Multi-attribute seismic waveform classification

Details Multi-attribute seismic waveform classification Multi-attribute seismic waveform classification

1999-10-20

2001-04-24

McElheny, Jr., Donald E. (Department: 2862)

Data processing: measuring, calibrating, or testing

Measurement system in a specific environment

Earth science

Reexamination Certificate

active

06223126

ABSTRACT:

BACKGROUND OF THE INVENTION
For many years seismic exploration for oil and gas has involved the use of a source of seismic energy and its reception by an array of seismic detectors, generally referred to as geophones. When used on land, the source of seismic energy can be a high explosive charge electrically detonated in a borehole located at a selected point on a terrain, or another energy source having capacity for delivering a series of impacts or mechanical vibrations to the earths surface. Offshore, air gun sources and hydrophone receivers are commonly used. The acoustic waves generated in the earth by these sources are transmitted back from strata boundaries and/or other discontinuities and reach the earth's surface at varying intervals of time, depending on the distance traversed and the characteristics of the subsurface traversed. On land these returning waves are detected by the geophones, which function to transduce such acoustic waves into representative electrical analog signals, which are generally referred to as traces. In use on land, an array of geophones is laid out along a grid covering an area of interest to form a group of spaced apart observation stations within a desired locality to enable construction of three dimensional (3D) views of reflector positions over wide areas. The source, which is offset a desired distance from the geophones, injects acoustic signals into the earth, and the detected signals at each geophone in the array are recorded for later processing using digital computers, where the analog data is generally quantized as digital sample points, e.g., one sample every two milliseconds, such that each sample point may be operated on individually. Accordingly, continuously recorded seismic field traces are reduced to vertical cross sections, or volume representations, or horizontal map views which approximate subsurface structure. The geophone array is then moved along to a new position and the process is repeated to provide a seismic survey. A 3D seismic survey is data gathered at the surface and presented as a volume representation of a portion of the subsurface.
After exploration of an area is completed, data relating to energy detected at a plurality of geophones will have been recorded, where the geophones are located at varying distances from the shotpoint. The data is then reorganized to collect traces from data transmitted at various shotpoints and recorded at various geophone locations, where the traces are grouped such that the reflections can be assumed to have been reflected from a particular point within the earth, i.e., a common midpoint. The individual records or “traces” are then corrected for the differing distance the seismic energy travels through the earth from the corresponding shotpoints, to the common midpoint, and upwardly to the various geophones. This step includes correction for the varying velocities through rock layers of different types and changes in the source and receiver depths. The correction for the varying spacing of shotpoint/geophone pairs is referred to as “normal move out.” After this is done the group of signals from the various midpoints are summed. Because the seismic signals are of a sinusoidal nature, the summation process serves to reduce noise in the seismic record, and thus increasing its signal-to-noise ratio. This process is referred to as the “stacking ” of common midpoint data, and is well known to those skilled in the art. Accordingly, seismic field data undergoes the above-mentioned corrections, and may also undergo migration, which is an operation on uninterpreted data and involves rearranging of seismic information so that dipping horizons are plotted in their true location. Other more exotic known processing techniques may also be applied, which for example enhance display of faults, stratigraphic features, amplitude versus offset (AVO) or some attribute such as peak amplitude, instantaneous frequency or phase, polarity etc., before the continuously recorded traces are reduced to vertical or horizontal cross sections or horizontal map views.
In the course of seismic exploration, control points may be established by boreholes that penetrate a strata of interest. Quite often, however, the boreholes are widely separated, and only at such sparse control points can the seismic observations be calibrated by comparison of the selected seismic attributes with the texture and composition of the target strata. The seismic survey, having relatively closely spaced observation points that are distributed between the sparse control points, has the potential for providing data to estimate reservoir conditions as they extend away from the wellbore.
Reconnaissance seismic analysis often uses trends of common seismic waveforms to identify prospective geological features. Many of the current processes use simple attributes such as peak amplitude, mean frequency, etc. to show a real trends, with peak amplitude most commonly used since it can represent a “bright-spot ” which may be a direct indication of hydrocarbon. More recent methods have used an actual waveform to show a real trends away from the wellbore. There remains, however, a need to include waveform shape analysis using widespread trace-to-trace similarities which compare multi-attributes of the seismic waveforms, for estimating a variation in rock type or texture, i.e. a change in lithology along a given stratum and/or hydrocarbon effects such as oil or gas in the pore space.
Accordingly, it is an object of this invention to allow an interpreter to survey variation of multi-attributes along an interpreted horizon.
Another more specific object of this invention is to analyze waveform shapes of traces in multiple seismic stacks (e.g. stacks of near and far offset) and to classify waveforms, and map any seismic waveform trends that are discovered.
Yet another object of this invention is to allow selection by an interpreter of the number of waveform classes that are assigned for classifying traces in a seismic data volume.
Another more specific object is to utilize a robust method for comparison of trace-to-trace similarities that considers both absolute amplitude and shape of the waveform being compared.
Yet another object of this invention is to extract additional lithological information, which is available from the seismic data volume.
SUMMARY OF THE INVENTION
According to the present invention the foregoing and other objects and advantages are attained in a method and apparatus for using a subset of seismic traces selected from a trace volume for identifying trends of multiple seismic attributes along a subsurface horizon. The method involves comparing the full waveform of each trace in the subset to every other trace in the subset over a predefined zone (or time interval) extending both above and below the subsurface horizon of interest in order to determine a similarity factor for each trace in the subset. The traces in the subset are separated into a number of waveform classes based on the range of variance of the similarity factors for the individual traces in the subset. Finally, the traces are grouped according to the waveform classes so as to show a real trends of multiple seismic attributes in the subsurface.
In a preferred embodiment the method first selects a gathered, stacked and migrated seismic trace volume, in which the interpreted horizon is identified, and further the method selects one or more alternately processed volumes (e.g. near stack, far stack, AVO intercept or gradient, coherency, p-wave section, s-wave section etc.) having the identified same horizon. A corresponding subset of representative traces is then selected from each volume to be processed according to this invention. A robust algorithm, which considers both shape and absolute amplitude of two waveforms being compared, is applied for comparing pairs of traces in the subset where each trace in the subset is compared to every other trace in the subset. Traces in corresponding subsets of all of the volumes to be processed are likewise compared. A similarity

Affiliated with

Also associated with

Rating

Multi-attribute seismic waveform classification does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Multi-attribute seismic waveform classification, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multi-attribute seismic waveform classification will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2458239