Ex. PGS 2016
`
`EX. PGS 2016
`
`
`
`
`
`
`
`

`
`GEOPHYSICS, VOL. 57, NO.4 (APRIL 1992); P. 543-553, 15 FIGS .. 2 rABLES.
`
`Incomplete AVO near salt structures
`
`Christopher P. Ross*
`
`ABSTRACT
`
`Amplitude versus offset (AVa) measurements for
`deep hydrocarbon-bearing sands can be compromised
`when made in close proximity to a shallow salt pierce(cid:173)
`ment structure. Anomalous responses are observed.
`particularly on low acoustic impedance bright spots.
`CMP data from key seismic profiles traversing the
`bright spots do not show the expected Class 3 offset
`responses. On these CMPs, significant decrease of far
`trace energy is observed. CMP data from other seismic
`profiles off-structure do exhibit
`the Class 3 offset
`responses, implying that structural complications may
`be interfering with the offset response. A synthetic
`AVO gather was generated using well log data, which
`supports the off-structure Class 3 responses, further
`reinforcing the concept of structurally-biased Ava
`responses. Acoustic, pseudo-spectral modeling of the
`structure substantiates the misleading Ava response.
`Pseudo-spectral modeling results suggest that signal
`degradation observed on the far offsets is caused by
`wavefield refraction-a shadow zone, where the
`known hydrocarbon-bearing sands are not completely
`illuminated. Such shadow zones obscure the correct
`AVO response, which may have bearing on explonl(cid:173)
`tion and development.
`
`OVERVIEW
`
`Ostrander (1982) presented the first integrated discussion
`on the delineation of hydrocarbon reservoirs using amplitude
`versus offset (AVa). Ostrander's SEG presentation and
`subsequent paper (1984) linked the effects of Poisson's ratio
`with gas saturation (Koefoed, 1955; and Domenico, 1974;
`1976; and 1977). The use of AVO techniques as a higher(cid:173)
`order refinement
`to direct hydrocarbon indicator (DHI)
`technology has since been developed and used successfully
`in clastic sedimentary basin sequences around the world
`
`(Rutherford and Williams, 1989; Hilterman, 1990; Mazzotti,
`1990). However, not all known hydrocarbon-charged reser(cid:173)
`voirs exhibit Ava responses indicative of hydrocarbons.
`Reasons for not observing appropriate Ava responses are
`varied. For one example, Hilterman (1990) showed that in
`some circumstances, higher-order petrophysical variations
`of clays in reservoir rock can cause discrepancies in Ava
`character. For another example,
`the focus of this paper,
`discrepancies in Ava character are generated by improper
`seismic illumination of the subsurface near complex geologic
`structures.
`In general, most published studies involving AVO have
`focused on the more common successes in relatively shal(cid:173)
`low, gently dipping strata. This is because shallow sediments
`typically exhibit greater ranges of porosity and larger con(cid:173)
`trasts in Poisson's ratio, permitting a more definitive AVO
`response. At shallow depths,
`the broader spectral band,
`made possible by minimized attenuation affects, enables
`better
`resolution.
`In addition, source-receiver distances
`greater than the objective depth allow wide-angle apertures
`for analysis, and structures with limited dip reduce the need
`for complex processing steps that may introduce artifacts
`that modify the Ava signature.
`Nonetheless, deeper plays exist and can be associated
`with more complex geologic structures. Plays such as these
`inherently have greater economic risk. Therefore, an under(cid:173)
`standing of Ava responses (in these more complex areas) is
`necessary to know when and why Ava techniques will and
`will not agree with the theoretical responses expected for
`hydrocarbon-charged reservoirs. These AVO response fun(cid:173)
`damentals are especially important for known hydrocarbon(cid:173)
`bearing sands that are often used as analogies for potential
`reservoir sands.
`
`A REVIEW OF AVO TECHNIQUES
`
`Shuey (1985) has simplified the mathematics of Zoeppritz
`(in Aki and Richards, 1980) by regrouping the necessary
`
`Manuscript received by the Editor March 21,1991; revised manuscript received September 19,1991.
`*Orxy Energy Company, P.O. Box 2880, Dallas, TX 75221-2880.
`© 1992 Society of Exploration Geophysicists. All rights reserved.
