Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Earth science
Reexamination Certificate
2002-04-22
2004-02-17
Hoff, Marc S. (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Earth science
C702S006000
Reexamination Certificate
active
06694263
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method for facilitating monitoring, in the course of time, of the evolution of seismic events in a zone of interest in an underground formation (a reservoir zone for example), by compared analysis of a certain number n of seismic record sets obtained respectively after successive 3D seismic surveys (technique referred to as 4D seismic method).
2. Description of the Prior Art
Seismic measurements are conventionally used to provide additional information, in relation to drilling data, on the variations of the subsoil formations: lithologic, petrophysical or fluid saturation variations. In particular, within the scope of hydrocarbon reservoir production, it has become quite frequent to record seismic measurements repeatedly and then to interpret the seismic measurement variations in connection with the saturation and pressure variations due to reservoir production phenomena.
A conventional method of using these records analyzes directly the amplitude (or any other seismic attribute) differences between the various surveys. Interpretation is then often backed up by modeling the acoustic behavior of the subsoil according to the estimated changes in the physical properties thereof. An example of this approach is described in:
Johnston, D., 2000, “Using Legacy Seismic Data in an Integrated Time-Lapse Study: Lena Field, Gulf of Mexico”, The Leading Edge, 19, No.3.
It may sometimes be difficult to interpret the data based on amplitude (or any other trace attribute) differences. In fact, variations in the physical properties of the rock in the course of time, related for example to the reservoir production, to the use of enhanced recovery methods, etc., lead to variations in the amplitude of the seismic traces in the reservoir, and also to time lags (lengthening or shortening of the trace). The difference between two seismic traces may therefore be difficult to interpret since it results from amplitude changes as well as from time lags which eclipse these amplitude variations as can be seen in FIG.
1
.
Another approach consists in using statistical pattern recognition techniques allowing classification of the seismic events into various categories according to the different physical states of the reservoir. These approaches are for example described in:
Dumay, J., Fournier, F., 1988, “Multivariate Statistical Analyses Applied to Seismic Facies Recognition”, Geophysics, 53, No.9, pp.1151-1159.
They can be applied, in the case of the interpretation of repeated seismic surveys, to the seismic amplitudes of the various seismic surveys, to any attribute deriving from the seismic trace or to the amplitude differences between surveys. These pattern recognition techniques can be used with or without learning, as already described and implemented in U.S. Pat. No. 6,052,651 and patent application EN-0,011,618, both filed in the name of the assignee.
Within the scope of repeated seismic surveys, an example of seismic events classification with learning, where learning has been carried out using the seismic data of a first survey and the classification applied independently to this first survey, then to a repeated survey, can be found in:
Sonneland, L., Veire, H. F., Raymond, B., Signer, C., Pedersen, L., Ryan, S., Sayers, C., 1997, “Seismic Reservoir Monitoring on Gullfalks”, The Leading Edge, 16, No.9, pp.1247-1252.
SUMMARY OF THE INVENTION
In order to be free from the artifacts related to the calculation of the amplitude (or any other seismic attribute) differences and to analyze the evolution of seismic events in the course of time in its entirety, the invention is a method allowing classification of these events according to their overall pattern, while simultaneously analyzing as a whole the seismic measurements obtained from the various surveys.
The invention provides studying and interpretation of an evolution of the seismic records which are related to the evolution of the physical properties of the zone as a result of production mechanisms.
Seismic events are understood to be seismic trace portions taken in the zone of interest from the successive record sets or traces. The seismic events to be classified are characterized by seismic parameters or attributes. These attributes can be of different types. They can consist, for example, of the succession of the amplitudes of the seismic trace portions (contained in the seismic window studied).
The method of the invention detects the physical changes undergone in the course of time by a subsoil zone, by analysis of the changes observable in seismic events characterized each by seismic attributes, recorded within a time window, on the seismic traces of several data sets obtained respectively during successive seismic surveys (repeated or 4D seismic surveys), comprising using a pattern recognition technique to classify the seismic events. The method comprises:
forming an analysis set comprising all the seismic events recorded on the traces of the various seismic trace sets, with identification of each seismic event by means of its spatial position in the zone and by the number of the trace set to which the seismic event belongs,
forming a learning base comprising learning classes each comprising a certain number of seismic events that can be associated with common physical properties,
constructing a calibrated classification function on the defined learning classes, and
applying to all of the seismic events the calibrated classification function so as to assign at least part of the seismic events of the set to the various learning classes.
Construction of a calibrated classifying function is for example carried out by means of a discriminant analysis technique, or by a neural network technique.
According to an implementation mode, the learning base is formed for example from seismic events measured in the vicinity of wells drilled through the formation studied, by defining therefrom learning classes corresponding to different rock natures or to different fluid contents.
According to another implementation mode, the learning classes are formed for example by non-supervised classification of the seismic events.
In particular, the modes of a multivariate probability density function calculated from all of the seismic events characterized by the associated attributes can be used.
According to another implementation mode, the learning base is formed by selecting the seismic traces in the most representative parts of the various estimated physical states of the zone, and of their variations, obtained for example with a numerical flow and production simulation model.
To analyze the results, it is possible, for example, to create classification difference charts from repeated surveys, allowing better highlighting of the class changes of a seismic event from one survey to the next.
It is thus possible to detect, in the course of time, changes in the overall pattern of the seismic trace, or a contrario stabilities of the seismic events in certain zones which are either out of reach of the recovery process used, or seismically insensitive to the physical state variations of the reservoir.
REFERENCES:
patent: 5444619 (1995-08-01), Hoskins et al.
patent: 6052651 (2000-04-01), Fournier
patent: 6272444 (2001-08-01), Thevoux-Chabuel et al.
patent: 6374185 (2002-04-01), Taner et al.
Fournier F. et al; “A Statistical Methodology for Deriving Reservoir Properties From Seismic Data” Geophysics of Exploration Geophysicists. Tulsa, US, vol. 60, No. 5, Sep. 1, 1995, pp. 1437-1450 XP002070858; ISSN: 0016-8033.
Dumay J. et al ; “Multivariate Statistical Analyses Applied to Seismic Facies Recognition” Geophysics, Sep. 1988, USA, vol. 53, No. 9, pp. 1151-1159, XP002187537; ISSN: 0016-8033.
Sonneland Lars et al., “Seismic Reservoir Monitoring on Gullfaks” Leading Edge; Leading Edge (Tulsa, OK) Sep. 1997 SOC of Exploration Geophysicists, Tulsa, OK, USA, vol. 16, No. 9, Sep. 1997, pp. 1247-1252, XP002187538.
Burns, C. Scott: “4D Seismic Interpretive Processing: Optimizing Differentia
Fournier Frederique
Lucet Nathalie
Antonelli Terry Stout & Kraus LLP
Gutierrez Anthony
Hoff Marc S.
Institut Francais du Pe'trole
LandOfFree
Method for facilitating monitoring, in the course of time,... 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 facilitating monitoring, in the course of time,..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for facilitating monitoring, in the course of time,... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3291490