown in Figure 5.011 . assuming straight-path reflection upper curve of Figure The deep primary Distance 1070 215 I 1 the stepout shown by the middle curve. The detector spacing is chosen to provide the desired overall difference in stepout between the nearest and the farthest common-reflection-point channels recorded. The relative the two reflections reflection path of the multiple zone. Harry Mayne & multiplicity. sumed.5 to 1. The second(simple multiple) from a shallow strength of order reflection reflector at 3. -4n objective vertical two-way of major interest is postuTable 4.300 sec. More attenuation can be obtained by using only five channels as shown.014 .021 . Similar arrangements with the shotpoint at one end of the spread are preferred for use in attenuating multiple reflections. The lower curve shows the difference in stepout between the two events. .300 sec.932 W.006 .

Since the differential-stepout-versus-distance curve is not linear.035 sec. One problem is apparent from Figure 5. Of equal importance is the fact that the amount and direction of dip of either reflector is . our station interval should be 4.180 v = 6000 RECORD + .2E time =2. If the required 0.500/21 or approximately 215 ft. Corrections for the moveout of the primary would be applied. but become increasingly unwieldy. and the five channels combined to producea singleresultant. Stepout as a function of shotpoint-to-detectordistancefor a primary reflection and a multiple with the same total travel time Assumed velocity function is typical of the Gulf Coast of Texas.300 not significant if common-reflection point geometry is applicable.007 set difference in stepout between the successive paths. Attenuation behavior will be the same as for a five-unit pattern array.028 set stepout is developed between a station 1 and a station 21 (Table 3).I60 SEC. assuming sinusoidal waves.500 ft. Suppose that data are to be recorded using the spread arrangement of Figure 4. Obviously paths with a common reflection point must be used if excessive subsurface averaging is to be avoided. / . the desired station intervals are nonuniform. These approximations yield better than a 4. Longer distances will not seriously affect the attenuation.5/S or 0. Shorter distances will seriously reduce the attenuation and should not be considered.028 set between the extremes if adequate attentuation is to be obtained. The lower curve represents the difference in stepout between the multiple and the primary at each distance. Hence we must have a minimum of 0.Horizontal Data Stacking Techniques 933 sec.5to-l theoretical improvement in the primary-to-multiple ratio. or a total of 0.I40 Qc Oq / / 0 500 1000 1500 2000 TO 2500 3000 3500 4000 4500 5000 5500 6( SHOT DETECTOR DISTANCE-FEET FIG. 5. Xote that the distance to the fifth or farthest channel in this example is 4. . Table 4 shows the comparison between the theoretical stepout desired. The estimated distances required to establish the necessary differences are indicated in Figure 5 by the circled points numbered one through five on the lower curve.03. There is so little stepout difference between the nearest channels (1 thru 4) and their next corresponding common-reflection-point channels (5 thru 8) that better attenuation is . The circled points numbered 1 thru 5 on the lower curve show the distance at which channels should he recorded to obtain infinite theoretical attenuation of a sinusoidal multiple with a period of 0. and the actual values obtained.

fold multiplicity in the Powder River Basin of Wyoming. . Shot and detector patterns. Comparison of the same traverse recorded with conventional single coverage and stacked 12. Harry Mayne CONVENTIONAL SINGLE COVERAGE FLTER: OUT42 MIXING: 40% TAPERED HDRlZONlALLY STACKED TWELVE FOLD COVER&E FILTER:OUT42 MWNG:40% TAPERED FIG. 6.934 W. and playback settings were identical for both sections.

Horizontal bata Stacking Techniques .

Harry Mayne w .W.