`
`543
`
`Downloaded 03/06/15 to 108.45.99.150. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`Ex. PGS 2016
`
`

`
`544
`
`Ross
`
`reflection coefficient terms into elements that correspond to
`the near,
`intermediate, and far offset ranges. Given the
`premise of small variations in petrophysical properties be(cid:173)
`tween reflecting layers, Shuey's equation can then be ex(cid:173)
`pressed as equations (1) through (4) as shown below, where
`
`R(e) = R o + [AoRo + Ilu 2] sin2e
`(1 - u)
`
`(1)
`
`(2)
`
`(3)
`
`and
`
`A = A o +
`
`I
`Ilu
`2 - '
`(1 - u) R o
`
`Vp , p, and u are the average values of both media, and e is
`the average angle of reflection and transmission. An exami(cid:173)
`nation of equation (1) for acceptable values for velocity,
`density, and Poisson's ratio, shows that large contraSots in
`Poisson's ratio result in significant variations in AVO (Koe(cid:173)
`foed, 1955; Ostrander, 1984). Domenico (1974, 1976) and
`others have demonstrated through laboratory experiments
`that sediments with gas-charged pore fluids have notable
`contrasts in Poisson's ratio when compared to sediments
`with brine-filled pores. Clean, gas-charged sandstones along
`and in the Gulf of Mexico have Poisson's ratio ranges of 0.10
`to 0.16, while their brine counterparts exhibit ranges of 0.35
`to 0.42. (See Table 1.) Variations in Poisson's ratio can be
`attributed to age, compaction, and volumetric impurities
`such as clays (Hilterman, 1990).
`Equation (3) consists of two terms that control the initial
`decrease or increase in AVO (gradient) for small angles of
`incidence. The first term assumes a constant Poisson's ratio,
`which results in an amplitude decay with offset, while the
`second term incorporates differences in Poisson's ratio.
`By neglecting the higher-order terms for larger angles of
`incidence in equation (1) and multiplying both sides of
`equation (3) by R o, substitution of equation (3) into equation
`(1) yields:
`
`R(e) = Ro + ARo sin 2e.
`
`(4)
`
`AVO analysis in this text will refer to Ra(e) to denote
`actual AVO measurements of seismic amplitudes, as op(cid:173)
`posed to R(e) for reflection coefficient AVO measurements.
`For simplification, Po will be used for R o and G in lieu of
`ARo. With these nomenclature substitutions equation (4) is
`now represented as equation (4a), which is graphically
`depicted in Figure 1.
`
`CMP data analysis for a particular event is addressed by
`plotting the observed AVO response [R a (e)] for pertinent
`angles of incidence (6). This permits an amplitude gradient
`(G) to be estimated via linear regression, as well as an
`extrapolated normal-incidence amplitude (Po). (See Figure
`1.) Po may be different from Ro in equation (2) since Po is
`estimated from G over a select swath of pertinent incident
`angles. Po is typically multiplied by G, yielding an AVO
`that
`is often diagnostic of hydrocarbon(cid:173)
`product (P*G)
`charged sands containing at least 5 percent gas by volume.
`By definition, low-impedance gas-charged sands have Class
`3 offset responses, and the AVO products of Class 3 re(cid:173)
`sponses are positive. AVO responses for clastic reservoirs
`have been listed in Table 2.
`
`Table 1. Poisson's ratio values for various Gulf of Mexico
`lithologies. Data compiled from in-situ field measurements and
`literature.
`
`Lithology/Pore Fluid
`
`Poisson's Ratio
`
`Unconsolidated shale
`High porosity sandstone
`Gas-charged sandstone
`Consolidated shale
`Limestone
`Salt
`
`0.38-0.45
`0.35-0.42
`0.10--0.16
`0.28--0.34
`0.28--0.32
`0.24-0.28
`
`Ra (8) = Po+ G sm2 8
`
`Role)
`
`* *
`
`* Observed
`
`--- L-2Fit
`
`(4a)
`
`FIG. 1. Graphical relationship of the AVO terms in equation
`(4a). The normal incidence intercept (Po) and gradient (G).
`
`Downloaded 03/06/15 to 108.45.99.150. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`Ex. PGS 2016
`
`

`
`Incomplete AVO
`
`545
`
`Table 2. AVO characterizations for various clastic reservoirs (modified from Rutherford and Williams, 1989).
