Communications – electrical: acoustic wave systems and devices – Seismic prospecting – Land-reflection type
Reexamination Certificate
2003-03-12
2004-10-05
Moskowitz, Nelson (Department: 3663)
Communications, electrical: acoustic wave systems and devices
Seismic prospecting
Land-reflection type
C367S021000, C367S028000, C367S045000, C367S073000, C702S013000, C702S017000
Reexamination Certificate
active
06801473
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to the field of seismic data processing. Specifically, the invention is a method for the estimation and removal of artifact noise from seismic data.
2. Description of the Related Art
Geophysical prospecting is the activity of searching for deposits of valuable minerals or hydrocarbons, such as oil and gas. Seismic surveys are used in geophysical prospecting to determine the location of potential reservoirs of hydrocarbon deposits in subterranean rock formations. Seismic surveys measure the travel time of seismic waves from the starting locations at seismic sources through reflections off the interfaces between layers in rock formations to the ending locations at seismic receivers. Seismic processing combines the many travel times between source and receiver pairs and attempts to remove all undesired noise in the recorded seismic signal. The desired result is an image of the locations of the rock formations.
At present, seismic data processing techniques result in seismic images that are degraded by noise, or by seismic artifacts. A seismic artifact is any distortion in the seismic data that can impede the ability to accurately estimate reservoir properties of interest from seismic data. An example of a seismic artifact is the interference by a shallower object in the seismic survey of a deeper target. This type of interference, in general, includes the effect of the overburden.
FIG. 1
shows an elevation view illustrating a model of a seismic signal with this type of artifact. The interfering shallow object
101
could be any type of near surface geologic feature, such as shallow gas (shown), gas filled channel complexes, hydrate beds, or salt domes. The interfering shallow object
101
is located between a seismic survey line
102
on the surface and a deeper subsurface target
103
, such as a potential hydrocarbon reservoir. The seismic survey line
102
could be located on a surface of the water in a marine seismic survey, or on a surface of the earth in a land seismic survey. The seismic survey line
102
contains the seismic sources
105
and the seismic receivers
106
.
These artifacts may arise in several forms. Artifacts can be caused by acquisition parameters, processing parameters, and geologic factors. An acquisition artifact may arise from survey design and execution irregularities and is often spatially periodic or semi-periodic. A geologic artifact may arise from formations above a target horizon having geologic variation or an irregular shape and is often spatially non-periodic. A processing artifact may arise from imperfect processing algorithms and parameter choices, and may be either periodic or non-periodic.
Artifacts with the same spatial periodicity as the acquisition geometry are termed acquisition-related. The spatial periodicity of acquisition-related artifacts can be modeled using conventional modeling techniques. Once the periodicity is identified, that is, the wavenumbers have been determined, several conventional filtering techniques are available to remove (or, more generally, to manipulate) the energy with the same periodicity as the artifacts. However, determining the exact magnitude of the artifacts is beyond current modeling technology. Therefore, removal of the artifacts is often not possible with conventional techniques.
Other types of artifacts are spatially irregular. For example, surface obstacles encountered during data acquisition or geological features of the overburden positioned above a target interval may cause irregularities or shadowing effects in the data. In general, contaminated seismic data can lead to a distorted interpretation of formation properties, which can in turn lead to missed prospects, dry holes, or uneconomic development wells. Thus, it is desirable to be able to remove the effects of these types of artifacts from seismic data. No conventional technique is available for filtering random or pseudo-random artifacts.
The demand for precision in the measurement of formation properties from seismic data is intensifying. Seismic amplitude analysis has long been a key component of successful exploration and exploitation strategies. Increasingly subtle variations in the amplitudes of the seismic data are now being analyzed to achieve detailed areal delineation, estimate pay thickness and net/gross ratio, determine details of the depositional environment, and predict reservoir fluid content, lithology, and migration pathways. However, even within a given seismic data set, acquisition geometry and processing steps can produce significant artifacts and cause variability in seismic amplitudes that are unrelated to the formation properties of interest. The evaluation, quantification, and removal of artifacts are critical to realizing the full potential of quantitative seismic attribute analysis. By developing and applying techniques to remove the artifacts (or perturbations caused by conditions in the overburden) from the seismic data, the accuracy of predictions based on seismic attributes can be improved, and the associated risk in subsequent exploration activity can be lowered.
Currently, artifacts are filtered from seismic amplitude data using two broad approaches: empirical and deterministic. Empirical approaches examine the data in order to identify and manipulate the portion of the signal that appears non-geologic. Such approaches can significantly improve the interpretability of amplitude maps that are contaminated by artifacts such as water bottom multiples. Unfortunately, empirical approaches are susceptible to making merely cosmetic changes to amplitude maps without sufficiently improving the accuracy of the measurements of formation properties. Deterministic artifact mitigation approaches systematically derive corrections based on the causes of the artifacts. These approaches can lead to accurate corrections if all of the true causes of the artifacts are known and correctable. This, however, is rarely the case.
Three empirical methods are commonly used to remove periodic amplitude artifacts from post stack seismic data. These three methods are the moving window average, the discrete wavelet transform low-pass filter, and the fast Fourier transform filter. The latter two filter methods attempt to remove or correct all of the energy at affected periodicities (wavenumbers), but sometimes only deal with a portion of this energy. A comparison of the resolution of these three empirical methods is presented in the following paragraphs after a model of a seismic signal with a periodic amplitude artifact is defined.
The simplest form of a seismic signal with a periodic amplitude artifact is an additive artifact. In one dimension, a signal along a seismic line with an additive artifact can be defined in its most general form by the equation
f
(
x
)=
g
(
x
)+
p
(
x
), (1)
where
x is the Common Depth Point (CDP) location,
f(x) is the recorded seismic signal,
g(x) is the true geologic signal, and
p(x) is the noise signal.
FIG. 1
shows an example in which a seismic signal may have an amplitude artifact such as may be characterized by Equation (1). The seismic signal f(x), recorded at receiver
106
, consists of the seismic representation of the geologic environment at the target
103
plus the imprint from local artifact
101
.
The target or true geologic signal g(x) is the seismic signal that would be recorded if the artifacts created by the acquisition, processing, or geologic imprint were eliminated. It is desired to either solve for or estimate this geologic signal g(x). The perturbation noise signal p(x) represents the artifact noise created by the acquisition, processing, or geologic imprint, and may be either random, semi-periodic, or periodic. The amplitude of perturbation p(x) is unknown, although the overall character of its spatial variation or periodicity is known or may be estimated empirically. Thus p(x) can be written as a product of an unknown amplitude modulation factor and the known or estimated spatia
Matteucci Gianni
Wang Yuan
Zimmerman Linda J.
ExxonMobil Upstream Research Company
Moskowitz Nelson
Plummer J. Paul
LandOfFree
Method for estimating and removing artifact noise from... 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 estimating and removing artifact noise from..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for estimating and removing artifact noise from... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3328742