They are preferable because reflection character and apparent damping are improved. Spreads should be as long as practicable.are and PaIo Duro Basins of Texas. and 410 ft is desirable if conditions permit. Corrections for moveout.035 set. this illustrates the order of subsurface averaging nhich will probably occur. weathering and elevation must be accurately determined and precisely applied. and it is then combined with any normal polarity channel having similar waveform and amplitude. horizontal-datastacking technique has added a new order of magnitude to the usable dimensions of multipath pattern array geometry. wider-band filters can generally be used. and the resultant primary reflection pulse will be broadened bg one-half cycle. i\ttenuation of the multiple reflection will be limited only by the degree of waveform identity between the two signals. Preliminary traverses in an area should be recorded with greater multiplicity than may be necessary. and xvi11 also make the process independent of the dips involved. and the required correction precision is somewhat reduced. Although fortuitous dip relations might permit combination of channels closer together. While experience The common-reflection-point. Figures 7 and 8 show a similar comparison in Southern Oklahoma. Figure 5 indicates that channels separated by a spread distance of approximately 3. the Delav. Filter: Because of improv-ed attenuation of “ground roll” and other extraneous lowfrequency noise. The most effective shot and detector patterns permitted by economic considerations should be employed. The longer spreads are much less vulnerable to multiple reflections. the deep Frio and Wilcox trends of the Texas Gulf Coast. and the precision of the phase and amplitude adjustments. Field effectiveness has been demonstrated in Mississippi. Shot pattern. The added multiplicity available with this technique should be used as a supplement to normal good practice and not in substitution thereof. CONCLUSION Figure 6 illustrates a conventional traverse in the Powder River Basin of Wyoming compared to the same traverse with 12. The common-reflection-point technique is also advantageous when inverted-polarity mixing is used to attenuate multiple reflections. the Powder River Basin of . After adjustment the polarity of one channel is reversed. Test processing can then be performed to select the most economical arrangement. but use of the commonreflection-point technique 141 eliminate subsurface averaging. 4. Thus. 2. but data quality is enhanced. DATA COMPARISONS has varied somewhat from place to place. South Louisiana. certain common considerations have become apparent. and deletion of obviously poor data can be of great benefit. 5.Horizontal Data Stacking Techniques 937 obtained if channels 5 thru 8 are omitted. the thrust areas of Southern Oklahoma. Approximately the same daily production can be maintained using this method as would be attained \vith conventional operations. This form of skip mixing is effected by first adjusting the multiples to time coincidence. Destructive interference will also attenuate the primary reflections unless they happen to be approximately one-half cycle out of phase on the channels \vhich are combined. Not only are they more efficient from a production standpoint. Detector stations should be at least 220 ft apart. Signal-to-noise ratios have been enchanced well beyond the saturation point of conventional pattern methods. Use of all six channels reduces the corresponding ratio to2. geophone arrangement and playback settings were identical. Careful editing of corrections. for a primary reflection period of 0. This means that the differential stepout between the multiple and the primary reflections must be one-half the period of the primary reflections.5 to 1.fold multiplicity. The multiple reflection will be attenuated as before. and complement the multiplicity obtainable with conventional pattern arrays. 3. FIELD EXPERIENCE Considerable field experience in a wide variety of terrain has been accumulated. 6. This excessive subsurface averaging will be avoided if common-reflection-point channels rvhich possessthe proper difference in stepout are used. 1.500 ft must be selected to avoid attenuation of the primary reflection.

25. Harry. p. D. O. pp. Shot hole array for elimmating horizontally traveling waves: U. . (Abstract: Geophysics. velocity determination bv means of reflection drofiles: Geo&sics. Mayne. 0. 689). Jr. Klipsch.. 15. v. The wholehearted support of the Petty Company and all my co-workers is gratefully acknowledged. Woods. Design factors for multiple array? of geophones and shot holes: Oil and Gas Journal. Frank. C. 1956.. ACKNOWLEDGMENTS The writer sincerely appreciates the confidence and cooperation of the many client companies who have accepted this development. C. H. G.146 (application 1950). F. 295.Seismic surveying:U.. 50. v. 145. A new reflection system with controlled directional sensitivity: Geophysics. (Abstract: Geophysics. Poulter.. Hewitt. and Fu. I’. Waterman. v. 365. 1955. v. B. John... F.. C. S.Seismicrecording on magnetictape: Geophysics. Sloat. 21. W. J.906 (application 1950)) (Abstract: Geophysics. ‘ homas C. k. Johnson. 283. 954). 18. 32. Ellsworth. 1. n. Lester. v. 181. p.. Green.. Y. Reynolds. 1954. 1954. 19X!.. 19.642. 1936. So&e aspedts of’ ultiple recording & in seismic prospecting: Geophysics. 1938. Itlner. 104. p.. v. Paul I. S. p. F. H. 1. 1948.. REFERENCES Loper. B. 20.. 856): Parr.732.939 W. Hansen. a symposium: Geophysics. Dix. C.53). p. R... P. The Poulter seismic method T of geophysical exploration: Geophysics. Walling. Seismic data enhancement-a Wyoming and Montana. Gutenberg. Multiple reflections. v. T. and Pittman.698. Seismic surveying: C. and Colombia. Patent 2. v. Rieber. South casehistory: Geophysics. J. 97. P. p. America.. v. no.. R. Jr. J. Joseph D.. 1936. 1953. and assisted in its application. Dean? Dresbach. Curtis H.‘.927 (application 19. v. 1-85. . . Patent 2. 3. v. p. Patent 2. 1960. p. 13. R. Harry Mayne Graebner. C. S.

Sign up to vote on this title
UsefulNot useful