`
`Designation
`
`Reservoir Attribute
`
`AVO Attribute
`
`Class I
`Class 2
`
`high-impedance sands
`near-zero impedance sands
`
`Class 3
`
`low-impedance sands
`
`large Ro and diminishing amplitude with offset
`near-zero Ro and increasing amplitude with offset or near-zero Ro with
`amplitude reversing polarity/phase and increasing at intermediate
`offsets
`large Ro and increasing amplitude with offset
`
`AN AVO ANALYSIS EXAMPLE
`
`Seismic reflection profiles used in this analysis and the
`areal extent of the upper seismic anomaly are shown in
`Figure 2. Seismic Lines A and B are shown in Figure 3 and
`Figure 4, respectively, and illustrate the amplitude anoma(cid:173)
`lies. Reflectors of interest are at an approximate depth of
`4900 m and reservoir traps are three-way closures against the
`flank of a shallow salt piercement structure. Tomlinson
`Geophysical Services (TGS) speculative data were collected
`with a maximum 4800 m offset, using 225 channels and a
`
`.++++++++
`
`Salt
`...
`...
`
`+
`
`..
`
`...
`
`--_~
`
`...
`
`Line B
`
`Line C
`
`~StudyAn.
`
`Gulf of Mexico
`
`• Selected CMPs
`
`I 1 Km I
`
`FIG. 2. DHI outline with prominent structural elements.
`Please note that
`there is no evidence of salt overhang
`associated with this salt diapir. Selected CMPs for each line
`have been annotated.
`
`group interval of 20 m. Data were processed to zero-phase
`(deterministically) after dip-moveout (DMO), prestack mi(cid:173)
`gration, and spectral balancing algorithms were applied.
`Both events have been penetrated by several wells. Well
`data have been correlated to the reflection seismic profiles
`via synthetic seismograms, and the anomalies are considered
`true DHIs (direct hydrocarbon indicator). Logs from a well
`penetrating the events were used to generate a synthetic
`CMP gather using the SYNAVO algorithm.
`SYNAVO is a synthetic AVO modeling program that uses
`ray theory instead of a total elastic solution. In areas where
`the velocity gradient is small (such as the Gulf of Mexico),
`the ray tracing algorithm in SYNAVO matches the output
`from total elastic synthetic seismograms (Hilterman, 1990).
`Synthetic CMP gather results are shown in Figure 5 with
`the anomalies of interest visible at 4.59 and 4.7 s. Figure 5
`illustrates the increase in trough and peak amplitude with
`offset as expected for the respective top and bottom of a
`low-impedance reservoir. AVO analysis was performed on
`the synthetic CMP and is shown in Figure 6. Gradient and
`estimated normal incidence values for the base of the upper
`anomaly are positive (Figure 6) as is the AVO product. The
`gradient was determined using L-2 regression and the Po
`value determined via gradient extrapolation. This increasing
`amplitude with offset response [RaUl)] from the SYNAVO
`the
`example is typical for known pay sands throughout
`trend.
`Unlike synthetic data, signal enhancements are typically
`required for geophysical analysis. For signal-to-noise reduc(cid:173)
`tion, trace-mixing was employed to construct "super-gath(cid:173)
`ers" for Lines A, B, and C. Since AVO analysis is per(cid:173)
`formed prior to the stacking algorithm,
`the random noise
`elements have not yet been addressed. Ostrander (1982)
`recommends partial stacking within each CMP gather for
`signa! enhancements. In areas with relatively gentle dip,
`super-gathers can be generated by trace-mixing several
`common source-receiver offsets in adjacent CMP gathers. In
`these examples, source-receiver offsets from five adjacent
`CMPs were summed with equal weighting. Sporadic noise
`was eliminated and signal-to-noise ratios were improved.
`These gathers are shown in Figures 7 and 8 for Lines A and
`B. respectively.
`Notice the strong coherent events at 4.5 through 4.8 s, and
`the diminished amplitude response occurring in the far 18 to
`20 traces of each record. These events correspond to the
`DHIs seen on the stacked sections (Figures 2 and 3), and
`differ from the SYNAVO results in Figure 6. Angles of
`incidence are estimated using interval velocities derived
`
`Downloaded 03/06/15 to 108.45.99.150. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`Ex. PGS 2016
`
`

`
`546
`
`Ross
`
`6756
`
`6792
`
`6844
`
`s
`CMP 6650
`3.0
`
`3.5
`
`-~ 4.0
`
`CD
`E
`j::
`
`4.5
`
`5.0
`
`5.5
`
`200m
`
`N
`
`7100
`3.0
`
`3.5
`
`4.0
`
`4.5
`
`5.0
`
`5.5
`
`FIG. 3. Migrated relative amplitude section-Line A. Notice the two well-developed DHls between 4.5 and 4.7 s, each with
`prolific trough-peak character typical of zero-phase data and thick blocky sands.
