Communications – electrical: acoustic wave systems and devices – Seismic prospecting – Land-reflection type
Reexamination Certificate
2003-05-15
2004-06-29
Moskowitz, Nelson (Department: 3663)
Communications, electrical: acoustic wave systems and devices
Seismic prospecting
Land-reflection type
C367S052000, C367S059000, C367S049000, C702S014000, C702S017000
Reexamination Certificate
active
06757216
ABSTRACT:
FIELD OF THE INVENTION
This invention relates generally to the field of seismic prospecting and, more particularly, to seismic data processing and interpretation. Specifically, this invention improves the ability of geoscientists to perform quantitative AVO (Amplitude Versus Offset) analysis on seismic data for exploration, development and production scenarios by allowing the interpreter to correct post-processed seismic amplitudes that are corrupted due to the stacking of non-flat gathers. The corrected amplitudes allow accurate construction and manipulation of AVO attributes.
BACKGROUND OF THE INVENTION
In many settings, hydrocarbons can be directly detected from seismic data by a trained observer. Where hydrocarbons are present, the density and velocity of a typical reservoir is reduced, which in most cases increases the magnitude of the reflection coefficient and hence the resulting seismic reflection amplitudes. The effect can often be seen as a bright spot on routinely processed seismic data. Such bright spots are one type of direct hydrocarbon indicator, or “DHI”. DHI is the name given to nearly any characteristic of seismic data that is sensitive to, and often diagnostic of, fluid type. Conventional DHIs are those that can be found in conventionally processed seismic data, i.e., full or near-offset stacks.
DHIs have several limitations. First, high porosity in the reservoir is necessary if reflection coefficients are to be increased to the detectable threshold. Sandstones at depths greater than ten thousand feet and carbonate reservoirs at any depth seldom have such porosity. Second, the difference in velocity and density between water and oil may be slight, and free gas may be required to produce DHIs in many geologic settings. Unfortunately, a small amount of gas has the same amplitude effect as a large amount, and many wells drilled on DHIs produce only fizz water. Most importantly, the presence of hydrocarbons is only one of a great many things that can cause amplitude and phase changes in seismic reflections.
One method of overcoming the limitations of conventional DHIs is to analyze prestack (or multi-offset/multi-angle) seismic data. When a compressional wave impinges obliquely on an interface, part of the energy is converted into shear waves. Since the amount of energy that is converted to shear waves depends on the angle of incidence (or offset, the horizontal distance between seismic source and receiver), compressional reflection amplitude also depends on angle or offset. This is the basis for AVO analysis.
AVO depends strongly on pore-fluid type and amount, as well as on the type of rocks above and below the reflecting interface. From a geophysical perspective, AVO is determined by contrasts in impedance and Poisson's ratio or ratio of P-wave to S-wave velocity. Class
1
reservoirs have a high impedance contrast, class
2
reservoirs have a near-zero impedance contrast, and class
3
reservoirs have a low impedance contrast. By exploiting more than one physical property of the reservoir, one can extract reservoir information that is not contained in conventional DHIs extracted from fully stacked data.
A difficulty with AVO analysis is the need to analyze more information than simply the stacked data. The process is often simplified by fitting AVO signatures or curves with a line and then working with only that line's slope and intercept. Since two points describe a line, AVO analysis may be performed on two partial stacks of the seismic data, one at far offsets and one at near offsets. Each partial stack combines several, adjacent seismic traces with a common mid-point but different, consecutive offsets. The far and near stacks give two points from which a linear amplitude relationship with offset may be estimated.
Accurate seismic information is critical for quantitative AVO analyses. However, even with a state-of-the-art AVO processing stream, non-flat gathers, due to inadequate normal moveout correction, may have been stacked to form partial stack seismic data (i.e. near and far stack volumes). (When reflections with different offsets but from the same reflector point are combined [stacked], a correction called the normal moveout correction must be made to compensate for the different travel times caused by the different path lengths.) This sub-optimal stacking manifests itself in many ways. The offset stack volumes are not aligned in time, and seismic amplitudes are reduced due to the destructive interference of traces within the partial stack.
This effect is illustrated in
FIGS. 1A-1D
using synthetically generated seismic traces.
FIGS. 1A
,
1
B and
1
C each depict a gather of nine traces (traces No's.
39
through
47
) having a common reflection point, each figure with a different amount of residual normal moveout (“RNMO”) and representing traces that might be stacked to form a far (partial) stack. The nine traces are plotted sequentially vs. depth as measured by the travel time for seismic wave propagation in the medium. Normal moveout (“NMO”) correction has already been applied to these data, as a seismic processing step. The NMO correction is key to the efficiency of common mid-point stacking. The recorded travel time for each trace depends on its offset. The greater the offset, the farther the energy must travel from source to receiver, and hence the greater the travel time. The NMO correction is a processing step intended to adjust the two-way travel time of reflections in each trace to be the same as that of its corresponding zero-offset trace. The NMO correction is calculated using the travel times for a given offset and a known or assumed seismic wave velocity. In practice, the effectiveness of the NMO correction varies, and residual normal moveout (RNMO) is a measure of the amount of under- or over-correction.
In
FIG. 1A
, the NMO correction is perfect. The peaks and troughs from each trace are aligned at the same depth, or vertical travel time, as indeed they should after normal moveout correction because they represent reflections from the same subterranean point. The perfectly flat reflections of
FIG. 1A
have 0 milliseconds (ms) RNMO.
FIGS. 1B and 1C
show cases where the seismic processing has not fully flattened the data. In
FIG. 1B
, trace No.
47
images the reflections approximately 0.01 seconds, or 10 ms, deeper than trace No.
39
. In
FIG. 1C
, the difference is approximately 16 ms. Thus, RNMO=10 Ms for
FIG. 1B
, and RNMO=16 ms for FIG.
1
C.
FIG. 1D
shows an expanded view of the normalized stacked amplitude for each of the three RNMO cases, covering the travel time range 0 ms to approximately 47 ms. Curve
1
represents the normalised sum of the traces in
FIG. 1A
, curve
2
corresponds similarly to
FIG. 1B
, and curve
3
corresponds to
FIG. 1C
, with the amplitude being progressively smaller because of increasing destructive interference (cancellation) between traces at increasing RNMO. Only curve
1
has preserved the correct amplitude. The other two curves have smaller amplitudes than they should, due to inadequate flattening (NMO correction) of the seismic data. In addition to reduced amplitudes, curves
2
and
3
also suffer from phase delay compared to curve
1
.
Because travel times are greater at large offsets, the corresponding NMO corrections are also greater. Greater NMO corrections tend to produce greater RNMO, and consequently the resulting amplitude distortion tends to be greater for the far stack than for the near stack.
The seismic interpreter typically receives a far and a near stack volume of data from the seismic processor. The interpreter also gets “mute window” information, which is the description of which traces are included in the two partial stacks and at what travel times, e.g., trace No's.
39
-
47
in the far stack example of
FIGS. 1A-1D
. Due to data handling and storage limitations, the interpreter receives little else from the volumes of seismic data that are originally acquired. Currently, there is no established method for assessing how
Eastwood John E.
Hussenoeder Stefan
Lazaratos Spyridon K.
Varnai Peter
West Brian P.
ExxonMobil Upstream Research Company
Moskowitz Nelson
Plummer J. Paul
LandOfFree
Method for post processing compensation of amplitude for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for post processing compensation of amplitude for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for post processing compensation of amplitude for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3365850