`
`Downloaded 03/06/15 to 108.45.99.150. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`Ex. PGS 2016
`
`

`
`Incomplete AVO
`
`547
`
`34753439
`
`3375
`
`E
`
`3200
`3.0
`
`3.5
`
`4.0
`
`4.5
`
`5.0
`
`5.5
`
`w
`
`-en-CD
`
`E
`i=
`
`3.5
`
`4.0
`
`4.5
`
`5.0
`
`5.5
`
`200m
`
`FIG. 4. Migrated relative amplitude section-Line B. Notice the lack of coherent energy in the salt mass updip from the DHls
`at 4.5 and 4.6 s.
`
`Downloaded 03/06/15 to 108.45.99.150. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`Ex. PGS 2016
`
`

`
`548
`
`Ross
`
`Ape~ for AVO Analysis
`
`......
`
`"" Ut.
`-H-ttt+H-t+<>4+t-H-~+H-H-1f++-f-H--I-tfJ.tl++-f-++t+'-+-! !-JIIJ-I+.II'! I I I
`..
`\ll':~
`
`CMP
`Offset (m)
`1.5
`
`325
`
`1603
`
`2880
`
`4658
`1.5
`
`2.0
`
`2.5
`
`-en-CDe
`
`i=
`
`5.0
`
`"o(e)
`
`• • • • • • • I i
`
`. . ft. t t • • • • • • + •• '.
`
`11'
`
`IS'
`
`zs' e
`ADgIe of Incidence at 4.60 •
`
`~
`a I.'
`::0""1 '.1
`I 1.1
`""~ ".1
`S0
`Z
`
`.1.1
`
`FIG. 6. AVO analysis plot of the SYNAVO model between
`incidence angles of I) degrees to 25 degrees. The base (peak)
`of the upper gas sand (4.6 s) exhibits a steady increase in
`amplitude with offset as shown in graphical form at
`the
`hottom of the figure. RaUl) is the normalized offset response
`for the selected event (arrow).
`
`6756
`CMP
`OII1'lIm) 35
`1.5
`
`..E
`
`i=
`
`2.0
`
`2.5
`
`3.0
`
`3.5
`
`4.0
`
`4.5
`
`5.0
`
`2.0
`
`2.5
`
`3.0
`
`3.5
`
`4.0
`
`4.5
`
`5.0
`
`2.0
`
`2.5
`
`3.0
`
`3.5
`
`4.0
`
`4.5
`
`5.0
`
`FIG. 5. Synthetic AVO model (SYNAVO). Upper and lower
`pay zones at 4.59 and 4.7 s have been marked with arrows.
`The thickness of each pay sand is approximately 30 m.
`
`Far __~Rela::'~~D~~c_-=e
`
`Near
`
`FIG. 7. Selected CMP super-gathers from Line A. Corre(cid:173)
`sponding CMP locations are marked on Figure 3. Relatiye
`proximity to the salt has been labeled for reference WIth
`Figure 2.
`
`Downloaded 03/06/15 to 108.45.99.150. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`Ex. PGS 2016
`
`

`
`Incomplete AVO
`
`549
`
`CMP
`3375
`Ofhll(m)1. 3~7
`
`16;47
`
`28,86
`
`3439
`472~ ~87 16;47
`
`2.
`
`~.,
`E
`;::
`
`2.0
`
`2.5
`
`3.0
`
`3.5
`
`4.0
`
`4.5
`
`5.0
`
`FIG. 8. Selected CMP super-gathers from Line B. Corre(cid:173)
`sponding CMP locations are marked in Figure 4.
`
`u
`
`~~mffif_mootu
`
`------_...\
`
`OmlftH
`
`n'
`
`.~.
`
`from seismic RMS velocities. For a far offset trace of 4754 m
`(Line A), the angle of incidence is 25 degrees (at 4.5-4.8 s)
`which is unfortunate and rather restrictive. In most AVO
`analyses, it would be preferable to have greater angles of
`incidence, permitting larger contrasts between near offset
`and far offset amplitudes.
`Results of interactive AVO analysis are shown for a
`Interactive
`super-gather CMP from Line A in Figure 9.
`analysis involves: refinement of RMS velocities; range lim(cid:173)
`iting the aperture of analysis by defining an incident angle
`range to eliminate near source bias; removing residual RMS
`velocity errors by static correcting the event in question; and
`eliminating erroneous traces with anomalous samples. A
`gradient and zero-offset amplitude are estimated via a best fit
`to Shuey's equation and presented graphically in the lower
`portion of Figure 9. Comparison of Figure 6 with Figure 9
`would imply that the physical properties of the model are not
`congruous with the observed anomalies.
`The diminished amplitudes occurring on the CMPs from
`Lines A and B at offsets greater than 3000 m are enigmatic.
`Data processing artifacts were first considered in attempting
`to explain the discordance in AVO for Lines A and B. Both
`DMO and prestack migration processes were removed with
`no detectable change in the AVO response, thereby elim(cid:173)
`inating the algorithms as suspect.
`In an attempt to explain this incongruity, a third line (Line
`C) with identical acquisition parameters was processed. Line
`C (Figure 10) is a strike line that images the events in a
`slightly downdip position. Figure 12 shows the interactive
`AVO analysis of a type super-gather from Line C as shown
`in Figure II. Note the concordance between Line C's AVO
`analysis in Figure 12 with the SYNAVO model in Figure 6.
`After examining the positive AVO response on CMPs
`from Line C, processing artifacts were no longer considered
`to be causes in the AVO disagreement between the observed
`and theoretical responses. The proximity to the salt and the
`acquisition geometry of the seismic lines were therefore
`suspect.
`
`2-D PSEUDO-SPECTRAL MODELING
`
`After constructing a 3-D model, a 2-D cross-section was
`extracted along Line A, and is shown in Figure 13. One low
`acoustic-impedance gas sand (40 m thick) was inserted at an
`equivalent depth using petrophysical properties from the
`well logs.
`is the primary
`interference of the salt
`Since structural
`purpose of the modeling, an acoustic algorithm was used in
`lieu of a full-elastic algorithm for reduced computational
`cost. Several synthetic shot records were acquired with the
`acoustic pseudo-spectral algorithm using a 300 x 240 numer(cid:173)
`ical grid with a node separation of 33 m. Data were acquired
`at 33 m intervals with a spread length of 4818 m and a near
`trace offset of 300 m. Shot source locations (labeled S l' S 2'
`and S 3) and their respective receiver locations are shown in
`Figure 13. with the synthetic shot records displayed in
`Figure 14. Figure 14 illustrates complete illumination of the
`reflector on the first and second shot gather, while the third
`shot gather does not completely illuminate the reflector at
`
`u-
`
`• •• •
`
`omtftH
`
`.~
`
`"alel
`
`1.1
`
`..
`....a
`::l'"~ I.'•~ ...
`~ ....
`il
`
`".
`
`,..
`
`e
`2'·
`21-
`Angle of Incidence at 4.615
`
`-1.1
`
`~0
`
`Z
`
`FIG. 9. AVO analysis plot of super-gather 6756 from Line A.
`Notice the strong negative gradient and subsequent "nega(cid:173)
`tive" AVO response for the base of the upper DHI (arrow).
`The incident angles in the analysis have been restricted
`between 12 degrees and 23 degrees because of near-source
`bias. Shaded areas indicate particular offsets that were
`omitted because of poor signal.
`
`Downloaded 03/06/15 to 108.45.99.150. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`Ex. PGS 2016
`
`

`
`550
`
`Ross
`
`w
`eMP 3601
`3.0 I
`
`3.5
`
`-en 4.0
`-CDe
`
`j::
`
`4.5
`
`5.0
`
`5.5
`
`200m
`
`3395 3367
`
`3299
`
`E
`
`3151
`13.0
`
`3.5
`
`4.0
`
`4.5
`
`5.0
`
`5.5
`
`FIG. 10. Migrated relative amplitude section-Line C. The two DHI anomalies are seen at 4.6 and 4.~ s.
`
`Downloaded 03/06/15 to 108.45.99.150. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`Ex. PGS 2016
`
`

`
`Incomplete AVO
`
`551
`
`10.Um
`I
`
`II
`
`S
`
`~
`oS...~
`
`t:l
`
`0.0
`I
`
`Distance (Km)
`
`5,,1
`
`5,1 I
`
`II,
`
`R,
`
`"4 I
`
`0.0
`
`R,
`
`1825 mil
`2.03g/cm3
`
`2.5
`
`2100 mil
`2.10
`
`...
`
`2700 mil
`2.15g1c'"
`
`5.0
`
`2600 mil
`2.20glc'"
`
`.
`
`.
`
`.
`.
`
`· · .
`·
`.2.14g1c.....
`
`4600 mil ••
`
`2150 mil
`2.00g/''''
`
`.
`
`·
`· ·
`.
`·
`·
`·
`
`.
`
`h
`. Km
`..
`
`·
`·
`
`·
`
`•
`
`FIG. 13. Representative 2-D velocity profile of Line A with
`source and receiver locations for pseudo-spectral forward
`modeling.
`
`3395
`4726367
`
`1647
`
`2886
`
`4726
`1.5
`
`2.0
`
`2.5
`
`3.0
`
`3.5
`
`4.0
`
`4.5
`
`5.0
`
`2.0
`
`2.5
`
`3.0
`
`3.5
`
`4.0
`
`4.5
`
`5.0
`
`~
`
`..
`
`E
`i=
`
`Far
`
`Relag~~D.f:J'f'''!'.--Farther
`
`FIG. 11. Selected CMP super-gathers from Line C. Corre(cid:173)
`sponding CMP locations are marked in Figure 10.
`
`Aperture for AVO Analysis
`
`,..
`Ie
`
`Shot
`Ottut(m)
`
`2.5
`
`3.0 --ffi1!r-+-n+
`
`3.5
`
`E 4.0
`
`..e
`
`i=
`
`3
`4818 330 1650 2870
`
`2870
`
`4618
`
`:-'r+.'"1rl'''---;
`
`3.5
`
`4.0
`
`4.5
`
`5.0
`
`I, 5.5
`
`4.5
`
`"lei
`
`OmlftH
`
`.'
`
`....
`
`..
`
`...
`
`* •••
`
`..
`
`".
`
`".
`
`e
`teO
`tr
`Angle of IncideDce at 4.66 s
`
`~
`
`'0a 1.'
`:::l...
`~ ...
`...
`~ ....
`1l
`
`e~
`
`@0
`
`Z F
`
`IG. 12. AVO analysis plot of super-gather 3367 from Line C.
`Note the strong positive gradient between the angles of 10
`degrees and 23 degrees for the base of the upper DHI
`(arrow). Shading indicates offsets that have been omitted
`because of poor signal.
`
`-1.1
`
`Relative Distance
`Far---= from salt ~---Near
`
`FIG. 14. Acoustic pseudo-spectral seismogram for shot(cid:173)
`receiver locations in Figure 13. Shot gathers 1,2, and 3 are
`shown left to right. Note the modeled horizon of interest
`(arrows) at approximately 5.2 s. Shot 3 illustrates the incom(cid:173)
`plete illumination of the DHI. Shots 1 and 2 have complete
`seismic coverage of the reflector but do not completely see
`the DHI updip.
`
`Downloaded 03/06/15 to 108.45.99.150. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`Ex. PGS 2016
`
`

`
`552
`
`Ross
`
`the far offsets. (Even though two of the three synthetic
`gathers image the reflector completelY, they are not imping(cid:173)
`ing on enough of the DHI to enhance the amplitude strength
`significantly.)
`Incomplete illumination on the third shot
`gather is due to the proximity of the salt. Upgoing reflected
`energy entering the salt is critically refracted.
`A snapshot of the acoustic wavefield at 2.6 s (Figure 15)
`embodies the effect of the salt mass. Reflected primary and
`first-order multiple wavefronts are readily visible in the
`sediment but are not seen in the salt diapir. The only
`observable energy in the salt is the transmitted primary and
`first-order multiple. No coherent reflected energy is seen in
`the salt for any snapshot of the wavefield. Therefore, the
`energy is interpreted to be critically refracted along the salt
`interface. The critical angle is 26 degrees to 35 degrees for
`the shallow sediments truncated against
`the salt, while
`angles of incidence for the reflected upgoing energy are
`much greater.
`
`Hence, the proximity of the salt may for some shooting
`geometries and line orientations result in incomplete CMP
`illumination, consequently altering the AVO response.
`
`CONCLUSIONS
`
`The lack of agreement between estimated and observed
`AVO responses for seismic lines in proximity to the salt
`mass are related to incomplete seismic imaging. Affected
`data in this study are the far offset source-receiver pairs that
`lack reflection energy due to critically refracted energy along
`the salt diapir/sediment
`interface. Missing or attenuated
`reflection energy on the far offsets can result in biased and
`erroneous AVO responses that do not adhere to the pre(cid:173)
`dicted hydrocarbon response. Such problems are further
`complicated at depths where angles of incidence for AVO
`analysis are limited.
`Although the conventional stacked data appear to illumi(cid:173)
`it
`is inadequate for AVO
`nate the structure sufficiently,
`
`9
`Trace_ 1
`
`Distance (Km)
`5;0
`
`2.5
`
`7.5
`
`2.5
`
`5.0
`
`7.5
`
`Km
`
`Trough (-)
`
`Peak (+)
`
`FIG. 15. Snapshot of pseudo-spectral model at 2.6 s from shot 2. Downward-traveling wavefronts have been labeled P (primary)
`and P (primary multiples). Upward-traveling wavefronts have been labeled Rand R m for reflections and reflected multiples.
`The primary salt reflection has been labeled with an S. Notice the lack of any coherent upgoing energy in the salt diapir.
`
`Downloaded 03/06/15 to 108.45.99.150. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`Ex. PGS 2016
`
`

`
`Incomplete AVO
`
`553
`
`analysis. Therefore, it is strongly recommended that CMP
`data be scrutinized near complex structures to confirm that
`substantial seismic coverage has been obtained for Ava
`analysis.
`The following is suggested for CMP examination near
`complex structures prior to AVO analysis:
`
`1) Generate synthetic Ava models for hydrocarbon and
`nonhydrocarbon-bearing reservoirs
`from available
`petrophysical data.
`2) Examine CMP data for continuity. Compare CMP data
`with Ava synthetic data and check CMP data for
`complete seismic coverage of key reflectors.
`3) Examine prestack processing algorithms in the proc(cid:173)
`essing sequence that may affect amplitude versus offset
`responses.
`4) Employ structural seismic modeling in areas where
`there is disagreement between Ava synthetic seismo(cid:173)
`grams and CMP data.
`
`These procedures may avoid a similar dilemma as afore(cid:173)
`mentioned, bias the AVO analysis, and complicate economic
`risk analysis.
`
`ACKNOWLEDGMENTS
`
`I wish to express my thanks to Turgut Ozdenvar for his
`time and help with our pseudo-spectral modeling software.
`
`Tim Long, Maynard Redeker, Paul Beale and Dave Cefola
`for their discussions and review. I also express my gratitude
`to TGS for the release of their seismic data for this publica(cid:173)
`tion, and to Oryx Energy Company for their time, resources,
`and permission to publish this paper.
`
`REFERENCES
`
`Aki. K. L.. and Richards, P. G., 1980, Quantitative seismology: W.
`H. Freeman & Co.
`Domenico, S. N., 1974, Effect of water saturation on seismic
`reflectivity of sand reservoirs encased in shale: Geophysics, 39,
`759-769.
`- - 1976, Effect of brine-gas mixture on velocity in an unconsol(cid:173)
`idated sand reservoir: Geophysics, 41, 882-894.
`- - 1977, Elastic properties of unconsolidated porous sand
`reservoirs: Geophysics, 42, 1339-1368.
`Hilterman, F. J., 1990, Is AVO the seismic signature oflithology? A
`case history of Ship Shoal Addition: The Leading Edge, 9, 15-22.
`Koefoed, 0., 1955, On the effect of Poisson's ratios of rock strata on
`the reflection coefficients of plane waves: Geophys. Prosp., 3,
`381-387.
`Mazzotti, A., 1990, Prestack amplitude analysis methodology and
`application to seismic bright spots in the Po Valley, Italy: Geo(cid:173)
`physics, 55, 157-166.
`Ostrander, W. J.. 1982, Plane-wave reflection coefficients for gas
`sands at non-normal angles of incidence: 52nd Ann. Internat.
`Mtg., Soc. Expl. Geophys., Expanded Abstracts, 216-218.
`- - 1984, Plane-wave reflection coefficients for gas sands at
`non-normal angles of incidence: Geophysics, 49, 1637-1648.
`Rutherford, S. R., and Williams, R. H., 1989, Amplitude-versus(cid:173)
`offset variations in gas sands: Geophysics, 54, 680-688.
`Shuey, R. Too 1985, A simplification of the Zoeppritz equations:
`Geophysics, 50, 609-614.
`
`Downloaded 03/06/15 to 108.45.99.150. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`Ex. PGS 2016